Stanley P. Leong · S.

David Nathanson
Jonathan S. Zager Editors

Cancer Metastasis
Through the
Lymphovascular
System
Cancer Metastasis Through the Lymphovascular
System
Cover illustration by Ian Wood, MD, resident in the general surgery program at the Henry Ford Health System
(HFHS). The artwork for the book cover is supported by the Nathanson/Rands Chair in Breast Cancer Research
of the Henry Ford Cancer Institute (HFCI).
Stanley P. Leong • S. David Nathanson
Jonathan S. Zager
Editors

Cancer Metastasis Through

the Lymphovascular System
Editors
Stanley P. Leong S. David Nathanson
Department of Surgery and Melanoma Center Henry Ford Cancer Institute
California Pacific Medical Center and Research Detroit, MI, USA
Institute
San Francisco, CA, USA

Jonathan S. Zager
Moffitt Cancer Center
Tampa, FL, USA

ISBN 978-3-030-93083-7    ISBN 978-3-030-93084-4 (eBook)

https://doi.org/10.1007/978-3-030-93084-4

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and
retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter
developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed
to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been
made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
This book is dedicated to our patients, who are the ones who bravely endure the
fight with cancer, and to our colleagues and team members, who help treat
cancer every day forging the relationships and fighting the good fight, although
sometimes losing the battle. However, when the battle is won, the feeling of
accomplishment, of helping someone in need, and victory over this vicious
disease could be overwhelming. We want to thank all of our patients, all of our
families for their continuous support and encouragement as we help in the fight
against cancer, all of the researchers for laying down the groundwork to
advance the field, and all of the clinical team members for helping to provide
the treatments. Working all together, there is no limit to what can be achieved.
Stanley P. Leong, MD
S. David Nathanson, MD
Jonathan S. Zager, MD

We also dedicate this book to our colleague Dale Han, MD, who served
tirelessly as a section editor for this book. Dr. Han graduated with a medical
degree from the State University of New York at Stony Brook School of
Medicine in 2003. Dr. Han did his general surgery residency at the University
of Pennsylvania School of Medicine and his fellowship at Moffitt Cancer
Center in Tampa. Dr. Han dedicated his life to treating melanoma and other
skin and soft-tissue malignancies. He worked with trainees and other faculty
members to publish outcome-based studies, some of which changed the way
we look at and treat these malignancies. Dr. Han was devoted to teaching,
working with trainees, mentoring them clinically, enabling them to do
research, and enriching their lives. He was the recipient of numerous awards.
Most importantly to the authors in this book, and to ourselves, he was a
beloved and loyal friend. When Dr. Han passed away, we all lost something
in the cancer world—a friend, a colleague, an exemplary researcher, and a
clinician who excelled on all fronts. We will truly miss him. This book is
dedicated to his memory and to his relentless commitment to excellence in
clinical care, research, and education.
Giorgos Karakousis, MD
Jonathan S. Zager, MD
Preface

Advances in molecular biology over the past several decades have firmly established the obser-
vations that cancer is a heterogeneous disease arising from genomic alterations or mutations.
Thus, within a cancer cell population, cellular subclones may develop resulting in heterogene-
ity within a tumor and between different metastatic deposits. The cancer microenvironment,
including immune responses, exerts a selective force akin to Darwinian “natural selection,”
which promotes the development of clones that invade and metastasize from the primary site
to the distant sites, according to the principles of Paget’s seed and soil hypothesis for cancer
spread. One critical question is whether the initial pathway for cancer metastasis is through
tumor-associated lymphatic vessels to the sentinel lymph nodes or through tumor-associated
blood vessels, or via both pathways. This important background drove the style and format of
this textbook on cancer metastasis.
Molecular and genetic mechanisms of cancer proliferation and metastasis are important in
the further definition and development of more targeted and successful treatment modalities.
Imaging techniques for the detection of early-stage cancer and cancer metastases and applica-
tion of liquid biopsies in sarcoma, prostate, gastrointestinal, and lung cancers are highlighted.
Several specialized topics, such as the latest findings on cancer biomarkers and personalized
cancer therapy, role of stem cells in cancer metastasis, HPV-induced cancer, impact of micro-
biome on cancer metastasis, and application of artificial intelligence in cancer metastasis
research, are discussed in separate chapters.
While many cancer textbooks concentrate on the systemic management of measurable and
microscopic metastases, this textbook focuses on molecular, genetic, diagnostic, and surgical
aspects of cancer metastasis. Several cancer types are selected to illustrate the common pat-
terns of progression and spread of metastases. The textbook is richly infused with discussions
on how cancer should be evaluated in the multidisciplinary setting and ultimately treated with
multimodality therapies including surgery, radiotherapy, and systemic therapies emphasized,
such as immunotherapy and targeted therapy.

San Francisco, CA, USA Stanley P. Leong

Detroit, MI, USA  S. David Nathanson
Tampa, FL, USA  Jonathan S. Zager

vii
Contents

Part I Genetic and Molecular Pathways of Cancer Growth and Metastasis

1 Hallmarks of Cancer: Molecular Underpinnings ���������������������������������������������������   3

Dhananjay A. Chitale
2 Unifying Concept of Genomic Changes: The Mutational
Landscape of Cancers������������������������������������������������������������������������������������������������� 15
Laura Favazza DO
3 Tumor Microenvironment: Coconspirator in Tumorigenesis��������������������������������� 21
Zhiqiang Wang
4 Hallmarks of Metastasis: Molecular Underpinnings����������������������������������������������� 29
Juan C. Gomez-Gelvez and Dhananjay A. Chitale
5 Hereditary Cancer Syndromes and Cancer Metastasis ����������������������������������������� 37
Brandon M. Shaw and Olena Kis

Part II Pathology of Cancer Growth and Metastasis

6 Pathologic Assessment of Lymph Node Metastasis������������������������������������������������� 55

James Isom and Jane L. Messina
7 Pathologic Assessment of Systemic or Distant Metastasis��������������������������������������� 63
Igor Katsyv and Andrew Turk
8 The Role of Angiotropic Extravascular Migratory Metastasis in Metastases������� 73
Raymond Barnhill and Claire Lugassy

Part III Dormant Cancer Cells and Their Reactivation in the Host Cancer
Microenvironment

9 Tumor Dormancy and Relapse Regulated by the Extracellular Matrix��������������� 89

Hyuna Kim, Rebecca E. Huber, Rita Das Mahapatra, Ning-Hsuan Tseng,
and Shelly R. Peyton
10 Therapy-Induced Dormancy and Residual Disease������������������������������������������������ 97
Ashley V. DiMarco, Nina Marie G. Garcia, and James V. Alvarez

Part IV The Microenvironment of Site-Specific Metastasis

11 The Microenvironment of Site-Specific Metastasis������������������������������������������������� 107

Isaac P. Witz and Sivan Izraely

ix
x Contents

Part V Circulating Cancer Cells and Liquid Biopsy

12 Circulating Tumor Cells and ctDNA in Sarcomas��������������������������������������������������� 121

Camille Jubelin, Denis Cochonneau, Emilie Moranton, Javier Muñoz-Garcia,
and Dominique Heymann
13 Liquid Biopsy Using Cell-Free Tumor DNA for Gastrointestinal Cancers����������� 129
Takafumi Nakano, Tadashi Abe, Seiichiro Takao, Hideyuki Saito,
Takaaki Masuda, and Koshi Mimori
14 Circulating Tumor Cells and ctDNA in Prostate Cancer ��������������������������������������� 139
Nikolas H. Stoecklein and Rui P. L. Neves
15 Circulating Tumor Cells in Lung Cancer����������������������������������������������������������������� 147
Paul Hofman
16 CTCs/ctDNA and Brain Metastasis�������������������������������������������������������������������������� 157
S. Ray Kenney and Dario Marchetti
17 Epigenetics and Liquid Biopsy in Oncology: Role in Metastasis
and Clinical Utility����������������������������������������������������������������������������������������������������� 167
Aitor Rodriguez-Casanova, Aida Bao-Caamano, Nicolás Costa-Fraga,
Laura Muinelo-Romay, and Angel Diaz-Lagares
18 Metabolic Reprogramming of Circulating Tumor Cells for Metastasis ��������������� 175
Ziyuan Zhang and Qihui Shi

Part VI Lymphatic System and Cancer Metastasis

19 Lymphatic System Biology, Pathobiology, and Relation

to Cancer Metastasis��������������������������������������������������������������������������������������������������� 187
Marlys H. Witte and Sarah K. Daley
20 Lymphatic Specification and Development, EMT-MET, and Cancer Spread ����� 199
Xin Geng and R. Sathish Srinivasan
21 Lymphangiogenesis: Lymphatic System and Lymph Nodes; Cancer
Lymphangiogenesis and Metastasis�������������������������������������������������������������������������� 209
Stanley P. Leong and Marlys H. Witte
22 Immune Cell Trafficking in the Lymphatics, Hyaluronan Biology
and Tumour Metastasis ��������������������������������������������������������������������������������������������� 231
David G. Jackson
23 Lymphovascular Genomics and Proteomics, Clinical Syndromes,
and Cancer Metastasis����������������������������������������������������������������������������������������������� 241
Robert P. Erickson and Michael T. Dellinger
24 Imaging of the Lymphatic System with Relevance to Cancer and Cancer
Metastasis��������������������������������������������������������������������������������������������������������������������� 249
Russell S. Witte and Michael Bernas
25 Lymphedema: General Pathophysiology, Prevention, and
Management in Invasive Cancer������������������������������������������������������������������������������� 261
Angelika Chachaj, Neil Piller, Francesco Boccardo, and Andrzej Szuba
Contents xi

Part VII Radiological Imaging of Early Cancer and Cancer Metastases

26 Primary Tumor Staging and Detection of Common Sites of

Distant Metastatic Disease����������������������������������������������������������������������������������������� 275
Kerry L. Thomas, A. Ahmed, and B. Morse
27 Imaging Modalities in the Detection and Diagnosis of Metastatic Disease����������� 283
S. Ali, S. Fenerty, and P. Jonnalagadda
28 Imaging of Bone Metastases��������������������������������������������������������������������������������������� 295
Colleen M. Costelloe, Raul Fernando Valenzuela, Hubert H. Chuang,
and John E. Madewell
29 Imaging of Lung and Soft Tissue Metastases����������������������������������������������������������� 309
William A. Biché, James F. McLoughlin, Vanesa Carlota Andreu-Arasa,
Stephan W. Anderson, and Christina A. LeBedis
30 Image-Guided Biopsy and Intervention for Metastatic Disease����������������������������� 323
Joao R. T. Vicentini, Sina Habibollahi, Ambrose J. Huang,
and Connie Y. Chang

Part VIII Sentinel Lymph Node Surgery for Solid Cancer

31 Sentinel Lymph Node Biopsy for Primary Cutaneous Malignancy����������������������� 339

Marc Moncrieff and Howard Peach
32 Sentinel Lymph Node Biopsy in Breast Cancer������������������������������������������������������� 353
Emily Siegel, John Kiluk, Armando Giuliano, and Brian Czerniecki
33 Sentinel Node Navigation Surgery for Upper Gastrointestinal Cancer����������������� 361
Shuhei Mayanagi and Yuko Kitagawa
34 Sentinel Lymph Node Mapping in Non-­small Cell Lung, Colon,
and Thyroid Carcinomas������������������������������������������������������������������������������������������� 369
Joshua K. Kays and Mark B. Faries

Part IX Surgical Treatment of Primary and Metastatic Cancer

35 The Efficacy and Evolution of Surgical Management Based

on Cancer Biology������������������������������������������������������������������������������������������������������� 377
Richard J. Straker III, Hayley Standage, Giorgos C. Karakousis, and Dale Han
36 The Role of Surgery in Managing Primary and Metastatic Melanoma ��������������� 385
Kristen E. Rhodin, Kirsten Baecher, Winta T. Mehtsun, Mike Lowe,
Genevieve Boland, and Georgia M. Beasley
37 The Role of Surgery in Managing Primary and Metastatic Breast Cancer ��������� 395
Alicia M. Terando, Azadeh Carr, Tina J. Hieken, Mara A. Piltin,
Bindupriya Chandrasekaran, and Carla S. Fisher
38 The Role of Surgery in Managing Primary and
Metastatic Colorectal Cancer ����������������������������������������������������������������������������������� 407
Richard J. Straker III, Hunter D. D. Witmer, Benjamin Deschner,
David Shibata, Kiran K. Turaga, and Najjia N. Mahmoud
39 The Role of Surgery in Management of Gastric Cancer����������������������������������������� 421
Andrew J. Sinnamon, Jose M. Pimiento, and Robert E. Roses
xii Contents

40 The Role of Surgery in Managing Primary and Metastatic

Hepatopancreaticobiliary Cancers��������������������������������������������������������������������������� 431
Lyonell B. Kone, Christopher Javadi, Jessica M. Keilson, Shishir K. Maithel,
George Poultsides, and Ajay V. Maker

Part X Cancer Metastasis to the Lung

41 Metastatic Disease of the Lung ��������������������������������������������������������������������������������� 447

Z. Hammoud, A. Popoff, Chinmayee Potti, and H. Nasser
Part XI Cancer Metastasis to the Liver
42 Tumor Resection and Ablation as a Means of Controlling Hepatic Metastases��� 463
Brian D. Griffith and Timothy L. Frankel
43 Regional Arterial Infusional Therapy as a Means of Controlling Hepatic
Metastases������������������������������������������������������������������������������������������������������������������� 469
Zachary Brown, Stanley Kalata, and Alex C. Kim

Part XII Cancer Metastasis to the Bone

44 Molecular Mechanisms of Metastasis to the Bone��������������������������������������������������� 483

Jeremy S. Frieling and Conor C. Lynch
45 Surgical Management of Metastatic Disease to the Upper Extremity������������������� 493
Allison C. Greene, Michael T. Torchia, Daniel C. Austin, John-Erik Bell,
and Eric R. Henderson
46 Surgical Management of Metastatic Disease to the Lower Extremity������������������� 505
Nathan W. Mesko and Lukas M. Nystrom
47 Surgical Management of Metastatic Disease to the Pelvis ������������������������������������� 515
Timothy J. Evans, Odion Binitie, and David M. Joyce
48 Surgical Management of Metastatic Disease to the Spine��������������������������������������� 523
Christopher P. Wang, Amanda Brisco, and James K. C. Liu
49 Interventional Radiology Techniques for Management of Skeletal Metastases��� 533
Nainesh Parikh and Altan Ahmed

Part XIII Cancer Metastasis to the Brain and the Nervous System

50 Brain Metastases: Overview and Molecular Mechanisms ������������������������������������� 541

S. Haider, J. Snyder, and I. Lee
51 Indications and Techniques for Surgical Intervention in Patients
with Metastatic Brain Tumors����������������������������������������������������������������������������������� 547
Jacob A. Pawloski, Omar Awan, Mateo Ziu, and Adam M. Robin
52 Brain Metastases: Current and Future Pharmacological Treatment ������������������� 559
Erika Santos Horta and Tobias Walbert
53 Radiotherapy for CNS Metastases ��������������������������������������������������������������������������� 567
Alexander N. Slade, Mark Ashamalla, and Samuel Ryu
54 Leptomeningeal Carcinomatosis������������������������������������������������������������������������������� 575
Sunny R. K. Singh, Sindhu J. Malapati, and Ahmad Mattour
Contents xiii

Part XIV Urologic Cancer and First Patterns of Metastasis

55 Urologic Cancer and the First Patterns of Metastasis��������������������������������������������� 587

Samuel L. Washington III, Peter R. Carroll, and Sima P. Porten

Part XV Biology and Clinical Aspects of Sarcoma Progression

56 Soft Tissue Sarcoma��������������������������������������������������������������������������������������������������� 595

Scott Kizy and Ricardo J. Gonzalez
57 Bone Sarcoma������������������������������������������������������������������������������������������������������������� 601
Akash A. Shah, Howard Y. Park, and Francis J. Hornicek
58 Liposarcoma Metastasis��������������������������������������������������������������������������������������������� 611
Luke V. Selby, Raphael Pollock, and Valerie Grignol
59 Desmoid Tumors��������������������������������������������������������������������������������������������������������� 619
Gaya Spolverato and Alessandro Gronchi
60 Atypical Patterns of Metastases: How Do Sarcomas Metastasize? ����������������������� 629
Pia van der Laan, Fabio Tirotta, Valeriya Pankova, Samuel Ford, Paul Huang,
and Winan J. van Houdt
61 Tumor Immune Microenvironment of Soft Tissue Sarcoma����������������������������������� 639
Tom Wei-Wu Chen, Sheng-Fang Su, and William W. Tseng

Part XVI The Role of Radiation in the Treatment of Primary and Metastatic Cancer

62 Basic Principles of Radiobiology and Cancer Metastasis Prevention������������������� 653

William L. Harryman and Anne E. Cress
63 Technical Innovations in the Delivery of Radiation Therapy��������������������������������� 661
Russell J. Hamilton
64 Radiation Therapy in the Definitive Treatment of Cancer������������������������������������� 671
Alexander N. Garcia and Baldassarre Stea
65 Radiation Therapy for Extracranial Oligometastatic Disease������������������������������� 681
Q. A. Ho and C. C. Hsu
66 Spine Radiotherapy ��������������������������������������������������������������������������������������������������� 697
Mark Ashamalla, Alexander N. Slade, Kartik Mani, and Samuel Ryu

Part XVII Immunotherapy of Solid Malignancies

67 Immunotherapy in Melanoma and Merkel Cell Cancer����������������������������������������� 709

Melissa Chow, Elizabeth A. Sangalang, Christine Chow, and Adil I. Daud
68 Landscape of Immunotherapy in Lung Cancer������������������������������������������������������� 719
Nirali Sanghavi, Umme Farwa, Faisal Khurshid, and Hatim Husain
69 Immuno-Oncologic Treatment of Genitourinary Malignancies����������������������������� 729
Przemyslaw Twardowski

Part XVIII Current Topics

70 Personalized Systemic Cancer Therapy������������������������������������������������������������������� 739

Kevin B. Kim and Mohammed Kashani-Sabet
xiv Contents

71 Cancer Stem Cells and Their Role in Metastasis����������������������������������������������������� 749

Ruby Ghadially, Richard W. Kim, and Alexandra Charruyer-Reinwald
72 HPV-Induced Cancers ����������������������������������������������������������������������������������������������� 757
Brittney L. Dickey, Jennifer M. Binning, and Julie Rathwell
73 Microbiome and Cancer Metastasis������������������������������������������������������������������������� 767
Michael G. White, Jennifer A. Wargo, and Jennifer L. McQuade
74 Application of Artificial Intelligence in Research on Cancer and
Its Metastasis��������������������������������������������������������������������������������������������������������������� 775
Benjamin Franc
Glossary������������������������������������������������������������������������������������������������������������������������������� 787
Index������������������������������������������������������������������������������������������������������������������������������������� 799
Editors and Contributors

About the Editors

Stanley P. Leong is the Chief of the Cutaneous Oncology Department at the Melanoma
Center of the California Pacific Medical Center and Senior Investigator of the California
Pacific Medical Center Research Institute in San Francisco, California. He is also Emeritus
Professor of Surgery at the University of California, San Francisco. He has published several
books on cancer metastasis with Springer and over 210 peer-reviewed publications. He is the
current President of the Sentinel Node Oncology Foundation. Dr. Leong has been a co-­
organizer and co-chair of the International Congress on Cancer Metastasis taking place in San
Francisco biannually since 2005.

S. David Nathanson is the Professor of Surgical Oncology, Wayne State University, and the
Chair of Breast Cancer Research at the Henry Ford Cancer Institute, Detroit, Michigan.
Besides his activities as a surgical oncologist, Dr. Nathanson teaches medical students, resi-
dents, fellows, nurses, physician assistants, and laboratory scientists at his institute, and he is
actively involved in translational research into the mechanisms of lymph node metastasis. He
is the author of numerous research articles and book chapters on metastasis and three books.

Jonathan S. Zager is the Chief Academic Officer at Moffitt Cancer Center and Senior
Member and Director of Regional Therapies in the Department of Cutaneous Oncology at
Moffitt. Dr. Zager is also the Chair of the Department of Oncologic Sciences and Professor of
Surgery at the University of South Florida Morsani College of Medicine. Dr. Zager has pub-
lished over 300 peer-reviewed articles, invited reviews, and book chapters. He is the editor of
three textbooks. Dr. Zager has given over 400 podium presentations, invited talks, keynote
lectures, and poster presentations. He is a co-organizer and co-chair of the Biannual International
Congress on Cancer Metastasis.

xv
xvi Editors and Contributors

Contributors

Tadashi Abe Kyushu University Beppu Hospital, Beppu, Japan

A. Ahmed H. Lee Moffitt Cancer Center, Tampa, FL, USA
Altan Ahmed Department of Diagnostic Imaging and Interventional Radiology, H. Lee
Moffitt Cancer Center and Research Institute, Tampa, FL, USA
S. Ali Temple University Hospital, Philadelphia, PA, USA
James V. Alvarez Department of Pharmacology and Cancer Biology, Duke University School
of Medicine, Durham, NC, USA
Stephan W. Anderson Department of Radiology, Boston University Medical Center, Boston
University School of Medicine, Boston, MA, USA
Vanesa Carlota Andreu-Arasa Department of Radiology, Boston University Medical
Center, Boston University School of Medicine, Boston, MA, USA
Mark Ashamalla Department of Radiation Oncology, Stony Brook University Hospital,
Stony Brook, NY, USA
Daniel C. Austin Dartmouth Hitchcock Medical Center, Lebanon, NH, USA
Omar Awan Inova Neurosciences and Spine Institute, University of Virginia Medical School,
Falls Church, VA, USA
Kirsten Baecher Department of Surgery, Emory University, Atlanta, GA, USA
Aida Bao-Caamano Cancer Epigenomics, Translational Medical Oncology Group
(Oncomet), Health Research Institute of Santiago (IDIS), University Clinical Hospital of
Santiago (CHUS/SERGAS), Santiago de Compostela, Spain
Raymond Barnhill Department of Translational Research, Institut Curie, Paris, France
University of Paris Réné Descartes Faculty of Medicine, Paris, France
Georgia M. Beasley Department of Surgery, Duke University, Durham, NC, USA
John-Erik Bell Dartmouth Hitchcock Medical Center, Lebanon, NH, USA
Michael Bernas Department of Medical Education, TCU School of Medicine, Fort Worth,
TX, USA
William A. Biché Department of Radiology, Boston University Medical Center, Boston
University School of Medicine, Boston, MA, USA
Odion Binitie H. Lee Moffitt Cancer Center, Tampa, FL, USA
Jennifer M. Binning Moffitt Cancer Center, Tampa, USA
Francesco Boccardo San Martino Hospital, University of Genoa, Genoa, Italy
Genevieve Boland Department of Surgery, Massachusetts General Hospital, Boston, MA,
USA
Amanda Brisco H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL, USA
Zachary Brown Division of Surgical Oncology, Department of Surgery, The James Cancer
Hospital and Solove Research Institute, The Ohio State University Wexner Medical Center and
James Cancer Hospital, Columbus, OH, USA
Azadeh Carr University of Southern California, Keck School of Medicine, Los Angeles,
CA, USA
Editors and Contributors xvii

Peter R. Carroll University of California, San Francisco, CA, USA

Angelika Chachaj Department of Angiology, Wroclaw Medical University, Wroclaw, Poland
Bindupriya Chandrasekaran Indiana University School of Medicine, Indianapolis, IN,
USA
Connie Y. Chang Division of Musculoskeletal Imaging and Intervention, Department of
Radiology, Massachusetts General Hospital, Boston, MA, USA
Alexandra Charruyer-Reinwald Department of Dermatology, Eli and Edythe Broad Center
of Regeneration Medicine and Stem Cell Research, University of California San Francisco,
San Francisco, CA, USA
Tom Wei-Wu Chen Graduate Institute of Oncology, College of Medicine, National Taiwan
University, Taipei, Taiwan
Department of Oncology, National Taiwan University Hospital, Taipei, Taiwan
Dhananjay A. Chitale Department of Pathology and Laboratory Medicine, Henry Ford
Health System, Detroit, MI, USA
Christine Chow Department of Medicine, Division of Hematology Oncology, University of
California, San Francisco, San Francisco, CA, USA
Melissa Chow Department of Medicine, Division of Hematology Oncology, University of
California, San Francisco, San Francisco, CA, USA
Hubert H. Chuang Department of Nuclear Medicine, The University of Texas MD Anderson
Cancer Center, Houston, TX, USA
Denis Cochonneau Nantes Université, CNRS, US2B, UMR 6286, Institut de Cancérologie
de l’Ouest, Blvd Jacques Monod, Saint-Herblain, France
Nicolás Costa-Fraga Cancer Epigenomics, Translational Medical Oncology Group
(Oncomet), Health Research Institute of Santiago (IDIS), University Clinical Hospital of
Santiago (CHUS/SERGAS), Santiago de Compostela, Spain
Colleen M. Costelloe Department of Musculoskeletal Imaging, The University of Texas MD
Anderson Cancer Center, Houston, TX, USA
Anne E. Cress Department of Cellular and Molecular Medicine, The University of Arizona
Cancer Center, Tucson, AZ, USA
Brian Czerniecki Breast Department, H. Lee Moffitt Cancer Center, Tampa, FL, USA
Sarah K. Daley Advanced Pathology Research Fellow in Lymphology, Department of
Surgery, University of Arizona College of Medicine, Tucson, AZ, USA
Adil I. Daud Department of Medicine, Division of Hematology Oncology, University of
California, San Francisco, San Francisco, CA, USA
Michael T. Dellinger Division of Surgical Oncology, Department of Surgery and the Hamon
Center for Therapeutic Oncology Research, UT Southwestern Medical Center, Dallas, TX,
USA
Benjamin Deschner Department of Surgery, Mountain Area Health Education Center,
Asheville, NC, USA
Angel Diaz-Lagares Cancer Epigenomics, Translational Medical Oncology Group
(Oncomet), Health Research Institute of Santiago (IDIS), University Clinical Hospital of
Santiago (CHUS/SERGAS), Santiago de Compostela, Spain
Centro de Investigación Biomédica en Red Cáncer (CIBERONC), Madrid, Spain
xviii Editors and Contributors

Brittney L. Dickey Moffitt Cancer Center, Tampa, USA

Ashley V. DiMarco Department of Pharmacology and Cancer Biology, Duke University
School of Medicine, Durham, NC, USA
Robert P. Erickson Department of Pediatrics, University of Arizona College of Medicine,
Tucson, AZ, USA
Timothy J. Evans H. Lee Moffitt Cancer Center, Tampa, FL, USA
Mark B. Faries The Angeles Clinic and Research Institute and Cedars-Sinai Medical Center,
Los Angeles, CA, USA
Umme Farwa University of California San Diego, La Jolla, CA, USA
Laura Favazza DO Pathology and Laboratory Medicine, Henry Ford Hospital, Detroit, MI,
USA
S. Fenerty Temple University Hospital, Philadelphia, PA, USA
Carla S. Fisher Indiana University School of Medicine, Indianapolis, IN, USA
Samuel Ford Department of Sarcoma Surgery, Queen Elizabeth Hospital, Birmingham, UK
Benjamin Franc Department of Radiology, Division of Nuclear Medicine and Molecular
Imaging, Stanford University, Stanford, CA, USA
Timothy L. Frankel Department of Surgery, Michigan Medicine, University of Michigan,
Ann Arbor, MI, USA
Jeremy S. Frieling Department of Tumor Biology, H. Lee Moffitt Cancer Center & Research
Institute, Tampa, FL, USA
Alexander N. Garcia University of Radiation Oncology, Tucson, USA
Nina Marie G. Garcia Department of Pharmacology and Cancer Biology, Duke University
School of Medicine, Durham, NC, USA
Xin Geng Cardiovascular Biology Research Program, Oklahoma Medical Research
Foundation, Oklahoma City, OK, USA
Department of Cell Biology, University of Oklahoma Health Sciences Center, Oklahoma City,
OK, USA
Ruby Ghadially Department of Dermatology, Eli and Edythe Broad Center of Regeneration
Medicine and Stem Cell Research, University of California San Francisco, San Francisco, CA,
USA
Armando Giuliano Department of Surgery, Cedars-Sinai Medical Center, Los Angeles, CA,
USA
Juan C. Gomez-Gelvez Department of Pathology and Laboratory Medicine, Henry Ford
Health System, Detroit, MI, USA
Ricardo J. Gonzalez Moffitt Cancer Center, Tampa, FL, USA
Allison C. Greene Albany Medical Center, Albany, NY, USA
Brian D. Griffith Department of Surgery, Michigan Medicine, University of Michigan, Ann
Arbor, MI, USA
Valerie Grignol Division of Surgical Oncology, Department of Surgery, The Ohio State
University Wexner Medical Center, Columbus, OH, USA
Alessandro Gronchi Department of Surgery, Fondazione IRCCS Istituto Nazionale dei
Tumori, Milan, Italy
Editors and Contributors xix

Sina Habibollahi Division of Musculoskeletal Imaging and Intervention, Department of

Radiology, Massachusetts General Hospital, Boston, MA, USA
S. Haider Henry Ford Health System, Detroit, USA
Russell J. Hamilton University of Arizona, Tucson, AZ, USA
Z. Hammoud Department of thoracic surgery Ascension Providence Hospital, Southfield,
MI, USA
Department of surgery, Wayne State University School of Medicine, Detroit, MI, USA
Dale Han Division of Surgical Oncology, Oregon Health and Science University, Portland,
OR, USA
William L. Harryman The University of Arizona Cancer Center, Tucson, AZ, USA
Eric R. Henderson Dartmouth Hitchcock Medical Center, Lebanon, NH, USA
Dominique Heymann Nantes Université, CNRS, US2B, UMR 6286, Institut de Cancérologie
de l’Ouest, Blvd Jacques Monod, Saint-Herblain, France
Department of Oncology and Metabolism, University of Sheffield, Sheffield, UK
Tina J. Hieken Mayo Clinic, Rochester, MN, USA
Paul Hofman Université Côte d’Azur, CNRS, INSERM, IRCAN, Centre Antoine Lacassagne,
FHU OncoAge, Team 4, Nice, France
Université Côte d’Azur, CHU Nice, FHU OncoAge, Laboratory of Clinical and Experimental
Pathology, Pasteur Hospital, Nice, France
Université Côte d’Azur, CHU Nice, FHU OncoAge, Hospital-Integrated Biobank BB-0033-
00025, Nice, France
Q. A. Ho Department of Radiation Oncology, The University of Arizona, Tucson, AZ, USA
Francis J. Hornicek Department of Orthopaedics, Sarcoma Multidisciplinary Care Group,
Sylvester Comprehensive Cancer Center, University of Miami Miller School of Medicine,
Los Angeles, CA, USA
Department Orthopaedics, University Miami Miller School of Medicine, Miami, FL, USA
Erika Santos Horta University of Arkansas for Medical Sciences, Little Rock, AR, USA
C. C. Hsu Department of Radiation Oncology, The University of Arizona, Tucson, AZ, USA
Ambrose J. Huang Division of Musculoskeletal Imaging and Intervention, Department of
Radiology, Massachusetts General Hospital, Boston, MA, USA
Paul Huang Division of Molecular Pathology, The Institute of Cancer Research, London, UK
Rebecca E. Huber University of Massachusetts Amherst, Amherst, MA, USA
Hatim Husain University of California San Diego, La Jolla, CA, USA
James Isom Department of Dermatology and Cutaneous Surgery, University of South Florida
Morsani College of Medicine, Tampa, FL, USA
Sivan Izraely The Shmunis School of Biomedicine and Cancer Research, George S. Wise
Faculty of Life Science, Tel Aviv University, Tel Aviv, Israel
David G. Jackson MRC Human Immunology Unit, MRC Weatherall Institute of Molecular
Medicine, University of Oxford, Oxford, UK
Christopher Javadi Stanford University, Stanford, CA, USA
P. Jonnalagadda Temple University Hospital, Philadelphia, PA, USA
xx Editors and Contributors

David M. Joyce H. Lee Moffitt Cancer Center, Tampa, FL, USA

Camille Jubelin Nantes Université, CNRS, US2B, UMR 6286, Institut de Cancérologie de
l’Ouest, Blvd Jacques Monod, Saint-Herblain, France
Stanley Kalata Section of General Surgery, Department of Surgery, University of Michigan
Medical School, Ann Arbor, MI, USA
Giorgos C. Karakousis Division of Endocrine and Oncologic Surgery, Department of
Surgery, University of Pennsylvania, Philadelphia, PA, USA
Mohammed Kashani-Sabet California Pacific Medical Center Research Institute, San
Francisco, CA, USA
Igor Katsyv Department of Pathology and Cell Biology, Columbia University, New York,
NY, USA
Joshua K. Kays Cedars-Sinai Medical Center, Los Angeles, CA, USA
Jessica M. Keilson Emory University, Atlanta, GA, USA
S. Ray Kenney RareCyte, Inc., Seattle, WA, USA
Faisal Khurshid University of California San Diego, La Jolla, CA, USA
John Kiluk Breast Department, H. Lee Moffitt Cancer Center, Tampa, FL, USA
Alex C. Kim Division of Surgical Oncology, Department of Surgery, The James Cancer
Hospital and Solove Research Institute, The Ohio State University Wexner Medical Center,
Columbus, OH, USA
Hyuna Kim University of Massachusetts Amherst, Amherst, MA, USA
Kevin B. Kim California Pacific Medical Center Research Institute, San Francisco, CA, USA
Richard W. Kim School of Medicine, University of California San Francisco, San Francisco,
CA, USA
Olena Kis Department of Pathology and Laboratory Medicine, Henry Ford Health System,
Detroit, MI, USA
Yuko Kitagawa Department of Surgery, Keio University School of Medicine, Tokyo, Japan
Scott Kizy Moffitt Cancer Center, Tampa, FL, USA
Lyonell B. Kone University of Illinois at Chicago—Metropolitan Group Hospitals, Chicago,
IL, USA
Christina A. LeBedis Department of Radiology, Boston University Medical Center, Boston
University School of Medicine, Boston, MA, USA
I. Lee Henry Ford Health System, Detroit, USA
Stanley P. Leong Department of Surgery and Melanoma Center, California Pacific Medical
Center and Research Institute, San Francisco, CA, USA
James K. C. Liu H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL, USA
Mike Lowe Department of Surgery, Emory University, Atlanta, GA, USA
Claire Lugassy Department of Translational Research, Institut Curie, Paris, France
Conor C. Lynch Department of Tumor Biology, H. Lee Moffitt Cancer Center & Research
Institute, Tampa, FL, USA
John E. Madewell Department of Musculoskeletal Imaging, The University of Texas MD
Anderson Cancer Center, Houston, TX, USA
Editors and Contributors xxi

Rita Das Mahapatra University of Massachusetts Amherst, Amherst, MA, USA

Najjia N. Mahmoud Division of Colon and Rectal Surgery, Department of Surgery, University
of Pennsylvania, Philadelphia, PA, USA
Shishir K. Maithel Emory University, Atlanta, GA, USA
Ajay V. Maker Department of Surgery, University of California San Francisco, San Francisco,
CA, USA
Sindhu J. Malapati Department of Hematology/Oncology, Ascension St. John Hospital,
Detroit, MI, USA
Kartik Mani Department of Radiation Oncology, Stony Brook University Hospital, Stony
Brook, NY, USA
Dario Marchetti Department of Internal Medicine, Division of Molecular Medicine, The
University of New Mexico Health Sciences Center, Albuquerque, NM, USA
Takaaki Masuda Kyushu University Beppu Hospital, Beppu, Japan
Ahmad Mattour Department of Hematology/Oncology, Henry Ford Cancer Institute, Detroit,
MI, USA
Shuhei Mayanagi Division of Esophageal Surgery, Shizuoka Cancer Center Hospital,
Shizuoka, Japan
James F. McLoughlin Department of Radiology, Boston University Medical Center, Boston
University School of Medicine, Boston, MA, USA
Jennifer L. McQuade Department of Melanoma Medical Oncology, The University of Texas
MD Anderson Cancer Center, Houston, TX, USA
Winta T. Mehtsun Department of Surgery, Massachusetts General Hospital, Boston, MA,
USA
Department of Surgery, Brigham and Women’s Hospital, Boston, MA, USA
Dana Farber Cancer Institute, Boston, MA, USA
Nathan W. Mesko Cleveland Clinic, Cleveland, OH, USA
Jane L. Messina Department of Dermatology and Cutaneous Surgery, University of South
Florida Morsani College of Medicine, Tampa, FL, USA
Departments of Pathology and Cutaneous Oncology, Moffitt Cancer Center, Tampa, FL, USA
Departments of Pathology and Cell Biology and Oncologic Sciences, University of South
Florida Morsani College of Medicine, Tampa, FL, USA
Koshi Mimori Kyushu University Beppu Hospital, Beppu, Japan
Marc Moncrieff Norfolk & Norwich University Hospital NHS Trust, Norwich, UK
Emilie Moranton Nantes Université, CNRS, US2B, UMR 6286, Institut de Cancérologie de
l’Ouest, Blvd Jacques Monod, Saint-Herblain, France
B. Morse H. Lee Moffitt Cancer Center, Tampa, FL, USA
Laura Muinelo-Romay Liquid Biopsy Analysis Unit, Translational Medical Oncology
Group (Oncomet), Health Research Institute of Santiago (IDIS), University Clinical Hospital
of Santiago (CHUS/SERGAS), Santiago de Compostela, Spain
Centro de Investigación Biomédica en Red Cáncer (CIBERONC), Madrid, Spain
Javier Muñoz-Garcia Nantes Université, CNRS, US2B, UMR 6286, Institut de Cancérologie
de l’Ouest, Blvd Jacques Monod, Saint-Herblain, France
xxii Editors and Contributors

Takafumi Nakano Kyushu University Beppu Hospital, Beppu, Japan

H. Nasser Henry Ford Hospital, Detroit, MI, USA
Rui P. L. Neves Experimental Surgical Oncology, Department of General, Visceral and
Pediatric Surgery, University Hospital and Medical Faculty of the Heinrich-Heine University
Düsseldorf, Düsseldorf, Germany
Lukas M. Nystrom Cleveland Clinic, Cleveland, OH, USA
Valeriya Pankova Division of Molecular Pathology, The Institute of Cancer Research,
London, UK
Nainesh Parikh Department of Diagnostic Imaging and Interventional Radiology, H. Lee
Moffitt Cancer Center and Research Institute, Tampa, FL, USA
Howard Y. Park Department of Orthopaedic Surgery, David Geffen School of Medicine at
UCLA, Los Angeles, CA, USA
Jacob A. Pawloski Department of Neurosurgery, Hermelin Brain Tumor Center, Henry Ford
Hospital, Detroit, MI, USA
Howard Peach Leeds Teaching Hospitals NHS Trust, Leeds, UK
Shelly R. Peyton University of Massachusetts Amherst, Amherst, MA, USA
Neil Piller Flinders Medical Centre, Department of Surgery, School of Medicine, Bedford
Park, SA, Australia
Mara A. Piltin Mayo Clinic, Rochester, MN, USA
Jose M. Pimiento Moffitt Cancer Center, Tampa, FL, USA
Raphael Pollock Division of Surgical Oncology, Department of Surgery, The Ohio State
University Wexner Medical Center and James Cancer Hospita, Columbus, OH, USA
A. Popoff Henry Ford Hospital, Detroit, MI, USA
Sima P. Porten University of California, San Francisco, CA, USA
C. Potti Henry Ford Hospital, Detroit, MI, USA
George Poultsides Stanford University, Stanford, CA, USA
Julie Rathwell Moffitt Cancer Center, Tampa, USA
Kristen E. Rhodin Department of Surgery, Duke University, Durham, NC, USA
Adam M. Robin Department of Neurosurgery, Hermelin Brain Tumor Center, Henry Ford
Hospital, Detroit, MI, USA
Aitor Rodriguez-Casanova Cancer Epigenomics, Translational Medical Oncology Group
(Oncomet), Health Research Institute of Santiago (IDIS), University Clinical Hospital of
Santiago (CHUS/SERGAS), Santiago de Compostela, Spain
Roche-Chus Joint Unit, Translational Medical Oncology Group (Oncomet), Health Research
Institute of Santiago (IDIS), Santiago de Compostela, Spain
Robert E. Roses University of Pennsylvania, Philadelphia, PA, USA
Samuel Ryu Department of Radiation Oncology, Stony Brook University, Stony Brook, NY,
USA
Hideyuki Saito Kyushu University Beppu Hospital, Beppu, Japan
Elizabeth A. Sangalang Department of Medicine, Division of Hematology Oncology,
University of California, San Francisco, San Francisco, CA, USA
Editors and Contributors xxiii

Nirali Sanghavi University of California San Diego, La Jolla, CA, USA

Luke V. Selby The Ohio State University James Cancer Center, Columbus, OH, USA
Akash A. Shah Department of Orthopaedic Surgery, David Geffen School of Medicine at
UCLA, Los Angeles, CA, USA
Brandon M. Shaw Department of Pathology and Laboratory Medicine, Henry Ford Health
System, Detroit, MI, USA
David Shibata Department of Surgery, University of Tennessee Health Science Center,
Memphis, TN, USA
Qihui Shi Institutes of Biomedical Sciences and Minhang Hospital, Fudan University,
Shanghai, China
Emily Siegel Breast Department, H. Lee Moffitt Cancer Center, Tampa, FL, USA
Sunny R. K. Singh Department of Hematology/Oncology, Henry Ford Cancer Institute,
Detroit, MI, USA
Andrew J. Sinnamon Moffitt Cancer Center, Tampa, FL, USA
Alexander N. Slade Department of Radiation Oncology, Stony Brook University Hospital,
Stony Brook, NY, USA
J. Snyder Henry Ford Health System, Detroit, USA
Gaya Spolverato Department of Surgical, Oncological and Gastroenterological Sciences,
University of Padova, Padova, Italy
R. Sathish Srinivasan Cardiovascular Biology Research Program, Oklahoma Medical
Research Foundation, Oklahoma City, OK, USA
Department of Cell Biology, University of Oklahoma Health Sciences Center, Oklahoma City,
OK, USA
Hayley Standage Division of Surgical Oncology, Oregon Health and Science University,
Portland, OR, USA
Baldassarre Stea University of Arizona, Tucson, USA
Nikolas H. Stoecklein Experimental Surgical Oncology, Department of General, Visceral and
Pediatric Surgery, University Hospital and Medical Faculty of the Heinrich-Heine University
Düsseldorf, Düsseldorf, Germany
Richard J. Straker III Division of Endocrine and Oncologic Surgery, Department of Surgery,
University of Pennsylvania, Philadelphia, PA, USA
Division of Colon and Rectal Surgery, Department of Surgery, University of Pennsylvania,
Philadelphia, PA, USA
Sheng-Fang Su Graduate Institute of Oncology, College of Medicine, National Taiwan
University, Taipei, Taiwan
YongLin Institute of Health, National Taiwan University (National Taiwan University YongLin
Scholar), Taipei, Taiwan
Andrzej Szuba Department of Angiology, Wroclaw Medical University, Wroclaw, Poland
Seiichiro Takao Kyushu University Beppu Hospital, Beppu, Japan
Alicia M. Terando University of Southern California, Keck School of Medicine, Los Angeles,
CA, USA
xxiv Editors and Contributors

Kerry L. Thomas Clinical Associate Professor, Department of Radiology, University of

North Carolina at Chapel Hill, School of Medicine, Tampa, FL, USA
Fabio Tirotta Department of Sarcoma Surgery, Queen Elizabeth Hospital, Birmingham, UK
Michael T. Torchia Dartmouth Hitchcock Medical Center, Lebanon, NH, USA
Ning-Hsuan Tseng University of Massachusetts Amherst, Amherst, MA, USA
William W. Tseng Department of Surgery, Division of Breast, Endocrine and Soft Tissue
Surgery, Keck School of Medicine of University of Southern California, Los Angeles, CA,
USA
Kiran K. Turaga Division of Surgical Oncology, Department of Surgery, University of
Chicago, Chicago, IL, USA
Andrew Turk Department of Pathology and Cell Biology, Columbia University, New York,
NY, USA
Przemyslaw Twardowski Medical Oncology and Urologic Oncology, Department of
Urologic Oncology, John Wayne Cancer Institute, Santa Monica, CA, USA
Raul Fernando Valenzuela Department of Musculoskeletal Imaging, The University of
Texas MD Anderson Cancer Center, Houston, TX, USA
P. van der Laan Department of Surgical Oncology, Sarcoma and Melanoma Unit, The
Netherlands Cancer Institute − Antoni van Leeuwenhoek, Amsterdam, The Netherlands
Winan J. van Houdt Department of Surgical Oncology, Sarcoma and Melanoma Unit, The
Netherlands Cancer Institute − Antoni van Leeuwenhoek, Amsterdam, The Netherlands
Joao R. T. Vicentini Division of Musculoskeletal Imaging and Intervention, Department of
Radiology, Massachusetts General Hospital, Boston, MA, USA
Tobias Walbert Hermelin Brain Tumor Center, Henry Ford Health System, Detroit, MI, USA
Christopher P. Wang H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL, USA
Zhiqiang Wang Department of Pathology and Laboratory Medicine, Henry Ford Hospital,
Detroit, MI, USA
Jennifer A. Wargo Department of Surgical Oncology, The University of Texas MD Anderson
Cancer Center, Houston, TX, USA
Samuel L. Washington III University of California, San Francisco, CA, USA
Michael G. White Department of Surgical Oncology, The University of Texas MD Anderson
Cancer Center, Houston, TX, USA
Hunter D. D. Witmer Division of Surgical Oncology, Department of Surgery, University of
Chicago, Chicago, IL, USA
Marlys H. Witte Department of Surgery, Neurosurgery, and Pediatrics, University of Arizona
College of Medicine, Tucson, AZ, USA
Russell S. Witte Department of Medical Imaging, University of Arizona, Tucson, AZ, USA
Isaac P. Witz The Shmunis School of Biomedicine and Cancer Research, George S. Wise
Faculty of Life Science, Tel Aviv University, Tel Aviv, Israel
Ziyuan Zhang Institutes of Biomedical Sciences and Minhang Hospital, Fudan University,
Shanghai, China
Mateo Ziu Inova Neurosciences and Spine Institute, University of Virginia Medical School,
Falls Church, VA, USA
 Part I
Genetic and Molecular Pathways of
Cancer Growth and Metastasis

Dhananjay Chitale

In the United States, cancer is the second most leading cause of death after cardiovascular
diseases [1]. In addition to high mortality, there is a lot of human emotional and physical suf-
fering inflicted by cancer. The essence of finding new targeted therapies to achieve complete
remission is to complement the present mostly anatomic-based treatment protocols with a full
understanding of the neoplasms at the molecular genetic level.
It is well known that some primary malignant tumors have a higher propensity to aggres-
sively invade local sites while others metastasize at an early stage. Pathologists have been
classifying tumors using morphological correlates such as differentiation or grade based on the
closeness of the tumor to its normal counterparts. Nuclear pleomorphism, mitosis, necrosis,
infiltration, and lymphatic and vascular invasion are microscopic features that, when correlated
with clinical outcomes, predict the aggressiveness of the tumor. The anatomic pathology stag-
ing is an integral part of AJCC and TNM staging classifications that consider the tumor size,
local lymph node status, and metastasis outside the local, regional anatomic lymphatic drain-
age. Overall, a larger tumor size is typically associated with a higher chance of metastasis to
locoregional lymph nodes or distant sites. These parameters are vital for surgeons and oncolo-
gists to tailor therapies based on grade and stage. However, we know that tumor behaviors vary
significantly, even within the same AJCC stages. Especially in the early-stage tumors where
there is no evidence of local or distant spread of the tumor, it is often difficult to predict the
biological potential of the malignant tumor based purely on histologic and radiologic param-
eters. Therefore, the knowledge of molecular programs of these tumors is fundamental to
guide therapies. The current testament to that concept is reflected in the evolution of the stag-
ing checklists that are slowly getting transformed from purely anatomic pathology-driven
items to a combination of anatomic and molecular biomarkers that have been introduced in the
recent 8th AJCC staging classifications to fine-tune tumor stages further to direct targeted
therapeutics. These stratification strategies will help separate more aggressive tumors within
the same anatomic stage for optimal treatments and outcomes. Breast cancer AJCC staging [2,
3] is one such example where biomarker status such as estrogen receptor (ER), progesterone
receptor (PgR), and human epidermal growth factor receptor 2 (Her2) expression status are
included along with multigene signature panels such as 21-gene Oncotype Dx Recurrence
Score, MammaPrint, the 12-gene EndoPredict, or the PAM50 test making its way into clinical
decision-making.
According to Willis, a neoplasm is defined as an abnormal mass of tissue, the growth of
which exceeds and is uncoordinated with that of the normal tissues and persists in the same
excessive manner after the cessation of the stimuli which evoked the change [4]. From the
original descriptions of neoplasm by Willis, after decades of clinical and experimental obser-
vations, we know that neoplasms are a mixture of heterogeneous cells that start from “clonal”
progeny of a single cell that acquires or less commonly inherits mutations, thus gaining sur-
vival advantage through natural selection. The tumor cells sequentially acquire the hallmarks
2 Genetic and Molecular Pathways of Cancer Growth and Metastasis

of cancer from a premalignant transformed state to a widely metastatic disease through a multistep
process.
The past 40+ years of cancer research has given deep insights into molecular characteristics and under-
pinnings that drive tumor biology and propensity towards metastasis. There is a large body of knowledge
that has further underscored the complexity of cancer as a disease driven by dynamic changes in the genome
through a multistep acquisition of large numerical or structural changes to the chromosomal complement or
subtle point mutations in oncogenes leading to gain of function and tumor-suppressor genes with loss of
function [5]. With the advent of molecular profiling and understanding of tumors at the genomic level, we
now know that tumor is not a uniform collection of cancer cells but rather a mixture of different types of
cells. It includes tumor cells, immune cells, supporting stromal cells, blood vessels, stem cells, and stromal
matrix, all working together as an autonomous organ. Consolidation of the current knowledge of different
pathways in human tumorigenesis led to the first conceptual seminal paper by Hanahan and Weinberg in the
year 2000 that described the hallmarks of cancer as an organizing principle of oncology [6]. The hallmarks
of cancer comprised six biological capabilities acquired by cancer cells through a multistep process. These
were updated in 2011 where two additional key properties to a total of 8 and 2 additional functional roles
attributed to the tumor microenvironment were added [7]. We have tried to deconstruct these complex hall-
marks of cancers described in the following chapters in detail.
In this section of the book, different chapters will describe the hallmarks of cancer with molecular under-
pinnings of primary tumors as a starting point to set the stage for subsequent chapters, followed by a unify-
ing concept of genomic changes that will elucidate the mutational landscape of cancers. The next chapter
will be on changes in the tumor microenvironment as a co-conspirator in tumorigenesis. Then we will deep
dive into the hallmarks of metastasis with its molecular underpinnings. And finally, we will summarize
hereditary cancer syndromes and their role in cancer metastasis. Through these systematic descriptions of
cancer genomics and microenvironments, we hope to give the readers insights and understanding of cancer
as a “system biology” where treating the entire system rather than individual components is necessary to
defeat this deadly disease effectively, hopefully soon.

References
1. Xu J, Murphy SL, Kockanek KD, Arias E. Mortality in the United States, 2018. NCHS Data Brief.
2020;355:1–8. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/32487294
2. Giuliano AE, Edge SB, Hortobagyi GN. Eighth edition of the AJCC cancer staging manual: breast can-
cer. Ann Surg Oncol. 2018;25(7):1783–5. https://doi.org/10.1245/s10434-­018-­6486-­6.
3. Jang N, Choi JE, Kang SH, Bae YK. Validation of the pathological prognostic staging system proposed
in the revised eighth edition of the AJCC staging manual in different molecular subtypes of breast can-
cer. Virchows Arch. 2019;474(2):193–200. https://doi.org/10.1007/s00428-­018-­2495-­x.
4. Willis RA. The spread of tumours in the human body. 2nd ed. London: Butterworth & Co.; 1952. p. ix.
447–28.
5. Kinzler KW, Vogelstein B. Lessons from hereditary colorectal cancer. Cell. 1996;87(2):159–70. https://
doi.org/10.1016/s0092-­8674(00)81333-­1.
6. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000;100(1):57–70. https://doi.org/10.1016/
s0092-­8674(00)81683-­9.
7. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144(5):646–74. ­https://
doi.org/10.1016/j.cell.2011.02.013.
 Hallmarks of Cancer: Molecular
Underpinnings 1
Dhananjay A. Chitale

Abstract Hanahan and Weinberg initially published in their seminal

Cancer, a genetic disease, is an autonomous organ com- paper published in 2000 [1]. They initially described six such
posed of heterogeneous cancer cell clones supported by properties, viz. self-sufficiency in growth signals, insensitiv-
its microenvironment. The cancer cells gain a set of prop- ity to growth-inhibitory (antigrowth) signals, evasion of pro-
erties and characteristics called hallmarks of cancer and grammed cell death (apoptosis), limitless replicative
enabling characteristics, respectively, that highjack the potential, sustained angiogenesis, and tissue invasion and
normal cellular mechanisms for their purpose. These are metastasis. These properties are acquired by the cancer cell
thought to be acquired in a stepwise fashion, in no par- with an inherent assumption that normal cells progressively
ticular order, giving cancer cells a survival advantage over transform into a neoplastic state through stepwise gain of
normal cells, making them self-sustainable, invade and mutations that confer survival advantage and disseminate at
metastasize to distant locations. This chapter provides a distant sites. The stages or steps of these changes are neces-
brief overview and key concepts on the molecular under- sarily not linear, and different tumors may have different
pinnings of cancer cells composed of eight hallmarks of sequential gains in these properties. After a decade or so,
cancer and two enabling characteristics. with additional accumulation of molecular literature and
understanding of cancer, in 2011, Hanahan and Weinberg
revised the original hallmarks and added two new ones to the
Learning Objectives original six, and expanded on the functional roles and contri-
• To understand the molecular genomic alterations in butions made by the tumor microenvironment [2].
primary cancers conceptually as Hallmarks of In this chapter, using Hanahan and Weinberg’s publica-
Cancer as the basis for cancer metastasis tions as the backbone [1, 2], we focus on detailed descriptions
• Based on the understanding of genomic underpin- of these properties that produce the cellular phenotype and
nings of cancers, learn diagnostic, prognostic, and briefly elucidate the molecular underpinnings to gain such
predictive therapeutic strategies properties in the primary tumors with some examples. The
details of underlying molecular mechanisms such as point
mutations, gene translocations/fusions, deletions, insertions,
amplification, chromosomal aneuploidy, transcriptional
1.1 Introduction changes (mRNA profiles), epigenetic changes, MicroRNA
profiles, and regulating DNA structural regulatory changes
Cancer is a genetic disease underpinned by mutations, either are described in the next chapter titled “Unifying Concept of
somatic or rarely genetically inherited, layered with epigen- Genomic Changes: The Mutational Landscape of Cancers.”
etic alterations that lead to an autonomous organ that is under
its own control, defying normal cellular signaling processes.
The molecular genetic characteristics of cancer impart tumor 1.2 Sustaining Proliferative Signaling
cells a set of properties called hallmarks of cancer that
Normal tissues regulate the proliferation of cells tightly con-
D. A. Chitale (*) trolled through growth factor-induced cell signaling, thus
Department of Pathology and Laboratory Medicine, controlling the tissue architecture and function. This process
Henry Ford Health System, Detroit, MI, USA is orchestrated through a growth factor binding to its specific
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 3

S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_1
4 D. A. Chitale

receptor transiently, activating a cascade of several down- through constitutive activation of components of the signal-
stream cytoplasmic signal-transduction proteins transmitting ing pathways downstream of the receptors and independent
the signal to the nucleus. This, in turn, induces and activates of growth factor ligand activation. Many recurrent somatic
transcription factors and epigenetic alterations initiating mutations have been detected by high-throughput DNA
DNA transcription. Genes encoding proteins that promote sequencing that predicts the constitutive activation of sig-
cell cycle progression ultimately result in cell division. The nal transduction. Recurrent activating mutations in the
expression of other genes at the same time also leads to BRAF gene, usually due to point mutations, have been noted
changes in metabolic activity (energy metabolism) and pro- in many solid tumors [9], including 40% of malignant mela-
teins supporting cell survival that are needed for optimal nomas [10]. These mutations lead to constitutively active
growth. The knowledge about the proliferative growth factor signaling through mitogen-activated protein kinase(MAPK)
signaling through the paracrine mechanism remains limited pathway that regulates a wide variety of cellular processes
[2]. However, in cancer cells, the most fundamental trait is such as proliferation, differentiation, apoptosis, and stress
their ability to sustain proliferative signals, many through responses [11]. Another canonical signaling pathway that is
tyrosine kinase activity, even in the absence of growth fac- frequently upregulated due to mutations in the catalytic sub-
tors. This self-sustaining property is achieved through sev- units of phosphoinositide 3-kinase (PI3K) is involved in cel-
eral molecular mechanisms involving one or multiple parts lular functions such as cell growth, proliferation,
of the cell signaling pathways. Primary mechanisms include differentiation, motility, survival, and intracellular traffick-
mutations in the oncogenes that constitutively activate cell ing [12, 13]. The tumor cells can also attain self-sufficiency
signaling pathways and deregulation of inhibitory feed- due to the deregulation of nuclear regulatory proteins.
back loops, enhancing cell signaling. Transcriptional activators, e.g., MYC, a master transcription
factor that regulates genes needed for rapid cell growth, are
overexpressed either through translocation (Burkitt’s lym-
1.2.1 Mutation in Oncogenes phoma) [14] or through amplification (breast carcinoma,
lung adenocarcinoma, colon adenocarcinoma, prostate ade-
Normal cellular genes that promote cell proliferation called nocarcinoma) [9] in aggressive tumors. Dysregulation of
proto-oncogenes are mutated or overexpressed autono- cell cycle regulators such as cyclins often occurs in some
mously in many cancers, which are then termed as onco- tumors. For example, cyclin D1 (CCDN1) is upregulated due
genes. An array of mechanisms alters the function of the to translocation (CCND1/IGH) in mantle cell lymphoma,
oncogenes by constitutive activation and resistance to con- plasma cell myeloma, making the tumor cells proliferate
trol by normal external signals free from checkpoints. The constitutively.
molecular underpinnings of this hallmark of cancer can
involve growth factors, growth factor receptors, proteins
involved in signal transduction, nuclear regulatory proteins, 1.2.2  eregulation of Negative Feedback
D
and cell cycle regulator. Loops
One pathway is through increasing growth factor pro-
duction. The cancer cells may produce growth factors them- Normal cellular mechanisms of cell signaling include nega-
selves (autocrine capability); for example, glioblastomas tive feedback loops that inactivate the activated signaling to
overexpress platelet-derived growth factor (PDGF) and maintain homeostasis. In tumors, the genes controlling this
PDGF receptor (PDGFR) [3]. Tumor cells may send signals negative feedback are downregulated in diverse tumor
to stimulate the normal cells in the tumor microenvironment, types, thus enhancing the mitogenic and proliferative sig-
which supply the growth factors (paracrine function). nals. For example, many carcinomas of solid organs such as
Amplification of FGF3 growth factor is observed in many the colon, pancreas, lung etc., harbor activating mutations
cancers (stomach, bladder, breast) [4]. Another mechanism in the KRAS gene [9]. The effects on downstream signaling
is producing high levels of receptor proteins, such as Her2 are not due to direct activation. But the constitutive activa-
(ERB1B2) gene amplification in breast cancers. In this case, tion of RAS protein through these mutations leads to loss of
receptor signaling can be deregulated and sensitized even in intrinsic GTPas activity of RAS, thus disrupting the auto-
the presence of normal amounts of ligand [5–8]. In lung ade- regulatory negative-feedback mechanism [2]. Another
nocarcinomas, discovery of activating hot spot mutations example is heterogeneous groups of enzymes called Dual-
in the tyrosine kinase domain encoded by exon 19 and 21 of specificity phosphatases (DUSP) that regulate diverse cel-
ERB1B1 (EGFR) receptor has revolutionized the way lung lular signaling cascades, including much-researched
adenocarcinomas are treated using small molecular targeted MAPK phosphatases. Dysregulation of DUSPs has been
therapies (tyrosine kinase inhibitors-TKI). Another well-­ implicated as major modulators of critical signaling path-
known means of cancer signaling upregulation is achieved ways in various tumors [15].
1 Hallmarks of Cancer: Molecular Underpinnings 5

Another well-known example is phosphatase and tensin tivated by somatic mutations [19]. Large-scale genomic
homolog (PTEN), which degrades phosphatidylinositol sequencing has identified similar somatic RB1 gene muta-
3,4,5-phosphate (PIP3), thus dampening PI3K signaling. tions in subsets of the lung, breast, bladder carcinoma, and
Loss of PTEN function due to either somatic mutations such glioblastoma [9], and among the top 5 somatically altered
as deletion or gene silencing by hypermethylation leads to genes in metastatic tumors [20]. The current model is that
amplification of the PI3K signaling in many solid tumors loss of normal cell cycle control occurs due to dysregulation
[12, 13]. Germline mutations in PTEN in Cowden syndrome, of at least one of the four key regulators: CDKN2A, CCND1,
an autosomal dominant inherited condition, predispose indi- CDK4, and RB1 in many human cancers [17]. The loss of
viduals to many benign and malignant tumors of the skin, tumor suppressor function of RB1 occurs directly due to
breast, thyroid, endometrium, colorectum, and kidney [16]. mutation in the RB1 gene or indirectly by the gain of func-
tion CDK/Cyclin D activity keeping RB1 protein in the
hyperphosphorylated state, thus inhibiting the binding and
1.3 Insensitivity to Growth Suppressors sequestering of E2F transcription factors. Release of E2F
leads to transcriptional activation and cell cycle
In the normal cellular proliferation machinery, set of genes progression.
categorized as tumor suppressors control the checkpoints to TP53, also known as Guardian of The Genome, is also a
prevent uncontrolled cell growth. These antiproliferative tumor suppressor gene regulating cell cycle progression,
mechanisms lead to quiescence or permanent cell cycle DNA repair, cellular senescence, and apoptosis. TP53 is
arrest, depending on the scenario. Several tumor suppressor mutated in 50% of all cancers, including lung, colon, breast,
genes are part of this regulatory network limiting cell prolif- ovary, and glioblastoma. Unlike RB1, the loss of function is
eration or survival responding to activated signaling primarily somatic and rarely seen as germline inheritable
(Table 1.1). A similar phenomenon occurs in cancer cells genetic mutation seen in Li-Fraumeni syndrome. Like RB1,
with overactive cell signaling that leads to a nonproliferative the loss of function occurs by somatic mutations in both the
state called oncogene-induced senescence [18]. Like mito- alleles or in cases of germline mutation, and other allele is
genic signaling, growth inhibitory signals also arise outside inactivated by somatic mutation. Another analogy to RB1
of the cells (paracrine) orchestrated through receptors, signal loss of function is that the functional loss of p53 protein can
transducers, and nuclear transcription regulators. Cancer occur in the absence of TP53 mutations. MDM2 gene ampli-
cells evade these checkpoints, thus gaining survival advan- fication observed in 33% of human sarcomas leads to p53
tage and cell proliferation. These can be categorized based deficiency due to accelerated degradation. Transforming pro-
on the molecular underpinnings, viz. genes encoding inhibi- teins of several DNA viruses bind to p53 and promote degra-
tors of mitogenic signaling pathways, inhibitors of cell cycle dation. With the loss of p53 protein function, DNA damage
progression, inhibitors of pro-growth programs of metabo- and, therefore, driver mutations accumulate in oncogenes,
lism and angiogenesis, inhibitors of invasion and metastasis, ultimately leading to the malignant transformation of cells.
enablers of genomic stability, DNA repair factors, and some RB protein transduces growth-inhibitory signals originat-
with unknown mechanisms [17]. It is thought that tumor sup- ing externally and internally and decides whether a cell
pressors have a broader inhibitory effect on many hallmarks should proceed through its growth-and-division cycle. In
of cancer. The key prototypical tumor suppressor proteins comparison, TP53 receives inputs from stress and abnormal-
include RB1 and TP53, both playing a central role as check- ity sensors intracellular signaling. Depending on the degree
point controls for a cell to proliferate or activate growth of genomic damage, the levels of nucleotide pools, growth-­
arrest, senescence, or activate apoptosis. These are described promoting signals, glucose, or oxygenation, TP53 can halt
in detail below. cell cycle progression until normalization. Alternatively, if
Much of our understanding and concepts regarding tumor there is irreparable damage to cellular subsystems, TP53 can
suppressor genes and their function are derived from studies trigger apoptosis [2]. In the absence of both these proteins,
on the first tumor suppression gene discovered, the retino- the cell proliferating characteristic of cancer cells remains
blastoma (RB1) gene, the so-called Governor of The Cell unimpeded.
Cycle. Based on observations between hereditary retinoblas-
tomas vs. sporadic, Knudson proposed the “two-hit” hypoth-
esis of oncogenesis that states that both the alleles (one from 1.3.1 Evasion of Contact Inhibition
mother and one from father) need to be inactivated by similar
or different molecular mechanisms to cause malignancies. In Normal cells stop proliferating upon contact with other cells.
inherited germline RB1 mutation, the second normal allele is This contact inhibition that is part of tissue homeostasis is
inactivated by somatic mutation silencing the gene function. lost in tumors. E-cadherin encoded by the CDH1 gene is a
While in sporadic retinoblastomas, both the alleles are inac- cell adhesion molecule that plays a vital role in
6 D. A. Chitale

Table 1.1 Selected tumor suppressor genes and associated familial syndrome and cancers grouped by cancer hallmarks are listed in this table
Gene Protein Function Familial syndromes Sporadic cancers
Inhibitors of mitogenic signaling pathways
APC Adenomatous Inhibitor of WNT signaling Familial colonic polyps and Carcinomas of stomach, colon,
polyposis coli protein carcinomas pancreas; melanoma
NF1 Neurofibromin-1 Inhibitor of RAS/MAPK Neurofibromatosis type 1 Neuroblastoma, juvenile myeloid
signaling (neurofibromas and malignant leukemia
peripheral nerve sheath tumors)
NF2 Merlin Cytoskeletal stability, hippo Neurofibromatosis type 2 Schwannoma, meningioma
pathway signaling (acoustic schwannoma and
meningioma)
PTCH Patched Inhibitor of hedgehog signaling Gorlin syndrome (basal cell Basal cell carcinoma,
carcinoma, medulloblastoma, medulloblastoma
several benign tumors)
PTEN Phosphatase and tensin Inhibitor of PI3K/AKT signaling Cowden syndrome (variety of Diverse cancers, particularly
homologue benign skin, GI, and CNS carcinomas and lymphoid tumors
growths; breast, endometrial, and
thyroid carcinoma)
SMAD2, SMAD2, SMAD4 Component of the TGF-β signaling Juvenile polyposis Frequently mutated (along with
SMAD4 pathway, repressors of MYC and other components of the TGF-β
CDK4 expression, inducers of signaling pathway) in colonic
CDK inhibitor expression and pancreatic carcinoma
Inhibitors of cell cycle progression
RB Retinoblastoma (RB) Inhibitor of G1/S transition Familial retinoblastoma syndrome Retinoblastoma; osteosarcoma;
protein during cell cycle progression (retinoblastoma, osteosarcoma, carcinomas of breast, colon, lung
other sarcomas)
CDKN2A p16/INK4a and p14/ p16: Negative regulator of Familial melanoma Pancreatic, breast, and
ARF cyclin-dependent kinases; p14, esophageal carcinoma;
an indirect activator of p53 melanoma; certain leukemias
Inhibitors of pro-growth programs of metabolism and angiogenesis
VHL von Hippel–Lindau Inhibitor of hypoxia-induced von Hippel–Lindau syndrome Renal cell carcinoma
(VHL) protein transcription factors (e.g., (cerebellar hemangioblastoma,
HIF1α) retinal angioma, renal cell
carcinoma)
STK11 Liver kinase B1 Activator of AMPK family of Peutz-Jeghers syndrome (GI Diverse carcinomas (5%–20% of
(LKB1) or STK11 kinases; suppresses cell growth polyps, GI cancers, pancreatic cases, depending on the type)
when cell nutrient and energy carcinoma, and other carcinomas)
levels are low
SDHB, Succinate TCA cycle, oxidative Familial paraganglioma, familial Paraganglioma
SDHD dehydrogenase phosphorylation pheochromocytoma
complex subunits B
and D
Inhibitors of invasion and metastasis
CDH1 E-cadherin Cell adhesion, inhibition of cell Familial gastric cancer Gastric carcinoma, lobular breast
motility carcinoma
Enablers of genomic stability
TP53 p53 protein Cell cycle arrest and apoptosis in Li-Fraumeni syndrome (diverse Most human cancers
response to DNA damage cancers)
DNA repair factors
BRCA1, Breast cancer-1 and Repair of double-stranded breaks Familial breast and ovarian Rare
BRCA2 breast cancer-2 in DNA carcinoma; carcinomas of the
(BRCA1 and BRCA2) male breast; chronic lymphocytic
leukemia (BRCA2)
MSH2, MSH1, MLH1, MSH6 DNA mismatch repair Hereditary nonpolyposis colon Colonic and endometrial
MLH1, carcinoma carcinoma
MSH6
Unknown mechanisms
WT1 Wilms tumor-1 (WT1) Transcription factor Familial Wilms tumor Wilms tumor, certain leukemias
MEN1 Menin Transcription factor Multiple endocrine neoplasia-1 Pituitary, parathyroid, and
(MEN1) (pituitary, parathyroid, pancreatic endocrine tumors
and pancreatic endocrine tumors)
CNS Central nervous system, GI Gastrointestinal, TCA Tricarboxylic acid
Adapted from Robbins and Cotran Pathologic Basis of Diseases, tenth Edition, Kumar, V., Abbas, A. K., Aster, J. C., Turner, J. R., Robbins, S. L.,
& Cotran, R. S., Page 292, Table 7.7, Copyright Elsevier, 2021 [17]
1 Hallmarks of Cancer: Molecular Underpinnings 7

contact-­ mediated growth inhibition of epithelial cells. through two significant circuits—extrinsic to cell (death
Disruption of E-cadherin is noted in many solid tumors such receptor—e.g., Fas ligand / Fas receptor) and intrinsic (mito-
as gastric signet ring cell carcinomas and invasive lobular chondrial) cellular programs, and also perforin/granzyme
carcinoma of the breast. This allows cells to easily disaggre- pathway. Intrinsic and extrinsic pathways activate a proteo-
gate due to loss of “stickiness” and promotes distant metas- lytic cascade of caspases (caspase 8, 9, and 3) that destroy
tasis. Another function of E-cadherin is that it binds and the cell in an organized manner to form apoptotic bodies
sequesters β-catenin. Disruption of this complex leads to the observed and reported in pathology literature many decades.
translocation of β-catenin to the nucleus, like the WNT path- Perforin / Granzyme pathway is activated by cytotoxic T
way that promotes proliferation. cells either by activating caspase 10 or directly lead to DNA
Another mechanism by which cells regulate proliferation cleavage. In the intrinsic pathway, there is disruption of
is through the NF2 gene. Merlin, the NF2 gene product, mitochondrial membrane activating BAX/BAK pro-­
­promotes contact inhibition by coupling cell-surface adhe- apoptotic proteins that increase the permeability of mito-
sion molecules such as E-cadherin to transmembrane recep- chondrial membrane that allows cytochrome c to leak into
tor tyrosine kinase such as EGFR. This adhesion strengthening the cytoplasm where it activates caspase-9 through APAF-1
also limits the emission of mitogenic signals [21, 22]. binding [25].
One of the complex underpinnings enhancing the tumor Abrogation of the apoptotic program in cancer cells is pos-
suppressor pathway in tumors (e.g., colon, pancreas) is dys- sible by two central mechanisms and is associated with high-
regulation of the WNT signaling pathway. WNT pathway grade tumors and developing resistance to therapies. The tumor
activation facilitates β-catenin activity by blocking the for- cells counterbalance pro-apoptotic signaling by upregulating
mation of a “destruction complex” composed of APC and anti-apoptotic proteins such as Bcl-2 family of proteins [26, 27].
other proteins. β-catenin, in turn, regulates cell proliferation For example, in follicular lymphoma, the molecular underpin-
by modulating transcription factors such as MYC and cyclin ning is the t(14;18)(q32;q21) translocation leading to constitu-
D1. β-Catenin activity is kept in check by this destruction tive activation of Bcl2 by transcriptionally active IGH promotor
complex composed of APC and other proteins. Mutations in region. The IGH-Bcl2 fusion help escape apoptosis. Another
the APC gene occur in 70–80% of early colonic adenomas, mechanism of escape from apoptosis is achieved through the
and carcinomas lead to unchecked activity of β-catenin. loss of TP53 function. Loss of TP53 prevents upregulation of a
BH3-only protein called PUMA (BH3 proteins neutralize
actions of Bcl2 related anti-apoptotic proteins). The evasion of
1.3.2  ownregulation of Cell Proliferation
D apoptotic circuitry is complex in different tumors and different
Inhibitor—Transforming Growth Factor mechanisms are derailed to evade apoptosis.
β (TGF-β) Pathway

In many cancers such as colon, stomach, and endometrium, 1.4.1 Autophagy

there are loss-of-function mutations in TGF-β receptors lead-
ing to loss of inhibitory control on cellular proliferation. In a state of severe nutrient deficiency, cells activate signal-
There is downregulation of SMAD genes in pancreatic ing circuits that either arrest cell growth or even cannibalize
tumors, downstream signal transducers in TGF-β, thus cor- their own organelles, proteins, and membranes for energy
rupting the TGF-β pathway. In addition to promoting tumor production. In a rapidly growing tumor, the cells may become
growth, depending on the cellular context, altered TGF-β sig- dormant activating autophagy due to scarcity of nutrients. A
naling is found to activate epithelial-mesenchymal transition similar phenomenon may be used by cancer cells under ther-
(EMT) that makes the cancer cell more aggressive [23, 24]. apeutic pressure, thus making tumor resistant to therapies
and leading to therapeutic failures. The role of autophagy in
cancer is still highly debatable and is based on the state of the
1.4 Resisting Cell Death tumor cell; it can be viewed as tumor protective in certain
circumstances and adverse in others [28].
Apoptosis, programmed cell death dismantling cells into
components, is activated in various cellular conditions
orchestrated by many signaling circuitry, including TP53 1.5  nabling Replicative Immortality:
E
and RB1 checkpoints. The signaling circuitry governing this The Stem-Cell-Like Properties
program is triggered by physiologic stresses such as exag-
gerated oncogenic signaling, DNA damage due to chemo- Over the last two decades, literature on cancer stem cells has
therapy, radiation, or hyperproliferative state of the cancer significantly evolved through intense cancer research, albeit
cell. The activation of the apoptotic program is thus achieved still under debate. The first descriptions of cancer stem cells
8 D. A. Chitale

were published on acute myeloid leukemia by Bonnet and oncogenes and tumor suppressor genes. The telomerase
Dick in 1997 [29]. The basic property of a cancer as having activity is also thought to be upregulated in cancer stem cells.
replicative immortality is conceptualized through its ability Tissue stem cells and germ cells retain the capability to
to self-replicate and differentiate through cancer stem-cell, a divide indefinitely through numerous cell divisions. By
cell that is thought to be like a normal stem-cell counterpart. expressing telomerase, they resist mitotic crises and the
The debate continues whether the “stemness” arises through accumulation of mutations. As stem cells divide, during
the transformation of a normal stem cell or through the abil- embryogenesis or when a cell is under stress, either both
ity of the cancer cells to transform into stem-like state (symmetric) or one (asymmetric) remains as daughter cells
through the acquisition of mutation through the selection [34]. Cancer stem cells are thought to retain their limitless
process. In either pathway, the molecular underpinnings are proliferative capacity and immortality, although debate con-
thought to be by evading senescence and mitotic crisis and tinues as to the mechanism and number of cancer stem cells
self-renewal ability. in different types of cancers. In summary, a small proportion
Normal cells typically divide 50 to 70 times (“Hayflick of cancer cells retain or acquire stem-like properties that
limit”) [30, 31], after which the cells become senescent, i.e., inactivate senescence signals, reactivate telomerase, and
lose their ability to divide again. But cancer cells gain ability enable replicative immortality.
to evade senescence in part by abnormalities in RB and TP53
proteins. Maintaining RB protein in hypophosphorylated
(active) state and downregulation of TP53 leads to disruption 1.6 Angiogenesis
of G1/S cell cycle checkpoint, thus keeping the cell in a pro-
liferative mode. For tumors to grow and sustain proliferative advantage, an
Another mechanism by which cancer cells achieve senes- adequate supply of oxygen, nutrients, and metabolic waste
cence evasion is through the maintenance of telomerase, management needs an adequate blood supply. Cancers stim-
which restores/“replenishes” the telomere sequences. ulate neoangiogenesis to sustain its growth. The angiogene-
Normally replicative capacity of a cell is lost after each cell sis is observed under the microscope in some tumors
division due to the shortening of the telomere, a repetitive extensive vascular network, for example, primary and meta-
nucleotide sequence at each end of the chromosome protect- static renal cell carcinoma [35], while in some other tumors,
ing the chromosomal deterioration and truncation. few blood vessels such as scirrhous invasive ductal carci-
Eventually, the cell loses its replicative capacity completely noma of the breast (Fig. 1.1). The current understanding of
and becomes senescent. But in many of tumors, human angiogenesis is controlled by a balance between promoters
telomerase reverse transcriptase (TERT) upregulation leads and inhibitors of vascular proliferation [36, 37]. In the physi-
to overexpression of telomerase enzyme, thus defying ologic states, such as female endometrial cycling and wound
mitotic crisis and senescence [32, 33]. The gain of telomer- healing, “angiogenic switch” is triggered transiently.
ase function may be induced due to cellular mitotic crisis However, in cancers as the tumor grows in size, the quiescent
from oncogenic program activation in cancer cells, thus stage is terminated, and vascular proliferation is initiated by
maintaining the replicative potential and the damage to increased local production of angiogenic factors and/or loss

Fig. 1.1 Invasive breast carcinoma with dense desmoplasia and clear cell carcinoma with rich capillary network
1 Hallmarks of Cancer: Molecular Underpinnings 9

of angiogenic inhibitors [17]. The angiogenic factors are cells become dis-cohesive and reduce interactions
locally produced by tumor cells, inflammatory cells, or stro- (“Loosening up” of tumor cells). This is followed by degra-
mal cells. Some of the inflammatory cells and, in particular, dation of extracellular matrix (ECM), attachment to “remod-
endothelial progenitor cells are thought to be bone marrow-­ eled” ECM components, migration, and invasion of tumor
derived in some cases [38]. The molecular underpinnings of cells. In the first step, the well-known alteration in cancer
angiogenesis in tumor cells include upregulation of many cells is loss of adhesion molecule E-cadherin encoded by the
genes, the key genes include hypoxia-inducible transcription CDH1 gene through a process called epithelial-­mesenchymal
factor 1α (HIF1α), vascular endothelial growth factor transition (EMT). EMT, thought to be an integral part of
(VEGF), as angiogenesis inducers, and thrombospondin-1 metastasis, especially in breast and prostate cancers, is con-
(TSP-1) as an inhibitor. The pro-angiogenic factors include: trolled by a combination of Snail, Slug, Twist, and Zeb1/2
Hypoxia led stabilization of HIF1α that upregulates VEGF transcription factors, favoring promigratory properties. The
gene expression. VEGF is also upregulated by oncogene sig- next step is the degradation of the basement membrane and
naling. This stimulates endothelial growth towards the tumor. interstitial connective tissues. Through autocrine or para-
Normally, TP53 upregulates TSP-1 expression that inhibits crine secretions, proteolytic enzymes such as matrix metal-
angiogenesis by repressing VEGF. TP53 mutations lead to loproteinase (MMP), cathepsin D, and urokinase plasminogen
p53 protein loss that permits angiogenesis. activator help cancer cells invade. The final step of invasion
Earlier it was thought that angiogenesis sets in fully when is the migration of cancer cells by locomotion through base-
the tumors proliferate. But recent literature has shown that ment membranes and areas of proteolysis by the contractile
the histologic changes of angiogenesis, such as microvessel actin cytoskeleton. Finally, through vascular dissemination,
density, are noted in the pre-malignant stages of neoplasms homing, and colonization, a few cancer cells are successful
in various organs, including squamous cell lesions in the oral in distant metastasis. Some tumors metastasize to specific
mucosa, skin, uterine cervix, vulva, and anal canal [39, 40]. sites, thus showing organ tropism. The underlying molecular
The fundamental properties of pre-malignant conditions for mechanism is thought to be through the expression of adhe-
prevention were first described in 1976 by Sporn [41]. Using sion or chemokine receptors, whose ligands are expressed by
the knowledge of the molecular basis of hallmarks of cancer, endothelial cells at the metastatic site [46].
Ryan and Faupel-Badger correlated these concepts with The details of molecular underpinnings of metastasis are
these properties of pre-malignant conditions [42], further described in a separate chapter in this section, and only high-­
supporting the acquisition of many of the hallmarks of can- level key discussion points are described above.
cers in the early stages of development.

1.8  merging Hallmarks and Enabling

E
1.7 Invasion and Metastasis Characteristics

One of the most aggressive properties of cancer cells is the With mounting literature support, two additional hallmarks
invasion and destruction of local tissues and metastasis. The of cancer, including Deregulating Cellular Energetics and
original concept of a pattern of cancer metastasis was Avoiding Immune Destruction and enabling characteristics
described by an English surgeon Stephen Paget in 1889 pub- Genomic Instability and Mutation and Tumor-promoting
lished as “seed and soil” hypothesis [43] [44]. Paget intro- Inflammation were introduced [2]. Each of these are
duced the concept of the spread of tumor cells through described below.
interactions between the cancer cells (seed) and the host
organ (soil). Even after a century, this concept is upheld by
extensive research with “seed” now identified as the cancer 1.9  rowth Promoting Altered Cellular
G
stem cell [45]. It has been shown that this hallmark of cancer Metabolism (Deregulating Cellular
occurs through a complex multistep progression called the Energetics)
invasion-metastasis cascade.
The major steps in this process include invasion of the To support the autonomous growth of cancer cells and get
extracellular matrix (ECM), angiolymphatic dissemination, oxygen and nutrients through neoangiogenesis, cancer cells
extravasation and tissue homing, and colonization. From pri- reprogram the energy metabolism, which is recently
mary site to metastatic site, the journey of cancer cells described as a key hallmark of cancer [2]. Cancer metabo-
includes breaching the supporting basement membrane, nav- lism research in cancer biology predates the discovery of
igating the interstitial connective tissues, and penetrating the oncogenes and tumor suppressor genes [47]. There is over-
vascular wall to gain access to vessels. During homing and whelming research support that cancer cells reprogram
colonization, this process is reversed. In the first step, the metabolism to improve cellular sustainability and resilience
10 D. A. Chitale

to gain selective advantage. The cancer cells change their Macfarlane Burnet then consolidated this hypothesis and
metabolic pathway to the glycolytic pathway as the preferred termed it as “immune surveillance.” This phenomenon is
source of energy where glucose is converted to lactose to thought to serve as a normal function of the immune system
generate ATP, even under aerobic conditions. This was first by constantly scanning the body for tumor cells and elimi-
described by Otto Warburg in 1930 and is now known as the nating them [17, 50]. This concept was tested by several
Warburg effect or aerobic glycolysis [48]. Although ATP experiments and observations and was initially abundant due
generation is lower than the mitochondrial oxidative meta- to a lack of sufficient evidence. Burnet suggested that the
bolic pathway through the tricarboxylic acid (TCA) cycle, cancer cells’ expression of tumor-specific neo-antigens
this pathway is preferred as its metabolic intermediates are induced an immunological response, thus eliminating such
used by the rapidly proliferating cells to synthesize cellular cells before clinical manifestations. We now know that cer-
components to fulfill the biosynthetic requirement. tain tumors are heavily infiltrated by lymphocytes and other
Biosynthesis of macromolecules or anabolic pathways is an immune cells. Such tumors often harbor numerous driver as
integral part of cancer metabolism to support growth. well as passenger mutations that have been discovered
The molecular underpinnings of the altered energy metab- through high-throughput whole-genome sequencing. This
olism include upregulation of oncogenes in receptor tyrosine “hyper mutator” genotype results in the production of neo-­
kinase/PI3K-AKT-mTOR signaling that enhances the activ- antigens that induces a robust host CD8+ cytotoxic T-cell
ity of glucose transporters and multiple glycolytic enzymes cell response, as reported in many malignant melanomas,
promoting lipid, protein, and nucleic acid biosynthesis. many colorectal cancers, lung cancers, among others [51]
Overexpression of transcription factor MYC leads to upregu- (Fig. 1.2). Direct demonstration of tumor-specific T cells and
lation of genes involved in anabolic metabolism through gly- antibodies in patients with cancers was noted in a few. Some
colytic enzymes and glutaminase. On the tumor suppressor experiments challenged the hypothesis of immune surveil-
side, STK11, an antagonist of oxidative glycolysis, is down- lance, which showed that not all the immunodeficient or
regulated to promote cell growth. immunosuppressed patients develop cancers. However, in
As described above, autophagy is activated in nutrient-­ immunocompromised patients (acquired or hereditary), viral
deficient conditions where cellular organelles and other com- associated tumors are prevalent (HPV, EBV, HHV8), and so
ponents are digested and reused. Depending on the scenario, are nonviral-associated solid tumors [52].
the tumor inactivates autophagy signaling to promote growth. Dunn and Schreiber developed a concept of “cancer
On the other side, tumor cells may use autophagy to go into immunoediting,” which includes three stages of “sculpting
a “hibernation” state due to external pressures, such as che- of tumor” by the immune system. The immune surveillance
motherapy, to become dormant and prevent cell death. hypothesis was proposed to be a general part of the cancer
Recent sequencing data in certain tumors such as glio- editing process and is considered the first stage called the
mas, acute myeloid leukemia, intrahepatic cholangiocarci- elimination of tumor cells. The next step is called the equi-
noma, and central chondrosarcoma has shown somatic librium phase, which appears to be the longest that leads to
mutations in Krebs cycle (TCA) genes, viz. isocitrate dehy- the formation of new variants in the cancer cells with differ-
drogenase genes-1 and 2 (IDH1, IDH2). Point mutations in ent mutations that resist the immune pressure. Different
IDH1 or IDH2 genes lead to mutant IDH protein, which molecular mechanisms are triggered when cancer cells go
loses its normal function and converts α-ketoglutarate to into quiescence or the slow-cycling phase described above.
2-hydrozyglutarate. This protein is considered an oncome- Cancer cells likely activate senescence gene signaling to
tabolite which acts as an inhibitor of epigenetic regulatory avoid cell death induced by immune cells. From this stage,
enzymes leading to cancer gene signaling pathway activa- some clines emerge and escape this immune surveillance
tion. Similarly, other enzymes from the TCA cycle that are and evade immune destruction. The molecular underpin-
mutated in other cancers include succinate dehydrogenase nings include selective outgrowth of antigen-negative vari-
(SDH), fumarate hydratase (FH), and L-2-hydroxyglutarate ants, loss or reduced expression of major histocompatibility
dehydrogenase. These oncometabolites present a potential (MHC) class I molecules escaping an attack from CD8+
therapeutic target [49]. cytotoxic T cells (CTL). Another mechanism that cancer
cells use to evade immune destruction is by inhibiting T cell
activation. The tumor cells upregulate inhibitory receptor
1.10  vasion of Host Defense
E CTLA-4 on tumor-specific T cells preventing sensitization
(Avoiding Immune Destruction) and extended escape from the tumor-specific T cells. Another
mechanism includes overexpression of Programmed death
Paul Ehrlich first hypothesized the role of host defense by the ligand (PD-L)-1 and PD-L2 cell surface proteins that inacti-
immune system in eliminating the tumor cells by recogniz- vates CD8+ CTL by docking with PD-1 receptor. Thus,
ing them as “foreign.” Lewis Thomas introduced and tumor cells avoid detection for elimination and escape
1 Hallmarks of Cancer: Molecular Underpinnings 11

a b

Fig. 1.2 Invasive colonic adenocarcinoma with dense Crohn’s-like inflammatory response, (a) stars marking peritumoral dense lymphocytic
infiltrate (100×), (b) depicting intra-epithelial lymphocytes (200×)
Ki
Pi

dn
lo

G
cy

Ly L Br s

Ki ney Ov
ey
M

lio

Lu un add
Ki

Lu Eso ctu
H
tic

dn P a
N
ed

m ow ea
m
C

ea Sto rus

ng g
eu

ng ph m
G as L

ph G st
a

ey ap ry
ul

hr
As

L
C
lio to

d m
ro
lo

Sq Ade r
om A

om ra

M mo
Sm ag
tro

C illa

ol rv
Pa tom

an ac
bl

bl CL
bl m

M -ce

el u
Pr ma

le ry

or ix

ua o
Bl ell
op L

Th be

a de
as L

as a
cy

ye ll

d h

an s
nc a

al us
U er
ar

C
os

e
B

N
to

to

ho

te

lC
yr

lo

Li

e
re

om
A

C
ta

ec
m

m
M

v
oi
L

n
a

el

e
te
a

a
k
1,000
l
(Number Mutations Per Megabase)

100
Somatic Mutation Prevalence

10

1.0

0.1

0.01

0.001

Fig. 1.3 Prevalence of somatic mutations per megabase across human cancer types in ascending order from left to right [53]

immune surveillance. The anti-PD-L1 and anti-CTLA-4 1.11 Genomic Instability

therapies block this inhibition in some solid tumors that
overexpress PD-L1 protein (melanoma, lung cancers) and Genomic instability as a distinct enabling characteristics was
renders them vulnerable to the CD8+ CTL. These therapies added to the original six hallmarks in 2011 [2]. The success
(antibodies) that overcome immune evasion by the tumors of all the hallmarks of cancer depends profoundly on the
are called immune checkpoint inhibitors (Fig. 1.3). underlying genomic alterations, directly or indirectly.
Detailed molecular underpinnings of this process is Through multistep gain of mutations, the cancer cell clones
described in the immune microenvironment chapter of this get a survival advantage to become more autonomous. This
section. phenomenon is nurtured and amplified by inactivation of
12 D. A. Chitale

genome maintenance system to detect and resolve these as modifying the tumor microenvironment described in the
defects, either due to somatic alterations (e.g., epigenetic latter part of this section. Few key characteristics include the
alterations in DNA mutation and histone modifications) or release of factors that promote proliferation, removal of
inherited mutations (e.g., TP53 (Li-Fraumeni syndrome) or growth suppressors, enhanced resistance to cell death, induc-
Lynch (HNPCC) syndrome due to mutations in the mismatch ing angiogenesis, activating invasion and metastasis, and
repair (MMR) genes). Broad categories that are part of this evading immune destruction [17].
genome maintenance system include genes that produce pro-
teins involved in detecting DNA damage and activating the
repair machinery, proteins directly repairing damaged DNA, 1.13 Conclusion
and proteins inactivating or intercepting mutagenic mole-
cules before they have damaged the DNA [54, 55]. Cancer development is a complex biological process that is
Genomic instability is acquired due to the variable loss of underpinned by the multistep acquisition of molecular
these functions. TP53 gene pauses the cell function when genetic aberrations imparting the above-described eight hall-
there is DNA damage to allow DNA repair to occur. TP53 marks of cancers and two enabling traits. These are gained
gene mutations are most frequently observed in many tumors either through small changes such as single nucleotide poly-
described in the sections above, leading to further accumula- morphisms, insertions/deletions, and amplification or
tion of mutations. DNA MMR deficiency is another mecha- through large structural genetic changes such as transloca-
nism that leads to the accumulation of numerous mutations tions, aneuploidy, and genomewide methylation, to name a
(hypermutator phenotype) producing many tumors in many few. These changes confer a survival advantage and ability to
organs (e.g., proximal colonic adenocarcinomas, endome- invade, recur, and metastasize to distant sites. Several molec-
trial carcinomas). Loss-of-function mutations in genes ularly targeted therapies are currently available that are the
encoding nucleoside excision repair system lead to inability standard of practice, while some molecular markers part of
to repair UV radiation induced cross-linking of pyrimidine AJCC cancer staging. Understanding the molecular basis of
residues thwarting replication, as seen in xeroderma cancer will further help healthcare professionals, researchers
pigmentosum. Defects in the homologous recombination
­ design assays for diagnostic, prognostic, and predictive pur-
DNA repair system constitute a group of disorders that poses and the pharmaceutical industry to discover newer tar-
includes Bloom syndrome, ataxia-telangiectasia, and geted therapeutics to pursue personalized precision
Fanconi anemia. BRCA1 and BRCA2 involved in DNA medicine.
repair are mutated in familial breast and ovarian cancers,
prostate cancers to name a few. DNA polymerase mutations
lead to loss of proofreading capabilities that lead to genomic References
instability, as noted in some colonic and endometrial carci-
nomas. Finally, many lymphoid neoplasms show a high level 1. Hanahan D, Weinberg RA. The hallmarks of can-
cer. Cell. 2000;100(1):57–70. https://doi.org/10.1016/
of genomic instability due to errors occurring during immu- s0092-­8674(00)81683-­9.
noglobulin and T cell receptor gene rearrangements (RAG1, 2. Hanahan D, Weinberg RA. Hallmarks of cancer: the next gen-
RAG2, activation-induced cytosine deaminase). eration. Cell. 2011;144(5):646–74. https://doi.org/10.1016/j.
cell.2011.02.013.
3. Westermark B. Platelet-derived growth factor in glioblastoma-­
driver or biomarker? Ups J Med Sci. 2014;119(4):298–305. https://
1.12 Cancer-Enabling Inflammation doi.org/10.3109/03009734.2014.970304.
4. Xie Y, Su N, Yang J, Tan Q, Huang S, Jin M, et al. FGF/FGFR
In pathology literature, observations of dense inflammatory signaling in health and disease. Signal Transduct Target Ther.
2020;5(1):181. https://doi.org/10.1038/s41392-­020-­00222-­7.
infiltrate in certain solid tumors have been described for many 5. Hynes NE, MacDonald G. ErbB receptors and signaling pathways
decades. The density of the infiltrate, from dense to very few in cancer. Curr Opin Cell Biol. 2009;21(2):177–84. https://doi.
noticeable cells, as well as the composition of cells—acute org/10.1016/j.ceb.2008.12.010.
vs. chronic inflammatory cells (innate and adaptive), vary in 6. Lemmon MA, Schlessinger J. Cell signaling by receptor tyrosine
kinases. Cell. 2010;141(7):1117–34. https://doi.org/10.1016/j.
different morphologic subtypes. Such immune response was cell.2010.06.011.
thought to be a favorable phenomenon of the immune system 7. Perona R. Cell signalling: growth factors and tyrosine kinase recep-
attempting to eliminate tumor cells, as described above, shap- tors. Clin Transl Oncol. 2006;8(2):77–82. https://doi.org/10.1007/
ing the newer clones of tumor cells that eventually escape s12094-­006-­0162-­1.
8. Witsch E, Sela M, Yarden Y. Roles for growth factors in cancer pro-
destruction. But there is enough evidence that the inflamma- gression. Physiology (Bethesda). 2010;25(2):85–101. https://doi.
tory response may have a paradoxical effect of enabling can- org/10.1152/physiol.00045.2009.
cer progression [2]. Inflammation enables several hallmarks 9. Consortium, A. P. G. AACR project GENIE: powering preci-
of cancer by having a direct effect on the tumor cells as well sion medicine through an international consortium. Cancer
1 Hallmarks of Cancer: Molecular Underpinnings 13

Discov. 2017;7(8):818–31. https://doi.org/10.1158/2159-­8290. 31. Hayflick L, Moorhead PS. The serial cultivation of human dip-
CD-­17-­0151. loid cell strains. Exp Cell Res. 1961;25:585–621. https://doi.
10. Davies MA, Samuels Y. Analysis of the genome to personalize org/10.1016/0014-­4827(61)90192-­6.
therapy for melanoma. Oncogene. 2010;29(41):5545–55. https:// 32. Heidenreich B, Rachakonda PS, Hemminki K, Kumar R. TERT
doi.org/10.1038/onc.2010.323. promoter mutations in cancer development. Curr Opin Genet Dev.
11. Guo YJ, Pan WW, Liu SB, Shen ZF, Xu Y, Hu LL. ERK/ 2014;24:30–7. https://doi.org/10.1016/j.gde.2013.11.005.
MAPK signalling pathway and tumorigenesis. Exp Ther Med. 33. Maciejowski J, de Lange T. Telomeres in cancer: tumour suppression
2020;19(3):1997–2007. https://doi.org/10.3892/etm.2020.8454. and genome instability. Nat Rev Mol Cell Biol. 2017;18(3):175–86.
12. Osaki M, Oshimura M, Ito H. PI3K-Akt pathway: its functions and https://doi.org/10.1038/nrm.2016.171.
alterations in human cancer. Apoptosis. 2004;9(6):667–76. https:// 34. Batlle E, Clevers H. Cancer stem cells revisited. Nat Med.
doi.org/10.1023/B:APPT.0000045801.15585.dd. 2017;23(10):1124–34. https://doi.org/10.1038/nm.4409.
13. Yuan TL, Cantley LC. PI3K pathway alterations in cancer: varia- 35. Zee YK, O’Connor JP, Parker GJ, Jackson A, Clamp AR, Taylor
tions on a theme. Oncogene. 2008;27(41):5497–510. https://doi. MB, et al. Imaging angiogenesis of genitourinary tumors. Nat Rev
org/10.1038/onc.2008.245. Urol. 2010;7(2):69–82. https://doi.org/10.1038/nrurol.2009.262.
14. Magrath I. Epidemiology: clues to the pathogenesis of Burkitt 36. Baeriswyl V, Christofori G. The angiogenic switch in carcino-
lymphoma. Br J Haematol. 2012;156(6):744–56. https://doi. genesis. Semin Cancer Biol. 2009;19(5):329–37. https://doi.
org/10.1111/j.1365-­2141.2011.09013.x. org/10.1016/j.semcancer.2009.05.003.
15. Patterson KI, Brummer T, O'Brien PM, Daly RJ. Dual-specificity 37. Bergers G, Benjamin LE. Tumorigenesis and the angiogenic
phosphatases: critical regulators with diverse cellular tar- switch. Nat Rev Cancer. 2003;3(6):401–10. https://doi.org/10.1038/
gets. Biochem J. 2009;418(3):475–89. https://doi.org/10.1042/ nrc1093.
bj20082234. 38. Patenaude A, Parker J, Karsan A. Involvement of endothe-
16. Tan MH, Mester JL, Ngeow J, Rybicki LA, Orloff MS, Eng lial progenitor cells in tumor vascularization. Microvasc Res.
C. Lifetime cancer risks in individuals with germline PTEN 2010;79(3):217–23. https://doi.org/10.1016/j.mvr.2010.01.007.
mutations. Clin Cancer Res. 2012;18(2):400–7. https://doi. 39. Menakuru SR, Brown NJ, Staton CA, Reed MW. Angiogenesis
org/10.1158/1078-­0432.CCR-­11-­2283. in pre-malignant conditions. Br J Cancer. 2008;99(12):1961–6.
17. Kumar V, Abbas AK, Aster JC, Turner JR, Robbins SL, Cotran https://doi.org/10.1038/sj.bjc.6604733.
RS. Robbins & Cotran pathologic basis of disease; 2021. 40. Raica M, Cimpean AM, Ribatti D. Angiogenesis in pre-­malignant
18. Collado M, Serrano M. Senescence in tumours: evidence from conditions. Eur J Cancer. 2009;45(11):1924–34. https://doi.
mice and humans. Nat Rev Cancer. 2010;10(1):51–7. https://doi. org/10.1016/j.ejca.2009.04.007.
org/10.1038/nrc2772. 41. Sporn MB. Approaches to prevention of epithelial cancer during
19. Knudson AG. Mutation and cancer: statistical study of retino- the preneoplastic period. Cancer Res. 1976;36(7 PT 2):2699–702.
blastoma. Proc Natl Acad Sci. 1971;68(4):820–3. https://doi. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/1277177.
org/10.1073/pnas.68.4.820. 42. Ryan BM, Faupel-Badger JM. The hallmarks of premalig-
20. Robinson DR, Wu YM, Lonigro RJ, Vats P, Cobain E, Everett J, nant conditions: a molecular basis for cancer prevention.
et al. Integrative clinical genomics of metastatic cancer. Nature. Semin Oncol. 2016;43(1):22–35. https://doi.org/10.1053/j.
2017;548(7667):297–303. https://doi.org/10.1038/nature23306. seminoncol.2015.09.007.
21. Curto M, Cole BK, Lallemand D, Liu CH, McClatchey AI. Contact-­ 43. Talmadge JE, Fidler IJ. AACR centennial series: the biol-
dependent inhibition of EGFR signaling by Nf2/Merlin. J Cell Biol. ogy of cancer metastasis: historical perspective. Cancer Res.
2007;177(5):893–903. https://doi.org/10.1083/jcb.200703010. 2010;70(14):5649–69. https://doi.org/10.1158/0008-­5472.
22. Okada T, Lopez-Lago M, Giancotti FG. Merlin/NF-2 mediates con- CAN-­10-­1040.
tact inhibition of growth by suppressing recruitment of Rac to the 44. Akhtar M, Haider A, Rashid S, Al-Nabet A. Paget’s “seed and
plasma membrane. J Cell Biol. 2005;171(2):361–71. https://doi. soil” theory of cancer metastasis: an idea whose time has come.
org/10.1083/jcb.200503165. Adv Anat Pathol. 2019;26(1):69–74. https://doi.org/10.1097/
23. Ikushima H, Miyazono K. TGFbeta signalling: a complex web in PAP.0000000000000219.
cancer progression. Nat Rev Cancer. 2010;10(6):415–24. https:// 45. Langley RR, Fidler IJ. Tumor cell-organ microenvironment inter-
doi.org/10.1038/nrc2853. actions in the pathogenesis of cancer metastasis. Endocr Rev.
24. Massague J. TGFbeta in cancer. Cell. 2008;134(2):215–30. https:// 2007;28(3):297–321. https://doi.org/10.1210/er.2006-­0027.
doi.org/10.1016/j.cell.2008.07.001. 46. Riggi N, Aguet M, Stamenkovic I. Cancer metastasis: a reap-
25. Elmore S. Apoptosis: a review of programmed cell praisal of its underlying mechanisms and their relevance to treat-
death. Toxicol Pathol. 2007;35(4):495–516. https://doi. ment. Annu Rev Pathol. 2018;13:117–40. https://doi.org/10.1146/
org/10.1080/01926230701320337. annurev-­pathol-­020117-­044127.
26. Adams JM, Cory S. The Bcl-2 apoptotic switch in cancer devel- 47. DeBerardinis RJ, Chandel NS. Fundamentals of cancer metabo-
opment and therapy. Oncogene. 2007;26(9):1324–37. https://doi. lism. Sci Adv. 2016;2(5):e1600200. https://doi.org/10.1126/
org/10.1038/sj.onc.1210220. sciadv.1600200.
27. Lowe SW, Cepero E, Evan G. Intrinsic tumour suppression. Nature. 48. Koppenol WH, Bounds PL, Dang CV. Otto Warburg’s contribu-
2004;432(7015):307–15. https://doi.org/10.1038/nature03098. tions to current concepts of cancer metabolism. Nat Rev Cancer.
28. Saha S, Panigrahi DP, Patil S, Bhutia SK. Autophagy in health 2011;11(5):325–37. https://doi.org/10.1038/nrc3038.
and disease: a comprehensive review. Biomed Pharmacother. 49. Collins RRJ, Patel K, Putnam WC, Kapur P, Rakheja
2018;104:485–95. https://doi.org/10.1016/j.biopha.2018.05.007. D. Oncometabolites: a new paradigm for oncology, metabolism,
29. Bonnet D, Dick JE. Human acute myeloid leukemia is organized as and the clinical laboratory. Clin Chem. 2017;63(12):1812–20.
a hierarchy that originates from a primitive hematopoietic cell. Nat https://doi.org/10.1373/clinchem.2016.267666.
Med. 1997;3(7):730–7. https://doi.org/10.1038/nm0797-­730. 50. Ribatti D. The concept of immune surveillance against tumors.
30. Burnet Sir M. Intrinsic mutagenesis: a genetic approach to ageing. The first theories. Oncotarget. 2017;8(4):7175–80. https://doi.
Dordrecht: Springer Netherlands; 2012. org/10.18632/oncotarget.12739.
14 D. A. Chitale

51. Schumacher TN, Schreiber RD. Neoantigens in cancer immuno- cancer. Nature. 2013;500(7463):415–21. https://doi.org/10.1038/
therapy. Science. 2015;348(6230):69–74. https://doi.org/10.1126/ nature12477.
science.aaa4971. 54. Ciccia A, Elledge SJ. The DNA damage response: making it safe
52. Dunn GP, Bruce AT, Ikeda H, Old LJ, Schreiber RD. Cancer to play with knives. Mol Cell. 2010;40(2):179–204. https://doi.
immunoediting: from immunosurveillance to tumor escape. Nat org/10.1016/j.molcel.2010.09.019.
Immunol. 2002;3(11):991–8. https://doi.org/10.1038/ni1102-­991. 55. Negrini S, Gorgoulis VG, Halazonetis TD. Genomic instabil-
53. Alexandrov LB, Nik-Zainal S, Wedge DC, Aparicio SA, Behjati ity—an evolving hallmark of cancer. Nat Rev Mol Cell Biol.
S, Biankin AV, et al. Signatures of mutational processes in human 2010;11(3):220–8. https://doi.org/10.1038/nrm2858.
 Unifying Concept of Genomic Changes:
The Mutational Landscape of Cancers 2
Laura Favazza DO

Abstract
• Learn how DNA hypomethylation and DNA hyper-
The genomic landscape of cancer is complex, can be methylation affects genetic transcription.
described in multiple different ways, and is still evolving. • Recognize the effects of miRNA on transcription in
As technology evolves and more tumors are sequenced, we tumors.
discover more each day about patterns of DNA alterations,
large- and small-scale chromosomal and copy number
changes, epigenetic and methylome modifications, and sub-
sequent RNA alterations among different tumor types. In 2.1 Introduction
this chapter we explore the global molecular alterations in
different cancer types and how these alterations are ulti- Across cancer there are multiple ways that the human
mately related to one another. The molecular alterations dis- genome, transcriptome, and proteosome may be altered in
cussed include chromosomal changes, smaller-scale copy cancer. The genetic alterations in cancer may range from
number alterations, DNA mutations, epigenetic profiles, large chromosomal changes, which may be observed via
mRNA and miRNA variations, and finally we will briefly karyotype, to small single nucleotide DNA substitutions, that
discuss protein expression in a multitude of different tumor can only be seen via DNA sequencing. Regardless of the size
types. These findings may be used to help us understand of the change, these variations can lead to cell increased cell
tumorigenesis and tumor progression. They also can help us cycling which promote tumor growth and progression. In
further classify tumors based on molecular findings, rather this chapter we will explore the general global genomic
than the traditional, purely histologic findings, which may alterations seen across these areas and across different tumor
or may not provide valuable information regarding predic- types.
tive factors and prognostic information. This chapter goes
into detail with examples of several different types of
molecular alterations in multiple different tumor types, how 2.2 Chromosomal Aneuploidy
these may be used to further classify tumors, and how the
various alterations are associated to one another. While traditional chromosomal cytogenetics is less com-
monly performed in the era of next-generation sequencing,
this was one of the first genetic classifications of tumors.
Although we use more sensitive and specific techniques
Learning Objectives
presently, classifications and observations have been made
• Understand the types of chromosomal copy number
through this type of analysis. Chromosomal aneuploidy
changes that can occur in cancer.
refers to whole chromosome losses and gains, and in cancer
• Discuss the different types of mutational signatures
genetics chromosomal arm losses and gains are also included.
and their prevalence in different cancer types.
Chromosomal instability is the process that gives rise to
aneuploidy and promotes tumor evolution and intratumoral
heterogeneity [1]. Chromosomal instability, for example, is
L. Favazza DO (*)
the loss or gain of small genes or clusters of nearby genes
Pathology and Laboratory Medicine, Henry Ford Hospital, Detroit,
MI, USA and is also frequently associated with loss of p53 tumor sup-
e-mail: [email protected] pressive function [2]. Chromosomal instability is associated

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 15

S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_2
16 L. Favazza DO

with aneuploidy; however, some tumors can have aneuploidy lost. In this case the genetic material has moved to different
without chromosomal instability. Aneuploidy may be seen in locations with different promoters. Thus, using only tradi-
approximately 80% of solid tumor neoplasms [3]. tional DNA sequencing, this type of translocation can be dif-
Tumors can be segregated into those with many altera- ficult to detect. The second group is also simple but involves
tions (copy number alterations high), moderate number of a copy-and-paste structural variant as in a tandem duplica-
alterations, and those with few alterations [4]. Approximately tion [10]. Depending on the size of the variants listed above,
one-third of cancers show relatively sparse number of chro- detection can be rather difficult with targeted sequencing
mosomal alterations. These tumor types include thyroid car- panels. Among these structural variants, deletions have been
cinoma, acute myelocytic leukemia, prostate carcinomas, descried as the most common followed by tandem duplica-
and thymomas [4]. tions, and finally unbalanced translocations. Remarkably
Patterns of aneuploidy can also be tumor-type specific. reciprocal translocations and inversions are the least com-
This is seen in testicular germ cell tumors which characteris- mon events [10]. Multiple types of these structural variants
tically show chromosome 12p gains and also in IDH wild- are found within specific tumors and tumor types. For exam-
type gliomas that have chromosome 7 gain and chromosome ple, a study found that esophageal adenocarcinomas har-
10 loss. An exception to this is seen in endometrial cancers. bored several deletions and many complex rearrangements,
Endometrial serous-like carcinomas have high copy number while ovarian cancers harbored frequent tandem duplica-
alterations and can be separated from other endometrial car- tions and deletions with fewer complex clustered rearrange-
cinomas [2]. Tumors from separate but similar origins also ments [10].
show recurring arm losses and gains. Gains of 8q, 13q, and The last category involves complex cut-and-paste struc-
chromosome 20 are seen in gastrointestinal tumors including tural variants, which occur by multiple cycles of DNA dam-
stomach, pancreas, colorectal, and esophageal adenocarcino- age via end-to-end sister chromatid fusions, mitotic bridge
mas [2]. Squamous carcinomas, particularly esophageal, fusions, and unstable DNA leading to further breakage [10].
lung, and head and neck, characteristically show gain of 3q This group contains two types of chromosomal changes. The
and loss of 3p [4]. first, chromoplexy, is defined as many simultaneous double-­
Aneuploidy may be prognostic in multiple tumor types. stranded DNA breaks in several chromosomes which are
Hyperdiploid karyotype in multiple myeloma and acute lym- then reconnected improperly, leading to multiple balanced
phoblastic lymphoma has a favorable overall survival [5]. translocations. This phenomenon is frequently seen in pros-
While the hypodiploid subgroup is an adverse prognostic tate cancers [10]. The second mechanism is called
factor in acute lymphoblastic lymphoma [6]. Losses of chromothripsis. Chromothripsis is considered a single cata-
1p19q is disease defining for oligodendrogliomas, which has strophic event, where the chromosomes shatter and rear-
a favorable prognosis [7]. For many solid tumors, a high range, resulting in multiple copy number changes and few to
degree of aneuploidy is associated with adverse features, hundreds of chromosomal breakpoints [10].
including disease progression, decreased overall survival,
and decreased progression-free survival [1]. While extreme
aneuploidy can sometimes make tumors more sensitive to 2.3 Somatic DNA Variants
chemotherapy, as a general rule, increased aneuploidy is usu-
ally associated with increased resistance to chemotherapy Molecular variants can occur via multiple mechanisms
[8]. Interestingly when comparing the number of mutations including single-base substitutions, double-base substitu-
to aneuploidy, those that are hypermutated (high microsatel- tions, and complex insertions and/or deletions. Those that
lite instability or POL ε mutation) show low aneuploidy [9]. cause either activation of oncogenes or deleterious effects on
However, when hypermutated tumors are excluded, increased tumor suppressor genes are considered pathogenic or likely
number of mutation and increased aneuploidy are frequently pathogenic [11].
seen [2]. These patterns may be classified in many ways. One way
There are four major ways to break down smaller struc- is via mutational signatures, of which there are now greater
tural changes in cancer [10]. The first group is simple cut-­ than 30 signatures [12]. Most cancers display greater than
and-­paste structural variants. These can be further broken one mutational signature. These signatures are based on pat-
down into a deletion of chromosome material, a reciprocal terns of mutational changes. For example, the single base
inversion, a fold-back inversion, an unbalanced translocation pair substitutions are characterized by C > A, C > G, C > T,
of chromosomal material from one chromosome to another, T > A, T > C, and T > G. There are also characterizations for
and finally a reciprocal translocation [10]. Obviously, in the double-base substitutions and insertions/deletions.
deletion of chromosomal material there is a copy number Signatures specificity of mutation type and sequence con-
change, however, with a balanced translocation there is no text (signature 10), equal amount of all 96 mutations (signa-
such change in copy number since no material is gained or ture 3), C > T predominant (1A/B, 6, 7, 11, 15, 19), C > A
2 Unifying Concept of Genomic Changes: The Mutational Landscape of Cancers 17

predominant (signatures 4, 8, 18), T > C predominant (sig- The number of mutations identified in a tumor type and
natures 5, 12, 16, 21), and T > G predominant (signatures 9, across tumor types can vary greatly [15]. Typically, a tumor
17) [13]. will have a number of driver mutations leading to oncogene
Several of these signatures have been well characterized. activation as well as a number of loss of function mutations
They include 1A/B also known as the age signature. This is in tumor suppressor genes. In addition, multiple passenger
the endogenous mutational process in normal and neoplastic mutations may be seen. These passenger mutations may not
cells that occurs due to aging and involves deamination of have a big impact on protein function but can accumulate
5-methyl-cytosine [13]. Signatures 2 and 13 are also called due to an unstable genome, loss of DNA repair mechanisms,
APOBEC (apolipoprotein B mRNA editing enzyme, cata- and rapid cell cycling, among others. Lung cancers typically
lytic polypeptide-like) signature and are seen in multiple have a high tumor mutation burden due to the accumulation
tumor types and lead to cytosine to uracil substitutions from of genetic alterations from DNA damage caused by smoking.
dysfunctional cytidine deaminases [13, 14]. Mutational sig- This type of phenomenon is also frequently seen in cutane-
nature 3 includes those with BRCA1/2 mutations, whether ous malignant melanomas, in which the mutations accumu-
they are germline or somatic. BRCA1/2 mutations are most late over time due to damage from ultraviolet rays from
frequently encountered in breast cancers, but are also not unprotected sun exposure. Conversely, papillary thyroid car-
uncommonly seen in prostate and pancreatic cancers. This cinomas typically have a low tumor mutation burden, with
signature is involved in defects in homologous recombination-­ few passenger mutations along with mutations seen in driver
based DNA double-stranded break repair [13]. oncogenes. Interestingly, tumor mutation burden typically
Additional signatures are associated with defects in DNA increases in older patients when compared to younger
repair. Those include signature 6, associated with DNA mis- patients, particularly children, who usually have a very low
match repair deficiency, and signature 10, associated with mutational burden [15]. The significance of a high tumor
Pol E mutations, which result in tumors harboring numerous mutation burden may have a therapeutic impact and is dis-
mutations. DNA mismatch repair deficient tumors usually cussed in a later chapter. A mechanism in which a high tumor
have somatic hypermutation and may arise either via germ- mutation burden is commonly seen is in tumors with loss of
line or somatic mutations of MLH1, PMS2, MSH2, MSH6, function mutations in mismatch repair genes, as mentioned
and EPCAM. These are most commonly associated with above. These genes, which work in pairs, are responsible for
colorectal carcinoma as well as endometrial carcinomas. proofreading and fixing errors in DNA particularly in micro-
Inactivation of MLH1 can also occur sporadically via pro- satellite regions, or regions in the genome with short one to
moter hypermethylation, and are also associated with two base pair repeats. The ultra-mutator phenotype has been
colorectal carcinomas and endometrial carcinomas. described with inactivating hot spot mutations in the gene
Mutations in POL ε typically occur via hot spot mutations in POL ε, which encodes the catalytic part of the protein poly-
endometrial carcinomas, and lead to somatic ultra-mutator merase epsilon responsible for DNA repair. When this gene
phenotype. is inactivated, the polymerase makes frequent mistakes dur-
Signatures 4 and 7 are associated with exogenous carcin- ing replication, resulting in a multitude of mutations through-
ogens. Tobacco use is associated with signature 4, as is fre- out the genome.
quently seen in lung cancers and head and neck cancers, Noncoding driver mutations are not common but also
while exposure to ultraviolet light/radiation is associated have been reported [16]. Approximately one-quarter of
with signature 7, most commonly seen in cutaneous melano- tumors have at least one noncoding driver point mutation. Of
mas and squamous cell carcinomas. Signatures 8, 12, and 16 those tumors, about one-third of them affect the TERT pro-
show transcription strand bias associated with bulky exoge- moter. Other promoters and enhancers are less frequently
nous carcinogens. A specific signature associated with can- seen [16]. But the TERT promoter mutation is seen relatively
cer treatment has also been discovered, signature 11, frequently. And usually bears a poor prognosis. This finding
secondary to treatment with temozolomide. Signature 9 is has been described in liver, prostate, bladder, and thyroid
seen frequently in lymphomas and chronic lymphocytic leu- carcinomas [10, 17].
kemia/lymphoma as it is associated with immunoglobulin
gene hypermutation. This leads to errors in AID-induced
cytidine deamination in the error-prone polymerase, poly- 2.4 DNA Methylation
merase η [13]. Multiple signatures including 5, 8, 10, 12, and
16 show transcription-coupled nucleotide excision repair Epigenetic genes are involved in cancer by regulating and
defects which lead to strand bias. Signature 15 shows a large changing the expression of certain genes responsible for
number of substitutions and small insertions/deletions. cell proliferation and differentiation. These genes can be
Several signatures including 14, 15, and 21 are still not well separated into three groups. Epigenetic modulators send
characterized or understood yet [13]. signals to the epigenetic regulators which change the status
18 L. Favazza DO

of the genome. Epigenetic mediators alter their roles based shown to occur with loss of DNA methylation. Amplifications
on these signals [18]. Many of these genes are controlled by of MYC are commonly seen in tumor progression and more
methylation. DNA methylation occurs when a methyl aggressive tumors. When this occurs, there is loss of regula-
group is added to the carbon-5 position of the cytosine tion of cell pathways, alteration of chromatin organization,
base, known as 5-methylcytosine. It most frequently occurs and the overall genome [19].
at cytosine-­guanine dinucleotides, known as CpG, via DNA In addition to the loss of 5-methylcytosine, global loss of
methyltransferase enzymes [19]. These enzymes are known 5-hydroxymethylcytosine may be seen in many different types
as the DMNTs. Interestingly, 5-methylcytosine may also be of carcinomas [19]. For example, in breast, colon, prostate,
altered by ten-eleven-translocation enzymes (TET), an liver, pancreatic, and lung cancers and melanomas, there are
enzyme which leads to oxidation (DNA hydroxymethyl- significantly decreased levels of 5-­ hydroxymethylcytosine.
ation) and the product 5-hydroxymethylcytosine, which This decrease is frequently associated with mutations in the
can lead to widespread alterations in the DNA methylation TET genes leading to decreased function of TET enzymes. If
landscape [19]. no TET mutations can be found, particularly in leukemia,
Global DNA methylation patterns are commonly altered mutations leading to abnormal activity of isocitrate dehydro-
in cancer. DNA hypomethylation events occur at oncogenes genase 1 and 2 may be seen. This is a cofactor involved in TET
(causing the gene to turn on) and at centromeres. DNA enzymatic activity. When patients have IDH1/2 mutations,
hypermethylation typically leads to loss of regulation of TET activity is inhibited, leading to increased CpG island
enhancers and promoters and silencing of tumor suppressor hypermethylation [19].
genes. These global alterations in the promoters of tumor In one study, tumors arising from the same organ system
suppressor genes and oncogenes lead to tumorigenesis. typically cluster by cancer specific hypermethylation [4]. For
Studies have shown that multiple recurrent loci are hyper- example, squamous cell carcinoma from the head and neck,
methylated in different tumor types. About 90% of prostate esophagus, lung, and cervix shares similar methylation pat-
cancers have hypermethylation of GSTP, in 20% of lung can- terns. As such, adenocarcinomas origination from the esoph-
cers hypermethylation occurs at p16INK4a, and at BRCA1 in agus, stomach, and colorectal also shares a distinct set of
around 10% of breast and ovarian cancers [19]. methylation patterns [4].
As tumors progress, studies have shown there is an
increased frequency of p16INK4a and GSTP1 hypermethyl-
ation, thus suggesting that DNA hypermethylation changes 2.5 RNA and miRNA Alterations
may be predictive of disease progression [19]. CpG islands
are susceptible to DNA methyltransferase, and thus can Alterations of RNA transcriptions in cancer frequently occur
become hypermethylated. This phenotype, known as the due to somatic alterations in genomes. RNA alterations may
CpG Island Methylator Phenotype (CIMP), was first occur via splice site alterations, gene fusions, and overex-
described in glioblastoma and colon cancer. With this dis- pression [21]. On average, tumors have approximately 20%
covery, tumors may be classified by subtypes using the meth- more frequent alternate splicing events when compared to
ylation signature [19]. DNMT enzyme expression is also normal tissues [22]. Many of these changes occur as single
commonly altered in cancer. When this occurs, the DNA nucleotide variants occurring at critical areas for binding of
methylation patterns also change, which may lead to splice site machinery [23]. Frameshift mutations in mRNA
increased somatic mutations in DNA [19]. resulting from intron retention occur most commonly in
Global DNA hypomethylation occurs in the CpG-poor tumor suppressor genes, while in-frame exonic single nucle-
regions. It is seen in multiple tumor types and is predicted to otide variants, altering exons, are more frequently seen in
lead to genomic instability and increased chromosomal oncogenes [23].
aneuploidy [20]. When there is a decrease in the amount of Splice site creating mutations may lead to alternate pro-
DNA methyltransferase 1, mutation rates increase, aneu- tein production and have been associated with higher expres-
ploidy is more frequent, and tumor development and pro- sion of T cell associated genes, as well as mRNAs which
gression occurs. This suggests that there is a role of DNA encode immune-checkpoint blockade molecules (such as
hypomethylation possibly leading to chromosomal breakage PD-1 and PD-L1). These alterations may have a significant
[19]. Global hypomethylation is associated with abnormal therapeutic impact [23].
chromosome and chromatin organization, which results in Driver mutations have been discovered in over 119 splic-
structural variation and genome instability [19]. However, ing factor genes [23]. Hotspot mutations in the RNA-splicing
causation has not been definitively shown at this time. factors SF3B1, SRSF2, and U2AF1 have been described to
Abnormal expression of oncogenes, repeat elements, and occur more frequently in myeloid leukemias, both acute and
transposable elements, such as the MYC gene, has been chronic, as well as uveal and mucosal melanomas [23].
2 Unifying Concept of Genomic Changes: The Mutational Landscape of Cancers 19

SF3B1 is the most commonly mutated across all cancers. 2.6 Analysis of Proteins
The effect of this mutation on prognosis varies between
tumor types. For example, SF3B1 mutations in myeloid dys- Protein analysis is not used routinely in clinical practice.
plastic syndrome are associated with a favorable prognosis, However, in a study using reverse phase protein array, tumors
while SF3B1 mutations in CLL, which are typically sub- were segregated into ten groups [4]. The first group (P1)
clonal, are increasingly seen in patients with aggressive dis- included gliomas, both low and high grade [4]. The second
ease [23]. RNA-splicing mutations are relatively uncommon group (P2) included multiple tumors types and a multiple
in many solid tumors. However, they are seen much more array of tumors including diffuse large B cell lymphoma,
frequently in several types, including bladder cancer and multiple types of soft tissue sarcomas thymomas, and cuta-
lung adenocarcinoma [23]. neous melanomas [4]. Another set including P3 and P4
MicroRNA (miRNA) are short RNA sequences, about 22 accounted for tumors driven by estrogen, including luminal
nucleotides each, which regulate mRNA molecules either by A and B breast cancers, gynecological carcinomas, and some
degradation or stopping translation [24]. Each sequence of liver cancers with increased estrogen receptor expression,
miRNA may be responsible for controlling hundreds of androgen receptor expression, and increased IGFBP2 expres-
genes. Changes in miRNA expression play a critical role in sion [4]. Then, there was a pan-kidney tumor group, called
cancer development and progression [24]. Depending on the P6, as well as a pan-gastrointestinal tumor group known as
circumstance, miRNAs can function either as promoters of P8 [4]. This shows that analysis of proteins may also be used
oncogenes or can lead to the silencing of tumor suppressor to further classify and define tumors.
genes [24]. The first miRNA discovered to be oncogenic is
called miR-17-92. It is found to be amplified in B cell lym-
phomas and has oncogenic activity [24]. 2.7 Conclusion
Different tumor types have unique miRNA expression
profiles. These differences may be used to help distinguish Complex computer algorithms can take all of the information
one tumor type from another. These signatures can help listed above and further compile it into more nuanced classifi-
further subclassify subtypes of cancer within the same cations to give more information on tumor prognosis and pre-
tumor type, and even give important prognostic informa- diction. While it may appear as though many of these alterations
tion [24]. The use of miRNA expression can help distin- happen separately from one another, that is not the case, as
guish precursor lesions, like tubular adenomas in the colon, several of the alterations are associated with one another. For
from invasive colon cancers [24]. Some of these markers, instance, global DNA hypomethylation is associated with
for example, miR-25 and miR-223, which are elevated in genomic instability and chromosomal aneuploidy, while hyper-
non-small cell lung carcinoma, can be detected in serum mutated tumors tend to have low aneuploidy. However, at this
[24]. miRNA expression has been used to help classify the time, comprehensive molecular profiling including microarray,
aggressiveness of acute leukemias, lung adenocarcinomas, whole genome and/or whole exome sequencing, methylation
and outcomes of colorectal carcinoma. These types of array, RNA sequencing, and proteomics is not routinely per-
assays can help classify types of lung, prostate, and hepato- formed as it is cost prohibited, takes a lot of time to analyze this
cellular cancers [24]. High levels of miR-155 and/or low information, and is not necessary for current clinical practice.
levels of let-7a-2 miRNAs have been correlated with poor This opens up many questions for the future of somatic
survival in lung adenocarcinomas [24]. While, loss of miR- molecular testing in cancer.
122 expression in hepatocellular carcinomas have been
associated with tumor progression and the ability to metas-
tasize [24]. In colorectal adenocarcinoma, high miR-21 Open Questions
expression has been linked to poor survival [24]. Thus, • What types of molecular studies will be recom-
miRNA profiles may help provide prognostic information mended for a routine comprehensive cancer work up?
in cancer [24]. • Will a biopsy sample suffice for the amount of test-
Finally, the profiling of miRNA signatures may also aid ing required (how much neoplastic material will be
as predictive biomarkers in cancer. miR-21 expression in necessary for testing)?
non-­small cell lung cancer has been shotwn to be highly • If some of these findings are interrelated, is it neces-
expressed in patients with a shorter disease-free survival, as sary to run all of the studies to confirm the findings,
well as those who have platinum-based chemotherapy or will we proceed in more of a stepwise fashion?
resistance. Additional markers have been found to predict • Will any of this apply to preneoplastic tumors to
chemotherapy responses in advanced colorectal carcino- help predict which ones will progress? How much
mas. And different miRNA panels have been used to help do these features change in the metastatic setting,
predict a tumor’s response to different types of chemother- and does it matter?
apy [24].
20 L. Favazza DO

References The repertoire of mutational signatures in human cancer. Nature.

2020;578:94–101.
13. Alexandrov LB, Nik-Zainal S, Wedge DC, Aparicio SA, Behjati
1. Ben-David U, Amon A. Context is everything: aneuploidy in can- S, Biankin AV, Bignell GR, Bolli N, Borg A, Børresen-Dale AL,
cer. Nat Rev Genet. 2020;21:44–62. et al. Signatures of mutational processes in human cancer. Nature.
2. Taylor AM, Shih J, Ha G, Gao GF, Zhang X, Berger AC, 2013;500:415–21.
Schumacher SE, Wang C, Hu H, Liu J, et al. Genomic and func- 14. Cho RJ, Alexandrov LB, den Breems NY, Atanasova VS,
tional approaches to understanding cancer aneuploidy. Cancer Cell. Farshchian M, Purdom E, Nguyen TN, Coarfa C, Rajapakshe K,
2018;33:676–689.e673. Prisco M, et al. APOBEC mutation drives early-onset squamous
3. Duesberg P, Li R, Fabarius A, Hehlmann R. The chromosomal cell carcinomas in recessive dystrophic epidermolysis bullosa. Sci
basis of cancer. Cell Oncol. 2005;27:293–318. Transl Med. 2018;10:eaas9668.
4. Hoadley KA, Yau C, Hinoue T, Wolf DM, Lazar AJ, Drill E, Shen 15. Chalmers ZR, Connelly CF, Fabrizio D, Gay L, Ali SM, Ennis R,
R, Taylor AM, Cherniack AD, Thorsson V, et al. Cell-of-origin pat- Schrock A, Campbell B, Shlien A, Chmielecki J, et al. Analysis
terns dominate the molecular classification of 10,000 tumors from of 100,000 human cancer genomes reveals the landscape of tumor
33 types of cancer. Cell. 2018;173:291–304.e296. mutational burden. Genome Med. 2017;9:34.
5. Manier S, Salem KZ, Park J, Landau DA, Getz G, Ghobrial 16. Rheinbay E, Nielsen MM, Abascal F, Wala JA, Shapira O, Tiao
IM. Genomic complexity of multiple myeloma and its clinical G, Hornshøj H, Hess JM, Juul RI, Lin Z, et al. Analyses of non-­
implications. Nat Rev Clin Oncol. 2017;14:100–13. coding somatic drivers in 2,658 cancer whole genomes. Nature.
6. Shago M. Recurrent cytogenetic abnormalities in acute lympho- 2020;578:102–11.
blastic leukemia. Methods Mol Biol. 2017;1541:257–78. 17. Consortium, I.T.P.-C.A.o.W.G. Pan-cancer analysis of whole
7. Halani SH, Brat DJ. Defining neoplastic diseases differently: an genomes. Nature. 2020;578:82–93.
emerging paradigm from the cancer genome atlas lower-grade glio- 18. Lee CJ, Ahn H, Jeong D, Pak M, Moon JH, Kim S. Impact of muta-
mas project. Mol Cell Oncol. 2016;3:e1074333. tions in DNA methylation modification genes on genome-wide
8. Hieronymus H, Murali R, Tin A, Yadav K, Abida W, Moller H, methylation landscapes and downstream gene activations in pan-­
Berney D, Scher H, Carver B, Scardino P, et al. Tumor copy number cancer. BMC Med Genet. 2020;13:27.
alteration burden is a pan-cancer prognostic factor associated with 19. Skvortsova K, Stirzaker C, Taberlay P. The DNA methylation land-
recurrence and death. Elife. 2018:7. scape in cancer. Essays Biochem. 2019;63:797–811.
9. Schukken KM, Foijer F. CIN and aneuploidy: different con- 20. Ehrlich M, Lacey M. DNA hypomethylation and hemimethylation
cepts, different consequences. Bioessays. 2018;40 https://doi. in cancer. Adv Exp Med Biol. 2013;754:31–56.
org/10.1002/bies.201700147. 21. Calabrese C, Davidson NR, Demircioğlu D, Fonseca NA, He
10. Li Y, Roberts ND, Wala JA, Shapira O, Schumacher SE, Kumar Y, Kahles A, Lehmann KV, Liu F, Shiraishi Y, Soulette CM,
K, Khurana E, Waszak S, Korbel JO, Haber JE, et al. Patterns of et al. Genomic basis for RNA alterations in cancer. Nature.
somatic structural variation in human cancer genomes. Nature. 2020;578:129–36.
2020;578:112–21. 22. Kahles A, Lehmann KV, Toussaint NC, Hüser M, Stark SG,
11. Li MM, Datto M, Duncavage EJ, Kulkarni S, Lindeman NI, Roy Sachsenberg T, Stegle O, Kohlbacher O, Sander C, Rätsch G, et al.
S, Tsimberidou AM, Vnencak-Jones CL, Wolff DJ, Younes A, Comprehensive analysis of alternative splicing across tumors from
et al. Standards and guidelines for the interpretation and reporting 8,705 patients. Cancer Cell. 2018;34:211–224.e216.
of sequence variants in cancer: a joint consensus recommendation 23. Obeng EA, Stewart C, Abdel-Wahab O. Altered RNA processing in
of the Association for Molecular Pathology, American Society of cancer pathogenesis and therapy. Cancer Discov. 2019;9:1493–510.
Clinical Oncology, and College of American Pathologists. J Mol 24. Li MH, Fu SB, Xiao HS. Genome-wide analysis of microRNA
Diagn. 2017;19:4–23. and mRNA expression signatures in cancer. Acta Pharmacol Sin.
12. Alexandrov LB, Kim J, Haradhvala NJ, Huang MN, Tian Ng AW, 2015;36:1200–11.
Wu Y, Boot A, Covington KR, Gordenin DA, Bergstrom EN, et al.
 Tumor Microenvironment:
Coconspirator in Tumorigenesis 3
Zhiqiang Wang

Abstract
• Comprehend the importance of endothelial cells
The tumor microenvironment is pivotal for tumor estab- and angiogenesis in tumorigenesis
lishment, evasion of host immune surveillance, growth, • Know different types of immune cells, e.g., tumor-­
invasion, and metastasis by providing a nutrient mesh- associated macrophages (TAM), myeloid-derived
work rich in lympho-vasculature, cellular and metabolic suppressor cells (MDSC), and regulatory T cells
coconspirators of the tumor, and a signaling network for (Treg), and their respective roles and functions in
cross talks. Of note are a number of players that support tumorigenesis and clinical immunotherapy
or impede tumor growth directly by themselves or indi-
rectly through their products. Given the immensity of the
topic and limitation of the scope of this chapter, we choose
to discuss the following well-studied players and products 3.1 Introduction
that are the focus of research and resource commitment:
(1) fibroblasts and cancer-associated fibroblasts, (2) endo- Tumor invasion features the departure of tumor cells from
thelial cells with angiogenesis, (3) immune cells, and a the primary normal microanatomic compartment after in situ
couple of cellular products, i.e., (4) secretomes, and (5) proliferation to infringe into neighboring territory. During
exosomes. We recognize that under each of the five sub- this invasion process, the tumor cells are supported by the
titles lies a world of evidence that embodies numerous surrounding specialized environment to support their growth.
never-ending topics. Each of them deserves separate in-­ Such an environment is frequently referred to as tumor
depth exploration. We hence set out to briefly introduce microenvironment (TME). TME is the compartment where
their basic biological functions, discuss their roles in tumor cells survive and thrive. By coordinating the interac-
tumor growth and mechanisms of interaction with tumor tions and communications among different components via
cells and among themselves, summarize representative complicated mechanisms, the TME is essential in creating
research and milestones of achievements, and finally primary niches for the tumor cells and is pivotal for sustained
address their relevant clinical implications and advances. tumor growth, evasion of host immune surveillance, genera-
tion of neo-vasculatures, promotion of invasion, and
metastasis.
TME consists of physical and chemical components in
Learning Objectives addition to the tumor cells. TME denotes the stroma with
• Understand the notion of the tumor microenviron- lymph vasculature and immune cells. Stroma is the connec-
ment (TME) and the basic roles in tumorigenesis tive tissue that comprises both noncellular extracellular
• Learn the names and tumorigenic roles of cancer-­ matrix (ECM) and cellular components. Chemically, the
associated fibroblasts (CAF), secretome, and noncellular components of ECM consist of ground sub-
exosome stances that form the framework for tumoral and non-tumoral
tissue to organize and network with each other. The noncel-
lular components of ECM also consist of functional mole-
Z. Wang (*) cules, including enzymes, metabolic products, and signaling
Department of Pathology and Laboratory Medicine, molecules such as chemokines, auto/juxta/paracrine hor-
Henry Ford Hospital, Detroit, MI, USA mones, and growth factors. In aggregate, these proteins are
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 21

S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_3
22 Z. Wang

called secretomes. In contrast, the recently emerging concept Although most CAFs are believed to be derived from
of exosomes represents portions of cellular compartments local fibroblast activation or genetic alterations, other cells of
that are shed into the ECM. The cellular components of TME origins have also been described. Several cells are able to
consist of multiple cell types, i.e., fibroblasts, endothelial “de-differentiate” into CAFs, which include adipocytes [5],
cells, and immune cells [1]. endothelial cells [6], and bone marrow-derived mesenchy-
In this chapter, we limit our review to the fundamental roles mal stem cells [7]. Epithelial cell as the origin of CAFs has
of the most relevant components of the TME as coconspirators been one of the most extensively studied mechanisms called
in tumorigenesis. These coconspirators include fibroblasts and epithelial-mesenchymal transition (EMT). EMT is associ-
cancer-associated fibroblasts (CAFs), secretomes and exo- ated with tumor progression and metastasis [8]. Both benign
somes released by CAFs and tumor cells, endothelial cells and malignant epithelial cells can generate CAFs through
with angiogenesis, and the immune cells. We also describe the EMT [9]. Although the transformation from cancer cells to
interactions among these different coconspirators and address CAFs is well described, it is interesting to note the nonover-
issues relating to the molecular mechanisms and events. lapping genetic alterations between cancer cells and CAFs
Lastly, we discuss their potential clinical implications. [10]. Such discrepancies, however, appear to be due to
genetic alterations occurring at early precursor stages with
loss or gain of these alterations in subsequent subclonal evo-
3.1.1 Fibroblast and CAFs lution of tumor cells.
Analogous as they may appear, CAFs are different from
Fibroblasts are the most cellular component of the stroma. In normal fibroblasts. Two experimental models are most
contrast to fibrocyte, the suffix “-blast” implies immaturity exemplary that describe the differences: (1) In their mouse
and the potential to function as precursors to differentiate model of prostate cancer, Olumi et al. [11] reported that
into other cell types. Embryonically, the fibroblast is derived CAFs directed tumor progression by inducing prostatic
from mesoderm with a subset derived from the neural crest. intraepithelial neoplasia in normal prostate epithelial cells.
Functionally, fibroblasts synthesize a variety of proteins, Under the same experimental conditions, such tumor pro-
including collagens, fibronectin, and laminin, to constitute gression was not observed by normal fibroblasts. These
the ECM and basement membrane. In addition, they produce findings were displayed by both in vivo tissue recombina-
various types of ground substances (glycosaminoglycans, tion system and in vitro co-culture system. (2) In their breast
proteoglycans, and glycoproteins) and release cytokines, cancer experimental model systems, Dumont et al. [12]
juxta- or paracrine hormones or growth factors, proteases, showed that CAFs were protumorigenic and supported
etc., to maintain and modulate the matrix framework. tumor dissemination by inducing a mesenchymal-like phe-
Fibroblasts can also migrate and proliferate in response to notype of both premalignant and malignant mammary epi-
different signals, e.g., leukotrienes and growth factors. thelial cells. In comparison, fibroblasts isolated from benign
Although they lack specific markers, the fibroblasts are mor- reduction mammaplasty helped maintain the tumor cells
phologically recognized by their spindle shape, oval to elon- epithelial morphology and hindered tumor growth and
gated nuclei, and their microanatomical location in an organ. dissemination.
When the fibroblasts are activated and/or proliferate, the Once in place, CAFs exert their roles in regulating tumor
cytoplasm is plump, bright eosinophilic on H & E stained growth, invasion, and metastasis in various ways through
sections as collagen laying fibroblasts in diseased conditions. interactions with other ECM components and tumor cells.
The fibroblasts are known for their plasticity. Transformation One school of thought that contributed to the pool of evi-
into myofibroblasts has been well described in tissue repair. dence was by implicating the role of physical forces in the
Lastly, as it has been well known, in vitro experimental mod- tumorigenic process. Using the breast as the model organ,
els also showed that human fibroblasts could be reversely Paszek and Weaver [13] illustrated that the “tumors are rigid
transformed into pluripotent stem cells [2]. because they have a stiff stroma”. The stiffness of the stroma
Since the first descriptions of CAF were brought forth [3], was created by increased interstitial tension via infiltrating
there has been an explosion in knowledge regarding CAFs’ immune cells and activated resident fibroblasts that will
roles in ECM remodeling, angiogenesis, signal networking increase the production of macromolecules, e.g., collagens,
and transduction, harnessing and modulating host immune fibronectin, laminin, and ground substances. The stiffness
system, and clinical potential in therapy. While progress con- was further enhanced by cross-linking enzymes [14]. In
tinues to be made in this field, challenges remain. Recently, a response, such increased tension would feedback to the
group of experts in the field of CAF biology summarized and tumor cells and increase their invasiveness [15]. One CAF
issued a Consensus Statement to elucidate relevant issues of product, i.e., the lysyl oxidase (LOX) family, deserves atten-
CAFs, including definition, tissue generation, functions, het- tion for their roles in contributing to the stiffness of tumor
erogeneity and plasticity, and potential clinical benefits [4]. stroma, tumor growth, and progression [16].
3 Tumor Microenvironment: Coconspirator in Tumorigenesis 23

CAFs’ functions are diverse, which require heterogeneity logical and pathological processes, including enzymes, met-
and plasticity among CAFs to carry out different functions abolic products, chemokines, auto/juxta/paracrine hormones,
under different conditions. Interestingly, there have been growth factors, etc. In the context of TME, the Matrix
both experimental and clinical model systems that have pro- MetalloProteinases (MMPs) family and TGF-beta (TGFβ)
vided supporting evidence of such specialization. are the two major secretory proteins of the secretome that
Investigators showed that some CAFs could mutually con- have remained at the center stage of attention over the past
vert their functional phenotypes under different instruction decades.
signals released by different tumor subclones [17]. The Since its first description in early 1960 by Gross and
molecular events that are implicated behind the functional Lapiere [24], extensive knowledge has accumulated on
switch of CAFs phenotypes are still unclear. MMPs over the past decades. These achievements ranged
Various strategies have been developed to target CAFs at from bioengineering of the first cDNA clone, to protein crys-
the cellular level for clinical therapeutic purposes based on tal structure characterization, mouse model establishment,
the knowledge accumulated over the past decades. By target- and approval of the first MMP inhibitor for clinical applica-
ing fibroblast activation protein (FAP), which CAFs selec- tion. MMPs are a family of secreted zinc-dependent endopep-
tively produce, Ostermann et al. [18] demonstrated the tidases, 24 identified so far in humans, comprising
antitumor effect of monoclonal antibody (mAb) FAP5-DM1 collagenases, stromelysins, matrilysins, gelatinases, and oth-
on xenograft immunodeficient mouse models for lung, pan- ers. MMPs are expressed by both tumor cells and CAFs and
creas, and head and neck cancers. Sibrotuzumab, a mAb, was are known for their roles in tumor invasion as groundbreakers
tried on patients with advanced colorectal cancer for clinical to degrade ECM components at the leading edge of the tumor.
phase I and early phase II trials, although eventually was not Additionally, MMPs also have regulatory functions on cell
successful [19, 20]. However, the setbacks did not stop inves- adhesion molecules processing which eventually impact cell
tigators from further pursuing alternative means. By conju- migration and invasion. The roles of MMPs at different stages
gating promelittin-containing FAP-cleavable sequences to of cancer survival, growth, invasion, and metastasis were
pegylated phospholipids and anchoring them to reduced gra- reviewed by Gonzalez-Avila et al. [25] in depth.
phene oxide (rGO) nanosheets, Kim et al. [21] reported that Given the roles of MMPs in cancer invasion and metasta-
the resulting nanosheets, PL-rGO, showed tumor inhibitory sis, MMPs have been applied as targets for MMPs inhibitors
effect both on in vitro culturing and in vivo mouse models. in clinical trials and as biomarkers for prognostic purposes.
On the other hand, by taking advantage of CAFs heterogene- Dozens of clinical trials have been conducted, some showing
ity and plasticity, other innovative methods are being devised encouraging preclinical data. However, progress on MMP
to reprogram and salvage CAFs back to their “normal” phe- inhibitors has been slow, with none demonstrating antitumor
notypes to help achieve tumor suppression effects [4]. effects in clinical trials. The value of MMPs for their clinical
Progress made in molecular genetics has helped advance application has been best reflected as biomarkers for prog-
our understanding of CAFs and their cross-interactions with nostic purposes. Two assays exemplify MMPs biomarkers’
cancer cells. The role of Braf, its associated ERK–MAPK sig- roles. One of the assays is the 70-gene MammaPrint assay
naling pathway, and clinically approved the application of Braf [26]. MammaPrint was approved by the US Food and Drug
inhibitor for melanoma have been well known [22]. The find- Administration and regulators in the European Union as an
ings that Hirata et al. [23] published on BRAF mutant mela- adjunct prognostic assay for patients with stage I/II breast
noma cells’ response to PLX4720 was most revealing. After cancer who are below 61 years of age with negative or 1–3
the initial response of melanoma cells to PLX4720, a rapid lymph nodes positive disease. MMP9 is listed as 1 of the 5
tolerance was developed. Such tolerance was because of genes under “altered extracellular matrix adhesion and
PLX4720’s effect on TME, leading to “paradoxical” activation remodeling” of the “Tissue Invasion and Metastasis” cate-
of CAFs with more matrix production and remodeling. As a gory. In addition, MMP9 is also included as 1 of 6 genes
result, an increased integrin β1/FAK/Src signaling was seen in under “altered expression of known angiogenesis effectors”
melanoma cells. After adding a FAK inhibitor to the anti-Braf of the “Sustained Angiogenesis” category. The other assay is
regimen, the investigators showed effective suppression of the Oncotype DX assay [27]. The Oncotype DX assay has
ERK signaling, and hence more effective tumor treatment. been most widely used and well received by clinicians
mainly because of the convenience of using formalin-fixed
paraffin-embedded tissues and its easy-to-interpret recur-
3.1.2 Secretome rence score-based reporting. MMP11 is included as one of
the 16 cancer-related genes.
Secretome denotes the collection of proteins released by a Of the many growth factors, TGF-beta (TGFβ) has been
cell into the extracellular space as defined by TME. These known as a central player for its complex roles in homeosta-
proteins are of different functional types involved in physio- sis and tumorigenesis by its signaling via canonical SMAD
24 Z. Wang

pathway and several non-canonical pathways. On the one olites and secretion of exosomes. Exosomes and transfer of
hand, TGFβ promotes the transdifferentiation of fibroblasts information between cancer cells and other cellular compo-
to CAF and recruits fibroblasts to the tumor. On the other nents of TME have emerged as a novel and important mecha-
hand, CAFs secrete large amount of TGFβ to affect tumor nism that has fueled many research opportunities in recent
growth by enhancing cross talk between cancer cells and years. It is worthwhile to mention that given the key roles
CAF, and by mediating metabolic activities of tumor. exosomes play, the term “Exosomics” was used to help carry
Similarly, another paracrine growth factor, hepatocyte the notion and its associated importance.
growth factor (HGF), also deserves attention for its contribu- Exosomes are known as a type of small extracellular ves-
tory roles to promote cancer progression. By overexpression icles (30–150 nm) that form via the cellular process of endo-
of HGF and TGFβ, Kuperwasser et al. [28] showed that nor- cytosis. The endosomes thus formed are shed by the parent
mal and malignant human breast tissues could be recon- cells into ECM with a double-layered lipid membrane. As
structed in mice. Therapeutically, it is not difficult to find part of the parent cell, exosome represents a portion of the
clinical trials targeting the CAF-induced signaling mole- cellular cytoplasm and membrane containing parent pro-
cules, especially TGFβ and HGF inhibitors and mAb. The teins, lipids, and nucleic acids. The proteins in exosomes
results are promising but mixed [29]. consist of endocytic proteins and protein molecules that are
In contrast, there has been evidence showing biphasic both membranous (receptors, adhesion molecules, channels)
roles of TGFβ in both promoting and hindering tumor growth and cytoplasmic (cytoskeletons, enzymes, players of differ-
and progress. In their study on skin tumors using transgenic ent pathways). Interestingly, on examination of the lipid con-
mouse models, Cui et al. [30] found that TGFβ inhibited tents, Llorente et al. [34] found significant differences
benign tumor formation and enhanced invasive carcinomas’ between exosomes and parent cells. The exosomes lipids
progression. Other CAF secretory proteins also demon- were highly enriched in a subset of lipids, e.g., glycosphin-
strated such inhibitory roles on cancer growth. In the study golipids, sphingomyelin, cholesterol, and phosphatidylser-
on colon cancer secretomes using in vitro cell culture, xeno- ine. Other non-enriched lipid species were selectively
graft mouse models, and clinically resected colon cancers, included. Such a pattern of lipid composition helped raise
Chen et al. [31] revealed novel molecular signature secretory the possibility of a yet-to-be elucidated sorting mechanism
proteins that exhibited negative modulatory effects on colon and potential tumor biomarkers.
cancer cell growth. Although largely cytoplasmic, it was demonstrated that
In addition to proteinase enzymes and signaling mole- exosomes contain different types of DNA species, i.e.,
cules, other aspects of secretomes impacting cancer biology single-­stranded DNA, mitochondrial DNA, and double-­
include senescence-associated secretome, inflammatory sec- stranded nuclear DNA. Thakur et al. [35] convincingly dem-
retome, angiogenesis secretome. The functions of these onstrated the presence of double-stranded genomic DNA in
diverse secretomes are intertwined. For instance, senescence exosomes and provided further experimental evidence show-
is a mechanism that prevents and blocks cancer cell prolif- ing that exosomal DNA represents the entire genome and
eration and, therefore, is regarded as a tumor suppressor carries detectable mutations that reflect parent tumor cells.
mechanism. As reviewed in depth by da Cunha [32], driven Such findings were received with enthusiasm and opened
by NF-ĸB, the secretome of cells in a senescent state can doors for many research opportunities, notably in non-small
contain as many as 80 bioactive molecules. These molecules cell lung carcinomas. Taking advantage of commercial DNA
include pro-senescent, pro-apoptotic, and antiangiogenic isolation products that have been made available, researchers
protein molecules. In association with p16 and SASP were able to detect major EGFR mutations with improved
(senescence-­ associated secretory phenotype) that incite sensitivity on liquid biopsy and pleural effusions using
immune cells, these secretomes facilitate tissue regeneration exosomal-­derived DNA.
and senescent cell removal. On the other hand, however, In the recent years, exosomal RNA has been extensively
these molecules also include aforementioned MMPs and investigated due to technical advances and the availability of
growth factors, plus pro-inflammatory leukotrienes and next-generation sequencing-based RNA sequencing. Various
angiogenic factors. As stated by Rodier and Campisi [33], RNA species have been characterized, including coding
cellular senescence might take part in complex biological mRNAs and noncoding RNAs. Through deep RNA sequenc-
processes that include tumor promotion. ing, Nolte-’t Hoen et al. [36] uncovered a variety of noncod-
ing RNAs, including microRNAs (miRNA), small nuclear
RNAs (snRNAs), circular RNAs (circRNAs), etc.
3.1.3 Exosome Interestingly, like lipids, a perplexed sorting mechanism also
exists by selectively including certain RNA species but
Other mechanisms involving CAF interactions and cross talk excluding others in the exosomes [37]. There have been dif-
among themselves and cancer cells include effects of metab- ferent schools of thought trying to address this issue, includ-
3 Tumor Microenvironment: Coconspirator in Tumorigenesis 25

ing the presence of recognizable RNA signal sequences and cells. In turn, endothelial cells can exhibit different pheno-
regulatory pathways. types, including “dedifferentiating” into CAFs as aforemen-
Research activities exploring the roles of exosomal RNA tioned under the CAFs section [6]. Behind the cellular
have been advancing on different fronts from neurodegener- components are multiple factors that impact the angiogenic
ative diseases to different types of cancer. In one transla- process, including hypoxia, tumor cell burden, inflamma-
tional study aiming for therapeutic purposes against tion, genetic alterations of tumor cells, and signaling mole-
pancreatic ductal adenocarcinoma (PDAC), Kamerkar et al. cules. Among the many known factors is vascular endothelial
[38] demonstrated state-of-the-art design of their experiment growth factor (VEGF) and integrin.
by taking advantage of the unique features of exosome as a Of the prolific angiogenic growth factors (VEGFs, FGF,
delivery system and molecular features of PDAC. As men- PDGFs), VEGF has been the focus of attention because of its
tioned above, exosomes are a portion of cellular cytoplasm pivotal roles in angiogenesis. Bevacizumab (Avastin), the
and membrane that contains transmembrane proteins. CD47 mAb against VEGF, is the most widely used therapeutics in
is an integrin-associated transmembrane protein that func- oncology and other FDA-approved anti-VEGF drugs. The
tions to protect the cells from being phagocytosed by mono- family of VEGF have several members, i.e., VEGF A-E and
cytes and macrophages. It, therefore, enhances endocytosis PIGF (Placenta Growth Factor), and a number of isoforms
and delivery of the cytoplasmic content to the recipient cells due to alternative splicing. VEGF-A is the major player in
[39]. As PDAC frequently harbor mutations in KRAS gene, regulation of angiogenesis and thus is frequently referred to
investigators developed exosomes derived from normal as VEGF. VEGF-C and VEGF-D are primarily implicated in
fibroblast-like mesenchymal cells to carry short interfering regulation of lymphangiogenesis, which is recognized as
RNA to functionally knock down this oncogenic KRAS another mode of vascularization in tumors and as alternate
G12D mutation. This exosome-based treatment by combined route for cancer cell dissemination [41]. Overexpression of
use of these features successfully achieved tumor suppres- the VEGF genes was detected in both tumor and CAFs and
sion in mouse models of PDAC with a significant increase in the proteins are secreted as soluble growth factors in the
overall survival. ECM that function via binding with VEGF receptors on the
The creation of a dedicated program met the enthusiasm surface of endothelial cells or other target cells. The binding
in extracellular RNA research under the US National triggers a cascade of signaling events in the VEGF regulated
Institutes of Health, the Extracellular RNA Communication pathways that eventually lead to proliferation of endothelial
Consortium (ERCC). Another program is the International cells and formation of new blood vessels. It is worth men-
Society for Extracellular Vesicles (ISEV), which surveyed tioning, however, that the tumor neovasculature thus formed
extracellular vesicles and RNA. The ISEV survey showed are irregular and defective. As a result, the defective vascula-
that although relatively new, the field has generated great ture will expose multiple cell types to the TME with various
interest and commitment both in resource and funding. stimulating signals and eventually trigger a cascade of events
However challenges remain in elucidating the mechanisms leading to more aggressive tumor behavior [42].
behind RNA production, transfer, and function in recipient Knowledge of VEGF led to an investigation using mAB
cells. to block its function, which was successfully tested on xeno-
graft mouse models by Kim et al. [43]. The preclinical suc-
cess led to clinical trials using a humanized murine antibody,
3.1.4 Endothelial Cells and Angiogenesis i.e., A.4.6.1 and eventual FDA approval of Bevacizumab in
2004 for previously untreated metastatic colorectal cancer
Angiogenesis and lymphangiogenesis is an integral part of and subsequently for patients with non-squamous non-small
TME. The clinical implications of TME can be dated back cell lung carcinoma among other malignancies. The success
about half a century ago when Dr. Judah Folkman published also provided opportunities for FDA approval of other VEGF
his article “ Tumor angiogenesis: therapeutic implications” inhibitors, including multiple tyrosine kinase inhibitors
[40]. Briefly, tumor growth is accompanied by neovascular- (TKI); Ramucirumab, which targets the VEGF receptor
ization, which is defined by angiogenesis and lymphangio- (VEGFR2); Ziv-aflibercept, which targets a chimeric VEGF
genesis essential to supply the necessary nutrients and (VEGF-A, VEGF-B, PIGF)
oxygen to support tumor growth, proliferation, and metasta- Integrin is another TME molecule that deserves attention
sis. Angiogenesis is a dynamic process in response to angio- because of its roles as hub for crosstalk between cells and
genic stimulation involving the participation of multiple cell ECM as well as roles in modulating tumor angiogenesis.
types and factors in the TME. Central to the angiogenic pro- Integrin comprises two subunits, i.e., α and β subunits, that
cess is the endothelial cells that can come from existing ves- form a heterodimer to combine with target receptors found in
sels from which new vessels sprout or originate from bone substances of ECM, which include fibronectin, vitronectin,
marrow-derived mesenchymal stem cells or even tumor stem tenascin, fibrillin, etc. Detection of integrin αvβ3 in endothe-
26 Z. Wang

lial cells helped link its role to angiogenesis and offered notypes in the tumorigenic process; i.e., M1 macrophages
research opportunities for targeted therapy in several differ-are pro-inflammatory and cytotoxic to tumor cells, M2 mac-
ent tumors, including glioblastoma (GBM) [44]. GBM has rophages are anti-inflammatory and are friendly to tumor
rich vasculature and high-level integrin αvβ3 expression and cells. It is worth mentioning that M1 and M2 macrophages
matrix protein vitronectin. Taking advantage of these fea- should not be simply viewed as physically distinct and sepa-
tures, MacDonald et al. [45] demonstrated antiangiogenic rate subclonal cell populations, but instead, they represent
efficacy of αv integrin antagonist in mouse models of functional plasticity that can switch from one phenotype to
GBM. The successful story helped devise Cilengitide, the another upon different signaling they receive from TME at
first integrin antagonist, and clinical trials. Although the different tumorigenic stages. The concept that TAMs pheno-
Phase III clinical trial failed to show improved therapeutic type can mutually switch was supported by DeNardo et al.
outcome in newly diagnosed GBM with methylated MGMT [47] using their mouse models in which suppression of M2
promoter, findings accumulated highlighted the potential of function of TAM helped improve the tumor toxicity role of
targeting integrins for the treatment of GBM [46] M1 type.
The unique polarizing phenotypes of TAMs also caught
investigators’ attention to extrapolate their inherent clinical
3.1.5 Immune Cells significance and implications. Progress is being made on
several fronts including prognosticating different types of
Dense inflammatory infiltration, tumor infiltrating lympho- solid tumors by quantifying TAMs’ density and M1/M2 ratio
cytes (TILs) in many solid tumors have been observed and using established immunohistochemical markers or by deter-
studied for a long time. The TILs are thought to be an mining TAMs’ locations to correlate with the tumor’s clini-
immune response against tumor cells as part of immune sur- cal behavior [48]. Advances in therapeutics are also
veillance and elimination of tumor cells that are foreign to noticeable, including suppressing tumor growth by inhibit-
the immune system. This hypothesis is supported by obser- ing M2 TAMs or promoting switching of M2 to M1 TAMs;
vations such as patients with immunosuppression (AIDS; immunotherapies by PD1/PD-L1 signaling blockade; tar-
organ transplant) developing certain types of cancers. On the geted therapies by mAb and inhibitors; genetic modifica-
other side, inflammation has been known to be tumorigenic, tions, etc.
as exemplified by association of inflammatory bowel disease MDSCs have been increasingly recognized as a major
and colorectal adenocarcinoma, viral hepatitis and hepato- regulator of host immune response to evade host immuno-
cellular carcinoma, H. pylori gastritis and lymphoma, etc. surveillance. They are mobilized and infiltrate tumors to
Such polarity highlights the intricacy between the host’s disrupt or hinder the processes of TAMs polarization, CTL
immune system and cancer. and NK cells cytotoxicity, and dendritic cells antigen pre-
Along with tumor cells, CAFs, endothelial cells, and sentation. MDSCs were discovered from an early study of
other mesenchymal cells, immune cells represent one of the inflammation on tumor-bearing mouse models in which a
significant cellular types in the TME. They include neutro- subset of myeloid cells expressing CD11b and Gr-1 were
phils, tumor-associated macrophages (TAM), myeloid-­ revealed [49]. Since there is no mouse Gr-1 ortholog in
derived suppressor cells (MDSC), dendritic cells (DC), humans, MDSC is functionally and phenotypically classi-
natural killer (NK) cells, or lymphocytes as adaptive immune fied into two major subsets, polymorphonuclear (PMN)
response. The immune cells can be resident to the TME or and monocytic (M)-MDSC. Criteria to define different
recruited into TME by migration via different signaling path- types of MDSCs were suggested based on a combination
ways and mechanisms. Functionally, immune cells can be of a set of differentially expressed surface markers, includ-
cytotoxic to tumor cells, e.g., CD8+ cytotoxic T lympho- ing CD11b, CD14, CD15, etc. Functionally, MDSCs’
cytes (CTL) and NK cells, or they can be immunosuppres- capabilities to suppress NK cell, lymphocytes, especially
sive and protumorigenic. Since tumor-suppressive or CTLs, were demonstrated by a number of experimental
cytotoxic functions of immune cells including neutrophils, models. MDSCs also produce several molecules to help
DCs, CTL, and NK cells have been well documented in the suppress immune responses, e.g., arginases, NO, and
literature, in the following paragraphs, we focus our discus- ROS. Although knowledge on MDSCs has been improving
sion on three types of immune cells that possess immuno- in recent years, clinical advances in utilizing MDSCs as a
modulatory functions: (1) TAMs, (2) MDSCs, and (3) prognostic marker and as a therapeutic target by immuno-
Regulatory T (Treg) cells. modulatory therapies have not been significant. In addi-
TAMs are believed to differentiate from monocytes and tion, the involvement of MDSCs is also being studied in
are recruited from blood to TME by tumor emitted or CAF-­ infectious disease, autoimmunity, aging, obesity, preg-
derived signals. TAMs are known for their polarizing phe- nancy, and transplantation.
3 Tumor Microenvironment: Coconspirator in Tumorigenesis 27

Treg cells are another immune cell type in the TME that associated secretomes, exosomes, endothelial cells, and
possess immune-modulatory functions and functionally immune cells, with emphasis on their fundamental biological
demonstrate some overlap with MDSCs in regard to the sup- roles and functions in tumorigenesis. While each of these
pression of tumor-associated antigen presentation and cyto- coconspirators appears to function as an independent entity,
toxic T cell function. The discovery of Treg cells represented their functions are intertwined and overlap in promoting
efforts of a generation of immunologists that started with tumorigenesis. The field is still evolving, but the definitions,
the hypothesis in the 1970s that a type of suppressive T cells basic functions, and roles of the conspirators in TME in
needs to be present to regulate and maintain homeostasis of tumorigenesis were highlighted. We have tried to review the
the host immune system. The breakthrough came with the most current research models, milestones, cutting-edge tech-
finding of CD25 as the surface marker for Treg in 1995. nologies in targeted therapies, and their roles in providing
With the subsequent discovery of Treg cells specific fork- personalized therapies and potentials for clinical
head helix transcription factor (Foxp3) as necessary and suf- applications.
ficient for the immune suppressive function, the
CD4+CD25+Foxp3+ lymphocytes were therein established
as Treg cells [50].
Study Questions
Treg cells are a heterogeneous population and can be
• With the accumulation of knowledge on CAFs,
either naturally occurring/thymus-derived or induced/periph-
what directions do investigators need to pursue in
erally derived. Several mechanisms have been identified on
the future?
how Treg cells exert their suppressive functions on effector T
• What can clinical scientists make the best use of
cells by intercellular interactions, secretome, metabolic
CAFs’ secretome?
interruptions, or by regulating FOXP3 expression. The
• How does the notion of Exosomics enlighten us?
improved understanding of Treg cells led to many clinical
• What aspects of tumor angiogenesis remain to be
trials. In organ transplantation medicine, researchers took
further investigated?
advantage of Treg cells’ immune-modulatory functions to
• With the advent of immunotherapy, immune cells in
infuse patients with ex vivo expanded Treg cells to prevent
TME are getting the attention and resource commit-
graft-versus-host disease [51]. On the other hand, investiga-
ment they deserve. What do we have to do to main-
tors also utilized the essence of Treg cells’ protumorigenic
tain the momentum?
effects to devise immunotherapeutic strategies by depletion
of Treg cells in the hope of augmenting antitumor immune
responses by effector T cells. Such depletion strategies can
be achieved by direct infusion of Treg-depleted donor
­lymphocytes to patients with hematopoietic malignancies or References
by functionally decreasing Treg cells via blocking their
migration to TME, targeting immune checkpoint (CTLA-4, 1. Liotta LA, Kohn EC. The microenvironment of the tumour-host
interface. Nature. 2001;411:375–9.
PD-1, etc.) and other molecules of Treg cells. With FDA 2. Takahashi K, Tanabe K, Ohnuki M, et al. Induction of pluripotent
approval of ipilimumab, an anti-CTLA-2 mAb, as the first- stem cells from adult human fibroblasts by defined factors. Cell.
line treatment of several solid tumors, including non-small 2007;131:861–72.
cell lung cancer, therapies targeting immune-modulatory 3. Kalluri R, Zeisberg M. Fibroblasts in cancer. Nat Rev Cancer.
2006;6:392–401.
cells hold great promise. However, given the heterogeneity, 4. Sahai E, Astsaturov I, Cukierman E, et al. A framework for advanc-
plasticity, and diverse effects of Treg cells on tumor, more ing our understanding of cancer-associated fibroblasts. Nat Rev
studies are needed on their tumorigenic mechanisms in order Cancer. 2020;20:174–86.
to provide new strategies for tumor immunotherapy and 5. Bochet L, Lehuédé C, Dauvillier S, et al. Adipocyte-derived fibro-
blasts promote tumor progression and contribute to the desmoplas-
develop more effective and less toxic antitumor compounds. tic reaction in breast cancer. Cancer Res. 2013;73:5657–68.
6. Piera-Velazquez S, Li Z, Jimenez SA. Role of endothelial-­
mesenchymal transition (EndoMT) in the pathogenesis of fibrotic
3.2 Conclusion disorders. Am J Pathol. 2011;179:1074–80.
7. Quante M, Tu SP, Tomita H, et al. Bone marrow-derived myofibro-
blasts contribute to the mesenchymal stem cell niche and promote
There have been significant advances in the field of TME tumor growth. Cancer Cell. 2011;19:257–72.
with many new discoveries. This chapter set out to cover the 8. Hay ED. An overview of epithelio-mesenchymal transformation.
most important achievements in both historical perspective Acta Anat (Basel). 1995;154:8–20.
9. Petersen OW, Nielsen HL, Gudjonsson T, et al. Epithelial to mesen-
and recent progress in TME literature. Discussions were chymal transition in human breast cancer can provide a nonmalig-
focused on five key coconspirators in TME, namely CAFs, nant stroma. Am J Pathol. 2003;162:391–402.
28 Z. Wang

10. Kurose K, Hoshaw-Woodard S, Adeyinka A, et al. Genetic model tinib in patients with advanced non-small cell lung cancer. Cancer.
of multi-step breast carcinogenesis involving the epithelium and 2017;123:2936–44.
stroma: clues to tumour microenvironment interactions. Hum Mol 30. Cui W, Fowlis DJ, Bryson S, et al. TGFbeta1 inhibits the formation
Genet. 2001;10:1907–13. of benign skin tumors: but enhances progression to invasive spindle
11. Olumi AF, Grossfeld GD, Hayward SW, et al. Carcinoma-­ carcinomas in transgenic mice. Cell. 1996;86:531–42.
associated fibroblasts direct tumor progression of initiated human 31. Chen SX, Xu XE, Wang XQ, et al. Identification of colonic fibro-
prostatic epithelium. Cancer Res. 1999;59:5002–11. blast secretomes reveals secretory factors regulating colon cancer
12. Dumont N, Liu B, Defilippis RA, et al. Breast fibroblasts modu- cell proliferation. J Proteome. 2014;110:155–71.
late early dissemination, tumorigenesis, and metastasis through 32. da Cunha BR, Domingos C, Stefanini ACB, et al. cellular inter-
alteration of extracellular matrix characteristics. Neoplasia. actions in the tumor microenvironment: the role of secretome. J
2013;15:249–62. Cancer. 2019;10:4574–87.
13. Paszek MJ, Weaver VM. The tension mounts: mechanics meets 33. Rodier F, Campisi J. Four faces of cellular senescence. J Cell Biol.
morphogenesis and malignancy. J Mammary Gland Biol Neoplasia. 2011;192:547–56.
2004;9:325–42. 34. Llorente A, Skotland T, Sylvänne T, et al. Molecular lipidomics of
14. Kechagia JZ, Ivaska J, Roca-Cusachs P. Integrins as biomechani- exosomes released by PC-3 prostate cancer cells. Biochim Biophys
cal sensors of the microenvironment. Nat Rev Mol Cell Biol. Acta Mol Cell Biol Lipids. 2013;1831:1302–9.
2019;20:457–73. 35. Thakur BK, Zhang H, Becker A, et al. Double stranded DNA
15. Mohammadi H, Sahai E. Mechanisms and impact of altered tumour in exosomes: a novel biomarker in cancer detection. Cell Res.
mechanics. Nat Cell Biol. 2018;20:766–74. 2014;24:766–9.
16. Wang TH, Hsia SM, Shieh TM. Lysyl oxidase and the tumor micro- 36. Nolte-’t Hoen EN, Buermans HP, Waasdorp M, et al. Deep sequenc-
environment. Int J Mol Sci. 2016;18:62. ing of RNA from immune cell-derived vesicles uncovers the selec-
17. Biffi G, Oni TE, Spielman B, et al. IL1-induced JAK/STAT sig- tive incorporation of small non-coding RNA biotypes with potential
naling is antagonized by TGFbeta to shape CAF heterogeneity regulatory functions. Nucleic Acids Res. 2012;40:9272–85.
in pancreatic ductal adenocarcinoma. Cancer Discov. 2019;9: 37. Squadrito ML, Baer C, Burdet F, et al. Endogenous RNAs modulate
282–301. microRNA sorting to exosomes and transfer to acceptor cells. Cell
18. Ostermann E, Garin-Chesa P, Heider KH, et al. Effective immuno- Rep. 2014;8:1432–46.
conjugate therapy in cancer models targeting a serine protease of 38. Kamerkar S, Lebleu VS, Sugimoto H, et al. Exosomes facilitate
tumor fibroblasts. Clin Cancer Res. 2008;14:4584–92. therapeutic targeting of oncogenic KRAS in pancreatic cancer.
19. Hofheinz RD, Al-Batran SE, Hartmann F, et al. Stromal antigen Nature. 2017;546:498–503.
targeting by a humanised monoclonal antibody: an early phase II 39. Brown EJ, Frazier WA. Integrin-associated protein (CD47) and its
trial of sibrotuzumab in patients with metastatic colorectal cancer. ligands. Trends Cell Biol. 2001;11:130–5.
Onkologie. 2003;26:44–8. 40. Folkman J. Tumor angiogenesis: therapeutic implications. N Engl J
20. Scott AM, Wiseman G, Welt S, et al. A phase I dose-escalation Med. 1971;285:1182–6.
study of sibrotuzumab in patients with advanced or metastatic 41. Alitalo A, Detmar M. Interaction of tumor cells and lymphatic ves-
fibroblast activation protein-positive cancer. Clin Cancer Res. sels in cancer progression. Oncogene. 2012;31:4499–508.
2003;9:1639–47. 42. Apte RS, Chen DS, Ferrara N. VEGF in signaling and disease:
21. Kim MG, Shon Y, Kim J, et al. Selective activation of anticancer beyond discovery and development. Cell. 2019;176:1248–64.
chemotherapy by cancerassociated fibroblasts in the tumor micro- 43. Kim KJ, Li B, Winer J, et al. Inhibition of vascular endothelial
environment. J Natl Cancer Inst. 2017;109(1):djw186. https://doi. growth factor-induced angiogenesis suppresses tumour growth
org/10.1093/jnci/djw186. in vivo. Nature. 1993;62:841–4.
22. Dhomen N, Reis-Filho JS, da Rocha DS, et al. Oncogenic Braf 44. Brooks PC, Clark RA, Cheresh DA. Requirement of vascular inte-
induces melanocyte senescence and melanoma in mice. Cancer grin αvβ3 for angiogenesis. Science. 1994;264:569–71.
Cell. 2009;15:294–303. 45. MacDonald TJ, Taga T, Shimada H, et al. Preferential susceptibil-
23. Hirata E, Girotti MR, Amaya Viros A, et al. Intravital imaging ity of brain tumors to the antiangiogenic effects of an a(v) integrin
reveals how BRAF inhibition generates drug- tolerant microen- antagonist. Neurosurgery. 2001;48:151–7.
vironments with high integrin beta1/FAK signaling. Cancer Cell. 46. Stupp R, Hegi ME, Gorlia T, et al. Cilengitide combined with stan-
2015;27:574–88. dard treatment for patients with newly diagnosed glioblastoma with
24. Gross J, Lapiere CM. Collagenolytic activity in amphibian tis- methylated MGMT promoter (CENTRIC EORTC 26071-22072
sues: a tissue culture assay. Proc Natl Acad Sci U S A. 1962;47: study): a multicentre, randomised, open-label, phase 3 trial. Lancet
1014–22. Oncol. 2014;15:1100–8.
25. Gonzalez-Avila G, Sommer B, Mendoza-Posada DA, et al. Matrix 47. DeNardo DG, Brennan D, Rexhapaj E, et al. Leukocyte complexity
metalloproteinases participation in the metastatic process and their in breast cancer predicts overall survival and functionally regulates
diagnostic and therapeutic applications in cancer. Crit Rev Oncol response to chemotherapy. Cancer Discov. 2011;1:54–67.
Hematol. 2019;137:57–83. 48. Jayasingam SD, Marimuthu Citartan M, Thang TH, et al. Evaluating
26. Mook S, Schmidt MK, Viale G, et al. The 70-gene prognosis-­ the polarization of tumor-associated macrophages into M1 and M2
signature predicts disease outcome in breast cancer patients with phenotypes in human cancer tissue: technicalities and challenges in
1-3 positive lymph nodes in an independent validation study. Breast routine clinical practice. Front Oncol. 2019;9:1512.
Cancer Res Treat. 2009;116:295–302. 49. Bronte V, Wang M, Overwijk WW, et al. Apoptotic death of CD8
27. Paik S, Shak S, Tang G, et al. A multigene assay to predict recur- þ T lymphocytes after immunization: induction of a suppressive
rence of tamoxifen-treated, node-negative breast cancer. N Engl J population of mac-1 þ /gr-1 þ cells. J Immunol. 1998;161:5313–20.
Med. 2004;351:2817–26. 50. Sakaguchi S, Yamaguchi T, Nomura T, et al. Ono M. regulatory T
28. Kuperwasser C, Chavarria T, Wu M, et al. Reconstruction of func- cells and immune tolerance. Cell. 2008;133:775–87.
tionally normal and malignant human breast tissues in mice. Proc 51. Brunstein CG, Miller JS, Cao Q, et al. Infusion of ex vivo expanded
Natl Acad Sci U S A. 2004;101:4966–71. T regulatory cells in adults transplanted with umbilical cord blood:
29. Tarhini AA, Rafique I, Floros T, et al. Phase 1/2 study of rilotu- safety profile and detection kinetics. Blood. 2011;117:1061–70.
mumab (AMG 102), a hepatocyte growth factor inhibitor, and erlo-
 Hallmarks of Metastasis: Molecular
Underpinnings 4
Juan C. Gomez-Gelvez and Dhananjay A. Chitale

Abstract metastatic tumor [1]. Cancer cells must initially leave their
Metastasis is one of the most important hallmarks in can- primary site to enter and circulate within the bloodstream,
cer biology responsible for the majority of cancer-related which includes acquiring invasion properties as well as the
deaths in patients. Cancer cells undergo a series of pro- ability of resisting physical stress within the blood vessels.
cesses orchestrated by genetic changes in order to acquire Subsequently, cancer cells must infiltrate distant tissue, a
invasive properties that allow them to migrate to second- step of the process that is facilitated by alterations that enable
ary sites and become a metastatic focus. In this chapter, formation of a premetastatic niche. Finally, cancer cells must
we will review the molecular bases of the most relevant adapt to the new microenvironment of the metastatic site and
steps in metastasis, including invasion, intravasation, cir- evade destruction by cellular immunity.
culation, extravasation, and metastatic colonization. Although advances in molecular biology have contributed
to identify key steps of the process, the genetic mechanisms
of metastasis remain poorly understood. An improved under-
Learning Objectives standing of the dynamic of this process is crucial to identify
• To understand the molecular genetic bases of the possible therapeutic targets that may help stop metastasis
most important processes contributing to metastasis. and tumor progression. In this chapter, we will review the
• To recognize the signaling pathways commonly molecular bases of the most relevant steps in metastasis.
involved in metastasis.
• To identify the interactions of cancer cells with other
tumor constituents (e.g., tumor microenvironment 4.2 Dissemination and Invasion
and extracellular matrix) that enable metastasis.
Dissemination and invasion refer to the ability of cancer cells
to infiltrate surrounding tissue and constitutes one of the
most important steps in malignant progression. The potential
4.1 Introduction of cancer cells to disseminate and invade is influenced by
numerous factors capable of modifying their growth, prolif-
Metastasis is defined as the development of secondary eration, motility, and/or migratory ability. These factors are
tumors in other parts of the body different from the primary the results of a close interaction between different tumor
tumor site. Metastasis is currently the leading cause of mor- constituents such as cancer cells, tumor microenvironment,
bidity, mortality, and treatment failure in patients with can- and extracellular matrix (ECM).
cer. Although millions of cancer cells are released into the Initially, cancer cells are in an epithelial-like state where
blood stream daily, metastasis is actually a very inefficient they are static and bound to other epithelial cells as well as to
process. As an example, studies performed on animal models the ECM. As tumor progresses, cancer cells acquired the
have demonstrated that only a vast minority of melanoma ability to detach and escape from the primary tumor through
cells (approximately 0.01%) have the potential to become a a mechanism that mimics a developmental process known as
epithelial-mesenchymal transition (EMT) [2]. This process
J. C. Gomez-Gelvez (*) · D. A. Chitale is crucial for embryogenesis and wound healing; however, in
Department of Pathology and Laboratory Medicine, the context of cancer, EMT involves multiple biochemical
Henry Ford Health System, Detroit, MI, USA and morphological changes that grant cancer cells a more
e-mail: [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 29

S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_4
30 J. C. Gomez-Gelvez and D. A. Chitale

mesenchymal-like phenotype with enhanced migratory and through different intracellular signaling pathways (such as
invasive capabilities [2]. This process that is crucial for TGF-β, Wnt, and Notch pathways) [2]. Cells within the
growth and invasion at the primary site as well as at second- tumor microenvironment that play an important role in acti-
ary/metastatic sites represents a spectrum of transitional phe- vation of EMT include cancer-associated fibroblasts, tumor-
notypes between epithelial and mesenchymal characteristics associated macrophages, myeloid-derived suppressor cells,
rather than a complete progression in which cancer cell and various subtypes of T lymphocytes (i.e., CD4+ helper T
acquires one or the other [3]. Evidence indicates that EMT is cells, CD8+ cytotoxic T cells, CD4+/CD25+ regulatory T
operated as a epigenetic process that is not dependent on cells) [2]. Cancer-associated fibroblasts can secrete chemo-
somatic mutations in normal or neoplastic cells [4]. In fact, kines and growth factors such as TGF-β, IL-6, EGF, VEGF,
EMT is controlled by multiple EMT-associated transcription and HGF that have been demonstrated to activate EMT in
factors, including the zinc-finger E-box-binding homeobox cancer cells [2]. Tumor-associated macrophages activate
factors ZEB1 and ZEB2, SNAIL (also known as SNAI1), EMT via secretion of TGF-β, TNF-α, CSF1, CCL18, and
SLUG (also known as SNAI2), and the basic helix-loop-­ IL-6 [2]. Myeloid-­derived suppressor cells are a heteroge-
helix factors TWIST1 and TWIST2 [2]. These factors regu- nous group of immature cells that are recruited by primary
late the expression of each other, repress the expression of tumors and have immunosuppressive properties. These cells
genes associated to the epithelial state, and simultaneously accumulate at the invasive front where they promote activa-
induce the expression of genes associated to the mesenchy- tion of EMT via secretion of TGF-β, EGF, and HGF [11].
mal state. These changes in gene expression result in disrup- Similarly, CD8+ cytotoxic T cells and CD4+ helper T cells
tion of cell-to-cell junctions, loss of cell polarity, can induce EMT with loss of E-cadherin expression and
reorganization of cytoskeletal proteins, degradation of the gain of vimentin expression in pancreatic ductal epithelia
underlying basement membrane, and reorganization of cells and pancreatic intraepithelial neoplasia regions,
ECM. SNAIL represses CDH1, the gene encoding for epi- respectively [12]. Although the exact mechanism used by T
thelial cadherin (E-cadherin), by binding to E boxes in the lymphocytes for induction of EMT has not been elucidated,
CDH1 promoter and recruiting the polycomb repressive cytokines such as IL-6, TNF-­α, and TGF-β have been sug-
complex [2, 5, 6]. Likewise, ZEB1 represses expression of gested to play an important role [13].
CDH1 while inducing expression of genes encoding vimen- EMT is not only significant for tumor dissemination and
tin and neural cadherin (N-cadherin) through recruitment of invasion. This process has also been associated to acquisition
chromatin modifiers to their respective promoters [2, 7]. of stem cell-like properties and resistance to therapy in can-
SNAIL and ZEB1 can also repress the expression of proteins cer cells, which is significant for tumor progression and
associated to tight junctions and apical-basal polarity of epi- metastasis. EMT transcription factors can generate cancer
thelial cell such as crumbs (a transmembrane protein), stem cells with a combination of epithelial and mesenchymal
PALS1-associated tight junction protein (PATJ; a scaffolding traits suggesting that they reside at an intermediate state
protein), and lethal giant larvae (LGL; a cytoskeletal protein) within the EMT spectrum; however, the precise molecular
[2, 8]. Furthermore, SNAIL and ZEB2 induce expression of mechanisms have not been established [2]. The mechanism
matrix metalloproteinases (MMP) which are important for underlying therapy resistance involves regulation of genes
degradation of basement membranes and cell invasion [2, 9, associated with cell survival and stem cell maintenance [2].
10]. Other important proteins associated to epithelial pheno- ZEB1, SNAIL, and SLUG confer resistance to oxaliplatin
type that are downregulated by EMT transcription factors and cisplatin-based chemotherapies in breast, ovarian colon,
include occludins, claudins, α6β4integrins, cytokeratins, and and pancreatic cancers [2]. The loss of E-cadherin has been
desmoplakin [2]. Similarly, other important proteins associ- associated with resistance to growth factor and kinase inhibi-
ated to the mesenchymal phenotype that are upregulated by tors in several carcinoma types [2, 14]. The role of EMT in
EMT transcription factors include fibronectin, as well as β1 therapy resistance is further supported by the finding that
and β3 integrins [2]. Once cancer cells have acquired intrinsic molecular mechanisms capable of reversing EMT
mesenchymal-­like morphology and phenotype, they display (such as miR-200 family member which directly repress
spindle shape with front-to-back polarity and reorganized expression of ZEB1 and ZEB2) have been associated with
cytoskeleton, which provides motility and invasive reversal of resistance to therapy [2]. Interestingly, SNAIL
capabilities. expression has also been associated with resistance to immu-
EMT is a process activated by tumor microenvironment notherapy in murine models of melanoma and breast carci-
rather than by cancer cells themselves. The tumor microen- nomas [2, 15].
vironment, comprised of a mixture of stromal cells and The importance of EMT in malignant progression has
immune cells, secrete a variety of chemokines and growth recently been a matter of debate since definitive experimen-
factors that act on cancer cells to activate EMT and induce tal proof of the causal roles of EMT in metastasis formation
the expression of EMT-associated transcription factors is limited. Recent studies support the notion that EMT tran-
4 Hallmarks of Metastasis: Molecular Underpinnings 31

scription factors such as ZEB1, SNAIL, and SLUG can con- static environment in lungs and bones of patients with breast
vert non-metastatic carcinoma cells into cells capable of cancer [20]. Clinically, hypoxia is associated to HIF activa-
invading and metastasizing. However, research in lung and tion, metastasis, resistance to therapy, and poor prognosis
pancreas cancer has suggested that EMT might not be essen- [20]. Expression of HIF-1α and HIF-2α is also associated to
tial for metastasis in all types of carcinoma, although it still patient mortality [23]. Similarly, tumors with more extensive
contributes to chemoresistance [16, 17]. Notably, EMT is hypoxic and anoxic areas have higher rate of metastasis [24].
currently envisioned as a process that generates a spectrum
of cellular states lying between epithelial and mesenchymal
poles. Therefore, while some metastasizing carcinomas may 4.3 Intravasation and Circulation
exhibit overt mesenchymal properties, other cases may not
require the same EMT-associated characteristics [18]. With malignant progression, cancer cells will eventually
Further evidence is needed to clarify the role of EMT in can- invade into the lumen of either normal vasculature in sur-
cer progression and metastasis. rounding tissues or neovasculature within tumor microenvi-
Multiple genes have been reported to determine invasive ronment. This process is denominated intravasation and
potential in cancer cells, suggesting that primary tumor cells generates circulating cancer cells (CTC) that have the abil-
may exhibit a metastatic mutational profile [3]. Integrative ity to reach distant organs where they may seed new meta-
clinical genomics has shown that the most prevalent genes static colonies. Although tumor vasculature is structurally
somatically altered in metastatic cancer include TP53, and functionally abnormal, different factors contributing to
CDKN2A, PTEN, PIK3CA, and RB1 [19]. Interestingly, intravasation have been documented beyond cancer cell
putative pathogenic germline variants were present in 12.2% invasiveness. Intravasation requires at least partial degrada-
of metastasis specimens, of which 75% were related to tion of the ECM and basement membrane underlying endo-
defects in DNA repair [19]. Nonetheless, attempts to iden- thelial cells. Tumor-associated macrophages enhance
tify genes that are specifically associated with metastasis invasion of cancer cells towards vessels through EGF and
have not been successful in spite of the continuous improve- CSF1 chemotaxis whereas VEGF causes downregulation of
ment of sequencing technologies. This may suggests that endothelial cell–cell junctions, thus increasing intravasation
metastasis is driven by largely non-genetic processes such [25, 26]. Using a micro-fluidic-based approach to investi-
as EMT [2]. gate cancer cell intravasation, a study showed that the endo-
Oxygen hemostasis in tumor microenvironment influ- thelial barrier permeability to cancer cells can be modified
ences expression of genes in cancer cells that are involved in by biochemical factors, such as TNF-α, and macrophages to
biologic processes related not only to EMT but also to many facilitate intravasation and regulate the tumor–endothelial
other aspects of metastasis such as invasion, migration, cell interactions [27]. Additionally, live-cell fluorescence
intravasation, extravasation, establishment of the premeta- microcopy provided evidence for a mitosis-mediated mech-
static niche, as well as survival and growth at the metastatic anism where tumor cells located along the vessel periphery
site [20]. Hypoxia is a fairly common phenomenon seen in are able to disrupt the endothelium through cell division and
approximately 50% to 60% of solid tumors and is the conse- detach into circulation [28]. The mechanical stress imposed
quence of an imbalance between oxygen delivery (e.g., by the tumor on invading cancer cells through passive push-
abnormalities in tumor vasculature) and oxygen consump- ing and pulling forces also plays a role in intravasation [3].
tion (e.g., increased demand by proliferating cancer cells) Granulocyte macrophage colony-stimulating factor and
[20, 21]. Under hypoxic conditions, tumor microenviron- various types of cytokines, including IL-5, IL-6, IL-8,
ment employs multiple mechanisms to activate hypoxia-­ IL-1β, and IL-17, have also demonstrated a significant role
inducible factors (HIF). These factors (HIF-1, HIF-2, and on metastasis [3].
HIF-3) are transcription regulators of hundreds of genes that Cancer cells can reach systemic circulation by intravasa-
help cancer cells adapt to low oxygen tensions and promote tion into either blood vessels or lymphatics. The latter is hav-
metastasis. HIF are capable of inducing EMT directly by ing a chief role in lymph node metastasis, which is often the
regulating expression of EMT transcription factors such as first site of metastasis, and is critical for tumor staging and
ZEB1, SNAIL, and TWIST [20]. HIF can also induce EMT prognosis. For some cancers, it is the lymphatic vessels
indirectly through diverse signaling pathways including rather than blood vessels contributing to the initial dissemi-
Notch, TGF-β, integrin-linked kinase, tyrosine kinase recep- nation of cancer cells. In fact, carcinomas and melanomas
tors, Wnt, and Hedgehog [20]. AXL, a receptor tyrosine tend to develop lymph node metastases more frequently than
kinase, has been identified as a novel HIF-target driving sarcomas [29]. Such disparity may be explained by physical
EMT, invasion and metastasis [22]. Additionally, HIF signal- or structural differences between lymphatics and blood ves-
ing stimulates secretion of lysyl oxidase (LOX), LOX-like sels; however, more active recruitment, including molecular
proteins (LOXL2 and 4), and exosomes to create a premeta- mechanisms, can also play a role. One of the main structural
32 J. C. Gomez-Gelvez and D. A. Chitale

differences is that lymphatic capillaries lack the tight inter- specifically, the adhered platelets inhibit the activity of nat-
endothelial junctions present in blood vessels. Additionally, ural killer cells by secreting TGF-β and PDGF, as well as
lymphatic capillaries lack the surrounding layers of peri- by physically shielding cancer cells [18]. Platelets can also
cytes, smooth muscle cells, and basement membranes [30, alter intracellular signaling pathways of cancer cells that
31]. These differences make lymphatic capillaries an easier ultimately affect their ability to metastasize. Notably, plate-
path of intravasation for cancer cells. It has also been sug- let-derived TGF-β and direct platelet-­tumor cell contact
gested that the flow of interstitial fluid influences the move- synergistically activate the TGF-β/Smad and NF-κB path-
ment of cancer cells towards lymphatics at the periphery of ways in cancer cells. These pathways induce EMT-like
the tumor, which passively increases the chance of intravasa- changes in CTC and enhance metastasis [42].
tion into lymphatics [32, 33]. CTC also interact with neutrophils and macrophages to
Molecular mechanisms have also been shown to play a form hetero-aggregates that sustain adhesion to the endothe-
role in lymphatic intravasation of cancer cells. For example, lium and contribute to metastasis. Neutrophils can adopt
vascular endothelial growth factor receptor-3 (VEGFR3), different functional states and in certain instances they have
which is exclusively expressed on lymphatic endothelial been found to inhibit metastasis [43]. However, the predom-
cells, and its ligand vascular endothelial growth factor inant role of neutrophils appears to be the increase of meta-
(VEGFC), which is overexpressed by cancer cells, influence static potential of CTC [3, 18]. Neutrophils can prolong the
the process of lymphangiogenesis and may contribute as an retention time of CTC within the vasculature and facilitate
attracting mechanism [31]. However, some studies did not adhesive interactions of CTC with sinusoids that contributes
find a significant association between VEGFC and lymph to extravasation [18]. Neutrophils can also exert immuno-
node metastasis, which may be attributed to the probable role suppressive functions by inhibiting the actions of both cyto-
of other factors and the heterogeneity of the study samples toxic CD8+ T cells and natural killer cells [18]. Among
[34]. Active recruitment of cancer cells towards lymphatics patients with advanced breast cancer, further advancement
may also occur via chemotaxis of EGF secreted by macro- of disease was faster in individuals with CTC-neutrophils
phages, found in the proximity of lymphatics, as well as lym- clusters than those with them [44]. Additionally, disease-
phatic stromal cells. The latter have also been reported to be free mice that had been injected with CTC-­ neutrophils
a source of EGF and IGF-I [35, 36]. Similarly, secretion of showed and higher number of metastasis as compared to
chemokines CCL1, CCL19, and CCL21 by lymphatic those that were injected with CTC alone [44]. Similarly, in
­endothelial cells may attract cancer cells expressing their mice with breast cancer, eradication of neutrophils was
respective receptors CCR8 (CCL1) and CCR7 (CCL19 and associated to decreased CTC-neutrophil clusters and
CCL21), mechanism that has been associated to lymph node delayed metastasis to lungs as compared to mice with breast
metastasis [37–39]. Expression of CCR7 on melanoma cells that did not have their neutrophils depleted [44].
has been shown to increase metastasis to draining lymph
nodes, an effect that was blocked by neutralizing anti-CCL21
antibodies [40]. Expression of CCR8 on melanoma cells has 4.4 Extravasation and Colonization
also been shown to be crucial for lymph node metastasis fol-
lowing CCL1, which is detected in lymph node lymphatic Extravasations consists of CTC leaving circulation and
sinuses but not in peripheral lymphatics [37]. entering the parenchyma of an organ. Initially, CTC arrest in
Once in circulation, CTC are found either as single cells small vessels as a consequence of either physical constrains
or, less frequently, as clusters. As compared to single cells, or adhesion to the intimal layer. Cancer cells must then
clusters have an increased metastatic potential since they undergo a process known as transendothelial migration, and
are more resistant to apoptosis and are more likely to be finally penetrate the basement membrane in order to form a
trapped in small capillaries of distal organs where they may metastatic colony. In some organs, such as the liver, spleen,
initiate metastatic growth [41]. Additionally, clusters con- and bone, sinusoidal vessels are more permeable allowing
tains stromal cells and immune components from the origi- for higher rate of metastasis; however, in other organs these
nal tumor microenvironment that contribute to the vessels are more difficult to trespass, therefore being more
heterogeneity and enhance their survival as well as meta- dependent on genetic and molecular mediation for metasta-
static efficiency [3]. Once in circulation, CTC rapidly asso- sis to take place [3]. Clinical observation suggests that most
ciate with platelets, an interaction that is triggered by tissue tumors metastasize to specific rather than random organs.
factor displayed on the surface of cancer cells [18]. This This concept, known as metastatic organotropism, is influ-
interaction facilitates metastasis through formation of a enced by multiple factors such as genetic alterations in can-
coating shield of platelets around clusters, which prevents cer cells different from those that mediate malignant
detection by immune cells and provides the necessary transformation, host microenvironment and the adaptive pro-
structure to bear physical stress in circulation [3]. More cess of invading cancer cells [3].
4 Hallmarks of Metastasis: Molecular Underpinnings 33

Extravasation is a complex process that involves ligand– increased Src activity and expression of CXCR4 are able to
receptor interactions, chemokines, and circulating non-tumor activate pro-survival pathways in response to bone-derived
cells. Although the specific details of this process remain CXCL12. On activation of these pathways, cancer cells can
unclear, integrins, a key factor in nearly every step of tumor resist TNF-related apoptosis-inducing ligand (TRAIL)
progression, play an especially important role in extravasa- induced apoptosis, which is a tissue defense mechanism dis-
tion [3]. Endothelial cells in each tissue express unique tinctively expressed in bone metastasis microenvironment.
markers that CTC can recognize and adhere to in a selective Metastatic cells without Src activity would otherwise be
manner. This selectivity is thought to be among the mecha- eliminated in this microenvironment [49]. Dormant cancer
nisms leading to metastatic organotropism [45]. Both in vivo cells can also survive due to their ability to resist anoikis via
and in vitro studies show that initial attachment of cancer expression of the tyrosine kinase receptor TrkB or via non-­
cells occurs preferentially at endothelial cell junctions, fol- canonical Wnt signaling mediated by WNT2 [18].
lowed by endothelial retraction and adhesion of cancer cells Additionally, dormant cell benefit from activation of prolif-
to the underlying basement membrane [45]. Cancer cells erative signals by interacting with the ECM. This interaction
have also been demonstrated to adhere preferably to areas of is mediated by the formation of filopodium-like protrusions
inflammation, which may explain why metastasis often that are coated with integrin β1, and activates proliferation
arises at sites of tissue injury [45]. Several proteins secreted programs driven by Fak, Src, and Erk signaling [18]. As
by cancer cells, including angiopoietin-like 4 (ANGPTL4), such, combined inhibition of these pathways induced apop-
VEGF, MMP, and ADAM12, have been reported to function tosis in dormant cells and delayed metastatic colonization
as disruptors or vascular integrity as well as enhancers of [50]. Metabolic homeostasis in dormancy is reached by
both intravasation and extravasation [18]. Recruitment of downregulating two of the most important signaling path-
CCR2+ inflammatory monocytes in response to CCL2 secre- ways activated in cancer, RAS/MAPK and PI3K/AKT path-
tion by cancer or host cells can also facilitate the extravasa- ways [51].
tion and metastatic growth in lung parenchyma [46, 47]. Several dormancy-inducing signals have been described
These inflammatory monocytes may differentiate into in the microenvironment of specific organs. Members of
metastasis-­ associated macrophages, which contributes to bone morphogenetic proteins (BMP) ligand family, such as
seeding, survival, and growth of cancer cells in the lung BMP7, induce dormancy in cancer cells by activating the
parenchyma through the release of VEGF [46]. CCL2 can metastatic suppressor gene N-myc downstream-regulated
also act directly on endothelial cells to enhance vascular per- gene 1 (NDGR1) leading to activation of the p38 MAPK
meability [47]. pathway, cell cycle inhibitor p21 expression, and cell cycle
Metastatic colonization is the final step of the invasion-­ arrest [52]. Activation of the p38 MAPK pathway along with
metastasis cascade. Although a significant proportion of absence of mitogenic signals promote an ERKlow/p38high state
CTC manage to extravasate, metastatic colonization is a in cancer cells that causes arrest in the G0/G1 phase of the
highly inefficient process since the cancer cells are exposed cell cycle and associate quiescence [18, 53]. Interestingly,
to challenging conditions at the targeted organ which makes dormant cells may reside in specific niches such as the end-
survival very difficult. In fact, inefficient colonization is the osteal niche during bone metastasis, where they compete
primary reason for the inefficiency of metastasis seen in ani- with hematopoietic stem cells that normally occupy this
mal models [1]. location [54]. These specialized niches can support their sur-
The vast majority of extravasated cancer cells seem vival, restrain their proliferation, and possibly provide resis-
almost invariably destined to either be eliminated from the tance to therapy [18]. Finally, cancer cells undergoing
tissue parenchyma or enter into a state of dormancy, in which reactivation after dormancy are nursed by specialized ECM
they persist in an indolent state as single disseminated tumor niches that support positive signals, including Wnt and Notch
cells or as small micrometastasis clusters [1, 18]. Of note, signaling pathways, and attenuated negative signals, such as
dormant cells at secondary sites are thought to be the source BMP [48].
of metastasis occurring after successful initial therapy of Before the arrival of any cancer cell, the metastatic
patients with cancer [48]. Dormancy in metastatic clusters is microenvironment, known as premetastatic niche, is primed
the result of mechanisms that control the proliferation of by a number of systemic signaling tumor-derived factors
cancer cells such as the inability to induce angiogenesis and and bone marrow-derived cells to create supportive condi-
active suppression by the immune system [18]. Additionally, tions for metastatic growth. The formation of this niche
dormancy programs can be initiated by intrinsic programs in includes actions of VEGFR+ bone marrow progenitors,
cancer cells, as well as by interactions with the microenvi- myeloid-­ derived suppressor cells, or neutrophils [18].
ronment that resemble those regulating the self-renewal Additionally, cancer cells from primary tumors can secrete
capacity of adult stem cells [48]. For example, breast cancer exomes containing biologic material (DNA, messenger
cells that have lodged in the bone marrow and possess RNA, microRNA, as well as proteins such as chemokines
34 J. C. Gomez-Gelvez and D. A. Chitale

and growth factors) that can modify the host microenviron- • The premetastatic niche is primed by tumor-derived fac-
ment and prepare the premetastatic niche [3, 18]. More tors and bone marrow-derived cells to create supportive
recently, a process used to kill microbes in which neutro- conditions for metastatic growth.
phils externalize their DNA has been associated to nonin- • The vast majority of extravasated cancer cells are either
fectious inflammatory conditions like cancer and metastasis, eliminated from the tissue parenchyma or enter a state of
and observed to occur at sites where cancer cells colonize dormancy, which makes metastatic colonization a highly
more efficiently [55]. inefficient process.
Exosomes also contribute to colonization by improving • Recognizing and understanding the molecular underpin-
the metastatic properties of cancer cells. For example, malig- nings of metastasis will contribute to the development of
nant tumor cells secrete exosomes containing mRNA tran- targeted therapies capable of stopping metastasis and
scribed from genes that are involved in cell migration and tumor progression; however, the genetic mechanisms of
metastasis. These exomes are then taken up by less malig- this hallmark of cancer remain poorly understood and
nant tumor cells, located either at the same tumor or at dis- require further investigation before its complex nature can
tant tumors, which acquire enhanced migratory behavior and be entirely elucidated.
metastatic capacity [56]. Furthermore, exosomes have the
ability to remodel the ECM by interacting with fibroblasts,
stromal cells, and endothelial cells to degrade protease-­ References
associated components such as collagen, laminin, and fibro-
nectin. Exosome-altered ECM shows increased stromal cell 1. Luzzi KJ, MacDonald IC, Schmidt EE, Kerkvliet N, Morris VL,
Chambers AF, Groom AC. Multistep nature of metastatic inefficiency:
proliferation, cancer cell migration and survival, and cancer dormancy of solitary cells after successful extravasation and limited
cell resistance to apoptotic signals [3, 57]. survival of early micrometastases. Am J Pathol. 1998;153(3):865–
Tumor-derived exosomes play also an important role in 73. https://doi.org/10.1016/S0002-­9440(10)65628-­3.
metastatic organotropism because they can fuse preferen- 2. Dongre A, Weinberg RA. New insights into the mechanisms of
epithelial-mesenchymal transition and implications for cancer. Nat
tially with cells at the metastatic site (e.g., lung fibroblasts Rev Mol Cell Biol. 2019;20(2):69–84. https://doi.org/10.1038/
and epithelial cells, liver Kupffer cells, and brain endothelial s41580-­018-­0080-­4.
cells) and help prepare the premetastatic niche [58]. Exosome 3. Fares J, Fares MY, Khachfe HH, Salhab HA, Fares Y. Molecular
proteomics demonstrated distinct integrin expression pat- principles of metastasis: a hallmark of cancer revisited. Signal
Transduct Target Ther. 2020;5(1):28. https://doi.org/10.1038/
terns associated with metastasis at specific organs [58]. s41392-­020-­0134-­x.
Exosome integrins α6β4 and α6β1 were associated with lung 4. Tam WL, Weinberg RA. The epigenetics of epithelial-­mesenchymal
metastasis, while exosome ανβ5 integrin was associated to plasticity in cancer. Nat Med. 2013;19(11):1438–49. https://doi.
liver metastasis, suggesting that exosome integrins could be org/10.1038/nm.3336.
5. Cano A, Perez-Moreno MA, Rodrigo I, Locascio A, Blanco MJ,
used to predict organ-specific metastasis [58]. Exosome inte- del Barrio MG, Portillo F, Nieto MA. The transcription factor
grin uptake activated Src phosphorylation and pro-­ snail controls epithelial-mesenchymal transitions by repressing
inflammatory S100 gene expression [58]. In line with these E-cadherin expression. Nat Cell Biol. 2000;2(2):76–83. https://doi.
findings, targeting integrins α6β4 and ανβ5 decreased exo- org/10.1038/35000025.
6. Herranz N, Pasini D, Diaz VM, Franci C, Gutierrez A, Dave N,
some uptake, as well as lung and liver metastasis [58]. Escriva M, Hernandez-Munoz I, Di Croce L, Helin K, Garcia de
Herreros A, Peiro S. Polycomb complex 2 is required for E-cadherin
repression by the Snail1 transcription factor. Mol Cell Biol.
2008;28(15):4772–81. https://doi.org/10.1128/MCB.00323-­08.
4.5 Conclusions and Perspectives 7. Sanchez-Tillo E, Lazaro A, Torrent R, Cuatrecasas M, Vaquero EC,
Castells A, Engel P, Postigo A. ZEB1 represses E-cadherin and
• Metastasis is the leading cause of morbidity and mortality induces an EMT by recruiting the SWI/SNF chromatin-­remodeling
protein BRG1. Oncogene. 2010;29(24):3490–500. https://doi.
patients with cancer. org/10.1038/onc.2010.102.
• Cancer cells modify their phenotype through a mecha- 8. Spaderna S, Schmalhofer O, Wahlbuhl M, Dimmler A, Bauer K,
nism known as epithelial-mesenchymal transition (EMT) Sultan A, Hlubek F, Jung A, Strand D, Eger A, Kirchner T, Behrens
in order to acquire motility and invasive capabilities. J, Brabletz T. The transcriptional repressor ZEB1 promotes metasta-
sis and loss of cell polarity in cancer. Cancer Res. 2008;68(2):537–
• Cancer cells with invasive properties infiltrate into the 44. https://doi.org/10.1158/0008-­5472.CAN-­07-­5682.
lumen of vessels and become circulating cancer cells 9. Miyoshi A, Kitajima Y, Kido S, Shimonishi T, Matsuyama S,
(CTC) that have the ability to reach distant organs where Kitahara K, Miyazaki K. Snail accelerates cancer invasion by
they may seed metastatic colonies. upregulating MMP expression and is associated with poor prog-
nosis of hepatocellular carcinoma. Br J Cancer. 2005;92(2):252–8.
• Extravasation and transendothelial migration of CTC are https://doi.org/10.1038/sj.bjc.6602266.
complex processes that occur through selective mecha- 10. Miyoshi A, Kitajima Y, Sumi K, Sato K, Hagiwara A, Koga Y,
nisms contributing to metastatic organotropism. Miyazaki K. Snail and SIP1 increase cancer invasion by upregulat-
4 Hallmarks of Metastasis: Molecular Underpinnings 35

ing MMP family in hepatocellular carcinoma cells. Br J Cancer. 22. Rankin EB, Fuh KC, Castellini L, Viswanathan K, Finger EC, Diep
2004;90(6):1265–73. https://doi.org/10.1038/sj.bjc.6601685. AN, LaGory EL, Kariolis MS, Chan A, Lindgren D, Axelson H,
11. Toh B, Wang X, Keeble J, Sim WJ, Khoo K, Wong WC, Kato M, Miao YR, Krieg AJ, Giaccia AJ. Direct regulation of GAS6/AXL
Prevost-Blondel A, Thiery JP, Abastado JP. Mesenchymal tran- signaling by HIF promotes renal metastasis through SRC and
sition and dissemination of cancer cells is driven by myeloid-­ MET. Proc Natl Acad Sci U S A. 2014;111(37):13373–8. https://
derived suppressor cells infiltrating the primary tumor. PLoS doi.org/10.1073/pnas.1404848111.
Biol. 2011;9(9):e1001162. https://doi.org/10.1371/journal. 23. Semenza GL. Cancer-stromal cell interactions mediated by
pbio.1001162. hypoxia-inducible factors promote angiogenesis, lymphangiogen-
12. Goebel L, Grage-Griebenow E, Gorys A, Helm O, Genrich G, Lenk esis, and metastasis. Oncogene. 2013;32(35):4057–63. https://doi.
L, Wesch D, Ungefroren H, Freitag-Wolf S, Sipos B, Rocken C, org/10.1038/onc.2012.578.
Schafer H, Sebens S. CD4(+) T cells potently induce epithelial-­ 24. Noman MZ, Messai Y, Muret J, Hasmim M, Chouaib S. Crosstalk
mesenchymal-­transition in premalignant and malignant pancreatic between CTC, immune system and hypoxic tumor microenvi-
ductal epithelial cells-novel implications of CD4(+) T cells in pancre- ronment. Cancer Microenviron. 2014;7(3):153–60. https://doi.
atic cancer development. Onco Targets Ther. 2015;4(4):e1000083. org/10.1007/s12307-­014-­0157-­3.
https://doi.org/10.1080/2162402X.2014.1000083. 25. Harney AS, Arwert EN, Entenberg D, Wang Y, Guo P, Qian
13. Cohen EN, Gao H, Anfossi S, Mego M, Reddy NG, Debeb B, BZ, Oktay MH, Pollard JW, Jones JG, Condeelis JS. Real-­
Giordano A, Tin S, Wu Q, Garza RJ, Cristofanilli M, Mani SA, time imaging reveals local, transient vascular permeability, and
Croix DA, Ueno NT, Woodward WA, Luthra R, Krishnamurthy S, tumor cell Intravasation stimulated by TIE2hi macrophage-­
Reuben JM. Inflammation mediated metastasis: immune induced derived VEGFA. Cancer Discov. 2015;5(9):932–43. https://doi.
epithelial-to-mesenchymal transition in inflammatory breast cancer org/10.1158/2159-­8290.CD-­15-­0012.
cells. PLoS One. 2015;10(7):e0132710. https://doi.org/10.1371/ 26. Roussos ET, Condeelis JS, Patsialou A. Chemotaxis in cancer. Nat
journal.pone.0132710. Rev Cancer. 2011;11(8):573–87. https://doi.org/10.1038/nrc3078.
14. Singh A, Settleman J. EMT, cancer stem cells and drug resis- 27. Zervantonakis IK, Hughes-Alford SK, Charest JL, Condeelis JS,
tance: an emerging axis of evil in the war on cancer. Oncogene. Gertler FB, Kamm RD. Three-dimensional microfluidic model for
2010;29(34):4741–51. https://doi.org/10.1038/onc.2010.215. tumor cell intravasation and endothelial barrier function. Proc Natl
15. Kudo-Saito C, Shirako H, Takeuchi T, Kawakami Y. Cancer metas- Acad Sci U S A. 2012;109(34):13515–20. https://doi.org/10.1073/
tasis is accelerated through immunosuppression during snail-­ pnas.1210182109.
induced EMT of cancer cells. Cancer Cell. 2009;15(3):195–206. 28. Wong AD, Searson PC. Mitosis-mediated intravasation in a
https://doi.org/10.1016/j.ccr.2009.01.023. tissue-engineered tumor-microvessel platform. Cancer Res.
16. Fischer KR, Durrans A, Lee S, Sheng J, Li F, Wong ST, Choi H, 2017;77(22):6453–61. https://doi.org/10.1158/0008-­5472.
El Rayes T, Ryu S, Troeger J, Schwabe RF, Vahdat LT, Altorki CAN-­16-­3279.
NK, Mittal V, Gao D. Epithelial-to-mesenchymal transition is not 29. Friedl P, Wolf K. Tumour-cell invasion and migration: diversity and
required for lung metastasis but contributes to ­chemoresistance. escape mechanisms. Nat Rev Cancer. 2003;3(5):362–74. https://
Nature. 2015;527(7579):472–6. https://doi.org/10.1038/ doi.org/10.1038/nrc1075.
nature15748. 30. Alitalo K, Carmeliet P. Molecular mechanisms of lymphangiogen-
17. Zheng X, Carstens JL, Kim J, Scheible M, Kaye J, Sugimoto H, esis in health and disease. Cancer Cell. 2002;1(3):219–27. https://
Wu CC, LeBleu VS, Kalluri R. Epithelial-to-mesenchymal transi- doi.org/10.1016/s1535-­6108(02)00051-­x.
tion is dispensable for metastasis but induces chemoresistance in 31. Wong SY, Hynes RO. Lymphatic or hematogenous dissemi-
pancreatic cancer. Nature. 2015;527(7579):525–30. https://doi. nation: how does a metastatic tumor cell decide? Cell Cycle.
org/10.1038/nature16064. 2006;5(8):812–7. https://doi.org/10.4161/cc.5.8.2646.
18. Lambert AW, Pattabiraman DR, Weinberg RA. Emerging biologi- 32. Aleskandarany MA, Sonbul SN, Mukherjee A, Rakha
cal principles of metastasis. Cell. 2017;168(4):670–91. https://doi. EA. Molecular mechanisms underlying lymphovascular invasion in
org/10.1016/j.cell.2016.11.037. invasive breast cancer. Pathobiology. 2015;82(3–4):113–23. https://
19. Freedland SJ, Aronson WJ. Commentary on “Integrative clinical doi.org/10.1159/000433583.
genomics of advanced prostate cancer”. Robinson D, Van Allen 33. Dadiani M, Kalchenko V, Yosepovich A, Margalit R, Hassid Y,
EM, Wu YM, Schultz N, Lonigro RJ, Mosquera JM, Montgomery Degani H, Seger D. Real-time imaging of lymphogenic metastasis
B, Taplin ME, Pritchard CC, Attard G, Beltran H, Abida W, Bradley in orthotopic human breast cancer. Cancer Res. 2006;66(16):8037–
RK, Vinson J, Cao X, Vats P, Kunju LP, Hussain M, Feng FY, 41. https://doi.org/10.1158/0008-­5472.CAN-­06-­0728.
Tomlins SA, Cooney KA, Smith DC, Brennan C, Siddiqui J, Mehra 34. Ran S, Volk L, Hall K, Flister MJ. Lymphangiogenesis and lymphatic
R, Chen Y, Rathkopf DE, Morris MJ, Solomon SB, Durack JC, metastasis in breast cancer. Pathophysiology. 2010;17(4):229–51.
Reuter VE, Gopalan A, Gao J, Loda M, Lis RT, Bowden M, Balk https://doi.org/10.1016/j.pathophys.2009.11.003.
SP, Gaviola G, Sougnez C, Gupta M, Yu EY, Mostaghel EA, Cheng 35. LeBedis C, Chen K, Fallavollita L, Boutros T, Brodt P. Peripheral
HH, Mulcahy H, True LD, Plymate SR, Dvinge H, Ferraldeschi lymph node stromal cells can promote growth and tumorigenicity
R, Flohr P, Miranda S, Zafeiriou Z, Tunariu N, Mateo J, Perez-­ of breast carcinoma cells through the release of IGF-I and EGF. Int
Lopez R, Demichelis F, Robinson BD, Schiffman M, Nanus DM, J Cancer. 2002;100(1):2–8. https://doi.org/10.1002/ijc.10481.
Tagawa ST, Sigaras A, Eng KW, Elemento O, Sboner A, Heath EI, 36. Schoppmann SF, Birner P, Stockl J, Kalt R, Ullrich R, Caucig C,
Scher HI, Pienta KJ, Kantoff P, de Bono JS, Rubin MA, Nelson Kriehuber E, Nagy K, Alitalo K, Kerjaschki D. Tumor-­associated mac-
PS, Garraway LA, Sawyers CL, Chinnaiyan AM. Cell. 21 May rophages express lymphatic endothelial growth factors and are related
2015;161(5):1215–1228. Urol Oncol. 2017;35(8):535. https://doi. to peritumoral lymphangiogenesis. Am J Pathol. 2002;161(3):947–
org/10.1016/j.urolonc.2017.05.010. 56. https://doi.org/10.1016/S0002-­9440(10)64255-­1.
20. Rankin EB, Giaccia AJ. Hypoxic control of metastasis. Science. 37. Das S, Sarrou E, Podgrabinska S, Cassella M, Mungamuri SK, Feirt
2016;352(6282):175–80. https://doi.org/10.1126/science.aaf4405. N, Gordon R, Nagi CS, Wang Y, Entenberg D, Condeelis J, Skobe
21. Vaupel P, Mayer A. Hypoxia in cancer: significance and impact M. Tumor cell entry into the lymph node is controlled by CCL1
on clinical outcome. Cancer Metastasis Rev. 2007;26(2):225–39. chemokine expressed by lymph node lymphatic sinuses. J Exp Med.
https://doi.org/10.1007/s10555-­007-­9055-­1. 2013;210(8):1509–28. https://doi.org/10.1084/jem.20111627.
36 J. C. Gomez-Gelvez and D. A. Chitale

38. Homey B, Muller A, Zlotnik A. Chemokines: agents for the immu- tied to Src-dependent survival signals. Cancer Cell. 2009;16(1):67–
notherapy of cancer? Nat Rev Immunol. 2002;2(3):175–84. https:// 78. https://doi.org/10.1016/j.ccr.2009.05.017.
doi.org/10.1038/nri748. 50. El Touny LH, Vieira A, Mendoza A, Khanna C, Hoenerhoff MJ,
39. Korbecki J, Grochans S, Gutowska I, Barczak K, Baranowska-­ Green JE. Combined SFK/MEK inhibition prevents metastatic out-
Bosiacka I. CC chemokines in a tumor: a review of pro-cancer and growth of dormant tumor cells. J Clin Invest. 2014;124(1):156–68.
anti-cancer properties of receptors CCR5, CCR6, CCR7, CCR8, https://doi.org/10.1172/JCI70259.
CCR9, and CCR10 ligands. Int J Mol Sci. 2020;21(20) https://doi. 51. Yeh AC, Ramaswamy S. Mechanisms of cancer cell dormancy—
org/10.3390/ijms21207619. another Hallmark of cancer? Cancer Res. 2015;75(23):5014–22.
40. Wiley HE, Gonzalez EB, Maki W, Wu MT, Hwang ST. Expression https://doi.org/10.1158/0008-­5472.CAN-­15-­1370.
of CC chemokine receptor-7 and regional lymph node metastasis of 52. Kobayashi A, Okuda H, Xing F, Pandey PR, Watabe M, Hirota S,
B16 murine melanoma. J Natl Cancer Inst. 2001;93(21):1638–43. Pai SK, Liu W, Fukuda K, Chambers C, Wilber A, Watabe K. Bone
https://doi.org/10.1093/jnci/93.21.1638. morphogenetic protein 7 in dormancy and metastasis of prostate
41. Aceto N, Bardia A, Miyamoto DT, Donaldson MC, Wittner BS, cancer stem-like cells in bone. J Exp Med. 2011;208(13):2641–55.
Spencer JA, Yu M, Pely A, Engstrom A, Zhu H, Brannigan BW, https://doi.org/10.1084/jem.20110840.
Kapur R, Stott SL, Shioda T, Ramaswamy S, Ting DT, Lin CP, 53. Sosa MS, Avivar-Valderas A, Bragado P, Wen HC, Aguirre-Ghiso
Toner M, Haber DA, Maheswaran S. Circulating tumor cell clus- JA. ERK1/2 and p38alpha/beta signaling in tumor cell quiescence:
ters are oligoclonal precursors of breast cancer metastasis. Cell. opportunities to control dormant residual disease. Clin Cancer
2014;158(5):1110–22. https://doi.org/10.1016/j.cell.2014.07.013. Res. 2011;17(18):5850–7. https://doi.org/10.1158/1078-­0432.
42. Labelle M, Begum S, Hynes RO. Direct signaling between platelets CCR-­10-­2574.
and cancer cells induces an epithelial-mesenchymal-like transition 54. Shiozawa Y, Pedersen EA, Havens AM, Jung Y, Mishra A, Joseph
and promotes metastasis. Cancer Cell. 2011;20(5):576–90. https:// J, Kim JK, Patel LR, Ying C, Ziegler AM, Pienta MJ, Song J, Wang
doi.org/10.1016/j.ccr.2011.09.009. J, Loberg RD, Krebsbach PH, Pienta KJ, Taichman RS. Human
43. Granot Z, Henke E, Comen EA, King TA, Norton L, Benezra prostate cancer metastases target the hematopoietic stem cell
R. Tumor entrained neutrophils inhibit seeding in the premetastatic niche to establish footholds in mouse bone marrow. J Clin Invest.
lung. Cancer Cell. 2011;20(3):300–14. https://doi.org/10.1016/j. 2011;121(4):1298–312. https://doi.org/10.1172/JCI43414.
ccr.2011.08.012. 55. Cedervall J, Zhang Y, Olsson AK. Tumor-induced NETosis
44. Szczerba BM, Castro-Giner F, Vetter M, Krol I, Gkountela S, as a risk factor for metastasis and organ failure. Cancer Res.
Landin J, Scheidmann MC, Donato C, Scherrer R, Singer J, Beisel 2016;76(15):4311–5. https://doi.org/10.1158/0008-­5472.
C, Kurzeder C, Heinzelmann-Schwarz V, Rochlitz C, Weber WP, CAN-­15-­3051.
Beerenwinkel N, Aceto N. Neutrophils escort circulating tumour 56. Zomer A, Maynard C, Verweij FJ, Kamermans A, Schafer R,
cells to enable cell cycle progression. Nature. 2019;566(7745):553– Beerling E, Schiffelers RM, de Wit E, Berenguer J, Ellenbroek
7. https://doi.org/10.1038/s41586-­019-­0915-­y. SIJ, Wurdinger T, Pegtel DM, van Rheenen J. In vivo imaging
45. Welch DR, Hurst DR. Defining the hallmarks of metastasis. Cancer reveals extracellular vesicle-mediated phenocopying of metastatic
Res. 2019;79(12):3011–27. https://doi.org/10.1158/0008-­5472. behavior. Cell. 2015;161(5):1046–57. https://doi.org/10.1016/j.
CAN-­19-­0458. cell.2015.04.042.
46. Qian BZ, Li J, Zhang H, Kitamura T, Zhang J, Campion LR, 57. Tarbe N, Losch S, Burtscher H, Jarsch M, Weidle UH. Identification
Kaiser EA, Snyder LA, Pollard JW. CCL2 recruits inflamma- of rat pancreatic carcinoma genes associated with lymphogenous
tory monocytes to facilitate breast-tumour metastasis. Nature. metastasis. Anticancer Res. 2002;22(4):2015–27.
2011;475(7355):222–5. https://doi.org/10.1038/nature10138. 58. Hoshino A, Costa-Silva B, Shen TL, Rodrigues G, Hashimoto
47. Wolf MJ, Hoos A, Bauer J, Boettcher S, Knust M, Weber A, A, Tesic Mark M, Molina H, Kohsaka S, Di Giannatale A, Ceder
Simonavicius N, Schneider C, Lang M, Sturzl M, Croner RS, S, Singh S, Williams C, Soplop N, Uryu K, Pharmer L, King T,
Konrad A, Manz MG, Moch H, Aguzzi A, van Loo G, Pasparakis Bojmar L, Davies AE, Ararso Y, Zhang T, Zhang H, Hernandez J,
M, Prinz M, Borsig L, Heikenwalder M. Endothelial CCR2 signal- Weiss JM, Dumont-Cole VD, Kramer K, Wexler LH, Narendran
ing induced by colon carcinoma cells enables extravasation via the A, Schwartz GK, Healey JH, Sandstrom P, Labori KJ, Kure EH,
JAK2-Stat5 and p38MAPK pathway. Cancer Cell. 2012;22(1):91– Grandgenett PM, Hollingsworth MA, de Sousa M, Kaur S, Jain M,
105. https://doi.org/10.1016/j.ccr.2012.05.023. Mallya K, Batra SK, Jarnagin WR, Brady MS, Fodstad O, Muller V,
48. Giancotti FG. Mechanisms governing metastatic dormancy and Pantel K, Minn AJ, Bissell MJ, Garcia BA, Kang Y, Rajasekhar VK,
reactivation. Cell. 2013;155(4):750–64. https://doi.org/10.1016/j. Ghajar CM, Matei I, Peinado H, Bromberg J, Lyden D. Tumour
cell.2013.10.029. exosome integrins determine organotropic metastasis. Nature.
49. Zhang XH, Wang Q, Gerald W, Hudis CA, Norton L, Smid M, 2015;527(7578):329–35. https://doi.org/10.1038/nature15756.
Foekens JA, Massague J. Latent bone metastasis in breast cancer
 Hereditary Cancer Syndromes
and Cancer Metastasis 5
Brandon M. Shaw and Olena Kis

Abstract 5.1 Introduction

Hereditary cancer syndromes, also known as familial or
inherited cancer syndromes, are inherited genetic disor- Hereditary cancer syndromes are caused by inherited patho-
ders that are associated with increased risk for certain genic variants that lead to an increased risk for cancer in the
types of cancer. Approximately 5–10% of all cancers individuals carrying these anomalies. Retinoblastoma is the
result from an inherited genetic change (mutation), a best studied hereditary cancer syndrome that Alfred
defect in a gene that has been passed on from generation G. Knudson first described in 1971. Knudson noted a differ-
to generation within a family. Individuals and families ence in the age of onset in individuals with the inherited form
affected with a hereditary cancer syndrome typically of retinoblastoma versus those with the sporadic form of the
present with a specific pattern of cancer within a family, disease. This observation led to the development of the two-­
such as the same type of cancer in close relatives, cancer hit hypothesis of tumorigenesis. The gene responsible for
diagnoses at unusually young ages, or multiple different hereditary retinoblastoma, RB1, was discovered in 1986.
cancers in the same person. This chapter provides a brief Inherited genetic pathogenic variants account for 5–10%
overview of hereditary cancer syndromes, their genetic of all cancer cases. Most hereditary cancers are associated
causes, and clinical presentation, with a specific focus on with a germline genetic alteration that was inherited from
the role of cancer metastasis in hereditary cancer syn- one of the parents via sperm or egg cell and is consequently
dromes such as Hereditary Breast and Ovarian Cancer present in every cell in the body, making this individual a
(HBOC) syndrome, Lynch syndrome, and Multiple carrier of a pathogenic (defective) allele in a gene that plays
Endocrine Neoplasia. In addition, this chapter reviews a vital role in protecting against cancer (tumor suppressor
emerging knowledge on the role of the genetic back- gene). Typically, one defective copy of the gene is insuffi-
ground in the development of metastatic disease and the cient to cause cancer, but it makes the individual more sus-
contribution of specific genes and genetic susceptibility ceptible to developing cancer in the future (increased risk for
loci to tumor progression and the risk for metastasis. cancer). Suppose any cell in the body loses the second copy
of this tumor suppressor gene by acquiring a new genetic
alteration (somatic mutation) or loss of the normal allele. In
Learning Objectives that case, this constitutes the “second hit” resulting in loss of
• To understand the differences between hereditary both copies of the tumor suppressor gene in this cell and
and nonhereditary cancers. making it permissive to cancerous growth. The type of cell
• To learn the main clinical features of several major that acquires this “second hit” somatic mutation will define
hereditary cancer syndromes. the type of cancer this individual will develop. Since the
• To understand the impact of inherited genetic muta- inherited genetic variant is present in every cell in the body,
tions on cancer metastasis. carrier individuals are at risk of having multiple types of can-
cer in their lifetime. They are more likely to develop cancer
earlier in life than noncarriers who require two sets of
acquired mutations to achieve a similar biallelic loss of this
B. M. Shaw (*) · O. Kis tumor suppressor gene.
Department of Pathology and Laboratory Medicine, Henry Ford In addition to the inherited mutations in high-penetrance
Health System, Detroit, MI, USA cancer predisposition genes that directly increase the risk for
e-mail: [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 37

S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_5
38 B. M. Shaw and O. Kis

cancer, recent studies have begun to examine the impact of genetic testing of DNA isolated from peripheral blood speci-
other germline mutations in low-penetrance genes or metas- men to identify a deleterious (pathogenic) mutation that can
tasis susceptibility loci on tissue susceptibility to metastasis. explain the pattern of cancer observed in the family. Since
The emerging data suggest that germline mutations play a malignancies are prevalent in the population, clinical criteria
much larger role in tumorigenesis than simply acting as pre- for genetic testing for hereditary cancer syndromes have
dispositions and that inherited germline mutations not only been established to help guide the testing and ensure that the
increase cancer risk susceptibility but also act as modulators indication for molecular genetic analysis is carefully
when shaping tumor evolution. Although the penetrance of considered.
other germline genetic defects is very small, collectively, Many genes associated with hereditary cancer predisposi-
these small changes may explain the apparent heritability in tion are involved in DNA repair, cell proliferation, cell death,
tumor progression and susceptibility to metastasis influenced or signaling pathways. Due to the variation in the roles of
by the genetic background of the individual. This book chap- these different genes, each cancer syndrome is different in
ter will review the impact of inherited genetic variants on the mechanisms leading to malignant transformation and dif-
tumor microenvironment, transcriptional differences, splice fers in the types of cancers that are likely to occur (Fig. 5.1).
isoforms and expression profiles in various tumors, and In addition, each hereditary cancer syndrome has a particular
tumor immune cell infiltration, and their effect on cancer characteristic spectrum of tumors. To date, there are ~50
progression and metastasis. hereditary cancer syndromes described in the literature, and
presumably, other genes associated with cancer predisposi-
tion are yet to be discovered. Table 5.1 provides an overview
5.2 Hereditary Cancer Syndromes of the selected hereditary cancer syndromes with known
genetic causes, have a well-characterized spectrum of clini-
Patients who are found to be carriers of a germline patho- cal features, and are associated with moderate to high risk for
genic mutation in a cancer predisposition gene are diagnosed developing specific types of cancer (high penetrance). Some
with hereditary cancer syndrome (see Table 5.1 for examples of the more common syndromes linked to high risk for devel-
of genes and their associated Hereditary Cancer Syndromes). oping cancer include hereditary breast and ovarian cancer
These individuals are at high risk for developing multiple syndrome, Lynch syndrome, also known as hereditary non-­
types of cancer and often at a younger age. They may benefit polyposis colorectal cancer (HNPCC) syndrome, and the
from more frequent cancer screening to help identify future multiple endocrine neoplasia type 1 and 2. These syndromes
cancer at an early stage, allow earlier targeted treatment of will be discussed in more detail, highlighting their genetic
second cancer, and provide additional risk-reducing surgery underpinnings, mechanisms of cancer development and pro-
options or other clinical management options. In addition, gression, and the role of cancer metastasis in disease pro-
some syndromes have additional clinical manifestations gression and treatment.
unrelated to cancer (e.g., cardiac defects, immunodeficiency,
diabetes, or hearing loss) that may require medical treatment
or increased screening. Once the individual is diagnosed 5.2.1  ereditary Breast and Ovarian Cancer
H
with a genetic syndrome, testing close relatives for the iden- Syndrome
tified pathogenic familial variant can also help identify other
at-risk family members. The World Health Organization lists breast cancer as the
Some hereditary cancer syndromes can be diagnosed most common cancer in women with 34% of new diagnoses
without genetic testing, for example, establishing the diag- occurring after cancer has disseminated to other locations
nosis of familial adenomatous polyposis based on the endo- within the body. Metastasis in breast cancer is responsible
scopic findings or presence of bone marrow failure and for most of the deaths from the disease. Because of this, a
chromosome breakage in individuals with Fanconi anemia. better understanding of the individual genetic background
However, other cancer predisposition syndromes do not have and the role it plays in the development of metastatic disease
any distinctive clinical manifestations and have a nonspecific is expected to provide important diagnostic and prognostic
pattern of cancer in the family that can be explained by mul- information and aid in the management of an individual’s
tiple genetic and nongenetic causes. For example, a family treatment regimen. While death rates have been declining
history of breast, ovarian, and prostate cancer can be linked due to improved treatments, better screening, and early
to several different hereditary cancer syndromes (see detection, breast cancer is still one of the leading causes of
Table 5.1) or may be due to environmental causes or sporadic death in women. Hereditary breast and ovarian cancer syn-
mutations (acquired). In many cases, diagnosis of a cancer drome (HBOC) is characterized by an increased risk for
predisposition syndrome requires multigene sequencing female breast cancer (40–80%), male breast cancer (1–10%),
using germline DNA from noncancerous cells such as ovarian cancer (up to 40%), and to a lesser extent several
5 Hereditary Cancer Syndromes and Cancer Metastasis 39

Table 5.1 Overview of selected highly penetrant hereditary cancer syndromes

Population
Hereditary cancer syndrome Gene(s) Inheritance incidence Associated cancer risk and other clinical features
Ataxia-Telangiectasia (OMIM# ATM AR 1/40,000– Increased risk for leukemia (ALL), lymphoma
208900) 1/100,000 (non-Hodgkin lymphoma), and breast cancer.
Other clinical features: Progressive cerebellar
ataxia from early childhood, telangiectasias, short
stature, frequent infections, immunodeficiency
Bloom Syndrome (OMIM# BLM AR 1/48,000 High risk for multiple malignancies (leukemias
210900) (AML, ALL), lymphoma, oropharyngeal cancers,
colorectal cancer, esophageal and gastric cancer,
breast cancer, genitourinary cancers, skin cancer,
and Wilms tumor. Other clinical features: Severe
pre- and postnatal growth deficiency, short
stature, immune abnormalities, sensitivity to
sunlight, telangiectases, cutaneous abnormalities
(café au lait macules, hyper- or
hypopigmentation), insulin resistance
Cowden Syndrome/PTEN PTEN AD 1/200,000 Increased risk for breast, thyroid, endometrial,
Hamartoma Tumor Syndrome and renal cancer. Elevated risk for colorectal
(OMIM# 158350) cancer (up to 9%) and melanoma. Other clinical
features: Macrocephaly, hamartomatous
overgrowth of tissues, skin hamartomas, uterine
fibroids, breast fibrosis, lipomas,
ganglioneuromas, trichilemmomas, papillomatous
papules, connective tissue, and epidermal nevi.
Familial Adenomatous Polyposis APC AD 1/5000– Multiple adenomatous colon polyps and
(FAP) and Attenuated FAP 1/33,000 increased risk for colorectal cancer (up to 80%),
(OMIM# 175100) duodenal polyposis and cancer, increased risk for
thyroid cancer (up to 12%), pancreatic cancer,
hepatoblastoma, CNS tumors (medulloblastoma),
gastric polyps and cancer. Other clinical features:
Osteomas, dental abnormalities, benign
cutaneous lesions, lipomas, desmoid tumors
Fanconi anaemia FANCA, FANCB, FANCC, AR, XLR, 1/100,000– Bone marrow failure and increased risk for
(OMIM# 605724, 227650, BRCA2 (FANCD1), AD 1/160,000 malignancy at a young age, most frequently
615272, 300514, 227645, FANCD2, FANCE, FANCF, leukemia (AML), MDS, solid tumors of the head
227646, 600185, 608111, FANCG, FANCI, BRIP1 and neck, esophagus, liver, and cervix. Other
603467, 609644, 609054, (FANCDJ), FANCL, clinical features: Physical abnormalities (short
610832, 600901, 614083, FANCM, PALB2 (FANCN), stature, abnormal skin pigmentation, skeletal
613390, 614082, 609053, 61951, RAD51C (FANCO), SLX4 malformations, microcephaly, others), cutaneous
617883, others) (FANCP), ERCC4 (FANCQ), abnormalities (café au lait macules, hyper- or
BRCA1 (FANCS) hypopigmentation).
Hereditary breast and ovarian BRCA1, BRCA2 AD 1/500– Increased risk for breast cancer in females (up to
cancer (HBOC) syndrome 1/1000 87%) and males (1–9%), ovarian cancer (up to
(OMIM# 113705, 600185) 54%), including fallopian tube and primary
peritoneal cancers, prostate cancer (up to 20%),
pancreatic cancer (1–7%), and elevated risk for
melanoma (cutaneous and ocular), cervical and
uterine cancer.
Hereditary Diffuse Gastric CDH1 AD Very rare Increased risk for gastric cancer (up to 80%),
Cancer (OMIM# 137215) lobular breast cancer (39–52%), and colorectal
cancer.
(continued)
40 B. M. Shaw and O. Kis

Table 5.1 (continued)

Population
Hereditary cancer syndrome Gene(s) Inheritance incidence Associated cancer risk and other clinical features
Hereditary Paraganglioma-­ SDHB, SDHD, SDHC, AD 1/1000,000 Characterized by benign or malignant
Pheochromocytoma (PGL/PCC) SDHAF2, SDHA, MAX, paragangliomas (neuroendocrine tissue tumors)
syndromes (OMIM# 115310, TMEM127 and pheochromocytomas (paragangliomas that
168000, 605373, 601650, are confined to the adrenal medulla). Extra-­
614165, 171300) adrenal sympathetic paragangliomas have an
increased likelihood of malignant transformation
and metastasis. Risk for developing other cancer
types including GIST, pulmonary chondromas,
renal clear cell carcinoma, papillary thyroid
carcinoma, pituitary adenomas, and
neuroendocrine tumors.
Juvenile Polyposis Syndrome, SMAD4, BMPR1A AD 1/100,000 Multiple hamartomatous polyps in the GI tract,
JPS (OMIM# 174900) increased risk for colorectal cancer (up to 70%)
Juvenile polyposis syndrome/ and gastric cancer (up to 20%). Other clinical
hereditary hemorrhagic features: Gastrointestinal bleeding, anemia.
telangiectasia, JPS/HHT Individuals with JPS/HHT dues to SMAD4
(OMIM# 175050) mutation have additional HHT-related
abnormalities such as epistaxis, telangiectases,
heart or brain abnormalities, arteriovenous
malformations, intestinal malrotation, digital
clubbing, and anomalies of the genitalia or
urinary tract.
Li-Fraumeni syndrome (OMIM# TP53 AD 1/5000– Lifetime risk for cancer >90% in women and
151623) 1/20,000 >70% in men. Most common cancer types: Breast
cancer (up to 80%), soft-tissue sarcomas,
osteosarcomas, adrenocortical carcinomas, CNS
tumors (glioblastomas, astrocytomas), leukemias
(ALL, AML), MDS, lymphoma, and GI tract
cancers.
Lynch syndrome (Hereditary MLH1, MSH2, MSH6, AD 1/400 Increased risk for colorectal cancer (up to 80%),
non-polyposis colon cancer) PMS2, EPCAM deletion endometrial cancer (up to 55%), ovarian cancer
(OMIM# 609310, 120,435, (up to 25%), stomach and small bowel cancer,
614,337, 614,350) urinary tract cancers (ureter, renal pelvis,
bladder), biliary tract cancer, pancreatic cancer,
prostate cancer (up to 30%), CNS tumors
(glioblastoma), breast cancer, and sebaceous
neoplasms. Other clinical features: Sebaceous
adenoma, tumors with microsatellite instability
(MSI).
Multiple endocrine neoplasia MEN1 AD 1/30,000 Increased risk for tumors of the parathyroid gland
type 1, MEN1 (OMIM# 131100) (up to 100%), pituitary gland (30–40%),
well-differentiated neuroendocrine tumors of the
pancreas and the GI tract (up to 70%), elevated
risk for carcinoid tumors, adrenocortical tumors,
and non-endocrine tumors (facial angiofibromas,
collagenomas, lipomas, meningiomas,
ependymomas, and leiomyomas)
Multiple endocrine neoplasia RET AD 1/35,000 Increased risk for medullary thyroid carcinoma
type 2, MEN2A/FMTC (OMIM# (up to 100%), pheochromocytoma (cancer of the
171400, 162,300) adrenal gland), parathyroid adenoma/hyperplasia,
hyperparathyroidism
MUTYH-Associated Polyposis MUTYH AR 1/7000– Multiple adenomatous colon polyps and
(OMIM# 608456) 1/22,000 increased risk for colorectal cancer (up to 90%),
duodenal adenomas and cancer, ovarian cancer
(6–14%), bladder cancer, breast cancer (up to
25%), elevated risk for thyroid, pancreatic,
gastric, and skin cancer.
5 Hereditary Cancer Syndromes and Cancer Metastasis 41

Table 5.1 (continued)

Population
Hereditary cancer syndrome Gene(s) Inheritance incidence Associated cancer risk and other clinical features
Neurofibromatosis type 1 NF1 AD 1/3000 Increased risk for optic nerve and other CNS
(OMIM# 162200) gliomas, malignant peripheral nerve sheath
tumors. Other clinical features: Multiple café au
lait spots, axillary and inguinal freckling,
multiple benign cutaneous neurofibromas,
plexiform neurofibromas, iris Lisch nodules,
choroidal freckling, musculoskeletal
abnormalities (scoliosis, tibial dysplasia),
vasculopathy and learning disabilities.
Neurofibromatosis type 2 NF2 AD 1/33,000– Characterized by the growth of noncancerous
(OMIM# 101000) 1/40,000 tumors in the nervous system, most frequently
bilateral vestibular schwannomas or acoustic
neuromas. Risk for schwannomas of other cranial
and peripheral nerves, meningiomas,
ependymomas, glioma and astrocytomas. Other
clinical features: Skin plaques at birth,
meningioma, retinal hamartoma,
mononeuropathy, neurofibroma, cortical wedge
cataracts, unilateral or bilateral hearing loss,
balance dysfunction, seizures.
Nevoid Basal Cell Carcinoma or PTCH1, SUFU AD 1/31,000– High risk for developing multiple basal cell
Gorlin syndrome (OMIM# 1/250,000 carcinomas (the most common form of skin
109400) cancer) and several other types of tumors
including benign jaw keratocysts, malignant brain
tumor (medulloblastoma), and cardiac and
ovarian fibromas. Other clinical features:
Congenital malformations and bone anomalies,
macrocephaly, distinct facial features, pits on the
palms of the hands or soles of the feet.
Nijmegen Breakage Syndrome NBN AR 1/100,000 Increased risk for cancer at a young age, most
(OMIM# 251260) frequently non-Hodgkin’s lymphoma. Risk for
other cancer types including brain tumors
(medulloblastoma, glioma) and muscle tissue
tumors (rhabdomyosarcoma). Other clinical
features: Microcephaly, growth retardation,
distinctive craniofacial features, respiratory tract
infections, immunodeficiency, intellectual
disability.
Peutz-Jeghers syndrome (PJS) STK11 AD 1/25,000– Characterized by the development of
(OMIM# 175200) 1/300,000 noncancerous hamartomatous polyps (Peutz-­
Jeghers-­type) in the gastrointestinal tract, most
commonly in the small intestine. Increased risk
for colorectal cancer (up to 40%), gastric cancer
(up to 30%), pancreatic cancer (11–36%), small
intestinal cancer (13%), early-onset breast cancer
(32–54%), ovarian sex cord tumors with annular
tubules (up to 20%), cervix and uterine cancer
(up to 10%), lung cancer (7–17%), and testicular
cancer (9%). Other clinical features: Melanocytic
macules (dark brown/dark blue).
Retinoblastoma, Familial RB1 AD 1/15,00– Characterized by malignant tumor of the
(OMIM# 180200) 1/25,000 developing retina that occurs in children at a
young age. Increased risk of developing bilateral
retinoblastoma and non-ocular tumors including
cancer of the pineal gland in the brain
(pineoblastoma), bone cancer (osteosarcomas),
soft tissue sarcomas (leiomyosarcomas and
rhabdomyosarcomas), and melanomas.
(continued)
42 B. M. Shaw and O. Kis

Table 5.1 (continued)

Population
Hereditary cancer syndrome Gene(s) Inheritance incidence Associated cancer risk and other clinical features
Tuberous sclerosis (OMIM# TSC1, TSC2 AD 1/6000 Characterized by benign tumors in many parts of
191100, 191,092) the body including skin, brain, kidneys, and other
organs. Increased risk for several types of cancer
including CNS tumors (subependymal giant cell
astrocytomas) and renal cell carcinomas as well
as benign tumors (cardiac rhabdomyomas and
renal angiomyolipomas). Other clinical features:
Abnormalities of the skin (hypomelanotic
macules, confetti skin lesions, facial
angiofibromas, shagreen patches, fibrous cephalic
plaques, ungual fibromas), renal cysts, brain
anomalies (subependymal nodules, cortical
dysplasias), heart arrhythmias, eye abnormalities
(retinal lesions), seizures, intellectual disability,
developmental delay, and psychiatric illness.
Von Hippel-Lindau syndrome VHL AD 1/36,000 Increased risk for a variety of malignant and
(OMIM# 193300) benign neoplasms, most frequently retinal,
cerebellar and spinal hemangioblastoma, renal
cell carcinoma (RCC), pancreatic cancer,
pheochromocytoma (adrenal gland), and tumors
of the reproductive tract. Other clinical features:
Cysts in many different parts of the body
Abbreviations: AD Autosomal dominant, ALL Acute lymphoblastic leukemia, AML Acute myeloid leukemia, AR Autosomal recessive, CNS
Central nervous system, GI Gastrointestinal, GIST Gastrointestinal stromal tumors, MDS Myelodysplastic syndromes, XLR X-linked recessive
Data Sources: OMIM®, Online Mendelian Inheritance in Man (https://www.omim.org/); GeneReviews (https://www.ncbi.nlm.nih.gov/books/
NBK1116/); NIH—Genetic and Rare Diseases Information Center (rarediseases.info.nih.gov); MedlinePlus—Genetics (https://medlineplus.gov/
genetics/)

other types of cancer such as prostate cancer, pancreatic can-BRCA2 genes account for approximately 20–25% of heredi-
cer, melanoma in BRCA2-mutation carriers, and several tary breast cancers and about 5–10% of all breast cancers as
other types of cancer. This section will discuss how inheritedwell as approximately 15% of ovarian cancers. Germline
mutations in BRCA1 and BRCA2 affect the development, mutations in other homologous recombination genes, includ-
progression, and metastasis of breast and ovarian cancers in ing PALB2, RAD51C, RAD51D, BARD1, and BRIP1, have
females and prostate cancer in males. also been linked to increasing the risk for breast and ovarian
cancer.
5.2.1.1 Genetic Causes Double-strand breaks simultaneously disrupt both strands
Hereditary breast and ovarian cancer syndrome is inherited of the DNA double helix leading to chromatid breaks. Repair
in an autosomal dominant manner and is caused by germline of double-strand breaks can happen in two ways, either
heterozygous mutations in the BReast CAncer susceptibility through homologous recombination (HR), a preferred, mostly
genes 1 and 2 (i.e., BRCA1 and BRCA2). BRCA1 and BRCA2 error-free pathway, or by non-homologous end joining
are tumor suppressor genes that encode proteins involved in (NHEJ), an error-prone process that may result in illegitimate
DNA damage response and DNA repair pathways. BRCA1 is end-joining potentially leading to intrachromosomal or inter-
involved in several cellular pathways that maintain genomic chromosomal rearrangements, deletion of segments of the
stability, including DNA damage repair, cell-cycle check- chromosome or other genomic abnormalities. Cells with
point control, transcriptional regulation of genes, chromatin BRCA1 or BRCA2 germline mutation acquire additional
remodeling, and protein ubiquitination [1]. BRCA2 plays an somatic alterations such as loss of the wild-type BRCA1 or
important role in DNA double-stranded break repair by BRCA2 allele (i.e., loss of heterozygosity, LOH), leading to a
homologous recombination and contributes to several other defective HR-mediated repair and forcing the cell to use
cellular processes that help maintain genome stability, NHEJ pathway for double-strand breaks repair. This results in
including cell cycle regulation, telomere homeostasis, tran- the accumulation of genomic alterations that promote malig-
scription, and DNA damage response [2]. Proteins encoded nant progression [5]. Inactivation of other tumor suppressor
by BRCA1 and BRCA2 form complexes with other tumor genes such as TP53, ATM or CHEK2 is also commonly seen
suppressor proteins, encoded by RAD51 and PALB2, to initi- in BRCA-deficient tumors allowing cells to bypass check-
ate homologous recombination at sites of DNA double-­ point controls and evade apoptosis, leading to malignant
strand breaks [3, 4]. Inherited mutations in BRCA1 and transformation. High co-occurrence of TP53 mutations in
5 Hereditary Cancer Syndromes and Cancer Metastasis 43

Fig. 5.1 Contributions of germline genomes (i.e., heredity), bad luck Image was used with permission from Springer Nature and adapted
(i.e., replicate), and external factors (i.e., environmental) to tumorigen- from Jiang X, Asad M, Li L et al. Germline genomes have a dominant-
esis and metastasis heritable contribution to cancer immune evasion and immunotherapy
The degree of contributions for each factor to tumorigenesis and metas- response. Quant Biol 8, 216–227 (2020). https://doi.org/10.1007/
tasis in each organ (bone, bone marrows, brain, thymus, respiratory s40484-­020-­0212-­7, Page 223, Table 7.7, Copyright Springer Nature,
tract, lung, stomach, colon, kidney, liver, ovary, and urinary system) is 2020 (Jiang et al., 2020)
illustrated by the color depth (from light pink to deep red)

breast tumors with BRCA1 and BRCA2 mutations suggests (PR), or human epidermal growth factor receptor 2 (HER2),
that TP53 loss of function may be a necessary step in the referred to as triple-negative breast cancer (ER-/PR-/HER2-).
tumorigenesis of BRCA-associated carcinomas [6]. BRCA2-associated breast cancer tumors typically overex-
presses ER (ER+) and in some cases PR (ER+/PR+) but typi-
5.2.1.2 Breast Cancer cally do not overexpress the HER2 receptor. Different breast
Female carriers of BRCA1 or BRCA2 mutation have a higher cancer subtypes exhibit distinct gene expression profiles, bio-
risk for developing breast cancer in their lifetime (46–87% logical behaviors, and patterns of metastasis [16].
and 38–84%, respectively) compared to females who are non- A high likelihood of CNS metastasis (50–53%) has been
carriers (12%) based on cumulative data from multiple large reported in both BRCA1- and BRCA2-mutation carriers com-
cohorts [7–12]. In addition, after the initial breast cancer pared to noncarriers (25%), and the presence of CNS metas-
diagnosis, carriers of inherited BRCA1 or BRCA2 mutation tasis was associated with increased mortality [16]. CNS
have a higher likelihood of developing contralateral breast metastasis in BRCA1-mutation carriers may be due to the
cancer with estimated cumulative risk by age 70 of 83% in higher prevalence of triple-negative breast cancer, which has
BRCA1-mutation carriers and 62% for BRCA2-­mutation car- been documented to increase the risk for CNS involvement
riers [11]. Breast tumors developing in BRCA1- or BRCA2- [17, 18]. However, BRCA2 germline mutation was found to
mutation carriers have been shown to have different be an independent predictor of CNS metastasis [16]. Breast
morphologic and molecular features [13–15]. BRCA1- tumors in BRCA1-mutation carriers also had frequent lung
associated tumors are typically high-grade carcinomas that and distant lymph node metastases (~50%), compared to
do not express estrogen receptor (ER), progesterone receptor predominant bone metastasis observed in BRCA2-mutation
44 B. M. Shaw and O. Kis

carriers (75%) and noncarriers (53%) [16]. In addition, BRCA2 germline mutation, prophylactic oophorectomy has
patients with metastatic cancer who are carriers of a germ- shown to result in a significant risk reduction for breast and
line BRCA1 or BRCA2 mutations are more likely to benefit ovarian cancer.
from treatment with Poly (ADP) ribose polymerase (PARP)
inhibitors. Several FDA-approved PARP inhibitors (e.g., 5.2.1.4 Prostate Cancer
olaparib, talazoparib) are included in the NCCN guidelines Male carriers of germline BRCA1 or BRCA2 mutation are at
for systemic treatment in patients with recurrent or meta- an increased lifetime risk for prostate cancer. Compared to
static breast cancer harboring BRCA1 or BRCA2 germline noncarriers, BRCA1- and BRCA2-mutation carriers have 3.7-­
mutations [19]. Over 65–70% of patients with advanced fold and 8.6-fold higher risk, respectively, for developing
breast cancer disease develop skeletal metastasis. Carriers of prostate cancer in their lifetime, with strongest association
the BRCA1 or BRCA2 mutation with metastatic breast cancer observed for advanced or metastatic prostate cancer [28–33].
have been found to respond better to PARP inhibitor therapy Based on a large cohort of >2000 patients with prostate can-
compared to noncarriers [20]. cer, individuals with germline BRCA1 or BRCA2 mutation
were found to have significantly higher rates of aggressive
5.2.1.3 Ovarian Cancer prostate cancer (Gleason score ≥8), higher risk of nodal
The risk of ovarian cancer is also very high in female carriers involvement and distant metastasis, and poor survival out-
of a BRCA1 (39–63%) or a BRCA2 (16–27%) mutation com- comes compared to noncarriers [34]. Other studies showed
pared to noncarriers (1–2%), and presence of a germline that prostate cancer arising in BRCA2-mutation carriers is
BRCA1 or BRCA2 mutation is also associated with a higher associated with a higher histologic grade and poorer overall
likelihood of serous adenocarcinoma pathology [7, 8, 10, survival [35, 36]. Collectively, these studies have demon-
21]. High-grade serous adenocarcinomas frequently seen in strated that nodal involvement and distant metastasis are
BRCA1 or BRCA2 mutation carriers typically exhibit promi- more common in patients with prostate cancer who have
nent intraepithelial lymphocytes, marked nuclear atypia, and germline BRCA1 or BRCA2 mutations than in noncarriers,
abundant mitoses [22]. BRCA1- and BRCA2-associated and that those carriers with local disease develop metastasis
high-grade serous carcinoma demonstrate characteristic his- earlier and may require tailored clinical management or ear-
topathological features and growth patterns in metastatic lier treatment intervention. FDA-approved PARP inhibitors
sites and are associated with either pushing pattern metasta- (e.g., olaparib, rucaparib) are included in guidelines as
ses or infiltrative metastases composed only of micropapil- second-­line and subsequent therapy for patients with meta-
lae. They differed significantly from cases lacking BRCA1 or static castration-resistant prostate cancer harboring patho-
BRCA2 abnormalities, which have infiltrative metastases genic mutations in BRCA1 or BRCA2.
with combinations of papillary, glandular, cribriform, and
micropapillary architecture [23].
Ovarian cancer in individuals with germline BRCA1 or 5.2.2 Lynch Syndrome
BRCA2 mutation is associated with increased risk for vis-
ceral metastasis outside the pelvis as well as a higher risk for Lynch syndrome, also known as hereditary non-polyposis
CNS metastasis than noncarriers [24–26]. However, CNS colorectal cancer (HNPCC), accounts for approximately
metastasis in BRCA1 or BRCA2 mutation carriers were more 1–3% of colon cancers, and 0.8–1.4% of endometrial can-
likely to occur without intracranial disease and were found to cers. With a population prevalence of approximately 1:440,
have a favorable prognosis. In addition, patients with meta- Lynch syndrome is the most common cause of hereditary
static cancer who are carriers of a germline BRCA1 or colorectal cancer and hereditary endometrial cancer in
BRCA2 mutations are more likely to benefit from treatment women. It is also associated with increased risk for several
with Poly (ADP) ribose polymerase (PARP) inhibitors, with other types of cancer, including ovarian, gastrointestinal,
several FDA-approved PARP inhibitors available for the renal, hepatobiliary tract, urinary tract, brain, and skin
treatment of metastatic ovarian and breast cancer in patients cancer.
with germline BRCA1 or BRCA2 mutation. Several studies
show improved response to PARP inhibitors in patients with 5.2.2.1 Genetic Causes
a germline pathogenic BRCA1 or BRCA2 mutation [27]. Lynch syndrome is inherited in an autosomal dominant man-
Some PARP inhibitors (e.g., niraparib, velaparib) have been ner and is caused by germline heterozygous mutation in
found to cross the blood-brain barrier in animal models and genes involved in the mismatch repair (MMR) pathway, such
may have efficacy against CNS metastasis in patients with as MLH1, MSH2, MSH6, or PMS2, or in rare cases by dele-
BRCA1 and BRCA2 germline mutations [26]. Given the high tion within EPCAM gene leading to epigenetic silencing of
risk for high-grade ovarian cancer (including fallopian tube MSH2 by promoter hypermethylation. The MLH1, MSH2,
and primary peritoneal cancers) in women with BRCA1 or MSH6, and PMS2 genes are involved in repairing DNA rep-
5 Hereditary Cancer Syndromes and Cancer Metastasis 45

lication mistakes that occur when DNA is copied during cell tivate the mismatch repair genes (e.g., MLH1 promoter
division. The MMR pathway functions to correct DNA mis- hypermethylation).
matches through the interaction of MMR proteins as het-
erodimers. MSH2 typically heterodimerizes with MSH6 to 5.2.2.2 Colorectal Cancer
form MutSa complex, which recognizes single base-pair Lynch syndrome is the most common hereditary cause of
mismatches and small insertion/deletion loops and creates a early-onset colorectal cancer accounting for approximately
sliding clamp around the DNA. The MutSa complex then 2–4% of all colorectal cancers. In individuals diagnosed with
binds with the MLH1-PMS2 (MutLa) complex. Interactions Lynch syndrome, the relative risk for developing colorectal
of this four-protein MMR complex with exonuclease-1, cancer varies depending on the causative gene (typically
PCNA, and DNA polymerase allow excision of the mismatch MLH1, MSH2, MSH6, or PMS2). The relative risk of devel-
to occur. In addition, another MMR protein, MSH3, plays a oping colorectal cancer is significantly higher in individuals
role in repair of larger insertion/deletion loops. When MSH2 with germline pathogenic mutation in MLH1 (46–49%) or
heterodimerizes with MSH3 it forms MutSb complex which MHS2 (43–52%) compared to carriers of a pathogenic vari-
can recognize larger insertion/deletion loops. MutSb com- ant in MSH6 (15–44%) or PMS2 (12–20%) [37–42]. Overall,
plements the function of MutSa, which recognizes single all four genes result in an increased risk for colorectal cancer
pair mismatches and small insertion/deletion loops. Several than the general population risk of 4.5%. In addition, carriers
other MutL heterodimers can also form since MLH1 can het- of MLH1 or MSH2 mutation typically develop colorectal
erodimerize with PMS2 (most common), PMS1, or MLH3. cancer at a younger age (median age of onset of 43–45 years)
Defects in any of the mismatch repair genes can result in the compared to carriers of mutations in MSH6 (51–63 years) or
disruption of the MMR complex leading to accumulation of PMS2 (47–66 years) and imply the need for different cancer
DNA errors as the abnormal cells continue to divide which surveillance strategies depending on the causative genetic
result in “slippage” errors in areas of repetitive DNA defect [40]. Colorectal cancer tumors that develop in indi-
sequences, known as microsatellites. These accumulated viduals with Lynch syndrome are poorly differentiated,
replication errors lead to microsatellite instability (MSI) excessively mucinous with a signet ring or cribriform histol-
genotype in MMR-deficient tumors such as Lynch syndrome ogy, and are characterized by frequent localization in the
tumors. proximal colon, tendency to develop multiple primary carci-
While MSI is a defining feature of Lynch syndrome nomas, and high levels of tumor infiltrating lymphocytes.
tumors, this molecular finding is not specific for the inherited Poorly differentiated tumors are typically associated with a
Lynch syndrome and occurs more frequently in sporadic poor prognosis. However, Lynch-related colorectal cancer
forms of colorectal and endometrial cancer. MSI is detected has a better prognosis than sporadic colorectal cancer and is
in approximately 15–20% of colorectal cancers and 20% of associated with less invasion and metastases than sporadic
endometrial cancers, most of which are caused by hyper- colorectal cancer. Although the mechanism behind this sur-
methylation of the MLH1 gene promoter leading to MLH1 vival advantage is not fully understood, several possible
gene inactivation through transcriptional silencing, or due to explanations have been proposed. Immunological response
biallelic somatic mutations in the MMR genes. MLH1 pro- of intraepithelial cytotoxic T lymphocytes against tumor
moter hypermethylation occurs sporadically and accounts cells in Lynch syndrome or epithelial-mesenchymal transi-
for 70–80% of MSI-High colorectal tumors. Thus, careful tion implicated in the development and progression of Lynch
evaluation of patients with MSI-High tumors using multiple syndrome makes it less susceptible to metastases [43].
testing methodologies is typically needed to differentiate Colorectal cancers with high microsatellite instability (MSI-­
between patients with sporadic cancer and those with inher- High) tend to have a better prognosis than microsatellite-­
ited Lynch syndrome who may benefit from increased cancer stable (MSS) colorectal cancers. This improved prognosis
surveillance, earlier and more extensive cancer screening or may be related to active anti-tumor immune responses.
risk-reducing medical interventions. Presence of germline Moreover, MSI-High/MMR deficient status in colorectal
inherited mutation in MLH1 and MSH2 accounts for approx- cancer is considered a predictive biomarker of response to
imately 90% of pathogenic variants in families with Lynch immunotherapy associated with favorable outcomes com-
syndrome, while a smaller proportion is attributed to patho- pared with microsatellite stable (MSS)/MMR proficient
genic mutations in MSH6 (7–10%) or PMS2 (<5%). In addi- colorectal cancer. The risk of recurrent colorectal cancer,
tion, germline deletions in EPCAM gene located near MSH2 however, is increased in individuals with Lynch syndrome.
result in epigenetic silencing of MSH2 by promoter hyper- Specifically, the risk of metachronous colorectal cancer is
methylation and are reported in approximately 1% of indi- reported to be highest (up to 43%) in individuals with an
viduals with Lynch syndrome. Deficiency in the mismatch MLH1 or MSH2 pathogenic variant who have segmental
repair system may also occur sporadically and can be caused resection but lower in carriers of MSH6 or PMS2 germline
by somatic genetic mutation or epigenetic changes that inac- mutation [44].
46 B. M. Shaw and O. Kis

5.2.2.3 Endometrial Cancer MEN1-associated endocrinopathy [47, 49, 50]. This tumor
Mutations in MMR repair genes are also associated with type typically presents with hypercalcemia with symptoms
increased risk for endometrial cancer. The highest risks for including osteoporosis, lathery, and depression. Pituitary
endometrial cancer is reported in carriers of pathogenic tumors are observed more commonly in females than males
mutations in MLH1 (43–57%), MSH2 (21–57%), or MSH6 with prolactinoma being the most common of the pituitary
(17–46%), and to a lesser extent PMS2 (0–15%) [37–42]. tumors occurring in 20% of cases. Clinical presentation of
Furthermore, the mean age at endometrial cancer diagnosis pituitary tumors varies based on both the tumor size and the
is earlier for MLH1 and MSH2-mutation carriers (47– hormone produced. Neuroendocrine tumors occur in 30–70%
49 years) compared to individuals with an inherited mutation of cases with gastrinoma and insulinoma being the most
in MSH6 (53–55 years) or PMS2 (49–56 years) and could common at 40% and 10%, respectively [47, 49, 50].
impact cancer surveillance [40]. Immunotherapy strategies Gastrinomas present with Zollinger-Ellison syndrome and
against tumors with high microsatellite instability (MSI-­ are usually malignant at the time of diagnosis and tend to
High) have been evaluated in multiple studies including metastasize to the lymph nodes or the liver. Carcinoid tumors
comparison of MSI-High colorectal and endometrial cancers are non-hormone-secreting tumors and occur in over 10% of
caused by either sporadic MLH1 promoter methylation or cases and are also known to be at increased risk of metasta-
inherited Lynch syndrome. MSI-High/MMR deficient status sis. Currently, the risk of metastasis correlates with tumor
is thought to be associated with high somatic mutational size with 25–40% of individuals with tumors larger than
load, which causes increased production of tumor-specific 4 cm developing liver metastases and 50–70% developing
neoantigens and increased tumor infiltration lymphocytes, lymph node metastases if the tumor is 2–3 cm in size [50].
and is associated with favorable response to immunotherapy Roughly one in four patients develop distant metastases
[45, 46]. resulting in death. Non-endocrine tumors include facial
angiofibromas (can be visible on the lips), collagenomas,
lipomas, meningiomas, ependymomas, and leiomyomas.
5.2.3  ultiple Endocrine Neoplasia
M
Syndromes 5.2.3.2 Multiple Endocrine Neoplasia Type 2
Multiple endocrine neoplasia type 2 (MEN2) results from an
The multiple endocrine neoplasia (MEN) syndromes are a activating germline mutation in the RET gene located at chro-
group of autosomal dominant inherited cancers character- mosome position 10q11.2. RET encodes for a tyrosine kinase
ized by the development of tumors in the endocrine system receptor located on the surface of cell membranes and func-
[47]. MEN is separated into multiple endocrine neoplasia tions in cell growth and differentiation. Activating mutations
type 1 (MEN1), multiple endocrine neoplasia type 2 (MEN2) in the RET gene are associated with MEN2 and result in the
and MEN4 based on the gene involved and the location of development of medullary thyroid carcinoma (MTC), pheo-
the tumors present. chromocytomas (PCC), and primary hyperparathyroidism
(PHPT). MEN2 is separated into Multiple Endocrine
5.2.3.1 Multiple Endocrine Neoplasia Type 1 Neoplasia 2A, MEN2A (MEN2), Multiple Endocrine
Multiple endocrine neoplasia type 1 (MEN1) syndrome Neoplasia 2B, MEN2B (MEN3), and familial medullary thy-
occurs both sporadically and as an inherited autosomal domi- roid cancer (FMTC) based on the RET mutation present,
nant disorder. The incidence of MEN1 is around 1 in 30,000 aggressiveness of the medullary thyroid carcinoma (MTC),
with ~90% of the cases being inherited. MEN1 syndrome and if other endocrine tumors are present. The American
results from pathogenic variants in the MEN1 gene which Thyroid Association has published guidelines for manage-
encodes for the protein menin. Menin acts as a tumor sup- ment of medullary thyroid carcinoma and in those guidelines
pressor gene and has a role in controlling the cell cycle. The created subcategories within MEN2A. These subcategories
criteria for diagnosing a patient with MEN1 include a patient include classical MEN2A (MEN2A), MEN2A and cutaneous
with two or more MEN1-associated tumors or one MEN1 lichen amyloidosis (MEN2A/CLA), and MEN2A and
associated tumor and a first degree relative with MEN1 [48, Hirschsprung disease (MEN2A/HD) [51]. For this discus-
49]. Additionally, a germline pathogenic variant in one of the sion, MEN2A and its subtypes will be discussed as a single
genes causing a MEN syndrome in an asymptomatic individ- entity as the incidence of MTC, PHEO, and PHPT is similar.
ual is considered diagnostic [49]. The age-related penetrance MEN2A most commonly results from pathogenic vari-
for clinical features is 50% by age 20 and 95% by age 40. ants in codons 609, 611, 618, and/or 620 of exon 10 or in
Individuals with MEN1 have the potential to develop one codon 634 of exon 11 of the RET gene [52]. Patients with
or a combination of over 20 different endocrine and non-­ MEN2A typically present with MTC which develops in
endocrine tumors. Parathyroid tumors occur in 90% of indi- ~95% of cases while ~50% of cases develop PCC and around
viduals between the ages of 20–25 and are the main 25%–30% will develop PHPT [53]. While most patients
5 Hereditary Cancer Syndromes and Cancer Metastasis 47

develop MTC, the presence of PCC and PHPT appears to monitor patients [51]. Individuals with MTC have a similar
depend on the specific mutations present. Patients with muta- pattern of metastases as individuals with sporadic
tions in exon 10 have a lower penetrance of PHEO’s than MTC. When a thyroid nodule is present, 75% of cases have
those with other RET mutations. Additionally, the RET metastases occurring in the regional lymph nodes and
mutations in codon 634 have a reportedly higher incidence of 10–15% have developed distant metastasis [63]. In infants
HPTH compared to mutations in codons 609, 611, 618, and with MEN2B, cases have been reported with metastasis to
620 [54, 55]. Management of MEN2A is dependent on the the lymph nodes at the time of prophylactic thyroidectomy
mutation present and calcitonin levels. Generally, individu- [64] further demonstrating the aggressive nature of this dis-
als with mutations in 609, 611, 618, 620, 630, and 631 (class ease. Metastases from MTC tend to develop in the lungs,
B/moderate risk) have a less aggressive phenotype than indi- bone, liver, and less commonly in the brain (1–5% of cases).
viduals with a mutation in codon 634 (class C/high risk). Metastases to the lungs typically involve multiple sites and
Patients with class B mutations can have an age of onset of can be associated with the mediastinal lymph nodes. Liver
MTC as early as 5 years of age. Due to this, monitoring of metastases occur in 45% of cases and can also involve mul-
calcitonin levels and physical examination should be per- tiple sites; thus resection is not usually possible. When dis-
formed before the age of 5 with monitoring for PHEO usu- tant metastases are present, the treatment options are mainly
ally starting at age 8. Patients with class C mutations can palliative in nature [65].
have MTC develop within the first years of life, so screening
is recommended to begin by age three in these individuals
[52, 56]. Prophylactic thyroidectomy can be considered 5.3 Metastasis Susceptibility Loci
before the age of 5.
MEN2B is also characterized by MTC (100% of cases) In addition to the genes known to be associated with heredi-
and the presence of PCC (50% of cases); however, primary tary cancer syndromes, recent studies have begun to examine
hyperparathyroidism (PHPT) is not present. Patients instead the impact of an individual’s germline mutations and the
exhibit a phenotype of mucosal neuromas of the lips and impact these changes may have on the development and
tongue giving the lips a bumpy appearance, ganglioneuroma- metastasis of tumors (Fig. 5.1). Studies in animal models
tosis of the gastrointestinal tract and marfanoid habitus. The have shown the role of the genetic background in the devel-
majority of MEN2B patients (95%) have a pathogenic mis- opment of metastatic disease. On such animal model, mouse
sense variant in exon 16 at codon 918 resulting in a Met to mammary tumor virus-driven polyomavirus middle T-antigen
Thr change [57, 58]. The next commonly observed patho- (MMTV-PyMT), uses MMTV-LTR to drive expression of a
genic variant (<5%) is in codon 883 and is also a missense mammary gland specific polyomavirus middle T-antigen
mutation which results in an Ala to Phe change [59, 60]. which results in development of highly metastatic tumors
Individuals with MEN2B have a more aggressive clinical [66]. These mice were crossed with multiple other strains of
course with the development of MTC within the first year of mice to generate animal models of breast cancer. The data
life and PHEO as early as 8 years of age. Due to the aggres- from these crossbreeding experiments show when MMTV-­
siveness of MTC, patients with inherited MEN2B should PyMT is crossed with different strains, the progeny of these
have a prophylactic thyroidectomy in the first months of life mating’s show significant variation in the development of
[52]. Those with de novo MEN2B are usually not diagnosed pulmonary metastasis from the primary mammary tumor
until a thyroid nodule is detected; however, earlier interven- [67]. This difference was as large as 40fold in some cases
tion is possible if the phenotype associated with MEN2B is [68]. The authors attribute this observed variation to inher-
recognized as it typically precedes MTC in these patients ited polymorphisms which changed the disease process in
[51]. It is also critical that PHEO are identified prior to before these animals. Other studies in mice have identified potential
thyroidectomy due to increased morbidity and death due to modifier genes which include genes known to lead to suscep-
an undiagnosed PHEO. tibility to metastasis in mammary cancer. Furthermore,
mouse models used in the study of prostate cancer have led
5.2.3.3 Metastasis in MEN2A and MEN2B to identification of potential metastasis susceptibility genes
Individuals with MEN2A and MEN2B are at an increased suggesting this susceptibility could be due to the genetic
risk to develop metastasis at an early age. While the mutation background of the individual [68]. Recent studies have been
present can provide some guidance as to the aggressiveness published in the literature which supports the impact of
of the disease, there is a high degree of variability among germline variation on tumor progression. This variation is
patients with similar mutations and even within families with influenced by amino acid changes which alter expression of
the same mutation [61, 62]. Due to this variability, overall neighboring genes which then increases the chance a somatic
tumor size and calcitonin levels are also good predictors of mutation will arise in the same gene or pathway. Germline
metastasis risk in these individuals and should be used to variants have also been tied to transcriptional differences,
48 B. M. Shaw and O. Kis

splice isoforms, and expression profiles in various tumors. association with tumor progression. Specifically, a nonsyn-
These germline mutations, therefore, lead to genomic insta- onymous SNP g.1421 C>T, p.Pro436Leu (rs9306160) was
bility, potential for more somatic pathogenic variants, associated with reduced risk of metastasis risk. This variant
increased proliferation, lack of apoptosis, and changes to the has an allele frequency of ~0.42 in the general population. In
tumor microenvironment leading to an increased risk of both cohorts, the T allele was associated with a significant
metastasis. difference in metastasis-free survival and overall survival.
The effect was determined to be greater in patients with ER+
tumors. It was also noted when rs9306160 (T) is combined
5.4 Metastasis Genes/Loci with the rs248490 (A;A) in the SIPA1 gene the risk to develop
distant metastasis was 2.5-fold lower than individual with
5.4.1 Signal-Induced Proliferation-­ the rs9306160 (C;C) and rs2448490(G) carriers [74].
Associated 1 (SIPA1)

SIPA1 encodes for a protein with a C-terminal leucine zipper 5.4.3 Interleukin-6 (IL-6)
motif and an N-terminal GTPase activating protein. The pro-
tein interacts with member of the Ras family of GTPases and IL-6 encodes a cytokine that is involved in variety of biologi-
is thought to play a role in regulating cell adhesion. Crawford cal functions including inducing expression of VEGF which
et al. identified three germline variants, −313G>A results in increased angiogenesis. IL-6 has been implemented
(rs931127), 545C>T (rs3741378), and 2760G>A (rs746429) in a wide range of disease processes including Kaposi’s sar-
in the SIPA1 gene which were associated with metastasis in coma, diabetes, Hodgkin’s lymphoma, hypertension, gastric
breast cancer [69]. The SNP’s rs746429 and rs91127 were cancer, neuroblastoma, ovarian cancer, breast cancer, and
also shown to be significantly associated with lymph node prostate cancer [75–79]. Most of these studies involving can-
involvement while rs3741378 was observed more commonly cer link IL6 expression to poor outcomes largely due to treat-
in estrogen receptor negative tumors and with progesterone ment resistance.
negative tumors [69]. Functional studies showed SIPA1 is A SNP within the promotor region of IL-6 (rs1800795;
highly expressed in the breast cell cancer line MDA-MB-231 G;G) has been linked to increased risk of distant metastasis
and interacts with the promoter of the ITGB1 gene regulat- in two separate breast cancer cohorts [80]. This SNP has
ing expression of integrin β1 which has been shown to result been previously linked to a poor outcome in several cancer
in cell invasion and migration thru the FAK/AKT-MMP9 types providing further information regarding its role in can-
signaling pathway [70]. cer. Abana et al. also noted an increased incidence of the
While most research involving SIPA 1 has focused on rs1800795 G:G allele in black and Asian populations. The
Breast cancer, this gene has been shown to potentially lead to higher incidence of G:G individuals in the black population
susceptibility to metastasis in other cancer types including may offer an explanation for the increase in brain metastasis
prostate cancer, oral squamous cell carcinoma, colorectal and overall poorer outcome in this population compared to
cancer, cervical cancer, lung cancer, gastric cancer, and mel- that observed in the white population [80].
anoma [71–73]). While the available evidence suggests the
SIPA1 gene may be a valuable prognostic indicator and
potential therapeutic target, there is further work that needs 5.4.4  ryl Hydrocarbon Receptor Nuclear
A
to be done to determine the significance of this gene and how Translocator Like 2 (ARNTL2)
it relates to metastatic disease.
ARNTL2 encodes a regulator of transcription that is a core
component of the circadian clock. The circadian clock is
5.4.2  ibosomal RNA-Processing Protein 1B
R responsible for regulation of metabolism and behavioral
(RRP1B) rhythms. The circadian clock establishes these rhythms
through regulation of gene expression throughout a 24-h cycle.
RRP1B is predicted to be involved in 60S ribosomal RNA Two separate studies suggest overexpression of ARNTL2
processing and has been shown to interact with SIPA1 in is strongly associated with metastasis. ER- breast cancer and
yeast two hybrid immunoprecipitation assays as well as in lung adenocarcinoma. The first study examined the role of
functional assays. Extracellular matrix (ECM) in vitro assays ARNTL2 in ER- breast cancer utilizing mouse models and a
showed that altered expression of RRP1B results in tumor patient cohort of individuals with ER-breast cancer. Studies
growth and dissemination in metastasis assays [74]. in ARNTL2 knockout mice showed a reduction in pulmo-
Two separate breast cancer cohorts’ constitutional poly- nary metastasis while overexpression resulted in increased
morphisms within the RRP1B gene showed a significant pulmonary metastasis. Neither of the changes resulted in a
5 Hereditary Cancer Syndromes and Cancer Metastasis 49

change in primary tumor growth [81]. Ha et al. also per- bility” situation not dissimilar to that observed in the search
formed human expression quantitative trait locus (eQTL) for genes involved in height and other genetic diseases [84].
analysis in two ER-breast cancer cohorts. This analysis
showed polymorphisms in the promoter region of ARNTL2
affect expression levels and in conjunction with the mouse 5.6 Future Directions
models suggest a significant effect on metastasis in ER-breast
cancer [81]. The second study examined the role of Future studies will be tasked with determining the overall
ARNTL2 in lung adenocarcinoma and showed this gene is effect of individual germline variants on metastasis as well
overexpressed in metastases from primary lung tumors com- as cumulative effects of multiple variants in a single indi-
pared to non-metastatic lung primary tumors. The authors vidual. Currently the field has the capability to do whole
concluded the data from both mouse and human lung adeno- genome analysis on both tumor and normal sample from the
carcinoma is consistent with ARTNL2 overexpression result- individual; however, there are limitations of doing these
ing in metastases compared to normal expression levels [82]. studies such as the ability to consistently detect copy number
variants and balanced changes, and rapid data analysis.

5.4.5 BLM RecQ like Helicase (BLM)

Open Questions
The BLM gene encodes the DNA helicase RecQl3, an • What technologies will be necessary to allow for
enzyme involved in maintenance of genomic integrity. consistent detection of copy number variants and
Homozygous pathogenic variants lead to Bloom syndrome. balanced changes?
The associated phenotype includes pre- and postnatal growth • Currently long-read next-generation sequencing
retardation, facial sun-sensitive telangiectatic erythema, and optical genome mapping show promise in this
increased susceptibility to infections, and cancer predisposi- area. Additionally, what new genomic changes may
tion with leukemia/lymphoma and adenocarcinoma being be discovered utilizing these newer technologies?
the most common. • Data analysis pipelines have been developed to
Carriers of pathogenic variants in the BLM gene do not analyze large amounts of data; however, what
display a phenotype associated with the syndrome yet may adjustments/improvements are necessary to speed
have in increased risk for breast cancer in females and colon up and improve data analysis of large data sets in a
cancer. A recent study of patients with advanced prostate clinical setting where answers are critical to
cancer showed a relative risk of metastasis to be 3.4 in carri- patient care?
ers of BLM germline mutations [83]. Based on the data, the • Lastly, how can we solve the “missing heritability”
possibility exists that carriers of pathogenic loss of function issue, where many variants may have a discernable dis-
variants in the BLM gene may convey susceptibility to ease effect yet the impact of each is so miniscule as to
metastasis in prostate cancer. Interestingly, the study also be undetectable by current data analysis techniques?
noted the individuals with loss of function germline variants
in the BLM gene had somatic BRCA2 inactivating mutations
in the tumors suggesting a possible interaction between these References
two genes [83].
1. Wu J, Lu LY, Yu X. The role of BRCA1 in DNA damage
response. Protein Cell. 2010;1(2):117–23. https://doi.org/10.1007/
s13238-­010-­0010-­5.
5.5 Conclusion 2. Fradet-Turcotte A, Sitz J, Grapton D, Orthwein A. BRCA2 func-
tions: from DNA repair to replication fork stabilization. Endocr
Relat Cancer. 2016;23(10):T1–t17. https://doi.org/10.1530/
In conclusion, there is a growing body of evidence in the erc-­16-­0297.
literature which supports the role of germline changes in sus- 3. Moynahan ME, Pierce AJ, Jasin M. BRCA2 is required for
ceptibility to metastasis. However, the majority of the studies homology-­directed repair of chromosomal breaks. Mol Cell.
conducted to this point cite small sample sizes and the need 2001;7(2):263–72. https://doi.org/10.1016/s1097-­2765(01)001
74-­5.
to do studies with larger data sets. Additionally, the genes 4. Zhao W, Steinfeld JB, Liang F, Chen X, Maranon DG, Jian Ma
involved in susceptibility to metastasis are likely to have a C, et al. BRCA1-BARD1 promotes RAD51-mediated homolo-
cumulative effect with each variant/gene contributing to the gous DNA pairing. Nature. 2017;550(7676):360–5. https://doi.
development of metastases but not being solely responsible. org/10.1038/nature24060.
5. Roy R, Chun J, Powell SN. BRCA1 and BRCA2: different roles
The search for the other involved genes may prove arduous in a common pathway of genome protection. Nat Rev Cancer.
and researchers could find themselves in a “missing herita- 2011;12(1):68–78. https://doi.org/10.1038/nrc3181.
50 B. M. Shaw and O. Kis

6. Sensi E, Tancredi M, Aretini P, Cipollini G, Naccarato AG, 21. McLaughlin JR, Rosen B, Moody J, Pal T, Fan I, Shaw PA, et al.
Viacava P, et al. p53 inactivation is a rare event in familial breast Long-term ovarian cancer survival associated with mutation in
tumors negative for BRCA1 and BRCA2 mutations. Breast BRCA1 or BRCA2. J Natl Cancer Inst. 2013;105(2):141–8. https://
Cancer Res Treat. 2003;82(1):1–9. https://doi.org/10.1023/ doi.org/10.1093/jnci/djs494.
B:BREA.0000003836.91844.b5. 22. Fujiwara M, McGuire VA, Felberg A, Sieh W, Whittemore AS,
7. Chen S, Iversen ES, Friebel T, Finkelstein D, Weber BL, Eisen A, Longacre TA. Prediction of BRCA1 germline mutation status
et al. Characterization of BRCA1 and BRCA2 mutations in a large in women with ovarian cancer using morphology-based crite-
United States sample. J Clin Oncol. 2006;24(6):863–71. https://doi. ria: identification of a BRCA1 ovarian cancer phenotype. Am
org/10.1200/jco.2005.03.6772. J Surg Pathol. 2012;36(8):1170–7. https://doi.org/10.1097/
8. Easton DF, Ford D, Bishop DT. Breast and ovarian cancer inci- PAS.0b013e31825d9b8d.
dence in BRCA1-mutation carriers. Breast cancer linkage consor- 23. Reyes MC, Arnold AG, Kauff ND, Levine DA, Soslow
tium. Am J Hum Genet. 1995;56(1):265–71. RA. Invasion patterns of metastatic high-grade serous carcinoma
9. Ford D, Easton DF, Bishop DT, Narod SA, Goldgar DE. Risks of of ovary or fallopian tube associated with BRCA deficiency.
cancer in BRCA1-mutation carriers. Breast cancer linkage con- Mod Pathol. 2014;27(10):1405–11. https://doi.org/10.1038/
sortium. Lancet. 1994;343(8899):692–5. https://doi.org/10.1016/ modpathol.2013.237.
s0140-­6736(94)91578-­4. 24. Jernigan AM, Mahdi H, Rose PG. Epithelial ovarian cancer meta-
10. Ford D, Easton DF, Stratton M, Narod S, Goldgar D, Devilee P, static to the central nervous system and a family history concerning
et al. Genetic heterogeneity and penetrance analysis of the BRCA1 for hereditary breast and ovarian cancer--a potential relationship.
and BRCA2 genes in breast cancer families. The breast cancer link- Int J Gynecol Cancer. 2015;25(7):1232–8. https://doi.org/10.1097/
age consortium. Am J Hum Genet. 1998;62(3):676–89. https://doi. igc.0000000000000489.
org/10.1086/301749. 25. Ratner E, Bala M, Louie-Gao M, Hazard S, Brastianos P. 946P -
11. Mavaddat N, Peock S, Frost D, Ellis S, Platte R, Fineberg E, et al. brain metastases in primary ovarian cancer: real-world data. Annals
Cancer risks for BRCA1 and BRCA2 mutation carriers: results of oncology, 29, viii338. 2018; https://doi.org/10.1093/annonc/
from prospective analysis of EMBRACE. J Natl Cancer Inst. mdy285.154.
2013;105(11):812–22. https://doi.org/10.1093/jnci/djt095. 26. Stasenko M, Cybulska P, Feit N, Makker V, Konner J, O’Cearbhaill
12. Satagopan JM, Offit K, Foulkes W, Robson ME, Wacholder S, Eng RE, et al. Brain metastasis in epithelial ovarian cancer by BRCA1/2
CM, et al. The lifetime risks of breast cancer in Ashkenazi Jewish mutation status. Gynecol Oncol. 2019;154(1):144–9. https://doi.
carriers of BRCA1 and BRCA2 mutations. Cancer Epidemiol org/10.1016/j.ygyno.2019.05.004.
Biomark Prev. 2001;10(5):467–73. 27. Borella F, Bertero L, Morrone A, Gambella A, Bovetti M, Cosma
13. Honrado E, Benítez J, Palacios J. The molecular pathology of S, et al. Brain metastases from ovarian cancer: current evidence in
hereditary breast cancer: genetic testing and therapeutic implica- diagnosis, treatment, and prognosis. Cancers (Basel). 2020;12(8)
tions. Mod Pathol. 2005;18(10):1305–20. https://doi.org/10.1038/ https://doi.org/10.3390/cancers12082156.
modpathol.3800453. 28. Abida W, Armenia J, Gopalan A, Brennan R, Walsh M, Barron D,
14. Mavaddat N, Barrowdale D, Andrulis IL, Domchek SM, Eccles et al. Prospective genomic profiling of prostate cancer across dis-
D, Nevanlinna H, et al. Pathology of breast and ovarian cancers ease states reveals germline and somatic alterations that may affect
among BRCA1 and BRCA2 mutation carriers: results from the clinical decision making. JCO Precis Oncol. 2017;2017 https://doi.
consortium of investigators of modifiers of BRCA1/2 (CIMBA). org/10.1200/po.17.00029.
Cancer Epidemiol Biomark Prev. 2012;21(1):134–47. https://doi. 29. Agalliu I, Gern R, Leanza S, Burk RD. Associations of high-grade
org/10.1158/1055-­9965.Epi-­11-­0775. prostate cancer with BRCA1 and BRCA2 founder mutations. Clin
15. Musolino A, Bella MA, Bortesi B, Michiara M, Naldi N, Zanelli P, Cancer Res. 2009;15(3):1112–20. https://doi.org/10.1158/1078-
et al. BRCA mutations, molecular markers, and clinical v­ ariables ­0432.Ccr-­08-­1822.
in early-onset breast cancer: a population-based study. Breast. 30. Leongamornlert D, Mahmud N, Tymrakiewicz M, Saunders E,
2007;16(3):280–92. https://doi.org/10.1016/j.breast.2006.12.003. Dadaev T, Castro E, et al. Germline BRCA1 mutations increase
16. Song Y, Barry WT, Seah DS, Tung NM, Garber JE, Lin prostate cancer risk. Br J Cancer. 2012;106(10):1697–701. https://
NU. Patterns of recurrence and metastasis in BRCA1/BRCA2- doi.org/10.1038/bjc.2012.146.
associated breast cancers. Cancer. 2020;126(2):271–80. https:// 31. Na R, Zheng SL, Han M, Yu H, Jiang D, Shah S, et al. Germline
doi.org/10.1002/cncr.32540. mutations in ATM and BRCA1/2 distinguish risk for lethal and
17. Hung MH, Liu CY, Shiau CY, Hsu CY, Tsai YF, Wang YL, et al. indolent prostate cancer and are associated with early age at
Effect of age and biological subtype on the risk and timing of brain death. Eur Urol. 2017;71(5):740–7. https://doi.org/10.1016/j.
metastasis in breast cancer patients. PLoS One. 2014;9(2):e89389. eururo.2016.11.033.
https://doi.org/10.1371/journal.pone.0089389. 32. Pritchard CC, Mateo J, Walsh MF, De Sarkar N, Abida W, Beltran
18. Lin NU, Vanderplas A, Hughes ME, Theriault RL, Edge SB, H, et al. Inherited DNA-repair gene mutations in men with meta-
Wong YN, et al. Clinicopathologic features, patterns of recur- static prostate cancer. N Engl J Med. 2016;375(5):443–53. https://
rence, and survival among women with triple-negative breast doi.org/10.1056/NEJMoa1603144.
cancer in the national comprehensive cancer network. Cancer. 33. van Asperen CJ, Brohet RM, Meijers-Heijboer EJ, Hoogerbrugge
2012;118(22):5463–72. https://doi.org/10.1002/cncr.27581. N, Verhoef S, Vasen HF, et al. Cancer risks in BRCA2 families:
19. Kaufman B, Shapira-Frommer R, Schmutzler RK, Audeh MW, estimates for sites other than breast and ovary. J Med Genet.
Friedlander M, Balmaña J, et al. Olaparib monotherapy in 2005;42(9):711–9. https://doi.org/10.1136/jmg.2004.028829.
patients with advanced cancer and a germline BRCA1/2 muta- 34. Castro E, Goh C, Olmos D, Saunders E, Leongamornlert D,
tion. J Clin Oncol. 2015;33(3):244–50. https://doi.org/10.1200/ Tymrakiewicz M, et al. Germline BRCA mutations are associ-
jco.2014.56.2728. ated with higher risk of nodal involvement, distant metastasis,
20. Stickeler E, Aktas B, Behrens A, Belleville E, Ditsch N, Fasching and poor survival outcomes in prostate cancer. J Clin Oncol.
PA, et al. Update breast cancer 2021 part 1 - prevention and early 2013;31(14):1748–57. https://doi.org/10.1200/jco.2012.43.1882.
stages. Geburtshilfe Frauenheilkd. 2021;81(5):526–38. https://doi. 35. Gallagher DJ, Gaudet MM, Pal P, Kirchhoff T, Balistreri L, Vora K,
org/10.1055/a-­1464-­0953. et al. Germline BRCA mutations denote a clinicopathologic subset
5 Hereditary Cancer Syndromes and Cancer Metastasis 51

of prostate cancer. Clin Cancer Res. 2010;16(7):2115–21. https:// 50. Thakker RV, Newey PJ, Walls GV, Bilezikian J, Dralle H, Ebeling
doi.org/10.1158/1078-­0432.Ccr-­09-­2871. PR, et al. Clinical practice guidelines for multiple endocrine neo-
36. Thorne H, Willems AJ, Niedermayr E, Hoh IM, Li J, Clouston plasia type 1 (MEN1). J Clin Endocrinol Metab. 2012;97(9):2990–
D, et al. Decreased prostate cancer-specific survival of men with 3011. https://doi.org/10.1210/jc.2012-­1230.
BRCA2 mutations from multiple breast cancer families. Cancer 51. Wells SA Jr, Asa SL, Dralle H, Elisei R, Evans DB, Gagel RF, et al.
Prev Res (Phila). 2011;4(7):1002–10. https://doi.org/10.1158/1940- Revised American Thyroid Association guidelines for the manage-
­6207.Capr-­10-­0397. ment of medullary thyroid carcinoma. Thyroid. 2015;25(6):567–
37. Baglietto L, Lindor NM, Dowty JG, White DM, Wagner A, Gomez 610. https://doi.org/10.1089/thy.2014.0335.
Garcia EB, et al. Risks of lynch syndrome cancers for MSH6 muta- 52. Raue F, Frank-Raue K. Genotype-phenotype correlation in
tion carriers. J Natl Cancer Inst. 2010;102(3):193–201. https://doi. multiple endocrine neoplasia type 2. Clinics (Sao Paulo),
org/10.1093/jnci/djp473. 67 Suppl. 2012;1(Suppl 1):69–75. https://doi.org/10.6061/
38. Bonadona V, Bonaïti B, Olschwang S, Grandjouan S, Huiart L, clinics/2012(sup01)13.
Longy M, et al. Cancer risks associated with germline mutations 53. Neumann HPH, Young WF Jr, Eng C. Pheochromocytoma and
in MLH1, MSH2, and MSH6 genes in lynch syndrome. JAMA. Paraganglioma. N Engl J Med. 2019;381(6):552–65. https://doi.
2011;305(22):2304–10. https://doi.org/10.1001/jama.2011.743. org/10.1056/NEJMra1806651.
39. Møller P, Seppälä TT, Bernstein I, Holinski-Feder E, Sala P, Gareth 54. Frank-Raue K, Rybicki LA, Erlic Z, Schweizer H, Winter A, Milos
Evans D, et al. Cancer risk and survival in path_MMR carriers by I, et al. Risk profiles and penetrance estimations in multiple endo-
gene and gender up to 75 years of age: a report from the prospec- crine neoplasia type 2A caused by germline RET mutations located
tive lynch syndrome database. Gut. 2018;67(7):1306–16. https:// in exon 10. Hum Mutat. 2011;32(1):51–8. https://doi.org/10.1002/
doi.org/10.1136/gutjnl-­2017-­314057. humu.21385.
40. Ryan NAJ, Morris J, Green K, Lalloo F, Woodward ER, Hill J, et al. 55. Herfarth KK, Bartsch D, Doherty GM, Wells SA Jr, Lairmore
Association of Mismatch Repair Mutation with age at cancer onset TC. Surgical management of hyperparathyroidism in
in lynch syndrome: implications for stratified surveillance strate- patients with multiple endocrine neoplasia type 2A. Surgery.
gies. JAMA Oncol. 2017;3(12):1702–6. https://doi.org/10.1001/ 1996;120(6):966–73; discussion 973–964. https://doi.org/10.1016/
jamaoncol.2017.0619. s0039-­6060(96)80042-­0.
41. Senter L, Clendenning M, Sotamaa K, Hampel H, Green J, Potter 56. Machens A, Gimm O, Hinze R, Höppner W, Boehm BO, Dralle
JD, et al. The clinical phenotype of lynch syndrome due to germ-­ H. Genotype-phenotype correlations in hereditary medullary thy-
line PMS2 mutations. Gastroenterology. 2008;135(2):419–28. roid carcinoma: oncological features and biochemical proper-
https://doi.org/10.1053/j.gastro.2008.04.026. ties. J Clin Endocrinol Metab. 2001;86(3):1104–9. https://doi.
42. Ten Broeke SW, van der Klift HM, Tops CMJ, Aretz S, Bernstein I, org/10.1210/jcem.86.3.7290.
Buchanan DD, et al. Cancer risks for PMS2-associated lynch syn- 57. Carlson KM, Dou S, Chi D, Scavarda N, Toshima K, Jackson
drome. J Clin Oncol. 2018;36(29):2961–8. https://doi.org/10.1200/ CE, et al. Single missense mutation in the tyrosine kinase cata-
jco.2018.78.4777. lytic domain of the RET protooncogene is associated with mul-
43. Gu GL, Zhu XQ, Wei XM, Ren L, Li DC, Wang SL. Epithelial-­ tiple endocrine neoplasia type 2B. Proc Natl Acad Sci U S A.
mesenchymal transition in colorectal cancer tissue of patients with 1994;91(4):1579–83. https://doi.org/10.1073/pnas.91.4.1579.
lynch syndrome. World J Gastroenterol. 2014;20(1):250–7. https:// 58. Eng C, Smith DP, Mulligan LM, Nagai MA, Healey CS, Ponder
doi.org/10.3748/wjg.v20.i1.250. MA, et al. Point mutation within the tyrosine kinase domain of
44. Anele CC, Adegbola SO, Askari A, Rajendran A, Clark SK, the RET proto-oncogene in multiple endocrine neoplasia type 2B
Latchford A, Faiz OD. Risk of metachronous colorectal can- and related sporadic tumours. Hum Mol Genet. 1994;3(2):237–41.
cer following colectomy in lynch syndrome: a systematic review https://doi.org/10.1093/hmg/3.2.237.
and meta-analysis. Color Dis. 2017;19(6):528–36. https://doi. 59. Gimm O, Marsh DJ, Andrew SD, Frilling A, Dahia PL, Mulligan
org/10.1111/codi.13679. LM, et al. Germline dinucleotide mutation in codon 883 of the RET
45. Howitt BE, Shukla SA, Sholl LM, Ritterhouse LL, Watkins proto-oncogene in multiple endocrine neoplasia type 2B without
JC, Rodig S, et al. Association of Polymerase e-mutated and codon 918 mutation. J Clin Endocrinol Metab. 1997;82(11):3902–
microsatellite-­instable endometrial cancers with neoantigen load, 4. https://doi.org/10.1210/jcem.82.11.4508.
number of tumor-infiltrating lymphocytes, and expression of 60. Smith DP, Houghton C, Ponder BA. Germline mutation of
PD-1 and PD-L1. JAMA Oncol. 2015;1(9):1319–23. https://doi. RET codon 883 in two cases of de novo MEN 2B. Oncogene.
org/10.1001/jamaoncol.2015.2151. 1997;15(10):1213–7. https://doi.org/10.1038/sj.onc.1201481.
46. Pakish JB, Zhang Q, Chen Z, Liang H, Chisholm GB, Yuan Y, 61. Mian C, Barollo S, Zambonin L, Pennelli G, Bernante P, Pelizzo
et al. Immune microenvironment in microsatellite-instable endo- MR, et al. Characterization of the largest kindred with MEN2A due
metrial cancers: hereditary or sporadic origin matters. Clin Cancer to a Cys609Ser RET mutation. Familial Cancer. 2009;8(4):379–82.
Res. 2017;23(15):4473–81. https://doi.org/10.1158/1078-­0432. https://doi.org/10.1007/s10689-­009-­9250-­z.
Ccr-­16-­2655. 62. Szinnai G, Meier C, Komminoth P, Zumsteg UW. Review of
47. Thakker RV. Multiple endocrine neoplasia type 1 (MEN1) and type multiple endocrine neoplasia type 2A in children: therapeu-
4 (MEN4). Mol Cell Endocrinol. 2014;386(1–2):2–15. https://doi. tic results of early thyroidectomy and prognostic value of codon
org/10.1016/j.mce.2013.08.002. analysis. Pediatrics. 2003;111(2):E132–9. https://doi.org/10.1542/
48. Giusti F, Brandi ML. Multiple endocrine neoplasia type 1. peds.111.2.e132.
GeneReviews 1993–2021; 2005, 2017 Dec 14. https://www.ncbi. 63. Moley J. Medullary thyroid carcinoma: management of lymph node
nlm.nih.gov/books/NBK1538/ metastases. J Natl Compr Cancer Netw. 2010;8:549–56. https://doi.
49. Niederle B, Selberherr A, Bartsch D, Brandi ML, Doherty GM, org/10.6004/jnccn.2010.0042.
Falconi M, et al. Multiple endocrine neoplasia type 1 (MEN1) and 64. Zenaty D, Aigrain Y, Peuchmaur M, Philippe-Chomette P,
the pancreas - diagnosis and treatment of functioning and non-­ Baumann C, Cornelis F, et al. Medullary thyroid carcinoma identi-
functioning pancreatic and duodenal neuroendocrine neoplasia fied within the first year of life in children with hereditary multiple
within the MEN1 syndrome - an international consensus statement. endocrine neoplasia type 2A (codon 634) and 2B. Eur J Endocrinol.
Neuroendocrinology. 2020; https://doi.org/10.1159/000511791. 2009;160(5):807–13. https://doi.org/10.1530/eje-­08-­0854.
52 B. M. Shaw and O. Kis

65. Wells SA. Advances in the management of MEN2: from improved 75. DeMichele A, Gray R, Horn M, Chen J, Aplenc R, Vaughan WP,
surgical and medical treatment to novel kinase inhibitors. Endocr Tallman MS. Host genetic variants in the interleukin-6 promoter
Relat Cancer. 2018;25(2):T1–T13. https://doi.org/10.1530/ predict poor outcome in patients with estrogen receptor-positive,
erc-­17-­0325. node-positive breast cancer. Cancer Res. 2009;69(10):4184–91.
66. Guy CT, Cardiff RD, Muller WJ. Induction of mammary tumors by https://doi.org/10.1158/0008-­5472.Can-­08-­2989.
expression of polyomavirus middle T oncogene: a transgenic mouse 76. Hefler LA, Grimm C, Ackermann S, Malur S, Radjabi-Rahat AR,
model for metastatic disease. Mol Cell Biol. 1992;12(3):954–61. Leodolter S, et al. An interleukin-6 gene promoter polymorphism
https://doi.org/10.1128/mcb.12.3.954-­961.1992. influences the biological phenotype of ovarian cancer. Cancer Res.
67. Hunter K. The intersection of inheritance and metastasis: the role 2003;63(12):3066–8.
and implications of germline polymorphism in tumor dissemina- 77. Hohaus S, Giachelia M, Di Febo A, Martini M, Massini G, Vannata
tion. Cell Cycle. 2005;4(12):1719–21. https://doi.org/10.4161/ B, et al. Polymorphism in cytokine genes as prognostic markers
cc.4.12.2258. in Hodgkin’s lymphoma. Ann Oncol. 2007;18(8):1376–81. https://
68. Hunter KW, Amin R, Deasy S, Ha NH, Wakefield L. Genetic doi.org/10.1093/annonc/mdm132.
insights into the morass of metastatic heterogeneity. Nat Rev 78. Lagmay JP, London WB, Gross TG, Termuhlen A, Sullivan N, Axel
Cancer. 2018;18(4):211–23. https://doi.org/10.1038/nrc.2017.126. A, et al. Prognostic significance of interleukin-6 single nucleotide
69. Crawford NP, Ziogas A, Peel DJ, Hess J, Anton-Culver H, Hunter polymorphism genotypes in neuroblastoma: rs1800795 (promoter)
KW. Germline polymorphisms in SIPA1 are associated with metas- and rs8192284 (receptor). Clin Cancer Res. 2009;15(16):5234–9.
tasis and other indicators of poor prognosis in breast cancer. Breast https://doi.org/10.1158/1078-­0432.Ccr-­08-­2953.
Cancer Res. 2006;8(2):R16. https://doi.org/10.1186/bcr1389. 79. Ruzzo A, Catalano V, Canestrari E, Giacomini E, Santini D, Tonini
70. Zhang Y, Gong Y, Hu D, Zhu P, Wang N, Zhang Q, et al. Nuclear G, et al. Genetic modulation of the interleukin 6 (IL-6) system in
SIPA1 activates integrin β1 promoter and promotes invasion of patients with advanced gastric cancer: a background for an alter-
breast cancer cells. Oncogene. 2015;34(11):1451–62. https://doi. native target therapy. BMC Cancer. 2014;14:357. https://doi.
org/10.1038/onc.2014.36. org/10.1186/1471-­2407-­14-­357.
71. Brooks R, Kizer N, Nguyen L, Jaishuen A, Wanat K, Nugent E, 80. Abana CO, Bingham BS, Cho JH, Graves AJ, Koyama T, Pilarski
et al. Polymorphisms in MMP9 and SIPA1 are associated with RT, et al. IL-6 variant is associated with metastasis in breast cancer
increased risk of nodal metastases in early-stage cervical cancer. patients. PLoS One. 2017;12(7):e0181725. https://doi.org/10.1371/
Gynecol Oncol. 2010;116(3):539–43. https://doi.org/10.1016/j. journal.pone.0181725.
ygyno.2009.09.037. 81. Ha NH, Long J, Cai Q, Shu XO, Hunter KW. The circadian rhythm
72. Gdowicz-Kłosok A, Giglok M, Drosik A, Suwiński R, gene Arntl2 is a metastasis susceptibility gene for estrogen receptor-­
Butkiewicz D. The SIPA1 -313A>G polymorphism is associ- negative breast cancer. PLoS Genet. 2016;12(9):e1006267. https://
ated with prognosis in inoperable non-small cell lung cancer. doi.org/10.1371/journal.pgen.1006267.
Tumour Biol. 2015;36(2):1273–8. https://doi.org/10.1007/ 82. Brady JJ, Chuang CH, Greenside PG, Rogers ZN, Murray
s13277-­014-­2753-­8. CW, Caswell DR, et al. An Arntl2-driven Secretome enables
73. Xie C, Yang L, Yang X, Yang R, Li Y, Qiu F, et al. Sipa1 pro- lung adenocarcinoma metastatic self-sufficiency. Cancer Cell.
moter polymorphism predicts risk and metastasis of lung cancer 2016;29(5):697–710. https://doi.org/10.1016/j.ccell.2016.03.003.
in Chinese. Mol Carcinog. 2013;52(Suppl 1):E110–7. https://doi. 83. Ledet EM, Antonarakis ES, Isaacs WB, Lotan TL, Pritchard C,
org/10.1002/mc.22039. Sartor AO. Germline BLM mutations and metastatic prostate
74. Crawford NP, Qian X, Ziogas A, Papageorge AG, Boersma BJ, cancer. Prostate. 2020;80(2):235–7. https://doi.org/10.1002/
Walker RC, et al. Rrp1b, a new candidate susceptibility gene pros.23924.
for breast cancer progression and metastasis. PLoS Genet. 84. Maher B. Personal genomes: the case of the missing heritability.
2007;3(11):e214. https://doi.org/10.1371/journal.pgen.0030214. Nature. 2008;456(7218):18–21. https://doi.org/10.1038/456018a.
 Part II
Pathology of Cancer Growth and Metastasis

Jane L. Messina

The early medical literature is rife with clinical descriptions of patients with advanced cancer,
from the observation in the Edwin Smith Papyrus that a bulging tumor of the breast was a grave
disease for which there is no treatment to the observations by Aulus Celsus that advanced
breast cancers have a propensity to spread to the armpit, and that spreading to distant organs
may cause death [1]. In 1842, William Budd was the first to note that not only tumor could
spread to lymph nodes in the region of the tumor, but also cancer was possessed of such powers
that it could be preserved “unimpaired when detached from the parent growth, and carried to
distant parts of the organism” [2]. 150 years later, Moreno et al. described a novel pattern of
spread of melanoma, described as demonstrating a “striking tropism for small and medium-­
sized vessels, primarily veins … proliferation within the vessel walls, expanding and replacing
the muscle layer but sparing the endothelia,” for which they proposed the term angiotropic
malignant melanoma [3].
In the ensuing years, knowledge about the mechanisms of the three mainstays of cancer
dissemination—lymphatic metastasis with nodal involvement, systemic metastasis through
vasculature, and angiotropic metastasis along the perivascular spaces—exploded. The first
chapters of this textbook describe the genetics, molecular pathways, and biology that govern
each type of metastasis.
This section will focus on the pathologic underpinnings of each of these three types of can-
cer growth and spread. Beginning with the clinical presentation and significance of the respec-
tive types of spread, their role in the staging and prognostication of cancer will be described.
The histologic hallmarks of tumor in lymph nodes, systemic/visceral spread, and perivascular
spaces will be described. Since many tumors present with metastatic disease in lymph nodes
and/or distant organs, the pathologic workup of these tumors will be described. Specific can-
cers have specific propensity to metastasize by certain routes, and knowledge of these can help
navigate diagnostic dilemmas. The pathologic workup and reporting of nodal disease identi-
fied at the time of surgical resection of a primary tumor will be addressed, including details of
sentinel node reporting for both melanoma and breast cancer. The workup of a tumor present-
ing as an unknown primary tumor, both as nodal disease and systemically, will be discussed.
The chapter on systemic metastasis contains an exhaustive description of methods of histo-
logic and immunohistochemical presentation that can be extrapolated to all cancer types. More
recently developed molecular methods of lymph node and systemic metastases will be
addressed.
Finally, the mechanisms of extravascular migratory metastasis (“angiotropic metastasis”)
will be dissected. Although it is the least common form of metastatic spread, it represents a
unique method in which tumor cells, particularly melanoma, can spread along low-resistance
vascular channels, similar to migration of melanoblasts along the neural crest. Identification of
54 Pathology of Cancer Growth and Metastasis

the distinctive pathologic pattern of this type of spread has been shown to be as prognostically significant
as nodal and systemic spread in cutaneous and uveal melanoma. Notably, it is the most common pattern of
melanoma involvement of the liver. Furthermore, this mechanism, which is slower than intravascular metas-
tasis, may explain a subset of tumors with delayed metastatic presentation. It is our hope that this section
provides the clinician with a glimpse of the role of a pathologist in diagnosis and reporting of metastatic
disease in its myriad presentation.

References
1. Hajdu SI. A note from history: landmarks in history of cancer, part 1. Cancer. 2011;117(5):1097–102.
2. Budd W. Remarks on the pathology and causes of cancer. Lancet. 1842;2:266–70.
3. Moreno A, Español I, Romagosa V. Angiotropic malignant melanoma. Report of two cases. J Cutan
Pathol. 1992;19(4):325–9.
 Pathologic Assessment of Lymph Node
Metastasis 6
James Isom and Jane L. Messina

Abstract
Learning Objectives
The pathologic assessment and evaluation of lymph node • Learn the frequency of nodal involvement in
status plays a central role in the diagnosis, staging, and patients with malignancy, both at presentation and
management of most malignancies. The pathologic status during the course of their disease.
of the lymph node plays a central role in the staging, and • Learn the clinical relevance of detection of nodal
thus subsequent treatment of diseases, which is codified disease at the time of diagnosis as well as during the
in the AJCC/UICC staging system. Lymph node metasta- course of disease.
sis may be subclinical and detected microscopically dur- • Describe the most common tumors to metastasize
ing surgical resection of a sentinel node or regional to lymph nodes and the pathologic workup of these
lymphadenectomy for initial therapy, or it may present tumors.
clinically as the first manifestation of a metastatic malig- • Explore newer methods of evaluation of lymph
nancy. Different protocols exist for the pathologic evalua- nodes.
tion of lymph nodes in these distinct settings. These
involve systematic gross assessment, careful light micro-
scopic evaluation, judicious and sometimes stepwise
application of immunohistochemical methods, and accu- 6.1 Introduction
rate pathologic reporting. The evaluation of a lymph node
metastasis in a patient with unknown primary malignancy The clinical importance of metastatic spread of cancer has
requires knowledge of the relative frequencies of malig- been recognized for centuries, and the importance of nodal
nancies, their most common sites of presentation, and an involvement looms large in many historical descriptions of
algorithmic approach to immunohistochemical testing. cancer. For example, the “fatal black tumor” described by
Molecular analysis by gene expression profiling offers Hippocrates in 5000 BC [1] that was later termed “mela-
modest improvement in diagnostic accuracy, but these nose” by Rene Laennec in 1804 was recognized to have the
improvements have not yet been translated into clinical propensity to metastasize by William Norris in 1820 [2]. And
improvements. while the prognosis of melanoma was uniformly acknowl-
edged to be dire, Samuel Cooper described surgical removal
J. Isom as having the potential to improve prognosis [3]. After this,
Department of Dermatology and Cutaneous Surgery, in 1898 Herbert Snow was the first to recognize the potential
University of South Florida Morsani College of Medicine, clinical benefit of removing clinically normal lymph nodes at
Tampa, FL, USA
the time of initial cancer surgery. In describing “anticipatory
J. L. Messina (*) gland excision,” he noted that “it is essential to remove,
Department of Dermatology and Cutaneous Surgery,
University of South Florida Morsani College of Medicine, whenever possible, those lymph glands which first receive
Tampa, FL, USA the infective protoplasm, and bar its entrance into the blood,
Departments of Pathology and Cutaneous Oncology, before they have undergone increase in bulk” [4]. These
Moffitt Cancer Center, Tampa, FL, USA early studies solidified the central role of lymph node
Departments of Pathology and Cell Biology and Oncologic involvement in tumor staging and prognostication. They also
Sciences, University of South Florida Morsani College of set the stage for the debate that continues today, with persis-
Medicine, Tampa, FL, USA tent questions. Are regional lymph nodes the first stop for
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 55

S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_6
56 J. Isom and J. L. Messina

metastases (“incubator” hypothesis) or does their involve- groin in 35% of the 80 patients with penile squamous cell
ment serve as an indicator of aggressive disease with inher- carcinoma in this study; patients with involved SLN had a
ent metastatic potential (“marker” hypothesis) [5]? Is there a worse five-year overall survival compared to those with neg-
better way than nodal staging to improve prediction of dis- ative nodes (70% v 90%) [10]. 15 years later, Donald Morton
ease outcome? Does lymph node removal improve disease provided the first description of the operative technique in
outcome? early-stage melanoma patients [11]. Over the ensuing
This chapter will address clinically important questions 28 years, the technique of preoperative lymphoscintigraphy
related to cancer metastasis to lymph nodes, focusing on the using a radiolabeled tracer (typically technetium-99 labeled
clinical relevance of metastatic nodal disease, and the clini- sulfur colloid) followed by perioperative injection of blue
copathologic workup of patients who present with disease in dye and intraoperative gamma probe detection has become
their lymph nodes, either clinically or microscopically the standard of care in the initial treatment of patients with
detected via sentinel node biopsy. The detection and signifi- clinically localized melanoma stage 1b (>0.8 mm in thick-
cance of nodal metastasis discovered through sentinel lymph ness) or greater [12]. According to this procedure, all lymph
node (SLN) evaluation will also be covered. A diagnostic nodes that exceed institutionally set thresholds for radioac-
algorithm for evaluation of patients presenting with nodal tivity and/or blue coloration are surgically removed and
metastasis is presented. Recently developed ancillary tech- labeled as a SLN.
niques for evaluation of lymph nodes will be covered, and SLN examination in patients with breast cancer became
pathologic changes in nodes harboring metastasis will be standard of care in parallel with its use in melanoma patients,
considered. and the technical details are similar and have been exhaus-
tively described elsewhere [13]. SLN biopsy has also been
applied to a variety of other tumors, chiefly those of skin,
6.2 Importance of the Lymph Node lung, and upper and lower gastrointestinal tract [14]. In these
in Tumor Staging instances, patients can also derive the benefits of improved
staging and identification of non-classic draining nodal
The centrality of lymph node status in accurate tumor stag- basins. Further, the refinement of nodal identification meth-
ing was solidified in the development of the tumor-node-­ ods using novel tracers such as fluorescent dyes continues to
metastasis (TNM) staging system by Pierre Denoix in 1952 drive knowledge in the field and expansion of the indications
[6]. This staging system was first adopted by the International for this procedure [15, 16].
Union Against Cancer (UICC), which then unified with the Currently, tumor staging using the results of SLN biopsy
American Joint Committee on Cancer (AJCC) in 1987 [7]. is employed in breast carcinoma, melanoma, and Merkel cell
Current TNM classification schemes for all tumors in which carcinoma [8]. For other tumors undergoing surgical resec-
nodal resection is part of the initial surgical therapy require tion, such as head and neck, lung, gastrointestinal, genitouri-
quantification of the number of regional lymph nodes with nary, and gynecologic tumors, lymph node staging continues
tumor involvement. Most staging schema contain pathologic to be performed in the form of regional node dissection at the
subclassifications describing the extent of lymph node time of initial surgery. In all of these cases, the evaluation of
involvement after surgical resection or biopsy sampling, lymph nodes is a crucial component of cancer staging using
such as tumor size, and whether tumor was identified by the TNM classifications. In addition, the College of American
immunohistochemical or molecular methods [8]. Considering Pathologists cancer reporting protocols, which include the
the importance of lymph status for prognosis, therefore, American Joint Committee on Cancer’s most recent recom-
almost all attempts at curative surgical resection of malig- mendations, require evaluation and reporting of lymph node
nancy include removal of both clinically negative and posi- status [17].
tive regional lymph node basins [9]. However, the morbidity
of these procedures, and the ability of tumors to metastasize
to non-predicted sites, led to attempts to refine surgery for 6.3 Tumors Most Frequently Involving
regional lymph node involvement. In 1977, Ramon Cabanas Lymph Nodes
was the first to utilize the term “sentinel lymph node” (SLN)
in his seminal work demonstrating that lymphangiography The incidence of lymph node involvement at the time of
could be used to identify patterns of nodal drainage from the definitive primary surgical excision for the most common
penis, and hypothesizing that this “lymph node center” cancers are: prostate (5%) [18], breast (25%) [19], lung and
appears to be the primary site of metastasis from penile car- bronchus (15%) [20], colorectal (33%) [21], urinary bladder
cinoma. A tumor-involved lymph node was found in the (25%) [22], corpus and uterus (21%) [23], thyroid (35%)
6 Pathologic Assessment of Lymph Node Metastasis 57

Table 6.1 Location at presentation of lymph node metastases and lymph nodes may include the use of immunohistochemical
most common anatomic site of primary tumor, adapted from [50, 67] evaluation, which has been shown to increase detection of
Site of lymph node Most common primary site tumor in a variety of settings, including breast [30], mela-
Cervical-upper Head and neck noma [31], uterine cervix [32], Merkel cell carcinoma [33],
Lymphoma
and many others. However, current AJCC guidelines do not
Cervical-lower Lymphoma
Lung require specialized immunohistochemical or molecular tech-
Stomach niques for routine examination of lymph node resections [8].
Pancreas Accurate staging of lymph node status requires reporting not
Axillary Breast only nodal positivity or negativity, but the total number of
Melanoma
Lung nodes identified as positive and negative. For each tumor-
Thorax Lung involved lymph node, the size of the largest metastasis
Breast should be assessed and further subcategorizes the N status of
Abdomen Pancreas the tumor. For example, nodal metastases of breast, colorec-
Lower gastrointestinal
tal, and gynecologic cancers use a tumor size of 0.2 cm for
Upper gastrointestinal
Kidney different subcategories of nodal involvement. Head and neck
Inguinal Gynecologic cancers use a 3 cm cutoff, and testicular tumors have 2 and
Genitourinary 5 cm nodal metastasis size categories [8]. The presence of
Melanoma extracapsular tumor extension must always be noted. Of
note, the definition of extracapsular extension is a subject of
[24], and melanoma (18%) [25]. It is worth noting that carci- debate. Most definitions include the requirement that the
noma is the dominant cancer morphology in each of these tumor extends through the lymph node capsule into the sur-
sites, in keeping with the theory that carcinoma favors lym- rounding connective tissue, with or without a stromal reac-
phatic over hematogenous spread. By contrast, hematoge- tion. However, in cases where incomplete lymph node
nous spread is believed to be the preferred pathway of sections do not demonstrate the capsule, or there is a fibrous
sarcoma, and lymph node metastases are exceedingly rare in reaction to extranodal tumor simulating a lymph node cap-
sarcoma cases [26]. sule, this can be difficult [34].
Certain primary tumors are more likely to be found in dis- Since pathologic evaluation of SLN in melanoma and
tinct anatomic locations (Table 6.1). Among tumors present- breast cancer is so pivotal to prognostication and refinement
ing as lymph node metastases, the most common causes of of treatment, immunohistochemical examination is consid-
death are lung cancer (60%), followed by pancreatic, upper ered a routine part of the workup. However, in the United
aerodigestive tract, liver, and breast cancer [27]. States there is wide variation among institutions with respect
to protocols for examination of SLN, including gross exami-
nation, number of sections examined, and number and type
6.4  athologic Analysis of Lymph Nodes
P of immunohistochemical stain performed. In melanoma,
in the Setting of a Known Primary S-100 is most commonly used, followed by Melan-A/Mart-1
Tumor [35, 36]. Current AJCC/UICC guidelines recognize two cat-
egories of SLN involvement in melanoma: clinically occult
This section will address the pathologic evaluation of lymph micrometastasis and clinically apparent macrometastasis
nodes containing metastasis in two situations: (1) evaluation [17]. Micrometastatic melanoma is defined as the presence
of lymph node(s) removed during definitive resection of the of morphologically malignant cells positive for at least one
primary tumor and (2) evaluation of the SLN. Histologic immunohistochemical marker (S-100, HMB-45, Melan-A/
analysis of lymph nodes harboring metastatic disease has Mart1, Sox10) and/or malignant melanoma detectable on
informed our knowledge of how disease spreads via lym- H&E staining alone [17]. One important pitfall in the diag-
phatics. The earliest pattern of nodal metastasis occurs when nosis of SLN in patients with melanoma is the presence of
tumor enters the lymph node via lymphatics, breaches the benign melanocytes in the connective tissue capsule or tra-
sinusoidal walls, and appears in the subcapsular sinus. Once beculae of the node. Known as nodal nevi or nevus cell
inside the lymph node, tumors that escape immune surveil- aggregates, these may be found in up to 6.3% of all regional
lance set up an immunosuppressive niche, adapt metaboli- lymph nodes evaluated by melanocytic immunohistochemi-
cally, and can continue to expand [28, 29]. cal stains [37] and up to 22% of melanoma cases [38]. They
Evaluation of lymph nodes resected in the course of surgi- are characteristically located in the fibrous capsule or intra-
cal treatment of a malignancy follows a set of general rules nodal fibrous trabeculae, although they may occasionally be
for all types of cancers. The pathologic evaluation of these found in the nodal parenchyma [39]. The origin of benign
58 J. Isom and J. L. Messina

melanocytes in lymph nodes has been debated, and is thought involvement, while in some a single focus of disease is
to either represent rests of embryonic origin or “benign found. In these latter cases, the most common sites of pre-
metastases.” While they may be difficult to distinguish from sentation are lymph nodes, lung, bone, and liver [47].
metastatic melanoma, differential immunohistochemical Approximately 10–40% of patients with CUP have lymph
expression of antibodies typically negative in melanoma node involvement at presentation [48]. The etiology of this
(p16 [40]) or positive in melanoma but not nevi (HMB-45 phenomenon has been debated. The most common hypothe-
[38], PRAME [41]) is useful. Interestingly, some have sis is that primary tumors contain a mixture of cells with
hypothesized that nodal nevi may be a source of primary varying degrees of motility, and that in patients who present
nodal metastatic melanoma presenting as metastasis of as CUP, the more stationary cells regressed through immu-
unknown primary, but this has never been proven [42]. nologic surveillance at the primary site, while the more
In contrast, breast cancer SLN involvement is catego- motile ones present at the metastatic site [42]. The prognosis
rized as either micrometastatic or macrometastatic based on of patients with CUP is poor, with a median overall survival
size of metastasis and number of cells. There are three of 3–10 months [49]. It is noteworthy that patients with
pathologic categories of SLN metastases in breast carci- lymph node involvement alone have a better prognosis than
noma. Macrometastases are defined as tumor deposits patients with multiple sites of involvement [50]. In addition
>2 mm, while micrometastases detected by H&E and/or to patients who present with nodal metastasis, there are sev-
IHC are >0.2–2 mm and/or >200 cells; both are staged as eral well-defined subsets of CUP, constituting about 20% of
N1. Lymph nodes are staged as N0(i-) when no metastases patients overall, that have a more favorable prognosis. Nodal
are seen histologically or with IHC, and N0(i+) when seen presentations with a more favorable outcome include poorly
on IHC alone. Per AJCC/UICC guidelines, tumor cell must differentiated carcinoma involving midline lymph nodes,
be enumerated in a single histologic section, regardless of and squamous cell carcinoma in cervical nodes or as isolated
whether multiple sections were performed, rather than inguinal nodes [51, 52].
added together. The pathology report must also state whether There is a well-defined clinical and diagnostic algorithmic
or not IHC was performed [43]. Numerous studies have approach to patients who present with CUP, including nodal
documented the increased yield of metastasis when serial metastasis. An understanding of the relative frequency of
sectioning and immunohistochemical staining, predomi- malignancies overall and their most common patterns of
nantly with broad spectrum anti-cytokeratin antibodies, are spread is essential to the evaluation of these patients and
employed [44]. The current controversies concerning the effective deployment of this algorithm. Improvements in
clinical relevance of micrometastases as well as the utility diagnostic testing now allow a definitive identification of the
of ALND in this scenario have led to confusion concerning primary tumor in up to 75% of patients [53]. In one system-
the optimum pathologic evaluation of the SLN. Since the atic review of 884 CUP patients evaluated at autopsy or with
presence of micrometastases/ITC may neither influence gene expression profiling, a site of primary origin could be
decision for further therapy nor impact survival, it has been identified in 73% of cases. The most common sites of origin
argued that serial sections and IHC have no use in routine were lung (24%) and pancreas (24%), followed by hepatobili-
evaluation of breast SLN. Since it has been proven that a ary tree (8%), kidneys (8%), bowel, genital system, and stom-
single H&E-­stained full-face section from each submitted ach [54]. Interestingly, the incidence of a CUP presentation
SLN block can identify macrometastases and a high propor- has decreased in recent years. Improvements in diagnostic
tion of micrometastases, current CAP and ASCO guidelines testing are the most likely cause for this recent decrease; for
do not recommend of the SLN by multilevel node sectioning example, age-standardized incidence rates in a Switzerland
or require specialized techniques for evaluation of the SLN cancer registry rose from 10.3 to 17.6 per 100,000 from 1981
at this time [45]. to 1997, but thereafter decreased to 5.8 in 2014 [55].
Lymph node metastasis of unknown primary origin are
identified because of clinical and imaging findings suspi-
6.5  valuation of the Lymph Node
E cious for malignancy with a resulting lymph node biopsy, or
Metastasis in Patients with Unknown as an unexpected finding during the histologic evaluation of
Primary Tumors a specimen for an unrelated pathology. The preferred method
is core needle biopsy or fine needle aspiration with a cell
According to the American Cancer Society, 1.7% of newly block [56] [57]. The most common tumor morphologies are
diagnosed cancer cases in the United States in 2021 pre- carcinoma, sarcoma, lymphoma, and melanoma. Beyond
sented as a cancer of unknown primary origin (CUP) [46]. these, less common morphologies include germ cell, meso-
Greater than 50% of patients present with multiple sites of thelioma, and pheochromocytoma/paraganglioma. In the set-
6 Pathologic Assessment of Lymph Node Metastasis 59

ting of an undifferentiated malignant neoplasm, the tumor is patient’s condition and urgency of a diagnosis, as well as
overwhelmingly likely to be carcinoma (80%)—according pathologist training. The tiered approach is generally recom-
to a recent study [58]. Discerning the location of the primary mended. The first round is employed to determine the gen-
lesion can be extremely challenging. In such cases, a defini- eral lineage of the tumor (carcinoma, sarcoma, lymphoma,
tive diagnosis is not always possible, but a limited differen- melanoma, germ cell tumor, and mesothelioma [61]. If nec-
tial can help clinical colleagues with further workup. essary, the second round is performed to refine the primary
When assessing a lymph node with metastasis from site or subtype of sarcoma or lymphoma. Once the diagnosis
unknown primary, the growth pattern and morphology of is established, additional biomarker staining may also be
the tumor cells is the first feature to be examined, and in useful, such as hormone receptor assays and HER2 evalua-
well-­differentiated lesions, it can be sufficient. Broadly, tion in breast carcinoma. Table 6.2 lists the most commonly
there are three morphologies: epithelioid (cohesive, large) used Tier 1 and Tier 2 markers.
cell, spindle cell, and small round cell. Adenocarcinoma is In summary, the pathologist has two primary functions
the prototype of epithelioid cell, and the most common with regard to lymph node evaluation: identifying lymph
subtype of carcinoma to metastasize, followed by squa- node involvement and determining the primary lesion if it is
mous cell carcinoma [59]. Well-differentiated adenocarci- unknown. The former requires a meticulous work at all
noma may form glands, tubules, and/or papillae. The stages, including gross examination, dissection, embedding,
presence of mucin (cytoplasmic or extracellular) can aid in staining, and histologic examination—and immunohisto-
the diagnosis of adenocarcinoma [58]. The accumulation chemical studies in some cases. The latter requires no less
of keratin within the cytoplasm of squamous cell carci- but also demands a deep understanding of tumor biology and
noma may confer an opaque eosinophilic “glassy” appear- immunoprofiles.
ance. Additional features such as the presence of
intercellular bridges and keratin pearls can further help
identify squamous cell carcinoma. Histologic clues to the 6.6  lternative Methods of Lymph Node
A
diagnosis of melanoma include cytoplasmic melanin, Evaluation
nested growth pattern, eosinophilic nucleoli and nuclear
pseudoinclusions, but melanin production can appear in Recent advances in molecular analysis have provided addi-
other tumors such as psammomatous melanotic schwan- tional diagnostic and therapeutic tools for use in the setting
noma. The many histomorphologies of melanoma indicate of patients with lymph node metastasis of unknown origin.
that it should be included in nearly every differential diag- There are several commercially available diagnostic tests
nosis when a tumor of unknown primary origin is being that employ gene expression profiling to examine similarity
assessed [60]. Typically, however, pure morphologic of the mRNA of a patient’s tumor to a panel of tumors with
assessment is insufficient to establish a diagnosis. In these known sites of origin. Commonly used tests include
cases, immunohistochemical studies are employed. Recent CancerTypeID (Biotheranostics) and Tissue of Origin
years have seen an explosion of markers touted to be lin- (Cancer Genetics). A recent review of validation analyses for
eage specific, and the repertoire of stains available has these tests shows sensitivity in the range of ~80–90% and
greatly expanded. A systematic approach to malignancy of sensitivity of >98% for this test using cases with known
unknown primary origin requires a stepwise approach to diagnoses [62]. The accuracy of CancerTypeID was signifi-
performing these stains, combined with knowledge of the cantly better than immunohistochemistry using a blinded set
most common tumors arising in that region. of 122 tumors of known origin, with an accuracy of 79%
It has been estimated that well-performed and adequately compared to 60% (p = 0.019) [63].The Tissue of Origin test
interpreted IHC panels can aid in identification of the pri- accuracy was more similar to immunohistochemistry in a set
mary site of origin in 75% of cases [53]. Two general schools of well-differentiated tumors (85.3 v 86.8%) but fared better
of practice exist: a stepwise, directed approach that by defini- in poorly or undifferentiated tumors (94% vs. 79.1%) [64].
tion may take several rounds of staining, or a “shotgun” Despite these improvements in diagnostic accuracy, how-
approach with large numbers of stains ordered simultane- ever, evidence that these tests result in improved clinical out-
ously. The choice of either method may be influenced by the comes is slim [65, 66].
60 J. Isom and J. L. Messina

Table 6.2 Algorithmic immunohistochemical evaluation for evaluation of undifferentiated cancer of unknown primary origin presenting as lymph
node metastasis
Step 1-Determination of major lineage of tumor
Immunohistochemical Histology/diagnosis
stains
Step 1-evaluation of lineage of tumor
Sox10 Melanoma
Pankeratin (AE1/AE3/ Carcinoma
CAM5.2)
CK5/6, p63, or p40 Squamous cell carcinoma
LCA Lymphoma
OCT 3/4 or SALL4 Germ cell tumor
Step 2-evaluation of carcinoma of unknown origin with CK7 and CK20 (80% of cases)
CK7+/CK20- Breast. lung, thyroid, Mullerian (endometrial, endocervical, ovarian
serous), bladder, upper GI, pancreatobiliary, mucinous ovarian
CK7+/CK20+ Bladder, upper GI, pancreatobiliary, mucinous ovarian, occ. rectum
and lung
CK7-/CK20+ Colon, Merkel cell, occ. upper GI
CK7-/CK20- Prostate, SCC, renal, adrenocortical, hepatocellular
Step 3-additional markers to clarify origin of most common tumors
CK7+/CK20- Breast carcinoma GATA3, GCDFP-15, mammaglobin, ER/
PR
Lung adenocarcinoma TTF1, Napsin 1
GATA3, GCDFP-15, mammaglobin, ER,
PR
Thyroid carcinoma TTF-1, PAX8
Endocervical adenocarcinoma P16, PAX8
Endometrial adenocarcinoma Vimentin, PAX8
Urothelial carcinoma (subset) GATA3, p63/p40, CK5/6, uroplakin II
Upper GI CDX2, villin
Pancreatobiliary
Serous ovarian PAX8/WT1
Salivary gland carcinoma CK5/6, p63, GATA3, AR
CK7+/CK20 variable Mucinous ovarian PAX8+/−, CDX2+/−
Pancreatic adenocarcinoma (~2/3) CDX2, CK19, SMAD4 loss CEA, CA19–9,
MUC5, CDX2 (variable)
CK7-/CK20+ Lower GI tract adenocarcinoma (small intestine, appendix, colon, CDX2, villin, SATB2
rectum)
Merkel cell carcinoma (CK20 perinuclear dots) Chromogranin, synaptophysin, CM2B4,
TTF1-negative
CK7+/CK20+ Urothelial P63, CK4/5, uroplakin, GATA3
Mucinous ovarian PAX8/CDK2
CK7-/CK20- Prostatic adenocarcinoma NKX3.-1, PSAP, PSA
Squamous cell carcinoma CK5/6 or p63 or p40
Renal cell carcinoma PAX2, PAX8, CAIX, EMA, vimentin,
CD10
Adrenocortical carcinoma SF-1, Melan-A, inhibin
Hepatocellular carcinoma (may be CK7 or CDK20+) Arginase-1, HepPar-1, Glypican-3, CD30,
pericanalicular CEA
6 Pathologic Assessment of Lymph Node Metastasis 61

16. Vuijk FA, et al. Fluorescent-guided surgery for sentinel lymph node
Conclusions and Future Perspectives detection in gastric cancer and carcinoembryonic antigen targeted
fluorescent-guided surgery in colorectal and pancreatic cancer. J
• Accurate assessment and characterization of lymph Surg Oncol. 2018;118(2):315–23.
node involvement by metastatic disease is an inte- 17. Cancer Protocol Templates. 2021 [cited 2021; Available from:
gral part of patient staging and a key driver of https://www.cap.org/protocols-­and-­guidelines/cancer-­reporting-
patient outcome. ­tools/cancer-­protocol-­templates.
18. Narayanan R, Wilson TG. Sentinel node evaluation in prostate can-
• Focused examination of the sentinel lymph node cer. Clin Exp Metastasis. 2018;35(5–6):471–85.
provides a precise and accurate method of lymph 19. Cotarelo CL, et al. Improved detection of sentinel lymph node
node staging in patients with breast, melanoma, and metastases allows reliable intraoperative identification of patients
Merkel cell carcinoma. with extended axillary lymph node involvement in early breast can-
cer. Clin Exp Metastasis. 2021;38(1):61–72.
• The lymph node is a common site of metastasis in 20. Osarogiagbon RU, et al. Survival implications of variation in the
patients presenting with a malignancy of unknown thoroughness of pathologic lymph node examination in American
primary origin. College of Surgeons Oncology Group Z0030 (Alliance). Ann
• An algorithmic approach utilizing knowledge of Thorac Surg. 2016;102(2):363–9.
21. Gleisner AL, et al. Nodal status, number of lymph nodes exam-
frequent sites of clinical presentation and a tiered ined, and lymph node ratio: what defines prognosis after resection
approach to immunohistochemical analysis can aid of colon adenocarcinoma? J Am Coll Surg. 2013;217(6):1090–100.
in accurate diagnosis. 22. Wright JL, Lin DW, Porter MP. The association between
• The role of molecular testing in evaluation of extent of lymphadenectomy and survival among patients with
lymph node metastases undergoing radical cystectomy. Cancer.
patients with lymph node metastasis is an area for 2008;112(11):2401–8.
future investigation. 23. Chan JK, et al. Metastatic gynecologic malignancies:
advances in treatment and management. Clin Exp Metastasis.
2018;35(5–6):521–33.
24. Mansour J, et al. Prognostic value of lymph node ratio in metastatic
References papillary thyroid carcinoma. J Laryngol Otol. 2018;132(1):8–13.
25. Mozzillo N, et al. Sentinel node biopsy in thin and thick melanoma.
1. Urteaga O, Pack GT. On the antiquity of melanoma. Cancer. Ann Surg Oncol. 2013;20(8):2780–6.
1966;19(5):607–10. 26. Leong SP, et al. Clinical patterns of metastasis. Cancer Metastasis
2. Gorantla VC, Kirkwood JM. State of melanoma: an historic Rev. 2006;25(2):221–32.
overview of a field in transition. Hematol Oncol Clin North Am. 27. Hemminki K, et al. Site-specific cancer deaths in cancer of unknown
2014;28(3):415–35. primary diagnosed with lymph node metastasis may reveal hidden
3. Cooper S. The first lines of the theory and practice of surgery. primaries. Int J Cancer. 2013;132(4):944–50.
London: Longman; 1840. 28. Kawaguchi T. Pathological features of lymph node metasta-
4. Snow HM. Melanoma cancerous disease. Lancet. sis. 2 from morphological aspects. Nihon Geka Gakkai Zasshi.
1892;140:869–922. 2001;102(6):440–4.
5. Faries MB, et al. Lymph node metastasis in melanoma: a debate on 29. Zhou H, Lei PJ, Padera TP. Progression of metastasis through lym-
the significance of nodal metastases, conditional survival analysis phatic system. Cell. 2021;10(3)
and clinical trials. Clin Exp Metastasis. 2018;35(5–6):431–42. 30. Cote RJ, et al. Role of immunohistochemical detection of lymph-­
6. Denoix P. Nomenclature classification des cancers. Bull Inst Nat node metastases in management of breast cancer. International
Hyg (Paris). 1952;7:743–8. Breast Cancer Study Group. Lancet. 1999;354(9182):896–900.
7. Hutter RV. At last—worldwide agreement on the staging of cancer. 31. Messina JL, et al. Pathologic examination of the sentinel lymph node
Arch Surg. 1987;122(11):1235–9. in malignant melanoma. Am J Surg Pathol. 1999;23(6):686–90.
8. Amin MB. AJCC cancer staging manual. 8th ed. Switzerland: 32. Euscher ED, et al. Ultrastaging improves detection of metastases in
Springer Nature; 2017. sentinel lymph nodes of uterine cervix squamous cell carcinoma.
9. Surgery basic science and clinical evidence. 2nd ed. Springer Am J Surg Pathol. 2008;32(9):1336–43.
Nature; 2008. 33. Su LD, et al. Immunostaining for cytokeratin 20 improves detection
10. Cabanas RM. An approach for the treatment of penile carcinoma. of micrometastatic Merkel cell carcinoma in sentinel lymph nodes.
Cancer. 1977;39(2):456–66. J Am Acad Dermatol. 2002;46(5):661–6.
11. Morton DL, et al. Technical details of intraoperative lymphatic 34. Abdel-Halim CN, et al. Histopathological definitions of extra-
mapping for early stage melanoma. Arch Surg. 1992;127(4):392–9. nodal extension: a systematic review. Head Neck Pathol.
12. Fayne RA, et al. Evolving management of positive regional lymph 2021;15(2):599–607.
nodes in melanoma: past, present and future directions. Oncol Rev. 35. Dekker J, Duncan LM. Lack of standards for the detection of mel-
2019;13(2):433. anoma in sentinel lymph nodes: a survey and recommendations.
13. Zeitoun J, Babin G, Lebrun JF. Sentinel node and breast can- Arch Pathol Lab Med. 2013;137(11):1603–9.
cer: a state-of-the-art in 2019. Gynecol Obstet Fertil Senol. 36. Cole CM, Ferringer T. Histopathologic evaluation of the sentinel
2019;47(6):522–6. lymph node for malignant melanoma: the unstandardized process.
14. Faries MB. Application of senitnel lymph node surgery outside of Am J Dermatopathol. 2014;36(1):80–7.
melanoma and breast cancer. Clin Exp Metastasis. 2021; 37. Bautista NC, Cohen S, Anders KH. Benign melanocytic nevus
15. Burghgraef TA, et al. In vivo sentinel lymph node identification cells in axillary lymph nodes. A prospective incidence and immu-
using fluorescent tracer imaging in colon cancer: a systematic review nohistochemical study with literature review. Am J Clin Pathol.
and meta-analysis. Crit Rev Oncol Hematol. 2021;158:103,149. 1994;102(1):102–8.
62 J. Isom and J. L. Messina

38. Carson KF, et al. Nodal nevi and cutaneous melanomas. Am J Surg 54. Pentheroudakis G, Golfinopoulos V, Pavlidis N. Switching bench-
Pathol. 1996;20(7):834–40. marks in cancer of unknown primary: from autopsy to microarray.
39. Biddle DA, et al. Intraparenchymal nevus cell aggregates in lymph Eur J Cancer. 2007;43(14):2026–36.
nodes: a possible diagnostic pitfall with malignant melanoma and 55. Binder C, et al. Cancer of unknown primary-epidemiological trends
carcinoma. Am J Surg Pathol. 2003;27(5):673–81. and relevance of comprehensive genomic profiling. Cancer Med.
40. Koh SS, Cassarino DS. Immunohistochemical expression of p16 in 2018;7(9):4814–24.
melanocytic lesions: an updated review and meta-analysis. Arch 56. Ettinger DS, et al. NCCN Clinical Practice Guidelines Occult pri-
Pathol Lab Med. 2018;142(7):815–28. mary. J Natl Compr Cancer Netw. 2011;9(12):1358–95.
41. Lezcano C, et al. Immunohistochemistry for PRAME in the distinc- 57. Bochtler T, Löffler H, Krämer A. Diagnosis and management of
tion of nodal nevi from metastatic melanoma. Am J Surg Pathol. metastatic neoplasms with unknown primary. Semin Diagn Pathol.
2020;44(4):503–8. 2018;35(3):199–206.
42. Kamposioras K, et al. Malignant melanoma of unknown primary 58. Bellizzi AM. An algorithmic Immunohistochemical approach
site. To make the long story short. A systematic review of the litera- to define tumor type and assign site of origin. Adv Anat Pathol.
ture. Crit Rev Oncol Hematol. 2011;78(2):112–26. 2020;27(3):114–63.
43. Lester S et al. Protocol for the examination of specimens from 59. Houston KA, et al. Patterns in lung cancer incidence rates and
patients with invasive carcinoma of the breast; December 2013. trends by histologic type in the United States, 2004-2009. Lung
44. Lyman GH, et al. Sentinel lymph node biopsy for patients with early-­ Cancer. 2014;86(1):22–8.
stage breast cancer: American Society of Clinical Oncology clinical 60. Banerjee SS, Harris M. Morphological and immunopheno-
practice guideline update. J Clin Oncol. 2014;32(13):1365–83. typic variations in malignant melanoma. Histopathology.
45. Wiatrek R, Kruper L. Sentinel lymph node biopsy indications and 2000;36(5):387–402.
controversies in breast cancer. Maturitas. 2011;69(1):7–10. 61. Pavlidis N. Cancer of unknown primary: biological and clinical
46. Siegel RL, et al. Cancer statistics, 2021. CA Cancer J Clin. characteristics. Ann Oncol. 2003;14(Suppl 3)):11–8.
2021;71(1):7–33. 62. Lee MS, Sanoff HK. Cancer of unknown primary. BMJ.
47. Buttar A, et al. Cancers of unknown primary. In: Pieters R, editor. 2020;371:m4050.
Cancer concepts: a guidebook for the non-oncologist. University of 63. Weiss LM, et al. Blinded comparator study of immunohistochemi-
Massachusetts Medical School; 2015. cal analysis versus a 92-gene cancer classifier in the diagnosis of
48. Hemminki K, et al. Survival in cancer of unknown primary site: the primary site in metastatic tumors. J Mol Diagn. 2013;15(2):
population-based analysis by site and histology. Ann Oncol. 263–9.
2012;23(7):1854–63. 64. Handorf CR, et al. A multicenter study directly comparing the
49. Losa F, et al. 2018 consensus statement by the Spanish Society of diagnostic accuracy of gene expression profiling and immunohisto-
Pathology and the Spanish Society of Medical Oncology on the chemistry for primary site identification in metastatic tumors. Am J
diagnosis and treatment of cancer of unknown primary. Clin Transl Surg Pathol. 2013;37(7):1067–75.
Oncol. 2018;20(11):1361–72. 65. Hainsworth JD, et al. Molecular gene expression profiling to pre-
50. Chorost MI, et al. Unknown primary. J Surg Oncol. dict the tissue of origin and direct site-specific therapy in patients
2004;87(4):191–203. with carcinoma of unknown primary site: a prospective trial of
51. Hainsworth JD, Fizazi K. Treatment for patients with unknown the Sarah Cannon research institute. J Clin Oncol. 2013;31(2):
primary cancer and favorable prognostic factors. Semin Oncol. 217–23.
2009;36(1):44–51. 66. Hayashi H, et al. Randomized phase II trial comparing site-specific
52. Jereczek-Fossa BA, Jassem J, Orecchia R. Cervical lymph node treatment based on gene expression profiling with carboplatin and
metastases of squamous cell carcinoma from an unknown primary. paclitaxel for patients with cancer of unknown primary site. J Clin
Cancer Treat Rev. 2004;30(2):153–64. Oncol. 2019;37(7):570–9.
53. Kandalaft PL, Gown AM. Practical applications in immunohisto- 67. Pavlidis N, Pentheroudakis G. Cancer of unknown primary site.
chemistry: carcinomas of unknown primary site. Arch Pathol Lab Lancet. 2012;379(9824):1428–35.
Med. 2016;140(6):508–23.
 Pathologic Assessment of Systemic
or Distant Metastasis 7
Igor Katsyv and Andrew Turk

Abstract prognosis for patients with disseminated disease remains

Pathologists’ microscopic examination of tissue is the poor, particularly for those with cancers of unknown primary
gold standard for cancer diagnosis. It begins with lesion (CUP) [2]. Rapid identification of tumor origin and selection
biopsy or resection and subsequently involves a time- and of optimal therapy is therefore critical. This chapter will
labor-intensive process of tissue processing, slide prepa- focus on the histologic and molecular methods that, together
ration, and ancillary testing before a final diagnosis can be with clinical and radiologic data, factor into the pathological
rendered. This ancillary testing includes characterization workup for systemic metastases.
of cancer- and organ-specific protein expression and
molecular alterations using immunohistochemical (IHC)
stains, and, increasingly, massively parallel sequencing to 7.2 Clinicopathologic Considerations
probe a tumor’s genomic, transcriptomic, and epigenomic and Correlation
landscape. Pathologists integrate these modalities into
unified diagnoses that form the basis for patients’ onco- For every case, consideration of factors related to pathogen-
logic care. esis occurs simultaneously with, and is inseparable from,
morphologic review of a specimen’s gross and microscopic
properties. Clinicopathologic features of specific cancers
affect the likelihood of various scenarios and inform a
Learning Objectives
pathologist’s differential diagnosis. Pertinent considerations
• Understand processing of tissue that occurs between
include the routes along which certain cancers metastasize
specimen procurement and histologic evaluation.
and the manner in which some metastases demonstrate a pre-
• Understand role of ancillary testing, such as immu-
dilection for stereotypical tissue types and/or anatomic sites.
nohistochemistry and molecular assays, in tumor
In terms of available mechanisms of metastasis, cancers
characterization and final diagnosis.
travel via lymphatic channels, vasculature (veins and arter-
• Appreciate the emergence of novel specimen types.
ies), and/or along nerves. Among these methods, some can-
cers spread preferentially by predictable avenues rather than
others. For instance, follicular carcinoma of the thyroid
7.1 Introduction migrates via hematogenous passageways only, and virtually
never involves lymph nodes [3, 4]. By contrast, Hurthle cell
Despite advances in early detection of cancer, patients often carcinoma—another malignancy derived from the same (fol-
present with metastatic disease at the time of initial diagno- licular epithelial) cell—travels via lymphatic as well as
sis. The proportion of stage IV disease at diagnosis varies hematogenous spread [3, 4]. Other cancers, such as prostatic
with malignancy, from less than 10% in prostate and female adenocarcinoma and certain salivary gland tumors, exhibit
breast cancers to more than 50% in cancers of the lung and high propensity for perineural invasion.
bronchus, pancreas, and ovary [1]. With few exceptions, the In addition to differences between canonical migratory
pathways, several cancers also demonstrate recurrent and
I. Katsyv · A. Turk (*) predictable metastatic sites. For example, prostate cancers
Department of Pathology and Cell Biology, Columbia University, tend to metastasize to bones, whereas breast cancers metas-
New York, NY, USA tasize to liver, bones, lung, pleura, and lymph nodes;
e-mail: [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 63

S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_7
64 I. Katsyv and A. Turk

melanomas tend to have increased risk for metastasis to the cation, with evaluation of tissue architecture. The pathologist
central nervous system, spleen, and heart compared to other then advances to medium- and high-power views, with
malignancies [5]. Lung cancers represent the most common examination of cellular and nuclear morphology in greater
metastases to central nervous system, and colorectal cancers detail. This procedure comprises a series of interrogations,
are the most frequently observed liver and, in men, perito- beginning with the most global and incrementally becoming
neal metastases [2, 5, 6]. The liver is also a common site of more granular. The first assessment distinguishes normal
uveal melanoma metastases [7, 8]. Among all solid tumors, from abnormal tissue. With abnormal specimens, the pathol-
hepatic and lymph node metastases are most common, fol- ogist then determines what category of disease accounts for
lowed by lung, bone, and pleura [5]. the irregularity, i.e., whether the lesion represents a neoplas-
Knowledge of these generalities (as well as the specific tic, inflammatory, or traumatic process, et cetera. For neopla-
clinical scenario of each particular case) is important prior to sia, subsequent distinctions parse primary versus metastatic,
and during review of tissue specimens, as these circumstances and (for the former) benign versus malignant tumors. Finally,
guide pathologists’ assessment, workup, and interpretation. before proceeding to ancillary testing, the pathologist gleans
any apparent indication of tumor type and/or primary site. In
general terms, broad categories of malignancy include carci-
7.3 Histopathology noma (of epithelial origin), sarcoma (of mesenchymal deri-
vation), melanoma, and lymphoma, among others. These
7.3.1 Specimen and Slide Preparation lineages exhibit stereotypical architectural and cytologic fea-
tures, albeit to variable degrees. Carcinoma, particularly
Morphologic examination of biopsy specimens for evaluation adenocarcinoma, may form glandular structures and/or pro-
of suspected metastasis first requires preparation of hema- duce mucin. Sarcoma characteristically displays spindle-­
toxylin- and eosin-stained slides. This process comprises sev- shaped cytology. Melanoma demonstrates diverse
eral steps, beginning with placement of sampled tissue in a architectural patterns and cytologic morphologies, but may
fixative, generally 10% formalin (which contains 4% formal- produce pigment. Lymphoma usually appears discohesive,
dehyde, a cross-linking agent). Other preservatives include relative to these other tumor types. For metastatic carcinoma,
oxidizing agents, such as KMnO4 and OsO4. Formalin-fixed other properties may permit preliminary assessment of pri-
tissue is then embedded in paraffin. This phase entails removal mary site, because carcinoma arising in certain organs pos-
of water from tissues via sequential immersion in successive sesses various histologic hallmarks. Colorectal carcinoma,
ethyl alcohol solutions of increasing concentration (70% then for instance, typically shows elongated, “pencillate” nuclei
95% then 100%), effectively dehydrating tissues and replac- and abundant grungy necrosis; pancreatic ductal adenocarci-
ing water with alcohol. The specimen is then cleansed of noma frequently contains neoplastic cells with clear cyto-
alcohol by immersion in xylene (or a xylene substitute such plasm; et cetera. Familiarity with these largely reproducible
as toluene or benzene), followed by placement into paraffin. properties facilitates generation of a differential diagnosis,
Once embedded in paraffin, tissues are sectioned via microt- refinement of which then depends upon ancillary testing.
omy and applied to glass slides. Staining of paraffin-embed-
ded tissues essentially involves reversal of the embedding
process. Immersion of tissue affixed to glass slides in xylene 7.4 Immunohistochemistry
first eliminates paraffin, then ethanol solutions of decreasing
concentration (100% then 95% then 70%) remove xylene, IHC is the primary ancillary method used by pathologists to
and tissues are ultimately rehydrated in water. Staining is then characterize cancers beyond their tissue architectural and
performed. Prior to microscopy, however, water must again cytologic features. IHC enables the spatial staining and qual-
be removed, due to differences between the refractive indices itative microscopic evaluation of protein expression in cells
of water and glass. Accordingly, stained tissues are washed in and tissues while preserving tissue architecture and tumor
the same sequence of ethanol concentrations, then xylene. microenvironment. The spatial resolution of IHC is particu-
Finally, the stained glass slide is prepared for examination by larly critical when tumors contain extensive stromal and
application of a coverslip using mounting media. immune cell infiltration. This is exemplified by highly infil-
trative tumors, where rare single cells, often detectable only
with cell-specific immunostains, are hidden among a sea of
7.3.2 Microscopic Evaluation intratumoral lymphocytes. Expression of tissue- and cancer-­
specific proteins enables the diagnostic classification of
Histologic review entails a structured visual algorithm, with malignancies, identification of tissue of origin, evaluation of
successive phases occurring at sequentially higher micro- expression of therapeutic targets and their modifiers, and
scopic resolution. The process begins at low-power magnifi- prognostication. IHC staining patterns can also serve as
7 Pathologic Assessment of Systemic or Distant Metastasis 65

surrogates of molecular alterations that impact protein Table 7.1 Select immunohistochemical markers
expression or stability, such as in TP53, ATRX, IDH1/2, and Cell lineage
MDM2, among many others. Epithelial Cytokeratins, EMA
Endothelial CD31, CD34, D2-40, ERG
Smooth Myocytic SMA, Caldesmon
Skeletal Myocytic Myogenin, MyoD1, Desmin
7.4.1 Immunohistochemistry Methodology
Neuroendocrine Synaptophysin, Chromogranin A,
CD56
Frozen or formalin-fixed, paraffin-embedded (FFPE) tissue Hematopoietic CD45, CD3, CD20, CD1a, CD68,
is sectioned using a cryostat or microtome, respectively, to MPO
sections measuring approximately 4–5 𝜇m in thickness and Melanocytic S100, SOX10, Melan-A, HMB-45,
MITF
subsequently mounted on glass slides. In the case of FFPE,
Malignancies
mounted tissue is the subject to physical, enzymatic, or Lung Adenocarcinoma CK7, TTF-1, NAPSIN
chemical antigen retrieval (AR) processes that disrupt pro- Gastric Adenocarcinoma CDX2, CK7, CK20
tein cross-links and expose epitopes recognized by target-­ Colon Adenocarcinoma CDX2, CK20, SATB2
specific primary antibodies. The primary antibody is Breast Invasive Ductal ER, PR, GATA-3, GCDFP-15,
subsequently applied at an optimized and clinically validated Carcinoma MAM
concentration to maximize target epitope binding while min- Prostate Adenocarcinoma PSA, NKX3.1, AR
Endometrial Adenocarcinoma CK7, CA-125, ER, PR
imizing nonspecific background binding. An enzyme-labeled
Clear Cell Renal Cell CA-IX, PAX8
secondary antibody raised against the immunoglobulin used Carcinoma
in the primary antibody is then used for signal detection and Squamous Cell Carcinoma p40/p63, CK5/6, HMWK
amplification. Horseradish peroxidase (HRP) and alkaline Small Cell Carcinoma Synaptophysin, Chromogranin A,
phosphatase (AP), the most commonly used enzymatic CD56
labels, produce a visible color (e.g., brown or red) when a
chromogen is applied. A counterstain, such as hematoxylin,
labels nuclei of all cells in the section. For quality control, ity. Examples include cytokeratins 7 and 20 (denoted as CK7
IHC interpretation includes positive control tissue that is and CK20), which accordingly represent useful targets dur-
known to express the target protein of interest and a negative ing evaluation of carcinoma arising from an unknown pri-
control (e.g., tissue minus the primary antibody) [9]. The mary location. The combinatorial result of stains for CK7
IHC-stained tissue is then interpreted for the presence and and CK20 facilitates distinction between possible origin
subcellular localization, or absence, of target expression in sites, as summarized by Wang et al. [11]. For instance, uro-
the cells of interest. The intensity of staining can serve as a thelial carcinoma generally stains positively for both mark-
surrogate for the degree of expression. In many cases, IHC-­ ers (CK7+ CK20+). Lung adenocarcinoma and mammary
based detection and interpretation of several protein markers, carcinoma usually show positive staining for CK7, whereas
such as estrogen receptor (ER𝛼), human epidermal growth staining for CK20 is negative (CK7+ CK20−); the opposite
factor receptor 2 (HER2), Ki-67, and phosphoproteins, can result (CK7− CK20+) is characteristic of colorectal carci-
be strongly influenced by tissue ischemia and fixation [10]. noma. Both stains are typically negative (CK7− CK20−) in
hepatocellular carcinoma, renal cell carcinoma, and prostatic
adenocarcinoma. In addition to intermediate filaments, other
7.4.2 Utility of Specific Markers antigens demonstrate tropism for certain lineages and, con-
sequently, value as immunohistochemical targets. Examples
The utility of immunohistochemistry in surgical pathology are listed in Table 7.1.
derives from the association between various antigens (pro-
teins) and certain tissues. These associations range from very
broad to highly specific. Broad associations include the rela- 7.5 Molecular Approaches
tionships between some cell populations and their character-
istic intermediate filaments, such as keratins (the predominant 7.5.1 Tissue-Based Methods
intermediate filament of epithelium), vimentin (an interme-
diate filament found in fibroblasts and other mesenchymal Detection of molecular alterations within a cancer specimen
cells), and desmin (in muscle). The abundance of these pro- provides insights beyond the information garnered from
teins within their respective cell types makes them valuable morphologic and immunohistochemical evaluation. In some
immunohistochemical markers of the corresponding tissues. tumor types, results of molecular analysis may have diagnos-
Among cytokeratins, some members of this intermediate fil- tic, prognostic, and/or predictive implications. Notably,
ament class demonstrate greater organ and cellular specific- PCR-based methods are not feasible following tissue
66 I. Katsyv and A. Turk

decalcification in acid. For biopsies of suspected bone metas- logic features. For instance, detection of an activating muta-
tasis upon which molecular testing may be indicated, alter- tion affecting exons 18–21 of EGFR within malignant but
native decalcifying agents such as EDTA are therefore otherwise unclassifiable cells suggests a diagnosis of meta-
recommended for this step. static lung adenocarcinoma, in the proper clinical context.
In terms of diagnostic utility, identification of oncogenic Whereas many cancers show recurrent alterations of onco-
mutations may facilitate more precise characterization of genes such as KRAS and PIK3CA, and/or tumor suppressors
some cases that do not permit definitive diagnosis by mor- like TP53, EGFR mutations are more characteristic of lung
phology alone. This prospect is especially pertinent in cases adenocarcinoma than many other tumor types. A third poten-
containing scant cellular material, and/or specimens obtained tial diagnostic application of molecular pathology involves
by invasive procedures after which repeat sampling would be the scenario of suspected metastasis in a patient whose pri-
inopportune. Three instances illustrate the potential diagnos- mary tumor contains a known mutation. The case depicted in
tic value of molecular pathology. First, the presence of Fig. 7.1 demonstrates this situation. A patient with history of
unequivocally oncogenic mutations may implicate malig- melanoma harboring a mutation of NRAS (c.37G>C, p.
nancy in a biopsy specimen that otherwise does not satisfy G13R) sustained a humeral fracture, suspected to represent a
qualitative morphologic criteria for designation as neoplas- pathologic fracture. Due to limited and suboptimal tissue
tic. Second, in the setting of metastatic cancer arising from sampling, however, histologic review did not yield an
an unknown source, detection of certain mutations may unequivocal diagnosis of melanoma. Subsequent molecular
implicate a specific diagnosis, despite ambiguous morpho- testing (which was feasible because the specimen had been

Fig. 7.1 Biopsy (upper panel) and molecular testing (lower panel) of a tered atypical cellular forms. Next-generation sequencing showed an
suspected humeral melanoma metastasis. Histologic review did not NRAS mutation (c.37G>C, p.G13R), identical to that within the ante-
yield an unequivocal diagnosis of melanoma, despite presence of scat- cedent resection specimen, facilitating definitive diagnosis
7 Pathologic Assessment of Systemic or Distant Metastasis 67

decalcified in EDTA) exhibited mutation of NRAS identical transcript-­level alterations in cancer driver genes. A signifi-
to that within the antecedent resection specimen. Interpreted cant advantage of NGS over more traditional methods is that
together with the morphologic findings and clinical history, it does not require a priori knowledge of genomic alterations.
molecular analysis of this case permitted definitive diagno- Furthermore, the high-throughput nature of NGS enables
sis, which would not have been possible if based solely on comprehensive, whole-genome assessment and identifica-
morphology. tion of novel and/or rare genomic variants, although often
Beyond these diagnostic considerations, molecular such variants are of limited or unknown clinical significance.
pathology also offers prognostic and predictive information Tissue foci of interest, typically areas of viable tumor cells,
in many tumor types. Examples of prognostically significant are identified on tissue mounted on glass slides and micro-
findings include KRAS mutations in lung adenocarcinoma, as dissected. DNA or RNA is then extracted from this microdis-
these alterations portend shorter survival in early-stage and sected tissue, and cDNA (complementary DNA) libraries are
locally advanced disease (reviewed by Jänne et al. [12]). generated for sequencing. Identified genomic variants are
Predictive contributions relate to mutations associated with subsequently filtered for those with known or predicted
targeted therapies, as well as alterations that confer resis- impact on tumor suppressor or oncogene function, or those
tance to certain agents. The former category currently predictive of therapeutic response, such as to receptor tyro-
includes targetable EGFR, BRAF, and MET mutations (sin- sine kinase inhibitors [14].
gle nucleotide variants and in-frame insertions/deletions), as Mutations in common cancer drivers such as TP53,
well as rearrangements involving ALK, ROS1, and RET in BRAF, PIK3CA, and APC are shared by many cancers [15].
lung adenocarcinoma; BRAF mutations in melanoma; Mutational profiling alone may therefore not offer the neces-
FGFR2 and FGFR3 fusions, and FGFR3 mutations in uro- sary resolution to fully characterize a tumor, particularly in
thelial carcinoma; and many other examples. Other altera- cases in which the primary site is not known. Here, analysis
tions may predict drug insensitivity and/or resistance, such of the tumor DNA methylome, though not yet widely used,
as KRAS and BRAF mutations in lung adenocarcinoma, and may be an attractive alternative. DNA methylation occurs
KRAS/NRAS/BRAF/PIK3CA mutations in colorectal carci- early in carcinogenesis, when CpG dinucleotides are methyl-
noma, among many other examples. More information is ated by DNA methyltransferases, modulating gene transcrip-
available via the FDA website (https://www.cancer.gov/ tion [16, 17]. DNA methylation is also tissue- and
about-­cancer/treatment/drugs). Overall, the diverse applica- disease-specific, and can serve as a molecular fingerprint of
tions of molecular methods in terms of diagnostic, p­ rognostic, cell state [18, 19]. These modifications can be detected by
and predictive impact demonstrate the tremendous utility of bisulfite treatment, in which unmethylated cytosines are
these ancillary tools. converted to uracils while methylated cytosines remain
­
As with protein expression, IHC is typically the first such unchanged, followed by NGS.
ancillary tool used to investigate a cancer’s molecular pro-
file. Mutation- or fusion-specific antibodies can be used to
detect the expression and/or aberrant localization of mutant 7.5.2 Cell-Free DNA
proteins. In situ hybridization (ISH) uses labeled DNA
probes complementary to RNA or DNA sequences to detect 7.5.2.1 Introduction
gene amplifications, deletions, fusions, and translocations. Although tissue remains the gold standard for cancer diag-
PCR uses a DNA polymerase and target-specific DNA prim- nosis, cell-free DNA (cfDNA), specifically circulating
ers to amplify a DNA region of interest. When monitored in tumor DNA (ctDNA), in body fluids is emerging as a reli-
real time, such as in quantitative real-time PCR (qPCR), the able and minimally invasive method of tumor diagnosis and
abundance or expression of the DNA or mRNA of interest monitoring. Fragments of double-stranded, nucleosome-
can be quantified. PCR primers spanning specific exon junc- associated DNA are released by secretion or following cell
tions can be used to detect pathogenic transcript splice vari- death (Fig. 7.2). This process occurs as part of normal cell
ants. Similarly, PCR primers spanning known fusion genes, turnover, as well as in pathological conditions such as
such as FGFR3-TACC3 in bladder cancers, can be used to trauma, infections, and malignancy [15, 20]. cfDNA profil-
amplify and detect tumor-specific gene fusions [13]. ing is currently used in prenatal care, for example, to detect
Next-generation sequencing (NGS) enables high-­ fetal trisomies in maternal plasma [17]. In the case of
throughput, ultrasensitive, massively parallel, and precise malignancies, these DNA fragments can be analyzed to
detection of genetic alterations. As it provides single nucleo- monitor a patient’s tumor burden and response to therapy,
tide resolution, NGS has proven indispensable for detecting detect molecular alterations that may be directly therapeu-
single-nucleotide variations, insertions and deletions, tically targetable, and monitor evolution of distinct tumor
fusions, translocations, copy number alterations, and clones [15].
68 I. Katsyv and A. Turk

Apoptosis
Apoptotic bodies

Point mutations

Copy number
Necrosis alterations

Rearrangements

CH3

Methylation
Secretion changes
CH3

Exosomal DNA

Fig. 7.2 Cellular sources of cfDNA and detectable alterations. Figure reproduced with permission from Wan et al. [15]

7.5.2.2 Isolation and Detection tage of ctDNA over standard methods of tumor monitoring is
cfDNA has been isolated from a variety of body fluids, its relatively short half-life on the order of minutes to hours.
including blood plasma, cerebrospinal fluid (CSF), saliva, In contrast, serological tumor makers such as cancer antigen
urine, stool, pleural and ascitic fluid, luminal gastrointestinal 15–3 (CA15–3/MUC1), cancer antigen 125 (CA-125/
tract brushings, cervical smears, and uterine lavage [15]. MUC16), and carcinoembryonic antigen (CEA) have half-­
Analyte fluid is collected in tubes containing a PCR-­ lives of weeks to days. ctDNA analysis may thus allow for
compatible anticoagulant, such as EDTA, and centrifuged to significantly higher temporal resolution, sensitivity, and
remove cells prior to cfDNA extraction. The isolated cfDNA specificity in monitoring disease status and therapeutic
is then profiled using polymerase chain-reaction (PCR)- response [15].
based and/or next-generation sequencing (NGS)-based Furthermore, cfDNA profiling may enable detection of
methods to identify tumor-specific DNA alterations such as minimal residual disease and risk stratification of patients
point mutations, structural rearrangements, and altered DNA following surgical resection [15]. In 55 patients who under-
methylation [15]. went neoadjuvant chemotherapy and surgical resection for
early-stage breast cancer, detection of ctDNA at postsurgical
7.5.2.3 Clinical Utility of cfDNA follow-up was associated with increased risk of metastatic
In healthy individuals, most cfDNA in blood comes primar- relapse [hazard ratio 25.1 (95% CI 4.08–130.5, logrank
ily from normal hematopoietic cells; each milliliter of plasma p < 0.0001)] [25]. Among 230 patients with early-stage colon
contains thousands of genome equivalents of cfDNA [20– cancer, detection of ctDNA following surgical resection was
22]. Patients with malignancies have altered profiles and similarly associated with increased risk of recurrence [haz-
increased concentration of plasma cfDNA, and those with ard ratio 18 (95% CI 7.9–40, logrank p < 0.001)] [26]. In
metastatic disease have as much as 100-fold higher levels of multiple studies, relapse detected by ctDNA preceded clini-
ctDNA compared to patients with early-stage tumors [15, 20, cal relapse by as much as 11 months [15]. Elevated ctDNA
21, 23, 24]. ctDNA levels may initially increase in patients may even precede radiographic progression, and thus may
whose tumors respond to therapy, but subsequently fall, serve as an early marker of treatment response [20, 24].
whereas ctDNA levels remained unchanged or increase in In a given patient, different tumors, different regions
those with poor therapeutic response [24]. An added advan- within a tumor, or even different cells in a single region of a
7 Pathologic Assessment of Systemic or Distant Metastasis 69

tumor may show significant molecular heterogeneity corre- 7.5.2.5 DNA Methylation
sponding to clones with divergent biological behavior and cfDNA is chemically stable, and cancer-specific DNA
treatment responsiveness. Compared to biopsy of a single methylation patterns can be detected in ctDNA with
mass, ctDNA may provide a more comprehensive and global improved specificity compared to genetic methods [18, 19].
portrait of this intra- and inter-tumor heterogeneity and poten- DNA methylation profiling distinguished benign pelvic
tially enable earlier identification of tumor clones with resis- masses from high-grade serous ovarian cancers with a
tance to particular treatments. Longitudinal monitoring of specificity of greater than 90%, and identified with a speci-
ctDNA can help detect the emergence of therapy-resistant ficity of 88% women who within 2 years developed ovarian
clones and obviate the need for repeated invasive tissue biop- cancer [17]. ctDNA methylation markers specific to hepa-
sies. For example, serial monitoring of colorectal cancer tocellular carcinoma (HCC) were correlated with tumor
patients undergoing treatment with the anti-EGFR antibody burden, stage, and treatment response [30]. Methylation
cetixumab showed emergence and positive selection of profiling of ctDNA has also shown promise in other malig-
mutant KRAS clones, with subsequent decay following anti- nancies, such as papillary thyroid carcinoma and lung can-
body discontinuation [27]. Similarly, ctDNA monitoring of cers [31, 32].
lung cancer patients receiving the EGFR inhibitor erlotinib
enabled detection of a resistance conferring T790M mutation 7.5.2.6 Concordance with Tissue-Based Methods
clone nearly a full year before clinical progression of disease A potential current limitation of cfDNA in tumor monitoring
[28]. The U.S. Food and Drug Administration (FDA) and the is the high variability in ctDNA concentration among
European Medicines Agency (EMA) have approved the use patients, even after matching for malignancy type and stage.
of PCR-based ctDNA assays for EGFR mutations prior to ini- Much of this variability is likely due to differences in disease
tiating therapy with anti-EGFR agents, and additional ctDNA- burden between patients. In patients with low tumor burden
based assays are being evaluated in clinical trials [15, 20, 29]. or with increased release of normal DNA from blood cells,
tumor-associated alterations in overall cfDNA can be below
7.5.2.4 cfDNA Fragmentation Profile the limit of detection and result in low concordance between
Cancer and non-cancer cells show significant differences in molecular alterations identified in ctDNA and those identi-
chromatin structure, some of which underlie transcriptional fied in tissue biopsies from the same patients. Studies with
programs necessary for carcinogenesis. These differences in matched tissue and ctDNA samples show a sensitivity of
chromatin structure may be detected as altered cfDNA frag- ctDNA analysis of greater than 60% and up to 98% for
mentation profiles in patients with cancer compared to detecting tumor-associated mutations [15]. Use of ultrasensi-
healthy controls. CfDNA profiles reflect both structural tive detection methods, such as droplet digital PCR; beads,
alterations, such as DNA copy number changes, as well as emulsion, amplification, magnetics (BEAM) PCR; and tar-
epigenetic modifications that regulate chromatin and nucleo- geted and whole-genome NGS can substantially improve
some structure. Differences in sequencing coverage of spe- sensitivity of ctDNA-based methods and improve their con-
cific genomic positions can thus distinguish ctDNA from cordance with tissue-based assays [15, 20].
cfDNA derived from normal cells and outperforms DNA
copy number and mitochondrial DNA-based methods in dis-
tinguishing cfDNA from healthy or cancer patients [21]. In a Conclusion and Future Perspective
cohort of 126 patients with carcinomas of the breast, bile Pathologists make use of multiple modalities when
duct, colon, stomach, lung, or ovary, a classifier based on evaluating a metastatic tumor. When the primary lesion
cfDNA fragmentation identified 82% of patients with can- is known, histologic, immunohistochemical, and
cer; when combined with mutational analysis, that number molecular profiling of metastatic lesions can identify
rate increased to 91%, and to as high as 100% for patients prognostic and treatment-altering genomic, epig-
with metastatic disease. In patients undergoing treatment enomic, and phenotypic changes. When a primary
with EGFR or ERBB2 inhibitors, changes in cfDNA frag- lesion is not known, this workup can often identify a
mentation profiles paralleled changes in EGFR or ERBB2 primary site. Although these assays are currently done
mutant allele frequencies, suggesting ctDNA fragmentation on tissue specimens, advances in ctDNA isolation and
may serve as a potential surrogate marker of treatment analysis may enable minimally invasive and global
response [21]. Furthermore, differences in ctDNA fragmen- profiling of a patient’s systemic disease burden. As
tation and the underlying differences in nucleosome occu- pathological diagnosis of cancer evolves, the following
pancy among different cancers may aid in accurate diagnosis questions will drive significant changes in the field:
in cancers of unknown primary [15].
70 I. Katsyv and A. Turk

11. Wang N, Zee S, Zarbo R, Bacchi C, Gown A. Coordinate expres-

• To what extent will DNA methylation studies sion of Cytokeratins 7 and 20 defines unique subsets of carcinomas.
Appl Immunohistochem. 1995;3(2):99–107.
improve diagnostic resolution and classification of 12. Jänne PA, Engelman JA, Johnson BE. Epidermal growth factor
cancers? receptor mutations in non–small-cell lung cancer: implications for
• Will more cfDNA-based methods supplant with tis- treatment and tumor biology. J Clin Oncol. 2005;23(14)
sue and serological tumor markers? 13. El-Deiry WS, Goldberg RM, Lenz H, Shields AF, Gibney GT,
Tan AR, Brown J, Eisenberg B, Heath EI, Phuphanich S, Kim E,
• Will integration of molecular findings from Brenner AJ, Marshall JL. The current state of molecular testing in
population-­based gene interaction networks, such the treatment of patients with solid tumors, 2019. Cancer J Clin.
as those based on the ARACNE algorithm [33], fur- 2019:caac.21560.
ther improve disease classification based on tran- 14. Goodwin S, McPherson JD, McCombie WR. Coming of age: ten
years of next-generation sequencing technologies. Nat Rev Genet.
scriptional dependencies in cancers? 2016;17(6)
• Relatedly, will artificial intelligence-based tools 15. Wan JC, Massie C, Garcia-Corbacho J, Mouliere F, Brenton JD,
augment the pathologist’s ability to integrate clini- Caldas C, Pacey S, Baird R, Rosenfeld N. Liquid biopsies come of
cal, histologic, and molecular data with improved age: towards implementation of circulating tumour DNA. Nat Rev
Cancer. 2017;17(4):223–38.
diagnostic accuracy? 16. Locke WJ, Guanzon D, Ma C, Liew YJ, Duesing KR, Fung KY,
• Will in vivo imaging methods advance sufficiently Ross JP. DNA methylation cancer biomarkers: translation to the
to provide histologic and cytologic resolution? clinic. Front Genet. 2019;10
17. Widschwendter M, Zikan M, Wahl B, Lempiäinen H, Paprotka T,
Evans I, Jones A, Ghazali S, Reisel D, Eichner J, Rujan T, Yang
Z, Teschendorff AE, Ryan A, Cibula D, Menon U, Wittenberger
T. The potential of circulating tumor DNA methylation analysis
for the early detection and management of ovarian cancer. Genome
References Med. 2017;9(1):116.
18. Fernandez AF, Assenov Y, Martin-Subero JI, Balint B, Siebert R,
1. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, Taniguchi H, Yamamoto H, Hidalgo M, Tan AC, Galm O, Ferrer I,
2021. CA Cancer J Clin. 2021;71(1) https://doi.org/10.3322/caac. Sanchez-Cespedes M, Villanueva A, Carmona J, Sanchez-Mut JV,
21654. Berdasco M, Moreno V, Capella G, Monk D, Ballestar E, Ropero S,
2. Riihimäki M, Thomsen H, Hemminki A, Sundquist K, Hemminki Martinez R, Sanchez-Carbayo M, Prosper F, Agirre X, Fraga MF,
K. Comparison of survival of patients with metastases from known Grana O, Perez-Jurado L, Mora J, Puig S, Prat J, Badimon L, Puca
versus unknown primaries: survival in metastatic cancer. BMC AA, Meltzer SJ, Lengauer T, Bridgewater J, Bock C, Esteller M. A
Cancer. 2013;13(1) https://doi.org/10.1186/1471-­2407-­13-­36. DNA methylation fingerprint of 1628 human samples. Genome
3. LiVolsi V, Nallib A, Baloch Z, Bartolazzi A, Chan J, DeLellis R, Res. 2012;22(2):407–19.
El-Naggar A, Eloy C, Eng C, Fagin J, Ghossein R, Giordano T, 19. Lokk K, Modhukur V, Rajashekar B, Märtens K, Mägi R, Kolde
Kondo T, Lloyd R, Mete O, Nikiforov Y, Nonaka D, Paschke R, R, Koltšina M, Nilsson TK, Vilo J, Salumets A, Tõnisson N. DNA
Perren A, Rosai J, Sadow P, Schneider A, Sobrinho S, Tallini G, methylome profiling of human tissues identifies global and tissue-­
Williams M (2017b) Follicular thyroid carcinoma. In: Lloyd RV, specific methylation patterns. Genome Biol. 2014;15(4):3248.
Osamura R, Kloppel G, Rosai J, editors. WHO classification 20. Corcoran RB, Chabner BA. Application of cell-free DNA analysis
of tumours of endocrine organs, vol. 10, 4th ed, World Health to cancer treatment. N Engl J Med. 2018;379(18):1754–65.
Organization (IARC), Lyon. 21. Cristiano S, Leal A, Phallen J, Fiksel J, Adleff V, Bruhm DC,
4. LiVolsi V, Baloch Z, Sobrinho S, Tallini G (2017a) Hurthle (onco- Jensen S, Medina JE, Hruban C, White JR, Palsgrove DN, Niknafs
cytic) cell tumours. In: Lloyd R, Osamura R, Kloppel G, Rosai J, N, Anagnostou V, Forde P, Naidoo J, Marrone K, Brahmer J,
editors. WHO classification of tumours of endocrine organs, vol 10, Woodward BD, Husain H, van Rooijen KL, Ørntoft MBW, Madsen
4th ed, World Health Organization (IARC), Lyon. AH, van de Velde CJ, Verheij M, Cats A, Punt CJ, Vink GR, van
5. Budczies J, von Winterfeld M, Klauschen F, Bockmayr M, Lennerz Grieken NC, Koopman M, Fijneman RJ, Johansen JS, Nielsen HJ,
JK, Denkert C, Wolf T, Warth A, Dietel M, Anagnostopoulos I, Meijer GA, Andersen CL, Scharpf RB, Velculescu VE. Genome-­
Weichert W, Wittschieber D, Stenzinger A. The landscape of meta- wide cell-free DNA fragmentation in patients with cancer. Nature.
static progression patterns across major human cancers. Oncotarget. 2019;570(7761):385–9.
2015;6(1) https://doi.org/10.18632/oncotarget.2677. 22. Phallen J, Sausen M, Adleff V, Leal A, Hruban C, White J,
6. Riihimäki M, Thomsen H, Sundquist K, Sundquist J, Hemminki Anagnostou V, Fiksel J, Cristiano S, Papp E, Speir S, Reinert T,
K. Clinical landscape of cancer metastases. Cancer Med. 2018;7(11) Orntoft MBW, Woodward BD, Murphy D, Parpart-Li S, Riley D,
https://doi.org/10.1002/cam4.1697. Nesselbush M, Sengamalay N, Georgiadis A, Li QK, Madsen MR,
7. Lorigan JG, Wallace S, Mavligit GM. The prevalence and loca- Mortensen FV, Huiskens J, Punt C, van Grieken N, Fijneman R,
tion of metastases from ocular melanoma: imaging study in 110 Meijer G, Husain H, Scharpf RB, Diaz LA, Jones S, Angiuoli S,
patients. Am J Roentgenol. 1991;157(6):1279–81. Ørntoft T, Nielsen HJ, Andersen CL, Velculescu VE. Direct detec-
8. Rajpal S, Moore R, Karakousis CP. Survival in metastatic ocular tion of early-stage cancers using circulating tumor DNA. Sci Transl
melanoma. Cancer. 1983;52(2) Med. 2017;9(403):eaan2415.
9. Magaki S, Hojat SA, Wei B, So A, Yong WH. An introduction to 23. Bettegowda C, Sausen M, Leary RJ, Kinde I, Wang Y, Agrawal
the performance of immunohistochemistry. Methods Mol Biol. N, Bartlett BR, Wang H, Luber B, Alani RM, Antonarakis ES,
2019;1897:289–98. Azad NS, Bardelli A, Brem H, Cameron JL, Lee CC, Fecher LA,
10. Kim SW, Roh J, Park CS. Immunohistochemistry for pathologists: Gallia GL, Gibbs P, Le D, Giuntoli RL, Goggins M, Hogarty MD,
protocols, pitfalls, and tips. J Pathol Transl Med. 2016;50(6):411–8. Holdhoff M, Hong SM, Jiao Y, Juhl HH, Kim JJ, Siravegna G,
7 Pathologic Assessment of Systemic or Distant Metastasis 71

Laheru DA, Lauricella C, Lim M, Lipson EJ, Marie SKN, Netto 29. Dawson SJ, Tsui DW, Murtaza M, Biggs H, Rueda OM, Chin
GJ, Oliner KS, Olivi A, Olsson L, Riggins GJ, Sartore-Bianchi A, SF, Dunning MJ, Gale D, Forshew T, Mahler-Araujo B, Rajan S,
Schmidt K, Shih LM, Oba-Shinjo SM, Siena S, Theodorescu D, Tie Humphray S, Becq J, Halsall D, Wallis M, Bentley D, Caldas C,
J, Harkins TT, Veronese S, Wang TL, Weingart JD, Wolfgang CL, Rosenfeld N. Analysis of circulating tumor DNA to monitor meta-
Wood LD, Xing D, Hruban RH, Wu J, Allen PJ, Schmidt CM, Choti static breast cancer. N Engl J Med. 2013;368(13):1199–209.
MA, Velculescu VE, Kinzler KW, Vogelstein B, Papadopoulos N, 30. Xu R, Wei W, Krawczyk M, Wang W, Luo H, Flagg K, Yi S, Shi W,
Diaz LA. Detection of circulating tumor DNA in early- and late Quan Q, Li K, Zheng L, Zhang H, Caughey BA, Zhao Q, Hou J,
stage human malignancies. Sci Transl Med. 2014;6(224):224ra24. Zhang R, Xu Y, Cai H, Li G, Hou R, Zhong Z, Lin D, Fu X, Zhu J,
24. Leon SA, Shapiro B, Sklaroff DM, Yaros MJ. Free DNA in the Duan Y, Yu M, Ying B, Zhang W, Wang J, Zhang E, Zhang C, Li O,
serum of cancer patients and the effect of therapy. Cancer Res. Guo R, Carter H, Zhu J, Hao X, Zhang K. Circulating tumour DNA
1977;37(3):646–50. methylation markers for diagnosis and prognosis of hepatocellular
25. Garcia-Murillas I, Schiavon G, Weigelt B, Ng C, Hrebien S, Cutts carcinoma. Nat Mater. 2017;16(11):1155–61.
RJ, Cheang M, Osin P, Nerurkar A, Kozarewa I, Garrido JA, 31. Khatami F, Teimoori-Toolabi L, Heshmat R, Nasiri S, Saffar
Dowsett M, Reis-Filho JS, Smith IE, Turner NC. Mutation tracking H, Mohammadamoli M, Aghdam MH, Larijani B, Tavangar
in circulating tumor DNA predicts relapse in early breast cancer. SM. Circulating ctDNA methylation quantification of two DNA
Sci Transl Med. 2015;7(302):133–302. methyl transferases in papillary thyroid carcinoma. J Cell Biochem.
26. Tie J, Wang Y, Tomasetti C, Li L, Springer S, Kinde I, Silliman N, 2019;120(10):17,422–37.
Tacey M, Wong HL, Christie M, Kosmider S, Skinner I, Wong R, 32. Liang W, Zhao Y, Huang W, Gao Y, Xu W, Tao J, Yang M, Li L, Ping
Steel M, Tran B, Desai J, Jones I, Haydon A, Hayes T, Price TJ, W, Shen H, Fu X, Chen Z, Laird PW, Cai X, Fan JB, He J. Non-­
Strausberg RL, Diaz LA, Papadopoulos N, Kinzler KW, Vogelstein invasive diagnosis of earlystage lung cancer using high-throughput
B, Gibbs P. Circulating tumor DNA analysis detects minimal resid- targeted DNA methylation sequencing of circulating tumor DNA
ual disease and predicts recurrence in patients with stage II colon (ctDNA). Theranostics. 2019;9(7):2056–70.
cancer. Sci Transl Med. 2016;8(346):92–346. 33. Alvarez MJ, Subramaniam PS, Tang LH, Grunn A, Aburi M,
27. Siravegna G, Mussolin B, Buscarino M, Corti G, Cassingena A, Rieckhof G, Komissarova EV, Hagan EA, Bodei L, Clemons
Crisafulli G, Ponzetti A, Cremolini C, Amatu A, Lauricella C, PA, Dela Cruz FS, Dhall D, Diolaiti D, Fraker DA, Ghavami A,
Lamba S, Hobor S, Avallone A, Valtorta E, Rospo G, Medico Kaemmerer D, Karan C, Kidd M, Kim KM, Kim HC, Kunju LP,
E, Motta V, Antoniotti C, Tatangelo F, Bellosillo B, Veronese S, Langel Li Z, Lee J, Li H, LiVolsi V, Pfragner R, Rainey AR, Realubit
Budillon A, Montagut C, Racca P, Marsoni S, Falcone A, Corcoran RB, Remotti H, Regberg J, Roses R, Rustgi A, Sepulveda AR, Serra
RB, Di Nicolantonio F, Loupakis F, Siena S, Sartore-Bianchi A, S, Shi C, Yuan X, Barberis M, Bergamaschi R, Chinnaiyan AM,
Bardelli A. Clonal evolution and resistance to EGFR blockade in the Detre T, Ezzat S, Frilling A, Hommann M, Jaeger D, Kim MK,
blood of colorectal cancer patients. Nat Med. 2015;21(7):795–801. Knudsen BS, Kung AL, Leahy E, Metz DC, Milsom JW, Park YS,
28. Sorensen BS, Wu L, Wei W, Tsai J, Weber B, Nexo E, Meldgaard Reidy-Lagunes D, Schreiber S, Washington K, Wiedenmann B,
P. Monitoring of epidermal growth factor receptor tyrosine kinase Modlin I, Califano A. A precision oncology approach to the phar-
inhibitor-sensitizing and resistance mutations in the plasma DNA macological targeting of mechanistic dependencies in neuroendo-
of patients with advanced non–small cell lung cancer during treat- crine tumors. Nat Genet. 2018;50(7):979–89.
ment with erlotinib. Cancer. 2014;120(24):3896–901.
 The Role of Angiotropic Extravascular
Migratory Metastasis in Metastases 8
Raymond Barnhill and Claire Lugassy

Abstract Learning Objectives

• The phenomenon of angiotropism, pericyte mim-
The phenomenon of angiotropism, pericyte mimicry
icry (PM), and extravascular migratory metastasis
(PM), and extravascular migratory metastasis (EVMM)
(EVMM) has questioned the concept that mela-
has questioned the concept that melanoma cells metasta-
noma cells metastasize exclusively within vascular
size exclusively within vascular channels. During this
channels.
phenomenon of cancer spread and metastasis, melanoma
• The long interval to recurrence after resection and
cells attach to the abluminal vascular surfaces i.e. angiot-
other therapies and the development of resistant
ropism and spread without intravasation via continuous
metastases have proved very problematic.
migration, replacing pericytes, i.e. pericyte mimicry
Dormancy remains poorly understood and, as a
(PM). Angiotropism, PM and EVMM have mainly been
result, effective therapeutic interventions have not
studied in melanoma but also occur in other cancer types.
been formulated.
PM and EVMM appear to be a reversion to an
• During EVMM, melanoma cells attach to the ablu-
embryogenesis-­derived program. There are many analo-
minal vascular surfaces, i.e., angiotropism and
gies between the neural crest development and cancer
spread without intravasation via continuous migra-
progression, including the important role of laminins,
tion, replacing pericytes, i.e., pericyte mimicry
epithelial-mesenchymal transition, and the re-activation
(PM) toward regional and distant secondary sites.
of embryonic signals by cancer cells. Furthermore, there
• There are many analogies between the neural crest
is no circulation of blood during the first trimester of
development and cancer progression, including the
embryogenesis, despite the fact that there is extensive
important role of laminins, epithelial-mesenchymal
migration of cells to distant sites and formation of organs
transition, and the re-activation of embryonic sig-
and tissues during this period. Embryonic migration
nals by cancer cells.
therefore is a continuous extravascular migration as are
• Importantly, there is no circulation of blood during
PM and EVMM, supporting the concept that these embry-
the first trimester of embryogenesis, despite the fact
onic migratory events appear to recur abnormally during
that there is extensive migration of cells to distant
the metastatic process. This new paradigm may greatly
sites and the formation of organs and tissues during
influence the development of new effective therapeutics
this period. Embryonic migration therefore is a con-
for metastasis.
tinuous extravascular migration as are PM and
EVMM.
• Via PM/EVMM, angiotropic cancer cells will fol-
low indefinitely, in a dysregulated way, “embryonic
seed and soil” mechanisms in the invaded organ.
R. Barnhill
• A more comprehensive understanding of the
Department of Translational Research, Institut Curie, Paris, France mechanism(s) of tumor spread is greatly needed,
University of Paris Réné Descartes Faculty of Medicine,
and angiotropism and PM/EVMM may profoundly
Paris, France influence and stimulate the development of more
C. Lugassy (*)
effective therapeutics.
Department of Translational Research, Institut Curie, Paris, France
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 73

S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_8
74 R. Barnhill and C. Lugassy

8.1 Introduction noma cells in contact with vessels” [4, 5]. Electron micros-
copy was necessary to confirm this direct contact of tumor
Intravascular dissemination of cancer is still considered the cells with the basal lamina of endothelial cells. This ultra-
principal if not exclusive route of metastasis. For many years, structural relationship was termed “the angio-tumoral com-
the relationships between a tumor and its vasculature have plex,” which in turn corresponds to angiotropism of tumor
been thought to principally involve (i) angiogenesis required cells at the light microscopic level.
for tumor growth at both primary and secondary sites and (ii)
intravascular circulation of tumor cells to metastatic sites.
However, the precise mechanisms of tumor growth and 8.2.1 Angiotropism
tumor cell migration and in particular the role of the tumor
vasculature in these processes remain unclear [1]. Thus, specifically in the cancer literature, angiotropism
Although it had been assumed that tumor cells in contact refers to tumor cells exclusively localized to the external sur-
with vessel abluminal surfaces were in the process of intrav- faces of the endothelial cells of blood vessels in a pericytic
asation, it was shown in the late 1990s that such vessel– location in the absence of intravasation (Figs. 8.1 and 8.2)
tumor interactions could have a completely different [6]. The term angiotropism has also been employed for an
biological interpretation. In particular, angiotropism, peri- uncommon lymphoma—“angiotropic lymphoma,” which
cyte mimicry (PM), and extravascular migratory metastasis presents almost exclusively with intravascular rather extra-
(EVMM). This new field has raised important questions vascular tumor. This lymphoma is not recognized as a dis-
about cancer biology. tinct entity in the Revised European-American Classification
Research on angiotropism, pericyte mimicry, and extra- of Lymphoid Neoplasms.
vascular migratory.
metastasis or EVMM, opened by Lugassy and Barnhill
and reviewed in [2], has questioned the assumption that all 8.2.2 Pericytic Mimicry
tumor cells metastasize exclusively within vascular lumina.
Angiotropism, PM, and EVMM are three complementary Since 1999, Lugassy and Barnhill have shown that angiotro-
terms that refer to an embryologically derived mechanism of pism is a prognostic factor and a microscopic marker of
continuous extravascular tumor cell migration without tumor cell migration along the abluminal vascular surface in
entrance into vascular channels. In addition to migration a pericytic location [2], described as “pericytic-like” migra-
along vessels as angiotropism and PM, EVMM can also tion. In 2013, they have replaced “pericytic-like migration”
occur along other anatomical tracks such as nerves in neurot- with the term “pericytic mimicry” (PM) [7, 8]. PM describes
ropism [2]. well this biological process of tumor spread along the perien-
Angiotropic tumor cells localize along the abluminal vas- dothelial, abluminal vascular surface, competition with, and
cular surfaces by competition with and mimicking of peri- mimicking of pericytes [2]. This mode of tumor spread along
cytes (or other perivascular mesenchymal stem cells): (i) vessels of all types (including larger vessels, the sinusoids of
tumor cells are recruited along the abluminal vascular sur- the liver, etc.) seems to connote tumor cells replacing and
face instead of pericytes and replace pericytes; (ii) tumor mimicking not only pericytes but also other perivascular
cells migrate along the abluminal surface (Fig. 8.1). However, mesenchymal stem cells [6].
tumor cells never stabilize neovessels nor the vascular base- Angiotropism/PM has been shown to occurs along preex-
ment membrane matrix as do pericytes. PM allows tumor isting vessels; however, several data indicate that it could
cells to progress along vessels with all of the needed nutri- also occur along neovessels (see below).
ents and oxygen, and such a process avoids the destruction of
tumor cells observed in hematogenous dissemination [3].
8.2.3 Extravascular Migratory Metastasis

Extravascular migratory metastasis (EVMM) was originally

8.2  ngiotropism, Pericytic Mimicry (PM),
A coined by Lugassy and Barnhill to describe tumor cell
and Extravascular Migratory spread by way of pericytic mimicry from a primary tumor to
Metastasis (EVMM) : Terminology secondary metastatic sites [9]. This additional mechanism
and Definitions of metastasis specifically involves tumor cell migration
which is a slower, progressive, mode of tumor dissemina-
In the work initiated by Lugassy, direct contact between tion (Fig. 8.1) versus the ultra-rapid circulation of tumor
tumor cells and endothelium was observed via a basal lamina cells in lymphatic or blood vascular channels [3, 10].
rich in laminin “which may promote the migration of mela- Angiotropism, PM, and EVMM may be considered three
8 The Role of Angiotropic Extravascular Migratory Metastasis in Metastases 75

a
3
c

2
f

1 e

3 d

Tumor

Fig. 8.1 Diagram of PM and EVMM red vessels. (e) Double immunostaining of a corresponding human his-
(a) Overview diagram of human body. (b) Diagram of PM: tumor cells topathological image at the advancing front of a liver metastasis.
migrate in a pericytic location along the abluminal vascular surface at Melanoma cells express HMB45 along endothelial cell expressing
the advancing front of the tumor. (c) Corresponding human histopatho- CD31 (DAB brown chromogen). F. Human melanoma brain metastasis.
logical image of angiotropism/PM at the advancing front of the primary Striking angiotropism of pigmented melanoma cells in a pericytic loca-
melanoma. Double immunostaining of melanoma cells by melan-A tion along microvessels containing erythrocytes
(red chromogen) and vascular channels by CD31 (brown-black chro- Green arrowheads indicate tumor cells, red arrowheads indicated ves-
mogen). (d) Diagram of the replacement histopathological growth pat- sels. Bicolor arrows indicate a schematic view of PC. (1) indicates PM
tern in liver metastasis. Note at the advancing front of the metastasis the at the advancing front of the primary tumor. (2) indicates EVMM via
progression of green tumor cells replacing hepatocytes (orange) along PM throughout the body. (3) indicates metastases sites [2]

complementary terms describing this mechanism of contin-

uous tumor cell migration toward distant sites (Fig. 8.1).
8.3  he Biological and Prognostic
T
Tumors cells are nourished during their migration along
Significance of Angiotropism
vessel surfaces by obtaining their requisite nutrients and
in Melanoma
oxygen. Such tumor cell migration also affords the advan-
tage of avoiding destruction upon entry into the blood vas-
8.3.1 Biological Significance
cular circulation [11, 12].
Lugassy and Barnhill first described the true biological sig-
EVMM can involve other anatomic structures as tracks
nificance of angiotropism as the specific histopathological
for tumor cell migration, as, e.g., nerves, i.e., neurotropism
marker of PM and EVMM [4, 5]. Angiotropism is defined as
in many tumors and white matter tracks in the brain with
tumor cells aligned along the external surfaces of vascular
glioma invasion [13–15].
76 R. Barnhill and C. Lugassy

b c

Fig. 8.2 Angiotropism in primary cutaneous melanoma the advancing front. Note the angiotropic melanoma cells (black
(a) Low magnification of the tumor. Encircled area: angiotropism at the arrows) associated with vascular channels (red arrow) [81]
advancing front (the base) of the lesion. (b, c) Higher magnification of

channels in a pericytic location (PM) (Fig. 8.1a–c). Micro- site traverse and extend beyond the sclera along the ablumi-
and macrovascular channels may show comparable angiotro- nal surfaces of vascular channels (including the endothelial-­
pism. Larger vessels are characterized by mimicry of lined channels of Schlemm) [16]. No evidence of intravascular
adventitial mesenchymal stem cells. Angiotropism may con- or intra-lymphatic tumor cells are observed.
sist of both single tumor cells or “collective” aggregates of
tumor cells and is most frequently detected at the advancing 8.3.1.1 Melanoma liver Metastases
front of the tumor or in nearby tissue and cannot be reliably Liver metastases from colorectal and breast cancer [17] and
recognized within a tumoral mass per se [6]. Angiotropism melanoma, both uveal [18] and cutaneous [19], show distinc-
occurs in both primary and metastatic tumors (Fig. 8.1). tive “histopathological growth patterns” (HGPs) as described
Although studied thus far mainly in melanoma, angiotropism above in the co-option section: “replacement,” “desmoplas-
is present not only in other tumors of neural crest origin such tic,” “pushing,” and two rare variants—the “sinusoidal” and
as glioma but also in many other neoplasms (see below). “portal” HGPs. In particular, the replacement HGP is defined
Angiotropism in melanomas and in other neoplasms is by tumor cells progressively spreading along the abluminal
recognized at the advancing front of a tumor or in the nearby surfaces of sinusoidal vessels (in the space of Disse) or
tissue (usually at least 0.5 to 2 mm away from the tumor) venules of the portal tracts [17–19]. Since hepatic stellate
(Fig. 8.2), and in loco-regional microsatellite, satellite, and cells, which are considered to be pericytes, are located in this
in transit metastases in nearby skin [2, 6]. In uveal mela- space of Disse [20, 21], tumor cells appear to engage in peri-
noma, melanoma cells from the primary intraocular uveal cytic mimicry, i.e., competing with and replacing these peri-
8 The Role of Angiotropic Extravascular Migratory Metastasis in Metastases 77

cytes. The replacement HGP represents approximately 75% sis has discovered 20 genes related to (i) cell migration
of the liver metastases derived from uveal melanoma [18] (ICAM1, CCL2, PDGFB, RGNEF, RANBP9, IL6), (ii) can-
and 50% of metastases associated with cutaneous melanoma cer progression (ICAM1, SELE, CCL2, CXCL6, IL6,
[19] (Fig. 8.1d,e). SERPINB2, TRAF1, BLID, UPF1, PLAA, MALT1, ZXDC,
RGNEF), (iii) EMT (ICAM1, CCL2, PDGFB IL6,), (iv)
inflammation (ICAM1, SERPINB2, CCL2, CXCL6, SELE,
8.3.2 Prognostic Significance SLC7A2, C2CD4B, PDGFB, IL6, TRAF1), (v) cancer
stem/embryonic cell properties (CCL2, PDGFB, EVX1,
Angiotropism is an independent prognostic factor for the CFDP1 and RANBP9, (v) pericyte recruitment (PDGFB).
development of metastases and diminished survival in both In addition, pericyte-marker expression by C8161 mela-
cutaneous and uveal melanoma, and this includes micro- and noma cells included CD146, NG2, CD44, CD73, CD105,
macroscopic satellites, in transit metastases, lymph node and CD144 [6].
metastases, and distant metastases [2]. In addition, the replace-
ment HGP in uveal and cutaneous melanoma liver metastases,
which exhibits angiotropism, is predictive of poor survival 8.4.2 Non-melanoma Tumor Models
with both univariate and multivariate analysis [18, 19]. As
with liver metastases from other neoplasms, the replacement Glioma propagation in the brain has been studied with the
HGP in liver metastases from cutaneous and uveal melanoma in vitro co-culture model of PM (see angiotropism/PM and
is associated with a poorer prognosis than the desmoplastic co-option above; for molecular data, see the Co-option sec-
HGP [17–19]. In addition, angiotropism appears to be a factor tion above) [30]. In a similar fashion, the in vitro co-culture
in resistance to BRAF inhibitor therapy [22]. and an organotypic brain slice culture models have demon-
strated PM by metastatic cells from several different tumor
types [31]. Metastatic colonization in various organs via PM
8.4 Experimental Models and Molecular occurs after tumor cells are injected into mice. In the latter
Findings studies, L1CAM (cell adhesion molecule L1) increases β1
integrin-ILK signaling and in turn YAP nuclear localization
8.4.1 Melanoma Experimental Models [31]. It thus appears that L1CAM and other adhesion mole-
cules are utilized by tumor cells of various origins to facili-
In vitro, ex vivo, and in vivo models have been developed for tate attachment to the abluminal vascular surface during PM
the study of angiotropism and PM of melanoma cells [2]. [31]. A recent study employing a murine leukemia model
These include a 3D in vitro co-culture of tumor cells with suggests that leukemia cells migrate directly along microvas-
endothelial tubules [23]; an ex vivo model of co-culture of cular channels into the central nervous system (CNS) via PM
melanoma cells with neovessels in rat aortic rings [24, 25], [32] circumventing the need to traverse the blood-brain bar-
an ex vivo ear tissue invasion model [26]; a modified shell-­ rier. PM has also been implicated in prostate cancer progres-
less chick chorioallantoic membrane (CAM) model [27]; sion [33]: CD44, which is a marker of both pericytes and
and a genetically engineered murine melanoma model. In the cancer stem cells, appears to promote migration and invasion
latter, UV irradiation elicited neutrophil-rich inflammation of docetaxel-resistant prostate cancer cells via induction of
and increased angiotropic melanoma metastases in these Hippo-YAP signaling [34].
murine melanomas [25] (Fig. 8.3a–c). In another murine
brain melanoma model, fluorescence-labeled melanoma
cells showed progressive outgrowth into the surrounding 8.5 Laminin in Angiotropism and PM
brain along the abluminal surfaces of preexisting vessels
[28] (Fig. 8.3d–f). A new mouse melanoma ear skin model Laminins constitute a large family of αβγ heterotrimeric gly-
has demonstrated angiotropism and PM by intravital imag- coproteins which serve as the principal component of the
ing of human melanoma cells [26]. Finally, PM and perivas- BM. Laminins seem to have first appeared about 500 million
cular migration of human cutaneous and uveal melanoma years ago with the development of multicellular organisms
cells have been visualized in a zebrafish melanoma [29] and are considered to be the first extracellular proteins [35].
(Fig. 8.3g,h). BM laminins regulate cellular phenotypes and differentia-
Gene expression studies of human angiotropic primary tion during organogenesis and cancer progression [6, 35–37].
cutaneous melanoma have identified 15 genes potentially By remodeling BM and adhesion proteins, embryonic and
relevant to angiotropism and PM [6]. Utilization of our cancer stem cells are able to develop, grow, and migrate [38].
in vitro model of PM for gene expression microarray analy- This will modify cell signaling pathways and in turn the
78 R. Barnhill and C. Lugassy

a b c

d e f

g h
8 The Role of Angiotropic Extravascular Migratory Metastasis in Metastases 79

cytoskeleton by modulating molecules in cryptic sites of the ses of breast carcinoma [45]. Furthermore, the γ2 laminin
extracellular matrix (ECM). Therefore there is a “remodel- chain, which plays a role in cancer cell motility, is strongly
ing” and not a “degradation” of the ECM and BM during expressed at the invasion front, and this correlates with bad
embryogenesis and cancer progression. Indeed, at the ultra- prognosis [46, 47]. α4-Laminins, including laminins 411 and
structural level, conventional, stable BM is organized in a 421, can be expressed by vessels and tumor cells [48]. In co-­
supramolecular structure, while during embryogenesis and option and PM progression of glial brain tumors, laminin-421
cancer progression, remodeled BM loses its organizing is switched to laminin-411 (α4β1γ1) [48], and this is associ-
structure [39]. ated with higher tumor grade and with expression CSC
In PM, tumor cells are present along the vascular ablumi- markers including CD133, Nestin, c-Myc, and Notch path-
nal surface, i.e., along a vascular BM which contains laminin. way members [43, 48].
Ultrastructural studies of angiotropism/PM in melanoma The interaction of leukemia cells and the abluminal vas-
have shown an BM material lacking the classical supramo- cular surfaces of bridging vessels triggers the expression
lecular structure between tumor and endothelial cells [4, 39]. of α6 integrin by leukemia cells with vascular laminin,
This material contains mainly laminin, and could expose lam- promoting the invasion of the central nervous system via
inin cryptic sites increasing cell migration [40]. The detection PM [32].
of such alterations may require ultrastructural examination or In addition, laminins containing the α5 chain play a role
other more refined analysis. Additional studies are needed to in cell-cell adhesion and migration in several cancers, nota-
clarify the interactions of tumor cells with the BM, in particu- bly via Lutheran (Lu) (also named basal cell adhesion mol-
lar, potential structural and functional alterations of laminins ecule [49]. Lu/BCAM is a laminin α5 receptor, a subunit of
and other associated molecules in PM. LM-511. Importantly, laminin α5 is essential during cranial
The role for laminins in cancer progression is well known. NC migration [50]. Recently, it has been also shown that dur-
Notably, experimental data show that PM in melanoma over-­ ing PM, L1CAM enables the adhesion of metastasis-­
expressed laminin chains and receptors [41, 42]. A laminin initiating cells to laminin-rich BM [51].
γ1 chain synthetic peptide (peptide C16) significantly pro- Concerning the replacement pattern in liver metastases, it
motes cancer progression via adhesion, metastasis, and is important to emphasize that in normal liver, there is a non-­
migration [43]. C16 also enhances PM of human melanoma electron dense BM-like structure with reticular collagen 4
cells [27]. In addition, melanoma cells produce several lam- and laminin between the endothelium and the hepatocytes
inin isoforms and migrate on the α5 laminin chain [44]. [21] and thus the two cell types are in discontinuous close
Laminin-332 is implicated in tumor migration and metasta- contact. Importantly, stellate cells, corresponding to peri-
sis. Laminin-332 in lungs may stimulate pulmonary metasta- cytes, are directly aligned along endothelial cells.

Fig. 8.3 In vivo studies of PM cross-section of the murine brain vessel confirming that melanoma cells
(a–c) PM and angiotropism in a genetically engineered mouse model. are external to the vessels, without intravasation. (e) Expression of
A. Macroscopically visible melanoma cell expansion along 25 mm of a Serpin B2 in the murine brain melanoma model. Strong co-localization
dermal blood vessels (arrows). B-C. Histologic(above) and immunohis- of Serpin B2 (purple) to the angiotropic GFP melanoma cells (green)
tochemical (below) analysis of angiotropism in the murine model (b), along vessels (red tomato lectin). Scale bars, 50 μm. (f) Expression of
in a human primary melanoma (c). Note the similar angiotropic images Serpin B2 by angiotropic melanoma cells in a human brain melanoma
in b and c. (d–f) Human melanoma cells in a murine melanoma model metastasis. Striking expression of SERPIN B2 (dark red) by melanoma
brain model. (d) Montage juxtaposing PM in the murine brain mela- cells is observed predominately along the abluminal surfaces of micro-
noma model (main picture) and a human sample of a melanoma meta- vascular channels. (g, h) Intravital observation of PM of melanomas in
static to the brain. On the left part of the image, in the murine model at a zebrafish xenograft. (g) Xenograft with GFP green human cutaneous
4 weeks, PM of green GFP-labeled melanoma cells are visible along melanoma cells. Angiotropism/PM of green GFP tumor cells along the
red vessels (red tomato lectin). Juxtaposed to the right of the same external surface of the caudal vein (red tomato lectin). On the right:
image: human sample of a melanoma metastatic to the brain. time-lapse images of the angiotropic cell in white square taken at 0, 4,
Microvessels (filled by red blood cells) are extensively coated by a and 8 h after the beginning of the imaging. (h) Xenograft with GFP
layer of melanin-containing melanoma cells, without intravasation, human uveal melanoma cells. Similar image along the external surface
demonstrating significant angiotropism and PM. Note the correspon- of red intersegmental vessels. Note single (g and h) and collective (h,
dence of the two juxtaposed images. Scale bars: 100 μm. Insert: XZ white square) tumor cell migration along vessels [2]
80 R. Barnhill and C. Lugassy

8.6 Reversion to Embryogenesis-Derived 8.6.1.2 S  tem Cells and Competition

Programs: Implications in Angiotropic in Embryogenesis and Cancer
EVMM Progression
CSCs are a subpopulation of highly plastic tumor cells via
In the mid-nineteenth Rudolf Virchow observed that “neo- EMT to MET switch [55, 56]. CSC properties include self-­
plasms arise in accordance with the same laws that regulate renewal, migratory, metastasizing, and drug resistance [26].
embryonic development” [52] emphasizing for the first time In the in vitro model of PM, the interaction of tumor cells
this important relationship between tumor development and with endothelial tubules induces the stem-embryonic-like
embryonic development. Embryogenesis involves major characteristics of tumor cells [7], like tumor cells in vascular
morphological changes and cellular migrations that are imi- niches [26].
tated and repeated in cancer development [2, 53]. Cell competition in development is described as a “fitness-­
Furthermore, re-activation of embryonic cellular signaling in sensing embryonic mechanism” resulting in the replacement
adult cells, as a consequence of mutational and epigenetic of cells less fit by those with greater fitness. Cell competition
programming, is prototypic of cancer progression [54]. can be hijacked by cancer cells tumor progression [60],
including competition for a niche [60], which is relevant to
co-option and PM. In particular, angiotropic cancer cells
8.6.1 Epithelial-to-Mesenchymal Transition may compete with pericytes (or other perivascular mesen-
in the Embryo and in Cancer chymal stem in larger or specialized vessels) during co-­
option and PM in order to reach the external vascular surfaces
The epithelial-to-mesenchymal transition (EMT) is a critical of preexisting vessels. The mechanism may also involve
process during embryonic development and organogenesis nascent vascular structures during neoangiogenesis, as sug-
[55]. In EMT, epithelial cells lose their epithelial properties, gested by histological and experimental observations [4, 24].
including both apical-basal polarity and cell-to-cell adhe-
sion, and acquire migratory and invasive mesenchymal cell 8.6.1.3 T he Seed and the Soil: Relevance
characteristics in order to spread from their tissue site of ori- to the Embryo and Cancer Development
gin. As described below, PM and EVMM share identical The “seed and soil” theory as originally described for cancer
properties. In a similar fashion, a switch between EMT and cells by Paget [58, 61, 62] is unquestionably relevant to
mesenchymal-epithelial-transition (MET) is fundamental organogenesis. Embryonic stem cells are the “seeds” migrat-
during both embryonic and cancer progression. MET facili- ing throughout the embryo to reach their ultimate phenotype-­
tates metastatic outgrowth and probably vascular co-option specific “soils” in order to grow and develop organs
within distant organs [55]. (organogenesis) [2]. This is particularly relevant to the devel-
EMT/MET in metastasis, as in embryogenesis, may result opment of the neural crest. The neural crest is a highly migra-
in a spectrum of phenotypes between epithelial and mesen- tory embryonic cell population corresponding to the “seeds”
giving rise to numerous lineages developing in their specific
chymal cell opposite poles of the spectrum [56]. This revers-
ible “phenotype switch” model of metastasis may be driven “soils” [57, 63].
by BRN2 in melanoma metastasis [57]. Notably, neural crest cells migrate along developing
intersomitic vessels at multiple locations along their travel
8.6.1.1 Gene Expression in the Embryo [64]. This mechanism of neural crest cell migration is very
and Cancer similar to that of tumor cells during angiotropic EVMM.
It is well known that human tumors express the early gene Furthermore, human metastatic melanoma cells, when
expression of development, for example, Wnt, Hedgehog, transplanted into the chick embryonic neural crest, respond
and Notch pathways [26]. YAP is linked to PM [58], and the to signals from the surrounding microenvironment and
Hippo pathway is essential in cell migration during organo- migrate along stereotypical migratory routes of neural crest
genesis [59]. cells [65]. Thus it appears that “melanoma revives an embry-
As already mentioned, during melanoma PM in vitro onic migration program to promote plasticity and invasion”
there is an expression of genes associated with EMT, embry- and reversion to an aberrant neural crest genes may promote
onic, and cancer stem cells (CSC) properties [7, 8]. melanoma plasticity and invasiveness [66]. Likewise, the
Furthermore, development, cell movement, cancer, and human gliomas also appear to recapitulate the migratory
embryonic development are among the most significantly behavior of glial progenitors, as oligodendrocyte precursors
enriched functional groups analyzed by Ingenuity-Pathway-­ migrating along the vascular system in the developing ner-
Analysis [7]. vous system [67].
8 The Role of Angiotropic Extravascular Migratory Metastasis in Metastases 81

Other stages of development have also demonstrated the it take for the tumor cells to reach secondary sites without
same analogies, for example, gastrulation [68]. Importantly, intravascular dissemination? Migration modes and dynamics
cancer cells in an embryo form a “normal mosaic” organ- can be observed by intravital imaging of invasive human
ism [69]. tumor cells [77]. For example, the in vivo velocity (in cm per
year) of human melanoma varies from 22 to 525 cm/year,
8.6.1.4 A  bsence of a Hematogenous Pathway while human breast cancer cells migrate from 13 to 220 cm/
during Organogenesis year [77]. These speeds are theoretically compatible with the
In non-hematogenous cancer migration, or PM/EVMM, the long-time intervals between the recognition of the primary
“seed” tumor cells migrate toward secondary sites (“soil”), cancer and the formation of metastases and could be appli-
while in non-angiogenic tumor growth, the “seed” tumor cable to PM/EVMM [2]. These long periods of time for PM/
cells grow in primary and secondary organs (“soils”). EVMM may explain in part the “dormancy” period observed
Intravascular cell migration to reach their soil (“inefficient” with many tumors. Interestingly, neural crest cell migration
hematogenous metastasis) [3]) doesn’t exist during organo- rates are comparable to migrating tumor cells [78].
genesis. Indeed, the maternal circulation to the placenta is Taken together, these data may indicate that PM/EVMM
not present before the end of the first trimester [70, 71]. of angiotropic melanoma cells represent a reversion to
Therefore neural crest cell “migrate as “seeds” during an embryonic forms of migration.
extensive continuous migration through the embryo to reach
their “soils” without entering the vascular lumen,” i.e.,
using among other mechanisms PM/EVMM [2]. 8.8 Conclusion and Perspectives
Embryonic signals activated in cancer are well described
[53, 54, 68]. It will be important to consider the similarities Given (i) the well-recognized analogies between neural crest
between PM/EVMM in cancers with different embryonic and melanoma cell migration, (ii) the similarities between
origins and the respective pathways of their progenitors dur- neural crest and cancer cell migratory pathways, (iii) the
ing embryonic development. For example, the replacement absence of circulation during embryonic organogenesis, and
growth pattern in liver metastases has strong analogies with therefore migration of stem cells via PM/extravascular
liver organogenesis [72] and embryonic cell competition migration, all in concert support the concept that these extra-
[60]. Indeed, early endothelial cells and nascent vascular vascular embryonic migratory mechanisms recur abnormally
structures interact with newly derived hepatic endoderm, during melanoma metastasis. Therefore, embryonic factors
prior to liver bud emergence, and promote hepatic morpho- linked to migration and cancer metastasis can become rele-
genesis. Notably, hepatic sinusoids normally lack a continu- vant targets for future therapies.
ous BM, although membrane-like structure with reticular Importantly, for several decades, “reductionist” approaches
collagen 4 and laminin is present [73]. In liver diseases, lam- to cancer chemotherapies including recent immunotherapy
inin deposits in BM membrane. Concomitantly, the sinusoi- for solid tumors have resulted in outcome failure rates of 90%
dal endothelium loses its fenestrae and is transformed into (±5) according to governmental agencies and industry” [79].
vascular type endothelium [73]. Such changes in liver metas- A more comprehensive understanding of the mechanism(s) of
tases may represent a reversion toward embryonic sinusoids, tumor spread is greatly needed. Since angiotropic EVMM is
used as tracks for cancer cells [73]. now recognized as a migratory mechanism of cancer spread
Taken together, these data strongly suggest that angiotro- [80, 81], this may profoundly influence and stimulate the
pism during PM/EVMM represents a reversion to embryonic development of more powerful and effective therapeutics.
migratory pathways.

Future Directions in the Field of Angiotropism, PM, and

8.7 Chronology and Latency EVMM
• What is the fundamental biological and molecular
Cancer dormancy is a term used to explain the extensive mechanism of angiotropic extravascular tumor cell
periods of time in which patients remain asymptomatic prior migration?
to relapse [74]. Dormancy, which refers to the presence in • What is the biological basis of resistance associated
metastatic niches of dormant CSC/circulating tumor cells with angiotropic melanoma (cancer) cells?
after extravasation, is a loosely defined phenomenon that • What governs the “switch” between extravascular
remains poorly understood [74]. In addition, it is possible tumor cell migration and tumor cell proliferation?
that the dormancy period corresponds to angiotropic cancer • What strategies may be used for the inhibition of vas-
cells migrating toward the metastatic sites [75]. PM/EVMM cular adhesion and angiotropic tumor migration?
represents a mesenchymal-type slow movement process [76] • What other therapeutic initiatives may be developed
and provides an alternative explanation for the delayed for- for targeting PM and EVMM?
mation of metastases in many cancer cases. How long would
82 R. Barnhill and C. Lugassy

References for scoring the histopathological growth patterns of liver metasta-

sis. Br J Cancer. 2017;117(10):1427–41.
18. Barnhill R, Vermeulen P, Daelemans S, van Dam PJ, Roman-­
1. Chitty JL, Filipe EC, Lucas MC, Herrmann D, Cox TR, Timpson
Roman S, Servois V, et al. Replacement and desmoplastic histo-
P. Recent advances in understanding the complexities of metas-
pathological growth patterns: a pilot study of prediction of outcome
tasis. F1000Res. 2018;7:F1000 Faculty Rev-1169. https://doi.
in patients with uveal melanoma liver metastases. J Pathol Clin Res.
org/10.12688/f1000research.15064.2.
2018;4(4):227–40.
2. Lugassy C, Kleinman HK, Vermeulen PB, Barnhill
19. Barnhill R, van Dam PJ, Vermeulen P, Champenois G, Nicolas A,
RL. Angiotropism, pericytic mimicry and extravascular migratory
Rawson RV, et al. Replacement and desmoplastic histopathologi-
metastasis: an embryogenesis-derived program of tumor spread.
cal growth patterns in cutaneous melanoma liver metastases: fre-
Angiogenesis. 2020;23(1):27–41. https://doi.org/10.1007/s10456-­
quency, characteristics, and robust prognostic value. J Pathol Clin
019-­09695-­9. Epub 2019 Nov 12.
Res. 2020;6(3):195–206.
3. Talmadge JE, Fidler IJ. AACR centennial series: the biol-
20. Kordes C, Sawitza I, Götze S, Herebian D, Häussinger D. Stellate
ogy of cancer metastasis: historical perspective. Cancer Res.
cells are mesenchymal stem cells. Eur J Med Res. 2014;19(Suppl
2010;70(14):5649–69. https://doi.org/10.1158/0008-­5472.CAN-­
1):S6.
10-­1040. Epub 2010 Jul 7.
21. Häussinger D, Kordes C. Space of Disse: a stem cell niche in the
4. Lugassy C, Eyden BP, Christensen L, Escande JP. Angio-tumoral
liver. Biol Chem. 2019;401(1):81–95.
complex in human malignant melanoma characterised by free
22. Lugassy C, Scolyer R, Long G, Menzies A, Mischel P. Barnhill RL
laminin: ultrastructural and immunohistochemical observations. J
PDGFBR expression in anti-BRAF resistant melanoma: are angio-
Submicrosc Cytol Pathol. 1997;29(1):19–28.
tropic melanoma cells a source of BRAF resistance and disease
5. Lugassy C, Dickersin GR, Christensen L, Karaoli T, LeCharpentier
progression? J Cutan Pathol. 2014;41:159–60.
M, Escande JP, Barnhill RL. Ultrastructural and immunohisto-
23. Lugassy C, Kleinman HK, Fernandez PM, Patierno SR, Webber
chemical studies of the periendothelial matrix in human mela-
MM, Ghanem G, et al. Human melanoma cell migration along
noma: evidence for an amorphous matrix containing laminin. J
capillary-like structures in vitro: a new dynamic model for
Cutan Pathol. 1999;26(2):78–83. https://doi.org/10.1111/j.1600-
studying extravascular migratory metastasis. J Invest Dermatol.
­0560.1999.tb01806.x.
2002;119(3):703–4.
6. Barnhill RL, Lugassy C. Angiotropic malignant melanoma and
24. Lugassy C, Kleinman HK, Engbring JA, Welch DR, Harms JF,
extravascular migratory metastasis: description of 36 cases with
Rufner R, et al. Pericyte-like location of GFP-tagged melanoma
emphasis on a new mechanism of tumour spread. Pathology.
cells: ex vivo and in vivo studies of extravascular migratory metas-
2004;36(5):485–90.
tasis. Am J Pathol. 2004;164(4):1191–8.
7. Lugassy C, Wadehra M, Li X, Corselli M, Akhavan D, Binder
25. Bald T, Quast T, Landsberg J, Rogava M, Glodde N, Lopez-­
SW, Péault B, Cochran AJ, Mischel PS, Kleinman HK, Barnhill
Ramos D, et al. Ultraviolet-radiation-induced inflammation
RL. Pilot study on “pericytic mimicry” and potential embryonic/
promotes angiotropism and metastasis in melanoma. Nature.
stem cell properties of angiotropic melanoma cells interact-
2014;507(7490):109–13.
ing with the abluminal vascular surface. Cancer Microenviron.
26. Lee G, Hall RR 3rd, Ahmed AU. Cancer stem cells: cellular plas-
2013;6(1):19–29.
ticity, niche, and its clinical relevance. J Stem Cell Res Ther.
8. Lugassy C, Péault B, Wadehra M, Kleinman HK, Barnhill
2016;6(10):pii: 363.
RL. Could pericytic mimicry represent another type of melanoma
27. Lugassy C, Kleinman HK, Vernon SE, Welch DR, Barnhill RL. C16
cell plasticity with embryonic properties? Pigment Cell Melanoma
laminin peptide increases angiotropic extravascular migration of
Res. 2013;26(5):746–54.
human melanoma cells in a shell-less chick chorioallantoic mem-
9. Lugassy C, Barnhill RL, Christensen L. Melanoma and extravascu-
brane assay. Br J Dermatol. 2007;157(4):780–2.
lar migratory metastasis. J Cutan Pathol. 2000;27(9):481.
28. Bentolila LA, Prakash R, Mihic-Probst D, Wadehra M, Kleinman
10. Celià-Terrassa T, Kang Y. Metastatic niche functions and therapeu-
HK, Carmichael TS, et al. Imaging of angiotropism/vascular co-­
tic opportunities. Nat Cell Biol. 2018;20(8):868–77.
option in a murine model of brain melanoma: implications for
11. Mitchell MJ, Denais C, Chan MF, Wang Z, Lammerding J,
melanoma progression along extravascular pathways. Sci Rep.
King MR. Lamin a/C deficiency reduces circulating tumor
2016;6:23,834.
cell resistance to fluid shear stress. Am J Physiol Cell Physiol.
29. Fornabaio G, Barnhill RL, Lugassy C, Bentolila LA, Cassoux N,
2015;309(11):C736–46.
Roman-Roman S, et al. Angiotropism and extravascular migra-
12. Moose DL, Krog BL, Kim TH, Zhao L, Williams-Perez S, Burke
tory metastasis in cutaneous and uveal melanoma progression in a
G, et al. Cancer cells resist mechanical destruction in circulation
zebrafish model. Sci Rep. 2018;8(1):10,448.
via RhoA/actomyosin-dependent mechano-adaptation. Cell Rep.
30. Cheng L, Huang Z, Zhou W, Wu Q, Donnola S, Liu JK, et al.
2020;30(11):3864–3874.e6.
Glioblastoma stem cells generate vascular pericytes to support ves-
13. Friedl P, Alexander S. Cancer invasion and the microenvironment:
sel function and tumor growth. Cell. 2013;153(1):139–52.
plasticity and reciprocity. Cell. 2011;147(5):992–1009.
31. Er EE, Valiente M, Ganesh K, Zou Y, Agrawal S, Hu J, et al.
14. Gritsenko PG, Ilina O, Friedl P. Interstitial guidance of cancer inva-
Pericyte-like spreading by disseminated cancer cells activates
sion. J Pathol. 2012;226(2):185–99.
YAP and MRTF for metastatic colonization. Nat Cell Biol.
15. Beunk L, Brown K, Nagtegaal I, Friedl P, Wolf K. Cancer invasion
2018;20(8):966–78.
into musculature: mechanics, molecules and implications. Semin
32. Yao H, Price TT, Cantelli G, Ngo B, Warner MJ, Olivere L, et al.
Cell Dev Biol. 2019;93:36–45.
Leukaemia hijacks a neural mechanism to invade the central ner-
16. Barnhill RL, Ye M, Batistella A, Stern MH, Roman-Roman S,
vous system. Nature. 2018;560(7716):55–60.
Dendale R, et al. The biological and prognostic significance of
33. Lugassy C, Vernon SE, Warner JW, Le CQ, Manyak M, Patierno SR,
angiotropism in uveal melanoma. Lab Investig. 2017; https://doi.
et al. Angiotropism of human prostate cancer cells: implications for
org/10.1038/labinvest.2017.16.
extravascular migratory metastasis. BJU Int. 2005;95(7):1099–103.
17. van Dam PJ, van der Stok EP, Teuwen LA, Van den Eynden GG,
34. Lai CJ, Lin CY, Liao WY, Hour TC, Wang HD, Chuu CP. CD44
Illemann M, Frentzas S, et al. International consensus guidelines
promotes migration and invasion of docetaxel-resistant prostate
8 The Role of Angiotropic Extravascular Migratory Metastasis in Metastases 83

cancer cells likely via induction of hippo-Yap signaling. Cell. 55. Campbell K. Contribution of epithelial-mesenchymal transitions
2019;8(4) to organogenesis and cancer metastasis. Curr Opin Cell Biol.
35. Ober EA, Lemaigre FP. Development of the liver: insights into 2018;55:30–5.
organ and tissue morphogenesis. J Hepatol. 2018;68(5):1049–62. 56. Saitoh M. Involvement of partial EMT in cancer progression. J
36. Yap L, Tay HG, Nguyen MTX, Tjin MS, Tryggvason K. Laminins Biochem. 2018;164(4):257–64.
in cellular differentiation. Trends Cell Biol. 2019;29(12):987–1000. 57. Fane ME, Chhabra Y, Smith AG, Sturm RA. BRN2, a POUerful
37. Engbring JA, Kleinman HK. The basement membrane matrix in driver of melanoma phenotype switching and metastasis. Pigment
malignancy. J Pathol. 2003;200(4):465–70. Cell Melanoma Res. 2019;32(1):9–24.
38. Qin Y, Rodin S, Simonson OE. Hollande F laminins and cancer 58. Ribatti D, Mangialardi G, Vacca A. Stephen Paget and the 'seed
stem cells: partners in crime? Semin Cancer Biol. 2017;45:3–12. and soil' theory of metastatic dissemination. Clin Exp Med.
39. Rousselle P, Scoazec JY. Laminin 332 in cancer: when the extracel- 2006;6(4):145–9.
lular matrix turns signals from cell anchorage to cell movement. 59. Wu Z, Guan KL. Hippo signaling in embryogenesis and develop-
Semin Cancer Biol. 2020;62:149–65. ment. Trends Biochem Sci. 2021;46(1):51–63.
40. Sekiguchi R, Yamada KM. Basement membranes in development 60. Di Gregorio A, Bowling S, Rodriguez TA. Cell competition and its
and disease. Curr Top Dev Biol. 2018;130:143–91. role in the regulation of cell fitness from development to cancer.
41. Giannelli G, Falk-Marzillier J, Schiraldi O, Stetler-Stevenson Dev Cell. 2016;38(6):621–34.
WG. Quaranta V induction of cell migration by matrix metallopro- 61. Paget S. The distribution of secondary growths in cancer of the
tease-­2 cleavage of laminin-5. Science. 1997;277(5323):225–8. breast. Lancet. 1889;133(3421):571–3.
42. Lugassy C, Torres-Muñoz JE, Kleinman HK, Ghanem G, Vernon 62. Lugassy C, Barnhill RL. Concepts of cancer and metastasis: his-
SE, Barnhill RL. Over-expression of malignancy-associated torical perspective. from biology to target. Cancer Res. 2020;80(19
laminins and laminin receptors by angiotropic human melanoma Suppl):IA06.
cells in a chick chorioallantoic membrane model. J Cutan Pathol. 63. Dupin E, Calloni GW, Coelho-Aguiar JM, Le Douarin NM. The
2009;36(12):1237–43. issue of the multipotency of the neural crest cells. Dev Biol.
43. Kuratomi Y, Nomizu M, Tanaka K, Ponce ML, Komiyama S, 2018;444(Suppl 1):S47–59.
Kleinman HK, et al. Laminin gamma 1 chain peptide, C-16 64. George L, Dunkel H, Hunnicutt BJ, Filla M, Little C, Lansford
(KAFDITYVRLKF), promotes migration, MMP-9 secretion, and R, et al. In vivo time-lapse imaging reveals extensive neural crest
pulmonary metastasis of B16-F10 mouse melanoma cells. Br J and endothelial cell interactions during neural crest migration and
Cancer. 2002;86(7):1169–73. formation of the dorsal root andsympathetic ganglia. Dev Biol.
44. Oikawa Y, Hansson J, Sasaki T, Rousselle P, Domogatskaya A, 2016;413(1):70–85.
Rodin S, et al. Melanoma cells produce multiple laminin isoforms 65. Kulesa PM, Kasemeier-Kulesa JC, Teddy JM, Margaryan NV, Seftor
and strongly migrate on α5 laminin(s) via several integrin recep- EA, Seftor RE, et al. Reprogramming metastatic melanoma cells to
tors. Exp Cell Res. 2011;317(8):1119–33. assume a neural crest cell-like phenotype in an embryonic microen-
45. Carpenter PM, Sivadas P, Hua SS, Xiao C, Gutierrez AB, Ngo T, vironment. Proc Natl Acad Sci U S A. 2006;103(10):3752–7.
et al. Migration of breast cancer cell lines in response to pulmonary 66. Bailey CM, Morrison JA, Kulesa PM. Melanoma revives an embry-
laminin 332. Cancer Med. 2017;6(1):220–34. onic migration program to promote plasticity and invasion. Pigment
46. Tsubota Y, Ogawa T, Oyanagi J, Nagashima Y, Miyazaki Cell Melanoma Res. 2012;25(5):573–83.
K. Expression of laminin gamma2 chain monomer enhances 67. Tsai HH, Niu J, Munji R, Davalos D, Chang J, Zhang H, et al.
invasive growth of human carcinoma cells in vivo. Int J Cancer. Oligodendrocyte precursors migrate along vasculature in the devel-
2010;127(9):2031–41. oping nervous system. Science. 2016;351(6271):379–84.
47. Teng Y, Wang Z, Ma L, Zhang L, Guo Y, Gu M, et al. Prognostic 68. Cofre J, Saalfeld K, Abdelhay E. Cancer as an embryologi-
significance of circulating laminin gamma2 for early-stage non-­ cal phenomenon and its developmental pathways: a hypothesis
small-­cell lung cancer. Onco Targets Ther. 2016;9:4151–62. regarding the contribution of the noncanonical Wnt pathway.
48. Sun T, Patil R, Galstyan A, Klymyshyn D, Ding H, Chesnokova A, ScientificWorldJournal. 2019;2019:4714781.
et al. Blockade of a Laminin-411-notch Axis with CRISPR/Cas9 69. Mintz B, Illmensee K. Normal genetically mosaic mice produced
or a nanobioconjugate inhibits glioblastoma growth through tumor-­ from malignant teratocarcinoma cells. Proc Natl Acad Sci U S A.
microenvironment cross-talk. Cancer Res. 2019;79(6):1239–51. 1975;72(9):3585–9.
49. Kikkawa Y, Ogawa T, Sudo R, Yamada Y, Katagiri F, Hozumi K, 70. Burton GJ, Watson AL, Hempstock J, Skepper JN, Jauniaux
et al. The lutheran/basal cell adhesion molecule promotes tumor E. Uterine glands provide histiotrophic nutrition for the human
cell migration by modulating integrin-mediated cell attachment to fetus during the first trimester of pregnancy. J Clin Endocrinol
laminin-511 protein. J Biol Chem. 2013;288(43):30990–1001. Metab. 2002;87(6):2954–9.
50. Hutchins EJ, Bronner ME. Draxin alters laminin organization dur- 71. Burton GJ, Jauniaux E. Development of the human placenta and
ing basement membrane remodeling to control cranial neural crest fetal heart: synergic or independent? Front Physiol. 2018;9:373.
EMT. Dev Biol. 2019;446(2):151–8. 72. Matsumoto K, Yoshitomi H, Rossant J, Zaret KS. Liver organo-
51. Ganesh K, Basnet H, Kaygusuz Y, Laughney AM, He L, Sharma genesis promoted by endothelial cells prior to vascular function.
R, et al. L1CAM defines the regenerative origin of metastasis-­ Science. 2001;294(5542):559–63.
initiating cells in colorectal cancer. Nat Cancer. 2020;1(1):28–45. 73. Mak KM, Mei R. Basement membrane type IV collagen and lam-
52. Virchow RLK. Cellular pathology, 1859 special ed. London: John inin: an overview of their biology and value as fibrosis biomarkers
Churchill; 1978. p. 204–7. of liver disease. Anat Rec (Hoboken). 2017;300(8):1371–90.
53. Cofre J, Abdelhay E. Cancer is to embryology as mutation is to 74. Yeh AC, Ramaswamy S. Mechanisms of cancer cell dormancy—
genetics: hypothesis of the cancer as embryological phenomenon. another hallmark of cancer? Cancer Res. 2015;75(23):5014–22.
ScientificWorldJournal. 2017;2017:3578090. https://doi.org/10.1158/0008-­5472.CAN-­15-­1370. Epub 2015 Sep 9.
54. Aiello NM, Stanger BZ. Echoes of the embryo: using the devel- 75. Kienast Y, von Baumgarten L, Fuhrmann M, Klinkert WE,
opmental biology toolkit to study cancer. Dis Model Mech. Goldbrunner R, Herms J, Winkler F. Real-time imaging reveals
2016;9(2):105–14. the single steps of brain metastasis formation. Nat Med.
84 R. Barnhill and C. Lugassy

2010;16(1):116–22. https://doi.org/10.1038/nm.2072. Epub 2009 therapeutic efficacy and unsustainable costs. Clin Transl Med.
Dec 20. 2018;7(1):11. https://doi.org/10.1186/s40169-­018-­0185-­6.
76. Friedl P, Wolf K. Proteolytic and non-proteolytic migration of 80. Bald T, Quast T, Landsberg J, Rogava M, Glodde N, Lopez-Ramos
tumour cells and leucocytes. Biochem Soc Symp. 2003;70:277–85. D, Kohlmeyer J, Riesenberg S, van den Boorn-Konijnenberg D,
https://doi.org/10.1042/bss0700277. Hömig-Hölzel C, Reuten R, Schadow B, Weighardt H, Wenzel
77. Clark AG, Vignjevic DM. Modes of cancer cell invasion and the D, Helfrich I, Schadendorf D, Bloch W, Bianchi ME, Lugassy C,
role of the microenvironment. Curr Opin Cell Biol. 2015;36:13–22. Barnhill RL, Koch M, Fleischmann BK, Förster I, Kastenmüller
https://doi.org/10.1016/j.ceb.2015.06.004. Epub 2015 Jul 14. W, Kolanus W, Hölzel M, Gaffal E, Tüting T. Ultraviolet-radiation-­
78. Li Y, Vieceli FM, Gonzalez WG, Li A, Tang W, Lois C, Bronner induced inflammation promotes angiotropism and metastasis in
ME. In vivo quantitative imaging provides insights into trunk neu- melanoma. Nature. 2014;507(7490):109–13.
ral crest migration. Cell Rep. 2019;26(6):1489–1500.e3. https:// 81. Massagué J, Ganesh K. Metastasis-initiating cells and ecosystems.
doi.org/10.1016/j.celrep.2019.01.039. Cancer Discov. 2021;11(4):971–94.
79. Maeda H, Khatami M. Analyses of repeated failures in cancer
therapy for solid tumors: poor tumor-selective drug delivery, low
 Part III
Dormant Cancer Cells and Their Reactivation in the
Host Cancer Microenvironment

Maria Soledad Sosa

Metastasis and its effects on the patients’ homeostatic functions are the main causes of cancer-­
related deaths. Clinical and experimental evidence supports that disseminated cancer cells
(DCCs) are the source of metastasis in many cancer types [1–5]. The data also support that
DCCs remain in a reversible dormant state for prolonged periods of time resisting therapies
until, eventually, they reactivate and form overt metastasis.
The program that regulates the cell life cycle of single DCCs is named cellular dormancy.
This program involves the ability of DCCs to persist most of the time in a cell cycle G0/G1
arrest which is initiated and/or maintained by the interactions of DCCs with the new niches at
secondary organs [6–8]. In addition, several studies suggest that the cellular dormancy pro-
grams may also be initiated within the primary tumor itself [9, 10] and then maintained or
augmented by secondary organ niche cues. Once at secondary organs, through epigenetic,
metabolic, and adaptive stress responses DCCs are reprogramed to evade the immune system
and survive the new niche-derived stress signals as well as therapies. Because dormancy is
reversible the inherent outcome of this life cycle is the emergence from the G0/G1 arrest and
the formation of metastatic outgrowth. Two excellent chapters in this book touch on important
aspects on how the pressure of treatment impacts dormancy and reactivation phenotypes and
the niches that control DCC dormancy and underlying mechanisms.
The cellular dormancy program can be activated spontaneously in several types of tumors,
even in the absence of a detectable primary tumor [11–14] and before any therapeutic interven-
tions. However, experimental and clinical data [15–24] support that therapy-induced dormancy
might occur in response to chemotherapy or targeted therapies used in the adjuvant setting,
albeit further investigation is required. In this regard, Garcia et al.’s review, among other fea-
tures, adapted metabolic reprogramming and pro-survival pathways acquired by residual dor-
mant cancer cells after therapies. This chapter provides evidence that many pathways are
differentially regulated in residual dormant cancer cells in response to therapies to allow for
the survival and further expansion of these tumor cells into metastasis.
As mentioned before, the dormant DCC life cycle has different stages including adaptation
to a new microenvironment by interactions with the new extracellular matrix (ECM) [1,
25–28]. Along these lines, Huber et al. review the cellular fates of dormant cancer cell in the
presence of fibronectin, collagens, growth factors associated to ECM, proteolysis, and stiffness
of ECM. Some of these ECM-derived signals constitute an appropriate niche for dormancy and
survival of DCCs whereas others are important for the reactivation of dormant DCCs into
metastasis.
Expanding our knowledge on how DCCs survive in a dormant state in their new microenvi-
ronments before and after treatments, how therapies modify DCC niches including ECM, and
what are the signals that trigger their reactivation programs may likely improve the prediction
of patient survival rates and increase the opportunities to design therapies that will extend the
remission phase.
86 Dormant Cancer Cells and Their Reactivation in the Host Cancer Microenvironment

References
1. Ghajar CM, Peinado H, Mori H, Matei IR, Evason KJ, Brazier H, et al. The perivascular niche regulates
breast tumour dormancy. Nat Cell Biol. 2013;15:807–17. https://doi.org/10.1038/ncb2767.
2. Malladi S, Macalinao DG, Jin X, He L, Basnet H, Zou Y, et al. Metastatic latency and immune evasion
through autocrine inhibition of WNT. Cell. 2016;165:45–60. https://doi.org/10.1016/j.cell.2016.02.025.
3. Gao H, Chakraborty G, Lee-Lim AP, Mo Q, Decker M, Vonica A, et al. The BMP inhibitor coco reac-
tivates breast cancer cells at lung metastatic sites. Cell. 2012;150:764–79. https://doi.org/10.1016/j.
cell.2012.06.035.
4. Albrengues J, Shields MA, Ng D, Park CG, Ambrico A, Poindexter ME, et al. Neutrophil extracellular
traps produced during inflammation awaken dormant cancer cells in mice. Science. 2018;361 https://
doi.org/10.1126/science.aao4227.
5. Sosa MS, Parikh F, Maia AG, Estrada Y, Bosch A, Bragado P, et al. NR2F1 controls tumour cell dor-
mancy via SOX9- and RARbeta-driven quiescence programmes. Nat Commun. 2015;6:6170. https://
doi.org/10.1038/ncomms7170.
6. Phan TG, Croucher PI. The dormant cancer cell life cycle. Nat Rev Cancer. 2020;20:398–411. https://
doi.org/10.1038/s41568-­020-­0263-­0.
7. Aguirre-Ghiso JA. Models, mechanisms and clinical evidence for cancer dormancy. Nat Rev Cancer.
2007;7:834–46.
8. Sosa MS, Bragado P, Aguirre-Ghiso JA. Mechanisms of disseminated cancer cell dormancy: an awak-
ening field. Nat Rev Cancer. 2014;14:611–22. https://doi.org/10.1038/nrc3793.
9. Fluegen G, Avivar-Valderas A, Wang Y, Padgen MR, Williams JK, Nobre AR, et al. Phenotypic hetero-
geneity of disseminated tumour cells is preset by primary tumour hypoxic microenvironments. Nat Cell
Biol. 2017;19:120–32. https://doi.org/10.1038/ncb3465.
10. Kim RS, Avivar-Valderas A, Estrada Y, Bragado P, Sosa MS, Aguirre-Ghiso JA, et al. Dormancy signa-
tures and metastasis in estrogen receptor positive and negative breast cancer. PLoS One. 2012;7:e35569.
https://doi.org/10.1371/journal.pone.0035569.
11. Harper KL, Sosa MS, Entenberg D, Hosseini H, Cheung JF, Nobre R, et al. Mechanism of early dis-
semination and metastasis in Her2(+) mammary cancer. Nature. 2016; https://doi.org/10.1038/
nature20609.
12. Hosseini H, Obradovic MM, Hoffmann M, Harper KL, Sosa MS, Werner-Klein M, et al. Early dissemi-
nation seeds metastasis in breast cancer. Nature. 2016; https://doi.org/10.1038/nature20785.
13. Linde N, Casanova-Acebes M, Sosa MS, Mortha A, Rahman A, Farias E, et al. Macrophages orches-
trate breast cancer early dissemination and metastasis. Nat Commun. 2018;9:21. https://doi.org/10.1038/
s41467-­017-­02481-­5.
14. Wang J, Wang G, Cheng D, Huang S, Chang A, Tan X, et al. Her2 promotes early dissemination of
breast cancer by suppressing the p38-MK2-Hsp27 pathway that is targetable by Wip1 inhibition.
Oncogene. 2020;39:6313–26. https://doi.org/10.1038/s41388-­020-­01437-­2.
15. Quesnel B. Tumor dormancy: long-term survival in a hostile environment. Adv Exp Med Biol.
2013;734:181–200. https://doi.org/10.1007/978-­1-­4614-­1445-­2_9.
16. Sharma SV, Lee DY, Li B, Quinlan MP, Takahashi F, Maheswaran S, et al. A chromatin-mediated
reversible drug-tolerant state in cancer cell subpopulations. Cell. 2010;141:69–80. https://doi.
org/10.1016/j.cell.2010.02.027.
17. Schewe DM, Aguirre-Ghiso JA. Inhibition of eIF2alpha dephosphorylation maximizes bortezomib effi-
ciency and eliminates quiescent multiple myeloma cells surviving proteasome inhibitor therapy. Cancer
Res. 2009;69:1545–52.
18. Fox DB, Garcia NMG, McKinney BJ, Lupo R, Noteware LC, Newcomb R, et al. NRF2 activation
promotes the recurrence of dormant tumour cells through regulation of redox and nucleotide metabo-
lism. Nat Metab. 2020;2:318–34. https://doi.org/10.1038/s42255-­020-­0191-­z.
19. El Touny LH, Vieira A, Mendoza A, Khanna C, Hoenerhoff MJ, Green JE. Combined SFK/MEK inhi-
bition prevents metastatic outgrowth of dormant tumor cells. J Clin Invest. 2014;124:156–68. https://
doi.org/10.1172/JCI70259.
Dormant Cancer Cells and Their Reactivation in the Host Cancer Microenvironment 87

20. Saleh T, Bloukh S, Carpenter VJ, Alwohoush E, Bakeer J, Darwish S, et al. Therapy-induced senes-
cence: an "old" friend becomes the enemy. Cancers (Basel). 2020;12 https://doi.org/10.3390/
cancers12040822.
21. Alvarez JV, Pan TC, Ruth J, Feng Y, Zhou A, Pant D, et al. Par-4 downregulation promotes breast can-
cer recurrence by preventing multinucleation following targeted therapy. Cancer Cell. 2013;24:30–44.
https://doi.org/10.1016/j.ccr.2013.05.007.
22. Kurppa KJ, Liu Y, To C, Zhang T, Fan M, Vajdi A, et al. Treatment-induced tumor dormancy through
YAP-mediated transcriptional reprogramming of the apoptotic pathway. Cancer Cell. 2020;37:104–
122 e112. https://doi.org/10.1016/j.ccell.2019.12.006.
23. Thompson VC, Day TK, Bianco-Miotto T, Selth LA, Han G, Thomas M, et al. A gene signature identi-
fied using a mouse model of androgen receptor-dependent prostate cancer predicts biochemical relapse
in human disease. Int J Cancer. 2012;131:662–72. https://doi.org/10.1002/ijc.26414.
24. Demicheli R, Desmedt C, Retsky M, Sotiriou C, Piccart M, Biganzoli E. Late effects of adjuvant che-
motherapy adumbrate dormancy complexity in breast cancer. Breast. 2020;52:64–70. https://doi.
org/10.1016/j.breast.2020.05.002.
25. Aguirre-Ghiso JA, Liu D, Mignatti A, Kovalski K, Ossowski L. Urokinase receptor and fibronectin
regulate the ERK(MAPK) to p38(MAPK) activity ratios that determine carcinoma cell proliferation or
dormancy in vivo. Mol Biol Cell. 2001;12:863–79.
26. Barkan D, El Touny LH, Michalowski AM, Smith JA, Chu I, Davis AS, et al. Metastatic growth from
dormant cells induced by a col-I-enriched fibrotic environment. Cancer Res. 2010;70:5706–16.
27. Boyerinas B, Zafrir M, Yesilkanal AE, Price TT, Hyjek EM, Sipkins DA. Adhesion to osteopontin in
the bone marrow niche regulates lymphoblastic leukemia cell dormancy. Blood. 2013;121:4821–31.
https://doi.org/10.1182/blood-­2012-­12-­475483.
28. Bragado P, Estrada Y, Parikh F, Krause S, Capobianco C, Farina HG, et al. TGF-beta2 dictates dissemi-
nated tumour cell fate in target organs through TGF-beta-RIII and p38alpha/beta signalling. Nat Cell
Biol. 2013;15:1351–61. https://doi.org/10.1038/ncb2861.
 Tumor Dormancy and Relapse
Regulated by the Extracellular Matrix 9
Hyuna Kim, Rebecca E. Huber, Rita Das Mahapatra,
Ning-Hsuan Tseng, and Shelly R. Peyton

Abstract 9.1 Introduction

Metastases account for 90% of cancer-related deaths, and
many patients experience relapse in distant organs, even Even after apparently successful therapy, disseminated
after successful initial therapy. This is due to dissemi- tumor cells can remain dormant for many years before out-
nated cells that remain dormant for many years until they growth. Breast cancer is a prominent example, where the
regrow in a distant site. There is increasing evidence presence of disseminated, quiescent tumor cells in the bone
showing that extracellular matrix (ECM) plays multiple marrow is a marker of poor prognosis [1, 2], indicating that
roles in promoting and maintaining dormancy, cell prolif- dormant disseminated tumor cells contribute to and predict
eration, and tumor outgrowth. For example, collagen can poor outcomes. Dormant metastases are kept quiescent either
induce a transition from dormancy to outgrowth in breast because individual cells are not proliferative (cellular dor-
cancer, fibronectin facilitates dormant cell survival, and mancy) or because an equilibrium is established between
other ECM proteins including TGFB-1 and periostin can proliferation and death across the bulk of the tumor mass
accelerate dormant tumor outgrowth. This chapter will (tumor dormancy) [3]. These dormant cancer cells are noto-
give an overview of such studies to elucidate the roles of riously difficult to treat, largely because they are not actively
these ECM proteins, other factors found in ECM, and the cycling, and so cannot be killed by the traditional chemo-
mechanical properties of ECM in regulating dormancy therapies typically given to patients with metastatic disease.
and reactivation of disseminated tumor cells. Much of the literature on this topic has focused on the transi-
tion of indolent micrometastases to overt metastases follow-
ing alterations to the tumor microenvironment [4–6].
Immunohistochemistry of fixed clinical and in vivo speci-
Learning Objectives mens can provide insight into localization of dormant or pro-
• To understand the known links between extracellu- liferating disseminated tumor cells (DTCs) within the
lar matrix and dormancy/relapse of epithelial extracellular matrix (ECM) via Ki-67 (a commonly used
cancers. proliferation marker) [7, 8], and studies have begun to iden-
• Understand the connection between collagen-­ tify the factors from the ECM that coincide with tumor out-
mediated matrix stiffening and cancer cell growth. Intravital imaging can observe cell dynamics from
dormancy. pre-implanted locations [9–11], biomaterials and chip tech-
• Understand the dual role fibronectin plays in pro- nologies have started to address these questions in vitro [12].
moting both dormancy and reactivation. These studies and others provide overwhelming evidence
that cell–ECM interactions regulate metastasis in vitro,
in vivo, and in humans [13, 14], and of the role of the ECM in
maintaining cell or tumor dormancy. There have been hints of
certain extracellular factors regulating dormancy, and the
H. Kim · R. E. Huber · R. D. Mahapatra · N.-H. Tseng “dormancy program” includes signals downstream of integrin
S. R. Peyton (*) engagement to ECM proteins (for reviews, see [15, 16]).
University of Massachusetts Amherst, Amherst, MA, USA
e-mail: [email protected]; [email protected]; Genetic screens have identified factors that permit dormant
[email protected]; [email protected]; cells to outgrow [17–19]. In related studies, cytokines [20],
[email protected] matrix proteins [7], and immune evasion [8] can promote cel-

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 89

S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_9
90 H. Kim et al.

lular quiescence, and current work is clarifying how dissemi- cell proliferation and migration. Deposition of collagen I in
nated cells eventually overcome these dormancy-­promoting breast tissue induces a transition from dormancy to metastatic
signals and transition to overt metastatic outgrowth. outgrowth [4]. In this study, they used two mouse mammary
Although not yet directly linked to dormancy, altered cancer cell lines: D2.0R (dormant) and D2A1 (metastatic).
expression of other matrix proteins has been shown to facili- These cell lines were derived in the Miller lab (D2A1: [28]
tate metastatic survival and outgrowth. As three examples, and D2.0R: [29]). These are commonly used cell lines for
fibronectin is produced by the stroma to support survival of dormancy studies, because both were derived from tumors
incoming disseminated tumor cells [21], tumor cells trigger arising from implants of the same D2 hyperplastic alveolar
stromal periostin secretion to metastasize to the lung [22], nodule line. D2A1 cells form metastatic lesions after a few
and breast cancer cell-derived tenascin C promotes lung weeks in mice, but D2.0R cells remain dormant for months
metastasis [23]. Understanding this shift from quiescence to with rare formation of metastatic lesions. In a three-dimen-
proliferation, alongside concurrent changes in the extracel- sional culture system, Type I Collagen induces proliferation
lular microenvironment, could provide oncologists with new in otherwise quiescent D2.0R cells via integrin β1 activation
avenues for therapy. This chapter will give an overview of of Src and focal adhesion kinase (FAK), leading to extracel-
the seminal contributions of researchers that have begun to lular signal-regulated kinase (ERK)-dependent myosin light
uncover the role of these ECM proteins and other ECM-­ chain (MLC) phosphorylation by myosin light chain kinase
associated factors in regulating dormancy and relapse of dis- and actin stress fiber formation (Fig. 9.1a). They also showed
seminated tumor cells, as well as introduce emerging that the induction of fibrosis in lung, associated with deposi-
technologies in the field. tion of Type I Collagen in the in vivo metastatic microenvi-
ronment, induces dormant D2.0R cells to form proliferative
cc Extracellular matrix (ECM): The scaffold of sugars, metastatic lesions through integrin β1.
proteins, and proteoglycans that surrounds cells in tissues. Although these and other studies demonstrate that collagen
The ECM provides structure, is a depot for growth factors deposition supports cancer cell outgrowth and proliferation,
and chemokines, provides cell binding sites, and imparts enriched collagen does not always support cancer cell out-
biophysical signaling to the cells within. growth. For example, in melanoma, an environment rich in
fibrillar collagen I induces quiescence of melanoma cells [30].
cc Integrins: The transmembrane heterodimeric proteins Multiple other studies show that tumor cell interaction with
that allow cells to bind to specific ECM proteins at different fibrillar collagen suppresses cell proliferation and keeps cells
tissue sites and initiate biochemical and biophysical signaling dormant [31, 32]. In this study, disorganized (from monomer)
within cells upon ECM binding. collagen gels and organized fibrillar collagen gels were used
to study human melanoma cell growth. When cells were on
cc Cellular dormancy: When a single cancer cell is disorganized collagen surfaces, cells proliferated. However,
quiescent for several years. organized collagen gels led to cell growth arrest in the G0/G1
phase of the cell cycle. When they treated the cells on orga-
cc Tumor dormancy: When the proliferation and death nized collagen gels with the tyrosine kinase inhibitor genis-
rates of a tumor cell population roughly equal, such that a tein, discoidin domain receptor 2 (DDR2) was downregulated
tumor is neither growing nor shrinking. and stimulated cell proliferation. Their study provides evi-
dence that fibrillar collagen induces G0/G1 cell cycle arrest
cc Relapse: When a tumor initiates proliferation after a period and restricts proliferation of tumor cells by a direct receptor-
of dormancy, resulting in a detectable, growing tumor mass. mediated mechanism that involves DDR2 (Fig. 9.1b).

9.3 Fibronectin
9.2 Collagens
Fibronectin (FN) is a glycoprotein found in the ECM of all
Collagens are the predominant structural proteins found in tissues [33]. It forms both linear and branched structures and
connective tissues. It has a triple helix structure, derived from exists as individual fibrils, complex matrices, or clustered
its primary structure: every third amino acid is a glycine resi- bundles. Converting cell-secreted FN molecules into more
due [24]. There are multiple isoforms of collagen, and many complex structures requires both FN–cell and FN–FN inter-
of them are known to influence cancer cell behaviors, such as actions. Secreted FN molecules first bind to each other,
invasion, proliferation, and metastasis [25]. Several studies dimerizing via disulfide bonds on the C terminus end of each
have found that high densities of collagen I in the epidermis molecule. Cells bind to FN via αVβ3 and α5β1 integrins on cell
[26] and collagen IV in the pancreas [27] stimulate cancer surfaces via its RGD and synergy domains, and both of these
9 Tumor Dormancy and Relapse Regulated by the Extracellular Matrix 91

a Deposition of collagen I induces cell proliferation c Increased matrix stiffness as a result of LOX release
Collagen I LOX
Collagen Crosslinked
collagen
2 1 2 1 2 1 2 1 2 1 2 1
Hypoxia induced LOX 2 1 2 1 2 1
release

FAK Increased pSrc pFAK

Src Src FAK LOX-mediated
deposition of
collagen pSrc pFAK
collagen I
crosslinking
pMEK/ERK
pMEK/ERK
Stiffer
environment pMLC
Action stress Cell proliferation
fiber formation

b Fibrillar collagen I induces G0/G1 cell cycle arrest via DDR2

Increased proliferation induced by
Soft matrix induces dormancy
Monomeric collagen Fibrillar collagen LOX on stiffer environment
(disorganized) (organized)
DDR2 DDR2

pERK p21 pERK p21 Collagen

Fibronectin
Cell Cell growth
growth arrest

Fig. 9.1 Collagen in ECM can alter cell proliferation or cell cycle arrest, is increased. When cells are in disorganized monomeric colla-
arrest. (a) Increased deposition of collagen I in ECM of breast tissue, gen, DDR2 is downregulated and thus cells can grow. Also, when cells
the collagen allows cells to metastatically outgrow, escaping from a are in fibrillar collagen but are treated with genistein, which inhibits
dormant state. Collagen I induces cell proliferation via integrin β1 acti- DDR2, cells enter the cell cycle even in the presence of fibrillar colla-
vation, which activates Src and focal adhesion kinase (FAK) signaling gen. Thus, fibrillar collagen in ECM environment induces cell cycle
pathway in its downstream, leading to extracellular signal-regulated arrest of melanoma cells by DDR2-mediated mechanism. (c) ECM
kinase (ERK)-dependent myosin light chain (MLC) phosphorylation. stiffness promotes outgrowth from dormant state. On the left, α2β1 inte-
MLC phosphorylation induces actin stress fiber formation, which then grin interacts to Collagen type I. Hypoxia induces expression of LOX
promotes cell stiffening and proliferation of preinvasive breast cancer enzymes. On the right, LOX mediates collagen cross-linking, resulting
cells. (b). When melanoma cells are in fibrillar collagen-rich environ- in increase in ECM stiffness. The increase in ECM stiffness allows dor-
ment, it activates discoidin domain receptor 2 (DDR2). When DDR2 is mant cells to proliferate via β1 activation which activates Src and FAK
activated, it downregulates ERK and thus p21, which triggers cell cycle signaling pathway. Created with BioRender.com

domains must be bound for fibril formation [34]. Once simultaneously, which drives ERK activity [35]. With this
bound, integrin-mediated signaling initiates cell contraction same cell line, they found that when uPAR is upregulated, it
via RhoA-ROCK, exposing cryptic FN-FN binding sites. interacts with α5β1 to produce more FN fibrils to amplify the
Integrin clustering also brings FN molecules closer together FN signal, increasing ERK activation and suppressing p38,
to promote these FN–FN interactions, which ultimately further shifting the balance toward proliferation [35]
results in a stiffer, insoluble, fibrillar matrix [35]. (Fig. 9.2). The authors then expanded that study to include
Apart from integrins, FN also binds to urokinase plasmin- multiple cancer cell lines, representing several cancer types,
ogen activator receptor (uPAR) which plays a major role in including breast, prostate, melanoma, and fibrosarcoma.
regulating the mitogen-activated protein kinase (MAPK) They found that with the exception of triple negative breast
pathways. Several studies have shown that a careful balance cancer cell line MDA-MB-231, the uPAR/integrin/FN com-
of MAPK signaling through ERK and p38MAPK (p38) regu- plex regulates ERK activity and the formation of this com-
lates the shift between dormancy and proliferation in cancer plex is a “perfect” predictor of proliferation vs. dormancy
cells [35–37]. Studies have shown that persistent activation [37].
of ERK causes tumor proliferation, and activation of p38 In addition to binding preexisting FN in the ECM, stromal
induces dormancy [35, 37]. FN–uPAR interactions are vital and cancer cells can also produce FN, impacting dormancy
to the orchestration of this balance. In head and neck cancer and relapse. In one study using an in vitro culture system to
(HEP3 cells), authors found cells bind to FN via α5β1 and study an array of cancer cell types, authors found that cells
uPAR (a non-integrin vitronectin cell adhesion receptor) able to produce a FN matrix had better overall survival dur-
92 H. Kim et al.

Dormant cancer cells Proliferating cancer cells on

on fibronectin matrix fibronectin monomers and fibrils

FN matrix degraded by
MMP-2 and MMP-9

pp38 pERK
uPAR

pERK pp38
uPAR

Fig. 9.2 Known fibronectin (FN) control of cellular dormancy. Left: ber and secrete FN fibrils as they proliferate. Illustrations of cell below
Dormant cancer cells and neighboring stroma produce a dense FN highlight major molecular pathways resulting from cell–FN interac-
matrix that supports dormant cell survival. The transition from dor- tions implicated in regulating a dormancy-to-proliferation switch.
mancy to proliferation prompts cancer cells to secrete MMP-2 and Created with BioRender.com
MMP-9 which degrade the FN matrix. Right: cancer cells grow in num-

ing dormancy than those that did not produce FN. The effect expression of TF led to tumor growth via recruitment of vas-
of the FN matrix persevered until the cells secreted matrix cular (CD105+) and myeloid (CD11b+ and F4/80+) cells.
metalloprotease 2 (MMP-2) and MMP-9 to degrade the When they first injected TF-lacking U373 cells in vivo, they
matrix, resulting in tumor proliferation [38]. This mirrors a were not immediately eliminated from immune-deficient
similar study in the lung, which found that cancer cell inter- mice. Rather, they remained viable and metabolically active at
action with stroma spurs production of FN by the dormant the injection site, but without evidence of palpable tumor
cancer cells and supports survival [39]. However, there are growth. The role of TF was not to initiate cancer cell prolifera-
some reports of FN regulating cancer dormancy in a contra- tion directly, but rather to recruit vascular and myeloid cells.
dictory manner. For example, a study compared the produc-
tion of FN between dormant and proliferative breast cell
lines from the D2A series, and they found the appearance of 9.4.1 ECM Proteolysis
FN stimulated proliferation [16].
The ECM is subject to continuous turnover, including break-
down of existing protein fibers and scaffolds by ECM prote-
9.4  ther ECM and ECM-Associated
O ases and rebuilding of these same scaffolds via ECM protein
Proteins deposition. It remains unclear how protease activity and
ECM turnover regulate the balance between dormancy and
Several studies have identified other individual matrix proteins relapse of cancer cells. Several studies have linked activation
that support dormancy and suppress overt metastasis, includ- and proliferation of dormant lung cancer cells to chronic
ing thrombospondin 1 [7] and e-selectin [40] on the microvas- inflammation in the lung [36, 44]. Albrengues et al. demon-
cular endothelium, and osteopontin in the bone marrow [41]. strated that chronic inflammation in lungs caused by infec-
Integrin engagement with these and other matrix-associated tion or cigarette smoke results in the formation of neutrophil
proteins, such as VWF and VCAM1, can protect these dor- extracellular traps (NETs, Fig. 9.3) [45]. Two NET-associated
mant, disseminated tumor cells from chemotherapy [42]. proteases, neutrophil elastase (NE) and MMP9, cleave lam-
Tissue factor, a cancer cell associated initiator of coagulation, inin in the local microenvironment, which further induces
can regulate dormancy in glioma [43]. They observed that a proliferation of dormant cells through activating α3β1 integ-
glioma cell line (U373, which normally do not produce tumors rin signaling, resulting in aggressive metastatic tumors.
in mice) deficient for TF remained viable but permanently Experiments with breast cancer cell lines (murine D2.0R and
dormant at an injection site for nearly a year, whereas inducing human MCF-7) injected intravenously into syngeneic or
9 Tumor Dormancy and Relapse Regulated by the Extracellular Matrix 93

Undetectable disseminated dormant cancer

Laminin Basement
membrane
Integrins
Awakening and cancer growth

Collagen
Dormant cancer cell

Inflammation caused by
e.g. tobacco smoke exposure

Awakening is prevented

NE and MMP9 associated with Neutrophil

Neutrophil Extracellular Traps
Cryptic
epitope

Fig. 9.3 NETs formed during lung inflammation awaken dormant tion of the previously dormant cancer cells. Antibodies blocking the
cancer cells. Lung inflammation—caused, for instance, by tobacco NET-generated laminin epitope prevent inflammation-induced awaken-
smoke exposure—leads to NET formation. Two proteases on the NETs ing. Reproduced with permission from The American Association of
sequentially cleave the extracellular matrix protein laminin. This cleav- Science [45]
age generates an epitope that activates integrin signaling and prolifera-

immunodeficient nude mice, respectively, established a may escape growth regulation in this “sub-niche.” Tumor
model of cancer cells dormancy-­to-­proliferation in the lungs, cells grow faster around neovascular tips, which are rich in
regulated by NETs. tumor-promoting factors such as active transforming growth
factor beta 1 (TGF-β1) and periostin.
Studies by Chen et al. revealed that abnormal expression
9.4.2 Growth Factors in the ECM of VCAM-1 in breast cancer cells is associated with relapse
in the lung [46]. To investigate whether VCAM-1 functions
Apart from direct binding to cells, the ECM also provides a as a mediator of metastasis, they used short hairpin RNA
binding depot for otherwise soluble growth factors in the (shRNA) interference to reduce its expression in a VCAM-1-
local microenvironment. DTCs dormant in these environ- overexpressing lung metastatic cell line (MDA231-LM2-4175)
ments could be kept quiescent in this local growth factor and implanted these cells into mammary glands of immuno-
milieu or awakened through the acute release of growth fac- deficient mice. They found VCAM-1 tethers metastasis-­
tors during ECM remodeling. To understand this, Ghajar associated macrophages to cancer cells via α4-containing
et al. [7] created organotypic models of lung- and bone integrins. Clustering of cell surface VCAM-1 triggers Akt
marrow-­microvascular niches where dormant DTCs reside activation via Ezrin and protects cancer cells from the pro-­
upon the microvasculature of lung, bone marrow, and brain. apoptotic cytokine TRAIL.
They show that endothelial cells induce and maintain breast The lung has a uniquely high expression of the TGF-β
cancer DTC quiescence. They identified thrombospondin-1 superfamily member BMP4, which promotes the dormancy
as an endothelium-derived tumor suppressor, which dimin- of DTCs within this tissue. Gao et al. showed that in breast
ishes near sprouting neovasculature, suggesting that tumors cancer mouse models, DTCs can overcome the suppressive
94 H. Kim et al.

microenvironment of the lung by expressing Coco, a secreted mancy regulated by matrix stiffness [53–55]. These models
antagonist of TGF-β ligands [47]. Coco binds to BMP4, pre- have proven successful in allowing dormancy-incapable
venting the activation of cognizant receptors, which triggers cells to proliferate and maintaining dormancy-capable cells
relapse of dormant cells. quiescent. However, still much in vivo research needs to be
done to validate these in vitro findings, and to better under-
stand how ECM stiffness is correlated to maintain dormancy
9.4.3 Lysyl Oxidase and ECM Stiffening and their proliferative growth after several years.

Lysyl oxidase (LOX) is a copper-dependent amine oxidase,

which catalyzes the covalent cross-linking of collagen and 9.5 Conclusions
elastin, thereby increasing matrix stiffness [48]. Previous
reports show that a fibrotic microenvironment and remodel- The role of the ECM in regulating tumor growth and metas-
ing of ECM accounts for the switch of dormant tumor cells tasis has been an active field for decades, but only recently
to metastatic growth [4, 16] (Fig. 9.1c). LOX family mem- have groups begun to establish a link between tissue ECM
bers enable a switch from dormancy to proliferation by and dormancy/relapse. Because this field is so new, several
establishing a fibrotic environment [4, 49]. Increased expres- questions remain:
sion of LOX has been found in several cancer types (breast,
prostate, colorectal, lung, bladder) [4, 48, 50], and hypoxia-­
• What triggers ECM remodeling that initiates tumor
inducible factor-1 causes the expression of several members
cell growth and relapse at a distant site?
of the LOX family in breast cancer cells, including LOX and
• Does the ECM at certain tissues help maintain dor-
Lysyl oxidase Like 2 (LOXL2).
mancy while other ECMs encourage relapse?
LOXL2 facilitates an epithelial to mesenchymal transi-
• Do certain preexisting conditions that affect ECM
tion (EMT) and induces the outgrowth of dormant breast
composition such as fibrosis and osteoporosis facil-
cancer cells [51]. In this study, dormant MCF-7 cells express-
itate dormancy or relapse in different patients?
ing LOXL2 (MCF-7-LOXL2#12) underwent EMT, and
• How can cell–ECM interactions be targeted to pre-
downregulation of LOXL2 expression in MCF-7-LOXL2#12
vent escape from dormancy in patients?
cells restored their epithelial phenotype. In vitro, a
• Can ECM-targeting drug regimens be combined
mammosphere assay confirmed that EMT induced by
­
with chemotherapy or hormone receptor-­targeting
LOXL2 expression is associated with the cancer stem-like
drugs to be used as long-term maintenance thera-
phenotype. Here, MCF-7-LOXL2#12 cells exhibited a sig-
pies for patients to prevent relapse?
nificant increase in the sphere-forming capacity whereas
control cells remained as single cells, suggesting MCF-7-
LOXL2#12 cells overcome the proliferation barrier. In vivo,
tail vein injection of nude mice showed MCF-7-LOXL2#12 References
cells colonizing the lung transitioned from dormancy to met-
astatic outgrowth, whereas control MCF-7 cells remained 1. Rosen LS, Gordon D, Tchekmedyian NS, Yanagihara R, Hirsh V,
Krzakowski M, Pawlicki M, De Souza P, Zheng M, Urbanowitz
dormant. MCF-7-LOXL2#12 cells on the surface of the G, Reitsma D, Seaman J. Long-term efficacy and safety of zole-
lungs were imaged after 40 days, which revealed only 20% dronic acid in the treatment of skeletal metastases in patients with
of the lung lesions remained as single tumor cells and 80% nonsmall cell lung carcinoma and other solid tumors: a random-
formed multi-cellular lesions. All metastatic lesions from ized, phase III, double-blind, placebo-controlled trial. Cancer.
2004;100(12):2613–21. Epub 2004/06/16. https://doi.org/10.1002/
control cells persisted as single quiescent cells (100%). cncr.20308.
A stiff mechanical environment induces B16 melanoma 2. Saad F, Gleason DM, Murray R, Tchekmedyian S, Venner P,
tumor-repopulating cells (TRC) to enter dormancy in vivo Lacombe L, Chin JL, Vinholes JJ, Goas JA, Zheng M. Zoledronic
and in vitro [52]. In vitro experiments on fibrin gels of three acid prostate cancer study G. long-term efficacy of zoledronic
acid for the prevention of skeletal complications in patients with
different stiffness values (90, 450, and 1050 Pa) have shown metastatic hormone-refractory prostate cancer. J Natl Cancer Inst.
that growth of TRCs into spheroids is fastest in more compli- 2004;96(11):879–82. Epub 2004/06/03. https://doi.org/10.1093/
ant environments, and the most compliant fibrin environ- jnci/djh141.
ments also maximized proliferation in vivo in subcutaneous 3. Dittmer J. Mechanisms governing metastatic dormancy in
breast cancer. Semin Cancer Biol. 2017;44:72–82. https://doi.
tumors. org/10.1016/j.semcancer.2017.03.006.
Currently available in vivo dormancy models still have 4. Barkan D, El Touny LH, Michalowski AM, Smith JA, Chu I,
some limitations, which if overcome could lead to quicker Davis AS, Webster JD, Hoover S, Simpson RM, Gauldie J, Green
advancement in the field. Bioengineers are developing 3D JE. Metastatic growth from dormant cells induced by a col-I-­
in vitro systems to study dormancy, often focusing on dor-
9 Tumor Dormancy and Relapse Regulated by the Extracellular Matrix 95

enriched fibrotic environment. Cancer Res. 2010;70(14):5706–16. 19. Yates LR, Knappskog S, Wedge D, Farmery JHR, Gonzalez S,
https://doi.org/10.1158/0008-­5472.CAN-­09-­2356. Martincorena I, Alexandrov LB, Van Loo P, Haugland HK, Lilleng
5. Lu X, Mu E, Wei Y, Riethdorf S, Yang Q, Yuan M, Yan J, Hua Y, Tiede PK, Gundem G, Gerstung M, Pappaemmanuil E, Gazinska P,
BJ, Haffty BG, Pantel K, Massague J, Kang Y. VCAM-1 promotes Bhosle SG, Jones D, Raine K, Mudie L, Latimer C, Sawyer E,
osteolytic expansion of indolent bone micrometastasis of breast can- Desmedt C, Sotiriou C, Stratton MR, Sieuwerts AM, Lynch AG,
cer by engaging alpha4beta1-positive osteoclast progenitors. Cancer Martens JW, Richardson AL, Tutt A, Lonning PE, Campbell
Cell. 2011;20(6):701–14. https://doi.org/10.1016/j.ccr.2011.11.002. PJ. Genomic evolution of breast cancer metastasis and relapse.
6. Shiozawa Y, Pedersen EA, Patel LR, Ziegler AM, Havens AM, Jung Cancer Cell. 2017;32(2):169–84 e7. https://doi.org/10.1016/j.
Y, Wang J, Zalucha S, Loberg RD, Pienta KJ, Taichman RS. GAS6/ ccell.2017.07.005.
AXL axis regulates prostate cancer invasion, proliferation, and 20. Bragado P, Estrada Y, Parikh F, Krause S, Capobianco C, Farina
survival in the bone marrow niche. Neoplasia (New York, NY). HG, Schewe DM, Aguirre-Ghiso JA. TGF-beta2 dictates dissemi-
2010;12(2):116–27. nated tumour cell fate in target organs through TGF-beta-RIII and
7. Ghajar CM, Peinado H, Mori H, Matei IR, Evason KJ, Brazier H, p38alpha/beta signalling. Nat Cell Biol. 2013;15(11):1351–61.
Almeida D, Koller A, Hajjar KA, Stainier DY, Chen EI, Lyden D, https://doi.org/10.1038/ncb2861.
Bissell MJ. The perivascular niche regulates breast tumour dor- 21. Kaplan RN, Riba RD, Zacharoulis S, Bramley AH, Vincent L, Costa
mancy. Nat Cell Biol. 2013;15(7):807–17. https://doi.org/10.1038/ C, MacDonald DD, Jin DK, Shido K, Kerns SA, Zhu Z, Hicklin
ncb2767. D, Wu Y, Port JL, Altorki N, Port ER, Ruggero D, Shmelkov SV,
8. Malladi S, Macalinao DG, Jin X, He L, Basnet H, Zou Y, de Jensen KK, Rafii S, Lyden D. VEGFR1-positive haematopoietic
Stanchina E, Massague J. Metastatic latency and immune evasion bone marrow progenitors initiate the pre-metastatic niche. Nature.
through autocrine inhibition of WNT. Cell. 2016;165(1):45–60. 2005;438(7069):820–7. https://doi.org/10.1038/nature04186.
https://doi.org/10.1016/j.cell.2016.02.025. 22. Malanchi I, Santamaria-Martinez A, Susanto E, Peng H, Lehr
9. Lawson MA, McDonald MM, Kovacic N, Hua Khoo W, Terry HA, Delaloye JF, Huelsken J. Interactions between cancer stem
RL, Down J, Kaplan W, Paton-Hough J, Fellows C, Pettitt JA, cells and their niche govern metastatic colonization. Nature.
Neil Dear T, Van Valckenborgh E, Baldock PA, Rogers MJ, Eaton 2012;481(7379):85–9. https://doi.org/10.1038/nature10694.
CL, Vanderkerken K, Pettit AR, Quinn JM, Zannettino AC, Phan 23. Oskarsson T, Acharyya S, Zhang XH, Vanharanta S, Tavazoie SF,
TG, Croucher PI. Osteoclasts control reactivation of dormant Morris PG, Downey RJ, Manova-Todorova K, Brogi E, Massague
myeloma cells by remodelling the endosteal niche. Nat Commun. J. Breast cancer cells produce tenascin C as a metastatic niche com-
2015;6:8983. https://doi.org/10.1038/ncomms9983. ponent to colonize the lungs. Nat Med. 2011;17(7):867–74. https://
10. Boyerinas BZM, Yesilkanal AL, Price TT, Hyjek EM, Sipkins doi.org/10.1038/nm.2379.
DA. Adhesion to osteopontin in the bone marrow niche regulates 24. Shoulders MD, Raines RT. Collagen structure and stability. Annu
lymphoblastic leukemia cell dormancy. Blood. 2013;121(24):4821– Rev Biochem. 2009;78:929–58. https://doi.org/10.1146/annurev.
31. https://doi.org/10.1182/blood-­2012-­12-­. biochem.77.032207.120833.
11. Price TT, Sivan A, Warner MJ, Cheng R, Lee CH, Olivere L, 25. Xu S, Xu H, Wang W, Li S, Li H, Li T, Zhang W, Yu X, Liu
Comatas K, Magnani J, Lyerly HK, Cheng Q, McCall CM, Sipkins L. The role of collagen in cancer: from bench to bedside. J Transl
DA. Dormant breast cancer micrometastases reside in specific bone Med. 2019;17(1):309. Epub 2019/09/16. https://doi.org/10.1186/
marrow niches that regulate their transit to and from bone. Sci s12967-­019-­2058-­1.
Transl Med. 2016;8(340):340ra73. 26. Samuel MS, Lopez JI, McGhee EJ, Croft DR, Strachan D, Timpson
12. Hassell BA, Goyal G, Lee E, Sontheimer-Phelps A, Levy O, Chen P, Munro J, Schroder E, Zhou J, Brunton VG, Barker N, Clevers
CS, Ingber DE. Human organ chip models recapitulate orthotopic H, Sansom OJ, Anderson KI, Weaver VM, Olson MF. Actomyosin-­
lung cancer growth, therapeutic responses, and tumor dormancy mediated cellular tension drives increased tissue stiffness and
in vitro. Cell Rep. 2017;21(2):508–16. https://doi.org/10.1016/j. beta-catenin activation to induce epidermal hyperplasia and tumor
celrep.2017.09.043. growth. Cancer Cell. 2011;19(6):776–91. Epub 2011/06/15. https://
13. Pickup MW, Mouw JK, Weaver VM. The extracellular matrix mod- doi.org/10.1016/j.ccr.2011.05.008.
ulates the hallmarks of cancer. EMBO Rep. 2014;15(12):1243–53. 27. Ohlund D, Franklin O, Lundberg E, Lundin C, Sund M. Type IV
https://doi.org/10.15252/embr.201439246. collagen stimulates pancreatic cancer cell proliferation, migration,
14. Jinka R, Kapoor R, Sistla PG, Raj TA, Pande G. Alterations in and inhibits apoptosis through an autocrine loop. BMC Cancer.
cell-extracellular matrix interactions during progression of cancers. 2013;13:154. https://doi.org/10.1186/1471-­2407-­13-­154.
y3er. 2012;2012:219196. https://doi.org/10.1155/2012/219196. 28. Miller FR, McEachern D, Miller BE. Growth regulation of mouse
15. Sosa MS, Bragado P, Aguirre-Ghiso JA. Mechanisms of dissemi- mammary tumor cells in collagen gel cultures by diffusible factors
nated cancer cell dormancy: an awakening field. Nat Rev Cancer. produced by normal mammary gland epithelium and stromal fibro-
2014;14(9):611–22. https://doi.org/10.1038/nrc3793. blasts. Cancer Res. 1989;49(21):6091–7. Epub 1989/11/01
16. Barkan D, Green JE, Chambers AF. Extracellular matrix: a gate- 29. Rak JW, McEachern D, Miller FR. Sequential alteration of peanut
keeper in the transition from dormancy to metastatic growth. agglutinin binding-glycoprotein expression during progression of
Eur J Cancer. 2010;46(7):1181–8. Epub 2010/03/23. https://doi. murine mammary neoplasia. Br J Cancer. 1992;65(5):641–8. Epub
org/10.1016/j.ejca.2010.02.027. 1992/05/01. https://doi.org/10.1038/bjc.1992.138.
17. Gao H, Chakraborty G, Lee-Lim AP, Mavrakis KJ, Wendel HG, 30. Ossowski L, Aguirre-Ghiso JA. Dormancy of metastatic melanoma.
Giancotti FG. Forward genetic screens in mice uncover media- Pigment Cell Melanoma Res. 2010;23(1):41–56. Epub 2009/10/22.
tors and suppressors of metastatic reactivation. Proc Natl Acad https://doi.org/10.1111/j.1755-­148X.2009.00647.x.
Sci U S A. 2014;111(46):16,532–7. https://doi.org/10.1073/ 31. Wall SJ, Zhong ZD, DeClerck YA. The cyclin-dependent kinase
pnas.1403234111. inhibitors p15INK4B and p21CIP1 are critical regulators of
18. Gao H, Chakraborty G, Zhang Z, Akalay I, Gadiya M, Gao fibrillar collagen-induced tumor cell cycle arrest. J Biol Chem.
Y, Sinha S, Hu J, Jiang C, Akram M, Brogi E, Leitinger B, 2007;282(33):24471–6. Epub 2007/06/08. https://doi.org/10.1074/
Giancotti FG. Multi-organ site metastatic reactivation mediated jbc.M702697200.
by non-canonical discoidin domain receptor 1 signaling. Cell. 32. Wall SJ, Werner E, Werb Z, DeClerck YA. Discoidin domain recep-
2016;166(1):47–62. https://doi.org/10.1016/j.cell.2016.06.009. tor 2 mediates tumor cell cycle arrest induced by fibrillar collagen. J
96 H. Kim et al.

Biol Chem. 2005;280(48):40187–94. Epub 2005/09/28. https://doi. Epub 2017/05/17. https://doi.org/10.1158/0008-­5472.CAN-­

org/10.1074/jbc.M508226200. 17-­0068.
33. Singh P, Carraher C, Schwarzbauer JE. Assembly of fibro- 45. Albrengues J, Shields MA, Ng D, Park CG, Ambrico A, Poindexter
nectin extracellular matrix. Annu Rev Cell Dev Biol. ME, Upadhyay P, Uyeminami DL, Pommier A, Kuttner V, Bruzas
2010;26:397–419. Epub 2010/08/10. https://doi.org/10.1146/ E, Maiorino L, Bautista C, Carmona EM, Gimotty PA, Fearon DT,
annurev-­cellbio-­100109-­104020. Chang K, Lyons SK, Pinkerton KE, Trotman LC, Goldberg MS,
34. Sottile J, Hocking DC, Langenbach KJ. Fibronectin polymeriza- Yeh JT, Egeblad M. Neutrophil extracellular traps produced dur-
tion stimulates cell growth by RGD-dependent and -independent ing inflammation awaken dormant cancer cells in mice. Science.
mechanisms. J Cell Sci. 2000;113(Pt 23):4287–99. 2018;361(6409) Epub 2018/09/29 https://doi.org/10.1126/science.
35. Aguirre-Ghiso JA, Liu D, Mignatti A, Kovalski K, aao4227.
Ossowski L. Urokinase receptor and fibronectin regulate the 46. Chen Q, Zhang XH, Massague J. Macrophage binding to recep-
ERK(MAPK) to p38(MAPK) activity ratios that determine car- tor VCAM-1 transmits survival signals in breast cancer cells
cinoma cell proliferation or dormancy in vivo. Mol Biol Cell. that invade the lungs. Cancer Cell. 2011;20(4):538–49. Epub
2001;12(4):863–79. Epub 2001/04/11. https://doi.org/10.1091/ 2011/10/22. https://doi.org/10.1016/j.ccr.2011.08.025.
mbc.12.4.863. 47. Gao H, Chakraborty G, Lee-Lim AP, Mo Q, Decker M, Vonica A,
36. Park SY, Nam JS. The force awakens: metastatic dormant cancer Shen R, Brogi E, Brivanlou AH, Giancotti FG. The BMP inhibitor
cells. Exp Mol Med. 2020;52(4):569–81. Epub 2020/04/18. https:// Coco reactivates breast cancer cells at lung metastatic sites. Cell.
doi.org/10.1038/s12276-­020-­0423-­z. 2012;150(4):764–79. https://doi.org/10.1016/j.cell.2012.06.035.
37. Aguirre-Ghiso JA, Estrada Y, Liu D, Ossowski L. ERK(MAPK) 48. Barker HE, Cox TR, Erler JT. The rationale for targeting the
activity as a determinant of tumor growth and dormancy; regu- LOX family in cancer. Nat Rev Cancer. 2012;12(8):540–52. Epub
lation by p38(SAPK). Cancer Res. 2003;63(7):1684–95. Epub 2012/07/20. https://doi.org/10.1038/nrc3319.
2003/04/03 49. Barry-Hamilton V, Spangler R, Marshall D, McCauley S,
38. Barney LE, Hall CL, Schwartz AD, Parks AN, Sparages C, Galarza Rodriguez HM, Oyasu M, Mikels A, Vaysberg M, Ghermazien H,
S, Platt MO, Mercurio AM, Peyton SR. Tumor cell-organized fibro- Wai C, Garcia CA, Velayo AC, Jorgensen B, Biermann D, Tsai D,
nectin maintenance of a dormant breast cancer population. Sci Adv. Green J, Zaffryar-Eilot S, Holzer A, Ogg S, Thai D, Neufeld G,
2020;6(11):eaaz4157. Epub 2020/03/21. https://doi.org/10.1126/ Van Vlasselaer P, Smith V. Allosteric inhibition of lysyl oxidase-­
sciadv.aaz4157. like-­2 impedes the development of a pathologic microenvironment.
39. Montagner M, Bhome R, Hooper S, Chakravarty P, Qin X, Sufi Nat Med. 2010;16(9):1009–17. Epub 2010/09/08. https://doi.
J, Bhargava A, Ratcliffe CDH, Naito Y, Pocaterra A, Tape CJ, org/10.1038/nm.2208.
Sahai E. Crosstalk with lung epithelial cells regulates Sfrp2-­ 50. Baker AM, Bird D, Lang G, Cox TR, Erler JT. Lysyl oxidase
mediated latency in breast cancer dissemination. Nat Cell Biol. enzymatic function increases stiffness to drive colorectal cancer
2020;22(3):289–96. Epub 2020/02/26. https://doi.org/10.1038/ progression through FAK. Oncogene. 2013;32(14):1863–8. Epub
s41556-­020-­0474-­3. 2012/05/30. https://doi.org/10.1038/onc.2012.202.
40. Price TT, Burness ML, Sivan A, Warner MJ, Cheng R, Lee CH, 51. Weidenfeld K, Schif-Zuck S, Abu-Tayeh H, Kang K, Kessler O,
Olivere L, Comatas K, Magnani J, Kim Lyerly H, Cheng Q, McCall Weissmann M, Neufeld G, Barkan D. Dormant tumor cells express-
CM, Sipkins DA. Dormant breast cancer micrometastases reside in ing LOXL2 acquire a stem-like phenotype mediating their transi-
specific bone marrow niches that regulate their transit to and from tion to proliferative growth. Oncotarget. 2016;7(44):71,362–77.
bone. Sci Transl Med. 2016;8(340):340ra73. Epub 2016/05/27. Epub 2016/09/23. https://doi.org/10.18632/oncotarget.12109.
https://doi.org/10.1126/scitranslmed.aad4059. 52. Liu Y, Lv J, Liang X, Yin X, Zhang L, Chen D, Jin X, Fiskesund R,
41. Boyerinas B, Zafrir M, Yesilkanal AE, Price TT, Hyjek EM, Tang K, Ma J, Zhang H, Dong W, Mo S, Zhang T, Cheng F, Zhou
Sipkins DA. Adhesion to osteopontin in the bone marrow Y, Xie J, Wang N, Huang B. Fibrin stiffness mediates dormancy
niche regulates lymphoblastic leukemia cell dormancy. Blood. of tumor-repopulating cells via a Cdc42-driven Tet2 epigenetic
2013;121(24):4821–31. Epub 2013/04/17. https://doi.org/10.1182/ program. Cancer Res. 2018;78(14):3926–37. Epub 2018/05/17.
blood-­2012-­12-­475483. https://doi.org/10.1158/0008-­5472.CAN-­17-­3719.
42. Carlson P, Dasgupta A, Grzelak CA, Kim J, Barrett A, Coleman 53. Fang JY, Tan SJ, Wu YC, Yang Z, Hoang BX, Han B. From com-
IM, Shor RE, Goddard ET, Dai J, Schweitzer EM, Lim AR, Crist petency to dormancy: a 3D model to study cancer cells and drug
SB, Cheresh DA, Nelson PS, Hansen KC, Ghajar CM. Targeting responsiveness. J Transl Med. 2016;14:38. Epub 2016/02/06.
the perivascular niche sensitizes disseminated tumour cells to che- https://doi.org/10.1186/s12967-­016-­0798-­8.
motherapy. Nat Cell Biol. 2019;21(2):238–50. Epub 2019/01/22. 54. Preciado JA, Reátegui E, Azarin SM, Lou E, Aksan
https://doi.org/10.1038/s41556-­018-­0267-­0. A. Immobilization platform to induce quiescence in dormancy-­
43. Magnus N, Garnier D, Meehan B, McGraw S, Lee TH, Caron M, capable cancer cells. Technology. 2017;05(03):129–38. https://doi.
Bourque G, Milsom C, Jabado N, Trasler J, Pawlinski R, Mackman org/10.1142/S2339547817500078.
N, Rak J. Tissue factor expression provokes escape from tumor dor- 55. Pradhan S, Hassani I, Seeto WJ, Lipke EA. PEG-fibrinogen hydro-
mancy and leads to genomic alterations. Proc Natl Acad Sci U S gels for three-dimensional breast cancer cell culture. J Biomed
A. 2014;111(9):3544–9. Epub 2014/02/13. https://doi.org/10.1073/ Mater Res A. 2017;105(1):236–52. https://doi.org/10.1002/
pnas.1314118111. jbm.a.35899.
44. Manjili MH. Tumor dormancy and relapse: from a natural byprod-
uct of evolution to a disease state. Cancer Res. 2017;77(10):2564–9.
 Therapy-Induced Dormancy
and Residual Disease 10
Ashley V. DiMarco, Nina Marie G. Garcia,
and James V. Alvarez

Abstract several years following adjuvant therapy are independent

In many common epithelial cancers, such as breast and markers for recurrence risk [4–6]. Furthermore, the extent of
prostate cancer, the majority of deaths are caused by the residual cancer present locally in the breast following neoad-
survival and eventual recurrence of residual tumor cells juvant therapy is strongly predictive of recurrence [7].
after therapy. In many cases, residual tumors can survive Together, these clinical data suggest that the survival and
for years or even decades prior to recurrence, suggesting persistence of residual cancer cells following therapy is a
that these tumors are in a dormant state. Because recur- major obstacle to curing breast cancer. Here, we review the
rent tumors are often incurable, it is critical to understand clinical significance of residual dormant cells and then dis-
how cancer cells survive treatment, persist as residual cuss some of the properties of dormant tumors, with a focus
cells during a prolonged dormant period, and eventually on dormant cell metabolism and the dormant tumor
recur. In this chapter, we review the clinical evidence sug- microenvironment.
gesting that dormant residual tumors are an important res-
ervoir for tumor recurrence, and then discuss some of the
molecular and cellular features of dormant residual 10.2  isseminated Tumor Cells (DTCs)
D
tumors. Predict Risk of Recurrence

In a seminal paper published over 20 years ago, Klaus Pantel

and colleagues examined the bone marrow of breast cancer
10.1 Introduction patients for the presence of cancer cells [8]. Importantly,
these patients did not have gross evidence of metastatic dis-
Tumor recurrence following potentially curative therapy is ease, and so their cancers were classified as Stages I–III. They
the leading cause of death in many types of cancer, including found that, despite the absence of clinically evident metasta-
some of the most common epithelial malignancies like breast sis, nearly half of these patients had detectable cancer cells in
cancer, prostate cancer, and melanoma [1]. In many of these their bone marrow. These cells were referred to as dissemi-
cancer types, tumors can recur after a prolonged period. For nated tumor cells, or DTCs. These DTCs were present both
instance, most breast cancer deaths result from tumor recur- as isolated cancer cells, as well as in DTC clusters. While the
rence following therapy, and tumors can recur as long as frequency of DTCs was highest in Stage III cancers, even a
20 years after treatment [1]. This suggests that residual tumor fraction of patients with early-stage, Stage I breast cancer,
cells can persist in a dormant state for decades prior to had detectable DTCs in their bone marrow. To assess the
reemerging as a recurrent tumor [2] [3]. Consistent with this, clinical significance of these DTCs, the authors examined
disseminated tumor cells (DTCs) are present in the bone their association with prognosis. As described above, meta-
marrow of up to 40% of breast cancer patients at diagnosis, static recurrence is the main cause of death from breast can-
and the number of DTCs at diagnosis and their persistence cer. This group found that the presence of DTCs in the bone
marrow greatly increased the risk of metastasis and death
from breast cancer. This indicates that DTCs are clinically
A. V. DiMarco · N. M. G. Garcia · J. V. Alvarez (*)
Department of Pharmacology and Cancer Biology, significant, and suggests that these cells may serve as a res-
Duke University School of Medicine, Durham, NC, USA ervoir for metastatic recurrence.
e-mail: [email protected]; [email protected];
[email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 97

S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_10
98 A. V. DiMarco et al.

These findings have been followed up on and validated by shown to be associated with a reduced risk of recurrence
a number of other groups [9–13]. For instance, a pooled anal- [19–22]. Similarly, using a different measure, Residual
ysis of patients from several academic hospitals confirmed Cancer Burden, it was shown that the presence of residual
these results, and extended them in an important way [9]. tumor cells following neoadjuvant chemotherapy is associ-
This study found that, in addition to the presence of DTCs at ated with increased risk of relapse [7]. It is important to keep
diagnosis, the persistence of DTCs for several years after in mind that the response of the primary tumor to therapy is
therapy was also predictive of eventual recurrence and sur- actually a surrogate for what is more important, namely the
vival [9]. This provides further evidence that DTCs serve as response of any disseminated tumor cells or micrometastases
a functionally important reservoir for disease relapse. to therapy, since the primary tumor will be surgically excised
Evidence that DTCs are dormant has come from several following neoadjuvant therapy. These studies make clear that
routes, both direct and indirect. Indirect evidence that DTCs the survival of residual cancer cells following therapy is
are dormant include the clinical observations that the number strongly associated with eventual recurrence.
of DTCs can be quite stable, even in patients who do not
develop overt metastases [14–16]. Additionally, there are
several case studies where dormant DTCs have reactivated 10.4 Therapy-Induced Dormancy
upon organ transplant, giving rise to metastases in the organ
recipient [14]. Direct analysis of the proliferation marker 10.4.1 Chemotherapy
Ki67 revealed that the majority of DTCs in breast cancer are
negative for Ki67 [17, 18]. Additionally, the dormancy A number of anticancer therapies can induce tumor cells to
marker NR2F1 was shown to mark the majority of DTCs, stop proliferating and enter into a dormant or quiescent state.
and NR2F1-positive DTCs were associated with good out- Pioneering work from Julio Aguirre-Ghiso and colleagues
come in breast cancer patients with DTCs [17]. These data found that cancer cell dormancy is dictated by the ratio
indicate that DTCs frequently exist in a dormant or slowly between activation of the ERK and p38 pathways. A high
proliferative state. ERK:p38 ratio promotes proliferation, while a low ERK:p38
Taken together, these studies demonstrate that DTCs rep- ratio induces dormancy [23]. Because many cellular stresses
resent a clinically relevant population of minimal residual can activate p38, this provides one mechanism of stress-­
disease, and provide strong clinical evidence that the survival induced dormancy. For instance, chemotherapy, by inducing
and persistence of residual tumor cells predicts relapse risk. p38 activation, can induce cancer cell dormancy [24]. p38
activation, in turn, can lead to induction of the unfolded pro-
tein response (UPR) and activation of the stress-responsive
10.3  esidual Disease Following
R transcription factor ATF6 [25]. This stress-response network
Neoadjuvant Therapy Predicts Risk therefore links therapy response to cell cycle arrest, dor-
of Recurrence mancy, and survival.

A separate line of clinical investigation has provided evi-

dence that the survival of cancer cells in response to therapy 10.4.2 Targeted Therapy
predicts eventual relapse. These studies focused on patients
receiving neoadjuvant therapy—that is, adjuvant therapy Targeted therapy against oncogenes can also induce dor-
prior to surgical removal of the tumor. The specific therapy mancy. Many oncogenes induce cell proliferation. For
given to patients in the neoadjuvant setting varies depending instance, receptor tyrosine kinases and growth factor recep-
upon the tumor characteristics, but most frequently includes tors signal through the Ras-MAPK and PI3K-Akt pathways
chemotherapy, such as Adriamycin, cyclophosphamide, and to induce both proliferation and survival. Many of the signal-
paclitaxel. For Her2-amplified tumors, neoadjuvant therapy ing proteins in these pathways are protooncogenes, and when
might include the anti-Her2 therapies such as Trastuzumab. mutated or overexpressed can induce hyperactivation of
Notwithstanding the specific therapies given, the benefit of these signaling pathways. Consequently, it makes sense that
neoadjuvant therapy is that it allows for an assessment of drugs targeting these oncogenes will induce both cell death
how the tumor responds to therapy, and understanding how and a loss of proliferation. In fact, the efficacy of many tar-
this response is related to disease outcome. geted therapies, including those against receptor tyrosine
A number of studies and meta-analyses have examined kinases such as EGFR and Her2, or targeting intracellular
the relationship between response to neoadjuvant therapy kinases like Raf, PI3K, or Akt, is likely due to the combina-
and various clinical endpoints, including local recurrence, tion of cell death and proliferation arrest. However, in many
metastatic recurrence, and survival. For instance, pathologic cases some cells can survive these therapies, but remain in a
complete response (pCR) in the breast has repeatedly been non-proliferative state. For instance, non-small cell lung can-
10 Therapy-Induced Dormancy and Residual Disease 99

cer (NSCLC) cells with mutations in the epidermal growth drugs. They can persist in this intermediate resistance state
factor receptor (EGFR) respond to EGFR inhibitors like gefi- until they acquire additional changes, most frequently
tinib and erlotinib. However, some cells can survive inhibitor genetic mutations, that render them fully resistant [26]. A
treatment and persist in a slowly proliferative state resem- number of studies have shown that the drug-tolerant state is
bling dormancy [26–28]. As described in more detail below, mediated by epigenetic changes in tumor cells [27]. For
these surviving cells have been referred to as drug-tolerant instance, in drug-tolerant persistent NSCLC cells following
persister cells. A similar scenario was observed in basal cell EGFR inhibition, the histone demethylase KDM5A leads to
carcinoma cells treated with a Smoothened inhibitor, where loss of H3K4 methylation at cell-cycle-related genes to pre-
tumors regress but a population of dormant cells seeds vent their expression. Similarly, the DNA demethylase TET2
relapse [29]. promotes dormancy in colon cancer by regulating the expres-
A number of insights on the mechanism of therapy-­ sion of proliferation genes [37].
induced dormancy have come from conditional genetically
engineered mouse models. In these models, doxycycline
(dox) administration induces expression of an oncogene— 10.5 Dormant Tumor Cell Metabolism
such as Her2/neu, Wnt1, or Myc—leading to mammary
tumor formation. Subsequent withdrawal of dox induces The unique metabolic features of cancer cells were recog-
oncogene downregulation and complete tumor regression, nized nearly 100 years ago, when Warburg noted that cancer
mimicking response to targeted therapies. However, a popu- cells exhibit elevated levels of glycolysis, producing lactate
lation of residual cells survives oncogene downregulation even under aerobic conditions. The past decade has seen a
and resides in a dormant, non-proliferative state in the mam- renewed interest in understanding the metabolic require-
mary gland, and these cells eventually reinitiate proliferation ments of tumor cell proliferation. A current model suggests
to form a recurrent tumor. Importantly, the residual dormant that tumor cells must adopt a glycolytic phenotype to pro-
cells that survive oncogene downregulation can be isolated duce the necessary precursors for anabolic growth to support
and characterized [30]. These models recapitulate residual their rapid proliferation [38]. Indeed, the requirements of
disease and recurrence as it occurs in breast cancer patients metabolic pathways to support proliferation, and the regula-
in several key respects. First, analogous to residual DTCs in tion of these pathways by tumor suppressors and oncogenes,
human patients, the residual cells that survive in these mod- have been well established. The metabolic demands of dor-
els are non-proliferative but remain viable [31]. Second, the mant cells are likely to be quite different from those of pro-
dormant period in these models is long (6–9 months) relative liferating cells: they do not need to provide precursors for
to the lifetime of the mouse, similar to what is observed in anabolic growth or proliferation, such as nucleotides to syn-
human breast cancer. Third, recurrent tumors arising in these thesize DNA or lipids to synthesize new cell membranes. In
models do not express the initiating oncogene, reminiscent contrast to the extensive literature on the metabolism of rap-
of the finding that HER2-amplified tumors are often HER2-­ idly proliferating tumor cells, virtually nothing is known
negative upon relapse [32]. about the metabolic requirements of dormant tumor cells, or
Using these models, it was found that loss of Her2 [33], how metabolic pathways are regulated during dormancy.
Myc [34], Wnt1 [35], and Ras [36] can induce apoptosis and Similarly, very little is known about the metabolism of recur-
tumor regression. The residual cells that survive oncogene rent tumors that form when dormant tumor cells reenter the
downregulation are non-proliferative, and are thought to cell cycle or reinitiate rapid proliferation. Considering the
reside in the G0 phase of the cell cycle. Thus, targeted thera- intimate relationship between cancer cell proliferation and
pies directed against oncogenes can induce dormancy in sur- metabolism, it is likely that the metabolism of cancer cells
viving residual tumor cells, and these cells serve as a reservoir may change dramatically during these transitions between
for eventual relapse. proliferation and dormancy. However, it is not known
whether sustained activity of particular metabolic pathways
is functionally important in regulating tumor dormancy and
10.4.3 Drug-Tolerant Persister Cells recurrence. Given the clinical importance of these cells, and
the fact that they can persist in cancer patients for decades,
Cells that can survive chemotherapy or targeted therapy are understanding what nutrient sources they depend upon and
analogous in some ways to so-called “drug-tolerant persister how their metabolic pathways are wired may be critical in
cells.” This concept, originally developed in relation to bac- therapeutically targeting them.
teria that survive antibiotic treatment, posits that some cells Several recent studies have begun to address these issues.
can survive therapy and persist in a quasi-resistant state. For instance, several studies showed that inhibition of diverse
These cells (or bacteria) are not fully resistant to anticancer oncogenes—including Her2, Ras, Myc, and Wnt—causes an
therapy (or antibiotics), but are viable in the presence of immediate drop in glucose uptake [36, 39, 40]. In response to
100 A. V. DiMarco et al.

decreased glucose metabolism, cells that survive the loss of both targeted therapy and chemotherapy [30]. Loss of Par-4
oncogenic Her2 [40] and KRAS [36] signaling upregulate expression through epigenetic silencing promotes the long-­
fatty acid oxidation (FAO) pathways ([36, 40, 41]. term survival of dormant cells following Her2 inhibition in
Additionally, both decreased glucose metabolism and breast cancer cells [30, 53, 54]. Elevated Wnt signaling is
increased FAO following oncogene inhibition results in responsible for dormant cell survival in basal cell carcinoma
increased levels of reactive oxygen species (ROS) [36, 40– cells treated with a Hedgehog inhibitor [29]. In EGFR-­
42], which can impair tumor cell viability [41, 43, 44]. mutant NSCLC, the transcription factor YAP promotes cell
Similarly, inhibitors targeting the critical cell cycle regulator survival by repressing the pro-apoptotic protein BMF [28].
cyclin D3-CDK6 leads to altered cell metabolism and induc- In all of these instances, targeting the pro-survival pathway
tion of ROS [45]. relied on by dormant cells is effective at eliminating these
These results suggest that tumor cells that survive therapy dormant cells. This illustrates the principle that preventing
and eventually develop resistance must adapt to overcome recurrence following treatment of the primary tumor may
therapy-induced metabolic stress. A number of pathways require a two-pronged strategy. In the first arm, the primary
have been identified that promote metabolic adaptations dur- tumor is treated by inhibiting the driving oncogene, for
ing tumor dormancy and recurrence remain. In breast tumors instance Her2 in breast cancer or EGFR in lung cancer.
following loss of Her2 signaling, activation of the antioxi- However, a second therapy is needed to eliminate the dor-
dant transcription factor NRF2 is an adaptive response to mant cells that survive and persist following the primary
elevated ROS [41]. Interestingly, NRF2 remains activated in treatment.
dormant tumors following Her2 inhibition, and NRF2 activa-
tion is both necessary and sufficient for these dormant tumor
cells to reinitiate proliferation to form recurrent tumors [41]. 10.7 Epithelial-to-Mesenchymal (EMT)
Mechanistically, NRF2 functions in dormant tumors to pro- Transition in Dormancy
mote nucleotide synthesis and maintain redox homeostasis.
Cells that survive therapy have metabolic dependencies Several lines of evidence have shown that dormant residual
that are unique from the untreated tumor cells. For example, tumor cells frequently undergo an epithelial-to-­mesenchymal
AMPK is activated by diverse therapies [41, 46] consistent transition (EMT), and that EMT contributes to tumor recur-
with the notion that tumor cells experience bioenergetic rence following therapy. A clinical study that tracked the
stress resulting from therapy. AMPK activation suppresses evolution of EGFR-mutant lung tumors as they acquired
anabolic metabolism and promotes catabolic metabolism. resistance to EGFR inhibitors found that a fraction of tumors
Pharmacologic activation of AMPK with metformin pro- underwent EMT [55]. In conditional GEM models of breast
motes the survival of dormant breast cancer cells [46]. cancer recurrence, recurrent tumors almost universally
Consistent with this, lysosomal degradation of proteins and undergo EMT, and experimental induction of EMT acceler-
organelles via autophagy is induced by chemotherapy [47] ates recurrence [33]. The dormant residual tumors that sur-
and mTOR inhibition [48–51], and inhibiting autophagy vive Her2 downregulation also exhibit features of EMT, and
synergizes with these therapies [48, 49]. Autophagy is a cen- barcoding analysis suggests that the induction of EMT is an
tral cellular response to nutrient deprivation that serves, in adaptive response to Her2 inhibition [56]. Several studies
part, to salvage high energy biomolecules for catabolic pro- have found that many cancer cells exhibit a drug-tolerant
cesses to maintain bioenergetic requirements. Autophagy state characterized by high expression of mesenchymal gene
also promotes dormant cell survival in response to unfavor- signatures [57, 58]. Interestingly, this drug-tolerant mesen-
able cues from the extracellular matrix [52]. These results chymal state can be targeted by inhibitors of the lipid peroxi-
provide further support that therapies cause metabolic stress dase GPX4 [57, 58]. Taken together, these studies indicate
and highlight the importance of metabolic adaptation for that EMT promotes dormant tumor survival and suggest that
dormant cell survival. drugs targeting the mesenchymal state may prevent
recurrence.

10.6  urvival Pathways in Dormant

S
Tumor Cells 10.8 The Dormant Tumor
Microenvironment
The fact that dormant tumor cells are able to survive therapy
and persist for long time periods suggests that these cells The tumor microenvironment is a critical determinant of
have activated pro-survival pathways or disabled apoptotic tumor growth, progression, and response and resistance to
pathways. The pro-apoptotic protein Par-4 (encoded by the therapy. Recent work has focused on characterizing the dor-
Pawr gene) is upregulated in response to diverse therapies, mant tumor microenvironment and understanding how the
10 Therapy-Induced Dormancy and Residual Disease 101

microenvironment contributes to tumor recurrence. A study

of NSCLC patients used single-cell RNA-sequencing to Future Directions
study how tumors change in response to targeted therapy and • Can therapeutic approaches be devised to target
in residual disease [59]. This study revealed both tumor cell-­ residual tumor cells?
intrinsic changes in gene expression and changes to the • Is it preferable to awaken dormant cells to sensitize
tumor immune microenvironment in residual tumors. these cells to chemotherapy, or attempt to keep
Residual tumor cells exhibited an alveolar-regenerative cell these cells dormant to prevent recurrence?
signature, suggesting that therapy induces a cell-state transi- • Are therapy-induced dormancy and metastatic dor-
tion to a more de-differentiated state. The immune composi- mancy regulated by similar cellular and molecular
tion of residual tumors was markedly different from primary pathways?
lung tumors, with increased T cell and decreased macro- • How do chemotherapy and targeted therapy differ-
phage infiltration as compared to primary tumors [59]. entially induce dormancy?
In a mouse model of Burkitt’s lymphoma, chemotherapy • How does therapy-induced dormancy differ
leads to secretion of IL-6 and Timp-1 as a response to DNA between different metastatic sites with distinct
damage, creating a chemoresistant niche [60]. This pro-­ tumor microenvironments?
survival niche permits a population of minimal residual
tumor cells to persist and serve as a reservoir for tumor
relapse. Interestingly, thymic endothelial cells serve as one References
source of the secreted IL-6, highlighting the contribution of
the tumor microenvironment to dormant cell survival. 1. Saphner T, Tormey DC, Gray R. Annual hazard rates of recur-
rence for breast cancer after primary therapy. J Clin Oncol.
In a conditional mouse model of breast cancer, Her2 1996;14:2738–46.
downregulation leads to secretion of TNF, which acts in an 2. Aguirre-Ghiso JA. Models, mechanisms and clinical evidence for
autocrine manner to stimulate the NFkB pathway [61]. cancer dormancy. Nat Rev Cancer. 2007;7:834–46.
Interestingly, an autocrine TNF pathway was also induced by 3. Sosa MS, Bragado P, Aguirre-Ghiso JA. Mechanisms of dissemi-
nated cancer cell dormancy: an awakening field. Nat Rev Cancer.
EGFR inhibition in lung cancer [62]. In both breast cancer 2014;14:611–22.
and lung cancer, TNF acts to promote the survival of residual 4. Braun S, Vogl FD, Naume B, Janni W, Osborne MP, Coombes RC,
cells in response to therapy. Interestingly, in breast cancer Schlimok G, Diel IJ, Gerber B, Gebauer G, Pierga JY, Marth C,
this was mediated through increased expression of CCL5 Oruzio D, Wiedswang G, Solomayer EF, Kundt G, Strobl B, Fehm
T, Wong GY, Bliss J, Vincent-Salomon A, Pantel K. A pooled anal-
and consequent recruitment of macrophages to dormant ysis of bone marrow micrometastasis in breast cancer. N Engl J
tumors [61]. Dormant tumors had abundant macrophages, Med. 2005;353:793–802.
and these macrophages led to collagen deposition in dormant 5. Janni W, Vogl FD, Wiedswang G, Synnestvedt M, Fehm T,
tumors to promote tumor relapse. These studies illustrate Juckstock J, Borgen E, Rack B, Braun S, Sommer H, Solomayer
E, Pantel K, Nesland J, Friese K, Naume B. Persistence of dissemi-
how therapy can induce adaptations in tumor cells that pro- nated tumor cells in the bone marrow of breast cancer patients pre-
mote dormant cell survival through both cell autonomous dicts increased risk for relapse--a European pooled analysis. Clin
mechanisms and via crosstalk with the microenvironment. Cancer Res. 2011;17:2967–76.
6. Pantel K, Muller V, Auer M, Nusser N, Harbeck N, Braun
S. Detection and clinical implications of early systemic tumor cell
dissemination in breast cancer. Clin Cancer Res. 2003;9:6326–34.
10.9 Conclusions and Perspective 7. Symmans WF, Peintinger F, Hatzis C, Rajan R, Kuerer H, Valero
V, Assad L, Poniecka A, Hennessy B, Green M, Buzdar AU,
• The presence of disseminated tumor cells (DTCs) in Singletary SE, Hortobagyi GN, Pusztai L. Measurement of residual
breast cancer burden to predict survival after neoadjuvant chemo-
patients and their persistence after treatment are associ- therapy. J Clin Oncol. 2007;25:4414–22.
ated with increased risk of tumor recurrence 8. Braun S, Pantel K, Muller P, Janni W, Hepp F, Kentenich CR,
• The survival of residual cancer cells following neoadju- Gastroph S, Wischnik A, Dimpfl T, Kindermann G, Riethmuller G,
vant therapy predicts poor prognosis Schlimok G. Cytokeratin-positive cells in the bone marrow and sur-
vival of patients with stage I, II, or III breast cancer. N Engl J Med.
• Dormant residual cancer cells are the reservoir for even- 2000;342:525–33.
tual tumor recurrence 9. Janni W, Rack B, Schindlbeck C, Strobl B, Rjosk D, Braun S,
• Chemotherapy and oncogene-targeted therapy can induce Sommer H, Pantel K, Gerber B, Friese K. The persistence of
cancer cells into a dormant or quiescent state isolated tumor cells in bone marrow from patients with breast
carcinoma predicts an increased risk for recurrence. Cancer.
• Dormant residual tumors have altered metabolism com- 2005;103:884–91.
pared to primary tumors 10. Mathiesen RR, Borgen E, Renolen A, Lokkevik E, Nesland JM,
• The dormant tumor microenvironment contributes to Anker G, Ostenstad B, Lundgren S, Risberg T, Mjaaland I,
tumor recurrence Kvalheim G, Lonning PE, Naume B. Persistence of disseminated
102 A. V. DiMarco et al.

tumor cells after neoadjuvant treatment for locally advanced breast kinase to drug resistance of dormant carcinoma cells. Cancer Res.
cancer predicts poor survival. Breast Cancer Res. 2012;14:R117. 2006;66:1702–11.
11. Naume B, Synnestvedt M, Falk RS, Wiedswang G, Weyde K, 25. Schewe DM, Aguirre-Ghiso JA. ATF6alpha-Rheb-mTOR signaling
Risberg T, Kersten C, Mjaaland I, Vindi L, Sommer HH, Saetersdal promotes survival of dormant tumor cells in vivo. Proc Natl Acad
AB, Rypdal MC, Bendigtsen Schirmer C, Wist EA, Borgen Sci U S A. 2008;105:10519–24.
E. Clinical outcome with correlation to disseminated tumor cell 26. Hata AN, Niederst MJ, Archibald HL, Gomez-Caraballo M, Siddiqui
(DTC) status after DTC-guided secondary adjuvant treatment with FM, Mulvey HE, Maruvka YE, Ji F, Bhang HE, Krishnamurthy
docetaxel in early breast cancer. J Clin Oncol. 2014;32:3848–57. Radhakrishna V, Siravegna G, Hu H, Raoof S, Lockerman E, Kalsy
12. Wiedswang G, Borgen E, Karesen R, Kvalheim G, Nesland JM, Qvist A, Lee D, Keating CL, Ruddy DA, Damon LJ, Crystal AS, Costa
H, Schlichting E, Sauer T, Janbu J, Harbitz T, Naume B. Detection C, Piotrowska Z, Bardelli A, Iafrate AJ, Sadreyev RI, Stegmeier
of isolated tumor cells in bone marrow is an independent prognostic F, Getz G, Sequist LV, Faber AC, Engelman JA. Tumor cells can
factor in breast cancer. J Clin Oncol. 2003;21:3469–78. follow distinct evolutionary paths to become resistant to epidermal
13. Wiedswang G, Borgen E, Karesen R, Qvist H, Janbu J, Kvalheim G, growth factor receptor inhibition. Nat Med. 2016;22:262–9.
Nesland JM, Naume B. Isolated tumor cells in bone marrow three 27. Sharma SV, Lee DY, Li B, Quinlan MP, Takahashi F, Maheswaran
years after diagnosis in disease-free breast cancer patients predict S, McDermott U, Azizian N, Zou L, Fischbach MA, Wong KK,
unfavorable clinical outcome. Clin Cancer Res. 2004;10:5342–8. Brandstetter K, Wittner B, Ramaswamy S, Classon M, Settleman
14. Klein CA. Framework models of tumor dormancy from patient-­ J. A chromatin-mediated reversible drug-tolerant state in cancer
derived observations. Curr Opin Genet Dev. 2011;21:42–9. cell subpopulations. Cell. 2010;141:69–80.
15. Meng S, Tripathy D, Frenkel EP, Shete S, Naftalis EZ, Huth JF, 28. Kurppa KJ, Liu Y, To C, Zhang T, Fan M, Vajdi A, Knelson EH, Xie
Beitsch PD, Leitch M, Hoover S, Euhus D, Haley B, Morrison L, Y, Lim K, Cejas P, Portell A, Lizotte PH, Ficarro SB, Li S, Chen
Fleming TP, Herlyn D, Terstappen LW, Fehm T, Tucker TF, Lane T, Haikala HM, Wang H, Bahcall M, Gao Y, Shalhout S, Boettcher
N, Wang J, Uhr JW. Circulating tumor cells in patients with breast S, Shin BH, Thai T, Wilkens MK, Tillgren ML, Mushajiang M, Xu
cancer dormancy. Clin Cancer Res. 2004;10:8152–62. M, Choi J, Bertram AA, Ebert BL, Beroukhim R, Bandopadhayay
16. Weckermann D, Polzer B, Ragg T, Blana A, Schlimok G, Arnholdt P, Awad MM, Gokhale PC, Kirschmeier PT, Marto JA, Camargo
H, Bertz S, Harzmann R, Klein CA. Perioperative activation of dis- FD, Haq R, Paweletz CP, Wong KK, Barbie DA, Long HW, Gray
seminated tumor cells in bone marrow of patients with prostate can- NS, Janne PA. Treatment-induced tumor dormancy through YAP-­
cer. J Clin Oncol. 2009;27:1549–56. mediated transcriptional reprogramming of the apoptotic pathway.
17. Borgen E, Rypdal MC, Sosa MS, Renolen A, Schlichting E, Lonning Cancer Cell. 2020;37:104–122.e112.
PE, Synnestvedt M, Aguirre-Ghiso JA, Naume B. NR2F1 stratifies 29. Sanchez-Danes A, Larsimont JC, Liagre M, Munoz-Couselo E,
dormant disseminated tumor cells in breast cancer patients. Breast Lapouge G, Brisebarre A, Dubois C, Suppa M, Sukumaran V, Del
Cancer Res. 2018;20:120. Marmol V, Tabernero J, Blanpain C. A slow-cycling LGR5 tumour
18. Pantel K, Schlimok G, Braun S, Kutter D, Lindemann F, Schaller population mediates basal cell carcinoma relapse after therapy.
G, Funke I, Izbicki JR, Riethmuller G. Differential expression of Nature. 2018;562:434–8.
proliferation-associated molecules in individual micrometastatic 30. Alvarez JV, Pan TC, Ruth J, Feng Y, Zhou A, Pant D, Grimley JS,
carcinoma cells. J Natl Cancer Inst. 1993;85:1419–24. Wandless TJ, Demichele A, Investigators IST, Chodosh LA. Par-4
19. Esserman LJ, Berry DA, DeMichele A, Carey L, Davis SE, Buxton downregulation promotes breast cancer recurrence by prevent-
M, Hudis C, Gray JW, Perou C, Yau C, Livasy C, Krontiras H, ing multinucleation following targeted therapy. Cancer Cell.
Montgomery L, Tripathy D, Lehman C, Liu MC, Olopade OI, 2013;24:30–44.
Rugo HS, Carpenter JT, Dressler L, Chhieng D, Singh B, Mies C, 31. Feng Y, Pan TC, Pant DK, Chakrabarti KR, Alvarez JV, Ruth JR,
Rabban J, Chen YY, Giri D, van’t Veer L, Hylton N. Pathologic Chodosh LA. SPSB1 promotes breast cancer recurrence by potenti-
complete response predicts recurrence-free survival more effec- ating c-MET signaling. Cancer Discov. 2014;4:790–803.
tively by cancer subset: results from the I-SPY 1 TRIAL--CALGB 32. Lindstrom LS, Karlsson E, Wilking UM, Johansson U, Hartman J,
150007/150012, ACRIN 6657. J Clin Oncol. 2012;30:3242–9. Lidbrink EK, Hatschek T, Skoog L, Bergh J. Clinically used breast
20. Kuerer HM, Newman LA, Smith TL, Ames FC, Hunt KK, Dhingra cancer markers such as estrogen receptor, progesterone recep-
K, Theriault RL, Singh G, Binkley SM, Sneige N, Buchholz TA, tor, and human epidermal growth factor receptor 2 are unstable
Ross MI, McNeese MD, Buzdar AU, Hortobagyi GN, Singletary throughout tumor progression. J Clin Oncol. 2012;30:2601–8.
SE. Clinical course of breast cancer patients with complete 33. Moody SE, Perez D, Pan TC, Sarkisian CJ, Portocarrero CP, Sterner
pathologic primary tumor and axillary lymph node response to CJ, Notorfrancesco KL, Cardiff RD, Chodosh LA. The transcrip-
doxorubicin-­ based neoadjuvant chemotherapy. J Clin Oncol. tional repressor snail promotes mammary tumor recurrence. Cancer
1999;17:460–9. Cell. 2005;8:197–209.
21. Ogston KN, Miller ID, Payne S, Hutcheon AW, Sarkar TK, Smith I, 34. Boxer RB, Jang JW, Sintasath L, Chodosh LA. Lack of sus-
Schofield A, Heys SD. A new histological grading system to assess tained regression of c-MYC-induced mammary adenocarcino-
response of breast cancers to primary chemotherapy: prognostic mas following brief or prolonged MYC inactivation. Cancer Cell.
significance and survival. Breast. 2003;12:320–7. 2004;6:577–86.
22. Sataloff DM, Mason BA, Prestipino AJ, Seinige UL, Lieber CP, 35. Gunther EJ, Moody SE, Belka GK, Hahn KT, Innocent N, Dugan
Baloch Z. Pathologic response to induction chemotherapy in locally KD, Cardiff RD, Chodosh LA. Impact of p53 loss on reversal and
advanced carcinoma of the breast: a determinant of outcome. J Am recurrence of conditional Wnt- induced tumorigenesis. Genes Dev.
Coll Surg. 1995;180:297–306. 2003;17:488–501.
23. Sosa MS, Avivar-Valderas A, Bragado P, Wen HC, Aguirre-Ghiso 36. Viale A, Pettazzoni P, Lyssiotis CA, Ying H, Sanchez N, Marchesini
JA. ERK1/2 and p38alpha/beta signaling in tumor cell quiescence: M, Carugo A, Green T, Seth S, Giuliani V, Kost-Alimova M, Muller
opportunities to control dormant residual disease. Clin Cancer Res. F, Colla S, Nezi L, Genovese G, Deem AK, Kapoor A, Yao W,
2011;17:5850–7. Brunetto E, Kang Y, Yuan M, Asara JM, Wang YA, Heffernan TP,
24. Ranganathan AC, Zhang L, Adam AP, Aguirre-Ghiso JA. Functional Kimmelman AC, Wang H, Fleming JB, Cantley LC, DePinho RA,
coupling of p38-induced up-regulation of BiP and activation Draetta GF. Oncogene ablation-resistant pancreatic cancer cells
of RNA-dependent protein kinase-like endoplasmic reticulum depend on mitochondrial function. Nature. 2014;514:628–32.
10 Therapy-Induced Dormancy and Residual Disease 103

37. Puig I, Tenbaum SP, Chicote I, Arques O, Martinez-Quintanilla protein kinase connects energy sensing to mitophagy. Science.
J, Cuesta-Borras E, Ramirez L, Gonzalo P, Soto A, Aguilar S, 2011;331:456–61.
Eguizabal C, Caratu G, Prat A, Argiles G, Landolfi S, Casanovas 51. Kim J, Kundu M, Viollet B, Guan KL. AMPK and mTOR regulate
O, Serra V, Villanueva A, Arroyo AG, Terracciano L, Nuciforo autophagy through direct phosphorylation of Ulk1. Nat Cell Biol.
P, Seoane J, Recio JA, Vivancos A, Dienstmann R, Tabernero J, 2011;13:132–41.
Palmer HG. TET2 controls chemoresistant slow-cycling can- 52. Vera-Ramirez L, Vodnala SK, Nini R, Hunter KW, Green
cer cell survival and tumor recurrence. J Clin Invest. 2018;128: JE. Autophagy promotes the survival of dormant breast cancer cells
3887–905. and metastatic tumour recurrence. Nat Commun. 2018;9:1944.
38. Vander Heiden MG, Cantley LC, Thompson CB. Understanding 53. Damrauer JS, Phelps SN, Amuchastegui K, Lupo R, Mabe NW,
the Warburg effect: the metabolic requirements of cell prolifera- Walens A, Kroger BR, Alvarez JV. Foxo-dependent Par-4 upregula-
tion. Science. 2009;324:1029–33. tion prevents long-term survival of residual cells following PI3K-­
39. Alvarez JV, Belka GK, Pan TC, Chen CC, Blankemeyer E, Alavi Akt inhibition. Mol Cancer Res. 2018;16:599–609.
A, Karp J, Chodosh LA. Oncogene pathway activation in mam- 54. Mabe NW, Fox DB, Lupo R, Decker AE, Phelps SN, Thompson
mary tumors dictates [18F]-FDG-PET uptake. Cancer Res. JW, Alvarez JV. Epigenetic silencing of tumor suppressor Par-4
2014;74:7583–98. promotes chemoresistance in recurrent breast cancer. J Clin Invest.
40. Havas KM, Milchevskaya V, Radic K, Alladin A, Kafkia E, 2018;128:4413–28.
Garcia M, Stolte J, Klaus B, Rotmensz N, Gibson TJ, Burwinkel 55. Sequist LV, Waltman BA, Dias-Santagata D, Digumarthy S, Turke
B, Schneeweiss A, Pruneri G, Patil KR, Sotillo R, Jechlinger AB, Fidias P, Bergethon K, Shaw AT, Gettinger S, Cosper AK,
M. Metabolic shifts in residual breast cancer drive tumor recur- Akhavanfard S, Heist RS, Temel J, Christensen JG, Wain JC, Lynch
rence. J Clin Invest. 2017;127:2091–105. TJ, Vernovsky K, Mark EJ, Lanuti M, Iafrate AJ, Mino-Kenudson
41. Fox DB, Garcia NMG, McKinney BJ, Lupo R, Noteware LC, M, Engelman JA. Genotypic and histological evolution of lung
Newcomb R, Liu J, Locasale JW, Hirschey MD, Alvarez JV. NRF2 cancers acquiring resistance to EGFR inhibitors. Sci Transl Med.
activation promotes the recurrence of dormant tumour cells 2011;3:75ra26.
through regulation of redox and nucleotide metabolism. Nat Metab. 56. Walens A, Lin J, Damrauer JS, McKinney B, Lupo R, Newcomb R,
2020;2:318–34. Fox DB, Mabe NW, Gresham J, Sheng Z, Sibley AB, De Buysscher
42. Krall EB, Wang B, Munoz DM, Ilic N, Raghavan S, Niederst MJ, Yu T, Kelkar H, Mieczkowski PA, Owzar K, Alvarez JV. Adaptation
K, Ruddy DA, Aguirre AJ, Kim JW, Redig AJ, Gainor JF, Williams and selection shape clonal evolution of tumors during residual dis-
JA, Asara JM, Doench JG, Janne PA, Shaw AT, McDonald Iii RE, ease and recurrence. Nat Commun. 2020;11:5017.
Engelman JA, Stegmeier F, Schlabach MR, Hahn WC. KEAP1 loss 57. Hangauer MJ, Viswanathan VS, Ryan MJ, Bole D, Eaton JK, Matov
modulates sensitivity to kinase targeted therapy in lung cancer. A, Galeas J, Dhruv HD, Berens ME, Schreiber SL, McCormick F,
Elife. 2017;6:e18970. McManus MT. Drug-tolerant persister cancer cells are vulnerable
43. Chio IIC, Tuveson DA. ROS in cancer: the burning question. Trends to GPX4 inhibition. Nature. 2017;551:247–50.
Mol Med. 2017;23:411–29. 58. Viswanathan VS, Ryan MJ, Dhruv HD, Gill S, Eichhoff OM,
44. Schafer ZT, Grassian AR, Song L, Jiang Z, Gerhart-Hines Z, Irie Seashore-Ludlow B, Kaffenberger SD, Eaton JK, Shimada K,
HY, Gao S, Puigserver P, Brugge JS. Antioxidant and oncogene Aguirre AJ, Viswanathan SR, Chattopadhyay S, Tamayo P, Yang
rescue of metabolic defects caused by loss of matrix attachment. WS, Rees MG, Chen S, Boskovic ZV, Javaid S, Huang C, Wu X,
Nature. 2009;461:109–13. Tseng YY, Roider EM, Gao D, Cleary JM, Wolpin BM, Mesirov
45. Wang H, Nicolay BN, Chick JM, Gao X, Geng Y, Ren H, Gao JP, Haber DA, Engelman JA, Boehm JS, Kotz JD, Hon CS, Chen
H, Yang G, Williams JA, Suski JM, Keibler MA, Sicinska E, Y, Hahn WC, Levesque MP, Doench JG, Berens ME, Shamji
Gerdemann U, Haining WN, Roberts TM, Polyak K, Gygi SP, AF, Clemons PA, Stockwell BR, Schreiber SL. Dependency of a
Dyson NJ, Sicinski P. The metabolic function of cyclin D3-CDK6 therapy-­resistant state of cancer cells on a lipid peroxidase pathway.
kinase in cancer cell survival. Nature. 2017;546:426–30. Nature. 2017;547:453–7.
46. Hampsch RA, Wells JD, Traphagen NA, McCleery CF, Fields JL, 59. Maynard A, McCoach CE, Rotow JK, Harris L, Haderk F, Kerr
Shee K, Dillon LM, Pooler DB, Lewis LD, Demidenko E, Huang DL, Yu EA, Schenk EL, Tan W, Zee A, Tan M, Gui P, Lea T, Wu
YH, Marotti JD, Goen AE, Kinlaw WB, Miller TW. AMPK acti- W, Urisman A, Jones K, Sit R, Kolli PK, Seeley E, Gesthalter Y, Le
vation by metformin promotes survival of dormant ER(+) breast DD, Yamauchi KA, Naeger DM, Bandyopadhyay S, Shah K, Cech
cancer cells. Clin Cancer Res. 2020;26:3707–19. L, Thomas NJ, Gupta A, Gonzalez M, Do H, Tan L, Bacaltos B,
47. Yun CW, Lee SH. The roles of autophagy in cancer. Int J Mol Sci. Gomez-Sjoberg R, Gubens M, Jahan T, Kratz JR, Jablons D, Neff
2018;19:3466. N, Doebele RC, Weissman J, Blakely CM, Darmanis S, Bivona
48. Aveic S, Pantile M, Polo P, Sidarovich V, De Mariano M, Quattrone TG. Therapy-induced evolution of human lung cancer revealed by
A, Longo L, Tonini GP. Autophagy inhibition improves the cyto- single-cell RNA sequencing. Cell. 2020;182:1232–1251.e1222.
toxic effects of receptor tyrosine kinase inhibitors. Cancer Cell Int. 60. Gilbert LA, Hemann MT. DNA damage-mediated induction of a
2018;18:63. chemoresistant niche. Cell. 2010;143:355–66.
49. Egan DF, Chun MG, Vamos M, Zou H, Rong J, Miller CJ, Lou 61. Walens A, DiMarco AV, Lupo R, Kroger BR, Damrauer JS, Alvarez
HJ, Raveendra-Panickar D, Yang CC, Sheffler DJ, Teriete P, Asara JV. CCL5 promotes breast cancer recurrence through macrophage
JM, Turk BE, Cosford ND, Shaw RJ. Small molecule inhibition of recruitment in residual tumors. elife. 2019;8
the autophagy kinase ULK1 and identification of ULK1 substrates. 62. Gong K, Guo G, Gerber DE, Gao B, Peyton M, Huang C, Minna
Mol Cell. 2015;59:285–97. JD, Hatanpaa KJ, Kernstine K, Cai L, Xie Y, Zhu H, Fattah FJ,
50. Egan DF, Shackelford DB, Mihaylova MM, Gelino S, Kohnz Zhang S, Takahashi M, Mukherjee B, Burma S, Dowell J, Dao K,
RA, Mair W, Vasquez DS, Joshi A, Gwinn DM, Taylor R, Asara Papadimitrakopoulou VA, Olivas V, Bivona TG, Zhao D, Habib
JM, Fitzpatrick J, Dillin A, Viollet B, Kundu M, Hansen M, AA. TNF-driven adaptive response mediates resistance to EGFR
Shaw RJ. Phosphorylation of ULK1 (hATG1) by AMP-activated inhibition in lung cancer. J Clin Invest. 2018;128:2500–18.
 Part IV
The Microenvironment of Site-Specific Metastasis

Isaac P. Witz

Cancer research in the seventies and eighties of the last century was dominated by the “cancer-­
centric” view postulating that genomic alterations are the only and exclusive determinants of
carcinogenesis and progression.
At the same time it became increasingly clear to a minority group of cancer researchers that
the cancer-centric view is incapable to solve the many challenges of the pathobiology of cancer
including providing a satisfactory mechanism for the progression towards metastasis. It was
also realized that cancer cells are part of an ecosystem, that they are “surrounded” by the host
(cancer patients or tumor-bearing animals), and that host components take part in tumor
progression.
The cancer ecosystem incorporating host-derived cellular and molecular components as
well as the landscape they create is defined as the tumor microenvironment [1].
Metastasis is the major killer of cancer patients.
One of the key elements of metastasis is its organ specificity. Tumor cells spread to the
vasculature of many organs but they form overt metastases in only a select few to which they
have a marked predilection.
Stephen Paget about 130 years ago was the first scientist to recognize that metastasis is site
specific. Describing this phenomenon in botanical terms he wrote:
“When a plant goes to seed, its seeds are carried in all directions; but they can only live and
grow if they fall on congenial soil.” In contemporary terms, the seeds are the cancer cells and
the soil is the organ that preferentially and specifically hosts future metastasis [1].
A significant step towards the comprehension of the mechanism underlying the formation
of metastasis in specific sites was the discovery of the pre-metastatic niche (PMN) by David
Lyden’s group [2].
Soluble factors and extracellular vesicles released from the primary tumor induce, in distant
organs, the formation of microenvironments that are hospitable for circulating metastasizing
cells before their arrival at these organs. These microenvironments are termed pre-metastatic
niches [3].
The chapter, The Microenvironment of Site-Specific Metastasis, deals with three main sub-
jects: principles underlying microenvironmental interactions of cancer cells with non-cancer
cells, metastatic cascade, and targeting the TME for cancer therapy.

References
1. Maman S, Witz IP. A history of exploring cancer in context. Nat Rev Cancer.
2018;18:359–76.
2. Kaplan RN, Riba RD, Zacharoulis S, Bramley AH, Vincent L, Costa C, MacDonald DD,
Jin DK, Shido K, Kerns SA, Zhu Z, Hicklin D, Wu Y, Port JL, Altorki N, Port ER, Ruggero
D, Shmelkov SV, Jensen KK, Rafii S, Lyden D. VEGFR1-positive haematopoietic bone
marrow progenitors initiate the pre-­metastatic niche. Nature. 2005;438(7069):820–7.
3. Wortzel I, Dror S, Kenific CM, Lyden D. Exosome-mediated metastasis: communication
from a distance. Dev Cell. 2019;49:347–60.
 The Microenvironment of Site-Specific
Metastasis 11
Isaac P. Witz and Sivan Izraely

Abstract
Learning Objectives
Solid tumors can be visualized as an ecosystem composed • Define “Tumor Microenvironment” (TME) and
of cancer and non-cancerous cells and their products, review principles of tumor-TME cross talk.
extra cellular matrix, metabolites, and numerous other • Describe the progression pathway from primary
factors. The cellular components of this ecosystem tumor to metastasis.
(termed tumor microenvironment) form an evolving sig- • Identify TME-associated factors that determine
naling network. The cross talk between cancer and non-­ site-specificity of metastasis.
cancerous cells brings about a progressing phenotype • Review strategies to modulate the cross talk
reprograming of the interacting cells culminating either in between the tumor and its microenvironment for the
tumor demise, dormancy, or progression. The progression development of cancer therapy.
of solid tumors on the road to metastasis is a stepwise
process comprising well-defined steps. An early step is
the formation of a pre-metastatic niche in selected organs
that will ultimately host metastasis. Other progression 11.1 Introduction
steps are: acquisition of a migratory capacity by a sub-
population of tumor cells resulting in their detachment Stephen Paget, over a century ago, was the first to conceive
from the tumor mass, penetration into the circulation, the idea that tumor formation and especially tumor progres-
extravasation into organs hosting a pre-metastatic niche, sion are largely determined by interactions of cancer cells
and finally the establishment of a metastatic lesion. with their microenvironment. This concept, lying dormant
Particular cancers usually metastasize to multiple but until the early seventies of last century, was awakened by
favored organ sites. The factors that determine organ-­ studies on angiogenesis and on the immune landscape of
specific metastasis are: reciprocal recognition elements, tumors [1] conducted by a handful number of researchers.
such as integrins expressed by tumor cells as well as by The comprehension of principles underlying signal trans-
the endothelium of the target organ and survival and duction and of mechanisms by which signaling networks con-
growth factors expressed by this organ. trol gene expression [2, 3] gradually led to the realization that
The realization that interactions between tumors and tumor cells communicate with cells in their immediate and dis-
their microenvironments play pivotal roles in driving or tant environments and that the genome of cancer cells and that
inhibiting tumor progression is being translated to the of non-tumor cells present in the tumor microenvironment
clinical setting in efforts to boost tumor-­microenvironment (TME) is reciprocally regulated by cues derived from these
interactions that inhibit tumor progression and/or oppose interaction partners. It became clear that this cross talk is a
interactions that promote tumor progression. major driver of tumor progression on the road to metastasis [4].
This concept is now well accepted by the cancer-research
community [5].
In addition to tumor cells the TME is composed of host-­
derived resident and recruited cells such as macrophages,
I. P. Witz (*) · S. Izraely
The Shmunis School of Biomedicine and Cancer Research, lymphocytes or fibroblasts, secreted products of these cells
George S. Wise Faculty of Life Science, Tel Aviv University, such as extracellular vesicles and soluble factors such as
Tel Aviv, Israel growth factors, antibodies, components of the extracellular
e-mail: [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 107
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_11
108 I. P. Witz and S. Izraely

matrix (ECM) and metabolites. Moreover the TME of solid originating in a single primary tumor but metastasizing to
tumors is characterized by hypoxia and low pH. different organs (i.e., different microenvironments) will
All the components of the TME create a dynamic interac- express different phenotypes. Studies on the functions of the
tion arena in which a web of signaling networks leads to 3 metastatic microenvironment in resisting or promoting site-­
alternative pathways: tumor progression, tumor regression, specific metastasis are now in the forefront of TME research.
or tumor dormancy. This chapter will review the principles of tumor-TME
The current trends in TME research are mainly focused interactions, the steps of the metastatic cascade, and the prin-
on exploring pathways leading to metastasis. Targeting inter- ciples of site-specific metastasis. Cancer therapy modalities
active tumor-TME pathways such as angiogenesis and involving targeting the TME will also be reviewed (Fig. 11.1).
immune responses as well as TME constituents such as
inflammation has developed into a critical strategy aimed at
the development of novel modalities of cancer therapy. 11.2  he Tumor and Its Microenvironment:
T
One of the major characteristics of metastasis is their Principles
organ specificity [6, 7].
The fact that the tumor-TME cross talk shapes the pheno- Listed below are general features of tumor cells in the con-
type of both tumor and non-tumor cells infers that metastases text of their immediate milieu.

e a

d b

Fig. 11.1 The principles of site-specific metastasis. Site-specific metastatic niche for future metastatic colonization. (d) Specific adhe-
metastasis is manifested by several key principals: (a) Tumor cells sive recognition between tumor cells and the endothelium of the target
derived from a single tumor, due to their phenotypic diversity, metasta- organ is mediated by adhesion molecules such as selectins and integrins
size to multiple organs, that are characteristic targets to each cancer and their ligands, as well as by chemokines and their receptors, that are
type. (b) Alterations in the genetic, epigenetic, and proteomic profiles tumor-type/microenvironment-specific. (e) Binding of survival and
of metastasizing cancer cells largely determine the specificity of target growth factors derived from the metastatic microenvironment to their
organs. (c) Extracellular vesicles (EVs) and soluble factors released cognate receptors expressed by tumor cells enable the survival and
from primary tumor cells induce the release of progenitor cells (bone propagation of tumor cells in the target organ site, and inhibit cell death,
marrow-derived cells (BDMCs)) from bone marrow and their targeted through various signaling cascades
migration to a specific future metastatic site, conditioning the pre-­
11 The Microenvironment of Site-Specific Metastasis 109

1. The TME is jointly created by the tumor and by the host. 5. Tumor-TME interactions are dynamic
This point was referred to in Sect. 11.1. The tumor-TME cross talk reprograms the phenotype of
2. Tumor heterogeneity creates diverse evolutionary path- these interaction partners and leads to a continuous emer-
ways of tumor cells in a single TME. gence of novel tumor variants.
Intratumor heterogeneity involving genetic, epigenetic, 6. The tumor microenvironment may operate as a double
and transcriptomic modifications of tumor cells residing agent
in a specific TME is a pivotal factor that determines tumor Single microenvironmental components may exert pro-
evolution in this microenvironment [8]. or anti-malignancy functions [15].
Numerous and diverse TME-originating signals are The following 2 examples, illustrating the Yin Yang
delivered simultaneously to the heterogeneous population functions of TME components on tumor progression, are
of tumor cells present in a given TME. The outcome of appropriate illustrations of this reality.
this reality is that distinct tumor evolution pathways take (a) In breast cancer tumor infiltrating B cells enhance
place at the same time in a single tumor ecosystem [9, 10]. anti-tumor immune responses but may also create an
3. Tumors having a similar histology but developing in dif- immune-suppressive microenvironment [16].
ferent organs express diverse phenotypes. This point is (b) Interferon gamma (IFNγ) stimulates both anti-tumor
very obvious: For example: epithelial tumors (carcino- immune activities and immune evasion of tumors. In
mas) may develop in different organs expressing features addition, IFNγ may act directly upon tumor cells and
specific to that organ. Since, as mentioned above, the reprograms the expression of molecules that either
microenvironment shapes the phenotype of tumor cells drive or inhibit the malignant phenotype resulting in
and vice versa, carcinomas in different organ microenvi- opposing effects on tumor progression [17].
ronments are different. This point will be highlighted 7. Tumor cells co-opt (hijack) normal biologic processes
when discussing site-specific metastasis (see below). such as mechanisms that regulate immune responses,
4. Reciprocal phenotype reprograming between tumor cells thereby defending themselves from immune attacks [18,
and non-tumor microenvironmental cells. The interac- 19] or chemokine-chemokine axes to direct their traffick-
tions between tumor cells and non-tumor cells in the ing to secondary sites [20].
TME are reciprocal and bidirectional. Each interaction Another physiological process hijacked by tumors is
partner shapes the transcriptomic, proteomic, metabolo- epithelial to mesenchymal transition (EMT) being of
mic, secretomic, etc. phenotype of the other partner. This prime significance in embryonic development [21].
mutual reprograming may drive or block tumor progres- This process confers upon epithelial tumor cells plas-
sion. Several examples illustrate this point. ticity, migratory and invasive and metastasizing capaci-
(a) Tumor cells and the extracellular matrix modulate ties [22].
each other’s phenotype [11], thereby driving tumor
progression.
(b) Cancer cells activate fibroblasts to become tumor-­ 11.3 The Metastatic Cascade
associated fibroblasts which, in turn, promote (or
inhibit) tumor progression and metastasis by a vari- The metastatic cascade is a sequential process in which a
ety of mechanisms [12]. subpopulation of primary tumor cells acquires the capacity
(c) Tumor-derived chemokines promote tumor progres- to migrate, reprogram the extracellular matrix, and invade.
sion by recruiting suppressor cells to the TME, These functions endow the cells with the ability to reach
thereby inhibiting immune responses. Such tumor-­ lymph and/or blood vessels and invade them. Circulating
derived chemokines may also function as autocrine tumor cells (CTC) in the lymph may reach lymph nodes
growth factors [13]. TME-derived chemokines may while CTC circulating in the blood may reach specific target
direct cancer cells expressing the corresponding organs (see below—section on site-specific metastasis). The
receptors to specific organs from which such chemo- CTC may then form micrometastases undetectable by regu-
kines are released [14]. lar imaging techniques, and/or clinically detectable
The balance between opposing effects of tumor-TME metastasis.
cross talk determines the extent of malignancy. If the Each step of the metastatic cascade is jointly controlled
impact of tumor drivers is stronger than that of inhibitors, by tumor-intrinsic factors and by those originating in the
the tumor will progress. If the opposite occurs, the tumor TME [23].
will regress. If these factors cancel each other, the tumor While reviewing the steps of the metastatic cascade we
will enter into a state of dormancy. should keep in mind several realities:
110 I. P. Witz and S. Izraely

• The first steps of the metastatic cascade are common for The EMT empowers the tumor cells to invade neighbor-
different types of tumors while other steps are organ ing tissues, reach the circulation, spread throughout the body,
specific. and subsequently metastasize to specific organs. EMT is
• Metastasis although being an inefficient process is the driven by genetic and epigenetic alterations and by TME-­
leading cause of cancer patient death [24]. derived paracrine cues delivered from resident or infiltrating
• Most cancer patients harbor metastasis by the time of non-tumor cells such as fibroblasts, macrophages, or immu-
diagnosis [25]. nocytes [36].
• The cancer lesion is heterogeneous, comprising a diverse The molecular programs that drive EMT function via
assortment of cellular variants with distinct genetic and microenvironmental multi-signaling pathways upregulate
transcriptomic characteristics. Such variants may prog- several transcription factors such as TWIST, ZEB1, and
ress at different speeds so that at a given time point an SNAIL which are identified as the EMT core signature
individual tumor would contain cells at different stages of [37, 38].
the metastatic cascade [26]. The EMT process follows distinct intermediate stages.
Multiple tumor subpopulations expressing phenotypes rang-
ing from a complete epithelial to a complete mesenchymal
11.3.1  ite-Specific Metastasis: Basic
S one may co-inhabit single solid tumors [39].
Principles In addition to effects on the tumor cells as specified above,
EMT also remodels the extracellular matrix and affects vari-
The predilection of different cancer types to metastasize to ous phenotypic traits of microenvironmental cells. These
particular organs and the organ specificity of metastasis was effects could impact tumor progression [38, 40].
first described by Paget in 1889 [27].
The concept of site-specific metastasis lay dormant for
about 100 years when the basic principles of site-specific 11.3.3 Intravasation and the Generation
metastasis started to be appreciated [6, 28]. of Circulating Tumor Cells

• Different cancer types usually metastasize to multiple but Cancer cell intravasation is the process by which cancer cells
favored organ sites [29] implying that metastases from a penetrate the vessel wall and enter the circulation. While
single tumor but developing in different organs would tumor cells may passively enter the circulation into damaged
express diverse phenotypes. vessels, they may also actively intravasate into intact vessels.
• The genetic, epigenetic, and proteomic profile of the Microenvironmental cells such as macrophages may partici-
metastasizing cancer cells as well as factors derived from pate in the intravasation process by secreting cytokines and
the metastatic microenvironment govern site-specific proteases that increase the permeability of blood and lym-
metastasis [30, 31]. phatic vessels. The density and diameter of vessels play also
• Tumor cells reach the vasculature of all organs but metas- a role in intravasation [41].
tasis develops only in selected organs that host a pre-­ Cancer cells may enter lymphatic vessels, migrate to the
metastatic niche. Specific adhesive recognition between tumor-draining lymph node (referred to as the Sentinel
cancer cells and the endothelium of the target organ, as Lymph Node—SLN), spread to other lymph nodes, develop
well as nonspecific anatomical or mechanical factors such into lesions in these nodes, and then enter the hematogenous
as circulation mechanics are involved in the arrest of cir- vasculature to spread to secondary organs [42, 43].
culating cancer cells at the target organ [32, 33]. The SLN, described for the first time in 1960 [44], is the
• The ability of metastasizing tumor cells to successfully first “foreign” microenvironment encountered by tumor cells
establish metastasis depends to a large extent by survival and following their emigration from the primary tumor [45].
growth factors from the metastatic microenvironment [34]. The primary tumor plays important functions in the
establishment and maintenance of metastasis in the SLN
by inducing immunosuppression [46] and lymphangiogen-
11.3.2  he Acquisition of a Migratory
T esis in this node, thereby generating a pre-metastatic niche
and Invasive Phenotype by Epithelial [47–49].
Cancer Cells Similar to other TMEs, the microenvironment of the
SLN may contribute to distant metastasis by shaping and
The metastatic cascade is initiated when epithelial cancer intensifying the malignant phenotype of tumor cells that
cells acquire invasive, migratory, and stem cell properties. migrated to this node as well as by serving as a resource for
The driver of this phase is a process termed Epithelial to such cells. This issue is however still under debate and
Mesenchymal Transition (EMT) [35]. investigation [43].
11 The Microenvironment of Site-Specific Metastasis 111

Another unsettled issue is whether tumor cells invade the 11.3.5 Colonization
systemic circulation directly from the primary tumor or indi-
rectly through the intermediate residence in the SLN [50]. CTCs that transmigrated the endothelium of specific target
Once in the blood stream, CTC have to survive coagula- organs may succumb to hostile factors (e.g., immune effec-
tion as well as physical and oxidative stress and to escape tors); enter a state of dormancy and form micrometastasis or
damage inflicted by circulating immunocytes such as NK proliferate and progress to form clinically detectable metas-
cells. Indeed, most CTCs die in the circulation. Two major tasis. Factors originating in the metastatic microenvironment
mechanisms defend CTCs from these impediments. The first (MME) mediate these scenarios.
is CTC coating with platelets which provides a physical As mentioned above, TME-derived factors may exert
shield from the destructive effects of shear stress and from anti- and pro-malignancy functions. The same holds true for
being killed by NK cells. The platelets also transduce protec- the MME. A case in point are brain metastases. Microglia
tive signals and may enhance adhesion to blood vessels of (the brain-residing macrophages) in a tumor cell-infiltrated
secondary organs, some of which will host metastasis [51]. brain may secrete both factors that trigger death of invading
The second major mechanism that contributes to the sur- tumor cells and factors that promote propagation of such
vival of CTCs in the circulation is cell clustering. Clustered cells [59].
CTCs are protected from apoptosis caused by lack of contact Cell death. Local immunity such as tumor infiltrating T
with other cells (anoikis) and thus have a survival advantage cells is the major mechanism restricting the propagation of
over solitary CTCs. Clustered CTCs also adhere more effi- CTCs which invaded a secondary organ [60].
ciently to blood vessels of secondary organs [52]. Other MME factors may also be involved in killing infil-
trating tumor cells. For example, microenvironmental
Apo2L/TRAIL, a member of the TNF superfamily, induces
11.3.4  xtravasation of CTCs to Future
E apoptosis of metastatic cells of different tumor types [61].
Metastatic Sites

CTCs may be trapped in small capillaries and invade from 11.3.6 Dormancy and micrometastasis
there to secondary organs [33]. Formation
However the specific interactions between CTCs and the
endothelium of the secondary organ are the major facilitating Tumor dormancy is expressed either as cellular dormancy
element of site-specific extravasation into organs that are (quiescent single cells or small micrometastases or as mass
homes for a pre-metastatic niche [53]. dormancy (an equal rate of proliferation and cell death)
Endothelial cells lining a particular organ often express a [62].
different profile of cell surface proteins compared to endo- Induction and maintenance of dormancy of cancer cells in
thelial cells of other organs [54]. primary tumors and in metastatic sites are controlled by the
This difference accounts for the selective binding of TME and by the MME, respectively [63].
CTCs to the vasculature of particular organs and for their Below are a few examples:
extravasation into these organ sites [55]. Bone-derived factors such as TGFβ2 induce dormancy of
Extravasation is a multistep process employed primarily bone-metastasizing prostate cancer cells [64].
by leukocytes to migrate through the endothelium and home The beta subunit of human hemoglobin (HBB) produced
to inflammatory sites. This process is mediated by interac- by alveolar epithelial and endothelial cells mediated growth
tions of selectins, integrins, cytokines, and chemokines with arrest of lung-metastasizing neuroblastoma metastases [65].
their ligands [56]. The underlying mechanism of these cases of MME-driven
Endothelial selectins, by interacting with corresponding dormancy is a creation of a high p38 and a low ERK signal-
ligands expressed by circulating leukocytes, decelerate their ing activity in cancer cells, thereby inducing dormancy in
flow in the blood causing their rolling on the endothelial layer. these cells [66].
Leukocyte-integrins activated by endothelial chemokines MME-derived factors also awaken dormant tumor cells
interact with their endothelial ligands. This interaction triggers [60].
a firm adhesion between the leukocytes and the endothelial In addition to their capacity to induce dormancy, MME-­
cells. Transendothelial migration (diapedesis) is the final step derived signals are also able to awaken dormant tumor cells.
in the migration of leukocytes into the target organ. For example, dormant human breast cancer cells are awak-
Tumor cells hijacked the extravasation strategy of leuko- ened by IL-6,
cytes and use it to invade specific organs and generate metas- IL-8, TNFα, and TGFβ-mediated stromal inflammation
tasis [57, 58]. [67].
112 I. P. Witz and S. Izraely

11.3.7 Proliferation and Formation Starting in the 1980s, the notion that the immune micro-
of Metastasis environment can be enlisted to curb malignant processes
gained increasing attention. Adoptive T cell therapy employ-
Newly metastasizing as well as awakened dormant tumor ing tumor-infiltrating lymphocytes (TILs) that kill cancer
cells need the support of the MME to proliferate and gener- cells (mostly CD8-positive T cells) is being successfully
ate metastasis. This support is mediated by factors sustaining used to treat certain cancer patients [78, 79].
survival, fostering proliferation, and neutralizing unfavor- Immune checkpoint blockade is a very promising and
able and hostile microenvironmental factors [24, 68]. successful modality of immunotherapy. The 2018 Nobel
Below are several examples of mechanisms by which Prize for Medicine was awarded for the identification of
MME-derived factors support, facilitate, and drive Programmed Cell Death Protein 1 (PD-1) and for conceiving
metastasis. the inhibition of the immune checkpoint strategy hijacked by
VCAM-1 expressing breast cancer cells bind to tumor cells.
α4-integrin expressing MME-associated macrophages. This The physiological function of immune checkpoints is to
binding protects the cancer cells from apoptosis [69]. regulate immune responses, especially autoimmune ones.
The chemokine CXCL12 and tumor necrosis factor-­ This regulation is accomplished by interactions between
related apoptosis-inducing ligand (TRAIL) expressed in the PD-1 expressed by T cells and its PD-L1 ligand expressed by
MME of the bone promote the survival of breast cancer cells antigen-presenting cells. Tumor cells acquired the ability to
in bones [70] and the CX3CL1-CX3CR1 axis plays an express PD-L1 and thereby to interact with the correspond-
important role in organ-specific peritoneal colonization of ing receptor on T cells. This interaction blocks the ability of
ovarian cancer. Reducing the interaction of the chemokine cytotoxic T cells that infiltrated the tumor microenvironment
CX3CL1 with its receptor reduces metastasis [71]. to kill the tumor cells. Inhibiting the PD-L1-PD-1 interaction
Neutrophils in the lung MME support the colonization of by antibodies directed either to the receptor or to its ligand
metastasizing breast cancer cells [72], and IL-1β-expressing blocks this checkpoint and permits the killing of the tumor
cells in the bone MME promoted bone metastasis formation cells [80].
by prostate cancer cells [73]. Targeting interactions between tumor cells and TME cells
The local immune system is the major resistance mecha- that drive tumor progression such as tumor-associated mac-
nism against tumor formation and its progression towards rophages (TAMs), cancer-associated fibroblasts (CAFs), and
metastasis [60]. various immune regulatory cells (T-regs, myeloid-derived
However, tumors have developed several mechanisms to suppressor cells-MDSC) is also studied and applied in the
escape from anti-tumor immunity. The major immunity clinic [1, 81].
escape mechanisms are: creation of immunosuppressive Targeting tumor-TME interactions that promote angio-
tumor and metastatic microenvironments; downregulation of genesis, a crucial event in tumor survival and progression,
MHC class 1 and tumor-associated antigens; recruiting vari- was one of the first efforts to target the TME [82]. Several
ous types of suppressor cells into the TME or MME; and anti-angiogenesis drugs have been approved by the FDA and
hijacking immune checkpoint strategies [74, 75]. are being used in the clinic. Inhibiting interactions between
receptor tyrosine kinases (RTKs) expressed by tumor cells
and TME ligands that stimulate tumor proliferation by RTK
11.4 Targeting Tumor-TME Cross Talk inhibitors or antibodies turned out to be an effective means to
halt tumor progression [83]. Other modalities are in different
The majority of cancer therapy modalities target only cancer stages of clinical development [84].
cells [76]. An exciting possibility to shift the balance of tumor-TME
However both tumors and their microenvironments play interactions to tumor inhibition is the possibility to “normal-
pivotal roles in shaping the malignant phenotype of tumors ize” the TME, for example, by restoring the abnormal and
and in metastasis formation (see above). leaky cancer vasculature or the metabolic defects in the TME
The huge database on the TME (up to 2020, about 48,000 [85, 86]. Two caveats have to be taken into account when
publications) lead the scientific community to employ dis- considering the targeting of the tumor-TME cross talk. The
coveries in this area for the benefit of cancer patients. The first is that the diverse targeting approaches are mostly gen-
two main avenues in such endeavors are to boost tumor-­ erally applicable and not restricted by organ specificity. The
microenvironment interactions that antagonize tumor pro- second is that there are quite many challenges in applying
gression and/or block interactions that promote tumor TME targeting. The following two examples illustrate such
progression [77]. Enhancing local anti-tumor immunity is challenges. In view of the yin-yang functions of many micro-
the most pursued contemporary avenue. environmental factors (see above), targeting these factors
11 The Microenvironment of Site-Specific Metastasis 113

may result in opposite results from those expected. Acknowledgments The authors thank their colleagues Dr. Orit Sagi-
Assif, Tsipi Meshel and Shlomit Ben-Menachem, the Laboratory of
Furthermore, due to the complexity of tumor-TME cross Tumor Microenvironment & Metastasis Research for their valuable
talk, TME-targeted treatments may instigate unintended help and support. The research conducted by the authors is supported
adverse side effects [1]. by the Dr. Miriam and Sheldon G. Adelson Medical Research
Foundation (Needham, MA, USA), the Sara and Natan Blutinger
Foundation (West Orange, NJ, USA), the Fred August and Adele
Wolpers Charitable Fund (Clifton, NJ, USA) and the James and Rita
11.5 Conclusions and Perspective Leibman Endowment Fund for Cancer Research (New York, NY, USA).

• Tumor cells and non-tumor cells in the tumor microenvi-

ronment maintain ongoing and dynamic interactions,
thereby reprograming each other’s phenotype. References
• Tumors having a similar histology (e.g., carcinomas) but
developing in different organs express diverse 1. Maman S, Witz IP. A history of exploring cancer in context. Nat
Rev Cancer. 2018;18(6):359–76.
phenotypes. 2. Schlessinger J. Signal transduction by allosteric receptor oligomer-
• Microenvironmental components may under certain cir- ization. Trends Biochem Sci. 1988;13(11):443–7.
cumstances promote tumor progression and under differ- 3. Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam
ent circumstances block it. S. The protein kinase complement of the human genome. Science.
2002;298(5600):1912–34.
• The sequence of events starting with the detachment of 4. Witz IP. Tumor-microenvironment interactions: dangerous liaisons.
tumor cells from the tumor mass and culminating in the Adv Cancer Res. 2008;100:203–29.
formation of frank metastasis in a distant organ is called 5. Hanahan D, Weinberg RA. Hallmarks of cancer: the next genera-
the metastatic cascade. tion. Cell. 2011;144(5):646–74.
6. Fidler IJ. Cancer metastasis. Br Med Bull. 1991;47(1):157–77.
• Metastasis is site-specific. Certain cancer types have a 7. Obenauf AC, Massague J. Surviving at a distance: organ-specific
marked predilection for metastasis to specific organs. metastasis. Trends Cancer. 2015;1(1):76–91.
Tumor cells reach the endothelium of most organs; how- 8. McGranahan N, Swanton C. Clonal heterogeneity and tumor evolu-
ever they can form metastasis in only select ones. tion: past, present, and the future. Cell. 2017;168(4):613–28.
9. Zhang M, Rosen JM. Developmental insights into breast cancer
• Metastasis is formed in organs hosting a pre-metastatic Intratumoral heterogeneity. Trends Cancer. 2015;1(4):242–51.
niche. 10. de Looff M, de Jong S, Kruyt FAE. Multiple interactions between
• As discussed above, the tumor and the TME shape each cancer cells and the tumor microenvironment modulate TRAIL
other’s phenotype. In view of the fact that the microenvi- signaling: implications for TRAIL receptor targeted therapy. Front
Immunol. 2019;10:1530.
ronment of diverse organs is different, metastases origi- 11. Malik R, Lelkes PI, Cukierman E. Biomechanical and biochemi-
nating in a single tumor but residing in diverse organ sites cal remodeling of stromal extracellular matrix in cancer. Trends
are different. Biotechnol. 2015;33(4):230–6.
• In view of the fact that the tumor microenvironment takes 12. Kalluri R. The biology and function of fibroblasts in cancer. Nat
Rev Cancer. 2016;16(9):582–98.
part in the formation of metastasis efforts to target the 13. Schneider GP, Salcedo R, Welniak LA, Howard OM, Murphy
microenvironment are increasingly employed in the clinic. WJ. The diverse role of chemokines in tumor progression: pros-
pects for intervention (review). Int J Mol Med. 2001;8(3):235–44.
14. Klein A, Sagi-Assif O, Meshel T, Telerman A, Izraely S, Ben-­
Menachem S, Bayry J, Marzese DM, Ohe S, Hoon DSB, Erez N,
Questions Witz IP. CCR4 is a determinant of melanoma brain metastasis.
• What phenotypic changes may occur in tumor cells Oncotarget. 2017;8(19):31079–91.
following their interaction with microenvironmen- 15. Witz IP. Yin-yang activities and vicious cycles in the tumor micro-
tal components? environment. Cancer Res. 2008;68(1):9–13.
16. Zhang Z, Zhu Y, Wang Z, Zhang T, Wu P, Huang J. Yin-yang effect
• Why do epithelial tumors of the ovaries and of the of tumor infiltrating B cells in breast cancer: from mechanism to
colon express different phenotypic characteristics? immunotherapy. Cancer Lett. 2017;393:1–7.
• What determines the organ specificity of metastasis 17. Ni L, Lu J. Interferon gamma in cancer immunotherapy. Cancer
originating in a single tumor? Med. 2018;7(9):4509–16.
18. Ribas A. Adaptive immune resistance: how cancer protects from
• Please describe the process of tumor cell extravasa- immune attack. Cancer Discov. 2015;5(9):915–9.
tion. Is this process restricted exclusively to tumor 19. Strazza M, Mor A. The complexity of targeting chemokines to pro-
cells? mote a tumor immune response. Inflammation. 2020;43(4):1201–8.
• Can tumor-TME interactions be targeted? If yes, 20. Muller A, Homey B, Soto H, Ge N, Catron D, Buchanan ME,
McClanahan T, Murphy E, Yuan W, Wagner SN, Barrera JL, Mohar
how? Please provide 2 examples for clinical studies A, Verastegui E, Zlotnik A. Involvement of chemokine receptors in
employing the strategy of targeting tumor-TME breast cancer metastasis. Nature. 2001;410(6824):50–6.
interactions. 21. Nieto MA, Huang RY, Jackson RA, Thiery JP. Emt: 2016. Cell.
2016;166(1):21–45.
114 I. P. Witz and S. Izraely

22. Derynck R, Weinberg RA. EMT and cancer: more than meets the factor-C increases metastasis by increasing delivery of can-
eye. Dev Cell. 2019;49(3):313–6. cer cells to lymph nodes: therapeutic implications. Cancer Res.
23. Valastyan S, Weinberg RA. Tumor metastasis: molecular insights 2006;66(16):8065–75.
and evolving paradigms. Cell. 2011;147(2):275–92. 43. Zhou H, Lei PJ, Padera TP. Progression of metastasis through lym-
24. Massague J, Obenauf AC. Metastatic colonization by circulating phatic system. Cell. 2021:10(3).
tumour cells. Nature. 2016;529(7586):298–306. 44. Gould EA, Winship T, Philbin PH, Kerr HH. Observations on a
25. Fidler IJ, Kripke ML. The challenge of targeting metastasis. Cancer “sentinel node” in cancer of the parotid. Cancer. 1960;13:77–8.
Metastasis Rev. 2015;34(4):635–41. 45. Sleeman JP. The lymph node as a bridgehead in the metastatic dis-
26. Dagogo-Jack I, Shaw AT. Tumour heterogeneity and resistance to semination of tumors. Recent Results Cancer Res. 2000;157:55–81.
cancer therapies. Nat Rev Clin Oncol. 2018;15(2):81–94. 46. Jones D, Pereira ER, Padera TP. Growth and immune evasion of
27. Paget S. The distribution of secondary growths in cancer of the lymph node metastasis. Front Oncol. 2018;8:36.
breast. 1889. Cancer Metastasis Rev. 1989;8(2):98–101. 47. Qian CN, Berghuis B, Tsarfaty G, Bruch M, Kort EJ, Ditlev J,
28. Hart IR. 'Seed and soil' revisited: mechanisms of site-specific Tsarfaty I, Hudson E, Jackson DG, Petillo D, Chen J, Resau JH,
metastasis. Cancer Metastasis Rev. 1982;1(1):5–16. Teh BT. Preparing the “soil”: the primary tumor induces vascula-
29. Nguyen DX, Bos PD, Massague J. Metastasis: from dis- ture reorganization in the sentinel lymph node before the arrival of
semination to organ-specific colonization. Nat Rev Cancer. metastatic cancer cells. Cancer Res. 2006;66(21):10365–76.
2009;9(4):274–84. 48. Hood JL, San RS, Wickline SA. Exosomes released by melanoma
30. Horak E, Darling DL, Tarin D. Analysis of organ-specific effects on cells prepare sentinel lymph nodes for tumor metastasis. Cancer
metastatic tumor formation by studies in vitro. J Natl Cancer Inst. Res. 2011;71(11):3792–801.
1986;76(5):913–22. 49. Olmeda D, Cerezo-Wallis D, Riveiro-Falkenbach E, Pennacchi
31. Maman S, Witz IP. The metastatic microenvironment. In: Shurin PC, Contreras-Alcalde M, Ibarz N, Cifdaloz M, Catena X, Calvo
MR, Umansky V, Malyguine A, editors. The tumor immunoenvi- TG, Canon E, Alonso-Curbelo D, Suarez J, Osterloh L, Grana O,
ronment. New York, NY: Springer; 2013. p. 15–38. Mulero F, Megias D, Canamero M, Martinez-Torrecuadrada JL,
32. Tarin D, Price JE. Influence of microenvironment and vascu- Mondal C, Di Martino J, Lora D, Martinez-Corral I, Bravo-Cordero
lar anatomy on "metastatic" colonization potential of mammary JJ, Munoz J, Puig S, Ortiz-Romero P, Rodriguez-Peralto JL, Ortega
tumors. Cancer Res. 1981;41(9 Pt 1):3604–9. S, Soengas MS. Whole-body imaging of lymphovascular niches
33. Wirtz D, Konstantopoulos K, Searson PC. The physics of cancer: identifies pre-­metastatic roles of midkine. Nature. 2017;546(7660):
the role of physical interactions and mechanical forces in metasta- 676–80.
sis. Nat Rev Cancer. 2011;11(7):512–22. 50. Nathanson SD. ASO author reflections: is breast cancer dissemina-
34. Radinsky R. Molecular mechanisms for organ-specific colon carci- tion lymphatic, Hematogenous, or both; and does it matter? Ann
noma metastasis. Eur J Cancer. 1995;31A(7–8):1091–5. Surg Oncol. 2020;27(Suppl 3):711–2.
35. Yang J, Antin P, Berx G, Blanpain C, Brabletz T, Bronner M, 51. Micalizzi DS, Maheswaran S, Haber DA. A conduit to
Campbell K, Cano A, Casanova J, Christofori G, Dedhar S, metastasis: circulating tumor cell biology. Genes Dev.
Derynck R, Ford HL, Fuxe J, Garcia de Herreros A, Goodall GJ, 2017;31(18):1827–40.
Hadjantonakis AK, RJY H, Kalcheim C, Kalluri R, Kang Y, Khew-­ 52. Fabisiewicz A, Grzybowska E. CTC clusters in cancer progression
Goodall Y, Levine H, Liu J, Longmore GD, Mani SA, Massague and metastasis. Med Oncol. 2017;34(1):12.
J, Mayor R, McClay D, Mostov KE, Newgreen DF, Nieto MA, 53. Peinado H, Zhang H, Matei IR, Costa-Silva B, Hoshino A,
Puisieux A, Runyan R, Savagner P, Stanger B, Stemmler MP, Rodrigues G, Psaila B, Kaplan RN, Bromberg JF, Kang Y, Bissell
Takahashi Y, Takeichi M, Theveneau E, Thiery JP, Thompson MJ, Cox TR, Giaccia AJ, Erler JT, Hiratsuka S, Ghajar CM, Lyden
EW, Weinberg RA, Williams ED, Xing J, Zhou BP, Sheng G, D. Pre-metastatic niches: organ-specific homes for metastases. Nat
Association EMTI. Guidelines and definitions for research on Rev Cancer. 2017;17(5):302–17.
epithelial-­
mesenchymal transition. Nat Rev Mol Cell Biol. 54. Augustin HG, Koh GY. Organotypic vasculature: From descrip-
2020;21(6):341–52. tive heterogeneity to functional pathophysiology. Science.
36. Hass R, von der Ohe J, Ungefroren H. The intimate relationship 2017;357(6353)
among EMT, MET and TME: A T(ransdifferentiation) E(nhancing) 55. Orr FW, Wang HH, Lafrenie RM, Scherbarth S, Nance
M(ix) to be exploited for therapeutic purposes. Cancers (Basel). DM. Interactions between cancer cells and the endothelium in
2020;12(12) metastasis. J Pathol. 2000;190(3):310–29.
37. Peinado H, Olmeda D, Cano A. Snail, Zeb and bHLH factors in 56. Vestweber D. How leukocytes cross the vascular endothelium. Nat
tumour progression: an alliance against the epithelial phenotype? Rev Immunol. 2015;15(11):692–704.
Nat Rev Cancer. 2007;7(6):415–28. 57. Miles FL, Pruitt FL, van Golen KL, Cooper CR. Stepping out of
38. Dongre A, Weinberg RA. New insights into the mechanisms of the flow: capillary extravasation in cancer metastasis. Clin Exp
epithelial-mesenchymal transition and implications for cancer. Nat Metastasis. 2008;25(4):305–24.
Rev Mol Cell Biol. 2019;20(2):69–84. 58. Strilic B, Offermanns S. Intravascular survival and extravasation of
39. Pastushenko I, Brisebarre A, Sifrim A, Fioramonti M, Revenco tumor cells. Cancer Cell. 2017;32(3):282–93.
T, Boumahdi S, Van Keymeulen A, Brown D, Moers V, Lemaire 59. Wu SY, Watabe K. The roles of microglia/macrophages in tumor
S, De Clercq S, Minguijon E, Balsat C, Sokolow Y, Dubois C, De progression of brain cancer and metastatic disease. Front Biosci
Cock F, Scozzaro S, Sopena F, Lanas A, D’Haene N, Salmon I, (Landmark Ed). 2017;22:1805–29.
Marine JC, Voet T, Sotiropoulou PA, Blanpain C. Identification 60. Angelova M, Mlecnik B, Vasaturo A, Bindea G, Fredriksen T,
of the tumour transition states occurring during EMT. Nature. Lafontaine L, Buttard B, Morgand E, Bruni D, Jouret-Mourin
2018;556(7702):463–8. A, Hubert C, Kartheuser A, Humblet Y, Ceccarelli M, Syed N,
40. Paolillo M, Schinelli S. Extracellular matrix alterations in meta- Marincola FM, Bedognetti D, Van den Eynde M, Galon J. Evolution
static processes. Int J Mol Sci. 2019:20(19). of metastases in space and time under immune selection. Cell.
41. Chiang SP, Cabrera RM, Segall JE. Tumor cell intravasation. Am J 2018;175(3):751–65. e16
Physiol Cell Physiol. 2016;311(1):C1–C14. 61. Mace TA, Yamane N, Cheng J, Hylander BL, Repasky
42. Hoshida T, Isaka N, Hagendoorn J, di Tomaso E, Chen YL, EA. The potential of the tumor microenvironment to influ-
Pytowski B, Fukumura D, Padera TP, Jain RK. Imaging steps of ence Apo2L/TRAIL induced apoptosis. Immunol Investig.
lymphatic metastasis reveals that vascular endothelial growth 2006;35(3–4):279–96.
11 The Microenvironment of Site-Specific Metastasis 115

62. Sosa MS, Bragado P, Aguirre-Ghiso JA. Mechanisms of dissemi- 72. Wculek SK, Malanchi I. Neutrophils support lung coloni-
nated cancer cell dormancy: an awakening field. Nat Rev Cancer. zation of metastasis-initiating breast cancer cells. Nature.
2014;14(9):611–22. 2015;528(7582):413–7.
63. Linde N, Fluegen G, Aguirre-Ghiso JA. The relationship between 73. Shahriari K, Shen F, Worrede-Mahdi A, Liu Q, Gong Y, Garcia FU,
dormant cancer cells and their microenvironment. Adv Cancer Res. Fatatis A. Cooperation among heterogeneous prostate cancer cells
2016;132:45–71. in the bone metastatic niche. Oncogene. 2017;36(20):2846–56.
64. Yu-Lee LY, Yu G, Lee YC, Lin SC, Pan J, Pan T, Yu KJ, Liu B, 74. Drake CG, Jaffee E, Pardoll DM. Mechanisms of immune evasion
Creighton CJ, Rodriguez-Canales J, Villalobos PA, Wistuba II, de by tumors. Adv Immunol. 2006;90:51–81.
Nadal E, Posas F, Gallick GE, Lin SH. Osteoblast-secreted factors 75. Schaller J, Agudo J. Metastatic colonization: escaping immune sur-
mediate dormancy of metastatic prostate cancer in the bone via acti- veillance. Cancers (Basel). 2020;12(11)
vation of the TGFbetaRIII-p38MAPK-pS249/T252RB pathway. 76. Bashraheel SS, Domling A, Goda SK. Update on targeted cancer
Cancer Res. 2018;78(11):2911–24. therapies, single or in combination, and their fine tuning for preci-
65. Maman S, Sagi-Assif O, Yuan W, Ginat R, Meshel T, Zubrilov I, sion medicine. Biomed Pharmacother. 2020;125:110,009.
Keisari Y, Lu W, Lu W, Witz IP. The Beta subunit of hemoglobin 77. Sung SY, Chung LW. Prostate tumor-stroma interaction: molecu-
(HBB2/HBB) suppresses neuroblastoma growth and metastasis. lar mechanisms and opportunities for therapeutic targeting.
Cancer Res. 2017;77(1):14–26. Differentiation. 2002;70(9–10):506–21.
66. Aguirre-Ghiso JA, Liu D, Mignatti A, Kovalski K, 78. Hinrichs CS, Rosenberg SA. Exploiting the curative poten-
Ossowski L. Urokinase receptor and fibronectin regulate the tial of adoptive T-cell therapy for cancer. Immunol Rev.
ERK(MAPK) to p38(MAPK) activity ratios that determine car- 2014;257(1):56–71.
cinoma cell proliferation or dormancy in vivo. Mol Biol Cell. 79. Paijens ST, Vledder A, de Bruyn M, Nijman HW. Tumor-infiltrating
2001;12(4):863–79. lymphocytes in the immunotherapy era. Cell Mol Immunol. 2020;
67. Tivari S, Lu H, Dasgupta T, De Lorenzo MS, Wieder R. Reawakening 80. Topalian SL, Drake CG, Pardoll DM. Immune checkpoint block-
of dormant estrogen-dependent human breast cancer cells by ade: a common denominator approach to cancer therapy. Cancer
bone marrow stroma secretory senescence. Cell Commun Signal. Cell. 2015;27(4):450–61.
2018;16(1):48. 81. Laplagne C, Domagala M, Le Naour A, Quemerais C, Hamel D,
68. Shibue T, Weinberg RA. Metastatic colonization: settlement, adap- Fournie JJ, Couderc B, Bousquet C, Ferrand A, Poupot M. Latest
tation and propagation of tumor cells in a foreign tissue environ- advances in targeting the tumor microenvironment for tumor sup-
ment. Semin Cancer Biol. 2011;21(2):99–106. pression. Int J Mol Sci. 2019;20(19)
69. Chen Q, Zhang XH, Massague J. Macrophage binding to recep- 82. Ribatti D. History of research on angiogenesis. Chem Immunol
tor VCAM-1 transmits survival signals in breast cancer cells that Allergy. 2014;99:1–14.
invade the lungs. Cancer Cell. 2011;20(4):538–49. 83. Ghosh S, Marrocco I, Yarden Y. Roles for receptor tyrosine kinases
70. Zhang XH, Wang Q, Gerald W, Hudis CA, Norton L, Smid in tumor progression and implications for cancer treatment. Adv
M, Foekens JA, Massague J. Latent bone metastasis in breast Cancer Res. 2020;147:1–57.
cancer tied to Src-dependent survival signals. Cancer Cell. 84. Rajabi M, Mousa SA. The role of angiogenesis in cancer treatment.
2009;16(1):67–78. Biomedicine. 2017;5(2)
71. Gurler Main H, Xie J, Muralidhar GG, Elfituri O, Xu H, Kajdacsy-­ 85. Jain RK. Normalizing tumor microenvironment to treat cancer:
Balla AA, Barbolina MV. Emergent role of the fractalkine axis in bench to bedside to biomarkers. J Clin Oncol. 2013;31(17):2205–18.
dissemination of peritoneal metastasis from epithelial ovarian car- 86. Quail DF, Joyce JA. Microenvironmental regulation of tumor pro-
cinoma. Oncogene. 2017;36(21):3025–36. gression and metastasis. Nat Med. 2013;19(11):1423–37.
 Part V
Circulating Cancer Cells and Liquid Biopsy

Catherine Alix-Panabières and Klaus Pantel

Research on disseminating tumor cells in cancer patients and animal models has opened new
avenues for better understanding of the biology of metastasis as the leading cause of cancer-­
related death. This fundamental work has led to a novel diagnostic concept designated “liquid
biopsy” [1] which included first the analysis of circulating tumor cells (CTCs) and has been
now extended to cell-free tumor-derived material (in particular circulating tumor DNA,
ctDNA) released by tumor cells in the peripheral blood of cancer patients [2].
Over the past decade, various methods have been developed to detect CTCs and ctDNA in
the peripheral blood of cancer patients. While reliable information can be easily obtained in
patients with advanced disease, early-stage cancer patients usually present with very low con-
centrations of CTCs and ctDNA [3]. At present, most CTC assays rely on epithelial markers
and the majority of CTCs detected are single isolated cells. The clinical relevance of “mesen-
chymal” CTCs lacking any epithelial markers as well as CTC clusters are still under investiga-
tion [4–6]. Although most published studies have been performed on patients with carcinomas
and melanomas, CTCs have also been detected in the peripheral blood of patients with primary
brain tumors (glioblastomas) despite the blood-brain barrier [7].
Liquid biopsy assays are currently being validated for early detection of cancer, which is
supposed to reduce cancer-related mortality [8]. Despite remarkable progress, liquid biopsy-­
based detection of early stages of cancer remains a challenge, in particular in breast cancer.
New blood-based biomarkers for early detection currently validated in clinical trials include
miRNAs, exosomes, tumor-associated proteins, and tumor-educated platelets.
In patients with diagnosed cancer, CTCs and ctDNA analyses can obtain independent infor-
mation on prognosis in early and advanced stages of disease. In particular, CTC counts at ini-
tial diagnosis are able to refine the current risk stratification by TNM staging in early-stage
breast cancer [9]. Moreover, early detection of relapse by sequential ctDNA (or CTCs) analy-
sis of blood samples obtained post-surgery during the follow-up [10] is possible and may be
used in future trials to stratify patients to “post-adjuvant” therapies [11].
Another key application of liquid biopsy is to identify therapeutic targets or mechanisms of
resistance of metastatic cells in individual patients [2]. While the analysis of ctDNA focuses
on mutations relevant for cancer therapy (e.g., EGFR, KRAS, or ESR1 mutations), CTCs offer
a wide spectrum of analyses at the DNA, RNA, and protein levels. Metastatic cells might have
unique characteristics that can differ from the bulk of cancer cells in the primary tumor cur-
rently used for stratification of patients to systemic therapy. Moreover, monitoring of CTCs
and ctDNA before, during, and after systemic therapy (e.g., chemotherapy, hormonal therapy,
antibody therapy) might provide unique information for the future clinical management of the
individual cancer patients and might serve as a surrogate marker for response to therapy. In the
context of recent success in antibody-mediated blockade of immune checkpoint control mol-
ecules, expression of the PD-L1 on CTCs might be of interest as a potential predictive marker
[12, 13]. Moreover, the expression of androgen receptor variant 7 in CTCs may predict resis-
118 Circulating Cancer Cells and Liquid Biopsy

tance to anti-androgen therapy in prostate cancer [14], while mutations in the estrogen receptor gene (ESR1)
provide information on resistance to hormone therapy in breast cancer [15].
Additional therapeutic targets detected on CTCs in cancer patients include the estrogen receptor and
HER-2 oncogene [2]. Single-cell RNAseq analysis of CTCs may provide more comprehensive information
on relevant pathways.
Considering to introduce liquid biopsy into clinical practice, its clinical utility must be proven [16]. For
the first time in this field, clinical utility of CTCs has been proven based on the CTC METABREAST study
(NCT01710605) [17]. Indeed, the CTC count is a reliable biomarker method for choosing between chemo-
therapy and endocrine therapy as the first-line treatment in patients with a metastatic cancer HR+ HER2−.
For functional analysis of CTCs, the development of in vitro and in vivo test systems has started, which
might also serve as models for drug testing. In particular, the development of cell lines and xenografts
derived from CTCs can provide novel insights into the biology of tumor cell dissemination and may be used
to discover new pathways to target specifically metastatic cells [18, 19].
Besides CTCs and ctDNA the analysis of circulating microRNAs, exosomes, tumor-associated proteins,
or tumor-educated platelets may provide complementary information as “liquid biopsy.” For example, the
integrin composition of exosomes seems to determine the organ site of metastatic niches and the RNA
expression pattern of blood platelets reveals information on tumors in cancer patients.
Sensitive methods have also been developed to capture disseminated tumor cells (DTCs) in the bone
marrow in cancer patients [11], which provide new insights into the process of “cancer dormancy.” The
nature of dormant breast cancer cells and the mechanisms leading to their outgrowth are poorly understood.
Efforts to unravel the nature of cancer dormancy have been hampered by the lack of sensitive methods to
detect dormant cells in cancer patients. The development of novel therapies designed to kill dormant resid-
ual tumor cells, or maintain them in a quiescent state, represents a highly attractive approach to prevent late
recurrence. Such an approach, however, would require a far more detailed understanding of tumor dor-
mancy and recurrence than that which exists today, as well as biomarkers to enable monitoring of this
process and predict recurrence. Analysis of DTCs leads to the discovery of new molecules relevant to the
biology of metastasis such as the putative metastasis-suppressor RAI2 [20].
In conclusion, liquid biopsy analysis can be used to obtain new insights into metastasis biology, and as
companion diagnostics to improve the stratification of therapies and to obtain insights into therapy-induced
selection of cancer cells. Different approaches such as CTC or ctDNA analysis will provide complementary
information [16]. Technical and clinical assay validation is very important and can be achieved in interna-
tional consortia such as the European Liquid Biopsy Society (www.elbs.eu) or the US-based BloodPAC [21].
This book provides a comprehensive overview of current research on cancer metastasis including experi-
mental and translational studies covering the entire spectrum of investigations from discovery of novel
genes and pathways to implementation of clinical biomarkers in personalized medicine.

References
1. Pantel K, Alix-Panabieres C. Circulating tumour cells in cancer patients: challenges and perspectives.
Trends Mol Med. 2010;16:398–406. https://doi.org/10.1016/j.molmed.2010.07.001.
2. Alix-Panabieres C, Pantel K. Liquid biopsy: from discovery to clinical application. Cancer Discov.
2021;11:858–73. https://doi.org/10.1158/2159-­8290.CD-­20-­1311.
3. Pantel K, Alix-Panabieres C. Liquid biopsy and minimal residual disease - latest advances and implica-
tions for cure. Nat Rev Clin Oncol. 2019;16:409–24. https://doi.org/10.1038/s41571-­019-­0187-­3.
4. Szczerba BM, et al. Neutrophils escort circulating tumour cells to enable cell cycle progression. Nature.
2019;566:553–7. https://doi.org/10.1038/s41586-­019-­0915-­y.
5. Gkountela S, et al. Circulating tumor cell clustering shapes DNA methylation to enable metastasis
seeding. Cell. 2019;176:98–112 e114. https://doi.org/10.1016/j.cell.2018.11.046.
6. Krol I, et al. Detection of clustered circulating tumour cells in early breast cancer. Br J Cancer. 2021;
https://doi.org/10.1038/s41416-­021-­01327-­8.
7. Muller C, et al. Hematogenous dissemination of glioblastoma multiforme. Sci Transl Med.
2014;6:247ra101. https://doi.org/10.1126/scitranslmed.3009095.
8. Pantel K. Blood tests for early detection of lung cancer: challenges and promises. Lancet Respir Med.
2020;8:654–6. https://doi.org/10.1016/S2213-­2600(20)30115-­6.
Circulating Cancer Cells and Liquid Biopsy 119

9. Bidard FC, et al. Circulating tumor cells in breast cancer patients treated by neoadjuvant chemotherapy:
a meta-analysis. J Natl Cancer Inst. 2018;110:560–7. https://doi.org/10.1093/jnci/djy018.
10. Trapp E, et al. Presence of circulating tumor cells in high-risk early breast cancer during follow-up and
prognosis. J Natl Cancer Inst. 2019;111:380–7. https://doi.org/10.1093/jnci/djy152.
11. Pantel K, Hayes DF. Disseminated breast tumour cells: biological and clinical meaning. Nat Rev Clin
Oncol. 2017; https://doi.org/10.1038/nrclinonc.2017.174.
12. Jacot W, et al. Clinical correlations of programmed cell death ligand 1 status in liquid and standard
biopsies in breast cancer. Clin Chem. 2020;66:1093–101. https://doi.org/10.1093/clinchem/hvaa121.
13. Sinoquet, L. et al. Programmed cell death ligand 1-expressing circulating tumor cells: a new prognostic
biomarker in non-small cell lung cancer. Clin Chem In press 2021.
14. El-Heliebi A, et al. In situ detection and quantification of AR-V7, AR-FL, PSA, and KRAS point muta-
tions in circulating tumor cells. Clin Chem. 2018;64:536–46. https://doi.org/10.1373/
clinchem.2017.281295.
15. Turner NC, et al. ESR1 mutations and overall survival on Fulvestrant versus Exemestane in advanced
hormone receptor-positive breast cancer: a combined analysis of the phase III SoFEA and EFECT tri-
als. Clin Cancer Res. 2020;26:5172–7. https://doi.org/10.1158/1078-­0432.CCR-­20-­0224.
16. Alix-Panabieres C. The future of liquid biopsy. Nature. 2020;579:S9. https://doi.org/10.1038/
d41586-­020-­00844-­5.
17. Bidard FC, et al. Efficacy of circulating tumor cell count-driven vs clinician-driven first-line therapy
choice in hormone receptor-positive, ERBB2-negative metastatic breast cancer: the STIC CTC ran-
domized clinical trial. JAMA Oncol. 2021;7:34–41. https://doi.org/10.1001/jamaoncol.2020.5660.
18. Koch C, et al. Characterization of circulating breast cancer cells with tumorigenic and metastatic capac-
ity. EMBO Mol Med. 2020;12:e11908. https://doi.org/10.15252/emmm.201911908.
19. Cayrefourcq L, et al. Selective treatment pressure in colon cancer drives the molecular profile of resis-
tant circulating tumor cell clones. Mol Cancer. 2021;20:30. https://doi.org/10.1186/
s12943-­021-­01326-­6.
20. Werner S, et al. Suppression of early hematogenous dissemination of human breast cancer cells to bone
marrow by retinoic acid-induced 2. Cancer Discov. 2015;5:506–19. https://doi.org/10.1158/2159-­8290.
CD-­14-­1042.
21. Connors D, et al. International liquid biopsy standardization alliance white paper. Crit Rev Oncol
Hematol. 2020;156:103112. https://doi.org/10.1016/j.critrevonc.2020.103112.
 Circulating Tumor Cells and ctDNA
in Sarcomas 12
Camille Jubelin, Denis Cochonneau, Emilie Moranton,
Javier Muñoz-Garcia, and Dominique Heymann

Abstract
Learning Objectives
Sarcomas are clustered in two oncological entities • To get a global overview of the principal methods
named bone and soft tissue sarcomas. Both are rare can- available for isolating circulating tumor cells in
cers originating from the mesenchyme, characterized by sarcomas.
their propensity to induce the development of lung • To gain insight into the biological value of circulat-
metastases. Sarcoma cells escaping from the primary ing tumor DNA in sarcoma patients.
tumor site spread to the pulmonary tissue through the • To benefit of a brief summary of the mimimal resid-
bloodstream where they found a favorable microenvi- ual disease in sarcomas.
ronment to establish metastatic foci. The low number of
patients, the high histological, genetic, and molecular
heterogeneity of sarcomas combined with the absence of
specific markers expressed by cancer cells make the 12.1 Introduction
detection and follow-­up of the minimal residual disease
challenging. Over the last decade, tremendous techno- Sarcomas belong to a large family of malignant entities of
logical progress has been made towards the detection of mesenchymal origin [1]. These rare cancers can be classified
recurrent diseases. The literature is now enriched of in two large categories, bone sarcomas and soft tissues sarco-
information describing the use of liquid biopsies in clin- mas (STS), according to their anatomical site of occurrence.
ical care of sarcoma patients. This chapter aims to give a More than 60 subtypes have been defined based on their
brief overview of the most recent data available on the molecular characteristics. Indeed, in addition to their com-
detection of circulating tumor cells and circulating mon origin, sarcomas share a high heterogeneity in terms of
tumor DNA in sarcomas. molecular/genetic and histological profile, leading to a con-
stant increase of the number of entities increase in parallel
with technological advances [2]. The three main subtypes of
bone sarcomas are osteosarcoma, Ewing sarcoma, and chon-
drosarcoma [3]. The two first are mostly observed in adoles-
cent and young adults whereas the peak of incidence of
C. Jubelin · D. Cochonneau · E. Moranton · J. Muñoz-Garcia
chondrosarcoma is principally seen in the 4 decades of age.
Nantes Université, CNRS, US2B, UMR 6286, Institut de
Cancérologie de l’Ouest, Blvd Jacques Monod, These diseases are generally treated using surgery and poly-­
Saint-Herblain, France chemotherapeutic approaches; however, despite the improve-
e-mail: [email protected]; ment of multimodal therapies, around one-third of patients
[email protected];
still succumb to sarcomas. Nowadays, the overall survival of
[email protected];
[email protected] osteosarcoma and Ewing sarcoma patients is around 70% at
5 years with localized disease. This survival rate drops to
D. Heymann (*)
Nantes Université, CNRS, US2B, UMR 6286, Institut de 30% when metastases are detected at the time of diagnosis.
Cancérologie de l’Ouest, Blvd Jacques Monod, Among STS, leiomyosarcoma, liposarcoma, and undifferen-
Saint-Herblain, France tiated pleomorphic sarcoma are the most common malignant
Department of Oncology and Metabolism, entities. Similar to bone sarcomas, the STS treatment
University of Sheffield, Sheffield, UK approach mixes surgery and drug administration (e.g., che-
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 121
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_12
122 C. Jubelin et al.

motherapy, tyrosine kinase inhibitor). The 5-year overall sur- cific markers expressed at the surface of sarcoma cells. This
vival for STS is estimated around 60%, depending on the restriction is mainly related to the low number of patients,
disease stage, the anatomical site, and histological/molecular the high intratumoral heterogeneity of sarcomas, their high
subtype [2]. diversity, and the absence of universal cell markers expressed
The natural history of sarcomas is frequently associated by most cancer cells.
with the development of metastases, which differ signifi- In the absence of specific cell markers, researchers have
cantly from those observed in carcinoma. Indeed, while car- proposed several methods for isolating CTCs from sarco-
cinomas use the lymphatic vasculature for spreading and mas. Defined as rare cell events in blood (1 CTC for 105–108
invading the lymph node chain located at proximity of the leukocytes) the isolation of CTCs requires two main techni-
tumor site, sarcoma cells predominantly spread through the cal steps: (i) a first step of pre-enrichment leading to the
blood vasculature [4]. Minimal residual disease (MDR) is enrichment of blood samples, followed by (ii) a second step,
defined as malignant cells that are resistant to treatment, and corresponding to the final cell isolation (Fig. 12.1a).
that remain in patients after remission, leading to relapse and Enrichment methods (Step 1) include physical based
metastasis. MDR is composed of drug-resistant tumor cells, approaches, as well as biological based techniques. The high
dynamically presented as persister/dormant/quiescent/can- size and low deformability of cancer cells are two important
cer cells in residual tumors [5, 6]. Circulating tumor cells physical features used for pre-enriching CTCs. Indeed, con-
(CTCs) in peripheral blood as well as disseminated tumor sidering these two characteristics, sarcoma cells can easily
cells in the bone marrow and other metastatic sites are com- be trapped in filter and in microcapillaries (microfluidics)
mon examples of MDR. MDR behavior is consequently [11]. In contrast, biological approaches are mainly based on
impacted by the respective microenvironment of the cancer immunomagnetic isolation techniques (e.g., negative deple-
cells concerned (e.g., cytokines/hormones, cells, epigenetic tion, nonspecific immunostaining). Isolation steps (Step 2)
events, therapies). The current biological methods for detect- use various technical devises including flow cytometry,
ing MDR (including local tumor recurrence and developing dielectrophoretic array, and microfluidics with antibody-­
metastatic foci) are based on the detection of CTCs and cir- based precoated systems (Fig. 12.1b–d). By using these two-­
culating tumor DNA (ctDNA). Whether such approaches step-­based enrichment/selection techniques, CTCs have
have overtaken the proof-of-concept in carcinomas, the clini- been successfully isolated in sarcomas [10, 12, 13]. For
cal significance of liquid biopsies as a prognostic or predic- instance, CTCs were previously detected in bone [14, 15]
tive tool in sarcoma is currently uncertain and under and STS patients [16–18] (Table 12.1).
investigation [7–10]. Liquid biopsies are non- or weakly The high plasticity of cancer cells makes the identification
invasive methods based on blood collection that might be of specific and universal CTCs markers difficult to diagnose.
useful for the diagnosis and clinical follow-up of cancer. Epithelial cells have been frequently isolated by using epithe-
This chapter gives a brief overview of the most recent data lial biomarkers (e.g., EpCAM, cytokeratin). However, there is
available in the literature about the isolation of CTCs and strong evidence that cancer cells isolated only represent a low
detection of ctDNA and discusses the clinical relevance of proportion of CTCs and constitute the emerged part of the
such approaches in the context of sarcomas. iceberg. Indeed, epithelial mesenchymal transition and cell
plasticity contribute to the loss or acquisition of epithelial and
mesenchymal markers according to the location of cancer
12.2  irculating Tumor Cells in Sarcoma:
C cells. Sarcoma cells originate from the mesenchyme but
Methods of Detection and Biological exhibit similar plasticity to carcinoma cells complicating the
Value identification of specific markers. CD99 is a heavily
O-glycosylated transmembrane protein expressed by Ewing
Cancer cells can escape from tumor foci and intravasate for sarcoma cells and spindle cell tumors such as synovial sarco-
migrating to distant sites before to extravasate and establish mas. Ewing sarcoma is one of the first sarcomas from which
metastases. CTCs are then cells released from primary and CTCs were initially isolated thanks to their CD99 expression
metastatic tumor foci, which migrate to secondary location (Table 12.1). In this context, research groups have been look-
through the peripheral blood. In contrast to epithelial tumors, ing for universal sarcoma markers in order to improve the
publications describing the isolation of CTCs in sarcomas sensitivity and specificity of CTC isolation. Recently, the
are currently limited [10]. Over the last decade, accumulat- cell-surface vimentin has emerged as a potential universal
ing data has described the isolation of CTCs in sarcoma biomarker of human sarcoma CTCs [25, 33]. Cell-surface
patients. However, researchers have been confronted to sev- vimentin was expressed by a high diversity of sarcoma sub-
eral obstacles during the diagnostics of CTC, including the types (bone sarcomas: osteosarcoma, Ewing sarcoma; soft
restricted number of patients, the high histological/genetic/ tissue sarcomas: angiosarcoma, leiomyosarcoma, undifferen-
molecular heterogeneity of the disease, and the lack of spe- tiated pleomorphic sarcoma). Indeed, the majority of the sar-
12 Circulating Tumor Cells and ctDNA in Sarcomas 123

a - Size and deformability (filtration, microfluidics)

- Density: density gradient (Ficoll-Hypaque, Percoll,...)
- Membrane electric charges (DEPArray)

Physicochemical
parameters Blood - Pre-enrichment step
samples

Biological
- Cell markers (membranous and cytoplasmic) parameters
Flow cytometry, CellSearch, DEPArray, etc)
- Functional properties (EPIDROP, Telomerscan, etc)

- Isolation step
b c

d
DAPI

Vimentin

CD99

Fig. 12.1 Detection of circulating tumor cells in sarcomas. (a) was prepared by PARSORTIX, followed by a cysospin and HE stain-
Isolation methods of CTCs combine a first pre-enrichment step fol- ing. (c) CTCs from an undifferentiated epithelioid sarcoma. Vimentin:
lowed by an isolation step. These steps are based on physical or/and red; DAPI: blue; (d) CTCs isolated for an Ewing sarcoma. Ten mL of
biological properties of CTCs. (b) CTCs isolated from blood patient blood were pre-enriched by Parsortix followed by CTCs isolation using
suffering from a pleomorphic liposarcoma. CTC pre-enriched fraction DEPArray after vimentin/CD99 immunostaining and DAPI staining

coma cell lines studied show high expression levels of series of blood sarcoma patients [25]. In a series of 77 gastro-
cell-surface vimentin. This characteristic was positively used intestinal stromal tumors, Li et al. identified CTCs, express-
by Satelli et al., who designed a 84-1 monoclonal antibody ing specific macrophage biomarkers (CD14 and CD68) [33].
targeting cell-surface vimentin for isolating CTCs in a short This new class of CTCs were CD45 negative but showed cell-
124 C. Jubelin et al.

Table 12.1 Detection of circulating tumor cells in sarcomas: methods and main results
Bone sarcomas
Detection/isolation
Sarcoma type method Biomarker Main conclusions References
Ewing sarcoma RT-PCR EWS-FLI-1/ERG CTCs detection in patients with localized [19, 20]
disease and association with poor outcome.
RT-PCR EWS-FLI-1/ERG Correlation between CTC detection and [21]
disease progression
RT-PCR EWS-FLI-1/ERG CTC detection related to reduced survival. [22]
RT-PCR EWS-FLI-1/ERG No prognostic data [23, 24]
Immunomagnetic Isolation of CD99+ cells Feasibility of finding CTCs in PB of ES [14]
separation with EWSR1/FLI1 followed patients by immunoseparation with CD99
microbeads by RT-PCR antibody and magnetic microbead
Immunomagnetic Cell-surface vimentin Identification of the first universal and [25]
separation with specific CTC marker described for
microbeads enumerating CTCs from different types of
sarcoma
Osteosarcoma RT-PCR mRNA of osteoblast Correlation of collagen type I expression and [26]
related genes the development of metastases.
PCR-ELISA Osf2 mRNA Correlation of Osf2 mRNA and metastasis in [27]
a mouse model
DEPArray GFP-cells Detection of CTCs before any detectable [28]
tumor mass in a mouse model. Modulation of
CTCs number by chemotherapy? High
heterogenetity of CTCs
ISET Cell size CTCs in patients with high-grade sarcoma [15]
Immunomagnetic Cell-surface vimentin Detection of cell surface vimentin+ CTCs in [13]
separation with cryopreserved peripheral blood mononuclear
microbeads cells.
Ewing sarcoma and CellSieve Cell size Detection of CTCs in bone sarcoma patients [12]
osteosarcoma
Giant cell tumor of bone Immunomagnetic Cell surface vimentin Detection of CTCs in patient blood [29]
separation with
microbeads + on Chip sort
Soft tissue sarcomas
Sarcoma type Detection method Marker Main conclusions References
Alveolar soft part sarcoma RT-PCR ASPSCR1-TFE3 Detection of CTCs in patient blood [30]
ApoStream Vimentin+, cytokeratin+/− Detection of CTCs in patient blood [31]
DEP field and β-catenin+ +
ASPSCR1-TFE3 by
FISH
Rhabdomyosarcoma RT-PCR PAX3-FKHR Detection of CTCs in patient blood [32]
Myxofibrosarcoma Immunomagnetic Cell surface vimentin Detection of CTCs in patient blood [12]
separation with
microbeads + on Chip sort
Leiomyosarcoma, Immunomagnetic Cell-surface vimentin Cell-surface vimentin as a universal sarcoma [25]
angiosarcoma, pleomorphic separation with CTC marker by using a monoclonal antibody
sarcoma microbeads
Synovial sarcoma, ISET Cell size CTCs in patients with high-grade sarcoma [15]
Dedifferentiated
liposarcoma
Gastrointestinal stromal Immunomagnetic Cell-surface vimentin/ Detection of cell surface vimentin+/CD68+ [33]
tumors separation with CD68 macrophage-like CTCs
microbeads
Gastrointestinal stromal Immunomagnetic Cell-surface vimentin Detection of cell surface vimentin+ CTCs in [13]
tumors, liposarcoma separation with cryopreserved peripheral blood mononuclear
microbeads cells.
Soft Tissue sarcoma Aptamer Cell-surface vimentin Feasibility in spiking experiments [34]
Soft tissue sarcoma Epic Sciences Cytokeratin−, vimentin+ No association was observed between the [35]
(clinical trial) expression and image CTC enumeration and clinical outcome.
analysis algorithm
CTC Circulating tumor cells, DEP Dielectophoresis, ISET Isolation by size
12 Circulating Tumor Cells and ctDNA in Sarcomas 125

surface vimentin, CD14 and CD68 positivity in addition to cancer cells composing the tumor tissue and fluctuate along
DOG1 and cKIT specific gastrointestinal stromal tumor tumor growth, as well as during tumor treatment. In this con-
markers [33]. Interestingly, the number of these vimentin/ text, ctDNAs have been used as diagnostic, prognostic, and
macrophage CTCs was correlated with metastatic status of predictive biomarkers of cancer. More specifically, based on
patients in contrast to CTCs lacking macrophage markers that their clinical interest, ctDNA has, over the last decade, been
failed to predict metastasis formation [33]. Cell-surface assessed in various sarcomas [39, 40].
vimentin could be extended to carcinoma cells that under- The presence of ctDNA was previously detected in plasma
went epithelial mesenchymal transition [36]. In a preclinical patients suffering from bone sarcoma [41–43] and STS [44–
model of osteosarcoma, CTCs appeared to be an early bio- 48]. Although ctDNA can be released from both healthy and
marker, detectable before any palpable/detectable tumors cancer cells, the ctDNA levels were higher in sarcoma
[28]. The number of CTCs fluctuated along the chemotherapy patients than in the cancer-free control group. Interestingly,
and was increased after ifosfamide administration. mutations related to cancer initiation and/or development
Surprisingly, this study showed an inverse correlation between were identified from ctDNA analysis. Mutated genes
CTCs enumeration and tumor growth, as well as between included TP53, PIK3CA IDH1, fusion oncogenes (e.g.,
CTC number and metastatic disease [28]. These results dem- SS18-SSX1/2), or cytokine receptors. cKIT and PDGFR are
onstrated that all CTCs did not possess intrinsic properties of indeed frequently mutated in gastrointestinal of stromal
metastases as revealed by the higher number of CTCs in ifos- tumors (GIST) and these mutations are not only detectable in
famide-treated mice. CTCs enumeration should indeed not be ctDNA but mutated ctDNA is quantitatively correlated with
the single criteria of their biological value but their molecular clinical progression [49]. The identification of the mutational
profiling appeared mandatory. Overall, these recent studies profile of ctDNA represents an interesting tool for predicting
highlighted the clinical potential of CTCs as prognostic or the functional resistance of sarcomas to tyrosine-kinase
predictive markers in sarcomas. Recently, multiple center inhibitors in clinic as observed in GIST patients [50].
phase 1b clinical trials investigated the feasibility to enumer- Mutational profile can be also predicted from ctDNA in plas-
ate CTCs in 51 patients with potentially resectable soft tissue mas of bone sarcoma patients. For instance, mutated IDH1
sarcoma treated with olaratumab. This monoclonal antibody detected in chondrosarcoma were correlated with tumor
was used to target platelet-derived growth factor receptor grade and poor prognosis [42]. Next-generation sequencing
(PDGFR)-α for one cycle. The first cycle was followed then was used to look for somatic single nucleotide variants,
by up to 6 cycles of olaratumab combined with doxorubicin insertions and deletions, as well as structural variants in
[35]. The isolation methods of CTCs relied on multiple stain- osteosarcoma patients [41, 48]. In Ewing sarcoma, somatic
ing, especially of cellular intracytoplasmic elements like mutations observed from ctDNA were correlated with infe-
cytokeratins and intermediate filaments. The expression pro- rior outcomes [48]. Ewing sarcomas are often characterized
files were analyzed using EPIC Sciences algorithm that mea- by their fusion gene expression in cancer cells, which can be
sured around 90 cell parameters (e.g., immunostaining, quantified from plasma ctDNA. Mutated ctDNAs represent
morphology, size). CTCs were detectable in collected blood valuable biomarkers for the diagnostic assessment and thera-
samples; their number increased at cycle 1 and decreased by peutic follow-up of cancer [51]. Cell-free EWSR1-ETS
cycle 3. Interestingly, the decrease of CTCs in patients treated fusion sequences have been assessed in Ewing sarcoma
with olaratumab monotherapy was higher in controlled dis- patients. Their copy-numbers correlated with patient’s risk
ease. These results showed the potential clinical value of factors including tumor volume and metastatic status. In
CTCs in sarcoma. Longitudinal clinical trials with large series addition, a rapid drop of their plasma levels was observed
of patients are now mandatory (Table 12.1). during treatment phases. In contrast, an increased level of
their copy-number was a clinical mark of tumor relapse [51].
More recently, studies have detected circulating nucleo-
12.3 Detection of Circulating Tumor DNA somes in blood, revealing that ctDNA retains at least some
in Sarcoma characteristics of the nuclear chromatin, a useful information
for determining the origin of cell-free ctDNA. Whole genome
In healthy and cancer tissues, apoptotic- and necrosis-related sequencing of ctDNA can give a genome-wide map of
cell death results in the release of DNA fragments into the nucleosome occupancy, allowing the identification of the
bloodstream, which can be easily detected by PCR. Tissue cell types that contribute to circulating ctDNA [52, 53]. Such
homeostasis is often associated with the physiological observation highly relevant to Ewing sarcoma led to follow
release of cell fragments and contributes to the circulating the chromatin state for monitoring therapeutic response in
DNA [37, 38]. When this circulating free DNA fraction is real time [54–56]. Sheffield et al. demonstrated that DNA
released from a tumor mass, it is entitled circulating tumor methylation heterogeneity defined a disease spectrum in
DNA (ctDNA). These ctDNAs reflect the genetic heritage of Ewing sarcoma [55]. Complementary investigations are
126 C. Jubelin et al.

mandatory to determine the clinical value of ctDNA meth- Patient Information and Guidelines
ylation heterogeneity. • EuSARC (European network for SARComa): https://
Overall, the detection and characterization of ctDNA in eusarc.com/
sarcomas similarly to epithelial cancers represent a promis- • NIH website for information to health professional,
ing prognostic/diagnostic tool. Investigation of larger bio- related to bone cancer: https://www.cancer.gov/types/
logical sarcoma cohorts with complete clinical annotations is bone/hp
currently needed to validate their clinical values. • World sarcoma network: http://www.worldsarcomanet-
work.com/

12.4 Conclusion Acknowledgments The authors are also grateful to M. Clément

Heymann for reviewing the English style.
Based on several emerging evidence, it is now convention-
ally approved that liquid biopsies, including CTCs and
ctDNA, represent very efficient tools for the clinical follow- References
up in oncology. The biological value of CTCs was initially
assessed in carcinoma patients and used to compare the 1. WHO Classification of Tumours Editorial Board. Soft tissue and
bone tumours, WHO classification of tumours series, vol. 3. 5th ed.
genetic/epigenetic profiles of primary tumors and CTCs Lyon: International Agency for Research on Cancer; 2020.
[57]. More recently, such approaches have been extended to 2. Grünewald TG, Alonso M, Avnet S, Banito A, Burdach S, Cidre-­
sarcoma. Overall, these revealed that CTCs partly reflect the Aranaz F, Di Pompo G, Distel M, Dorado-Garcia H, Garcia-Castro
mutational landscape of the primary and metastatic tumor J, González-González L, Grigoriadis AE, Kasan M, Koelsche
C, Krumbholz M, Lecanda F, Lemma S, Longo DL, Madrigal-­
foci [57–60] and that CTCs should be considered as a snap- Esquivel C, Morales-Molina Á, Musa J, Ohmura S, Ory B, Pereira-­
shot of the tumor tissue heterogeneity at a given time [58, Silva M, Perut F, Rodriguez R, Seeling C, Al Shaaili N, Shaabani
61]. Recent methods allow to better identify the origin of S, Shiavone K, Sinha S, Tomazou EM, Trautmann M, Vela M,
detected ctDNA. Both ctDNA and CTCs could be an added Versleijen-Jonkers YM, Visgauss J, Zalacain M, Schober SJ, Lissat
A, English WR, Baldini N, Heymann D. Sarcoma treatment in the
value for longitudinal patient monitoring. Although the era of molecular medicine. EMBO Mol Med. 2020;12:e11131.
detection of ctDNA appears easier than CTCs, CTCs are 3. Brown HK, Schiavone K, Gouin F, Heymann MF, Heymann
advantageous for characterizing the heterogeneity of the D. Biology of bone sarcomas and new therapeutic developments.
mutational spectrum and orientating the future therapeutic Calcif Tissue Int. 2018;102(2):174–95.
4. Pennacchioli E, Tosti G, Barberis M, De Pas TM, Verrecchia F,
options towards the development of new platform for drug Menicanti C, Testori A, Mazzarol G. Sarcoma spreads primarily
screening [62]. through the vascular system: are there biomarkers associated with
vascular spread? Clin Exp Metastasis. 2012;29(7):757–73.
5. Nicolini A, Rossi G, Ferrari P, Carpi A. Minimal residual disease in
advanced or metastatic solid cancers: The G0-G1 state and immu-
Open Questions notherapy are key to unwinding cancer complexity. Semin Cancer
• Which is the best method for isolating CTCs in Biol. 2020;21:1054.
sarcomas? 6. Vallette FM, Olivier C, Lézot F, Oliver L, Cochonneau D, Lalier
L, Cartron PF, Heymann D. Dormant, quiescent, tolerant and per-
• What is the prognostic value of CTCs and ctDNA in sister cells: four synonyms for the same target in cancer. Biochem
sarcomas? Pharmacol. 2019;162:169–76.
• Do CTCs reflect the tumor heterogeneity in 7. Alix-Panabières C, Schwarzenbach H, Pantel K. Circulating tumor
sarcomas? cells and circulating tumor DNA. Annu Rev Med. 2012;63:199–215.
8. Cortés-Hernández LE, Eslami-S Z, Alix-Panabières C. Circulating
• What is the clonal evolution of CTCs along tumor cells as the functional aspect of liquid biopsy to understand the
treatments? metastatic cascade in solide cancer. Mol Asp Med. 2020;72:100816.
• Do CTCs have the same capacity to induce meta- 9. Coombs CC, Dickherber T, Crompton BD. Chasing ctDNA
static foci? in patients with sarcoma. Am Soc Clin Oncol Educ Book.
2020;40:e351–60.
• Do CTCs have dormant/quiescent/stemness 10. Tellez-Gabriel M, Brown HK, Young R, Heymann MF, Heymann
properties? D. The challenges of detecting circulating tumor cells in sarcoma.
Front Oncol. 2016;6:202.
12 Circulating Tumor Cells and ctDNA in Sarcomas 127

11. Gabriel MT, Calleja LR, Chalopin A, Ory B, Heymann D. Circulating 26. Wong IH, Chan AT, Johnson PJ. Quantitative analysis of circulat-
tumor cells: a review of non-EpCam-based approaches for cell ing tumor cells in peripheral blood of osteosarcoma patients using
enrichment and isolation. Clin Chem. 2016;62(4):571–81. osteoblast-specific messenger RNA markers: a pilot study. Clin
12. Hayashi M, Zhu P, McCarty G, Meyer CF, Pratilas CA, Levin A, Cancer Res. 2000;6(6):2183–8.
Morris CD, Albert CM, Jackson KW, Tang CM, Loeb DM. Size-­ 27. Hatano H, Kawashima H, Ogose A, Hotta T, endo N. A PCR-­
based detection of sarcoma circulating tumor cells and cell clusters. ELISA assay for the detection of disseminated osteosarcoma cells
Oncotarget. 2017;8(45):78,965–77. in a mouse metastatic model. J Orthop Sci. 2001;6(3):269–75.
13. Li H, Meng QH, Noh H, Batth IS, Somaiah N, Torres KE, Xia X, 28. Chalopin A, Tellez-Gabriel M, Brown HK, Vallette F, Heymann
Wang R, Li S. Detection of circulating tumor cells from cryopre- MF, Gouin F, Heymann D. Lsolation of circulating tumor cells in a
served human sarcoma peripheral blood mononuclear cells. Cancer preclinical model of osteosarcoma: effect of chemotherapy. J Bone
Lett. 2017;403:216–23. Oncol. 2018;12:83–90.
14. Benini S, Gamberi G, Cocchi S, Garbetta J, Alberti L, Righi A, 29. Hasegawa N, Nakamura IT, Ueno T, Kojima S, Kawazu M, Akaike
Gambarotti M, Picci P, Ferrari S. Detection of circulating tumor K, Okubo T, Takagi T, Suehara Y, Hayashi T, Saito T, Kaneko K,
cells in liquid biopsy from Ewing sarcoma patients. Cancer Manag Kohsaka S. Detection of circulating sarcoma tumor cells using a
Res. 2018;10:49–60. microfluidic chip-type cell sorter. Sci Rep. 2019;9(1):20047.
15. Chinen LTD, Mello CAL, Abdallah EA, Ocea LM, Buuim 30. Hoshino M, Ogose A, Kawasima H, Izumi T, Hotta T, Hatano H,
ME, Breve NM, Junior JLG, Fanelli MF, Paterlini-Brechot Morita T, Otsuka H, Yanoma S, Tsukuda M, Endo N. Molecular
P. Isolation, detection, and immunomorphological characteriza- analyses of cell origin and detection of circulating tumor cells in
tion of circulating tumor cells (CTCs) from patients with dif- the peripheral blood in alveolar soft part sarcoma. Cancer Genet
ferent types of sarcoma using isolation by size of tumor cells: a Cytogenet. 2009;190(2):75–80.
window on sarcoma-­ cell invasion. Onco Targets Ther. 2014;7: 31. Balasubramanian P, Kinders RJ, Kummar S, Gupta V, Hasegawa
1609–17. D, Menachery A, Lawrence SM, Wang L, Ferry-Galow K, Davis
16. Braun AC, de Mello CAL, Corassa M, Abdallah EA, Urvanegia D, Parchment RE, Tomaszewski JE, Doroshow JH. Antibody-­
AC, Alves VS, Flores BCTCP, Díaz M, Nicolau UR, Silva VSE, independent capture of circulating tumor cells of non-epithelial ori-
Calsavara V, Paterlini-Brechót P, Chinen LTD. EGFR expression in gin with the ApoStream® system. PLoS One. 2017;12(4):e0175414.
circulating tumor cells from high-grade metastatic soft tissue sarco- 32. Kelly KM, Womer RB, Barr FG. Minimal disease detection in
mas. Cancer Biol Ther. 2018;19(6):454–60. patients with alveolar rhabdomyosarcoma using a reverse transcrip-
17. Mihály D, Nagy N, Papp G, Pápai Z, Sápi Z. Release of circulat- tase-polymerase chain reaction method. Cancer. 1996;78(6):1320–7.
ing tumor cells and cell-free nucleic acids is an infrequent event in 33. Li H, Meng QH, Noh H, Somaiah N, Torres KE, Xia X, Batth IS,
synovial sarcoma: liquid biopsy analysis of 15 patients diagnosed Joseph CP, Liu M, Wang R, Li S. Cell-surface vimentin-positive
with synovial sarcoma. Diagn Pathol. 2018;13(1):81. macrophage-like circulating tumor cells as a novel biomarker of
18. Przybyl, van de Rijn M, Rutkowski P. Detection of SS18-SSX1/2 metastatic gastrointestinal stromal tumors. Onco Targets Ther.
fusion transcripts in circulating tumor cells of patients with syno- 2018;7(5):e1420450.
vial sarcoma. Diagn Pathol. 2019;14(1):24. 34. Zheng Y, Zhang J, Huang M, Wang T, Qu X, Wu L, Song J, Wang
19. West DC, Grier HE, Swallow MM, Demetri GD, Granowetter W, Song Y, Yang C. selection of aptamers against vimentin for iso-
L, Sklar J. Detection of circulating tumor cells in patients with lation and release of circulating tumor cells undergoing epithelial
Ewing's sarcoma and peripheral primitive neuroectodermal tumor. mesenchymal transition. Anal Chem. 2020;92(7):5178–84.
J Clin Oncol. 1997;15(2):583–8. 35. Martín-Broto J, Pousa AL, Brohl AS, Van Tine BA, Powers B,
20. Schleiermacher G, Peter M, Oberlin O, Philip T, Rubie H, Stacchiotti S, Blay JY, Hu JS, Oakley GJ 3rd, Wang H, Szpurka
Mechinaud F, Sommelet-Olive D, Ladman-Parker J, Bours D, AM, Levy DE, Mo G, Ceccarelli M, Jones RL. Circulating tumor
Michon J, Delattre O. Société Française de Pédiatrie. Increased risk cells and biomarker modulation with Olaratumab monotherapy fol-
of systemic relapses associated with bone marrow micrometastasis lowed by Olaratumab plus doxorubucin : phase Ib study in patients
and circulating tumor cells in localized Ewing tumor. J Clin Oncol. with soft-tissue sarcoma. Mol Cancer Ther. 2021;20:132–41.
2003;21(1):85–91. 36. Satelli A, Brownlee Z, Mitra A, Meng QH, Li S. Circulating tumor
21. Avigad S, Cohen IJ, Ziberstein J, Liberzon E, Goshen Y, Ash S, cell enumeration with a combination of epithelial cell adhesion
Meller I, Kollender Y, Issakov J, Zaizov YI. The predictive poten- molecule- and cell-surface vimentin-based methods for monitoring
tial of molecular detection in the nonmetastatic Ewing family of breast cancer therapeutic response. Clin Chem. 2015;61(1):259–66.
tumors. Cancer. 2004;100(5):1053–8. 37. Chen M, Zhao H. Next-generation sequencing in liquid biopsy:
22. Fagnou C, Michon J, Peter M, Bernoux A, Oberlin O, Zucker JM, cancer screening and early detection. Hum Genomics. 2019;13:34.
Magdelenat H, Delattre O. Presence of tumor cells in bone marrow 38. Volckmar AL, Sültmann H, Riediger A, Fioretos T, Schirmacher P,
but not in blood is associated with adverse prognosis in patients Endris V, Stenzinger A, Dietz S. A field guide for cancer diagnos-
with Ewing’s tumor. Société Française d'Oncologie Pédiatrique. J tics using cell-free DNA: from principles to practice and clinical
Clin Oncol. 1998;16(5):1707–11. applications. Genes Chromosomes Cancer. 2018;57:123–39.
23. Peter M, Madgelenat H, Michon J, Melot T, Oberlin O, Zucker JM, 39. Coombs CC, Dicherber T, Crompton BD. Chasing ctDNA in patients
Thomas G, Delattre O. Sensitive detection of occult Ewing’s cells with sarcoma. Am Soc Clin Oncol Educ Book. 2020;40:e351–60.
by the reverse transcriptase-polymerase chain reaction. Br J Cancer. 40. Namlos HM, Boye K, Meza-Zepeda LA. Cell-free DNA in blood as
1995;72(1):96–100. a noninvasive insight into the sarcoma genome. Mol Aspect Med.
24. Zoubek A, Ladenstein R, Windhager R, Amann G, Fischmeister G, 2020;72:100827.
Kager L, Jugovic D, Ambros PF, Gadner H, Kovar H. Predictive 41. Barris DM, Weiner SB, Dubin RA, Fremed M, Zhang X, Piperdi S,
potential of testing for bone marrow involvement in Ewing tumor Zhang W, Maqbool S, Gill J, Roth M, Hoang B, Geller D, Gorlick
patients by RT-PCR: a preliminary evaluation. Int J Cancer. R, Weiser DA. Detection of circulating tumor DNA in patients with
1998;79(1):56–60. osteosarcoma. Oncotarget. 2018;9:12695–704.
25. Satelli A, Mitra A, Cutrera JJ, Devarie M, Xia X, Ingram DR, Dibra 42. Gutteridge A, Rathbone VM, Gibbons R, Bi M, Archard N, Davies
D, Somaiah N, Torres KE, Ravi V, Ludwig JA, Kleinerman ES, Li KEJ, Brown J, Plagnol V, Pillay N, Amary F, O’Donnell GM,
S. Universal marker and detection tool for human sarcoma circulat- Tirabisco R, Flanagan AM, Forshew T. Digital PCR analysis of
ing tumor cells. Cancer Res. 2014;74(6):1645–50. circulating tumor DNA: a biomarker for chondrosarcoma diagno-
128 C. Jubelin et al.

sis, prognostication, and residual disease detection. Cancer Med. 52. Snyder MW, Kircher M, Hill AJ, Daza RM, Shendure J. Cell-free
2017;6:2194–202. DNA comprises an in vivo nucleosome footprint that informs its
43. Shukla NN, Patel JA, Magnan H, Zehir A, You D, Tang J, Meng tissues-of-origin. Cell. 2016;164:57–68.
F, Samoila A, Slotkin EK, Ambati SR, Chou AJ, Wexler LH, 53. Ulz P, Auer M, Heitzer E. Detection of circulating tumor DNA in
Meyers PA, Peerschke EI, Viale A, Berger MF, Ladanyi M. Plasma the blood of cancer patients: an important tool in cancer chemopre-
DNA-based molecular diagnosis, prognostication, and monitor- vention. Methods Mol Biol. 2016;1379:45–68.
ing of patients with EWSR1 fusion-positive sarcomas. JCO Precis 54. Riggi N, Knoechel B, Gillespie SM, Rheinbay E, Boulay G, Suvà
Oncologia. 2017;2017:PO.16.00028. ML, Rossetti NE, Boonseng WE, Oksuz O, Cook EB, Formey A,
44. Boonstra PA, Ter Elst A, Tibbesma M, Bosman LJ, Mathijssen R, Patel A, Gymrek M, Thapar V, Deshpande V, Ting DT, Hornicek
Atrafi F, van Coevorden F, Steeghs N, Farag S, Gelderblom H, van FJ, Nielsen GP, Stamenkovic I, Aryee MJ, Bernstein BE, Rivera
der Graaf WTA, Desar OME, Maier J, Overbosch J, Suurmeijer MN. EWS-FLI1 utilizes divergent chromatin remodeling mecha-
AJH, Gitema J, Schuuring E, Reuners AKL. A single digital drop- nisms to directly activate or repress enhancer elements in Ewing
let PCR assay to detect multiple KIT exon 11 mutations in tumor sarcoma. Cancer Cell. 2014;26(5):668–81.
and plasma from patients with gastrointestinal stromal tumors. 55. Sheffield NC, Pierron G, Klughammer J, Datlinger P, Schönegger
Oncotarget. 2018;9:13870–83. A, Schuster M, Hadler J, Surdez D, Guillemot D, Lapouble E,
45. Eastley NC, Ottolini B, Neumann R, Luo JL, Hastings RK, Khan I, Freneaux P, Champigneulle J, Bouvier R, Walder D, Ambros IM,
Moore DA, Esler CP, Shaw JA, Royle NJ, Ashford RU. Circulating Hutter C, Sorz E, Amaral AT, de Alava E, Schallmoser K, Strunk D,
tumour-derived DNA in metastatic soft tissue sarcoma. Oncotarget. Rinner B, Liegl-Atzwanger B, Huppertz B, Leithner A, de Pinieux
2018;9:10549–60. G, Terrier P, Laurence V, Michon J, Ladenstein R, Holter W,
46. Namløs HM, Boye K, Mishkin SJ, Barøy T, Lorenz S, Bjerkehagen Windhager R, Dirksen U, Ambros PF, Delattre O, Kover H, Bock
B, Stratford EW, Munthe E, Kudlow BA, Myklebost O, Meza-­ C, Tomazou EM. DNA methylation heterogeneity defines a disease
Zepeda LA. Noninvasive detection of ctDNA reveals intratumor spectrum in Ewing sarcoma. Nat Med. 2017;23:386–95.
heterogeneity and is associated with tumor burden in gastrointesti- 56. Tomazou EM, Sheffield NC, Schmidl C, Schuster M, Schönegger
nal stromal tumor. Mol Cancer Ther. 2018;17:2473–80. A, Datlinger P, Kubicek S, Bock C, Kovar H. Epigenome mapping
47. Ogino S, Konishi H, Ichikawa D, Hamada J, Shoda K, Arita T, reveals distinct modes of gene regulation and widespread enhancer
Komatsu S, Shiozaki A, Okamoto K, Yamazaki S, Yaskuwa S, reprogramming by the oncogenic fusion protein EWS-FLI1. Cell
Konishi E, Otsuji E. Detection of fusion gene in cell-free DNA Rep. 2015;10:1082–95.
of a gastric synovial sarcoma. World J Gastroenterol. 2018;24: 57. Keller L, Pantel K. Unravelling tumour heterogeneity by single-­
949–56. cell profiling of circulating tumour cells. Nat Rev Cancer.
48. Shulman DS, Klega K, Imamovic-Tuco A, Clapp A, Nag A, 2019;19(10):553–67.
Thorner AR, Van Allen E, Ha G, Lessnick SL, Gorlick R, et al. 58. Brown HK, Tellez-Gabriel M, Cartron PF, Vallette FM, Heymann
Detection of circulating tumour DNA is associated with inferior MF, Heymann D. Characterization of circulating tumor cells as a
outcomes in Ewing sarcoma and osteosarcoma: a report from reflection of the tumor heterogeneity: myth or reality? Drug Discov
the Children’s Oncology Group. Br J Cancer. 2018;2018(119): Today. 2019;24(3):763–72.
615–21. 59. Paoletti C, Cani AK, Larios JM, Hovelson DH, Aung K, Darga EP,
49. Maier J, Lange T, Kerle I, Specht K, Bruegel M, Wickenhauser Cannell EM, Baratta PJ, Liu CJ, Chu D, Yazdani M, Blevins AR,
C, Jost P, Niederwieser D, Peschel C, Duyster J, von Bubnoff Sero V, Tokudome N, Thomas DG, Gersch C, Schott AF, Wu YM,
N. Detection of mutant free circulating tumor DNA in the Lonigro R, Robinson DR, Chinnaiyan AM, Bischoff FZ, Johnson
plasma of patients with gastrointestinal stromal tumor harbor- MD, Park BH, Hayes DF, Rae JM, Tomlins SA. Comprehensive
ing activating mutations of CKIT or PDGFRA. Clin Cancer Res. mutation and copy number profiling in archived circulating breast
2013;19:4854–67. cancer tumor cells. Cancer Res. 2018;78(4):1110–22.
50. Yoo C, Ryu MH, Na YS, Ryoo BY, Park SR, Kang YK. Analysis 60. Wu ZJ, Tan JC, Qin X, Liu B, Yuan ZC. Significance of circulating
of serum protein biomarkers, circulating tumor DNA, and dovitinib tumor cells in osteosarcoma patients treated by neoadjuvant chemo-
activity in patients with tyrosine kinase inhibitor-refractory gastro- therapy and surgery. Cancer Manag Res. 2018;10(10):3333–9.
intestinal stromal tumors. Ann Oncol. 2014;25:2272–7. 61. Tellez-Gabriel M, Heymann MF, Heymann D. Circulating tumor
51. Krumbholz M, Hellberg J, Steif B, Bäuerle T, Gillmann C, Fritscher cells as a tool for assessing tumor heterogenetity. Theranostics.
T, Agaimy A, Frey B, Juengert J, Wardelmann E, Hartmann W, 2019;9(16):4580–94.
Juergens H, Dirksen U, Metzler M. Genomic EWSR1 fusion 62. Tellez-Gabriel M, Cochonneau D, Cadé M, Jubellin C, Heymann
sequence as highly sensitive and dynamic plasma tumor marker in MF, Heymann D. Circulating tumor cell-derived pre-clinical mod-
Ewing sarcoma. Clin Cancer Res. 2016;22:4356–65. els for personalized medicine. Cancers (Basel). 2018;11(1):19.
 Liquid Biopsy Using Cell-Free Tumor
DNA for Gastrointestinal Cancers 13
Takafumi Nakano, Tadashi Abe, Seiichiro Takao,
Hideyuki Saito, Takaaki Masuda, and Koshi Mimori

Abstract 13.1 Introduction

Since the advent of precision medicine in 2015, personal-
ized genomic profiles of mutated DNA from tumor tissues According to global epidemiological data, the mortality rate
have been applied for the selection of compatible drugs related to gastrointestinal (GI) tract malignancies is increas-
for the clinical treatment of genomic aberrations. As a ing [1]. We believe that early diagnosis of postoperative
more convenient and less invasive method of precision recurrence, even as minimal residual disease (MRD), will
medicine, liquid biopsy that analyzes mutated cell-free improve clinical outcomes. In current clinical practice, we
tumor DNA (ctDNA) can be used as a biomarker for post- measure conventional serum tumor markers, such as carcino-
operative recurrence of genomic aberrations and their embryonic antigen, carbohydrate antigen 19-9, and squa-
appropriate treatment. In this chapter, we explore liquid mous cell carcinoma antigen. However, due to the limited
biopsies for gastrointestinal (GI) cancers in detail. We detectability of these cell surface proteins, we cannot use
review the literature on ctDNA-related issues, including them as tumor markers for early subclinical diagnosis; thus,
the mechanism of ctDNA identification in plasma, the there is a need for minimally invasive, convenient, and pre-
methodology for identifying ctDNA in profiles and tar- cise markers and approaches for cancer diagnosis.
geted sequencing assays, intratumor heterogeneity, recent A precision medicine initiative was launched and imple-
technological innovations in ctDNA analysis in GI can- mented following the 2015 State of the Union Address by the
cers, and the limitations of current approaches. We also former President of the United States Barack Obama. Liquid
describe the clinical aspects of GI cancer, such as action- biopsy using mutated cell-free tumor DNA (ctDNA) analysis
able targets and clinical trials using mutated ctDNA anal- should be the recommended method for precision medicine,
ysis, the significance of longitudinal observation of as it can identify early markers of postoperative recurrence
ctDNA for clonal evolution after targeted therapy, and the and actionable targets for appropriate drugs. A recent clinical
detection of minimal residual disease by ctDNA analysis. trial in Japan revealed that the outcomes of metastatic
Finally, we summarize a few recent studies on ctDNA in colorectal cancer (CRC) patients with actionable alterations
colorectal, gastric, and esophageal cancers. (192) were better than those without them [2].
In this chapter, we review the following aspects of ctDNA
as it evolves into the most clinically feasible approach for
precision medicine in GI cancers: (1) the basic knowledge of
Learning Objectives
ctDNA and its mechanism in the circulating system, (2) clin-
• Identification of origin of clones excreting ctDNA.
ical applications of ctDNA, and (3) recent studies on ctDNA
• We could realize the difference in detectability of
in GI cancers.
mutations in ctDNA among cancers.
• Clinical significance is equivalent between tissue
and ctDNA, i.e., actionable targets and prognosis.

T. Nakano · T. Abe · S. Takao · H. Saito · T. Masuda

K. Mimori (*)
Kyushu University Beppu Hospital, Beppu, Japan
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 129
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_13
130 T. Nakano et al.

13.1.1 B
 asic Knowledge of ctDNA in GI mutated genes in ctDNA in all samples was significantly
Cancers associated with the expression of regulatory T cells and
exhausted T cell-related genes in metastatic sites. This para-
13.1.1.1  eneral Statement of Liquid Biopsy
G doxical finding was due to increased immunogenicity caused
in GI Cancers by upregulated immune-resistant molecules in metastatic
Liquid biopsies have enabled the collection of information sites.
regarding malignancies using blood or fluids from cancer
patients without using invasive procedures. This methodol- 13.1.1.3  wo Major Approaches
T
ogy was developed gradually to improve its diagnostic accu- for the Detection of Mutated ctDNA
racy to be at par with conventional biopsy procedures. During
the metastasis of cancer cells from primary locations, there 1. Profiling of mutated ctDNA by exome sequencing:
are several methods to obtain diverse data reflecting the met- Broad coverage assays use NGS-based approaches and
astatic processes, including circulating tumor cells, cell-free have the capability to detect many variants in multiple
DNA, cancer-specific transcripts, various noncoding RNAs, genes, often over 50 genes, and can be applied to different
and exosomes. These cancer-specific data can evaluate post- tumor types. In CRC cases, primary advanced cancer is
operative recurrence and susceptibility to cancer therapy. derived from numerous minor mutations that result in
The discovery of ctDNA has its origins in CRC. ctDNA in intratumor heterogeneity. However, driver mutations are
circulating fluid is considered the most informative and expressed ubiquitously in primary tumors in a neutral
robust material for the diagnosis of cancers at the subclinical manner [10], thereby reflecting major pathogenic action-
level. In 1948, Mandel and Métais reported “cell-free DNA” able drivers to be diagnosed or cured by molecular tar-
(cfDNA). Sidransky et al. identified KRAS mutations in geted therapy.
stool samples from nine CRC patients and proposed early In contrast, precancerous lesions or early CRC can
detection of CRC in a noninvasive manner [3]. In 1994, contain several driver genes in tumors. A few bona fide
detection of KRAS-mutated DNA in the peripheral blood drivers must show positive mutations to progress through
from CRC patients enabled the identification of cancer in the Darwinian evolution [11]. Therefore, in identifying
patients using genomic aberrations in cfDNA from body flu- patients with early cancer using mutated driver ctDNA in
ids and excrements (e.g., blood, stool, and urine) [4, 5]. plasma, it cannot be determined which altered genes are
Although cfDNA analysis is unable to detect less than 10 ng actionable pathogenic drivers. It is more efficient to sur-
of cfDNA/mL of plasma in a healthy patient [6], the levels vey the whole profile of mutated ctDNA in the early
are increased in cancer patients, thereby enabling easy detec- phase of cancer.
tion [7]. The use of next-generation sequencing (NGS) can 2. Targeted sequencing for detection of cancers: Targeted
detect low levels of ctDNA; thus, the technique has pro- sequencing for a single or small number of variants is
gressed into a highly feasible method for clinical liquid aimed at greater coverage (1000–10,000×) in comparison
biopsies. to exome sequencing (100–200×). Targeted sequencing
uses a “cancer panel” of mutated driver genes from the
13.1.1.2 Circulation of ctDNA Cancer Genome Atlas Pan-Cancer Analysis Project, and a
In general, increased levels of circulating DNA fragments in small amount of ctDNA (10 ng) is amplified to obtain
the peripheral blood are derived from apoptotic and necrotic sequencing results [12]. This assay is appropriate for the
cells in tissues by pathological processes, such as malignant longitudinal analysis of a limited number of mutated
and benign lesions, inflammatory diseases, stroke, trauma, genes using personalized monitoring targets. In esopha-
and sepsis [8]. In healthy conditions, necrotic or apoptotic geal cancer, recurrence can be identified 3–6 months ear-
cells are eliminated by macrophages; however, the exponen- lier by ctDNA analysis than by any other conventional
tial growth of malignant cells simultaneously induces several method at the subclinical level [13]. In addition, ctDNA
apoptotic and necrotic cells, which form a major source of analysis has a rapid turnaround time, and results can be
ctDNA [9]. obtained earlier than tissue analysis (18 days for tumor
Recently, we identified that the detectability of ctDNA tissue analysis and 7 days for ctDNA analysis) [14]. In
was determined not by host immunity in the cancer microen- addition, several solid tumors, such as intrahepatic chol-
vironment in metastatic CRC. In our study, the mutated angiocellular carcinoma, harbor heterogeneous ­mutations;
genes in ctDNA were derived from malignant clones with an therefore, these cancers might also benefit from a panel
inactivated immune response due to a low binding affinity with personalized targets instead of using a commercial
(50 < IC50) for estimated neoantigen peptides by a calcu- multipurpose cancer panel.
lated pipeline netMHCpan even though high expression of For a specific target, a digital PCR or BEAMing assay
tumor-specific mutated RNA. In addition, the frequency of can yield high sensitivity (variant allele frequency; VAF
13 Liquid Biopsy Using Cell-Free Tumor DNA for Gastrointestinal Cancers 131

<0.1%) and specificity in the detection of altered 13.1.1.5 New Technologies for Detection
sequences. Although they are clinically feasible of ctDNA in GI Cancers (Table 13.1)
approaches with high performance, there are concerns As listed in Table 13.1, Cohen et al. reported a CancerSEEK
about contamination from nucleic acids in the air, and assay using a combination of ctDNA and protein markers to
only a limited number of mutated loci can be detected detect cancer cells in 1005 patients with non-metastatic
simultaneously. colorectal, gastric, esophageal, pancreatic, liver, lung,
breast, and ovarian cancers [20]. Cristiano et al. reported
13.1.1.4  uperiority of ctDNA Analysis
S that the DELFI method detects cancer- and organ-specific
Compared with Tissue Analysis fragment patterns using low-coverage sequencing of the
in Overcoming Intratumor whole genome with 91% sensitivity and 98% specificity in
Heterogeneity multiple cancers [21].
Considering intratumor heterogeneity, the analysis of one
biopsy specimen might pose a risk for a false-negative diag- 13.1.1.6 Limitations of ctDNA Analysis
nosis. Moreover, several studies have reported that ctDNA The precise method to determine the allele frequency of
analysis is better than tissue analysis for GI cancers. Parikh mutated ctDNA without heterogeneity among examinees
et al. described the therapeutic implications and potential and diverse preanalytical variables (e.g., timing of sampling,
advantages of cfDNA over tissue biopsy in the setting of nucleic acid extraction, transportation system, processing,
acquired resistance [15]. They suggested that a single-lesion and storage conditions) is still controversial. In addition,
tumor biopsy alone is inadequate to detect acquired resis- various patient factors might affect the results, such as
tance and that ctDNA analysis regularly identifies heteroge- inflammation, trauma, injuries, metabolites of glucose or fat,
neous, clinically relevant resistance alterations critical to the immunity, food, and lifestyle.
selection of subsequent therapy. Huang et al. reported the The ability to detect ctDNA varies among cancers; it is
clinical benefits of ctDNA assays in the management of lower in glioma, thyroid cancer, renal cell carcinoma, prostate
hepatocellular carcinoma [16, 17]. Pectasides et al. showed cancer, and medulloblastoma, while it is higher in bladder
that ctDNA sequencing demonstrated the feasibility of cancer, CRC, and gastroesophageal cancer [22]. Additional
detecting genomic amplifications not detected by tissue sam- information from protein assays was added to the mutated
pling of gastroesophageal adenocarcinoma [18]. Multiple ctDNA data in the CancerSEEK study by Cohen et al.
genomic changes have been observed during clonal evolu- According to their findings, the utility of liquid biopsy was
tion owing to the pressure of therapy in cancer, resulting in different among various malignancies; it was low in breast
acquired secondary resistance [19]. cancer, lung cancer, and CRC and high in ovarian cancer [20].

Table 13.1 Approaches of liquid biopsy using ctDNA in GI cancers

Cancer Alteration Platform Sensitivity and specificity Author
Multiple ctDNA and protein CancerSEEK assay 1005 cases Cohen et al. [20]
cancers Sensitivity: 69–98%
Specificity: 99%
HCC p16INK4 methylation in qRT-PCR Tumor burden after surgery Wong et al. [55]
ctDNA
CRC Cancer-specific methylation in Methylation-specific dPCR 1493 CRC cases Luo et al. [56]
ctDNA Forms the prediction model Sensitivity: 89.7%
Specificity: 86.8%
Multiple Evaluate the size of ctDNA Deep sequence Focus on 90–150 bp of ctDNA Mouliere et al. [57]
cancers 95% of cases show improved
analysis
Multiple Fragment pattern analysis of DELFI Sensitivity 91%; specificity 98% Cristiano et al. [21]
cancers ctDNA
Multiple Extracellular vesicle in plasma Cancer panel for proteome assay Tissue: Sensitivity 90%; specificity Hoshino et al. [58]
cancers with AI analysis 94%
Plasma: Sensitivity 95%;
specificity 90%
Multiple ctDNA Integration of variant reads High sensitivity; >one/one million Wan et al. [59]
cancers pipeline
132 T. Nakano et al.

Furthermore, somatic mutations accumulate in healthy both somatic and KRAS mutations across a 641 cancer-­
hematopoietic cells during aging, which can drive the clonal associated gene panel in a single ctDNA assay. Actionable
expansion of hematopoietic cells in the absence of dysplasia. mutations were identified in 41 of 100 patients, and 11 of
These mutations are called clonal hematopoiesis of indeter- these patients received matched therapy [34]. In Japan, a
minate potential (CHIP) [23]. CHIP may cause false-positive large cohort study was conducted in advanced GI tract can-
ctDNA results in cases where high sensitivity is required, cers to evaluate the utility of ctDNA assays compared to tis-
such as MRD and early cancer detection. sue assays [2]. In a comparison between the GI-SCREEN
study (advanced GI cancers, tumor tissue genotyping using
Oncomine Comprehensive Assay v1 (143-gene) or v3 (161-­
13.1.2 Clinical Application of ctDNA Analysis gene)) and the GOZILA study (advanced solid tumors,
including GI cancers, plasma ctDNA genotyping using
13.1.2.1 Actionable Targets in ctDNA Guardant360, a 74-gene ctDNA assay, Central lab: Guardant
for the Selection of Anticancer Agents Health (Redwood City, CA)), the trial enrollment rate in
There are studies comparing PCR-based ctDNA assays to patients with actionable alterations was significantly higher
detect RAS mutations and tissue-based RAS testing [24, 25]. in GOZILA (41/340, 12.1%) than in GI-SCREEN (85/11714,
The OncoBEAM™ RAS CRC Kit, which detects RAS muta- 5.0%) (p < 0.0001) (Nakamura et al., ASCO Annual Meeting,
tions in ctDNA derived from mCRC by a BEAMing method, Virtual, May 29–31, 2020). In addition, the objective
is the only ctDNA assay approved in Europe and Japan for response rate in tissue was 18.9% (16/85), compared to
ctDNA genomic biomarker testing in GI cancers. Analysis of 17.1% (7/41) in plasma, suggesting that the efficacy of the
ctDNA sequences detected genomic alterations in advanced matched clinical trials was equal.
GI cancers, including RAS and BRAF mutations in CRC and
HER2 amplification in esophageal cancer [26, 27]. 13.1.2.3  ongitudinal Observation of ctDNA to
L
Considering immune checkpoint inhibitor treatment, NGS Detect Clonal Evolution After Targeted
analysis of ctDNA could determine the microsatellite insta- Therapy
bility status in GI cancers, with a high concordance with Several drugs have been approved to target EGFR mutations,
tissue-­based testing [28, 29]. Putative germline BRCA1/2 but tumors inevitably develop resistance to them. In half of
mutations were found in 8 out of 332 (2.4%) patients with the patients in one study, resistance was caused by a T790M
pancreatic cancer, suggesting that ctDNA analysis can dis- mutation in EGFR; osimertinib was applied, but people who
tinguish germline mutations from somatic mutations by responded to it tended to relapse within a year [35]. Therapy
approximately 50% of the mutation allele frequency [30]. was decided by monitoring the resistance and by tracking
blood ctDNA; if there is a reduction in T790M mutation,
13.1.2.2 c tDNA Analyses in Clinical Trials of GI treatment was switched to a first-generation EGFR inhibitor;
Cancers however, if the T790M levels rise, the treatment was switched
Patient selection based on ctDNA has recently been con- back to osimertinib [35]. A similar finding was observed in
ducted. Metastatic gastroesophageal cancer with FGFR2 GI tract cancers, in which longitudinal ctDNA surveillance
amplification was evaluated by ctDNA analysis and FGFR2b detected clonal evolution with minimal invasiveness. After
overexpression by immunohistochemical analysis (phase I/ an increased number of RAS mutations, temporal withdrawal
III NCT03343301; FIGHT). In the HERACLES study, which of an anti-EGFR antibody resulted in a decline in KRAS
used trastuzumab plus lapatinib for HER2-positive mCRC, mutant allele levels, and eventually, the sensitivity to anti-­
the plasma copy number of HER2 amplification in ctDNA EGFR antibody was restored [36, 37].
was correlated with the best objective response and For HER2/neu-amplified gastric cancer cases, the ToGA
progression-­free survival [31]. In the VIKTORY study of study indicated treatment with trastuzumab in combination
gastric cancer, ctDNA analysis demonstrated a better corre- with chemotherapy. In this study, the proportion of NF1
lation between high MET copy number and response to mutations emerged to dominate ctDNA during disease pro-
savolitinib than the tissue NGS MET copy number [32]. In gression in resistant cases [26].
the CRICKET study, after longitudinal evaluation of KRAS In clinical trials, new molecule-targeting agents have been
mutant ctDNA, challenging with anti-EGFR was effective in identified by ctDNA analysis, including PARP1 inhibitors for
metastatic CRC cases with KRAS-null mutant clones [33]. BRCA1-mutated pancreatic cancer [38], TRK inhibitors for
TARGET is a molecular profiling program with the primary NTRK-mutated cases of TRK fusion-positive GI cancer [39],
aim of matching patients with a broad range of advanced and FGFR inhibitors for FGFR2 mutations in patients with
cancers to early-phase clinical trials based on analysis of FGFR2 fusion-positive cholangiocarcinoma [40].
13 Liquid Biopsy Using Cell-Free Tumor DNA for Gastrointestinal Cancers 133

13.1.2.4 ctDNA in MRD reported that the presence of ctDNA non-shedders is a limi-
Analysis of ctDNA for the identification of predictive tation of this method [48]. Anderson et al. identified a novel
genomic biomarkers for GI cancers has the potential to detect cancer-specific methylated DNA marker that undergoes
MRD and early cancers with no clinically evident disease. aberrant methylation during oncogenesis [49].
Most patients in whom ctDNA was detectable at the first
follow-up visit (13–56 days after surgery) had recurrence, 13.1.3.3 Esophageal Cancer
while patients with undetectable ctDNA had no recurrence In esophageal cancer cases, genomic DNA from tumors and
[41]. Tie et al. used an NGS-based method (Safe-SeqS) and ctDNA from plasma were obtained from 13 patients under-
reported that in 178 stage II CRC patients, ctDNA-mutated going treatment for newly diagnosed esophageal cancer.
cases showed 18-fold higher risk of recurrence than non-­ NGS of ctDNA in plasma was used to identify mutations in
mutated cases [42]. 53 cancer-related genes in which recurrent mutations were
previously detected in ESCC. We identified more than one
concordant somatic mutation in ctDNA and primary tumor
13.1.3 S
 ummary of Recent Studies of cfDNA samples from 10 patients (83.3%) and the ctDNA and recur-
in GI Cancers (Table 13.2) rent tumor samples from one patient (100%). Furthermore,
to identify accurate biomarkers for identifying ESCC
13.1.3.1 CRC patients, we assessed the usability of the four genes (TP53,
Bettegowda et al. conducted droplet digital PCR (ddPCR) to FAT3, MLL3, and AJUBA) in which mutations were the most
detect primary tumors and achieved high sensitivity and frequently detected. Mutations in the four genes were identi-
specificity [22]. Normanno et al. reported that the early fied with 78.9% sensitivity, 100% specificity, and 92.3%
detection of KRAS mutations in ctDNA in chemotherapy-­ accuracy in tumor and ctDNA samples from patients [13].
free patients was correlated with primary resistance to anti-­ A recent study demonstrated a proof-of-principle for
EGFR mAbs, while both the objective response rate and the ctDNA detection of biomarkers in head and neck cancers
overall survival were improved in patients harboring wild-­ [50]. Studies have also correlated the abundance of ctDNA
type RAS in ctDNA [43]. Using the OncoBEAM RAS CRC with tumor size and stage, with as much as 40% in metastatic
assay, Vidal et al. reported plasma ctDNA RAS mutation cancers but as low as 0.11% in premalignant or early-stage
analysis results for the diagnosis and treatment monitoring disease [51]. Thus, a highly sensitive detection system is
of metastatic CRC patients [25]. required for the detection of ctDNA. With the improvement
In terms of methylation of septin 9 (SEPT9), there were of NGS technologies, it is possible to employ targeted deep-­
three intriguing studies using qPCR. Warren et al. reported sequencing analysis to detect ctDNA with high sensitivity.
that the FDA granted premarket approval for the first ctDNA-­ Iwaya et al. reported that the clinical validity of ctDNA mon-
based liquid biopsy for cancer screening using methylated itoring was confirmed in 91% (31/34) of patients with ESCC
SEPT9 DNA (mSEPT9) [44]. Song et al. evaluated the per- using 31 mutated genes as a panel [52]. Azad et al. detected
formance of the SEPT9 gene methylation assay and com- ctDNA after chemoradiotherapy, which precedes radio-
pared it with other CRC screening tests in a meta-analysis graphic evidence of tumor progression, and the combined
[45]. In addition, Church et al. reported that a blood-based ctDNA and metabolic imaging analyses accurately predicted
mSEPT9 test showed that CRC signals in blood could be tumor progression in 100% of the patients [53]. Maron et al.
detected in asymptomatic average-risk individuals undergo- demonstrated that their approach using ctDNA provided
ing screening. However, the utility of the test for population clinical benefits not only in HER2/neu-amplified cases but
screening for CRC requires improved sensitivity for the also in EGFR-amplified cases [54].
detection of early and advanced cancers [46].

13.1.3.2 Gastric Cancer 13.1.4 Conclusion

Frankell et al. showed that clinical ctDNA-NGS testing holds
promise for gastroesophageal adenocarcinoma for both the In this review, we introduced a liquid biopsy using ctDNA
detection of MRD in early-stage disease and as a serial tumor analysis as the recommended powerful method for precision
marker [47]. Wang et al. monitored trastuzumab resistance to medicine to identify early markers of postoperative recur-
determine the mechanisms underlying the limited response rence and actionable targets for appropriate drugs. This
rate and rapid emergence of resistance in HER2+ metastatic approach is superior to conventional serum tumor markers.
gastric cancer (mGC) and showed that longitudinal ctDNA In this chapter, we review the several intriguing aspects of
sequencing provided novel insights into gene alterations ctDNA. At first, we summarized the basic knowledge of
underlying trastuzumab resistance in HER2+ mGC [26]. ctDNA and its mechanism in the circulating system. The
Kim et al. monitored cancer-specific rearrangements and increased levels of circulating DNA fragments in the periph-
134 T. Nakano et al.

Table 13.2 Summary of liquid biopsy using ctDNA in GI cancers

Platform Target alterations Cases Clinical application Findings Authors
Colorectal ddPCR KRAS mutation 12 Detection of primary Sensitivity: 87.2% Bettegowda
cancer tumors Specificity: 100% et al. [22]
Digital PCR KRAS and NRAS 340 Prediction of therapy RAS mut. 33/92 (35.9%) Normanno
mutations et al. [43]
qPCR SEPT9 methylation 50 Detection of primary Sensitivity: 90% Warren et al.
tumors [44]
qPCR SEPT9 methylation 2613 CRC screening tests Song et al.
cases6030 (meta-analysis) [45]
controls
qPCR SEPT9 methylation 7941 Detection of asymptomatic Sensitivity: I: 35.0%, II: Church et al.
CRC(multicenter study) 63.0%, III: 46.0%, IV: [46]
77.4%
Specificity: 91.5%
OncoBEAM™ RAS mutation 115 Diagnosis and treatment Vidal et al.
RAS CRC monitoring [25]
assay
HTSeq-count CNV 24 Diagnosis and treatment Sensitivity: Molparia et al.
monitoring 79%specificity: 100% [60]
HiSeq 2500 CNV Normal 35, Staging, diagnosis AUC I: 0.53, II: 0.54 Li et al. [61]
(Illumina) Adenoma 20 III: 0.61, IV: 0.60
CRC 80
dPCR TP53, KRAS, 22 Early detection of Sensitivity: Furuki et al.
PIK3CA, metastasis 89%specificity: 100% [62]
APC, FBXW7,
BRAF
Gastric NGS Mutation profile 1630 Frankell et al.
Cancer [47]
Target-Seq PIK3CA, ERBB2 78 Monitoring trastuzumab 35.3% 29.4% Wang et al.
resistance [26]
WGS Personalized 19 Monitoring cancer-specific Recurrence Kim et al. [48]
cancer-specific rearrangements
Rearrangements
Target-Seq Molecular tumor 173 samples Resistant to trastuzumab Sensitivity: 94% Wang et al.
Burden index (39 cases) [63]
(mTBI)
Methylation ELMO1, ZNF569, Novel methylated DNA Sensitivity: 86% Anderson
PCR C13orf18 markers (MDM) Specificity: 95% et al. [49]
CORD assay RUNX3 T:50 N:61 Diagnostic tool Sensitivity: 50 Hideura et al.
Specificity: 80.3 [64]
ddPCR HER2 copy number T:60 N:30 ddPCR-based HER2 copy Sensitivity: 73.3 Shoda et al.
number Specificity: 93.3 [65]
Esophageal Target-Seq 53 genes 13 Early detection of Sensitivity: 10/13 (83%) Ueda et al.
cancer recurrence [13]
dPCR 31 genes 34 Early intervention Validated for 91% (31/34) Iwaya et al.
[52]
CTC + ctDNA CTC/ctDNA 57 Serial monitoring for OS Ko et al. [66]
CAPP-Seq 607 genes SNV 45 ctDNA and metabolic Azad et al.
imaging [53]
Target-Seq 483 genes mutation 17 Disease biomarker Lower frequency Meng [67]
In post-surgery
ctDNA-NGS 73 genes 203 Clinical decision-making Benefit in HER2 or Maron et al.
EGFR-amplified patients [54]
Target-Seq 96 genes 38 Diagnosis of early-stage 18/38 (47%) Egyud et al.
EAC [68]
Target-Seq 54–73 genes 55 Genomic landscape ERBB2 alterations poor Kato et al.
OS [69]
Nano-hmC-­ 5hmC 5mC 150 Biomarkers for MRD Sensitivity: 3.75% Tian et al. [70]
seal Specificity: 85.71%
Method
13 Liquid Biopsy Using Cell-Free Tumor DNA for Gastrointestinal Cancers 135

eral blood are derived from apoptotic and necrotic cells in References
tissues by pathological processes. However, the abundant
amount of cell-free DNA was derived from cancer cells 1. Ferlay J, Colombet M, Soerjomataram I, et al. Estimating the global
cancer incidence and mortality in 2018: GLOBOCAN sources and
detected by exome sequencing and/or target sequencing. In
methods. Int J Cancer. 2019;144(8):1941–53.
terms of clinical applications of ctDNA, we demonstrated 2. Nakamura Y, Taniguchi H, Ikeda M, et al. Clinical utility of cir-
recent intriguing studies on ctDNA in GI cancers. culating tumor DNA sequencing in advanced gastrointestinal can-
In CRC cases, as primary tumors were comprised of cer: SCRUM-Japan GI-SCREEN and GOZILA studies. Nat Med.
2020;26(12):1859–64.
major driver genes as clonal genomic aberrations, such as
3. Sidransky D, Tokino T, Hamilton SR, et al. Identification of ras
APC, KRAS, and TP53, we can realize the earlier detection oncogene mutations in the stool of patients with curable colorectal
of recurrence and/or evaluate the MRD precisely and con- tumors. Science. 1992;256(5053):102–5.
tinuously during variant therapies by target sequencing. 4. Sorenson GD, Pribish DM, Valone FH, et al. Soluble normal and
mutated DNA sequences from single-copy genes in human blood.
However, in the respect of the early detection of primary site,
Cancer Epidemiol Biomark Prev. 1994;3(1):67–71.
susceptibility to detect cancer cells by capturing mutated 5. Vasioukhin V, Anker P, Maurice P, et al. Point mutations of the
ctDNA within mucosal layer is not enough clinically, further N-ras gene in the blood plasma DNA of patients with myelodys-
studies will be required. In GC cases, the amplification of plastic syndrome or acute myelogenous leukaemia. Br J Haematol.
1994;86(4):774–9.
HER2/neu is the main driver gene to cause GC. It is worth
6. Mouliere F, El Messaoudi S, Pang D, et al. Multi-marker analy-
describing that some cases were applicable to Trastuzumab, sis of circulating cell-free DNA toward personalized medicine for
which was determined not by tissue biopsy analysis but by colorectal cancer. Mol Oncol. 2014;8(5):927–41.
ctDNA sequencing. This discrepancy will be provoked by 7. Spindler KL, Appelt AL, Pallisgaard N, et al. Cell-free DNA in
healthy individuals, noncancerous disease and strong prognostic
the heterogeneity of the HER2 expression in tissues, while
value in colorectal cancer. Int J Cancer. 2014;135(12):2984–91.
the liquid biopsy can overcome the genomic diversity in tis- 8. Schwarzenbach H, Hoon DS, Pantel K. Cell-free nucleic acids as
sues. In esophageal cancer cases, just four genes (TP53, biomarkers in cancer patients. Nat Rev Cancer. 2011;11(6):426–37.
FAT3, MLL3, and AJUBA) were the most frequently detected. 9. Jahr S, Hentze H, Englisch S, et al. DNA fragments in the blood
plasma of cancer patients: quantitations and evidence for their origin
If we improve the clinical outcome in the treatment of the
from apoptotic and necrotic cells. Cancer Res. 2001;61(4):1659–65.
current intractable disease, the early diagnosis and treatment 10. Uchi R, Takahashi Y, Niida A, et al. Integrated multiregional analy-
must be required. However, the predisposition of the affected sis proposing a new model of colorectal cancer evolution. PLoS
esophageal mucosa was comprised of NOTCH1, NFE2L2, Genet. 2016;12(2):e1005778.
11. Saito T, Niida A, Uchi R, et al. A temporal shift of the evolutionary
CDKN2A, and FBXW7, which were different from advanced
principle shaping intratumor heterogeneity in colorectal cancer. Nat
cancers, and we have to bespoke the appropriate cancer panel Commun. 2018;9(1):2884.
for target sequencing. 12. Reinert T, Scholer LV, Thomsen R, et al. Analysis of circulating
However, there are inherent limitations to the current tumour DNA to monitor disease burden following colorectal cancer
surgery. Gut. 2016;65(4):625–34.
methodologies. Therefore, clinical oncologists, including
13. Ueda M, Iguchi T, Masuda T, et al. Somatic mutations in plasma
surgeons, should consider the specific favorable and unfavor- cell-free DNA are diagnostic markers for esophageal squamous cell
able traits of a given assay before applying it to clinical carcinoma recurrence. Oncotarget. 2016;7(38):62,280–91.
studies. 14. Bachet JB, Bouche O, Taieb J, et al. RAS mutation analysis in
circulating tumor DNA from patients with metastatic colorectal
In terms of the future of this field, we would like to
cancer: the AGEO RASANC prospective multicenter study. Ann
address the following queries: Oncol. 2018;29(5):1211–9.
15. Parikh AR, Leshchiner I, Elagina L, et al. Liquid versus tissue
biopsy for detecting acquired resistance and tumor heterogeneity in
gastrointestinal cancers. Nat Med. 2019;25(9):1415–21.
Open Questions 16. Chabon JJ, Simmons AD, Lovejoy AF, et al. Circulating tumour
• How can detection sensitivity be improved to enable DNA profiling reveals heterogeneity of EGFR inhibitor resistance
the diagnosis of early-stage cancer? mechanisms in lung cancer patients. Nat Commun. 2016;7:11815.
• What is the significance of detecting epigenomic 17. Huang A, Zhang X, Zhou SL, et al. Detecting circulating tumor
DNA in hepatocellular carcinoma patients using droplet digital
aberrations in ctDNA? PCR is feasible and reflects intratumoral heterogeneity. J Cancer.
• Can we identify copy number aberrations/structural 2016;7(13):1907–14.
variants in ctDNA? 18. Pectasides E, Stachler MD, Derks S, et al. Genomic heterogeneity
• Which sample is appropriate: urine or blood? as a barrier to precision medicine in gastroesophageal adenocarci-
noma. Cancer Discov. 2018;8(1):37–48.
• What mutated genes should be monitored for longi- 19. McGranahan N, Swanton C. Biological and therapeutic impact
tudinal analysis? of intratumor heterogeneity in cancer evolution. Cancer Cell.
2015;27(1):15–26.
136 T. Nakano et al.

20. Cohen JD, Li L, Wang YX, et al. Detection and localization of sur- 40. Goyal L, Saha SK, Liu LY, et al. Polyclonal secondary FGFR2
gically resectable cancers with a multi-analyte blood test. Science. mutations drive acquired resistance to FGFR inhibition in patients
2018;359(6378):926. with FGFR2 fusion-positive cholangiocarcinoma. Cancer Discov.
21. Cristiano S, Leal A, Phallen J, et al. Genome-wide cell- 2017;7(3):252–63.
free DNA fragmentation in patients with cancer. Nature. 41. Diehl F, Schmidt K, Choti MA, et al. Circulating mutant DNA to
2019;570(7761):385–9. assess tumor dynamics. Nat Med. 2008;14(9):985–90.
22. Bettegowda C, Sausen M, Leary RJ, et al. Detection of circulating 42. Tie J, Wang Y, Tomasetti C, et al. Circulating tumor DNA analysis
tumor DNA in early- and late-stage human malignancies. Sci Transl detects minimal residual disease and predicts recurrence in patients
Med. 2014;6(224):224ra24. with stage II colon cancer. Sci Transl Med. 2016;8(346):346ra92.
23. Steensma DP, Bejar R, Jaiswal S, et al. Clonal hematopoiesis of 43. Normanno N, Esposito Abate R, Lambiase M, et al. RAS testing of
indeterminate potential and its distinction from myelodysplastic liquid biopsy correlates with the outcome of metastatic colorectal
syndromes. Blood. 2015;126(1):9–16. cancer patients treated with first-line FOLFIRI plus cetuximab in
24. Thierry AR, Mouliere F, El Messaoudi S, et al. Clinical validation the CAPRI-GOIM trial. Ann Oncol. 2018;29(1):112–8.
of the detection of KRAS and BRAF mutations from circulating 44. Warren JD, Xiong W, Bunker AM, et al. Septin 9 methylated DNA
tumor DNA. Nat Med. 2014;20(4):430–5. is a sensitive and specific blood test for colorectal cancer. BMC
25. Vidal J, Muinelo L, Dalmases A, et al. Plasma ctDNA RAS muta- Med. 2011;9:133.
tion analysis for the diagnosis and treatment monitoring of meta- 45. Song L, Jia J, Peng X, et al. The performance of the SEPT9 gene
static colorectal cancer patients. Ann Oncol. 2017;28(6):1325–32. methylation assay and a comparison with other CRC screening
26. Wang DS, Liu ZX, Lu YX, et al. Liquid biopsies to track trastu- tests: a meta-analysis. Sci Rep. 2017;7(1):3032.
zumab resistance in metastatic HER2-positive gastric cancer. Gut. 46. Church TR, Wandell M, Lofton-Day C, et al. Prospective evalua-
2019;68(7):1152–61. tion of methylated SEPT9 in plasma for detection of asymptomatic
27. Wang H, Li B, Liu Z, et al. HER2 copy number of circulating tumour colorectal cancer. Gut. 2014;63(2):317–25.
DNA functions as a biomarker to predict and monitor trastuzumab 47. Frankell AM, Smyth EC. ctDNA in gastric and gastroesophageal
efficacy in advanced gastric cancer. Eur J Cancer. 2018;88:92–100. cancer: prognostic, predictive, or preliminary? Clin Cancer Res.
28. Mayrhofer M, De Laere B, Whitington T, et al. Cell-free DNA pro- 2019;25(23):6893–5.
filing of metastatic prostate cancer reveals microsatellite instability, 48. Kim YW, Kim YH, Song Y, et al. Monitoring circulating tumor
structural rearrangements and clonal hematopoiesis. Genome Med. DNA by analyzing personalized cancer-specific rearrange-
2018;10(1):85. ments to detect recurrence in gastric cancer. Exp Mol Med.
29. Willis J, Lefterova MI, Artyomenko A, et al. Validation of micro- 2019;51(8):1–10.
satellite instability detection using a comprehensive plasma-based 49. Anderson BW, Suh YS, Choi B, et al. Detection of gastric cancer
genotyping panel. Clin Cancer Res. 2019;25(23):7035–45. with novel methylated DNA markers: discovery, tissue validation,
30. Slavin TP, Banks KC, Chudova D, et al. Identification of inciden- and pilot testing in plasma. Clin Cancer Res. 2018;24(22):5724–34.
tal germline mutations in patients with advanced solid tumors who 50. Wang Y, Springer S, Mulvey CL, et al. Detection of somatic muta-
underwent cell-free circulating tumor DNA sequencing. J Clin tions and HPV in the saliva and plasma of patients with head and neck
Oncol. 2018;36:JCO1800328. squamous cell carcinomas. Sci Transl Med. 2015;7(293):293ra104.
31. Siravegna G, Sartore-Bianchi A, Nagy RJ, et al. Plasma HER2 51. Perdomo S, Avogbe PH, Foll M, et al. Circulating tumor DNA
(ERBB2) copy number predicts response to HER2-targeted detection in head and neck cancer: evaluation of two different
therapy in metastatic colorectal cancer. Clin Cancer Res. detection approaches. Oncotarget. 2017;8(42):72621–32.
2019;25(10):3046–53. 52. Iwaya T, Endo F, Takahashi F, et al. Frequent tumor burden moni-
32. Lee J, Kim ST, Kim K, et al. Tumor genomic profiling guides toring of esophageal squamous cell carcinoma with circulating
patients with metastatic gastric cancer to targeted treatment: the tumor DNA using individually designed digital polymerase chain
VIKTORY umbrella trial. Cancer Discov. 2019;9(10):1388–405. reaction. Gastroenterology. 2021;160:463–465.e4.
33. Cremolini C, Antoniotti C, Lonardi S, et al. Activity and safety of 53. Azad TD, Chaudhuri AA, Fang P, et al. Circulating tumor
Cetuximab plus modified FOLFOXIRI followed by maintenance DNA analysis for detection of minimal residual disease
with Cetuximab or bevacizumab for RAS and BRAF wild-type after Chemoradiotherapy for localized esophageal cancer.
metastatic colorectal cancer: a randomized phase 2 clinical trial. Gastroenterology. 2020;158(3):494–505 e6.
JAMA Oncol. 2018;4(4):529–36. 54. Maron SB, Chase LM, Lomnicki S, et al. Circulating tumor DNA
34. Rothwell DG, Ayub M, Cook N, et al. Utility of ctDNA to support sequencing analysis of gastroesophageal adenocarcinoma. Clin
patient selection for early phase clinical trials: the TARGET study. Cancer Res. 2019;25(23):7098–112.
Nat Med. 2019;25(5):738–43. 55. Wong IH, Zhang J, Lai PB, et al. Quantitative analysis of tumor-­
35. Willyard C. Cancer therapy: an evolved approach. Nature. derived methylated p16INK4a sequences in plasma, serum, and
2016;532(7598):166–8. blood cells of hepatocellular carcinoma patients. Clin Cancer Res.
36. Morelli MP, Overman MJ, Dasari A, et al. Characterizing the pat- 2003;9(3):1047–52.
terns of clonal selection in circulating tumor DNA from patients 56. Luo H, Zhao Q, Wei W, et al. Circulating tumor DNA methylation
with colorectal cancer refractory to anti-EGFR treatment. Ann profiles enable early diagnosis, prognosis prediction, and screening
Oncol. 2015;26(4):731–6. for colorectal cancer. Sci Transl Med. 2020;12(524):eaax7533.
37. Parseghian CM, Loree JM, Morris VK, et al. Anti-EGFR-resistant 57. Mouliere F, Chandrananda D, Piskorz AM, et al. Enhanced detec-
clones decay exponentially after progression: implications for anti-­ tion of circulating tumor DNA by fragment size analysis. Sci Transl
EGFR re-challenge. Ann Oncol. 2019;30(2):243–9. Med. 2018;10(466):eaat4921.
38. Shroff RT, Hendifar A, McWilliams RR, et al. Rucaparib mono- 58. Hoshino A, Kim HS, Bojmar L, et al. Extracellular vesicle
therapy in patients with pancreatic cancer and a known deleterious and particle biomarkers define multiple human cancers. Cell.
BRCA mutation. JCO Precis Oncologia. 2018;2018:PO.17.00316. 2020;182(4):1044–61 e18.
39. Russo M, Misale S, Wei G, et al. Acquired resistance to the 59. Wan JCM, Heider K, Gale D, et al. ctDNA monitoring using
TRK inhibitor Entrectinib in colorectal cancer. Cancer Discov. patient-specific sequencing and integration of variant reads. Sci
2016;6(1):36–44. Transl Med. 2020;12(548):eaaz8084.
13 Liquid Biopsy Using Cell-Free Tumor DNA for Gastrointestinal Cancers 137

60. Molparia B, Oliveira G, Wagner JL, et al. A feasibility study of 66. Ko JMY, Ng HY, Lam KO, et al. Liquid biopsy serial monitoring of
colorectal cancer diagnosis via circulating tumor DNA derived treatment responses and relapse in advanced esophageal squamous
CNV detection. PLoS One. 2018;13(5):e0196826. cell carcinoma. Cancers (Basel). 2020;12(6)
61. Li J, Dittmar RL, Xia S, et al. Cell-free DNA copy number varia- 67. Meng P, Wei J, Geng Y, et al. Targeted sequencing of circulating
tions in plasma from colorectal cancer patients. Mol Oncol. cell-free DNA in stage II-III resectable oesophageal squamous cell
2017;11(8):1099–111. carcinoma patients. BMC Cancer. 2019;19(1):818.
62. Furuki H, Yamada T, Takahashi G, et al. Evaluation of liquid biop- 68. Egyud M, Tejani M, Pennathur A, et al. Detection of circulating
sies for detection of emerging mutated genes in metastatic colorec- tumor DNA in plasma: a potential biomarker for esophageal adeno-
tal cancer. Eur J Surg Oncol. 2018;44(7):975–82. carcinoma. Ann Thorac Surg. 2019;108(2):343–9.
63. Wang Y, Zhao C, Chang L, et al. Circulating tumor DNA analy- 69. Kato S, Okamura R, Baumgartner JM, et al. Analysis of circulating
ses predict progressive disease and indicate trastuzumab-­ tumor DNA and clinical correlates in patients with esophageal, gas-
resistant mechanism in advanced gastric cancer. EBioMedicine. troesophageal junction, and gastric adenocarcinoma. Clin Cancer
2019:43261–9. Res. 2018;24(24):6248–56.
64. Hideura E, Suehiro Y, Nishikawa J, et al. Blood free-circulating 70. Tian X, Sun B, Chen C, et al. Circulating tumor DNA
DNA testing of methylated RUNX3 is useful for diagnosing early 5-­
hydroxymethylcytosine as a novel diagnostic biomarker for
gastric cancer. Cancers (Basel). 2020;12(4) esophageal cancer. Cell Res. 2018;28(5):597–600.
65. Shoda K, Ichikawa D, Fujita Y, et al. Monitoring the HER2 copy
number status in circulating tumor DNA by droplet digital PCR in
patients with gastric cancer. Gastric Cancer. 2017;20(1):126–35.
 Circulating Tumor Cells and ctDNA
in Prostate Cancer 14
Nikolas H. Stoecklein and Rui P. L. Neves

Abstract 14.1 Introduction

The rapidly expanding treatment options for patients with
castration resistant prostate cancer (mCRCP) are promis- In men, prostate cancer (PC) is the most common non-­
ing. However, there is a need for biomarkers to help clini- cutaneous cancer in Western countries and responsible for
cians delivering the right treatment at the right time and around 10% of all cancer-related deaths. PC therapy is
sequence in the final stage of this common male cancer. adapted to the stage of the disease. PC begins as an organ-­
As the majority of patients present with bone metastases, confined cancer, and evolves through a state displaying ris-
high-quality tissue biopsies for molecular diagnostics are ing levels of prostate-specific antigen (PSA) [1] until
often not easy to obtain and are not suitable for longitudi- reaching a metastatic state. Since androgens are major stimu-
nal testing. In this situation, CTC and ctDNA assays can lants of PC cell growth, PCs are hormone-sensitive and
seamlessly complement tissue biopsies and provide clini- androgen-deprivation therapies (ADT) have been the corner-
cally relevant information for rational therapeutic stone therapies during these initial PC phases [2]. Finally, PC
decision-­making in real time can capture the evolutionary enters a state termed metastatic castration-resistant PC
dynamics of the disease. In fact, the first liquid biopsy (mCRCP), in which the cancer cells become unresponsive to
assays started to enter clinical routine diagnostics and it is previous ADT, and that is lethal for most men. However, dur-
conceivable that this development continues. However, in ing the last decade, the development of a new generation of
view of the low concentration of the liquid biopsy ana- drugs further interfering with the androgen signaling axis
lytes in blood, their detection and characterization remain has led to significantly prolonged survival in PC patients pre-
challenging and newly developed assays require rigorous viously unresponsive to ADT [3, 4]. It was reported that
technical and clinical validation. together with earlier hormonal and chemotherapy adminis-
tration, the novel inhibitors of the androgen axis could
improve survival up to 5 years [5]. Nevertheless, while the
Learning Objectives 5-year survival rates of non-metastatic PC are nearly 100%,
• To be able to define the major applications of CTC metastatic PC, and especially mCRCP, remain incurable and
and ctDNA liquid biopsies in PC. lethal with overall 5-year survival rates of less than 30% [6].
• To understand the current state of CTC-based and However, with rapidly expanding therapeutic options for
ctDNA-based liquid biopsies for clinical decision-­ mCRCP patients there is realistic hope to prolong survival.
making in prostate cancer. Besides androgen receptor signaling inhibitors (ARSi; e.g.
• To be able to describe in which clinical situations abiraterone and enzalutamide), drugs targeting mitotic-­
the CTC and ctDNA tests are currently used in PC. spindle microtubules (docetaxel, cabazitaxel), radioactive
calcium mimetics targeting bone metastases (radium-223),
stimulators of the immune system (sipuleucel-T), or inhibi-
tors of the poly ADP-ribose polymerase (PARP; olaparib), an
enzyme involved in DNA repair, have become available for
N. H. Stoecklein (*) · R. P. L. Neves
mCRPC patients [3, 7]. However, selecting the patients that
Experimental Surgical Oncology, Department of General, Visceral
and Pediatric Surgery, University Hospital and Medical Faculty of benefit from these therapies is still challenging [8]. This is
the Heinrich-Heine University Düsseldorf, Düsseldorf, Germany further complicated by the heterogenous nature of PC and
e-mail: [email protected]; the evolutionary dynamics of the disease [9–11]. Selective
[email protected]­duesseldorf.de

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 139
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_14
140 N. H. Stoecklein and R. P. L. Neves

pressures of systemic therapies further shape genetic diver- early as week four is associated with longer survival [23]. A
sity of tumor cells and, consequently, the response to subse- recent re-analysis using individual patient data of five pro-
quent therapies [12]. In fact, recent data indicated that one spective randomized phase III trials, which enrolled a total of
dominant PC cell lineage emerges, which initiates and drives 6081 patients (COU-AA-301, AFFIRM, ELM-PC-5,
systemic progression. Nonetheless, subsequent polyclonal ELM-PC-4, and COMET-1), established a negative CTC
seeding between different metastatic sites appears to be com- CellSearch test (CTC0) as easily recognizable and clinically
mon [9, 13]. Importantly, the potential genetic drivers of meaningful endpoint. Since this outcome can be determined
metastasis distinguish the metastatic subclones from those shortly after the treatment initiation and indicating benefit of
remaining locally confined in the prostate. In this context, it the therapy, the use of CTC0 as a response endpoint was rec-
is interesting to note that the metastatic lineage appears to be ommended for early-phase clinical trials [24]. However, the
represented in liquid biopsy samples [13]. This suggests that predictive value (i.e., information on the probability of
important information on molecular characteristics of the response to a particular therapy) of CTC enumeration when
cancer can be provided by liquid biopsies in real time, com- measured before treatment could not be established yet,
plementary to routine tissue biopsies [14]. This can be par- which limits currently the routine use in the clinic, despite its
ticularly important, to capture evolutionary developments by proven strong prognostic value.
longitudinal sampling, which is barley possible with stan-
dard tissue biopsies. Another aspect is that approximately
90% of the metastatic PC patients suffer from bone metasta- 14.2.2 Molecular Characterization
ses [15]. Bone biopsies can be uncomfortable and contain of CTCs in PC
frequently (in up to 20% of the cases) material insufficient
for molecular testing [16]. Thus, given the increasing treat- Molecular characterization of enriched or individual CTCs is
ment options and the associated challenges of selecting the versatile and the potential to provide information for rational
right treatment at the right time and in the right sequence, the therapeutic decision-making is high [25]. In this regard, further
less invasive liquid biopsies have the potential to improve the phenotypic characterization of CTCs with the CellSearch plat-
care of metastatic PC patients. The goal of this section is to form has been applied in PC to assess potential therapeutic
provide an overview on the status of CTC and ctDNA analy- targets on CTCs as pharmacodynamic endpoint in early-­phase
sis in advanced PC, with a focus on liquid biopsies closest to clinical trials (e.g., [26]) or for prediction (e.g., for PSMA,
routine use. [27]). In terms of therapeutic prediction, the androgen receptor
splice variant-7 (AR-V7) has become a major focus of interest.
Resistance to ARSi is often associated with elevated expres-
14.2 CTCs in Prostate Cancer sion of both full-length AR (AR-FL) and AR-V7, encoding a
truncated protein lacking the C-terminal ligand-binding
14.2.1 Enumeration of Circulating Tumor domain, which can lead to ligand-­independent transcriptional
Cells in Prostate Cancer activity [28–30]. Detecting AR-V7 in CTCs might therefore be
helpful to predict ARSi resistance and improve therapy selec-
Metastatic prostate cancer is one of three indications tion in mCRCP. While several methods for detection of
(together with breast and colorectal cancer) for which the AR-V7 in CTCs have been established [1], the initial study
CellSearch system was cleared by the US Food and Drug was based on the Adna-Test platform: a multiplexed reverse-
Administration (FDA) for therapeutic monitoring. Since transcription polymerase-chain-reaction (qRT-PCR) with
back then in 2008, the CellSearch system has been techni- primers to detect the presence of CTCs as well as custom prim-
cally unchanged and the only CTC detection platform with ers designed to detect mRNAs for AR-FL and AR-V7 [31].
regulatory clearance. The focus of most CellSearch studies This detection assay does not allow CTC enumeration since it
in PC Patients was mCRPC, in which the prognostic impact is based on mRNA expression of blood cells enriched for
of CellSearch CTCs was well established through multiple CTCs by magnetic particles conjugated with a combination of
large phase II and III trials. In PC, unfavorable counts of antibodies recognizing prostate cancer cells. However, the data
CTCs (≥5/7.5 mL blood) before treatment or acquired dur- of the initial study indicated the AR-V7 CTC signal to be asso-
ing treatment correlate with reduced disease-free and overall ciated with resistance to ARSi. This was validated in a larger
survival [17–21]. In combination with increased serum lac- mCRPC patient cohort of 202 patients. In this study, the
tate dehydrogenase (LDH, a biomarker indicating aggressive mRNA signal was corrected for the presence of CellSearch
malignancy) at week 12 of treatment, CTC counts fulfilled CTCs and it could be demonstrated that CTC AR-V7 was
all four Prentice criteria as an individual patient-level surro- prognostically independent of the CTC enumeration by
gate for survival [22]. For patients with ≥5 CTCs in 7.5 mL CellSearch [32]. Another far-­ developed method to assess
blood before therapy, a ≥30% decline in the CTC counts as AR-V7 positive CTCs was established on Epic sciences “no
14 Circulating Tumor Cells and ctDNA in Prostate Cancer 141

cell left behind” platform. This automated image-based assay Similar comprehensive single-cell profiling studies in PC
can identify individual CTCs with nuclear localization of patients on CTCs isolated with different CTC-methods have
AR-V7. Two studies indicated clinical utility of the Epic been performed with next-generation sequencing (NGS)
Sciences assay in mCRPC patients undergoing treatment including whole-exome sequencing (WES), whole-genome
change to select patients for therapy with taxanes instead of sequencing (WGS), and RNA sequencing. Collectively,
ARSi [33, 34]. Notably, despite the lack of a randomized con- these studies demonstrated the feasibility of comprehensive
trolled trial to establish the predictive value for selection of CTC sequencing on the single-cell level with implications
taxanes vs. ARSi, the federal US health insurance program for potential clinical diagnostics and to identify new bio-
Medicare covers now the AR-V7 protein assay, commercial- markers [40].
ized as Oncotype DX AR-V7 Nucleus Detect test, for mCRCP
patients [35] and will therefore likely enter clinical practice.
Supporting its clinical application, the recent PROPHECY 14.2.3 Functional Analysis of CTCs in PC
study (NCT02269982) demonstrated that both mRNA- and
protein-­based CTC AR-V7 assays are valuable to select The versatility and potential power of CTC-based liquid biop-
mCRCP patients to ARSi or taxane treatment in the context of sies can be best illustrated by the possibility to generate cell
other clinical parameters [36]. The authors stressed that the culture models that allow functional analysis. This became a
CTC AR-V7 status can explain some of the resistance to AR major interest in the CTC-field [41], but has turned out as tech-
therapy, but AR-V7 heterogeneity within patients and over nically extremely challenging with so far only anectodical
time suggests that additional strategies are needed to overcome success in most cancer types, including PC [42, 43]. This is
resistance in such mCRCP patients. A further notion of care despite the fact that the viability of PC CTCs could be demon-
concerning the relevance of AR-V7 for ARSi resistance come strated already more than 15 years ago via the EPISPOT assay
from the so far largest sequencing effort (n = >400) in mCRPC [44]. Here, enriched CTCs are plated and short-term cultured
tumor tissue [37]. In this investigation, an increased expression on a membrane coated with antibodies to capture secreted
of AR-V7 and other AR splice variants was observed in sam- prostate-specific antigen (PSA). With this approach, CTCs
ples from patients receiving taxanes or ARSi, but this was not could be identified in metastatic as well as non-metastatic PC
correlated with therapeutic resistance or survival. (83% vs. and 42% positivity rate), apparently, with a higher
In order to establish new predictive biomarkers with the sensitivity when compared to CellSearch. More advanced
help of CTCs, broader molecular characterizations of indi- functional testing would require CTC-­derived cell lines or
vidual CTCs have been performed. The Epic Sciences plat- xenografts. An important lesson learned was that high num-
form has been used to investigate the relevance of phenotypic bers of non-apoptotic CTCs are required for successful expan-
CTC heterogeneity to predict therapeutic resistance [38]. For sion of CTCs [45, 46]. This can be enabled by harvesting
this, 9225 morphologic features of individual CTCs from 179 higher volumes of peripheral blood. Therefore, diagnostic leu-
mCRPC patients were recorded and subjected to heterogene- kapheresis (DLA) might be a key to achieve this [47]. DLA is
ity analysis applying the Shannon-Index. It was observed that based on the clinically safe leukapheresis, a method to collect
a higher CTC diversity was linked to worse overall survival in white blood cells with a certain density by continuous cen-
mCRCP patients treated with an ARSi, while survival was not trifugation from liters of peripheral blood. Since epithelial
affected by the degree of diversity in patients treated with tax- cancer cells have similar densities, CTCs become co-collected
anes [38]. The cell-to-cell heterogeneity of mCRCP CTCs during the process. DLA does not only increase dramatically
could also be observed on the genomic level. CTC detection but could indeed facilitate the generation of
CTC-derived organoid culture and xenografts in PC [39, 48].
Such DLA-derived PC organoid cultures have the potential to
Example serve as relevant pre-­clinical models for metastatic PC or even
as a personalized drug screening platform [49] in the future.
Lambros and colleagues generated genome-wide CNA
profiles from 185 individual CTCs isolated from 14
mCRCP patients via FACS after their detection by 14.3 ctDNA in Prostate Cancer
CellSearch. The authors could confirm the typical
genomic landscape of advanced prostate cancer for the Inferring genomic information from plasma samples is tech-
CellSearch detected CTCs and revealed complex intra- nically less complex compared to CTCs. Since the advent of
patient, intercellular genomic heterogeneity missed by NGS, cell-free DNA (cfDNA) analysis could have been
bulk biopsy [39]. ◄ swiftly optimized for routine diagnostic use [50]. Initial
NGS-driven cfDNA studies in PC investigated low-coverage
142 N. H. Stoecklein and R. P. L. Neves

whole-genome sequencing (WGS), also known as shallow or AKT-inhibition is used to therapeutically target PI3K-driven
low-pass WGS, and could establish tumor-specific copy cancers and data indicate that a subset of PC patients might
number alterations (CNA) from the samples [51]. Along with benefit from a combination of AKT inhibitors and anti-andro-
CNAs typical for PC it was possible to detect high-level copy gens, ctDNA analysis for PI3K/AKT pathway genes may
number gains in the AR locus in mCRPC patients, which serve as predictive assay [57]. Another set of signaling path-
might be of diagnostic use in the context of ADT resistance. ways commonly altered in mCRCP are belonging to the DNA
Since circulating tumor DNA (ctDNA) represents only a damage response (DDR). Around 25% of mCRPCs carry
fraction of cfDNA and low-pass WGS require high ctDNA mutations in genes of the DDR and this can be therapeutically
fractions, these assays are limited to metastatic PC stages in exploited. Poly (ADP-ribose) polymerase (PARP) inhibitors
which high tumor DNA concentrations in the plasma are induce programmed cell death in DDR-impaired tumor cells,
more common [52]. It is of note that higher total ctDNA lev- and monotherapy with the PARP inhibitor olaparib has antitu-
els are correlated with poor prognosis and may have a role as mor activity in mCRCP, delays significantly disease progres-
biomarker to indicate aggressive cancers or higher disease sion, and improves survival. These could be very recently
burden [53, 54]. On the other hand, dropping levels of detect- established in the TOPARP A/B [65, 66] and PROfound trials
able ctDNA during therapy are correlated with therapeutic [67]. Importantly, clinically relevant response to olaparib was
response and good prognosis [55, 56], qualifying ctDNA as associated with mutations in BRCA2, BRCA1, ATM, and
potential biomarker for therapy monitoring. PALB2 genes [65]. The DDR mutations can be well captured
Similar to the CNA profiles, ctDNA point mutations were by ctDNA analysis and were concordant with contemporane-
largely concordant with tumor tissue, indicating ctDNA as ously taken tumor tissue. Importantly, ctDNA analysis
surrogate material for tumor biopsies [57]. revealed the emergence of new mutations at disease progres-
Detection of somatic ctDNA mutations in TP53 and AR sion. Based on the available data the FDA approved On
genes, in genes of the PI3K pathway, as well as in DNA dam- November 6, 2020, a commercially available assay to identify
age repair genes might have immediate clinical importance mutations in BRCA1, BRCA2, and ATM genes [68] as a com-
for prognosis and prediction. The TP53 gene is very fre- panion diagnostic device in patients with mCRPC eligible for
quently mutated in mCRPC (around 50% of the cases) and treatment with olaparib. Thus, DDR gene testing on ctDNA
has been associated with poor prognosis [58]. In ctDNA, material became the first liquid biopsy test for therapy selec-
somatic TP53 mutations were an independent prognostic tion with regulatory approval in prostate cancer.
factor outperforming the prognostic value of AR alterations
[59]. While genetic alterations of AR are usually not detected
in primary tumors or hormone-naïve metastases, copy num- 14.4 Conclusion
ber gain of AR can be detected in mCRCP in 30% and point
mutations in approximately 20% of the tumors cases after With CTC-based AR-V7 testing and the DDR gene ctDNA
ADT [57]. Copy number gains of AR were associated with assay, liquid biopsy just became clinical reality in mCRCP, the
resistance to ARSi [60] and were associated with poor prog- lethal advanced stage of PC. With the rapidly expanding treat-
nosis [56, 61, 62]. In contrast to AR copy number gains, spe- ment armamentarium it is crucial for the clinicians to select the
cific AR point mutations (T878A or L702H) emerged in 13% right treatment at the right time in the correct sequence for the
of the cases during progression on abiraterone [53, 56]. patients. With their less invasiveness and the possibility for lon-
Although detection of AR amplifications did not outperform gitudinal testing, CTC- and ctDNA-based liquid biopsies appear
standard prognostic biomarkers in a recent study, structural to seamlessly complement classic tissue biopsies in PC. These
rearrangements truncating the ligand-binding domain of AR are often not easy to obtain in PC, as the majority of patients
were identified in several patients with primary resistance present with bone metastases. Nonetheless, as technology
[63]. While the prognostic data of AR alterations in ctDNA advances, it is conceivable that CTC-based liquid biopsies in
remain somewhat inconclusive, their detection via ctDNA particular will continue to evolve with their versatility and yield
might have clinical utility for treatment selection. more clinically relevant assays. mCRPC is also quite suitable
Approximately 50% of mCRCPs display alterations in for the implementation of such liquid biopsies, as both CTC
PIK3CA/B, PTEN, or AKT, making the phosphatidylinositol-­3 number and ctDNA concentration are often high, making the
kinase (PI3K)/AKT pathway one of the most frequently assays well applicable. However, in the early stages of the dis-
deregulated signaling pathways in PC [64]. Mutations in these ease, the low abundance of the cancer-derived analytes limits
genes are significantly more prevalent in mCRPC than in pri- currently their application. Therefore, more sensitive assays or
mary PC (50% vs. <20%) suggesting an important role of this better enrichment methods for the analytes are required. Most
pathway in PC progression [57]. The somatic mutations in the importantly, after rigorous technical and clinical validation of
PI3K pathway genes can be traced by ctDNA assays and have new liquid biopsy assays for PC patients, suitable prospective
been associated with a poor response to ARSi [63]. Since clinical trials are needed to demonstrate their clinical utility.
14 Circulating Tumor Cells and ctDNA in Prostate Cancer 143

tate cancer. J Clin Invest. 2013;123(11):4918–22. https://doi.

Open Questions org/10.1172/jci70354.
11. Kumar A, Coleman I, Morrissey C, Zhang X, True LD, Gulati R,
• Can molecular CTC profiling on single-cell level Etzioni R, Bolouri H, Montgomery B, White T, Lucas JM, Brown
improve therapeutic management in the clinical LG, Dumpit RF, DeSarkar N, Higano C, Yu EY, Coleman R,
routine? Schultz N, Fang M, Lange PH, Shendure J, Vessella RL, Nelson
• Does current technology enable the implementation PS. Substantial interindividual and limited intraindividual genomic
diversity among tumors from men with metastatic prostate cancer.
of CTC-derived organoids as individual drug-test- Nat Med. 2016;22(4):369–78. https://doi.org/10.1038/nm.4053.
ing platform? 12. Burrell RA, Swanton C. Tumour heterogeneity and the evolution
• Have CTC and ctDNA assays a role for clinical of polyclonal drug resistance. Mol Oncol. 2014;8(6):1095–111.
decision-making in earlier PC stages? https://doi.org/10.1016/j.molonc.2014.06.005.
13. Woodcock DJ, Riabchenko E, Taavitsainen S, Kankainen M,
• Can ctDNA-based liquid biopsies capture the clini- Gundem G, Brewer DS, Ellonen P, Lepistö M, Golubeva YA,
cally relevant genetic heterogeneity of the systemic Warner AC, Tolonen T, Jasu J, Isaacs WB, Emmert-Buck MR,
disease? Nykter M, Visakorpi T, Bova GS, Wedge DC. Prostate cancer evo-
• What is the best strategy to allow a fast and inde- lution from multilineage primary to single lineage metastases with
implications for liquid biopsy. Nat Commun. 2020;11(1):5070.
pendent technical and clinical validation of com- https://doi.org/10.1038/s41467-­020-­18843-­5.
mercialized CTC- and ctDNA liquid biopsy assays? 14. Alix-Panabieres C, Pantel K. Liquid biopsy: from discovery to
clinical application. Cancer Discov. 2021;11(4):858–73. https://
doi.org/10.1158/2159-­8290.CD-­20-­1311.
15. Bubendorf L, Schöpfer A, Wagner U, Sauter G, Moch H, Willi N,
Gasser TC, Mihatsch MJ. Metastatic patterns of prostate cancer: an
References autopsy study of 1,589 patients. Hum Pathol. 2000;31(5):578–83.
https://doi.org/10.1053/hp.2000.6698.
1. Pantel K, Hille C, Scher HI. Circulating tumor cells in prostate can- 16. Lorente D, Omlin A, Zafeiriou Z, Nava-Rodrigues D, Pérez-López
cer: from discovery to clinical utility. Clin Chem. 2019;65(1):87– R, Pezaro C, Mehra N, Sheridan E, Figueiredo I, Riisnaes R,
99. https://doi.org/10.1373/clinchem.2018.287102. Miranda S, Crespo M, Flohr P, Mateo J, Altavilla A, Ferraldeschi R,
2. Chang AJ, Autio KA, Roach M 3rd, Scher HI. High-risk pros- Bianchini D, Attard G, Tunariu N, de Bono J. Castration-resistant
tate cancer-classification and therapy. Nat Rev Clin Oncol. prostate cancer tissue acquisition from bone metastases for molecu-
2014;11(6):308–23. https://doi.org/10.1038/nrclinonc.2014.68. lar analyses. Clin Genitourin Cancer. 2016b;14(6):485–93. https://
3. Sartor O, de Bono JS. Metastatic prostate cancer. N Engl J Med. doi.org/10.1016/j.clgc.2016.04.016.
2018a;378(7):645–57. https://doi.org/10.1056/NEJMra1701695. 17. de Bono JS, Scher HI, Montgomery RB, Parker C, Miller MC, Tissing
4. Sartor O, de Bono JS. Metastatic prostate cancer. N Engl J Med. H, Doyle GV, Terstappen LW, Pienta KJ, Raghavan D. Circulating
2018b;378(17):1653–4. https://doi.org/10.1056/NEJMc1803343. tumor cells predict survival benefit from treatment in metastatic cas-
5. Carioli G, Bertuccio P, Boffetta P, Levi F, La Vecchia C, Negri E, tration-resistant prostate cancer. Clin Cancer Res. 2008;14(19):6302–
Malvezzi M. European cancer mortality predictions for the year 9. https://doi.org/10.1158/1078-­0432.CCR-­08-­0872.
2020 with a focus on prostate cancer. Ann Oncol. 2020;31(5):650– 18. Goldkorn A, Ely B, Quinn DI, Tangen CM, Fink LM, Xu T,
8. https://doi.org/10.1016/j.annonc.2020.02.009. Twardowski P, Van Veldhuizen PJ, Agarwal N, Carducci MA, Monk
6. Damodaran S, Kyriakopoulos CE, Jarrard DF. Newly diagnosed JP 3rd, Datar RH, Garzotto M, Mack PC, Lara P Jr, Higano CS,
metastatic prostate cancer: has the paradigm changed? Urol Hussain M, Thompson IM Jr, Cote RJ, Vogelzang NJ. Circulating
Clin North Am. 2017;44(4):611–21. https://doi.org/10.1016/j. tumor cell counts are prognostic of overall survival in SWOG
ucl.2017.07.008. S0421: a phase III trial of docetaxel with or without atrasentan
7. Swami U, McFarland TR, Nussenzveig R, Agarwal N. Advanced for metastatic castration-resistant prostate cancer. J Clin Oncol.
prostate cancer: treatment advances and future directions. 2014;32(11):1136–42. https://doi.org/10.1200/JCO.2013.51.7417.
Trends Cancer. 2020;6(8):702–15. https://doi.org/10.1016/j. 19. Goodman OB Jr, Fink LM, Symanowski JT, Wong B, Grobaski
trecan.2020.04.010. B, Pomerantz D, Ma Y, Ward DC, Vogelzang NJ. Circulating
8. Valenca LB, Sweeney CJ, Pomerantz MM. Sequencing cur- tumor cells in patients with castration-resistant prostate cancer
rent therapies in the treatment of metastatic prostate cancer. baseline values and correlation with prognostic factors. Cancer
Cancer Treat Rev. 2015;41(4):332–40. https://doi.org/10.1016/j. Epidemiol Biomarkers Prev. 2009;18(6):1904–13. https://doi.
ctrv.2015.02.010. org/10.1158/1055-­9965.epi-­08-­1173.
9. Gundem G, Van Loo P, Kremeyer B, Alexandrov LB, Tubio JMC, 20. Scher HI, Jia X, de Bono JS, Fleisher M, Pienta KJ, Raghavan D,
Papaemmanuil E, Brewer DS, Kallio HML, Högnäs G, Annala Heller G. Circulating tumour cells as prognostic markers in progres-
M, Kivinummi K, Goody V, Latimer C, O’Meara S, Dawson sive, castration-resistant prostate cancer: a reanalysis of IMMC38
KJ, Isaacs W, Emmert-Buck MR, Nykter M, Foster C, Kote- trial data. Lancet Oncol. 2009;10(3):233–9.
Jarai Z, Easton D, Whitaker HC, Neal DE, Cooper CS, Eeles 21. Vogelzang NJ, Fizazi K, Burke JM, De Wit R, Bellmunt J, Hutson
RA, Visakorpi T, Campbell PJ, McDermott U, Wedge DC, Bova TE, Crane E, Berry WR, Doner K, Hainsworth JD, Wiechno PJ,
GS. The evolutionary history of lethal metastatic prostate cancer. Liu K, Waldman MF, Gandhi A, Barton D, Jungnelius U, Fandi
Nature. 2015;520(7547):353–7. https://doi.org/10.1038/nature A, Sternberg CN, Petrylak DP. Circulating tumor cells in a phase
14347. 3 study of docetaxel and prednisone with or without lenalido-
10. Haffner MC, Mosbruger T, Esopi DM, Fedor H, Heaphy CM, mide in metastatic castration-resistant prostate cancer. Eur Urol.
Walker DA, Adejola N, Gürel M, Hicks J, Meeker AK, Halushka 2017;71(2):168–71. https://doi.org/10.1016/j.eururo.2016.07.051.
MK, Simons JW, Isaacs WB, De Marzo AM, Nelson WG, 22. Scher HI, Heller G, Molina A, Attard G, Danila DC, Jia X, Peng
Yegnasubramanian S. Tracking the clonal origin of lethal pros- W, Sandhu SK, Olmos D, Riisnaes R, McCormack R, Burzykowski
T, Kheoh T, Fleisher M, Buyse M, de Bono JS. Circulating tumor
144 N. H. Stoecklein and R. P. L. Neves

cell biomarker panel as an individual-level surrogate for survival 7 in circulating tumor cells as a predictive biomarker for castration-­
in metastatic castration-resistant prostate cancer. J Clin Oncol. resistant prostate cancer. JAMA Oncol. 2018;4(9):1179–86. https://
2015;33(12):1348–55. https://doi.org/10.1200/JCO.2014.55.3487. doi.org/10.1001/jamaoncol.2018.1621.
23. Lorente D, Olmos D, Mateo J, Bianchini D, Seed G, Fleisher M, 34. Scher HI, Lu D, Schreiber NA, Louw J, Graf RP, Vargas HA,
Danila DC, Flohr P, Crespo M, Figueiredo I, Miranda S, Baeten K, Johnson A, Jendrisak A, Bambury R, Danila D, McLaughlin B,
Molina A, Kheoh T, McCormack R, Terstappen LW, Scher HI, de Wahl J, Greene SB, Heller G, Marrinucci D, Fleisher M, Dittamore
Bono JS. Decline in circulating tumor cell count and treatment out- R. Association of AR-V7 on circulating tumor cells as a treatment-­
come in advanced prostate cancer. Eur Urol. 2016a;70(6):985–92. specific biomarker with outcomes and survival in castration-­
https://doi.org/10.1016/j.eururo.2016.05.023. resistant prostate cancer. JAMA Oncol. 2016;2(11):1441–9. https://
24. Heller G, McCormack R, Kheoh T, Molina A, Smith MR, Dreicer doi.org/10.1001/jamaoncol.2016.1828.
R, Saad F, de Wit R, Aftab DT, Hirmand M, Limon A, Fizazi K, 35. CMS.gov. Local coverage determination (LCD): MolDX: andro-
Fleisher M, de Bono JS, Scher HI. Circulating tumor cell number as gen receptor variant (AR-V7) protein test (L37701); 2020. https://
a response measure of prolonged survival for metastatic castration-­ www.cms.gov/medicare-­c overage-­d atabase/details/lcd-­d etails.
resistant prostate cancer: a comparison with prostate-specific anti- aspx?LCDId=37701. Accessed 13 Jan 2021.
gen across five randomized phase III clinical trials. J Clin Oncol. 36. Armstrong AJ, Halabi S, Luo J, Nanus DM, Giannakakou P,
2018;36(6):572–80. https://doi.org/10.1200/jco.2017.75.2998. Szmulewitz RZ, Danila DC, Healy P, Anand M, Rothwell CJ,
25. Keller L, Pantel K. Unravelling tumour heterogeneity by single-­ Rasmussen J, Thornburg B, Berry WR, Wilder RS, Lu C, Chen
cell profiling of circulating tumour cells. Nat Rev Cancer. Y, Silberstein JL, Kemeny G, Galletti G, Somarelli JA, Gupta S,
2019;19(10):553–67. https://doi.org/10.1038/s41568-­019-­0180-­2. Gregory SG, Scher HI, Dittamore R, Tagawa ST, Antonarakis ES,
26. de Bono JS, Attard G, Adjei A, Pollak MN, Fong PC, Haluska George DJ. Prospective multicenter validation of androgen recep-
P, Roberts L, Melvin C, Repollet M, Chianese D, Connely M, tor splice variant 7 and hormone therapy resistance in high-risk
Terstappen LW, Gualberto A. Potential applications for circulating castration-­resistant prostate cancer: the PROPHECY study. J Clin
tumor cells expressing the insulin-like growth factor-I receptor. Clin Oncol. 2019;37(13):1120–9. https://doi.org/10.1200/jco.18.01731.
Cancer Res. 2007;13(12):3611–6. https://doi.org/10.1158/1078- 37. Abida W, Cyrta J, Heller G, Prandi D, Armenia J, Coleman I,
­0432.ccr-­07-­0268. Cieslik M, Benelli M, Robinson D, Van Allen EM, Sboner A,
27. Gorges TM, Riethdorf S, von Ahsen O, Nastał YP, Röck K, Boede M, Fedrizzi T, Mosquera JM, Robinson BD, De Sarkar N, Kunju LP,
Peine S, Kuske A, Schmid E, Kneip C, König F, Rudolph M, Pantel Tomlins S, Wu YM, Nava Rodrigues D, Loda M, Gopalan A, Reuter
K. Heterogeneous PSMA expression on circulating tumor cells: a VE, Pritchard CC, Mateo J, Bianchini D, Miranda S, Carreira S,
potential basis for stratification and monitoring of PSMA-directed Rescigno P, Filipenko J, Vinson J, Montgomery RB, Beltran H,
therapies in prostate cancer. Oncotarget. 2016;7(23):34,930–41. Heath EI, Scher HI, Kantoff PW, Taplin ME, Schultz N, deBono
https://doi.org/10.18632/oncotarget.9004. JS, Demichelis F, Nelson PS, Rubin MA, Chinnaiyan AM, Sawyers
28. Antonarakis ES, Armstrong AJ, Dehm SM, Luo J. Androgen recep- CL. Genomic correlates of clinical outcome in advanced prostate
tor variant-driven prostate cancer: clinical implications and thera- cancer. Proc Natl Acad Sci U S A. 2019;116(23):11,428–36. https://
peutic targeting. Prostate Cancer Prostatic Dis. 2016;19(3):231–41. doi.org/10.1073/pnas.1902651116.
https://doi.org/10.1038/pcan.2016.17. 38. Scher HI, Graf RP, Schreiber NA, McLaughlin B, Jendrisak A,
29. Hu R, Dunn TA, Wei S, Isharwal S, Veltri RW, Humphreys E, Han Wang Y, Lee J, Greene S, Krupa R, Lu D, Bamford P, Louw JE,
M, Partin AW, Vessella RL, Isaacs WB, Bova GS, Luo J. Ligand-­ Dugan L, Vargas HA, Fleisher M, Landers M, Heller G, Dittamore
independent androgen receptor variants derived from splicing of R. Phenotypic heterogeneity of circulating tumor cells informs
cryptic exons signify hormone-refractory prostate cancer. Cancer clinical decisions between AR signaling inhibitors and Taxanes
Res. 2009;69(1):16–22. https://doi.org/10.1158/0008-­5472. in metastatic prostate cancer. Cancer Res. 2017;77(20):5687–98.
can-­08-­2764. https://doi.org/10.1158/0008-­5472.can-­17-­1353.
30. Watson PA, Chen YF, Balbas MD, Wongvipat J, Socci ND, Viale 39. Lambros MB, Seed G, Sumanasuriya S, Gil V, Crespo M, Fontes
A, Kim K, Sawyers CL. Constitutively active androgen recep- MS, Chandler R, Mehra N, Fowler G, Ebbs B, Flohr PR, Miranda
tor splice variants expressed in castration-resistant prostate S, Yuan W, Mackay A, Ferreira A, Pereira R, Bertan C, Figueiredo
cancer require full-length androgen receptor. Proc Natl Acad I, Riisnaes R, Nava Rodrigues D, Sharp A, Goodall J, Boysen
Sci U S A. 2010;107(39):16759–65. https://doi.org/10.1073/ G, Carreira S, Bianchini D, Rescigno P, Zafeiriou Z, Hunt J,
pnas.1012443107. Moloney D, Hamilton L, Neves RPL, Swennenhuis JF, Andree KC,
31. Antonarakis ES, Lu C, Wang H, Luber B, Nakazawa M, Roeser Stoecklein NH, Terstappen LW, de Bono JS. Single cell analyses of
JC, Chen Y, Mohammad TA, Chen Y, Fedor HL, Lotan TL, Zheng prostate cancer liquid biopsies acquired by apheresis. Clin Cancer
Q, De Marzo AM, Isaacs JT, Isaacs WB, Nadal R, Paller CJ, Res. 2018; https://doi.org/10.1158/1078-­0432.CCR-­18-­0862.
Denmeade SR, Carducci MA, Eisenberger MA, Luo J. AR-V7 40. Lu YT, Delijani K, Mecum A, Goldkorn A. Current status of liq-
and resistance to enzalutamide and abiraterone in prostate cancer. uid biopsies for the detection and management of prostate cancer.
N Engl J Med. 2014;371(11):1028–38. https://doi.org/10.1056/ Cancer Manag Res. 2019;11:5271–91. https://doi.org/10.2147/
NEJMoa1315815. cmar.s170380.
32. Antonarakis ES, Lu C, Luber B, Wang H, Chen Y, Zhu Y, 41. Pantel K, Alix-Panabières C. Functional studies on viable circu-
Silberstein JL, Taylor MN, Maughan BL, Denmeade SR, Pienta lating tumor cells. Clin Chem. 2016;62(2):328–34. https://doi.
KJ, Paller CJ, Carducci MA, Eisenberger MA, Luo J. Clinical sig- org/10.1373/clinchem.2015.242537.
nificance of androgen receptor splice Variant-7 mRNA detection in 42. Gao D, Vela I, Sboner A, Iaquinta PJ, Karthaus WR, Gopalan
circulating tumor cells of men with metastatic castration-resistant A, Dowling C, Wanjala JN, Undvall EA, Arora VK, Wongvipat
prostate cancer treated with first- and second-line Abiraterone and J, Kossai M, Ramazanoglu S, Barboza LP, Di W, Cao Z, Zhang
enzalutamide. J Clin Oncol. 2017;35(19):2149–56. https://doi. QF, Sirota I, Ran L, MacDonald TY, Beltran H, Mosquera JM,
org/10.1200/jco.2016.70.1961. Touijer KA, Scardino PT, Laudone VP, Curtis KR, Rathkopf DE,
33. Scher HI, Graf RP, Schreiber NA, Jayaram A, Winquist E, Morris MJ, Danila DC, Slovin SF, Solomon SB, Eastham JA, Chi
McLaughlin B, Lu D, Fleisher M, Orr S, Lowes L, Anderson A, P, Carver B, Rubin MA, Scher HI, Clevers H, Sawyers CL, Chen
Wang Y, Dittamore R, Allan AL, Attard G, Heller G. Assessment Y. Organoid cultures derived from patients with advanced pros-
of the validity of nuclear-localized androgen receptor splice variant
14 Circulating Tumor Cells and ctDNA in Prostate Cancer 145

tate cancer. Cell. 2014;159(1):176–87. https://doi.org/10.1016/j. 53. Carreira S, Romanel A, Goodall J, Grist E, Ferraldeschi R,
cell.2014.08.016. Miranda S, Prandi D, Lorente D, Frenel JS, Pezaro C, Omlin
43. Rossi E, Rugge M, Facchinetti A, Pizzi M, Nardo G, Barbieri V, A, Rodrigues DN, Flohr P, Tunariu N, SdB J, Demichelis F,
Manicone M, De Faveri S, Chiara Scaini M, Basso U, Amadori Attard G. Tumor clone dynamics in lethal prostate cancer. Sci
A, Zamarchi R. Retaining the long-survive capacity of circulating Transl Med. 2014;6(254):254ra125. https://doi.org/10.1126/
tumor cells (CTCs) followed by xeno-transplantation: not only from scitranslmed.3009448.
metastatic cancer of the breast but also of prostate cancer patients. 54. Wyatt AW, Annala M, Aggarwal R, Beja K, Feng F, Youngren J,
Onco Targets Ther. 2014;1(1):49–56. https://doi.org/10.18632/ Foye A, Lloyd P, Nykter M, Beer TM, Alumkal JJ, Thomas GV,
oncoscience.8. Reiter RE, Rettig MB, Evans CP, Gao AC, Chi KN, Small EJ,
44. Alix-Panabières C, Rebillard X, Brouillet JP, Barbotte E, Iborra F, Gleave ME. Concordance of circulating tumor DNA and matched
Segui B, Maudelonde T, Jolivet-Reynaud C, Vendrell JP. Detection metastatic tissue biopsy in prostate cancer. J Natl Cancer Inst.
of circulating prostate-specific antigen-secreting cells in pros- 2017;109(12) https://doi.org/10.1093/jnci/djx118.
tate cancer patients. Clin Chem. 2005;51(8):1538–41. https://doi. 55. Goodall J, Mateo J, Yuan W, Mossop H, Porta N, Miranda S, Perez-­
org/10.1373/clinchem.2005.049445. Lopez R, Dolling D, Robinson DR, Sandhu S, Fowler G, Ebbs B,
45. Baccelli I, Schneeweiss A, Riethdorf S, Stenzinger A, Schillert A, Flohr P, Seed G, Rodrigues DN, Boysen G, Bertan C, Atkin M,
Vogel V, Klein C, Saini M, Bauerle T, Wallwiener M, Holland-Letz Clarke M, Crespo M, Figueiredo I, Riisnaes R, Sumanasuriya
T, Hofner T, Sprick M, Scharpff M, Marme F, Sinn HP, Pantel K, S, Rescigno P, Zafeiriou Z, Sharp A, Tunariu N, Bianchini D,
Weichert W, Trumpp A. Identification of a population of blood cir- Gillman A, Lord CJ, Hall E, Chinnaiyan AM, Carreira S, de Bono
culating tumor cells from breast cancer patients that initiates metas- JS. Circulating cell-free DNA to guide prostate cancer treatment
tasis in a xenograft assay. Nat Biotechnol. 2013;31(6):539–44. with PARP inhibition. Cancer Discov. 2017;7(9):1006–17. https://
https://doi.org/10.1038/nbt.2576. doi.org/10.1158/2159-­8290.cd-­17-­0261.
46. Franken A, Driemel C, Behrens B, Meier-Stiegen F, Endris 56. Wyatt AW, Azad AA, Volik SV, Annala M, Beja K, McConeghy B,
V, Stenzinger A, Niederacher D, Fischer JC, Stoecklein NH, Haegert A, Warner EW, Mo F, Brahmbhatt S, Shukin R, Le Bihan
Ruckhaeberle E, Fehm T, Neubauer H. Label-free enrichment and S, Gleave ME, Nykter M, Collins CC, Chi KN. Genomic alterations
molecular characterization of viable circulating tumor cells from in cell-free DNA and enzalutamide resistance in castration-resistant
diagnostic leukapheresis products. Clin Chem. 2019;65(4):549–58. prostate cancer. JAMA Oncol. 2016;2(12):1598–606. https://doi.
https://doi.org/10.1373/clinchem.2018.296814. org/10.1001/jamaoncol.2016.0494.
47. Fischer JC, Niederacher D, Topp SA, Honisch E, Schumacher 57. Gasi Tandefelt D, de Bono J. Circulating cell-free DNA: trans-
S, Schmitz N, Zacarias Fohrding L, Vay C, Hoffmann I, lating prostate cancer genomics into clinical care. Mol Asp Med.
Kasprowicz NS, Hepp PG, Mohrmann S, Nitz U, Stresemann 2020;72:100837. https://doi.org/10.1016/j.mam.2019.100837.
A, Krahn T, Henze T, Griebsch E, Raba K, Rox JM, Wenzel F, 58. Boerrigter E, Groen LN, Van Erp NP, Verhaegh GW, Schalken
Sproll C, Janni W, Fehm T, Klein CA, Knoefel WT, Stoecklein JA. Clinical utility of emerging biomarkers in prostate cancer liquid
NH. Diagnostic leukapheresis enables reliable detection of cir- biopsies. Expert Rev Mol Diagn. 2020;20(2):219–30. https://doi.
culating tumor cells of nonmetastatic cancer patients. Proc Natl org/10.1080/14737159.2019.1675515.
Acad Sci U S A. 2013;110(41):16,580–5. https://doi.org/10.1073/ 59. De Laere B, Oeyen S, Mayrhofer M, Whitington T, van Dam
pnas.1313594110. PJ, Van Oyen P, Ghysel C, Ampe J, Ost P, Demey W, Hoekx L,
48. Faugeroux V, Pailler E, Oulhen M, Deas O, Brulle-Soumare L, Schrijvers D, Brouwers B, Lybaert W, Everaert EG, De Maeseneer
Hervieu C, Marty V, Alexandrova K, Andree KC, Stoecklein NH, D, Strijbos M, Bols A, Fransis K, Beije N, de Kruijff IE, van Dam
Tramalloni D, Cairo S, NgoCamus M, Nicotra C, Terstappen L, V, Brouwer A, Goossens D, Heyrman L, Van den Eynden GG,
Manaresi N, Lapierre V, Fizazi K, Scoazec JY, Loriot Y, Judde Rutten A, Del Favero J, Rantalainen M, Rajan P, Sleijfer S, Ullén A,
JG, Farace F. Genetic characterization of a unique neuroendocrine Yachnin J, Grönberg H, Van Laere SJ, Lindberg J, Dirix LY. TP53
transdifferentiation prostate circulating tumor cell-derived eXplant outperforms other androgen receptor biomarkers to predict abi-
model. Nat Commun. 2020;11(1):1884. https://doi.org/10.1038/ raterone or enzalutamide outcome in metastatic castration-resistant
s41467-­020-­15426-­2. prostate cancer. Clin Cancer Res. 2019;25(6):1766–73. https://doi.
49. Mout L, van Dessel LF, Kraan J, de Jong AC, Neves RPL, Erkens-­ org/10.1158/1078-­0432.ccr-­18-­1943.
Schulze S, Beaufort CM, Sieuwerts AM, van Riet J, Woo TLC, 60. Azad AA, Volik SV, Wyatt AW, Haegert A, Le Bihan S, Bell RH,
de Wit R, Sleijfer S, Hamberg P, Sandberg Y, Te Boekhorst PAW, Anderson SA, McConeghy B, Shukin R, Bazov J, Youngren J,
van de Werken HJG, Martens JWM, Stoecklein NH, van Weerden Paris P, Thomas G, Small EJ, Wang Y, Gleave ME, Collins CC,
WM, Lolkema MP. Generating human prostate cancer organoids Chi KN. Androgen receptor gene aberrations in circulating cell-free
from leukapheresis enriched circulating tumour cells. Eur J Cancer. DNA: biomarkers of therapeutic resistance in castration-resistant
2021;150:179–89. https://doi.org/10.1016/j.ejca.2021.03.023. prostate cancer. Clin Cancer Res. 2015;21(10):2315–24. https://
50. Neumann MHD, Bender S, Krahn T, Schlange T. ctDNA and doi.org/10.1158/1078-­0432.ccr-­14-­2666.
CTCs in liquid biopsy—current status and where we need to prog- 61. Conteduca V, Wetterskog D, Sharabiani MTA, Grande E,
ress. Comput Struct Biotechnol J. 2018;16:190–5. https://doi. Fernandez-Perez MP, Jayaram A, Salvi S, Castellano D, Romanel
org/10.1016/j.csbj.2018.05.002. A, Lolli C, Casadio V, Gurioli G, Amadori D, Font A, Vazquez-­
51. Heitzer E, Ulz P, Belic J, Gutschi S, Quehenberger F, Fischereder Estevez S, González Del Alba A, Mellado B, Fernandez-Calvo O,
K, Benezeder T, Auer M, Pischler C, Mannweiler S, Pichler M, Méndez-Vidal MJ, Climent MA, Duran I, Gallardo E, Rodriguez
Eisner F, Haeusler M, Riethdorf S, Pantel K, Samonigg H, Hoefler A, Santander C, Sáez MI, Puente J, Gasi Tandefelt D, Wingate A,
G, Augustin H, Geigl JB, Speicher MR. Tumor-associated copy Dearnaley D, Demichelis F, De Giorgi U, Gonzalez-Billalabeitia E,
number changes in the circulation of patients with prostate can- Attard G. Androgen receptor gene status in plasma DNA associates
cer identified through whole-genome sequencing. Genome Med. with worse outcome on enzalutamide or abiraterone for castration-­
2013;5(4):30. https://doi.org/10.1186/gm434. resistant prostate cancer: a multi-institution correlative biomarker
52. Kolinsky MP, Stoecklein N, Lambros M, Gil V, Rodrigues DN, study. Ann Oncol. 2017;28(7):1508–16. https://doi.org/10.1093/
Carreira S, Zafeiriou Z, de Bono JS. Genetic analysis of circulating annonc/mdx155.
tumour cells. Recent Results Cancer Res. 2020;215:57–76. https:// 62. Romanel A, Gasi Tandefelt D, Conteduca V, Jayaram A, Casiraghi
doi.org/10.1007/978-­3-­030-­26439-­0_3. N, Wetterskog D, Salvi S, Amadori D, Zafeiriou Z, Rescigno P,
146 N. H. Stoecklein and R. P. L. Neves

Bianchini D, Gurioli G, Casadio V, Carreira S, Goodall J, Wingate Seed G, Jain S, Ralph C, Protheroe A, Hussain S, Jones R, Elliott
A, Ferraldeschi R, Tunariu N, Flohr P, De Giorgi U, de Bono JS, T, McGovern U, Bianchini D, Goodall J, Zafeiriou Z, Williamson
Demichelis F, Attard G. Plasma AR and abiraterone-resistant pros- CT, Ferraldeschi R, Riisnaes R, Ebbs B, Fowler G, Roda D, Yuan
tate cancer. Sci Transl Med. 2015;7(312):312re310. https://doi. W, Wu YM, Cao X, Brough R, Pemberton H, A'Hern R, Swain
org/10.1126/scitranslmed.aac9511. A, Kunju LP, Eeles R, Attard G, Lord CJ, Ashworth A, Rubin
63. Annala M, Vandekerkhove G, Khalaf D, Taavitsainen S, Beja K, MA, Knudsen KE, Feng FY, Chinnaiyan AM, Hall E, de Bono
Warner EW, Sunderland K, Kollmannsberger C, Eigl BJ, Finch D, JS. DNA-repair defects and Olaparib in metastatic prostate cancer.
Oja CD, Vergidis J, Zulfiqar M, Azad AA, Nykter M, Gleave ME, N Engl J Med. 2015;373(18):1697–708. https://doi.org/10.1056/
Wyatt AW, Chi KN. Circulating tumor DNA genomics correlate NEJMoa1506859.
with resistance to Abiraterone and enzalutamide in prostate cancer. 66. Mateo J, Porta N, Bianchini D, McGovern U, Elliott T, Jones
Cancer Discov. 2018;8(4):444–57. https://doi.org/10.1158/2159- R, Syndikus I, Ralph C, Jain S, Varughese M, Parikh O, Crabb
­8290.cd-­17-­0937. S, Robinson A, McLaren D, Birtle A, Tanguay J, Miranda S,
64. Robinson D, Van Allen EM, Wu YM, Schultz N, Lonigro RJ, Figueiredo I, Seed G, Bertan C, Flohr P, Ebbs B, Rescigno P,
Mosquera JM, Montgomery B, Taplin ME, Pritchard CC, Attard G, Fowler G, Ferreira A, Riisnaes R, Pereira R, Curcean A, Chandler
Beltran H, Abida W, Bradley RK, Vinson J, Cao X, Vats P, Kunju R, Clarke M, Gurel B, Crespo M, Nava Rodrigues D, Sandhu S,
LP, Hussain M, Feng FY, Tomlins SA, Cooney KA, Smith DC, Espinasse A, Chatfield P, Tunariu N, Yuan W, Hall E, Carreira
Brennan C, Siddiqui J, Mehra R, Chen Y, Rathkopf DE, Morris S, de Bono JS. Olaparib in patients with metastatic castration-
MJ, Solomon SB, Durack JC, Reuter VE, Gopalan A, Gao J, Loda resistant prostate cancer with DNA repair gene aberrations
M, Lis RT, Bowden M, Balk SP, Gaviola G, Sougnez C, Gupta M, (TOPARP-B): a multicentre, open-­ label, randomised, phase 2
Yu EY, Mostaghel EA, Cheng HH, Mulcahy H, True LD, Plymate trial. Lancet Oncol. 2020;21(1):162–74. https://doi.org/10.1016/
SR, Dvinge H, Ferraldeschi R, Flohr P, Miranda S, Zafeiriou Z, S1470-­2045(19)30684-­9.
Tunariu N, Mateo J, Perez-Lopez R, Demichelis F, Robinson BD, 67. de Bono J, Mateo J, Fizazi K, Saad F, Shore N, Sandhu S, Chi KN,
Sboner A, Schiffman M, Nanus DM, Tagawa ST, Sigaras A, Eng Sartor O, Agarwal N, Olmos D, Thiery-Vuillemin A, Twardowski
KW, Elemento O, Sboner A, Heath EI, Scher HI, Pienta KJ, Kantoff P, Mehra N, Goessl C, Kang J, Burgents J, Wu W, Kohlmann A,
P, de Bono JS, Rubin MA, Nelson PS, Garraway LA, Sawyers Adelman CA, Hussain M. Olaparib for metastatic castration-­
CL, Chinnaiyan AM. Integrative clinical genomics of advanced resistant prostate cancer. N Engl J Med. 2020;382(22):2091–102.
prostate cancer. Cell. 2015;162(2):454. https://doi.org/10.1016/j. https://doi.org/10.1056/NEJMoa1911440.
cell.2015.06.053. 68. fda.gov. FDA approves liquid biopsy NGS companion diagnostic
65. Mateo J, Carreira S, Sandhu S, Miranda S, Mossop H, Perez-­ test for multiple cancers and biomarkers; 2020. https://www.fda.
Lopez R, Nava Rodrigues D, Robinson D, Omlin A, Tunariu N, gov/drugs/fda-­approves-­liquid-­biopsy-­ngs-­companion-­diagnostic-­
Boysen G, Porta N, Flohr P, Gillman A, Figueiredo I, Paulding C, test-­multiple-­cancers-­and-­biomarkers. Accessed 13 Jan 2021.
 Circulating Tumor Cells in Lung Cancer
15
Paul Hofman

Abstract
• To inform on the different interests of using CTCs
The detection and identification of circulating tumor cells in lung cancer patients as a prognostic and as a
(CTCs) and the analysis of the information they provide diagnostic biomarker, including for lung cancer
are often the subject of discussion in thoracic oncology, in screening.
particular for routine clinical use. However, CTCs are • To show the strong interest of combining the analy-
being used for an increasing range of applications in sev- ses of CTCs and other blood components (including
eral domains of lung carcinoma. Thus, CTCs can partici- the circulating free DNA) to optimize the value of
pate in better defining not only the prognosis and diagnosis using liquid biopsies for targeted therapy and
of lung cancer but also screening for the disease. CTCs immunotherapy in lung cancer patients.
can also contribute to the major therapeutic orientations • To highlight the importance of translational research
through not only the study of the efficacy of targeted mol- projects based on CTCs assessment for better
ecules but also of immunotherapy. In addition, the study understanding the pathophysiology of lung cancer
of CTCs provides a better understanding of the natural analyses.
history of lung cancer and of the mechanisms of metasta- • To inform on the next challenges of using CTCs in
sis. This review analyzes the different orientations and thoracic oncology.
therapeutic approaches associated with the study of CTCs
in the domain of thoracic oncology.

15.1 Introduction
Learning Objectives
• To cite the main current challenges for the detection Lung cancer is the first cause of death from cancer in the
and the characterization of CTCs in lung cancer world [1]. Despite progress into early diagnosis and the
patients. emergence of novel therapeutics that improve the overall
• To develop the potential interest of looking for the survival of patients, the incidence of cancer has increased in
presence of CTCs in lung cancer patients in routine the majority of countries [1]. Thus, it is urgent to make an
clinical practice. effort to slow down a disease that concerns more and more
young patients, and so has a major impact on health costs
worldwide too. To improve the diagnostic and care effective-
ness of this cancer a number of approaches are possible
including treating better patients and thereby increasing the
chances of remission or of prolonging life, even at a meta-
P. Hofman (*)
Université Côte d’Azur, CNRS, INSERM, IRCAN, Centre Antoine static stage. Among the different approaches, the develop-
Lacassagne, FHU OncoAge, Team 4, Nice, France ment of liquid biopsies has revolutionized in recent years
Université Côte d’Azur, CHU Nice, FHU OncoAge, Laboratory several domains of lung cancer patient taking care, including
of Clinical and Experimental Pathology, Pasteur Hospital, the potential capacity of screening, early diagnosis, identifi-
Nice, France cation of factors of prognosis and of the predictive response
Université Côte d’Azur, CHU Nice, FHU OncoAge, or resistance to treatment as well as the detection of a resid-
Hospital-­Integrated Biobank BB-0033-00025, Nice, France ual cancer (or a « minimal residual disease »). To this aim
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 147
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_15
148 P. Hofman

several components of blood have been studied but it is the 15.2.2  he Challenges in Using CTCs in Daily
T
analysis of circulating free nucleic acids in plasma that in Practice for Lung Cancer Patients
recent years has been the main domain of daily life applica-
tion of liquid biopsies in thoracic oncology. More precisely, Therefore, the usefulness of CTCs in routine clinical practice
the analysis of genomic tumor circulating DNA in plasma for lung cancer patients is still under debate, since most
has progressively become one of the routine clinical applications, recommendations, and guidelines (such as
approaches in thoracic oncology. This noninvasive approach those provided by the International Association for the Study
also provides more and more rapid results, notably for the of Lung Cancer/IASLC), for use of a liquid biopsy for better
choice of targeted therapy [2–4]. care of lung cancer patients, are mainly based on evaluation
Other than nucleic acids in plasma several components of cfDNA [11]. However, many investigators highlight the
and compartments of blood, including circulating tumor cells fact that CTCs are a major component for assessment of lung
(CTCs) are being investigated in some very important studies cancer patients and have strong added value for many differ-
in the area of lung cancer [5, 6]. Paradoxically, CTCs were ent clinical and research applications [12–15] (Fig. 15.1).
the first to be studied with liquid biopsies but their use in clin- The issue of the added value will be dealt with below and
ical routine and the interest they hold for oncologist in the should also be considered in comparison with the potential
area of lung cancer have not followed the same momentum use of cfDNA. In fact, some of the challenges in using CTCs
and development as the analysis of circulating free DNA in routine clinical practice include, at least, the cost effec-
(cfDNA). This can be linked to a number of constraints that tiveness of the assay, the reproducibility of the results, the
will be discussed in this chapter. However, beyond its limits, difficulty in obtaining accreditation for this biological test,
the analysis of CTCs provides essential information and is and good mastering of the pre-analytical phase. For this lat-
indispensable to the area of lung cancer. Moreover, the infor- ter point, and since the detection of CTCs is a rare event, it
mation obtained can be complementary to that provided by requires optimal preservation of the blood for several hours.
the study of other exploitable components of liquid biopsies.
The issues linked to the detection and characterization of
CTCs in lung cancer and the perspectives in this domain will 15.3 Circulating Tumor Cells
be discussed. as a Prognostic Biomarker for Lung
Cancer

15.2  irculating Tumor Cells in Lung

C Several studies have demonstrated the interest in detecting
Cancer: Advantages and Limitations and quantifying CTCs of solid tumors, in particular of lung
cancers, and of correlating the number with a more or less
15.2.1 Current Issue unfavorable prognosis [16–18]. These studies were per-
formed on patients at an early stage (stage I-IIIA) or at an
Despite the many improvements in assays and technologies advanced or metastatic stage (stage IIIB/IV) [19–21]. A cor-
for CTC detection and characterization, the main challenge relation between a high number of CTCs isolated from blood
certainly lies in the selection of the best method according to and poor prognosis was observed irrespective of the histo-
the project and the different goals associated with CTC logical type of the lung cancer, non-small cell or small cell
assessment [7–10]. The multiple methods and approaches to lung cancer [19–22]. The different analyses were performed
the isolation, identification, and characterization of CTCs in either at diagnosis or on tumor progression on treatment [19–
lung oncology have made the selection of a technique par- 21]. Finally, some studies have shown an association between
ticularly difficult. This has also brought in constraints linked a high number of CTCs isolated with the CellSearch method
to the comparison of the different studies into validation and from blood of pulmonary veins before surgical resection and
reproducibility, placing these methods in competition, nota- a high frequency of postoperative recurrence [23, 24].
bly for their specificity and sensitivity. Thus, the major chal- Different methods have been used to detect and quantify
lenge today is to use CTCs as a biomarker in routine lung CTCs in lung cancer, in particular the CellSearch method or
oncology, which has not yet been adopted. The only method the ISET which is based on blood filtration on polycarbonate
that has been approved by the Food and Drug Administration filters [20, 21, 25, 26]. Each method has its advantages and
(FDA) in the USA is the CellSearch method, but only for disadvantages [7]. The CellSearch method may lack sensitiv-
specific targeted clinical situations to define prognostic fac- ity due to the use of an anti-cytokeratin antibody for capture
tors for prostate, colon, and breast metastatic cancers. So, no of CTCs since a certain number of these cells do not express,
method of detection of CTCs has been approved by either the or express a low level of, cytokeratins as a result of epithelial-­
FDA or a European agency for management of lung cancer mesenchymal phenomenon, which reduces the sensitivity of
and thus for lung cancer patient care. this method [25–27]. Blood filtration can identify the CTCs
15 Circulating Tumor Cells in Lung Cancer 149

Prognosis in early stage cancers

Prognosis in late stage cancers

Predictive in immuno-oncology
Predictive in targeted therapy
3

Diagnosis of non-
CTCs detection
small cell lung carcinoma vs
and characterization
small-cell lung carcinoma
Adenocarcinoma vs squamous
4
cell carcinoma

Screening tests
Benign vs malignant nodules
5

Improvement of
our knowledge in
lung cancer pathophysiology

Fig. 15.1 Add value of exploring the circulating tumor cells (CTCs) in lung cancer patients

on the surface of the filters. However, the characterization sitivity of the CellSearch and ISET methods using samples
and identification of these CTCs can be fastidious, since it is from the same lung cancer patients. Overall, the ISET
very time consuming and requires a strong expertise of a method appears to be more sensitive than the CellSearch
cytopathologist [28]. Some of the cells on the filters have method in these populations [25, 26]. Moreover, it cannot be
cytopathological aspects that do not allow them to be distin- excluded that some small circulating tumor cells may escape
guished with certainty, in the absence at least of immunocy- detection due to passage through the pores of the filters and
tochemistry (ICC), between non-tumor epithelial cells, giant so are not counted.
macrophages or large monocytes, activated lymphocytes or These studies demonstrated that the detection of CTCs
even circulating megakaryocytes [28, 29]. Thus, the cyto-­ before tumor resection can be associated with a poor progno-
morphological characterization of CTCs on filters is difficult sis and that postoperative recurrence may occur, even after
and requires considerable expertise along with comparative complete tumor resection during surgery [19]. The presence
analyses for validation [19]. Studies have compared the sen- of CTCs in patients with a locally advanced stage of disease
150 P. Hofman

and treated with radio-chemotherapy was associated with few recent studies orientated to this type of combination, CTCs
early recurrence [30]. The presence and high number of are not used in daily practice to evaluate the EGFR status of
CTCs detected in the blood of patients at a metastatic stage patients with metastatic lung cancer. This is probably due to the
before treatment also correlated with poor prognosis, for difficulty in standardization of the technological approaches,
both non-small cell and small cell lung carcinoma [20, 21, the heterogeneity of the CTCs population, and also the quanti-
31]. In contrast, a rapid decrease in the number of CTCs in tative variations in the number of CTCs from patient to patient.
treated patients, particularly during chemotherapy, has been Thus, at present the detection of mutations in EGFR is done
shown to be a factor of positive response to treatment and of routinely with cfDNA in plasma using either targeted sequenc-
favorable prognosis [21]. This was observed for both non-­ ing technologies or NGS [2–4].
small cell and small cell lung carcinomas [21, 31, 32]. Several studies have shown that it is possible to evaluate the
Finally, patients can also be monitored by CTC detection status of gene rearrangements (ALK, ROS1, and RET fusions)
after surgery or during treatment for the possible presence of with CTCs when using ICC and/or FISH [42, 43]. Most stud-
a residual tumor and therefore potentially proposed a new ies have shown that the detection of ALK rearrangements with
therapeutic option [33]. free circulating nucleic acids has a low sensitivity (in the range
Other than their number, certain phenotypes of CTCs of 40–50%) compared to analyses performed with tissue sam-
identified with different biomarkers (anti-cytokeratin, anti-­ ples from the same patient [38]. A recent multi-centric study
vimentin, or anti-PD-L1 antibodies, etc.) were associated performed in France validated the possibility of examining the
with poorer prognosis, notably in the case of epithelial-­ ALK status by ICC or FISH with CTCs isolated with filtration
mesenchymal transition phenomenon [34–36]. More techniques [43]. Other studies demonstrated the interest in
recently, it was shown that CTCs circulating in clumps (or in detecting certain types of profiles of ALK amplification in
« clusters ») rather than isolated were associated with a CTCs as a biomarker of efficacy and resistance to therapy tar-
poorer prognosis for patients with lung cancer [37]. geting ALK [42]. The status of ROS1 has also been evaluated
with CTCs isolated by filtration of blood, but a multi-centric
validation study and comparison with RT-PCR or NGS from
15.4  irculating Tumor Cells as a Predictive
C free nucleic acids in plasma are certainly necessary to deter-
Biomarker for Lung Cancer mine the sensitivity of this approach [42].
Other biomarkers predictive of the response to targeted
Certain biomarkers identified on CTCs can be predictive of a therapy have been evaluated with CTCs. Thus, the protein
therapeutic response to either targeted therapies or immuno- expression of MET has been quantified with CTCs isolated
therapy [38, 39]. These biomarkers can be detected with and characterized after filtration on polycarbonate filters and
immunocytochemistry, fluorescent in situ hybridization the level of expression was compared to that of tissue sam-
(FISH), or, though less frequently, with complex molecular ples obtained from the same patients [44].
analyses after extraction of nucleic acids and then analysis On tumor progression in patients treated with ALK inhibi-
by targeted sequencing or next-generation sequencing (NGS) tors the mechanisms of resistance are more often analyzed with
[38, 39]. circulating free nucleic acids (in particular ALK mutations)
[45]. However, some studies have also shown that amplifica-
tions in MET, one of the mechanisms of resistance to ALK
15.4.1 Biomarkers for Targeted Therapies inhibitors, can be detected on progression with CTCs [46].

Thanks to molecular analyses performed with CTCs the inter-

est in detecting mutations in EGFR in blood was reported for 15.4.2 Biomarkers of Immunotherapy
the first time for patients with non-small cell lung cancer.
Consequently, patient treatment could be adjusted with admin- Some biomarkers have been evaluated or used as predictive
istration of inhibitors of tyrosine kinase targeting these muta- factors of response to immune checkpoint inhibitors [39].
tions [40]. This approach has also been tested in more recent Both tissue and blood samples are used to detect these bio-
studies [41]. However, this approach using CTCs was rapidly markers [36, 39, 47]. The predictive biomarker the most ana-
replaced by the analysis of these mutations using cfDNA, lyzed and validated to date to predict the response to the
because certain technologies provided more reproducible anti-PD-L1/PDL1 immune checkpoint inhibitors is the pro-
results, were more easily performed and less costly [2, 3]. It is tein marker PD-L1 [47]. This biomarker is evaluated by
interesting to note that an integrative approach associating immunohistochemistry with different anti-PD-L1 clones, but
detection of EGFR mutations in CTCs and cfDNA from the has also been evaluated by ICC on different cytological sam-
same patients increased the possibility of detecting these ples [48]. Moreover, several studies have shown good corre-
genomic alterations. Nonetheless it is clear that aside from a lation between the expression of PD-L1 on tissue and
15 Circulating Tumor Cells in Lung Cancer 151

cytological samples from the same patients [48]. In this con- obstructive pulmonary disease or emphysema), with low-­
text, the expression of PD-L1 has been evaluated by cytology dose computed tomography (LDCT) [51]. Thus, the
with blood and thus on CTCs obtained with different National Lung Screening Trial (NLST) and Nelson study
approaches including methods of filtration [35]. It has been have shown that screening can reduce the annual mortality
shown that it is possible to identify PD-L1 on CTCs but the from lung cancer for a few thousand individuals [51].
prognostic significance and the predictive value of this Despite these important findings, several hurdles have
expression is controversial [35, 49]. A recent study demon- slowed down the setup of screening with LDCT in several
strated that using the CellSearch system, it was possible to countries, including France. While minimally invasive this
detect PD-L1 in a subset of CTCs associated with worse test can appear arduous for patients and patient compliance
overall survival of late stage NSCLC [36]. is sometimes low, as in the USA. The CT images can also
Different immunotherapy clinical trials in lung cancer give false negative or positive results resulting into need-
patients have evaluated the tumor mutation burden (TMB) less surgery with a risk of comorbidity and mortality.
using circulating nucleic acids in plasma [39]. Several clinical Knowing this it is of strong interest to develop or associate
trials have shown that a high level of TMB in plasma can be other tests for early screening of lung cancer including
associated with a good response to immunotherapy [39]. Only blood tests [50]. The detection of CTCs has aroused hope
a few studies have evaluated the prognostic value of the analy- as a screening approach, in particular for patients at risk. A
sis of the TMB with cytological samples but an approach with mono-centric study was thus able, in the absence of CT
CTCs may appear to date to have major technological issues images of tumors, to predict the emergence of a lung cancer
and very difficult to perform in clinical practice. The number several months before detection by imagery [52]. This
of CTCs detected before immunotherapy has been associated promising study has received substantial enthusiasm from
with a poor response to immune checkpoint inhibitors [39]. the media but needs to be discussed while integrating sev-
However, at present, this parameter is not taken into account eral parameters. It concerns a retrospective study performed
when modifying the treatment in clinical practice. on a relatively small cohort of patients. A multi-centric
validation study involving several French hospitals per-
formed did not totally confirm the initial results [53]. This
15.5 Circulating Tumor Cells discordance can be explained by several elements, in par-
as a Diagnostic Biomarker of Lung ticular the variation in the pre-analytical and analytical
Cancer parameters depending on the hospital, but also a slight dif-
ference of the design of the two studies [52, 53].
One of the challenges in thoracic oncology is to diagnose at
an early stage a lung cancer using methods that are robust,
sensitive and specific, reproducible and noninvasive and with 15.7  tudying Circulating Tumor Cells
S
a low cost [6]. Another challenge is to propose to a population to Better Understand
of eligible patients an early diagnosis that potentially leads to the Pathophysiology of Lung Cancer
the detection of a resectable tumor [50]. In this regard, many
trials have been developed with the aim of characterizing a The different analyses performed with CTCs, in particular
molecular biological signature that provides information into for patients presenting with small cell or non-small cell lung
the primitive origin of a cancer [50]. These approaches are carcinoma, have improved knowledge into the natural his-
based mostly on characterization of circulating nucleic acids tory of these cancers and into the molecular anomalies,
or microRNAs, or protein signatures [50]. At present, it is which have led to envisage novel targeted therapeutics [12,
more difficult to use CTCs to orientate the diagnosis toward a 14]. These studies relied on many technological approaches
specific type of solid tumor. Some biomarkers can nonethe- and innovative methods.
less be tested with CTCs (P40 and TTF1 detected with ICC,
for example) in providing information concerning the primi-
tive pulmonary origin of these CTCs [8]. 15.7.1  etting up Different Mice Models
S
Using CTCs

15.6  irculating Tumor Cells as a Screening

C It was under the impetus of the team of Caroline Dive at the
Tool for Lung Cancer Paterson Cancer Institute in Manchester (UK) that models of
mice implanted with CTCs (or CDXs) from patients with
Recent studies have shown that it is possible to screen for small cell metastatic lung cancer were developed [31]. These
lung cancer, notably in a population at higher risk (subjects mice models have been used to study tumor progression and
aged over 50 years, heavy smokers, developing chronic response to certain treatments.
152 P. Hofman

15.7.2 Genomic Analysis of a Single CTC 15.8 Perspectives and New Issues

The capture of CTCs makes possible the detection of certain Studies into CTCs have opened up several short- and
genomic alterations of these cells and also the demonstration mid-term prospects for patients with lung cancer. Routine
of the heterogeneity of the circulating population of cancer detection and characterization of CTCs has been rapidly
cells, in terms of both phenotype and genotype [54]. Thus, a replaced for different goals, such as genomic alteration
number of applications have been developed with CTCs, in assessment, by the use of circulating nucleic acid in plasma
particular for small cell lung cancer [54]. Certain genetic since the tests on cfDNA can be automated, are reproducible,
profiles can be distinguished between chemo-sensitive and are in general sensitive and specific as well as less costly
and -resistant tumors [54]. [57]. The development of NGS approaches with cfDNA is
now possible and can be performed in clinical routine [4].
However, a few perspectives are now open for using CTCs in
15.7.3  nalysis of Interactions Between CTCs
A thoracic oncology.
and Circulating Hematological Cells

It is well established that CTCs interact, either directly after 15.8.1  etup of Procedures for Accreditation
S
cell contact or indirectly through the intermediary of secreted and Harmonization
molecules, with circulating hematological cells [55]. Different
populations of non-tumor cells circulate in the blood, includ- The International Liquid Biopsy Standardization Alliance
ing myeloid and lymphoid cells (T lymphocytes, natural (ILSA) is composed of organizations and foundations that
killer cells), monocytes, and macrophages, but also circulat- recognize the importance of working toward the global use
ing free endothelial cells and fibroblasts [55]. A strong cellu- of liquid biopsy in oncology to support clinical decision-­
lar interaction between CTCs and platelets also exists [55]. making and regulatory considerations and to seek to pro-
The different actors present in the blood circulation can also mote it in the community. An independent liquid biopsy- and
interact in a complex way with each other through direct standardization-­based program has been initiated by ILSA
physical contact via receptors or cytokines, but also by means [58]. The procedures of accreditation, notably the ISO
of transfer of microvesicles such as exosomes [55]. Thus, it 15189 norm, can be adapted to methods of detection of
may concern “a three- or four-way cell partnership” involving CTCs but the challenge strongly varies according to the
CTCs, myeloid cells, monocytes/macrophages, and platelets technique used. Some of the procedures of the ISO 15189
[55]. These cellular interactions may facilitate the progres- norm are certainly difficult to apply to clinical validation of
sion of CTCs and can protect them from anoikis or from the the detection of CTCs. These procedures require that an
cytotoxic effect of some circulating lymphoid cells (such as external quality assessment (EQA) be performed. Several
CD8 lymphocytes, natural killer cells, and T reg lympho- constraints can arise when standardizing methods of detec-
cytes). Moreover, this allows CTCs to adhere to the wall of tion and characterization of CTCs. One of the main chal-
vessels and favors their intravasation and extravasation with lenges certainly concerns mastering the pre-analytical
the development of metastases in different organs [55]. phase. It is also necessary to demonstrate the efficacy and
The newly discovered knowledge into interactions between added value of detection of CTCs in the clinic by perform-
CTCs and different cells present at the parenchymal level have ing comparative analyses and studies with different tests of
provided a better understanding of the implantation of CTCs in detection.
certain target organs and of the phenomenon of “self-seeding”
with enrichment of the primitive tumor environment, which
leads to renewed growth of the tumor or renewed progression 15.8.2 I ntegrate the Data on CTCs with that
of the tumor at the initial site of surgical resection. Obtained from the Studies of Other
Blood Components

15.7.4 Functional Analysis of CTCs Analyses of free circulating nucleic acids and the more dif-
ficult to perform detection and characterization of CTCs are
It is possible to evaluate the function of CTCs using certain often presented in opposition [4, 57]. Integrative biology is
approaches, in particular Epispot method [56]. Epispot can certainly a way to optimize and make some approaches more
detect proteins secreted/released/shed from single epithelial attractive and robust, notably for early detection of lung can-
cells, which distinguishes, for example, these viable cells cer, by associating radiological images and algorithms based
from apoptotic CTCs. on artificial intelligence [51].
15 Circulating Tumor Cells in Lung Cancer 153

15.8.3  emonstrate the Contribution of CTCs

D References
Analysis to the Clinical Routine
1. Zhang Y, Luo G, Etxeberria J, Hao Y. Global patterns and trends in
lung cancer incidence: a population-based study. J Thorac Oncol.
It is essential to demonstrate the added value of routine 2021;16(6):933–44. https://doi.org/10.1016/j.jtho.2021.01.1626.
detection of CTCs in thoracic oncology and in particular for 2. Heeke S, Benzaquen J, Hofman V, Ilié M, Allegra M, Long-Mira E,
improved care of patients. Notably, the question of integrat- Lassalle S, Tanga V, Salacroup C, Bonnetaud C, Fayada J, Gazoppi
ing the detection of CTCs in clinical trials has been raised L, Ribeyre L, Castelnau O, Garnier G, Cattet F, Nanni I, de Fraipont
F, Cohen C, Berthet JP, Leroy S, Poudenx M, Marquette CH, Denis
[59]. An important consideration concerns the tumor hetero- MG, Barlesi F, Hofman P. Critical assessment in routine clinical
geneity [33]. In fact, there exists a degree of disagreement practice of liquid biopsy for EGFR status testing in non-small-cell
between the NGS results obtained with circulating nucleic lung cancer: a single-laboratory experience (LPCE, Nice, France).
acids and tissue biopsies from the same patients. A comple- Clin Lung Cancer. 2020;21(1):56–65.e8. https://doi.org/10.1016/j.
cllc.2019.07.010.
mentary CTC analysis can provide additional information 3. Heeke S, Hofman V, Benzaquen J, Otto J, Tanga V, Zahaf K, Allegra
concerning the heterogeneity and the discordant results [41]. M, Long-Mira E, Lassalle S, Marquette CH, Ilie M, Hofman
One of the goals of using CTCs in daily practice could be to P. Detection of EGFR mutations from plasma of NSCLC patients
detect and monitor the minimal residual disease in early-­ using an automatic cartridge-based PCR system. Front Pharmacol.
2021;12:657743. https://doi.org/10.3389/fphar.2021.657743.
stage patients or patients after treatment. 4. Hofman P. Next-generation sequencing with liquid biopsies
from treatment-Naïve non-small cell lung carcinoma patients.
Cancers (Basel). 2021;13(9):2049. https://doi.org/10.3390/
15.9 Conclusion cancers13092049.
5. Alix-Panabières C. The future of liquid biopsy. Nature.
2020;579(7800):S9. https://doi.org/10.1038/d41586-­020-­00844-­5.
Studies into CTCs have improved the understanding of the 6. Pantel K. Blood tests for early detection of lung cancer: challenges
pathophysiology of lung cancer and the development of dif- and promises. Lancet Respir Med. 2020;8(7):654–6.
ferent treatments for patients. It is essential to integrate in a 7. Hofman V, Ilie M, Long E, Guibert N, Selva E, Washetine K,
Mograbi B, Mouroux J, Vénissac N, Reverso-Meinietti J, Milano
global perspective the analysis of the different components G, Mazières J, Marquette CH, Paterlini-Bréchot P, Hofman
circulating in the blood of patients with lung cancer and to P. Detection of circulating tumor cells from lung cancer patients in
not limit analyses to circulating free nucleic acids in plasma. the era of targeted therapy: promises, drawbacks and pitfalls. Curr
Improvements need to be made in the standardization of Mol Med. 2014;14(4):440–56.
8. Hofman VJ, Ilie M, Hofman PM. Detection and characterization
CTCs detection methods, and the characterization and quan- of circulating tumor cells in lung cancer: why and how? Cancer
tification of CTCs, in particular for thoracic oncology. The Cytopathol. 2016;124(6):380–7.
analysis of CTCs will provide a better understanding of the 9. Riethdorf S, O’Flaherty L, Hille C, Pantel K. Clinical applications
natural history of lung cancer and lead to the development of of the CellSearch platform in cancer patients. Adv Drug Deliv Rev.
2018;125:102–21.
innovative treatments. To this aim, it must be kept in mind 10. Xu M, Liu W, Zou K, Wei S, Zhang X, Li E, Wang Q. Design and
that the creation of biological collections associating CTCs clinical application of an integrated microfluidic device for circulat-
will allow different analyses and collaborative studies to be ing tumor cells isolation and single-cell analysis. Micromachines
performed [60]. (Basel). 2021;12(1):49.
11. Rolfo C, Mack PC, Scagliotti GV, Baas P, Barlesi F, Bivona TG,
Herbst RS, Mok TS, Peled N, Pirker R, Raez LE, Reck M, Riess
JW, Sequist LV, Shepherd FA, Sholl LM, Tan DSW, Wakelee
HA, Wistuba II, Wynes MW, Carbone DP, Hirsch FR, Gandara
Open Questions
DR. Liquid biopsy for advanced non-small cell lung cancer
• How to optimize the methods for a better detection (NSCLC): a statement paper from the IASLC. J Thorac Oncol.
and quantification of CTCs in thoracic oncology? 2018;13(9):1248–68.
• How to determine the metastatic potential of CTCs 12. Alix-Panabières C, Pantel K. Liquid biopsy: from discovery to
clinical application. Cancer Discov. 2021;11(4):858–73. https://
in thoracic oncology? doi.org/10.1158/2159-­8290.CD-­20-­1311.
• How to be able to better obtain CDXs models from 13. Cortés-Hernández LE, Eslami-S Z, Alix-Panabières C. Liquid
CTCs isolated from NSCLC patients? biopsy to detect circulating tumor cells: is it ready for a value propo-
• How to develop some tests combining at the same sition in laboratory medicine? J Appl Lab Med. 2020;5(5):1027–37.
14. Cortés-Hernández LE, Eslami-S Z, Pantel K, Alix-Panabières
time the analyses of CTCs and other blood compo- C. Molecular and functional characterization of circulating
nents for a better taking care of lung cancer tumor cells: from discovery to clinical application. Clin Chem.
patients? 2020;66(1):97–104.
• How to set up validation methods and how to obtain 15. Ross K, Pailler E, Faugeroux V, Taylor M, Oulhen M, Auger N,
Planchard D, Soria JC, Lindsay CR, Besse B, Vielh P, Farace F. The
an accreditation associated with CTCs detection in potential diagnostic power of circulating tumor cell analysis for non-­
thoracic oncology? small-­cell lung cancer. Expert Rev Mol Diagn. 2015;15(12):1605–
29. https://doi.org/10.1586/14737159.2015.1111139.
154 P. Hofman

16. Chemi F, Mohan S, Guevara T, Clipson A, Rothwell DG, Dive radical surgery for non-small-cell lung carcinoma: comparison
C. Early dissemination of circulating tumor cells: biological of the efficacy of the CellSearch Assay™ and the isolation by
and clinical insights. Front Oncol. 2021;11:672195. https://doi. size of epithelial tumor cell method. Int J Cancer. 2011;129(7):
org/10.3389/fonc.2021.672195. 1651–60.
17. Lindsay CR, Faugeroux V, Michiels S, Pailler E, Facchinetti F, Ou 27. Alix-Panabières C, Mader S, Pantel K. Epithelial-mesenchymal
D, Bluthgen MV, Pannet C, Ngo-Camus M, Bescher G, Caramella plasticity in circulating tumor cells. J Mol Med (Berl). 2017;95:133–
C, Billiot F, Remon J, Planchard D, Soria JC, Besse B, Farace 42. https://doi.org/10.1007/s00109-­016-­1500-­6.
F. A prospective examination of circulating tumor cell profiles 28. Hofman V, Long E, Ilie M, Bonnetaud C, Vignaud JM, Fléjou JF,
in non-small-cell lung cancer molecular subgroups. Ann Oncol. Lantuejoul S, Piaton E, Mourad N, Butori C, Selva E, Marquette
2017;28(7):1523–31. CH, Poudenx M, Sibon S, Kelhef S, Vénissac N, Jais JP, Mouroux J,
18. Lindsay CR, Blackhall FH, Carmel A, Fernandez-Gutierrez F, Molina TJ, Vielh P, Hofman P. Morphological analysis of circulat-
Gazzaniga P, Groen HJM, Hiltermann TJN, Krebs MG, Loges S, ing tumour cells in patients undergoing surgery for non-small cell
López-López R, Muinelo-Romay L, Pantel K, Priest L, Riethdorf lung carcinoma using the isolation by size of epithelial tumour cell
S, Rossi E, Terstappen L, Wikman H, Soria JC, Farace F, Renehan (ISET) method. Cytopathology. 2012;23(1):30–8.
A, Dive C, Besse B, Michiels S. EPAC-lung: pooled analysis of cir- 29. Manjunath Y, Mitchem JB, Suvilesh KN, Avella DM, Kimchi
culating tumour cells in advanced non-small cell lung cancer. Eur J ET, Staveley-O’Carroll KF, Deroche CB, Pantel K, Li G, Kaifi
Cancer. 2019;117:60–8. JT. Circulating Giant tumor-macrophage fusion cells are inde-
19. Hofman V, Bonnetaud C, Ilie MI, Vielh P, Vignaud JM, Fléjou JF, pendent prognosticators in patients with NSCLC. J Thorac Oncol.
Lantuejoul S, Piaton E, Mourad N, Butori C, Selva E, Poudenx 2020;15(9):1460–71.
M, Sibon S, Kelhef S, Vénissac N, Jais JP, Mouroux J, Molina 30. Chinniah C, Aguarin L, Cheng P, Decesaris C, Cutillo A, Berman
TJ, Hofman P. Preoperative circulating tumor cell detection using AT, Frick M, Doucette A, Cengel KA, Levin W, Hahn S, Dorsey JF,
the isolation by size of epithelial tumor cell method for patients Simone CB 2nd, Kao GD. Early detection of recurrence in patients
with lung cancer is a new prognostic biomarker. Clin Cancer Res. with locally advanced non-small-cell lung cancer via circulating
2011;17(4):827–35. tumor cell analysis. Clin Lung Cancer. 2019;20(5):384–90.
20. Hou JM, Krebs MG, Lancashire L, Sloane R, Backen A, Swain RK, 31. Blackhall F, Frese KK, Simpson K, Kilgour E, Brady G, Dive
Priest LJ, Greystoke A, Zhou C, Morris K, Ward T, Blackhall FH, C. Will liquid biopsies improve outcomes for patients with small-­
Dive C. Clinical significance and molecular characteristics of cir- cell lung cancer? Lancet Oncol. 2018;19(9):e470–e81.
culating tumor cells and circulating tumor microemboli in patients 32. Shishido SN, Carlsson A, Nieva J, Bethel K, Hicks JB, Bazhenova
with small-cell lung cancer. J Clin Oncol. 2012;30(5):525–32. L, Kuhn P. Circulating tumor cells as a response monitor in stage IV
https://doi.org/10.1200/JCO.2010.33.3716. non-small cell lung cancer. J Transl Med. 2019;17(1):294.
21. Krebs MG, Sloane R, Priest L, Lancashire L, Hou JM, Greystoke 33. Pawlikowska P, Faugeroux V, Oulhen M, Aberlenc A, Tayoun T,
A, Ward TH, Ferraldeschi R, Hughes A, Clack G, Ranson M, Pailler E, Farace F. Circulating tumor cells (CTCs) for the nonin-
Dive C, Blackhall FH. Evaluation and prognostic significance of vasive monitoring and personalization of non-small cell lung can-
circulating tumor cells in patients with non-small-cell lung can- cer (NSCLC) therapies. J Thorac Dis. 2019;11(Suppl 1):S45–56.
cer. J Clin Oncol. 2011;29(12):1556–63. https://doi.org/10.1200/ https://doi.org/10.21037/jtd.2018.12.80.
JCO.2010.28.7045. 34. Ge B, Wang Y, Lei S, Zhang J. Effect of phenotypic detection
22. Foy V, Lindsay CR, Carmel A, Fernandez-Gutierrez F, Krebs of circulating tumor cells marked by epithelial-mesenchymal
MG, Priest L, Carter M, Groen HJM, Hiltermann TJN, de Luca transformation on the prognosis of lung cancer: a protocol for
A, Farace F, Besse B, Terstappen L, Rossi E, Morabito A, Perrone systematic review and meta-analysis. Medicine (Baltimore).
F, Renehan A, Faivre-Finn C, Normanno N, Dive C, Blackhall F, 2020;99(44):e22960.
Michiels S. EPAC-lung: European pooled analysis of the prognos- 35. Ilié M, Szafer-Glusman E, Hofman V, Chamorey E, Lalvée S, Selva
tic value of circulating tumour cells in small cell lung cancer. Transl E, Leroy S, Marquette CH, Kowanetz M, Hedge P, Punnoose E,
Lung Cancer Res. 2021;10(4):1653–65. https://doi.org/10.21037/ Hofman P. Detection of PD-L1 in circulating tumor cells and white
tlcr-­20-­1061. blood cells from patients with advanced non-small-cell lung cancer.
23. Chemi F, Rothwell DG, McGranahan N, Gulati S, Abbosh C, Ann Oncol. 2018;29(1):193–9.
Pearce SP, Zhou C, Wilson GA, Jamal-Hanjani M, Birkbak N, 36. Sinoquet L, Jacot W, Gauthier L, Pouderoux S, Viala M, Cayrefourcq
Pierce J, Kim CS, Ferdous S, Burt DJ, Slane-Tan D, Gomes F, L, Quantin X, Alix-Panabières C. Programmed cell death ligand
Moore D, Shah R, Al Bakir M, Hiley C, Veeriah S, Summers Y, 1-expressing circulating tumor cells: a new prognostic biomarker
Crosbie P, Ward S, Mesquita B, Dynowski M, Biswas D, Tugwood in non-small cell lung cancer. Clin Chem. 2021;6:hvab131. https://
J, Blackhall F, Miller C, Hackshaw A, Brady G, Swanton C, Dive C, doi.org/10.1093/clinchem/hvab131.
TRACERx Consortium. Pulmonary venous circulating tumor cell 37. Murlidhar V, Reddy RM, Fouladdel S, Zhao L, Ishikawa MK,
dissemination before tumor resection and disease relapse. Nat Med. Grabauskiene S, Zhang Z, Lin J, Chang AC, Carrott P, Lynch WR,
2019;25(10):1534–9. Orringer MB, Kumar-Sinha C, Palanisamy N, Beer DG, Wicha
24. Crosbie PA, Shah R, Krysiak P, Zhou C, Morris K, Tugwood J, MS, Ramnath N, Azizi E, Nagrath S. Poor prognosis indicated by
Booton R, Blackhall F, Dive C. Circulating tumor cells detected venous circulating tumor cell clusters in early-stage lung cancers.
in the tumor-draining pulmonary vein are associated with disease Cancer Res. 2017;77(18):5194–206. https://doi.org/10.1158/0008-
recurrence after surgical resection of NSCLC. J Thorac Oncol. ­5472.CAN-­16-­2072.
2016;11(10):1793–7. https://doi.org/10.1016/j.jtho.2016.06.017. 38. Hofman P. Liquid biopsy and therapeutic targets: present and future
25. Farace F, Massard C, Vimond N, Drusch F, Jacques N, Billiot F, issues in thoracic oncology. Cancers (Basel). 2017;9(11):154.
Laplanche A, Chauchereau A, Lacroix L, Planchard D, Le Moulec 39. Hofman P, Heeke S, Alix-Panabières C, Pantel K. Liquid biopsy
S, André F, Fizazi K, Soria JC, Vielh P. A direct comparison of in the era of immuno-oncology: is it ready for prime-time use for
CellSearch and ISET for circulating tumour-cell detection in patients cancer patients? Ann Oncol. 2019;30(9):1448–59.
with metastatic carcinomas. Br J Cancer. 2011;105(6):847–53. 40. Maheswaran S, Sequist LV, Nagrath S, Ulkus L, Brannigan
26. Hofman V, Ilie MI, Long E, Selva E, Bonnetaud C, Molina T, B, Collura CV, Inserra E, Diederichs S, Iafrate AJ, Bell DW,
Vénissac N, Mouroux J, Vielh P, Hofman P. Detection of circu- Digumarthy S, Muzikansky A, Irimia D, Settleman J, Tompkins RG,
lating tumor cells as a prognostic factor in patients undergoing Lynch TJ, Toner M, Haber DA. Detection of mutations in EGFR in
15 Circulating Tumor Cells in Lung Cancer 155

circulating lung-cancer cells. N Engl J Med. 2008;359(4):366–77. of NSCLC patients and correlation with response to PD-1/PD-L1
https://doi.org/10.1056/NEJMoa0800668. inhibitors. Cancers (Basel). 2019;11(6):835.
41. Liu HE, Vuppalapaty M, Wilkerson C, Renier C, Chiu M, Lemaire 50. Calabrese F, Lunardi F, Pezzuto F, Fortarezza F, Vuljan SE,
C, Che J, Matsumoto M, Carroll J, Crouse S, Hanft VR, Jeffrey SS, Marquette C, Hofman P. Are there new biomarkers in tissue and liq-
Di Carlo D, Garon EB, Goldman J, Sollier E. Detection of EGFR uid biopsies for the early detection of non-small cell lung cancer? J
mutations in cfDNA and CTCs, and comparison to tumor tissue Clin Med. 2019;8(3):414. https://doi.org/10.3390/jcm8030414.
in non-small-cell-lung-cancer (NSCLC) patients. Front Oncol. 51. Benzaquen J, Boutros J, Marquette C, Delingette H, Hofman
2020;10:572895. P. Lung cancer screening, towards a multidimensional approach:
42. Catelain C, Pailler E, Oulhen M, Faugeroux V, Pommier AL, Farace why and how? Cancers (Basel). 2019;11(2):212. https://doi.
F. Detection of gene rearrangements in circulating tumor cells: org/10.3390/cancers11020212.
examples of ALK-, ROS1-, RET-rearrangements in non-small-cell 52. Ilie M, Hofman V, Long-Mira E, Selva E, Vignaud JM, Padovani B,
lung cancer and ERG-rearrangements in prostate cancer. Adv Exp Mouroux J, Marquette CH, Hofman P. “sentinel” circulating tumor
Med Biol. 2017;994:169–79. cells allow early diagnosis of lung cancer in patients with chronic
43. Ilie M, Mazieres J, Chamorey E, Heeke S, Benzaquen J, Thamphya B, obstructive pulmonary disease. PLoS One. 2014;9(10):e111597.
Boutros J, Tiotiu A, Fayada J, Cadranel J, Poudenx M, Moro-­Sibilot 53. Marquette CH, Boutros J, Benzaquen J, Ferreira M, Pastre J, Pison
D, Barlesi F, Thariat J, Clement-Duchene C, Tomasini P, Hofman C, Padovani B, Bettayeb F, Fallet V, Guibert N, Basille D, Ilie M,
V, Marquette CH, Hofman P, STALKLUNG01 Study Consortium Hofman V, Hofman P. AIR project study group. Lancet Respir Med.
Investigators. Prospective multi-center validation of the detection 2020;8(7):709–16.
of ALK rearrangements of circulating tumor cells for non-invasive 54. Carter L, Rothwell DG, Mesquita B, Smowton C, Leong HS, Fernandez-
longitudinal management of patients with advanced non-small Gutierrez F, Li Y, Burt DJ, Antonello J, Morrow CJ, Hodgkinson CL,
cell lung cancer. J Thorac Oncol. 2021;S1556-­0864(21):01664–6. Morris K, Priest L, Carter M, Miller C, Hughes A, Blackhall F, Dive C,
https://doi.org/10.1016/j.jtho.2021.01.1617. Brady G. Molecular analysis of circulating tumor cells identifies distinct
44. Ilie M, Szafer-Glusman E, Hofman V, Long-Mira E, Suttmann copy-number profiles in patients with chemosensitive and chemorefrac-
R, Darbonne W, Butori C, Lalvée S, Fayada J, Selva E, Yu W, tory small-cell lung cancer. Nat Med. 2017;23(1):114–9.
Marquette CH, Shames DS, Punnoose E, Hofman P. Expression 55. Heeke S, Mograbi B, Alix-Panabières C, Hofman P. Never travel
of MET in circulating tumor cells correlates with expression in alone: the crosstalk of circulating tumor cells and the blood micro-
tumor tissue from advanced-stage lung cancer patients. Oncotarget. environment. Cell. 2019;8(7):714.
2017;8(16):26112–21. 56. Alix-Panabières C. EPISPOT assay: detection of viable DTCs/
45. Hofman P. Detecting resistance to therapeutic ALK inhibitors in CTCs in solid tumor patients. Recent Results Cancer Res.
tumor tissue and liquid biopsy markers: an update to a clinical rou- 2012;195:69–76. https://doi.org/10.1007/978-­3-­642-­28160-­0_6.
tine practice. Cell. 2021;10(1):168. 57. Ilie M, Hofman V, Long E, Bordone O, Selva E, Washetine K,
46. Berger LA, Janning M, Velthaus JL, Ben-Batalla I, Schatz S, Falk Marquette CH, Hofman P. Current challenges for detection of
M, Iglauer P, Simon R, Cao R, Forcato C, Manaresi N, Bramlett circulating tumor cells and cell-free circulating nucleic acids,
K, Buson G, Hanssen A, Tiemann M, Sauter G, Bokemeyer C, and their characterization in non-small cell lung carcinoma
Riethdorf S, Reck M, Pantel K, Wikman H, Loges S. Identification patients. What is the best blood substrate for personalized medi-
of a high-level MET amplification in CTCs and cfTNA of an ALK-­ cine? Ann Transl Med. 2014;2(11):107. https://doi.org/10.3978/j.
positive NSCLC patient developing evasive resistance to Crizotinib. issn.2305-­5839.2014.08.11.
J Thorac Oncol. 2018;13(12):e243–e6. 58. Connors D, Allen J, Alvarez JD, Boyle J, Cristofanilli M, Hiller C,
47. Ilie M, Hofman V, Dietel M, Soria JC, Hofman P. Assessment Keating S, Kelloff G, Leiman L, McCormack R, Merino D, Morgan
of the PD-L1 status by immunohistochemistry: challenges and E, Pantel K, Rolfo C, Serrano MJ, Pia Sanzone A, Schlange T,
perspectives for therapeutic strategies in lung cancer patients. Sigman C, Stewart M. International liquid biopsy standardization
Virchows Arch. 2016;468(5):511–25. https://doi.org/10.1007/ alliance white paper. Crit Rev Oncol Hematol. 2020;156:103112.
s00428-­016-­1910-­4. 59. Massard C, Borget I, Farace F, Aspeslagh S, Le Deley MC, Le
48. Gosney JR, Boothman AM, Ratcliffe M, Kerr KM. Cytology for Tourneau C, Bidard FC, Pierga JY, Dieras V, Hofman P, Spano JP,
PD-L1 testing: a systematic review. Lung Cancer. 2020;141:101–6. Ferte C, Lacroix L, Soria JC. RECIST response and variation of
https://doi.org/10.1016/j.lungcan.2020.01.010. circulating tumour cells in phase 1 trials: a prospective multicentric
49. Janning M, Kobus F, Babayan A, Wikman H, Velthaus JL, study. Eur J Cancer. 2017;83:185–93.
Bergmann S, Schatz S, Falk M, Berger LA, Böttcher LM, Päsler 60. Nejlund S, Smith J, Kraan J, Stender H, Van MN, Langkjer ST,
S, Gorges TM, O’Flaherty L, Hille C, Joosse SA, Simon R, Nielsen MT, Sölétormos G, Hillig T. Cryopreservation of circulat-
Tiemann M, Bokemeyer C, Reck M, Riethdorf S, Pantel K, Loges ing tumor cells for enumeration and characterization. Biopreserv
S. Determination of PD-L1 expression in circulating tumor cells Biobank. 2016;14(4):330–7. https://doi.org/10.1089/bio.2015.007.
 CTCs/ctDNA and Brain Metastasis
16
S. Ray Kenney and Dario Marchetti

Abstract

An estimated 20–30% of all patients with cancer will Learning Objectives

develop brain metastasis, and this percentage increases at • Introduction of biological and clinical challenges
autopsy, with advanced stage and aggressive cancers aug- overcoming BM onset.
menting the risk. However, in spite of the severity of brain • Description of latest advances of CTC/ctDNA
metastasis as a daunting clinical problem, advances in assays and CTCs/ctDNA implications for clinical
systemic therapies have been limited in scope and effec- monitoring of BM and their therapy.
tiveness. Brain metastases occur via seeding of circulat- • Perspectives in applying liquid biopsy concepts by
ing tumor cells (CTCs) in the brain microvasculature. analyses of CTCs and ctDNA for predictive and
This biology guides increasing interest towards interro- preventive targets of BM.
gating CTCs and circulating tumor DNA (ctDNA) as • Overview on CTCs/ctDNA profiling to guide novel
tumor components of body fluids like blood and cerebral BM therapeutic strategies.
spinal fluid (CSF). Here, we provide a review on research
advances implicating CTCs and ctDNAs in relation to the
onset of brain metastasis, particularly in breast cancer.
Further, we introduce key biomarker discoveries on CTCs 16.1 Introduction
associated with breast cancer brain metastasis, and pro-
vide overviews of similar roles played by ctDNA as stud- Brain metastasis (BM) represents a significant clinical and
ies reported in the literature. Lastly, we discuss the biological challenge with highly devastating and lethal con-
scientific and technical challenges that remain to be over- sequences. An estimated 20% of all cancer patients will
come to implement the concept of liquid biopsy in the develop brain metastasis [1, 2]. However, this estimate is
clinic, notably using liquid biopsy to predict brain metas- likely an underestimate, since autopsy studies have sug-
tasis onset as guidance for a specific treatment, or to pre- gested a higher incidence, approximately 40%, of intracra-
vent secondary brain metastasis with more effective nial metastasis in patients with cancer [2, 3]. For example,
therapies. brain metastasis frequently occurs as an initial site in meta-
static melanoma with diagnosis in approximately 60% of
patients, and up to 80% of patients at autopsy [3]. Lung,
melanoma, and breast cancer are the cancer types that most
frequently metastasize to brain, although others include
colorectal, pancreatic, renal, and prostate carcinomas [1, 2].
Notably, the overall 2- and 5-year survival rates for patients
diagnosed with BM remain dismal at 8.1% and 2.4%, respec-
S. R. Kenney
tively, with a median overall survival of only 4 months from
RareCyte, Inc., Seattle, WA, USA
e-mail: [email protected] diagnosis [4, 5]. To date, localized BM therapies include
neurological resection of BM lesions or stereotactic radio-
D. Marchetti (*)
Department of Internal Medicine, Division of Molecular Medicine, surgery, while widespread disease includes whole brain
The University of New Mexico Health Sciences Center, radiotherapy, targeted therapy, or immunotherapy. While the
Albuquerque, NM, USA latter has shown some significant activity, it is often less than
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 157
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_16
158 S. R. Kenney and D. Marchetti

what observed for patients with only extracranial metastases Similarly, alternative approaches to CellSearch based on
[6, 7]. Furthermore, standard treatments result in high neuro- procedures involving density, immunomagnetic separation,
toxicity (45% of patients), exacerbated by the fact that the size exclusion, and/or flow manipulation in microfluidic
blood-brain barrier (BBB) impairs the ability of chemothera- devices are limited by the heterogeneous nature of CTCs,
peutic approaches to achieve therapeutic concentrations in often not capturing EpCAM− CTCs [10]. This is because a
the brain/central nervous system (CNS). significant fraction of CTCs that extravasate and are able to
Over the past several years, significant advances have generate distant metastases may have lost EpCAM expres-
been made in the understanding of the biology of circulating sion and undergone epithelial-mesenchymal transition
tumor cells (CTCs), the smallest functional units of cancer (EMT), expressing several EMT and/or cancer stem cell bio-
and “seeds” of fatal metastatic disease [8], and circulating markers [10]. Further, it has been reported that CellSearch
cell-free tumor DNA (ctDNA) [9–12]. Notably, newly dis- is unable to capture CTCs in 30–35% of metastatic breast
covered biomarkers of CTC subsets and/or ctDNA signa- cancer patients and in more than 60% of patients with breast
tures may be associated with BM. The objective of this cancer BM (BCBM) [13, 14]. To this end, we developed new
review is to outline current CTC/ctDNA-based approaches approaches to identify and characterize EpCAM− CTCs in
linked to BM onset and biology. We also describe the patients clinically diagnosed with or without BCBM, detect-
advances made that might foster the implementation of these able by MRI as standard of care [15].
technologies in new clinical trials designed to evaluate the Our focus has been to investigate biomarkers expressed
efficacy of CTC/ctDNA biomarkers as novel assays, thus by those EpCAM− CTCs in their abilities to induce BCBM
implementing the concept of liquid biopsy (LB) that can be [16, 17]; moving beyond CTC enumeration to demonstrate
applied either to predict or to prevent BM through targeted that CTCs can be isolated, cause new tumors, and be charac-
therapies. terized as brain-homing CTC subsets. We reported the isola-
tion of EpCAM− CTCs from patients with metastatic breast
cancer using multiparametric flow cytometry, then assessed
16.2  TCs Induce Breast Cancer Brain
C these EpCAM− CTCs for metastatic competence by select-
Metastasis: The BMSM Signature ing for biomarkers HPSE+/Notch1+/EGFR+/HER2+, which
we termed the Brain Metastasis Selected Marker (BMSM)
CTCs are essential components of the metastatic cascade, signature. We showed that CTCs selected for this signature
and crucial not only for understanding mechanisms of cancer are predisposed to metastasize to the brain and lungs once
metastasis but also for developing new liquid biopsy assays injected in immunodeficient animals [16, 17] (Fig. 16.1).
and applying them for initial diagnosis, monitoring of treat- The CTC BMSM signature is significant for two reasons: (1)
ment efficacy, and for assessment of therapy efficacy. These It is the result of a critical evaluation of CTC biomarkers cur-
concepts are particularly important in BM onset, where the rently in use; (2) It demonstrates that many biomarkers cur-
known characteristics of brain as target organ, including the rently used in the clinic, e.g., EpCAM, may not precisely
presence of a high microvascular network, circulatory arrest identify all clinically relevant CTCs. This suggests that
due to decreased blood flow in capillaries, specific CTC-­brain CTCs possessing brain specific metastatic competence must
endothelial interactions favoring adhesion to/invasion into be examined to better understand mechanisms of CTC-­
brain microenvironment, can facilitate an increased exposure driven metastasis vs. clinical dormancy.
to CTCs. However, the comprehensive ­characterization of
CTCs inducing BM remains incomplete, owing to CTC rar-
ity and fragility, high phenotypic heterogeneity, and a lack of 16.3  he Characterization of CTCs Related
T
universal CTC biomarkers; coupled with CTC technologies to Breast Cancer Dormancy
which cannot capture and evaluate the entire spectrum of
BM-relevant CTCs. For example, the CellSearch™ platform, The molecular switch to differentiate quiescence vs. meta-
the only CTC in vitro diagnostic platform approved for use by static competence of CTCs depends on the cross talk between
the U.S. Food and Drug Administration, relies on antibodies CTCs and the tumor microenvironment. We posited that spe-
targeting the epithelial cell adhesion molecule (EpCAM) and cific EpCAM-negative CTC subpopulations, shed from the
pan-cytokeratin (CK). Therefore, it is only capable of captur- primary or metastatic tumors, avoid organ arrest with extreme
ing EpCAM+/CK+ CTCs, but not CTCs that are EpCAM/ efficiency by the concomitant presence of stem cell and qui-
CK-independent, e.g., EpCAM-low or EpCAM-negative. escence properties. For example, previous studies have
16 CTCs/ctDNA and Brain Metastasis 159

established the presence of two neoplastic biomarkers, uro- CD24−) by multiparametric flow cytometry with alternative
kinase plasminogen activator receptor (uPAR) and integrin expression of uPAR and ITGβ1: uPAR+/β1+, uPAR−/β1−,
β1 (ITGβ1), promoting tumor cell growth and proliferation uPAR+/β1−, or uPAR−/β1+ [20, 21]. We investigated these
when they interact with the extracellular brain microenviron- CTC subsets for cell adhesion and proliferation properties,
ment, with loss of uPAR and ITGβ1expression, causing a and for their abilities to generate in vitro 3D CTC tumor-
shift from an invasive/metastatic to a dormant state [18–20]. spheres, and invade the extracellular matrix. We discovered
We have reported the isolation of EpCAM-negative breast that these properties were dependent upon uPAR/ITGβ1
cancer CTC subsets containing stem-cell properties (CD44+/ expression, and propose the idea of uPAR and ITGβ1 as bio-

a ~ 50 mls Peripheral Blood

7.5 ml 40ml

PBMC Isolation

ALDH1 CK 5/G/18 Multiparametric FACS (12 ch) BioViewTM

EpCAM+ CTC enumeration

EpCAM+ CTC profiling 400x 400x
EGFR+ EGFR+ EGFR+ EGFR+ EGFR
DAPI Merge

HER2+ HER2+ HER2– HER2–

Merge CK DAPI CD45 EpCAM
ALDH1+ ALDH1– ALDH1+ ALDH1–
400x 400x etc. etc. etc. etc.
EpCAM Vimentin

ZR-75(EpCAM+) EpCAM–CTCs
400x 400x

New CTC RT-PCR

CellSearchTM
231BR
CTC-1

CTC-2

CTC-3
MDA-

MDA-
231P

MW
(kDa)
50 CK14

EpCAM- CTC injection in xenografts

AE1

CK16
36 CK19

42 β-actin

CTC metastatic competency

IF/Westerns

Fig. 16.1 The isolation of EpCAM-negative CTCs from blood of tion (EMT), cytokeratins (CK), and breast cancer stem cell biomarkers
metastatic breast cancer patients using multiparametric flow (CD44high/CD24low, and CD44 variant v8). PBMCs isolated from
cytometry. (a) Panel shows the isolation of peripheral blood mononu- patients with (patient) or without (control) breast cancer were used as
clear cells (PBMCs) from patients’ blood samples which was followed negative controls for the BMSM CTC signature. In addition, human
by multiparametric flow sorting using the negative selection of brain metastatic [MDA-MB-231BRHER2 (brain tropic clone derived
EpCAM-expressing PBMCs, in synergy with the positive selection of from MDA-MB231 parental and HER2 transduced], poorly brain meta-
biomarkers (HER2, EGFR), known to relate either with BCBM, or to static (MDA-MB-231 parental), and non-­metastatic (MCF-7) breast
breast cancer cell stemness (ALDH1), and combinatorial. This strategy cancer cell lines were used as BMSM biomarkers controls.
resulted in the isolation of EpCAM-negative CTCs which were subse- Glyceraldehyde-3-phosphate dehydrogenase was the loading control.
quently shown to be of neoplastic origin by genetic analyses, and for (Bottom panel) Relative gene expression analyses of the BMSM CTC
survival/growth abilities as in vitro cultures. CTC cultures (CTC-1, profile in EpCAM-negative CTCs. MDA-MB-231BR, MDA-MB-
CTC-2, CTC-3) were then interrogated applying a newly developed 231P, and SKBR3 cell lines were used as BMSM CTC profile controls.
CTC-PCR assay (Panel b). Several positive/negative controls were per- (c) (Top left panels) Representative images of BCBM (insets) induced
formed to demonstrate EpCAM-negative CTC selection. They con- by EpCAM-negative BMSM+ CTCs when injected intracardially into
sisted to either use parallel CellSearch profiling to capture mice. BCBM sections underwent analyses for tumor burden by the CRi
EpCAM-positive CTCs within the same blood sample, the visualiza- Vectra Intelligent system, which highlights and enumerates single
tion of BCBM CTC biomarkers by immunofluorescence microscopy tumor cells (in green). (Bottom left panels) Representative immunos-
(IF)/Western blotting, or the interrogation of EGFR overexpression by taining for BMSM biomarker EGFR in CTC-driven BCBM by IHC of
BioView™ platform. EpCAM-negative CTC cultures expressing brain macro/micrometastasis. Specific staining for metastatic lesions
BCBM biomarkers were then injected in immunodeficient mice to eval- vs. adjacent brain tissue (negative control) is shown. The same results
uate their metastatic competence, notably to brain [16]. (b) (Top panel) were found for the other biomarkers of the BMSM CTC profile [16].
RT-PCR analyses of isolated EpCAM-negative CTC cultures (CTC-1, Quantification of IHC staining intensity for the BMSM CTC profile in
CTC-2, CTC-3). Selected genes were classified into four groups: BCBM compared to lung metastasis within the same CTC-injected ani-
HPSE/Notch1/EGFR/HER2 gene expression profile (termed the mal. Data are means ± SEM (n = 8 sections per mouse; 10 mice per
BMSM CTC signature), biomarkers for epithelial mesenchymal transi- CTC culture)
160 S. R. Kenney and D. Marchetti

b
The BMSM CTC Profile EMT Cytokeratin Stem cell

EpCAM
No primer

Notch1

Vimentin

CD44 v8
EGFR
HPSE
GAPDH

HER2

CD45

CD24

CD44
CK18

CK19

CK20
Twist

CK8
CTC-1

CTC-2

CTC-3
Patient PBMCs

Control PBMCs

MDA-MB-231BRHER2

MDA-MB-231P

MCF-7

30
HPSE Notch1 EpCAM EGFR HER2
Rel Gene Exp

15

0
CTC-1 CTC-2 CTC-3 MDA-MB-231 BR MDA-MB-231P SKBR3

c
Breast Cancer Brain Metastasis

Brain Lung

4
BMSM CTC-1 BMSM CTC-2 BMSM CTC-3
IHC staining intensity

200X 200X

40X 40X 0
BMSM CTC-1 BMSM CTC-2 BMSM CTC-3

100X 100X

Fig. 16.1 (continued)

markers of the CTC “dormancy axis.” Cell adhesion, prolif- genetic mutations at the single CTC level confirmed the high
eration, and invasion properties of EpCAM− CTC subsets heterogeneity of CTCs, it can provide a better approach to
were found to be relevant to BCBM onset. Of note, because evaluate the biology of CTCs by targeting these mutations
heterogeneous populations of CTCs harbor genetic and epi- and assessing their impact. These findings are significant to
genetic changes at single-cell level and exhibit distinct breast clinical dormancy, since CTCs exhibit various phenotypes
cancer phenotypes, we used the DEPArray™ single-cell iso- via mechanisms of expansion which are still unclear. These
lation platform to isolate individual CTCs derived from phenotypes are dependent on the presence of specific bio-
BCBM vs. No BCBM, along with applying MassARRAY™ markers, such as the expression of uPAR and ITGβ1 axis, in
mutation analyses. We detected the common COSMIC muta- combination with other markers. Multiple CTC phenotypes
tion PRKCB G785T in patient-derived CTCs, regardless of must exist: they resist apoptosis, undergo evolution and
the expression markers analyzed (uPAR/ITGβ1/HER2, etc.) clonal selection via DNA damage and active DNA repair
or BCBM status [20]. Accordingly, while the variability of pathways, and avoid arrest and adhesion to target organs
16 CTCs/ctDNA and Brain Metastasis 161

with extreme efficiency. Specific CTC clones within the 27]. The CTC gene signature was uniformly increased across
uPAR/ITGβ1 dormancy axis may therefore undergo prolif- pBC subtypes. These findings were reflected in cellular func-
eration and expansion for a long-term niche pool, while CTC tions as CTC upregulated genes were involved in cell death,
clustering can occur leading to secondary tumorsphere for- apoptosis, and cell survival with a concomitant decrease in
mation. Accordingly, the expansion of investigations estab- cellular proliferation and invasive properties, thus substanti-
lishing consistent in vitro CTC cultures for effective drug ating the concept that most CTCs exist in a state of reduced
screening/testing rather than in vivo transplantation will be a metabolic and/or mitotic activity, enabling them to overcome
clinically useful tool in applications of personalized stress and survive in circulation [22, 23].
medicine. Second, considering that Ki67 is routinely used for
assessing the proliferative index of cancer tissues, and is the
single most important prognostic factor for breast cancer
16.4  he Identification of Biomarker
T brain metastasis [28], immunostaining of Ki67 in isolated
Properties and the Proliferative State CTCs revealed distinct, nuclear-localized puncta, in both
of CTCs Associated with BCBM PanCK+ and CD44+/CD24− CTC subtypes. This staining
was quantified at the single-cell level using the DEPArray
Because of findings which have established the relevance of platform. Next, because of our investigations into the dor-
EpCAM– CTCs to BCBM onset, we expanded upon the mancy axis using uPAR/ITGβ1 in EpCAM-negative CTCs
characterization of de-­ differentiated, EpCAM-agnostic [20, 21], we evaluated their cell surface expression. Despite
CTCs, including both epithelial and stem-like breast cancer the significant concordance between Ki67 and uPAR/ITGβ1
CTC subsets. We devised a multiparametric workflow made expression in CTCs, there was a low level of discordance
of three sequential steps—(i) doublet discrimination and between Ki67 and uPAR/ITGβ1 expression in some indi-
dead cell elimination, (ii) depletion of hematopoietic-lineage vidual CTCs. Considering that both immunostaining and the
positive cells normally present in the peripheral blood, and DEPArray platform are only semi-quantitative, we designed
(iii) positive selection of PanCK+ (epithelial) or CD44+/ a single-­step flow cytometry method to enumerate Ki67+ and
CD24− (stem-like) CTCs. Immunocytochemistry confirmed uPAR+/ITGβ1+ CTCs in patient blood. Data analyses dem-
the selected CTC subsets from this multiparametric flow onstrated that BCBM patients had twice as many Ki67High
cytometry approach expressed appropriate biomarkers used CTCs (ratio of Ki67High:Ki67Low ~ 2:1), whereas patients
for selection. Isolated single CTCs were further interrogated without BCBM had 60% greater Ki67Low CTCs (ratio of
using the DEPArray platform, evaluating the expression of Ki67High:Ki67Low ~ 1:1.6) [22, 23]. We also compared the
PanCK and CD44, with an absence of CD45 and CD24. The ratio of uPAR-/ITGβ1− vs. uPAR+/ITGβ1+ CTCs in these
identity of isolated CTCs was validated by genomic and patient samples and found a 2.4-fold increase of uPAR+/
mRNA analyses of these cells [22, 23]. By employing these ITGβ1+ CTCs in BCBM patients. Consistent with our previ-
approaches, we isolated between 101 and 839 CTCs/8 mL of ous CTC analyses, we found that the Ki67 status of 20–30%
patient blood, whereas CellSearch identified between 0 and CTCs do not correlate with their uPAR/ITGβ1expression and
88 CTCs/7.5 mL blood when processed in parallel, implying vice versa, emphasizing the complementary roles of these
an ability to capture a larger CTC pool that would have oth- biomarkers defining distinct CTC subsets. Collectively, dif-
erwise remained undetected. Of note, transcriptional profil- ferences in biomarker expression of these CTC subsets indi-
ing on CTCs from BC patients with or without BCBM, as cate that at least a portion of CTCs in BCBM patient blood
confirmed by MRI, was compared to curated data, using the were unique from the remaining population in regard to their
same platform and pathway analysis, from primary patient biology and behavior [22, 23].
samples of various BC subtypes. We were able to make three Third, we postulated that gene expression profile differ-
remarkable observations from these analyses. ences between these two groups would arise specifically
First, 29,758 genes were downregulated in CTCs com- from brain metastatic CTCs. Using the presence of MRI
pared to 1972 genes upregulated (fold-change <−2 or >2, identified brain metastases as the clinical discriminator of
ANOVA p-value <0.05), indicative of generalized low CTC the study, CTC gene signatures from BCBM vs. No BCBM
transcriptional activity, suggesting that most CTCs are quies- patients clustered according to respective clinical groups. We
cent or undergoing apoptosis. Interestingly, none of the discovered a unique 126-gene signature which was signifi-
CTCs clustered with their corresponding primary breast can- cantly altered between these two groups, which we desig-
cer (pBC) molecular subtype [24, 25]. Conversely, CTCs nated the BCBM CTC signature (Fig. 16.2). This signature
clustered together as a single group, suggesting they resem- not only contained the four-gene BMSM CTC profile previ-
ble each other more closely than their pBC counterparts [26, ously identified [16], but also that individual Ki67+ CTCs
162 S. R. Kenney and D. Marchetti

a CTC transcriptomes b
BCBM NO BCBM
Ingenuity
-log Functions Activation
Canonical Ratio Molecules p-Value Molecules
(p-value) Annotation z-score
Pathways
3.21 19.81
Communication
between CXCL8, cell movement ADAM17,AIF1,CD86,
Innate and 3.09E00 3.37E-02 CD86, of mononuclear 2.09E-04 2.244 CXCL8,CXCR4,IFNGR1,
Adaptive IGHA1 leukocytes MAP3K2
NOTCH1 Immune Cells ADAM17,AIF1,
NODAL chemotaxis of
2.14E-03 2.129 ARHGEF7,CXCL,
ADAM17 leukocytes
CXCR4
RBPJ Notch Signaling 2.55E00 5.26E-02 ADAM17, ADAM17,AIF1,AREG,
CXCL8 RBPJ ARHGEF,CD86,CXCL8,
CXCR4 migration of cells 7.69E-04 2.1
CXCR4,DDIT4,IFN
CD86 GR1,MAP3K2,
Hematopoiesis MARCKS,NAMPT,
CXCL8,
from Pluripotent 2.37E00 4.26E-02 SLC3A2,TNFAIP3
IGHA1
Stem Cells

ADAM17,ARHGEF7,
Primary CD86,CXCL8,CXCR4,
IGHA1, cellular
Immunodeficiency 2.35E00 4.17E-02 4.12E-04 2.069 DDIT4,GABARAPL1,IF
TAP2 homeostasis
Signaling NGR1,mir-506,NAMPT,
RBPJ,SLC3A2,TNFAIP3

Nur77 Singaling CD86, Lymphocyte ADAM17,CD86,CXCL8,

in T Lymphocytes 2.21E00 3.51E-02 MAP3K2 migration 2.50E-04 2.018 CXCR4,IFNGR1,MAP3K2

Analyses performed with IPA software (Qiagen) using genes with <>2 fold -change.

Fig. 16.2 Difference in transcriptomic signatures of CTCs isolated bers of the BMSM profile but also for the Notch pathway/cell stemness
from blood of breast cancer patients diagnosed with or without biomarkers, along with chemokines involved in communications with
brain metastasis (BCBM/No BCBM). (a) Heat map showing CTC immune system cells (displayed in red). (b) Table shows activated
transcriptomes in BCBM vs. No BCBM and the discovery of a 126-­ canonical pathways and cellular functions in BCBM CTCs. Integrated
gene CTC signature found to be upregulated between BCBM and No pathway analyses indicate highest activation of pathways involved in
BCBM, using the Affymetrix HTA2.0 gene panel. Clustering was per- communications with cells of the immune system: Notch signaling and
formed using genes with fold change <−2 or >2, ANOVA p-value pluripotent stem cells [20, 22, 37]
<0.05. Of note, highest gene expression was detected not only for mem-

recapitulate this 126-gene signature. Pathway enrichment inroads have been recently made of a comprehensive analysis
analyses revealed higher activation of known CTC pathways, for distinct blood circulatory cell populations from patients
e.g., Notch, along with the discovery of novel hematopoietic diagnosed with metastatic/brain-metastatic BC. This was per-
and immune evasion signaling pathways in CTCs derived formed at the single-cell level (10x Genomics Chromium), not
from blood of BCBM patients. This was further validated by only identifying distinct immune cell clusters but also a cluster
performing Notch1 immunocytochemistry on patient-­ containing circulatory neoplastic cells, that is, CTCs [29].
derived CTCs isolated by DEPArray. A total of 22 CTCs These advances extend the definition of what a CTC is, and
from 3 patients diagnosed with BCBM and 17 CTCs from 3 suggest that the neoplastic cell capacity is distributed across
patients with No BCBM were analyzed. In the BCBM group, heterogeneous cell profiles in circulation. They will be highly
~72% (16/22) CTCs were positive for Notch staining, instrumental to replace rigid definitions (eg, presence/absence
whereas 24% (4/17) CTCs were positive for Notch1 staining of EpCAM, etc.) with more solid notions on how a CTC func-
in the No BCBM group. tions in their metastatic/brain-metastatic potency [29].
These results suggest that upregulation of Notch activity is
a feature of BCBM CTCs rather than CD44+/CD24− CTCs.
Importantly, cellular functional annotations associated with 16.5 The Characterization of ctDNA
distant metastases such as cell migration and chemotaxis, in Brain Metastasis
which were downregulated in the 10-patient CTC-pBC cohort,
were significantly activated in the BCBM vs. No BCBM CTC Blood plasma cell-free DNA (cfDNA) was first reported in
cohorts; along with an increase of pro-­inflammatory chemo- 1948, and the correlation between increased cfDNA and can-
kines (TNF, IL1β, and NF-κB), immunomodulatory networks cer was discovered in 1977 [11, 12]. Circulating tumor DNA
(CXCL8, CXCR4, CD86), and mitogenic growth factors (ctDNA), a subset of cfDNA, is derived from tumors as
(PDGF-BB). Taken together, these findings suggest that active secretion in extracellular vesicles, leakage after tumor
growth-arrested CTCs represent a substantial portion of the necrosis or apoptosis, lysis of CTC during micrometastasis,
CTC population, accounting for the anti-proliferative signa- and macrophage digestion of tumor DNA fragments and
ture of CTC cohorts. However, further segregation of the par- concentration varies due to a number of factors including
ent cohort into BCBM vs. No BCBM highlighted the gene stage of disease progression, tumor location, and vascular-
signature of transcriptionally and mitotically active CTCs ization [12, 30]. While it is known that higher levels of circu-
associated with the BCBM phenotype. To this end, important lating tumor DNA (ctDNA) at diagnosis predict poorer
16 CTCs/ctDNA and Brain Metastasis 163

patient outcome [16, 25, 30, 31], there has been significant another study, looking at paired detection of CTCs by cytol-
advancement in the use of ctDNA as prognostic or monitor- ogy and ctDNA in patients with primary or secondary brain
ing tools in several cancer types, including breast, lung, and lesions, ctDNA was present in all 58 patient samples, while
cancers of the central nervous system (CNS) [32, 33]. CTCs were only present in 76% [23]. The use of NGS to
The use of plasma ctDNA presents advantages, as well as detect mutant allelic frequency (MAF) had two important
limitations, as a diagnostic tool. First, plasma can be isolated findings. First, 68 of the 143 cancer associated genes had
from the same sample as CTCs, presenting low additional mutations in at least one BM CSF sample, but all patient
risk to patients, and allowing for a genomic comparison of samples carried at least one mutation. Second, CSF ctDNA
cfDNA, CTC DNA, and specific ctDNA [33, 34]. These highly correlated with number of CTCs, indicating that CSF
comparisons allow for longitudinal tracking of disease pro- ctDNA provides a better marker for disease progression than
gression and genetic changes at metastatic sites. A second CTCs. Melanoma patients with progressive BM disease pos-
aspect of ctDNA that must be considered is stability and sessed BRAF and MRAS mutations in CSF and 30% of the
turnover. Estimates of ctDNA clearance from the body range patients interpreted negative by cytology were positive
from 1 to 13 h as detected by next-generation sequencing through ctDNA analysis [15, 34]. Similarly, in patients with
(NGS) [30]. Although ctDNA is rapidly cleared, the high BC or melanoma, quantification of CTCs or ctDNA was
levels of cell death and turnover in tumors contribute a steady similarly associated with secondary LM disease and a ctDNA
supply of ctDNA. Patients with various localized cancer concentration of 0.022 ng/mL was correlated with an
types present with ctDNA approximately 55% of the time increased risk of death [1]. In a recent study, researchers at
[34]. In patients with localized breast, colon, pancreas, or Toronto Hospital’s Neuroscience Center generated cell-free
gastroesophageal cancers, detectable ctDNA was present in methylated DNA immunoprecipitation (cfMeDIP-seq), a
49–78% of patients, compared to 86–100% of patients with high-throughput sequencing profile of plasma samples from
known metastatic disease. However, in patients with primary patients presenting with a variety of brain cancers, including
brain lesions, detection of plasma ctDNA fell to approxi- BM [36]. They demonstrated not only that this platform can
mately 20%. In the case of primary brain lesions, plasma distinguish brain tumors from extracranial cancer types and
ctDNA was only detectable if there were visceral metastases healthy controls, but also that it can discriminate between
[31]. In patients with localized brain lesions only, ctDNA different tumors of the CNS that often look undistinguish-
was only detectable in cerebrospinal fluid (CSF). In individ- able by brain scan.
uals with primary or metastatic lesions within the CNS, While multiple different techniques have been employed
detection of ctDNA within the CSF is more sensitive and to detect ctDNA, they have all been shown to be similar in
precise than that of plasma [31, 34, 35]. These studies sug- sensitivity [34]. However, these techniques are not without
gest that liquid biopsy of CSF is better suited for the detec- drawbacks, as most are highly specific and look at individual
tion and monitoring of primary and metastatic lesions within cancer-related mutations [10, 15, 35], or the need to have a
the CNS. biopsy or CTCs presence to confirm the source of ctDNA
Quantification of ctDNA from plasma correlates with [15, 34]. There are advantages in ctDNA testing. For exam-
patient survival [32, 34, 35]. In a study of early-stage BC ple, detection of CTCs from brain metastasis can fail [13–15,
patients, detection of ctDNA was positive and prognostic in 34], likely due to segregation within the CSF. Conversely,
a multivariate model, with ~80% of patients with detectable ctDNA is detectable in plasma at low levels, even in the
ctDNA developing progressive disease, and levels of ctDNA absence of CTCs. Further, in the presence of brain lesions,
correlating with subtype. For example, TNBC patients had CSF ctDNA concentrations are much higher than in plasma
highest levels of ctDNA (4.96 copies/mL), HER2+ had inter- [15, 31, 34]. Taken together, these studies provide evidence
mediate levels (0.81 copies/mL), and ER+/HER2− had low- of clinical validity for the use of ctDNA in progressive dis-
est (0 copies/mL) [35]. In a study of non-small cell lung ease towards or within the brain. However, similar to the use
cancer, patients had lower ctDNA levels at first follow-up of CTCs, they have yet to demonstrate clinical utility towards
and best radiological response, than at baseline, while BM detection, monitoring, or therapies.
patients with stable or progressive disease showed no change
in ctDNA levels, indicating ctDNA levels change in response
to treatment [32]. Recently, quantification of ctDNA isolated 16.6 Conclusion
from CSF has been shown to be similarly prognostic of
patient outcome [1, 10, 23, 31]. Additionally, quantification Although the application of novel strategies for cancer diag-
of CSF ctDNA is more sensitive and accurate than plasma nosis, treatment, and prevention has resulted in many suc-
ctDNA in detection of brain lesions [1, 15, 23, 31, 34]. The cesses, converting some neoplastic conditions from incurable
use of next-generation sequencing (NGS) detected ctDNA in to curable, many challenges still remain. The concept of liq-
all patient CSF samples and allowed for clustering of molec- uid biopsy to confront the problem of early detection of can-
ular subtypes which correlated with overall survival [10]. In cer and its prevention was born on this premise. Liquid
164 S. R. Kenney and D. Marchetti

Fig. 16.3 Translational CTC Burden

perspectives for future
applications of novel CTC
tests to predict or to
monitor/prevent BCBM.
The figure shows the clinical
implications of BCBM-­
associated CTC biomarkers
Ki67+/uPAR+/ITGβ1+ and/or
Notch1+, and their
enumeration in relation to
BCBM. Enumeration of
BCBM CTCs can be used as a
sensitive screening method
for micrometastasis detection
Blood
in patients at risk of
developing BCBM or who
have BCBM prior to detection
with MRI. Conversely, serial
analyses of BCBM-associated
CTC biomarkers can be used
as a clinically useful tool to Brain
evaluate responses to therapy
for patients with overt BCBM
disease MRI Treatment

Multi-clonal cell Nascent brain Overt brain Monitoring

seeding in brain metastases metastases therapy efficacy
MRI undetectable MRI detectable

biopsy is particularly appealing and has remarkable poten- conclusion, we foresee that the extension of CTC/ctDNA
tial, considering the many platforms and technologies that analyses beyond BCBM and their application will equip cli-
have evolved allowing its implementation [8, 11, 12, 29]. nicians with critical information on the “BM state” in the
Complementing our investigations on BCBM-associated patient, allowing BM detection in its nascent stages. The
CTCs and signaling mechanisms potentiating CTC charac- information from liquid biopsy tests will aid in formulating
teristics which allow colonization of the brain, the studies rational therapies targeted specifically for brain metastasis
presented here have important clinical implications, which and evaluate their efficacy for the benefit of the patient.
can be further enhanced and/or complemented by ctDNA
analyses relating to BM onset. Building upon these studies,
we present two important directions for liquid biopsy: (1) the
application of novel liquid biopsy tests in the clinic as a sen- Open Questions
sitive screening method for detection of micrometastatic • How can CTC/ctDNA testing be improved and
brain disease which is undetectable by MRI, and (2) the use complemented for the prediction or prevention of
of the newly identified CTC/ctDNA signatures as tools for BM?
measuring responses to therapy for patients with MRI-­ • Can CTC/ctDNA testing be associated directly with
detectable BCBM (Fig. 16.3). As outcomes of our CTC stud- responses to treatment modalities such as focal or
ies and after further validation of the BCBM CTC signature, whole-brain radiation therapy?
CTC tests could be implemented in the clinic to correlate • Is there a common BM CTC/ctDNA profile across
CTCs/biomarkers of the CTC signature with the detection of all cancer types which increase the risk of BM,
brain micrometastasis, thus alerting the oncologist of the brain relapse or minimal residual disease?
potential for BCBM development at the time of pBC treat- • How can CTC/ctDNA testing better define inter-
ment or early stages of BCBM. Conversely, in clinical sce- plays between BM-associated dormancy vs. meta-
narios where BCBM is already detected, a reduction of static competence?
Ki67+/uPAR+/ITGβ1+ and/or Notch1+ CTC numbers by • Are CTC/ctDNA profiles promoting BM consistent
longitudinal monitoring can correlate with a successful ther- across all cancer types? Are there differences in
apy (Fig. 16.3). The latter can be significant in relation to patient populations that may be exploited or limit
predicting or preventing further secondary metastasis to viable liquid biopsy tests from being developed?
brain or other organ sites (“metastasis of metastasis”). In
16 CTCs/ctDNA and Brain Metastasis 165

References Evaluating circulating tumor DNA from the cerebrospinal fluid of

patients with melanoma and leptomeningeal disease. J Neuropathol
Exp Neurol. 2018;77:628–35. https://doi.org/10.1093/jnen/nly046.
1. Nevel KS, DiStefano N, Lin X, Skakodub A, Ogilvie SQ, Reiner
16. Zhang L, Ridgway LD, Wetzel MD, Ngo J, Yin W, Kumar D, Good-
AS, Pentsova E, Boire A. A retrospective, quantitative assessment
man JC, Groves MD, Marchetti D. The identification and charac-
of disease burden in patients with leptomeningeal metastases from
terization of breast cancer CTCs competent for brain metastasis.
non-small-cell lung cancer. Neuro-Oncology. 2019;22:675–83.
Sci Transl Med. 2013;5:180ra148. https://doi.org/10.1126/scitrans-
https://doi.org/10.1093/neuonc/noz208.
lmed.3005109.
2. Simonelli M, Dipasquale A, Orzan F, Lorenzi E, Persico P, Navarria
17. Li J, Jiang W, Wei J, Zhang J, Cai L, Luo M, Wang Z, Sun W,
P, Pessina F, Nibali MC, Bello L, Santoro A, et al. Cerebrospinal
Wang S, Wang C, et al. Patient specific circulating tumor DNA
fluid tumor DNA for liquid biopsy in glioma patients’ management:
fingerprints to monitor treatment response across multiple tumors.
close to the clinic? Crit Rev Oncol Hematol. 2020;146:102.879.
J Transl Med. 2020;18:293. https://doi.org/10.1186/s12967-­020-­
https://doi.org/10.1016/j.critrevonc.2020.102879.
02449-­y.
3. Baumgartner JM, Riviere P, Lanman RB, Kelly KJ, Veerapong J,
18. Aguirre-Ghiso JA. Models, mechanisms and clinical evidence for
Lowy AM, Kurzrock R. Prognostic utility of pre- and postopera-
cancer dormancy. Nat Rev Cancer. 2007;7:834–46. https://doi.
tive circulating tumor DNA liquid biopsies in patients with peri-
org/10.1038/nrc2256.
toneal metastases. Ann Surg Oncol. 2020;27:3259–67. https://doi.
19. Alcaide M, Cheung M, Hillman J, Rassekh SR, Deyell RJ, Batist
org/10.1245/s10434-­020-­08331-­x.
G, Karsan A, Wyatt AW, Johnson N, Scott DW, et al. Evaluating
4. Dagogo-Jack I, Carter SL, Brastianos PK. Brain metastasis: clinical
the quantity, quality and size distribution of cell-free DNA by mul-
implications of branched evolution. Trends Cancer. 2016;2:332–7.
tiplex droplet digital PCR. Sci Rep. 2020;10:12,564. https://doi.
https://doi.org/10.1016/j.trecan.2016.06.005.
org/10.1038/s41598-­020-­69432-­x.
5. Klein-Scory S, Wahner I, Maslova M, Al-Sewaidi Y, Pohl M, Mika
20. Vishnoi M, Peddibhotla S, Yin W, Scamardo AT, George GC, Hong
T, Ladigan S, Schroers R, Baraniskin A. Evolution of RAS muta-
DS, Marchetti D. The isolation and characterization of CTC subsets
tional status in liquid biopsies during first-line chemotherapy for
related to breast cancer dormancy. Sci Rep. 2015;5:17,533. https://
metastatic colorectal cancer. Front Oncol. 2020;10:1115.
doi.org/10.1038/srep17533.
6. Ribas A, Hamid O, Daud A, Hodi FS, Wolchok JD, Kefford R,
21. Lee S-Y, Chae D-K, Lee S-H, Lim Y, An J, Chae CH, Kim BC,
Joshua AM, Patnaik A, Hwu W-J, Weber JS, et al. Association of
Bhak J, Bolser D, Cho D-H. Efficient mutation screening for cervi-
pembrolizumab with tumor response and survival among patients
cal cancers from circulating tumor DNA in blood. BMC Cancer.
with advanced melanoma. JAMA. 2016;315:1600–9. https://doi.
2020;20:694. https://doi.org/10.1186/s12885-­020-­07161-­0.
org/10.1001/jama.2016.4059.
22. Boral D, Vishnoi M, Liu HN, Yin W, Sprouse ML, Scamardo A,
7. Vad-Nielsen J, Meldgaard P, Sorensen BS, Nielsen AL. Cell-free
Hong DS, Tan TZ, Thiery JP, Chang JC, et al. Molecular character-
chromatin immunoprecipitation (cfChIP) from blood plasma can
ization of breast cancer CTCs associated with brain metastasis. Nat
determine gene-expression in tumors from non-small-cell lung
Commun. 2017;8:196. https://doi.org/10.1038/s41467-­017-­00196-­1.
cancer patients. Lung Cancer. 2020;147:244–51. https://doi.
23. Zhao Y, He JY, Cui JZ, Meng Z-Q, Zou YL, Guo XS, Chen X,
org/10.1016/j.lungcan.2020.07.023.
Wang X, Yan L-T, Han WX, et al. Detection of genes mutations in
8. Pantel K, Alix-Panabières C. Liquid biopsy and minimal residual
cerebrospinal fluid circulating tumor DNA from neoplastic men-
disease—latest advances and implications for cure. Nat Rev Clin
ingitis patients using next generation sequencing. BMC Cancer.
Oncol. 2019;16:409–24. https://doi.org/10.1038/s41571-­019-­
2020;20:690. https://doi.org/10.1186/s12885-­020-­07172-­x.
0187-­3.
24. Sørlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Johnsen H, Has-
9. Alix-Panabieres C, Pantel K. Clinical applications of circulating
tie T, Eisen MB, van de Rijn M, Jeffrey SS, et al. Gene expression
tumor cells and circulating tumor DNA as liquid biopsy. Cancer
patterns of breast carcinomas distinguish tumor subclasses with
Discov. 2016;6:479–91. https://doi.org/10.1158/2159-­8290.CD-­
clinical implications. Proc Natl Acad Sci. 2001;98:10,869. https://
15-­1483.
doi.org/10.1073/pnas.191367098.
10. Li Y-S, Zheng M-M, Jiang B-Y, Tu H-Y, Yang J-J, Zhang X-C, Wu
25. Tang K, Gardner S, Snuderl M. The role of liquid biopsies in pedi-
Y-L. Association of cerebrospinal fluid tumor DNA genotyping
atric brain tumors. J Neuropathol Exp Neurol. 2020;79:934–40.
with survival among patients with lung adenocarcinoma and central
https://doi.org/10.1093/jnen/nlaa068.
nervous system metastases. JAMA Netw Open. 2020;3:–e209077.
26. Duan H, Hu J-L, Chen Z-H, Li J-H, He Z-Q, Wang Z-N, Zhang
https://doi.org/10.1001/jamanetworkopen.2020.9077.
G-H, Guo X-Y, Liang L, Mou Y-G. Assessment of circulating
11. Heitzer E, Haque IS, Roberts CES, Speicher MR. Current and
tumor DNA in cerebrospinal fluid by whole exome sequencing to
future perspectives of liquid biopsies in genomics-driven oncology.
detect genomic alterations of glioblastoma. Chin Med J. 2020;133
Nat Rev Genet. 2019;20:71–88. https://doi.org/10.1038/s41576-­
27. Steinert G, Schölch S, Niemietz T, Iwata N, García SA, Behrens
018-­0071-­5.
B, Voigt A, Kloor M, Benner A, Bork U, et al. Immune escape and
12. Diaz LA Jr, Bardelli A. Liquid biopsies: genotyping circulat-
survival mechanisms in circulating tumor cells of colorectal can-
ing tumor DNA. J Clin Oncol. 2014;32:579–86. https://doi.
cer. Cancer Res. 2014;74:1694. https://doi.org/10.1158/0008-­5472.
org/10.1200/JCO.2012.45.2011.
CAN-­13-­1885.
13. Mego M, De Giorgi U, Dawood S, Wang X, Valero V, Andreopou-
28. Sartori D, Bari M, Pappagallo GL, Rosetti F, Olsen S, Vinante
lou E, Handy B, Ueno NT, Reuben JM, Cristofanilli M. Charac-
O. Brain metastases in breast cancer: different survival by bio-
terization of metastatic breast cancer patients with nondetectable
logical subtype and Ki67 expression. J Clin Oncol. 2009;27:1069.
circulating tumor cells. Int J Cancer. 2011;129:417–23. https://doi.
https://doi.org/10.1200/jco.2009.27.15_suppl.1069.
org/10.1002/ijc.25690.
29. Pauken CM, Kenney SR, Brayer KJ, Guo Y, Brown-Glaberman
14. Wehrle J, Philipp U, Jolic M, Follo M, Hussung S, Waldeck S, Deu-
UA, Marchetti D. Heterogeneity of circulating tumor cells neoplas-
ter M, Rassner M, Braune J, Rawluk J, et al. Personalized treatment
tic subpopulations outlined by single-cell transcriptomics. Cancers.
selection and disease monitoring using circulating tumor DNA
2021;4885:1–21. https://doi.org/10.3390/cancers13194885.
profiling in real-world cancer patient management. Diagnostics
30. Alimirzaie S, Bagherzadeh M, Akbari MR. Liquid biopsy in breast
(Basel). 2020;10 https://doi.org/10.3390/diagnostics10080550.
cancer: a comprehensive review. Clin Genet. 2019;95:643–60.
15. Ballester LY, Glitza Oliva IC, Douse DY, Chen MM, Lan C,
https://doi.org/10.1111/cge.13514.
Haydu LE, Huse JT, Roy-Chowdhuri S, Luthra R, Wistuba II, et al.
166 S. R. Kenney and D. Marchetti

31. De Mattos-Arruda L, Mayor R, Ng CKY, Weigelt B, Martinez-­ 35. Garcia-Murillas I, Chopra N, Comino-Méndez I, Beaney M, Tovey
Ricarte F, Torrejon D, Oliveira M, Arias A, Raventos C, Tang J, H, Cutts RJ, Swift C, Kriplani D, Afentakis M, Hrebien S, et al.
et al. Cerebrospinal fluid-derived circulating tumour DNA better Assessment of molecular relapse detection in early-stage breast
represents the genomic alterations of brain tumours than plasma. cancer. JAMA Oncol. 2019;5:1473–8. https://doi.org/10.1001/jam-
Nat Commun. 2015;6:8839. https://doi.org/10.1038/ncomms9839. aoncol.2019.1838.
32. Hyun MH, Lee ES, Eo JS, Kim S, Kang EJ, Sung JS, Choi YJ, 36. Nassiri F, Chakravarthy A, Feng S, Shen SY, Nejad R, Zuccato JA,
Park KH, Shin SW, Lee SY, et al. Clinical implications of circulat- Voisin MR, Patil V, Horbinski C, Aldape K, et al. Detection and
ing cell-free DNA quantification and metabolic tumor burden in discrimination of intracranial tumors using plasma cell-free DNA
advanced non-small cell lung cancer. Lung Cancer. 2019;134:158– methylomes. Nat Med. 2020;26:1044–7. https://doi.org/10.1038/
66. https://doi.org/10.1016/j.lungcan.2019.06.014. s41591-­020-­0932-­2.
33. Lin C-C, Huang W-L, Wei F, Su W-C, Wong DT. Emerging plat- 37. Sprouse LM, Welte T, Boral D, Liu NH, Yin W, Vishnoi M, Gos-
forms using liquid biopsy to detect EGFR mutations in lung cancer. wami-Sewell D, Li L, Pei G, Jia P, et al. PMN-MDSCs enhance
Expert Rev Mol Diagn. 2015;15:1427–40. https://doi.org/10.1586/ CTC metastatic properties through reciprocal interactions via
14737159.2015.1094379. ROS/notch/nodal signaling. Int J Mol Sci. 2019;20 https://doi.
34. Bettegowda C, Sausen M, Leary RJ, Kinde I, Wang Y, Agrawal N, org/10.3390/ijms20081916.
Bartlett BR, Wang H, Luber B, Alani RM, et al. Detection of cir-
culating tumor DNA in early- and late-stage human malignancies.
Sci Transl Med. 2014;6:224ra224. https://doi.org/10.1126/scitrans-
lmed.3007094.
 Epigenetics and Liquid Biopsy
in Oncology: Role in Metastasis 17
and Clinical Utility

Aitor Rodriguez-Casanova, Aida Bao-Caamano,

Nicolás Costa-Fraga, Laura Muinelo-Romay,
and Angel Diaz-Lagares

Abstract
Learning Objectives
The deregulation of epigenetic mechanisms in tumor cells • Describe the concept of epigenetics, the different
is implicated in multiple steps of cancer development and epigenetic mechanisms, the cross talk between
progression. These epigenetic alterations are tightly asso- them, and the interest of epigenetic therapy.
ciated to the metastatic process, which is one of the hall- • Explain how epigenetic mechanisms are deregu-
marks of cancer and the main cause of cancer-related lated in cancer and metastasis.
deaths. By means of lymphatic system or blood vessels, • Define the concept of liquid biopsy and describe the
tumors can release different types of tumor material to cir- different types of tumor material that can circulate
culation, including circulating tumor cells (CTCs), nucleic in body fluids.
acids (cNAs), and extracellular vesicles (EVs). These cir- • Describe the presence of epigenetic alterations in
culating tumor elements can present tumor-­specific epi- liquid biopsy and their relation to cancer dissemina-
genetic alterations with relevant biological information tion and metastasis.
about tumor evolution and dissemination, and with clinical • Summarize the role of the epigenetic alterations as
utility as tumor biomarkers. In fact, the analysis of these tumor biomarkers and their clinical implications in
epigenetic biomarkers by liquid biopsies has shown prom- cancer.
ising results in different types of tumors for the personal-
ized management of cancer patients. Here, we provide an
overview of the main epigenetic alterations in cancer, and
describe the implications of these epigenetic marks in can- 17.1 Introduction
cer dissemination and metastasis, highlighting their clini-
cal utility as circulating tumor biomarkers. Cancer dissemination and metastasis of tumors remain the
leading cause of cancer-related deaths [1]. The metastatic
Aitor Rodriguez-Casanova, Aida Bao-Caamano, and Nicolás Costa-­ process of primary tumors is based on the acquisition of a
Fraga share first authorship. long series of sequential and interrelated steps, including

A. Rodriguez-Casanova L. Muinelo-Romay
Cancer Epigenomics, Translational Medical Oncology Group Liquid Biopsy Analysis Unit, Translational Medical Oncology
(Oncomet), Health Research Institute of Santiago (IDIS), Group (Oncomet), Health Research Institute of Santiago (IDIS),
University Clinical Hospital of Santiago (CHUS/SERGAS), University Clinical Hospital of Santiago (CHUS/SERGAS),
Santiago de Compostela, Spain Santiago de Compostela, Spain

Roche-Chus Joint Unit, Translational Medical Oncology Centro de Investigación Biomédica en Red Cáncer (CIBERONC),
Group (Oncomet), Health Research Institute of Santiago (IDIS), Madrid, Spain
Santiago de Compostela, Spain A. Diaz-Lagares (*)
A. Bao-Caamano · N. Costa-Fraga Cancer Epigenomics, Translational Medical Oncology Group
Cancer Epigenomics, Translational Medical Oncology (Oncomet), Health Research Institute of Santiago (IDIS),
Group (Oncomet), Health Research Institute of Santiago (IDIS), University Clinical Hospital of Santiago (CHUS/SERGAS),
University Clinical Hospital of Santiago (CHUS/SERGAS), Santiago de Compostela, Spain
Santiago de Compostela, Spain Centro de Investigación Biomédica en Red Cáncer (CIBERONC),
Madrid, Spain
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 167
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_17
168 A. Rodriguez-Casanova et al.

tumor vascularization (angiogenesis), local invasion of the facilitates the migration and invasion of tumor cells, and
host stroma, the entry in the systemic circulation usually by colonization at distant sites. In addition, the alteration of
means of the lymphovascular system, extravasation to the epigenetic mechanisms has a relevant role in cancer dis-
surrounding tissue, and proliferation within the organ paren- semination from the site of primary tumor growth to distant
chyma [2]. Other relevant mechanisms have been identified tissues via the vessels and lymph nodes (LNs) of the lym-
in relation to the development of metastasis. Thus, tumor phatic system. LNs have a fundamental role in metastasis,
cells are able to escape the action of the immune system cre- acting as a “bridgehead” where subpopulations of tumor
ating an immunosuppressive microenvironment that facili- cells are selected according to their properties to colonize
tates tumor growth and dissemination [3]. distant sites [4, 5]. In fact, in most of the tumors, the dis-
All these metastatic steps are tightly regulated by molec- ease progression occurs first through the lymphatic system.
ular changes, such as deregulation of the epigenetic mecha- However, this is not a general rule, since tumor cells can
nisms: DNA methylation, posttranslational modifications enter directly the blood bypassing the LNs. Both dissemi-
of histones, and the expression of noncoding RNAs nation paths can lead to venous circulation, as lymphatics
(ncRNAs). These epigenetic marks are implicated in very drain into blood and lymphatic fluid is filtered by LNs [6].
relevant metastatic features, such as epithelial-mesenchy- Therefore, in blood we can find tumor material representa-
mal transition (EMT), which represents a reversible cellu- tive of both dissemination mechanisms, including circulat-
lar transition process that allows epithelial cells to adopt ing tumor cells (CTCs), circulating tumor DNA (ctDNA),
characteristics of mesenchymal cells and the properties of and other circulating elements, such as tumor-­ derived
tumor-initiating cells (TICs). Of note, this EMT process extracellular vesicles (EVs) [7] (Fig. 17.1).

Fig. 17.1 Epigenetic alterations in liquid biopsy and clinical utility clinical utility as dynamic tumor biomarkers in different clinical con-
in cancer. Primary tumors and metastasis can release different types of texts: early detection, prognosis, monitoring, therapy selection, and
tumor material, including circulating tumor cells (CTCs), nucleic acids evaluation of therapy response. The analysis of these circulating epi-
(cNAs), and extracellular vesicles (EVs), that can reach the circulation genetic biomarkers represents a promising tool for personalized oncol-
via lymphatic system or blood vessels. These circulating tumor ele- ogy. Created with BioRender.com
ments are able to present tumor-specific epigenetic alterations, showing
17 Epigenetics and Liquid Biopsy in Oncology: Role in Metastasis and Clinical Utility 169

In this chapter, we provide an overview of the main epi- DNA methylation, which is the best characterized
genetic alterations that can be found in cancer, describing epigenetic mechanism, consists of the incorporation of
their implications in the dissemination of tumor cells and a methyl group (CH3) to the 5′ carbon of cytosines in
metastasis, and highlighting the clinical utility of these epi- cytosine-­phosphate-­guanine (CpG) dinucleotides to produce
genetic modifications as tumor biomarkers for cancer 5-­methylcytosine (5mC). This methylation process is medi-
patients in the era of personalized medicine. ated by the enzymes DNA methyltransferases (DNMTs) and
usually occurs in CpG islands (CGIs), which are genomic
regions with a high density of CpGs. These CGIs are often
17.2 Epigenetic Mechanisms located in the promoter regions of genes but they can also
be present in intragenic and intergenic regions. Importantly,
The term epigenetics refers to the study of heritable modifi- DNA methylation can be reversed through the use of
cations in the activity and expression of genes that do not epigenetic-­based drugs (epidrugs), such as DNMT inhibitors
involve alterations of the DNA sequence [8]. The epigenetic (DNMTi) [9].
machinery includes several mechanisms (Fig. 17.2), such as Different types of reversible posttranslational modifica-
DNA methylation, histone modifications, and ncRNAs. tions (PTM) have been described in histone tails, mainly
These mechanisms regulate the normal gene expression in acetylation and methylation, which are related to gene acti-
cells and their deregulation has important relevance in cancer vation or silencing [10]. These PTMs play an important role
development and progression [9]. in both normal development and pathogenesis. The combina-

Fig. 17.2 Epigenetic mechanisms. There are several interconnected ulation has shown utility in the clinic as epigenetic biomarkers and
epigenetic layers that normally regulate the gene expression of cells: therapeutic targets in multiple types of tumors. 5mC 5-Methylcytosine,
DNA methylation, posttranslational modifications of histones, and non- C Cytosine, lncRNA Long noncoding RNA, miRNA MicroRNA, cir-
coding RNAs. These epigenetic mechanisms can undergo alterations cRNA Circular RNA. Created with BioRender.com
inducing cancer development, progression, and metastasis. This dereg-
170 A. Rodriguez-Casanova et al.

tion of histone PTMs (“histone code”) is regulated by differ- Importantly, several epigenetic events have been described
ent types of enzymes. In the case of acetylation and in relation to the metastatic process. Thus, promoter hyper-
methylation, the balance of these histone marks depends on methylation of CDH1 is able to silence the expression of this
histone acetyltransferases (HATs), histone deacetylases gene, leading to the activation of the EMT program and the
(HDACs), histone methyltransferases (HMTs), and histone promotion of invasiveness [13]. Another member of the cad-
demethylases (HDMs). In addition, these marks can be herin family, cell-adherence gene Cadherin-11 (CDH11),
reversed by using epidrugs, including HDAC and HMT was found to be unmethylated in primary tumors but hyper-
inhibitors (HDACi and HMTi) [9, 11]. methylated and silenced in metastatic tumor cells from LN,
NcRNAs are also involved in the epigenetic regulation of suggesting that the epigenetic modulation of this gene could
gene expression. These noncoding transcripts are classified disrupt cell-to-cell contacts, allowing tumor cells intravasa-
according to their nucleotide length into: (i) small ncRNAs tion into lymphatic vessels [4]. On the other hand, the meth-
(sncRNAs), with less than 200 nt, which include microRNAs ylation status of some genomic regions, such as long
(miRNA), small interfering RNAs (siRNA), and piwi-­ interspersed nuclear elements (LINE-1), has been also asso-
interacting RNAs (piRNA); and (ii) long ncRNAs (lncRNA), ciated with dissemination of tumor cells. The hypermethyl-
with more than 200 nt, which encompass long intergenic ation of LINE-1 contributes to genomic stability, and its
ncRNA (lincRNA) and long intronic ncRNA (intronic hypomethylation is associated with tumor invasion and the
lncRNA). MiRNAs are the most widely described ncRNAs, presence of positive LN [14]. Importantly, some agents can
and exert their function by binding to mRNA, blocking its induce hypomethylation by reversing the tumorigenic effect
transcription or triggering mRNA degradation. Thus, each of hypermethylation. DNMTis, such as Decitabine and
miRNA can regulate the expression of hundreds of genes, Azacitidine, have been described as epidrugs and approved
modulating a lot of key pathways. On the other hand, it has by U.S. Food and Drug Administration (FDA) [11].
been described that lncRNAs represent the majority of non- Metastasis is also characterized by the disruption of his-
coding transcripts, and they are also involved in the regula- tone modification patterns. The alteration of histone acetyla-
tion of gene expression [12]. tion levels can influence target genes implicated in migration,
Importantly, the different layers of epigenetic modifica- the development of EMT program and metastasis. HDAC
tions form a complex regulatory network, where there is a deregulation plays an essential role in silencing CHD1 dur-
cross talk between the different epigenetic players [9]. ing the metastatic process and in the upregulation of mesen-
chymal genes, such as SNAIL [15]. Similarly, the alteration
of HMTs can produce an aberrant methylation of histones
17.3 Deregulation of Epigenetic leading to the development of metastasis. This is the case of
Mechanisms in Cancer and Metastasis enhancer of zeste homolog 2 (EZH2), an enzyme that cata-
lyze trimethylation of histone H3 at lysine 27 (H3K27me3)
Epigenetic mechanisms are deregulated in multiple types of contributing to transcriptional silencing of target genes, and
tumors and they are involved in all the steps of carcinogene- whose overexpression is linked to lymphovascular dissemi-
sis, participating in the initiation, progression, and dissemi- nation and cancer progression [16]. Due to the interplay
nation of tumor cells. Importantly, the disruption of the between histone enzymes with metastasis, these epigenetic
different layers of the epigenetic machinery has been pro- players represent promising targets for antimetastatic ther-
posed as a hallmark of cancer [11]. apy. Nowadays, some epidrugs that target the action of the
The hypermethylation of CGIs in promoters is usually histone enzyme machinery have been approved by FDA and
linked to silencing of both coding and noncoding tumor sup- are available as standard-of-care treatment for some tumor
pressor genes, while the hypomethylation of CpG-poor types (e.g., HDAC and EZH2 inhibitors), and many others
regions has been associated with proto-oncogene expression, are under development [11].
genomic instability, and malignant transformation of tumors The deregulation of ncRNA expression levels has also
[12]. On the other hand, the histone PTMs, and/or the shown association with invasion and motility of tumor
enzymes that regulate these modifications, are often altered cells, as well as with LN metastasis. The overexpression of
in tumor cells, leading to an imbalance of histone marks [9]. some miRNAs in tumor cells, such as miR-10b, is able to
NcRNAs can also have a tumorigenic effect, either by induce tumor invasion and metastasis in several tumor
repressing or promoting tumor properties. The normal cross types [17]. Importantly, the prometastatic effect of this type
talk among the different epigenetic layers is also disrupted in of ncRNAs can be reversed using targeted therapies, pre-
cancer, where, for example, the alteration of specific meth- venting, for example, the metastatic spread from the pri-
ylation and histone modification patterns can modify the mary tumor to the LNs and arresting the metastatic growth
expression of different types of ncRNAs [9, 12]. in the already formed LN metastasis [18]. In addition to
17 Epigenetics and Liquid Biopsy in Oncology: Role in Metastasis and Clinical Utility 171

miRNAs, the deregulation in the expression levels of other 17.5  pigenetic Alterations in Liquid
E
ncRNAs, such as lncRNAs (e.g., HOTAIR), has shown to Biopsy and Their Association
be implicated in tumor dissemination and LN metastasis of with Metastasis
cancer patients [19].
During the last decade, different studies have demonstrated
the alteration of epigenetic mechanisms in tumor circulating
17.4 Liquid Biopsy elements, mainly in CTCs and ctDNA, linked to the several
steps of tumor generation and dissemination (Fig. 17.1) [24].
All the body fluids that contain tumor material suitable for Pioneering works characterized the methylation status of
molecular characterization are encompassed under the term tumor-associated genes in CTCs from breast cancer with the
liquid biopsy. Although blood is the most used sample, other goal of identifying new mechanisms that favor this tumor
fluids, such as urine, pleural effusion, ascitic fluid, cerebro- spread. From these studies we learned that CST6, BRMS1,
spinal fluid, saliva, and stool are also of great value as liquid and SOX17 promoters were hypermethylated in CTCs, as a
biopsies. This tumor material present in body fluids can be mechanism for promoting the tumor dissemination [25].
released from primary tumors or metastasis and it mainly More recently, a specific methylation program which regu-
consists of circulating tumor cells (CTCs) and circulating lates stemness-associated transcription factors (OCT4,
nucleic acids (cNAs). Both elements constitute a valuable NANOG, SOX2, and SIN3A) has been identified in CTCs-­
source of information about the molecular mechanisms clusters and linked to the metastasis formation in breast can-
underlying tumor evolution and dissemination and also of cer [26]. In addition to the stem properties, EMT is another
noninvasive biomarkers [20]. important mechanism favoring CTCs dissemination. In fact,
The use of cNAs has started to be applied for the therapy in colorectal cancer (CRC) patients, hypermethylation of the
selection. These analyses are mainly centered on circulating SFRP2 promoter and VIM has been described and associated
cell-free DNA (cfDNA), which is released as a consequence with a more mesenchymal phenotype of this cells and the
of cell death and is present highly fragmentated in body flu- promotion of metastasis [27]. The epigenetic regulation of
ids (size, 160–180 bp). This cfDNA generally contains, in a EMT in CTCs from prostate cancer has also been described
range as low as 0.01–1%, circulating tumor DNA (ctDNA) in and associated with the blood-borne dissemination [28].
patients with advanced tumors [20]. This cfDNA can be ana- Besides, in patients with advanced prostate cancer CTCs
lyzed in a noninvasive and comprehensive way to study dif- showed hypermethylation patterns in genes with relevant
ferent genetic alterations such as point mutations, copy roles for angiogenesis and apoptosis [29].
number variations, small indels, and translocations, together Different methylation patterns have also been identified in
with epigenetic modifications [20]. In addition, numerous cfDNA from cancer patients and associated with the dissem-
noncoding RNAs (ncRNAs) can be detected in liquid biopsy, ination process. Thus, hypomethylation of LINE-1 in plasma
and have been described as regulators of specific gene cfDNA of CRC patients was correlated with the disease pro-
expression signatures in tumor cells and stroma, controlling gression, especially, with the presence of LN affectation and
cell-to-cell communication [21]. the generation of distant metastasis [30]. Contrary, the hyper-
On the contrary of cNAs, the study of CTCs has shown methylation of other gene promoters such as RUNX3,
limited translation into the routine clinical practice so far. GATA5, TFPI2, SFRP2, SHOX2, and SEPT9 in cfDNA of
These cells are present in very low numbers in blood and CRC patients has been implicated in the lymphatic invasion
require the implementation of highly sensitive and specific and tumor recurrence [31–34]. Similarly, in plasma from
strategies for their enrichment and detection. Although the breast cancer patients the methylation of SOX17 has been
most common strategy to isolate CTCs was immune enrich- described associated with a more aggressive disease and
ment, this strategy has not been widely adopted as a clinical clearly linked to the presence of lymphovascular dissemina-
tool, mainly due to the lack of clear benefits for treatment tion [35]. High levels of methylation of different genes such
decision-making. Other antigen-independent technologies as PCDH8 were also associated with high preoperative
for the isolation and characterization of CTCs have been prostate-­specific antigen (PSA) levels and LN metastasis in
developed to isolate a more heterogeneous CTC population patients with advanced prostate cancer [36].
[22]. Despite the limitation for their isolation, molecular In addition to DNA methylation, other epigenetic mecha-
characterization of CTCs has provided valuable information nisms, such as histone modifications, have been described in
to understand how tumor cells disseminate and implant at cfDNA and associated with the process of carcinogenesis.
distant locations. Their molecular study, including epigene- For example, it has been described that H3K27me3 altera-
tic mechanisms, is important to unravel the biological aspects tions detected in plasma represent an epigenetic mark in
of cancer and identify new clinical and therapeutic biomark- prostate carcinogenesis, being associated with the presence
ers to manage this disease [23]. of disseminated disease after surgery [37].
172 A. Rodriguez-Casanova et al.

On the other hand, different miRNA signatures have posed as prognosis biomarkers [12]. Besides, the methyla-
shown an impact on cancer progression. The presence of tion of ITIH5 or RASSF1 in CTCs was also related to poor
miRNAs related to migration and invasion have been found survival rates [12]. Hypermethylation of other genes such as
altered in CTCs in breast tumors [38]. Besides, different ESR1 have shown an impact on the response to targeted ther-
miRNAs contained in circulating exosomes of CRC patients, apy in hormone sensitive tumors [46]. In addition, the meth-
such as miR-25-3p, are increased in patients with metastatic ylation of several gene panels, such as being APC, RARβ2,
disease [39]. This miRNA regulates the expression of target and RASSF1A, has demonstrated high accuracy to detect
genes, promoting vascular permeability, angiogenesis, and breast cancer [45]. Besides, the analysis of RASSF1A and
the formation of liver and lung metastasis in preclinical mod- PITX2 methylation in cfDNA have shown value to predict
els [39]. Furthermore, recent results have indicated that the overall survival in advanced breast tumors [45].
decreased levels of serum exosomal miR-638 [40] and miR-­ In bronchial aspirates, pleural effusions, and plasma from
548c-­5p [41] are associated with liver metastasis in CRC patients with lung cancer, SHOX2 methylation was found
patients. Besides, several circulating miRNAs have shown clearly increased and associated with lymphovascular affec-
association with cancer-related thrombosis, which promotes tation, supporting the development of a diagnostic test (Epi
the generation of metastasis [42]. proLung BL Reflex Assay) [44]. Beyond the diagnosis, the
Other circulating ncRNAs have been implicated in the promoter methylation of DCLK1, SHP1P2, or SOX17 has
promotion of cancer spread. For example, the lncRNA been described as a factor associated with lower survival
MALAT1 is well known as an enhancer of the expansion and rates in lung tumors [45]. Besides, in prostate cancer the
migration of tumor cells. This lncRNA has been found methylation of GSTP1 in cfDNA was described as a specific
upregulated in serum and plasma of epithelial ovarian can- mark and showed clinical value to predict PSA progression
cer, clearly associated with the presence of lymphovascular and overall survival in patients with castration resistant
affectation [43]. tumors [45].
Notably, recently genome-wide methylation approaches
on cfDNA have shown great value to detect the presence of
17.6  linical Implications of Circulating
C different tumor types, alone or in combination with the
Epigenetic Biomarkers in Liquid detection of genomic alterations and specific epigenetic pat-
Biopsy terns. Moreover, the combination of NGS with machine
learning allowed the methylation analysis of cfDNA and the
Alterations of epigenetic patterns in liquid biopsies have development of a test capable of detecting more than 50
shown great utility as tumor biomarkers for early detection, tumor types [47, 48].
prognosis, monitoring, and evaluation of therapeutic Besides, histone modifications have also been proposed
response at different stages of cancer disease (Fig. 17.1). In as circulating epigenetic biomarkers. Thus, for example, low
the context of the early detection and screening of CRC, levels of the circulating histone marks H4K20me3, and
several tests focused on the analysis of methylation signa- H3K27me3, have shown value for CRC detection [49].
tures have been developed. Epi proColon, based on the NcRNAs are also of great interest as cancer biomarkers.
detection of methylation in the promoter region of the Several circulating miRNA signatures have been proposed as
SEPT9 gene in cfDNA, was the first test commercialized. biomarkers for early detection of colon, breast, and lung can-
The combined analysis of the methylation status of SEPT9 cer, among others [21]. Thus, circulating miR-21, miR-155,
and SDC2 led to the development of a new test miR-125b, or miR-19a had shown to be promising diagnos-
(ColoDefense), which improved the detection of CRC and tic biomarkers for breast cancer detection. In lung cancer, the
adenomas, and represent a valuable tool for CRC screening most consistent miRNAs biomarkers for the disease diagno-
and early detection [44]. sis and prognosis are miR-125b and miR-21. In CRC, the
The analysis of epigenetic signatures in cfDNA and CTCs analysis of miR-21 levels has been also proposed for early
from breast cancer patients has also provided potential clini- detection and as a prognostic factor. In addition, increased
cal biomarkers. Thus, specific methylation profiles of cfDNA levels of miR-375 and miR-141 in plasma have been corre-
have been described as interesting tools for the early diagno- lated with the presence of metastatic disease in prostate can-
sis and the prediction of the therapy response [12, 45]. The cer. Finally, several circulating lncRNAs such as HOTAIR,
hypermethylation of BRMS1, SOX17, and CST6 in CTCs has MALAT1, or PCAT18 have an increasing interest as diag-
been associated with advanced breast cancer stages and pro- nostic and prognostic biomarkers of cancer [21].
17 Epigenetics and Liquid Biopsy in Oncology: Role in Metastasis and Clinical Utility 173

Acknowledgments This work was co-funded by the ISCIII

Conclusions and Future Perspectives (PI18/00307) and the European Regional Development Fund (FEDER).
AD-L is funded by a contract “Juan Rodés” (JR17/00016) and AB-C by
The epigenetic characterization of circulating tumor a predoctoral contract PFIS (FI19/00240) co-funded by “Fondo Social
material has generated a lot of interest in recent years Europeo” (FSE) from ISCIII. AR-C is supported by Roche-Chus Joint
due to its relevant implication in cancer development Unit (IN853B 2018/03) and GAIN. NC-F is funded by a predoctoral
contract from “Xunta de Galicia” (IN606A-2020/004). LM-R is sup-
and dissemination. The study of epigenetics in liquid ported by AECC.
biopsy represents an emerging field with great poten-
tial to elucidate the main biological and clinical aspects
associated with cancer metastasis. Thus, the alteration
of epigenetic mechanisms has shown great relevance References
in the main steps of the metastatic process, such as
1. Gupta GP, Massague J. Cancer metastasis: building a framework.
migration through the EMT process and the Cell. 2006;127(4):679–95.
lymphovascular-­ blood dissemination, among others. 2. Fidler IJ. The pathogenesis of cancer metastasis: the ‘seed and soil’
Epigenetic marks, including DNA methylation, his- hypothesis revisited. Nat Rev Cancer. 2003;3(6):453–8.
3. Soundararajan R, Fradette JJ, Konen JM, et al. Targeting the inter-
tone modifications, and ncRNAs, can be detected in
play between epithelial-to-mesenchymal-transition and the immune
liquid biopsy and have shown very promising results system for effective immunotherapy. Cancers. 2019;11(5)
as tumor biomarkers for early detection, prognosis, 4. Carmona FJ, Villanueva A, Vidal A, et al. Epigenetic disrup-
monitoring, and evaluation of therapeutic response at tion of cadherin-11 in human cancer metastasis. J Pathol.
2012;228(2):230–40.
different time points of cancer disease. Importantly,
5. Dongre A, Weinberg RA. New insights into the mechanisms of
epigenetic alterations can be reversed by using epi- epithelial-mesenchymal transition and implications for cancer. Nat
drugs, some of them approved by FDA, and many oth- Rev Mol Cell Biol. 2019;20(2):69–84.
ers under development. 6. Wong SY, Hynes RO. Lymphatic or hematogenous dissemi-
nation: how does a metastatic tumor cell decide? Cell Cycle.
Overall, the study of epigenetics in liquid biopsy
2006;5(8):812–7.
represents an emerging field with enormous perspec- 7. Lianidou E, Pantel K. Liquid biopsies. Genes Chromosomes
tives for the future. The characterization of epigenetic Cancer. 2019;58(4):219–32.
marks in circulating tumor material with “omic” and 8. Berger SL, Kouzarides T, Shiekhattar R, et al. An operational defi-
nition of epigenetics. Genes Dev. 2009;23(7):781–3.
very sensitive technologies will provide a great view of
9. Mari-Alexandre J, Diaz-Lagares A, Villalba M, et al. Translating
the epigenetic landscape to discover new biomarkers cancer epigenomics into the clinic: focus on lung cancer. Transl
and therapeutic targets, in a noninvasive and dynamic Res. 2017;189:76–92.
way. Besides, exploring new epigenetic modifications 10. Kouzarides T. Chromatin modifications and their function. Cell.
2007;128(4):693–705.
in liquid biopsy, such as DNA hydroxymethylation and
11. Bates SE. Epigenetic therapies for cancer. N Engl J Med.
different types of ncRNAs, and the analysis of RNA 2020;383(7):650–63.
modifications (epitranscriptomic marks), will offer 12. Bao-Caamano A, Rodriguez-Casanova A, Diaz-Lagares
valuable information about cancer biology. In a clini- A. Epigenetics of circulating tumor cells in breast cancer. Adv Exp
Med Biol. 2020;1220:117–34.
cal context, the use of epigenetic modifications as non-
13. Lombaerts M, van Wezel T, Philippo K, et al. E-cadherin transcrip-
invasive tumor biomarkers and therapeutic targets will tional downregulation by promoter methylation but not mutation is
provide the opportunity to personalize the manage- related to epithelial-to-mesenchymal transition in breast cancer cell
ment of cancer patients. However, some relevant ques- lines. Br J Cancer. 2006;94(5):661–71.
14. Sunami E, de Maat M, Vu A, et al. LINE-1 hypomethyl-
tions must be addressed in the close future to improve
ation during primary colon cancer progression. PLoS One.
the clinical use of circulating epigenetic biomarkers: 2011;6(4):e18884.
15. von Burstin J, Eser S, Paul MC, et al. E-cadherin regulates
• How specific are the epigenetic marks identified in metastasis of pancreatic cancer in vivo and is suppressed by a
SNAIL/HDAC1/HDAC2 repressor complex. Gastroenterology.
liquid biopsies for being used as diagnostic tools?
2009;137(1):361–71, 371 e361-365.
• Are all the circulating tumor elements equally use- 16. Nienstedt JC, Schroeder C, Clauditz T, et al. EZH2 overexpression
ful to characterize the epigenetic marks? in head and neck cancer is related to lymph node metastasis. J Oral
• Can we use the analysis of epigenetic biomarkers to Pathol Med. 2018;47(3):240–5.
17. Ma L, Teruya-Feldstein J, Weinberg RA. Tumour invasion and
guide the potential treatment of tumors with
metastasis initiated by microRNA-10b in breast cancer. Nature.
epidrugs? 2007;449(7163):682–8.
• Will we be able to translate the growing knowledge 18. Yigit MV, Ghosh SK, Kumar M, et al. Context-dependent dif-
about the epigenetic mechanisms that promote the ferences in miR-10b breast oncogenesis can be targeted for the
prevention and arrest of lymph node metastasis. Oncogene.
tumor dissemination into therapeutic targets?
2013;32(12):1530–8.
• How are the different epigenetic mechanisms inter- 19. Cai B, Wu Z, Liao K, et al. Long noncoding RNA HOTAIR can
connected to promote tumor dissemination? serve as a common molecular marker for lymph node metastasis: a
meta-analysis. Tumour Biol. 2014;35(9):8445–50.
174 A. Rodriguez-Casanova et al.

20. Siravegna G, Marsoni S, Siena S, et al. Integrating liquid biop- tion patterns lead to prognostic and predictive classifiers. Oncogene.
sies into the management of cancer. Nat Rev Clin Oncol. 2019;38(18):3387–401.
2017;14(9):531–48. 36. Lin YL, Li YL, Ma JG. Aberrant promoter methylation of
21. Anfossi S, Babayan A, Pantel K, et al. Clinical utility of cir- Protocadherin8 (PCDH8) in serum is a potential prognostic
culating non-coding RNAs—an update. Nat Rev Clin Oncol. marker for low Gleason score prostate cancer. Med Sci Monitor.
2018;15(9):541–63. 2017;23:4895–900.
22. Barbazan J, Alonso-Alconada L, Elkhatib N, et al. Liver metastasis 37. Deligezer U, Yaman F, Darendeliler E, et al. Post-treatment circu-
is facilitated by the adherence of circulating tumor cells to vascular lating plasma BMP6 mRNA and H3K27 methylation levels dis-
fibronectin deposits. Cancer Res. 2017;77(13):3431–41. criminate metastatic prostate cancer from localized disease. Clin
23. Cortes-Hernandez LE, Eslami SZ, Pantel K, et al. Molecular and Chim Acta. 2010;411(19–20):1452–6.
functional characterization of circulating tumor cells: from discov- 38. Markou A, Zavridou M, Sourvinou I, et al. Direct comparison of
ery to clinical application. Clin Chem. 2020;66(1):97–104. metastasis-related miRNAs expression levels in circulating tumor
24. Thomas ML, Marcato P. Epigenetic modifications as biomarkers cells, corresponding plasma, and primary tumors of breast cancer
of tumor development, therapy response, and recurrence across the patients. Clin Chem. 2016;62(7):1002–11.
cancer care continuum. Cancers. 2018;10(4) 39. Zeng Z, Li Y, Pan Y, et al. Cancer-derived exosomal miR-25-3p pro-
25. Chimonidou M, Strati A, Tzitzira A, et al. DNA methylation of motes pre-metastatic niche formation by inducing vascular perme-
tumor suppressor and metastasis suppressor genes in circulating ability and angiogenesis. Nat Commun. 2018;9(1):5395.
tumor cells. Clin Chem. 2011;57(8):1169–77. 40. Yan S, Han B, Gao S, et al. Exosome-encapsulated microR-
26. Gkountela S, Castro-Giner F, Szczerba BM, et al. Circulating tumor NAs as circulating biomarkers for colorectal cancer. Oncotarget.
cell clustering shapes DNA methylation to enable metastasis seed- 2017;8(36):60,149–58.
ing. Cell. 2019;176(1–2):98–112 e114. 41. Peng ZY, Gu RH, Yan B. Downregulation of exosome-encapsulated
27. Lyberopoulou A, Galanopoulos M, Aravantinos G, et al. Identification miR-548c-5p is associated with poor prognosis in colorectal cancer.
of methylation profiles of cancer-related genes in circulating tumor J Cell Biochem. 2018; https://doi.org/10.1002/jcb.27291.
cells population. Anticancer Res. 2017;37(3):1105–12. 42. Xue J, Xie VK, Wang P, et al. Interrelationships of circulating
28. Pixberg CF, Raba K, Muller F, et al. Analysis of DNA methylation tumor cells with metastasis and thrombosis: role of microRNAs.
in single circulating tumor cells. Oncogene. 2017;36(23):3223–31. Curr Pharm Des. 2014;20(33):5298–308.
29. Friedlander TW, Ngo VT, Dong H, et al. Detection and charac- 43. Qiu JJ, Lin XJ, Tang XY, et al. Exosomal metastasis associated
terization of invasive circulating tumor cells derived from men lung adenocarcinoma transcript 1 promotes angiogenesis and pre-
with metastatic castration-resistant prostate cancer. Int J Cancer. dicts poor prognosis in epithelial ovarian cancer. Int J Biol Sci.
2014;134(10):2284–93. 2018;14(14):1960–73.
30. Jung G, Hernandez-Illan E, Moreira L, et al. Epigenetics of 44. Beltran-Garcia J, Osca-Verdegal R, Mena-Molla S, et al. Epigenetic
colorectal cancer: biomarker and therapeutic potential. Nat Rev IVD tests for personalized precision medicine in cancer. Front
Gastroenterol Hepatol. 2020;17(2):111–30. Genet. 2019;10:621.
31. Nishio M, Sakakura C, Nagata T, et al. RUNX3 promoter methyla- 45. Constancio V, Nunes SP, Henrique R, et al. DNA methylation-based
tion in colorectal cancer: its relationship with microsatellite insta- testing in liquid biopsies as detection and prognostic biomarkers for
bility and its suitability as a novel serum tumor marker. Anticancer the four major cancer types. Cell. 2020;9(3)
Res. 2010;30(7):2673–82. 46. Mastoraki S, Strati A, Tzanikou E, et al. ESR1 methylation: a liquid
32. Tang D, Liu J, Wang DR, et al. Diagnostic and prognostic value biopsy-based epigenetic assay for the follow-up of patients with
of the methylation status of secreted frizzled-related protein 2 in metastatic breast cancer receiving endocrine treatment. Clin Cancer
colorectal cancer. Clin Invest Med. 2011;34(2):E88–95. Res. 2018;24(6):1500–10.
33. Zhang X, Song YF, Lu HN, et al. Combined detection of plasma 47. Liu L, Toung JM, Jassowicz AF, et al. Targeted methylation
GATA5 and SFRP2 methylation is a valid noninvasive biomarker sequencing of plasma cell-free DNA for cancer detection and clas-
for colorectal cancer and adenomas. World J Gastroenterol. sification. Ann Oncol. 2018;29(6):1445–53.
2015;21(9):2629–37. 48. Liu MC, Oxnard GR, Klein EA, et al. Sensitive and specific multi-­
34. Bergheim J, Semaan A, Gevensleben H, et al. Potential of quan- cancer detection and localization using methylation signatures in
titative SEPT9 and SHOX2 methylation in plasmatic circulat- cell-free DNA. Ann Oncol. 2020;31(6):745–59.
ing cell-free DNA as auxiliary staging parameter in colorectal 49. Gezer U, Yoruker EE, Keskin M, et al. Histone methylation marks
cancer: a prospective observational cohort study. Br J Cancer. on circulating nucleosomes as novel blood-based biomarker in
2018;118(9):1217–28. colorectal cancer. Int J Mol Sci. 2015;16(12):29,654–62.
35. Panagopoulou M, Karaglani M, Balgkouranidou I, et al. Circulating
cell-free DNA in breast cancer: size profiling, levels, and methyla-
 Metabolic Reprogramming
of Circulating Tumor Cells for Metastasis 18
Ziyuan Zhang and Qihui Shi

Abstract
Learning Objectives
Despite the leading cause of cancer-associated deaths, • Understand metabolic heterogeneity in primary
metastasis is a complex, multi-step, and inefficient bio- tumors and how such heterogeneity promotes can-
logical process for cancer cells, including detachment cer metastasis.
from the primary tumor site, survival in circulation, and • Understand metabolic plasticity of CTCs during
colonization at the distant sites. Metastatic cancer cells metastatic spread and how they survive in system-
thus face a daunting series of challenges, such as anoikis, atic circulation and successfully seed distant metas-
nutrient deprivation, shear stress in circulation, and meta- tases by metabolic adaption to the hostile
bolic stress at distant organs. Cancer cells rewire meta- environment.
bolic pathways to rapidly generate energy, cellular • Understand metabolic-specific inhibitors targeting
building blocks and antioxidants for supporting tumor therapeutic vulnerability from altered metabolic
growth and fueling metastasis. Inherent metabolic hetero- circuits in cancer cells and the limitations.
geneity and additional acquisition of metabolic features
during tumor spread promote circulating tumor cell
(CTC) survival and seeding metastases by adapting to
hostile environments during metastatic cascades. Unlike 18.1 Introduction
tumor tissue samples, metabolic assessment of CTCs in
blood and other types of body fluids is technologically Cancer metastasis accounts for the major cause (~90%) of
challenging because these liquid biopsies contain many cancer-associated deaths worldwide [1, 2]. Metastasis is a
different types of normal cells, requiring development of complex multi-step biological process, including detach-
efficient and specific markers for accurate CTC identifica- ment from the primary tumor site, survival in circulation, and
tion. In addition to clinical tumor imaging (e.g., positron successful colonization at the distant sites [3]. Each step
emission tomography), metabolic reprogramming charac- poses a significant barrier to the metastatic cascade. Anoikis,
teristic of cancer cells represents attractive therapeutic shear stress in circulation, and metabolic stress in coloniza-
targets for anticancer treatment. Although many tion at the distant organ are extremely challenging for cancer
metabolism-­specific inhibitors are under developing or cells to seed distant metastases [4, 5]. Thus, metastasis is an
tested in clinical trials, none of them were approved to inefficient process for cancer cells [3]. It has been estimated
enter routine clinical practice due to metabolic plasticity that only less than 0.02% metastatic cancer cells successfully
of cancer cells and narrow therapeutic window between seed metastases from their primary site [3, 4, 6]. To date,
normal proliferating cells and cancer cells. how cancer cells detach from primary lesion, how they sur-
vive in circulation, and how they select distant sites and cope
with hostile microenvironments for full-blown metastasis are
largely unknown [1, 7]. In particular, accumulating evidence
reveals that cancer cells are able to disseminate to distant
sites early during primary tumor growth [8], indicating
Z. Zhang · Q. Shi (*) intrinsic metastatic potential of cancer cells.
Institutes of Biomedical Sciences and Minhang Hospital, Fudan For mechanistic understanding of cancer metastasis,
University, Shanghai, China many efforts have been devoted to identify metastasis-­
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 175
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_18
176 Z. Zhang and Q. Shi

specific mutations but were unsuccessful [9]. Metastatic can- progression and metastasis [22–25]. Compared with lung
cer cells are indistinguishable from their non-metastatic and colorectal cancer exhibiting elevated glycolytic states,
counterparts in primary tumors or in circulation by genetic breast cancer and glioblastoma are largely viewed as oxida-
signatures. Due to the plasticity of cancer cells, reversible tive tumors [26].LeBieu and coworkers demonstrated inva-
mechanisms (e.g., metabolic, epigenetic) rather than irre- sive breast cancer cells specifically favored mitochondrial
versible mechanisms (e.g., mutation) are likely used by can- respiration for increased ATP production by using the tran-
cer cells to adapt to hostile environments during cancer scription coactivator PGC-1α [27].Cai and coworkers
metastasis. Metabolic reprogramming is known to be one of showed crucial role of pyruvate dehydrogenase complex
the fundamental hallmarks of cancer cells and plays a vital (PDHc) and TCA cycle maintained by AMPK (AMP-­
role in the cancer progression and metastasis [10]. Unlike activated protein kinase) for promoting cancer metastasis by
normal cells, cancer cells often utilize aerobic glycolysis, a adapting breast cancer cells to metabolic and oxidative
phenomenon termed Warburg effect, to generate energy, cel- stresses [28].This AMPK-PDHc axis, activated in advanced
lular building blocks, and antioxidants to support cell trans- breast cancer, was found essential for developing full-blown
formation, to fight oxidative stress, and to fuel the metastatic metastatic cancer and associated with poor metastasis-free
cascade [11]. Complex interplay between the tumor muta- survival [28]. In addition, Porportato and coworkers reported
tional landscape, epigenetic modifications, and microenvi- that both mitochondrial overload and mild respiration dys-
ronmental influences contributes to rewiring of metabolic function promoted metastasis by enhanced mitochondrial
pathways within cancer cells [7]. Understanding the tumor superoxide production [29].
metabolic features, especially the acquired metabolic fea- Interestingly, Tasdogan and coworkers showed that meta-
tures during metastasis, and metabolic vulnerabilities might bolic differences among melanoma cells conferred different
open novel diagnostic and therapeutic approaches to improve metastatic potential as a result of differences in the function of
the efficacy of current anticancer treatments. the MCT1 transporter [30]. In patient-derived xenografts, effi-
In this chapter, we will discuss recent work in understand- ciently metastasizing melanomas were found to have higher
ing the metabolic reprogramming of cancer cells during can- levels of MCT1, whereas inefficiently metastasizing melano-
cer metastasis. We first discuss the metabolic heterogeneity in mas exhibited low or negative MCT1 levels. Despite similar
primary tumors and how such heterogeneity promotes metas- tumor-forming capacities between MCT1high and MCT1−/low
tasis. Secondly, we discuss the metabolic plasticity of dis- cells, MCT1high melanoma cells formed more metastases after
seminating tumor cells during metastatic spread and how they intravenous injection. Elevated expression of MCT1 enhanced
survive in systematic circulation and successfully colonizing uptake of circulating lactate by cancer cells and this mechanism
distant metastases by metabolic adaption. Last, we review was used by metastasizing cells to manage oxidative stress and
potential drugs targeting metabolism-related ­therapeutic vul- promote metastasis. Thus, metabolic differences among cancer
nerability of cancer cells, especially those in metastasis. cells conferred differences in metastatic potential.
The diverse metabolic dependencies have been also
observed between the primary and metastatic lesions of the
18.2  etabolic Heterogeneity in Primary
M same patient, as well as within distinct regions of the same
Tumors Relevant to Metastasis tumor [31, 32].Metabolic heterogeneity arises from genetic
and epigenetic diversity, redundant metabolic pathways, and
In the 1920s, Otto Warburg introduced the concept of meta- altered microenvironmental conditions [33], and thereby
bolic reprogramming that cancer cells depend on aerobic endows cancer cells with enhanced flexibility for rapidly
glycolysis for energy production due to the damaged mito- adapting changing environmental conditions during tumor
chondrial respiration [12–14]. Given that mitochondria are a progression and metastasis. Dupuy and coworkers showed
main source of reactive oxygen species (ROS) generation, extensive metabolic heterogeneity of primary breast cancer
mitochondrial dysfunction may also reduce oxidative stress cells and that distinct metabolic programs were associated
in cancer cells. The phenomenon that cancer cells are with site-specific metastases [34]. Surprisingly, breast cancer
addicted to aerobic glycilysis, termed as Warburg effect, pro- cells with similar genetic backgrounds but distinct metabolic
vides a new approach for detecting cancer cells and paves the characteristics exhibited preferential metastatic tropism to
way for the development of clinical tumor imaging: fluoro- different sites. Bone- or lung-metastatic breast cancer cells
deoxyglucose (18F)-positron emission tomography (18F-FDG were found engaging in OXPHOS metabolic profile whereas
PET) [15, 16]. However, such concept was debated because liver-metastatic cells exhibited a glycolytic-dependent meta-
recent studies demonstrated fully functional mitochondria bolic program [34]. Unlike cancer cells, proliferating cells in
activity in cancers [17–21], and equally important roles for tumor microenvironment have shown a uniform metabolic
mitochondrial metabolism—the tricarboxylic acid (TCA) signature as revealed by a multi-isotope imaging mass spec-
cycle and oxidative phosphorylation (OXPHOS) in tumor trometry with glucose and glutamine tracers [35].
18 Metabolic Reprogramming of Circulating Tumor Cells for Metastasis 177

18.3 Metabolic Heterogeneity such as EpCAM and keratins (CKs) for patients with epithe-
of Disseminating Tumor Cells During lial tumors [40]. Yang and coworkers developed a novel meta-
Metastatic Spread bolic function-associated marker, hexokinase 2 (HK2), for
CTC detection and phenotyping in peripheral blood of
Disseminating tumor cells present in the lymph nodes, blood, NSCLC patients [41]. As the first enzymatic step of glycoly-
and other body fluids offer a direct approach for evaluating sis, hexokinase phosphorylates glucose and traps it in the cell
metabolic states and heterogeneity of cancer cells during [42]. Among hexokinase isoforms, only HK2 is found highly
metastatic spread, especially the acquisition of specific meta- expressed in a wide range of cancers, including epithelial and
bolic features during the different metastatic cascade. mesenchymal origins, and exhibits restricted distribution of
However, these samples (lymph nodes or body fluids) also expression in normal adult tissues [42–45]. As a metabolic
contain a large number and many different types of normal function-associated marker, HK2 overcomes the limitation of
cells. Thus, disseminating tumor cells should be distinguish- traditional epithelial CTC markers that is inherently biased
able from these normal cells before metabolic evaluation, toward identifying CTCs with epithelial traits, leading to the
requiring discovery of markers of disseminating tumor cells capability of detecting CTCs experiencing epithelial-to-­
in lymph nodes and liquid biopsies. mesenchymal transition (EMT) with downregulation of epi-
Pleural effusion is a body fluid that associates with many thelial markers or lineage plasticity [46–49]. For this reason,
types of malignant diseases. Pleural effusion containing can- HK2 enables resolving an underrepresented HK2high/CK−
cer cells shed from primary tumor is termed malignant pleu- CTC population in peripheral blood of NSCLC patients [41].
ral effusion (MPE) that denotes an advanced stage of diseased Although CK− CTCs were reported in previous studies [49–
with metastasis (M1a stage). For this reason, MPE offers a 53], Yang and coworkers surprisingly found that CK−CTCs
direct evaluation of metabolic states of disseminating tumor were a prevalent population in half of NSCLC patients with
cells and their metabolic heterogeneity at the single-cell level. positive CTC counts [41]. At the molecular level, CK−CTCs
Li and coworkers employed two fluorescent metabolic probes exhibited consistent genomic copy number alternations but
(2-NBDG: glucose analog; resazurin: redox indicator) to distinct transcriptome signatures compared to the CK+ coun-
assay all nucleated cells in MPE from non-small cell lung terpart. HK2 as a metabolic-function associated marker
cancer (NSCLC) patients for identifying metabolically active allows to identify CK−CTCs, leading to classification of CTC
disseminating tumor cells and resolving their ­metabolic phe- based on CK expression. Meanwhile, HK2high and HK2low
notypes with single-cell resolution [36, 37]. Single-­cell CTCs represented different CTC subgroups with distinct met-
sequencing was performed to confirm the malignancy of abolic phenotype but similar genomic signatures [41].
these metabolically active, CD45 negative cells identified by
2-NBDG and resazurin markers. Three distinct metabolic
phenotypes of disseminating tumor cells were identified in 18.4  etabolic Adaption of CTCs
M
MPE. The first subset (2-NBDGhigh/resazurinlow) was raven- to Promote Their Survival
ous for glucose with high glycolysis but limited mitochon- and Metastasis
drial oxidation. The second subset (resazurinhigh/2-NBDGlow)
had limited glucose uptake and exhibited high mitochondrial To successfully seed distant metastasis, cancer cells detached
oxidation activity. The third subgroup that was a minority from primary sites as metastatic precursors face a daunting
phenotype (2-NBDGhigh/resazurinhigh) showed enhanced gly- series of challenges including loss of matrix adhesion, nutri-
colysis and mitochondrial oxidation. This is the first study ent deprivation, shear stress in circulation, and metabolic
that reveals the metabolic heterogeneity of disseminating stress at distant organs, requiring additional metabolic adap-
tumor cells at the single-cell level. Despite harboring same tion of cancer cells to these hostile environments. During
driver mutations, the difference in metabolic phenotypes of extracellular matrix detachment, untransformed epithelial
these cancer cells was found to associate with different ther- cells undergo cell death as a result of reduced glucose uptake,
apy response and survival of NSCLC patients [37]. Metabolic ATP depletion, and oxidative stress. In the suspended condi-
signature of disseminating tumor cells in MPE is capable of tion, epithelial cells were found to undergo metabolic repro-
predicting therapy resistance and patient survival. gramming by markedly upregulating pyruvate dehydrogenase
Compared with pleural effusion, disseminating tumor (PDH) kinase 4 (PDK4) for attenuating mitochondrial respi-
cells circulating in blood, termed as circulating tumor cells ration and oxidative stress [54]. Decreased glucose oxidation
(CTCs), become subject of intense investigation because following cell detachment is essential for anoikis resistance.
CTCs enable seeding metastases in distant organs [38, 39]. Unlike untransformed cells, most cancer cells exhibit
Metabolic states and heterogeneity of cancer cells in metasta- increasing glucose uptake and flux but reduced glucose oxi-
sis can be evaluated by assaying CTCs in blood samples. dation due to oncogenic signaling. This inherently possess a
Traditional CTC detection methods rely on epithelial markers survival advantage when suspended [54]. The Warburg effect
178 Z. Zhang and Q. Shi

is thus believed to be a defense mechanism protecting cancer pharmacological inhibition of FAO or YAP deletion inhibited
cells from oxidative stress and promoting anoikis resistance lymph node metastasis, but not blood-borne metastasis, indi-
and metastasis by increased production of antioxidants (e.g., cating that the metabolic shift toward FAO was preferentially
NADPH, glutathione) [11, 20, 55]. CTCs traveling in the required for lymph node metastasis.
bloodstream accumulate excessive ROS within cells due to
decreased NADPH and ATP production and are thus more
susceptible to cell death [55]. Metastasis-competent CTCs 18.5 Targeting Metabolic Transformation
need to manage the oxidative stress by restoring the skewed for Improving Cancer Therapeutics
redox balance for prolonging cell survival.
Piskounova and coworkers demonstrated that the oxida- The altered metabolic pathway of cancer cells has been
tive stress was a significant hurdle for melanoma dissemina- exploited for visualizing tumor in vivo by imaging the
tion and thereby limited distant metastasis in vivo [56].During uptake of radiolabeled glucose analog 18F-FDG. This meta-
metastasis, ROS levels increased in circulating melanoma bolic distinction also represents attractive therapeutic tar-
cells. To combat ROS and to maintain cell survival, success- gets (e.g., key metabolic enzymes) for anticancer treatment
fully metastasizing CTCs underwent reversible metabolic [60–63]. Anti-glycolytic treatment (2-deoxyglucose, 2-DG)
changes that increased NADPH and conferred oxidative was first tested in tumor model [64] and used to treat cancer
stress resistance by promoting the folate pathway in one-car- patients [65] in the 1950s. 2-DG as a glucose analog is phos-
bon metabolism. NADPH-generating enzymes ALDH1L2 phorylated by HK to 2-DG-phosphate, which cannot be fur-
and MTHFD1 in the folate pathway were used as suppressors ther metabolized. Accumulation of 2-DG in cells thus
of ROS-mediated toxicity. ALDH1L2 or MTHFD1 knock- inhibits glycolysis causing ATP depletion, cell cycle inhibi-
down was found to inhibit distant metastasis without signifi- tion, and cell death. Although there are no ongoing clinical
cantly affecting the subcutaneous tumors [56]. However, it is trials using 2-DG as a single agent, combining 2-DGwith
also possible that AMPK activation in CTCs reduces NADPH radiation or chemotherapeutic treatments were found to
consumption by fatty acid synthesis (FAS) and regenerates potentiate the tumor-destroying effects and enhance the
NADPH by induction of fatty acid oxidation (FAO), leading clinical efficacy [66].
to NADPH homeostasis and reduction of elevated ROS. Besides increasing glucose uptake, cancer cells facilitate
Zheng and coworkers compared single-cell RNA-seq pro- the first enzymatic step in elevated glucose metabolism by
files of CTCs from breast, prostate, and lung cancers, and markedly inducing upregulation of HK2 in addition to HK1,
found consistent induction of β-globin that was overex- which is already expressed in normal cells. Both HK1 and
pressed to suppress intracellular ROS within CTCs [57]. HK2 bind to mitochondria in a voltage-dependent anion
β-Globin expression within CTCs protected them from ROS-­ channel (VDAC)-dependent manner, and AKT promotes
induced apoptosis, promoted survival in circulation, and HK1/HK2 association with mitochondria [67]. HK1 and
contributed to their ability to initiate distant metastases. HK2 possess two tandem catalytic domains, but only one is
Tasdogan and coworkers found that upregulation of cell sur- active in HK1, whereas both are active in HK2 [68]. HK2
face MCT1 transporter in circulating melanoma cells induction in cancer cells is driven by multiple oncogenic
enhanced uptake of circulating lactate for managing oxida- pathways and is transcriptionally upregulated by MYC [43,
tive stress of CTCs [30]. MCT1 inhibition reduced lactate 69–71]. However, why cancer cells induce HK2 rather than
uptake and increased ROS levels, resulting in depletion of further increasing HK1 expression is largely unknown. Due
CTCs and decreased metastatic burden in patient-derived to the structural similarities between HK1 and HK2, it could
xenografts and in mouse melanomas [30]. Chen and cowork- be challenging for developing inhibitors that preferentially
ers established a CTC-derived, brain metastatic cell line and inhibit HK2. To date, a few HK2 inhibitors were reported but
these brain metastatic cells showed enhanced mitochondrial with limited success [72–74]. Recent studies demonstrated
respiratory pathways for energy production and upregulation that HK1+/HK2+ cancers could tolerate HK2 silencing or
of pentose phosphate pathway and the glutathione system selective inhibition, but the subset of HK1−/HK2+ tumors
which can minimize production of ROS resulting from an was found sensitive to therapeutic HK2 inhibition [75–78].
enhanced oxidative metabolism [58]. These metabolic For this reason, stratification of tumors that express HK2, but
changes were associated with strongly enhanced tumor cell not the housekeeping HK1 isozyme, should identify tumors
survival and proliferation in the brain microenvironment. treatable with HK2 inhibitors. An interesting example is
In addition to metabolic adaption for blood-borne metas- hepatocellular carcinoma (HCC). Normal hepatocytes typi-
tasis, Lee and coworkers identified a metabolic shift of mela- cally express HK4 rather than HK1. HCC cells exhibit sup-
noma cells toward FAO during their lymph node metastasis pression of HK4 expression and predominant induction of
[59]. Yes-associated protein (YAP), a downstream mediator HK2 expression [79]. Thus, systemic delivery of an HK2
of the Hippo pathway, was found to be selectively activated inhibitor could selectively target HCC cells but not normal
in lymph node metastases and thus induce FAO. Interestingly, hepatocytes.
18 Metabolic Reprogramming of Circulating Tumor Cells for Metastasis 179

Despite accelerated aerobic glycolysis in cancer cells, proliferating cells and cancer cells remains a major challenge
normal proliferating cells are known to use the same meta- in the development of cancer therapy targeting metabolic
bolic pathway and glycolytic enzymes as cancer cells. Thus, transformation. Table 18.1 shows metabolic-specific inhibi-
the inhibition of glycolytic enzymes for treating cancer raises tors that are specifically targeting the altered metabolic cir-
concerns on the risk of adverse and undesirable effects on cuits of tumors under developing or in clinical trials. Until
normal cells. Finding a therapeutic window between normal now, none of the tested metabolic therapies were approved to

Table 18.1 Compounds targeting cancer metabolism in clinical studies

Targeted Clinical Trials. gov
metabolism Metabolic inhibitors Targeted metabolic enzymes Cancer types Study phase reference
Glycolysis 2-deoxy-D-glucose (2-DG) HK Advanced solid tumors Phase I NCT00096707a
Phase I/II NCT00633087b
Lonidamine(LND) HK Benign prostatic Phase III NCT00435448b
hyperplasia Phase II NCT00237536b
3-Bromopyruvate (3-BrPA) HK/GAPDH Pancreatic cancer, Phase I N/A
metastatic melanoma
Silibin-Phytosome GLUTs Prostate cancer Phase II NCT00487721a
Phloretin GLUTs Colon cancer, leukemia Preclinical N/A
WZB117 GLUTs Lung cancer, breast Preclinical N/A
cancer
TLN-232 Pyruvate kinase Melanoma Phase II NCT00735332b
Oxamate LDHA Breast cancer Preclinical N/A
FX11 LDHA Lymphoma Preclinical N/A
Citric acid cycle Dichloroacetate (DCA) PDK Many tumor types Phase II NCT01029925b
Phase II NCT01386632a
Phase I NCT01111097a
CPI-613 PDH/α-ketoglutarate Lymphoma, leukemia Phase II NCT03793140c
dehydrogenase
AG-120 (Ivosidenib) Mutant IDH1 Advanced solid tumors Phase II NCT04195555c
IDH305 MutantIDH1 Advanced malignancies Phase I NCT02381886d
FT-2102 MutantIDH1 Hepatobiliary tumors Phase I/II NCT03684811d,e
AG-221 (Enasidenib) Mutant IDH2 Hematologic neoplasms Phase I/II NCT01915498d
AG-881 Pan-mutant IDH Advanced hematologic Phase I NCT02492737a
malignancies
AZD3965 MCTs Adult solid tumors Phase I NCT01791595d
Oxidative Metformin Mitochondrial AMPK Colorectal cancer Phase III NCT02614339c
phosphorylation
Arginine pathway Arginine deiminase ASS1 Melanoma, carcinoma Phase II NCT00450372a
(ADI-PEG-20) Phase I NCT00029900a
Phase II NCT00056992a
INCB001158 Arginase Advanced-stage solid Phase I/II NCT02903914d,e
tumors
Asparagine Asparaginase Depletion of circulating Acute lymphoblastic Phase III NCT00165178a,e
pathway asparagine leukemia Phase III NCT00400946a,e
Glutamine Phenylacetate Glutamine synthetase and Brain and central Phase II NCT00003241a
pathway glutaminase nervous system tumors Phase II NCT00006450a
Phase I NCT00001565a
CB-839 Glutaminase Solid tumors Phase I/II NCT03875313c,e
Phase I/II NCT02771626e,f
BHV-4157 (Trigriluzole) EAAT Metastatic or Phase I NCT03229278a,e
Unresectable solid
tumors or lymphoma
Fatty acid Omeprazole Proton pump Breast cancer Phase II NCT02595372d
synthesis (H+/K + -ATPase)
TVB-2640 FASN Solid malignant tumors Phase I/II NCT03808558c
NCT02223247a
Orlistat FASN N/A Preclinical N/A
C75 FASN N/A Preclinical N/A
(continued)
180 Z. Zhang and Q. Shi

Table 18.1 (continued)

Targeted Clinical Trials. gov
metabolism Metabolic inhibitors Targeted metabolic enzymes Cancer types Study phase reference
Nucleotide 5-fluorouracil(5-FU) Thymidylate synthase Many tumor types US FDA >100
synthesis approved
Cytarabine DNA polymerase/ Many tumor types US FDA >100
ribonucleotide reductase approved
Methotrexate DHFR Advanced colorectal Phase II NCT00952016b
cancer
Pemetrexed DHFR/thymidylate synthase Head and neck cancer Phase II NCT00293579a
/GARFT
6-mercaptopurine (6-MP) 5-phosphoribosyl Acute lymphoblastic Phase IV >30
and 6-thioguanine (6-TG) pyrophosphate (PRPP) leukemia(ALL)
amidotransferase
Leflunomide DHODH Brain and central Phase III NCT00003293a,e
nervous system tumors
pH regulation SLC0111 Carbonic anhydrase Solid tumors Phase I NCT02215850a
Phase I/II NCT03450018c,e
E7070 (Indisulam) Carbonic anhydrase Gastric cancer Phase I NCT00060567a,e
Phase I/II NCT00165594b
Akt/mTOR/P13K Halofuginone hydrobromide Akt/mTORC1 Kaposi sarcoma Phase II NCT00064142a
signaling
Rapamycin mTORC1 Prostate cancer Phase I/II NCT00311623a
Temsirolimus and Everlimus mTORC1 Solid tumors US FDA >100
approved
Ridaforolimus and related mTORC1 Solid tumors Phase I >10
compounds
BEZ235 PI3K/mTOR Malignant solid tumors Phase I NCT01343498a
GSK2110183 (Afuresertib) PI3K/Akt Many tumor types Phase II NCT01531894a
HIF signaling PX-478 HIF1α Advanced solid tumors Phase I NCT00522652a
IGF signaling MK-0646 IGF1R Non-small cell lung Phase II NCT00799240a,e
cancer
BIIB022 IGF1R Solid tumors Phase I NCT00555724a
AVE1642 IGF1R Liver carcinoma Phase I/II NCT00791544b,e
IDO signaling INCB024360 (Epacadostat) IDO1 Urothelial cancer Phase III NCT03361865a,e
BMS-986205 (Linrodostat) IDO1 Advanced cancers Phase I/II NCT03459222c,e
Indoximod IDO1/2 Melanoma Phase II NCT03301636b,e
Angiogenesis Bevacizumab and related Hypoxia/VEGF/VEGFR Solid tumors US FDA >100
compounds approved
Aflibercept Angiogenesis factors and Metastatic colorectal Phase I/II NCT01661972a,e
receptors cancer
Regorafenib Angiogenesis factors and Metastatic colorectal Phase II NCT02466009d
receptors cancer Phase II NCT02699073a
Metabolic Imprime PGG (β- glucan) Innate immune cells Advanced melanoma, Phase II NCT02981303d,e
reprogramming triple-negative breast
cancer
Tirapazamine Hypoxia Cervical cancer, lung Phase I NCT00098995a,e
cancer Phase III NCT00017459a,e
HK Hexokinase, GAPDH Glyceraldehyde-3-phosphate dehydrogenase, GLUT Glucose transporter, LDHA Lactate dehydrogenase, PDK Pyruvate
dehydrogenase kinase, PDH Pyruvate dehydrogenase, IDH Isocitrate dehydrogenase, MCT Monocarboxylate transporter, AAS1 Argininosuccinate
synthase 1, EAAT: Excitatory amino acid transporters, FASN Fatty acid synthase, DHFR Dihydrofolate reductase, GARFT Glycinamide ribonucle-
otide formyltransferase, DHODH Dihydroorotate dehydrogenase
a
Completed
b
Terminated
c
Recruiting
d
Active, not recruiting
e
In combination with other drugs
f
Unknown
18 Metabolic Reprogramming of Circulating Tumor Cells for Metastasis 181

enter routine clinical practice. Due to the metabolic plasticity 4. Chambers AF, Groom AC, MacDonald IC. Dissemination and
of cancer cells, metabolism-inhibiting treatment may require growth of cancer cells in metastatic sites. Nat Rev Cancer.
2002;2(8):563–72.
inhibiting multiple metabolic pathways simultaneously or 5. Senft D, Ronai ZEA. Adaptive stress responses during tumor
coupled with standard chemo-, targeted or immunotherapy. metastasis and dormancy. Trends Cancer. 2016;2(8):429–42.
6. Luzzi KJ, MacDonald IC, Schmidt EE, Kerkvliet N, Morris VL,
Chambers AF, Groom AC. Multistep nature of metastatic inef-
ficiency: dormancy of solitary cells after successful extravasa-
18.6 Conclusions tion and limited survival of early micrometastases. Am J Pathol.
1998;153(3):865–73.
Metabolic reprogramming of cancer cells has been exten- 7. Elia I, Doglioni G, Fendt SM. Metabolic hallmarks of metastasis
sively studied over the past century and is now recognized to formation. Trends Cell Biol. 2018;28(8):673–84.
8. Chemi F, Rothwell DG, McGranahan N, et al. Pulmonary venous
be a common feature of many cancer cells. The inherent circulating tumor cell dissemination before tumor resection and
metabolic heterogeneity of cancer cells and acquired meta- disease relapse. Nat Med. 2019;25(10):1534–9.
bolic adaptations during dissemination promote CTC sur- 9. Valastyan S, Weinberg RA. Tumor metastasis: molecular insights
vival in circulation and seeding metastases at distant organs. and evolving paradigms. Cell. 2011;147(2):275–92.
10. Hanahan D, Weinberg RA. Hallmarks of cancer: the next genera-
CTC-derived cell lines [59, 80–83] and xenografts [84, 85] tion. Cell. 2011;144(5):646–74.
have been established from only hundreds of CTCs, as com- 11. Hamanaka RB, Chandel NS. Cell biology. Warburg effect and
pared to enormous challenges to establish cell lines from redox balance. Science. 2011;334(6060):1219–20.
millions of primary tumor cells. This might be explained by 12. Warburg O, Posener K, Negelein E. Ueber den stoffwechsel der
tumoren. Biochem Z. 1924;152(1):319–44.
the advantages of CTCs in survival and growth. Although 13. Warburg O, Wind F, Negelein E. The metabolism of tumors in the
culturing CTCs in vitro or expanding in vivo using xeno- body. J Gen Physiol. 1927;8(6):519–30.
grafts have significant clinical applications, successful 14. Warburg O. On the origin of cancer cells. Science.
examples are still rare because of low numbers of CTCs in 1956;123(3191):309–14.
15. Som P, Atkins HL, Bandoypadhyay D, et al. A fluorinated glucose
patients. In addition to analyzing CTCs temporally and spa- analog, 2-fluoro-2-deoxy-D-glucose (F-18): nontoxic tracer for
tially (at different vascular sites), new analytical approaches rapid tumor detection. J Nucl Med. 1980;21(7):670–5.
such as single-cell, quantitative metabolome profiling would 16. Hsu PP, Sabatini DM. Cancer cell metabolism: Warburg and
allow us to map the entire metabolic circuits of metastatic beyond. Cell. 2008;134(5):703–7.
17. Weinhouse S. On respiratory impairment in cancer cells. Science.
cancer cells for exploring the diagnostic and therapeutic 1956;124(3215):267–9.
potential [86, 87]. 18. Warburg O. On respiratory impairment in cancer cells. Science.
1956;124(3215):269–70.
19. Corbet C, Feron O. Cancer cell metabolism and mitochondria:
nutrient plasticity for TCA cycle fueling. Biochim Biophys Acta
Open Questions Rev Cancer. 2017;1868(1):7–15.
20. Vander HMG, Cantley LC, Thompson CB. Understanding the
• How do cancer cells acquire additional metabolic Warburg effect: the metabolic requirements of cell proliferation.
features to adapt to hostile environment during Science. 2009;324(5930):1029–33.
metastasis? 21. Gatenby R, Gillies R. Why do cancers have high aerobic glycoly-
• How to perform metabolic phenotyping of CTCs sis? Nat Rev Cancer. 2004;4(11):891–9.
22. Ahn CS, Metallo CM. Mitochondria as biosynthetic factories for
for prognosis and prediction of therapy response? cancer proliferation. Cancer Metab. 2015;3(1):1.
• How to identify and selectively kill CTCs with high 23. Pavlova NN, Thompson CB. The emerging hallmarks of cancer
metastatic potential by targeting therapeutic vulner- metabolism. Cell Metab. 2016;23(1):27–47.
ability from altered metabolic circuits? 24. Hensley CT, Wasti AT, DeBerardinis RJ. Glutamine and cancer:
cell biology, physiology, and clinical opportunities. J Clin Invest.
• How to identify early metastasis in cancer patients 2013;123(9):3678–84.
by globally detecting CTCs with different 25. Wang MD, Wu H, Fu GB, et al. Acetyl-coenzyme a carboxyl-
phenotypes? ase alpha promotion of glucose-mediated fatty acid synthesis
enhances survival of hepatocellular carcinoma in mice and patients.
Hepatology. 2016;63(4):1272–86.
26. Lehuédé C, Dupuy F, Rabinovitch R, Jones RG, Siegel
PM. Metabolic plasticity as a determinant of tumor growth and
metastasis. Cancer Res. 2016;76(18):5201–8.
References 27. LeBleu VS, O’Connell JT, Gonzalez Herrera KN, et al. PGC-1α
mediates mitochondrial biogenesis and oxidative phosphory-
1. Gupta GP, Massague J. Cancer metastasis: building a framework. lation in cancer cells to promote metastasis. Nat Cell Biol.
Cell. 2006;127(4):679–95. 2014;16(10):992–1003.
2. Steeg PS. Targeting metastasis. Nat Rev Cancer. 2016;16(4):201–18. 28. Cai Z, Li CF, Han F, et al. Phosphorylation of PDHA by AMPK
3. Vanharanta S, Massagué J. Origins of metastatic traits. Cancer Cell. drives TCA cycle to promote cancer metastasis. Mol Cell.
2013;24(4):410–21. 2020;80(2):263–78.
182 Z. Zhang and Q. Shi

29. Porporato PE, Payen VL, Pérez-Escuredo J, et al. A mitochondrial 50. Mohme M, Riethdorf S, Pantel K. Circulating and disseminated
switch promotes tumor metastasis. Cell Rep. 2014;8(3):754–66. tumour cells - mechanisms of immune surveillance and escape. Nat
30. Tasdogan A, Faubert B, Ramesh V, et al. Metabolic heterogene- Rev Clin Oncol. 2017;14(3):155–67.
ity confers differences in melanoma metastatic potential. Nature. 51. Agerbæk MØ, Bang-Christensen SR, Yang MH, et al. The
2020;577(7788):115–20. VAR2CSA malaria protein efficiently retrieves circulating
31. Hensley CT, Faubert B, Yuan Q, et al. Metabolic heterogeneity in tumor cells in an EpCAM-independent manner. Nat Commun.
human lung tumors. Cell. 2016;164(4):681–94. 2018;9(1):3279.
32. Sengupta D, Pratx G. Imaging metabolic heterogeneity in cancer. 52. Yu M, Bardia A, Wittner BS, et al. Circulating breast tumor cells
Mol Cancer. 2016;15(1):1–12. exhibit dynamic changes in epithelial and mesenchymal composi-
33. Gerlinger M, Rowan AJ, Horswell S, et al. Intratumor heterogene- tion. Science. 2013;339(6119):580–4.
ity and branched evolution revealed by multiregion sequencing. N 53. Pecot CV, Bischoff FZ, Mayer JA, et al. A novel platform for detec-
Engl J Med. 2012;366(10):883–92. tion of CK+ and CK- CTCs. Cancer Discov. 2011;1(7):580–6.
34. Dupuy F, Tabaries S, Andrzejewski S, et al. PDK1-dependent meta- 54. Kamarajugadda S, Stemboroski L, Cai Q, Simpson NE, Nayak
bolic reprogramming dictates metastatic potential in breast cancer. S, Tan M, Lu J. Glucose oxidation modulates anoikis and tumor
Cell Metab. 2015;22(4):577–89. metastasis. Mol Cell Biol. 2012;32(10):1893–907.
35. Zhang Y, Guillermier C, De Raedt T, et al. Imaging mass spec- 55. Schafer ZT, Grassian AR, Song L, et al. Antioxidant and oncogene
trometry reveals tumor metabolic heterogeneity. iScience. rescue of metabolic defects caused by loss of matrix attachment.
2020;23(8):101355. Nature. 2009;461(7260):109–13.
36. Tang Y, Wang Z, Li Z, et al. High-throughput screening of rare 56. Piskounova E, Agathocleous M, Murphy MM, et al. Oxidative
metabolically active tumor cells in pleural effusion and periph- stress inhibits distant metastasis by human melanoma cells. Nature.
eral blood of lung cancer patients. Proc Natl Acad Sci U S A. 2015;527(7577):186–91.
2017;114(10):2544–9. 57. Zheng Y, Miyamoto DT, Wittner BS, et al. Expression of β-globin
37. Li Z, Wang Z, Tang Y, et al. Liquid biopsy-based single-cell meta- by cancer cells promotes cell survival during blood-borne dissemi-
bolic phenotyping of lung cancer patients for informative diagnos- nation. Nat Commun. 2017;8(1):14344.
tics. Nat Commun. 2019;10(1):3856. 58. Chen EI, Hewel J, Krueger JS, et al. Adaptation of energy
38. Keller L, Pantel K. Unravelling tumour heterogeneity by single-­ metabolism in breast cancer brain metastases. Cancer Res.
cell profiling of circulating tumour cells. Nat Rev Cancer. 2007;67(4):1472–86.
2019;19(10):553–67. 59. Lee CK, Jeong SH, Jang C, et al. Tumor metastasis to lymph
39. Pantel K, Alix-Panabières C. Liquid biopsy and minimal residual nodes requires YAP-dependent metabolic adaptation. Science.
disease—latest advances and implications for cure. Nat Rev Clin 2019;363(6427):644–9.
Oncol. 2019;16(7):409–24. 60. Zhao Y, Butler EB, Tan M. Targeting cellular metabolism to
40. Werner S, Keller L, Pantel K. Epithelial keratins: biology and improve cancer therapeutics. Cell Death Dis. 2013;4(3):e532.
implications as diagnostic markers for liquid biopsies. Mol Asp 61. Vander Heiden MG. Targeting cancer metabolism: a therapeutic
Med. 2020;72(5):100817. window opens. Nat Rev Drug Discov. 2011;10(9):671–84.
41. Yang L, Yan X, Chen J, et al. Hexokinase 2 discerns a novel 62. Tennant DA, Duran RV, Gottlieb E. Targeting metabolic transfor-
circulating tumor cell population associated with poor prog- mation for cancer therapy. Nat Rev Cancer. 2010;10(4):267–77.
nosis in lung cancer patients. Proc Natl Acad Sci U S A. 63. Hay N. Reprogramming glucose metabolism in cancer: can it be
2021;118(11):e2012228118. exploited for cancer therapy. Nat Rev Cancer. 2016;16(10):635–49.
42. Mathupala SP, Ko YH, Pedersen PL. Hexokinase II: cancer’s double-­ 64. Ely JO. 2-deoxy-D-glucose as an inhibitor of cancerous growth in
edged sword acting as both facilitator and gatekeeper of malignancy animals. J Frankl Inst. 1954;258(2):157–60.
when bound to mitochondria. Oncogene. 2006;25(34):4777–86. 65. Landau BR, Laszlo J, Stengle J, Burk D. Certain metabolic and
43. Patra KC, Wang Q, Bhaskar PT, et al. Hexokinase 2 is required pharmacologic effects in cancer patients given infusions of
for tumor initiation and maintenance and its systemic dele- 2-deoxy-D-glucose. J Natl Cancer Inst. 1958;21(3):485–94.
tion is therapeutic in mouse models of cancer. Cancer Cell. 66. Dwarakanath B, Jain V. Targeting glucose metabolism with
2013;24(2):213–28. 2-deoxy-D-glucose for improving cancer therapy. Fut Oncol.
44. Wang L, Xiong H, Wu F, et al. Hexokinase 2-mediated Warburg 2009;5(5):581–5.
effect is required for PTEN- and p53-deficiency-driven prostate 67. Gottlob K, Majewski N, Kennedy S, Kandel E, Robey RB, Hay
cancer growth. Cell Rep. 2014;8(5):1461–74. N. Inhibition of early apoptotic events by Akt/PKB is dependent
45. DeWaal D, Nogueira V, Terry AR, et al. Hexokinase-2 depletion on the first committed step of glycolysis and mitochondrial hexoki-
inhibits glycolysis and induces oxidative phosphorylation in hepa- nase. Genes Dev. 2001;15(11):1406–18.
tocellular carcinoma and sensitizes to metformin. Nat Commun. 68. Tsai HJ, Wilson JE. Functional organization of mammalian hexo-
2018;9(1):446. kinases: both N- and C-terminal halves of the rat type II isozyme
46. Keller L, Werner S, Pantel K. Biology and clinical relevance of possess catalytic sites. Arch Biochem Biophys. 1996;329(1):17–23.
EpCAM. Cell Stress. 2019;3(6):165–80. 69. Katabi MM, Chan HL, Karp SE, Batist G. Hexokinase type II: a
47. Gires O, Pan M, Schinke H, Canis M, Baeuerle PA. Expression and novel tumor-specific promoter for gene-targeted therapy differen-
function of epithelial cell adhesion molecule EpCAM: where are tially expressed and regulated in human cancer cells. Hum Gene
we after 40 years? Cancer Metastasis Rev. 2020;39(3):969–87. Ther. 1999;10(2):155–64.
48. Quintanal-Villalonga Á, Chan JM, Yu HA, Pe’er D, Sawyers CL, 70. Pedersen PL, Mathupala S, Rempel A, Geschwind JF, Ko
Sen T, Rudin CM. Lineage plasticity in cancer: a shared pathway of YH. Mitochondrial bound type II hexokinase: a key player in the
therapeutic resistance. Nat Rev Clin Oncol. 2020;17(6):360–71. growth and survival of many cancers and an ideal prospect for thera-
49. Stoecklein NH, Fischer JC, Niederacher D, Terstappen peutic intervention. Biochim Biophys Acta. 2002;1555(1–3):14–20.
LW. Challenges for CTC-based liquid biopsies: low CTC frequency 71. Kim JW, Gao P, Liu YC, Semenza GL, Dang CV. Hypoxia-­
and diagnostic leukapheresis as a potential solution. Expert Rev inducible factor 1 and dysregulated c-Myc cooperatively induce
Mol Diagn. 2016;16(2):147–64. vascular endothelial growth factor and metabolic switches hexo-
18 Metabolic Reprogramming of Circulating Tumor Cells for Metastasis 183

kinase 2 and pyruvate dehydrogenase kinase 1. Mol Cell Biol. ble implications in novel therapeutic strategies. Cancers (Basel).
2007;27(21):7381–93. 2020;12(6):1668.
72. Lin H, Zeng J, Xie R, et al. Discovery of a novel 2,6-disubstituted 80. Cayrefourcq L, Mazard T, Joosse S, et al. Establishment and char-
glucosamine series of potent and selective hexokinase 2 inhibitors. acterization of a cell line from human circulating colon cancer cells.
ACS Med Chem Lett. 2016;7(3):217–22. Cancer Res. 2015;75(5):892–901.
73. Behar V, Pahima H, Kozminsky-Atias A, et al. A hexokinase 2 81. Pantel K, Alix-Panabières C. Cell lines from circulating tumor
modulator for field-directed treatment of experimental actinic kera- cells. Onco Targets Ther. 2015;2(10):815–6.
toses. J Invest Dermatol. 2018;138(12):2635–43. 82. Zhang L, Ridgway LD, Wetzel MD, et al. The identification and
74. Li W, Zheng M, Wu S, et al. Benserazide, a dopadecarboxylase characterization of breast cancer CTCs competent for brain metas-
inhibitor, suppresses tumor growth by targeting hexokinase 2. J Exp tasis. Sci Transl Med. 2013;5(180):180ra48.
Clin Canc Res. 2017;36(1):58. 83. Yu M, Bardia A, Aceto N, et al. Ex vivo culture of circulating
75. Xu S, Catapang A, Doh HM, et al. Hexokinase 2 is targetable for breast tumor cells for individualized testing of drug susceptibility.
HK1-negative, HK2-positive tumors from a wide variety of tissues Science. 2014;345(6193):216–20.
of origin. J Nucl Med. 2018:jnumed.118.212365. 84. Simpson KL, Stoney R, Frese KK, et al. A biobank of small cell
76. Xu S, Catapang A, Braas D, et al. A precision therapeutic strat- lung cancer CDX modelselucidates inter- and intratumoral pheno-
egy for hexokinase 1-null, hexokinase 2-positive cancers. Cancer typicheterogeneity. Nat Cancer. 2020;1(4):437–51.
Metab. 2018;6(1):7. 85. Baccelli I, Schneeweiss A, Riethdorf S, et al. Identification of a
77. Xu S, Herschman HR. A tumor agnostic therapeutic strategy for populationof blood circulating tumor cells from breast cancer
hexokinase 1-null/hexokinase 2-positive cancers. Cancer Res. patientsthat initiates metastasis in a xenograft assay. NatBiotechnol.
2019;79(23):5907–14. 2013;31(6):539–44.
78. Xu S, Zhou T, Doh HM, et al. An HK2 antisense oligonucleotide 86. Alix-Panabières C, Pantel K. Challenges in circulatingtumour cell
induces synthetic lethality in HK1-HK2+ multiple myeloma. research. Nat Rev Cancer. 2014;14(9):623–31.
Cancer Res. 2019;79(10):2748–60. 87. Sun YF, Wu L, Liu SP, et al. Dissecting spatial heterogeneity and
79. Mossenta M, Busato D, Dal Bo M, Toffoli G. Glucose metabolism theimmune-evasionmechanism of CTCs by single-cellRNA-seq
and oxidative stress in hepatocellular carcinoma: role and possi- in hepatocellular carcinoma. Nat Commun. 2021;12(1):4091.
 Part VI
Lymphatic System and Cancer Metastasis

Marlys H. Witte

Historically, recognition that cancers (tumors) spread through the lymphatic system—as evi-
denced by enlarged regional lymph nodes observed in dying cancer patients—long preceded
Virchow’s explanation of the origin of cancer in aberrant cells. There was then not even the
most fundamental understanding of lymphatic system function in general and specifically its
relation to the involved nodes. Indeed, the view held at the time was that cancers arose in
“lymph”—or tissue fluid, a notion that has regained traction recently in expanded investigation
of the role of the “tumor microenvironment” in the initiation and spread of cancer—a topic
covered in other sections of this textbook.
Gaspar Asellius of Padua discovered the chylous lymph circulation (“lacteis venis”) in a
postprandial dog in 1627, remarkably the same year that William Harvey described the blood
circulation. In the late 1890s, Ernest Starling’s elucidation of the laws governing transcapillary
fluid movement, which he termed “lymph formation,” ushered in the era of lymphatic “syste-
momics” (Fig. 1) in the first half of the twentieth century—embryonic development and growth
and regrowth of the lymphatics (currently termed “lymphvasculogenesis” and “lymphangio-
genesis”), transcapillary leakage and recirculation of plasma proteins, lymphocyte trafficking,
dietary fat absorption, and immune network. Specific distinguishing features characteristic of
lymphatic vessels—ultrastructure, intrinsic contractility, valve function, and lymphatic-venous
communications—were described.
Meanwhile, lymphatic system imaging was advancing beyond visual vital dye lymphangi-
ography to include new modalities of direct conventional oil contrast lymphography followed
by radionuclide lymphoscintigraphy and fluorescence microlymphangiography. These first
two techniques allowed dynamic visualization of the peripheral and central lymphatic vascu-
lature and lymph nodes in continuity in the living human along with characteristic abnormali-
ties arising in disease states, particularly lymphomas and metastatic process but also
lymphedemas and lymphatic malformations. They also laid the groundwork for image-guided
access for interventional therapies enhanced by magnetic resonance imaging and computed
tomography. Further insights into disturbances of lymph circulation resulted from thoracic
duct cannulation, lymph sampling, drainage, and also lymph diversion and lymphatic decom-
pression under conditions of lymph overload and lymphatic obstruction.
In the late 1990s, the era of molecular lymphology dawned with the discovery of specific
vascular growth factor ligand-receptor combinations (VEGF and angiopoietin families) and a
cascade of lymphatic-specific or predominantly lymphatic-directed molecules incorporated in
the signaling pathways for lymphvasculogenesis and lymphangiogenesis. Lymphatic systemo-
mics is now being reenergized as these molecules and pathways find expression in clinical
syndromes and particularly hereditary lymphedema syndromes, where genotype-phenotype
studies are uncovering previously unrecognized and unsuspected links of the lymphatic system
to a wide variety of vital processes and disorders, including cancer and its spread.
186 Lymphatic System and Cancer Metastasis

Fig. 1 Lymphatic systemomics encompasses lymphatic system functions as a vasculature, circulation of extravascular extracellular fluid, route of
entry into the body, and (image according to Sir Gus Nossal) center of the immune network. Each component can promote or defend against cancer
spread

In this section, we will lay the groundwork for understanding the lymphatic system as a vasculature, a
circulation of extravascular extracellular fluid, a route of entry and transport of foreign particles and abnor-
mal cells, and the center of the immune system itself. In that light, we will relate the structure and function
of the lymphatic system to the events in and manifestations of cancer metastasis and how this system influ-
ences both positively and negatively the fate of the cancer cell and ultimately the fate of the host.
 Lymphatic System Biology,
Pathobiology, and Relation to Cancer 19
Metastasis

Marlys H. Witte and Sarah K. Daley

Abstract
lymph formation–lymph absorption balance and the
Although described since ancient times, the lymphatic imbalances in edema/lymphedema.
system was not fully recognized as an integrated vascula- • Name factors/forces involved in lymph flow and
ture, circulation, and center of the immune network until propulsion.
the discipline of lymphology was established a little more • List several ways lymphatic biology and pathobiol-
than half a century ago. This chapter reviews its distinc- ogy relate to cancer theranostics, spread, and
tive macroscopic and microscopic anatomy and the physi- complications.
ologic principles governing lymph formation, absorption,
and propulsion. In addition, we discuss the pathobiology
and therapeutic implications underlying common lym-
phatic disorders. Finally, links to cancer, its complica-
tions, and spread are highlighted. 19.1 Introduction

Scarcely visible during life and collapsed after death, lym-

Learning Objectives phatic vessels, in contrast to blood vessels, captured little
• Distinguish the vascular, circulatory, and immuno- attention following their discovery in ancient times and even
logic components of the lymphatic system. after Gaspar Asellius described the milky, chyliferous lymph
• Define the general macroscopic anatomy and distri- circulating in a well-fed dog [1]. On the other hand, Wil-
bution of the peripheral and central lymphatic liam Harvey’s discovery of the blood circulation published
vasculature. that same year (1627) excited an explosion of interest and
• Enumerate the distinctive microscopic features of spurred rapidly accumulating knowledge [2]. Nonetheless,
lymphatic vessels and their specific immunostain- the centuries that followed witnessed meticulous postmor-
ing properties. tem anatomic dissections establishing the nearly ubiquitous
• Define Starling’s law of transcapillary fluid presence of lymphatic vessels throughout the body. In 1896
exchange (lymph formation) and the principle of and the decades thereafter, the principles of transcapillary
fluid exchange (lymph formation) and transcapillary protein
movement and recirculation were elucidated along with the
processes of lymphatic propulsion, growth, and develop-
ment.
Beginning during the last half of the twentieth century,
with the ability to dynamically image lymphatic vessels and
M. H. Witte (*) lymph nodes in the living human, lymphology as a discipline
Department of Surgery, Neurosurgery, and Pediatrics, University
finally emerged. Not merely “lymph nodes held together by
of Arizona College of Medicine, Tucson, AZ, USA
e-mail: [email protected] strings,” the lymphatic system came to be fully recognized
as a distinctive vasculature, fluid circulation, route of entry
S. K. Daley
Advanced Pathology Research Fellow in Lymphology, Department into the body, and center of the immune network (“lymphatic
of Surgery, University of Arizona College of Medicine, systemomics”), thereby playing a vital role in both health
Tucson, AZ, USA and disease (Figure in Overview, Lymphatic Systemomics).
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 187
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_19
188 M. H. Witte and S. K. Daley

The close relationships between the growth of cancer, both lymph suspension of lymphocytes through lymph vessels
its containment and spread, and resulting complications is and lymph nodes. As a specialized subcompartment of the
one such example. extracellular space, therefore, the lymphatic system closes
The lymph circulation is a unidirectional vascular system the “blood–lymph loop” of the circulation by returning liq-
that transports surplus tissue fluid back to the bloodstream. uid, macromolecules, trafficking cells, and other blood ele-
Furthermore, this network stabilizes the mobile intercellular ments that “escape” or “leak” from blood capillaries.
liquid and extracellular matrix microenvironment to ensure
parenchymal cellular integrity and function. The lymphatic
system is composed of vascular conduits: lymphoid organs 19.2 Lymphatic Vasculature: Structural–
including the lymph nodes and cellular elements circulating Functional Aspects (Fig. 19.1) [3, 4]
in the liquid lymph, such as lymphocytes and macrophages.
These cells, migrating in the blood vascular system, cross the Identification of lymphatic vessels was long hampered by
blood capillary barrier along with immunoglobulins, poly- lack of readily identifiable structures. Subsequent anatomic
peptides, plasma protein complexes, and cytokines and enter dissections established the presence of lymphatics through-
the lymphatics to return, mixed with specific lymph “absor- out the body accompanying venous trunks everywhere except
bents” such as chylomicra, to the bloodstream. Whereas in the central nervous system and cortical bony skeleton. The
body water circulates very rapidly as a plasma suspension latter have their own special lymph drainage systems.
of red blood cells within the blood vascular compartment, it In general, lymph from the lower torso and viscera
percolates slowly outside the bloodstream as a tissue fluid-­ enters the bloodstream via the thoracic duct at the left sub-

a Normal b

Right Left
Lymphatic-venous Lymphatic-venous Lumen
junction (LVJ) junction (LVJ)
Central
veins

R lymph
duct Thoracic
duct

Cephalad Lymph
Cisterna Transport (LT)
chyli
Anchoring filaments

Basal Collagen
lamina fibrils
Regional
Contractile lymph Cephalad Lymph
lymphangion nodes Transport (LT)
Lymphatic
valves
Fibroblast

Lymph
/ absorption (LA) .

Fig. 19.1 Anatomy of the lymphatic system. (a) Schematic depicting the adjacent collagen bundles, thereby forming a firm connection
the normal macroscopic anatomy and direction of flow of the peripheral between the capillary wall and the surrounding interstitium. Reproduced
and central lymphatic system. (b) Lymphatic capillary reconstructed by with permission of Leak (1970) Microvasc Res 2: 361–391 [49]. See
collated electron micrographs The anchoring filaments originate from text for further details
the abluminal surface of the lymphatic endothelial cells and extend into
19 Lymphatic System Biology, Pathobiology, and Relation to Cancer Metastasis 189

clavian–jugular venous junction (Fig. 19.1a). Lymphatics relatively impervious to macromolecules even albumin. A
from the head and neck and upper extremities enter cen- variety of cells pass through specialized lymphatic endothe-
tral veins either independently or via a common supracla- lial junctions as well as migrate transcellularly.
vicular cistern. Numerous interconnections exist within this It is difficult to distinguish between blood and lymph ves-
rich vascular network. The bulk of cardiac and pulmonary sels at the tissue level without special stains, although the lat-
lymph empties into the great veins on the right side of the ter are usually thin-walled and tortuous, have a wider, more
neck whereas intestinal lymph transporting cholesterol, fat-­ irregular lumen, and are largely devoid of red blood cells.
soluble vitamins, and long-chain triglycerides as chylomicra Many staining features have been advocated to differentiate
courses retroperitoneally to the aortic hiatus, then joins with between blood and lymph microvasculature but it was not
other visceral and retroperitoneal lymphatics to form the cis- until recently that histochemical specificity has been refined
terna chyli and thoracic duct. Most liver lymph flows retro- [7]. One of the most commonly used specific markers cur-
grade or countercurrent to portal blood and joins intestinal rently in use is LYVE-1, with D-240 widely used to stain
lymph collectors just before the origin of the thoracic duct. clinical specimens. LYVE-1 has been applied to tissues rang-
Recognition of these topographic features and their variants ing from early mouse embryo to adult human and highlights
is essential for understanding patterns of cancer spread. initial collecting vessels and lymphatic capillaries but not
Although the retina and brain do not technically have larger-caliber lymphatics. The LYVE-1 receptor appears to
lymphatic apparatuses, they possess analogous circulations, be a key entry point for dendritic cells to selectively access
such as the aqueous humor canal of Schlemm (anterior the lymphatic system [8]. Another strong marker localizing
chamber of the eye) or the cerebrospinal fluid/subarachnoid to the nucleus of lymphatic endothelial cells and adaptable
villus (pacchionian bodies) connections (the brain). Natural to multiple tissues is the transcription factor PROX1 which
vasomotion of cerebral intraparenchymal arteries/arterioles regulates podoplanin, a transmembrane glycoprotein in
generates a force that moves fluid from the perivascular lymphatics but not blood vessel endothelial cells. D2-40,
space into the brain parenchyma crossing the glial perivas- the selective antibody toward podoplanin, used routinely in
cular sheet via water pores present in the membrane. Glial human tissue immunohistochemistry, also sharply distin-
elements with their astrocytic foot processes sweeping tis- guishes lymphatic from blood vessel endothelium and larger
sue fluid and nonendothelial-lined intracerebral perivascu- collecting lymphatics. This feature has been useful in iden-
lar (Virchow–Robin) spaces mix with cerebrospinal fluid to tifying preexisting and new lymphatics in tumors although
form the glymphatic system [5]. The latter empties into the there is overlap of cell types. These special stains have con-
endothelial-lined meningeal lymphatic capillaries and thence tributed greatly to delineating lymphatics within cancers and
to the contractile collecting cervical lymphatics, which pass their relationship to cancer metastasis.
through the cervical lymph nodes to enter the central lymph Ultrastructurally, lymph capillaries display both open
and bloodstream. (“button”) and “closed” (tight) (“zipper”) endothelial junc-
Regarding development of the lymphatic vasculature, tions and can dramatically adjust their shape and lumen size
vigorous controversy has persisted since the early 1900s [9, 10]. Unlike blood capillaries, a basal lamina (basement
about lymphvasculogenesis and how endothelial precursors membrane) is tenuous or altogether lacking. Moreover, com-
or stem cells, such as lymphangioblasts, differentiate and plex elastic fibrils, termed anchoring filaments, tether the
proliferate into a primitive tubular network. This network outer portions of the endothelium to a fibrous gel matrix
then expands by subsequent lymphangiogenesis (sprouting in the interstitium [11] (Fig. 19.1b). These filaments allow
from existing vessels) whether derived largely from central lymph microvessels to open wide causing a sudden increase
veins with growth centrifugally but also arising indepen- in tissue fluid load and pressure, in contrast to the simultane-
dently from tissue mesenchyme in peripheral tissues with ous collapse of adjacent blood capillaries. Just beyond the
growth centripetally. This subject is extensively reviewed in lymph capillaries, in contrast to more proximal and larger
the following chapter by Geng and Srinivathan along with its lymph collectors and trunks, the terminal lymphatics with a
relation to epi(endo)thelial-mesenchymal transition (EMT-­ continuous “zipper” structure are devoid of smooth muscle,
MET) and recapitulation in cancer metastasis [6]. although the endothelial lining is rich in the contractile pro-
Under light microscopy, efferent lymphatics originate tein actin. Intraluminal bicuspid valves are also prominent
within the interstitium as specialized capillaries or less com- features that partition the lymphatic vessels into discrete
monly (e.g., liver, brain) from nonendothelialized precapil- contractile segments termed lymphangions. These special-
lary channels. Lymphatic capillaries are remarkably porous ized microscopic features support the delicate functions of
and readily permit entry of large macromolecules (molecular absorbing and transporting various elements including large
weight > 1000 kD). In this respect, they resemble the uniquely protein moieties and cells from the bloodstream but also for-
“leaky” fenestrated hepatic sinusoidal blood capillaries but eign agents (e.g., viruses, bacteria) that gain access to the
are in distinct contrast to most blood capillaries, which are interstitial space.
190 M. H. Witte and S. K. Daley

Progressive development of techniques to image the of fluid escaping from the bloodstream (lymph formation)
dynamic in vivo structure and function of lymphatic vessels depends primarily on the transcapillary balance of hydro-
built on centuries of technical advances. After first visual- static and protein osmotic (oncotic) pressure gradients as
izing small channel “streamers” from intracutaneous vital modified by the character (i.e., permeability, surface area,
dyes, direct oil contrast lymphography (LAG) followed in the solute reflection coefficient) of the filtering microvascular
mid-1950s, opening a window on the lymphatic system and surface (depicted in Fig. 19.2, top). Normally, a small excess
its participation in a variety of disorders [12]. Initial interest of tissue fluid forms continuously (net capillary filtration),
focused on primary tumors of the lymphatic system (lympho- and this surplus enters the lymphatic system and returns to
mas) and cancer primaries and metastases along with staging the venous system (here, referred to as “lymph absorption”).
of these conditions. Modern imaging techniques (discussed In contrast to blood, which flows in a circular pattern at
at length in a following Chapter by R Witte and Bernas [13]) several liters per minute, lymph flows entirely in one direc-
include the “gold standard” whole-body lymphangioscintig- tion and at rest at a rate of only 1.5–2.5 L/24 h. This limited
raphy (LAS) after subcutaneous or intradermal injection of volume derives from a slight hydrodynamic imbalance that
protein-bound radiotracers [14]. LAS provides a dynamic favors movement of fluid, salt, and macromolecules from
image of the peripheral and central lymphatic system non- plasma into tissue spaces. Although blood capillary beds
invasively. Single Photon Emission Computed Tomography vary in hydraulic conductance, in general, disturbances in
(SPECT-CT) greatly improves the sensitivity and spatial the transcapillary hydrostatic and protein osmotic pressure
localization with 3D higher resolution LAS images. Other gradients (Starling forces) tend to promote edema that is low
agents used for indirect lymphography include various flu- in protein content (<1.0 g/dL [10 g/L]), whereas impedance
orescent or magnetic particles, infrared particles and dyes, to lymph flow (lymph stasis) promotes (lymph) edema that is
immunoglobulin conjugates, and microbubbles for detection high in protein content (>1.5-3.5 g/dL [15–35 g/L]), reflected
with fluorescent microscopes, optical imaging systems, com- in parallel alterations in tissue oncotic pressure.
puted tomography (CT), magnetic resonance imaging (MRI; Although lymphatic vessels, like veins, are thin-walled
with and without contrast), and ultrasound including in com- flexible conduits that return liquid to the heart, the flow–
bination with light (photoacoustics). Most important clini- pressure relationships in the venous system and the lym-
cally has been the refinement of percutaneous MR imaging phatic system are different. The energy to drive blood in the
with direct puncture of central lymphatics and also inguinal venous system derives primarily from the thrust of the heart.
node injection of contrast to open up the central lymphatic The cardiac propulsive boost maintains a pressure head
system to clear view and with it, the opportunity to intervene through the arteries and blood capillaries into the veins. In
therapeutically [15, 16]. contrast, lymph vessels in tissues are not directly contiguous
with the blood vasculature, and the chief source of energy
for propulsion of lymph emanates from the intrinsic lym-
19.3  ymph Circulation: Lymph Formation,
L phatic truncal wall contractions (propulsor lymphaticum).
Absorption, and Propulsion (Fig. 19.2) To a lesser extent, extrinsic haphazard forces, such as yawn-
ing, breathing, muscular squeezing, and transmitted arterial
Any protein which leaves these vessels…is lost for the time to the pulsations, contribute to lymph propulsion. A vestige of the
vascular system…it must be collected by lymphatics and restored
to the vascular system by way of the thoracic or right lymphatic
powerful amphibian lymph hearts, mammalian lymphatic
duct. —Physiologist Ernest Starling, 1909 [17]. smooth muscle cells contract in a rhythmically and coordi-
nated fashion. Contractile waves propagate rapidly from one
As a fine adjuster of the tissue microenvironment, the lym- edge of the lymphatic trunk to the other, in the direction of
phatic system is neglected in most treatises on vascular dis- tissue-to-lymph node (afferent lymphatic vessels) or lymph
eases. Yet this seemingly delicate system plays a vital role node-to-venous system (efferent lymphatic vessels). These
in maintaining the liquid, protein, and osmotic equilibrium waves, coupled to a well-developed intraluminal valve sys-
around cells and aids in absorption and distribution of nutri- tem, facilitate transport of lymph. In a sense, the lymphatic
ents, disposal of wastes, and exchange of oxygen and carbon structures function as micropumps that respond to fluid chal-
dioxide in the local milieu intérieur. lenges, and small rises in intraluminal pressure can lead to
Two-thirds of the body is composed of water, and most of dramatic increases in contractile frequency. For example, in
this liquid volume is contained within cells. The remainder popliteal afferent lymphatic vessels, raising the intraluminal
outside cells continuously circulates. In a series of epochal pressure 1 mmHg triples contractile frequency purportedly
experiments conducted more than a century ago, Starling improving movement of lymph. Ordinarily, resistance to
outlined the pivotal factors that regulate partitioning of the flow in the lymphatic vessels is relatively high in comparison
extracellular fluid. In brief, the distribution of fluid between to the low resistance in the venous system, but the pumping
the blood vascular compartment and tissues and the net flux capacity of the lymphatics is able to overcome this imped-
19 Lymphatic System Biology, Pathobiology, and Relation to Cancer Metastasis 191

Normal
Lymph Formation = Lymph Absorption
(LF) (LA)
Pc = Capillary Pressure
Pt = Tissue Pressure
πp = Plasma Oncotic Pressure
πt = Tissue Oncotic Pressure
IFV = Interstitial Fluid Volume
Kf = Capillary Filtration Coefficient Jv
σ = Solution Reflaction Coefficient
Jv = Transcapillary Water Flux Pc IFV
Pt
Kf σ
πp
πt

Blood Tissue
Lymph
Kf[(Pc - Pt)-σ(πp - πt)] = Jv

High Output Failure Edema Low Output Failure (Lymph)Edema

LF↑↑ > LA ↑ Lymph Formation > Lymph Absorption ↓

IFV↑

IFV↑ Jv↑ Jv↓

Pc
Pt
↑Pc Kf σ
πp
← Pt↑ πt
↑Kf σ↓
↓πp Blood Lymph
→πt↑
Blood Tissue
Lymph
Tissue Normal > ↓Jv

Fig. 19.2 Physiologic factors involved in lymph formation and lymph lymphedema, or a mixture of both pathomechanisms. Treatment aims
absorption. Normally (top), these two processes are in balance. Edema to restore the balance between the two processes by either reducing
appears when lymph formation exceeds lymph absorption either lymph formation or enhancing lymph absorption or both. Examples of
because excessive lymph formation overwhelms the lymphatic absorp- high output as well as low output failure of the lymph circulation during
tive capacity—high output failure of the lymph circulation (bottom, the course of cancer spread are common clinical manifestations as illus-
left) or lymph absorption is defective and cannot handle even a normal trated in Fig. 19.4
lymphatic load—low output failure (bottom, right) as in classical

ance by generating intraluminal pressures of 30–50 mmHg in the wall of lymphatic vessels can act as a pacemaker,
and sometimes even equaling or exceeding arterial pressure. and the rhythm is set by the cell that is faster in firing an
This formidable lymphatic ejection force is modulated not action potential, which propagates through the lymphan-
only by filling pressure but also by temperature, sympatho- gion wall [19]. Although muscular contraction and external
mimetic agents, by-products of arachidonic acid metabolism massage clearly accelerate lymph return in the presence of
(thromoxanes, prostaglandins), neurogenic stimuli, circu- edema, under normal conditions, peripheral lymph flow is
lating hormones, and locally released paracrine and auto- regulated primarily by spontaneous contraction of the lym-
crine cytokine secretions and other products of endothelial phatics themselves. In peripheral lymphatics, unlike periph-
metabolism [18, 19]. The pacemaking activity present in eral veins, the column of liquid is incomplete. Accordingly,
lymphatic vessels is responsible for coordinating contrac- with normal intralymphatic pressure, external compression
tions and setting their rhythm, consequently affecting lymph is ineffective in propelling lymph onward, although it may
flow. Lymphatic contractions are a result of action potentials increase the frequency and amplitude of lymphatic contrac-
that originate within the lymphangion, likely due to activ- tions. During lymphatic obstruction and persistent lymph
ity of pressure-sensitive chloride ion channels in lymphatic stasis, the fluid column in the lymphatics becomes continu-
smooth muscle cells. Theoretically, any smooth muscle cell ous, and skeletal muscle or forceful external compression
192 M. H. Witte and S. K. Daley

then becomes an effective pumping mechanism that aids tious agents. Lymphoid organs also include the spleen, Peyer
lymph transport. Assuming an erect position sharply raises patches, thymus, and nasopharyngeal tonsils. The impor-
distal venous pressure, but peripheral intralymphatic pres- tance of immune cell trafficking through the blood–lymph
sure is unaffected, even though lymphatic truncal pulsation loop is now well recognized for health and in a wide variety
increases in both frequency and amplitude, favoring removal of disorders. However, the regulation and intricate mecha-
of tissue fluid. nisms involved, including in cancer progression and defense,
are just beginning to be unraveled and are discussed at length
in the following Section Chapter by Jackson [22].
19.4 Lymphatic Transport Route

The lymphatic route of transport of liquid, protein, and 19.6 Molecular Lymphology
macromolecules, particularly chylomicra originating in the
small intestine, is well recognized albeit underappreciated. Advances in molecular biology and the development of new
The lymphogenous route of cancer spread forms the con- techniques from the Human Genome Project have ushered in
ceptual basis for most current treatment approaches to solid the era of molecular lymphology.
organ cancers. The role of lymphatic uptake and clearance of
microorganisms in the clinical manifestations (e.g., lymphan-
gitic streaking), pathogenesis, inflammatory and other tissue 19.6.1 Lymphangiogenesis
responses, and immune defense has been underemphasized,
particularly in recent years. From the 1930s to the 1960s, a In contrast to uncontrolled new growth (lymphangiosar-
number of carefully performed studies examined the critical coma), programmed proliferation of lymphatic endothelium
lymphatic route in the initial and later phases of viral and with tube formation (lymphangiogenesis) is critical to a host
bacterial infection and the dynamic physiology of the affer- of physiologic and pathologic processes. During the past
ent and efferent arms of the immune response. Lympholo- several decades, since the phenomenon of angiogenesis was
gist Joseph Yoffey laid the groundwork for recognition of first reproduced in endothelial cell and mixed vascular tis-
the importance of the “inert” lymphocyte and the discovery sue cultures, considerable attention has been directed toward
of lymphoid cell trafficking, which are fundamental to the understanding this process, but largely in the context of blood
principles of modern immunology [4]. Recasting these older vessels (‘hemangiogenesis”) [23]. The lymphatic counter-
investigations with more modern physiologic techniques part (i.e., lymphangiogenesis) had received scant attention
and molecular tools may yield new translational insight into until relatively recently, although vigorous (re)generation of
disorders ranging from AIDS and tuberculosis to parasitic lymphatics has been recognized since the early 1900s. In the
disorders and disturbances in the external (skin) or internal 1980s, lymphatic endothelial cells were isolated from lym-
(gastrointestinal, respiratory, and genitourinary tract) micro- phatic tumors, collecting ducts, and since, from a variety of
biomes while also elucidating the lymphatic spread of cancer other tissues and passed in culture with demonstration of
[20]. It is likely that SARS-Cov2 (COVID-­19) will also turn “lymphangiogenesis in vitro” [24–27]. Essential to health,
out to be an important but currently unrecognized lymph- lymphatic growth, when uncontrolled, underlies a variety of
borne infection [21]. uncommon, often disfiguring, and sometimes life- and limb-
threatening conditions [28]. On the other hand, lymphan-
giogenesis may contribute to or protect against many other
19.5  ymphatic Center of Immune
L common disorders. This topic is discussed at length in the
Network: Lymph Nodes, Lymphoid following Section Chapters and particularly in the Chapter
Organs, and Trafficking Cells by Detmar [29].

In addition to their central immunologic role, lymph nodes

are a potential site of impediment to the free flow of lymph. 19.6.2 Lymphvascular Genomics–Proteomics
Unlike frogs, which lack lymph nodes but possess four or
more strategically placed lymph hearts that propel large Lymphangiogenesis is clearly under genetic control and
quantities of peripheral lymph back to the bloodstream, multiple genes are continuing to be found specifically asso-
mammals possess immunoreactive lymph nodes, which, ciated with human lymphedema syndromes and lymphatic
when swollen, fibrotic, or atrophic, may act to initiate or per- malformations on nearly every chromosome and also in
petuate lymph stasis. On the other hand, they also perform chromosomal deletions, translocations, and aneuploidy as
an important “barrier function” first recognized in the 1860s summarized in the Section Chapter by Erickson and Del-
in regard to the spread of tumors, particulates, and infec- linger [30]. Many of these same genes and related proteins
19 Lymphatic System Biology, Pathobiology, and Relation to Cancer Metastasis 193

have been implicated in cancer growth and metastasis, and portal venous pressure) or accelerating lymph drainage or
many more have been identified putatively in mice and other both [40–42]. Events at the thoracic duct–venous junction in
animal species with uncertain application to human disor- the neck such as central venous hypertension or a pinchcock
ders. These include growth factor genes, their products, effect from limited distensibility may be critical in interfer-
and transcription factors, most of which were discovered ing with the free flow of overproduced lymph [42]. In this
in human studies by reverse and forward genetics and more situation, thoracic duct decompression through shunting or
recently by next-generation sequencing (whole genome and stenting into lower pressure veins may be beneficial.
whole exome sequencing) [31–33]. Because congenital absence and radical excision of
regional lymph nodes (and hence, lymphatics) are associated
with edema, it has seemed straightforward that lymphedema
19.6.3 Growth/Inhibitory Factors is simply the direct result of insufficient lymphatic drainage
(“low output failure” of the lymphatic circulation) (Fig. 19.2,
Although the underlying mechanisms governing lymphan- bottom, right). However, persistent peripheral lymphedema
giogenesis are only beginning to be understood, a variety of has proved both hard to reproduce experimentally and chal-
stimulatory [most importantly vascular endothelial growth lenging to treat. Initial experimental attempts to simulate the
factor C (VEGF-C) discovered by Alitalo [34] and its signal- clinical condition by lymphatic sclerosis and radical exci-
ing pathway] and inhibitory factors have been discovered to sion were unsuccessful and revealed a remarkable capacity
play a key role. Other factors are context-dependent, such as of obstructed lymphatics to regenerate and “bridge the gap”
the angiopoietin-1–angiopoietin-2 yin-yang interaction [35], or bypass the induced blockage with spontaneous opening
and these also modulate growth, development, and remodel- of auxiliary lymphatic–venous shunts. It is now nonetheless
ing of the lymphatic vasculature. Function/dysfunction of the clear that nonpitting, brawny extremity edema can arise from
lymphatic vasculature may be influenced in common with or lymph stasis alone and by the unremitting accumulation of
distinct from effects on the blood vasculature [36]. Invaluable protein-rich fluid in the extracellular matrix. Progressive
insights have been gained from transgenic, spontaneous, and truncal tortuosity and dilatation give way to marked lym-
chemical- and/or radiation-induced mutant mouse models. phangiectasia, valvular incompetence, and retrograde flow
These angiomodulatory factors are crucial to understanding (dermal backflow) [26]. Serial microscopy discloses mono-
clinical conditions as diverse as cystic hygroma, Klippel– nuclear cell infiltration, intramural destruction of lymphatic
Trénaunay syndrome, (lymph)angiosarcoma, elephantiasis, collectors, and collagen and adipose deposition through-
and a constellation of peripheral and visceral disorders char- out the soft tissues. Eventually, the lymph trunks lose their
acterized by swelling, scarring, immunodysregulation, and distinctive smooth muscle and endothelial lining, and the
malnutrition. boundary lines between lymph collectors and the surround-
ing matrix progressively blur. Pathophysiology and clinical
aspects of lymphedema are discussed at length in the Section
19.7  athobiology of Lymph Formation–
P Chapter by Szuba, Boccardo, and Piller as well as in refer-
Lymph Absorption: The Formula enced textbooks on lymphology [3, 12, 43].
of Edema (Fig. 19.2) Chylous and non-chylous lymph reflux can also occur
including as a complication of cancer and cancer metasta-
Edema or tissue swelling, a reflection of lymph formation– sis. Abnormal retrograde transport of lymph is termed reflux.
lymph absorption imbalance, is one of the commonest mani- When the lymph derives from outside the intestine (e.g.,
festations of disease, ranging from a cosmetic imperfection from the limbs, where it is generally straw-colored to watery
as in transient foot swelling after a long flight to the immedi- clear), it is termed non-chylous reflux, whereas lymph aris-
ate cause of death when surrounding the heart. The common ing from the intestine (generally milky in color) and flow-
local or generalized edemas (“high-output” lymph flow fail- ing retrograde is termed chylous reflux. Because cholesterol
ures) (Fig. 19.2, bottom, left), where increased lymph forma- and long-chain triglycerides in the form of chylomicra are
tion outpaces enhanced but inadequate lymph absorption, are absorbed exclusively by the lymphatic system, dysfunction
seen after venous occlusion in congestive heart failure, and (in the form of disruption, compression, obstruction, or fis-
in hepatic cirrhosis-associated portal venous hypertension. tulization) of mesenteric lacteals, the cisterna chyli, and the
Here, increased lymph formation results from an imbalance thoracic duct is directly linked to chylothorax, chylous asci-
in the Starling forces (elevated blood capillary pressure) tes, chyluria, chylometrorrhagia, and protein-losing enterop-
along with progressive impediments to increased lymph flow athy. In some patients with high-grade blockage of intestinal
[37–39]. Accordingly, lymphatic overload is approached lymph flow, the peripheral lymphatics gradually dilate, and
therapeutically by restoring lymph balance by either reduc- with progressive valvular incompetence, lactescent lymph
ing lymph formation (e.g., lowering peripheral, central, or refluxes into the soft tissues of the pelvis, scrotum, and lower
194 M. H. Witte and S. K. Daley

extremities (chylous vesicles and chyledema) or even into protein kinase (MEK), has been reported to be effective in
the bronchial tree, manifesting as chyloptysis [44]. controlling and even reversing the condition.
The aforementioned pathologic processes can be imaged Tumor angiogenesis, which generally refers to heman-
and intervention undertaken by physical methods, endovas- giogenesis, has been extensively studied since Folkman’s
cular image-guided techniques, and surgery to decompress original observations and has also become a major therapeu-
or reroute the obstructed or overloaded lymphatic system. tic target. Tumor lymphangiogenesis, a phenomenon whose
To ameliorate edema including lymphedema, either lymph very existence until relatively recently was questioned, has
formation must be reduced or lymph drainage enhanced to now become a “hot topic,” discussed at length in the Section
bring the two processes back into balance. This principle Chapter by Detmar [29].
and these therapeutic approaches apply as well to edema, Novel and potential approaches to modulate these pro-
effusions, and chylous reflux during the course of primary cesses include gene therapy to promote lymphangiogenesis
cancers, cancer metastasis, and cancer-associated treatment when inadequate (as in primary lymphedema) or inhibit
complications. ­lymphatic growth when excessive (as in lymphangiomato-
sis). Stem cell therapy to grow new vessels from undifferen-
tiated embryonic or adult stem cells (e.g., lymphangioblasts)
19.8 Lymphangiotumorigenesis and new biomaterials for artificial lymphatics and scaffold-
and Tumor Lymphatics ing to drain the tissue also hold promise for the future.

A rare but revealing sequela of long-standing peripheral

lymphedema is the occurrence of lymphangiosarcoma and/ 19.9  ymphatic Systemomics and Cancer
L
or angiosarcoma and perhaps also other opportunistic neo- Metastasis (Figs. 19.3 and 19.4)
plasms. This aggressive vascular malignancy was once
thought to arise exclusively after radical mastectomy and When an axillary gland becomes cancerous after disease of the
mamma..during a long period remains diseased without the
irradiation for local control of breast cancer (Stewart–Treves glands next in succession or other organs becoming affected..the
syndrome) (illustrated in Fig. 19.4b, f). However, lymphan- gland collects the hurtful ingredients absorbed from the breast
giosarcoma has now been documented in other secondary and affords protection to the body... but at length proves insuffi-
lymphedemas and even in congenital or primary lymph- cient, perhaps itself becomes a new independent source of fur-
ther propagation of the poisonous matter from the diseased
edema. More recently, Kaposi sarcoma (Fig. 19.4c), a vas- parts of the gland.—Rudolf Virchow, 1868 [46]
cular tumor akin to Stewart–Treves syndrome and allied
closely with AIDS, has been linked to an origin from virally This remarkably prescient passage from Virchow, the origina-
transformed lymphatic endothelium. Perhaps local or gen- tor of the “cellular theory” of cancer displacing the “lymph
eralized immunodeficiency/dysregulation underlies a wide theory” that cancer arises in an altered interstitium [47], elu-
range of common and bizarre vasoproliferative and lympho- cidates the barrier function of lymph nodes. He had dem-
logic syndromes, including hemolymphangioma (presenting onstrated this phenomenon previously by subcutaneously
“angiotumorigenesis”), Klippel–Trénaunay syndrome, Gor- injecting particulates. In addition, the passage also suggests the
ham–Stout syndrome and multiple lymphangiomatosis (rep- sentinel node concept and points to the further lymphogenous
resenting as bony lymphangiomatosis or “disappearing bone route of dissemination of propagated cancer from the lymph
disease”), angiofollicular hyperplasia (epithelioid hemangi- node. We now know that the lymphatic system is intimately
oma), lymphangioleiomyomatosis, Kaposiform hemangio- involved in all phases of cancer, from its development, progres-
endotheliosis, tumor lymphangiogenesis, and lymphangitic sion, and spread to evaluation, prognosis, complications, treat-
metastatic carcinomatosis. Specific genes have been identi- ment, treatment-related sequelae, prevention, containment,
fied in the RAS and MAPK pathway that underlie some of and opportunities for cure [48]. Specifically, at the biological
the lymphatic malformations. Moreover, lymphatic-directed level (Fig. 19.3), blood capillary hyperpermeability promotes
angioinhibitory therapy, such as targeting mitogen-activated excess lymph formation, lymphangiogenesis is stimulated, and
19 Lymphatic System Biology, Pathobiology, and Relation to Cancer Metastasis 195

Lymphatic System and Cancer

Epithelium
Epi(endo)thelial-
mesenchymal
rbc
Blood vessel transition
(EMT MET)

Lymph
Formation Lymphocyte
Parenchyma trafficking
Development
Regeneration Extracellular Matrix
(MET EMT) Tumor
Lymphvasculo/
angiogenesis Fibrosis
ECM

Tumor lymphangiogensis
Superficial lymphatics

Jugular lymph sac Lymph Lymphangiotumorigenesis

Subclavian lymph sac
Lymph node
Absorption- Lymphomagenesis
Deep lymphatics
Thoracic duct
Transport
Retroperitoneal Lymphogenesis
lymph sac
Cisterna chyli dissemination
Posterior lymph sac
Lymph node Mucosa
Superficial lymphatics Lymphatic
External Lymph node
Environment

Blood-Lymph
circulatory loop

systemic organs

Fig. 19.3 Points of interplay between the developing cancer and ongo- the reverse process (MET) in development/regeneration and neoplasia
ing processes within the interstitium and lymphatic system: lymph- (green) are identified. These complex structural–functional interactions
edema, lymphangiogenesis, lymph absorption-transport, lymphogenous participate in the pathogenesis, clinical manifestations, evaluation, and
tumor spread, lymphomagenesis, tumor-generated immune response. prognosis as well as the treatment of cancer. Reproduced with permis-
Potential sites of epi/endothelial-mesenchymal transition (EMT) and sion of Witte, M et al. (2012) Clin Exp Mets 29: 707–712 [45]

lymph drainage may be compromised leading to lymph sta- the main pathway for lymphogenous spread to regional lymph
sis and resultant edema/effusions, fibrosis, tissue overgrowth, nodes (sentinel nodes and beyond). The lymphatic system in
and ­immunodysregulation. Primary lymphatic malignancies its entirety—vessels, lymph, lymph nodes and lymphocytes—
arising in lymphatic endothelial cells (highly aggressive lym- may also function as an immunologic barrier to further spread
phangiosarcomas—Stewart–Treves syndrome and Kaposi sar- or alternatively, as a nidus for continued proliferation and fur-
coma) and in lymphocytes (lymphomas) also occur. Lymphatic ther lymphogenous spread through central lymphatic channels
involvement not only alters the local tumor microenvironment to the systemic bloodstream and beyond. It is this dual and con-
affecting crucial physical forces (solid stress, fluid pressure, trasting role of the lymphatic system that is a target for both
stiffness, and microarchitecture in the interstitium) [48] but is current and future cancer therapy.
196 M. H. Witte and S. K. Daley

a b c d

60
e f g h
50

Ascitic 40
Fluid
Protein
Content 30
(%Plasma)
20

10

11 5
0
Cancer
P<,001

Fig. 19.4 Spectrum of clinical manifestations of lymphatic system showing nest of intralymphatic breast cancer cells in capsule of involved
disorders in patients with cancer metastasis (a–d, top row) and underly- draining lymph node. Reproduced with permission of Nagle et al.
ing pathology/pathophysiology of these clinical manifestations (e–h, (1987) Lymphology 20: 20–24 [50]. (f) Corresponding histopathology
lower row). (a) Massive lymphedema of left arm and shoulder from of lymphangiosarcoma in patient in (b). (g) One minute flow of thoracic
untreated invasive left breast cancer with (inset) corresponding left duct lymph in cannulated control patient (~1 ml per minute) compared
upper extremity lymphangioscintigram displaying complete obstruc- to markedly reduced minute flow of “jaundiced” (high bilirubin con-
tion of lymphatic channels with extensive dermal reflux of radiotracer. tent) lymph in patient with massive ascites (edema fluid in abdominal
(b) Cutaneous spread of aggressive lymphangiosarcoma (Stewart– cavity) secondary to extensive peritoneal lymphatic involvement from
Treves syndrome) with associated histopathology in E, secondary to biliary tract cancer (example of low flow lymphatic failure). (h) illus-
long-standing upper extremity lymphedema following initial breast trates that cancer spread can be associated with a high output failure of
cancer treatment by radical mastectomy. (c) Extensive disseminated the lymph circulation and a resultant low-protein transudative ascites
AIDS-Kaposi sarcoma-associated anasarca (generalized body edema) due to portal vein obstruction producing portal hypertension (elevated
with inset, upper and lower extremity lymphangioscintigrams display- portal venous pressure) secondary to invasive pancreatic cancer (right
ing bizarre dilated lymphatic collector channels along all four extremi- bar) compared to more typical high protein exudative ascites associated
ties, reflecting the origin of this tumor in lymphatic endothelial cells. with low-output failure of the lymph circulation from lymphatic
(d) Severe facial lymphedema acutely developing after radical neck dis- obstruction due to peritoneal carcinomatosis (left bar). Reproduced
section for invasive throat cancer—vital blue dye injected into swollen with permission of Witte, et al. (1972) JAMA 221:1380–1381 [51]. See
tongue gradually disappeared subsequently over several weeks as lym- Fig. 19.2 for pathomechanisms and text for further details
phatic vessels regenerated and edema subsided. (e) Histopathology

19.10 Conclusions central bloodstream to complete the “blood–lymph loop” of

the extracellular fluid circulation.
For lymphatic vessels to function effectively, lymph formed Lymphatic dysfunction is associated with disturbances
from blood capillary filtrate, cell products, and trafficking in transcapillary exchange of liquid, macromolecules, and
cells must be absorbed from the interstitium, carried through cells across intact and abnormal microvessels. Deranged
non-leaky valveless initial lymphatics and pre-collectors, lymphatic growth patterns and lymph kinetics are individu-
and then propelled through patent intrinsically contractile ally and together seen when tissue swelling and its sequelae
valved peripheral collecting channels and interposed lymph occur. Low output failure of the lymph circulation mani-
nodes to reach the cisterna chyli draining visceral lymph in fested as peripheral lymphedema is characteristically indo-
the abdomen. Lymph finally arrives at the left and right tho- lent for many years before lymphatic insufficiency and
racic duct, where it passes through a valved entry into the edema accelerate and become persistent. Chronic lymph-
19 Lymphatic System Biology, Pathobiology, and Relation to Cancer Metastasis 197

edema is characterized by trapping of fluid and extravasated

plasma proteins and other macromolecules in the skin and fluid pressure, tissue stiffness, and microarchitec-
subcutaneous tissues and also in the viscera and body cavi- tural alterations and also to epithelial-mesenchymal
ties, with an increased propensity for superimposed infection transition (EMT)?
and a progressive pro-growth microenvironment. • Could “liquid biopsy” sampling central lymph
On the other hand, still more common is lymphatic over- (noninvasive endovascularly now) be more infor-
load (high output failure of the lymph circulation) as occurs mative about cancer identity and spread than liquid
in hepatic cirrhosis with portal hypertension and in right heart biopsy of peripheral blood?
failure from central venous hypertension, leading to increased • What features explain the fine line between benign
lymph formation from an imbalance of Starling forces along metastasizing tumors (e.g., lymphangiomatosis)
with progressive resistance to increased lymph drainage, pro- and frank malignancies, including malignant pro-
cesses that need to be brought back into balance. gression/transdifferentiation in some of these
Transdifferentiation and transformation of endothelium benign conditions?
and other vascular accessory cells associated with lymph sta-
sis may also be pivotal factors in a wide range of dysplastic
and neoplastic vascular disorders, including Stewart–Treves
syndrome, AIDS-associated Kaposi sarcoma, recurrent References
lymphangiomatosis, and lymphangitic metastatic carcino-
1. Asellius G. De lactibus sive lacteis venis quarto vasorum mesarai-
matosis. These phenomena have their origin in controlled corum genere novo inuento. Mediolani: JB Biddellium; 1627.
and uncontrolled lymphangiogenesis and apparently are reg- 2. Harvey W. De Motu Cordis. Germany: Frankfurt; 1627.
ulated at specific sites across the genome. Their modulation 3. Földi M, Földi E, editors. Földi’s textbook of lymphology. 3rd ed.
may offer new avenues of therapy for both primary and sec- Munich: Urban & Fischer; 2012. 672 pp.
4. Yoffey JM, Courtice FC. Lymphatics, lymph and lymphomyeloid
ondary lymphedema and related angiodysplasia syndromes, complex. London/New York: Academic Press Inc; 1970. 942 pp
as well as a variety of benign and malignant neoplasms. 5. Nedergaard M, Goldman SA. Glymphatic failure as a final common
Finally, while pathobiologic features of lymphatic sys- pathway to dementia. Science. 2020;370(6512):50–6.
tem dysfunction are general ones, they also apply directly to 6. Geng X, Srinivasan S. Lymphatic specification and development,
EMT-MET, and cancer spread. Section 18, Witte MH (Section ed).
understanding lymphogenous spread of cancer, its complica- In: Leong SPL, Nathanson SD, Zager JS, editors. Cancer metasta-
tions as well as approaches to intervene and even prevent sis through the lymphovascular system. Switzerland AG: Springer
cancer metastasis. Nature; 2021. (in press).
7. Evangelou E, Kyzas PA, Trikalinos TA. Comparison of the diag-
nostic accuracy of lymphatic endothelium markers: Bayesian
approach. Mod Pathol. 2005;18(11):1490–7.
Unanswered Questions 8. Johnson LA, Banerji S, Lawrance W, et al. Dendritic cells enter
lymph vessels by hyaluronan-mediated docking to the endothelial
• Do metastases metastasize, through which route, receptor LYVE-1. Nat Immunol. 2017;18(7):762–70.
and under what conditions? 9. Casley-Smith JR, Florey HW. The structure of normal small lym-
• Which genetic or microenvironmental influences phatics. Q J Exp Physiol. 1961;46:101–6.
explain why some cancers stay localized, others 10. Baluk P, Fuxe J, Hashizume H, et al. Functionally specialized junc-
tions between endothelial cells of lymphatic vessels. J Exp Med.
metastasize widely, and still others metastasize to 2007;204(10):2349–62.
only single or selected organs? 11. Leak LV. Electron microscopic observation on lymphatic capillar-
• What functional changes occur in the local and ies and the structural components of the connective tissue-lymph
regional lymphatics during cancer spread and how interface. Microvasc Res. 1970;2(4):361–91.
12. Kinmonth JB. The lymphatic diseases, lymphography, and surgery.
do these contribute to controlling or favoring spread? Baltimore: Williams & Wilkins; 1972.
• Could local lymphatic obstruction or deficiency 13. Witte RS, Bernas MJ. Lymphatic imaging and intervention: appli-
impede cancer spread while promoting primary cations to cancer metastasis. Section 18, Witte MH (Section ed).
growth? In: Leong SPL, Nathanson SD, Zager JS, editors. Cancer metasta-
sis through the lymphovascular system. Switzerland AG: Springer
• What factors, including lymphatic, explain delayed Nature; 2021. (in press).
cancer dormancy, recurrence, and spread? 14. The Diagnosis and Treatment of Peripheral Lymphedema. 2020
• How does lymphatic dysfunction in the tumor area Consensus Document of the International Society of Lymphology.
contribute to physical forces influencing cancer Lymphology. 2020;53(1):3–19.
15. Itkin M, Nadolski GJ. Modern techniques of lymphangiography and
growth, containment, and spread—i.e., solid stress, interventions: current status and future development. Cardiovasc
Intervent Radiol. 2018;41(3):366–76.
198 M. H. Witte and S. K. Daley

16. Schwartz FR, James O, Kuo PH, et al. Lymphatic imaging: cur- 34. Karaman S, Leppänen VM, Alitalo K. Vascular endothelial growth
rent noninvasive and invasive techniques. Semin Intervent Radiol. factor signaling in development and disease. Development.
2020;37(3):1–13. 2019;145((14):dev151019.
17. Starling EH. The fluids of the body. In: The herter lecture. Chicago: 35. Gale NW, Thurston G, Hackett SF, et al. Angiopoietin-2 is required
WT Keener; 1909. p. 81. for postnatal angiogenesis and lymphatic patterning, and only the
18. McHale NG, Roddie IC, Thornbury KD. Nervous modulation latter role is rescued by angiopoietin-1. Dev Cell. 2002;3(3):411–23.
of spontaneous contractions in bovine mesenteric lymphatics. J 36. Dellinger MT, Witte MH. Lymphangiogenesis, lymphatic syste-
Physiol. 1980;309:461–72. momics, and cancer: context, advances and unanswered questions.
19. Scallan JP, Zawieja SD, Castorena-Gonzalez JA, et al. Lymphatic Clin Exp Metastasis. 2018;35(5-6):419–24.
pumping: mechanics, mechanisms and malfunction. J Physiol. 37. Dumont AE, Witte MH. Clinical usefulness of thoracic duct can-
2016;594(20):5749–68. nulation. In: Stollerman GH, editor. Advances in internal medicine,
20. Cole WR, Witte MH, Witte CL. Lymph culture: a new tool for vol. XV. Chicago: Year Book Medical Publ., Inc.; 1969. p. 51–72.
the investigation of human infections. Ann Surg. 1969;170(4): 38. Witte MH, Dumont AE, Cole WR, et al. Lymph circulation in
705–14. hepatic cirrhosis: effect of portacaval shunt. Ann Intern Med.
21. Witte MH, Daley SK. SARS-CoV-2/COVID-19, lymphatic ves- 1969;70(2):303–10.
sels, lymph, and lymphology. [Editorial]. Lymphology. 2020;53(3): 39. Witte MH, Dumont AE, Clauss RH, et al. Lymph circulation in
97–8. congestive heart failure: effect of external thoracic duct drainage.
22. Jackson D. Immune cell trafficking in the lymphatic system, hyal- Circulation. 1969;39(6):723–33.
uronan biology, and cancer metastasis. Section 18, Witte MH 40. James H, Witte MH, Bernas M, et al. Proposal for prevention or
(Section ed). In: Leong SPL, Nathanson SD, Zager JS, editors. alleviation of protein/lymph-losing enteropathy (PLE/LLE) after
Cancer metastasis through the lymphovascular system. Switzerland Fontan circulation treatment of univentricular hearts: restoration of
AG: Springer Nature; 2021. (in press). lymph balance with “lymphatic right-to-left shunt”. Lymphology.
23. Folkman J, Haudenschild C. Angiogenesis in vitro. Nature. 2016;49(3):114–27.
1980;288:551–6. 41. Kreutzer C, Klinger DA, Chiostri B, et al. Lymphatic decompres-
24. Witte MH, Way DL, Witte CL, et al. Lymphangiogenesis: mecha- sion concomitant with Fontan/Kreutzer procedure: early experi-
nisms, significance and clinical implications. In: Goldberg IV, ence. World J Pediatr Congenit Heart Surg. 2020;11(3):284–92.
Rosen EM, editors. Regulation of angiogenesis. Basel: Birkhäuser; 42. Witte M. Thoracic duct decompression: an idea whose time has
1997. p. 65–112. come – again. Lymphology. 2020;53(2):51–4.
25. Bowman CA, Witte MH, Witte CL, et al. Cystic hygroma reconsid- 43. Chachaj A, Piller N, Boccardo F, et al. Lymphedema: general
ered: hamartoma or neoplasm? Primary culture of an endothelial pathophysiology, prevention and management in invasive cancer.
cell line from a massive cervicomediastinal cystic hygroma with Section 18, Witte MH (Section ed). In: Leong SPL, Nathanson SD,
bony lymphangiomatosis. Lymphology. 1984;17(1):15–22. Zager JS, editors. Cancer metastasis through the lymphovascular
26. Witte MH, Witte CL. Lymphangiogenesis and lymphologic syn- system. Switzerland AG: Springer Nature; 2021. (in press).
dromes. Lymphology. 1986;19(1):21–8. 44. Campisi C, Boccardo F, Witte M, et al. Lymphatic surgery and sur-
27. Witte MH, Bernas MJ, Martin CP, et al. Lymphangiogenesis and gery of lymphatic disorders. In: Dieter RS, Dieter Jr RA, Dieter III
lymphangiodysplasias: from molecular to clinical lymphology. RA, editors. Venous and lymphatic diseases, Chapter 42. New York,
Microscop Res Tech. 2001;55(2):122–45. NY: McGraw Hill; 2011. p. 607–29.
28. Oliver G, Kipnis J, Randolph GJ, et al. The lymphatic vasculature 45. Witte MH, Dellinger MT, Papendieck CM, et al. Overlapping
in the 21st century: novel functional roles in homeostasis and dis- biomarkers, pathways, processes and syndromes in lymphatic
ease. Cell. 2020;182(2):270–96. development, growth and neoplasia. Clin Exp Metastasis.
29. Detmar M. Lymphangiogenesis and tumor lymphangiogenesis. 2012;29(7):707–27.
Section 18, Witte MH (Section ed). In: Leong SPL, Nathanson SD, 46. Virchow R (1868) Cellular pathology as based upon physiologi-
Zager JS, editors. Cancer metastasis through the lymphovascular cal and pathological histology. Trans. 2nd ed. of original by Frank
system. Switzerland AG: Springer Nature; 2021. (in press). Chance. NY: RM De Witt.
30. Erickson RP, Dellinger MT. Lymphvascular genomics and pro- 47. Triola VA. Nineteenth century foundations of cancer research.
teomics, clinical syndromes, and cancer metastasis. Section 18, Advances in tumor pathology, nomenclature, and theories of onco-
Witte MH (Section ed). In: Leong SPL, Nathanson SD, Zager JS, genesis. Cancer Res. 1965;25:75–106.
editors. Cancer metastasis through the lymphovascular system. 48. Nia H, Munn L, Jain R. Physical traits of cancer. Science.
Switzerland AG: Springer Nature; 2021. (in press). 2020;370:eaaz0868.
31. Brouillard P, Boon L, Vikkula M. Genetics of lymphatic anomalies. 49. Leak LV, Burke F. Fine structure of the lymphatic capillary and the
J Clin Invest. 2014;124(3):898–904. adjoining connective tissue area. Am J Anat. 1966;118(3):785–809.
32. Michelini S, Paolacci S, Manara E, et al. Genetic tests in lym- 50. Nagle RB, Witte MH, Martinez AP, et al. Factor VIII-associated
phatic vascular malformations and lymphedema. J Med Genet. antigen in human lymphatic endothelium. Lymphology.
2018;55(4):222–32. 1987;20(1):20–4.
33. Gordon K, Varney R, Keeley V, et al. Update and audit of the St 51. Witte MH, Witte CL, Davis WM, et al. Peritoneal transudate:
George's classification algorithm of primary lymphatic anoma- a diagnostic clue to portal system obstruction in patients with
lies: a clinical and molecular approach to diagnosis. J Med Genet. intraabdominal malignancy or peritonitis. J Am Med Assoc.
2020;57(10):653–9. 1972;221(12):1380–3.
 Lymphatic Specification
and Development, EMT-MET, 20
and Cancer Spread

Xin Geng and R. Sathish Srinivasan

Abstract
• Learn how defects in lymphatic vascular develop-
Lymphatic vasculature is an integral part of the circula- ment could lead to diseases such as lymphedema
tory system that regulates interstitial fluid homeostasis, and chylous ascites.
inflammation, and lipid absorption. Lymphatic vessels are • Understand the roles of lymphatic vasculature in
also a major route for tumor metastasis. Lymphatic ves- cancer progression.
sels were previously considered as passive players that • Learn how tumor cells undergo epithelial to mesen-
are manipulated by tumors. However, recent evidences chymal transition (EMT) and mesenchymal to epi-
suggest that the lymphatics play more complex and con- thelial transition (MET) during metastasis.
flicting roles in tumor progression. Lymphatic vessels
could recruit tumor cells and provide a suitable environ-
ment for their survival. Nevertheless, lymphatic vessels
could also modulate immune cell behavior and either 20.1 Introduction
­promote or inhibit immunotherapy. Hence, a better under-
standing of the mechanisms that regulate the growth and Lymphatic vasculature regulates fluid homeostasis, lipid
behavior of lymphatic vasculature might reveal innova- absorption, and inflammation [1, 2]. Lymphatic vascular
tive opportunities for clinical intervention. In this chapter dysfunction leads to lymphedema, metabolic disorders such
we will summarize our current understanding of the as obesity and inflammation. Additionally, lymphatic vessels
mechanisms that regulate lymphatic vascular develop- are a major route for tumor metastasis. Tumors promote lym-
ment. We will also briefly describe the epithelial to mes- phatic vessel growth (lymphangiogenesis) within or around
enchymal transition (EMT) and mesenchymal to epithelial them. Invasive tumor cells enter these nearby lymphatic ves-
transition (MET) that are part of the tumor metastasis sels and migrate to lymph nodes (LNs) and further upstream
process. to the blood circulation. Subsequently, the tumor cells exit
the blood circulation in the lungs, brain, or liver and form
secondary tumors. Evidences suggest that the mechanisms
that regulate the development of lymphatic vessels are also
Learning Objectives necessary for tumor lymphangiogenesis. Hence, a better
• Understand the stepwise development of the lym- understanding of stepwise development of lymphatic vascu-
phatic vasculature. lature could reveal novel strategies to treat cancer.
• Know the critical regulators of lymphatic vascular
development.
20.2  ymphatic Endothelial Cell
L
Specification

X. Geng · R. S. Srinivasan (*) Imaging analysis and lineage tracing experiments have
Cardiovascular Biology Research Program, Oklahoma Medical shown that lymphatic endothelial cells (LECs) originate
Research Foundation, Oklahoma City, OK, USA predominantly from embryonic veins [1, 2]. The homeobox
Department of Cell Biology, University of Oklahoma Health transcription factor PROX1 is the first unique marker that
Sciences Center, Oklahoma City, OK, USA distinguishes LECs from blood vascular endothelial cells
e-mail: [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 199
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_20
200 X. Geng and R. S. Srinivasan

a b c

d e

Fig. 20.1 Stepwise development of lymphatic vasculature. (a) LEC location of LVVs. (d) Lymphatic plexus formation: Lymphatic vessels
specification and differentiation: A subset of embryonic venous endothe- sprout from LS and PLLV to form the lymphatic plexus. The tip cells
lial cells express PROX1 to become specified into LEC progenitors. (yellow) sense and migrate towards the VEGF-C gradient. Tip cells are
These LEC progenitors upregulate the expression of VEGFR3 and podo- devoid of S1PR1 activity, lymph flow, and LSS. The more mature lym-
planin and migrate out of the veins as clusters of differentiated LECs. (b) phatic plexus (green) experience LSS and have S1PR1 activity. LSS
Lymph sac and LVV formation: The LECs that originate from the ante- enhances VEGF-C signaling, which is buffered by S1PR1. Together they
rior portions of the embryos associate with each other to form a pair (left maintain the quiescence of lymphatic vessels in the presence of low lev-
and right) of LS. One of the LS is schematically depicted. Immediately els of VEGF-C. Vessels that have detached from the rest of the vascula-
after its formation LS establishes connection with the anterior cardinal ture will regress due to the absence of LSS. (e) Lymphatic vascular
vein through a pair of LVVs. One LVV is depicted in the figure for sim- maturation: The lymphatic plexus undergoes maturation to become lym-
plicity. LVVs express unique molecules such as GATA2 and FOXC2. (c) phatic capillaries or collecting lymphatic vessels. Capillaries have but-
Primitive lymphatic structures: LS and PLLV are primitive lymphatic ton-like junctions to allow fluid entry. Collecting lymphatic vessels have
structures from which the lymphatic vessels of the anterior portion of the zipper-like junctions to prevent fluid leakage. In addition, collecting
body (skin, head, arms, heart, and lungs) arise. Their overall organiza- lymphatic vessels have LV to regulate the unidirectional flow of fluid
tion on one side of the embryo is shown. The red arrow indicates the and LMCs to regulate contractility

(BECs) (Fig. 20.1a) [3]. Hence, LEC specification is defined veins, common cardinal vein, intersomitic veins) increase
as the activation of PROX1 expression in endothelial cells. substantially until E12.5 [5, 6]. LECs originate from other
PROX1 starts to be expressed in a polarized manner in the embryonic veins such as azygous veins and iliac veins [7].
anterior cardinal veins at Theiler stage 15–16 in mamma- Therefore, PROX1 is likely expressed in these veins as well.
lian embryos. This stage corresponds to embryonic day (E) The expression of PROX1 is directly regulated by the
9.75 in mice or ~ E40 in humans. These PROX1+ cells are transcription factors COUP-TFII and SOX18 [8, 9]. COUP-­
known as LEC progenitors as they give rise to most of the TFII is a nuclear hormone receptor that specifies LEC pro-
lymphatic vasculature at subsequent stages [4]. The num- genitors by initiating PROX1 expression through a highly
ber of PROX1+ LEC progenitors in veins (anterior cardinal conserved 16-nucleotide long regulatory element [9]. How-
20 Lymphatic Specification and Development, EMT-MET, and Cancer Spread 201

ever, COUP-TFII is expressed in all venous endothelial cells. from VEGFR3 does not affect lymph sac formation [16].
Hence, COUP-TFII is thought to synergize with additional These observations suggest that VEGF-C could indirectly
transcription factors to regulate the expression of PROX1. activate VEGFR3. The VEGFR3 co-receptor neuropilin-2 is
SOX18 is one such transcription factor that is specifically one such molecule that could associate with VEGF-C [17].
expressed in a subset of venous endothelial cells that eventu- Integrin-β1 is another molecule that could enhance VEGFR3
ally turn on PROX1 [8]. Importantly, SOX18 is necessary for signaling through a cross talk with mechanical signals [18].
the specification of PROX1 as LEC progenitors are absent in Interstitial fluid pressure increases around the embryonic
mice lacking SOX18. veins prior to the formation of LS. This interstitial fluid
Prox1+/− embryos lacking one allele of Prox1 have pressure activates integrin-β1, which then associates with
approximately half the number of LEC progenitors that are VEGFR3 to promote the downstream signaling cascade and
observed in control littermates [10]. This observation sug- LEC proliferation.
gested that PROX1 regulates itself in LEC progenitors. Additional molecules also regulate PROX1 and VEGF-C
Indeed, the interaction between PROX1 and COUP-TFII signaling. The transcription factor GATA2 is necessary for
is necessary to maintain Prox1 expression in LEC progeni- upregulating the expression of VEGFR3 in migrating LECs
tors [9]. Additionally, an exquisite feedback loop operates [19]. GATA2 could also regulate the expression of PROX1
between PROX1 and vascular endothelial growth factor-C [20]. Consequently, mice lacking GATA2 from all endo-
(VEGF-C) signaling [11, 12]. Expression of vascular endo- thelial cells have hypoplastic LS [19, 21]. Adrenomedullin
thelial growth factor receptor-3 (VEGFR3) is regulated by is a peptide hormone that signals through the G protein-
PROX1. In turn, PROX1 expression is positively regulated coupled receptor (GPCR) calcitonin receptor-like recep-
by VEGF-C signaling. Consequently, LEC progenitors dis- tor (CALCRL) and its receptor activity modifying protein
appear rapidly in Vegfc−/− embryos and are dramatically RAMP2. Adrenomedullin signaling regulates the expression
reduced in Vegfc+/− embryos [12]. of PROX1 in LECs. Consequently, deletion of adrenomedul-
lin, CALCRL or RAMP2 from mice results in hypoplastic
LS [22].
20.3  ymphatic Endothelial Cell Migration
L
from Veins and Lymph Sac Formation
20.4 Development of Lymphovenous
Starting at E10.5 in mouse embryos the LEC progenitors Valves and Establishment of Blood-­
upregulate the expressions of VEGFR3, which is the cognate Lymph Separation
receptor of VEGF-C, and the glycoprotein podoplanin and
migrate out of the veins as strings of interconnected cells At E12.0 the developing LS establish connections with blood
(Fig. 20.1a) [5, 6]. These migrating cells are considered as circulation at the junction of jugular and subclavian veins [10,
fully differentiated LECs. 23]. Lymphovenous valves (LVVs) rapidly develop at these
At E12.0 the LECs migrating from the anterior locations sites of blood/lymph interaction (Fig. 20.1b, c). The primary
of embryos aggregate to form four bilaterally located primi- role of LVVs is to permit lymph return to blood circulation
tive lymphatic structures (Fig. 20.1b, c). Jugular lymph sac while preventing the entry of blood into the lymphatic vas-
(LS) is located close to the anterior cardinal vein and periph- culature. LVV-forming endothelial cells (LVV-ECs) strongly
eral longitudinal lymphatic vessel (PLLV) is located dorsally express the transcription factors PROX1, FOXC2, and
near the dermis [3, 5]. LS and PLLV are connected to each GATA2. Integrin- α9 and gap junction molecule connexin-­37
other through an anterior isthmus. LECs that originate from are other molecules that are expressed in LVV-ECs. PROX1,
posterior veins have not received the attention of recent stud- FOXC2, and Wnt/β-catenin signaling are necessary for the
ies. However, they are known to assemble and form retro- differentiation of LVV-forming endothelial cells (LVV-ECs)
peritoneal lymph sacs and cisterna chyli. [23, 24]. GATA2, VEGF-C signaling, the transcriptional co-
VEGF-C signaling is essential for the migration of LECs factors YAP/TAZ, and the Ras GTPase activating protein
from veins [13]. VEGF-C is produced in an inactive form RASA1 are necessary for the maintenance of LVV-ECs [21,
in the mesenchymal cells and it is proteolytically processed 25, 26]. The receptor tyrosine kinase EPHB4 and the gap
into a fully active form by CCBE1 and ADAMTS3 [14, 15]. junction molecule connexin-37 regulate the morphogenesis
Mice lacking VEGF-C, CCBE1, or ADAMTS3 also do not of LVV-ECs into LVVs [27, 28]. Mice lacking most of the
have LECs, LS, or PLLV. Although VEGF-C could signal abovementioned molecules have dilated LS or edema due to
through both VEGFR3 and VEGFR2, VEGFR3 is the pri- defective lymph drainage caused by the loss of LVVs.
mary mediator of VEGF-C signaling during LEC migration Although LECs originate from veins, lymphatic vessels are
and lymph sac formation. Mice lacking the kinase activity of devoid of blood cells except for leukocytes that migrate in and
VEGFR3 (Vegfr3KD/KD) are completely devoid of lymph sacs out of lymph nodes (LNs) to regulate immune response. Plate-
[16]. Intriguingly, deletion of the VEGF-C binding domain lets are important for generating this blood/lymph separation.
202 X. Geng and R. S. Srinivasan

Platelet-expressed CLEC2 interacts with LEC-expressed 41, 42]. Laminar shear stress (LSS) inhibits Notch signal-
podoplanin to generate transient clots at the sites of LEC ing in LECs to induce proliferation. In addition, LSS induces
migration from veins and LVVs to prevent blood from enter- the expression of VEGF-C from LECs and also enhances the
ing LS without blocking lymphatic drainage [29–31]. sensitivity of LECs to VEGF-C [33, 41]. Statically cultured
In summary, LVVs are the first valves to develop outside LECs were unable to activate MAPK pathway in response
the heart and numerous molecules regulate their develop- to low (for example, 1 ng/ml) concentrations of VEGF-C
ment. LVVs are the gatekeepers of lymphatic vascular func- [33]. In contrast, LECs that were previously exposed to LSS
tioning and integrity. for 24 hrs could activate MAPK pathway in response to the
same low concentration of VEGF-C. Thus, LSS enhances the
response of LECs to VEGF-C.
20.5 Lymphatic Vascular Growth Mechanisms are also in place to specifically antagonize
and Patterning LSS-enhanced VEGF-C signaling. The GPCR sphingosine
1-phosphate receptor 1 (S1PR1) maintains the quiescence
The purpose of vascular patterning is to minimize the total energy of lymphatic vessels by specifically antagonizing LSS-­
that is required to efficiently transport fluid. This goal is achieved enhanced VEGF-C signaling in LECs [33]. We propose that
by minimizing the number of branch points (and therefore the S1PR1 buffers the shear stress-induced VEGF-C signaling
transportation distance) that are necessary to provide nutrients or to keep it within a range that is necessary for the survival of
collect interstitial fluid from the body. This outcome is achieved LECs, but not for growth.
during lymphatic vascular development in steps. Lymphatic vessel maturation: The final step of lymphatic
Lymphatic vascular growth and plexus formation: The vessel patterning is maturation. During this step some ves-
first step of lymphatic vascular patterning is the formation sels will specialize into lymphatic capillaries that absorb
of lymphatic plexus. Lymphatic vessels sprout from LS and interstitial fluid while others become collecting lymphatic
PLLV to populate the tissues and organs. The sprouting vessels that transport lymph (Fig. 20.1e).
lymphatic vessels branch in a stereotypic manner at regular Lymphatic capillaries are thinner, shorter, and more
intervals to form the lymphatic plexus (Fig. 20.1d). VEGF-C prevalent than collecting lymphatic vessels. In addition,
signaling is the most important pro-lymphangiogenic mol- lymphatic capillaries strongly express molecules such as
ecule. Vegfc+/− and Vegfr3+/KD mice are almost completely LYVE1, PROX1, and VEGFR3 and are devoid of continu-
devoid of dermal lymphatic plexus [13, 32]. Dermal lym- ous basement membrane. Furthermore, lymphatic capillaries
phatic vascular hypoplasia is also observed in Nrp2−/− mice lack mural cell coverage and acquire button-like junctions
lacking the VEGFR3 co-receptor neuropilin-2 [17]. Delta that are characterized by discontinuous localization of junc-
like 4 (DLL4) and angiopoietin-2 enhance VEGF-C signal- tional molecules such as VE-Cadherin, PECAM-1, and clau-
ing through unidentified mechanisms [33, 34]. Consistent din-5 [43]. The LEC boundaries overlap to prevent lymph
with these findings Dll4+/− and Angpt2−/− embryos also have leakage. LECs of capillaries are attached to the surround-
hypoplastic lymphatic vessels [33, 35]. ing connective tissues by anchoring filaments. The overlap-
Molecules that regulate VEGF-C signaling such as TGF-­ ping LEC boundaries are stretched open by the action of
β, GATA2, YAP/TAZ, and PKD1 are also necessary for the anchoring filaments to allow the interstitial fluid to percolate
patterning of lymphatic plexus. TGF-β and GATA2 positively between LECs and into the lymphatic capillaries. Relax-
regulate the expression of VEGFR3 [19, 36]. YAP and TAZ are ation of anchoring filaments results in the closure of LEC
able to both positively and negatively regulate the expression junctions. Subsequently, lymph from the capillaries is trans-
of PROX1 in response to VEGF-C signaling [25, 37]. While ported upstream by yet unknown mechanisms.
PKD1 is necessary for the sprouting of LECs in response to Lymph from numerous capillaries drains into bigger
VEGF-C, the mechanisms are not fully defined [38, 39]. collecting lymphatic vessels. Collecting lymphatic ves-
Additional regulators of lymphatic vascular patterning sels express lower levels of LYVE1, PROX1, and VEGFR3
include the transcription factor FOXC2, and β-catenin [24, and have zipper-like LEC junctions that are characterized
40]. Mice lacking these molecules have dilated lymphatic by continuous localization of VE-Cadherin, PECAM1, and
vessels with abnormal number of branches. FOXC2 is a tar- claudin-5 between LECs that prevent lymph leakage [43].
get of Wnt/β-catenin signaling [24]. FOXC2 and ephrin-B2/EPHB4 signaling are necessary for
Lymphatic vascular pruning and quiescence: Lymphatic the maintenance of LEC junctions in the collecting lym-
vessels of the plexus that do not efficiently transport fluid phatic vessels [44–46]. FOXC2 maintains junctional integ-
are removed. Fluid flow generates shear stress on endothelial rity at locations that are exposed to oscillatory shear stress
cells. Therefore, shear stress is likely the signal that dictates (OSS) by preventing cell proliferation. In addition, FOXC2
the decision between pruning and survival. Indeed, we and inhibits MAPK signaling and regulates the cytoskeleton [45,
others have identified shear stress as an enhancer of cell pro- 47]. Ephrin-B2/EPHB4 signaling activates Rac1 and inhibits
liferation and VEGF-C signaling in LECs (Fig. 20.1d) [33, RhoA activities [44].
20 Lymphatic Specification and Development, EMT-MET, and Cancer Spread 203

Collecting lymphatic vessels recruit specialized mural 20.6 Tumor Lymphangiogenesis

cells known as lymphatic muscle cells (LMCs) for sup-
port. Unlike the mural cells that surround the arteries and Tumors secrete VEGF-C directly or indirectly through mac-
veins, LMCs have intrinsic contractile activity to propel rophages to promote the sprouting of preexisting lymphatic
lymph [48]. Ephrin-B2, reelin, and PDGF-B regulate vessels. Lymphatic vessels are observed mostly around the
the recruitment of LMCs to collecting lymphatic vessels periphery of tumors although intratumoral lymphatic vessels
[49–51]. are observed in melanoma and head and neck cancer [55].
Lymphatic valves (LVs) develop within collecting lym- Tumors could also promote lymphangiogenesis in the drain-
phatic vessels to regulate the unidirectional flow of lymph ing lymph nodes and in distant organs, which later become
[52]. LVs and LVVs share a similar molecular profile and sites of secondary tumor formation. Additionally, tumors
require almost an identical set of genes for their develop- could transform the identity of LN LECs by downregulating
ment. Shear stress plays critical roles in the morphogenesis the expression of molecules such as JAM-C and upregulat-
of LVs [52–54]. LVs develop at sites such as branch point of ing ITGB2A [56]. Such changes are thought to create pre-­
vessels or in the middle of long vessels that are exposed to metastatic lymphovascular niche to support metastatic cells
OSS. Importantly, OSS enhances the expression of FOXC2, when they arrive. The heparin-binding chemokine midkine is
GATA2, connexin-37, ephrin-B2, and integrin-α9 in LECs. at least one of the factors that are involved in pre-metastatic
The expression of FOXC2 and GATA2 is dependent on niche formation [57].
PROX1 and OSS-induced secretion of Wnt ligands from Lymphatic vessels play dual roles in regulating tumor
LECs. Furthermore, OSS-induced Ca2+ flow between LECs progression. The lipid-rich environment of lymphatic ves-
through gap junctions is necessary for the nuclear localiza- sels protects melanoma cells from oxidative stress and helps
tion of NFATC1 and FOXC2/NFATC1 complex formation. them proliferate and metastasize [58]. Furthermore, LN
In addition, syndecan-4-induced inhibition of the planar LECs could inactivate tumor-specific T cells and provide an
cell polarity molecule VANGL2 in response to LSS is nec- immune-compromised environment for the tumors to grow
essary for the morphogenesis of LVs. Thus, complex cross [59]. However, lymphatic vessels could also reduce intratu-
talk between mechanical and biochemical signals regulates moral fluid pressure and permit chemotherapeutic agents to
LV formation within collecting lymphatic vessels. enter the tumor microenvironment and reduce tumor burden
In summary, we described the stepwise development of [60]. Additionally, lymphatic vessels drain tumor antigens to
lymphatic vasculature. In Table 20.1 we have listed some of lymph nodes where anti-tumor inflammatory response is acti-
the genes that are frequently mutated in human patients with vated. Hence, an increase in lymphatic vessel density is asso-
congenital lymphedema or lymphatic disorders. ciated with better prognosis of glioblastoma and colorectal
cancer in mouse models or human patients, respectively [61,
62]. These discrepancies could be due to tissue- or tumor-
Table 20.1 Genes commonly associated with lymphatic disorders and specific differences in the lymphatic vascular function.
their mechanisms of action Tumor cells undergo two-transition states, epithelial to
Dev steps in mesenchymal transition (EMT) and mesenchymal to epi-
Gene Disease mice thelial transition (MET), in order to metastasize to distant
SOX18 Hypotrichosis-lymphedema-­ 1 organs [63]. Primary tumor cells are attached to each other
telengiectasia
through cell junction molecules such as cadherins and clau-
VEGFR3 Milroy disease 1, 2, 3, 4, 5
dins. Tumor cells are also attached to the basement mem-
VEGFC Milroy-like disease 1, 2, 3, 4, 5
ADAMTS3 Hennekam syndrome 2, 3 brane through integrins. These cell–cell and cell–ECM
CCBE1 Hennekam syndrome 2, 3 interactions provide apical-basal polarity and reduce their
FAT4 Hennekam syndrome 3, 4, 5 migration ability. Frequently (but not always) malignant
FOXC2 Lymphedema-distichiasis syndrome 3, 4, 5 cells need to partially or completely downregulate cell
GATA2 Emberger syndrome 2, 3, 4, 5 junction molecules such as E-cadherin in order to become
RASA1 Capillary malformation; arteriovenous 3, 4, 5 metastatic. In addition, malignant cells start expressing mol-
malformation
PIEZO1 Generalized lymphatic dysplasia of 5
ecules such as FSP1 and vimentin, which are characteristic
Fotiou of migratory mesenchymal cells. EMT is the process through
EPHB4 Lymphatic-related hydrops fetalis 5 which tumor cells lose their cell junctions and polarity and
ITGA9 Fetal chylothorax 5 acquire mesenchymal characteristics. Signaling molecules
CALCRL Nonimmune hydrops fetalis 2 such as TGF-β1, transcription factors such as Snail, Slug,
Step 1: LEC specification ZEB1, ZEB2, and FOXC2 are necessary for tumor cells to
Step 2: LEC differentiation and migration
successfully undergo EMT (Fig. 20.2a). Detachment of cells
Step 3: Lymphatic vascular patterning and/or integrity
Step 4: Lymphatic vascular maturation from the basement membrane could trigger a form of pro-
Step 5: LV and LVV morphogenesis grammed cell death known as anoikis. Therefore, tumor cells
204 X. Geng and R. S. Srinivasan

Fig. 20.2 Mechanisms of tumor metastasis through the lymphatic vas- could exit from LNs via HEVs or continue upstream and enter the blood
culature. (a) EMT: Primary tumors undergo EMT to become invasive. circulation through LVVs. (d) Tumor extravasation: Tumor cells exit
The molecular events associated with EMT are presented. (b) Tumor the blood circulation through capillaries at pre-metastatic niches. (e)
intravasation: Tumors secrete VEGF-C or VEGF-D to attract the lym- Secondary tumor formation: Tumor cells undergo MET, which is the
phatic vessels. LECs secrete factors such as CCL21 that attract tumor reversal of EMT, to gain cell junctions and become secondary tumors
cells that express CCR7. (c) Migration of tumor cells: Tumor cells
20 Lymphatic Specification and Development, EMT-MET, and Cancer Spread 205

should develop additional adaptations to survive the subse- Acknowledgments This work is supported by NIH/NHLBI
(R01HL131652 and R01HL133216 to RSS), NIH/NIGMS COBRE
quent stages of metastasis. (P20 GM103441 to XG; PI: Dr. Rodger McEver), and Oklahoma
Lymphatic capillaries with their button-like junctions are Center for Adult Stem Cell Research (4340 to RSS). We apologize to
conducive for the intravasation (entry) of metastatic tumor the authors of numerous articles that we are unable to cite due to space
cells (Fig. 20.2b). LEC-secreted CCL21 could play an inad- restrictions.
Disclosures: None.
vertent role in recruiting and maintaining CCR7+ tumor stem
cells in the lymphatic vessels.
Tumor cells exit the lymphatic vasculature and enter
the blood circulation through the high endothelial venules References
(HEVs) of LNs (Fig. 20.2c) [64, 65]. Tumor cells could also
circumnavigate the entire lymphatic vasculature and enter 1. Aspelund A, Robciuc MR, Karaman S, Makinen T, Alitalo
K. Lymphatic system in cardiovascular medicine. Circ Res.
the bloodstream through LVVs. Finally, tumor cells extrava- 2016;118:515–30.
sate from the vasculature at pre-metastatic niche, downregu- 2. Oliver G, Kipnis J, Randolph GJ, Harvey NL. The lymphatic vas-
late mesenchymal genes, and regain epithelial cell markers culature in the 21(st) century: Novel functional roles in homeostasis
through MET (Fig. 20.2d, e). MET has been found to be and disease. Cell. 2020;182:270–96.
3. Wigle JT, Oliver G. Prox1 function is required for the development
necessary for the survival of some tumors and the proper of the murine lymphatic system. Cell. 1999;98:769–78.
localization of some others. Nevertheless, the mechanisms 4. Srinivasan RS, Dillard ME, Lagutin OV, Lin FJ, Tsai S, Tsai MJ,
of MET remain to be fully defined. Samokhvalov IM, Oliver G. Lineage tracing demonstrates the
venous origin of the mammalian lymphatic vasculature. Genes Dev.
2007;21:2422–32.
5. Hagerling R, Pollmann C, Andreas M, Schmidt C, Nurmi H, Adams
Summary and Unanswered Questions RH, Alitalo K, Andresen V, Schulte-Merker S, Kiefer F. A novel
• LECs originate primarily from embryonic veins as multistep mechanism for initial lymphangiogenesis in mouse
a homogeneous population of cells. embryos based on ultramicroscopy. EMBO J. 2013;32:629–44.
6. Yang Y, Garcia-Verdugo JM, Soriano-Navarro M, Srinivasan RS,
• LECs aggregate to form LS, which sprout to form Scallan JP, Singh MK, Epstein JA, Oliver G. Lymphatic endothelial
the lymphatic plexus. progenitors bud from the cardinal vein and intersomitic vessels in
• Lymphatic plexus undergoes maturation to form mammalian embryos. Blood. 2012;120:2340–8.
lymphatic capillaries, collecting lymphatic vessels 7. van der Putte SC. The early development of the lymphatic sys-
tem in mouse embryos. Acta Morphol Neerl Scand. 1975;13:
and lymphatic valves. 245–86.
• Lymph is returned to blood through LVVs. 8. Francois M, Caprini A, Hosking B, Orsenigo F, Wilhelm D, Browne
• Tumor cells undergo EMT and enter the lymphatic C, Paavonen K, Karnezis T, Shayan R, Downes M, Davidson T,
capillaries. Tutt D, Cheah KS, Stacker SA, Muscat GE, Achen MG, Dejana E,
Koopman P. Sox18 induces development of the lymphatic vascula-
• Tumor cells exit the lymphatic vasculature and ture in mice. Nature. 2008;456:643–7.
enter blood circulation either from LNs or through 9. Srinivasan RS, Geng X, Yang Y, Wang Y, Mukatira S, Studer
LVVs. M, Porto MP, Lagutin O, Oliver G. The nuclear hormone recep-
• Lymphatic vasculature plays a dual role in tumor tor coup-tfii is required for the initiation and early maintenance
of prox1 expression in lymphatic endothelial cells. Genes Dev.
progression. 2010;24:696–707.
• The signals that regulate LEC specification remain 10. Srinivasan RS, Oliver G. Prox1 dosage controls the number of lym-
to be identified. phatic endothelial cell progenitors and the formation of the lympho-
• The signals that coordinate with VEGF-C to regu- venous valves. Genes Dev. 2011;25:2187–97.
11. Koltowska K, Lagendijk AK, Pichol-Thievend C, Fischer JC,
late lymphatic plexus formation remain to be thor- Francois M, Ober EA, Yap AS, Hogan BM. Vegfc regulates bipo-
oughly identified. tential precursor division and prox1 expression to promote lym-
• The mechanisms of lymphatic vessel maturation phatic identity in zebrafish. Cell Rep. 2015;13:1828–41.
are not fully understood. 12. Srinivasan RS, Escobedo N, Yang Y, Interiano A, Dillard ME,
Finkelstein D, Mukatira S, Gil HJ, Nurmi H, Alitalo K, Oliver
• The role of shear stress remains to be dissected. G. The prox1-vegfr3 feedback loop maintains the identity and
• Whether there are any tissue-specific differences in the number of lymphatic endothelial cell progenitors. Genes Dev.
the identity of LECs remains to be determined. 2014;28:2175–87.
• Whether the cross talk between tumors and lym- 13. Karkkainen MJ, Haiko P, Sainio K, Partanen J, Taipale J, Petrova
TV, Jeltsch M, Jackson DG, Talikka M, Rauvala H, Betsholtz C,
phatic vessels depends on tissue-specific differ- Alitalo K. Vascular endothelial growth factor c is required for
ences in LECs needs to be investigated. sprouting of the first lymphatic vessels from embryonic veins. Nat
Immunol. 2004;5:74–80.
206 X. Geng and R. S. Srinivasan

14. Bos FL, Caunt M, Peterson-Maduro J, Planas-Paz L, Kowalski J, inactivating mutations cause autosomal dominant lymphatic-related
Karpanen T, van Impel A, Tong R, Ernst JA, Korving J, van Es hydrops fetalis. J Clin Invest. 2016;126:3080–8.
JH, Lammert E, Duckers HJ, Schulte-Merker S. Ccbe1 is essen- 28. Munger SJ, Geng X, Srinivasan RS, Witte MH, Paul DL, Simon
tial for mammalian lymphatic vascular development and enhances AM. Segregated foxc2, nfatc1 and connexin expression at normal
the lymphangiogenic effect of vascular endothelial growth factor-c developing venous valves, and connexin-specific differences in the
in vivo. Circ Res. 2011;109:486–91. valve phenotypes of cx37, cx43, and cx47 knockout mice. Dev
15. Bui HM, Enis D, Robciuc MR, Nurmi HJ, Cohen J, Chen M, Biol. 2016;412:173–90.
Yang Y, Dhillon V, Johnson K, Zhang H, Kirkpatrick R, Traxler 29. Crosswhite PL, Podsiadlowska JJ, Curtis CD, Gao S, Xia L,
E, Anisimov A, Alitalo K, Kahn ML. Proteolytic activation defines Srinivasan RS, Griffin CT. Chd4-regulated plasmin activation
distinct lymphangiogenic mechanisms for vegfc and vegfd. J Clin impacts lymphovenous hemostasis and hepatic vascular integrity. J
Invest. 2016;126:2167–80. Clin Invest. 2016;126:2254–66.
16. Zhang L, Zhou F, Han W, Shen B, Luo J, Shibuya M, He Y. Vegfr-3 30. Hess PR, Rawnsley DR, Jakus Z, Yang Y, Sweet DT, Fu J,
ligand-binding and kinase activity are required for lymphangiogen- Herzog B, Lu M, Nieswandt B, Oliver G, Makinen T, Xia L,
esis but not for angiogenesis. Cell Res. 2010;20:1319–31. Kahn ML. Platelets mediate lymphovenous hemostasis to main-
17. Yuan L, Moyon D, Pardanaud L, Breant C, Karkkainen MJ, Alitalo tain blood-lymphatic separation throughout life. J Clin Invest.
K, Eichmann A. Abnormal lymphatic vessel development in neuro- 2014;124:273–84.
pilin 2 mutant mice. Development. 2002;129:4797–806. 31. Uhrin P, Zaujec J, Breuss JM, Olcaydu D, Chrenek P, Stockinger H,
18. Planas-Paz L, Strilic B, Goedecke A, Breier G, Fassler R, Lammert Fuertbauer E, Moser M, Haiko P, Fassler R, Alitalo K, Binder BR,
E. Mechanoinduction of lymph vessel expansion. EMBO J. Kerjaschki D. Novel function for blood platelets and podoplanin
2012;31:788–804. in developmental separation of blood and lymphatic circulation.
19. Frye M, Taddei A, Dierkes C, Martinez-Corral I, Fielden M, Blood. 2010;115:3997–4005.
Ortsater H, Kazenwadel J, Calado DP, Ostergaard P, Salminen M, 32. Karkkainen MJ, Saaristo A, Jussila L, Karila KA, Lawrence EC,
He L, Harvey NL, Kiefer F, Makinen T. Matrix stiffness controls Pajusola K, Bueler H, Eichmann A, Kauppinen R, Kettunen MI,
lymphatic vessel formation through regulation of a gata2-­dependent Yla-Herttuala S, Finegold DN, Ferrell RE, Alitalo K. A model for
transcriptional program. Nat Commun. 2018;9:1511. gene therapy of human hereditary lymphedema. Proc Natl Acad Sci
20. Kazenwadel J, Betterman KL, Chong CE, Stokes PH, Lee YK, U S A. 2001;98:12677–82.
Secker GA, Agalarov Y, Demir CS, Lawrence DM, Sutton DL, 33. Geng X, Yanagida K, Akwii RG, Choi D, Chen L, Ho Y, Cha B,
Tabruyn SP, Miura N, Salminen M, Petrova TV, Matthews JM, Mahamud MR, Berman de Ruiz K, Ichise H, Chen H, Wythe J,
Hahn CN, Scott HS, Harvey NL. Gata2 is required for lym- Mikelis CM, Hla T, Srinivasan RS. S1pr1 regulates the quiescence
phatic vessel valve development and maintenance. J Clin Invest. of lymphatic vessels by inhibiting laminar shear stress-dependent
2015;125:2979–94. vegf-c signaling. JCI. Insight. 2020;5:e137652.
21. Mahamud MR, Geng X, Ho YC, Cha B, Kim Y, Ma J, Chen L, 34. Zheng W, Nurmi H, Appak S, Sabine A, Bovay E, Korhonen EA,
Myers G, Camper S, Mustacich D, Witte M, Choi D, Hong YK, Orsenigo F, Lohela M, D'Amico G, Holopainen T, Leow CC,
Chen H, Varshney G, Engel JD, Wang S, Kim TH, Lim KC, Dejana E, Petrova TV, Augustin HG, Alitalo K. Angiopoietin 2
Srinivasan RS. Gata2 controls lymphatic endothelial cell junctional regulates the transformation and integrity of lymphatic endothelial
integrity and lymphovenous valve morphogenesis through mir-126. cell junctions. Genes Dev. 2014;28:1592–603.
Development. 2019;146 35. Dellinger M, Hunter R, Bernas M, Gale N, Yancopoulos G,
22. Fritz-Six KL, Dunworth WP, Li M, Caron KM. Adrenomedullin Erickson R, Witte M. Defective remodeling and maturation of the
signaling is necessary for murine lymphatic vascular development. lymphatic vasculature in angiopoietin-2 deficient mice. Dev Biol.
J Clin Invest. 2008;118:40–50. 2008;319:309–20.
23. Geng X, Cha B, Mahamud MR, Lim KC, Silasi-Mansat R, Uddin 36. James JM, Nalbandian A, Mukouyama YS. Tgfbeta signaling is
MK, Miura N, Xia L, Simon AM, Engel JD, Chen H, Lupu F, required for sprouting lymphangiogenesis during lymphatic net-
Srinivasan RS. Multiple mouse models of primary lymphedema work development in the skin. Development. 2013;140:3903–14.
exhibit distinct defects in lymphovenous valve development. Dev 37. Cho H, Kim J, Ahn JH, Hong YK, Makinen T, Lim DS, Koh
Biol. 2016;409:218–33. GY. Yap and taz negatively regulate prox1 during developmental
24. Cha B, Geng X, Mahamud MR, Fu J, Mukherjee A, Kim Y, Jho and pathologic lymphangiogenesis. Circ Res. 2019;124:225–42.
EH, Kim TH, Kahn ML, Xia L, Dixon JB, Chen H, Srinivasan 38. Coxam B, Sabine A, Bower NI, Smith KA, Pichol-Thievend C,
RS. Mechanotransduction activates canonical wnt/beta-catenin Skoczylas R, Astin JW, Frampton E, Jaquet M, Crosier PS, Parton
signaling to promote lymphatic vascular patterning and the devel- RG, Harvey NL, Petrova TV, Schulte-Merker S, Francois M, Hogan
opment of lymphatic and lymphovenous valves. Genes Dev. BM. Pkd1 regulates lymphatic vascular morphogenesis during
2016;30:1454–69. development. Cell Rep. 2014;7:623–33.
25. Cha B, Ho YC, Geng X, Mahamud MR, Chen L, Kim Y, Choi D, 39. Outeda P, Huso DL, Fisher SA, Halushka MK, Kim H, Qian F,
Kim TH, Randolph GJ, Cao X, Chen H, Srinivasan RS. Yap and taz Germino GG, Watnick T. Polycystin signaling is required for
maintain prox1 expression in the developing lymphatic and lym- directed endothelial cell migration and lymphatic development.
phovenous valves in response to vegf-c signaling. Development. Cell Rep. 2014;7:634–44.
2020; https://doi.org/10.1242/dev.195453. 40. Norrmen C, Ivanov KI, Cheng J, Zangger N, Delorenzi M, Jaquet
26. Chen D, Geng X, Lapinski PE, Davis MJ, Srinivasan RS, King M, Miura N, Puolakkainen P, Horsley V, Hu J, Augustin HG, Yla-­
PD. Rasa1-driven cellular export of collagen iv is required for Herttuala S, Alitalo K, Petrova TV. Foxc2 controls formation and
the development of lymphovenous and venous valves in mice. maturation of lymphatic collecting vessels through cooperation
Development. 2020;147 with nfatc1. J Cell Biol. 2009;185:439–57.
27. Martin-Almedina S, Martinez-Corral I, Holdhus R, Vicente A, 41. Choi D, Park E, Jung E, Seong YJ, Hong M, Lee S, Burford
Fotiou E, Lin S, Petersen K, Simpson MA, Hoischen A, Gilissen C, J, Gyarmati G, Peti-Peterdi J, Srikanth S, Gwack Y, Koh CJ,
Jeffery H, Atton G, Karapouliou C, Brice G, Gordon K, Wiseman Boriushkin E, Hamik A, Wong AK, Hong YK. Orai1 activates pro-
JW, Wedin M, Rockson SG, Jeffery S, Mortimer PS, Snyder MP, liferation of lymphatic endothelial cells in response to laminar flow
Berland S, Mansour S, Makinen T, Ostergaard P. Ephb4 kinase-­ through kruppel-like factors 2 and 4. Circ Res. 2017;120:1426–39.
20 Lymphatic Specification and Development, EMT-MET, and Cancer Spread 207

42. Choi D, Park E, Jung E, Seong YJ, Yoo J, Lee E, Hong M, Lee S, 54. Sabine A, Saygili Demir C, Petrova TV. Endothelial cell responses
Ishida H, Burford J, Peti-Peterdi J, Adams RH, Srikanth S, Gwack to biomechanical forces in lymphatic vessels. Antioxid Redox
Y, Chen CS, Vogel HJ, Koh CJ, Wong AK, Hong YK. Laminar flow Signal. 2016;25:451–65.
downregulates notch activity to promote lymphatic sprouting. J 55. Karaman S, Detmar M. Mechanisms of lymphatic metastasis. J
Clin Invest. 2017;127:1225–40. Clin Invest. 2014;124:922–8.
43. Baluk P, Fuxe J, Hashizume H, Romano T, Lashnits E, Butz S, 56. Commerford CD, Dieterich LC, He Y, Hell T, Montoya-Zegarra JA,
Vestweber D, Corada M, Molendini C, Dejana E, McDonald Noerrelykke SF, Russo E, Rocken M, Detmar M. Mechanisms of
DM. Functionally specialized junctions between endothelial cells tumor-induced lymphovascular niche formation in draining lymph
of lymphatic vessels. J Exp Med. 2007;204:2349–62. nodes. Cell Rep. 2018;25:3554–63. e4
44. Frye M, Stritt S, Ortsater H, Hernandez Vasquez M, Kaakinen 57. Olmeda D, Cerezo-Wallis D, Riveiro-Falkenbach E, Pennacchi PC,
M, Vicente A, Wiseman J, Eklund L, Martinez-Torrecuadrada JL, Contreras-Alcalde M, Ibarz N, Cifdaloz M, Catena X, Calvo TG,
Vestweber D, Makinen T. Ephrinb2-ephb4 signalling provides rho-­ Canon E, Alonso-Curbelo D, Suarez J, Osterloh L, Grana O, Mulero
mediated homeostatic control of lymphatic endothelial cell junction F, Megias D, Canamero M, Martinez-Torrecuadrada JL, Mondal C,
integrity. elife. 2020;9 Di Martino J, Lora D, Martinez-Corral I, Bravo-Cordero JJ, Munoz
45. Norden PR, Sabine A, Wang Y, Demir CS, Liu T, Petrova TV, J, Puig S, Ortiz-Romero P, Rodriguez-Peralto JL, Ortega S, Soengas
Kume T. Shear stimulation of foxc1 and foxc2 differentially regu- MS. Whole-body imaging of lymphovascular niches identifies pre-­
lates cytoskeletal activity during lymphatic valve maturation. elife. metastatic roles of midkine. Nature. 2017;546:676–80.
2020;9 58. Ubellacker JM, Tasdogan A, Ramesh V, Shen B, Mitchell EC,
46. Sabine A, Bovay E, Demir CS, Kimura W, Jaquet M, Agalarov Martin-Sandoval MS, Gu Z, McCormick ML, Durham AB, Spitz
Y, Zangger N, Scallan JP, Graber W, Gulpinar E, Kwak BR, DR, Zhao Z, Mathews TP, Morrison SJ. Lymph protects metastasiz-
Makinen T, Martinez-Corral I, Ortega S, Delorenzi M, Kiefer F, ing melanoma cells from ferroptosis. Nature. 2020;585:113–8.
Davis MJ, Djonov V, Miura N, Petrova TV. Foxc2 and fluid shear 59. Lund AW, Wagner M, Fankhauser M, Steinskog ES, Broggi MA,
stress stabilize postnatal lymphatic vasculature. J Clin Invest. Spranger S, Gajewski TF, Alitalo K, Eikesdal HP, Wiig H, Swartz
2015;125:3861–77. MA. Lymphatic vessels regulate immune microenvironments in
47. Fatima A, Wang Y, Uchida Y, Norden P, Liu T, Culver A, Dietz WH, human and murine melanoma. J Clin Invest. 2016;126:3389–402.
Culver F, Millay M, Mukouyama YS, Kume T. Foxc1 and foxc2 60. Padera TP, Kadambi A, di Tomaso E, Carreira CM, Brown EB,
deletion causes abnormal lymphangiogenesis and correlates with Boucher Y, Choi NC, Mathisen D, Wain J, Mark EJ, Munn LL, Jain
erk hyperactivation. J Clin Invest. 2016;126:2437–51. RK. Lymphatic metastasis in the absence of functional intratumor
48. Muthuchamy M, Zawieja D. Molecular regulation of lymphatic lymphatics. Science. 2002;296:1883–6.
contractility. Ann N Y Acad Sci. 2008;1131:89–99. 61. Mlecnik B, Bindea G, Kirilovsky A, Angell HK, Obenauf AC,
49. Lutter S, Xie S, Tatin F, Makinen T. Smooth muscle-endothelial cell Tosolini M, Church SE, Maby P, Vasaturo A, Angelova M,
communication activates reelin signaling and regulates lymphatic Fredriksen T, Mauger S, Waldner M, Berger A, Speicher MR, Pages
vessel formation. J Cell Biol. 2012;197:837–49. F, Valge-Archer V, Galon J. The tumor microenvironment and
50. Makinen T, Adams RH, Bailey J, Lu Q, Ziemiecki A, Alitalo immunoscore are critical determinants of dissemination to distant
K, Klein R, Wilkinson GA. Pdz interaction site in ephrinb2 is metastasis. Sci Transl Med. 2016;8:327ra26.
required for the remodeling of lymphatic vasculature. Genes Dev. 62. Song E, Mao T, Dong H, Boisserand LSB, Antila S, Bosenberg
2005;19:397–410. M, Alitalo K, Thomas JL, Iwasaki A. Vegf-c-driven lymphatic
51. Wang Y, Jin Y, Mae MA, Zhang Y, Ortsater H, Betsholtz C, Makinen drainage enables immunosurveillance of brain tumours. Nature.
T, Jakobsson L. Smooth muscle cell recruitment to lymphatic 2020;577:689–94.
vessels requires PDGFB and impacts vessel size but not identity. 63. Bakir B, Chiarella AM, Pitarresi JR, Rustgi AK. Emt, met, plastic-
Development. 2017;144:3590–601. ity, and tumor metastasis. Trends Cell Biol. 2020;30:764–76.
52. Geng X, Cha B, Mahamud MR, Srinivasan RS. Intraluminal 64. Pereira ER, Kedrin D, Seano G, Gautier O, Meijer EFJ, Jones D,
valves: development, function and disease. Dis Models Mech. Chin SM, Kitahara S, Bouta EM, Chang J, Beech E, Jeong HS,
2017;10:1273–87. Carroll MC, Taghian AG, Padera TP. Lymph node metastases can
53. Sabine A, Agalarov Y, Maby-El Hajjami H, Jaquet M, Hagerling invade local blood vessels, exit the node, and colonize distant
R, Pollmann C, Bebber D, Pfenniger A, Miura N, Dormond O, organs in mice. Science. 2018;359:1403–7.
Calmes JM, Adams RH, Makinen T, Kiefer F, Kwak BR, Petrova 65. Brown M, Assen FP, Leithner A, Abe J, Schachner H, Asfour G,
TV. Mechanotransduction, prox1, and foxc2 cooperate to control Bago-Horvath Z, Stein JV, Uhrin P, Sixt M, Kerjaschki D. Lymph
connexin37 and calcineurin during lymphatic-valve formation. Dev node blood vessels provide exit routes for metastatic tumor cell dis-
Cell. 2012;22:430–45. semination in mice. Science. 2018;359:1408–11.
 Lymphangiogenesis: Lymphatic System
and Lymph Nodes; Cancer 21
Lymphangiogenesis and Metastasis

Stanley P. Leong and Marlys H. Witte

Abstract then, to distant sites. Occasionally, cancer cells may

escape into the blood vessels for distant metastasis. In
The lymphatic system is considered the major conduit
general, the SLN may be considered as the major gate-
for cancer metastasis. Two key recent developments have
way for cancer metastasis. Micrometastasis in the SLN is
made cancer lymphangiogenesis and cancer metastasis
an excellent biomarker for a poor clinical outcome.
more understandable on a molecular and translational
Based on experimental and clinical evidence, the SLN
level. These include the discovery of lymphatic markers
may be preconditioned by factors produced by cancer
such as VEGF-C, LYVE-1, podoplanin, and Prox-1 and
cells in the primary site to form a pre-metastatic niche
the recognition of the clinical significance of sentinel
that facilitates the arrival of cancer cells to take residence
lymph nodes (SLNs) in melanoma and breast cancer. The
in the SLN. Molecular mechanisms of how cancer cells
lymphatic system develops during the fifth week of
dissociate from the primary site, follow the lymphatic
embryological life based on Sabin’s centrifugal theory
vessels to reach the SLN, and then, spread beyond to the
and on a minor extent relating to contrasting centripetal
distant sites remain under investigation.
theory. The extracellular fluid that escapes at the blood
capillaries is absorbed through the lymphatic vessels and
passes through multiple levels of lymph nodes. The
lymph from most of the body subsequently drains in a
large part into the thoracic duct and then to the left sub- Learning Objectives
clavian vein. On the upper right side of the body, drain- • Describe steps in the development of and growth of
age is largely into the right subclavian vein. Thus, the lymphatic vessels (lymphangiogenesis) and lymph
excess extracellular fluid transported as lymph through nodes.
the lymphatic vessels and filtered through the lymph • Profile the various cell populations in lymph nodes,
nodes becomes sterile from bacteria or other microorgan- their location, and function.
isms prior to reentry into the circulatory system. Although • Define a SLN.
cancer angiogenesis involving blood vessels (hemangio- • Identify the locations relative to the primary tumor
genesis) occurs in the tumor microenvironment, its prog- and metastatic site where tumor lymphangiogenesis
nostic value is unclear. On the other hand, cancer has been documented.
lymphangiogenesis has been extensively studied and it is • Describe the postulated relationships between
associated significantly with poor prognosis. Cancer tumor lymphangiogenesis and cancer metastasis.
cells in the primary tumor site have the propensity to
migrate to the SLN of the tumor draining nodal basin,

21.1 Introduction
S. P. Leong (*)
Department of Surgery and Melanoma Center, California Pacific
Medical Center and Research Institute, San Francisco, CA, USA Since the lymphatic system is the major conduit for cancer
e-mail: [email protected] metastasis, it is important to understand the development of
M. H. Witte the lymphatic system including the lymph nodes and associ-
Department of Surgery, Neurosurgery and Pediatrics, University of ated cell populations. The relationship between the lym-
Arizona College of Medicine-Tucson, Tucson, AZ, USA phatic system and cancer invasion and metastasis will be
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 209
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_21
210 S. P. Leong and M. H. Witte

discussed in detail in this chapter and by others in this sec- They then join the cisterna chyli and cut off their connections
tion. Recent discovery of several lymphatic-specific bio- with adjacent veins. All lymph sacs except the anterior por-
markers has opened new vistas in the understanding of the tion from which the cisterna chyli develops become invaded
lymphatic system and its pervasiveness in health and multi- by mesenchymal cells and are converted into groups of lymph
system disease. The development of the SLN concept and its nodes. The other lymphoid structures include the spleen,
clinical utility in the staging of cancer such as melanoma and which develops from mesenchymal cells from the dorsal
breast cancer have underscored the significance of the lym- mesentery of the stomach and the thymus, which arises as an
phatic system in cancer metastasis. This chapter focuses on outgrowth of the third pharyngeal pouch [5].
the development of the lymphatic system, the process of In contrast, the centripetal theory as suggested by
lymphangiogenesis, and its relationship to cancer invasion Kampmeier [6] and others [7] states that the lymphatic ves-
and metastasis. Maldevelopment and malformations of the sels arise from the mesenchymal tissues in the periphery and
lymphatic system will not be addressed here but are covered develop toward the central portion of the body to join the
in Chap. 23, which goes into more detail on lymphovascular central venous system to complete the circulatory loop by
genomics and molecular mechanisms in lymphatic delivering the lymph fluid from the final thoracic duct into
development. the left internal jugular vein. This centripetal origin contrib-
utes to the final network of lymphatic vessels arising in dis-
crete organs.
21.2 Development of the Lymphatic
System
21.3 Major Breakthroughs
Although lymphatics were rediscovered by Gaspar Asellius in the Advancement of Research
in 1627 [1, 2], study on the lymphatics did not advance much in the Lymphatic System
beyond postmortem dissections compared to the immediate
interest and rapid progress in exploring the blood circulation The vascular endothelial growth factor (VEGF) family mem-
being associated with William Harvey’s discovery in 1628 bers and their receptors were discovered in 1983 [8] and the
[3]. In 1902, Florence Sabin was the first to propose that the gene was subsequently cloned in 1989 [9, 10]. VEGF-A has
lymphatic system was formed by the sprouting of endothelial been found from fish to mammals. It plays a major role in the
cells from embryonic veins. This process resulted in the for- formation of blood vessels during embryonic vasculogenesis
mation of primordial lymph sacs with lymphatic endothelial and angiogenesis. It is also involved for neovascularization
cells then branching by sprouting into surrounding tissues in cancer and other diseases [11]. VEGF-A is a member of a
thus forming mature lymphatic networks [4]. family of related growth factors including placental growth
By the end of the fifth week of embryonic development, factor (PLGF), VEGF-B, VEGF C, and VEGF-D, and the
lymphatic vessels begin to develop from lymph sacs arising viral VEGF-Es [12–14]. VEGF-C, discovered in 1996 as a
from developing veins (Fig. 21.1), which are derived from the ligand for the orphan receptor VEGFR-3, is crucial in lym-
mesoderm, according to Sabin [4]. The lymph sacs first phangiogenesis for the growth and proliferation of lymphatic
appear as paired jugular lymph sacs at the junction of the vessels [15].
internal jugular and subclavian veins. Thus, the centrifugal In 1999, lymphatic vessel endothelial hyaluronan recep-
theory of Sabin maintains that the lymphatic plexuses spread tor 1 (LYVE1) was discovered. It is a type I intergral mem-
from the jugular lymph sacs to the thorax, upper limbs, neck, brane glycoprotein and encoded by the LYVE1 gene [16].
and head. At least one connection is established between each LYVE1 binds to hyaluronic acid (HA), homologous to
jugular lymph sac and its jugular vein. The left jugular sac CD44, the main HA receptor. The LYVE-1 molecule binds
develops into the superior portion of the thoracic duct. Each both soluble and immobilized HA, like CD44. However, it is
jugular lymph sac retains at least one connection with its jug- different from CD44, it binds with HA on the luminal face of
ular vein, the left one developing into the superior portion of the lymphatic vessel wall. Further, it is completely absent
the thoracic duct. The unpaired retroperitoneal lymph sac at from blood vessels, although it has been found also in other
the root of the mesentery of the intestine, developing from the tissues [17–19]. Thus, LYVE-1 is a very useful specific bio-
primitive vena cava and mesonephric veins sprout and spread marker for lymphatic vessels of smaller caliber [16].
from the retroperitoneal lymph sac to form lymphatic vessels Another molecule that is expressed by lymphatic endo-
and drain the abdominal viscera and diaphragm. The sac thelial cells (LECs) and not by blood vascular endothelial
establishes connections with the cisterna chyli but loses con- cells is podoplanin, which is a small, extensively
nections with neighboring veins. The paired lower posterior O-glycosylated, type I transmembrane glycoprotein, first
lymph sacs arise from the iliac veins and form the lymphatic discovered in rats 1997 [20]. Like podoplanin, Prox-1 is also
vessels of the abdominal wall, pelvic region, and lower limbs. a mucin-type transmembrane protein being expressed in
21 Lymphangiogenesis: Lymphatic System and Lymph Nodes; Cancer Lymphangiogenesis and Metastasis 211

system while independent of the vascular system ultimately

joins the blood circulatory system at the thoracic duct/sub-
clavian vein junction [29] and prompted Witte and Witte to
ask the question: How and when are lymphatics different
from the blood vessels and under what circumstances do
Superficial lymphatics they merge [30]. As noted above, Fig. 21.1 shows the embry-
ological development of the lymphatic system [31].
Jugular lymph sac

Subclavian lymph sac

21.4 Development of the Lymphatic
Lymph node Vessels and Lymph Nodes
Deep lymphatics

Thoracic duct As mentioned previously, lymph sacs formed by the mesen-

chymal cells become elongated to form lymphatic vessels.
Retroperitoneal The lymph sac is surrounded by mesenchymal cells, which
lymph sac then invade the sac and give rise to the lymph node stroma.
Along the track of the lymphatic vessels, lymph nodes
Cisterna chyli formed from the lymph sacs are situated as they are clustered
Posterior lymph sac in the nodal basin in the axilla, groin, neck, chest, and
Lymph node
abdomen.
Superficial lymphatics According to the centrifugal theory, the first lymph sacs
arise from the paired jugular lymph sacs at the junction of the
Fig. 21.1 The developing lymphatic system showing the primary internal jugular and subclavian veins. At the root of the mes-
lymph sacs and the thoracic duct primordium, as well as several entery of the intestine from the primitive vena cava and
lymph nodes of a 2-month-old human embryo. Figure and legend
reproduced with permission from https://thoracickey.com/lymphatic-­ mesonephric veins, the unpaired retroperitoneal lymph sac
pathophysiology/ [31] appears next. The posterior lymph sacs being paired from the
iliac veins are the last to develop. They give rise to capillary
plexuses and lymphatic vessels of the abdominal wall, pelvic
LECs. The Prox-1 gene was first cloned by homology to the region, and lower limbs. All lymph sacs become invaded by
Drosophila melanogaster gene, prospero, and associated mesenchymal cells and converted into groups of lymph
with the development of the lymphatic system [21, 22]. It has nodes, except for the anterior part of the sac from which the
been found that podoplanin expression is regulated by cisterna chyli develops, which is a dilated sac just distal to
Prox-1 in the LECs at the transcriptional level [23]. Both the thoracic duct into which lymph from the upper body, the
podoplanin and PROX-1 are excellent markers for the iden- intestinal trunk, and two lumbar lymphatic trunk drains. The
tification of lymphatic vessels. spleen develops from mesenchymal cells within the layers of
Thus, there are now several very useful lymphatic vessel the stomach on the dorsal mesentery [5].
biomarkers available to date including VEGFR-3, LYVE-1, The lymphatic system consisting of the lymphatic vessels
podoplanin, and Prox-1, as described above. These have and lymph nodes is complex. The lymphatic drainage sys-
been used extensively to study the steps in the development tems of the stomach, pancreas, viscera, pelvis, and lower
of the lymphatic system. Prox-1 has been found to induce the extremities drain into the final pathway to the cisterna chyli
formation of lymphatic vessels from blood vascular endothe- and then to the thoracic duct [32]. In the head and neck, the
lium [24, 25]. lymphatic system drains into the left and right jugular lym-
Using the receptor to the VEGFR-3, Kaipainen et al. [26] phatic trunks with the left emptying into the thoracic duct
have demonstrated the sprouting of lymphatic sacs from the and the right into the right subclavian vein. In the right upper
venous endothelium in the mouse embryo. These studies using extremity, the lymphatic system drains into the right subcla-
the lymphatic biomarkers have supported the centrifugal the- vian trunk, which returns the lymph at the right internal jugu-
ory of lymphatic development. However, Wilting et al. [27, lar and subclavian vein junction, while the left upper
28], using the chimeric quail-chick embryos in the engrafted extremity drains into the left subclavian trunk via the tho-
wing lymphatic system, have shown the centripetal mecha- racic duct. In the upper right portion of the body, the lymph
nism from the lymphangioblasts of the mesenchymal system. drains into the right lymphatic duct, which returns the fluid
Perhaps, both the centrifugal and centripetal processes through the right subclavian vein. In the upper left portion of
contribute to the formation of the lymphatic system and sub- the body, the lymph drains into the thoracic duct, which
sequent remodeling. The final outcome is that the lymphatic returns the fluid into the left subclavian vein. Furthermore,
212 S. P. Leong and M. H. Witte

the lymphatic system includes the mucosa-associated lym- medullary cords. The medullary sinuses are vessel-like
phoid tissues like tonsils, adenoids, Peyer’s patches in the spaces that drain the lymph fluid into the efferent lymphatic
small bowel, and even the appendix [32]. There are about vessels. A lymph follicle (Fig. 21.5b) is shown with its ger-
600 lymph nodes in the human body within the lymphatic minal center with B-cells in the center and T-cells in the
system. About 17 L of the 20 L of the blood at the arteriove- paracortical area. The proliferating B cells in the germinal
nous capillary junctions return through the venous capillar- center are activated by helper T cells in the paracortical area.
ies daily. About 2–3 L of fluid (15%) without cellular Fig. 21.6 shows the distribution of macrophages, dendritic
components of the blood enter into the extracellular space cells, and lymphocytes in the subcapsular follicles. Dendritic
and drain into the lymphatic vessels with the lymph fluid car- cells are present in peripheral tissues such as the skin
rying cellular debris, macromolecules of proteins, excess Langerhans cells and in the inner lining of the nose, lungs,
water, and toxins [33]. stomach, and intestines. Activated dendritic cells migrate to
The lymphatic vessel has unidirectional semilunar valves the lymph nodes where they interact with T cells and B cells
with an average diameter of 100 μm. The lymphatic vessel to initiate adaptive immune response. A significant number
wall is thin and lined by endothelial cells surrounded by of dendritic cells are found in the T-cell areas of peripheral
smooth muscle with the outermost layer being the adventitia lymphoid organs such as the spleen, lymph node, and Peyer’s
anchoring the lymphatic vessels to the surrounding tissue, patch. High-endothelial venules (HEVs) are specialized
mainly the adipose tissue [34]. Fig. 21.2a, b shows the postcapillary venules (Fig. 21.7). They are termed “high”
detailed drawing to illustrate the microanatomy of the struc- because of the thickness of the endothelial cells of the low
ture of the lymphatic vessel [35]. cuboidal type in contrast to the typical squamous epithelial
Lymphedema manifests when the extracellular fluid is type of the endothelial cells. HEVs are vascular sites for the
not adequately drained. It is discussed in detail in Chap. entry and exit of lymphocytes from the secondary lymphoid
25. The lymph fluid is filtered through multiple lymph organs. HEV endothelial cells have ligands that bind lym-
nodes and finally drains predominantly into the thoracic phocytes, facilitating their adhesion and subsequent extrava-
duct and subclavian veins as normally nearly sterile lymph sation into the lymph node. Thus, an adult lymph node
fluid to rejoin the blood circulation. For a more detailed consists of primarily the capsule and trabeculae, both are of
review of the anatomy and physiology of the lymphatic the dense connective tissue, reticular tissue with reticular
system, please see separate reviews [35–37]. The struc- cells, and fibers forming the supporting meshwork and mac-
ture of the lymphatics and lymphatic vessels is shown in rophages, which process antigen and mingle in with the
Fig. 21.2. reticular cells [38–41].
Figure 21.3 shows the structure of a mature lymph node. Lymphoid stromal cells are best known as the architec-
Fig. 21.4 shows the hematoxylin and eosin staining of a tural cells of lymphoid organs [42]. It is now known that the
lymph node showing the capsule, subcapsular sinus, cortex, stromal cells are primarily responsible for the adaptive
lymph follicle, medulla, and medullary sinus. B cells are pri- immune response. Fibroblastic reticular cells are derived
marily located in the outer cortex where they are grouped from mesoderm to form the stroma consisting mainly of the
together as follicular B cells in lymphoid follicles. Mature B extracellular matrix (type III collagen) with supporting net-
cells become plasma cells to secrete antibodies against for- work of cells and extracellular matrix in secondary lym-
eign antigens by means of humoral immunity. The medulla is phoid organs namely lymph nodes, spleen, and thymus.
the innermost of the lymph node consisting of large blood The fibers crisscross within the sinuses. They produce
vessels, sinuses, and medullary cordis, where antibody-­ reticular fibers to form a network between the capsule and
secreting plasma cells, B-cells, and macrophages are found. trabeculae allowing the interaction and movement of T
Macrophages are phagocytic with the ability to engulf and cells, B cells, dendritic cells (DCs), plasma cells, and mac-
destroy bacteria and other harmful organisms. They may also rophages [43]. These myofibroblasts differentiate into
present antigens to T cells and release inflammatory cyto- fibroblastic reticular cells in the T cell zone of the cortical
kines. Macrophages take up the carbon such as from tattoos and medullar regions. They bind dendritic cells which pres-
and appear black. Fig. 21.5a shows the germinal center of a ent antigens to T cells as well as induce remodeling reticu-
lymph node and lymphoid follicles by hematoxylin and lar cells in lymph nodes. Further, the reticular cells may
eosin staining. Fig. 21.5b shows a close-up of a germinal regulate the traffic of lymphocytes and activation of spe-
center and paracortical T cell zone. B cells are primarily cific lymphocyte populations [44]. The myofibroblast
clustered in structures called lymphoid follicles, whereas T markers consist of desmin, vimentin, CD90, CD73, CD103,
cells are found mainly in the paracortex. The T cells contrib- smooth muscle actin (SMA) and ERTR7 antigen, podo-
ute to the cellular immunity by recognizing the foreign anti- planin, platelet-derived growth factor receptor, and genes
gen on the infected cell and destroying it. The medullary from antigen presentation and cytokine response pathways
sinuses or sinusoids are vessel-like spaces separating the but lacking CD45 and CD31.
21 Lymphangiogenesis: Lymphatic System and Lymph Nodes; Cancer Lymphangiogenesis and Metastasis 213

a Interstitial Tumour Immune

Highly
fluid cell cell Intercellular Flap
permeable
junction

Blind-
ended
sacs
Direction of
lymphatic flow
lymphatic
Initial

Minimal
BM
Thin, discontinuous walls Adherens and
tight junctions
No supporting cells
Pre-collecting

Occasional valves
lymphatic

Discontinuous BM Fluid flow

Some SMC
coverage
Lymphatic endothelial cells
Frequent valves
Collecting
lymphatic

Lymph node

SMC layer
Complete
BM

b Initial lymphatic Blood capillary

Pericyte
BEC

LEC
Elastic fibre
BM (incomplete)
BM (complete)
Anchoring
filament

Nature Reviews | Cancer

Fig. 21.2 (a and b) Structure of the lymphatic vasculature and lym- vascular capillary is shown for comparison. The anchoring filaments
phatic endothelial cells from initial lymphatics. The hierarchy of lym- that are associated with initial lymphatic vessels connect the ablumi-
phatic vessel subtypes is shown in part a, with the characteristics of nal membrane of LECs to surrounding elastic fibers—these filaments
the subtypes indicated. The association between lymphatic endothe- are short, but for visual clarity, they are shown larger than to scale.
lial cells (LECs) in initial lymphatics is shown in the two pull-out Anchoring filaments are involved in controlling the entry of cells and
boxes, with the contact regions between LECs in the vessels being interstitial fluid into the initial lymphatics. It should be noted that
indicated by the brown lines in the first pullout box. Fluid is thought cells can cross lymphatic endothelial layers via a transcellular route,
to enter (as shown by the arrows in the second pullout box) via the tips as well as via the paracellular route that is shown here. BEC, blood
of the flaps of LECs, without compromising the integrity of the inter- vascular endothelial cell; BM, basement membrane; SMC, smooth
cellular junctions on the sides of the flaps. The structure of LECs from muscle cell. Figure and legend reprinted by permission from Springer
initial lymphatics is shown in part b, illustrating the entry of fluid and Nature: Nature Reviews Cancer, Lymphangiogenesis and lymphatic
cells, which is favored when the interstitial fluid pressure is high, vessel remodeling in cancer [35]
causing stretching of the extracellular matrix. The structure of a blood
214 S. P. Leong and M. H. Witte

Afferent Lymph Primary

Lymphatic Flow Follicle
Vessel

Secondary
Follicle
Capsule

Germinal
Center
Marginal
Sinus

Medullary
Sinus

Cortex Mantle
of Small B
Lymphocytes

Paracortical
Area (T Cell Zone)
Senescent
Germinal Center

Medulla Efferent
Lymphatic Blood
Vessel Vessels

Fig. 21.3 Structure of a lymph node with arterial supply and venous cells, and plasmas cells. Surrounding the germinal center is the mature
drainage. The diagram shows the afferent and efferent lymphatic ves- B cells with the cortex. The T cells reside in the paracortical area, the
sels. The lymph node is surrounded by a fibrous capsule. Medulla is space between the cortex and the medulla. Figure and legend repro-
toward the center of the lymph node. Multiple follicles or lymphatic duced with permission from https://www.iheartpathology.net/post/
nodules are noted. Within the lymphatic follicle, there is the germinal lymph-­nodes-­101 [39]
center (Fig. 21.4) with lymphoblasts, macrophages, follicular dendritic

Dendritic cells or antigen-presenting cells originate from [29, 36, 47]. During cancer treatment, secondary lymph-
the same common progenitor as monocytes. Their role in edema may arise [48]. These topics are beyond the scope of
adaptive immunity consists of presenting antigens and induc- this chapter but are discussed in Chap. 25.
ing a primary immune response in resting naïve T cells [45].
Recently it has been shown that recirculation of antigen-­
bearing dendritic cells in remote lymphoid organs can prime 21.5 Trafficking of the Lymphocytes
T cells in other locations [46].
During the development of the lymphatic system, so-­ Lymphocytes circulate continuously between the blood
called lymphangiogenesis, as described in Section 21.9, mal- and secondary lymphoid tissues such as the lymph nodes,
development and malfunction may be encountered. Thus, Peyer’s patches, and the spleen, where antigens and
primary lymphedema and lymphangiodysplasia may occur antigen-­presenting dendritic cells are selectively present.
21 Lymphangiogenesis: Lymphatic System and Lymph Nodes; Cancer Lymphangiogenesis and Metastasis 215

Capsule

Cortex Fat

Blood vessel

Medullary sinus

Medulla

Sub-capsular sinus

Fig. 21.4 Histologic hematoxylin and eosin staining of part of a lymph contains mostly T-cell lymphocytes. B-cell lymphocytes respond to
node (low power magnification) showing the cortex (B cells) and para- antigens that enter the lymph node and produce antibodies once they are
cortex space of T cells between the cortex and medulla. Lymph fluid activated. T-cell lymphocytes respond to cell-bound antigens as cyto-
enters the subcapsular sinus gap between the clusters of lymphocytes toxic T cells against infected cells with viral particles or even cancer
and the capsule wall. The lymph is filtered through the node. The filter- cells. On the other hand, helper T cells may assist B-cells in antibody
ing process ensures anything nonself that is found in the lymph is production. The paracortical region may become expanded in a T cell-
selected out and processed by the lymphocytes. The cortex of the lymph dominant immunological response. Courtesy of Ed Uthman MD,
node is mostly populated by B-cells, and in the paracortex area (slightly Flickr, originally published under CC BY [38]
deeper region that is sandwiched between the cortex and the medulla)

The naive lymphocytes first bind to and migrate across

HEVs. The HEVs are distinctly different from the normal 21.6  evelopment of the Sentinel Lymph
D
venules in, at least, two ways. First, HEVs express unique Node Concept
adhesion molecules, so-called vascular addressins that
act as ligands for homing receptors of the lymphocytes. When Halstad developed radical mastectomy including a
Second, chemokines and chemokine-binding molecules complete axillary lymph node dissection for breast cancer
are produced within the extracellular matrix in and [50] and Snow championed en bloc resection of the primary
around the HEVs. These chemokines activate integrins melanoma along with a radical regional lymph node dissec-
on circulating lymphocytes and attract them into the tion [51, 52], they assumed that cancer spread from the pri-
lymph nodes and Peyer’s patches. Thus, the lymphocytes mary site then to the regional nodal basin before metastasizing
can navigate themselves into the lymph nodes or Peyer’s to the distant sites. Halsted’s influence was very profound so
patches to launch effective immune responses with the that radical lymph node dissections became standard for dif-
help from the antigen-presenting dendritic cells [49]. The ferent types of cancer such as the esophagus, gastric, colon,
trafficking of lymphocytes is schematically shown in gynecological cancers, and others. Haagensen et al. described
Fig. 21.8. Detailed molecular mechanisms of lymphocyte the lymphatics in cancer extensively [53]. Although adequate
trafficking and immune responses are areas under exten- resection of the primary site and simultaneous radical
sive investigation. regional lymph node dissection was performed, systemic
216 S. P. Leong and M. H. Witte

a Capsule dilemma of staging the regional lymph node by a simpler and

more reliable procedure to remove the SLN for evaluation
Adipose tissue
Germinal Center
and yet to avoid a radical lymph node dissection associated
with significant morbidity. Thus, if the SLN biopsy is nega-
tive, a radical lymph node dissection may be spared.
Lymphoid follicles
Furthermore, melanoma [59] and breast cancer [60] patients
with a negative SLN biopsy enjoy a significantly improved
survival.
In the post-sentinel node era, we have learned that, in
most cases such as melanoma and breast cancer, cancer
spreads in an orderly process [61] from the primary site to
b the SLNs primarily and then beyond to the distant sites.
Mantle
A recent multicenter SLN Working Group study showed
that the distant metastatic rate was 4.7% [62] for 4059 mela-
noma patients with a negative SLN biopsy with a median
follow-up of 31.2 months. In another retrospective study, it
was found that 126 of 1277 melanoma patients with a nega-
Germinal Center tive SLN biopsy developed distant disease (9.9%) [59].
Similarly, for breast cancer patients with a negative SLN
biopsy, their recurrence rate is also very low, less than 5%
[60]. Thus, most patients with a negative SLN biopsy were
disease-free. The patients with a negative SLN biopsy of
about 5–10% may bypass the SLN, likely through other
T-cell Zone
missed SLNs, lymphatic-venous communications, or blood
vessels, to spread to the distant sites.
Overall, melanoma and breast cancer patients with a neg-
Fig. 21.5 (a) Subcapsular location of germinal centers and lymphoid ative SLN biopsy manifest much better disease-free and
follicles by hematoxylin and eosin staining (low power magnification). overall survival than those with a positive SLN biopsy [63,
Figure and legend reproduced with permission from https://www. 64]. Thus, the spectrum theory of Hellman seems to be most
iheartpathology.net/post/lymph-­nodes-­101 [39]. (b) Histologic hema-
compatible with the clinical outcome of patients undergoing
toxylin and eosin staining (high power magnification) of the germinal
center within a lymph follicle. The germinal center consists of lym- SLN staging procedure. The spectrum theory is applicable
phoblasts, macrophages, follicular dendritic cells, and plasmas cells. for melanoma patients with minimal tumor burden in the
Surrounding the germinal center is the mature B cells with the cortex. SLN. When the micrometastasis has been found to be less
The T cells reside in the paracortical area, the space between the cortex
than 0.1 mm, their clinical outcome is the same as those
and the medulla. Figure reproduced with permission from http://med-
cell.med.yale.edu/systems_cell_biology/lymphatics_lab.php [40] patients with a negative SLN biopsy, according to van Akkooi
et al. from the Rotterdam criteria [65, 66]. In breast cancer
patients, similar findings have been found in that if the breast
recurrence still occurred. Using breast cancer as a model, cancer SLN is found to harbor isolated tumor cells (ITC;
Fisher proposed the systemic theory stating that clinically ≤0.2 mm), the clinical outcome of these breast cancer
apparent cancer is a systemic disease [54] in that cancer dis- patients is the same as those with a negative SLN biopsy
semination occurs even when clinically the cancer burden is [67].
small. Accordingly, treatment of local or regional disease When an SLN is positive with micrometastasis, is it an
should not affect overall survival. Thus, Cady concluded that incubator for cancer cells to proliferate before spreading to
the lymph node was the marker but not the governor of can- the distant sites or is it a marker for systemic disease [59]
cer metastasis [55]. An alternate spectrum theory by Hellman Perhaps, there is a spectrum of the SLN from an incubator
was developed to indicate that cancer development was pro- then to a marker, when the tumor burden reaches a certain
gressive. Cancer may be compartmentalized initially before volume with associated distant metastasis. The cutoff level
becoming widespread. Thus, oligometastatic cancer is local- of tumor burden for this transition has not been determined.
ized. Appropriate treatment such as surgery or radiation ther- In the post-sentinel node era, we have learned that, in most
apy may render extermination of the metastasis either alone cases such as melanoma and breast cancer, cancer in the
or combined with systemic therapy [56]. early stage spreads in an orderly process [61] from the pri-
The pioneering work on the development of the SLN con- mary site to the SLNs primarily and then beyond to the dis-
cept by Cabanas [57] and Morton [58] has resolved the tant sites.
21 Lymphangiogenesis: Lymphatic System and Lymph Nodes; Cancer Lymphangiogenesis and Metastasis 217

Fig. 21.6 High power view

of the subcapsular follicles Capsule
and sinus of a lymph node.
The mature lymphocytes
Dendritic cell
appear to be more in
comparison to other cells,
such as phagocytic cells like
macrophages and dendritic
cells, which are in the
subcapsular sinus. Most of the
other irregular looking cells
are lymphocytes in their
different developmental
stages. Figure reproduced Lymphocytes
with permission from https://
histologyblog.
Lymphocytes
com/2012/04/30/ Macrophage
histoquarterly-­lymph-­node/
[41]

100 µm

Fig. 21.7 Lumens of the

high endothelial venule
(HEV), high power view. One
significant feature of the
paracortex T-cell zone of the
lymph node is the HEV,
where circulating
lymphocytes leave the
bloodstream to enter the node.
These postcapillary HEVs can
be distinguished by their
cuboidal endothelial cells. High Endothelial Vessels
Adhesion molecules called
selectins and integrins on the
surfaces of HEVs and
lymphocytes mediate the
attachment of lymphocytes to
endothelial cells. Figure
reproduced with permission
from http://medcell.med.yale.
edu/systems_cell_biology/
lymphatics_lab.php [40]

21.7 Trafficking of Cancer Cells a single cancer cell is slightly larger than that of a lympho-
cyte with an average diameter of 15 μm, thus, the diameter of
The cancer cells utilize the lymphatic system to spread from a cancer cell may be up to 20 μm or more, and it can comfort-
the primary site as they dissociate themselves from their ably pass through the 100 μm lymphatic vessel. Figure 21.9
microenvironment and enter the lymphatic vessels toward shows that cancer may spread through the lymphatic and
the lymph nodes and beyond to the distant sites. The size of blood vessels.
218 S. P. Leong and M. H. Witte

Afferent lymphatic B Subcapsular

Cortex sinus
T-cell zone
FRC conduit
B Medulla

Artery

HEV Vein

B Efferent
lymphatic

Dendritic cells
Lymphocytes B
Antigen
Humoral factors Lymphocytes

HEV = high endothelial venule PMID: 15122201 Nature Reviews | Immunology

Fig. 21.8 Trafficking of lymphocytes in and out of the lymph node. the lymph node. B—B-cell zones. DC—dendritic cells. FRC—fibro-
Unfiltered lymph fluid with cells as lymphoctyes and dendritic cells blastic reticular cells. HEV—high endothelial venues. Figure and leg-
flow into the lymph node through the afferent (towards lymph node) end reprinted by permission from Springer Nature: Nature Reviews
lymphatic vessels. Efferent (away from lymph node) lymphatic vessels Immunology, Lymphocyte trafficking across high endothelial venules:
flow out of a lymph node and carry filtered lymph fluid. Lymphocytes Dogmas and enigmas [49]
use either afferent lymph or high endothelial venules for migration into

Systemic that systemic metastasis in breast cancer is most often

Sites owing to SLN metastasis and only occasionally directly
into the bloodstream [69]. Once the breast cancer cells
reach the SLN, they may gain direct access to the sys-
Vascular temic circulation to the distant sites by invasion into the
channel Sentinel Lymph Node
Vascular
Sentinel Lymph Node channel
veins in the lymph node [69]. Figure 21.10 shows the
possible mechanism of cancer metastasis to the SLN and
then to the distant sites through the venous channel
within the lymph node [70].
Thus, a critical question is whether cancer clones that
Lymphatic Channel spread through the blood vessels are the same clones that
Primary spread through the lymphatic vessels. In other words, are the
Cancer Site cancer cells dependent on lymphatic or blood vessels for
Non-Sentinel
Lymph Node metastasis? Induction tumor angiogenesis provides a possi-
ble explanation for how cancer cells escape their original site
Fig. 21.9 Two major routes of cancer metastasis: one through the lym- by invading the newly formed blood vascular bed [71].
phatic vessels to the SLNs as the primary gateway and the other through Tumor vascularization has been observed in human mela-
the vascular channels directly to the distant sites bypassing the SLNs. noma both experimentally [72, 73] and clinically [74]. The
Figure reprinted by permission from Springer: Clinical and
Experimental Metastasis, Front page (Special Issue on: The 2017 sev-
importance of tumor angiogenesis for the prognosis of pri-
enth International Cancer Metastasis Symposium; Cancer Metastasis mary malignant melanomas of the skin, however, has
through the Lymphovascular System: Biology and Treatment [68]) remained controversial [75]. Whereas several studies found
that tumor microvessel density was associated with decreased
As described previously based on the clinical out- disease-­free and overall survival [76, 77], other reports did
come of melanoma and breast cancer patients with a not detect any significant differences of tumor blood
negative SLN biopsy being very favorable with low rate microvessel density between metastatic and nonmetastatic
of distant metastasis, it can be induced that the SLN may melanomas [78]. Therefore, the potential prognostic value of
be the major gateway in early cancer development for tumor vascularization in malignant melanoma remains at
metastasis to the distant sites. Nathanson et al. suggest present unclear [79].
21 Lymphangiogenesis: Lymphatic System and Lymph Nodes; Cancer Lymphangiogenesis and Metastasis 219

Left Middle Right

Lymp
Efferent
lymphatic

h flow
vessel

Local
blood
vessel

Lymph
Fatty acids Midkine

flow
CD36 VEGFC
Extracellular vesicles Afferent
PD-L1 lymphatic
Dendritic cells vessel
MHC I/II NK cells
Neutrophils
FABP5 Macrophages
Lymphatic endothelial cells

Fig. 21.10 Cancer cells survive in lymph nodes and spread to other metastasis. Although the lymph node is filled with various immune cell
organs. (Left) To adapt to the lipid-rich environment in lymph nodes, populations, the microenvironment is immunosuppressive. Tolerogenic
cancer cells enhance their reliance on fatty acid metabolism by upregu- DCs, M2-like TAMs, and CD15+ TANs were enriched in metastatic
lating CD36 and FABP5. In addition, to avoid immune destruction, can- lymph nodes. These cells contribute to the inhibition of cytotoxic T
cer cells increase expression of the immunosuppressive checkpoint cells. The lytic activity of NK cells is also reduced in metastatic lymph
molecule PD-L1, while decreasing MHC I/II expression. (Middle) nodes. (Right) After surviving in the lymph node, cancer cells can
Midkine and VEGF-C secreted by cancer cells induce lymphangiogen- invade the lymph node blood vasculature and further spread to distant
esis in draining lymph nodes prior to seeding, generating a pre-­ organs. Cancer cells can also exit through the efferent lymphatic vessel.
metastatic niche. Similarly, cancer cells also release extracellular Figure and legend originally published under CC BY 4.0 [70] with per-
vesicles to interact with B cells in the cortex, promoting lymph node mission to be reprinted here

It is apparent that the lymphatic system draining various microenvironment of the site to allow it to support the cancer
body organs is different, some with more complex networks cells which may spread to such a site [80]. Other processes
than the others. The SLN drainage of melanoma and breast may include recruitment of certain cell types that support
cancer follows more reliable routes than the internal organs metastatic implant, induction of growth factors, and cyto-
such as the lungs, stomach, pancreas, and other organs, kines to prepare for the pre-metastatic niche and restructuring
which show complicated lymphatic drainage patterns. of the extracellular matrix [81, 82]. Perhaps, the seed and soil
Certainly, melanoma and breast cancer allow us to appreci- concept of cancer metastasis of Paget [83] may be explained
ate the biology of early cancer arrival in the SLNs and may by the fact that the soil is “pre-conditioned” as the pre-meta-
form a prototype for cancer spread to the SLNs and beyond static niche for the seed to survive and grow. Evidence for the
to the non-sentinel lymph node compartments [64] and pre-metastatic niche is extensive including pre-metastatic
beyond to the distant sites. morphological and histological changes, immunological
alterations, tumor-induced immunosuppression in lymph
nodes, vascular remodeling, tumor-secreted conditional fac-
21.8  vidence of Metastatic Niche
E tors, and remodeling of extracellular matrix [82].
in Sentinel Lymph Nodes Primary tumors may secrete soluble factors to induce
the pre-metastatic niche in the tumor draining or SLNs
A microenvironment that facilitates the survival and progres- [82]. In melanoma, immunosuppressive microenvironment
sion of disseminated cancer cells is termed a pre-metastatic suppressor cell activity has been shown in tumor-­draining
niche. Cancer cells may secrete factors to precondition the lymph nodes [84–86]. Similarly, in breast cancer SLNs,
220 S. P. Leong and M. H. Witte

immunosuppressive activity was noted [87–89]. Further, genesis in vitro) [96, 99, 100]. The recognition of tumor
immunomodulatory cytokines have been found in mela- lymphangiogenesis did not occur until more than a decade
noma SLNs [90, 91]. later. Indeed, following the elucidation and extensive inves-
Thus, Cochran et al. provocatively asked Is SLN suscep- tigation of blood vessel angiogenesis and its relation to
tibility to metastases related to nodal immune modulation tumor growth, the presence and potential significance of the
[92]. On a molecular basis, Takeuchi et al. have demonstrated phenomenon was still questioned and controversial [101–
a potential mechanism by tumor-draining lymph nodes 104]. Subsequently, particularly with the development of
expressing CCL21/SLC for recruitment and homing of lymphatic-specific markers to identify lymphatics and dis-
CCR7-positive metastatic melanoma cells [93]. tinguish them from blood vessels on histologic sections, it
Of course, as SLNs are the closest draining lymph nodes has been demonstrated that tumor lymphangiogenesis not
to the primary tumor, they may be more effectively precon- only occurs within the primary tumor with intratumoral
ditioned by factors secreted by the primary tumor to develop lymphatics but also peritumorally, in draining lymph nodes
pre-metastatic niche resulting in preference of the primary and even in distant metastatic niches as cancer cells spread
tumor to invade the SLNs. To a lesser extent, the non-SLN through the lymphatic system. Indeed, tumor lymphangio-
and distant sites are much less affected by their lower gradi- genesis seems to be a superior predictor of poor prognosis
ent of receiving the “pre-­metastatic niche” inducing factors. and cancer progression [79] than blood vessel angiogenesis
Perhaps, the orderly progression of cancer metastasis from (hemangiogenesis).
the primary site to the SLN [61], then to the non-SLNs and Thus, it is now well-recognized that in the cancer micro-
distant sites follow this gradient phenomenon. This “gradi- environment, cancer cells release VEGF A/C, which pro-
ent” may explain and reflect the rationale of the tumor nodal motes local lymphangiogenesis, a process by which new
metastasis (TNM) staging of cancer. lymphatic vessels are formed. Cancer cells may also undergo
Since the tumor-draining node or SLN is the critical epithelial–mesenchymal transition to boost their invasive-
microenvironment for the cancer cells to invade and pro- ness. Further, they can upregulate the expression of CCR7/8
liferate, it should be the target of therapeutic manipula- and CXCR4/5, which respond to the corresponding CCL1/21
tion. Granulocyte/macrophage colony-stimulating factor CXCL10/12 from the lymph node, respectively.
(GM-CSF) has been found to reverse immunosuppression in Cancer cells respond to these stimuli by migrating to lym-
melanoma SLNs [84]. A recent study showed that nanoparti- phatic vessels through chemotaxis. Some cancer cells pro-
cle targeting of dendritic cells in tumor-­draining lymph nodes duce ALOX15, which functions as a catalyst to convert
stimulated maturation of dendritic cells, which then induced a arachidonic acid to 12[S]-HETE and 15[S]-HETE. These
TH1 response. Such a response resulted in increased numbers metabolites cause circular defects on lymphatic endothelial
of tumor antigen-specific CD8+ T cells with resultant slower cells (LECs), allowing cancer cells to enter the lymphatic
tumor growth [94]. Thus, targeting the SLNs to reverse the vessels. In addition, interstitial fluid flow forces the cancer
immunosuppression and increase antitumor response may be cells mechanically into the lymphatic vessels. TWIST1 plays
potentially a viable strategic option. A more extensive discus- an important role in metastasis, like angiogenesis, invadopo-
sion of pre-metastatic niche for disseminated cancer cells to dia, and extravasation. SNAIL1/2 downregulates E-cadherin
establish it as the soil to grow is addressed in a review by and induces epithelial to mesenchymal transition in the pro-
Sleeman [82]. cess of cancer metastasis. The flow also assists cancer cells
to produce and respond to the autocrine chemokine gradients
to lymphatic vessels [70]. Figure 21.11 shows the migration
21.9  ancer Lymphangiogenesis: A Marker
C and invasion of lymphatic vessels by cancer cells following
for Cancer Metastasis lymphangiogenesis.
Cancer-induced lymphatics by lymphangiogenesis may
Understanding the growth and development of the lym- become dilated [105]. It may be speculated that a cluster of
phatic system has increased dramatically since it was first cancer cells may accumulate at the valve of the lymphatic
described in the developing embryo by Sabin in the early vessel [106] and eventually a colony of cancer cells may
1900s. Lymphangiogenesis was subsequently investigated grow on the inner wall of the lymphatic vessel with subse-
in vivo during regeneration after severing, inflammation, quent invasion into the surrounding fat tissue as in-transit
and ligation [95]. However, the term lymphangiogenesis metastasis, such as the case in melanoma as shown in
was not introduced until the early 1980s when lymphatic Fig. 21.12. The other possibility is that a specific receptor on
endothelial cells (LECs) were first isolated in vitro [96–98], the cancer cell may attach to the inner wall of the lymphatic
and branching structures were seen in vitro (lymphangio- vessel to establish a nidus to grow. Although the ­clinical sig-
21 Lymphangiogenesis: Lymphatic System and Lymph Nodes; Cancer Lymphangiogenesis and Metastasis 221

Left Right

Interstitial fluid flow

ALOX15
VEGFA/C
CCL1/19/21
Tumor Lymphatic capillary
CXCL10/12
CCR7/8

CXCR4/5
TWIST1

SNAIL1/2

Fig. 21.11 Migration and invasion of cancer cells into the lymphatic vessels following lymphangiogenesis. Figure and legend originally pub-
lished under CC BY 4.0 [70] with permission to be reprinted here

nificance of SLNs in melanoma [107] and breast cancer with increased intratumoral lymphatics. A relative lym-
[108] has been well established, the molecular mechanism as phatic vessel area of >1.5% was significantly correlated
to how the cancer cells migrate to the SLNs has not been well with poor disease-free and overall survival, but there were
characterized. no differences in the density of tumor-associated blood ves-
Skobe and Detmar have found that VEGF-C induces can- sels. They proposed that intratumoral lymphangiogenesis
cer lymphangiogenesis associated with breast cancer metas- was a novel prognostic marker for the risk of lymph node
tasis [109]. Following orthotopic transplantation of breast metastasis in melanoma [79]. Numerous studies have
cancer in nude mice, they have found intratumoral lymphan- shown that increased lymphangiogenesis associated with
giogenesis within human breast cancers using LYVE-1. increased intratumoral lymphatics and lymphatic invasion
Since VEGF-C overexpression in breast cancer cells was is correlated significantly with reduced overall survival in
associated with increased intratumoral lymphangiogenesis melanoma, breast, colorectal [110, 111] and lung cancer
associated with the lymph node and lung metastases, they [112, 113]. It has been suggested that lymphangiogenesis
concluded that VEGF-C could be considered as a molecular and the remodeling of existing lymphatics may play an
link between cancer lymphangiogenesis and metastasis. important role in cancer metastasis [35].
Dual staining with LYVE-1 and the panvascular marker LECs in the lymph node maintain peripheral tolerance by
CD31 was used to study lymphangiogenesis in cutaneous directly inhibiting T-cells against autologous antigens under
melanoma by Dadras and Detmar [79]. Nineteen melanoma normal conditions. Using two distinct experimental tumor
patients with no evidence of metastatic disease and 18 mel- models, Dieterich et al. have found that PD-L1, a T-cell
anoma patients with early nodal metastasis were closely inhibitory molecule, is expressed by cancer-associated LECs
matched for histological and clinical features. Their sam- similar to the LECs in the lymph node. However, PDL-1 is
ples were examined by dual staining as mentioned previ- downregulated in cancer-associated blood vessels. It was
ously. It was found that the incidence of intratumoral speculated that the upregulation of PDL-1 in the LECs was
lymphatics was significantly higher in patients with meta- probably a result of IFN-g released by tumor microenviron-
static melanoma. Poor disease-free survival was associated ment stromal cells. Further, the authors have found that when
222 S. P. Leong and M. H. Witte

melanoma was correlated with poor outcomes (unpublished

data from the authors). Using a transgenic murine model in
which vascular endothelial growth factor C (VEGF-C) was
induced in the lung, the authors found more growth of pul-
monary metastases associated with greater dissemination to
other organs. Thus, they have shown that lymphangiogene-
sis is present in metastatic sites with further metastasis to
other sites [117]. As described previously, cancer-induced
lymphangiogenesis facilitates cancer metastasis through the
lymphatic vessels. VEGF receptor-3 (VEGFR-3) expression
on tumor cells with autocrine signaling by VEGF-3 has been
found to be associated with increased metastatic potential.
Also, lymphatic-homing chemokine receptors, particularly
C–C chemokine receptor 7 (CCR7), which is expressed in B
Fig. 21.12 Metastatic melanoma is trapped in the lymphatic vessel and T cells as well as mature dendritic cells, may be involved
within the subcutaneous tissue. Figure and legend reprinted by permis-
sion from Springer: Clinical and Experimental Metastasis, Molecular in lymph node metastasis. Issa et al. have found that VEGF-C
mechanisms of cancer metastasis via the lymphatic versus the blood acts to increase secretion of CCL21 by lymphatic tissue,
vessels [114] another lymphoid homing chemokine, which interacts with
CCR7 to initiate cancer invasion toward lymphatics. The
PDL-1 is blocked, T-cell stimulation by antigen-presenting CCL21/CCR7 axis leads the CCR7 lymphocytes to the
LECs was increased in vitro. Based on these findings, the tumor-draining lymph node [93] and renders it immunosup-
authors assert that tumor-induced lymphatic vessels are criti- pressive [118]. VEGF-C also enhances tumor invasiveness
cal in the initiation of metastasis and modulation of immune through autocrine signaling by increasing the proteolytic
responses against autologous tumor by transporting antigen activity and motility of tumor cells in a three-­dimensional
from the tumor microenvironment to tumor-draining lymph matrix. Further, VEGF-C induces production of lymphatic
nodes. The T-cell inhibition by LECs is like that of periph- CCL21 in vivo when VEGF-C protein was injected intrader-
eral tolerance by LECs in lymph nodes. Since PDL-1 may be mally in the mouse. Thus, the authors conclude that their
involved in tumor tolerance, checkpoint inhibitors such as data support the fact that VEGF-C promotes tumor invasion
anti-PD-1 or anti-PDL-1 may be useful to reverse such a toward lymphatics by both autocrine and CCR7-associated
tumor tolerance [115]. paracrine signaling mechanisms in addition to lymphangio-
Further study on the pre-metastatic niche in draining genesis initiated by VEGF-C. Both of these effects are
lymph nodes was carried out in two murine models with the VEGFR-3 dependent based on the 3-D matrix model [119].
4 T1 breast cancer and B16F10 melanoma by Commerford Using the 3D matrix model, Shields et al. have further
et al. [116]. They have found that LEC expansion in tumor-­ shown that CCR7 creates a gradient across the cells of
draining lymph nodes representing lymphangiogenesis was increased interstitial flow to enhance tumor cell migration
noted as early as 4 days after tumor injection. Altered tran- into the lymphatic vessels like that of dendritic cell traffick-
scriptional profile of LECs from tumor-draining lymph ing [120].
nodes as compared to naive lymph nodes was observed with Although lymphangiogenesis by VEGF-C is associated
RNA sequencing in both tumor models. Integrin aIIb upreg- with poor prognosis, Fankhauser and Swartz have found that
ulated in LECs of tumor-draining lymph nodes. Since integ- VEGF receptor-3 (VEGFR-3) blocking antibody was able to
rin aIIb was shown to mediate LEC adhesion to fibrinogen reverse immunosuppression in a murine model as VEGF-C
in vitro and LEC-associated fibrinogen was also identified in was found to enhance the effect of antibody immunotherapy
tumor-draining lymph nodes in vivo, the authors suggested [121]. VEGF-C induced CCL21, which caused tumor infil-
that aIIb could be involved in lymphatic remodeling with tration of naive T cells. When the mice were treated with
possible formation of pre-metastatic niche ahead of the checkpoint inhibition immunotherapy, the antitumor
arrival of cancer cells [116]. response was significantly increased. Thus, even though
In a separate mouse metastasis model, Ma et al. found VEGF-C causes lymphangiogenesis being associated with
that lymphangiogenesis occurred in distant lung metastases poor prognosis, its induction of CCL21 may be exploited for
and that some metastatic tumor cells were located in the a positive antitumor effect by checkpoint inhibition immuno-
lymphatic vessels and draining lymph nodes of the distant therapy. The authors further completed a clinical trial to
metastatic sites [117]. In patients with metastatic melanoma, show that gene expression of VEGF-C was strongly associ-
increased lymphatic density was noted within and around ated with T cell inflammation. Further, serum VEGF-C con-
the pulmonary metastases. Invasion of lymphatic vessels by centrations were correlated with T cell activation and
21 Lymphangiogenesis: Lymphatic System and Lymph Nodes; Cancer Lymphangiogenesis and Metastasis 223

expansion as well as clinical response to peptide tumor vac- is that cancer cells may have a specific receptor to attach to
cination and checkpoint inhibition immunotherapy. The the lymphatic vessel inner wall and set up a residence. The
authors propose that serum VEGF-C may serve as a predic- lymphatic system draining various body organs is different,
tive biomarker for immunotherapy response [121]. some are more complex than other. Spread through the SLN
Interstitial flow and lymphatic drainage within the tumor can be best illustrated by melanoma and breast cancer, which
microenvironment are increased in the edge of the tumor follow more reliable routes in an orderly fashion [61] than
[122, 123]. The increased interstitial flow may facilitate can- the internal organs such as the lungs, stomach, pancreas, and
cer cells to migrate in the direction by chemotaxis from che- others, which show complicated lymphatic drainage pat-
mokines secreted by cancer cells [120]. Interstitial flow may terns. Certainly, melanoma and breast cancer allow us to
also affect stromal cells, resulting in cell and matrix align- appreciate the biology of early cancer arrival in the SLNs
ment [124, 125], increase fibroblast motility via matrix with minimal tumor burden. It has been shown that mela-
metalloproteinase-1 [126], and induce myofibroblast differ- noma patients with a negative SLN biopsy have low recur-
entiation via transforming growth factor (TGF)-b1 [124]. rence rate of 5–10% [59, 62].
Stromal cells in the cancer microenvironment including For breast cancer patients with a negative SLN biopsy,
fibroblasts produce TGF-beta 1. It transforms fibroblast to their recurrence rate is also very low, less than 5% [60].
myofibroblast (cancer-associated fibroblast; [127]), which Overall, melanoma or breast cancer patients with a negative
causes remodeling of fibroblasts. When cancer cell lines SLN biopsy fare much better than those with a positive SLN
were cocultured with dermal fibroblasts in a 3D collagen biopsy [63, 64]. Studying the biology of cancer trafficking
matrix, Shieh et al. were able to show that cancer invasion and lymph node metastasis in early melanoma and breast
was enhanced by fibroblasts when interstitial flow was pres- cancer may allow us to understand the mechanisms of cancer
ent. Thus, the authors describe a novel mechanism in which spread from the primary site to lymph nodes and beyond to
increased interstitial flow in the tumor microenvironment the distant sites. The routes of spread by the lymphatic ver-
may induce fibroblast remodeling and migration by TGF-­ sus the blood vessels will be addressed in other review arti-
beta1 activation and collagen degradation, thus, facilitating cles [69, 114]. The relative orderly progression of melanoma
tumor invasion into the lymphatic vessels [128]. Swartz and and breast cancer metastasis through the SLNs, in contrast to
Lund have linked lymphangiogenesis, interstitial flow, the complicated lymphatic system of the stomach, pancreas,
mechanobiology, and immunity within the cancer microen- and other organs, is like Mendel’s discovery of the indepen-
vironment [129]. Thus, according to the authors, while lym- dent segregation of the dominant and recessive genetic
phangiogenesis is associated with cancer invasion and poor alleles in the pea model with respect to seven features of the
prognosis, it may be of therapeutic benefit with checkpoint peas [130]. If Mendel would have used the human hair color
inhibition immunotherapy. Figure 21.13a, b summarizes the variations to study heredity, he would not have reached such
linkage between mechanobiology, lymphangiogenesis, and clear and discrete conclusions as there are multiple crossover
antitumor immunity. genes involving the heredity of hair color. Thus, the SLN
model for melanoma and breast cancer, being relatively sim-
ple, will help us to understand the cancer metastasis to the
21.10 Conclusion: The Lymphatic System SLN and its interaction within the lymphatic vessels.
as a Conduit for Cancer Metastasis Angiogenesis involving blood vessels induced by cancer
has been well established as a means for cancer to support its
The cancer cells exploit the lymphatic system to “hitchhike” growth and perhaps for it to gain into the circulation to
it to travel from the primary site as they dissociate them- spread to distant sites [71]. When melanoma transplants
selves from their microenvironment and enter the lymphatic were placed in the hamster cheek pouch, new blood vessels
vessels toward the lymph nodes and beyond to the distant were formed [72, 73]. Clinically angiogenesis has been
sites. As a cancer cell is slightly larger than a lymphocyte found to be associated with melanoma progression [74].
with an average diameter of 15 μm, the average diameter of Tumor microvessel density has been correlated with
a cancer cell may be up to 20 μm or more. The size of a sin- decreased disease free and overall survival in several studies
gle cancer cell can easily travel within the lymphatic vessel [76, 77]. On the other hand, other reports found no differ-
of 100 μm. Lymphatic vessels induced by cancer by means ences in tumor microvessel density in metastasizing and
of lymphangiogenesis may become dilated [105]. nonmetastasizing melanomas [78]. Thus, the prognostic
It is speculative that a cluster of cancer cells may become value of tumor hemangiogenesis in melanoma has remained
trapped in the lymphatic vessel at the valve of the lymphatic controversial [75] and the potential prognostic value of such
vessel [106] and eventually grow on the lymphatic vessel tumor angiogenesis in melanoma is not clear at this time
inner wall as a colony to invade the surrounding fat tissue as [79]. On the other hand, based on experimental and clinical
in-transit metastasis, as in melanoma. The other speculation data, cancer lymphangiogenesis has been found to be signifi-
224 S. P. Leong and M. H. Witte

a Efferent
lymph

TDLN

Afferent
High IFP Increased lymph
Tumour lymph flow
angiogenesis
Lymphangiogenesis

Lymphangiogenesis
Interstitial flow

b
Interstitial flow

Stromal ECM
Tumour antigens
alignment and
and cytokines
stiffening
Flow

TGFb Tumour DC, iDC Macrophage

cell or regDC or MDSC
Contraction VEGFA
Tumour
cell invasion VEGFC ECM TReg cell CAF

Fig. 21.13 (a and b) Linking mechanobiology, lymphangiogenesis, tion and remodeling that is dependent on transforming growth factor-β
and antitumor immunity. Lymphangiogenesis, mechanobiology, and (TGFβ). Matrix stiffening promotes tumor cell invasion through CAF-­
immunity in the tumor microenvironment contain many interdependent led collective migration and activation of mechanically sensitive stro-
features. (a) Fluid pathways in the tumor microenvironment are shown. mal components. Interestingly, stromal features of the tumor can mimic
Angiogenic, immature tumor vessels (shown in red) are hyperperme- those of the TDLN. Immunological changes include biasing of the
able, driving heightened interstitial fluid pressure (IFP) in the tumor. tumor-infiltrating lymphocyte populations by local factors. Tumor-­
This creates steep pressure gradients at the tumor margin that drive promoting cytokines promote alternatively activated (M2) macrophages
interstitial flow through the stroma and into peritumoral lymphatic ves- and myeloid-derived suppressor cells (MDSCs) hinder the maturation
sels (green), which are often expanded and hyperplastic. These lym- of immature dendritic cells (iDCs) and stimulate regulatory DCs (reg-
phatic vessels carry tumor interstitial fluid to the sentinel or DCs), which in turn promote regulatory T (TReg) cells and suppress
tumor-draining lymph node (TDLN), where lymphangiogenesis is also local cytotoxic T cell activity. Similar cell populations are biased in the
seen. (b) The heightened interstitial and lymphatic flows (green) in the TDLN. VEGF, vascular endothelial growth factor. Figure and legend
tumor microenvironment coincide with biomechanical (red) and immu- reprinted by permission from Springer Nature: Nature Reviews Cancer,
nological (purple) changes in the tumor stroma. Biomechanical changes Lymphatic and interstitial flow in the tumor microenvironment: Linking
include stiffening and alignment of the extracellular matrix (ECM) of mechanobiology with immunity [129]
the tumor stroma, owing to cancer-associated fibroblast (CAF) contrac-

cantly correlated with poor prognosis as presented in this Furthermore, the strong clinical data that SLN from mela-
chapter. Such strong evidence that poor prognosis is associ- noma and breast cancer are a significant prognostic bio-
ated with cancer lymphangiogenesis indicates that the lym- marker, supports the view that the lymphatic system may
phatic system may be the major conduit for cancer metastasis. serve as the major gateway for cancer to spread.
21 Lymphangiogenesis: Lymphatic System and Lymph Nodes; Cancer Lymphangiogenesis and Metastasis 225

References
Future Perspectives and Unanswered Questions
The concept and utility of SLNs have been well estab- 1. Asellius G. De lactibus sive lacteis venis. Milan: Mediolani; 1627.
2. Irschick R, Siemon C, Brenner E. The history of anatomical
lished in the clinical arena, particularly in the manage-
research of lymphatics — from the ancient times to the end of
ment of melanoma [107] and breast cancer [108] with the European renaissance. Ann Anat Anatomischer Anzeiger.
SLN surgery as a reliable staging modality for the 2019;223:49–69. https://doi.org/10.1016/j.aanat.2019.01.010.
regional nodal basin [32, 114]. The microenvironment 3. Harvey W. Exercitation Anatomica de Motu Cordis et Sanguinis
in Animalibus. Frankfurt; 1628.
with the SLN in relationship to the cancer growth and
4. Sabin FR. On the origin of the lymphatic system from the veins
spread awaits further study [131]. and the development of the lymph hearts and thoracic duct in the
Cancer lymphangiogenesis in the primary site and pig. Am J Anat. 1902;1(3):367–89.
the tumor-draining node or the SLN has been well estab- 5. Hill MA (2019) Embryology lymph node development. https://
embryology.med.unsw.edu.au/embryology/index.php/Lymph_
lished as a poor prognosis marker [132] with detailed dis-
Node_Development. Accessed June 15, 2021.
cussion in this chapter. Molecular mechanisms of cancer 6. Kampmeier OF. Evolution and comparative morphology of the
spread to the SLN and then to the distant sites are being lymphatic system. Springfield: Charles C. Thomas; 1969.
pursued in many cancer centers. The release of factors by 7. Huntington GS. The development of the mammalian jugular
lymphsac, of the tributary primitive ulnar lymphatic and of the
the primary tumor to the SLN to render it a pre-metastatic
thoracic ducts from the view point of recent investigations of ver-
niche is gaining evidence. A recent review article on pro- tebrate lymphatic ontogeny, together with a consideration of the
gressive role of the breast cancer SLN with anticancer genetic relations of lymphatic and haemal vascular channels in the
immunity to develop into a lymph node with immuno- embryos of amniotes. Am J Anat. 1914;16(3):259–316.
8. Senger DR, Galli SJ, Dvorak AM, Perruzzi CA, Harvey
suppressive microenvironment allowing the cancer cells
VS, Dvorak HF. Tumor cells secrete a vascular permeabil-
to migrate to the SLN and grow in this immunosuppres- ity factor that promotes accumulation of ascites fluid. Science.
sive state is intriguing [133]. 1983;219(4587):983–5. https://doi.org/10.1126/science.6823562.
9. Keck PJ, Hauser SD, Krivi G, Sanzo K, Warren T, Feder J,
Connolly DT. Vascular permeability factor, an endothelial cell
• When does the SLN has the capacity to kill cancer
mitogen related to PDGF. Science. 1989;246(4935):1309–12.
cells with anticancer immunity? https://doi.org/10.1126/science.2479987.
• When does the SLN become immunosuppressive? 10. Leung DW, Cachianes G, Kuang WJ, Goeddel DV, Ferrara
• What causes such a change? N. Vascular endothelial growth factor is a secreted angio-
genic mitogen. Science. 1989;246(4935):1306–9. https://doi.
• Does it mean that when a SLN harbors cancer cells,
org/10.1126/science.2479986.
the lymph node microenvironment has already 11. Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its
become immunosuppressive? receptors. Nat Med. 2003;9(6):669–76. https://doi.org/10.1038/
• What is the contribution of T cell repertoire of each nm0603-­669.
12. Holmes DI, Zachary I. The vascular endothelial growth factor
patient in such changes in the SLN microenviron-
(VEGF) family: angiogenic factors in health and disease. Genome
ment? Biol. 2005;6(2):209. https://doi.org/10.1186/gb-­2005-­6-­2-­209.
• What is the spectrum of heterogeneity with respect 13. Li X, Eriksson U. Novel VEGF family members: VEGF-B,
to cancer both within the cancer population of each VEGF-C and VEGF-D. Int J Biochem Cell Biol. 2001;33(4):421–
6. https://doi.org/10.1016/s1357-­2725(01)00027-­9.
patient and between patients?
14. Shibuya M. Vascular endothelial growth factor receptor-2: its
• What is the spectrum of heterogeneity in the T cell unique signaling and specific ligand, VEGF-E. Cancer Sci.
repertoire of each patient? 2003;94(9):751–6. https://doi.org/10.1111/j.1349-­7006.2003.
• What is the dynamic interaction between cancer tb01514.x.
15. Rauniyar K, Jha SK, Jeltsch M. Biology of vascular endothe-
cells and T cells in the background of such hetero-
lial growth factor C in the morphogenesis of lymphatic vessels.
geneity? Front Bioeng Biotechnol. 2018;6:7. https://doi.org/10.3389/
fbioe.2018.00007.
Perhaps, elucidation of molecular mechanisms of 16. Banerji S, Ni J, Wang S-X, Clasper S, Su J, Tammi R, Jones M,
Jackson DG. LYVE-1, a new homologue of the CD44 glycopro-
these changes can explain the conundrum that a SLN
tein, is a lymph-specific receptor for hyaluronan. J Cell Biol.
while it should be protective against the invasion 1999;144(4):789–801.
of cancer cells to the lymph node, but then, it may 17. Carreira CM, Nasser SM, di Tomaso E, Padera TP, Boucher Y,
become the gateway for distant metastasis with the Tomarev SI, Jain RK. LYVE-1 is not restricted to the lymph
vessels: expression in normal liver blood sinusoids and down-­
development of an immunosuppressive microenviron-
regulation in human liver cancer and cirrhosis. Cancer Res.
ment to allow the arrival and subsequent proliferation 2001;61(22):8079–84.
of cancer cells in the lymph node to spread. Under- 18. Gordon EJ, Gale NW, Harvey NL. Expression of the hyaluronan
standing these important molecular mechanisms of receptor LYVE-1 is not restricted to the lymphatic vasculature;
LYVE-1 is also expressed on embryonic blood vessels. Dev Dyn.
lymph node metastasis may allow us to develop new
2008;237(7):1901–9. https://doi.org/10.1002/dvdy.21605.
strategy and novel therapy in the treatment of cancer 19. Maruyama K, Asai J, Ii M, Thorne T, Losordo DW, D'Amore
metastasis. PA. Decreased macrophage number and activation lead to reduced
lymphatic vessel formation and contribute to impaired diabetic
226 S. P. Leong and M. H. Witte

wound healing. Am J Pathol. 2007;170(4):1178–91. https://doi. Mariani G, Vidal-Sicart S, Valdes Olmos RA, editors. Atlas of
org/10.2353/ajpath.2007.060018. lymphoscintigraphy and sentinel node mapping: a pictorial case-
20. Breiteneder-Geleff S, Matsui K, Soleiman A, Meraner P, Poc- based approach. 2nd ed. Cham, Switzerland: Springer Nature
zewski H, Kalt R, Schaffner G, Kerjaschki D. Podoplanin, Switzerland AG; 2020. p. 1–6. https://doi.org/10.1007/978-­3-­
novel 43-kd membrane protein of glomerular epithelial cells, 030-­45296-­4.
is down-­ regulated in puromycin nephrosis. Am J Pathol. 38. Uthman E (2008) Normal lymph node. https://www.flickr.com/
1997;151(4):1141–52. photos/euthman/3113417731/. Accessed August 20 2021.
21. Oliver G, Sosa-Pineda B, Geisendorf S, Spana EP, Doe CQ, Gruss 39. Graham T (2021) Lymph nodes 101. https://www.iheartpathology.
P. Prox 1, a prospero-related homeobox gene expressed during net/post/lymph-­nodes-­101. Accessed August 20 2021.
mouse development. Mech Dev. 1993;44(1):3–16. https://doi. 40. Lymphatics Lab. (2021). http://medcell.med.yale.edu/systems_
org/10.1016/0925-­4773(93)90012-­m. cell_biology/lymphatics_lab.php. Accessed August 20 2021.
22. Wigle JT, Oliver G. Prox1 function is required for the develop- 41. Thomas D (2012) HistoQuarterly: lymph node. https://histolo-
ment of the murine lymphatic system. Cell. 1999;98(6):769–78. gyblog.com/2012/04/30/histoquarterly-­lymph-­node/. Accessed
https://doi.org/10.1016/s0092-­8674(00)81511-­1. August 20 2021.
23. Pan Y, Wang WD, Yago T. Transcriptional regulation of podo- 42. Gentek R, Bajenoff M. Lymph node stroma dynamics
planin expression by Prox1 in lymphatic endothelial cells. and approaches for their visualization. Trends Immunol.
Microvasc Res. 2014;94:96–102. https://doi.org/10.1016/j. 2017;38(4):236–47. https://doi.org/10.1016/j.it.2017.01.005.
mvr.2014.05.006. 43. Fletcher AL, Acton SE, Knoblich K. Lymph node fibroblas-
24. Hong YK, Detmar M. Prox1, master regulator of the lymphatic tic reticular cells in health and disease. Nat Rev Immunol.
vasculature phenotype. Cell Tissue Res. 2003;314(1):85–92. 2015;15(6):350–61. https://doi.org/10.1038/nri3846.
https://doi.org/10.1007/s00441-­003-­0747-­8. 44. Alvarenga HG, Marti L. Multifunctional roles of reticular
25. Hong YK, Harvey N, Noh YH, Schacht V, Hirakawa S, Detmar M, fibroblastic cells: more than meets the eye? J Immunol Res.
Oliver G. Prox1 is a master control gene in the program specify- 2014;2014:402038. https://doi.org/10.1155/2014/402038.
ing lymphatic endothelial cell fate. Dev Dyn. 2002;225(3):351–7. 45. Liu K, Nussenzweig MC. Origin and development of dendritic
https://doi.org/10.1002/dvdy.10163. cells. Immunol Rev. 2010;234(1):45–54. https://doi.org/10.1111/
26. Kaipainen A, Korhonen J, Mustonen T, van Hinsbergh VW, j.0105-­2896.2009.00879.x.
Fang GH, Dumont D, Breitman M, Alitalo K. Expression of the 46. Hervouet C, Luci C, Bekri S, Juhel T, Bihl F, Braud VM, Czerkin-
fms-like tyrosine kinase 4 gene becomes restricted to lymphatic sky C, Anjuere F. Antigen-bearing dendritic cells from the sublin-
endothelium during development. Proc Natl Acad Sci U S A. gual mucosa recirculate to distant systemic lymphoid organs to
1995;92(8):3566–70. https://doi.org/10.1073/pnas.92.8.3566. prime mucosal CD8 T cells. Mucosal Immunol. 2014;7(2):280–
27. Wilting J, Papoutsi M, Schneider M, Christ B. The lymphatic 91. https://doi.org/10.1038/mi.2013.45.
endothelium of the avian wing is of somitic origin. Dev Dyn. 47. Witte MH, Jones K, Wilting J, Dictor M, Selg M, McHale N,
2000a;217(3):271–8. https://doi.org/10.1002/(SICI)1097-­ Gershenwald JE, Jackson DG. Structure function relationships in
0177(200003)217:3<271::AID-­DVDY5>3.0.CO;2-­2. the lymphatic system and implications for cancer biology. Cancer
28. Wilting J, Schneider M, Papoutski M, Alitalo K, Christ B. An Metastasis Rev. 2006;25(2):159–84.
avian model for studies of embryonic lymphangiogenesis. Lym- 48. Bernas M, Thiadens SRJ, Stewart P, Granzow J. Secondary
phology. 2000b;33(3):81–94. lymphedema from cancer therapy. Clin Exp Metastasis. 2021;
29. Witte MH, Bernas MJ, Martin CP, Witte CL. Lymphangiogenesis https://doi.org/10.1007/s10585-­021-­10096-­w.
and lymphangiodysplasia: from molecular to clinical lymphology. 49. Miyasaka M, Tanaka T. Lymphocyte trafficking across high
Microsc Res Tech. 2001;55(2):122–45. https://doi.org/10.1002/ endothelial venules: dogmas and enigmas. Nat Rev Immunol.
jemt.1163. 2004;4(5):360–70.
30. Witte M, Witte C, Ohhashi T. Are lymphatics different from blood 50. Halsted WS. I. the results of operations for the cure of cancer of
vessels. Lymphology. 1987;20(4):169. the breast performed at the Johns Hopkins Hospital from June,
31. Witte MH, Bernas MJ (2016) Lymphatic pathophysiology. https:// 1889, to January, 1894. Ann Surg. 1894;20(5):497–555. https://
thoracickey.com/lymphatic-­pathophysiology/. Accessed July 27 doi.org/10.1097/00000658-­189407000-­00075.
2021. 51. Neuhaus SJ, Clark MA, Thomas JM. Dr. Herbert Lumley Snow
32. Leong SP, Pissas A, Scarato M, Gallon F, Pissas MH, Amore M, (1847–1930): the original champion of elective lymph node dis-
Wu M, Faries M, Lund A. The lymphatic system and sentinel section in melanoma. Ann Surg Oncol. 2004;11(9):875–8. https://
lymph nodes: conduit for cancer metastasis. Clin Exp Metasta- doi.org/10.1245/ASO.2004.02.031.
sis. 2021. https://doi.org/10.1007/s10585-021-10123-w. PMID: 52. Snow H. Melanotic cancerous disease. Lancet. 1892;2:872–4.
34651243. 53. Haagensen CD, Feind CR, Herter FP. The lymphatics in cancer.
33. FuseSchool (2019) The lymphatic system. YouTube.com, https:// Philadelphia: WB Saunders Co; 1972.
www.youtube.com/watch?v=o0-­1OknbO3M 54. Fisher B, Ravdin RG, Ausman RK, Slack NH, Moore GE,
34. Breslin JW, Yang Y, Scallan JP, Sweat RS, Adderley SP, Murfee Noer RJ. Surgical adjuvant chemotherapy in cancer of the
WL. Lymphatic vessel network structure and physiology. breast: results of a decade of cooperative investigation. Ann
Compr Physiol. 2018;9(1):207–99. https://doi.org/10.1002/cphy. Surg. 1968;168(3):337–56. https://doi.org/10.1097/00000658-­
c180015. 196809000-­00004.
35. Stacker SA, Williams SP, Karnezis T, Shayan R, Fox SB, 55. Cady B. Lymph node metastases. Indicators, but not gover-
Achen MG. Lymphangiogenesis and lymphatic vessel remodel- nors of survival. Arch Surg. 1984;119(9):1067–72. https://doi.
ling in cancer. Nat Rev Cancer. 2014;14(3):159–72. https://doi. org/10.1001/archsurg.1984.01390210063014.
org/10.1038/nrc3677. 56. Hellman S, Weichselbaum RR. Oligometastases. J Clin Oncol.
36. Dellinger MT, Witte MH. Lymphangiogenesis, lymphatic sys- 1995;13(1):8–10. https://doi.org/10.1200/JCO.1995.13.1.8.
temomics, and cancer: context, advances and unanswered 57. Cabanas RM. An approach for the treatment of penile carci-
­questions. Clin Exp Metastasis. 2018;35(5–6):419–24. https://doi. noma. Cancer. 1977;39(2):456–66. https://doi.org/10.1002/1097-­
org/10.1007/s10585-­018-­9907-­9. 0142(197702)39:2<456::aid-­cncr2820390214>3.0.co;2-­i.
37. Nieweg OE, Tanis PJ, Leong SP. Anatomy and physiology of 58. Morton DL, Wen DR, Wong JH, Economou JS, Cagle LA,
lymphatic circulation: application to lymphatic mapping. In: Storm FK, Foshag LJ, Cochran AJ. Technical details of intra-
21 Lymphangiogenesis: Lymphatic System and Lymph Nodes; Cancer Lymphangiogenesis and Metastasis 227

operative lymphatic mapping for early stage melanoma. noma. N Engl J Med. 1991;324(1):1–8. https://doi.org/10.1056/
Arch Surg. 1992;127(4):392–9. https://doi.org/10.1001/arch- NEJM199101033240101.
surg.1992.01420040034005. 72. Hubler WR Jr, Wolf JE Jr. Melanoma. Tumor angiogenesis
59. Morton DL, Hoon DS, Cochran AJ, Turner RR, Essner R, Takeu- and human neoplasia. Cancer. 1976;38(1):187–92. https://doi.
chi H, Wanek LA, Glass E, Foshag LJ, Hsueh EC, Bilchik AJ, org/10.1002/1097-­0142(197607)38:1<187::aid-­cncr2820380129
Elashoff D, Elashoff R. Lymphatic mapping and sentinel lymph- >3.0.co;2-­d.
adenectomy for early-stage melanoma: therapeutic utility and 73. Warren B, Shubik P. The growth of the blood supply to mela-
implications of nodal microanatomy and molecular staging for noma transplants in the hamster cheek pouch. Transplantation.
improving the accuracy of detection of nodal micrometastases. 1966;4(5):652.
Ann Surg. 2003;238(4):538–49.; discussion 549–550. https://doi. 74. Barnhill RL, Fandrey K, Levy MA, Mihm MC Jr, Hyman
org/10.1097/01.sla.0000086543.45557.cb. B. Angiogenesis and tumor progression of melanoma. Quantifica-
60. Kapoor NS, Sim MS, Lin J, Giuliano AE. Long-term outcome of tion of vascularity in melanocytic nevi and cutaneous malignant
patients managed with sentinel lymph node biopsy alone for node-­ melanoma. Lab Investig. 1992;67(3):331–7.
negative invasive breast cancer. Arch Surg. 2012;147(11):1047– 75. Streit M, Detmar M. Angiogenesis, lymphangiogenesis, and mel-
52. https://doi.org/10.1001/archsurg.2012.1563. anoma metastasis. Oncogene. 2003;22(20):3172–9. https://doi.
61. Reintgen D, Cruse CW, Wells K, Berman C, Fenske N, Glass F, org/10.1038/sj.onc.1206457.
Schroer K, Heller R, Ross M, Lyman G. The orderly progression 76. Srivastava A, Laidler P, Davies RP, Horgan K, Hughes LE. The
of melanoma nodal metastases. Ann Surg. 1994;220(6):759–67. prognostic significance of tumor vascularity in intermediate-­
https://doi.org/10.1097/00000658-­199412000-­00009. thickness (0.76–4.0 mm thick) skin melanoma. A quantitative his-
62. Thomas DC, Han G, Leong SP, Kashani-Sabet M, Vetto J, Pockaj tologic study. Am J Pathol. 1988;133(2):419.
B, White RL, Faries MB, Schneebaum S, Mozzillo N, Charney KJ, 77. Straume O, Salvesen HB, Akslen LA. Angiogenesis is prognosti-
Sondak VK, Messina JL, Zager JS, Han D. Recurrence of mela- cally important in vertical growth phase melanomas. Int J Oncol.
noma after a negative sentinel node biopsy: predictors and impact 1999;15(3):595–9. https://doi.org/10.3892/ijo.15.3.595.
of recurrence site on survival. Ann Surg Oncol. 2019;26(7):2254– 78. Busam KJ, Berwick M, Blessing K, Fandrey K, Kang S, Karaoli
62. https://doi.org/10.1245/s10434-­019-­07369-­w. T, Fine J, Cochran AJ, White WL, Rivers J. Tumor vascularity is
63. Andersson Y, Frisell J, Sylvan M, de Boniface J, Bergkvist not a prognostic factor for malignant melanoma of the skin. Am J
L. Breast cancer survival in relation to the metastatic tumor bur- Pathol. 1995;147(4):1049.
den in axillary lymph nodes. J Clin Oncol. 2010;28(17):2868–73. 79. Dadras SS, Paul T, Bertoncini J, Brown LF, Muzikansky A, Jackson
https://doi.org/10.1200/JCO.2009.24.5001. DG, Ellwanger U, Garbe C, Mihm MC, Detmar M. Tumor lym-
64. Rios-Cantu A, Lu Y, Melendez-Elizondo V, Chen M, Gutierrez-­ phangiogenesis: a novel prognostic indicator for cutaneous mela-
Range A, Fadaki N, Thummala S, West-Coffee C, Cleaver J, noma metastasis and survival. Am J Pathol. 2003;162(6):1951–60.
Kashani-Sabet M, Leong SP. Is the non-sentinel lymph node com- https://doi.org/10.1016/S0002-­9440(10)64328-­3.
partment the next site for melanoma progression from the sentinel 80. Kaplan RN, Riba RD, Zacharoulis S, Bramley AH, Vincent L, Costa
lymph node compartment in the regional nodal basin? Clin Exp C, MacDonald DD, Jin DK, Shido K, Kerns SA, Zhu Z, Hicklin
Metastasis. 2017;34(5):345–50. https://doi.org/10.1007/s10585-­ D, Wu Y, Port JL, Altorki N, Port ER, Ruggero D, Shmelkov SV,
017-­9854-­x. Jensen KK, Rafii S, Lyden D. VEGFR1-positive haematopoietic
65. van Akkooi AC, de Wilt JH, Verhoef C, Schmitz PI, van Geel bone marrow progenitors initiate the pre-metastatic niche. Nature.
AN, Eggermont AM, Kliffen M. Clinical relevance of melanoma 2005;438(7069):820–7. https://doi.org/10.1038/nature04186.
micrometastases (<0.1 mm) in sentinel nodes : Are these nodes 81. Sleeman JP. The metastatic niche and stromal progression. Cancer
to be considered negative? Ann Oncol. 2006;17(10):1578–85. Metastasis Rev. 2012;31(3–4):429–40. https://doi.org/10.1007/
https://doi.org/10.1093/annonc/mdl176. s10555-­012-­9373-­9.
66. van Akkooi AC, Nowecki ZI, Voit C, Schäfer-Hesterberg G, 82. Sleeman JP. The lymph node pre-metastatic niche. J Mol Med
Michej W, de Wilt JH, Rutkowski P, Verhoef C, Eggermont (Berl). 2015;93(11):1173–84. https://doi.org/10.1007/s00109-­
AM. Sentinel node tumor burden according to the Rotterdam cri- 015-­1351-­6.
teria is the most important prognostic factor for survival in mela- 83. Paget S. The distribution of secondary growths in cancer of the
noma patients: a multicenter study in 388 patients with positive breast. Lancet. 1889;133(3421):571–3.
sentinel nodes. Ann Surg. 2008;248(6):949–55. 84. Cochran AJ, Huang RR, Lee J, Itakura E, Leong SP, Essner
67. Reed J, Rosman M, Verbanac KM, Mannie A, Cheng Z, Tafra R. Tumour-induced immune modulation of sentinel lymph nodes.
L. Prognostic implications of isolated tumor cells and microme- Nat Rev Immunol. 2006;6(9):659–70. https://doi.org/10.1038/
tastases in sentinel nodes of patients with invasive breast cancer: nri1919.
10-year analysis of patients enrolled in the prospective East Caro- 85. Cochran AJ, Wen DR, Farzad Z, Stene MA, McBride W, Lana
lina University/Anne Arundel medical center sentinel node mul- AM, Hoon DS, Morton DL. Immunosuppression by melanoma
ticenter study. J Am Coll Surg. 2009;208(3):333–40. https://doi. cells as a factor in the generation of metastatic disease. Anticancer
org/10.1016/j.jamcollsurg.2008.10.036. Res. 1989;9(4):859–64.
68. Special Issue on: The 2017 7th International Cancer Metastasis 86. Hoon DS, Bowker RJ, Cochran AJ. Suppressor cell activity in
Symposium; Cancer Metastasis through the Lymphovascular melanoma-­draining lymph nodes. Cancer Res. 1987;47(6):1529–33.
System: Biology and Treatment (2018). Volume 35; Number 87. Huang RR, Wen DR, Guo J, Giuliano AE, Nguyen M, Offo-
5–6; June–August 2018 edn. Springer, Clinical and Experimental dile R, Stern S, Turner R, Cochran AJ. Selective modulation of
Metastasis. paracortical dendritic cells and T-lymphocytes in breast cancer
69. Nathanson SD, Detmar M, Padera TP, Yates LR, Welch DR, sentinel lymph nodes. Breast J. 2000;6(4):225–32. https://doi.
Beadnell TC, Scheid AD, Wrenn ED, Cheung K. Mechanisms of org/10.1046/j.1524-­4741.2000.98114.x.
breast cancer metastasis. Clin Exp Metastasis. 2021; https://doi. 88. Matsuura K, Yamaguchi Y, Ueno H, Osaki A, Arihiro K, Toge
org/10.1007/s10585-­021-­10090-­2. T. Maturation of dendritic cells and T-cell responses in senti-
70. Zhou H, P-j L, Padera TP. Progression of metastasis through lym- nel lymph nodes from patients with breast carcinoma. Cancer.
phatic system. Cell. 2021;10(3):627. 2006;106(6):1227–36. https://doi.org/10.1002/cncr.21729.
71. Weidner N, Semple JP, Welch WR, Folkman J. Tumor angio- 89. Mohos A, Sebestyén T, Liszkay G, Plótár V, Horváth S, Gaudi
genesis and metastasis--correlation in invasive breast carci- I, Ladányi A. Immune cell profile of sentinel lymph nodes in
228 S. P. Leong and M. H. Witte

patients with malignant melanoma – FOXP3+ cell density in 107. Faries MB, Thompson JF, Cochran AJ, Andtbacka RH, Mozzillo
cases with positive sentinel node status is associated with unfavor- N, Zager JS, Jahkola T, Bowles TL, Testori A, Beitsch PD, Hoek-
able clinical outcome. J Transl Med. 2013;11(1):43. https://doi. stra HJ, Moncrieff M, Ingvar C, Wouters M, Sabel MS, Levine
org/10.1186/1479-­5876-­11-­43. EA, Agnese D, Henderson M, Dummer R, Rossi CR, Neves RI,
90. Botella-Estrada R, Dasi F, Ramos D, Nagore E, Herrero MJ, Trocha SD, Wright F, Byrd DR, Matter M, Hsueh E, MacKenzie-
Gimenez J, Fuster C, Sanmartin O, Guillen C, Alino S. Cyto- Ross A, Johnson DB, Terheyden P, Berger AC, Huston TL, Wayne
kine expression and dendritic cell density in melanoma sen- JD, Smithers BM, Neuman HB, Schneebaum S, Gershenwald JE,
tinel nodes. Melanoma Res. 2005;15(2):99–106. https://doi. Ariyan CE, Desai DC, Jacobs L, McMasters KM, Gesierich A,
org/10.1097/00008390-­200504000-­00003. Hersey P, Bines SD, Kane JM, Barth RJ, McKinnon G, Farma JM,
91. Leong SP, Peng M, Zhou YM, Vaquerano JE, Chang JW. Cyto- Schultz E, Vidal-Sicart S, Hoefer RA, Lewis JM, Scheri R, Kelley
kine profiles of sentinel lymph nodes draining the primary mela- MC, Nieweg OE, Noyes RD, Hoon DSB, Wang HJ, Elashoff DA,
noma. Ann Surg Oncol. 2002;9(1):82–7. https://doi.org/10.1245/ Elashoff RM. Completion dissection or observation for sentinel-­
aso.2002.9.1.82. node metastasis in melanoma. N Engl J Med. 2017;376(23):2211–
92. Cochran AJ, Huang RR, Su A, Itakura E, Wen DR. Is sentinel 22. https://doi.org/10.1056/NEJMoa1613210.
node susceptibility to metastases related to nodal immune mod- 108. Giuliano AE. The evolution of sentinel node biopsy for breast can-
ulation? Cancer J. 2015;21(1):39–46. https://doi.org/10.1097/ cer: personal experience. Breast J. 2020;26(1):17–21. https://doi.
PPO.0000000000000094. org/10.1111/tbj.13729.
93. Takeuchi H, Fujimoto A, Tanaka M, Yamano T, Hsueh E, Hoon 109. Skobe M, Hawighorst T, Jackson D, Prevo R, Janes L, Velasco P,
DS. CCL21 chemokine regulates chemokine receptor CCR7 bear- Riccardi L, Alitalo K, Claffey K, Detmar M. Induction of tumor
ing malignant melanoma cells. Clin Cancer Res. 2004;10(7):2351– lymphangiogenesis by VEGF-C promotes breast cancer metasta-
8. https://doi.org/10.1158/1078-­0432.ccr-­03-­0195. sis. Nat Med. 2001;7(2):192–8.
94. Thomas SN, Vokali E, Lund AW, Hubbell JA, Swartz MA. Tar- 110. Huang C, Chen Y. Lymphangiogenesis and colorectal cancer.
geting the tumor-draining lymph node with adjuvanted Saudi Med J. 2017;38(3):237–44. https://doi.org/10.15537/
nanoparticles reshapes the anti-tumor immune response. Bioma- smj.2017.3.16245.
terials. 2014;35(2):814–24. https://doi.org/10.1016/j.biomateri- 111. Sundlisaeter E, Dicko A, Sakariassen PO, Sondenaa K, Enger PO,
als.2013.10.003. Bjerkvig R. Lymphangiogenesis in colorectal cancer--prognostic
95. Witte MH, Way DL, Witte CL, Bernas M. Lymphangiogen- and therapeutic aspects. Int J Cancer. 2007;121(7):1401–9. https://
esis: mechanisms, significance and clinical implications. EXS. doi.org/10.1002/ijc.22996.
1997;79:65–112. https://doi.org/10.1007/978-­3-­0348-­9006-­9_5. 112. Sun JG, Wang Y, Chen ZT, Zhuo WL, Zhu B, Liao RX, Zhang
96. Bowman CA, Witte MH, Witte CL, Way DL, Nagle RB, Copeland SX. Detection of lymphangiogenesis in non-small cell lung can-
JG, Daschbach CC. Cystic hygroma reconsidered: hamartoma or cer and its prognostic value. J Exp Clin Cancer Res. 2009;28:21.
neoplasm? Primary culture of an endothelial cell line from a mas- https://doi.org/10.1186/1756-­9966-­28-­21.
sive cervicomediastinal hygroma with bony lymphangiomatosis. 113. Werynska B, Dziegiel P, Jankowska R. Role of lymphangiogene-
Lymphology. 1984;17(1):15–22. sis in lung cancer. Folia Histochem Cytobiol. 2009;47(3):333–42.
97. Gnepp DR, Chandler W. Tissue culture of human and canine tho- https://doi.org/10.2478/v10042-­009-­0090-­3.
racic duct endothelium. In Vitro Cell Dev Biol. 1985;21(4):200–6. 114. Leong SP, Naxerova K, Keller L, Pantel K, Witte M. Molecular
https://doi.org/10.1007/BF02620930. mechanisms of cancer metastasis via the lymphatic versus the vas-
98. Johnston MG, Walker MA. Lymphatic endothelial and smooth-­ cular channels. Clin Exp Metastasis. in press-a;
muscle cells in tissue culture. In Vitro. 1984;20(7):566–72. https:// 115. Dieterich LC, Ikenberg K, Cetintas T, Kapaklikaya K, Hutmacher
doi.org/10.1007/BF02639772. C, Detmar M. Tumor-associated lymphatic vessels upregulate
99. Leak LV, Jones M. Lymphangiogenesis in vitro: formation of PDL1 to inhibit T-cell activation. Front Immunol. 2017;8:66.
lymphatic capillary-like channels from confluent monolay- https://doi.org/10.3389/fimmu.2017.00066.
ers of lymphatic endothelial cells. In Vitro Cell Dev Biol Anim. 116. Commerford CD, Dieterich LC, He Y, Hell T, Montoya-Zegarra
1994;30A(8):512–8. JA, Noerrelykke SF, Russo E, Rocken M, Detmar M. Mechanisms
100. Way D, Hendrix M, Witte M, Witte C, Nagle R, Davis J. Lym- of tumor-induced Lymphovascular niche formation in draining
phatic endothelial cell line (CH3) from a recurrent retroperito- lymph nodes. Cell Rep. 2018;25(13):3554–63. e3554. https://doi.
neal lymphangioma. In Vitro Cell Dev Biol. 1987;23(9):647–52. org/10.1016/j.celrep.2018.12.002.
https://doi.org/10.1007/BF02621074. 117. Ma Q, Dieterich LC, Ikenberg K, Bachmann SB, Mangana J,
101. Folkman J. Clinical applications of research on angiogenesis. N Proulx ST, Amann VC, Levesque MP, Dummer R, Baluk P,
Engl J Med. 1995;333(26):1757–63. McDonald DM, Detmar M. Unexpected contribution of lymphatic
102. Folkman J. Tumor angiogenesis and tissue factor. Nat Med. vessels to promotion of distant metastatic tumor spread. Sci Adv.
1996;2(2):167–8. https://doi.org/10.1038/nm0296-­167. 2018;4(8):eaat4758. https://doi.org/10.1126/sciadv.aat4758.
103. van Netten JP, Cann SA, van der Westhuizen NG. Angiogenesis 118. Shields JD, Kourtis IC, Tomei AA, Roberts JM, Swartz MA. Induc-
and tumor growth. N Engl J Med. 1996;334(14):920–1. author tion of lymphoidlike stroma and immune escape by tumors that
reply 921 express the chemokine CCL21. Science. 2010;328(5979):749–52.
104. Witte MH, Witte CL. On tumor (and other) lymphangiogenesis. https://doi.org/10.1126/science.1185837.
Lymphology. 1997;30(1):1–2. 119. Issa A, Le TX, Shoushtari AN, Shields JD, Swartz MA. Vascu-
105. He Y, Rajantie I, Pajusola K, Jeltsch M, Holopainen T, Yla-­ lar endothelial growth factor-C and C-C chemokine receptor 7 in
Herttuala S, Harding T, Jooss K, Takahashi T, Alitalo K. Vascular tumor cell-lymphatic cross-talk promote invasive phenotype. Can-
endothelial cell growth factor receptor 3-mediated activation of cer Res. 2009;69(1):349–57. https://doi.org/10.1158/0008-­5472.
lymphatic endothelium is crucial for tumor cell entry and spread CAN-­08-­1875.
via lymphatic vessels. Cancer Res. 2005;65(11):4739–46. https:// 120. Shields JD, Fleury ME, Yong C, Tomei AA, Randolph GJ, Swartz
doi.org/10.1158/0008-­5472.CAN-­04-­4576. MA. Autologous chemotaxis as a mechanism of tumor cell hom-
106. Karpanen T, Alitalo K. Lymphatic vessels as targets of tumor ther- ing to lymphatics via interstitial flow and autocrine CCR7 signal-
apy? J Exp Med. 2001;194(6):F37–42. ing. Cancer Cell. 2007;11(6):526–38.
21 Lymphangiogenesis: Lymphatic System and Lymph Nodes; Cancer Lymphangiogenesis and Metastasis 229

121. Fankhauser M, Broggi MAS, Potin L, Bordry N, Jeanbart L, Lund Circ Physiol. 2009;297(4):H1225–34. https://doi.org/10.1152/
AW, Da Costa E, Hauert S, Rincon-Restrepo M, Tremblay C, ajpheart.00369.2009.
Cabello E, Homicsko K, Michielin O, Hanahan D, Speiser DE, 127. Shi X, Young CD, Zhou H, Wang X. Transforming growth
Swartz MA. Tumor lymphangiogenesis promotes T cell infiltra- factor-beta signaling in fibrotic diseases and cancer-associated
tion and potentiates immunotherapy in melanoma. Sci Transl fibroblasts. Biomol Ther. 2020;10(12) https://doi.org/10.3390/
Med. 2017;9(407) https://doi.org/10.1126/scitranslmed.aal4712. biom10121666.
122. Dafni H, Israely T, Bhujwalla ZM, Benjamin LE, Neeman 128. Shieh AC, Rozansky HA, Hinz B, Swartz MA. Tumor cell inva-
M. Overexpression of vascular endothelial growth factor 165 sion is promoted by interstitial flow-induced matrix priming by
drives peritumor interstitial convection and induces lymphatic stromal fibroblasts. Cancer Res. 2011;71(3):790–800.
drain: magnetic resonance imaging, confocal microscopy, and 129. Swartz MA, Lund AW. Lymphatic and interstitial flow in the
histological tracking of triple-labeled albumin. Cancer Res. tumour microenvironment: linking mechanobiology with immu-
2002;62(22):6731–9. nity. Nat Rev Cancer. 2012;12(3):210–9. https://doi.org/10.1038/
123. Harrell MI, Iritani BM, Ruddell A. Tumor-induced sentinel lymph nrc3186.
node lymphangiogenesis and increased lymph flow precede mela- 130. Mendel G (2016) Experiments on plant hybrids (1866). http://
noma metastasis. Am J Pathol. 2007;170(2):774–86. https://doi. www.bshs.org.uk/bshs-­translations/mendel. Accessed 6 May
org/10.2353/ajpath.2007.060761. 2021.
124. Ng CP, Hinz B, Swartz MA. Interstitial fluid flow induces myofi- 131. Kim R, Emi M, Tanabe K, Arihiro K. Immunobiology of the
broblast differentiation and collagen alignment in vitro. J Cell Sci. sentinel lymph node and its potential role for antitumour immu-
2005;118(Pt 20):4731–9. https://doi.org/10.1242/jcs.02605. nity. Lancet Oncol. 2006;7(12):1006–16. https://doi.org/10.1016/
125. Ng CP, Swartz MA. Fibroblast alignment under interstitial fluid S1470-­2045(06)70975-­5.
flow using a novel 3-D tissue culture model. Am J Physiol Heart 132. Hirakawa S. From tumor lymphangiogenesis to lymphvascular
Circ Physiol. 2003;284(5):H1771–7. https://doi.org/10.1152/ajp- niche. Cancer Sci. 2009;100(6):983–9. https://doi.org/10.1111/
heart.01008.2002. j.1349-­7006.2009.01142.x.
126. Shi ZD, Ji XY, Qazi H, Tarbell JM. Interstitial flow promotes vas- 133. Jana S, Muscarella R, Jones D. The multifaceted effects of breast
cular fibroblast, myofibroblast, and smooth muscle cell motility in cancer on tumor-draining lymph nodes. Am J Pathol. 2021;
3-D collagen I via upregulation of MMP-1. Am J Physiol Heart
 Immune Cell Trafficking
in the Lymphatics, Hyaluronan Biology 22
and Tumour Metastasis

David G. Jackson

Abstract
• The biological features of hyaluronan and how
As an integral part of the immune system, the lymphatics immune cells employ hyaluronan and its receptors
enable immune surveillance by shepherding the migration to enter lymphatic vessels and migrate to draining
of antigen-laden immune cells to draining lymph nodes lymph nodes.
(dLNs), as well as handling the clearance of macrophages • The distinctive endothelial junctions through which
and neutrophils from sites of injury and infection. In addi- immune cells enter lymphatics.
tion, the lymphatics play key roles as conduits for tumour • How tumours exploit hyaluronan and its receptors
metastasis. The molecular mechanisms controlling entry to metastasize via lymph.
to the lymphatics in these different contexts have yet to be
fully elucidated. However, it is clear that in most cases
they involve coordination between the processes of che-
motaxis and cell adhesion within the specialized endothe-
lial junctions of initial lymphatic capillaries for transit to 22.1 Introduction
the vessel lumen. A fundamental mediator of such adhe-
sion is the large matrix glycosaminoglycan hyaluronan, The lymphatics play multiple roles in maintaining tissue
and growing evidence indicates anchorage of hyaluronan homeostasis through the uptake and drainage of interstitial
to the surface of migrating cells and appropriate engage- fluids and the transport of electrolytes, lipids, cholesterol,
ment with LYVE-1, its’ primary receptor in lymphatic extracellular matrix and other dissolved macromolecules for
endothelium are critical to the entry process. The present appropriate degradation or delivery to the systemic blood
chapter outlines the main properties of hyaluronan and circulation [1]. Of equal importance, lymphatics are integral
how they support the trafficking of leucocytes for immu- to the immune system, insofar as they also act as conduits for
nity and inflammation. In addition, it discusses the evi- the migration of immune cells from tissues to the draining
dence that hyaluronan may also aid the invasion and lymph nodes for the generation, maintenance and modula-
systemic dissemination of metastasizing tumour cells. tion of antigen-specific immune responses [2]. In addition,
lymphatic vessels are involved in the timely clearance of
phagocytes, apoptotic cell bodies and tissue debris from
Learning Objectives injured and inflamed tissues after the processes of repair
• The main immune cell populations that traffic in and re-modelling. In the context of cancer, however, these
lymph and their functions in innate and adaptive otherwise beneficial properties of lymphatics are co-opted
immunity. by tumours that exploit them to support their own uncon-
trolled growth and metastasis, first to draining lymph nodes
and then to distant organs with inevitable consequences for
patient morbidity and mortality (Pathology of Lymph Node
Metastases, Messina, J.).
Despite their undisputed importance, the lymphatics have
D. G. Jackson (*) until recently been relegated to the role of Cinderellas, with
MRC Human Immunology Unit, MRC Weatherall Institute of most of the limelight being given to the blood vasculature,
Molecular Medicine, University of Oxford, Oxford, UK and how immune cells traffic and tumour cells disseminate
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 231
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_22
232 D. G. Jackson

within it (Hallmarks of Metastasis: Molecular Underpin- them eventually return to the circulation for continued sur-
nings, Gomez-Gelvez, J.). This situation is now changing veillance and maintenance of the memory T-cell pool [8].
and in the case of leucocytes, much has been learned about In addition, such migration may allow them to amplify and/
the mechanisms by which the various different populations or polarize immune responses in draining lymph nodes
of these cells enter and migrate within lymphatics and how through their capacity for cytokine release, or even enable
they are often distinct from the mechanisms used to extrava- their antigen responsiveness to be downmodulated or toler-
sate from blood vessels. In particular, evidence is emerging ized by T-cell immune checkpoint receptor ligands such as
for the fundamental involvement of the pro-adhesive extra- PDL-1 expressed in the endothelium of lymph node sinuses
cellular matrix glycosaminoglycan hyaluronan [3]. This and tumour-associated lymphatics (Immune microenvi-
ubiquitous molecule, which plays a well-documented role in ronment: Chen, T.). In addition, a significant proportion
both immune cell adhesion/migration and tumour metastasis (≤25%) of the CD4 population in normal afferent lymph are
in the blood vasculature via its primary receptor CD44, is dedicated FOXP3+ T regulatory cells (TREGs), an important
now also recognized to play a key role in immune cell entry lymphocyte subset that can suppress other T-cell responses,
to the lymphatics, by mediating adhesion to endothelium via and one that plays a key role in maintaining tolerance to
the related receptor LYVE-1 [2, 4, 5]. Indeed, preliminary self-antigens as well as the suppression of anti-tumour
evidence suggests HA and its receptors may also support responses. Following inflammation, and the resulting influx
tumour metastasis via the lymphatics. The present chapter of T-cells to the tissues from the circulation, the numbers
presents the current state of knowledge in this relatively of T-cells in afferent lymph increase, along with those of
new area of lymphatic biology. Starting with an overview certain innate immune cell populations as described sub-
of the main leucocyte populations that migrate in lymph and sequently.
the novel architecture of the lymphatic capillaries through After T-cells, the next most numerous leucocyte popula-
which they enter, I go on to describe the particular molecular tion in normal afferent lymph are DCs, the key mediators
mechanisms by which HA and its receptors mediate immune of immune surveillance that patrol tissues and endocytose,
cell trafficking and potentially tumour metastasis via the process and present antigens as bound MHC complexes
lymphatics. to stimulate T-cell immune responses [9]. Historically, the
migration of these cells has been tracked following their in
situ labelling by FITC skin painting, ex vivo labelling and
22.2 I mmune Cell Populations that Traffic adoptive transfer, or in specialized cases after microneedle
Via Lymph injection into lymphatic capillaries. Again, only small num-
bers of DCs migrate in afferent lymph under normal steady-
Within tissues, the continuous flow of lymph and the state conditions in non-inflamed tissues, partly to maintain
steady migration of immune cells through afferent capil- peripheral tolerance to self-antigens. However, in common
laries to immediate downstream lymph nodes creates a with TEM cells, the numbers of these DCs increase dramati-
local neighbourhood watch system that maintains uninter- cally in inflammation due to increased lymph flow and the
rupted ­surveillance for injury and the presence of microbial release of cytokines such as IL-1, TNFα and INFγ within the
or viral infection. In normal resting tissue, the numbers of tissues that induce DCs to mobilize, differentiate and upreg-
such cells are small, comprising almost exclusively T-cells ulate expression of the chemokine receptor CCR7 for vessel
and dendritic cells (DCs). Few if any are naive T lympho- entry (see below).
cytes, as these enter lymph nodes directly from the circu- Other leucocyte populations that may be present in sig-
lation via high endothelial venules, and the majority are nificant numbers in afferent lymph are macrophages and
instead antigen-experienced cells of the so-­called effector neutrophils. Recruited to tissues from the blood circulation
memory TEM subset. Having originally entered the tissues in response to injury or infection, these innate immune cells
in search of cognate antigen, the majority of these previ- phagocytose tissue debris and repair and re-model the extra-
ously primed and activated TEM cells then migrate via affer- cellular matrix (ECM). The short-lived neutrophil popula-
ent lymphatics to the draining lymph nodes, licensed by tion subsequently undergoes programmed cell death, and the
their selective expression of CCR7, the receptor for the apoptotic corpses are taken up by macrophages through effe-
critical lymph-tropic chemokine CCL21 [6, 7]. They con- rocytosis and cleared from the resolving tissue sites via affer-
stitute predominantly CD4 cells that provide help for B cell ent lymph. In addition, some neutrophils that phagocytose
antibody responses as opposed to CD8 cells that engage protozoal parasites and bacteria including Mycobacterium
in target killing and which largely remain within the tis- bovis BCG have extended tissue survival, and can migrate to
sues as immotile CCR7−ve resident memory (TRM) cells. The dLNs more rapidly than either T-cells or macrophages where
passage of TEM cells through afferent lymph to downstream they activate and polarize T-cell responses through cytokine
lymph nodes and subsequent exit in efferent lymph allows release and crosstalk with DCs [10, 11].
22 Immune Cell Trafficking in the Lymphatics, Hyaluronan Biology and Tumour Metastasis 233

In contrast to afferent lymph, the efferent lymph that section, the location of LYVE-1 above such portals and its
flows from lymph nodes back to the circulation via the tho- ligand binding characteristics endow this receptor with the
racic duct contains predominantly naive T and B cells. Hav- ability to mediate vessel entry through engagement with HA
ing entered the nodes from the circulation, these migrate in the incoming leucocyte glycocalyx. In addition to such
along fibroblast reticular cell conduits in the cortex to probe adhesion, the process of immune cell intravasation is driven
resident DCs for antigen, before exiting through medullary by actomyosin-mediated motility, directed by the chemokine
sinuses to the efferent lymphatics, in response to rising gra- CCL21, which is constitutively synthesized and secreted
dients of the sinus-derived chemotactic lipid sphingosine-­1-­ by LECs to form long-range haptotactic gradients which
phosphate (S1P [12]). Notably, during the onset of infection are detected by migrating DCs, TEM cells, macrophages and
or inflammation, the efflux of this recirculating population some neutrophils through the receptor CCR7. Intriguingly,
from the lymph node is halted transiently (3–4 days) so as contact between transiting DCs and vessel endothelium can
to prolong their residence time and increase the efficiency of also trigger rapid “on demand” discharge of CCL21 from
immune recognition. pre-stored depots in trans Golgi vesicles to sites of entry on
the basolateral surface of the endothelium by a Ca2+ triggered
exocytic mechanism that involves microtubules and fibrillar
22.3  he Distinctive Features of Lymphatic
T actin [16].
Vessels, Inter-Endothelial Junctions While the foregoing studies indicate immune cells enter
and Sites of Immune Cell Entry lymphatics via overlapping “buttoned” junctions under rest-
ing conditions or during the initial response to inflamma-
The lymphatic network differs from the blood vasculature tory stimuli (i.e. brief exposure to LPS), it is highly likely
in terms of both architecture and physiology, reflecting the that additional modes of entry are used during conditions
distinct functional requirements of the two vasculatures, and of chronic inflammation, given that newly sprouting lym-
in particular the necessity for lymphatics to handle the pas- phatic vessels generated by inflammation-induced lymphan-
sive uptake of fluid and macromolecules as well as selective giogenesis have zippered junctions (which lack LYVE-1)
entry of appropriate immune cell traffic. In addition to being rather than buttoned junctions [17]. Furthermore, exposure
a one-­way rather than a circulatory system, the lymphatics of existing lymphatic capillaries to inflammatory cytokines
fulfill these dual requirements through local variations in leads to the upregulated expression of ICAM-1, VCAM-1
vessel structure and function. Accordingly, the initial blind- and E-selectin—receptors that are known to mediate leuco-
ended capillaries possess characteristically loose, overlap- cyte transit across zippered endothelial junctions in inflamed
ping endothelial junctions and a rudimentary basement blood vessels [18]. It may be that the number of default entry
membrane, conducive to fluid absorption and cell entry [13]. sites through button junctions becomes limiting in inflam-
Just downstream of these capillaries, the pre-collectors and mation, and that induction of ICAM-1 and VCAM-1 is nec-
valved collectors have more conventional tight endothelial essary to provide for additional entry sites in downstream
junctions, akin to those of blood vessels, and a dense base- zippered junctions.
ment membrane more appropriate to the containment of
lymph and its transport by oscillatory smooth muscle con-
traction. As shown by seminal studies of DC migration in 22.4  yaluronan, its Key Biological
H
acutely inflamed mouse trachea, the majority of immune cells Functions and its Involvement
entering the lymphatics target the first few millimetres of the in Immune Cell Migration Via Lymph
initial capillaries [13]. Here, the constituent endothelial cells
have a distinctive oakleaf shape, and their interdigitation cre- HA, a large (≤ 5MDa) unbranched glycosaminoglycan, is a
ates scalloped borders in which the overlapping membrane co-polymer made up from the repeating di-saccharide [N-ace-
flaps are “buttoned” together at their sides by the adherens-­ tyl glucosamine (β-1,4) glucuronic acid β-1,3)]n, and consti-
junction protein VE–cadherin and tight junction proteins tutes a ubiquitous component of the interstitial matrix [19].
including Claudin-5, ZO-1 (zonula occludens-1), ESAM Among the many functions of HA in soft tissues is the provi-
(endothelial selective adhesion molecule) and JAM-A [13, sion of structural support and a low resistance environment
14]. In contrast, between these buttons, the flaps are coated permissive for cell migration, roles that are enabled by its
exclusively with PECAM-1 (CD31), and most notably, the capacity to absorb water equivalent to many times its mass,
lymphatic endothelial HA receptor LYVE-1 [5, 15], which fill tissue spaces and form hydrated gel-like meshworks. Syn-
gives them the flexibility to remain free and untethered, form- thesized by one of three HA synthases (HAS1-3) located at
ing small apertures or “portals” of approximately 0.5–1 μm the inner surface of the plasma membrane, these enzymes
diameter in the endothelium through which immune cells extrude the nascent HA chains to the cell surface for secre-
can squeeze (Fig. 22.1; [13, 14]). As detailed in the following tion as free polymers that are subsequently sequestered by
234 D. G. Jackson

Initial lymphatic vessel

Interstitium

Overlapping LYVE-1
junctions VE-Cadherin
DC

DC

LEC
LEC

Vessel Lumen
DC entry via transmigratory
cup in buttoned junction Key LYVE-1 HA

ICAM-1 β2 integrin

DC HA (£ 20µm)

HA

PSGL-1
(50nm)
10nm P-Selectin

β2 Integrin
(21nm) ICAM-1

α β

COOH COOH COOH COOH

Fig. 22.1 Diagram depicting a dendritic cell entering an initial lym- lower panel shows a confocal microscopic image of the DC HA glyco-
phatic capillary (upper panels) showing the distinctive overlapping calyx and the dimensions of a typical HA chain relative to underlying
junctions between individual oakleaf shaped endothelial cells and the adhesion receptors on the DC surface
formation of an LYVE-1 transmigratory cup (modified from [3]). The
22 Immune Cell Trafficking in the Lymphatics, Hyaluronan Biology and Tumour Metastasis 235

proteoglycans of the Link superfamily including Aggrecan, generates long-chain high molecular weight polymers, and
Versican and Link protein, which together with type I colla- a proportion of these is retained by CD44 in the surface gly-
gen and fibronectin form cross-linked matrices in most organs cocalyx as a dense corona of some 200–300 nm thickness
[20]. Notably, the metabolism of HA within tissues requires (Fig. 22.1; [28]). Curiously, assembly of the glycocalyx in
the participation of the lymphatics [21]. Under normal rest- DCs appears to proceed coordinately with the biosynthesis
ing conditions, HA in tissues such as skin undergoes regular and processing of CD44, insofar as the nascent HA chains
turnover (T1/2 ≥ 48 h) by cell-associated hyaluronidases, and associate with the receptor intracellularly and the bound
the intermediate cleavage products are rapidly removed and complexes are then exported to the cell surface. As a result,
transported in afferent lymph to draining lymph nodes where CD44 gene-deleted DCs fail to elaborate a glycocalyx and
they are terminally degraded through uptake by the endocytic instead accumulate the glycosaminoglycan within cytoplas-
HA receptor HARE (HA Receptor for Endocytosis) in sinu- mic vesicles [28]. Crucial evidence that the engagement of
soidal endothelium [22]. However, in inflammation or injury, HA with LYVE-1 is important for vessel entry in vivo came
the local turnover of HA is increased and this generates frag- from studies tracking the migration of fluorescently labelled
ments that are detected by the innate immune system as dan- DCs in oxazolone hypersensitized skin. These demonstrated
ger signals through binding to toll-like receptors TLR2 and that Lyve1 gene deletion and mAb blockade led to a peri-­
TLR4, and the leucocyte HA receptor CD44 [23, 24]. Unlike lymphatic accumulation of DCs in the dermis and a conse-
intact high molecular weight (0.5–5 MDa) HA which is gener- quent delay or inhibition of their trafficking to downstream
ally anti-inflammatory, these smaller (< 500 KDa) degradation axial LNs, respectively [4]. Likewise, deletion of CD44 or
products are pro-inflammatory and induce cytokine and che- depletion of the HA glycocalyx using hyaluronidase and the
mokine release as well as promoting both angiogenesis and HA biosynthetic inhibitor 4-MU (Hymecromone®) resulted
lymphangiogenesis and increased lymphatic vessel perme- in a dramatic reduction in DC nodal migration compared to
ability [25]. The molecular basis for discrimination between controls. Moreover, the significance of LYVE-1 • HA medi-
high and low MW HA is a hotly debated issue within the field, ated DC migration for in vivo immune function was under-
but is thought by many to reflect the influence of chain length scored by experiments with mice immunized intradermally
on receptor clustering and downstream signaling events. with ovalbumin as a model antigen. These showed that both
Arguably the most important function of HA is the sup- Lyve1 gene deletion and mAb blockade disrupted the genera-
port of cell:cell and cell:matrix adhesion during migration, tion of ova-specific CD4 and CD8 T-cell responses in down-
achieved by means of stable and/or transient interactions stream lymph nodes, confirming that in vivo, the process of
with receptors on neighbouring static and migratory cells LYVE-1 • HA mediated lymphatic entry can be rate-limiting
respectively. Integral to this is the capacity of HA to form a for protective immunity [4].
surface glycocalyx in appropriate cell types through reten- How engagement of the HA glycocalyx with LYVE-
tion of the polymer in the plasma membrane by its primary 1 mediates vessel entry has been elucidated from rigorous
receptor CD44. Indeed, in post-capillary blood venules, such in vitro studies with cultured monolayers of primary dermal
an HA glycocalyx on the luminal surface of the vascular lymphatic endothelial cells. These show that contact with the
endothelium serves to capture inflammatory cells includ- DC glycocalyx triggers redistribution of LYVE-1 into ring-­
ing T lymphocytes and neutrophils from flow via their own like assemblages termed transmigratory cups (Fig. 22.1)
expression of CD44, thus facilitating their extravasation to beneath the adherent leukocytes, similar to the structures that
the underlying tissue through formation of a CD44 • HA • form around T-cells during their transit across blood vascular
CD44 sandwich (reviewed in [26]). As intimated in the pre- endothelium ( [4] and reviewed in [5]). As observed by video
vious section, the discovery that HA plays an analogous role microscopy, the DCs attach to these LYVE-1 cups via their
in the intravasation of immune cells from tissue to lymphatic posterior pole or uropod, while extending a lamellipodium at
vessels came from the identification of the closely related the leading edge to explore the endothelial surface for poten-
HA receptor LYVE-1 in the button-like junctions of lym- tial sites of transit. Whether such transit occurs through the
phatic capillaries that serve as leucocyte entry portals [13, endothelial cell body (transcellular migration) or at nearby
15]. Prior to this, some earlier reports had suggested DCs inter-endothelial junctions is not yet clear, although both
and activated T-cells can synthesize and sequester cell sur- modes of transmigration are clearly possible [28]. Moreover,
face HA, but its role in lymphatic entry and migration was the density of the HA glycocalyx can influence the strength
overlooked [27]. HA is synthesized in DCs and monocyte/ of adhesion to endothelium, and hence in vivo it is likely that
macrophages primarily by HA synthase-2 (HAS2) which changes in glycocalyx organization, e.g. through enhanced
236 D. G. Jackson

HA or CD44 biosynthesis or surface clustering in response response to tissue injury. Curiously, it has also been reported
to inflammatory cytokines, may tune the rate of leucocyte that this lipoxin-mediated process for endothelial junctional
trafficking as appropriate to local tissue requirements [28]. retraction is exploited by certain mammary and colorectal
In addition to its involvement in DC trafficking, the tumours for lymphatic metastasis [34].
LYVE-1 • HA axis also mediates the trafficking of mac- Hence, transmigration by means of an HA glycocalyx is
rophages in lymph, a process that enables their clearance not a mechanism adopted by all leucocytes for lymphatic
from inflamed tissues, and one that is vital for resolution entry. In addition, though engagement of the HA glycocalyx
of inflammation and the restoration of normal homeosta- with LYVE-1 and transmigratory cup formation are criti-
sis. Like DCs, monocyte-­ derived macrophages assemble cal steps in the entry process, these are only the first in a
a surface HA glycocalyx, and their capacity to synthesize series of adhesion events required for diapedesis and it is
HA increases markedly following their differentiation to evident that further receptor ligand pairs are also involved.
an inflammatory M1-like phenotype that can migrate via Although a full description is beyond the scope of the present
lymph [29]. Indeed, such macrophages were shown to discourse, these include ICAM-1, VCAM-1 and E-selectin
adhere and transmigrate inflamed LEC monolayers in vitro, as mentioned previously, as well as ALCAM/CD166 (acti-
in an LYVE-1 and HA-dependent manner [29]. The in vivo vated leucocyte cell adhesion molecule), L1CAM, CD99,
importance of the LYVE-1 • HA axis for macrophage macrophage mannose receptor (MMR), 4-1BB/CD137 and
migration is underlined by recent studies in mouse models CLEVER-1 (combined lymphatic endothelial and vascu-
of tissue injury [30]. Firstly, in a ligation-induced model of lar endothelial receptor), several of which are selectively
myocardial infarction, a disease in which ischaemia leads upregulated by inflammatory cytokines, and all of which
to infiltration of the myocardium by neutrophils and macro- been reported to participate in immune cell adhesion and/
phages, it was shown that the latter populations are cleared or transmigration (see e.g. [2]. for review). Significantly,
via cardiac lymphatics to mediastinal lymph nodes, and both ICAM-1 and VCAM-1 are present and appear to co-­
furthermore, that stimulation of cardiac lymphangiogen- distribute with LYVE-1 in endothelial transmigratory cups
esis through local administration of recombinant VEGF-C in vitro, and blocking antibodies to ICAM-1 and its β2 integ-
not only enhanced clearance, but also reduced subsequent rin ligand LFA-1 inhibit the entry of both DCs and T-cells to
fibrotic scarring. Significantly, macrophage exit was attenu- lymphatics in oxazolone-hypersensitized and TNFα treated
ated in Lyve1−/− mice, which exhibited aggravated inflam- mouse dermis, respectively [4, 35]. One possibility is that
mation and delayed cardiac repair [30]. these come into play after appropriate disengagement or
Whether an HA glycocalyx is utilized for vessel entry by disruption of HA • LYVE-1 interactions, so as to provide
the TEM cell population that migrates in afferent lymph has extra traction for the transit of immune cells or their sub-
yet to be resolved. Although T-cells have been reported to sequent intraluminal crawling towards downstream collec-
have HA biosynthetic capacity [31], the levels of the glycos- tors. In DCs, the HA glycocalyx appears to remain attached
aminoglycan detectable on the cell surface appear to be an to the cell surface following transendothelial migration [4].
order of magnitude below those of DCs and macrophages. Nevertheless, it could conceivably undergo either discrete
Moreover, TEFF cells and TREGs are known to enter lymphatics re-­modelling by hyaluronidases, or regulated redistribution
in response to S1P and lymphotoxin (LTα1β3) and its recep- to uncover access to integrins or other underlying adhesion
tor LTβR, so the likelihood remains that these rather than the receptor ligands during diapedesis. Determining the precise
LYVE-1 • HA axis are used for their transit [32]. choreography of such events is a topic of ongoing research
Lastly, it is now evident that neutrophils lack a HA glyco- in the field of lymphatic trafficking, and one that will require
calyx and employ a mechanism that is distinct from that of much further research.
other immune cells to enter lymphatic capillaries. Recruited
exclusively to inflamed tissues, these professional phagocytes
can in appropriate circumstances exit through the l­ ymphatics 22.5 What Is So Special About Hyaluronan?
and migrate to draining lymph nodes by adhering to endo-
thelium via β2 integrins and then transmigrating through the Why immune cells should have evolved to utilize an HA
combined actions of the matrix metalloproteinases MMP8 glycocalyx for adhesion and transit across lymphatic endo-
and MMP9, the serine protease neutrophil elastase and the thelium, rather than more conventional adhesion receptors,
arachidonate-derived chemorepellent, 12,hydroxyeicosa- may be explained by the unique physical characteristics of
tetraenoate (12(S)HETE) that together promote junctional HA polymers and the novel binding properties of its primary
retraction [33]. The rationale behind this unconventional lymphatic receptor LYVE-1. First, the extreme size of HA
mechanism may be that it allows a more rapid mode of lym- molecules (comprising chains of up to 20,000 di-saccharides
phatic entry than transit through button junctions, in keeping with contour lengths of up to 1 μm) and their composition
with the strict requirement for these cells to mount a rapid from charged sugar residues bestows them with the capacity
22 Immune Cell Trafficking in the Lymphatics, Hyaluronan Biology and Tumour Metastasis 237

to form an extended polyanionic coat that reaches far beyond clear that lymphatic vessels and not blood vessels are prefer-
underlying adhesion receptors on the leucocyte surface such entially targeted by such tumours for early systemic spread
as β2 integrins (approx. 20 nm, see Fig. 22.1). This poises (Pathology of Distant/Systemic Metatases: Turk, A.). Yet in
the glycocalyx as the first key component on the immune spite of these insights, the molecular mechanisms that are
cell surface to make adhesive contact with endothelium and employed by tumours to enter the lymphatics and exit across
may necessitate its displacement to enable subsequent inter- subcapsular sinuses to access lymph node HEVs are poorly
actions with other adhesion receptors and their ligands as understood, and the degree to which they overlap with the
noted above. Second, because the receptor binding footprint transit mechanisms used by immune cells remains uncertain.
is relatively short (8–20 sugar residues), each polymer chain Within the primary tumour mass, cancer cells are initially
can accommodate multiple LYVE-1 molecules, thus gener- attracted towards lymphatic vessels by chemotaxis, pro-
ating avidity-dependent binding, the magnitude of which can moted by the same endothelial-derived chemokines CCL21
be controlled by changes in either LYVE-1 or CD44 den- and CXCL12 that direct DC and TMEM cell entry, while in
sity through surface clustering [5, 29]. Third, the mechanics draining lymph nodes their transit across the subcapsular
of the binding interaction between LYVE-1 and HA appear sinus to the parenchyma is directed by CCL1, a chemokine
to be optimally tuned for a role in facilitating immune that mediates similar transit of T-helper cells and DCs via its
cell entry. As demonstrated by atomic force spectroscopy, receptor CCR8 during inflammation [39]. Individual tumour
LYVE-­1 binds HA through a novel interaction that allows cells or emboli then initiate metastasis by invading either
the receptor to slide along the polymer chain, as distinct pre-existing or newly sprouting vessels (Lymphangiogenesis
from the conventional “sticking” bonds mediated by CD44, and Tumor Lymphangiogenesis: Detmar, M.) generated by
potentially enabling a leucocyte to crawl over the surface of secretion of the lymphangiogenic growth factor VEGF-C
the lymphatic capillary via its glycocalyx prior to transit [5, and to a lesser extent VEGF-D, both by the tumour and by
36]. Clearly, such characteristics are ideally suited to the low macrophages in its’ local microenvironment (Tumor Micro-
shear environment within the tissues from which immune environment: Co-conspirator in Tumorigenesis: Wang, Z.).
cells enter lymphatics in response to chemokine-driven hap- This expansion of tumour-associated lymphatics increases
totaxis as opposed to being captured from rapid laminar flow, the number of potential sites for tumour entry, while their
as is the case for leucocytes exiting the blood circulation. greater leakiness makes them more amenable for infiltration,
either through button junctions in the manner of immune
cells, or through zippered pre-collectors/collectors rendered
22.6 Hyaluronan and Tumour Metastasis more permeable by VEGF-C induced vessel dilation. One
might therefore anticipate that tumours with the capacity to
The majority of cancers, and in particular carcinomas, assemble a HA glycocalyx could co-opt LYVE-1 for entry to
metastasize via the lymphatic vasculature, and lymph node initial lymphatic capillaries, and as discussed subsequently,
involvement is one of the first clinical signs of tumour pro- there is now evidence suggesting this is indeed the case.
gression (Overview of Principles of Basic Lymphology and The association of HA with tumour growth and invasion
Relation to Cancer Metastasis: Witte, M.), (Primary Tumor has been recognized for many years. A majority of different
Staging and Detection of Common Sites of Distant Meta- tumour cell lines have elevated levels of both HA and CD44,
static Disease: Thomas, K., Ahmed, A., Morse, B.). Never- and upregulation of HAS gene expression and HA deposi-
theless, the impact of lymphatic metastasis remains a topic tion correlates with aggressive and metastatic cancers [40].
for debate, and the common assumption is that dissemination Moreover, upregulated synthesis of HA and CD44 is hall-
by the lymphatic route enables seeding of distant metasta- mark of epithelial to mesenchymal transition, a key process
ses only indirectly, after sequential spread through sentinel in carcinoma progression during which tumour cells acquire
and post-sentinel nodes followed by transport in efferent enhanced migratory capacity and in some cases develop-
lymph to the ­ systemic circulation via the thoracic duct. ment of cancer stem cells (Hallmarks of Cancer: Molecular
Recent studies, however, have presented persuasive evidence Underpinnings: Chitale, D.).
that tumours can metastasize directly from the first drain- While studies have linked high levels of CD44 expres-
ing node by entering the circulation through the invasion of sion with metastasis and disease progression in multiple
high endothelial venules (HEVs) in the lymph node medulla different cancers, their focus on lymphatic metastasis has
[37, 38]. Notably, ductal breast carcinomas microinjected been mostly indirect. For example, in a histological analysis
into pre-nodal lymphatics have been shown to enter the sys- of gastric carcinoma, most tumour cells showed histologi-
temic vasculature by such a route without accessing succes- cal positivity for HA, and a high proportion of such cells
sive downstream nodes, seeding distant lung metastases at a was significantly associated with lymphatic invasion, nodal
much faster rate than would be achieved by invasion of local metastasis, poor differentiation grade and adverse prognosis
blood vessels at the primary tumour site [38]. Hence, it is [41]. Similarly, studies of colorectal carcinoma revealed a
238 D. G. Jackson

strong correlation between the proportion of tumour cells stitium was associated with both peritumoral lymphatic
with HA positivity and poor survival, and between HA invasion and lymph node metastasis [50].
staining intensity and advanced metastatic dissemination
[42]. Indeed, in breast cancer metastasis, the appearance
of a CD44-associated pericellular HA matrix was observed 22.7 Conclusion and Perspective
around tumour emboli present within newly invaded lym-
phatic vessels [43]. This has also been substantiated by The present chapter has detailed how the extracellular matrix
in vitro studies demonstrating that breast tumour HS578T polysaccharide HA forms a glycocalyx on the surface of many
and MCF7 cells synthesize an HA pericellular coat and can immune cells, and how this structure mediates their entry to
adhere via HA to LYVE-1 in transfected fibroblasts [44]. lymphatic capillaries by engaging the key HA receptor LYVE-1
Furthermore, preliminary work in the authors laboratory in lymphatic endothelium for subsequent trafficking to down-
has revealed that the 4 T1 breast carcinoma line which can stream lymph nodes during immune surveillance and the reso-
metastasize directly to the blood circulation from sentinel lution of tissue injury. In addition, it outlines how an equivalent
lymph nodes also assemble an HA glycocalyx, and can glycocalyx on certain cancer cells may facilitate their invasion
adhere and transit across primary lymphatic endothelial and metastasis via lymphatics, and how HA oligosaccharides
cells via LYVE-1 transmigratory cups. Moreover, when generated by high turnover likely enhance such invasion by pro-
injected orthotopically into syngeneic mice, their capac- moting peri-tumoral lymphangiogenesis and increased lym-
ity to invade local lymphatics and metastasize to draining phatic permeability. These properties of HA and its novel
inguinal lymph nodes is disrupted in Lyve1−/− mice (Johnson binding interaction with LYVE-1 appear specifically adapted to
LA and Jackson DG unpublished). Whether this truly repre- the trafficking of both cell populations in the low shear environ-
sents the same mechanism used by DCs and macrophages to ment around lymphatic vessels, and may even offer new targets
enter initial lymphatics will require careful validation. How- for appropriate adjunct therapy of metastatic cancers.
ever, independent observations using EM, that transgenic
mouse prostate adenocarcinoma tumours, colon carcinoma
Key Questions for Future Research
cells and B16 melanomas extend filopodia-like protrusions
• What are the events subsequent to engagement of
towards the endothelium during lymphatic invasion, and
the HA glycocalyx that enable immune cells to
that endothelial cells simultaneously extend transmigratory
enter lymphatic vessels, and which additional
cup-like projections around the transiting tumour cells, add
receptor and ligands are involved?
further credence to the notion [45]. It is also likely that the
• Does the HA glycocalyx also mediate the transit of
tumour cell HA glycocalyx plays a role in downstream inva-
immune cells and tumour cells across lymph node
sion of lymph node subcapsular or medullary sinuses as a
sinuses?
prelude to systemic metastasis (Pathology of Distant/Sys-
• How widespread is HA glycocalyx assembly among
temic Metastases: Turk, A.). Notably, both LYVE-1 and the
different cancer types, and how closely does it cor-
endocytic HA receptor HARE are expressed in such sinuses,
relate with metastatic proclivity?
and administration of HARE blocking mAbs in an ortho-
• Could the HA glycocalyx have other functions dur-
topic mouse model of prostate cancer has been shown to
ing tumour growth and survival, e.g. by providing
dramatically reduce the metastatic burden in tumour drain-
camouflage to protect against immune recognition?
ing para-aortic lymph nodes [46].
• Could targeted inhibition of HA synthesis or LYVE-­
Finally, the secretion of small HA oligosaccharides by
1 antibody blockade be exploited as an adjunct for
tumour cells as a consequence of the rapid HA turnover
metastatic cancer therapy?
in neoplastic tissue may also aid lymphatic entry through
effects on vessel permeability and growth. Specifically, it
has been shown that binding of low molecular weight HA
to LYVE-1 not only decreases endothelial barrier func- References
tion in vitro, by transducing signals for the junctional
­disassembly of VE–cadherin, but also promotes lymphan- 1. Randolph GJ, Ivanov S, Zinselmeyer BH, Scallan JP. The lym-
giogenesis in vitro as well as in vivo, most likely through phatic system: integral roles in immunity. Annu Rev Immunol.
collaboration with tyrosine kinase linked growth factor 2017;35:31–52.
2. Jackson DG. Leucocyte trafficking via the lymphatic vasculature—
receptors [47–49]. Indeed, in studies on a cohort of human mechanisms and consequences. Front Immunol. 2019;10:471.
colorectal carcinoma patients, an increase in the concentra- 3. Jackson DG. Immunological functions of hyaluronan and its recep-
tion of 12-50mer HA oligosaccharides in the tumour inter- tors in the lymphatics. Immunol Rev. 2009;230(1):216–31.
22 Immune Cell Trafficking in the Lymphatics, Hyaluronan Biology and Tumour Metastasis 239

4. Johnson LA, Lawrance W, Roshorn YM, Hanke T, Banerji S, 23. Jiang D, Liang J, Noble PW. Hyaluronan as an immune regulator in
Jackson DG. The lymphatic vessel endothelial receptor LYVE-1 human diseases. Physiol Rev. 2011;91(1):221–64.
mediates dendritic cell entry to the afferent lymphatics via transmi- 24. Termeer C, Benedix F, Sleeman J, Fieber C, Voith U, Ahrens T,
gratory cups that engage the leukocyte hyaluronan glycocalyx. Nat et al. Oligosaccharides of Hyaluronan activate dendritic cells via
Immunol. 2017;18:762–70. toll-like receptor 4. J Exp Med. 2002;195(1):99–111.
5. Jackson DG. Hyaluronan in the lymphatics: The key role of the 25. Petrey AC, de la Motte CA. Hyaluronan, a crucial regulator of
hyaluronan receptor LYVE-1 in leucocyte trafficking. Matrix Biol. inflammation. Front Immunol. 2014;5:101.
2019;78–79:219–235. 26. Pure E, Cuff CA. A crucial role for CD44 in inflammation. Trends
6. Bromley SK, Thomas SY, Luster AD. Chemokine receptor CCR7 Mol Med. 2001;7(5):213–21.
guides T cell exit from peripheral tissues and entry into afferent 27. Mummert ME, Mummert D, Edelbaum D, Hui F, Matsue H,
lymphatics. Nat Immunol. 2005;6(9):895–901. Takashima A. Synthesis and surface expression of hyaluronan
7. Bromley SK, Yan S, Tomura M, Kanagawa O, Luster AD. by dendritic cells and its potential role in antigen presentation. J
Recirculating memory T cells are a unique subset of CD4+ T Immunol. 2002;169(8):4322–31.
cells with a distinct phenotype and migratory pattern. J Immunol. 28. Johnson L, Banerji S, Lagerholm BC, Jackson DG. Dendritic cell
2013;190(3):970–76. entry to lymphatic capillaries is regulated by CD44 and the hyal-
8. Forster R, Braun A, Worbs T. Lymph node homing of T cells uronan glycocalyx. Life Sci Alliance. 2021;4(5):e202000908.
and dendritic cells via afferent lymphatics. Trends Immunol. 29. Lawrance W, Banerji S, Day AJ, Bhattacharjee S, Jackson DG.
2012;33(6):271–80. Binding of Hyaluronan to the native lymphatic vessel endothelial
9. Randolph GJ, Angeli V, Swartz MA. Dendritic-cell trafficking receptor LYVE-1 is critically dependent on receptor clustering
to lymph nodes through lymphatic vessels. Nat Rev Immunol. and Hyaluronan organization. J Biol Chem. 2016;291(15):8014–
2005;5(8):617–28. 30.
10. Abadie V, Badell E, Douillard P, Ensergueix D, Leenen PJ, 30. Vieira JM, Norman S, Villa Del Campo C, Cahill TJ, Barnette DN,
Tanguy M, et al. Neutrophils rapidly migrate via lymphatics after Gunadasa-Rohling M, et al. The cardiac lymphatic system stimu-
Mycobacterium bovis BCG intradermal vaccination and shuttle live lates resolution of inflammation following myocardial infarction. J
bacilli to the draining lymph nodes. Blood. 2005;106(5):1843–50. Clin Invest. 2018;128(8):3402–12.
11. Hampton HR, Bailey J, Tomura M, Brink R, Chtanova T. Microbe-­ 31. Mahaffey CL, Mummert ME. Hyaluronan synthesis is
dependent lymphatic migration of neutrophils modulates lympho- required for IL-2-mediated T cell proliferation. J Immunol.
cyte proliferation in lymph nodes. Nat Commun. 2015;6:7139. 2007;179(12):8191–99.
12. Pham TH, Baluk P, Xu Y, Grigorova I, Bankovich AJ, Pappu R, 32. Brinkman CC, Iwami D, Hritzo MK, Xiong Y, Ahmad S, Simon T,
et al. Lymphatic endothelial cell sphingosine kinase activity is et al. Treg engage lymphotoxin beta receptor for afferent lymphatic
required for lymphocyte egress and lymphatic patterning. J Exp transendothelial migration. Nat Commun. 2016;7:12021.
Med. 2010;207(1):17–27. 33. Rigby DA, Ferguson DJ, Johnson LA, Jackson DG. Neutrophils
13. Baluk P, Fuxe J, Hashizume H, Romano T, Lashnits E, Butz S, et rapidly transit inflamed lymphatic vessel endothelium via integrin-­
al. Functionally specialized junctions between endothelial cells of dependent proteolysis and lipoxin-induced junctional retraction. J
lymphatic vessels. J Exp Med. 2007;204(10):2349–62. Leukoc Biol. 2015;98:897–912.
14. Zhang F, Zarkada G, Yi S, Eichmann A. Lymphatic endothelial cell 34. Kerjaschki D, Bago-Horvath Z, Rudas M, Sexl V, Schneckenleithner
junctions: molecular regulation in physiology and diseases. Front C, Wolbank S, et al. Lipoxygenase mediates invasion of intrameta-
Physiol. 2020;11:509. static lymphatic vessels and propagates lymph node metastasis of
15. Banerji S, Ni J, Wang SX, Clasper S, Su J, Tammi R, et al. LYVE-­ human mammary carcinoma xenografts in mouse. J Clin Invest.
1, a new homologue of the CD44 glycoprotein, is a lymph-specific 2011;121(5):2000–12.
receptor for hyaluronan. J Cell Biol. 1999;144(4):789–801. 35. Johnson LA, Prevo R, Clasper S, Jackson DG. Inflammation-
16. Vaahtomeri K, Brown M, Hauschild R, De Vries I, Leithner AF, induced uptake and degradation of the lymphatic endothelial hyal-
Mehling M, et al. Locally triggered release of the chemokine uronan receptor LYVE-1. J Biol Chem. 2007;282(46):33671–80.
CCL21 promotes dendritic cell transmigration across lymphatic 36. Bano F, Banerji S, Howarth M, Jackson D, Richter R. A Single mol-
endothelia. Cell Rep. 2017;19(5):902–9. ecule assay to probe monovalent and multivalent bonds between
17. Yao LC, Baluk P, Srinivasan RS, Oliver G, McDonald DM. hyaluronan and its key leukocyte receptor CD44 under force.
Plasticity of button-like junctions in the endothelium of air- Scientific Rep. 2016;6:34176.
way lymphatics in development and inflammation. Am J Pathol. 37. Padera TP, Kadambi A, di Tomaso E, Carreira CM, Brown EB,
2012;180(6):2561–75. Boucher Y, et al. Lymphatic metastasis in the absence of functional
18. Johnson LA, Clasper S, Holt AP, Lalor PF, Baban D, Jackson intratumor lymphatics. Science. 2018;296(5574):1883–6.
DG. An inflammation-induced mechanism for leukocyte trans- 38. Brown M, Assen FP, Leithner A, Abe J, Schachner H, Asfour
migration across lymphatic vessel endothelium. J Exp Med. G, et al. Lymph node blood vessels provide exit routes
2006;203(12):2763–77. for metastatic tumor cell dissemination in mice. Science.
19. Monslow J, Govindaraju P, Pure E. Hyaluronan—a functional 2018;359(6382):1408–11.
and structural sweet spot in the tissue microenvironment. Front 39. Das S, Sarrou E, Podgrabinska S, Cassella M, Mungamuri SK,
Immunol. 2015;6:231. Feirt N, et al. Tumor cell entry into the lymph node is controlled
20. Weigel PH, DeAngelis PL. Hyaluronan synthases: a decade-plus by CCL1 chemokine expressed by lymph node lymphatic sinuses. J
of novel glycosyltransferases. J Biol Chem. 2007;282(51):36777– Exp Med. 2013;210(8):1509–28.
81. 40. Liu M, Tolg C, Turley E. Dissecting the dual nature of Hyaluronan
21. Armstrong SE, Bell DR. Relationship between lymph and tissue in the tumor microenvironment. Front Immunol. 2019;10:947.
hyaluronan in skin and skeletal muscle. Am J Physiol Heart Circ 41. Setala LP, Tammi MI, Tammi RH, Eskelinen MJ, Lipponen PK,
Physiol. 2002;283(6):H2485–94. Agren UM, et al. Hyaluronan expression in gastric cancer cells is
22. Harris EN, Kyosseva SV, Weigel JA, Weigel PH. Expression, pro- associated with local and nodal spread and reduced survival rate. Br
cessing, and glycosaminoglycan binding activity of the recombi- J Cancer. 1999;79(7–8):1133–38.
nant human 315-kDa hyaluronic acid receptor for endocytosis 42. Ropponen K, Tammi M, Parkkinen J, Eskelinen M, Tammi R,
(HARE). J Biol Chem. 2007;282(5):2785–97. Lipponen P, et al. Tumor cell-associated hyaluronan as an unfa-
240 D. G. Jackson

vorable prognostic factor in colorectal cancer. Cancer Res. 47. Wu M, Du Y, Liu Y, He Y, Yang C, Wang W, et al. Low molecular
1998;58(2):342–7. weight hyaluronan induces lymphangiogenesis through LYVE-1-­
43. Williams CS, Leek RD, Robson AM, Banerji S, Prevo R, Harris mediated signaling pathways. PLoS One. 2014;9(3):e92857.
AL, et al. Absence of lymphangiogenesis and intratumoural 48. Bauer J, Rothley M, Schmaus A, Quagliata L, Ehret M, Biskup M,
lymph vessels in human metastatic breast cancer. J Pathol. et al. TGFbeta counteracts LYVE-1-mediated induction of lym-
2003;200(2):195–206. phangiogenesis by small hyaluronan oligosaccharides. J Mol Med.
44. Du Y, Liu Y, Wang Y, He Y, Yang C, Gao F. LYVE-1 enhances the 2018;96(2):199–209.
adhesion of HS-578T cells to COS-7 cells via hyaluronan. Clin 49. Hou WH, Liu IH, Tsai CC, Johnson FE, Huang SS, Huang JS.
Invest Med. 2011;34(1):E45–54. CRSBP-1/LYVE-1 ligands disrupt lymphatic intercellular adhe-
45. Azzali G. On the transendothelial passage of tumor cell from sion by inducing tyrosine phosphorylation and internalization of
extravasal matrix into the lumen of absorbing lymphatic vessel. VE-cadherin. J Cell Sci. 2011;124(Pt 8):1231–44.
Microvasc Res. 2006;72(1–2):74–85. 50. Schmaus A, Klusmeier S, Rothley M, Dimmler A, Sipos B, Faller
46. Simpson MA, Weigel JA, Weigel PH. Systemic blockade of the G, et al. Accumulation of small hyaluronan oligosaccharides in
hyaluronan receptor for endocytosis prevents lymph node metasta- tumour interstitial fluid correlates with lymphatic invasion and
sis of prostate cancer. Int J Cancer. 2012;131(5):E836–40. lymph node metastasis. Br J Cancer. 2014;111(3):559–67.
 Lymphovascular Genomics
and Proteomics, Clinical Syndromes, 23
and Cancer Metastasis

Robert P. Erickson and Michael T. Dellinger

Abstract 23.1 Introduction

The number of genetic mutations that cause lymphedema
and other diseases of the lymphatic system has signifi- Over the past 30 years, substantial progress has been made in
cantly grown in recent years. This information has shed the field of molecular lymphology. Molecular markers of
light on the molecular mechanisms controlling the devel- lymphatics have been identified, genes that control lym-
opment of the lymphatic system and has increased our phatic development have been uncovered, and treatments for
understanding of the etiology of inherited and non-­ lymphatic diseases have been discovered. Additionally, lym-
inherited diseases of the lymphatic system. Here, we phatics have been shown to serve a critical role in the tumor
briefly describe the different genetic mutations that cause microenvironment. Cancer cells can stimulate the growth of
diseases of the lymphatic system, some of which are also lymphatics and use these newly formed lymphatics to escape
mutated in some cancers. Additionally, we discuss the the primary tumor. The growth of lymphatics involves cell
similarities between lymphatic diseases and cancer. proliferation and migration, two tightly regulated processes
that are dysregulated in cancer cells. Interestingly, the molec-
ular pathways controlling these behaviors in cancer cells
serve overlapping functions in lymphatic endothelial cells
Learning Objectives (LECs). There is also growing evidence that the genetic
By the end of this chapter, readers will be able to: mutations that cause cancer can also cause rare lymphatic
malformations. Thus, a better understanding of the molecu-
• Name pathways that serve a critical role in the lar mechanisms underlying rare diseases of the lymphatic
development of lymphatic vessels in health and system could yield novel therapies to block tumor lymphan-
disease. giogenesis and/or halt the progression of cancer. In this
• Name genes that control the development of lym- chapter, we briefly review the genetics of human lymphatic
phatic vessels. diseases, discuss the surprising similarities between some
• List several characteristics of different lymphatic lymphatic diseases and cancer, and point to directions for
diseases. studies of lymphatics in metastasis.
• Give examples of treatments for lymphatic
diseases.
23.2 Early Genetic Discoveries

Lymphedema is abnormal swelling due to accumulated inter-

stitial fluid that generally occurs in the arms or legs. The most
R. P. Erickson (*) common form of lymphedema is secondary lymphedema. In
Department of Pediatrics, University of Arizona College
the Western world, this form of lymphedema is frequently
of Medicine, Tucson, AZ, USA
e-mail: [email protected] observed in cancer patients that have had lymph nodes
removed for the staging of their disease. Primary lymph-
M. T. Dellinger
Division of Surgical Oncology, Department of Surgery edema occurs less frequently and is caused by inherited
and the Hamon Center for Therapeutic Oncology Research, UT genetic mutations. Our evolving knowledge of genes in which
Southwestern Medical Center, Dallas, TX, USA mutations cause primary lymphedema has depended on syn-
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 241
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_23
242 R. P. Erickson and M. T. Dellinger

ergistic findings in both man and mice. The first major dis- of the vessel, and then collectively invade the vessel to form
coveries were based on classical, human genetic approaches a valve comprised of two layers of LECs. Importantly, muta-
using linkage studies in large families to locate, and then tions in FOXC2 and GATA2 have been shown to affect lym-
clone, genes involved in lymphatic development and func- phatic valve development in mice and cause lymphedema in
tion. Mutations in VEGFR3 (FLT4, the receptor for the major humans [9, 11, 12]. The planar cell polarity (PCP) gene
lymphatic growth factor, VEGFC) in Milroy’s, FOXC2 (a CELSR1 is also essential for normal lymphatic valve devel-
transcription factor important in the development of lymphat- opment. PCP refers to the coordinated orientation of cells in
ics) in lymphedema/distichiasis syndrome, and SOX18 epithelia in the direction perpendicular to the apical-basal
(another transcription factor) in hypotrichosis–lymphedema– orientation. Loss of Celsr1 in LECs in mice impairs lym-
telangiectasia syndrome are early examples [1–3]. phatic valve development and mutations in CELSR1 have
been identified in families with lymphedema [9, 13].
Connexins are gap junction proteins that are important for
23.3  utations Affecting Early Stages
M lymphatic development and three have been found to be
of Lymphatic Development expressed by lymphatic valves: Cx37, Cx43, and Cx47 [14].
In mice, knockouts of Cx37 and Cx43 disrupt lymphatic
Multiple genes are involved in the early development of lym- valve development and result in embryonic lymphedema and
phatics. The discovery of prospero-related homeobox-1 chylothorax with markedly reduced postnatal survival [15].
(Prox1) provided the first marker for early lymphatic develop- Ferrell and colleagues first reported GJC2 (Cx47) mutations
ment [4]. It is required for the development and maintenance in two families with dominantly inherited lymphedema but
of LECs [5]. CCBE1 (collagen and calcium-binding EGF “uncomplicated lymphedema of the leg or hand” was the
domain-containing protein 1), FAT4 (encoding an atypical only nearly constant feature and non-penetrance occurred in
cadherin), and ADAMTS3 (encoding a metalloproteinase) also one individual [16]. Other families followed and one family
affect early lymphatic development and mutations in these with a mutation in GJA1 (Cx43) was found [17]. Recently,
genes cause Hennekam’s syndrome, an autosomal recessive the CRISPR/Cas9 system was used to create mice that har-
disorder characterized by lymphangiectasia/lymphedema [6]. bor a Cx47 mutation found in humans. These mice showed a
Rare mutations in VEGFC, the major stimulator of lymphatic lymphatic phenotype of increased numbers of lymph nodes
growth and development, can also cause lymphedema [7]. and lymphatic channels and decreased lymphatic contractil-
VEGF-C also stimulates tumor lymphangiogenesis and metas- ity [18]. Genetic variation in CX47 [19] and VEGFR3 [20]
tasis and it was recently reported that VEGF-C-mediated lym- has been implicated in susceptibility to post-breast cancer
phatic drainage enables immunosurveillance of brain tumors surgery lymphedema.
[8]. VEGF-C activates VEGFR3, a receptor tyrosine kinase
that promotes PI3K/AKT/mTOR and RAS/MAPK signaling
in LECs. Importantly, these pathways have many components 23.5  utations Affecting Flow Sensing
M
in which mutations are associated with disorders that exhibit and Signaling in Lymphatics
lymphedema. They include Noonan’s syndrome (PTPN11,
SOS1, RAF1, KRAS, BRAF, SHOC2, and CBL), Costello syn- After the initiation of lymphangiogenesis, laminar fluid flow
drome (HRAS), cardio-facial-cutaneous syndrome (BRAF), serves a function in expanding the growing lymphatic net-
lymphedema–choanal atresia syndrome (PTPN14), and rare work. ORAI1 is a pore subunit of the calcium release–acti-
cases of chylothorax (RASA1) [9]. vated calcium channel that promotes LEC proliferation in
response to laminar fluid flow [21]. Mechanistically, ORAI1
induces LEC proliferation by upregulating Krüppel-like factor
23.4  utations Affecting Lymphatic Valve
M (KLF)-2 and KLF4 in flow-activated LECs. These transcrip-
Formation tion factors upregulate VEGF-A, VEGF-C, and FGFR3 and
downregulate the cell cycle inhibitor, p57 [21]. PIEZO1 is also
Lymphatic valves are bicuspid structures that prevent the ret- an ion channel whose activity is regulated by fluid flow.
rograde flow of lymph. In general, mutations in genes that Interestingly, loss of Piezo1 in mice impairs the development
regulate lymphatic valve development cause lymphangiecta- of lymphatic valves, and mutations in PIEZO1 can cause
sia and lymph reflux. Lymphatic valve development begins lymphedema in humans [22]. Fluid shear stress signaling in
when clusters of LECs upregulate the transcription factors LECs is also mediated by a mechanosensory complex com-
PROX1, FOXC2, and GATA2 [10]. These valve-forming prised of PECAM-1 and VE-cadherin and loss of either gene
cells elongate, reorient with respect to the longitudinal axis impairs the development of lymphatic valves in mice [23, 24].
23 Lymphovascular Genomics and Proteomics, Clinical Syndromes, and Cancer Metastasis 243

23.6 Polygenic Causes of Lymphedema phatic anomalies. Sporadic lymphatic anomalies are non-
inherited diseases caused by errors in the development of
Several chromosomal disorders are associated with lymph- the lymphatic system. According to the International
edema. Turner’s syndrome (45,XO) frequently has infantile Society for the Study of Vascular Anomalies (ISSVA), there
generalized lymphedema (so much so that, before the chromo- are several different types of sporadic lymphatic anomalies
somal cause was known, it was considered a separate syn- [33]. These include common cystic lymphatic malforma-
drome, Bonnevie-Ullrich) and in children/adults it can reoccur. tions (CCLMs), generalized lymphatic anomaly (GLA),
While an X-chromosomal location has been found, a specific central conducting lymphatic anomaly (CCLA), kaposi-
gene has not been pinpointed and it is likely that several loci’s form lymphangiomatosis (KLA), and Gorham–Stout dis-
gene products contribute to the phenotype [25]. Similarly, tri- ease (GSD) [33]. Patients with these diseases can have
somy 21 (and the rarer trisomies 13 and 18) is frequently asso- tortuous dilated lymphatics, large disfiguring lymphatic
ciated with lymphedema and these chromosomal mutations malformations (LMs), multifocal LMs that infiltrate sur-
are sometimes diagnosed in utero because of ultrasound- rounding tissue, and ectopic lymphatics in bone [34]. The
detected nuchal edema (posterior lateral neck swellings overlapping signs, symptoms, and complications of these
thought to be related to enlargement of the cervical lymphatic diseases make them difficult to diagnose and the etiology of
sacs). Again, multiple genes are likely to be involved. The not- these diseases is poorly understood [34]. However, recent
infrequently diagnosed Prader–Willi syndrome, which genetic studies have begun to shed light on the genetic
involves abnormal imprinting of a portion of chromosome 15 basis of sporadic lymphatic anomalies. In the following
caused by gene or chromosomal mutations, often shows sections, we briefly review the latest advances in research
lymphedema along with other syndromic features [26]. on the genetic underpinnings and treatment of sporadic
lymphatic anomalies.

23.7  dditional Pathways Regulating

A
Lymphangiogenesis
23.9  ommon Cystic Lymphatic
C
While the emphasis in the preceding section was on the Malformations (CCLMs)
VEGFC/VEGFR3 pathway, there are many other signaling and Generalized Lymphatic
pathways involved in lymphatic development and function. Anomaly (GLA)
These pathways control LEC growth, proliferation, m­ igration,
adhesion, differentiation, and apoptosis. EphB4 is a receptor CCLMs are isolated lesions comprised of irregular lymphatic
tyrosine kinase that is activated by the transmembrane ligand vessels [35, 36]. Based on radiographic imaging, these mal-
EphrinB2 [27]. Interestingly, engagement of EphB4 by formations are classified as being either macrocystic, micro-
EphrinB2 causes “reverse signaling” from the cytoplasmic cystic, or mixed [35, 36]. A majority of CCLMs are apparent
tail of EphrinB2. This unusual form of signaling serves a at birth and affect the head and neck [35, 36]. Unfortunately,
critical role in the maturation of lymphatics in mice [27]. CCLMs can severely disfigure affected areas and impair nor-
Additionally, EphB4 “forward signaling” has been shown to mal organ function by obstructing or compressing structures
function in lymphatic valve development in mouse and man [35, 36]. GLA (formerly called lymphangiomatosis) is a
[28, 29]. Another pathway highly important for lymphatic related disorder that is characterized by multiple diffuse or
development is the Angiopoietin-Tie pathway. Angiopoietin-2 multifocal LMs [37]. Patients with GLA frequently display
(Angpt2) activates Tie2 on LECs and loss of either of the fac- lymphatic abnormalities in their soft tissues and abdominal
tor impairs lymphatic development in mice [30, 31]. and thoracic viscera. Interestingly, they can also develop lym-
Underscoring the importance of this pathway in lymphatic phatics in multiple noncontiguous bones [37]. Recently,
development is the recent observation that mutations in somatic activating mutations in PIK3CA have been shown to
ANGPT2 may cause primary lymphedema in humans [32]. cause both CCLM and GLA [35, 36, 38, 39]. PIK3CA encodes
for p100α, which is the catalytic subunit of phosphatidylino-
sitol 3-kinase (PI3K). The PIK3CA mutations in CCLM and
23.8  verview of Sporadic Lymphatic
O GLA patients are “hotspot” mutations that occur in cancer
Anomalies and cause hyperactive PI3K/AKT/mTOR signaling [39].
Importantly, the FDA-approved mTOR inhibitor, sirolimus,
The previous sections focused on germline mutations and has been reported to reduce lesion size in patients with CCLM
inherited diseases of the lymphatic system. The following and to reduce signs and symptoms of disease in patients with
sections focus on somatic mutations and sporadic lym- GLA [39–42].
244 R. P. Erickson and M. T. Dellinger

23.10 C
 entral Conducting Lymphatic in cell-free DNA from patients [47, 48]. However, not all
Anomaly (CCLA) patients with KLA have a mutation in NRAS. Recently, a
causative mutation in CBL was identified in a patient with
CCLA (formerly called lymphangiectasia) is characterized KLA [49]. Mutations in CBL also increase RAS/MAPK sig-
by the pathologic dilation of lymphatic vessels [34]. Common naling and the KLA patient with the CBL mutation had a
complications of CCLA include chylothorax, chylous asci- dramatic improvement in their disease following treatment
tes, and protein-losing enteropathy [34]. Despite intense with trametinib [49].
research efforts to identify effective therapies for CCLA, it
remains an extremely challenging condition to treat. To gain
a better understanding of the etiology of CCLA, affected tis- 23.12 Gorham–Stout Disease (GSD)
sues from two patients were analyzed by next-generation
sequencing [43]. This revealed that both patients had a GSD (also known as vanishing bone disease) is a rare disease
somatic mutation in ARAF (p.S214P) [43]. ARAF is a serine/ that is characterized by the development of lymphatics in bone
threonine kinase and the p.S214P mutation increased MAPK and progressive osteolysis. J.B.S. Jackson is credited with
signaling in LECs and caused aberrant lymphatic develop- publishing the first case of GSD in 1838 [50]. However, the
ment in zebrafish [43]. Importantly, one patient with an clinical and histologic features of the disease were not defined
ARAF (p.S214P) mutation was treated with trametinib (an until Gorham and Stout published their seminal paper in 1955
FDA-approved MEK1/2 inhibitor) and their lung function [51]. Since 1955, more than 300 cases of GSD have been
and lymphedema significantly improved while on drug [43]. described in the literature [50]. These reports have revealed
Whether trametinib is effective in other patients with CCLA that GSD does not display clear sex predilection or inheritance
remains to be determined. pattern. Although GSD can involve any bone in the body, it
typically affects the ribs and vertebrates [50]. Areas of bone
resorption can arise in a single bone, but usually occur in mul-
23.11 Kaposiform Lymphangiomatosis tiple contiguous bones [50]. In severe cases of the disease,
(KLA) osteolysis continues until entire bones are resorbed and
replaced by fibrous tissue [50]. A somatic activating KRAS
KLA is an aggressive lymphatic anomaly that is typically mutation (p.Q61R) was recently identified in one patient with
diagnosed in children and has a 5-year survival rate of 51% GSD [52]. The KRAS (p.Q61R) variant is a known “hotspot”
[44]. Patients with KLA exhibit diffuse lymphatic abnor- mutation that frequently occurs in cancer and causes excessive
malities in multiple tissues with the mediastinum, lung, RAS/MAPK signaling [52]. The identification of this muta-
pleura, spleen, and skeleton being the most frequently tion raises the possibility that inhibitors of the RAS/MAPK
affected sites [44]. Similar to GLA, bone involvement in pathway could be novel therapies for GSD.
KLA is associated with the loss of medullary bone, not cor-
tical bone [44]. However, hemorrhagic pleural or pericardial
effusions are more common in KLA than GLA [44]. 23.13 P
 recision Medicine for Sporadic
Thrombocytopenia is also more common in KLA than GLA Lymphatic Anomalies
[44]. KLA is diagnosed based on clinical and histological
observations. KLA has a unique histology with spindle- Sporadic lymphatic anomalies are confusing and devastating
shaped LECs residing near abnormal lymphatics [44]. diseases. If left untreated, patients suffer from a poor quality
Several therapeutics have been used to treat KLA with the of life and can develop life-threatening complications. The
most common being vincristine, interferon-α, and sirolimus reports described above show that sporadic lymphatic anom-
[44, 45]. However, these treatments are inadequate for some alies are caused by actionable mutations in oncogenes and
patients. A better understanding of the genetic causes of respond to FDA-approved cancer therapies [39, 43, 49].
KLA could lead to new treatments for patients. To identify a Therefore, supportive care should no longer be the standard
disease-causing mutation for KLA, LECs were isolated of care for patients. Clinicians who encounter patients should
from a KLA patient and analyzed by whole-exome sequenc- attempt to obtain a molecular diagnosis because targeted
ing [46]. This revealed that the cells harbored a “hotspot” therapies can produce miraculous results when administered
mutation in NRAS (p.Q61R) [46]. The p.Q61R mutation to patients. Future clinical trials with molecularly diagnosed
occurs in cancer and causes hyperactive RAS/MAPK sig- patients and targeted therapies could ultimately lead to the
naling. Interestingly, this same NRAS variant has been repurposing of FDA-approved cancer therapies for sporadic
observed in affected tissues from 10 other KLA patients and lymphatic anomalies.
23 Lymphovascular Genomics and Proteomics, Clinical Syndromes, and Cancer Metastasis 245

23.14 T
 argeted Therapies to Block Tumor
Lymphangiogenesis • What is the full complement of mutations that cause
lymphatic disease in humans?
Cancer patients are frequently treated with drugs that block • Are there tissue-specific differences in the develop-
either PI3K/AKT/mTOR or RAF/MEK/MAPK signaling ment and function of lymphatics?
because these pathways serve a critical role in tumorigene- • Can molecular-based therapies be used to treat pri-
sis. These pathways, when active in LECs, also stimulate mary lymphedema?
lymphangiogenesis (discussed in previous sections). • Why do mutations in oncogenes in LECs cause spo-
Therefore, PI3K/AKT/mTOR and RAF/MEK/MAPK inhib- radic lymphatic anomalies instead of cancer?
itors could potentially be used to inhibit tumor lymphangio- • Can synthetically lethal genetic interactions be
genesis and metastasis to lymph nodes. Future investigations exploited to develop novel therapies for sporadic
could reveal the effect of targeted therapies on tumor lym- lymphatic anomalies?
phangiogenesis and metastasis. • Do targeted therapies inhibit lymphangiogenesis in
pathological settings (e.g., tumor lymphangiogene-
sis in cancer)?
23.15 A
 dditional Links between Lymphatics
and Cancer Importantly, sustained research efforts will allow
scientists to fill these gaps in knowledge and increase
There are several additional links between lymphatics, lym- the ability of clinicians to treat patients suffering from
phatic diseases, and cancer. As mentioned previously, cancer primary lymphedema or a sporadic lymphatic anom-
cells metastasize to lymph nodes by traveling through lym- aly. They could also result in new treatments to block
phatic vessels. Sporadic lymphatic anomaly patients can also the spread and growth of cancer.
have multifocal disease. Whether this is a result of LEC
“metastasis” remains to be determined. Additionally, some
secondary lymphedema patients go on to develop an aggres-
sive cancer called lymphangiosarcoma. The mechanisms by References
which the altered microenvironment in lymphedematous
limbs facilitate this transformation is poorly understood. 1. Fang J, Dagenais SL, Erickson RP, Arlt MF, Glynn MW, Gorski
JL, Seaver LH, Glover TW. Mutations in FOXC2 (MFH-1),
Lastly, many of the genes that serve a role in the development a forkhead family transcription factor, are responsible for the
of the lymphatic system function in processes that form the hereditary lymphedema-distichiasis syndrome. Am J Hum Genet.
biological hallmarks of cancer. These processes include prolif- 2000;67(6):1382–8. https://doi.org/10.1086/316915.
erative signaling, cell invasion, regulation of cellular energet- 2. Ferrell RE, Levinson KL, Esman JH, Kimak MA, Lawrence
EC, Barmada MM, Finegold DN. Hereditary lymphedema: evi-
ics, and angiogenesis. Thus, advances in the study of lymphatic dence for linkage and genetic heterogeneity. Hum Mol Genet.
diseases could impact our understanding of the biological hall- 1998;7(13):2073–8.
marks of cancer and the treatment of cancer patients. 3. Irrthum A, Devriendt K, Chitayat D, Matthijs G, Glade C, Steijlen
PM, Fryns JP, Van Steensel MA, Vikkula M. Mutations in the tran-
scription factor gene SOX18 underlie recessive and dominant forms
of hypotrichosis-lymphedema-telangiectasia. Am J Hum Genet.
Conclusion and Perspective 2003;72(6):1470–8. https://doi.org/10.1086/375614.
4. Wigle JT, Oliver G. Prox1 function is required for the develop-
Traditional genetic approaches and next-generation ment of the murine lymphatic system. Cell. 1999;98(6):769–78.
sequencing studies have led to the identification of doi:S0092-­8674(00)81511-­1 [pii]
germline and somatic mutations that cause diseases of 5. Johnson NC, Dillard ME, Baluk P, McDonald DM, Harvey NL,
the lymphatic system. This information has shed light Frase SL, Oliver G. Lymphatic endothelial cell identity is revers-
ible and its maintenance requires Prox1 activity. Genes Dev.
on the molecular mechanisms controlling the develop- 2008;22(23):3282–91. https://doi.org/10.1101/gad.1727208.
ment of the lymphatic system and has increased our 6. Scheuerle AE, Sweed NT, Timmons CF, Smith ED, Alcaraz WA,
understanding of the etiology of these disabling, dis- Shinde DN. An additional case of Hennekam lymphangiectasia-­
figuring, and life-threatening diseases. Mutations lymphedema syndrome caused by loss-of-function mutation in
ADAMTS3. Am J Med Genet A. 2018;176(12):2858–61. https://
occurring in sporadic lymphatic abnormalities also doi.org/10.1002/ajmg.a.40633.
occur in some cancers and the lymphatic system is 7. Gordon K, Schulte D, Brice G, Simpson MA, Roukens MG, van
highly implicated in tumor spread. Although substan- Impel A, Connell F, Kalidas K, Jeffery S, Mortimer PS, Mansour
tial progress has been made in the field, many ques- S, Schulte-Merker S, Ostergaard P. Mutation in vascular endothe-
lial growth factor-C, a ligand for vascular endothelial growth fac-
tions still exist surrounding the development of tor receptor-3, is associated with autosomal dominant Milroy-like
lymphatic vessels in health and disease. primary lymphedema. Circ Res. 2013;112(6):956–60. https://doi.
org/10.1161/CIRCRESAHA.113.300350.
246 R. P. Erickson and M. T. Dellinger

8. Song E, Mao T, Dong H, Boisserand LSB, Antila S, Bosenberg activated ion channel PIEZO1 is required for lymphatic valve for-
M, Alitalo K, Thomas JL, Iwasaki A. VEGF-C-driven lym- mation. Proc Natl Acad Sci U S A. 2018;115(50):12817–22. https://
phatic drainage enables immunosurveillance of brain tumours. doi.org/10.1073/pnas.1817070115.
Nature. 2020;577(7792):689–94. https://doi.org/10.1038/ 23. Wang Y, Baeyens N, Corti F, Tanaka K, Fang JS, Zhang J, Jin Y,
s41586-­019-­1912-­x. Coon B, Hirschi KK, Schwartz MA, Simons M. Syndecan 4 con-
9. Michelini S, Paolacci S, Manara E, Eretta C, Mattassi R, Lee BB, trols lymphatic vasculature remodeling during mouse embryonic
Bertelli M. Genetic tests in lymphatic vascular malformations development. Development. 2016;143(23):4441–51. https://doi.
and lymphedema. J Med Genet. 2018;55(4):222–32. https://doi. org/10.1242/dev.140129.
org/10.1136/jmedgenet-­2017-­105064. 24. Yang Y, Cha B, Motawe ZY, Srinivasan RS, Scallan
10. Sabine A, Saygili Demir C, Petrova TV. Endothelial cell responses JP. VE-cadherin is required for lymphatic valve formation and
to biomechanical forces in lymphatic vessels. Antioxid Redox maintenance. Cell Rep. 2019;28(9):2397–412. e2394. https://doi.
Signal. 2016;25(7):451–65. https://doi.org/10.1089/ars.2016.6685. org/10.1016/j.celrep.2019.07.072.
11. Kazenwadel J, Betterman KL, Chong CE, Stokes PH, Lee YK, 25. Boucher CA, Sargent CA, Ogata T, Affara NA. Breakpoint anal-
Secker GA, Agalarov Y, Demir CS, Lawrence DM, Sutton DL, ysis of turner patients with partial Xp deletions: implications for
Tabruyn SP, Miura N, Salminen M, Petrova TV, Matthews JM, the lymphoedema gene location. J Med Genet. 2001;38(9):591–8.
Hahn CN, Scott HS, Harvey NL. GATA2 is required for lym- https://doi.org/10.1136/jmg.38.9.591.
phatic vessel valve development and maintenance. J Clin Invest. 26. Lee BB, Andrade M, Antignani PL, Boccardo F, Bunke N, Campisi
2015;125(8):2979–94. https://doi.org/10.1172/JCI78888. C, Damstra R, Flour M, Forner-Cordero I, Gloviczki P, Laredo J,
12. Petrova TV, Karpanen T, Norrmen C, Mellor R, Tamakoshi T, Partsch H, Piller N, Michelini S, Mortimer P, Rabe E, Rockson
Finegold D, Ferrell R, Kerjaschki D, Mortimer P, Yla-Herttuala S, Scuderi A, Szolnoky G, Villavicencio JL, International Union
S, Miura N, Alitalo K. Defective valves and abnormal mural cell of P (2013) Diagnosis and treatment of primary lymphedema.
recruitment underlie lymphatic vascular failure in lymphedema Consensus document of the International Union of Phlebology
distichiasis. Nat Med. 2004;10(9):974–81. https://doi.org/10.1038/ (IUP)-2013. Int Angiol 32 (6):541–574.
nm1094nm1094. 27. Makinen T, Adams RH, Bailey J, Lu Q, Ziemiecki A, Alitalo
13. Erickson RP, Lai LW, Mustacich DJ, Bernas MJ, Kuo PH, K, Klein R, Wilkinson GA. PDZ interaction site in ephrinB2 is
Witte MH. Sex-limited penetrance of lymphedema to females required for the remodeling of lymphatic vasculature. Genes Dev.
with CELSR1 haploinsufficiency: a second family. Clin Genet. 2005;19(3):397–410. https://doi.org/10.1101/gad.330105.
2019;96(5):478–82. https://doi.org/10.1111/cge.13622. 28. Li D, Wenger TL, Seiler C, March ME, Gutierrez-Uzquiza A,
14. Kanady JD, Simon AM. Lymphatic communication: connexin Kao C, Bhoj E, Tian L, Rosenbach M, Liu Y, Robinson N, Behr
junction, what's your function? Lymphology. 2011;44(3):95–102. M, Chiavacci R, Hou C, Wang T, Bakay M, Pellegrino da Silva
15. Kanady JD, Dellinger MT, Munger SJ, Witte MH, Simon R, Perkins JA, Sleiman P, Levine MA, Hicks PJ, Itkin M, Dori Y,
AM. Connexin37 and Connexin43 deficiencies in mice disrupt lym- Hakonarson H. Pathogenic variant in EPHB4 results in central con-
phatic valve development and result in lymphatic disorders includ- ducting lymphatic anomaly. Hum Mol Genet. 2018;27(18):3233–
ing lymphedema and chylothorax. Dev Biol. 2011;354(2):253–66. 45. https://doi.org/10.1093/hmg/ddy218.
https://doi.org/10.1016/j.ydbio.2011.04.004. 29. Martin-Almedina S, Martinez-Corral I, Holdhus R, Vicente A,
16. Ferrell RE, Baty CJ, Kimak MA, Karlsson JM, Lawrence EC, Fotiou E, Lin S, Petersen K, Simpson MA, Hoischen A, Gilissen C,
Franke-Snyder M, Meriney SD, Feingold E, Finegold DN. GJC2 Jeffery H, Atton G, Karapouliou C, Brice G, Gordon K, Wiseman
missense mutations cause human lymphedema. Am J Hum Genet. JW, Wedin M, Rockson SG, Jeffery S, Mortimer PS, Snyder MP,
2010;86(6):943–8. https://doi.org/10.1016/j.ajhg.2010.04.010. Berland S, Mansour S, Makinen T, Ostergaard P. EPHB4 kinase-­
17. Brice G, Ostergaard P, Jeffery S, Gordon K, Mortimer PS, Mansour inactivating mutations cause autosomal dominant lymphatic-­
S. A novel mutation in GJA1 causing oculodentodigital syndrome related hydrops fetalis. J Clin Invest. 2016;126(8):3080–8. https://
and primary lymphoedema in a three generation family. Clin Genet. doi.org/10.1172/JCI85794.
2013;84(4):378–81. https://doi.org/10.1111/cge.12158. 30. Dellinger M, Hunter R, Bernas M, Gale N, Yancopoulos G,
18. Mustacich DJ, Kylat RI, Bernas MJ, Myles RJ, Jones JA, Kanady Erickson R, Witte M. Defective remodeling and maturation of
JD, Simon AM, Georgieva TG, Witte MH, Erickson RP, Pires PW. the lymphatic vasculature in Angiopoietin-2 deficient mice.
Abnormal lymphatic phenotype in a crispr mouse model of the Dev Biol. 2008;319(2):309–20. https://doi.org/10.1016/j.
human lymphedema-causing connexin47 R260c point mutation. ydbio.2008.04.024.
Lymphology. 2021;54:78–91. 31. Souma T, Thomson BR, Heinen S, Carota IA, Yamaguchi S, Onay
19. Finegold DN, Baty CJ, Knickelbein KZ, Perschke S, Noon SE, T, Liu P, Ghosh AK, Li C, Eremina V, Hong YK, Economides AN,
Campbell D, Karlsson JM, Huang D, Kimak MA, Lawrence EC, Vestweber D, Peters KG, Jin J, Quaggin SE. Context-dependent
Feingold E, Meriney SD, Brufsky AM, Ferrell RE. Connexin 47 functions of angiopoietin 2 are determined by the endothelial phos-
mutations increase risk for secondary lymphedema following phatase VEPTP. Proc Natl Acad Sci U S A. 2018;115(6):1298–303.
breast cancer treatment. Clin Cancer Res. 2012;18(8):2382–90. https://doi.org/10.1073/pnas.1714446115.
https://doi.org/10.1158/1078-­0432.CCR-­11-­2303. 32. Leppanen VM, Brouillard P, Korhonen EA, Sipila T, Jha SK,
20. Newman B, Lose F, Kedda MA, Francois M, Ferguson K, Janda M, Revencu N, Labarque V, Fastre E, Schlogel M, Ravoet M, Singer A,
Yates P, Spurdle AB, Hayes SC. Possible genetic predisposition to Luzzatto C, Angelone D, Crichiutti G, D'Elia A, Kuurne J, Elamaa
lymphedema after breast cancer. Lymphat Res Biol. 2012;10(1):2– H, Koh GY, Saharinen P, Vikkula M, Alitalo K. Characterization
13. https://doi.org/10.1089/lrb.2011.0024. of ANGPT2 mutations associated with primary lymphedema. Sci
21. Choi D, Park E, Jung E, Seong YJ, Hong M, Lee S, Burford Transl Med. 2020;12(560) https://doi.org/10.1126/scitranslmed.
J, Gyarmati G, Peti-Peterdi J, Srikanth S, Gwack Y, Koh CJ, aax8013.
Boriushkin E, Hamik A, Wong AK, Hong YK. ORAI1 activates pro- 33. Wassef M, Blei F, Adams D, Alomari A, Baselga E, Berenstein
liferation of lymphatic endothelial cells in response to laminar flow A, Burrows P, Frieden IJ, Garzon MC, Lopez-Gutierrez JC, Lord
through Kruppel-like factors 2 and 4. Circ Res. 2017;120(9):1426– DJ, Mitchel S, Powell J, Prendiville J, Vikkula M, Committee.
39. https://doi.org/10.1161/CIRCRESAHA.116.309548. IBaS. Vascular anomalies classification: recommendations from the
22. Nonomura K, Lukacs V, Sweet DT, Goddard LM, Kanie A, Whitwam International Society for the Study of vascular anomalies. Pediatrics.
T, Ranade SS, Fujimori T, Kahn ML, Patapoutian A. Mechanically 2015;136(1):e203–14. https://doi.org/10.1542/peds.2014-­3673.
23 Lymphovascular Genomics and Proteomics, Clinical Syndromes, and Cancer Metastasis 247

34. Trenor CC 3rd, Chaudry G. Complex lymphatic anomalies. 43. Li D, March ME, Gutierrez-Uzquiza A, Kao C, Seiler C, Pinto E,
Semin Pediatr Surg. 2014;23(4):186–90. https://doi.org/10.1053/j. Matsuoka LS, Battig MR, Bhoj EJ, Wenger TL, Tian L, Robinson
sempedsurg.2014.07.006. N, Wang T, Liu Y, Weinstein BM, Swift M, Jung HM, Kaminski
35. Boscolo E, Coma S, Luks VL, Greene AK, Klagsbrun M, Warman CN, Chiavacci R, Perkins JA, Levine MA, Sleiman PMA, Hicks PJ,
ML, Bischoff J. AKT hyper-phosphorylation associated with PI3K Strausbaugh JT, Belasco JB, Dori Y, Hakonarson H. ARAF recur-
mutations in lymphatic endothelial cells from a patient with lym- rent mutation causes central conducting lymphatic anomaly treat-
phatic malformation. Angiogenesis. 2015;18(2):151–62. https:// able with a MEK inhibitor. Nat Med. 2019;25(7):1116–22. https://
doi.org/10.1007/s10456-­014-­9453-­2. doi.org/10.1038/s41591-­019-­0479-­2.
36. Osborn AJ, Dickie P, Neilson DE, Glaser K, Lynch KA, Gupta 44. Croteau SE, Kozakewich HP, Perez-Atayde AR, Fishman SJ,
A, Dickie BH. Activating PIK3CA alleles and lymphangiogenic Alomari AI, Chaudry G, Mulliken JB, Trenor CC 3rd. Kaposiform
phenotype of lymphatic endothelial cells isolated from lymphatic lymphangiomatosis: a distinct aggressive lymphatic anom-
malformations. Hum Mol Genet. 2015;24(4):926–38. https://doi. aly. J Pediatr. 2014;164(2):383–8. https://doi.org/10.1016/j.
org/10.1093/hmg/ddu505. jpeds.2013.10.013.
37. Lala S, Mulliken JB, Alomari AI, Fishman SJ, Kozakewich 45. Ozeki M, Fujino A, Matsuoka K, Nosaka S, Kuroda T, Fukao
HP, Chaudry G. Gorham-Stout disease and generalized lym- T. Clinical features and prognosis of generalized lymphatic anom-
phatic anomaly-clinical, radiologic, and histologic differen- aly, Kaposiform Lymphangiomatosis, and Gorham-Stout disease.
tiation. Skelet Radiol. 2013;42:917–24. https://doi.org/10.1007/ Pediatr Blood Cancer. 2016;63(5):832–8. https://doi.org/10.1002/
s00256-­012-­1565-­4. pbc.25914.
38. Luks VL, Kamitaki N, Vivero MP, Uller W, Rab R, Bovee JV, 46. Manevitz-Mendelson E, Leichner GS, Barel O, Davidi-Avrahami I,
Rialon KL, Guevara CJ, Alomari AI, Greene AK, Fishman SJ, Ziv-Strasser L, Eyal E, Pessach I, Rimon U, Barzilai A, Hirshberg
Kozakewich HP, Maclellan RA, Mulliken JB, Rahbar R, Spencer A, Chechekes K, Amariglio N, Rechavi G, Yaniv K, Greenberger
SA, Trenor CC 3rd, Upton J, Zurakowski D, Perkins JA, Kirsh A, S. Somatic NRAS mutation in patient with generalized lym-
Bennett JT, Dobyns WB, Kurek KC, Warman ML, SA MC, Murillo phatic anomaly. Angiogenesis. 2018;21(2):287–98. https://doi.
R. Lymphatic and other vascular malformative/overgrowth dis- org/10.1007/s10456-­018-­9595-­8.
orders are caused by somatic mutations in PIK3CA. J Pediatrics. 47. Barclay SF, Inman KW, Luks VL, McIntyre JB, Al-Ibraheemi
2015;166(4):1048–54. e1041–1045. https://doi.org/10.1016/j. A, Church AJ, Perez-Atayde AR, Mangray S, Jeng M, Kreimer
jpeds.2014.12.069. SR, Walker L, Fishman SJ, Alomari AI, Chaudry G, Trenor Iii
39. Rodriguez-Laguna L, Agra N, Ibanez K, Oliva-Molina G, Gordo G, CC, Adams D, Kozakewich HPW, Kurek KC. A somatic activat-
Khurana N, Hominick D, Beato M, Colmenero I, Herranz G, Torres ing NRAS variant associated with kaposiform lymphangioma-
Canizalez JM, Rodriguez Pena R, Vallespin E, Martin-Arenas R, tosis. Genet Med. 2019;21(7):1517–24. https://doi.org/10.1038/
Del Pozo A, Villaverde C, Bustamante A, Ayuso C, Lapunzina P, s41436-­018-­0390-­0.
Lopez-Gutierrez JC, Dellinger MT, Martinez-Glez V. Somatic 48. Ozeki M, Aoki Y, Nozawa A, Yasue S, Endo S, Hori Y, Matsuoka
activating mutations in PIK3CA cause generalized lymphatic K, Niihori T, Funayama R, Shirota M, Nakayama K, Fukao
anomaly. J Exp Med. 2019;216(2):407–18. https://doi.org/10.1084/ T. Detection of NRAS mutation in cell-free DNA biological fluids
jem.20181353. from patients with kaposiform lymphangiomatosis. Orphanet J Rare
40. Adams DM, Trenor CC 3rd, Hammill AM, Vinks AA, Patel MN, Dis. 2019;14(1):215. https://doi.org/10.1186/s13023-­019-­1191-­5.
Chaudry G, Wentzel MS, Mobberley-Schuman PS, Campbell LM, 49. Foster JB, Li D, March ME, Sheppard SE, Adams DM, Hakonarson
Brookbank C, Gupta A, Chute C, Eile J, McKenna J, Merrow AC, H, Dori Y. Kaposiform lymphangiomatosis effectively treated
Fei L, Hornung L, Seid M, Dasgupta AR, Dickie BH, Elluru RG, with MEK inhibition. EMBO Mol Med. 2020:e12324. https://doi.
Lucky AW, Weiss B, Azizkhan RG. Efficacy and safety of Sirolimus org/10.15252/emmm.202012324.
in the treatment of complicated vascular anomalies. Pediatrics. 50. Dellinger MT, Garg N, Olsen BR. Viewpoints on vessels and
2016;137(2):1–10. https://doi.org/10.1542/peds.2015-­3257. vanishing bones in Gorham-Stout disease. Bone. 2014;63:47–52.
41. Ricci KW, Hammill AM, Mobberley-Schuman P, Nelson SC, Blatt https://doi.org/10.1016/j.bone.2014.02.011.
J, Bender JLG, McCuaig CC, Synakiewicz A, Frieden IJ, Adams 51. Gorham LW, Stout AP. Massive osteolysis (acute spontane-
DM. Efficacy of systemic sirolimus in the treatment of general- ous absorption of bone, phantom bone, disappearing bone);
ized lymphatic anomaly and Gorham-Stout disease. Pediatr Blood its relation to hemangiomatosis. J Bone Joint Surg Am.
Cancer. 2019;66(5):e27614. https://doi.org/10.1002/pbc.27614. 1955;37-A(5):985–1004.
42. Strychowsky JE, Rahbar R, O'Hare MJ, Irace AL, Padua H, Trenor 52. Nozawa A, Ozeki M, Niihori T, Suzui N, Miyazaki T, Aoki Y. A
CC 3rd. Sirolimus as treatment for 19 patients with refractory cervi- somatic activating KRAS variant identified in an affected lesion of
cofacial lymphatic malformation. Laryngoscope. 2018;128(1):269– a patient with Gorham-Stout disease. J Hum Genet. 2020; https://
76. https://doi.org/10.1002/lary.26780. doi.org/10.1038/s10038-­020-­0794-­y.
 Imaging of the Lymphatic System
with Relevance to Cancer and Cancer 24
Metastasis

Russell S. Witte and Michael Bernas

Abstract 24.1 Introduction

Imaging plays a crucial role in the fundamental under-
standing of lymphatic system structure and function There are many techniques currently in use for imaging the
throughout the body both in health and a wide variety of lymphatic system, and they all focus on producing images to
clinical disorders. It also has application to identify pri- help understand the structure and function of the lymphatic
mary and metastatic cancer spread to regional nodes/ system in health and disease. These various imaging tech-
organs as well as its importance in cancer assessment/ niques and technologies have unique strengths and weak-
treatment planning using the sentinel node protocol. nesses, including their application in the setting of cancer
There are multiple different techniques utilizing associ- metastasis.
ated tracers for in vivo dynamic imaging in both humans Although in some cases lymphatic vessels and nodes can
and experimental models, and the future continues to be visualized without contrast, most techniques require con-
offer advancements including theranostics. Each tech- trast agents of relatively large molecules (the size of albumin
nique has strengths and weaknesses for specific applica- and greater), which are selectively taken up by lymphatic
tions. Individual techniques (and sometimes in capillaries rather than blood vessels when deposited intra-
combination) produce images for the diagnosis, treatment dermally, subcutaneously, interstitially, directly into lymph
monitoring, and even to develop new treatment strategies nodes, or by image-guided direct puncture of large channels.
for cancer and cancer-related sequelae. Tracers have different and unique characteristics depending
on the modality utilized and the individual energy sources
(X-ray, gamma ray, near infrared, magnetic, and ultrasound)
Learning Objectives as they produce images of the lymphatic system at different
• Describe more than five techniques currently used spatial resolutions, fields-of-view, and penetration depths.
to image the lymphatic system. Images depict both functional and dynamic flow through the
• List at least one advantage and one disadvantage to lymphatics as well as the anatomical structure of nodes and
each of the techniques. channels. The most common imaging techniques and appli-
• Delineate at least four applications of lymphatic cations are described below.
imaging in the management of patients with meta-
static cancer.
24.2 Modalities

24.2.1 X-Ray or CT Lymphography (LAG)

24.2.1.1 Technique Summary

R. S. Witte (*)
Classical LAG is performed by first identifying a pedal lym-
Department of Medical Imaging, University of Arizona,
Tucson, AZ, USA phatic through the use of blue dye and then cannulation for
e-mail: [email protected] instilling the contrast agent [1, 2]. The agent (ethiodized oil
M. Bernas or injecting iodized water-soluble solution [300 mg iodine/
Department of Medical Education, TCU School of Medicine, mL]) is injected at a very slow rate (6 mL/h) in order to not
Fort Worth, TX, USA cause rupture of the fine peripheral lymphatics. Patients must
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 249
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_24
250 R. S. Witte and M. Bernas

remain relatively immobile for 60–90 min. A drawback of cases of chylous or non-chylous reflux where a surgeon or
this procedure is the invasive cut down needed to cannulate interventional radiologist can subsequently perform the
the lymphatic vessels as well as the time for patient immobi- treatment to inhibit leakage. Recent work using intranodal
lization. Whereas early studies required X-ray systems for injection has clearly demonstrated the thoracic duct includ-
imaging, CT has become more widely used due to its 3D ing identification of areas of leakage [3] (Fig. 24.1) and has
capability particularly for those cases involving chylous greatly impacted the field of lymphatic imaging and inter-
leakage. An obvious advantage of this technique is that both ventions [5]. Using US guidance, the placement of a needle
X-ray and CT systems are widely available throughout the directly into a node allows injection of the oil-based iodin-
world. However, there require exposure to ionizing radiation. ated contrast. These advances have been adopted worldwide
LAG can also be performed by an intranodal injection of the and helped bring lymphatic interventional treatment proce-
contrast agent for access to the deeper lymphatic structures dures to the forefront [6]. A drawback is that the oil-based
of the cisterna chyli and thoracic duct [3, 4]. contrast agent has increased viscosity which can inhibit clear
visualization of a leakage.
24.2.1.2 Imaging Summary
From its discovery in the 1950s conventional bipedal oily
contrast direct lymphography (LAG) was widely used and 24.2.2 Lymphoscintigraphy (LAS)
considered safe for staging cancers particularly lymphomas
to determine lymph nodal involvement prior to surgery or 24.2.2.1 Technique Summary
chemotherapy, to deliver endolymphatic chemotherapy, and LAS is a primary imaging modality to study and quantify
also to delineate other nonneoplastic lymphadenopathic dis- lymphatic drainage and flow to help diagnose and stage lym-
orders [2]. LAG produces images of fine anatomical detail phatic disorders, as well as to assess and monitor treatment.
(Fig. 24.1). Although these images are static, real-time fluo- This minimally invasive technique uses a single small
roscopy can demonstrate the function of the lymphatic sys- (0.05 mL) injection into each hand or foot. The standard
tem. Intranodal LAG is particularly invaluable for precise radiotracer (in the United States) is Tc99m filtered sulfur
anatomic depiction of the lymphatic vessels and nodes in colloid with particle sizes of approximately 50 nm.

a b c

Fig. 24.1 Lymphography. Classical pedal lymphography (a and b) in lower image, “packets” of lymph are clearly demonstrated depicting the
a man (a) clearly displaying normal lymphatic imaging in the upper noncontinuous column of lymphatic fluid in normal flow. (c)
thigh with discrete lymphatic vessels (arrowheads) and lymph nodes Fluoroscopic image of the thoracic duct (arrow) following injection
(arrows). Lymphography (captured by fluoroscopy) in experimental intranodal injection of the contrast. The injected contrast is seen being
sheep imaging demonstrating thoracic duct in the upper chest (upper transported into the subclavian vein at thoracic duct to subclavian vein
image arrows) as well as in the abdomen (lower image arrows). In the junction (arrowhead). (a modified from [2], c modified from [3])
24 Imaging of the Lymphatic System with Relevance to Cancer and Cancer Metastasis 251

Technetium has a half-life of only 6 h, making it available to staging as well as helping to delineate potential approaches
almost any patient with few restrictions. It also has an excel- for future treatment in patients with complications such as
lent safety profile (total radiation dose is approximately chylous ascites or chylothorax (either due to situations
1/20th of a chest X-ray), readily available, and relatively involving cancer or not) [7]. See Ref. [11] for an extensive
inexpensive. A drawback concerning LAS is that it delivers review of LAS imaging depicting multiple patients with vari-
poor anatomical resolution that can limit its role in both ous lymphatic abnormalities and clinical presentations as
diagnostics and pre-treatment procedural planning. A gamma well as their corresponding LAS images. The main limita-
camera is utilized for imaging and it can visualize the func- tion of standard LAS is poor spatial resolution and lack of
tional lymphatic system from the distal extremities to the anatomical landmarks. SPECT/CT (below) can be added to
deep thoracic duct. Most systems can now produce a single provide such landmarks and regional localization of lym-
“whole body” image following bilateral simultaneous injec- phatic transport of the tracer. There have been investigators
tions (Fig. 24.2). A dual-head camera is typically used to who have proposed quantitative evaluation algorithms for
acquire early static and delayed (~3 h) dynamic images. LAS images [12, 13] as well as classification systems for
lymphedema based on LAS imaging [14].
24.2.2.2 Imaging Summary SPECT/CT addresses limitations with lymphoscintigra-
LAS is the current gold standard for imaging of the lym- phy, which produces only planar images without clear ana-
phatic system as recommended by the Consensus Document tomical landmarks. SPECT/CT overcomes this limitation by
[10]. The technique produces images for visualization of taking advantage of state-of-the-art gamma cameras com-
site(s) of obstruction (no or reduced flow), backflow, and bined with CT tomography [8, 15]. Cameras are placed at
leakage which are valuable for determining diagnosis and different angles to produce multiple 2D images and a tomo-

a b c d

Fig. 24.2 (a, b) Lymphscintigraphy. (a) 2-D whole body image of nor- was transported through the lymphatic system. (c) SPECT-CT image in
mal late (3.5 h) image of the legs depicting almost complete tracer a child demonstrating advantageous 3D localization compared to 2D
clearance through the lymphatic vessels following injection into both planar LAS and also co-registration of SPECT and CT for anatomic
feet. Symmetric, bilateral inguinal nodes, and retroperitoneal nodes are correlation. Low radioactivity appears as blue and highest radioactivity
clearly seen. Markers (m) are at the knees, pubis, xiphoid, and sterna as orange with intermediate in yellow. Patient has severe protein-losing
notch. Liver (L) can be seen in this later image as the tracer has been enteropathy. Note radiotracer highlighting the duodenal loop location
successfully transported through the lymphatic system into the blood of leakage. (d) Panoramic views of ICG fluorescence lymphography in
pool. (b) Late image of a patient with swelling of the right upper thigh a healthy volunteer demonstrate injection site (arrow) and transport of
following treatment for sarcoma. The patient developed a seroma which the tracer up to the groin. (a and b modified from Bernas [7], c modified
was being treated with a drain (arrow) and both collected tracer as it from Kuo [8], d modified from Unno [9])
252 R. S. Witte and M. Bernas

graphic reconstruction produces a 3D image with the local- 24.2.4 Magnetic Resonance
ized radioactive signal (typically produced by 99 m-Tc) Lymphangiography (MRL)
displaying lymphatic structure and function, including leak-
age (Fig. 24.2). This method is quite useful for mapping of 24.2.4.1 Technique Summary
SLNs before resection and surgical treatment for several MRL is increasingly used to visualize lymphatic vessels, cis-
types of cancer. terna chyli, and the thoracic duct in humans instead of inva-
sive lymphangiography. This is due to its increased spatial
resolution and the availability of improved equipment and
24.2.3 NIR (ICG) Imaging newer imaging sequences. MR lymphagiography can largely
be divided into two types of imaging contrast-enhanced and
24.2.3.1 Technique Summary non-contrast images. They each have their strengths and
Fluorescent contrast agents have been utilized to image the weaknesses and can produce images of high quality
anatomy and function of the lymphatic system. Typically, (Fig. 24.3).
they are injected intradermally to access the initial lymphat- MR lymphangiography without contrast can be per-
ics beneath the epidermis and then are transported through formed at most facilities with a 1.5-T or 3-T clinical MR
the lymphatic system. Some of the earliest work in humans machine. Although specific sequences vary, in general a cor-
was performed by Bollinger and his colleagues [16] using onal or sagittal heavy T2 three-dimensional (3D) volume
FITC-labeled dextran. They were able to show lymphatic sequence [15] or turbo-spin-echo (TSE) pulse train with long
microvessels since the techniques did not allow deeper imag- echo time is utilized [25]. There has also been work pub-
ing. The fluorescein sodium is excited by visible light and lished using a 3-T system and a noninvasive arterial spin
the fluorescent photons are absorbed by many common tis- labeling (ASL) MR method for measuring blood flow to
sues (blood, water, etc.) which also inhibits its usefulness for evaluate lymphatic flow [26]. The advantage of not using
deeper structures [17]. contrast agents is balanced by potential interference of large
A newer technique using indocyanine green (ICG) has pleural effusions that may be related to a lymphatic disorder
become much more widely used. ICG is excited by near-­ on image quality and respiratory artifact. In addition, visual-
infrared light (700–900 nm) and that is why these techniques ization of smaller lymphatic vessels in the periphery is not as
are termed near-infrared (NIR) imaging (also sometimes successful without contrast [15].
referred to as ICG imaging) [18]. Since near-infrared light Dynamic contrast-enhanced MR lymphangiography was
can penetrate 3–4 cm beneath the tissue surfaces, there is initially described in 2015 [27]. It is performed by direct
minimal autofluorescence and, as a result, NIR imaging with intranodal injection of diluted gadolinium into the inguinal
ICG has been more recently employed for imaging the initial lymph nodes using ultrasound guidance [3, 28]. It provides
lymphatic system as well as imaging of deeper lymphatic superior information on lymphatic flow dynamics and is best
structures. There are also multiple commercial cameras and for suspected lymphatic flow pathologies and lymphatic
imaging systems, which are relatively low-cost and in use in malformations that may be seen in the setting of cancer.
a variety of settings. A drawback of ICG imaging is its lack Opposed to non-contrast imaging, dynamic contrast-­
of whole-body imaging due to relatively small fields of view enhanced MR lymphangiography of the extremities can be
from the cameras and limited depth resolution. obtained using a subdermal injection of gadolinium contrast
media [15, 29].
24.2.3.2 Imaging Summary
NIR imaging using ICG has become more common to assist 24.2.4.2 Imaging Summary
in the diagnosis of lymphedema [9, 19], and imaging pat- Clear images of central lymphatic structures can be obtained
terns suggesting different levels of lymphatic dysfunction in using either contrasted or non-contrast MR lymphangiogra-
lymphedema have been defined [20]. There have also been phy (Fig. 24.3). Imaging of the central structures and their
publications showing lymphatic system changes in the set- pathology (leaks, obstructions, lymphatic aplasia, hypo- and
ting of cancer [21, 22] as well as used to visualize and quan- hyperplasia, etc.) has opened the window into designing and
titate in vivo lymphatic function [18]. Additionally, ICG performing various treatments. The first percutaneous lym-
imaging has moved into the operating room as a planning phatic intervention in the setting of the chylous leak was per-
tool for microsurgical anastomosis of lymphatic to venous formed by Cope in 2002 [30]. Since that time and with the
vessels for the treatment of lymphedema and surgical resec- newer imaging, multiple interventional treatments have been
tion of sentinel lymph nodes [23, 24]. carried out for a variety of lymphatic abnormalities [3, 4, 31].
24 Imaging of the Lymphatic System with Relevance to Cancer and Cancer Metastasis 253

a b

Fig. 24.3 Contrasted and non-contrasted MR lymphangiography. (a) with gadolinium-based contrast agent in a dynamic contrast-enhanced
Non-contrasted MR imaging using heavily T2-weighted sequence MR lymphangiography (DCRML) displays visualization of a normal
clearly displays the lower portion of the thoracic duct (small arrows) single thoracic duct (white arrow). (a modified from Schwartz et al.
and the cisterna chyli (arrowhead). Contrasted MR lymphangiography [15], b modified from Itkin et al. [3])

24.2.5 U
 ltrahigh Frequency Ultrasound poor specificity for discriminating normal, malignant, and
(UHFUS) Imaging metastatic tissue. For example, a second Multicenter
Selective Lymphadenectomy Trial (MSLT-II) revealed a sen-
24.2.5.1 Technique Summary sitivity of 6.6%, increasing only slightly with higher risk pri-
For decades, pulse-echo ultrasound has been used as a non- maries [36]. However, as the ultrasound frequency increases,
invasive screening tool to assess lymph node status at the so does the spatial resolution. New commercial systems like
submillimeter scale. Conventional ultrasound detects struc- the Vevo MD (Visualsonics by Fujifilm) offer UFUS probes
tural changes suggestive of lymph node metastases, such as (>50 MHz) for high-resolution imaging of lymphatic chan-
enlargement, irregular borders, and hypoechogenicity nels approaching the scale of single cells (<30 μm). The pri-
[32–35]. While standard B mode ultrasound is sensitive to mary disadvantage of UFUS is that the attenuation increases
subtle differences in acoustic backscatter, it is not commonly dramatically at frequencies >50 MHz, limiting the depth of
used for defining lymph node status because it typically has penetration to ~10 mm.
254 R. S. Witte and M. Bernas

24.2.5.2 Imaging Summary 24.2.6.2 Imaging Summary

Resolution improves dramatically as the ultrasound fre- CEUS with microbubbles has great potential to identify senti-
quency increases. A recent paper evaluated ultrahigh fre- nel lymph nodes and aid with axillary staging in breast cancer
quency ultrasound (UHFUS, 48 and 70 MHz) for assessing patients. The method can be performed noninvasively with a
the structure and degeneration status of human peripheral standard clinical scanner without the need for radioactive con-
lymphatic vessels prior to surgical resection [37]. There was trast agents. A recent study in rabbits improved the feasibility
good agreement between UHFUS captured on the Vevo MD of CEUS for mapping peripheral lymphatic vessels and nodes
system (Visualsonics, Fujifilm) and subsequent histological by implementing high frame rates to reduce tissue artifacts,
examination of the lumen inner and outer diameters and wall limit bubble disruption, and improve SNR for detecting slow
thickness. The data support UHFUS as a noninvasive tech- flow with Doppler-based imaging techniques [42]. A similar
nique for planning and guiding lymphatic surgery by helping approach has recently been applied to diagnose central lymph
to identify lymphatic channels suitable for anastomosis. node metastasis in papillary thyroid cancer [42]. Another
Although UHFUS is suitable for mapping superficial struc- group demonstrated deep contrast-­enhanced ultrasound imag-
tures, penetration depth is limited to <10 mm. ing of lymphatic channels and sentinel nodes with Sonzazoid
in a porcine model [43]. In recent years, there has been an
increased interest in applying CEUS to evaluate lymphatic
24.2.6 C
 ontrast-Enhanced Ultrasound (CEUS) channels and sentinel lymph nodes in breast cancer patients
Imaging for presurgical planning and needle biopsy guidance [44–46].
Other investigators have combined CEUS, B mode, and
24.2.6.1 Technique Summary Doppler flow imaging to improve sensitivity and specificity
Contrast-enhanced ultrasound (CEUS), on the other hand, for detecting lymph node metastasis. Figure 24.4 provides an
employs a subcutaneous injection of microbubbles (2–6 μm) example of an intravenous injection of SonoVue microbubbles
to enhance the contrast of the lymphatic system. CEUS pro- for comparing metastatic and nonmetastatic cervical lymph
vides a real-time functional assessment of transport into and nodes using CEUS in patients with papillary thyroid carci-
out of peripheral lymphatic vessels and nodes [38]. Although noma [47].
microbubble contrast agents are FDA-approved for cardiac
imaging and other clinical applications, CEUS is not a stan-
dard for imaging the human lymphatic system, and the 24.2.7 Multispectral Photoacoustics
method remains primarily a preclinical technique for inves-
tigating function and dysfunction of the lymphatic system 24.2.7.1 Technical Summary
and developing novel therapeutics. An alternative approach Photoacoustic imaging (PAI) is a hybrid technique that uses
for CEUS uses an intravenous (IV) injection of microbub- a nanosecond laser pulse to transiently heat the tissue and,
bles combined with color Doppler to enhance the blood due to thermoelastic expansion, generate ultrasonic waves
vascular network and help classify lymph nodes as that are detected to produce images proportional to the opti-
metastatic, tuberculous, cystic, or enlarged secondary to cal absorption of the underlying structures. PAI is a scalable
lymphoma [39–41]. technology that can provide real-time, high resolution

a b

Fig. 24.4 CEUS and B mode US images of the thyroid and cervical Doppler to identify nodal metastasis and develop predictive models of
lymph node from a patient with nonmetastatic (a) and another with metastatic disease based on preoperative clinical and ultrasonographic
metastatic (b) papillary thyroid carcinoma confirmed by histology. The features [47]
authors of this study combined features from CEUS, B mode, and
24 Imaging of the Lymphatic System with Relevance to Cancer and Cancer Metastasis 255

(<200 μm) images deep into tissue (>20 mm), including vol- cles were injected into the footpad and within 24 h filled the
umetric images of the human breast within a single breath sentinel lymph node that was not tumor-bearing metastatic.
hold [48, 49]. Without contrast agents, PAI is able to map However, the nanoparticles were not able to accumulate inside
blood vessels and blood oxygen saturation (SO2) based on nodes that were expressed metastatic disease (Fig. 24.5). The
the optical absorption properties of hemoglobin. Melanoma high specificity of the results in mice suggests that PAI could
cells also have strong near-infrared absorption and can be be a valuable tool for both diagnosing metastatic disease and
identified and tracked with PAI. Near-infrared absorbing monitoring effects of treatment (i.e., theranostics).
tracers, such as FDA-approved dyes ICG and Methylene A large-field-of-view photoacoustic platform has recently
Blue, are required to visualize lymphatic vessels and nodes been developed for mapping lymphatic vessels in the human
with PAI [50]. By sweeping the optical wavelength, multi- hand, foot, or limb [54, 55]. ICG was injected intradermally or
spectrum PAI with spectral unmixing is able to display both subdermally into the webspace of the foot in both healthy sub-
blood and lymphatic vessels in the same image. Because it is jects and patients with secondary lymphedema (Fig. 24.6).
inherently a 3D modality, PAI has better depth resolution Photoacoustic lymphangiography revealed 3D complex lym-
than fluorescent imaging with better depth penetration. phatic structures near regions of dermal backflow (DBF),
However, PAI requires custom or commercial hardware that which was confirmed on lymphoscintigraphy and NIR fluores-
integrates a tunable pulsed laser with ultrasound technology, cent imaging. Compared to other modalities, the system pro-
making it cost-prohibitive at the present time compared to vided 3D high-resolution images (<200 μm) with a penetration
standard NIR fluorescent imaging with ICG. PAI also typi- depth > 20 mm. In addition, the platform was able to identify
cally has worse sensitivity than fluorescent and nuclear small blood vessels near the lymphatic structures and distin-
imaging techniques for detecting contrast agents. guish arteries from veins based on the intrinsic NIR absorption
of hemoglobin and the dependence of oxygen on the absorp-
24.2.7.2 Imaging Summary tion spectrum. One of the clinical applications of the imaging
Preclinical studies have demonstrated the capability of real-­ system is for postoperative assessment of anastomoses.
time PAI to quantify lymphatic flow, pumping, and pharmaco- Clinical studies have also been performed using multi-
kinetics using ICG tracers and encapsulated nanoparticles spectral photoacoustic imaging to evaluate metastatic status
[51]. Another study in mice developed a targeted contrast in sentinel lymph nodes (SLNs) based on the presence of
agent for immunofunctional photoacoustic imaging of lymph melanoma cells [56, 57]. However, questions remain regard-
node metastasis [52]. Glycol-chitosan-coated gold nanoparti- ing the sensitivity of the system, and large clinical trials are

a Naïve/Control
d % of SLN with GC-AU NPs (after 24 hr)
100 %

90 %
**
Nanoparticle-to-tissue ratio in SLN (%)

80 %
***
70 %
b Non-Metastatic 60 %

50 %

40 %

30 %

20 %
c Metastatic
10 %

0%
Naive mouse Tumor bearing Tumor bearing
Non-metastatic Lymph node metastatic

Fig. 24.5 Detection of lymph node metastasis with targeted gold pulse-echo image displaying anatomical structure; (d) Nanoparticle-to-­
nanoparticles and photoacoustic imaging in mice. (a) PAI (hot colors) tissue ratio for each type of mouse, indicating that mice with metastatic
of sentinel lymph node and afferent lymphatic vessel 24 h after subcu- disease had poorly functioning SNLs compared to the other categories
taneous injection of GC-AuNPs in the footpad in (a) control; (b) non- (modified from [52, 53])
metastatic; and (c) metastatic mice. PAI is superimposed with grayscale
256 R. S. Witte and M. Bernas

a b c

Fig. 24.6 Wide-field photoacoustic lymphangiography (PAL) in male ing a subdermal injection of ICG) and veins identified by the photo-
patient with secondary lymphedema. (a) Top view photograph of upper acoustic signal corresponding to deoxygenated hemoglobin. (c)
limb before lymphaticovenular anastomosis surgery with red and black Zoomed in representation of dotted oval (in a and b) clearly illustrating
marks denoting approximate location of lymphatic vessels identified the high resolution (<200 μm) and deep imaging capability (>20 mm)
with NIR and small veins identified with B-mode US, respectively. (b) of PAL. Compared to control images, patients with secondary lymph-
Corresponding 3D representation of the forearm captured with PAL edema manifested a much higher quantity of lymphatic vessels, which
indicating complex network of lymphatic channels (enhanced follow- also appeared twisted and tortuous [55]

necessary to determine the efficacy of noninvasive PAI for the book. Applications are found in basic science investiga-
detecting SLN metastases compared to the gold standard. tions of in vitro as well as in vivo phantom and animal
Another approach applies real-time, label-free photoacoustic models to delineate mechanisms, discover and improve
detection and tracking of circulating tumor cells (CTCs) and imaging techniques, and test potential therapeutic targets.
distant metastasis [58]. The photoacoustic device is called These findings as well as clinical and translational research
the Cytophone and provides a liquid biopsy based on the can then be utilized in clinical settings to determine (singu-
intrinsic contrast of circulating melanoma cells, red blood larly or in multimodal combinations) whether, when, and
cells, and other intrinsic optical absorbers. The system was where: specific cancers metastasize; the impact and altera-
also combined as a theranostic tool by counting CTCs before tions of the lymphatic system and flow; and potential ave-
and after photodynamic therapy consisting of continuous nues and points for the treatment and monitoring of both
laser exposure in the cubital vein for 1 h. the lymphatic and cancer alterations [59]. These treatments
could entail image-­guided therapy (and monitoring) as well
as in select instances (i.e., high-frequency ultrasound) use
24.3  pplication and Use in Cancer
A of imaging technology itself to treat primary cancer and
Metastasis metastases.
Table 24.1 summarizes the most common and emerging
Lymphatic imaging is invaluable to broadly understanding, imaging modalities and their attributes for studying the lym-
evaluating, and treating cancer metastasis as illustrated phatic system with relevance to cancer and cancer
throughout this chapter as well as in most other chapters of metastasis.
24 Imaging of the Lymphatic System with Relevance to Cancer and Cancer Metastasis 257

Table 24.1 Common and emerging imaging modalities and their attributes for studying the lymphatic system with relevance to cancer and cancer
metastasis
Modality Clinical tracer Basic Clinical Spatial resolution Contrast resolution Penetration Sensitivity
Lymphography-classical Lipidol X + +++ +++ +++
Lymphography- intra-nodal Lipidol X + +++ +++ +++
Lymphscintigraphy 99mTc X X + +++ +++ +++
Lymphscintigraphy- CT-SPECT Lipidol, 99mTc X ++ +++ +++ +++
MRI Gd (optional) X X ++ ++ +++ ++
NIR ICG X X +++ +++ + +++
US Microbubbles X X +++ ++ +++ ++
PAI ICG, meth. Blue X X +++ +++ ++ +

24.4 Conclusions • Which techniques (or combinations) are optimal for

imaging primary and metastatic cancers?
Imaging of the lymphatic system is crucial for depicting both • Can central (thoracic duct) access be perfected to
lymphatic structure and function in health and a wide variety facilitate stenting into venous system to reverse tho-
of clinical disorders. The multiple different techniques racic duct obstruction, i.e., for decompression after
described utilize various tracers and techniques for humans tumor invasion and lymph leakage? For sampling
and experimental models and each have their own strengths, lymph noninvasively as “liquid biopsy” to detect
weaknesses, and specific applications (summarized in and characterize cancers?
Table 24.1). Beginning with a history of postmortem dissec- • Would endolymphatic therapy provide a better
tions and contrast injections centuries ago, these techniques approach to regional chemotherapy in certain cancers?
have evolved from simple intradermal visualization of blue • Can the lymphatic system itself (absorption and
dye-colored lymphatic “streamers” and invasive oil-contrast transport of various agents) be utilized for cancer
lymphography to radionuclide lymphoscintigraphy, single treatment?
photon-emission computed tomography, magnetic resonance • What other techniques are being developed that use
imaging with and without contrast, a variety of fluorescent various energies (nuclear, light, sound, or magne-
imaging, and most-recently, photoacoustic applications. tism) and tracers (e.g., dyes, nanoparticles) for
These techniques are now used in combinations and particu- imaging the lymphatic system?
larly for image-guided procedures and to directly intervene • Does multispectral photoacoustic imaging have
in an expanding number of lymphatic disorders. Applications sufficient sensitivity and specificity to assess lymph
to cancer and cancer metastasis are extensive, ranging from node status and reliably detect metastasis?
detection and monitoring of cancer spread using molecular
tools to endolymphatic administration of anticancer drugs
and decompression of central lymphatic obstruction. The
future is encouraging as the field continues to offer advance- References
ments and additional applications.
1. Kinmonth JB, Taylor GW, Harper RK. Lymphangiography; a tech-
nique for its clinical use in the lower limb. BMJ. 1955;1(4919):940–2.
2. Kinmonth JB. The lymphatics: diseases, lymphography and sur-
gery. Baltimore: Williams & Wilkins; 1972.
Unanswered Questions
3. Itkin M, Nadolski GJ. Modern techniques of lymphangiography and
• There are many unanswered questions as well as interventions: current status and future development. Cardiovasc
points to consider for advancing the field in the Intervent Radiol. 2018;41(3):366–76. https://doi.org/10.1007/
future. Some of these include: s00270-­017-­1863-­2. Epub 2017 Dec 18
4. Dori Y. Novel lymphatic imaging techniques. Tech Vasc
• How can devices, e.g., gamma camera, MRI be
Interv Radiol. 2016;19(4):255–61. https://doi.org/10.1053/j.
made more portable and less expensive to facilitate tvir.2016.10.002.
point-of-care use? 5. Guermazi A, Brice P, Hennequin C, Sarfati E. Lymphography: an old
• How can spatial resolution and depth penetration be technique retains its usefulness. Radiographics. 2003;23:1541–60.
6. Peterss S, Gratz KF, Khaladj N, Hagl C, Warnecke G, Fischer S,
improved for each imaging modality?
et al. Lymphoscintigraphic localization of a high-output chylous
• What multimodal combinations are optimal for leak after bilateral lung transplantation. J Heart Lung Transpl.
imaging specific lymphatic disorders and for inter- 2011;30:481–3.
vention in reversing pathologic processes?
258 R. S. Witte and M. Bernas

7. Bernas M. Imaging of the lymphatic system. In: Lymphedema 22. Aldrich MB, Rasmussen JC, Fife CE, Shaitelman SF, Sevick-­
management: the comprehensive guide for practitioners. 3rd ed. Muraca EM. The development and treatment of lymphatic dys-
New York: Thieme; 2013. function in cancer patients and survivors. Cancers (Basel).
8. Kuo PH, Barber BJ, Kylat RI, Klewer SE, Behan S, Lau-Braunhut 2020;12(8):2280. https://doi.org/10.3390/cancers12082280.
S, Bernas MJ, Moedano L, Bedrick AD, Mustacich DJ, Witte 23. Chang EI, Skoracki RJ, Chang DW. Lymphovenous anastomosis
MH. Whole-body lymphangioscintigraphy and SPECT/CT in chil- bypass surgery. Semin Plast Surg. 2018;32(1):22–7. https://doi.
dren with lymphatic complications after surgery for complex con- org/10.1055/s-­0038-­1636510.
genital heart disease. Lymphology. 2019;52(4):157–65. 24. Hackethal A, Hirschburger M, Eicker SO, Mücke T, Lindner C,
9. Unno N, Nishiyama M, Suzuki M, Yamamoto N, Inuzuka K, Sagara Buchweitz O. Role of Indocyanine green in fluorescence imag-
D, Tanaka H, Konno H. Quantitative lymph imaging for assessment ing with near-infrared light to identify sentinel lymph nodes, lym-
of lymph function using indocyanine green fluorescence lymphog- phatic vessels and pathways prior to surgery: a critical evaluation of
raphy. Eur J Vasc Endovasc Surg. 2008;36(2):230–6. https://doi. options. Geburtshilfe Frauenheilkd. 2018;78(1):54–62. https://doi.
org/10.1016/j.ejvs.2008.04.013. Epub 2008 Jun 4 org/10.1055/s-­0043-­123937.
10. International Society of Lymphology. The diagnosis and treatment 25. Crescenzi R, Donahue PMC, Hartley KG, Desai AA, Scott AO,
of peripheral lymphedema: 2020 consensus document of the inter- Braxton V, Mahany H, Lants SK, Donahue MJ. Lymphedema eval-
national society of lymphology. Lymphology. 2020;53:3–19. uation using noninvasive 3T MR lymphangiography. J Magn Reson
11. Witte CL, Witte MH, Unger EC, Williams WH, Bernas MJ, McNeill Imaging. 2017;46(5):1349–60. https://doi.org/10.1002/jmri.25670.
GC, Stazzone AM. Advances in imaging of lymph flow disorders. Epub 2017 Feb 28
Radiographics. 2000;20(6):1697–719. 26. Rane S, Donahue PMC, Towse T, Ridner S, Chappell M, Jordi
12. Kleinhans E, Baumeister RG, Hahn D, Siuda S, Büll U, Moser J, Gore J, Donahue MJ. Clinical feasibility of noninvasive visu-
E. Evaluation of transport kinetics in lymphoscintigraphy: alization of lymphatic flow with principles of spin labeling MR
­follow-­up study in patients with transplanted lymphatic vessels. imaging: implications for lymphedema assessment. Radiology.
Eur J Nucl Med. 1985;10(7–8):349–52. https://doi.org/10.1007/ 2013;269:893–902.
BF00251310. 27. Krishnamurthy R, Hernandez A, Kavuk S, Annam A, Pimpalwar
13. Cambria RA, Gloviczki P, Naessens JM, Wahner HW. Noninvasive S. Imaging the central conducting lymphatics: initial experience with
evaluation of the lymphatic system with lymphoscintigraphy: a pro- dynamic MR lymphangiography. Radiology. 2015;274(03):871–8.
spective, semiquantitative analysis in 386 extremities. J Vasc Surg. 28. Chavhan GB, Amaral JG, Temple M, Itkin M. MR lymphangiog-
1993;18(5):773–82. https://doi.org/10.1067/mva.1993.50510. raphy in children: technique and potential applications. Radiogr
14. Pecking AP, Alberini JL, Wartski M, Edeline V, Cluzan Radiol Soc N Am. 2017:1775–90.
RV. Relationship between lymphoscintigraphy and clinical findings 29. Pieper CC. Nodal and Pedal MR lymphangiography of the cen-
in lower limb lymphedema (LO): toward a comprehensive staging. tral lymphatic system: techniques and applications. Semin
Lymphology. 2008;41(1):1–10. Epub 2008/06/28 Intervent Radiol. 2020;37(3):250–62. https://doi.org/10.1055/s-­-
15. Schwartz FR, James O, Kuo PH, Witte MH, Koweek LM, Pabon-­ 0040-­1713442. Epub 2020 Jul 31
Ramos WM. Lymphatic imaging: current noninvasive and invasive 30. Cope C, Kaiser LR. Management of unremitting chylothorax by
techniques. Semin Intervent Radiol. 2020;37(3):237–49. https:// percutaneous embolization and blockage of retroperitoneal lym-
doi.org/10.1055/s-­0040-­1713441. Epub 2020 Jul 31 phatic vessels in 42 patients. J Vasc Interv Radiol. 2002;13:1139–
16. Bollinger A, Jäger K, Sgier F, Seglias J. Fluorescence micro- 48. 374
lymphography. Circulation. 1981;64(6):1195–200. https://doi. 31. Dori Y, Keller MS, Fogel MA, Rome JJ, Whitehead KK, Harris
org/10.1161/01.cir.64.6.1195. MA, Itkin M. MRI of lymphatic abnormalities after functional
17. O'Donnell TF Jr, Rasmussen JC, Sevick-Muraca EM. New diagnos- single-ventricle palliation surgery. AJR. 2014;203:426–31.
tic modalities in the evaluation of lymphedema. J Vasc Surg Venous 32. Vassallo P, Wernecke K, Roos N, Peters PE. Differentiation of
Lymphat Disord. 2017;5(2):261–73. https://doi.org/10.1016/j. benign from malignant superficial lymphadenopathy: the role
jvsv.2016.10.083. Epub 2017 Jan 16. Epub 2017 Jan 16. PMID: of high-resolution US. Radiology. 1992;183(1):215–20. (ISSN:
28214496; PMCID: PMC5325714 0033–8419)
18. Rasmussen JC, Tan IC, Marshall MV, Fife CE, Sevick-Muraca 33. Tregnaghi A, De Candia A, Calderone M, Cellini L, Rossi CR,
EM. Lymphatic imaging in humans with near-infrared fluorescence. Talenti E, Blandamura S, Borsato S, Muzzio PC, Rubaltelli
Curr Opin Biotechnol. 2009;20(1):74–82. https://doi.org/10.1016/j. L. Ultrasonographic evaluation of superficial lymph node metas-
copbio.2009.01.009. tases in melanoma. Eur J Radiol. 1997;24(3):216–21. ISSN 0720-­
19. Yamamoto T, Matsuda N, Doi K, Oshima A, Yoshimatsu H, 048X. https://doi.org/10.1016/S0720-­048X(96)01102-­3.
Todokoro T, Ogata F, Mihara M, Narushima M, Iida T, Koshima 34. Sureshkannan P, Vijayprabhu JR. Role of ultrasound in detection of
I. The earliest finding of indocyanine green lymphography in metastatic neck nodes in patients with oral cancer. Indian J Dent Res.
asymptomatic limbs of lower extremity lymphedema patients sec- 2011;22(3):419–23. https://doi.org/10.4103/0970-­9290.87064.
ondary to cancer treatment: the modified dermal backflow stage 35. Maxwell FM, de Margerie Mellon C, Bricout M, Cauderlier E,
and concept of subclinical lymphedema. Plast Reconstr Surg. Chapelier M, Albiter M, Bourrier P, Espié M, de Kerviler E, de
2011a;128(4):314e–21e. Bazelaire C. Diagnostic strategy for the assessment of axil-
20. Yamamoto T, Yamamoto N, Doi K, Oshima A, Yoshimatsu H, lary lymph node status in breast cancer. Diagn Interv Imaging.
Todokoro T, Ogata F, Mihara M, Narushima M, Iida T, Koshima 2015;96(10):1089–101. ISSN 2211-5684
I. Indocyanine green-enhanced lymphography for upper extremity 36. Thompson JF, Haydu LE, Uren RF, et al. Preoperative ultrasound
lymphedema: a novel severity staging system using dermal back- assessment of regional lymph nodes in melanoma patients does not
flow patterns. Plast Reconstr Surg. 2011b;128(4):941–7. https:// provide reliable nodal staging: results from a large multicenter trial.
doi.org/10.1097/PRS.0b013e3182268cd9. Ann Surg. 2019; Epub 7 Jun 2019
21. Rasmussen JC, Kwon S, Sevick-Muraca EM, Cormier JN. The 37. Bianchia A, Viscontib G, Hayashic A, Santorod A, Longoe V,
role of lymphatics in cancer as assessed by near-infrared fluores- Salgarellof M. Ultra-high frequency ultrasound imaging of lymphatic
cence imaging. Ann Biomed Eng. 2012;40:408–21. https://doi. channels correlates with their histological features: a step forward in
org/10.1007/s10439-­011-­0476-­1. lymphatic surgery. J Plast Reconstr Aesthet Surg. 2020;73:1622–9.
24 Imaging of the Lymphatic System with Relevance to Cancer and Cancer Metastasis 259

38. Choi SH, Kono Y, Corbeil J, Lucidarme O, Mattrey RF. Model ultrasound features in preoperative prediction of central lymph node
to quantify lymph node enhancement on indirect sonographic metastases of different sized papillary thyroid carcinomas. Cancer
lymphography. AJR Am J Roentgenol. 2004;183(2):513–7. Manag Res. 2021;13:3403–15. Published 2021 Apr 19. https://doi.
https://doi.org/10.2214/ajr.183.2.1830513. PMID: 15269049. org/10.2147/CMAR.S299157.
39. Steinkamp HJ, Wissgott C, Rademaker J, Felix R. Current status 48. Lin L, Hu P, Shi J, Appleton CM, Maslov K, Li L, Zhang R, Wang
of power Doppler and color Doppler sonography in the differential LV. Single-breath-hold photoacoustic computed tomography of the
diagnosis of lymph node lesions. Eur Radiol. 2002;12(7):1785–93. breast. Nat Commun. 2018;9(1):2352.
https://doi.org/10.1007/s003300101111. Epub 2001 Sep 7. PMID: 49. Witte RS, Karunakaran C, Zuniga AN, Schmitz H, Arif H. Frontiers
12111070. of cancer imaging and guided therapy using ultrasound, light, and
40. Moritz JD, Ludwig A, Oestmann JW. Contrast-enhanced color microwaves. Clin Exp Metastasis. 2018;
Doppler sonography for evaluation of enlarged cervical lymph 50. Garcia-Uribe A, Erpelding TN, Krumholz A, Ke H, Maslov K,
nodes in head and neck tumors. Am J Roentgenol. 2000;174:1279. Appleton C, Margenthaler JA, Wang LV. Dual-modality photo-
41. Yang L, Zhao H, He Y, Zhu X, Yue C, Luo Y, Ma B. Contrast- acoustic and ultrasound imaging system for noninvasive senti-
Enhanced Ultrasound in the Differential Diagnosis of Primary nel lymph node detection in patients with breast cancer. Sci Rep.
Thyroid Lymphoma and Nodular Hashimoto’s Thyroiditis in 2015;5:15748.
a Background of Heterogeneous Parenchyma. Front Oncol. 51. Forbrich A, Heinmiller A, Zemp RJ. Photoacoustic imaging of lym-
2021;10:597975. https://doi.org/10.3389/fonc.2020.597975. PMID: phatic pumping. J Biomed Opt. 2017;22(10):106003.
33489895; PMCID: PMC7817885. 52. Dumani DS, Sun IC, Emelianov SY. Ultrasound-guided immu-
42. Zhu J, Lin S, Leow CH, Rowland EM, Riemer K, Harput S, nofunctional photoacoustic imaging for diagnosis of lymph node
Weinberg PD, Tang MX. High frame rate contrast-enhanced metastases. Nanoscale. 2019;11(24):11649–59.
ultrasound imaging for slow lymphatic flow: influence of ultra- 53. Witte RS, Tamimi EA. Emerging photoacoustic and thermoacous-
sound pressure and flow rate on bubble disruption and image tic imaging technologies for detecting primary and metastatic can-
­persistence. Ultrasound Med Biol. 2019;45(9):2456–70. https://doi. cer and guiding therapy. Clin Exp Metastasis. 2021; https://doi.
org/10.1016/j.ultrasmedbio.2019.05.016. Epub 2019 Jul 3 org/10.1007/s10585-­021-­10095-­x.
43. Kawai Y, Ajima K, Nagai T, Kaidoh M, Ohhashi T. Real-time imag- 54. Kajita H, Suzuki Y, Sakuma H, Imanishi N, Tsuji T, Jinzaki M, Aiso
ing of the lymphatic channels and sentinel lymph nodes of the S, Kishi K. Visualization of lymphatic vessels using photoacous-
stomach using contrast-enhanced ultrasonography with Sonazoid tic imaging. Keio J Med. 2021; https://doi.org/10.2302/kjm.2020-­
in a porcine model. Cancer Sci. 2011;102(11):2073–81. https://doi. 0010-­OA. Epub ahead of print
org/10.1111/j.1349-­7006.2011.02048.x. Epub 2011 Aug 24 55. Nagae K, Asao Y, Sudo Y, Murayama N, Tanaka Y, Ohira K, Ishida
44. Wang Y, Zhou W, Li C, Gong H, Li C, Yang N, Zha X, Chen L, Y, Otsuka A, Matsumoto Y, Saito S, Furu M, Murata K, Sekiguchi H,
Xia T, Liu X, Wang M, Ding Q. Variation of sentinel lymphatic Kataoka M, Yoshikawa A, Ishii T, Togashi K, Shiina T, Kabashima
channels (SLCs) and sentinel lymph nodes (SLNs) assessed by K, Toi M, Yagi T. Real-time 3D photoacoustic visualization sys-
contrast-enhanced ultrasound (CEUS) in breast cancer patients. tem with a wide field of view for imaging human limbs. F1000Res.
World J Surg Oncol. 2017;15(1):127. https://doi.org/10.1186/ 2018;7:1813. https://doi.org/10.12688/f1000research.16743.2.
s12957-­017-­1195-­3. 56. Sinnamon AJ, Neuwirth MG, Song Y, Schultz SM, Liu S, Xu X,
45. Cui Q, Dai L, Li J, Xue J. Accuracy of CEUS-guided sentinel et al. Multispectral photoacoustic imaging for the detection of sub-
lymph node biopsy in early-stage breast cancer: a study review and clinical melanoma. J Surg Oncol. 2019;119(8):1070–6.
meta-analysis. World J Surg Oncol. 2020;18(1):112. https://doi. 57. Stoffels I, Morscher S, Helfrich I, et al. Metastatic status of sen-
org/10.1186/s12957-­020-­01890-­z. tinel lymph nodes in melanoma determined noninvasively with
46. Li J, Lu M, Cheng X, Hu Z, Li H, Wang H, Jiang J, Li T, multispectral optoacoustic imaging. Sci Transl Med. 2015;
Zhang Z, Zhao C, Ma Y, Tan B, Liu J, Yu Y. How Pre-operative 7(317):317.
Sentinel Lymph Node Contrast-Enhanced Ultrasound Helps 58. Galanzha EI, et al. In vivo liquid biopsy using Cytophone platform
Intra-operative Sentinel Lymph Node Biopsy in Breast Cancer: for photoacoustic detection of circulating tumor cells in patients
Initial Experience. Ultrasound Med Biol. 2019;45(8):1865–73. with melanoma. Sci Transl Med. 2019;11(496):eaat5857.
https://doi.org/10.1016/j.ultrasmedbio.2019.04.006. Epub 2019 59. Bernas M, Thiadens SRJ, Stewart P, Granzow J. Secondary lymph-
May 14. PMID: 31101445. edema from cancer therapy. Clin Exp Metastasis. 2021; https://doi.
47. Wang Y, Nie F, Wang G, Liu T, Dong T, Sun Y. Value of combin- org/10.1007/s10585-­021-­10096-­w. Published online May 05
ing clinical factors, conventional ultrasound, and contrast-enhanced
 Lymphedema: General
Pathophysiology, Prevention, 25
and Management in Invasive Cancer

Angelika Chachaj, Neil Piller, Francesco Boccardo,

and Andrzej Szuba

Abstract exists for lymphedema. A range of drugs, including

anti-T cell, anti-inflammatory, and antibiotics add to the
Lymphedema is a common complication in cancer survi-
potential of pharmacological treatment/management of
vors. It results from lymphatic system injury due to
lymphedema. Coumarin, studied years ago, shows it has a
lymphadenectomy, radio-, and chemotherapy. It also can
potential worth further exploration.
result from lymphatic obstruction by malignant cells in
advanced cancer. Reported incidence of CRL varies
widely depending on the type of malignancy, lymph-
edema diagnostic criteria, and follow-up period. CRL is a Learning Objectives
debilitating condition and significantly impairs the qual- • To identify patients at the highest risk for develop-
ity of life in cancer survivors in all, physical, emotional, ing cancer-related lymphedema (CRL).
social, and economic aspects of daily life. Prospective • To recognize the main clinical signs and symptoms
surveillance of persons after cancer surgery helps early of CRL.
recognition and treatment. The decongestive lymphatic • To be aware of significant physical and psychoso-
therapy is a gold standard in conservative treatment of cial consequences of CRL.
lymphedema in the acute phase and custom compression • To know the prevention and treatment methods in
garments are used in the maintenance therapy. CRL.
Microsurgical methods are gaining more popularity and • To become more aware of drugs that may help
have the potential to restore lymphatic flow in affected lymphedema treatment.
extremities both to prevent lymphedema and improve • To know how to prevent lymphedema following
existing lymphedema. Excess of subcutaneous fat in nodal dissection by microsurgical lymphatic venous
lymphedematous limbs can be effectively removed using bypass technique (LYMPHA procedure), indica-
liposuction; however, continuous compression with gar- tions, tips, and tricks.
ments is necessary. No specific, FDA-approved drug • To know how to treat chylous leakage following
lymph nodal dissection for the treatment of abdomi-
Sections 7.1–7.5.1 have been prepared by Andrzej Szuba and Angelika
nal organ tumors, diagnostics, and management.
Chachaj, Section 7.52 has been prepared by Neil Piller and Sections
7.53 and 7.54 have been prepared by Francesco Boccardo.

A. Chachaj · A. Szuba (*) 25.1 Introduction

Department of Angiology, Wroclaw Medical University,
Wroclaw, Poland
e-mail: [email protected] Lymphedema is a chronic and progressive condition caused
by lymph stasis due to lymphatic vessels dysfunction [1].
N. Piller
Flinders Medical Centre, Department of Surgery, School of Primary lymphedema results from abnormal development
Medicine, Bedford Park, SA, Australia of lymphatic vasculature due to germinal or somatic genetic
e-mail: [email protected] mutations. Secondary or acquired lymphedema is a result of
F. Boccardo lymphatic injury, which can be iatrogenic, post-traumatic, or
San Martino Hospital, University of Genoa, Genoa, Italy caused by infection. Filariasis is the most prevalent cause of
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 261
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_25
262 A. Chachaj et al.

secondary lymphedema worldwide. Lymph node metastases drained and transported solely by lymphatic vessels.
and/or lymphatic vessels’ obliteration by neoplastic cells in Reabsorption into capillary venules at the level of microcir-
advanced stage of cancer may also cause lymph stasis and culation is transient and minimal [8]. Impairment of lym-
lymphedema (malignant lymphedema). The most common phatic drainage results in lymph stasis and lymphedema [1].
cause of secondary lymphedema in countries not affected by Regional lymph node dissection (axillary, inguinofemo-
filariasis is oncological therapy [2]. With improved progno- ral, pelvic, neck dissection) and/or radiotherapy results in
sis for cancer survivors, the number of patients who suffer disruption of lymphatic vessels and failure in drainage of
from cancer-related lymphedema (CRL) may increase in the appropriate lymphatic territory. Malignant lymphedema
coming years [3]. The incidence of CRL varies with the type develops when neoplastic cells cause an obstruction of
of cancer, the treatment used, measurement methods, and the lymph nodes and/or lymphatic vessels [9].
duration of follow-up [4, 5]. CRL most frequently develops Lymphatic vessels’ disruption leads to a rise in lymph
within 2 years after oncological treatment, but it can also pressure inside collecting lymphatic vessels and progressive
appear years or even decades later [4]. structural changes—from initial dilation to fibrosis and com-
CRL affects areas drained by dysfunctional/dissected plete obliteration [10]. Lymph stasis is accompanied by the
lymph nodes: upper and lower limbs, head, and neck, but accumulation of immune cells, glycosaminoglycans, meta-
also breast, trunk, pubic area, groin, and scrotum. CRL is bolic products, and inflammatory mediators leading to
usually limited to the skin and subcutaneous tissue; however, chronic inflammation. The subsequent steps in lymphedema
head and neck lymphedema may affect deeper tissues and development are proliferation of keratinocytes, fibroblasts
organs, e.g., tongue, larynx, and throat. and adipocytes, destruction of elastic fibers, and collagen
Breast cancer survivors are the major group of patients deposition [9].
affected with CRL. According to large meta-analysis of 72 Lymphedema, initially reversible, becomes irreversible at
studies, the overall incidence of arm lymphedema in this the later stages. Skin thickens and hardens and it is impossi-
group of patients was estimated to be approximately 17%. ble to pinch a fold of skin at the base of the second toe. This
Data from prospective cohort studies suggest higher inci- symptom is known as positive Stemmer’s sign, considered
dence of arm lymphedema in breast cancer survivors—21%. pathognomonic for lymphedema. Lymph stasis impairs local
Arm lymphedema is nearly four times more prevalent in immune response leading to recurrent infections of the skin
women after axillary lymph node dissection (ALND) than in and subcutaneous tissue (dermatolymphangioadenitis).
those who had sentinel lymph node dissection (SLND) [4]. Recurrent infections cause further damage of skin lymphat-
Lymphedema is a common complication of oncological ics and aggravation of lymphedema [9].
treatment in all solid malignancies including melanoma, International Society of Lymphology distinguishes four
gynecological and genitourinary tumors, head and neck can- stages of lymphedema progression: stage 0–subclinical lymph
cers, and sarcoma (Fig. 25.1). The overall prevalence of CRL stasis, stage 1—reversible edema, stage 2—irreversible
excluding breast cancer was estimated as 15% and varied by edema, stage 3—irreversible edema with limb deformation
malignancy (Table 25.1) [5]. and skin changes (elephantiasis) (ISL Consensus 2020) [11].

Definition
*The section on “General pathophysiology of cancer-­
Cancer-related lymphedema (CRL)—edema caused
related lymphedema” was solely prepared by Angelika
by lymphatic system injury and lymph stasis due to
Chachaj and Andrzej Szuba
oncological treatment (lymph nodes dissection, radio,
and chemotherapy) or by tumor spread into lymphat-
ics. Lymphedema is characterized by progressive
Chronic lymphedema is a risk factor of secondary malig-
swelling, fibrosis, and abnormal fat deposition of the
nancy–lymphedema-associated angiosarcoma (earlier called
affected region.
lymphangiosarcoma). Initially described in women with
breast cancer–related lymphedema (Stewart–Treves syn-
drome), this is an extremely aggressive tumor that originated
from the lymphatic endothelium, with extremely poor prog-
25.2  eneral Pathophysiology of Cancer-­
G nosis. Five-year survival is less than 10% [12]. Other malig-
Related Lymphedema* nancies with increased prevalence in lymphedema include
Kaposi’s sarcoma, squamous cell carcinoma, malignant lym-
The cardinal function of the lymphatic system is absorption phoma, and melanoma [13].
and transport of interstitial fluid from extracellular space. Extremity lymphedema leads to remarkable functional
Under physiological conditions, the extracellular fluid is impairment depending on its severity. Asymmetric limb
25 Lymphedema: General Pathophysiology, Prevention, and Management in Invasive Cancer 263

a b

c d

Fig. 25.1 (a–d) Examples of cancer-related lymphedema: Right-arm cal cancer (surgery + radiotherapy) (c), left leg lymphedema secondary
lymphedema secondary to breast cancer treatment (a), left arm “malig- to treatment of vulvar cancer (d)
nant” lymphedema (b), right leg lymphedema after treatment of cervi-
264 A. Chachaj et al.

Table 25.1 Cancer-related lymphedema (CRL) prevalence according therapy (radiation therapy, but also chemotherapy taxanes),
to the literature data
and skin infections in the peri- and postoperative period are
Average (range) the most important risk factors for lymphedema develop-
prevalence of CRL
Type of cancer [%] References
ment in cancer patients [4, 5, 11, 15]. Increased body weight
Breast cancer 17 (8–21) [4] (particularly >25) and being insufficiently active are also
 – with axillary lymph node 20 associated with higher risk, however, the exact pathophysio-
dissection (ALND) logical mechanism mediating this association is not known
 – with sentinel lymph node 6 [4]. Generally, the possibility of developing CRL increases
dissection (SLND)
Melanoma [5]
with the number of risk factors [11].
 – with axillary lymph node 3 (1–39) Despite similar risk factors, only some patients develop
dissection (ALND) CRL. Genetic predisposition may also play a role. Published
 – with inguinofemoral 18 (6–66) studies found associations between genetic factors and the
lymph node dissection
development of breast CRL for variations in the following
Gynecologic malignancies 25 (0–73) [5]
 – cervical 27 (2–49) genes: HGF, MET, GJC2, IL1A, IL4, IL6, IL10, IL13,
 – vulvar 30 (0–73) VEGF-C, NFKB2, LCP-2, NRP-2, SYK, VCAM1, FOXC2,
 – endometrial 1 VEGFR2, VEGFR3, and RORC [16]. Other possible risk fac-
Genitourinary malignancies 11 (1–23) [5] tors are lower efficiency of lymphatic drainage and transport
 – penile cancer 21 (20–21)
 – bladder cancer 4 (15–23) and concomitant chronic venous insufficiency. Women who
 – prostate cancer 16 (1–18) developed breast CRL had higher pumping pressures and
Head/neck cancer [6] lymph transport before cancer treatment, what could indicate
 – external lymphedema 90 initial lymphatic overload [17]. Individual anatomical arrange-
 – internal lymphedema 89
ment of lymph nodes, specific lymphatic and venous network,
 – internal and external 81
lymphedema higher peripheral filtration rates within microcirculation, and
Sarcoma 30 [7] individual weakness or underdevelopment of lymphatics and
veins could constitute additional risk factors [18–20].

weight causes significant strain on the musculoskeletal sys-

tem and peripheral nerves. Lymphedema complications 25.4 Prevention of Cancer-Related
include among others: brachial plexopathy, cervical and Lymphedema
lumbar radiculopathy, carpal tunnel syndrome, degenerative
spine, and joint changes [1]. People with advanced lymph- Primary prevention of CRL is based on modification of risk
edema have difficulties to move around and have limited factors, especially maintaining a proper body weight and
ability to work and to carry out daily activities [14]. performing regular physical activity. Avoiding skin injuries
CRL is also one of the most important conditions leading to (e.g., using protective gloves when working in the kitchen or
psychosocial dysfunction. Lymphedema is a chronic, lifelong, in the garden) and insect bites using repellants is also advised
and incurable condition. It is visible, hence can be stigmatiz- [11, 21]. Prophylactic compression garments, physical exer-
ing. It constantly reminds the patient of cancer disease. cises, or self-massage after cancer treatment are often sug-
Lymphedema makes it difficult to find suitable clothes or gested in CRL prevention, however, these methods have
shoes. It often causes withdrawal from social life, or leads to been not sufficiently evaluated in clinical studies. Avoiding
loss of job. Anxiety and depressive symptoms are common. some lifestyle-associated behaviors and medical procedures
Therefore, lymphedema significantly contributes to an overall on at-risk limb, such as air travel, effect of extreme tempera-
poor quality of life. It affects every aspect of daily life: physi- tures (e.g., hot tubs or sauna), intravenous blood collection
cal, emotional, social, and financial [14]. or injections, and ipsilateral arm blood pressure measure-
Figure 25.1 presents examples of patients with cancer-­ ments is widely advised to cancer survivors; however, their
related lymphedema. significance in CRL prevention has also been insufficiently
investigated [9, 11, 15].
Early diagnosis of CRL gives the opportunity for greater
25.3  isk Factors of Cancer-Related
R effectiveness of conservative treatment and in some patients of
Lymphedema applying surgical methods, e.g., creation of lymphatic-­venous
anastomoses. Therefore, postoperative follow-up evaluation
The advanced stage of cancer, extensive oncological treat- of patients assures early diagnosis (e.g., in 3 months intervals
ment, including more extensive surgical procedures, removal during the first post-cancer treatment and then less frequently)
of significant number of lymph nodes, receipt of adjuvant and is highly recommended [11].
25 Lymphedema: General Pathophysiology, Prevention, and Management in Invasive Cancer 265

25.5 Management of Cancer-Related utes to relief in lymphedema-associated symptoms, improve

Lymphedema body image, quality of life, and insulin control. Thus, weight
loss should be recommended for patients with CRL [11, 21].
Treatment of lymphedema should be complex and multi- Psychological support and patient education in self-­
modal. Optimally, it should involve cooperation of medical compression therapy, MLD, infection prevention, exercises,
team consisting of lymphologist and physiotherapist and in and diet for weight control are also crucial in successful CRL
some patients also of nurse, surgeon, dietician, and psychol- therapy [11].
ogist. Therapy of peripheral lymphedema is divided into
conservative (nonoperative) and surgical methods [11].
25.5.2 T
 he Pharmacological Management
of Lymphedema and its Risk Factors
25.5.1 Conservative Treatment
There are no FDA-approved drugs for lymphedema.
Complete decongestive therapy (CDT) is widely recognized Lymphedema has an inflammatory basis. The factors leading
as primary and effective conservative treatment. It can be to its precipitation/worsening are infection and increasing
used in both children and adults. CDT may be also applied as fibrosis. By exploring/using drugs affecting these processes,
palliative method in cases of tumor-blocked lymphatics to we could expect increased benefits from the traditional core
control volume of malignant lymphedema [11, 22]. treatment modalities, reducing health system/patient costs,
CDT includes skincare (protection from injuries, moistur- and improving functional outcomes.
izing, active treatment of possible ulcerations), use of manual
lymphatic drainage (MLD) to augment lymphatic flow, com-
The Section “The Pharmacological Management of
pression therapy for volume reduction, and decongestive
Lymphoedema and its risk factors” was solely pre-
exercises, both aerobic and resistive. The first phase of CDT
pared by Neil Piller
is applied to improve the condition of the skin and to reduce
limb volume. The second phase of CDT aims to maintain and
optimize the effect achieved during the first phase [9, 11, 22].
The most important component of CDT is compression 25.5.2.1 Anti T-Cell drugs
therapy. The mechanism of its effectiveness in reducing Regulatory T cells ensure that our immune responses do not
edema volume depends on producing gradual pressure—the overshoot the target. Lymphedema can be kept under control
highest at the distal part of the limb and the lowest proxi- by suppressing inflammation, but the danger in this is that
mally. Major medical contraindications to compression ther- tumor growth may be promoted.
apy include heart failure, peripheral arterial disease, acute In a mouse model, increasing the number of T-reg cells
infections, and some skin disorders. Compression therapy reduced lymphedema development and reduced edema,
during the first phase of CDT is provided using multilayered inflammation, and fibrosis, and increased lymphatic drainage
bandage wrapping with the low-stretch bandages worn for in established lymphedema [23].
23 h a day. After achieving the maximum reduction in vol- Ketoprofen (an anti-inflammatory drug) was found to
ume, compression therapy in the second stage of CDT may improve lymphatic function in a small trial [24].
be replaced by compression garments: preferably Ubenimex (an inhibitor of Leukotriene LT (LTB4) synthe-
­custom-­made flat-knitted compression stockings or sleeves. sis) was investigated in a double-blind, placebo-controlled
The compression garments should be worn every day [11]. study with a 6-month treatment period, in adults with pri-
Compression systems with Velcro were recently reported to mary/secondary lymphedema, but there was no improve-
as effective as low-stretch bandages in the treatment of ment over placebo in some primary and secondary endpoints
lymphedema [23]. This type of compression garment is eas- [25].
ier to put on by patients and therapists and can effectively Tacrolimus (an anti-T-cell immunosuppressive drug),
reduce limb volume in ambulatory patients who have no applied externally prevented lymphedema development and
access to CDT. improved established lymphedema. This study showed
Intermittent pneumatic compression, lymphotaping, low-­ reduced swelling, T-cell infiltration and tissue fibrosis, and
level laser therapy and far infrared radiation therapy, play a increased formation of lymphatic collaterals. There was min-
supporting role in the management of lymphedema [11]. imal systemic absorption of the drug [26]. These studies
There is limited evidence to support weight loss in lymph- raise the possibility of the use of this range of drugs in the
edema treatment. However, higher body mass index is a risk treatment of secondary lymphedema but further human test-
factor of lymphedema development and weight loss contrib- ing is essential.
266 A. Chachaj et al.

25.5.2.2 Other Drugs feasible, but some uncertainty remains about the link between
A new and relatively untrialed drug with potential anti-­ the genotype, metabolic phenotype, and the exact gene prod-
inflammatory/anti-fibrotic activities is deupirfenidone. It is a ucts involved, these are being elucidated and once clarified,
deuterated form of the synthetic antifibrotic agent pirfeni- a targeted pharmaco-genomic approach may be appropriate.
done. Animal studies have shown it inhibits a variety of pro-­
inflammatory mediators, such as interleukin-6 (IL-6), tumor 25.5.2.4 Conclusion
necrosis factor-alpha (TNF-α), and transforming growth A range of drugs, including anti-T cell, anti-inflammatory,
factor-beta (TGF-β). This means it has the potential to reduce and antibiotics, adds to the potential of pharmacological
fibrosis and inflammation and may facilitate improvements treatment/management of lymphedemas becoming main-
in impaired lymphatic flow. stream. But refocusing on Coumarin shows it has a potential
as a pharmacological means of targeting edemas where pro-
25.5.2.3 Coumarin and the Benzopyrones tein levels are elevated. Its core action is the stimulation of
However, the focus of this section of this chapter revolves proteolysis by tissue macrophages. By removing protein,
around changing the pathophysiology of the progression of there is less osmotic force holding the fluids, freeing them
lymphedema. This has led to a focus on the Benzopyrones. from the accumulation sites, reducing associated chronic
All are derived from an alpha or gamma pyrone united with inflammation, and fibrosis formation.
a benzene ring. There are two main groups, Coumarin and its A barrier to its common use (hepatotoxicity) can probably
derivatives and Flavones and derivatives. be removed through improved screening.
They have been shown to have a degree of clinical effi- In combination with (or pre and post use) with other treat-
cacy in lymphedema management. Some cases of idiosyn- ments such as compression and massage synergistic
cratic hepatotoxicity, however, were documented leading to improved outcomes might be expected, but need to be further
a reluctance to use them. However, by considering their met- explored.
abolic pathways and identifying relevant alleles needed to
take a targeted pharmacogenetic approach in their future use
this risk can be minimized. 25.5.3 S
 urgical Approaches to Lymphedema
Benzopyrones have reasonably high levels of evidence to Prevention and Treatment
indicate a benefit in reducing primary/secondary lymph- in Oncological Patients*
edema [27, 28]. They reduce high protein edema by stimulat-
ing proteolysis by tissue macrophages [29]. Removing/ 25.5.3.1 Background
reducing protein means a reduced osmotic force so they are A side effect of axillary lymph node excision and radiation
freed to leave the accumulation sites. Since these often con- therapy for breast cancer is arm lymphedema, which occurs
tain a range of inflammatory mediators, its removal reduces in approximately 25% of patients (13% to 52%). Sentinel
chronic inflammation, fibrosis formation, and the risk of sec- lymph node (SLN) biopsy reduced the severity of swelling in
ondary infection [27]. nearly 6% of patients (2% to 7%) but, in the case of positive
However, reports of hepatotoxic effects led to their use for SLN, complete axillary dissection (AD) is still required. This
lymphedema management being curtailed. The cytotoxic is why the Axillary Reverse Mapping (ARM) method was
outcomes relate to its metabolism. Coumarin has two main developed, with the aim of identifying and preserving the
hepato-­
­ detoxification pathways. First, the nontoxic lymphatic vessels that drain the arm [32–34]. It consists of
7-­hydroxylation catalyzed by CYP2A6 and the toxic hetero- injecting a small amount (1–2 mL) of blue dye intradermally
cyclic “ring-­splitting” epoxidation pathways catalyzed likely and subcutaneously on the medial surface of the arm, which
by CYP1A and CYP2E. Acetaldehyde dehydrogenase helps to localize the lymphatic tracts of the draining arm.
(ALDH) clears toxic aldehyde intermediates. Leaving the lymph nodes related to lymphatic drainage in
Dysfunctional polymorphisms of the above genes are sig- the arm in place would reduce the risk of arm lymphedema,
nificantly more prevalent in Eastern Asian populations and but the main risk is to leave the metastatic disease in the
relatively uncommon in Caucasian populations. axilla. On the contrary, the lymphatic pathways of the arm
The problem of coumarin toxicity can relatively easily be that enter the armpit are not considered as a usual site of
solved by considering and knowing its metabolic pathways breast cancer and their preservation would certainly lead to a
and identifying relevant alleles. Knowing these precautions, significant decrease in the onset rate of lymphedema.
a targeted pharmaco-genetic approach can be taken in its use
[30, 31]. CYP2A6 polymorphism screening methods, includ-
*The section on “Surgical Approaches to Lymphedema
ing genotyping, by real-time polymerase chain reaction, and
Prevention and Treatment in Oncological Patients”
chromatography–mass spectroscopy functional metabolite
was solely prepared by Francesco Boccardo
assays are available and increasingly reliable and clinically
25 Lymphedema: General Pathophysiology, Prevention, and Management in Invasive Cancer 267

25.5.3.2 Clinical Applications 25.5.3.3 Discussion

Based on long-term experience in lymphatic-venous anasto- Recent advances in breast cancer treatment, particularly
moses (LVA) for the treatment of lymphedema [35, 36], we regarding the prevention of lymphatic complications follow-
conceived and carried out a preventive LVA during nodal dis- ing sentinel lymph node biopsy and axillary dissection, have
section, which consists of anastomosing the lymphatics of led to the proposal of a new technique to primarily prevent
the arm to a collateral branch of the axillary vein (Lymphatic lymphedema by microsurgical lymphatic-venous anastomo-
Microsurgical Preventing Healing Approach—LYMPHA sis (LYMPHA). LVA has been shown not only to prevent
technique) [37]. Indications for the LYMPHA technique are lymphedema but also to reduce early lymphatic complica-
based on clinical and lymphoscintigraphic parameters. tions (e.g., lymphorrhea, lymphocele) thanks to the reduced
Obesity is known to predispose to lymphedema and there- regional intralymphatic pressure. The proposed surgical
fore patients with BMI greater than 30 are considered to be technique for patients with operable breast cancer requiring
at high risk of lymphedema and are candidates for axillary dissection consisted of performing LVA [38, 39]
LYMPHA. Patients with normal BMI are studied with lym- between the identified arm lymphatic vessels by simultane-
phoscintigraphy which is to detect latent lymphatic impair- ously injecting the blue dye into the arm and a branch of the
ment, not yet clinically evident. Radiotherapy causes axillary vein (Fig. 25.2). The incidence of secondary arm
temporary arm edema but do not jeopardize the LVA. lymphedema after the LYMPHA technique is about 4.05%,

a b

c d

Fig. 25.2 (a–e) Injection of BPV dye to the arm (a), isolation of the of the axillary vein (c), lymphatic-venous anastomosis at the end (d),
brachial lymphatics during lymph node dissection and identification of LYMPHA technique at higher magnification (e)
the same using metal clips (b), isolation and preparation of a collateral
268 A. Chachaj et al.

compared to the incidence rate after complete axillary dis- modernly performed through mini-invasive procedures. There
section alone reported in the literature, which varies from is still not a consensus on which are the specific indications for
13% to 65%, depending on the criteria used and the evalua- each technique, how to evaluate results, pros, and cons. The
tion methods. aim is to give an outline of the state of the art of surgical treat-
ment of lymphedema. Various review articles and published
25.5.3.4 LyMPHA Technique Beyond Breast clinical experiences from different groups of surgeons devoted
Cancer to lymphatic surgery report about the usefulness of lymphatic
The LYMPHA technique is also useful in patients with trunk reconstructive microsurgery in early stages, where there are
melanoma and vulvar cancer, in whom preventive lymphatic-­ less fibrotic tissue and lymphatic changes (8). It is much more
venous anastomosis (LVA) can be performed simultaneously difficult to find good lymphatics in more advanced stages,
with inguinal lymphadenectomy [40] (Fig. 25.3). Patients at because of the severe sclerotic changes of the tissues and the
risk can be assessed peroperitively with lymphoscintigraphy aponeurotic fasciae, even though searching for both superfi-
and be referred to preventive measures. cial and deep lymphatic collectors, actually a few numbers of
lymphatic vessels may be found. It is mandatory to perform a
25.5.3.5 Lymphedema Surgical Options proper preoperative diagnostic assessment through superficial
Different surgical techniques are nowadays proposed for the and deep lymphoscintigraphy and echo-duplex venous scan.
treatment of lymphedema. Most of them are reconstructive Lymphangio-MR is used in the most advanced stages to study
microsurgical techniques and some of them are debulking, the morphology of lymphatic drainage. ICG lymphangiogra-

a b

c d e f

Fig. 25.3 (a–f) Iliac-obturator and femoro-inguinal lymph node dis- the limb lymphatic tracts by fluorescence (c, d), lymphatic-venous
section for trunk melanoma (a), injection of BPV dye and indocyanine anastomosis at the end, with control of patency by fluorescence (e, f)
green (ICG) distal to the surgical area (b), identification and isolation of
25 Lymphedema: General Pathophysiology, Prevention, and Management in Invasive Cancer 269

phy is used peroperitively together with the blue dye (BPV) to 25.5.4.3 Treatment
find sound lymphatic collectors during surgery. Results are Chylous pathologies are initially treated by an adequate
evaluated by clinical assessment and lymphoscintigraphy. compensation of the metabolic conditions, with an appropri-
Minimally p­ erformed debulking techniques are used in the ate dietary regime, based on protein reintegration and limit-
most advanced cases to reduce the fibrotic tissue after CDP ing the lipid intake exclusively to fats based on medium-chain
plus microsurgery. triglycerides (MCT). The surgical timing will start articulat-
The best results are obtained by microsurgery in the ear- ing itself on the basis of the result obtained with the conser-
lier stages and in secondary lymphedemas in terms of reduc- vative therapy. The surgical treatment will be modulated
tion of lymphangitis and less necessity of postoperative according to the case and the different possible surgical pro-
physical treatment and elastic garments. Microsurgery may cedures will involve:
be of great help also in late stages, which may however need
proper postoperative treatment to improve the outcome. –– drainage of the chyloperitoneum;
The role of surgery in lymphedema treatment is of utmost –– identification and closure of the sites of the chylorrhagia;
importance mainly for secondary lymphedemas and obstruc- –– removal of chylomas;
tive primary ones. The meaning of microsurgical techniques –– ligatures of dysplastic chyliferous vessels;
is to open the tap and allow the lymph to pass through a –– derivative microsurgical techniques (chyle-venous
lymphatic-venous bypass as it happens in arterial surgery anastomosis);
and in both cases, the vein is the queen. –– application of fibrin glue or platelet gel.

25.5.3.6 Final Considerations

Lymphedema is a consequence of cancer treatment and **The section on “Treatment of chylous effusions
patients should be informed of the prevention options of from metastatic disease” was solely prepared by
lymphedema. The use of blue dye, Indocyanine Green (ICG), Francesco Boccardo
and LVA helps to solve the problem of the prevention of sec-
ondary lymphedema of the arms and legs while maintaining
the oncological radicality. LYMPHA, therefore, could repre-
sent a rational approach to the prevention of lymphedema Finally, alternative therapy to the above-described tech-
and reduce other lymphatic complications after axillary and niques is represented by the peritoneo-jugular shunt (Denver
inguinal surgery in the treatment of breast cancer, melanoma, shunt), in nonresponsive cases.
vulvar cancer, and other cancers.

25.6 Conclusion and Perspective

25.5.4 T
 reatment of Chylous Effusions
from Metastatic Disease* • Cancer-related lymphedema (CRL) is a common conse-
quence of malignancy and cancer treatment.
25.5.4.1 Background • The frequency of this complication will probably increase
Chylous disorders can manifest themselves with chyloperi- as a result of improvement in effectiveness of cancer
toneum, chyledema of the lower limbs and/or of the external treatment.
genitalia, chyluria, chylothorax, chylometrorrea, chylarthro- • CRL is a debilitating condition which significantly
sis, variously associated with one another and involve a impairs the quality of life in cancer-survivors and demands
demanding diagnostic and therapeutic process [41]. lifelong therapy.
• Screening of cancer survivors toward lymphedema is
25.5.4.2 Diagnostics essential in early diagnosis and successful treatment of
The main diagnostic investigations include an accurate CRL.
hematochemical study (including protein electrophoresis, • Understanding the pathophysiology, recognition of risk
lipidogram, calcemia), echotomography, lymphoscintigra- factors and predisposition, and knowledge about preva-
phy, lymphangio-CT, lymphangio-MR, and laparoscopy, lent methods of CRL treatment will provide the basis for
which is able to provide precise data on the clinical situation, the development of new preventive, diagnostic, and thera-
severity, and extent of the disease [42]. peutic methods in the management of CRL.
270 A. Chachaj et al.

14. Chachaj A, Malyszczak K, Pyszel K, Lukas J, Tarkowski R,

Open Questions Pudelko M, et al. Physical and psychological impairments of
women with upper limb lymphedema following breast cancer treat-
• Can we improve cancer treatment to avoid/mini- ment. Psychooncology. 2010;19(3):299–305.
mize the risk of treatment-related lymphedema? 15. Asdourian MS, Skolny MN, Brunelle C, Seward CE, Salama L,
• What prophylactic measures should be mandatory Taghian AG. Precautions for breast cancer-related lymphoedema:
in cancer patients at risk of lymphedema? risk from air travel, ipsilateral arm blood pressure measurements,
skin puncture, extreme temperatures, and cellulitis. Lancet Oncol.
• Should early microsurgical restoration of lymphatic 2016;17(9):e392–405.
flow be recommended for all cancer patients with 16. Visser J, van Geel M, Cornelissen AJM, van der Hulst R, Qiu
early lymphedema? SS. Breast cancer-related lymphedema and genetic predisposi-
• Is therapeutic lymphangiogenesis a valid and safe tion: a systematic review of the literature. Lymphat Res Biol.
2019;17(3):288–93.
option for lymphedema treatment? 17. Cintolesi V, Stanton AW, Bains SK, Cousins E, Peters AM,
• What exactly are protective mechanisms in cancer Purushotham AD, et al. Constitutively enhanced lymphatic pump-
survivors, who do not develop limb lymphedema ing in the upper limbs of women who later develop breast cancer-­
after treatment? related lymphedema. Lymphat Res Biol. 2016;14(2):50–61.
18. Szuba A, Chachaj Z, Koba-Wszedybylb M, Hawro R, Jasinski
R, Tarkowski R, et al. Axillary lymph nodes and arm lymphatic
drainage pathways are spared during routine complete axillary
Patient Information and Guidelines clearance in majority of women undergoing breast cancer surgery.
Lymphology. 2011;44(3):103–12.
Support groups and resources can be found at the National 19. Pain SJ, Barber RW, Ballinger JR, Solanki CK, Mortimer PS,
Lymphedema Network. Purushotham AD, et al. Tissue-to-blood transport of radiolabelled
immunoglobulin injected into the web spaces of the hands of nor-
mal subjects and patients with breast cancer-related lymphoedema.
J Vasc Res. 2004;41(2):183–92.
References 20. Britton TB, Solanki CK, Pinder SE, Mortimer PS, Peters AM,
Purushotham AD. Lymphatic drainage pathways of the breast and
1. Szuba A, Rockson SG. Lymphedema: anatomy, physiology and the upper limb. Nucl Med Commun. 2009;30(6):427–30.
pathogenesis. Vasc Med. 1997;2(4):321–6. 21. Shaw C, Mortimer P, Judd PA. Randomized controlled trial com-
2. Szuba A, Rockson SG. Lymphedema: classification, diagnosis and paring a low-fat diet with a weight-reduction diet in breast cancer-­
therapy. Vasc Med. 1998;3(2):145–56. related lymphedema. Cancer. 2007;109(10):1949–56.
3. Brayton KM, Hirsch AT, O Brien PJ, Cheville A, Karaca-Mandic 22. Lasinski BB, McKillip Thrift K, Squire D, Austin MK, Smith KM,
P, Rockson SG. Lymphedema prevalence and treatment benefits in Wanchai A, et al. A systematic review of the evidence for complete
cancer: impact of a therapeutic intervention on health outcomes and decongestive therapy in the treatment of lymphedema from 2004 to
costs. PLoS One 2014;9(12):e114597. 2011. PM R. 2012;4(8):580–601.
4. DiSipio T, Rye S, Newman B, Hayes S. Incidence of unilateral arm 23. Gousopoulos E, Proulx ST, Bachmann SB, Scholl J, Dionyssiou
lymphoedema after breast cancer: a systematic review and meta-­ D, Demiri E, et al. Regulatory T cell transfer ameliorates lymph-
analysis. Lancet Oncol. 2013;14(6):500–15. edema and promotes lymphatic vessel function. JCI Insight.
5. Cormier JN, Askew RL, Mungovan KS, Xing Y, Ross MI, Armer 2016;1(16):e89081.
JM. Lymphedema beyond breast cancer: a systematic review and 24. Rockson SG, Tian W, Jiang X, Kuznetsova T, Haddad F, Zampell J,
meta-analysis of cancer-related secondary lymphedema. Cancer. et al. Pilot studies demonstrate the potential benefits of antiinflam-
2010;116(22):5138–49. matory therapy in human lymphedema. JCI. Insight. 2018;3(20)
6. Ridner SH, Dietrich MS, Niermann K, Cmelak A, Mannion K, 25. ULTRA Study EP. www.eigerbio.com/eiger-­biopharmaceuticals-­
Murphy B. A prospective study of the lymphedema and fibrosis announces-­p hase-­2 -­u ltra-­r esults-­o f-­u benimex-­i n-­l ower-­l eg-­
continuum in patients with head and neck cancer. Lymphat Res lymphedema-­study-­did-­not-­meet-­primary-­or-­secondary-­endpoint
Biol. 2016;14(4):198–205. http://www.eigerbio.com/eiger-­biopharmaceuticals-­announces-­
7. Robinson MH, Spruce L, Eeles R, Fryatt I, Harmer CL, Thomas phase-­2 -­u ltra-­r esults-­o f-­u benimex-­i n-­l ower-­l eg-­l ymphedema-­
JM, et al. Limb function following conservation treatment of adult study-­did-­not-­meet-­primary-­or-­secondary-­endpoint. 2017.
soft tissue sarcoma. Eur J Cancer. 1991;27(12):1567–74. 26. Gardenier JC, Kataru RP, Hespe GE, Savetsky IL, Torrisi JS, Nores
8. Levick JR, Michel CC. Microvascular fluid exchange and the GD, et al. Topical tacrolimus for the treatment of secondary lymph-
revised Starling principle. Cardiovasc Res. 2010;87(2):198–210. edema. Nat Commun. 2017;8:14345.
9. Rockson SG, Keeley V, Kilbreath S, Szuba A, Towers A. Cancer-­ 27. Casley-Smith JR, Morgan RG, Piller NB. Treatment of lymph-
associated secondary lymphoedema. Nat Rev Dis Primers. edema of the arms and legs with 5,6-benzo-[alpha]-pyrone. N Engl
2019;5(1):22. J Med. 1993;329(16):1158–63.
10. Mihara M, Hara H, Hayashi Y, Narushima M, Yamamoto T, 28. Badger C, Preston N, Seers K, Mortimer P. Benzo-pyrones for
Todokoro T, et al. Pathological steps of cancer-related lymph- reducing and controlling lymphoedema of the limbs. Cochrane
edema: histological changes in the collecting lymphatic vessels Database Syst Rev. 2004;(2):CD003140.
after lymphadenectomy. PLoS One. 2012;7(7):e41126. 29. Piller NB. The induction of controlled proteolysis in high protein
11. The diagnosis and treatment of peripheral lymphedema. 2020 oedemas by coumarin. Z Lymphol. 1979;3(2):110–3.
Consensus Document of the International Society of Lymphology. 30. Farinola N, Piller NB. CYP2A6 polymorphisms: is there a role for
Lymphology. 2020;53(1):3–19. pharmacogenomics in preventing coumarin-induced hepatotoxicity
12. Sharma A, Schwartz RA. Stewart-Treves syndrome: pathogenesis in lymphedema patients? Pharmacogenomics. 2007;8(2):151–8.
and management. J Am Acad Dermatol. 2012;67(6):1342–8. 31. Hu M, Piller NB. Strategies for avoiding benzopyrone hepatotox-
13. Peyron N, Dandurand M, Guillot B. Malignant tumors as complica- icity in lymphedema management: the role of pharmacogenet-
tions of lymphedema. J Mal Vasc. 1993;18(4):293–8.
25 Lymphedema: General Pathophysiology, Prevention, and Management in Invasive Cancer 271

ics, metabolic enzyme gene identification, and patient selection. 37. Boccardo FM, Casabona F, Friedman D, Puglisi M, De Cian F,
Lymphat Res Biol. 2017;15(4):317–23. Ansaldi F, et al. Surgical prevention of arm lymphedema after
32. Thompson M, Korourian S, Henry-Tillman R, Adkins L, Mumford breast cancer treatment. Ann Surg Oncol. 2011;18(9):2500–5.
S, Westbrook KC, et al. Axillary reverse mapping (ARM): a new 38. Boccardo F, Fulcheri E, Villa G, Molinari L, Campisi C, Dessalvi S,
concept to identify and enhance lymphatic preservation. Ann Surg et al. Lymphatic microsurgery to treat lymphedema: techniques and
Oncol. 2007;14(6):1890–5. indications for better results. Ann Plast Surg. 2013;71(2):191–5.
33. Nos C, Lesieur B, Clough KB, Lecuru F. Blue dye injection in the 39. Witte MH, Dellinger MT, Papendieck CM, Boccardo
arm in order to conserve the lymphatic drainage of the arm in breast F. Overlapping biomarkers, pathways, processes and syndromes in
cancer patients requiring an axillary dissection. Ann Surg Oncol. lymphatic development, growth and neoplasia. Clin Exp Metastasis.
2007;14(9):2490–6. 2012;29(7):707–27.
34. Ponzone R, Mininanni P, Cassina E, Sismondi P. Axillary reverse 40. Boccardo F, Valenzano M, Costantini S, Casabona F, Morotti M,
mapping in breast cancer: can we spare what we find? Ann Surg Sala P, et al. LYMPHA Technique to prevent secondary lower limb
Oncol. 2008;15(1):390–1. author reply 92-3 lymphedema. Ann Surg Oncol. 2016;23(11):3558–63.
35. Boccardo F, Casabona F, De Cian F, Friedman D, Villa G, Bogliolo 41. Boccardo F, Bellini C, Eretta C, Pertile D, Da Rin E, Benatti E, et al.
S, et al. Lymphedema microsurgical preventive healing approach: The lymphatics in the pathophysiology of thoracic and abdominal
a new technique for primary prevention of arm lymphedema after surgical pathology: immunological consequences and the unex-
mastectomy. Ann Surg Oncol. 2009;16(3):703–8. pected role of microsurgery. Microsurgery. 2007;27(4):339–45.
36. Francesco B, Corrado C, Giuseppe M, Emanuela B, Chiara B, 42. Dessalvi S, Boccardo F, Molinari L, Spinaci S, Campisi C, Ferrari
Francesco P, et al. Prevention of lymphatic injuries in surgery. GM. Chyloperitoneum: diagnostic and therapeutic options.
Microsurgery. 2010;30(4):261–5. Lymphology. 2016;49(1):1–7.
 Part VII
Radiological Imaging of Early Cancer
and Cancer Metastases

Jamie T. Caracciolo

Diagnostic imaging is a critical component in the ongoing fight against cancer and is widely
utilized throughout all phases of patient management and cancer care from early tumor detec-
tion through long-term patient surveillance. A multitude of diagnostic imaging examinations
and image-guided interventions are available to primary care providers and specialized cancer
care physicians alike in the detection and diagnosis of various types and stages of malignant
diseases, surgical and medical treatment planning and execution, assessment of treatment
response, and post-treatment follow-up of patients living with cancer. Diagnostic and interven-
tional radiologists working closely with primary care physicians, surgical specialists, medical
and radiation oncologists, and pathologists in a multidisciplinary approach to cancer diagnosis
and treatment provide for optimal patient management and clinical care. Ideally performed in
close consultation with referring physicians, diagnostic imaging provides clinicians with criti-
cal data to inform many aspects of the clinical decision-making process.
Diagnostic imaging examinations including radiography, mammography, ultrasound, com-
puted tomography (CT), magnetic resonance imaging (MRI), and nuclear medicine examina-
tions, such as positron-emission tomography (PET), allow for detection and characterization
of primary malignant tumors, initial tumor staging, identification of distant metastatic disease,
and assessment of treatment response. Screening examinations of small high-risk patient
groups or large patient populations are performed in an effort to detect earlier stage disease
allowing for earlier initiation of treatment which often leads to improved patient outcomes
such as disease-free survival and overall survival. Following detection of a suspicious abnor-
mality, image-guided biopsies are available for tissue sampling to determine tumor histology
and, in many cases, confirm the presence or absence of distant metastatic disease. Diagnostic
imaging also informs many important decisions throughout the course of therapy. Cross-­
sectional imaging is widely available to assess the local extent of primary diseases helping to
determine surgical resectability or the need for neoadjuvant therapy. CT, MRI, and PET/CT are
often used to assess tumor response to chemotherapy in patients with non-operable or meta-
static disease. Numerous targeted therapies delivered under fluoroscopic guidance, such as
hepatic chemo- or radio-embolization, are currently available in the treatment of primary and
metastatic disease. Cross-sectional imaging is also typically performed throughout the course
of long-term surveillance of previously treated patients to detect recurrent disease. These are
several of the many important and diverse roles diagnostic and interventional radiologists play
in the management of patients with malignant diseases.
274 Radiological Imaging of Early Cancer and Cancer Metastases

The following chapters focus on several important topics in diagnostic and interventional radiology
relevant to the detection, diagnosis, and management of the most common primary malignancies in the
United States and associated patterns of metastatic disease. First, we review several currently available
and commonly used imaging modalities and examinations to detect, characterize, and stage malignant
tumors, and present emerging techniques such as radiomics and artificial intelligence which promise to
become part of the routine evaluation of malignant diseases in the future. We discuss typical imaging
findings of the most common primary malignancies in the US emphasizing primary tumor characteriza-
tion, important findings defining the local extent of disease which direct surgical or medical manage-
ment, and common sites and patterns of distant metastatic disease for each primary tumor type. We
expand our discussion of distant metastatic disease with dedicated chapters highlighting pertinent imag-
ing findings in pulmonary metastatic disease, visceral and lymph node metastatic disease, and osseous
metastatic disease. We also discuss relevant site-specific risks associated with metastatic disease, some
of the available treatment options, and imaging findings reflective of response or resistance to therapy.
Finally, we discuss the role of interventional radiology in the initial diagnosis and treatment of metastatic
disease including image-guided biopsy and targeted interventions for metastatic disease.
 Primary Tumor Staging and Detection
of Common Sites of Distant Metastatic 26
Disease

Kerry L. Thomas, A. Ahmed, and B. Morse

Abstract detection and characterization of primary tumors, subse-

Diagnostic imaging is a cornerstone of cancer diagnosis, quent tumor staging, and surveillance of cancer. Depending
staging, assessment of treatment response, and long-term on the primary site of disease, various imaging studies and
surveillance. The most appropriate choice of imaging modalities are available to characterize the primary tumor
modality for primary tumor staging and detection of the and to assess for locoregional and distant metastatic disease.
metastatic disease depends on the tumor type and com- This chapter discusses several of the most common primary
mon sites of distant disease. This chapter reviews several malignancies affecting patients in the United States and how
of the most encountered cancers in the United States, dis- diagnostic imaging can be used in primary tumor assessment
cusses typical patterns of metastatic disease, and suggests and detection of metastatic disease.
appropriate imaging examinations to evaluate primary
and metastatic disease.
26.2 Breast Cancer

Breast cancer is expected to be the leading site of all new

Learning Objectives
cancer cases and the second most common cause of cancer
After reviewing this chapter, the reader with better
death in women in the United States in 2020. There will be
understand:
an estimated 276,480 new cases and 42,170 new deaths
• Most common malignancies in men and women. related to breast cancer this year. These estimates account for
• Staging and prognosis of the malignancies reviewed. 30% of all new cancer cases in women and 15% of cancer
• Common sites of metastatic disease. deaths in women for 2020 [2, 3]. The 5-year relative survival
rate for breast cancer is 90%, and the lifetime risk of devel-
oping breast cancer is about 13% [2].
Staging of breast cancer follows the tumor, node, metas-
tasis (TNM) staging system set forth by the American Joint
26.1 Introduction Committee on Cancer (AJCC) [4]. Historically, the TNM
staging system stages cancers by anatomic extent of disease,
Cancer is a leading cause of death in the United States, sec- including tumor size, presence or absence of locoregional
ond only to heart disease, according to the Centers for nodal metastases, and the presence or absence of distant
Disease Control, accounting for nearly 600,000 deaths in metastases. However, for breast cancer staging, the eighth
2017 [1]. Diagnostic imaging plays a crucial role in initial edition of the AJCC Cancer Staging Manual also includes a
prognostic stage, which incorporates biologic factors, includ-
ing tumor grade and hormone receptor status [4].
K. L. Thomas (*)
Diagnostic imaging used in initial staging of invasive
Clinical Associate Professor, Department of Radiology,
University of North Carolina at Chapel Hill, School of Medicine, breast cancer includes diagnostic mammogram or breast
Tampa, FL, USA tomosynthesis and ultrasound as necessary for primary
e-mail: [email protected] tumor assessment [5, 6]. Breast magnetic resonance imaging
A. Ahmed · B. Morse (MRI) is considered optional and may be helpful for defining
H. Lee Moffitt Cancer Center, Tampa, FL, USA tumor extent in patients with multifocal, multicentric, and
e-mail: [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 275
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_26
276 K. L. Thomas et al.

occult or poorly defined tumors by other imaging modalities clinically and pathologically with Gleason score and group
[5, 6]. score, as well as serum prostate-specific antigen (PSA) [11].
The presence or absence of lymph node metastasis is an Transrectal ultrasound (TRUS) is used for guiding pros-
important predictor of overall survival and recurrence. tate biopsy in clinically suspected prostate cancer.
Regional lymph nodes for the breast include axillary, ipsilat- Multiparametric MRI of the prostate gland is useful in local
eral internal mammary, and ipsilateral supraclavicular lymph T staging of prostate cancer and MRI may also be used for
nodes [4]. Axillary lymph nodes levels are determined by targeted biopsy [11, 12]. The T stage of the tumor reflects
their relationship to the pectoralis minor muscle and include local extent of disease within the prostate gland and any
three levels: level I: lateral to pectoralis minor; level II: extra-prostatic extension. While CT may demonstrate gross
between the muscular borders of the pectoralis minor; and extracapsular extension, in general, CT is not considered suf-
level III: medial to the pectoralis minor. Imaging assessment ficient to assess the prostate gland, but it is helpful in evaluat-
of regional axillary lymph nodes can be performed with ing for nodal metastasis [13]. Nodal (N) staging for prostate
diagnostic ultrasound and, if indicated, ultrasound-guided cancer is less extensive than other diseases, whereby nodal
biopsy. However, controversy remains as to whether or not staging consists of only three categories: not assessed, no
the routine use of axillary ultrasound should be employed positive nodes, or presence of positive nodes. Nodes refer to
preoperatively [4, 5]. MRI, when performed for evaluating regional lymph nodes in the true pelvis. However, it should
primary tumor extent, is also helpful in evaluating the axilla be mentioned, that the NCCN does not recommend nodal
for lymph node involvement as it is included in the field of staging assessment in very low and low-risk patients. Nodal
view. MRI also provides information about internal mam- assessment for disease is reserved for patients in intermedi-
mary and supraclavicular lymph nodes (level III), which ate or higher risk categories where there is a greater probabil-
should be assessed if there is evidence of level I or level II ity (>10% risk) of nodal disease [13].
adenopathy. If MRI is not performed, internal mammary and Distant sites of metastatic disease in prostate cancer most
supraclavicular nodes may be assessed with contrast-­ commonly occur in bones and lymph nodes, followed by
enhanced chest computed tomography (CT). liver, lung and pleura, and adrenal glands. Other sites are
National Comprehensive Cancer Network (NCCN) rec- reported but are less common [14, 15]. Because bones are the
ommendations suggest that imaging evaluation for distant most common site of distant metastatic disease, skeletal
metastatic disease only be performed if there are other clini- imaging is the mainstay of distant tumor detection. However,
cal signs, symptoms, or laboratory values that indicate not all patients with prostate cancer should undergo bone
metastasis [6]. imaging at initial presentation. Guidelines from the NCCN,
The most common sites of distant metastatic disease for American Urologic Association (AUA), and American
breast cancer include bone, liver, lung, and brain, and College of Radiology Appropriateness Criteria recommend
patients may have multiple sites of metastatic disease [7, 8]. bone imaging in cases of clinical stage T2 and PSA >10 ng/
Imaging studies recommended by the NCCN for metastatic mL, baseline PSA >20 ng/mL, clinical stage T3 or higher, or
evaluation, when indicated, includebone scan or sodium Gleason score of 8 or greater [12, 13]. Osseous metastases
fluoride PET/CT to assess for bone metastases, chest CT for are usually osteoblastic, and Tc-99 whole-body bone scan
pulmonary metastases, and either CT or MRI of the abdo- has been standard to assess for osseous metastases. However,
men and pelvis for visceral metastases. FDG PET/CT is there are several novel radiotracers with promising results
optional [6]. for improved accuracy compared with conventional bone
scan. These tracers, which may be performed with both PET/
CT and PET/MRI include 18-F sodium fluoride, 11-C
26.3 Prostate Cancer Choline, and Fluciclovine [16]. Other imaging studies under
investigation, including 68-Ga PMSA (prostatic membrane
Prostate cancer is expected to be the leading site of all new specific antigen), may prove useful in initial staging and
cancer cases and the second most common cause of cancer detection of recurrent disease [16].
death in men in the United States in 2020. It is estimated that
there will be 191,930 new cases and 33,330 deaths from
prostate cancer this year. This accounts for 21% of all new 26.4 Bronchogenic/Lung Cancer
cancer cases and 10% of cancer related death in men [3, 9].
The 5-year relative survival rate for prostate cancer is 97.8% Lung cancer remains the leading cause of cancer deaths in
and the lifetime risk of developing prostate cancer is approx- the United States. Nearly 25% of all oncologic deaths occur
imately 12% [9, 10]. as a result of primary lung carcinoma, and it accounts for
Like other tumors, prostate cancer is staged utilizing the more mortality than colon, breast, and prostate cancers com-
AJCC TNM staging system, but T stage is assessed both bined [17]. Lung carcinomas are generally divided into two
26 Primary Tumor Staging and Detection of Common Sites of Distant Metastatic Disease 277

major categories: small cell and non-small cell carcinomas the liver and adrenal glands. MRI is not typically used in
(NSCLC). Most lung malignancies, nearly 80–85%, are of staging of lung cancers. However, it may be useful in estab-
the non-small cell type. Adenocarcinomas are the most com- lishing the relationship of the tumor and chest wall, confirm
mon non-small cell cancers and the prevalent histologic type invasion of mediastinal structures, neural axis involvement,
found in smokers. They are found predominantly in the or determine the involvement of the brachial plexus in cases
periphery of the lung, owing to their origin in mucus-­ of Pancoast tumors [29, 30].
secreting cells lining the alveolar sacs.
Lymphogenic spread is the most common route of spread
for lung malignancies. This is reflected in the TNM staging 26.5 Colorectal Cancer
criteria which are the prevalent means of classifying the ana-
tomic extent of tumor, where the N designates regional nodal Colorectal cancer (CRC) is one of the most common causes
involvement. N1 implies ipsilateral peribronchial or hilar of cancer-related death in the United States (third leading
lymph nodal spread, i.e., a right lung cancer with spread to a for both men and women) [31]. Worldwide CRC accounts
right hilar node. N2 implies spread to ipsilateral mediastinal for approximately 10% of newly diagnosed cancers and
and/or subcarinal lymph nodes. N3 indicates tumor spread to cancer-­related deaths [32]. The rate of newly diagnosed
contralateral mediastinal, contralateral hilar, or supraclavic- CRC cases has dropped in the United States since the 1980s,
ular nodal stations or any combination thereof [18]. a trend thought to be secondary to improved screening and
Most lung cancers spread to ipsilateral hilar nodes before population changes of modifiable risk factors [31].
eventually reaching mediastinal nodes. Lung cancers in the Unfortunately, a new trend is increasing incidence in those
lower lobes and middle lobe usually drain via the posterior younger than age 55 (increased 2% per year from 2007 to
mediastinum to subcarinal and subaortic nodal stations. 2016) [31]. CRC is more common in developed countries
Tumors in right upper lobe mostly drain to the superior and worldwide cases are expected to increase with improved
mediastinum, particularly right paratracheal nodes. Tumors living standards in less developed countries over the next
in the left upper lobe drain typically to the anterior mediasti- several decades [32].
num and involve the subaortic station, as well as to the upper CRC is staged using the TNM system. A key component
mediastinum in nearly 1/3 of all cases. Retrograde lymphatic of primary tumor (T) staging is determination of local extent
drainage and subsequently tumor spread to the pleura can of disease and depth of mural invasion of the colon or rec-
occur, especially from peripheral tumors. Hematogenous tum. This distinction is less important with colon cancer than
spread to other more distant sites can also occur, with spread rectal cancer. In general, colon cancer will be resected unless
to other sites in the lungs, bone, brain, and liver being the there is bulky nodal disease or invasion of adjacent organs
most commonly observed sites of metastatic spread in [33]. Rectal cancer has a more nuanced approach. With rec-
NSCLC [19, 20]. tal cancer, contrast-enhanced pelvic MRI is commonly used
NCCN guidelines recommend initial staging with to determine whether the tumor is confined to the rectal wall
contrast-­ enhanced chest CT. Pretreatment evaluation for or extends into the surrounding soft tissues. Imaging findings
resectability involves nodal measurement at different at pelvic MRI will inform timing and type of surgery, assist
­stations, with suspected nodal involvement warranting a in the selection of patients for neoadjuvant therapy, and iden-
multi-­modality treatment approach rather than a primary tify tumors with high-risk features [34]. Tumors confined to
surgical approach. A relatively low sensitivity of 57% has the rectum can proceed to upfront surgery while those
been reported when using a 1 cm short axis measurement extending beyond the wall may be downsized prior to sur-
threshold for determining metastatic involvement [21–24]. gery with chemotherapy and radiation therapy.
Hence, CT is useful only as an initial tool to guide further N staging is based on the number and location of lymph
investigation into nodal involvement, potentially providing nodes involved by disease. Nodes outside the regional colon
targets for guiding endobronchial ultrasound (EBUS) sam- and/or rectal lymph node basin are considered metastatic.
pling which is ultimately necessary in most cases for accu- The M component of staging refers to distant metastatic dis-
rate staging. PET/CT may be performed as an initial staging ease. Common sites of metastatic disease for CRC are the
tool or in surveillance after treatment. It has a higher liver, lung, and peritoneum [35]. Treatment varies based on
reported sensitivity for detection of nodal metastases [25] the degree and location of metastatic disease, but in general
and has shown the ability to alter management in a signifi- there is a role for surgery in cases of limited metastatic dis-
cant number of patients, potentially avoiding mediastinos- ease. In most cases, initial assessment for metastatic disease
copy and nodal sampling in a smaller subset [26, 27]; in colorectal carcinoma is accomplished with contrast-­
however, this last assertion remains controversial [28]. enhanced CT of the chest, abdomen, and pelvis. PET/CT is
Contrast-enhanced abdominal CT is also typically per- not routinely used for initial staging, but could be considered
formed to evaluate common sites of distant disease such as when evaluating cases with limited metastatic disease to
278 K. L. Thomas et al.

assist with therapy planning [33]. Contrast-enhanced abdom- the American College of Radiology Appropriateness Criteria
inal MRI may be used to evaluate indeterminate liver lesions and the NCCN to perform multiphasic unenhanced and
detected by other imaging modalities or surgical planning contrast-­enhanced CT or MRI (CT urography or MR urogra-
when CT is deemed inadequate. phy) for lesion characterization including unenhanced,
nephrographic, and excretory phase imaging.
Abdominopelvic CT or MRI is useful for assessing the pri-
26.6 Anal Cancer mary tumor as well as tumor staging. Kidney cancers are
staged according to the TNM classification from the AJCC. T
Anal cancer is relatively uncommon although the worldwide staging is determined by tumor size as well as local invasion
incidence has been rising for many years [36]. In the United of the adjacent adrenal gland, perirenal soft tissues, and pres-
States, the lifetime risk of anal cancer is approximately 1 in ence of renal vein and IVC tumor thrombus. Nodal staging is
500 [36]. Most cases of anal cancer are now thought to be determined by the number of regional lymph nodes: none
secondary to infection with oncogenic subtypes of the human (N0); 1 node (N1); or multiple nodes (N2) [40].
papillomavirus (HPV). Worldwide it is estimated that 90% Renal cell carcinoma (RCC) is more likely to spread
of anal cancer cases are secondary to HPV infection [37]. hematogenously than lymphatically. Common sites of meta-
Risk factors for men include anal receptive sexual inter- static disease include the lungs and bones [39, 40]. Brain
course and in women risk increases with number of sexual metastases can be seen in up to 17% of renal cancer patients
partners and early age at first intercourse [36]. [42]. As such, imaging for staging of RCC is aimed at these
Anal cancer can be staged using the TNM system. The more common sites of disease. Guidelines recommend chest
primary tumor stage (T) is determined by the size of the radiography as part of initial staging with consideration of
tumor and whether it has invaded adjacent structures. N stag- chest CT when clinical presentation suggests distant disease.
ing is based on the number and location of involved lymph Similarly, bone scan may also be considered if there is clini-
nodes. Non-regional lymph nodes qualify as metastatic dis- cal concern of osseous metastasis, whereby CT or MRI of
ease. Common sites of metastatic disease (M) include the the head/brain is useful to screen for brain metastasis. At
liver and lungs [38]. Therapy is generally chemotherapy and present, the role of PET/CT in the staging of renal cancer is
radiation therapy with surgery reserved for very small tumors unclear and often limited by excreted radiotracer, and thus,
at initial detection or in a salvage setting. not usually recommended [39].
The main imaging modality used for staging of anal can-
cer is contrast-enhanced CT. Pelvic MRI can be useful to
determine if there is invasion of adjacent structures with 26.8 Urinary Bladder Cancer
large or aggressive tumors. PET/CT can be considered at
baseline or to evaluate for viable tumor after therapy. Recent Bladder cancer is the sixth most common cancer in the
work has found PET/CT can be useful in initial staging for United States [43]. It is most commonly diagnosed later in
characterization of metastatic lymph nodes compared to CT life (mean age of 73 years) [43]. Risk factors for bladder
alone [38]. cancer include male sex, tobacco exposure, exposure to
some industrial chemicals and medications, certain birth
defects and genetic syndromes, and chronic bladder irrita-
26.7 Kidney and Renal Pelvis Cancer tion [44].
Bladder cancer is staged using the TNM staging system.
It is estimated that in 2020, there will be 73,750 new cases of Primary tumor (T) staging is defined by depth of mural inva-
kidney and renal pelvis cancer diagnosed in the United States sion and whether the tumor extends to the muscular layer of
[3]. Kidney and renal pelvis cancers represent the sixth most the bladder wall. Non-muscle invasive bladder cancer can be
common cancer in men and eighth most common cancer in treated with local transurethral resection and intravesical
women in 2020, accounting for 5% and 3% of all new can- instillation of chemotherapy. When tumor invades the mus-
cers cases, respectively [1, 3]. The rate of new cases of kid- cular layer of the bladder wall, cystectomy is necessary for
ney and renal pelvis cancer has been on the rise, thought at complete tumor removal and local tumor control. T staging
least in part due to increased incidental detection with newer is often performed using transurethral resection and patho-
imaging techniques [39, 40]. According to the American logical evaluation of the depth of invasion, although recent
Cancer Society, the overall 5-year relative survival rate of advances may increase the role of MRI in the future.
kidney and renal pelvis cancer is 75% for all stages com- Contrast-enhanced pelvic MRI including high-resolution
bined, although the 5-year relative survival rate in those who T2-weighted imaging is gaining acceptance in the initial
present with localized disease is 93% [41]. evaluation of bladder cancer including muscular invasion
While several imaging modalities are available to charac- [45]. Several studies have also shown the ability of MRI to
terize indeterminate renal masses, it is recommended by both detect residual or recurrent tumor after therapy [45].
26 Primary Tumor Staging and Detection of Common Sites of Distant Metastatic Disease 279

Complete staging of bladder cancer typically includes whole-body PET/CT scanning [50, 51]. PET/CT has a rela-
contrast-enhanced abdominopelvic CT to assess the size, tively high sensitivity for the detection of metastatic disease,
location, and extent of the primary bladder tumor; assess- particularly in more advanced disease and may result in dis-
ment of locoregional lymph nodes in the pelvis; evaluate the ease management changes in nearly a third of patients [52].
kidneys and upper urinary tracts for synchronous genitouri-
nary tumors; and identify distant metastatic disease to
abdominal viscera or retroperitoneum. N staging is based on 26.10 Pancreatic Cancer
number and location of locoregional lymph nodes with non-­
regional nodes classified as metastatic disease. Common While not among the most common cancers, pancreatic
sites of metastatic disease include liver, lungs, and bones cancer still accounts for nearly 8% of all cancer deaths and
[35]. Chest imaging can be performed with either radio- only has a 10% combined 5-year survival rate [3]. Some of
graphs or CT based on clinical suspicion of pulmonary meta- the challenges in effective management of this deadly can-
static disease (whereby higher suspicion warrants chest CT) cer include a relatively large median tumor size of 30 mm
[46]. Whole-body bone scan may also be performed in cases at the time of presentation and metastatic disease affecting
with increased clinical concern for bony metastases [46]. a majority of patients [53]. While resectability numbers
PET/CT is generally not indicated but may be used as a are dismal for tumors 30 mm or greater at detection, these
problem-­solving tool. have been reported as high as 83% for tumors that are
smaller than 2 cm [54]. Larger tumors also have a higher
propensity to grow outside the confines of the pancreas via
26.9 Melanoma perineural spread, precluding curative intent via surgery,
and resulting in a high recurrence rate [55]. Hence, early
According to the American Cancer Society, over 100,000 new detection is the largest factor in improving survival from
cases of melanoma will be diagnosed in the United States in pancreatic cancer.
2020, resulting in nearly 7000 deaths. The risk for developing Staging of pancreatic cancer is via the AJCC TNM clas-
melanoma increases with age, however, it is not an uncom- sification. NCCN guidelines have been established for strati-
mon cancer in young adults, particularly females [3]. fying nonmetastatic tumors into primary resectable,
Five-year survival rates for melanoma vary greatly based borderline resectable, or nonresectable categories based on
on classification into localized, regional, and distant stages. the degree of vascular encasement [56]. Tumors without any
A 99% 5-year survival rate is reported for localized cancers, vascular involvement are deemed primary resectable, while
while that percentage drops to 66% for regional spread and there is some debate about the borderline resectable and non-
only 27% for distant metastatic disease [47]. Accurate initial resectable categories [57, 58]. Despite differences between
staging is important not only for risk stratification and prog- institutions on management, an accurate assessment of the
nosis but also for guiding treatment. anatomic relation of the tumor to critical vessels remains key
For detection of melanoma nodal spread, nuclear lympho- in determining the surgical outcome.
scintigraphy has become standard of care for identification Contrast-enhanced multi-detector CT utilizing a pancre-
of the sentinel draining node as well as in-transit nodes. atic protocol and multiplanar reconstructions is the primary
These represent embryonic lymphatic rests and can also har- means of staging pancreatic cancer and guiding treatment.
bor tumor along lymphatic channels. 99mTc sulfur colloid is CT has a relatively high sensitivity for detection of tumor
the FDA-approved radiopharmaceutical injected and imaged vascular encasement [59] and may also be able to discern
both dynamically and intraoperatively to identify these sites, extra-pancreatic and metastatic disease which would pre-
with performance superior to vital dyes [48, 49]. clude an R0 or microscopically negative surgical resection,
NCCN guidelines do not recommend baseline imaging and patients typically may receive neoadjuvant therapy fol-
for stage I or stage II melanoma, defined as local disease. lowed by reassessment of resectablility in this scenario.
Baseline imaging including contrast-enhanced CT of the Limitations for CT include an inability to diagnose or accu-
chest, abdomen, and pelvis should be performed for stage III rately assess the size of isoattenuating tumors, limited detec-
or regional melanoma, defined as spread either to adjacent tion of nodal metastases with enlarged nodes being frequently
skin, soft tissues or draining nodal basin. In more advanced reactive and small nodes harboring micrometastases, and
stage III disease, a brain MRI should also be considered. poor sensitivity for small liver metastases [60]. MRI is a
Baseline ultrasound is recommended in cases where nodal helpful adjunct to CT with better detection of isoattenuating
disease is equivocal by physical examination; otherwise, tumors and small liver metastases and should be performed
sonography is typically utilized in nodal basin surveillance preoperatively in surgical candidates [61] as well in patients
after resection. Stage IV metastatic disease should also be undergoing neoadjuvant therapy. PET/CT may also serve as
staged by similar CT and MRI exams, with consideration to an adjunct to CT staging, with increased sensitivity for dis-
280 K. L. Thomas et al.

tant metastatic disease with the caveat that nearly 30% of all 9. Prostate Cancer—Cancer Stat Facts [Internet]. [cited 2020 Oct 31].
Available from: https://seer.cancer.gov/statfacts/html/prost.html
tumors are not PET avid [62]. 10. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer
There is ongoing research in increasing tumor conspicu- J Clin. 2020;70:7–30. https://doi.org/10.3322/caac.21590. Wiley
ity with spectral CT utilizing low energy acquisition and split 11. Buyyounouski MK, Choyke PL, McKenney JK, Sartor O, Sandler
contrast bolus protocols [63, 64]. Newer MR-guided radia- HM, Amin MB, et al. Prostate cancer—major changes in the
American Joint Committee on Cancer eighth edition cancer stag-
tion therapy systems have been developed with superior tis- ing manual. CA Cancer J Clin. 2017;67:245–53. Available from:
sue contrast to guide real-time tumor demarcation and track https://pubmed.ncbi.nlm.nih.gov/28222223/
normal tissue motion with higher radiation doses [65]. 12. Coakley FV, Oto A, Alexander LF, Allen BC, Davis BJ, Froemming
AT, et al. ACR appropriateness criteria® prostate cancer—pre-
treatment Detection, surveillance, and staging. J Am Coll Radiol.
2017;14:S245–57. Available from: https://pubmed.ncbi.nlm.nih.
Conclusion and Perspective gov/28473080/
13. Freedman-Cass D, Shead DA, Schaeffer E, Antonarakis ES,
Armstrong AJ, et al. NCCN Guidelines Version 2. 2020 Prostate
• Imaging plays a key role in the diagnosis and stag- Cancer 2020.
ing of cancer. 14. Bubendorf L, Schöpfer A, Wagner U, Sauter G, Moch H, Willi N,
• Will improvements in imaging technology result in et al. Metastatic patterns of prostate cancer: an autopsy study of
1,589 patients. Hum Pathol. 2000;31:578–83. WB Saunders
earlier cancer detection?
15. Vinjamoori AH, Jagannathan JP, Shinagare AB, Taplin M-E, Oh
• Will advances in medical imaging help in the devel- WK, Van den Abbeele AD, et al. Atypical metastases from prostate
opment of more effective anticancer agents? cancer: 10-year experience at a single institution. Am J Roentgenol.
• Will functional imaging (e.g., PET/CT) play a 2012;199:367–72. https://doi.org/10.2214/AJR.11.7533.
16. Tanaka T, Yang M, Froemming AT, Bryce AH, Inai R, Kanazawa S,
larger role in the detection and staging of cancer in
et al. Current imaging techniques for and imaging spectrum of prostate
the future? cancer recurrence and metastasis: a pictorial review. Radiographics.
• How will the growing field of artificial intelligence 2020;40:709–26. Radiological Society of North America Inc.
affect cancer imaging? Available from: https://pubmed.ncbi.nlm.nih.gov/32196428/
17. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA A
• Can improvements in medical imaging decrease the
Cancer J Clin. 2020;70:7–30. https://doi.org/10.3322/caac.21590.
financial cost of cancer care? 18. Rami-Porta R, Asamura H, Travis WD, et al. Lung cancer - major
changes in the American joint committee on cancer eighth edi-
tion cancer staging manual. CA A Cancer J Clin. 2017;67:138–55.
https://doi.org/10.3322/caac.21390.
19. Hess KR, Varadhachary GR, Taylor SH, et al. Metastatic pat-
References terns in adenocarcinoma. Cancer. 2006;106:1624–33. https://doi.
org/10.1002/cncr.21778.
1. FastStats—Leading Causes of Death [Internet]. [cited 2020 Oct 20. Viadana E, Au K-L. Patterns of metastases in adenocarcinomas of
30]. Available from: https://www.cdc.gov/nchs/fastats/leading-­ man. An autopsy study of 4728 cases. J Med. 1975;6:1–14.
causes-­of-­death.htm 21. Toloza EM, Harpole L, McCrory DC. Noninvasive staging of non-­
2. Female Breast Cancer—Cancer Stat Facts [Internet]. [cited 2020 small cell lung cancer*: a review of the current evidence. Chest.
Oct 30]. Available from: https://seer.cancer.gov/statfacts/html/ 2003;123:137–46. https://doi.org/10.1378/chest.123.1_suppl.137S.
breast.html 22. Fritscher-Ravens A, Bohuslavizki KH, Brandt L, et al. Mediastinal
3. American Cancer Society. Annual Cancer Facts and Figures 2020. lymph node involvement in potentially resectable lung cancer:
[cited 2020 Oct 30]. Available from: https://www.cancer.org/con- comparison of CT, positron emission tomography, and endoscopic
tent/dam/cancer-­org/research/cancer-­facts-­and-­statistics/annual-­ ultrasonography with and without fine-needle aspiration. Chest.
cancer-­facts-­and-­figures/2020/cancer-­facts-­and-­figures-­2020.pdf 2003;123:442–51. https://doi.org/10.1378/chest.123.2.442.
4. Kalli S, Semine A, Cohen S, Naber SP, Makim SS, Bahl M. American 23. Arita T, Matsumoto T, Kuramitsu T, et al. Is it possible to differentiate
joint committee on cancer’s staging system for breast cancer, malignant mediastinal nodes from benign nodes by size? Reevaluation
eighth edition: what the radiologist needs to know. Radiographics. by CT, transesophageal echocardiography, and nodal specimen.
2018;38:1921–33. https://doi.org/10.1148/rg.2018180056. Chest. 1996;110:1004–8. https://doi.org/10.1378/chest.110.4.1004.
5. Lee SC, Jain PA, Jethwa SC, Tripathy D, Amashita MWY. 24. Prenzel KL, Mönig SP, Sinning JM, et al. Lymph node size
Radiologists’ role in breast cancer staging: providing key infor- and metastatic infiltration in non-small cell lung cancer. Chest.
mation for clinicians. Radiographics. 2014;34:330–42. Available 2003;123:463–7. https://doi.org/10.1378/chest.123.2.463.
from: www.rsna.org/rsnarights 25. Steinert HC, Hauser M, Allemann F, et al. Non-small cell lung can-
6. Lurie RH, Anderson BO, Abraham J, Aft R, Agnese D, Allison KH, cer: nodal staging with FDG PET versus CT with correlative lymph
et al. NCCN Guidelines Version 6. 2020 Breast Cancer 2020. node mapping and sampling. Radiology. 1997;202:441–6. https://
7. DiSibio G, French SW. Metastatic patterns of cancers : results doi.org/10.1148/radiology.202.2.9015071.
from a large autopsy study. Arch Pathol Lab Med. 2008;132:931–9. 26. Stamatis G. Staging of lung cancer: the role of noninvasive, mini-
Available from: https://pubmed.ncbi.nlm.nih.gov/18517275/ mally invasive and invasive techniques. Eur Respir J. 2015;46:521–
8. Chen MT, Sun HF, Zhao Y, Fu WY, Yang LP, Gao SP, et al. 31. https://doi.org/10.1183/09031936.00126714.
Comparison of patterns and prognosis among distant metastatic 27. Schimmer C, Neukam K, Elert O. Staging of non-small cell lung
breast cancer patients by age groups: a SEER population-based cancer: clinical value of positron emission tomography and medi-
analysis. Sci Rep. 2017;7. Available from: https://pubmed.ncbi. astinoscopy. Interact Cardiovasc Thorac Surg. 2006;5:418–23.
nlm.nih.gov/28835702/ https://doi.org/10.1510/icvts.2006.129478.
26 Primary Tumor Staging and Detection of Common Sites of Distant Metastatic Disease 281

28. Xia Y, Zhang B, Zhang H, et al. Evaluation of lymph node metastasis Experience. Ann Surg Oncol. 2000;7:469–74. https://doi.
in lung cancer: who is the chief justice? J Thorac Dis. 2015;7:231– org/10.1007/s10434-­000-­0469-­z.
7. https://doi.org/10.3978/j.issn.2072-­1439.2015.11.63. 49. Intenzo CM, Truluck CA, Kushen MC, et al. Lymphoscintigraphy
29. Komaki R, Putnam JB Jr, Walsh G, et al. The management of supe- in cutaneous melanoma: an updated Total body atlas of senti-
rior sulcus tumors. Semin Surg Oncol. 2000;18:152–64. https://doi. nel node mapping. Radiographics. 2009;29:1125–35. https://doi.
org/10.1002/(sici)1098-­2388(200003)18:2<152:aid-­ssu9>3.0.co;2-­h. org/10.1148/rg.294085745.
30. MacDonald SLS, Hansell DM. Staging of non-small cell lung can- 50. Freeman M, Laks S. Surveillance imaging for metastasis in high-­
cer: imaging of intrathoracic disease. Eur J Radiol. 2003;45:18–30. risk melanoma: importance in individualized patient care and sur-
https://doi.org/10.1016/s0720-­048x(02)00300-­5. vivorship. Melanoma Manag. 2019;6:12. https://doi.org/10.2217/
31. Colorectal Cancer Statistics | How Common Is Colorectal Cancer? mmt-­2019-­0003.
[Internet]. [cited 2020 Oct 31]. Available from: https://www.cancer. 51. Coit DG, Thompson JA, Albertini MR, et al. Cutaneous melanoma,
org/cancer/colon-­rectal-­cancer/about/key-­statistics.html Version 2.2019, NCCN clinical practice Guidelines in oncol-
32. Dekker E, Tanis PJ, Vleugels JLA, Kasi PM, Wallace MB. Colorectal ogy. J Natl Compr Cancer Netw. 2019;17:367–402. https://doi.
cancer [internet]. Lancet. 2019:1467–80. Lancet Publishing Group. org/10.6004/jnccn.2019.0018.
Available from: https://pubmed.ncbi.nlm.nih.gov/31631858/ 52. Krug B, Crott R, Lonneux M, et al. Role of PET in the initial staging
33. Benson AB, Al-Hawary MM, Arain MA, Chen Y-J, Ciombor KK, of cutaneous malignant melanoma: systematic review. Radiology.
Cohen S, et al. NCCN Guidelines Version 42020 Colon Cancer 2008;249:836–44. https://doi.org/10.1148/radiol.2493080240.
Continue NCCN Guidelines Panel Disclosures; 2020. 53. Hidalgo M. Pancreatic cancer. N Engl J Med. 2010;362:1605–17.
34. Horvat N, Rocha CCT, Oliveira BC, Petkovska I, Gollub MJ. MRI https://doi.org/10.1056/NEJMra0901557.
of rectal cancer: tumor staging, imaging techniques, and manage- 54. Agarwal B, Correa AM, Ho L. Survival in pancreatic carci-
ment. Radiographics. 2019;39:367–87. https://doi.org/10.1148/ noma based on tumor size. Pancreas. 2008;36:15–20. https://doi.
rg.2019180114. org/10.1097/mpa.0b013e31814de421.
35. Metastatic Cancer—National Cancer Institute [Internet]. [cited 55. Chang ST, Jeffrey RB, Patel BN, et al. Preoperative multidetec-
2020 Oct 31]. Available from: https://www.cancer.gov/types/ tor CT diagnosis of extrapancreatic perineural or duodenal inva-
metastatic-­cancer sion is associated with reduced postoperative survival after
36. Key Statistics for Anal Cancer [Internet]. [cited 2020 Oct 31]. pancreaticoduodenectomy for pancreatic adenocarcinoma: pre-
Available from: https://www.cancer.org/cancer/anal-­cancer/about/ liminary experience and implications for patient care. Radiology.
what-­is-­key-­statistics.html 2016;281:816–25. https://doi.org/10.1148/radiol.2016152790.
37. Valvo F, Ciurlia E, Avuzzi B, Doci R, Ducreux M, Roelofsen F, 56. Tempero MA, Malafa MP, Chiorean EG, et al. Pancreatic adenocar-
et al. Cancer of the anal region [internet]. Crit Rev Oncol Hematol. cinoma, version 1.2019. J Natl Compr Cancer Netw. 2019;17:202–
2019:115–27. Elsevier Ireland Ltd.; Available from: https:// 10. https://doi.org/10.6004/jnccn.2019.0014.
pubmed.ncbi.nlm.nih.gov/30819440/ 57. Varadhachary GR, Tamm EP, Abbruzzese JL, et al. Borderline
38. Durot C, Dohan A, Boudiaf M, Servois V, Soyer P, Hoeffel C. Cancer resectable pancreatic cancer: definitions, management, and role of
of the anal canal: diagnosis, staging and follow-up with MRI. Korean preoperative therapy. Ann Surg Oncol. 2006;13:1035–46. https://
J Radiol. 2017;18:946–56. Korean Radiological Society; Available doi.org/10.1245/ASO.2006.08.011.
from: https://pubmed.ncbi.nlm.nih.gov/29089827/ 58. Khorana AA, Mangu PB, Berlin J, et al. Potentially curable pancre-
39. Vikram R, Beland MD, Blaufox MD, Moreno CC, Gore JL, atic cancer: American Society of Clinical Oncology clinical prac-
Harvin HJ, et al. ACR Appropriateness criteria renal cell carci- tice guideline update. J Clin Oncol. 2017;35:2324–8. https://doi.
noma ­staging. J Am Coll Radiol. 2016;13:518–25. Available from: org/10.1200/JCO.2017.72.4948.
https://pubmed.ncbi.nlm.nih.gov/27016804/ 59. Zamboni GA, Kruskal JB, Vollmer CM, et al. Pancreatic adeno-
40. Ng CS, Wood CG, Silverman PM, Tannir NM, Tamboli P, Sandler carcinoma: value of multidetector CT angiography in preoperative
CM. Renal cell carcinoma: diagnosis, staging, and surveillance. evaluation. Radiology. 2007;245:770–8. https://doi.org/10.1148/
AJR Am J Roentgenol. 2008:1220–32. Available from: https:// radiol.2453061795.
pubmed.ncbi.nlm.nih.gov/18806169/ 60. Zins M, Matos C, Cassinotto C. Pancreatic adenocarcinoma
41. Kidney Cancer Early Detection, Diagnosis, and Staging. [cited staging in the era of preoperative chemotherapy and radiation
2020 Oct 31]; Available from: www.cancer.org/treatment/ therapy. Radiology. 2018;287:374–90. https://doi.org/10.1148/
understanding-­your-­diagnosis/tests/imaging-­radiology-­ radiol.2018171670.
42. Shuch B, La Rochelle JC, Klatte T, Riggs SB, Liu W, Kabbinavar 61. Kim HW, Lee J-C, Paik K-H, et al. Adjunctive role of preoperative
FF, et al. Brain metastasis from renal cell carcinoma. Cancer. liver magnetic resonance imaging for potentially resectable pancre-
2008;113:1641–8. https://doi.org/10.1002/cncr.23769. John Wiley & atic cancer. Surgery. 2017;161:1579–87. https://doi.org/10.1016/j.
Sons, Ltd surg.2016.12.038.
43. Degeorge KC, Holt HR, Hodges SC. Bladder cancer: diagnosis and 62. Rijkers AP, Valkema R, Duivenvoorden HJ, et al. Usefulness of
treatment. Am Fam Physician. 2017; Available from: www.aafp. F-18-fluorodeoxyglucose positron emission tomography to confirm
org/afp suspected pancreatic cancer: a meta-analysis. Eur J Surg Oncol.
44. Bladder Cancer Risk Factors [Internet]. [cited 2020 Oct 31]. 2014;40:794–804. https://doi.org/10.1016/j.ejso.2014.03.016.
Available from: https://www.cancer.org/cancer/bladder-­cancer/ 63. Macari M, Spieler B, Kim D, et al. Dual-source dual-energy
causes-­risks-­prevention/risk-­factors.html MDCT of pancreatic adenocarcinoma: initial observations with
45. Gurram S, Muthigi A, Egan J, Stamatakis L. Imaging in localized data generated at 80 kVp and at simulated weighted-average 120
bladder cancer: can current diagnostic modalities provide accurate kVp. Am J Roentgenol. 2010;194:27–32. https://doi.org/10.2214/
local tumor staging? Curr Urol Rep. 2019;20(12):1–7. Springer; AJR.09.2737.
Available from: https://pubmed.ncbi.nlm.nih.gov/31781871/ 64. Brook OR, Gourtsoyianni S, Brook A, et al. Split-bolus spectral
46. Flaig TW, Spiess PE, Chair V, Agarwal N, Bangs R, Patient Advocate multidetector CT of the pancreas: assessment of radiation dose
M, et al. NCCN Guidelines Version 6.2020 Bladder Cancer 2020. and tumor conspicuity. Radiology. 2013;269:139–48. https://doi.
47. National Cancer Institute Surveillance, Epidemiology, and End org/10.1148/radiol.13121409.
Results Database. Cancer Stat Facts: Melanoma of the Skin. 65. Kupelian P, Sonke J-J. Magnetic resonance-guided adaptive radio-
48. Cascinelli N, Belli F, Santinami M, et al. Sentinel lymph node therapy: a solution to the future. Semin Radiat Oncol. 2014;24:227–
biopsy in cutaneous melanoma: the WHO Melanoma Program 32. https://doi.org/10.1016/j.semradonc.2014.02.013.
 Imaging Modalities in the Detection
and Diagnosis of Metastatic Disease 27
S. Ali, S. Fenerty, and P. Jonnalagadda

Abstract 27.1 Introduction

Imaging is a critical tool in the evaluation and workup of
patients with known or suspected metastatic disease and Metastatic disease is an indication of advanced tumor and
in the follow-up of disease progression after surgery, che- generally indicates a poor prognosis. Historically, before the
motherapy, and radiation therapy. The imaging modality advent of medical imaging, patients frequently presented
utilized is dependent on the primary tumor and location of with advanced disease at the time of diagnosis. With the
the metastatic deposit(s). Traditional imaging modalities advent of plain film radiography little changed, as the sensi-
include plain film radiography, computed tomography tivity was such that only advanced disease was detectable,
(CT), ultrasound, magnetic resonance imaging (MRI), and generally only advanced bone or pulmonary metastases
nuclear medicine, and more recently fused imaging such were diagnosed. CT imaging revolutionized the management
as PET-CT and PET-MRI. A single modality may be used of cancer patients as the spatial and contrast resolution far
in isolation; however, it is more common that various exceeds that of plain films. For the first time, abdominal, pel-
combinations are used depending on the specific tumor. vic, and central nervous system structures could be imaged,
Interventional radiology complements these diagnostic tumors staged, and more appropriate therapy started. MRI,
imaging modalities for diagnosis (biopsy and tissue sam- with its superior contrast resolution, was another step in ear-
pling) as well as therapy (ablation, intravascular chemo- lier disease diagnosis, particularly in the central nervous sys-
therapy) and preoperative or palliative tumor embolization. tem, abdomen, and pelvis. Fusion of cross-sectional imaging
In this chapter, we outline the basics of physics and the with nuclear medicine (PET-CT and recently PET-MRI) sub-
utility of each modality and indicate future directions. sequently enabled a functional imaging overlay of the meta-
bolic information provided by PET with the anatomic detail
of CT and MRI. This has reduced the amount of unnecessary
Learning Objectives biopsies and allows for follow-up after treatment. Historically,
• Understand the basic physics of the respective nuclear medicine proved most useful in metastatic bone
imaging modalities. tumor assessment, but more recently has afforded significant
• Understand the benefits and limitations of each advances in the early detection and diagnosis of other tumors.
modality.
• Understand radiation dose for different modalities.
• Apply knowledge of the various modalities to max- 27.2 Plain Film Radiography
imize benefit.
Radiography relies on X-ray radiation to produce a two-­
dimensional image. X-rays are a form of ionizing radiation,
which means they possess sufficient energy to potentially
remove electrons from an atom, thus giving it a charge and
creating an ion. X-rays were discovered by German physicist
Wilhelm Roentgen in 1895 when he attempted to see if cath-
S. Ali (*) · S. Fenerty · P. Jonnalagadda
Temple University Hospital, Philadelphia, PA, USA ode rays could pass through glass and realized that the rays
e-mail: [email protected]; could pass through the heavy black paper covering the cath-
[email protected]; ode tube. He named these mysterious rays “X” rays and soon
[email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 283
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_27
284 S. Ali et al.

ascertained that the rays could also pass through human tis- patients. Multiplanar reconstructions can aid in visualizing
sue, hence the term X-rays, which has become widely known pathology and in surgical planning. The administration of
in the medical literature. intravenous iodinated contrast media increases contrast reso-
A typical X-ray unit consists of a source, patient, image lution and therefore diagnostic accuracy. More recent formu-
receptor, processor, and viewbox or PACS (Picture Archiving lations of low osmolar contrast media have resulted in
and Communication System) [1]. The source is the X-ray significantly fewer side effects than prior formulations and
tube, which emits a beam containing a spectrum of photon are generally well tolerated [5]. Oral contrast media in
energy (the X-ray beam). When the beam enters the body, a abdominal/pelvic imaging also serves the same purpose,
portion of the radiation is absorbed in a process known as although this may not be routinely required in oncology fol-
attenuation. The remnant beam passes through the patient low-­up imaging [6]. One caveat is that there is an increased
and is detected by radiographic film or, as is much more radiation dose compared to plain films, especially when
common in current practice, a digital plate detector. Due to patients undergo multiple scans as is typical for oncologic
differential absorption of the X-ray photons, areas of image imaging. However, the benefits usually greatly outweigh the
receptor receive different amounts of radiation, resulting in risks in these patients.
an image with areas of contrasting density [2]. Those por- The fundamental principle of CT imaging is the use of a
tions of the image receptor that receive the most radiation rotating X-ray tube within a gantry. Detectors located 180°
(i.e., portions of the beam experiencing the least attenuation) from the tube perform the same function as a conventional
will be more exposed and are processed to appear darker. radiograph. The rotation of the tube allows multiple X-ray
Conversely, those receiving the least radiation will appear measurements of the same point from different angles which
lighter. The denser the structure encountered by the beam, can be reconstructed using computer algorithms to generate
the greater the photon attenuation. This is why bone, which an image or “slice” of tissue. The density of the tissue is
is high density, appears white, while lungs, which contain detected and scaled using Hounsfield Units (HU), ranging
mostly air, appear black, with soft tissues appearing varying from −1000 HU for air to +2000 HU for very dense bone and
shades of gray. Therefore, if the tumor destroys bone then the greater than 3000 for metal. Over the past three decades, the
affected area will attenuate fewer photons and appear darker use of a two-dimensional array of detector elements (multide-
(lytic) on a background of whiter normal bone. Conversely, tector CT) compared to the linear array seen in conventional
osteoblastic metastases will attenuate more photons and or helical scanners has dramatically increased the speed and
appear whiter than normal bone. Similarly, lung metastases resolution of these scans. Multidetector CT reduces the effect
in a background of very low-density air will appear as denser of patient motion and increases patient throughput without
white lesions. significantly increasing radiation dose if key imaging param-
In the past, image intensifying screens were used to eters are observed, including the use of automatic exposure
decrease the radiation dose. With the introduction of digital control (AEC) [7]. When using AEC, care must be taken in
plates and computerized radiography systems, intensifying obese patients or when imaging areas with metallic hardware,
screens are no longer necessary [3]. Processing of films is which can result in paradoxical dose increases [8]. More
also of historical interest, as the images generated by the digi- recent advances include dual-energy CT, where source tech-
tal plate are now transmitted electronically to the PACS sys- niques include the use of two X-ray tubes or fast Kev switch-
tem and viewed and stored electronically. The format used for ing (among other techniques), or retrospective dual-layer
storage and transfer universally is called DICOM (Digital technology where the energy separation is detector based
Imaging and Communications in Medicine). Image storage with the outer layer absorbing higher energy photons and the
can be done locally, however, with the vast amount of data not inner layer lower energy photons [9]. This technology pro-
just from radiographs but also other imaging that is currently vides additional functional information including detection of
generated by nationwide radiology departments, data is also bone marrow edema, which can be extremely useful in sus-
usually backed up by large third-party vendors [4]. pected metastatic disease where the morphologic CT appear-
ance is normal although bone marrow edema is present, and
for metal artifact reduction which can be especially useful in
27.3 Computed Tomography (CT) postoperative cancer imaging [10].

Computed tomography (CT), formerly known as computed

axial tomography (CAT), has probably been the greatest 27.4 X-Ray Tomography
advancement in the detection and staging of metastatic dis-
ease. Its exquisite spatial and contrast resolution has revolu- The most relevant application of tomosynthesis is digital
tionized the field of oncologic imaging (Fig. 27.1). These breast tomosynthesis (DBT), a technique that generates a
scans are quick, fairly cost-effective, and well tolerated by volumetric reconstruction of the whole breast from a set
27 Imaging Modalities in the Detection and Diagnosis of Metastatic Disease 285

a b

Fig. 27.1 56-year-old man with melanoma. (a) Plain films are normal. mass consistent with metastasis. This illustrates the increasing contrast
(b) Noncontrast CT shows subtle effacement of the left adductor mus- resolution from plain films through CT and MRI
cle fat relative to the right. (c) Axial T2W MRI shows a large soft tissue

number of two-dimensional projections obtained at different 27.5 Ultrasound

X-ray tube angles, which is emerging as the standard of care
for breast imaging in both screening and diagnostic settings. Ultrasound (US) imaging uses a probe containing multiple
Conventional mammography is two dimensional; thus, acoustic transducers to send pulses of high-frequency sound
­overlap of breast tissue, particularly in women with denser (“ultrasound” waves, generally ranging from 2 to 20 MH),
breasts, may both obscure cancers and simulate abnormali- into the body [12]. When the sound waves encounter mate-
ties where none are present. In DBT, the X-ray tube pivots in rial of a different density (acoustic impedance), a portion of
an arc (ranging from 15° to 60°) over the breast while acquir- the wave is reflected back to the transducer and detected as
ing multiple low-dose images, with each image delivering a an echo, which is then converted back into electrical impulses
fraction of the radiation dose provided in a standard mam- by the transducer crystals and processed to form the ultra-
mogram and obtaining structural information for each layer. sound image displayed on the screen. A greater difference in
Through a computer algorithm, these are reconstructed into acoustic impedance (the resistance an ultrasound beam
a series of stacked images, not dissimilar to CT “slices,” pro- encounters as it passes through tissue) results in a higher pro-
viding localization information and improved lesion charac- portion of reflected sound and proportionate decrease in the
terization [11] (Fig. 27.2), which may decrease or eliminate amount of transmitted sound. With significant differences in
the need for additional diagnostic workup, particularly in tissue density, sound is entirely reflected, resulting in com-
those with dense breasts.
286 S. Ali et al.

a b c

Fig. 27.2 55 year old woman for screening mammogram. (a) 2D Right Ultrasound shows area of shadowing and distortion corresponding to
MLO shows scattered fibroglandular tissue with no discrete mammo- tomosynthesis finding, biopsied to reveal invasive mammary carci-
graphic abnormality (b) Corresponding tomosynthesis view shows noma. Images courtesy of Mindy Yang, MD
architectural distortion in the upper outer right breast (arrow) (c)

plete acoustical shadowing and the inability to see deeper, as tions of US include its greater dependence on operator skill,
occurs behind bones, dense calcifications, and air. limited reproducibility, and potential for inadequate acousti-
Four different modes of ultrasound are used in medical cal windows (often due to bowel gas or patient obesity).
imaging: Incomplete visualization of lesions due to large lesion size or
acoustic shadowing from calcifications, variable sonographic
• A-mode (amplitude mode), the simplest type of ultra- appearance of malignancies, and poor sensitivity in detection
sound, in which a single transducer scans a line through of small or isoechoic tumors are other potential limitations.
the body with the echoes plotted on screen as a function Furthermore, US examinations are typically targeted to an
of depth. This is used for echoencephalography and area of interest, and hence are seldom the primary modality
echo-ophthalmoscopy. for comprehensive staging, but may be a useful adjunct in the
• B-mode (brightness mode, 2D mode, or gray-scale imag- focused assessment of certain organs (e.g., detection of liver
ing), in which a linear array of transducers simultane- or ovarian metastases or residual axillary nodal disease in
ously scan a plane through the body which is displayed as breast cancer), can assess vascular tumor involvement, and
a two-dimensional image on screen. demonstrate malignant ascites when present. In some
• M-mode (motion mode), a time motion display of the instances, US examination performed for underlying symp-
ultrasound wave along a chosen ultrasound line, used for toms (e.g., abdominal pain) may reveal a previously undiag-
analyzing moving body parts, typically used in cardiac, nosed malignancy (Fig. 27.3).
fetal cardiac, and lung US. Doppler imaging can provide important directional blood
• Doppler mode, typically used to assess blood flow, allow- flow information, but it is most effective for evaluating large
ing visualization of flow direction and velocity (color vessels such as carotid arteries, leg veins, and major visceral
Doppler) and yielding graphical representations of flow vessels of the abdomen, including hepatic artery, central por-
velocity over time (spectral Doppler). tal, and hepatic veins. The ability of Doppler imaging to
detect blood flow at the perfusional level is limited. Contrast-­
US is a convenient, cost-effective, and widely available enhanced US (CEUS), which involves the intravenous
imaging tool, free of ionizing radiation and capable of being administration of microbubble contrast agents (microbubbles
performed bedside or brought into the operating room of perfluorocarbon, nitrogen gas, or sulfur hexafluoride sta-
through the use of portable and point-of-care ultrasound bilized in a phospholipid membrane) and specialized imag-
machines. Most importantly, it can provide accurate diagnos- ing techniques, can provide dynamic real-time imaging
tic information comparable to CT and MRI. The main limita- blood flow and tissue perfusion information with high spatial
27 Imaging Modalities in the Detection and Diagnosis of Metastatic Disease 287

a b

c d

Fig. 27.3 67 year old man underwent abdominal ultrasound for evalu- hancing masses in the liver (c) and large mass involving pancreatic tail/
ation of abdominal pain, which showed multiple hyperechoic liver splenic hilum (d). Subsequent biopsy of the depicted right hepatic lobe
masses, one of which is shown in (a); and a mass abutting the splenic mass showed metastatic pancreatic adenocarcinoma
hilum (b). Correlative portal venous phase CT images show hypoen-

and temporal capability, thus allowing ultrasound to play a dence for use of microbubble contrast agents includes their
competitive role relative to CT and MRI in the evaluation of ease of use, positive contribution to patient care, lack of
both solid and hollow organ pathology. Vascular CEUS con- nephrotoxicity, and favorable safety profile [14].
trast agents show distinct enhancement phases much like CT Variations of CEUS such as dynamic contrast-enhanced
and MRI contrast media, permitting dynamic evaluation of ultrasound (DCE-US) and ultrasound molecular imaging
the vascularity of a target lesion, most commonly in the liver (targeted CEUS), which utilizes molecularly targeted ultra-
or kidney, which may aid in diagnosis [13]. Supporting evi- sound contrast agents, are emerging techniques for quantita-
288 S. Ali et al.

tive monitoring of treatment effects and earlier cancer is used to produce MR images. This alignment (or magneti-
detection. CEUS has an advantage over contrast-enhanced zation) is subsequently perturbed by introducing an external
MRI and CT in patients with contraindications such as renal radio frequency (RF) energy to the body. The nuclei return to
failure or iodinated contrast allergy and also allows for their resting alignment through various relaxation processes,
dynamic and repeat examinations. In the United States, a in the process emitting RF energy, which is detected by a
greater availability of CT and MRI and limited availability of radiofrequency receiver. Different tissues realign at different
US contrast agents have contributed to its lesser role in onco- rates and image acquisition relies on subtle differences in the
logic imaging, although in Europe and Canada, where it is intrinsic behavior of hydrogen protons bound to different
widely used, CEUS has shown high (87–91%) accuracy in soft tissues and fluids to generate image contrast [18]. Using
detection and characterization of liver lesions [15]. Fourier transformation, the frequency information from each
US guidance is also a useful tool for biopsy of suspected location in the imaged plane is converted to corresponding
malignancies, while intraoperative US (IOUS) can be impor- intensity levels, which are displayed as shades of gray in a
tant in localization of tumors for ablative techniques or to matrix arrangement of pixels. By varying the sequence of RF
guide intraoperative biopsy or surgical resection [16]. In pulses, different types of images are created.
many centers with equivalent expertise in US and CT, US The main magnetic field of an MRI scanner is generated
guidance is the method of choice for percutaneous interven- by a large electric current flowing through wires formed into
tional procedures if the lesion is easily visible and accessible a loop in the magnet of the imaging system in which the
sonographically, as it provides real-time imaging, does not patient lies. These wires have zero resistance to electrical cur-
use ionizing radiation, is readily available, and inexpensive. rent when maintained at superconducting temperatures
US fusion imaging, which involves the co-registered display (−269 °C), which is achieved by immersing the coil in liquid
of real-time US with a reference series from another modal- helium. A power supply is used to inject an electric current
ity, such as CT, MRI, or PET, arranged side-by-side or over- into the coils until the desired magnetic field is reached. The
laid on the US monitor, can further aid in the detection and power supply can then be removed. Once established, the cur-
localization of lesions with low conspicuity on standard rent continues in the closed-loop coil for years without sig-
B-mode US [17]. It is also possible to indicate the desired nificant decline, provided that liquid helium levels are
needle trajectory for guidance during the procedure. US maintained. Consequently, the magnetic field is always pres-
fusion imaging can be associated with the use of different ent. The majority of MRI systems in clinical use have a mag-
ultrasound techniques such as color Doppler US, elastogra- netic field strength of 1.5 T (tesla) or 3 T, although field
phy, and CEUS, for better localization and characterization strengths of 7 T are now commercially available [19]. One
of lesions. During biopsy, the use of color Doppler US allows Tesla is equal to 10,000 gausses (G); in comparison, the
for identification of major vascular structures which should earth’s magnetic field is only about 0.5 Gauss. There are
preferably be avoided during needle insertion, while CEUS unique potential risks associated with MR imaging such as
can increase the conspicuity of lesions to be treated and vas- projectile injury, translational force and torque from the main
cular structures to be avoided. magnet, burns and conductive loops from radiofrequency
coils, and acoustic injury from the gradient coils. In order to
avoid MRI-related accidents, there is an MRI safety zoning
27.6 Magnetic Resonance Imaging system (zones 1–4) with progressive restriction of entry and
increased supervision for higher zones closer to the magnet.
Unlike CT and radiographs, magnetic resonance imaging The basic MRI sequences are T1-weighted (T1W) and
(MRI) uses nonionizing radiation to yield diagnostic images. T2-weighted (T2W) sequences, which are predominately
MRI is based on the magnetization properties of atomic determined by intrinsic T1 and T2 properties of tissue,
nuclei, taking advantage of the high prevalence of hydrogen respectively. A detailed discussion of MRI physics is beyond
in the body and the magnetic properties of protons in a the scope of this section, however, in general, fluid is bright
hydrogen atom. Each hydrogen proton acts like a tiny mag- (or white) on T2W images and dark on T1W images. Fat is
net which rotates along its long axis and generates a small bright on both but can be suppressed (darkened) using fat
magnetic field. These are randomly oriented so that their suppression techniques to create appropriate contrast. T1W
magnetic fields cancel out. When subject to powerful, uni- imaging can also be performed with intravenous Gadolinium-­
form, external magnetic field (referred to as B0), protons will based contrast agents. Gadolinium is a paramagnetic contrast
either align either parallel (spin up) or antiparallel (spin enhancement agent which shortens T1 relaxation time, thus
down) to the direction of the magnetic field, with a slight demonstrating increased signal on T1W sequences. Fluid-­
excess in the lower energy parallel direction, producing a attenuated inversion recovery (FLAIR) is a sequence com-
“net magnetization” parallel to the main magnetic field. This monly used in CNS imaging, which removes or nulls the
net magnetization becomes the source of our MR signal and signal of cerebrospinal fluid, making it easier to differentiate
27 Imaging Modalities in the Detection and Diagnosis of Metastatic Disease 289

a b c

Fig. 27.4 65 year old man with prostate cancer and increasing left hip pain. (a) AP radiograph is negative. (b) Coronal CT image shows no focal
lesion. (c) Coronal T1W MRI shows diffuse and focal tumor infiltration consistent with metastases

abnormal hyperintense lesions. Diffusion-weighted imaging normal fatty marrow by malignant cells (Fig. 27.4). On a
(DWI) is another commonly utilized sequence based on T2W fat-saturated sequence, osteolytic metastases usually
measuring the random Brownian motion of water molecules demonstrate hyperintense signal due to elevated water con-
and is particularly useful in tumor characterization and cere- tent and enhancement on post-contrast study due to increased
bral ischemia. In simplified terms, highly cellular tissues, or vascularity [21]. Osteoblastic metastases are typically low
those with cellular swelling, will exhibit lower diffusion signal (dark) on both sequences.
coefficients.
While CT has better spatial resolution, the soft-tissue
contrast resolution of MRI (or ability to distinguish between 27.7 Nuclear Medicine
differences in image intensity) is much better than CT. MRI
is therefore often better able to delineate the extent of tumor Nuclear medicine techniques quantitatively assess the func-
invasion and assess for residual disease after resection or tion of tissue or tumor cells as compared to morphologic
treatment. MRI is highly sensitive in the detection of many imaging techniques such as plain films, CT, and MRI where
osseous and abdominopelvic metastases and is often used as only the anatomy of a lesion is demonstrated. Two radionu-
a problem-solving tool for equivocal imaging findings on CT clide functional techniques, planar scintigraphy and 18F-­
or US. Although the MRI appearance of metastatic disease is FDG positron emission tomography (PET), with or without
variable and typically dependent on the primary tumor, cer- CT/MRI integration, play a major role in the diagnosis and
tain MR enhancement characteristics, such as an early ring evaluation of tumors including skeletal metastasis [22].
enhancement and peripheral washout of liver lesions, are
considered specific to metastases and allow differentiation of
metastases from benign lesions. Metastatic disease may be 27.7.1 Bone Scan
visible on MRI prior to other imaging modalities. Osseous
metastases, for example, are detectable on MRI before radio- To illustrate the concept of planar scintigraphy, we will use
nuclide bone scan [20]. Normal bone marrow demonstrates osseous metastases as an example. Bone scan is a highly sen-
high signal on T1W sequence due to a high percentage of fat. sitive, frequently performed, relatively inexpensive, and
On T1W sequences most osseous metastases demonstrate widely available radionuclide technique for the detection of
discrete foci or diffuse low signal due to the replacement of these lesions. A major advantage of a bone scan is that imag-
290 S. Ali et al.

a b c

Fig. 27.5 73-year-old man with prostate cancer. (a) Planar bone scan demonstrates focal intense tracer uptake in midthoracic spine. (b, c) SPECT
CT aids in anatomic localization of the lesion allowing for targeted treatment. Images courtesy of Bhishak Kamat, MD

ing of the entire skeleton can be performed. The procedure is

performed after intravenous administration of technetium- 27.8 PET-CT and PET-MRI
­99 m labeled methylene diphosphonate. Images are obtained
2–6 hours later. The tracer rapidly accumulates in bone and PET-CT is a fusion of functional planar radionuclide imag-
emits radiation (gamma rays), which are detected by a ing and cross-sectional anatomic imaging (CT).
gamma camera from which planar anterior and posterior Fluorodeoxyglucose F18 (18F-FDG) is most commonly used
images are generated. The labeled diphosphonates enable in oncology because it is a radiopharmaceutical analog of
visualization of metastases in metabolically active tumors glucose and there is high glucose metabolism of malignant
such as prostate and breast cancers. A two or three-phase cells. This radiotracer is produced in a cyclotron. The patient
protocol at earlier time points can provide additional infor- is injected intravenously and is imaged approximately
mation regarding blood flow and soft tissue uptake. 60 min following the injection of 18F-FDG. The tracer is
Metastases that are not intensely hypermetabolic may not taken up by metabolically active tumor cells, and unlike glu-
be detected in planar imaging and the detection of these cose, the breakdown product does not undergo further
metastases is markedly improved by the addition of single-­ metabolism and is trapped in metabolically active cells.
photon emission computed tomography (SPECT) or fused Positrons emitted from the F18 annihilate electrons releasing
SPECT-CT (Fig. 27.5). The sensitivity and specificity for two annihilation photons at 511 keV traveling in opposite
detection of metastasis increase over 90% when SPECT-CT directions that are detected by PET detectors using a tech-
is used, especially for prostate metastases [23]. nique called coincidence detection.
27 Imaging Modalities in the Detection and Diagnosis of Metastatic Disease 291

a b

Fig. 27.6 50-year-old man with prostate cancer and rising PSA. (a) CT shows predominantly lytic lesion in right iliac bone. (b) Bone scan shows
no tracer uptake. (c) PET-CT F-18 Axumin shows focal increased uptake in the lesion consistent with metastasis

The PET scanner is located behind the CT scanner in the can provide an objective measure of the response to treat-
same gantry. 18F-FDG PET-CT acquires PET and CT data in ment. It is more sensitive and specific in certain cancers and
the same imaging session, therefore allow accurate anatomi- primarily aids in staging, restaging, and recurrence [24]. 18F-­
cal localization of lesions detected on 18F-FDG PET scan. FDG PET-CT Axumin (Fluciclovine) demonstrates intense
The assessment of radiotracer uptake by normal physiologic focal uptake in lytic prostate cancer metastases and is a prob-
and pathologic tissues is performed by visual inspection and lem solver in those patients with a negative conventional
the standardized uptake value (SUV) which is a semiquanti- bone scan due to lack of osteoblastic activity (Fig. 27.6).
tative method. Normal physiologic uptake of the agent is PET-MRI is a more recent imaging modality where the
seen in the brain, thymus, bone marrow, intestines, and exquisite contrast resolution of MRI is combined with func-
brown fat. Increased SUV values are seen in tumors. A major tional data of 18F-FDG, similar to PET-CT, however, with
advantage of 18F-FDG PET is the ability to compare the max- better anatomic detail. Its clinical efficacy is currently being
imum SUV value of metastases on sequential studies which evaluated [25] (Fig. 27.7).
292 S. Ali et al.

a c

Fig. 27.7 70-year-old woman with history of breast cancer and DVT. fused image increases diagnostic accuracy. Biospy confirmed intramus-
(a) Fused PET- MRI image shows hypermetabolic mass (SUV 2.69) in cular myxoma. Images courtesy of Andrew Newberg, MD and Anthony
left thigh corresponding to T2 hyperintense mass on axial T2W MRI Bazzan, MD
image (b) and hypermetabolic focus on planar PET image (c). The
27 Imaging Modalities in the Detection and Diagnosis of Metastatic Disease 293

7. Mahesh M. MDCT physics: the basics—technology, image qual-

Conclusion ity and radiation dose. Philadelphia, PA: Lippincott Williams &
Wilkins; 2009.
Early diagnosis of metastatic disease allows for appro- 8. Raman SP, Mahesh M, Blasko RV, Fishman EK. CT scan param-
priate and targeted treatments and can reduce morbid- eters and radiation dose: practical advice for radiologists. J Am Coll
ity and mortality. Current treatment planning is heavily Radiol. 2013;10(11):840–6.
9. Johnson TR. Dual-energy CT: general principles. AJR Am J
dependent on imaging as is follow-up. Recent advances Roentgenol. 2012;1999(Suppl. 5):S3–8.
in diagnostic imaging have progressed at an exponen- 10. Rajiah P, Sundaram M, Subhas N. Dual-energy CT in musculoskel-
tial pace with faster and more accurate imaging. etal imaging: what is the role beyond gout? AJR Am J Roentgenol.
Functional imaging and functional overlay of diagnos- 2019;213(3):493–505.
11. Sujlana PS, Mahesh M, Vedantham S, Harvey SC, Mullen LA,
tic scans (PET-CT and PET-MRI) have revolutionized Woods RW. Digital breast tomosynthesis: image acquisition prin-
the field and continue to grow. Although beyond the ciples and artifacts. Clin Imaging. 2019;55:188–95.
scope of this chapter, the field of radiogenomics is rap- 12. Hoskins PR, Martin K, Thrush A. Diagnostic ultrasound. 2nd ed.
idly advancing and is likely to play an integral role in Cambridge University Press; 2010.
13. Kloth C, Kratzer W, Schmidberger J, Beer M, Clevert DA, Graeter
tumor imaging. T. Ultrasound diagnostics & therapy: on the way to multimodal
Ongoing and future research will strive to answer: ultrasound: contrast-enhanced ultrasound (CEUS), microvas-
• Can radiation dose be decreased by increasing the cular Doppler techniques, fusion imaging, Sonoelastography,
use of modalities utilizing nonionizing radiation Interventional Sonography. In: RöFo-Fortschritte auf dem Gebiet
der Röntgenstrahlen und der bildgebenden Verfahren, vol. 2020.
(such as ultrasound and MRI) particularly in cases Georg Thieme Verlag; 2020. https://doi.org/10.1055/a-­1217-­7400.
where traditionally radiography and CT have been 14. Wilson SR, Greenbaum LD, Goldberg BB. Contrast-enhanced
utilized? ultrasound: what is the evidence and what are the obstacles? Am J
• Plain film radiography is the most commonly per- Roentgenol. 2009;193(1):55–60.
15. Kaur H, Hindman NM, Al-Refaie WB, et al. ACR appropriate-
formed imaging study. Will further refinements in ness criteria® suspected liver metastases. Available at https://www.
computerized radiography further decrease radia- jacr.org/article/S1546-­1440(17)30144-­8/pdf. American College of
tion dose? Radiology.
• Will functional imaging play a greater role in tumor 16. Pomfret EA, Washburn K, Wald C, et al. Report of a national con-
ference on liver allocation in patients with hepatocellular carci-
diagnosis and follow-up? noma in the United States. Liver Transpl. 2010;16:262–78.
• What is the future role of contrast-enhanced ultra- 17. European Society of Radiology (ESR). Abdominal applications of
sound in tumor follow-up? ultrasound fusion imaging technique: liver, kidney, and pancreas.
• What is the future potential of radiogenomics? Insights. Imaging. 2019;10(1):6.
18. Currie S, Hoggard N, Craven IJ, et al. Understanding MRI: basic
MR physics for physicians. Postgrad Med J. 2013;89:209–23.
19. Tsai LL, Grant AK, Mortele KJ, Kung JW, Smith MP. A practi-
cal guide to MR imaging safety: what radiologists need to know.
Radiographics. 2015;35(6):1722–37.
References 20. O'Sullivan GJ, Carty FL, Cronin CG. Imaging of bone metastasis:
an update. World J Radiol. 2015;7(8):202–11.
1. Ritenour ER. Physics overview of screen-film radiography. 21. Yu HH, Tsai YY, Hoffe SE. Overview of diagnosis and management
Radiographics. 1996;16(4):903–16. of metastatic disease to bone. Cancer Control. 2012;19(2):84–91.
2. Bushberg JT. The essential physics of medical imaging. 2nd ed. 22. Cuccurullo V, Cascini GL, Tamburrini O, Rotondo A, Mansi
Philadelphia: Lippincott Williams & Wilkins; 2002. p. 210. L. Bone metastases radiopharmaceuticals: an overview. Curr
3. Bansal GJ. Digital radiography. A comparison with modern con- Radiopharm. 2013;6(1):41–7.
ventional imaging. Postgraduate Med J. 2006;82(969):425428. 23. Heindel W, Gübitz R, Vieth V, Weckesser M, Schober O, Schäfers
4. Ratib O. From PACS to the clouds. Eur J Radiol. 2011;78(2):161–2. M. The diagnostic imaging of bone metastases. Dtsch Arztebl Int.
5. Hunt CH, Hartman RP, Hesley GK. Frequency and severity of 2014;111(44):741–7.
adverse effects of iodinated and gadolinium contrast materials: 24. Vassiliou V, Andreopoulos D, Frangos S, Tselis N, Giannopoulou
retrospective review of 456,930 doses. AJR Am J Roentgenol. E, Lutz S. Bone metastases: assessment of therapeutic response
2009;193:1124–12. through radiological and nuclear medicine imaging modalities.
6. Harieaswar S, Rajesh A, Griffin Y, Tyagi R, Morgan B. Routine Clin Oncol (R Coll Radiol). 2011;23(9):632–45.
use of positive oral contrast material is not required for oncol- 25. Çelebi F. What is the diagnostic performance of 18F-FDG-PET/
ogy patients undergoing follow-up multidetector CT. Radiology. MRI in the detection of bone metastasis in patients with breast can-
2009;250(1):246–53. cer? Eur J Breast Health. 2019;15(4):213–6. Published 2019 Oct 1
 Imaging of Bone Metastases
28
Colleen M. Costelloe, Raul Fernando Valenzuela,
Hubert H. Chuang, and John E. Madewell

Abstract 28.1 Introduction

Bone metastases have become more common due to the
successful prolongation of the lives of many cancer Patients are living longer due to the successful treatment of
patients. Imaging plays an integral role in the staging of their malignancies. This encouraging phenomenon has, nev-
newly diagnosed malignancies and is essential for the ertheless, led to an increase in the incidence of bone metas-
follow-up of these patients throughout their therapeutic tases [1]. Standard imaging modalities such as radiography,
course. The roles of standard imaging such as radiogra- computed tomography (CT), conventional magnetic reso-
phy, CT, conventional MRI; functional imaging such as nance imaging (MRI); functional imaging such as skeletal
skeletal scintigraphy and positron emission tomography scintigraphy/bone scan (BS), and positron-emission tomog-
(PET/CT and PET/MRI); and advanced imaging such as raphy (PET) with fused CT (PET/CT) or MRI (PET/MRI),
whole-body MRI and diffusion-weighted imaging will be and advanced imaging such as whole-body MRI (WBMRI)
discussed. and diffusion-weighted imaging (DWI) play important roles
in the initial and on-going assessment of patients with
malignancies.
Learning Objectives
• Learn how conventional imaging can be used for
the staging of bone metastases. 28.2 Conventional Imaging
• Understand the advantages of indirect and direct
functional imaging. Radiographs are highly specific for the detection and identi-
• Learn that radiopharmaceuticals can be used to treat fication of bone metastases [2], but a minimum of 50% of
bone metastases. bone loss must occur for many lytic metastases to be detect-
• Understand diffusion-weighted imaging with able on radiographs [3] and lesions in the axial skeleton can
respect to bone metastases. be obscured by overlapping structures. When the clinical
question is answered on radiography, it is a highly cost-­
effective method of diagnosis, making it a first-line imaging
modality [4] (Fig. 28.1a).
Radiography is also a cost-effective method of assess-
ing pathologic fracture risk in the long bones and Mirel’s
scoring system [5] can be used to quantify the risk. Points
are given based on four critical factors including the site of
metastatic disease (upper limb = 1, lower limb = 2, peritro-
C. M. Costelloe (*) · R. F. Valenzuela · J. E. Madewell chanteric femur = 3), pain (mild = 1, moderate = 2, aggra-
Department of Musculoskeletal Imaging, The University of Texas
vated by function = 3), lesion density (blastic = 1,
MD Anderson Cancer Center, Houston, TX, USA
e-mail: [email protected]; [email protected]; mixed = 2, lytic = 3), and size with respect to the diameter
[email protected] of the bone (<1/3 = 1, 1/3–2/3 = 2, >2/3 = 3). Based on a
H. H. Chuang study of 61 lesions, Mirel et al. found that a score of 8
Department of Nuclear Medicine, The University of Texas MD indicated a 15% probability of fracture, while a score of 9
Anderson Cancer Center, Houston, TX, USA indicated a 33% probability. A high score, such as 8, can
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 295
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_28
296 C. M. Costelloe et al.

a b c CT is an excellent modality for evaluating mineralized

structures. The cross-sectional nature of CT with multiplanar
reformations (sagittal, coronal, and oblique) makes it more
effective than radiography for assessing the axial skeleton
and irregular bones. MRI provides superior soft-tissue con-
trast resolution in comparison to CT. MRI is commonly used
to delineate the extent of disease inside of the medullary cav-
ity and adjacent soft tissues [4], and as a first-line modality
for assessing commonly involved anatomy in high-risk
patients (spine and pelvis). Conventional pulse sequences for
evaluation of bone metastases include fast spin-echo (FSE)
T1, fat-saturated (FS) T2, and FS FSE T1-weighted images
with intravenous gadolinium contrast [6].
The behavior of bone metastases following therapy can be
assessed using the MD Anderson bone metastases response
criteria (MDA criteria) [7] which provide for quantitative
and nonquantitative evaluation. The nonquantitative portion
of the criteria allows the use of factors other than size to
influence the response evaluation. For example, complete
response (CR) is defined as a return to normal bone density
or complete osteoblastic (sclerotic) fill-in of previously lytic
or mixed lesions on radiography or CT, normalization of
tracer uptake on bone scan (BS), and normalization of signal
intensity on MRI. Partial response (PR) is defined as the
development of a sclerotic rim and/or partial sclerotic fill-in
of lytic metastases on radiography (Fig. 28.2) or CT
(Fig. 28.3), a decrease of 50% in unmeasurable lesions or of
tracer uptake on BS. Progressive disease (PD) is defined as
an increase of 25% or the development of new lesions.
Quantitatively, the perpendicular sum of the greatest
diameters of well-defined intramedullary lesions can be
used. PR requires a 50% or greater decrease on radiography,
CT, or MRI, PD a 25% or greater increase, and stable disease
Fig. 28.1 (a) Anteroposterior (AP) radiograph of the fifth metatarsal (SD) does not meet these requirements.
demonstrates a lytic, permeative lesion (arrows) with pathologic frac- The MDA criteria have been shown to be clinically effica-
ture (arrowhead). It is likely a bone metastasis in this patient with renal
cell carcinoma. Biopsy is recommended to confirm the diagnosis in the
cious [8–10]. A study of 55 patients that assessed the effect
absence of other bone metastases or if infection or other etiologies are of radiation therapy on vertebral bone metastases found that
suspected. No advanced imaging is necessary unless MRI is needed to sclerotic fill-in demonstrated a significant trend with a posi-
further assess the extent of the disease. The lesion and fracture can be tive outcome (reduced pain, p = 0.021), [8]. In a study that
followed with radiography. (b) AP radiograph of the femur in a patient
with metastatic urothelial carcinoma demonstrates a lytic metastasis
compared the World Health Organization and International
with soft-tissue extension (chevrons). According to Mirel’s system, this Union Against Cancer criteria to the MDA criteria in 41
lesion scores 11 (lower limb = 2, pain aggravated by function = 3, breast cancer patients with bone-only metastases found a
lytic = 3, size >2/3 of the diameter of the bone = 3), indicating a very significant correlation to progression-free survival (PFS)
high fracture risk. (c) Treatment with resection arthroplasty
(p = 0.025), [9]. In another study of 29 breast cancer patients
with bone and measurable soft tissue metastases [10], the
be used as an indication for i­ntervention, such as prophy- MDA criteria indicated PFS at 6 months (p = 0.002). The
lactic internal fixation, or resection with reconstruction criteria have also been incorporated into clinical protocols
(Fig. 28.1b). [11, 12]. The MDA criteria can be used as a formal or an
28 Imaging of Bone Metastases 297

a b

Fig. 28.2 (a) Unremarkable pre-therapeutic radiograph of the proxi- 1 year later shows an osteoblastic metastasis, indicating interval healing
mal femur. (b) Simultaneous T1 coronal MRI demonstrates a radio- sclerosis. Without the benefit of the prior MRI, the lesion would appear
graphically occult lung cancer metastasis (arrow). (c) Radiography to represent new disease (PD rather than PR)

a b

Fig. 28.3 (a) Initial CT of faintly sclerotic prostate cancer metastases in bilateral superior pubic rami (arrows). (b) Increased density following
therapy indicates PR (arrowheads). Knowledge of the therapeutic history is essential for proper image interpretation
298 C. M. Costelloe et al.

informal guide to the interpretation of the imaging of bone [19]. However, lesion detection can be radiopharmaceutical
metastases following therapy. specific (Fig. 28.5), so the choice of the imaging agent is
important.
Responses with direct tumor imaging agents can be seen
28.3 Functional Imaging much more rapidly, and flare phenomenon has not been
reported. Thus, direct tumor imaging agents may be ideal for
Functional imaging of bone metastases can be divided detecting and assessing response in bone metastases
into direct and indirect mechanisms. Direct detection relies (Fig. 28.4).
upon direct uptake by tumor, such as with 18F-FDG,
123
I-mIBG, somatostatin analogs (e.g., 68Ga-DOTA-TATE),
18
F-Fluciclovine, and PSMA agents. BS, using 18F-NaF or 28.4 PET/MRI
skeletal scintigraphy, indirectly detect lesions through osteo-
blastic reaction. Each method has consequences for lesion Hybrid imaging with PET/MRI would seem to be the ideal
detection and response assessment. method for identifying osseous metastases, by taking
BS with diphosphonates like 99mTc-MDP is widely avail- ­advantage of the specificity of PET tracers and combining it
able, images the whole body, and is relatively inexpensive. with the anatomic sensitivity of MRI. Catalano et al. reported
However, uptake can be nonspecific and related to tumors, higher detection of osseous lesions in breast cancer patients
inflammation, or trauma. Tomographic imaging with single-­ (96% vs. 85%) [20]. Botsikas et al. found similar sensitivities
photon emission tomography (SPECT), combined with ana- (92% vs. 69%) during the staging of breast cancer, but also
tomic imaging into SPECT/CT scans, increases lesion reported more false-positive PET/MRI findings (7 vs. 0),
contrast and improves localization and characterization, [21]. Beiderwellen et al. also reported better detection of
resulting in higher diagnostic accuracy [13]. benign and metastatic bone lesions in a mix of patients with
18
F-NaF PET/CT has higher spatial resolution and gener- PET/MRI (48/48) compared to PET/CT (45/48 or 94%) [22].
ates whole-body tomographic images, resulting in higher However, not all studies have found greater detection of osse-
lesion detection [14], making this the ideal indirect detection ous lesions by PET/MRI. Guberina et al. reported equivalent
method. However, it is not as widely available, is more expen- detection of 23 bone metastases in prostate cancer patients
sive, and more complex to interpret as compared to BS. [23], and Eiber et al. reported no significant difference in
Bone-seeking agents are better for detecting sclerotic/ detection of bone metastases in a cohort of 119 patients with
osteoblastic metastases as compared to lytic metastases [15]. different cancers [24]. Moreover, Ishii et al. reported the iden-
However, bone metastases can undergo sclerotic change with tification of 34/36 bone lesions by PET/MRI but 35/36 lesions
therapy [7, 16], resulting in paradoxical “scintigraphic flare” on PET/CT [25]. PET/MRI is a relatively new modality, and
on BS (Fig. 28.4), similar to the “sclerotic flare” seen on CT although there appears to be a small improvement in detec-
scans and paralleling clinical flare (with increased bone tion of bone metastases in certain groups, it is unclear if this
markers and bone pain), which can be misinterpreted as pro- improved lesion detection is enough to make significant
gressive disease [17]. Moreover, after therapy, activity with changes in overall patient management.
bone remodeling may take years to normalize [18]. Thus,
indirect detection of bone metastases may result in a signal
that persists after the eradication of tumor. 28.5 Radiopharmaceutical Therapy
Direct detection of tumor is better at identifying lytic or
exclusively marrow-based metastases as compared to scle- Radiopharmaceuticals, Strontium-89, and Samarium-153
rotic lesions. The most experience is with 18F-FDG PET/CT, can be used for the palliative treatment of painful osseous
and the increased detection often results in treatment metastases. About 25–80% of patients report reduced bone
changes, highlighting its utility. A meta-analysis reported pain that lasts about 6 months, with minimal toxicity [26].
higher sensitivity and specificity of 18F-FDG PET for bone Radium-223 also improves survival (14.9 vs. 11.3 months)
metastases (90% and 97%) compared to CT (73% and 95%) in metastatic castration-resistant prostate cancer patients,
or BS (86% and 81%), but similar to MRI (91% and 95%) despite potential lack of PSA response [27]. However,
28 Imaging of Bone Metastases 299

Fig. 28.4 (a) BS shows

a b c
osseous metastases from
breast cancer. (b–d) PET/CT
the same day shows
metastatic disease involving
bones and liver, reported as
progressive. Two months
following new therapy, (e) BS
shows increased uptake,
indicative of progressive
disease versus scintigraphic
flare. (f–h) PET/CT shows d
decreased uptake in hepatic
and osseous metastases, but
little anatomic change
(arrows). After 2 more months
of therapy (i) BS shows
persistently increased uptake,
consistent with scintigraphic
flare because (j–l) PET/CT
shows near-complete
metabolic response and e f g
healing sclerotic change
(arrow)

i j k

l
300 C. M. Costelloe et al.

c­ombining radium-223 with other therapies may result in tion regarding the cellularity of musculoskeletal lesions.
increased complications or toxicity without survival benefit, Tumor foci demonstrate increased signal intensity on DWI
so caution is advised [28]. [29]. Also, the amount of signal loss from restricted free
diffusion across the intact cell membranes of viable malig-
nant cells is quantified as correspondingly low measured
28.6 Advanced Imaging apparent diffusion coefficient (ADC), [30]. The ADC value
increases as cell membranes break down, allowing free
Magnetic resonance imaging (MRI) with functional water diffusion following tumor kill from effective
sequences including diffusion-weighted imaging (DWI) therapy.
increases diagnostic accuracy by permitting the assess- Whole-body DWI is often utilized for tumor detection
ment of metastatic tumor burden, treatment response, and (Fig. 28.6), and localized DWI for characterizing lesions and
can help differentiate metastases from other lesions, when assessing treatment response. The typical spinal marrow
conventional MRI may be inconclusive. DWI is a nonen- ADC value in children and young adults is approximately
hanced functional MRI technique offering useful informa- 0.5 × 10−3 mm2/s. In the elderly population, due to normal

a b

Fig. 28.5 (a–c) 18F-FDG PET/CT negative for active tumor in this patient with prostate cancer. (d) 18F-Fluciclovine PET/CT reveals pelvic nodal
metastasis (arrow) in addition to bone metastases (e–f)
28 Imaging of Bone Metastases 301

d e

Fig. 28.5 (continued)

increased marrow fat content, ADC value drops to approxi- 12 weeks of systemic treatment, the measurement of the
mately 0.4 × 10−3 mm2/s. Most cellular malignant tumors mean ADC from a whole-lesion ROI (region of interest) may
demonstrate ADC values ranging between 0.7 and not be the most appropriate approach, because the changing
1.0 × 10−3 mm2/s [29]. proportions of bone marrow and tumor within the same ROI
ADC can be useful in assessing response to systemic ther- may have opposing effects. Instead, whole-lesion ADC his-
apy. In metastatic prostate carcinoma, resistance to therapy and tograms can accurately demonstrate the proportions of both
shorter time to progression is indicated when a significant per- normal and diseased marrow, helping to discriminate
centage of tumor voxels continues to demonstrate low ADC between responding and progressing patients [32].
(below 1.0 × 10−3 mm2/s) after 1 month of antiandrogenic treat- Most viable malignant tumor components tend to demon-
ment [31]. Findings on conventional imaging can be nonspe- strate a concentration of pixels with ADC values of
cific while ADC increases only after effective therapy (Fig. 28.7). 1.0 × 10−3 mm2/s or less on the left side of the histogram
Normal marrow fat demonstrates low ADC values. During curve. Bharwani et al. demonstrated that patients in whom
treatment, bone metastases can undergo yellow marrow the tumor ADC histogram showed a higher number of pixels
reconversion, leading to post-therapeutic ADC measure- within the lowest 25th percentile values at baseline have a
ments below the pre-therapeutic values. This ­counterintuitive reduced overall survival [33]. Typically, a posttreatment dis-
lowering of ADC values represents a positive treatment placement of the histogram to the right indicates decreased
response [29]. Therefore, when assessing response after cellularity and positive treatment response [29] (Fig. 28.8).
302 C. M. Costelloe et al.

Fig. 28.6 WB-MRI-DWI

a c
demonstrates one right
femoral (long arrows) and
multiple nodal (short arrows)
breast cancer metastases in
the neck and chest on a single
whole-body study. (a)
WB-MRI Dixon fat-only
coronal, (b) WB-DWI b800,
(c) WB-STIR coronal, (d)
WB-DWI b800 coronal MIP,
and (e) DWI b800 axial
images

e
28 Imaging of Bone Metastases 303

a b

c d

e f

g h

Fig. 28.7 Medullary thyroid carcinoma patient with multiple pelvic arrows), and axial ADC maps (black arrows). (a–d) Pretreatment exam-
bone lesions, without size or appearance change during treatment on ination. (e–h) 2 months after the beginning of therapy, and (i–l) at
conventional MRI. A metastasis is demonstrated in the left sacrum. 4 months. ADC values increased from 0.8 × 10−3 mm2/s at baseline (d)
From left to right in three sets of four images: T1 axials (white arrows), to 1.0 × 10−3 mm2/s at 2 months (h) and 1.5 × 10−3 mm2/s at 4 months
FST2 axials (long gray arrows), FST post-contrast axials (short gray (l), indicative of treatment response
304 C. M. Costelloe et al.

i j

k l

Fig. 28.7 (continued)

With respect to radiation therapy, DWI/ADC can assess can- on therapeutic assessment [36] (Fig. 28.9). WB-MRI-DWI
cer cell kill, allowing an objective assessment of radiation effect benefits from improved tumor to background contrast on
[34]. Decreased ADC values, indicative of increased lesion cel- high b-value images, permitting the creation of multiple met-
lularity and progression of disease, are seen only in 1.72% of rics for tumor analysis such as the global apparent diffusion
radiated bone metastases, while the vast majority demonstrate coefficient (gADC) and total diffusion volume (tDV) which
increased postradiation ADC values [35]. Multifractionated reflect tissue cellularity and total tumor burden, respectively.
radiation therapy shows a higher increase in ADC values and a Responding patients show a more considerable increase in
lower retreatment rate compared with single-fraction radiation gADC and nonresponding patients demonstrate an increase
therapy, despite equal pain control rates [34]. in tDV [37].
In thyroid tumors and melanoma bone metastases, the
WB-MRI-DWI shows higher sensitivity than 18F-FDG PET/
28.7 WB-MRI-DWI CT and BS for lesion detection. In lung cancer bone metas-
tases, WB-MR-DWI demonstrates similar sensitivity to 18F-­
Unlike BS and CT, which rely primarily on an osteoblastic FDG PET/CT [38]. For the detection of prostate cancer bone
response for the detection of bone metastases, whole-body metastases, WB-MRI shows a higher specificity but lower
magnetic resonance with diffusion-weighted imaging sensitivity than 18F-NaF PET/CT, as the latter is considered a
(WB-MRI-DWI) provides an evaluation of tumor viability. highly sensitive modality [39]. As WB-MRI-DWI becomes
WB-MRI-DWI plays an increasing role in assessing the more widely available, it may replace BS as the most
response to therapy of tumors with a high prevalence of bone ­common method of evaluating the entire skeleton for metas-
disease, such as prostate and breast cancer, allowing the tases on a single imaging study. BS will likely remain a cost-­
early detection of nonresponders, with a significant impact effective study.
28 Imaging of Bone Metastases 305

Fig. 28.8 The same patient

as Fig. 28.7. Sacral bone
a
lesion (arrows) with interval
increased ADC values from
(a) 0.7 × 10−3 mm2/s at
baseline to (c)
1.5 × 10−3 mm2/s at 4 months,
indicative of positive
treatment response. (b)
Pretreatment histogram
demonstrates most pixels
under 1.0 × 10−3 mm2/s, with
curve centered at
0.9 × 10−3 mm2/s (arrow), (d)
posttreatment histogram
demonstrates a right-sided
curve shift, with most pixels b
above 1.0 × 10−3 mm2/s,
centered at 1.8 × 10−3 mm2/s
(arrow)

d
306 C. M. Costelloe et al.

Fig. 28.9 (a–c) Coronal CT

a d
reformations, displaying
widespread sclerotic prostate
cancer metastases, such as in
the right iliac bone and
acetabulum (arrows). (d)
99m
Tc MDP BS demonstrates
increase radiotracer uptake in
the bilateral pelvis (long
arrows). These findings can
remain unchanged for a
prolonged period despite b
therapeutic response or lack
thereof. (e–g) Coronal pelvic
DWI reformations
demonstrate low signal,
indicating that the disease is
active in the left hemipelvis
and sacrum (small arrows) but
not in portions of the right
hemipelvis despite persistent
MDP uptake on BS c

g
28 Imaging of Bone Metastases 307

7. Hamaoka T, Madewell JE, Podoloff DA, Hortobagyi GN, Ueno

Conclusion NT. Bone imaging in metastatic breast cancer. J Clin Oncol.
2004;22(14):2942–53.
• The MDA criteria can be used as a guide to the 8. Nakata E, Sugihara S, Kataoka M, Yamashita N, Furumatsu T,
interpretation of the conventional imaging of bone Takigawa T, et al. Early response assessment of re-ossification after
metastases following therapy. Future research palliative conventional radiotherapy for vertebral bone metastases.
endeavors can investigate the addition of functional J Orthop Sci. 2019;24(2):332–6.
9. Hamaoka T, Costelloe CM, Madewell JE, Liu P, Berry DA, Islam R,
imaging and advanced imaging to increase the effi- et al. Tumour response interpretation with new tumour response cri-
cacy of the criteria. teria vs the World Health Organisation criteria in patients with bone-­
• Bone scans are indirect tumor imaging agents that only metastatic breast cancer. Br J Cancer. 2010;102(4):651–7.
are more sensitive for sclerotic than lytic metasta- 10. Hayashi N, Costelloe CM, Hamaoka T, Wei C, Niikura N, Theriault
RL, et al. A prospective study of bone tumor response assessment
ses, but may show paradoxical flare after initiation in metastatic breast cancer. Clin Breast Cancer. 2013;13(1):24–30.
of successful therapy. In contrast, direct tumor 11. Tsai YC, Lee HL, Kuo CC, Chen CY, Hsieh KL, Wu MH, et al.
imaging agents are better at identifying lytic or Prognostic and predictive factors for clinical and radiographic
marrow-based metastases and response to therapy, responses in patients with painful bone metastasis treated with mag-
netic resonance-guided focused ultrasound surgery. Int J Hyperth.
but detection is dependent on the imaging agent. 2019;36(1):932–7.
Development of an imaging agent that both accu- 12. Yu T, Choi CW, Kim KS. Treatment outcomes of stereotactic
rately identifies and reflects response in all types of ablative radiation therapy for non-spinal bone metastases: focus
bone metastases would be ideal. on response assessment and treatment indication. Br J Radiol.
2019;92(1099):20181048.
• Radium-223 is an alpha-emitting bone-seeking 13. Donohoe KJ, Cohen EJ, Giammarile F, Grady E, Greenspan BS,
radionuclide that can increase the survival of Henkin RE, et al. Appropriate use criteria for bone scintigraphy
castration-­
resistant prostate cancer patients with in prostate and breast cancer: summary and excerpts. J Nucl Med.
osseous metastases. Its utility in treating osseous 2017;58(4):14N–7N.
14. Even-Sapir E, Metser U, Mishani E, Lievshitz G, Lerman H,
metastases from other cancer types is under Leibovitch I. The detection of bone metastases in patients with
investigation. high-risk prostate cancer: 99mTc-MDP Planar bone scintigraphy,
• While conventional MRI can be inconclusive in single- and multi-field-of-view SPECT, 18F-fluoride PET, and
assessing bone metastases for treatment response, 18F-fluoride PET/CT. J Nucl Med. 2006;47(2):287–97.
15. Cook GJ, Houston S, Rubens R, Maisey MN, Fogelman I. Detection
DWI has the potential to evaluate more reliably the of bone metastases in breast cancer by 18FDG PET: differing meta-
effect of systemic cytotoxic and radiation therapy. bolic activity in osteoblastic and osteolytic lesions. J Clin Oncol.
• Future refinements in post-processing may grant 1998;16(10):3375–9.
WB-MRI-DWI a central role in oncologic imaging 16. Costelloe CM, Chuang HH, Madewell JE. FDG PET for the detec-
tion of bone metastases: sensitivity, specificity and comparison
for total tumor burden and therapy response with other imaging modalities. PET Clin. 2010;5(3):281–95.
assessments. 17. Coleman RE, Mashiter G, Whitaker KB, Moss DW, Rubens RD,
Fogelman I. Bone scan flare predicts successful systemic therapy
for bone metastases. J Nucl Med. 1988;29(8):1354–9.
18. Matin P. The appearance of bone scans following fractures,
including immediate and long-term studies. J Nucl Med.
1979;20(12):1227–31.
References 19. Yang HL, Liu T, Wang XM, Xu Y, Deng SM. Diagnosis of bone
metastases: a meta-analysis comparing (1)(8)FDG PET, CT, MRI
1. Langer C, Hirsh V. Skeletal morbidity in lung cancer patients with and bone scintigraphy. Eur Radiol. 2011;21(12):2604–17.
bone metastases: demonstrating the need for early diagnosis and 20. Catalano OA, Nicolai E, Rosen BR, Luongo A, Catalano M,
treatment with bisphosphonates. Lung Cancer. 2010;67(1):4–11. Iannace C, et al. Comparison of CE-FDG-PET/CT with CE-FDG-­
2. Hortobagyi GN, Libshitz HI, Seabold JE. Osseous metastases of PET/MR in the evaluation of osseous metastases in breast cancer
breast cancer. Clinical, biochemical, radiographic, and scintigraphic patients. Br J Cancer. 2015;112(9):1452–60.
evaluation of response to therapy. Cancer. 1984;53(3):577–82. 21. Botsikas D, Bagetakos I, Picarra M, Da Cunha Afonso Barisits AC,
3. Galasko CS. Skeletal metastases and mammary cancer. Ann R Coll Boudabbous S, Montet X, et al. What is the diagnostic performance
Surg Engl. 1972;50(1):3–28. of 18-FDG-PET/MR compared to PET/CT for the N-and M- stag-
4. Costelloe CM, Rohren EM, Madewell JE, Hamaoka T, Theriault ing of breast cancer? Eur Radiol. 2019;29(4):1787–98.
RL, Yu TK, et al. Imaging bone metastases in breast cancer: 22. Beiderwellen K, Huebner M, Heusch P, Grueneisen J, Ruhlmann
techniques and recommendations for diagnosis. Lancet Oncol. V, Nensa F, et al. Whole-body [(1)(8)F]FDG PET/MRI vs. PET/
2009;10(6):606–14. CT in the assessment of bone lesions in oncological patients: initial
5. Mirels H. Metastatic disease in long bones: a proposed scoring results. Eur Radiol. 2014;24(8):2023–30.
system for diagnosing impending pathologic fractures. Clin Orthop 23. Guberina N, Hetkamp P, Ruebben H, Fendler W, Grueneisen J,
Relat Res. 1989;2003(Suppl. 415):S4–13. Suntharalingam S, et al. Whole-body integrated [(68)Ga]PSMA-­
6. Costelloe CM, Amini B, Madewell JE. Risks and benefits of 11-­PET/MR imaging in patients with recurrent prostate cancer:
gadolinium-­based contrast-enhanced MRI. Semin Ultrasound CT comparison with whole-body PET/CT as the standard of reference.
MR. 2020;41(2):170–82. Mol Imaging Biol. 2020;22(3):788–96.
308 C. M. Costelloe et al.

24. Eiber M, Takei T, Souvatzoglou M, Mayerhoefer ME, Furst S, 32. Messiou C, Collins DJ, Giles S, de Bono JS, Bianchini D, de Souza
Gaertner FC, et al. Performance of whole-body integrated 18F-­ NM. Assessing response in bone metastases in prostate cancer with
FDG PET/MR in comparison to PET/CT for evaluation of malig- diffusion weighted MRI. Eur Radiol. 2011;21(10):2169–77.
nant bone lesions. J Nucl Med. 2014;55(2):191–7. 33. Bharwani N, Miquel ME, Powles T, Dilks P, Shawyer A, Sahdev A,
25. Ishii S, Shimao D, Hara T, Miyajima M, Kikuchi K, Takawa M, et al. Diffusion-weighted and multiphase contrast-enhanced MRI
et al. Comparison of integrated whole-body PET/MR and PET/CT: as surrogate markers of response to neoadjuvant sunitinib in meta-
Is PET/MR alternative to PET/CT in routine clinical oncology? static renal cell carcinoma. Br J Cancer. 2014;110(3):616–24.
Ann Nucl Med. 2016;30(3):225–33. 34. Musio D, De Francesco I, Galdieri A, Marsecano C, Piciocchi A,
26. Pandit-Taskar N, Batraki M, Divgi CR. Radiopharmaceutical ther- Napoli A, et al. Diffusion-weighted magnetic resonance imaging
apy for palliation of bone pain from osseous metastases. J Nucl in painful bone metastases: using quantitative apparent diffusion
Med. 2004;45(8):1358–65. coefficient as an indicator of effectiveness of single fraction versus
27. Parker C, Nilsson S, Heinrich D, Helle SI, O'Sullivan JM, Fossa multiple fraction radiotherapy. Eur J Radiol. 2018;98:1–6.
SD, et al. Alpha emitter radium-223 and survival in metastatic pros- 35. Reginelli A, Silvestro G, Fontanella G, Sangiovanni A, Conte M,
tate cancer. N Engl J Med. 2013;369(3):213–23. Nuzzo I, et al. Validation of DWI in assessment of radiotreated
28. Smith M, Parker C, Saad F, Miller K, Tombal B, Ng QS, et al. bone metastases in elderly patients. Int J Surg. 2016;33(Suppl
Addition of radium-223 to abiraterone acetate and prednisone or 1):S148–53.
prednisolone in patients with castration-resistant prostate cancer and 36. Padhani AR, Makris A, Gall P, Collins DJ, Tunariu N, de Bono
bone metastases (ERA 223): a randomised, double-blind, placebo-­ JS. Therapy monitoring of skeletal metastases with whole-body
controlled, phase 3 trial. Lancet Oncol. 2019;20(3):408–19. diffusion MRI. J Magn Reson Imaging. 2014;39(5):1049–78.
29. Valenzuela RF, Madewell JE, Kundra V. Advanced imaging in mus- 37. Blackledge MD, Collins DJ, Tunariu N, Orton MR, Padhani AR,
culoskeletal oncology: moving away from RECIST and embrac- Leach MO, et al. Assessment of treatment response by total tumor
ing advanced bone and soft tissue tumor imaging (ABASTI)-Part I: volume and global apparent diffusion coefficient using diffusion-­
Novel functional imaging techniques. Semin Ultrasound CT MRI. weighted MRI in patients with metastatic bone disease: a feasibility
2020:1–14. study. PLoS One. 2014;9(4):e91779.
30. Lecouvet FE, Talbot JN, Messiou C, Bourguet P, Liu Y, de Souza 38. Stecco A, Trisoglio A, Soligo E, Berardo S, Sukhovei L, Carriero
NM, et al. Monitoring the response of bone metastases to treatment A. Whole-body MRI with diffusion-weighted imaging in bone
with magnetic resonance imaging and nuclear medicine techniques: metastases: a narrative review. Diagnostics (Basel). 2018;8(3)
a review and position statement by the European Organisation for 39. Mosavi F, Johansson S, Sandberg DT, Turesson I, Sorensen J,
Research and Treatment of Cancer imaging group. Eur J Cancer. Ahlstrom H. Whole-body diffusion-weighted MRI compared
2014;50(15):2519–31. with (18)F-NaF PET/CT for detection of bone metastases in
31. Reischauer C, Koh DM, Froehlich JM, Patzwahl R, Binkert CA, patients with high-risk prostate carcinoma. AJR Am J Roentgenol.
Gutzeit A. Pilot study on the detection of antiandrogen resistance 2012;199(5):1114–20.
using serial diffusion-weighted imaging of bone metastases in pros-
tate cancer. J Magn Reson Imaging. 2016;43(6):1407–16.
 Imaging of Lung and Soft Tissue
Metastases 29
William A. Biché, James F. McLoughlin,
Vanesa Carlota Andreu-Arasa, Stephan W. Anderson,
and Christina A. LeBedis

Abstract imaging has continued to expand, with studies like RECIST

Excepting our surgery and pathology colleagues, most [1] formalizing imaging as a primary method of monitoring
physicians only “see” cancer via medical imaging. This cancer response to therapy. In this chapter, we will review
evaluation is key to the diagnosis and staging of nearly the myriad appearances of metastatic disease in the soft tis-
every cancer, driving the choice of therapeutic approach sues, proceeding by organ.
and prognosis. While radiologists should be available for It should be noted as a warning to the reader that these
consultation, the ability to recognize potential metastatic descriptions are not comprehensive and serve only as a guide
disease on imaging is a valuable skillset for any practitio- to stereotypical and common appearances. For one, the met-
ner, particularly those most involved in oncology. This astatic disease can present in more ways than any book could
chapter focuses on non-osseous metastases, introducing a describe. For two, metastasis should be on the differential for
framework and lexicon aimed at non-radiologists to assist any unusual lesion in a patient with known malignancy.
in accurate diagnosis and appropriate therapy selection. While this chapter should serve as a guide toward familiarity
We cover numerous modalities, including MR, CT, and with the most typical presentations, it should not dissuade
ultrasound, and move organ by organ, describing the practitioners from consulting their colleagues in radiology
imaging features of metastases to each. for further insight into a given lesion.

Learning Objectives 29.1 Intracranial Metastases

• Describe several imaging findings concerning met-
astatic disease in the central nervous system, chest, 29.1.1 Introduction
abdomen, and pelvis.
• Know the most common primary cancers for metas- Approximately 15–20% of patients with cancer will develop
tases found in various organs. intracranial metastases [2]. The true incidence may be even
• Correctly distinguish between normal lymph nodes higher, as screening guidelines for the majority of solid
and those concerning metastasis based on imaging tumors do not recommend routine brain scans. Intracranial
characteristics. metastatic lesions may involve the parenchyma—the cere-
bral hemispheres, cerebellum, and the brain stem—and/or
the meninges.
While pathology remains the gold standard for the diagnosis Although any primary tumor can metastasize to the brain
of malignancy, imaging plays a critical role in screening, parenchyma or the meninges, several large series have iden-
diagnosis, and monitoring. In the latter case, the role of tified lung cancer as universally the most common solid pri-
mary tumor resulting in intracranial metastasis, seen in
William Biche and James McLoughlin are co-first authors, contributed 18–64% of cases, followed by breast cancer (2–21%), mela-
equally to this chapter. noma (4–16%), and colorectal cancer (2–11%). When
included, hematologic cancers cause approximately 10% of
W. A. Biché · J. F. McLoughlin · V. C. Andreu-Arasa
S. W. Anderson · C. A. LeBedis (*) intracranial metastases, primarily to the meninges [3].
Department of Radiology, Boston University Medical Center, Within the brain parenchyma, metastases occur in approx-
Boston University School of Medicine, Boston, MA, USA imately 85% of cases to the cerebral hemispheres, 15% to the
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 309
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_29
310 W. A. Biché et al.

cerebellum, and 3% to the brainstem, in proportion to blood all parenchymal metastases demonstrate avid enhancement
flow to these structures [4, 5]. Less commonly, tumors may on post-contrast imaging. Enhancement patterns may be
metastasize to the choroid plexus, ventricles, or pituitary solid, nodular, or ring-enhancing, although specific enhance-
gland. The most common location for parenchymal metasta- ment pattern offers little insight into the primary tissue of
sis is the gray–white matter junction due to the significant origin. Contrast-enhanced computed tomography (CECT)
change in the size of arterioles at this location. Retroperitoneal may be used to screen for metastases if magnetic resonance
tumors such as colorectal, genitourinary, and uterine cancer imaging (MRI) is contraindicated or unavailable, however is
metastasize disproportionately to the infratentorial region less sensitive than contrast-enhanced MRI [11, 12].
(posterior fossa and spine), which may be due to retrograde Metastases involving the meninges may not have any vis-
dissemination to this location via Batson’s venous plexus [4]. ible signs on NECT, especially when not large enough to
Leptomeningeal carcinomatosis represents metastases, cause mass effect of the brain parenchyma. When large
which grow within the leptomeninges. Metastases may enough to be detectable, they present as extra-axial biconvex
spread to leptomeninges via different mechanisms, including hyperdensities. On CECT, metastases to the meninges may
direct seeding from the brain parenchyma, seeding from the appear as enhancing biconvex masses displacing the brain
dura and skull, endoneural/perineural invasion, and hema- parenchyma. Involvement of the calvarium would be more
togenous seeding. It is an uncommon and late complication suggestive of dural-bases metastasis than leptomeningeal
seen in 5–8% of cases of solid tumors and 5–15% of cases of involvement. MRI is the preferable modality for the assess-
hematological cancers [6]. The incidence of leptomeningeal ment of metastasis involving the meninges.
carcinomatosis varies among different cancer types, but is
diagnosed most frequently in patients with breast or lung 29.1.2.2 Magnetic Resonance (MR)
cancer (10–35%), as well as melanoma (5–25%) [7, 8]. MR with contrast is the most sensitive screening test avail-
Dural metastases are less common than leptomeningeal able for detecting intracranial metastasis. It is also useful to
metastases and are most frequently associated with contigu- further characterize mass lesions found on CT in order to
ous extension from calvarial metastases in approximately refine the differential diagnosis. In fact, CT may underesti-
61% of cases. Isolated dural metastases can also be seen, but mate the number of parenchymal metastatic lesions in up to
are much less common, occurring in 4% of cases. Particular 20% of cases, even when contrast medium is used [13].
cancers that are associated with dural-based metastases On MR, intraparenchymal brain metastases are usually
include breast (34%), prostate (17%), and lung (13%). iso- or hypointense on T1, hyperintense on T2, and exhibit
Mechanisms of spread to the dura are similar to the leptomen- avid contrast enhancement. Commonly there will be sur-
inges and include hematogenous spread, lymphatic spread, or rounding T2/FLAIR hyperintense vasogenic edema
retrograde spread via the vertebral venous plexus [9]. (Fig. 29.1). Melanoma metastases can be T1 hyperintense due
to the free radical content of melanin. Hemorrhagic metasta-
ses can also be T1 hyperintense, depending on the age of
29.1.2 Imaging Findings hemorrhage. Susceptibility-weighted images (SWI) are
highly sensitive to the deposition of hemosiderin and other
29.1.2.1 Computed Tomography (CT) blood breakdown products, and the presence of blooming
Non-enhanced computed tomography (NECT) may be the artifact on these sequences aids in differentiating melanin
first imaging modality a patient with brain metastasis under- from hemorrhage. Low T2 signal is seen in some metastatic
goes, either in detecting a previously unknown lesion or with neoplasms and may offer a clue to tissue of origin. Mucin-
the development of new neurologic findings. NECT alone is secreting metastases, which can be seen in GI and GU adeno-
not sensitive enough to accurately characterize cerebral carcinomas, and at times lung, may demonstrate low T2
metastases, but findings can suggest the presence of a lesion signal. Restricted diffusion (high signal on diffusion-­weighted
or signs of mass effect. imaging with reduced apparent diffusion coefficient maps)
Brain parenchymal metastases on CT appear as solitary or may also occasionally be seen in mucinous-producing lesions
multiple well-circumscribed round mass lesions, which are and metastasis from other highly cellular primary neoplasms
present a majority of the time at the gray–white matter inter- such as breast, colon, testicular, and small and non-small cell
face. They are typically small and surrounded by digitiform lung carcinoma [5, 14].
hypodensity representing vasogenic edema. Metastases may Nearly all parenchymal metastases demonstrate enhance-
be hypodense, isodense, or hyperdense to the brain paren- ment on contrast-enhanced MR due to lack of the normal
chyma. Approximately 20% of metastases are hemorrhagic, blood-brain barrier architecture. Like CT, enhancement pat-
and a hemorrhagic parenchymal neoplasm is more likely to terns may demonstrate solid, nodular, or ring-like enhance-
be metastatic than primary. Melanoma metastases tend to be ment, however provide little insight into the primary site of
hyperdense even in the absence of hemorrhage [10]. Almost tumor. Although the vast majority of contrast-enhancing lesions
29 Imaging of Lung and Soft Tissue Metastases 311

a b c d

e f g

Fig. 29.1 (a–d) A 63-year-old male with melanoma. (a and b) Axial stage IV small cell lung carcinoma. (e) Axial postcontrast T1-WI dem-
pre and postcontrast T1-WI demonstrate cortical thickening and hyper- onstrates multiple small nodular foci of enhancement consistent with
intensity on pre-contrast images (arrows). (c) Axial SWI shows foci of metastasis (arrow and arrowhead). (f) Axial SWI show susceptibility
susceptibility artifact (arrowheads) from hemosiderin deposition due to artifact (arrowhead) in some of the lesion representing blood products.
hemorrhage. (d) Axial FLAIR shows significant FLAIR hyperintensity (g) Coronal post-contrast T1-WI again shows multiple scattered meta-
representing vasogenic edema (arrows). (e–g) A 67-year-old male with static lesions (arrows)

in a patient with a primary tumor are metastases, the differen- ents as parenchymal hyperintensity on FLAIR sequences.
tial diagnosis includes primary brain tumor, abscess, infarction, Dural metastases enhance avidly, and like meningiomas, an
or radiation necrosis in a previously treated patient [5]. enhancing dural tail may be present in nearly half of cases.
MR is more sensitive for detecting intracranial metastasis There may be calvarial involvement of an enhancing lesion,
to the leptomeninges compared to CT and has a specificity of suggestive of contiguous extension into the skull [9, 16].
approximately 77–100% [6]. Leptomeningeal metastases
present as either smooth or nodular thickening of the lepto-
meninges. Lesions are typically T1 iso- or hypointense rela- 29.2 Extra-Osseous Spine Metastases
tive to the gray matter, and T2/FLAIR sequences demonstrate
hyperintensity following the sulci and CSF spaces. 29.2.1 Introduction
Leptomeninges demonstrate avid enhancement, which is
usually diffuse but may include areas of focal nodularity. Metastases to the spine can involve the bone, epidural space,
Common locations to be affected by leptomeningeal metas- leptomeninges, and/or spinal cord. Extradural lesions
tasis include the basilar cisterns and cerebellar folia [14, 15]. account for up to 95% of spinal lesions, and can be divided
Dural metastases appear as thickening of the dura, which into pure epidural lesions, and those originating from the
may be smooth or nodular, and following the contour of the vertebrae and subsequently impinging on the thecal sac.
calvarium. Like leptomeningeal metastases, lesions may Intradural extramedullary and intramedullary metastases are
appear T1 iso- or hypointense and variable on T2-weighted less common than extradural metastases, accounting for
images. They may cause mass effect on the underlying 5–6% and 0.5–1% of spinal metastases, respectively [17].
parenchyma with associated vasogenic edema, which pres- Routes of tumor dissemination to the spinal cord and menin-
312 W. A. Biché et al.

ges include hematogenous spread via Batson’s venous on T1-weighted images. On post-contrast imaging, lesions
plexus, arterial spread, perineural and perivascular migration typically demonstrate avid enhancement. In cases where there
from systemic tumors, contiguous extension from the verte- is diffuse enhancement coating the spinal cord, it is referred as
brae, and spread through the CSF as drop metastases. “sugar coating” due to the appearance of extensive bright
The vast majority of epidural metastases originate in enhancement along the spinal cord, brainstem, and sulci [17].
either the vertebrae (85%) or the paravertebral tissues (10– Intradural intramedullary metastases are located deep to
15%) with extension into the epidural space, with the remain- the pia mater, typically within the substance of the cord
der originating in the epidural space itself [18]. As tumors itself. These are typically small, well-circumscribed spheri-
grow in the epidural space, they can encroach on the thecal cal, or ovoid lesions with surrounding edema seen as hyper-
sac and compresses the spinal cord, causing neurological intensity on T2-weighted images that may be out of
disturbance. proportion to the size of the lesion. Post-contrast images
In adults, intradural extramedullary metastases of the demonstrate avid enhancement of the lesion within the spinal
spine have an incidence as high as 4–15% in patients with cord. Intramedullary metastases can be hemorrhagic, and in
solitary tumors not part of the CNS [19]. They are most com- these cases, demonstrate similar MR findings to hemorrhagic
monly found in primary tumors of the breast, lung, and mela- brain lesions. A thin line of increased enhancement sur-
noma; melanoma and lung cancer have the highest rates of rounding the margins of the lesion (“rim sign”) with ill-­
metastasis to this location at 20% and 11%, respectively [20]. defined triangular extension along the cranial and caudal
Intramedullary spinal metastases are a rare but devastating aspect of the lesion (“flame sign”) on sagittal post-contrast
manifestation of systemic malignancy. Lung cancer is the images is considered highly specific for intramedullary
most common primary sources for intramedullary metastasis metastasis [16, 17, 21].
(51.4%), followed by breast cancer (18.6%), renal cell carci-
noma (8.6%), and melanoma (3.9%) [21]. The presence of an
intramedullary spinal metastatic lesion suggests a biologi- 29.3 Lung Metastases
cally aggressive form of cancer. The mechanism of metastatic
spread from the primary tumor to the spinal cord is thought to 29.3.1 Introduction
be hematogenous via an arterial route in the majority of cases.
Lung metastases are seen in 20–54% of extrathoracic malig-
nancies. The lungs are the second most frequent site of
29.2.2 Imaging Findings metastasis, after the liver. Approximately 20% of metastatic
disease is isolated to just the lungs. For these reasons, assess-
Although CT can be useful for assessing the osseous struc- ment of lung imaging is of paramount importance in the
tures of the spine, it is inadequate for the assessment of pos- management and staging of cancer patients [22, 23].
sible meningeal and intramedullary metastatic disease. MR A retrospective study conducted in 288 patients with con-
is used for this purpose and can aid in the diagnosis of metas- firmed metastatic lung nodules revealed the primary sites of
tasis, characterization of levels of involvement, and the diag- tumor were as follows: colorectal 25.8%, head and neck 19.4%,
nosis of potential associated cord compression. urologic (kidneys, ureter, bladder, prostate, testes) 14.7%, gas-
Extradural metastases are located external to the dura trointestinal non-colorectal 10.9%, breast 10.5%, melanoma
mater as the epidural space is extradural. Typically, epidural 6.5%, gynecologic (ovarian, endometrial, cervical) 6.1%, and
metastases appear as a region of iso- or hypointense T1 sig- other sites (e.g., sarcoma, thyroid) in 6.1% of cases [24].
nal, increased T2 signal, and contrast enhancement within Malignancies can reach the lung parenchyma by hematog-
the epidural space. Involvement may be focal or multilevel enous spread through the pulmonary or bronchial arteries,
depending on the extent of disease. Increased T2 signal lymphatic spread, or direct invasion. The most common path
intensity in the spinal cord in this setting is suggestive com- by far is hematogenous spread—metastases by this route tend
pressive edema or infarction from vascular compression. to be located in the more highly vascularized basal and periph-
Intradural extramedullary metastases are located inside the eral regions of the lungs. In extensive metastatic disease, the
pachymeninges but outside the spinal cord. Most lesions are distribution becomes more random and multifocal [25].
located within the subarachnoid space and involve the lepto-
meninges. Lesions demonstrate enhancing smooth or nodular
thickening along the surface of the spinal cord and/or the cord 29.3.2 Imaging Findings
roots, which can be focal or diffuse. Involvement can be iso-
lated to just the spine or can involve both the intracranial struc- 29.3.2.1 Radiography
tures and the spine (Fig. 29.2). Metastatic lesions may be seen Chest radiographs are known to have low sensitivity com-
as small T2 isointense nodules, but are generally poorly seen pared to computed tomography (CT) imaging in the detection
29 Imaging of Lung and Soft Tissue Metastases 313

a b c

d e

Fig. 29.2 A 31-year-old female with breast carcinoma. (a and b) Axial contrast administration shows extensive enhancement coating the spi-
post-contrast T1-WI shows a large enhancing mass in the right temporal nal cord (arrowhead) and cauda equina nerve roots (arrowheads), which
fossa (a) that was thought to be extra-axial, surgically removed, and is also demonstrated on axial post-contrast T1-WI at the lower thoracic
demonstrated to be a parenchymal lesion (arrow). In addition, there is and lumbar spine levels (d and e, respectively). Also, lung nodules are
diffuse enhancement along the bilateral internal auditory canals noted (d) consistent with lung metastasis (circles)
(arrows) representing leptomeningeal disease. (c) Sagittal T1-WI after

of metastatic nodules. Radiographs fail to depict metastatic metastases. Typical findings include multiple nodules of dif-
lesions smaller than 7 mm and are limited by overlapping ferent sizes, diffuse thickening of the interstitium (lymphan-
structures and relatively low contrast. In addition, radio- gitic carcinomatosis), and involvement of the pleura. Atypical
graphs are known to underestimate metastases compared to patterns of lung metastasis can confound the differential and
CT; a solitary lesion seen with radiograph may be associated include diffuse miliary seeding, cavitation, and calcified
with multiple nodules on CT [26, 27]. metastases [28] (Fig. 29.3).
The most common CT pattern of pulmonary metastases is
29.3.2.2 Computed Tomography (CT) the presence of multiple solid round pulmonary nodules.
CT is the modality of choice for the detection and follow-up Nodules have soft-tissue attenuation and can be smooth or
of pulmonary metastases because of its high spatial resolu- irregular. When nodules are numerous, they are distributed
tion. No contrast is required for the detection of pulmonary diffusely throughout the lungs without any anatomic
314 W. A. Biché et al.

a b c

d e f

Fig. 29.3 (a, b) 82-year-old female with thyroid cancer. PA chest which also involve the lateral aspect of the mediastinum and encase part
radiograph demonstrates opacity of the superior mediastinum causing of the aortic arch. Such findings are most suspicious for pleural metas-
right deviation of the trachea (arrow). There are multifocal nodular-like tases. (e) 68-year-old female with uterine sarcoma. Axial CT image of
opacities within the bilateral lungs. A follow-up axial CT of the chest the chest with contrast demonstrates multiple smooth-walled nodules
with contrast demonstrates innumerable solid pulmonary nodules in a throughout the lungs. In addition, there is a cavitary lesion with thick,
random distribution. The majority of the nodules are smooth but few irregular walls (arrow). (f) 60-year-old female with stage IV adenocar-
have irregular margins, highly suspicious for metastatic disease. (c, d) cinoma of the left lung. Axial CT images of the chest with contrast
54-year-old female with history of metastatic melanoma. PA chest demonstrate diffuse bilateral interlobular septal thickening with nodu-
radiograph demonstrates multiple nodular opacities, which occupy the larity and irregularity along the interlobular septa, consistent with lym-
majority of the left hemithorax, favoring the periphery. Axial CT images phangitic carcinomatosis. Loss of lung volume and pleural effusion is
of the chest with contrast demonstrate multiple pleural-based masses, noted in the left lung

distribution; when nodules are few or solitary, they are pre- most cases, there is the preservation of the underlying lung
dominantly peripheral and within the lung bases. It is also architecture [30].
not unusual for nodules to be of different sizes due to multi- While the classic imaging appearances of lung metastases
ple different episodes of tumor embolization [25, 29]. are extremely helpful in narrowing the differential down to
Lymphangitic carcinomatosis refers to the spread of a metastasis, especially with a known primary, atypical fea-
neoplasm through the lymphatics. It is most commonly seen tures can make it difficult to distinguish lung metastases
in adenocarcinomas, particularly primary tumors of the lung, from other lung pathologies.
stomach, breast, pancreas, uterus, rectum, and prostate. The Miliary nodules refer to innumerable small randomly distrib-
mechanisms of metastatic involvement of the lymphatics uted nodular opacities approximately <5 mm in size that resem-
include antegrade lymphatic invasion through the diaphragm bles millet seeds. Miliary metastases can be seen in cancers of
or pleural surfaces or retrograde spread from hilar metasta- the thyroid, kidney, breast, and pancreas, as well as more rarely
ses. The typical appearance is characterized by interlobular in malignant melanoma and osteosarcoma. The differential for
septal thickening, with areas of nodularity or irregularity miliary patterns on chest imaging includes tuberculosis, fungal
along the interlobular septa extending toward the hilum. In infections, sarcoidosis, pneumoconiosis [31].
29 Imaging of Lung and Soft Tissue Metastases 315

Cavitation is seen in approximately 5% of lung metasta- false positives caused by infectious or inflammatory etiolo-
sis. Tumor necrosis of the central portion of the lesion and gies also demonstrate avidity on FDG-PET. False-­negative
subsequent discharge of the necrotic material is the mecha- FDG-PET pulmonary nodules are usually caused by lesions
nism which causes the cavitating appearance. Cavitating smaller than 1 cm. False negatives can also occur due to
lesions generally feature thick and irregular walls. Squamous malignant pulmonary lesions which are not FDG avid, which
cell carcinomas are the most common primary tumor to include renal and testicular cancers. For these reasons, a neg-
cause cavitation, although adenocarcinomas (GI tract, ative FDG-PET scan in the presence of suspicious findings on
breast), transitional cell tumors, and sarcomas are also chest CT cannot reliably exclude metastatic disease [36].
known to cavitate. In addition, primary lung cancers can also
cavitate, which confounds the differential when there is a
solitary cavitating pulmonary nodule present instead of mul- 29.4 Abdomen Metastases
tiple lesions [32].
Calcification of metastatic nodules is often seen only on 29.4.1 Liver, Biliary Tree, and Pancreas
CT scans. Calcification is seen in metastases from osteosar-
coma, chondrosarcoma, giant cell tumor of the bone, and The liver is one of the most common sites of metastatic dis-
mucinous adenocarcinoma. Reasons for calcification vary ease in the body, second only to lymph nodes [37]. Almost
but can include bone formation in primary bone tumors, dys- any tumor is capable of finding its way to the hepatic paren-
trophic calcification, and mucoid calcifications [28]. chyma, either via the organ’s prodigious blood supply or via
In addition to tumor disseminating to the lung paren- its complex lymphatic network. The most common primary
chyma, the pleura can also be involved. Pleural metastases malignancy is, of course, adenocarcinoma of the colon, but
originate from hematogenous spread to the pleura, but can most GI malignancies are likely to metastasize to the liver.
also be caused by lymphangitic spread, or contiguous spread This is facilitated by the portal venous system and thus is less
of tumor to the chest wall or mediastinum from established commonly seen in cirrhotic patients where blood is shunted
metastases. Pleural metastases are present on imaging as away from the liver.
nodularities along the pleura. There may or may not be an Metastatic lesions in the liver are diverse in appearance.
associated malignant pleural effusion. Typically pleural Most classically, they appear as well-defined hypoattenuat-
metastases are visible on non-contrast CT as irregularities of ing lesions, which demonstrate mild peripheral enhancement
the pleura, which do not typically require contrast, however on portal venous imaging [38] (Fig. 29.4a, h). This appear-
without contrast may be obscured by effusions or other air- ance is most associated with hypovascular colon cancer
space disease [33]. metastases. Hypervascular metastases generally enhance
avidly on arterial phase imaging (Fig. 29.4e), and may even
29.3.2.3 Magnetic Resonance Imaging (MR) be undetectable on portal venous imaging. Neuroendocrine
MR has lower sensitivity than CT in the detection of meta- tumors, choriocarcinomas, and melanomas are common, but
static pulmonary nodules and for that reason is not routinely thyroid, RCC, and a small percentage of breast cancers may
used in the evaluation of lung metastases. Metastatic nodules also develop hypervascular liver metastases. Liver metasta-
have low or intermediate intensities on T1-weighted images ses may also be cystic or necrotic, with nonenhancing, low,
and higher intensity on T2-weighted images. T1 weighted, central attenuation. This appearance is more commonly seen
T2 weighted, and STIR sequences can help distinguish meta- in lung cancer, melanoma, mucinous colon cancer, or occa-
static lesions from other lung lesions which may confound sionally carcinoid.
the diagnosis such as tuberculoma, bronchocele, mucin-­ On MRI, metastatic lesions are most often T1 hypoin-
containing tumors, hamartoma, and aspergilloma [34]. tense and T2 hyperintense [39] (Fig. 29.4g), although this
can flip in metastases with significant hemorrhage or mela-
29.3.2.4 Nuclear Imaging nin. Metastatic lesions also tend to restrict diffusion. One
Fluorodeoxyglucose positron emission tomography (FDG-­ particular advantage to MR imaging of the liver is the advent
PET) is the primary scintigraphic modality used in the evalu- of the use of contrast agents with hepatobiliary secretion,
ation of pulmonary metastases. FDG-PET is important in adding another tool in the characterization of hepatic lesions.
differentiating between benign and malignant nodules, espe- As one would predict, metastatic lesions in the liver do not
cially in the setting of a solitary pulmonary nodule. It can aid contain functional hepatocytes, and thus do not enhance with
in raising suspicion of malignancy without proceeding to these agents.
more invasive diagnostic tests such as biopsy. Hepatic metastases most typically have a targetoid
FDG-PET has been shown to have a sensitivity of 96% appearance on ultrasound (Fig. 29.4f), with a hyperechoic or
and specificity of 88% for diagnosing a pulmonary nodule as isoechoic center and a hypoechoic periphery [40]. This
malignant [35]. The positive predictive value is lower because appearance is most typical in colon cancer, but HCC and
316 W. A. Biché et al.

a b c d

e f g h

Fig. 29.4 Hepatic metastases. (a and b) A 51-year-old male with pan- 63-year-old female with small bowel neuroendocrine tumor. Axial CT
creatic cancer metastatic to the liver. (a) Axial CT image with IV con- with IV contrast in the arterial phase shows an enhancing hypervascular
trast showing multiple rim enhancing hypoattenuating masses in the metastasis (arrow). (f) A 71-year-old male with mucinous colon adeno-
liver (arrows). (b) Ultrasound of hypoechoic metastasis in right hepatic carcinoma. US of the liver shows a hyperechoic lesion with a hypoechoic
lobe (arrow) (c) 76-year-old female with colon cancer. Axial CT image halo, a classic “target” metastasis. (g and h) A 65-year-old man with
with IV contrast shows a large hypoattenuating mass in the liver with rectal cancer. Right hepatic lobe metastasis manifesting as a hyperin-
psammomatous calcifications (d) 57-year-old female with breast cancer tensity on T2 fat-suppressed imaging (arrow, g). This lesion demon-
metastatic to liver. Axial CT image with IV contrast with liver pseudo- strates continuous peripheral contrast enhancement which was
cirrhosis, caused by innumerable small treated metastases. (e) A persistent on T1-delayed phase imaging (arrow, h)

lymphoma can also present in this manner. Ultrasound can pancreas, it has typically seeded other organs first. The most
also appreciate the hypervascularity and cystic nature of cer- common primaries are melanoma, RCC, lung carcinoma,
tain cancers (as above). Another feature, appreciable on and breast, and less commonly, colon [42]. These lesions are
ultrasound, is calcification. While this can be seen in the set- typically round, well marginated, and heterogeneously
ting of chronic granulomatous changes, metastases from enhancing on cross-sectional imaging (Fig. 29.7a, b). While
mucinous tumors of the GI tract, breast, and ovaries may also multiple lesions are always suggestive of metastases, solitary
show calcifications, which may also be appreciated on CT lesions are more common. On MRI, these metastases are
and MRI (Fig. 29.4c). Lastly, in patients with diffuse small typically T1 hypointense and T2 hyperintense (Fig. 29.7c)
metastases throughout the liver, the parenchyma may appear compared to healthy pancreatic parenchyma and they appear
diffusely heterogeneous on ultrasound; this is notably seen in solid and hypoechoic on ultrasound.
some metastatic breast cancers. Once these are treated, they
may involute, causing nodularity of the liver known as pseu-
docirrhosis (Fig. 29.4d). 29.4.2 Spleen, Kidneys, and Ureters
Metastases to the gallbladder and biliary tree are quite
rare, although are occasionally seen in aggressive melanoma, Despite its considerable blood flow, the spleen is an uncom-
gastric, or breast cancers. The route is typically hematoge- mon site for metastatic disease. Only 2–7% of cadavers with
nous. These lesions generally enhance, and demonstrate advanced metastatic disease have splenic lesions, and the
either infiltrative or polypoid morphology, sometimes with metastatic disease usually presents elsewhere first [43]. Most
internal foci of hemorrhage [41]. These lesions are commonly they are from melanoma (>50%), breast, lung,
­hyperechoic on ultrasound with minimal acoustic shadowing ovary, gastric, or prostate cancer. The lesions are classically
and may show internal vascularity. ill-defined, soft tissue attenuating masses with peripheral
Pancreatic metastases are uncommon. Cadaver studies enhancement (Fig. 29.7d). They may be cystic, particularly
report approximately 3–12% of patients with advanced in melanoma. MRI shows T1 hypointensity and T2 hyperin-
malignancy have metastatic disease to the pancreas, and tensity, with variable enhancement [44]. Of note, implanta-
metastases are outnumbered by primary lesions 30:1. tions to the serosa from intraperitoneal seeding (most
Unsurprisingly, by the time metastatic disease reaches the commonly from gastric, ovarian, or pancreatic cancers) may
29 Imaging of Lung and Soft Tissue Metastases 317

a b

Fig. 29.5 Adrenal metastases. A 60-year-old female with lung cancer metastatic to left adrenal gland. (a) Axial CT with IV contrast showing large
enhancing left adrenal nodule (arrow). (b) Nodule demonstrates hypermetabolic activity (arrow) on PET

cause scalloping (Fig. 29.6a) as the masses tug at the splenic phology, and may bleed or calcify. Smaller lesions, in con-
capsule. trast, tend to be more homogenous and well defined, making
Metastases to the kidneys are relatively common, being them harder to differentiate from lipid-poor benign adeno-
seen in ~7% of all patients with extrarenal cancers on mas. This diagnostic difficulty is compounded by several
autopsy. Despite this, they tend to be involved later in the similar imaging characteristics including diffusion restric-
course of disease relative to organs like the lungs or liver. tion on MRI and enhancement. Metastases do washout more
The primary cancers are most commonly lung, breast, or GI slowly compared to benign lesions, with absolute washout
adenocarcinomas. When seen on imaging, they are typically <60% and relative washout <40% [47]. Both lipid-poor ade-
multiple and bilateral, with low attenuation on CT. Metastases nomas and metastases are also usually T1 hypointense and
tend to be solid and infiltrative, with cystic appearance being T2 hyperintense [48], although the latter moderately favors
uncommon. Relative to the kidney parenchyma, they tend to metastases. In uncertain scenarios, the best imaging differen-
be lower attenuation, with variable enhancement. tiator may simply be growth over time.

29.4.3 Adrenal Glands 29.4.4 Gastrointestinal Tract

The adrenal glands are a very common location for metastatic The GI tract is most commonly involved in metastatic dis-
disease, representing the fifth most common site in the body ease through the peritoneal spread of malignancy, from ovar-
(after the liver, lungs, bones, and lymph nodes). They are most ian, gastric, or colonic primaries. In this setting, small, soft
frequently involved in the lung (especially bronchogenic car- tissue attenuation lesions may be identified in the serosa,
cinoma), breast, colon adenocarcinoma, and melanoma, but which frequently causes adhesions to nearby loops of the
RCC, thyroid, and lymphoma are also common [45]. bowel, predisposing to obstruction or encasement. They are
Adrenal metastases are usually bilateral and mostly atten- very rarely solitary and frequently innumerable.
uate higher than the normal adrenal parenchyma on CT Less commonly, it may be seeded hematogenously, most
(Fig. 29.5a, b), with lesions >43 Hounsfield units being very commonly by melanoma, breast, or lung cancers. In con-
likely metastatic [46]. Exceptions are RCC and rare HCC trast to peritoneal spread, these lesions tend to be submuco-
metastases, which may be hypoattenuating. When large sal or intramural, with contrast enhancement. In these
(>3 cm), this attenuation may be more heterogeneous, while locations, they may cause intussusception, in addition to
the lesion itself may have a poorly defined, lobulated mor- obstructions.
318 W. A. Biché et al.

a b

Fig. 29.6 Peritoneal/omental metastases. (a) A 38-year-old female and spleen (arrowheads), consistent with serosal deposits. (b) A
with metastatic mucinous appendiceal carcinoma. Axial CT with IV 27-year-old male with colon cancer metastatic to peritoneum. Axial CT
contrast through the upper abdomen shows mucinous metastases with IV contrast shows omental thickening and soft tissue deposits
throughout the peritoneum consistent with pseudomyxoma peritonei. (arrows) consistent with “omental caking”
Note the “scalloping” of the stomach (hollow arrow), liver (arrows),

a b e

c d

Fig. 29.7 Pancreatic and ovarian metastases. (a–d) A 54-year-old cancer. Axial CT with IV contrast and numerous hypoattenuating cystic
male with renal cell carcinoma metastatic to the pancreatic head. The metastases in the spleen (arrows). Image credit Ashley Davidoff, MD.
metastasis demonstrated arterial enhancement on axial CT (a, arrow) (e) 51-year-old female with gastric cancer metastatic to ovaries
and T1 sequence MRI (b, arrow). Note internal hypoattenuation sug- (Krukenberg tumor). Coronal CT with IV contrast shows large hypoat-
gestive of necrosis. This lesion was hyperintense on T2 fat-suppressed tenuating polycystic masses in the bilateral ovaries with mild ascites
MR imaging (c, arrow). (d) A 65-year-old female with small cell lung
29 Imaging of Lung and Soft Tissue Metastases 319

29.4.5 Peritoneum 29.6 Lymph Node Metastases

Peritoneal metastases are soft tissue typically small, soft-­ Irregularities of the regional lymph nodes are frequently the
tissue attenuating, and demonstrate contrast enhancement. first radiologic sign of metastatic disease, particularly in car-
Similar to most metastatic lesions, they show T1 hypoin- cinomas and melanomas [56]. Tumors nearly always spread
tensity and T2 hyperintensity on MRI. They are often from their most proximal, or “sentinel,” node more centrally
accompanied by ascites, and where they involve the through the lymphatic tree [57]. While the involvement of
omentum they may be mass-like, or broad and flat any specific set of nodes is thus widely variable based on the
(Fig. 29.6b). site and lymphatic drainage of the primary tumor, numerous
suspicious features remain common between them. It also
should be noted that while imaging is a necessary step in
29.5 Pelvic Metastases evaluating lymph nodes, conventional imaging such as CT
and MR remain moderately limited in both sensitivity and
The bladder is infrequently involved in metastatic disease, specificity for metastatic involvement, and that here, as
and, where present, it tends to be direct invasion from other always, tissue diagnosis remains king.
primary pelvic malignancies like prostate or ovarian cancers The most commonly noted finding for lymphadenopathy
[49]. These cases tend to be less equivocal, with irregular is size (Fig. 29.8a) usually reported as maximum short-axis
enhancing masses infiltrating into and distorting normal axial diameter. Specific size cutoffs are both controversial
bladder wall contours. Rarely, hematogenous spread may and vary by lymph node site, ranging from 11 mm for sub-
bring gastric, breast, or lung cancers to intramural locations carinal nodes to 5 mm in the lateral retropharyngeal group
within the bladder wall, where they form small, contrast-­ [58]. Unfortunately, limits of imaging dictate that any
enhancing, soft tissue–attenuating lesions [50]. Bladder improvement in sensitivity will sacrifice specificity and vice
involvement is typically late, with metastatic disease pre- versa. That said, a commonly used rule of thumb is 10 mm or
senting elsewhere first. Metastatic nodules may be indistin- larger, at which point most nodes should be looked at with
guishable from primary bladder cancer on imaging, requiring suspicion.
tissue sampling for diagnosis. Notably, the bladder may be Location of irregular lymph nodes should also be consid-
involved in lymphoma, where imaging may identify a solid ered when gauging the likelihood of their involvement in
solitary intramural mass. metastatic disease. As tumor spread through lymphatics is
Metastatic disease to the female reproductive tract favors circumscribed by the route those lymphatics travel, the like-
the ovaries, with uterine metastases being extremely rare. lihood of remote lymphatic involvement with normal proxi-
Ovarian metastases are often bilateral on presentation, with mal nodes is very unlikely [59]. Further, diffuse
T1/T2 hypointensity on MRI, and usually are at least par- lymphadenopathy is an uncommon sign of metastatic (non-­
tially cystic [51] (Fig. 29.7e). Primaries include GI tract lymphoma) cancer. In these instances, alternative etiologies
tumors, lungs, and endometrium [52]. On ultrasound, they for lymphadenopathy should be considered.
are heterogeneous, generally with hypervascularity of their Lastly, and perhaps most importantly, is consideration of
solid components. Metastases to the ovary may demonstrate the internal nodal architecture. Typically, lymph nodes are
a desmoplastic reaction, with fibrous connective tissue bands soft tissue attenuation on CT and are isointense to muscle on
pulling on the surrounding tissues. They also commonly T1-weighted MR sequences, with isointensity to mild hyper-
present with ascites. intensity on T2 [60]. Normal lymph nodes demonstrate a
The male reproductive tract is also a rare target of meta- reniform morphology with a fatty hilum. As metastatic dis-
static disease. Case reports of metastases from colon, renal, ease distorts physiologic nodal architecture, this hilum may
and bladder cancers exist in the literature, but what few be lost, a sign of metastatic involvement. Lymph nodes do
case series exist point toward the prostate as the most com- show mild homogeneous enhancement physiologically, but
mon primary, representing approximately 15% of metasta- heterogenous or avid enhancement (usually corresponding to
ses [53]. In these rare cases, ultrasound examination tended that of the primary tumor) may be a sign of metastasis.
to show heterogeneous testicles with multiple distinct Lymph node necrosis, demonstrated by heterogeneous low
masses and occasional calcifications [54]. The prostate is attenuation and increased central T2 signal (Fig. 29.8c), or
likewise a rare target, with occasional metastases from heterogenous calcification (Fig. 29.8b), should also raise
colon cancer [55]. In this setting, case reports have suspicion [61]. It should be noted, however, that necrosis
described a hypoattenuating mass with minimal contrast may be seen as a sign of tumor response to therapy, and gran-
enhancement. ulomatous infection may cause calcification.
320 W. A. Biché et al.

a b

Fig. 29.8 Lymph node metastases. (a) A 76-year-old female with neously calcified mediastinal lymph node. (c) A 63-year-old male with
colon cancer. Axial CT images with IV contrast demonstrating bulky rectal cancer metastatic to retroperitoneal lymph nodes. Axial CT
retroperitoneal lymph nodes. Note rounded shape and loss of fatty hila. images with IV contrast show an enlarged, irregularly shaped lymph
(b) 67-year-old male with metastatic recurrence of thyroid cancer. node with internal hypoattenuation consistent with necrosis​​
Axial CT images with IV contrast demonstrate enlarged heteroge-

Unfortunately, the reality is that these single features are • Imaging remains the primary metric to gauge response to
rarely both sensitive and specific for metastatic involvement therapy.
per se, and so acceptable diagnostic accuracy is only achieved • As novel immunologics and “theranostics” enter clinical
by considering them together. practice, both traditional imaging and novel imaging
modalities (including specialized PETs) will require fur-
ther development to monitor treatment response.
29.7 Conclusion

• Knowledge of typical metastatic patterns of specific can- Disclosure The authors have no relevant financial disclosures related
to this topic.
cers can aid in their detection.
• CT, US MRI, and PET play complementary roles with
respect to lesion detection, characterization, and targeting
for tissue sampling). References
• Radiology will continue to provide critical diagnostic and 1. Schwartz LH, Litière S, de Vries E, et al. RECIST 1.1-Update and
prognostic indicators for the foreseeable future as the pre- clarification: from the RECIST committee. Eur J Cancer Oxf Engl.
mier noninvasive tool. 1990;2016(62):132–7. https://doi.org/10.1016/j.ejca.2016.03.081.
29 Imaging of Lung and Soft Tissue Metastases 321

2. Johnson JD, Young B. Demographics of brain metastasis. Neurosurg 21. Goyal A, Yolcu Y, Kerezoudis P, Alvi MA, Krauss WE, Bydon
Clin N Am. 1996;7(3):337–44. M. Intramedullary spinal cord metastases: an institutional review
3. Posner JB, Chernik NL. Intracranial metastases from systemic can- of survival and outcomes. J Neuro-Oncol. 2019;142(2):347–54.
cer. Adv Neurol. 1978;19:579–92. https://doi.org/10.1007/s11060-­019-­03105-­2.
4. Delattre JY, Krol G, Thaler HT, Posner JB. Distribution of 22. Mohammed T-LH, Chowdhry A, Reddy GP, et al. ACR
brain metastases. Arch Neurol. 1988;45(7):741–4. https://doi. Appropriateness criteria® screening for pulmonary metastases.
org/10.1001/archneur.1988.00520310047016. J Thorac Imaging. 2011;26(1):W1–3. https://doi.org/10.1097/
5. Lignelli A, Khandji AG. Review of imaging techniques in the RTI.0b013e3182010bf9.
diagnosis and management of brain metastases. Neurosurg 23. Stella GM, Kolling S, Benvenuti S, Bortolotto C. Lung-seeking
Clin N Am. 2011;22(1):15–25. https://doi.org/10.1016/j. metastases. Cancers. 2019;11:7. https://doi.org/10.3390/
nec.2010.09.002. cancers11071010.
6. Batool A, Kasi A. Leptomeningeal carcinomatosis. In: StatPearls. 24. Caparica R, Mak MP, Rocha CH, et al. Pulmonary nodules in patients
StatPearls Publishing; 2020. Accessed December 17, 2020. http:// with nonpulmonary cancer: not always metastases. J Glob Oncol.
www.ncbi.nlm.nih.gov/books/NBK499862/. 2016;2(3):138–44. https://doi.org/10.1200/JGO.2015.002089.
7. Clarke JL, Perez HR, Jacks LM, Panageas KS, Deangelis 25. Hirakata K, Nakata H, Nakagawa T. CT of pulmonary metasta-
LM. Leptomeningeal metastases in the MRI era. ses with pathological correlation. Semin Ultrasound CT MR.
Neurology. 2010;74(18):1449–54. https://doi.org/10.1212/ 1995;16(5):379–94. https://doi.org/10.1016/0887-­2171(95)
WNL.0b013e3181dc1a69. 90027-­6.
8. El Shafie RA, Böhm K, Weber D, et al. Outcome and prognostic 26. Girvin F, Ko JP. Pulmonary nodules: detection, assessment, and
factors following palliative craniospinal irradiation for leptomenin- CAD. AJR Am J Roentgenol. 2008;191(4):1057–69. https://doi.
geal carcinomatosis. Cancer Manag Res. 2019;11:789–801. https:// org/10.2214/AJR.07.3472.
doi.org/10.2147/CMAR.S182154. 27. Ketai L, Malby M, Jordan K, Meholic A, Locken J. Small nodules
9. Lyndon D, Lansley JA, Evanson J, Krishnan AS. Dural masses: detected on chest radiography: does size predict calcification? Chest.
meningiomas and their mimics. Insights Imaging. 2019;10 https:// 2000;118(3):610–4. https://doi.org/10.1378/chest.118.3.610.
doi.org/10.1186/s13244-­019-­0697-­7. 28. Seo JB, Im JG, Goo JM, Chung MJ, Kim MY. Atypical pulmo-
10. McGann GM, Platts A. Computed tomography of cranial met- nary metastases: spectrum of radiologic findings. Radiogr Rev Publ
astatic malignant melanoma: features, early detection and Radiol Soc N Am Inc. 2001;21(2):403–17. https://doi.org/10.1148/
unusual cases. Br J Radiol. 1991;64(760):310–3. https://doi. radiographics.21.2.g01mr17403.
org/10.1259/0007-­1285-­64-­760-­310. 29. Gross BH, Glazer GM, Bookstein FL. Multiple pulmonary nod-
11. Schellinger PD, Meinck HM, Thron A. Diagnostic accuracy of MRI ules detected by computed tomography: diagnostic implica-
compared to CCT in patients with brain metastases. J Neuro-Oncol. tions. J Comput Assist Tomogr. 1985;9(5):880–5. https://doi.
1999;44(3):275–81. https://doi.org/10.1023/a:1006308808769. org/10.1097/00004728-­198509000-­00007.
12. Yokoi K, Kamiya N, Matsuguma H, et al. Detection of brain metas- 30. Bruce DM, Heys SD, Eremin O. Lymphangitis carcinomatosa: a
tasis in potentially operable non-small cell lung cancer: a com- literature review. J R Coll Surg Edinb. 1996;41(1):7–13.
parison of CT and MRI. Chest. 1999;115(3):714–9. https://doi. 31. Pillai S, Khan A, Khan S. Adenocarcinoma of the lung presenting
org/10.1378/chest.115.3.714. with intrapulmonary miliary metastasis. Cureus. 2019;11(8):e5430.
13. Hanssens P, Karlsson B, Yeo TT, Chou N, Beute G. Detection of https://doi.org/10.7759/cureus.5430.
brain micrometastases by high-resolution stereotactic magnetic res- 32. Chaudhuri MR. Cavitary pulmonary metastases. Thorax.
onance imaging and its impact on the timing of and risk for distant 1970;25(3):375–81. https://doi.org/10.1136/thx.25.3.375.
recurrences. J Neurosurg. 2011;115(3):499–504. https://doi.org/10. 33. Purandare NC, Rangarajan V. Imaging of lung cancer: impli-
3171/2011.4.JNS101832. cations on staging and management. Indian J Radiol Imaging.
14. Fink KR, Fink JR. Imaging of brain metastases. Surg Neurol 2015;25(2):109–20. https://doi.org/10.4103/0971-­3026.155831.
Int. 2013;4(Suppl. 4):S209–19. https://doi.org/10.4103/2152-­ 34. Biederer J, Hintze C, Fabel M. MRI of pulmonary nod-
7806.111298. ules: technique and diagnostic value. Cancer Imaging Off
15. Wang N, Bertalan MS, Brastianos PK. Leptomeningeal metastasis Publ Int Cancer Imaging Soc. 2008;8:125–30. https://doi.
from systemic cancer: review and update on management. Cancer. org/10.1102/1470-­7330.2008.0018.
2018;124(1):21–35. https://doi.org/10.1002/cncr.30911. 35. Patz EF, Lowe VJ, Hoffman JM, et al. Focal pulmonary abnor-
16. Rykken JB, Diehn FE, Hunt CH, et al. Rim and flame signs: post- malities: evaluation with F-18 fluorodeoxyglucose PET scan-
gadolinium MRI findings specific for non-CNS intramedullary spi- ning. Radiology. 1993;188(2):487–90. https://doi.org/10.1148/
nal cord metastases. AJNR Am J Neuroradiol. 2013;34(4):908–15. radiology.188.2.8327702.
https://doi.org/10.3174/ajnr.A3292. 36. Roh J-L, Ryu CH, Kim JS, et al. Clinical significance of intratho-
17. Shah LM, Salzman KL. Imaging of spinal metastatic disease. Int racic lesions detected by 18F-fluorodeoxyglucose positron emis-
J Surg Oncol. 2011;2011:769753. https://doi.org/10.1155/2011/ sion tomography in the management of patients with head and neck
769753. cancer. Oral Oncol. 2007;43(8):757–63. https://doi.org/10.1016/j.
18. Yáñez ML, Miller JJ, Batchelor TT. Diagnosis and treatment of oraloncology.2006.09.004.
epidural metastases. Cancer. 2017;123(7):1106–14. https://doi. 37. Roth CG, Deshmukh S. Fundamentals of body MRI.; 2017. Accessed
org/10.1002/cncr.30521. December 13, 2020. http://search.ebscohost.com/login.aspx?direct
19. Kaplan JG, DeSouza TG, Farkash A, et al. Leptomeningeal =true&scope=site&db=nlebk&db=nlabk&AN=1288236
metastases: comparison of clinical features and laboratory data 38. Sica GT, Ji H, Ros PR. CT and MR imaging of hepatic metas-
of solid tumors, lymphomas and leukemias. J Neuro-Oncol. tases. AJR Am J Roentgenol. 2000;174(3):691–8. https://doi.
1990;9(3):225–9. org/10.2214/ajr.174.3.1740691.
20. Amer MH, Al-Sarraf M, Baker LH, Vaitkevicius VK. Malignant 39. Namasivayam S, Martin DR, Saini S. Imaging of liver metastases:
melanoma and central nervous system metastases: incidence, diag- MRI. Cancer Imaging Off Publ Int Cancer Imaging Soc. 2007;7:2–
nosis, treatment and survival. Cancer. 1978;42(2):660–8. https:// 9. https://doi.org/10.1102/1470-­7330.2007.0002.
doi.org/10.1002/1097-­0142(197808)42:2<660::aid-­cncr28204202 40. Wernecke K, Vassallo P, Bick U, Diederich S, Peters PE. The
37>3.0.co;2-­e. distinction between benign and malignant liver tumors on sonog-
322 W. A. Biché et al.

raphy: value of a hypoechoic halo. AJR Am J Roentgenol. 52. Lee S-J, Bae J-H, Lee A-W, Tong S-Y, Park Y-G, Park
1992;159(5):1005–9. https://doi.org/10.2214/ajr.159.5.1329454. J-S. Clinical characteristics of metastatic tumors to the ovaries.
41. Choi WS, Kim SH, Lee ES, et al. CT findings of gallbladder J Korean Med Sci. 2009;24(1):114–9. https://doi.org/10.3346/
metastases: emphasis on differences according to primary tumors. jkms.2009.24.1.114.
Korean J Radiol. 2014;15(3):334–45. https://doi.org/10.3348/ 53. Turk A, Graff JM, Memo M. Case report of metastatic pros-
kjr.2014.15.3.334. tate cancer to testicles: an ominous sign of advanced disease.
42. Klein KA, Stephens DH, Welch TJ. CT characteristics of Urol Case Rep. 2019;26:100935. https://doi.org/10.1016/j.
metastatic disease of the pancreas. Radiogr Rev Publ Radiol eucr.2019.100935.
Soc N Am Inc. 1998;18(2):369–78. https://doi.org/10.1148/ 54. Ulbright TM, Young RH. Metastatic carcinoma to the testis: a
radiographics.18.2.9536484. clinicopathologic analysis of 26 nonincidental cases with empha-
43. Compérat E, Bardier-Dupas A, Camparo P, Capron F, Charlotte sis on deceptive features. Am J Surg Pathol. 2008;32(11):1683–93.
F. Splenic metastases: clinicopathologic presentation, differential https://doi.org/10.1097/PAS.0b013e3181788516.
diagnosis, and pathogenesis.Arch Pathol Lab Med. 2007;131(6):965– 55. Youssef FR, Hunt L, Meiring PD, Taraporewalla DR, Gupta R,
9. https://doi.org/10.1043/1543-­2165(2007)131[965:SMCPDD]2. James MJ. Metastasis of a cecal adenocarcinoma to the prostate
0.CO;2. five years after a right hemicolectomy: a case report. J Med Case
44. Elsayes KM, Narra VR, Mukundan G, Lewis JS, Menias CO, Rep. 2011;5:223. https://doi.org/10.1186/1752-­1947-­5-­223.
Heiken JP. MR imaging of the spleen: spectrum of abnormali- 56. Wong SY, Hynes RO. Lymphatic or hematogenous dissemination:
ties. Radiogr Rev Publ Radiol Soc N Am Inc. 2005;25(4):967–82. how does a metastatic tumor cell decide? Cell Cycle Georget Tex.
https://doi.org/10.1148/rg.254045154. 2006;5(8):812–7. https://doi.org/10.4161/cc.5.8.2646.
45. Lam K-Y, Lo C-Y. Metastatic tumours of the adrenal glands: 57. Nathanson SD. Insights into the mechanisms of lymph node
a 30-year experience in a teaching hospital. Clin Endocrinol. metastasis. Cancer. 2003;98(2):413–23. https://doi.org/10.1002/
2002;56(1):95–101. https://doi.org/10.1046/j.0300-­0664.2001. cncr.11464.
01435.x. 58. Lan M, Huang Y, Chen C-Y, et al. Prognostic value of cervical
46. Richard WW. Fundamentals of body Ct. 5th ed. Elsevier; 2019. nodal necrosis in nasopharyngeal carcinoma: analysis of 1800
47. Johnson PT, Horton KM, Fishman EK. Adrenal mass imaging patients with positive cervical nodal metastasis at MR imag-
with multidetector CT: pathologic conditions, pearls, and pitfalls. ing. Radiology. 2015;276(2):536–44. https://doi.org/10.1148/
Radiogr Rev Publ Radiol Soc N Am Inc. 2009;29(5):1333–51. radiol.15141251.
https://doi.org/10.1148/rg.295095027. 59. Paño B, Sebastià C, Buñesch L, et al. Pathways of lymphatic
48. Elsayes KM, Mukundan G, Narra VR, et al. Adrenal masses: MR spread in male urogenital pelvic malignancies. Radiogr Rev Publ
imaging features with pathologic correlation. Radiogr Rev Publ Radiol Soc N Am Inc. 2011;31(1):135–60. https://doi.org/10.1148/
Radiol Soc N Am Inc. 2004;24(Suppl. 1):S73–86. https://doi. rg.311105072.
org/10.1148/rg.24si045514. 60. Koh DM, Hughes M, Husband JE. Cross-sectional imag-
49. Modh RA, Corbyons KA, Yeung LL. A rare metastasis to ing of nodal metastases in the abdomen and pelvis. Abdom
the bladder. Case Rep Urol. 2013;2013:789039. https://doi. Imaging. 2006;31(6):632–43. https://doi.org/10.1007/
org/10.1155/2013/789039. s00261-­006-­9022-­2.
50. Karjol U, Jonnada P, Cherukuru S, Chandranath A. Bladder 61. Yang WT, Lam WW, Yu MY, Cheung TH, Metreweli
metastasis from breast cancer: a systematic review. Cureus. C. Comparison of dynamic helical CT and dynamic MR imaging
2020;12(3):e7408. https://doi.org/10.7759/cureus.7408. in the evaluation of pelvic lymph nodes in cervical carcinoma. AJR
51. Yada-Hashimoto N, Yamamoto T, Kamiura S, et al. Metastatic ovar- Am J Roentgenol. 2000;175(3):759–66. https://doi.org/10.2214/
ian tumors: a review of 64 cases. Gynecol Oncol. 2003;89(2):314– ajr.175.3.1750759.
7. https://doi.org/10.1016/s0090-­8258(03)00075-­1.
 Image-Guided Biopsy and Intervention
for Metastatic Disease 30
Joao R. T. Vicentini, Sina Habibollahi, Ambrose J. Huang,
and Connie Y. Chang

Abstract
and (3) place the correct orders so adequate testing
The field of image-guided musculoskeletal intervention is performed.
includes a wide variety of diagnostic and therapeutic pro- • Most image-guided ablation techniques involve
cedures in the management of metastatic disease. This thermal energy, and the main applications for osse-
chapter provides an overview of percutaneous musculo- ous metastases are pain palliation and treatment of
skeletal biopsies, including information on preprocedural oligometastatic disease. The choice of technique
planning, indications, technique, strategies to maximize will depend mostly on location of the lesion and
diagnostic yield, and potential risks and complications. In resource availability.
recent years, new image-guided treatment options have
been developed for the management of metastases-related
pain and oligometastatic disease. These techniques are
also reviewed, with a focus on radiofrequency ablation
and cryoablation, and additional information on micro- 30.1 Introduction
wave ablation, laser ablation, and MR imaging-guided
extracorporeal focused ultrasound. Image-guided percutaneous biopsy has become the first-line
option for pathologic confirmation of the metastatic disease
in the musculoskeletal system. When compared to surgical
biopsy, image-guided procedures are less invasive, safer, and
Learning Objectives lower cost, while maintaining similar diagnostic yield [1].
• Preprocedural planning is the most important step Image-guided intervention can also be used for therapeutic
of every image-guided musculoskeletal interven- purposes, and percutaneous ablation has been expanding as
tion and should include a review of the indications, an alternative option for treatment and pain palliation in the
a plan for the safest approach, and anticipation of setting of metastatic disease. This chapter offers an overview
potential issues, with the help of a multidisciplinary of the main image-guided diagnostic and therapeutic proce-
team whenever possible. dures currently available for musculoskeletal metastases.
• Biopsy technique, including the choice of imaging
modality, will vary depending on the type and loca-
tion of the lesion, but the main goals are similar for 30.2 Percutaneous Musculoskeletal Biopsy
each case: (1) secure a safe access to the target; (2)
obtain a satisfactory number of viable specimens; 30.2.1 Preprocedural Planning

30.2.1.1 Imaging Review and Biopsy Indication

The first step when planning a biopsy is reviewing all the
available imaging to confirm that the procedure is necessary.
Many incidental lesions found during staging turn out to be
J. R. T. Vicentini · S. Habibollahi · A. J. Huang · C. Y. Chang (*) benign or non-related to the primary malignancy. The
Division of Musculoskeletal Imaging and Intervention, Department ­radiologist can make that determination when classic benign
of Radiology, Massachusetts General Hospital, Boston, MA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 323
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_30
324 J. R. T. Vicentini et al.

features are present, or when long-term stability can be both methods. Benign lesions such as lipomas and ganglion
ensured by comparison with older studies. cysts can be easily identified due to their typical imaging fea-
A common source of uncertainty during imaging workup tures. Hibernomas are a distinct type of adipocytic lesion
is the differentiation of osteoblastic metastases and enostoses composed primarily of brown fat, which can be found in
or bone islands. Enostoses tend to be smaller (1 mm–2 cm) typical locations for brown fat deposits, such as the neck,
and usually remain stable in size over time, although slow axilla, shoulder, chest wall, retroperitoneum, and thighs.
growth has been described [2, 3]. Irregular or spiculated mar- Hibernomas have characteristics similar to lipomas but tend
gins are a common feature, and some can show scintigraphic to be more heterogeneous, with low-signal serpentine or
activity on 99 m-Tc-methylene diphosphate (MDP) bone branching vascular structures, and can also show increased
scan [3]. Since enostoses are composed of compact cortical 18
F-FDG avidity on PET [8].
bone, these lesions are usually very dense, and that charac- Radiographs can be helpful during workup of soft tissue
teristic can be used to differentiate them from metastatic masses. For example, hemangiomas and other types of vas-
lesions. On CT, threshold values of 885 HU for mean attenu- cular lesions can show radiopaque phleboliths [9].
ation and 1060 HU for maximal attenuation have been asso- Radiographs can also show the typical peripheral ossification
ciated with 95% sensitivity and 96% specificity for pattern associated with myositis ossificans, which may
differentiating benign enostoses from untreated osteoblastic mimic an aggressive lesion on MRI. Although CT scans are
metastases [2]. For borderline values or variable appearance, not typically part of the diagnostic workup of soft tissue
correlation with other imaging modalities, such as MRI or masses, CT can help with pre-procedure planning.
bone scan, may be helpful. Degenerative changes may also When the imaging review of an osseous or soft tissue
present with sclerosis and increased uptake on radionuclide lesion cannot definitively rule out malignancy, tissue sam-
bone scans, but the typical pattern of distribution makes pling is indicated. Then, the next step is to determine the best
them less likely to be mistaken for a neoplastic process. way to obtain an adequate sample.
Infection is another mimicker of bone malignancy
(Fig. 30.1). Osteomyelitis can present with an indolent course 30.2.1.2 Planning the Approach
and cause progressive bone destructive changes and periosteal The goal of preprocedural planning is to determine the safest
reaction, making it difficult to rule out an underlying marrow approach to the target and increase the chances of obtaining
replacing lesion [4]. In many cases, clinical data may point a diagnostic sample. This process includes the choice of
toward the correct diagnosis; however, classic signs of infec- imaging modality for guidance, which is mainly influenced
tion may not be present in chronic osteomyelitis, and sampling by the type of targeted lesion (osseous vs. soft tissue) and its
of the affected area may be necessary for confirmation. location. If there are multiple lesions, which is the case for
Incidental lucent lesions need to be differentiated from many metastatic tumors, the choice of a biopsy target usually
osteolytic metastases. The broad differential diagnosis of involves a combined evaluation of safety profile, technical
osseous lucent lesions is beyond the scope of this chapter, feasibility, and size of the lesion. For example, pelvic or long
but as a general rule, nonaggressive imaging features suggest bones are usually preferred over vertebrae given safer access
benignity. These features include well-defined margins, through the pelvic soft tissues. Larger lesions offer the pos-
­narrow transition zone with the normal bone, presence of sibility of obtaining more samples but may also contain areas
intralesional fat, and preservation of adjacent trabecular and of hemorrhage or necrosis.
cortical bone [3]. Common incidental benign lucent lesions In general, the areas with more aggressive features are
include intraosseous lipomas and ganglion cysts. selected for biopsy to avoid underestimating tumor grade
­Non-­ossifying fibromas are benign lesions that can show [10]. Some lesions will demonstrate non-enhancing areas,
hypermetabolism on 18F-fluorodeoxyglucose positron emis- which can reflect nonviable, necrotic tissue; therefore, the
sion tomography (18F-FDG PET) [5]. MRI can be helpful to enhancing solid components are usually chosen for sampling
help identify the different components of each lesion. to increase the likelihood of obtaining diagnostic tissue
Metastases to subcutaneous tissues and skeletal muscle (Fig. 30.2) [1, 11, 12]. Although necrotic tissue does not con-
are uncommon, with reported rates below 2% among patients tribute to the final histologic diagnosis, documenting the
investigated for a soft tissue mass [6]. Most authors believe presence of necrosis may be helpful given its association
this is likely related to variable blood flow and decreased with worse prognosis in some types of sarcoma [13, 14].
capillary density when compared to other common sites of Pathologic fractures should be avoided whenever possible.
metastatic disease such as the liver and lungs. The most com- Sampling the fractured area makes histopathologic evalua-
mon primary tumor associated with soft tissue metastasis is tion more challenging given the presence of granulation tis-
lung cancer [6, 7]. Imaging workup of soft tissue masses can sue and immature osteoid as part of the healing process,
be performed with ultrasound, MRI, or a combination of which can mimic bone-forming tumors [10].
30 Image-Guided Biopsy and Intervention for Metastatic Disease 325

a b

c d

Fig. 30.1 A 33-year-old male with history of HCV admitted after low-­ concern for associated matrix-forming lesion (arrowhead). (c) Sampling
grade trauma. (a) Axial and (b) sagittal CT images of the lumbar spine of disc material and (d) bone was performed and sent for surgical
show fractures with fragmentation of the vertebral endplates at L3–L4 pathology and microbiology. Results were consistent with discitis–
(arrows). Mineralization at the right anterolateral aspect of L3 raised osteomyelitis with cultures positive for Pseudomonas aeruginosa
326 J. R. T. Vicentini et al.

a b

c d

Fig. 30.2 A 65-year-old male presented with chronic left hip pain. (a) distinguished through ultrasound. (c, d) Patient returned for a CT-guided
Axial T2-weighted MR image shows a large mass centered within the biopsy with the use of intravenous iodinated contrast. (c) Contrast-­
adductor musculature in the left proximal thigh (arrow). (b) Ultrasound-­ enhanced CT image showed subtle enhancement along the peripheral
guided core biopsy was performed (black arrowhead) and frozen portion of the mass (white arrowheads). (d) This area was targeted dur-
pathology showed predominantly nondiagnostic necrotic tissue. ing the biopsy, and pathology was consistent with undifferentiated pleo-
Additional samples were obtained; however, these were also nondiag- morphic sarcoma
nostic due to extensive necrosis. Necrotic and viable tissue could not be

Another important factor when considering the approach controversial topic in the literature, most studies suggest that
in musculoskeletal biopsies is the risk of seeding the biopsy the risk of needle tract seeding is likely below 1% [15–17].
track, particularly in primary sarcomas. Although this is a For metastatic disease, this is a less important issue, since
30 Image-Guided Biopsy and Intervention for Metastatic Disease 327

treatment will rarely involve surgical intervention, so lesions 30.2.2 Technique

with the highest yield potential and safest approach are pri-
oritized. It is still good practice to communicate with the 30.2.2.1 Ultrasound Guidance
referring provider to discuss any potential options and pref- When the target is mostly composed of soft tissue in a super-
erences for each patient, especially for solitary lesions that ficial location, ultrasound is the preferred modality for image
may turn out to be a synchronous primary malignancy. As a guidance. Real-time visualization allows the operator to find
general rule, when performing a percutaneous needle biopsy, the most promising area based on appearance and size and
one should avoid traversing uninvolved compartments, dif- also helps to avoid damage to adjacent structures, including
ferent joint spaces, and neurovascular bundles [1, 10, 11, blood vessels, which can be evaluated with color Doppler,
18]. Another strategy to minimize the risk of needle tract and tendons and nerves. Differentiation between solid and
seeding is the use of coaxial systems, further discussed in the cystic components can also be helpful, since in most cases
technique section. the solid component is targeted for sampling [11]. Superficial,
small lesions are also better assessed under ultrasound when
30.2.1.3 Orders compared to CT. Another advantage is the lack of ionizing
The final step in biopsy planning is placing adequate lab radiation, which is especially important for younger patients.
orders and anticipating the need for specific tests or analyses Limitations of ultrasound guidance include limited access to
besides standard surgical pathology. If a lesion shows a cys- deep lesions, poor evaluation of osseous lesions, and limited
tic component from which fluid can be aspirated, the obtained correlation with other cross-sectional studies, such as MRI
material can also be sent to cytology. When infection is and PET-CT.
among the differential diagnoses, additional samples should
be obtained and sent for microbiology analysis. Other spe- 30.2.2.2 CT Guidance
cific tests or requests may be included when they can influ- CT is the modality of choice for purely intraosseous lesions
ence the treatment plan, such as genetic testing and dosage of or those located within deep planes. Advantages of CT
hormone receptors. Research biopsies are also common for include the larger field of view and the possibility to access
patients with metastatic disease and can be used to guide difficult locations. Disadvantages include the use of ionizing
treatment and to determine eligibility for a specific therapy radiation and technical difficulties for small, superficial
or clinical trial. lesions [11]. Gantry size may also influence the final posi-
tioning and choice of imaging modality. The total length of
30.2.1.4 Plan for Sedation the external portions of the biopsy tools should be added to
Most musculoskeletal biopsies can be performed with local the diameter of the selected body region and then compared
anesthesia. This is a common choice for soft tissue lesions with the CT gantry size [10]. This is especially important for
and lytic osseous lesions with cortical breakthrough. In some obese patients and procedures in the thoracoabdominal
cases, however, conscious sedation can be helpful, especially region.
for purely intraosseous lesions, when there is surrounding
inflamed tissue, for lesions of difficult access or in proximity 30.2.2.3 Other Imaging Modalities
to vital structures, or if the patient experiences extreme anxi- Fluoroscopy can be used for biopsy guidance, and common
ety. General anesthesia is rarely needed for most musculo- uses include bone marrow aspiration and biopsy of large
skeletal procedures and is reserved for special cases, such as osseous lesions in the pelvis or appendicular skeleton.
for pediatric patients or image-guided ablation. Advantages include higher flexibility in terms of position-
ing, potentially shorter procedure times, and less radiation
30.2.1.5 Multidisciplinary Approach exposure when compared to CT [1]. MRI and PET-CT guid-
Multidisciplinary review is key whenever planning any kind ance are potential options for CT occult bone lesions; how-
of intervention for musculoskeletal tumors, primary or meta- ever, these modalities are usually associated with higher
static. Some factors that can be discussed ahead of time cost and longer procedure times and may not be readily
include clinical findings, need for further imaging before the available in many institutions [19, 20]. A more common use
procedure, biopsy approach, plan for frozen pathology, of MRI and PET-CT is to help establish landmarks for sam-
potential complications, specific tests to be ordered, and pling lesions that are not well seen on CT. Although this
treatment plans. Multidisciplinary teams should include all approach may be difficult for smaller findings, biopsy of CT
the specialties involved in patient care, including orthopedic occult lesions using anatomical landmarks has been shown
surgeons or surgical oncologists, radiologists, medical to result in similar diagnostic yield when compared to CT
oncologists, pathologists, and radiation oncologists. evident lesions [1, 21].
328 J. R. T. Vicentini et al.

30.2.2.4 Coaxial Technique a plateau at four samples for soft tissue lesions and three for
For many osseous and deep soft tissue lesions, it is helpful to bone lesions [29].
secure direct access to the biopsy target. This can be achieved Fine needle aspiration (FNA) alone has lower accuracy
with coaxial technique, which involves the use of an intro- when compared to core needle biopsy and may underesti-
ducer device that is positioned next to the area to be biopsied. mate tumor grading due to lack of information on tissue
The biopsy needle can then be advanced through the intro- architecture and appearance of adjacent structures [23, 27,
ducer to obtain the samples without creating new biopsy 30]. However, it can be used as an adjunct during a core nee-
tracks. This practice avoids the need for multiple needle dle biopsy, especially in the investigation of a potentially
passes from the skin through the soft tissues into the lesion, metastatic lesion, when the visualization of cells with malig-
minimizing the risk of injury to adjacent structures [1]. nant features is usually enough to confirm the diagnosis.
Larger diameter introducers can accommodate different
types of biopsy needles. They can also be used for aspiration
and placement of probes for ablation therapy. 30.2.4 Complications

30.2.2.5 Biopsy Device The complication rate of percutaneous needle biopsies is

The choice of which biopsy device to use is usually influ- considered low, and most studies report complication rates
enced by the composition of the targeted lesion (bone or soft below 10%, and some as low as 1.1%, versus up to 16% for
tissue), location, operator preference, and resource availabil- open surgical biopsies [9, 11, 18]. Medication-related com-
ity. Sampling of lesions arising in the soft tissues or of the plications include allergic reaction to local anesthetic or
extraosseous soft tissue component of bone lesions can be medications used for conscious sedation [11]. Minor techni-
done with a spring-loaded side-cutting biopsy needle. For cal complications include increased pain, swelling, and
osseous lesions with an intact cortex, a manual or battery-­ bruising. Major complications include bleeding, infection,
powered drill can be used to insert an introducer or cannula and damage to adjacent structures. In spine biopsies, there is
through the cortex. Then, a biopsy needle can be inserted a potential risk of hematoma formation and compression of
coaxially through the introducer to obtain the samples. Fine nerve roots or the spinal cord. For biopsies of the thoracic
needle aspiration (FNA) can also be performed through the spine, ribs, or sternum, pneumothorax is also a potential
introducer. FNA can be especially helpful for lesions in chal- complication. Given the low incidence of hemorrhagic com-
lenging locations, such as in the craniofacial and cervical plications, in some scenarios, anticoagulation does not need
regions. to be suspended. For patients on warfarin, international nor-
malized ratio of 1.5 or less is considered safe, and platelet
count equal or above 50.000/μL is usually recommended [1].
30.2.3 Accuracy

Reported accuracy of image-guided biopsy of musculoskel- 30.3  herapeutic Options: Percutaneous

T
etal lesions ranges from 74% to 97% [11, 22–27]. Even when Ablation
a biopsy is not necessarily accurate in terms of identifying
the histopathologic features and tumor grade, the procedure Since the first report describing the use of radiofrequency
can still be considered successful when it allows the refer- ablation (RFA) for treatment of osteoid osteoma in 1992, new
ring provider to make a decision about treating or monitoring techniques have been developed, and the indications for abla-
a lesion, or at least to rule out a more urgent concern, such as tion have expanded to include treatment of metastatic malig-
infection. Accuracy is higher for malignant and high-grade nancy [31, 32]. The main principle of ablative therapy is to
lesions, large tumors, osteolytic lesions, and metastases [1, eradicate tumor cells in situ without the need for surgical
11, 22, 23]. Factors that may decrease biopsy success are resection. When compared to surgery, percutaneous ablation
usually related to difficulty in obtaining adequate samples, is less invasive and usually associated with lower cost, shorter
including densely sclerotic lesions or presence of necrosis. hospitalization times, and similar or better safety profile [32].
Lower success rate has also been reported for benign or low-­
grade tumors, spine lesions, lack of soft tissue component,
myxoid lesions, and infection [11, 22, 23, 28]. 30.3.1 Indications and Patient Selection
The size and number of samples may also influence the
biopsy success rate. Wu et al. reported higher diagnostic Percutaneous ablation has shown many benefits in the treat-
yield with samples measuring at least 10 mm in length and ment of metastatic disease in the musculoskeletal system
noticed increased yield with additional specimens, reaching [32, 33]. Indications include chronic pain that failed to
30 Image-Guided Biopsy and Intervention for Metastatic Disease 329

a b c e

Fig. 30.3 A 64-year-old female with a history of metastatic breast can- lytic and blastic lesion (arrows). Heat packs were placed on the skin
cer presents with moderate right knee pain. (a) Anteroposterior radio- surface during the cryoablation to protect the skin from freezing (*).
graph of the right distal femur shows a radiodense metastasis in the Biopsy performed during the cryoablation confirmed metastatic breast
medial femoral condyle that had been previously irradiated (arrow). (b) cancer. (e) Coronal MIP image from a PET performed 7 months after
Coronal maximum intensity projection (MIP) image from a PET shows the cryoablation shows resolution of the FDG activity within the right
increased FDG activity within the right medial femoral condyle lesion, medial femoral condyle lesion. The patient also reported decreased
consistent with recurrence (arrow). (c, d) Axial CT images of the right knee pain after the cryoablation
medial femoral condyle show two cryoablation probes within the mixed

improve with standard therapy (Fig. 30.3), risk of increased and at least two multicenter trials [37, 38]. To be an adequate
morbidity with progression of disease (e.g., higher risk for a candidate for treatment, the patient should be experiencing at
pathologic fracture), adjuvant therapy before surgical resec- least moderate pain, usually a visual analog scale score of 4
tion, and oligometastatic disease in which surgery is not indi- out of 10, according to most authors [32, 39]. Ideally, symp-
cated (Fig. 30.4). toms should also be localized in an area that correlates to the
Pain palliation is the most common indication for percuta- presence of a metastatic lesion on cross-sectional imaging.
neous image-guided ablation in the setting of metastatic dis- Lesions with osteolytic or soft tissue components are associ-
ease. Up to 50–79% of cancer patients experience ated with better treatment response [39, 40]. The location of
metastasis-­related pain, but only 20–30% achieve long-­lasting the targeted area also plays a role in patient selection, since
pain relief with conventional therapies, which usually involve proximity to other important structures, such as peripheral
external beam radiation therapy or chemotherapy [34, 35]. nerves or the spinal cord, may preclude the use of certain
Ablation may help decrease pain by destroying nerve end- ablation techniques due to the risk of thermal injury. Patients
ings, decreasing tumor volume, reducing neurostimulating should also be able to tolerate conscious sedation or anesthe-
cytokines from tumor cells, and inhibiting osteoclastic activ- sia, which is required in many cases.
ity at the interface between normal bone and tumor [34]. In a Oligometastatic disease is another indication for percuta-
multicenter trial with 61 patients undergoing cryoablation for neous image-guided ablation and is defined as the presence
treatment of painful musculoskeletal metastases, 75% of of 5 or fewer metastatic lesions. Studies have demonstrated
them achieved 90% or higher pain relief at some point during rates of local control varying from 36% to 97% with the use
follow-up, and pain recurrence was observed in only 14% of of different ablation techniques [41, 42]. The most common
the cases [36]. RFA has also been proven to be an effective primary malignancies in these studies were renal, breast, and
tool for the treatment of painful metastases on prior reports prostate cancer.
330 J. R. T. Vicentini et al.

a b c

d e

Fig. 30.4 A 73-year-old female with a history of thyroid cancer, status image during radiofrequency ablation shows the radiofrequency probe
post 100 mCi I-131 treatment, presents with rising thyroglobulin. (a) within the left proximal femur lesion (arrow). Biopsy at the time of
Fused axial image from a PET-CT shows an FDG-avid lesion in the left ablation confirmed metastatic thyroid cancer. (e) Fused axial image
proximal femur (arrow), concerning metastatic disease. Axial (b) from a PET-CT performed 2.5 years after the radiofrequency ablation
T1-weighted and (c) T2-weighted fat-suppressed MR images show the shows increased attenuation of the left proximal femur lesion, repre-
low T1 signal intensity, high T2 signal intensity lesion within the left senting post-ablation change, and resolution of the FDG activity
proximal femur (arrows), concerning metastatic disease. (d) Axial CT

30.3.2 Radiofrequency Ablation (RFA) (b) Technical details

As with biopsies, all the available imaging is reviewed
(a) Background prior to the procedure in order to identify the ablation
RFA uses an alternating electric current to generate target and surrounding structures. Proximity to nerves or
heat within the targeted tissue. Most systems include a vital organs may influence the choice of probe size and
generator, dispersive grounding pads, and a probe with planned ablation area. For pain palliation, a successful
an exposed tip, which is placed at the area of treatment. ablation may prioritize the bone–tumor interface with-
The generator creates a high-frequency alternating cur- out necessarily covering the entire lesion, especially if
rent that is transmitted to the probe and dispersed by the safety to surrounding structures is a concern. On the
grounding pads. Ions in the vicinity of the probe tip will other hand, for oligometastatic disease, it is important to
oscillate in an attempt to align with the alternating cur- be able to ablate the lesion plus a margin in order to
rent, which creates an elliptical-shaped area of frictional achieve local control [32]. Large vessels should be
heat centered at the tip of the probe [34]. The high tem- avoided because they act as heat sinks and dissipate the
peratures lead to protein denaturation and cell mem- heat away from the ablation zone.
brane damage, eventually leading to cell death. The choice of access device, as well as size and num-
Irreversible cell damage occurs in 4–6 min for tempera- ber of probes, will depend on size and location of the
tures around 50 °C, and it is instantaneous for tempera- lesion. Imaging guidance is usually performed with
tures in the range of 60–100 °C [34]. CT. However, ultrasound and MRI can also be used
30 Image-Guided Biopsy and Intervention for Metastatic Disease 331

when compatible equipment is available. For larger ball with CT or MR imaging throughout the procedure.
lesions, an array of probes may be used to increase the Larger tumors may require more than one session to
size of the ablation zone. Once the probe tip is posi- complete the treatment.
tioned, the system is monitored until the target tempera- (c) Complications
ture is achieved. This should be maintained for at least Complications are rare in cryoablation, in part due to
5 min and up to 15 min. Larger tumors may require addi- adequate patient selection. A multicenter trial reported a
tional ablation cycles. 2% complication rate (1 out of 61 patients) with the
(c) Complications development of osteomyelitis of the treated area [36]. In
Post-ablation syndrome presenting predominantly a study with 40 patients who underwent cryoablation for
with malaise has been described, likely due to the release treatment of oligometastatic disease, 2 patients (5%)
of cytokines following tumor cell death [34]. Paradoxical developed major complications including avascular
increase in pain can also occur in the first few days after osteonecrosis of the femoral head and ureteral stricture
the ablation. More important complications such as [42]. Permanent skin damage is a potential complication
adjacent nerve damage, skin burns, and subsequent frac- that should be addressed for superficial lesions.
tures have been reported in a minority of patients [38].

30.3.4 Other Techniques

30.3.3 Cryoablation
30.3.4.1 Microwave Ablation
Microwave ablation (MWA) is a relatively new, yet promis-
(a) Background ing technique for the treatment of musculoskeletal tumors.
Cryoablation promotes cell death by freezing the Similar to RFA, MWA utilizes thermal energy to cause cell
affected area. Different cellular death mechanisms hap- death. However, instead of using an alternating electric cur-
pen simultaneously during cryoablation. Differences in rent, MWA devices use an electromagnetic field. Malignant
solute concentration resulting from extracellular ice for- tumors are well perfused and usually show high water con-
mation lead to cell dehydration and subsequent damage tent, which makes them especially susceptible to the effects
to enzymatic machinery and the cell membrane. of the electromagnetic field. Because microwave propaga-
Intracellular ice crystals formed during this process also tion does not depend on electric conductivity, heat distribu-
damage the cell components through mechanical shear tion is more uniform, and larger ablation zones (up to
injury. Lastly, the lower temperatures promote ischemia 5–7 cm) can be achieved more rapidly than with RFA, even
and endothelial damage in the affected tissues [34]. in tissues with high impedance, such as bone and lung [43].
(b) Technical details Another advantage is that microwave systems do not require
Most cryoablation systems use pressurized argon gas grounding pads or other additional equipment. The micro-
as a coolant and pressurized helium for thawing. Cell wave antenna design, microwave frequency used, and power
death usually occurs at sustained temperatures of −40 °C applied are the main determinants of the final size and shape
or lower. Studies have demonstrated that this process is of the ablation zone. Disadvantages include the lack of
more effective using cycles of freeze-thaw–freeze rather smaller gauge antennas in some MWA systems, which can
than continuous freezing only [34]. Most protocols make the treatment of smaller lesions more challenging [44].
involve freeze-thaw–freeze cycles lasting 10, 5, and Potential complications include posttreatment instability of
10 min, respectively [32]. Each cryoprobe produces an the treated area with a higher risk of pathologic fracture,
ice ball whose size depends on the length of the non-­ which has led some practices to combine cement osteoplasty
insulated probe tip, and multiple cryoprobes can be following MWA therapy [45].
arranged so that the ice balls they generate cover a larger
area. For pain palliation, the bone–tumor interface is 30.3.4.2 MR Imaging-Guided Extracorporeal
­prioritized when the entire lesion cannot be safely treated Focused Ultrasound (MRgFUS)
because of its proximity to important structures. To MRgFUS combines MRI guidance and extracorporeal
achieve local control of a tumor, the ice ball should ide- focused ultrasound energy to generate heat and promote the
ally extend at least 3 mm beyond its margins [40]. This destruction of periosteal innervation. It has shown promising
is necessary because at the margins of the ice ball, the results in pain palliation for osseous metastatic disease with-
temperature remains around 0 °C, which would still out significant adverse effects [46]. This indication was
allow some cells to be viable after the procedure. One of approved by the Food and Drug Administration (FDA) in the
the main advantages of cryoablation when compared to United States in 2012 and is now reimbursed by many insur-
RFA is the possibility of monitoring the size of the ice ance companies [47]. Because of the high acoustic energy
332 J. R. T. Vicentini et al.

absorption by osseous tissue, energy levels as low as 30% of References

that used for soft tissue lesions can be employed on bone
lesions with similar effects [48]. Additionally, specific MR 1. Tomasian A, Hillen TJ, Jennings JW. Bone biopsies: what radiolo-
gists need to know. Am J Roentgenol. 2020;215:523–33.
techniques can be used for real-time thermometry and nonin- 2. Ulano A, Bredella MA, Burke P, Chebib I, Simeone FJ, Huang
vasive temperature monitoring during the procedure. AJ, et al. Distinguishing untreated osteoblastic metastases from
MRgFUS is truly noninvasive and can be used in areas previ- enostoses using CT attenuation measurements. Am J Roentgenol.
ously exposed to radiation therapy. Potential complications 2016;207:362–8.
3. Bernard S, Walker E, Raghavan M. An approach to the evaluation
include skin burns and thermal damage to adjacent struc- of incidentally identified bone lesions encountered on imaging
tures. Sonication-related pain has also been reported as a studies. AJR Am J Roentgenol. 2017;208:960–70.
limiting factor for some patients. 4. Moser T, Ehlinger M, Chelli Bouaziz M, Fethi Ladeb M, Durckel
J, Dosch J-C. Pitfalls in osteoarticular imaging: how to distin-
guish bone infection from tumour? Diagn Intervent Imaging.
30.3.4.3 Laser Ablation 2012;93:351–9.
Another alternative to thermal ablation is the use of infrared 5. Goodin GS, Shulkin BL, Kaufman RA, McCarville MB. PET/
light energy or laser ablation. The laser energy is adminis- CT Characterization of fibroosseous defects in children: 18
tered using optic fibers, and the equipment consists of a com- F-FDG uptake can mimic metastatic disease. Am J Roentgenol.
2006;187:1124–8.
mercially available laser, such as diode or Nd:YAG laser, a 6. Plaza JA, Perez-Montiel D, Mayerson J, Morrison C, Suster
disposable applicator, and a built-in cooling system. S. Metastases to soft tissue: a review of 118 cases over a 30-year
Advantages of laser ablation include MRI compatibility, period. Cancer. 2008;112:193–203.
rapid rate of heating, which makes the procedure time-­ 7. Sammon J, Jain A, Bleakney R, Mohankumar R. Magnetic reso-
nance imaging appearance of soft-tissue metastases: our experience
efficient, and the possibility of creating temperature maps to at an orthopedic oncology center. Skelet Radiol. 2017;46:513–21.
monitor the treatment zones [39, 49]. To achieve a sizable 8. Murphey MD, Carroll JF, Flemming DJ, Pope TL, Gannon FH,
ablation zone, more than one optic fiber is usually needed. Kransdorf MJ. From the archives of the AFIP: benign musculoskel-
Potential complications are similar to other types of thermal etal lipomatous lesions. Radiographics. 2004;24:1433–66.
9. Wu JS, Hochman MG. Soft-tissue tumors and tumorlike lesions: a
ablation and usually involve damage to adjacent structures. systematic imaging approach. Radiology. 2009;253:297–316.
10. Meek RD, Mills MK, Hanrahan CJ, Beckett BR, Leake RL, Allen
H, et al. Pearls and pitfalls for soft-tissue and bone biopsies: a
cross-institutional review. Radiographics. 2020;40:266–90.
Conclusion 11. Huang AJ, Kattapuram SV. Musculoskeletal neoplasms: biopsy and
Review of the indications and careful preprocedural intervention. Radiol Clin N Am. 2011;49:1287–305. vii
planning are vital to increase the chances of success of 12. Filippiadis DK, Charalampopoulos G, Mazioti A, Keramida K,
diagnostic and therapeutic image-guided interventions. Kelekis A. Bone and soft-tissue biopsies: what you need to know.
Semin Intervent Radiol. 2018;35:215–20.
A multidisciplinary approach is the best option to ensure 13. Kim MS, Lee S-Y, Cho WH, Song WS, Koh J-S, Lee JA, et al.
an effective and safe procedure. The wide variety of Tumor necrosis rate adjusted by tumor volume change is a better
biopsy tools and ablation techniques are being continu- predictor of survival of localized osteosarcoma patients. Ann Surg
ally improved, and new innovations are expected. Oncol. 2008;15:906–14.
14. Rakheja R, Makis W, Tulbah R, Skamene S, Holcroft C, Nahal A,
et al. Necrosis on FDG PET/CT correlates with prognosis and mor-
• Can artificial intelligence improve our accuracy by tality in sarcomas. Am J Roentgenol. 2013;201:170–7.
helping select the most promising targets for 15. Oliveira MP, Lima PM de A, Silva HJ da, Mello RJV de. Neoplasm
biopsy? seeding in biopsy tract of the musculoskeletal system. A systematic
review. Acta Ortop Bras 2014;22:106–110.
• Will there be new tools or testing for intraproce- 16. UyBico SJ, Motamedi K, Omura MC, Nelson SD, Eilber FC,
dural specimen analysis, decreasing the need for Eckardt J, et al. Relevance of compartmental anatomic guidelines
frozen pathology? for biopsy of musculoskeletal tumors: retrospective review of 363
• Can new biopsy devices increase safety to allow biopsies over a 6-year period. J Vasc Intervent Radiol. 2012;23:511–
518.e2.
sampling in even more challenging locations? 17. Siddiqi MA, Kim H-S, Jede F, Han I. Association of core needle
• With the development of improved systemic treat- biopsy tract resection with local recurrence in extremity soft tissue
ment regimens to increase survival, will percutane- sarcoma. Skelet Radiol. 2017;46:507–12.
ous ablations become a standard option for the 18. Espinosa LA, Jamadar DA, Jacobson JA, DeMaeseneer MO,
Ebrahim FS, Sabb BJ, et al. CT-guided biopsy of bone: a radiolo-
treatment of oligometastatic disease? gist’s perspective. AJR Am J Roentgenol. 2008;190:W283–9.
• Will new artificial intelligence algorithms be able to 19. Kumar R, Mittal BR, Bhattacharya A, Singh H, Bal A, Prakash G,
predict the ablation success rate based on pre- or et al. 18F-FDG PET/CT-Guided real-time automated robotic arm–
posttreatment appearance of the lesions? assisted needle navigation for percutaneous biopsy of hypermeta-
bolic bone lesions: diagnostic performance and clinical Impact. Am
J Roentgenol 2019;212:W10–W18.
The possibilities are many and are likely to con- 20. Carrino JA, Khurana B, Ready JE, Silverman SG, Winalski
tinue expanding in future years. CS. Magnetic resonance imaging-guided percutaneous biopsy of
musculoskeletal lesions. J Bone Joint Surg Am. 2007;89:2179–87.
30 Image-Guided Biopsy and Intervention for Metastatic Disease 333

21. Hillen TJ, Talbert RJ, Friedman MV, Long JR, Jennings JW, Wessell 36. Callstrom MR, Dupuy DE, Solomon SB, Beres RA, Littrup
DE, et al. Biopsy of CT-occult bone lesions using anatomic land- PJ, Davis KW, et al. Percutaneous image-guided cryoablation
marks for CT guidance. AJR Am J Roentgenol. 2017;209:214–21. of painful metastases involving bone: multicenter trial. Cancer.
22. Omura MC, Motamedi K, UyBico S, Nelson SD, Seeger 2013;119:1033–41.
LL. Revisiting CT-guided percutaneous core needle biopsy of 37. Dupuy DE, Liu D, Hartfeil D, Hanna L, Blume JD, Ahrar K, et al.
musculoskeletal lesions: contributors to biopsy success. AJR Am J Percutaneous radiofrequency ablation of painful osseous metas-
Roentgenol. 2011;197:457–61. tases: a multicenter American College of Radiology Imaging
23. Hau A, Kim I, Kattapuram S, Hornicek FJ, Rosenberg AE, Gebhardt Network trial. Cancer. 2010;116:989–97.
MC, et al. Accuracy of CT-guided biopsies in 359 patients with 38. Goetz MP, Callstrom MR, Charboneau JW, Farrell MA, Maus TP,
musculoskeletal lesions. Skelet Radiol. 2002;31:349–53. Welch TJ, et al. Percutaneous image-guided radiofrequency abla-
24. Stoker DJ, Cobb JP, Pringle JA. Needle biopsy of musculoskel- tion of painful metastases involving bone: a multicenter study. J
etal lesions. A review of 208 procedures. J Bone Joint Surg Br. Clin Oncol. 2004;22:300–6.
1991;73:498–500. 39. Kurup AN, Callstrom MR. Ablation of skeletal metastases: current
25. Hryhorczuk AL, Strouse PJ, Biermann JS. Accuracy of CT-guided status. J Vasc Intervent Radiol. 2010;21:S242–50.
percutaneous core needle biopsy for assessment of pediatric mus- 40. Callstrom MR, Kurup AN. Percutaneous ablation for bone
culoskeletal lesions. Pediatr Radiol. 2011;41:848–57. and soft tissue metastases—why cryoablation? Skelet Radiol.
26. Jelinek JS, Murphey MD, Welker JA, Henshaw RM, Kransdorf 2009;38:835–9.
MJ, Shmookler BM, et al. Diagnosis of primary bone tumors with 41. Kurup AN, Morris JM, Callstrom MR. Ablation of musculoskeletal
image-guided percutaneous biopsy: experience with 110 tumors. metastases. American J Roentgenol. 2017;209:713–21.
Radiology. 2002;223:731–7. 42. McMenomy BP, Kurup AN, Johnson GB, Carter RE, McWilliams
27. Rimondi E, Rossi G, Bartalena T, Ciminari R, Alberghini M, RR, Markovic SN, et al. Percutaneous cryoablation of musculo-
Ruggieri P, et al. Percutaneous CT-guided biopsy of the musculo- skeletal oligometastatic disease for complete remission. J Vasc
skeletal system: results of 2027 cases. Eur J Radiol. 2011;77:34–42. Intervent Radiol. 2013;24:207–13.
28. Datir A, Pechon P, Saifuddin A. Imaging-guided percutaneous 43. Hinshaw JL, Lubner MG, Ziemlewicz TJ, Lee FT, Brace
biopsy of pathologic fractures: a retrospective analysis of 129 CL. Percutaneous tumor ablation tools: microwave, radiofrequency,
cases. AJR Am J Roentgenol. 2009;193:504–8. or cryoablation—what should you use and why? Radiographics.
29. Wu JS, Goldsmith JD, Horwich PJ, Shetty SK, Hochman MG. Bone 2014;34:1344–62.
and soft-tissue lesions: what factors affect diagnostic yield of 44. Lubner MG, Brace CL, Hinshaw JL, Lee FT. Microwave tumor
image-guided core-needle biopsy? Radiology. 2008;248:962–70. ablation: mechanism of action, clinical results, and devices. J Vasc
30. Yang YJ, Damron TA. Comparison of needle core biopsy and Intervent Radiol. 2010;21:S192–203.
fine-needle aspiration for diagnostic accuracy in musculoskeletal 45. Deib G, Deldar B, Hui F, Barr JS, Khan MA. Percutaneous micro-
lesions. Arch Pathol Lab Med. 2004;128:759–64. wave ablation and cementoplasty: clinical utility in the treatment of
31. Rosenthal DI, Alexander A, Rosenberg AE, Springfield D. Ablation painful extraspinal osseous metastatic disease and myeloma. Am J
of osteoid osteomas with a percutaneously placed electrode: a new Roentgenol. 2019;212:1377–84.
procedure. Radiology. 1992;183:29–33. 46. Liberman B, Gianfelice D, Inbar Y, Beck A, Rabin T, Shabshin
32. Rosenthal D, Callstrom MR. Critical review and state of the art in N, et al. Pain palliation in patients with bone metastases using
interventional oncology: benign and metastatic disease involving MR-guided focused ultrasound surgery: a multicenter study. Ann
bone. Radiology. 2012;262:765–80. Surg Oncol. 2009;16:140–6.
33. Callstrom MR, Charboneau JW, Goetz MP, Rubin J, Atwell TD, 47. Scipione R, Anzidei M, Bazzocchi A, Gagliardo C, Catalano C,
Farrell MA, et al. Image-guided ablation of painful metastatic bone Napoli A. HIFU for Bone metastases and other musculoskeletal
tumors: a new and effective approach to a difficult problem. Skelet applications. Semin Intervent Radiol. 2018;35:261–7.
Radiol. 2006;35:1–15. 48. Tempany CMC, McDannold NJ, Hynynen K, Jolesz FA. Focused
34. Rybak LD. Fire and ice: thermal ablation of musculoskeletal ultrasound surgery in oncology: overview and principles.
tumors. Radiol Clin N Am. 2009;47:455–69. Radiology. 2011;259:39–56.
35. Simon CJ, Dupuy DE. Percutaneous minimally invasive therapies 49. Ahrar K, Stafford RJ. Magnetic resonance imaging-guided laser
in the treatment of bone tumors: thermal ablation. Sem Musculoskel ablation of bone tumors. Tech Vasc Interv Radiol. 2011;14:177–82.
Radiol. 2006;10:137–44.
 Part VIII
Sentinel Lymph Node Surgery for Solid Cancer

Mark B. Faries

“The danger lies in the diffusion of malignant cell particles from this primary focus; these always impli-
cate the nearest lymph glands, which intercept them for a time. Eventually they pass beyond such “traps”
into the blood current; then death, with multiple visceral metastases, ensues. Palpable enlargement of
these glands is unfortunately but a late symptom of deposit therein; by the time it occurs there is almost
always implication of deeper organs or tissues … We further see the paramount importance of securing,
whenever possible, the perfect eradication of those lymph glands which will necessarily be first infected;
before enlargement takes place radical removal of such organs in the axilla, groin, surface regions of the
neck &c, before they have undergone appreciable increase in bulk, is a safe and easy measure, which,
under the conditions indicated, should never be neglected.”
-Herbert Snow MD, Lancet, 1892 [1]

For many solid tumors, the importance of lymphatic spread and lymph node metastases has
been abundantly clear from the time of their earliest descriptions. Regional lymph nodes rep-
resent the initial site of metastasis for most solid tumors and the presence of disease in those
nodes is generally associated with a marked decrease in prognosis. However, although these
central observations are self-evident, their implications for clinical medicine and the treatment
of patients with cancer have long been the source of great controversy and remain the subject
of extensive research. Initial proposals, such as that of Herbert Snow above, were for early and
radical excision of draining regional lymph nodes automatically, even before any evidence of
tumor spread to those locations.
This reaction was intuitive, particularly in an era when most primary tumors were relatively
advanced at diagnosis. However, the price of such a radical approach was also apparent in the
morbidity associated with complete dissections. Patients undergoing groin or axillary dissec-
tions are at risk for pain, nerve dysfunction, and, most concerningly, lymphedema of the adja-
cent extremity. This last toxicity, though treatable, is often permanent at least to some degree.
As primary tumors were diagnosed earlier, the frequency of nodal involvement decreased and
the wisdom of complete nodal dissection for all patients was increasingly questioned. In breast
cancer and melanoma, randomized trials were conducted to evaluate the value of immediate or
elective lymph node dissection in the absence of clinically apparent metastases [2, 3]. In gas-
tric, lung, and esophageal cancers, more extensive nodal dissections were compared to less
extensive ones, and debates raged in surgical and oncologic meetings around the world [4–6].
Into this swirl of controversy Donald Morton, Alistair Cochran, and their colleagues at the
UCLA introduced a novel technique for intraoperative mapping of the lymphatic drainage
from the primary tumor site to the initial node(s) along the pathway [7]. The pathologic status
of these sentinel lymph nodes (SLN) was found to be representative of that of the entire basin.
This technique has gone on to revolutionize the care of not only melanoma but also breast
cancer, and it is being productively explored in multiple other malignancies as outlined in this
section.
The concept of a “sentinel” node was not new. The idea of a node that was the specific site
of drainage from a tumor had arisen several times, likely independently, in the past [8]. Cabanas
proposed the idea for penile cancers as did Gould for the parotid and Kett in breast cancer
336 Sentinel Lymph Node Surgery for Solid Cancer

[9–11]. Similar ideas regarding specific patterns of lymphatic drainage were put forward as early as 1923
by Braithwaite, while Virchow described the drainage of tattoo ink in the skin to regional nodes as early as
the nineteenth century. However, a practical, reproducible, and reliable surgical technique for finding the
critical node eluded the medical community until Morton described it.
With SLN biopsy it became possible to ascertain the pathologic status of the regional nodal basin with a
minimally invasive, outpatient procedure. A revolution was launched that has had far-reaching implications
that have benefitted countless patients around the world.
Why was SLN biopsy such an advance? It offers substantial advantages over complete dissection of an
entire nodal basin. First it is less morbid than full dissection. It can be completed through much smaller
incisions with much less dissection. Nerves, blood vessels, and lymphatic channels that are transected in
dissections are largely preserved in SLN biopsy. The risk of long-term toxicity, including lymphedema, is
also markedly reduced. Second, this decrease in the extent of surgery is not accompanied by any reduction
in the accuracy of regional staging. In fact, the opposite is true. Since the number of nodes that must be
examined by the pathologist is greatly reduced, the intensity of evaluation of the SLN is able to be increased.
Routine use of immunohistochemical markers is possible in SLN, making small metastases more readily
detected. In melanoma, as an example, this has resulted in greatly improved prognostic discrimination for
node-negative and node-positive patients, which remarkably improved staging in the current 8th Edition of
the American Joint Commission on Cancer staging system [12].
Finally, it appears that SLN biopsy may be therapeutic in some respects, and not merely a staging pro-
cedure on the way to a therapeutic nodal dissection. The concept appears to be most clearly evaluated for
melanoma, where in the second Multicenter Selective Lymphadenectomy Trial (MSLT-II) patients with
SLN metastases were randomized to either completion dissection or nodal observation [13]. One of the less
appreciated results of the trial was that among observation arm patients, all of whom had nodal disease,
only a minority ever developed regional nodal recurrence and never required full dissection. So for about
three-quarters of the patients, all nodal disease was eradicated with the SLN biopsy alone. From MSLT-I we
also know that early removal of nodal disease likely results in improved long-term melanoma-­specific sur-
vival [14]. The degree to which these observations apply in other diseases will likely vary based on the
biology of those malignancies. In breast cancer, as a different example, regional nodal disease may be more
likely to be controlled with nonsurgical measures, making a therapeutic impact less likely [15]. Additional
study will be required to fully elucidate the relative therapeutic value of SLN biopsy in different
malignancies.
SLN biopsy continues to be developed in other malignancies, including upper gastrointestinal cancer,
colon cancer, and lung cancer. These diseases have proved more challenging both technically and in terms
of evaluation of efficacy. Advances in technology have increased the likelihood of accurate and practical
application of lymphatic mapping beyond breast cancer and melanoma. These advances include the use of
fluorescent dyes that may be visualized during laparoscopic or robotic surgical approaches [16].
Ferromagnetic mapping agents may also come into increased use as research continues. SPECT/CT may
enhance the accuracy of mapping in anatomically difficult regions such as the head and neck [17]. These
developments are likely to continue to open avenues of investigation to attempt to improve care for patients
in the future.
A final area of active investigation relates to the increasingly multidisciplinary nature of oncology in the
modern era. In the past, the pathway between the surgeon and the medical oncologist was a one-way street.
With the advent of neoadjuvant approaches to many cancers, treatments have become better and more
sophisticated, albeit more complicated. In breast cancer, where SLN biopsy and neoadjuvant systemic treat-
ment have had perhaps the most interaction, it appears that lymphatic mapping is more challenging than it
is in the de novo setting. Work continues to optimize the interaction of the two techniques to render treat-
ment that is both accurate and minimally morbid.

References
1. Snow HM. Melanotic cancerous disease. Lancet. 1892;140:869–922.
2. Fisher B, Jeong JH, Anderson S, et al. Twenty-five-year follow-up of a randomized trial comparing
radical mastectomy, total mastectomy, and total mastectomy followed by irradiation. N Engl J Med.
2002;347:567–75.
Sentinel Lymph Node Surgery for Solid Cancer 337

3. Balch CM, Cascinelli N, Sim FH. Elective lymph node dissection: results of prospective randomized
surgical trials. In: Balch CM, Houghton AN, Sober AJ, et al., editors. Cutaneous melanoma. Saint
Louis, MO: Quality Medical Publishing; 2003. p. 379–95.
4. Songun I, Putter H, Kranenbarg EM, et al. Surgical treatment of gastric cancer: 15-year follow-up
results of the randomised nationwide Dutch D1D2 trial. Lancet Oncol. 2010;11:439–49.
5. Fok M, Siu KF, Wong J. A comparison of transhiatal and transthoracic resection for carcinoma of the
thoracic esophagus. Am J Surg. 1989;158:414–9.
6. Izbicki JR, Thetter O, Habekost M, et al. Radical systematic mediastinal lymphadenectomy in non-­
small cell lung cancer: a randomized controlled trial. Br J Surg. 1994;81:229–35.
7. Morton D, Wen D, Wong J, et al. Technical details of intraoperative lymphatic mapping for early stage
melanoma. Arch Surg. 1992;127:392–9.
8. Nieweg OE, Uren RF, Thompson JF. The history of sentinel lymph node biopsy. Cancer J.
2015;21:3–6.
9. Cabanas R. An approach for the treatment of penile carcinoma. Cancer. 1977;39:456–66.
10. Gould EA, Winship T, Philbin PH, et al. Observations on a "sentinel node" in cancer of the parotid.
Cancer. 1960;13:77–8.
11. Kett K, Lukacs L. Diagnosis of regional lymph node metastases in cancer of the breast. A study of the
lymphatic system of the diseased breast. Am J Surg. 1974;128:398–401.
12. Gershenwald JE, Scolyer RA, Hess KR, et al. Melanoma staging: evidence-based changes in the
American joint committee on cancer eighth edition cancer staging manual. CA Cancer J Clin.
2017;67:472–92.
13. Faries MB, Thompson JF, Cochran AJ, et al. Completion dissection or observation for sentinel-node
metastasis in melanoma. N Engl J Med. 2017;376:2211–22.
14. Morton DL, Thompson JF, Cochran AJ, et al. Final trial report of sentinel-node biopsy versus nodal
observation in melanoma. N Engl J Med. 2014;370:599–609.
15. Giuliano AE, Ballman K, McCall L, et al. Locoregional recurrence after sentinel lymph node dissection
with or without axillary dissection in patients with sentinel lymph node metastases: long-term follow-
­up from the American College of Surgeons oncology group (Alliance) ACOSOG Z0011 randomized
trial. Ann Surg. 2016;264:413–20.
16. Ito N, Fukuta M, Tokushima T, et al. Sentinel node navigation surgery using indocyanine green in
patients with lung cancer. Surg Today. 2004;34:581–5.
17. Chapman BC, Gleisner A, Kwak JJ, et al. SPECT/CT improves detection of metastatic sentinel lymph
nodes in patients with head and neck melanoma. Ann Surg Oncol. 2016;23:2652–7.
 Sentinel Lymph Node Biopsy
for Primary Cutaneous Malignancy 31
Marc Moncrieff and Howard Peach

Abstract
Learning Objectives
Sentinel lymph node biopsy (SLNB) is the international
standard of care for providing prognostic information in This chapter will inform the reader about:
cutaneous melanoma. This chapter describes the ratio-
nale behind development of sentinel node biopsy in this • Why elective lymph node dissection was initially
role, as well as the important practical considerations advocated, what the clinical concerns about it were
for lymphoscintigraphic imaging. It chronicles the sci- and how it helped develop the role of SLNB.
entific evidence gleaned from large international clini- • How the new procedure of SLNB was developed,
cal trials and how this evidence was used to shape introduced into clinical practice and how it impacts
current clinical practice. This includes detailed analysis on prognosis.
of the MSLT-1, MSLT-2 and DeCOG trials, which has • The logic of using the current radio-labelled tracers
permitted the safe reduction in recommended surgical and how the use of SPECT-CT improves localisa-
procedures. There is a distillation of the evidence on the tion accuracy.
relevance of microscopic deposits within the node and • The results of the MSLT-1, -2 and DeCOG studies.
other pathological findings and how the extent of this It explains the controversy around particularly the
disease impacts on prognosis. It concludes with a sum- MSLT-1 study and how each trial has helped define
mary of current recommended practice. the role of SLNB and the cohort of patients who are
Finally, the role of SLNB in the context of other skin most likely to benefit.
cancers is less transparent. There is an explanation why • The importance of micrometastatic disease pathol-
the natural biology and the clinical presentation of other ogy and how it impacts on predicted future mela-
skin cancers questions the validity of SLNB in these noma recurrence.
pathologically different cancers. • Which cohort of melanoma patients and patients
with other skin malignancies should be offered
SLNB.

31.1 Introduction

M. Moncrieff (*) Primary cutaneous malignancies, particularly melanoma,

Norfolk & Norwich University Hospital NHS Trust, readily metastasise to the regional lymph nodes by accessing
Norwich, United Kingdom
e-mail: [email protected] the rich lymphovascular network afforded to the skin. Unlike
other cancers, the lymphatic drainage pattern of the primary
H. Peach
Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom tumour cannot be easily predicted based on its anatomical
e-mail: [email protected] location, making elective lymph node dissection an unreli-

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 339
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_31
340 M. Moncrieff and H. Peach

able and ineffective treatment strategy. Sentinel lymph node primary tumour. [4–7] However, analyses from these trials
biopsy (SLNB) is a technique developed to identify the lym- and others [8] suggested that a subgroup of patients with
phatic drainage of the primary tumour and to stage the intermediate thickness melanoma may benefit from early
regional lymph nodes to identify patients in need of further nodal intervention. The Intergroup Melanoma Surgical Trial
treatment. Since its development, SLNB has evolved to [9] randomised patients with melanomas 1–4 mm thick to
become the standard of care for primary cutaneous mela- ELND or observation. This was the first trial to mandate
noma and is incorporated in the current AJCC classification lymphatic mapping using preoperative lymphoscintigraphy
system. [1] This chapter describes the use of SLNB for the to direct the nodal dissection and found that, without this
staging and treatment of melanoma, and compares its appli- information, ELND would have been misdirected in more
cation to other cutaneous malignancies, such as Merkel cell than 50% of patients. Whilst the trial failed to show an over-
and squamous cell carcinomas. all survival benefit for ELND, a planned subgroup analysis
found an increased survival benefit for patients with thinner
melanomas (1.0–2.0 mm thick) and patients younger than
31.2  volution of Sentinel Lymph Node
E 60 years old. This was the first trial to show a survival benefit
Biopsy for Cutaneous Malignancy for early removal of microscopic nodal disease and, with
preoperative lymphoscintigraphy, became the foundation of
Primary cutaneous melanoma readily metastasises through the sentinel node biopsy paradigm.
the draining lymphatic system to the regional lymph nodes.
The risk of this event is directly proportionate to the depth of
invasion of the primary into the skin, termed Breslow thick- 31.4 Sentinel Lymph Node Biopsy
ness. Whilst the probability of nodal metastasis given a par-
ticular Breslow thickness is consistent across the literature, Donald Morton developed the technique for identifying the
[2] the process is almost certainly stochastic, meaning that a draining nodal basin and identifying the sentinel node. In
nodal metastasis may not be predicted precisely in any given the landmark paper, [10] the initial procedure termed,
patient with the current biomarkers available to clinicians at ‘intraoperative lymphatic mapping and selective lymphad-
present. enectomy’ incorporated preoperative lymphoscintigraphy
to identify the draining regional nodes and intraoperative
injection of blue dye to identify the sentinel nodes visually.
31.3 Elective Lymph Node Dissection This was the culmination of work from animal studies dem-
onstrating variable nodal drainage in the same nodal field
Clinicians have been aware for decades that the most impor- from different primary sites. [11] Crucially they also dem-
tant predictor of patient outcome or survival is the nodal sta- onstrated a predictable, repeatable and orderly nodal drain-
tus at the time of presentation of the primary melanoma. age pattern via the sentinel node. Using this technique in
Accordingly, once a nodal metastasis has been identified, humans, Morton’s team identified sentinel nodes in 82% of
then the contribution of the prognostic factors of the primary patients and had a false-negative rate of 1%. These data led
melanoma seemingly have little or no further influence on them to conclude that, ‘this technique identifies, with a
the patient’s prognosis. Prior to the 1990s, elective lymph high degree of accuracy, patients with early stage mela-
node dissection (ELND) was routinely practiced using the noma who have nodal metastases’. These results confirmed
rationale that removing microscopic disease may reduce the that the sentinel node is the initial site of regional lymphatic
risk of distant spread and accurately stage the patient. node metastases.
Unfortunately, the majority of the patients did not harbour Shortly afterwards, Morton and others described the
microscopic disease and underwent a morbid procedure for use of intraoperative radio-localisation of the sentinel node
which they received no clinical benefit. Clinicians were using a handheld gamma probe. [12] This technique was
aware that the risk of regional nodal metastasis increased rapidly validated by other centres internationally. [13–15]
according to the Breslow thickness, and that a dual-track risk Subsequently, Morton published a multicentre experience
of nodal and distant metastasis appeared to operate whereby, of the technique which described both a ‘learning phase’ of
the risk of distant metastasis outweighed the risk of nodal 30 cases for a centre to become proficient at the technique
metastasis for tumours >4 mm in thickness but the situation and a case volume of at least 25 cases per annum to main-
was reversed in the early or intermediate thickness tumours tain that proficiency. [16] Furthermore, the paper cemented
between 0.75 and 4 mm thick. [3] the current technique of dual-localisation using preoperative
Several international randomised controlled trials failed lymphoscintigraphy and intraoperative blue dye and radio-
to show any survival benefit for melanoma patients who localisation as the most accurate method for identifying the
underwent ELND at the time of surgical management of the sentinel node.
31 Sentinel Lymph Node Biopsy for Primary Cutaneous Malignancy 341

31.5 Technical Details 31.6  utcomes of Sentinel Lymph Node

O
of Lymphoscintigraphy Biopsy for Melanoma

The technical aspects for a modern sentinel lymph node The purpose of developing SLNB was to identify those
practice are covered in comprehensive detail by two review patients harbouring micrometastatic disease in the draining
articles from North America [17] and Europe [18] and it is regional lymph nodes to subsequently offer a lymphadenec-
unnecessary to replicate those details here. Whilst the surgi- tomy to prevent distant relapse and death. There were signals
cal technique has not changed substantially over the last from previous ELND studies that this may be a beneficial
decade, there have been significant improvements in the pre- therapeutic strategy for a specific subgroup of patients, par-
operative sentinel node localisation, brought about by ticularly those with intermediate thickness melanomas, 1.2–
changes in the imaging protocols and the choice of tracer 3.5 mm thick. [8, 9, 25] Accordingly, Morton and colleagues
agents. [19] devised the Multicenter Selective Lymphadenectomy Trial-1
The European Association of Nuclear Medicine (EANM) (MSLT-1) to investigate the therapeutic utility of SLNB with
states that, ‘hybrid imaging with SPECT and CT including subsequent completion lymph node dissection (CLND) for
anatomical information improves the localisation of SLNs micrometastatic nodal disease.
and reduces misinterpretation of images. Images obtained by
SPECT/CT are three-dimensional and have better contrast
and spatial resolution than planar images. For SPECT/CT, a
higher overall SLN detection rate and better detection of 31.6.1 MSLT-1 Study
SLNs located next to the injection site have been reported’.
A large prospective multicentre trial found that SPECT/CT The MSLT-1 study [26] can rightly be described as a land-
is also able to modify clinical decision-making. The mark trial in the surgical management of cutaneous mela-
International Atomic Energy Agency Sentinel Node Trial noma [27]. The study recruited 2001 patients with primary
demonstrated that SPECT/CT had modified the surgical cutaneous melanoma and randomised patients to either
approach in 37% of patients with melanoma, with greatest SLNB with subsequent CLND for a positive SLNB (SN+) or
improvement in accuracy seen in the head and neck and trun- nodal observation and subsequent therapeutic lymph node
cal regions. [20] dissection (TLND) for patients who developed palpable
Novel tracer agents have entered into mainstream use to lymphadenopathy during follow-up. The principal target
improve accuracy of localisation both preoperatively. One of cohort was intermediate thickness tumours (1.2–3.5 mm) but
the significant details of the procedure is timing regarding eligibility criteria allowed for melanomas >3.5 mm thick-
the decay in gamma radiation of 99mTc which constrains the ness, or <1.2 mm thickness if the tumour was Clark’s level
surgery to within, at most, 24 h of the preoperative localisa- IV or V. The randomisation cohorts were allocated in a 60:40
tion. Further concerns arise with temporal migration of the manner in favour of the SLNB group to power the study
tracer agent to the second echelon nodes, theoretically appropriately for nodal events in the SLNB field.
increasing the risk of harvesting an adjacent non-sentinel The study results showed a non-significant survival dif-
node, with the consequent risk of a false-negative biopsy. A ference in the primary endpoint of melanoma-specific sur-
recent European multicentre study found a significantly vival (MSS) for the intermediate thickness group. This was
reduced survival (hazard ratio 1.36, 95% CI 1.05–1.74; unsurprising given that the incidence of SN+ was 16.0% and
p = 0.018) and an increased nodal relapse rate (4.5% vs. only 20% developed nodal metastases in the observation
2.5%; p = 0.008) when using 99mTc-nanocolloid, in patients group. Accordingly, 80% of patients in either group could
who underwent SLNB >12 h after their preoperative SLN not have benefitted from any intervention, and this majority
localisation, compared to those who had their SLNB within meant that the study was not statistically powered to detect a
12 h. [21] Novel tracer agents, such as 99mTc-tilmanocept, survival difference, though at 10 years, there was a non-­
have been approved as alternatives to address these issues. significant 4% higher survival in the SLNB arm.
99m
Tc-tilmanocept is composed of a dextran backbone, Some commentators seized upon the negative outcome of
attached to units of mannose and DTPA and avidly binds to the primary endpoint and dismissed the therapeutic utility of
a receptor specific to reticuloendothelial cells. [22] SLNB and the trial as a result. However, the subgroup analy-
Alternative tracer agents, indocyanine green and MRI-­ sis of the data revealed significant findings regarding the
visible, magnetic nanoparticles have been investigated to biology of the tumour, which merit further discussion. One
replace the blue dye and radiolabelled 99mTC tracer agents of the major findings demonstrated the importance of SLNB
with promising results. [23, 24] as a staging tool for primary cutaneous melanoma. For SN+
342 M. Moncrieff and H. Peach

Fig. 31.1 (a) Pre-operative SPECT-CT imaging demonstrating lym- demonstrating lymphatic channels and sentinel nodes stained with pat-
phatic mapping of a patient with a melanoma on the left posterior scalp. ent blue dye. (Photographs courtesy of Howard Peach)
(b) Intraoperative photograph of a sentinel nodes in the right axilla

patients the 10-year melanoma-specific survival rate was groups. The mean number of nodes involved, for the inter-
62.1 ± 4.8% compared with 85.1 ± 1.5% for SN-negative mediate thickness cohort, was 1.4 ± 0.1 in the SLNB group
patients (hazard ratio for death from melanoma, 3.09; 95% versus 3.2 ± 0.4 in the observation group (p = 0.0001) for the
CI, 2.12–4.49; p < 0.001). Multivariate analysis revealed that intermediate thickness cohort. Patients in the SN group who
SLNB was the most important prognostic indicator for MSS subsequently relapsed in the nodal field (termed ‘false nega-
(HR: 2.40; 95% CI: 1.61–3.56; p < 0.001) (Fig. 31.1). tive’) had a similar nodal burden to the observation group
Another major finding was the highly significant differ- (4.3 ± 1.3; p < 0.0001). The cumulative incidence of nodal
ence in nodal disease burden and behaviour between the two metastasis in the SLNB and observation groups respectively
31 Sentinel Lymph Node Biopsy for Primary Cutaneous Malignancy 343

were (21.9% vs. 19.5%; p = NS) for the intermediate thick- positive sentinel node (about 90%) and were subject to the
ness cohort and identical (42.0% vs. 41.4%; p = NS)—Fig. additional morbidity of the procedure, from which they
31.2a, b in the thick melanoma cohort. The disease-free sur- would derive no benefit. Moreover, CLND was mandated at
vival for both the intermediate and thick melanoma cohorts that time, for most large phase III adjuvant systemic therapy
was significantly better in the SLNB groups, which was clinical trials, whilst those patients who demonstrated addi-
attributable to the reduced incidence of nodal relapse. Finally, tional metastases in their non-sentinel nodes (NSN) had an
the median time to nodal relapse was the same for the false-­ overall poor prognosis, for which there were no effective
negative SLNB group and the observation groups in both the systemic therapeutic options. [27]
intermediate thickness and thick melanoma cohorts, though To resolve these concerns, Dr. Morton and colleagues
the for the thick melanoma cohort overall the time to relapse devised the MSLT-2 study which randomised patients to
was approximately half the time of the intermediate thick- either observation of the metastatic nodal field with four-­
ness cohort overall (9.2 vs. 19.2 months). monthly targeted ultrasound (and TLND on clinical nodal
The most plausible interpretation of these data is that a recurrence) or CLND at the time of diagnosis of the positive
nodal micrometastasis will progress to palpable disease SLNB. [29] The study also incorporated a screening phase,
without intervention, and that intervention (i.e. SLNB) pre- with histologically negative SLNBs being screened for dis-
vents the nodal disease progressing to second echelon nodes ease using reverse transcription polymerase chain reaction
within the nodal field, thereby preventing further nodal (RT-PCR) testing. Those patients who were RT-PCR positive
relapse. Figure 31.2c, d shows the survival curves for the were then randomised to CLND or nodal observation. The
intermediate-thickness cohort, which was a planned sub- primary endpoint was melanoma-specific survival (MSS) at
group analysis. The analysis demonstrates that among 10 years. At approximately the same time, the Dermatologic
patients with nodal metastases from intermediate-thickness Cooperative Oncology Group (DeCOG) opened a trial with
melanomas, the 10-year melanoma-specific survival rate was a very similar trial design, [30] with minor differences that
62.1 ± 4.8% in the SLNB group as compared with included patients being screened with ultrasound every
41.5 ± 5.6% in the observation group (hazard ratio for death 3 months with a serum S100B measurement performed at the
from melanoma, 0.56; 95% CI, 0.37–0.84; p = 0.006). same time, and with head and neck melanomas being
Similarly, there was no difference in MSS between the false-­ excluded. Both trials excluded metastatic nodes demonstrat-
negative group and the observation group (HR, 1.15 (95% ing extracapsular spread (syn. extracapsular or extranodal
CI, 0.68–1.94); p = 0.60). Whilst, the comparison is statisti- extension) and patients who were immunocompromised.
cally problematic, since the randomisation took place at the The major difference between the two trials was that the
time of treatment of the primary, the data make a highly com- DeCOG-SLT had distant disease-free survival (distant-DFS)
pelling argument for a survival benefit for a small group of as the primary endpoint.
patients undergoing SLNB who have their focus of For the MSLT-2 study, [29] a total of 3531 patients were
­micrometastatic disease excised in the draining node. A fur- enrolled in the screening phase and 1939 patients underwent
ther, important therapeutic benefit, and a survivorship issue, randomisation, with 1713 pathology positive and 226
was the significant reduction in lymphoedema rates seen in RT-PCR positive. The median size of the deposits in the
the patients who underwent CLND in the SLNB cohort com- SLNB was 0.61 mm in the dissection group and 0.67 mm in
pared to those who underwent TLND in the observation the observation group. The majority of patients had only one
cohort (20.4% vs. 12.4%, p = 0.04). [28] positive sentinel node: 72.3% in the CLND group and 69.1%
in the observation group. There was no difference in MSS
between the two groups (86 ± 1.3% and 86 ± 1.2%; p = 0.42).
31.6.2 C
 ompletion Lymph Node Dissection The rate of disease-free survival was slightly higher in the
for Sentinel Node Positive Patients: dissection group than in the observation group (68 ± 1.7%
MSLT-2 and DeCOG Studies and 63 ± 1.7%, respectively; p = 0.05) at 3 years. The authors
assumed that this was based on the significantly reduced
The original rationale for undertaking SLNB was to identify regional nodal relapse rate in the CLND group at 3 years
patients who were harbouring micrometastatic disease in the (92 ± 1.0% vs. 77 ± 1.5%; P < 0.001). It is noteworthy that
regional nodal field and who would subsequently benefit the incidence of ‘node-only’ relapses in the observation
from completion lymph node dissection. However, it became group was 7.7% and the overall incidence was 4.7%. The
rapidly apparent that there was a equipoise regarding the presence of non-SLN metastases in the CLND arm was an
need for the CLND. On the one hand, patients benefitted independent predictor of MSS (HR 1.78; 95% CI 1.19–2.68,
from improved regional control and the extra prognostic p = 0.005). Subgroup analyses did not identify any groups
information that the procedure afforded, whilst on the other, with significant survival benefit from CLND. However,
the majority of patients had no further disease other than the CLND was associated with a significantly lymphoedema
344 M. Moncrieff and H. Peach

a Cumulative 10-Yr Incidence of Metastasis, Intermediate- b Cumulative 10-Yr Incidence of Nodal Metastasis, Thickness Melanornas
Thickness Melanornas
Rate (%) Rate (%)
Yr 3 Yr 5 Yr 7 Yr 9 Yr 10 Yr 3 Yr 5 Yr 7 Yr 9 Yr 10
OBS 13.7±1.58 16.3±1.71 17.0±1.75 18.8±1.86 19.5±1.91 OBS 33.2±4.5 38.7±4.8 40.0±4.8 41.4±4.9 41.4±4.9
SNB 19.5±1.44 20.3±1.46 20.7±1.48 21.2±1.50 21.9±1.54 SNB 40.4±3.8 41.1±3.8 42.0±3.8 42.0±3.8 42.0±3.8
50 50 Clinically detected recurrence
Positive
(false neg. SNB)
45 45 SNB 42.0%
(N = 12)
(N = 57)
40 40 41.4%
Cumulative Incidence

Cumulative Incidence
of Nodal Metastasis

of Nodal Metastasis
35 Clinically detected recurrence 35 Clinically detected recurrence
(false neg. SNB) after OBS
30 Positive (N = 31) 30
(N = 44)
25 SNB 25
(N=122) 21.9%
20 20
19.5%
15 15
Clinically detected recurrence
10 after OBS 10
5 (N = 87) 5
0 0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Years Years

No. at Risk No. at Risk

OBS 500 380 338 305 260 175 OBS 117 63 50 47 39 25
SNB 770 551 484 432 364 245 SNB 173 86 68 53 43 26

d Metastasis-Specific Survival, Thickness Melanornas

c Metastasis-Specific Survival, Intermediate-Thickness Melanornas
No. of Events/
No. of Events/ Total No. Rate (%)
Total No. Rate (%) Yr 5 Yr 10
Yr 5 Yr 10 OBS, no nodal recurrence 16/73 76.1±5.2 76.1±5.2
OBS, no nodal recurrence 48/413 92.0±1.4 86.6±1.8 OBS, nodal recurrence 23/44 53.8±7.6 45.8±7.8
OBS, nodal recurrence 49/87 57.5±5.4 41.5±5.6 SNB, true neg. 27/104 76.0±4.4 69.8±5.0
SNB, true neg. 63/612 92.3±1.1 88.0±1.4 SNB, pos. 28/57 60.8±6.6 48.0±7.0
SNB, pos. 41/122 69.8±4.4 62.1±4.8 SNB, false neg. 9/12 19.4±12.2 —
SNB, false neg. 20/31 45.2±8.9 34.4±8.7
1.00
1.00
Probability of Melanoma-Specific
Probability of Melanoma-Specific

1
0.75 1
0.75 2

2
Survival
Survival

0.50
0.50 3
3
4
4
0.25 0.25

0.00 0.00
0 2 4 6 8 10 12 0 2 4 6 8 10 12

Years Years

No. at Risk No. at Risk

OBS, no nodal 413 375 339 311 282 175 4 OBS, no nodal 73 62 52 47 43 25 0
recurrence recurrence
OBS, nodal 87 73 51 40 36 16 0 OBS, no nodal 44 32 24 21 14 9 0
recurrence recurrence
SNB, true neg. 612 570 511 448 395 243 5 SNB, true neg. 104 92 78 61 46 26 0
SNB, pos. 122 100 81 68 60 31 0 SNB, pos. 57 43 34 28 22 15 0
SNB, false neg. 31 26 15 12 10 6 0 SNB, false neg. 12 8 3 2 2 0 0

1. SNB, neg. vs. OBS, no nodal recurrence: HR, 0.89 (95% CI, 0.61–1.29); 1. SNB, neg. vs. OBS, no nodal recurrence: HR, 1.18 (95% CI, 0.63–2.18);
P=0.54 P=0.61
2. SNB, pos. vs. SNB, neg.: HR, 3.93 (95% CI, 2.65–5.83); P<0.001 2. SNB, pos. vs. SNB, neg.: HR, 2.20 (95% CI, 1.29–3.73); P=0.004
3. SNB, pos. vs. OBS, nodal recurrence: HR, 0.56 (95% CI, 0.37–0.84); 3. SNB, pos. vs. OBS, nodal recurrence: HR, 0.92 (95% CI, 0.53–1.60);
P=0.006 P=0.78
4. SNB, false neg. vs. OBS, nodal recurrence: HR, 1.15 (95% CI, 0.68–1.94); 4. SNB, false neg. vs. OBS, nodal recurrence: HR, 1.96 (95% CI, 0.90–4.27);
P=0.60 P=0.09
31 Sentinel Lymph Node Biopsy for Primary Cutaneous Malignancy 345

rate was more common in patients in the CLND arm (24% nodal burden is skewed. Both the MSLT-2 studies demon-
vs. 6%, p < 0.001). strated that the N1 disease comprises ~80% of the SN+
Similar results were seen in the DeCOG-SLT study. [30] cases, with N3 disease only 1%. Furthermore, the nodal bur-
Although the study closed early due to a reduced accrual rate den of the SN+ and the Breslow thickness of the primary are
and unexpectedly low event rate, the distant metastasis-free not independent of each other. [2, 33, 34] These issues are
survival (DMFS) was identical between the two treatment highlighted in the AJCC eighth edition classification system
groups at 3 years: 77.0 ± 5.1% in the observation group and for melanoma. [1] Inspection of the survival curves reveals
74.9 ± 5.4% in the CLND group. The same outcomes were that there is no clinically relevant survival difference between
seen for both overall survival (81.7 ± 4.9% vs. 81.2 ± 5.1%) the N1 and N2 stages (Fig. 31.3a, b), which pertains to 99%
and recurrence-free survival (67.4 ± 4.8% vs. 66.8 ± 5.9%). of the SN+ cohort. In addition, the classification system for
The node-only recurrence rate was 4.9% at 3 years with no stage III disease takes into account the Breslow thickness
significant difference seen between the two groups. A total of and ulceration status of the primary melanoma, highlighting
65.8% patients had SLNB deposits ≤1 mm and this group the competing and intertwined risks of primary tumour bur-
demonstrated a significantly improved DMFS in multivariate den and nodal tumour burden in determining overall patient
analysis (hazard ratio 2.33; 90% CI: 1.63–3.34; p < 0.0001). prognosis.

31.7  lassification of the Positive Sentinel

C 31.7.2 Rotterdam Criteria
Lymph Node Biopsy
The Rotterdam criteria uses a very simple method of measur-
The MSLT-1, MSLT-2 and DeCOG-SLT studies utilised the ing the maximum diameter of the largest deposit in the SN+s
crude binary outcome of SN+ or SN− for analysis or entry and sub-classifying the measurement into three groups;
criteria, however, centres providing SLNB services rapidly <0.1 mm, 0.1–1 mm and >1 mm. [35, 36] Using this classifi-
began to appreciate that more refined prognostic information cation, a multicentre multicentre study involving 388 SLN-­
can be obtained from a more granular analysis of the pheno- positive melanoma patients reported that among the 10% of
typical appearance of the tumour deposits within the SLNBs. patients with SLN micrometastases of less 0.1 mm, only 3%
These additional measurements were investigated in the later had additional positive non-SLNs. Furthermore, patients
surgical trials and have become incorporated into recent iter- with SN+ <0.1 mm had the same prognosis as the SN-patients.
ations of the AJCC classification system for melanoma. [36] The prognostic relevance of the Rotterdam criteria was
also validated in the DeCOG-SLT, with a diameter >1 mm
being a significant and independent predictor for worse
31.7.1 TNM (N-Stage) DMFS (HR: 2.33; 90% CI: 1.63–3.34; p < 0.0001), overall
survival (HR: 3.46; 90% CI: 1.93–6.21; p < 0.0001) and
The simplest classification system for SN+ disease is the recurrence-free survival (HR: 1.85; 90% CI: 1.34–2.55;
absolute number of nodes. This gives a relatively crude mea- p = 0.002) on multivariate analysis. [30] However, these
surement of nodal tumour burden and was the first refine- observations were not replicated in the analysis undertaken
ment to be incorporated into the AJCC classification system. in the John Wayne Cancer Institute, where the incidence of
[31, 32] Ostensibly, the classification system gives a good subclinical SN+ detected by RT-PCR was 30%. [37] This
separation of survival outcomes, with the N-stage being cat- group had a significantly worse recurrence-free and for over-
egorised as N1, N2 or N3 for involvement of 1, 2–3 or ≥4 all survival (hazard ratio: 11.42; 95% CI, 3.17–41.1;
nodes, respectively. The issue is that the distribution of the p = 0.0002) on multivariate analysis. Data from the AJCC

Fig. 31.2 Estimated 10-year incidence of nodal metastasis and melanoma-­specific survival, according to study group, melanoma thickness and
presence or absence of nodal recurrence. Panels (a) and (b) show the cumulative incidence of nodal metastasis at 10 years among patients with
intermediate-thickness melanomas and those with thick melanomas, respectively. Data are from the per-protocol analysis; patients who left the
study or were lost to follow-up were excluded. In the biopsy group, nodal recurrence in patients whose sentinel nodes were negative for tumour
(i.e. false-negative biopsy results) was assessed. Plus–minus values are means ±SE for the estimated rate of nodal metastasis. The abbreviation neg.
denotes negative, and pos. positive. Panels (c) and (d) show the probability of melanoma-specific survival (i.e. survival until death from melanoma)
among patients with intermediate-thickness melanomas and those with thick melanomas, respectively. Numbers (1–4) to the right of the survival
curves refer to the numbered comparisons below each graph. Plus–minus values are means ±SE for the estimated probability of melanoma-specific
survival. (From: Morton DL, Thompson JF, Cochran AJ, et al. Final Trial Report of Sentinel-Node Biopsy versus Nodal Observation in Melanoma.
N Engl J Med. 2014;370(7):599–609 Copyright © (2014) Massachusetts Medical Society. Reprinted with permission)
346 M. Moncrieff and H. Peach

1.0
0.8
Melanoma-Specific Survival Probability

0.6

N 5-YR 10-YR
0.4

N1 2669 82% 75%

N2 1354 76% 68%
N3 685 57% 47%
0.2
0.0

1 2 3 4 5 6 7 8 9 10
Years Since Diagnosis

b
1.0
0.8
Melanoma-Specific Survival Probability

0.6

N 5-YR 10-YR
N1a 1817 84% 75%
0.4

N1b 323 76% 71%

N1c 529 81% 75%
N2a 850 79% 71%
N2b 248 71% 71%
0.2

N2c 256 69% 59%

N3a 164 60% 46%
N3b 187 64% 57%
N3c 334 52% 43%
0.0

1 2 3 4 5 6 7 8 9 10
Years Since Diagnosis

Fig. 31.3 Kaplan–Meier melanoma-specific survival curves according 2–4 mm). From the Eighth Edition International Melanoma Database.
to (a) N categories, (b) subcategories and (c) maximum dimension of (From: Gershenwald JE, Scolyer RA, Hess KR, et al. Melanoma stag-
sentinel lymph node metastatic focus (mm) from the eighth edition ing: Evidence-based changes in the American Joint Committee on
international melanoma database. (Note that there were insufficient Cancer eighth edition cancer staging manual. CA Cancer J Clin.
data (<10 cases) to estimate 10-year melanoma-specific survival for 2017;67(6):472–492. © 2017 American Cancer Society. Reprinted with
patients who had a maximum sentinel lymph node metastatic focus of permission)
31 Sentinel Lymph Node Biopsy for Primary Cutaneous Malignancy 347

1.0
0.8
Melanoma-Specific Survival Probability

0.6

N 5-YR 10-YR
0.01-0.2 228 96% 96%
0.4

>0.2-0.2 230 93% 88%

>0.5-1.0 240 90% 90%
>1.0-2.0 210 89% 89%
0.2

>2.0-4.0 147 86% -

>4.0-15.0 383 72% 64%
>15.0 282 66% 57%
0.0

1 2 3 4 5 6 7 8 9 10
Years Since Diagnosis

Fig. 31.3 (continued)

eighth edition classification system (Fig. 31.3c) demon- Rotterdam criteria was good to excellent, the agreement for
strated that the nodal tumour burden has significant prognos- the identification of subcapsular non-capsular deposits was
tic implications. The 10-year melanoma-specific survival for only fair. [40] A separate study [39] showed that the several
patients with tumour deposits <0.2 mm and 1.01–2 mm were parameters of SN tumour burden including intranodal loca-
96% and 89%, respectively. tion, tumour-penetrating depth and maximum size provided
prognostic information, but their prognostic significance var-
ied considerably between the different centres, leading the
31.7.3 Dewar’s Microanatomic Classification authors to conclude that further studies are required, ‘to
establish a classification system for SN tumour burden that
Dewar et al. developed a separate classification system to consistently and accurately stratifies patients into meaning-
determine patients eligible for CLND. [38] The SN+ was ful prognostic groups’.
assessed according to its microanatomic location within the
node (subcapsular, combined subcapsular and parenchymal,
parenchymal, multifocal or extensive), and this was corre- 31.7.4 Extracapsular Spread
lated with the presence of involved non-sentinel nodes. They
identified that the risk of a positive non-SLN increased with Data from two UK centres showed that SN+ can be alterna-
changing microanatomical involvement of the node 0% tively stratified according to the presence of extracapsular
(subcapsular), 11.1% (combined subcapsular and parenchy- spread (ECS). [41] ECS is a well-known risk factor for mela-
mal), 18.8% (parenchymal), 36.8% (multifocal), 42.1% noma recurrence in macroscopic disease, but was an exclu-
(extensive—extracapsular or metastasis >5 mm). A separate sion criterion for both the MSLT-2 and DeCOG-SLT studies.
study combined both the Rotterdam criteria and Dewar’s Accordingly, clinical data and treatment guidance were
classification which identified low-risk patients in particular missing for this subgroup. In a sample of 265 SN+ patients,
with a SN+, where the Kaplan–Meier 5- and 10-year MSS the incidence of ECS was 16.2%. When the cohort was com-
rates were both 95% for patients with Rotterdam criteria pared with a contemporaneous sample of 250 patients with
<0.1 mm, confined to the subcapsular space. [39] Concerns palpable disease, the researchers found that, where ECS was
have been raised about the reproducibility of Dewar’s clas- absent, there was a significantly increased median number of
sification. One study identified that, whilst the interobserver nodes in the palpable disease cohort (1 node vs. 2 nodes;
agreement for quantifiable nodal measurements, such as the p < 0.0001), as with the MSLT-1 study. Whereas, the cohort
348 M. Moncrieff and H. Peach

where ECS was present, the median number of nodes was produce previous iterations of the AJCC staging system for
the same for both SN+ and the palpable disease cohorts (3, melanoma, [31, 54] that the SLNB status is the single most
both). Furthermore, when the MMS curves were compared important prognostic biomarker for cutaneous melanoma.
between the two groups, the presence of ECS in the micro- Accordingly, the current challenge remains in identifying
scopic involved nodal group, upstaged these N+ patients to subgroups of patients who are at significant risk of harbour-
the same outcome as the palpable disease group, in a similar ing a micrometastasis in their sentinel node. The current
fashion seen in the AJCC classification system curves when AJCC staging system has suggested that the threshold for
analysing the presence of ulceration of the primary tumour. offering SLNB in thinner melanomas is where the risk of
Multivariate analysis demonstrated that the presence of ECS SN+ is ≥5%, as indicated by the T1b subcategory. [1]
was the most important independent predictor of overall sur- The prognostic significance of SLNB is particularly true
vival (HR: 2.47; 95% CI 1.87–3.26; p < 0.0001). These find- for thin melanoma (0.8–1.0 mm, AJCC IB), and for interme-
ings led the researchers to hypothesise that the presence of diate thickness melanoma, (1.0–4 mm thick, AJCC IB-IIA).
ECS in the sentinel node is a biomarker for a high-risk geno- [1] For this group, the outcome of the SLNB has significant
type and led them to coin the term ‘nodal ulceration’ to management implications, since the risk of distant relapse or
describe the entity and its behaviour. Finally, the authors death for the sentinel node negative (SN−) patients is
suggest that substitution of the N-Stage for ECS status in the extremely low, such that a negative SLNB result in the 0.8–
AJCC eighth edition stage III classification system identifies 1.0 mm group indicates a downstaging to AJCC IA. [1] An
low-risk, intermediate Breslow thickness melanoma, SN+ ongoing phase III surgical RCT reported its interim results,
patients, which may guide further clinical management. [55] which demonstrated that a low-frequency, patient-­
directed follow-up programme for SN- patients had a very
high satisfaction rate (>93%) and no difference in quality of
31.7.5 Molecular Analysis life scores when compared with a nationally directed, higher
frequency follow-up regimen. Importantly, there was no dif-
The treatment of metastatic melanoma with systemic therapy ference in recurrence detection rates, which was by the
is an established standard of care for AJCC stage III and IV patient in 70% of cases, regardless of which follow-up regi-
disease. [42] Phase III clinical trials have also demonstrated men was employed. This study, and its reciprocal study in
a relapse-free survival benefit for resected stage III patients the Netherlands, [56] has important implications for service
treated with adjuvant systemic therapy, either with targeted planning and resource utilisation for cancer centres manag-
MAP-Kinase pathway inhibitors for BRAF-mutant melano- ing melanoma patients, particularly since the MSLT-1 study
mas [43] or with checkpoint inhibitors. [44, 45] The inci- demonstrated that the great majority of SLNB are negative
dence of BRAF mutations in metastatic melanoma is and the prognosis for this group is excellent.
~45–50% [46] and the incidence of NRAS mutations is ~15– The current management of SN+ in the thin and intermedi-
20%, [47] though the latter are typically classified as BRAF ate thickness group is still evolving. The MSLT-1 study dem-
wild type, despite the known true mutation. Limited studies onstrated that the risk of recurrence is increased, which would
have suggested that the presence of BRAF mutation in suggest that a regular systemic imaging schedule is indicated,
resected stage III disease is associated with a worse progno- [57] though even this is controversial for some. Whilst the
sis. [48–50] Current guidelines recommend testing to deter- MSLT-II and DeCOG-SLT studies utilised regional ultra-
mine BRAF mutation status on the SLNB metastasis, to sound as the follow-up imaging modality in their protocols,
direct future adjuvant systemic therapy. [51] Rapid and reli- the efficacy of this investigation is questionable given that
able results can be achieved with immunohistochemistry both studies also found the incidence of ‘node-only’ relapse
(IHC) markers specific for BRAF V600E mutations, with was very low. The MSLT-1 study was highly suggestive of a
molecular testing reserved for the IHC-negative samples to therapeutic benefit from removing the micrometastatic focus
ensure V600K mutations are not overlooked. [52, 53] in the sentinel node for a subgroup of patients in the interme-
diate risk group (post hoc analysis 1–4 mm). It is clear from
the MSLT-2 and DeCOG-SLT studies, that no further thera-
31.8  urrent Role of Sentinel Lymph Node
C peutic benefit can be derived by undertaking a CLND, except
Biopsy in the Management in extenuating circumstances. Accordingly, one of the mod-
of Cutaneous Melanoma ern challenges for clinicians remains to identify those SN+
patients with thin or intermediate thickness melanomas who
What information does the outcome data from these three have been adequately treated with surgery alone, whilst con-
‘landmark’ surgical trials provide for the surgeon managing sidering adjuvant systemic therapy for those patients in this
primary cutaneous melanoma? The MSLT-1 study confirmed group who are at risk of further relapse. Data from the
the findings of several studies, including the datasets used to DeCOG-SLT study would suggest that patients with tumour
31 Sentinel Lymph Node Biopsy for Primary Cutaneous Malignancy 349

deposit(s) greater than 1 mm are at increased risk of relapse, little RCT data exist on this disease; however, retrospective
though data from the AJCC eighth edition analysis would case series from large institutions suggest that either surgery
suggest that 10-year survival of patients with tumour deposits or primary radiotherapy is equally effective in the manage-
0.2–0.5 mm, 0.5–1 mm and 1—2 mm are identical. [1] An ment of both the primary disease and nodal metastases. [62–
alternative strategy taking into account the ECS status of the 64] SLNB for MCC is a controversial subject, [65] but large
SN+ has been suggested instead. [41] retrospective case series indicate that the incidence of SN+
The role of SLNB for thick and high-risk melanomas disease is common (~20–40%) and the finding of a SN+
would appear to be a more straightforward process of identi- helps to guide adjuvant therapy in the context of the patient’s
fying patients to be referred for adjuvant systemic therapy. underlying health status. On a technical level, one would
This treatment strategy for this group may also evolve, how- caution the accuracy of the SLNB in the head and neck
ever, based on the future outcome of data from two phase III region and radiological screening with PET-CT is recom-
adjuvant systemic immunotherapy trials for AJCC IIB-IIC mended for surveillance. [66] Furthermore, consideration
(eighth edition) patients (‘KeyNote-716’, ClinicalTrials.gov should be given to omitting intraoperative blue dye injection,
Identifier: NCT03553836 and ‘CheckMate-76 K’, which causes highly visible dermal staining, if treatment of
ClinicalTrials.gov Identifier: NCT04099251). The MSLT-1 the primary site with radiotherapy alone is likely.
study showed no survival benefit from removing the positive
sentinel node(s) for patients with primary melanomas thicker
than 3.5 mm. The likely explanation is that the SN+ nodal 31.9.2 Cutaneous SCC
status is a biomarker for systemic disease in this subgroup of
patients, particularly if this is associated with ECS [41]. Cutaneous squamous cell carcinoma (cSCC) is the second
Accordingly, the emphasis is placed on identifying all patients most common skin malignancy. It routinely affects the
at higher risk of metastatic disease, and the role of SLNB has elderly and most commonly develops on the head and neck
become one that now influences access to adjuvant systemic region. Multiple primaries are very common and rapid
therapy. [51, 58] Furthermore, the last two iterations of the growth of the primary resulting in large lesions at presenta-
AJCC systems classified the presence of microsatellites at the tion are frequently seen. Unlike MCC and melanoma, the
primary site as stage IIIB-C disease, an indication for direct overall incidence of lymph node metastases is low, though
referral for adjuvant systemic therapy. Accordingly, it is hard nodal metastases are associated with a poor prognosis. When
to justify the use of SLNB for this group of patients. targeting high-risk head and neck cSCC, the incidence of
SN+ disease was ~15% in two single-centre studies. [67, 68]
The incidence of false negativity was 22–45% indicating the
31.9  entinel Node Biopsy for Other Skin
S limitations of SLNB accuracy in the head and neck region.
Cancers Given the limitations of accuracy of SLNB for cSCC, the
lack of high-level evidence for the utility of the procedure,
The overwhelming majority of the literature describing SLNB and the lack of consensus regarding the management of a
for cutaneous malignancy is dedicated to melanoma. The SN+ and the lack of consensus regarding the definition of a
main reasons that SLNB is an effective treatment strategy for high-risk primary, [69] there are no current standardised
melanoma are that the primary is discrete measuring <10 mm guidelines for SLNB for cSCC. [70]
in diameter, it is a common tumour, nearly 50% of patients
diagnosed are <65 years old, and it is very unusual for patients
to have more than one melanoma in their lifetime. 31.10 Conclusion

High-level evidence and consistent international experience

31.9.1 Merkel Cell Carcinoma has firmly established the role of SLNB for the staging and
management of primary cutaneous melanoma, though the
Merkel cell carcinoma (MCC) is a very rare tumour that usu- paradigm has permanently shifted from a role of identifying
ally develops on the head and neck region. It commonly patients suitable for regional surgery to one of identifying
affects the elderly with a mortality rate of 40%. [59] Rapid patients eligible for adjuvant systemic therapy. SLNB for
growth of the primary is common, with the tumour usually Merkel cell carcinoma currently has a role for directing
>15 mm in diameter at diagnosis. [60] UV-B radiation expo- regional therapy to the draining nodal field, though high-­
sure and polyomavirus virus infection are proposed as the level evidence for the efficacy of the procedure is currently
two major aetiologies. [59] Unlike melanoma, MCC is a lacking. Evidence is lacking for the utility of SLNB for cuta-
radiosensitive tumour, which affords the advantage of further neous SCC, accordingly it is not routinely recommended for
treatment options in this often-frail demographic. [61] Very this group of patients at present.
350 M. Moncrieff and H. Peach

13. Thompson JF, McCarthy WH, Bosch CM, et al. Sentinel lymph
Open Questions node status as an indicator of the presence of metastatic melanoma
in regional lymph nodes. Melanoma Res. 1995;5(4):255–60.
• Can methods of patient selection for sentinel lymph 14. Reintgen D, Cruse CW, Wells K, et al. The orderly progression of
node biopsy in melanoma be refined with standard melanoma nodal metastases. Ann Surg. 1994;220(6):759–67.
clinicopathologic variables? 15. Albertini JJ, Cruse CW, Rapaport D, et al. Intraoperative radio-­
• Can methods of patient selection for patients with lympho-­scintigraphy improves sentinel lymph node identification
for patients with melanoma. Ann Surg. 1996;223(2):217–24.
sentinel node melanoma metastases be refined to 16. Morton DL, Thompson JF, Essner R, et al. Validation of the
identity low risk patients who can be observed and accuracy of intraoperative lymphatic mapping and sentinel
high risk patients who may benefit from adjuvant lymphadenectomy for early-stage melanoma: a multicenter trial.
systemic therapy? Multicenter Selective Lymphadenectomy Trial Group. Ann Surg.
1999;230(4):453–63; discussion 463–465.
• What is the role of sentinel node lymph node biopsy 17. Wong SL, Balch CM, Hurley P, et al. Sentinel lymph node biopsy
for patients with primary Merkel cell carcinoma? for melanoma: American Society of Clinical Oncology and Society
• What is the role of sentinel lymph node biopsy in of Surgical Oncology joint clinical practice guideline. J Clin Oncol.
cutaneous squamous cell carcinoma and in which 2012;30(23):2912–8.
18. Bluemel C, Herrmann K, Giammarile F, et al. EANM practice
patients? guidelines for lymphoscintigraphy and sentinel lymph node biopsy
in melanoma. Eur J Nucl Med Mol Imaging. 2015;42(11):1750–66.
19. Quartuccio N, Siracusa M, Pappalardo M, Arnone A, Arnone
G. Sentinel node identification in melanoma: current clinical
impact, new emerging SPECT radiotracers and technological
References advancements. An update of the last decade. Curr Radiopharm.
2020;13(1):32–41.
1. Gershenwald JE, Scolyer RA, Hess KR, et al. Melanoma stag- 20. Jimenez-Heffernan A, Ellmann A, Sado H, et al. Results of a pro-
ing: evidence-based changes in the American Joint Committee on spective multicenter International Atomic Energy Agency sentinel
Cancer eighth edition cancer staging manual. CA Cancer J Clin. node trial on the value of SPECT/CT over planar imaging in various
2017;67(6):472–92. malignancies. J Nucl Med. 2015;56(9):1338–44.
2. Sondak VK, Taylor JMG, Sabel MS, et al. Mitotic rate and younger 21. Moncrieff MD, O’Leary FM, Beadsmoore CJ, et al. Effect of
age are predictors of sentinel lymph node positivity: lessons learned delay between nuclear medicine scanning and sentinel node biopsy
from the generation of a probabilistic model. Ann Surg Oncol. on outcome in patients with cutaneous melanoma. Br J Surg.
2004;11(3):247–58. 2020;107(6):669–76.
3. Balch CM. Surgical management of regional lymph nodes in cuta- 22. Sondak VK, King DW, Zager JS, et al. Combined analysis of phase
neous melanoma. J Am Acad Dermatol. 1980;3(5):511–24. III trials evaluating [99mTc]tilmanocept and vital blue dye for iden-
4. Veronesi U, Adamus J, Bandiera DC, et al. Stage I melanoma of tification of sentinel lymph nodes in clinically node-negative cuta-
the limbs. Immediate versus delayed node dissection. Tumori. neous melanoma. Ann Surg Oncol. 2013;20(2):680–8.
1980;66(3):373–96. 23. Knackstedt R, Couto RA, Ko J, Cakmakoglu C, Wu D, Gastman
5. Veronesi U, Adamus J, Bandiera DC, et al. Inefficacy of immediate B. Indocyanine green fluorescence imaging with lymphoscintigra-
node dissection in stage 1 melanoma of the limbs. N Engl J Med. phy for sentinel node biopsy in melanoma: increasing the sentinel
1977;297(12):627–30. lymph node-positive rate. Ann Surg Oncol. 2019;26(11):3550–60.
6. Sim FH, Taylor WF, Pritchard DJ, Soule EH. Lymphadenectomy 24. Anninga B, White SH, Moncrieff M, et al. Magnetic technique for
in the management of stage I malignant melanoma: a prospective sentinel lymph node biopsy in melanoma: the MELAMAG trial.
randomized study. Mayo Clin Proc. 1986;61(9):697–705. Ann Surg Oncol. 2016;23(6):2070–8.
7. Cascinelli N, Morabito A, Santinami M, MacKie RM, Belli 25. Morton DL, Wanek L, Nizze JA, Elashoff RM, Wong JH. Improved
F. Immediate or delayed dissection of regional nodes in patients long-term survival after lymphadenectomy of melanoma metastatic
with melanoma of the trunk: a randomised trial. WHO Melanoma to regional nodes. Analysis of prognostic factors in 1134 patients
Programme. Lancet. 1998;351(9105):793–6. from the John Wayne Cancer Clinic. Ann Surg. 1991;214(4):491–9;
8. Essner R, Conforti A, Kelley MC, et al. Efficacy of lymphatic map- discussion 499–501.
ping, sentinel lymphadenectomy, and selective complete lymph 26. Morton DL, Thompson JF, Cochran AJ, et al. Final trial report of
node dissection as a therapeutic procedure for early-stage mela- sentinel-node biopsy versus nodal observation in melanoma. N
noma. Ann Surg Oncol. 1999;6(5):442–9. Engl J Med. 2014;370(7):599–609.
9. Balch CM, Soong S-J, Bartolucci AA, et al. Efficacy of an elective 27. Bello DM, Faries MB. The landmark series: MSLT-1, MSLT-2
regional lymph node dissection of 1 to 4 mm thick melanomas for and DeCOG (management of lymph nodes). Ann Surg Oncol.
patients 60 years of age and younger. Ann Surg. 1996;224(3):255. 2020;27(1):15–21.
10. Morton DL, Wen DR, Wong JH, et al. Technical details of intraop- 28. Faries MB, Thompson JF, Cochran A, et al. The impact on mor-
erative lymphatic mapping for early stage melanoma. Arch Surg. bidity and length of stay of early versus delayed complete lymph-
1992;127(4):392–9. adenectomy in melanoma: results of the multicenter selective
11. Wong JH, Cagle LA, Morton DL. Lymphatic drainage of lymphadenectomy trial (I). Ann Surg Oncol. 2010;17(12):3324–9.
skin to a sentinel lymph node in a feline model. Ann Surg. 29. Faries MB, Thompson JF, Cochran AJ, et al. Completion dissection
1991;214(5):637–41. or observation for sentinel-node metastasis in melanoma. N Engl J
12. Alex JC, Weaver DL, Fairbank JT, Rankin BS, Krag DN. Gamma-­ Med. 2017;376(23):2211–22.
probe-­guided lymph node localization in malignant melanoma. 30. Leiter U, Stadler R, Mauch C, et al. Complete lymph node dis-
Surg Oncol. 1993;2(5):303–8. section versus no dissection in patients with sentinel lymph node
31 Sentinel Lymph Node Biopsy for Primary Cutaneous Malignancy 351

biopsy positive melanoma (DeCOG-SLT): a multicentre, ran- anoma: implications for melanoma staging and adjuvant therapy.
domised, phase 3 trial. Lancet Oncol. 2016;17(6):757–67. Eur J Cancer. 2014;50(15):2668–76.
31. Balch CM, Soong S-J, Gershenwald JE, et al. Prognostic factors 49. Picard M, Pham Dang N, D’Incan M, et al. Is BRAF a prognos-
analysis of 17,600 melanoma patients: validation of the American tic factor in stage III skin melanoma? A retrospective study of
Joint Committee on Cancer melanoma staging system. J Clin 72 patients after positive sentinel lymph node dissection. Br J
Oncol. 2001;19(16):3622–34. Dermatol. 2014;171(1):108–14.
32. Balch CM, Buzaid AC, Soong SJ, et al. Final version of the 50. Heppt MV, Siepmann T, Engel J, et al. Prognostic significance of
American Joint Committee on cancer staging system for cutaneous BRAF and NRAS mutations in melanoma: a German study from
melanoma. J Clin Oncol. 2001;19(16):3635–48. routine care. BMC Cancer. 2017;17(1):536.
33. Sabel MS, Griffith K, Sondak VK, et al. Predictors of nonsentinel 51. Coit DG, Thompson JA, Albertini MR, et al. Cutaneous melanoma,
lymph node positivity in patients with a positive sentinel node for version 2.2019, NCCN clinical practice guidelines in oncology. J
melanoma. J Am Coll Surg. 2005;201(1):37–47. Natl Compr Cancer Netw. 2019;17(4):367–402.
34. Quaglino P, Ribero S, Osella-Abate S, et al. Clinico-pathologic 52. Lo MCI, Paterson A, Maraka J, et al. A UK feasibility and valida-
features of primary melanoma and sentinel lymph node predictive tion study of the VE1 monoclonal antibody immunohistochemistry
for non-sentinel lymph node involvement and overall survival in stain for BRAF-V600E mutations in metastatic melanoma. Br J
melanoma patients: a single centre observational cohort study. Surg Cancer. 2016;115(2):223–7.
Oncol. 2011;20(4):259–64. 53. Long GV, Wilmott JS, Capper D, et al. Immunohistochemistry is
35. van Akkooi ACJ, Nowecki ZI, Voit C, et al. Sentinel node tumor highly sensitive and specific for the detection of V600E BRAF
burden according to the Rotterdam criteria is the most important mutation in melanoma. Am J Surg Pathol. 2013;37(1):61–5.
prognostic factor for survival in melanoma patients: a multicenter 54. Balch CM, Gershenwald JE, Soong S-J, et al. Final version of
study in 388 patients with positive sentinel nodes. Ann Surg. 2009 AJCC melanoma staging and classification. J Clin Oncol.
2008;248(6):949–55. 2009;27(36):6199–206.
36. van Akkooi ACJ, de Wilt JHW, Verhoef C, et al. Clinical relevance of 55. Moncrieff MD, Underwood B, Garioch JJ, et al. The MelFo study
melanoma micrometastases (<0.1 mm) in sentinel nodes: are these UK: effects of a reduced-frequency, stage-adjusted follow-up
nodes to be considered negative? Ann Oncol. 2006;17(10):1578– schedule for cutaneous melanoma 1B to 2C patients after 3-years.
85. https://doi.org/10.1093/annonc/mdl176. Ann Surg Oncol. 2020;27(11):4109–19. https://doi.org/10.1245/
37. Takeuchi H, Morton DL, Kuo C, et al. Prognostic significance of s10434-­020-­08758-­2.
molecular upstaging of paraffin-embedded sentinel lymph nodes in 56. Deckers EA, Hoekstra-Weebers JEHM, Damude S, et al. The
melanoma patients. J Clin Oncol. 2004;22(13):2671–80. MELFO study: a multicenter, prospective, randomized clinical
38. Dewar DJ, Newell B, Green MA, Topping AP, Powell BWEM, trial on the effects of a reduced stage-adjusted follow-up sched-
Cook MG. The microanatomic location of metastatic melanoma ule on cutaneous melanoma IB-IIC patients-results after 3 years.
in sentinel lymph nodes predicts nonsentinel lymph node involve- Ann Surg Oncol. 2020;27(5):1407–17. https://doi.org/10.1245/
ment. J Clin Oncol. 2004;22(16):3345–9. s10434-­019-­07825-­7.
39. van der Ploeg APT, van Akkooi ACJ, Rutkowski P, et al. Prognosis 57. Stodell M, Thompson JF, Emmett L, Uren RF, Kapoor R, Saw
in patients with sentinel node-positive melanoma is accurately RPM. Melanoma patient imaging in the era of effective systemic
defined by the combined Rotterdam tumor load and Dewar topog- therapies. Eur J Surg Oncol. 2017;43(8):1517–27.
raphy criteria. J Clin Oncol. 2011;29(16):2206–14. 58. Peach H, Board R, Cook M, et al. Current role of sentinel lymph
40. Murali R, Cochran AJ, Cook MG, et al. Interobserver reproducibil- node biopsy in the management of cutaneous melanoma: a UK
ity of histologic parameters of melanoma deposits in sentinel lymph consensus statement. J Plast Reconstr Aesthet Surg. 2020;73(1):
nodes: implications for management of patients with melanoma. 36–42.
Cancer. 2009;115(21):5026–37. 59. Becker JC, Stang A, Hausen AZ, et al. Epidemiology, biology and
41. Lo M, Robinson A, Wade R, et al. Extracapsular spread in mela- therapy of Merkel cell carcinoma: conclusions from the EU project
noma lymphadenopathy: prognostic implications, classification, IMMOMEC. Cancer Immunol Immunother. 2018;67(3):341–51.
and management. Ann Surg Oncol. 2021;28(3):1642–53. https:// 60. Smith FO, Yue B, Marzban SS, et al. Both tumor depth and diam-
doi.org/10.1245/s10434-­020-­09099-­w. eter are predictive of sentinel lymph node status and survival in
42. Baetz TD, Fletcher GG, Knight G, et al. Systemic adjuvant therapy Merkel cell carcinoma. Cancer. 2015;121(18):3252–60.
for adult patients at high risk for recurrent melanoma: a systematic 61. Strom T, Carr M, Zager JS, et al. Radiation therapy is associated
review. Cancer Treat Rev. 2020;87:102032. with improved outcomes in Merkel cell carcinoma. Ann Surg
43. Dummer R, Hauschild A, Santinami M, et al. Five-year analysis of Oncol. 2016;23(11):3572–8.
adjuvant Dabrafenib plus Trametinib in stage III melanoma. N Engl 62. Gunaratne DA, Howle JR, Veness MJ. Sentinel lymph node
J Med. 2020;383(12):1139–48. biopsy in Merkel cell carcinoma: a 15-year institutional experi-
44. Weber J, Mandala M, Del Vecchio M, et al. Adjuvant Nivolumab ver- ence and statistical analysis of 721 reported cases. Br J Dermatol.
sus Ipilimumab in resected stage III or IV melanoma. N Engl J Med. 2016;174(2):273–81.
2017;377:1824–35. https://doi.org/10.1056/NEJMoa1709030. 63. Perez MC, Oliver DE, Weitman ES, et al. Management of senti-
45. Eggermont AMM, Blank CU, Mandala M, et al. Adjuvant nel lymph node metastasis in Merkel cell carcinoma: comple-
Pembrolizumab versus placebo in resected stage III melanoma. N tion lymphadenectomy, radiation, or both? Ann Surg Oncol.
Engl J Med. 2018;378(19):1789–801. 2019;26(2):379–85.
46. Greaves WO, Verma S, Patel KP, et al. Frequency and spectrum of 64. Lee JS, Durham AB, Bichakjian CK, et al. Completion lymph
BRAF mutations in a retrospective, single-institution study of 1112 node dissection or radiation therapy for sentinel node metas-
cases of melanoma. J Mol Diagn. 2013;15(2):220–6. tasis in Merkel cell carcinoma. Ann Surg Oncol. 2019;26(2):
47. Adler NR, Wolfe R, Kelly JW, et al. Tumour mutation status and 386–94.
sites of metastasis in patients with cutaneous melanoma. Br J 65. Hruby G, Guminski A, Thompson JF. Management of regional
Cancer. 2017;117(7):1026–35. lymph nodes in patients with Merkel cell carcinoma following
48. Barbour AP, Tang YH, Armour N, et al. BRAF mutation status is an a positive sentinel node biopsy: less may be more, but is either
independent prognostic factor for resected stage IIIB and IIIC mel- enough? Ann Surg Oncol. 2019;26(2):315–7.
352 M. Moncrieff and H. Peach

66. Bichakjian CK, Olencki T, Aasi SZ, et al. Merkel cell carcinoma, and neck. JAMA Otolaryngol Head Neck Surg. 2016;142(12):
version 1.2018, NCCN clinical practice guidelines in oncology. J 1171–6.
Natl Compr Cancer Netw. 2018;16(6):742–74. 69. Ahadiat O, Higgins S, Sutton A, Ly A, Wysong A. SLNB in cutane-
67. Gore SM, Shaw D, Martin RCW, et al. Prospective study of sentinel ous SCC: a review of the current state of literature and the direction
node biopsy for high-risk cutaneous squamous cell carcinoma of for the future. J Surg Oncol. 2017;116(3):344–50.
the head and neck. Head Neck. 2016;38(Suppl 1):E884–9. 70. Allen JE, Stolle LB. Utility of sentinel node biopsy in patients with
68. Durham AB, Lowe L, Malloy KM, et al. Sentinel lymph node high-risk cutaneous squamous cell carcinoma. Eur J Surg Oncol.
biopsy for cutaneous squamous cell carcinoma on the head 2015;41(2):197–200.
 Sentinel Lymph Node Biopsy in Breast
Cancer 32
Emily Siegel, John Kiluk, Armando Giuliano,
and Brian Czerniecki

Abstract 32.1 Introduction

Lymphatic metastasis is a strong predictor of poor out-
come in breast cancer patients. Most often, lymphatic Lymphatic metastasis is an important prognostic indicator
metastasis occurs through the axillary lymph nodes. for outcomes in breast cancer. The evaluation of lymph nodes
While axillary dissection (ALND) was previously rou- and axillary staging is a vital tool in the treatment of patients
tine, sentinel lymph node biopsy (SLNB) has become the with breast cancer and can be accomplished in multiple
standard of care for clinically and pathologically node-­ ways. The surgical assessment of lymphatic metastasis for
negative breast cancer given its reduced morbidity. SLNB breast cancer has evolved significantly since first introduced
has replaced ALND for node-positive disease in patients in the 1970s and can now be achieved in many circumstances
with small tumors undergoing breast conservation ther- with axillary sentinel lymph node biopsy.
apy and is rapidly evolving to replace ALND in patients
who are converted from node positive to node negative
with the use of neoadjuvant chemotherapy. Controversy 32.2  istory of SLNB and Technical
H
remains about whether axillary staging, and therefore Considerations
SLNB may be omitted in a certain subset of patients, such
as the elderly and those with exceptional responses to Lymphatic metastasis is a strong prognostic indicator of poor
neoadjuvant therapy. outcome in breast cancer. Nodal metastasis occurs through a
variety of mechanisms and in a stepwise pattern which
begins with tumor lymphangiogenesis. Cancer cells then
invade from the tumor to lymphatic vessels at the site of pri-
Learning Objectives mary disease, from which they travel to the sentinel lymph
• To be able to describe the effectiveness of SLNB to node. There, the cancer cells survive through immunogenic-
accurately stage the axilla. ity and nodal immunosuppression [1].
• To understand the utility of SLNB in lymph node-­ Given its prognostic indication, axillary nodal status is an
positive breast cancer patients. important factor for guiding adjuvant treatment decisions.
• To understand the current controversies in the scope Surgical identification of axillary lymph nodes has long been
of SLNB compared to ALND. the gold standard of axillary staging, though the scope of
axillary surgery in breast cancer has changed dramatically
over the past century. In 1977, Morton et al. used cutaneous
lymphoscintigraphy to identify lymphatic drainage to nodal
basins of melanomas [2]. They also showed that preoperative
E. Siegel · J. Kiluk · B. Czerniecki (*) lymphoscintigraphy could demonstrate a single lymph node
Breast Department, H. Lee Moffitt Cancer Center,
receiving drainage [3]. The ability to safely and accurately
Tampa, FL, USA
e-mail: [email protected]; [email protected]; perform sentinel lymph node biopsy (SLNB) in breast can-
[email protected] cer was then demonstrated in 1994 by Giuliano and Morton
A. Giuliano [4]. The initial studies defined the technique and demon-
Department of Surgery, Cedars-Sinai Medical Center, strated a significant learning curve, with the final patients
Los Angeles, CA, USA enrolled exhibiting a 100% predictive rate of negative senti-
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 353
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_32
354 E. Siegel et al.

nel nodes. A prospective trial by the same group then demon- performed using a combination of radiopharmaceutical and
strated that in addition to technical feasibility, SLNB was blue dye in 79.4% of cases, blue dye alone in 14.8% cases,
safe and highly accurate [5]. Since that time, multiple large-­ and radiopharmaceutical alone in 5.7% of cases, with an
scale studies have validated the technique and feasibility of overall success rate of 98.7%. The type of dye or radioactive
SLNB [6, 7]. The low rate of locoregional tumor recurrence isotope did not affect failure rates as unsuccessful identifica-
in patients with negative SLN made a tremendous impact on tion with blue dye only (1.4%), radiopharmaceutical only
axillary surgery, further promoting the use of SLNB [8, 9]. (2.3%), and combined (1.2%) were not statistically different
To improve the learning curve and increase identification, (p = 0.28). Further evaluation of isosulfan blue compared to
Cox et al. demonstrated the technique of combination blue-­ methylene blue was performed and no significant differences
dye and radiocolloid [14]. In early single-centered trials, the were seen in identification rates and adverse reactions [16].
combination technique showed a quick learning curve and After proving technical feasibility, sentinel lymph node
low false-negative rate [15]. Studies of particle kinetics and biopsy became the preferred nodal staging procedure for
comparison studies of different commercially available trac- patients undergoing surgery for breast cancer. SLNB is pre-
ers including albumin, sulfur colloid, and albumin colloid, ferred over complete axillary lymph node dissections
were completed to classify the best technique [10, 11]. The (ALND) given decreased rates of morbidity. Initial reports in
mechanism by which these tracers are retained in sentinel 2000 on small numbers of patients suggested significantly
lymph nodes was demonstrated by Faries et al. Their group higher patient-reported rates of lymphedema and pain in
demonstrated that antigen-presenting cells rapidly take up those who received ALND [20] compared to those treated
radiotracer [12]. These cells, expressing major histocompat- with SLNB. Further retrospective data confirmed these find-
ibility class II proteins, then appear in sentinel lymph nodes ings, suggesting that patients with SLNB had 3.2-fold lower
with increasing amounts over time [12]. Since that time, risk of pain, five-fold lower risk of lymphedema, and 7.7-­
other receptor-binding radiopharmaceuticals have appeared fold lower risk of numbness [21]. Prospective multicenter
on the market. Lymphoseek, specifically designed for SLN studies followed. One which included 395 patients showed
mapping, was compared to sulfur colloid, and showed equal significantly higher arm volume (and therefore potential for
uptake in sentinel nodes with faster injection site clearance, lymphedema) and increased shoulder and arm pain in
though it is not in widespread use given the need for a 2-day patients with ALND compared to SLNB [22]. The multi-
protocol and cost associated [13]. center randomized National Surgical Adjuvant Breast and
Attention then turned to the location of injection. Initial Bowel Project (NSABP) B-32 trial randomized 5611 patho-
studies advocated for the use of peritumoral injection sites logically node-negative patients to either SLNB alone or
[17, 18]. However, this technique requires preoperative SLNB and completed axillary dissection. Patients in both
imaging of the breast and injection under radioguidance. In groups had similar disease-free and overall survival.
addition, injections for tumors in the upper outer quadrant Furthermore, there was no difference in axillary or regional-­
create difficulty in determining which activity is in the nodes node recurrences between groups [23]. Given the utility in
versus in the tumor bed, called “shine through.” Therefore, node-negative patients, interest in SLNB for clinically node-­
interest in periareolar injections surfaced. In 2004, the first negative but pathologically node-positive patients grew.
prospective multicenter trial using periareolar injections was
published. In this study, patients underwent SLNB and com-
pleted axillary dissection. The SLN was successfully identi- 32.2.2 A
 dequacy of Sentinel Node Biopsy
fied in 99.3% of patients who received subareolar injections in Node-Positive Patients
and 95.6% of patients who underwent periareolar injections,
significantly higher than the 91.1% of patients who under- The landmark ACOSOG Z0011 trial aimed to determine if
went peritumoral injection [19]. Thus, the current technique SLNB could replace ALND for node-positive patients under-
to identify axillary SLN became periareolar or subareolar going breast conservation therapy (BCT). Inclusion criteria
injections. were female patients with T1–T2 invasive breast cancer
without palpable adenopathy and 1–2 positive SLN identi-
fied during surgery. Patients were randomized to either
32.2.1 Increasing Use of SLNB in Clinically SLNB alone or ALND with >10 nodes removed at the time
Node-Negative Patients of surgery. Importantly, all patients underwent adjuvant radi-
ation treatment. With a median follow-up of 6.3 years at the
The American College of Surgeons Oncology Groups time of initial publication, there was no statistically signifi-
(ACOSOG) Z0010 trial was then completed, a multi-­ cant difference in 5-year overall survival (92.5% vs. 91.8%
institutional study evaluating both the accuracy and prognos- in ALND and SLNB groups, respectively), with a low axil-
tic value of SLNB. In this study, identification of SLN was lary recurrence rate of fewer than 1% in both groups. There
32 Sentinel Lymph Node Biopsy in Breast Cancer 355

was also no difference in outcomes between patients which was initially concerning, with thoughts that lymphatic map-
were estrogen receptor (ER) +/Progesterone Receptor (PR)+, ping would be disrupted by excessive fibrosis of tumor lym-
and those who were ER−/PR− [24]. Criticism of the phatics or blocking of lymphatic channels with cellular
ACOSOG z0011 trial focused on its short-term follow-up, material or necrotic material created during chemotherapeu-
generally low-risk patients (given high percentage of ER+/ tic destruction of the primary tumor and/or axillary metasta-
PR+), and lack of blinding of sentinel node status to radia- sis. Indeed, initial studies of SLNB after NAC had high rates
tion oncologists. In response, the long-term 10-year follow- of failure to identify the node, up to 20%, with up to 16%
up was published in 2017. With a median follow-up of
­ false-negative rates [28, 29]. However, these initial studies
9.3 years, SLNB was non-inferior to ALND for overall sur- included patients with initial bulky axillary metastasis and
vival (survival rate 86.3% SLNB alone vs. 83.6% ALND, generally had a wide range of inclusion criteria. Patients with
non-inferiority p = 0.02). The 10-year disease-free survival clinically node-negative axilla prior to initiating NAC had
was not statistically significantly different between groups improved ability to identify SLN. Initial retrospective data
(80.2% SLNB alone vs. 78.2% ALND, p = 0.32). A subset showed comparable rates of identification (IR) and false neg-
analysis was performed on those patients with specific radia- ativity (FNR) in patients undergoing NAC as those who had
tion field data available. Two hundred and twenty-eight surgery upfront, with IR of 99% and 97% and FNR of 4%
patients had this data available, which showed no difference and 6% [30]. Improved identification rates were seen in
in disease-free survival, overall survival, or locoregional those patients whose lymphatics were mapped using dual
recurrence for those treated with nodal-field irradiation com- tracer [31]. These improved IR and FNR rates were validated
pared with those who did not receive irradiation [25]. in meta-analyses [32].
ACOSOG Z0011 allowed SLNB to replace routine ALND in The rates of negative SLNB were consistently higher in
node-positive patients with T1–T2 invasive breast cancer, no patients who received NAC in both single institution and
palpable axillary lymphadenopathy undergoing breast con- multi-institutional studies [30, 33, 34]. Increasing node-­
servation therapy. The ACOSOG Z0011 trial was further negativity, the ability to potentially downstage the axilla, and
validated with the “Radiotherapy or surgery of the axilla therefore avoid ALND and/or nodal radiation led to increase
after a positive sentinel node in breast cancer (EORTC use of SLNB in patients who were initially node positive.
10981-22023 AMAROS)” trial which again evaluated sur- These attempts were supported by low rates of axillary recur-
gery or radiotherapy of the axilla after a positive SLN. Unlike rence in those patients who achieved nodal pCR [35]. Many
the ACOSOG Z0011 trial, patients undergoing mastectomy prospective multi-institutional studies then proceeded to
were included in the AMAROS trial. There was no differ- examine the ability to accurately identify SLN in patients
ence in axillary recurrence in patients who were randomized who were clinically node positive prior to NAC. The
to axillary dissection compared to those in the axillary radio- ACOSOG 1071 trial was a phase II clinical trial which
therapy group (0.43% vs. 1.19%, respectively), though over- occurred in centers across the United States. Women with
all number of patients undergoing mastectomy was small biopsy confirmed axillary nodal disease undergoing NAC
[26]. Whether or not patients who are node positive and were included. Patients with a previous history of axillary
treated with mastectomy can spare full ALND or can undergo surgery or prior SLNB were excluded. Those enrolled in
radiation therapy remains unstudied in large randomized ACOSOG 1071 underwent lymphatic mapping with dual
groups. tracer following the completion of NAC, and all SLN were
excised prior to proceeding with ALND. The FNR in patients
with two or more nodes removed was 13%, while the FNR
32.3  LNB in Patients Receiving
S with three or more nodes removed fell to 9%, however, only
Neoadjuvant Chemotherapy a small percentage (56%) of patients in the trial met this cri-
terion [36]. The SENTInel NeoAdjuvant (SENTINA) trial
Further attempts at integrating SLNB into practice were then was another multicenter prospective study investigating
made. As neoadjuvant chemotherapy (NAC) became more patients undergoing NAC. The primary endpoint of the study
widely adopted in the treatment of breast cancer, timing and was the accuracy and FNR of patients converting from posi-
technical feasibility of SLNB became widely debated. It was tive SLN to negative SLN after NAC. FNR of these patients
initially suggested that operative SLNB be performed prior was over 14% and the detection rate was only 80% [37].
to administration of neoadjuvant chemotherapy in order to Further attempts at accurately detecting SLN after NAC con-
accurately stage the axilla and determine the need for nodal tinued. The Sentinel Node Biopsy Following Neoadjuvant
radiation [27]. However, if axillary metastasis were eradi- Chemotherapy (SN FNAC) study also enrolled patients with
cated by neoadjuvant chemotherapy, performing SLNB after biopsy-proven node-positive patients, attempting to reduce
the chemotherapy regimen could eliminate the need for a full FNR to under 10%. However, this study was stopped early
ALND. The technical ability to perform SLNB after NAC when the ACOSOG Z1071 and other trials failed to meet
356 E. Siegel et al.

their criteria of FNR <10%. Prior to ending, the SN FNAC negative at surgery are randomized to either receive nodal
showed an IR of 88% and FNR of 8% [38]. radiation or not. The use of targeted dissection and subse-
Given these unacceptably high FNR, further attempts to quent results of the NSABP B-51 trial also have the potential
improve SLNB after NAC for patients initially node positive for dramatic shifts in the treatment of breast cancer patients
ensued. The addition of placing a clip in the biopsied node after neoadjuvant therapy.
and subsequently performing targeted SLNB was examined.
Initial retrospective reviews showed IR of 92% when adding
a marker and wire-localization of the biopsied axillary node 32.4 When to Omit SLNB
to the operation [39]. A prospective trial performed by MD
Anderson Cancer Center enrolled patients with biopsy con- Given a still-present 1–3% risk of lymphedema with SLNB,
firmed axillary metastasis who underwent clip placement in and a potential for axillary nodal status to not affect treat-
that node. After NAC, these patients underwent localization ment in certain groups, attempts have been made to omit
with placement of an iodine-125 radioactive seed. FNR in SLNB certain clinical situations. This is true in patients
these patients was significantly lower, 4.2% for the clipped undergoing NAC, specifically those with triple negative (TN)
node. The FNR of the combination of SLNB and targeted or HER2-positive tumors with clinical complete response to
dissection of the clipped node reduced FNR to 1.4%. In this NAC. A retrospective analysis of 646 T1–4, N0–2 patients
study, 23% of patients were unable to have their clipped node undergoing NAC showed that HER2 positive and TN tumors
identified. The only factor which was associated with this who started as clinically node negative remain node negative
discordance was having >3 abnormal nodes on initial ultra- if they achieve an in-breast pathologic complete response
sound [40, 41]. (pCR) [43]. A prospective study from the group at MD
The long-term oncologic safety and axillary recurrence Anderson also examined the role of SLNB in patients who
rates in patients who undergo targeted axillary dissection were so-called exemplary responders to NAC. A total of 527
remain to be seen. Recent data from a single institution ret- patients were included, 290 of whom were node-negative by
rospective study comparing patients who were initially clini- ultrasound prior to NAC. Of those, 116 (40%) had a breast
cally node negative to those who were clinically node pCR, 100% of whom also had no evidence of axillary lymph
positive with negative SLNB treated with ALND, and to node metastasis. Those with initial axillary metastasis
those who were node positive with negative SLNB without showed an in-breast pCR rate of 32% (N = 77), of whom
ALND showed an IR of 98.3%. There was no significant dif- 89.6% also had negative nodal disease [44]. Given these
ference in axillary recurrence-free survival (p = 0.12), studies, the authors suggest that patients may be able to be
disease-­ free survival (p = 0.27), and overall survival spared SLNB if they are initially node negative and have an
(p = 0.15) between groups, suggesting oncologic safety. in-breast pathologic complete response. However, there is no
Though there remains a paucity of prospective data on axil- strong or randomized data to suggest that an in-breast com-
lary or distant recurrence in those with negative SLNB (and plete pathologic response is associated with a nodal com-
no subsequent ALND) after NAC, the technique has become plete pathologic response. In addition, long-term safety data
widely adopted [42]. Further studies examining the rates of are not yet available.
locoregional recurrence in these patients will be necessary Additional controversy exists on whether elderly patients
long term. benefit from SLNB. Prospective trials on patients aged
Currently, a large-scale multicenter randomized trial is 65–80 with clinically node-negative axilla have been per-
accruing patients who have received NAC to determine formed in attempts to clarify its utility. Omission of any
whether ALND is necessary or whether patients can be axillary staging in patients aged 65–80 years old resulted in
treated with SLNB. The Alliance A011202 trial is a phase III very low 15-year axillary recurrence and that breast-cancer-
study comparing patients who are clinically node positive specific mortality is not affected by axillary staging.
prior to treatment who then undergo NAC. Those patients However, these patients had small, ER+ tumors [45]. Other
then undergo surgery with SLNB. Patients with a positive single-institution retrospective studies have evaluated
SLN after NAC at the time of surgery are then randomized to patients over the age of 70 with small (T1), ER+, ductal
either ALND with nodal radiation or radiation to both nodal tumors. Christian et al. showed that in this group of patients,
fields and the axillary basin. The results of this study stand to omission of SLNB increased with time from 2009 to the end
be groundbreaking and practice changing should nodal radi- of their study period, 2017 [46]. A meta-analysis and sys-
ation have similar outcomes to ALND in patients who are tematic review examined the differences in risk of axillary
node-positive after NAC. In addition, the NSABP B-51 trial recurrence in elderly patients either undergoing or omitting
is examining whether SLNB can inform decisions about axillary staging. While the risk of axillary recurrence was
radiation. In this multicenter randomized trial, patients who lower in patients who underwent axillary staging (RR 0.24,
are node-positive undergo NAC and are subsequently SLN CI 0.06–0.95, p = 0.04), there was no significant difference
32 Sentinel Lymph Node Biopsy in Breast Cancer 357

in in-­breast recurrence or distant recurrence (RR 1.20, 95% superior to the mastectomy incision in the skin flap. When
CI: 0.55–2.64, I2 = 62%, p = 0.65, RR 1.17, 95% CI: 0.75– performed using this technique, SLNB was successful in 65%
1.82, I2 = 0%, p = 0.48, respectively), as well as no differ- of cases [50]. Small data sets also examined patients undergo-
ence in overall or breast-cancer specific mortality (RR 0.99, ing repeat axillary staging after ALND. Indeed, redo SLNB
95% CI: 0.79–1.24, I2 = 0%, p = 0.92, RR 1.07, 95% CI: was more successful after SLNB than after ALND, with suc-
0.72–1.57, I2 = 0%, p = 0.75, respectively), inciting ques- cess rates of 74% versus 38%, respectively [51]. Subsequently,
tions on the clinical significant of axillary staging in this systematic review and meta-­analyses confirmed these initial
group of patients [45]. findings. SLNB was successful in 81% (95% CI 76–85.2%)
Conversely, a large-scale retrospective analysis using the of patients who had previously undergone SLNB, and 52.2%
National Cancer Database (NCDB) showed a potential role (95% CI 46.6–57.8%) of patients with previous ALND [52].
for lymph node surgery in the elderly (>70 years old). A negative redo SLN has an excellent negative predictive
Patients with clinical T1–T3 and clinical N0 status were value, 96.5%, suggesting a significant benefit to patients in
divided into two groups—those with lymph nodes evaluated this setting [53].
and those without. Over 133,700 patients were included in However, whether to stage the axilla in the context of
the analysis. Seventy-six percent of those underwent nodal recurrent breast cancer remains best considered in a multidis-
surgery, and those were more likely to receive chemotherapy, ciplinary context. Small retrospective studies showed patients
radiation, and endocrine therapy. Those who did not receive who underwent re-operative axillary staging had similar rates
nodal surgery had worse overall survival (hazard ratio 1.66, of subsequent radiation and systemic therapy. In addition,
95% confidence interval 1.61–1.7) [47]. However, selection there was no statistically significant difference in times to
bias weighs heavily, as those patients who underwent che- recurrence, axillary failure, or non-axillary recurrence, nor
motherapy and radiation may have had overall better base- was there a difference in rates of distant metastasis or death,
line status and potential longer term survival from other though follow-up time was relatively short (4.2 years) [54].
conditions. Given mixed data, guidelines have yet to explic- Further multicenter trials are needed to determine the overall
itly adopt this recommendation. The Society of Surgical utility of axillary staging for recurrent disease.
Oncology “Choosing Wisely” guidelines state that routine
use of sentinel lymph node biopsy can be omitted in women
older than 70 with early stage, hormone receptor positive, 32.6 Alternative Drainage Sites
and HER2 negative disease. The NCCN states that “the per-
formance of axillary staging may be considered optional” for While the overwhelming amount of data for SLNB is for
elderly patients “in the absence of definitive data.” those nodes in the ipsilateral axilla, it is important to note
that alternative lymphatic drainage sites for primary breast
tumors exist. Drainage to these alternative sites is rare but
32.5  LNB in Patients with Previous
S not impossible. In a series of 549 patients undergoing lym-
Axillary Surgery phoscintigraphy, drainage to internal mammary nodes (IMN)
occurred in 15.7% of patients, and other non-axillary and
When the use of SLNB continued to expand, concerns arose non-internal mammary nodes in 8% of patients [55].
about whether it was technically feasible in patients who had However, the decision to remove these non-axillary nodes
previously undergone breast or axillary surgery. Initial con- remains controversial. A systematic review and meta-­
cern arose from case reports which demonstrated disrupted analysis were performed examining the utility of IMN exci-
lymphatics after ipsilateral breast surgery. However, further sion. A total of 2427 patients and 18 articles were included.
studies thereafter showed more promising results. Using a The pooled IMN positivity rate was 15%, and axillary nodal
dual-tracer technique, patients who underwent repeat SLNB metastasis was a strong, independent predictor of IMN
after an ipsilateral SLNB were able to be mapped in over 80% involvement (odds ratio 6.01, 95% CI, 3.49, 10.34). However,
of cases. Oncologic safety was also demonstrated, with no little data exist on whether removal of these nodes either
axillary recurrences seen [48]. Although thought to be techni- improves survival or decreases recurrence, and isotope drain-
cally impractical, repeat SLNB was also shown to be possible age to the IMN nodes may also depend on the depth of injec-
after ipsilateral ALND. A retrospective analysis of 45 patients tion site within the breast [56–58].
who underwent breast conservation therapy and axillary As the utility of SLNB has grown, so has the interest in
lymph node dissection was performed. Of those, 29% had the biology of these metastases. While the goal was origi-
successful lymphatic mapping and SLNB [49]. Indeed, even nally to identify tumor cells within the lymph node, current
studies looking at SLNB after mastectomy showed promise. research is focused on the biology of these cells and whether
A small review of 20 patients who had previously undergone they match the primary tumor. Recent publications have shed
mastectomy utilized radioactive tracer and blue dye injected light on the unique characteristics of these metastases, show-
358 E. Siegel et al.

ing a distinct genomic profile in comparison to the primary References

tumor, with upregulation of chemotaxis genes and downreg-
ulation of those which regulate cell–cell interactions includ- 1. Jones D, Pereira ER, Padera TP. Growth and immune evasion of
lymph node metastasis. Front Oncol. 2018;8:36.
ing collagen and metallopeptidase genes [59]. Therefore, 2. Robinson DS, Sample WF, Fee HJ, et al. Regional lymphatic drain-
treatment of axillary or nodal metastases may have different age in primary malignant melanoma of the trunk determined by
targets than the primary breast lesion. Management of the colloidal gold scanning. Surg Forum. 1977;28:147–8.
axilla continues to progress. From formal axillary dissec- 3. Morton DL, Wen DR, Foshag LJ, et al. Intraoperative lymphatic
mapping and selective cervical lymphadenectomy for early-stage
tions for every patient to decisions based on SLNB, the spec- melanomas of the head and neck. J Clin Oncol. 1993;11:1751–6.
trum of surgical intervention has changed dramatically in the 4. Giuliano AE, Kirgan DM, Guenther JM, Morton DL. Lymphatic
past 30 years. As such, further interventions will occur and mapping and sentinel lymphadenectomy for breast cancer. Ann
the extent of axillary surgery will continue to evolve. Surg. 1994;220:391–8; discussion 398–401.
5. Giuliano AE, Jones RC, Brennan M, Statman R. Sentinel lymphad-
enectomy in breast cancer. J Clin Oncol. 1997;15:2345–50.
6. Morrow M, Rademaker AW, Bethke KP, et al. Learning sentinel
32.7 Conclusion node biopsy: results of a prospective randomized trial of two tech-
niques. Surgery. 1999;126:714–20; discussion 720–712.
7. Veronesi U, Paganelli G, Galimberti V, et al. Sentinel-node biopsy
The evolution of axillary staging has been dramatic over the to avoid axillary dissection in breast cancer with clinically negative
past half-century, with a striking shift from ALND to lymph-nodes. Lancet. 1997;349:1864–7.
SLNB. Indeed, ALND has been abandoned for DCIS, and 8. Langer I, Marti WR, Guller U, et al. Axillary recurrence rate in
SLNB has replaced ALND for node-negative patients under- breast cancer patients with negative sentinel lymph node (SLN) or
SLN micrometastases: prospective analysis of 150 patients after
going both BCT and mastectomy, in addition to those patients SLN biopsy. Ann Surg. 2005;241:152–8.
with node-positive, small primary tumors undergoing 9. Smidt ML, Janssen CM, Kuster DM, et al. Axillary recurrence after
BCT. SLNB has been shown to be technically feasible with a negative sentinel node biopsy for breast cancer: incidence and
low false-negative rates in patients following neoadjuvant clinical significance. Ann Surg Oncol. 2005;12:29–33.
10. Strand SE, Persson BR. Quantitative lymphoscintigraphy I: basic
chemotherapy treatment, and its long-term utility is being fur- concepts for optimal uptake of radiocolloids in the parasternal
ther evaluated in patients who have been converted from lymph nodes of rabbits. J Nucl Med. 1979;20:1038–46.
node-positive to node-negative. Further trials are also ongo- 11. Wong JH, Terada K, Ko P, Coel MN. Lack of effect of particle size
ing to determine whether SLNB with radiation can replace on the identification of the sentinel node in cutaneous malignancies.
Ann Surg Oncol. 1998;5:77–80.
ALND in patients who remain node-positive after neoadju- 12. Faries MB, Bedrosian I, Reynolds C, et al. Active macromolecule
vant systemic therapy. As the extent of axillary surgery con- uptake by lymph node antigen-presenting cells: a novel mecha-
tinues to fall, there will likely be further groups of patients nism in determining sentinel lymph node status. Ann Surg Oncol.
who do not receive SLNB, even as its utility continues to 2000;7:98–105.
13. Wallace AM, Hoh CK, Limmer KK, et al. Sentinel lymph node
expand. accumulation of lymphoseek and Tc-99m-sulfur colloid using a
“2-day” protocol. Nucl Med Biol. 2009;36:687–92.
14. Cox CE. Lymphatic mapping in breast cancer: combination tech-
nique. Ann Surg Oncol. 2001;8:67S–70S.
15. Cox CE, Salud CJ, Cantor A, et al. Learning curves for breast can-
Questions for Future Directions cer sentinel lymph node mapping based on surgical volume analy-
• Can we rely only on SLNB and omit ALND for sis. J Am Coll Surg. 2001;193:593–600.
patients treated with mastectomy with favorable 16. Eldrageely K, Vargas MP, Khalkhali I, et al. Sentinel lymph node
tumor biology found to be SLN positive? mapping of breast cancer: a case-control study of methylene blue
tracer compared to isosulfan blue. Am Surg. 2004;70:872–5.
• When can SLNB be omitted for recurrent breast 17. Giuliano AE. Sentinel lymphadenectomy in primary breast carci-
cancer and does that answer depend not only on noma: an alternative to routine axillary dissection. J Surg Oncol.
length of time from primary disease but also on 1996;62:75–7.
receptor status? 18. Krag DN, Weaver DL, Alex JC, Fairbank JT. Surgical resection and
radiolocalization of the sentinel lymph node in breast cancer using
• What are the long-term outcomes of omitting a gamma probe. Surg Oncol. 1993;2:335–9; discussion 340.
ALND in patients who are treated with neoadjuvant 19. Chagpar A, Martin RC 3rd, Chao C, et al. Validation of subareolar
chemotherapy and converted from node-positive to and periareolar injection techniques for breast sentinel lymph node
node-negative on targeted axillary dissection? biopsy. Arch Surg. 2004;139:614–8; discussion 618–620.
20. Schrenk P, Rieger R, Shamiyeh A, Wayand W. Morbidity following
• What are the long-term outcomes of omitting SLNB sentinel lymph node biopsy versus axillary lymph node dissection
in patients who are “extraordinary” responders to for patients with breast carcinoma. Cancer. 2000;88:608–14.
neoadjuvant chemotherapy? 21. Schijven MP, Vingerhoets AJ, Rutten HJ, et al. Comparison of mor-
• Given differences in gene expressions between pri- bidity between axillary lymph node dissection and sentinel node
biopsy. Eur J Surg Oncol. 2003;29:341–50.
mary tumors and axillary metastases, will targeted 22. Husted Madsen A, Haugaard K, Soerensen J, et al. Arm morbidity
therapies to axillary metastases improved locore- following sentinel lymph node biopsy or axillary lymph node dis-
gional control and potentially overall survival? section: a study from the Danish Breast Cancer Cooperative Group.
Breast. 2008;17:138–47.
32 Sentinel Lymph Node Biopsy in Breast Cancer 359

23. Krag DN, Anderson SJ, Julian TB, et al. Sentinel-lymph-node staging in node positive breast cancer following neoadjuvant che-
resection compared with conventional axillary-lymph-node dissec- motherapy. Eur J Surg Oncol. 2020;46:1028–33.
tion in clinically node-negative patients with breast cancer: overall 40. Caudle AS, Yang WT, Krishnamurthy S, et al. Improved axillary
survival findings from the NSABP B-32 randomised phase 3 trial. evaluation following neoadjuvant therapy for patients with node-­
Lancet Oncol. 2010;11:927–33. positive breast cancer using selective evaluation of clipped nodes:
24. Giuliano AE, Hunt KK, Ballman KV, et al. Axillary dissection implementation of targeted axillary dissection. J Clin Oncol.
vs no axillary dissection in women with invasive breast cancer 2016;34:1072–8.
and sentinel node metastasis: a randomized clinical trial. JAMA. 41. Choi HJ, Kim I, Alsharif E, et al. Use of sentinel lymph node
2011;305:569–75. biopsy after neoadjuvant chemotherapy in patients with axil-
25. Giuliano AE, Ballman KV, McCall L, et al. Effect of axillary dis- lary node-­ positive breast cancer in diagnosis. J Breast Cancer.
section vs no axillary dissection on 10-year overall survival among 2018;21:433–41.
women with invasive breast cancer and sentinel node metastasis: 42. Vugts G, Maaskant-Braat AJ, de Roos WK, et al. Management of
the ACOSOG Z0011 (Alliance) Randomized Clinical Trial. JAMA. the axilla after neoadjuvant chemotherapy for clinically node posi-
2017;318:918–26. tive breast cancer: a nationwide survey study in the Netherlands.
26. Donker M, van Tienhoven G, Straver ME, et al. Radiotherapy Eur J Surg Oncol. 2016;42:956–64.
or surgery of the axilla after a positive sentinel node in breast 43. Esgueva A, Siso C, Espinosa-Bravo M, et al. Leveraging the
cancer (EORTC 10981-22023 AMAROS): a randomised, mul-
­ increased rates of pathologic complete response after neoadjuvant
ticentre, open-label, phase 3 non-inferiority trial. Lancet Oncol. treatment in breast cancer to de-escalate surgical treatments. J Surg
2014;15:1303–10. Oncol. 2021;123(1):71–9.
27. Sabel MS, Schott AF, Kleer CG, et al. Sentinel node biopsy prior to 44. Tadros AB, Yang WT, Krishnamurthy S, et al. Identification of
neoadjuvant chemotherapy. Am J Surg. 2003;186:102–5. patients with documented pathologic complete response in the
28. van Deurzen CH, Vriens BE, Tjan-Heijnen VC, et al. Accuracy of breast after neoadjuvant chemotherapy for omission of axillary sur-
sentinel node biopsy after neoadjuvant chemotherapy in breast can- gery. JAMA Surg. 2017;152:665–70.
cer patients: a systematic review. Eur J Cancer. 2009;45:3124–30. 45. Liang S, Hallet J, Simpson JS, et al. Omission of axillary staging
29. Nason KS, Anderson BO, Byrd DR, et al. Increased false nega- in elderly patients with early stage breast cancer impacts regional
tive sentinel node biopsy rates after preoperative chemotherapy for control but not survival: a systematic review and meta-analysis. J
invasive breast carcinoma. Cancer. 2000;89:2187–94. Geriatr Oncol. 2017;8:140–7.
30. Hunt KK, Yi M, Mittendorf EA, et al. Sentinel lymph node sur- 46. Christian N, Heelan Gladden A, Friedman C, et al. Increasing
gery after neoadjuvant chemotherapy is accurate and reduces the omission of radiation therapy and sentinel node biopsy in elderly
need for axillary dissection in breast cancer patients. Ann Surg. patients with early stage, hormone-positive breast cancer. Breast J.
2009;250:558–66. 2020;26:133–8.
31. Xing Y, Foy M, Cox DD, et al. Meta-analysis of sentinel lymph 47. Tamirisa N, Thomas SM, Fayanju OM, et al. Axillary nodal evalua-
node biopsy after preoperative chemotherapy in patients with breast tion in elderly breast cancer patients: potential effects on treatment
cancer. Br J Surg. 2006;93:539–46. decisions and survival. Ann Surg Oncol. 2018;25:2890–8.
32. Shirzadi A, Mahmoodzadeh H, Qorbani M. Assessment of senti- 48. Cox CE, Furman BT, Kiluk JV, et al. Use of reoperative sentinel
nel lymph node biopsy after neoadjuvant chemotherapy for breast lymph node biopsy in breast cancer patients. J Am Coll Surg.
cancer in two subgroups: initially node negative and node positive 2008;207:57–61.
converted to node negative—a systemic review and meta-analysis. 49. Kaur P, Kiluk JV, Meade T, et al. Sentinel lymph node biopsy in
J Res Med Sci. 2019;24:18. patients with previous ipsilateral complete axillary lymph node dis-
33. Fisher B, Brown A, Mamounas E, et al. Effect of preoperative che- section. Ann Surg Oncol. 2011;18:727–32.
motherapy on local-regional disease in women with operable breast 50. Karam A, Stempel M, Cody HS 3rd, Port ER. Reoperative senti-
cancer: findings from National Surgical Adjuvant Breast and Bowel nel lymph node biopsy after previous mastectomy. J Am Coll Surg.
Project B-18. J Clin Oncol. 1997;15:2483–93. 2008;207:543–8.
34. Bear HD, Anderson S, Brown A, et al. The effect on tumor response 51. Port ER, Garcia-Etienne CA, Park J, et al. Reoperative sentinel
of adding sequential preoperative docetaxel to preoperative doxo- lymph node biopsy: a new frontier in the management of ipsilateral
rubicin and cyclophosphamide: preliminary results from National breast tumor recurrence. Ann Surg Oncol. 2007;14:2209–14.
Surgical Adjuvant Breast and Bowel Project protocol B-27. J Clin 52. Maaskant-Braat AJ, Voogd AC, Roumen RM, Nieuwenhuijzen
Oncol. 2003;21:4165–74. GA. Repeat sentinel node biopsy in patients with locally recurrent
35. Mamounas EP, Anderson SJ, Dignam JJ, et al. Predictors of locore- breast cancer: a systematic review and meta-analysis of the litera-
gional recurrence after neoadjuvant chemotherapy: results from ture. Breast Cancer Res Treat. 2013;138:13–20.
combined analysis of National Surgical Adjuvant Breast and Bowel 53. Poodt IGM, Vugts G, Schipper RJ, Nieuwenhuijzen GAP. Repeat
Project B-18 and B-27. J Clin Oncol. 2012;30:3960–6. sentinel lymph node biopsy for ipsilateral breast tumor recurrence:
36. Boughey JC, Suman VJ, Mittendorf EA, et al. Sentinel lymph node a systematic review of the results and impact on prognosis. Ann
surgery after neoadjuvant chemotherapy in patients with node-­ Surg Oncol. 2018;25:1329–39.
positive breast cancer: the ACOSOG Z1071 (Alliance) Clinical 54. Ugras S, Matsen C, Eaton A, et al. Reoperative sentinel lymph node
Trial. JAMA. 2013;310:1455–61. biopsy is feasible for locally recurrent breast cancer, but is it worth-
37. Kuehn T, Bauerfeind I, Fehm T, et al. Sentinel-lymph-node biopsy while? Ann Surg Oncol. 2016;23:744–8.
in patients with breast cancer before and after neoadjuvant che- 55. Tanis PJ, Nieweg OE, Valdes Olmos RA, et al. Impact of non-­
motherapy (SENTINA): a prospective, multicentre cohort study. axillary sentinel node biopsy on staging and treatment of breast
Lancet Oncol. 2013;14:609–18. cancer patients. Br J Cancer. 2002;87:705–10.
38. Boileau JF, Poirier B, Basik M, et al. Sentinel node biopsy after 56. Tan C, Caragata R, Bennett I. Is sentinel node biopsy of the internal
neoadjuvant chemotherapy in biopsy-proven node-positive breast mammary lymph nodes relevant in the management of breast can-
cancer: the SN FNAC study. J Clin Oncol. 2015;33:258–64. cer? Breast J. 2017;23:410–4.
39. Balasubramanian R, Morgan C, Shaari E, et al. Wire guided locali- 57. Qiu PF, Zhao RR, Wang W, et al. Internal mammary sentinel lymph
sation for targeted axillary node dissection is accurate in axillary node biopsy in clinically axillary lymph node-positive breast can-
360 E. Siegel et al.

cer: diagnosis and implications for patient management. Ann Surg 59. Srour MK, Gao B, Dadmanesh F, et al. Gene expression compari-
Oncol. 2020;27:375–83. son between primary triple-negative breast cancer and paired axil-
58. Gong J, Yu Y, Wu G, et al. Should internal mammary lymph node lary and sentinel lymph node metastasis. Breast J. 2020;26:904–10.
sentinel biopsy be performed in breast cancer: a systematic review
and meta-analysis. World J Surg Oncol. 2019;17:135.
 Sentinel Node Navigation Surgery
for Upper Gastrointestinal Cancer 33
Shuhei Mayanagi and Yuko Kitagawa

Abstract
cer: this method uses radioisotopes and dye, and
Radical resection with systematic lymph node dissection involves dissection of the lymphatic basin defined
is traditionally performed in patients with upper gastroin- as the lymphatic drainage area along the main gas-
testinal cancer because the lymphatic system of these can- tric arteries.
cers is complex and skip metastases are often present. • For clinical T1 gastric cancer <3–4 cm, large-scale
Although sentinel lymph node mapping for upper gastro- prospective clinical trials for sentinel lymph node
intestinal cancer is still in clinical trials, the sentinel node navigation surgery are ongoing, with overall sur-
concept could be applied for early-stage cancers. In early vival as the primary endpoint.
gastric cancer, in particular, cancer metastasis can be
identified with high diagnostic accuracy using the dual
tracer method and sentinel basin dissection. The useful-
ness of sentinel node navigation surgery for early gastric
cancer has been verified in multicenter prospective stud- 33.1 Introduction
ies. In the near future, it will likely be possible to safely
perform radical surgery with a minimal extent of gastrec- The presence and extent of lymph node metastasis are impor-
tomy based on the sentinel node concept. tant prognostic factors for upper gastrointestinal cancers,
such as gastric cancer, esophagogastric junction cancer, and
esophageal cancer, and systematic regional lymph node dis-
section is the standard surgical procedure for these patients.
Learning Objectives Even today, with the development of chemotherapy and radi-
• The sentinel lymph node in upper gastrointestinal ation therapy, surgery still plays a very important role in the
cancer is not necessarily the node that is located treatment of upper gastrointestinal cancers. In particular,
anatomically closest to the primary lesion and is not early gastric cancer (EGC) in which the invasion of the pri-
necessarily limited to a single lymphatic drainage mary lesion is limited to the submucosa (T1), is curable by
area. gastrectomy with lymph node dissection alone. Several fac-
• The dual tracer method is the gold standard for sen- tors contribute to the favorable outcome of surgery for EGC,
tinel lymph node navigation surgery for gastric can- the main reason being the small number of lymph node
metastases. The frequency of lymph node metastasis in EGC
is 1–8% in mucosal cancer and 5–19% in submucosal inva-
sive cancer [1]. Even if lymph node metastasis is present, the
majority of patients with EGC are cured by appropriate lymph
node dissection. However, patients who undergo standard
S. Mayanagi
gastrectomy with systematic lymph node dissection often suf-
Division of Esophageal Surgery, Shizuoka Cancer Center Hospital,
Shizuoka, Japan fer from postgastrectomy syndromes, including heartburn,
e-mail: [email protected] early satiety, and abnormal bowel movements [2, 3].
Y. Kitagawa (*) In recent years, most surgeries for upper gastrointestinal
Department of Surgery, Keio University School of Medicine, cancers have been performed endoscopically, laparoscopi-
Tokyo, Japan cally, or thoracoscopically, and the quality of life of patients
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 361
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_33
362 S. Mayanagi and Y. Kitagawa

has improved, especially in the early postoperative period. 33.2 Gastric Cancer
The main difference between laparoscopic surgery and open
surgery is the trauma to the abdominal wall, while the extent 33.2.1 Sentinel Node Mapping Procedure
of gastrectomy and lymph node dissection is the same in
both; therefore, there is almost no difference in long-term Unlike malignancies on the surface of the body, such as mel-
outcomes, such as postgastrectomy syndromes, between the anoma and breast cancer, several points require attention in
two types of surgery. Standard gastrectomy and systematic sentinel lymph node mapping for upper gastrointestinal can-
lymph node dissection have produced excellent overall sur- cers that are deep in the body cavity.
vival for EGC, but with significant risk for postoperative First, regarding the tracer, as with other cancers, there are
dysfunction; thus the next logical goal is trying to achieve two main methods for identifying sentinel lymph nodes in
similar cure rates without the requirement for extensive gas- EGC: the radio-guided method and dye method. Technetium-­
trectomy and lymph node dissection. 99m colloid is frequently used as a tracer in the radio-guided
Historically, it is essential to develop research on the method, and isosulfan blue, patent blue, and indocyanine
lymph flow around the stomach in order to establish the green (ICG) are frequently used in the dye method. However,
extent of lymph node dissection for gastric cancer. In line each method has advantages and disadvantages. For instance,
with this, a systematic report on the lymphatic system of the the radio-guided method is able to confirm the complete har-
stomach was published in the early 1900s [4]. Since then, the vest of sentinel nodes by gamma probing, even deep inside
lymphatic system, based on lymph node metastasis and the body cavity; however, its use is severely restricted due to
recurrence in clinical practice, has been extensively studied radiation exposure. While the dye method allows real-time
for more than 100 years. According to the Japanese observation of the visualized lymphatic vessels, it may be
Classification of Gastric Cancer, regional lymph nodes of difficult to see with dye in deep adipose tissue. The improve-
gastric cancer are defined as the “lymph node group targeted ment of the detection rate and diagnostic accuracy using a
for dissection,” and are classified into three groups [5]. In combined method has been reported in lymphatic mapping
this classification, the frequency of lymph node metastasis for both breast cancer and malignant melanoma, and conse-
and the therapeutic effect of lymph node dissection are con- quently, it has become the gold standard for gastric cancer.
sidered in addition to the anatomical position of the lymph The technical errors that are present using the single map-
nodes: The perigastric lymph nodes are classified into Group ping agent approach can be reduced by adding a different
(Compartment) 1, along the main trunk arteries into Group 2, approach for lymphatic mapping. Therefore, a combination
and periaortic and deep lymph nodes into Group 3. Depending of the dye and radio-guided methods is recommended for
on the clinical stage of the primary tumor, the extent of the systematic lymphatic mapping of EGC. In our institute, we
lymph node groups that should be removed is determined have chosen to use technetium-99m tin colloid, which has a
prophylactically, e.g., D2 dissection for removal of up to relatively large particle size [6]. In our experience, tin colloid
Group 2 lymph nodes. with a diameter of 500 nm migrates into the sentinel nodes
Systematic lymph node dissection, represented by D2 dis- within 2 h and remains there for more than 20 h following
section, sacrifices the major arteries that feed the stomach and phagocytosis by macrophages in the sentinel nodes. The
results in extensive gastrectomy. For instance, a distal gas- radio-guided method allows quantitative detection of senti-
trectomy with D2 lymph node dissection for gastric cancer nel nodes due to the stable incorporation of the radioactive
located at the antrum of the stomach requires resection of the tracer, even in blind areas. Moreover, the use of a small
distal two-thirds of the stomach. Therefore, minimization of gamma probe is also useful for sentinel node identification in
lymph node dissection is essential to reduce the extent of gas- laparoscopic surgery. In contrast, ICG, which is frequently
trectomy. As mentioned previously, the rate of lymph node used as a dye tracer for sentinel node identification, has an
metastasis in EGC is low, so routine systematic lymph node extremely small particle size and is effectively taken up into
dissection and extensive gastrectomy are not always required lymphatic vessels, and is therefore suitable for real-time
for EGCs. However, since most lymph node metastases in observation of lymphatic flow during surgery [7]. In ­addition,
EGC are microscopic (histological lymph node metastasis), a method of irradiating near-infrared light to capture the flu-
there is currently no reliable method of nonsurgical diagnos- orescence emitted by ICG under a laparoscope has been
tic for the detection of lymph node metastasis. It is unclear developed, and good results have been reported that surpass
which of the myriad regional lymph nodes should be biopsied the dye method that relies on visual recognition [8, 9]. In this
in order to make a rapid intraoperative pathological diagnosis method, a 2.0 mL solution of technetium-99m tin colloid
of lymph node metastasis; this is in contrast to malignant (150 MBq) is injected into the submucosal layer around the
tumors on the body surface, such as breast cancer. The senti- primary tumor in all four abdominal quadrants the day before
nel lymph node biopsy represents the long-awaited method surgery, and ICG is also initially injected using intraopera-
required for a breakthrough in EGC treatment. tive endoscopy via an approach similar to that used for radio-
33 Sentinel Node Navigation Surgery for Upper Gastrointestinal Cancer 363

isotopes. Stained lymphatic vessels and nodes can be prospective multicenter trial [13]. Sentinel basin dissection
detected visually within 15 min of injecting the dye, and a is considered a minimally focused lymphadenectomy method
hand-held gamma probe is simultaneously used to identify for EGC with a reasonable safety net that avoids recurrence
the radioactive sentinel nodes. after a false-negative sentinel node biopsy.
Second, two types of sentinel lymph node sampling pro-
cedures have been described in the treatment of EGC, the
pickup method or lymphatic basin dissection (Fig. 33.1). The 33.2.2 Indication for Sentinel Node Mapping
pickup method is a well-established and simple method that
is currently used to assess breast cancer and melanoma; how- One of the most common reasons for a false-negative result of
ever, this method is not considered to be suitable for EGC sentinel node mapping in gastric cancer is a lymphatic vessel
given its diverse and complex lymphatic drainage. To address obstructed by cancer invasion. In these cases, the adminis-
this issue, the concept of lymphatic basin dissection has been tered tracer cannot migrate into the real sentinel nodes and its
developed in EGC. The lymphatic basin is defined as the flow instead takes the path of least resistance, leading it to the
lymphatic zone, which is divided by the stream of dye map- second-echelon, or false, sentinel nodes. Thus, clinically pos-
ping [10]. The proximal border of the lymphatic basin is the itive node cases and advanced primary tumor cases are not
fatty tissue attached to the stomach wall, and the distal bor- indications for sentinel node mapping. The purpose of this
der of the basin is the front of the stained node furthest from technique is to identify clinically undetectable lymph node
the stomach. All sentinel lymph nodes exist within the lym- involvement. Although muscular invasive tumor (T2) is the
phatic basin, and the lymphatic basins are thought to be the indication in many studies as well as T1 tumors, false-nega-
primary lymphatic drainage areas, with one or a few basin(s) tive cases were reported in patients with T2 gastric cancer in
normally present in EGC. Miwa and Kinami proposed the a previous prospective study [13]. Therefore, this technique is
basin concept and classified the basins into five directions now excluded from indications in order to eliminate false
along the following main arteries: left gastric artery, right positives. Moreover, lateral spread is also an important factor
gastric artery, left gastroepiploic artery, right gastroepiploic in the indication, and initially, it was also indicated for tumors
artery, and posterior gastric artery in which sentinel node that were spread widely in the lateral direction. However, a
basins contained true positive nodes [10, 11] (Fig. 33.2). large tumor size >4 cm strongly correlated with skip metasta-
Compared to the pickup method, basin dissection has been ses [14]. In short, clinical T1 tumors with a primary lesion
reported to be superior in terms of detection rate, sensitivity, diameter <4 cm are considered as suitable indications for sen-
and specificity [12] and it was also shown to be valid in a tinel node mapping for gastric cancer.

Fig. 33.1 Sentinel lymph

node mapping procedures:
Pickup method versus
lymphatic basin dissection.
Pickup method is not
considered to be suitable for
early gastric cancer given its
diverse and complex
lymphatic drainage. The
lymphatic basins are thought
to be the primary lymphatic
drainage areas, with one or a
few basin(s) normally present
in early gastric cancer.
364 S. Mayanagi and Y. Kitagawa

Fig. 33.2 Gastric lymphatic

basins. The sentinel basins are
classified into five directions 2
along the following main
arteries: left gastric artery, Left gastric artery Posterior gastric artery
right gastric artery, left 1
4sa
gastroepiploic artery, right
gastroepiploic artery, and 9 11d
posterior gastric artery 7 11p

Right gastric artery 4sb

8
3a
Left gastroepiploic artery
5
3b

4d

Right gastroepiploic artery

33.3 Current Status and Future Prospects lymph node of gastric cancer is not necessarily the node that
is located anatomically closest to the primary lesion and is
The generalization of the sentinel node concept is remark- not necessarily limited in a single lymphatic drainage area.
able in both malignant melanoma and breast cancer, and Furthermore, in gastrointestinal cancer, the sentinel node
lymph node dissection using biopsy is currently omitted in may be both patient- and lesion-specific. Sentinel nodes with
these cancers. In contrast, the development of the sentinel multiple distributions are not a major problem for lymphatic
node concept in EGC has been delayed and is still in the mapping. In addition, the development of local gastrectomy
clinical trial setting. There are several reasons for this. First, performed at the same time as sentinel lymph node biopsy
gastric cancer may be cured by thorough lymph node dissec- has also progressed. With the development of laparoscopic
tion, even in cases with a wide extent of lymph node metas- endoscopic cooperative surgery, such as nonexposed endo-
tases, compared to breast cancer where there is no evidence scopic wall inversion surgery [24], endoscopic full-thickness
that axillary lymph node dissection improves prognosis. resection has become safer and more reliable, and the extent
Therefore, there was initially little motivation to omit lymph of gastrectomy has become smaller. Furthermore, it has
node dissection. Second, the lymphatic system of the stom- become possible to realize function-preserving minimizing
ach is complex and spreads in multiple directions, and skip surgery for EGC using sentinel lymph node biopsy.
metastases are observed at a frequency of approximately A large-scale multicenter prospective clinical trial was
5–10% [15, 16]. Furthermore, there was a misunderstanding conducted by the Japanese Society for Sentinel Node
surrounding whether or not the sentinel lymph node concept Navigation Surgery (SNNS) in order to verify the validity of
would hold true in the context of EGC. Finally, there was sentinel node mapping for EGC [13]. Of the 433 cT1 or cT2
skepticism regarding the application of the sentinel node gastric cancer patients, 397 were considered eligible based
concept in EGC because the usefulness of surgery without on surgical findings. Sentinel node biopsies were performed
extensive gastrectomy and lymph node dissection for func- in all patients, with a detection rate of 97.5% (387/397).
tion preservation was unclear. There were no serious adverse effects associated with endo-
However, despite this initial skepticism concerning this scopic tracer infusion. Of the 57 patients with conventional
concept in the management of EGC, a number of single insti- hematoxylin and eosin-stained lymph node metastases, 93%
tutional studies supporting the validity of the sentinel node (53 of 57) were sentinel node positive, and the accuracy of
concept for EGC have been reported with the development the lymph node evaluation of metastases was 99% (383 of
of the dual tracer method and basin dissection mentioned 387). Only four false-negative sentinel node biopsies were
above [6, 17–23]. It is now hypothesized that the sentinel observed, and pathological analysis revealed that three of
33 Sentinel Node Navigation Surgery for Upper Gastrointestinal Cancer 365

these biopsies had pT2 or tumor >4 cm. It has been sug- 33.4 Esophageal Cancer
gested that the sentinel node concept is sufficiently clinically
applicable to relatively small cT1 gastric cancer of ≤4 cm Sentinel node mapping for thoracic esophageal cancer is
using the dual tracer method. In addition, the short-term out- technically more difficult than that for gastric cancer.
comes of another large-scale prospective multicenter ran- Because the esophagus is deep in the mediastinum, even
domized controlled trial (the SENORITA trial) were reported though surgical mobilization of the esophagus is essential for
in South Korea [25]. This trial was designed to compare esophagectomy, the mobilization can interfere with active
laparoscopic standard gastrectomy and laparoscopic SNNS lymphatic flow from the primary tumor site. For this reason,
using the dual tracer method with 3-year disease-free sur- the radio-guided method is preferred over the dye method to
vival in EGC tumors <3 cm. A total of 580 patients were identify sentinel lymph nodes in esophageal cancer. Although
randomly enrolled in this trial. In the SNNS group, sentinel several types of radioisotope tracers have been used in senti-
node basins were detected in 96.7%, of the 245 patients: 127 nel mapping of esophageal cancer, radioactive colloids with
(52%) had 1, 89 (36%) had 2, and 21 (9%) had 3 sentinel relatively large particle sizes have long deposits of tracers in
basins. Stomach-preserving surgery, such as wedge resection the lymph nodes and are effective in identifying sentinel
and segmental resection, was carried out in 210 of 258 lymph nodes. In our institute, we use technetium-99m tin
patients, and the rate and severity of postoperative complica- colloid that is endoscopically injected around the primary
tions are comparable to those after standard gastrectomy. tumor in the same way as is performed for gastric cancer
Meanwhile, an unacceptably high false-negative rate of [29]. Preoperative lymphatic scintigraphy is usually per-
46.4% (13/28) was reported in the multicenter prospective formed 3–4 h after injection of the tracer. Lymphatic scintig-
study conducted by the Japan Clinical Oncology Group raphy makes it easy to understand the distribution of the
(JCOG0302) [26]. This study was conducted to evaluate the sentinel lymph nodes, while gamma probing can identify the
feasibility and accuracy of diagnosis using sentinel node sentinel lymph node intraoperatively.
biopsy using the only dye method with intraoperative subse- Although few studies have shown the validity of the sen-
rosal direct injection of ICG. The most serious limitations of tinel lymph node concept in early esophageal cancer com-
this study were the inclusion criteria for the participating pared to that in EGC, some single-center studies have shown
institutes in terms of prior experience with sentinel node map- acceptable outcomes not only in thoracic esophageal squa-
ping for gastric cancer. In the field of breast cancer surgery, mous cell carcinoma but also in esophagogastric junctional
the significance of the technical learning curve for this tech- adenocarcinoma [30–34]. We have also previously reported a
nique has already been clearly demonstrated. For sentinel feasibility study on radio-guided sentinel mapping for 75
node mapping of gastric cancer, the experience of 26 cases patients with cT1 or cT2 esophageal cancer [35]. Our study
has been suggested as the minimum requirement for high demonstrated acceptable sentinel node detection in 95% and
detection rates [27]. These studies showed that SNNS for diagnostic accuracy in 94% of nodal metastases based on
EGC within 3–4 cm can be performed with high diagnostic sentinel node status. Interestingly, the distribution of SN
accuracy using the dual tracer method in a facility proficient showed a widespread from cervical to abdominal areas in
in sentinel mapping. A relatively high incidence of skip patients with thoracic esophageal cancer. Even in lower tho-
metastases in EGC is not an obstacle to the clinical applica- racic esophageal cancer, the sentinel nodes were mainly
tion of the sentinel node concept. On the contrary, this tech- located in the lower and middle mediastinum and the abdom-
nique is essential for the accurate detection of micrometastases inal area, although some patients showed sentinel nodes in
in EGC due to the variations in the lymphatic drainage routes. the upper mediastinum. In more than 85% of patients with
It has also been suggested that the sentinel concept could thoracic esophageal cancer, at least one sentinel node was
be applied to lesions after endoscopic resection [28]. As a found to be located in the Group 2 or 3 regional lymph nodes.
next step, it may be possible to perform further minimizing A systematic review and meta-analysis reported the diagnos-
surgery by combining endoscopic resection with sentinel tic value of sentinel node mapping in esophageal cancer [36].
biopsy in order to preserve the entire stomach. At present, the A meta-analysis of 23 related studies including 623 patients,
enrollment of patients in clinical trial to verify non-­inferiority the detection rate of sentinel node and diagnostic accuracy
with overall survival as the primary outcome has been com- was 93% and 88%, respectively. There was no significant
pleted in Japan as well as in the SENORITA trial. Based on difference in histological type between squamous cell carci-
the results of these large-scale prospective studies, it is noma or adenocarcinoma and it was concluded that SN map-
expected that SNNS will improve the quality of life for ping was technically feasible for patients with esophageal
patients with EGC in the future. cancer.
366 S. Mayanagi and Y. Kitagawa

Radical esophagectomy with extended three-field lymph-

adenectomy, including the neck, mediastinum, and abdo- Open Questions
men, has been established in Japan as a potential curative • To what extent can function-preserving limited gas-
surgery for patients with thoracic esophageal squamous cell trectomy be applied to gastric cancer using lym-
carcinoma. Esophagectomy with three-field lymphadenec- phatic mapping and the sentinel lymph node
tomy may be appropriate due to the wide distribution of sen- concept?
tinel nodes and the presence of unpredictable metastatic • Can training of surgeons in performing lymphatic
lymph nodes. Therefore, the application of the sentinel node mapping and sentinel lymph node biopsy be
concept to thoracic esophageal squamous cell carcinoma improved and standardized to allow accurate per-
may be limited. However, it may be possible to select cases formance of the technique outside of current centers
and apply sentinel mapping with the aim to improve the cur- of excellence?
ability of surgery and reduce postoperative complications. • Will ongoing clinical trials validate acceptable sur-
For example, if a sentinel node is identified along the recur- vival with improved morbidity with endoscopic
rent laryngeal nerves in the upper mediastinum and is posi- resection and sentinel lymph node mapping in gas-
tive for cancer metastasis in a patient with lower esophageal tric cancer?
cancer, extended lymphadenectomy for the upper mediasti- • Which patients with esophageal cancer are optimal
num might be considered; this may be the case even in Ivor candidates for lymphatic mapping and sentinel
Lewis esophagectomy, which does not normally require lymph node biopsy?
upper mediastinum lymph node dissection. Also, if sentinel
nodes are only identified in the abdominal area, and are neg-
ative for metastasis in patients with adenocarcinoma of the
distal esophagus, it may be possible to treat with limited References
resection of the distal esophagus without the need for exten-
sive mediastinal lymphadenectomy. In addition, if a combi- 1. Japanese Gastric Cancer Association. Japanese gastric cancer treat-
ment guidelines 2010 (ver. 3). Gastric Cancer. 2011;14(2):113–23.
nation of endoscopic submucosal dissection and sentinel 2. Eagon JC, Miedema BW, Kelly KA. Postgastrectomy syndromes.
lymph node biopsy can be performed simultaneously, as in Surg Clin North Am. 1992;72(2):445–65.
gastric cancer, esophageal-preserving treatment may be pos- 3. Nakada K, Ikeda M, Takahashi M, Kinami S, Yoshida M, Uenosono
sible in patients with early cancer located in the thoracic Y, Kawashima Y, Oshio A, Suzukamo Y, Terashima M, et al.
Characteristics and clinical relevance of postgastrectomy syndrome
esophagus or at the esophagogastric junction. However, no assessment scale (PGSAS)-45: newly developed integrated ques-
multicenter prospective trial has been conducted to evaluate tionnaires for assessment of living status and quality of life in post-
the sentinel node concept of esophageal cancer. Although gastrectomy patients. Gastric Cancer. 2015;18(1):147–58.
further accumulation of evidence based on multicenter trials 4. Jamieson JK, Dobson JF. Lectures on the lymphatic system of the
stomach. Lancet. 1907;169(4364):1061–6.
using standard protocols for sentinel node mapping is 5. Japanese Gastric Cancer Association. Japanese classification
needed, SNNS has the potential to be an ideal and minimally of gastric carcinoma—2nd English edition. Gastric Cancer.
invasive personalized treatment for esophageal cancer. 1998;1(1):10–24.
6. Kitagawa Y, Fujii H, Mukai M, Kubota T, Otani Y, Kitajima
M. Radio-guided sentinel node detection for gastric cancer. Br J
Surg. 2002;89(5):604–8.
33.5 Conclusion 7. Takeuchi H, Kitagawa Y. New sentinel node mapping technologies
for early gastric cancer. Ann Surg Oncol. 2013;20(2):522–32.
SNNS for early-stage upper gastrointestinal cancer is one of 8. Nimura H, Narimiya N, Mitsumori N, Yamazaki Y, Yanaga K,
Urashima M. Infrared ray electronic endoscopy combined with
the most promising treatment options, as long as it is adapted indocyanine green injection for detection of sentinel nodes of
by an appropriate method for the indication. With the devel- patients with gastric cancer. Br J Surg. 2004;91(5):575–9.
opment of procedures and devices, it is expected that SNNS 9. Tajima Y, Yamazaki K, Masuda Y, Kato M, Yasuda D, Aoki T, Kato
will become more sophisticated and will be of considerable T, Murakami M, Miwa M, Kusano M. Sentinel node mapping
guided by indocyanine green fluorescence imaging in gastric can-
benefit to patients with EGC. cer. Ann Surg. 2009;249(1):58–62.
33 Sentinel Node Navigation Surgery for Upper Gastrointestinal Cancer 367

10. Miwa K, Kinami S, Taniguchi K, Fushida S, Fujimura T, Nonomura node biopsy using radioactive tracer in gastric cancer. Surgery.
A. Mapping sentinel nodes in patients with early-stage gastric car- 2005;138(5):899–904.
cinoma. Br J Surg. 2003;90(2):178–82. 24. Goto O, Takeuchi H, Kawakubo H, Sasaki M, Matsuda T, Matsuda
11. Kinami S, Fujimura T, Ojima E, Fushida S, Ojima T, Funaki H, S, Kigasawa Y, Kadota Y, Fujimoto A, Ochiai Y, et al. First case
Fujita H, Takamura H, Ninomiya I, Nishimura G, et al. PTD clas- of non-exposed endoscopic wall-inversion surgery with sentinel
sification: proposal for a new classification of gastric cancer loca- node basin dissection for early gastric cancer. Gastric Cancer.
tion based on physiological lymphatic flow. Int J Clin Oncol. 2015;18(2):434–9.
2008;13(4):320–9. 25. Eom BW, Yoon HM, Min JS, Cho I, Park JH, Jung MR, Hur H, Kim
12. Lee YJ, Ha WS, Park ST, Choi SK, Hong SC, Park JW. Which YW, Park YK, Nam BH, et al. Prospective multicenter feasibility
biopsy method is more suitable between a basin dissection and study of laparoscopic Sentinel Basin dissection after endoscopic
pick-up biopsy for sentinel nodes in laparoscopic sentinel-node submucosal dissection for early gastric cancer: SENORITA 2 trial
navigation surgery (LSNNS) for gastric cancer? J Laparoendosc protocol. J Gastric Cancer. 2019;19(2):157–64.
Adv Surg Tech A. 2008;18(3):357–63. 26. Miyashiro I, Hiratsuka M, Sasako M, Sano T, Mizusawa J,
13. Kitagawa Y, Takeuchi H, Takagi Y, Natsugoe S, Terashima M, Nakamura K, Nashimoto A, Tsuburaya A, Fukushima N. High
Murakami N, Fujimura T, Tsujimoto H, Hayashi H, Yoshimizu N, false-negative proportion of intraoperative histological examina-
et al. Sentinel node mapping for gastric cancer: a prospective mul- tion as a serious problem for clinical application of sentinel node
ticenter trial in Japan. J Clin Oncol. 2013;31(29):3704–10. biopsy for early gastric cancer: final results of the Japan Clinical
14. Lee SE, Lee JH, Ryu KW, Cho SJ, Lee JY, Kim CG, Choi IJ, Kook Oncology Group multicenter trial JCOG0302. Gastric Cancer.
MC, Nam BH, Park SR, et al. Sentinel node mapping and skip 2014;17(2):316–23.
metastases in patients with early gastric cancer. Ann Surg Oncol. 27. Lee JH, Ryu KW, Lee SE, Cho SJ, Lee JY, Kim CG, Choi IJ, Kook
2009;16(3):603–8. MC, Kim MJ, Park SR, et al. Learning curve for identification of
15. Maruyama K, Sasako M, Kinoshita T, Sano T, Katai H. Can sentinel sentinel lymph node based on a cumulative sum analysis in gastric
node biopsy indicate rational extent of lymphadenectomy in gastric cancer. Dig Surg. 2009;26(6):465–70.
cancer surgery? Fundamental and new information on lymph-node 28. Mayanagi S, Takahashi N, Mitsumori N, Arigami T, Natsugoe S,
dissection. Langenbecks Arch Surg. 1999;384(2):149–57. Yaguchi Y, Suda T, Kinami S, Ohi M, Kawakubo H, et al. Sentinel
16. Siewert JR, Sendler A. Potential and futility of sentinel node node mapping for post-endoscopic resection gastric cancer: mul-
detection for gastric cancer. Recent Results Cancer Res. ticenter retrospective cohort study in Japan. Gastric Cancer.
2000;157:259–69. 2020;23(4):716–24.
17. Hiratsuka M, Miyashiro I, Ishikawa O, Furukawa H, Motomura 29. Takeuchi H, Kitagawa Y. Sentinel node navigation surgery for
K, Ohigashi H, Kameyama M, Sasaki Y, Kabuto T, Ishiguro S, esophageal cancer. Gen Thorac Cardiovasc Surg. 2008;56(8):393–6.
et al. Application of sentinel node biopsy to gastric cancer surgery. 30. Yasuda S, Shimada H, Chino O, Tanaka H, Kenmochi T, Takechi
Surgery. 2001;129(3):335–40. M, Nabeshima K, Okamoto Y, Kato Y, Kijima H, et al. Sentinel
18. Hayashi H, Ochiai T, Mori M, Karube T, Suzuki T, Gunji Y, lymph node detection with Tc-99m tin colloids in patients with
Hori S, Akutsu N, Matsubara H, Shimada H. Sentinel lymph esophagogastric cancer. Jpn J Clin Oncol. 2003;33(2):68–72.
node mapping for gastric cancer using a dual procedure with 31. Burian M, Stein HJ, Sendler A, Piert M, Nährig J, Feith M, Siewert
dye- and gamma probe-guided techniques. J Am Coll Surg. JR. Sentinel node detection in Barrett’s and cardia cancer. Ann Surg
2003;196(1):68–74. Oncol. 2004;11(3 Suppl):255s–8s.
19. Ichikura T, Morita D, Uchida T, Okura E, Majima T, Ogawa T, 32. Lamb PJ, Griffin SM, Burt AD, Lloyd J, Karat D, Hayes N. Sentinel
Mochizuki H. Sentinel node concept in gastric carcinoma. World J node biopsy to evaluate the metastatic dissemination of oesopha-
Surg. 2002;26(3):318–22. geal adenocarcinoma. Br J Surg. 2005;92(1):60–7.
20. Ohdaira H, Nimura H, Mitsumori N, Takahashi N, Kashiwagi H, 33. Bhat MA, Naikoo ZA, Dass TA, Lone RA, Dar AM. Role of intra-
Yanaga K. Validity of modified gastrectomy combined with senti- operative sentinel lymph node mapping in the management of carci-
nel node navigation surgery for early gastric cancer. Gastric Cancer. noma of the esophagus. Saudi J Gastroenterol. 2010;16(3):168–73.
2007;10(2):117–22. 34. Kim HK, Kim S, Park JJ, Jeong JM, Mok YJ, Choi YH. Sentinel
21. Gretschel S, Bembenek A, Hünerbein M, Dresel S, Schneider W, node identification using technetium-99m neomannosyl
Schlag PM. Efficacy of different technical procedures for senti- human serum albumin in esophageal cancer. Ann Thorac Surg.
nel lymph node biopsy in gastric cancer staging. Ann Surg Oncol. 2011;91(5):1517–22.
2007;14(7):2028–35. 35. Takeuchi H, Fujii H, Ando N, Ozawa S, Saikawa Y, Suda K, Oyama
22. Lee JH, Ryu KW, Kim CG, Kim SK, Lee JS, Kook MC, Choi IJ, T, Mukai M, Nakahara T, Kubo A, et al. Validation study of radio-­
Kim YW, Chang HJ, Bae JM. Sentinel node biopsy using dye and guided sentinel lymph node navigation in esophageal cancer. Ann
isotope double tracers in early gastric cancer. Ann Surg Oncol. Surg. 2009;249(5):757–63.
2006;13(9):1168–74. 36. Nagaraja V, Eslick GD, Cox MR. Sentinel lymph node in oesopha-
23. Zulfikaroglu B, Koc M, Ozmen MM, Kucuk NO, Ozalp N, geal cancer—a systematic review and meta-analysis. J Gastrointest
Aras G. Intraoperative lymphatic mapping and sentinel lymph Oncol. 2014;5(2):127–41.
 Sentinel Lymph Node Mapping
in Non-­small Cell Lung, Colon, 34
and Thyroid Carcinomas

Joshua K. Kays and Mark B. Faries

Abstract 34.1 Introduction

Lymphatic mapping and sentinel lymph node (SLN)
biopsy is now a standard procedure for melanoma and Lymph node involvement is the single most important
breast cancer, providing the most accurate staging in locoregional prognostic factor in nearly all solid tumors.
those diseases. This technique is now also being applied Sentinel lymph node (SLN) mapping is a technique that
to other malignancies, including non-small cell lung car- involves injection of a dye or tracer at the primary tumor site
cinoma, colon cancer, and thyroid cancer, although with- in order to map the lymphatic drainage of that site and deter-
out widespread acceptance as yet. This chapter reviews mine the location of the lymph node(s) receiving direct
the use of SLN biopsy in those malignancies including drainage from the tumor. This technique has proven to be
technical details, SLN detection rates, frequency of nodal successful in breast cancer and melanoma, revolutionizing
upstaging, and potential clinical applications and the way these malignancies are treated. It has allowed for
limitations. more precise staging, which has ultimately resulted in
improved treatments and outcomes for patients with these
cancers. More recently, this technique has been applied to
other solid tumors with similar goals.
Learning Objectives
The concept of an SLN as a site of specific lymphatic
• Understand the importance of accurate lymph node drainage from an organ or body site of interest has been
staging for NSCLC, colon, and thyroid cancer prog- speculated and investigated for many years [1]. However,
nosis and treatment. broad applicability remained elusive until Morton and col-
• Appreciate the variability in lymphatic drainage leagues described a surgical technique involving injection of
and how it contributes to the limitations of tradi- vital blue dye at the site of a primary melanoma and identifi-
tional lymph node harvesting and pathologic exam- cation of the blue, SLN in the draining nodal basin [2]. This
ination for NSCLC, colon, and thyroid cancer. technique proved accurate and reproducible, allowing the
• Describe the various techniques and success rates determination of the pathologic status of the nodal basin
for SLN mapping. without the need for a complete nodal dissection. Further,
• Understand how SLN mapping improves patho- this study showed the likelihood of a non-sentinel node har-
logic staging compared to traditional methods. boring cancer with a negative sentinel lymph node was less
• Appreciate SLN mapping’s role in the ever-­ than 1%.
changing care of NSCLC, colon, and thyroid cancer Given the advantages of SLN biopsy that were demon-
patients. strated in melanoma, it was natural that the technique was
investigated in other solid tumors. In breast cancer, it has
also become a standard component of initial therapy. In other
malignancies, though, several challenges have kept the tech-
J. K. Kays
Cedars-Sinai Medical Center, Los Angeles, CA, USA nique from broad application. However, in non-small cell
e-mail: [email protected] carcinoma of the lung (NSCLC), colon cancer, and thyroid
M. B. Faries (*) cancer, the technique remains under active investigation. Its
The Angeles Clinic and Research Institute and Cedars-Sinai use in other diseases, including cancers of the upper gastro-
Medical Center, Los Angeles, CA, USA intestinal organs, is discussed elsewhere in this text.
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 369
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_34
370 J. K. Kays and M. B. Faries

34.2  otential Benefits of Sentinel Lymph

P without clinically apparent lymph node involvement. The
Node Biopsy in General primary treatment for these tumors is surgical resection with
mediastinal lymph node sampling to stage the lymph nodes.
Prior to SLN mapping, regional lymph node basins were Patients without lymph node involvement are considered to
staged by removal of the entire basin and histopathologic be free of disease and undergo surveillance, while those with
examination. This approach not only engenders the risk of lymph node involvement are generally treated with adjuvant
surgical morbidity for a larger procedure, it also burdens the chemotherapy with the goals of decreasing recurrence and
pathologist with an increased amount of material to examine. improving survival.
In breast cancer and melanoma, the risks of dissection The fact that patients with negative lymph node involve-
include lymphedema, which is probably one of the most ment by traditional pathologic methods have a 20–30%
feared complications. While that particular risk is not a major recurrence rate contradicts the notion that these patients are
concern for nodal dissections for NSCLC, colon cancer or completely disease free. [3] It is suspected that occult lymph
thyroid cancer, dissections for these cancers also present node metastases lead to this relapse. Furthermore, the pres-
risks for increased morbidity to the patient. As discussed ence of locoregional lymph node disease decreases 5-year
subsequently, the scrutiny a pathologist can practically place survival by approximately 50% [4]. SLN biopsy has the
on a full lymph node dissection is less than that which can be potential to improve staging accuracy in NSCLC as it has in
used for a single, or small number of, SLNs. This enhanced other malignancies, but technical issues that are particular to
pathologic analysis may also increase the accuracy of stag- intrathoracic malignancies make mapping for these cancers
ing, in addition to decreasing morbidity. more challenging. These challenges include the close prox-
imity of many lymph nodes to the primary tumor site, the
physically confined nature of the thoracic cavity, and the
34.3 Non-small Cell Lung Carcinoma anthracotic nature of pulmonary lymph nodes making blue
dye more difficult to see.
Lung cancer is the second most common cancer diagnosis in Table 34.1 shows an overview of studies conducted with
the United States and is the number one cause of cancer mor- corresponding SLN identification success rates. The first
tality. NSCLC represents over 80% of all lung cancers. published study using an SLN technique in NSCLC was per-
Early-stage NSCLC can generally be defined as NSCLC formed by Little et al. in 1999 [5]. They performed intraop-

Table 34.1 Sentinel lymph node detection rates

Non-small cell lung carcinoma
Study N Tracer Injection method SLN detection
Little et al. [5] 36 ISB PT 47% (17/36)
Liptay et al. [6] 45 99m
Tc-SC PT 82% (37/45)
Norori et al. [7] 47 99m
Tc tin colloid PT 87% (40/47)
Schmidt et al. [8] 31 99m
Tc -SC+ ISB Intratumoral 80% (25/31)
Sugi et al. [9] 48 Indocyanine green Intra-op PT 6.3% (1/16)
ISB Intra-op PT 50% (9/18)
99m
Tc tin colloid CT-guided PT 64.3% (9/14)
Melfi et al. [10] 26 99m
Tc-SC Intra-op PT ora 96% (25/26)
CT-guided PT
Lardinois et al. [11] 20 99m
Tc-SC Endoscopic injection 95% 919/20)
Nakagawa et al. [12] 38 Ferumoxides Intra-op PT 82% (31/38)
Faries et al. [13] 28 ISB + 99m Tc-SC Intra-op PT 100% (28/28)
Colon cancer
Study N Tracer Injection method SLN detection
Bilchik et al. [14] 120 ISB and 99m Tc-SC Intra-op, in vivo + ex vivo, PT 96% (115/120)
Saha et al. [15] 86 ISB Intra-op PT 99% (85/86)
Patten et al. [16] 57 ISB + 99m Tc sulfur Intra-op PT 98% (56/57)
Blichik et al. [17] 100 ISB Intra-op, in vivo + ex vivo, PT 97% (97/100)
Saha et al. [18] 209 ISB/fluorescein + 99m Tc-SC 100% (209/209)
Wong et al. [19] 124 ISB Ex vivo, PT 98% (121/124)
Wood et al. [20] 75 ISB Intra-op, in vivo + ex vivo, PT 96% (72/75)
Bertagnolli et al. [21] 72 ISB Intra-op PT 92% (66/72)
Weisse, et al. [35] 83 ISB Intra-op PT 99% (82/83)
ISB isosulfan blue, 99mTc-SC technetium-99m sulfur colloid, PT peritumoral
a—did not differentiate rates between methods/Tc-99m—Technetium 99m colloid
34 Sentinel Lymph Node Mapping in Non-small Cell Lung, Colon, and Thyroid Carcinomas 371

erative peritumoral injections of isosulfan blue. While they and improved 5-year DFS and OS when compared to those
showed a relatively low SLN detection rate of 42%, they had who were deemed to be nodal negative by traditional medi-
a 100% negative predictive value, as no patient with a nega- astinal lymph node sampling. While this could be attributed
tive SLN was found to have disease in any other node after to a sampling error, other studies have shown similar results.
systematic mediastinal lymphadenectomy. Kuboschok et al. demonstrated a 21.6% upstaging rate of
More recent studies have shown significantly improved previously negative SLN after IHC [23]. Those discovered to
SLN detection rates using different tracers, most notably have disseminated tumor cells by IHC had decreased DFS
technetium-99 colloids. Melfi et al. used pre- and intra-­ and OS on univariate analysis. Furthermore, multivariate
operative peritumoral injections of technetium-99 albumin analysis showed IHC positive nodes were associated with a
nanocolloid to achieve 96% SLN detection rate [10], while 2.7 times increased risk of recurrence and a 2.5 times
Lardinois used transbronchial injections technetium-99 just increased risk of shorter survival.
prior to incision to obtain a 95% SLN detection rate with a To date, SLN biopsy for NSCLC remains confined to cer-
90% negative predictive value of the SLN [11]. Several oth- tain specialized centers with an interest in the technique.
ers have shown similar results. More clinical data will likely be required to demonstrate the
The ipsilateral peribronchial, hilar, and intrapulmonary utility of the technique for this cancer before the adoption
nodes are the closest lymph node basins to a lung tumor. becomes more widespread.
These are designated as the N1 nodal station and represent
the most likely first place for the lymphatic spread of
NSCLC. However, studies have shown that between 13 and 34.4 Colon Cancer
21% of SLN in NSCLC were in locations other than the N1
nodal station [11, 13]. This opens up the possibility of having Colon cancer is the third most common cancer diagnosis and
skip metastases, which are metastases in nodes beyond the the second leading cause of cancer deaths. Similar to
N1 station despite the most proximal nodes being negative. NSCLC, early-stage colon cancer is generally defined as
These skip metastases may be responsible for the high rate of colon cancer without lymph node involvement. The primary
recurrence of node-negative NSCLC. SLN mapping is a sys- treatment is colectomy with removal of the adjacent mesen-
tematic way to identify aberrant lymph node drainage and tery and the lymph nodes it contains. Examination of at least
may lead to more reliable identification of skip metastases 12 lymph nodes is required as a minimum for accurate stag-
than the current standard of mediastinal lymph node ing. Patients with negative lymph nodes are typically man-
sampling. aged with surveillance, while those with positive nodes are
Another significant value of SLN mapping is that it allows treated with adjuvant chemotherapy. However, patients with
for “ultrastaging” of the regional lymph nodes. Traditional “high-risk” Stage II colon cancers, including those with
histopathologic evaluation involves hematoxylin and eosin inadequate nodal assessment are considered for adjuvant
(H&E) staining to allow examination of the cell structure. chemotherapy as well.
Immunohistochemistry (IHC) uses antibodies targeting cell-­ In the same manner that NSCLC has a high recurrence
specific antigens to identify tissues and has been shown to rate among node-negative patients, a significant number of
accurately detect even isolated tumor cells. IHC, however, is patients with early-stage colon cancer will also experience
more costly and time-consuming than traditional H&E stain- recurrence of the disease. Approximately 1/3 of node-­
ing, and its application to all lymph nodes harvested from a negative patients will have disease recurrence and node posi-
regional lymph node basin is not practical. SLN mapping tivity is associated with a relative decrease in survival by
allows for targeted use of IHC, which has been shown to about 50% [24–26].
discover metastatic disease in lymph nodes previously Table 34.1 shows an overview of studies conducted on
declared disease negative by H&E staining. SLN mapping in colon cancer and the corresponding rates of
Upstaging rates for NSCLC from N0 to N+ disease using SLN identification. While nearly all studies use intraopera-
this ultrastaging technique have been reported as high as tive peritumoral injections, mapping agents vary. The overall
30%. Having the nodes upstaged to N+ also appears to have success rate of SLN identification is between 58 and 100%
significant implication on patient outcomes as studies have with most studies reporting >80% success rates. This is in
shown those with SLN involvement discovered by IHC have contrast to mapping for rectal cancers, which has not been
increased risk of recurrence, shorter recurrence-free survival found to have acceptable accuracy in most studies.
(RFS), disease-free survival (DFS), and overall survival (OS). Variability in lymphatic drainage with aberrant or unex-
Digesu et al. showed that SLN mapping with IHC was pected SLN locations has also been encountered in colon
able to detect occult metastatic disease in 30% of all SLN cancer. While most studies show the SLN to be within the
specimens in patients with NSCLC [22]. Those with nega- nodes in most immediate proximity to the tumor, there are
tive SLN determined by IHC had lower rates of recurrence studies that have demonstrated a not insignificant number of
372 J. K. Kays and M. B. Faries

patients to have aberrant nodal drainage to locations that negative rate with 13% of patients with a negative SLN
might not be removed in standard colectomy procedures. having non-SLN with metastases [32]. Keleman et al.
Saha et al. used peritumoral injections of blue dye at the showed that 11.8% of patients who underwent SLN map-
time of colon resection and found that 22% of patients had ping had SLN in the internal jugular region in addition to
aberrant lymphatic drainage that required a change in opera- the paratracheal SLN [33]. Finally, Balasubrammania
tion to ensure removal of these nodes [27]. In addition, nodal et al. presented an analysis of the theoretical cost and ben-
positivity was higher in patients who required a change in efit of SLN biopsy suggested the procedure would only
operation, and 10% of the positive SLN were found in aber- save seven patients from central neck lymph node dissec-
rant locations. tion in a 5-year period at their institution [34].
Ultrastaging sentinel lymph nodes with IHC have been
shown to have the potential to upstage colon cancer as well.
Weixler et al. showed that SLN mapping with blue dye fol- 34.6  linical Application and Future
C
lowed by IHC examination resulted in a 26% upstaging rate Direction
to node-positive disease and that micrometastases and ITC
were associated with decreased survival [28]. Protic et al. The data on SLN mapping for NSCLC and colon cancer
showed significantly lower disease recurrence rates in clearly demonstrate that it is technically feasible, readily
patients with IHC negative nodes (2.6% vs. 16.7%) and a reproducible, and has a high success rate of SLN identifica-
significantly longer DFS (92.9 months vs. 71.8 months) tion. Adding IHC to SLN that were determined to be nega-
compared to patients with IHC positive nodes [29]. While tive by traditional pathologic methods appears to identify a
Bilchik et al. found a lower rate of upstaging after IHC at significant proportion of patients that may be upstaged to
only 8%, they did show that 83% of recurrences at 25 months node-positive and may benefit from adjuvant therapy to
had lymph node metastases that were only detected by IHC decrease recurrence and improve survival. Conversely,
[30]. They concluded that detection of SLN metastases by patients with colon cancer who have negative SLN may have
IHC after being deemed node negative by traditional staining a sufficiently favorable prognosis that adjuvant systemic
methods may improve selection of patients for adjuvant che- therapy and its attendant morbidity is not necessary for them.
motherapy and those with SLN negative by IHC are likely to Future trials enrolling patients with micrometastases and
be cured with surgery alone. ITC discovered by SLN mapping and IHC will need to be
performed to determine the true benefit. As of this writing
the benefit of SLN mapping in thyroid cancer is less clear
34.5 Thyroid Cancer and additional studies are needed to establish its utility.

While thyroid cancer is not nearly as deadly as NSCLC or

colon cancer, nodal staging is a standard component of treat- 34.7 Limitations of SLN Mapping
ment for many patients and lymph node involvement is a
poor prognostic feature, particularly among older patients. One of the major limitations preventing the widespread
Nodal involvement also requires additional treatment, which application of SLN mapping in NSCLC and colon cancer is
in the case of thyroid cancer may involve a complete central the lack of standardization with the technique. Among the
or lateral lymph node dissection. While complete central published studies, various techniques have been used with
and/or lateral neck dissection is generally a safe procedure, various results; preoperative versus intraoperative injection,
there are risks including nerve injury or inadvertent removal peritumoral versus CT-guided versus endoscopic, blue dye
of parathyroid glands. SLN mapping has been investigated versus technetium-99 colloids versus ICG. A standardized
as a means to allow the identification of patients who should approach may result in more acceptance among clinicians
have a neck dissection performed and those in whom it may and widespread usage. As technology improves, optimal
be possible to omit this procedure. techniques may emerge in each of these malignancies.
The data on the utility of SLN mapping in thyroid can- Another factor limiting interest in SLN biopsy for these
cer are not as extensive as it is for other previously men- malignancies is the paucity of data proving clinical benefit.
tioned malignancies. Albers et al. found an SLN detection While it has been shown SLN mapping is feasible and that it
rate that was 67% by preoperative scintigraphy, but can upstage disease, studies have yet to show a clinical ben-
dropped to 46% when using intraoperative gamma probe efit with regard to decreased recurrence or improved sur-
[31]. Cabera et al. showed a much higher rate of detection vival. Further studies need to be performed to demonstrate
of an SLN by scintigraphy at 91%. However, the final the clinical benefit before SLN mapping will be likely to be
pathologic examination showed a not insignificant false- widely adopted for these cancers.
34 Sentinel Lymph Node Mapping in Non-small Cell Lung, Colon, and Thyroid Carcinomas 373

34.8 Conclusions cancer: the sentinel node technique. J Thorac Cardiovasc Surg.
1999;117(2):220–4.
6. Liptay MJ, Masters GA, Winchester DJ, Edelman BL, Garrido BJ,
While regional nodal assessment using lymphatic mapping Hirschtritt TR, Perlman RM, Fry WA. Intraoperative radioisotope
and sentinel lymph node biopsy has proven to be an invalu- sentinel lymph node mapping in non-small cell lung cancer. Ann
able technique in several common malignancies, its role in Thorac Surg. 2000;70(2):384–9.
7. Normori H, Horio H, Naruke T, Orikasa H, Yamazaki K, Suemasu
the lung, colon, and thyroid cancer is still being defined. In K. Use of technetium-99m tin colloid for sentinel lymph node iden-
NSCLC, improvements in technology and continued clinical tification in non-small cell lung cancer. J Thorac Cardiovasc Surg.
research evaluations will help improve the applicability of 2002;124(3):486–92.
the technique there. In colon cancer, adaptation of lymphatic 8. Schmidt FE, Woltering EA, Webb WR, Garcia OM, Cohen JE,
Rozans MH. Sentinel nodal assessment in patients with carcinoma
mapping to the minimally invasive techniques that are more of the lung. Ann Thorac Surg. 2002;74(3):870–5.
commonly employed for that disease will likely be essential 9. Sugi K, Fukuda M, Nakamura H, Kaneda Y. Comparison of three
to broadening the utility of sentinel node evaluation in that tracers for detecting sentinel lymph nodes in patients with clinical
disease. In particular, if the accuracy of nodal staging can be N0 lung cancer. Lung Cancer. 2003;39(1):37–40.
10. Melfi FM, Chella A, Menconi GF, Givigliano F, Boni G, Mariani G,
improved with SLN mapping, more patients may be able to Sbragia P, Angeletti CA. Intraoperative radioguided sentinel lymph
avoid the toxicity of adjuvant chemotherapy. The biology of node biopsy in non-small cell lung cancer. Eur J Cardiothorac Surg.
thyroid cancer may limit the utility of SLN mapping in that 2003;23(2):214–20.
disease, but additional research may demonstrate popula- 11. Lardinois D, Brack T, Gaspert A, Spahr T, Schneiter D, Steinert
HC, Weder W. Bronchoscopic radioisotope injection for sentinel
tions for whom the technique is of particular value. lymph node mapping in potentially resectable non-small cell lung
cancer. Eur J Cardiothorac Surg. 2003;23(5):824–7.
12. Nakagawa T, Minamiya Y, Katayose Y, Saito H, Taguchi K, Imano
H, Watanabe H, Enomoto K, Sageshima M, Ueda T, Ogawa JI. A
Future Perspective Questions novel method for sentinel lymph node mapping using magnetite in
• Can prognostic evaluation of patients with NSCLC patients with non-small cell lung cancer. J Thorac Cardiovasc Surg.
be improved by SLN mapping? 2003;126(2):563–7.
• Will technologies, including fluorescence imaging 13. Faries MB, Bleicher RJ, Ye X, Essner R, Morton DL. Lymphatic
mapping and sentinel lymphadenectomy for primary and
in minimally invasive thoracic surgery enable more metastatic pulmonary malignant neoplasms. Arch Surg.
widespread application of SLN mapping in that 2004;139(8):870–7.
disease? 14. Bilchik AJ, Nora DT, Sobin LH, Turner RR, Trocha S, Krasne
• Can ultrastaging of regional lymph nodes be made D, Morton DL. Effect of lymphatic mapping on the new tumor-­
node-­metastasis classification for colorectal cancer. J Clin Oncol.
feasible using SLN mapping and would such stag- 2003;21(4):668–72.
ing enable more patients to safely avoid adjuvant 15. Saha S, Wiese D, Badin J, Beutler T, Nora D, Ganatra BK, Desai
chemotherapy? D, Kaushal S, Nagaraju M, Arora M, Singh T. Technical details of
• Which patients with thyroid cancer may benefit sentinel lymph node mapping in colorectal cancer and its impact on
staging. Ann Surg Oncol. 2000;7:120–4.
from nodal staging using SLN mapping? 16. Patten LC, Berger DH, Rodriguez-Bigas M, Mansfield P,
Delpassand E, Cleary KR, Fagan SP, Curley SA, Hunt KK, Feig
BW. A prospective evaluation of radiocolloid and immunohisto-
chemical staining in colon carcinoma lymphatic mapping. Cancer.
References 2004;100(10):2104–9.
17. Blichik AJ, Nora D, Tollenaar RA, Van de Velde CJ, Wood T,
Turner R, Morton DL, Hoon DS. Ultrastaging of early colon can-
1. Kett K, Varga G, Lukacs L. Direct lymphography of the breast.
cer using lymphatic mapping and molecular analysis. Eur J Cancer.
Lymphology. 1970;3(1):3–12.
2002;38(7):977–85.
2. Morton DL, Wen DR, Wong JH, Economou JS, Cagle LA,
18. Saha S, Dan AG, Beutler T, Wiese D, Schochet E, Badin J, Branigan
Storm FK, Foshag LJ, Cochran AJ. Technical details of intraop-
T, Ng P, Bassily N, David D. Sentinel lymph node mapping tech-
erative lymphatic mapping for early stage melanoma. Arch Surg.
nique in colon cancer. Semin Oncol. 2004;31(3):374–81.
1992;127(4):392–9.
19. Wong JH, Johnson DS, Namiki T, Tauchi-Nishi P. Validation
3. Lou F, Huang J, Sima CS, Dycoco J, Rusch V, Bach PB. Patterns
of ex vivo lymphatic mapping in hematoxylin-eosin node-­
of recurrence and second primary lung cancer in early stage lung
negative carcinoma of the colon and rectum. Ann Surg Oncol.
cancer survivors followed with routine CT surveillance. J Thorac
2004;11(8):772–7.
Cardiovasc Surg. 2013;145(1):75–82.
20. Wood TF, Saha S, Morton DL, Tsioulias GJ, Rangel D, Hutchinson
4. Rusch VW, Hawes D, Decker PA, Martin SE, Abati A, Landreneau
W, Foshag LJ, Bilchik AJ. Validation of lymphatic mapping in
RJ, Patterson GA, Inculet RI, Jones DR, Malthaner RA, Cohen
colorectal cancer: in vivo, ex vivo and laparoscopic techniques.
RG. Occult metastases in lymph nodes predict survival in resect-
Ann Surg Oncol. 2001;8(2):150–7.
able non-small cell lung cancer: report of the ACOSOG Z0040 trial.
21. Bertagnolli M, Miedema B, Redston M, Dowell J, Niedzwiecki D,
J Clin Oncol. 2011;29(32):4313.
Fleshman J, Bem J, Mayer R, Zinner M, Compton C. Sentinel node
5. Little AG, DeHoyos A, Kirgan DM, Arcomano TR, Murray
staging of resectable colon cancer. Results of a multicenter study.
KD. Intraoperative lymphatic mapping for non-small cell lung
Ann Surg. 2004;240(4):624.
374 J. K. Kays and M. B. Faries

22. Digesu CS, Hachey KJ, Gilmore DM, Khullar OV, Tsukada H, results and identifies patients with decreased survival: a prospective
Whang B, Chirieac LR, Padera RF, Jaklitsch MT, Colson YL. Long single-center trial. World J Surg. 2017;41(9):2378–86.
term outcomes after near-infrared sentinel lymph node map- 29. Protic M, Stojadinovic A, Nissan A, Wainberg Z, Steele SR, Chen
ping in non-small cell lung cancer. J Thorac Cardiovasc Surg. DC, Avital I, Bilchik AJ. Prognostic effect of ultra-staging node-­
2018;155(3):1280–91. negative colon cancer without adjuvant chemotherapy: a prospec-
23. Kubuschok B, Passlick B, Izbicki JR, Thetter O, Pantel tive national cancer institute-sponsored clinical trial. J Am Coll
K. Disseminated tumor cells in lymph nodes as a determinant for Surg. 2015;221(3):643–51.
survival in surgically resected non-small cell lung cancer. J Clin 30. Bilchik AJ, Hoon DS, Saha S, Turner RR, Wiese D, DiNome M,
Oncol. 1999;17(1):19–9. Koyanagi K, McCarter M, Shen P, Iddings D, Chen SL. Prognostic
24. Le Voyer TE, Sigurdson ER, Hanlon AL, Mayer RJ, Macdonald JS, impact of micrometastases in colon cancer. Interim results of a pro-
Catalano PJ, Haller DG. Colon cancer survival is associated with spective multicenter trial. Ann Surg. 2007;246(4):568–77.
increasing number of lymph nodes analyzed: a secondary survey of 31. Albers MB, Nordenstrom E, Wohlfahrt J, Bergenfelz A, Almquist
intergroup trial INT-0089. J Clin Oncol. 2003;21(15):2912–9. M. Sentinel lymph node biopsy in thyroid cancer. World J Surg.
25. Cohen AM, Kelsen D, Saltz L, Minsky BD, Nelson H, Farouk 2020;44(1):142–7.
R, Gunderson LL, Michelassi F, Arenas RB, Schilsky RL, Willet 32. Cabrera RN, Chone CT, Zantut-Wittmann D, Matos P, Perreira
CG. Adjuvant therapy for colorectal cancer. Curr Probl Surg. DM, Pereira PS, Ferrari RJ, Santos AO, Crespo AN, Etchebehere
1997;34(8):601–76. EC. Value of sentinel lymph node biopsy in papillary thyroid can-
26. Cohen AM, Tremiterra S, Candela F, Thaler HT, cer: initial results of a prospective trial. Eur Arch Otorhinolaryngol.
Sigurdson ER. Prognosis of nodal positive cancer. Cancer. 2015;272(4):971–9.
1991;67(7):1859–61. 33. Keleman PR, Van Herle AJ, Giuliano AE. Sentinel lymph-
27. Saha S, Elgamal M, Cherry M, Buttar R, Pentapati S, Mukkamala adenectomy in thyroid malignant neoplasms. Arch Surg.
S, Devisetty K, Kaushal S, Alnounou M, Singh T, Grewal 2008;133(3):288–92.
S. Challenging the conventional treatment of colon cancer by 34. Balasubramanian SP, Brignall J, Lin HY, Stephenson TJ, Wadsley J,
sentinel lymph node mapping and its role of detecting micro- Harrison BJ, Craig WL, Smart L, Krukowski. Sentinel node biopsy
metastases for adjuvant chemotherapy. Clin Exp Metastasis. in papillary thyroid cancer—what is the potential? Langenbecks
2018;36(5–6):463–9. Arch Surg. 2014;399(2):245–51.
28. Weixler B, Rickenbacher A, Raptis DA, Viehl CT, Guller U, Rueff 35. Wiese DA, Saha S, Badin J, Ng PS, Gauthier J, Ahsan A, Yu
J, Zetti A, Zuber M. Sentinel lymph node mapping with isosul- L. Pathologic evaluation of sentinel lymph nodes in colorectal car-
fan blue or indocyanine green in colon cancer shows comparable cinoma. Arch Pathol Lab Med. 2000;124(12):1759–63.
 Part IX
Surgical Treatment of Primary
and Metastatic Cancer

Giorgos Karakousis and Dale Han

The surgical management of cancer has undergone remarkable changes over the past century
based on an improved understanding of tumor biology, the development of effective systemic
therapies, and the introduction of technologies to detect patient-specific molecular aberrations
and therapeutic targets, thereby allowing for individualized, precision oncologic care. In con-
trast, prior to the mid-twentieth century, most oncologic surgeries were performed with a rudi-
mentary understanding of cancer biology and a lack of other treatment options. The ability to
control cancer and provide a durable cure was relatively limited during this early period of
oncologic care and the prognosis for cancer patients was generally poor. However, the intro-
duction of radical surgery by William Stewart Halsted in the late nineteenth century revolu-
tionized the thought process of the treatment of tumors and ultimately helped to pave the way
for the development of newer theories on cancer biology, such as the systemic theory and later
the spectrum theory, which challenged the Halstedian view of orderly cancer progression. In
parallel with an improved understanding of cancer biology, there were significant develop-
ments in other modalities for cancer treatment, including radiation therapy and systemic thera-
pies. In particular, systemic therapy has seen an evolution from a relatively unselective
approach with uniform usage of cytotoxic chemotherapeutics to more recently the develop-
ment of targeted therapies taking advantage of specific genetic aberrations in certain cancers,
and finally to harnessing the power of the innate host immune system to eradicate cancers and
provide durable regression of disease.
The efficacy of surgical therapy for cancers is now determined by several factors, and this
section will focus on the development and evolution of surgical treatments in specific cancer
types based on cancer biology and the availability of other therapies that can be used as multi-­
modality therapy in conjunction with surgery. Drs. Georgia Beasley, Michael Lowe, and
Genevieve Boland will specifically discuss the role of surgery in managing primary and meta-
static melanoma, while Drs. Alicia Terando, Mara Piltin, Tina Hieken, and Carla Fisher will
evaluate the role of surgery in managing primary and metastatic breast cancer. Our focus will
then change to a discussion of the role of surgery in treating several intra-abdominal cancers
with Drs. Najjia Mahmoud, Kiran Turaga, and David Shibata discussing the role of surgery in
managing primary and metastatic colorectal cancer and Drs. Robert Roses and Jose Pimiento
assessing the role of surgery in managing primary and metastatic gastric cancer. Finally, Drs.
Ajay Maker, George Poultsides, and Shishir Maithel will discuss the role of surgery in specifi-
cally managing primary and metastatic hepatopancreaticobiliary cancers.
In demonstrating how the role of surgery has evolved and developed in these various
cancer types, our goals are to demonstrate how our armamentarium of therapy options has
increased as our comprehension of cancer biology has evolved, and show that while surgery
remains an integral component to oncologic treatment, it is but one of many treatment
376 Surgical Treatment of Primary and Metastatic Cancer

options we have at hand which should be used in conjunction with, not exclusively from, other available
treatment modalities. In light of these developments and in line with the shift from a more radical to
conservative approach, the role of surgery in the treatment of cancer remains essential but is evolving,
with surgical oncologists now playing an active role in all stages of the oncologic treatment spectrum,
including preventive counseling and screening, initial cancer diagnosis, collaboration with multidisci-
plinary teams to determine the optimal sequence of treatment and interventions, postoperative surveil-
lance monitoring, end-of-life care, and high-quality research.
 The Efficacy and Evolution of Surgical
Management Based on Cancer Biology 35
Richard J. Straker III, Hayley Standage,
Giorgos C. Karakousis, and Dale Han

Abstract
Learning Objectives
The surgical management of cancer has changed dramati- • To understand the different theories on cancer biol-
cally over the past century. As our understanding of can- ogy that have developed over time.
cer biology has evolved, surgical techniques have also • To understand the historical and current role of sys-
changed to align with the concepts of tumor growth and temic therapy in the management of cancer.
metastasis. William Stewart Halsted’s theory of radical • To understand the evolution of surgical manage-
surgery at the end of the nineteenth century represented ment of cancers in light of developing theories of
early methods for curative therapy for cancers. However, cancer biology and the development of effective
the introduction of alternative theories, including the sys- systemic therapies.
temic theory and later the spectrum theory, has helped to • To understand the modern molecular techniques
change our approach to cancer surgery, such that some being used to diagnose, identify, and treat cancer in
cancers are treated with more radical surgery, while other the setting of evaluating patients for surgical
tumors may benefit from more conservative approaches. management.
Furthermore, the development of effective systemic thera-
pies and precision molecular medicine has also influenced
how certain cancers are managed, and while the role and
timing of surgery may be changing, it remains an integral 35.1 Introduction
component of modern cancer care. As such, a universal
treatment algorithm cannot be mandated, but each cancer The surgical management of cancer has undergone consider-
must be approached individually and treated with the best able changes over the past century based on an improved
combination of available therapies at hand for that spe- understanding of tumor biology and the development of
cific malignancy. effective systemic therapies. Prior to the mid-twentieth cen-
tury, most oncologic surgeries were performed for palliation
[1]. Other treatment options, such as radiation and systemic
therapies, either were not available or were novel approaches
Richard J. Straker and Hayley Standage contributed equally to the man-
uscript and share first authorship.
that were not generally efficacious. The ability to control
cancer and provide a durable cure was relatively limited, and
Giorgos C. Karakousis and Dale Han contributed equally to the the prognosis for cancer patients was generally poor [2].
manuscript and share senior authorship.
However, the introduction of radical surgery by William
Stewart Halsted in the late nineteenth century revolutionized
R. J. Straker III (*) · G. C. Karakousis the thought process of the treatment of tumors [3]. This phi-
Division of Endocrine and Oncologic Surgery, Department of losophy dominated oncology therapy for most of the next
Surgery, University of Pennsylvania, Philadelphia, PA, USA century.
e-mail: [email protected]; giorgos.
Oncologic care in the latter half of the twentieth century
[email protected]
evolved as adjunct treatments such as radiation and systemic
H. Standage · D. Han
therapies demonstrated beneficial outcomes for cancer
Division of Surgical Oncology, Oregon Health and Science
University, Portland, OR, USA patients. Alternative theories of cancer biology also influ-
e-mail: [email protected]; [email protected] enced the view of cancer development and treatment. For

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 377
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_35
378 R. J. Straker III et al.

instance, Bernard Fisher challenged the Halstedian paradigm cer but also adequate surrounding tissue to control that level
with the systemic theory of cancer biology, which postulated (e.g., radical surgery) could prevent cancer from spreading to
that the systemic spread of cancer could occur earlier in the the subsequent level. This idea was first introduced in his
disease process [4]. Radical surgery would therefore be inad- discussion of the radical mastectomy [3] and was then
equate to halt the spread of cancer. The acceptance of the applied to other cancer types. The philosophy of radical sur-
systemic theory of cancer biology led to fundamental gical control of cancer dominated oncology therapy for much
changes in the management of breast cancer. However, it has of the next century.
since been realized that the biology of cancers and even dif- The Halstedian paradigm was challenged in the latter half
ferent tumors of the same cancer category can vary tremen- of the twentieth century by Dr. Bernard Fisher, who theo-
dously. It is now theorized that cancer biology is a rized that breast cancer was a systemic disease. Fisher
heterogeneous spectrum. Tumors may remain localized observed evidence of cancer recurrence at various sites in
throughout the disease course, while other cancers may many patients despite radical surgery. In animal models, he
spread systemically relatively early. Therefore, treatment of found cancer cells in circulation when the tumor otherwise
a specific tumor should be tailored to the specific biology of appeared clinically localized. The Fisher theory of cancer
that cancer and may include surgery or a multimodal biology, also known as the systemic theory, postulated that
approach with a combination of surgery, radiation therapy, cancers did not progress in an orderly fashion and could
and systemic therapy [4]. instead spread both via the lymphatic system and hematog-
The efficacy of surgical therapy for cancers is determined enously. Therefore, the systemic spread of cancer could
by several factors, such as the biology of a specific tumor and occur earlier in the disease process. A radical surgical
the availability and effectiveness of therapies that can be uti- approach would have limited utility to halt the progression of
lized in conjunction with surgery to optimize outcomes [5]. cancer that was already present at distant sites at disease pre-
For example, surgical treatment of a primary tumor may pro- sentation [6]. Fisher’s theory led to fundamental changes in
vide more benefit in cancers with a greater tendency to breast cancer treatment. For instance, less radical surgeries
remain localized or in cancer types in which effective sys- were completed to treat the primary tumor (e.g., lumpec-
temic therapies are available. In contrast, metastasectomy tomy), and additional therapies such as radiation therapy
may have limited utility in cases in which there is more were used to further control the primary site. Systemic ther-
aggressive biology or in which systemic therapies are less apy was utilized in cases that appeared clinically localized
effective in controlling the metastatic disease. This section but were at high risk for disease recurrence.
will focus on the development and evolution of surgical The Fisher model of cancer spread was controversial and
treatments in specific cancer types based on cancer biology met with resistance, but the preponderance of data support-
and the availability of other therapies that can be used as ing the doctrines of this theory, through the National Surgical
multi-modality therapy in conjunction with surgery. Adjuvant Breast and Bowel Project clinical trials, led to gen-
eral acceptance of his theory and the development of new
management paradigms in breast cancer [6]. Subsequently,
35.2  istorical Evolution of Theories
H however, it was observed that the biology of cancers appeared
on Cancer Biology to vary according to different cancer types and to be far more
complex than what could be explained by the models devel-
An understanding of cancer biology was limited until the oped by Halsted and Fisher. In 1994, Hellman et al. intro-
mid-twentieth century. Early surgical treatment for cancer duced the “spectrum theory” of metastases, which theorized
was based on anatomic principles and an early rudimentary that cancer biology was a heterogeneous spectrum, with
understanding of tumor and disease spread. As such, sur- some tumors remaining localized throughout the disease
geons simply excised the affected anatomical structures. course and other tumors presenting with systemic disease at
Additionally, adjunctive options for disease control such as first detection [7]. The biology of a specific cancer was the
radiation therapy or chemotherapy were not yet available [1]. function of that specific type of tumor growth and progres-
As a result, operations were often palliative, and the progno- sion. Persistence of disease increased the risk for cancer
sis was generally poor [2]. In the late nineteenth century, Dr. spread and progression at secondary sites. For instance,
William Stewart Halsted introduced the concept of radical lymph node metastasis was a marker of more aggressive
surgery that transformed the approach to cancer treatment. biology but could also serve as a nidus for further disease
The Halstedian model of cancer progression stated that metastases. Therefore, locoregional therapy could poten-
tumors progressed by spread through contiguous structures tially have an important role in the prevention of disease pro-
in an orderly fashion from the primary site to the level of the gression and decrease opportunities for cancer cells to
regional lymph nodes, and then to the systemic level at dis- develop into metastatic lesions. Moreover, the importance of
tant sites. Appropriate therapy that removed not only the can- the host, particularly the immune system, in influencing the
35 The Efficacy and Evolution of Surgical Management Based on Cancer Biology 379

progression of cancers continues to be an area of active known antifolates and is very prevalent in the modern
investigation [8]. treatment of several lymphoma subtypes, osteosarcoma, and
As the biological theories of cancer continue to evolve, choriocarcinoma [9]. Expanding on the work with antifo-
the approach to treatment is also adapting. Early surgeons lates, numerous additional classes of antimetabolites were
focused on anatomical resections to provide palliation. These created, mostly functioning via the inhibition of purine or
surgeries were subsequently modified to optimize surgical pyrimidine production, and having major impacts on both
outcomes and survival. However, alternative theories based solid and hematologic malignancies. Five-fluorouracil in
on biological principles, which seemed radical at the time, particular, a pyrimidine antimetabolite, has revolutionized
challenged widely accepted models, and provided important the treatment of gastrointestinal malignancies, especially
foundations for the modern understanding of cancer biology. colorectal cancer [16].
Furthermore, the development of efficacious treatment As these aforementioned categories of drugs were being
options like radiation and systemic therapies, the addition of modified and expanded upon, the next few decades wit-
cancer screening tools, and innovations in molecular science nessed the introduction of several novel cytotoxic chemo-
have changed how we view cancer biology and treatment. therapeutic classes. These included the Vinca alkaloids
(vincristine, vinblastine), which function to destabilize
microtubules and are used today for treating ALL, lympho-
35.3 Evolution of Systemic Therapy mas, metastatic breast cancer and small cell lung cancer,
topoisomerase inhibitors (irinotecan), which inhibit DNA
First introduced in the mid-1900s, initial systemic chemo- replication and are used today for treating colorectal cancer
therapeutic agents were broadly used for many types of can- and ovarian cancer, and taxanes (paclitaxel, docetaxel),
cer. These drugs were nondiscriminatory against which cells which impair microtubule dynamics and are frequently used
they acted upon, resulting in severe toxicities and poor toler- for treating patients with metastatic breast or pancreatic can-
ance [9]. As our knowledge of cellular biology progressed in cer, non-small cell lung cancer, and gastric cancer [17, 18].
parallel with medical technological advancements, these Antibiotics with antitumor effects were discovered and
treatments have evolved into the systemic agents of the mod- include anthracyclines (doxorubicin, daunorubicin), which
ern era which have the ability to precisely target individual act on DNA topoisomerase, and bleomycin, which is com-
cellular components and can be modified to fit patient-­ monly used today in treating lymphomas, testicular cancer,
specific genetic mutations. and cervical cancer [19].
Systemic chemotherapy was first studied during World Starting in the 1980s and continuing through the present
War I when scientists discovered on autopsy evaluations that time, rapid advances in technology led to landmark discover-
mustard gas, at that time being utilized as a chemical warfare ies in the fields of immunology, biochemistry, and cellular
agent, led to extensive myelosuppression in those poisoned biology. These findings have allowed for the highly sophisti-
[10]. Further research identified these nitrogen and sulfur cated understanding of the molecular mechanisms driving
mustard compounds to be DNA-alkylating agents, which cancer development that we hold today, and have permitted
formed covalent bonds with purines and induced apoptosis the identification of numerous molecular compounds that act
[11]. These and other nitrogen mustards, including cyclo- as therapeutic targets including growth factors, signaling
phosphamide, became first-line agents for treating lympho- molecules, and cell-cycle proteins. Key to these new drugs is
mas, leukemias, and sarcomas, and some of these drugs are their ability to pinpoint the products of mutated oncogenes
still used today [12, 13]. Discovery of additional alkylating and tumor suppressor genes while avoiding the wild-type
agents, comprising the well-known classes of alkyl sulfates products, thereby markedly reducing the toxic effects of sys-
(busulfan), triazenes (dacarbazine), and platinum com- temic therapy that are observed with cytotoxic agents [9, 20].
pounds, ensued, and many of these agents remain mainstay One of the most influential breakthroughs at the begin-
components of modern systemic treatment regimens [14]. ning of the targeted therapeutic era was the development of
The next major breakthrough came in the form of anti- the tyrosine kinase inhibitor imatinib mesylate (Gleevec),
metabolite therapy, which was initially developed in the late which targets the BCR-ABL receptor, c-KIT, and platelet-­
1940s. Antifolate compounds were the first antimetabolites derived growth factor receptor. Used to treat chronic myelog-
studied and demonstrated the ability to induce remission in enous leukemia and gastrointestinal stromal tumors, this
patients with acute lymphoblastic leukemia (ALL) [15]. drug was shown to be slowly metabolized, well-absorbed
Initially not understood, the mechanism behind antifolate when taken orally, and incited minimal adverse events [21].
therapy was eventually discovered to be permanent inhibi- It was the prototypical example for all future targeted thera-
tion of dihydrofolate reductase (DHFR), leading to the pies to be modeled after and paved the way for the develop-
inability of cells to complete purine production with subse- ment of agents targeting epidermal growth factor receptor,
quent apoptosis. Methotrexate is one of the most well- MAPK/MEK pathway, vascular endothelial growth factor
380 R. J. Straker III et al.

receptor, mTOR, and many more that are used to treat a vari- ACOSOG Z0011 trial demonstrated no survival benefit for
ety of both solid and hematologic malignancies [18]. performing lymphadenectomy in certain positive sentinel
The next epochal turning point was the development of node patients, thereby further reducing the role of surgery in
immune-mediated therapies in the form of monoclonal anti- the locoregional management of specific breast cancer
bodies directed at cellular checkpoint pathways, which func- patients [6]. In addition, the introduction of screening mam-
tion to regulate host immune responses. Termed immune mography at the end of the twentieth century enabled earlier
checkpoint inhibitors, the most common agents act against detection of breast cancers [10] and identified patients more
cytotoxic T-lymphocyte-associated antigen 4 (ipilimumab) likely to have disease restricted to the primary site, thus lead-
and programmed cell death protein 1 (nivolumab, pembroli- ing to the “spectrum theory” in which breast cancer was con-
zumab), both of which lead to the activation of pathways that sidered a heterogeneous disease with a variety of
downregulate the host immune system. By blocking these presentations. Tumors may remain locoregional or possess a
inhibitory pathways, these immune checkpoint inhibitors higher propensity for spread and progression [27].
prevent tumors from taking advantage of them and evading For melanoma, early surgical management was also pal-
immune detection [22]. Ipilimumab, the first approved of liative with techniques that ranged from cauterization to
these agents, is currently used alone or in combination with amputation [27, 28]. Extreme margins of 4–5 cm were used
nivolumab for unresectable or metastatic melanoma, and is to control the primary site initially, but these margins were
also being evaluated in the management of renal cancer, non-­ subsequently decreased to 1–2 cm based on several impor-
small cell lung cancer, and other types of cancer [23]. tant clinical trials that showed no significant differences
Nivolumab and pembrolizumab are frequently used for the in local recurrence rates with narrower margins. In addition,
treatment of locoregionally advanced and metastatic mela- in the late nineteenth century, surgeons recognized that mel-
noma and for non-small cell lung cancer, along with urothe- anomas metastasized to regional lymph nodes. Herbert Snow
lial carcinoma [9, 24]. recommended the addition of lymphadenectomy in all mela-
noma excisions in 1892 [27]. Analogous to Halsted’s radical
mastectomy, the en bloc approach for melanoma, in which
35.4  volution of Surgical Management
E the primary melanoma was removed along with the draining
of Cancer nodal basin, whether therapeutically or electively, remained
the standard of care for many decades [29].
The surgical management of various cancer types has The development of sentinel lymph node biopsy (SLNB)
evolved and changed as our understanding of cancer biology was another important advancement in melanoma surgery.
has improved and adjunct therapies have become available. Based on the theory of lymphatic spread, Donald Morton
For instance, the introduction of the radical mastectomy by demonstrated that determining the status of the SLN pro-
Halsted in the late nineteenth century represented an impor- vided the status of the entire draining nodal basin using a less
tant shift in breast cancer management. Halsted’s extensive technique. Positive SLN patients underwent com-
­predecessors performed a variety of anatomical excisions for plete lymph node dissection (CLND), whereas negative SLN
breast tumors, but these operations had high rates of compli- patients were spared unnecessary lymphadenectomies.
cations and disease recurrence [25]. As a result, early breast However, the biology of the SLN remained debated with two
cancer surgeries were generally palliative. Halsted found that theories that were developed [30]. The “incubator hypothe-
many patients developed cancer recurrences near the initial sis” theorized that tumor cells spread in an orderly fashion
primary tumor site, and he theorized that cancer spread con- from the primary site to the regional lymph nodes and then to
tiguously from local, to regional, and then to distant sites [6, distant sites. Therefore, control of disease at certain levels
25]. He subsequently developed the radical, or en bloc, could potentially prevent further disease progression. In con-
method to remove all vulnerable tissue for locoregional con- trast, the “marker hypothesis” suggested that tumor cells in
trol before the disease spread elsewhere [6, 25]. Halsted’s an SLN served as an early marker of systemic disease spread.
patients had lower rates of local recurrence [6], and as a In this theory, control of nodal disease would not affect the
result, the radical mastectomy was considered the standard ultimate outcome of patients with nodal metastases. DeCOG-­
of care for much of the next century [7]. However, surgeons SLT and Multicenter Selective Lymphadenectomy Trial-II
found that improved locoregional control did not necessarily provided support for the marker hypothesis by demonstrat-
improve survival [26]. ing no differences in survival between positive SLN patients
Fisher’s challenging of Halsted’s theory of cancer spread who were treated with CLND versus nodal observation [29].
with his own systemic theory resulted in the shift of breast Despite these results, the biological relevance of the SLN
cancer surgery from a radical approach to more conservative continues to be debated. More recently, metastasectomy is
techniques like lumpectomy in conjunction with radiother- being used more frequently in appropriately selected patients
apy and systemic treatments [6]. Subsequently, results of the with stage IV melanoma, given the availability of effective
35 The Efficacy and Evolution of Surgical Management Based on Cancer Biology 381

systemic therapies, particularly immune checkpoint inhibi- dures. Mikulicz devised a two-stage colon resection that
tors [31]. Even so, surgery for early-stage melanoma, like for included mesenteric lymphadenectomy based on his theory
breast cancer, has continued to become less extensive both at of lymphatic spread. Subsequently, in 1908, Ernest Miles
the primary site and at the level of the regional lymph nodes published a more extended approach of lymphadenectomy
as we have increased our understanding of cancer biology. with abdominoperineal resection (APR) to include proximal
The theory of lymphatic spread also guided the evolution lymph nodes, which he postulated were responsible for local
of the radical gastrectomy. From Theodor Billroth’s first suc- recurrences [35]. However, surgeons challenged that APR
cessful distal gastrectomy in 1881, surgeons described vary- was too radical for all cases of colorectal carcinoma. This led
ing degrees of anatomical resection of the stomach. Most to the development of more conservative techniques that
based surgical techniques on the idea of disease progression included Henri Hartmann’s end colostomy to salvage the dis-
through the stomach wall layers to adjacent lymphatics [31]. tal rectum and Claude Dixon’s creation of the sphincter-­
Throughout the next several decades, the concept of radical sparing anterior resection [37, 38].
resection would continue to develop as surgeons performed Consistent with the importance of locoregional control in
en bloc operations to include adjacent organs, such as the en bloc resections, Bill Heald introduced total mesorectal
pancreas and spleen. Despite these innovations, survival excision (TME) for treating rectal cancers in 1982. TME pre-
rates for patients with gastric cancer remained poor. It was served the embryological plane in the mesentery to prevent
not until the mid-1900s that surgeons determined that these disruption of lymphatics and the release of tumor cells [38].
risks exceeded the safety and benefit of gastrectomy. For Although colon and rectal cancers were historically
instance, randomized controlled trials in the late twentieth approached in a similar manner, the inclusion of adjunct
century demonstrated that there was no survival difference in treatments has led to their acknowledgment as two distinct
patients who underwent total versus subtotal gastrectomy, entities. For instance, further locoregional control for pri-
which supported a less radical surgical approach. mary rectal carcinoma can be achieved with neoadjuvant
Furthermore, clinical trials evaluated the extent of lymphad- radiation therapy [37], and may alter the extensiveness of the
enectomy for gastric cancer. Fifteen-year follow-up of the planned surgical resection. Finally, hepatic metastasectomy
Dutch D1D2 trial showed significant improvements in recur- may be appropriate in select cases of limited-stage IV
rence and death rates after a D2 resection, consisting of a colorectal cancer and can be performed simultaneously or
more extensive lymph node dissection, over a D1 resection metachronously with primary colon surgery [38]. Adjuvant
[32, 33]. However, the extent of lymphadenectomy for gas- systemic chemotherapy is also utilized and has been shown
tric cancer continues to be controversial with the develop- to improve survival in colon cancer patients with nodal
ment of D3 lymphadenectomies, which are often performed involvement [39, 40].
in Asian countries [34]. Pelvic exenteration is one of the most radical forms of
Another pioneering en bloc operation by Halsted was the surgery that is still performed in the modern world. It was
first radical resection for periampullary carcinoma in 1898. first introduced by Brunschwig for recurrent cervical carci-
Allen Whipple subsequently modeled his approach on noma in 1948. This en bloc resection of multiple pelvic
Halsted’s theories in a two-step operation that was later organs, employed for locally advanced tumors, was initially
amended to a more radical one-step procedure [33]. The lat- associated with high rates of mortality and was performed
ter, named the standard Whipple, was successfully com- for palliation [39, 40]. However, advances in perioperative
pleted by Alexander Brunschwig in 1937. In contrast, care and imaging techniques have assisted with operative
Traverso and Longmire proposed a similar procedure with planning and patient selection, ultimately leading to
reduced margins, known as the pylorus-preserving pancre- improved postoperative outcomes. Similarly, the radical hys-
aticoduodenectomy. Despite concerns that microscopic dis- terectomy was a mainstay in the treatment of cervical cancer
ease may persist with decreased margins, trials showed no for many decades [41]. Although modified forms of this
differences in morbidity, mortality, or survival between the operation continue to be utilized for locoregional treatment
standard Whipple and a total pancreatectomy. Advances in of select cases of gynecologic cancers, the addition of or pri-
technique, such as en bloc vascular resection with recon- mary use of radiation therapy has allowed for more conser-
struction, as well as the use of neoadjuvant therapies in bor- vative treatment approaches.
derline resectable cases, have led to enhanced locoregional Surgery for genitourinary cancers has also undergone a
control and identified more patients who are amenable to similar transformation. For instance, radical prostatectomy,
surgical resection [35, 36]. first performed in 1904, was the standard surgical treatment
Similarly, early surgical management of colorectal can- for prostate cancer for several decades [42]. Moreover, the
cers was initially palliative due to high recurrence rates. retropubic approach, suggested by Terrence Millin in 1945,
However, more radical versions of these operations would enabled surgeons to access pelvic lymph nodes. Subsequently,
transform colorectal surgery into potentially curative proce- transurethral resection of the prostate was introduced as a
382 R. J. Straker III et al.

less extensive method for treating certain prostate cancers intraoperative imaging modalities, and detecting new and
[42]. However, unlike other cancers of the pelvis, many pros- downstream activating mutations in currently targeted
tate cancers are now treated medically, specifically with molecular pathways, which may be responsible for treatment
androgen-ablation therapy, and surgery has a lesser role in resistance [45].
the management of prostate cancer [42]. As previously discussed, our understanding of the biology
underlying cancer development continues to progress. In con-
trast to Halsted’s theory, it is abundantly clear that simply
35.5 Future Perspectives of Surgery removing the malignant disease is not sufficient in many
in Cancer Management cases, as this does not address microscopic metastatic disease
which may be present and is ultimately the cause of most
Monumental innovations in human genetics and biomedical cancer-related deaths. Likewise, using toxic chemicals at con-
technology have led to the development of cutting-edge tech- centrations necessary to eradicate every malignant cell causes
niques for the diagnosis and characterization of cancer. intolerable toxicities to their nonmalignant counterparts. New
Remarkably accurate, these novel assays can identify patient-­ and emerging treatments will need to focus on exploiting
to-­patient variations for specific cancers at the molecular and molecular and genetic differences between malignant and
genomic level with extreme precision. This allows for indi- nonmalignant cells, such as blocking constitutively activated
vidually tailored treatment protocols based on pathologic cellular receptors on cancer cells, preventing cancer cells
responses to therapies or dynamic changes in levels of bio- from utilizing growth factors or cellular machinery, and
markers. In light of these developments and in line with the blocking cancer cells from hijacking regulatory immune
shift from a more radical to conservative approach, the role mechanisms to evade detection. However, nonimmune sys-
of surgery in the treatment of cancer remains essential but is temic therapies to date generally fall short of demonstrating
evolving. the capacity to completely exterminate all areas of malignant
The ability to sequence the human genome led to the disease [46].
development of the DNA microarray. First described in As we enter the modern era of cancer treatment, evidence
1995, this groundbreaking diagnostic assay allows for the is emerging that combination therapy, both in the form of
characterization of genome-wide gene expression at any multidrug chemotherapy regimens and multimodal treatment
given point in time. Initial studies with yeast and bacteria approaches combining systemic treatment, radiation therapy,
demonstrated the wide-reaching potential of this new tech- and surgery, are associated with improved outcomes [47].
nology, including its capability to detect induction and Neoadjuvant therapy is also becoming increasingly com-
repression of genes due to environmental stresses and its mon, with subsequent therapies dependent on the response to
ability to bring to light novel cellular regulatory pathways by the initial treatment [48]. As all aspects of cancer manage-
the detection of transcription and other regulatory factors ment are evolving, so too is the role of the surgeon in the
based on upstream gene expression patterns [43]. cancer patient’s care. While surgical resection is still the
Some of the most noteworthy applications of DNA micro- mainstay of treatment for most early-stage malignancies, the
arrays have been with its use to define cancers by their gene scope of the modern surgeon is no longer simply that of a
expression profiles (GEPs). Gene profiling allows for accu- technician tasked with excising a malignancy. He or she now
rate identification of tumors of unknown etiology and can has an active role in all stages of the treatment spectrum,
detect subtle molecular differences between two otherwise including preventive counseling and screening, initial cancer
seemingly similar cancers, thereby leading to individualized diagnosis, collaboration with multidisciplinary teams to
treatment regimens and significant improvements in out- determine the optimal sequence of treatment and interven-
comes [44]. Changes in GEPs following therapy administra- tions, postoperative surveillance monitoring, end-of-life
tion can provide important prognostic information, guide care, and high-quality research [49]. Trials have already
therapy, and identify novel biomarkers and therapeutic drug begun studying the extent of surgical resection based on
targets. In modern practice, gene profiling has been instru- pathologic response to neoadjuvant treatment, and as the
mental in the management of breast cancer and is commonly efficacy of systemic therapy continues to improve, more
utilized in the treatment of lung, prostate, and colon cancers, patients who initially would have had unresectable disease
among many others [44]. Future research with gene profiling will now become surgical candidates [50]. Just as with the
and other molecular techniques will likely focus on combin- other disciplines of care for the cancer patient, the role of
ing these already established methods with high-throughput surgical therapy will continue to evolve alongside the entire
technologies, integrating tumor molecular patterns with field of oncology.
35 The Efficacy and Evolution of Surgical Management Based on Cancer Biology 383

35.6 Conclusions References

• As our comprehension of cancer biology continues to 1. Onuigbo WI. Historical trends in cancer surgery. Med Hist.
1962;6:154–61. https://doi.org/10.1017/s0025727300027125.
evolve, so too does our armamentarium of therapy options. 2. Halsted WS. I. The results of operations for the cure of cancer of
• We now know that cancer does not only ascribe to the the breast performed at the Johns Hopkins Hospital from June,
initial Halstedian dogma of strict local to regional to dis- 1889, to January, 1894. Ann Surg. 1894;20(5):497–555. https://doi.
tant spread; yet, the Fisher model of cancer as an inher- org/10.1097/00000658-­189407000-­00075.
3. Fisher B. Laboratory and clinical research in breast cancer—a per-
ently systemic disease at presentation falls short of sonal adventure: the David A. Karnofsky memorial lecture. Cancer
accurately capturing the complexity of cancer biology for Res. 1980;40(11):3863.
all tumor types. 4. Leong SP, Tseng WW. Micrometastatic cancer cells in lymph
• In accordance with the spectrum theory, it is likely that nodes, bone marrow, and blood: clinical significance and biologic
implications. CA Cancer J Clin. 2014;64(3):195–206. https://doi.
each unique cancer type, and subtypes within these cate- org/10.3322/caac.21217.
gories, has specific molecular and biological attributes, 5. Lawrence W, Lopez MJ. Radical surgery for cancer: a historical
allowing it to develop and progress in a distinct fashion; perspective. Surg Oncol Clin N Am. 2005;14(3):441–6., , v. https://
progression of a particular cancer is also likely contingent doi.org/10.1016/j.soc.2005.05.003.
6. Özmen V. Paradigm shift from Halstedian radical mastectomy to
upon the host (immune) factors. personalized medicine. J Breast Health. 2017;13(2):50–3. https://
• Surgery remains an integral component to oncologic doi.org/10.5152/tjbh.2017.312017.
treatment, but it is important to remember that it is but one 7. Hellman S. Karnofsky memorial lecture. Natural history of small
of many treatment options we have at hand, and should be breast cancers. J Clin Oncol. 1994;12(10):2229–34. https://doi.
org/10.1200/JCO.1994.12.10.2229.
used in conjunction with, not exclusively from, other 8. Wilky BA. Immune checkpoint inhibitors: the linchpins of mod-
available treatment modalities. ern immunotherapy. Immunol Rev. 2019;290(1):6–23. https://doi.
• The development of technology to detect patient-specific org/10.1111/imr.12766.
molecular aberrations and identify novel biomarker and 9. Chabner BA, Roberts TG. Timeline: chemotherapy and the war on
cancer. Nat Rev Cancer. 2005;5(1):65–72. https://doi.org/10.1038/
therapeutic targets will allow for individualized, precision nrc1529.
oncologic care. 10. Gilman A, Philips FS. The biological actions and therapeu-
• Ongoing and future research will strive to answer: tic applications of the B-Chloroethyl amines and sulfides.
Science. 1946;103(2675):409–36. https://doi.org/10.1126/
science.103.2675.409.
11. Brookes P, Lawley PD. The reaction of mustard gas with nucleic
acids in vitro and in vivo. Biochem J. 1960;77(3):478–84. https://
Open Questions
doi.org/10.1042/bj0770478.
• Is there a synergistic effect of combining immune-­ 12. Rhoads CP. Nitrogen mustards in the treatment of neoplastic dis-
mediated or targeted molecular therapies together ease; official statement. J Am Med Assoc. 1946;131:656–8. https://
or with traditional chemotherapeutic agents? doi.org/10.1001/jama.1946.02870250010003.
13. Emadi A, Jones RJ, Brodsky RA. Cyclophosphamide and can-
• How will the use of intraoperative or improved cer: golden anniversary. Nat Rev Clin Oncol. 2009;6(11):638–47.
more sensitive preoperative molecular imaging https://doi.org/10.1038/nrclinonc.2009.146.
techniques change the landscape of oncologic 14. Puyo S, Montaudon D, Pourquier P. From old alkylating agents to
surgery? new minor groove binders. Crit Rev Oncol Hematol. 2014;89(1):43–
61. https://doi.org/10.1016/j.critrevonc.2013.07.006.
• How will pathologic response to neoadjuvant treat- 15. Farber S, Diamond LK. Temporary remissions in acute leukemia
ment be used to influence subsequent cancer treat- in children produced by folic acid antagonist, 4-aminopteroyl-­
ment including extensiveness of surgical resection? glutamic acid. N Engl J Med. 1948;238(23):787–93. https://doi.
• How will the discovery of novel biomarkers reshape org/10.1056/NEJM194806032382301.
16. Carrillo E, Navarro SA, Ramírez A, et al. 5-Fluorouracil deriva-
treatment regimens and therapy duration? tives: a patent review (2012–2014). Expert Opin Ther Pat.
• Will the development of more precise and 2015;25(10):1131–44. https://doi.org/10.1517/13543776.2015.105
molecular-­ based systemic therapies allow more 6736.
patients with advanced-stage disease to become 17. Rosen LS. Irinotecan in lymphoma, leukemia, and breast, pan-
creatic, ovarian, and small-cell lung cancers. Oncology (Williston
candidates for curative surgical resection? What Park). 1998;12(8 Suppl 6):103–9.
will the role of surgery be for the removal of cancer 18. Falzone L, Salomone S, Libra M. Evolution of cancer phar-
sites resistant to systemic therapies? macological treatments at the turn of the third millennium.
Review. Front Pharmacol. 2018;9:1300. https://doi.org/10.3389/
fphar.2018.01300.
19. Bolzán AD, Bianchi MS. DNA and chromosome damage
Disclosures The authors declare no conflicts of interest or external induced by bleomycin in mammalian cells: an update. Mutat Res.
funding for this publication. 2018;775:51–62. https://doi.org/10.1016/j.mrrev.2018.02.003.
384 R. J. Straker III et al.

20. Pucci C, Martinelli C, Ciofani G. Innovative approaches for 36. Lirici MM, Hüscher CGS. Techniques and technology evolution
cancer treatment: current perspectives and new challenges. of rectal cancer surgery: a history of more than a hundred years.
Ecancermedicalscience. 2019;13:961. https://doi.org/10.3332/ Minim Invasive Ther Allied Technol. 2016;25(5):226–33. https://
ecancer.2019.961. doi.org/10.1080/13645706.2016.1198381.
21. Waller CF. Imatinib mesylate. Recent Results Cancer Res. 37. Rentsch M, Schiergens T, Khandoga A, Werner J. Surgery
2018;212:1–27. https://doi.org/10.1007/978-­3-­319-­91439-­8_1. for colorectal cancer—trends, developments, and future
22. Seidel JA, Otsuka A, Kabashima K. Anti-PD-1 and anti-CTLA-4 perspectives. Visc Med. 2016;32(3):184–91. https://doi.
therapies in cancer: mechanisms of action, efficacy, and limitations. org/10.1159/000446490.
Front Oncol. 2018;8:86. https://doi.org/10.3389/fonc.2018.00086. 38. Tamas K, Walenkamp AM, de Vries EG, et al. Rectal and colon
23. Sakamuri D, Glitza IC, Betancourt Cuellar SL, et al. Phase I cancer: not just a different anatomic site. Cancer Treat Rev.
dose-escalation study of anti-CTLA-4 antibody ipilimumab and 2015;41(8):671–9. https://doi.org/10.1016/j.ctrv.2015.06.007.
lenalidomide in patients with advanced cancers. Mol Cancer 39. Brown KGM, Solomon MJ, Koh CE. Pelvic Exenteration surgery:
Ther. 2018;17(3):671–6. https://doi.org/10.1158/1535-­7163. the evolution of radical surgical techniques for advanced and recur-
MCT-­17-­0673. rent pelvic malignancy. Dis Colon Rectum. 2017;60(7):745–54.
24. Palumbo MO, Kavan P, Miller WH, et al. Systemic cancer ther- https://doi.org/10.1097/DCR.0000000000000839.
apy: achievements and challenges that lie ahead. Front Pharmacol. 40. Sevin BU, Koechli OR. Pelvic exenteration. Surg Clin
2013;4:57. https://doi.org/10.3389/fphar.2013.00057. North Am. 2001;81(4):771–9. https://doi.org/10.1016/
25. Sakorafas GH, Safioleas M. Breast cancer surgery: s0039-­6109(05)70164-­x.
an historical narrative. Part II. 18th and 19th centu- 41. Dursun P, Gultekin M, Ayhan A. The history of radical hyster-
ries. Eur J Cancer Care. 2010;19(1):6–29. https://doi. ectomy. J Low Genit Tract Dis. 2011;15(3):235–45. https://doi.
org/10.1111/j.1365-­2354.2008.01060.x. org/10.1097/LGT.0b013e31820eb038.
26. Cotlar AM, Dubose JJ, Rose DM. History of surgery for breast can- 42. Denmeade SR, Isaacs JT. A history of prostate cancer treatment.
cer: radical to the sublime. Curr Surg. 2003;60(3):329–37. https:// Nat Rev Cancer. 2002;2(5):389–96. https://doi.org/10.1038/nrc801.
doi.org/10.1016/S0149-­7944(02)00777-­8. 43. Brewster JL, Beason KB, Eckdahl TT, Evans IM. The micro-
27. Rebecca VW, Sondak VK, Smalley KS. A brief history of melanoma: array revolution: perspectives from educators. Biochem
from mummies to mutations. Melanoma Res. 2012;22(2):114– Mol Biol Educ. 2004;32(4):217–27. https://doi.org/10.1002/
22. https://doi.org/10.1097/CMR.0b013e328351fa4d. bmb.2004.494032040362.
28. Gonzalez A. Sentinel lymph node biopsy: past and present impli- 44. Pierouli K, Mitsis T, Papakonstantinou E, Vlachakis D. Introductory
cations for the management of cutaneous melanoma with nodal chapter: gene profiling in cancer in the era of metagenomics and
metastasis. Am J Clin Dermatol. 2018;19(1):24–30. https://doi. precision medicine; 2019.
org/10.1007/s40257-­018-­0379-­0. 45. Szalat R, Avet-Loiseau H, Munshi NC. Gene expression pro-
29. Faries MB, Han D, Reintgen M, Kerivan L, Reintgen D, Caracò files in myeloma: ready for the real world? Clin Cancer Res.
C. Lymph node metastasis in melanoma: a debate on the signifi- 2016;22(22):5434–42. https://doi.org/10.1158/1078-­0432.
cance of nodal metastases, conditional survival analysis and clini- CCR-­16-­0867.
cal trials. Clin Exp Metastasis. 2018;35(5–6):431–42. https://doi. 46. Somarelli JA, Gardner H, Cannataro VL, et al. Molecular biology
org/10.1007/s10585-­018-­9898-­6. and evolution of cancer: from discovery to action. Mol Biol Evol.
30. Lee DJ, Faries MB. Practical manual for dermatologic and surgical 2020;37(2):320–6. https://doi.org/10.1093/molbev/msz242.
melanoma management. 1st ed. Springer International Publishing; 47. Wolmark N, Colangelo L, Wieand S. National surgical adju-
2020. vant breast and bowel project trials in colon cancer. Semin
31. Zhang CD, Yamashita H, Seto Y. Gastric cancer surgery: his- Oncol. 2001;28(1 Suppl 1):9–13. https://doi.org/10.1016/
torical background and perspective in Western countries versus s0093-­7754(01)90245-­3.
Japan. Ann Transl Med. 2019;7(18):493. https://doi.org/10.21037/ 48. Imyanitov EN, Yanus GA. Neoadjuvant therapy: theoreti-
atm.2019.08.48. cal, biological and medical consideration. Chin Clin Oncol.
32. Are C, Dhir M, Ravipati L. History of pancreaticoduode- 2018;7(6):2.
nectomy: early misconceptions, initial milestones and the 49. Elder K, Barber M. The role of the surgeon in cancer care.
pioneers. HPB (Oxford). 2011;13(6):377–84. https://doi. Surgery (Oxford). 2018;36. https://doi.org/10.1016/j.
org/10.1111/j.1477-­2574.2011.00305.x. mpsur.2017.12.005.
33. Griffin JF, Poruk KE, Wolfgang CL. Pancreatic cancer surgery: 50. Reijers ILM, Dimitriadis P, Rozeman EA, et al. Personalized com-
past, present, and future. Chin J Cancer Res. 2015;27(4):332–48. bination of neoadjuvant domatinostat, nivolumab and ipilimumab
https://doi.org/10.3978/j.issn.1000-­9604.2015.06.07. in macroscopic stage III melanoma patients stratified according
34. Douridas GN, Pierrakakis SK. Is there any role for D3 lymphad- to the interferon-gamma signature: the DONIMI study. J Clin
enectomy in gastric cancer? Front Surg. 2018;5:27. https://doi. Oncol. 2020;38(15_suppl):TPS10087. https://doi.org/10.1200/
org/10.3389/fsurg.2018.00027. JCO.2020.38.15_suppl.TPS10087.
35. Galler AS, Petrelli NJ, Shakamuri SP. Rectal cancer surgery: a brief
history. Surg Oncol. 2011;20(4):223–30. https://doi.org/10.1016/j.
suronc.2010.01.001.
 The Role of Surgery in Managing
Primary and Metastatic Melanoma 36
Kristen E. Rhodin, Kirsten Baecher, Winta T. Mehtsun,
Mike Lowe, Genevieve Boland, and Georgia M. Beasley

Abstract
Learning Objectives
This chapter describes updates in the surgical manage- After reading this chapter, readers should be able to:
ment of melanoma including primary melanoma, regional
melanoma, and distant metastatic disease. Surgical man- • Describe appropriate margins for primary
agement is discussed in the context of a multidisciplinary melanoma
approach. Specifically, surgical therapy needs to be con- • Understand indications for sentinel lymph node
sidered in the context of new effective systemic therapy biopsy
agents for melanoma. With the approval of multiple new • List options for management of regionally advanced
treatment strategies for melanoma, the surgical approach melanoma
has changed dramatically in the last 10 years. Here we • Describe neoadjuvant approaches for resectable
review the rationale for appropriate contemporary surgi- melanoma.
cal approaches to melanoma including minimizing mor-
bidity from surgery and introduction of neoadjuvant
therapy approaches.
36.1 Introduction

The incidence of cutaneous melanoma has been rapidly

increasing in recent years [1]. In the last decade (2010–
2020), there have also been dramatic changes in both the sur-
gical and medical management of patients with melanoma.
Given the introduction of effective systemic therapy for mel-
K. E. Rhodin · G. M. Beasley (*) anoma in the last decade, the current approach to surgery
Department of Surgery, Duke University, Durham, NC, USA needs to be viewed as just one component of an effective
e-mail: [email protected]; [email protected] strategy to minimize disease mortality.
K. Baecher · M. Lowe
Department of Surgery, Emory University, Atlanta, GA, USA
e-mail: [email protected]; [email protected]
36.2 Background
W. T. Mehtsun
Department of Surgery, Massachusetts General Hospital,
A number of distinct features of the primary melanoma are
Boston, MA, USA
well known to be prognostic of outcomes—notably Breslow
Department of Surgery, Brigham and Women’s Hospital,
thickness and the presence of ulceration [2]. Metastatic
Boston, MA, USA
spread of cutaneous melanoma usually occurs as regional
Dana Farber Cancer Institute, Boston, MA, USA
lymph node metastases, locoregional metastases, or distant
e-mail: [email protected]
metastases (in 50%, 20%, and 30% of patients with recurrent
G. Boland
melanoma, respectively) [3]. The presence of lymph node
Department of Surgery, Massachusetts General Hospital,
Boston, MA, USA metastases either detected clinically or via sentinel lymph
e-mail: [email protected] node biopsy (SLNB), and the presence of distant metastases

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 385
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_36
386 K. E. Rhodin et al.

is also critical components in the staging system with increas- nomas greater than 2 mm in depth. Tissue removed should
ing disease burden associated with worse outcomes. include skin and subcutaneous tissue to the level of the fascia
Increasingly, host immunity and molecular profiling are which contains all local lymphatic tissue. There are ongoing
also recognized as key determinants of melanoma biology. efforts to determine if 1 cm margins may be adequate in
Recently, several somatic mutations have been identified patients regardless of Breslow depth. Long-term data from a
causing deregulation of key intracellular signaling pathways large randomized trial in the United Kingdom comparing
important in the proliferation of melanoma, including acti- 1–3 cm margins for patients with primary melanomas at least
vating mutations in the BRAF serine/threonine protein 2 mm in thickness found that melanoma-specific survival
kinase in approximately 50% of cases [4], activating muta- was significantly worse for those in the 1 cm group com-
tions in NRAS, and C-KIT mutations. This has led to the use pared to the 3 cm margin group (unadjusted hazard ratio,
and success of targeted therapy for appropriate melanoma 1.24; 95% CI 1.01–1.53; p = 0.041) [11]. However, surgical
patients, including combination BRAF/MEK inhibitors [5]. complications were nearly double in the 3 cm group (8% vs.
Additionally, the degree of host lymphoid infiltration into the 15%) and SLN biopsy (now considered standard of care for
primary tumor and into metastatic lesions may be important this thickness group) was not routinely performed [11]. To
in disease control and response to modern therapies, which further investigate the optimal margins for thicker melano-
include immune checkpoint inhibitors (ICI) [6]. Since 2011, mas, the MelmarT-I study was conducted as an international
multiple new immune therapies and targeted therapies have randomized study comparing 1–2 cm margins in patients
been FDA (United States Food and Drug Administration) with greater than 1 mm thick tumors [12]. Patients in the
approved for melanoma based on encouraging clinical trial 2 cm margin group more often required reconstruction (34.9
results for those therapies [7]. vs. 13.6%; p < 0.0001) and had an increased wound necrosis
While complete surgical resection of melanoma offered rate (3.6% vs. 0.5%; p = 0.036) compared to the 1 cm margin
one of the few options of effective disease control in the past, group. The MelMarT-II study (NCT03860883) is ongoing.
surgeons managing melanoma in the modern era now have Patients with >2 mm thickness or 1- to 2-mm thickness with
updated evidence-based guidelines for the surgical manage- ulceration are being randomized to either 1 or 2 cm margins.
ment of the primary melanoma lesion and the regional nodal The primary outcome is disease-free survival with secondary
basin and for the current role of metastasectomy. In this outcomes that include local recurrence, melanoma-specific
chapter, we will highlight key recent updates, especially with survival, and surgery-related adverse events. This trial should
regard to the surgical management of the draining nodal further clarify the application of 1 cm margins to patients
basin. with melanoma regardless of Breslow depth.
Certain anatomic sites including but not limited to digits,
ears, the face, and the plantar surface of the foot may not be
36.2.1 Management of Primary Melanoma as easily amenable to rigid surgical criteria without signifi-
cant concerns for poor cosmetic and functional outcomes.
In 1907, Dr. Handley described extensive surgical resection For subungual lesions, conservative amputation at the most
of skin and subcutaneous tissue for melanoma [8]. For distal interphalangeal joint is appropriate. Certain melanoma
decades, a wide variation existed in surgical practice in part subtypes also may prompt special consideration. Wider mar-
due to wide variability in melanoma biology. Several retro- gins (2 cm) are generally considered for desmoplastic mela-
spective studies then found that resection margins of 2 cm or noma, particularly for the pure subtype of desmoplastic
less in melanomas with Breslow depth <1 mm were associ- melanoma, given that 1 cm margins have been associated
ated with low rates of local recurrence [8]. The World Health with higher rates of local recurrence [13].
Organization published the first prospective randomized trial Among patients with clinical stage I and II melanoma,
supporting adequate disease control with narrow margins metastasis to the sentinel node is one of the strongest prog-
(1 cm) compared to 3 cm margins for melanomas less than nostic factors for melanoma [14]. The application of SLNB
2 mm in thickness [9]. Disease-free and overall survival rates to melanoma by Morton in 1992 proved to be practice chang-
were similar between the two groups. Other randomized tri- ing. In the multicenter selective lymphadenectomy trial 1
als completed around the world also supported more conser- (MSLT-1), patients undergoing wide excision for melanoma
vative margins [10]. A Cochrane review supported that were randomized to receive either SLNB with immediate
margins greater than 2 cm could safely be avoided [10]. completion lymph node dissection (CLND) for a positive
Based on data from these clinical trials, current National SLN or observation with interval therapeutic lymph node
Comprehensive Care Network (NCCN) guidelines recom- dissection (TLND) for a nodal relapse. There was no differ-
mend wide excision with 1 cm margins for melanomas ence in 10-year melanoma-specific survival between the two
1.0 mm or less, wide excision with 1–2 cm margin for mela- arms; however, in secondary analysis, SLNB did improve the
nomas 1.01–2 mm in thickness, and 2 cm margins for mela- 10-year rate of disease-free survival in patients with interme-
36 The Role of Surgery in Managing Primary and Metastatic Melanoma 387

diate thickness (1.2–3.5 mm) melanoma [14]. In patients median of 18 months after surgery and is associated with
with intermediate-thickness melanomas, the 10-year worse survival compared to those with localized disease
melanoma-­specific survival rate was 62.1 ± 4.8% among [19].
those with SLN metastasis versus 85.1 ± 1.5% for those ITM can become symptomatic with patients often report-
without SLN metastasis [14]. On multivariate analysis, SLN ing discomfort, pruritus, bleeding, and infection; however,
node status was the strongest predictor of both disease recur- its heterogeneous presentation has made its management
rence and death from melanoma. Recently, both the American challenging. Historically, surgical excision has had a limited
Society of Clinical Oncology and Society of Surgical role in the management of ITM—limited primarily to resec-
Oncology reviewed available data and published guidelines tion of few and small lesions. Unfortunately, ITM lesions are
for current SLNB indications [15]. SLNB is recommended frequently not as amenable to metastasectomy as isolated
for intermediate thickness T2 or T3 melanomas (1.0–4.0 mm distant metastases, with resection often being morbid or
thickness) and may be considered in T1b thin melanomas impractical [19]. The utility of SLNB for ITM is debated,
(0.8–1.0 mm or <0.8 mm thick with ulceration) [15]. although SLNB may refine the stage III subgroup staging
Furthermore, SLNB may be performed for patients with T4 and thus carry some minor prognostic value. In the modern
thick melanomas (>4.0 mm thickness) because SLNB era, patients with ITM are eligible for newer systemic ther-
appears to still provide prognostic data and a positive SLNB apy agents, given the Stage III classification regardless of
in the current era renders these patients eligible for effective SLN status.
systemic adjuvant therapy [7]. Although the rate of SLN For the majority of patients who are unable to undergo
metastasis is generally less than 5% for patients with lesions resection of ITM, its superficial nature lends accessibility for
less than 0.8 mm in thickness (T1a), there may be select regional and intralesional therapies. Regional chemotherapy
patients at higher risk for whom SLN may be considered. remains an option for ITM limited to an extremity. This strat-
Additional recent efforts have focused on identifying patients egy involves isolating the affected limb’s blood supply and
with greater than 1 mm thick tumors for whom the risk of administering high-dose chemotherapy. Two variants of this
SLN metastases is low, thereby potentially allowing these include hyperthermia isolated limb perfusion (HILP) and
patients to forego SLNB [16]. isolated limb infusion (ILI). While both provide high-dose
Approximately 10–20% of patients with tumors >1 mm regional chemotherapy, they differ in the invasiveness of vas-
in depth will have a positive SLN [14]. SLNB-based man- cular access, limb-threatening toxicity, and operative times
agement of patients resulted in improved melanoma-specific [19]. In HILP, the major vessels of the extremity are dis-
survival among patients with nodal metastases and interme- sected and cannulated, followed by the application of a tour-
diate depth melanomas in the MSLT-1 trial (10-year niquet. The extremity is heated to 41 °C prior to infusion of
melanoma-­specific survival: 62.1% ± 4.8% vs. 41.5% ± 5.6% melphalan, and a bypass circuit is used to maintain oxygen-
in observation group). In addition, complication rates for ation. Alternatively, ILI utilizes percutaneous catheterization
SLNB are much lower than what is seen for lymph node dis- and infusion of melphalan plus actinomycin D at normother-
sections [17]. For example, in the multicenter selective mic to slightly hyperthermic temperatures under progres-
lymphadenectomy trial 2 (MSLT-2), 24.1% of patients in sively hypoxic conditions. HILP has superior complete
the CLND group experienced lymphedema compared to response rates that must be balanced with higher rates of
only 6.3% of patients who had SLNB only [18]. In addition limb-threatening toxicity [20]. Proponents of ILI favor its
to biologic factors of Breslow depth and ulceration, the use as it is less invasive and can be performed multiple times.
decision to perform SLNB should also be based on overall While no studies have demonstrated superior survival with
patient status and how results can or will impact further HILP versus ILI (or prospectively in a randomized fashion
management. directly comparing these two forms of regional therapy),
based on the results of Chai and colleagues’ multi-­
institutional findings, it has been proposed that ILI be uti-
36.3 Management of In-Transit Disease lized upfront, except in patients with bulky disease, while
saving HILP as a salvage strategy for those with high-­volume
In-transit melanoma (ITM) is a distinct clinical entity disease, progression on ILI, or rapid recurrence [20].
whereby melanoma tumor cells recur as dermal or subcuta- Alternatively, intralesional therapies have been an active
neous nodules arising between the primary site and the area of investigation for the management of ITM [19]. These
draining regional lymph node basin. This definition encom- therapies are hypothesized to incite a local inflammatory
passes a wide array of presentations, with ITM lesions rang- response that induces antitumor immunity in the local envi-
ing in size from 0.2 to 3 cm and in quantity from single to ronment and perhaps systemically. The accessibility of
hundreds of lesions [19]. Despite optimal primary resection superficial lesions makes intralesional therapy a feasible and
of melanoma, ITM arises in 4–12% of patients within a attractive strategy. Further, local therapies have less systemic
388 K. E. Rhodin et al.

adverse effects and lower toxicity risk to patients. Several sidering treatment for patients with regional disease. The
agents that have been trialed in this setting include, but are diagnosis of microscopic nodal disease (N1a, N2a, N3a in
not limited to, interleukin-2 (IL-2), interferon-alpha, Bacillus the eighth edition of the staging system) is made through
Calmette–Guerin (BCG), and PV-10 (Rose Bengal) [21]. SLNB, while macroscopic and in transit/satellite disease is
Many of these therapies demonstrated promising responses diagnosed clinically or radiographically. Management of
in injected lesions, and for some agents such as PV-10, patients with macroscopic nodal disease has largely remained
responses were also seen in noninjected bystander lesions; unchanged, and these patients should undergo TLND,
however, in trials, significant survival benefits have not been although the extent of lymphadenectomy is evolving and
consistently shown. neoadjuvant approaches are being evaluated.
Oncolytic viruses have become a promising intralesional Upon diagnosis of microscopic nodal disease by SLNB
therapy for ITM. Notably, Talimogene Laherparepvec or of macroscopic disease clinically or radiographically,
(T-VEC), a genetically modified herpes simplex virus type 1, management includes some combination of surgery, sys-
has demonstrated efficacy in patients with advanced mela- temic therapy (ICIs and BRAF/MEK inhibitors), and poten-
noma and was FDA approved in 2015. T-VEC, like other tially radiation therapy in select cases, depending on the
oncolytic viruses, has a direct effect by lysing tumor cells extent of disease in the nodal basin. The intent of adjuvant
[22]. T-VEC also acts by producing granulocyte-macrophage therapy in stage III melanoma is to “reduce the recurrence
colony-stimulating factor (GM-CSF), which recruits immune rate of radically operated patients at high risk… and to
cells to the tumor microenvironment. While T-VEC has a potentially improve survival.” [27] A thorough understand-
durable response rate of 16.3% (objective response lasting at ing of the mechanisms underlying the development of nodal
least 6 months), it is important to note that it has not demon- metastases is critical to determining which therapy or com-
strated a survival benefit when utilized as a single therapy bination of therapies may be most beneficial to patients with
[23]. Further exploration into other oncolytic viruses and nodal disease.
intratumoral agents, including ONCOS-102, toll-like recep-
tors, and many others, is ongoing [21].
Although patients with isolated ITM are candidates for 36.5  urgical Management of the Nodal
S
the aforementioned regional and intralesional therapies, the Basin
majority of patients with ITM will develop nodal or distant
metastasis. Furthermore, even for patients with isolated ITM Prior to the development of SLNB, elective lymph node dis-
in the absence of distant metastases, no regional therapy has section (ELND) was routinely performed to evaluate for
demonstrated improvements in overall survival, although nodal metastases and theoretically prevent the spread of mel-
select patients who respond to therapy may achieve a sur- anoma beyond the lymph nodes. However, due to the rela-
vival benefit. Conversely, systemic therapy with ICI or tively low prevalence of microscopic nodal metastasis in
BRAF/MEK inhibitor therapy has been associated with patients with primary cutaneous melanoma, which is approx-
durable response rates and improved survival in patients with imately 20% [28], routine ELND subjected a majority of
unresectable stage III melanoma, which includes patients patients with no metastatic disease to the complications of
with ITM [24]. As such, in the current era, in patients pre- ELND without deriving any measurable benefit. With the
senting with unresectable ITM, systemic therapy should be advent of lymphoscintigraphy and SLNB, the prognostic
strongly considered as the first-line treatment. Regional and benefit of lymph node dissection could be achieved without
intralesional therapies can be used for patients who have a subjecting the patient to the morbidities associated with
low volume of ITM only, have contraindications to systemic ELND. However, the question still remained: in a patient
therapy, or have failed systemic therapies. One exciting with a positive SLN, should we return to the operating room
approach is to combine intralesional or regional chemother- to perform a CLND?
apy with systemic therapy, and multiple trials are ongoing Following the validation of SLNB, CLND was nearly uni-
exploring these combination therapies [25, 26]. formly adopted as the standard of care for patients with a
positive SLN. However, several conceptual questions under-
lie this issue. Does the result of SLNB correlate with the
36.4  valuation of the Nodal Basin
E likelihood of non-sentinel lymph node (NSLN) disease? Do
and Management of Node-Positive SLNs become positive before NSLNs, and if so, does
Disease removal of the SLN protect against disease progression in
the nodal basin and beyond? We know that only 10–30% of
The eighth edition of the AJCC guidelines attempted to take patients with a positive SLN will have positive NSLNs [29].
into account the recent advancements in both the understand- Various systems (e.g., Rotterdam–Dewar) have attempted to
ing and the management of cutaneous melanoma when con- predict NSLN status based on features (size, location) of the
36 The Role of Surgery in Managing Primary and Metastatic Melanoma 389

SLN metastasis [30]. A 2008 study of stage I and II melano- In general, these represent patient populations that were
mas found that positive SLNs with micrometastases (metas- underrepresented in the clinical trials and are thought to be at
tases less than 2 mm in diameter) only were associated with higher risk for harboring additional regional nodal disease.
negative NSLNs [31]. However, we have also learned that The decision to perform CLND in these populations should
patients who do have positive NSLNs have a worse progno- be determined on an individual basis and take into consider-
sis than patients with only positive SLNs, regardless of the ation the risks of the specific operation (e.g., higher risk of
number of overall positive nodes [32]. This suggests that lymphedema with inguinal vs. axillary or cervical
positive NLSNs may reflect a different tumor biology and lymphadenectomies).
carry prognostic significance beyond just increasing the For those patients with clinically positive nodes, TLND
nodal metastatic burden. For these reasons, some proposed remains the standard of care. With clinically positive
that the prognostic value of the NSLNs merits any invasive nodes, the rates of local and systemic recurrence are sig-
means necessary to find this prognostic information, further nificantly increased, and overall survival is significantly
cementing CLND as a necessary step in the management of lower [37]. These patients should be strongly considered
SLN-positive patients. Later studies investigated additional for adjuvant systemic therapy and possibly radiation ther-
noninvasive methods of determining the prognostic informa- apy. In determining adjuvant treatment strategies, the
tion of NSLNs; for example, the N-SNORE (Non sentinel patient-specific prognosis must be assessed. Factors that
node risk score) was a tool developed in 2010 to predict the portend a poorer prognosis, and therefore warrant discus-
likelihood of NSLN disease based on characteristics derived sions of adjuvant therapy, include stage IIIA disease with
from the SLNs [33]. Since then it has been demonstrated that SLN tumor deposit >1 mm and stage IIIB/IIIC/IIID dis-
even in patients with positive NSLNs, the melanoma meta- ease. Patients with stage IIIA melanoma with <1 mm
static burden within those NSLNs are significantly diluted as tumor deposit in the SLN may be offered surveillance
compared to their SLNs [34]. Overall, data suggest that without adjuvant therapy. Stage III melanomas should be
while a small proportion of patients may have some NSLN evaluated for mutational profile, specifically BRAF muta-
disease left behind after SLNB, the disease burden is likely tions status, to assess for eligibility of BRAF/MEK inhibi-
to be lower than in the SLNs. Furthermore, although use of tor targeted therapy.
the presence of positive NSLNs may aid in prognosis and
risk of recurrence, prognosis and risk can also be ascertained
based on known variables without subjecting the patient to 36.6 Adjuvant Systemic Immunotherapy
additional surgery.
The role of CLND as simply a prognostic tool was fore- Checkpoint inhibitors are a recent development within the
shadowed when several studies showed no improvement in past decade and have revolutionized the management of
melanoma-specific survival with CLND [18]. The second many cancers, including cutaneous melanoma. As a class,
Multicenter Selective Lymphadenectomy Trial (MSLT-II) these monoclonal antibodies are directed against proteins
and DeCOG-SLT provided definitive evidence confirming involved in downregulating the T-cell-dependent immune
that melanoma-specific survival is not significantly different activity of the patient. Programmed cell death protein 1 (PD-­
between patients with positive SLNs who underwent CLND 1) and cytotoxic T lymphocyte antigen 4 (CTLA-4) are T-cell
or nodal observation [35, 36]. Since that time, CLND is no membrane proteins that affect immune self-tolerance via
longer routinely recommended for patients with microscopic T-cell suppression; PD-L1 is a ligand for PD-1. Melanomas
nodal stage III disease. The preferred alternative for patients are able to evade these T-cell-dependent immune pathways
with SNLB-positive nodes per NCCN guidelines is clinical by constitutively expressing proteins that promote self-­
follow-up and nodal basin ultrasound surveillance, as well as tolerance, such as PD-L1 [38]. The anti-CTLA-4 drug ipili-
consideration for adjuvant therapy (notably all adjuvant ther- mumab was the first molecule to gain FDA approval for use
apy trials were completed mandating CLND). in metastatic melanoma; FDA approval for adjuvant use in
While CLND has been abandoned for most patients with stage III disease was granted in 2015 [39] based on a phase 3
positive SLNs, some indications may remain for CLND. The randomized controlled trial that demonstrated improved sur-
main indication for CLND is the inability of the patient to vival with ipilimumab versus placebo in patients with
undergo routine surveillance. Additional indications are resected stage III disease [40]. Since that time, the anti-PD-1
more controversial, and include head and neck location, drugs nivolumab and pembrolizumab have also been
presence of more than 1 mm of disease in an SLN node and approved for use in the adjuvant setting [7]. Direct compari-
having more than one positive SLN. For example, the authors son between nivolumab and ipilimumab demonstrated
of MSLT-II declined to recommend against CLND in patients improved recurrence-free survival and fewer adverse events
with positive SLNs with tumor deposits greater than 1 mm. with nivolumab [41]; consequently, in a 2020 clinical prac-
390 K. E. Rhodin et al.

tice guideline, the American Society of Clinical Oncology 36.8  he Role of Neoadjuvant Therapy
T
recommended that adjuvant nivolumab or pembrolizumab in Managing Advanced Stage
should be offered to patients with stage III melanoma, and Melanoma
cautioned against routine use of ipilimumab in the adjuvant
setting. The success of neoadjuvant approaches in other solid tumor
malignancies underscores its potential application in mela-
noma patients. Pathologic response is an early readout of
36.7  djuvant Systemic Targeted Therapy:
A patient outcomes, correlating with subsequent recurrence-­
BRAF, MEK Inhibitors free survival and is a marker of patient-specific therapy
response, potentially informing the choice of future thera-
Melanoma is also able to evade the immune response by pies. Furthermore, neoadjuvant approaches enable a deeper
manipulating the mitogen-activated protein kinase (MAPK) understanding of mechanisms of resistance and response
signaling pathway. Activating mutations in the downstream through the identification of biomarkers. There is also evi-
molecules RAF and MAPK-kinase (MEK) are classic cul- dence that T-cell checkpoint blockade given as neoadjuvant
prits in this deregulation, allowing for the loss of cell cycle therapy induces a stronger and broader tumor-specific T-cell
regulation and promotion of cell cycle progression. response than adjuvant therapy [43]. Neoadjuvant trials in
Consequently, agents targeting these mutated proteins have melanoma predate the introduction of ICIs and targeted ther-
been developed and are now approved adjuvant medications apy; however, for the purposes of this section, we will focus
for treating melanoma. Vemurafenib was the first mutant on the recent targeted therapy and immunotherapy trials
BRAF inhibitor to be approved for use in melanoma. In a (Table 36.1).
randomized controlled trial in 2013, vemurafenib was dem-
onstrated to be superior to dacarbazine (a chemotherapeutic
agent) in unresectable stage III and stage IV melanoma with 36.8.1 Targeted Therapy
BRAF V600E mutations [5]. Since that time, newer BRAF
inhibitors such as dabrafenib have been utilized in combina- Amaria et al. examined patients with bulky stage III or
tion with MEK inhibitors such as trametinib to overcome resectable stage IV BRAF V600E/K mutated melanoma and
melanoma’s resistance to single-agent therapy [42]. At pres- randomized them in a 1:2 ratio to either standard of care
ent, specifically for patients whose melanoma carries the (SOC) surgery and adjuvant therapy versus neoadjuvant dab-
V600E/K BRAF mutation, a combination of adjuvant dab- rafenib plus trametinib therapy for 8 weeks followed by sur-
rafenib and trametinib may be offered as an alternative to gery and up to 44 weeks of adjuvant dabrafenib plus
adjuvant ICI in patients with BRAF-mutated melanoma trametinib therapy [44]. Fifty-eight percent of the neoadju-
based on results of the COMBI-AD trial [42]. vant group who underwent surgery achieved a pathologic

Table 36.1 Neoadjuvant targeted therapy and immunotherapy trials for locally advanced melanoma
Study Population Regimen N pCR (%) Median RFS (mo) Median FU (mo)
NCT02231775 Stage II, IV Dab/Tram x 8w-surgery- 21 58 19.7 18.6
Amaria (2018) [18] BRAF V600E/K Dab/Tram x 44w
NCT01972347 Stage III Dab/Tram x 12w-surgery 35 49 23.0 27.0
Long (2019) [19] BRAF V600E/K Dab/Tram x 40w
NCT02437279 Stage III I3N1 x 2-surgery-I3N1 x 2 10 33 NR 32
Blank (2018) [20]
NCT02519322 Stage III, IV (resectable) A: Nivo x 4-surgery-Nivo x 13 A: 12 25 NR 20
Amaria (2019) [22] B: I3N1 x 3-surgery-Nivo x 13 B: 11 45 NR
NCT02519322 Stage III, IV (resectable) Pembro x 1- surgery-Pembro x 17 30 19 NR 18
Huang (2019) [23]
NCT02977052 Stage III A: I3N1 x 2 surgery A: 30 B: 57 NR 8.3
Rozeman (2019) [24] B: I3N1 x 2 surgery B: 30
C: Ipi x 2-Nivo x 2- surgery C: 26
NR not reported, Dab dabrafenib, Tram trametinib, I ipilimumab, N nivolumab, Nivo nivolumab, Pembro pembrolizumab, pCR pathologic com-
plete response, RFS recurrence-free survival, FU follow up
36 The Role of Surgery in Managing Primary and Metastatic Melanoma 391

complete response (pCR). Patients in the neoadjuvant group arms (ipilimumab 3 mg/kg and nivolumab 1 mg/kg, ipilim-
had a 60-fold reduced risk of relapse after surgery in com- umab 1 mg/kg and nivolumab 3 mg/kg, and sequential
parison with patients who underwent SOC (median event-­ ­ipilimumab 3 mg/kg followed by nivolumab 3 mg/kg). Based
free survival was 19.7 months vs. 2.9 months, respectively, on response and toxicities, ipilimumab 1 mg/kg and
hazard ratio 0.016, p < 0.0001), and the trial was stopped nivolumab 3 mg/kg (called “flip dose”) were determined to
early because it reached the primary objective of improved be the best regimen to study in future trials. The MeMaloc
event-free survival with neoadjuvant treatment [44]. Long substudy of OpACIN-neo investigated the feasibility of tar-
et al. examined patients with bulky stage IIIB/C BRAF geted removal of a single pathological (index) node and
V600E/K mutated melanoma in a single-arm phase II trial whether this node was a reliable indicator of pathologic
where they administered dabrafenib and trametinib for response in the entire lymph node basin [49]. The index node
12 weeks prior to surgery followed by 40 weeks of dab- was marked with a magnetic seed, removed surgically, and
rafenib and trametinib [45]. At the time of surgical evalua- analyzed separately from the rest of the lymph node basin
tion, 17 patients achieved a pCR (49%). No progression of after two cycles of therapy. Results showed that the index
disease was observed during the neoadjuvant therapy period. node was congruent with the entire lymph node basin in 12
At a median follow-up period of 12.1 months (95% CI 8.8– out of 12 nodes, which prompted further investigation.
14.8 months), recurrence of disease was observed in 12 Personalized Response-driven Adjuvant therapy after
patients (36%), with 4 patients having recurrence while Combination of Ipilimumab and Nivolumab (PRADO) is an
undergoing adjuvant targeted therapy [45]. In summary, both extension cohort of the OpACIN-neo study which is actively
these trials demonstrated that patients with a pCR had sig- examining three objectives: (1) the pathologic response rate
nificantly longer distant metastasis-free survival. and safety of ipilimumab 1 mg/kg + nivolumab 3 mg/kg
combination scheme, (2) whether patients with a pCR or
near-pCR in the index lymph node can be safely spared a
36.9 Immunotherapy Trials lymphadenectomy without affecting recurrence-free sur-
vival, and (3) whether recurrence-­free survival can be pro-
Blank et al. examined adjuvant versus neoadjuvant combina- longed among pathological nonresponders with additional
tion therapy (ipilimumab 3 mg/kg + nivolumab 1 mg/kg) in adjuvant therapy. Patients receive two cycles of neoadjuvant
patients presenting with clinical stage III melanoma ipilimumab 1 mg/kg and nivolumab 3 mg/kg after marker
(OpACIN phase 1b trial) [43]. After a median follow-up placement in the index lymph node. Patients that achieve a
32 months, none of the seven patients that achieved a patho- pathologic response in the index lymph node do not undergo
logic response in the neoadjuvant arm relapsed. At 30 months, lymphadenectomy; patients with partial response (10–50%
estimated recurrence-free survival rates were 80% for the viable tumor cells) undergo lymphadenectomy; and patients
neoadjuvant arm and 60% for the adjuvant arm, while overall with no pathologic response undergo lymphadenectomy and
survival rates were 90% and 67%, respectively [43]. Amaria receive adjuvant nivolumab or targeted therapy for 52 weeks
et al. examined a combination immunotherapy approach in a +/− radiotherapy. Preliminary data confirm the high patho-
phase II randomized trial of neoadjuvant nivolumab 3 mg/kg logic response rate and safety observed previously in the
versus neoadjuvant ipilimumab 3 mg/kg and nivolumab OpACIN-neo arm B (ipilimumab 1 mg/kg + nivolumab
1 mg/kg in clinical stage III and oligometastatic stage IV 3 mg/kg), with a pathologic response rate of 71% and grade
melanoma patients [46]. Improved relapse-free survival was 3–4 adverse event rate of 22% in the first 12 weeks.
shown in patients achieving a pCR, but the trial was stopped Lymphadenectomy was omitted in 59 (60%) patients [46].
early due to high rates of disease progression in the mono- Longer follow-up is needed to report safety and recurrence-
therapy arm and high rates of toxicities in the combination free survival when lymphadenectomy is omitted in patho-
arm. Huang et al. examined monotherapy neoadjuvant pem- logic responders.
brolizumab in clinical stage III and oligometastatic stage IV The dosing and duration of therapy have been heteroge-
melanoma patients [47]. After a single dose of neoadjuvant neous in the above-highlighted trials. The International
pembrolizumab, 5 of 27 patients achieved a pCR, and all Neoadjuvant Melanoma Consortium (INMC) recommends
patients with a pCR or major pathologic response remained that neoadjuvant protocols are 6–8 weeks in the preoperative
relapse free. This study showed that pathological response phase and that trials should complete an adjuvant phase to
and immunological analyses after a single neoadjuvant dose total a year of systemic therapy [50]. The degree of patho-
can predict clinical outcome and unveil underlying mecha- logical response may direct the inclusion of an adjuvant
nisms in checkpoint blockade. phase, as is occurring in the PRADO study, but this approach
The OpACIN-neo study sought to identify the best dosing is still very much under review. Data on surgical resectabil-
and scheduling of neoadjuvant immunotherapy in clinical ity, complications, and delays of surgical intervention have
stage III melanoma [48]. Patients were randomized to three not been robustly collected in the above-mentioned trials, but
392 K. E. Rhodin et al.

some preliminary data demonstrate that rates of surgical Disclosures GMB (Beasley) served on a scientific advisory board for
complications in the neoadjuvant setting are no different Regeneron (2019) and has currently clinical trial funding paid to insti-
tution from Oncosec, Istari oncology, Replimune, and Delcath. GMB
than upfront surgery. The long-term results of the PRADO (Boland) served on a scientific advisory board and currently serves on a
study will shed key insights on our approach to a more steering committee for Nektar Therapeutics. She served on a scientific
focused versus more extensive lymph node dissection and on advisory board and as a speaker for Novartis. She has sponsored
the use of radiographic markers to assess and monitor dis- research agreements with Olink proteomics and Palleon
Pharmaceuticals. MCL serves on an advisory board for BMS and
ease burden as outlined in the PRADO cohort. These results receives research support from Vaccinex, BMS, Delcath, and Amgen.
may also further inform treatment strategies for satellite, in-­
transit, or oligometastatic disease within the neoadjuvant
trial setting. The INMC Surgery Working Group is preparing
a specific surgical assessment tool that will standardize sur- References
gical outcomes collected and implemented in future trials.
1. Whiteman DC, Green AC, Olsen CM. The growing burden of
invasive melanoma: projections of incidence rates and numbers
of new cases in six susceptible populations through 2031. J Invest
36.10 Conclusion Dermatol. 2016;136:1161–71.
2. Gardner LJ, Ward M, Andtbacka RHI, Boucher KM, Bowen GM,
Multiple new systemic therapies have been approved for Bowles TL, Cohen AL, Grossmann K, Hitchcock YJ, Holmen SL,
Hyngstrom J, Khong H, McMahon M, Monroe MM, Ross CB,
patients with melanoma based on results showing efficacy. Suneja G, Wada D, Grossman D. Risk factors for development
Along with these new agents, landmark surgical trials have of melanoma brain metastasis and disease progression: a single-­
been published demonstrating no survival benefit for routine center retrospective analysis. Melanoma Res. 2017;27:477–84.
CLND for a positive SLNB. Therefore, the current role for PMC5583702.
3. Leiter U, Meier F, Schittek B, Garbe C. The natural course of cuta-
surgery in managing melanoma must take into account these neous melanoma. J Surg Oncol. 2004;86:172–8.
new results. However, surgical therapy for primary mela- 4. Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S,
noma is relatively unchanged, although efforts to define Teague J, Woffendin H, Garnett MJ, Bottomley W, Davis N,
appropriate margins are ongoing. SLNB is still utilized pri- Dicks E, Ewing R, Floyd Y, Gray K, Hall S, Hawes R, Hughes
J, Kosmidou V, Menzies A, Mould C, Parker A, Stevens C, Watt
marily for prognostication, and is now increasingly to direct S, Hooper S, Wilson R, Jayatilake H, Gusterson BA, Cooper C,
effective adjuvant treatment, but SLNB may harbor a possi- Shipley J, Hargrave D, Pritchard-Jones K, Maitland N, Chenevix-­
ble therapeutic benefit in some cases. Surgery will likely Trench G, Riggins GJ, Bigner DD, Palmieri G, Cossu A, Flanagan
continue to play a role in the management of select patients A, Nicholson A, Ho JW, Leung SY, Yuen ST, Weber BL, Seigler
HF, Darrow TL, Paterson H, Marais R, Marshall CJ, Wooster R,
with isolated metastatic disease in conjunction with systemic Stratton MR, Futreal PA. Mutations of the BRAF gene in human
therapy and will also allow for assessment of pathologic cancer. Nature. 2002;417:949–54.
response after neoadjuvant therapy. 5. Chapman PB, Hauschild A, Robert C, Haanen JB, Ascierto P, Larkin
Ongoing and future research will strive to answer: J, Dummer R, Garbe C, Testori A, Maio M, Hogg D, Lorigan P,
Lebbe C, Jouary T, Schadendorf D, Ribas A, O’Day SJ, Sosman JA,
Kirkwood JM, Eggermont AM, Dreno B, Nolop K, Li J, Nelson B,
Hou J, Lee RJ, Flaherty KT, McArthur GA. Improved survival with
vemurafenib in melanoma with BRAF V600E mutation. N Engl J
Med. 2011;364:2507–16. PMC3549296.
• What are standard margins for a patient with early-­
6. Ayers M, Lunceford J, Nebozhyn M, Murphy E, Loboda A,
stage localized melanoma? Kaufman DR, Albright A, Cheng JD, Kang SP, Shankaran V, Piha-­
• What is the appropriate surgery for a patient with Paul SA, Yearley J, Seiwert TY, Ribas A, McClanahan TK. IFN-­
clinically involved regional lymph nodes and no gamma-­related mRNA profile predicts clinical response to PD-1
blockade. J Clin Invest. 2017;127:2930–40. PMC5531419.
other distant disease?
7. Kwak M, Farrow NE, Salama AKS, Mosca PJ, Hanks BA, Slingluff
• Which patients should be offered a sentinel lymph CL, Beasley GM. Updates in adjuvant systemic therapy for mela-
node biopsy? noma. J Surg Oncol. 2019;119:222–31. PMC6330126.
• Is neoadjuvant therapy standard of care for patients 8. Hewitt BB, Bradley E, Berger AC. Margins of melanoma exci-
sion and modifications to standards. Surg Oncol Clin N Am.
with advanced melanoma?
2020;29:339–47.
• Which intralesional agents or regional therapies 9. Veronesi U, Cascinelli N, Adamus J, Balch C, Bandiera D, Barchuk
should be considered for patients with metastatic A, Bufalino R, Craig P, De Marsillac J, Durand JC, et al. Thin stage
dermal and in-transit melanoma and how can they I primary cutaneous malignant melanoma. Comparison of excision
with margins of 1 or 3 cm. N Engl J Med. 1988;318:1159–62.
be combined with systemic therapies?
10. Sladden MJ, Balch C, Barzilai DA, Berg D, Freiman A, Handiside
• When should surgery be considered for patients T, Hollis S, Lens MB, Thompson JF. Surgical excision margins
with distant metastases? for primary cutaneous melanoma. Cochrane Database Syst Rev.
2009:CD004835.
36 The Role of Surgery in Managing Primary and Metastatic Melanoma 393

11. Hayes AJ, Maynard L, Coombes G, Newton-Bishop J, Timmons of Talimogene Laherparepvec therapy: a multi-institutional experi-
M, Cook M, Theaker J, Bliss JM, Thomas JM. Wide versus nar- ence. J Am Coll Surg. 2019;228:644–9.
row excision margins for high-risk, primary cutaneous melano- 23. Andtbacka RH, Kaufman HL, Collichio F, Amatruda T, Senzer
mas: long-term follow-up of survival in a randomised trial. Lancet N, Chesney J, Delman KA, Spitler LE, Puzanov I, Agarwala
Oncol. 2016;17:184–92. PMC4737890. SS, Milhem M, Cranmer L, Curti B, Lewis K, Ross M, Guthrie
12. Moncrieff MD, Gyorki D, Saw R, Spillane AJ, Thompson JF, Peach T, Linette GP, Daniels GA, Harrington K, Middleton MR, Miller
H, Oudit D, Geh J, Dziewulski P, Wilson E, Matteucci P, Pritchard-­ WH Jr, Zager JS, Ye Y, Yao B, Li A, Doleman S, VanderWalde A,
Jones R, Olofsson Bagge R, Wright FC, Crampton N, Cassell O, Gansert J, Coffin RS. Talimogene Laherparepvec improves durable
Jallali N, Berger A, Kelly J, Hamilton S, Durrani A, Lo S, Paton response rate in patients with advanced melanoma. J Clin Oncol.
E, Henderson MA. 1 versus 2-cm excision margins for pT2-pT4 2015;33:2780–8.
primary cutaneous melanoma (MelMarT): a feasibility study. Ann 24. Robert C, Long GV, Brady B, Dutriaux C, Giacomo AMD, Mortier
Surg Oncol. 2018;25:2541–9. PMC6097739. L, Rutkowski P, Hassel JC, McNeil CM, Kalinka EA, Lebbé C,
13. Maurichi A, Miceli R, Camerini T, Contiero P, Patuzzo R, Tragni G, Charles J, Hernberg MM, Savage KJ, Chiarion-Sileni V, Mihalcioiu
Crippa F, Romanidis K, Ruggeri R, Carbone A, Santinami M. Pure C, Mauch C, Arance A, Cognetti F, Ny L, Schmidt H, Schadendorf
desmoplastic melanoma: a melanoma with distinctive clinical D, Gogas H, Zoco J, Re S, Ascierto PA, Atkinson V. Five-year out-
behavior. Ann Surg. 2010;252:1052–7. comes with nivolumab in patients with wild-type braf advanced
14. Morton DL, Thompson JF, Cochran AJ, Mozzillo N, Nieweg OE, melanoma. J Clin Oncol. 2020;38(33):3937–46.
Roses DF, Hoekstra HJ, Karakousis CP, Puleo CA, Coventry BJ, 25. Ribas A, Dummer R, Puzanov I, VanderWalde A, Andtbacka RHI,
Kashani-Sabet M, Smithers BM, Paul E, Kraybill WG, McKinnon Michielin O, Olszanski AJ, Malvehy J, Cebon J, Fernandez E,
JG, Wang HJ, Elashoff R, Faries MB, Group M. Final trial report Kirkwood JM, Gajewski TF, Chen L, Gorski KS, Anderson AA,
of sentinel-node biopsy versus nodal observation in melanoma. N Diede SJ, Lassman ME, Gansert J, Hodi FS, Long GV. Oncolytic
Engl J Med. 2014;370:599–609. PMC4058881. virotherapy promotes intratumoral T cell infiltration and improves
15. Wong SL, Faries MB, Kennedy EB, Agarwala SS, Akhurst TJ, anti-PD-1 immunotherapy. Cell. 2017;170:1109–1119.e10.
Ariyan C, Balch CM, Berman BS, Cochran A, Delman KA, Gorman 26. Ariyan CE, Lefkowitz RA, Panageas K, Callahan MK, Misholy
M, Kirkwood JM, Moncrieff MD, Zager JS, Lyman GH. Sentinel O, Bello D, Fisher C, Kotin A, Siegelbaum R, Coit DG, Wolchok
lymph node biopsy and management of regional lymph nodes in JD, Brady MS. Safety and clinical activity of combining systemic
melanoma: American Society of Clinical Oncology and Society ipilimumab with isolated limb infusion in patients with in-transit
of Surgical Oncology clinical practice guideline update. Ann Surg melanoma. J Clin Oncol. 2014;32:9078.
Oncol. 2018;25:356–77. 27. Testori AAE, Ribero S, Indini A, Mandalà M. Adjuvant treatment
16. Lo SN, Ma J, Scolyer RA, Haydu LE, Stretch JR, Saw RPM, of melanoma: recent developments and future perspectives. Am J
Nieweg OE, Shannon KF, Spillane AJ, Ch’ng S, Mann GJ, Clin Dermatol. 2019;20:817–27.
Gershenwald JE, Thompson JF, Varey AHR. Improved risk predic- 28. Morton DL, Wen DR, Wong JH, Economou JS, Cagle LA,
tion calculator for sentinel node positivity in patients with mela- Storm FK, Foshag LJ, Cochran AJ. Technical details of intraop-
noma: the Melanoma Institute Australia Nomogram. J Clin Oncol. erative lymphatic mapping for early stage melanoma. Arch Surg.
2020;38(24):2719–27. 1992;127:392–9.
17. Morton DL, Thompson JF, Cochran AJ, Mozzillo N, Elashoff 29. McMasters KM, Wong SL, Edwards MJ, Chao C, Ross MI,
R, Essner R, Nieweg OE, Roses DF, Hoekstra HJ, Karakousis Noyes RD, Viar V, Cerrito PB, Reintgen DS. Frequency of non-
CP, Reintgen DS, Coventry BJ, Glass EC, Wang HJ. Sentinel-­ sentinel lymph node metastasis in melanoma. Ann Surg Oncol.
node biopsy or nodal observation in melanoma. N Engl J Med. 2002;9:137–41.
2006;355:1307–17. 30. Dewar DJ, Newell B, Green MA, Topping AP, Powell BW, Cook
18. Faries MB, Thompson JF, Cochran AJ, Andtbacka RH, Mozzillo N, MG. The microanatomic location of metastatic melanoma in sen-
Zager JS, Jahkola T, Bowles TL, Testori A, Beitsch PD, Hoekstra tinel lymph nodes predicts nonsentinel lymph node involvement. J
HJ, Moncrieff M, Ingvar C, Wouters M, Sabel MS, Levine EA, Clin Oncol. 2004;22:3345–9.
Agnese D, Henderson M, Dummer R, Rossi CR, Neves RI, Trocha 31. Glumac N, Hocevar M, Zadnik V, Snoj M. Sentinel lymph node
SD, Wright F, Byrd DR, Matter M, Hsueh E, MacKenzie-Ross micrometastasis may predict non-sentinel involvement in cutane-
A, Johnson DB, Terheyden P, Berger AC, Huston TL, Wayne JD, ous melanoma patients. J Surg Oncol. 2008;98:46–8.
Smithers BM, Neuman HB, Schneebaum S, Gershenwald JE, 32. Rossi CR, Mocellin S, Campana LG, Borgognoni L, Sestini
Ariyan CE, Desai DC, Jacobs L, McMasters KM, Gesierich A, S, Giudice G, Caraco C, Cordova A, Solari N, Piazzalunga D,
Hersey P, Bines SD, Kane JM, Barth RJ, McKinnon G, Farma JM, Carcoforo P, Quaglino P, Caliendo V, Ribero S. Prediction of non-­
Schultz E, Vidal-Sicart S, Hoefer RA, Lewis JM, Scheri R, Kelley sentinel node status in patients with melanoma and positive senti-
MC, Nieweg OE, Noyes RD, DSB H, Wang HJ, Elashoff DA, nel node biopsy: an Italian Melanoma Intergroup (IMI) Study. Ann
Elashoff RM. Completion dissection or observation for sentinel-­ Surg Oncol. 2018;25(1):271–9.
node metastasis in melanoma. N Engl J Med. 2017;376:2211–22. 33. Murali R, Desilva C, Thompson JF, Scolyer RA. Non-sentinel node
PMC5548388. risk score (N-SNORE): a scoring system for accurately stratify-
19. Perone JA, Farrow N, Tyler DS, Beasley GM. Contemporary ing risk of non-sentinel node positivity in patients with cutane-
approaches to in-transit melanoma. J Oncol Pract. 2018;14:292–300. ous melanoma with positive sentinel lymph nodes. J Clin Oncol.
20. Chai CY, Deneve JL, Beasley GM, Marzban SS, Chen YA, Rawal 2010;28:4441–9.
B, Grobmyer SR, Hochwald SN, Tyler DS, Zager JS. A multi-­ 34. Ulmer A, Dietz K, Werner-Klein M, Häfner H-M, Schulz C, Renner
institutional experience of repeat regional chemotherapy for recur- P, Weber F, Breuninger H, Röcken M, Garbe C, Fierlbeck G, Klein
rent melanoma of extremities. Ann Surg Oncol. 2012;19:1637–43. CA. The sentinel lymph node spread determines quantitatively
PMC4515968. melanoma seeding to non-sentinel lymph nodes and survival. Eur J
21. Farrow NE, Leddy M, Landa K, Beasley GM. Injectable therapies Cancer. 2018;91:1–10.
for regional melanoma. Surg Oncol Clin N Am. 2020;29:433–44. 35. Faries MB, Cochran AJ, Elashoff RM, Thompson JF. Multicenter
PMC7274146. selective lymphadenectomy trial-I confirms the central role of senti-
22. Louie RJ, Perez MC, Jajja MR, Sun J, Collichio F, Delman KA, nel node biopsy in contemporary melanoma management: response
Lowe M, Sarnaik AA, Zager JS, Ollila DW. Real-world outcomes to ‘No survival benefit for patients with melanoma undergoing
394 K. E. Rhodin et al.

sentinel lymph node biopsy: critical appraisal of the multicenter 44. Amaria RN, Prieto PA, Tetzlaff MT, Reuben A, Andrews MC,
selective lymphadenectomy trial-I final report’. Br J Dermatol. Ross MI, Glitza IC, Cormier J, Hwu WJ, Tawbi HA, Patel SP, Lee
2015;172:571–3. PMC4437591. JE, Gershenwald JE, Spencer CN, Gopalakrishnan V, Bassett R,
36. Leiter U, Stadler R, Mauch C, Hohenberger W, Brockmeyer N, Simpson L, Mouton R, Hudgens CW, Zhao L, Zhu H, Cooper ZA,
Berking C, Sunderkötter C, Kaatz M, Schulte K-W, Lehmann P, Wani K, Lazar A, Hwu P, Diab A, Wong MK, McQuade JL, Royal
Vogt T, Ulrich J, Herbst R, Gehring W, Simon J-C, Keim U, Martus R, Lucci A, Burton EM, Reddy S, Sharma P, Allison J, Futreal PA,
P, Garbe C. Complete lymph node dissection versus no dissection Woodman SE, Davies MA, Wargo JA. Neoadjuvant plus adjuvant
in patients with sentinel lymph node biopsy positive melanoma dabrafenib and trametinib versus standard of care in patients with
(DeCOG-SLT): a multicentre, randomised, phase 3 trial. Lancet high-risk, surgically resectable melanoma: a single-centre, open-­
Oncol. 2016;17:757–67. label, randomised, phase 2 trial. Lancet Oncol. 2018;19:181–93.
37. Nessim C, Law C, McConnell Y, Shachar S, McKinnon G, Wright 45. Long GV, Saw RPM, Lo S, Nieweg OE, Shannon KF, Gonzalez M,
F. How often do level III nodes bear melanoma metastases and does Guminski A, Lee JH, Lee H, Ferguson PM, Rawson RV, Wilmott JS,
it affect patient outcomes? Ann Surg Oncol. 2013;20:2056–64. Thompson JF, Kefford RF, Ch'ng S, Stretch JR, Emmett L, Kapoor
38. Byrne EH, Fisher DE. Correlates of metastatic melanoma response R, Rizos H, Spillane AJ, Scolyer RA, Menzies AM. Neoadjuvant
to checkpoint inhibition. Cancer. 2018;123(Suppl 11):2143–53. dabrafenib combined with trametinib for resectable, stage IIIB-C,
39. Weber J, Mandala M, Vecchio MD, Gogas HJ, Arance AM, BRAF(V600) mutation-positive melanoma (NeoCombi): a
Cowey CL, Dalle S, Schenker M, Chiarion-Sileni V, Marquez-­ single-­arm, open-label, single-centre, phase 2 trial. Lancet Oncol.
Rodas I, Grob J-J, Butler MO, et al. ftCC: adjuvant Nivolumab 2019;20:961–71.
versus Ipilimumab in resected stage III or IV melanoma. NEJM. 46. Amaria RN, Reddy SM, Tawbi HA, Davies MA, Ross MI, Glitza
2017;377:1824–35. IC, Cormier JN, Lewis C, Hwu WJ, Hanna E, Diab A, Wong MK,
40. Eggermont AMM, Chiarion-Sileni V, Grob JJ, Dummer R, Wolchok Royal R, Gross N, Weber R, Lai SY, Ehlers R, Blando J, Milton
JD, Schmidt H, Hamid O, Robert C, Ascierto PA, Richards JM, DR, Woodman S, Kageyama R, Wells DK, Hwu P, Patel SP, Lucci
Lebbe C, Ferraresi V, Smylie M, Weber JS, Maio M, Hosein F, de A, Hessel A, Lee JE, Gershenwald J, Simpson L, Burton EM,
Pril V, Kicinski M, Suciu S, Testori A. Adjuvant ipilimumab versus Posada L, Haydu L, Wang L, Zhang S, Lazar AJ, Hudgens CW,
placebo after complete resection of stage III melanoma: long-term Gopalakrishnan V, Reuben A, Andrews MC, Spencer CN, Prieto V,
follow-up results of the European Organisation for Research and Sharma P, Allison J, Tetzlaff MT, Wargo JA. Neoadjuvant immune
Treatment of Cancer 18071 double-blind phase 3 randomised trial. checkpoint blockade in high-risk resectable melanoma. Nat Med.
Eur J Cancer. 2019;119:1–10. 2018;24:1649–54. PMC6481682.
41. Weber J, Mandala M, Del Vecchio M, Gogas HJ, Arance AM, Cowey 47. Huang AC, Orlowski RJ, Xu X, Mick R, George SM, Yan PK,
CL, Dalle S, Schenker M, Chiarion-Sileni V, Marquez-Rodas I, Manne S, Kraya AA, Wubbenhorst B, Dorfman L, D’Andrea K,
Grob JJ, Butler MO, Middleton MR, Maio M, Atkinson V, Queirolo Wenz BM, Liu S, Chilukuri L, Kozlov A, Carberry M, Giles L,
P, Gonzalez R, Kudchadkar RR, Smylie M, Meyer N, Mortier L, Kier MW, Quagliarello F, McGettigan S, Kreider K, Annamalai
Atkins MB, Long GV, Bhatia S, Lebbe C, Rutkowski P, Yokota K, L, Zhao Q, Mogg R, Xu W, Blumenschein WM, Yearley JH,
Yamazaki N, Kim TM, de Pril V, Sabater J, Qureshi A, Larkin J, Linette GP, Amaravadi RK, Schuchter LM, Herati RS, Bengsch B,
Ascierto PA. Adjuvant Nivolumab versus Ipilimumab in resected Nathanson KL, Farwell MD, Karakousis GC, Wherry EJ, Mitchell
stage III or IV melanoma. N Engl J Med. 2017;377(19):1824–35. TC. A single dose of neoadjuvant PD-1 blockade predicts clini-
42. Long GV, Hauschild A, Santinami M, Atkinson V, Mandala cal outcomes in resectable melanoma. Nat Med. 2019;25:454–61.
M, Chiarion-Sileni V, Larkin J, Nyakas M, Dutriaux C, Haydon PMC6699626.
A, Robert C, Mortier L, Schachter J, Schadendorf D, Lesimple 48. Rozeman EA, Blank CU, Akkooi ACJV, Kvistborg P, Fanchi L,
T, Plummer R, Ji R, Zhang P, Mookerjee B, Legos J, Kefford Thienen JVV, Stegenga B, Lamon B, Haanen JBAG, Schumacher
R, Dummer R, Kirkwood JM. Adjuvant Dabrafenib plus T. Neoadjuvant ipilimumab + nivolumab (IPI+NIVO) in palpable
Trametinib in stage III BRAF-mutated melanoma. N Engl J Med. stage III melanoma: updated data from the OpACIN trial and first
2017;377:1813–23. immunological analyses. J Clin Oncol. 2017;35:9586.
43. Blank CU, Rozeman EA, Fanchi LF, Sikorska K, van de Wiel B, 49. Schermers B, Franke V, Rozeman EA, van de Wiel BA, Bruining A,
Kvistborg P, Krijgsman O, van den Braber M, Philips D, Broeks A, Wouters MW, van Houdt WJ, Ten Haken B, Muller SH, Bierman
van Thienen JV, Mallo HA, Adriaansz S, Ter Meulen S, Pronk LM, C, Ruers TJM, Blank CU, van Akkooi ACJ. Surgical removal of
Grijpink-Ongering LG, Bruining A, Gittelman RM, Warren S, van the index node marked using magnetic seed localization to assess
Tinteren H, Peeper DS, Haanen J, van Akkooi ACJ, Schumacher response to neoadjuvant immunotherapy in patients with stage III
TN. Neoadjuvant versus adjuvant ipilimumab plus nivolumab in melanoma. Br J Surg. 2019;106:519–22. PMC6593699.
macroscopic stage III melanoma. Nat Med. 2018;24:1655–61. 50. Amaria R, Caffrey M, et al. Am J Manag Care. 2019;25:sp43–4.
 The Role of Surgery in Managing
Primary and Metastatic Breast Cancer 37
Alicia M. Terando, Azadeh Carr, Tina J. Hieken,
Mara A. Piltin, Bindupriya Chandrasekaran,
and Carla S. Fisher

Abstract
Learning Objectives
Modern breast cancer surgery began with Dr. William • Understand the biological subtypes of breast
Halsted’s hypothesis that breast cancer spreads in a step- cancer.
wise fashion from local to regional to distant sites and his • Understand the differences between breast-­
description of the radical mastectomy in 1894. With the conserving surgery and mastectomy in the surgical
introduction of antineoplastic chemotherapy and radia- management of early-stage breast cancer.
tion therapy, and improved knowledge of the biology and • Define standard resection margins in the treatment
behavior of breast cancer, the treatment of breast cancer of breast cancer.
has evolved substantially. It is now understood that breast • Understand appropriate surgical management of T4
cancer metastasis is driven by tumor biology, hence en breast cancer.
bloc resection does not improve outcomes over operations • Describe the surgical management of breast tumors
that extirpate cancer while minimizing sacrifice of normal following neoadjuvant chemotherapy.
tissue. In turn, efficacious new systemic therapies, includ- • Understand the evolution of the surgical manage-
ing biologically targeted therapies, permit further de-­ ment of the axillary lymph nodes.
escalation of surgical therapy. For patients presenting • Understand the role of surgery in the management
with an intact primary breast tumor and synchronous dis- of stage IV breast cancer.
tant metastases, recent data suggest that surgery for the
primary site does not improve overall survival, providing
further evidence that tumor biology cannot be overcome
by surgery.
37.1 Introduction

Modern breast cancer surgery began with Dr. William

Halsted’s hypothesis that breast cancer spreads in a stepwise
fashion from local to regional to distant sites and his descrip-
tion of the radical mastectomy in 1894. With the introduction
of antineoplastic chemotherapy and radiation therapy, and
improved knowledge of the biology and behavior of breast
cancer, the treatment of breast cancer has evolved substan-
tially. It is now understood that breast cancer metastasis is
A. M. Terando (*) · A. Carr
driven by tumor biology, hence en bloc resection does not
University of Southern California, Keck School of Medicine,
Los Angeles, CA, USA improve outcomes over operations that extirpate cancer
e-mail: [email protected]; [email protected] while minimizing sacrifice of normal tissue. In turn, effica-
T. J. Hieken · M. A. Piltin cious new systemic therapies, including biologically targeted
Mayo Clinic, Rochester, MN, USA therapies, permit further de-escalation of surgical therapy.
e-mail: [email protected]; [email protected] For patients presenting with an intact primary breast tumor
B. Chandrasekaran · C. S. Fisher and synchronous distant metastases, recent data suggest that
Indiana University School of Medicine, Indianapolis, IN, USA surgery for the primary site does not improve overall ­survival,
e-mail: [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 395
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_37
396 A. M. Terando et al.

providing further evidence that tumor biology cannot be endocrine therapy is often used alone except in situations of
overcome by surgery. high-risk multigene expression assay results, grade 3 dis-
ease, or involvement of multiple lymph nodes. Luminal B
tumors have a worse prognosis than Luminal A and are most
37.2 Breast Cancer Tumor Biology often recommended cytotoxic systemic therapy along with
endocrine therapy [10]. Due to the aggressive nature of Her2/
Beyond the key markers used to stratify breast cancer sub- neu-amplified tumors, anti-Her2/neu therapy in combination
types (estrogen receptor [ER], progesterone receptor [PR], with cytotoxic chemotherapy is recommended for tumors
and human epidermal growth factor [Her2/neu]), more ≥0.5 cm and/or those with nodal metastases. Basal-like
sophisticated gene expression profiling is now used to pre- TNBC is generally treated with cytotoxic chemotherapy due
dict tumor behavior and selection of adjuvant therapies even to the lack of known targetable proteins or mutations.
beyond that of the four main breast cancer subtypes: Luminal Interestingly, there is a higher incidence of basal-like tumors
A, Luminal B, Her2/neu amplified, and basal like. Due to in African Americans [11].
variation in prognosis and treatment sensitivity, knowledge Different molecular subtypes of breast cancer also behave
of these subtypes is important in guiding treatment. Tumors differently as far as the risk of early recurrence and metasta-
may be classified into intrinsic subtypes by multigene expres- sis [10, 12, 13]. Luminal A tumors, which are the most com-
sion molecular assays (PAM50—predictor analysis of mon, carry the best prognosis of all breast cancer subtypes
Microarray 50). Since these tests are not readily available, [10, 14, 15]. They are associated with a lower overall risk of
pathology-based surrogate definitions were proposed and distant metastasis, although metastases have been known to
have since been revised to further distinguish Luminal A and present much later, on the order of 5–15 years following
Luminal B subtypes [1–3]. Luminal A is defined as ER and completion of primary therapy [12, 16, 17]. On the other
PR positive (≥20%), Her2/neu non-amplified, Ki-67 low hand, basal like and Her2/neu amplified/ER-negative tumors
(<14%), with a low recurrence risk based on multigene are more likely to metastasize early, within 5 years after
expression assay results, if available. Luminal B is defined as diagnosis and treatment [18]. Additional studies of Her2/neu
ER positive, Her2/neu non-amplified, and at least one of the amplified tumors have identified significant heterogeneity
following: Ki67 high (>14%), PR negative or low positive within this group, with some Her2/neu amplified tumors
(<20%), with a high risk of recurrence based on multigene exhibiting luminal A features or basal-like features, which
expression assay results. Other studies have recommended may affect outcomes and response to treatment [19–21].
higher cutoffs for Ki 67 of ≥20% [4]. Tumors that are Her2/ Disease-free survival and specific sites of metastasis may
neu-amplified and ER/PR positive are also classified as also differ based on tumor subtype. A retrospective review of
Luminal B. Basal-like tumors are ER/PR negative, and Her2/ 1951 patients with node-negative, early-stage breast cancer
neu non-amplified, commonly referred to as triple-negative showed that disease-free survival at 10 years was higher for
breast carcinoma (TNBC). However, not all TNBCs are Luminal A breast cancers (86%) compared to Luminal B
basal like: uncommon types of TNBC such as medullary and (76%), Her2/neu amplified (73%), and triple-negative breast
adenoid cystic carcinoma are not considered basal like and cancers (71%, p < 0.001) [16]. Luminal B tumors are more
have a low risk of distant recurrence [1, 5]. Conversely, likely than other subtypes to have first recurrences in the
metaplastic breast cancer, while being triple-negative, has a bones. A retrospective review of 3762 patients with a median
worse prognosis than classic basal-like TNBC and responds follow-up of 14.8 years found, on multivariate analysis,
differently to systemic therapies [6]. which compared with luminal A tumors, luminal B Her2/neu
The choice of systemic treatment for breast cancer is amplified, and Her2/neu amplified tumors had significantly
driven by subtype. Endocrine therapy is an integral part of higher rates of brain, liver, and lung metastases [12]. Basal-­
treatment for all patients with hormone receptor-positive like tumors had higher rates of brain, lung, and distant nodal
tumors. There are a number of prognostic assays available metastases, but lower rates of liver and bone metastases. In a
for Luminal A and Luminal B tumors (not Her2/neu ampli- study of breast cancers diagnosed from 2010 to 2014 and
fied) that estimate recurrence risk (OncotypeDx®, reported in the SEER database, the rates of concurrent dis-
EndoPredict, PAM50, Breast Cancer Index, and tant metastases were 3.28% for bone, 1.52% for lung, 1.2%
MammaPrint), however only OncotypeDx® Recurrence for liver, and 0.35% for the brain [13]. On multivariate analy-
Score is validated for predicting adjuvant chemotherapy ben- sis, when compared to hormone receptor-positive, Her2/neu
efit in reducing the risk of recurrence and thus included in non-amplified tumors (luminal A and Her2/neu non-­
National Comprehensive Cancer Network Treatment amplified luminal B), Her2/neu amplified subtypes, whether
Guidelines for Breast Cancer [7–9]. For Luminal A tumors, hormone receptor-positive or not, had significantly higher
37 The Role of Surgery in Managing Primary and Metastatic Breast Cancer 397

rates of liver and lung metastasis. TNBCs had a higher rate risk of regional and distant recurrence for a given tumor and
of the brain (OR 1.95), liver (OR, 1.35), and lung (OR, 1.34) is influenced by biologic subtype and adjuvant systemic ther-
metastasis but significantly lower rate of bone metastasis apies to a greater extent than the choice for or against breast
than hormone receptor-positive, Her2/neu non-amplified. preservation when appropriate surgical care is provided.
Overall prognosis stratified by metastatic site was also worse Therefore, shared decision-making is a key component of
for triple-negative subtypes. In a cohort study of patients breast surgical care. When an adequate oncologic surgical
treated with breast-conserving therapy for early-stage breast resection can be performed, the correct choice is one made
cancer, the overall rate of brain metastasis was less than 2% by an informed patient reflecting their own values and pref-
at 5 years and less than 5% at 10 years. However, luminal B erences [28].
(12%), luminal B Her2/neu amplified (7%), Her2/neu ampli- Technical standards for the conduct of breast oncologic
fied (12%), and basal-like (7%) tumors were more likely to surgery have been published under the auspices of the Cancer
develop brain metastasis [22]. Even for patients whose Programs of the American College of Surgeons and provide
tumors fall into the same subtype, there remains consider- an excellent resource for the surgeon [28].
able heterogeneity as far as clinical outcomes. Further refine-
ment of breast cancer subtypes is an area of active research
which may help better determine the prognosis for individual 37.3.2 P
 athologic Margins for Breast Cancer
patients in the future [23]. Surgery

One critical technical standard for breast cancer operations is

37.3  urgery for the Primary Site in Breast
S resection to achieve microscopically negative margins with
Cancer care taken to avoid violating the tumor. The surgeon must
properly orient the specimen for the pathologist who then
Early breast cancer refers to breast cancer limited to the reports on both gross and microscopic margins [8, 28]. Based
breast and/or regional lymph nodes that are presumed cur- on a meta-analysis of 33 studies published in 2014 [29], in an
able. Notwithstanding notable advances in systemic therapy, effort to decrease secondary operations for margin re-­excision
surgery remains an essential component of treatment for which averaged 30% at the time [30], the Society of Surgical
early breast cancer. Oncology and the American Society of Radiation Oncology
published consensus guidelines regarding appropriate nega-
tive margin width in the setting of breast-­conserving surgery
37.3.1 Breast Conservation Versus with planned whole breast radiation. These guidelines, subse-
Mastectomy quently endorsed by the American Society for Clinical
Oncology, state that no-tumor-on-ink is an acceptable margin
There are two main options for surgical management of for patients with stage I and II breast cancer treated with
breast tumors, namely mastectomy or breast-conserving sur- lumpectomy and adjuvant radiation [31, 32].
gery. Mastectomy is recommended for multicentric disease No consensus guidelines exist for pathologic margins
and for large or unfavorably located tumors when resection when mastectomy is performed for breast cancer treatment.
with negative margins would preclude acceptable cosmesis. A recent systematic review of 34 studies concluded that posi-
Mastectomy can be done with or without immediate recon- tive margins were associated with a 2.6-fold increased risk of
struction (first stage or definitive) thus enabling preservation local recurrence, but there was no clear association with mar-
of an envelope of the breast skin and, in selected cases, even gin width. Acknowledging that margin width was defined
the nipple–areolar complex [8, 24, 25]. For the remainder of differently across the studies, similar hazard ratios were seen
patients, either approach is acceptable, and decision-making for positive/close margins within 2 mm [33]. A study of 1206
is multifaceted. In 2002, publication of the 20-year follow- mastectomies for cancer with meticulous pathologic margin
­up from two large randomized controlled clinical trials com- assessment, including 434 skin- or nipple-sparing mastecto-
paring lumpectomy with whole breast radiation to mies performed from 2006 to 2010 with a minimum 5-year
mastectomy undisputedly established the oncologic safety of follow-up, reported a 3.1% local recurrence rate at 5 years, a
breast preservation showing no difference in overall or lower local recurrence rate with negative margins of >2 mm
disease-­free survival [26] [27]. In-breast tumor recurrence versus closer margins, a linear decrease in recurrence for
(IBTR) is higher among patients with a retained breast, as negative margin width up to 10 mm and no difference
would be expected, but the risk of local recurrence is not nil between patients treated with total mastectomy versus skin-
with mastectomy. The risk of local recurrence parallels the or nipple-sparing mastectomy [34].
398 A. M. Terando et al.

37.3.3 Surgical Treatment of T4 Breast Cancers 37.4  urgery for Nodal Metastases
S
in Breast Cancer
T4 tumors of the breast include those with invasion of the
chest wall beyond the pectoralis major muscle (T4a), skin 37.4.1 Axillary Node Dissection for All
ulceration or nodules (T4b), or inflammatory breast cancer
(T4d) defined clinically as erythema or edema of less than or Until the 1990s, axillary lymph node dissection (ALND)
equal to one-third of the breast for ≤6 months with biopsy-­ remained an important part of the surgical treatment for all
confirmed underlying invasive breast cancer. Notably, a skin patients with breast cancer because of the pathological status
biopsy to assess tumor emboli in dermal lymphatics neither of the lymph nodes contributed to breast cancer staging and
confirms nor excludes the diagnosis [8, 35]. These tumors prognostication. Additionally, at the time, removal of the
pose local management challenges for the breast surgical axillary lymph nodes was felt to be necessary to achieve
oncologist. The standard-of-care recommendation in these locoregional control. The presence of axillary metastasis is
situations is for total mastectomy with en bloc resection of known to be associated with reduced disease free and overall
the affected chest wall for T4a tumors, and resection includ- survival compared to patients without nodal disease [39].
ing skin nodules or ulcers for T4b tumors, although breast ALND involves removal of nodal tissue between the ana-
conservation may be considered for highly selected patients tomic landmarks of the axillary vein superiorly, the chest
with favorable biology and noninflammatory T4 disease with wall medially, and the thoracodorsal bundle and latissimus
limited skin and/or chest wall involvement [8, 35, 36]. dorsi muscle laterally, generally yielding approximately
Neoadjuvant systemic therapy is recommended for inflam- 10–20 nodes. While ALND, with or without mastectomy, has
matory breast cancer and often for other T4 tumors as well. less morbidity than the radical mastectomy, it is still associ-
The often-widespread dermal lymphatic involvement seen in ated with significant morbidity including paresthesia,
inflammatory breast cancer, along with high rates (~90%) of seroma, lymphedema, axillary web formation, pain,
regional nodal involvement at presentation, has thus far sty- decreased shoulder mobility and rarely, nerve palsy [40–42].
mied efforts at breast and/or axillary conservation with high Thus, surgeons sought to develop an alternative procedure to
false-negative rates reported for sentinel node surgery. stage breast cancer to avoid the morbidity of ALND.
De-escalating surgical treatment in exceptional responders
to neoadjuvant systemic therapy might be investigated in the
future as novel and increasingly efficacious regimens are 37.4.2 I ntroduction of the Sentinel Lymph
developed. Node Biopsy

The concept of the sentinel lymph node (SLN) was first

37.3.4 M
 anagement of the Primary Breast developed for melanoma surgery in the late 1980s [43]. The
Tumor After Neoadjuvant Therapy SLN is defined as the first lymph node to receive lymphatic
drainage from a tumor and therefore should be the first site of
For patients for whom chemotherapy is indicated at the time lymphatic metastases. Identification of these nodes was per-
of diagnosis, administration of chemotherapy in the neoadju- formed via injection of radioactive tracers and/or vital blue
vant setting can allow de-escalation of surgical treatment. dye into the breast tissue. The sentinel lymph node biopsy
The index tumor marked with a clip may substantially (SLNB) technique was subsequently adapted for patients
decrease in size following neoadjuvant therapy to permit with breast cancer by Giuliano et al. who published their
breast-conserving surgery, especially in patients with experience with the procedure in 114 patients with breast
ER-negative-Her2/neu non-amplified or Her2/neu amplified cancer in 1994 [44]. The SLNB was performed in patients
tumors [37]. With pathologic complete response rates in the with clinically negative axillae, meaning that there was no
breast as high as 50% for some tumor types, clinical trials evidence of metastatic nodal disease identified on physical
testing whether or not surgery might be avoided altogether in examination or by imaging. For patients with clinically posi-
selected patients have been initiated. The NRG BR005 trial tive nodes, a failed SLNB, or SLNB that returned showing
evaluated the use of image-guided core needle biopsies of metastatic breast cancer, a complete ALND was performed.
the primary breast tumor bed following neoadjuvant SLNB in patients with early-stage breast cancer was shown
chemotherapy. Recently reported results from this study
­ to be an oncologically safe procedure to establish selective
showed a low sensitivity (50%) for detection of residual dis- application of complete axillary node dissection in clinically
ease in the breast, suggesting further refinements are needed node-negative patients in two large randomized controlled
before surgical treatment can be safely omitted for these trials: the NSABP B-32 and the Axillary Lymphatic Mapping
patients [38]. Against Nodal Axillary Clearance (ALMANAC) trials [45–
37 The Role of Surgery in Managing Primary and Metastatic Breast Cancer 399

47]. Not surprisingly, patients undergoing SLNB had lower These findings remained true in a 10-year follow-up analysis
rates of lymphedema, pain, and improved arm mobility com- of the original Z0011 trial [55]. While the conduct and results
pared to ALND [48, 49]. of the Z0011 trial were initially met with skepticism, many
studies since its original publication have confirmed that
ALND can be safely omitted for patients meeting “Z0011
37.4.3 S
 entinel Lymph Node Biopsy criteria.” The findings of the Z0011 trial were quickly
and Micrometastases accepted by surgical oncologists, leading to a profound and
sweeping change in clinical practice that was soon reflected
Although the SLNB technique was a great advancement for in national guidelines supporting the omission of ALND in
patients and surgeons, it introduced a new pathologic finding patients meeting the inclusion criteria of the Z0011 trial [56].
which needed to be addressed. With more thorough examina-
tion of the SLNs and advances in pathology, there was
enhanced detection of very small amounts of disease in the 37.4.5 N
 odal Evaluation After Neoadjuvant
lymph nodes. Micrometastatic disease in lymph nodes is Chemotherapy
defined as disease measuring between 0.2 and 2 mm. In a
review of eight studies with at least 100 patients each and The increasing use of neoadjuvant chemotherapy (NAC) for
5 years follow-up, Tjan-Heijen et al. concluded that there breast cancer has led to an increasing and substantial number
was insufficient evidence to confirm that micrometastases of patients who require axillary nodal evaluation. Initially,
were of prognostic utility [50]. The International Breast there were no data to guide how and when axillary nodal
Cancer Study Group (IBCSG 23-01) published results com- evaluation should be performed relative to neoadjuvant treat-
paring T1–T2 patients with isolated tumor cells and micro- ment. For several years, the management of the axilla in
metastatic nodal disease who underwent ALND with patients these patients was determined by surgeon preference: some
who did not undergo ALND and showed no difference in surgeons favored SLNB prior to chemotherapy and others
disease-free survival between the two groups [51, 52]. This espoused complete ALND after chemotherapy. In the
trial confirmed that ALND can be safely omitted in patients absence of data, both approaches were criticized. On view
with micrometastatic nodal disease. was that SLNB prior to chemotherapy did not give patients
an opportunity to “downstage” following chemotherapy,
obligating patients with a positive sentinel lymph node to a
37.4.4 A
 merican College of Surgeons complete axillary lymph node dissection. On the other hand,
Oncology Group (ACOSOG) Z0011 Trial there were concerns related to the feasibility and accuracy of
SLNB after chemotherapy, related to fibrosis in lymphatic
As experience with SLNB grew, the question of the need for channels following chemotherapy that may result in altered
ALND in all patients with node positivity was raised. Many lymphatic drainage and potentially a nonuniform response to
patients with a positive sentinel node or nodes (micro- and treatment in the axillary nodes that would cause an unaccept-
macro-metastatic disease) had no further axillary nodal ably high false-negative rate (FNR) for SLNB after
involvement at the time of ALND. Moreover, as medical NAC. Several studies, including NSABP B-27, demonstrated
therapy was increasingly dictated by the receptor profile of that sentinel node identification rates and FNRs after chemo-
the tumor as well as molecular assays, the morbidity associ- therapy were not significantly different from the pre-­
ated with ALND became hard to justify for the small amount neoadjuvant therapy scenario [57, 58]. In clinically
of additional information gleaned from completion axillary node-negative patients, SLNB following NAC was soon con-
dissection. sidered to be reliable and became the standard of care.
The Z0011 trial enrolled patients with T1–T2 breast can- However, if the SLNB was positive or unable to be identi-
cers who were clinically node negative and planned to fied, a completion ALND was performed.
undergo breast-conserving surgery (BCS) and whole breast While SLNB became routine after neoadjuvant therapy
radiation, and who were found to have one or two positive for patients who were clinically node negative at presenta-
SLNs. These patients were randomized to completion ALND tion, all patients with nodal involvement at initial presenta-
or no further surgery. All patients received whole breast tion continued to undergo complete ALND following NAC,
­radiation and adjuvant endocrine and/or chemotherapy. The regardless of clinical or radiologic response. However, it
study showed no differences in overall survival, disease-free began to be appreciated that a substantial number of patients
survival, or locoregional recurrence between the two groups with clinical N1 disease at presentation experienced a com-
and concluded that there was no clinical benefit to ALND in plete response to therapy. Specifically, two studies reported
patients with low-volume axillary nodal disease [53, 54]. pathologic complete responses in the lymph nodes in
400 A. M. Terando et al.

12–65% of clinically node-positive patients who received 37.5.1 Treatment of Distant Nodal Metastases
neoadjuvant systemic therapy [59, 60]. The pathologic com-
plete response rates were highest in triple-negative and Her2/ Metastatic involvement of contralateral axillary lymph nodes
neu amplified cancers. These impressive rates of pathologic has traditionally been considered a distant metastatic disease
complete response led to a reevaluation of the need for com- process. However, there is evidence to suggest that in breast
plete ALND in patients who have a clinical response to neo- cancer this may be more appropriately considered local–
adjuvant therapy. Several studies were performed to examine regional disease. Contralateral lymph node recurrence may
the safety and accuracy of SLNB following NAC in patients be the result of aberrant lymphatic drainage from the ipsilat-
who were clinically node positive at presentation [61, 62]. eral breast to the contralateral nodal basin in light of previous
The key outcome for these studies was an acceptable breast and axillary surgery. A prospective study of sentinel
FNR. While the reported FNRs in these trials were felt to be node biopsy in recurrent breast cancer found that 58.9% of
prohibitively high, further reviews noted that FNR could be patients had aberrant drainage pathways—with a signifi-
lowered further by removing at least two sentinel nodes, cantly higher rate seen in patients who underwent prior
using dual tracers, and by excising the lymph node that was ALND (79.3%) compared to previous SLNB (25.0%,
initially biopsy proven to contain metastatic disease, which p < 0.001) [67]. Of 95 patients with successful lymphatic
could be identified via use of a clip placed at the time of ini- mapping, 14 (14.7%) mapped to the contralateral axilla, all
tial biopsy [63]. Piltin et al. showed that omission of ALND of whom all had undergone previous ALND. A systematic
for patients with nodal pathologic complete response to neo- literature review was recently performed of patients with an
adjuvant systemic therapy did not worsen 2-year oncologic isolated contralateral nodal recurrence of breast cancer fol-
outcomes [64]. lowing prior curative treatment, with or without ipsilateral
Unfortunately, there remains a substantial number of breast cancer recurrence [68]. The majority of these patients
patients with nodal involvement at the time of presentation received surgical and systemic treatment with curative intent
who do not have a pathologic complete response to chemo- rather than for palliation. Although the numbers were small
therapy in the lymph nodes. At the time of this writing, the and the study was retrospective, the overall survival for these
standard of care in this scenario is to perform a complete patients was higher than would generally be expected for
ALND followed by comprehensive nodal radiation. As a metastatic breast cancer (82.6% after a mean of 50 months).
continuation of the trend for de-escalation of surgical ther- While these results should be interpreted with caution, they
apy, it is now being questioned whether complete ALND suggest that surgery may be considered for contralateral axil-
provides any benefit over comprehensive nodal radiation in lary nodal metastases in the absence of other distant
these patients. The clinical trial, ALLIANCE A11202 [65], metastases.
will specifically examine the clinical outcomes of patients
with residual nodal disease following neoadjuvant systemic
therapy who undergo comprehensive nodal radiation with 37.5.2 Treatment of Oligometastatic Disease
ALND or comprehensive nodal radiation alone.
Treatment of distant metastases in breast cancer is aimed at
prolonging survival and alleviating possible associated
37.5  urgery for Distant Metastases
S symptoms while maintaining an appropriate quality of life.
in Breast Cancer Individual responses to treatment and outcomes are depen-
dent upon a number of factors, including tumor subtype, site
Patients who present with or develop metastatic breast can- of disease, and overall disease burden. Patients with visceral
cer are unlikely to be cured of their disease. The develop- metastases and short disease-free intervals following initial
ment of newer treatment agents over the past 3 decades has therapy will generally have a poorer outcome compared to
led to significant improvements in overall survival for patients with soft tissue and bone metastases [69]. Because
patients with metastatic breast cancer [66]. Generally, systemic treatment is directed by tumor subtype, a repeat
patients who develop metastatic disease are treated with sys- biopsy to establish the diagnosis and to reevaluate the recep-
temic therapy but there may be situations in which surgical tor status should be considered at the time potential metasta-
treatment of the primary or distant disease may be consid- ses are identified.
ered. At this time, there is no clear evidence that local treat- In specific situations, local intervention for isolated
ment improves survival, although some retrospective trials metastases may be considered. Patients are considered to
have suggested a possible benefit. When surgery for meta- have oligometastatic disease when a limited number of meta-
static disease is considered, a thorough evaluation for addi- static sites are identified which may be considered amenable
tional sites of disease is imperative. to local treatment [70]. Patients may be considered suitable
37 The Role of Surgery in Managing Primary and Metastatic Breast Cancer 401

for local treatment based on a limited extent of disease, an goal being to preserve liver function as much as possible and
overall good performance status, a longer disease-free inter- to obtain a complete resection/treatment. A systematic retro-
val between primary definitive therapy and the identification spective review of 553 patients in 19 studies who underwent
of metastatic disease, and the likelihood of achieving com- liver resection for metastatic breast cancer was reported in
plete resection. These situations are different from instances 2011 [86]. The median post-hepatectomy survival for
when urgent intervention may be performed to alleviate patients in this study was 40 months, with a 5-year survival
symptoms causing significant pain or loss of function such as of 40%, which is longer than would typically be expected for
cord compression or pathologic fracture. The goal in these patients with metastatic breast cancer. A cost-utility analysis
settings is palliation of symptoms and not to improve sur- of liver resection performed for metastatic breast cancer
vival. Patients with bone disease are generally considered for found that liver resection followed by postoperative conven-
local treatment if there is evidence of a fracture or impending tional systemic therapy compared to long-term treatment
fracture, debilitating pain, joint restriction, or spinal cord with systemic therapy was cost-effective [87]. Poor prognos-
compression. Local treatment is generally not recommended tic features include involved surgical margins and hormone
for asymptomatic patients with bony metastases due to the receptor-negative disease. Other local therapy options may
general slow growth and progression of bone disease in include SBRT and radiofrequency ablation (RFA) as well as
breast cancer [71]. various other less studied liver-directed therapies, which
An estimated 10–30% of women with breast cancer will have been used for colorectal cancer liver metastases, but the
develop brain metastases [72]. As discussed previously, data are limited on their utility in patients with breast
patients with Her2/neu amplified and basal-like breast can- cancer.
cer are at higher risk of developing brain metastasis [12, 73].
Treatment options include surgery, whole-brain radiother-
apy, and stereotactic radiosurgery [74]. The Breast Graded 37.5.3 M
 anagement of the Primary Site
Prognostic Assessment (GPA) is a prognostic index for in Patients with Distant Metastatic
breast cancer patients with brain metastases that have under- Disease at Presentation
gone a number of subsequent modifications [75]. This index
may be used to determine prognosis and help direct individu- A number of studies have attempted to determine whether
alized treatment. There has been conflicting evidence on the surgical resection of the primary breast tumor is of benefit
benefit of surgery followed by whole-brain radiation therapy for patients with metastatic breast cancer. Several retrospec-
versus whole brain radiation alone [76–78]. However, guide- tive studies have suggested a survival advantage for patients
lines do recommend considering surgical resection in with metastatic breast cancer who undergo surgery of the
patients with a limited number of brain metastases, espe- primary site. However, this has not been clearly demon-
cially when other systemic disease is absent and the patient strated in prospective studies. Retrospective studies include
has a good performance score [79]. reviews of the National Cancer Database [88] and the
Patients with lung metastases will rarely become symp- Surveillance, Epidemiology, and End Results database [89]
tomatic until there is significant, unresectable tumor burden as well as institutional reviews [90], each of which found
[80]. In fact, most lung metastases are asymptomatic and survival advantages for patients who received local extirpa-
found incidentally [81]. In addition to metastatic disease, tive therapy as well as systemic therapy. However, matched-­
lung lesions identified in breast cancer patients may also rep- pair analyses have found similar survival outcomes for
resent primary lung cancer or benign, inflammatory, or infec- patients treated with surgery or without [91, 92], suggesting
tious processes. Depending on the size and location of the that the survival benefit demonstrated in the forgoing retro-
lesion, image-guided biopsy or surgical excision may be spective studies is due to selection bias.
used to obtain a tissue diagnosis [82]. Retrospective studies Two international prospective trials have attempted to eval-
have demonstrated an improved survival for patients under- uate the potential benefit of surgery for the intact primary site
going pulmonary metastasectomy, although selection bias in patients with metastatic breast cancer. The first trial ran-
likely plays a role in these findings [83, 84]. domized 350 women with stage IV breast cancer who had par-
While liver metastases occur in half of the patients with tial or complete responses to chemotherapy to surgery for the
metastatic breast cancer, there is limited data on the benefit primary site versus no locoregional treatment, finding no dif-
of liver resection. Only a small number of patients, (5–25%) ference in overall survival between the groups [93]. Notably,
will have isolated liver metastases [85]. Urgent indications however, this study did not use standard chemotherapy and
for local therapy include pain, uncontrolled bleeding, or bili- endocrine therapy regimens. A second study randomized
ary obstruction. Treatment options include hepatic resection patients with metastatic breast cancer and intact primary
and stereotactic body radiation therapy (SBRT), with the tumors to surgery first followed by systemic therapy versus
402 A. M. Terando et al.

systemic therapy alone [94]. At 3 years, no significant differ- Author Disclosures Alicia M. Terando, MD has no disclosures.
Azadeh Carr, MD has no disclosures. Tina J. Hieken, MD receives
ence could be seen between the groups, but at 5 years there research support from Genentech, Breast Cancer Research Foundation
was improved survival in patients who received local therapy and SkylineDx. Mara A. Piltin, DO has no disclosures. Bindupriya
(46.4% vs. 26.4%; HR 0.66, 95% CI 0.49–0.88). It should be Chandrasekaran, MD has no disclosures. Carla S. Fisher, MD, MBA
noted, however, that the local therapy group had few patients has no disclosures
with TNBC and visceral metastases. Furthermore, surgery for
the primary site was performed before systemic therapy so
patients’ response to chemotherapy could not be ascertained. References
A recent randomized phase III trial presented at the American
Society of Clinical Oncology Annual meeting in 2020— 1. Goldhirsch A, Winer EP, Coates AS, Gelber RD, Piccart-Gebhart
M, Thurlimann B, Senn HJ, Panel members. Personalizing the
Eastern Cooperative Oncology Group E2108—evaluated sys- treatment of women with early breast cancer: highlights of the St
temic therapy plus early local therapy versus systemic therapy Gallen international expert consensus on the primary therapy of
alone in women with de novo stage IV breast cancer [95]. early breast cancer 2013. Ann Oncol. 2013;24(9):2206–23. https://
Enrolled patients received optimal systemic therapy up-front doi.org/10.1093/annonc/mdt303.
2. Goldhirsch A, Wood WC, Coates AS, Gelber RD, Thurlimann B,
and those who did not progress during 4–8 months of treat- Senn HJ, Panel members. Strategies for subtypes—dealing with the
ment were randomized to local therapy versus none. There diversity of breast cancer: highlights of the St. Gallen international
were no significant differences observed in overall survival or expert consensus on the primary therapy of early breast cancer
progression-free survival. The current data suggest that locore- 2011. Ann Oncol. 2011;22(8):1736–47. https://doi.org/10.1093/
annonc/mdr304.
gional treatment for metastatic breast cancer does not show a 3. Prat A, Pineda E, Adamo B, Galvan P, Fernandez A, Gaba L, Diez
clear survival benefit and should be considered only in patients M, Viladot M, Arance A, Munoz M. Clinical implications of the
with good response to systemic therapy, oligometastatic dis- intrinsic molecular subtypes of breast cancer. Breast. 2015;24(Suppl
ease, and as part of a multidisciplinary treatment plan. 2):S26–35. https://doi.org/10.1016/j.breast.2015.07.008.
4. Maisonneuve P, Disalvatore D, Rotmensz N, Curigliano G,
Colleoni M, Dellapasqua S, Pruneri G, Mastropasqua MG, Luini
A, Bassi F, Pagani G, Viale G, Goldhirsch A. Proposed new clini-
37.6 Conclusion copathological surrogate definitions of luminal A and luminal B
(HER2-negative) intrinsic breast cancer subtypes. Breast Cancer
Res. 2014;16(3):R65. https://doi.org/10.1186/bcr3679.
Surgery plays a critical role in the multidisciplinary manage- 5. Lehmann BD, Bauer JA, Chen X, Sanders ME, Chakravarthy
ment of breast cancer. Appropriate adherence to best clinical AB, Shyr Y, Pietenpol JA. Identification of human triple-negative
practices, as they evolve, from the standpoint of selection of breast cancer subtypes and preclinical models for selection of tar-
the appropriate surgical treatment plan, sequencing with geted therapies. J Clin Invest. 2011;121(7):2750–67. https://doi.
org/10.1172/JCI45014.
other therapies, and the technical conduct of breast cancer 6. Aydiner A, Sen F, Tambas M, Ciftci R, Eralp Y, Saip P, Karanlik
operations will ensure optimal patient outcomes. H, Fayda M, Kucucuk S, Onder S, Yavuz E, Muslumanoglu
M, Igci A. Metaplastic breast carcinoma versus triple-negative
breast cancer: survival and response to treatment. Medicine
(Baltimore). 2015;94(52):e2341. https://doi.org/10.1097/
Open Questions MD.0000000000002341.
• How might consideration of tumor biologic subtype 7. Krop I, Ismaila N, Andre F, Bast RC, Barlow W, Collyar DE,
further refine the choice of operation for the breast Hammond ME, Kuderer NM, Liu MC, Mennel RG, Van Poznak C,
Wolff AC, Stearns V. Use of biomarkers to guide decisions on adju-
and axillary lymph nodes? vant systemic therapy for women with early-stage invasive breast
• Will tumor biologic subtype be further refined by cancer: American Society of Clinical Oncology clinical practice
enhanced genomic profiling? If so, will this be uti- guideline focused update. J Clin Oncol. 2017;35(24):2838–47.
lized in clinical practice? https://doi.org/10.1200/JCO.2017.74.0472.
8. NCCN (2020) National Comprehensive Cancer Network Clinical
• How will a clinical complete response to neoadju- Practice Guidelines in Oncology. Breast Cancer v6.2020.
vant chemotherapy be defined in the future? 9. Sparano JA, Gray RJ, Makower DF, Pritchard KI, Albain KS,
• Will surgery always be required for patients who Hayes DF, Geyer CE Jr, Dees EC, Goetz MP, Olson JA Jr, Lively
have a complete clinical response to neoadjuvant T, Badve SS, Saphner TJ, Wagner LI, Whelan TJ, Ellis MJ, Paik
S, Wood WC, Ravdin PM, Keane MM, Gomez Moreno HL,
chemotherapy? Reddy PS, Goggins TF, Mayer IA, Brufsky AM, Toppmeyer DL,
• Can axillary radiation replace axillary dissection in Kaklamani VG, Berenberg JL, Abrams J, Sledge GW Jr. Adjuvant
patients with residual nodal disease following neo- chemotherapy guided by a 21-gene expression assay in Breast can-
adjuvant chemotherapy? cer. N Engl J Med. 2018;379(2):111–21. https://doi.org/10.1056/
NEJMoa1804710.
37 The Role of Surgery in Managing Primary and Metastatic Breast Cancer 403

10. Voduc KD, Cheang MC, Tyldesley S, Gelmon K, Nielsen TO, and characteristics by biologic subtype. Breast Cancer Res Treat.
Kennecke H. Breast cancer subtypes and the risk of local and 2012;136(1):153–60. https://doi.org/10.1007/s10549-012-2243-x.
regional relapse. J Clin Oncol. 2010;28(10):1684–91. https://doi. 23. Rueda OM, Sammut SJ, Seoane JA, Chin SF, Caswell-Jin JL,
org/10.1200/JCO.2009.24.9284. Callari M, Batra R, Pereira B, Bruna A, Ali HR, Provenzano E,
11. Millikan RC, Newman B, Tse CK, Moorman PG, Conway K, Liu B, Parisien M, Gillett C, McKinney S, Green AR, Murphy L,
Dressler LG, Smith LV, Labbok MH, Geradts J, Bensen JT, Jackson Purushotham A, Ellis IO, Pharoah PD, Rueda C, Aparicio S, Caldas
S, Nyante S, Livasy C, Carey L, Earp HS, Perou CM. Epidemiology C, Curtis C. Dynamics of breast-cancer relapse reveal late-recurring
of basal-like breast cancer. Breast Cancer Res Treat. 2008;109 ER-positive genomic subgroups. Nature. 2019;567(7748):399–404.
(1):123–39. https://doi.org/10.1007/s10549-007-9632-6. https://doi.org/10.1038/s41586-019-1007-8.
12. Kennecke H, Yerushalmi R, Woods R, Cheang MC, Voduc D, 24. Sada A, Day CN, Hoskin TL, Degnim AC, Habermann EB, Hieken
Speers CH, Nielsen TO, Gelmon K. Metastatic behavior of breast TJ. Mastectomy and immediate breast reconstruction in the elderly:
cancer subtypes. J Clin Oncol. 2010;28(20):3271–7. https://doi. trends and outcomes. Surgery. 2019;166(4):709–14. https://doi.
org/10.1200/JCO.2009.25.9820. org/10.1016/j.surg.2019.05.055.
13. Xiao W, Zheng S, Yang A, Zhang X, Zou Y, Tang H, Xie X. Breast 25. Young WA, Degnim AC, Hoskin TL, Jakub JW, Nguyen MD, Tran
cancer subtypes and the risk of distant metastasis at initial diagno- NV, Harless CA, Manrique OJ, Boughey JC, Hieken TJ. Outcomes
sis: a population-based study. Cancer Manag Res. 2018;10:5329– of >1300 nipple-sparing mastectomies with immediate recon-
38. https://doi.org/10.2147/CMAR.S176763. struction: the impact of expanding indications on complications.
14. Sorlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Johnsen H, Ann Surg Oncol. 2019;26(10):3115–23. https://doi.org/10.1245/
Hastie T, Eisen MB, van de Rijn M, Jeffrey SS, Thorsen T, Quist s10434-­019-­07560-­z.
H, Matese JC, Brown PO, Botstein D, Lonning PE, Borresen-­ 26. Fisher B, Anderson S, Bryant J, Margolese RG, Deutsch M, Fisher
Dale AL. Gene expression patterns of breast carcinomas dis- ER, Jeong JH, Wolmark N. Twenty-year follow-up of a random-
tinguish tumor subclasses with clinical implications. Proc Natl ized trial comparing total mastectomy, lumpectomy, and lumpec-
Acad Sci U S A. 2001;98(19):10869–74. https://doi.org/10.1073/ tomy plus irradiation for the treatment of invasive breast cancer.
pnas.191367098. N Engl J Med. 2002;347(16):1233–41. https://doi.org/10.1056/
15. Sotiriou C, Pusztai L. Gene-expression signatures in breast can- NEJMoa022152.
cer. N Engl J Med. 2009;360(8):790–800. https://doi.org/10.1056/ 27. Veronesi U, Cascinelli N, Mariani L, Greco M, Saccozzi R, Luini
NEJMra0801289. A, Aguilar M, Marubini E. Twenty-year follow-up of a randomized
16. Metzger-Filho O, Sun Z, Viale G, Price KN, Crivellari D, Snyder study comparing breast-conserving surgery with radical mastec-
RD, Gelber RD, Castiglione-Gertsch M, Coates AS, Goldhirsch A, tomy for early breast cancer. N Engl J Med. 2002;347(16):1227–
Cardoso F. Patterns of recurrence and outcome according to breast 32. https://doi.org/10.1056/NEJMoa020989.
cancer subtypes in lymph node-negative disease: results from inter- 28. Nelson H, Hunt K, Veeramachaneni N, Blair S, Chang G, Halverson
national breast cancer study group trials VIII and IX. J Clin Oncol. A, Katz M, Posner M, editors. Operative standards for cancer sur-
2013;31(25):3083–90. https://doi.org/10.1200/JCO.2012.46.1574. gery, Breast, lung, pancreas, colon, vol. 1. Philadelphia: Lippincott,
17. Minicozzi P, Bella F, Toss A, Giacomin A, Fusco M, Zarcone Williams & Wilkins; 2015.
M, Tumino R, Falcini F, Cesaraccio R, Candela G, La Rosa F, 29. Houssami N, Macaskill P, Marinovich ML, Morrow M. The associ-
Federico M, Sant M. Relative and disease-free survival for breast ation of surgical margins and local recurrence in women with early-­
cancer in relation to subtype: a population-based study. J Cancer stage invasive breast cancer treated with breast-conserving therapy:
Res Clin Oncol. 2013;139(9):1569–77. https://doi.org/10.1007/ a meta-analysis. Ann Surg Oncol. 2014;21(3):717–30. https://doi.
s00432-­013-­1478-­1. org/10.1245/s10434-­014-­3480-­5.
18. Savci-Heijink CD, Halfwerk H, Hooijer GK, Horlings HM, 30. Boughey JC, Hieken TJ, Jakub JW, Degnim AC, Grant CS, Farley
Wesseling J, van de Vijver MJ. Retrospective analysis of metastatic DR, Thomsen KM, Osborn JB, Keeney GL, Habermann EB. Impact
behaviour of breast cancer subtypes. Breast Cancer Res Treat. of analysis of frozen-section margin on reoperation rates in women
2015;150(3):547–57. https://doi.org/10.1007/s10549-­015-­3352-­0. undergoing lumpectomy for breast cancer: evaluation of the
19. Carey LA, Berry DA, Cirrincione CT, Barry WT, Pitcher BN, Harris National Surgical Quality Improvement Program data. Surgery.
LN, Ollila DW, Krop IE, Henry NL, Weckstein DJ, Anders CK, 2014a;156(1):190–7. https://doi.org/10.1016/j.surg.2014.03.025.
Singh B, Hoadley KA, Iglesia M, Cheang MC, Perou CM, Winer 31. Buchholz TA, Somerfield MR, Griggs JJ, El-Eid S, Hammond ME,
EP, Hudis CA. Molecular heterogeneity and response to neoadju- Lyman GH, Mason G, Newman LA. Margins for breast-conserving
vant human epidermal growth factor receptor 2 targeting in CALGB surgery with whole-breast irradiation in stage I and II invasive breast
40601, a randomized phase III trial of paclitaxel plus Trastuzumab cancer: American Society of Clinical Oncology endorsement of
with or without Lapatinib. J Clin Oncol. 2016;34(6):542–9. https:// the Society of Surgical Oncology/American Society for Radiation
doi.org/10.1200/JCO.2015.62.1268. Oncology consensus guideline. J Clin Oncol. 2014;32(14):1502–6.
20. Llombart-Cussac A, Cortes J, Pare L, Galvan P, Bermejo B, Martinez https://doi.org/10.1200/JCO.2014.55.1572.
N, Vidal M, Pernas S, Lopez R, Munoz M, Nuciforo P, Morales S, 32. Moran MS, Schnitt SJ, Giuliano AE, Harris JR, Khan SA, Horton
Oliveira M, de la Pena L, Pelaez A, Prat A. HER2-enriched subtype J, Klimberg S, Chavez-MacGregor M, Freedman G, Houssami N,
as a predictor of pathological complete response following trastu- Johnson PL, Morrow M. Society of Surgical Oncology-American
zumab and lapatinib without chemotherapy in early-stage HER2-­ Society for Radiation Oncology consensus guideline on margins for
positive breast cancer (PAMELA): an open-label, single-group, breast-conserving surgery with whole-breast irradiation in stages I
multicentre, phase 2 trial. Lancet Oncol. 2017;18(4):545–54. and II invasive breast cancer. Ann Surg Oncol. 2014;21(3):704–16.
https://doi.org/10.1016/S1470-­2045(17)30021-­9. https://doi.org/10.1245/s10434-­014-­3481-­4.
21. Prat A, Carey LA, Adamo B, Vidal M, Tabernero J, Cortes J, Parker 33. Bundred J, Michael S, Bowers S, Barnes N, Jauhari Y, Plant D,
JS, Perou CM, Baselga J. Molecular features and survival outcomes Maishman T, Cutress R, Holleczek B, Dodwell D, Bundred N. Do
of the intrinsic subtypes within HER2-positive breast cancer. J Natl surgical margins matter after mastectomy? A systematic review.
Cancer Inst. 2014;106(8):dju152. https://doi.org/10.1093/jnci/dju152. Eur J Surg Oncol. 2020;46(12):2185–94. https://doi.org/10.1016/j.
22. Arvold ND, Oh KS, Niemierko A, Taghian AG, Lin NU, Abi-Raad ejso.2020.08.015.
RF, Sreedhara M, Harris JR, Alexander BM. Brain metastases 34. Glorioso JM, Gonzalez Juarrero AB, Rodysill BR, Harmsen
after breastconserving therapy and systemic therapy: incidence WS, Habermann EB, Carter JM, Mutter RW, Degnim AC, Jakub
404 A. M. Terando et al.

JW. Margin proximity correlates with local recurrence after mas- 48. Purushotham AD, Upponi S, Klevesath MB, Bobrow L, Millar K,
tectomy for patients not receiving adjuvant radiotherapy. Ann Myles JP, Duffy SW. Morbidity after sentinel lymph node biopsy
Surg Oncol. 2017;24(11):3148–56. https://doi.org/10.1245/ in primary breast cancer: results from a randomized controlled
s10434-­017-­5968-­2. trial. J Clin Oncol. 2005;23(19):4312–21. https://doi.org/10.1200/
35. Murphy BL, Hoskin TL, Boughey JC, Degnim AC, Carter JM, JCO.2005.03.228.
Glazebrook KN, Hieken TJ. Contemporary operative management 49. Veronesi U, Paganelli G, Viale G, Luini A, Zurrida S, Galimberti
of T4 breast cancer. Surgery. 2016;160(4):1059–69. https://doi. V, Intra M, Veronesi P, Robertson C, Maisonneuve P, Renne G,
org/10.1016/j.surg.2016.06.030. De Cicco C, De Lucia F, Gennari R. A randomized comparison of
36. Hieken TJ, Murphy BL, Boughey JC, Degnim AC, Glazebrook KN, sentinel-node biopsy with routine axillary dissection in breast can-
Hoskin TL. Influence of biologic subtype of inflammatory breast cer. N Engl J Med. 2003;349(6):546–53. https://doi.org/10.1056/
cancer on response to neoadjuvant therapy and cancer outcomes. NEJMoa012782.
Clin Breast Cancer. 2018;18(4):e501–6. https://doi.org/10.1016/j. 50. Tjan-Heijnen VC, Buit P, de Widt-Evert LM, Ruers TJ, Beex
clbc.2017.10.003. LV. Micro-metastases in axillary lymph nodes: an increasing classi-
37. Leon Ferre R, Hieken TJ, Boughey JC. The landmark series: fication and treatment dilemma in breast cancer due to the introduc-
neoadjuvant chemotherapy for breast cancer. Ann Surg Oncol. tion of the sentinel lymph node procedure. Breast Cancer Res Treat.
2020;28(4):2111–9. 2001;70(2):81–8. https://doi.org/10.1023/a:1012938825396.
38. Basik M, Cecchini RS, De Los Santos JF. Abstract GS5-05: pri- 51. Galimberti V, Cole BF, Zurrida S, Viale G, Luini A, Veronesi
mary analysis of NRG-BR005, a phase II trial assessing accuracy P, Baratella P, Chifu C, Sargenti M, Intra M, Gentilini O,
of tumor bed biopsies in predicting pathologic complete response Mastropasqua MG, Mazzarol G, Massarut S, Garbay JR, Zgajnar J,
(pCR) in patients with clinical/radiological complete response Galatius H, Recalcati A, Littlejohn D, Bamert M, Colleoni M, Price
after neoadjuvant chemotherapy (NCT) to explore the feasibil- KN, Regan MM, Goldhirsch A, Coates AS, Gelber RD, Veronesi U,
ity of breast-conserving treatment without surgery. Cancer Res. International Breast Cancer Study Group Trial 23-01 Investigators.
2020;80 (4). Axillary dissection versus no axillary dissection in patients
39. Dings PJ, Elferink MA, Strobbe LJ, de Wilt JH. The prognos- with sentinel-node micrometastases (IBCSG 23-01): a phase 3
tic value of lymph node ratio in node-positive breast cancer: ­randomised controlled trial. Lancet Oncol. 2013;14(4):297–305.
a Dutch nationwide population-based study. Ann Surg Oncol. https://doi.org/10.1016/S1470-­2045(13)70035-­4.
2013;20(8):2607–14. https://doi.org/10.1245/s10434-­013-­2932-­7. 52. Hieken TJ, Boughey JC. Axillary dissection versus no axillary
40. Ashikaga T, Krag DN, Land SR, Julian TB, Anderson SJ, Brown dissection in patients with sentinel-node micrometastases: com-
AM, Skelly JM, Harlow SP, Weaver DL, Mamounas EP, Costantino mentary on the IBCSG 23-01 trial. Gland Surg. 2013;2(3):128–32.
JP, Wolmark N, National Surgical Adjuvant Breast, Bowel Project. https://doi.org/10.3978/j.issn.2227-­684X.2013.07.04.
Morbidity results from the NSABP B-32 trial comparing sentinel 53. Giuliano AE, Hunt KK, Ballman KV, Beitsch PD, Whitworth
lymph node dissection versus axillary dissection. J Surg Oncol. PW, Blumencranz PW, Leitch AM, Saha S, McCall LM, Morrow
2010;102(2):111–8. https://doi.org/10.1002/jso.21535. M. Axillary dissection vs no axillary dissection in women with
41. Warmuth MA, Bowen G, Prosnitz LR, Chu L, Broadwater G, invasive breast cancer and sentinel node metastasis: a randomized
Peterson B, Leight G, Winer EP. Complications of axillary lymph clinical trial. JAMA. 2011;305(6):569–75. https://doi.org/10.1001/
node dissection for carcinoma of the breast: a report based on a jama.2011.90.
patient survey. Cancer. 1998;83(7):1362–8. https://doi.org/10.1002/ 54. Giuliano AE, McCall L, Beitsch P, Whitworth PW, Blumencranz P,
(sici)1097-­0142(19981001)83:7<1362::aid-­cncr13>3.0.co;2-­2. Leitch AM, Saha S, Hunt KK, Morrow M, Ballman K. Locoregional
42. Zintel HA, Nay HR. Postoperative complications of radical mas- recurrence after sentinel lymph node dissection with or without
tectomy. Surg Clin North Am. 1964;44:313–23. https://doi. axillary dissection in patients with sentinel lymph node metastases:
org/10.1016/s0039-­6109(16)37231-­0. the American College of Surgeons Oncology Group Z0011 ran-
43. Morton DL, Wen DR, Wong JH, Economou JS, Cagle LA, Storm FK, domized trial. Ann Surg. 2010;252(3):426–32; discussion 432-423.
Foshag LJ, Cochran AJ. Technical details of intraoperative lymphatic https://doi.org/10.1097/SLA.0b013e3181f08f32.
mapping for early stage melanoma. Arch Surg. 1992;127(4):392–9. 55. Giuliano AE, Ballman KV, McCall L, Beitsch PD, Brennan MB,
https://doi.org/10.1001/archsurg.1992.01420040034005. Kelemen PR, Ollila DW, Hansen NM, Whitworth PW, Blumencranz
44. Giuliano AE, Kirgan DM, Guenther JM, Morton DL. Lymphatic PW, Leitch AM, Saha S, Hunt KK, Morrow M. Effect of axillary dis-
mapping and sentinel lymphadenectomy for breast cancer. Ann section vs no axillary dissection on 10-year overall survival among
Surg. 1994;220(3):391. women with invasive breast cancer and sentinel node metastasis:
45. Clarke D, Khonji NI, Mansel RE. Sentinel node biopsy in breast the ACOSOG Z0011 (Alliance) randomized clinical trial. JAMA.
cancer: ALMANAC trial. World J Surg. 2001;25(6):819–22. 2017;318(10):918–26. https://doi.org/10.1001/jama.2017.11470.
https://doi.org/10.1007/s00268-­001-­0011-­x. 56. Black DM, Mittendorf EA. Landmark trials affecting the surgi-
46. Krag DN, Anderson SJ, Julian TB, Brown AM, Harlow SP, cal management of invasive breast cancer. Surg Clin North Am.
Costantino JP, Ashikaga T, Weaver DL, Mamounas EP, Jalovec 2013;93(2):501–18. https://doi.org/10.1016/j.suc.2012.12.007.
LM, Frazier TG, Noyes RD, Robidoux A, Scarth HM, Wolmark 57. Hunt KK, Yi M, Mittendorf EA, Guerrero C, Babiera GV, Bedrosian
N. Sentinel-lymph-node resection compared with conventional I, Hwang RF, Kuerer HM, Ross MI, Meric-Bernstam F. Sentinel
axillary-lymph-node dissection in clinically node-negative patients lymph node surgery after neoadjuvant chemotherapy is accu-
with breast cancer: overall survival findings from the NSABP B-32 rate and reduces the need for axillary dissection in breast cancer
randomised phase 3 trial. Lancet Oncol. 2010;11(10):927–33. patients. Ann Surg. 2009;250(4):558–66. https://doi.org/10.1097/
https://doi.org/10.1016/S1470-­2045(10)70207-­2. SLA.0b013e3181b8fd5e.
47. Mansel RE, Fallowfield L, Kissin M, Goyal A, Newcombe RG, 58. Mamounas EP, Brown A, Anderson S, Smith R, Julian T, Miller
Dixon JM, Yiangou C, Horgan K, Bundred N, Monypenny I, B, Bear HD, Caldwell CB, Walker AP, Mikkelson WM, Stauffer
England D, Sibbering M, Abdullah TI, Barr L, Chetty U, Sinnett JS, Robidoux A, Theoret H, Soran A, Fisher B, Wickerham DL,
DH, Fleissig A, Clarke D, Ell PJ. Randomized multicenter trial Wolmark N. Sentinel node biopsy after neoadjuvant chemotherapy
of sentinel node biopsy versus standard axillary treatment in in breast cancer: results from National Surgical Adjuvant Breast
operable breast cancer: the ALMANAC trial. J Natl Cancer Inst. and Bowel Project Protocol B-27. J Clin Oncol. 2005;23(12):2694–
2006;98(9):599–609. https://doi.org/10.1093/jnci/djj158. 702. https://doi.org/10.1200/JCO.2005.05.188.
37 The Role of Surgery in Managing Primary and Metastatic Breast Cancer 405

59. Boughey JC, McCall LM, Ballman KV, Mittendorf EA, Ahrendt oligometastatic cancers: where will the evidence lead? Front Oncol.
GM, Wilke LG, Taback B, Leitch AM, Flippo-Morton T, Hunt 2019;9:543. https://doi.org/10.3389/fonc.2019.00543.
KK. Tumor biology correlates with rates of breast-conserving 71. Briasoulis E, Karavasilis V, Kostadima L, Ignatiadis M, Fountzilas
surgery and pathologic complete response after neoadjuvant G, Pavlidis N. Metastatic breast carcinoma confined to bone: por-
chemotherapy for breast cancer: findings from the ACOSOG trait of a clinical entity. Cancer. 2004;101(7):1524–8. https://doi.
Z1071 (Alliance) prospective multicenter clinical trial. Ann org/10.1002/cncr.20545.
Surg. 2014b;260(4):608–14; discussion 614–606. https://doi. 72. Lin NU, Bellon JR, Winer EP. CNS metastases in breast cancer.
org/10.1097/SLA.0000000000000924. J Clin Oncol. 2004;22(17):3608–17. https://doi.org/10.1200/
60. Hennessy BT, Hortobagyi GN, Rouzier R, Kuerer H, Sneige N, JCO.2004.01.175.
Buzdar AU, Kau SW, Fornage B, Sahin A, Broglio K, Singletary 73. Gabos Z, Sinha R, Hanson J, Chauhan N, Hugh J, Mackey JR,
SE, Valero V. Outcome after pathologic complete eradication of Abdulkarim B. Prognostic significance of human epidermal
cytologically proven breast cancer axillary node metastases follow- growth factor receptor positivity for the development of brain
ing primary chemotherapy. J Clin Oncol. 2005;23(36):9304–11. metastasis after newly diagnosed breast cancer. J Clin Oncol.
https://doi.org/10.1200/JCO.2005.02.5023. 2006;24(36):5658–63. https://doi.org/10.1200/JCO.2006.07.0250.
61. Boileau JF, Poirier B, Basik M, Holloway CM, Gaboury L, Sideris 74. Raghunath A, Desai K, Ahluwalia MS. Current treatment options
L, Meterissian S, Arnaout A, Brackstone M, McCready DR, Karp for Breast cancer brain metastases. Curr Treat Options in Oncol.
SE, Trop I, Lisbona A, Wright FC, Younan RJ, Provencher L, 2019;20(3):19. https://doi.org/10.1007/s11864-­019-­0618-­5.
Patocskai E, Omeroglu A, Robidoux A. Sentinel node biopsy after 75. Sperduto PW, Mesko S, Li J, Cagney D, Aizer A, Lin NU, Nesbit
neoadjuvant chemotherapy in biopsy-proven node-positive breast E, Kruser TJ, Chan J, Braunstein S, Lee J, Kirkpatrick JP, Breen
cancer: the SN FNAC study. J Clin Oncol. 2015;33(3):258–64. W, Brown PD, Shi D, Shih HA, Soliman H, Sahgal A, Shanley R,
https://doi.org/10.1200/JCO.2014.55.7827. Sperduto W, Lou E, Everett A, Boggs DH, Masucci L, Roberge D,
62. Kuehn T, Bauerfeind I, Fehm T, Fleige B, Hausschild M, Helms G, Remick J, Plichta K, Buatti JM, Jain S, Gaspar LE, Wu CC, Wang
Lebeau A, Liedtke C, von Minckwitz G, Nekljudova V, Schmatloch TJC, Bryant J, Chuong M, Yu J, Chiang V, Nakano T, Aoyama
S, Schrenk P, Staebler A, Untch M. Sentinel-lymph-node biopsy H, Mehta MP. Beyond an updated graded prognostic assessment
in patients with breast cancer before and after neoadjuvant che- (Breast GPA): a prognostic index and trends in treatment and
motherapy (SENTINA): a prospective, multicentre cohort study. survival in Breast cancer brain metastases from 1985 to today.
Lancet Oncol. 2013;14(7):609–18. https://doi.org/10.1016/ Int J Radiat Oncol Biol Phys. 2020;107(2):334–43. https://doi.
S1470-­2045(13)70166-­9. org/10.1016/j.ijrobp.2020.01.051.
63. Boughey JC, Ballman KV, Le-Petross HT, McCall LM, Mittendorf 76. Mintz AH, Kestle J, Rathbone MP, Gaspar L, Hugenholtz
EA, Ahrendt GM, Wilke LG, Taback B, Feliberti EC, Hunt H, Fisher B, Duncan G, Skingley P, Foster G, Levine M. A
KK. Identification and resection of clipped node decreases the randomized trial to assess the efficacy of surgery in addi-
false-negative rate of sentinel lymph node surgery in patients pre- tion to radiotherapy in patients with a single cerebral metas-
senting with node-positive Breast cancer (T0-T4, N1-N2) who tasis. Cancer. 1996;78(7):1470–6. https://doi.org/10.1002/
receive neoadjuvant chemotherapy: results from ACOSOG Z1071 (sici)1097-­0142(19961001)78:7<1470::aid-­cncr14>3.0.co;2-­x.
(Alliance). Ann Surg. 2016;263(4):802–7. https://doi.org/10.1097/ 77. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ, Maruyama Y,
SLA.0000000000001375. Kryscio RJ, Markesbery WR, Macdonald JS, Young B. A ran-
64. Piltin MA, Hoskin TL, Day CN, Davis J Jr, Boughey JC. Oncologic domized trial of surgery in the treatment of single metastases
outcomes of sentinel lymph node surgery after neoadjuvant che- to the brain. N Engl J Med. 1990;322(8):494–500. https://doi.
motherapy for node-positive breast cancer. Ann Surg Oncol. org/10.1056/NEJM199002223220802.
2020;27(12):4795–801. https://doi.org/10.1245/s10434-­020-­08900-­0. 78. Vecht CJ, Haaxma-Reiche H, Noordijk EM, Padberg GW,
65. NCT01901094 (2014) Comparison of axillary lymph node dissec- Voormolen JH, Hoekstra FH, Tans JT, Lambooij N, Metsaars
tion with axillary radiation for patients with node-positive breast JA, Wattendorff AR, et al. Treatment of single brain metastasis:
cancer treated with chemotherapy. https://ClinicalTrials.gov/show/ radiotherapy alone or combined with neurosurgery? Ann Neurol.
NCT01901094. 1993;33(6):583–90. https://doi.org/10.1002/ana.410330605.
66. Caswell-Jin JL, Plevritis SK, Tian L, Cadham CJ, Xu C, Stout NK, 79. Soffietti R, Abacioglu U, Baumert B, Combs SE, Kinhult S, Kros
Sledge GW, Mandelblatt JS, Kurian AW. Change in survival in met- JM, Marosi C, Metellus P, Radbruch A, Villa Freixa SS, Brada M,
astatic breast cancer with treatment advances: meta-analysis and Carapella CM, Preusser M, Le Rhun E, Ruda R, Tonn JC, Weber
systematic review. JNCI Cancer Spectr. 2018;2(4):pky062. https:// DC, Weller M. Diagnosis and treatment of brain metastases from
doi.org/10.1093/jncics/pky062. solid tumors: guidelines from the European Association of Neuro-­
67. Maaskant-Braat AJ, Roumen RM, Voogd AC, Pijpers R, Luiten Oncology (EANO). Neuro Oncol. 2017;19(2):162–74. https://doi.
EJ, Rutgers EJ, Nieuwenhuijzen GA. Sentinel node and recur- org/10.1093/neuonc/now241.
rent Breast cancer (SNARB): results of a nationwide registration 80. Yousefi M, Nosrati R, Salmaninejad A, Dehghani S, Shahryari
study. Ann Surg Oncol. 2013;20(2):620–6. https://doi.org/10.1245/ A, Saberi A. Organ-specific metastasis of breast cancer: molecu-
s10434-­012-­2625-­7. lar and cellular mechanisms underlying lung metastasis. Cell
68. Moossdorff M, Vugts G, Maaskant-Braat AJ, Strobbe LJ, Voogd AC, Oncol (Dordr). 2018;41(2):123–40. https://doi.org/10.1007/
Smidt ML, Nieuwenhuijzen GA. Contralateral lymph node recur- s13402-­018-­0376-­6.
rence in breast cancer: regional event rather than distant metastatic 81. Rashid OM, Takabe K. The evolution of the role of sur-
disease. A systematic review of the literature. Eur J Surg Oncol. gery in the management of breast cancer lung metastasis.
2015;41(9):1128–36. https://doi.org/10.1016/j.ejso.2015.05.015. J Thorac Dis. 2012;4(4):420–4. https://doi.org/10.3978/j.
69. Leone BA, Vallejo CT, Romero AO, Machiavelli MR, Perez JE, issn.2072-­1439.2012.07.16.
Leone J, Leone JP. Prognostic impact of metastatic pattern in stage 82. Nichols FC. Pulmonary metastasectomy. Thorac Surg Clin.
IV breast cancer at initial diagnosis. Breast Cancer Res Treat. 2012;22(1):91–9, vii. https://doi.org/10.1016/j.thorsurg.2011.08.017.
2017;161(3):537–48. https://doi.org/10.1007/s10549-­016-­4066-­7. 83. Meimarakis G, Ruttinger D, Stemmler J, Crispin A, Weidenhagen
70. Al-Shafa F, Arifin AJ, Rodrigues GB, Palma DA, Louie AV. A R, Angele M, Fertmann J, Hatz RA, Winter H. Prolonged over-
review of ongoing trials of stereotactic ablative radiotherapy for all survival after pulmonary metastasectomy in patients with
406 A. M. Terando et al.

breast cancer. Ann Thorac Surg. 2013;95(4):1170–80. https://doi. Ann Surg Oncol. 2013;20(6):1893–9. https://doi.org/10.1245/
org/10.1016/j.athoracsur.2012.11.043. s10434-­012-­2844-­y.
84. Yoshimoto M, Tada K, Nishimura S, Makita M, Iwase T, Kasumi 91. Cady B, Nathan NR, Michaelson JS, Golshan M, Smith BL. Matched
F, Okumura S, Sato Y, Nakagawa K. Favourable long-term results pair analyses of stage IV breast cancer with or without resection of
after surgical removal of lung metastases of breast cancer. Breast primary breast site. Ann Surg Oncol. 2008;15(12):3384–95. https://
Cancer Res Treat. 2008;110(3):485–91. https://doi.org/10.1007/ doi.org/10.1245/s10434-­008-­0085-­x.
s10549-­007-­9747-­9. 92. Dominici L, Najita J, Hughes M, Niland J, Marcom P, Wong YN,
85. Caralt M, Bilbao I, Cortes J, Escartin A, Lazaro JL, Dopazo C, Carter B, Javid S, Edge S, Burstein H, Golshan M. Surgery of the
Olsina JJ, Balsells J, Charco R. Hepatic resection for liver metas- primary tumor does not improve survival in stage IV breast can-
tases as part of the “oncosurgical” treatment of metastatic breast cer. Breast Cancer Res Treat. 2011;129(2):459–65. https://doi.
cancer. Ann Surg Oncol. 2008;15(10):2804–10. https://doi. org/10.1007/s10549-­011-­1648-­2.
org/10.1245/s10434-­008-­0072-­2. 93. Badwe RA, Kataria K, Srivastava A. Surgical resection of Phyllodes
86. Chua TC, Saxena A, Liauw W, Chu F, Morris DL. Hepatic tumour: a radical approach as a safeguard against local recur-
resection for metastatic breast cancer: a systematic review. rence. Indian J Surg. 2015;77(2):161–3. https://doi.org/10.1007/
Eur J Cancer. 2011;47(15):2282–90. https://doi.org/10.1016/j. s12262-­013-­0935-­z.
ejca.2011.06.024. 94. Soran A, Ozmen V, Ozbas S, Karanlik H, Muslumanoglu M, Igci
87. Spolverato G, Vitale A, Bagante F, Connolly R, Pawlik TM. Liver A, Canturk Z, Utkan Z, Ozaslan C, Evrensel T, Uras C, Aksaz E,
resection for breast cancer liver metastases: a cost-utility analy- Soyder A, Ugurlu U, Col C, Cabioglu N, Bozkurt B, Uzunkoy A,
sis. Ann Surg. 2017;265(4):792–9. https://doi.org/10.1097/ Koksal N, Gulluoglu BM, Unal B, Atalay C, Yildirim E, Erdem
SLA.0000000000001715. E, Salimoglu S, Sezer A, Koyuncu A, Gurleyik G, Alagol H,
88. Khan SA, Stewart AK, Morrow M. Does aggressive local ther- Ulufi N, Berberoglu U, Dulger M, Cengiz O, Sezgin E, Johnson
apy improve survival in metastatic breast cancer? Surgery. R. Randomized trial comparing resection of primary tumor with no
2002;132(4):620–6; discussion 626–627. https://doi.org/10.1067/ surgery in stage IV Breast cancer at presentation: protocol MF07-­
msy.2002.127544. 01. Ann Surg Oncol. 2018;25(11):3141–9. https://doi.org/10.1245/
89. Gnerlich J, Jeffe DB, Deshpande AD, Beers C, Zander C, s10434-­018-­6494-­6.
Margenthaler JA. Surgical removal of the primary tumor increases 95. Khan SA, Zhao F, Solin LJ, Goldstein LJ, Cella D, Basik M, Golshan
overall survival in patients with metastatic breast cancer: analysis M, Julian TB, Pockaj BA, Lee CA, Razaq W, Sparano JA, Babiera
of the 1988-2003 SEER data. Ann Surg Oncol. 2007;14(8):2187– GV, Dy IA, Jain S, Silverman P, Fisher C, Tevaarwerk AJ, Wagner
94. https://doi.org/10.1245/s10434-­007-­9438-­0. LI, Sledge GW. A randomized phase III trial of systemic therapy
90. Lang JE, Tereffe W, Mitchell MP, Rao R, Feng L, Meric-Bernstam plus early local therapy versus systemic therapy alone in women
F, Bedrosian I, Kuerer HM, Hunt KK, Hortobagyi GN, Babiera with de novo stage IV breast cancer: a trial of the ECOG-ACRIN
GV. Primary tumor extirpation in breast cancer patients who pres- Research Group (E2108). J Clin Oncol. 2020;38(18_suppl):LBA2.
ent with stage IV disease is associated with improved survival. https://doi.org/10.1200/JCO.2020.38.18_suppl.LBA2.
 The Role of Surgery in Managing
Primary and Metastatic Colorectal 38
Cancer

Richard J. Straker III, Hunter D. D. Witmer,

Benjamin Deschner, David Shibata, Kiran K. Turaga,
and Najjia N. Mahmoud

Abstract
Learning Objectives
The surgical management of metastatic colorectal cancer • To understand the current literature describing the
(CRC) has evolved dramatically over the last few decades. importance of lymph node harvest and its associa-
Proper lymph node evaluation is essential, as not only does tion with survival in patients with colorectal
detection of regional lymph node metastases carry impor- cancer.
tant prognostic implications, but it also dictates the need for • To understand the different lymph node dissection
adjuvant therapy following surgical resection. Attainment techniques currently practiced for patients with
of an adequate number of lymph nodes is necessary for colorectal cancer and describe the advantages and
complete staging and is used as a quality care metric, but disadvantages of a more versus less extensive
debate continues to exist about the clinical significance of dissection.
harvesting greater or fewer nodes than this minimum • To understand the advantages and disadvantages of
required amount. Data have shown that with careful patient synchronous versus staged resection of colorectal
selection, resection of hepatic CRC metastases can provide cancer hepatic metastases.
a survival benefit, but the ideal timing and sequence of • To understand that current data on cytoreductive
resection remain unclear. The use of cytoreductive surgery, surgery/heated intraperitoneal chemotherapy for
with or without hyperthermic intraperitoneal chemother- colorectal cancer peritoneal metastases.
apy, for the management of peritoneal metastases continues • To understand the potential role of advanced molec-
to be an area of active ongoing research, as well. As we ular and genetic analyses in the management of
enter the era of precision medicine, growing evidence sup- metastatic colorectal cancer.
ports the prognostic implications provided by genetic anal-
ysis of tumor specimens, and emerging molecular
techniques may offer more accurate identification of tumor
metastases within harvested lymph nodes. 38.1 Introduction

Modern paradigms of tumor metastases continue to evolve as

Richard J. Straker and Hunter D. D. Witmer contributed equally to the our biological understanding of cancer expands. Newer theo-
manuscript and share first authorship. ries of metastases have emerged out of advances in biotech-
Kiran K. Turaga and Najjia N. Mahmoud contributed equally to the nology and analytics. The diaspora theory [1] consists of
manuscript and share senior authorship. three distinct primary considerations: (a) the suitability of

R. J. Straker III (*) · N. N. Mahmoud B. Deschner

Division of Colon and Rectal Surgery, Department of Surgery, Department of Surgery, Mountain Area Health Education Center,
University of Pennsylvania, Philadelphia, PA, USA Asheville, NC, USA
e-mail: [email protected];
D. Shibata
[email protected]
Department of Surgery, University of Tennessee Health Science
H. D. D. Witmer · K. K. Turaga Center, Memphis, TN, USA
Division of Surgical Oncology, Department of Surgery, University e-mail: [email protected]
of Chicago, Chicago, IL, USA
e-mail: [email protected];
[email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 407
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_38
408 R. J. Straker III et al.

the primary tumor’s microenvironment, (b) the adaptability 38.2  he Significance of Lymph Node
T
and proficiency of various tumor clonal populations to Harvest in Colorectal Cancer
migrate to other sites, and (c) the suitability of the microen-
vironment at remote, potential metastatic sites (Fig. 38.1). For patients with CRC without distant metastases, histopath-
These variables help to explain some of the driving forces ologic lymph node status represents the most important prog-
underpinning the mechanisms of metastatic spread, such as nostic factor for survival [4]. Estimated 5-year survival
the relative degree of selective pressure for epithelial to mes- decreases from 70 to 90% for those without lymph node
enchymal transition and the inherent propensity for malig- involvement to 30–60% for those with spread to regional
nant cells to shed into the lymphovascular system. The nodes, and survival rates further vary based on the number of
diaspora theory also provides a conceptual framework for nodes involved [5]. Standard of care for patients with lymph
the proclivity of disseminated tumor cell subpopulations for node-positive disease entails adjuvant chemotherapy with a
certain tissues irrespective of patterns of anatomic drainage fluoropyrimidine-based regimen with or without oxaliplatin,
or circulation and for variations in the subsequent aggres- which has been shown to both reduce disease recurrence and
siveness at these sites. The related theory of oligometastases improve overall survival (OS) [6, 7]. To ensure adequate stag-
holds that there are intermediate steps between initial tumor ing among node-negative patients, current guidelines recom-
cell dispersion and fulminant metastatic disease. Moreover, mend a minimum of 12 lymph nodes be evaluated in both
metastatic potential is limited to certain sites, and, by impli- colon and rectal cancer [8]. In the absence of adequate lymph
cation, is potentially curable through localized treatment [2, node sampling, positive lymph nodes may not be detected,
3]. Given the substantial incidence of metastatic colorectal thus leading to patients potentially being under-­staged [9].
cancer (CRC), we outline some of the nuances of surgical Given the clinically important correlations between lymph
management of patients with regional and distant metastases node involvement and outcomes, numerous studies have eval-
in the context of these paradigms. uated the prognostic implications of the number of nodes har-
vested, with many studies finding a positive association
between higher nodal retrieval and improved survival [10].
Specifically among patients with stage II colon cancer, Chen
et al. demonstrated that patients with 15 or more lymph nodes
evaluated had better survival than those in which fewer nodes
were assessed, and similar results have been seen with rectal
cancer, as well [11, 12]. While this would lead one to believe
that the standard should shift toward harvesting more than the
recommended 12 nodes, significant debate exists regarding
whether these aforementioned studies are capturing the true
reasons behind the association between node number and sur-
vival. Higher numbers of harvested nodes likely lead to a
Cancer cells
stage migration effect, resulting in the appearance of improved
Host cells survival among stage I/II patients as more stage III patients
are placed into the correct staging category [13].
Soluble factors However, critics of this, “Will Rogers phenomenon,” the-
(e.g. exosomes, ory argues that more extensive lymph node yields don’t nec-
cytokines, hormones) essarily correlate to higher rates of node positivity, suggesting
that other confounding factors may be present [14]. Baxter
et al. showed that among patients with pathologic T3 colon
cancers, staging is improved with higher nodal yields when
the nodal count is low (<5–7 nodes), but that this effect is
mitigated at higher nodal counts [15]. An alternative theory
gaining much traction postulates that host immune responses
could be the dominant variable responsible for the associa-
tion between lymph node count and survival, discussed in
Fig. 38.1 Illustrative representation of the many complex interactions more detail subsequently [13]. Additionally, as discussed in
of cancer progression under the diaspora theory. The microenvironment the following sections, surgical excision is just one of the
at both the primary tumor site as well as in other tissues, and the inter-
play between them, may determine where metastases develop and their
many factors that ultimately dictates the number of lymph
behavior at these sites nodes evaluated in a given resection specimen. As such,
38 The Role of Surgery in Managing Primary and Metastatic Colorectal Cancer 409

these uncertainties surrounding the association between the identification of lymph nodes and is highly operator
lymph node count and survival have led some to propose that dependent. Studies have shown that marked variation exists
the lymph node ratio, calculated as the ratio of metastatic across institutions in the methods utilized and time spent dis-
lymph nodes to total lymph nodes harvested, may be a more secting out lymph nodes, with one study demonstrating that
reliable prognostic factor than the absolute number of lymph an additional 20 min of time spent searching led to the dis-
nodes retrieved [16]. covery of a mean of six additional lymph nodes [22, 23].
Finally, the retrieval of at least 12 nodes for pathologic Much data are emerging linking the host immune system
evaluation has emerged as a metric of quality care and a sur- with lymph node yield and outcomes in CRC. Enhanced
rogate marker for an adequate surgical resection in colon can- immunologic reactions against the tumor can result in lymph
cer [17]. Much enthusiasm exists for its use as a quality node hyperplasia and enlargement allowing for easier detec-
indicator due to its relative ease of measurement and compa- tion, and lymph node hyperplasia has been shown to be associ-
rability across institutions [18]. Early studies showed that a ated with improved survival, as well [13]. Larger tumors may
majority of institutions failed to achieve compliance with this induce more profound immune responses in the mesentery,
12-node metric, but more recent data demonstrate improve- leading to a higher absolute number of lymph nodes being
ments in the proportion of hospitals meeting this benchmark identified for pathologic evaluation from the resected mesen-
[4, 14, 19]. However, improvements in clinical outcomes teric specimen. The more advanced bowel wall penetration
have not reliably paralleled this increased compliance, and exhibited by higher T-stage tumors may also induce greater
many investigators have outlined the numerous limitations immune responses, again leading to increased numbers of
associated with using lymph node harvest in this capacity, nodes retrieved following resection. Studies have also shown a
including the fact that surgical resection is only one of many positive correlation between the presence of tumor-­infiltrating
known factors associated with lymph node yield [18]. As we lymphocytes and lymph node retrieval, and likewise improved
continue to parse out the true associations between lymph survival when increasing numbers of these immunologic cells
node yield and survival, the role of this metric in the assess- are present. Finally, the established association between mic-
ment of quality surgical care may come to change. rosatellite instability and improved survival is thought, at least
in part, to be due to immunologic factors [13]. Based on this
evidence, the immune system appears to play an increasingly
38.2.1 N
 onsurgical Factors Affecting Lymph prevalent role in both the absolute number of lymph nodes
Node Evaluation harvested and the association between lymph node yield and
survival in colorectal cancer.
While surgical excision plays an obvious role in the quantity A number of patient-specific factors have been shown to
of lymph nodes available for analysis, the histopathologic influence lymph node harvest. Numerous reports have iden-
methods by which lymph nodes are defined and examined, as tified age as an important factor affecting lymph node yield,
well as immunologic, patient, and tumor-specific character- with increasing age consistently being associated with lower
istics also influence the number of lymph nodes that are ulti- recovery counts [24, 25]. This may be related to decreased
mately available for evaluation. surgical aggressiveness in older patients, or alternatively
Significant alterations to the pathologic classification of a may be due to a diminished immune response with older age,
lymph node metastasis have occurred over successive edi- leading to smaller and less easily detectable lymph nodes
tions of the TNM staging system. Although minimal changes [13]. The association between gender and body mass index
were made from the previously used seventh edition to the (BMI) on lymph node recovery is less clear, with some stud-
currently used eighth edition, the seventh edition introduced ies finding that females and lower BMI patients have higher
the N1c stage as a separate category for tumor deposits lymph node harvests, while others have shown higher BMI
­present in the subserosa, mesentery, or nonperitonealized leads to lower yields [26, 27].
pericoloc or perirectal/mesorectal tissues without concomi- Several tumor-specific factors have been associated with
tant lymph node involvement [5, 20]. Prior editions defined lymph node recovery, as well, including tumor location,
tumor deposits differently, and many of these deposits were T-stage, and lymph node size. Overwhelming evidence has
classified as regional lymph node metastases. However, the demonstrated that right-sided colon cancers yield signifi-
seventh edition no longer identified these deposits as lymph cantly higher lymph node counts compared to left-sided
nodes with tumor involvement, resulting in a significant rise tumors. Many explanations for this observation exist, includ-
in the number of cases with less than 12 lymph nodes detect- ing anatomic differences between the right and left colon, as
able for evaluation, and thus increasing the proportion of well as a higher rate of microsatellite instability among right-­
patients not meeting the minimum lymph node requirement sided cancers, a factor that has also been shown to correlate
for adequate staging [21]. Thorough and complete dissection with higher lymph node counts [24, 25, 27–29]. As previ-
of the surgical specimen by the pathologist is also critical for ously mentioned, higher T-stage tumors and larger lymph
410 R. J. Straker III et al.

node size are also consistently found to be associated with nodes, which are involved in an estimated 3–11% of tumors,
greater lymph node yields, possibly due to more robust and may also lead to the removal of skip metastases (node-­
immunologic responses against more advanced tumors, lead- positive main nodes in the absence of positive intermediate
ing to easier detection [24, 25, 27, 28, 30]. nodes), both of which would have otherwise been left behind
had an extended lymphadenectomy been deferred [35].
However, despite the theoretical benefits of these extended
38.2.2 Extent of Lymph Node Dissection lymphadenectomies, data have not indisputably demonstrated
improved oncologic outcomes following a more extensive
In addition to removal of the malignant lesion with appropri- dissection. Clinical trials have failed to show a survival
ate bowel margins, surgical resection also includes mesen- advantage even with high ligation of the inferior mesenteric
teric excision which provides the lymph nodes for staging. artery (IMA) in patients with left-sided colon or rectal cancer,
Much debate exists regarding whether there is a direct thera- and lateral pelvic nodal dissection in these patients has not
peutic benefit from this lymph node dissection, what the been shown to improve survival, either [37, 38]. Moreover,
optimal extent of this excision is, and whether outcomes are extended lymphadenectomy techniques are associated with
significantly improved when more extended dissections are increased morbidity, with higher rates of direct injury to
employed. neighboring structures, including the duodenum, ureters,
Indirect evidence for a therapeutic benefit to lymphade- spleen, and nerve plexuses in some reports [38, 39]. Concern
nectomy may be derived from the Intergroup Trial INT-0089. also continues to exist over ischemia-related anastomotic
While the authors of this trial found that higher lymph node complications due to decreased blood flow following high
counts in general were associated with improved survival, ligation, although the evidence surrounding this specific topic
this improvement was also seen specifically among subgroup is controversial [40]. Given the unproven benefit coupled
analyses of patients with node-positive disease, in which with the potential for increased morbidity, current evidence is
stage N2 patients with >35 nodes removed had a 5-year sur- insufficient to recommend extended lymphadenectomy, i.e.,
vival of 71% compared to a 51% survival for N2 patients D3 dissection, as the standard of care [32].
with ≤35 nodes [31]. Given that all patients within this anal-
ysis had N2 disease, stage migration, and more accurate
staging incompletely explain these results, suggesting that 38.2.3 P
 recision Medicine and Lymph Node
there may have been a therapeutic component to the higher Harvest
lymph node yield. However, the data remain conflicting,
with other reports finding incomplete evidence to support a As our comprehension of cancer biology continues to evolve,
direct therapeutic benefit [32]. we are gaining more insight into the importance of molecular
Whether a therapeutic benefit is derived from a more pathogenesis, not just in CRC, but within the entire disci-
extensive lymphadenectomy directly relates to the debate on pline of oncology. Genetic analyses and precision medicine
what is the optimal extent of the mesenteric dissection is. are revolutionizing the modern treatment of cancer, and it is
Vascular ligation combined with lymphadenectomy up to the likely that these techniques will have a pivotal role in the
root of the main feeding vessel is defined as a D2 resection harvest and evaluation of lymph nodes in CRC, too.
and results in the removal of the pericolic and intermediate Oncogenic mutations in KRAS and TP53 have been shown to
level nodes, leaving behind the main/apical level nodes that correlate with higher rates of regional lymph node metasta-
are present along the length of the main feeding vessel. The sis, while mutations in APC and PIK3CA have not. Similarly,
more extended D3 resection, commonly practiced in Japan, certain transcription factors and copy number variations have
includes removal of these main level nodes, requiring dissec- been demonstrated to confer an increased risk for nodal dis-
tion along the length of the feeding vessels in order to excise ease [41]. Zhang et al. correlated the level of lymph node
the lymphoadipose tissue they reside in [33]. Similar to the metastases with underlying genetic mutations, allowing for
D3 resection, Hohenberger et al. described the complete the discovery of mutations which confer a higher risk for
mesocolic excision (CME) with central vascular ligation development of skip metastases [42]. Identification of these
(CVL), although this technique has been shown to have sig- molecular variants may allow for classification of patients as
nificantly longer lateral resection margins than are seen with high or low risk for lymph node involvement and allow for a
the D3 method [34]. Proponents of these extended dissec- personalized approach to deciding who will benefit from a
tions suggest that the higher nodal yields achieved are asso- higher nodal harvest or more extensive lymphadenectomy.
ciated with improved survival as evidenced by studies Advanced molecular techniques have also been applied to
demonstrating low recurrence and improved survival rates the evaluation of harvested lymph nodes and can more pre-
with employment of these techniques [35, 36]. Furthermore, cisely identify the presence of cancer within the lymph node,
these central dissections allow for removal of the apical as well as quantify the amount of tumor present [43]. Further
38 The Role of Surgery in Managing Primary and Metastatic Colorectal Cancer 411

research is needed to determine whether these advanced A 2013 consensus statement from the Americas Hepato-­
analyses will become mainstay in the management of Pancreato-­Biliary Association (AHPBA), Society of Surgical
colorectal cancer. Oncology (SSO), and Society for Surgery of the Alimentary
Tract (SSAT) further specified that patients with limited and
resectable EHD or a reasonable expectation for long-term
38.3 Hepatic Metastases control with adjuvant treatment could be considered for
hepatic resection [45].
38.3.1 C
 riteria for Surgical Resection of Liver In terms of prognostication, Fong et al. [48] established a
Metastases clinically relevant scoring system for predicting survival in
patients undergoing resection of CRC liver metastases in
Current category 2A recommendations from the National which one point is assigned for each of the following risk
Comprehensive Cancer Network (NCCN) [6] state resection factors: lymph node-positive primary tumor, preoperative
as the treatment of choice for liver metastases from CRC carcinoembryonic antigen (CEA) level > 200, disease-free
when (a) complete resection is feasible, (b) adequate postop- interval of <12 months, >1 metastasis, and largest liver
erative liver function is expected, (c) the primary tumor has metastasis >5 cm. In the authors’ validation series of 1001
undergone an R0 resection, and (d) no unresectable extrahe- patients, median survival in months was 74, 51, 47, 33, 20,
patic disease (EHD) is present (Table 38.1). Other guidelines and 22 for scores of 0 through 5, respectively.
recommend synchronous resection when the hepatic phase
of the operation is expected to be less extensive (i.e., when
liver metastases are small, peripheral, and/or the expected 38.3.2 S
 ynchronous, Staged and Liver-First
hepatic specimen is <50% of liver volume) and that patients Resections
with peritoneal metastases (PM) can also be considered for
curative surgical therapy [44]. Prospective randomized data There has been substantial interest in the optimal timing of
and meta-analyses also show laparoscopic resection to be hepatic metastasis resection given the number of patients
safe and oncologically sound, and without any differences in who present with synchronous primary and isolated hepatic
disease-free survival (DFS) or OS [46, 47]. metastases and the risk-benefit assessment in the context of

Table 38.1 Society and national guidelines for resection of hepatic CRC metastases
NCCN guidelines [6] Chinese guidelines [44] AHPBA/SSO/SSAT guidelines [45]
Technical – Feasibility of complete resection – R0 resection expected – R0 resection expected
considerations – Maintenance of adequate hepatic – Adequate FLR volume: – Ability to preserve two contiguous
function  • >40% for synchronous resection segments
– Consider PVE or staged resection  • >30% for staged resection – Adequate vascular inflow/outflow
when insufficient FLR is expected – Preservation of at least one hepatic & biliary drainage
vein – Adequate FLR volume (20–30%)
– Assessment of patient intolerance – Appropriate regenerative response
after PVE in patients with
marginal FLR
Considerations for – Prior R0 resection of primary tumor – Prior, or expected, R0 resection of – Resectable EHD or reasonable
EHD – No unresectable EHD primary tumor expectation of long-term control
– No unresectable EHD or not with adjuvant therapy
amenable local ablative treatment – Defer resection in the setting of
metastatic disease progression
while on first-line systemic
therapy
Synchronous versus Synchronous or staged depending on Recommend synchronous resection Not mentioned
staged resection complexity, comorbidities, surgical when:
exposure, and surgeon expertise – Non-emergent
– Small and/or peripherally located
hepatic metastases or confined to
single liver lobe
– Expected hepatectomy volume
< 50%
– Suspicious portal lymph nodes,
intra-abdominal or other distant
metastases can be completely
resected
EHD Extrahepatic disease, FLR Future liver remnant, PVE Portal vein embolization
412 R. J. Straker III et al.

multimodal treatment options. While a staged approach may 38.3.4 A

 ssociated Liver Partition and Portal
reduce surgical morbidity and avert mechanical complica- Vein Ligation for Staged Hepatectomy
tions associated with primary tumors, it may also prolong
the time from diagnosis to initiation of systemic chemo- A newer option for patients with marginal or insufficient pro-
therapy [49]. Much of the literature is retrospective, yet sup- jected postoperative future liver remnant (FLR) has emerged
ports a synchronous over a staged approach despite a higher and is dubbed associated liver partition and portal vein liga-
associated morbidity and mortality, especially with more tion for staged hepatectomy (ALPPS). This approach is
substantial hepatectomies [50–53]. Recent prospective data divided into two stages. The initial operation involves mobi-
suggest that OS improves with synchronous resection com- lization of the liver and ligation of the portal vein branch
pared to a delayed approach without an increase in complica- supplying the diseased side at the bifurcation without dis-
tions [54]. There is also an ongoing clinical trial (RESECT, rupting the arterial or biliary systems. The liver parenchyma
NCT02954913) seeking to further delineate the complica- is subsequently divided along the intended resection plane
tion profile associated with synchronous resection. (i.e., the in situ splitting phase) and the de-portalized seg-
Another sequencing option, known as the liver-first (or ments of liver are then left in place and the abdomen closed.
reverse) approach, has also been accepted as a valid treat- The second stage, or completion of resection, occurs after a
ment strategy in which patients receive systemic chemother- delay of 1–2 weeks allowing for compensatory hypertrophy
apy and undergo resection of liver metastases first, while the of the FLR [59]. Parenchymal division in the first stage is
primary tumor is removed in a delayed fashion [55]. The thought to promote more rapid compensatory hypertrophy
predominant advantages of this approach are performing by preventing any collateral portal blood flow to the de-­
metastasectomy prior to progression that may preclude portalized side [60], an effect not possible through percuta-
resectability and a reduction in time to initiating systemic neous approaches to portal exclusion. Although this approach
chemotherapy; however, local complications related to the can facilitate curative resection of hepatic metastases in
primary tumor can still occur in the surgical interval, albeit patients with borderline predicted FLR, it is associated with
infrequently [49]. substantial morbidity and mortality and has not been proven
to be superior to staged resections in retrospective analysis
[61–64].
38.3.3 Sequential Hepatic Resections

The surgical community has also sought innovative 38.3.5 Administration of Chemotherapy
approaches to increasing the number of patients eligible for Before and After Liver Resection
resection, and therefore cure, without untenable risks of
complications such as postoperative liver failure. One such The timing of chemotherapy with respect to hepatic metasta-
technique is the two-stage hepatectomy, classically for sectomy is also a germane consideration given that periop-
patients with bi-lobar metastases in whom a single step erative chemotherapy has been shown to improve 3-year
resection would be impossible. In the first operation, dis- progression-free survival (PFS) in patients with resectable
ease in one hemi-liver is resected while the remaining liver hepatic metastases undergoing surgery, but is also associated
parenchyma is left to undergo compensatory hypertrophy. with higher rates of postoperative complications [65].
The remaining metastatic disease is then resected in the Unfortunately, the survival benefit in this population does
second stage. This can be done alone, with simultaneous not appear to be durable [66]. Several agents used in standard
resection of the primary, and/or in conjunction with abla- of care chemotherapeutic regimens are associated with spe-
tive or embolization techniques [56–58]. Although the mor- cific hepatic complications: the use of irinotecan is associ-
bidity associated with two major hepatic resections is high, ated with higher rates of steatohepatitis and postoperative
the majority of patients undergoing two-stage hepatectomy mortality, whereas oxaliplatin has been linked to vascular
reported in the literature ultimately underwent the second lesion development in non-tumor-bearing liver specimens
stage and had R0 resections [56]. Five-year OS is reported and more significant intraoperative transfusion requirements,
to be an encouraging 32–64% considering that these but no difference in postoperative mortality [67–69]. The use
patients traditionally would have been deemed to have of neoadjuvant chemotherapy also necessitates a larger FLR
unresectable stage IV disease and only offered palliative volume in order to mitigate the risks of postoperative liver
chemotherapy [56]. failure due to impaired liver function [70, 71].
38 The Role of Surgery in Managing Primary and Metastatic Colorectal Cancer 413

Evaluation of the clinical response to neoadjuvant chemo- evidence of metastatic disease, short course RT is attractive
therapy may also aid in the identification of patients who will in reducing this treatment interval without foregoing the ben-
benefit from subsequent metastasectomy. Those who demon- efits of improved local control. Any risks or benefits in delay-
strate disease progression while on upfront chemotherapy are ing liver-directed therapies in favor of long-course RT remain
less likely to benefit from resection, and the decision for sur- to be delineated, and individualized treatment decisions
gical intervention can be reconsidered, whereas those with should be made with a multidisciplinary tumor board
stable disease or who respond well to chemotherapy can then discussion.
still undergo their planned surgical resection [72, 73].
However, even in patients who appear to have a complete
radiographic response to neoadjuvant treatment, the majority 38.4 Peritoneal Metastases
are still likely to have residual disease, and therefore, com-
plete clinical response should not be equated with a cure [74]. The estimated incidence of CRC PM is 8.3–26.6% and var-
Neoadjuvant therapy may also be administered through a ies in the literature depending on study type and inclusion
regional approach via hepatic arterial infusion chemotherapy criteria [80]. Autopsy series show that in patients who die
[75]. Direct administration of chemotherapy to the liver from CRC cancer, 20% with adenocarcinoma, 48% with
allows delivery and exposure of greater drug concentrations mucinous subtypes, and 51% with signet ring cell tumors
directly to the tumor over an extended period of time while ultimately develop PM, while nearly 15% have isolated PM
limiting systemic side effects through implantation of a regardless of histology [80, 81]. Cytoreductive surgery
hepatic artery infusion pump. High objective radiologic (CRS), with or without hyperthermic intraperitoneal chemo-
response rates have been observed with this method of therapy (HIPEC), therefore continues to be an attractive
administration, but a definitive survival advantage has not treatment modality in the setting of synchronous and meta-
been demonstrated [75]. Hepatic toxicity and technical com- chronous PM and is an area of ongoing research [82].
plications related to the pump or accompanying ports/cathe-
ters limit this treatment approach, and its use is mostly
confined to specialized centers with extensive experience in 38.4.1 Clinical Evidence
this treatment modality [76].
The highest quality evidence for CRS/HIPEC in the setting of
PM from CRC comes from a randomized clinical trial of
38.3.6 T
 iming of Radiation Therapy for Rectal patients with colorectal adenocarcinoma without solid organ
Cancer with Metastases metastases who were randomized to either systemic chemo-
therapy alone or CRS/HIPEC with mitomycin C and adjuvant
Neoadjuvant radiation therapy (RT) currently plays a signifi- systemic chemotherapy [83]. Both treatment arms received
cant role in treatment algorithms for locally advanced rectal systemic fluorouracil and leucovorin alone, which is no lon-
cancer in order to improve long-term local control; however, ger current first-line therapy [84]. The addition of CRS/
appropriate sequencing of treatment is less clear in the set- HIPEC improved both PFS (12.6 vs. 7.7 months, p = 0.020)
ting of synchronous liver metastases [77]. NCCN guidelines and disease-specific survival (DSS) (22.2 vs. 12.6 months,
for nonmetastatic rectal cancer advocate the use of long p = 0.028) after long-term follow-up, and represents a sub-
course RT (45–50 Gy in 25–28 fractions) in conjunction with stantial improvement compared to the expected survival with
radio-sensitizing fluoropyrimidine-based chemotherapy fol- chemotherapy alone [85–87]. Furthermore, patients who
lowed by resection 4–6 weeks after completion [7]. These underwent an R0–1 resection had a median survival of
guidelines also stipulate that patients with T3 tumors can be 48 months and a 5-year OS of 45%, demonstrating a durable
considered for a short course regimen (25 Gy in 5 fractions) survival benefit. Notably, this study did not parse out the indi-
followed by surgery within 3–7 days, or after 4–6 weeks, vidual contributions of HIPEC versus CRS (which was evalu-
based on current evidence [78]. The concern with long ated in the PRODIGE 7 trial, discussed subsequently), and
course RT directed at the primary tumor in the setting of also did not include combinations of chemotherapy regimens
hepatic metastases is the prolonged delay in treatment of the currently used today. Therefore, it is plausible that the sys-
metastatic disease, which may no longer be resectable within temic therapy alone arm would have significantly better out-
the recommended surgical window for the primary tumor comes with modern chemotherapeutics.
(4–6 weeks after completing RT). Short course RT has shown More recently, preliminary data from the unpublished
equivalency in terms of 5-year local recurrence, OS, and late PRODIGE 7 [88] trial (NCT00769405) presented in 2018
toxicity for patients with T3 tumors within 12 cm of the anal included 265 patients with synchronous PM from CRC and a
verge, and T3/T4 tumors without sphincter involvement [78, peritoneal cancer index ≤25 who were randomized to either
79]. Although neither of these studies involved patients with CRS/HIPEC with oxaliplatin or CRS alone. Both groups
414 R. J. Straker III et al.

received standard neoadjuvant and adjuvant systemic che- 38.4.3 Ongoing Clinical Trials
motherapy for a total of 6 months. Interestingly, the addition
of oxaliplatin-HIPEC to CRS did not prolong OS (41.2 vs. There are currently several ongoing trials seeking to further
41.7 months, p = 0.995) or relapse-free survival (11.1 vs. assess the role and approach to CRS/HIPEC in patients with
13.1 months, p = 0.486), which implies that the comparative CRC. The ICARuS trial (NCT01815359) is a multicenter,
survival advantage for this population over systemic chemo- phase II study comparing CRS/HIPEC with mitomycin C to
therapy alone may derive primarily from the resection of CRS with early postoperative intraperitoneal chemotherapy
their metastatic disease. (EPIC) with 5-fluorodeoxyuridine and leucovorin in the
treatment of PM for CRC and appendiceal cancers. Two
other ongoing phase III trials in Spain (PROMENADE,
38.4.2 A
 djuvant HIPEC and Second Look NCT02974556) and China (APEC, NCT02965248) are also
Procedures investigating the role of HIPEC at the time of primary resec-
tion despite the aforementioned evidence against prophylac-
Two recent clinical trials investigated the role of adjuvant tic HIPEC in high-risk patients. Additionally, CAIRO6
HIPEC in patients without synchronous PM but at high risk (NCT02758951) will assess the oncologic efficacy of CRS/
for subsequent metachronous disease due to perforated or HIPEC alone versus in addition to standard systemic chemo-
T4 primary CRC tumors. In the COLOPEC trial [89] therapeutic regimens for patients with PM from non-­
(NCT02231086), patients underwent resection of the pri- appendiceal CRC.
mary tumor and received either HIPEC with oxaliplatin
(460 mg/m2) and routine adjuvant systemic chemotherapy
or the standard of care with adjuvant systemic chemother- 38.5  vidence for Hepatic Resection
E
apy alone. Ultimately, no differences in 18-month PM-free in the Setting of Extrahepatic Disease
survival, DFS, or OS were observed between groups.
Criticisms of this trial include that only 87% of patients in Despite the traditional paradigm regarding the futility of
the experimental arm actually underwent HIPEC, some of resection in the setting of multisite metastatic disease, the
which were performed concurrent with their initial resec- benefits of performing hepatic resection in this setting are
tion, whereas others were delayed 5–8 weeks, and that increasingly supported by the literature [91]. In patients
patients in the control arm were eligible for cross over if undergoing resection of EHD, OS appears to depend on the
they developed PM during surveillance. Ultimately, the extrahepatic metastatic site with longer median survival
results of this trial do not support the adjuvant use of HIPEC associated with metastasis to the lungs compared to portal or
in patients with high-risk CRC primary tumors without syn- para-aortic lymph nodes or the peritoneum [92]. This obser-
chronous PM. vation challenges the paradigm of surgical futility in the
The other trial (PROPHYLOCHIP [90], NTC01226394) treatment of oligometastatic disease. Only one prospective
included patients with perforated primary tumors and randomized trial comparing pulmonary metastasectomy to
those with isolated PM resected with the primary in order active surveillance has been published and was terminated
to assess the efficacy of routine delayed CRS/HIPEC in early due to poor accrual, thereby precluding significant
patients at high risk for metachronous peritoneal disease. interpretation [93]. Some retrospective evidence shows that
All patients received adjuvant first-line systemic chemo- long-term survival for patients undergoing pulmonary and
therapy following their initial resection and had negative hepatic metastasectomy is comparable to that seen with iso-
CT scan and tumor markers at 6 months. Randomization lated hepatic resection, although it is unclear to what extent
was then to either continued standard surveillance or a this is a reflection of treatment versus tumor biology [94, 95].
second-look surgery with CRS/HIPEC (with either oxali- Available data suggest that no medical intervention is as
platin and irinotecan, oxaliplatin alone, or mitomycin C efficacious in increasing survival as surgical resection for
alone) and simultaneous intravenous fluorouracil and leu- both isolated hepatic metastases and hepatic metastases with
covorin. At a median follow-up of 50.8 months, there were EHD. Even so, the median survival reported in patients
no differences in recurrence rates, 5-year DFS, or OS undergoing resection of multisite EHD is comparable to that
between groups, therefore, arguing against routine second- seen with chemotherapy alone and is rarely curative [96, 97].
look surgery. However, it should be noted that, there was Proposed scoring systems for OS and recurrence-free sur-
significant heterogeneity in the HIPEC chemotherapeutic vival (RFS) based on preoperatively known factors include
agents employed and the trial was not designed to assess largest liver metastasis >3 cm, >5 liver metastases, and unfa-
the efficacy of prophylactic HIPEC in high-risk patients vorable EHD site (portal/retroperitoneal lymph nodes or
without PM. multiple sites) with each criterion associated with worse out-
38 The Role of Surgery in Managing Primary and Metastatic Colorectal Cancer 415

comes [91, 98]. Abstracting any uniform recommendation tions having a more deleterious association with outcomes in
from the present literature is challenging due to substantial younger compared with older patients undergoing resection
heterogeneity in patient selection, perioperative chemother- of CRC liver metastases compared to older ones [108], and
apy regimens, adjunctive treatment modalities, and possible increasing prognostic precision when considering combina-
reporting bias [62, 92, 96, 98]. tions of coexisting mutations [109]. Such factors may ulti-
mately be incorporated into risk-benefit stratification
paradigms.
38.6 Clinical Prognostication Through
Histologic and Genetic Factors
38.6.3 Future Directions
38.6.1 Immunoscores
Promising avenues of investigation include the use of cell-­
Growing translational research continues to offer useful free DNA, not only in the earlier detection of disease pro-
prognostic information in patients with metastatic CRC gression but also prediction of which patients will maximally
based on histologic factors of the primary and metastatic benefit from resection of oligometastatic disease based on
tumors [99]. For example, cellular expression of COX-2, genetic profile [110]. Further research is needed in order to
HLA, and PD-1 in resected metastatic deposits is indepen- delineate the effects of primary tumor and/or metastatic
dently associated with prolonged RFS and OS after systemic resection on responsiveness to immunotherapy or other sys-
chemotherapy [100]. A general indication of the host immune temic treatments [111].
reaction to the tumor can be determined through assessment
of T-cell infiltration at the tumor center and invasive margin,
which is then used to calculate the immunoscore of a patho- 38.7 Conclusions
logic specimen [101]. Some data suggest that these scores
prognostically outperform current clinical staging systems in • Histopathologic lymph node status remains the most
terms of DSS and OS, with a more robust immune response important prognostic factor for patients diagnosed with
associated with better outcomes [99, 100, 102]. Furthermore, colorectal cancer without distant metastases.
activation of known oncogenic signaling pathways, such as • Although absolute lymph node count is currently used as
WNT/β-catenin, dampens T-cell recruitment and reduces the a metric of quality surgical care, a complex interplay of
efficacy of T-cell targeted monoclonal antibody therapy in a multiple factors is likely responsible for the lymph node
mouse model [103]. Even so, an increasingly sophisticated yield harvested at the time of surgery, bringing into ques-
understanding of the mechanisms of metastasis, tumor mod- tion the utility of nodal count as a quality metric.
ulation of the cellular microenvironment, and the maladap- • More extensive nodal dissections lead to higher nodal
tive host response is leading to promising new counts, but whether these extended dissections lead to
immunotherapies. Indeed, immune checkpoint blockade has improved survival remains unclear, and must be weighed
already shown efficacy for patients with microsatellite insta- against the associated increased morbidity observed when
bility–high metastatic colorectal cancer [104, 105]. they are performed.
• With careful patient selection, resection of colorectal can-
cer hepatic metastases is associated with improved sur-
38.6.2 P
 atient Selection Through Mutational vival, but the optimal timing and sequence of
Analysis metastasectomy continue to be debated and must be
weighed against surgical morbidity that can lead to delays
Growing evidence also supports genetic analysis of tumor in systemic therapies.
specimens as an emerging prognostic factor in CRC. As • While cytoreductive surgery may provide a survival ben-
described previously, older clinical scoring systems incorpo- efit for patients with colorectal cancer peritoneal metasta-
rating clinical markers can be determined preoperatively to ses, studies are ongoing to determine the exact role of
infer oncologic benefit from resection of CRC liver metasta- cytoreductive surgery/heated intraperitoneal chemother-
ses [48]. However, these systems are now being modified to apy in the management of colorectal cancer.
include mutational analyses in specific signaling pathways • Advanced molecular evaluations and genetic analyses
that are associated with OS, such as RAS-RAF and SMAD may allow for a more targeted and individualized approach
[106, 107]. Other nuanced associations with potential clini- to the management of metastatic colorectal cancer.
cal implications continue to be described such as RAS muta- • Ongoing and future research will strive to answer:
416 R. J. Straker III et al.

10. Chang GJ, Rodriguez-Bigas MA, Skibber JM, Moyer VA. Lymph
Open Questions node evaluation and survival after curative resection of colon can-
cer: systematic review. J Natl Cancer Inst. 2007;99(6):433–41.
• Is there a better metric of quality surgical care than https://doi.org/10.1093/jnci/djk092.
absolute lymph node harvest for patients undergoing 11. Chen SL, Bilchik AJ. More extensive nodal dissection improves
surgical resection of primary colorectal cancer? survival for stages I to III of colon cancer: a population-based
study. Ann Surg. 2006;244(4):602–10. https://doi.org/10.1097/01.
• Do more extensive lymph node dissections lead to sla.0000237655.11717.50.
improved survival? 12. Tepper JE, O’Connell MJ, Niedzwiecki D, et al. Impact of num-
• Can patients undergo major resections of primary ber of nodes retrieved on outcome in patients with rectal can-
colorectal cancer and hepatic metastases in a syn- cer. J Clin Oncol. 2001;19(1):157–63. https://doi.org/10.1200/
JCO.2001.19.1.157.
chronous fashion without suffering increased mor- 13. Märkl B. Stage migration vs immunology: the lymph node count
bidity in comparison to a staged approach? story in colon cancer. World J Gastroenterol. 2015;21(43):12218–
• What is the optimal combination of systemic ther- 33. https://doi.org/10.3748/wjg.v21.i43.12218.
apy and cytoreductive surgery/heated intraperito- 14. van Erning FN, Crolla RM, Rutten HJ, Beerepoot LV, van Krieken
JH, Lemmens VE. No change in lymph node positivity rate despite
neal chemotherapy for patients with colorectal increased lymph node yield and improved survival in colon can-
cancer peritoneal metastases? cer. Eur J Cancer. 2014;50(18):3221–9. https://doi.org/10.1016/j.
• Can the surgical management of metastatic colorec- ejca.2014.10.011.
tal cancer be stratified based on genetic and molec- 15. Baxter NN, Ricciardi R, Simunovic M, Urbach DR, Virnig
BA. An evaluation of the relationship between lymph node num-
ular evaluations? ber and staging in pT3 colon cancer using population-based data.
Dis Colon Rectum. 2010;53(1):65–70. https://doi.org/10.1007/
DCR.0b013e3181c70425.
Disclosures The authors declare no conflicts of interest or external 16. Parnaby CN, Scott NW, Ramsay G, et al. Prognostic value of
funding for this publication. lymph node ratio and extramural vascular invasion on survival for
patients undergoing curative colon cancer resection. Br J Cancer.
2015;113(2):212–9. https://doi.org/10.1038/bjc.2015.211.
17. American College of Surgeons. Commission on cancer quality
of care measures. https://www.facs.org/quality-­programs/cancer/
References ncdb/qualitymeasures. Accessed 18 Nov 2020
18. Baxter NN. Is lymph node count an ideal quality indicator for can-
1. Pienta KJ, Robertson BA, Coffey DS, Taichman RS. The cancer cer care? J Surg Oncol. 2009;99(4):265–8. https://doi.org/10.1002/
diaspora: metastasis beyond the seed and soil hypothesis. Clin jso.21197.
Cancer Res. 2013;19(21):5849–55. https://doi.org/10.1158/1078- 19. Parsons HM, Tuttle TM, Kuntz KM, Begun JW, McGovern
­0432.CCR-­13-­2158. PM, Virnig BA. Association between lymph node evalua-
2. Weichselbaum RR, Hellman S. Oligometastases revisited. Nat tion for colon cancer and node positivity over the past 20
Rev Clin Oncol. 2011;8(6):378–82. https://doi.org/10.1038/ years. JAMA. 2011;306(10):1089–97. https://doi.org/10.1001/
nrclinonc.2011.44. jama.2011.1285.
3. Pitroda SP, Khodarev NN, Huang L, et al. Integrated molecu- 20. Amin M, Edge S, Greene F et al AJCC cancer staging manual,
lar subtyping defines a curable oligometastatic state in colorec- 8th edn. Springer International Publishing: American Joint
tal liver metastasis. Nat Commun. 2018;9(1):1793. https://doi. Commission on Cancer; 2017.
org/10.1038/s41467-­018-­04278-­6. 21. Jin M, Roth R, Rock JB, Washington MK, Lehman A, Frankel
4. Baxter NN, Virnig DJ, Rothenberger DA, Morris AM, Jessurun J, WL. The impact of tumor deposits on colonic adenocarci-
Virnig BA. Lymph node evaluation in colorectal cancer patients: noma AJCC TNM staging and outcome. Am J Surg Pathol.
a population-based study. J Natl Cancer Inst. 2005;97(3):219–25. 2015;39(1):109–15. https://doi.org/10.1097/PAS.0000000000
https://doi.org/10.1093/jnci/dji020. 000320.
5. Edge S, Byrd D, Compton C, Fritz A, Greene F, Trotti A. AJCC 22. Evans MD, Barton K, Rees A, Stamatakis JD, Karandikar SS. The
cancer staging manual 7th edition. 7th edn. Springer; 2010. impact of surgeon and pathologist on lymph node retrieval in
6. National Comprehensive Cancer Network. Colon cancer, version colorectal cancer and its impact on survival for patients with
4.2020, NCCN clinical practice guidelines in oncology. https:// Dukes’ stage B disease. Color Dis. 2008;10(2):157–64. https://
www.nccn.org/professionals/physician_gls/pdf/colon_blocks.pdf. doi.org/10.1111/j.1463-­1318.2007.01225.x.
Accessed 17 Nov 2020. 23. de Burlet KJ, van den Hout MF, Putter H, Smit VT, Hartgrink
7. National Comprehensive Cancer Network. Rectal cancer, version HH. Total number of lymph nodes in oncologic resections, is there
6.2020, NCCN clinical practice guidelines in oncology. https:// more to be found? J Gastrointest Surg. 2015;19(5):943–8. https://
www.nccn.org/professionals/physician_gls/pdf/rectal_blocks.pdf. doi.org/10.1007/s11605-­015-­2764-­9.
Accessed 17 Nov 2020 24. Stocchi L, Fazio VW, Lavery I, Hammel J. Individual surgeon,
8. Nelson H, Petrelli N, Carlin A, et al. Guidelines 2000 for colon pathologist, and other factors affecting lymph node harvest in
and rectal cancer surgery. J Natl Cancer Inst. 2001;93(8):583–96. stage II colon carcinoma. Is a minimum of 12 examined lymph
https://doi.org/10.1093/jnci/93.8.583. nodes sufficient? Ann Surg Oncol. 2011;18(2):405–12. https://doi.
9. Joseph NE, Sigurdson ER, Hanlon AL, et al. Accuracy of deter- org/10.1245/s10434-­010-­1308-­5.
mining nodal negativity in colorectal cancer on the basis of 25. Gonsalves WI, Kanuri S, Tashi T, et al. Clinicopathologic fac-
the number of nodes retrieved on resection. Ann Surg Oncol. tors associated with lymph node retrieval in resectable colon
2003;10(3):213–8. https://doi.org/10.1245/aso.2003.03.059. cancer: a Veterans’ Affairs Central Cancer Registry (VACCR)
38 The Role of Surgery in Managing Primary and Metastatic Colorectal Cancer 417

database analysis. J Surg Oncol. 2011;104(6):667–71. https://doi. colorectal surgery. Scand J Surg. 2012;101(1):21–5. https://doi.
org/10.1002/jso.21886. org/10.1177/145749691210100105.
26. Kuo YH, Lee KF, Chin CC, Huang WS, Yeh CH, Wang JY. Does 41. Kandagatla P, Maguire LH, Hardiman KM. Biology of nodal
body mass index impact the number of LNs harvested and influ- spread in colon cancer: insights from molecular and genetic
ence long-term survival rate in patients with stage III colon cancer? studies. Eur Surg Res. 2018;59(5–6):361–70. https://doi.
Int J Color Dis. 2012;27(12):1625–35. https://doi.org/10.1007/ org/10.1159/000494832.
s00384-­012-­1496-­5. 42. Zhang C, Zhang L, Xu T, et al. Mapping the spreading routes of
27. Mekenkamp LJ, van Krieken JH, Marijnen CA, van de Velde lymphatic metastases in human colorectal cancer. Nat Commun.
CJ, Nagtegaal ID, Pathology Review Committee and the 2020;11(1):1993. https://doi.org/10.1038/s41467-­020-­15886-­6.
Co-operative Clinical Investigators. Lymph node retrieval in rec- 43. Güller U, Zettl A, Worni M, et al. Molecular investigation of
tal cancer is dependent on many factors—the role of the tumor, lymph nodes in colon cancer patients using one-step nucleic
the patient, the surgeon, the radiotherapist, and the pathologist. acid amplification (OSNA): a new road to better staging?
Am J Surg Pathol. 2009;33(10):1547–53. https://doi.org/10.1097/ Cancer. 2012;118(24):6039–45. https://doi.org/10.1002/
PAS.0b013e3181b2e01f. cncr.27667.
28. Ahmadi O, Stringer MD, Black MA, McCall JL. Influence of age 44. Xu J, Fan J, Qin X, et al. Chinese guidelines for the diagnosis and
and site of disease on lymph node yield in colorectal cancer. N Z comprehensive treatment of colorectal liver metastases (version
Med J. 2014;127(1395):31–40. 2018). J Cancer Res Clin Oncol. 2019;145(3):725–36. https://doi.
29. Søreide K, Nedrebø BS, Søreide JA, Slewa A, Kørner H. Lymph org/10.1007/s00432-­018-­2795-­1.
node harvest in colon cancer: influence of microsatellite instabil- 45. Adams RB, Aloia TA, Loyer E, et al. Selection for hepatic
ity and proximal tumor location. World J Surg. 2009;33(12):2695– resection of colorectal liver metastases: expert consensus
703. https://doi.org/10.1007/s00268-­009-­0255-­4. statement. HPB (Oxford). 2013;15(2):91–103. https://doi.
30. Märkl B, Rößle J, Arnholdt HM, et al. The clinical significance of org/10.1111/j.1477-­2574.2012.00557.x.
lymph node size in colon cancer. Mod Pathol. 2012;25(10):1413– 46. Fretland Å, Dagenborg VJ, Bjørnelv GMW, et al. Laparoscopic
22. https://doi.org/10.1038/modpathol.2012.92. versus open resection for colorectal liver metastases: the OSLO-­
31. Le Voyer TE, Sigurdson ER, Hanlon AL, et al. Colon cancer COMET randomized controlled trial. Ann Surg. 2018;267(2):199–
survival is associated with increasing number of lymph nodes 207. https://doi.org/10.1097/SLA.0000000000002353.
analyzed: a secondary survey of intergroup trial INT-0089. 47. Pan L, Tong C, Fu S, et al. Laparoscopic procedure is associ-
J Clin Oncol. 2003;21(15):2912–9. https://doi.org/10.1200/ ated with lower morbidity for simultaneous resection of colorec-
JCO.2003.05.062. tal cancer and liver metastases: an updated meta-analysis.
32. Ong ML, Schofield JB. Assessment of lymph node involvement in World J Surg Oncol. 2020;18(1):251. https://doi.org/10.1186/
colorectal cancer. World J Gastrointest Surg. 2016;8(3):179–92. s12957-­020-­02018-­z.
https://doi.org/10.4240/wjgs.v8.i3.179. 48. Fong Y, Fortner J, Sun RL, Brennan MF, Blumgart LH. Clinical
33. Brouwer NPM, Hugen N, Nagtegaal ID. More extensive lymph- score for predicting recurrence after hepatic resection for meta-
adenectomy in colon cancer; how far are we willing to go for a static colorectal cancer: analysis of 1001 consecutive cases.
biomarker? Tech Coloproctol. 2020;24(7):761–4. https://doi. Ann Surg. 1999;230(3):309–18; discussion 318–21. https://doi.
org/10.1007/s10151-­020-­02239-­0. org/10.1097/00000658-­199909000-­00004.
34. West NP, Kobayashi H, Takahashi K, et al. Understanding optimal 49. Lillemoe HA, Vauthey JN. Surgical approach to synchronous
colonic cancer surgery: comparison of Japanese D3 resection and colorectal liver metastases: staged, combined, or reverse strat-
European complete mesocolic excision with central vascular liga- egy. Hepatobiliary Surg Nutr. 2020;9(1):25–34. https://doi.
tion. J Clin Oncol. 2012;30(15):1763–9. https://doi.org/10.1200/ org/10.21037/hbsn.2019.05.14.
JCO.2011.38.3992. 50. Martin RC 2nd, Augenstein V, Reuter NP, Scoggins CR,
35. Kanemitsu Y, Komori K, Kimura K, Kato T. D3 lymph node dis- McMasters KM. Simultaneous versus staged resection for syn-
section in right hemicolectomy with a no-touch isolation technique chronous colorectal cancer liver metastases. J Am Coll Surg.
in patients with colon cancer. Dis Colon Rectum. 2013;56(7):815– 2009;208(5):842–50; discussion 850–2. https://doi.org/10.1016/j.
24. https://doi.org/10.1097/DCR.0b013e3182919093. jamcollsurg.2009.01.031.
36. Hohenberger W, Weber K, Matzel K, Papadopoulos T, Merkel 51. Thelen A, Jonas S, Benckert C, et al. Simultaneous versus staged
S. Standardized surgery for colonic cancer: complete mesocolic liver resection of synchronous liver metastases from colorec-
excision and central ligation—technical notes and outcome. tal cancer. Int J Color Dis. 2007;22(10):1269–76. https://doi.
Color Dis. 2009;11(4):354–64; discussion 364–5. https://doi. org/10.1007/s00384-­007-­0286-­y.
org/10.1111/j.1463-­1318.2008.01735.x. 52. Reddy SK, Pawlik TM, Zorzi D, et al. Simultaneous resections
37. Rouffet F, Hay JM, Vacher B, et al. Curative resection for left of colorectal cancer and synchronous liver metastases: a multi-­
colonic carcinoma: hemicolectomy vs. segmental colectomy. A institutional analysis. Ann Surg Oncol. 2007;14(12):3481–91.
prospective, controlled, multicenter trial. French Association for https://doi.org/10.1245/s10434-­007-­9522-­5.
Surgical Research. Dis Colon Rectum. 1994;37(7):651–9. https:// 53. Feng Q, Wei Y, Zhu D, et al. Timing of hepatectomy for resect-
doi.org/10.1007/BF02054407. able synchronous colorectal liver metastases: for whom simul-
38. Nagawa H, Muto T, Sunouchi K, et al. Randomized, controlled taneous resection is more suitable—a meta-analysis. PLoS
trial of lateral node dissection vs. nerve-preserving resection in One. 2014;9(8):e104348. https://doi.org/10.1371/journal.
patients with rectal cancer after preoperative radiotherapy. Dis pone.0104348.
Colon Rectum. 2001;44(9):1274–80. https://doi.org/10.1007/ 54. Boudjema K, Locher C, Sabbagh C, et al. Simultaneous versus
BF02234784. delayed resection for initially resectable synchronous colorec-
39. Bertelsen CA, Neuenschwander AU, Jansen JE, et al. Short-term tal cancer liver metastases: a prospective, open-label, random-
outcomes after complete mesocolic excision compared with ‘con- ized, controlled trial. Ann Surg. 2021;273(1):49–56. https://doi.
ventional’ colonic cancer surgery. Br J Surg. 2016;103(5):581–9. org/10.1097/SLA.0000000000003848.
https://doi.org/10.1002/bjs.10083. 55. Mentha G, Majno PE, Andres A, Rubbia-Brandt L, Morel P,
40. Tsujinaka S, Kawamura YJ, Tan KY, et al. Proximal bowel Roth AD. Neoadjuvant chemotherapy and resection of advanced
necrosis after high ligation of the inferior mesenteric artery in synchronous liver metastases before treatment of the colorectal
418 R. J. Straker III et al.

primary. Br J Surg. 2006;93(7):872–8. https://doi.org/10.1002/ 69. Aloia T, Sebagh M, Plasse M, et al. Liver histology and surgi-
bjs.5346. cal outcomes after preoperative chemotherapy with fluoro-
56. Lam VW, Laurence JM, Johnston E, Hollands MJ, Pleass uracil plus oxaliplatin in colorectal cancer liver metastases.
HC, Richardson AJ. A systematic review of two-stage hepa- J Clin Oncol. 2006;24(31):4983–90. https://doi.org/10.1200/
tectomy in patients with initially unresectable colorectal liver JCO.2006.05.8156.
metastases. HPB (Oxford). 2013;15(7):483–91. https://doi. 70. Narita M, Oussoultzoglou E, Fuchshuber P, et al. What is a safe
org/10.1111/j.1477-­2574.2012.00607.x. future liver remnant size in patients undergoing major hepatec-
57. Jaeck D, Oussoultzoglou E, Rosso E, Greget M, Weber JC, tomy for colorectal liver metastases and treated by intensive pre-
Bachellier P. A two-stage hepatectomy procedure combined with operative chemotherapy? Ann Surg Oncol. 2012;19(8):2526–38.
portal vein embolization to achieve curative resection for initially https://doi.org/10.1245/s10434-­012-­2274-­x.
unresectable multiple and bilobar colorectal liver metastases. Ann 71. Kishi Y, Zorzi D, Contreras CM, et al. Extended preoperative che-
Surg. 2004;240(6):1037–49; discussion 1049–51. https://doi. motherapy does not improve pathologic response and increases
org/10.1097/01.sla.0000145965.86383.89. postoperative liver insufficiency after hepatic resection for
58. Karoui M, Vigano L, Goyer P, et al. Combined first-stage hepa- colorectal liver metastases. Ann Surg Oncol. 2010;17(11):2870–6.
tectomy and colorectal resection in a two-stage hepatectomy https://doi.org/10.1245/s10434-­010-­1166-­1.
strategy for bilobar synchronous liver metastases. Br J Surg. 72. Kopetz S, Vauthey JN. Perioperative chemotherapy for resectable
2010;97(9):1354–62. https://doi.org/10.1002/bjs.7128. hepatic metastases. Lancet. 2008;371(9617):963–5. https://doi.
59. Schnitzbauer AA, Lang SA, Goessmann H, et al. Right portal vein org/10.1016/S0140-­6736(08)60429-­8.
ligation combined with in situ splitting induces rapid left lateral 73. Khan K, Wale A, Brown G, Chau I. Colorectal cancer with liver
liver lobe hypertrophy enabling 2-staged extended right hepatic metastases: neoadjuvant chemotherapy, surgical resection first or
resection in small-for-size settings. Ann Surg. 2012;255(3):405– palliation alone? World J Gastroenterol. 2014;20(35):12391–406.
14. https://doi.org/10.1097/SLA.0b013e31824856f5. https://doi.org/10.3748/wjg.v20.i35.12391.
60. Selby K, Hernandez-Alejandro R. Two-stage hepatectomy for liver 74. Benoist S, Brouquet A, Penna C, et al. Complete response of
metastasis from colorectal cancer. CMAJ. 2014;186(15):1163–6. colorectal liver metastases after chemotherapy: does it mean cure?
https://doi.org/10.1503/cmaj.131022. J Clin Oncol. 2006;24(24):3939–45. https://doi.org/10.1200/
61. Bertens KA, Hawel J, Lung K, Buac S, Pineda-Solis K, JCO.2006.05.8727.
Hernandez-Alejandro R. ALPPS: challenging the concept of 75. Bartlett DL, Berlin J, Lauwers GY, Messersmith WA, Petrelli
unresectability—a systematic review. Int J Surg. 2015;13:280–7. NJ, Venook AP. Chemotherapy and regional therapy of hepatic
https://doi.org/10.1016/j.ijsu.2014.12.008. colorectal metastases: expert consensus statement. Ann
62. Stewart CL, Warner S, Ito K, et al. Cytoreduction for colorectal Surg Oncol. 2006;13(10):1284–92. https://doi.org/10.1245/
metastases: liver, lung, peritoneum, lymph nodes, bone, brain. s10434-­006-­9018-­8.
When does it palliate, prolong survival, and potentially cure? 76. Chouliaras K, Russell G, Levine E, et al. Hepatic arterial infu-
Curr Probl Surg. 2018;55(9):330–79. https://doi.org/10.1067/j. sion chemotherapy for colorectal liver metastases revisited.
cpsurg.2018.08.004. HPB (Oxford). 2020;22(9):1265–70. https://doi.org/10.1016/j.
63. Ratti F, Schadde E, Masetti M, et al. Strategies to increase the hpb.2019.12.003.
resectability of patients with colorectal liver metastases: a multi-­ 77. Sauer R, Liersch T, Merkel S, et al. Preoperative versus postopera-
center case-match analysis of ALPPS and conventional two-stage tive chemoradiotherapy for locally advanced rectal cancer: results
hepatectomy. Ann Surg Oncol. 2015;22(6):1933–42. https://doi. of the German CAO/ARO/AIO-94 randomized phase III trial after
org/10.1245/s10434-­014-­4291-­4. a median follow-up of 11 years. J Clin Oncol. 2012;30(16):1926–
64. Moris D, Ronnekleiv-Kelly S, Kostakis ID, et al. Operative results 33. https://doi.org/10.1200/JCO.2011.40.1836.
and oncologic outcomes of associating liver partition and portal 78. Ngan SY, Burmeister B, Fisher RJ, et al. Randomized trial of
vein ligation for staged hepatectomy (ALPPS) versus two-stage short-course radiotherapy versus long-course chemoradiation
hepatectomy (TSH) in patients with unresectable colorectal liver comparing rates of local recurrence in patients with T3 rec-
metastases: a systematic review and meta-analysis. World J Surg. tal cancer: trans-Tasman radiation oncology group trial 01.04.
2018;42(3):806–15. https://doi.org/10.1007/s00268-­017-­4181-­6. J Clin Oncol. 2012;30(31):3827–33. https://doi.org/10.1200/
65. Nordlinger B, Sorbye H, Glimelius B, et al. Perioperative JCO.2012.42.9597.
chemotherapy with FOLFOX4 and surgery versus surgery 79. Bujko K, Nowacki MP, Nasierowska-Guttmejer A, Michalski W,
alone for resectable liver metastases from colorectal cancer Bebenek M, Kryj M. Long-term results of a randomized trial com-
(EORTC Intergroup trial 40983): a randomised controlled trial. paring preoperative short-course radiotherapy with preoperative
Lancet. 2008;371(9617):1007–16. https://doi.org/10.1016/ conventionally fractionated chemoradiation for rectal cancer. Br J
S0140-­6736(08)60455-­9. Surg. 2006;93(10):1215–23. https://doi.org/10.1002/bjs.5506.
66. Nordlinger B, Sorbye H, Glimelius B, et al. Perioperative 80. Jacobson R, Sherman SK, Dahdaleh F, Turaga KK. Peritoneal
FOLFOX4 chemotherapy and surgery versus surgery alone metastases in colorectal cancer. Ann Surg Oncol. 2018;25(8):2145–
for resectable liver metastases from colorectal cancer (EORTC 51. https://doi.org/10.1245/s10434-­018-­6490-­x.
40983): long-term results of a randomised, controlled, phase 81. Hugen N, van de Velde CJH, de Wilt JHW, Nagtegaal
3 trial. Lancet Oncol. 2013;14(12):1208–15. https://doi. ID. Metastatic pattern in colorectal cancer is strongly influenced
org/10.1016/S1470-­2045(13)70447-­9. by histological subtype. Ann Oncol. 2014;25(3):651–7. https://
67. Vauthey JN, Pawlik TM, Ribero D, et al. Chemotherapy regi- doi.org/10.1093/annonc/mdt591.
men predicts steatohepatitis and an increase in 90-day mor- 82. Ellis RJ, Schlick CJR, Yang AD, Barber EL, Bilimoria KY,
tality after surgery for hepatic colorectal metastases. J Clin Merkow RP. Utilization and treatment patterns of cytoreduction
Oncol. 2006;24(13):2065–72. https://doi.org/10.1200/ surgery and intraperitoneal chemotherapy in the United States.
JCO.2005.05.3074. Ann Surg Oncol. 2020;27(1):214–21. https://doi.org/10.1245/
68. Abdalla EK, Vauthey JN. Steatosis as a risk factor in liver sur- s10434-­019-­07492-­8.
gery. Ann Surg. 2007;246(2):340–1; author reply 341. https://doi. 83. Verwaal VJ, van Ruth S, de Bree E, et al. Randomized trial
org/10.1097/SLA.0b013e31811ea9d2. of cytoreduction and hyperthermic intraperitoneal chemo-
38 The Role of Surgery in Managing Primary and Metastatic Colorectal Cancer 419

therapy versus systemic chemotherapy and palliative surgery vival outcomes. HPB (Oxford). 2016;18(3):209–20. https://doi.
in patients with peritoneal carcinomatosis of colorectal cancer. org/10.1016/j.hpb.2015.12.004.
J Clin Oncol. 2003;21(20):3737–43. https://doi.org/10.1200/ 97. de Jong MC, Pulitano C, Ribero D, et al. Rates and patterns of
JCO.2003.04.187. recurrence following curative intent surgery for colorectal liver
84. Chicago Consensus Working Group. The Chicago consensus metastasis: an international multi-institutional analysis of 1669
on peritoneal surface malignancies: management of colorectal patients. Ann Surg. 2009;250(3):440–8. https://doi.org/10.1097/
metastases. Ann Surg Oncol. 2020;27(6):1761–7. https://doi. SLA.0b013e3181b4539b.
org/10.1245/s10434-­020-­08315-­x. 98. Leung U, Gonen M, Allen PJ, et al. Colorectal cancer liver metas-
85. Verwaal VJ, Bruin S, Boot H, van Slooten G, van Tinteren H. tases and concurrent extrahepatic disease treated with resec-
8-year follow-up of randomized trial: cytoreduction and hyper- tion. Ann Surg. 2017;265(1):158–65. https://doi.org/10.1097/
thermic intraperitoneal chemotherapy versus systemic chemother- SLA.0000000000001624.
apy in patients with peritoneal carcinomatosis of colorectal cancer. 99. Bruni D, Angell HK, Galon J. The immune contexture and
Ann Surg Oncol. 2008;15(9):2426–32. https://doi.org/10.1245/ immunoscore in cancer prognosis and therapeutic efficacy.
s10434-­008-­9966-­2. Nat Rev Cancer. 2020;20(11):662–80. https://doi.org/10.1038/
86. Jayne DG, Fook S, Loi C, Seow-Choen F. Peritoneal carcinoma- s41568-­020-­0285-­7.
tosis from colorectal cancer. Br J Surg. 2002;89(12):1545–50. 100. Moretto R, Corallo S, Belfiore A, et al. Prognostic impact of
https://doi.org/10.1046/j.1365-­2168.2002.02274.x. immune-microenvironment in colorectal liver metastases resected
87. Glehen O, Kwiatkowski F, Sugarbaker PH, et al. Cytoreductive after triplets plus a biologic agent: a pooled analysis of five
surgery combined with perioperative intraperitoneal chemo- prospective trials. Eur J Cancer. 2020;135:78–88. https://doi.
therapy for the management of peritoneal carcinomatosis from org/10.1016/j.ejca.2020.04.045.
colorectal cancer: a multi-institutional study. J Clin Oncol. 101. Angell HK, Bruni D, Barrett JC, Herbst R, Galon J. The
2004;22(16):3284–92. https://doi.org/10.1200/JCO.2004.10.012. immunoscore: colon cancer and beyond. Clin Cancer Res.
88. Quenet F, Elias D, Roca L, Goere D, Ghouti L, Pocard M, Facy 2020;26(2):332–9. https://doi.org/10.1158/1078-­0432.
O, Arvieux C, Lorimier G, Pezet D, Marchal F, Loi V, Meeus P, CCR-­18-­1851.
De Forges H, Stanbury T, Paineau L, Glehen O. A UNICANCER 102. Mlecnik B, Tosolini M, Kirilovsky A, et al. Histopathologic-based
phase III trial of hyperthermic intra-peritoneal chemotherapy for prognostic factors of colorectal cancers are associated with the
colorectal peritoneal carcinomatosis: PRODIGE 7. J Clin Oncol. state of the local immune reaction. J Clin Oncol. 2011;29(6):610–
2018;36(abstr LBA3503.) 8. https://doi.org/10.1200/JCO.2010.30.5425.
89. Klaver CEL, Wisselink DD, Punt CJA, et al. Adjuvant hyper- 103. Spranger S, Bao R, Gajewski TF. Melanoma-intrinsic beta-­
thermic intraperitoneal chemotherapy in patients with locally catenin signalling prevents anti-tumour immunity. Nature.
advanced colon cancer (COLOPEC): a multicentre, open-label, 2015;523(7559):231–5. https://doi.org/10.1038/nature14404.
randomised trial. Lancet Gastroenterol Hepatol. 2019;4(10):761– 104. Yu X, Zhu L, Liu J, Xie M, Chen J, Li J. Emerging role of immu-
70. https://doi.org/10.1016/S2468-­1253(19)30239-­0. notherapy for colorectal cancer with liver metastasis. Onco Targets
90. Goere D, Glehen O, Quenet F, et al. Second-look surgery plus Ther. 2020;13:11645–58. https://doi.org/10.2147/OTT.S271955.
hyperthermic intraperitoneal chemotherapy versus surveillance in 105. Biller LH, Schrag D. Diagnosis and treatment of metastatic
patients at high risk of developing colorectal peritoneal metastases colorectal cancer: a review. JAMA. 2021;325(7):669–85. https://
(PROPHYLOCHIP-PRODIGE 15): a randomised, phase 3 study. doi.org/10.1001/jama.2021.0106.
Lancet Oncol. 2020;21(9):1147–54. https://doi.org/10.1016/ 106. Lang H, Baumgart J, Heinrich S, et al. Extended molecular profil-
S1470-­2045(20)30322-­3. ing improves stratification and prediction of survival after resection
91. Elias D, Sideris L, Pocard M, et al. Results of R0 resection for of colorectal liver metastases. Ann Surg. 2019;270(5):799–805.
colorectal liver metastases associated with extrahepatic disease. https://doi.org/10.1097/SLA.0000000000003527.
Ann Surg Oncol. 2004;11(3):274–80. https://doi.org/10.1245/ 107. Brudvik KW, Jones RP, Giuliante F, et al. RAS mutation clini-
aso.2004.03.085. cal risk score to predict survival after resection of colorectal
92. Hwang M, Jayakrishnan TT, Green DE, et al. Systematic review liver metastases. Ann Surg. 2019;269(1):120–6. https://doi.
of outcomes of patients undergoing resection for colorec- org/10.1097/SLA.0000000000002319.
tal liver metastases in the setting of extra hepatic disease. Eur 108. Jacome AA, Vreeland TJ, Johnson B, et al. The prognos-
J Cancer. 2014;50(10):1747–57. https://doi.org/10.1016/j. tic impact of RAS on overall survival following liver resec-
ejca.2014.03.277. tion in early versus late-onset colorectal cancer patients. Br
93. Treasure T, Farewell V, Macbeth F, et al. Pulmonary metasta- J Cancer. 2021;124(4):797–804. https://doi.org/10.1038/
sectomy versus continued active monitoring in colorectal can- s41416-­020-­01169-­w.
cer (PulMiCC): a multicentre randomised clinical trial. Trials. 109. Kawaguchi Y, Kopetz S, Newhook TE, et al. Mutation status of
2019;20(1):718. https://doi.org/10.1186/s13063-­019-­3837-­y. RAS, TP53, and SMAD4 is superior to mutation status of RAS
94. Marudanayagam R, Ramkumar K, Shanmugam V, et al. Long-­ alone for predicting prognosis after resection of colorectal liver
term outcome after sequential resections of liver and lung metasta- metastases. Clin Cancer Res. 2019;25(19):5843–51. https://doi.
ses from colorectal carcinoma. HPB (Oxford). 2009;11(8):671–6. org/10.1158/1078-­0432.CCR-­19-­0863.
https://doi.org/10.1111/j.1477-­2574.2009.00115.x. 110. Phallen J, Sausen M, Adleff V, et al. Direct detection of early-­
95. Miller G, Biernacki P, Kemeny NE, et al. Outcomes after resection stage cancers using circulating tumor DNA. Sci Transl Med.
of synchronous or metachronous hepatic and pulmonary colorec- 2017;9(403):eaan2415. https://doi.org/10.1126/scitranslmed.
tal metastases. J Am Coll Surg. 2007;205(2):231–8. https://doi. aan2415.
org/10.1016/j.jamcollsurg.2007.04.039. 111. Nakayama I, Hirota T, Shinozaki E. BRAF mutation in colorec-
96. Hadden WJ, de Reuver PR, Brown K, Mittal A, Samra JS, Hugh tal cancers: from prognostic marker to targetable mutation.
TJ. Resection of colorectal liver metastases and extra-hepatic dis- Cancers (Basel). 2020;12(11):3236. https://doi.org/10.3390/
ease: a systematic review and proportional meta-analysis of sur- cancers12113236.
 The Role of Surgery in Management
of Gastric Cancer 39
Andrew J. Sinnamon, Jose M. Pimiento,
and Robert E. Roses

Abstract grammed death receptor immune checkpoint pathway are

under active investigation and show promise for appropri-
Gastric adenocarcinoma is an aggressive tumor with sig-
ately selected patients.
nificant global burden. Incidence is decreasing in Western
countries but gastric cancer remains a major cause of can-
cer death in East Asian and developing countries. Gastric Learning Objectives
adenocarcinomas have been historically subdivided • Understand classic histopathologic classification of
according to dominant histology by the Lauren classifica- gastric adenocarcinoma into intestinal- and diffuse-­
tion into gland-forming intestinal-type lesions and infil- type and correlation with molecular characteristics.
trative, non-gland forming diffuse-type lesions. Loss of • Appreciate lymphatic drainage patterns of gastric
E-cadherin expression encoded by the CDH1 gene plays tumors.
a major role in carcinogenesis of diffuse-type cancers • Be able to define the extent of surgical lymphade-
while Helicobacter pylori infection contributes to nectomy and identify evidence supporting D2 ver-
intestinal-­
type cancers. Gastric adenocarcinoma has a sus D1 and D3 lymphadenectomy.
proclivity for early invasion into lymphatic channels with • Recognize the implications of the anatomic epicen-
rates of lymph node involvement in the range of 20–30% ter of a gastric tumor.
with submucosal or deeper involvement by the primary • Identify the options for multimodality therapy
tumor. Adequate lymphadenectomy in potentially resect- including perioperative chemotherapy, adjuvant
able cases has therefore been a topic of substantial study. chemoradiotherapy, adjuvant chemotherapy alone,
A surgical approach accounting for the typical lymphatic as well as an emerging role for immune therapy.
drainage of gastric tumors has been well described by
Japanese surgeons and D2 lymphadenectomy, encom-
passing surgical clearance of the extended regional basin,
has increasingly been adopted worldwide. Multimodal 39.1 Introduction
therapy plays an important role in clinical stage II–III
cancers, with perioperative chemotherapy gaining Gastric cancer—specifically adenocarcinoma—is an aggres-
momentum in recent years following the publication of sive entity with a propensity for early lymphatic spread. It is
randomized trial data supporting this approach. The role one of the more common cancers worldwide, but there is
of adjuvant chemoradiotherapy compared to adjuvant marked variation in geographic distribution related to envi-
chemotherapy alone remains a subject of controversy. ronmental and genetic risk factors. While the incidence con-
Therapy directed at the HER2 receptor and the pro- tinues to decrease in Western countries, including the United
States and Europe, it is common relative to other cancers in
East Asian countries, most notably Japan, Korea, and China,
as well as in other developing countries. While understand-
A. J. Sinnamon · J. M. Pimiento
Moffitt Cancer Center, Tampa, FL, USA ing of the underlying biology continues to evolve, gastric
e-mail: [email protected]; [email protected] cancers are generally classified into two histologic sub-
R. E. Roses (*) types—diffuse and intestinal. Surgery plays a critical role in
University of Pennsylvania, Philadelphia, PA, USA potentially curable gastric cancer as well as in select cases
e-mail: [email protected] for palliative intent. The merits of multimodality therapy are

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 421
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_39
422 A. J. Sinnamon et al.

well supported by a growing body of evidence, with surgery, beginning with infection and ending with gastric adenocarci-
chemotherapy, radiotherapy, and immune therapy playing noma over the course of decades has been described [7]. This
roles depending on the clinical scenario. process begins with active superficial gastritis related to H.
pylori infection. In a subset of individuals, this progresses to
chronic atrophic gastritis with loss of normal gastric glandu-
39.2 Histology and Pathogenesis lar tissue. The next step in this gastric carcinogenesis path-
way is intestinal metaplasia, first resembling small intestinal
Gastric adenocarcinomas are commonly classified as either epithelium and later resembling colonic epithelium. It is pos-
of two histologic subtypes according to the Lauren classifi- sible that during this sequence, the stomach may be colo-
cation [1]. Intestinal-type tumors maintain a gland-forming nized by bacteria more likely to produce carcinogenic
morphology, while diffuse-type cancers lack cohesion nitrosamines. It should be noted that while this cascade of
between cancer cells and thus do not exhibit gland-forming events is thought to result from H. pylori infection, a very
morphology. The overarching difference between these sub- small fraction of infected individuals develops gastric can-
types is the tendency for the intercellular adhesion responsi- cer. However, given the worldwide prevalence of H. pylori
ble for the formation of recognizable gland structures. This infection, the cumulative burden of disease is substantial.
feature is maintained in intestinal-type tumors and lost in dif- Efforts have been made to define gastric cancers according
fuse cancers. Lacking cohesion, cells from diffuse carcino- to molecular signature as improvements in genomic technolo-
mas tend to be more infiltrative into surrounding tissues, gies evolve. Pooling tumors with similar genetic and epigen-
which may result in the clinical entity known as linitis plas- etic patterns into subtypes may allow for better targeted, more
tica. Linitis plastica is characterized by a rigid, leathery con- effective treatment approaches. For example, The Cancer
sistency of the stomach on a macroscopic level. Moreover, Genome Atlas has defined four molecular subtypes of gastric
signet cells are often seen within diffuse-type carcinomas. adenocarcinoma [8]. These include chromosomally unstable
The subdivision by histologic appearance of gastric can- gastric cancers (~50% of gastric cancers), genomically stable
cers is now thought to reflect genetic pathways unknown to gastric cancers (~20%), microsatellite unstable tumors
Lauren at the time of description in 1965. The E-cadherin (~20%), and EBV-associated tumors (10%). Chromosomally
(“epithelial cadherin”) protein is encoded by the CDH1 gene unstable cancers are typically intestinal-type by histology.
and contributes to cell-to-cell adhesion and thus maintenance Genetically stable tumors are frequently diffuse-type by his-
of normal epithelial architecture [2, 3]. Loss of E-cadherin tology, and CDH1-­associated tumors make up about a third of
protein function has been shown to result in abnormal tissue these. Microsatellite unstable (MSI) tumors have a deficiency
architecture with unregulated cell growth and invasion into in the mismatch repair machinery (dMMR), which com-
adjacent tissues. A variety of genetic and epigenetic altera- monly results from epigenetic hypermethylation and silenc-
tions can lead to a defect in CDH1 genetic function. A fre- ing of gene expression. Inherited dMMR is known as Lynch
quently observed mechanism of such loss of function is syndrome, which most commonly includes colorectal and
promoter region hypermethylation. Hereditary diffuse g­ astric endometrial cancers but can be associated with gastric cancer
cancer (HDGC) is a manifestation of these genetics. Families as well.
with HDGC typically carry a pathogenic variant in the CDH1 Other molecular markers that have drawn particular inter-
gene, although a number of genetic and epigenetic variants est because of their therapeutic implications include HER2
have been described which result in the same phenotype [4, and PD-L1 overexpression. HER2 (human epidermal growth
5]. Lifetime risk of developing diffuse-type gastric cancer is factor receptor 2) is a receptor tyrosine kinase in the EGFR
very high and often occurs at an early age; cumulative inci- family, and overexpression of HER2 is thought to drive
dence is 70% among men and 56% among women, with an tumor growth. HER2 overexpression is more commonly
average age of onset of 38 years. Individuals with HDGC are seen in proximal and intestinal-type adenocarcinomas [9].
also at increased risk of lobular breast cancer, with a cumula- Tumors with PD-L1 (ligand of programmed cell death recep-
tive incidence of 42% for women. tor 1, PD-1) overexpression are capable of evading immune
Intestinal-type gastric adenocarcinoma, in contrast to clearance [10, 11]. PD-L1 on the tumor cell binds PD1 on
diffuse-­type, is not thought to be related to a single identifi- T-cells, thus causing T-cell inhibition, also known as immune
able genetic signature and is often associated with checkpoint activation. This phenomenon is under active
Helicobacter pylori infection. Anatomically, these cancers study in a variety of cancers and is of growing interest in
typically arise from the gastric body, fundus, and antrum. In gastric cancer as well.
fact, it is estimated that worldwide, 89% of non-cardia gas- Histologic depth of invasion defines T stage. In the eighth
tric cancers are attributable to H. pylori, making this patho- edition of the American Joint Committee on Cancer (AJCC)
gen a major global cause of cancer [6]. A multistep process Staging Manual, T1 tumors are subdivided into T1a lesions
39 The Role of Surgery in Management of Gastric Cancer 423

which invade the lamina propria or muscularis mucosae and below follow patterns seen in the development of gastric
T1b lesions which invade the submucosa [12]. T2 tumors cancers.
invade the muscularis propria and T3 tumors invade the sub- The Siewert classification is a commonly used framework
serosal connective tissue without invasion of adjacent struc- when considering GEJ adenocarcinomas [17]. Tumors are
tures or serosa. T4 tumors that penetrate the serosa only are divided into three types based on the location of their epicen-
classified as T4a lesions while invasion through the serosa ter. Siewert type I tumors arise from 1 to 5 cm above the
and into adjacent structures are classified as T4b. T stage is anatomic esophagogastric junction. Siewert type II lesions
correlated with the rate of lymphatic spread (N stage). This straddle the junction and are located from 1 cm above to
has been demonstrated in multiple surgical series. For 2 cm below the GEJ. Lastly, Siewert type III tumors extend
example, Maruyama and colleagues meticulously dissected 2–5 cm below the GEJ. In their original studies, Siewert and
specimens from 1931 gastric resections and identified colleagues suggested that the pattern of lymphatic drainage
increasing rates of regional nodal involvement in a stepwise of type II and III tumors is primarily directed inferiorly,
fashion for tumors with invasion into muscularis propria, toward the paracardial, lesser curvature, and left gastric
serosa, and into adjacent structures [13]. Similarly, in an artery nodes [18]. This is consistent with analyses performed
analysis of 920 gastric cancer patients undergoing resection, by Nigro and colleagues, which further demonstrated GEJ
the rate of N+ disease was shown to increase with higher T tumors with intramuscular invasion had a higher likelihood
stage [14]. Among patients with primary tumors confined to of spreading to abdominal lymph nodes along the left gastric
the mucosa (T1a), the rate of lymph node metastases was artery and greater curvature [19].
15.0% (n = 3/20). This increased to 23.4% (n = 15/64) with
T1b disease, 48.2% (n = 52/108) with T2 disease, 69.8%
(n = 280/358) with T3 disease, and 75.7% (n = 212/280) 39.4  atterns of Lymphatic Drainage
P
with T4a disease. These series may be somewhat limited and Clinical Assessment
with respect to early gastric cancers because of underrepre-
sentation of early tumors. As frequency of upper endoscopy Lymphatic drainage of the stomach includes two different
has increased over time, earlier cancers with submucosal networks [20]. The intrinsic network is located between the
only invasion have been increasingly detected. Rate of muscularis mucosae and the submucosa, and the extrinsic
lymph node metastases is approximately 10% in mucosal network follows the arterial blood supply to the stomach,
cancers (T1a) and 20–30% in submucosal (T1b) cancers which defines the regional nodal basin. Gastric lymphatics
[15, 16]. Independent risk factors for lymph node metasta- are designated by the arteries they follow and often drain to
ses include tumor size, depth of submucosal invasion, poor lymph nodes along these arteries. The final lymph nodes col-
differentiation, and lymphovascular invasion. Depth of sub- lecting stomach lymph are the preaortic (also “celiac”) nodes
mucosal invasion can be further separated into sm1, sm2, found in front of the aorta around the celiac trunk as it arises
and sm3 lesions, which represent invasion into the upper, from the abdominal aorta. Lymph from the gastric and duo-
middle, and lower third of the submucosal layer. Patients denal regions is collected from gastric, gastro-epiploic, pan-
with multiple risk factors may have appreciable rates of creaticosplenic, pyloric, and hepatic nodes that pass lymph
lymph node metastases despite only submucosal invasion of into the celiac nodes, intestinal lymph trunks, union of
the primary lesion [15]. abdominal lymph trunks and finally, to the thoracic duct.
The perigastric lymph nodes are thought to be the first
echelon nodes to harbor disease. These include the right and
39.3 Anatomic Considerations left paracardial nodes superiorly, nodes along the lesser and
greater curvature, and the supra- and infrapyloric nodes.
As described previously, carcinogenesis is often related to Much of the lymphatic drainage, particularly in the proximal
the anatomic site of the tumor, with non-cardia cancers com- stomach, coalesces around the left gastric, common hepatic,
monly resulting from H. pylori infection. In contrast, tumors and splenic arteries, as well as those nodes within the splenic
of the gastric cardia and gastroesophageal junction (GEJ) hilum and hepatoduodenal ligament. This lymphatic drain-
may represent an overlap region, both in terms of anatomy age paradigm has been described by the Japanese Gastric
and disease pathogenesis. Adenocarcinomas above this point Cancer Association, which defines the gastric cancer lymph
are thought to arise from the intestinal metaplasia sequence node stations [21]. However, it has been reported that lymph
seen in Barrett’s esophagus, while adenocarcinomas arising node metastases can “skip” nodal basins.
424 A. J. Sinnamon et al.

D1 D1
4sb
D1+ 4d 4sb D1+ 4d
D2 D2

4sa

6 6
3 2 3
1 1
5 5
7 7
12a 9 11d 8a 9
8a 11p 12a 11p

Total gastrectomy Distal gastrectomy

Figure courtesy of Japanese Gastric Cancer Association. Japanese gastric cancer treatment guidelines 2018 (fifth edition). Gastric Cancer 24, 1–21
(2021). https://doi.org/10.1007/s10120-­020-­01042-­y. Open access material.

Gastric cancer may spread beyond regional lymph nodes racy of EUS is enhanced with the addition of fine-needle
to involve the superior mesenteric, middle colic, retropan- aspiration (FNA) of suspicious lymph nodes for histopatho-
creatic, pancreaticoduodenal, peripancreatic, para-aortic, logic confirmation [25, 26].
and other retroperitoneal lymph nodes. These nodes are not Cross-sectional CT of the chest abdomen and pelvis is
considered part of the regional nodal basin. Thus, involve- critical for staging in gastric cancer and is recommended in
ment of these lymph nodes constitutes M1 disease accord- all patients by the NCCN. Its primary utility is in the detec-
ing to the Eighth Edition of the AJCC Cancer Staging tion of distant metastatic disease, as its accuracy in assessing
Manual. T stage and N stage may be eclipsed by EUS. Small regional
Involvement of the regional lymph nodes has important lymph nodes harboring metastatic disease may not be
implications for use of adjuvant and perioperative multimo- detected by CT, while conversely, enlarged nodes due to
dality therapy (to be discussed subsequently). Accurate inflammatory response may be falsely identified as positive.
lymph node staging is therefore critical in clinical practice Sensitivity, specificity, and overall accuracy for regional
and for treatment selection. A variety of diagnostic tech- nodal staging are reported to be 86–97%, 65–76%, and
niques exist with varying sensitivity and specificity for 58–87%, respectively [27–29]. The sensitivity for CT for the
regional nodal staging. Endoscopic ultrasound (EUS), cross-­ detection of nodal metastases increases with greater N stage
sectional CT, and FDG-PET each may play a role in nodal and number of metastatic lymph nodes. The use of FDG-­
staging for newly diagnosed gastric cancer. PET may be helpful in individualized cases when CT and
Endoscopic ultrasound is an effective diagnostic modality EUS are equivocal [30].
for the assessment of depth of mural invasion of the primary
tumor (T stage), more so than cross-sectional imaging, and is
therefore recommended for routine staging by the National 39.5  undamentals of Surgery for Gastric
F
Comprehensive Cancer Network (NCCN) and European Cancer
Society for Medical Oncology (ESMO) [22, 23]. Its accu-
racy for assessing nodal involvement (N stage), however, is Surgical resection remains a cornerstone of management for
less clear. In a large meta-analysis including 44 individual clinically localized gastric cancer. The goals of surgery are
studies reporting diagnostic accuracy assessment of regional rooted in the biologic behavior of the disease; namely, ade-
nodal disease in over 3500 patients, sensitivity and specific- quate assessment for occult metastatic disease, excision of
ity of EUS were determined to be 83% and 67%, respectively the primary tumor with adequate margins, and adequate
[24]. Importantly, as EUS is a highly user-dependent tech- lymphadenectomy, given the propensity for early regional
nique due to both endoscopic capability and radiographic spread. In patients with an inherited predisposition for the
interpretation, an appreciable amount of inter-study hetero- development of gastric cancer, mitigation of risk for clini-
geneity was observed in this meta-analysis. Diagnostic accu- cally significant cancer is an additional goal.
39 The Role of Surgery in Management of Gastric Cancer 425

39.5.1 Role of Staging Laparoscopy a Billroth II or a Roux-en-Y configuration. More proximal

tumors generally require a total gastrectomy with Roux-­
Several modalities are useful in the proper staging of a newly en-­ Y esophagojejunostomy. Proximal gastrectomies with
diagnosed gastric cancer, including endoscopy with biopsy esophagogastric reconstructions have been advocated in the
for molecular classification, EUS, cross-sectional CT of the East, but in general, poor functional outcomes have limited
chest, abdomen, and pelvis, and PET scan. While these stud- the adoption of this surgical technique in the West.
ies are complementary to one another and provide a thor- Optimal management of GEJ tumors remains controver-
ough locoregional and metastatic evaluation, one limitation sial given overlapping but distinct recommendations for sur-
is the possibility of failing to detect occult peritoneal disease. gical and multimodality approaches in pure esophageal and
Laparoscopy has emerged as an effective remedy to this limi- gastric cancers. Therefore, when approaching these tumors,
tation. Peritoneal carcinomatosis that is occult on cross-­ careful attention to endoscopy is critical in order to attempt
sectional imaging can be found and biopsied at the time of to identify the epicenter of the cancer. Siewert type I and II
laparoscopy. Additionally, peritoneal washings may be per- tumors are more often treated as esophageal cancers, while
formed for cytologic examination. Presence of malignant type III lesions are more often treated as gastric cancers. In
cells on cytology constitutes M1 disease and generally pre- the latter, total gastrectomy is often warranted, usually with
cludes surgical resection for curative intent. For this reason, en bloc resection of the distal esophagus in order to obtain an
laparoscopy is best performed as a part of the staging workup adequate margin. Frozen section pathology may be obtained
and before initiation of systemic therapy. Generally, the yield on the proximal margin in order to ensure adequate clear-
of diagnostic laparoscopy—in terms of detecting previously ance. In cases where proximal clearance cannot be achieved
unknown peritoneal disease—increases with clinical T stage without more extensive esophagectomy, a colonic interposi-
(typically determined on EUS). The yield of laparoscopy has tion or jejunal conduit may be considered in order to obtain
been reported to be up to 36%, with approximately 20% with enteral continuity. However, one should carefully balance
macroscopic peritoneal disease seen on laparoscopy and an the likelihood of a curative operation for such poor disease
additional 10–15% with microscopic disease found on cyto- biology with the higher morbidity of such an approach.
logic washings [31, 32]. The NCCN currently recommends Minimally invasive resections utilizing laparoscopic and
consideration of laparoscopy with peritoneal cytology for robotic techniques are being increasingly performed.
lesions that are thought to be T1b or higher. Minimally invasive gastrectomy—as compared to open—
appears to be associated with decreased length of hospital-
ization without compromise of the R0 resection rate or
39.5.2 Extent of Gastric Resection immediate or overall survival [35]. In an analysis using the
National Cancer Data Base, comparing robotic and laparo-
Curative intent resection for gastric adenocarcinoma is gen- scopic gastrectomy, the average number of lymph nodes
erally felt to require ample tissue resection to obtain a nega- sampled was higher among patients undergoing robotic
tive microscopic margin. Classically, a gross margin of resection (19.6 vs. 17.4 lymph nodes, respectively).
approximately 5 cm was required to ensure adequate clear- Furthermore, the proportion of resections in which at least
ance. This requirement seems to stand for early gastric can- 15 lymph nodes were removed was higher in the robotic
cers but not for more advanced disease wherein the risk of group compared with the laparoscopic group (64% vs. 58%,
systemic disease drives survival. For example, an analysis by respectively; [36]). This may be related to improved visual-
the US Gastric Cancer Collaborative, focusing on distal gas- ization during surgery or maybe an epiphenomenon of care
tric cancers, demonstrated 3.1–5.0 cm margins to be associ- at expert centers.
ated with the same favorable overall survival as >5.0 cm
margins for stage I disease, whereas T stage and nodal
involvement dictated overall survival and not margins 39.5.3 Extent of Lymphadenectomy
>5.0 cm for stage II–III disease [33]. In an analysis from
Memorial Sloan-Kettering, focusing on cancers of the GEJ, A focus of surgical therapy for gastric cancer is the perfor-
a margin length of 3.8 cm ex vivo was identified as a target mance of adequate lymphadenectomy given the propensity
for resection [34]. It should also be noted that the rate of for lymphatic spread early in the disease process. The extent
incomplete resection remains appreciable even in the clinical of lymphadenectomy required during gastrectomy has been
trial setting as illustrated in the recent FLOT4 trial (R0 resec- a topic of great study and debate, which is a consequence of
tion rate of 78% and 85% by arm, to be discussed further an evolution of balancing aggressiveness of lymph node dis-
subsequently). Distal tumors are generally thought to be section with perioperative morbidity and mortality. Extent of
adequately resected with distal, or subtotal, gastrectomy. lymphadenectomy is denoted with the prefix letter D and
Reconstruction may be performed with gastrojejunostomy in may be defined as D0, D1, or D2. In general, D1 and D2
426 A. J. Sinnamon et al.

resections correlate with the removal of N1 and N2 lymph mortality. The results from these trials initially shed doubt on
node basins, respectively. These lymphadenectomies vary the routine performance of D2 lymphadenectomy for gastric
according to the extent of gastric resection being performed cancer.
(e.g., total vs. distal gastrectomy). A more aggressive D3 However, long-term follow-up data from these studies
resection has historically been proposed and prospectively suggested oncologic benefit with extended lymphadenec-
studied in Japanese literature. However, it is no longer tomy. In the Dutch trial, gastric cancer-related death was sig-
defined by the NCCN and is generally not currently part of nificantly higher in the D1 group than the D2 group at
the framework for surgical management. 15 years of follow-up (48% vs. 37%, respectively; [39]).
A D1 lymphadenectomy for a total gastrectomy is defined Similarly, local recurrence rates were higher with D1 com-
by the NCCN as the removal of the nodes of the greater and pared to D2 resections (22% vs. 12%, respectively).
lesser omenta, which consist of the nodes of the right and left Importantly, surgical techniques improved during this time-
cardia, greater and lesser curvature, and supra- and infrapy- frame allowing for safer spleen-preserving methods of gas-
loric areas. The Japanese classification of a D1 resection trectomy with D2 lymphadenectomy. Increased postoperative
additionally includes resection of the nodes of the left gastric morbidity and mortality from the MRC and Dutch trials have
artery (station 7 nodes for a complete removal of stations been attributed to the performance of splenectomy and distal
1–7). A D1 resection during a distal gastrectomy may omit pancreatectomy in posttrial analyses, and more modern
removal of the left paracardial nodes (station 2). A D2 resec- spleen- and pancreas-preserving resections should function
tion is defined by the NCCN as a D1 resection with addi- to mitigate this increase [40]. To this point, the authors of the
tional removal of the nodes of the left gastric, common Dutch trial ultimately felt their results supported the use of
hepatic, celiac, and splenic arteries, which corresponds to the D2 lymphadenectomy despite initial trial results.
Japanese classification of a D2 resection as removal of sta- These trials suggest a modest advantage of D2 lymphad-
tions 1–9, 11, and 12. Again, for a distal gastrectomy, left enectomy given that an extended length of follow-up—up to
paracardial nodes (station 2) may be omitted. A D3 lymph- 15 years—was required to overcome the immediate
adenectomy includes peripancreatic, superior mesenteric, increased surgical morbidity and mortality. Importantly, the
mesocolic, and para-aortic nodes. increased surgical morbidity and mortality of D2 resection
The debate regarding the importance of D2 versus D1 are significantly lower in modern series than in these early
lymphadenectomy for gastric cancer must be considered in trials. This was demonstrated by the Italian Gastric Cancer
historical context. Japanese series published in the 1970s and Study Group, which randomized patients to D1 or D2 resec-
1980s, which reported outcomes following D2 resections, tion in specialized centers and demonstrated no significant
generally demonstrated favorable oncologic outcomes with differences in immediate postoperative morbidity (12% vs.
acceptable rates of morbidity and mortality. However, at that 18%, respectively) or 30-day mortality (3.0% vs. 2.2%,
time, randomized prospective studies comparing this respectively; [41]).
approach to D1 lymphadenectomy had not been performed. Extended lymphadenectomy beyond D2, including the
Western surgeon investigators sought to perform such trials paraaortic nodes (D3 resection), has also been prospectively
in the 1980s. Initial results from these endeavors did not sup- studied and shown to be of no benefit. The large Japan
port D2 lymphadenectomy. In the UK Medical Research Clinical Oncology Group (JCOG) study 9501 randomly
Council (MRC) Gastric Cancer Surgical Trial, patients were assigned patients undergoing curative intent gastrectomy to
randomized to undergo gastric resection with D1 versus D2 either D2 or D3 lymphadenectomy [42]. Postoperative mor-
lymphadenectomy, with the D2 group having significantly bidity was higher in the D3 group versus D2 group (28% vs.
higher and unacceptably high overall postoperative mortality 21%, respectively), with no difference in operative mortality.
[37]. Importantly, rates of distal pancreatectomy and sple- Importantly, at 5 years, recurrence-free survival and overall
nectomy were high among the D2 resection group, and in survival were no differences between the D2 and D3 groups.
subsequent analysis, these combined operations were felt to This data would suggest D3 lymphadenectomy carries little
be the primary cause of increased mortality compared to the or no oncologic benefit and is associated with a higher com-
more limited D1 resection. Contemporaneously, Bonenkamp plication rate. A smaller randomized study by Yonemura and
and colleagues similarly randomized patients to undergo colleagues produced similar results [43].
gastrectomy with D1 versus D2 lymphadenectomy in what is Accounting for the data produced in the aforementioned
known as the Dutch trial [38]. Again, the D2 group had a studies, the NCCN currently recommends D1 lymphad-
significantly higher rate of complications and consequent enectomy plus the resection of the associated named vessel
mortality (10% vs. 4%) when compared to the D1 group. lymph nodes or what could be considered a modified D2
Rates of pancreatectomy and splenectomy were again high resection, with a goal of examining at least 16 lymph nodes.
in the D2 arm (30% and 37%, respectively). Splenectomy The NCCN emphasizes that D2 lymphadenectomy should be
was once again found to be associated with a higher rate of performed by experienced surgeons in high-volume centers.
39 The Role of Surgery in Management of Gastric Cancer 427

Furthermore, the NCCN recommends against routine distal juvant and adjuvant settings; however, the optimal regimen
pancreatectomy and cautions against splenectomy except in remains unclear. Furthermore, choice of therapy should be
instances with splenic tumor involvement or extensive hilar tailored to the individual patient after multidisciplinary dis-
lymphadenopathy. cussion. Several landmark trials have established effective
Individuals found to be carriers of a pathogenic CDH1 neoadjuvant, adjuvant, and perioperative regimens.
variant should be considered for prophylactic total gastrec- The use of perioperative chemotherapy was established
tomy given their substantial lifetime risk of developing gas- with the Medical Research Council Adjuvant Gastric
tric cancer. Goals of surgical therapy in this setting are to Infusional Chemotherapy—or MAGIC—trial [47]. In this
remove all gastric mucosa, which may involve intraoperative landmark trial, patients with potentially curable at least clini-
pathologic assessment of proximal and distal resection mar- cal stage II gastric, distal esophageal, or GEJ adenocarcino-
gins to ensure presence of esophageal and duodenal tissue, mas were randomly assigned to undergo surgery alone or
respectively. As gastrectomy in this setting is prophylactic, a surgery with perioperative (both preoperative and postopera-
D2 lymphadenectomy is not required. Endoscopy should be tive) chemotherapy. The regimen used was epirubicin, cispl-
performed preoperatively to assess for the presence of a clin- atin, and fluorouracil (ECF), with three cycles administered
ically significant cancer, as this would change management before surgery and three given after. The majority of the
from a prophylactic approach to a therapeutic one. The patients (~74%) had true gastric cancers. Overall survival
NCCN currently recommends that gastrectomy be offered to was demonstrated to be superior for the patients undergoing
these patients after the age of 18 years. Effective surveillance perioperative therapy compared with surgery alone (36% vs.
in these patients can be challenging but may be considered 23%, respectively, at 5 years). This was despite the fact that
selectively [44]. only 42% of the patients in the perioperative chemotherapy
arm completed the full regimen, underscoring the frequent
difficulties in completing postoperative therapy in practice.
39.5.4 Role of Sentinel Lymph Node Biopsy The chemotherapy regimen in the MAGIC trial (ECF)
was subsequently compared to a regimen of docetaxel, oxali-
While a “standard oncologic lymphadenectomy” is currently platin, and leucovorin/fluorouracil (FLOT) given in four
recommended during gastrectomy for gastric cancer, the cycles preoperatively and four cycles postoperatively
potential role of sentinel lymph node biopsy to specifically (“FLOT4”; [48]). In this trial, the ECF arm also included
identify draining lymph nodes from a gastric tumor has been patients treated with oral capecitabine in place of infusional
studied. Sentinel lymph node biopsy could play a role for fluorouracil (ECX). Approximately one-quarter of these
selective lymphadenectomy when performing endoscopic or patients had Siewert type 1 lesions and were treated with
local surgical resection of early gastric cancer to determine transthoracic esophagectomy. Median overall survival was
the need for formal gastrectomy with regional lymphadenec- longer among patients undergoing FLOT perioperative ther-
tomy. A multicenter Japanese trial included 397 patients with apy than for ECF/ECX (50 months vs. 35 months, respec-
early gastric cancers with performance of sentinel lymph tively). The FLOT regimen currently is preferred by the
node biopsy using both preoperative endoscopic injection of NCCN for the use of perioperative therapy in medically fit
radiolabeled technetium and intraoperative injection of blue patients who are felt to be able to tolerate it.
dye [45]. The sentinel lymph nodes were positive in 93% (53 Adjuvant chemotherapy is also an option for gastric can-
of 57) of patients with involved lymph nodes, and the overall cer in addition to perioperative therapy. Multiple regimens
accuracy of sentinel lymph node staging was 99%. Mueller are acceptable. The use of oral capecitabine and oxaliplatin
and colleagues have shown promising preliminary results in (CAPOX) was established in the CLASSIC trial
the West, also using a dual tracer technique to identify the (“capecitabine and oxaliplatin adjuvant study in stomach
sentinel lymph nodes [46]. cancer”), where patients who underwent curative gastrec-
tomy with D2 lymphadenectomy for stages II–IIIB tumors
were randomized to eight cycles of CAPOX versus surgery
39.6 Adjuncts to Surgery only. In the initial trial report, receipt of CAPOX was associ-
and Multimodality Therapy ated with improved progression-free survival [49]. With lon-
for Gastric Cancer ger follow-up, improved overall survival was also seen with
the adjuvant treatment compared with surgery alone (78%
Given the propensity of gastric cancer to metastasize, vs. 69%, respectively; [50]). In Japan and Europe, S-1 is an
adjuncts to surgery to treat and prevent local, regional, and option for adjuvant therapy following gastrectomy with D2
distant recurrences play an important role in management. lymphadenectomy for patients with stage II/III disease. This
Use of systemic therapy, radiotherapy, and the combination agent, not available in the United States, was shown to be
of the two have been studied prospectively in both the neoad-
428 A. J. Sinnamon et al.

associated with improved overall survival at 5 years when tomy is currently recommended. Multimodal therapy for
compared to surgery alone in the ACTS-GC trial [51]. clinical stage II–III disease has gained traction and is sup-
The use of adjuvant chemoradiotherapy following resec- ported by several prospective randomized trials, with periop-
tion of gastric and GEJ cancers was established with the erative chemotherapy, adjuvant chemotherapy, and adjuvant
Intergroup 0116 trial [52]. Patients were randomized to sur- chemoradiotherapy all accepted regimens.
gery alone or a course of fluorouracil-based chemoradiation.
Overall survival and relapse-free survival were both
improved with chemoradiotherapy. However, a major criti- Future Directions
cism of this trial is that surgical resections included fairly • How will individualized treatment paradigms
limited lymphadenectomies, with only 10% of patients evolve as the understanding of molecular classifica-
undergoing a D2 resection, and over half of the patients tion improves?
undergoing a D0 resection. Several subsequent trials sought • What is the role of adjuvant chemoradiotherapy
to answer whether combined chemoradiotherapy in the adju- relative to adjuvant chemotherapy alone in the set-
vant setting provides benefit over chemotherapy alone, ting of an adequate surgical lymphadenectomy?
including the ARTIST, ARTIST 2, and CRITICS trials [53– • What is the preferred systemic regimen for HER2-­
55]. These trials included more extensive surgical lymphad- overexpressing tumors and what is the best sequence
enectomy than the Intergroup trial and thus may be more of therapy relative to surgery?
appropriate to assess whether an adjuvant locoregional ther- • What is the role of sentinel lymph node biopsy for
apy (radiation) provides benefit following adequate lymph- selective lymphadenectomy in the setting of endo-
adenectomy. However, none of these trials showed benefit of scopic or local resections for early gastric cancers?
chemoradiotherapy over chemotherapy alone. It therefore • Do immunotherapies using PD1-PDL1 inhibitors
remains an open question of whether chemoradiotherapy provide benefit in the perioperative and adjuvant
provides a significant benefit over chemotherapy alone in the setting?
adjuvant setting following an adequate clearance of regional
lymph nodes with surgery.
The addition of therapy targeting the HER2 receptor (i.e., Authorship Disclosures None.
trastuzumab, pertuzumab) or inhibiting the PD1–PDL1
immune checkpoint (i.e., pembrolizumab, nivolumab) to
systemic therapy regimens is currently under investigation. References
In a recent randomized phase II trial, the addition of trastu-
zumab and pertuzumab to perioperative FLOT therapy for 1. Lauren P. The two histological main types of gastric carcinoma: dif-
HER2 overexpressing gastric cancers significantly improved fuse and so-called intestinal-type carcinoma. An attempt at a histo-­
pathologic complete response rates when compared to peri- clinical classification. Acta Pathol Microbiol Scand. 1965;64(31)
2. Berx G, Becker KF, Hofler H, van Roy F. Mutations of the human
operative FLOT only (35% vs. 12%, respectively) and E-cadherin (CDH1) gene. Hum Mutat. 1998;12:226.
pathologic nodal negativity rates (68% vs. 39%, respec- 3. Graziano F, Humar B, Guilford P. The role of the E-cadherin gene
tively) in the resected surgical specimens [56]. Several other (CDH1) in diffuse gastric cancer susceptibility: from the laboratory
trials are currently underway investigating the addition of to clinical practice. Ann Oncol. 2003;14(12):1705.
4. Hansford S, et al. Hereditary diffuse gastric cancer syndrome:
HER2-­directed therapy to systemic regimens for potentially CDH1 mutations and beyond. JAMA Oncol. 2015;1(1):23.
resectable gastric cancers with HER2 overexpression. 5. van der Post RS, et al. Hereditary diffuse gastric cancer: updated
Similarly, while PD1–PDL1 inhibitor immunotherapies are clinical guidelines with an emphasis on germline CDH1 mutation
approved for use in the metastatic setting, their role in the carriers. J Med Genet. 2015;52(6):361–74.
6. Plummer M, et al. Global burden of gastric cancer attributable to
perioperative and adjuvant setting is currently under Helicobacter pylori. Int J Cancer. 2015;136(2):487.
investigation. 7. Correa P. Human gastric carcinogenesis: a multistep and mul-
tifactorial process—First American Cancer Society Award
Lecture on Cancer Epidemiology and Prevention. Cancer Res.
1992;52(24):6735.
39.7 Conclusions 8. Cancer Genome Atlas Research Network. Comprehensive
molecular characterization of gastric adenocarcinoma. Nature.
Gastric cancer is an aggressive entity with early spread to 2014;513(7517):202–9.
lymph nodes. Surgical resection with an adequate lymphad- 9. Bang YJ, et al. Trastuzumab in combination with chemother-
apy versus chemotherapy alone for treatment of HER2-positive
enectomy is a cornerstone of therapy. Accounting for long-­ advanced gastric or gastro-oesophageal junction cancer (ToGA):
term follow-up data in the landmark surgical a phase 3, open-label, randomised controlled trial. Lancet.
lymphadenectomy trials, a D2 or modified D2 lymphadenec- 2010;376(9742):687.
39 The Role of Surgery in Management of Gastric Cancer 429

10. Dong H, et al. B7-H1, a third member of the B7 family, co-­ 32. Ikoma N, et al. Yield of staging laparoscopy and lavage cytology
stimulates T-cell proliferation and interleu-kin-10 secretion. Nat for radiologically occult peritoneal carcinomatosis of gastric can-
Med. 1999;5:1365–9. cer. Ann Surg Oncol. 2016;23(13):4332–7.
11. Topalian SL, Drake CG, Pardoll DM. Targeting the PD-1/ 33. Squires MH 3rd, et al. Is it time to abandon the 5-cm margin
B7-H1(PD-L1) pathway to activate anti-tumor immunity. Curr rule during resection of distal gastric adenocarcinoma? A multi-­
Opin Immunol. 2012;24:207–12. institution study of the U.S. Gastric Cancer Collaborative. Ann
12. Amin MB et al. (Eds.). AJCC cancer staging manual, 8th edn. Surg Oncol. 2015;22(4):1243–51.
Springer International Publishing: American Joint Commission on 34. Barbour AP, et al. Adenocarcinoma of the gastroesophageal junc-
Cancer; 2017. tion: influence of esophageal resection margin and operative
13. Maruyama K, et al. Lymph node metastases of gastric cancer. approach on outcome. Ann Surg. 2007;246(1):1–8.
General pattern in 1931 patients. Ann Surg. 1989;210(5):596–602. 35. Sweigert PJ, et al. Total gastrectomy in patients with gastric
14. Zuo CH, et al. Characterization of lymph node metastasis and its adenocarcinoma: is there an advantage to the minimally invasive
clinical significance in the surgical treatment of gastric cancer. Mol approach? Surgery. 2019;166(4):623–31.
Clin Oncol. 2014;2(5):821–6. 36. Ryan S, et al. Robotic versus laparoscopic gastrectomy for gas-
15. Tajima Y, et al. Risk factors for lymph node metastasis from tric adenocarcinoma: propensity-matched analysis. Surg Innov.
gastric cancers with submucosal invasion. Ann Surg Oncol. 2020;27(1):26–31.
2010;17(6):1597–604. 37. Cuschieri A et al. Postoperative morbidity and mortality after D1
16. Ohashi S, Okamura S, Urano F, Maeda M. Clinicopathological and D2 resections for gastric cancer: preliminary results of the MRC
variables associated with lymph node metastasis in submucosal randomised controlled surgical trial. The Surgical Cooperative
invasive gastric cancer. Gastric Cancer. 2007;10(4):241–50. Group. Lancet 1996;347(9007):995–9.
17. Siewert JR, Stein HJ. Classification of adenocarcinoma of the 38. Bonenkamp JJ, et al. Extended lymph-node dissection for gastric
oesophagogastric junction. Br J Surg. 1998;85(11):1457–9. cancer. N Engl J Med. 1999;340:908–14.
18. Siewert JR, Feith M, Werner M, Stein HJ. Adenocarcinoma of the 39. Songun I, et al. Surgical treatment of gastric cancer: 15-year fol-
esophagogastric junction: results of surgical therapy based on ana- low-­up results of the randomised nationwide Dutch D1D2 trial.
tomical/topographic classification in 1,002 consecutive patients. Lancet Oncol. 2010;11:439–49.
Ann Surg. 2000;232(3):353–61. 40. Cuschieri A, et al. Patient survival after D1 and D2 resections for
19. Nigro JJ, et al. Prevalence and location of nodal metastases in distal gastric cancer: long-term results of the MRC randomized surgical
esophageal adenocarcinoma confined to the wall: implications for trial. Br J Cancer. 1999;79(9–10):1522–30.
therapy. J Thorac Cardiovasc Surg. 1999;117(1):16–23. 41. Degiuli M, et al. Morbidity and mortality in the Italian Gastric
20. Alexander JS, Ganta VC, Jordan PA, Witte MH. Gastrointestinal Cancer Study Group randomized clinical trial of D1 versus D2
lymphatics in health and disease. Pathophysiology. resection for gastric cancer. Br J Surg. 2010;97(5):643–9.
2010;17(4):315–35. 42. Sasako M, et al. D2 lymphadenectomy alone or with para-­
21. Japanese Gastric Cancer Association. Japanese gastric cancer aortic nodal dissection for gastric cancer. N Engl J Med.
treatment guidelines 2018, (5th edition). Gastric Cancer. 2020;24: 2008;359(5):453–62.
1–21. 43. Yonemura Y, et al. Randomized clinical trial of D2 and extended
22. National Comprehensive Cancer Network (NCCN). NCCN clini- paraaortic lymphadenectomy in patients with gastric cancer. Int J
cal practice guidelines in oncology; n.d.. https://www.nccn.org/ Clin Oncol. 2008;13(2):132–7.
professionals/physician_gls. 44. Blair VR, et al. Hereditary diffuse gastric cancer: updated clinical
23. Smyth EC et al; ESMO Guidelines Committee. Gastric cancer: practice guidelines. Lancet Oncol. 2020;21(8):e386–97.
ESMO clinical practice guidelines for diagnosis, treatment and 45. Kitagawa Y, et al. Sentinel node mapping for gastric can-
follow-up. Ann Oncol. 2016;27(suppl 5):v38–v49. cer: a prospective multicenter trial in Japan. J Clin Oncol.
24. Mocellin S, Pasquali S. Diagnostic accuracy of endoscopic ultra- 2013;31(29):3704–10.
sonography (EUS) for the preoperative locoregional staging 46. Mueller CL, et al. Sentinel lymph node sampling for early gastric
of primary gastric cancer. Cochrane Database Syst Rev. 2015, cancer-preliminary results of a north American prospective study. J
2015;(2):CD009944. Gastrointest Surg. 2019;23(6):1113–21.
25. Chang KJ, et al. Endoscopic ultrasound-guided fine-needle aspira- 47. Cunningham D, et al. Perioperative chemotherapy versus sur-
tion. Gastrointest Endosc. 1994;40(6):694. gery alone for resectable gastroesophageal cancer. N Engl J Med.
26. Pollack BJ, Chak A, Sivak MV Jr. Endoscopic ultrasonography. 2006;355(1):11–20.
Semin Oncol. 1996;23(3):336. 48. Al-Batran SE, et al. Perioperative chemotherapy with fluoro-
27. D'Elia F, Zingarelli A, Palli D, Grani M. Hydro-dynamic CT uracil plus leucovorin, oxaliplatin, and docetaxel versus fluoro-
preoperative staging of gastric cancer: correlation with patho- uracil or capecitabine plus cisplatin and epirubicin for locally
logical findings. A prospective study of 107 cases. Eur Radiol. advanced, resectable gastric or gastro-oesophageal junction
2000;10(12):1877. adenocarcinoma (FLOT4): a randomised, phase 2/3 trial. Lancet.
28. Sohn KM, et al. Comparing MR imaging and CT in the staging of 2019;393(10184):1948–57.
gastric carcinoma. AJR Am J Roentgenol. 2000;174(6):1551. 49. Bang YJ, et al. Adjuvant capecitabine and oxaliplatin for gas-
29. Yan C, et al. Value of multidetector-row computed tomography in tric cancer after D2 gastrectomy (CLASSIC): a phase 3 open-
the preoperative T and N staging of gastric carcinoma: a large-scale label, randomised controlled trial. Lancet. 2012;379(9813):
Chinese study. J Surg Oncol. 2009;100(3):205. 315–21.
30. Yun M, Lim JS, Noh SH, Hyung WJ, Cheong JH, Bong JK, Cho 50. Noh SH, et al. Adjuvant capecitabine plus oxaliplatin for gas-
A, Lee JD. Lymph node staging of gastric cancer using (18) tric cancer after D2 gastrectomy (CLASSIC): 5-year follow-
F-FDG PET: a comparison study with CT. J Nucl Med. 2005 up of an open-label, randomised phase 3 trial. Lancet Oncol.
­
Oct;46(10):1582–8. 2014;15(12):1389–96.
31. Bentrem D, et al. The value of peritoneal cytology as a preoperative 51. Sakuramoto S, et al. Adjuvant chemotherapy for gastric cancer
predictor in patients with gastric carcinoma undergoing a curative with S-1, an oral fluoropyrimidine. N Engl J Med. 2007;357(18):
resection. Ann Surg Oncol. 2005;12(5):347–53. 1810–20.
430 A. J. Sinnamon et al.

52. Macdonald JS, et al. Chemoradiotherapy after surgery compared gastric cancer after D2 resection: the ARTIST 2 trial. Ann Oncol.
with surgery alone for adenocarcinoma of the stomach or gastro- 2021;32(3):368–74.
esophageal junction. N Engl J Med. 2001;345(10):725–30. 55. Cats A, et al. Chemotherapy versus chemoradiotherapy after sur-
53. Lee J, et al. Phase III trial comparing capecitabine plus cis- gery and preoperative chemotherapy for resectable gastric cancer
platin versus capecitabine plus cisplatin with concurrent (CRITICS): an international, open-label, randomised phase 3 trial.
capecitabine radiotherapy in completely resected gastric cancer Lancet Oncol. 2018;19(5):616–28.
with D2 lymph node dissection: the ARTIST trial. J Clin Oncol. 56. Hofheinz RD, et al. Perioperative trastuzumab and pertuzumab in
2012;30(3):268–73. combination with FLOT versus FLOT alone for HER2-positive
54. Park SH, et al. A randomized phase III trial comparing adjuvant resectable esophagogastric adenocarcinoma: final results of the
single-agent S1, S-1 with oxaliplatin, and postoperative chemo- PETRARCA multicenter randomized phase II trial of the AIO. J
radiation with S-1 and oxaliplatin in patients with node-positive Clin Oncol. 2020;38S:ASCO #4502.
 The Role of Surgery in Managing
Primary and Metastatic 40
Hepatopancreaticobiliary Cancers

Lyonell B. Kone, Christopher Javadi, Jessica M. Keilson,

Shishir K. Maithel, George Poultsides, and Ajay V. Maker

Abstract 40.1 Introduction

This chapter focuses on the effectiveness of surgery in the
management of primary and metastatic hepatopancreati- Hepatopancreaticobiliary (HPB) cancers are defined by their
cobiliary (HPB) cancers. Tumor biology is reviewed, as is complex and aggressive tumor biology. Our understanding
the role of lymphovascular spread and how surgery is tai- of these inherent properties dictates surgical management as
lored to treat each HPB cancer. The importance of surgi- well as the need for adjuvant therapies to control disease pro-
cal intervention to achieve long-term survival is addressed gression and limit recurrence. In pancreaticobiliary cancers,
in the context of its significant limitations to control the complexities of lymphovascular invasion (LVI) highlight
microscopic disease and metastasis. Recent and upcom- both the need for aggressive surgical management and its
ing adjunctive therapies hold great promise for the man- limitations. As locoregional spread occurs via lymphatics for
agement of HPB patients and are also addressed in the several malignancies, LVI may serve as a window into the
context of increasing eligibility and efficacy of surgical dissemination of disease beyond its origin. In colorectal liver
intervention for these complex cancers. metastasis and hepatocellular carcinoma, tumor-specific
molecular and genetic subclassification, as well as patient-­
specific factors such as organ dysfunction and functional sta-
Learning Objectives
tus, guide our surgical approach. Surgical resection provides
At the conclusion of this chapter, readers will be able to:
the only curative-intent option for these diseases; however,
the efficacy of these approaches is limited by tumor-intrinsic
• describe the complex tumor biology and lympho-
factors that drive disease aggressiveness, underscoring the
vascular spread of hepatopancreaticobiliary cancers
need for adjunctive therapies.
• identify the benefits and limitations of surgical ther-
Below we will discuss, in a site-specific manner, tumor
apy for hepatopancreaticobiliary cancers
biology and disease progression, surgical approaches and
• recognize the benefits of adjunctive therapies in man-
their limitations, and the role of adjunctive therapies in man-
aging microscopic and distant metastatic disease in
aging disease progression and recurrence.
the setting of hepatopancreaticobiliary cancers

L. B. Kone 40.2 Gallbladder carcinoma

University of Illinois at Chicago—Metropolitan Group Hospitals,
Chicago, IL, USA
Gallbladder carcinoma (GBC) is a rare disease most often
e-mail: [email protected]
detected at advanced stages, with an estimated 5-year sur-
C. Javadi · G. Poultsides
vival rate of 5–13%. The gallbladder is drained by a plexus
Stanford University, Stanford, CA, USA
e-mail: [email protected]; [email protected] of lymphatic vessels within the subserosa that runs along the
anterior and posterior surfaces, following a similar distribu-
J. M. Keilson · S. K. Maithel
Emory University, Atlanta, GA, USA tion to the arterial system and ultimately emptying into lym-
e-mail: [email protected]; [email protected] phatic collecting ducts located on the inferior surface of the
A. V. Maker (*) gallbladder. Ito and colleagues describe the lymphatic drain-
Department of Surgery, University of California San Francisco, age of the gallbladder as following three distinct routes: (1)
San Francisco, CA, USA the cholecystoretropancreatic pathway, running along the
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 431
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_40
432 L. B. Kone et al.

anterior and posterior surfaces of the common bile duct to improved stage stratification and disease-specific survival in
converge at the retroportal nodes; (2) the cholecystoceliac node-positive patients to 33 months with a total LN count ≥6
pathway, through the hepatoduodenal ligament to the celiac compared to 15 months for a total LN count <6 [8]. However,
nodes; and (3) the cholecystomesenteric pathway, running the use of absolute LN count introduces a certain degree of
anterior to the portal vein toward the posterior pancreatico- bias as it relies on adequate LN retrieval during surgery and
duodenal nodes, all of which ultimately converge at the aor- accurate histolopathologic examination. To mitigate this
tocaval and paraaortic nodes [1]. Lymph also drains directly bias, some investigators have suggested the utilization of LN
into the lymphatics of the liver through pathways within the ratio to more accurately reflect disease burden and to better
gallbladder fossa. As surgical resection confers the only prognosticate following curative resection [9]. More recent
chance of cure in GBC, radical resection is recommended, studies have proposed a ratio of the number of positive LNs
including regional lymphadenectomy dictated by lympho- to the number of negative LNs, the log odds of positive LNs,
vascular drainage patterns. Hepatic resection of segments as a more reliable marker of outcome than either LN ratio or
IVB and V is indicated for lesions with subserosal invasion AJCC N-stage [10].
[2], whereas simple cholecystectomy is sufficient for the Indeed, LVI and LN positivity strongly correlate with sur-
treatment of T1a disease [3]. vival outcomes; however, the role of lymphadenectomy
T stage has a strong correlation with LN positivity. Up to extends beyond prognostication and may provide therapeutic
20% of patients with T1b disease are found to be lymph node benefits. In patients with early-stage disease, where 24% of
(LN) positive, underscoring the importance of completing T1b and 45% of T2 lesions demonstrate LN involvement,
regional lymphadenectomy at the time of resection even in radical liver resection or cholecystectomy alone with LN
the setting of superficial disease. While the value of resection evaluation improves cancer-specific and overall survival
in patients with known LN metastases is debatable, up to over resection of the primary tumor alone [11]. This same
20% 5-year survival rate is possible in patients with LN finding is seen in patients with T3N0 disease; however, T3N1
metastases following resection [4]. Regional and paraaortic and T4 disease do not demonstrate improved outcomes with
lymphadenectomy does not improve outcomes in patients radical resection [12]. Thus, while LN evaluation is a requi-
with the paraaortic disease and confers survival equivalent to site factor in disease staging and prognosis, lymphadenec-
that of distant metastatic disease [5]. However, among tomy may provide therapeutic value for earlier stage
patients with regional N1 or N2 disease who underwent sys- disease.
tematic lymphadenectomy, survival was equivalent to
patients without nodal disease. Thus, resection remains stan-
dard of care for patients with locoregional disease. 40.3 Cholangiocarcinoma
In the setting of incidental GBC identified following cho-
lecystectomy, T-stage has been largely associated with the Cholangiocarcinoma (CCA) is a rare malignancy of the bili-
presence of residual disease and thus decreased survival. ary tract that also carries a poor prognosis and high mortality.
However, prognostic risk scores incorporating LVI have Complete resection is the standard of care and remains the
demonstrated superior prediction of locoregional and distant only potentially curative approach in highly selected patients.
disease compared to T-stage alone [6]. Therefore, the pres- As in GBC, LN-positive disease is associated with worse
ence of LVI detected after the index cholecystectomy may survival outcomes. This is reflected in the AJCC staging sys-
potentially guide the use of preoperative therapy prior to re-­ tem where LN-positive disease was considered stage IV in
resection for incidental GBC. the seventh edition and is currently stage IIIb in the eighth
The eighth edition American Joint Committee on Cancer edition.
(AJCC) tumor TNM staging system defines N stage based on Intrahepatic cholangiocarcinoma (ICCA) arises from the
the number of metastatic LNs, an advancement from the sev- bile ducts within the hepatic parenchyma and accounts for
enth edition which stratified node-positive disease based on approximately 20% of all CCA. LN metastasis is the most
the location of involved LNs [7]. Resection and evaluation of significant predictor of poor survival after resection [13],
at least six LNs, including hilar, celiac, periduodenal, peri- with up to 34% of patients with ICCA having LN-positive
pancreatic, superior mesenteric, and posterior pancreatic disease. The lymphatic system of the liver includes a deep
head nodes, improve stage stratification. Distant LN involve- lymphatic system, which is contained within the portal triad
ment, such as paraaortic LNs, is still regarded as stage IV and communicates with the hilar and peripancreatic LNs or
metastatic disease. A number of studies have demonstrated runs along the hepatic venous system toward the inferior
increased survival following regional LN dissection for vena cava to drain into the general lymphatic system, and a
node-positive GBC following R0 resection. Liu et al. showed superficial lymphatic system, which runs within the subsero-
40 The Role of Surgery in Managing Primary and Metastatic Hepatopancreaticobiliary Cancers 433

sal connective tissue of the liver and drains into the general postoperative mortality associated with more extensive
lymphatic system or distant LNs [14]. Tumors originating resections, including vascular en bloc resections or hepatico-
from the right hepatic lobe spread to hilar, periduodenal, and pancreatoduodenectomy, against a reduction in locoregional
peripancreatic LNs, whereas left hepatic lobe tumors typi- recurrence rates [18]. In retrospective series, total hepatec-
cally spread to hilar, inferior phrenic, and gastrohepatic LNs. tomy with orthotopic liver transplant, the most extreme
Though the current AJCC eighth edition offers no recom- option to mitigate the incidence of residual disease, has been
mendation for the optimal number of LNs for accurate stag- associated with improved recurrence-free and overall sur-
ing, a major limitation in LN evaluation for ICCA is that vival compared to resection alone [18, 19]. However, these
portal lymphadenectomy often contains less than three LNs patients are highly selected and receive 6 months of high-­
[15]. Further, up to 13% of patients reported to have LN neg- dose radiation and chemotherapy prior to transplantation,
ative disease following resection have occult portal LN suggesting that decreased recurrence rates and improved sur-
metastases on detailed pathologic reexamination [16]. This vival outcomes may be in part a consequence of patient
suggests that the rate of LN positivity is not only influenced selection and inherent tumor biology, as well as this addi-
by the extent of LN harvest but also by the detail with which tional therapy received prior to transplantation.
LNs are evaluated following resection. As such, several Distal cholangiocarcinoma (DCCA) involves the distal
authors have suggested LVI as an independent prognostic extrahepatic, or intrapancreatic, portion of the bile duct.
factor for reduced survival and a surrogate marker for aggres- Similar to ICCA and HCCA, surgical resection is the only
sive tumor biology that can be utilized in the setting of inad- potentially curative option. As with other biliary malignan-
equate LN sampling or as an indication for adjuvant cies, LN metastasis is a prognostic indicator of decreased
chemotherapy [15]. survival. The frequency of LN-positive disease among
Hilar cholangiocarcinoma (HCCA) accounts for 50–60% patients with DCCA has been reported as high as 68% [20],
of CCA. Lymphatic drainage courses along the common thus assessment of at least 12 LNs is suggested for accurate
hepatic artery to the celiac LNs, the common bile duct toward staging [7]. As surgical resection of DCCA necessitates pan-
the posterior pancreatic head and paraaortic LNs, and the creatoduodenectomy, adequate lymphadenectomy is often
portal vein and superior mesenteric artery to drain into the achieved.
superior mesenteric LNs [17]. Due to its infiltrative growth LVI has also been shown to be independently associated
pattern, extrahepatic bile duct resection for HCCA is often with decreased overall survival in DCCA. Thus, a new
associated with high positive margin rates, which carries a T-stage system that incorporates both tumor size >3 cm and
survival rate similar to that of metastatic disease [15]. presence of LVI has been proposed [21].
Node-positive disease is an independent prognostic factor
for both locoregional and distant recurrence among
HCCA. The incidence of nodal involvement in patients with 40.4  djuvant Therapy for Biliary Tract
A
HCCA following resection is as high as 31–58% [17], and Disease
occurs most commonly around the common bile duct and
portal vein. Retrieval of LNs along the hepatoduodenal liga- Surgical resection is the only potentially curative option for
ment and posterior pancreatic head is considered the stan- biliary tract cancers; however, these cancers are often
dard of care. LN involvement beyond these nodal basins is detected at advanced stages and carry a high risk of recur-
considered M1 metastatic disease and is a contraindication rence, underscoring the need for effective adjuvant
to resection. treatments.
Recurrence rates in patients with resected HCCA are esti- The landmark BILCAP (Adjuvant Capecitabine for
mated to be as high as 76% and have been demonstrated up Biliary Tract Cancer) phase III randomized trial demon-
to 8 years following resection [18]. Locoregional recurrence strated improved overall survival with adjuvant capecitabine
is typically the first site of disease recurrence, resulting from compared to observation alone in a prespecified intention-to-­
positive margins along the proximal or distal bile duct mar- treat analysis adjusting for nodal status, disease grade, and
gins or within the peribiliary tissue. Tumor spread along the sex [22]. Adjuvant capecitabine is therefore recommended
length of the bile duct or within the surrounding tissue often for patients with resected biliary tract cancers for a duration
occurs through the lymphovascular system. As a result, stan- of 6 months. However, gemcitabine plus cisplatin remains an
dard of care includes hemi-hepatectomy dictated by the effective and acceptable adjuvant treatment regimen.
extent of tumor and Bismuth classification in order to clear The SWOG 0809 non-randomized phase II trial of
any tissue with lymphovascular involvement and to increase capecitabine plus gemcitabine followed by capecitabine-­
R0 resection rates. There remains controversy in balancing based chemoradiation demonstrated comparable disease-­
434 L. B. Kone et al.

Table 40.1 Retrospective studies examining the impact on survival of extending pancreatic ductal adenocarcinoma resections to convert positive
margins based on intraoperative frozen section to negative
Median overall P (R1-to-R0
Author Institution Period Margin status N survival (mos) vs. R1)
Hernandez U. of South Florida 1995–2006 R1-to-R0 17 11 0.91
(2009) R1 44 13
Fatima Mayo Clinic 1981–2007 R1-to-R0 57 18 NS
(2010) R1 127 15
Pang (2014) Royal North Shore Hospital, Australia 2007–2012 R1-to-R0 7 16 0.31
R1 67 23
Mathur Florida Hospital Tampa, U. of South Florida 1995–2012 R1-to-R0 40 14 0.19
(2014) R1 110 12
Kooby (2014) Emory, U. of Wisconsin, Northwestern, U. of 2000–2012 R1-to-R0 72 12 0.98
Cincinnati, U. of North Carolina, U. of Louisville, R1 131 14
Vanderbilt, Washington U.

free and median overall survival between R0 and R1 40.5.1 E

 xtending the Resection to Obtain
subgroups as well as tolerability in patients with extrahepatic Negative Margins
CCA and GBC [23]. Based on the 2020 ASCO Clinical
Practice Guidelines for resected biliary tract cancers, patients The status of the surgical margin is an important prognostic
with microscopically positive margins following resection factor for PDAC and is considered positive if the tumor is
for extrahepatic CCA and GBC can be offered chemoradio- within 1 mm of the cut surface. The uncinate margin is at the
therapy [24]. While there remains limited evidence on the highest risk for residual disease for pancreatic head tumors
role of radiotherapy for biliary tract cancers, chemoradiation and consists of the connective tissue separating the uncinate
in this setting may be efficacious in controlling local and dis- process from the superior mesenteric artery (SMA) and the
tant disease recurrence. neural and lymphatic tissue surrounding the celiac trunk
(CA) [28].
Multiple studies have found improved survival in patients
40.5 Pancreatic Adenocarcinoma with microscopically negative tumor margins. However,
multiple retrospective studies have examined the utility of
In the United States, pancreatic ductal adenocarcinoma re-resection of the pancreatic neck to achieve an R0 pancre-
(PDAC) is expected to become the second leading cause of atic margin, revealing no association with survival compared
cancer-related death by 2030 [25]. The 5-year survival over to an R1 resection. Patients with negative initial frozen sec-
the last several decades has improved only modestly, from tions had significantly longer survival than re-resection R0
3% in 1975, to 5% in 2005, and to 10% in 2010. For patients and R1 patients, suggesting that a positive margin is a reflec-
with localized disease, surgery followed by systemic chemo- tion of aggressive tumor biology rather than a driver of dis-
therapy remains the only potentially curative approach. ease progression (Table 40.1).
However, even after a successful, curative-intent operation, Furthermore, the dominant pattern of recurrence, regard-
up to a third of patients with PDAC die within a year from less of margin status, is distant disease. A large series from
surgery [26]. These early recurrences, largely secondary to Belgium found that neither an R1 resection nor a tumor-free
microscopic distant disease at the time of surgery, have margin less than 1 mm resulted in a greater number of locore-
fueled debates on the extent of surgical resection, particu- gional recurrences. In their study, only 8% of patients who
larly in regard to margins and the appropriate extent of recurred had an isolated local recurrence [29]. Collectively,
lymphadenectomy. The biological issue with these tumors is these studies suggest that subclinical systemic disease deter-
that the first tumor driving mutations are often present two mines the survival for the overwhelming majority of PDAC
decades before the radiographic appearance of the tumor patients, not local recurrence.
[27]. Thus by the time surgery is considered, tumor cells are
often already circulating systemically, and the window upon
which to select patients with localized disease has often 40.5.2 Lymph Node Status Predicts Prognosis
passed. Extensive genomic analyses have revealed the exis-
tence of distinct molecular subtypes of pancreatic cancer that Five-year survival after pancreaticoduodenectomy is approx-
may enhance clinical stratification of patients for both per- imately 30% for node-negative patients but only 10% for
sonalized treatments and surgical selection in the future, but patients with nodal involvement, even if only one node is
until this time, surgical treatment consists of pancreatectomy positive [30, 31]. Moreover, an increasing number of positive
with lymphadenectomy. lymph nodes correlates with decreased survival. In a study of
40 The Role of Surgery in Managing Primary and Metastatic Hepatopancreaticobiliary Cancers 435

811 patients who underwent pancreatoduodenectomy for bidity, particularly if the celiac artery and SMA are skeleton-
PDAC between 2001 and 2012, median survival was ized circumferentially, resulting in high rates of postoperative
26.1 months for patients with two to three positive nodes, diarrhea with malnutrition.
21.9 months for cases with four to seven positive nodes, and
18.3 months for patients with greater than eight positive
nodes [32]. 40.5.4 Systemic Therapy for PDAC
An analysis of SEER data found that among pathologi-
cally node-negative patients, there is improved survival asso- The more widespread use of systemic agents, including neo-
ciated with an increasing number of examined nodes [33]. adjuvant and adjuvant chemotherapy, and radiation has
Examination of 15 lymph nodes appears to be optimal for become a major focus of cancer centers treating pancreatic
accurate staging of node-negative adenocarcinoma of the ductal adenocarcinoma. ESPAC-4 found a prolonged sur-
pancreas after pancreaticoduodenectomy. The median sur- vival in patients who received adjuvant gemcitabine and
vival for the group of patients with pN0 disease who had 15 capecitabine, with 5-year survival rates of 30% for all
or more nodes examined was 27 months, whereas the median patients, rising to 40% in those with an R0 resection margin,
survival for the group with fewer than 15 nodes examined and to nearly 50% in node-negative patients [37]. The
was 19 months. Thus, adequate surgery includes a lymphad- PRODIGE 24 trial demonstrated an impressive survival ben-
enectomy, the extent of which has been studied by multiple efit of adjuvant oxaliplatin, irinotecan, fluorouracil, and leu-
groups, although these studies remain steeped in institutional covorin (mFOLFIRINOX), with a median OS reaching
biases. 54.4 months [38]. Other trials, ESPAC-5F and SWOG
S1505, have shown a benefit of neoadjuvant or perioperative
approaches using either mFOLFIRINOX or gemcitabine
40.5.3 Randomized Trials of Extended plus nanoalbumin-bound (nab)-paclitaxel [39, 40].
Lymphadenectomy

Pathologic assessment after extended lymphadenectomy 40.6 Pancreatic Neuroendocrine Tumors

(LAD) revealed that 30–40% of patients with resectable pan-
creatic head tumors had involved lymph nodes outside the Pancreatic neuroendocrine tumors (PNETs) are a rare group
confines of a standard Whipple operation, typically in the of neoplasms arising from pancreatic islet cells that have a
region between the CA and the SMA. Early reports sug- wide spectrum of biologic behavior. Traditionally, PNETs
gested that extended LAD of the CA, SMA, middle colic, are more indolent than their exocrine counterpart PDACs,
and para-aortic nodes was associated with decreased local but they can exert locoregional effects due to location and
recurrence and increased survival [34–36]. However, since size, and can metastasize to regional lymph nodes and distant
these early reports, extended LAD has been investigated in metastatic sites. The incidence of PNETs is rising due to
five prospective, randomized controlled trials (RCTs) increased use of cross-sectional imaging, endoscopic ultra-
(Table 40.2). Despite differences in study design and surgical sound, and nuclear medicine imaging (Ga-DOTATATE-­
outcomes, these RCTs showed that extended LAD does not PET), which has led to the detection of smaller tumors at an
improve R0 resection rates, overall recurrence rates, or long-­ earlier stage [41].
term survival compared with standard pancreatectomy. The 2019 World Health Organization classification
Extended LAD, however, is associated with increased mor- divides PNETs into well-differentiated NETs and poorly

Table 40.2 Randomized trials examining the impact of extended lymphadenectomy on post-resection survival for pancreatic ductal
adenocarcinoma
Lymph-­ Nodes Median OS
Author Institution Period adenectomy N retrieved (mos) P Morbidity
Pedrazzoli Italy, Germany, 1991–1994 Standard 40 13 11 0.65 No difference
(1998) USA Extended 41 20 16
Yeo (2002) Johns Hopkins 1996–2001 Standard 81 17 21 0.79 Extended LAD had increased LOS,
Extended 82 28 20 DGE, fistulae, and infection rate
Farnell (2005) Mayo Clinic 1997–2003 Standard 40 15 26 0.32 Extended LAD had increased
Extended 39 34 19 transfusion rate and OR time
Nimura Multicenter, 2000–2003 Standard 51 13 20 0.12 Extended LAD had higher OR time and
(2012) Japan Extended 50 40 14 intra-operative blood loss
Jang (2014, South Korea 2006–2009 Standard 83 17 19 0.39 Two extended patients had SMA
2017) Extended 86 34 16 aneurysms
LAD lymphadenectomy, LOS length of stay, DGE delayed gastric emptying, OS overall survival, SMA superior mesenteric artery
436 L. B. Kone et al.

d­ ifferentiated neuroendocrine carcinomas (NECs) based on Using a multi-institutional database along with SEER data,
tumor morphology, with further classification of well-­ Zhang et al. found that eight or more lymph nodes were
differentiated NETs into grades 1–3 based on the Ki67 pro- required for optimal N staging [48].
liferation index and mitotic index. Surgical resection remains Unlike PDAC, there are no prospective RCTs investigat-
the mainstay of treatment for functional tumors, larger local- ing a survival benefit of regional lymphadenectomy for
ized tumors, and even for highly selected patients with meta- PNETs. In a retrospective review of 2664 patients in the
static disease. National Cancer Database, of whom 80.6% received lymph-
adenectomy with a median of nine nodes resected, overall
survival was similar between patients who did and did not
40.6.1 Observation receive lymphadenectomy [49]. Furthermore, evidence for
pancreas-preserving approaches (enucleation and central
The natural history of small, well-differentiated PNETs is pancreatectomy) and spleen preservation, in which fewer
commonly one of little to no growth nor metastases; there- lymph nodes are typically harvested, also suggests that a for-
fore, many advocate for surveillance for tumors <2 cm. A mal lymphadenectomy may not be necessary in all patients.
study of 95 patients with small nonfunctioning PNETs Enucleations generally are performed for PNETs away from
(NF-PNETs) at the Mayo Clinic found a median growth rate the pancreatic duct as a way to avoid transection. Central
of 0.03 mm/month. No patients progressed beyond resect- pancreatectomy may also be considered for tumors at the
ability or developed metastatic disease during the observa- pancreatic neck or proximal body in order to preserve
tion period [42]. A European study found observation to be a parenchyma.
safe strategy for small NF-PNETs in 327 patients, none of Two meta-analyses have compared the outcomes of
whom had any disease-related mortality [43]. patients who underwent either standard resection or enucle-
ation. Enucleation was associated with a significantly
reduced incidence of exocrine and endocrine insufficiency,
40.6.2 Surgical Resection but the pancreatic fistula rate was higher [50, 51]. Similarly,
a recent meta-analysis concluded that central pancreatec-
Upfront surgical resection is often considered with curative tomy was associated with preserved pancreatic function
intent for localized PNETs that are >2 cm, <grade 3, symp- compared with distal pancreatectomy, but at the expense of
tomatic, functional, or with evidence of regional lymph node higher postoperative complications and length of stay [52].
metastases in patients who are fit for surgery. Population-­ Since neither enucleation nor central pancreatectomy typi-
based studies have demonstrated significantly improved sur- cally removes regional lymph nodes, these pancreas-­
vival for patients with localized tumors who have undergone preserving approaches are recommended for tumors at low
surgical resection versus no surgery, but these studies have risk of lymph node metastasis, such as insulinomas or small
been limited by significant selection bias [44, 45]. A large ret- NF-PNETs. Among patients who are candidates for
rospective review found that the association of survival with pancreas-­preserving resections, regional lymph node dis-
resection of NF-PNETs increased as tumor size increased, section and resection of suspicious lymph nodes on preop-
with a cutoff of >1.5 cm [46]. Most current guidelines recom- erative imaging should be considered. Radiographic criteria
mend observation for asymptomatic NF-PNETs < 1 cm, resec- and scoring systems that include size, grade, and location,
tion for appropriate surgical candidates with PNETs >2 cm, have been proposed to aid in identifying patients at highest
and to individualize the decision for resection versus observa- risk for lymph node metastases who may benefit from
tion for PNETs 1–2 cm based on patient age, comorbidities, lymphadenectomy [53].
tumor location (head vs. body/tail), grade, radiographic char-
acteristics (pancreatic duct dilatation, presence of calcifica-
tions, solid morphology), and patient preference. 40.7 Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is the fourth leading cause

40.6.3 Role of Lymphadenectomy of cancer-related death worldwide, and unlike many other
cancers, the overall burden of disease has been rising over
A recent meta-analysis of 13,374 patients with resected time. In the United States, the incidence of HCC has risen
PNETs found a strong association between lymph node two- to threefold over the past three decades, attributed prin-
metastasis and decreased overall survival, with a hazard ratio cipally to hepatitis C infection in the mid-twentieth century,
of 3.87 for all PNETs and 4.98 for NF-PNETs [47]. The total and the more recent rise in obesity-related non-alcohol ste-
number of nodes evaluated also impacts adequate staging. atohepatitis (NASH).
40 The Role of Surgery in Managing Primary and Metastatic Hepatopancreaticobiliary Cancers 437

40.7.1 Molecular Subtype gical approach, they can help guide prognosis after surgery,
and may help guide precision medicine approaches to sys-
HCC is the most common form of primary liver cancer, and temic therapies.
usually occurs in the setting of chronic liver disease (>90%),
namely viral hepatitis B and C infections, alcoholic cirrho-
sis, and non-alcoholic fatty liver disease (NAFLD). 40.7.2 Staging and Management
Dysplastic nodules, which consist of clonal hepatocytes with
premalignant features, arise in patients with chronic liver The surgical and clinical management of HCC is complex
disease. A series of genetic and epigenetic changes then pro- and depends on resource availability, local expertise, liver
mote a progression of low-grade to high-grade dysplasia to function, and patient condition. The Barcelona Clinic Liver
invasive cancer, also known as the dysplasia-carcinoma Cancer (BCLC) staging system can help determine appropri-
sequence. Telomerase reactivation is the principal molecular ate patients for surgery and takes into account tumor burden,
driver of this sequence. In the non-cirrhotic liver, however, liver function, and patient performance status (PS) (Figs. 40.1
HCC development arises in an adenoma-carcinoma and 40.2). Liver function is graded with the Child–Pugh sys-
sequence. The initial mutation in this sequence generally tem and PS is determined using the ECOG scale that ranges
occurs in the proto-oncogene CTNNB1, which encodes for from 0 (fully active) to 4 (in bed or chair at all times with
B-catenin. TERT mutation also plays a role, but arises later need for complete assistance with all tasks). In patients with
in the sequence [54]. Given the complex molecular heteroge- limited and compensated liver disease, the goals of surgical
neity of HCC development, subclassification is challenging, resection are to remove HCC with negative margins, while
but a popular classification scheme based on genetic signa- preserving an adequate functional liver remnant. Therefore,
ture divides HCC into proliferation and nonproliferation major resections are often avoided in favor of parenchymal
subtypes. The proliferation subtype is associated with TP53 preserving techniques and performed either via an open lapa-
mutation, chromosomal instability, hepatitis B infection, rotomy or minimally invasive technique. Techniques to
high AFP levels, high tumor grade, high vascular invasion, increase the size of the functional liver remnant may be con-
and worse clinical outcome. The nonproliferation subtype is sidered, e.g., portal vein embolization, though these are care-
associated with CTNNB1 mutation, immune exclusion, hep- fully considered on a patient-by-patient basis. As opposed to
atitis C infection, alcohol use, low AFP levels, low tumor liver resection in non-cirrhotic cases, operations on patients
grade, low vascular invasion, and better clinical outcomes with compensated cirrhosis are potentially complicated by
[55]. While these current subtypes do not yet dictate the sur- portal hypertension, thrombocytopenia, bleeding, and poor

Fig. 40.1 BCLC HCC staging system. (Adapted from “Diagnosis, Heimbach, 2018, Hepatology, 68(2):723–750. Reprinted with permis-
Staging, and Management of Hepatocellular Carcinoma: 2018 Practice sion). BCLC Barcelona Clinic Liver Cancer, HCC Hepatocellular
Guidance by the American Association for the Study of Liver Diseases” Carcinoma, ECOG Eastern Cooperative Oncology Group, PS
by Jorge A Marrero, Laura M Kulik, Claude B Sirlin, Andrew X Zhu, Performance status
Richard S Finn, Michael M Abecassis, Lewis R Roberts, Julie K
438 L. B. Kone et al.

BARCELONA STAGE

STAGE 0 STAGE A STAGE B STAGE C STAGE D

Level of
Evidence

1 Resection TACE Sorafenib (1L)

Lenvatinib (1L)
Regorafenib (2L)
Cabozantinib (2L)

2 RFA Resection TARE Nivolumab (2L) OLT

MWA OLT Downsize OLT BSC
RFA
MWA
TARE
TACE
SBRT
3 TARE

Fig. 40.2 Treatment recommendations according to BCLC Stage. Reprinted with permission). BCLC Barcelona Clinic Liver Cancer,
(Adapted from “Diagnosis, Staging, and Management of Hepatocellular MWA microwave ablation, RFA radiofrequency ablation, OLT
Carcinoma: 2018 Practice Guidance by the American Association for Orthotopic liver transplant, TARE Transarterial radioembolization,
the Study of Liver Diseases” by Jorge A Marrero, Laura M Kulik, TACE Transarterial chemoembolization, SBRT Stereotactic Body
Claude B Sirlin, Andrew X Zhu, Richard S Finn, Michael M Abecassis, Radiation Therapy, BSC best supportive care, 1 L first-line therapy, 2 L
Lewis R Roberts, Julie K Heimbach, 2018, Hepatology, 68(2):723–750. second-line therapy

liver function of the remnant liver, which make technical 40.7.3 Systemic Therapy
considerations of parenchymal sparing techniques, hemo-
static agents, and the use of ablation or energy devices all the Unfortunately, long-term survival, even after complete surgi-
more pertinent. cal resection or ablation, remains elusive, with half of
Patients with early HCC may also be eligible for ortho- patients experiencing recurrences within 18 months, and up
topic liver transplant. The Milan criteria identifies patients to 80% recurring within 5 years. Thus, including the 30–40%
with a single lesion less than 5 cm, or two-to-three lesions of patients that present with advanced disease, most patients
each less than 3 cm, as eligible for transplant with an expected will not be eligible for curative intent surgical options such
5-year overall survival greater than 70%. Transarterial che- as liver transplant or resection. Before 2007, there was no
moembolization (TACE) consists of intra-arterial infusion of effective systemic therapy for patients with advanced-stage
chemotherapy and embolizing particles of the tumor-feeding HCC. However, since that time there has been regained
artery, while TARE is based on the intra-arterial infusion of enthusiasm due to the success of the multikinase inhibitor
radioactive microspheres loaded with Yttrium-90 in the sorafenib in improving survival compared to best supportive
tumor-feeding artery. There are currently no major differ- care. In recent years, additional multikinase inhibitors (len-
ences in overall survival between the two modalities and vatinib and regorafenib) have gained approval for treatment
both remain important tools in the armamentarium for con- of advanced HCC, which have shown marginal advantages
trol of unresectable or multifocal disease and as a bridge to over sorafenib but without improving overall survival.
transplant. Though these agents have redefined the role of systemic ther-
40 The Role of Surgery in Managing Primary and Metastatic Hepatopancreaticobiliary Cancers 439

apy, the survival benefit remains measured in weeks. With CRLM not only in prognosis but also for treatment since
the great hope inspired by the success of PD-1/PD-L1 check- recent trials have demonstrated durable antitumor responses
point blockade immunotherapy in other tumor types, immu- to checkpoint blockade immunotherapy in MSI CRC.
notherapy for HCC has also been explored. Unfortunately, Recent advances in high throughput sequencing technol-
the PD-1 inhibitor nivolumab demonstrated no survival ben- ogy may enable subclassification pertinent to directed treat-
efit compared to sorafenib in the CheckMate040 trial, and ment of metastatic CRC. Based on 1290 CRCs, Sadanandam
pembrolizumab was only evaluated in the second-line set- et al. subclassified tumors into six groups associated with
ting. Since vascular endothelial growth factor (VEGF) response to systemic therapies. For instance, stem-like and
expression in the tumor microenvironment both promotes goblet-like subgroups may benefit from FOLFIRI combina-
angiogenesis and enhances evasion of tumors from the tion chemotherapy, while the transit-amplifying subgroup
immune system via increased expression of PD-1, combina- may be more responsive to either cetuximab or cMET inhibi-
tion therapy of atezolizumab (PD-L1 inhibitor) and bevaci- tors depending on the sensitivity subtype [62]. High through-
zumab (anti-VEGF) was explored and was recently put analysis also improves the ability to predict long-term
demonstrated to be superior to first-line sorafenib. As of May outcomes. Indeed, Pitroda et al. combined clinical data with
2020, this combination is recommended as first-line treat- integrated molecular analysis of colorectal liver metastases
ment for unresectable or metastatic HCC in appropriate and were able to create three subgroup risk strata among
patients [56]. patients who were eligible for surgical resection, with an
estimated 10-year overall survival of 94% (low risk), 45%
(intermediate risk), and 19% (high risk) [63]. Recent studies
40.8 Colorectal Liver Metastasis of BRAF, RAS, TP53, and SMAD4 mutations in resected
CRLM have also determined pathways associated with early
Colorectal cancer (CRC) is the second leading cause of recurrence after surgery [64]. It is possible that these muta-
cancer-­related death worldwide, with 20% presenting with tional signatures will be incorporated into the surgical
synchronous colorectal liver metastases (CRLM) and another decision-­making process of liver metastatic disease in the
20–30% eventually presenting with metachronous CRLM future. Furthermore, it has become evident that the immuno-
[57, 58]. In selected patients with liver-limited disease and logic landscape of CRLM tumor microenvironment is a pre-
an appropriate functional liver remnant, surgical resection dictor of outcome. It has been demonstrated that the presence
has been demonstrated to provide a greater than 20% 10-year of activated and proliferating tumor-infiltrating lymphocytes
survival. The clinical risk score (CRS) was one of the first is associated with improved survival [65]. Moreover, the
reliable predictors of survival in patients with CRLM that presence of regulatory T-cells is associated with worse sur-
helped select patients for surgical resection. Predictors of vival [66], and the ratio of cytotoxic T-cells to regulatory
poor outcome in this score include positive margins, extrahe- T-cells is predictive of outcomes. These studies will help
patic disease, node-positive primary, disease-free interval determine patients most likely to benefit from new immuno-
<12 months, >1 hepatic tumor, largest tumor >5 cm, and therapeutic strategies that enhance T-cell infiltration and sup-
CEA > 200 ng/mL [59]. More recently, the GAME score was press immunosuppressive signals.
created utilizing only preoperative data and was demon-
strated to be a superior predictor of outcome compared to the
CRS [60]. 40.8.2 Surgical Management

Surgical resection of CRLM is currently the only curative-­

40.8.1 Molecular Subtype intent treatment and it provides the best chance of long-term
survival in highly selected patients. Historically, a 1-cm mar-
CRC has been classified into three classical pathways. The gin was deemed necessary, however, it is now accepted that
most common is the chromosomal instability pathway, which the width of the margin is not a predictor of outcome [67].
follows the adenoma-carcinoma sequence starting with APC Thus, similar to HCC, parenchymal sparing hepatectomy is
mutation, followed by KRAS mutation, and TP53 inactiva- preferred to anatomical hepatectomies where appropriate. In
tion. Second is the microsatellite instability (MSI) pathway patients with bilateral CRLM, surgical resection may still be
resulting from inactivation of mismatch repair genes (MLH1, pursued with specific strategies. In a two-staged hepatec-
MSH2). Third is the CpG island methylation phenotype tomy (TSH), one side of the liver can be surgically cleared of
(CIMP) pathway, which often occurs concurrently with the tumor through resection and/or ablation, and portal venous
MSI pathway, and results in inactivating methylation of ligation/embolization of the contralateral vessel is utilized to
tumor suppressor genes (MINT1, MLH1, TIMP3) [61]. increase the functional liver remnant prior to resection of dis-
These pathways become important in consideration of ease in the contralateral lobe. Another less common approach
440 L. B. Kone et al.

performed in highly selected patients is associating liver par- 40.9 Conclusion and Perspective
tition and portal vein ligation for staged hepatectomy
(ALPPS) , which was first described in 2011 and appears to Tumor biology, surgical management, and adjunctive ther-
increase liver growth rate and volume in a shorter period of apy for HPB cancers were reviewed. Surgery alone, or as a
time. The LIGRO trial (NCT02215577) demonstrated a part of an adjunctive treatment plan, remains the best and
higher resection rate with ALLPS over TSH and equal safety only curative-intent strategy to enable long-term survival.
profile, but these procedures remain isolated to a few special- Survival after surgical resection is dependent on tumor biol-
ized centers around the world due to their morbidity and ogy and lymphovascular invasion/involvement. Advances in
mortality. In the setting of synchronous disease, the timing of systemic, ablative, and radiation therapy are likely to both
hepatic resection with respect to the primary tumor resection increase operative eligibility and improve outcomes.
is also an area of study. While there is no strong data to sup-
port liver-first versus colorectal-first versus combined resec-
tion strategies, some have advocated that combined resection Future Directions in this Field will Likely be Driven by
be reserved for healthy, non-rectal cancer patients requiring Answers to the Following Questions
minor hepatectomies. Interestingly, the SECA-2 trial • What is the role of neoadjuvant chemotherapy in
(NCT01479608) in Norway reported a 5-year overall sur- improving survival in gallbladder and biliary
vival of 83% after liver transplant in patients with unresect- cancers?
able CRLM and good response to chemotherapy by RECIST • How can pancreatic cancer be diagnosed and treated
criteria (10% decrease). However, these are highly selected at an earlier stage?
patients and liver transplant for CRLM is uncommon in the • Can we biologically determine which pancreatic
United States, compounded by the low availability of liver neuroendocrine tumors, regardless of size, have
donors. malignant potential?
• Is it possible to stop or reverse the biologic path-
ways that lead to progressive dysplasia in cirrhotic
40.8.3 Perioperative Therapy livers?
• How can we harness the immune system to recog-
Numerous studies have demonstrated that the timing of sys- nize and attack colon cancers that have metasta-
temic chemotherapy (neoadjuvant or adjuvant) does not sized to the liver?
affect long-term survival in resectable CRLM. Furthermore,
perioperative chemotherapy has not been shown to impact
overall survival when compete extirpation of disease is
Authors Disclosures Lyonell Kone, Christopher Javadi, Jessica
­performed. Regardless, systemic chemotherapy plays a sig- Keilson, Shishir Maithel, George Poultsides, and Ajay Maker have no
nificant role in the surgical treatment of CRLM by reducing disclosures to report.
the risk of relapse, assessing for chemosensitivity, testing the
biology of disease, and by potentially converting an unre-
sectable to a resectable tumor. This is at the expense of a
potentially disappearing small tumor and chemotoxicity, References
specifically steatosis and steatohepatitis.
Patients with unresectable disease that receive chemo- 1. Ito M, Mishima Y, Sato T. An anatomical study of the lymphatic
drainage of the gallbladder. Surg Radiol Anat. 1991;13(2):89–104.
therapy to extend survival will benefit from regimens that https://doi.org/10.1007/BF01623880.
include combination chemotherapy FOLFOX or FOLFIRI, 2. Chijiiwa K, Nakano K, Ueda J, Noshiro H, Nagai E,
anti-EGFR antibodies (Panitumumab, Cetuximab), and/or Yamaguchi K, Tanaka M. Surgical treatment of patients with
the addition of anti-VEGF (Bevacizumab) therapy. Intra-­ T2 gallbladder carcinoma invading the subserosal layer. J
Am Coll Surg. 2001;192(5):600–7. https://doi.org/10.1016/
arterial therapies (TACE and TARE) are also adjuncts for the s1072-­7515(01)00814-­6.
management of unresectable CRLM, with some reporting 3. Lee SE, Jang JY, Lim CS, Kang MJ, Kim SW. Systematic review
conversion to resectable CRLM as high as 52% [68]. This on the surgical treatment for T1 gallbladder cancer. World J
data builds on the long-standing data reporting the efficacy Gastroenterol. 2011;17(2):174–80. https://doi.org/10.3748/wjg.
v17.i2.174.
of Floxuridine-based (FUDR) hepatic artery infusion pumps 4. D’Angelica M, Dalal KM, DeMatteo RP, Fong Y, Blumgart LH,
in improving survival, response rates, and conversion to Jarnagin WR. Analysis of the extent of resection for adenocarci-
resectable disease in bilateral CRLM [69]. noma of the gallbladder. Ann Surg Oncol. 2009;16(4):806–16.
https://doi.org/10.1245/s10434-­008-­0189-­3.
40 The Role of Surgery in Managing Primary and Metastatic Hepatopancreaticobiliary Cancers 441

5. Kondo S, Nimura Y, Hayakawa N, Kamiya J, Nagino M, Uesaka DJ, Kingham TP, Jarnagin WR, van Gulik TM. Recurrence rate
K. Regional and para-aortic lymphadenectomy in radical surgery and pattern of perihilar cholangiocarcinoma after curative intent
for advanced gallbladder carcinoma. Br J Surg. 2000;87(4):418–22. resection. J Am Coll Surg. 2015;221(6):1041–9. https://doi.
https://doi.org/10.1046/j.1365-­2168.2000.01384.x. org/10.1016/j.jamcollsurg.2015.09.005.
6. Ethun CG, Postlewait LM, Le N, Pawlik TM, Buettner S, Poultsides 19. Ethun CG, Lopez-Aguiar AG, Anderson DJ, Adams AB, Fields
G, Tran T, Idrees K, Isom CA, Fields RC, Jin LX, Weber SM, Salem RC, Doyle MB, Chapman WC, Krasnick BA, Weber SM,
A, Martin RC, Scoggins C, Shen P, Mogal HD, Schmidt C, Beal E, Mezrich JD, Salem A, Pawlik TM, Poultsides G, Tran TB, Idrees
Hatzaras I, Shenoy R, Merchant N, Cardona K, Maithel SK. A novel K, Isom CA, Martin RCG, Scoggins CR, Shen P, Mogal HD,
pathology-based preoperative risk score to predict locoregional Schmidt C, Beal E, Hatzaras I, Shenoy R, Cardona K, Maithel
residual and distant disease and survival for incidental gallbladder SK. Transplantation versus resection for hilar cholangiocarcinoma:
cancer: a 10-institution study from the U.S. Extrahepatic Biliary an argument for shifting treatment paradigms for resectable dis-
Malignancy Consortium. Ann Surg Oncol. 2017;24(5):1343–50. ease. Ann Surg. 2018;267(5):797–805. https://doi.org/10.1097/
https://doi.org/10.1245/s10434-­016-­5637-­x. SLA.0000000000002574.
7. Amin MB, Greene FL, Edge SB, Compton CC, Gershenwald JE, 20. Ito K, Ito H, Allen PJ, Gonen M, Klimstra D, D’Angelica MI,
Brookland RK, Meyer L, Gress DM, Byrd DR, Winchester DP. The Fong Y, DeMatteo RP, Brennan MF, Blumgart LH, Jarnagin
eighth edition AJCC cancer staging manual: continuing to build a WR. Adequate lymph node assessment for extrahepatic bile duct
bridge from a population-based to a more “personalized” approach adenocarcinoma. Ann Surg. 2010;251(4):675–81. https://doi.
to cancer staging. CA Cancer J Clin. 2017;67(2):93–9. https://doi. org/10.1097/SLA.0b013e3181d3d2b2.
org/10.3322/caac.21388. 21. Postlewait LM, Ethun CG, Le N, Pawlik TM, Buettner S, Poultsides
8. Liu GJ, Li XH, Chen YX, Sun HD, Zhao GM, Hu SY. Radical G, Tran T, Idrees K, Isom CA, Fields RC, Krasnick B, Weber SM,
lymph node dissection and assessment: impact on gallbladder can- Salem A, Martin RCG, Scoggins C, Shen P, Mogal HD, Schmidt C,
cer prognosis. World J Gastroenterol. 2013;19(31):5150–8. https:// Beal E, Hatzaras I, Vitiello G, Cardona K, Maithel SK. Proposal for
doi.org/10.3748/wjg.v19.i31.5150. a new T-stage classification system for distal cholangiocarcinoma: a
9. Negi SS, Singh A, Chaudhary A. Lymph nodal involvement as 10-institution study from the U.S. Extrahepatic Biliary Malignancy
prognostic factor in gallbladder cancer: location, count or ratio? Consortium. HPB (Oxford). 2016;18(10):793–9. https://doi.
J Gastrointest Surg. 2011;15(6):1017–25. https://doi.org/10.1007/ org/10.1016/j.hpb.2016.07.009.
s11605-­011-­1528-­4. 22. Primrose JN, Fox RP, Palmer DH, Malik HZ, Prasad R, Mirza
10. Xiao Z, Shi Z, Hu L, Gao Y, Zhao J, Liu Y, Xu Q, Huang D. A D, Anthony A, Corrie P, Falk S, Finch-Jones M, Wasan H, Ross
new nomogram from the SEER database for predicting the prog- P, Wall L, Wadsley J, Evans JTR, Stocken D, Praseedom R, Ma
nosis of gallbladder cancer patients after surgery. Ann Transl Med. YT, Davidson B, Neoptolemos JP, Iveson T, Raftery J, Zhu S,
2019;7(23):738. https://doi.org/10.21037/atm.2019.11.112. Cunningham D, Garden OJ, Stubbs C, Valle JW, Bridgewater J,
11. Jensen EH, Abraham A, Habermann EB, Al-Refaie WB, Vickers BILCAP Study Group. Capecitabine compared with observation in
SM, Virnig BA, Tuttle TM. A critical analysis of the surgical man- resected biliary tract cancer (BILCAP): a randomised, controlled,
agement of early-stage gallbladder cancer in the United States. J multicentre, phase 3 study. Lancet Oncol. 2019;20(5):663–73.
Gastrointest Surg. 2009;13(4):722–7. https://doi.org/10.1007/ https://doi.org/10.1016/S1470-­2045(18)30915-­X.
s11605-­008-­0772-­8. 23. Ben-Josef E, Guthrie KA, El-Khoueiry AB, Corless CL, Zalupski
12. Foster JM, Hoshi H, Gibbs JF, Iyer R, Javle M, Chu Q, Kuvshinoff MM, Lowy AM, Thomas CR Jr, Alberts SR, Dawson LA, Micetich
B. Gallbladder cancer: defining the indications for primary KC, Thomas MB, Siegel AB, Blanke CD. SWOG S0809: a phase
radical resection and radical re-resection. Ann Surg Oncol. II intergroup trial of adjuvant capecitabine and gemcitabine fol-
2007;14(2):833–40. https://doi.org/10.1245/s10434-­006-­9097-­6. lowed by radiotherapy and concurrent capecitabine in extrahepatic
13. Murakami Y, Uemura K, Sudo T, Hashimoto Y, Nakashima A, cholangiocarcinoma and gallbladder carcinoma. J Clin Oncol.
Kondo N, Sakabe R, Ohge H, Sueda T. Prognostic factors after 2015;33(24):2617–22. https://doi.org/10.1200/JCO.2014.60.2219.
surgical resection for intrahepatic, hilar, and distal cholangiocarci- 24. Shroff RT, Kennedy EB, Bachini M, Bekaii-Saab T, Crane C,
noma. Ann Surg Oncol. 2011;18(3):651–8. https://doi.org/10.1245/ Edeline J, El-Khoueiry A, Feng M, Katz MHG, Primrose J,
s10434-­010-­1325-­4. Soares HP, Valle J, Maithel SK. Adjuvant therapy for resected bili-
14. Morine Y, Shimada M. The value of systematic lymph node dis- ary tract cancer: ASCO clinical practice guideline. J Clin Oncol.
section for intrahepatic cholangiocarcinoma from the viewpoint of 2019;37(12):1015–27. https://doi.org/10.1200/JCO.18.02178.
liver lymphatics. J Gastroenterol. 2015;50(9):913–27. https://doi. 25. Rahib L, Smith BD, Aizenberg R, Rosenzweig AB, Fleshman JM,
org/10.1007/s00535-­015-­1071-­2. Matrisian LM. Projecting cancer incidence and deaths to 2030:
15. Patel SH, Kooby DA, Staley CA 3rd, Sarmiento JM, Maithel the unexpected burden of thyroid, liver, and pancreas cancers in
SK. The prognostic importance of lymphovascular invasion in chol- the United States. Cancer Res. 2014;74(11):2913–21. https://doi.
angiocarcinoma above the cystic duct: a new selection criterion for org/10.1158/0008-­5472.CAN-­14-­0155.
adjuvant therapy? HPB (Oxford). 2011;13(9):605–11. https://doi. 26. Winter JM, Brennan MF, Tang LH, D’Angelica MI, Dematteo RP,
org/10.1111/j.1477-­2574.2011.00335.x. Fong Y, Klimstra DS, Jarnagin WR, Allen PJ. Survival after resec-
16. Ocuin LM, Bagci P, Fisher SB, Patel SH, Kooby DA, Sarmiento tion of pancreatic adenocarcinoma: results from a single institution
JM, Cardona K, Russell MC, Staley CA, Volkan Adsay N, Maithel over three decades. Ann Surg Oncol. 2012;19(1):169–75. https://
SK. Discordance between conventional and detailed lymph node doi.org/10.1245/s10434-­011-­1900-­3.
analysis in resected biliary carcinoma at or above the cystic duct: 27. Yachida S, Jones S, Bozic I, Antal T, Leary R, Fu B, Kamiyama
are we understaging patients? Ann Surg Oncol. 2013;20(13):4298– M, Hruban RH, Eshleman JR, Nowak MA, Velculescu VE, Kinzler
304. https://doi.org/10.1245/s10434-­013-­3143-­y. KW, Vogelstein B, Iacobuzio-Donahue CA. Distant metastasis
17. Li J, Zhou MH, Ma WJ, Li FY, Deng YL. Extended lymphadenec- occurs late during the genetic evolution of pancreatic cancer. Nature.
tomy in hilar cholangiocarcinoma: what it will bring? World J 2010;467(7319):1114–7. https://doi.org/10.1038/nature09515.
Gastroenterol. 2020;26(24):3318–25. https://doi.org/10.3748/wjg. 28. Evans DB, Farnell MB, Lillemoe KD, Vollmer C, Strasberg SM,
v26.i24.3318. Schulick RD. Surgical treatment of resectable and borderline resect-
18. Groot Koerkamp B, Wiggers JK, Allen PJ, Besselink MG, Blumgart able pancreas cancer: expert consensus statement. Ann Surg Oncol.
LH, Busch OR, Coelen RJ, D’Angelica MI, DeMatteo RP, Gouma 2009;16(7):1736–44. https://doi.org/10.1245/s10434-­009-­0416-­6.
442 L. B. Kone et al.

29. Van den Broeck A, Sergeant G, Ectors N, Van Steenbergen W, Aerts pancreatic ductal adenocarcinoma. Ann Surg. 2020;272(3):481–6.
R, Topal B. Patterns of recurrence after curative resection of pan- https://doi.org/10.1097/SLA.0000000000004155.
creatic ductal adenocarcinoma. Eur J Surg Oncol. 2009;35(6):600– 40. Ghaneh P, Palmer DH, Cicconi S, Halloran C, Psarelli EE,
4. https://doi.org/10.1016/j.ejso.2008.12.006. Rawcliffe CL, Sripadam R, Mukherjee S, Wadsley J, Al-Mukhtar
30. Allen PJ, Kuk D, Castillo CF-D, Basturk O, Wolfgang CL, Cameron A, Jiao LR, Wasan HS, Carter R, Graham JS, Ammad F, Evans
JL, Lillemoe KD, Ferrone CR, Morales-Oyarvide V, He J, Weiss MJ, J, Tjaden C, Hackert T, Buchler MW, Neoptolemos JP, European
Hruban RH, Gönen M, Klimstra DS, Mino-Kenudson M. Multi-­ Study Group for Pancreatic Cancer (ESPAC). ESPAC-5F: four-arm,
institutional validation study of the American Joint Commission prospective, multicenter, international randomized phase II trial of
on cancer (8th edition) changes for T and N staging in patients immediate surgery compared with neoadjuvant gemcitabine plus
with pancreatic adenocarcinoma. Ann Surg. 2017;265(1):185–91. capecitabine (GEMCAP) or FOLFIRINOX or chemoradiotherapy
https://doi.org/10.1097/SLA.0000000000001763. (CRT) in patients with borderline resectable pancreatic cancer. J
31. Showalter TN, Winter KA, Berger AC, Regine WF, Abrams RA, Clin Oncol. 2020;38(15_suppl):4505. https://doi.org/10.1200/
Safran H, Hoffman JP, Benson AB, MacDonald JS, Willett CG. The JCO.2020.38.15_suppl.4505.
influence of total nodes examined, number of positive nodes, and 41. Dasari A, Shen C, Halperin D, Zhao B, Zhou S, Xu Y, Shih T,
lymph node ratio on survival after surgical resection and adjuvant Yao JC. Trends in the incidence, prevalence, and survival out-
chemoradiation for pancreatic cancer: a secondary analysis of comes in patients with neuroendocrine tumors in the United States.
RTOG 9704. Int J Radiat Oncol Biol Phys. 2011;81(5):1328–35. JAMA Oncol. 2017;3(10):1335–42. https://doi.org/10.1001/
https://doi.org/10.1016/j.ijrobp.2010.07.1993. jamaoncol.2017.0589.
32. Strobel O, Hinz U, Gluth A, Hank T, Hackert T, Bergmann 42. Yohanathan L, Dossa F, St Germain AT, Golbafian F, Moulton
F, Werner J, Büchler MW. Pancreatic adenocarcinoma: num- C-A, McGilvray ID, Greig PD, Serra S, Wei AC, Jhaveri KS,
ber of positive nodes allows to distinguish several N catego- Gallinger S, Cleary SP. Management and surveillance of non-func-
ries. Ann Surg. 2015;261(5):961–9. https://doi.org/10.1097/ tional pancreatic neuroendocrine tumours: retrospective review.
SLA.0000000000000814. Pancreatology. 2019;19(2):360–6. https://doi.org/10.1016/j.
33. Tomlinson JS, Jain S, Bentrem DJ, Sekeris EG, Maggard pan.2019.02.008.
MA, Hines OJ, Reber HA, Ko CY. Accuracy of staging node-­ 43. Partelli S, Cirocchi R, Crippa S, Cardinali L, Fendrich V, Bartsch
negative pancreas cancer: a potential quality measure. Arch Surg. DK, Falconi M. Systematic review of active surveillance versus
2007;142(8):767–23; discussion 773–764. https://doi.org/10.1001/ surgical management of asymptomatic small non-functioning pan-
archsurg.142.8.767. creatic neuroendocrine neoplasms. Br J Surg. 2017;104(1):34–41.
34. Cubilla AL, Fortner J, Fitzgerald PJ. Lymph node involve- https://doi.org/10.1002/bjs.10312.
ment in carcinoma of the head of the pancreas area. Cancer. 44. Assi HA, Mukherjee S, Kunz PL, Machiorlatti M, Vesely S, Pareek
1978;41(3):880–7. V, Hatoum H. Surgery versus surveillance for well-differentiated,
35. Fortner JG. Regional resection of cancer of the pancreas: a new sur- nonfunctional pancreatic neuroendocrine tumors: an 11-year analy-
gical approach. Surgery. 1973;73(2):307–20. papers3://publication/ sis of the National Cancer Database. Oncologist. 2020;25(2):e276–
uuid/502CAAC8-A199-4028-9419-52EFF08CE121. 83. https://doi.org/10.1634/theoncologist.2019-­0466.
36. Ishikawa O, Ohigashi H, Imaoka S, Furukawa H, Sasaki Y, Fujita 45. Hill JS, McPhee JT, McDade TP, Zhou Z, Sullivan ME, Whalen GF,
M, Kuroda C, Iwanaga T. Preoperative indications for extended Tseng JF. Pancreatic neuroendocrine tumors: the impact of surgi-
pancreatectomy for locally advanced pancreas cancer i­nvolving cal resection on survival. Cancer. 2009;115(4):741–51. https://doi.
the portal vein. Ann Surg. 1992;215(3):231–6. https://doi. org/10.1002/cncr.24065.
org/10.1097/00000658-­199203000-­00006. 46. Zhang IY, Zhao J, Fernandez-Del Castillo C, Braun Y, Razmdjou
37. Neoptolemos JP, Palmer DH, Ghaneh P, Psarelli EE, Valle JW, S, Warshaw AL, Lillemoe KD, Ferrone CR. Operative versus non-
Halloran CM, Faluyi O, O'Reilly DA, Cunningham D, Wadsley operative management of nonfunctioning pancreatic neuroendo-
J, Darby S, Meyer T, Gillmore R, Anthoney A, Lind P, Glimelius crine tumors. J Gastrointest Surg. 2016;20(2):277–83. https://doi.
B, Falk S, Izbicki JR, Middleton GW, Cummins S, Ross PJ, org/10.1007/s11605-­015-­3043-­5.
Wasan H, McDonald A, Crosby T, Ma YT, Patel K, Sherriff D, 47. Tanaka M, Heckler M, Mihaljevic AL, Probst P, Klaiber U, Heger
Soomal R, Borg D, Sothi S, Hammel P, Hackert T, Jackson R, U, Schimmack S, Büchler MW, Hackert T. Systematic review and
Buchler MW, European Study Group for Pancreatic Cancer. metaanalysis of lymph node metastases of resected pancreatic neu-
Comparison of adjuvant gemcitabine and capecitabine with gem- roendocrine tumors. Ann Surg Oncol. 2020;21:9512. https://doi.
citabine monotherapy in patients with resected pancreatic can- org/10.1245/s10434-­020-­08850-­7.
cer (ESPAC-4): a multicentre, open-label, randomised, phase 3 48. Zhang X-F, Xue F, Dong D-H, Lopez-Aguiar AG, Poultsides G,
trial. Lancet. 2017;389(10073):1011–24. https://doi.org/10.1016/ Makris E, Rocha F, Kanji Z, Weber S, Fisher A, Fields R, Krasnick
S0140-­6736(16)32409-­6. BA, Idrees K, Smith PM, Cho C, Beems M, Lv Y, Maithel SK,
38. Conroy T, Hammel P, Hebbar M, Ben Abdelghani M, Wei AC, Pawlik TM. New nodal staging for primary pancreatic neuro-
Raoul JL, Chone L, Francois E, Artru P, Biagi JJ, Lecomte T, endocrine tumors: a multi-institutional and national data analy-
Assenat E, Faroux R, Ychou M, Volet J, Sauvanet A, Breysacher sis. Ann Surg. 2019;274(1):e28–35. https://doi.org/10.1097/
G, Di Fiore F, Cripps C, Kavan P, Texereau P, Bouhier-Leporrier K, SLA.0000000000003478.
Khemissa-Akouz F, Legoux JL, Juzyna B, Gourgou S, O’Callaghan 49. Mao R, Zhao H, Li K, Luo S, Turner M, Cai J-Q, Blazer D. Outcomes
CJ, Jouffroy-Zeller C, Rat P, Malka D, Castan F, Bachet JB, of lymph node dissection for non-metastatic pancreatic neuroendo-
Canadian Cancer Trials Group and the Unicancer-GI–PRODIGE crine tumors: a propensity score-weighted analysis of the National
Group. FOLFIRINOX or gemcitabine as adjuvant therapy for pan- Cancer Database. Ann Surg Oncol. 2019;26(9):2722–9. https://doi.
creatic cancer. N Engl J Med. 2018;379(25):2395–406. https://doi. org/10.1245/s10434-­019-­07506-­5.
org/10.1056/NEJMoa1809775. 50. Chua TC, Yang TX, Gill AJ, Samra JS. Systematic review and meta-­
39. Ahmad SA, Duong M, Sohal DPS, Gandhi NS, Beg MS, Wang-­ analysis of enucleation versus standardized resection for small
Gillam A, Wade JL 3rd, Chiorean EG, Guthrie KA, Lowy AM, pancreatic lesions. Ann Surg Oncol. 2016;23(2):592–9. https://doi.
Philip PA, Hochster HS. Surgical outcome results from SWOG org/10.1245/s10434-­015-­4826-­3.
S1505: a randomized clinical trial of mFOLFIRINOX versus gem- 51. Hüttner FJ, Koessler-Ebs J, Hackert T, Ulrich A, Büchler MW,
citabine/Nab-paclitaxel for perioperative treatment of resectable Diener MK. Meta-analysis of surgical outcome after enucleation
40 The Role of Surgery in Managing Primary and Metastatic Hepatopancreaticobiliary Cancers 443

versus standard resection for pancreatic neoplasms. Br J Surg. 62. Sadanandam A, Lyssiotis CA, Homicsko K, Collisson EA, Gibb
2015;102(9):1026–36. https://doi.org/10.1002/bjs.9819. WJ, Wullschleger S, Ostos LC, Lannon WA, Grotzinger C, Del Rio
52. Dragomir MP, Sabo AA, Petrescu GED, Li Y, Dumitrascu M, Lhermitte B, Olshen AB, Wiedenmann B, Cantley LC, Gray
T. Central pancreatectomy: a comprehensive, up-to-date meta-­ JW, Hanahan D. A colorectal cancer classification system that
analysis. Langenbeck’s Arch Surg. 2019;404(8):945–58. https:// associates cellular phenotype and responses to therapy. Nat Med.
doi.org/10.1007/s00423-­019-­01829-­3. 2013;19(5):619–25. https://doi.org/10.1038/nm.3175.
53. Poultsides GA, Huang LC, Chen Y, Visser BC, Pai RK, Jeffrey RB, 63. Pitroda SP, Khodarev NN, Huang L, Uppal A, Wightman SC, Ganai
Park WG, Chen AM, Kunz PL, Fisher GA, Norton JA. Pancreatic S, Joseph N, Pitt J, Brown M, Forde M, Mangold K, Xue L, Weber
neuroendocrine tumors: radiographic calcifications correlate with C, Segal JP, Kadri S, Stack ME, Khan S, Paty P, Kaul K, Andrade
grade and metastasis. Ann Surg Oncol. 2012;19(7):2295–303. J, White KP, Talamonti M, Posner MC, Hellman S, Weichselbaum
https://doi.org/10.1245/s10434-­012-­2305-­7. RR. Integrated molecular subtyping defines a curable oligo-
54. Craig AJ, von Felden J, Garcia-Lezana T, Sarcognato S, Villanueva metastatic state in colorectal liver metastasis. Nat Commun.
A. Tumour evolution in hepatocellular carcinoma. Nat Rev 2018;9(1):1793. https://doi.org/10.1038/s41467-­018-­04278-­6.
Gastroenterol Hepatol. 2020;17(3):139–52. https://doi.org/10.1038/ 64. Kawaguchi Y, Kopetz S, Newhook TE, De Bellis M, Chun YS,
s41575-­019-­0229-­4. Tzeng CD, Aloia TA, Vauthey JN. Mutation status of RAS, TP53,
55. Villanueva A. Hepatocellular carcinoma. N Engl J Med. and SMAD4 is superior to mutation status of RAS alone for pre-
2019;380(15):1450–62. https://doi.org/10.1056/NEJMra1713263. dicting prognosis after resection of colorectal liver metastases. Clin
56. Finn RS, Qin S, Ikeda M, Galle PR, Ducreux M, Kim TY, Kudo M, Cancer Res. 2019;25(19):5843–51. https://doi.org/10.1158/1078-
Breder V, Merle P, Kaseb AO, Li D, Verret W, Xu DZ, Hernandez ­0432.CCR-­19-­0863.
S, Liu J, Huang C, Mulla S, Wang Y, Lim HY, Zhu AX, Cheng AL, 65. Maker AV, Ito H, Mo Q, Weisenberg E, Qin LX, Turcotte S,
Investigators IM. Atezolizumab plus Bevacizumab in unresectable Maithel S, Shia J, Blumgart L, Fong Y, Jarnagin WR, DeMatteo
hepatocellular carcinoma. N Engl J Med. 2020;382(20):1894–905. RP, D’Angelica MI. Genetic evidence that intratumoral T-cell pro-
https://doi.org/10.1056/NEJMoa1915745. liferation and activation are associated with recurrence and sur-
57. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal vival in patients with resected colorectal liver metastases. Cancer
A. Global cancer statistics 2018: GLOBOCAN estimates of inci- Immunol Res. 2015;3(4):380–8. https://doi.org/10.1158/2326-
dence and mortality worldwide for 36 cancers in 185 countries. ­6066.CIR-­14-­0212.
CA Cancer J Clin. 2018;68(6):394–424. https://doi.org/10.3322/ 66. Katz SC, Bamboat ZM, Maker AV, Shia J, Pillarisetty VG, Yopp
caac.21492. AC, Hedvat CV, Gonen M, Jarnagin WR, Fong Y, D’Angelica
58. Leporrier J, Maurel J, Chiche L, Bara S, Segol P, Launoy G. A MI, DeMatteo RP. Regulatory T cell infiltration predicts out-
population-based study of the incidence, management and prog- come following resection of colorectal cancer liver metastases.
nosis of hepatic metastases from colorectal cancer. Br J Surg. Ann Surg Oncol. 2013;20(3):946–55. https://doi.org/10.1245/
2006;93(4):465–74. https://doi.org/10.1002/bjs.5278. s10434-­012-­2668-­9.
59. Fong Y, Fortner J, Sun RL, Brennan MF, Blumgart LH. Clinical 67. Pawlik TM, Scoggins CR, Zorzi D, Abdalla EK, Andres A, Eng
score for predicting recurrence after hepatic resection for meta- C, Curley SA, Loyer EM, Muratore A, Mentha G, Capussotti L,
static colorectal cancer: analysis of 1001 consecutive cases. Ann Vauthey JN. Effect of surgical margin status on survival and site
Surg. 1999;230(3):309–18; discussion 318–321. https://doi. of recurrence after hepatic resection for colorectal metastases.
org/10.1097/00000658-­199909000-­00004. Ann Surg. 2005;241(5):715–22; discussion 722-714. https://doi.
60. Margonis GA, Sasaki K, Gholami S, Kim Y, Andreatos N, Rezaee org/10.1097/01.sla.0000160703.75808.7d.
N, Deshwar A, Buettner S, Allen PJ, Kingham TP, Pawlik TM, 68. Datta J, Narayan RR, Kemeny NE, D'Angelica MI. Role of hepatic
He J, Cameron JL, Jarnagin WR, Wolfgang CL, D'Angelica MI, artery infusion chemotherapy in treatment of initially unresectable
Weiss MJ. Genetic And Morphological Evaluation (GAME) colorectal liver metastases: a review. JAMA Surg. 2019;154(8):768–
score for patients with colorectal liver metastases. Br J Surg. 76. https://doi.org/10.1001/jamasurg.2019.1694.
2018;105(9):1210–20. https://doi.org/10.1002/bjs.10838. 69. Thiels CA, D’Angelica MI. Hepatic artery infusion pumps. J Surg
61. Markowitz SD, Bertagnolli MM. Molecular origins of can- Oncol. 2020;122(1):70–7. https://doi.org/10.1002/jso.25913.
cer: molecular basis of colorectal cancer. N Engl J Med.
2009;361(25):2449–60. https://doi.org/10.1056/NEJMra0804588.
 Part X
Cancer Metastasis to the Lung

Chinmayee Potti, Hassan Nasser

Andrew Popoff and Zane Hammoud

Lung cancer kills more people than breast, colon, and prostate cancer combined. The American
Cancer Society estimates that for 2020, 229,000 Americans will be diagnosed with lung cancer
and roughly 136,000 deaths will be attributed to the disease. The majority of lung cancers pres-
ent at advanced stages or are metastatic (stage IV) at the time of diagnosis. The prognosis in
advanced disease is dramatically worse than when it is detected early. Lung cancer screening
programs, now well established in Western Europe, the United States, Canada, and elsewhere,
aim to shift this survival curve by detecting more early, curable disease. The penetrance of lung
cancer screening, however, remains low at present.
Knowledge of the anatomy, pathophysiology, and management of advanced and metastatic
disease is crucial for appropriate, guideline-directed clinical care. This chapter provides an
overview of anatomy, presentation, and prognostic implications of metastatic disease, and
management of various clinical scenarios that cancer clinicians may encounter. Less common
clinical situations, such as the management of oligometastatic disease, as well as combined
modality and palliative approaches to the care of metastatic disease, are also discussed. Other
primary malignancies, such as colorectal, renal cell, and uterine cancers with metastases to the
lung, and pulmonary metastasectomy are also briefly addressed.
Scientific advances in the treatment of advanced and metastatic disease, such as targeted
therapies directed at specific tumor genetics, as well as immunotherapy, either alone or in con-
cert with conventional chemotherapy, have strengthened the armamentarium of clinicians who
routinely care for patients with metastatic disease. It is likely that as more refined therapeutics
become available, such as chimeric antigen receptor T-cell (CAR T-cell) therapy for lung can-
cer and other solid tumors, the arsenal of treatment options will continue to evolve. This, and
other novel modalities, remains an area of intense interest in the future to have an even greater
impact on the treatment of a disease that carries what has traditionally been a dismal
prognosis.
 Metastatic Disease of the Lung
41
Z. Hammoud, A. Popoff, Chinmayee Potti, and H. Nasser

Abstract –– Histological tropism of lung metastases to cer-

Lung cancer remains a malignancy in which patients tain organs based
with metastatic disease have a much poorer prognosis –– Molecular basis of metastasis to and from the
compared to those without metastatic disease. With the lung
majority of patients presenting with advanced or meta- Why does lung cancer metastasize to certain
static disease, knowledge of the process of metastasis and organs, and why do tumors from elsewhere
clinical management is critical. With advancements in metastasize to the lungs?
therapy, numerous guideline changes have occurred in the –– Sites of lung cancer metastasis
past decade. The main aim of this chapter is to shed light Lymph nodes
on some of the less commonly answered questions about Hematogenous to other organs
lung cancer metastasis. The chapter discusses the pro- –– Diagnosing and staging metastatic lung cancer
cesses behind lung cancer metastasis and the tendency for –– Management of lung cancer metastasis
certain histological types to metastasize to certain organs. Non-small cell lung cancer
The chapter then focuses on diagnosing and managing • Treatment options
lung cancer metastasis to lymph nodes and to distant • Oligometastasis
organs, with emphasis on small cell versus non-small cell Small cell lung cancer
lung carcinoma. The chapter ends with a brief description –– Surgical management of metastasis to the lung:
of oncological indications for pulmonary metastasectomy. Pulmonary metastasectomy
Oncological indications and recommendations

Learning Objectives
• Relevant clinical anatomy of the lung 41.1 Introduction
• Lymph node mapping
• Understanding lung metastases Lung cancer is a major public health problem throughout the
–– Incidence and implications world. Worldwide, lung cancer is the most common cancer,
–– Morbidity and mortality both in incidence and mortality. In the United States, lung
cancer kills more people than breast, colon, and prostate can-
cer combined. Lung cancer is the leading cause of cancer
death in both men and women. The American Cancer Society
estimates that for 2020, 229,000 Americans will be diagnosed
Z. Hammoud (*)
Department of thoracic surgery Ascension Providence Hospital, with lung cancer and roughly 136,000 deaths will be attrib-
Southfield, MI, USA uted to the disease [3]. The majority of lung cancers present
Department of surgery, Wayne State University School of at advanced stages or are metastatic (Stage IV) at the time of
Medicine, Detroit, MI, USA diagnosis, thereby limiting curative treatment options. Despite
e-mail: [email protected] notable improvements in detection and advances in treatment,
A. Popoff · C. Potti · H. Nasser overall 5-year survival remains dismal. Like other cancers,
Henry Ford Hospital, Detroit, MI, USA the prognosis in advanced disease is dramatically worse than
e-mail: [email protected]; [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 447
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_41
448 Z. Hammoud et al.

when it is detected early. Therefore, efforts aimed at earlier 41.2 Section I

detection have been undertaken. Lung cancer screening pro-
grams, now well established in Western Europe, the United 41.2.1 Lung and Lymph: Anatomy
States, Canada, and elsewhere aim to shift this survival curve
by detecting more early, curable disease. The penetrance of 41.2.1.1 Clinical Anatomy of the Lung
lung cancer screening, however, remains low at present. The functional design of the thorax facilitates the complex
Tobacco use is well established as the predominant risk process of respiration—the exchange of oxygen and carbon
factor for the development of lung cancer, accounting for dioxide. Each lung is covered by pleura, a serous membrane
up to 90% of lung cancers [3]. While not a main topic of arranged as a closed invaginated sac. The parietal pleura is
this chapter, no discussion of lung cancer can be considered constituted by costovertebral, diaphragmatic, cervical (pleu-
complete without mention of its main risk factor. Both the ral dome), and mediastinal pleura. The visceral pleura is
number of cigarettes and the duration of smoking contribute closely adherent to the lung and follows the interlobar fissures
to the risk of lung cancer. Furthermore, tobacco use is asso- to eventually continue over the hila as the parietal pleura.
ciated with certain histologies of lung cancer. For instance, The right lung has three lobes (upper, middle, and lower)
there is a strong association between tobacco use and small and two fissures (oblique and horizontal). The left lung has
cell lung cancer while nonsmokers tend to be afflicted pre- two lobes (upper and lower) separated by the oblique fis-
dominantly by adenocarcinoma. Efforts at curbing tobacco sure. The lobes are further divided into bronchopulmonary
use, such as widespread smoking cessation programs, have segments, with each segment having its own bronchus and
been met with limited success. Other causes of lung cancer blood supply.
include secondhand smoke exposure, exposure to radon gas The lungs have a dual blood supply via the pulmonary
as well as other inhaled chemicals, and asbestos exposure and bronchial vessels. Arising from the right ventricle, the
among others. main pulmonary artery divides into left and right branches
Knowledge of the anatomy, pathophysiology, and manage- at the left of midline, lying above the left main bronchus.
ment of advanced and metastatic disease is crucial for appro- Further branches tend to lie close to the segmental bron-
priate, guideline-directed clinical care. This chapter provides chi and their branches. The bronchial circulation con-
an overview of anatomy, the presentation, prognostic impli- sists of a paired group of left bronchial branches arising
cations of metastatic disease, and management of various directly from the descending thoracic aorta, and a right
clinical scenarios cancer clinicians may encounter. Accurate bronchial branch arising from a right-sided posterior inter-
staging of the patient with lung cancer is crucial to the man- costal artery, most commonly the third posterior intercos-
agement and a brief discussion of staging is undertaken in the tal artery [1]. Through an extensive capillary plexus, the
chapter. Less common clinical situations, such as the manage- pulmonary arterial system is responsible for alveolar gas
ment of oligometastatic disease, as well as combined modality exchange whereas the high-­pressure bronchial circulation,
and palliative approaches to the care of metastatic disease are which is a part of the systemic circulation, is responsible
also discussed. Other primary malignancies, such as colorec- for the nutrient supply of the parenchyma, visceral pleura,
tal, renal cell, and uterine cancers with metastases to the lung, and tracheobronchial tree. While the pulmonary vein tribu-
and pulmonary metastasectomy are also briefly addressed. taries are constituted by draining capillaries of bronchial
Surgical management of various scenarios is discussed. and pulmonary arteries, the bronchial veins drain the
Indications for other therapies are also briefly mentioned. larger bronchi.
Scientific advances in the treatment of advanced and met- Lying beneath the visceral pleura, lymphatics in the inter-
astatic disease, such as targeted therapies directed at specific lobular septa drain proximally toward the hilum, containing
tumor genetics, as well as immunotherapy, either alone or smooth muscle and valves as they get larger. On occasion,
in concert with conventional chemotherapy, have strength- lymphatics can bypass lung lymph nodes to enter the medi-
ened the armamentarium of clinicians who routinely care for astinum, likely the basis of “skip metastases” [1]. Similarly,
patients with metastatic disease. Agents directed at the epi- while bronchial lymphatics usually remain ipsilateral, they
dermal growth factor receptor (EGFR) and tyrosine kinase can cross to the contralateral mediastinum in front of lower
inhibitors (TKI) have shown efficacy in the treatment of trachea and carina. Located either within the parenchyma or
advanced and/or metastatic disease. It is likely that as more underneath the visceral pleura, intrapulmonary lymph nodes
refined therapeutics become available, such as chimeric anti- are further connected to bronchopulmonary lymph nodes
gen receptor T-cell (CAR T-cell) therapy for lung cancer and which are found at the bifurcations of segmental, lobar bron-
other solid tumors, the arsenal of treatment options will con- chi, and bifurcations of the pulmonary artery. These lymph
tinue to evolve. This, and other novel modalities, remain an nodes are connected to the lobar lymph nodes, which in turn
area of intense interest in the future to have an even greater are connected to the mediastinal lymph nodes. Mediastinal
impact on the treatment of a disease that carries what has lymph nodes consist of anterior (prevascular), tracheobron-
traditionally been a dismal prognosis.
41 Metastatic Disease of the Lung 449

chial, paratracheal, and posterior nodes in the posterior metastases,” where there is direct drainage to the mediastinal
mediastinum. lymph nodes without the involvement of the hilar and inter-
lobar lymph nodes, occur in up to 25% of cases [1].
41.2.1.2 Lymph Node Mapping
Numerous studies have mapped the lymphatic drainage of
various lung segments (see Fig. 41.1). Right upper lobe 41.2.2 Understanding Lung Metastasis
tumors usually drain to the right paratracheal lymph nodes;
left upper lobe tumors drain to the periaortic and subaor- 41.2.2.1 Incidence and Implications
tic lymph nodes; and middle and lower lobe tumors drain The 5-year survival rate in metastatic lung cancer (5%) is
to the subcarinal and right paratracheal lymph nodes. “Skip much lower than colon (64.5%), breast (89.6%), and pros-

Fig. 41.1 A map of regional

lymph nodes for the
determination of the N
descriptor during tumor, node, Supraclavicular zone
and metastasis staging of lung 1 Low cervical, supraclavicular, and
cancer, published by the sternal notch nodes
International Association for
the Study of Lung Cancer.
Abbreviations: Ao aorta; AP Superior mediastinal nodes
aortopulmonary; Eso
esophagus; mPA main Upper zone
pulmonary artery; SVC 2R Upper paratracheal (right)
superior vena cava; T trachea. 2L Upper paratracheal (left)
(Reprinted with permission
3a Prevascular
from the International
Association for the Study of 3p Retrotracheal
Lung Cancer, Staging Manual 4R Lower paratracheal (right)
in Thoracic Oncology, Orange 4L Lower paratracheal (left)
Park, FL: Editorial Rx Press,
2009) [2]

Aortic nodes

AP zone
5 Subaortic
6 Para-aortic (ascending aorta or
phrenic)

Inferior mediastinal nodes

Subcarinal zone
7 Subcarinal

Lower zone
8 Paraesophageal (below carina)
9 Pulmonary ligament

N1 nodes

Hilar / interlobar zone

10 Hilar
11 Interlobar

Peripheral zone
12 Lobar
13 Segmental
14 Subsegmental
450 Z. Hammoud et al.

tate (98.2%) cancer. Only 16% of lung cancers are diagnosed some mechanisms behind reduced aggressiveness of malig-
in the early stages [3]. Lung cancer represents a typical nancies in elderly patients have also been identified: cell pro-
example for which metastatic patients tend to have extraor- liferation is slower, the concentration of hormones and the
dinarily poorer prognosis than nonmetastatic cases [4]. responsiveness of cells to hormones are altered, and angio-
Overall, 70–95% of lung cancer patients present with locally genesis is impaired [4].
advanced or metastatic disease at diagnosis [5]. In the United In a study of 159,241 lung cancer cases by Wang et al. [4],
States, the proportions of early-stage, locally advanced, and the frequency of bone metastasis was the highest for adeno-
advanced/metastatic cases were 15.9%, 22.0%, and 57.0%, carcinoma, squamous cell carcinoma, LCLC, and NSCLC/
respectively [6]. NOS, while the liver was the most common metastatic site
in SCLC. More than 30% of adenocarcinoma and LCLC
41.2.2.2 Histological Tropism and Sites patients demonstrated distant metastasis, while squamous
of Metastasis cell cancer had the lowest rate of distal metastasis. Moreover,
Tumor hallmarks, metastatic patterns, and prognostic out- among all histological types of lung cancer, SCLC had the
comes differ greatly among different histological types highest frequency of extrathoracic metastasis, especially to
of lung cancer. Histologically, non-small cell lung cancer the liver, which is consistent with the reported data in previ-
(NSCLC), including adenocarcinoma, squamous cell carci- ous studies [13, 14]. Although the majority of brain metas-
noma, large-cell lung cancer (LCLC), and others that are not tasis was noted to be from adenocarcinoma, this has been
otherwise specified (NOS), accounts for more than 80% of attributed to the higher incidence of adenocarcinoma, while
all lung cancers [7]. Small-cell lung cancer (SCLC), account- small cell carcinoma has a higher predilection and risk for
ing for less than 20% of cases, is the most aggressive form brain metastasis [10, 12]. While large and small intestine
of lung cancer and is characterized by rapid growth and early metastasis was most commonly noted from large cell carci-
invasive spread [8]. noma, adrenal metastasis mostly arose from adenocarcinoma
Although lung cancer metastasizes hematogenously as [12]. Wang et al. further note that all single-site metastases
well as by lymphatics, Popper et al. [9] report that careful were independent of prognostic factors [4]. Multisite metas-
evaluation of resected lung carcinomas revealed that vascu- tasis studies in lung cancer established that that bone prefer-
lar invasion is often seen in low-stage tumors, which usu- entially tended to co-metastasize with liver and distal lymph
ally results in increased incidence of recurrence as well as nodes while liver metastasis was significantly correlated
shortened survival of the patient. While metastasis via lym- with distant lymph node metastasis. Combined metastases
phatics usually takes longer until distant metastases are pres- resulted in worse prognosis than the separated single-organ
ent, hematogenous spread leads to early distant metastases. metastasis [4].
Preferentially, lung carcinomas metastasize to bones, brain,
liver, and adrenal glands. The frequency of lung cancer 41.2.2.3  olecular Basis of Cancer Metastasis:
M
metastasis to other organs is as follows: bone 34.3%, lung From and to the Lung
32.1%, brain 28.4%, adrenals 16.7%, and liver 13.4% [10]. Many theories have been proposed to explain organ-specific
Clinicopathologically, metastatic patients tended to have metastasis. However, only two have withstood the test of
poorer tumor differentiation, larger tumor size, and a higher time [15]. Stephen Paget’s seed and soil hypothesis, reported
rate of regional lymph node involvement, an indicator of a in 1889, proposed that metastasis occurs as a result of favor-
more aggressive and invasive tumor biology [4]. able interactions between the cancer cells themselves (seed)
The tropism of different histological cancer types for cer- and their specific organ microenvironment (soil). Thus, only
tain tissues or organs is now better recognized. According when the seed and soil are compatible will metastasis occur.
to previously reported data, bone and brain were two lead- In 1929, James Ewing challenged this theory, postulating
ing distant targets for metastasis in lung cancer [11]. While that metastatic spread occurs strictly as a result of mechani-
hematogenous metastases were most common to liver, bone, cal factors inherent to the circulatory system and that prefer-
brain, and adrenals, the lymphatic spread was most prevalent ential metastasis can be accounted for by circulatory patterns
in hilar lymph nodes. Understanding metastatic patterns of alone [15]. There is likely an existential interplay between
lung cancer is crucial to clinical management and therapy. the two theories when applied to lung-specific metastasis.
In an autopsy study by Milovanovic et al. [12], lung adeno- A comprehensive review by Popper et al. [9] gives
carcinoma was noted to be the most frequent histological insight into the science behind this histological tro-
type of metastasizing primary lung cancer, with metastasis pism, essentially contingent on tumor microenvironment.
more common in younger patients. Biologically aggressive Metastasis requires cell migration toward higher oxygen
tumor forms in younger patients, age-dependent metabolic tension, which is based on changing the structure of the cell
dynamic differences, and survival rate differences are pos- (epithelial–mesenchymal transition), orientation within the
tulated hypotheses to explain this finding [4]. Furthermore, stroma and stromal interaction, and communication with
41 Metastatic Disease of the Lung 451

the immune system to avoid attack. Once in the blood- increases, ranging from 2 to 40% [21, 22]. There are sev-
stream, cells must survive trapping by the coagulation sys- eral well-­known risk factors for lymph node metastasis in
tem, shear stress in small blood vessels, and must find the NSCLC. These include degree of differentiation, visceral
right location for extravasation. Once outside in the meta- pleural invasion, amount of ground-glass opacity, lympho-
static locus, tumor cells have to learn the communication vascular invasion, high SUVmax on PET scan [23], and
with the “foreign” stroma cells to establish vascular supply tumor size, especially the solid tumor size [24]. Rather
and again express molecules, which induce immune toler- than acting individually, these factors interact together
ance. Although detailed description is beyond the scope and influence lymph node metastasis. Despite multiple
of this chapter, different expression of genes and proteins previous publications with differing opinions, Riquet
including upregulation of MLH 1, downregulation of COX et al. [25] concluded that lymphatic spread of NSCLC
1, FGFR4, SMAD3, p120, serine protease fibroblast acti- cannot be considered lobe-­specific. Therefore, systematic
vation protein, fibroblast growth factor receptor 1 in bone ipsilateral mediastinal lymph node dissection is necessary
metastases, cytokine, and adhesion molecule regulation to accurately determine nodal status and to interrupt lym-
are involved in metastatic patterns. Macrophage signature phatic pathways.
seems to play a key role in prognosis in metastatic lung
cancer [9, 16]. Prolonged patient survival has been corre- 41.3.1.2 Hematogenous Spread
lated with a high density of M1 macrophages, while poor
prognosis correlated with a high density of M2 macro- Bone
phages in tumor islets [17]. Mechanism of tumor migration The skeletal system is one of the most common distant
differs among histological variants of lung cancer, with sin- metastatic sites in patients with lung cancer, with 30–40%
gle-cell or small cell cluster movement noted in small cell of those with advanced lung cancer developing skeletal
and undifferentiated NSCLC, whereas movement by large metastases, mainly to the spine and ribs [26]. The reported
clusters of organized cells noted in acinar adenocarcinoma prevalence of bone metastasis at the time of lung cancer
and some cases of squamous cell carcinoma [9]. diagnosis is 15–40% [27]. Such patients may present with
metabolic disorders such as hypercalcemia, pathologic
fractures, and spinal cord compression, any of which
41.3 Section II reduces median survival from 4.4 months to 2.8 months in
one study [26]. Sugiura et al. [28] estimated the survival
The overall median survival of metastatic lung cancer was rates in the patients with bone metastases at 9.7 months
noted to be 5 months, ranging from less than 3 months to after diagnosis. Better prognosis was recorded in women,
over 12 months [18]. In one of the largest studies on meta- in individuals where adenocarcinoma was the histological
static locations, 18% of patients had lung metastasis, 16% type, in cases of solitary bone metastasis, better perfor-
bone metastasis, 12% brain metastasis, 7% liver metastasis, mance status, lower number of skeletal metastases, use of
and 6% adrenal gland metastasis at diagnosis [19]. Adding an epithelial growth factor receptor inhibitor, as well as
to previous reports that have associated liver metastasis with in the individuals without pathological fractures. Worse
a short survival, Riihimaki et al. [20] reported comparable prognosis was noted in patients with elevated CA-125 and
survival between metastasis to the nervous system and to the alkaline phosphatase (ALP) [29]. A possible mechanism
respiratory system and adrenals, whereas survival was poor for spinal metastasis is access to the vertebral bodies in
in bone and liver metastasis, especially from LCLC. They the thoracic and lumbar spine through the plexus vertebral
also noted a better prognosis in younger patients and women. system and the high bone marrow flow of some skeletal
Younger cancer patients are more often diagnosed with elements [30].
advanced metastatic involvement to lymph nodes and other Bone metastasis from lung cancer are treated palliatively
organs, possibly because tumors need to be more aggressive not only because of poorer prognosis when compared to other
to escape immunosurveillance in the young immunocompe- cancers but also because of decreased pulmonary capacity
tent individual [20]. to tolerate surgical procedures in this patient population.
Systemic therapies are found to have a modest impact on
survival [30]. Receptor activator for nuclear factor-κβ ligand
41.3.1 Sites of Lung Cancer Metastases (RANKL) inhibitors have demonstrably superior efficacy
over bisphosphonates that block osteoclast activity in reduc-
41.3.1.1 Lymphatic Spread ing the incidence of skeletal-related events in NSCLC in a
Lymph nodes are among the first metastatic foci of lung randomized trial [26]. Radiotherapy is considered standard
carcinomas [9]. Generally, the frequency of lymph node for small lesions, whereas surgical management is reserved
metastasis has been shown to increase as the tumor size for fractures or impending fractures.
452 Z. Hammoud et al.

Brain of adenocarcinoma in the first year of diagnosis had adre-

Brain metastases are thought to occur via seeding of circu- nal metastases. This is attributed to the likelihood of serious
lating tumor cells into the brain microvasculature; within complications such as bleeding or adrenal insufficiency [12].
this unique microenvironment, tumor growth is promoted
and the penetration of systemic medical therapies is lim- Other Metastasis
ited. Development of brain metastases remains a substan- Other rare metastases, such as gastric, intestinal, and pancre-
tial contributor to overall cancer mortality [31]. About 25% atic metastases, were found in higher percentage in autopsy
of patients with lung carcinoma show brain metastases in studies but are rarely diagnosed in clinical settings [12].
advanced stages of the disease. Solitary brain metastasis Autopsy studies [12, 36] note that intestinal metastases were
may clinically present with seizures, symptoms of increased most often from LCLC, while adenocarcinoma and squa-
intracranial pressure (e.g., headache), or other nonspecific mous cell carcinoma were rarely the sources of intestinal
symptoms even prior to diagnosis of lung cancer [12]. These metastases. Patients with intestinal metastases died within
have a poor prognosis that strongly depends on the number the first year after clinical diagnosis of lung cancer.
of brain metastases, presence of metastases in other organs,
and the stage of the primary lung cancer [32]. Although more
cases of brain metastases have been reported previously with 41.4 Section III
adenocarcinoma given its higher incidence, SCLC has the
highest risk of brain metastases [32]. In recent years, brain 41.4.1 Diagnosing Lung Cancer Metastasis
metastases are increasingly seen in adenocarcinomas with
epidermal growth factor receptor (EGFR) mutations and In patients with NSCLC, determining the extent of disease
EML4ALK1 rearrangement, whereas squamous cell carci- will strongly impact management and prognosis. While pal-
nomas have a tendency to locally invade the thoracic wall liation is the objective of management of distant metasta-
[9]. ses, localized disease is treated with an intention to cure,
especially based on lymph node involvement. The American
Liver College of Chest Physicians guidelines recommends surgical
SCLC has been associated with an increased metastatic resection for patients with stage IA, IB, IIA, and IIB while
potential to liver and brain [12, 20, 33]. The incidence of patients stages III and IV rarely meet criteria for surgery
liver metastasis in newly diagnosed SCLC patients is as high [37]. In patients with clinical stage III or IV NSCLC, routine
as 17% compared to 4% in NSCLC patients [33]. Liver is a imaging of the brain with head MRI is recommended, even
frequent metastatic focus in other histological types [12, 20]. if they have a negative clinical evaluation.
However, in autopsy studies, the incidence of liver metastasis While the potential for surgical resection exists in patients
in patients with lung cancer has been reported to be 33–51% with NSCLC, patients with SCLC are rarely treated with
[12, 33]. Extensive and diffuse metastases of SCLC in the resection and are most often treated with systemic chemo-
liver parenchyma were reported to potentially cause fulmi- therapy plus/minus radiation therapy.
nant hepatic insufficiency (jaundice, coagulopathy, encepha- When there is a strong suspicion of lung cancer, stag-
lopathy, and organs dysfunction) with poor prognosis and ing workup may be pursued prior to or simultaneously
often quick lethal outcome [12]. Lung cancer may also cause while confirming diagnosis [38]. Extensive clinical evalu-
biliary tract obstruction by metastasizing to lymph nodes; ation for risk factors, history, and physical examination is
this should be differentiated from parenchymal liver failure the best predictor of distant metastasis. Site-specific symp-
as systemic chemotherapy might relieve this obstruction toms warrant directed evaluation with appropriate studies
while parenchymal involvement has a poor prognosis with [37]. Staging modalities are of varying sensitivities and
some chemotherapeutic agents worsening liver failure [33]. specificities (see Table 41.1). Positron emission tomogra-
phy–computed tomography (PET-CT) scan has emerged as
Adrenal a standard recommendation for the staging of lung cancer.
Following the initial diagnosis of NSCLC, metastasis to PET can diagnose unsuspected mediastinal and/or distant
the adrenal glands was reported in 10% of cases [34]. metastasis and thus is critical to determining accurate stage.
Furthermore, in advanced stages of the disease, NSCLC Patients with apparent peripheral early-stage (Ia) disease on
metastases are usually bilateral and appear in 40% of patients computed tomography (CT) or those with clear evidence
[35]. Survival in patients with adrenal metastasis is compa- of advanced disease or metastasis on CT may not require a
rable to patients with metastasis to the nervous system and PET scan or preoperative invasive mediastinal staging [37].
respiratory system from lung [20]. In an autopsy study, it If needle biopsy techniques are negative, surgical biopsy is
was noted that 60% of adrenal metastases originated from recommended. Tissue biopsy confirmation is recommended
lung adenocarcinoma, and more than 50% of those who died for all abnormal scans, with more evidence pointing toward
41 Metastatic Disease of the Lung 453

Table 41.1 Lymph node staging modalities Gelberg J, Grondin S, Pembrolizumab is also recommended for patients with meta-
Tremblay A. Mediastinal staging for lung cancer. Can Respir J.
static NSCLC and PD-L1 levels of 1–49% who cannot toler-
2014;21(3):159–161. doi:10.1155/2014/890108 (reprinted with per-
mission) [39] ate platinum-based chemotherapy; in patients with negative
%
test results for EGFR, ALK, ROS1, or BRAF genetic altera-
Lymph node
Staging modality access Sensitivitya Specificitya NPVa tions, regardless of their PD-L1 expression levels [41].
Noninvasive Multiple reports suggest that a subset of these patients
CT chest All 55 81 83 with oligometastatic disease (generally defined as five or
PET All 80 88 91 fewer metastatic sites) can benefit from definitive manage-
PET-CT All 62 90 90 ment of the disease with improved long-term survival [42].
Invasive Furthermore, synchronous versus metachronous presentation
2R, 2L, 4R,
Mediastinoscopyb 78 100c 91
(metachronous being more favorable), nodal status, number,
4L, 7
EBUSb 2R, 2L, 4R, 89 100c 91 and location of oligometastatic sites have all been shown to
4L, 7, 10R, impact survival [43].
10L, 11R, 11L
EUS 4L, 7, 5, 8, 9 89 100c 86 Intracranial Oligometastasis
Combined EBUS/ 2R, 2L, 4R, 91 100c 96
Patel et al. [42] reported that surgical resection of oligometa-
EUS 4L, 7, 10R,
10L, 11R, 11L, static brain metastases results in long-term survival rates of
5, 8, 9 up to 37%. Solitary brain metastasis should be considered for
Source: American College of Chest Physicians lung cancer guidelines, surgical resection [44]. Additionally, it is recommended that
3rd edn (1). CT computed tomography; EUS endoscopic ultrasound; patients with neurological symptoms undergo local therapy
PET positron emission tomography for intracranial disease in the form of whole-brain radiation
a
Sensitivity, specificity and negative predictive value (NPV) for each
staging modality were obtained from studies that may have different therapy (WBRT) or surgery as appropriate, while consider-
patient populations and direct comparison between modalities based on ing performance status and other prognostic factors [44].
these values alone should be avoided
b
Studies directly comparing endobrachial ultrasound (EBUS) versus
Extracranial Oligometastasis
mediastinoscopy are described in the text
c Improved long-term survival rates ranging from 23 to 86%
Results cannot be verified in studies reporting 100% specificity because
no follow-up test was performed to confirm the results with surgical resection of extracranial oligometastatic
NSCLC has been demonstrated, with higher survival in
metachronous than synchronous disease [42]. Additionally,
improved patient outcomes with more complete staging. the efficacy of less invasive surgery in the resection of adre-
Furthermore, significantly lower risk of death is reported nal and contralateral lung oligometastatic disease may allow
in patients with bi- and trimodality staging versus single-­ for fewer complications from such resection. Radiation ther-
modality staging [40]. apy has also shown a survival advantage in metachronous
oligometastatic disease [42]. However, performance status
should be carefully considered when deciding on therapeutic
41.4.2 Managing Lung Cancer Metastasis options.

41.4.2.1 Non-small Cell Lung Cancer (NSCLC) 41.4.2.2 Small Cell Lung Cancer (SCLC)
For management of metastatic NSCLC, the National SCLC is generally rapidly growing with a tendency for early
Comprehensive Cancer Network (NCCN) guidelines recom- metastases and high recurrence rates. It typically presents
mend first establishing a histological subtype, followed by with bulky symptomatic masses and mediastinal involve-
molecular testing for EGFR mutation, ALK, ROS1, BRAF, ment, often associated with a number of neurologic and endo-
and PD-L1. Chemotherapy with platinum-based agents crine paraneoplastic syndromes, with extrathoracic spread in
has emerged as the first line of management in metastatic 75–80% patients at presentation [45]. Limited-­stage small-
NSCLC. However, more focus is currently being placed on cell lung cancer (LS-SCLC) is defined as tumor confined to
upcoming immunotherapeutic agents. In the 2018 NCCN 1 hemithorax, with or without regional lymph-­node involve-
guidelines update: crizotinib was recommended in ROS1 ment, which can be safely encompassed in a tolerable radia-
rearrangements; alectinib, ceritinib, and crizotinib in ALK tion field. Extensive-stage small cell lung cancer (ES-SCLC)
rearrangement; dabrafenib with trametinib with or without is defined as disease that cannot be safely encompassed
cytotoxic agents in BRAF V600E mutations. In the 2020 in a tolerable radiation field [45]. The median survival for
update, NCCN recommends single-agent pembrolizumab LS-SCLC is 18–24 months with a 5-year survival of 20–25%.
as a first-line therapy option for certain patients with met- For ES-SCLC, the median survival is 9–10 months, with 10%
astatic NSCLC and PD-L1 expression levels of ≥50%. 2-year survival [45]. Chemotherapy and radiation form the
454 Z. Hammoud et al.

mainstay of treatment for SCLC, with very rare exception of cohort of patients maximally benefitting from lung metasta-
certain cases of LS-SCLC where surgical management with sectomy [48]. Active primary cancer, extrathoracic metasta-
intention to cure is preferred. Chest radiation and prophylactic ses, incomplete surgical resection, and mediastinal lymphatic
cranial radiation (25 Gy over 10 daily fractions) have shown spread are negative prognostic features [47].
increased survival. Immunotherapy, including nivolumab with In urological cancers, metastasectomy for single metas-
or without ipilimumab, is now available for refractory disease. tasis was statistically related to longer time to progression
A staging evaluation consists of medical history and especially in patients who previously responded to chemo-
physical examination, complete blood count (CBC), and therapy, with resectable tumor, and documented period of
comprehensive chemistry panel with renal and hepatic func- stability [48]. Lung metastasectomy has been shown to pro-
tion tests, CT of the chest and abdomen with intravenous long survival and should always be considered when feasible
contrast, MRI or CT of the brain, and bone scan. PET is usu- in patients with sarcoma, given that up to 20% with soft tis-
ally added as well. sue sarcoma and 40% with osteosarcoma present with iso-
LS-SCLC is typically treated with chemoradiation: a com- lated pulmonary metastases [48]. Younger age at diagnosis
bination of etoposide and cisplatin administered concomitantly and low-grade tumors with longer disease-free interval offer
with thoracic irradiation. The NCCN and American College the greatest survival advantage to radically resected patients
of Chest Physicians (ACCP) guidelines suggest that there [49]. Five-year survival after lung metastasectomy of head
may also be a role for surgery in the management of patients and neck cancers is reported to be 50–60% [50] with worse
with very early-stage SCLC (stage I) generally without nodal outcomes in older age, squamous cell carcinoma, and short
involvement, with some studies showing an improved 5-year disease-free interval and mediastinal lymph nodal involve-
survival [46]. Prophylactic cranial irradiation is recommended ment [51]. In gynecological cancers, NCCN recommends
in patients who had a response to initial treatment [46]. In surgical resection for resectable regional uterine cancer
LS-SCLC, with T3 or node positivity, concurrent chemoradia- metastasis, with a 5-year survival rate of 32.9–46.8% [48].
tion with prophylactic cranial irradiation is recommended in Overall, lung metastasectomy, when feasible, has shown to
patients who had a response to initial treatment. improve long-term survival and prognosis.
For ES-SCLC, systemic chemoradiation with a combi-
nation of platinum agent and etoposide and organ-specific
therapy (radiation or surgery) is recommended. Prophylactic 41.5 Conclusions
cranial irradiation is controversial in this patient population.
Although there is 40–70% objective response rate to che-
motherapy, 2-year and 5-year survival in ES-SCLC remains 1. Lung cancer remains the most common cause of cancer-
<5% and <2%, respectively [46]. related mortality among both men and women.
2. An understanding of the common sites of metastasis,
their detection, and their management is critical to
41.4.3 Resectability of Metastases to the Lung improving outcomes.
3. The lung is a common metastatic site for other types of
Lung is the second most common site of metastases, mak- cancers.
ing it important to define management and resectability. The 4. The understanding and management of metastases to the
present oncological criteria for pulmonary metastasectomy lung from other cancers may also lead to improved out-
include [47]: comes in these cancers.

(I) Primary cancer is controllable

(II) No extrathoracic metastasis that is not controlled or Open Questions
controllable • What diagnostic modalities will improve staging in
(III) All of the tumor must be resectable, with adequate pul- lung cancer?
monary reserve • What explains the tropism of certain lung cancers to
(IV) No alternative medical treatment options with lower specific organs?
morbidity. • Which genes/proteins are responsible for metastasis
in lung cancer?
Generally, favorable prognostic features in patients with • Which targeted agents will prove efficacious in
pulmonary metastases include one or few metastases, long treating lung cancer?
disease-free interval, and normal CEA in colorectal cancers • What makes the lung a suitable host organ for
[49]. Radical resection, histology, and disease-free inter- metastasis from other cancers?
val seem to be independent prognostic factors identifying a
41 Metastatic Disease of the Lung 455

References 20. Riihimäki M, Hemminki A, Fallah M, Thomsen H, Sundquist K,

Sundquist J, et al. Metastatic sites and survival in lung cancer. Lung
Cancer. 2014;86:78–84.
1. Susan Standring MBE, PhD, DSc, FKC, Hon FAS, Hon FRCS,
21. Oda M, Watanabe Y, Shimizu J, et al. Extent of mediastinal node
Gray’s Anatomy, Chapter 54, pp. 953–69.e2.
metastasis in clinical stage I non-small-cell lung cancer: the role of
2. The International Association for the Study of Lung Cancer,
systematic nodal dissection. Lung Cancer. 1998;22:23–30.
Staging Manual in Thoracic Oncology, Orange Park, FL: Editorial
22. Ohta Y, Oda M, Wu J, et al. Can tumor size be a guide for limited
Rx Press; 2009.
surgical intervention in patients with peripheral non-small cell lung
3. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer
cancer? Assessment from the point of view of nodal micrometasta-
J Clin. 2019;69:7–34. https://doi.org/10.3322/caac.21551.
sis. J Thorac Cardiovasc Surg. 2001;122:900–6.
4. Wang X, Wang Z, Pan J, et al. Patterns of extrathoracic metas-
23. Miyata Y, Tsutani Y, Okada M. Use of high-resolution computed
tases in different histological types of lung cancer. Front Oncol.
tomography and positron emission tomography/computed tomog-
2020;10:715. Published 2020 May 19. https://doi.org/10.3389/
raphy in the management of stage IA adenocarcinoma. Semin
fonc.2020.00715.
Thorac Cardiovasc Surg. 2012;24:267–74.
5. Araujo LH, Baldotto C, Castro G Jr, Katz A, Ferreira CG, Mathias
24. Seok Y, Yang HC, Kim TJ, Lee KW, Kim K, Jheon S, Cho
C. Lung cancer in Brazil. J Bras Pneumol. 2018;44(1):55–64.
S. Frequency of lymph node metastasis according to the size of
https://doi.org/10.1590/s1806-­37562017000000135.
tumors in resected pulmonary adenocarcinoma with a size of
6. Cancer Research UK. Lung cancer statistics. London: Cancer
30 mm or smaller. J Thorac Oncol. 2014 Jun;9(6):818–24. https://
Research UK; 2015.
doi.org/10.1097/JTO.0000000000000169.
7. Goldstraw P, Ball D, Jett JR, Le Chevalier T, Lim E, Nicholson
25. Riquet M, Rivera C, Pricopi C, Arame A, Mordant P, Foucault C,
AG, et al. Non-small-cell lung cancer. Lancet. 2011;378:1727–40.
Dujon A, Le Pimpec-Barthes F. Is the lymphatic drainage of lung
https://doi.org/10.1016/S0140-­6736(10)62101-­0.
cancer lobe-specific? A surgical appraisal. Eur J Cardiothorac Surg.
8. Rudin CM, Ismaila N, Hann CL, Malhotra N, Movsas B, Norris
2015;47(3):543–9. https://doi.org/10.1093/ejcts/ezu226.
K, et al. Treatment of small-cell lung cancer: American society of
26. Rosen LS, Gordon D, Tchekmedyian NS, Yanagihara R, Hirsh V,
clinical oncology endorsement of the American college of chest
Krzakowski M, Pawlicki M, De Souza P, Zheng M, Urbanowitz
physicians guideline. J Clin Oncol. 2015;33:4106–11. https://doi.
G, et al. Long-term efficacy and safety of zoledronic acid in the
org/10.1200/JCO.2015.63.7918.
treatment of skeletal metastases in patients with nonsmall cell lung
9. Popper HH. Progression and metastasis of lung cancer. Cancer
carcinoma and other solid tumors: a randomized, phase III, double-­
Metastasis Rev. 2016;35(1):75–91. https://doi.org/10.1007/
blind, placebo-controlled trial. Cancer. 2004;100(12):2613–21.
s10555-­016-­9618-­0.
https://doi.org/10.1002/cncr.20308.
10. Tamura T, Kurishima K, Nakazawa K, Kagohashi K, Ishikawa H,
27. Oliveira MB, Mello FC, Paschoal ME. The relationship between
Satoh H, Hizawa N. Specific organ metastases and survival in meta-
lung cancer histology and the clinicopathological characteristics of
static non-small-cell lung cancer. Mol Clin Oncol. 2015;3:217–21.
bone metastases. Lung Cancer. 2016;96:19–24.
11. Hattori A, Takamochi K, Oh S, Suzuki K. New revisions and cur-
28. Sugiura H, Yamada K, Sugiura T, Hida T, Mitsudomi T. Predictors
rent issues in the eighth edition of the TNM classification for non-­
of survival in patients with bone metastasis of lung cancer. Clin
small cell lung cancer. Jpn J Clin Oncol. 2019;49:3–11. https://doi.
Orthop Relat Res. 2008;466(3):729–36. https://doi.org/10.1007/
org/10.1093/jjco/hyy142.
s11999-­007-­0051-­0.
12. Milovanovic IS, Stjepanovic M, Mitrovic D. Distribution patterns
29. Zhou Y, Yu QF, Peng AF, et al. The risk factors of bone metasta-
of the metastases of the lung carcinoma in relation to histologi-
ses in patients with lung cancer. Sci Rep. 2017;7:8970. https://doi.
cal type of the primary tumor: an autopsy study. Ann Thorac Med.
org/10.1038/s41598-­017-­09650-­y.
2017;12(3):191–8. https://doi.org/10.4103/atm.ATM_276_16.
30. Cho YJ, Cho YM, Kim SH, Shin KH, Jung ST, Kim HS. Clinical
13. Cai H, Wang H, Li Z, Lin J, Yu J. The prognostic analysis of
analysis of patients with skeletal metastasis of lung cancer. BMC
different metastatic patterns in extensive-stage small-cell lung
Cancer. 2019;19(1):303. . Published 2019 Apr 3. https://doi.
cancer patients: a large population-based study. Future Oncol.
org/10.1186/s12885-­019-­5534-­3.
2018;14:1397–407. https://doi.org/10.2217/fon-­2017-­0706.
31. Achrol AS, Rennert RC, Anders C, Soffietti R, Ahluwalia MS,
14. Nakazawa K, Kurishima K, Tamura T, Kagohashi K, Ishikawa H,
Nayak L, Peters S, Arvold ND, Harsh GR, Steeg PS, Chang
Satoh H, et al. Specific organ metastases and survival in small cell
SD. Brain metastases. Nat Rev Dis Primers. 2019;5(1):5. https://
lung cancer. Oncol Lett. 2012;4:617–20. https://doi.org/10.3892/
doi.org/10.1038/s41572-­018-­0055-­y.
ol.2012.792.
32. Arrieta O, Villarreal-Garza C, Zamora J, Blake-Cerda M, de la Mata
15. Piaseczny MM, Allan AL. Why does breast cancer often spread to
MD, Zavala DG, et al. Long-term survival in patients with non-­
the lung? Womens Health. 2014;10:561–4. https://doi.org/10.2217/
small cell lung cancer and synchronous brain metastasis treated
WHE.14.45.
with whole-brain radiotherapy and thoracic chemoradiation. Radiat
16. Sarode P, Schaefer MB, Grimminger F, Seeger W, Savai
Oncol. 2011;6:166.
R. Macrophage and tumor cell cross-talk is fundamental for lung
33. Kagohashi K, Satoh H, Ishikawa H, Ohtsuka M, Sekizawa K. Liver
tumor progression: we need to talk. Front Oncol. 2020;10:324. .
metastasis at the time of initial diagnosis of lung cancer. Med
Published 2020 Mar 11. https://doi.org/10.3389/fonc.2020.00324.
Oncol. 2003;20:25–8.
17. Cao L, Che X, Qiu X, Li Z, Yang B, Wang S, et al. M2 macrophage
34. Allard P, Yankaskas BC, Fletcher RH, Parker LA, Halvorsen RA. Jr
infiltration into tumor islets leads to poor prognosis in non-small-­
Sensitivity and specificity of computed tomography for the detec-
cell lung cancer. Cancer Manag Res. 2019;11:6125–38. https://doi.
tion of adrenal metastatic lesions among 91 autopsied lung cancer
org/10.2147/CMAR.S199832.
patients. Cancer. 1990;66:457–62.
18. Socinski M, Morris D, Masters G, Lilenbaum R. Chemotherapeutic
35. Matthews MJ. Problems in morphology and behavior of broncho-
management of stage IV non-small cell lung cancer. Chest.
pulmonary malignant disease. In: Lung cancer: natural history,
2003;123:226S–43S.
prognosis and therapy. New York: Academic Press; 1976.
19. Oikawa A, Takahashi H, Ishikawa H, Kurishima K, Kagohashi K,
36. McNeill PM, Wagman LD, Neifeld JP. Small bowel metastases
Satoh H. Application of conditional probability analysis to distant
from primary carcinoma of the lung. Cancer. 1987;59:1486–9.
metastases from lung cancer. Oncol Lett. 2012;3:629–34.
456 Z. Hammoud et al.

37. Silvestri GA, Gonzalez AV, Jantz MA, Margolis ML, Gould MK, brain metastasis(es): an American Society for Radiation Oncology
Tanoue LT, Harris LJ, Detterbeck FC. Methods for staging non-­ evidence-based guideline. Pract Radiat Oncol. 2012;2:210–25.
small cell lung cancer: diagnosis and management of lung cancer, 45. Wang S, Zimmermann S, Parikh K, Mansfield AS, Adjei
3rd ed: American College of Chest Physicians evidence-based clini- AA. Current diagnosis and management of small-cell lung cancer.
cal practice guidelines. Chest. 2013 May;143(5 Suppl):e211S–50S. Mayo Clin Proc. 2019;94(8):1599–622. https://doi.org/10.1016/j.
https://doi.org/10.1378/chest.12-­2355. mayocp.2019.01.034.
38. Rivera MP, Mehta AC, Wahidi MM. Establishing the diagnosis of 46. Yu JB, Decker RH, Detterbeck FC, Wilson LD. Surveillance epide-
lung cancer: diagnosis and management of lung cancer, 3rd ed: miology and end results evaluation of the role of surgery for stage I
American College of Chest Physicians evidence-based clinical small cell lung cancer. J Thorac Oncol. 2010;5(2):215–9.
practice guidelines. Chest. 2013;143(5 suppl):e142S–65S. 47. Royal Brompton & Harefield NHS Foundation Trust, University of
39. Gelberg J, Grondin S, Tremblay A. Mediastinal staging for Cambridge, University of Sussex. A randomised Trial of Pulmonary
lung cancer. Can Respir J. 2014;21(3):159–61. https://doi. Metastasectomy in Colorectal Cancer (PulMiCC). https://clinical-
org/10.1155/2014/890108. trials.gov/ct2/show/NCT01106261.
40. Farjah F, Flum DR, Ramsey SD, Heagerty PJ, Symons RG, Wood 48. Treasure T, Milošević M, Fiorentino F, et al. Pulmonary metas-
DE. Multi-modality mediastinal staging for lung cancer among tasectomy: what is the practice and where is the evidence for
medicare beneficiaries. J Thorac Oncol. 2009;4(3):355–63. effectiveness? Thorax. 2014;69:946–9. https://doi.org/10.1136/
41. Ettinger DS, Wood DE, Aggarwal C, Aisner DL, Akerley W, thoraxjnl-­2013-­204528.
Bauman JR, Bharat A, Bruno DS, Chang JY, Chirieac LR, et al. 49. Giuliano K, Sachs T, Montgomery E, et al. Survival following lung
NCCN guidelines insights: non-small cell lung cancer, version metastasectomy in soft tissue sarcomas. Thorac Cardiovasc Surg.
1.2020. J Natl Compr Cancer Netw. 2019;17(12):1464–72. 2016;64:150–8. https://doi.org/10.1055/s-­0035-­1563538.
42. Patel AN, Simone CB 2nd, Jabbour SK. Risk factors and management 50. Haro A, Yano T, Yoshida T, et al. Results of a surgical resection of
of oligometastatic non-small cell lung cancer. Ther Adv Respir Dis. pulmonary metastasis from malignant head and neck tumor. Interact
2016;10(4):338–48. https://doi.org/10.1177/1753465816642636. Cardiovasc Thorac Surg. 2010;10:700–3. https://doi.org/10.1510/
43. Pirker R, Pereira J, Szczesna A, Von Pawel J, Krzakowski M, icvts.2009.219766.
Ramlau R, et al. Prognostic factors in patients with advanced non-­ 51. Galetta D, Petrella F, Leo F, et al. Treatment of pulmonary metastases
small cell lung cancer: data from the phase III FLEX study. Lung from primary intraosseous odontogenic carcinoma. Lancet Oncol.
Cancer. 2012;77:376–82. 2006;7:272–3. https://doi.org/10.1016/S1470-­2045(06)70620-­9.
44. Tsao M, Rades D, Wirth A, Lo S, Danielson B, Gaspar L, et al.
Radiotherapeutic and surgical management for newly diagnosed
 Part XI
Cancer Metastasis to the Liver

Alex C. Kim and Clifford S. Cho

Cancer metastasis involves a complex pathologic cellular process that allows for malignant
cells to travel from the primary tumor site to regional lymph nodes or distant organs. It is the
major contributor to host morbidity and is responsible for 90% of cancer-related mortality [1,
2]. The mechanism of metastatic cascade consisting of initiation, migration, colonization, and
re-propagation is a very complex process. Detailed description of the cascade is beyond the
scope of this chapter. The concepts introduced here are intended to provide a general overview
of hypotheses that are viewed as important in hepatic metastasis.
The liver is a unique, multifunctional organ responsible for metabolic homeostasis, bile
synthesis, and paracrine and endocrine secretions. Anatomically, the liver derives its blood
supply from both the splanchnic (portal) and systemic circulations. The portal system derives
its blood volume from intra-abdominal organs and is very rich in nutrients and metabolic by-­
products. The hepatic artery delivers well-oxygenated blood from the systemic circulation to
the liver. The liver is further organized into eight Couinaud segments with each possessing its
independent portal and systemic vascular inflow, hepatic outflow, and biliary pedicles. At the
level of the parenchyma, hepatocytes and cholangiocytes make up the majority of the liver
and accommodate for metabolic detoxification, bile synthesis, and endocrine function.
Additional mesenchymal and hematopoietic derived cells make up the rest of the liver, and
mediate liver microcirculation and regenerative, inflammatory, and immune functions. The
cells are arranged in a complex, dense network of polyhedral shaped hepatic lobules com-
posed of peripheral portal triads and a central vein. The portal triad contains the hepatic
arteriole branch, portal venous branch, bile ducts, lymphatics, and nerve branch. While the
microcirculation flows through the hepatic arterioles and the portal venules and drains through
the central vein, the lymphatic and bile flows are in the opposite direction. This evolutionarily
conserved complex microvascular network is a perfectly fertile environment for entry and
seeding of malignant cells, and provides a mechanism for tumor growth and proliferation and
for immune escape or tolerance. As such, the liver serves as an ideal niche for cancer metas-
tasis, especially for colorectal cancer.
One of the hallmarks of cancer is the ability to invade and to metastasize [2–4]. In order to
disseminate to distant tissue and to establish metastatic colonies, neoplastic cells commandeer
a complex biologic program known as epithelial-mesenchymal transition (EMT) [5–8]. EMT
is a dynamic molecular process that triggers a transient switch in cellular phenotype from a
polarized, differentiated epithelial state to a mobile, dedifferentiated mesenchymal state. This
process is readily evident during embryogenesis and wound healing, and allows for spatial and
temporal cellular distribution and generation of new organs or tissues [4, 6, 9]. During the
initiation of invasion, aberrant activation of EMT is hypothesized to be the principal mecha-
nism for the acquisition of metastatic phenotype by epithelial tumors [5, 6]. The transition
allows for nonmetastatic cells to suppress their epithelial/differentiated state (loss of cell-cell
458 Cancer Metastasis to the Liver

adhesion and polarity) while activating their mesenchymal/dedifferentiated state (gain of motility, invasion,
and stemness) [9]. The activation of EMT is dependent on a convergence of various cellular signals from
the tumor microenvironment, including the TGF-β signaling cascade [4, 7, 9, 10]. The resultant induction
of intracellular signaling pathways provokes expression and activation of multiple EMT transcription fac-
tors, such as TWIST1 and SNAI1, to trigger the cellular transition [4, 6, 9]. In addition, activation of these
factors, especially in colorectal cancer, can trigger an undifferentiated state through TGF-β signaling that
allows cells to behave as cancer stem cells [11]. These cells gain an ability to self-renew, produce differenti-
ated progeny, and re-metastasize to a new site. The cancer stem cell theory further explains that these cells
are thought to be resilient to cytotoxic and hypoxic conditions that allow survival until they arrive at their
destination organ [12]. Colorectal cancer cells with tumor-­propagating ability are characterized by cell
surface molecules including CD44, CD133, and LGR5 [12–16]. CD44 is a cell surface adhesion peptide
and consists of various isoforms stemming from alternate splicing products [17]. Particularly, CD44v6
gained wide interest due to its ability to bind to human growth factor (HGF) and vascular endothelial
growth factor (VEGF) [12, 18, 19]. These interactions allow for activation of downstream signal transduc-
tion pathways that cause tumor proliferation and angiogenesis. Increased CD44v6 expression is often cor-
related with advanced stages of colon cancer and is associated with poorer prognosis [20]. CD133 is another
molecule expressed particularly in colon cancer stem cells [12]. Cells positive for this molecule are able to
initiate tumor formation in in vivo models. Only a small percentage of cells are positive for these molecules,
supporting their representation of rare populations of cancer stem cells [12, 21]. LGR5 was identified as the
defining marker for colonic intestinal adult stem cell populations [16]. This molecule is a downstream target
for the WNT pathway and also serves as the receptor for R-spondin [16, 22]. It is hypothesized that the
constitutive activation of the β-catenin pathway in LGR5 results in the transformation of adult stem cells
into cancer stem cells [23]. Through EMT, these cells gain the ability to migrate and establish distant meta-
static colonies in favorable sites like the liver.
In 1889, Stephen Paget postulated that cancer metastasis requires an optimal interaction between “seeds,”
or the metastatic cells, and the “soil,” or the premetastatic niche [24]. The entry of the metastatic cells into
the liver and their subsequent colonization are a result of bidirectional interaction between malignant cells
and liver microenvironment. This is often referred to as a premetastatic niche, or a microenvironment that
is permissive for metastatic growth [25, 26]. The hepatic microcirculation consists of hepatic sinusoids,
Kupffer cells, and extracellular matrix (ECM), all representing an example of a premetastatic niche. In
murine models of metastatic pancreatic ductal adenocarcinoma, it was recently observed that exosomes and
IL-6 released from primary pancreas cancers, respectively, trigger signaling pathways and protein expres-
sion changes in distant Kupffer cells and hepatocytes that make the liver more receptive to infiltration and
growth of metastatic cancer cells—demonstrating a paracrine cross talk that evolves between primary
tumors and liver which favors the development of liver metastases [27, 28]. The initial implantation of can-
cer cells induces local inflammatory responses. Particularly, neutrophils are hypothesized to play a signifi-
cant role in the infiltration of cancer cells to the metastatic site. Activation of neutrophils at the site of injury
results in the release of highly adhesive neutrophil extracellular traps (NETs) composed of DNA and anti-
microbial enzymes. It is hypothesized that these NETs may facilitate early adhesion of cancer cells into the
metastatic niche of the liver [29, 30]. The subsequent ECM remodeling through enzymes such as tissue
inhibitor of metalloproteninases-1 (TIMP-1) and the increased expression of molecules including E-selectin
and vascular cell adhesion molecule (VCAM)-1 under the control of cytokine and signaling pathway activa-
tion by Kupffer cells are thought to enable further implantation of metastatic cells [28, 31]. Indeed, block-
ade of this inflammation-induced cell surface molecule expression has resulted in decreased liver metastasis
[32–34].
The next steps in hepatic metastasis involve early proliferation and cellular growth within the hepatic
microcirculation. Under normal conditions, hepatic stellate cells (HSC) are quiescent cells that contribute
to extracellular matrix turnover and participate in liver homeostasis through the production of stimulating
factors [35]. Upon contact with cancer cells, HSCs are activated and undergo differentiation into myofibro-
blasts, with the expression of a-SMA, vimentin, and PDGFR-a/b [36]. Activated myofibroblasts are in turn
responsible for remodeling of the microenvironment, with production of desmoplastic stroma. In addition,
Cancer Metastasis to the Liver 459

they promote metastasis through the secretion of potent mitogens including TGF-β, enabling further recruit-
ment of stromal cells for colony establishment [35, 37]. The activated TGF-β signaling pathway addition-
ally promotes increased expression of metalloproteinases that allow for metastatic cell migration and
invasion, promote antitumor responses by downregulating inflammatory cells including NK cells, and acti-
vate angiogenesis by increasing production of VEGF [37].
Upon establishment of the initial metastatic colony, the activation of angiogenesis becomes crucial for
the survival of these cells [38]. The local injury caused by cancer cell entrapment and disruption of micro-
vasculature results in paracrine stimulation from local cells including myofibroblasts. The resulting hypoxia
causes increased expression of hypoxia-inducible factor 1-α (HIF1-α), promoting the secretion of pro-­
angiogenic factors such as VEGF [39]. Subsequent interaction between HSC, myofibroblasts, and tumor
cells results in further secretion of IL-6, prostaglandin E2, VEGF, and MMP2 to promote endothelial cell
migration and angiogenesis that establishes new metastatic vascular circuits.
The last phase, or the established hepatic metastasis phase, results in further growth of the established
metastatic colony through further alteration of the local environment, tumor proliferation, and differentia-
tion. This growth phase would be impossible without evasion from and suppression of the immune system
through two generalized mechanisms [40]. First, downregulation of MHC class I on hepatic antigen-­
presenting cells and increased expression of immune inhibitory molecules such as PD-1 on hepatic lympho-
cytes result in immune evasion and suppression [40]. Next, the active suppression of adaptive immune
responses by expanding local populations of regulatory T cells within the metastasis enables further immune
escape. With these processes contributing to a growth circuit, the result is progression and enlargement of
the metastatic deposit.
Stemming from these complex processes of metastasis, it is easy to envision multiple mechanisms of
treatment failure resulting in recurrence, morbidity, and mortality. Treatment goals should target, in a mul-
timodal fashion, the disruption of individual processes of cancer metastasis development. The next two
chapters examine the role of locoregional therapies including surgical resection and cancer regional thera-
pies to control liver metastasis. However, it is prudent to realize that the application of locoregional therapy
alone is insufficient for disease eradication.

References
1. Chaffer CL, Weinberg RA. A perspective on cancer cell metastasis. Science. 2011;331:1559–64.
2. Seyfried TN, Huysentruyt LC. On the origin of cancer metastasis. Crit Rev Oncogenesis.
2012;18:43–73.
3. Thiery JP. Epithelial–mesenchymal transitions in tumour progression. Nat Rev Cancer.
2002;2:442–54.
4. Kalluri R. EMT: when epithelial cells decide to become mesenchymal-like cells. J Clin Invest.
2009;119:1417–9.
5. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144:646–74.
6. Tam W, Weinberg RA. The epigenetics of epithelial-mesenchymal plasticity in cancer. Nat Med.
2013;19:1438–49.
7. Fabregat I, Malfettone A, Soukupova J. New insights into the crossroads between EMT and Stemness
in the context of cancer. J Clin Medicine. 2016;5:37.
8. Chuang H-YY, et al. Aminopeptidase A initiates tumorigenesis and enhances tumor cell stemness via
TWIST1 upregulation in colorectal cancer. Oncotarget. 2017;8:21266–80.
9. Kalluri R, Weinberg RA. The basics of epithelial-mesenchymal transition. J Clin Invest.
2009;119:1420–8.
10. Xu J, Lamouille S, Derynck R. TGF-β-induced epithelial to mesenchymal transition. Cell Res.
2009;19:cr20095.
11. Nakano M, et al. Dedifferentiation process driven by TGF-beta signaling enhances stem cell properties
in human colorectal cancer. Oncogene. 2019;38:780–93.
12. Ricci-Vitiani L, et al. Identification and expansion of human colon-cancer-initiating cells. Nature.
2007;445:111–5.
460 Cancer Metastasis to the Liver

13. Dalerba P, et al. Single-cell dissection of transcriptional heterogeneity in human colon tumors. Nat
Biotechnol. 2011;29:1120–7.
14. Vermeulen L, et al. Single-cell cloning of colon cancer stem cells reveals a multi-lineage differentiation
capacity. Proc National Acad Sci. 2008;105:13427–32.
15. Dalerba P, et al. Phenotypic characterization of human colorectal cancer stem cells. Proc National Acad
Sci. 2007;104:10158–63.
16. Barker N, et al. Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature.
2007;449:1003–7.
17. Wielenga VJ, et al. Expression of CD44 variant proteins in human colorectal cancer is related to tumor
progression. Cancer Res. 1993;53:4754–6.
18. Tremmel M, et al. A CD44v6 peptide reveals a role of CD44 in VEGFR-2 signaling and angiogenesis.
Blood. 2009;114:5236–44.
19. Orian-Rousseau V, Chen L, Sleeman JP, Herrlich P, Ponta H. CD44 is required for two consecutive
steps in HGF/c-Met signaling. Genes Dev. 2002;16:3074–86.
20. Finke, Lothar H. et al. Colorectal cancer prognosis and expression of exon-v6-containing CD44 pro-
teins. Lancet. 1995;345:583–4.
21. Todaro M, et al. Colon cancer stem cells dictate tumor growth and resist cell death by production of
interleukin-4. Cell Stem Cell. 2007;1:389–402.
22. Carmon KS, Gong X, Lin Q, Thomas A, Liu Q. R-spondins function as ligands of the orphan receptors
LGR4 and LGR5 to regulate Wnt/β-catenin signaling. Proc National Acad Sci. 2011;108:11452–7.
23. Barker N, et al. Crypt stem cells as the cells-of-origin of intestinal cancer. Nature. 2009;457:608–11.
24. Paget S. The distribution of secondary growths in cancer of the breast. Lancet. 1889;133:571–3.
25. Liu Y, Cao X. Characteristics and significance of the pre-metastatic niche. Cancer Cell.
2016;30:668–81.
26. Peinado H, et al. Pre-metastatic niches: organ-specific homes for metastases. Nat Rev Cancer.
2017;17:302–17.
27. Costa-Silva B, et al. Pancreatic cancer exosomes initiate pre-metastatic niche formation in the liver.
Nature Cell Biol. 2015;17:816–26.
28. Lee JW, et al. Hepatocytes direct the formation of a pro-metastatic niche in the liver. Nature.
2019;567:249–52.
29. Cools-Lartigue J, et al. Neutrophil extracellular traps sequester circulating tumor cells and promote
metastasis. J Clin Invest. 2013;123:3446–58.
30. Erpenbeck L, Schön MP. Neutrophil extracellular traps: protagonists of cancer progression? Oncogene.
2017;36:2483–90.
31. Seubert B, et al. Tissue inhibitor of metalloproteinases (TIMP)-1 creates a premetastatic niche in the
liver through SDF-1/CXCR4-dependent neutrophil recruitment in mice. Hepatology.
2015;61:238–48.
32. Ham B, et al. TNF receptor-2 facilitates an immunosuppressive microenvironment in the liver to pro-
mote the colonization and growth of hepatic metastases. Cancer Res. 2015;75:5235–47.
33. Jiao S-F, et al. Inhibition of tumor necrosis factor alpha reduces the outgrowth of hepatic micrometas-
tasis of colorectal tumors in a mouse model of liver ischemia-reperfusion injury. J Biomed Sci.
2014;21:1.
34. Khatib A-M, Fallavollita L, Wancewicz EV, Monia BP, Brodt P. Inhibition of hepatic endothelial
E-selectin expression by C-raf antisense oligonucleotides blocks colorectal carcinoma liver metastasis.
Cancer Res. 2002;62:5393–8.
35. Tsuchida T, Friedman SL. Mechanisms of hepatic stellate cell activation. Nat Rev Gastroentero.
2017;14:397–411.
36. Tadashi T, Kiyoko M, Noboru T, Yasuo S, Yasuni N. Alpha-smooth muscle actin-positive stromal cells
in cholangiocarcinomas, hepatocellular carcinomas and metastatic liver carcinomas. J Hepatol.
1996;24:706–12.
37. Marvin DL, Heijboer R, ten Dijke P, Ritsma L. TGF-β signaling in liver metastasis. Clin Transl
Medicine. 2020;10:e160.
Cancer Metastasis to the Liver 461

38. Olaso E, et al. Proangiogenic role of tumor-activated hepatic stellate cells in experimental melanoma
metastasis. Hepatology. 2003;37:674–85.
39. Pezzuto A, Carico E. Role of HIF-1 in cancer progression: novel insights. A review. Curr Mol Med.
2019;18:343–51.
40. Butt AQ, Mills KHG. Immunosuppressive networks and checkpoints controlling antitumor immunity
and their blockade in the development of cancer immunotherapeutics and vaccines. Oncogene.
2014;33:4623–31.
 Tumor Resection and Ablation
as a Means of Controlling Hepatic 42
Metastases

Brian D. Griffith and Timothy L. Frankel

Abstract 42.1 Introduction

The liver is a common site for metastasis in colorectal
cancer. Hepatectomy has emerged as the gold standard for Colorectal cancer is the third most common malignancy
the treatment of colorectal liver metastases and is the only in the United States, with up to 145,000 new cases per
treatment modality that offers a potential for long-term year [1]. As many as 20% of patients present with meta-
cure. However, only a small proportion of patients are static disease, with liver metastasis present in almost 80%
candidates for surgical resection. Adjunctive treatments of patients with stage IV disease [2]. As surgical and anes-
like portal vein embolization, two-stage hepatectomy, thetic techniques for hepatectomy have improved over the
and/or neoadjuvant chemotherapy can be used to down- past decades, hepatic resection has emerged as the gold
stage disease or convert patients to surgical candidates. standard for the treatment of colorectal liver metastases
Less invasive therapies like ablation, radiotherapy, and (CLM). While operative mortality was initially as high
systemic chemotherapy have emerged as treatment as 30% in early series of hepatic resection for colorec-
options for poor surgical candidates. tal cancer [3], perioperative mortality has improved and
is now as low as 1–3% [4–6]. In appropriately selected
patients, hepatic resection provides excellent disease con-
trol when compared to systemic chemotherapy and can
Learning Objectives lead to a long-term cure [7]. A multidisciplinary approach
is needed to achieve the best long-term outcomes and
• To identify indications for resection of hepatic should involve medical oncology, radiology, anesthesia,
metastases and understand the appropriate preop- and surgery.
erative imaging
• To understand indications for adjunct therapies like
portal vein embolization, two-stage hepatectomy, 42.2 Hepatic Resection
or neoadjuvant chemotherapy in patients who are
not initially surgical candidates 42.2.1 Indications for Resection and Efficacy
• To understand the role of ablation, radiotherapy,
and systemic chemotherapy for the treatment of The essential tenets of resection of metastatic disease, first
hepatic metastasis in patients who are not surgical popularized by Alexander and Haigh, include a resectable
candidates primary tumor, a patient who can tolerate a major sur-
gery, and the ability to obtain a negative margin [8]. As
the effectiveness of systemic chemotherapy and the safety
of the operation have improved, indications for resection
of CLM have expanded. Multiple predictive models based
on large retrospective reviews have been developed to bet-
ter select patients for resection. Fong et al. [9] reviewed
B. D. Griffith · T. L. Frankel (*)
1001 patients undergoing liver resection over a 13-year
Department of Surgery, Michigan Medicine, University of
Michigan, Ann Arbor, MI, USA period and discovered that positive margin, presence of
e-mail: [email protected]; [email protected] extrahepatic disease, positive nodal status of the primary,

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 463
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_42
464 B. D. Griffith and T. L. Frankel

disease-free interval from primary to metastatic disease 42.2.2 Preoperative Imaging

<12 months, number of tumors >1, preoperative carcino-
embryonic antigen (CEA) level >200 ng/mL, and size of Common poor predictors of survival among patients with
largest tumor >5 cm were all significant and independent CLM include size of the primary tumor, presence of multiple
predictors of poor survival. By assigning one point to each hepatic lesions, and evidence of extrahepatic disease [5, 9].
of the latter five criteria, a score was created which was Appropriate preoperative cross-sectional imaging is essen-
highly predictive of cancer-specific outcome; low-score tial to best identify these factors as well as to plan the resec-
patients had a median survival of 74 months, while high- tion and improve the overall safety and efficacy of resection.
score patients had a median survival of only 22 months. Except in the case of two-staged hepatectomy, the overall
These scoring systems have been externally validated in goal of resection is to remove all metastatic disease, preserve
independent series [10], and other similar prognostic sys- unaffected tissue, and maintain the segmental blood supply
tems have been developed [4, 5]. A more recent examina- and biliary and venous drainage to the remaining liver rem-
tion of cancer survival by Nordlinger et al. [5] evaluated nant. Thus, when assessing the feasibility of a resection, one
1568 patients undergoing hepatic resection for CLM and must consider the number of involved segments as well as
found similar prognostic factors with the addition of age proximity to arteries, veins, and bile ducts.
and stage of the primary tumor. In addition to clinical fac- There are multiple imaging modalities, each with their
tors, recent attention has turned to aspects of tumor biology respective advantages and disadvantages. Multidetector
that impact prognosis. Genetic sequencing of tumors has computed tomography (MDCT) is routinely used in workup
revealed that the presence of both KRAS mutations [11] and surveillance of oncology patients, most often with the
and BRAF mutations [12] is associated with decreased addition of iodinated contrast agents. With contrast enhance-
survival in patients with CLM. As our understating of the ment, colorectal liver metastases appear hypovascular
biologic factors driving tumorigenesis advances, incorpo- with variable heterogeneity on portal venous phase [16].
ration into prognostic scoring will likely improve patient The addition of arterial phase imaging generally does not
selection for surgical resection. improve the detection of CLM [16], but is helpful for presur-
In patients with unresectable metastatic colon cancer, the gical planning to identify aberrant anatomy. CT angiography
median survival with systemic chemotherapy is approxi- is particularly important for planning hepatic artery pump
mately 21 months, with very few alive at 5 years [13]. In placement if hepatic artery infusion (HAI) chemotherapy is
comparison, the overall 5-year actuarial survival of patients planned. Positron emission tomography (PET) is an imaging
with CLM after hepatic resection with curative intent ranges modality often combined with CT that detects the uptake of
from 27 to 39% and 10-year actuarial survival of 17–22% a glucose analog, 18F-fluorodeoxyglucose (FDG) in hyper-
[7, 9, 14]. Tomlinson et al. [7] evaluated 612 patients with metabolic tumors and is most often used for detection of
CLM treated with hepatic resection and of the 102 patients extrahepatic metastases [17]. Metastases demonstrate high
that were alive at 10 years, 97% were disease free and only uptake on PET, but sensitivity in the liver is limited due to
one experienced a disease-specific death after 10 years, high physiologic background hepatic uptake [18].
­suggesting that those who survive 10 years were likely cured Magnetic resonance imaging (MRI), by comparison,
of the disease. offers improved soft-tissue contrast, particularly in the
While survival can be dramatically improved, liver presence of steatosis or fibrosis. In general, MR images
surgery can be associated with significant morbidity, and are obtained in T1 weighted (T1w), T2 weighted (T2w),
proper patient education and medical optimization are par- and diffusion-weighted images (DWI) before and after
amount. As cancer patients tend to be older, prior to any gadolinium-­ binding contrast agent, with CLM hypoin-
resection, patients should undergo evaluation to ensure tense on T1w, hyperintense on T2w, and hyperintense
appropriate medical fitness. Patients should be screened to on DWI. After IV contrast, CLM is hypovascular with
ensure adequate cardiac and pulmonary function, perfor- irregular rim enhancement. Over the past decade, hepa-
mance status, and sufficient future liver remnant (FLR). The tocyte-specific contrast agents have also been developed.
FLR is the volume of functional liver parenchyma present These contrast agents, including gadobenate dimeglumine
after the resection relative to total liver volume excluding (Gd-BOPTA) and gadoexitic acid (Gd-EOB-DTPA), are
cysts, tumors, and ablation cavities and is useful to predict taken up by functioning hepatocytes and excreted in bile
morbidity and mortality after hepatic resection. In general, [19]. This allows improved detection of metastatic disease,
an FLR of 25–30% for chemotherapy naive patients or 40% as normal parenchyma is hyperintense relative to liver meta-
for those with parenchymal damage like steatosis or fibro- static disease in T1-weighted delayed hepatobiliary phased
sis is associated with a low risk of postoperative risk of images. The combination of diffusion-weighted images and
liver failure [15]. hepatobiliary-­phased MR images allows the highest sensi-
42 Tumor Resection and Ablation as a Means of Controlling Hepatic Metastases 465

tivity for small CLM, particularly metastases less than 1 cm tion does not appear to be significantly different than open
in size, relative to other imaging modalities [20, 21]. resection in propensity-matched patients [26, 27].
As more is learned about the biology of metastatic disease
and with improvements in chemotherapy, an increasing pool
42.2.3 Surgical Technique of patients with initially unresectable disease can often be
converted to surgical candidates with curative intent. Two-­
The goal of hepatectomy for CLM is to obtain a complete stage hepatectomy is one such strategy to convert initially
resection with negative margins, while preserving healthy unresectable bilobar disease into resectable disease. Initially
parenchyma. Historically, resections for CLM were fre- described by Adam et al. [28], the goal of a staged hepatec-
quently performed as anatomic resections like segmen- tomy is to clear tumors from one lobe of the liver during
tectomies, hemi-hepatectomy, or tri-sectionectomy, as an initial operation followed by a period of convalescence
non-anatomics resection was felt to impose higher risk of where ipsilateral hypertrophy can occur. At a second opera-
a positive margin [22, 23]. However, as further retrospec- tion, the contralateral lobe of the liver is resected or cleared
tive reviews have suggested no difference in rates of posi- of remaining metastases. In cases where the FLR is predicted
tive margins, overall recurrence, or overall survival [24, 25], to be below threshold, preoperative portal vein emboliza-
nonanatomic, more limited resections are performed with tion (PVE) can be used to induce hypertrophy of the FLR.
increasing frequency. The specific resection performed is Occlusion of portal flow to the hepatic lobe has been shown
dependent on the location of the tumors in relation to the to cause compensatory hypertrophy to the contralateral lobe
artery and portal venous inflow pedicles, as well as the out- [29]. Preoperative portal vein embolization can be used to
flow veins. Frequently, operative plans are developed at a induce up to 30–50% increase in relative size of the contra-
multidisciplinary conference with the assistance of radiolo- lateral lobe at 6 weeks [30, 31]. Hypertrophy can continue
gists to identify tumors and anatomic structures, as well as for up to 6 months, but most often plateaus 6–8 weeks later.
medical oncologists and interventional radiologists to dis- Two-stage hepatectomy has been shown to be a safe proce-
cuss preoperative strategies to downsize tumors or treat with dure, with perioperative mortality of approximately 0–7%
adjunct therapies. and median overall survival of 37 months [32]. Series show
Open exploration for CLM begins with a midline laparot- that approximately 70–100% of patients scheduled for two-­
omy or subcostal incision large enough to allow for inspection stage hepatectomy were able to successfully complete the
of the entire liver and should consider any planned combined treatment [28, 33].
procedure for synchronous disease. The abdominal cavity,
including peritoneal surfaces and the portal lymph nodes,
should be inspected to exclude extrahepatic disease. Hepatic 42.2.4 Postoperative Care and Follow-Up
exploration is performed with palpation and intraopera-
tive ultrasound to identify deeper lesions and clearly define Postoperatively, patients should be monitored for evidence
the anatomy. Once the degree of disease burden has been of bleeding and adequate analgesia provided. Hepatic func-
established, inflow and outflow control can be e­ stablished tion can be measured with INR to assess liver synthetic
prior to hepatic transection. This is particularly important function and serum bilirubin to assess hepatic clearance.
if hemi-hepatectomy or tri-sectionectomy is planned and/ A trend of rising INR or bilirubin over the first few post-
or tumors are close to the major inflow or outflow vessels. operative days should raise concern for postoperative liver
Once inflow and outflow are controlled, the plane of tran- failure. Supportive measures such as phosphorus repletion
section is marked. Multiple techniques have been described and avoidance of hypotension, as well as imaging to rule out
for transection including the crush–clamp method, ultrasonic potentially correctable biliary and inflow or outflow vascular
vibrations, or radiofrequency ablation. Regardless of the issues, should be considered.
technique used, transection should be performed as a careful Common complications after major hepatectomy
dissection to avoid injury to inflow and outflow structures of include pleural effusion, wound breakdown, or ascites.
the remnant liver, with frequent reorientation or application Early identification of ascites, which can lead to wound
of the ultrasound probe to ensure an adequate margin and leakage, protein wasting, and early satiety if severe, and
reconfirm anatomic landmarks. After transection, hemostatic treatment with diuretics or aldosterone antagonists is pru-
agents or an argon beam coagulator can be used to ensure dent to minimize sequala. An important complication that
hemostasis and biliostasis. Although traditionally performed can lead to liver failure is unrecognized infection, as the
as an open resection, minimally invasive laparoscopic or systemic inflammatory response impairs liver regenera-
robot-assisted resection of CLM is increasingly performed, tion. Biliary leak should be suspected in the setting of ris-
particularly for small peripheral tumors. Studies have con- ing bilirubin and appropriate cross-sectional imaging is
firmed that overall long-term survival for laparoscopic resec- considered to identify a leak. The risk of venous throm-
466 B. D. Griffith and T. L. Frankel

boembolism is similar to that of other major abdominal larger tumors have high rates of local recurrence [42]. RFA is
operations, and prophylaxis should be considered [34]. also limited near major blood vessels where a heat sink effect
causes relatively low temperatures and poor tumor destruction.
The intense heat of ablation can cause damage to surrounding
42.2.5 Hepatic Artery Infusion Chemotherapy structures including bile ducts, the gallbladder as well as colon,
stomach, or diaphragm [43]. As such, the size of tumors, number
Hepatic artery infusion (HAI) chemotherapy is a treatment of tumors, and location of metastases are essential to consider
that can be used either as stand-alone therapy for liver domi- prior to ablation. RFA is generally recommended for patients
nant disease or to adjunctively deliver chemotherapy to the with the following characteristics: tumor size <3 cm, less than 5
functional liver remnant after metastasectomy, as the liver lesions, in which the tumor is not located near major bile ducts
is a common site of recurrence. The infusion catheter is or other structures at risk for injury [44].
most often placed into the gastroduodenal artery, and che- MWA is a second hyperthermic ablation technique that
motherapy, typically Fluorodeoxyuridine based, is delivered is gaining widespread adoption. Benefits over RFA include
intra-­arterially starting about 4 weeks after surgery. Hepatic a larger volume of necrosis, shorter procedure time, higher
metastases derive blood supply primarily from the hepatic temperatures within the target lesion, and less variability
artery, while normal liver cells derive it from the portal vein. within the treatment zone due to heat sink effect [45]. MWA
That biologic phenomenon, in addition to a high first-pass may offer improved local control for perivascular tumors,
metabolism, allows for the delivery of concentrated chemo- while the more controlled RFA may be preferred for peribili-
therapy with minimal systemic effects [35]. Studies on the ary lesions [46, 47]. However, specific indications for RFA
use of adjuvant HAI and systemic therapy versus systemic and MWA continue to evolve.
therapy alone found improved hepatic relapse-free survival When comparing ablation and surgery, a recent meta-­
and overall survival in patients with CLM treated with analysis found significantly shorter hospital stay and lower
regional chemotherapy [36, 37]. Additionally, HAI chemo- complication rate (13% vs. 26%) among the ablation group,
therapy can be effective in converting patients with unresect- driven primarily by lower wound complications, gastrointes-
able disease into surgical candidates [38]. tinal complications, and biliary complications [48]. However,
ablative techniques have been associated with higher rates of
local recurrence [48–50] and decreased overall survival [48,
42.3  lternatives to Resection for Local
A 51]. Comparisons between ablation and surgery are primar-
Control ily retrospective in nature, which may be subject to patient
section bias, as ablation is more often used for unresectable
42.3.1 Ablation disease or as an adjunct to surgery. To better compare abla-
tion and surgery in patients with CLM, the Colorectal Liver
Tumor ablation is a minimally invasive local therapy that Metastases: Surgery versus Thermal Ablation (COLLISION)
uses hyperthermic treatment modalities to treat metastases. trial is currently enrolling to evaluate whether thermal abla-
Options for ablation include radiofrequency ablation (RFA), tion with RFA or MWA is non-inferior to hepatic resection in
microwave ablation (MWA), and high-intensity focused patients with at least one resectable or ablatable CLM [52].
ultrasound (HIFU). The goal of treatment is to induce com-
plete necrosis of the metastasis and a safety margin similar to
what would be achieved with an R0 resection. While surgi- 42.3.2 Stereotactic Body Radiotherapy
cal resection remains the gold standard treatment for hepatic
metastases, a significant population of patients is not surgical Stereotactic body radiotherapy (SBRT) is an alternative,
candidates due to inadequate FLR, metastases location, or noninvasive treatment modality for treatment liver metasta-
medical comorbidities. A randomized control trial investi- ses, in which targeted radiation is delivered to a tumor in
gating systemic therapy with or without RFA in patients with a hypofractionation schedule while minimizing radiation to
unresectable CLM found a significantly improved long-term adjacent normal tissue. Radiotherapy has historically been
survival in the combined treatment group [39, 40]. Ablation limited by poor tolerance from liver parenchyma, but the
can also be used as an adjunct to surgery particularly for focused delivery in SBRT limits radiation-induced liver dis-
deep, small lesions where removal of a large volume of nor- ease and liver failure [53]. SBRT has safely been adminis-
mal liver would be needed for tumor resection. tered up to 75 gray for treatment of metastases [54]. Similar
In the past, RFA was the most commonly used ablative to the ablative techniques previously described, SBRT offers
therapy and relies on high-energy radio waves to heat tissues. an adjunct for local treatment of metastases in patients who
It suffers from a relative lack of tissue penetration and is typi- are not candidates for resection. Early evidence has shown
cally reserved for tumors less than 3 cm in diameter [41], as promising results for local control of 55–90%, and actuar-
42 Tumor Resection and Ablation as a Means of Controlling Hepatic Metastases 467

ial overall survival of 40–70% at 2 years, in patients with 4. Rees M, Tekkis PP, Welsh FK, O’Rourke T, John TG. Evaluation
of long-term survival after hepatic resection for metastatic colorec-
unresectable CLM [54, 55], with differences likely depen- tal cancer: a multifactorial model of 929 patients. Ann Surg.
dent on tumor volume and pathology, prior therapy, SBRT 2008;247(1):125–35.
doses, and fractionation regimens [56]. The role of SBRT for 5. Nordlinger B, Guiguet M, Vaillant JC, et al. Surgical resection of
local therapy and its comparison to other ablative techniques colorectal carcinoma metastases to the liver. A prognostic scor-
ing system to improve case selection, based on 1568 patients.
in treatment of metastasis has yet to be defined. However, Association Francaise de Chirurgie. Cancer. 1996;77(7):1254–62.
SBRT may provide local control when considered for treat- 6. Simmonds PC, Primrose JN, Colquitt JL, Garden OJ, Poston GJ,
ment of one to five metastases, which are <6 cm in diameter, Rees M. Surgical resection of hepatic metastases from colorec-
and in patients with good performance status and adequate tal cancer: a systematic review of published studies. Br J Cancer.
2006;94(7):982–99.
pre-­SBRT liver function [56]. 7. Tomlinson JS, Jarnagin WR, DeMatteo RP, et al. Actual 10-year
survival after resection of colorectal liver metastases defines cure. J
Clin Oncol. 2007;25(29):4575–80.
42.4 Conclusion 8. Alexander J, Haight C. Pulmonary resection for solitary
metastatic sarcomas and carcinomas. Surg Gynecol Obstet.
1947;85(2):129–46.
Colon cancer is the third most common malignancy in 9. Fong Y, Fortner J, Sun RL, Brennan MF, Blumgart LH. Clinical
the United States, and up to 20% of patients present with score for predicting recurrence after hepatic resection for metastatic
metastatic disease. Hepatectomy is the gold standard for colorectal cancer: analysis of 1001 consecutive cases. Ann Surg.
1999;230(3):309–18; discussion 318–321.
the treatment of colorectal liver metastases and is the most 10. Mala T, Bohler G, Mathisen O, Bergan A, Soreide O. Hepatic resec-
proven treatment option with regards to potential for a long- tion for colorectal metastases: can preoperative scoring predict
term cure. Adjunct therapies like portal vein embolization patient outcome? World J Surg. 2002;26(11):1348–53.
and two-stage hepatectomy have expanded the population 11. Vauthey JN, Zimmitti G, Kopetz SE, et al. RAS mutation status
predicts survival and patterns of recurrence in patients under-
of patients in whom safe resection can be performed. For going hepatectomy for colorectal liver metastases. Ann Surg.
patients who are poor surgical candidates, therapies like 2013;258(4):619–26; discussion 626–617.
ablation and radiotherapy, offer alternative approaches for 12. Pikoulis E, Margonis GA, Andreatos N, et al. Prognostic role of
tumor eradication which may improve survival when com- BRAF mutations in colorectal cancer liver metastases. Anticancer
Res. 2016;36(9):4805–11.
pared to systemic therapy alone. 13. Meyerhardt JA, Mayer RJ. Systemic therapy for colorectal cancer.
N Engl J Med. 2005;352(5):476–87.
14. Choti MA, Sitzmann JV, Tiburi MF, et al. Trends in long-term sur-
vival following liver resection for hepatic colorectal metastases.
Ann Surg. 2002;235(6):759–66.
Open Questions
15. Ferrero A, Vigano L, Polastri R, et al. Postoperative liver dysfunc-
• What tumor biomarkers are predictive of outcome, tion and future remnant liver: where is the limit? Results of a pro-
and what is the role of tumor biology in determin- spective study. World J Surg. 2007;31(8):1643–51.
ing candidacy for resection in patients with CLM? 16. Soyer P, Poccard M, Boudiaf M, et al. Detection of hypovascu-
lar hepatic metastases at triple-phase helical CT: sensitivity of
• Is there a role for ablation in the treatment of CLM phases and comparison with surgical and histopathologic findings.
that are surgically resectable, particularly in patients Radiology. 2004;231(2):413–20.
with significant comorbidities? 17. Sacks A, Peller PJ, Surasi DS, Chatburn L, Mercier G, Subramaniam
• What is the role of hepatic artery infusion chemo- RM. Value of PET/CT in the management of liver metastases, part
1. AJR Am J Roentgenol. 2011;197(2):W256–9.
therapy in unresectable CLM? 18. Kong G, Jackson C, Koh DM, et al. The use of 18F-FDG PET/
• What is the role of radiotherapy for both local con- CT in colorectal liver metastases—comparison with CT and liver
trol and treatment of unresectable CLM? MRI. Eur J Nucl Med Mol Imaging. 2008;35(7):1323–9.
• What is the role of resection of CLM in the setting 19. Seale MK, Catalano OA, Saini S, Hahn PF, Sahani
DV. Hepatobiliary-specific MR contrast agents: role in imaging the
of extra-hepatic metastatic disease? liver and biliary tree. Radiographics. 2009;29(6):1725–48.
20. Muhi A, Ichikawa T, Motosugi U, et al. Diagnosis of colorectal hepatic
metastases: comparison of contrast-enhanced CT, contrast-enhanced
US, superparamagnetic iron oxide-enhanced MRI, and gadoxetic acid-
enhanced MRI. J Magn Reson Imaging. 2011;34(2):326–35.
References 21. Seo HJ, Kim MJ, Lee JD, Chung WS, Kim YE. Gadoxetate
disodium-enhanced magnetic resonance imaging versus contrast-­
1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer enhanced 18F-fluorodeoxyglucose positron emission tomography/
J Clin. 2019;69(1):7–34. computed tomography for the detection of colorectal liver metasta-
2. Siegel RL, Ward EM, Jemal A. Trends in colorectal cancer inci- ses. Investig Radiol. 2011;46(9):548–55.
dence rates in the United States by tumor location and stage, 1992-­ 22. Welsh FK, Tekkis PP, O’Rourke T, John TG, Rees M. Quantification
2008. Cancer Epidemiol Biomark Prev. 2012;21(3):411–6. of risk of a positive (R1) resection margin following hepatic resec-
3. Foster JH, Berman MM. Solid liver tumors. Major Probl Clin Surg. tion for metastatic colorectal cancer: an aid to clinical decision-­
1977;22:1–342. making. Surg Oncol. 2008;17(1):3–13.
468 B. D. Griffith and T. L. Frankel

23. DeMatteo RP, Palese C, Jarnagin WR, Sun RL, Blumgart LH, 40. Ruers T, Van Coevorden F, Punt CJ, et al. Local treatment of unre-
Fong Y. Anatomic segmental hepatic resection is superior to wedge sectable colorectal liver metastases: results of a randomized phase
resection as an oncologic operation for colorectal liver metastases. II trial. J Natl Cancer Inst. 2017;109(9):djx015.
J Gastrointest Surg. 2000;4(2):178–84. 41. van Duijnhoven FH, Jansen MC, Junggeburt JM, et al. Factors
24. Kokudo N, Tada K, Seki M, et al. Anatomical major resection influencing the local failure rate of radiofrequency ablation of
versus nonanatomical limited resection for liver metastases from colorectal liver metastases. Ann Surg Oncol. 2006;13(5):651–8.
colorectal carcinoma. Am J Surg. 2001;181(2):153–9. 42. Shady W, Petre EN, Gonen M, et al. Percutaneous radiofrequency abla-
25. Sarpel U, Bonavia AS, Grucela A, Roayaie S, Schwartz ME, Labow tion of colorectal cancer liver metastases: factors affecting outcomes—a
DM. Does anatomic versus nonanatomic resection affect recur- 10-year experience at a single center. Radiology. 2016;278(2):601–11.
rence and survival in patients undergoing surgery for colorectal 43. Cirocchi R, Trastulli S, Boselli C, et al. Radiofrequency ablation in
liver metastasis? Ann Surg Oncol. 2009;16(2):379–84. the treatment of liver metastases from colorectal cancer. Cochrane
26. Lewin JW, O’Rourke NA, Chiow AKH, et al. Long-term survival Database Syst Rev 2012(6):CD006317.
in laparoscopic vs open resection for colorectal liver metastases: 44. Gillams A, Goldberg N, Ahmed M, et al. Thermal ablation of
inverse probability of treatment weighting using propensity scores. colorectal liver metastases: a position paper by an international
HPB (Oxford). 2016;18(2):183–91. panel of ablation experts, The Interventional Oncology Sans
27. Beppu T, Wakabayashi G, Hasegawa K, et al. Long-term and peri- Frontieres meeting 2013. Eur Radiol. 2015;25(12):3438–54.
operative outcomes of laparoscopic versus open liver resection 45. Brace CL. Radiofrequency and microwave ablation of the liver,
for colorectal liver metastases with propensity score matching: a lung, kidney, and bone: what are the differences? Curr Probl Diagn
multi-institutional Japanese study. J Hepatobiliary Pancreat Sci. Radiol. 2009;38(3):135–43.
2015;22(10):711–20. 46. Shady W, Petre EN, Do KG, et al. Percutaneous microwave ver-
28. Adam R, Laurent A, Azoulay D, Castaing D, Bismuth H. Two-stage sus radiofrequency ablation of colorectal liver metastases: ablation
hepatectomy: a planned strategy to treat irresectable liver tumors. with clear margins (A0) provides the best local tumor control. J
Ann Surg. 2000;232(6):777–85. Vasc Interv Radiol. 2018;29(2):268–275 e261.
29. Makuuchi M, Thai BL, Takayasu K, et al. Preoperative portal 47. van Tilborg AA, Scheffer HJ, de Jong MC, et al. MWA versus RFA
embolization to increase safety of major hepatectomy for hilar bile for perivascular and peribiliary CRLM: a retrospective patient-
duct carcinoma: a preliminary report. Surgery. 1990;107(5):521–7. and lesion-based analysis of two historical cohorts. Cardiovasc
30. van Lienden KP, van den Esschert JW, de Graaf W, et al. Portal Intervent Radiol. 2016;39(10):1438–46.
vein embolization before liver resection: a systematic review. 48. van Amerongen MJ, Jenniskens SFM, van den Boezem PB, Futterer
Cardiovasc Intervent Radiol. 2013;36(1):25–34. JJ, de Wilt JHW. Radiofrequency ablation compared to surgical
31. Abulkhir A, Limongelli P, Healey AJ, et al. Preoperative portal vein resection for curative treatment of patients with colorectal liver
embolization for major liver resection: a meta-analysis. Ann Surg. metastases—a meta-analysis. HPB (Oxford). 2017;19(9):749–56.
2008;247(1):49–57. 49. Otto G, Duber C, Hoppe-Lotichius M, Konig J, Heise M, Pitton
32. Lam VW, Laurence JM, Johnston E, Hollands MJ, Pleass HC, MB. Radiofrequency ablation as first-line treatment in patients with
Richardson AJ. A systematic review of two-stage hepatectomy in early colorectal liver metastases amenable to surgery. Ann Surg.
patients with initially unresectable colorectal liver metastases. HPB 2010;251(5):796–803.
(Oxford). 2013;15(7):483–91. 50. Aliyev S, Agcaoglu O, Aksoy E, et al. Efficacy of laparoscopic
33. Adam R, Miller R, Pitombo M, et al. Two-stage hepatectomy radiofrequency ablation for the treatment of patients with small
approach for initially unresectable colorectal hepatic metastases. solitary colorectal liver metastasis. Surgery. 2013;154(3):556–62.
Surg Oncol Clin N Am. 2007;16(3):525–36, viii. 51. Agcaoglu O, Aliyev S, Karabulut K, et al. Complementary use of
34. Tzeng CW, Katz MH, Fleming JB, et al. Risk of venous throm- resection and radiofrequency ablation for the treatment of colorec-
boembolism outweighs post-hepatectomy bleeding complications: tal liver metastases: an analysis of 395 patients. World J Surg.
analysis of 5651 National Surgical Quality Improvement Program 2013;37(6):1333–9.
patients. HPB (Oxford). 2012;14(8):506–13. 52. Puijk RS, Ruarus AH, Vroomen L, et al. Colorectal liver metasta-
35. Sigurdson ER, Ridge JA, Kemeny N, Daly JM. Tumor and liver ses: surgery versus thermal ablation (COLLISION)—a phase III
drug uptake following hepatic artery and portal vein infusion. J Clin single-blind prospective randomized controlled trial. BMC Cancer.
Oncol. 1987;5(11):1836–40. 2018;18(1):821.
36. Kemeny N, Jarnagin W, Gonen M, et al. Phase I/II study of hepatic 53. Dawson LA, Normolle D, Balter JM, McGinn CJ, Lawrence
arterial therapy with floxuridine and dexamethasone in combina- TS, Ten Haken RK. Analysis of radiation-induced liver disease
tion with intravenous irinotecan as adjuvant treatment after resec- using the Lyman NTCP model. Int J Radiat Oncol Biol Phys.
tion of hepatic metastases from colorectal cancer. J Clin Oncol. 2002;53(4):810–21.
2003;21(17):3303–9. 54. Scorsetti M, Comito T, Tozzi A, et al. Final results of a phase II trial
37. Kemeny N, Huang Y, Cohen AM, et al. Hepatic arterial infusion of for stereotactic body radiation therapy for patients with inoperable
chemotherapy after resection of hepatic metastases from colorectal liver metastases from colorectal cancer. J Cancer Res Clin Oncol.
cancer. N Engl J Med. 1999;341(27):2039–48. 2015;141(3):543–53.
38. Kemeny NE, Melendez FD, Capanu M, et al. Conversion to resect- 55. Chang DT, Swaminath A, Kozak M, et al. Stereotactic body radio-
ability using hepatic artery infusion plus systemic chemotherapy therapy for colorectal liver metastases: a pooled analysis. Cancer.
for the treatment of unresectable liver metastases from colorectal 2011;117(17):4060–9.
carcinoma. J Clin Oncol. 2009;27(21):3465–71. 56. Cacciola A, Parisi S, Tamburella C, et al. Stereotactic body radia-
39. Ruers T, Punt C, Van Coevorden F, et al. Radiofrequency ablation tion therapy and radiofrequency ablation for the treatment of
combined with systemic treatment versus systemic treatment alone liver metastases: how and when? Rep Pract Oncol Radiother.
in patients with non-resectable colorectal liver metastases: a ran- 2020;25(3):299–306.
domized EORTC Intergroup phase II study (EORTC 40004). Ann
Oncol. 2012;23(10):2619–26.
 Regional Arterial Infusional Therapy
as a Means of Controlling Hepatic 43
Metastases

Zachary Brown, Stanley Kalata, and Alex C. Kim

Abstract
Learning Objectives
Colorectal cancer (CRC) is one of the most prevalent
cancers diagnosed in the United States. Many patients • History and conception of hepatic artery infusional
are diagnosed with synchronous or metachronous liver therapy
metastasis. Liver is the most common site of metastasis • Understanding pharmacology of floxuridine
and is the leading contributing factor for cancer-related (FUDR) infusion
morbidity and mortality for CRC. The prognosis of • Introduction to operative approaches for hepatic
the patient is highly dependent on the ability to eradi- artery infusion pump placement
cate disease through surgery. In addition, modern • Recognition of pitfalls and complications from the
systemic chemotherapy has significantly contributed hepatic artery infusional therapy
to increased overall survival as well as conversion of • Assessment of critical trials behind hepatic artery
initially unresectable disease to resection in patients infusion pump therapy
with colorectal liver metastasis (CRLM). Despite these • Identification of future application of hepatic artery
improvements, there are still significant volume of infusional therapy
patients who either experience early recurrence fol-
lowing surgery or remain with unresectable disease.
Regional infusional therapy, such as the hepatic artery
pump therapy, is gaining wider acceptance with its role 43.1 Introduction
established in adjuvant therapy following resection in
high-risk patients, definitive treatment of high-volume Colorectal cancer (CRC) is the fourth most prevalent cancer
disease with a goal of conversion to resection, and pal- in the United States with 147,950 estimated cases and with
liation for unresectable disease. 53,200 estimated cancer deaths in 2020 [1]. Colorectal liver
metastasis (CRLM) is the most common site of distant dis-
ease, accounting for greater than 60% of all metastatic sites
and remains the top cause of cancer-related mortality [2–4].
It is believed that approximately 15–25% of patients present
with synchronous metastasis and additional 18–25% eventu-
Zachary Brown and Stanley Kalata contributed equally to this work.
ally develop metachronous disease within 5 years of initial
diagnosis [4, 5]. The patient’s prognosis is highly dependent
on the resectability of the CRLM, with R0 resection being
Z. Brown · A. C. Kim (*) potentially curative [6, 7]. The advent of modern systemic
Division of Surgical Oncology, Department of Surgery, The James chemotherapy (5-FU, leucovorin, and oxaliplatin (FOLFOX);
Cancer Hospital and Solove Research Institute, The Ohio State 5-FU, leucovorin, and irinotecan (FOLFIRI), with or without
University Wexner Medical Center, Columbus, OH, USA
targeted therapy such as bevacizumab (anti-VEGF) and
e-mail: [email protected]; [email protected]
cetuximab (anti-EGF)) has not only increased survival in
S. Kalata
patients with any metastatic disease but also allowed for the
Section of General Surgery, Department of Surgery, University of
Michigan Medical School, Ann Arbor, MI, USA conversion of unresectable CRLM patients to resection [8,
e-mail: [email protected] 9]. Despite these advances, a majority of patients are still

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 469
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_43
470 Z. Brown et al.

faced with unresectable metastatic disease with limited treat- complications including arterial thrombosis, sepsis, and air
ment options with only 15–25% of patients becoming candi- embolism. The percutaneous catheter required a connection
dates for hepatectomy [7, 10]. Recently, the hepatic artery to an extracorporeal pump device that was unwieldy and
infusion (HAI) pump therapy has gained more widespread restrictive for patients. One of the largest series from that era
acceptance with its role established in adjuvant therapy fol- showed an operative mortality of 12% and catheter-related
lowing resection in high-risk patients, definitive treatment of morbidity of 24% [21]. In 1980, Buchwald published the
high-volume disease with a goal of conversion to resection, first study on five patients who underwent placement of an
and palliation for unresectable disease. implantable infusion pump [22]. Subsequent studies com-
The concept of liver-directed arterial infusion therapy was paring implantable pumps versus ports found that unsalvage-
initially developed in the 1950s with a goal of treating pri- able complications occurred in 30% of pump patients and in
mary and secondary liver cancers. The rationale of direct 47% of port patients, with a complication-free survival of
delivery of chemotherapy to the tumor was to capitalize on 12.2 months in the pump group compared to 7.3 months in
the unique hepatic circulation. The liver is supplied by a dual the port group [23]. A metanalysis of over 3000 patients
vascular inflow consisting of the portal vein and the hepatic found that the complication rate was 16% with implantable
artery. The portal vein is a low-pressure system with a pumps and over 30% for port systems. Furthermore, the
70–80% oxygen saturation while the hepatic artery main- median number of cycles delivered in the pump group was
tains systemic arterial pressure and oxygen saturation [11]. 12 versus 9 in the port group which also suggested increased
Based on this fact, preclinical studies utilizing animal mod- durability and reliability of the implantable pump system
els examined and demonstrated a differential delivery of [24]. There are now several models commercially available
blood flow to either the native liver parenchyma or hepatic for implantation, with one of the most commonly used
tumors from the portal vein and hepatic artery. The injection pumps including a 20–50 mL titanium disk that uses two
of India ink through the portal vein resulted in preferential chambers to separate the drug infusion from the energy
native parenchymal uptake, while the injection through the source. The pump design allows the process of filling the
hepatic artery demonstrated preferential tumor uptake [12]. drug infusion chamber to recharge the chemical energy
Subsequent research established that the malignant tumor power source such that only periodic refills are required for
vasculature was mostly derived from the arterial system due maintenance [25].
to portal system invasion and occlusion by the tumor [12,
13]. In 1966, Nilsson et al. performed an in vivo study dem-
onstrating selective hepatic tumor necrosis following hepatic 43.2 Pharmacology
artery ligation and improved survival in the animal models
[14]. Furthermore, inadvertent and deliberate hepatic artery Once the delivery mechanism was established, extensive
interruptions in patients confirmed the preferential effect in studies to identify the optimal chemotherapy regimen
both primary and secondary liver cancers [14–18]. As such, ensued. The ideal agent for this type of therapy was required
the concept of liver regional therapy was conceived by the to exhibit a steep dose–response curve in which greater anti-
landmark study from Sullivan et al. In the study, the authors tumor and toxic effects directly correlate to increased dose
described their experience of placement of a catheter in the and/or concentration. The agent also needed to demonstrate
common hepatic artery and directly infusing chemotherapy a high first-pass hepatic metabolism and short plasma half-­
for 1–2 months to treat primary and secondary liver cancer life which limit the systemic toxicity and increase the thera-
[19]. The results of the study were astounding and found that peutic index [26]. Early candidate agents included
of 16 patients who received a full course of therapy, 10 antimetabolites [5-fluoruouracil (5-FU) and floxuridine
patients derived true clinical benefit from the treatment. (FUDR)] [27–29], mitomycin C, cisplatin [30], and doxoru-
These patients exhibited a reduction of liver tumor volume bicin [31]. FUDR was quickly identified as the leading can-
and significant improvement in liver function tests with min- didate. In vivo studies demonstrated that the infusion of
imal toxicity. The subsequent studies by groups at the FUDR resulted in a 15-fold higher drug concentration within
University of Wisconsin and Memorial Sloan Kettering tumor cells in comparison to normal hepatocytes. Moreover,
Cancer Center revealed similar results revealing safety and there was 92% first-pass extraction which limited the sys-
efficacy [13, 20]. temic exposure. The result was a significant hepatic-to-­
The early delivery systems for hepatic artery infusion systemic concentration of 400:1 [32].
(HAI) were fraught with many technical and mechanical Stemming from the unique delivery mechanism, the side
complications. These involved placing an external catheter effects of FUDR infusion are mainly limited to gastrointesti-
or port during an exploratory laparotomy or via a percutane- nal (GI) tract inflammation/ulceration and hepatotoxicity
ous approach. The delivery catheter itself had risks of clot- manifested by abnormalities in liver function tests.33As such,
ting, kinking, leaking, dislodgement, with more serious prophylaxis with PPI or H2 blockers and close laboratory
43 Regional Arterial Infusional Therapy as a Means of Controlling Hepatic Metastases 471

monitoring is crucial. Initially, it was hypothesized that the strategy was discovered after patients with ligated hepatic
upper GI inflammation may be the result of collateralization lobar arteries demonstrated uniform bilobar hepatic perfusion
of the blood vessels following the operation or possible for 100% of patients by postoperative day 5 [39]. In the rare
exposure of the tract to the FUDR metabolites in the bile case of a trifurcation of the left hepatic artery/right hepatic
[33]. While majority of these cases can be managed with GI artery/GDA off the common hepatic artery (CHA), it is still
prophylaxis, some patients warrant endoscopic or interven- preferable to ligate the smaller lobar vessel and insert the cath-
tional radiology procedures to appropriately treat them. eter in the GDA. If the GDA is absent or unable to be cannu-
Mild biliary toxicity, manifested by elevation of aspartate lated for catheter placement, retrograde cannulation via the
transaminase (AST), alkaline phosphatase (ALT), and biliru- splenic artery can be attempted with ligation of the GDA, right
bin, warrants dose reduction and/or infusion of dexametha- gastric artery, left gastric artery, and smaller collateral vessels.
sone through the pump. Due to the variability in liver enzyme Unfortunately, cannulating any vessel other than the GDA
derangement from metastatic disease burden, hepatic toxic- results in an increased risk of pump related complications and
ity is generally defined as at least doubling of alkaline phos- failure [40].
phatase, tripling of AST, and a 1.5-fold increase in bilirubin Staging laparoscopy to rule out extrahepatic disease should
levels. Occasionally, treatment cessation is necessary in a be considered for patients with unresectable liver masses to
small number of patients. Biliary sclerosis is a dreaded com- avoid committing patients to open exploration. A laparotomy
plication likely resulting from direct drug toxicity and isch- is performed via an upper midline or subcostal incision fol-
emic injury from decreasing the blood supply to the bile duct lowed by an extensive search for extrahepatic disease. If no
via ligation of the gastroduodenal artery collaterals. The extrahepatic disease is found, a portal lymphadenectomy is
addition of dexamethasone was shown to decrease rates of performed by dissecting inferolateral to the confluence of the
biliary sclerosis (30% FUDR alone vs. 9% FUDR/dexameth- cystic duct and common hepatic duct before dissecting the
asone) [34]. Systemic bevacizumab in conjunction with medial aspect of the bile duct. A cholecystectomy is then per-
FUDR therapy has been shown to significantly increase the formed to prevent chemical cholecystitis as the cystic artery
rates of biliary sclerosis [35]. As such, the addition of beva- originates from the hepatic artery branches. The CHA and
cizumab to the therapy is not recommended [35]. Maintaining GDA are palpated in the porta hepatis. The distal CHA, GDA,
a high index of suspicion is key when hepatic enzymes fail to and proper hepatic artery (PHA) are skeletonized circumfer-
improve after dose reduction and warrants radiographic entially and all collateral branches arising from these vessels
investigation of extrinsic compression of the biliary tree for at least 2 cm are ligated to prevent extrahepatic perfusion
from recurrence via CT with subsequent MRCP or ERCP to [41]. The right gastric artery is also ligated in a similar fashion.
investigate for biliary stricture. ERCP is particularly helpful It is also crucial to rule out critical stenosis at the origin of the
in diagnosing extrahepatic or intrahepatic biliary tree stric- celiac axis necessitating retrograde flow through the GDA by
tures not related to tumor progression and therefore stenting/ clamping the GDA and palpating pulses in the PHA. A 2–3 cm
dilation or external biliary drainage can be pursued [36]. stump of the GDA is necessary to safely secure the infusion
catheter. After the target arterial anatomy is sufficiently iso-
lated, a pump pocket is created by making an 8 cm transverse
43.3 Operation incision and forming a pocket superficial to the abdominal
wall fascia. The pump pocket should be chosen so that the
The operative steps to hepatic artery infusion pump placement pump will not touch the anterior superior iliac spine or costal
remain relatively similar for both open and minimally invasive margin, and the pump must easily be filled with a needle
approaches. Careful preoperative assessment of the hepatic through the abdominal wall. If the patient is obese with a great
arterial anatomy and its collaterals is essential. ­High-­resolution deal of subcutaneous fat placement on the chest may be neces-
CT angiography remains the preferred diagnostic modality in sary to ensure subsequent filling.
most major centers. The presence of aberrant hepatic artery Prior to cannulating the GDA, the catheter is flushed and
anatomy is usually not a contraindication to pump placement. pump filled with heparinized saline. An aperture is made at
A retrospective review of 200 angiograms of patients with the center of the pump pocket to facilitate catheter placement
hepatic artery infusion catheters placed found that 70% dem- into the abdominal cavity. The pump is then secured to the
onstrated typical anatomy with the gastroduodenal artery pocket with stay sutures ensuring the catheter is positioned
(GDA) proximal to the bifurcation of the hepatic artery [37]. behind it. Proximal and distal control of the GDA is obtained
For patients with an aberrant hepatic artery, dual lumens by ligating the distal GDA with nonabsorbable sutures and
pumps were used to cannulate the GDA and aberrant artery for placement of vascular clamps on the CHA and PHA. Another
uniform delivery to the liver [38]. These pumps are no longer technique that has been performed is to place a clamp directly
manufactured, so the operative strategy has evolved to ligating on the GDA at its takeoff from the CHA. The latter technique
the anomalous vessel and placing the pump in the GDA. This prevents the inadvertent advancement of the catheter into the
472 Z. Brown et al.

CHA. An arteriotomy on the GDA is created and the catheter between 40% and 60% [32, 47], with improvement in rates
is advanced and secured with two or three nonabsorbable to 16–22% over time [48]. The largest review of a single-­
ties. The catheter tip should lie exactly at the junction of the center experience was published by Allen et al. in 2005
CHA and GDA. Catheter placement in the CHA may result which demonstrated 22% of patients developed a pump-­
in turbulent flow and thrombosis while placement in the dis- related complication with a resultant 45% rate of pump sal-
tal GDA will leave this vessel segment directly exposed to vage to complete hepatic artery infusion therapy [40].
cytotoxic drugs with minimal flow leading to sclerosis, pseu- Arterial thrombosis can result acutely intraoperatively dur-
doaneurysm formation, or thrombosis [42]. With the catheter ing insertion or due to chronic exposure to cytotoxic therapy
secure, methylene blue (using visible light) or fluorescein with resultant sclerosis/thrombosis. A systematic review of
(using a Wood’s lamp) is injected into the pump, and the liver 1371 implantable pump placements found a thrombosis rate
is assessed for uniform dye distribution while extrahepatic of 6.6%, while the use of thrombolytic therapy and/or antico-
perfusion or dye extravasation is evaluated. If any abnormal- agulation results in salvage rates of approximately 30% [24].
ities are identified, an extensive search for potential acces- Infection associated with the pump pocket occurs in 2–3% of
sory hepatic arteries or collateral vessels should be performed patients and typically treated with oral antibiotics for simple
and those vessels should be ligated. Once the perfusion test cellulitis. The failure to improve on antibiotics, presence of
is adequate, the surgical incisions should be closed, and the infected collections, or pump extrusion necessitates explora-
procedure is complete. Although not universally performed, tion. Operative exploration typically involves a local opera-
a radionucleotide pump flow study can be done on postop- tion within the pump bed to remove the infected pump and
erative day 3–5 by intravenously injecting radiolabeled sul- catheter, splicing the old catheter and new catheter together,
fur colloid followed by a technetium-labeled macroaggregated and creating a new pump pocket. Early postoperative hemor-
albumin injection via the hepatic artery pump to ensure uni- rhage or rapidly expanding hematomas warrant immediate
formity of hepatic perfusion and confirm the absence of reexploration to control the source of bleeding.
extrahepatic perfusion. Hepatic arterial chemotherapy is Incomplete hepatic perfusion occurs in about 2% of cases
typically initiated 7–14 days after the operation and requires from failure to ligate an aberrant hepatic artery or failure of col-
biweekly labs with monthly clinical examination. lateralization after lobar artery ligation. Given the concern for a
Percutaneous catheter placement via a transaxillary decrease in treatment efficacy, these patients can be followed
approach has been attempted to avoid a major abdominal with repeat imaging in 2–4 weeks to determine if incomplete
operation, and while the length of stay was shorter, catheter-­ hepatic perfusion persists. Nearly 100% of patients have a reso-
related complications resulting in a delay or cessation of lution of this finding, while a persistent patent accessory artery
treatment was seen in 43% of percutaneous catheter patients can be managed with angioembolization [49]. A more worri-
compared to 7% in the standard laparotomy group [43]. some finding is extrahepatic perfusion of the stomach, duode-
Minimally invasive platforms including both laparoscopic num, or pancreas, which occurs from collateral vessels arising
and robotic-assisted operation have been utilized and shown distal to the catheter or catheter malposition. This presents as
to have a shorter length of stays, less pain, and more immedi- severe epigastric pain/diarrhea with infusions from ulcers or pan-
ate commencement of regional chemotherapy [44]. These creatitis. After the diagnosis is confirmed with endoscopic or
are becoming increasingly popular as concomitant proce- pump flow studies, embolizing the culprit vessel allows for pump
dures to address primary and secondary tumors are almost salvage. A study by Sofocleous found 43 (9%) of patients had
universally attempted laparoscopically. Unfortunately, the signs of extrahepatic perfusion and 19 of 24 (79%) patients with
complexity of the minimally invasive maneuvers required a vessel to target were able to undergo embolization with clinical
limits this approach to specialized high-volume centers. success [50]. Pump malfunction is typically discovered early in
Robotic-assisted surgery for HAI pump placement has been the postoperative setting and can universally be salvaged with
gaining popularity with multiple single center experience replacing the pump. Catheter occlusion/dislodgement and pump
studies proving them to be safe and feasible [45, 46]. Further migration have also been described with an 11–36% rate of sal-
studies are needed to determine the durability and generaliz- vage, while catheter erosion necessitates pump removal.
ability of the results outside of these specific institutions.

43.5 Clinical Outcomes

43.4 Technical Complications
43.5.1 Unresectable Colorectal Cancer Liver
The most common complications after HAI pump placement Metastasis
are thrombosis, infection, hemorrhage, incomplete or extra-
hepatic perfusion, and catheter kinking/migration. Early In patients with CRLM, long-term survival is achievable in
studies as late as the 1990s reported pump complication rates properly selected patients after resection of isolated, low-­
43 Regional Arterial Infusional Therapy as a Means of Controlling Hepatic Metastases 473

volume disease with a 5-year survival up to 70% [51–53]. with less than 20% tumor burden where a median survival
Unfortunately, 80–90% of patients initially present with was not reached at the end of the study. Importantly, this
unresectable liver disease [3, 54–56]. HAI pump therapy for investigation defined the adverse events stemming from the
patients with unresectable colorectal liver metastasis was HAI pump therapy. Severe GI side effects, such as gastroin-
originally adopted in the 1980s where Kemeny et al. uti- testinal ulcers, diffuse gastritis or duodenitis, were present
lized FUDR with promising results [33] (Table 43.1). The in 46% of patients as well as 71% of patients having hepatic-
study demonstrated a partial response in 12 of 23 patients enzyme abnormalities [33]. Nonetheless, this monumental
(52%) with a mean duration of response of 8 months. study ushered the era for hepatic artery infusion pump
Interestingly, patients previously treated with systemic che- therapy.
motherapy exhibited a smaller response rate of 3 of 18 Building on this trial, Chang et al. hypothesized that
(17%) patients with a mean duration of 4 months. Although regional delivery of FUDR was more effective compared to
the study cohort size was small, the therapeutic responses systemic delivery [57]. The authors randomized 64 patients,
were promising especially in previously untreated patients. 32 for regional intra-arterial infusion and 32 for intravenous
It was hypothesized that the degree of tumor burden in the systemic infusion of FUDR. It was noted that there was a
liver had an impact of survival, especially in patients with significantly improved response for patients who received
more than 60% hepatic involvement, resulting in median intra-arterial infusion of FUDR versus systemic FUDR (62%
survival of 13 months. This is in comparison to patients vs. 17%, respectively) (p < 0.0003). The trial did not demon-

Table 43.1 Studies of hepatic artery infusion therapy and unresectable colorectal liver metastasis
Sample
Study Treatment regimen Control arm size Results
Kemeny et al. HAI FUDR None 41 PR: 12 of 23 (52%) previously treated patients
1984 Mean duration of response: 8 months
PR: 3 of 18 (17%) previously treated patients
Mean duration of response: 4 months
Chang et al. HAI FUDR Systemic 64 Significantly improved response for HAI FUDR (62%)
1987 FUDR compared to systemic (17%) p < 0.003
No difference is OS, HAI (22%) and systemic (15%), p = 0.27
Kemeny et al. HAI FUDR + dex Systemic 5-FU 135 HAI with longer OS (24.4 vs. 20 months) p = 0.0034, greater
2006 and leucovorin response rates (47% vs. 24%) p = 0.012, longer time to
(CALGB progression in the liver (9.8 vs. 7.3 months) p = 0.034 and a
9481) shorter time to extrahepatic progression (7 vs. 14.8) p = 0.29
Kemeny et al. Systemic oxaliplatin and None 49 8% CR, 84% PR
2009 irinotecan with HAI FUDR + dex 47% of patients were able to undergo resection
Median OS chemo-naive 50.8, previously treated 35 months
Ammori et al. All patients treated with HAI and None 373 Patients without extrahepatic disease had a median survival of
2012 systemic chemotherapy at a single 32 months compared to 16 months in patients with extrahepatic
institution from 2000 to 2009 disease p < 0.001
Ammori et al. All patients treated with HAI and None 373 25% of patient underwent subsequent resection or ablation
2013 systemic chemotherapy at a single Median 5-year survival from time of HAI pump placement was
institution from 2000 to 2009 59 months in patients who underwent resection or ablation and
16 months in patients who did not (p < 0.001)
D’Angelica Patients who received <2 cycles of None 49 Overall response rate 76%
et al. 2015 oxaliplatin with no symptoms 47% of patients achieved conversion to resection
received oxaliplatin, irinotecan, Median OS 38 months
and bev Median PFS 13 months
Patients who received >2 cycles of Conversion to resection only factor associated with prolonged
oxaliplatin received irinotecan, OS and PFS
5-FU, leucovorin, and bev Bev increased biliary toxicity and stopped
Combined with HAI FUDR + dex
Pak et al. Patients who received <2 cycles of None 64 54% of patients were converted to resection
2018 oxaliplatin with no symptoms Median PFS 13 months
received oxaliplatin, irinotecan, Median OS 38 months
and bev 5-year OS 36%
Patients who received >2 cycles of Conversion to resection only factor associated with prolonged
oxaliplatin received irinotecan, OS and PFS
5-FU, leucovorin, and bev
Combined with HAI FUDR + dex
Abbreviations: HAI hepatic artery infusion; FUDR floxuridine; PR partial response; CR complete response; OS overall survival; PFS progression-­
free survival; dex dexamethasone; 5-FU fluorouracil; bev bevacizumab
474 Z. Brown et al.

strate an improvement in overall survival (OS) with a 2-year with conversion to resectability included no prior chemo-
actuarial survival of 22% (intra-arterial) and 15% (intrave- therapy, a clinical risk score less than 3 (risk score including
nous), p = 0.27. The poor survival was attributed to patients lymph node-positive tumor, disease-free inter-
ultimately succumbing to systemic disease progression val < 12 months, >one hepatic metastasis, largest metastasis
despite good locoregional control. A subset analysis revealed >5 cm, and preoperative CEA > 200 ng/mL), treatment on a
that patients with negative portal nodes at the time of surgery clinical protocol, younger age, and treatment toward the end
demonstrated a significant improvement in survival of the study period [63, 64]. Additional studies confirmed
(p = 0.03). Ultimately, the study opened the next phase of these findings with conversion to resection being the only
trials examining the combined systemic and regional therapy significant factor associated with prolonged OS and PFS in
for patients with unresectable CRLM [57]. patients receiving HAI therapy [61, 62]. This finding remains
To address the efficacy of standard systemic therapy com- significant independent of method of resection including
pared to regional therapy, a randomized control trial (CALGB two-staged hepatectomy or repeated resection for recurrent
9481) was performed. The protocol randomized 135 patients disease [65, 66]. This principle holds true even for extrahe-
to either HAI therapy with FUDR or systemic 5-FU and leu- patic metastasis. The initial trial excluded patients with
covorin. Compared to systemic therapy, patients receiving extrahepatic disease including positive portal lymphadenop-
HAI experienced a longer OS (24.4 vs. 20 months), a greater athy. A study by Ammori et al. found that in patients with
response rate (47% vs. 24%), a longer time to progression in unresectable CRLM treated with HAI and systemic therapy,
the liver (9.8 vs. 7.3 months) but a shorter time to extrahe- patients without extrahepatic disease had a median survival
patic progression (7 vs. 14.8) [58]. Unfortunately, the trial of 32 months compared to 18 months in patients with one
was initiated prior to the establishment of modern chemo- site of extrahepatic disease or 9 months in patients with mul-
therapy consisting of oxaliplatin, irinotecan, or targeted ther- tiple sites of extrahepatic disease [67]. Concurrent resection
apies. Nonetheless, the use of 5-FU, the backbone of modern of both hepatic and extrahepatic disease can be associated
systemic chemotherapy, alone provided a greater response with long-term survival [68]. Again these are all based on
rate and time to liver progression thus providing evidence of reporting from a single institution study. In a meta-analysis
efficacy of combined systemic and regional therapies utiliz- of HAI for unresectable CRLM including 10 studies and
ing 5-FU. Although the application of this regimen was 1277 patients, there was no statistical difference in OS
questioned, it still established the role of systemic therapy in between HAI alone and systemic chemotherapy although
combination with regional therapy. HAI had a greater tumor response rate [69].
The ability to resect metastatic liver disease greatly influ-
ences survival. The establishment of modern systemic che-
motherapy with the incorporation of oxaliplatin and 43.5.2 Resectable Colorectal Cancer Liver
irinotecan has increased interest in interval downstaging of Metastasis
liver metastasis for potential resection. The modern regimen
allowed 30–40% of patients with previously unresectable Complete surgical resection of liver metastasis results in
liver disease to become surgical resection candidates [56, 5-year survival of 37–50% and remains the only curative
59]. Similarly, HAI therapy was viewed as a bridging ­therapy therapy [64, 70]. At a median follow-up of 30 months,
to surgical resection. Utilization of HAI therapy with FUDR approximately 60% of patients who underwent surgery for
in conjunction with systemic oxaliplatin and irinotecan curative intent for colorectal liver metastasis presented with
resulted in the conversion of 47% of previously unresectable recurrences with 43.2% of recurrences being isolated intra-
patients. A prolonged and improved prognosis was observed hepatic disease [70]. The failure of locoregional control was
particularly in chemo-naive patients, compared to chemo- predicted by R1 resection and the use of loco-ablative ther-
conditioned patients (50.8 vs. 35 months respectively, apy [70]. Conversely, the recurrence rates were significantly
p = 0.02) [60]. It is notable that the high-response rates seen lower in patients receiving adjuvant systemic chemotherapy
in this study may be due to the use of three distinct chemo- [71, 72]. With a goal to examine the efficacy of liver-directed
therapeutic agents. The addition of bevacizumab to systemic regional therapy for reduction in postoperative recurrence
therapy had no impact on survival outcomes. More impor- rates, Kemeny et al. examined the role of adjuvant HAI ther-
tantly, bevacizumab was identified as a risk factor for high apy plus intravenous systemic 5-FU compared to systemic
rates of biliary toxicity [61, 62]. chemotherapy alone following complete surgical resection
A retrospective review of a single institution experience [73] (Table 43.2). Although not statistically significant, a
revealed an improved median survival of 59 months in ini- trend toward improved (OS) with a median survival of
tially unresectable patients converted to either complete 72.2 months in the combination therapy group compared to
resection or ablation compared to 16 months in patients 59.3 months in the control group (p = 0.11) was observed.
remaining unresectable [63]. Identified factors associated Multivariate analysis validated the combination therapy as a
43 Regional Arterial Infusional Therapy as a Means of Controlling Hepatic Metastases 475

Table 43.2 Studies of hepatic artery infusion therapy and resected colorectal liver metastasis
Sample
Study Treatment regimen Control arm size Results
Kemeny et al. HAI FUDR + systemic 5-FU Systemic 5-FU 156 Median OS combination therapy 72.2 months and
1999 monotherapy 59.3 months (p = 0.21)
Median hepatic PFS combination not reached versus
monotherapy 42.7 months (p < 0.0001)
Median overall PFS combination 37.4 months versus
mono therapy 17.2 months (p = 0.06)
Kemeny et al. HAI FUDR + systemic 5-FU Systemic 5-FU 156 10-year OS combination 41.1 and monotherapy
2005 27.2 months
Kemeny et al. HAI FUDR + systemic 5-FU None 109 4-year recurrence-free rate (RFR) combination therapy
2002 45.7% versus no therapy 25.2% (p = 0.04)
4-year liver RFR combination therapy 66.9% versus no
therapy 43% (p = 0.03)
4-year OS combination therapy 61.5% versus no therapy
52.7% (p = 0.6)
House et al. HAI FUDR + systemic Systemic modern 250 5-year disease-specific survival combination therapy 76%
2011 modern chemotherapy chemotherapy versus systemic therapy 55% (p < 0.01)
5-year hepatic RFS combination therapy 79% versus
systemic therapy 55% (p < 0.001)
5-year RFS combination therapy 48% versus systemic
therapy 25% (p < 0.01)
Goéré et al. HAI FUDR + systemic 5-FU Systemic modern 498 5-year OS combination therapy 54% versus systemic
2013 chemotherapy therapy 52% (p = 0.34)
3-year DFS combination therapy 33% versus systemic
therapy 5% (p < 0.0001)
3-year hepatic DFS combination therapy 49% versus
systemic therapy 21%, (p = 0.0008)
Gholami et al. HAI FUDR No HAI 674 5-year OS (KRAS WT) HAI 78% vs no HAI 57% (p <
2019 0.001)
5-year OS (KRAS mut) HAI 59% versus no HAI 40% (p
< 0.001)
Abbreviations: HAI hepatic artery infusion; FUDR floxuridine; PR partial response; CR complete response; OS overall survival; PFS progression-­
free survival; dex dexamethasone; 5-FU fluorouracil; bev bevacizumab

predictor of improved OS (HR 2.34 (1.10–4.98), p = 0.027). improvement in OS was hypothesized to be secondary to the
Despite the lack of significant improvement in OS, the study absence of modern systemic therapy or due to extrahepatic
clearly demonstrated improved hepatic progression-free sur- disease progression following reduction of systemic therapy
vival (PFS) with hepatic PFS in the monotherapy arm being dosing. Nonetheless, HAI pump therapy was established as a
42.7 months while hepatic PFS was not reached for the com- multimodality adjuvant treatment option to reduce locore-
bination therapy (p < 0.001) [73]. At the 10-year follow-up, gional recurrence rates in high-risk patients. In a pooled
the authors were revealed a persistent and significant PFS in analysis of four consecutive protocols examining adjuvant
the combination therapy group (31.3 vs. 17.2 months, therapy with the combination of HAI and systemic therapy
p = 0.02) [74]. following liver resection, the 5- and 10-year median survival
Although the previous study introduced the concept of was found to be 66% and 48% respectively in patients treated
adjuvant HAI therapy, its true effect as an adjuvant therapy after 2003 [76].
was never examined. In 2002, Kemeny et al. examined the While modern adjuvant systemic therapy introduced
role of adjuvant HAI therapy plus intravenous 5-FU follow- oxaliplatin and irinotecan to the 5-FU backbone, these two
ing complete surgical resection versus surgical resection particular drugs were not previously evaluated for regional
alone [75]. The study demonstrated a 4-year recurrence rate therapy. With FUDR being the key component of HAI ther-
of 46% in the HAI group compared to 25% in the control apy, the question of use and efficacy of different HAI-based
arm (p = 0.04). Additionally, the HAI group had a significant chemotherapy was posed. Goere et al. sought to examine two
improvement in liver-specific recurrence-free rate compared key issues in their study. First, they examined the role of HAI
to the control arm (67% vs. 43% respectively, p = 0.03) [75]. pump therapy delivery of oxaliplatin in patients deemed high
With the goal of detecting liver-specific recurrence-free risk for recurrent disease following liver metastatectomy.
rates, the study was unfortunately not powered to detect [77] Second, they attempted to compare the role of regional
OS. Besides the study design, the failure to demonstrate delivery versus systemic delivery of modern chemotherapy
476 Z. Brown et al.

[77]. The study demonstrated a significantly prolonged In patients with KRAS wild-type tumors, the 5-year OS was
disease-­free survival in patients receiving adjuvant HAI plus 78% for patients receiving HAI therapy versus 57% in
systemic 5-FU versus modern systemic chemotherapy after patients without HAI (P < 0.001). Similarly, in patients with
resection of CRLM [77]. Similar to the previous two studies, KRAS mutant tumors, the 5-year OS with HAI was 59% ver-
the author faced difficulties in demonstrating significant sus 40% in patients treated without HAI therapy (p < 0.001)
improvement in OS. However, the benefits of oxaliplatin [82]. This indicates that although KRAS mutation is an indi-
infusion including a relatively quick infusion of oxaliplatin cator of poor tumor biology and overall worse prognosis
(2-h infusion every 2 weeks compared to a 14-day infusion long-term, regional therapy with HAI may trump the genetic
every 4 weeks), decreased rates of biliary toxicity, and a prognosis and may improve clinical outcomes.
potential decrease in the frequency of oxaliplatin-based neu-
ropathy was established. Although the study did not directly
examine the role of HAI + modern chemotherapy versus 43.6 Conclusion
modern chemotherapy, it provided evidence for potential
other chemotherapeutic options through HAI pump Additional studies are required to further define the role of
delivery. HAI in the treatment of CRLM. The implementation of HAI
The efficacy of modern systemic chemotherapy with HAI therapy has been shown to be safe but further efficacy still
compared to modern systemic chemotherapy was finally needs to be established [83, 84]. Based on available data,
examined by House et al. [78] This key study addressed the HAI appears to have a greater impact on disease progression
concerns of efficacy following the addition of HAI therapy to and recurrence without a significant OS benefit. Additionally,
modern systemic chemotherapy. The study demonstrated HAI is generally not an accepted treatment modality outside
that the combination therapy compared to systemic therapy of a clinical trial in patients with extrahepatic disease [67, 68,
was independently associated with significant improvement 85]. Basic fundamentals of tumor biology still reign: patients
in 3- and 5-year disease-specific survival (3 year: 86% vs. who have progressed on systemic chemotherapy prior to
76%, 5 year: 75% vs. 55%, p < 0.01) and 5-year hepatic HAI display a more aggressive tumor biology and have
recurrence-free survival (79% vs. 55%, p < 0.001). decreased survival compared to those who have not previ-
Interestingly, 5-year overall recurrence-free survival was ously received systemic therapy. In addition, the application
also significantly improved (48% vs. 25% p < 0.01), result- of HAI therapy to other primary or secondary liver tumors is
ing in the question of does avoidance of early liver recur- being investigated, such as intrahepatic cholangiocarcioma
rence potentially prevents other site recurrrence [78]. The (ICC). ICC is the second most common primary liver malig-
exact mechanism of this observation remains debatable. nancy following hepatocellular carcinoma (HCC). Resection
However, the authors explained that the regional disease with portal lymphadenectomy remains the only curative
control by HAI therapy may potentially contribute to the therapy. Unfortunately, most patients are initially diagnosed
improved survival benefit by providing long-term control of with either locally advanced or metastatic disease [86]. For
liver metastasis. Although this study hinted at the control of these patients, standard systemic therapy consisting of gem-
the extrahepatic recurrence with regional therapy, the citabine and cisplatin offers a median OS of 11.7 months,
­mechanism of this prevention remains unclear. It is hypoth- median PFS of 8 months, and a tumor control rate of 81.4%
esized that the exposure to systemic chemotherapy alone is [87]. These patients ultimately progress and succumb to their
the key to the suppression of extrahepatic metastasis. The disease due to ineffective therapies and limited treatment
suppression of extrahepatic recurrence was demonstrated options.
even in the setting of dose reduction of systemic chemother- HAI therapy has gained interest in the treatment of ICC.
apy in patients receiving HAI compared to standard dosing (Table 43.3). The ideal candidates for this therapy are evolving
[79]. As such, it indirectly suggested that the addition of IV but initial studies included patients with advanced, multifocal
systemic chemotherapy may be responsible for the suppres- disease confined to the liver or unresectable nature of the liver
sion of extrahepatic disease and not the HAI therapy. tumor. The initial phase II clinical trial included 34 patients
Currently, there is starting to be a greater understanding with unresectable primary liver cancer (26 ICC and 8 hepato-
on how tumor biology may influence response to therapy. cellular carcinoma (HCC)) for treatment with HAI-FUDR. Of
KRAS mutations are associated with worse survival and the 34 patients, there were 16 patients with a partial response,
higher recurrence risks after resection of CRLM [80, 81]. 14 patients had stable disease, and 3 patients had progressive
Examination of 674 patients stratified by KRAS mutational disease. Patients who responded to therapy had a median dis-
status (418 KRAS wild-type and 256 KRAS mutant), expo- ease-specific survival of 29.5 months. The patients with ICC
sure to adjuvant HAI therapy following resection revealed had a greater response rate than HCC (53.8% vs. 25%, respec-
that use of HAI was independently associated with improved tively) and thus established a potential role of HAI therapy for
survival regardless of KRAS mutational status (p < 0.002). ICC [88].
43 Regional Arterial Infusional Therapy as a Means of Controlling Hepatic Metastases 477

Subsequently, a retrospective review analyzed 104 therapy for ICC. In addition, we can aim to model additional
patients with unresectable ICC with disease confined to the trials to follow suit for CRLM. Lastly, with the application of
liver who were treated with combined HAI and systemic HAI for ICC and its comparison to resection, HAI therapy
therapy (gemcitabine-based) or systemic therapy alone [89]. becomes the medium to answer the question of the prognosis
The combination therapy demonstrated a response rate of role of HAI compared to resection.
59% compared to 39% in systemic therapy alone. Unlike the
CRLM studies, the HAI/systemic therapy group exhibited
improved OS (30.8 months vs. 18.4 months, p < 0.001)
regardless of nodal status [89].
Systemic gemcitabine and oxaliplatin were then investi- Open Questions
gated in combination with FUDR via HAI. The study The questions remain
included patients with unresectable ICC and also allowed for • Is there role of hepatic artery infusional therapy for
resectable regional lymphadenopathy. Eighty-four percent- other primary or secondary liver cancers?
age of patients achieved disease control at 6 months with • Are there other agents including small molecule
58% of patients showing a partial response. The 6-month inhibitors and immunotherapy that can be utilized
PFS was 67% for pretreated patients and 89% for chemo- as a regional therapy?
therapy-naive patients. The median OS was 25 months. • Can we establish a role of neoadjuvant regional
Similar to prior studies, nodal disease did not alter OS or therapy in patients with resectable disease?
PFS. Four patients experienced a significant enough response • Is there a difference in the efficacy of hepatic artery
to undergo resection of their tumor [90]. infusion therapy based on the sidedness of the pri-
Although there is a limitation in number of studies, the use mary colon cancer?
of HAI pump therapy, particularly in ICC is garnering • Is there a role for genetics in degerming candidacy
increased interest. Stemming from these studies, a random- and efficacy of the hepatic artery infusional therapy?
ized controlled trial is crucial to clearly establish a role of this

Table 43.3 Studies of hepatic artery infusion therapy and intrahepatic cholangiocarcinoma
Study Treatment regimen Control arm Sample size Results
Jarnagin et al. HAI FUDR + dex None 34 patients: 26 16 patients with PR
2009 ICC and 8 HCC 14 patients with stable disease
3 patients with progressive disease
Response rate: ICC 53.8% versus HCC 25%
2-year survival 67%
Kemeny et al. HAI FUDR with dex + bev None 22 patients 31.8% PR
2011 18 ICC and 4 68.2% stable disease
HCC Median survival: 31.1 months
PFS 8.45 months
Hepatic PFS 11.3 months
Adding bev appeared to increase toxicity
without clear improvement in outcomes
Konstantinidis Systemic gemcitabine-based Gemcitabine-based 104 Response rate: HAI + Chemo—59%, systemic
et al. 2016 chemotherapy with HAI chemotherapy therapy alone 39% (p = 0.11)
FUDR OS: HAI + Chemo—30.8 months, systemic
therapy alone—18.4 months (P < 0.001)
8 patients responded enough to undergo
resection with median OS of 37 months
Cercek et al. 2019 Systemic gemcitabine and None 38 58% with partial radiologic response
oxaliplatin with HAI FUDR 84% achieved disease control at 6 months
4 patients were able to undergo subsequent
resection
Median PFS 11.8 months
Median OS was 25 months
6-month progression-free rate was 67% for
pretreated patients and 89% for chemotherapy-­
naïve patients.
Abbreviations: HAI hepatic artery infusion; ICC intrahepatic cholangiocarcinoma; HCC hepatocellular carcinoma; FUDR floxuridine; PR partial
response; OS overall survival; PFS progression-free survival; dex dexamethasone; bev bevacizumab
478 Z. Brown et al.

References carcinoma using a totally implantable infusion pump. Cancer.

1980;45(5):866–9.
23. Heinrich S, Petrowsky H, Schwinnen I, et al. Technical com-
1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer plications of continuous intra-arterial chemotherapy with
J Clin. 2020;70(1):7–30. 5-­fluorodeoxyuridine and 5-fluorouracil for colorectal liver metas-
2. Fong Y, Cohen AM, Fortner JG, et al. Liver resection for colorectal tases. Surgery. 2003;133(1):40–8.
metastases. J Clin Oncol. 1997;15(3):938–46. 24. Bacchetti S, Pasqual E, Crozzolo E, Pellarin A, Cagol PP. Intra-­
3. Foster JH. Treatment of metastatic disease of the liver: a skeptic’s arterial hepatic chemotherapy for unresectable colorectal liver
view. Semin Liver Dis. 1984;4(2):170–9. metastases: a review of medical devices complications in 3172
4. van Gestel YR, de Hingh IH, van Herk-Sukel MP, et al. Patterns patients. Med Devices (Auckland, NZ). 2009;2:31.
of metachronous metastases after curative treatment of colorectal 25. Blackshear P, Rohde T, Dorman F, Buchwald H. An implantable
cancer. Cancer Epidemiol. 2014;38(4):448–54. pump for long-term intravascular drug infusion. Med Instrum.
5. van der Geest LG, Lam-Boer J, Koopman M, Verhoef C, Elferink 1981;15(4):226.
MA, de Wilt JH. Nationwide trends in incidence, treatment and 26. Burrows J, Talley R, Drake E, San Diego E, Tucker W. Infusion of
survival of colorectal cancer patients with synchronous metastases. fluorinated pyrimidines into hepatic artery for treatment of meta-
Clin Exp Metastasis. 2015;32(5):457–65. static carcinoma of liver. Cancer. 1967;20(11):1886–92.
6. Ito K, Govindarajan A, Ito H, Fong Y. Surgical treatment of hepatic 27. Clarkson B, Young C, Dierick W, et al. Effects of continuous hepatic
colorectal metastasis: evolving role in the setting of improving sys- artery infusion of antimetabolites on primary and metastatic cancer
temic therapies and ablative treatments in the 21st century. Cancer of the liver. Cancer. 1962;15(3):472–88.
J. 2010;16(2):103–10. 28. Ensminger WD, Rosowsky A, Raso V, et al. A clinical-­pharmacological
7. Akgül Ö, Çetinkaya E, Ersöz Ş, Tez M. Role of surgery in evaluation of hepatic arterial infusions of 5-fluoro-­2′-deoxyuridine and
colorectal cancer liver metastases. World J Gastroenterol. 5-fluorouracil. Cancer Res. 1978;38(11 Part 1):3784–92.
2014;20(20):6113–22. 29. Hohn DC, Rayner AA, Economou JS, Ignoffo RJ, Lewis BJ, Stagg
8. Gallagher DJ, Kemeny N. Metastatic colorectal can- RJ. Toxicities and complications of implanted pump hepatic arterial
cer: from improved survival to potential cure. Oncology. and intravenous floxuridine infusion. Cancer. 1986;57(3):465–70.
2010;78(3–4):237–48. 30. Kelsen D, Hoffman J, Alcock N, et al. Pharmacokinetics of cispla-
9. Adam R, Delvart V, Pascal G, et al. Rescue surgery for unresectable tin regional hepatic infusions. Am J Clin Oncol. 1982;5(2):173–8.
colorectal liver metastases downstaged by chemotherapy: a model 31. Garnick MB, Ensminger WD, Israel M. A clinical-pharmacological
to predict long-term survival. Ann Surg. 2004;240(4):644–57; dis- evaluation of hepatic arterial infusion of adriamycin. Cancer Res.
cussion 657–648. 1979;39(10):4105–10.
10. Adam R, Vinet E. Regional treatment of metastasis: surgery of 32. Ensminger W, Gyves J. Clinical pharmacology of hepatic arterial
colorectal liver metastases. Ann Oncol. 2004;15(Suppl 4):103–6. chemotherapy. Paper presented at Seminars in oncology; 1983.
11. Chu DZ, Hutchins L, Lang NP. Regional chemotherapy of liver 33. Kemeny N, Daly J, Oderman P, et al. Hepatic artery pump infu-
metastases from colorectal carcinoma: hepatic artery or portal vein sion: toxicity and results in patients with metastatic colorectal car-
infusion? Cancer Treat Rev. 1988;15(4):243–56. cinoma. J Clin Oncol. 1984;2(6):595–600.
12. Breedis C, Young G. The blood supply of neoplasms in the liver. 34. Kemeny N, Seiter K, Niedzwiecki D, et al. A randomized trial of
Am J Pathol. 1954;30(5):969–77. intrahepatic infusion of fluorodeoxyuridine with dexamethasone
13. Fortner JG, Beattie EJ Jr, Shiu MH, et al. Surgery in liver tumors. versus fluorodeoxyuridine alone in the treatment of metastatic
Curr Probl Surg. 1972;9(6):1–56. colorectal cancer. Cancer. 1992;69(2):327–34.
14. Nilsson LA. Therapeutic hepatic artery ligation in patients with 35. Cercek A, D’Angelica M, Power D, et al. Floxuridine hepatic
secondary liver tumors. Rev Surg. 1966;23(5):374–6. arterial infusion associated biliary toxicity is increased by concur-
15. Mori W, Masuda M, Miyanaga T. Hepatic artery ligation and tumor rent administration of systemic bevacizumab. Ann Surg Oncol.
necrosis in the liver. Surgery. 1966;59(3):359–63. 2014;21(2):479–86.
16. Almersjö O, Bengmark S, Engevik L, Hafström L, Loughridge B, 36. Kemeny MM, Battifora H, Blayney DW, et al. Sclerosing cholan-
Nilsson L. Serum enzyme changes after hepatic dearterialization in gitis after continuous hepatic artery infusion of FUDR. Ann Surg.
man. Ann Surg. 1968;167(1):9. 1985;202(2):176.
17. Plengvanit U. Treatment of primary carcinoma of the liver by hepatic 37. Daly JM, Kemeny N, Oderman P, Botet J. Long-term hepatic arterial
artery ligation: Preliminary report of 409 cases. Paper presented at infusion chemotherapy: anatomic considerations, operative tech-
International Association for the Study of the Liver. Liver Research; nique, and treatment morbidity. Arch Surg. 1984;119(8):936–41.
translation of the 3^< rd> International Symposium, Tokyo and 38. Kemeny MM, Hogan JM, Goldberg DA, et al. Continuous hepatic
Kyoto, 1966, Antwerp, Tijdschrift voor Gastroenterologie; 1967. artery infusion with an implantable pump: problems with hepatic
18. Madding GF, Kennedy PA, Sogemeier E. Hepatic artery ligation for artery anomalies. Surgery. 1986;99(4):501–4.
metastatic tumor in the liver. Am J Surg. 1970;120(1):95–6. 39. Rayner AA, Kerlan RK, Stagg RJ, Price DC, Hohn DC. Total
19. Sullivan RD, Norcross JW, Watkins E Jr. Chemotherapy of meta- hepatic arterial perfusion after occlusion of variant lobar ves-
static liver cancer by prolonged hepatic-artery infusion. N Engl J sels: implications for hepatic arterial chemotherapy. Surgery.
Med. 1964;270(7):321–7. 1986;99(6):708–15.
20. Wirtanen GW, Bernhardt LC, Mackman S, Ramirez G, Curreri 40. Allen PJ, Nissan A, Picon AI, et al. Technical complications
AR, Ansfield FJ. Hepatic artery and celiac axis infusion for and durability of hepatic artery infusion pumps for unresectable
the treatment of upper abdominal malignant lesions. Ann Surg. colorectal liver metastases: an institutional experience of 544 con-
1968;168(1):137–41. secutive cases. J Am Coll Surg. 2005;201(1):57–65.
21. Reed ML, Vaitkevicius VK, Al-Sarraf M, et al. The practicality of 41. Qadan M, D’Angelica MI, Kemeny NE, Cercek A, Kingham
chronic hepatic artery infusion therapy of primary and metastatic TP. Robotic hepatic arterial infusion pump placement. HPB.
hepatic malignancies: ten-year results of 124 patients in a prospec- 2017;19(5):429–35.
tive protocol. Cancer. 1981;47(2):402–9. 42. Leal JN, Kingham TP. Hepatic artery infusion chemotherapy for
22. Buchwald H, Grage TB, Vassilopoulos PP, Rohde TD, Varco RL, liver malignancy. Surg Oncol Clin. 2015;24(1):121–48.
Blackshear PJ. Intraarterial infusion chemotherapy for hepatic
43 Regional Arterial Infusional Therapy as a Means of Controlling Hepatic Metastases 479

43. Arru M, Aldrighetti L, Gremmo F, et al. Arterial devices for regional with unresectable hepatic metastases from colorectal cancer:
hepatic chemotherapy: transaxillary versus laparotomic access. J conversion to resection and long-term outcomes. Ann Surg.
Vasc Access. 2000;1(3):93–9. 2015;261(2):353–60.
44. Franklin ME, Gonzalez JJ. Laparoscopic placement of hepatic 62. Pak LM, Kemeny NE, Capanu M, et al. Prospective phase II trial
artery catheter for regional chemotherapy infusion: technique, of combination hepatic artery infusion and systemic chemotherapy
benefits, and complications. Surg Laparosc Endosc Percutan Tech. for unresectable colorectal liver metastases: long term results and
2002;12(6):398–407. curative potential. J Surg Oncol. 2018;117(4):634–43.
45. Dhir M, Zenati MS, Padussis JC, et al. Robotic assisted placement 63. Ammori JB, Kemeny NE, Fong Y, et al. Conversion to complete
of hepatic artery infusion pump is a safe and feasible approach. J resection and/or ablation using hepatic artery infusional chemo-
Surg Oncol. 2016;114(3):342–7. therapy in patients with unresectable liver metastases from colorec-
46. Dhir M, Magge D, Novak S, Bartlett DL, Zureikat AH. Robotic-­ tal cancer: a decade of experience at a single institution. Ann Surg
assisted placement of an hepatic artery infusion pump and cath- Oncol. 2013;20(9):2901–7.
eter for regional chemotherapy of the liver. Ann Surg Oncol. 64. Fong Y, Fortner J, Sun RL, Brennan MF, Blumgart LH. Clinical
2016;23(5):755–6. score for predicting recurrence after hepatic resection for metastatic
47. Campbell KA, Burns RC, Sitzmann JV, Lipsett PA, Grochow colorectal cancer: analysis of 1001 consecutive cases. Ann Surg.
LB, Niederhuber JE. Regional chemotherapy devices: effect 1999;230(3):309–18; discussion 318–321.
of experience and anatomy on complications. J Clin Oncol. 65. Brouquet A, Abdalla EK, Kopetz S, et al. High survival rate
1993;11(5):822–6. after two-stage resection of advanced colorectal liver metastases:
48. Scaife CL, Curley SA, Izzo F, et al. Feasibility of adjuvant hepatic response-based selection and complete resection define outcome. J
arterial infusion of chemotherapy after radiofrequency ablation Clin Oncol. 2011;29(8):1083–90.
with or without resection in patients with hepatic metastases from 66. Cardona K, Donataccio D, Kingham TP, et al. Treatment of exten-
colorectal cancer. Ann Surg Oncol. 2003;10(4):348–54. sive metastatic colorectal cancer to the liver with systemic and
49. Curley SA, Chase JL, Roh MS, Hohn DC. Technical considerations hepatic arterial infusion chemotherapy and two-stage hepatic resec-
and complications associated with the placement of 180 implantable tion: the role of salvage therapy for recurrent disease. Ann Surg
hepatic arterial infusion devices. Surgery. 1993;114(5):928–35. Oncol. 2014;21(3):815–21.
50. Sofocleous CT, Schubert J, Kemeny N, et al. Arterial embolization 67. Ammori JB, D’Angelica MI, Fong Y, et al. Hepatic artery infusional
for salvage of hepatic artery infusion pumps. J Vasc Interv Radiol. chemotherapy in patients with unresectable colorectal liver metas-
2006;17(5):801–6. tases and extrahepatic disease. J Surg Oncol. 2012;106(8):953–8.
51. Aloia TA, Vauthey JN, Loyer EM, et al. Solitary colorectal liver 68. Carpizo DR, Are C, Jarnagin W, et al. Liver resection for metastatic
metastasis: resection determines outcome. Arch Surg (Chicago, Ill: colorectal cancer in patients with concurrent extrahepatic disease:
1960). 2006;141(5):460–6; discussion 466–467. results in 127 patients treated at a single center. Ann Surg Oncol.
52. Hur H, Ko YT, Min BS, et al. Comparative study of resection and 2009;16(8):2138–46.
radiofrequency ablation in the treatment of solitary colorectal liver 69. Mocellin S, Pilati P, Lise M, Nitti D. Meta-analysis of hepatic arte-
metastases. Am J Surg. 2009;197(6):728–36. rial infusion for unresectable liver metastases from colorectal can-
53. Lee WS, Yun SH, Chun HK, et al. Clinical outcomes of hepatic cer: the end of an era? J Clin Oncol. 2007;25(35):5649–54.
resection and radiofrequency ablation in patients with solitary 70. de Jong MC, Pulitano C, Ribero D, et al. Rates and patterns of
colorectal liver metastasis. J Clin Gastroenterol. 2008;42(8):945–9. recurrence following curative intent surgery for colorectal liver
54. Van Cutsem E, Nordlinger B, Adam R, et al. Towards a pan-­ metastasis: an international multi-institutional analysis of 1669
European consensus on the treatment of patients with colorec- patients. Ann Surg. 2009;250(3):440–8.
tal liver metastases. Eur J Cancer (Oxford, England: 1990). 71. Nordlinger B, Sorbye H, Glimelius B, et al. Perioperative
2006;42(14):2212–21. FOLFOX4 chemotherapy and surgery versus surgery alone for
55. Kemeny N. Management of liver metastases from colorectal cancer. resectable liver metastases from colorectal cancer (EORTC 40983):
Oncology (Williston Park). 2006;20(10):1161–76, 1179; discus- long-term results of a randomised, controlled, phase 3 trial. Lancet
sion 1179–1180, 1185–1166. Oncol. 2013;14(12):1208–15.
56. Alberts SR, Horvath WL, Sternfeld WC, et al. Oxaliplatin, fluo- 72. Wang ZM, Chen YY, Chen FF, Wang SY, Xiong B. Peri-operative
rouracil, and leucovorin for patients with unresectable liver-only chemotherapy for patients with resectable colorectal hepatic metas-
metastases from colorectal cancer: a north central cancer treatment tasis: a meta-analysis. Eur J Surg Oncol. 2015;41(9):1197–203.
group phase II study. J Clin Oncol. 2005;23(36):9243–9. 73. Kemeny N, Huang Y, Cohen AM, et al. Hepatic arterial infusion of
57. Chang AE, Schneider PD, Sugarbaker PH, Simpson C, Culnane chemotherapy after resection of hepatic metastases from colorectal
M, Steinberg SM. A prospective randomized trial of regional cancer. N Engl J Med. 1999;341(27):2039–48.
versus systemic continuous 5-fluorodeoxyuridine chemother- 74. Kemeny NE, Gonen M. Hepatic arterial infusion after liver resec-
apy in the treatment of colorectal liver metastases. Ann Surg. tion. N Engl J Med. 2005;352(7):734–5.
1987;206(6):685–93. 75. Kemeny MM, Adak S, Gray B, et al. Combined-modality treatment
58. Kemeny NE, Niedzwiecki D, Hollis DR, et al. Hepatic arterial infu- for resectable metastatic colorectal carcinoma to the liver: surgi-
sion versus systemic therapy for hepatic metastases from colorectal cal resection of hepatic metastases in combination with continu-
cancer: a randomized trial of efficacy, quality of life, and molecular ous infusion of chemotherapy—an intergroup study. J Clin Oncol.
markers (CALGB 9481). J Clin Oncol. 2006;24(9):1395–403. 2002;20(6):1499–505.
59. Pozzo C, Basso M, Cassano A, et al. Neoadjuvant treatment of unre- 76. Kemeny NE, Chou JF, Boucher TM, et al. Updated long-term sur-
sectable liver disease with irinotecan and 5-fluorouracil plus folinic vival for patients with metastatic colorectal cancer treated with liver
acid in colorectal cancer patients. Ann Oncol. 2004;15(6):933–9. resection followed by hepatic arterial infusion and systemic chemo-
60. Kemeny NE, Melendez FD, Capanu M, et al. Conversion to resect- therapy. J Surg Oncol. 2016;113(5):477–84.
ability using hepatic artery infusion plus systemic chemotherapy 77. Goéré D, Benhaim L, Bonnet S, et al. Adjuvant chemotherapy after
for the treatment of unresectable liver metastases from colorectal resection of colorectal liver metastases in patients at high risk of
carcinoma. J Clin Oncol. 2009;27(21):3465–71. hepatic recurrence: a comparative study between hepatic arterial
61. DʼAngelica MI, Correa-Gallego C, Paty PB, et al. Phase II trial of infusion of oxaliplatin and modern systemic chemotherapy. Ann
hepatic artery infusional and systemic chemotherapy for patients Surg. 2013;257(1):114–20.
480 Z. Brown et al.

78. House MG, Kemeny NE, Gönen M, et al. Comparison of adjuvant alone in patients with low-risk resectable colorectal liver metas-
systemic chemotherapy with or without hepatic arterial infusional tases—the multicenter randomized controlled PUMP trial. BMC
chemotherapy after hepatic resection for metastatic colorectal can- Cancer. 2019;19(1):327.
cer. Ann Surg. 2011;254(6):851–6. 85. Karanicolas PJ, Metrakos P, Chan K, et al. Hepatic arterial infusion
79. Srouji RM, Narayan RR, Boerner T, et al. Extrahepatic recurrence pump chemotherapy in the management of colorectal liver metasta-
rates in patients receiving adjuvant hepatic artery infusion and sys- ses: expert consensus statement. Curr Oncol. 2014;21(1):e129–36.
temic chemotherapy after complete resection of colorectal liver 86. El-Diwany R, Pawlik TM, Ejaz A. Intrahepatic cholangiocarci-
metastases. J Surg Oncol. 2020;122(8):1536–42. noma. Surg Oncol Clin N Am. 2019;28(4):587–99.
80. Karagkounis G, Torbenson MS, Daniel HD, et al. Incidence and prog- 87. Valle J, Wasan H, Palmer DH, et al. Cisplatin plus gemcitabine
nostic impact of KRAS and BRAF mutation in patients undergoing versus gemcitabine for biliary tract cancer. N Engl J Med.
liver surgery for colorectal metastases. Cancer. 2013;119(23):4137–44. 2010;362(14):1273–81.
81. Kemeny NE, Chou JF, Capanu M, et al. KRAS mutation influences 88. Jarnagin WR, Schwartz LH, Gultekin DH, et al. Regional chemo-
recurrence patterns in patients undergoing hepatic resection of therapy for unresectable primary liver cancer: results of a phase II
colorectal metastases. Cancer. 2014;120(24):3965–71. clinical trial and assessment of DCE-MRI as a biomarker of sur-
82. Gholami S, Kemeny NE, Boucher TM, et al. Adjuvant hepatic vival. Ann Oncol. 2009;20(9):1589–95.
artery infusion chemotherapy is associated with improved survival 89. Konstantinidis IT, Groot Koerkamp B, Do RK, et al. Unresectable
regardless of kras mutation status in patients with resected colorec- intrahepatic cholangiocarcinoma: systemic plus hepatic arterial infu-
tal liver metastases: a retrospective analysis of 674 patients. Ann sion chemotherapy is associated with longer survival in comparison
Surg. 2020;272(2):352–6. with systemic chemotherapy alone. Cancer. 2016;122(5):758–65.
83. Buisman FE, Grünhagen DJ, Homs MYV, et al. Adjuvant hepatic 90. Cercek A, Boerner T, Tan BR, et al. Assessment of hepatic arterial
arterial infusion pump chemotherapy after resection of colorec- infusion of floxuridine in combination with systemic gemcitabine
tal liver metastases: results of a safety and feasibility study in the and oxaliplatin in patients with unresectable intrahepatic cholangio-
Netherlands. Ann Surg Oncol. 2019;26(13):4599–607. carcinoma: a phase 2 clinical trial. JAMA Oncol. 2019;6(1):60–7.
84. Buisman FE, Homs MYV, Grünhagen DJ, et al. Adjuvant hepatic
arterial infusion pump chemotherapy and resection versus resection
 Part XII
Cancer Metastasis to the Bone

David M. Joyce and Odion Binitie

Bones are the third most common site for metastatic disease following the lung and liver. Some
of the most common cancers that spread to bone include cancers of breast, prostate, lung, thy-
roid, and kidney as well as multiple myeloma. Other cancers likely to spread to bone that also
have a propensity for lymphatic spread include head and neck cancer, squamous cell, cervical,
uterine carcinomas, and melanoma. As people survive longer with cancer, more will likely
develop significant metastatic bone disease burden necessitating the involvement of specialties
such as orthopedics, spine surgery, and interventional radiology. Understanding the mecha-
nisms of cancer spread, the effect of metastatic disease on the bone microscopically and mac-
roscopically impacts the surgical, medical, and percutaneous treatment.
During the workup of a patient with suspected metastatic bone disease, orthogonal radio-
graphs of the affected site; a CT scan involving the chest, abdomen, and pelvis with IV con-
trast; a bone scan; and/or a skeletal survey (should there be concern for multiple myeloma) are
obtained. Cross-sectional imaging of the involved bone with the favored test being an MRI
with and without contrast to better evaluate the lesion is sometimes obtained. If a MRI is not
feasible a CT with contrast can be substituted. A biopsy will often be obtained for diagnosis
prior to any surgical intervention.
Traditionally surgeons have used a scoring system to determine if a patient is at risk for
fracture. Mirels score is based on several factors including the location of the lesion, amount
of pain, and bone destruction in long bones. A sum of 9 or greater was shown to be a bone at
risk for pathologic fracture; three-dimensional analysis of lesion size and extent with CT imag-
ing may improve risk assessment. Prevention of a fracture has the benefit of limiting morbidity
from the fracture, and some studies have demonstrated improved overall survival. Spinal insta-
bility neoplastic score (SINS) was introduced in 2010 to help guide clinicians, scoring patients
into groups of stable, potentially unstable, and unstable. However, there is limited evidence on
the use of SINS as a predictive tool for the progression of deformity or fracture.
In this section we review the factors surrounding the metastatic spread of cancer to the bone
and its management. Frieling and Lynch in their chapter on the molecular basis of cancer
metastasis discuss normal bone physiology, highlight how that homeostasis is affected by the
“vicious cycle” of metastatic disease, and review the therapeutic interventions that work to
disrupt this. The next group of chapters discuss the surgical management of metastatic bone
disease based on the anatomic site. Long bones of the upper (Greene et al.) and lower extremi-
482 Cancer Metastasis to the Bone

ties (Mesko and Nystrom) are reviewed with the indications for procedures based on the site of the bone
affected. The use of different forms of fixation, bone augmentation, resection, and reconstruction and their
associated complications are discussed. The pelvis, given its distinct anatomic considerations, is covered in
a separate chapter (Evans et al.), as is the spine (Wang et al.). The utilization of minimal invasive procedures
is highlighted in each surgical chapter and is focused on in the chapter on interventional radiology tech-
niques by Parikh and Ahmed.
 Molecular Mechanisms of Metastasis
to the Bone 44
Jeremy S. Frieling and Conor C. Lynch

Abstract Learning Objectives

Bone metastasis is an incurable and lethal aspect of many
advanced-stage malignancies that significantly impacts • Review normal bone physiology with respect to
patient quality of life. Although it is often associated with cancer-induced bone disease
advanced disease, bone metastasis can occur early in dis- • Understand the common genetic and molecular
ease progression and is dictated, in part, by a combination determinants of skeletal metastasis
of genetic and molecular factors from the heterogeneous • Identify the contributions of bone niches to meta-
primary tumor. In order to prepare future metastatic sites, static tumor dissemination and dormancy
signals from the primary tumor precondition the bone for • Dissect the vicious cycle interactions between
the arrival of metastatic cells. Bone marrow niches are tumor, osteoblasts, osteoclasts, and other cells in
further co-opted to assist with the recruitment and protec- the tumor–bone microenvironment
tion of metastatic cells, while the physiological activities • Describe the mechanisms of action for the current
of osteoclasts and osteoblasts, as well as other host bone standard of care therapies
cell types, can contribute to metastatic outgrowth and sur-
vival. The resulting metastases generate lesions of unique
osteolytic and/or osteoblastic pathology, depending on
the primary cancer type. Understanding the factors driv- 44.1 Introduction
ing bone metastasis has led to the application of current
standard of care antiresorptive therapeutics, and the con- Skeletal colonization marks a significant turning point in can-
tinued elucidation of additional mechanisms is aiding the cer progression for which prognosis is poor and clinical thera-
development of novel strategies to tackle specific aspects pies are limited. Many cancers metastasize to the bone,
of these lesions to more effectively inhibit disease including breast, prostate, lung, melanoma, kidney, and thy-
progression. roid. Other skeletal malignancies such as multiple myeloma
and lymphoma also colonize the systemic skeleton. For rea-
sons that we are only beginning to comprehend, breast and
prostate cancer are among the most prevalent with bone
metastases occurring in 65–90% of patients [1]. Yet, the result-
ing skeletal lesions for each cancer type are characterized by
J. S. Frieling · C. C. Lynch (*)
Department of Tumor Biology, H. Lee Moffitt Cancer Center & unique pathologies: osteoblastic, osteolytic, and mixed lesions.
Research Institute, Tampa, FL, USA These lesions are largely generated by the activities of bone-
e-mail: [email protected]; [email protected] building osteoblasts and bone-resorbing osteoclasts in

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 483
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_44
484 J. S. Frieling and C. C. Lynch

response to tumor signals. In addition to increased patient pled activities of osteoclasts and osteoblasts at a rate of
mortality, colonization of the skeleton significantly deterio- approximately 10% per year. Remodeling occurs in local-
rates the quality of life with complications such as pain and ized compartments consisting of osteoclasts, osteoblasts,
skeletal-related events (SREs) including pathological fracture, and osteocytes that preserve the tightly regulated process
hypercalcemia, and spinal cord and nerve compression. [4]. Osteoclasts are large, multinucleated myeloid lineage
Like metastasis to other organ sites, metastasis to bone is cells that adhere to the bone surface and are responsible
highly inefficient, and disseminated tumor cells (DTCs) for resorption of existing bone tissue. Mature osteoclasts
must overcome a sequence of challenges beginning with form by the fusion of myeloid precursors in response to
localized invasion and migration from the primary site and osteoblast-­derived (also osteocyte and immune cells) fac-
concluding with extravasation and secondary tumor out- tors receptor activator of nuclear factor kappa-B ligand
growth in the bone. Modulation of both primary and second- (RANKL) and macrophage colony-stimulating factor
ary organ microenvironments aids in the establishment of (M-CSF). Under appropriate environmental cues, such as
metastatic lesions, requiring coordination between the tumor biochemical signals from mechano-sensing osteocytes,
and resident osteoblasts, osteoclasts, and osteocytes, in addi- osteoclasts polarize to form a ruffled border, which
tion to endothelial and immune cell types. Bone metastasis increases their surface area and form a tight actin ring
ultimately usurps normal skeletal remodeling and homeosta- with the bone surface to create a covered space where
sis to encourage invasion, homing, and eventual outgrowth acid, collagenases, and proteases are secreted to deminer-
of occult metastases [2]. While our understanding of these alize the bone matrix. In doing so, components of the
processes remains incomplete, expanding the knowledge of bone including embedded growth factors are released into
specific underlying mechanisms has led to the development the bone microenvironment. Subsequent to osteoclasts,
of several standard of care therapies and ongoing research osteoblasts generate bone in remodeled areas. Osteoblasts
continues to generate novel treatments. are derived from the mesenchymal lineage and differenti-
ate in response to signaling cues such as bone morphoge-
netic proteins (BMPs), transforming growth factor beta
44.2 The Bone (TGF-β), Wnt/β-catenin, and parathyroid hormone-­related
protein (PTHrP). Mature osteoblasts are recruited to the
Understanding normal bone physiology and its cellular remodeling compartment where they secrete type I
players offers insights into the local and systemic changes collagen-­rich osteoid matrix. After building bone, a frac-
that result in bone metastasis. Bone is a dynamic tissue tion of osteoblasts become embedded into the bone matrix
which not only provides for support and movement but and terminally differentiate into mechano-sensing osteo-
also serves as a reservoir for minerals such as Ca2+ and cel- cytes which communicate via long cytoplasmic processes
lular energy generated by osteoblasts. It is largely com- that extend through passages in the bone known as cana-
prised of a type I collagen-­rich extracellular matrix that is liculi. Others reside on the bone surface as bone-lining
mineralized by deposition of hydroxyapatite, a combina- cells. Osteocytes respond to mechanical stimuli and sig-
tion of magnesium, calcium, and phosphate ions. This nals from adjacent cells to regulate bone remodeling by
composition produces a stiff, durable structure where the releasing sclerostin and/or Dickkopf-1 (Dkk1) [5]. The
collagen fibers impart flexibility to prevent brittleness or cellular composition of bone is highly diverse, comprised
fracture. Two primary types of bone tissue are found within of both resident and transit cells such as endothelial cells,
the skeleton: cortical and trabecular. Cortical bone is dense mesenchymal stem cells, adipocytes, platelets myeloid/
and serves primarily structural and mechanical purposes myeloid-derived suppressor cells (MDSC), T cells, and
whereas the lighter, more metabolically active trabecular nerve cells. Unique roles for many of these cell types and
bone is located with the bone marrow compartment and their activities in bone remodeling have been reviewed
assists the maintenance of hematopoietic stem cells elsewhere [6].
(HSCs) and mesenchymal stem cells (MSCs). Surrounding
trabecular bone is the bone marrow, a site for hematopoi-
esis, producing myeloid and lymphoid ­lineages [3]. In the 44.3 Cancer Bone Tropism
context of cancer, local factors embedded within the bone
matrix, together with systemic factors from the primary Current evidence indicates that organotropism is dictated
tumor, not only regulate the specific cell types in bone but by a combination of genetic and molecular factors, circu-
can influence the behavior of DTCs. lation patterns, and crosstalk between the primary tumor
The skeleton undergoes routine systematic remodeling and metastatic and host microenvironment of the bone
whereby old bone is resorbed and replaced via the cou- (Fig. 44.1).
44 Molecular Mechanisms of Metastasis to the Bone 485

a
MMPs
PTHrP
LOX OPN
EVs

b
Primary Tumor DTC
M
c lo
id HCAM
SC
Metastatic ye
M β3 miRNAs
Tumor Cells

CX
αv

CR
β1

2 C
XC
Lx
Osteolytic Factors Osteoblastic Factors

2
TGF-β↑
(IL-6, IL-1β, PTHrP, MMPs) (BMPs, FGFs, WNTs) HGF↑
CCL2↑
Osteoclasts Atf3↑ Osteoblast

Osteoclast Mesenchymal
Precursors Stem Cells
Bone-derived
Growth Factors
(TGF-β, IGFs, PDGF, BMPs)

RANKL

New Bone Lesion

Osteoblasts

Osteoclasts
Osteocytes

Fig. 44.1 (a) Metastatic tumor cells disseminate from primary organ pre-metastatic niche that offers a suitable environment for the arrival of
tumor masses via lymphatic and hematogenous vasculature. The pri- disseminated tumor cells (DTCs). (c) DTCs that successfully colonize
mary tumor provides signals in the form of secreted proteins and the bone microenvironment generate a habitat favorable to growth and
enzymes as well as cargo-shuttling extracellular vesicles such as exo- survival, where tumor-derived factors contribute to the formation of
somes. (b) Cues from the primary tumor mediate interactions with stro- osteolytic and/or osteoblastic lesions via the “vicious cycle.” Illustration
mal components in future distant sites of metastasis to prepare a created with BioRender

44.3.1 O
 steotropism and Gene Signatures/ Ewing, who posited that metastasis was dictated by the anat-
Determinants omy of the vasculature and lymphatics in which the dissemi-
nated cells traveled. The first organ encountered would be
The predilection of specific tumor types to metastasize to the site where most tumor cells arrest due to mechanical
select organs such as bone (“osteotropism”) has long been limitations of the diameter of the vessels. He speculated that
acknowledged. More than 150 years ago, German patholo- the large number of cells released into circulation from the
gist Rudolph Virchow proposed that the locations of meta- primary site would allow for a few physically trapped cells to
static outgrowths were dictated based on where the colonize [9].
disseminated cell lodged and arrested within the vasculature, Modern evidence suggests that these two theories are not
thus explaining why vasculature-rich organs such as the mutually exclusive, with bone offering a strong case in point.
lungs and bones are common sites of metastasis [7]. In the The red bone marrow has a high rate of blood flow, deliver-
late 1800s, observations of metastatic breast cancer cases led ing abundant numbers of DTCs into bone. DTCs invade the
English surgeon Stephen Paget to realize that the metastases sinusoids of the bone marrow cavity and migrate to the bone
occurred in a nonrandom manner. Paget proposed the “seed surface. Bone metastases are most commonly located in
and soil hypothesis,” which suggested that the challenges of bones rich in red marrow such as the vertebrae, ribs, pelvis,
metastasis begin before cells leave the primary tumor and and ends of long bones but are rarely found in bones of the
that the DTCs (the seed) could only establish and flourish in hands or feet [10]. The aforementioned sites are also rich in
a secondary site with a conducive microenvironment (the metabolically active trabecular bone, which would present a
soil) [8]. Paget’s hypothesis was not universally accepted, more conducive soil for growth. Further, tumor cells must
however, and was challenged by others including James express compatible adhesion molecules and produce angio-
486 J. S. Frieling and C. C. Lynch

genic and bone-resorbing factors to remodel the bone micro- type 4 (CXCR4) chemokine axis is one of the best-­
environment for successful outgrowth [11]. Therefore, in characterized recruitment mechanisms in bone. CXCL12 is
addition to arriving in bone via lymphatic and hematogenous expressed by osteoblasts and bone marrow stromal cells
vasculature, genetic and molecular determinants also con- while the receptor, CXCR4, is highly expressed by many
tribute to osteotropism. cancer types with bone metastatic ability [14]. The study of
The importance of genetic cooperation in giving rise to serially selected MDA-MB-231 human breast cancer cell
metastatic variants that successfully colonize bone was ele- subpopulations described above identified a gene signature
gantly demonstrated by Massagué and colleagues almost associated with increased capacity to metastasize to bone,
two decades ago. In this key study, subpopulations of and CXCR4 was among the genes in this signature [12]. In
MDA-MB-231 breast cancer cells in vivo with enhanced addition to homing, CXCR4/CXCL12 interactions may also
metastatic potential [12]. Microarray profiling of the bone promote proliferation of breast and prostate cancer cells in
metastatic subpopulations revealed a gene signature consist- bone [15]. Additional chemokine axes that facilitate bone
ing of IL11, connective tissue growth factor (CTGF), metastasis include CXCL12/CXCR7, CXCR2/CXCL5,
CXCR4, and matrix metalloproteinase-1 (MMP-1). Notably, macrophage chemoattract protein-1 (MCP-1)/C–C motif
these genes were associated with functions including inva- chemokine receptor 2 (CCR2), and CXCL16/CXCR6 [16].
sion, bone marrow homing/extravasation, angiogenesis, and The expression of integrins on the surface of tumor cells
osteolysis. A shortage of models that accurately recapitulate can also facilitate interactions between the tumor and extra-
bone metastasis in vivo has traditionally hampered similar cellular matrix in bone. For example, ανβ3 integrin can assist
investigations in prostate cancer. The recent development of tumor cell adhesion by interacting with osteopontin (OPN),
NPKEYFP mice (Nkx3.1CreERT2/+; Ptenflox/flox; KrasLSL-G12D/+; fibronectin, and vitronectin proteins in bone, and expression
R26R-CAG-LSL-EYFP/+) offers an in vivo model where DTCs of ανβ3 integrin by breast and prostate cancer cells is associ-
reach and develop metastases in bone as indicated by yellow ated with increased bone metastasis, osteolysis, and coloni-
fluorescent protein (YFP) imaging. Interrogating the bone zation [17]. Similarly, members of the β1 integrin family
metastases using this model revealed a MYC/RAS co-­ (α5β1, α2β1, and α4β1) have been shown to mediate interac-
activation signature specific to bone metastases in prostate tions between the bone marrow stroma and leukemia,
cancer [13]. Silencing of MYC by shRNA significantly myeloma, prostate, and breast cancer cells, thus contributing
reduced bone metastases in the spine, pelvis, femur, tibia, to increased tumor cell colonization and survival in bone.
and humerus but did not impact lung metastases. Notably, Expression of other cell adhesion factors such as VCAM1,
this signature correlates with human prostate cancer where selectins, and galectin 3 has also been implicated [18].
activation of MYC and RAS is frequently observed in
advanced tumors. Further comparative analysis of mouse
bone metastasis and human prostate cancer based on these 44.3.3 Pre-metastatic Niche
studies has established a human 16-gene signature that is
associated with time to metastasis and response to antiandro- The development of metastasis is often considered a late
gen treatment in metastatic disease and could possess clini- event in cancer progression, however, biological evidence
cal utility with continued development. In many cases, such as the detection of DTCs in the bones of 4- to 9-week-­
patients with advanced disease may already have DTCs in old breast cancer mouse models (HER2/neu and PyMT
the bone at the time of presentation. The development of pre- transgenic) indicates that the process can occur much earlier
dictive gene signatures holds potential for identifying [19]. Organotropism arises during the primary stages of
patients at risk of developing bone metastases. malignancy as tumor-derived factors systemically remodel
future sites of metastasis in distant organs to prepare a “pre-­
metastatic niche.” The pre-metastatic niche concept was first
44.3.2 Molecular Pathways/Homing demonstrated using in vivo models of melanoma where vas-
cular endothelial growth factor receptor 1 (VEGFR1) + bone
According to the seed and soil theory, DTCs must be suited marrow-derived hematopoietic precursor cells (HPCs)
to home to and survive in the secondary organ niche, while homed to sites of future metastasis in the lung, forming clus-
this niche must reciprocally be a fertile soil that can provide ters and increasing fibronectin production prior to the arrival
cues and nutrients to DTCs. These cues are often generated of metastatic cancer cells [20]. In bone metastasis, primary
via mutual interactions with bone marrow stromal cells tumors condition the bone marrow microenvironment via
through integrins and chemokines. Tumor cells use homing circulating factors that assist in making the bone more con-
mechanisms similar to those employed by normal cells, such ducive to colonization. For example, PTHrP produced by
as HSCs, in order to metastasize to the bone. The C–X–C prostate tumors promotes bone resorption and enhances the
motif chemokine 12 (CXCL12)/C–X–C chemokine receptor production of factors in the bone marrow such as chemokine
44 Molecular Mechanisms of Metastasis to the Bone 487

ligand 2 (CCL2) that in turn can facilitate tumor growth [21]. secondary site niches where they are less vulnerable to
Primary tumors have also been shown to secrete lysyl oxi- immune surveillance, radiation, and chemotherapies. A
dase (LOX) which could regulate ECM remodeling and number of dormancy mechanisms have been proposed to
recruiting bone marrow stromal cells to pre-metastatic date. For example, “tumor cell dormancy” is defined as soli-
niches. High LOX expression by primary breast tumors has tary cells in a quiescent state, harboring decreased expres-
also been shown to alter NFATc1 expression in osteoclasts, sion of proliferation markers. This form of dormancy can be
enhancing osteolytic activity in pre-metastatic niches [22]. induced by local microenvironment factors, systemic signals
Tumor-derived OPN is reported to contribute to pre-­ from the primary tumor, or in some cases, drug therapies. In
metastatic niche formation via the induction of growth and contrast, “tumor mass dormancy” arises from DTCs that ini-
chemotactic factors in the bone, namely hepatocyte growth tially proliferate at the metastatic site but do not develop
factor (HGF) and TGF-β [23] (Fig. 44.1a). macrometastases. It is hypothesized that poor angiogenesis
Emerging data indicate that pre-metastatic niche prepara- and/or active immunosurveillance results in tumor mass dor-
tion in bone depends on altering host immune responses mancy and micrometastases [29]. The steps leading to entry
[24]. Observations from a novel spontaneous murine model into and exit from dormancy in the bone are not fully charac-
of breast cancer (MMTV-NeuT transgenic) found transcrip- terized due to limited models to study the interactions that
tional reprogramming of the bone marrow hematopoietic cell take place in bone; however, new technologies and sophisti-
compartment prior to metastasis resulted in changes in the cated models have begun to reveal the key mechanisms
cellular composition and stromal architecture of the bone involved [30].
marrow. These changes included increases in Nestin,+ DTCs that arrest in bone are protected by specialized
CXCL12-expressing mesenchymal cells and CXCR4-­ niches: hematopoietic, endosteal, and vascular. These niches
expressing myeloid cells with noted decreases in erythroid can contribute to the regulation of dormancy and activation
and B lymphoid populations. Mechanistically, upregulation of arrested cells [31]. Dormancy in the hematopoietic and
of Atf3 in the bone marrow and epigenetic regulation by cir- endosteal niches is thought to exploit mechanisms that con-
culating miRNAs from the miR-29 and miR-23 families trol the quiescence of HSCs in the bone marrow. For instance,
were found to be responsible for the observed changes in this prostate tumor cells have been shown to mimic HSCs and
model [24]. bind to osteoblast-expressed annexin II. This in turn induces
Long-distance tumor–stroma communication and pre-­ the expression of AXL receptor tyrosine kinases. Subsequent
metastatic niche formation can be enhanced by the shedding engagement of AXL with osteoblast-expressed growth
of extracellular vesicles such as exosomes. Exosomes trans- arrest-specific 6 (GAS6) ligand induces dormancy [32].
port a variety of cargo, including functional proteins, mRNA, Heightened expression of AXL and dormancy has also been
and miRNA. Pioneering studies in melanoma demonstrated reported in multiple myeloma cells during in vitro coculture
that tumor-derived exosomes home to common secondary with osteoblasts and these dormant cells could be awakened
sites for melanoma metastasis, and once in the secondary by pharmacologically targeting AXL [33]. Additional mech-
site, horizontal transfer of the cargo could educate bone mar- anisms exist whereby prostate cancer cell contact with osteo-
row–derived cells to craft a pre-metastatic niche to support blasts activates Tank-binding kinase 1 (TBK-1) that leads to
the arrival of DTCs [25]. Prostate cancer-derived exosomes inhibition of mechanistic target of rapamycin (mTOR). This
promote pre-metastatic niche formation in the bone by trans- results in the arrest of protein synthesis and growth, drug
ferring pyruvate kinase M2 (PKM2) to bone marrow stromal resistance, and an increased stemness phenotype [34].
cells. This transfer upregulates CXCL12 via HIF1α and The most studied mechanism of dormancy to date involves
enhances prostate cancer seeding and growth in bone [26]. the balance between p38 mitogen-activated protein kinase
Prostate cancer-derived exosomes also stimulate osteoblast (MAPK) and extracellular signal-regulated kinase (ERK)
activity, possibly mediating pre-metastatic niche formation activity. A high p38/ERK ratio has been shown as a mediator
and osteoblastic pathology of prostate bone metastases [27]. of dormancy in many cancer types, including breast, pros-
Collectively, these data point to systemic changes occurring tate, melanoma, and fibrosarcoma. However, interactions
in the bone microenvironment prior to the arrival of cancer with microenvironment proteins such as collagen can stimu-
cells (Fig. 44.1b). late a switch to high ERK/p38 ratio and reverse dormancy
[35]. Bone marrow–derived TGF-β2 is also implicated in
maintaining dormancy via p38 activation, and inhibition of
44.3.4 Bone Niches and Activation/Dormancy either the TGF-β receptor, TGFβR1, or p38 led to awakening
and metastatic outgrowth [36]. Additional bone-derived fac-
The length of time from diagnosis to development of metas- tors such as Dkk3, BMP1, and growth differentiation factor
tasis is highly variable, ranging from months to decades [28]. 10 (GDF10) have also been shown to activate the p38/MAPK
This is attributed to DTCs entering states of dormancy in pathway and suppress the proliferation of various prostate
488 J. S. Frieling and C. C. Lynch

cancer cell lines both in vitro and in vivo [37]. Sophisticated sequestered factors such as TGF-β, VEGF, IGFs, and PDGF
in vitro techniques are emerging to study dormancy in bone, from the bone matrix, and these factors are potent stimulators
such as 3D bone-mimicking scaffolds. When cultured in a of tumor cell proliferation as well as osteoblast activities,
3D in vitro endosteal niche model comprised of endothelial, either directly or indirectly. The balance of specific factors
bone marrow stroma, and fetal osteoblasts, parental can dictate osteoblastic or osteolytic lesion pathology. For
MDA-MB-231 cells enter a dormant state in a p38 MAPK-­ example, tumor-derived factors such as interleukin 6 (IL-6)
dependent manner, strengthening the premise that p38 and IL-1β act predominantly on osteoclasts to promote an
MAPK is a major mediator of tumor dormancy in the bone osteolytic vicious cycle. Osteolytic bone disease is a common
microenvironment [38]. attribute of breast cancer and multiple myeloma and results
In addition to the endosteal niche, the bone marrow vascu- from tumor stimulation of osteoclast differentiation and activ-
lar niche is also supportive of dormancy. Dormant cells have ity rather than direct effect of cancer cells on the skeleton. In
been identified in perisinusoidal regions rich in E-selectin and addition to IL-6 and IL-1β, the osteolytic vicious cycle is
CXCL12 proteins, and blocking these proteins mobilized largely mediated by TGF-β, PTHrP, and RANKL. In contrast,
tumor cells [39]. Ghajar and colleagues have provided strong osteoblastic lesions tend to arise from the stimulation of
evidence for dormant DTCs in vascular niches of the bone osteoblasts mediated by growth factors such as BMPs, FGF,
where expression of thrombospondin-1 (TSP-1) by endothe- and Wnts. The factors that govern osteoclast and osteoblast
lial cells keeps DTCs in a dormant state. However, in regions biology are well reviewed here: [30, 41]. Osteoclast activity is
of vascular sprouting and angiogenesis, TSP-1 expression is likely required for bone metastasis even in osteoblastic dis-
lost as periostin and TGF-β expression increase thereby simu- ease. Increased concentrations of N-terminal telopeptides, a
lating dormancy exit [40]. These findings suggest that active molecular marker of bone turnover, have been detected in
angiogenesis can awaken cells from dormancy, whereas cells osteoblastic disease for instance [42]. Likewise, osteolytic
that settle near stable microvasculature are likely to be main- breast cancer metastases may feature pathological bone for-
tained in a quiescent state. Although derived from indepen- mation. Overall, however, there is still a substantial net loss of
dent studies, there is likely considerable overlap between bone in breast cancer skeletal metastasis whereas prostate
hematopoietic, endosteal, and vascular niches and their cancer metastasis results in a gain of poorly formed, woven
impact on tumor dormancy. Comprehensive characterization pathological bone. Bone-­ derived factors, such as TGF-β
of the spatial, temporal, and molecular niche dynamics released by osteoclasts in the vicious cycle, can also signifi-
remains a high priority to better understand the important and cantly contribute to other complications that impact patient
clinically significant problem of dormancy [30]. well-being, for example, muscle function. Osteolytic disease
often results in significant weakening of skeletal muscle
(cachexia), as has been observed in prostate, breast, lung, and
44.3.5 Invasion and Establishment multiple myeloma. Bone-derived TGF-β is a primary driver
of the Vicious Cycle of cancer-induced muscle weakness, upregulating expression
of NADPH oxidase 4 (Nox4) which leads to oxidation of the
The bone is abundant in growth factors such as TGF-β, insu- ryanodine receptor (RyR1) and calcium leakage [43].
lin growth factors (IGF), BMPs, platelet-derived growth fac- The vicious cycle continues to evolve as the molecular
tors (PDGF), fibroblast growth factors (FGF), vascular mechanisms that drive it are revealed. MMPs, for example,
endothelial growth factors (VEGF), and calcium (Ca2+). regulate the bioavailability of factors such as PTHrP, TGF-β,
Skeletal remodeling by osteoclasts releases many of these and RANKL [44–46]. MMPs can be produced by the tumor
factors into the microenvironment, and signals from the pri- cells themselves or secreted by other cell types within the
mary tumor can initiate the process. Metastatic tumor cells stromal compartment. Their processing of growth factors
exploit bone remodeling in order to promote survival and influences a host of other cells. Mesenchymal stem cells
growth. The resulting disruption of normal bone physiology (MSCs) comprise 0.001–0.01% of the bone marrow and
and decoupling of bone resorption and apposition has been respond to vicious cycle factors such as TGF-β, BMPs, and
described as the “vicious cycle” based on the cyclical inter- Wnts with potential to differentiate into osteoblasts, chon-
actions occurring between metastatic tumor cells and bone drocytes, and adipocytes [47]. The precise contributions of
stromal cells such as osteoblasts and osteoclasts [2] MSCs in bone metastasis are not fully understood, but recip-
(Fig. 44.1c). rocal interactions between tumor cells and MSCs appear to
In the most fundamental version of the vicious cycle, support the proliferation, differentiation, and survival of
tumor-derived factors stimulate the osteoblast compartment other stromal cells such as HPCs as well as metastatic tumor
to express RANKL that in turn promotes the fusion of osteo- cells, depending on disease and context. For example, MSCs
clast precursors into mature, bone-resorbing osteoclasts. recruited to prostate tumors via a CXCR6/CXCL16 axis pro-
Through the process of bone resorption, osteoclasts release mote growth and metastasis by secretion of CXCL12 [48].
44 Molecular Mechanisms of Metastasis to the Bone 489

MSCs can also suppress immune cells by secretion of factors macrophages (including osteal macrophages) in an immune-­
such as prostaglandin E2 (PGE2), IL-6, IL-10, and indole- competent mouse model inhibited prostate tumor-induced
amine 2,3 deoxygenase (IDO) [47]. Conversely, MSCs have bone formation [57]. Other myeloid lineage immune cells
been shown to have anti-tumorigenic effects as well, such as including neutrophils, megakaryocytes, platelets, and den-
downregulation of Wnt signaling by upregulating Dkk-1 dritic cells are also involved. For further review, see [58].
[49]. Further, secretion of BMP-2 by metastatic prostate can- The adaptive immune system is tightly linked to bone
cer cells can indirectly induce MSCs to differentiate into physiology and consequently is involved in bone metastasis.
osteoblasts and enhance osteoblast activity [50]. This could The percentage of T cells in bone is usually low compared to
contribute toward the osteoblastic phenotype characteristic other cell types. However, active CD4 and CD8 T cells have
of metastatic prostate cancer. MSC-derived adipocytes have been found in the bone marrow of breast cancer patients
also been shown to influence osteoblast and osteoclast activi- prior to treatment and were demonstrated to provide antitu-
ties in the vicious cycle with new roles emerging [51]. mor effects in bone metastases [58]. Patients with DTCs in
Increasing evidence has revealed vital contributions from bone present with increased numbers of activated NK cells
immune components in bone remodeling. The bone marrow and memory T cells in the bone marrow. The increase was
is involved in systemic immunity and therefore possesses a highest in patients with stage II disease and decreased in
unique immune environment that provides immune-­ stage III/IV, suggesting that immune responses may be sur-
privileged milieus for tumor cells. Naive immature myeloid mounted as bone metastatic disease progresses [59].
cells in the bone marrow can differentiate into macrophages, Interestingly, several studies have also reported roles for T
granulocytes, and neutrophils. In some instances, tumor-­ cells in stimulating osteoclast activity [6]. Conflicting reports
derived factors such as VEGF, stem cell factor (SCF), exist for the contributions of regulatory T cells (Tregs) in
granulocyte-­ macrophage colony-stimulating factor bone metastasis. Tregs produce RANKL, suggesting that
(GM-CSF), IL-6, cyclooxygenase-2 (COX2), and M-CSF they can promote osteoclast differentiation and as a conse-
expand a population of MDSCs, which possess the capacity quence, bone metastasis [58]. However, under certain condi-
to suppress both innate and adaptive immunity [52]. MDSCs tions, Tregs may inhibit osteoclast activity [60]. Increased
can also contribute to osteolytic lesion formation by serving numbers of Tregs have been detected in prostate cancer
as osteoclast precursors. For example, MDSCs derived from patients with bone metastasis, and the transfer of Tregs in a
the bone marrow of breast tumor-bearing mice have been xenograft mouse model of prostate cancer increased bone
shown to differentiate along an osteoclast lineage [53]. mineral density, whereas Treg depletion reduced bone den-
Increased numbers of MDSCs are found in bone metastases sity [61]. Interestingly, patients treated with bisphosphonates
compared to both primary breast tumors or lung metastases, that inhibit osteoclast activity have been shown to also
and this is regulated in part by interferon regulatory factor undergo an expansion and activation of γδ T cells. γδ T cells
(IRF)-7. Silencing IRF7 restricted immunosurveillance and possess functional characteristics of both innate and adaptive
promoted bone metastasis, whereas restoration of IRF7 immunity and can exert antitumor effects by releasing pro-­
expression decreased MDSCs while increasing CD8 T cells inflammatory cytokines. Their role in bone metastasis has
and NK cells, suggesting that certain patients may benefit not been thoroughly explored, but a European phase I clini-
from interferon (IFN)-based therapies [54]. MDSCs can also cal trial treating bone metastatic prostate cancer patients
indirectly promote tumor growth by activating tumor angio- with zoledronate plus interleukin-2 (IL-2) resulted in disease
genesis and lymphangiogenesis via secretion of VEGF, stabilization for six out of nine patients [58].
TNFα, and angiopoietin [55]. Macrophage infiltration is also Crosstalk between immune cells and other cells in the bone
observed in many cancer types and can contribute to metas- microenvironment is multidirectional. Osteoclast secretion of
tasis. Naive macrophages respond to microenvironmental chemokines such as CCL5, CXCL5, and CXCL1 recruits
cues to polarize into pro- or anti-inflammatory phenotypes, CD8+ T cells in transgenic OT-1 mice. The recruited T cells are
with the majority acquiring an anti-inflammatory phenotype stimulated by osteoclasts to proliferate and secrete IL-2, IL-6,
that is pro-tumorigenic. Tumor-derived factors like CCL2 and IFN-γ, however, the osteoclast-activated, FoxP3-expressing
promote macrophage recruitment to the bone which can cells were anergic which suppressed further priming of naïve
increase tumor growth and bone metastasis in vivo [56]. CD8+ T cells by dendritic cells [62]. Further, osteoclasts can
Bone is an abundant reservoir for osteal macrophages, which also induce CD4+ T regs or inflammatory CD4+ T cells
represent 20% of bone marrow cells. These cells can support depending on context, while osteoblast depletion led to an 80%
tumor growth and aid in immune escape as well as impact reduction of B cells in the bone specifically [63].
the bone microenvironment by stimulating osteoblast and In addition to changes in the tumor–bone microenviron-
osteoclast activity. For example, macrophages often associ- ment as discussed previously, bone metastases cause great
ate with formation of woven pathological bone in prostate pain that further contribute to patient morbidity. Many sensory
cancer bone metastasis specimens, and ablation of CD169+ nerve fibers innervate the bone and express acid-sensing ion
490 J. S. Frieling and C. C. Lynch

channels, including transient receptor potential cation channel osteoblastic lesions due to better uptake and incorporation
subfamily V member 1 (TRPV1), acid-sensing ion channel 1 into newly produced bone. Radionuclides offer better pain
(ASIC1), and ASIC3. These sensors are activated at pH 3–4 alleviation for diffuse pain resulting from multiple metastases
and thus respond to osteoclast-induced acidosis [64]. and are cleared quickly from nonskeletal sites. Radionuclides
Interestingly, nerve growth factor (NGF) blockade prevented emit either α or β radiation. β-emitters primarily provide pal-
increased sprouting of sensory nerve fibers and significantly liative pain relief, however, α-emitters also provide antitu-
reduced early and late-stage pain behaviors in murine models moral activity with minimal radiation damage to nearby
of prostate cancer, however, its beneficial effects did not healthy tissue due to their higher energy and soft tissue range
extend to tumor-induced bone remodeling [64, 65]. Innervation of less than 100 μm. In a study of men with bone metastatic
can also contribute to tumor invasion and metastasis via sig- prostate cancer previously treated with radiotherapy, the
naling of the autonomic nervous system. Prostate cancer xeno- α-emitter Radium-223 delayed time to first SRE and improved
grafts develop poorly in mice that are pretreated with chemical overall survival compared to placebo (14.0 months vs.
or surgical sympathectomy of the prostate gland or when stro- 11.2 months) leading to FDA approval for treatment of cas-
mal β2 and β3 adrenergic receptors were genetically deleted, tration-resistant bone metastatic prostate cancer without vis-
and tumor invasion and metastasis were reduced following ceral metastases [73, 74]. Based on molecular studies, new
blockade of type 1 muscarinic receptors. The findings of these agents are being generated and tested for the treatment of
studies suggest that sympathetic neonerves promote early bone metastases. For further review, see [75].
stages of tumorigenesis whereas parasympathetic nerves con-
tribute to dissemination and metastasis and offer unique
opportunities to develop therapeutic approaches that alleviate 44.4 Conclusion
pain and possibly limit disease progression [66].
Bone is a vital organ providing structure and support, but the
high metabolic and molecular activity combined with well-
44.3.6 T
 ackling the Vicious Cycle and vascularized nature and signals from the primary tumor
Cancer-­Induced Bone Disease appears to prepare a suitable soil for disseminated tumor
cells from numerous primary malignancies. Notably, depend-
Understanding the interactions from genetic and molecular ing on primary origin, the resulting metastatic lesions can
perspective can yield therapeutic opportunities. Antiresorptive have unique osteolytic, osteoblastic, or mixed pathologies,
agents such as bisphosphonates and anti-RANKL antibodies arising from the specific factors at play in a vicious cycle
remain a standard of care for patients with bone metastatic between tumor cells and bone. Recently, the importance of
disease. Antiresorptives protect skeletal integrity and allevi- the immune, nervous, and muscular systems has become evi-
ate pain by disrupting the vicious cycle, however, they have dent in this process as well, revealing the broad range of cel-
little impact on tumor growth and progression. lular players and processes governing bone metastasis. At
Bisphosphonates such as zoledronate bind hydroxyapatite present bone metastasis remains incurable and significantly
crystals in areas of active bone remodeling and are taken up impacts patient quality of life with pain, fracture, hypercal-
by bone-resorbing osteoclasts which induce their apoptosis cemia, and other skeletal complications. While molecular
[67]. Bisphosphonates increase time to first SRE, however, and genetic understanding has produced several leads for
no significant increases in overall survival have been demon- investigation, critical questions and areas remain to be
strated, and the development of new bone metastases and explored.
disease progression is seen in 30–50% of patients prescribed
bisphosphonates [68]. Although their intended target is
osteoclasts, bisphosphonates may also affect other cell types, Open Questions
such as γδ T cells, monocytes, MDSCs, endothelial cells, • How can we define the optimal therapeutic window
osteoblasts, and tumor cells [18, 69]. RANKL is a central to tackle dormancy?
mediator in bone turnover and metastasis. Denosumab is a • What are the pro- and anti-tumorigenic contribu-
fully humanized monoclonal antibody that binds RANKL in tions of an intact immune system in the bone micro-
order to prevent its binding to RANK on osteoclast precur- environment during skeletal colonization?
sors, thereby inhibiting osteoclast formation and activity • How does the bone microenvironment contribute to
[70]. Denosumab decreases skeletal complications and bone the evolution of resistant disease?
pain, but it did not increase overall survival in bone meta- • How do bone metastases regulate cachexia and
static breast cancer [71]. However, compared to zoledronic what strategies can be employed to minimize skel-
acid, denosumab significantly delayed the onset of SREs in etal muscle loss and improve patient quality of life?
bone metastatic prostate cancer patients [72]. • Are bone metastases the final destination or do bone
Bone-seeking radiopharmaceuticals have gained traction metastases seed tertiary metastatic sites?
recently for the treatment of bone metastasis, particularly for
44 Molecular Mechanisms of Metastasis to the Bone 491

25. Peinado H, et al. Melanoma exosomes educate bone marrow pro-

References genitor cells toward a pro-metastatic phenotype through MET. Nat
Med. 2012;18(6):883–91.
1. Ponzetti M, Rucci N. Switching homes: how cancer moves to bone. 26. Dai J, et al. Primary prostate cancer educates bone stroma through
Int J Mol Sci. 2020;21(11):4124. exosomal pyruvate kinase M2 to promote bone metastasis. J Exp
2. Mundy GR. Metastasis to bone: causes, consequences and thera- Med. 2019;216(12):2883–99.
peutic opportunities. Nat Rev Cancer. 2002;2(8):584–93. 27. Borel M, et al. Prostate cancer-derived exosomes promote osteoblast
3. Hvid I, et al. The biology of bone lengthening. J Child Orthop. differentiation and activity through phospholipase D2. Biochim
2016;10(6):487–92. Biophys Acta (BBA) - Mol Basis Dis. 2020;1866(12):165919.
4. Andersen TL, et al. A physical mechanism for coupling bone 28. Shiozawa Y, et al. Human prostate cancer metastases target the
resorption and formation in adult human bone. Am J Pathol. hematopoietic stem cell niche to establish footholds in mouse bone
2009;174(1):239–47. marrow. J Clin Invest. 2011;121(4):1298–312.
5. Manolagas SC. Birth and death of bone cells: basic regulatory 29. Aguirre-Ghiso JA. Models, mechanisms and clinical evidence for
mechanisms and implications for the pathogenesis and treatment of cancer dormancy. Nat Rev Cancer. 2007;7(11):834–46.
osteoporosis. Endocr Rev. 2000;21(2):115–37. 30. Mayhew V, Omokehinde T, Johnson RW. Tumor dormancy in bone.
6. Fornetti J, Welm AL, Stewart SA. Understanding the bone in cancer Cancer Rep. 2020;3(1):e1156.
metastasis. J Bone Miner Res. 2018;33(12):2099–113. 31. Braun S, et al. Cytokeratin-positive cells in the bone marrow and
7. Virchow R. Cellular pathology (1858). Thoemmes Continuum; survival of patients with stage I, II, or III breast cancer. N Engl J
2004. Med. 2000;342(8):525–33.
8. Paget G. Remarks on a case of alternate partial anaesthesia. Br Med 32. Shiozawa Y, et al. GAS6/AXL axis regulates prostate cancer
J. 1889;1(1462):1–3. invasion, proliferation, and survival in the bone marrow niche.
9. Neoplastic diseases: a treatise on tumours. By James Ewing, A.M., Neoplasia. 2010;12(2):116–27.
M.D., Sc.D., Professor of Pathology at Cornell University Medical 33. Khoo WH, et al. A niche-dependent myeloid transcriptome signa-
College, N.Y.; Pathologist to the Memorial Hospital. Third edition. ture defines dormant myeloma cells. Blood. 2019;134(1):30–43.
Royal 8vo. Pp. 1127, with 546 illustrations. 1928. Philadelphia and 34. Kim JK, et al. TBK1 regulates prostate cancer dormancy through
London: W. B. Saunders Co. Ltd. 63s. net. BJS (British Journal of mTOR inhibition. Neoplasia. 2013;15(9):1064–74.
Surgery). 1928;16(61):174–175. 35. Aguirre-Ghiso JA, et al. ERK(MAPK) activity as a determinant of
10. Krishnamurthy GT, et al. Distribution pattern of metastatic tumor growth and dormancy; regulation by p38(SAPK). Cancer
bone disease: a need for total body skeletal image. JAMA. Res. 2003;63(7):1684–95.
1977;237(23):2504–6. 36. Bragado P, et al. TGF-beta2 dictates disseminated tumour cell fate
11. Esposito M, Kang Y. Targeting tumor-stromal interactions in bone in target organs through TGF-beta-RIII and p38alpha/beta signal-
metastasis. Pharmacol Ther. 2014;141(2):222–33. ling. Nat Cell Biol. 2013;15(11):1351–61.
12. Kang Y, et al. A multigenic program mediating breast cancer metas- 37. Yu-Lee L-Y, et al. Bone secreted factors induce cellular quiescence
tasis to bone. Cancer Cell. 2003;3(6):537–49. in prostate cancer cells. Sci Rep. 2019;9(1):18635.
13. Arriaga JM, et al. A MYC and RAS co-activation signature in local- 38. McGrath J, et al. Identification of genes regulating breast cancer
ized prostate cancer drives bone metastasis and castration resis- dormancy in 3D bone endosteal niche cultures. Mol Cancer Res.
tance. Nature Cancer. 2020;1(11):1082–96. 2019;17(4):860–9.
14. Wang J, Loberg R, Taichman RS. The pivotal role of CXCL12 39. Price TT, et al. Dormant breast cancer micrometastases reside in
(SDF-1)/CXCR4 axis in bone metastasis. Cancer Metastasis Rev. specific bone marrow niches that regulate their transit to and from
2006;25(4):573–87. bone. Sci Transl Med. 2016;8(340):340ra73.
15. Lapteva N, et al. CXCR4 knockdown by small interfering RNA 40. Ghajar CM, et al. The perivascular niche regulates breast tumour
abrogates breast tumor growth in vivo. Cancer Gene Ther. dormancy. Nat Cell Biol. 2013;15(7):807–17.
2005;12(1):84–9. 41. Suva LJ, et al. Bone metastasis: mechanisms and therapeutic oppor-
16. Frieling JS, Basanta D, Lynch CC. Current and emerging thera- tunities. Nat Rev Endocrinol. 2011;7(4):208–18.
pies for bone metastatic castration-resistant prostate cancer. Cancer 42. Coleman RE, et al. Predictive value of bone resorption and
Control. 2015;22(1):109–20. formation markers in cancer patients with bone metastases
17. Schneider JG, Amend SR, Weilbaecher KN. Integrins and bone receiving the bisphosphonate zoledronic acid. J Clin Oncol.
metastasis: integrating tumor cell and stromal cell interactions. 2005;23(22):4925–35.
Bone. 2011;48(1):54–65. 43. Regan JN, et al. The role of TGFβ in bone-muscle crosstalk. Curr
18. Weilbaecher KN, Guise TA, McCauley LK. Cancer to bone: a fatal Osteoporos Rep. 2017;15(1):18–23.
attraction. Nat Rev Cancer. 2011;11(6):411–25. 44. Frieling JS, et al. Matrix metalloproteinase processing of PTHrP
19. Hüsemann Y, et al. Systemic spread is an early step in breast cancer. yields a selective regulator of osteogenesis, PTHrP1-17. Oncogene.
Cancer Cell. 2008;13(1):58–68. 2017;36:4498–507.
20. Kaplan RN, et al. VEGFR1-positive haematopoietic bone mar- 45. Thiolloy S, et al. An osteoblast-derived proteinase controls tumor
row progenitors initiate the pre-metastatic niche. Nature. cell survival via TGF-beta activation in the bone microenviron-
2005;438(7069):820–7. ment. PLoS One. 2012;7(1):e29862.
21. Li X, et al. A destructive cascade mediated by CCL2 facilitates 46. Lynch CC, et al. MMP-7 promotes prostate cancer-induced
prostate cancer growth in bone. Cancer Res. 2009;69(4):1685–92. osteolysis via the solubilization of RANKL. Cancer Cell.
22. VanHook AM. LOX induces a metastatic niche. Sci Signal. 2005;7(5):485–96.
2015;8(380):ec151. 47. Uccelli A, Moretta L, Pistoia V. Mesenchymal stem cells in health
23. Pang X, et al. Osteopontin as a multifaceted driver of bone metasta- and disease. Nat Rev Immunol. 2008;8(9):726–36.
sis and drug resistance. Pharmacol Res. 2019;144:235–44. 48. Jung Y, et al. Recruitment of mesenchymal stem cells into prostate
24. Chiodoni C, et al. Transcriptional profiles and stromal changes tumours promotes metastasis. Nat Commun. 2013;4:1795.
reveal bone marrow adaptation to early breast cancer in asso- 49. Qiao L, et al. Dkk-1 secreted by mesenchymal stem cells inhib-
ciation with deregulated circulating microRNAs. Cancer Res. its growth of breast cancer cells via depression of Wnt signalling.
2020;80(3):484–98. Cancer Lett. 2008;269(1):67–77.
492 J. S. Frieling and C. C. Lynch

50. Joseph J, et al. Disseminated prostate cancer cells can instruct 63. Zhu J, et al. Osteoblasts support B-lymphocyte commitment
hematopoietic stem and progenitor cells to regulate bone pheno- and differentiation from hematopoietic stem cells. Blood.
type. Mol Cancer Res. 2012;10(3):282–92. 2007;109(9):3706–12.
51. Cook LM, et al. Integrating new discoveries into the “vicious cycle” 64. Mantyh PW. Mechanisms that drive bone pain across the lifespan.
paradigm of prostate to bone metastases. Cancer Metastasis Rev. Br J Clin Pharmacol. 2019;85(6):1103–13.
2014;33(2–3):511–25. 65. Halvorson KG, et al. A blocking antibody to nerve growth factor
52. Gabrilovich DI, Nagaraj S. Myeloid-derived suppressor cells as reg- attenuates skeletal pain induced by prostate tumor cells growing in
ulators of the immune system. Nat Rev Immunol. 2009;9(3):162–74. bone. Cancer Res. 2005;65(20):9426–35.
53. Sawant A, et al. Myeloid-derived suppressor cells function as novel 66. Magnon C, et al. Autonomic nerve development contributes to
osteoclast progenitors enhancing bone loss in breast cancer. Cancer prostate cancer progression. Science. 2013;341(6142):1236361.
Res. 2013;73(2):672–82. 67. Rogers MJ, Watts DJ, Russell RG. Overview of bisphosphonates.
54. Bidwell BN, et al. Silencing of Irf7 pathways in breast cancer Cancer. 1997;80(8 Suppl):1652–60.
cells promotes bone metastasis through immune escape. Nat Med. 68. Saad F, et al. Long-term efficacy of zoledronic acid for the preven-
2012;18(8) tion of skeletal complications in patients with metastatic hormone-­
55. Murdoch C, et al. The role of myeloid cells in the promotion of refractory prostate cancer. J Natl Cancer Inst. 2004;96(11):879–82.
tumour angiogenesis. Nat Rev Cancer. 2008;8(8):618–31. 69. Bellido T, Plotkin LI. Novel actions of bisphosphonates in bone: pres-
56. Mizutani K, et al. The chemokine CCL2 increases prostate tumor ervation of osteoblast and osteocyte viability. Bone. 2011;49(1):50–5.
growth and bone metastasis through macrophage and osteoclast 70. www.cancer.gov, FDA Approval for Denosumab. National Cancer
recruitment. Neoplasia. 2009;11(11):1235–42. Institute; 2013.
57. Wu AC, et al. CD169(+) macrophages mediate pathological forma- 71. Stopeck AT, et al. Denosumab compared with zoledronic acid
tion of woven bone in skeletal lesions of prostate cancer. J Pathol. for the treatment of bone metastases in patients with advanced
2016;239(2):218–30. breast cancer: a randomized, double-blind study. J Clin Oncol.
58. Capietto A-H, Faccio R. Immune regulation of bone metastasis. 2010;28(35):5132–9.
BoneKEy Rep. 2014;3:600. 72. Fizazi K, et al. Denosumab versus zoledronic acid for treatment of
59. Feuerer M, et al. Enrichment of memory T cells and other profound bone metastases in men with castration-resistant prostate cancer: a
immunological changes in the bone marrow from untreated breast randomised, double-blind study. Lancet. 2011;377(9768):813–22.
cancer patients. Int J Cancer. 2001;92(1):96–105. 73. Choi JY. Treatment of bone metastasis with bone-targeting radio-
60. Buchwald ZS, et al. Osteoclast-induced Foxp3+ CD8 T-cells limit pharmaceuticals. Nucl Med Mol Imaging. 2018;52(3):200–7.
bone loss in mice. Bone. 2013;56(1):163–73. 74. Parker C, et al. Alpha emitter radium-223 and survival in metastatic
61. Zhao E, et al. Regulatory T cells in the bone marrow microen- prostate cancer. N Engl J Med. 2013;369(3):213–23.
vironment in patients with prostate cancer. Onco Targets Ther. 75. Clément-Demange L, Clézardin P. Emerging therapies in bone
2012;1(2):152–61. metastasis. Curr Opin Pharmacol. 2015;22:79–86.
62. Kiesel JR, Buchwald ZS, Aurora R. Cross-presentation by
osteoclasts induces FoxP3 in CD8+ T cells. J Immunol.
2009;182(9):5477–87.
 Surgical Management of Metastatic
Disease to the Upper Extremity 45
Allison C. Greene, Michael T. Torchia, Daniel C. Austin,
John-Erik Bell, and Eric R. Henderson

Abstract
Learning Objectives
Metastatic disease to bone is common and is best man-
aged by a multidisciplinary team of medical providers. • To understand general strategies for orthopedic sur-
Several interventional modalities exist for metastatic gical management of metastatic bone disease.
bone disease, including medical therapy, radiation, abla- • To understand patient- and disease-related consid-
tion, embolization, and surgery. The primary goals of erations that impact surgical planning.
these interventions are functional restoration, durable • To understand adjuvant management options that
local control, and pain relief. Orthopedic surgeons are may reduce the morbidity of surgery and/or improve
integral members of the treatment team. Operative strate- postsurgical outcomes.
gies for metastatic bone disease of the upper extremity are
unique due to the reduced weight-bearing requirements of
the arms compared to the legs. Appropriate selection of
an operative strategy requires the careful consideration of 45.1 Introduction
several characteristics of the bone and relevant lesion(s)
including: (1) is the metastasis isolated; (2) what is the Approximately 1.7 M new cases of cancer are diagnosed in the
anticipated response to treatment; (3) what is the quality United States each year, of which ~700,000 have a high pro-
of the available bone; (4) is there a fracture; (5) is the joint pensity for skeletal metastasis. Metastatic bone disease (MBD)
surface affected; (6) what are the functional requirements; robs patients of their independence, both by inducing substan-
(7) is embolization warranted? Prior to any interven- tial pain and leading to morbid bone destruction and fracture.
tion, local confirmation of metastatic disease should be Treatment goals for MBD include function restoration
obtained unless the patient is failing treatment. Operative and pain relief. When considering the best management of a
strategies include intralesional curettage, extramedul- new bone metastasis, providers must balance the individual
lary and intramedullary implants, conventional shoulder patient’s global needs and goals. While it is tempting for
arthroplasty, reverse shoulder arthroplasty, and segmen- the surgeon to venture to the operating room to improve a
tal implants. The best strategy will confer the most likely patient’s mobility and/or quality of life, surgery will require
attainment of goals with the least likelihood of morbidity. an interruption in the patient’s—usually—ongoing cancer
treatment. Decision-making about the surgical manage-
ment of MBD should, as with most oncological treatment,
A. C. Greene (*) be made by a multidisciplinary group including specialists
Albany Medical Center, Albany, NY, USA
e-mail: [email protected]
from medical oncology, radiation oncology, palliative care,
social work, and orthopedic surgery.
M. T. Torchia · J.-E. Bell · E. R. Henderson
Dartmouth Hitchcock Medical Center, Lebanon, NH, USA
Because the upper extremities are not primarily respon-
e-mail: [email protected]; sible for weightbearing and mobility, the reconstructive
[email protected]; [email protected] options available to surgeons are arguably more diverse than
D. C. Austin for lower extremity reconstructions, where the ability to bear
Dartmouth Hitchcock Medical Center, Lebanon, NH, USA the body’s weight is imperative.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 493
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_45
494 A. C. Greene et al.

This chapter is written from the perspective of an orthope- • What are the patient’s functional requirements? A patient
dic surgeon who is participating in the care of a patient with with compromised ability to fully bear weight on the legs
biopsy-proven metastatic bone disease, for which appropri- may require the use of their upper extremities for an assistive
ate staging and multidisciplinary treatment is underway. For device. Depending on the region of bone involved with meta-
a review of the workup of a new bone lesion or appropriate static disease, an increased requirement for weightbearing
staging protocols, please review the applicable, preceding may drive a surgeon toward arthroplasty.
chapters. • Is embolization warranted? Metastatic kidney and thy-
roid cancer lesions are often highly vascular. When inter-
vention is taken on axial or proximal appendicular lesions,
45.2 General Considerations preoperative embolization should be considered.

When considering operative management of a bone metas-

tasis, we recommend considering the following questions:
45.3 Nonsurgical Management
• Is it a metastasis and does it matter? While the odds
that a new, aggressive-appearing bone lesion in an adult Because of the lower weight-bearing requirement, nonopera-
patient is a metastasis are ~500:1, a metastatic process tive strategies for MBD are more tenable in the upper extrem-
should never be assumed unless the discovery of a new ities than the pelvis or lower extremities. In a treatment-­naive
primary bone cancer is believed to have no impact on the patient with a new diagnosis and the expectation of an excel-
patient’s survival whatsoever, for example, in the setting lent response, nonoperative management of upper extremity
of a hospice-bound patient with a new pathological frac- lesions, particularly in the absence of fracture, is reasonable.
ture. In any other situation, tissue confirmation of the
lesion as a metastasis is warranted.
• Is the metastasis isolated? The data supporting en bloc 45.3.1 Immobilization
resection of isolated bone metastases are not robust, how-
ever, this strategy is supported by the National Comprehensive In the setting of pathological fracture, casting and splinting
Cancer Network (NCCN) for solitary or oligometastatic dis- produce poor outcomes [3]. Immobilization of impending or
ease in the setting of renal cell carcinoma [1]. Before under- existing pathologic fractures is feasible in patients for whom
taking a substantial en bloc resection, ensure that staging surgery is contraindicated [4]. Additionally, patients without
scans are current to avoid unnecessary risk. an established diagnosis may require a immobilization prior
• What is the expected response to antineoplastic thera- to definitive intervention [5].
pies? Newly diagnosed patients who are expected to have
an excellent response to treatment may have years or
decades of life ahead of them with good potential for bone 45.3.2 Radiotherapy
reconstitution and fracture healing. In this setting, upper
extremity lesions are highly amenable to nonoperative Radiotherapy (RT) is employed for most skeletal metastases.
management with surveillance and surgical strategies in RT can decrease pain, maintain and restore function, reduce
these patients may rely more on bone reinforcement. the risk of pathologic fracture, and improve quality of life
Conversely, patients who are failing second-, third-, or [6]. Local tumor control is also possible, although highly
last-­line treatment have poor healing potential and may dependent on histological origin [7].
benefit from replacement strategies.
• What is the quality of the remaining bone stock? This 45.3.2.1 Photon-Beam Radiation Therapy
question must be weighed against the expected treatment Most cancer radiotherapy utilizes high-energy photons
response—see above—but in general greater degrees of (X-ray or gamma-ray). Ionizing radiation produces reactive
bone loss indicate a heavier reliance on replacement oxygen species in the targeted tissues, inducing DNA dam-
strategies. age in both malignant and healthy tissue. By their nature,
• Is there a fracture? Pathological fracture healing is cancer cells are less capable of DNA repair and therefore
always fraught; however, certain primary cancers portend this insult leads to cell death. Cellular damage is dependent
worse healing [2]. As above, anticipated treatment on dose; therefore, treatment strategies seek to maximize
response, and fracture healing, should be considered tumor dose and minimize adjacent tissue exposure. There is
when choosing a reconstruction strategy. no consensus regarding standardization of fractioning [8];
• Is a joint surface compromised? Deformation of the therefore, the trend has been to concentrate dose and reduce
joint surface usually mandates an arthroplasty procedure. fractions, resulting in improved treatment completion.
45 Surgical Management of Metastatic Disease to the Upper Extremity 495

45.3.2.2 Particle-Beam Radiation Therapy 45.3.3.3 Embolization

Heavy ions (protons, carbon ions) are also used for RT Metastatic tumors are often hypervascular [16]. Transarterial
with some advantage over photons. Whereas photon dose embolization (TAE) has been investigated as both a preop-
decreases exponentially upon entering tissue, heavy ions erative intervention to limit hemorrhage and as a treatment
deposit most energy toward the end of their trajectory, for local tumor control. As a hemorrhage control modality,
termed the Bragg peak [8]. Adjusting initial particle speed TAE has greatest utility for highly vascular metastases, spine
and beam intensity allows for precise targeting of a tumor metastases, and in cases requiring prolonged surgical time
while sparing surrounding tissues, conferring increased rela- [17]. Microcatheters are the preferred conduit for embolic
tive biological effectiveness (RBE) compared to photons agents, which include proprietary engineered particles, gel-
(approximately 3x higher) [8]. foam, liquid agents, and stainless steel or platinum coils
[17]. A recent systematic review reported that 50% of studies
(Level III) investigating TAE for metastatic renal cell carci-
45.3.3 Interventional Radiology noma in long bones (primarily femur and humerus) demon-
strate a significant reduction in perioperative blood loss [18].
In situations where radiotherapy and surgery are contrain- The evidence to support TAE for local tumor control is
dicated, percutaneous thermal ablation techniques may be inconclusive and high-quality randomized controlled trials
used to manage symptoms and achieve local tumor control. are necessary to ascertain its effectiveness [18].
Thermal ablation is a minimally invasive technique in which
an applicator is inserted directly into a tumor site under
image guidance in order to deliver temperatures (low or 45.4 Surgical Management
high) sufficient to induce tissue necrosis [9, 10].
Surgical management of upper extremity skeletal metastases
45.3.3.1 Radiofrequency Ablation should restore function and decrease pain [5]. Preoperative
Radiofrequency ablation (RFA) of musculoskeletal lesions planning is paramount to obtaining the desired outcome.
was first reported in 1992 and continues to be used widely Patients with skeletal metastases introduce particular con-
[11]. RFA uses high-frequency (375–500 kHz) alternat- siderations relevant to preoperative planning, specifically
ing current delivered via 15- to 17-gauge electrodes into a osteopenic bone with limited healing potential [2], which
tumor, inducing local ionic agitation, rapid heat formation therefore places increased loads on the construct. More
(>60 °C), resulting in protein denaturation and local hemo- fundamentally, an accurate diagnosis of the skeletal lesion,
stasis. Cooled saline is circulated periodically at the site which is predicated on a comprehensive workup, is critical to
to reduce charring, which can inhibit electrical conduction avoiding inadvertent fixation of primary osseous malignan-
[10]. Early RFA systems utilize monopolar electrodes with cies [19, 20].
a grounding pad applied to the patient; simultaneous use Within the upper extremity, the humerus is the most
of three monopolar electrodes has been described [9, 10]. common site of metastatic skeletal lesions [21], therefore,
Newer bipolar systems create more precise, larger ablation ­outcomes studies of humeral lesions are represented dis-
zones with faster heat production and shorter treatment proportionally in the literature. As mentioned previously,
duration [12]. because it is less responsible for weightbearing, there is a
broader range of acceptable operative strategies for MBD of
45.3.3.2 Cryoablation the upper extremity.
In contrast to RFA, cryoablation is a thermal technique
that relies on temperatures between −20 to −40 °C. A hol-
low cryoprobe with an uninsulated tip is inserted into the 45.4.1 Extra-articular, Intralesional Strategies
tumor. Pressurized argon and helium gases are delivered
at the tip and the resulting gas expansion produces rapid MBD lesions that do not involve an articular surface and
cooling. A single probe may treat a zone 5 cm in length and have reasonable bone stock are generally amenable to intra-
3.5 cm in diameter; multiple probes may be used simulta- lesional operative strategies. These include bone filler-only
neously to expand the treatment zone. Effective treatment options, extramedullary—plate-screw—implants, and intra-
requires >1 min of tissue freezing [10]. Treatment areas medullary implants (Fig. 45.1).
can be monitored on CT and MRI [10]. Cryoablation elic-
its effective MBD-related pain control [10, 13] and has a 45.4.1.1 Bone Filler-Only
low adverse event rate [14]. Cryoablation achieved local For limited indications, a bone filler—most commonly
tumor control in 82% of patients with metastatic renal cell polymethylmethacrylate (PMMA)—may be used without
carcinoma [15]. a metallic implant. PMMA offers strong resistance to com-
496 A. C. Greene et al.

a b e f g

c d

Fig. 45.1 Examples of various strategies of intralesional management lary fixation; (e, f) open reduction and internal fixation of a humerus
of metastatic bone disease: (a, b) curettage and polymethylmethacry- affected by metastatic endometrial cancer, showing bone consolidation
late cement packing alone without trauma implants; (c, d) intramedul- following chemotherapy and radiation management (g)

pression but does not provide much stability for bending improved at 1-year follow-up, there were 52 immediate
moments. Appropriate indications for use of PMMA only postoperative complications. Similarly, a 2011 review of
therefore are limited to subchondral or epiphyseal–metaphy- 63 metastatic bone tumors to the humerus stabilized with
seal locations. plate fixation and cement augmentation found that while all
patients reported pain relief and 50/63 (79%) regained inde-
45.4.1.2 Extramedullary Instrumentation pendence in activities of daily living, there were 14 compli-
The advent of locked plating has improved the stability of cations with seven requiring reoperation [31].
plate-screw constructs in osteopenic, and otherwise compro-
mised, bone [22]. Stability can be augmented with the use 45.4.1.3 Intramedullary Instrumentation
of PMMA, a strategy that has been used for several decades Intramedullary nails (IMNs) are load-sharing devices [32,
[23, 24]. PMMA may be used to augment screw pullout 33]. IMNs are designed to bear weight initially, with gradual
strength [25–27] and used as a delivery vehicle for antibiot- transfer to the healing bone. In the setting of poor bone qual-
ics and antitumoral drugs [28, 29]. ity and limited healing potential characteristic of skeletal
Advantages to plate-screw constructs include generally metastases, this transmission of load sharing may not occur,
greater stability than intramedullary options. Furthermore, sometimes resulting in implant failure.
because the humeral intramedullary space ends several cen- The humerus is the only upper extremity bone readily
timeters proximal to the elbow, plate-screw constructs are amenable to IMN. Use of IMNs in the humerus is also lim-
more appropriate when distal humeral fixation is necessary. ited by lesion location. Lesions suitable for IMNs are limited
Another advantage to open management of MBD is the abil- generally to well-contained proximal metadiaphyseal lesions
ity to debulk the lesion through curettage and oversee directly and mid-diaphyseal lesions [34, 35].
the introduction of PMMA. Plate-screw constructs are the Outcomes following IMN for prophylactic and fracture
preferred strategy for operative lesions of the forearm. fixation of humeral MBD are generally good, with reliable
Disadvantages to plate-screw constructs include more and effective pain relief with function restoration and low
extensive exposure and dissection and higher adverse event morbidity [36–38]. A comparison of IMN and IMN with
rates. A 2009 study prospectively evaluated 187 patients who PMMA for pathologic humerus fractures showed reduced
underwent locked plating of a pathologic proximal humerus pain 6 weeks following surgery in the group receiving
fracture [30]. Although clinical outcome scores significantly PMMA [39].
45 Surgical Management of Metastatic Disease to the Upper Extremity 497

In the absence of a clear contraindication, RT is generally et al. reported outcomes for 10 patients who underwent
recommended following intralesional surgical management RTSA and 37 with hemiarthroplasty with an average follow-
[40, 41]. ­up of 27 months. Postoperative motion was significantly bet-
ter with RTSA than hemiarthroplasty, with forward flexion
averages of 85° and 28°, respectively. Postoperatively, 42%
45.4.2 R
 econstruction: Proximal Humerus of hemiarthroplasty patients experienced an adverse event
Lesions compared to 10% of RTSA patients. Instability occurred
in 16% of patients with hemiarthroplasty and none of the
The proximal one-third of the humerus is the most common patients in the RTSA cohort [54].
site for metastatic disease in the upper extremity [40]. The
options for reconstruction of the proximal humerus depend
on the volume of diseased humeral bone, the location of 45.4.4 Addressing Bone Loss
the lesion, functional and oncological goals for resection,
and the age of the patient. Relatively localized subchondral Less conventional implants are required to reestablish
lesions may be treated with conventional arthroplasty tech- anatomical relationships when proceeding with shoulder
niques, while more extensive or distal lesions will require arthroplasty in the setting of segmental bone insufficiency.
more extensive resection and segmental reconstruction. Proximal bone loss may be addressed with either an APC
Several implant options are available to reconstruct seg- or an endoprosthesis, either of which can then be mated to
mental defects, including primarily metallic endoprostheses a hemiarthroplasty or RTSA articulation. The endoprosthe-
(EP), allograft-prosthetic composites (APCs), and osteoar- sis option provides what some consider to be a faster and
ticular allografts [42–44]. The duration following surgery more straightforward reconstruction [55] where the miss-
to achieve a good palliative and functional result should be ing bone is simply replaced with metal. Endoprostheses
factored into implant selection due to the often short survival present challenges including soft-tissue reattachment to the
of patients with MBD [45]. implant to provide stability and function, although creative
solutions such as wrapping the implant in a vascular mesh
have been proposed [56]. An allograft-prosthetic composite
45.4.3 A
 rthroplasty: Hemiarthroplasty versus (APC) is a construct in which a portion of the allograft bone
Reverse Total Shoulder Arthroplasty is cut to the appropriate length and then mated with either
a hemiarthroplasty or RTSA prosthesis embedded within it
Conventional hemiarthroplasty has been the traditional strat- [57–59]. Advantages of this include the ability to customize
egy for patients with MBD of the humeral head. This proce- the implant to any patient and to provide soft-tissue attach-
dure is effective for providing patients with pain relief but ments for repairing the native tendons, increasing stability
may result in limited motion [46, 47]. Challenges to postop- and functionality [60, 61]. Osteoarticular allografts provide
erative function arise commonly in the setting of MBD due an additional option for reconstruction of large defects and
to diseased bone at the footprint of the rotator cuff, creating utilize the osseous and articular portions of donor bone to
challenges to RTC repair with a traditional hemiarthroplasty recreate the missing anatomy. This allograft is then mated
implant [48] and poor outcomes when the rotator cuff is defi- to the distal host bone with an expectation of osteosynthe-
cient [49]. In a series of 36 patients reconstructed with APC sis over time [62, 63]. Advantages of this technique include
hemiarthroplasty for a variety of tumor-related indications, avoiding a metallic implant in younger patients as well as
the postoperative active flexion was only 56° overall [50]. A providing allograft sites for reattachment of native tendons.
similar study of 47 patients with endoprosthetic hemiarthro- Risks include fracture and infection. While these allografts
plasties reported average active flexion of 55° [46]. have been reported to provide excellent or good outcomes
Reverse total shoulder arthroplasty (RTSA) provides in two-­thirds of patients [62], they are marked by substantial
an alternative to hemiarthroplasty [49, 51]. This design is limitations in movement and lifting at the glenohumeral joint
effective for cuff-deficient patients, with high relevance to [63]. Despite this, they do provide pain relief and maintain
patients with MBD. Outcome reports following RTSA for hand functionality [63].
patients with MBD demonstrate active forward flexion of
163° and an average Constant Shoulder Scores of 76, values
similar to those reported for RTSA in the absence of MBD 45.4.5 Outcomes and Complications
[52]. Another series reports more modest outcomes with a
mean Constant score of 52 and flexion of 122° [53]. Shoulder function after tumor resection and reconstruction has
There are few direct comparisons of outcomes follow- traditionally been limited with a goal of achieving a stable, pain-
ing hemiarthroplasty and RTSA for tumor resection. Grosel free joint which allows for effective hand function. While the
498 A. C. Greene et al.

RTSA has improved range of motion at the shoulder itself, the is remote from both the proximal and distal joint surfaces.
overall goals of pain relief and stability remain important crite- After tumor removal, the resulting void can be bridged with
ria. When comparing specific bone loss constructs, a large-scale allograft, autograft, or an endoprosthesis. A recent case
retrospective cohort study and systematic review both indicate series of intercalary endoprostheses highlighted the reli-
that APCs, endoprostheses, and osteoarticular allografts provide ability of this technique in the humerus as no non-oncologic
similar overall functionality and pain relief [42, 64]. Importantly, failures were observed and patients reported satisfactory
these outcomes do not necessarily capture the various complica- functional outcomes scores [67]. Allograft reconstruction
tions that a patient may experience. Postoperative complications presents a biological alternative for filling diaphyseal voids
following tumor resection and reconstruction are common and and may be more appropriate for younger patients in whom
important to consider [45, 48, 54]. an endoprosthesis may not be a long-term solution [68]. In a
Nonunion at the interface between the host and donor series of 30 patients undergoing humeral reconstruction with
bone can cause construct failure following either an APC or allograft, the overall complication rate was 53%, most often
osteoarticular allograft. Rates of nonunion following APC due to a nonunion at the host–graft interface [68]. While this
reconstructions vary greatly in the literature from 0 to 26% technique avoids implantation of a prosthesis and can prove
[58, 59], highlighting the impact of differing surgical tech- durable if graft incorporation occurs, the overall success can
niques or patient populations between studies. Osteoarticular be limited. A final option for a diaphyseal void is a vascular-
allograft reconstructions can also develop nonunion; a sys- ized free-fibula transfer which may provide increased healing
tematic review highlighted that this complication was more potential for the patient. However, the literature highlights
common with osteoarticular allografts (11%) than APCs that these can also be associated with graft failure and frac-
(5.0%). When considering either the APC or osteoarticular ture, with a small series demonstrating a 75% reoperation
allograft techniques that rely on osteosynthesis, it is impor- rate for fracture, but ultimately a 100% union rate [69].
tant to recognize that chemotherapy can be a risk factor for
nonunion [60, 65] and may make an endoprosthesis a more
attractive option. Postoperative stability of the glenohumeral 45.4.7 Distal Humerus Reconstruction
joint is an important goal of any reconstruction as disloca-
tion is a common reason for reoperation [48, 54]. Theoretical The need for tumor resection and reconstruction of the distal
benefits of an APC or osteoarticular allograft include more humerus is rarer than the proximal humerus, but the options
attachment sites for native soft tissues to help provide stabil- for reconstruction are similar and include primary total elbow
ity, however, large-scale studies have not detected meaning- prostheses, endoprostheses, allograft–prosthetic composites,
ful differences between groups [42, 64]. RTSA provides a and osteoarticular allografts, with similar risks and benefits for
new, more constrained arthroplasty option, and preliminary each approach. Osteoarticular allografts again provide young
studies suggest a lower rate of instability when using RTSA patients with an alternative to a prosthesis but are marked with
versus hemiarthroplasty following tumor resection [54]. high complication rates including fractures and joint instability
Additional complications to consider include aseptic loos- [68]. When the area of resection is relatively limited, or when
ening of an implant and fracture adjacent to, or within, the combined with an allograft–prosthesis composite construct,
construct. Varying rates of these complications are presented in primary total elbow replacement components have been asso-
heterogeneous small case studies throughout the literature [45, ciated with good pain relief and function with a 20% revision
48, 50, 59] and it is necessary to look to multicenter studies rate [70]. More extensive reconstruction can be addressed with
[42] or systematic reviews [64] to observe meaningful trends. segmental ­endoprostheses, which have been found to result in
A large-scale study demonstrated that endoprosthesis and APC an acceptable 107° arc of motion and reasonable clinical out-
were generally equivalent in terms of the risk of loosening and comes, but can be associated with aseptic loosening requiring
fracture but that osteoarticular allografts were clearly inferior revision [71] (Figs. 45.2 and 45.3).
with a fracture rate of up to 49% [42, 64]. Poor long-term out-
comes with osteoarticular allografts in another recent study led
the authors to conclude that their use should be avoided [66]. 45.4.8 Amputation

Amputation is rarely performed for malignant lesions of the upper

45.4.6 I ntercalary Reconstruction: Diaphyseal extremity and is particularly uncommon for metastatic disease.
and Distal Humerus Lesions Amputation may be utilized for tumors presenting with exten-
sive neurovascular infiltration, ulceration or fungating growth,
In situations where a metastatic humerus lesion is limited or infection. Amputation may also be indicated for patients
to the diaphyseal region of the bone, resection and interca- with loss of limb function or unremitting pain. Amputations are
lary reconstruction can be an excellent option if the location characterized by level of limb involvement and include the fol-
45 Surgical Management of Metastatic Disease to the Upper Extremity 499

a b c d

Fig. 45.2 Examples of two strategies of wide-local excision of renal bicolumnar plating. (c) Showing a mid-diaphyseal humerus fracture
cell metastases of the humerus: (a) showing a distal diaphyseal humerus through a renal cell metastasis, (d) treated operatively with a cemented,
fracture through a renal cell metastasis, (b) treated operatively with intercalary implant

lowing: above-shoulder (forequarter), shoulder disarticulation, Acrometastases are exceedingly rare, accounting for
transhumeral, through-elbow, below-elbow, wrist disarticulation, just 0.1% of bone metastases [76]. These lesions are most
and at the level of the hand or digits. Resections adjacent to the commonly seen in patients with primary renal cell disease
elbow and wrist are uncommon and are not discussed here. and typically portend a poor prognosis. Surgical ablative
Forequarter (interscapulothoracic) was first described techniques for metastatic disease of the hand and dig-
in the management of major traumatic injuries of the arm its have been reported in the literature. Puhaindran et al.
and involves surgical resection of the entire upper extrem- report good functional and cosmetic outcomes following
ity, shoulder girdle, scapula, and a portion of the clavicle single ray resection for malignant tumors of the hand [77].
[72]. The technique was first utilized in the management of Not unexpectedly, double ray resection results in worse
malignancy in 1836 and has traditionally been used in the functional outcomes, though remains a feasible hand-spar-
treatment of high-grade primary bone sarcomas of the proxi- ing option [78].
mal humerus. Forequarter amputation is also indicated for
the management of unresectable metastatic lesions arising
from the scapula or proximal humerus. Such tumors have 45.5 Summary
failed to respond to adjuvant chemotherapy and/or radiation,
are locally aggressive, and lead to debilitating symptoms for MBD is common and may result in substantial patient
patients [73]. Individuals undergoing forequarter amputa- morbidity through the destruction of large portions of the
tion have a poor overall prognosis, and time to death follow- skeleton, causing function and disabling pain. Treatment
ing this procedure is short (3–24 months), leading some to plans should align with patient goals and be made in
advocate against radical amputation [74]. Merimsky et al., the setting of a multidisciplinary team, with attention to
however, demonstrate an improvement in performance status the individual patient’s global health status and ongo-
and quality of life in patients who underwent palliative fore- ing treatment. Weightbearing and mobility requirements
quarter amputation [74]. It is important to bear in mind the are less of a limitation in the upper extremities, thereby
potential psychological implications of this procedure [75]. expanding the treatment strategies available to patients
Forequarter amputations are disfiguring, and upper extremity with MBD of the upper extremities. In the setting of frac-
prostheses often do not provide favorable functionality when ture and/or treatment failure, surgical reconstruction is
compared to lower extremity prostheses. generally indicated.
500 A. C. Greene et al.

a b

c d

Fig. 45.3 Examples of two strategies of endoprosthesis used in the humeral replacement. (c) Showing extensive humeral head destruction
humerus for metastatic bone disease: (a) showing failed intralesional from metastatic lung cancer, (d) treated with a segmental endoprosthe-
management of a renal cell metastasis, (b) later converted to a distal sis with reverse total shoulder replacement
45 Surgical Management of Metastatic Disease to the Upper Extremity 501

Questions 5. When reconstructing large metaphyseal bone

defects following metastatic disease resection, mul-
1. What is the most common upper extremity location
tiple reconstruction options are available. Although
of skeletal metastatic disease?
studies have shown that functional outcomes are
A. Radius
similar between options, how do these compare in
B. Scaphoid
terms of complication rates?
C. Humerus
A. Endoprostheses, allograft–prosthetic compos-
D. Ulna
ites, and osteoarticular allografts are all similar
2. An appropriate indication for using polymethyl-
B. Endoprostheses are marked by higher compli-
methacrylate (PMMA) bone filler would be a meta-
cation rates
static lesion located in the:
C. Allograft–prosthetic composites are marked by
A. Diaphysis
higher complication rates
B. Meta-diaphyseal region
D. Osteoarticular allografts are marked by
C. Subchondral surface
higher complication rates
D. Articular surface
E. This question has not been previously studied
3. Which location of a metastatic lesion would be
in the literature
most amenable to intramedullary instrumentation?
A. Distal humerus
B. Proximal humerus
C. Diaphysis of the radius
D. Diaphysis of the humerus Future Directions
4. When the resection of a metastatic lesion from the
proximal humerus requires sacrificing the humeral • How will innovations in gene therapy change the
head and tuberosities, which reconstruction option surgical management of metastatic disease of bone
will provide the patient with the best range of in the upper extremity?
motion postoperatively? • How will enhanced understanding of metaphysical
A. Total shoulder arthroplasty wellness practices help patients navigate the physi-
B. Girdlestone cal limitations created by limb reconstruction
C. Reverse total shoulder arthroplasty procedures?
D. Hemiarthroplasty
E. Range of motion is similar between these
options

tases. Clin Oncol (R Coll Radiol). 2006;18:747–60. https://doi.

45.6 References org/10.1016/j.clon.2006.09.007.
8. Trikalinos TA, Terasawa T, Ip S, et al. Particle beam radiotherapies
1. National Comprehensive Cancer Network (NCCN). Clinical prac- for cancer. Agency for Healthcare Research and Quality (US); 2009.
tice guidelines; n.d. 9. Ahmed M, Solbiati L, Brace CL, et al. Image-guided tumor ablation:
2. Gainor BJ, Buchert P. Fracture healing in metastatic bone disease. standardization of terminology and reporting criteria—a 10-year
Clin Orthop Relat Res. 1983;178:297–302. update. Radiology. 2014;273:241–60. https://doi.org/10.1148/
3. Flemming JE, Beals RK. Pathologic fracture of the humerus. Clin radiol.14132958.
Orthop Relat Res. 1986:258–60. 10. Moynagh MR, Kurup AN, Callstrom MR. Thermal ablation of bone
4. Healey JH, Brown HK. Complications of bone metasta- metastases. Semin Intervent Radiol. 2018;35:299–308. https://doi.
ses. Cancer. 2000;88:2940–51. https://doi.org/10.1002/1097-­ org/10.1055/s-­0038-­1673422.
0142(20000615)88:12+<2940::AID-­CNCR10>3.0.CO;2-­W. 11. Volkmer D, Sichlau M, Rapp TB. The use of radiofrequency abla-
5. Anract P, Biau D, Boudou-Rouquette P. Metastatic fractures of long tion in the treatment of musculoskeletal tumors. J Am Acad Orthop
limb bones. Orthop Traumatol Surg Res. 2017;103:S41–51. https:// Surg. 2009;17:737–43.
doi.org/10.1016/j.otsr.2016.11.001. 12. Chang W, Lee JM, Lee DH, et al. Comparison of switching bipo-
6. De Felice F, Piccioli A, Musio D, Tombolini V. The role of lar ablation with multiple cooled wet electrodes and switching
radiation therapy in bone metastases management. Oncotarget. monopolar ablation with separable clustered electrode in treatment
2017;8:25691–9. https://doi.org/10.18632/oncotarget.14823. of small hepatocellular carcinoma: a randomized controlled trial.
7. Agarawal JP, Swangsilpa T, van der Linden Y, et al. The role of PLoS One. 2018;13:e0192173. https://doi.org/10.1371/journal.
external beam radiotherapy in the management of bone metas- pone.0192173.
502 A. C. Greene et al.

13. Susa M, Kikuta K, Nakayama R, et al. CT guided cryoablation for 32. Bong MR, Kummer FJ, Koval KJ, Egol KA. Intramedullary
locally recurrent or metastatic bone and soft tissue tumor: initial nailing of the lower extremity: biomechanics and b­iology.
experience. BMC Cancer. 2016;16:798. https://doi.org/10.1186/ J Am Acad Orthop Surg. 2007;15:97–106. https://doi.
s12885-­016-­2852-­6. org/10.5435/00124635-­200702000-­00004.
14. Auloge P, Cazzato RL, Rousseau C, et al. Complications of percu- 33. Cheung G, Zalzal P, Bhandari M, et al. Finite element analysis
taneous bone tumor cryoablation: a 10-year experience. Radiology. of a femoral retrograde intramedullary nail subject to gait load-
2019;291:521–8. https://doi.org/10.1148/radiol.2019181262. ing. Med Eng Phys. 2004;26:93–108. https://doi.org/10.1016/j.
15. Gardner CS, Ensor JE, Ahrar K, et al. Cryoablation of bone metas- medengphy.2003.10.006.
tases from renal cell carcinoma for local tumor control. J Bone Joint 34. Redmond BJ, Biermann JS, Blasier RB. Interlocking intra-
Surg. 2017;99:1916–26. https://doi.org/10.2106/JBJS.16.01182. medullary nailing of pathological fractures of the shaft of the
16. Kickuth R, Waldherr C, Hoppe H, et al. Interventional management humerus. J Bone Joint Surg Am. 1996;78:891–6. https://doi.
of hypervascular osseous metastasis: role of embolotherapy before org/10.2106/00004623-­199606000-­00012.
orthopedic tumor resection and bone stabilization. Am J Roentgenol. 35. Scolaro J, Lackman R. Surgical management of metastatic long
2008;191:W240–7. https://doi.org/10.2214/AJR.07.4037. bone fractures: principles and techniques. J Am Acad Orthop Surg.
17. Owen RJT. Embolization of musculoskeletal bone 2014; https://doi.org/10.5435/JAAOS-­22-­02-­90.
tumors. Semin Interv Radiol. 2010;27:111. https://doi. 36. Alvi HM, Damron TA. Prophylactic stabilization for bone metasta-
org/10.1055/s-­0030-­1253510. ses, myeloma, or lymphoma: do we need to protect the entire bone?
18. Geraets SEW, Bos PK, van der Stok J. Preoperative emboliza- Clin Orthop Relat Res. 2013;471:706–14. https://doi.org/10.1007/
tion in surgical treatment of long bone metastasis: a system- s11999-­012-­2656-­1.
atic literature review. EFORT Open Rev. 2020;5:17. https://doi. 37. Pretell J, Rodriguez J, Blanco D, et al. Treatment of pathological
org/10.1302/2058-­5241.5.190013. humeral shaft fractures with intramedullary nailing. A retrospec-
19. Adams SC, Potter BK, Mahmood Z, et al. Consequences and tive study. Int Orthop. 2010;34:559–63. https://doi.org/10.1007/
prevention of inadvertent internal fixation of primary osseous s00264-­009-­0833-­y.
sarcomas. Clin Orthop Relat Res. 2009;467:519–25. https://doi. 38. Spencer SJ, Holt G, Clarke JV, et al. Locked intramedullary nailing
org/10.1007/s11999-­008-­0546-­3. of symptomatic metastases in the humerus. J Bone Joint Surg Br.
20. Spence G, Dunning MT, Cannon SR, Briggs TWR. The hazard of 2010;92:142–5. https://doi.org/10.1302/0301-­620X.92B1.22399.
retrograde nailing in pathological fractures. Three cases involving 39. Laitinen M, Nieminen J, Pakarinen T-K. Treatment of pathologi-
primary musculoskeletal malignancy. Injury. 2002;33(6):533–8. cal humerus shaft fractures with intramedullary nails with or with-
https://doi.org/10.1016/s0020-­1383(02)00035-­9. out cement fixation. Arch Orthop Trauma Surg. 2011;131:503–8.
21. Hage W, Aj A, Dm A. Incidence, location, and diagnostic evalua- https://doi.org/10.1007/s00402-­010-­1172-­6.
tion of metastatic bone disease. The Orthopedic clinics of North 40. Frassica FJ, Frassica DA. Metastatic bone disease of the
America; 2000. https://pubmed.ncbi.nlm.nih.gov/11043092/. humerus. J Am Acad Orthop Surg. 2003;11:282–8. https://doi.
Accessed 1 Dec 2020. org/10.5435/00124635-­200307000-­00008.
22. Tejwani NC, Guerado E. Improving fixation of the osteoporotic 41. Leopold SS. Editor’s spotlight/take 5: prophylactic stabiliza-
fracture: the role of locked plating. J Orthop Trauma. 2011;25:S56. tion for bone metastases, myeloma, or lymphoma: do we need
https://doi.org/10.1097/BOT.0b013e31821b8a52. to protect the entire bone? (DOI 10.1007/s11999-012-2656-1).
23. Harrington KD, Johnston JO, Turner RH, Green DL. The use of Clin Orthop Relat Res. 2013;471:703–5. https://doi.org/10.1007/
methylmethacrylate as an adjunct in the internal fixation of malig- s11999-­012-­2714-­8.
nant neoplastic fractures. J Bone Joint Surg Am. 1972;54:1665–76. 42. Nota S, Teunis T, Kortlever J, et al. Functional outcomes and com-
24. Schatzer J, Ha’eri GB, Chapman M. Methylmethacrylate as an plications after oncologic reconstruction of the proximal humerus.
adjunct in the internal fixation of intertrochanteric fractures of the J Am Acad Orthop Surg. 2018;26:403–9. https://doi.org/10.5435/
femur. J Trauma. 1978;18(10):732–5. JAAOS-­D-­16-­00551.
25. Cameron HU, Jacob R, Macnab I, Pilliar RM. Use of polymethyl- 43. Potter BK, Adams SC, Pitcher DJJ, et al. Proximal humerus recon-
methacrylate to enhance screw fixation in bone. J Bone Joint Surg. structions for tumors. Clin Orthop Relat Res. 2009;467:1035–41.
1975;57:655–6. https://doi.org/10.1007/s11999-­008-­0531-­x.
26. Motzkin NE, Chao EY, An KN, et al. Pull-out strength of screws from 44. Rödl RW, Gosheger G, Gebert C, et al. Reconstruction of the proxi-
polymethylmethacrylate cement. J Bone Joint Surg Br. 1994;76:320–3. mal humerus after wide resection of tumours. J Bone Joint Surg Br.
27. Wittenberg RH, Lee K-S, Shea M, et al. Effect of screw diameter, 2002;84:1004–8. https://doi.org/10.1302/0301-­620x.84b7.12989.
insertion technique, and bone cement augmentation of pedicular 45. Kaa AKS, Jørgensen PH, Søjbjerg JO, Johannsen HV. Reverse
screw fixation strength. Clin Orthop Relat Res. 1993;296:278–87. shoulder replacement after resection of the proximal humerus
28. Wang HM, Galasko CSB, Crank S, et al. Methotrexate loaded for bone tumours. Bone Joint J. 2013;95-B:1551–5. https://doi.
acrylic cement in the management of skeletal metastases: biome- org/10.1302/0301-­620X.95B11.31545.
chanical, biological, and systemic effect. Clin Orthop Relat Res. 46. Kumar D, Grimer RJ, Abudu A, et al. Endoprosthetic replacement
1995;312:173–86. of the proximal humerus. Long-term results. J Bone Joint Surg Br.
29. Wasserlauf S, Warshawsky A, Arad-Yelin R, et al. The release 2003;85:717–22.
of cytotoxic drugs from acrylic bone cement. Bull Hosp Jt Dis. 47. Ross A, Wilson J, Scales J. Endoprosthetic replacement of the prox-
1993;53(1):68–74. https://europepmc.org/article/med/8374496. imal humerus. J Bone Joint Surg Br. 1987;69-B:656–61. https://doi.
Accessed 1 Dec 2020. org/10.1302/0301-­620X.69B4.3611177.
30. Südkamp N, Bayer J, Hepp P, et al. Open reduction and internal 48. Raiss P, Kinkel S, Sauter U, et al. Replacement of the proxi-
fixation of proximal humeral fractures with use of the locking prox- mal humerus with MUTARS tumor endoprostheses. Eur J
imal humerus plate. Results of a prospective, multicenter, observa- Surg Oncol (EJSO). 2010;36:371–7. https://doi.org/10.1016/j.
tional study. J Bone Joint Surg Am. 2009;91:1320–8. https://doi. ejso.2009.11.001.
org/10.2106/JBJS.H.00006. 49. Boileau P, Watkinson DJ, Hatzidakis AM, Balg F. Grammont
31. Weiss KR, Bhumbra R, Biau DJ, et al. Fixation of pathological humeral reverse prosthesis: design, rationale, and biomechanics. J
fractures by the cemented plate technique. J Bone Joint Surg Br. Shoulder Elb Surg. 2005;14:147S–61S. https://doi.org/10.1016/j.
2011;93:1093–7. https://doi.org/10.1302/0301-­620X.93B8.26194. jse.2004.10.006.
45 Surgical Management of Metastatic Disease to the Upper Extremity 503

50. Abdeen A, Hoang BH, Athanasian EA, et al. Allograft-prosthesis atic review. Clin Orthop Relat Res. 2014;472:2245–53. https://doi.
composite reconstruction of the proximal part of the humerus: func- org/10.1007/s11999-­014-­3474-­4.
tional outcome and survivorship. J Bone Joint Surg. 2009;91:2406– 65. Hazan EJ, Hornicek FJ, Tomford W, et al. The effect of adjuvant
15. https://doi.org/10.2106/JBJS.H.00815. chemotherapy on osteoarticular allografts. Clin Orthop Relat Res.
51. Walker M, Brooks J, Willis M, Frankle M. How reverse shoulder 2001;385:176–81.
arthroplasty works. Clin Orthop Relat Res. 2011;469:2440–51. 66. Bus MPA, van de Sande MAJ, Taminiau AHM, Dijkstra PDS. Is there
https://doi.org/10.1007/s11999-­011-­1892-­0. still a role for osteoarticular allograft reconstruction in musculoskel-
52. De Wilde L, Boileau P, Van der Bracht H. Does reverse shoulder etal tumour surgery? a long-term follow-up study of 38 patients and
arthroplasty for tumors of the proximal humerus reduce impair- systematic review of the literature. Bone Joint J. 2017;99-B:522–30.
ment? Clin Orthop Relat Res. 2011;469:2489–95. https://doi. https://doi.org/10.1302/0301-­620X.99B4.BJJ-­2016-­0443.R2.
org/10.1007/s11999-­010-­1758-­x. 67. Benevenia J, Kirchner R, Patterson F, et al. Outcomes of a modular
53. Bonnevialle N, Mansat P, Lebon J, et al. Reverse shoulder arthro- intercalary endoprosthesis as treatment for segmental defects of the
plasty for malignant tumors of proximal humerus. J Shoulder Elb femur, tibia, and humerus. Clin Orthop Relat Res. 2016;474:539–
Surg. 2015;24:36–44. https://doi.org/10.1016/j.jse.2014.04.006. 48. https://doi.org/10.1007/s11999-­015-­4588-­z.
54. Grosel TW, Plummer DR, Everhart JS, et al. Reverse total shoul- 68. Ogink PT, Teunissen FR, Massier JR, et al. Allograft reconstruction
der arthroplasty provides stability and better function than hemi- of the humerus: complications and revision surgery. J Surg Oncol.
arthroplasty following resection of proximal humerus tumors. J 2019;119:329–35. https://doi.org/10.1002/jso.25309.
Shoulder Elb Surg. 2019;28:2147–52. https://doi.org/10.1016/j. 69. Rose PS, Shin AY, Bishop AT, et al. Vascularized free fibula
jse.2019.02.032. transfer for oncologic reconstruction of the humerus. Clin
55. Guven MF, Aslan L, Botanlioglu H, et al. Functional outcome of Orthop Relat Res. 2005;438:80–4. https://doi.org/10.1097/01.
reverse shoulder tumor prosthesis in the treatment of proximal blo.0000179586.34727.5b.
humerus tumors. J Shoulder Elb Surg. 2016;25:e1–6. https://doi. 70. Athwal GS, Chin PY, Adams RA, Morrey BF. Coonrad-Morrey
org/10.1016/j.jse.2015.06.012. total elbow arthroplasty for tumours of the distal humerus and
56. Marulanda GA, Henderson E, Cheong D, Letson GD. Proximal elbow. J Bone Joint Surg. 2005;87-B:1369–74. https://doi.
and total humerus reconstruction with the use of an aortograft org/10.1302/0301-­620X.87B10.16569.
mesh. Clin Orthop Relat Res. 2010;468:2896–903. https://doi. 71. Weber KL, Lin PP, Yasko AW. Complex Segmental Elbow
org/10.1007/s11999-­010-­1418-­1. Reconstruction After Tumor Resection. Clin Orthop
57. Abdeen A, Healey JH. Allograft-prosthesis composite reconstruc- Relat Res. 2003;415:31–44. https://doi.org/10.1097/01.
tion of the proximal part of the humerus: surgical technique. J Bone blo.0000093894.12372.53.
Joint Surg. 2010;92:188–96. https://doi.org/10.2106/JBJS.J.00167. 72. Dimas V, Kargel J, Bauer J, Chang P. Forequarter amputation for
58. Chacon A, Virani N, Shannon R, et al. Revision arthroplasty with malignant tumours of the upper extremity: case report, techniques
use of a reverse shoulder prosthesis-allograft composite. J Bone Joint and indications. Can J Plast Surg. 2007;15(2):83–5. https://doi.
Surg Am. 2009;91:119–27. https://doi.org/10.2106/JBJS.H.00094. org/10.1177/229255030701500204.
59. Sanchez-Sotelo J, Wagner ER, Houdek MT. Allograft-prosthetic 73. Wittig JC, Bickels J, Kollender Y, et al. Palliative forequarter ampu-
composite reconstruction for massive proximal humeral bone loss tation for metastatic carcinoma to the shoulder girdle region: indi-
in reverse shoulder arthroplasty. JBJS Essent Surg Tech. 2018;8:e3. cations, preoperative evaluation, surgical technique, and results. J
https://doi.org/10.2106/JBJS.ST.17.00051. Surg Oncol. 2001;77(2):105–13.
60. Hejna MJ, Gitelis S. Allograft prosthetic composite replacement for bone 74. Merimsky O, Kollender Y, Inbar M, et al. Is forequarter amputation
tumors. Semin Surg Oncol. 1997;13:18–24. https://doi.org/10.1002/ justified for palliation of intractable cancer symptoms? Oncology.
(SICI)1098-­2388(199701/02)13:1<18::AID-­SSU4>3.0.CO;2-­7. 2001;60:55–9. https://doi.org/10.1159/000055297.
61. King JJ, Nystrom LM, Reimer NB, et al. Allograft-prosthetic 75. Elsner U, Henrichs M, Gosheger G, et al. Forequarter amputa-
composite reverse total shoulder arthroplasty for reconstruc- tion: a safe rescue procedure in a curative and palliative setting in
tion of proximal humerus tumor resections. J Shoulder Elb Surg. high-grade malignoma of the shoulder girdle. World J Surg Oncol.
2016;25:45–54. https://doi.org/10.1016/j.jse.2015.06.021. 2016;14:216. https://doi.org/10.1186/s12957-­016-­0973-­7.
62. Gebhardt MC, Roth YF, Mankin HJ. Osteoarticular allografts for 76. Sawada R, Shinoda Y, Niimi A, et al. Multiple acrometastases in a
reconstruction in the proximal part of the humerus after excision of patient with renal pelvic urothelial cancer. Case Reports in Urology;
a musculoskeletal tumor. J Bone Joint Surg Am. 1990;72:334–45. 2017. https://www.hindawi.com/journals/criu/2017/7830207/.
63. Getty PJ, Peabody TD. Complications and functional out- Accessed 1 Dec 2020.
comes of reconstruction with an osteoarticular allograft 77. Puhaindran ME, Healey JH, Athanasian EA. Single ray amputa-
after intra-articular resection of the proximal aspect of the tion for tumors of the hand. Clin Orthop Relat Res. 2010;468:1390.
humerus. J Bone Joint Surg Am. 1999;81:1138–46. https://doi. https://doi.org/10.1007/s11999-­009-­1019-­z.
org/10.2106/00004623-­199908000-­00009. 78. Puhaindran ME, Athanasian EA. Double ray amputation for tumors
64. Teunis T, Nota SPFT, Hornicek FJ, et al. Outcome after recon- of the hand. Clin Orthop Relat Res. 2010;468:2976. https://doi.
struction of the proximal humerus for tumor resection: a system- org/10.1007/s11999-­010-­1389-­2.
 Surgical Management of Metastatic
Disease to the Lower Extremity 46
Nathan W. Mesko and Lukas M. Nystrom

Abstract 46.1 Introduction

Metastatic skeletal disease of the lower extremity can lead
to significant morbidity and quality of life diminishment After the spine, the most common locations for metastatic
due to the stresses in the lower extremity during ambula- skeletal involvement are the pelvis, ribs, and proximal femur.
tion. In order to maintain mobility and improve pain con- Distant metastases in the tibia/foot (distal to the knee) are
trol, interventions to address lower extremity metastatic rare (<10% incidence) and are most commonly represented
lesions are oftentimes considered. These interventions by lung cancer and renal cell carcinoma [1]. Because of the
can include radiation therapy, minimally invasive solu- increased stress loads in the lower extremity, the risk for
tions, percutaneous stabilization, open curettage and fixa- pathologic fracture is oftentimes high, and surgical interven-
tion, or en bloc excision and reconstruction using tion is commonly considered. According to a large national
endoprosthetic joint replacement techniques. Multiple registry population analysis quantifying characteristics of
factors should be considered when determining the best surgically treated non-spinal metastases, the most common
solution for each patient situation. Future research will site of surgically treated skeletal metastases, the proximal
help guide better treatment algorithms as our ability to femur, humerus, pelvis, and tibia—with the most common
predict fracture risk and prognosis helps to better balance histologies involved being that of breast, prostate, kidney,
construct durability and recovery following an lung, and myeloma [2]. Pragmatically, 70% of patients with
intervention. metastatic bone disease will experience significant pain [3].
Moreover, up to 30% of patients with metastatic lesions in
the pelvis and lower extremity will sustain a pathologic frac-
ture, with 90% requiring surgical intervention [4].
Learning Objectives Due to the high incidence of skeletal metastatic involve-
• Understand the indications for different treatment ment in the lower extremity and the important role that the
modalities when treating metastatic disease of the lower extremity anatomy plays in facilitating ambulation, a
lower extremity. keen understanding of the balance between the stresses is
• Recognize the availability of different predictive placed on the human body with varying interventional, the
models in determining the risk for pathologic risk for fracture, and overall patient prognosis is mandatory.
fracture. This chapter will help explore the thought process for deter-
• Recognize the role that anatomic location of skele- mining the type of intervention in these lower extremity
tal disease plays in determining the type of recon- skeletal-related events, as well as understand various tech-
structive technique utilized. niques specific to location in the femur, tibia, and foot/ankle.
• Discuss broad categories of complications that can
occur when treating metastatic disease of the lower
extremities. 46.2 General Treatment Considerations

When considering treatment options, the prioritized goals

should drive decisions based on maintaining the highest pos-
N. W. Mesko (*) · L. M. Nystrom
Cleveland Clinic, Cleveland, OH, USA sible quality of life. The goals should aim to (1) restore func-
e-mail: [email protected]; [email protected] tion, (2) maintain/restore mobility, and (3) reduce pain—all

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 505
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_46
506 N. W. Mesko and L. M. Nystrom

while attempting to minimize recovery time given the often- with pain relief, with single-fraction external beam radiation
times limited prognosis that can accompany patients display- therapy (EBRT) showing equivalence to multi-fraction regi-
ing advanced skeletal disease burden. Given the importance mens [8]. More targeted, focal dosing techniques, such as
of the lower extremities with ambulation, maintaining mobil- intensity-modulated radiation therapy [IMRT] or stereotactic
ity and function directly affects ECOG (Eastern Cooperative body radiation therapy [SBRT] can be delivered in scenarios
Oncology Group) Performance Status measurements— where either collateral damage to normal tissue should be
which is a commonly used standard when determining a limited, or in histologies that are historically considered
patient’s candidacy for systemic therapy or clinical trial more radioresistant—though the surgeon should realize that
inclusion. In addition, the surgical solution should be a dura- radiation techniques rarely diminish the risk for pathologic
ble construct that takes into account the limited bony healing fracture in isolation [9, 10].
potential of a pathologic fracture, with the goals to perform a Focal ablation techniques have been shown to be effective
single operation and begin immediate mobilization. Open and durable techniques for the treatment of appropriately
and percutaneous surgical options, minimally invasive solu- selected lesions [6]. Thermal induced necrosis of the neoplas-
tions (such as radiofrequency ablation [RFA], cryoablation, tic tissue, either through heat (RFA) or cold (cryoablation) is
or cementoplasty), and nonoperative modalities all have a meant to gain local control of the tumor, while supplementa-
role in addressing both impending and progressed fractures tion with cementoplasty theoretically is meant to help stabi-
of the lower extremity. Determining the appropriate course lize surrounding microfractures [5]. These techniques
of treatment should draw from a variety of information described above have limitations, including potential collat-
sources, including surgeon experience, publicly available eral thermal necrosis damage or cement extravasation in peri-
risk predictor models, patient presentation, and a multidisci- articular locations or locations adjacent to major neurovascular
plinary discussion surrounding multimodal goals of treat- structures. Furthermore, unique characteristics of the lytic
ment. When specifically considering surgical versus lesion, such as lesser trochanter involvement or >30 mm of
nonsurgical scenarios, the important patient presentation fac- axial length cortical destruction have portended to poorer out-
tors should include patient prognosis, histology, and surgical comes with minimally invasive techniques [5, 6].
candidacy based on comorbidities. Surgery should be favored
in the lower extremity in the following scenarios:
46.4 Surgical Technical Options
1. Patients in whom multimodal conservative treatment for Treatment
measures fail (if appropriate)
2. Compromised structural support in the bony anatomy that Surgical techniques may vary, and should be dictated pri-
is unlikely to respond to conservative treatment marily by balancing the need for a durable construct that
3. Metastatic lesions with historically poor response to radi- will outlive expected prognosis and the need for an expe-
ation or chemotherapy (e.g., melanoma, renal cell, lung, dited recovery. Oftentimes, commencement of adjuvant
etc.) therapies such as immunotherapy, cytotoxic therapy, and
4. Expected prognosis >4–12 weeks (though this will be sur- radiation therapy is dictated by wound healing and perfor-
gery and patient situation specific). As a general rule, recov- mance status. Intralesional curettage techniques utilizing
ery should not take longer than anticipated prognosis. adjuvant modalities to aid with microscopic tumor burden
Surgery should not be discounted as a choice if pain control control (e.g., cryoablation, phenol, electrocautery, concen-
and OOB transfers can be enhanced with stabilization trated hydrogen peroxide, thermal ablation), typically
despite a limited prognosis measured in weeks. require larger incisions but allow for improved debulking
of gross/microscopic disease. Percutaneous intralesional
techniques, such as an intramedullary device, create the
advantage of small wounds at the sacrifice of limited tumor
46.3  onoperative and Minimally Invasive
N debulking ability. Intramedullary vacuum irrigator devices
Options for Treatment can facilitate some manner of tumor debulking and allow
for biopsy material to be obtained. En bloc excisions have
Nonoperative and minimally invasive solutions are appropri- a limited role in metastatic disease, leaving larger defects
ate in scenarios with limited prognosis, poor surgical candi- that require reconstruction with endoprosthetic or interca-
dacy, or amendable lesions [5–7]. Bisphosphonates and lary implants. Pathologic fracture involving a joint or
rank-ligand inhibitor drugs have been shown to aid with pain select scenarios with solitary skeletal metastases can por-
improvement and increase time to a skeletal-related event tend to the need for en bloc resection and reconstruction
(SRE) in multiple histologies. Radiation is effective in aiding techniques [11–13]. Recent literature would suggest as
46 Surgical Management of Metastatic Disease to the Lower Extremity 507

high as a fourfold longer survival in patients with oligo- 46.6 Femur

metastatic renal cell carcinoma who undergo en bloc exci-
sion (as compared to intralesional surgery) [12, 14]. When considering surgical intervention for femoral metasta-
ses it is most helpful to divide the femur into proximal,
diaphyseal, and distal segments. This division can aid in sur-
46.5 Pathologic Fracture Predictive Models gical decision-making and implant selection. As with all
metastatic patients, once the decision to operate has been
Multiple algorithms have been used historically in determin- made, the surgeon should select the procedure that is intended
ing a patient’s overall risk for fracture in the setting of a to be definitive in light of the patient’s local lesion and ulti-
pathologic lesion. Examples of such models include: mate prognosis. As an example, consider two radiographi-
cally identical lesions destroying the peritrochanteric femur.
• 1971—Snell and Beals criteria for femoral metastases in One patient with an isolated renal cell metastasis may be
the breast cancer population highlighted impending treated with a proximal femoral resection and reconstruction
pathologic fracture risk include lytic lesions >2.5 cm and with a proximal femur replacement, whereas another patient
cortical involvement >50% (circumference) [15]. with progressive widespread osseous and visceral metastases
• 1989—The Mirels scoring system predicts risk of patho- may be treated with an intramedullary nail to prevent patho-
logic fracture based on (1) lesion size, (2) anatomic loca- logic fracture.
tion, (3) radiographic appearance, and (4) the presence/
severity of pain. This system’ reliability and reproducibil-
ity have been called into question with multiple sugges- 46.6.1 Proximal Femur: Head/Neck
tions to improve, but still remains a staple for lower
extremity risk prediction [16, 17]. The main decision points in determining the appropriate
• 2005—A randomized trial looking at patients at increased treatment are: (1) Whether there is sufficient bone in the
risk for fracture with femoral metastases undergoing radio- head/neck for internal fixation options and (2) whether fixa-
therapy used multiple criteria, highlighting pain, lesion size tion options will adequately address the mechanical instabil-
>25 mm, a proximal femoral location, axial cortical ity in the bone and thus relieve the patient’s pain.
involvement >30 mm or circumferential involvement Lesions isolated to the femoral head and neck are reliably
>50% as factors that increase the risk for fracture [18]. treated with arthroplasty. Resection of the head/neck with
• 2015—CT-based rigidity analysis (CTRA) has been more arthroplasty allows for optimal local surgical management of
recently demonstrated to be a better predictor of fracture the disease and an immediately durable reconstruction.
than the Mirels scoring system. This system defines an There is debate about whether to perform long-stem arthro-
impending pathologic fracture based on greater than 35% plasty stabilization [25] versus standard length and also
reduction in axial, bending, or torsional rigidities. This sys- whether to perform hemiarthroplasty versus total hip arthro-
tem requires CT imaging for every fracture, which is not plasty [26].
often necessary in order to plan for interventions. Follow-­up Consideration can be given to performing internal fixa-
studies have also not shown ability to influence surgical tion for select femoral neck lesions depending on the loca-
decision-making in simulated modeling [19–21]. tion. Decision to perform this must take into account size/
location of the lesion, blood supply to the femoral head, and
In more recent times, publicly available, validated predic- durability of fixation.
tive nomogram models are available for use at the bedside to
assess risk for fracture when taking into account age, gender,
lab values, histology, ECOG status, visceral and lymph node 46.6.2 Proximal Femur: Pertrochanteric
metastatic burden, number of skeletal metastases, and the
physician’s estimate of patient survival [22]. Impending or actual pathologic fractures of the pertrochan-
As a simple radiographic predictive determinant, the pres- teric region become more interesting in terms of the discus-
ence of a lesser trochanter fracture is generally considered to sion of method of stabilization. The armamentarium includes
show a strong association with malignant etiology, with the intramedullary nail, proximal femoral replacement, and long
risk for pathologic fracture elevated due to the predicted stem arthroplasty. Perhaps extrapolating the orthopedic sur-
load-bearing ability being significantly reduced. Because of geon’s experience with management of intertrochanteric
this, involvement of the lesser trochanter and surrounding fractures in the non-neoplastic setting, the reflex consider-
posteromedial calcar anatomy serves as a strong indication ation is to internally fix the majority of these lesions [27].
for operative intervention [23, 24]. There is, however, data to suggest that there are better out-
508 N. W. Mesko and L. M. Nystrom

comes for patients treated with arthroplasty techniques in The main consideration in the femur will be the mode of
this setting as well [28]. fixation. Intramedullary nail fixation is often preferred for
As an example, a patient with a pertrochanteric lesion purposes of spanning the entire bone, most notably the femo-
(Fig. 46.1a, lower arrow) may be appropriately treated with an ral neck, for prophylactic fixation in the event of future
intramedullary device for stabilization. However, if that same pathologic lesions at a more distant site in the same bone.
patient also has lesions in the femoral head/neck (Fig. 46.1a, However, there is evidence that the theoretical benefit of pro-
upper arrows) and also subsequently develops a pathologic tecting the entire bone is rarely utilized [30, 31]. Given this
fracture (Fig. 46.1b) that will likely shift the balance in deci- evidence, it is entirely reasonable to use plate/screw con-
sion-making toward proximal femoral resection and endo- structs if one already has the entire bone exposed for the pur-
prosthetic reconstruction (Fig. 46.1c). While intramedullary poses of marginal/intralesional management of a metastatic
fixation could be considered, there would be substantial con- lesion.
cerns regarding purchase of fixation in the pathologic femoral The larger question, as it is in many instances of metastatic
head/neck and also ability of the bone to heal given the need disease, is whether to perform intralesional management of
for palliative radiotherapy for local control of the disease. the tumor simultaneously with the stabilization procedure.
The surgical treatment of the pertrochanteric femoral This is most often performed in situations where: (1) the pri-
metastasis should not be prescribed based on the knowledge mary histology dictates based on history of radioresistance
of the general location in the bone alone. It should be based (i.e., renal cell, lung adenocarcinoma) or (2) there is substan-
on the amount of bone destruction, location in relation to the tial destruction warranting intralesional resection and cement
calcar, soft tissue mass, tumor histology, anticipated response augmentation to further promote skeletal stability.
to radiotherapy/systemic therapy, the patient’s functional sta-
tus, and overall prognosis.
46.6.4 Distal Femur

46.6.3 Diaphyseal Femur As in the proximal femur, the distal femur considerations are
large whether there is the ability to salvage the function of the
The vast majority of diaphyseal lesions in the femur will be articular structures. If the joint can be salvaged, and predict-
treated with some form of internal fixation, although seg- ably maintained based on the natural history of the histologic
mental resection and intercalary prosthetics may be consid- subtype, then it is the favored approach (Fig. 46.2a). If it can-
ered in the setting of solitary or oligometastatic disease [29]. not, or the underlying condition of the joint (i.e., arthritis and/

a b c

Fig. 46.1 Examples of proximal and diaphyseal femur reconstruction. was subsequently treated with a proximal femoral endoprosthetic
A large metastatic lytic lesion burden in the calcar and femoral neck (a) replacement (c)
led to a pathologic subtrochanteric pathologic femur fracture (b). This
46 Surgical Management of Metastatic Disease to the Lower Extremity 509

or intra-articular extension of a pathologic fracture) renders it en bloc excision and reconstruction, or amputation. In the
a major source of disability then an arthroplasty technique is prophylactic setting or minimally displaced fracture with
preferred, most often distal femoral replacement (Fig. 46.2b). contained subchondral margins, stabilization utilizing a
plate/screw construct or intramedullary device is standard,
oftentimes coupled with open intralesional curettage, adju-
46.7 Tibia/Fibula vant therapy treatment (e.g., heat, cold, phenol, etc.) and
bone cement supplementation. In the setting of metastatic
Though metastatic disease distal to the knee (“acral” metas- disease, en bloc excision is a less commonly used technique.
tases) occurs at a <10% incidence, goals of treatment remain Amputation is generally used only in settings where recon-
similar [1]. The proximal tibial metaphyseal location is the struction techniques do not offer a durable longevity and
most common location of involvement distal to the knee. As pain is unable to be effectively addressed with nonoperative
with other locations, treatment categories include an intrale- measures. Examples of various reconstruction techniques
sional approach with supplemental stabilization constructs, can be found in Fig. 46.2.

a b c d

e f g h

Fig. 46.2 Examples of distal femur, tibia, and foot reconstruction. sis (d) are options, depending on subchondral bony involvement, then
Reconstructions of the knee and tibia/foot locations should address the contained defect present, and expected prognosis. Diaphyseal lesions
tumor burden while the technique should be tailored to anticipate are generally treated with a locked intramedullary device (e), though
patient prognosis and the required longevity of reconstruction. occasionally the burden of tumor leaves such poor bone tock quality
Curettage with plate/bone cement reconstruction can be used to pre- that an amputation is the best solution (f). Distal tibial lesions and foot
serve the knee joint (a) or a distal femur replacement may be required lesions are generally treated with intralesional procedures reinforced
when extensive bone replacement by tumor burden is encountered (b). with plates, screws, and/or bone cement, such as in this metastatic lung
Likewise, in the proximal tibia, intralesional procedures reinforced with cancer medial malleolus lesion (g) or this metastatic renal carcinoma
plate/bone cement techniques (c) or replacement with an endoprosthe- talus lesion (h)
510 N. W. Mesko and L. M. Nystrom

46.7.1 Proximal can be used. If en bloc excision is required, reconstruction of

the lateral collateral complex is necessary with hinged knee
The mainstay “workhorse” of the proximal tibia anatomy is brace protection postoperatively.
plate/screw reconstruction—utilizing open incisions that
facilitate combined techniques of high-speed burr curettage,
adjuvant therapy with adjuvant agents, and filling the defect 46.7.2 Diaphyseal
with cement prior to plate supplementation. Intramedullary
device solutions are less commonly employed given the lack While intramedullary nail placement generally is used for
of ability to gain adequate fixation through the compromised tibial diaphyseal lesions, plate/screw techniques can play a
metaphyseal bone, as well as the lack of ability to add reli- role. If adjacent implants or hardware are present, such as a
able support to the subchondral weight-bearing surface. diaphyseal tibial lesion below a total knee replacement, plat-
Arthroplasty solutions can be considered when (1) the ing techniques are more pragmatic to address impending or
subchondral bony anatomy is compromised to a point that outright pathologic fracture concerns. Utilizing cement aug-
would risk cement extravasation into the intra-articular space mentation in large defects can add structural stability to the
with plate/bone cement reconstruction, (2) a displaced patho- construct. As with en bloc excisions in the proximal anatomy,
logic fracture precludes the ability for a durable joint salvage this is a rarely employed technique in tibial diaphyseal anat-
reconstruction option, or (3) a peri-implant lesion occurs omy. If an en bloc excision is indicated, then a modular inter-
below a prior knee arthroplasty that can be addressed utiliz- calary implant can be utilized, utilizing cement fixation for the
ing a stemmed revision baseplate construct. When consider- stems with or without intercalary cross-locking screws.
ing arthroplasty solutions, if the rim of metaphyseal bone is Fibular diaphyseal lesions failing nonoperative palliative
present to create a “contained” defect after curettage, then modalities are treated with en bloc excision most
conventional tibial revision baseplates with cemented stems commonly.
can be used to bypass the defect, with potential metaphyseal
tibial cone augmentation. With large, uncontained defects or
displaced pathologic fractures, endoprosthetic reconstruc- 46.7.3 Distal
tions with a proximal tibial hinged knee replacement (PTR)
can be used—though these come with concerns of high As with the proximal anatomy, plate/screw supplementation
infection rates and soft tissue coverage issues [32–35]. with bone cement augmentation is the common of the proxi-
Whenever possible, if the opportunity is available to save the mal tibia plate/screw reconstruction—utilizing open inci-
extensor mechanism attachment (i.e., tibial tubercle) without sions that facilitate combined techniques of high-speed burr
compromising construct durability, this should be priori- curettage, adjuvant therapy with adjuvant agents, and filling
tized, in order to address the high rate of extensor lag deficits the defect with cement prior to plate supplementation.
following PTR [36]. Smaller contained defects may be addressed with percutane-
Indications for en bloc excision in the setting of meta- ous techniques, including screw placement with cement sup-
static disease are rare. These include oligometastatic disease plementation. With limited subcondral destruction in the
where potential improvement in overall survival may be setting of an uncontained defect, utilizing gel foam over the
afforded by complete excision of certain histologies (i.e., articular defect can help to contain cement during an intral-
renal cell carcinoma), large lesions that affect the articular esional curettage procedure and allow for joint salvage.
weight-bearing surface which does not allow for joint sal- Arthroplasty is not a reliable solution at the tibiotalar joint—
vage, or peri-articular pathologic fractures with minimal if subchondral anatomic destruction precludes joint salvage,
bone stock available for reconstruction. Unlike in the setting a below-knee amputation (BKA) will precipitate a higher
of primary bone malignancies, biologic reconstruction with degree of postoperative function in most cases. In rare sce-
bulk allograft (intercalary allograft, allograft–prosthetic narios, placement of a retrograde tibial-talar-calcaneal (TTC)
composite) or custom implant solutions are not used as stan- fusion nail with cement augmentation can be considered—
dard reconstructions given the need for limited weight bear- though a BKA may still afford better over functional out-
ing and long recovery requirements. Modular endoprosthesis comes given the known reliability of the procedure.
allowing for PTR with cement fixation is the standard solu-
tion that allows for the quickest mobilization.
Fibular lesions are very rare and treated nonoperatively, 46.7.4 S
 oft-Tissue Considerations in Tibial
commonly palliated with radiation and temporary protected Reconstruction
weight bearing as a general rule. In contained defects that
have failed non-interventional measures, percutaneous Given the subcutaneous anatomy of the tibia, coverage over a
cementoplasty or open curettage, and cement augmentation reconstruction construct is vital to prevent again postoperative
46 Surgical Management of Metastatic Disease to the Lower Extremity 511

wound-healing concerns leading to infection. In an intercalary with the surgical technique utilized (i.e., hip instability with
resection and modular implant reconstruction event, a regional hip arthroplasty, patellar maltracking with distal femoral
soleus or gastrocnemius flap combination can be considered. replacement, etc.).
More commonly, gastrocnemius neurovascular pedicled flaps While some complications are inherent to the procedure
have become a mainstay to supplement PTR, with the twofold being performed, the more discouraging complications are
advantage of decreasing infection rates and allowing for those that require reoperation and which may have been
improved extensor mechanism augmentation, aiding with post- avoided with optimizing the surgical plan.
operative improved extensor mechanism function [34, 36].
If soft-tissue coverage is compromised secondary to adju-
vant radiation therapy, and temporization of the soft tissue 46.10.1 Iatrogenic Fracture
envelope is needed prior to definitive stabilization and cover-
age, external fixation is a described potential technique [37]. When performing prophylactic internal fixation one of
the complications that may be encountered is iatrogenic
fracture. This can occur for a variety of reasons, not the
46.8 Patella least of which is the operating environment of pathologic
bone. It is important to ensure appropriate nail path to
The incidence of metastatic disease in the patella is rare, minimize stress on the pathologic bone. Furthermore, if a
described in case report literature [38–41]. In small lesions, pathologic fracture is identified it is of the utmost impor-
conservative therapy with antiresorptive agents, radiation, tance that it be appropriately managed. This may mean
and systemic therapy should be considered as the first line. In converting operative approaches to optimize reduction,
larger lesions or lesions that fail conservative therapy, sur- or potentially even switching implants. A poor reduction
gery can be performed. If there is a contained that does not of pathologic bone risks nonunion and implant failure
communicate with the joint, curettage with cement augmen- (Fig. 46.3a).
tation can be used. Patellectomy is reserved in large lesions
that give poor reconstructive options. This creates weakness
with extension given the loss of a native fulcrum, and an 46.10.2 Cortical Perforation
adaptive gait will need to be employed to adapt to quadriceps
weakness and altered muscle control. By definition, prophylactic internal fixation with an intra-
medullary device involves placing an intramedullary implant
in an unbroken bone. While modern implants have made
46.9 Foot efforts to normalize the radius of curvature of the implant to
most femora, there will be instances where there is a mis-
Acral metastases to the foot occur rarely, reported at <0.3% inci- match. Furthermore, if proper technique and starting points
dence [42–44]. The calcaneus and talus are the most commonly are not used there will be a risk of perforating the anterior
involved anatomic locations. Treatment combines a multimodal cortex of the distal femur. In order to minimize the risk of
approach to consist of radiation, temporary immobilization (if this complication, the starting point on the trochanter should
there is a chance for healing to occur, such as in myeloma, lym- be perfectly at the tip or biased anterior. If cortical perfora-
phoma, or breast cancer), orthotics, minimally invasive strategies tion does occur the nail should either be redirected and
(in larger bones of the hindfoot), and surgery. Involvement of the ensured that the defect is spanned or supplemental fixation
distal phalanges portends to the need for amputation, while should be added (Fig. 46.3b).
metatarsal, midfoot, and hindfoot anatomy may benefit from
open curettage and cement augmentation +/− plate and screw
fixation in contained lesions. As with distal tibial lesions, a BKA 46.10.3 Disease Progression/Device Failure
is a reproducible operation with reliable functional outcomes in
The overall risk of device failure is thought to be low,
patients where limb salvage surgery will not aid reliably with
pain relief or a durable construct. with revision rates reported ranging from 5% to 8.9%
[45, 46]. The low rate of revision is largely contributed
by the poor overall survival in patients with metastatic
46.10 Complications cancer. Factors that are felt to contribute to implant fail-
ure include renal carcinoma metastases [45], patients
The potential complications that can be experienced in the with actual fractures (as opposed to impending) [46], a
management of lower extremity metastatic disease are wide-­ history of radiotherapy prior to the surgical treatment
ranging and encompass all of the complications associated [45, 46].
512 N. W. Mesko and L. M. Nystrom

a b

Fig. 46.3 Complications. A poor reduction of a subtrochanteric patho- on the greater trochanter led to distal anterior cortical perforation of the
logic femur fracture spanned with an undersized cephalomedullary femur, requiring a separate distal incision in order to plate a bypassing
device culminated in early failure (a). A start point place too posterior plate construct protecting against cortical perforation stress risers

46.11 Conclusion
Open Questions
Lower extremity location for metastatic disease is a common • When is intralesional treatment of the metastatic
occurrence and patients often present with significant pain-­ lesion warranted and when is it not necessary?
causing ambulatory limitations. Understanding the “com- • What are appropriate indications and methods of
plete picture” that encompasses prognosis, tumor delivery for using adjuvant treatments to the lesion
characteristics, anatomic support and stress loading, and a site in order to offer an advantage in decreasing
multidisciplinary plan for cancer treatment is necessary to local recurrence rates? (i.e., topical application,
tailoring an appropriate algorithm for addressing skeletal-­ bone cement impregnation with adjuvant, histology-­
related events in the lower extremity. Predictive models are specific chemotherapeutics and/or bone-directed
available and being contemporarily refined to better deter- agents, etc.)
mine appropriate surgical and nonsurgical solutions for treat- • When does wide resection of oligometastatic dis-
ment. These models have limitations and do not substitute ease represent a survival advantage?
for multidisciplinary communication between the medical • Utilizing predictive models, what presentation
oncology, radiation oncology, radiology, pathology, surgical, characteristics (histology, prognosis, age, location,
palliative medicine, and physical therapy team for coordina- etc.) will accurately guide a surgeon toward choos-
tion of care and goals-directed therapy. ing between an endoprosthetic device versus an
Surgical planning should prioritize a durable construct intramedullary nail fixation versus plate + cement
that outlasts overall prognosis, and allow for immediate technique of a metastatic lesion?
weight-bearing and joint motion, utilizing the “least inva-
sive” approach that will accomplish this. Maintaining the
quality of life and alleviating pain should be the ultimate References
goals.
In the future, newer technologies that can allow for more 1. Leeson MC, Makley JT, Carter JR. Metastatic skeletal disease dis-
minimally invasive delivery of resection and reconstruction tal to the elbow and knee. Clin Orthop. 1986;206:94–9.
solutions will be considered. These can include robotic tech- 2. Ratasvuori M, Wedin R, Keller J, et al. Insight opinion to surgi-
cally treated metastatic bone disease: Scandinavian Sarcoma Group
nology, virtual/augmented reality, or better predictive mod- Skeletal Metastasis Registry report of 1195 operated skeletal metas-
els that more reliably stratify and match appropriate treatment tasis. Surg Oncol. 2013;22(2):132–8. https://doi.org/10.1016/j.
techniques with patient presentation. suronc.2013.02.008.
46 Surgical Management of Metastatic Disease to the Lower Extremity 513

3. Varadhachary GR, Abbruzzese JL, Lenzi R. Diagnostic strate- ct-based rigidity analysis. Clin Cancer Res. 2015;21(11):2514–9.
gies for unknown primary cancer. Cancer. 2004;100(9):1776–85. https://doi.org/10.1158/1078-­0432.CCR-­14-­2668.
https://doi.org/10.1002/cncr.20202. 21. Nazarian A, Entezari V, Villa-Camacho JC, et al. Does CT-based
4. Harrington KD. Orthopaedic management of extremity and pelvic rigidity analysis influence clinical decision-making in simula-
lesions. Clin Orthop. 1995;312:136–47. tions of metastatic bone disease? Clin Orthop. 2016;474(3):652–9.
5. Mesko NW, Lawrenz JM, Lietman SA, Joyce MJ, Winalski CS, https://doi.org/10.1007/s11999-­015-­4371-­1.
Ilaslan H. Minimally invasive techniques for pain palliation in 22. Home. PathFX. https://www.pathfx.org/. Accessed 19 Dec 2020.
extraspinal bone metastases: a review of cementoplasty and radio- 23. Shinoda Y, Sawada R, Fujiwara S, Inokuchi H, Karasawa Y, Haga
frequency ablation. Curr Orthop Pract. 2016;27(6):686–95. https:// N. Prediction of the pathological fracture risk during stance and
doi.org/10.1097/BCO.0000000000000440. fall-loading configurations for metastases in the proximal femur,
6. Lawrenz J, Ilaslan H, Lietman S, Joyce M, Winalski C, Mesko using a computed tomography-based finite element method. Ann
N. Minimally invasive techniques for pain palliation in extra- Phys Rehabil Med. 2018;61:e412–3. https://doi.org/10.1016/j.
spinal bone metastases: a review of conventional methods and rehab.2018.05.961.
cryoablation. Curr Orthop Pract. 2016;27(5):547–52. https://doi. 24. Cho HS, Lee Y-K, Yoon B-H, Park J-W, Ha Y-C, Koo K-H. Isolated
org/10.1097/BCO.0000000000000417. avulsion fracture of the lesser trochanter in adults. In Vivo.
7. Ogbata O, Portnoy D. Nonoperative management of patients with 2020;34(6):3519–26. https://doi.org/10.21873/invivo.12193.
carcinoma metastatic to bone. Curr Orthop Pract. 2014;25:520–4. 25. Patterson BM, Healey JH, Cornell CN, Sharrock NE. Cardiac
https://doi.org/10.1097/BCO.0000000000000175. arrest during hip arthroplasty with a cemented long-stem compo-
8. Lutz S, Chow E. Palliative radiotherapy: past, present and future-­ nent. A report of seven cases. J Bone Joint Surg Am. 1991;73(2):
where do we go from here? Ann Palliat Med. 2014;3(4):286–90. 271–7.
https://doi.org/10.3978/j.issn.2224-­5820.2014.10.04. 26. Reif TJ, Strotman PK, Kliethermes SA, Miller BJ, Nystrom
9. Francolini G, Detti B, Ingrosso G, et al. Stereotactic body radia- LM. No consensus on implant choice for oligometastatic disease
tion therapy (SBRT) on renal cell carcinoma, an overview of of the femoral head and neck. J Bone Oncol. 2018;12:14–8. https://
technical aspects, biological rationale and current literature. Crit doi.org/10.1016/j.jbo.2018.02.006.
Rev Oncol Hematol. 2018;131:24–9. https://doi.org/10.1016/j. 27. Steensma M, Healey JH. Trends in the surgical treatment of patho-
critrevonc.2018.08.010. logic proximal femur fractures among Musculoskeletal Tumor
10. Rühle A, Andratschke N, Siva S, Guckenberger M. Is there a Society members. Clin Orthop. 2013;471(6):2000–6. https://doi.
role for stereotactic radiotherapy in the treatment of renal cell org/10.1007/s11999-­012-­2724-­6.
­carcinoma? Clin Transl Radiat Oncol. 2019;18:104–12. https://doi. 28. Steensma M, Boland PJ, Morris CD, Athanasian E, Healey
org/10.1016/j.ctro.2019.04.012. JH. Endoprosthetic treatment is more durable for pathologic proxi-
11. Baloch KG, Grimer RJ, Carter SR, Tillman RM. Radical sur- mal femur fractures. Clin Orthop. 2012;470(3):920–6. https://doi.
gery for the solitary bony metastasis from renal-cell carci- org/10.1007/s11999-­011-­2047-­z.
noma. J Bone Joint Surg Br. 2000;82-B(1):62–7. https://doi. 29. Huang H-C, Hu Y-C, Lun D-X, et al. Outcomes of intercalary pros-
org/10.1302/0301-­620X.82B1.0820062. thetic reconstruction for pathological diaphyseal femoral fractures
12. Ouzaid I, Capitanio U, Staehler M, et al. Surgical metastasectomy secondary to metastatic tumors. Orthop Surg. 2017;9(2):221–8.
in renal cell carcinoma: a systematic review. Eur Urol Oncol. https://doi.org/10.1111/os.12327.
2019;2(2):141–9. https://doi.org/10.1016/j.euo.2018.08.028. 30. Alvi HM, Damron TA. Prophylactic stabilization for bone metas-
13. Ljungberg B. The role of metastasectomy in renal cell carcinoma tases, myeloma, or lymphoma: do we need to protect the entire
in the era of targeted therapy. Curr Urol Rep. 2013;14(1):19–25. bone? Clin Orthop. 2013;471(3):706–14. https://doi.org/10.1007/
https://doi.org/10.1007/s11934-­012-­0293-­6. s11999-­012-­2656-­1.
14. Ratasvuori M, Wedin R, Hansen BH, et al. Prognostic role of en-­ 31. Moon B, Lin P, Satcher R, Bird J, Lewis V. Intramedullary nail-
bloc resection and late onset of bone metastasis in patients with ing of femoral diaphyseal metastases: is it necessary to protect the
bone-seeking carcinomas of the kidney, breast, lung, and prostate: femoral neck? Clin Orthop. 2015;473(4):1499–502. https://doi.
SSG study on 672 operated skeletal metastases. J Surg Oncol. org/10.1007/s11999-­014-­4064-­1.
2014;110(4):360–5. https://doi.org/10.1002/jso.23654. 32. Jeys LM, Kulkarni A, Grimer RJ, Carter SR, Tillman RM,
15. Beals RK, Lawton GD, Snell WE. Prophylactic internal fixation of Abudu A. Endoprosthetic reconstruction for the treatment
the femur in metastatic breast cancer. Cancer. 1971;28(5):1350–4. of musculoskeletal tumors of the appendicular skeleton and
https://doi.org/10.1002/1097-­0142(1971)28:5<1350::aid-­cncr2820 pelvis. J Bone Joint Surg. 2008;90(6):1265–71. https://doi.
280539>3.0.co;2-­6. org/10.2106/JBJS.F.01324.
16. Mirels H. Metastatic disease in long bones. A proposed scoring sys- 33. Henderson ER, Groundland JS, Pala E, et al. Failure mode classifi-
tem for diagnosing impending pathologic fractures. Clin Orthop. cation for tumor endoprostheses: retrospective review of five insti-
1989;249:256–64. tutions and a literature review. JBJS. 2011;93(5):418–29. https://
17. Howard EL, Shepherd KL, Cribb G, Cool P. The validity of the doi.org/10.2106/JBJS.J.00834.
Mirels score for predicting impending pathological fractures 34. Myers GJC, Abudu AT, Carter SR, Tillman RM, Grimer RJ. The
of the lower limb. Bone Jt J. 2018;100-B(8):1100–5. https://doi. long-term results of endoprosthetic replacement of the proximal
org/10.1302/0301-­620X.100B8.BJJ-­2018-­0300.R1. tibia for bone tumours. J Bone Joint Surg Br. 2007;89-B(12):1632–
18. Van der Linden YM, Dijkstra PDS, Kroon HM, et al. Comparative 7. https://doi.org/10.1302/0301-­620X.89B12.19481.
analysis of risk factors for pathological fracture with femoral 35. Capanna R, Scoccianti G, Frenos F, Vilardi A, Beltrami G,
metastases. J Bone Joint Surg Br. 2004;86(4):566–73. Campanacci DA. What was the survival of megaprostheses in
19. Damron TA, Nazarian A, Entezari V, et al. CT-based struc- lower limb reconstructions after tumor resections? Clin Orthop.
tural rigidity analysis is more accurate than mirels scoring for 2015;473(3):820–30. https://doi.org/10.1007/s11999-­014-­3736-­1.
fracture prediction in metastatic femoral lesions. Clin Orthop. 36. Jentzsch T, Erschbamer M, Seeli F, Fuchs B. Extensor function
2016;474(3):643–51. https://doi.org/10.1007/s11999-­015-­4453-­0. after medial gastrocnemius flap reconstruction of the proximal
20. Nazarian A, Entezari V, Zurakowski D, et al. Treatment planning tibia. Clin Orthop. 2013;471(7):2333–9. https://doi.org/10.1007/
and fracture prediction in patients with skeletal metastasis with s11999-­013-­2851-­8.
514 N. W. Mesko and L. M. Nystrom

37. Oskouei SV, Walton ZJ. Treatment of metastatic disease to the tibia, 43. Hattrup SJ, Amadio PC, Sim FH, Lombardi RM. Metastatic tumors
ankle, and foot. Curr Orthop Pract. 2014;25(6):554–7. https://doi. of the foot and ankle. Foot Ankle. 1988;8(5):243–7. https://doi.
org/10.1097/BCO.0000000000000176. org/10.1177/107110078800800503.
38. Gasagranda B, Leeman K, Heller MT. Knee pain and swelling: an 44. Maheshwari AV, Chiappetta G, Kugler CD, Pitcher JD, Temple
atypical presentation of metastatic colon cancer to the patella. Radiol HT. Metastatic skeletal disease of the foot: case reports and lit-
Case Rep. 2014;9(1):e00031. https://doi.org/10.2484/rcr.v9i1.890. erature review. Foot Ankle Int. 2008;29(7):699–710. https://doi.
39. Benedek TG. Lysis of the patella due to metastatic carci- org/10.3113/FAI.2008.0699.
noma. Arthritis Rheum. 1965;8:560–7. https://doi.org/10.1002/ 45. Miller BJ, Soni EEC, Gibbs CP, Scarborough MT. Intramedullary
art.1780080410. nails for long bone metastases: why do they fail? Orthopedics.
40. Pauzner R, Istomin V, Segal-Lieberman G, Matetzky S, Farfel 2011;34(4) https://doi.org/10.3928/01477447-­20110228-­12.
Z. Bilateral patellar metastases as the clinical presentation of bron- 46. Willeumier JJ, Kaynak M, van der Zwaal P, et al. What factors are
chogenic adenocarcinoma. J Rheumatol. 1996;23(5):939–41. associated with implant breakage and revision after intramedullary
41. Li G, Shan C, Sun R, et al. Patellar metastasis from primary tumor. nailing for femoral metastases? Clin Orthop. 2018;476(9):1823–
Oncol Lett. 2018;15(2):1389–96. https://doi.org/10.3892/ol.2017.7478. 33. https://doi.org/10.1007/s11999.0000000000000201.
42. Ellington JK, Ellington K, Kneisl JS. Acrometastasis to the foot:
three case reports with primary colon cancer. Foot Ankle Spec.
2009;2(3):140–5. https://doi.org/10.1177/1938640009335271.
 Surgical Management of Metastatic
Disease to the Pelvis 47
Timothy J. Evans, Odion Binitie, and David M. Joyce

Abstract We will also discuss the variations in management based on

This chapter reviews surgical treatment of metastatic dis- those patients presenting with oligometastatic disease
ease to the bones of the pelvis from cancers that have a (though poorly defined, often <5 metastatic lesions) versus
proclivity for lymph node spread. The anatomy of the pel- diffuse metastatic disease (>5 metastatic lesions). This con-
vis is correlated to both nonsurgical and surgical manage- cept is especially important in patients with cervical, uterine,
ment. The surgical techniques are reviewed as well as melanoma, thyroid, and renal cell cancers whose prognosis
their reported outcomes. and treatment strategy will depend on the extent of meta-
static disease at the time of diagnosis [2–4]. The most com-
mon metastasis to all bones, including the pelvis are breast,
prostate, thyroid, lung, and kidney. Common cancers that
Learning Objectives specifically spread via lymphatic pathways include lym-
phoma, head and neck cancers, breast cancer, gynecological
• Outline common primary cancers that are com- cancers, melanoma, kidney cancer, and lung cancer. Renal
monly found to have both lymphovascular spread as cell carcinoma (RCC) is one of the most common primary
well as bony metastatic disease of the pelvis cancers to spread to lymph node (LN) and bone, with a 12%
• Discuss the initial workup of patients found to have risk of metastases to LN and 34% risk to bone at the time of
bony metastatic disease of the pelvis diagnosis. RCC metastases tend to be highly vascular tumors
• Relate the relevant anatomy of the pelvis to com- and often necessitate preoperative embolization in the pelvis
mon locations of bony metastatic disease and how to minimize intraoperative blood loss. Lymphoma, mela-
this helps direct treatment recommendations noma, and lung cancer can also present with large associated
• Discuss common nonoperative and operative inter- soft tissue masses in the pelvis and associated bone destruc-
ventions used in the management of this disease tion. Squamous cell carcinoma of the head and neck devel-
ops metastatic bone disease in about 20% of patients. The
lesions in these patients were found to be located in the pel-
vis or femur more than 50% of the time [5, 6].
47.1 Introduction

The pelvis, possessing large proportions of the body’s red 47.2 Pelvic Metastases and Lytic Pelvic
bone marrow, is one of the most common sites of metastatic Bone Lesion Workup
disease of bone [1]. In the following paragraphs, we will out-
line the common types of cancer that spread through lym- Evaluation of the patient with suspected pelvic metastases
phatic pathways and present with metastasis to the pelvis, the should begin with a comprehensive physical examination
particular characteristics of the disease type, and other fea- including evaluation of all four extremities with respect to
tures which have implications on surgical decision making. pain, sensorimotor and vascular function as well as range of
motion. An assessment of the spine, especially lumbar is
T. J. Evans · O. Binitie · D. M. Joyce (*) critical. Orthogonal radiographs should be obtained of any
H. Lee Moffitt Cancer Center, Tampa, FL, USA painful areas. In terms of radiographic evaluation of the pel-
e-mail: [email protected]; [email protected] vis, an anterior–posterior view should be the initial image

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 515
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_47
516 T. J. Evans et al.

obtained. The iliopectineal and ilioischial lines, as well as pelves via strong sacroiliac ligaments and connected to the
the anterior and posterior walls, and the sourcil (acetabular more caudal aspects of the pelvis via the sacrospinous and
roof) should be evaluated for continuity. Additional imaging sacrotuberous ligaments. Interestingly, the sacral bony archi-
depends on the area of concern within the pelvis, with Judet tecture is such that the densest bone lies in the trajectory of
views and inlet/outlet views being the most common. Judet the sacral pedicles and on either side of this dense bone,
views of the pelvis provide unobstructed views of the ante- there are regions with minimal cancellous bone. This archi-
rior and posterior columns, the anterior and posterior walls tecture is important to keep in mind when evaluating the
of the acetabulum. Inlet and outlet views provide orthogonal sacrum for lytic metastatic lesions as some of this perceived
imaging of the pelvic ring as well as additional views of the decreased density can be mistaken for lytic lesions. This
sacrum. understanding of the bone quality in the area often can sup-
Cross-sectional imaging is a valuable tool in the evalua- port nonoperative management.
tion and treatment of pelvic metastases due to the complex, The ilium provides the connection from the axial spine to
three-dimensional anatomy of the pelvis as well as the prox- the legs through the sciatic buttress, which forms the anterior
imity of major neurovascular structures. These scans should and superior walls of the greater sciatic notch and is respon-
be obtained in three planes—coronal, sagittal, and trans- sible for transmitting forces between the acetabulum and hip
verse—and the combination of findings on all three planes joint and the axial skeleton. It also serves as the origin of
helps determine the true nature of the metastatic lesions. CT most of the muscles required for mobilization. The sciatic
scans provide the best detail of bony anatomy and structural buttress is the thickest and strongest part of the iliac bone and
integrity of the pelvis. MRI is used to help delineate the neu- can be approximated by the pelvic brim, between the sacro-
rovascular structures and soft tissues. MRI can also evaluate iliac joint and superomedial acetabulum [7]. If the sciatic
for inflammation and infiltrative neoplasms that can explain buttress is uninvolved by metastasis nonoperative treatment
a patient’s symptoms when X-rays are inconclusive. T2 can be utilized to manage metastatic lesions in this area
hyperintense on fluid-weighted sequences and contrasted (Fig. 47.1). The ligamentous anatomy of the ilium has been
images can delineate tumor extent in the bone and soft tis- studied extensively in the orthopedic literature as it has sig-
sues. Radiographs, CT, MRI can help evaluate for fracture, nificant implications when determining stability after pelvic
tumor extent, and aid in intervention planning. ring fractures, [8] which in turn can help the orthopedic sur-
geon determine the need for stabilization. Even with signifi-
cant metastatic burden of the pelvis, and in large lesions, the
47.3 Pelvis Structure and Anatomic ligamentous anatomy can maintain the pelvic ring stability.
Considerations The acetabulum is made up of anterior and posterior col-
umns in an inverted-Y pattern in which the convalescence of
The hemipelvis is made up of three bones, all of which fuse the columns occurs in the supra-acetabular region. The ante-
at the triradiate cartilage during growth. Due to the dynamic rior, posterior, superior, and medial walls of the acetabulum
nature of the forces experienced by the pelvis, much of the all play important roles in maintaining the congruency and
stability comes from the ligamentous structures connecting stability of the hip joint. The competency of the medial and
the two hemipelves to each other and the sacrum. The pelvis superolateral acetabulum, as the weight-bearing dome,
is unique with respect to metastatic disease due to its com- helps determine if a lytic lesion can be treated non-opera-
plex, three-dimensional anatomy and its proximity to multi- tively or dictate the characteristics of the surgical
ple, critical neurovascular structures. The areas of the pelvis reconstruction.
that are directly involved in weight bearing include the The ischium provides the main origin for many of the
sacrum, sciatic buttress, and acetabulum while the remaining hamstring and adductor muscles. Similarly, the pubic bone
iliac wing and pubic rami do not contribute to force transmis- provides the anterior wall of the acetabulum and the insertion
sion for weight bearing and serve as muscle origin and inser- site for the abdominal muscles and origin of the adductor
tion locations. Knowledge of this anatomy is vital when musculature. The superior and inferior pubic rami along with
evaluating the radiographic characteristics of a lesion or the pubic symphysis form the obturator foramen. This part of
fracture and guides the decision for intervention and surgical the pelvis provides for insertions of the abdominal muscula-
techniques used. No definitive classification exists for meta- ture and the origins of the adductor muscle compartment
static disease in the pelvis, however, several classification along with some of the rotators of the femur. This part of the
systems can be used to guide one in the decision of surgical pelvis is not weight supporting. It serves only for origins and
versus nonsurgical treatment. insertions of muscles. Patients can have pain from these
The sacrum is a wedge-shaped bone that acts as the “key- lesions while walking, not because the involved bone is
stone” of the pelvis and essentially connects the upper body weight supporting, but because the muscles are pulling on
through the spine to the lower body. It is secured to the hemi- weakened bone causing pain.
47 Surgical Management of Metastatic Disease to the Pelvis 517

Fig. 47.1 Nonoperative management of pelvic lesions can be recom- and axial skeleton. Below this region (outlined by white arrows) there
mended when the region of the sciatic buttress and superior acetabulum is a large lytic lesion within the posterior column of the acetabulum.
are left unaffected. The black arrows outline the anatomic location of This patient was treated nonoperatively as the lesion did not compro-
the sciatic buttress, which is the caudal portion of the iliac bone and is mise the structural integrity of the weight-bearing hemipelvis
responsible for force transmission from the acetabulum to the sacrum

An understanding of this anatomy is imperative in the ment modalities that can be implemented and the invasive-
management of patients with metastatic disease of the pelvis, ness and risk associated with these methods vary considerably.
as the first step in their treatment is the determination of It is important the patient understands their diagnosis of
whether the location of the metastatic disease compromises metastatic disease and prognosis, especially in patients with
the structural integrity of the weight-bearing portions of the advanced cancer [9, 10]. Attempts have been made to develop
pelvis. Succinctly, if the axis of force transmission from the algorithms to predict survival in patients with metastatic dis-
acetabulum into the sacrum is maintained, the patient should ease, but none have been shown to be consistently reliable
still be able to safely bear weight on the extremity without an and/or widely applicable [9–12]. Thus, this conversation pro-
unacceptable risk of fracture (Fig. 47.1). We can affirm that vides valuable information to the surgeon regarding the
lesions within the superior and inferior pubic rami, iliac patient’s insight into their own disease and acceptance of
wing (not involving the sciatic buttress) and medial and infe- their diagnosis. The goal of any intervention in the pelvis is
rior aspects of the acetabulum do not significantly impact the to relieve pain and improve the patient’s mobility. Any surgi-
structural integrity of the hemipelvis. Thus, these regions can cal procedure in the pelvis for diffuse metastatic disease
be safely considered for nonoperative management from a should allow immediate weight bearing.
structural standpoint. Radiation therapy may be an option for patients who want to
avoid surgery, or for those who are otherwise deemed to unhealthy
to undergo surgery, as long as the patient is without a displaced
47.4 Noninvasive Treatment Options fracture and is able to mobilize safely. Radiation therapy is effec-
tive in palliating pain and is used often in metastatic disease of the
When presented with a patient with pelvic metastatic disease pelvis. External beam radiation therapy (EBRT) is the typical
of bone it is important to take a multidisciplinary approach option for nonoperative treatment for painful osseous metastases
involving surgical oncology, radiation oncology, medical and is effective in 60–70% of patients. There has been much inter-
oncology, and others to create a personalized, comprehen- est in the optimal radiation regimen for these patients and, accord-
sive plan. Equally important is the need to have a discussion ing to the American Society for Radiation Oncology, all these
with the patient regarding their goals of care. This is espe- regimens are equally efficacious [13]. A single fraction of 8 Gy to
cially important in this patient cohort for several reasons. As a bony metastasis has been shown to have the same pain relief as
will be outlined in the coming pages, there are multiple treat- higher doses split up into several fractions [13, 14]. This offers a
518 T. J. Evans et al.

more convenient option for many patients; however, it is impor- able pain relief as well as progression-free survival [20].
tant to inform patients interested in a single fraction therapy that With this technique, it should be noted that the location and
they do run a higher risk of having recurrent pain in the treatment size of the lesion are of utmost importance, as the thermal
field which would require repeat radiation therapy [13]. Patients ablation occurs in a centrifugal pattern and can cause dam-
with recurrence of their pain after single fraction treatment have age to nearby neurovascular structures. The percutaneous
been found to develop recurrent pain around 3–4 months after techniques chapter covers this in much detail.
treatment [15, 16]. Interest in stereotactic body radiotherapy Cementoplasty was based on the principles used for kypho-
(SBRT) is increasing given evidence that it has similar local con- plasty in the spinal column. These procedures offer several
trol and pain relief rates when compared to EBRT [17]. benefits over open techniques. The percutaneous approach
reduces the blood loss and operative time significantly without
sacrificing considerable pain relief postoperatively [21, 22].
47.5 Minimally Invasive Treatment Options Second, the risks of wound complications or infection are
reduced due to the size of the incisions and the minimal soft
Embolization can be used as either the primary treatment tissue dissection performed throughout the procedure. This, in
modality or as a neoadjuvant treatment for certain primary theory, could facilitate quicker transitions back to medical and
cancer types in preparation for future surgical management. radiation adjuvants postoperatively. Limitations in using this
This treatment is associated with a 95–97% clinical response treatment involve the treatment of uncontained defects, which
rate and there is often healing of the bone lesions in greater could allow extravasation of cement into the surrounding soft
than 50% of the cases [18, 19]. It should be noted that, if tissues or, potentially, into the joint, neural foramen, and pel-
performed as a neoadjuvant treatment, surgery should be vic space. One study quoted a 22–28% risk of extra-osseous
performed within 48 h of embolization to ensure there is no extravasation of cement in their cohort [23]. This extravasation
revascularization of the lesion prior to surgery. can result in radiculopathy from impingement on the sciatic
Percutaneous techniques exist that can be used in loca- nerve and/or pain with ambulation if cement spreads within
tions of the pelvis that are not amenable to surgery or radia- the hip joint itself. The location of the lesions can influence the
tion. Radiofrequency ablation (RFA) and cryoablation are ease of this procedure. Sacral lesions are traditionally very dif-
forms of image-guided tumor ablation that can provide pal- ficult to treat in this way given the difficulty with obtained
liation to patients with painful pelvic metastases. Both tech- adequate imaging and the three-dimensional anatomy in this
niques are similar in the sense that they use variations in region. This method has been taken one step further when per-
temperature to cause cell death within the metastatic lesion. cutaneous screws are inserted, as previously described by
These techniques have shown efficacy in the treatment of Starr et al. [24], prior to cement insertion in periacetabular
patients with oligometastatic disease and have led to accept- defects (Fig. 47.2). The percutaneous screws are thought to

Fig. 47.2 This patient illustrates the indications for percutaneous tech- was not eligible for acetabular reconstruction with a cage construct.
niques as shown in the postoperative image. In the preoperative CT Percutaneous cannulated 6.5 mm screws and cementoplasty were uti-
scan, there is extension of the lesion cranially into the sciatic buttress lized to provide pain relief and allow early ambulation for this patient
and medial ilium; however, due to the extent of the disease this patient
47 Surgical Management of Metastatic Disease to the Pelvis 519

provide a scaffold for cementation of the defects, adding sta- tion of the hip joint and fill the weakened, lytic areas of the
bility to the construct [22]. metastatic lesions.
Reconstruction cages come in many varieties; however,
the general concept is the same throughout all options
47.6 Invasive Treatment Options (Fig. 47.3). The cage serves as a foundation on which to
reconstruct the hip joint and the cage-screw construct act as a
The Harrington classification system was first proposed in sort of rebar to strengthen the cementation of the acetabulum.
1981, not only as a means of compartmentalizing and guid- Reconstruction cages are designed to have at least one point
ing surgical management of periacetabular lesions but also as of fixation on either side of the acetabulum—both cranial and
a way of proposing novel treatment strategies for these com- caudal to the joint [26–28]. The goal of this cage construct is
plex tumors [25]. The driving force behind this classification to provide fixation into healthier bone, distributing the joint
system is the presence and quality of bone within the peri- forces to the axial skeleton. The cranial aspect often consists
acetabular region of the pelvis and how this affects the load of a cluster of screws placed through one or several flanges
transmission from the acetabulum to the axial skeleton. The extending up the iliac wing. The caudal portion consists of an
classification system is based on the integrity of the medial, ischial hook, blade, or screw-hole cluster, which are all meant
lateral, and/or superior walls. This system is still used often to gain purchase in either the ischium or around the inferior
to describe lesions and help guide treatment within this aspect of the cotyloid fossa. Rapidly drying cement is then
region of the pelvis. However, the treatment strategies have placed into the cage to anchor the construct to the remaining
continued to evolve as our understanding of the disease pro- pelvis. If needed, a small corticotomy can be made in the
cess improves. outer table of the ilium to allow for cement filling of defects
Total hip arthroplasty in metastatic disease of the peri- in the iliac bone. A cemented femoral stem is typically used.
acetabular region is a challenging procedure for several rea- Although the Harrington reconstruction techniques are
sons. First, there is often significant bone loss which leads to still used by many surgeons to this day, there has been con-
lack of adequate support for conventional total hip arthro- tinued interest in finding a solution to the loosening seen
plasty components. In addition, many of these patients have with this type of reconstruction. As such, techniques used
undergone radiation and/or chemotherapeutic treatments to manage massive acetabular bone loss in revision hip
which have had an impact on the local bone’s ability to in arthroplasty have been adapted to oncologic scenarios.
grow into the implanted prostheses, increasing the risk of Porous tantalum shells are advantageous given their high
aseptic loosening postoperatively. Considering this, it is coefficient of friction which allows for an immediate
necessary to create a construct that reestablishes (or reduces scratch fit of the implant in the remaining acetabular bone,
the need for) column support and provides articular stabil- facilitating early weight bearing for these patients. In addi-
ity. The treatment for most acetabular lesions with signifi- tion, these augments are designed to accept multiple screws
cant metastatic defects is an acetabular reconstruction that can be inserted to increase the fixation to the bone and
involving a cage. The reconstruction cage is adjusted to fit improve the compression of the implant to native bone,
the patient’s anatomy and to provide as many fixation points theoretically enhancing bone ingrowth. In fact, these tanta-
as possible. These cages prevent medial or superior migra- lum implants have shown adequate amounts of bony

Fig. 47.3 Operative treatment of pelvic lesions can be considered reconstruction with a reconstruction cage, total hip arthroplasty with a
when the morphology of the lesion is such that it impacts the anatomy dual mobility liner, and a cemented stem. This allows the transmission
responsible for force transmission and increases the risk of fracture due of the weight-bearing forces from the acetabulum to the healthier,
to these high amounts of force. This patient underwent acetabular stronger portions of the iliac bone
520 T. J. Evans et al.

ingrowth even in the setting of previously irradiated bone. ity and weight-bearing precautions postoperatively for the
This technique is often used after radiation for fracture first several months. These restrictions are meant to maintain
deformity or arthritis with defects present with no gross the position of the prosthesis while the soft tissues heal and
disease present. When compared to the Harrington recon- scar in to provide added stability prior to resumption of nor-
structions, these uncemented reconstructions achieved sim- mal activity. Additionally, these patients commonly have
ilar outcome scores while having a lower rate of neuropraxias in the perioperative period secondary to change
loosening—0% rate in one study [29]. in their limb length [33, 34]. Proximal migration of the pros-
There are several options with regards to the type of con- thesis through the ilium is commonly seen. Unfortunately,
straint used in these periacetabular reconstructions. this particular type of prosthesis has not yielded the desired
Constrained polyethylene liners are used in a patient popula- results, with most patients failing to return to ambulation
tion who is at high risk of instability of the hip postopera- with minimal or no assistance; as such has fallen out of favor.
tively. However, due to the increased restraint, the Due to the poor functional outcomes related to resec-
polyethylene–cement interface experiences high amounts of tion arthroplasty for pelvic sarcomas there has always
shear stress and, thus, component loosening may be seen been an interest in endoprosthetic reconstruction. However,
mid-to-long term. As dual mobility technology has improved early attempts at developing these megaprostheses were
there has been increased interest in these patients. The dual unsuccessful mostly due to the aseptic loosening seen sec-
mobility design provides less constraint compared to the ondary to the high shear forces experienced at the prosthe-
tripolar constrained liners; however, it is more constrained sis-native bone interface as well as the limited fixation
than primary total hip components, in a way possessing some options. With the advent of 3D printing there is renewed
of the desirable aspects of both options. Similarly, the risk of interest in endoprosthetic reconstruction as these implants
instability after implantation of dual mobility components can now be custom-­made to accommodate the patient’s
appears to be similar to constrained liners [28]. unique anatomy. Another downstream effect of 3D print-
There are unique scenarios and disease types in which a ing is the improvement in ingrowth surfaces on these
single metastatic lesion can be resected in the pelvis to ren- implants. In certain scenarios, these prostheses can be used
der the patient a status of no evidence of disease (NED). In for specific disease types that have demonstrated disease
periacetabular lesions in which a resection of the acetabulum stability over time.
is required for treatment there are several reconstructive Resection alone for palliation can be used if other options
options available. The pedestal cup (also known as the “ice are not possible. This could be due to the lack of residual
cream cone” prosthesis) is a modular design composed of a bone to allow for reconstruction. Although reconstructing
stem, which is inserted into medial ilium, and a metal cup the acetabulum is often the goal of surgery to provide a stable
that is locked into the stem with the desired orientation. The weight-bearing construct these reconstructive procedures are
medial ilium along the sciatic buttress is ideal for insertion as technically demanding, time-consuming and associated with
it has the straightest geometry of the iliac bone, thus provid- a high rate of complications. Thus, resection arthroplasty,
ing the most predictable and reproducible insertion [30]. The internal and external hemipelvectomy remain viable options
stem can be inserted in a cemented or uncemented fashion. for pain control in the setting of single-site metastatic dis-
The design of this implant is such that it recreates parts of the ease. The proximal femur is stabilized to the remaining
resected sciatic buttress, facilitating native force transfer hemipelvis to create a pseudarthrosis, as a “flail limb.” This
from the hip joint into the medial pelvis and sacrum. An type of reconstruction does require enough remaining iliac
advantage lies in the fact that the cup can be adjusted to the bone in order to provide a stable platform to which the proxi-
desired version at the time of the surgery, helping the sur- mal femur will articulate. It also requires preoperative educa-
geon take measures to reduce the likelihood of postoperative tion of the patient regarding limitations of the extremity as
instability. Having said that, instability remains the most well as an expected limb length discrepancy
common complication of these implants, with dislocation postoperatively.
rates ranging from 15 to 39% with the majority of these dis- In a systematic review performed by Brown et al. [35],
locations occurring within the first year of the initial surgery they compared many of the reconstructive options. Their
[30, 31]. However, with the advent of dual mobility hip com- study showed that there was a 50% risk of complications
ponents, the dislocation rates have decreased significantly overall, with deep prosthetic infection and joint instability
[31, 32]. Saddle prostheses were designed to provide a quick being most common in all techniques. In addition, the
reconstruction option that allows for tensioning of the hip improvement in the MSTS score varied from 62 to 72%;
musculature and restoration of limb length. One disadvan- however, all techniques did show a sustained improvement in
tage of this prosthesis is that these patients have strict activ- pain and function at final follow-up.
47 Surgical Management of Metastatic Disease to the Pelvis 521

47.7 Conclusion 7. Ebraheim NA, et al. Anatomic consideration in the anterior approach
to the sacro-iliac joint. Spine (Phila Pa 1976). 1994;19(6):721–5.
Print.
• When there is a concern for metastatic disease of the pel- 8. Browner BD. Skeletal trauma: basic science, management, and
vis, the appropriate prognosis, imaging, and goals of care reconstruction. 6th ed. St. Louis, MO: Elsevier; 2019. Print.
should be determined prior to providing possible treat- 9. Kirkinis MN, et al. Metastatic bone disease: a review of survival,
prognostic factors and outcomes following surgical treatment of the
ment recommendations. appendicular skeleton. Eur J Surg Oncol. 2016;42(12):1787–97.
• Location of the metastatic lesions within the pelvis, and Print.
how these lesions compromise the structural integrity of 10. Kirkinis MN, et al. Metastatic bone disease of the pelvis and
the acetabulum and pelvic ring during weight bearing, extremities: rationalizing orthopaedic treatment. ANZ J Surg.
2017;87(11):940–4. Print.
will direct which treatment modalities are possible. 11. Chow E, et al. Validation of a predictive model for survival in
• There are nonoperative and operative treatment options patients with advanced cancer: secondary analysis of rtog 9714.
for each patient and the outcomes of each of these options World J Oncol. 2011;2(4):181–90. Print.
should be provided to help with patient education and 12. Paulino Pereira NR, et al. Development of a prognostic survival
algorithm for patients with metastatic spine disease. J Bone Joint
decision-making. Surg Am. 2016;98(21):1767–76. Print.
13. Lutz S, et al. Palliative radiation therapy for bone metastases:
update of an astro evidence-based guideline. Pract Radiat Oncol.
2017;7(1):4–12. Print.
Open Questions 14. Rich SE, et al. Update of the systematic review of palliative radia-
Future research in the area of surgical management of tion therapy fractionation for bone metastases. Radiother Oncol.
metastatic bone disease of the pelvis should address 2018;126(3):547–57. Print.
15. van der Linden YM, et al. Single fraction radiotherapy is effica-
the following questions:What is the quantitative limit cious: a further analysis of the dutch bone metastasis study control-
of nonoperative management of metastatic disease of ling for the influence of retreatment. Int J Radiat Oncol Biol Phys.
the pelvis? 2004;59(2):528–37. Print.
• What are the anatomic characteristics that necessi- 16. van der Velden JM, et al. Evaluation of effectiveness of palliative
radiotherapy for bone metastases: a prospective cohort study. J
tate the need for operative intervention? Radiat Oncol. 2018;7(4):325–33. Print.
• Likewise, what anatomic characteristics correlate 17. Spencer KL, et al. Systematic review of the role of stereo-
with the best postoperative outcomes of each of the tactic radiotherapy for bone metastases. J Natl Cancer Inst.
above-­mentioned surgical management options? 2019;111(10):1023–32. Print.
18. Rossi G, et al. Embolisation of bone metastases from renal cancer.
• What is the role of antiresorptive medication in the Radiol Med. 2013;118(2):291–302. Print.
management of patients with metastatic disease of 19. Rossi G, et al. Selective arterial embolisation for bone tumours:
the pelvis? experience of 454 cases. Radiol Med. 2011;116(5):793–808. Print.
• In the event of nonoperative management of these 20. Cazzato RL, et al. Complications following percutaneous image-­
guided radiofrequency ablation of bone tumors: a 10-year dual-­
patients, what is the appropriate surveillance center experience. Radiology. 2020;296(1):227–35. Print.
timeline? 21. Durfee RA, et al. Percutaneous acetabuloplasty for metastatic
lesions to the pelvis. Orthopedics. 2017;40(1):e170–e75. Print.
22. Yang R, et al. A novel tripod percutaneous reconstruction technique
in periacetabular lesions caused by metastatic cancer. J Bone Joint
References Surg Am. 2020;102(7):592–9. Print.
23. Weill A, Kobaiter H, Chiras J. Acetabulum malignancies: technique
1. Onken JS, et al. Metastatic dissemination patterns of different pri- and impact on pain of percutaneous injection of acrylic surgical
mary tumors to the spine and other bones. Clin Exp Metastasis. cement. Eur Radiol. 1998;8(1):123–9. Print.
2019;36(6):493–8. Print. 24. Starr AJ, Reinert CM, Jones AL. Percutaneous fixation of the
2. Lievens Y, et al. Defining oligometastatic disease from a radia- columns of the acetabulum: a new technique. J Orthop Trauma.
tion oncology perspective: an estro-astro consensus document. 1998;12(1):51–8. Print.
Radiother Oncol. 2020;148:157–66. Print. 25. Harrington KD. The management of acetabular insufficiency sec-
3. Lu X, et al. Oligometastatic state predicts a favorable outcome for ondary to metastatic malignant disease. J Bone Joint Surg Am.
renal cell carcinoma patients with bone metastasis under the treat- 1981;63(4):653–64. Print.
ment of sunitinib. Oncotarget. 2016;7(18):26879–87. Print. 26. Assi C, et al. The cross technique for the positioning of kerboull
4. Yin Z, et al. Impact of sites versus number of metastases on survival plate in acetabular reconstruction surgery. SICOT J. 2018;4:20.
of patients with organ metastasis from newly diagnosed cervical Print.
cancer. Cancer Manag Res. 2019;11:7759–66. Print. 27. Rowell P, et al. Is an acetabular cage and cement fixation suffi-
5. Basu D, et al. Detection of occult bone metastases from head and ciently durable for the treatment of destructive acetabular metasta-
neck squamous cell carcinoma: impact of positron emission tomog- ses? Clin Orthop Relat Res. 2019;477(6):1459–65. Print.
raphy computed tomography with fluorodeoxyglucose F 18. Arch 28. Wegrzyn J, et al. Total hip arthroplasty for periacetabular meta-
Otolaryngol Head Neck Surg. 2007;133(8):801–5. Print. static disease. an original technique of reconstruction according to
6. Bhandari V. Incidence of bone metastasis in carcinoma buccal the harrington classification. J Arthroplast. 2018;33(8):2546–55.
mucosa. Indian J Med Paediatr Oncol. 2016;37(2):70–3. Print. Print.
522 T. J. Evans et al.

29. Houdek MT, et al. Comparison of porous tantalum acetabular 33. Aljassir F, et al. Outcome after pelvic sarcoma resection
implants and harrington reconstruction for metastatic disease of the reconstructed with saddle prosthesis. Clin Orthop Relat Res.
acetabulum. J Bone Joint Surg Am. 2020;102(14):1239–47. Print. 2005;438:36–41. Print.
30. Bus MP, et al. Clinical outcome of pedestal cup endoprosthetic 34. Renard AJ, et al. The saddle prosthesis in pelvic primary and second-
reconstruction after resection of a peri-acetabular tumour. Bone ary musculoskeletal tumors: functional results at several postoperative
Joint J. 2014;96-B(12):1706–12. Print. intervals. Arch Orthop Trauma Surg. 2000;120(3–4):188–94. Print.
31. Bus MP, et al. Lumic((R)) endoprosthetic reconstruction after peri- 35. Brown TS, et al. Reconstruction of the hip after resection of peri-
acetabular tumor resection: short-term results. Clin Orthop Relat acetabular oncological lesions: a systematic review. Bone Joint J.
Res. 2017;475(3):686–95. Print. 2018;100-B(1 Supple A):22–30. Print.
32. Issa SP, et al. Pelvic reconstructions following peri-acetabular bone
tumour resections using a cementless ice-cream cone prosthesis
with dual mobility cup. Int Orthop. 2018;42(8):1987–97. Print.
 Surgical Management of Metastatic
Disease to the Spine 48
Christopher P. Wang, Amanda Brisco, and James K. C. Liu

Abstract
Learning Objectives
The surgical management of spinal metastases has
unique challenges not encountered in other aspects of • Readers will understand the indications and goals
spine surgery due to the needs of the oncologic patient. of surgical treatment for spinal metastases, includ-
Among the key steps in the surgical management is the ing strategies to assist in optimal surgical
decision-­making process, which must take into account intervention.
the systemic and adjuvant therapies that are also neces- • Readers will be able to understand the different sur-
sary for systemic and local control of the disease. gical approaches that can be employed for the surgi-
Decision-­making frameworks, such as the NOMS sys- cal management of metastatic disease to the spine.
tem, have been designed to take into account the multi- • Readers will develop an understanding of the appli-
ple factors that can affect the prognosis of the oncologic cation of radiation as an important adjunct to the
patient and reach an optimal surgical plan that incorpo- surgical management of spine metastases.
rates available adjuvant treatment, such as radiation
therapy. Surgical intervention for spinal metastases has
evolved from large reconstructive procedures to less
invasive techniques, which still allow for the preserva- 48.1 Introduction
tion of neural function and restoration of spinal stability,
but with the strategy of selecting the optimal interven- Metastatic spine tumors are the most common type of spinal
tion which can best synergize with available adjuvant column tumors [1]. The malignancies that most commonly
therapies for effective local disease control. Advancement develop metastases to the spine are cancers of the lung, breast,
of surgical strategies such as “separation surgery” and and kidney. Hematological malignancies can spread to the
minimally invasive percutaneous stabilization tech- spine and are often considered under the category of spinal
niques can achieve surgical goals with decreased mor- metastases and these are predominantly composed of multiple
bidity and shortened recovery times, which allows for myeloma and lymphoma [2]. The surgical treatment of spine
minimal disruption of adjuvant or systemic therapies. metastases requires an understanding of the unique needs of
The development of conformal radiation techniques the oncologic patient, which must be taken into account when
such as radiosurgery has greatly assisted in achieving determining the optimal surgical treatment and approach.
these goals. There are a variety of surgical techniques that can be tailored
to provide the most effective treatment strategy with minimal
disruption to the overall goal of continuation of systemic ther-
apy, including minimally invasive strategies as well as percu-
taneous stabilization, ablative, and radiation techniques. Here
we will outline the surgical approach to the metastatic spine
C. P. Wang · A. Brisco · J. K. C. Liu (*)
H. Lee Moffitt Cancer Center & Research Institute,
patient, from individualizing decision-making to various sur-
Tampa, FL, USA gical strategies and how each may be optimally used to treat
e-mail: [email protected]; [email protected]; [email protected] metastatic spine disease.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 523
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_48
524 C. P. Wang et al.

48.2 Surgical Decision-Making acterize the degree of spinal cord compression, the Spinal
Oncology Study Group (SOSG) devised the Epidural Spinal
Among the unique challenges that are encountered in the sur- Cord Compression (ESCC) grading scale, which provides
gical management of spinal metastases is selecting the inter- a 6-point grading system that describes the extent of spinal
vention that best incorporates into the treatment plan for the cord compression [7]. This scale quantifies the degree of
oncologic patient. Preserving neurologic function through epidural extension of tumor from isolation only within the
decompression of neural elements and achieving stability vertebral body, to compression on the spinal cord (Fig. 48.1).
remains the primary goal for surgical management. The type Grade 0 indicates disease confined to the bone. Grade 1 indi-
of surgical intervention that is employed to achieve those cates the presence of disease in the epidural space and is sub-
goals must take into account factors that affect the oncologic categorized in 1a–1c. 1a indicates epidural disease without
patient, including the state of systemic disease, prognosis, impingement on the thecal sac, 1b indicates the presence of
ongoing therapies, and options of adjuvant therapy that may compression of the thecal sac without compression of the
assist in the treatment of the metastatic disease. Several clini- spinal cord, and 1c indicates thecal sac compression and
cal decision-making tools have been established to provide abutment of the spinal cord without compression. Grade 2
guidelines in regard to surgical management of metastatic indicates compression of the spinal cord with CSF visible
spine disease. Early scoring systems provided a framework around the spinal cord and Grade 3 indicates compression
to suggest the degree of surgical intervention based on the of the spinal cord without evidence of surrounding CSF. The
aggressiveness of underlying malignancy. The modified ESCC provides a universal scale for assessing the degree of
Bauer scoring system provided a framework for surgical neurological compromise present, which may indicate pres-
decision-making based on the presence of visceral metasta- ent or potential neurologic deficits, but also dictates radia-
sis [3]. The Tomita scoring system considered the type of tion options. Clinical evidence of neural compromise also
primary malignancy, as well as the presence of visceral and contributes to surgical decision-making. This includes spi-
bony metastases [4]. The revised Tokuhashi scoring system nal cord compression resulting in myelopathy, manifesting
was similar to the Tomita system, but added number of ver- as hyperreflexia and clonus, to loss of sensation, bowel or
tebral body metastases, performance status, and neurological bladder dysfunction, and motor weakness. Compression of
deficits as factors to consider [5]. Based on the extent of the nerve roots may manifest as radiculopathy along a given
disease, these scoring systems would suggest palliative treat- dermatomal distribution, but also may be present only upon
ments that may not include surgery, to invasive excisional axial loading and is consistent with mechanical radiculopa-
surgery that may involve extensive tumor resection. The evo- thy, indicating potential instability of the spine. Together, the
lution of these scoring systems underscores the importance radiographic and clinical evidence of the “neurologic” status
of understanding the natural course of the primary malig- of the patient determines the potential need and urgency for
nancy and its potential response to systemic therapy, as well decompression of the neural elements.
as the extent of metastasis present to determine whether a “Oncologic,” is defined by the tumor histology of the
complex surgical intervention is indicated and beneficial to primary malignancy. This provides insight into the overall
the patient. prognosis of the patient as well as the responsiveness to radi-
ation. Radiation therapy has become an important adjuvant
therapy in the local treatment of spinal metastases but not all
48.2.1 NOMS pathologies respond equally [8]. Tumor pathologies can be
divided in their response to radiation therapy by radiosen-
Previous treatment decision tools were based on assessing sitive or radioresistant tumors. Tumors that are considered
how surgery alone can play a role in local management of radiosensitive, including lymphoma, myeloma, seminoma,
metastatic spine disease. The emergence of adjuvant thera- breast cancer, and prostate cancer, demonstrate good over-
pies, particularly stereotactic radiation therapy, has created all control to conventional external beam radiation therapy
additional options for local control of metastatic spine d­ isease (cEBRT), which consists of non-conformal radiation applied
and therefore shifted the approach for surgical management. to a general area with low risk of toxicity to adjacent neu-
Laufer et al. established the NOMS system to provide ral structures [9]. Many solid malignancies such as renal,
a clinical treatment framework for treating spinal metasta- gastrointestinal, non-small cell lung, thyroid, and sarcomas,
ses based on the four categories: Neurologic, Oncologic, demonstrate poor response to cEBRT [8]. These pathologies
Mechanical, and Systemic (NOMS) [6] (Fig. 48.1). require conformal radiation in the forms of hypo-­fractionated
“Neurologic,” refers to the degree of neurological compro- radiation therapy or stereotactic radiosurgery [10]. An under-
mise present, as measured radiographically by the amount of standing of the tumor histology and its responsiveness to
spinal cord compression by the tumor, as well as neurologi- radiation therapy will help to optimize surgical intervention
cal deficits present on physical examination. To help char- so that it can work with the radiation therapy available.
48 Surgical Management of Metastatic Disease to the Spine 525

a b c

d e

Fig. 48.1 (a, b) T2-weighted MRI demonstrating grade 3 epidural spi- mentation two levels above and below the level of metastasis. (d, e)
nal cord compression. (c) Intraoperative view of separation surgery, Postoperative X-rays demonstrating posterolateral instrumented fixa-
with the removal of the left facet joint and pedicle, demonstrating cir- tion followed decompression via separation surgery
cumferential decompression of the thecal sac and posterolateral instru-

“Mechanical” refers to the stability of the spine at the tion for intervention. To determine the level of instability and
level of disease and serves as an indication for surgical inter- indication for intervention, the Spinal Instability Neoplastic
vention. It is assessed based on the clinical assessment of Score (SINS) classification system was derived to provide a
mechanical and postural pain. Restoration of stability is one framework as to the degree of instability secondary to tumor
of the main goals in spine surgery and is a primary indica- involvement in the spine. It is composed of six elements:
526 C. P. Wang et al.

location of the lesion, mechanical/postural pain, type of bone majority of metastatic disease arises from the vertebral body
lesion, radiographic spinal alignment, vertebral body col- resulting in anterior neural compression, therefore posterior
lapse, and posterolateral involvement of the spinal elements. decompression alone may not provide adequate decom-
Each element is assigned a score, then tallied to provide a pression of the neural elements and would provide little in
total SINS score, which categorizes the spine as stable, overall disease control. These studies resulted in a shift of
unstable, or of intermediate instability [11]. In general, the the treatment paradigm away from surgical decompression
presence of mechanical pain represents spinal instability that to radiation alone. This approach remained the standard
likely requires stabilization. These can manifest as back pain until Patchell et al. demonstrated in 2005 that direct surgi-
on axial loading and movement and is relieved with rest. In cal decompression in addition to radiation therapy provided
addition to mechanical back pain, tumor involvement in the improvement in ambulatory status, steroid use, pain score,
posterior elements may manifest as pain when lying flat due and overall survival compared to radiation alone [15]. The
to loading of the facet joints, which may also be an indica- benefit of surgery in addition to radiation noted in this study
tion for stabilization. was likely due to the use of circumferential decompression,
“Systemic” characterizes the disease burden and related which was made possible due to the emergence of spinal
medical comorbidities and serves to assess a patient’s prog- instrumentation that allowed for the reconstruction of the
nosis and fitness for treatment [12]. Disease burden is deter- spine, which started in the mid-1980s [16].
mined by tumor staging, and the prognosis is estimated using With the ability to reconstruct the spinal column, more
a patient’s functional status, extent of disease, and options aggressive surgical approaches were taken to provide long-­
for systemic treatment. term local control for metastatic spine disease. Prognostic
Taken together, the NOMS framework helps to direct scoring systems such as the Tomita and Tokuhashi systems
treatment planning based using the optimal combination supported the use of en bloc tumor resections for certain
of therapies. For example, for radiosensitive tumors, such types of oligometastatic disease to achieve optimal local
as lymphoma, myeloma, and prostate cancer, conventional control [4, 5]. The challenges of en bloc resection are that
radiation therapy alone may be the ideal treatment regard- it is difficult to achieve in the spine given the extent of dis-
less of neurologic compromise or ESCC, given the rapid ease and proximity to neural structures, and that it is also
response of the tumor to conventional radiation therapy. For technically demanding and can involve significant morbidity,
radioresistant tumors, high-dose stereotactic radiation is which may not be well tolerated by oncologic patients [17].
needed to achieve local tumor control. Therefore, the type Along with the risk for distance recurrence in metastatic dis-
of surgical intervention will depend on the need for neural ease, en bloc resection may not prove to be worthwhile in the
decompression, based on the degree of ESCC or neurologi- long term [18].
cal symptoms, as well as the stability of the spine, measured An alternative to en bloc or marginal tumor resections
by the SINS score. In grade 0 or 1 ESCC, surgery is nec- requiring circumferential reconstruction of the spine for
essary only if there is evidence of instability. For grade 2 metastatic spine disease is “separation surgery,” which is a
or 3 ESCC, surgical decompression is necessary to provide surgical approach with the intent of neurological preserva-
adequate clearance of tumor from the spinal cord so that the tion, the establishment of spinal stability, and preparation for
appropriate radiation dosage can be applied without risk of adjunctive therapy. Instead of attempting to achieve marginal
toxicity to the spinal cord. The NOMS framework empha- resections, separation surgery involves decompression of the
sizes the importance of the histology of the tumor and its spinal cord to provide a 2–3 mm clearance of tumor away
responsiveness to radiation to determine whether surgery is from the spinal cord so that an adequate radiation dosage can
necessary for local control. be applied to the tumor without a risk of neural toxicity [10,
19]. This is achieved primarily through a posterior approach
to the spine, followed by a laminectomy to decompress the
48.3 Open Surgical Approaches posterior aspect of the thecal sac. Further decompression of
the lateral and anterior aspect of the spinal cord is performed
Surgery for spinal metastases in its infancy consisted pri- through unilateral or bilateral facetectomies, resection of the
marily of posterior decompression. This was prior to the pedicles, and sectioning of the anterior longitudinal ligament
availability of spinal instrumentation, which allowed for cir- to resect tumor adjacent to the spinal cord. Removal of the
cumferential decompression and reconstruction of the spine. facet joints and pedicles allows access to the anterior ver-
Several studies examined laminectomy for epidural spinal tebral body through a posterolateral approach, avoiding the
cord compression and found that decompressive laminectomy need for a direct lateral or anterior approach, which involves
in addition to conventional external beam radiation (cEBRT) much greater morbidity, while reducing surgical time and
did not demonstrate an advantage compared to radiotherapy blood loss. Removal of the facet joints requires instrumented
alone [13, 14]. These results were unsurprising given that the stabilization, often through pedicle screw fixation that
48 Surgical Management of Metastatic Disease to the Spine 527

extends two vertebral body levels above and below the level of percutaneous fixation is the lack of a long midline inci-
of decompression (Fig. 48.1). In cases of significant loss of sion with disruption of the paraspinal musculature, which
bony integrity in the anterior column secondary to erosion by allows for rapid restoration of systemic therapy or appli-
tumor involvement, the anterior column support can be aug- cation of local radiation therapy with little risk for wound
mented with placement of polymethylmethacrylate from a complications [26] (Fig. 48.2).
posterior approach or placement of an expandable cage from A hybrid approach consisting of percutaneous fixation
through a transpedicular or costotransversectomy approach, combined with “mini-open” laminectomies or unilateral
depending on the level and degree of bony destruction [20]. facetectomies can also be utilized when focused neural
decompression is also indicated [27]. A small midline or
paramedian incision can be utilized with or without tubular
48.4 Minimally Invasive Approaches retractors to allow for focused neural decompression, which
still prevents significant paraspinal injury compared to tradi-
The development of tubular retractor systems and percutane- tional open techniques.
ous screw placement techniques has led to an exponential
increase in the adoption of minimally invasive techniques
for the treatment of spinal pathologies. Smaller incisions 48.4.2 Percutaneous Cement Augmentation
with reduced degree of paraspinal muscle destruction have
allowed placement of instrumentation into the spine with Metastatic spinal disease contained primarily within the
decreased length of hospital stays and faster recovery follow- anterior vertebral body resulting in mechanical pain may
ing surgery compared to traditional open approaches [21]. be treated with cement augmentation. Cement augmenta-
By taking advantage of the effectiveness of radiation therapy, tion was originally applied to the spine for the treatment of
minimally invasive approaches can effectively achieve surgi- painful hemangiomas and then became widely used primar-
cal goals of care with minimal interruption of adjuvant and ily for osteoporotic compression fractures [28, 29]. Cement
systemic therapies. augmentation can also be effectively applied for the treat-
ment of pathologic vertebral compression fractures, or loss
of bone quality secondary to lytic metastatic disease. The
48.4.1 Percutaneous Pedicle Screw Fixation Cancer Patient Fracture Evaluation (CAFE) study was a
randomized controlled trial that involved 22 international
Percutaneous pedicle screw fixation can serve as an effec- sites, which compared balloon kyphoplasty to nonsurgical
tive alternative to open fixation techniques for unstable frac- fracture management of pathologic compression fractures
tures secondary to metastatic disease without the need for [30]. The study demonstrated definitively that kyphoplasty
neural decompression. Percutaneous fixation has been asso- provided rapid pain reduction and improvement in patient
ciated with lower postoperative pain scores and faster recov- function with low rates of adverse events. Subsequent stud-
ery time compared to open approaches [22, 23]. They allow ies have demonstrated the value of cement augmentation for
for stabilization of pathology affecting both the anterior malignant compression fractures resulting in improvement
and posterior columns, without disruption of the paraspinal in functional status with minimal risk [31].
musculature necessary in an open posterior approach, and The indication for cement augmentation is localized pain
therefore preserving the posterior tension brand. This allows secondary to metastatic disease secondary involving the
for placement of shorter segmental constructs achieving sta- anterior vertebral body without significant involvement of
bilization through fixation one level above and below the the posterior column. In addition to proper localization to
diseased level compared to conventional fixation strate- isolate the correct level as the pain generator, the morphol-
gies which necessitate longer constructs [24]. An adjunct ogy of the vertebral body dictates whether successful cement
to percutaneous fixation is the use of fenestrated pedicle augmentation can be performed. Relative contraindications
screws which allow for cement augmentation of the screw include disruptions in the cortical walls of the vertebral body,
tips within the vertebral bodies [25]. Patients with patho- particularly the posterior wall, which may lead to the extrav-
logical fractures secondary to metastatic spine disease often asation of cement into the spinal canal [32]. Augmentation
have multiple levels of vertebral body disease involvement can be performed in the presence of cortical wall breaches
preventing optimal screw purchase, necessitating longer in experienced hands by adjusting needle placement to mini-
constructs to obtain more points of fixation to decrease the mize extravasation. Although primarily performed in the
risk of hardware pullout. The development of fenestrated thoracic and lumbar levels, through a posterior transpedic-
screw systems which allow for cement augmentation allows ular approach, augmentation of the cervical levels through
improved bony purchase and short segment constructs even an open anterior approach can also be performed [33]. The
in patients with poor bone quality. A significant advantage most common form of cement used is polymethylmethac-
528 C. P. Wang et al.

a b c

d e f

Fig. 48.2 (a, b) Postoperative X-rays demonstrating percutaneously recess of the thecal sac and involvement of the left pedicle causing
placed pedicle screw fixation with cement augmentation of the pedicle instability and nerve root compression. (e, f) Postoperative X-rays fol-
screws and kyphoplasty at the level of metastatic disease. (c, d) lowing percutaneously placed pedicle screws and a mini-open laminec-
T2-weighted MRI demonstrated epidural tumor involves the left lateral tomy and a left facetectomy to decompress the exiting nerve root

rylate (PMMA), an acrylic bone cement. Polymerization of improvements [36]. An alternative to cavity creation with
PMMA results in an exothermic reaction within the vertebral a balloon tamp is the use of a directional osteotome, which
body which may also contribute to pain relief secondary to creates tracts within the vertebral body allowing cement to
necrosis of tumor tissue [34]. fill across the vertebral body through a unilateral approach
Percutaneous cement augmentation can be performed [37] (Fig. 48.3).
in the form of vertebroplasty or kyphoplasty. In vertebro- Despite being a percutaneous procedure, there is the
plasty, cement is directly injected into the vertebral body potential for significant complications. Cement extravasa-
under fluoroscopic visualization. In kyphoplasty, a balloon tion can occur in the presence of a cortical breach. Cement
tamp is inserted into the vertebral body and inflated to cre- extravasation rates are higher for pathologic fractures com-
ate a cavity, followed by injection of cement. The intention pared to osteoporotic fractures, likely due to the lytic nature
of the cavity creation is to allow for an area of low-pressure of the metastatic disease. Leakage rates can be as high as
cement filling and for potential expansion of the vertebral 37.9% and 13.6% in vertebroplasty and kyphoplasty for
body endplates for restoration of height [35]. Studies have metastatic spine disease, respectively, although symptomatic
demonstrated a slight advantage in kyphoplasty in terms of leaks are typically around 1% [31]. Cement extravasation
reduced rates of cement extravasation and improvement in into the vasculature is a rare but potentially devastating com-
the restoration of vertebral body height, although there was plication of cement augmentation. Most vascular extravasa-
no significant difference noted in pain or disability score tions are asymptomatic and remain in the distal vasculature
48 Surgical Management of Metastatic Disease to the Spine 529

a b c d

e f g h

Fig. 48.3 (a–d) Intraoperative fluoroscopy imaging demonstrating strating PMMA cement after kyphoplasty. (g, h) Intraoperative fluoros-
cannulation of the vertebral body with a Jamishidi needle (a, b), fol- copy imaging demonstrating expansion of a directional balloon tamp in
lowed by extention of a directional osteotome to create a tract to the the vertebral body for midline cavity creation prior to cement
contralateral vertebral body (d, e). (e, f) Postoperative CT scan demon- placement

near the vertebral bodies, but pulmonary embolism and car- in these areas is a strong indication that instrumentation
diac thrombus have been reported [38–40]. is required for stabilization of the posterior column [41].
Another indication for instrumented fixation is the presence
of mechanical radiculopathy, which is radiculopathy pres-
48.4.3 Selection of Percutaneous Techniques ent upon weight bearing and relieved with rest. This occurs
due to nerve root compression secondary to instability in
The introduction of minimally invasive techniques has the posterior elements as axial loading is applied and can
allowed spinal stabilization of metastatic spine disease often be alleviated with stabilization.
and symptomatic relief through percutaneous strategies.
Among the challenges is determining the optimal percu-
taneous strategy and whether cement augmentation alone 48.5 Percutaneous Ablative Procedures
will provide adequate stability and pain relief, or whether
the addition of percutaneous pedicle screw fixation is 48.5.1 Radiofrequency Ablation
needed. Although one of the components of the SINS score
is the presence of unilateral or bilateral involvement of the Radiofrequency ablation (RFA) is a percutaneous technique
posterior elements, it does not specifically address what that uses high-frequency alternative current to generate heat
comprises of tumor involvement. The presence of tumor to damage tumor cells at a temperature of 60–100 °C [42].
involvement visualized as signal changes in the posterior This can be used along or in conjunction with cement aug-
elements on MRI may not necessarily correlate with loss mentation, prior to cement delivery [43]. RFA is primar-
of bony integrity resulting in instability. CT imaging is ily used for the treatment of osteolytic lesions as sclerotic
therefore critical in properly assessing the degree of bony lesions possess higher levels of impedance which prevents
destruction of the posterior elements, particularly within reaching cytotoxic temperatures [44]. RFA combined with
the pedicles and facet joints. The presence of lytic lesions cement augmentation has demonstrated an improvement in
530 C. P. Wang et al.

pain as well as disability score, with local tumor control of field radiation in a fractionated dose, commonly 30 Gray in
up to 70% [43, 45]. 10 fractions. Fractionation allows tissue to repair after each
session, allowing for recovery from post-procedural toxic-
ity and a greater cumulative dose delivery. cEBRT delivers
48.5.2 Cryoablation radiation to a large nonspecific field, often encompassing
vertebral body levels above and below the intended target, as
Similar to RFA, cryoblation is a percutaneous technique well as surrounding normal tissue. The presence of the spinal
which results in cytotoxic tumor effect although through cord and other adjacent organs at risk limits the radiation
cold temperatures. Using liquid argon, a cryoprobe that is dosage. Studies have demonstrated that responsiveness to
placed into the tumor undergoes a rapid freeze, thaw, and cEBRT depends on the histology of the tumor being treated,
freeze cycle, which results in the formation of an enlarg- with lymphoma, seminoma, myeloma, and prostate demon-
ing ice ball. This results in the creation of an osmotic gra- strating the greatest radiosensitivity, and the majority of solid
dient which leads to cell membrane injury and death [42]. tumors demonstrating varying degrees of radioresistance [8].
Cryoablation is preferred for metastatic tumors with large The need to overcome this clinical challenge led to the
soft tissue components and preferred over RFA for osteo- emergence of stereotactic radiosurgery (SRS) , which allows
blastic lesions. for the delivery of high dosage radiation in a tightly con-
formal three-dimensional area in the brain or spine. In the
spine, rigid immobilization is important to allow for precise
48.5.3 Laser Interstitial Thermal Therapy dose delivery and often requires the placement of the patient
in an immobilizing frame. Once the radiation delivery path
Laser interstitial thermal therapy (LITT) is a technique of has been simulated, the machine can deliver ionizing pho-
percutaneous thermal ablation initially developed for the ton radiation in a three-dimensional plane to within 1–2 mm
treatment of malignant brain tumors [46]. This therapy has of the calculated margin. While extremely accurate, the
evolved to include metastatic tumors of the brain as well as resultant small overspill margin puts the spinal cord a risk
the spine [47]. The LITT system uses a nanometer wave- for radiation-induced myelopathy, and therefore separation
length diode that emits heat circumferentially or direc- surgery is needed to provide adequate clearance of critical
tionally, inducing DNA damage and tissue necrosis in the neural structures [49]. SRS has been shown to have much
applied area. Thermal ablation is applied, with multiple better local control rates compared to EBRT for metastatic
cycles of heat application under direct visualization with epidural spinal cord compression. Single high-dose SRS
MRI thermal imaging to track temperature changes within administration of 24 Gy has reported 1-year local control
the target region. LITT has been used as an alternative to rates of approximately 95% regardless of tumor histology
SRS in patients with epidural spinal cord compression. A [50]. Lower dose rates, such as 14–18 Gy, have a reported
recent review from a single cancer center demonstrated an local control rate of 90% [51]. The 1-year local control rate
8-month local tumor control rate of 81.7% with LITT for with EBRT has ranged from 61 to 86%, but can be as low as
varying grades of epidural spinal cord compression in a 46% for radioresistant solid tumors [52]. Similar to SRS is
cohort of 120 patients [48]. fractionated stereotactic radiotherapy (FSRT), which utilizes
the volumetric specificity of SRS to deliver contoured doses
over multiple fractions (24–30 Gy in 3–5 fractions). The
48.6 Radiation Therapy benefit of FSRT is the high dosage delivery with a decreased
risk of injuring adjacent structures. High-dose FSRT is asso-
Although not a form of surgical intervention, radiation therapy ciated with a 4% local progression rate after 1 year, com-
has nonetheless become an important interventional adjunct pared with 22% in low-dose hypo-fractionated SRS across
and has significantly shaped the surgical strategies employed all tumor histologies [10].
against metastatic spine disease and therefore deserves men-
tion here. As noted previously, the understanding of sensitiv-
ity to radiation therapy of different tumor histologies and the 48.6.1 Intraoperative Radiation Therapy
emergence of conformal radiation in the form of stereotactic
radiosurgery, has allowed the evolution of surgical interven- An alternative or supplement to traditional external radia-
tions to be tailored toward neural decompression and spinal tion therapy is the use of intraoperative radiation therapy
stabilization through more minimally invasive approaches or brachytherapy to the tumor itself or to peri-tumoral
which are better suited for oncologic patients. structures. These procedures are typically performed in
Traditional radiation therapy involves conventional exter- the operating room in conjunction with open procedures
nal beam radiation therapy (cEBRT), which delivers wide-­ and allow for more direct application of radiation to mini-
48 Surgical Management of Metastatic Disease to the Spine 531

mize toxicity to nearby critical structures. One type is References

plaque brachytherapy, which has been utilized with 192Ir-
and 90Y-based sources with a polycarbonate backing [53]. 1. Bollen L, et al. Prognostic factors associated with survival in
patients with symptomatic spinal bone metastases: a retrospective
Folkert et al. used a more flexible silicone-coated plaque cohort study of 1 043 patients. Neuro-Oncology. 2014;16(7):991–8.
chemically bound with 32P [54]. The plaques were applied 2. Groenen KHJ, et al. The Dutch national guideline on metastases
to the dural surface following separation surgery to reduce and hematological malignancies localized within the spine; a mul-
local recurrence. The authors noted that 32P plaques could tidisciplinary collaboration towards timely and proactive manage-
ment. Cancer Treat Rev. 2018;69:29–38.
directly deliver radiation without risk to nearby critical 3. Leithner A, et al. Predictive value of seven preoperative prog-
structures and could serve as an effective adjunct to radia- nostic scoring systems for spinal metastases. Eur Spine J.
tion therapy in patient who has already received failed pre- 2008;17(11):1488–95.
vious radiation. 4. Tomita K, et al. Surgical strategy for spinal metastases. Spine (Phila
Pa 1976). 2001;26(3):298–306.
More recently percutaneous intraoperative radiation ther- 5. Tokuhashi Y, et al. A revised scoring system for preoperative evalu-
apy combined with kyphoplasty has been tested for localized ation of metastatic spine tumor prognosis. Spine (Phila Pa 1976).
delivery of radiation therapy along with cement augmenta- 2005;30(19):2186–91.
tion (Kypho-IORT). A phase I/II trial demonstrated improve- 6. Laufer I, et al. The NOMS framework: approach to the treatment of
spinal metastatic tumors. Oncologist. 2013;18(6):744–51.
ment in pain score along with a >90% local progression-free 7. Bilsky MH, et al. Reliability analysis of the epidural spinal cord
survival [55]. compression scale. J Neurosurg Spine. 2010;13(3):324–8.
8. Gerszten PC, Mendel E, Yamada Y. Radiotherapy and radiosurgery
for metastatic spine disease: what are the options, indications, and
outcomes? Spine (Phila Pa 1976). 2009;34(22 Suppl):S78–92.
48.7 Conclusion and Perspective 9. Lovelock DM, et al. Correlation of local failure with mea-
sures of dose insufficiency in the high-dose single-fraction
• Surgical management for metastasis to the spine requires treatment of bony metastases. Int J Radiat Oncol Biol Phys.
an understanding of the unique requirements of the onco- 2010;77(4):1282–7.
10. Laufer I, et al. Local disease control for spinal metastases following
logic patient. “separation surgery” and adjuvant hypofractionated or high-dose
• The NOMS framework considers the neurologic, onco- single-fraction stereotactic radiosurgery: outcome analysis in 186
logic, mechanical, and systemic status of the patient to patients. J Neurosurg Spine. 2013;18(3):207–14.
determine the optimal treatment for spinal metastases. 11. Fisher CG, et al. A novel classification system for spinal instabil-
ity in neoplastic disease: an evidence-based approach and expert
• Separation surgery allows for neural decompression with- consensus from the Spine Oncology Study Group. Spine (Phila Pa
out significant reconstruction of the spine, combined with 1976). 2010;35(22):E1221–9.
conformal radiation therapy to achieve effective local 12. Newman WC, Laufer I, Bilsky MH. Neurologic, oncologic,
tumor control. mechanical, and systemic and other decision frameworks for spinal
disease. Neurosurg Clin N Am. 2020;31(2):151–66.
• The application of minimally invasive techniques in spi- 13. Young RF, Post EM, King GA. Treatment of spinal epidural metas-
nal oncology has provided strategies that achieve surgical tases. Randomized prospective comparison of laminectomy and
goals of care with decreased morbidity and rapid resump- radiotherapy. J Neurosurg. 1980;53(6):741–8.
tion of oncologic care. 14. Gilbert RW, Kim JH, Posner JB. Epidural spinal cord compres-
sion from metastatic tumor: diagnosis and treatment. Ann Neurol.
1978;3(1):40–51.
15. Patchell RA, et al. Direct decompressive surgical resection in the
treatment of spinal cord compression caused by metastatic cancer:
a randomised trial. Lancet. 2005;366(9486):643–8.
16. Steffee AD, Biscup RS, Sitkowski DJ. Segmental spine plates with
Open Questions
pedicle screw fixation. A new internal fixation device for disorders
• Can molecular profiling of tumor and targeted ther- of the lumbar and thoracolumbar spine. Clin Orthop Relat Res.
apies be incorporated into decision-making for sur- 1986;203:45–53.
gical treatment of spinal metastasis to minimize 17. Demura S, et al. Total en bloc spondylectomy for spinal metastases
in thyroid carcinoma. J Neurosurg Spine. 2011;14(2):172–6.
intervention? 18. Cloyd JM, et al. En bloc resection for primary and metastatic tumors
• Can endoscopic approaches be incorporated into of the spine: a systematic review of the literature. Neurosurgery.
decompression for metastatic disease? 2010;67(2):435–44; discussion 444–5.
• Can intraoperative radiation techniques provide 19. Al-Omair A, et al. Surgical resection of epidural disease improves
local control following postoperative spine stereotactic body radio-
adequate long-term control to obviate the need for therapy. Neuro-Oncology. 2013;15(10):1413–9.
external radiation therapy? 20. Bilsky MH, et al. Single-stage posterolateral transpedicle
approach for spondylectomy, epidural decompression, and cir-
532 C. P. Wang et al.

cumferential fusion of spinal metastases. Spine (Phila Pa 1976). 38. Krueger A, et al. Management of pulmonary cement embolism after
2000;25(17):2240–9; discussion 250. percutaneous vertebroplasty and kyphoplasty: a systematic review
21. Pennington Z, et al. Minimally invasive versus conventional spine of the literature. Eur Spine J. 2009;18(9):1257–65.
surgery for vertebral metastases: a systematic review of the evi- 39. Wang LJ, et al. Pulmonary cement embolism associated with percu-
dence. Ann Transl Med. 2018;6(6):103. taneous vertebroplasty or kyphoplasty: a systematic review. Orthop
22. Than KD, et al. Complication rates associated with open versus Surg. 2012;4(3):182–9.
percutaneous pedicle screw instrumentation among patients under- 40. Gosev I, et al. Right ventricular perforation and pulmonary embo-
going minimally invasive interbody fusion for adult spinal defor- lism with polymethylmethacrylate cement after percutaneous
mity. Neurosurg Focus. 2017;43(6):E7. kyphoplasty. Circulation. 2013;127(11):1251–3.
23. Wang H, et al. Comparison of open versus percutaneous pedicle 41. Barzilai O, et al. Minimal access surgery for spinal metastases: pro-
screw fixation using the sextant system in the treatment of traumatic spective evaluation of a treatment algorithm using patient-reported
thoracolumbar fractures. Clin Spine Surg. 2017;30(3):E239–e246. outcomes. World Neurosurg. 2018;120:e889–901.
24. Moussazadeh N, et al. Short-segment percutaneous pedicle screw 42. Tomasian A, et al. Percutaneous thermal ablation of spinal
fixation with cement augmentation for tumor-induced spinal insta- metastases: recent advances and review. AJR Am J Roentgenol.
bility. Spine J. 2015;15(7):1609–17. 2018;210(1):142–52.
25. Barzilai O, et al. Utility of cement augmentation via percutaneous 43. Wallace AN, et al. Radiographic local control of spinal metastases
fenestrated pedicle screws for stabilization of cancer-related spinal with percutaneous radiofrequency ablation and vertebral augmenta-
instability. Oper Neurosurg (Hagerstown). 2019;16(5):593–9. tion. AJNR Am J Neuroradiol. 2016;37(4):759–65.
26. Versteeg AL, et al. Complications after percutaneous pedicle screw 44. Singh S, Saha S. Electrical properties of bone. A review. Clin
fixation for the treatment of unstable spinal metastases. Ann Surg Orthop Relat Res. 1984;186:249–71.
Oncol. 2016;23(7):2343–9. 45. Reyes M, et al. Multicenter clinical and imaging evaluation
27. Barzilai O, Bilsky MH, Laufer I. The role of minimal access sur- of targeted radiofrequency ablation (t-RFA) and cement aug-
gery in the treatment of spinal metastatic tumors. Global Spine J. mentation of neoplastic vertebral lesions. J Neurointerv Surg.
2020;10(2 Suppl):79s–87s. 2018;10(2):176–82.
28. Dousset V, et al. Asymptomatic cervical haemangioma treated by 46. Sloan AE, et al. Results of the NeuroBlate System first-in-humans
percutaneous vertebroplasty. Neuroradiology. 1996;38:392–4. Phase I clinical trial for recurrent glioblastoma: clinical article. J
29. Amar AP, et al. Percutaneous transpedicular polymethylmethac- Neurosurg. 2013;118(6):1202–19.
rylate vertebroplasty for the treatment of spinal compression frac- 47. Tatsui CE, et al. Spinal laser interstitial thermal therapy: a novel
tures. Neurosurgery. 2001;49(5):1105–14; discussion 1114–5. alternative to surgery for metastatic epidural spinal cord compres-
30. Berenson J, et al. Balloon kyphoplasty versus non-surgical fracture sion. Neurosurgery. 2016;79(Suppl 1):S73–82.
management for treatment of painful vertebral body compression 48. Bastos DCA, et al. Spinal laser interstitial thermal therapy: single-­
fractures in patients with cancer: a multicentre, randomised con- center experience and outcomes in the first 120 cases. J Neurosurg
trolled trial. Lancet Oncol. 2011;12(3):225–35. Spine. 2020:1–10.
31. Sorensen ST, et al. Vertebroplasty or kyphoplasty as palliative treat- 49. Yamada Y, et al. The impact of histology and delivered dose on
ment for cancer-related vertebral compression fractures: a system- local control of spinal metastases treated with stereotactic radiosur-
atic review. Spine J. 2019;19(6):1067–75. gery. Neurosurg Focus. 2017;42(1):E6.
32. Lee MJ, et al. Percutaneous treatment of vertebral compression 50. Moulding HD, et al. Local disease control after decompressive sur-
fractures: a meta-analysis of complications. Spine (Phila Pa 1976). gery and adjuvant high-dose single-fraction radiosurgery for spine
2009;34(11):1228–32. metastases. J Neurosurg Spine. 2010;13(1):87–93.
33. Stangenberg M, et al. Cervical vertebroplasty for osteolytic metas- 51. Yamada Y, et al. High-dose, single-fraction image-guided intensity-­
tases as a minimally invasive therapeutic option in oncological sur- modulated radiotherapy for metastatic spinal lesions. Int J Radiat
gery: outcome in 14 cases. Neurosurg Focus. 2017;43(2):E3. Oncol Biol Phys. 2008;71(2):484–90.
34. Lieberman IH, Togawa D, Kayanja MM. Vertebroplasty and kypho- 52. Tseng CL, et al. Spine stereotactic body radiotherapy: indications,
plasty: filler materials. Spine J. 2005;5(6 Suppl):305S–16S. outcomes, and points of caution. Global Spine J. 2017;7(2):179–97.
35. Lieberman IH, et al. Initial outcome and efficacy of “kyphoplasty” 53. DeLaney TF, et al. Intraoperative dural irradiation by customized
in the treatment of painful osteoporotic vertebral compression frac- 192iridium and 90yttrium brachytherapy plaques. Int J Radiat
tures. Spine (Phila Pa 1976). 2001;26(14):1631–8. Oncol Biol Phys. 2003;57(1):239–45.
36. Chang X, et al. Vertebroplasty versus kyphoplasty in osteoporotic 54. Folkert MR, et al. Intraoperative 32P high-dose rate brachytherapy
vertebral compression fracture: a meta-analysis of prospective of the dura for recurrent primary and metastatic intracranial and
comparative studies. Int Orthop. 2015;39(3):491–500. spinal tumors. Neurosurgery. 2012;71(5):1003–10; discussion
37. Dalton BE, et al. Radiofrequency-targeted vertebral augmentation 1010–1.
versus traditional balloon kyphoplasty: radiographic and morpho- 55. Bludau F, et al. Phase I/II trial of combined kyphoplasty and
logic outcomes of an ex vivo biomechanical pilot study. Clin Interv intraoperative radiotherapy in spinal metastases. Spine J.
Aging. 2012;7:525–31. 2018;18(5):776–81.
 Interventional Radiology Techniques
for Management of Skeletal Metastases 49
Nainesh Parikh and Altan Ahmed

Abstract fluoroscopy may be used separately or in combination for

performing image-guided procedures. In addition to select-
Osseous events in patients with metastatic disease can
ing an imaging modality, several techniques are available for
contribute to significant impairments in quality of life.
the image-guided management of skeletal metastases. The
Advances in minimally invasive image-guided interven-
technique or techniques utilized is dependent on multidis-
tions have led to increasing utilization as adjunctive, pal-
ciplinary discussion, location and type of lesion (i.e., lytic,
liative, and occasionally definitive therapies for osseous
blastic, or mixed), and treatment intent. While transarterial
metastatic disease. In this chapter, we discuss the clinical
embolization of a hypervascular skeletal metastasis may be
applications, technique, and outcomes of percutaneous
utilized to reduce operative blood loss, it can also be used to
ablation, transarterial embolization, cementoplasty, and
palliate pain even if surgery is not an option. Similarly, per-
percutaneous osteosynthesis.
cutaneous ablation might be used for disease control/defini-
tive treatment in addition to palliation of pain, depending
on the clinical scenario. Finally, nonoperative percutaneous
consolidation (e.g., cementoplasty and/or osteosynthesis)
Learning Objectives might be utilized in patients in whom surgery is not feasible
• Understand the role of interventional radiology in to improve quality of life [2, 3].
the management of skeletal metastases Ultimately the goals of intervention, the ability to perform
• Learn indications for image-guided interventions the intervention safely and with image guidance, the pos-
for skeletal metastases sible requirement of general anesthesia, and the suitability
• Describe technical considerations and outcomes of of an image-guided intervention compared to other types of
percutaneous ablation, transarterial embolization, nonoperative intervention (such as systemic or radiotherapy)
cementoplasty, and percutaneous osteosynthesis indicate which type of nonoperative management is best for
the patient. This chapter reviews the types of image-guided
therapies, indications, and outcomes of these therapies in the
management of skeletal metastases.
49.1 Introduction

Although the field of Interventional Radiology is relatively 49.2 Percutaneous Ablation

young, rapid advancements of minimally invasive techniques
have strengthened the armamentarium of therapies available Percutaneous ablation is performed by using a single or com-
to patients [1]. Imaging modalities such as ultrasound (US), bination guidance modality such as US, CT, or MRI to place
computed tomography (CT), magnetic resonance (MR), and a needle or probe into a bony structure to deliver a cytotoxic
substance or thermal energy and cause cellular destruction.
Various substances and technologies are available including
N. Parikh (*) · A. Ahmed ethanol, radiofrequency (RF), microwave (MW), and cryo-
Department of Diagnostic Imaging and Interventional Radiology, therapy. Thermal ablation refers to the direct application of
H. Lee Moffitt Cancer Center and Research Institute,
Tampa, FL, USA
either extreme heat (>50 °C) or extreme cold (between −20
e-mail: [email protected]; [email protected] and −40 °C) to the target tissue. The choice of substance/

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 533
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_49
534 N. Parikh and A. Ahmed

technology depends on the underlying histology, morphol- geneity and confounding effect of other therapies, one study
ogy, and the location of adjacent critical structures. RF and reports median overall survival of 47 months and median
cryotherapy are the most frequently utilized forms of ablation disease-­free survival of 7 months after osseous ablation [5].
for skeletal metastases. RF ablation works by passing electri- Prognostic factors that influence the outcome of percuta-
cal current in the range of radiofrequency waves between the neous ablation include the size and number of metastases,
electrodes. This current produces heat around and between length of disease-free survival before the onset of metas-
the electrodes, resulting in tissue destruction. Cryoablation tases, adequacy of treatment of the primary tumor, and the
works by cycling inert gas down the shaft of a probe, rap- presence of multiple metastatic sites [6, 7].
idly producing ice and locally freezing the tissue. By cycling All ablation modalities are thought to relieve pain via
through freezing and thawing phases, the cells are disrupted similar multifactorial processes: the destruction of sensory
to such an extent that they are destroyed (Fig. 49.1) [2–4]. fibers that supply the periosteum, decompression of tumor
Ablation procedures are typically performed under CT volume, eradication of cytokine-producing tumor cells, and
guidance and with either general anesthesia or moderate seda- inhibition of osteoclast activity [8]. Reported outcomes vary;
tion (fentanyl and midazolam) depending on the patient per- however, most describe significant improvement in bone-­
formance status and associated comorbidities, as well as the related pain refractory to standard therapy [9, 10].
planned procedure. A detailed neurologic examination and
pain survey are paramount prior to and following the pro-
cedure. If the target ablation site is close to the skin, warm 49.3 Transarterial Embolization
or cold saline compresses can be applied for protection. Air,
fluid, or catheter-mounted balloons may also be utilized to dis- Many tumors demonstrate neoangiogenesis with varying
place critical structures in the vicinity of the planned ablation. degrees of vascularity. Metastases from melanoma, renal
Because the destructive properties of ablation are lim- cell, thyroid, breast, and hepatocellular carcinoma are con-
ited to the tissue adjacent to and/or in between probes, it sidered among the most vascular. TAE aims to devascular-
becomes challenging to provide definitive tumor control for ize tumors by blocking arterial blood flow to the capillary
large tumors. Small tumors, typically less than 5 cm, are ide- bed. The most common application of TAE is as a surgical
ally treated with ablative technology with the goals of local adjunct to reduce perioperative blood loss. Tumor histology,
control and/or palliation of symptoms. If “cancer pain” is type of surgery planned, and surgeon preference are factors
felt to be secondary to lytic metastasis, ablative technologies that should be considered in deciding whether adjunctive
are helpful for palliation. Most ablative procedures can be TAE should be performed. TAE may also be applied as a
performed on an outpatient basis, with the option to monitor palliative intervention for symptomatic metastases refractory
patients overnight for pain control. or not amenable to other treatments.
While skeletal metastases often imply advanced stage A thorough understanding of regional vascular anatomy
malignancy, oligometastatic disease (up to five sites) may is essential to safe and efficacious TAE. Pre-procedural
represent an opportunity to achieve tumor control. While cross-sectional imaging is useful to delineate tumor anatomy
the data are challenging to interpret due to patient hetero- and potential sources of vascularization. In some cases, there

a b c

Fig. 49.1 CT demonstrates a lytic renal cell carcinoma left scapular zone (b). Cementoplasty was subsequently performed due to subopti-
metastasis (arrow) (a). CT-guided cryoablation with three probes mal pain response to ablation alone (arrow) (c)
(arrows) was performed. The arrowhead marks the hypodense ablation
49 Interventional Radiology Techniques for Management of Skeletal Metastases 535

a b

Fig. 49.2 Digital subtraction angiogram demonstrates a vascular renal (a). The vascular supply was coiled (arrows) without significant resid-
cell carcinoma metastasis involving the lateral left clavicle (rectangle) ual tumor enhancement (b)
with arterial supply arising from branches of the thyrocervical trunk

may discordance between vascularity identified on cross-­ arteriovenous shunting, risk of nontarget embolization, and
sectional imaging versus that seen angiographically [11]. operator experience. There is insufficient data to support the
Contraindications to angiography and embolization include use of one embolic material over another.
uncorrectable coagulopathy, underlying infection, renal fail- Multiple studies have demonstrated the safety of trans-
ure, or inability to isolate tumoral blood supply from that of arterial embolization of skeletal metastases [11, 13–16]. In
critical tissue. addition, data suggest adjunctive TAE is associated with
TAE is typically performed under conscious sedation. lower operative blood loss [13, 14, 17–21]. There are no
Arterial access is typically obtained via the femoral or radial consistent data demonstrating significant differences in
artery. After placing a vascular sheath, a 4–5f diagnostic operative time between patients who undergo adjunctive
catheter is maneuvered over a guidewire to the area of inter- embolization and those who do not [19]. The optimal inter-
est. An initial non-selective angiogram is performed to delin- val between TAE and surgery has been reported as <3 days
eate vascular anatomy. It is critical to identify blood supply to reduce the formation of collateral circulation [17, 18, 22].
to both tumor and normal tissue as embolization of nontarget Further investigation is needed to refine selection criteria for
vessels can result in devastating clinical outcomes such as patients who may benefit from preoperative TAE of skeletal
sensorimotor deficits and tissue or limb loss. Once the tar- metastases [23, 24].
get vessel or vessels are identified, a coaxial microcatheter is In palliative settings, the mean duration of pain improve-
typically advanced as distal as possible from the parent ves- ment has been reported in the range of 1–15 months [15,
sel to reduce the risk of nontarget embolization. After con- 17, 25, 26]. Combination therapy with external beam therapy
firming catheter position, embolic material is administered may provide more durable pain relief [26].
in small aliquots under fluoroscopic observation (Fig. 49.2).
The angiographic end point of embolization in this setting
has been described as ≥70% reduction in tumor staining [12, 49.4 Percutaneous Cementoplasty
13]. Post-embolization syndrome characterized by fever, and Osteosynthesis
fatigue, pain, and/or nausea/vomiting is common after TAE
and is usually self-limited. Pathologic fractures, pain, and decreased mobility are
Embolic materials are categorized as either temporary or major morbidities of osseous metastatic disease [27].
permanent agents. The primary temporary agent is gelatin Cementoplasty (also referred to as vertebroplasty or kypho-
sponge. Permanent embolic materials include particulates plasty in the spine) involves the injection of polymethyl
(e.g., polyvinyl alcohol (PVA) and microspheres), coils, methacrylate (PMMA) into bone to palliate painful metas-
and liquid embolics (e.g., N-butyl cyanoacrylate and ethyl- tases and/or prevent pathologic fracture (Fig. 49.1). In areas
ene vinyl alcohol). Overall, liquid materials afford greater where torsional forces are prominent (e.g., long bones), the
penetration into the target vascular bed but require greater consolidative effect of cementoplasty alone is limited despite
operator experience and may carry a higher risk of nontarget pain reduction and may be supplemented by percutaneous
embolization. A detailed discussion of each embolic agent is osteosynthesis when appropriate [28]. The analgesic effect of
beyond the scope of this text. The choice of embolic mate- cementoplasty is predominantly thought to arise from frac-
rial is largely dependent on vascular anatomy, presence of ture consolidation. The exothermic polymerization reaction
536 N. Parikh and A. Ahmed

of PMMA may also contribute to analgesic effect through 49.5 Conclusion

thermal injury to local nociceptive fibers [29]. Patients with
limited osseous metastatic disease may benefit from a com-
• Percutaneous image-guided therapies comprise a number
bined approach to treat the tumor with thermal ablation,
of minimally invasive and efficacious interventions for
radiation, or systemic therapy followed by cementoplasty to
osseous metastatic disease
stabilize the bone [30].
• Regardless of technique or techniques selected, a multi-
Understanding whether compressive or torsional forces
disciplinary approach to patient selection is essential for
predominate in the area of concern is a key component of
successful outcomes
pre-procedural assessment in addition to quantifying base-
line pain and analgesic use. Cross-sectional imaging is
essential prior to intervention to establish the safest access
and delineate tumor and regional anatomy. Contraindications Open Questions
include uncorrectable coagulopathy, infection, allergy to Future controlled studies are needed to determine:
cement components, neurologic compromise, sagittal spinal • Which patients would benefit most from minimally
instability. Cementoplasty may be performed under con- invasive management of bone metastases?
scious sedation if a limited level vertebroplasty is planned. • Do differences exist between embolization materi-
However, general anesthesia is often necessary if patient sta- als or ablation modalities?
tus dictates and if or multiple sites or types of intervention • What is the role of percutaneous techniques for sta-
are planned. In the spine, the most common access routes are bilization of pathologic fractures?
transpedicular for the lumbar and lower thoracic, intercos-
topedicular for mid and upper thoracic, and anterior or tran-
soral for the cervical vertebrae. Unilateral or bilateral access
may be obtained. Outside the spine, access routes vary con- References
siderably depending on the target bone. A bone access needle
is advanced into the bone along the chosen tract under fluo- 1. Murphy TP, Soares GM. The evolution of interventional radi-
ology. Semin Intervent Radiol. 2005;22:6–9. https://doi.
roscopic and/or CT guidance. Once appropriate position is org/10.1055/s-­2005-­869570.
confirmed, PMMA is injected under observation to ensure 2. Barile A, Arrigoni F, Zugaro L, et al. Minimally invasive treatments
appropriate filling and to monitor for extravasation or intrav- of painful bone lesions: state of the art. Med Oncol. 2017;34:53.
asation [31–35]. https://doi.org/10.1007/s12032-­017-­0909-­2.
3. Munk PL, Legiehn GM. Musculoskeletal interventional radiology:
Percutaneous osteosynthesis utilizes similar principles applications to oncology. Semin Roentgenol. 2007;42:164–74.
of bone access. A coaxial needle is placed across a planned https://doi.org/10.1053/j.ro.2007.04.004.
trajectory followed by the advancement of a Kirschner wire. 4. Errani C, Bazzocchi A, Spinnato P, et al. What’s new in management
Alternatively, a Kirschner wire may be placed directly. After of bone metastases? Eur J Orthop Surg Traumatol. 2019;29:1367–
75. https://doi.org/10.1007/s00590-­019-­02446-­y.
measuring the required length, the appropriate cannulated 5. McMenomy BP, Kurup AN, Johnson GB, et al. Percutaneous
screw is advanced over the wire which is subsequently removed. cryoablation of musculoskeletal oligometastatic disease for com-
If percutaneous osteosynthesis is to be performed at the same plete remission. J Vasc Interv Radiol. 2013;24:207–13. https://doi.
time as cementoplasty, it can be done prior to the cement set- org/10.1016/j.jvir.2012.10.019.
6. Nicholas Kurup A, Callstrom MR. Ablation of musculoskeletal
ting using coaxial technique through the existing bone access metastases: pain palliation, fracture risk reduction, and oligometa-
or prior to cementoplasty using separate access [28]. static disease. Tech Vasc Interv Radiol. 2013;16:253–61. https://
Cementoplasty in the spine consistently demonstrates doi.org/10.1053/j.tvir.2013.08.007.
a durable pain response in the setting of spinal metastatic 7. Weichselbaum RR, Hellman S. Oligometastases revisited.
Nat Rev Clin Oncol. 2011;8:378–82. https://doi.org/10.1038/
disease [32, 33]. In subset analysis of a prospective series, nrclinonc.2011.44.
644 patients with vertebral metastatic disease reported a sig- 8. Callstrom MR, Charboneau JW, Goetz MP, et al. Painful metasta-
nificant pain reduction from 8.3 to 2.4 on an 11-point visual ses involving bone: feasibility of percutaneous CT- and US-guided
analog scale (VAS) [36]. There is growing evidence of simi- radio-frequency ablation. Radiology. 2002;224:87–97. https://doi.
org/10.1148/radiol.2241011613.
lar pain response in extra-spinal cementoplasty [37–39]. In 9. Dupuy DE, Liu D, Hartfeil D, et al. Percutaneous radiofre-
a series of 27 patients with painful osseous metastases or quency ablation of painful osseous metastases: a multicenter
impending pathologic fracture who underwent percutaneous American College of Radiology Imaging Network trial. Cancer.
osteosynthesis, Pusceddu et al. reported a mean decrease in 2010;116:989–97. https://doi.org/10.1002/cncr.24837.
10. Goetz MP, Callstrom MR, Charboneau JW, et al. Percutaneous
VAS score from 8 pre-treatment to 1.6 and 1.4 one and six image-guided radiofrequency ablation of painful metastases involv-
months posttreatment, respectively [40]. Similar findings ing bone: a multicenter study. J Clin Oncol. 2004;22:300–6. https://
have been reported in other studies [37, 41–44]. doi.org/10.1200/JCO.2004.03.097.
49 Interventional Radiology Techniques for Management of Skeletal Metastases 537

11. Prabhu VC, Bilsky MH, Jambhekar K, et al. Results of preopera- ease in cancer patients. Semin Intervent Radiol. 2018;35:238–47.
tive embolization for metastatic spinal neoplasms. J Neurosurg. https://doi.org/10.1055/s-­0038-­1673359.
2003;98:156–64. https://doi.org/10.3171/spi.2003.98.2.0156. 29. Deramond H, Wright NT, Belkoff SM. Temperature elevation
12. Kwon JH, Shin JH, Kim JH, et al. Preoperative transcath- caused by bone cement polymerization during vertebroplasty. Bone.
eter arterial embolization of hypervascular metastatic tumors 1999;25:17S–21S. https://doi.org/10.1016/S8756-­3282(99)00127-­1.
of long bones. Acta Radiol. 2010;51:396–401. https://doi. 30. Cazzato RL, Buy X, Grasso RF, et al. Interventional radiolo-
org/10.3109/02841851003660081. gist’s perspective on the management of bone metastatic disease.
13. Sun S, Lang EV. Bone metastases from renal cell carcinoma: pre- Eur J Surg Oncol. 2015;41:967–74. https://doi.org/10.1016/j.
operative embolization. J Vasc Interv Radiol. 1998;9:263–9. https:// ejso.2015.05.010.
doi.org/10.1016/S1051-­0443(98)70267-­2. 31. Anselmetti GC. Musculoskeletal interventions osteoplasty percu-
14. Pazionis TJC, Papanastassiou ID, Maybody M, Healey taneous bone cement injection beyond the spine. Semin Intervent
JH. Embolization of hypervascular bone metastases reduces intra- Radiol. 2010;27:199–208.
operative blood loss: a case-control study. Clin Orthop Relat Res. 32. Buy X, Catena V, Roubaud G, et al. Image-guided bone consolida-
2014;472:3179–87. https://doi.org/10.1007/s11999-­014-­3734-­3. tion in oncology. Semin Intervent Radiol. 2018;35:221–8. https://
15. Radeleff B, Eiers M, Lopez-Benitez R, et al. Transarterial embo- doi.org/10.1055/s-­0038-­1669468.
lization of primary and secondary tumors of the skeletal sys- 33. Katsanos K, Sabharwal T, Adam A. Percutaneous cemento-
tem. Eur J Radiol. 2006;58:68–75. https://doi.org/10.1016/j. plasty. Semin Intervent Radiol. 2010;27:137–47. https://doi.
ejrad.2005.12.008. org/10.1055/s-­0030-­1253512.
16. Rossi G, Mavrogenis AF, Rimondi E, et al. Selective emboliza- 34. Mastier C, Gjorgjievska A, Thivolet A, et al. Musculoskeletal
tion with N-butyl cyanoacrylate for metastatic bone disease. J metastases management: the interventional radiologist’s tool-
Vasc Interv Radiol. 2011;22:462–70. https://doi.org/10.1016/j. box. Semin Intervent Radiol. 2018;35:281–9. https://doi.
jvir.2010.12.023. org/10.1055/s-­0038-­1673420.
17. Barton PP, Waneck RE, Karnel FJ, et al. Embolization of bone 35. Yevich S, Tselikas L, Gravel G, et al. Percutaneous cement injec-
metastases. J Vasc Interv Radiol. 1996;7:81–8. https://doi. tion for the palliative treatment of osseous metastases: a techni-
org/10.1016/S1051-­0443(96)70738-­8. cal review. Semin Intervent Radiol. 2018;35:268–80. https://doi.
18. Gellad FE, Sadato N, Numaguchi Y, Levine AM. Vascular meta- org/10.1055/s-­0038-­1673418.
static lesions of the spine: preoperative embolization. Radiology. 36. Anselmetti GC, Marcia S, Saba L, et al. Percutaneous vertebro-
1990;176:683–6. https://doi.org/10.1148/radiology.176.3.2389026. plasty: multi-centric results from EVEREST experience in large
19. Geraets SEW, Bos PK, van der Stok J. Preoperative emboliza- cohort of patients. Eur J Radiol. 2012;81:4083–6. https://doi.
tion in surgical treatment of long bone metastasis: a systematic org/10.1016/j.ejrad.2012.07.005.
literature review. EFORT Open Rev. 2020;5:17–25. https://doi. 37. Deschamps F, de Baere T, Hakime A, et al. Percutaneous osteosyn-
org/10.1302/2058-­5241.5.190013. thesis in the pelvis in cancer patients. Eur Radiol. 2016;26:1631–9.
20. Ma J, Tullius T, Van Ha TG. Update on preoperative emboliza- https://doi.org/10.1007/s00330-­015-­3971-­1.
tion of bone metastases. Semin Intervent Radiol. 2019;36:241–8. 38. Deschamps F, Farouil G, Hakime A, et al. Cementoplasty of
https://doi.org/10.1055/s-­0039-­1693120. metastases of the proximal femur: is it a safe palliative option? J
21. Olerud C, Jónsson H, Löfberg AM, et al. Embolization of spinal Vasc Interv Radiol. 2012a;23:1311–6. https://doi.org/10.1016/j.
metastases reduces peroperative blood loss: 21 patients operated on jvir.2012.06.027.
for renal cell carcinoma. Acta Orthop. 1993;64:9–12. https://doi. 39. Wang Z, Zhen Y, Wu C, et al. CT fluoroscopy-guided percutaneous
org/10.3109/17453679308994517. osteoplasty for the treatment of osteolytic lung cancer bone metas-
22. Gottfried ON, Schloesser PE, Schmidt MH, Stevens tases to the spine and pelvis. J Vasc Interv Radiol. 2012;23:1135–
EA. Embolization of metastatic spinal tumors. Neurosurg Clin N 42. https://doi.org/10.1016/j.jvir.2012.06.007.
Am. 2004;15:391–9. https://doi.org/10.1016/j.nec.2004.04.003. 40. Pusceddu C, Fancellu A, Ballicu N, et al. CT-guided percutane-
23. Jernigan EW, Tennant JN, Esther RJ. Not all patients undergoing ous screw fixation plus cementoplasty in the treatment of pain-
stabilization of impending pathologic fractures for renal cell car- ful bone metastases with fractures or a high risk of pathological
cinoma metastases to the femur need preoperative embolization. fracture. Skelet Radiol. 2017;46:539–45. https://doi.org/10.1007/
Clin Orthop Relat Res. 2018;476:529–34. https://doi.org/10.1007/ s00256-­017-­2584-­y.
s11999.0000000000000058. 41. Amoretti N, Huwart L, Hauger O, et al. Percutaneous screw fixa-
24. Kickuth R, Waldherr C, Hoppe H, et al. Interventional management tion of acetabular roof fractures by radiologists under CT and fluo-
of hypervascular osseous metastasis: role of embolotherapy before roscopy guidance. Am J Roentgenol. 2013;200:447–50. https://doi.
orthopedic tumor resection and bone stabilization. Am J Roentgenol. org/10.2214/AJR.11.7968.
2008;191:240–7. https://doi.org/10.2214/AJR.07.4037. 42. Deschamps F, Farouil G, Hakime A, et al. Percutaneous stabili-
25. Forauer AR, Kent E, Cwikiel W, et al. Selective palliative trans- zation of impending pathological fracture of the proximal femur.
catheter embolization of bony metastases from renal cell car- Cardiovasc Intervent Radiol. 2012b;35:1428–32. https://doi.
cinoma. Acta Oncol (Madr). 2007;46:1012–8. https://doi. org/10.1007/s00270-­011-­0330-­8.
org/10.1080/02841860701280725. 43. Fischer S, Vogl TJ, Marzi I, et al. Percutaneous cannulated screw
26. Uemura A, Fujimoto H, Yasuda S, et al. Transcatheter arterial embo- fixation of sacral fractures and sacroiliac joint disruptions with
lization for bone metastases from hepatocellular carcinoma. Eur CT-controlled guidewires performed by interventionalists: single
Radiol. 2001;11:1457–62. https://doi.org/10.1007/s003300000792. center experience in treating posterior pelvic instability. Eur J Radiol.
27. Filippiadis DK, Tselikas L, Bazzocchi A, et al. Percutaneous man- 2015;84:290–4. https://doi.org/10.1016/j.ejrad.2014.11.017.
agement of cancer pain. Curr Oncol Rep. 2020;22:43. https://doi. 44. Strobl FF, Haeussler SM, Paprottka PM, et al. Technical and clini-
org/10.1007/s11912-­020-­00906-­x. cal outcome of percutaneous CT fluoroscopy-guided screw place-
28. Deschamps F, Yevich S, Gravel G, et al. Percutaneous fixation by ment in unstable injuries of the posterior pelvic ring. Skelet Radiol.
internal cemented screw for the treatment of unstable osseous dis- 2014;43:1093–100. https://doi.org/10.1007/s00256-­014-­1890-­x.
 Part XIII
Cancer Metastasis to the Brain and the Nervous
System

Ian Lee

Central nervous system (CNS) metastases are a common occurrence in the course of disease
for cancer patients, reported to affect between 9% and 17% of all cancer patients, although this
likely underestimates the true prevalence of disease [1]. Although combinatorial treatments
can effectively manage disease in many patients, CNS metastases still represent a source of
significant decrease in the quality of life, due to the effects of metastases as well as their treat-
ment. In addition, a sizable portion of cancer patients will succumb due to the CNS manifesta-
tions of their cancer. Some types of cancer, like metastatic melanoma, have a particularly poor
prognosis [2]. As patients are living longer with metastatic cancer due to improved treatments,
it is also likely that the incidence of CNS metastases is increasing over time [3]. Taken together,
CNS metastases represent a large burden of disease in cancer patients, although not necessarily
a terminal condition with the advent of new treatment modalities.
CNS metastases most typically occur in the brain parenchyma. Other less common sites
include the meninges, pituitary or pineal gland, and choroid plexus. Leptomeningeal dissemi-
nation of cancer through the cerebrospinal fluid carries an especially poor prognosis for most
patients, as current treatments still have limited efficacy [4]. The most common primary sites
of cancer are lung, breast, and melanoma. Other types of cancer such as pancreas and prostate
rarely cause brain metastases. The overall prevalence of lung and breast cancer likely contrib-
utes to the high prevalence of brain metastases from those sites. In comparison, the relatively
high incidence of CNS metastatic disease from certain histologies (despite their rarity) sug-
gests a CNS tropism depending on the cell type. For instance, in melanoma patients up to 50%
of patients with metastatic disease will develop CNS involvement [5]. Another example is the
inverse relationship of testicular cancer, which has a high incidence of brain metastases, com-
pared to prostate, which uncommonly spreads to the CNS [6, 7].
One of the hallmarks of CNS metastases is the simultaneous discovery of multiple CNS
lesions; however, a significant minority will present with a solitary lesion [8]. Some types of
cancer, like lung and melanoma, tend to present as multiple tumors. In contrast, other types
such as breast, renal, and colorectal are associated with solitary brain metastases [8, 9]. Most
patients will present with neurological symptoms specific to disease location (i.e., numbness,
weakness, or aphasia) or with increased intracranial pressure manifesting as headache or
decreased level of consciousness [10]. In addition, certain histologies like breast, lung, thyroid,
renal, and melanoma tend to present as hemorrhagic lesions in the brain, presenting both diag-
nostic and treatment challenges. In fact, in a patient who presents with multiple intracerebral
hemorrhages, metastatic cancer should be highly suspected [11].
With no treatment, prognosis is thought to be approximately 3 months. For patients with
oligometastatic disease, the combination of radiation and surgery can typically control disease.
However, overall prognosis remains poor, in the range of 4–6 months, likely as a result of
widespread disease outside the CNS [12]. In a subset of patients with single CNS metastases
and limited extracranial disease, life expectancy may be more favorable, exceeding a year [13].
Nonetheless, the development of CNS metastases is a harbinger of poor survival.
540 Cancer Metastasis to the Brain and the Nervous System

The following section will focus on the pathophysiology of CNS metastases, including the molecular
mechanisms of CNS tropism. We will review the molecular cascades underpinning the spread of cancer to
the CNS, which have been an active area of research. Furthermore, these mechanisms may present targets
for treatment. Next, the current mainstays of treatment, radiation, and/or surgery will be discussed. The
combination of radiation and surgery can typically control CNS disease, especially in cases of oligometas-
tases. However, they do have attendant risks due to their invasive nature as well as an expected decrement
in quality of life. As such, newer treatments such as stereotactic radiosurgery represent a paradigm shift in
the treatment of CNS metastatic disease in that it is efficacious and (relatively) spares the surrounding nor-
mal parenchyma, thus partially preventing the cognitive effects of whole-brain radiation therapy. Surgery is
typically reserved for larger tumors or those that have progressed despite radiation treatment. However,
with the advent of less invasive treatments like laser interstitial thermal ablation, the surgical indications for
CNS metastases have expanded.
Classically, CNS metastases are thought to be unresponsive to systemic chemotherapy due to the blood-­
brain barrier preventing adequate penetration of chemotherapeutics. However, this viewpoint is gradually
shifting with advances in the development of chemotherapeutic agents and especially with the introduction
of immunotherapy. Finally, the pathophysiology and treatment of leptomeningeal carcinomatosis, the most
dreaded form of CNS metastatic disease, will also be reviewed.

References
1. Nayak L, Lee EQ, Wen PY. Epidemiology of brain metastases. Curr Oncol Rep. 2012;14:48–54. https://
doi.org/10.1007/s11912-­011-­0203-­y
2. McWilliams RR, Brown PD, Buckner JC, et al. Treatment of brain metastases from melanoma. Mayo
Clin Proc. 2003;78:1529–36. https://doi.org/10.4065/78.12.1529
3. Bachmann C, Schmidt S, Staebler A, et al. CNS metastases in breast cancer patients: prognostic impli-
cations of tumor subtype. Med Oncol. 2015;32:400. https://doi.org/10.1007/s12032-­014-­0400-­2
4. Morris PG, Reiner AS, Szenberg OR, et al. Leptomeningeal metastasis from non-small cell lung can-
cer: survival and the impact of whole brain radiotherapy. J Thorac Oncol. 2012;7:382–5. https://doi.
org/10.1097/JTO.0b013e3182398e4f
5. Spagnolo F, Picasso V, Lambertini M, et al. Survival of patients with metastatic melanoma and brain
metastases in the era of MAP-kinase inhibitors and immunologic checkpoint blockade antibodies: a
systematic review. Cancer Treat Rev. 2016;45:38–45. https://doi.org/10.1016/j.ctrv.2016.03.003
6. Gandaglia G, Abdollah F, Schiffmann J, et al. Distribution of metastatic sites in patients with prostate
cancer: a population-based analysis. Prostate. 2014;74:210–6. https://doi.org/10.1002/pros.22742
7. Bredael JJ, Vugrin D, Whitmore WF. Autopsy findings in 154 patients with germ cell tumors of the
testis. Cancer. 1982;50:548–51. https://doi.
org/10.1002/1097-­0142(19820801)50:3<548::aid-cncr2820500327>3.0.co;2-2
8. Nussbaum ES, Djalilian HR, Cho KH, Hall WA. Brain metastases. Histology, multiplicity, surgery, and
survival. Cancer. 1996;78:1781–8.
9. Delattre JY, Krol G, Thaler HT, Posner JB. Distribution of brain metastases. Arch Neurol. 1988;45:741–
4. https://doi.org/10.1001/archneur.1988.00520310047016
10. Boire A. Metastasis to the central nervous system. CONTINUUM: Lifelong Learning in Neurology.
2020;26:1584. https://doi.org/10.1212/CON.0000000000000939
11. Boissonneau S, Graillon T, Rolland A, et al. Management of patients suffering from hemorrhagic intra-
cranial metastases: propositions to help the neurosurgeon in emergencies situations based on a litera-
ture review. Neurochirurgie. 2020; https://doi.org/10.1016/j.neuchi.2020.11.005.
12. Johnson JD, Young B. Demographics of brain metastasis. Neurosurg Clin N Am. 1996;7:337–44.
13. Stelzer KJ. Epidemiology and prognosis of brain metastases. Surg Neurol Int. 2013;4:S192–202.
https://doi.org/10.4103/2152-­7806.111296
 Brain Metastases: Overview
and Molecular Mechanisms 50
S. Haider, J. Snyder, and I. Lee

Abstract
Learning Objectives
Both the incidence and prevalence of brain metastases
(BM) have increased, partly due to improved radiographic • Appreciate the multistep model of tumor metastasis
detection of lesions and prolonged survival from primary • Intratumoral genetic heterogeneity and clonal
cancers. Overall prognosis of BM remains poor despite evolution promote driver mutations favoring
emerging therapies. Metastasis to the central nervous sys- neurotropism
tem (CNS) requires that a cancer cell harbor the unique • Cancer cells establish a favorable tumor microen-
molecular machinery necessary to overcome protective vironment by hijacking glial and stromal molecular
CNS mechanisms ordinarily designed to impede patho- machinery of the brain
gen invasion and to facilitate tissue repair. A primary
tumor’s ability to successfully invade and thrive in the
central nervous system is dependent on subclonal evolu-
tion of cancer cell populations, with genetic changes con- 50.1 Introduction
ferring selective advantages to overcome each step in the
multistep model of tumor metastasis. Organotropism Brain metastases portend a bleak prognosis in cancer
refers to the predilection of circulating tumor cells to tar- patients and negatively contribute to morbidity and mor-
get specific secondary sites, often involving a bidirec- tality. Extant systemic chemotherapy options offer limited
tional interaction where the primary tumor influences central nervous system (CNS) penetration owing to their
changes at a distant site to create an environment condu- intrinsic hydrophobicity; thus surgery or radiation often
cive to metastatic proliferation. Once in the brain, cancer serve as the mainstay of treatment for appropriately
cells hijack glial and CNS-stromal cellular machinery to selected patients [1] [Klotz 2020]. With 10–30% of cancer
help create a tumor microenvironment where cancer cells patients developing brain metastases and a median overall
continue to thrive. A better understanding of the molecu- survival of under 1 year from the time of diagnosis, there
lar mechanisms of BM will help drive future targeted exists an acute need to understand underlying mecha-
therapy as well as identifying those at greater risk for BM. nisms of brain metastasis to develop novel therapeutic
approaches.
We outline herein a brief overview of the multi-step model
of tumor spread from a primary site to the brain as well
S. Haider · J. Snyder · I. Lee (*)
Henry Ford Health System, Detroit, USA as idiosyncrasies of the central nervous system
e-mail: [email protected]; [email protected] microenvironment.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 541
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_50
542 S. Haider et al.

50.2  he Multistep Model of Tumor

T Despite improvements in extracranial disease control with
Metastasis cytotoxic therapy and next-generation targeted therapy [7–
9], many patients go on to develop CNS metastases, recur-
The pathophysiology of tumor proliferation from a primary rence, or otherwise disproportionate response for
to distant site is best described as a series of sequential steps, intracranial disease burden. While the BBB is partially
each exerting a selective pressure among a heterogeneous responsible, next-­generation therapies with excellent CNS
population of cancer cells. The CNS microenvironment is an penetration still have considerable intracranial progression
intrinsically privileged milieu by virtue of a tightly regulated despite extracranial efficacy [10, 11]. Studies that have
blood–brain barrier (BBB) that resists the entry of most cir- investigated the molecular makeup of cancer throughout a
culating immune and tumor cells. Thus, infiltration of the treatment course have further demonstrated that systemic
BBB plays a key role in the pathogenesis of metastatic therapy itself may lead to selective pressures that favor sub-
spread. Once breached, proliferation in the CNS microenvi- clonal expansion of treatment-resistant tumors [5, 12–14].
ronment still brings forth its own challenges for tumor cells. Genomic heterogeneity is therefore a key mechanism by
Cancer cells that have deposited in the brain employ vascular which tumor cells ultimately invade and proliferate in the
co-option and other changes to modify the BBB to a blood– CNS environment.
tumor barrier (BTB) that fosters neoangiogenesis and may
modify the tumor local microenvironment to a partially open
state. This BTB is likely a critical factor in drug delivery and/ 50.4  ropism for the Central Nervous
T
or tumor behavior and an evolving area of research [2]. System
The multistep process of tumor metastasis, first described
by Dr. Fidler [3], outlines eight key steps beginning with (1) The population of circulating tumor cells (CTC) in the lym-
tumor growth at the primary site, (2) neovascular recruitment phovascular system also has subtle genomic and phenotypic
(angiogenesis), (3) acquired motility of tumor cells, (4) lym- differences compared to tumor cells of the primary site.
phovascular infiltration, (5) intravasation into circulation, (6) Those differences are partly responsible for overcoming the
survival in circulation, (7) extravasation out of circulation, initial barriers to tumor spread. Yet still, not all CTC are des-
and ultimately (8) proliferation at a distant site with second- tined for CNS deposition. A growing body of evidence sug-
ary tumor growth [3, 4] [Fidler 1975, Hart and Fidler 1980]. gests that an additional subset of all CTC harbor the
The focus of subsequent discussions will be on extravasation molecular machinery necessary to overcome the BBB and
out of circulation and proliferation within the central nervous subsequently adapt and thrive in the brain microenviron-
system. ment. Vascular co-option in the CNS is postulated to be an
early initiating factor in BM when CTCs migrate to the ablu-
minal vessel wall where they may travel in a dormant state
50.3  lonal Evolution and Intermetastatic
C along vasculature or establish a focus of brain invasion [15,
Genomic Heterogeneity 16]. As this BM focus grows, neoangiogenesis may ensue
resulting in tortuous vessels, a hypoxic shift in the tumor
The genomic composition of brain metastases harbors key microenvironment (TME), and other changes outlined below
differences from cancer cells at the primary site of tumor. For that result in the heterogeneous blood–tumor barrier that is a
instance, breast cancer with synchronous bone and brain hallmark of BM disease [16].
metastases show genomic differences that lend a selective
advantage for proliferation at that particular target organ [1,
5]. One of the hallmarks of cancer is genomic instability 50.5 Tumor Microenvironment of the Brain
leading to dysregulation of replication and repair. The down-
stream result of these replicative aberrations creates genetic The unique microenvironment of the brain is characterized
heterogeneity, a minority of which will contribute to by a host of nonneuronal cell types that maintain physiologic
­mutations that offer a selective growth advantage in compari- homeostasis to support neurons. The glial cells of the brain,
son to healthy tissue. in order of decreasing frequency, include oligodendrocytes,
Many solid tumors and brain metastases exhibit a growth astrocytes, microglia, endothelial cells, pericytes, ependymal
pattern consistent with branched evolution; a process by cells, and resident/recruited immune cells [17]. Under nor-
which subclonal populations of cancer cells diverge from a mal circumstances, perceived threats to the CNS, such as
common ancestral population. The resultant intratumoral pathogens or tissue damage, initiate an inflammatory
heterogeneity results in genetic variation over time (tempo- response to sequester and thwart the pathogen or aid in tissue
ral heterogeneity) as well as an unbalanced metastatic dis- repair. Local astrocytes and microglia represent the first line
tribution across distant sites (spatial heterogeneity) [5, 6]. of defense and are extremely proficient at sequestering an
50 Brain Metastases: Overview and Molecular Mechanisms 543

Astrocyte

TGFβ (IL-2) NOS

MMP Cytokine
MHCI
release

Neurons
CD40

CD200 CD200R
CD11α
active
iNOS

NO
CD54
PAMP
receptor CD58
BDNF
etc MHCII
HIF1α4 VEGF, PDGF, FGF,
Angiopoietin, SD1α Activated Microglia APCs

Angiogenesis

Fig. 50.1 Adapted from: The role of the immune system in brain metastasis [20] from Curr Neurobiol and used under a CC-BY 3.0 license. A
depiction of tumor cells in circulation extravasating and interacting with glia and neuropil

insult (Fig. 50.1). When unsuccessful at stopping a pathogen Bidirectional signaling between tumor cells and reactive
or for large areas of tissue damage, the inflammatory cascade astrocytes creates an environment that both promotes tumor
intensifies with elevated proinflammatory cytokines and che- survival and enhances resistance to chemotherapy [22, 23].
mokines. The downstream recruitment of immune cells in While the functional interactions of specific cancer subtypes,
addition to persistent activation of astrocytes and microglia cytokines, and chemokines are incompletely understood,
promote increased blood vessel permeability, cytotoxicity, there is a significant body of direct and indirect evidence
and later astrogliosis to protect normal brain [18]. Extensive from preclinical models. In the example of lung cancer, pro-
preclinical studies, mostly on murine models, have demon- inflammatory cytokines TNF-a, IL-1beta, and IL-6 enhance
strated that metastatic cancer cells hijack this physiologic in vitro tumor proliferation [24]. Interleukin-1beta has been
inflammatory response to create a “metastatic niche” in implicated in breast cancer metastasis to the brain whereby
which cancer cells may thrive [19]. astrocytes activated by IL-1beta in turn enhanced cancer
stem-cell (CSC) self-renewal both in vitro and in vivo while
IL-1beta knockdown attenuated metastatic growth [25]. In a
50.6 Role of Astrocytes in the TME clinical context, breast cancer patients with lesions harboring
lower expression of IL-1beta also had longer brain metastasis-­
The initiation of the neuroinflammatory cascade involves free survival compared to those with lesions expressing
activation of astrocytes and upregulation of proinflammatory higher levels of IL-1beta [25]. Breast cancer cells have addi-
genes [21]. Morphological changes in reactive astrocytes tionally been shown to release prostaglandin COX2 to induce
promote scarring, known as gliosis, whereby elongated cell astrocytic release of CCL7, another potent promotor of CSC
processes form a functional demarcation around an area of renewal [26].
injury. Brain metastases similarly demonstrate astrogliosis at While the early astrocytic response to brain tumor inva-
the interface between normal parenchyma and tumor; how- sion is considered tumoricidal by inducing apoptosis via
ever, key astrocytic functions are parasitized to the advantage plasmin activator (PA) secretion, both breast and lung cancer
of the tumor. metastasis have been shown to secrete serpins (PA inhibi-
544 S. Haider et al.

tors) that in turn increase vascular co-option [19, 27]. The macrophages following chemotherapy in comparison to pre-
increased expression of serpins, potent protease inhibitors, in treatment samples. Thus not only do metastatic tumor cells
tumor cells, therefore, thwart cell death signals from astro- exhibit the ability to hijack normal macrophage-mediated
cytes and recruits a nourishing vascular network. Astrocytes proteolytic functions to promote extracellular matrix remod-
and cancer cells also communicate through direct gap junc- eling and tumor cell motility, but they additionally combat
tions by which proinflammatory cytokines further promote the efficacy of cytotoxic therapies [33]. Consequently, we
metastatic proliferation and chemoresistance by upregula- find that the density of tumor-associated macrophages in
tion of survival genes [28–30]. tumor masses positively correlates with tumor size and
While offering some direct and indirect evidence for the inversely correlates with prognosis [31, 34, 35].
role of astrocytes, the aforementioned examples of cytokines
and chemokines are certainly not exhaustive.
50.8 Organotropism for the Brain

50.7  ole of Macrophages and Microglia

R Cancer cells that have left the primary site but not yet left the
in the TME lymphovascular system are known as circulating tumor cells
(CTC). Extravasation and transmigration out of circulation
Macrophages and microglia are abundant within the typical with deposition at a distant site are mediated via cancer-­
neuroinflammatory cascade as well as the maladaptive specific and organic-specific mechanisms. Stephen Paget, an
inflammatory environment of metastatic lesions. English surgeon of the late nineteenth century, popularized
Macrophages recruited by metastases help create a hospita- the “seed and soil” hypothesis in which CTC is the “seed”
ble niche for tumor cells, induce trophic changes, and pro- and metastasis-permissive end organs are the “soil”; the
mote angiogenesis. Given that these innate immune cells analogy extends the concept that seeds may spread indis-
have a stable genome and do not typically undergo malignant criminately yet require a particular, favorable type of soil
mutations, they provide an attractive target for potential ther- (milieu) to prosper. The prevailing and competing sentiment
apeutic interventions in the intracranial compartment [31]. at the time, championed by Virchow, was that metastasis was
One particular protein, Cathepsin S, has been shown to be entirely explained by hematogenous spread alone. Paget,
expressed by both tumor cells and macrophages and exerts a however, in examining over 700 autopsies of breast cancer
strong influence in promoting BBB transmigration as well as patients found significant discrepancies between metastases
proliferation in the brain metastatic microenvironment. The to liver and bone with relative sparing of the spleen—a find-
early stages of metastatic invasion show high expression of ing incompatible with Virchow’s theory alone [36].
cathepsin S initially within tumor cells which subsequently More recent literature provides evidence that a complex,
declines in late-stage lesions where Cathepsin S expression bidirectional interplay exists between tumor cells and desti-
shifts to stromal macrophages. In a clinical context, high lev- nation organs. The notion of a premetastatic niche (PMN) by
els of cathepsin S at the primary tumor site are associated which soluble factors derived from a primary tumor site
with reduced brain metastasis-free survival. While macro- modify the microenvironment of distant sites to make them
phages are the dominant source of cathepsin S, knockdown more hospitable to CTCs has gained traction for multiple
models of Cathepsin S show a reduction in experimental cancer and target organ subtypes. An example of the pre-
brain metastasis only occurs when both tumor- and metastatic niche is provided by the observation that breast
macrophage-­derived cathepsin S are depleted [32]. The pro- cancer cells at the primary site secrete extracellular vesicles
teolytic network provided by tumor-associated macrophages containing noncoding DNA that modulate distant brain
additionally promotes remodeling of the stromal extracellu- endothelial cells to downregulate phosphoinositide-­
lar matrix to create an environment suitable for tumor expan- dependent kinase 1 (PDPK1), thereby weakening the BBB
sion [33]. and facilitating transmigration of CTCs [37, 38].
In the instance of breast cancer metastasis, tumor-­ The most common primary tumors to metastasize to brain
associated macrophages also exert a protective effect against include lung, melanoma, breast, renal, and colorectal cancer
cytotoxic chemotherapies such as paclitaxel, etoposide, and with significantly lower rates of skin and soft tissue metasta-
doxorubicin through both cathepsin-dependent and sis relative to population incidence [21]. Initiation of meta-
cathepsin-­independent mechanisms. Further evidence is sup- static colonization necessitates the expression of molecular
ported by the observation that cathepsin inhibition conse- machinery that (a) overcomes the BBB and (b) thrives in
quently improved the efficacy of paclitaxel and etoposide conditions of relative hypoxia and hypoglycemia [21, 39–
in vivo. In an orthotopic murine model using breast cancer 41]. Transcriptome analysis of brain-metastatic tumor cells
samples before and after neoadjuvant chemotherapy, investi- indeed reveals driver mutations unappreciated in matched
gators found a significant accumulation of tumor-associated primary tumors or extracranial metastatic cells [6]. Additional
50 Brain Metastases: Overview and Molecular Mechanisms 545

epigenetic alterations at distinct tumor-specific loci are also References

found in similar comparisons [21].
While our understanding of why a particular tumor may 1. Klotz R, Thomas A, Teng T, et al. Circulating tumor cells exhibit
metastatic tropism and reveal brain metastasis drivers. Cancer
metastasize to the brain more frequently than other sites is
Discov. 2020;10(1):86–103. https://doi.org/10.1158/2159-­8290.
still incomplete, studies have identified aberrant genetic and CD-­19-­0384.
epigenetic phenotypes geared toward homing CTCs specifi- 2. Lockman PR, Mittapalli RK, Taskar KS, et al. Heterogeneous
cally to the CNS and promoting their subsequent prolifera- blood-tumor barrier permeability determines drug efficacy in
experimental brain metastases of breast cancer. Clin Cancer
tion. Primary tumor subtypes have also been implicated in
Res. 2010;16(23):5664–78. https://doi.org/10.1158/1078-­0432.
differential organotropism for the brain; the epidermal CCR-­10-­1564.
growth factor receptor (EGFR) mutation and anaplastic lym- 3. Fidler IJ. Mechanisms of cancer invasion and metastasis. In:
phoma kinase (ALK) rearrangement show increased neurot- Becker FF, editor. Biology of tumors: surfaces, immunology, and
comparative pathology. Springer US; 1975. p. 101–31. https://doi.
ropism in non-small cell lung cancer [42, 43], BRAF
org/10.1007/978-­1-­4684-­2736-­3_3.
mutations driving the MAPK pathway in melanoma metasta- 4. Hart IR, Fidler IJ. Cancer invasion and metastasis. Q Rev Biol.
ses [42], and both triple-negative and Her-2 overexpression 1980;55(2):121–42. https://doi.org/10.1086/411730.
in breast cancer [44, 45]. The aforementioned cancer sub- 5. Dagogo-Jack I, Carter SL, Brastianos PK. Brain metastasis:
clinical implications of branched evolution. Trends in Cancer.
types harbor actionable mutations-specific targeted therapies
2016;2(7):332–7. https://doi.org/10.1016/j.trecan.2016.06.005.
currently available and under investigation can be found in 6. Brastianos PK, Carter SL, Santagata S, et al. Genomic characteriza-
subsequent chapters within this section. tion of brain metastases reveals branched evolution and potential
therapeutic targets. Cancer Discov. 2015;5(11):1164–77. https://
doi.org/10.1158/2159-­8290.CD-­15-­0369.
7. Kang J, Zhang C, Zhong W-Z. Neoadjuvant immunotherapy
50.9 Conclusion for non-small cell lung cancer: state of the art. Cancer Commun
(Lond). 2021;41(4):287–302. https://doi.org/10.1002/cac2.12153.
The prognosis for patients with brain metastases remains 8. Keilson JM, Knochelmann HM, Paulos CM, Kudchadkar RR, Lowe
MC. The evolving landscape of immunotherapy in solid tumors.
poor. A primary tumor’s ability to successfully invade and
J Surg Oncol. 2021;123(3):798–806. https://doi.org/10.1002/
thrive in the central nervous system is dependent on genetic jso.26416.
and epigenetic alterations conferring selective advantages to 9. Park YH, Park MJ, Ji SH, et al. Trastuzumab treatment improves
overcome each step in the multistep model of tumor metasta- brain metastasis outcomes through control and durable prolonga-
tion of systemic extracranial disease in HER2-overexpressing
sis. A better understanding of the mechanisms for neurotro-
breast cancer patients. Br J Cancer. 2009;100(6):894–900. https://
pism as well as key interactions between tumor and glia doi.org/10.1038/sj.bjc.6604941.
within the CNS tumor microenvironment holds promising 10. Ou S-HI, Ahn JS, De Petris L, et al. Alectinib in crizotinib-­refractory
avenues as the basis for developing novel therapeutics as ALK-rearranged non-small-cell lung cancer: a phase II global
study. J Clin Oncol. 2016;34(7):661–8. https://doi.org/10.1200/
well as identifying those at increased risk for brain
jco.2015.63.9443.
metastasis. 11. Niu J, Gelbspan D, Weitz D, Markman M, Quan W. HER2-positive,
trastuzumab-resistant metastatic esophageal cancer presenting with
brain metastasis after durable response to dual HER2 blockade: a
case report. J Gastrointest Oncol. 2014;5(6):E103–8. https://doi.
org/10.3978/j.issn.2078-­6891.2014.045.
12. Bhang HC, Ruddy DA, Krishnamurthy Radhakrishna V, et al.
Open Questions Studying clonal dynamics in response to cancer therapy using high-­
• How can circulating tumor cells be screened to complexity barcoding. Nat Med. 2015;21(5):440–8. https://doi.
org/10.1038/nm.3841.
determine the risk of brain metastasis?
13. Sequist LV, Waltman BA, Dias-Santagata D, et al. Genotypic
• What factors promote the survival of circulating and histological evolution of lung cancers acquiring resistance to
tumor cells? EGFR inhibitors. Sci Transl Med. 2011;3(75):75ra26. https://doi.
• How may resident immune cells of the central ner- org/10.1126/scitranslmed.3002003.
14. Yates LR, Gerstung M, Knappskog S, et al. Subclonal diversifica-
vous system be targeted to treat BM?
tion of primary breast cancer revealed by multiregion sequencing.
• What is the role of the immune system with respect Nat Med. 2015;21(7):751–9. https://doi.org/10.1038/nm.3886.
to the pre-metastatic niche? 15. Kienast Y, von Baumgarten L, Fuhrmann M, et al. Real-time imag-
• How may vascular co-option by tumor cells be tar- ing reveals the single steps of brain metastasis formation. Nat Med.
2010;16(1):116–22. https://doi.org/10.1038/nm.2072.
geted to treat BM?
16. Arvanitis CD, Ferraro GB, Jain RK. The blood–brain bar-
rier and blood–tumour barrier in brain tumours and metastases.
546 S. Haider et al.

Nat Rev Cancer. 2020;20(1):26–41. https://doi.org/10.1038/ 31. Condeelis J, Pollard JW. Macrophages: obligate partners for tumor
s41568-­019-­0205-­x. cell migration, invasion, and metastasis. Cell. 2006;124(2):263–6.
17. DiSabato DJ, Quan N, Godbout JP. Neuroinflammation: the devil is https://doi.org/10.1016/j.cell.2006.01.007.
in the details. J Neurochem. 2016;139(Suppl 2):136–53. https://doi. 32. Sevenich L, Bowman RL, Mason SD, et al. Analysis of tumour- and
org/10.1111/jnc.13607. stroma-supplied proteolytic networks reveals a brain-metastasis-­
18. Schwartz M, Raposo C. Protective autoimmunity: a unifying model promoting role for cathepsin S. Nat Cell Biol. 2014;16(9):876–88.
for the immune network involved in CNS repair. Neuroscientist. https://doi.org/10.1038/ncb3011.
2014;20(4):343–58. https://doi.org/10.1177/1073858413516799. 33. Shree T, Olson OC, Elie BT, et al. Macrophages and cathepsin pro-
19. Doron H, Pukrop T, Erez N. A blazing landscape: neuroinflam- teases blunt chemotherapeutic response in breast cancer. Genes Dev.
mation shapes brain metastasis. Cancer Res. 2019;79(3):423–36. 2011;25(23):2465–79. https://doi.org/10.1101/gad.180331.111.
https://doi.org/10.1158/0008-­5472.CAN-­18-­1805. 34. Medrek C, Pontén F, Jirström K, Leandersson K. The presence
20. Leibold AT, Monaco GN, Dey M. The role of the immune system in of tumor associated macrophages in tumor stroma as a prognos-
brain metastasis. Curr Neurobiol. 2019;10(2):33–48. tic marker for breast cancer patients. BMC Cancer. 2012;12:306.
21. Yuzhalin AE, Yu D. Brain metastasis organotropism. Cold Spring https://doi.org/10.1186/1471-­2407-­12-­306.
Harb Perspect Med. 2020;10(5):a037242. https://doi.org/10.1101/ 35. Morrison C. Immuno-oncologists eye up macrophage targets.
cshperspect.a037242. Nat Rev Drug Discov. 2016;15(6):373–4. https://doi.org/10.1038/
22. Langley RR, Fan D, Guo L, et al. Generation of an immortal- nrd.2016.111.
ized astrocyte cell line from H-2Kb-tsA58 mice to study the role 36. Fidler IJ, Poste G. The “seed and soil” hypothesis revis-
of astrocytes in brain metastasis. Int J Oncol. 2009;35(4):665–72. ited. Lancet Oncol. 2008;9(8):808. https://doi.org/10.1016/
https://doi.org/10.3892/ijo_00000378. S1470-­2045(08)70201-­8.
23. Lin Q, Balasubramanian K, Fan D, et al. Reactive astrocytes protect 37. Tominaga N, Kosaka N, Ono M, et al. Brain metastatic cancer cells
melanoma cells from chemotherapy by sequestering intracellular release microRNA-181c-containing extracellular vesicles capable
calcium through gap junction communication channels. Neoplasia. of destructing blood-brain barrier. Nat Commun. 2015;6:6716.
2010;12(9):748–54. https://doi.org/10.1593/neo.10602. https://doi.org/10.1038/ncomms7716.
24. Seike T, Fujita K, Yamakawa Y, et al. Interaction between lung 38. Zhou W, Fong MY, Min Y, et al. Cancer-secreted miR-105 destroys
cancer cells and astrocytes via specific inflammatory cytokines in vascular endothelial barriers to promote metastasis. Cancer Cell.
the microenvironment of brain metastasis. Clin Exp Metastasis. 2014;25(4):501–15. https://doi.org/10.1016/j.ccr.2014.03.007.
2011;28(1):13–25. https://doi.org/10.1007/s10585-­010-­9354-­8. 39. Peters A, Schweiger U, Pellerin L, et al. The selfish brain: competi-
25. Xing F, Kobayashi A, Okuda H, et al. Reactive astrocytes promote tion for energy resources. Neurosci Biobehav Rev. 2004;28(2):143–
the metastatic growth of breast cancer stem-like cells by activat- 80. https://doi.org/10.1016/j.neubiorev.2004.03.002.
ing Notch signalling in brain. EMBO Mol Med. 2013;5(3):384–96. 40. Winkler F. The brain metastatic niche. J Mol Med (Berl).
https://doi.org/10.1002/emmm.201201623. 2015;93(11):1213–20. https://doi.org/10.1007/s00109-­015-­1357-­0.
26. Wu K, Fukuda K, Xing F, et al. Roles of the cyclooxygenase 2 matrix 41. Lowery FJ, Yu D. Brain metastasis: unique challenges and open
metalloproteinase 1 pathway in brain metastasis of breast cancer. J opportunities. Biochim Biophys Acta Rev Cancer. 2017;1867(1):49–
Biol Chem. 2015;290(15):9842–54. https://doi.org/10.1074/jbc. 57. https://doi.org/10.1016/j.bbcan.2016.12.001.
M114.602185. 42. Ni W, Chen W, Lu Y. Emerging findings into molecular mechanism
27. Valiente M, Obenauf AC, Jin X, et al. Serpins promote cancer of brain metastasis. Cancer Med. 2018;7(8):3820–33. https://doi.
cell survival and vascular co-option in brain metastasis. Cell. org/10.1002/cam4.1667.
2014;156(5):1002–16. https://doi.org/10.1016/j.cell.2014.01.040. 43. Rangachari D, Yamaguchi N, VanderLaan PA, et al. Brain metasta-
28. Dai W, Zhu H, Chen G, et al. Orchestration of the crosstalk between ses in patients with EGFR-mutated or ALK-rearranged non-small-­
astrocytes and cancer cells affects the treatment and progno- cell lung cancers. Lung Cancer. 2015;88(1):108–11. https://doi.
sis of lung cancer sufferers with brain metastasis. J Thorac Dis. org/10.1016/j.lungcan.2015.01.020.
2016;8(11):E1450–4. https://doi.org/10.21037/jtd.2016.11.11. 44. Shen Q, Sahin AA, Hess KR, et al. Breast cancer with brain metas-
29. Kim S-J, Kim J-S, Park ES, et al. Astrocytes upregulate survival tases: clinicopathologic features, survival, and paired biomarker
genes in tumor cells and induce protection from chemotherapy. analysis. Oncologist. 2015;20(5):466–73. https://doi.org/10.1634/
Neoplasia. 2011;13(3):286–98. https://doi.org/10.1593/neo.11112. theoncologist.2014-­0107.
30. Hanibuchi M, Kim S-J, Fidler I, Nishioka Y. The molecular biology 45. Jin L, Han B, Siegel E, Cui Y, Giuliano A, Cui X. Breast cancer
of lung cancer brain metastasis: an overview of current comprehen- lung metastasis: molecular biology and therapeutic implications.
sions and future perspectives. J Med Investig. 2014;61(3.4):241– Cancer Biol Ther. 2018;19(10):858–68. https://doi.org/10.1080/15
53. https://doi.org/10.2152/jmi.61.241. 384047.2018.1456599.
 Indications and Techniques for Surgical
Intervention in Patients with Metastatic 51
Brain Tumors

Jacob A. Pawloski, Omar Awan, Mateo Ziu,

and Adam M. Robin

Abstract
Learning Objectives
Brain metastases occur in up to 30% of cancer patients
and may be the presenting problem in as many as 20%. • Describe the indications for surgical intervention in
Advances in our understanding of cancer biology and pre- the patient with a brain metastasis
cision therapeutics have resulted in improved systemic • Gain familiarity with surgical techniques and their
disease control for many cancers, making the treatment of applications for treatment for the patient with a
metastatic disease to the central nervous system (CNS) brain metastasis
more critical than ever. Surgery and adjuvant radiation • Understand the rationale for surgical adjuncts that
have long been the mainstay of treatment. Surgery can help surgeons optimize tumor removal and safety
provide a diagnosis, symptom relief, and ideally locally
curative benefits. Surgical innovations and techniques
such as single or multiple craniotomies for microsurgical
excision, intraoperative brain mapping, minimally inva- 51.1 Introduction
sive parafascicular surgery (MIPS), intraoperative image-­
guided surgical resection, and laser interstitial thermal Brain metastases occur in up to 30% of cancer patients and
therapy (LITT) allow for the optimization of tumor account for over 50% of intracranial tumors [1, 2]. Surgery
removal and patient safety as never before. In this chapter, can provide a diagnosis, symptom relief, and ideally locally
the reader will become familiar with the rationale for the curative benefits. Furthermore, the ability to obtain tissue for
surgical management of the patient with a brain metasta- histopathological and sequencing analysis is an important
sis and many of the techniques driving better surgical out- consideration. While patients with known brain metastatic
comes for our patients. disease are often found to have lesions which are more easily
accessible for tissue diagnosis, there is evidence to suggest
that brain metastases can contain therapeutically relevant
J. A. Pawloski · A. M. Robin (*) molecular alterations not found in the primary lesion [3].
Department of Neurosurgery, Hermelin Brain Tumor Center, Patients with known systemic disease may have other non-
Henry Ford Hospital, Detroit, MI, USA
e-mail: [email protected]; [email protected]
metastatic intracranial lesions (e.g., primary brain tumor or
abscess) for which diagnosis would prompt a shift in treat-
O. Awan · M. Ziu
Inova Neurosciences and Spine Institute, University of Virginia
ment. Surgical resection is the cornerstone of the manage-
Medical School, Falls Church, VA, USA ment of brain metastases and provides unmatched direct
e-mail: [email protected]; [email protected] cytoreduction of local disease. Craniotomy for resection can

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 547
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_51
548 J. A. Pawloski et al.

be a valuable tool for relief of symptomatic mass effect, The histology and location of the metastatic lesions are
malignant edema and intracranial pressure, or obstruction of both important considerations in determining treatment
cerebrospinal fluid (CSF) flow, which can all contribute to options for single or multiple metastases. Certain histologies
significant mortality and morbidity. In addition to tumor such as small cell lung cancers or CNS lymphomas may be
resection, surgery plays a critical role in a number of other effectively treated with radiation and/or chemotherapy alone,
ways. In the case of leptomeningeal disease, CSF diversion such that surgery is not indicated as a first-line therapy even
in the form of ventriculoperitoneal shunting (VPS) or endo- for large tumors while melanoma, renal cell carcinoma, and
scopic third ventriculostomy (ETV) can be employed to pro- sarcoma are relatively radioresistant and are more amenable
vide palliation for symptoms of hydrocephalus. Furthermore, to surgical resection. With regard to tumor location, the func-
surgical intervention can be used to implant intraventricular tional anatomy of the surrounding brain tissue must be con-
devices for infusion of intrathecal therapeutics or following sidered. Epidemiological studies have shown that 80–85% of
resection for placement of material imbued with chemothera- metastatic lesions are supratentorial and 10–15% affect the
peutics or radioactive isotopes for delivery of brachytherapy, cerebellum or brainstem [9]. Involvement of eloquent brain
allowing for higher drug concentrations in the CSF or peritu- areas may result in surgical morbidity that could delay or
moral region than can be achieved in the plasma or for local even preclude adjuvant therapies and offset any potential
radiotherapy in the setting of previously radiated lesions. survival benefit conferred by surgery [10]. As will be dis-
Over the last 20 years, innovations in the field of brain cussed later, neurosurgical advances now allow for access to
tumor surgery have provided the neurosurgeon with an lesions affecting almost any area of the brain safely but the
expansive armamentarium to treat brain metastases, allowing decision to operate should always be made within the appro-
for the design of surgical plans that are tailored to the patient priate clinical context including the patient’s tumor charac-
and their disease. Herein, we will review the role of surgical teristics, prognosis from their systemic disease, likelihood of
intervention in patients with brain metastases with a special neurologic improvement following surgery, and other
focus on techniques that are increasingly utilized to optimize comorbid medical conditions that may impact the risks of
tumor removal while minimizing perioperative morbidity. surgical intervention.
Discussion to this point has centered around surgical inter-
vention involving craniotomy and removal of lesions with the
51.2 Surgical Indications objective of decreasing mass effect and obtaining local con-
trol via direct cytoreduction. Alternatively, stereotactic biopsy
51.2.1 Single or Dominant Brain Metastasis is a minimally invasive procedure used to obtain tissue for
diagnosis only. Biopsy is indicated in instances where tumors
Routine use of craniotomy for resection of a solitary metastatic are inaccessible surgically, there is an unknown primary can-
lesion was first shown to be beneficial in the landmark random- cer, or craniotomy is contraindicated due to poor prognosis or
ized controlled trial performed by Patchell and colleagues [4]. overall health. Epidemiologically, brain metastasis is more
Surgery followed by whole-brain radiation therapy (WBRT) than twice as common as primary brain tumors yet in as many
improved overall survival, functional independence, and pro- as 15% of cases a primary original of brain metastasis cannot
gression-free survival compared to WBRT alone [5]. However, be identified [11, 12]. Further, almost 20% of patients can
studies have also shown that surgical resection alone is insuffi- present with brain metastasis as the first symptom of meta-
cient, with local recurrence rates ranging from 40 to 70% [6, 7]. static cancer [13]. Due to the relatively low extracranial tumor
Thus, surgery should be considered as a cornerstone of combina- burden, patients with brain metastasis associated with an
tion therapy with adjuvant radiation. unknown primary present with a therapeutic window where
Indications favoring surgical resection include tumor size early biopsy and tissue diagnosis may be associated with
>3 cm, location in the posterior fossa associated with or at improved outcomes when targeted therapy is available [13].
risk for development of obstructive hydrocephalus, and This point is underscored when treating lesions in which a
lesions causing significant symptoms due to mass effect. diagnosis of a chemo or radiation-­sensitive histology is sus-
Individuals with good functional status in whom the sys- pected, such as lymphoma or small-cell lung cancer.
temic disease is well-controlled are the best candidates for In summary, surgical resection remains the foundation of
surgical intervention, while those with limited life e­ xpectancy management for metastatic disease to the central nervous
due to their advanced oncologic disease or significant medi- system and has been shown to increase survival and decrease
cal comorbidities may benefit more from less invasive treat- rates of local recurrence, especially when used in combina-
ment. Current evidence suggests that stereotactic radiosurgery tion with adjuvant radiation therapy. Unless patient factors
(SRS) alone is as effective as surgery plus WBRT, and thus preclude open surgery, resective surgery can improve func-
for tumors smaller than 3 cm, surgery is often reserved as tional status and halt or improve progressive neurologic defi-
salvage therapy for those that recur following SRS [8]. cit in as many as 90% of cases [14, 15].
51 Indications and Techniques for Surgical Intervention in Patients with Metastatic Brain Tumors 549

51.2.2 Multiple Brain Metastases mass effect. While both tumor progression and RN can pres-
ent similarly, distinguishing one from the other may have
As many as half of patients with brain metastasis present profound implications on care. Moreover, with increasing
with multiple lesions [16]. For the most part, studies evaluat- recognition of the heterogeneity between brain metastases
ing the role of surgical resection for patients with multiple molecular profiling may inform selection of precision thera-
brain metastases have been limited. These studies have peutics. When encountering suspected local tumor recur-
shown that patients who had resection of all accessible rence after radiotherapy, surgical management is dependent
lesions had survival rates similar to patients with single on the size, location, and symptoms attributable to the sus-
lesion who underwent resection, although patients in these pected recurrence. If the lesion recurs with significant edema
studies typically had resection of three or fewer lesions [14, and mass effect in a patient who is a good surgical candidate,
17]. Patients with expected overall survival of less than craniotomy for resection of the recurrent tumor or RN is rea-
3 months due to systemic disease should not be considered sonable. Less impressive progressive enhancement and asso-
for surgical intervention. Instead, patients with limited or ciated edema on serial MRIs following radiation therapy
controlled systemic disease and accessible tumors should be should be considered for stereotactic biopsy to differentiate
considered for surgical resection—even when not all tumors recurrence from RN. In these cases, the addition of LITT has
can be resected, considering that complications from surgery become a valuable surgical tool as it has the capacity to treat
are as low as 7–9% [14, 17]. In general, the same selection both tumor recurrence and RN in a minimally invasive fash-
criteria for surgical candidacy for single lesion resection may ion with little surgical footprint beyond that which is required
also be applied to patients with multiple metastasis (namely, for stereotactic needle biopsy alone.
age, KPS < 70, and extracranial disease burden) with similar
outcomes even if all lesions are not amendable to resection.
Studies suggest that patients with more than three or four 51.3 Surgical Techniques
lesions, especially with poor preoperative performance sta-
tus, are unlikely to benefit from surgical resection and should 51.3.1 Decision-Making in the Surgical
be considered for radiation therapy instead. Treatment of Brain Metastases

When operating on cancer patients with brain metastases, the

51.2.3 P
 rogressive or Recurrent Brain risks of resective surgery must be balanced with the patient’s
Metastasis overall prognosis from their systemic disease, their likeli-
hood of neurologic improvement or decline with surgery,
As individuals with brain metastases continue to live longer and the alternative treatment options available, among other
after their initial treatment, cases of tumor recurrence or pro- factors. Maximal resection and adjuvant radiotherapy can
gression are becoming more common and are reported in provide durable local control and improved quality of life.
10–40% of individuals with brain metastases [18, 19]. Local Overall survival following resection of brain metastasis aver-
control is impacted by a variety of factors including histol- ages from 10 to 14 months, but can exceed 5 years in select
ogy, tumor size, adjuvant radiotherapy to the resection cav- patient populations [1, 10, 17, 21]. Likewise, neurologic
ity, as well as method of surgical removal (e.g., en bloc vs. complications from surgery can have significant negative
piecemeal resection) [19]. There is some scarce evidence consequences, with one study finding overall survival was
that reoperation can lead to long-term survival in select reduced from 12.6 to 2.4 months in individuals with new
patients: younger patients, good functional status, and a lon- postoperative neurologic deficits [10]. This underscores the
ger time from initial surgery to recurrence [15, 17, 20].. importance of using tailored surgical approaches and surgi-
Aggressive surgical resection of recurrent metastatic lesions cal adjuncts to minimize operative morbidity (Fig. 51.1).
is a reasonable option in patients who have exhausted
­noninvasive treatment options and retain good functional sta-
tus with controlled systemic disease. 51.3.2 C
 raniotomy for Solitary or Single
Brain biopsy of a suspected recurrent metastasis is useful Dominant Metastases in Eloquent
in patients who have previously received radiotherapy, as Locations
tumor recurrence and radiation necrosis (RN) can both mani-
fest with nearly identical clinical and radiographic presenta- Brain metastases occurring in speech or primary somatosen-
tions. Radiation necrosis can occur at almost any time from sory regions inherently carry a higher risk of neurologic dys-
a few months to years following radiation therapy and can function. The risk of neurologic deficit without treatment or
appear radiographically identical to disease recurrence with during treatment, particularly surgery can be quite high. SRS
increasing contrast enhancement, worsening edema, and is a reasonable option for tumors in eloquent regions, but due
550 J. A. Pawloski et al.

Brain
Metastasis
Diagnosis

Symptomatic
Known primary?
Yes No or large lesion Yes

No

Biopsy Biopsiable
Radiotherapy Radiosensitive? alternative systemic
Yes Yes
site lesions

No Surgical
Recurrence No Management

Large, Surgery for

Stereotactic
LITT or symptomatic, Recurrence?
biopsy +/–
Surgical No Mass effect,
radiotherapy
Resection hydrocephalus

Yes

Consider surgical
Systemic
adjuncts: iMRI, LITT,
Poor, KPS < 70, LMD Stable, KPS > 70
Disease Control brachytherapy, intra-op
chemo, clinical trials

Goals of care
discussion,
consider palliative Yes Non-eloquent Eloquent Deep-Seated Lesion 2+ symptomatic lesions
surgery or shunt for
LMD

Tractography
Standard Awake Multiple
with available
Confirm with Craniotomy Craniotomy Craniotomies
Large or Recurrence corridor
symptomatic PET, MR
Suspected
lesion Perfusion
Yes No

No

Minimally-Invasive
LITT on
Repeat radiotherapy, Parafascicular
recurrence
biopsy +/– LITT, clinical trials Surgery

Fig. 51.1 Recommended Treatment Paradigm for the patient presenting with a brain metastasis

to their location, these tumors often manifest with debilitating mapping can be completed using awake or asleep techniques
symptoms that are exacerbated initially with radiation ther- with intraoperative electrophysiologic monitoring and direct
apy. Moreover, radiation necrosis in eloquent regions cortical or subcortical stimulation, while speech mapping
can have catastrophic effects on neurologic function and requires awake craniotomy and the support of a speech and
quality of life. Increasingly, metastases in the primary senso- language pathologist. Various paradigms have been described
rimotor cortex are considered for surgical intervention. to ensure high fidelity results while mapping function in sur-
Successfully dealing with these tumors requires careful plan- gery. A combination of direct visualization, stereotactic
ning and patient evaluation. Patient improvement with preop- image-guided navigation, and ultrasound guides the approach
erative steroid administration is a good prognostic indicator. to the lesion (Fig. 51.2). In the case of a central sulcus metas-
Stereotactic image guidance, intraoperative ultrasound, and tasis, the authors prefer arachnoid dissection of the central
intraoperative neurophysiologic monitoring with or without vein allowing for a transsulcal approach to the tumor. Using a
awake craniotomy are surgical adjuncts that help to localize transsulcal approach obviates the need for dissection through
the tumor augment and tumor removal. The size of the crani- eloquent sensorimotor cortex. Care is taken to preserve criti-
otomy should be planned to provide adequate access for cal vascular structures within the sulcus. Use of intraoperative
resection of the lesion as well as enough exposure to ade- mapping can reliably reduce the associated morbidity of
quately map the adjacent motor or speech centers. Motor metastatic tumor resection in these regions [22, 23].
51 Indications and Techniques for Surgical Intervention in Patients with Metastatic Brain Tumors 551

a b c

Pre-Op

d e f

11 Months Post-Op

Fig. 51.2 Panel (a) T1 axial post-contrast MRI demonstrating a left- skull pins following a complete scalp block. Panel (c) demonstrates dis-
sided brain metastasis of unknown origin interposed between sensory section of the pia adjacent to the central sulcal vein. For reference, the
and motor cortices. Case example of a 48-year-old man with a central cotton patty pictured is 12 mm in size. Panel (d) shows the dissected
sulcus tumor and no known primary disease who underwent an awake central sulcus, splaying of the sensory and motor cortex, and mobiliza-
craniotomy for sensorimotor mapping and trans-central sulcus approach tion of the central vein and arterioles after opening the tumor capsule.
with complete resection. Panel (b) demonstrates the “lazy S” incision Panel (e) shows the post-operative appearance after mobilization of the
used for exposure of the sensorimotor region. Note the absence of a vein and sensory motor cortex to allow for resection of the brain
ventilator circuit and endotracheal tube. The patient’s head is fixed in metastasis

51.3.3 Craniotomy for Multiple Metastases resection for patients with good systemic disease control and
multiple metastases that are accessible via surgery [26].
The safety of surgical resection has improved, and in prac- Patient selection is crucial and should focus on patients
tice, individuals undergo craniotomy for multiple brain who have multiple lesions that are not amenable to SRS and
metastases even if not all lesions can be removed. Iwadate are likely to tolerate one or more craniotomies with a good
et al. reviewed 138 patients with single or multiple metasta- chance of stable or improved neurologic function after sur-
sis treated with surgery and WBRT and concluded that there gery. The authors do not recommend routinely performing
was no difference in overall survival in the single or multiple multiple operations in eloquent regions at the same time as
metastases subgroups [24]. Of note, patients with multiple the multiple sites required for craniotomy can interfere
lesions who underwent surgery but not all lesions were with the feasibility of brain mapping. Additionally, care
resected had shorter overall survival, however, treating the should be taken to accurately register stereotactic neuro-
remaining lesions with SRS is equally effective [25]. Based navigation for incision and craniotomy planning. Incisions
on these and other evidence, the most recent guidelines from must not further compromise blood flow and wound heal-
the Congress of Neurological Surgeons recommend surgical ing, particularly in patients with brain metastases who
552 J. A. Pawloski et al.

a b

Pre-Op DTI Fiber Tractography

c d

3D Surgical Plan Intra-op Ultrasound

e f

Tubular Retractor Microdissection Post-Op

Fig. 51.3 Case example of a 57-year-old woman with bifrontal breast surgical (MIPS) approach for complete resection of both lesions. 3D
metastases treated with multiple craniotomies in a single session utiliz- tractography/surgical planning was utilized before and during surgery
ing a traditional craniotomy and a minimally invasive parafascicular as was intraoperative ultrasound to optimize safety and tumor removal

likely will undergo multiple radiotherapy treatments and 51.3.4 M

 inimally Invasive Approaches
when possible, the authors utilize a single incision. As dis- to Metastatic Tumors
cussed at length in later sections, minimally invasive
options such as laser ablation and minimally invasive tubu- Metastatic tumors in deep-white matter areas pose a chal-
lar retractor systems can be especially useful in treating lenge to the surgeon due to the high risk of injuring travers-
multiple lesions with minimal manipulation of adjacent ing fiber tracts while accessing and resecting the tumor.
structures (Fig. 51.3). Radiation has been utilized in circumstances where tumors
51 Indications and Techniques for Surgical Intervention in Patients with Metastatic Brain Tumors 553

are not easily accessible via standard craniotomy. cal technique, increasingly utilized for patients with brain
Alternatively, laser ablation may be used to provide patho- metastases. The goal of tumor cytoreduction remains, only it
logic diagnosis as well as cytoreduction of deep-seated is achieved via laser ablation rather than excision, and the
lesions. Minimally invasive parafascicular microsurgery tumor is accessed with stereotactic placement of a laser fiber
(MIPS) has also been employed to retain the advantages and via stab incision and twist drill hole in the skull in lieu of a
flexibility of open craniotomy while reducing the amount of scalp flap and standard craniotomy (Fig. 51.4). Preoperative
white matter injury and retraction on adjacent or superficial planning is paramount and focuses on utilizing high-­
structures. Preoperative planning in these cases is important resolution anatomic imaging, tractography, three-­
and advances in MRI diffusion tensor imaging (DTI) and dimensional volumetric analysis with build out of volumes
fiber tractography allow for planning surgical approaches for tumor and other nearby eloquent structures to help ensure
with the intent of avoiding crucial white matter tracts. that the laser fibers are placed to maximize lesion ablation.
Much of the planning is based on surgeon experience
51.3.4.1  inimally Invasive Parafascicular
M whereby heat sinks such as the ventricular system, CSF cis-
Surgery (MIPS) terns, and large vessels are accounted for in order to optimize
Tubular retractor systems, also known as port-based or heat dispersion out from the laser in a controlled fashion to
channel-­based systems, are cylindrical or ovoid retractors, achieve cytoreduction and ablation of tumor or radiation
which allow for brain cannulation using a sulcus or small necrosis. Using multiple circumferentially oriented or direc-
corticotomy to provide access to deep-seated subcortical tionally oriented laser fibers and withdrawing the fibers dur-
lesions. Intensive preplanning utilizing high-resolution ana- ing the ablation process, the surgeon can define an ablation
tomic studies and DTI white matter tractography are critical zone that correlates well with lesions up to 10–20 cc and
for this technique. Algorithms that enable 3D visualization beyond. Briefly, the technique involves placement of a cath-
of important fiber bundles aid in the identification of safe eter using stereotactic guidance, through which a laser fiber
surgical corridors along white matter tracts. Accessing these is used to induce thermal injury to target tissue by controlled
corridors allows for safer access to lesions previously asso- heating. A critical step involves intraoperative MRI for con-
ciated with prohibitive morbidity using more traditional sur- firmation of adequate laser fiber position and for precise tem-
gical approaches. This can be achieved with minimal perature control using MR thermometry. The placement of
disruption and manipulation of the surrounding brain the fiber is nearly identical to that of placing a biopsy needle
(Fig. 51.3). A key benefit of the cylindrical retractors with an and ablative therapy can be performed during the same pro-
atraumatic cannulating tip is that they are minimally trau- cedure following tissue sampling.
matic during brain cannulation and distribute the retraction The majority of literature pertaining to the use of LITT
force circumferentially, preventing the sheering of white for metastatic disease is in the context of recurrence or neo-
matter and pressure-­related ischemia that can be associated adjuvant ablation following SRS. In treatment of recurrence
with bladed retractors used in a standard open craniotomy, following SRS, a situation complicated by relatively few
often with very favorable patient outcomes [27, 28]. A larger treatment options, one systematic review found median over-
series of 50 patients with metastatic and primary CNS all survival following complete ablation ranging from 6 to
lesions with median depth of 5.4 cm below the cortex under- 20 months [30]. Significantly, in this study no patient in
went port-based resection with 94% of patients having a whom complete ablation occurred suffered from disease pro-
stable or improved KPS postoperatively [29]. Compared to gression during the follow-up period. A prospective study of
laser ablation techniques, tubular retractors provide the patients undergoing LITT for metastatic tumor recurrence
added benefit of debulking, which makes them more suit- found an overall survival of 64.5% at 26 weeks [31].
able for lesions that present with significant mass effect. Several studies have shown the most significant limitation
Despite their diminutive size, some systems as small as regarding the success of LITT to be tumor size, and accord-
11 mm in diameter, there has been a proliferation of bayo- ingly the proportion of tumor which is within the ablated
netted, single-shaft instruments some of which are lighted, area. While multiple staged procedures may be undertaken to
allowing for two-handed microsurgery. Overall, tubular address larger tumors, this is less favorable as it can increase
retractor-based approaches are a promising new addition to the risk of procedural complications and delay initiation of
surgical management of metastatic disease by offering a adjuvant therapies. Although there are no specific guidelines
safer alternative for resection of deep-seated lesions as com- regarding tumor size, studies have shown that tumors as
pared to traditional craniotomy. large as 5 cm may be associated with benefit in terms of
progression-­free survival [32]. Further, stereotactic trajecto-
51.3.4.2 Laser Ablation for Brain Metastasis ries may be limited due to the patient’s anatomy. Larger and
Laser interstitial thermal therapy (LITT), also known as ste- more vascular lesions are considered to be poor targets for
reotactic laser ablation (SLA) is a minimally invasive surgi- thermal ablation and while generally considered as safe or
554 J. A. Pawloski et al.

a b c

Pre-Op Ablation Catheter Placement 2 Months Post-Op

d e f

4 Months Post-Op 7 Months Post-Op 4 Years Post-Op

Fig. 51.4 Case example of a 48-year-old man with 4-year radiographic apy (LITT). Biopsy at the time of LITT treatment was consistent with
follow-up in a patient with a melanoma metastasis after treatment with radiation necrosis
stereotactic radiosurgery (SRS) and then laser interstitial thermal ther-

safer than standard craniotomy, LITT can be associated with standard treatment, for deeply seated, surgically inaccessible
post-procedural hemorrhage or swelling, and is not as suit- lesions, and patients too frail to tolerate surgical stress. As it
able for treatment of lesions causing significant mass effect. offers significant local control and survival benefits in
In one institutional series, there was a 4% risk of hema- patients who have already failed radiation therapy and may
toma formation at the ablation site requiring evacuation [33]. not be candidates for craniotomy, LITT may be a valuable
Lesions with volume >10 cc are associated with an increased tool for salvage therapy. While promising for other indica-
risk of perioperative complications such as seizures or neu- tions and clinical scenarios, further study is yet necessary.
rologic deficit and a lower rate of gross total ablation [34,
35]. In these cases, there is evidence that adjuvant SRS to the
residual tumor within a month following surgery can 51.3.5 I ntraoperative MRI and Optical
decrease the rate of tumor progression [35]. Guidance Technology
Overall, LITT is a novel therapy which has been shown to
be particularly useful in treating patients suffering from There is substantial evidence demonstrating that gross total
recurrent metastatic disease, or disease that is refractory to resection of brain metastasis improves overall survival and
51 Indications and Techniques for Surgical Intervention in Patients with Metastatic Brain Tumors 555

decreases risk of local progression as compared to subtotal Leptomeningeal spread as a result of surgery is also a sig-
resection [8]. Considerable advances have been made in nificant concern that is more associated with piecemeal
intraoperative imaging technologies to increase the likeli- resection strategies, particularly in the posterior fossa. In a
hood of identifying and resecting all malignant tissue in the large series of patients with a single metastasis the patients
operative field. iMRI has been heavily utilized in glioma sur- who underwent piecemeal resection were 2.7 times more
gery to increase extent of resection, leading to improved out- likely to develop leptomeningeal disease (LMD) than those
comes in glioma patients [36, 37]. iMRI has not been utilized with lesions that were removed en bloc [41]. Melanoma in
as heavily for metastatic disease, likely because these lesions particular was nearly nine times more likely to progress to
are generally less infiltrative than intrinsic tumors and are LMD when resected via piecemeal versus en bloc
therefore more readily delineated. However, in cases of technique.
recurrent metastases or lesions in eloquent locations, iMRI
can be useful to increase the likelihood of gross total resec-
tion. Intraoperative MRI has also been utilized with good 51.3.7 S
 urgical Adjuncts and Palliative
results as an adjunct to MIPS to increase extent of resection Procedures
in deep-­seated lesions which may be difficult to visualize
intraoperatively [38]. 51.3.7.1 Surgical Treatment of Leptomeningeal
In addition to iMRI, there have been advancements in Disease
optical technologies that allow for greater visualization of Leptomeningeal disease (LMD), or carcinomatous menin-
tumor such as large-field florescence with 5-Aminolevulinic gitis, is a dreaded complication for cancer patients due to
acid (5-ALA) and fluorescein sodium, confocal microscopy, the poor prognosis it affords, with overall survival of
and Raman spectroscopy. While a detailed discussion of this 3–6 months after diagnosis of LMD. Current management
diverse group of optical imaging modalities is outside the of LMD focuses primarily on high-dose intrathecal and
scope of this discussion these technologies are potentially systemic chemotherapy, with or without radiotherapy.
transformative for their ability to provide instant tissue char- Surgical interventions for LMD have focused largely on
acterization in the operating room via in vivo or ex vivo eval- palliation of pain related to hydrocephalus or to provide a
uation [39]. route for the safe administration of intrathecal chemother-
apy. Ommaya reservoir (OR) placement simplifies intra-
thecal chemotherapy administration and allows for more
51.3.6 M
 ethods of Surgical Resection: En Bloc standard and uniform dosing as opposed to serial lumbar
Versus Piecemeal Removal punctures. CSF diversion via VPS can also be beneficial in
patients with an obstructive CSF flow pattern or symptom-
Resection of intra-axial lesions is typically done via an en atic communicating hydrocephalus. Patient selection for
bloc resection where the entire lesion is separated from the VPS placement is important as headaches are the most
surrounding parenchyma and removed as a contiguous mass common presenting symptom for LMD, and 40% of
or by a piecemeal resection in which the lesion is entered and patients have an opening pressure >20 cmH20 on initial
then taken out in pieces until all malignant tissue has been lumbar puncture [42, 43]. While symptoms of headache or
removed. Largely speaking, en bloc resection is more favor- nausea may be due to symptomatic hydrocephalus, direct
able when it is possible due to lower complication rates and dural and parenchymal invasion by malignant cells can
less risk of local recurrence and leptomeningeal spread [8]. also account for symptoms of LMD [44]. As with any sur-
The location and size of the mass may make en bloc resec- gical procedure, the decision to operate should be based on
tion impossible without causing significant retraction injury a holistic evaluation that includes the patient’s functional
or disruption of eloquent structures, and in those instances status, systemic disease control, overall prognosis, and
piecemeal removal may be the only option. One large series need/capacity for palliation. One series of patients under-
looking at complication rates between the two approaches going placement of a reservoir-shunt system for CSF
found that resection via the piecemeal approach increased diversion and intermittent chemotherapy administration
the overall 30-day complication rate by 6% across all tumor reported functional improvement in 80% of patients
types and sizes. Notably, in that study there was a significant that was sustained for 6 months postoperatively [45]. A
difference in tumor size as well as tumor functional grade, trial lumbar puncture, or “tap test” may be useful in
such that larger and more eloquent lesions were more likely ­differentiating patients with potential to benefit from CSF
to be removed via the piecemeal approach, suggesting that diversion from those that have headaches or cranial nerve
the en bloc subgroup consisted of tumors with lower risk palsies from direct leptomeningeal invasion (unpublished
characteristics [40]. data).
556 J. A. Pawloski et al.

51.3.7.2 Brachytherapy
Brachytherapy consists of internalized radiotherapy seeds Future Perspectives
deposited at the site of surgical resection, and relies on a
constant, low dose of radiation directly to the tumor bed to • To what degree will increased use of minimally-­
improve local control after surgery. I-125 and Cs-131 are invasive surgical techniques change how practitio-
two of the most common forms of brachytherapy implant, ners view surgically accessible versus inaccessible
however, I-125 is less widespread due to higher rates of tumors?
radiation necrosis reported up to 15% [46]. Cs-131 seeds • Which of the available intra-operative surgical
have similar efficacy with regard to local control and over- adjuncts (brachytherapy, 5-ALA, etc.) will have the
all survival with lower rates of radiation necrosis than greatest impact when utilized as salvage therapy for
I-125 or SRS [47]. Although there is a paucity of large recurrent metastases?
studies with long-term follow-up utilizing brachytherapy, • What factors can assist in differentiating patients
the available evidence suggests there is utility in prevent- with leptomeningeal spread that will benefit from
ing local recurrence and minimizing the toxic effects of CSF diversion via shunting?
traditional radiation. Brachytherapy may be useful as a sal- • How much can biopsy of a metastatic brain tumor
vage therapy, with one series of nine patients with recur- of known pathology for additional molecular char-
rent metastases who underwent repeat operation and acterization inform systemic chemotherapy and
Cs-131 seed placement demonstrating no cases of local radiation treatment regimens?
recurrence at a median follow-up of 10.3 months [48]. The • Are there certain tumor types in which laser abla-
role of brachytherapy in management of recurrent brain tion may be beneficial as first-line therapy instead
metastasis is particularly appealing in patients who have of radiotherapy for appropriately selected small
already received traditional radiation, as a means to re- brain metastases?
irradiate the tumor bed with less risk of radiation necrosis
[49]. Limitations of brachytherapy include decreased pro-
pensity to affect rates of distant recurrence as much as tra- Patient Information and Guidelines
ditional field radiation and availability and access of the American Cancer Society https://www.cancer.org
radioactive materials make broad adoption in clinical prac- Clinical Trials in Metastatic Brain Tumors https://www.can-
tice challenging. cer.gov/about-­cancer/treatment/clinical-­trials/adult-­metastatic-
­brain-­tumors
National Brain Tumor Foundation https://braintumor.org
51.4 Conclusion Family Caregiver Alliance National Center on Caregiving
https://www.caregiver.org/brain-­tumor
As cancer patients achieve better control of their systemic
disease, effective treatment of intracranial metastases is
crucial to improving quality of life and overall survival. References
Neurosurgical advances are paving the way for the next
generation of neuro-oncologic care in which the treatment 1. Bindal RK, Sawaya R, Leavens ME, Hess KR, Taylor
SH. Reoperation for recurrent metastatic brain tumors. J Neurosurg.
plan and surgical approach are tailored to the patient’s 1995;83(4):600–4. https://doi.org/10.3171/jns.1995.83.4.0600.
pathology, treatment goals, and tumor characteristics. 2. Barnholtz-Sloan JS, Sloan AE, Davis FG, Vigneau FD, Lai P,
Minimally invasive and neurophysiology-based approaches Sawaya RE. Incidence proportions of brain metastases in patients
now allow for resection of previously inaccessible lesions, diagnosed (1973 to 2001) in the Metropolitan Detroit Cancer
Surveillance System. J Clin Oncol. 2004;22(14):2865–72. https://
serial biopsy for tissue characterization and advanced tumor doi.org/10.1200/JCO.2004.12.149.
visualization technologies are helping to increase the effec- 3. Brastianos PK, Carter SL, Santagata S, et al. Genomic characteriza-
tiveness and safety of tumor resection. Selected patients tion of brain metastases reveals branched evolution and potential
with recurrent and multiple metastases can still hope to therapeutic targets. Cancer Discov. 2015;5(11):1164–77. https://
doi.org/10.1158/2159-­8290.CD-­15-­0369.
benefit from surgical intervention to improve their overall 4. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ, Maruyama Y,
survival and ­neurologic function. Further advances in laser- Kryscio RJ, Markesbery WR, Macdonald JS, Young B. A ran-
based treatment, optically-based intraoperative tumor visu- domized trial of surgery in the treatment of single metastases
alization technologies, and other intraoperative adjuncts to the brain. N Engl J Med. 1990;322(8):494–500. https://doi.
org/10.1056/NEJM199002223220802.
such as brachytherapy have remarkable potential to shape 5. Patchell RA, Tibbs PA, Walsh JW, et al. A randomized trial of
the next decade of neurosurgical treatment of brain surgery in the treatment of single metastases to the brain. N
metastasis.
51 Indications and Techniques for Surgical Intervention in Patients with Metastatic Brain Tumors 557

Engl J Med. 1990;322(8):494–500. https://doi.org/10.1056/ 22. Walter J, Kuhn SA, Waschke A, Kalff R, Ewald C. Operative
NEJM199002223220802. treatment of subcortical metastatic tumours in the central region.
6. Patchell RA, Tibbs PA, Regine WF, et al. Postoperative radio- J Neuro-Oncol. 2011;103(3):567–73. https://doi.org/10.1007/
therapy in the treatment of single metastases to the brain: A ran- s11060-­010-­0420-­5.
domized trial. J Am Med Assoc. 1998;280(17):1485–9. https://doi. 23. Kellogg RG, Munoz LF. Selective excision of cerebral metastases
org/10.1001/jama.280.17.1485. from the precentral gyrus. Surg Neurol Int. 2013;4:66. https://doi.
7. Mahajan A, Ahmed S, McAleer MF, et al. Post-operative stereo- org/10.4103/2152-­7806.112189.
tactic radiosurgery versus observation for completely resected 24. Iwadate Y, Namba H, Yamaura A. Significance of surgical resec-
brain metastases: a single-centre, randomised, controlled, phase 3 tion for the treatment of multiple brain metastases. Anticancer Res.
trial. Lancet Oncol. 2017;18(8):1040–8. https://doi.org/10.1016/ 2000;20(1B):573–7.
S1470-­2045(17)30414-­X. 25. Pollock BE, Brown PD, Foote RL, Stafford SL, Schomberg
8. Nahed BV, Alvarez-Breckenridge C, Brastianos PK, et al. Congress PJ. Properly selected patients with multiple brain metastases
of neurological surgeons systematic review and evidence-based may benefit from aggressive treatment of their intracranial dis-
guidelines on the role of surgery in the management of adults with ease. J Neuro-Oncol. 2003;61(1):73–80. https://doi.org/10.102
metastatic brain tumors. Neurosurgery. 2019;84(3):E152–5. https:// 3/a:1021262218151.
doi.org/10.1093/neuros/nyy542. 26. Ammirati M, Nahed BV, Andrews D, Chen CC, Olson JJ. Congress
9. Delattre JY, Krol G, Thaler HT, Posner JB. Distribution of of neurological surgeons systematic review and evidence-based
brain metastases. Arch Neurol. 1988;45(7):741–4. https://doi. guidelines on treatment options for adults with multiple meta-
org/10.1001/archneur.1988.00520310047016. static brain tumors. Neurosurgery. 2019;84(3):E180–2. https://doi.
10. Medikonda R, Jackson CM, Feghali J, Lim M. The effects of post- org/10.1093/neuros/nyy548.
operative neurological deficits on survival in patients with single 27. Marenco-Hillembrand L, Alvarado-Estrada K, Chaichana
brain metastasis. [Review]. Oper Neurosurg. 2020;19(6):628–34. KL. Contemporary surgical management of deep-seated metastatic
https://doi.org/10.1093/ons/opaa224. brain tumors using minimally invasive approaches. Front Oncol.
11. Balestrino R, Rudà R, Soffietti R. Brain metastasis from unknown 2018;8:558. https://doi.org/10.3389/fonc.2018.00558.
primary tumour: Moving from old retrospective studies to clinical 28. Bakhsheshian J, Strickland BA, Jackson C, et al. Multicenter
trials on targeted agents. Cancers. 2020;12(11):1–18. https://doi. investigation of channel-based subcortical trans-sulcal exoscopic
org/10.3390/cancers12113350. resection of metastatic brain tumors: a retrospective case series.
12. Schroeder T, Bittrich P, Kuhne JF, et al. Mapping distribution of Oper Neurosurg (Hagerstown). 2019;16(2):159–66. https://doi.
brain metastases: does the primary tumor matter? J Neuro-Oncol. org/10.1093/ons/opy079.
2020;147(1):229–35. https://doi.org/10.1007/s11060-­020-­03419-­6. 29. Gassie K, Wijesekera O, Chaichana KL. Minimally invasive tubu-
13. Füreder LM, Widhalm G, Gatterbauer B, et al. Brain metasta- lar retractor-assisted biopsy and resection of subcortical intra-axial
ses as first manifestation of advanced cancer: exploratory analy- gliomas and other neoplasms. J Neurosurg Sci. 2018;62(6):682–9.
sis of 459 patients at a tertiary care center. Clin Exp Metastasis. https://doi.org/10.23736/S0390-­5616.18.04466-­1.
2018;35(8):727–38. https://doi.org/10.1007/s10585-­018-­9947-­1. 30. Alattar AA, Bartek J, Chiang VL, et al. Stereotactic laser abla-
14. Schackert G, Steinmetz A, Meier U, Sobottka SB. Surgical man- tion as treatment of brain metastases recurring after stereotactic
agement of single and multiple brain metastases: Results of a radiosurgery: a systematic literature review. World Neurosurg.
retrospective study. Onkologie. 2001;24(3):246–55. https://doi. 2019;128:134–42. https://doi.org/10.1016/j.wneu.2019.04.200.
org/10.1159/000055087. 31. Ahluwalia M, Barnett GH, Deng D, et al. Laser ablation after ste-
15. Stark AM, Tscheslog H, Buhl R, Held-Feindt J, Mehdorn reotactic radiosurgery: a multicenter prospective study in patients
HM. Surgical treatment for brain metastases: prognostic factors and with metastatic brain tumors and radiation necrosis. J Neurosurg.
survival in 177 patients. Neurosurg Rev. 2005;28(2):115–9. https:// 2019;130(3):804–11. https://doi.org/10.3171/2017.11.JNS171273.
doi.org/10.1007/s10143-­004-­0364-­3. 32. Holste KG, Orringer DA. Laser interstitial thermal therapy. Neuro-­
16. Schackert G, Lindner C, Petschke S, Leimert M, Kirsch Oncology Adv. 2020;2(1):vdz035. https://doi.org/10.1093/noajnl/
M. Retrospective study of 127 surgically treated patients with mul- vdz035.
tiple brain metastases: Indication, prognostic factors, and outcome. 33. Ahluwalia M, Barnett GH, Deng D, et al. Laser ablation after ste-
Acta Neurochir. 2013;155(3):379–87. https://doi.org/10.1007/ reotactic radiosurgery: a multicenter prospective study in patients
s00701-­012-­1606-­8. with metastatic brain tumors and radiation necrosis. J Neurosurg.
17. Bindal RK, Sawaya R, Leavens ME, Lee JJ. Surgical treatment of 2018;130(3):804–11. https://doi.org/10.3171/2017.11.JNS171273.
multiple brain metastases. J Neurosurg. 1993;79(2):210–6. https:// 34. Salehi A, Kamath AA, Leuthardt EC, Kim AH. Management of
doi.org/10.3171/jns.1993.79.2.0210. intracranial metastatic disease with laser interstitial thermal therapy.
18. Ogiwara H, Kalakota K, Rakhra SS, et al. Intracranial relapse Front Oncol. 2018;8:499. https://doi.org/10.3389/fonc.2018.00499.
rates and patterns, and survival trends following post-resection 35. Ali MA, Carroll KT, Rennert RC, et al. Stereotactic laser abla-
cavity radiosurgery for patients with single intracranial metasta- tion as treatment for brain metastases that recur after stereotactic
ses. J Neuro-Oncol. 2012;108(1):141–6. https://doi.org/10.1007/ radiosurgery: a multiinstitutional experience. Neurosurg Focus.
s11060-­012-­0808-­5. 2016;41(4):E11. https://doi.org/10.3171/2016.7.FOCUS16227.
19. Patel AJ, Suki D, Hatiboglu MA, et al. Factors influencing the risk of 36. Kubben PL, ter Meulen KJ, Schijns OEMG, ter Laak-Poort MP,
local recurrence after resection of a single brain metastasis. J Neurosurg. van Overbeeke JJ, van Santbrink H. Intraoperative MRI-guided
2010;113(2):181–9. https://doi.org/10.3171/2009.11.JNS09659. resection of glioblastoma multiforme: a systematic review.
20. Schackert G, Schmiedel K, Lindner C, Leimert M, Kirsch Lancet Oncol. 2011;12(11):1062–70. https://doi.org/10.1016/
M. Surgery of recurrent brain metastases: retrospective analysis S1470-­2045(11)70130-­9.
of 67 patients. Acta Neurochir. 2013;155(10):1823–32. https://doi. 37. Senft C, Franz K, Blasel S, et al. Influence of iMRI-guidance on
org/10.1007/s00701-­013-­1821-­y. the extent of resection and survival of patients with glioblastoma
21. Weil RJ, Lonser RR. Selective excision of metastatic brain multiforme. Technol Cancer Res Treat. 2010;9(4):339–46.
tumors originating in the motor cortex with preservation of func- 38. Akbari SHA, Sylvester PT, Kulwin C, et al. Initial experience
tion. J Clin Oncol. 2005;23(6):1209–17. https://doi.org/10.1200/ using intraoperative magnetic resonance imaging during a trans-­
JCO.2005.04.124. sulcal tubular retractor approach for the resection of deep-seated
558 J. A. Pawloski et al.

brain tumors: a case series. Oper Neurosurg (Hagerstown). 44. Le Rhun E, Taillibert S, Chamberlain MC. Carcinomatous
2019;16(3):292–301. https://doi.org/10.1093/ons/opy108. meningitis: Leptomeningeal metastases in solid tumors.
39. Valdés PA, Jacobs V, Harris BT, et al. Quantitative fluorescence Surg Neurol Int. 2013;4(Suppl 4):S265–88. https://doi.
using 5-aminolevulinic acid–induced protoporphyrin IX biomarker org/10.4103/2152-­7806.111304.
as a surgical adjunct in low-grade glioma surgery. J Neurosurg. 45. Lin N, Dunn IF, Glantz M, et al. Benefit of ventriculoperitoneal cere-
2015;123(3):771–80. https://doi.org/10.3171/2014.12.JNS14391. brospinal fluid shunting and intrathecal chemotherapy in neoplas-
40. Patel AJ, Suki D, Hatiboglu MA, Rao VY, Fox BD, Sawaya tic meningitis: a retrospective, case-controlled study. J Neurosurg.
R. Impact of surgical methodology on the complication rate and 2011;115(4):730–6. https://doi.org/10.3171/2011.5.JNS101768.
functional outcome of patients with a single brain metastasis. J 46. Raleigh DR, Seymour ZA, Tomlin B, et al. Resection and brain
Neurosurg. 2015;122(5):1132–43. https://doi.org/10.3171/2014.9. brachytherapy with permanent iodine-125 sources for brain metas-
JNS13939. tasis. J Neurosurg. 2017;126(6):1749–55. https://doi.org/10.3171/2
41. Suki D, Hatiboglu MA, Patel AJ, et al. Comparative risk of lep- 016.4.JNS152530.
tomeningeal dissemination of cancer after surgery or stereotactic 47. Chitti B, Goyal S, Sherman JH, et al. The role of brachytherapy
radiosurgery for a single supratentorial solid tumor metastasis. in the management of brain metastases: a systematic review.
Neurosurgery. 2009;64(4):664–74; discussion 674–676. https://doi. J Contemp Brachytherapy. 2020;12(1):67–83. https://doi.
org/10.1227/01.NEU.0000341535.53720.3E. org/10.5114/jcb.2020.93543.
42. Hyun J-W, Jeong IH, Joung A, Cho HJ, Kim S-H, Kim 48. Xia Y, Mashouf LA, Baker BR, et al. Outcomes of metastatic brain
HJ. Leptomeningeal metastasis: clinical experience of 519 lesions treated with radioactive Cs-131 seeds after surgery: expe-
cases. Eur J Cancer. 2016;56:107–14. https://doi.org/10.1016/j. rience from one institution. Cureus. 2018;10(7):e3075. https://doi.
ejca.2015.12.021. org/10.7759/cureus.3075.
43. Lee SH, Kong DS, Seol HJ, Nam D-H, Lee J-I. Ventriculoperitoneal 49. Chidambaram S, Pannullo SC, Schwartz TH, Wernicke
shunt for hydrocephalus caused by central nervous system metas- AG. Reirradiation of recurrent brain metastases: where do we stand?
tasis. J Neuro-Oncol. 2011;104(2):545–51. https://doi.org/10.1007/ World Neurosurg. 2019;125:156–63. https://doi.org/10.1016/j.
s11060-­010-­0512-­2. wneu.2019.01.182.
 Brain Metastases: Current and Future
Pharmacological Treatment 52
Erika Santos Horta and Tobias Walbert

Abstract common intracranial malignancy [3, 4]. Brain metastases

Solid tumor metastasis to the brain is a frequent event that increase patient morbidity and mortality [4–6]. The solid
increases morbidity and mortality of patients suffering tumors that most commonly metastasize to the brain are
from an already life-threatening disease. Pharmacological melanoma, lung cancer, breast cancer, renal cancer, and
treatment efficacy is challenged by the difficulties of colorectal cancer [2, 7, 8].
delivering drugs to the central nervous system. Traditional Molecular pathology findings are revealing that some
chemotherapy has limited success. Fortunately, immuno- mutations can stratify a patient’s risk for developing brain
therapy and target-based therapies are broadening treat- metastases and offer a molecular target for specific therapies.
ment options with promising results in clinical trials and For example, mutations of the epidermal growth factor
in real-world settings. receptor (EGFR) or anaplastic lymphoma kinase (ALK) in
patients with non-small-cell lung cancer (NSCLC), KRAS/
RAS mutation in colon cancer, and human epidermal growth
factor receptor 2 (HER2) in breast cancer are risk factors for
Learning Objectives brain metastasis [2, 8–10].
• To understand the challenges of pharmacotherapy There are many challenges to the treatment of brain
when treating brain metastases. metastases, including the blood-brain and the blood-CSF
• To be able to evaluate the best pharmacotherapy barriers, different tumor clones, and tumor microenviron-
options for different types of brain metastases. ment, among others [6]. The blood–brain barrier (BBB) is a
• Be aware of the side effects of pharmacotherapy for functional structure formed by endothelial cells connected
brain metastases. by tight junctions with astrocytes and pericytes. The BBB
regulates the transport of molecules from the blood to the
brain parenchyma through the tight junctions and efflux
pumps in the endothelial cells. Doing so protects the brain
52.1 Introduction from toxins, as well as foreign substances, maintaining a
stable microenvironment [3, 6, 11]. The blood-CSF barrier is
Due to seeding of systemic cancers, 20–40% of patients will similar to the BBB but is formed by choroid plexus epithelial
eventually present with brain metastases, and the incidence cells and capillaries. Those barriers can alter the delivery,
is increasing due to better systemic control of the disease distribution, and accumulation of chemotherapies, targeted
and expansion of brain imaging [1, 2]. As a consequence, therapies, and antibodies [6]. Other challenges are tumor
metastases of solid tumors to the brain represent the most cell-astrocyte gap cell junctions that directly connect the
cytoplasm of two cells as well as alterations in microglia that
can alter the tumor microenvironment in the central nervous
E. S. Horta (*)
system (CNS). Another obstacle to treat brain metastases is
University of Arkansas for Medical Sciences,
Little Rock, AR, USA that they can be formed by different clones than the primary
e-mail: [email protected] tumor and therefore respond differently to treatment than the
T. Walbert primary tumor [3, 6]. For all of these reasons, the treatment
Hermelin Brain Tumor Center, Henry Ford Health System, of brain metastasis has focused on surgery and radiation
Detroit, MI, USA therapy, while pharmacotherapy is mostly used in the
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 559
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_52
560 E. S. Horta and T. Walbert

a­ djuvant setting [6]. On the other hand, target-based thera- ease, when radiation therapy and surgery are not an option,
pies have shown promising results when brain metastases asymptomatic metastases, and SCLC [12]. Traditional che-
demonstrate selective mutations [6]. motherapy protocols include platinum compounds (alkylat-
Traditional chemotherapy aims to achieve a cure or ing agents) in monotherapy or in combination with
decrease the size of a tumor by interfering in the cell cycle and pemetrexed, vinorelbine, or etoposide as first-line therapy [4,
therefore killing predominantly cancer cells. A common 12].
mechanism of action of those substances is to add errors to the
DNA of cells. Currently, cancer treatment has expanded with
the use of immunotherapy and target-based therapy. 52.2.2 Target-Based Therapy
Immunotherapy sensitizes the patient’s own immune system
to recognize and fight cancer through the presentation of can- Tyrosine kinase inhibitors (TKIs) targeting ALK inhibitor
cer antigens. Although virus therapy can also be used to and EGFR are now considered to be first-line therapy options
enhance the immune system to fight cancers, immunotherapy for patients with lung cancer bearing those mutations [9].
is mainly represented by antibodies. Target-based therapy tries ALK inhibitors have been FDA approved for NSCLC
to dismantle some of the intracellular pathways that cancerous therapy since 2011 and have been shown to be efficacious in
cells use by aiming at a particular mutation. Tyrosine kinase patients with brain metastases. Among ALK inhibitors,
inhibitors (TKIs) are a good example of target-­based therapies alectinib, brigatinib, ceritinib, and lorlatinib seem to be
as they block enzymes that work in intracellular pathways. more efficacious than crizotinib for CNS disease, due to bet-
Another advantage of TKIs and other targeted therapies is that ter ability to cross the BBB. In clinical trials, patients with
the molecules are smaller than the traditional chemotherapy NSCLC receiving treatment with ALK inhibitors for sys-
substances and therefore can cross the BBB [8, 11]. temic disease presented with CNS metastases later than
In this chapter, we will discuss pharmacotherapy (tradi- controls [4, 11, 13].
tional chemotherapy, target-based chemotherapies, and TKIs that act to reversibly inhibit EGFR, like erlotinib
immunotherapy represented by antibodies) for the most and osimertinib, have shown better response rates than tradi-
common brain metastases. tional chemotherapy and better overall survival (OS) in
patients with EGFR-mutant NSCLC brain metastases [1, 4].
In the phase 3 clinical trial AURA3, the osimertinib group
52.2 Lung Cancer showed a longer CNS progression-free survival (11.7
months) than the platinum-pemetrexed group (5.6 months)
Lung cancer is divided into small-cell lung carcinoma [14]. The percentage of patients alive in the 6 and 12 months
(SCLC) and non-small-cell lung cancer (NSCLC) which is (75% and 43%, respectively) in the osimertinib group was
further subdivided into adenocarcinoma, squamous cell car- also improved when compared with the chemotherapy group
cinoma, and large-cell carcinoma. Although SCLC (45% and 17%, respectively). The group with osimertinib
­metastases have more tropism to the brain than NSCLC, reached CNS response rate of 70% versus 31% in the chemo-
NSCLC metastases are more common due to the higher therapy group. Osimertinib is now one of the first-line treat-
prevalence of NSCLC [7]. Among the different subtypes of ments for EGFR-positive metastatic NSCLC, and it showed
NSCLC, adenocarcinoma and large-cell carcinoma have activity in patients with leptomeningeal disease [14–16].
more tropism to the brain. SCLC metastasizes to the brain Clinical trials have evaluated if adding radiation therapy
earlier than NSCLC and is more sensitive to standard chemo- to TKI therapy in patients with brain metastases increase OS
therapeutic agents. In most institutions, patients with SCLC or progression-free survival (PFS) in patients with brain
are treated with prophylactic whole-brain radiation therapy, metastases [2]. A recent meta-analysis concluded that there
but the increased sensitivity of SCLC can be further targeted is no significant increase of OS or PFS, but added toxicty [9].
when combining chemotherapy with radiotherapy [6, 12].
NSCLC is frequently associated with mutations in EGFR
and anaplastic lymphoma kinase (ALK) which can be tar- 52.2.3 Immunotherapy
geted with systemic therapy and targeted agents [9]. Those
mutations are also associated with better survival when com- Checkpoint inhibitors (anti-PD1, anti-PDL1, and anti-CTL4)
pared with wild-type tumors [9]. were studied in patients with NSCLC and brain metastasis
and showed sustained activity in patients with PDL1 expres-
sion over 1%. Intracranial response of 33–36% in
52.2.1 Chemotherapy chemotherapy-­naive patients and in pre-treated patients is
seen in NSCLC and in SCLC [1, 17, 18].
Response rates vary from 30–80% in SCLC to 10–30% in The potential benefits of using the anti-VEGF antibody
NSCLC. Chemotherapy is a good choice for multifocal dis- bevacizumab remain unclear. A recent meta-analysis in
52 Brain Metastases: Current and Future Pharmacological Treatment 561

patients with NSCLC concluded that bevacizumab might than monotherapy with trastuzumab [2, 20, 22]. Therefore,
protect patients from developing new brain metastases and it for HER2-­positive patients with brain metastases, HER2/
might control existing metastases when compared with con- EGFR TKIs are currently considered an effective therapeu-
ventional chemotherapy [19]. The medium progression-free tic option [1, 8].
survival of patients with brain metastases being treated with TKI have been associated with increased risks for radia-
bevacizumab was 7.1 months, and medium overall survival tion necrosis when added to radiation therapy [23]. Tucatinib
was 13.5 months. A pooled analysis showed better OS and might be associated with less adverse events than the other
PFS hazard ratio in patients with bevacizumab when com- dual TKIs [8].
pared with other traditional chemotherapies [19]. PARP inhibitors have been tested in phase 1 trials and
have been well tolerated, but they need to be further investi-
gated in patients with brain metastases. Those medications
52.3 Breast Cancer regulate BRCA1-promoter methylation which is thought to
play an important role in breast cancer pathogenesis in the
Breast cancer is the second most common cause of brain setting of BRCA1/2 mutation carriers [22, 24].
metastases [8]. It is often classified by the type of hormone CDK4/6 inhibitors (palbociclib, ribociclib, abemaciclib)
receptors (estrogen [ER], progesterone [PR], human epider- are FDA-approved medications for metastatic breast cancers
mal growth factor 2 [HER2]) because it guides pharmacother- that are HER2-negative and estrogen/progesterone-positive.
apy choices and predicts the specific tendency to metastasize Current clinical trials show that these medications are also
to the brain. For example, triple-negative breast cancers are efficient in patients with brain metastases [20, 25].
usually treated with traditional chemotherapy, as radiosensi-
tivity is low and there is no role for hormone therapy.
52.3.3 Immunotherapy

52.3.1 Chemotherapy Trastuzumab and pertuzumab are antibodies against the

HER2 cell receptor and are FDA approved for HER2-
The great challenge for breast cancer treatment is that the positive breast cancer therapy since 1998 and 2012, respec-
majority of chemotherapy agents used for systemic therapy tively. Trastuzumab in addition to chemotherapy increases
do not cross the BBB [2]. This is considered one of the rea- OS and PFS and, when conjugated with a cytotoxic micro-
sons the incidence of breast cancer metastasis to the brain is tubule molecule, has shown to decrease the size of brain
increasing, while therapy for systemic disease, especially in metastases [8, 20]. The CLEOPATRA clinical trial showed
hormone receptor-positive patients, has resulted in improved that pertuzumab added to trastuzumab and docetaxel delays
outcomes [20]. Options for chemotherapy include the time to onset of brain metastasis and increases the OS in
capecitabine, cyclophosphamide, vincristine, temozolomide, patients with brain metastasis from breast cancer [26].
and many others, with response rates varying from 30 to Based on these findings, the combination of pertuzumab and
60%, when chemotherapy was added to radiation therapy [2, trastuzumab should be offered to HER2 breast cancer
12, 20]. Unfortunately, median OS for breast cancer patients patients [12].
is less than one year in phase 2 and 3 trials [2, 20].
For ER-positive tumors, tamoxifen and aromatase inhibi-
tors should be offered for brain metastasis patients as first-­ 52.4 Melanoma
line therapy, as these agents reach high concentrations in the
brain parenchyma and increase patient survival [20, 21]. The Melanomas exhibit high levels of tropism to the brain. This
increased survival was also seen in some patients with lepto- is reflected by the fact that one third of all melanoma patients
meningeal disease [21]. present with brain metastases [2]. Before the advent of
monoclonal antibodies and target-based therapies, the out-
come for melanoma patients with brain metastases was dis-
52.3.2 Target-Based Therapy mal with an OS of only weeks [2]. BRAF and checkpoint
inhibitors are changing this trajectory with median overall
Dual HER2/EGFR TKIs block tyrosine phosphorylation survival of one year in phase 2 clinical trials [2, 6, 27].
and intracellular pathways with good brain parenchyma
penetrance. Lapatinib, a HER2/EGFR TKIs, is FDA
approved for treatment of breast cancer since 2007. Those 52.4.1 Chemotherapy
TKIs have been used in combination with capecitabine and
temozolomide with a response rate of 20–60%, decrease in Cytotoxic chemotherapy has failed to show efficacy and
6–10% in the volume of intracranial lesions, and better OS therefore has no significant role in the treatment of patients
562 E. S. Horta and T. Walbert

with melanoma brain metastases, even when combined with Research Foundation suggests that trials with checkpoint
radiation therapy [2, 12]. inhibitors are reasonable [28].

52.4.2 Target-Based Therapy 52.5 Renal Cancer

Melanoma frequently presents with BRAF V600 mutations, Renal cell carcinoma has a 5-year cumulative incidence of
and in patients with brain metastases, clinical trials using brain metastases of 10% with a mean overall survival of
BRAF inhibitors in monotherapy or associated with a MEK 4–11 months after resection, independently if WBRT or tra-
inhibitor showed intracranial response near 40–60% in ditional chemotherapy is used [2]. Currently, it is considered
treatment-­naïve patients and over 30% in previously treated a good practice to offer TKIs to patients with renal cell car-
patients [1, 6, 27]. Unfortunately, resistance to treatment has cinoma with brain metastases [12].
been seen after one year [6, 27]. Combination therapies of
BRAF plus MEK inhibitors have shown high response rates,
decrease in side effects, and without increase in OS [6, 27]. 52.5.1 Chemotherapy
After published case series and case reports summing up
over 20 cases of patients with leptomeningeal disease with Renal cell carcinoma is a chemotherapy- and radiation-­
response to TKIs, the 2019 Consensus of Melanoma resistant tumor as neither whole-brain radiation nor systemic
Research Foundation suggests that BRAF monotherapy or traditional chemotherapies have extended the overall sur-
the combination of BRAF and MEK inhibitors can be con- vival or progression-free survival [2]. Systemic chemother-
sidered in the treatment of patients with leptomeningeal dis- apy can be used in an attempt to control extra-CNS disease.
ease, but data is still too sparse to make stronger
recommendations [28].
52.5.2 Target-Based Therapy

52.4.3 Immunotherapy VEGFR-TKIs, like sunitinib and pazopanib, have brain pene-
trance and have shown the ability to stabilize brain metastases
Anti-CTL4 immunotherapies were FDA approved for meta- and increase overall survival in clinical studies, achieving a
static melanoma in 2011. In 2012, an open-label, phase 2 response rate of 4–13% and stable disease in 33–60% of
trial showed that ipilimumab could be an option for patients patients [33–35]. Efficacy seems to increase when combined
with brain metastasis [29]. Although ipilimumab cannot with stereotactic radiosurgery (SRS), at the cost of radiation
cross the BBB due to its particle size, it is thought to stimu- necrosis [12, 36]. Retrospective studies and expanded-access
late the immune system against melanoma, and the stimu- trials indicated that patients with renal cancer developed brain
lated CD8 T cells can then infiltrate and attack metastatic metastases later and in fewer numbers when using TKIs [33–
cancer cells in the CNS [30]. 37]. While we are waiting for the final results of clinical trials
Checkpoint inhibitors have been successful in monother- that are evaluating VEGFR-TKIs in patients with metastatic
apy for brain metastases and in combination with traditional renal cancer, it is considered a good practice to offer VEGFR-
chemotherapy or with other immunotherapy [28, 30, 31]. TKI to patients with renal cancer and brain metastases [35, 38].
Intracranial response rates have been reported to reach While mTOR TKIs like everolimus are approved for
44–56% [1, 32]. The response to therapies that incorporate advanced renal cell carcinoma and are considered safe in
these inhibitors seems to be more durable than target-based patients with brain metastases, they have not yet shown to be
therapies alone [28, 30, 31]. However, checkpoint inhibitors efficacious [35].
are less effective when steroids are used. The combination of
checkpoint inhibitors with radiation therapy, especially ste-
reotactic radiosurgery, is thought to increase efficacy but also 52.5.3 Immunotherapy
the risk of radiation necrosis [28, 30, 31]. The increase of
survival with checkpoint inhibitor and SRS might be higher Immunotherapy for brain metastases in renal cell carci-
for anti-PDL1 than for anti-CTL4 [29–31]. noma has had disappointing results. Although case reports
There are three ongoing clinical trials with checkpoint describe successful use of interleukin-2, this treatment was
inhibitors that are now allowing patients with leptomenin- evaluated in a cohort of patients with metastatic renal can-
geal disease to participate in, but there are case reports cer or melanoma [39]. Objective response was low, 5.6 %,
describing response and even complete response of lepto- with grade 3 or higher neurological toxicity seen in 7–16 %
meningeal disease. The 2019 Consensus of Melanoma of patients.
52 Brain Metastases: Current and Future Pharmacological Treatment 563

Although nivolumab is indicated for renal cell carcinoma, Corticosteroids, such as dexamethasone, are the first-
checkpoint inhibitors in monotherapy for brain metastases line treatment for increased intracranial pressure or symp-
have not shown any efficacy [1, 22]. tomatic peritumoral edema, and treatment and dosing
should be guided by clinical presentation and neurologi-
cal deficits rather than by MRI findings [12, 38]. A fast
52.6 Colon Cancer down titration is recommended after patients improve to
control side effects like diabetes, osteoporosis, insomnia,
Brain metastases are considered a late event of colon cancer increased blood pressure, osteonecrosis, gastrointestinal
and typically present when the patient has metastases to other hemorrhage, and immune-suppression [12, 38]. As alter-
organs like the liver and lungs resulting in only 30% of patients natives, Boswellia and osmolar agents can be considered,
with brain metastasis surviving more than one year [40, 41]. and a randomized trial using a short course of bevaci-
Risk factors for brain metastases are age <60 years, RAS zumab has been shown to decrease radiation necrosis-
mutation, and adenocarcinoma [40, 41]. Due to the lack of induced edema [12, 38].
clinical trials, there is little knowledge on the best pharmaceu- While brain metastases can cause seizures, anti-epilep-
tical treatment of brain metastasis in colorectal patients; there- tic medications are not recommended prophylactically
fore, protocols are based on surgery and radiation oncology. [12, 38]. Non-enzyme-inducer medications like leveti-
racetam and lamotrigine are preferred over phenytoin,
phenobarbital, and carbamazepine due to a lack of inter-
52.6.1 Chemotherapy actions with other medications and better side effect pro-
file [38].
Patients likely had already received chemotherapy before
having brain metastases; therefore, options are limited, and
protocols are usually a combination of two or three medica- 52.8 Treatment-Related Toxicity
tions [40]. Options include capecitabine, oxaliplatin, fluoro-
uracil, and irinotecan. When treating a patient, not only efficacy must be analyzed
but also the potential for treatment-related toxicity.
Traditional chemotherapies have many side effects, which
52.6.2 Target-Based Therapy include chemotherapy-induced nausea, hair loss, cognitive
dysfunction, cardiomyopathy, neuropathy, and many others
BRAF mutations are found in 5–15% of patients with [2, 5].
colorectal cancer. TKIs are being used more and more fre- Immunotherapy is also known to cause adverse events
quently in patients with colon cancer, but their safety and such as colitis, pneumonitis, and other endocrinological
efficacy are still being studied in the context of brain metas- and neurological symptoms [1, 43–45]. Some of those
tases [40]. adverse events differ depending on the type of immuno-
therapy. For example, myasthenia gravis and hypophysitis
are seen with anti-CTL4, anti-PD1, and anti-PDL1 [1, 4,
52.6.3 Immunotherapy 43].Five other adverse events like encephalopathy, head-
ache, and neuropathy are observed in many classes of
A retrospective study of patients with brain metastases immunotherapy [1, 43, 45]. Vitiligo secondary to immu-
showed that bevacizumab, an antiangiogenic monoclonal notherapy seems to be more prevalent in patients with
antibody against VEGF, improved survival from 59 to 151 melanoma. When combined with radiation therapy, immu-
days [41]. In another retrospective study from a single insti- notherapy can increase the likelihood of radiation necro-
tution, the time to develop brain metastases in patients with sis, encephalopathy, and seizures [9, 12, 24, 28, 43].
colorectal cancer treated with bevacizumab was longer than Target-based therapies are known to cause photosensibil-
in patients that did not received them [42]. ity, paronychia, diarrhea, cardiotoxicity, and neurological
adverse events [45].
The treatment of immunotherapy-induced toxicity
52.7 Symptomatic Treatment includes temporary or permanent discontinuation of therapy
and, when needed, associated corticosteroids in the form of
Symptomatic treatment of patients with brain metastases prednisone or dexamethasone [44, 45]. After the patient
focuses on the control of peritumoral edema, increased intra- improves, resuming therapy is possible and should be con-
cranial pressure, and the treatment of seizures. sidered on a case-by-case basis [43–45].
564 E. S. Horta and T. Walbert

52.9 Conclusions and Future Directions 6. Lowery FJ, Yu D. Brain metastasis: unique challenges and open
opportunities. Biochim Biophys Acta Rev Cancer. 2017; https://doi.
org/10.1016/j.bbcan.2016.12.001.
In the last ten years, basic research and clinical trials have 7. Cagney DN, et al. Incidence and prognosis of patients with brain
greatly increased the understanding of brain cancer metasta- metastases at diagnosis of systemic malignancy: a population-based
ses. Pharmacotherapy for the treatment of brain metastases study. Neuro-Oncology. 2017; https://doi.org/10.1093/neuonc/
nox077.
has increased and is now offering new options that have 8. Duchnowska R, Loibl S, Jassem J. Tyrosine kinase inhibitors for
shown meaningful increases of progression-free and overall brain metastases in HER2-positive breast cancer. Cancer Treat Rev.
survival. Traditional chemotherapy has been replaced by 2018;67:71–7.
immunotherapy and targeted therapies that are better toler- 9. Singh R, et al. Brain metastases from non-small cell lung cancer with
EGFR or ALK mutations: a systematic review and meta-analysis of
ated and have resulted in promising results. However, chal- multidisciplinary approaches. Radiother Oncol. 2020;144:165–79.
lenges specific to brain metastases such as limited BBB 10. Christensen TD, Spindler KLG, Palshof JA, Nielsen DL. Systematic
permeability remain, and newer drug delivery methods like review: brain metastases from colorectal cancer-incidence and
liposomes, nanoparticles, and pegylation or using mannitol patient characteristics. BMC Cancer. 2016; https://doi.org/10.1186/
s12885-­016-­2290-­5.
to increase therapeutic response need to be explored. 11. Dodson C, Richards TJ, Smith DA, Ramaiya NH. Tyrosine kinase
Ultrasound for blood BBB disruption and implantable bio- inhibitor therapy for brain metastases in non-small-cell lung can-
material devices are also being studied to increase delivery cer: a primer for radiologists. Am J Neuroradiol. 2020; https://doi.
of drugs to the brain parenchyma. Other lines of research are org/10.3174/AJNR.A6477.
12. Soffietti R, Abacioglu U, Baumert B. Diagnosis and treatment of
focused on the actual prevention of these deadly brain metastases from solid tumors: guidelines from the European
malignancies. association of neuro-oncology (EANO). Neuro-Oncology. 2017;
https://doi.org/10.1093/neuonc/now241.
13. Peters S, et al. Alectinib versus Crizotinib in untreated ALK -posi-
tive non–small-cell lung cancer. N Engl J Med. 2017; https://doi.
Questions to be answered after reading this chapter are: org/10.1056/nejmoa1704795.
• How can we overcome some of the challenges of 14. Wu YL, et al. CNS efficacy of osimertinib in patients with T-790M-­
treating brain metastases with systemic positive advanced nonsmall-cell lung cancer: data from a random-
ized phase III trial (AURA 3). J Clin Oncol. 2018;10:2702–9.
pharmacotherapies? 15. Reungwetwattana T, et al. CNS response to osimertinib versus
• How do pharmacotherapy options vary per type of standard epidermal growth factor receptor tyrosine kinase inhibi-
tumor? tors in patients with untreated EGFR-mutated advanced non-­
• Can we integrate pharmacotherapy with other types small-­cell lung cancer. J Clin Oncol. 2018; https://doi.org/10.1200/
JCO.2018.78.3118.
of treatment? 16. Planchard D, et al. Osimertinib vs standard of care (SoC) EGFR-­
• My patient has brain metastases. Should I start an TKI as first-line therapy in patients (pts) with untreated EGFRm
anti-seizure medication? advanced NSCLC: FLAURA post-progression outcomes. Ann
• What are the potential side effects of pharmacother- Oncol. 2018; https://doi.org/10.1093/annonc/mdy425.
17. Wang S, Hu C, Xie F, Liu Y. Use of programmed death receptor-1
apy for brain metastases? and/or programmed death ligand 1 inhibitors for the treatment of
brain metastasis of lung cancer. Onco Targets Ther. 2020;13:667–83.
18. Goldman JW, et al. Nivolumab (nivo) in patients (pts) with
advanced (adv) NSCLC and central nervous system (CNS)
metastases (mets). J Clin Oncol. 2016; https://doi.org/10.1200/
jco.2016.34.15_suppl.9038.
References 19. Liang P, et al. Bevacizumab for non-small cell lung cancer patients
with brain metastasis: a meta-analysis. Open Med. 2020;15:589–97.
1. Lauko A, Thapa B, Venur V, Ahluwalia M. Management of brain 20. Shah N, et al. Investigational chemotherapy and novel pharma-
metastases in the new era of checkpoint inhibition. Curr Neurol cokinetic mechanisms for the treatment of breast cancer brain
Neurosci. 2018;18:70. metastases. Pharmacol Res. 2018; https://doi.org/10.1016/j.
2. Lombardi G, Di Stefano AL, Farina P, Zagonel V, Tabouret phrs.2018.03.021.
E. Systemic treatments for brain metastases from breast cancer, 21. Bergen ES, et al. Continued endocrine therapy is associated with
non-small cell lung cancer, melanoma and renal cell carcinoma: improved survival in patients with breast cancer brain metasta-
an overview of the literature. Cancer Treat Rev. 2014; https://doi. ses. Clin Cancer Res. 2019; https://doi.org/10.1158/1078-­0432.
org/10.1016/j.ctrv.2014.05.007. CCR-­18-­1968.
3. Achrol AS, et al. Brain metastases. Nat Rev Dis Primers. 2019; 22. Costa R, et al. Developmental therapeutics for patients with breast
https://doi.org/10.1038/s41572-­018-­0055-­y. cancer and central nervous system metastasis: current landscape
4. Aizer AA, Lee EQ. Brain metastases. Neurol Clin. 2018; https:// and future perspectives. Ann Oncol. 2017; https://doi.org/10.1093/
doi.org/10.1016/j.ncl.2018.04.010. annonc/mdw532.
5. Walbert T, Horta E. Health-related quality of life related to toxic- 23. Lin NU, et al. Phase I study of lapatinib (L) in combination with
ity treatments in central nervous system metastases. In: Ahluwalia whole-brain radiation therapy (WBRT) in patients (pts) with brain
M, Metellus PSR, editors. Central nervous system metastasis. metastases from HER2-positive breast cancer. J Clin Oncol. 2010;
SpringerLink; 2020. p. 373–82. https://doi.org/10.1200/jco.2010.28.15_suppl.1154.
52 Brain Metastases: Current and Future Pharmacological Treatment 565

24. Mehta MP, et al. Veliparib in combination with whole brain radiation sion with other targeted therapies. Eur J Cancer. 2013; https://doi.
therapy in patients with brain metastases: results of a phase 1 study. org/10.1016/j.ejca.2013.06.003.
J Neuro-Oncol. 2015; https://doi.org/10.1007/s11060-­015-­1733-­1. 35. Matsui Y. Current multimodality treatments against brain metastases
25. Abraham J, et al. Use of cyclin-dependent kinase (CDK) 4/6 inhibi- from renal cell carcinoma. Cancer. 2020; https://doi.org/10.3390/
tors for hormone receptor-positive, human epidermal growth factor cancers12102875.
receptor 2-negative, metastatic breast cancer: a roundtable discus- 36. Juloori A, et al. Overall survival and response to radiation and
sion by The Breast Cancer Therapy Expert Group (BCTEG). Breast targeted therapies among patients with renal cell carcinoma brain
Cancer Res Treat. 2018;171:11–20. metastases. J Neurosurg. 2020; https://doi.org/10.3171/2018.8.
26. Baselga J, et al. Pertuzumab plus Trastuzumab plus Docetaxel JNS182100.
for metastatic breast cancer. N Engl J Med. 2012; https://doi. 37. Verma J, et al. Impact of tyrosine kinase inhibitors on the incidence
org/10.1056/nejmoa1113216. of brain metastasis in metastatic renal cell carcinoma. Cancer.
27. Davies MA, et al. Dabrafenib plus trametinib in patients with 2011;117:4958–65. https://doi.org/10.1002/cncr.26138.
BRAFV600-mutant melanoma brain metastases (COMBI-MB): a 38. Roth P, et al. Neurological and vascular complications of primary
multicentre, multicohort, open-label, phase 2 trial. Lancet Oncol. and secondary brain tumours: EANO-ESMO Clinical Practice
2017; https://doi.org/10.1016/S1470-­2045(17)30429-­1. Guidelines for prophylaxis, diagnosis, treatment and follow-up.
28. Glitza IC, et al. Leptomeningeal disese in melanoma patients: an Ann Oncol. 2021;32:171–82.
update to treatment, challenges, and future directions. Pigment 39. Guirguis L, et al. Safety and efficacy of high-dose interleukin-2
Cell Melanoma Res. 2020;33(4):527–41. https://doi.org/10.1111/ therapy in patients with brain metastases. J Immunother. 2002;25:
pcmr.12861. 82–7.
29. Margolin K, et al. Ipilimumab in patients with melanoma and 40. Mege D, Sans A, Ouaissi M, Iannelli A, Sielezneff I. Brain metasta-
brain metastases: an open-label, phase 2 trial. Lancet Oncol. ses from colorectal cancer: characteristics and management. ANZ J
2012;13:459–65. Surg. 2018; https://doi.org/10.1111/ans.14107.
30. Fonkem E, et al. Melanoma brain metastasis: overview of cur- 41. Chahine G, et al. Colorectal cancer and brain metastases: an aggres-
rent management and emerging targeted therapies. Expert Rev sive disease with a different response to treatment. Tumori. 2019;
Neurother. 2012; https://doi.org/10.1586/ern.12.111. https://doi.org/10.1177/0300891618765541.
31. van Opijnen M, Dirven L, Coremans I, Taphoorn M, Kapiteijn 42. Fountzilas C, et al. Clinical characteristics and treatment outcomes
E. The impact of current treatment modalities on the outcomes of of patients with colorectal cancer who develop brain metasta-
patients with melanoma brain metastases: a systematic review. Int J sis: a single institution experience. J Gastrointest Oncol. 2017;8:
Cancer. 2020;146:1479–89. 55–63.
32. Kluger HM, et al. Long-term survival of patients with melanoma 43. Horta ES, et al. Neural autoantibody clusters aid diagnosis of can-
with active brain metastases treated with pembrolizumab on a phase cer. Clin Cancer Res. 2014;20:3862–9.
II trial. J Clin Oncol. 2019; https://doi.org/10.1200/JCO.18.00204. 44. Choi J, Lee S. Clinical characteristics and treatment of immune-­
33. Gore ME, et al. Final results from the large sunitinib global related adverse events of immune checkpoint inhibitors. Immune
expanded-access trial in metastatic renal cell carcinoma. Br J Netw. 2020;20:e9.
Cancer. 2015; https://doi.org/10.1038/bjc.2015.196. 45. Baldo B, Pham N. Targeted drugs for cancer therapy: small mol-
34. Matrana MR, et al. Outcomes of patients with metastatic clear-cell ecules and monoclonal antibodies. Drug Allergy. 2020:595–644.
renal cell carcinoma treated with pazopanib after disease progres- https://doi.org/10.1007/978-­3-­030-­51740-­3_14.
 Radiotherapy for CNS Metastases
53
Alexander N. Slade, Mark Ashamalla, and Samuel Ryu

Abstract
• To understand potential management paradigms for
Brain metastases are common among cancer patients, patients with brain metastases which integrate radi-
estimated to impact nearly one in five patients throughout ation therapy.
the course of their disease. Radiation therapy, along with • To understand emerging paradigms in the use of
surgery, comprises the primary modalities used to manage radiation therapy for brain metastases, including
brain metastases. This chapter considers contemporary pre-operative stereotactic radiosurgery and hippo-
prognostic approaches to patients with brain metastases. campal avoidance whole brain radiation therapy.
We then discuss the evolution of radiation therapy in the
management of brain metastases from whole brain radia-
tion therapy to stereotactic radiosurgery, in which it is
now possible to deliver highly targeted radiation therapy
to brain metastases (or the surgery cavities thereof) while 53.1 Introduction
sparing normal structures. We also review recent advances
in limiting the neurocognitive toxicities in patients receiv- Brain metastasis represents the most common type of intra-
ing whole brain radiation, including hippocampal avoid- cranial malignancy and is estimated to affect approximately
ance strategies. Finally, we discuss a framework whereby one in five cancer patients [1]. While historically the pres-
various management strategies, ranging from conserva- ence of brain metastasis implied an extremely limited prog-
tive management to surgical resection and radiosurgery, nosis, the advent of an increasing number of systemic and
may be considered. targeted therapies, along with improved stratification for
patients diagnosed with metastatic disease to the brain, has
led to improvements in the survival of patients diagnosed
with brain metastases over time. However, there exist large
Learning Objectives
variations depending on cancer type [2–4].
• To understand commonly used prognostic classifi-
Whole brain radiation therapy (WBRT) has historically
cation systems in patients with brain metastases.
represented the standard of care for individuals with mul-
• To understand the role of CNS-directed radiation
tiple brain metastases. However, with technical progress of
therapy, including whole brain radiation therapy
focused radiation delivery and clinical evidence of cognitive
and stereotactic radiosurgery, both in the intact and
decline after WBRT, practice patterns have shifted to favor
postoperative setting, in reducing the risk of local
more focused radiation therapy such as stereotactic radio-
recurrence.
surgery (SRS), both in the intact and postoperative setting,
for patients with a limited number of metastases. In fact, the
American Society for Therapeutic Radiology and Oncol-
ogy (ASTRO) Choosing Wisely guidelines recommend the
avoidance of routine use of WBRT in patients receiving SRS
A. N. Slade · M. Ashamalla · S. Ryu (*)
for a limited number of metastases [5].
Department of Radiation Oncology, Stony Brook University,
Stony Brook, NY, USA Since brain metastasis constitutes a large spectrum of
e-mail: [email protected]; patients with significant oncologic and clinical heterogene-
[email protected]; ity, the possible management strategies for brain metastases
[email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 567
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_53
568 A. N. Slade et al.

can range from conservative management (e.g., steroids 53.3  ostsurgical Management of Brain
P
alone) up to surgical management and/or radiosurgery. Metastases: Evolution from Whole
This chapter is organized as follows: we first review prog- Brain Radiation Therapy
nostic approaches in patients with brain metastases and then to Radiosurgery
focus on the use of radiosurgery for brain metastasis, both
in the intact and postoperative setting. We also review some Surgery, along with radiation therapy, remains the mainstay
of the emerging data supporting the addition of hippocam- of treatment in patients with most solitary brain metastases or
pal avoidance and memantine to reduce the risk of cogni- large lesions causing neurological symptoms. While radiation
tive decline among patients receiving whole brain radiation therapy alone is an option for patients even with a single brain
therapy. metastasis, there are several clinical scenarios which may make
a surgical approach attractive. First, it can clarify pathologic
diagnosis, especially if a primary diagnosis does not exist or is
53.2  rognostic Approaches in Patients
P not consistent with the development of brain metastasis (e.g., a
with Brain Metastasis patient with prostate cancer after a prostatectomy with no evi-
dence of disease recurrence). It may also provide molecular or
Given how common brain metastases are in routine clini- genetic information toward the use of targeting agents when
cal practice, efforts have been made to stratify patients into the information is otherwise not available. Finally, surgical
prognostic groups. One of the original approaches was the management can relieve mass effect from symptomatic metas-
Recursive Partitioning Analysis (RPA) by Gaspar et al. [6] tases much faster than nonsurgical approaches.
who used three variables to create three classes of patients. Despite the benefits of surgical management of brain
The variables included Karnofsky Performance Status metastases, the risk of local failure without additional
(KPS), presence of extracranial disease and well-controlled metastasis-­ directed radiotherapy remains high in most
primary disease, and age to develop three classes (i.e., parti- series; approximately 50% of the resected brain metastasis
tions) of patients with brain metastasis. Patients meeting the ultimately develops local recurrence without adjuvant radio-
criteria of Class I, consisting of the most favorable charac- therapy [10]. Indeed, infiltrative growth beyond the metas-
teristics (KPS > = 70, age < 65, and well-controlled primary tasis boundary was detected in up to 63% of cases in one
and no extracranial disease), had a median overall survival pathological series of resected brain metastases [11]. This
of ≈7 months. Alternatively, patients meeting Class III crite- may partially explain the high recurrence rate after surgery
ria (KPS < 70) had a median overall survival of ≈2 months. alone and necessity of adjuvant treatment.
All other patients (Class II) had a median overall survival Radiotherapy to brain metastases, either in the intact or
of ≈4 months. Since then, many other groupings have been postoperative setting, is typically delivered in one of two
proposed. ways. WBRT treats both the surgical resection site and the
Sperduto et al. [7] further optimized stratification of remaining brain which has been thought to harbor subclinical
these patients by developing the Graded Prognostic Assess- metastatic or leptomeningeal disease and thus lowers the risk
ment (GPA), which incorporated age, KPS, systemic dis- of development of additional metastasis in the brain. Alterna-
ease, number of lesions, and volume of disease. Across tively, stereotactic radiosurgery delivers a very high dose of
all malignancies in this classification system, KPS is the radiation directed only to the metastasis (or surgical cavity),
strongest and most robust predictor for survival. As both without treating the uninvolved tissue. Indeed, randomized
prognostic and predictive mutations emerged for specific data has supported the role of postoperative WBRT in reduc-
cancers, cancer-­specific Graded Prognostic Assessment [2] ing the risk of new distant in-brain recurrences but without any
was defined to reflect factors such as disease site and histol- improvement of overall survival [12]. However, improved dis-
ogy, number of lesions, tumor gene status (e.g., EGFR and tant disease control is a potential benefit to WBRT over SRS,
ALK gene alterations in non-small-cell lung cancer [8] and as both leptomeningeal disease and distant brain failure can
BRAF mutation in melanoma [9]), and other clinical char- be common after postoperative radiosurgery [13]. However, it
acteristics (e.g., hemoglobin level in renal cell carcinoma). comes at the cost of several acute and long-term side effects,
Under the updated GPA classification, for instance, a patient most notably neurocognitive decline [14]. By using postop-
with excellent performance status and a more favorable his- erative radiosurgery (often with a single dose of 14–16 Gy
tologic profile (e.g., breast cancer) may have a survival in to the surgical cavity), most institutional series reported that
excess of 2 years. tumor control of the resected site improved to 80–90% [15,
53 Radiotherapy for CNS Metastases 569

16], and randomized data confirms similarly excellent levels cavity shrinkage was seen within 2 weeks postoperatively. As
of local control with postoperative SRS as well [17]. When such, this will typically limit the volume of radiation delivered
there is gross residual disease after surgical resection (which to both the cavity and normal brain parenchyma (Fig. 53.1,
often manifests as contrast-­enhancing lesions), the gross dis- panel B). However, especially for small lesions, it is possible
ease typically needs to be treated to higher doses. A common that the cavity may actually have a larger volume compared
approach to postoperative radiosurgical planning has been to pre-operatively [21]. In terms of timing of radiosurgery,
presented by Robbins et al. [18] and summarized in Fig. 53.1, we recommend commencing treatment relatively expediently
panel A. postoperatively soon after the patient recovered from surgery,
Increasing evidence has been put forth to support the rou- usually within 2–4 weeks following surgery.
tine use of radiosurgery (where possible) over whole brain An emerging strategy in the management of brain metas-
radiation therapy. A recent randomized clinical trial com- tases involves treating the intact tumor before surgical resec-
pared patients with ≤ 3 brain metastases to postoperative SRS tion. This approach allows for relatively straightforward
(12–20 Gy) vs. whole brain radiation therapy (30–37.5 Gy target delineation and also has the theoretical advantages of
in 10–15) fractions [19]. Overall, this study found a similar limiting leptomeningeal spread secondary to surgical inter-
overall survival of ≈1 year for both groups of patients, but a vention and sterilizing the surgical field. Indeed, institutional
lower rate of cognitive decline among patients receiving SRS series have reported promising results with similar rates of
compared to whole brain radiation. local control in pre-operative compared to postoperative
In addition to decompressing the mass effect of a large SRS [22], with lower rates of symptomatic radionecrosis and
metastasis and relieving neurological symptoms, surgery may leptomeningeal disease [23]. A common approach involves
render the cavity smaller compared to the pre-surgical tumor utilizing single-fraction radiosurgery with a reduction of the
volume. An early single-institutional series found that there treatment dose by 10–20%, with short interval from delivery
was an ≈29% reduction in the size of the surgical cavity (com- SRS to surgery (median of 2 days) [24].
pared to pre-operatively) [20]. This was most pronounced in Finally, an area of continued interest involves the role of
larger tumor sizes, and it was found that most of the surgical radiosurgery in potentiating an abscopal effect along with

Radiosurgery Target Dose

Surgical cavity + 2-3 mm margin Median dose 16 Gy (12-20 Gy),

or clinical decision prescribed to 90% isodose line

Gross residual C+ lesion ≥ 18 Gy

Pre-op Post-op SRS target 6 month post-SRS

Fig. 53.1 Postoperative treatment with stereotactic radiosurgery. dose ranging from 12 to 20 Gy (median 16 Gy) in the single fraction
Example of the use of stereotactic radiosurgery in the treatment of a setting. Gross residual disease (typically contrast enhancing) may
postoperative brain metastasis. Panel A depicts an example of a treat- require higher doses for local control. Panel B depicts serial MRI scans
ment plan in a postoperative cavity (left) and common target volumes showing a brain metastasis pre-operatively (left), the smaller postopera-
and doses (right). In general, the surgical cavity is contoured, and an tive cavity (middle), and a 6-month follow-up scan after postoperative
additional margin of 2–3 mm is added. This volume is prescribed to a radiosurgery
570 A. N. Slade et al.

systemic immunotherapy, in which an irradiated tumor may number of metastases that can be treated with radiosurgery
induce a regression in non-irradiated sites [25]. It is thought alone, the omission of WBRT is often preferable given its
that irradiation induces improved antigen expression and pre- morbidity and lack of survival benefit.
sentation, which combined with increased T cell function can While much of the existing literature focused on patients
result in an abscopal response [26]. There has been historic with limited numbers of brain metastases (e.g., four and
belief that extracranial abscopal effects from brain irradiation below), there is increasing credence to using radiosurgical
(either SRS or WBRT) would be uncommon due to a unique approaches in patients with a higher number and volume of
immune environment and the blood brain barrier. However, lesions. This is increasingly important with the advent of
several reports have been noted of this phenomenon, espe- sophisticated high-resolution MRI scans that can detect very
cially related to metastases from lung cancer [27, 28] and mel- small lesions. As treatment planning techniques have evolved
anoma [29]. It remains to be an area of active investigation. to allow very limited doses to surrounding brain parenchyma
and surrounding other critical structures, it is possible to safely
treat larger number of lesions. Indeed, institutional series con-
53.4  adiosurgery for Patients with Intact
R sidering treatment of >10 lesions with radiosurgery generally
Brain Metastases have found outcomes similar to series considering more lim-
ited number of metastases [36–38]. Indeed, in the JLGK0901
As with radiosurgery delivered postoperatively for resected which enrolled patients with 2–10 newly diagnosed brain
brain metastases, use of radiosurgery for intact metastases metastases treated with SRS alone, the median survival of
(treatment-naïve or unresected) confers excellent local con- patients with two to four metastases was not significantly dif-
trol, often ranging from 70 to 90% at 1 year in institutional ferent compared to those with 5–10 metastases (median over-
series [30, 31]. The early dose-escalation experience of the all survival was 10.8 months in both groups).
RTOG 90–05 trial provided a framework by which SRS doses
(in the single-fraction setting) are typically used with accept-
able rates of toxicity [32]. Based on the experience of this trial, 53.5  pproaches to Limit Cognitive
A
recommended doses are 24 Gy for tumors 2 cm or smaller Decline in Patients Receiving Whole
and 15 Gy for tumors between 3 and 4 cm. Early experiences Brain Radiotherapy
explored the value of SRS as a “boost” to whole brain radia-
tion therapy. RTOG 9508 considered patients with a limited While radiosurgery has been increasingly used in the manage-
number of brain metastases (<3 metastases) who were ran- ment of brain metastases, WBRT still has utility, especially in
domized between WBRT alone and WBRT alone + SRS patients with multiple metastases, and offers increased control
boost, which found a stable or improved performance status of subclinical and leptomeningeal disease. However, as dis-
at 6 months follow-up among patients with an SRS compared cussed above, a common complication of WBRT is cognitive
to those without (43% vs. 27%; p = 0.03) [33]. While survival decline, particularly with the memory recall [39]. The pre-
among the entire cohort was similar, patients with favorable cise mechanisms by which WBRT leads to delayed recall are
histologic status or RPA class 1 had improved overall survival. uncertain. However, there are two predominant theories that
More recently, efforts have been made to curtail the use have been proposed: destruction of hippocampal stem cells
of WBRT in patients who are otherwise good candidates [40, 41] and radiation-induced vascular damage [42].
for radiosurgery, such as patients with a good performance Two trials attempted to address both putative systemic and
status and a limited number of metastases. As such, studies technical interventions to potentially lower the risk of cog-
have investigated the role of SRS with or without WBRT. A nitive decline in WBRT. RTOG 0614 randomized patients
randomized multi-institutional clinical trial from Japan was receiving WBRT to placebo vs. memantine, an n-methyl-d-­
performed comparing WBRT+ radiosurgery versus radiosur- aspartate (NMDA) antagonist (classically used for dementia),
gery alone in patients with one to four brain metastases each on the trajectory of cognitive decline after patients receive
less than 3 cm [34]. This trial found similar tumor control WBRT [43]. Though the results for this endpoint were not
with SRS alone although distant intracranial relapse was statistically different between the placebo and memantine
more common; however, there was no difference in overall arms (p = 0.059), patients in the memantine arm did have a
survival. A more recent EORTC 22952–22,601 study ran- significantly slower time to cognitive decline. Alternatively,
domized patients with one to three brain metastases treated RTOG 0933 considered the role of radiation planning to
with either surgery or SRS to adjuvant WBRT vs. obser- avoid the hippocampus on cognitive decline, which found
vation. As with the experience from Japan, the addition of the absolute value of cognitive decline significantly lower
WBRT decreased the rate of intracranial relapse, but did than historical controls [5, 44]. Hippocampal avoidance may
not improve the duration of neurological independence or be technically achieved by delineating the hippocampus via
overall survival [35]. As such, among patients with limited image fusion of a T1-weighted MRI sequence to the planning
53 Radiotherapy for CNS Metastases 571

CT scan. Through intensity-­ modulated radiation therapy Life after Treatment for Brain Metastases (QUARTZ) trial
(IMRT) planning, the dose to the hippocampus can be lim- randomized patients deemed ineligible for radiosurgery or
ited. A common dose constraint (used in the NRG CC001 surgical resection (many with poor KPS) to optimal support-
trial discussed below) involves limiting 100% of the hippo- ive care with or without WBRT. The trial found no significant
campal volume to 9 Gy, with a maximum dose of 16 Gy. difference in overall survival or quality of life [47].
Despite the promising results of these phase II trials,
overall uptake of these modalities remained relatively lim-
ited across the United States, with a perceived need for 53.6 Conclusions
randomized, phase III results [45]. The recently completed
NRG CC001 trial incorporated both the use of memantine With improved systemic therapy and patient stratification,
and hippocampal avoidance into the study design [46]. In the management of brain metastases is not a “one size fits
this trial, patients were randomized to memantine with either all” approach; indeed, it represents a spectrum of diseases.
hippocampal-sparing WBRT or standard WBRT. The trial Patients with widespread brain metastases may be managed
found that at a median follow-up of 7.8 months, the risk with whole brain radiotherapy, with or without hippocampal
of cognitive failure was lower after hippocampal-sparing sparing, whereas patients with fewer brain metastases and
WBRT compared to conventional WBRT. improved performance status may be managed with stereo-
Finally, in patients with poor performance status and very tactic radiosurgery. Still others with few metastases may be
limited prognoses (e.g., KPS < 70), the omission of metastasis-­ managed surgically, especially if these metastases are symp-
directed treatment entirely, and using supportive care (includ- tomatic, followed by postoperative radiation therapy. The
ing steroids) alone, is a reasonable option. The Quality of spectrum of treatment options is displayed in Fig. 53.2.

Conservative
Hippocampal
management Whole Brian RT Radiosurgery Surgery + RS
sparing WBRT
(steroids alone)

Limited number of brain

One or multiple Multiple brain Multiple brain Limited number
metastasis, especially
brain metastases metastases metastases of brain metastases
if symptomatic

Poorer Any
Very poor Better performance Better performance
performance performance
performance status status status
status status

Non-invasive Non-invasive Non-invasive Non- or minimally Invasive surgery

invasive
Lower risk of
Lower risk of
No risk of RT Risk of cognitive Lower risk of cognitive decline,
cognitive decline,
related side effects decline cognitive decline radionecrosis, risk
radionecrosis
involved in surgery

Better performance status,

fewer co-morbidities, well controlled
extracranial disease, improved prognosis, increasingly symptomatic.

Least aggressive Most aggressive

Fig. 53.2 Contemporary management strategies for patients with symptomatic lesions may benefit from surgical management, which
brain metastases. The spectrum of management options for patients may provide rapid relief of symptoms and provide a smaller cavity vol-
with brain metastases range from no metastasis-directed therapy (i.e., ume for irradiation postoperatively. In general, less aggressive options
conservative management) to invasive surgical options followed by are preferred in patients with a poor prognosis, increased comorbidities,
postoperative stereotactic radiation therapy. Patients with large or and poorer performance status
572 A. N. Slade et al.

For some patients, clear demonstration of reducing the 7. Sperduto PW, et al. Summary report on the graded prognostic
assessment: an accurate and facile diagnosis-specific tool to esti-
risk of developing new metastasis justifies the option of mate survival for patients with brain metastases. J Clin Oncol Off J
WBRT. This may be motivated by the fear of symptomatic Am Soc Clin Oncol. 2012;30:419–25.
recurrence or the desire to also reduce the need for future 8. Sperduto PW, et al. Estimating survival in patients with lung cancer
treatment. For others, fear of irreversible toxicity such as and brain metastases: an update of the graded prognostic assess-
ment for lung cancer using molecular markers (lung-molGPA).
neurocognitive decline given the lack of survival benefit JAMA Oncol. 2017;3:827–31.
drives toward SRS provided all the known tumors are treated. 9. Sperduto PW, et al. Estimating survival in melanoma patients with
The clinical and pathologic heterogeneity of patients with brain metastases: an update of the graded prognostic assessment
brain metastases, combined with diverse treatment options for melanoma using molecular markers (melanoma-molGPA). Int J
Radiat Oncol Biol Phys. 2017;99:812–6.
available for its management, suggests a nuanced approach 10. Patel AJ, et al. Factors influencing the risk of local recur-
that requires shared decision-making between the patient rence after resection of a single brain metastasis. J Neurosurg.
and his/her surgical, medical, and radiation oncologists to 2010;113:181.
design the optimal treatment plan. 11. Baumert BG, et al. A pathology-based substrate for target definition
in radiosurgery of brain metastases. Int J Radiat Oncol Biol Phys.
2006;66:187–94.
12. Patchell RA, et al. Postoperative radiotherapy in the treatment
Future Directions of single metastases to the brain: a randomized trial. JAMA.
1998;280:1485–9.
• How will prognostic classifications evolve in the 13. Marcrom SR, et al. Focal management of large brain metastases and
future to more optimally stratify patients with brain risk of leptomeningeal disease. Adv Radiat Oncol. 2020;5:34–42.
metastases? 14. Brown PD, et al. Effect of radiosurgery alone vs radiosurgery with
• As technological advances improve, allowing for whole brain radiation therapy on cognitive function in patients
with 1 to 3 brain metastases: a randomized clinical trial. JAMA.
faster and more widespread delivery of stereotactic 2016;316:1–409.
radiosurgery, what will be the role for whole brain 15. Soltys SG, et al. Stereotactic radiosurgery of the postoperative
radiation therapy? resection cavity for brain metastases. Int J Radiat Oncol Biol Phys.
• Are there benefits from combining SRS with tar- 2008;70:187–93.
16. Gans JH, et al. The role of radiosurgery to the tumor bed after resec-
geted therapies and/or immunotherapy? tion of brain metastases. Neurosurgery. 2013;72:317–25. discus-
• What will be the optimal role for pre-operative sion 325-6
SRS, and what will be the optimal timing and dose 17. Mahajan A, et al. Post-operative stereotactic radiosurgery versus
parameters? observation for completely resected brain metastases: a single-­
centre, randomised, controlled, phase 3 trial. Lancet Oncol.
• What are potential future interventions to reduce 2017;18:1040–8.
the rates of toxicities in patients receiving radio- 18. Robbins JR, et al. Radiosurgery to the surgical cavity as adju-
therapy for brain metastases? vant therapy for resected brain metastasis. Neurosurgery.
2012;71:937–43.
19. Brown PD, et al. Postoperative stereotactic radiosurgery compared
with whole brain radiotherapy for resected metastatic brain disease
(NCCTG N107C/CEC·3): a multicentre, randomised, controlled,
phase 3 trial. Lancet Oncol. 2017;18:1049–60.
References 20. Atalar B, et al. Cavity volume dynamics after resection of brain
metastases and timing of postresection cavity stereotactic radiosur-
1. Sacks P, Rahman M. Epidemiology of brain metastases. Neurosurg gery. Neurosurgery. 2013;72:180–5. discussion 185
Clin N Am. 2020;31:481–8. 21. Jarvis LA, et al. Tumor bed dynamics after surgical resection of
2. Sperduto PW, et al. Survival in patients with brain metastases: sum- brain metastases: implications for postoperative radiosurgery. Int J
mary report on the updated diagnosis-specific graded prognostic Radiat Oncol Biol Phys. 2012;84:943–8.
assessment and definition of the eligibility quotient. J Clin Oncol. 22. Asher AL, et al. A new treatment paradigm: neoadjuvant radio-
2020;38:3773–84. surgery before surgical resection of brain metastases with analy-
3. American Society for Radiation Oncology. Choosing wisely: five sis of local tumor recurrence. Int J Radiat Oncol Biol Phys.
things physician and patients should question. American Society 2014;88:899–906.
for Radiation Oncology; 2014. https://www.choosingwisely.org. 23. Patel KR, et al. Comparing preoperative with postoperative ste-
Accessed 3/1/2022. reotactic radiosurgery for resectable brain metastases: a multi-­
4. Kavouridis VK, et al. Survival and prognostic factors in surgically institutional analysis. Neurosurgery. 2015;79:279–85.
treated brain metastases. J Neuro-Oncol. 2019;143:359–67. 24. Prabhu RS, et al. Preoperative stereotactic radiosurgery before
5. Gondi V, et al. Preservation of memory with conformal avoidance planned resection of brain metastases: updated analysis of effi-
of the hippocampal neural stem-cell compartment during whole-­ cacy and toxicity of a novel treatment paradigm. J Neurosurg.
brain radiotherapy for brain metastases (RTOG 0933): a phase II 2018:1–8.
multi-institutional trial. J Clin Oncol. 2014;32:3810–6. 25. Siva S, et al. Abscopal effects of radiation therapy: a clinical review
6. Gaspar L, et al. Recursive partitioning analysis (RPA) of prog- for the radiobiologist. Cancer Lett. 2015;356:82–90.
nostic factors in three radiation therapy oncology group 26. Reynders K, et al. The abscopal effect of local radiotherapy: using
(RTOG) brain metastases trials. Int J Radiat Oncol Biol Phys. immunotherapy to make a rare event clinically relevant. Cancer
1997;37:745–51. Treat Rev. 2015;41:503–10.
53 Radiotherapy for CNS Metastases 573

27. Lin X, et al. Extracranial abscopal effect induced by combining hippocampal sparing in the treatment of multiple brain metastases.
immunotherapy with brain radiotherapy in a patient with lung ade- World Neurosurg. 2020;135:e174–80.
nocarcinoma: a case report and literature review. Thoracic Cancer. 38. Rava P, et al. Survival among patients with 10 or more brain
2019;10:1272–5. metastases treated with stereotactic radiosurgery. J Neurosurg.
28. D'Andrea MA, Reddy GK. Extracranial abscopal effects induced 2013;119:457.
by brain radiation in advanced lung cancer. Am J Clin Oncol. 39. Chang EL, et al. Neurocognition in patients with brain metastases
2019;42:951–7. treated with radiosurgery or radiosurgery plus whole-brain irradia-
29. D'Andrea MA, Reddy GK. Brain radiation induced extracra- tion: a randomised controlled trial. Lancet Oncol. 2009;10:1037–44.
nial abscopal effects in metastatic melanoma. Am J Clin Oncol. 40. Monje ML, et al. Irradiation induces neural precursor-cell dysfunc-
2020;43:836–45. tion. Nat Med. 2002;8:955–62.
30. Minniti G, et al. Single-fraction versus multifraction (3Gy) stereo- 41. Gondi V, et al. Hippocampal dosimetry predicts neurocognitive
tactic radiosurgery for large brain metastases: a comparative analy- function impairment after fractionated stereotactic radiotherapy for
sis of local control and risk of radiation-induced brain necrosis. Int benign or low-grade adult brain tumors. Int J Radiat Oncol Biol
J Radiat Oncol Biol Phys. 2016;95:1142–8. Phys. 2012;83:e487–93.
31. Hasegawa T, et al. Brain metastases treated with radiosurgery 42. Brown WR, et al. Vascular damage after fractionated whole-brain
alone: an alternative to whole brain radiotherapy? Neurosurgery. irradiation in rats. Radiat Res. 2005;164:662–8.
2003;52:1318–26. 43. Brown P, et al. Memantine for the prevention of cognitive dys-
32. Shaw E, et al. Single dose radiosurgical treatment of recurrent function in patients receiving whole-brain radiotherapy: a ran-
previously irradiated primary brain tumors and brain metastases: domized, double-blind, placebo-controlled trial. Neuro-Oncology.
final report of RTOG protocol 90-05. Int J Radiat Oncol Biol Phys. 2013;15:1429–37.
2000;47:291–8. 44. Gondi V, et al. Memory preservation with conformal avoidance of
33. Andrews DW, et al. Whole brain radiation therapy with or without the hippocampus during whole-brain radiation therapy for patients
stereotactic radiosurgery boost for patients with one to three brain with brain metastases: primary endpoint results of RTOG 0933. Int
metastases: phase III results of the RTOG 9508 randomised trial. J Radiat Oncol Biol Phys. 2013;87:1186.
Lancet. 2004;363:1665–72. 45. Slade AN, Stanic S. The impact of RTOG 0614 and RTOG 0933
34. Aoyama H, et al. Stereotactic radiosurgery plus whole-brain trials in routine clinical practice: the US survey of utilization of
radiation therapy vs stereotactic radiosurgery alone for treat- memantine and IMRT planning for hippocampus sparing in patients
ment of brain MetastasesA randomized controlled trial. JAMA. receiving whole brain radiotherapy for brain metastases. Contemp
2006;295:2483–91. Clin Trials. 2016;47:74–7.
35. Kocher M, et al. Adjuvant whole-brain radiotherapy versus obser- 46. Brown PD, et al. Hippocampal avoidance during whole-brain radio-
vation after radiosurgery or surgical resection of one to three cere- therapy plus memantine for patients with brain metastases: phase
bral metastases: results of the EORTC 22952-26001 study. J Clin III trial NRG oncology CC001. J Clin Oncol. 2020;38:1019–29.
Oncol. 2010;29:134–41. 47. Mulvenna P, et al. Dexamethasone and supportive care with or
36. Grandhi R, et al. Stereotactic radiosurgery using the Leksell gamma without whole brain radiotherapy in treating patients with non-­
knife Perfexion unit in the management of patients with 10 or more small cell lung cancer with brain metastases unsuitable for resec-
brain metastases. J Neurosurg. 2012;117:237–45. tion or stereotactic radiotherapy (QUARTZ): results from a phase
37. Susko MS, et al. Stereotactic radiosurgery to more than 10 brain 3, non-inferiority, randomised trial. Lancet. 2016;388:2004–14.
metastases: evidence to support the role of radiosurgery for ideal
 Leptomeningeal Carcinomatosis
54
Sunny R. K. Singh, Sindhu J. Malapati,
and Ahmad Mattour

Abstract
Learning Objectives
The prevalence of leptomeningeal carcinomatosis (LC) • Discuss the epidemiology, pathophysiology, clini-
has increased due to improved survival of cancer patients; cal features, and outcomes of leptomeningeal
however, it continues to be associated with poor outcomes carcinomatosis.
despite multiple advances in cancer-directed therapy. • Describe the workup of leptomeningeal carcinoma-
Clinical presentation is highly variable, and its evaluation tosis and its limitations.
continues to be largely guided by clinical signs and symp- • Analyze the role and benefit of currently approved
toms, findings on craniospinal magnetic resonance imag- treatment options for leptomeningeal carcinomatosis.
ing (MRI), and cerebrospinal fluid (CSF) cytology. • Outline the recent advances both in terms of evalu-
Recently, the ability to detect circulating cancer cells and ation and possible future treatment options for lep-
tumor DNA in the CSF has improved the accuracy of tomeningeal carcinomatosis.
diagnosing LC in those with false-negative cytology. It
has also uncovered the potential of utilizing CSF for
molecular analysis using next-generation sequencing.
This not only furthers our understanding of this clinical
entity but also allows us make treatment decisions based 54.1 Introduction
on the presence of certain resistance mutations. To date,
there is no standard approach to the treatment of Leptomeningeal carcinomatosis (LC), also known as neo-
LC. Historically, treatment choices are limited to sys- plastic meningitis, is the spread of malignancy, within the
temic or intra-CSF cytotoxic chemotherapy and/or radia- pia mater and/or arachnoid components of the central ner-
tion. With the advent of novel agents, we now have several vous system (CNS), aka leptomeninges. It is usually a late
potential therapeutic agents being studied prospectively. complication of advanced malignancy with devastating con-
Also, the inclusion of patients with LC in recent clinical sequences on neurological outcomes, quality of life, and
trials marks the beginning of long overdue advances that survival. Without treatment, median overall survival has
would lead to improved outcomes in this cohort of been reported to be around 6–8 weeks. Even with adoption
patients. of multiple newer multimodal treatments, survival benefits
remain poor (3–9 months with intrathecal chemotherapy)
[1–3]. Standard of care for the management of LC is still
not well established, and the goal of treatment is largely pal-
liative. Treatment outcomes to be considered while deciding
S. R. K. Singh
therapy include treatment-associated toxicity, time to neu-
Department of Hematology/Oncology, Henry Ford Cancer
Institute, Detroit, MI, USA rologic progression (TTNP), impact on quality of life, and
overall survival.
S. J. Malapati
Department of Hematology/Oncology, Ascension St. John
Hospital, Detroit, MI, USA
A. Mattour (*)
Department of Hematology/Oncology, Henry Ford Cancer
Institute, Detroit, MI, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 575
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_54
576 S. R. K. Singh et al.

54.2 Epidemiology 54.4.1 Differential Diagnosis

The estimated frequency of leptomeningeal carcinomatosis This condition is frequently misdiagnosed due to vague pre-
is around 5–8% in solid malignancies and 5–15% in hema- senting symptoms and non-specific findings on neurological
tological malignancies [4], but this does not account for imaging. Oftentimes, LC is diagnosed at the same time as
the significant underdiagnosis of this condition. The com- parenchymal or dural involvement [8]. Differentials include
bination of newer, life-prolonging systemic therapies and meningitis (infectious or chemical), parenchymal brain
improved access to imaging modalities has resulted in what metastases at multiple locations, encephalopathy (toxic or
is described as long-term cancer patients and, inadvertently, metabolic), neuropathy secondary to chemotherapy, cord
higher incidences of late metastasis, including LC [5]. Solid compression, steroid myopathy, and paraneoplastic syn-
malignancies most commonly associated with LC are breast, dromes (limbic encephalitis, encephalomyelitis, paraneo-
lung, melanoma, and gastrointestinal cancers and hemato- plastic cerebellar degeneration).
logical malignancies.

54.4.2 Clinical Presentation

54.3 Pathogenesis
LC presentation can be highly variable. In cases with high
Several mechanisms are implicated in the development of LC, metastatic tumor burden, classical LC symptoms may be
including direct extension from surrounding structures (bone, more clearly defined. Occasionally LC could even be the ini-
brain, dura), hematologic spread through venous plexus, and tial manifestation of malignancy. The spectrum of symptoms
perineural extension [6]. It must be pointed out that lepto- remains broadly vague in nature, and a high index of sus-
meninges include the arachnoid and pia mater, but not the picion remains important to make a timely diagnosis [10].
dura mater (aka pachymeninges). Dura is not protected by the In a study of 103 patients, more than 80% had a neurologi-
blood-brain barrier and therefore can be accessed by systemic cal sign or symptom at diagnosis of leptomeningeal disease
agents, while the leptomeninges are limited in this regard. In [11]. Classically, the symptoms of LC have been categorized
a recent preclinical study, complement component 3 (C3) according to the site of neural involvement [3]:
was found to be upregulated and necessary for cancer growth
in the leptomeningeal space. C3 derived from cancer cells 1. Cerebral involvement usually presents with a variety of
activates the C3a receptor in choroid plexus epithelium and symptoms such as headaches, dizziness, nausea, seizures,
disrupts the BBB, allowing mitogens to enter the CSF from and encephalopathy. Another manifestation could be a
plasma and promote cancer cell proliferation [7]. Involve- hydrocephalus.
ment is very common in dependent portions of the CNS like 2. Symptoms due to involvement of the cranial nerves by
the cauda equina, base of the skull, posterior fossa, and also LC manifest with findings of diplopia, vertigo, hearing
basal cisterns. Despite extensive study, the exact mechanism loss, and facial numbness.
of development of LC is not known and is still being evalu- 3. Involvement of the spinal cord and nerve roots can pres-
ated. Malignancy and intrathecal therapies can cause local ent with symptoms of myelopathy such as limb weak-
inflammation resulting in impaired resorption of CSF and ness, loss of sensation, pain, and paresthesia.
obstruction of flow, and this may manifest with symptoms of
elevated intracranial pressure [8]. Proposed mechanisms leading to these symptoms include
parenchymal compression or invasion, ischemia secondary
to vascular involvement, blood-brain barrier disruption, and
54.4 Diagnostic Evaluation metabolic strain [12].

Diagnosis of LC is based on clinical features, findings on

craniospinal MRI, and cerebrospinal fluid (CSF) analysis. 54.4.3 Radiological Evaluation
While typical clinical and imaging signs are contributory
to making the diagnosis, the detection of tumor cells in the About 70–80% patient with leptomeningeal disease are
CSF is confirmatory in a patient with known malignancy. noted to have abnormalities on imaging [13]. The test of
Among those with known cancer and without CSF confirma- choice is a T1-weighted MRI of the brain and spine with
tion, diagnosis of LC is probable if both typical clinical and gadolinium contrast enhancement [14]. Cranial computed
imaging signs are present and possible if only one of them tomography can be used when there are contraindications for
is present [9]. the use of MRI. Typical findings of neoplastic meningitis on
54 Leptomeningeal Carcinomatosis 577

a contrast-enhanced MRI include enhancement or oblitera- LC have been defined based on MRI findings: linear lepto-
tion of sulci, linear ependymal enhancement, cranial nerve meningeal contrast enhancement only (type A), leptomen-
root enhancement, and leptomeningeal nodule enhancement, ingeal nodules only (type B), combination of both (type
notably of the cauda equina [15] (Fig. 54.1). Four types of C), and with no neuroimaging evidence of LC, except for

a b

c d

Fig. 54.1 Contrast-enhanced axial imaging of the supratentorial brain VII and VIII within the internal auditory canal and cerebellopontine
(a, b) demonstrates curvilinear enhancement of the pial surface of the angle (black block arrows), right cranial nerve V in lateral pontine cis-
brain from leptomeningeal disease (LMD). Extensive pial enhancement tern and Meckel’s cave (black arrows), as well as the folia of the cere-
involves the occipital and parietal cortex bilaterally (white arrows) as bellum (white block arrows). Republished with permission of Wiley for
well as a focal nodular deposit involving the right frontal cortex (white Glitza, IC, et al. Pigment Cell Melanoma Res. 2020; 33: 527–541; per-
curved arrow). Contrast-enhanced axial imaging of the posterior fossa mission conveyed through Copyright Clearance Center
(c, d) demonstrates nodular enhancing LMD involving cranial nerves
578 S. R. K. Singh et al.

Fig. 54.2 CSF cytological

examination (Short arrows:
MALIGNANT lymphocytes,
Long arrow: Normal
lymphocytes)

possibly hydrocephalus (type D) [9]. In order to avoid false of the leptomeninges and, if feasible, should be further ana-
positives, cerebrospinal MRI should be obtained before lum- lyzed with DNA sequencing and/or other molecular tech-
bar puncture or ventricular shunt placement because these niques [21]. Fluorescent in situ hybridization (FISH) is used
procedures could lead to contrast enhancement of the menin- in peripheral blood for the enrichment and identification of
ges. FDG-PET is not considered a standard diagnostic tool circulating tumor cells (CTCs) and isolation of tumor DNA,
in LC, but there are case reports describing its utility in this irrespective of cell size variation and specific surface marker
situation [16, 17]. expression [22]. This technique has recently been applied
to CSF. Circulating tumor DNA (ctDNA) has been shown
to be more abundant in CSF than in plasma, and this could
54.4.4 Cerebrospinal Fluid (CSF) Analysis translate to a better diagnostic yield. With increasing utiliza-
tion of driver mutation targeted therapy, it is very helpful that
Demonstrating tumor cells in the CSF remains the gold stan- driver and resistance mutations can also be identified in CSF
dard to diagnose LC (Fig. 54.2). Identifying cancer cells ctDNA in some patients with CNS metastases. This approach
using Papanicolaou staining has a yield of 50–60% in the had a sensitivity of 95.2% to detect LC, which was higher
first lumbar puncture (LP). Sensitivity can be increased to than that of thin prep cytology (57.1%), MRI (47.6%), and
80% with a second LP and by a further 2–5% in the third MRI plus cytology (90.5%), respectively [23–25]. Next-gen-
attempt. At least 5–10 ml of CSF should be drawn, hemor- eration sequencing of DNA from CSF tumor cells showed a
rhagic contamination should be avoided, and the sample high level of concordance with the primary tumor in terms
should be processed within 30 min after collection [18, 19]. of genetic profile [23]. Given the ease of access, CSF could
Flow cytometry to detect cancer cells is used primarily in be a potential source of a “liquid biopsy” with the option
the setting of hematological malignancies and was found to to investigate resistance mutations and determine optimal
have severalfold higher sensitivity to detect LC when com- treatment, especially when there is CNS progression despite
pared to conventional cytology. Also, it was found to be stable extra CNS lesions.
especially valuable in cases where cytology was equivocal
[20]. Even in the absence of pathologic cells, CSF altera-
tions can be observed in more than 90% of the patients with 54.4.5 Prognostic Factors
LC. These include increased opening pressure (> 200 mm
H2O) in 21–42%, increased CSF leukocyte count (> 4 mm−3) While prognosis of LC in general is poor, it is influenced
in 48–77%, elevated protein (> 50 mg/dl) in 56–91%, and by tumor-specific factors and treatment administered. The
decreased glucose (< 60 mg/dl) in 22–63% [13]. Meningeal INDEX score was recently proposed for patients with LC
biopsy is an option when there is a high pretest probabil- and a breast primary malignancy. This score correlated
ity with symptoms and/or imaging concerning for LC but overall survival with older age, performance status, lumi-
multiple negative CSF cytology testing. This is usually done nal subtype, and treatment intensity. Patients classified as
after ruling out an infectious or inflammatory etiology and having bad prognosis presented a median overall survival of
documented negative cytology on repeated CSF exams. 1.5 months compared to patients with good prognosis who
When biopsy is performed, it should be from a nodular area experienced a median overall survival of 9.6 months [26].
54 Leptomeningeal Carcinomatosis 579

Among those with LC in the setting of melanoma, factors 54.5.3 Chemotherapy

associated with poorer survival included ECOG perfor-
mance status greater than 0, neurological symptoms, ele- Pharmacokinetic studies involving systemic chemotherapy
vated LDH, and lack of concurrent systemic disease. Both during treatment of CNS tumors did not demonstrate evi-
treatment with targeted agents and intrathecal agents was dence of penetration of the BBB with effective, therapeutic
associated with improved outcomes, while radiation was concentrations [33]. Most chemotherapeutic agents are too
found to have no impact on survival [27]. In lung cancer large to pass across the BBB and enter the CNS. In addition
patients with LC, based on a retrospective review, good PS, to this, BBB is lined with drug efflux pumps, such as Pgp,
use of EGFR-­TKIs, and a nonsmoking status were associ- which actively transports toxic substances, including drugs,
ated with better outcomes [28]. out of the CNS. These highly functioning pumps ultimately
cause drug resistance in the CNS [34–36]. Animal models,
for example, have shown that only 3–5% of systemically
54.5 Treatment administered methotrexate penetrates the blood-brain barrier
(BBB) at a clearance halftime of 6 hours and requires days of
A generally accepted standard of care for treatment of LC continuous high-dose IV methotrexate infusions to achieve a
remains an elusive goal. The current management of LC therapeutic concentration (>1 μM) in CSF [37] [38]. Intra-
consists of single or multimodality treatment and is indi- thecal chemotherapy on the other hand aims to bypass the
vidualized on a case-by-case basis. The modalities available BBB, maximize drug exposure in the CSF, and limit sys-
include intrathecal chemotherapy, systemic therapy, radio- temic toxicity and remains the cornerstone of treatment of
therapy, or best-supportive care. LC. It achieves this by ensuring a higher drug concentration
despite a smaller dose due to the lower volume of distribu-
tion of the CSF compared to plasma (140 vs 3500 ml) [39].
54.5.1 Symptom Management For example, therapeutic concentration of methotrexate in
the CSF with minimal systemic absorption (<0.1 μM) can
There is no data supporting the use of primary seizure pro- be achieved for 48–72 hours with a single intraventricular
phylaxis in patients with LC [29]. For signs of increased ICP injection (6.25 mg/m2) [40]. Despite this being a high need
such as headache, nausea, or encephalopathy, high-volume area for research, only six randomized clinical trials evaluat-
lumbar puncture can be considered [8]. Ventriculoperitoneal ing treatment strategies for neoplastic meningitis have been
shunts also have a role to relieve symptomatic hydrocephalus completed. These have explored different regimens of intra-­
[30]. Opioids, steroids, and palliative focal radiation can be CSF chemotherapy or the role of adding intra-CSF chemo-
used to treat symptoms of meningeal irritation or radicular therapy to systemic treatment in patients with LC. Contrary
pain [8, 9]. to the observations in retrospective studies, OS benefit was
not seen in either prospective studies with the addition of IT
chemotherapy; however, PFS improvement was noted in one
54.5.2 Radiation Therapy [41–46].
Contrast enhancement in leptomeningeal lesion signal
Treatment of LC needs to be directed towards the entire towards the absence of an intact BBB and, therefore, sys-
neuroaxis to be effective, but this modality would be highly temic pharmacotherapy, if active, should be able to target
toxic. Whole-brain RT is therefore not an appropriate modal- these areas. Systemic agents should be selected according to
ity and has not been shown to improve survival, except when the primary tumor, clinical status, molecular characteristics,
there is co-existing brain metastasis [31]. However, it has a prior therapy, and how readily it crosses the BBB.
role in treating nodular/bulky disease to improve CSF flow Chemotherapy can be administered into the CSF
and in improving penetration of intra-CSF chemotherapy. It through either repeated lumbar punctures or ventricular
is also indicated as a palliative measure for symptomatic sites devices. Ventricular devices offer the benefit of comfort for
and cauda equina or skull-base syndromes despite absence of the patient, ability to achieve an even distribution of the
radiographic signs [32]. drug, and an extra layer of safety by avoiding inadvertent
580 S. R. K. Singh et al.

delivery of the drug into the epidural or subdural space. insufficient in itself to define progressive disease, especially
The most commonly placed device is the Ommaya reser- when it occurs in the presence of stable or improved clini-
voir, a subcutaneous reservoir attached to an intraventricu- cal or radiographic findings. It is recommended to get an
lar catheter. In a 616-patient study reporting an infection MRI every 2 months for 6 months and then every 3 months
rate of 5.5% following placement of Ommaya reservoir, if the disease shows stability. Treatment response based
infection was thought to be related to placement of the on CSF cytology is defined as the conversion of cytology
device in one third and due to subsequent accessing of from positive to negative, with negativity confirmed again
the reservoir in others [47]. Revision rates of ventricular after 4 weeks. Also, the proposal in its current format does
devices were found to be below 7.4% in different cohorts of not incorporate patient-reported outcomes and has yet to be
patients, and this highlights an overall favorable safety pro- validated [14]. Recently however, CSF tumor cell detection
file of these devices [48]. Before each intra-CSF injection, using FISH has provided us the ability to measure CSF tumor
the same amount of volume as that of the chemotherapy cells in a dynamic manner. The number of CSF tumor cells
should be removed and sent for analysis. One of the relative was shown to be concordant with headache alleviation and
contraindications of intra-CSF pharmacotherapy includes dynamic changes of intracranial pressure [22]. Also, changes
patients who have obstructive gross disease or symptom- in CSF tumor cell counts predicted treatment response [18].
atic hydrocephalus. A ventriculoperitoneal shunt could be Changes in CSF tumor cell counts can be used to monitor
considered in select such cases prior to therapy [49]. The disease progression and treatment response since this param-
most commonly used drugs on a routine basis for intrathe- eter correlates well with the leptomeningeal disease state.
cal administration include methotrexate, cytarabine, lipo-
somal cytarabine, and thiotepa. Historically, other agents
have been tried, but none have been shown to be better 54.5.5 R
 ecent Developments and Ongoing
than the abovementioned standard agents [12]. One of the Prospective Studies
side effects of intra-CSF administration of chemotherapy
includes aseptic chemical meningitis, and this is transient. There are currently many ongoing and recently completed
The risk is higher with agents characterized by long half- trials that include novel therapies in the management of lep-
life, for example, liposomal cytarabine, and can be reduced tomeningeal disease (Table 54.1; Source: www.clinicaltrials.
by the simultaneous administration of dexamethasone [32]. gov, last accessed on 11/29/2020). Full results and comple-
Intra-CSF use of novel agents and targeted therapies in tion of these trials are eagerly awaited in this population
patients with LC (e.g., trastuzumab in breast cancer, inter- which currently has limited options with a minimal to mod-
leukin-­2 in melanoma, and panitumumab in NSCLC) has erate efficacy at best.
been reported; however, long-term results from prospective
studies are awaited [50–52].
54.6 Conclusion

54.5.4 Assessment of Response to Treatment • The frequency of leptomeningeal carcinomatosis is esti-

mated to be at least 5–8% in those with solid malignancies
The difficulties with assessing treatment response in patients and 5–15% in those with hematological malignancies.
with LC include inconsistencies associated with CSF cytol- • A majority present with some neurological signs or
ogy, lack of quantitative measurement of the disease, and symptoms.
subjectivity associated with LC-related symptoms ([3, 44, • Sensitivity of CSF circulating tumor DNA in the diagno-
53, 54]. Some of this has been attempted to be addressed sis of LC is much higher than that of CSF cytology.
by the response assessment in Neuro-Oncology (RANO) • DNA sequencing of the CSF circulating tumor DNA
working group, which has drafted a consensus proposal for offers an attractive avenue for liquid biopsy and DNA
­evaluating response in patients with LC. The goal is to stan- sequencing.
dardize response definitions in forthcoming LC trials and per- • Treatment is based on the clinical context and is primarily
mit cross trial comparisons. The tools included in response through systemic therapy directed at the primary or intra-­
assessment include a standardized neurological exam, CSF CSF chemotherapy (methotrexate, cytarabine, thiotepa).
cytology or flow cytometry, and radiographic evaluation.
Based on the above, progressive disease is defined by either Many promising novel agents are currently under study
deteriorating findings on imaging or neurological exam. and will hopefully change the treatment landscape in the near
A persistently positive CSF cytology or flow cytometry is future and hopefully improve outcomes for these patients.
54 Leptomeningeal Carcinomatosis 581

Table 54.1 Recently completed and ongoing prospective studies for patients with Leptomeningeal carcinomatosis (IT Chemo: Intrathecal che-
motherapy, IFRT: Involved field radiotherapy, TKI: Tyrosine kinase inhibitor, ICI: Immune checkpoint inhibitor, mAbs: Monoclonal antibodies,
IV Chemo: Intravenous chemotherapy)
Treatment
Phase Trial identifier modality Study title Status
I NCT03101579 IT chemo Intrathecal Pemetrexed for recurrent Leptomeningeal metastases from non-small cell Completed
lung cancer
II NCT03082144 RT + IT chemo Concurrent involved-field radiotherapy and Intrathecal Chemotherapy for Completed
Leptomeningeal metastases from solid tumors
II NCT02616393 TKI Study of Tesevatinib in subjects with non-small cell lung cancer, EGFR activating Completed
mutation, prior treatment with a tyrosine kinase inhibitor, and brain metastases or
Leptomeningeal metastases
III NCT01645839 IT chemo Interest of Intrathecal Chemotherapy with liposomal Cytarabine (DepoCyte®) in Completed
meningeal metastasis of breast cancer (DEPOSEIN)
I NCT03661424 Cellular therapy Anti-CD3 x anti-Her2/Neu (Her2Bi) armed activated T cells (ATC) in patients with Recruiting
breast cancer Leptomeningeal metastases
I NCT03719768 ICI + RT Avelumab with radiotherapy in patients with Leptomeningeal disease Recruiting
I NCT03025256 IV + IT ICI Intravenous and Intrathecal Nivolumab in treating patients with Leptomeningeal Recruiting
disease
II NCT02886585 ICI Pembrolizumab in central nervous system metastases Recruiting
II NCT04233021 TKI Study of Osimertinib in patients with a lung cancer with brain or Leptomeningeal Recruiting
metastases with EGFR mutation (ORBITAL)
II NCT04425681 TKI + mAbs Osimertinib with Bevacizumab for Leptomeningeal metastasis from EGFR-mutation Recruiting
non-small cell lung cancer (OWBLM)
II NCT04343573 RT Proton Craniospinal radiation therapy vs. partial photon radiation therapy for Recruiting
Leptomeningeal metastasis from solid tumors
II NCT04420598 Antibody drug A multicenter, open-label, single-arm, multicohort phase II clinical trial of Recruiting
conjugate Trastuzumab Deruxtecan (DS-8201a) in human epidermal growth factor receptor 2
(HER2)- positive advanced breast cancer with brain metastases and/or
Leptomeningeal Carcinomatosis - the DEBBRAH study
II NCT03501979 TKI + mAbs + IV Tucatinib and Trastuzumab and Capecitabine for treatment of Leptomeningeal Recruiting
chemo metastases in HER2 positive breast cancer
II NCT02422641 IV chemo Prospective evaluation of high-dose systemic methotrexate in patients with breast Recruiting
cancer and Leptomeningeal metastasis
II/III NCT03275402 Radio-­ Intracerebroventricular Radioimmunotherapy using 131I-omburtamab for Recruiting
immunotherapy Neuroblastoma central nervous system/Leptomeningeal metastases

References
Future Direction
• Broaden the understanding of pathophysiology 1. Chamberlain MC. Leptomeningeal metastases: a review of evalua-
underlying development of LC in cancer patients. tion and treatment. J Neuro-Oncol. 1998;37:271–84. https://doi.org
• Further validation of CSF circulating tumor cells /10.1023/a:1005976926058.
and DNA as a diagnostic tool. 2. Gwak H-S, Joo J, Kim S, et al. Analysis of treatment outcomes of
intraventricular chemotherapy in 105 patients for leptomeningeal
• Application of next-generation sequencing to CSF carcinomatosis from non-small-cell lung cancer. J Thorac Oncol.
circulating tumor cells and DNA in order to identify 2013;8:599–605.
therapeutic targets. 3. Wasserstrom WR, Glass JP, Posner JB. Diagnosis and treatment of
• Standardization of response assessment tools leptomeningeal metastases from solid tumors: experience with 90
patients. Cancer. 1982;49:759–72. https://doi.org/10.1002/1097-­
and criteria across clinical trials and practice 0142(19820215)49:4<759::aid-­cncr2820490427>3.0.co;2-­7.
settings. 4. Beauchesne P. Intrathecal chemotherapy for treatment of lepto-
• Development of phase III clinical trials which meningeal dissemination of metastatic tumours. Lancet Oncol.
include patients with LC, since this continues to 2010;11:871–9. https://doi.org/10.1016/S1470-­2045(10)70034-­6.
5. Zhang XH-F, Giuliano M, Trivedi MV, et al. Metastasis dor-
be an exclusion criterion in most prospective mancy in estrogen receptor-positive breast cancer. Clin Cancer
studies. Res. 2013;19:6389–97. https://doi.org/10.1158/1078-­0432.CCR-­
13-­0838.
582 S. R. K. Singh et al.

6. Kokkoris CP. Leptomeningeal carcinomatosis. how does cancer Clin Cancer Res. 2017;23:5480–8. https://doi.org/10.1158/1078-
reach the pia-arachnoid? Cancer. 1983;51:154–60. https://doi. ­0432.CCR-­17-­0047.
org/10.1002/1097-­0142(19830101)51:1<154::aid-­cncr282051013 24. Murtaza M, Dawson S-J, Pogrebniak K, et al. Multifocal clonal
0>3.0.co;2-­k. evolution characterized using circulating tumour DNA in a case of
7. Boire A, Zou Y, Shieh J, et al. Complement component 3 metastatic breast cancer. Nat Commun. 2015;6:8760. https://doi.
adapts the cerebrospinal fluid for Leptomeningeal metasta- org/10.1038/ncomms9760.
sis. Cell. 2017;168:1101–113.e13. https://doi.org/10.1016/j. 25. Zhao Y, He J-Y, Zou Y-L, et al. Evaluating the cerebrospinal fluid
cell.2017.02.025. ctDNA detection by next-generation sequencing in the diagnosis of
8. Wang N, Bertalan MS, Brastianos PK. Leptomeningeal metastasis meningeal Carcinomatosis. BMC Neurol. 2019;19:331. https://doi.
from systemic cancer: review and update on management. Cancer. org/10.1186/s12883-­019-­1554-­5.
2018;124:21–35. https://doi.org/10.1002/cncr.30911. 26. Niwińska A, Pogoda K, Michalski W, et al. Determinants of pro-
9. Le Rhun E, Weller M, Brandsma D, et al. EANO-ESMO clini- longed survival for breast cancer patient groups with leptomenin-
cal practice guidelines for diagnosis, treatment and follow-up of geal metastasis (LM). J Neuro-Oncol. 2018;138:191–8. https://doi.
patients with leptomeningeal metastasis from solid tumours. Ann org/10.1007/s11060-­018-­2790-­z.
Oncol. 2017c;28:iv84–99. https://doi.org/10.1093/annonc/mdx221. 27. Ferguson SD, Bindal S, Bassett RLJ, et al. Predictors of survival in
10. Leal T, Chang JE, Mehta M, Robins HI. Leptomeningeal metasta- metastatic melanoma patients with leptomeningeal disease (LMD).
sis: challenges in diagnosis and treatment. Curr Cancer Ther Rev. J Neuro-Oncol. 2019;142:499–509. https://doi.org/10.1007/
2011;7:319–27. https://doi.org/10.2174/157339411797642597. s11060-­019-­03121-­2.
11. Le Rhun E, Taillibert S, Zairi F, et al. A retrospective case series of 28. Yan W, Yu J. The clinical characteristic and prognostic factors of
103 consecutive patients with leptomeningeal metastasis and breast leptomeningeal metastasis in patients with non-small-cell lung can-
cancer. J Neuro-Oncol. 2013b;113:83–92. https://doi.org/10.1007/ cer: a retrospective study from one single cancer institute. J Clin
s11060-­013-­1092-­8. Oncol. 2019;37:e13502. https://doi.org/10.1200/JCO.2019.37.15_
12. Le Rhun E, Taillibert S, Chamberlain MC. Carcinomatous menin- suppl.e13502.
gitis: leptomeningeal metastases in solid tumors. Surg Neurol Int. 29. Tremont-Lukats IW, Ratilal BO, Armstrong T, Gilbert MR. Anti-
2013a;4:S265–88. https://doi.org/10.4103/2152-­7806.111304. epileptic drugs for preventing seizures in people with brain
13. Franzoi MA, Hortobagyi GN. Leptomeningeal carcinomatosis in tumors. Cochrane Database Syst Rev. 2008:CD004424. https://doi.
patients with breast cancer. Crit Rev Oncol Hematol. 2019;135:85– org/10.1002/14651858.CD004424.pub2.
94. https://doi.org/10.1016/j.critrevonc.2019.01.020. 30. Nigim F, Critchlow JF, Kasper EM. Role of ventriculoperitoneal
14. Chamberlain M, Junck L, Brandsma D, et al. Leptomeningeal shunting in patients with neoplasms of the central nervous system:
metastases: a RANO proposal for response criteria. Neuro-Oncol- an analysis of 59 cases. Mol Clin Oncol. 2015;3:1381–6. https://
ogy. 2017;19:484–92. https://doi.org/10.1093/neuonc/now183. doi.org/10.3892/mco.2015.627.
15. Collie DA, Brush JP, Lammie GA, et al. Imaging features of lep- 31. Morris PG, Reiner AS, Szenberg OR, et al. Leptomeningeal metas-
tomeningeal metastases. Clin Radiol. 1999;54:765–71. https://doi. tasis from non-small cell lung cancer: survival and the impact of
org/10.1016/S0009-­9260(99)91181-­9. whole brain radiotherapy. J Thorac Oncol. 2012;7:382–5. https://
16. Short RG, Bal S, German JP, et al. Potential of F-18 PET/CT doi.org/10.1097/JTO.0b013e3182398e4f.
in the detection of Leptomeningeal metastasis. Neuroradiol J. 32. Le Rhun E, Taillibert S, Chamberlain MC. Neoplastic meningitis
2014;27:685–9. https://doi.org/10.15274/NRJ-­2014-­10098. due to lung, breast, and melanoma metastases. Cancer Control.
17. Ortapamuk H, Demir MK. Intracranial leptomeningeal carcino- 2017b;24:22–32. https://doi.org/10.1177/107327481702400104.
matosis in three cases from breast cancer demonstrated on F-18 33. Jacus MO, Daryani VM, Harstead KE, et al. Pharmacokinetic
Fluorodeoxyglucose positron emission tomography/computer- properties of anticancer agents for the treatment of central
ized tomography. Indian J Nucl Med. 2017;32:16–8. https://doi. nervous system tumors: update of the literature. Clin Pharma-
org/10.4103/0972-­3919.198454. cokinet. 2016;55:297–311. https://doi.org/10.1007/s40262-­015-­
18. Lee JS, Melisko ME, Magbanua MJM, et al. Detection of cere- 0319-­6.
brospinal fluid tumor cells and its clinical relevance in lepto- 34. Bowman KM, Kumthekar P. Medical management of brain metas-
meningeal metastasis of breast cancer. Breast Cancer Res Treat. tases and leptomeningeal disease in patients with breast carcinoma.
2015;154:339–49. https://doi.org/10.1007/s10549-­015-­3610-­1. Future Oncol. 2018;14:391–407. https://doi.org/10.2217/fon-­
19. Glantz MJ, Cole BF, Glantz LK, et al. Cerebrospinal fluid cytol- 2017-­0255.
ogy in patients with cancer: minimizing false-­ negative results. 35. Le Rhun E, Rudà R, Devos P, et al. Diagnosis and treatment pat-
Cancer. 1998;82:733–9. https://doi.org/10.1002/(sici)1097-­ terns for patients with leptomeningeal metastasis from solid tumors
0142(19980215)82:4<733::aid-­cncr17>3.0.co;2-­z. across Europe. J Neuro-Oncol. 2017a;133:419–27. https://doi.
20. Ahluwalia MS, Wallace PK, Peereboom DM. Flow cytometry as a org/10.1007/s11060-­017-­2452-­6.
diagnostic tool in lymphomatous or leukemic meningitis: ready for 36. Scott BJ, Oberheim-Bush NA, Kesari S. Leptomeningeal metasta-
prime time? Cancer. 2012;118:1747–53. https://doi.org/10.1002/ sis in breast cancer - a systematic review. Oncotarget. 2016;7:3740–
cncr.26335. 7. https://doi.org/10.18632/oncotarget.5911.
21. Glitza IC, Smalley KSM, Brastianos PK, et al. Leptomeningeal dis- 37. Shapiro WR, Young DF, Mehta BM. Methotrexate: distribution in
ease in melanoma patients: an update to treatment, challenges, and cerebrospinal fluid after intravenous, ventricular and lumbar injec-
future directions. Pigment Cell Melanoma Res. 2020;33:527–41. tions. N Engl J Med. 1975;293:161–6. https://doi.org/10.1056/
https://doi.org/10.1111/pcmr.12861. NEJM197507242930402.
22. Li X, Zhang Y, Ding J, et al. Clinical significance of detecting CSF-­ 38. Tetef ML, Margolin KA, Doroshow JH, et al. Pharmacokinetics
derived tumor cells in breast cancer patients with leptomeningeal and toxicity of high-dose intravenous methotrexate in the treatment
metastasis. Oncotarget. 2018;9:2705–14. https://doi.org/10.18632/ of leptomeningeal carcinomatosis. Cancer Chemother Pharmacol.
oncotarget.23597. 2000;46:19–26. https://doi.org/10.1007/s002800000118.
23. Jiang B-Y, Li Y-S, Guo W-B, et al. Detection of driver and resis- 39. Poplack D, Bleyer WA HM (1980) Pharmacology of antineo- plas-
tance mutations in Leptomeningeal metastases of NSCLC by next-­ tic agents in cerebrospinal fluid. https://doi.org/10.1007/978-1-
generation sequencing of cerebrospinal fluid circulating tumor cells. 4684-1039-6_39.
54 Leptomeningeal Carcinomatosis 583

40. Taillibert S, Laigle-Donadey F, Chodkiewicz C, et al. Leptomenin- 47. Mead PA, Safdieh JE, Nizza P, et al. Ommaya reservoir infections:
geal metastases from solid malignancy: a review. J Neuro-Oncol. a 16-year retrospective analysis. J Infect. 2014;68:225–30. https://
2005;75:85–99. https://doi.org/10.1007/s11060-­004-­8101-­x. doi.org/10.1016/j.jinf.2013.11.014.
41. Boogerd W, Van Den Bent MJ, Koehler PJ, et al. The relevance 48. Le Rhun E, Preusser M, van den Bent M, et al. How we treat patients
of intraventricular chemotherapy for leptomeningeal metastasis in with leptomeningeal metastases. ESMO Open. 2019;4:e000507.
breast cancer: a randomised study. Eur J Cancer. 2004;40:2726–33. https://doi.org/10.1136/esmoopen-­2019-­000507.
https://doi.org/10.1016/j.ejca.2004.08.012. 49. Lin N, Dunn IF, Glantz M, et al. Benefit of ventriculoperitoneal
42. Glantz MJ, Jaeckle KA, Chamberlain MC, et al. A random- cerebrospinal fluid shunting and intrathecal chemotherapy in neo-
ized controlled trial comparing intrathecal sustained-release plastic meningitis: a retrospective, case-controlled study. J Neuro-
cytarabine (DepoCyt) to intrathecal methotrexate in patients surg. 2011;115:730–6. https://doi.org/10.3171/2011.5.JNS101768.
with neoplastic meningitis from solid tumors. Clin Cancer Res. 50. Glitza IC, Rohlfs M, Guha-Thakurta N, et al. Retrospective review of
1999;5:3394–402. metastatic melanoma patients with leptomeningeal disease treated
43. Grossman SA, Finkelstein DM, Ruckdeschel JC, et al. Random- with intrathecal interleukin-2. ESMO Open. 2018;3:e000283.
ized prospective comparison of intraventricular methotrexate and https://doi.org/10.1136/esmoopen-­2017-­000283.
thiotepa in patients with previously untreated neoplastic men- 51. Hollebecque A, Levy A, Broutin S, et al. First case report of intra-
ingitis. J Clin Oncol. 1993;11:561–9. https://doi.org/10.1200/ thecal panitumumab for treatment of meningeal carcinomatousis
JCO.1993.11.3.561. in an EGFR mutant lung adenocarcinoma patient. Lung Cancer.
44. Hitchins RN, Bell DR, Woods RL, Levi JA. A prospective random- 2013;80:113–4.
ized trial of single-agent versus combination chemotherapy in men- 52. Zagouri F, Sergentanis TN, Bartsch R, et al. Intrathecal administra-
ingeal carcinomatosis. J Clin Oncol. 1987;5:1655–62. https://doi. tion of trastuzumab for the treatment of meningeal carcinomatosis
org/10.1200/JCO.1987.5.10.1655. in HER2-positive metastatic breast cancer: a systematic review and
45. Le Rhun E, Mailliez A, Wallet J, et al. CMET-39. Intra-csf liposo- pooled analysis. Breast Cancer Res Treat. 2013;139:13–22. https://
mal cytarabine plus systemic therapy as initial treatment of breast doi.org/10.1007/s10549-­013-­2525-­y.
cancer leptomeningeal metastasis: a randomised, open-label trial. 53. de Visser BW, Somers R, Nooyen WH, et al. Intraventricular
Neuro-Oncology. 2018;20:vi61-2. methotrexate therapy of leptomeningeal metastasis from breast
46. Shapiro WR, Schmid M, Glantz M, Miller JJ. A randomized carcinoma. Neurology. 1983;33:1565–72. https://doi.org/10.1212/
phase III/IV study to determine benefit and safety of cytarabine wnl.33.12.1565.
liposome injection for treatment of neoplastic meningitis. J Clin 54. Freilich RJ, Krol G, DeAngelis LM. Neuroimaging and cerebrospi-
Oncol. 2006;24:1528. https://doi.org/10.1200/jco.2006.24.18_ nal fluid cytology in the diagnosis of leptomeningeal metastasis. Ann
suppl.1528. Neurol. 1995;38:51–7. https://doi.org/10.1002/ana.410380111.
 Part XIV
Urologic Cancer and First Patterns of Metastasis

Samuel L. Washington III, Peter R. Carroll

and Sima P. Porten

In this section of Urologic Cancer and First Patterns of Metastasis, we focus the discussion on
two of the most common, and distinct, genitourinary malignancies, prostate cancer and bladder
cancer. These two cancers impact organs within close anatomic proximity, and thus share simi-
lar patterns of lymphatic drainage. For prostate cancer, we discuss the anatomic considerations
surrounding lymphatic drainage and prior efforts to map the lymphatic system of the prostate.
Next, we describe the correlation between our anatomic understanding and clinical practice.
For bladder cancer we examine the lymphatic drainage and clinical staging at the time of diag-
nosis. Next, we outline the evaluation for nodal disease and its implications for patients diag-
nosed with bladder cancer. The comparison of these two genitourinary malignancies affords us
the opportunity to gain a greater understanding of how the clinical implications differ despite
a similar pattern of lymphatic drainage and metastatic spread.
 Urologic Cancer and the First Patterns
of Metastasis 55
Samuel L. Washington III, Peter R. Carroll,
and Sima P. Porten

Abstract 55.2 Prostate Cancer

In this chapter we will describe the patterns of metastasis
for two of the most common genitourinary malignancies, Prostate cancer will account for an estimated 248,530 new
prostate cancer and bladder cancer. We will also discuss cancer diagnoses in 2021 and remains the second most com-
aspects of management for individuals with node-positive mon cause of cancer deaths in men in the United States [1].
disease to provide insight into the clinical impact of Although the majority will present with low-grade disease,
metastasis on oncologic outcomes. a subset of patients will present with more aggressive dis-
ease. The key principle of timely evaluation and diagnosis
of prostate cancer is to identify men with early stage disease
who harbor aggressive disease and are at risk of metastasis.
Learning Objectives Here we describe prior efforts to understand the pathways in
• To understand the lymphatic drainage of prostate which aggressive disease travels from the prostate cancer to
cancer and bladder cancer. regional lymph nodes and distant sites of metastasis.
• To appreciate the patterns of metastatic spread for
prostate cancer and bladder cancer.
• To review the current discussion surrounding the 55.3 Anatomic Considerations
surgical management of lymph node metastasis for for Lymphatic Drainage
prostate cancer and bladder cancer.
The lymphatic drainage of the prostate is thought to mir-
ror the vascular drainage system, with a similarly complex
system of the main and smaller branches. Numerous studies
55.1 Introduction have gradually improved our understanding of the lymphatic
drainage of the prostate and patterns of metastatic spread in
In this chapter we will review the patterns of metastatic men with prostate cancer [2, 3]. As a result, we now under-
spread for two of the most common genitourinary malignan- stand that the lymphatic drainage largely depends on the
cies, prostate cancer and bladder cancer. These cancers rep- location within the prostate and the tissue types present.
resent distinct histologic tumors in close anatomic proximity One such study examined five fresh prostate specimens
to one another within the pelvis. As a result, they share simi- from patients undergoing cystoprostatectomy for bladder
lar patterns of lymphatic drainage. cancer [2]. Specimens were sectioned and incubated with
a monoclonal antibody specific for desmoplakin, which
detects the endothelial layer of lymphatic channels. Five
fields within the central, transitional, and peripheral zones
were randomly selected to calculate the density of lymphat-
ics per region. Within the glandular prostate, no lymphatics
were in close proximity to the glandular acini or intra-acinar
connective tissue. A consistent pattern of lymphatic in the
S. L. Washington III (*) · P. R. Carroll · S. P. Porten
University of California, San Francisco, CA, USA interacinar connective tissue septum was noted in each of
e-mail: [email protected] the three prostatic zones. Some lymphatics ran parallel to the

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 587
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_55
588 S. L. Washington III et al.

prostatic capsule while other accompanied arteries or neu- nal iliac, and obturator arteries, which make up the standard
ral ganglia. Lymphatics were present between smooth mus- template for lymph node dissection.
cle bundles of the anterior fibromuscular stroma but rarely Leveraging prior efforts to map the anatomic pathways
detected in the fibrous stoma itself. The greatest density of of lymphatic drainage, the utility of a sentinel lymph dissec-
lymphatics ran parallel and in the longitudinal direction of tion remains a topic of ongoing investigation. The concept
the ejaculatory ducts compared to other areas within the of a sentinel lymph node rests on the idea of there being a
glandular prostate. The prostatic capsule harbored a slightly single lymph node in the direct drainage pathway from the
greater, yet not statistically significant different, density of organ of origin to the rest of the body [7]. Conceptually,
lymphatics compared to the glandular prostate itself. a sentinel lymph node would provide an accurate assess-
Broadly, larger lymphatic vessels within the gland drain ment of the likelihood of distant nodal metastasis and help
into a subcapsular network, which in turn drains a system justify whether a pelvic lymph node dissection should (or
of collecting vessels often pairing with arteries [3]. These should not) be performed. Yet this is not entirely applicable
lymphatic vessel networks are concentrated at the base of to prostate cancer, as we know the lymphatic drainage var-
the gland, where the highest density of vessels is present. ies based on the location within the gland. In addition, the
The lymphatic channels are thought to also drain the upper intrapelvic location of the prostate and relevant lymphatic
aspect of the posterior prostate, extending along the medial drainage makes it difficult to accurately visualize the poten-
aspect of the seminal vessels and extending towards the ure- tially affected lymph nodes [8]. Efforts to incorporate vari-
ters, then cross the umbilical artery, and travel laterally to the ous imaging modalities and lymph node dissection strategies
external iliac lymph nodes. The lymphatic channels drain- continue, but sentinel lymph node dissection remains largely
ing the inferior aspect of the prostate travel to the base of experimental.
the gland and then to the hypogastric lymph nodes. Some of
these channels communicate with those of the rectum, drain-
ing into the pararectal lymph nodes. The anterior prostate 55.5 Clinical Correlation with Anatomic
has lymphatic channels that run along the apex of the gland Understanding
and follow the internal pudendal artery, eventually joining
the internal iliac lymph nodes. Additional anterior and lateral Historically, surgical resection of the lymph nodes was uti-
lymphatics drain to the para-vesical lymph nodes, eventually lized to detect spread of cancer beyond the organ of origin.
draining into the bladder lymphatic system. Prior work leveraged autopsy as a primary means of iden-
tified patterns of spread in men with advanced disease. By
noting the landing sites of metastasis, we were able to gain
55.4 Mapping the Lymphatic System a greater understanding of the spread. Interestingly, autopsy
of the Prostate studies noted that nodal yield of the surgical dissection often
underrepresented the total lymph node burden of extrapelvic
Previous efforts have attempted to map the lymphatic sys- disease noted at autopsy [9]. Lymph node dissection eventu-
tem of the prostate using various methodologies. Prior stud- ally became a feasible method of sampling, balanced with
ies have injected contrast into human glands to patterns of the risk of morbidity, rather than a thorough examination of
lymphatic spread and have outlined the drainage along larger the potential full extent of disease. The extent of the dissec-
vessels, but these methods may have limited characteriza- tion evolved over time as surgical approach evolved from a
tion of smaller vessels that were unable to accommodate perineal to retropubic approach. This change in approach in
the contrast molecule and/or were too small to be visualized turn led to greater lymph node yield and nodal positivity. The
adequately on imaging. Other studies have identified three observed examination of nodal burden increased as lymph
main patterns of spread: superior and lateral to the internal-­ node dissections became more extensive, expanding beyond
external iliac lymph nodes; from perineal floor to internal the obturator and external iliac lymph nodes to include areas
pudendal vessels; and via sacral lymphatics (S2–5) [4, 5]. such as the internal iliac, common iliac, presacral, and pre-
Others have used radiotracers such as technetium to examine sciatic regions.
lymphatic spread based on the concept of lymphatic spread Autopsy study of 753 men with metastatic disease dem-
occurring through key sentinel lymph nodes, similar to lym- onstrated that 63% had a lymph node metastasis [10]. They
phatic spread for breast cancer [6]. This study was limited by noted that even in cases of metastasis confined to the lymph
the properties of the radiotracer, both in terms of time needed nodes, pelvic nodal involvement was not more frequent than
for dispersion and size limitations with respect to the lym- paraaortic nodal involvement. From this study, two patterns
phatic vessels it could travel through. The end result of such of metastatic spread were observed: lymphogenous spread
studies was the confirmation that the most common sites of to paraaortic lymph node metastasis with pelvic lymph
involvement were generally along the external iliac, inter- node involvement and hematogenous spread to paraaor-
55 Urologic Cancer and the First Patterns of Metastasis 589

tic lymph nodes without pelvic lymph node involvement. nodal disease at time of primary radical prostatectomy, up
Cases involving metastasis to a single organ were noted to to 29% may have an undetectable prostate-specific antigen
frequently harbor metastasis to the lumbar spine and para- level with minimal differences in survival among those that
aortic lymph nodes, largely thought to occur by way of the received adjuvant therapy or not. Those with a detectable
vertebral vein, as a method of skipping pelvic lymphatic prostate-specific antigen level, irrespective of nodal status,
drainage channels [11]. were noted to have worse recurrence-free survival [16]. An
Another autopsy study of 1589 men with prostate can- observational study of 1249 men with high-risk prostate
cer similarly noted lymphatic or hematogenous metastases cancer demonstrated that a greater lymph node burden at
in 39.7% of men [12]. Paraaortic lymph nodes were most time of surgery was associated with worse cancer-specific
frequently noted, followed by pelvic lymph nodes. A hema- survival [17]. In this setting, a pelvic lymph node dissec-
togenous metastatic pattern was the most common overall tion provides diagnostic value by providing an estimate of
(35%), with frequent sites being the bone (spine, 90%), the nodal burden at time of treatment. A systemic review
lung (46%), liver (25%), pleura (21%), and adrenals (13%). examined the risk-benefit ratio of different extents of lymph
Hematogenous metastasis was more frequent in those with node dissection, finding that the extent of lymph node dissec-
paraaortic lymph node involvement than in those with pel- tion impacted the rates of biochemical recurrence and clini-
vic lymph node involvement (88.9% vs. 63.9%, p < 0.001). cal recurrence with no significant impact on cancer survival
Spine metastasis was significantly associated with para- outcomes [18]. In addition, more extensive dissections were
aortic lymph node involvement, likely through a hematog- associated with greater perioperative risks of blood loss,
enous metastasis pattern. In this study, histologic grade was extended hospital stay, and postoperative complications. The
strongly associated with hematogenous metastatic spread, findings of this systematic review question the therapeutic
ranging from 7.5% prevalence in low-grade tumors to value of the lymph node dissection while highlighting the
60.7% in high-grade tumors. A similar pattern was noted diagnostic benefit.
with pathologic stage, with metastases found in 4.2% of Most recently, a report from a prospective single-center
men with pT2 cancers, 41.1% of pT3, and 80.3% of pT4 phase III randomized control trial sought to determine the
tumors (p < 0.001). Lymph node positivity is rarely associ- impact of an extended lymph node dissection on oncologic
ated with anteriorly located tumors, and 30–40% of lymph outcomes for men with intermediate- or high-risk clinically
node metastases are found to be contralateral to the domi- localized prostate cancer [19]. A total of 300 men were ran-
nant tumor [13]. domized to a limited or extended pelvic lymph node dissec-
Using a surgical approach to map nodal metastasis in tion. The median biochemical recurrence-free survival was
high-risk prostate cancer, 19 men underwent an extended 6.14 months in men who underwent a limited node dissec-
lymph node dissection and retroperitoneal lymph node dis- tion and had not yet been reached in those with an extended
section at the time of radical prostatectomy [14]. In this study approach. Yet median metastasis-free survival was not
the most common sites of nodal involvement were the obtu- reached in either group nor was cancer-specific survival data
rator (88%), external iliac (83.3%), common iliac (77%), available because no patient had died due to prostate can-
hypogastric (44.4%), and presacral (33.3%). All cases with cer by the cutoff date. Ultimately, the authors concluded that
common iliac lymph node involvement had positive pelvic based on early oncologic outcomes reporting, a more exten-
lymph nodes, and all cases with retroperitoneal disease had sive lymph node dissection offers better pathologic staging
positive common iliac lymph nodes. Positive retroperitoneal without demonstrating a clear benefit in oncologic outcomes.
lymph nodes were found only in cases of positive common
iliac lymph nodes.
Given the under-appreciation of the total nodal metastasis 55.6 Bladder Cancer
burden using a surgical approach, the role and extent of a pel-
vic lymph node dissection at time of radical prostatectomy Bladder cancer is the fourth most common cancer in the
continues to be debated. Discussions are commonly centered United States in men (less common in women) accounting
around appropriate patient selection, the extent of dissec- for about 83,730 new cases (64,280 in men and 19,450 in
tion, the morbidity of the procedure, and the diagnostic ver- women) and results in 17,200 deaths (12,260 in men and
sus therapeutic value. Utilization of pelvic node dissection 4940 in women) in 2020 [1]. The primary tumor stage is
has varied across the United States, with rates ranging from vital to patient outcomes, and indeed 70% of patients pres-
58.9% to 72.1% over time and with greater use in those of ent with non-muscle invasive bladder cancer (Tis, Ta, and
intermediate- and high-risk disease compared to those with T1) [1]. However, 20–40% of patients either present with or
low-risk features [15]. The diagnostic and therapeutic value progress to muscle-invasive cancer (T2–T4), which in turn
of a lymph node dissection is also intrinsically related to the has an increased risk of development of metastatic disease
prevalence of nodal disease in the population. For men with [20]. The primary pattern of metastasis in bladder cancer is
590 S. L. Washington III et al.

through lymphovascular spread to locoregional lymph nodes the pelvis is considered involved if it is ≥8 mm in short
before developing disease to distant sites (most commonly axis, and other features such roundness and necrosis
bone and lung) [21]. Clinically, these patients have a mark- can be considered in determining lymph node positivity.
edly worse prognosis with 5-year overall and disease-free Unfortunately, the sensitivity for detecting lymph node
survival rates of 35% and 31% compared to 89% and 69% involvement in bladder cancer is limited at 30–75%, with
for patients without lymph node involvement [22]. The pres- an overall accuracy of 61% [27, 28]. CT is limited by an
ence of nodal metastasis (whether occult micro-metastasis inability to detect micro-­ metastatic disease in normal-
or overt disease) is one of the most significant independent sized lymph nodes. Magnetic resonance imaging (MRI) is
prognostic factors of mortality [23]. a potential alternative to CT; however, it has similar limita-
tions in detecting the presence of disease in a normal-sized
lymph node. Despite the utilization of multiparametric
55.7 Lymphatic Drainage of the Bladder techniques, which have been revolutionary in character-
and Clinical Staging izing prostate cancer (T1/T2 signal intensity, diffusion
characteristics, and timing of enhancement of gadolinium-
A recent study by Roth et al. prospectively mapped the based contrast), pooled sensitivity was only between 56
primary landing sites (first area) for lymphatic drainage in and 86% [29, 30].
patients with muscle-invasive bladder cancer with 80% of Attempts to improve sensitivity and overall accuracy for
patients having regional lymph node drainage to the obtura- imaging have not proved as fruitful as other advances in
tor and internal iliac areas and external iliac region (bilater- imaging such as PSMA-PET for prostate cancer. Although
ally) [24]. Interestingly, some patients had primary landing metastatic lymph nodes have increased metabolic activity
sites (first area of drainage) to more distant lymph nodes and radiotracer uptake on PET/CT, spatial resolution limits
such as the common iliac (15%) or paraaortic (4%) areas, detection in normal-sized lymph nodes and has the added
and the primary location of the tumor can impact lymphatic downside of false-positive findings in the setting of inflam-
spread [24, 25]. mation [31]. Meta-analyses report low and varying sensi-
Similar to other cancers, clinical staging of nodal metas- tivities of 33–70% [32, 33]. Alternative radiotracers such as
tases is based on TNM guidelines set by the American Joint C11-choline have also been assessed with similarly low sen-
Committee on Cancer and reflects anatomical lymphatic sitivity of 66% in identifying nodal disease [34]. For MRI,
spread [26]. Nodal involvement is denoted as N0-M1a the use of intravenous lymphotropic ultrasmall superpara-
­disease with N0 disease defined as no lymph node involve- magnetic particles of iron oxide (USPIO) to identify benign
ment. N1 disease is defined as one regional pelvic lymph lymph nodes (where there is uptake), and thus uncover
node identified in the obturator, internal iliac, external iliac, normal-­sized lymph nodes harboring disease, improves sen-
perivesical, or presacral chain, and N2 disease is one or more sitivity to 75% in a small prospective trial [35]. Currently,
of regional lymph nodes involved. N3 disease is a positive USPIO is not approved for clinical use in many countries and
lymph node at either or both common iliac chains. Nodal often requires multiple exams.
metastases outside the true pelvis, which are located above Unlike other malignancies, sentinel lymph node biopsy
the common iliacs, are defined as true metastatic disease has not proven a useful tool in staging for patients with blad-
(M1a). Identification of nodal disease is based upon histo- der cancer nor for treatment decision-making due to a high
pathology (biopsy or excision) or unequivocal findings on false-negative rate. A meta-analysis including 336 patients
cross-sectional imaging, and the size of nodal involvement is was performed, and pooled sensitivity was 79% (95% CI
not a part of the current clinical staging system. 0.69–0.86%) [36]. Recently, Salehi et al. published a con-
secutive series of 41 patients with sentinel nodes detected in
85% [35] at time of cystectomy, regardless of prior systemic
55.8 Detection of Nodal Disease treatment [37]. However, although sentinel lymph node
and Clinical Implications biopsy is feasible, to make decision about proceeding with
standard lymphadenectomy, pathologic status for sentinel
At presentation with muscle-invasive disease, patients will node had to be performed for both sides of the hemipelvis
undergo cross-sectional imaging to determine if distant as contralateral involvement was common. Hence, to forgo
metastasis is present, which in turn guides primary treat- a potentially curative lymphadenectomy based on sentinel
ment and estimates of prognosis. CT (computed tomogra- lymph node biopsy is not aligned with current standard of
phy) is used most commonly and is performed with and care.
without intravenous contrast. Although distant metastasis The current cornerstone of treatment for muscle-invasive
can be detected robustly, characterizing lymph node-posi- bladder cancer is radical cystectomy (complete bladder
tive disease on imaging is more elusive. A lymph node in removal with the prostate in men and potentially uterus/cer-
55 Urologic Cancer and the First Patterns of Metastasis 591

vix/ovaries in women) with pelvic node dissection. Lymph

node removal is thought to be both diagnostic and therapeu- Open Questions
tic (for some). Although no single template is “standard,” • How can we improve diagnostic accuracy of imag-
extrapolation (based on the abovementioned anatomic stud- ing for patients with bladder cancer?
ies and staging principles) suggests that boundaries of an • Are there subsets of patients with prostate cancer
adequate lymph node dissection should be the genitofemoral for whom a pelvic lymph node dissection might
nerve laterally, the bladder wall medially, the inguinal liga- confer a therapeutic benefit?
ment distally/node of Cloquet, and the bifurcation of the • How can we advance imaging modalities to become
common iliac artery proximally [38]. Multiple retrospec- accurate noninvasive assessments of nodal disease
tive studies from institutional and national registries have to replace pelvic lymph node dissection on patients
reported that lymph node status (pathologic staging) along with prostate cancer?
with other quantitative measurements including lymph node • Are there histologic factors that harbor a greater
yield (number of lymph nodes taken), lymph node involve- risk of early distant metastasis in patient with blad-
ment (number of positive nodes present), and lymph node der cancer?
density (number of nodes positive per number taken) can • Will we develop biomarkers to accurately assess
affect patient outcomes. Hence, the rationale is that if you aggressive disease in prostate and bladder cancer?
remove a larger number of nodes (even if no nodes are • Should sentinel lymph node biopsy be performed
positive), and the number of nodes positive is significantly for prostate and bladder cancer?
less than the number of nodes taken, patient outcomes are
improved significantly by eradicating micro-metastatic dis-
ease [39–42]. Extended lymph node dissection (includes References
presacral lymph nodes and those up to the aortic bifurcation)
has been advocated with improvements in survival up to 25% 1. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2021.
[43, 44]. Recently, a large multicenter randomized phase III CA: Cancer J Clin. 2021;71(4):359. First Published online: April
19, 2021. https://doi.org/10.3322/caac.21654.
study evaluating the extent of pelvic lymph node dissec- 2. Fukuda H, Yamada T, Kamata S, Saitoh H. Anatomic distribution
tion demonstrated that extended lymphadenectomy was not of intraprostatic lymphatics: implications for the lymphatic spread
superior to a standard template based on recurrence-free of prostate cancer – a preliminary study. Prostate. 2000;44:322–7.
survival (hazard ratio [HR], 0.84; P = 0.36), cancer-specific https://doi.org/10.1002/1097-­0 045(20000901)44:4<322::aid-­
pros9>3.0.co;2-­8.
survival (HR, 0.70; P = 0.10), or overall survival (HR, 0.78; 3. Rouviere H. Anatomy of the human lymphatic system. Ann Arbor.
P = 0.12). However, patients who had an extended node dis- 1939;27:194–5. https://doi.org/10.1002/bjs.18002710531.
section had a higher complication rate, primarily with lym- 4. Raghavaiah NV, Jordan WP. Prostatic lymphography. J Urol.
phoceles requiring intervention (3.4% vs. 8.6%; P = 0.04) 1979;121:178–81. https://doi.org/10.1016/s0022-­5347(17)56712-­9.
5. Brossner C, Ringhofer H, Hernady T, Kuber W, Madersbacher S,
[45]. Another randomized phase III Southwest Oncology Pycha A. Lymphatic drainage of prostatic transition and periph-
Group (SWOG S1011) trial comparing standard to extended eral zones visualized on a three-dimensional workstation. Urology.
node dissection in patients with cystectomy was recently 2001;57:389–93. https://doi.org/10.1016/s0090-­4295(00)00921-­3.
completed and will provide more robust data on the role of 6. Paganelli G, De Cicco C, Cremonesi M, et al. Optimized sentinel
node scintigraphy in breast cancer. Q J Nucl Med. 1998;42:49–53.
extent of lymph node dissection. It remains to be seen if a 7. Nieweg OE, Tanis PJ, Kroon BB. The definition of a sentinel node.
surgical cure of metastatic disease at the primary landing site Ann Surg Oncol. 2001;8:538–41. https://doi.org/10.1007/s10434-­
(i.e., pelvic lymph nodes) is an attainable goal in patients 001-­0538-­y.
with invasive bladder cancer. Improvements in imaging may 8. Narayanan R, Wilson TG. Sentinel node evaluation in prostate can-
cer. Clin Exp Metastasis. 2018;35:471–85. https://doi.org/10.1007/
help to identify patients where this could be possible, and a s10585-­018-­9936-­4.
deeper understanding of the biology of metastasis in these 9. Swanson GP, Hubbard JK. A better understanding of lymphatic
patients could result in improved biomarkers of aggressive drainage of the prostate with modern imaging and surgical tech-
disease. niques. Clin Genitourin Cancer. 2013;11:431–40. https://doi.
org/10.1016/j.clgc.2013.04.031.
10. Saitoh H, Yoshida K-I, Uchijima Y, Kobayashi N, Suwata J, Kamata
S. Two different lymph node metastatic patterns of a prostatic
55.9 Conclusion cancer. Cancer. 1990;65:1843–6. https://doi.org/10.1002/1097-­
0142(19900415)65:8<1843::aid-­cncr2820650830>3.0.co;2-­4.
11. Batson OV. The function of the vertebral veins and their role in
At the conclusion of this chapter, you should have a greater the spread of metastases. Ann Surg. 1940;112:138–49. https://doi.
understanding of the anatomy of the lymphatic system, pat- org/10.1097/00000658-­194007000-­00016.
terns of metastatic spread, and current discussions surround- 12. Bubendorf L, Schöpfer A, Wagner U, et al. Metastatic patterns of
ing the clinical management of metastatic disease for both prostate cancer: an autopsy study of 1,589 patients. Hum Pathol.
2000;31:578–83. https://doi.org/10.1053/hp.2000.6698.
prostate and bladder cancer.
592 S. L. Washington III et al.

13. Tokuda Y, Carlino LJ, Gopalan A, et al. Prostate cancer topog- 30. Woo S, Suh CH, Kim SY, Cho JY, Kim SH. The diagnostic perfor-
raphy and patterns of lymph node metastasis. Am J Surg Pathol. mance of MRI for detection of lymph node metastasis in bladder
2010;34:1862–7. https://doi.org/10.1097/PAS.0b013e3181fc679e. and prostate cancer: an updated systematic review and diagnos-
14. Briganti A, Suardi N, Capogrosso P, et al. Lymphatic spread of tic meta-analysis. AJR Am J Roentgenol. 2018;210:W95–W109.
nodal metastases in high-risk prostate cancer: the ascending path- https://doi.org/10.2214/AJR.17.18481.
way from the pelvis to the retroperitoneum. Prostate. 2012;72:186– 31. Goodfellow H, Viney Z, Hughes P, et al. Role of fluorodeoxyglucose
92. https://doi.org/10.1002/pros.21420. positron emission tomography (FDG PET)-computed tomography
15. Nocera L, Sood A, Dalela D, et al. Rate and extent of pelvic lymph (CT) in the staging of bladder cancer. BJU Int. 2014;114:389–95.
node dissection in the US prostate cancer patients treated with 32. Soubra A, Hayward D, Dahm P, et al. The diagnostic accuracy of
radical prostatectomy. Clin Genitourin Cancer. 2018;16:e451–67. 18F-fluorodeoxyglucose positron emission tomography and com-
https://doi.org/10.1016/j.clgc.2017.10.003. puted tomography in staging bladder cancer: a single-institution
16. Washington SL, Cowan JE, Herlemann A, et al. Influence of pelvic study and a systematic review with meta-analysis. World J Urol.
lymph node dissection and node-positive disease on biochemical 2016;34:1229–37. https://doi.org/10.1007/s00345-­016-­1772-­z.
recurrence, secondary treatment, and survival after radical pros- 33. Jeong IG, Hong S, You D, Hong JH, Ahn H, Kim C-S. FDG PET-
tatectomy in men with prostate cancer. Prostate. 2021;81:102–8. CT for lymph node staging of bladder cancer: a prospective study of
https://doi.org/10.1002/pros.24085. patients with extended pelvic lymphadenectomy. Ann Surg Oncol.
17. Moris L, Van den Broeck T, Tosco L, et al. Impact of lymph node 2015;22:3150–6. https://doi.org/10.1245/s10434-­015-­4369-­7.
burden on survival of high-risk prostate cancer patients following 34. Kim S-J, Koo PJ, Pak K, Kim I-J, Kim K. Diagnostic accuracy of
radical prostatectomy and pelvic lymph node dissection. Front C-11 choline and C-11 acetate for lymph node staging in patients
Surg. 2016;3:416–9. https://doi.org/10.3389/fsurg.2016.00065. with bladder cancer: a systematic review and meta-­analysis. World J
18. Fossati N, Willemse P-PM, Van den Broeck T, et al. The benefits and Urol. 2018;36:331–40. https://doi.org/10.1007/s00345-­017-­2168-­4.
harms of different extents of lymph node dissection during radical 35. Birkhäuser FD, Studer UE, Froehlich JM, et al. Combined
prostatectomy for prostate cancer: a systematic review. Eur Urol. ultrasmall superparamagnetic particles of iron oxide-enhanced and
2017;72:84–109. https://doi.org/10.1016/j.eururo.2016.12.003. diffusion-weighted magnetic resonance imaging facilitates detec-
19. Lestingi JFP, Guglielmetti GB, Trinh QD, et al. Extended versus tion of metastases in normal-sized pelvic lymph nodes of patients
limited pelvic lymph node dissection during radical prostatectomy with bladder and prostate cancer. Eur Urol. 2013;64:953–60.
for intermediate- and high-risk prostate cancer: early oncological https://doi.org/10.1016/j.eururo.2013.07.032.
outcomes from a randomized phase 3 trial. Eur Urol. 2021;79:595– 36. Zarifmahmoudi L, Ghorbani H, Sadri K, et al. Sentinel node
604. https://doi.org/10.1016/j.eururo.2020.11.040. biopsy in Urothelial carcinoma of the bladder: systematic review
20. Cheng L, Weaver AL, Leibovich BC, et al. Predicting the survival and meta-analysis. Urol Int. 2019;103:373–82. https://doi.
of bladder carcinoma patients treated with radical cystectomy. org/10.1159/000497310.
Cancer. 2000;88:2326–32. https://doi.org/10.1002/(sici)1097-­ 37. Zarifmahmoudi L, Ghorbani H, Sadeghi R, et al. Sentinel lymph
0142(20000515)88:10<2326::aid-­cncr17>3.0.co;2-­t. node biopsy in muscle-invasive bladder cancer: single-center expe-
21. Shinagare AB, Ramaiya NH, Jagannathan JP, Fennessy FM, Tap- rience. Ann Nucl Med. 2020;34:718–24. https://doi.org/10.1007/
lin ME, Van den Abbeele AD. Metastatic pattern of bladder can- s12149-­020-­01502-­6.
cer: correlation with the characteristics of the primary tumor. AJR 38. Watts KL, Ristau BT, Yamase HT, Taylor JA. Prognostic implica-
Am J Roentgenol. 2011;196:117–22. https://doi.org/10.2214/ tions of lymph node involvement in bladder cancer: are we under-
AJR.10.5036. staging using current methods? BJU Int. 2011;108:484–92. https://
22. Stein JP, Lieskovsky G, Cote R, Groshen S, Feng AC, Boyd S, doi.org/10.1111/j.1464-­410X.2011.10330.x.
Skinner E, Bochner B, Thangathurai D, Mikhail M, Raghavan 39. Herr HW, Bochner BH, Dalbagni G, Donat SM, Reuter VE,
D, Skinner DG. Radical cystectomy in the treatment of invasive Bajorin DF. Impact of the number of lymph nodes retrieved on
bladder cancer: long-term results in 1,054 patients. J Clin Oncol. outcome in patients with muscle invasive bladder cancer. J Urol.
2001;19(3):666–75. https://doi.org/10.1200/JCO.2001.19.3.666. 2002;167:1295–8.
PMID: 11157016. 40. Konety BR, Joslyn SA, O'Donnell MA. Extent of pelvic lymph-
23. Shariat SF, Karakiewicz PI, Palapattu GS, et al. Outcomes of radical adenectomy and its impact on outcome in patients diagnosed with
cystectomy for transitional cell carcinoma of the bladder: a contem- bladder cancer: analysis of data from the surveillance, epidemiol-
porary series from the bladder cancer research consortium. J Urol. ogy and end results program data base. J Urol. 2003;169:946–50.
2006;176:2414–22–discussion2422. https://doi.org/10.1016/j. https://doi.org/10.1097/01.ju.0000052721.61645.a3.
juro.2006.08.004. 41. Stein JP, Cai J, Groshen S, Skinner DG. Risk factors for patients
24. Roth B, Wissmeyer MP, Zehnder P, et al. A new multimodality with pelvic lymph node metastases following radical cystectomy
technique accurately maps the primary lymphatic landing sites of with en bloc pelvic lymphadenectomy: concept of lymph node
the bladder. Eur Urol. 2010;57:205–11. https://doi.org/10.1016/j. density. J Urol. 2003;170:35–41. https://doi.org/10.1097/01.
eururo.2009.10.026. ju.0000072422.69286.0e.
25. Paño B, Sebastià C, Buñesch L, et al. Pathways of lymphatic 42. Kassouf W, Agarwal PK, Herr HW, et al. Lymph node density is
spread in male urogenital pelvic malignancies. Radiographics. superior to TNM nodal status in predicting disease-specific survival
2011;31:135–60. https://doi.org/10.1148/rg.311105072. after radical cystectomy for bladder cancer: analysis of pooled data
26. Paner GP, Stadler WM, Hansel DE, Montironi R, Lin DW, Amin from MDACC and MSKCC. J Clin Oncol. 2008;26:121–6. https://
MB. Updates in the eighth edition of the tumor-node-metastasis doi.org/10.1200/JCO.2007.12.9247.
staging classification for urologic cancers. Eur Urol. 2018;73:560– 43. Dhar NB, Klein EA, Reuther AM, Thalmann GN, Madersbacher
9. https://doi.org/10.1016/j.eururo.2017.12.018. S, Studer UE. Outcome after radical cystectomy with limited or
27. Horn T, Zahel T, Adt N, et al. Evaluation of computed tomography extended pelvic lymph node dissection. J Urol. 2008;179:873–8–
for lymph node staging in bladder cancer prior to radical cystec- discussion878. https://doi.org/10.1016/j.juro.2007.10.076.
tomy. Urol Int. 2016;96:51–6. https://doi.org/10.1159/000440889. 44. Leissner J, Ghoneim MA, Abol-Enein H, et al. Extended radi-
28. Ficarra V, Dalpiaz O, Alrabi N, Novara G, Galfano A, Artibani cal lymphadenectomy in patients with urothelial bladder cancer:
W. Correlation between clinical and pathological staging in a series of results of a prospective multicenter study. J Urol. 2004;171:139–44.
radical cystectomies for bladder carcinoma. BJU Int. 2005;95:786– https://doi.org/10.1097/01.ju.0000102302.26806.fb.
90. https://doi.org/10.1111/j.1464-­410X.2005.05401.x. 45. Gschwend JE, Heck MM, Lehmann J, et al. Extended versus limited
29. Salminen AP, Jambor I, Syvanen KT, Boström PJ. Update on novel lymph node dissection in bladder cancer patients undergoing radical
imaging techniques for the detection of lymph node metastases in cystectomy: survival results from a prospective randomized trial. Eur
bladder cancer. Itali J Urol Nephrol. 2016;68:138–49. Urol. 2019;75:604–11. https://doi.org/10.1016/j.eururo.2018.09.047.
 Part XV
Biology and Clinical Aspects of Sarcoma
Progression

William W. Tseng

“Sarcoma” derives from the Greek word meaning “fleshy mass.” Sarcoma encompasses a
group of rare malignancies that originate from the mesenchymal or connective tissues. These
tumors develop from the structural support within the body including the bones and soft tis-
sues. This is in contrast to the more common cancers (e.g., breast, colon carcinoma) which
arise from the epithelium that lines the surfaces of hollow organs and structures. Sarcoma is
also different on many levels from melanoma, which arises from the pigmented cells of the
skin.
With this background it is important to emphasize that sarcoma is therefore (1) not organ
specific and (2) includes a diversity of histologic subtypes based on the precise tissue of origin.
As an example, liposarcoma originates from fat, can occur anywhere in the body (extremities,
trunk, or abdomen), and can arguably be regarded as a separate malignancy from another sar-
coma subtype (e.g., leiomyosarcoma—originates from smooth muscle). Although recognition
of the tissue of origin is very important, ironically the answer is not entirely clear for some
subtypes (e.g., desmoid tumor, undifferentiated pleomorphic sarcoma).
While there are well-known exceptions, in general, bone and soft-tissue sarcomas are
aggressive malignancies with a high mortality rate. Overall, clinical management of sarcoma
can be very challenging. Patients with sarcoma are best managed by a specialist in this disease
and ideally in the setting of multidisciplinary team discussion. Recognition of the histologic
subtype, as mentioned, is important as a start. Surgery remains the mainstay of treatment for
localized disease; however, recurrence (both local and distant) remains a significant issue in
most patients. As such, other modalities for disease control such as radiation and systemic
therapy are important. Further research is needed to provide better treatments and novel
approaches to treatment that are desperately needed in patients with sarcoma.
This section is entirely dedicated to sarcoma. This includes chapters that provide overviews
of bone and separately soft-tissue sarcoma, followed by more focused chapters that explore
select subtypes. All of the chapters are written by sarcoma specialists and fundamentally
grounded with clinical experience: “What is important to the patient with sarcoma?” However,
with each chapter, whenever possible, the authors try to discuss disease behavior and clinical
observations that deserve further scientific investigation and specific data from translational
studies that may help to elucidate mechanisms for disease development and progression.
Finally, one specific chapter is dedicated to a relatively newer and promising “frontier” in sar-
coma research, the immunologic aspects of the tumor microenvironment. The hope is that the
content is relevant to both the practicing clinician and the research scientist and can encourage
dialogue between these two audiences.
594 Biology and Clinical Aspects of Sarcoma Progression

As the reader will quickly see, different subtypes of sarcoma have unique and biologically fascinating
behavior. Liposarcoma, which includes at least three distinct subtypes, can be regarded as a spectrum of
aggressiveness and metastatic potential. Desmoid tumors have no metastatic potential and current clini-
cal management includes observation, as a subpopulation of patients will experience spontaneous regres-
sion. While the lungs are overall the predominant site of sarcoma metastasis, some subtypes can “ignore”
this pattern of dissemination with clear predilection for specific organs (e.g., bone or brain) or even tis-
sues (e.g., fat). With rare exceptions, the vast majority of sarcoma subtypes do not spread to the lymph
nodes, in stark contrast to paradigms seen in carcinoma and melanoma. Interestingly, some of the sub-
types which do demonstrate lymphatic spread either are “epithelioid” or have melanoma markers (e.g.,
clear cell sarcoma)!
Thus, with this section, the reader is invited to immerse himself/herself in learning about sarcoma. While
doing this, the reader should note similarities and differences between sarcoma and other non-sarcoma
cancers. Based on this, the reader is also encouraged to formulate scientific hypotheses about this group of
rare malignancies that may ultimately—perhaps through interdisciplinary collaboration—help us to better
understand fundamental disease mechanisms across all cancers.
 Soft Tissue Sarcoma
56
Scott Kizy and Ricardo J. Gonzalez

Abstract 56.1 Introduction

Soft tissue sarcomas are rare mesenchymal malignancies
made up of several heterogeneous subtypes with distinct clin- Sarcomas are rare malignancies of mesenchymal origin.
icopathologic features. Lymphovascular metastases are rare More than 100 subtypes of sarcomas exist. 80% of sar-
in sarcoma. However, certain subtypes of sarcoma can comas arise from the soft tissue—i.e., fat, muscle, nerve
develop lymphatic metastases—including rhabdomyosar- sheath, and blood vessels. The other 20% arise from the
coma, synovial sarcoma, epithelioid sarcoma, clear cell sar- bone. Only about 1% of new cancer diagnoses are sarco-
coma, and angiosarcoma. Hematogenous spread is more mas. The American Cancer Society estimates that 13,130
common for most subtypes of sarcoma, with common sites of new soft tissue sarcomas will be diagnosed in 2020 in US,
spread being the lung and liver. Metastases in soft tissue sar- and 5350 people are expected to die of soft tissue sarcoma
coma portend a poor prognosis and require multimodal ther- [1]. According to the Surveillance, Epidemiology, and End
apy for management. The underlying mechanisms for Results (SEER) Program of the National Cancer Institute,
lymphatic and hematogenous spread are complex and under- the incidence of soft tissue sarcoma was 1.3 per 100,000 in
studied but likely include vascular endothelial growth factor 2017. The median age at diagnosis is 61 with a slight male
(VEGF) and hypoxia inducible factor (HIF) 1α pathways. preponderance [2].
Further study is necessary to identify potential therapeutic tar- As sarcomas arise from cells of mesenchymal origin, they
gets in the management of patients with soft tissue sarcoma. can occur at all anatomic body sites. The most common site
of presentation is the extremities, with the lower extrem-
ity accounting for 46% of tumors and the upper extremity
accounting for 13%. The torso and retroperitoneum account
Learning Objectives
for 18% and 13%, respectively. The head and neck account
• Soft tissue sarcomas are rare malignancies arising for 9% of soft tissue sarcomas [3].
from cells of mesenchymal origin with several sub- In adults, the most common histologic subtype of soft tis-
types with distinct clinicopathologic features. sue sarcoma is undifferentiated pleomorphic sarcoma, fol-
• Soft tissue sarcomas have a propensity for local recur- lowed by liposarcoma (20%), and leiomyosarcoma (14%).
rence and hematogenous spread (especially to the Other common subtypes of sarcoma include gastrointestinal
lungs and liver) but rarely spread to the lymph nodes. stromal tumors (9%), synovial sarcoma (5%), myxofibrosar-
• A combination of surgical resection, radiation ther- coma (5%), fibrosarcoma (3%), and malignant peripheral
apy, and chemotherapy has been used to treat soft nerve sheath tumors (2%). Each subtype has unique clinical
tissue sarcomas. pathologic characteristics which affect treatment and prog-
• Vascular endothelial growth factor (VEGF)- and nosis [4].
hypoxia inducible factor (HIF) 1α-dependent path- The cornerstone of treatment of soft tissue sarcoma
ways have been implicated in sarcoma metastases, remains surgical resection with adequate margins. Based
but more study is necessary. on variations in locoregional and distant recurrence
among histologic subtypes and disease sites, multimodal
therapy (including chemotherapy and radiation) is used
S. Kizy (*) · R. J. Gonzalez to supplement treatment for patients with soft tissue sar-
Moffitt Cancer Center, Tampa, FL, USA comas. Given the rarity of soft tissue sarcoma and the
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 595
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_56
596 S. Kizy and R. J. Gonzalez

heterogeneity of histologic subtypes, robust clinical data evaluation, and 1.5% of patients in that cohort had lymph node
from well-designed randomized prospective clinical trials metastases. In patients with high-risk subtypes (clear cell sar-
are difficult to complete. coma, epithelioid sarcoma, rhabdomyosarcoma, and angiosar-
An important factor when staging and treating patients coma), 21.3% of patients underwent lymphadenectomy, and
with a new or recurrent sarcoma is identifying the risk for 5.9% were identified to have lymphatic metastases. Lymphatic
distant metastases. Most soft tissue sarcomas carry a sig- metastases are associated with a hazard ratio of mortality of
nificant risk for local recurrence, which can be mitigated by 5.1 compared to similar patients without metastases. Lym-
adjuvant or neoadjuvant radiation therapy to the primary site. phatic spread was associated with younger age, tumor grade,
Distant metastases are primarily spread hematogenously, tumor size, and subtype [10]. Similar rates of lymph node
with the lungs being the most likely site for distant disease involvement have been identified in the European studies [11].
[5]. Hepatic metastases can also occur in patients with ret- In the 2017 revision of the American Joint Committee on
roperitoneal or visceral sarcomas. Lymphatic metastases are Cancer (AJCC), lymph node metastases in sarcomas orig-
quite rare, although they can occur with certain histologic inating in the trunk and extremities are classified as stage
subtypes—i.e., rhabdomyosarcoma, synovial sarcoma, epi- IV disease. For sarcomas of the retroperitoneum or viscera,
thelioid sarcoma, clear cell sarcoma, and vascular sarcoma N1 disease is classified as stage III, given the finding that
[6]. In this chapter we present relevant basic science and the prognosis of patients with nodal disease is slightly more
translational studies on mechanisms of lymphatic metasta- favorable compared to patients with systemic metastases
sis and disease progression in soft tissue sarcomas. Liposar- [12], although these studies may be biased due to the small
coma will be covered separately in the upcoming chapter. incidence of nodal disease in retroperitoneal sarcomas com-
pared to truncal or extremity sarcomas.
For patients with lymph node metastases or suspected
56.2 Clinical Implications lymphatic metastases, management decisions are complex.
Given its rarity, no prospective data exists on the manage-
Lymphatic metastases from soft tissue sarcomas are rare. Given ment of lymph node metastases in sarcoma. Based on retro-
this rarity, relatively little data exist on the clinical implica- spective database studies, omission of lymphatic evaluation
tions and management of lymphatic metastases from soft tis- in these high-risk cohorts (clear cell sarcoma, epithelioid sar-
sue sarcoma. A prospectively gathered database of Soft Tissue coma, rhabdomyosarcoma, and angiosarcoma) was found to
Sarcoma at Memorial Sloan Kettering Cancer Center found have negative effects on survival [10, 13, 14]. Based on these
that about 2.6% of patients develop lymph node metastases. retrospective studies, imaging assessment of the regional
In this study, tumor subtypes with the highest prevalence of lymph node basin is recommended, and surgical evaluation
lymph node metastases were angiosarcoma (13.5%), rhabdo- can be considered for patients with high-risk subtypes.
myosarcoma (13.6%), and epithelioid sarcoma (16.7%) [6]. A The role of imaging to identify lymphatic metastases in
recently published large institutional-­based database from the sarcoma is unclear. The NCCN guidelines propose that for
University of Toronto identified 4.5% of patients with extrem- high-risk histologies, imaging assessment of the regional
ity sarcoma had developed lymph node metastases. The most lymph node basin may be appropriate in the work-up and sur-
common histologic subtypes with lymph node metastases in veillance of patients with soft tissue sarcoma. The appropri-
their dataset were clear cell sarcoma (27.6%), epithelioid sar- ate imaging modality is yet to be identified. CT scan is the
coma (21.9%), rhabdomyosarcoma (17.3%), angiosarcoma most common imaging modality for staging or surveillance
(14%), and extraskeletal myxoid chondrosarcoma (9.3%). of patients with retroperitoneal or extremity sarcoma and can
Synovial sarcomas have been identified as having a high rate identify lymph nodes that are worrisome for metastatic dis-
of lymphatic metastases in other small institutional series, ease (Fig. 56.1). Its sensitivity and specificity in this capacity
but that was not found in the University of Toronto dataset are unknown. FDG-PET imaging has been proposed as a pos-
(2.7%) [7, 8]. Large database studies, including the SEER sible noninvasive method of lymphatic staging in such sub-
database, have found that lymphatic metastases in soft tissue types (Fig. 56.2). However, the sensitivity and specificity of
sarcoma carry a poor prognosis and signify aggressive tumor PET imaging for the identification of lymphatic metastases in
biology. An analysis of the SEER database found that 0.9% soft tissue sarcoma may actually be quite low, with one study
of patients with soft tissue sarcoma (without distant metasta- quoting a sensitivity of 57% and a specificity of 52% [15].
ses) had lymphatic metastases [9]. An analysis of the National Sentinel lymph node biopsy has been proposed as an option
Cancer Database (NCDB) found that 10.8% of patients with for the evaluation of the lymph node basin in patients with
resectable, non-­metastatic soft tissue sarcoma underwent LN high-risk sarcoma histology [16]. Several small series and a
56 Soft Tissue Sarcoma 597

meta-analysis found that patients with clear cell sarcoma may

benefit from sentinel lymph node biopsy given a high number
of occult metastases, but other subtypes had little benefit [16,
17] (Table 56.1). At our center, if a high-­risk lesion is ame-
nable to sentinel node biopsy (i.e., small enough to plan an
injection), we will perform such a technique.
For patients with soft tissue sarcoma with lymph node
involvement, radical lymphadenectomy may provide survival
benefit. The benefit is largely derived from historical single-
institution studies, including the MSKCC database study.
In this study, patients with lymph node involvement had a
median survival of 4.3 months if they did not undergo radi-
cal lymphadenectomy compared to 16.3 months in patients
who underwent lymph node dissection [6]. In the more recent
University of Toronto database study, a similar improvement
in survival was seen in patients treated with radical lymph-
Fig. 56.1 An enlarged right inguinal lymph node identified on CT scan
imaging in a patient with a rare prostate sarcoma adenectomy [7]. Given the retrospective nature of these stud-
ies, there is significant selection bias; however, based on
these historical series, the National Comprehensive Cancer
Network (NCCN) clinical practice guidelines recommend
consideration of regional lymph node dissection at the time
of primary surgery for patients with lymph node involvement.
As lymph node involvement is considered stage IV disease in
extremity sarcoma, and stage III disease in retroperitoneal sar-
coma, multimodal therapy, including regional nodal radiation
therapy, is also recommended by the NCCN18. With patients
without clinically evident lymphatic metastases found to have
a positive sentinel lymph node, there is currently very little
data to guide management. Unlike melanoma, where well
done prospective trials have been evaluated, it is unknown if
nodal observation can be pursued. Given the risk of further
lymphatic metastases, completion lymphadenectomy is gener-
ally performed although this is based on the above historical,
retrospective data. At this time, we would advocate comple-
tion node dissection in this population.
At our institution if a patient with a soft tissue sarcoma
is found to have lymphatic metastases, either via sentinel
lymph node biopsy or preoperative clinical exam/imaging,
most patients are offered a radical lymphadenectomy after
discussion at a multidisciplinary tumor board. Treatment
would include consideration for neoadjuvant systemic ther-
apy, followed by adjuvant radiation in patients with good
Fig. 56.2 FDG-PET imaging identifying a metabolically active
parotid lymph node in a patient with angiosarcoma of the scalp. This
performance status. Stage IV disease is a driver for systemic
was identified to be metastatic angiosarcoma after biopsy therapy prior to surgery.
598 S. Kizy and R. J. Gonzalez

Table 56.1 Sentinel lymph node positivity rate in patients with soft tissue sarcoma
Study Design N SLN positivity False-negative rate
Andreou et al. 2013 [17] Prospective 62 Synovial: 2/42 NR
Clear cell: 6/12
Epithelioid: 0/4
Rhabdomyosarcoma: 0/4
Wright et al. 2012 [29] Meta-analysis 114 Synovial: 2/34 Synovial: 33% (1/3)
Clear cell: 6/17 Clear cell: 25% (2/8)
Epithelioid: 0/17 Epithelioid: 100% (1/1)
Rhabdomyosarcoma: 5/22 Rhabdomyosarcoma: 43% (3/7)
Wagner et al. 2020 [15] Prospective 28 Epithelioid: 1/2 NR
Rhabdomyosarcoma 4/8
Other: 2/8
Parida et al. 2012 [30] Retrospective 23 Total: 1/23 (rhabdomyosarcoma) NR
Maduekwe et al. 2009 [31] Retrospective 29 Total: 3/29 NR

56.3 Mechanisms of Lymphatic Metastases cular endothelial growth factor (VEGF) signaling has been
identified as an important factor in developing lymph node
The molecular mechanisms underlying lymphatic metas- metastases. In particular VEGF-C and VEGF-D have been
tases in soft tissue sarcoma are not understood. Proposed identified in multiple tumor types to induce lymphangiogen-
mechanisms of lymphatic metastases include lymphatic esis, promote lymphatic vessel permeability, and correlate
­vessel development, tumor cell detachment, and survival with lymph node metastases. Certain sarcoma cell lines have
within the lymphatic microenvironment. Whether the lack of been shown to have reduced VEGF-C expression. Overex-
lymph node metastases in soft tissue sarcoma is due to lower pression of VEGF-C in a murine fibrosarcoma and human
lymphangiogenesis, an inability of tumor cells to detach and osteosarcoma cell line increased the ability of these cells to
move through the lymphatic system, or an inhospitable lym- metastasize in the lymph nodes. Patients with sarcoma lymph
phatic microenvironment is unknown. Considering its rarity, node metastases have also been shown to have a high VEGF-
understanding the reason for the resistance of lymph node D expression [22]. Unfortunately, this finding has not been
metastases in soft tissue sarcoma may prove informative in confirmed in other sarcoma human cell lines or tissues. Fur-
the management of patients with metastatic disease. thermore, many tumor subtypes, including those with rare
The density of intratumoral and peritumoral lymphatics lymphatic metastases (liposarcoma and leiomyosarcoma),
likely plays a role in the development of lymphatic spread in have been shown to express high levels of VEGF-C[20].
most human carcinomas. In soft tissue sarcoma, while intra- These findings have decreased the interest in VEGF-C as a
tumoral lymphatic vessel density has been observed to be modulator of lymph node metastases in soft tissue sarcoma.
similar to that of carcinomas, the correlation with lymphatic The role of the tumor microenvironment in sarcoma lym-
spread has not been identified. On the other hand, peritu- phatic metastases is also unknown and understudied. How-
moral lymphatic density has been identified to be lower in ever, the complex interaction between soft tissue sarcoma
soft tissue sarcoma than that of breast cancer. Moreover, tumors and the immune system is beginning to be elucidated.
patients with sarcoma lymph node metastases have been Generally, sarcomas are considered immune quiescent as
identified to have a higher density of peritumoral lymphatics. they commonly have low mutational burden. However, the
The decreased peritumoral lymphatic density in soft tissue complexity of the immune systems interactions with soft
sarcoma has been proposed as a potential cause for the lack tissue sarcoma has been underscored by advancements in
of clinical lymph node metastases [19, 20]. immunotherapy. More recently, soft tissue sarcomas have
The ability of tumor cells to detach and invade nearby been subclassified based on immune cell complement within
lymphatics is also an important component of lymph node the tumor microenvironment. Immune-­ type subclass has
metastases in human malignancy. Lymphovascular inva- identified several subgroups of sarcomas with diverse tumor
sion, or the identification of tumor cells within lymphatic histologies that exhibit different prognoses, virulence, and
vessels on pathologic examination, is commonly found in response to immunotherapy. Tumor subgroups with a higher
most malignancies with a propensity for lymphatic spread. B-cell lineage have improved overall survival and improved
Although inadequately studied, lymphovascular invasion in responsiveness to pembrolizumab, a PD-1 inhibitor [23].
soft tissue sarcoma is rare [21]. The reason for this is unclear, Another avenue of interest is the use of tumor-infiltrating
and more study is necessary. lymphocytes (TIL) in the treatment of certain sarcoma sub-
A variety of cytokines have been identified to play a role types. TIL harvest and culture has been shown to be feasible
in lymphangiogenesis and lymphovascular invasion. Vas- and demonstrated to have tumor-­specific reactivity in recent
56 Soft Tissue Sarcoma 599

studies [24]. As a promising new modality of treatment, [18]. Unfortunately, most studies to date have failed to show
clinical trials are ongoing evaluating the safety and efficacy improvement in outcomes in other sarcoma subtypes.
of TIL in the treatment of metastatic sarcoma. These find-
ings underscore the heterogeneity of soft tissue sarcoma and
complexity of management and may elucidate several mech- 56.5 Conclusions
anisms involved in both lymphatic and systemic metastases
of soft tissue sarcomas. Soft tissue sarcomas are rare tumors of mesenchymal origin.
Multiple histologic subtypes of soft tissue sarcomas have
been identified with varying clinical/pathologic characteris-
56.4 Systemic Metastasis tics and treatment strategies. Lymphatic metastases are rare
among soft tissue sarcomas; however, certain histologic sub-
Hematogenous spread of soft tissue sarcoma is far more types—clear cell sarcoma, epithelioid sarcoma, rhabdomyo-
common than lymphatic metastases. Several mechanisms sarcoma, and angiosarcoma—have a propensity for lymphatic
have been implicated in the ability of sarcoma tumors to spread. Given its rarity, the underlying mechanisms of lym-
metastasize. Owing to the heterogeneity of different sub- phatic spread in soft tissue sarcoma remain understudied.
types of sarcoma, several pathways have been identified to Certain molecular pathways including VEGF signaling,
be important in various histologies. HIF1α, and GP78 have been implicated in lymphatic and sys-
Hypoxia has been identified as an important factor in temic metastases in soft tissue sarcoma. Further study is nec-
the induction of metastatic pathways in multiple sarcoma essary to provide potential targeted therapeutic options for
subtypes. Low intratumoral oxygen level and high levels of patients with soft tissue sarcoma.
hypoxia inducible factor 1α (HIF1α) expression have been
identified as a key predictor of metastases in soft tissue sar-
coma [25]. This is likely dependent on procollagen-lysine,2-­ Open Questions
oxoglutarate 5-dioxygenase 2 (PLOD2)-regulated deposition • What is the role of sentinel lymph node biopsy for
of collagen fibers. PLOD2 seems to modify collagen in a histologic subtypes of soft tissue sarcoma at high
way to allow sarcoma cell migration and motility. In cell line risk for lymph node metastasis?
and murine models, minoxidil, a PLOD t2 inhibitor, has been • What signaling pathways underlie lymphatic and
found to decrease pulmonary metastases [26]. Underscoring systemic metastases in soft tissue sarcoma, and can
the importance of the extracellular matrix and collagen depo- these pathways be utilized to develop therapeutic
sition, matrix metalloproteinases have been identified to play interventions?
an important role in the metastatic potential of leiomyosar- • Is there a role for immunotherapy and immunomod-
comas of uterine origin [27]. ulation in the treatment and management of patients
Gp78, an ubiquitin ligase localized on the endoplasmic with metastatic soft tissue sarcoma?
reticulum, has also been shown to be implicated in sarcoma
metastases. This is likely due to its interaction (ubiquitina-
tion and downregulation) with KA11, a transmembrane
protein and known metastasis suppressor. This mechanism References
was identified in a highly metastatic sarcoma tumor cell line
but also identified in microarray studies of human sarcoma 1. American Cancer Society. Key statistics for soft tissue sarcomas.
tissues. KA11 is thought to induce apoptosis of metastatic 2020. https://www.cancer.org/cancer/soft-tissue-sarcoma/about/
cells, and its downregulation by Gp78 may allow for survival key-statistics.html.
2. Surveillance, Epidemiology, and End Results. Cancer Statiscs and
of tumor cells in its metastatic milieu [28]. Facts: Soft Tissue including heart cancer.
VEGF signaling has also been implicated in the ability for 3. Lawrence W, Donegan WL, Natarajan N, Mettlin C, Beart R, Win-
sarcomas to metastasize. VEGFA is a regulator of vascular- chester D. Adult soft tissue sarcomas: a pattern of care survey of
ization of tumors and contributes to tumor growth and metas- the American College of Surgeons. Ann Surg. 1987;205:349–59.
https://doi.org/10.1097/00000658-­198704000-­00003.
tasis. VEGFA isoform expression increases fibrosarcoma 4. Brennan MF, Antonescu K, Maki A, Maki R. Management of soft
metastases likely through direct tumor vascularization and tissue sarcoma. 2nd ed. Switzerland: Springer; 2016.
changes in the primary tumor microenvironment. Interest 5. Potter DA, Glenn J, Kinsella T, et al. Patterns of recurrence in
in utilizing a targeted VEGF inhibitor, bevacizumab, in the patients with high-grade soft-tissue sarcomas. J Clin Oncol.
1985;3:353–66. https://doi.org/10.1200/JCO.1985.3.3.353.
treatment of patients with metastatic sarcoma has increased. 6. Fong Y, Coit DG, Woodruff JM, Brennan MF. Lymph node
The NCCN currently recommends bevacizumab in certain metastasis from soft tissue sarcoma in adults: analysis of data
epithelioid tumors and malignant solitary fibrous tumors from a prospective database of 1772 sarcoma patients. Ann
600 S. Kizy and R. J. Gonzalez

Surg. 1993;217:72–7. https://doi.org/10.1097/00000658-­ 20. Lahat G, Lazar A, Wang X, et al. Increased vascular endothelial
199301000-­00012. growth factor-C expression is insufficient to induce lymphatic metas-
7. Basile G, Mattei J-C, Alshaygy I, et al. Curability of patients with tasis in human soft-tissue sarcomas. Clin Cancer Res. 2009;15:2637–
lymph node metastases from extremity soft-tissue sarcoma. Cancer. 46. https://doi.org/10.1158/1078-­0432.CCR-­08-­2442.
2020;126:5098–108. https://doi.org/10.1002/cncr.33189. 21. Ethun CG, Lopez-Aguiar AG, Switchenko JM, et al. The prognostic
8. Blazer DG, Lazar AJ, Xing Y, et al. Clinical outcomes of molecu- value of Lymphovascular invasion in Truncal and extremity soft tis-
larly confirmed clear cell sarcoma from a single institution and in sue sarcomas: an analysis from the National Cancer Database. Ann
comparison with data from the surveillance, epidemiology, and end Surg Oncol. 2019;26:4723–9. https://doi.org/10.1245/s10434-­019-­
results registry. Cancer. 2009;115:2971–9. https://doi.org/10.1002/ 07805-­x.
cncr.24322. 22. Yanagawa T, Shinozaki T, Watanabe H, Saito K, Raz A,
9. Johannesmeyer D, Smith V, Cole DJ, Esnaola NF, Camp ER. The Takagishi K. Vascular endothelial growth factor-D is a key
impact of lymph node disease in extremity soft-tissue sarcomas: a molecule that enhances lymphatic metastasis of soft tissue sar-
population-based analysis. Am J Surg. 2013;206:289–95. https:// comas. Exp Cell Res. 2012;318:800–8. https://doi.org/10.1016/j.
doi.org/10.1016/j.amjsurg.2012.10.043. yexcr.2012.01.024.
10. Ecker BL, Peters MG, McMillan MT, et al. Implications of lymph 23. Petitprez F, de Reyniès A, Keung EZ, et al. B cells are associated
node evaluation in the management of resectable soft tissue sar- with survival and immunotherapy response in sarcoma. Nature.
coma. Ann Surg Oncol. 2017;24:425–33. https://doi.org/10.1245/ 2020;577:556–60. https://doi.org/10.1038/s41586-­019-­1906-­8.
s10434-­016-­5641-­1. 24. Mullinax JE, Hall M, Beatty M, et al. Expanded tumor-­infiltrating
11. Behranwala KA, A’Hern R, Omar AM, Thomas JM. Prognosis of lymphocytes from soft tissue sarcoma have tumor-specific func-
lymph node metastasis in soft tissue sarcoma. Ann Surg Oncol. tion. J Immunother. 2021;44:63–70. https://doi.org/10.1097/
2004;11:714–9. https://doi.org/10.1245/ASO.2004.04.027. CJI.0000000000000355.
12. AJCC. Cancer staging manual. 8th ed; 2017. 25. El-Naggar AM, Veinotte CJ, Cheng H, et al. Translational activa-
13. Sherman KL, Kinnier CV, Farina DA, et al. Examination of national tion of HIF1α by YB-1 promotes sarcoma metastasis. Cancer Cell.
lymph node evaluation practices for adult extremity soft tissue 2015;27:682–97. https://doi.org/10.1016/j.ccell.2015.04.003.
sarcoma. J Surg Oncol. 2014;110:682–8. https://doi.org/10.1002/ 26. Eisinger-Mathason TSK, Zhang M, Qiu Q, et al. Hypoxia-dependent
jso.23687. modification of collagen networks promotes sarcoma metastasis.
14. Nelen SD, Vogelaar FJ, Gilissen F, Van der Linden JC, Boss- Cancer Discov. 2013;3:1190–205. https://doi.org/10.1158/2159-
cha K. Lymph node metastasis after a soft tissue sarcoma of the ­8290.CD-­13-­0118.
leg: a case report and a review of the literature. Case Rep Surg. 27. Bodner-Adler B, Bodner K, Kimberger O, Czerwenka K, Leodolter
2013:930361. https://doi.org/10.1155/2013/930361. S, Mayerhofer K. MMP-1 and MMP-2 expression in uterine leio-
15. Wagner LM, Kremer N, Gelfand MJ, et al. Detection of lymph node myosarcoma and correlation with different clinicopathologic
metastases in pediatric and adolescent/young adult sarcoma: senti- parameters. J Soc Gynecol Investig. 2003;10:443–6. https://doi.
nel lymph node biopsy versus fludeoxyglucose positron emission org/10.1016/s1071-­5576(03)00120-­5.
tomography imaging—a prospective trial. Cancer. 2017;123:155– 28. Tsai YC, Mendoza A, Mariano JM, et al. The ubiquitin ligase gp78
60. https://doi.org/10.1002/cncr.30282. promotes sarcoma metastasis by targeting KAI1 for degradation.
16. Blazer DG 3rd, Sabel MS, Sondak VK. Is there a role for sentinel Nat Med. 2007;13:1504–9. https://doi.org/10.1038/nm1686.
lymph node biopsy in the management of sarcoma? Surg Oncol. 29. Wright S, Armeson K, Hill EG, et al. The role of sentinel
2003;12:201–6. https://doi.org/10.1016/s0960-­7404(03)00030-­6. lymph node biopsy in select sarcoma patients: a meta-analysis.
17. Andreou D, Boldt H, Werner M, Hamann C, Pink D, Tunn Am J Surg. 2012;204:428–33. https://doi.org/10.1016/j.amj-
P-U. Sentinel node biopsy in soft tissue sarcoma subtypes with a surg.2011.12.019.
high propensity for regional lymphatic spread 2014;results of a 30. Parida L, Morrisson GT, Shammas A, et al. Role of lymphoscin-
large prospective trial. Ann Oncol. 2013;24:1400–5. https://doi. tigraphy and sentinel lymph node biopsy in the management of
org/10.1093/annonc/mds650. pediatric melanoma and sarcoma. Pediatr Surg Int. 2012;28:571–8.
18. National Comprehensive Cancer Network. Soft Tissue Sarcoma. https://doi.org/10.1007/s00383-­012-­3066-­x.
NCCN. 2020. 31. Maduekwe UN, Hornicek FJ, Springfield DS, et al. Role of senti-
19. Friedrichs N, Hahne JC, Pepper MS, et al. Immunohistochemi- nel lymph node biopsy in the staging of synovial, Epithelioid, and
cal quantification of lymph vessels, VEGF-C and VEGF receptor clear cell sarcomas. Ann Surg Oncol. 2009;16:1356. https://doi.
3 in human sarcomas. Histopathology. 2006;49:87–8. https://doi. org/10.1245/s10434-­009-­0393-­9.
org/10.1111/j.1365-­2559.2006.02438.x.
 Bone Sarcoma
57
Akash A. Shah, Howard Y. Park, and Francis J. Hornicek

Abstract
• Percutaneous biopsy is preferred for tissue diagnosis.
While relatively rare, primary bone sarcomas are associ- • Standard treatment for osteosarcoma is neoadju-
ated with high morbidity and mortality. A comprehensive vant chemotherapy, wide resection, and adjuvant
history, examination, and imaging work-up should be per- chemotherapy.
formed in any suspected bone sarcoma. Percutaneous • Standard treatment for chondrosarcoma is wide
biopsy is preferred to establish a tissue diagnosis. The resection.
MSTS system and AJCC system are employed for staging. • Ewing sarcoma is treated with neoadjuvant chemo-
Osteosarcoma is the most common bone sarcoma, with therapy, wide resection, and adjuvant chemotherapy.
peaks in the second decade of life and after 59 years. • Treatment for fibrous and vascular sarcomas is wide
Standard therapy consists of neoadjuvant chemotherapy, resection.
wide resection with limb salvage, and adjuvant chemother-
apy. Chondrosarcoma is the most common bone sarcoma
of adulthood, affecting patients between 40 and 75 years.
Wide resection is the mainstay of treatment, as it is rela- 57.1 Introduction
tively resistant to radiation and systemic therapy. Ewing
sarcoma is the second most common bone malignancy of Bone tumors can be classified on a spectrum from benign
childhood, occurring between 5 and 25 years. It is classi- to malignant disease. The vast majority of bone lesions is
cally associated with a t(11,22) translocation. It is treated benign. Benign lesions grow slowly, have a low risk of local
with neoadjuvant chemotherapy, wide resection, and adju- recurrence, and do not have metastatic potential.
vant chemotherapy. Fibrous and vascular sarcomas of the Primary malignant bone tumors grow rapidly, carry a
bone are exceedingly rare and treated with wide resection. high risk of local recurrence, and have metastatic potential.
Primary malignant tumors of the bone are typically of mes-
enchymal origin and represent a wide spectrum of disease
Learning Objectives
including sarcoma, lymphoma, and chordoma. There are
• Primary bone sarcomas are associated with high
an estimated 3600 new cases of primary bone malignan-
morbidity and mortality.
cies and 1720 deaths per year in the United States [1]. In
• Imaging work-up includes plain radiographs, CT,
patients 15–19 years of age, bone malignancies are respon-
and MRI.
sible for 19% and 18% of deaths from cancer in males and
females, respectively [2]. Primary bone sarcomas are rare,
accounting for <1% of all diagnosed malignancies per year.
A. A. Shah · H. Y. Park Nevertheless, they carry a high morbidity and mortality
Department of Orthopaedic Surgery, David Geffen School of
Medicine at UCLA, Los Angeles, CA, USA rate—afflicting children, young adults, and older adults.
F. J. Hornicek (*)
Department of Orthopaedics, Sarcoma Multidisciplinary Care
Group, Sylvester Comprehensive Cancer Center, University of
Miami Miller School of Medicine, Los Angeles, CA, USA
Department Orthopaedics, University Miami Miller School of
Medicine, Miami, FL, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 601
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_57
602 A. A. Shah et al.

57.2 Diagnosis 57.3 Staging

57.2.1 History and Physical Examination There are two commonly used staging systems for primary
bone sarcomas: the Musculoskeletal Tumor Society (MSTS)
Obtaining an accurate history is a critical component for system and the American Joint Committee on Cancer (AJCC)
evaluation of the patient with a suspected bone malignancy. system. Largely unchanged from the Enneking system pro-
The presence, nature, and exacerbating factors of pain should posed in 1980, the MSTS system assesses grade, site, and
be assessed. The patient should be asked about constitutional presence of metastases [5]. Low-grade tumors are stage I, and
symptoms. A recent history of trauma should be determined, high-grade tumors are stage II; A or B classifiers are assigned
since pathologic fracture is often the cause of presentation. if the tumor is confined within the cortex of the bone (intra-
In addition to a detailed past medical history, a family history compartmental) or extends beyond the cortex (extracompart-
of malignancy should be obtained. mental), respectively. If metastases are present, the mass is
A comprehensive physical examination should be per- stage III. While the factors employed in MSTS staging are
formed. If a mass is present, the clinician should examine intuitive to surgeons, it rarely impacts the extent of surgery
the mass for presence of a draining sinus or fungation. Size, since most tumors are classified as stage IIB [6].
warmth, tenderness, and firmness should be documented. The AJCC staging system is more consistent with the
Presence of lymphadenopathy should be assessed. A com- TNM convention employed for other tumor types [7]. Fur-
plete skin and abdominal examination should also be per- thermore, it undergoes regular evidence-based updates and
formed to assess for the presence of syndromes that may be includes factors outside the Enneking system. The AJCC
associated with musculoskeletal pathology. system provides different tumor staging criteria for the
appendicular skeleton, spine, and pelvis (Table 57.1). Nota-
bly, distant metastasis is divided into lung and non-lung
57.2.2 Imaging sites, acknowledging literature supporting that patients with
isolated lung metastasis carry a better prognosis than those
Plain radiographs with at least two orthogonal views must with bony metastasis [6].
be obtained; it is imperative to image the entire bone.
Computed tomography (CT) is useful in the pelvis and/
or sacrum to evaluate the degree of bony involvement and 57.4 Osteosarcoma
for pre-­operative planning. CT chest is also needed for
staging purposes, as the lung is the most common site of Osteosarcoma is the most common bone sarcoma. It is char-
metastasis of bone sarcomas. Magnetic resonance imag- acterized by the production of osteoid and bone by malignant
ing (MRI) is a necessary follow-up study if primary bone spindle cells. The peak incidence is in the second decade of
malignancy is suspected. MRI allows for evaluation of life; a second peak occurs after 59 years of age. The most
bone marrow, non-calcific soft tissue, and lymph node common presenting symptom is pain. Intramedullary osteo-
involvement. MRI of the entire involved bone should sarcomas develop in the medullary cavity; juxtacortical
be obtained to assess for skip metastases. A whole-body osteosarcomas arise on the surface of long bones and do not
bone scan may be performed to assess for widespread involve the medullary canal (Table 57.2).
bony metastases. On radiographs, a metaphyseal lesion with blastic or lytic
features and ill-defined borders is classically encountered
[8]. An MRI evaluates soft tissue involvement, bone marrow
57.2.3 Biopsy involvement, and presence of skip metastases. A summary of
osteosarcoma subtypes is provided in Table 57.3.
Biopsy is performed after other staging studies. Percutane-
ous biopsies are generally favored over open biopsies. There
is a higher association of biopsy tract seeding and subsequent 57.4.1 Genetics
local recurrence with open biopsy compared to percutaneous
[3]. Among percutaneous biopsy techniques, core biopsy is Most cases of osteosarcoma are sporadic. Increased expres-
preferred since it allows for assessment of lesion architecture sion of vascular endothelial growth factor is correlated with
compared to fine-needle aspiration [4]. Incisional biopsy is decreased overall survival [9]. Cyclin-dependent kinase
not recommended for evaluation of a suspected malignancy. 11 is essential for osteosarcoma cell growth and survival
Excisional biopsy may be appropriate for a lesion <3 cm or in vitro [10], while rare, genetic predisposition to osteosar-
in a location where wide resection is associated with signifi- coma involves mutations in tumor suppressor genes or DNA
cant morbidity. helicases [8].
57 Bone Sarcoma 603

Table 57.1 Definition of primary tumor in the American Joint Committee on Cancer (AJCC) staging system for bone sarcomas
Category Appendicular skeleton Spine Pelvis
TX Primary tumor cannot be assessed Primary tumor cannot be assessed Primary tumor cannot be assessed
T0 No evidence of primary tumor No evidence of primary tumor No evidence of primary tumor
T1 Tumor <8 cm in greatest Tumor confined to 1–2 adjacent Tumor confined to 1 segment without extraosseous
dimension segments extension
T1a – – Tumor <8 cm in greatest dimension
T1b – – Tumor >8 cm in greatest dimension
T2 Tumor >8 cm in greatest Tumor confined to 3 adjacent segments Tumor confined to 1 segment with extraosseous
dimension extension or 2 segments without extraosseous
extension
T2a – – Tumor <8 cm in greatest dimension
T2b – – Tumor >8 cm in greatest dimension
T3 Presence of skip metastases Tumor confined to >4 adjacent segments Tumor spanning 2 segments with extraosseous
or any nonadjacent segments extension
T3a – – Tumor <8 cm in greatest dimension
T3b – – Tumor >8 cm in greatest dimension
T4 – Extension into spinal canal or great Tumor spanning 3 segments or crossing sacroiliac
vessels joint
T4a – Extension into spinal canal Involves sacroiliac joint and extends medial to
sacral neuroforamina
T4b – Extension into great vessels or presence Tumor encasement of external iliac vessels or
of tumor thrombus presence of tumor thrombus

Table 57.2 Characteristics of common osteosarcoma subtypes

Low-grade High-grade
Conventional Telangiectatic Small cell intramedullary Parosteal Periosteal surface
Proportion of 80% <4% 1.5% 1–2% 1–6% 1–2% <1%
cases
Age Range 1st and 2nd decades 1st and 2nd 1st and 2nd 3rd and 4th 3rd decade 3rd decade 3rd decade
decades decades decades
Common Distal femur, Distal femur, Distal femur Distal femur, Posterior Distal femur, Femur, tibia
Locations proximal tibia, proximal tibia proximal tibia distal femur proximal tibia
proximal humerus
Location on Intramedullary Intramedullary Intramedullary Intramedullary Juxtacortical Juxtacortical Juxtacortical
bone
Frequent High grade High grade High grade Low grade Low grade Intermediate High grade
histologic grade
grade

Table 57.3 Characteristics of chondrosarcoma subtypes

Mesenchymal Clear cell Dedifferentiated
Conventional Chondrosarcoma chondrosarcoma Chondrosarcoma Chondrosarcoma
Precursor lesion Osteochondroma, enchondroma None None Conventional
Chondrosarcoma
Associated Ollier’s disease; Maffucci None None None
syndromes syndrome; MHE
Age range 40s–60s 50–60s Any age, peak 30s–50s Any age, peak 30s–50s
Common locations Pelvis, proximal femur Axial skeleton, extra-skeletal Femoral and humeral Pelvis, femur
(meninges) epiphyses
Frequent histologic Low grade High grade Low grade High grade
grade
Prognosis Good if low grade Poor Good Poor
Chemosensitivity Low Sensitive Low Low
Radiosensitive Low Sensitive Low Low
604 A. A. Shah et al.

57.4.2 Intramedullary Osteosarcoma cell osteosarcoma is characterized by small, round malignant

Subtypes cells within an osteoid matrix. It is differentiated from Ewing
sarcoma by the production of osteoid [11].
57.4.2.1 Conventional Osteosarcoma
Conventional osteosarcoma represents 80% of all osteosar-
comas. It affects patients in the first or second decade of life 57.4.3 Juxtacortical Osteosarcoma
and has a male predominance. The most commonly affected
locations are the distal femur, proximal tibia, and proximal 57.4.3.1 Parosteal Osteosarcoma
humerus. Conventional osteosarcoma tends to be high-grade Parosteal osteosarcoma is the most common juxtacortical
and originates in the metaphysis; diaphyseal or axial sites are osteosarcoma and makes up 1–6% of all osteosarcomas.
less commonly observed [8]. Peak incidence is in the third decade of life and has a female
On radiographs, lytic destruction with cortical disruption predominance [8]. Radiographs typically reveal a densely
is noted. A Codman triangle may be observed. As the tumor ossified mass on the posterior aspect of the distal femur that
extends into the soft tissue, a sunburst pattern of malignant has a “stuck-on” appearance [16]. On histology, a low-grade
bone may be seen. MRI of the entire bone must be evaluated lesion with well-differentiated fibrous stroma is observed
for the presence of skip metastases. [11].
The defining histopathologic feature of osteosarcoma
is the production of osteoid. Malignant mesenchymal cells 57.4.3.2 Periosteal Osteosarcoma
with pleomorphic nuclei and occasional mitotic figures are Periosteal osteosarcoma comprises 1–2% of all osteosarco-
typical for conventional osteosarcoma [11]. Conventional mas [8]. It is a typically radiolucent mass that most com-
osteosarcoma is further subcategorized by the extracellular monly involves the proximal tibia or distal femur [16]. An
matrix produced (e.g., osteoblastic, chondroblastic, fibro- aggressive periosteal reaction (e.g., sunburst appearance,
blastic) [8]. Codman triangle) may be present. Periosteal osteosarcoma
has mostly cartilaginous matrix on histology with small
57.4.2.2 Telangiectatic Osteosarcoma amounts of calcification and osteoid [11].
Typically encountered in the second decade of life, telangi-
ectatic osteosarcoma comprises <4% of osteosarcoma cases. 57.4.3.3 High-Grade Surface Osteosarcoma
A quarter of patients present with pathologic fracture [8]. High-grade surface osteosarcoma accounts for <1% of all
MRI should be obtained when telangiectatic osteosarcoma osteosarcomas [11]. It is a surface lesion usually involving
is suspected. Care must be taken when interpreting imaging, the femur or tibia with extension into the surrounding soft
as findings are similar to benign aneurysmal bone cysts: a tissue with disruption of the underlying cortex [16]. Similar
radiolucent lesion with fluid-fluid levels. A lesion with an to conventional osteosarcoma, high-grade spindle cells with
aggressive pattern of bone destruction with partial fluid-fluid varying amounts of osteoid are observed on histology [11].
levels should prompt core needle biopsy, demonstrating mul-
tiple hemorrhagic cavities with osteoid and malignant stro-
mal cells [11–13]. 57.4.4 Secondary Osteosarcoma

57.4.2.3 Low-Grade Intramedullary The second peak in osteosarcoma incidence in older patients
Osteosarcoma is due to malignant transformation in Paget’s disease. Pagetic
Low-grade intramedullary osteosarcoma affects adults in osteosarcomas tend to be high-grade intramedullary lesions.
the third or fourth decade of life, most commonly in the The pelvis, femur, and humerus are most commonly affected.
distal femur and proximal tibia. On plain radiographs, an Their histologic appearance reflects the chaotic remodeling
unaggressive-­ appearing lytic and/or blastic lesion with of Paget’s disease, with multiple atypical osteoblasts and the
septal sclerosis is encountered [14]. On histology, well-­ presence of osteoclast giant cells [17].
differentiated cells with woven microtrabeculae and fibrous Osteosarcoma may arise through a second hit mutation
stroma are observed [11]. in the setting of an existing genetic predisposition such as
Li-Fraumeni syndrome, inherited retinoblastoma, and Roth-
57.4.2.4 Small Cell Osteosarcoma mund-Thomson syndrome [8]. Radiation-induced osteosar-
Small cell osteosarcoma comprises 1.5% of cases. It has a comas are rare; the mean latency period between radiation
similar age distribution as conventional osteosarcoma. The and sarcoma occurrence is approximately 15 years. These
distal femur is most commonly involved. On radiographs, lesions are almost universally high-grade and carry a dismal
a lytic lesion with variable sclerosis is observed [15]. Small prognosis [18].
57 Bone Sarcoma 605

57.4.5 Treatment and Prognosis 57.5.1 Genetics

Conventional osteosarcoma requires neoadjuvant chemo- Most chondrosarcomas are sporadic. Hedgehog and Rb path-
therapy, wide resection, and adjuvant chemotherapy. In way alterations, matrix metalloproteinases, and hypoxia-­
addition to treating detected or presumed osteosarcoma, inducible factor 1 have been shown to be associated with
chemotherapy may decrease the mass size—allowing for chondrosarcoma progression [24, 25]. Syndromes associated
improved tumor demarcation and a less morbid resec- with secondary chondrosarcoma such as multiple hereditary
tion. MAP therapy is most commonly used. Chemotherapy exostoses (MHE), Ollier’s disease, and Maffucci syndrome
response is measured by percent tumor necrosis. Patients are due to mutations in EXT1/EXT2, IDH1, and IDH2,
with less than 90% necrosis at the time of surgery are at respectively.
increased risk of local recurrence and decreased overall sur-
vival [19]. There is no significant role for radiation therapy;
palliative radiation may be employed in the treatment of the 57.5.2 Primary Chondrosarcoma
painful metastases.
Wide resection with limb salvage is typically per- 57.5.2.1 Conventional Chondrosarcoma
formed; amputation is only considered in cases where a Conventional chondrosarcoma accounts for 85% of all cases,
functional limb cannot remain or in cases of recurrence. occurring most commonly in the proximal femur and pelvis.
It is imperative that negative margins be obtained. Posi- Conventional central chondrosarcoma arise de novo within
tive margins or < 2 mm are associated with increased the bone, while peripheral lesions arise from malignant trans-
risk of local recurrence and decreased survival [19, 20]. formation within a pre-existing benign cartilage cap [26].
If limb salvage is feasible, several reconstructive options Conventional chondrosarcoma can be assigned to three
exist including endoprosthetic reconstruction. Low-grade grades. Locally aggressive with low metastatic potential,
lesions such as parosteal and low-­grade intramedullary grade I chondrosarcoma contains abundant hyaline cartilage
subtypes can be treated with wide resection alone. Higher- matrix with low cellularity and no pleomorphism. Grade II
grade osteosarcomas such as periosteal, high-grade sur- chondrosarcoma has some pleomorphism and increased cellu-
face, and secondary subtypes are treated like conventional larity. Grade III lesions are extremely cellular with high pleo-
osteosarcoma. morphism; metastasis is observed in up to 70% of cases [27].
Long-term survival rates of patients with localized con-
ventional osteosarcoma range from 60% to 80%; this falls 57.5.2.2 Clear Cell Chondrosarcoma
to 20–30% in patients with metastasis [8]. While high-grade Clear cell chondrosarcoma is a low-grade variant, observed in
surface osteosarcoma has a similar prognosis to conventional patients between 20 and 50 years. Clear cell chondrosarcoma
osteosarcoma, the other juxtacortical subtypes have an excel- occurs in the epiphysis with metaphyseal extension, most
lent prognosis. Poor prognostic factors include metastatic commonly in the femoral head. The typical radiographic
disease, poor chemotherapy response, axial location, and appearance is a well-marginated lucency with sclerotic bor-
skip metastases [8, 19, 21]. ders [28]. Histologically, clear cell chondrosarcoma is char-
acterized the presence of clear cells with round nuclei [29].

57.5 Chondrosarcoma 57.5.2.3 Dedifferentiated Chondrosarcoma

Dedifferentiated chondrosarcoma is the most malignant vari-
Chondrosarcoma is the most common bone sarcoma of ant, accounting for approximately 10% of cases. It tends to
adulthood, with an incidence of 1 in 200,000 per year [22]. It occur in patients between 50 and 70 years. It is found in a
has a slight male predominance and affects patients between similar anatomic distribution to conventional lesions and is
40 and 75 years. Chondrosarcoma is most commonly found associated with a high risk of pathologic fracture. It is his-
in the appendicular skeleton; less than 10% occur in the axial tologically characterized by two distinct neighboring areas
skeleton. The most common presenting complaint is pain of low-grade conventional chondrosarcoma and high-grade
due to weakening of the cortex; pathologic fracture may be heterogeneous sarcoma [30].
encountered.
Radiographic signs of aggressive biology include endos- 57.5.2.4 Mesenchymal Chondrosarcoma
teal scalloping >50% of cortical width, cortical thinning, and Mesenchymal chondrosarcoma is a rare high-grade variant
lucency adjacent to an enchondroma [23]. CT demonstrates most commonly occurring in the jaw, spine, and ribs. It has a
a “rings and arcs” pattern of peripheral calcification. MRI biphasic histology with subtypes of small round cell compo-
assesses intramedullary and soft tissue involvement. nents with islands of well-differentiated chondroid tissue [31].
606 A. A. Shah et al.

57.5.3 Secondary Chondrosarcoma 57.6.1 Genetics and Histology

The majority of secondary chondrosarcomas arise as Ewing sarcoma is classically associated with translocations
osteochondromas in patients with MHE. Malignant between the EWS and ETS families of genes. In 80–95%
transformation occurs in 5–10% of cases; a cartilage of cases, the translocation is t(11;22) and produces the
cap thicker than 1–2 centimeters should raise suspicion. EWS-­FLI1 fusion protein [38]. EWS-FLI1 stimulates prolif-
Ollier’s disease carries a 10–40% risk of malignant trans- eration, differentiation, and survival of neural crest-derived
formation, and small studies have indicated that nearly cells [40]. Ewing sarcoma is a high-grade tumor histologi-
all patients with Maffucci syndrome develop chondrosar- cally characterized by sheets of small round blue cells with
coma [32]. a high nuclear-to-cytoplasm ratio; pseudo-rosettes may be
present [38]. CD99 is expressed almost universally.

57.5.4 Treatment and Prognosis

57.6.2 Treatment and Prognosis
Usually low-grade, chondrosarcoma is largely unresponsive
to radiation and chemotherapy. Radiation does not offer local Ewing sarcoma is treated with neoadjuvant chemotherapy
control and is only employed for unresectable tumors or pal- followed by resection and adjuvant chemotherapy. Wide
liation. Chondrosarcomas have poor vascularity and low resection with negative margins allows for local control with
associated pH, impeding the delivery of systemic therapy a low recurrence rate and also avoids the risk of a radiation-­
[24]. Furthermore, chondrosarcoma cells express high levels induced malignancy. In addition to superior local control
of P-glycoprotein—an efflux pump that removes chemother- and survival, surgical management provides the advantage
apeutic agents from cells [33]. of assessing tumor response to chemotherapy [38]. Adjuvant
The consensus treatment is wide resection for all grades radiation may be appropriate for unresectable disease or
of conventional chondrosarcoma [34]. Wide resection is those with poor chemotherapy response. Radiation of lung
indicated for dedifferentiated chondrosarcoma; young metastases improves survival.
patients may benefit from adjuvant chemotherapy [35]. Clear Five-year survival is 65–80% and 25–40% in patients with
cell chondrosarcoma carries a high risk of local recurrence localized and metastatic disease at presentation, respectively
and requires wide resection. Contrary to other subtypes, [38]. Factors associated with favorable prognosis include
mesenchymal chondrosarcoma is radiosensitive and chemo- peripheral site, normal LDH level at presentation, tumor vol-
sensitive. Treatment includes wide resection with adjuvant ume <100 ml, and localized disease [39]. The presence of
chemoradiation [36]. metastatic disease—especially bony metastasis—is the most
Appendicular chondrosarcomas have improved sur- significant risk factor for poor prognosis. Recurrent disease
vival and local recurrence rates compared to axial is associated with <10% 5-year survival [38].
lesions. Conventional, clear cell, and mesenchymal chon-
drosarcomas have five-year survival rates of 70%, 100%,
and 50% [37]. Most patients with dedifferentiated chon- 57.7 Other Bone Sarcomas
drosarcoma succumb to metastatic disease within 1 year
[22]. 57.7.1 Fibrosarcoma of the Bone

Fibrosarcoma incidence peaks in the fourth and sixth decades

57.6 Ewing Sarcoma of life. It most commonly occurs in the metaphyseal distal
femur. An eccentric lytic lesion with a permeative appearance
Ewing sarcoma is the second most common bone malig- is observed on plain radiographs. Fibrosarcoma is character-
nancy in children after osteosarcoma. It is typically seen ized by a malignant spindle cell population arranged in a “her-
between 5 and 25 years with a slight male predominance ringbone” pattern. The most significant prognostic factor is
[38]. The pelvis and femur are most commonly affected. histologic grade. Wide resection is the preferred treatment [41].
Pain and swelling are classic presenting symptoms. Unlike
other bone sarcomas, constitutional symptoms such as
fever, weight loss, and fatigue are often present [38, 57.7.2 Undifferentiated High-Grade
39]. Elevated serum lactate dehydrogenase (LDH) may Pleomorphic Sarcoma
be present. A lytic permeative lesion is often seen, often
accompanied by a periosteal reaction with an “onion-skin” Historically called malignant fibrous histiocytoma, undiffer-
appearance. entiated pleomorphic sarcoma is most commonly seen after
57 Bone Sarcoma 607

the fifth decade of life. On radiographs, a diffusely lytic tumor 57.9 Conclusion
with cortical destruction is seen in metaphyseal-­diaphyseal
bone. A high-grade tumor, it is histologically characterized • Primary bone sarcoma is associated with high morbidity
by a mixed population of spindle cells, malignant histiocytes, and mortality.
and multinucleated giant cells. Undifferentiated pleomorphic • A thorough history and physical examination is critical.
sarcoma has high rates of local recurrence and metastasis. • Radiographs, MRI, and CT are necessary to determine the
Treatment is wide resection and adjuvant chemotherapy [42]. extent of disease and for staging.
• Percutaneous core biopsy is preferred for tissue
diagnosis.
57.7.3 Angiosarcoma • Bone sarcomas can be staged using the MSTS or AJCC
staging systems.
Angiosarcoma is a rare malignancy of endothelial origin • Osteosarcoma is the most common bone sarcoma, typi-
most commonly involving the proximal femur. Approxi- cally affecting patients in the second decade of life. The
mately a third of cases are multifocal, usually in the same treatment is neoadjuvant chemotherapy, wide resection,
bone or limb. Angiosarcoma is composed of sheets of atypi- and adjuvant chemotherapy.
cal epithelioid cells with eosinophilic cytoplasm where vacu- • Chondrosarcoma is the most common bone sarcoma of
oles with erythrocyte fragments are present. Angiosarcoma adulthood, typically affecting patients 40–75 years. Wide
is often intermediate- or high-grade at presentation and car- resection is the mainstay of treatment for
ries a 30% five-year survival. Wide resection offers the best chondrosarcoma.
chance of local control. Multifocal disease may benefit from • Ewing sarcoma primarily occurs in patients between 5
adjuvant chemoradiation [43]. and 25 years. It is characterized by the t(11;22) transloca-
tion with production of the EWS-FLI1 fusion protein.
Standard therapy consists of neoadjuvant chemotherapy,
57.8  dvances in Translational Bone
A wide resection, and adjuvant chemotherapy.
Sarcoma Research • Fibrous and vascular sarcomas are treated with wide
resection.
In recent years, new investigations into the tumor microen- • Advances in understanding of the tumor microenviron-
vironment and molecular progression of primary bone sar- ment and molecular drivers of bone sarcoma progression
coma have led to advances in our understanding of potential provide potential targets for therapy.
therapeutic avenues. The tumor microenvironment plays
an important role in malignant transformation as well as
disease progression [44, 45]. Tumor-associated macro-
phages—the primary immune cells present in the tumor
microenvironment—are recruited by secretion of chemo- Open Questions
kines from osteosarcoma cells [46, 47]. M2 macrophages • Can the genetic drivers of bone sarcoma be better
in the tumor microenvironment promote osteosarcoma elucidated to allow for targeted therapy?
metastasis through secretion of matrix metalloproteinase 12, • What genetic characteristics of mesenchymal chon-
which subsequently degrades the extracellular matrix [48]. drosarcoma allow it to be more radiosensitive and
Tumor-­associated macrophages have also been implicated chemosensitive compared to other subtypes?
in chemotherapy resistance [49]. Targeting macrophage • What soft-tissue coverage options will allow for
activity for osteosarcoma immunotherapy is an active area decreased complications after wide resection for
of research. The immune microenvironment associated with primary bone sarcomas?
Ewing sarcoma also plays a critical role in disease progres- • Can advanced machine learning methods be used to
sion. Immunosuppressive fibrocytes present in the Ewing more accurately predict survival or outcomes after
sarcoma cell microenvironment induce angiogenesis and treatment for primary bone sarcomas?
mediate immunosuppression [50]. Multiple surface and • Can advances in radiation modalities allow for
intracellular tumor-specific antigens have been identified for more targeted therapy for historically radioresistant
Ewing sarcoma; these have been used as targets for mono- bone sarcomas?
clonal antibodies, chimeric antigen receptor T cell-mediated
approaches, and vaccine-based approaches [51].
608 A. A. Shah et al.

References 23. Marco R, Gitelis S, Brebach G, Healey J. Cartilage tumors: evalua-

tion and treatment. J Am Acad Orthop Surg. 2000;8(5):292–304.
24. Bovee J, Hogendoorn P, Wunder J, Alman B. Cartilage tumors and
1. Siegel R, Miller K, Jemal A. Cancer statistics, 2020. CA Cancer J
bone development: molecular pathology and possible therapeutic
Clin. 2020;70(1):7–30.
targets. Nat Rev Cancer. 2010;10(7):481–8.
2. Miller K, Fidler-Benaoudia M, Keegan T, Hipp H, Jemal A, Siegel
25. Schrage Y, Lam S, Jochemsen A, et al. Central chondrosar-
R. Cancer statistics for adolescents and young adults, 2020. CA
coma progression is associated with pRb pathway alterations:
Cancer J Clin. 2020;70(6):443–59.
CDK4 down-­regulation and p16 overexpression inhibit cell
3. Barrientos-Ruiz I, Ortiz-Cruz E, Serrano-Montilla J, Bernabeu-­
growth in vitro. J Cell Mol Med. 2009;13(9A):2843–52.
Taboada D, Pozo-Kreilinger J. Are biopsy tracts a concern for
26. Organization WH. Cartilage tumors. In: Fletcher C, Unni K,
seeding and local recurrence in sarcomas? Clin Orthop Relat Res.
Mertens F, editors. World Health Organization classification of
2017;475(2):511–8.
tumors. Pathology and genetics. Tumors of soft tissue and bone.
4. Mitsuyoshi G, Naito N, Kawai A, et al. Accurate diagnosis of mus-
Lyon, France: IARC Press; 2002. p. 234–57.
culoskeletal lesions by core biopsy. J Surg Oncol. 2006;94(1):21–7.
27. Angelini A, Guerra G, Mavorgenis A, Pala E, Picci P, Ruggieri
5. Enneking W, Spanier S, Goodman M. A system for the surgical staging
P. Clinical outcome of central conventional chondrosarcoma. J Surg
of musculoskeletal sarcoma. Clin Orthop Relat Res. 1980;153:106–
Oncol. 2012;106(8):929–37.
20. https://doi.org/10.1007/978-­1-­4471-­5451-­8_127.
28. Collins M, Koyama T, Swee R, Inwards C. Clear cell chondrosar-
6. Steffner R, Jang E. Staging of bone and soft-tissue sarcomas. J Am
coma: radiographic, computed tomographic, and magnetic reso-
Acad Orthop Surg. 2018;26:e269–78.
nance findings in 34 patients with pathologic correlation. Skelet
7. Kneisl J, Rosenberg A, Anderson P, et al. Bone. In: Amin M, Edge
Radiol. 2003;32(12):687–94.
S, Greene F, et al., editors. AJCC cancer staging manual. 8th ed.
29. Itala A, Leerapun T, Inwards C, Collins M, Scully S. An institu-
Springer International: Switzerland; 2017. p. 471–86.
tional review of clear cell chondrosarcoma. Clin Orthop Relat Res.
8. Messerschmitt P, Garcia R, Abdul-Karim F, Greenfield E, Getty
2005;440:209–12.
P. Osteosarcoma. J Am Acad Orthop Surg. 2009;17:515–27.
30. Dornauer K, Soder S, Inwards C, Boveee J, Aigner T. Matrix bio-
9. Wu H, Zhang J, Dai R, Xu J, Feng H. Transferrin receptor-1 and
chemistry and cell biology of dedifferentiated chondrosarcomas.
VEGF are prognostic factors for osteosarcoma. J Orthop Surg Res.
Pathol Int. 2010;60(5):365–72.
2019;14(1):296.
31. Ghafoor S, Hameed M, Tap W, Hwang S. Mesenchymal chondro-
10. Feng Y, Sassi S, Shen J, et al. Targeting CDK11 in osteosar-
sarcoma: imaging features and clinical findings. Skelet Radiol.
coma cells using the CRISPR-Cas9 system. J Orthop Res.
2021;50(2):333–41.
2015;33(2):199–207.
32. Lin P, Moussallem C, Deavers M. Secondary chondrosarcoma. J
11. Klein M, Siegal G. Osteosarcoma: anatomic and histologic vari-
Am Acad Orthop Surg. 2010;18(10):608–15.
ants. Am J Clin Pathol. 2006;125(4):555–81.
33. Trieb K, Sulzbacher I, Kubista B. Recurrence rate and progression
12. Zishan U, Pressney I, Khoo M, Saifuddin A. The differentia-
of chondrosarcoma is correlated with heat shock protein expres-
tion between aneurysmal bone cyst and telangiectatic osteosar-
sion. Oncol Lett. 2016;11(1):521–4.
coma: a clinical, radiographic and MRI study. Skelet Radiol.
34. Zoccali C, Baldi J, Attala D, et al. Intralesional versus extral-
2020;49(9):1375–86.
esional procedures for low-grade central chondrosarcoma: a
13. Hegde V, Burke Z, Park H, et al. Is core needle biopsy reliable in
systematic review of the literature. Arch Orthop Trauma Surg.
differentiating between aggressive benign and malignant radiolu-
2018;138(7):929–37.
cent bone tumors? Clin Orthop Relat Res. 2018;476(3):568–77.
35. Nooij M, Whelan J, Bramwell V, et al. Doxorubicin and cisplatin
14. Andresen K, Sundaram M, Unni K, Sim F. Imaging features of
chemotherapy in high-grade spindle cell sarcomas of the bone, other
low-­grade central osteosarcoma of the long bones and pelvis. Skelet
than osteosarcoma or malignant fibrous histiocytoma: a European
Radiol. 2004;33(7):373–9.
osteosarcoma intergroup study. Eur J Cancer. 2005;41(2):225–30.
15. Nakajima H, Sim F, Bond J, Unni K. Small cell osteosarcoma of
36. De Amorim BK, DeLaney T. Chordomas and chondrosarcomas -
bone: review of 72 cases. Cancer. 1997;79(11):2095–106.
the role of radiation therapy. J Surg Oncol. 2016;114(5):564–9.
16. Harper K, Sathiadoss P, Saifuddin A, Sheikh A. A review of imag-
37. Schneiderman B, Kliethermes S, Nystrom L. Survival in mesen-
ing of surface sarcomas of bone. Skelet Radiol. 2021;50:9–28.
chymal chondrosarcoma varies based on age and tumor location:
17. Hansen M, Seton M, Merchant A. Osteosarcoma in Paget’s disease
a survival analysis of the SEER database. Clin Orthop Relat Res.
of bone. J Bone Miner Res. 2007;21:P58–63.
2017;475:799–805.
18. Mavrogenis A, Pala E, Guerra G, Ruggieri P. Post-radiation
38. Maheshwari A, Cheng E. Ewing sarcoma family of tumors. J Am
sarcomas: clinical outcome of 52 patients. J Surg Oncol.
Acad Orthop Surg. 2010;18:94–107.
2012;105(6):570–6.
39. Biermann J, Chow W, Reed D, et al. NCCN guidelines insights:
19. Jeys L, Thorne C, Parry M, Gaston C, Sumathi V, Grimer J. A
bone cancer, version 2.2017. J Natl Compr Cancer Netw.
novel system for the surgical staging of primary high-grade osteo-
2017;15(2):155–67.
sarcoma: the Birmingham classification. Clin Orthop Relat Res.
40. Grunewald T, Bernard V, Gilardi-Hebenstreit P, et al. Chimeric
2017;476(3):842–50.
EWSR1-FLI1 regulates the Ewing sarcoma susceptibility gene
20. Bertrand T, Cruz A, Binitie O, Cheong D, Letson G. Do surgical
EGR2 via a GGAA microsatellite. Nat Genet. 2015;47(9):1073–8.
margins affect local recurrence and survival in extremity, non-
41. Papagelopoulos P, Galanis E, Trantafllidis P, Bascainos P, Sim F,
metastatic, high-grade osteosarcoma? Clin Orthop Relat Res.
Unni K. Clinicopathologic feature, diagnosis, and treatment of
2016;474(3):677–83.
fibrosarcoma of bone. Am J Orthop. 2002;31(5):253–7.
21. Sajadi K, Heck R, Neel M, et al. The incidence and progno-
42. Kalil R. Malignant fibrous histiocytoma of bone. In: Santini-Araujo
sis of osteosarcoma skip metastases. Clin Orthop Relat Res.
E, Kalil R, Bertoni F, Park Y, editors. Tumors and tumor-like lesions
2004;426:92–6.
of bone. London: Springer; 2015.
22. Giuffrida A, Burguneo J, Koniaris L, Gutierrez J, Duncan R, Scully
43. Kalil R. Angiosarcoma. In: Santini-Araujo E, Kalil R, Bertoni F,
S. Chondrosarcoma in the United States (1973 to 2003): an analy-
Park Y, editors. Tumors and tumor-like lesions of bone. London:
sis of 2890 cases from the SEER database. J Bone Joint Surg Am.
Springer; 2015.
2009;91:1063–72.
57 Bone Sarcoma 609

44. Binnewies M, Roberts E, Kersten K, Chan V, Fearon D, Merad polarization of tumor-associated macrophages. Cancer Immunol
M. Understanding the tumor immuune microenvironment (TIME) Res. 2017;5(7):547–59.
for effective therapy. Nat Med. 2018;24(5):541–50. 49. Liang X, Guo W, Ren T, Huang Y, Sun K, Zhang H. Macrophages
45. Quail D, Joyce J. Micorenvironmental regulation of tumor progres- reduce the sensitivity of osteosarcoma to neoadjuvant chemotherapy
sion and metastasis. Nat Med. 2013;19(11):1423–37. drugs by secreting Interleukin-1 beta. Cancer Lett. 2020;480:4–14.
46. Segaliny A, Mohamadi A, Dizier B, Lokajczyk A, Brion R, Lanel 50. Zhang H, Maric I, DiPrima M, et al. Fibrocytes represent a novel
R. Interleukin-34 promotes tumor progression and metastatic pro- MDSC subset circulating in patients with metastatic cancer. Blood.
cess in osteosarcoma through induction of angiogenesis and mac- 2013;122(7):1105–13.
rophage recruitment. Int J Cancer. 2015;137(1):73–85. 51. Morales E, Olson M, Iglesias F, Dahiya S, Luetkens T, Atanackovic
47. Cassetta L, Pollard J. Targeting macrophages: therapeutic D. Role of immunotherapy in Ewing sarcoma. J Immunother
approaches in cancer. Nat Rev Drug Discov. 2018;17(12):887–904. Cancer. 2020;8:e000653.
48. Zhou Q, Xian M, Xiang S, Xiang D, Shao X, Wang J. All-trans
retinoic acid prevents osteosarcoma metastasis by inhibiting M2
 Liposarcoma Metastasis
58
Luke V. Selby, Raphael Pollock, and Valerie Grignol

Abstract
Learning Objectives
Soft tissue sarcomas are rare tumors originating from
mesodermal tissues (fat, muscle connective tissue, and ves- • Liposarcoma is the most common subtype of soft
sels). In-depth molecular analysis has identified at least 50 tissue sarcomas.
subtypes, which are often grouped together into larger clas- • Liposarcoma exists along a spectrum of biological
sifications. Liposarcoma, the most common subtype, activity related to subtype and histological grade.
accounts for at least 20% of all sarcoma and occurs with a • Dedifferentiated liposarcoma is a locally aggressive
peak incidence of 50–65 years of age without gender predi- histology with metastatic potential.
lection. Liposarcoma itself has four subtypes, which occur • Well-differentiated liposarcoma and dedifferenti-
on a spectrum of biological activity and metastatic poten- ated liposarcoma secrete biologically active extra-
tial, though the predominant cause of sarcoma-specific cellular vesicles that influence the tumor
death is locoregional recurrence. Well-­differentiated lipo- microenvironment and the premetastatic niche to
sarcoma (WDLPS) is locally aggressive but incapable of make both more hospitable to tumor growth.
metastasis. Dedifferentiated liposarcoma, WDLPS’s more
aggressive counterpart, is locally aggressive and also capa-
ble of distant metastasis, primarily to the lung. Myxoid-
round cell liposarcoma is a single histology whose 58.1 Introduction
metastatic potential (to the spine, axillary fat pads, and the
retroperitoneal fat) occurs on a spectrum. Pleomorphic Soft tissue sarcomas (STSs) are a rare group of locally aggres-
liposarcoma is the most aggressive liposarcoma subtype. sive and potentially metastatic tumors. STSs are derived from
Pleomorphic liposarcoma metastasizes early and has a very tissues of mesodermal origin (fat, muscle, connective tissue,
poor prognosis. Ongoing research suggests metastasis of vessels) and depending on histology are either locally aggres-
well-differentiated and dedifferentiated liposarcoma may sive, metastatic, or both. Recent molecular classifications have
be mediated by extracellular vesicles, which contain active identified at least 50 different histological subtypes (WHO
genetic material and condition the local tumor microenvi- Classification of Tumours, 5th Edition, Volume 3. Soft Tissue
ronment and the distant premetastatic niche to support and Bone Tumours [1]), based primarily on the underlying
tumor growth. Further work is needed to further define the genetic mutations. Given their rarity, varied anatomic loca-
role of these extracellular vesicles and develop therapeutic tions, and the difficulty in making pathologic diagnoses, surgi-
strategies to interrupt this cascade. cal pathologists without significant experience making these
diagnoses should consult with experienced centers as neces-
sary [2, 3]. Benign tumors of these same tissue types, such as
lipomas, are at least 100× more common than their malignant
counterparts [4]. Malignant STSs are treated with a combina-
tion of surgery, radiation therapy, and systemic therapy, based
L. V. Selby · R. Pollock · V. Grignol (*)
on the tumor’s anatomic location, its histological subtype, and
Division of Surgical Oncology, Department of Surgery, The Ohio
State University Wexner Medical Center and James Cancer its biological behavior. With the development of effective sys-
Hospital, Columbus, OH, USA temic chemotherapy and immunotherapy, the role of systemic
e-mail: [email protected] therapy in these diseases is increasing [5–7]. Unfortunately,

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 611
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_58
612 L. V. Selby et al.

approximately 40% of patients with STS die from complica- 58.2 Liposarcoma
tions associated with local recurrence or distant metastatic
disease. Liposarcoma, the most common STS subtype, accounts for
Broadly, STSs are characterized by their size, location, and at least 20% of all STSs. Four distinct liposarcoma subtypes,
histological grade. Size is broken down into increments of based on tumor biology and driver mutations, have been
5 cm, with tumors <5 cm considered small, 5–10 cm consid- described. Well-differentiated (WDLPS) liposarcoma and
ered intermediate, and >10 cm considered large. Though his- dedifferentiated (DDLPS) liposarcoma exist along a spec-
tological subtype influences biological activity, STSs are trum of biological activity. WDLPS and DDLPS uniformly
broadly characterized into low-grade, intermediate-grade, and have an amplification of chromosome 12q13-15, which
high-grade tumors. Grade is a histological determination that involves MDM2, CDK4, and HMGA2, with other amplifica-
tracks a tumor’s biological activity and aggressiveness. Low- tions identified less commonly [10, 11]. Overexpression of
grade tumors are slow growing, and while they can be locally MDM2 DNA from this amplification leads to excess MDM2
aggressive and recurrent, they have very low risk of metastatic protein expression, which then overrides the patient’s wild-­
disease. High-grade tumors, on the other hand, are faster type p53 tumor suppression function [12]. WDLPS is a
growing and frequently metastasize. In addition to the risks of locally aggressive but non-metastatic tumor that occurs
local recurrence, large high-grade tumors can have rates of within the proximal extremities and the retroperitoneum.
metastasis as high as 50%. Unfortunately, liposarcoma in gen- WDLPS that occurs within the extremity is cytologically and
eral has a poor response to systemic chemotherapy [6, 8]. histologically identical to the retroperitoneal WDLPS but
With the exception of epithelioid sarcoma, rhabdomyo- due to much different biological behavior is often referred to
sarcoma, clear cell sarcoma, angiosarcoma, and some less as an atypical lipomatous tumor (ALT) (WHO Classification
frequent histologies, STS generally doesn’t metastasize to of Tumours, 5th Edition, Volume 3. Soft Tissue and Bone
the regional lymph nodes [9]. While metastatic location is Tumours). Patients with WDLPS or ALT present with a pain-
dependent on the particular tumor location and histological less slow-growing mass of the extremity or slowly increasing
type, sarcoma most commonly metastasizes to the lung, abdominal girth in the case of a retroperitoneal WDLPS
though myxoid-round cell liposarcoma also metastasizes to (Fig. 58.1). Retrospectively identified gastrointestinal symp-
the fat pads and bones. toms in patients with retroperitoneal tumors were often sub-

Fig. 58.1 Selected axial and coronal images from a CT scan showing a large retroperitoneal dedifferentiated liposarcoma. The patient underwent
resection of the mass without resection of adjacent organs
58 Liposarcoma Metastasis 613

tle and non-diagnostic. WDLPS of the retroperitoneum can to have lymphatic metastasis, it has traditionally been
be incidentally diagnosed on cross-sectional imaging or dur- thought that liposarcoma achieved distant metastasis via
ing the workup for an hernia if the mass has protruded hematogenous, not lymphatic, routes. Recent work from the
through a defect in the abdominal wall [13]. NCDB, however, suggests that the rate of lymph node metas-
WDLPS and DDLPS occur on a continuum of differen- tasis in liposarcoma, regardless of anatomic site of origin, is
tiation and in total account for approximately 60% of all approximately 2.5% [28]. Given the complexity of making a
liposarcomas. In combination, WDLPS/DDLPS is the post diagnosis in liposarcoma [2], and the inherent heterogeneity
common liposarcoma histology. DDLPS is identified when of large databases like the NCDB, it is hard to know if the
a tumor contains a region of non-lipogenic sarcoma of at patients with a lymph node metastasis were diagnosed with
least several millimeters in diameter. The transition from the correct histology. It was initially hypothesized that local
WDLPS to DDLPS is marked by invasion into, instead of recurrence following resection of retroperitoneal liposar-
displacement of, adjacent structures and the ability to coma was the result of incomplete tumor resection and sub-
develop distant metastasis. Radiologically, DDLPS shows sequent tumor regrowth. More recently, multiple institutional
both fatty and non-fatty components, while WDLPS series have suggested that retroperitoneal sarcoma often
appears as a mass of entirely fatty components [14]. occurs in the setting of a field defect within the retroperito-
DDLPS is more locally aggressive than WDLPS and is fre- neal fat, placing the entire retroperitoneal compartment at
quently diagnosed on re-­resection of a locally recurrent risk of tumor development [29, 30]. However, emerging evi-
WDLPS. Though it has less metastatic potential than other dence suggests that tumor produced extracellular vesicles
high-grade pleomorphic sarcomas, DDLPS has a high risk that contribute to tumor regrowth by conditioning the tumor
of distant metastasis, with series reporting distant recur- microenvironment.
rent rates of 20–30% [15].
Historically, myxoid liposarcoma and round-cell liposar-
coma were thought of as two distinct tumor types [16]; they 58.3 Extracellular Vesicles
are now recognized as a single entity with differing clinical
behaviors [17] and account for approximately 30–40% of all Extracellular vesicles (EVs) are membrane-bound vesicles
liposarcomas. They most commonly occur deep in the that contain cellular cytosol and genetic material and are
extremities, with at least 66% occurring in the thigh. More enclosed in a lipid bilayer [31]. Vesicles between 30 and
than 90% of all myxoid-round cell tumors have a character- 150 nm in size are classified as exosomes; those >150 nm in
istic t(12;16) (q13-14;p11) translocation [18]. Tumors with size are called microvesicles. Exosomes are produced within
high histological grade were previously classified as round-­ specific compartments of the cell for release and are marked
cell liposarcomas; they are defined as having a >5% round by Tsg101, Alix, CD9, CD63, and CD81 markers, while
cell component and have a much worse prognosis than the microvesicles are formed by cell budding and CD59, CD47,
pure myxoid subtype, which has 0–5% round cell compo- and CD55 [31–33]. Initially thought to contain only cell
nent [19]. High-grade tumors have a 50% 5-year survival, waste, EVs are known to contain mRNA and microRNA in
compared to the >90% 5-year survival of low-grade tumors. addition to other bioactive components including protein,
Like all liposarcomas, myxoid-round cell tumors metasta- lipids, nucleic acids, and various metabolites [34, 35].
size to the lungs. However, they also metastasize to the spine,
the axillary fat pads, and the retroperitoneal fat, even in the
absence of pulmonary metastasis [20–22]. Consequently, 58.3.1 EVs and the Tumor Microenvironment
their metastatic workup is more extensive and included CT
scan of the primary site, chest, abdomen, and pelvis and Since the demonstration that EVs can transfer their contents
finally MRI of the spine. Myxoid-round cell liposarcoma is to recipient cells, their role in cancer development has gar-
exquisitely sensitive to radiation therapy, ifosfamide, and tra- nered great attention (Fig. 58.2). Non-neoplastic cells near a
bectedin [23–25]. solid tumor make up the tumor microenvironment (TME),
Pleomorphic liposarcomas are high grade, highly malig- and the interaction between these stromal cells and the pri-
nant, and account for <5% of all liposarcomas [26]. They have mary tumor itself is essential to tumor survival and metasta-
high mitotic activity and due to their aggressiveness often have sis. Tumor-stromal communication occurs via direct cell-cell
internal hemorrhage and necrosis. These aggressive tumors interaction as well as indirectly via paracrine signaling
metastasize to the lungs early [27], and despite sensitivity to involving growth factors, chemokines, and proteases and
gemcitabine and ifosfamide-based chemotherapy regimens, also via EVs [36]. EVs, containing active RNA, can influ-
pleomorphic liposarcomas have high mortality rates. ence the tumor microenvironment to make the environment
Though synovial sarcoma, rhabdomyosarcoma, clear cell supportive of tumor growth and survival. EVs are known to
sarcoma, epithelioid sarcoma, and angiosarcoma are known shift fibroblasts and mesenchymal stem cells into a myofi-
614 L. V. Selby et al.

Fig. 58.2 Local and distant

effects of liposarcoma-derived
extracellular vesicles

MDM2

Extracellular
Vesicles
Liposarcoma
Tumor Tissues and
organs in the body

CAF differentiation Blood vessels

Angiogenesi
ECM degradation
Neutrophils
mobilization

broblast phenotype, which facilitates angiogenesis and The liposarcoma TME contains abundant preadipocytes,
metastasis [37]. Tumor angiogenesis is supported when EVs adipocytes, macrophages, inflammatory cells, endothelial
transfer specific signals and increase VEGF signaling [38, cells, and fibroblasts [45]. Importantly, trabectedin, a medi-
39]. As cells in the TME are transformed by EVs, these cells cation with activity against myxoid-round cell liposarcoma,
themselves then contribute to the continued development of acts against the tumor itself and also against the
the TME. EVs from regulatory T lymphocytes (Tregs) have TME. Trabectedin forms covalent bonds with the minor
been shown to transfer miRNA containing exosomes to other grove of the DNA double helix [46, 47]. As a result, trabect-
immune cells, suppressing the anticancer Th1 phenotype by edin modulates cytokine and chemokine production and dis-
changing cell proliferation and cytokine secretion [40, 41]. places oncogenic transcription factors from their targeted
DDLPS EVs contain DNA and RNA important in the promoters in tumor cells and cells in the TME. Trabectedin’s
development of malignancy and metastasis. miR-25-3p and anti-TME activity seems to come from its activity against
miR-92a-3p regulate their target mRNAs during both tran- tumor-associated macrophages (TAMs) which as quiescent
scription and translation, increasing NF-kB-mediated IL-6 cell are not targeted by traditional cytotoxic chemotherapies
secretion by TME macrophages. Cultured DDLPS cells [48]. TAMs are monocytes recruited from the circulating
exposed to macrophage condition medium enriched in IL-6 blood and activated into a TAM phenotype by the tumor and
show increased cell proliferation and invasion [42–44]. TME cells through the various mechanisms discussed above
DDLPS-derived EVs are known to contain amplified copies [49]. Importantly, trabectedin has activity against slow-­
of MDM2. MDM2 amplification is a core component of growing tumors, but to this point it has shown activity mostly
WDLPS and DDLPS and results in functional inactivation of for myxoid-round cell liposarcoma.
the p53 tumor suppressor protein. Transfer of MDM2 DNA
to preadipocytes in the TME leads to overexpression of
MDM2 in cells without genomic alterations in MDM2 copy 58.3.2 EVs and the Premetastatic Niche
number. These cells, genotypically normal preadipocytes
from the tumor microenvironment, are also known to pro- Just as EVs can condition the tumor microenvironment to sup-
duce excess levels of metalloprotease-2 (MMP2), an extra- port local tumor growth, so can they condition distant organ
cellular collagenase that is vital to metastasis formation [43]. sites to transform into a premetastatic nice (PMN). Across
Taken together, this suggests that EVs released by DDLPS tumor types, EVs have been shown to condition the PMN to
increase IL-6-mediated cell proliferation and invasion via allow metastasis survival. Breast carcinoma-derived EVs con-
miRNA-specific pathways, but they also inactivate wt p53 tribute to immunosuppression by pushing myeloid cells
and increase MMP2 production (which in turns also facili- towards a myeloid-derived suppressor cell phenotype [50],
tates cell migration and invasion) via an MDM2-specific while melanoma-derived EVs impair dendritic cell maturation
pathway. in the PMN and shift them towards an immunosuppressive
58 Liposarcoma Metastasis 615

phenotype [51]. Exosomes have been shown to stimulate Postoperative surveillance in retroperitoneal liposarcoma
MMP2 [43] and MMP9 [52] production in the PMN, facilitat- consists of physical exam and CT scans of the abdomen and
ing metastasis invasion and angiogenesis, which is also aided pelvis, with the chest added for grade II and III liposarcomas
by EV upregulation of VEGF [53]. In total, just as tumor- (DDLPS). In addition to this screening, patients with
derived EVs exert their influence on the TME to allow tumor myxoid-­round cells liposarcoma should undergo CT scans of
survival, through the same mechanisms they affect the distant the primary tumor site, chest, abdomen, and pelvis, as well
PMN and push it towards supporting metastasis [54]. as an MRI of the spine to screen for metastasis to the axillary
fat pads and retroperitoneum. MRI is exquisitely sensitive at
detecting extrapulmonary metastasis from myxoid-round
58.4 Surgery and Postoperative cell liposarcoma [63]. In the future, it is possible that periph-
Surveillance erally detected EVs will be used as biomarkers of disease in
WDLPS and DDLPS. Changes in EV levels, either decreased
Surgery, the gold standard of treatment for liposarcoma, is EV levels in response to systemic therapy or rising EV levels
often combined with either systemic therapy or radiation during surveillance, may allow patients to be more closely
therapy in disease-specific settings [55]. The goal of curative-­ monitored in between cross-sectional imaging.
intent surgery is function-preserving complete gross resec-
tion. In the retroperitoneum, patients who undergo R2
resection have worse outcome than those undergoing either 58.5 Conclusion
R0 or R1 resection. It is unclear if there is differential sur-
vival between those who undergo R0 or R1 resection [56]. Liposarcoma, the most common subtype of soft tissue sar-
Significant controversy exists surrounding the importance of coma, is made up of four distinct subtypes with unique tumor
extended compartmental resection to achieve a gross nega- biology. Liposarcoma metastasis is hematologic, not lym-
tive margin of at least 1 cm in the retroperitoneum. Patients phatic, and is a poorly understood phenomenon. Liposarcoma
undergoing compartmental resection have lower rates of predominantly metastasizes to the lungs, while myxoid-­
local recurrence, and may have improved overall survival, round cell liposarcoma also possesses the ability to metasta-
but higher rates of postoperative complications [56–58]. size to the retroperitoneal and axillary fat pads. WDLPS
Additionally, because large tumors lie as close to “unresect- local recurrence and DDLPS metastasis appear to be medi-
able” organs (the central vessels, liver, or duodenum) as they ated by tumor-derived extracellular vesicles that influence
lie to the “resectable” organs (a single kidney, small intes- the stroma of the TME and PMN to undergo a series of pre-
tine, large intestine), compartmental resections are generally metastasis changes that improve tumor survival. Further
used selectively when gross negative margins can be achieved work is required to more fully define the metastatic steps,
with an organ-sparing approach at the initial operation [57, and therapeutic interventions to halt metastasis, in these rare
59]. Rapid local recurrence and invasion into, instead of tumors.
abutment of, adjacent structures may reflect tumor biology,
obviating the benefit of aggressive local resection [60].
Additionally, even in the metastatic setting, most DDLPS
Perspective
death (and all WDLPS death) is the result of locoregional
recurrence, not distant metastatic disease.
• Is liposarcoma metastasis hematologic or
The principles of resection of extremity myxoid-round
lymphatic?
cell liposarcomas are identical to those of retroperitoneal
• Which liposarcoma histologies are prone to
tumors: function-sparing complete resection with gross neg-
metastasis?
ative margins. Rates of local recurrence for extremity sar-
• What role do liposarcoma-derived extracellular
coma are much lower than for retroperitoneal sarcoma [61,
vesicles play in local recurrence and metastasis?
62]. High-grade myxoid-round cell liposarcoma is extremely
• What medications can be used to affect the sarcoma
sensitive to radiation therapy. The decision to administer
tumor microenvironment?
radiation therapy in the neoadjuvant or adjuvant setting or
• What are the current first-line systemic therapies for
upon local recurrence is based on the location within the
liposarcoma?
thigh and weighing the risks and benefits of the sequencing
of these therapies.
616 L. V. Selby et al.

References 16. Orvieto E, Furlanetto A, Laurino L, Dei Tos AP. Myxoid and round
cell liposarcoma: a spectrum of myxoid adipocytic neoplasia.
Semin Diagn Pathol. 2001;18(4):267–73.
1. WHO Classification of Tumours, 5th Edition, Volume 3. Soft Tissue
17. Antonescu CR, Elahi A, Humphrey M, Lui MY, Healey JH,
and Bone Tumours; (n.d.).
Brennan MF, Woodruff JM, Jhanwar SC, Ladanyi M. Specificity of
2. Riddle ND, Gonzalez RJ, Bridge JA, Antonia S, Bui MM. A
TLS-CHOP rearrangement for classic myxoid/round cell liposar-
CD117 and CD34 immunoreactive sarcoma masquerading as a
coma: absence in predominantly myxoid well-differentiated lipo-
gastrointestinal stromal tumor: diagnostic pitfalls of ancillary stud-
sarcomas. J Mol Diagn. 2000;2(3):132–8. https://doi.org/10.1016/
ies in sarcoma. Cancer Control. 2011;18(3):152–9. https://doi.
s1525-­1578(10)60628-­9.
org/10.1177/107327481101800302.
18. Knight JC, Renwick PJ, Dal Cin P, Van den Berghe H, Fletcher
3. Thway K, Wang J, Mubako T, Fisher C. Histopathological diagnos-
CD. Translocation t(12;16)(q13;p11) in myxoid liposarcoma and
tic discrepancies in soft tissue tumours referred to a specialist centre:
round cell liposarcoma: molecular and cytogenetic analysis. Cancer
reassessment in the era of ancillary molecular diagnosis. Sarcoma.
Res. 1995;55(1):24–7.
2014;2014:686902. https://doi.org/10.1155/2014/686902.
19. Fiore M, Grosso F, Lo Vullo S, Pennacchioli E, Stacchiotti S, Ferrari
4. Rydholm A, Berg NO. Size, site and clinical incidence of
A, Collini P, Lozza L, Mariani L, Casali PG, Gronchi A. Myxoid/
lipoma. Factors in the differential diagnosis of lipoma and sar-
round cell and pleomorphic liposarcomas: prognostic factors and
coma. Acta Orthop Scand. 1983;54(6):929–34. https://doi.
survival in a series of patients treated at a single institution. Cancer.
org/10.3109/17453678308992936.
2007;109(12):2522–31. https://doi.org/10.1002/cncr.22720.
5. Bill KL, Casadei L, Prudner BC, Iwenofu H, Strohecker AM,
20. Dürr HR, Rauh J, Baur-Melnyk A, Knösel T, Lindner L, Roeder F,
Pollock RE. Liposarcoma: molecular targets and therapeutic
Jansson V, Klein A. Myxoid liposarcoma: local relapse and meta-
implications. Cell Mol Life Sci. 2016;73(19):3711–8. https://doi.
static pattern in 43 patients. BMC Cancer. 2018;18(1):304. https://
org/10.1007/s00018-­016-­2266-­2.
doi.org/10.1186/s12885-­018-­4226-­8.
6. Liebner DA. The indications and efficacy of conventional che-
21. Fuglø HM, Maretty-Nielsen K, Hovgaard D, Keller J, Safwat
motherapy in primary and recurrent sarcoma. J Surg Oncol.
AA, Petersen MM. Metastatic pattern, local relapse, and sur-
2015;111(5):622–31. https://doi.org/10.1002/jso.23866.
vival of patients with myxoid liposarcoma: a retrospective
7. McGovern Y, Zhou CD, Jones RL. Systemic therapy in metastatic
study of 45 patients. Sarcoma. 2013;2013:548628. https://doi.
or unresectable well-differentiated/dedifferentiated liposarcoma.
org/10.1155/2013/548628.
Front Oncol. 2017;7(292) https://doi.org/10.3389/fonc.2017.00292.
22. Spillane AJ, Fisher C, Thomas JM. Myxoid liposarcoma – the fre-
8. Anaya DA, Lahat G, Wang X, Xiao L, Tuvin D, Pisters PW, Lev
quency and the natural history of nonpulmonary soft tissue metas-
DC, Pollock RE. Establishing prognosis in retroperitoneal sarcoma:
tases. Ann Surg Oncol. 1999;6(4):389–94. https://doi.org/10.1007/
a new histology-based paradigm. Ann Surg Oncol. 2009;16(3):667–
s10434-­999-­0389-­5.
75. https://doi.org/10.1245/s10434-­008-­0250-­2.
23. Chung PW, Deheshi BM, Ferguson PC, Wunder JS, Griffin
9. Fong Y, Coit DG, Woodruff JM, Brennan MF. Lymph node metastasis
AM, Catton CN, Bell RS, White LM, Kandel RA, O’Sullivan
from soft tissue sarcoma in adults. Analysis of data from a prospec-
B. Radiosensitivity translates into excellent local control in extremity
tive database of 1772 sarcoma patients. Ann Surg. 1993;217(1):72–
myxoid liposarcoma: a comparison with other soft tissue sarcomas.
7. https://doi.org/10.1097/00000658-­199301000-­00012.
Cancer. 2009;115(14):3254–61. https://doi.org/10.1002/cncr.24375.
10. Gisselsson D, Höglund M, Mertens F, Johansson B, Dal Cin P, Van
24. Crago AM, Dickson MA. Liposarcoma: multimodality man-
den Berghe H, Earnshaw WC, Mitelman F, Mandahl N. The struc-
agement and future targeted therapies. Surg Oncol Clin N Am.
ture and dynamics of ring chromosomes in human neoplastic and
2016;25(4):761–73. https://doi.org/10.1016/j.soc.2016.05.007.
non-neoplastic cells. Hum Genet. 1999;104(4):315–25. https://doi.
25. Suarez-Kelly LP, Baldi GG, Gronchi A. Pharmacotherapy for lipo-
org/10.1007/s004390050960.
sarcoma: current state of the art and emerging systemic treatments.
11. Szymanska J, Tarkkanen M, Wiklund T, Virolainen M,
Expert Opin Pharmacother. 2019;20(12):1503–15. https://doi.org/1
Blomqvist C, Asko-Seljavaara S, Tukiainen E, Elomaa I,
0.1080/14656566.2019.1618271.
Knuutila S. Gains and losses of DNA sequences in liposarco-
26. Gebhard S, Coindre JM, Michels JJ, Terrier P, Bertrand G, Trassard M,
mas evaluated by comparative genomic hybridization. Genes
Taylor S, Château MC, Marquès B, Picot V, Guillou L. Pleomorphic
Chromosomes Cancer. 1996;15(2):89–94. https://doi.org/10.1002/
liposarcoma: clinicopathologic, immunohistochemical, and follow-­up
(sici)1098-­2264(199602)15:2<89::Aid-­gcc2>3.0.Co;2-­#.
analysis of 63 cases: a study from the French Federation of Cancer
12. Sirvent N, Coindre JM, Maire G, Hostein I, Keslair F, Guillou L,
Centers Sarcoma Group. Am J Surg Pathol. 2002;26(5):601–16.
Ranchere-Vince D, Terrier P, Pedeutour F. Detection of MDM2-­
https://doi.org/10.1097/00000478-­200205000-­00006.
CDK4 amplification by fluorescence in situ hybridization in 200
27. Downes KA, Goldblum JR, Montgomery EA, Fisher C. Pleomorphic
paraffin-embedded tumor samples: utility in diagnosing adipocytic
liposarcoma: a clinicopathologic analysis of 19 cases. Mod Pathol.
lesions and comparison with immunohistochemistry and real-­
2001;14(3):179–84. https://doi.org/10.1038/modpathol.3880280.
time PCR. Am J Surg Pathol. 2007;31(10):1476–89. https://doi.
28. Keung EZ, Chiang YJ, Voss RK, Cormier JN, Torres KE, Hunt
org/10.1097/PAS.0b013e3180581fff.
KK, Feig BW, Roland CL. Defining the incidence and clini-
13. Montgomery E, Fisher C. Paratesticular liposarcoma: a clinico-
cal significance of lymph node metastasis in soft tissue sarcoma.
pathologic study. Am J Surg Pathol. 2003;27(1):40–7. https://doi.
Eur J Surg Oncol. 2018;44(1):170–7. https://doi.org/10.1016/j.
org/10.1097/00000478-­200301000-­00005.
ejso.2017.11.014.
14. Messiou C, Morosi C. Imaging in retroperitoneal soft tissue sarcoma.
29. Neuhaus SJ, Barry P, Clark MA, Hayes AJ, Fisher C, Thomas
J Surg Oncol. 2018;117(1):25–32. https://doi.org/10.1002/jso.24891.
JM. Surgical management of primary and recurrent retroperi-
15. Tseng WW, Chopra S, Engleman EG, Pollock RE. Hypothesis:
toneal liposarcoma. Br J Surg. 2005;92(2):246–52. https://doi.
The intratumoral immune response against a cancer progenitor cell
org/10.1002/bjs.4802.
impacts the development of well-differentiated versus dedifferenti-
30. Tseng WW, Madewell JE, Wei W, Somaiah N, Lazar AJ, Ghadimi
ated disease in liposarcoma. Front Oncol. 2016;6:134. https://doi.
MP, Hoffman A, Pisters PW, Lev DC, Pollock RE. Locoregional
org/10.3389/fonc.2016.00134.
58 Liposarcoma Metastasis 617

disease patterns in well-differentiated and dedifferentiated retro- late liposarcoma progression. Cancer Res. 2017;77(14):3846–56.
peritoneal liposarcoma: implications for the extent of resection? https://doi.org/10.1158/0008-­5472.Can-­16-­2984.
Ann Surg Oncol. 2014;21(7):2136–43. https://doi.org/10.1245/ 46. D’Incalci M, Badri N, Galmarini CM, Allavena P. Trabectedin,
s10434-­014-­3643-­4. a drug acting on both cancer cells and the tumour microenviron-
31. Colombo M, Raposo G, Théry C. Biogenesis, secretion, and inter- ment. Br J Cancer. 2014;111(4):646–50. https://doi.org/10.1038/
cellular interactions of exosomes and other extracellular vesicles. bjc.2014.149.
Annu Rev Cell Dev Biol. 2014;30:255–89. https://doi.org/10.1146/ 47. D’Incalci M, Galmarini CM. A review of trabectedin (ET-743): a
annurev-­cellbio-­101512-­122326. unique mechanism of action. Mol Cancer Ther. 2010;9(8):2157–63.
32. Gould SJ, Raposo G. As we wait: coping with an imperfect nomen- https://doi.org/10.1158/1535-­7163.Mct-­10-­0263.
clature for extracellular vesicles. J Extracell Vesicles. 2013;2 48. Galmarini CM, D’Incalci M, Allavena P. Trabectedin and plitidep-
https://doi.org/10.3402/jev.v2i0.20389. sin: drugs from the sea that strike the tumor microenvironment. Mar
33. Simons M, Raposo G. Exosomes – vesicular carriers for intercel- Drugs. 2014;12(2):719–33. https://doi.org/10.3390/md12020719.
lular communication. Curr Opin Cell Biol. 2009;21(4):575–81. 49. Mantovani A, Sozzani S, Locati M, Allavena P, Sica
https://doi.org/10.1016/j.ceb.2009.03.007. A. Macrophage polarization: tumor-associated macrophages
34. Valadi H, Ekström K, Bossios A, Sjöstrand M, Lee JJ, Lötvall as a paradigm for polarized M2 mononuclear phagocytes.
JO. Exosome-mediated transfer of mRNAs and microRNAs is a Trends Immunol. 2002;23(11):549–55. https://doi.org/10.1016/
novel mechanism of genetic exchange between cells. Nat Cell Biol. s1471-­4906(02)02302-­5.
2007;9(6):654–9. https://doi.org/10.1038/ncb1596. 50. Xiang X, Poliakov A, Liu C, Liu Y, Deng ZB, Wang J, Cheng
35. Waldenström A, Gennebäck N, Hellman U, Ronquist Z, Shah SV, Wang GJ, Zhang L, Grizzle WE, Mobley J, Zhang
G. Cardiomyocyte microvesicles contain DNA/RNA and convey HG. Induction of myeloid-derived suppressor cells by tumor
biological messages to target cells. PLoS One. 2012;7(4):e34653. exosomes. Int J Cancer. 2009;124(11):2621–33. https://doi.
https://doi.org/10.1371/journal.pone.0034653. org/10.1002/ijc.24249.
36. Roma-Rodrigues C, Mendes R, Baptista PV, Fernandes 51. Valenti R, Huber V, Filipazzi P, Pilla L, Sovena G, Villa A,
AR. Targeting tumor microenvironment for cancer therapy. Int J Corbelli A, Fais S, Parmiani G, Rivoltini L. Human tumor-released
Mol Sci. 2019;20(4) https://doi.org/10.3390/ijms20040840. microvesicles promote the differentiation of myeloid cells with
37. Zhang X, Yuan X, Shi H, Wu L, Qian H, Xu W. Exosomes in can- transforming growth factor-beta-mediated suppressive activity
cer: small particle, big player. J Hematol Oncol. 2015;8:83. https:// on T lymphocytes. Cancer Res. 2006;66(18):9290–8. https://doi.
doi.org/10.1186/s13045-­015-­0181-­x. org/10.1158/0008-­5472.Can-­06-­1819.
38. Al-Nedawi K, Meehan B, Micallef J, Lhotak V, May L, Guha A, 52. Hiratsuka S, Nakamura K, Iwai S, Murakami M, Itoh T, Kijima
Rak J. Intercellular transfer of the oncogenic receptor EGFRvIII H, Shipley JM, Senior RM, Shibuya M. MMP9 induction by
by microvesicles derived from tumour cells. Nat Cell Biol. vascular endothelial growth factor receptor-1 is involved in lung-­
2008;10(5):619–24. https://doi.org/10.1038/ncb1725. specific metastasis. Cancer Cell. 2002;2(4):289–300. https://doi.
39. Al-Nedawi K, Meehan B, Rak J. Microvesicles: messengers and org/10.1016/s1535-­6108(02)00153-­8.
mediators of tumor progression. Cell Cycle. 2009;8(13):2014–8. 53. Grange C, Tapparo M, Collino F, Vitillo L, Damasco C, Deregibus
https://doi.org/10.4161/cc.8.13.8988. MC, Tetta C, Bussolati B, Camussi G. Microvesicles released from
40. Okoye IS, Coomes SM, Pelly VS, Czieso S, Papayannopoulos V, human renal cancer stem cells stimulate angiogenesis and forma-
Tolmachova T, Seabra MC, Wilson MS. MicroRNA-containing tion of lung premetastatic niche. Cancer Res. 2011;71(15):5346–
T-regulatory-cell-derived exosomes suppress pathogenic T helper 56. https://doi.org/10.1158/0008-­5472.Can-­11-­0241.
1 cells. Immunity. 2014;41(3):503. https://doi.org/10.1016/j. 54. Lobb RJ, Lima LG, Möller A. Exosomes: Key mediators of metas-
immuni.2014.08.008. tasis and pre-metastatic niche formation. Semin Cell Dev Biol.
41. Smyth LA, Ratnasothy K, Tsang JY, Boardman D, Warley A, 2017;67:3–10. https://doi.org/10.1016/j.semcdb.2017.01.004.
Lechler R, Lombardi G. CD73 expression on extracellular vesicles 55. Mansfield SA, Pollock RE, Grignol VP. Surgery for abdominal
derived from CD4+ CD25+ Foxp3+ T cells contributes to their well-differentiated liposarcoma. Curr Treat Options in Oncol.
regulatory function. Eur J Immunol. 2013;43(9):2430–40. https:// 2018;19(1):1. https://doi.org/10.1007/s11864-­018-­0520-­6.
doi.org/10.1002/eji.201242909. 56. Bonvalot S, Rivoire M, Castaing M, Stoeckle E, Le Cesne A, Blay
42. Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and JY, Laplanche A. Primary retroperitoneal sarcomas: a multivari-
function. Cell. 2004;116(2):281–97. https://doi.org/10.1016/ ate analysis of surgical factors associated with local control. J Clin
s0092-­8674(04)00045-­5. Oncol. 2009;27(1):31–7. https://doi.org/10.1200/jco.2008.18.0802.
43. Casadei L, Calore F, Braggio DA, Zewdu A, Deshmukh AA, 57. Bonvalot S, Miceli R, Berselli M, Causeret S, Colombo C, Mariani
Fadda P, Lopez G, Wabitsch M, Song C, Leight JL, Grignol VP, L, Bouzaiene H, Le Péchoux C, Casali PG, Le Cesne A, Fiore M,
Lev D, Croce CM, Pollock RE. MDM2 derived from dedifferenti- Gronchi A. Aggressive surgery in retroperitoneal soft tissue sar-
ated liposarcoma extracellular vesicles induces MMP2 production coma carried out at high-volume centers is safe and is associated
from preadipocytes. Cancer Res. 2019;79(19):4911–22. https://doi. with improved local control. Ann Surg Oncol. 2010;17(6):1507–14.
org/10.1158/0008-­5472.Can-­19-­0203. https://doi.org/10.1245/s10434-­010-­1057-­5.
44. Denli AM, Tops BB, Plasterk RH, Ketting RF, Hannon 58. Gronchi A, Lo Vullo S, Fiore M, Mussi C, Stacchiotti S, Collini P,
GJ. Processing of primary microRNAs by the Microprocessor Lozza L, Pennacchioli E, Mariani L, Casali PG. Aggressive sur-
complex. Nature. 2004;432(7014):231–5. https://doi.org/10.1038/ gical policies in a retrospectively reviewed single-institution case
nature03049. series of retroperitoneal soft tissue sarcoma patients. J Clin Oncol.
45. Casadei L, Calore F, Creighton CJ, Guescini M, Batte K, Iwenofu 2009;27(1):24–30. https://doi.org/10.1200/jco.2008.17.8871.
OH, Zewdu A, Braggio DA, Bill KL, Fadda P, Lovat F, Lopez G, 59. Gronchi A, Pollock R. Surgery in retroperitoneal soft tissue sar-
Gasparini P, Chen JL, Kladney RD, Leone G, Lev D, Croce CM, coma: a call for a consensus between Europe and North America.
Pollock RE. Exosome-derived miR-25-3p and miR-92a-3p stimu- Ann Surg Oncol. 2011;18(8):2107–10. https://doi.org/10.1245/
s10434-­011-­1746-­8.
618 L. V. Selby et al.

60. Singer S, Antonescu CR, Riedel E, Brennan MF. Histologic subtype 62. Hoffman A, Ghadimi MP, Demicco EG, Creighton CJ, Torres K,
and margin of resection predict pattern of recurrence and survival Colombo C, Peng T, Lusby K, Ingram D, Hornick JL, Wang WL,
for retroperitoneal liposarcoma. Ann Surg. 2003;238(3):358– Ravi V, Lazar AJ, Lev D, Pollock RE. Localized and metastatic
70.; discussion 370–351. https://doi.org/10.1097/01. myxoid/round cell liposarcoma: clinical and molecular observations.
sla.0000086542.11899.38. Cancer. 2013;119(10):1868–77. https://doi.org/10.1002/cncr.27847.
61. Cahlon O, Brennan MF, Jia X, Qin LX, Singer S, Alektiar KM. A 63. Gorelik N, Reddy SMV, Turcotte RE, Goulding K, Jung S, Alcindor
postoperative nomogram for local recurrence risk in extremity T, Powell TI. Early detection of metastases using whole-body
soft tissue sarcomas after limb-sparing surgery without adjuvant MRI for initial staging and routine follow-up of myxoid liposar-
radiation. Ann Surg. 2012;255(2):343–7. https://doi.org/10.1097/ coma. Skelet Radiol. 2018;47(3):369–79. https://doi.org/10.1007/
SLA.0b013e3182367aa7. s00256-­017-­2845-­9.
 Desmoid Tumors
59
Gaya Spolverato and Alessandro Gronchi

Abstract 59.1 Introduction

Desmoid tumors (DT) are a rare disease with no metastatic
potential, characterized by a local aggressive monoclonal Desmoid tumors (DT) are a rare disease with no metastatic
fibroblastic proliferation. DT are characterized by a mutu- potential, characterized by a local aggressive monoclonal
ally exclusive CTNNB1 or APC mutation: CTNNB1 muta- fibroblastic proliferation. The incidence of DT ranges
tions are somatic and occur only in the sporadic form, between 5 and 6 cases per million inhabitants per annum. DT
while APC mutations are generally germline and associ- have a female predominance and are more common in the
ated with the familiar variant (Gardner syndrome). Surgery third and fourth life decades. Although DT do not spread dis-
has been the standard of care for many years; however, tantly, they can be locally aggressive, causing chronic pain,
recently we observed a paradigm shift toward a more con- functional deficit, and psychological problems, affecting the
servative management. We present herein the available evi- overall quality of life. Intestinal obstruction and perforation
dences on the management of sporadic and familiar DT, can also be a complication of intra-abdominal desmoid
with a particular focus on active surveillance, available tumors and life-threatening conditions, mostly in the case of
treatments, and assessment of response. multifocal disease and/or familial adenomatous polyposis
(FAP)-associated DT.
Desmoids occur in three locations:

1. Extra-abdominal sites: most commonly extremities, limb

Learning Objectives
girdles (mostly periscapular), thoracic wall, breast, head
• Epidemiology and presentation of desmoid tumors. and neck.
• Molecular signatures of desmoid tumors. 2. Abdominal wall: mostly in women, especially during or
• Indication for active surveillance of sporadic and after pregnancy.
familiar desmoids. 3. Intra-abdominal sites, such as the bowel wall and mesen-
• Indication for active treatment and treatment tery: often associated with FAP.
algorithm.
• Assessment of response to treatment and assess- Sporadic extra-abdominal desmoids can occasionally be
ment of tumor growth during observation. multifocal, but mostly confined to one limb. On the contrary
FAP-associated desmoids are more often multifocal, and
while they predominantly affect the mesentery and abdomi-
nal wall, they can be located in multiple extra-abdominal
G. Spolverato sites (Fig. 59.1).
Department of Surgical, Oncological and Gastroenterological
Unfortunately, the indolent behavior of desmoids is not
Sciences, University of Padova, Padova, Italy
supported by specific molecular signature, gene mutations,
A. Gronchi (*)
or microenvironment characteristics. No previous studies
Department of Surgery, Fondazione IRCCS Istituto Nazionale dei
Tumori, Milan, Italy investigated the metastatic behavior of desmoids and the bio-
e-mail: [email protected] logical basis of the absent metastatic potential.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 619
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_59
620 G. Spolverato and A. Gronchi

a b c

Fig. 59.1 Panel a, b: contrast-enhanced T1-weighted MRI of the right abdomen, axial view, showing masses (white arrows) in the abdominal
lower limb, coronal view, showing multiple nodules (white arrows) in wall (panel c) and mesentery (panel d) in a 31-year-old female with
the thigh (panel a) and leg (panel b) in a 26-year-old female with no Gardner syndrome
associated syndromes. Panel c, d: contrast-enhanced CT scan of the

59.2 Molecular Genetics report that S45F-mutated DT tend to have a higher recur-
rence rate after surgery, while others report conflicting results
DT are characterized by a mutually exclusive CTNNB1 or [5–10]. A recent meta-analysis evaluating the impact of
APC mutation. CTNNB1 mutations are somatic and occur CTNNB1 mutation type on recurrence and recurrence-free
only in the sporadic form. Occasionally also APC mutations survival in adult patients undergoing surgery for DT [4]
can be somatic and can occur in the sporadic form. However, found that S45F mutation has a higher risk of recurrence
they are generally germline and associated with the familiar compared to T41A, S45P, and wild-type DT, but the effect
variant (Gardner syndrome). Hence, if a somatic CTNNB1 is modification between tumor size and mutation type suggests
reported, a syndromic condition can be excluded, while in that the association between mutation and prognosis is medi-
case of CTNNB1 wild-type status, FAP should be investi- ated by tumor size.
gated with extensive clinical workup (including a colonos- The prognostic impact of APC mutations in FAP
copy or a germline test). CTNNB1 mutation determines a patients is unclear, while the effects of APC mutation on
disruption of Wnt/beta-catenin signaling, involved in cell-­ the risk of developing desmoids have been extensively
cell adhesion [1]. Several activating mutations of CTNNB1 investigated. Mutations between codons 1444 and 1578
are known, all of them in exon 3, causing the following were found to correlate with DT, but this observation was
amino acid changes: T41A, S45F, and S45P. In a minority of not compared to a reference population. In a following
patients, DT result from a germline or sporadic loss of APC, systematic review of the genotype-phenotype associa-
which is a negative regulator of beta-catenin stability. A tions of DT in patients with FAP, APC mutations between
germline mutation of the adenomatous polyposis coli gene codons 543–713 and 1310–2011 were found to confer an
falls in the context of FAP—Gardner syndrome. The muta- increased risk for desmoid tumors in individuals with
tion in the APC gene on chromosome 5 is known for the compared to a reference population of FAP patients with
development of hundreds of colonic polyps. DT, manifesting APC gene mutations [11].
mainly with intra-abdominal multifocal tumors, occur in Among the epigenetic factors associated with the risk
about 20% of FAP patients [2, 3]. of developing DT, the surgical injury of open colectomy
The effect of CTNNB1 mutation on long-term oncologi- was found to be an independent predictor of DT, and
cal outcome and in particular on recurrence after surgery has patients receiving laparotomic prophylactic colectomy
been extensively evaluated in sporadic DT [4]. Some studies were found to have a 5-year cumulative risk of develop-
59 Desmoid Tumors 621

ing DT threefold higher than those operated with a lapa- lar, Lev et al., comparing a large series of desmoid patients
roscopic approach [12]. treated at the University of Texas M.D. Anderson Cancer
In the era of active surveillance for sporadic and FAP-­ Center to a previously published series from the same insti-
related DT, it is even more important to define the impact of tution, showed that, even if the rate of macroscopic residual
mutation type on risk of progression in order to anticipate the disease was higher and the rate of positive microscopic mar-
tumor behavior and modulate the treatment plan. The effect gins was equivalent overtime, 5-year local recurrence rate
of APC or CTNNB1 mutation on tumor aggressiveness or improved from 30% to 20% [16]. Similarly, Crago et al.,
response to medical treatment is still unknown. The correla- reporting on a large series of DT patient treated at Memorial
tion between the efficacy of meloxicam treatment and status Sloan Kettering Cancer Center, confirmed that margin status
of CTNNB1 mutations was analyzed in a cohort of 33 was not associated with local recurrence-free survival and
patients with extraperitoneal sporadic DT, and S45F muta- excluded it from the prognostic nomogram [17].
tion was found to be associated with a poor response to treat- Active surveillance was investigated by different authors
ment [13]. Similarly, the same mutation was found to in retrospective studies reporting on DT in several locations.
correlate with response to imatinib, further supporting the The first report dates back in 2007, when Bonvalot et al.
idea of a potential predictive impact of the mutation status on reported on two groups of patients: those who underwent
DF treatment decision-making [14]. surgery with or without radiotherapy and those who were
managed with nonsurgical strategies, such as systemic treat-
ment alone or active surveillance. Patients with microscopi-
59.3 Active Surveillance cally negative margins resection had similar outcome
compared to patients who did not undergo surgery. Growth
59.3.1 Sporadic Desmoid arrest occurred in two thirds of non-operated patients, and
strategy (surgery versus no-surgery) was not correlated with
Surgery has historically been the primary treatment for the oncological outcome [18]. Similarly, Fiore et al., report-
patients with resectable DT. Surgical recommendations have ing on a series of patients who were initially treated with a
traditionally been based on retrospective studies in which frontline deliberately conservative policy, showed that half
surgery was offered when feasible; thus, the conservative of patients achieved medium-term stable disease. Perhaps
approach was neglected for a long time. A cutting-edge study more importantly patients managed with active surveillance
was published by Brennan et al. in 1999, comparing major had a progression-free survival similar to patients managed
amputation to observation among patients with recurrent with medical treatment (mainly hormonal therapy and che-
extremity desmoids. In a group of 22 patients who had dis- motherapy) [19].
ease requiring amputation, 7 underwent amputation, and 15 The criteria for selecting patients for active surveillance
were managed with a conservative approach. Among them, have been largely discussed. There is consensus that abdomi-
four patients received systemic treatment, and six were just nal wall tumors tend to be more indolent. In these patients,
observed. Within a follow-up of 25–92 months, none or surgery can allow for local control [20–25]; however, the
insignificant tumor progression was observed, and three post-surgical morbidity associated with abdominal recon-
patients in the observation group had spontaneous tumor struction needs to be considered. An important study on pri-
regression. DT were found to be enigmatic: the reason why mary abdominal wall desmoids by Bonvalot et al. [26]
some tumors continue to grow and others could regress was reported on 147 patients undergoing either frontline surgery
unclear. For the first time, an accent was put on the need to or initial observation/medical treatment. Within 36 months
account for function and structure preservation, while avoid- of follow-up, approximately one third of patients that did not
ing aggressive attempts to achieve negative resection mar- undergo surgery had stable disease, and one third experi-
gins, resulting in unnecessary complications. In parallel with enced spontaneous regression (Fig. 59.2a, b). The authors
this observation, it became clear over those years how recommended an initial nonsurgical approach to abdominal
­microscopic surgical margins were not associated to a higher wall DT, followed by surgery in case of tumor growth. The
risk of local recurrence, shedding some light on a different safety of a nonsurgical approach was also confirmed in extra-­
angle about the presence of a relatively indolent subgroup of abdominal wall DT. In a study by Colombo et al., analyzing
diseases. One of the first large studies analyzing the impact the natural history of patients with extra-abdominal wall DT,
of margins on long-term oncological outcome dates back in among the 122 patients treated with a non-operative manage-
2003. Gronchi et al., reporting on a series of 203 patients ment, 5% required surgery, 51% other treatments, while the
undergoing surgery for primary or recurrent DT, showed that 20% experienced spontaneous regression (Fig. 59.2c, d)
microscopically positive margins did not affect disease-free [27]. Tumor site was found to be a major prognostic factor
survival in patients with extra-abdominal DT [15]. These for event-free survival in a following study [28]. The survival
findings were confirmed by the following studies. In particu- of patients with DT in favorable locations, such as the
622 G. Spolverato and A. Gronchi

a b

c d

e f

Fig. 59.2 Panel a, b: contrast-enhanced T1-weighted MRI, axial view, right buttock desmoid at presentation (panel c) and 7 years after active
of a left abdominal wall desmoid at presentation (panel a) and 6 years surveillance (panel d), showing spontaneous regression; histology
after active surveillance (panel b), showing spontaneous regression. image of the desmoid tumor (panel e, H&E; panel f, IHC)
Panel c, d: contrast-enhanced T1-weighted MRI, coronal view, of a

abdominal wall; intra-abdominal, breast, and digestive vis- the patients were initially managed with active surveillance
cera; and lower limb, was independent from the treatment [28]. Finally, a randomized clinical trial comparing sorafenib
strategy (operative versus non-operative). Interestingly, the to placebo in patients with progressing DT showed that
outcome of patients with unfavorable locations was better if tumor stabilization or regression occurred in two thirds of
59 Desmoid Tumors 623

patients in the placebo group, further supporting a non-­ FAP-related DT, which tend to be more aggressive and
operative first-line approach [29]. ­interest patients who underwent major colorectal surgery.
The natural history of DT during active observation is diffi- However, the switch to an active treatment may need to be
cult to describe, since in most of the studies patients are offered considered sooner, given the higher risk of complications
to drop out from active observation early in the management. In related to the multifocality, anatomic site, and biological
a study from Briand et al., DT were found to be able to stabilize behavior.
spontaneously after 1 year of evolution and within 3 years for the
vast majority of patients, while the cumulative probability of fail-
ure during active treatment was approximately 10% at 10 years, 59.4 Selection for an Active Treatment
further confirming active surveillance as a valid first-line man-
agement for patients with primary or recurrent extra-abdominal The type of intervention mostly depends on tumor site
DT [30]. Of note, regressions have been noted also after initial (Fig. 59.3). Abdominal wall is the only site where surgery is
progressions in prospective observational studies on active sur- the first option in patients failing observation. Surgical mor-
veillance [31]. The final results along with the correlation bidity is generally limited, and local control is close to 95%
between mutational status and progression are awaited. [26]. If surgery is the treatment of choice, the primary goal is
Based on these findings, the Desmoid Tumor Working to achieve R0 margins [26].
Group reached a consensus about key aspects of the manage- For all other locations, surgery is reserved in case of fail-
ment of DT [32]. According to the most recent guidelines by ure of non-operative management, such as hormonal thera-
the Desmoid Tumor Working Group, active surveillance, pies, chemotherapy, and molecular target therapies. Local
defined as the continuous monitoring of DT patients with modalities such as isolated limb perfusion, radiotherapy,
magnetic resonance imaging (MRI), initially performed cryotherapy, and HIFU can also be considered depending on
within 1–2 months and then at 3–6 months interval, does not location, age, and symptoms.
influence the efficacy of subsequent treatments. Since up to Several medical treatments are available, but the lack of
50–60% of DT do not grow after diagnosis and 20–30% may comparative studies does not allow for a well-defined
shrink after initial progression [29], initial management of sequence of systemic options (Table 59.1). Among the sys-
DT should consider active surveillance while sparing unnec- temic therapies, we count: hormonal therapies, nonsteroidal
essary surgery. The presence of symptoms should not modify anti-inflammatory drugs (NSAIDs), tyrosine kinase inhibi-
the tumor-directed treatment approach, since pain control tors (TKIs), low-dose chemotherapy with iv methotrex-
can be achieved with a conservative approach. Perhaps more ate + vinblastine/vinorelbine or oral vinorelbine alone, and
importantly, an extended surgical treatment risks to nega- conventional chemotherapy including liposomal doxorubi-
tively affect pain control and quality of life. cin [29, 32, 34–41]. It is intuitive to employ less toxic ther-
Among the abandoned indication for upfront surgery, we apy followed by more toxic ones in a stepwise fashion.
recognize pain, as previously stated, resection of positive mar- Several factors need to be considered when defining the
gins, and pregnancy/postpartum. In particular, during preg- treatment plan, such as (1) level of evidence, (2) overall
nancy, due to the secretion of growth factors and hormones, response rate, (3) progression-free survival rate, (4) the ease
DT can grow or firstly appear. However, the tumor tends to of administration, and (5) the expected toxicity, and have
decrease after pregnancy; thus, surgery should not be offered. been summarized in a five-dimensional model proposed by
The biological basis beyond DT behavior has been investi- the Desmoid Tumor Working Group. After several retro-
gated, and in particular the characterization of the tumor spective studies testing tamoxifen or toremifene alone or in
immune microenvironment and immune checkpoints has been combination with NSAIDs and one phase II prospective
recently described by Siozopoulou et al. [33] Although the study showing limited activity of this regimen, there is no
authors did not define a molecular and immune pattern charac- evidence to consider antihormonal therapies and NSAIDs in
terizing patients experiencing tumor regression under observa- patients with DT [42–45]. Differently, clear clinical benefit
tion, they were the first to show that DT have a strong immune was found for TKIs, in particular for sorafenib and pazo-
infiltration at the tumor margins, characterized by lymphoid panib compared to imatinib. This effect is achievable even
aggregates, and to demonstrate that there is no PD-L1-driven with low dosages, limiting the potential side effects (such as
immune suppression present in the tumor cells. fatigue, rash, hypertension, gastrointestinal symptoms) [29,
36, 37, 46–48]. Of note, now that sorafenib and pazopanib
are largely offered to DT patients, the long-term risk of
59.3.2 FAP-Related Desmoid hypertension and/or hypothyroidism should be taken into
account. These side effects are not negligible since DT
While most of the available evidence is based on sporadic patients are often young. Chemotherapy options can include
DT, active surveillance is even more appropriate in case of low-dose methotrexate plus vinblastine or vinorelbine
624 G. Spolverato and A. Gronchi

Diagnosis (core needle biopsy)

If Stabilization / Regression:
Front-line approach: Active Surveillance (1-2 years)
Active Surveillance

In case of Progression
(consider - if clinically possible - to wait until 3 subsequent progression)

Intraabdominal / Extremity / girdles / Head & neck /

Abdominal wall
Retroperitoneal / pelvic chest wall intrathoracic

Sx MTx MTx MTx

Sx* /RTx or Sx* or

MTx (or RTx) MTx RTx MTx RTx or Sx* + RTx
both ILP

Investigational treatments, ...

Abbreviations: Sx:Surgery; Sx*: Surgery is an option if morbidity is limited; MTx: Medical treatment; RTx: Radiotherapy; ILP: Isolated limb perfusion.

Fig. 59.3 Treatment algorithm of DT. Published with permission from European Journal of Cancer127 (2020) 96–107

Table 59.1 Available evidences for the existing tyrosine kinase, γ-secretase inhibitors, and cytotoxic agents for desmoid tumor treatment
Article Pt Inclusion criteria Drug Duration Response rate 12 mo PFS
Heinrich 2006 19 Heavily pretreated patients Imatinib (800 mg) 325 days 16 36
Penel 2011 35 Radiological evidence for progression Imatinib (400 mg) 1 year 11 67
Chugh 2010 49 Locally advanced disease Imatinib (200–600 mg) Until progression 6 66
Kasper 2017 38 RECIST progression Imatinib (800 mg) 2 years 19 59
Gounder 2018 50 Progressive or symptomatic Sorafenib (400 mg) Until progression 33 89
Toulmonde 2019 48 RECIST progression Pazopanib (800 mg) 1 year 37 86
Kummar 2017 17 Progressive or symptomatic Nirogacestat (300 mg) Until progression 29 100
Toulmonde 2019 24 RECIST progression MTX + VBL 1 year 25 79
Palassini 2017 75 Progressive disease MTX + VBL MTX + VIN 14 months 48 60

(including oral vinorelbine alone) or conventional anthracy- Nirogacestat (PF-03084014) is an orally available, revers-
cline-based regimens or liposomal doxorubicin. The ible γ-secretase inhibitor, which proved to be highly active
response to low-dose chemotherapy has proved to occur sev- and safe. γ-Secretase cleaves intracellular Notch and
eral months after the start of therapy, to be durable even after results in pathway activation [34]. In a phase I trial, niro-
treatment withdrawal, and to occur in up to 50–70% of gacestat allowed to achieve partial response in five of
patients with acceptable toxicity [29, 36, 37, 46–48]. Two seven patients [49]. This result was confirmed by a follow-
regimens including doxorubicin plus dacarbazine and low- ing phase II trial in which the 29% of patients achieved
dose methotrexate plus vinca alkaloids showed to be effec- partial response and the 71% stable disease [50]. Finally,
tive in FAP-associated DT. tegavivint is currently being tested in the first in-human
A new treatment option for DT patients is nirogacestat, trial as a new therapeutic option, since it inhibits the Wnt
a safe and active drug with effect on the Notch signaling. and beta-catenin pathway.
59 Desmoid Tumors 625

TNF-α and melphalan-based isolated limb perfusion These results need to be confirmed in a larger series of DT,
(TM-ILP), considered the standard of care in Europe for possibly of different sites, to confirm T2 signaling as a valid
patients with limb-threatening sarcomas, are also offered to tool in DT surveillance [57].
patients with locally advanced aggressive fibromatosis to
avoid limb amputation. The evidences come from a study
conducted in three European centers, where 16/25 patients 59.6 Open Questions
treated with TM-ILP achieved complete or partial response,
22/25 had their limb spared, and the large majority had good There are still some unresolved issues on DT that require
functional outcomes [51, 52]. further investigations. First, although a paradigm shift in the
RT is another option to achieve local control, mostly after management of DT occurred in the last decade, it is neces-
failure of all available medical treatments, in the elderly, in sary to clearly define when to consider upfront surgery.
patients with several comorbidities, or in case of rapidly Possible complications, such as obstruction, perforation, and
growing disease threatening vital organs. Definitive RT at mesenteric ischemia can require upfront surgery; however, if
moderate doses (i.e., 50 Gy) allows obtaining a long-term the tumor is not safely resectable with negative margins, one
progression arrest in up to 70% of the cases. Long-term side could well consider to manage the complication and start a
effects such as fibrosis, fractures, and radiation-induced medical treatment, which—in case of disease control/
tumors should be considered when offering RT to young response—may avoid an R2 resection. Second, the prognos-
patients with DT [52]. tic role of CTNNB1 mutational status among patients under-
Among the other local options, cryotherapy and HIFU are going surveillance needs to be clarified in order to integrate
gaining popularity in the treatment of DT. However, few case the available prognostic tools with reliable biological bio-
series with a short follow-up and without a matched control markers. Similarly, the impact of APC mutations on the
group are available to support their indication. In particular development of DT in FAP patients and on their prognosis
cryotherapy has been recently proved to be safe, well toler- needs to be further evaluated. Perhaps more importantly,
ated, and effective in a series of 14 patients [53]. Cryoablation none of the available studies proved a correlation between
demonstrated feasibility in progressive DT patients enrolled specific APC mutations and the biological aggressiveness of
in a recent prospective, open-label, non-randomized, non-­ the disease and in particular the lack of metastatic potential.
comparative, multicenter study (CRYODESMO-01) [54]. Third, although the armamentarium of treatment options
Similarly, magnetic resonance-guided focused ultrasound for DT is growing, in the absence of comparative studies,
has been used to treat extra-abdominal DT due to its nonin- the treatment sequence in patients who fail active surveil-
vasiveness, lack of ionizing radiation, low morbidity, and lance has to be defined. Finally, the management of recur-
good safety profile [55]. However, since their promising rent DT is still debated. To which extent the management of
effects are still under investigation, they should be offered in primary DT can be applied to recurrent disease is unclear.
the context of clinical trials. Data on recurrent disease are needed to further understand
the behavior and the clinical management of these diverse
entities.
59.5 Assessment of Response

Assessment of response to treatment and assessment of 59.7 Conclusions

tumor growth during observation are still matters of concern.
RECIST criteria often fail to answer clinical questions, In conclusion, desmoids are rare and locally aggressive
because tissue response to treatment and in particular cellu- tumors, with low metastatic potential and often unpredict-
larity are difficult to define. MRI can be applied as a surro- able clinical course. Surgery has been the standard of care
gate for cellular content, since T2 sequences are able to for many years; however, recently we observed a paradigm
discriminate between neoplastic tissue, fibrous tissue, and shift toward a more conservative management. New treat-
extracellular components [56]. Response to systemic thera- ment modalities are awaited for patients who fail active sur-
pies has shown to be associated with decrease in T2 signal veillance and for recurrent disease. Health-related quality of
intensity, and recently hyperintense T2 was also found to life should be adopted as a metric to compare different treat-
correlate with progression-free survival [57]. In the study by ment approaches; validation of a specific tool is in progress.
Cassidy et al., reporting on 37 patients with extra-abdominal Prognostic and predictive biomarkers could serve as valid
DT managed with observation, 1-year PFS of T2 signal in tools to stratify patients for progression and response to
≥90% was 55%, compared to 94% of the T2 <90% group. treatment.
626 G. Spolverato and A. Gronchi

12. Vitellaro M, Sala P, Signoroni S, et al. Risk of desmoid tumours

Open Questions after open and laparoscopic colectomy in patients with familial
adenomatous polyposis. Br J Surg. 2014;101(5):558–65.
• Although a paradigm shift in the management of 13. Hamada S, Futamura N, Ikuta K, et al. CTNNB1 S45F mutation
DT occurred in the last decade, when is it necessary predicts poor efficacy of meloxicam treatment for desmoid tumors:
a pilot study. PLoS One. 2014;9(5):e96391.
to consider upfront surgery? 14. Kasper B, Gruenwald V, Reichardt P, Bauer S, Hohenberger P,
• What is the prognostic role of CTNNB1 mutational Haller F. Correlation of CTNNB1 mutation status with progres-
status among patients undergoing surveillance? sion arrest rate in RECIST Progressive desmoid-type fibromatosis
• What is the impact of APC mutations on the devel- treated with imatinib: translational research results from a phase 2
study of the German Interdisciplinary Sarcoma Group (GISG-01).
opment of DT in FAP patients and on their Ann Surg Oncol. 2016;23(6):1924–7.
prognosis? 15. Gronchi A, Casali PG, Mariani L, et al. Quality of surgery and
• What is the treatment sequence in patients who fail outcome in extra-abdominal aggressive fibromatosis: a series of
active surveillance? patients surgically treated at a single institution. J Clin Oncol.
2003;21(7):1390–7.
• To which extent the management of primary DT 16. Lev D, Kotilingam D, Wei C, et al. Optimizing treatment of des-
can be applied to recurrent disease? moid tumors. J Clin Oncol. 2007;25(13):1785–91.
• What is the biological and mutational basis under 17. Crago AM, Denton B, Salas S, et al. A prognostic nomogram
the lack of metastatic potential of DT? for prediction of recurrence in desmoid fibromatosis. Ann Surg.
2013;258(2):347–53.
18. Bonvalot S, Eldweny H, Haddad V, et al. Extra-abdominal primary
fibromatosis: aggressive management could be avoided in a sub-
group of patients. Eur J Surg Oncol. 2008;34(4):462–8.
References 19. Fiore M, Rimareix F, Mariani L, et al. Desmoid-type fibromato-
sis: a front-line conservative approach to select patients for surgical
1. Crago AM, Chmielecki J, Rosenberg M, et al. Near univer- treatment. Ann Surg Oncol. 2009;16(9):2587–93.
sal detection of alterations in CTNNB1 and Wnt pathway 20. Peng PD, Hyder O, Mavros MN, et al. Management and recurrence
regulators in desmoid-­ type fibromatosis by whole-exome patterns of desmoids tumors: a multi-institutional analysis of 211
sequencing and genomic analysis. Genes Chromosomes Cancer. patients. Ann Surg Oncol. 2012;19(13):4036–42.
2015;54(10):606–15. 21. Stoeckle E, Coindre JM, Longy M, et al. A critical analysis of treat-
2. Heiskanen I, Jarvinen HJ. Occurrence of desmoid tumours in famil- ment strategies in desmoid tumours: a review of a series of 106
ial adenomatous polyposis and results of treatment. Int J Color Dis. cases. Eur J Surg Oncol. 2009;35(2):129–34.
1996;11(4):157–62. 22. Bonvalot S, Desai A, Coppola S, et al. The treatment of desmoid
3. Bertario L, Russo A, Sala P, et al. Genotype and phenotype factors tumors: a stepwise clinical approach. Ann Oncol. 2012;23(Suppl
as determinants of desmoid tumors in patients with familial adeno- 10):x158–66.
matous polyposis. Int J Cancer. 2001;95(2):102–7. 23. Bertani E, Chiappa A, Testori A, et al. Desmoid tumors of the ante-
4. Timbergen MJM, Colombo C, Renckens M, et al. The prognos- rior abdominal wall: results from a monocentric surgical experience
tic role of beta-catenin mutations in desmoid-type fibromatosis and review of the literature. Ann Surg Oncol. 2009;16(6):1642–9.
undergoing resection only: a meta-analysis of individual patient 24. Catania G, Ruggeri L, Iuppa G, Di Stefano C, Cardi F, Iuppa
data. Ann Surg. 2021;273(6):1094–01; https://doi.org/10.1097/ A. Abdominal wall reconstruction with intraperitoneal prosthesis
SLA.0000000000003698. in desmoid tumors surgery. Updat Surg. 2012;64(1):43–8.
5. Lazar AJ, Tuvin D, Hajibashi S, et al. Specific mutations in the 25. Sutton RJ, Thomas JM. Desmoid tumours of the anterior abdominal
beta-catenin gene (CTNNB1) correlate with local recurrence in wall. Eur J Surg Oncol. 1999;25(4):398–400.
sporadic desmoid tumors. Am J Pathol. 2008;173(5):1518–27. 26. Bonvalot S, Ternes N, Fiore M, et al. Spontaneous regression of
6. Mullen JT, DeLaney TF, Rosenberg AE, et al. beta-Catenin muta- primary abdominal wall desmoid tumors: more common than pre-
tion status and outcomes in sporadic desmoid tumors. Oncologist. viously thought. Ann Surg Oncol. 2013;20(13):4096–102.
2013;18(9):1043–9. 27. Colombo C, Miceli R, Le Pechoux C, et al. Sporadic extra abdomi-
7. van Broekhoven DL, Verhoef C, Grunhagen DJ, et al. Prognostic nal wall desmoid-type fibromatosis: surgical resection can be safely
value of CTNNB1 gene mutation in primary sporadic aggressive limited to a minority of patients. Eur J Cancer. 2015;51(2):186–92.
fibromatosis. Ann Surg Oncol. 2015;22(5):1464–70. 28. Penel N, Le Cesne A, Bonvalot S, et al. Surgical versus non-­surgical
8. Colombo C, Miceli R, Lazar AJ, et al. CTNNB1 45F mutation approach in primary desmoid-type fibromatosis patients: a nation-
is a molecular prognosticator of increased postoperative primary wide prospective cohort from the French Sarcoma Group. Eur J
desmoid tumor recurrence: an independent, multicenter validation Cancer. 2017;83:125–31.
study. Cancer. 2013;119(20):3696–702. 29. Gounder MM, Mahoney MR, Van Tine BA, et al. Sorafenib
9. Garvey PB, Booth JH, Baumann DP, et al. Complex reconstruc- for advanced and refractory desmoid tumors. N Engl J Med.
tion of desmoid tumor resections does not increase desmoid tumor 2018;379(25):2417–28.
recurrence. J Am Coll Surg. 2013;217(3):472–80. 30. Briand S, Barbier O, Biau D, et al. Wait-and-see policy as a first-­
10. Domont J, Salas S, Lacroix L, et al. High frequency of beta-­ line management for extra-abdominal desmoid tumors. J Bone
catenin heterozygous mutations in extra-abdominal fibromatosis: Joint Surg Am. 2014;96(8):631–8.
a potential molecular tool for disease management. Br J Cancer. 31. Colombo C LVS, Fiore M, Grignani G, Palesandro E, Boccone B,
2010;102(6):1032–6. D'Ambrosio L, Bianco A, Collini P, Palassini E, Stacchiotti S, Paolo
11. Slowik V, Attard T, Dai H, Shah R, Septer S. Desmoid tumors com- Dei Tos A, Casali P, Perrone F, Mariani L, Gronchi A. Can wait and
plicating familial adenomatous polyposis: a meta-analysis mutation see be the standard of care for initial approach to primary sporadic
spectrum of affected individuals. BMC Gastroenterol. 2015;15:84. desmoid tumors? Preliminary data from an Italian sarcoma group
59 Desmoid Tumors 627

prospective study. Connective Tissue Oncology Society Annual sulindac: a single-center long-term observational study in 134
Meeting; 2018; Rome. patients. Familial Cancer. 2016;15(1):31–40.
32. Desmoid Tumor Working G. The management of desmoid tumours: 45. Skapek SX, Anderson JR, Hill DA, et al. Safety and efficacy of
A joint global consensus-based guideline approach for adult and high-dose tamoxifen and sulindac for desmoid tumor in children:
paediatric patients. Eur J Cancer. 2020;127:96–107. results of a Children's Oncology Group (COG) phase II study.
33. Siozopoulou V, Marcq E, Jacobs J, et al. Desmoid tumors dis- Pediatr Blood Cancer. 2013;60(7):1108–12.
play a strong immune infiltration at the tumor margins and no 46. Kasper B, Gruenwald V, Reichardt P, et al. Imatinib induces sus-
PD-L1-driven immune suppression. Cancer Immunol Immunother. tained progression arrest in RECIST progressive desmoid tumours:
2019;68(10):1573–83. Final results of a phase II study of the German Interdisciplinary
34. Kasper B, Raut CP, Gronchi A. Desmoid tumors: to treat or not to Sarcoma Group (GISG). Eur J Cancer. 2017;76:60–7.
treat, that is the question. Cancer. 2020;126(24):5213–21. 47. Gounder MM, Lefkowitz RA, Keohan ML, et al. Activity of
35. Heinrich MC, McArthur GA, Demetri GD, et al. Clinical and Sorafenib against desmoid tumor/deep fibromatosis. Clin Cancer
molecular studies of the effect of imatinib on advanced aggressive Res. 2011;17(12):4082–90.
fibromatosis (desmoid tumor). J Clin Oncol. 2006;24(7):1195–203. 48. Szucs Z, Messiou C, Wong HH, et al. Pazopanib, a promising
36. Toulmonde M, Pulido M, Ray-Coquard I, et al. Pazopanib or option for the treatment of aggressive fibromatosis. Anti-Cancer
methotrexate-­ vinblastine combination chemotherapy in adult Drugs. 2017;28(4):421–6.
patients with progressive desmoid tumours (DESMOPAZ): a non-­ 49. Messersmith WA, Shapiro GI, Cleary JM, et al. A Phase I, dose-­
comparative, randomised, open-label, multicentre, phase 2 study. finding study in patients with advanced solid malignancies of the
Lancet Oncol. 2019;20(9):1263–72. oral gamma-secretase inhibitor PF-03084014. Clin Cancer Res.
37. Chugh R, Wathen JK, Patel SR, et al. Efficacy of imatinib in 2015;21(1):60–7.
aggressive fibromatosis: Results of a phase II multicenter Sarcoma 50. Gounder MM. Notch inhibition in desmoids: “Sure it works in prac-
Alliance for Research through Collaboration (SARC) trial. Clin tice, but does it work in theory?”. Cancer. 2015;121(22):3933–7.
Cancer Res. 2010;16(19):4884–91. 51. van Broekhoven DL, Deroose JP, Bonvalot S, et al. Isolated
38. Penel N, Le Cesne A, Bui BN, et al. Imatinib for progressive and limb perfusion using tumour necrosis factor alpha and melpha-
recurrent aggressive fibromatosis (desmoid tumors): an FNCLCC/ lan in patients with advanced aggressive fibromatosis. Br J Surg.
French Sarcoma Group phase II trial with a long-term follow-up. 2014;101(13):1674–80.
Ann Oncol. 2011;22(2):452–7. 52. Keus RB, Nout RA, Blay JY, et al. Results of a phase II pilot study
39. Palassini E, Frezza AM, Mariani L, et al. Long-term efficacy of moderate dose radiotherapy for inoperable desmoid-type fibro-
of methotrexate plus vinblastine/vinorelbine in a large series matosis – an EORTC STBSG and ROG study (EORTC 62991-­
of patients affected by desmoid-type fibromatosis. Cancer J. 22998). Ann Oncol. 2013;24(10):2672–6.
2017;23(2):86–91. 53. Saltiel S, Bize PE, Goetti P, et al. Cryoablation of extra-abdominal
40. Azzarelli A, Gronchi A, Bertulli R, et al. Low-dose chemother- desmoid tumors: a single-center experience with literature review.
apy with methotrexate and vinblastine for patients with advanced Diagnostics (Basel). 2020;10(8):556.
aggressive fibromatosis. Cancer. 2001;92(5):1259–64. 54. Kurtz JE, Buy X, Deschamps F, et al. CRYODESMO-O1: A prospec-
41. Gega M, Yanagi H, Yoshikawa R, et al. Successful chemotherapeu- tive, open phase II study of cryoablation in desmoid tumour patients
tic modality of doxorubicin plus dacarbazine for the treatment of progressing after medical treatment. Eur J Cancer. 2020;143:78–87.
desmoid tumors in association with familial adenomatous polypo- 55. Griffin MO, Kulkarni NM, O'Connor SD, Sudakoff GS, Lea WB,
sis. J Clin Oncol. 2006;24(1):102–5. Tutton SM. Magnetic resonance-guided focused ultrasound: a brief
42. Fiore M, Colombo C, Radaelli S, et al. Hormonal manipulation review with emphasis on the treatment of extra-abdominal desmoid
with toremifene in sporadic desmoid-type fibromatosis. Eur J tumors. Ultrasound Q. 2019;35(4):346–54.
Cancer. 2015;51(18):2800–7. 56. Sundaram M, McGuire MH, Herbold DR, Beshany SE, Fletcher
43. Hansmann A, Adolph C, Vogel T, Unger A, Moeslein G. High-dose JW. High signal intensity soft tissue masses on T1 weighted pulsing
tamoxifen and sulindac as first-line treatment for desmoid tumors. sequences. Skelet Radiol. 1987;16(1):30–6.
Cancer. 2004;100(3):612–20. 57. Cassidy MR, Lefkowitz RA, Long N, et al. Association of
44. Quast DR, Schneider R, Burdzik E, Hoppe S, Moslein G. Long-­ MRI T2 signal intensity with desmoid tumor progression dur-
term outcome of sporadic and FAP-associated desmoid tumors ing active observation: a retrospective cohort study. Ann Surg.
treated with high-dose selective estrogen receptor modulators and 2020;271(4):748–55.
 Atypical Patterns of Metastases: How
Do Sarcomas Metastasize? 60
Pia van der Laan, Fabio Tirotta, Valeriya Pankova,
Samuel Ford, Paul Huang, and Winan J. van Houdt

Abstract 60.1 Introduction

Soft tissue sarcoma is a rare and heterogeneous group of
cancers and comprises around 1% of all malignancies. Soft tissue sarcoma (STS) is a heterogeneous group of rare
This group of tumors consists of more than 60 different solid tumors that arise in mesenchymal tissues and accounts
subtypes, each varying in biology and clinical behavior. for 1% of all malignancies in adults [1]. STS consists of a
In general, between 25% and 50% of the patients develop broad spectrum of tumors in terms of location, histologic sub-
local or distant recurrences. However, the propensity of type, treatment strategy, age at diagnosis, and prognosis.
sarcoma to metastasize depends on many different fac- Nowadays, over 60 subtypes have been described, with undif-
tors, such as histological subtype, anatomical site, size, ferentiated pleiomorphic sarcoma (28%), Gastrointestinal
and grade. Most of the STS subtypes have a tendency to stromal tumor (18%), liposarcoma (15%), leiomyosarcoma
spread hematogenously with a predilection for pulmonary (12%), and synovial sarcoma (10%) being some of the most
seeding, while lymph node metastases are rarely seen. frequently diagnosed subtypes [2, 3]. Examples of more rare
However, there are several subtypes with deviating meta- subtypes are clear cell sarcoma, epithelioid sarcoma, and alve-
static patterns, and in this chapter, the different atypical olar soft part sarcoma. Soft tissue sarcoma can occur at all
metastatic patterns are being described. ages and at any anatomic location; however, the majority of
STS arise in the extremities, more frequently in the lower
compared to the upper extremity [4]. Surgical resection, often
with (neo-)adjuvant radiotherapy, is the mainstay of treatment
Learning Objectives
with curative intent or primary STS, while for specific sub-
• Most soft tissue sarcoma subtypes have a predilec-
types and for some high-risk sarcoma, also neo-adjuvant che-
tion for pulmonary metastases.
motherapy is given [5]. However, between 25% and 50% of
• Specific sarcoma subtypes show a spectrum of
patients develop distant metastases, and the dogma is that they
atypical metastatic patterns.
usually arise as pulmonary metastases without any lymph
• Lymph node metastasis occurs only in a small sub-
node metastases. Although a dogma becomes a dogma for a
set of sarcomas, according to the CARES
reason, it has become more clear over the years that metastatic
acronym.
patterns vary widely between different sarcoma subtypes and
• Hypoxia or specific oncogenes might play a role in
that for some sarcoma subtypes there are atypical metastatic
driving sarcoma metastasis.
patterns. This is important to realize for surveillance purposes
and also for proper staging before starting any treatment. In
this chapter, the aim is to clarify these atypical patterns to pro-

P. van der Laan · W. J. van Houdt (*) V. Pankova · P. Huang

Department of Surgical Oncology, Sarcoma and Melanoma Unit, Division of Molecular Pathology, The Institute of Cancer
The Netherlands Cancer Institute – Antoni van Leeuwenhoek, Research, London, UK
Amsterdam, The Netherlands
e-mail: [email protected]
F. Tirotta · S. Ford
Department of Sarcoma Surgery, Queen Elizabeth Hospital,
Birmingham, UK

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 629
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_60
630 P. van der Laan et al.

vide more insight in tumor biology of the different soft tissue per subtype and per country/hospital, but the most common
sarcoma subtypes. chemotherapeutic therapies for sarcoma are doxorubicin-­
based regimens. Also, for patients with oligometastatic dis-
ease with longer disease-free intervals, local treatment can
60.2 Metastases in Sarcoma: General be beneficial which traditionally used to be resection of
Patterns, Risk Factors, and Prognosis mostly pulmonary metastases [10]. For pulmonary metasta-
ses, even repeated metastasectomy has been associated with
Despite often adequate locoregional treatment for primary prolonged overall survival [11]. More recently, also other
tumors, STS frequently metastasizes, which is the main local treatment options such as stereotactic radiotherapy, as
cause of death in many STS patients. Around 10% of patients well as ablative techniques, have been shown to be effective
have detectable metastases at the time of diagnosis of the in this setting.
primary tumor, whereas as many as 25–40% will develop
metastases at a later time point reflecting the high metastatic
potential of STS [4]. The majority of these metastases occurs 60.3 Subtypes with Alternative
within 2 years after treatment of the primary tumor [6]. In Metastasizing Patterns
general, most sarcoma subtypes have a tendency to metasta-
size to the lungs as the only or at least the first site of distant Although many sarcoma subtypes have a preference for pul-
disease. Therefore, in most treatment and follow-up guide- monary metastases, there are several sarcoma types with
lines and local protocols, pulmonary surveillance by chest deviating metastasizing patterns. To illustrate this, the most
CT or chest X-ray is recommend after treatment of the pri- important atypical metastatic patterns of STS subtypes are
mary tumor. being described below, without claiming to be complete
However, in the last decades our knowledge of and insight since there are several ultra-rare sarcomas with atypical pat-
in the biological behavior of the different sarcoma subtypes terns as well. It is currently poorly understood why many
have been increased. Also, better prediction tools for the risk subtypes have the predominance for lung metastases,
of metastases have been developed. Several patient and whereas others show some completely different metastatic
tumor characteristics have been identified as risk factors for patterns.
distant metastases. The risk of metastases varies widely per
subtype. The location of the primary tumor has also shown to
be an important factor in the risk of distant recurrence, since 60.3.1 Leiomyosarcoma
patients with extremity sarcoma have a lower risk of recur-
rence compared to patients with primary tumors elsewhere. Soft tissue leiomyosarcoma (LMS) originates from cells show-
Tumors with a higher histological grade, a larger tumor size, ing smooth muscle differentiation and predominantly arise in the
and deeply seated tumors are also more likely to metastasize extremities, abdomen, or retroperitoneum, from which a major
[7]. Recently, two prediction models have been developed part originates from large blood vessels [4]. LMS is a heteroge-
and made available for everyone with an app, the Sarculator neous disease, ranging from low-grade tumors located in the der-
and the Persarc apps, which are both available in the App mis that behave relatively benign to very aggressive lesions with
Store or Google Play Store. These models are based on a a highly aggressive metastatic potential [12, 13]. Grade and size
large number of patients and take into account many patient- have shown to be the most important factors affecting distant
and tumor-specific characteristics to better personalize cal- recurrence and survival [14].
culations for the risk of metastases [8, 9]. Metastatic patterns of LMS are quite distinct from other
Since the efficiency of systemic treatment options for subtypes. Although the lung is among the most common
metastatic STS is limited, survival remains poor. Median sites of metastasis, metastases in the liver and in soft tissue
survival after detection of metastases is around 1 year [10]. like subcutis or muscles are frequently seen, with or without
However, there is a wide variety in metastatic presentation, concomitant pulmonary metastases [4, 12]. A study of Wile
and prognosis is usually better after a long disease-free inter- et al. describe lung and liver metastases as most frequent in
val between the primary tumor and development of metasta- LMS each accounting for 30% of metastases [15].
ses. Also, oligometastatic disease carries a better prognosis Leiomyosarcoma is also the most common source of sar-
compared to multiple metastatic lesions. Treatment of meta- coma metastases to the skin [16]. Typical examples are cuta-
static disease varies widely between per patients and per sub- neous or subcutaneous metastases, anywhere on the body but
type, but the mainstay of treatment is usually systemic even on the scalp, sometimes clinically mimicking a seba-
treatment with older chemotherapy regimens, newer tyrosine ceous cyst. Treatment of systemic disease will often include
kinase inhibitors, or other more modern drugs. The choice of chemotherapy regimens such as doxorubicin-DTIC or newer
first line and further lines of systemic therapy varies widely agents like pazopanib. Metastasectomy of pulmonary and
60 Atypical Patterns of Metastases: How Do Sarcomas Metastasize? 631

hepatic metastases has shown to increase long-term survival to 20% of the patients are metastasized at initial diagnosis
in selected cases [17–19]. [27]. Metastases occur predominantly intra-abdominally
Follow-up imaging after treatment of the primary tumor (most frequently involving the liver), while lymph node
should not only consist of chest imaging, but at least a CT metastases or metastases outside the abdomen such as in the
chest/abdomen/pelvis is advisable given the aberrant meta- lungs, bone, or brain are rare, except for some rare GIST
static patterns. subtypes such as SDH-deficient GISTs [27, 28]. Therefore,
imaging of the abdomen and pelvis will be sufficient in most
cases.
60.3.2 Myxoid Liposarcoma The tyrosine kinase inhibitor (TKI) imatinib is the stan-
dard first-line treatment for advanced GIST. The role of sur-
Myxoid liposarcoma (MLS) accounts for 30–50% of liposar- gery in the management of metastatic GIST is controversial:
comas, and most cases are characterized by a t(12,16) resection of limited disease after good response to imatinib
(q13;p11) translocation resulting in the FUS-CHOP product treatment can be considered, whereas surgery for GIST pro-
[20]. MLS is in general classified as a low-grade sarcoma gressing at several sites is not recommended [26]. Resection
with a relatively good clinical outcome. However, a subset of of oligo-progressive metastases after progression on TKI
MLS tumors is considered as a high-grade version of MLS, treatment could be considered; however, the median
containing “hypercellular” or “round cell” areas, showing to progression-­free survival is fairly short (8–9 months), and
have an increased potential to metastasize [4, 21]. Spillane the 2-year progression-free survival is only 9.7% [29].
et al. demonstrated that the presence of a round cell compo-
nent in MLS tumors was predictive of a 5-year metastasis
rate of 58%, whereas those without a round cell component 60.3.4 Alveolar Soft-Part Sarcoma
had a 5-year metastasis rate of 16% [22].
MLS tumors tend to metastasize in a non-pulmonary Alveolar soft-part sarcoma (ASPS) is a rare form of STS
fashion. In a study of Shinoda et al., involving 48 metastatic accounting for less than 1% of the STS. It often affects young
MLS patients, 45 patients (93.8%) having extrapulmonary adults and in a majority of patients originates from the lower
lesions were identified including lesions in the bone, abdo- extremities. Although it is known for its relative indolent
men, retroperitoneum, and other soft tissues. From this group behavior, ASPS has the greatest metastatic potential com-
of patients, the majority showed metastases without the pared to other types of STS [30]. Metastases sometimes
lungs being involved (81.3%). Combining literature and occur decades after resection of the primary tumor often
their data, they suggest lung and liver metastases as potential without evidence of local recurrence [31].
prognostic factors of shorter survival [23]. Because of the The most common metastatic sites are the lung, bone, and
distinct pattern of metastasis in MLS, patients should brain. ASPS metastasizes to the brain more commonly com-
undergo imaging studies of the chest, abdomen, spine, and pared to other types of STS [32]. Lieberman et al. published
pelvis during staging or follow-up [20]. Sheah et al. propose follow-up data of 91 ASPS patients and showed a median
MRI for the early detection of metastases, since MRI has overall survival of 11 years in patients without metastasis at
shown to be the most sensitive modality to detect soft tissue diagnosis and 3 years in patients with metastatic disease [33].
and bone lesions [24]. A total body MRI could be considered ASPS is one of the few sarcoma subtypes that has shown some
[25]. Metastatic patients can be treated with many different promising responses to anti-PD1 immunotherapy [34].
multidisciplinary strategies, varying from radiotherapy for
this radiosensitive tumor or resections, and for oligometa-
static or oligoregressive lesions, to systemic therapy such as 60.3.5 Low-Grade Fibromyxoid Sarcoma
doxorubicin-ifosfamide or trabectedin.
Low-grade fibromyxoid sarcomas (LGFMS) are tumors his-
tologically consisting of alternating collagenous and myxoid
60.3.3 Gastrointestinal Stroma Tumor areas. These tumors are genetically characterized by either a
t(7,16) or t(11,16) translocation resulting in a FUS-CREB3L2
Gastrointestinal stromal tumor (GIST) represents 20% of all or FUS-CREB3L1 fusion gene [35]. It commonly arises in
STS, making it the largest group of single-type STS. GIST the proximal extremities and trunk and typically occurs in
tumors are usually caused by activating mutations in KIT young adults with a slight male predilection [4, 36].
and PDGFRA genes, but some other rare molecular aberra- Although LGFMS appears histologically indolent and
tions have also been described (so-called wild-type GISTs) shows low metastatic rates in the first years after resection, the
[26]. Primary GIST develops throughout the gastrointestinal metastatic pattern of this subtype is atypical since this tumor is
tract, most commonly in the stomach or small intestine. Up known for its extreme late metastases with intervals up to
632 P. van der Laan et al.

45 years [37]. Prolonged follow-up is therefore recommended achieved in patients with isolated regional LNM [45, 51–53] and
in these often relatively young patients. Overall metastatic rate questioned whether a more reliable classification would be
varies between 6% and 26%, and the main sites are the lung, achieved by separating LNM from the other sites. To support this
pleura, chest wall, and other soft tissues [35, 36]. Surgical data, other studies [41, 44, 47] demonstrated a limited impact on
excision is usually performed for locally recurrent lesions and prognosis in patients with isolated LNM when treated properly
also for late oligometastatic disease [37]. For advanced with resection, compared to those with pulmonary or other sys-
LGFMS, efficacy of chemotherapy is limited [38]. temic metastases. Keung et al. showed that patients with LNM
have a poorer prognosis in the absence of distant metastases
(N1M0) compared to patients with primary disease only (N0M0)
60.4 Lymph Node Metastases in Sarcoma and that LNM affect histologies differently. Interestingly, patients
who had favorable histologies and LNM were those showing
Lymph node metastases (LNM) from STS are uncommon, worse outcomes [44].
with an overall estimated incidence of approximately 5% An important and unanswered question is how and when to
[39–46]. Their pathologic evaluation is inconsistent making screen lymph nodes in STS. Clinical examination, CT or MRI
their real incidence difficult to determine [44], although it is scan, and sentinel lymph node biopsy (SNLB) can be used.
accepted that specific histologies, such as angiosarcoma, The former seems to be particularly effective for clear cell sar-
clear cell sarcoma, epithelioid sarcoma, and small cell sar- coma but less so for other histologies [54]. Ecker et al. demon-
coma, have a higher tendency to give LNM [41–44]. Hence, strated a possible higher risk of death when lymph node
studies reporting higher incidence rates of LNM are those evaluations were omitted in patients with epithelioid and clear
which considered only high-risk subtype for LNM [42, 43, cell sarcoma [43]. Although there is no clear evidence that rou-
46]. The low incidence of LNM in sarcoma is usually tine evaluation of lymph nodes improves survival outcomes in
explained by the difference in cell of origin, where most all subtypes or if this should be limited to high-risk histologies
malignancies have an epithelial origin, while sarcoma is a only, it does seem logical to at least evaluate for LNM in his-
mesenchymal tumor which is probably less likely to spread tologies known to be at higher risk of developing LNM.
to lymphatic vessel lying underneath epithelial layers. Where LNM are present without further metastatic dis-
LNM from STS can affect any anatomical site. Although ease (N1M0), the question of whether to perform a locore-
previously reported series at a single institution described an gional lymphadenectomy still needs to be addressed. In a
incidence of LNM in the trunk/extremities up to 70% of recent report, Basile et al. demonstrated improved survival
cases at the time of presentation [39, 41], two large retro- outcomes for patients who underwent excision of isolated
spective studies – one using the Surveillance, Epidemiology, LNM in extremity STS [45]. This study, along with others
and End Results database and the other using the National [39, 41], recommended treating isolated LNM with en bloc
Cancer Data Base – found opposite results, reporting an inci- surgical resection. In contrast, Sawamura reported possible
dence of approximately 2% [44, 47]. Indeed, in the absence benefits for lymphadenectomy at 1.5 years but no difference
of synchronous distant metastases, LNM in extremities were in survival at 5 years [48]. The role of lymphadenectomy
described in just 1% of cases [47]. Moreover, the same therefore warrants further large, multicenter, and prospective
authors identified rates of LNM for head and neck, intratho- evaluation, but in the meantime locoregional lymph node
racic, and intra-abdominal/retroperitoneal of 4.5%, 3.1%, dissections for LNM in the absence of further distant metas-
and 4.5%, respectively. This wide difference in distribution tases are considered standard of care and are also incorpo-
across sites may be related to an uneven distribution of the rated in the NCCN guidelines [55].
histologies and to a different number of lymph nodes
removed during surgery depending on the location, as non-­
deliberate removal of lymph nodes may occur more often in 60.4.1 Angiosarcoma
other sites than in the extremities. In addition, a different
staging/treatment related to the anatomical location may also Angiosarcoma represents 1% of all soft tissue sarcomas. In
be responsible for such a wide difference. half of all cases, it arises from the scalp, originating from
The clinical impact of LNM in STS is unclear. Several studies deep soft tissues in only 10% of cases [56, 57]. The overall
have reported worse survival outcomes in patients with LNM, incidence of LNM for angiosarcoma ranges from 6% to
suggesting that it can represent a preliminary manifestation of 13.5% [39, 44] although a recent study found that regional
systemic disease [48–50]. According to the eighth edition of the LNM from the scalp, which is the most frequent site, can
AJCC staging system, LNM in the extremities/trunk are consid- reach 52.5% [58]. Distant metastases were less frequent in
ered prognostically comparable to distant metastases [51]. This the absence of lymph nodal spread, and LNM seemed to be
newest edition has been largely criticized for this reason, and able to predict systemic spread within 3 and 6 months.
different reports have highlighted that a better prognosis is Nevertheless, overall survival did not differ in the presence
60 Atypical Patterns of Metastases: How Do Sarcomas Metastasize? 633

of LNM. Other anatomical sites such as the breast can be prognosis compared to other histologies, even when a high-­
involved, where the number of cases of radiation-induced grade component is present [62]. LNM from myxofibrosar-
angiosarcoma is rising. Noticeably, lymphadenectomy is not coma are usually rare, <10% [44, 45]. Sanfilippo et al.
recommended as lymph nodes are rarely involved [59]. reported that three out of four patients with regional node
The subtypes more likely to develop lymph node metasta- disease developed distant metastases and one out of five
ses are more epitheiloid like compared to other sarcoma sub- patients had positive LNM before their metastatic spread,
types or are developing in the skin (cutaneous angiosarcoma). suggesting a potential role in predicting metastatic spread in
This might partly explain why these subtypes more resemble this specific subtype [62].
the lymphogenic metastatic patterns of malignancies with an
epithelial origin.
60.4.5 Rhabdomyosarcoma

60.4.2 Epithelioid Sarcoma LNM in pediatric cases and young adults is particularly rel-
evant in patients with rhabdomyosarcoma, which constitute
Epithelioid sarcoma represents less than 1% of all STS and half of all STS during this age [63]. Different subtypes of
have a high tendency to metastasize to lymph nodes [60] rhabdomyosarcoma have been described, including embryo-
with an incidence of LNM of approximately 10% [43, 44]. In nal, alveolar, and pleomorphic rhabdomyosarcomas.
contrast to angiosarcoma, lymph nodal evaluation seems to Particularly alveolar rhabdomyosarcoma has a distinct
affect survival outcomes, which may be related to a direct genetic background including two patterns of translocations:
benefit from the lymphadenectomy or to better disease stag- t(2,13) (q35;q14) creating the PAX3-FOXO1 fusion gene
ing [43]. The role of lymphadenectomy requires further eval- and t(1,13) (p36;q14) with the PAX7-FOXO1 fusion gene
uation. In fact, some studies [43, 47] do not show clear [64]. The presence of LNM in rhabdomyosarcoma is approx-
evidence that extended lymphadenectomy improves OS in imately 20% and associated with worse prognosis [65, 66].
patients with extremity STS. However, other studies [39] In this specific case, sentinel node biopsy is safe and effec-
demonstrated opposing results, showing that long-term sur- tive [63] and can be used in order to potentially offer extended
vival was achieved in those patients who underwent radical lymphadenectomy to minimize the risk of recurrence and/or
resection of involved LNM. Nevertheless, lymph node evalu- metastatic disease. However, the role of SNB for soft tissue
ation should routinely be performed in these patients [43]. sarcoma in general, but also for the subtypes more likely to
develop LNM, is not established or proven to be beneficial.

60.4.3 Clear Cell Sarcoma

60.4.6 Others
Clear cell sarcoma is a rare type of sarcoma, previously
described as “melanoma of soft parts” for its histological Historically, the abovementioned subtypes were combined
similarity and the propensity for lymph nodal spread, which with synovial sarcoma and incorporated in the acronym
tends to occur in 10–15% of cases [44, 45]. Genetically, clear SCARE (synovial sarcoma, clear cell sarcoma, angiosarcoma,
cell sarcoma is marked by the translocation t(12,22) rhabdomyosarcoma, and epithelioid sarcoma), also called
(q13;q12), found in 70–90% of the cases [4]. In common CARES. These subtypes were always considered to be at high
with other high-risk histologies, the prognostic impact of risk for LNM. However, more recent and larger studies dem-
LNM and the potential benefits of nodal staging and treat- onstrated that the real incidence of LNM in synovial sarcoma
ment appear to be restricted to the subset of patients with is low, representing approximately <5% [43, 45, 67], and this
isolated N1 disease in the absence of M0 disease [44, 45]. In calls into question the accuracy of previously reported series
a study involving only extremities STS, Bianchi et al. that classified this histology as high risk for LNM.
reported a 10-year specific survival rate of 41% for all
patients, 53% for localized, and 0% for patients with metas-
tasis, regardless of lymph node involvement [61]. This was 60.5 Sarcoma Metastases: Biological
consistent with what was reported in larger series [44, 45]. Mechanisms

60.5.1 Overview of the Metastatic Cascade

60.4.4 Myxofibrosarcoma and Organotropism

Myxofibrosarcoma is one of the most common STS, espe- The development of metastasis is a multistep process referred
cially in the elderly [62]. One of its peculiarities is the high to as the metastatic cascade, which enables dissociation of
risk of local recurrence despite the fact that it has a better cancer cells from the primary site and secondary outgrowth
634 P. van der Laan et al.

at distant organs. The metastatic cascade follows sequential dicted risk of metastasis in undifferentiated pleomorphic
steps of local invasion of the primary tissue, intravasation of sarcoma (UPS) [76]. Under hypoxic conditions, the tran-
cancer cells into vessels, survival of circulating cancer cells, scription factor hypoxia-inducible factor 1-alpha (HIF1A)
extravasation into the secondary sites, and colonization and induces expression of multiple genes to increase angiogen-
growth at distant sites [68]. All of these steps must be suc- esis and enable cellular metabolism in oxygen-deprived
cessfully completed for a detectable lesion to develop; how- conditions [77]. In carcinomas, HIF1A is known to pro-
ever, in reality metastasis is a highly inefficient process, with mote migration and cancer cell invasion. Procollagen-
<0.01% of tumor cells in the systemic circulation giving rise lysine,2-oxoglutarate 5-dioxygenase 2 (PLOD2) is a
to macro metastatic lesions [69]. Tumors consist of clones downstream target of HIF1A and is an enzyme that cata-
with variable metastatic potential, and it has been argued that lyzes the hydroxylation of lysine residues, which is a type
there are pre-existing clones present in the primary sarcoma of posttranslational modification essential for maturation of
tumor that possess traits necessary to complete all steps in collagen [78]. Eisinger-Mathason et al. found that HIF1A
the metastatic cascade resulting in distant metastases which and PLOD2 were highly expressed in primary tumors from
are clonal [70]. UPS patients who went on to develop metastasis versus
Most cancers, including sarcomas, preferentially spread those who had low expression levels [79]. Knockdown of
to certain organs in a process known as metastatic organotro- either HIF1A or PLOD2 expression reduced sarcoma cell
pism [71]. The reasons for these recurrent patterns of meta- migration in vitro and lung metastasis development in a
static spread have long been a matter of debate. In 1889 murine model of UPS. Moreover, PLOD2 expression in
Steven Paget proposed the “seed and soil” theory of meta- these models was associated with dense and disorganized
static dissemination suggesting that the spread of metastatic collagen network, which could potentially form migration
cells relies on the compatibility of cancer cells (“seeds”) and tracks for metastasizing sarcoma cells. It has been further
distant organ or tissue (“soil”) [72]. Paget’s view was chal- shown by Lewis et al. that in UPS, hypoxia modifies the
lenged by James Ewing who proposed that metastatic spread mechanical properties of collagen through increased
is governed purely by mechanical factors such as anatomical PLOD2 expression, and there is a feedback loop between
structure of vasculature and patterns of blood flow. This mechanical properties and PLOD2 expression generated
“mechanical” spread allows initial delivery and arrest of can- through TGFβ-SMAD2 signaling that enhances sarcoma
cer cells at organs first encountered based on blood-flow pat- cell migration and lung metastasis [80]. These data demon-
terns from the primary tumor. Growing evidence suggests strate that hypoxia and the mechanical properties of colla-
that these two concepts are not mutually exclusive, and both gen may be important in promoting UPS metastatic spread,
may play complementary roles in successful metastatic but more investigation is required in order to establish if
spread and development [73]. hypoxia plays a role in determining preferential spread to
specific organs in UPS.

60.5.2 Molecular Drivers of Soft Tissue

Sarcoma Metastasis 60.5.4 Myxoid Liposarcoma

While the biological mechanisms of metastasis have been In contrast to the majority of STS subtypes that metastasize
extensively studied in bone sarcomas, less is known about to the lung, myxoid liposarcomas (MLS) have atypical pat-
the molecular drivers of STS metastasis. To date, studies terns of metastasis, with preference for the retroperitoneum,
have been largely undertaken in undifferentiated pleomor- limbs, and bones [81–83]. A genetic hallmark of MLS is a
phic sarcoma (UPS), myxoid liposarcoma (MLS), and syno- translocation which fuses FUS (FUsed in Sarcoma) to CHOP
vial sarcoma (SS). Here we summarize our current (C/EBP HOmologous Protein), and the FUS-CHOP onco-
understanding of the extrinsic and intrinsic molecular factors gene has been implicated in driving metastasis in this disease
that play a role in driving metastasis in these subtypes. [84]. Overexpression of FUS-CHOP increased the migration
and invasion of liposarcoma and fibrosarcoma cells in vitro
as well as promoted the formation of liver and lung metasta-
60.5.3 Undifferentiated Pleomorphic Sarcoma sis in the chicken embryo metastasis (CAM) assay in vivo
[85]. In another study by Tornin et al., the authors found that
Hypoxia occurs when a tumor outgrows its blood supply in MLS cells, FUS-CHOP activated the proto-oncogene
resulting in insufficient tumor oxygenation that has been tyrosine-protein kinase Src (SRC) and increased the levels of
linked to distant metastasis development in STS [74, 75]. focal adhesion kinase (FAK), which enhanced MLS cell
Moreover, a gene expression signature consisting of upreg- invasion and extravasation [86]. In MLS cells, inhibition of
ulated hypoxia-related genes (e.g., HIF1A, PLOD2) pre- SRC or FAK reduced 3D collagen spheroid invasion and
60 Atypical Patterns of Metastases: How Do Sarcomas Metastasize? 635

extravasation in the CAM assay. These studies have shed and metastatic organotropism might pave the way for more
light on the role of the fusion protein and its downstream effective treatment strategies for STS patients to either pre-
effectors in driving metastasis; however, it remains to be vent metastases or more effectively treat metastatic disease.
determined if the FUS-CHOP-induced pathways govern the
atypical metastatic patterns commonly seen in MLS.
Open Questions
Future directions in research in this field should focus
60.5.5 Synovial Sarcoma on the many open questions that remain, such as:
• Why is there a preference for pulmonary metastases
A few studies suggest the involvement of epithelial-to-­
in most STS subtypes, independent of anatomical
mesenchymal transition (EMT) in the synovial sarcoma
location of the primary tumor?
(SS) metastatic process [87–89]. EMT has been exten-
• What explains the variety in metastatic patterns
sively described in carcinomas which increases the pro-
between STS subtypes?
pensity to disseminate once cancer cells acquire the
• Why does a minority of specific STS subtypes
mesenchymal phenotype [90]. As sarcomas are mesenchy-
develop lymphogenic metastases, in contrast to
mal tumors, there remains a debate as to whether EMT
most other subtypes?
exists in these tumors. SS exhibit some epithelial differen-
• Can we identify biological mechanisms driving
tiation, particularly biphasic SS which have glandular
STS metastasis?
structures, and expression of epithelial markers such as
• How do we personalize treatment and surveillance
E-cadherin and cytokeratin is seen in SS tumors [91, 92].
strategies based on metastatic pattern differences?
In particular, high E-cadherin expression associates with
better outcome in SS [93, 94]. Peng et al. showed that in
SS cells, the SHC SH2-domain binding protein 1
(SHCBP1) protein enhances adhesion, migration, and Patient Information and Guidelines
invasion through promotion of EMT (decreased levels of ESMO Clinical Practical Guidelines - Soft Tissue and
E-cadherin and increased levels of vimentin) [87]. In Visceral Sarcomas
another study by Xu et al., E-cadherin was shown to be https://www.esmo.org/guidelines/sarcoma-and-gist/soft-
targeted and negatively regulated by the microRNA tissue-and-visceral-sarcomas
mir-9 in SS [88]. Overexpression of mir-9 enhanced lung
metastasis in vivo by downregulating E-cadherin. EMT is
a complex process with several layers of regulation, and References
further investigation of the role of this process in driving
1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer
metastasis in SS and other STS exhibiting some degree of J Clin. 2019;69(1):7–34.
epithelial differentiation is necessary. 2. Coindre J-M, et al. Predictive value of grade for metastasis devel-
opment in the main histologic types of adult soft tissue sarcomas:
a study of 1240 patients from the French Federation of Cancer
Centers Sarcoma Group. Cancer. 2001;91(10):1914–26.
60.6 Conclusion 3. Ducimetière F, et al. Incidence of sarcoma histotypes and molecular
subtypes in a prospective epidemiological study with central pathol-
Unfortunately, metastasis is a common phenomenon in STS ogy review and molecular testing. PLoS One. 2011;6(8):e20294.
https://doi.org/10.1371/journal.pone.0020294. PMID: 21826194;
patients given the high metastatic rate. Although most meta- PMCID: PMC3149593.
static STS patients will have pulmonary metastases, several 4. WHO Classification of Tumours Editorial Board. WHO classifica-
STS subtypes show atypical patterns of metastases leading to tion of tumours of soft tissue and bone. 5th ed. Lyon, France: IARC
different strategies for staging, surveillance, and treatment. In Press; 2020.
5. Roland CL, van Houdt W, Gronchi A. The landmark series: mul-
this chapter, we have highlighted different metastatic patterns timodality treatment of extremity sarcoma. Ann Surg Oncol.
in terms of anatomical location and timing of metastases in the 2020;27(10):3672–82.
course of the disease. We also showed the role of lymph node 6. Gadd MA, et al. Development and treatment of pulmonary metas-
staging and treatment of LNM in a subset of STS. tases in adult patients with extremity soft tissue sarcoma. Ann Surg.
1993;218(6):705–12.
Despite the relatively poor survival of STS patients com- 7. Pisters PWT, et al. Analysis of prognostic factors in 1,041 patients
pared to many other cancers, metastatic mechanisms are an with localized soft tissue sarcomas of the extremities. J Clin Oncol.
underexplored area in STS biology. Therefore, metastatic 1996;14(5):1679–89.
patterns of sarcoma are still poorly understood. Determining 8. van Praag VM, et al. A prediction model for treatment decisions
in high-grade extremity soft-tissue sarcomas: Personalised sarcoma
the biological mechanisms underpinning STS dissemination care (PERSARC). Eur J Cancer. 2017;83:313–23.
636 P. van der Laan et al.

9. Callegaro D, et al. Development and external validation of a 33. Lieberman PH, et al. Alveolar soft-part sarcoma. A clinico-­
dynamic prognostic nomogram for primary extremity soft tissue pathologic study of half a century. Cancer. 1989;63(1):1–13.
sarcoma survivors. EClinicalMedicine. 2019;17:100215. 34. Italiano A, Bellera C, D’Angelo S. PD1/PD-L1 targeting in
10. Billingsley KG, et al. Multifactorial analysis of the survival of advanced soft-tissue sarcomas: a pooled analysis of phase II trials.
patients with distant metastasis arising from primary extremity sar- J Hematol Oncol. 2020;13(1):55. https://doi.org/10.1186/s13045-
coma. Cancer. 1999;85(2):389–95. 020-00891-5. PMID: 32430039; PMCID: PMC7236113.
11. Chudgar NP, et al. Is repeat pulmonary metastasectomy indicated 35. Guillou L, et al. Translocation-positive low-grade fibromyxoid sar-
for soft tissue sarcoma? Ann Thorac Surg. 2017;104(6):1837–45. coma: clinicopathologic and molecular analysis of a series expand-
12. Gladdy RA, et al. Predictors of survival and recurrence in primary ing the morphologic spectrum and suggesting potential relationship
leiomyosarcoma. Ann Surg Oncol. 2013;20(6):1851–7. to sclerosing epithelioid fibrosarcoma: a study from the French
13. Pijpe J, et al. The relation between histological, tumor-biological Sarcoma Group. Am J Surg Pathol. 2007;31(9):1387–402.
and clinical parameters in deep and superficial leiomyosarcoma and 36. Folpe AL, et al. Low-grade fibromyxoid sarcoma and hyalinizing
leiomyoma. Sarcoma. 2002;6(3):105–10. https://doi.org/10.1080/1 spindle cell tumor with giant rosettes a clinicopathologic study of
357714021000065404. PMID: 18521340; PMCID: PMC2395483. 73 cases supporting their identity and assessing the impact of high-­
14. Miyajima K, et al. Clinicopathological prognostic factors in soft grade area. Am J Surg Pathol. 2000;24(10):1353–60.
tissue leiomyosarcoma: a multivariate analysis. Histopathology. 37. Evans HL. Low-grade fibromyxoid sarcoma: a clinicopathologic
2002;40:353–9. study of 33 cases with long-term follow-up. Am J Surg Pathol.
15. Wile AG, Evans HL, Romsdahl MM. Leiomyosarcoma of soft tis- 2011;35(10):1450–62.
sue: a clinicopathologic study. Cancer. 1981;48(4):1022–32. 38. Chamberlain F, et al. Low-grade fibromyxoid sarcoma: treat-
16. Wang WL, et al. Sarcoma metastases to the skin: a clinicopatho- ment outcomes and efficacy of chemotherapy. In Vivo.
logic study of 65 patients. Cancer. 2012;118(11):2900–4. 2020;34(1):239–45.
17. Lang H, et al. Hepatic metastases from leiomyosarcoma: a single-­ 39. Fong Y, et al. Lymph node metastasis from soft tissue sarcoma in
center experience with 34 liver resections during a 15-year period. adults. Analysis of data from a prospective database of 1772 sar-
Ann Surg. 2000;231(4):500–5. coma patients. Ann Surg. 1993;217(1):72–7.
18. DeMatteo RP, et al. Results of hepatic resection for sarcoma meta- 40. Daigeler A, et al. Lymph node metastases in soft tissue sarcomas - a
static to liver. Ann Surg. 2001;234(4):540–7. single center analysis of 1,597 patients. Langenbeck’s Arch Surg.
19. Burt BM, et al. Repeated and aggressive pulmonary resections for 2009;394(2):321–9.
leiomyosarcoma metastases extends survival. Ann Thorac Surg. 41. Behranwala KA, et al. Prognosis of lymph node metastasis in soft
2011;92(4):1202–7. tissue sarcoma. Ann Surg Oncol. 2004;11(7):714–9.
20. Durr HR, et al. Myxoid liposarcoma: local relapse and metastatic 42. Mazeron JJ, Suit HD. Lymph nodes as sites of metastases from sar-
pattern in 43 patients. BMC Cancer. 2018;18(1):304. comas of soft tissue. Cancer. 1987;60(8):1800–8.
21. Smith TA, Easley KA, Goldblum JR. Myxoid/round cell liposar- 43. Ecker BL, et al. Implications of lymph node evaluation in the
coma of the extremities. A clinicopathologic study of 29 cases with management of resectable soft tissue sarcoma. Ann Surg Oncol.
particular attention to extent of round cell liposarcoma. Am J Surg 2017;24(2):425–33.
Pathol. 1996;20(2):171–80. 44. Keung EZ, et al. Defining the incidence and clinical significance
22. Spillane AJ, Fisher C, Meirion Thomas J. Myxoid liposarcoma— of lymph node metastasis in soft tissue sarcoma. Eur J Surg Oncol.
the frequency and the natural history of nonpulmonary soft tissue 2018;44(1):170–7.
metastases. Ann Surg Oncol. 1999;6(4):389–94. 45. Basile G, et al. Curability of patients with lymph node
23. Shinoda Y, et al. Prognostic factors of metastatic myxoid liposar- metastases from extremity soft-tissue sarcoma. Cancer.
coma. BMC Cancer. 2020;20(1):883. 2020;126(23):5098–108.
24. Sheah K, et al. Metastatic myxoid liposarcomas: imaging and his- 46. Sherman KL, et al. Examination of national lymph node evalua-
topathologic findings. Skelet Radiol. 2008;37(3):251–8. tion practices for adult extremity soft tissue sarcoma. J Surg Oncol.
25. Visgauss JD, et al. Staging and surveillance of myxoid liposarcoma: 2014;110(6):682–8.
follow-up assessment and the metastatic pattern of 169 patients sug- 47. Johannesmeyer D, et al. The impact of lymph node disease in
gests inadequacy of current practice standards. Ann Surg Oncol. extremity soft-tissue sarcomas: a population-based analysis. Am J
2021;28(12):7903–11. https://doi.org/10.1245/s10434-021-10091-1. Surg. 2013;206(3):289–95.
PMID: 33961173. 48. Sawamura C, et al. Lymphadenectomy and histologic subtype
26. Joensuu H, Hohenberger P, Corless CL. Gastrointestinal stromal affect overall survival of soft tissue sarcoma patients with nodal
tumour. Lancet. 2013;382(9896):973–83. metastases. Clin Orthop Relat Res. 2013;471(3):926–31.
27. von Mehren M, Joensuu H. Gastrointestinal stromal tumors. J Clin 49. Blazer DG, Sabel MS, Sondak VK. Is there a role for sentinel
Oncol. 2018;36(2):136–43. lymph node biopsy in the management of sarcoma? Surg Oncol.
28. DeMatteo RP, et al. Two hundred gastrointestinal stromal tumors: 2003;12(3):201–6.
recurrence patterns and prognostic factors for survival. Ann Surg. 50. Skinner KA, Eilber FR. Soft tissue sarcoma nodal metastases: bio-
2000;231(1):51–8. logic significance and therapeutic considerations. Surg Oncol Clin
29. Mussi C, et al. Post-imatinib surgery in advanced/metastatic GIST: N Am. 1996;5(1):121–7.
is it worthwhile in all patients? Ann Oncol. 2010;21(2):403–8. 51. Cates JMM. The AJCC 8th edition staging system for soft tissue
30. Paoluzzi L, Maki RG. Diagnosis, prognosis, and treat- sarcoma of the extremities or trunk: a cohort study of the SEER
ment of alveolar soft-part sarcoma: a review. JAMA Oncol. database. J Natl Compr Canc Netw. 2018;16(2):144–52.
2019;5(2):254–60. 52. Fisher SB, et al. Comparative performance of the 7th and 8th edi-
31. Pennacchioli E, et al. Alveolar soft part sarcoma: clinical presenta- tions of the american joint committee on cancer staging systems for
tion, treatment, and outcome in a series of 33 patients at a single soft tissue sarcoma of the trunk and extremities. Ann Surg Oncol.
institution. Ann Surg Oncol. 2010;17(12):3229–33. 2018;25(5):1126–32.
32. Portera CA, et al. Alveolar soft part sarcoma: clinical course and 53. Ashamalla M, et al. Challenging AJCC 8 staging for soft tis-
patterns of metastasis in 70 patients treated at a single institution. sue sarcoma using the NCDB. Int J Radiat Oncol Biol Phys.
Cancer. 2001;91(3):585–91. 2019;105(2):338–45.
60 Atypical Patterns of Metastases: How Do Sarcomas Metastasize? 637

54. Andreou D, et al. Sentinel node biopsy in soft tissue sarcoma sub- 75. Lewis DM, et al. Intratumoral oxygen gradients mediate sarcoma
types with a high pro- pensity for regional lymphatic spread-results cell invasion. Proc Natl Acad Sci U S A. 2016;133(33):9292–7.
of a large prospective trial. Ann Oncol. 2013;24(5):1400–5. 76. Francis P, et al. Diagnostic and prognostic gene expression sig-
55. Network, N.C.C. Soft Tissue Sarcoma. 2016; Available from: natures in 177 soft tissue sarcomas: hypoxia-induced transcrip-
https://www.nccn.org/professionals/physician_gls/pdf/sarcoma. tion profile signifies metastatic potential. BMC Genomics.
pdf. 2007;8(1):73.
56. Gaballah AH, et al. Angiosarcoma: clinical and imaging features 77. Masoud GN, Li W. HIF-1α pathway: role, regulation and interven-
from head to toe. Br J Radiol. 2017;90(1075):20170039. tion for cancer therapy. Acta Pharm Sin B. 2015;5(5):378–89.
57. Young RJ, et al. Angiosarcoma. Lancet Oncol. 2010;11(10):983–91. 78. Du H, et al. PLOD2 in cancer research. Biomed Pharmacother.
58. Kang Y, et al. Regional lymph node metastasis of scalp angiosar- 2017;90:670–6.
coma: a detailed clinical observation study of 40 cases. Ann Surg 79. Eisinger-Mathason TSK, et al. Hypoxia-dependent modification of
Oncol. 2020;27(8):3018–27. collagen networks promotes sarcoma metastasis. Cancer Discov.
59. Torres KE, et al. Long-term outcomes in patients with 2013;3(10):1190–205.
radiation-­associated angiosarcomas of the breast following 80. Lewis DM, et al. A feedback loop between hypoxia and matrix
surgery and radiotherapy for breast cancer. Ann Surg Oncol. stress relaxation increases oxygen-axis migration and metastasis in
2013;20(4):1267–74. sarcoma. Cancer Res. 2019;79(8):1981–95.
60. Touati N, et al. European organisation for research and treatment 81. Estourgie SH, Nielsen GP, Ott MJ. Metastatic patterns of extrem-
of cancer soft tissue and bone sarcoma group experience with ity myxoid liposarcoma and their outcome. J Surg Oncol.
advanced/metastatic epithelioid sarcoma patients treated in pro- 2002;80(2):89–93.
spective trials: clinical profile and response to systemic therapy. 82. Asano N, et al. Metastatic patterns of myxoid/round cell liposar-
Clin Oncol. 2018;30(7):448–54. coma: a review of a 25-year experience. Sarcoma. 2012;2020:1–6.
61. Bianchi G, et al. Clear cell sarcoma of soft tissue: a retrospec- 83. Lee ATJ, et al. Clinical and molecular spectrum of liposarcoma. J
tive review and analysis of 31 cases treated at Istituto Ortopedico Clin Oncol. 2018;36(2):151–9.
Rizzoli. Eur J Surg Oncol. 2014;40(5):505–10. https://doi. 84. Crozat A, et al. Fusion of CHOP to a novel RNA-binding protein in
org/10.1016/j.ejso.2014.01.016. human myxoid liposarcoma. Nature. 1993;363(3430):640–4.
62. Sanfilippo R, et al. Myxofibrosarcoma: prognostic factors and sur- 85. Patil N, et al. A mechanistic study on the metastasis inducing func-
vival in a series of patients treated at a single institution. Ann Surg tion of FUS-CHOP fusion protein in liposarcoma. Int J Cancer.
Oncol. 2011;18(3):720–5. 2014;134(12):2808–19.
63. Wagner LM, et al. Detection of lymph node metastases in pediatric 86. Tornin J, et al. FUS-CHOP promotes invasion in myxoid lipo-
and adolescent/young adult sarcoma: Sentinel lymph node biopsy sarcoma through a SRC/FAK/RHO/ROCK-dependent pathway.
versus fludeoxyglucose positron emission tomography imaging-A Neoplasia. 2018;20(1):44–56.
prospective trial. Cancer. 2017;123(1):155–60. 87. Peng C, et al. SHCBP1 promotes synovial sarcoma cell metastasis
64. Nakano K, Takahashi S. Translocation-related sarcomas. Int J Mol via targeting TGF-β1/Smad signaling pathway and is associated
Sci. 2018;19(12):3784. with poor prognosis. J Exp Clin Cancer Res. 2017;36(1):141.
65. Rodeberg DA, et al. Prognostic significance and tumor biology 88. Xu X-Z, et al. MiR-9 promotes synovial sarcoma cell migration
of regional lymph node disease in patients with rhabdomyosar- and invasion by directly targeting CDH1. Int J Biochem Cell Biol.
coma: a report from the Children’s Oncology Group. J Clin Oncol. 2019;112:61–71.
2011;29(10):1304–11. 89. Qi Y, et al. Transforming growth factor-β1 signaling promotes
66. Meza JL, et al. Analysis of prognostic factors in patients epithelial-mesenchymal transition-like phenomena, cell motil-
with nonmetastatic rhabdomyosarcoma treated on Intergroup ity, and cell invasion in synovial sarcoma cells. PLoS One.
Rhabdomyosarcoma Studies III and IV: the Children’s Oncology 2017;12(8):e0182680.
Group. J Clin Oncol. 2006;24(24):3844–51. 90. Dongre A, Weinberg RA. New insights into the mechanisms of
67. Ashamalla M, et al. Clinical presentation and patterns of care epithelial-mesenchymal transition and implications for cancer. Nat
in SCARE Soft Tissue Sarcoma. Int J Radiat Oncol Biol Phys. Rev Mol Cell Biol. 2019;20(2):69–84.
2017;99(2S):E749. 91. Olsen SH, Thomas DG, Lucas DR. Cluster analysis of immunohis-
68. Lambert AW, Pattabiraman DR, Weinberg RA. Emerging biologi- tochemical profiles in synovial sarcoma, malignant peripheral nerve
cal principles of metastasis. Cell. 2017;168(4):670–91. sheath tumor, and Ewing sarcoma. Mod Pathol. 2006;19(5):659–68.
69. Fidler IJ. Metastasis: quantitative analysis of distribution and fate 92. Sato H, et al. Expression of cadherins and their undercoat proteins
of tumor emboli labeled with 125 I-5-iodo-2′-deoxyuridine. J Natl (alpha-, beta-, and gamma-catenins and p120) and accumulation of
Cancer Inst. 1970;45(4):773–82. beta-catenin with no gene mutations in synovial sarcoma. Virchows
70. Tang YJ, et al. Tracing tumor evolution in sarcoma reveals clonal Arch. 2001;438(1):23–30.
origin of advanced metastasis. Cell Rep. 2019;28(1):2837–50. 93. Qi Y, et al. The correlation between morphology and the expression
71. Gao Y, et al. Metastasis organotropism: redefining the congenial of TGF-β signaling pathway proteins and epithelial-mesenchymal
soil. Dev Cell. 2019;49(3):375–91. transition-related proteins in synovial sarcomas. Int J Clin Exp
72. Paget S. The distribution of secondary growths in cancer of the Pathol. 2013;6(12):2787–99.
breast. Lancet. 1889;133(3421):571–3. 94. Saito T, et al. Prognostic value of the preserved expression
73. Weiss L. Comments on hematogenous metastatic patterns in humans of the E-cadherin and catenin families of adhesion molecules
as revealed by autopsy. Clin Exp Metastasis. 1992;10(3):191–9. and of beta-catenin mutations in synovial sarcoma. J Pathol.
74. Brizel DM, et al. Tumor oxygenation predicts for the likelihood 2000;192(3):342–50.
of distant metastases in human soft tissue sarcoma. Cancer Res.
1996;56(5):941–3.
 Tumor Immune Microenvironment
of Soft Tissue Sarcoma 61
Tom Wei-Wu Chen, Sheng-Fang Su, and William W. Tseng

Abstract Learning Objectives

Soft tissue sarcoma (STS) is a group of cancers trans- • The presence of B cells and tertiary lymphoid struc-
formed from mesenchymal cells. The treatment out- tures in the tumor immune microenvironment of
come of immunotherapy has not been as successful in soft tissue sarcoma are associated with improved
STS, and our understanding of the tumor immune prognosis and is considered as the engine of anti-­
microenvironment (TIME) may help us in the search of tumor immune response.
better treatment. This chapter aims to describe the • Tumor-associated macrophages are the most abun-
tumor immune microenvironment of soft tissue sarcoma dant immune cell types in the tumor immune micro-
in detail in terms of specific immune cell types as well environment of various soft tissue sarcoma
as specific histologies. The topics included B cells and histologies. Anti-macrophage targets such as CD47
tertiary lymphoid structures, T cells, macrophages, and and CSF-1R are under intensive investigations.
the effects of anticancer treatments on the TIME. In the • Understanding the alternative effect of anticancer
latter part of this chapter, we will also discuss about the treatments on the tumor immune microenvironment
TIME of leiomyosarcoma, liposarcoma, and epigenetic modulation and dynamics provides a broader per-
regulation-­ associated STS and ways to improve the spective for future combination treatments with
treatment. immunotherapy.
• Different histologies in soft tissue sarcoma, such as
leiomyosarcoma, liposarcoma, and epigenetic
change-related sarcomas, may need different mech-
anisms to overcome the immunosuppressive tumor
immune microenvironment.
T. W.-W. Chen (*)
Graduate Institute of Oncology, College of Medicine, National
Taiwan University, Taipei, Taiwan
Department of Oncology, National Taiwan University Hospital,
Taipei, Taiwan 61.1 Introduction
e-mail: [email protected]
S.-F. Su Soft tissue sarcomas (STS) are tumors that arise from the
Graduate Institute of Oncology, College of Medicine, National mesenchymal cells. It comprised about 1–2% of all solid
Taiwan University, Taipei, Taiwan tumors but could occur in any body parts. Furthermore, STS
YongLin Institute of Health, National Taiwan University (National is not a single disease entity but is composed of over more
Taiwan University YongLin Scholar), Taipei, Taiwan than 70 different histologies. The advancement in molecular
e-mail: [email protected] biology has provided useful tools in understanding the
W. W. Tseng underlying causes of various STS histologies. In terms of
Department of Surgery, Division of Breast, Endocrine and Soft genomic alterations, STS could be parsed into those with
Tissue Surgery, Keck School of Medicine of University of
Southern California, Los Angeles, CA, USA simple pathognomonic genetic rearrangements that could
e-mail: [email protected] define a specific histology, for instance, the presence of c-kit

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 639
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_61
640 T. W.-W. Chen et al.

and platelet-derived growth factor receptor (PDGFR) in gas- popular for TIME studies; the openness and public availabil-
trointestinal stromal tumors (GIST), SS18-SSX gene rear- ity of the transcriptomic material also made inter-study com-
rangement in synovial sarcoma, or MDM2 gene amplification parisons achievable. Although many algorithms for
in well-differentiated and dedifferentiated liposarcoma transcriptomic-based TIME analysis exist, these algorithms
(LPS) or those with a complex genomic alteration or aneu- could be categorized into two conceptually distinct
ploidy such as leiomyosarcoma and angiosarcoma. However, approaches: the marker-gene-based or deconvolution-based
tumor is not made of only cancer cells; we now understand approach [5]. Marker-gene-based approach utilizes a pre-­
that the tumor microenvironment is also vital to tumorigen- specified set of genes that were validated as characteristics of
esis as well as treatment resistance. Compared to our under- the representative immune cells. Nevertheless, marker-gene-­
standing of the molecular mechanisms that drive STS based algorithms are mostly semiquantitative, allowing
cancers, what we comprehend about the tumor microenvi- inter-samples comparisons of a specific immune cell type but
ronment, or more specifically, the tumor immune microenvi- not intra-sample comparison of different immune cell types
ronment (TIME), is still very rudimentary. This is reflected [5]. Deconvolution methods use a system of equations to
in the fact that, as compared to other cancer types, immuno- decipher the sample gene expression profile as the weighted
therapy, or mostly immune checkpoint inhibitor (ICI), treat- sum of expression profiles of different immune cell types. In
ments in STS have not been as successful. To improve the principle, deconvolution method allows the generation of
outcome of immunotherapy in STS, deeper understanding of absolute scores that could be interpreted as different immune
the TIME cannot be overemphasized as it has been proven to cell population fractions, providing detail intratumor immune
be correlated with finding a better solution for ICIs in many cell type proportions that could be compared between sam-
other cancer types [1]. This chapter aims to provide the cur- ples as fractions but not absolute values [6]. Although the
rent understanding of the TIME of STS in terms of specific analytical results of different algorithms have shown satis-
immune cell types, tertiary lymphoid structures, and factory correlation on CD8+ T cells, B cells, and NK cells,
histologies. the correlation of other immune cell types such as regulatory
and non-regulatory CD4+ T cells and dendritic cells remains
to be improved [5]. Overall, it is essential to understand the
61.2  ommon Methods Used for Tumor
C computational algorithms used for the analysis of TIME
Immune Microenvironment before we can fully interpret and compare the different
Investigation results.

It is essential to first understand the methodologies used for

the investigation of TIME, as every methodology has its 61.3  verview of Tumor Immune
O
strengths and pitfalls. The most common and easily Microenvironment in Soft Tissue
approached method to assess the TIME is by immunohisto- Sarcoma
chemistry (IHC). Two methods are commonly used: a full
tissue section examination or tissue microarray. The full tis- Much heterogeneity is noted in the TIME of any cancer type,
sue section is a more conventional approach that resembles but STS are more likely to have a “cold” TIME that lacks
clinical pathological examination but is limited by resolution immune cell infiltrations. One of the reasons is that STS have
for multiple chromogens and the lack of standards in inter- in average a lower tumor mutation burden (TMB) as com-
pretation. Tissue microarray uses a tiny proportion of the tis- pared to other carcinoma cancer types [7]. A higher TMB is
sue but permits high-throughput analysis that could involve associated with an increased chance of presenting neoanti-
tens or hundreds of specimens in arraying them into paraffin gens and hence inducing adaptive immunity [8]. The mean
blocks. Tissue microarray saves precious tumor material but somatic mutation in The Cancer Genome Atlas (TCGA)
may be biased by intratumoral heterogeneity in determina- SARC cohort was 1.06 per MB of DNA, while in contrast
tion of the true outcome [2]. Recently, multispectral IHC or the TMB of melanoma on average was 14 per Mb of DNA
immunofluorescence staining with up to seven colors on a [9]. In addition, STS are more likely to have unbalanced
single slide has gained popularity and in use for TIME analy- copy number alterations (CNA), which have been associated
sis [3, 4]. Although flow cytometry is also an important with a lower cytotoxic immune cell infiltration in the TIME
method to determine immune cell type in the lab, the require- [10]. The genomic feature of STS histology is also associ-
ment of fresh tissue and laborious effort to filter the ated with different types of TIME. STS with simple karyo-
­tumor-­infiltrating immune cells have limited its clinical or types or genomic alterations such as synovial sarcoma or
translational applicability. alveolar rhabdomyosarcoma commonly have a “cold” TIME
The advent and reduced cost of next-generation sequenc- with fewer immune cell infiltrations [11, 12]. On the con-
ing have made transcriptomic-based (RNA-Seq) analysis trary, STS with a complex karyotype or genomic alterations
61 Tumor Immune Microenvironment of Soft Tissue Sarcoma 641

such as undifferentiated pleomorphic sarcoma (UPS) or productions and, along with dendritic cells and T follicular
angiosarcoma (AS) are more likely to feature a TIME with helper (Tfh) cells, to facilitate the maturation of effector T
higher infiltrating lymphocytes [13, 14]. cells [18].
A recent joint collaborative study by sarcoma research- A positive effect of B cells on STS outcomes was first
ers from France, Taiwan, and the United States that amassed reported by Sorbye et al. They used tissue microarray to
936 STS patients’ transcriptomic or IHC data provided a examine the TIME of 249 primary STS. In this study, CD45,
broader view of the TIME of STS [15]. Based on the level CD3, CD4, CD8, and CD20 were used as markers for leuko-
of different immune cell types and endothelial cell marker cytes, T cells, CD4+ T cell, cytotoxic T cells, and B cells,
expressions in the TIME as computed by the respectively. In the multivariate model, CD20 positivity was
Microenvironment Cell Populations-counter (MCP- the only significant positive prognostic marker associated
counter) [16], STS could be categorized into five distinct with better disease-specific survival for STS patients who
Sarcoma Immune Classes (SIC). SICs A, C, and E are more had a wide resection margin (HR 5.5 95% CI 1.6–18.6,
homogeneous and stand for “immune-desert,” “vascular- p = 0.006) [19]. Another study from Karolinska Institute also
ized,” and “immune and tertiary lymphoid structure (TLS) supported the favorable role of B cells in STS. Primary STS
high.” The overall survival (OS) was significantly better for patients with higher CD20+ B cells were associated with bet-
SIC E group as compared to SIC A group. Around 70% of ter metastasis-free survival (61 vs 29 months, p = 0.009) and
STS patients (including SICs A, B, and C) were considered overall survival (OS; 60 vs 34, months, p = 0.022). However,
to have a “cold” TIME with few immune cell infiltrates. in the validation TCGA cohort, the favorable prognostic role
Immune checkpoint molecules were also reflective of the of B cells was only observed in STS tumor with low M2
level of inflammation in each SIC, with only the SIC E hav- (protumoral) macrophages, represented by low IL-10, low
ing concomitantly a high expression of multiple immune PTGS2, or low CD163 expression [20].
checkpoint markers including CTLA-4, PD-1, LAG3, and The concept that in STS the presence of active B cells is
TIM3. Intriguingly, the number of non-­synonymous muta- critical to support the cytotoxic T cell function is further
tions in the STS patients were not correlated with the SIC strengthened in the study by Petitprez et al. [15]. Using the
groups, suggesting that TMB may not be the only factor to transcriptomic data from 496 STS patients, we confirmed
determine the TIME of STS. that the B lineage signature (including genes BANK1, CD19,
In the following sections, we discuss in more detail the CD22, CD79A, CR2, FCRL2, IGKC, MS4A1, and PAX5)
roles of different immune cell types and specific immune plays a more prominent role than the T cell signature (includ-
modulating molecules in the TIME and our understandings ing genes CD28, CD3D, CD3G, CD5, CD6, CHRM3-AS2,
in specific STS histologies. CTLA4, FLT3LG, ICOS, MAL, PBX4, SIRPG, THEMIS,
TNFRSF25, and TRAT1) or cytotoxic lymphocyte signature
(including genes CD8A, EOMES, FGFBP2, GNLY, KLRC3,
61.4  he Role of Specific Immune Cell
T KLRC4, KLRD1) in determining the OS outcome. Patients
Types in the Tumor Microenvironment with high cytotoxic lymphocytes/low B lineage or high T
cell/low B lineage cells had a worse OS than patients with
61.4.1 B Cells/Tertiary Lymphoid Structures high cytotoxic lymphocytes/high B lineage (p = 0.015) or
high T cells/high B lineage (p = 0.02), respectively [15].
It may be unorthodox to start with a discussion about B cells A plausible explanation for the positive impact of B cells
in TIME rather than the T cells. After all, the cytotoxic T in the TIME of STS is its association with tertiary lymphoid
cells are considered the top-tier immune cells responsible for structure (TLS). TLS are observed in chronic inflammation
tumor killing. B cells are mostly responsible for the secretion sites and malignant cancers. The formation of TLS requires
of antibody as an adaptive immunity mechanism. The anti- both lymphoid tissue organizer (LTo) cells and lymphoid tis-
body binds to the surface antigens and induces downstream sue inducer (LTi) cells. Where stromal tissues in chronic
complement system activation or antibody-dependent cellu- inflamed sites act as LTo, tissue-resident monocytes, B cells,
lar cytotoxicity to kill the targeted cells. However, B cells and Th17 cells could function as LTi in the same microenvi-
also play an important role as an antigen-presenting cell that ronment. The interaction between LTo and LTi stimulates the
could facilitate the maturation of adaptive cytotoxic CD8 T secretion of chemokines that function as lymphocyte attrac-
cells. The maturation process mostly occurs in the germinal tant (CCL21 and CXCL12) and also molecules that govern
center (GC) in the secondary lymphoid organs such as the the structure formation of TLS (CXCL13, CCL19, adhesion
lymph node, tonsils, or spleen. When these GC arise in molecules) [18]. Within TLS microenvironment, mature
organs of chronic inflammation or tumorigenesis, these lym- dendritic cells (DC) present major histocompatibility com-
phoid follicles are referred to as tertiary lymphoid structures plex (MHC) class II peptides to CD4+ T cells and facilitate
(TLS) [17]. It is postulated that TLS is responsible for B cell their differentiation into effector T cells and Tfh cells. Mature
642 T. W.-W. Chen et al.

Tfh cells express CXCL13 receptor CXCR5, facilitating its have a consensus interpretation [25]. For instance, the prog-
migration into the CXCL13-rich B cell zone. The contact of nostic impact of CD4+ or CD8+ T cells has remained diver-
Tfh cells and B cells induces the activation and differentia- gent and unsettled in STS [19, 26, 27]. However, it is
tion of B cells into plasmablasts and the formation of GC in generally accepted that STS, as compared to other immuno-
which the B cells undergo class switch recombination [18]. genic cancers such as melanoma or lung cancer, has fewer
The prognostic impact of TLS has been demonstrated in CD8+ T cells [25, 28]. In addition, CD8+ T cells are more
various cancer types including STS [21]. In a cohort of 93 likely to be higher in genomic-complex STS than simple
STS samples including dedifferentiated liposarcoma, leio- fusion gene-associated STS [12]. Memory effector T cells
myosarcoma, and UPS, 12% (11/93) had at least 1 TLS as are matured T cells that could rapidly respond to antigen
defined by the presence of CD20+ B cell follicles juxtaposed stimulation and has been shown to be associated with a better
to a CD3+ T cell aggregates containing at least 1 DC-LAMP+ prognosis in a variety of cancer types [29, 30]. In STS, pleo-
mature dendritic cell as determined by IHC [15]. It was con- morphic sarcoma and myxofibrosarcoma had the highest
firmed that all but 1 STS patient with TLS was in the SIC E percentage of memory effector T cells among CD8+ T cells,
group and 9 out of 11 patients in the SIC E group had TLS, while GIST had the lowest memory effector T cells [31]. A
suggesting the potential of TLS as a surrogate biomarker for comprehensive discussion of all the subtypes of T cells is out
the identification of SIC E in STS. Indeed, in the same vali- of the scope of this chapter, but it is safe to say we need to
dation cohort, the association of TLS and the presence of identify the most important T cell subtype that is associated
anti-tumor immunity were confirmed: STS with TLS had a with better prognosis in each STS histology and understand
significantly higher CD3+ cell density (p = 4.0 × 10–5), CD8+ the methods to enhance its infiltration or trafficking into the
cell density (p = 1.8 × 10–4), and CD20+ cell density TIME.
(p = 1.5 × 10–7) as compared to those without TLS [15]. The
favorable prognostic impact of TLS was also confirmed in
GIST [22] and retroperitoneal sarcomas [23]. 61.4.3 Macrophages
Different stages of TLS maturation may also signal differ-
ent levels of anti-tumor immunity [18]. Based on the expres- The most common myeloid immune cell type found in the
sion of two different markers, CD21 on follicular dendritic TIME of STS is macrophage [32]. Under normal physiologi-
cells and CD23 on both B cells and dendritic cells, the matu- cally, macrophages are responsible for pathogen killing.
ration stages of TLS have been proposed as early lymphoid However, in the TIME, many macrophages differentiate into
aggregates (E-TLS) that lack both CD21 and CD23 expres- tumor-associated macrophages (TAMs), retaining features of
sion, primary follicle-like TLS (PFS-TLS) with CD21-­ M2 macrophages such as promoting angiogenesis, breaking
positive but CD23-negative expression, and secondary down of matrix proteins to facilitate tumor metastasis, and
follicle-like TLS (SFL-TLS) containing both matured CD21-­ inducing an immunosuppressive TIME by the secretion of
positive dendritic cells and CD23-positive B cells in the GC IL10 and TGF-β [32, 33].
[18]. Although it is not common practice to differentiate TLS Studies using either IHC or tissue microarrays have
in terms of its maturity stage, the differential impact of the shown that TAMs are the most abundant immune cell type in
maturity stages of TLS has been shown in non-small-cell the TIME of many STS histologies [11, 34]. Although the
lung cancer and colon cancer [24], as well as in GIST [22]. typical representative marker remains debatable, CD68 and
In the study by Petitprez et al., E-TLS, PFL-FLS, and SFL-­ CD163 are the two most common markers to identify TAMs.
TLS represented 60.5%, 21.1%, and 18.3% of TLS exam- While CD68 is a marker for pan-macrophage identification,
ined, and the most common histology with SFL-TLS was its specificity is low as stromal cells and sarcoma cells could
UPS (p = 7.76 × 10–5) [15] . also express CD68 [32]. CD163 is a membranous marker
In summary, B cells and TLS in the TIME of STS affect associated with M2 macrophage which is the common form
the prognosis. What remains to be investigated is how to of TAM. In addition, functional studies have demonstrated
induce the formation of TLS in different STS histologies and that mice with CD163-deficient TAM had significantly
prospective studies to determine the surrogacy of TLS as a decreased sarcoma tumor growth as compared with mice
biomarker for ICI treatment success. with CD163 wild-type TAM. In this model, the production of
IL6 from CD163-positive macrophage is the determinant
chemokine in facilitating tumor growth [35]. Based on these
61.4.2 T Cells evidences, CD163 is likely a more specific marker to deter-
mine M2-like TAM in the TIME of STS.
Although T cells remain an integral part in assessing the anti-­ In the TIME of STS tumors, the quantity of TAM varies.
tumor response in the field of oncology, distinct and partial UPS patients are most likely to have the highest number of
overlapping assessment methods make it a difficult task to TAMs per tumor area or the highest ratio of CD163+ to
61 Tumor Immune Microenvironment of Soft Tissue Sarcoma 643

CD68+ macrophages, but the variation is also wide [11, Despite the same histology and similar TMB of the trans-
34]. Other STS histologies with higher TAMs included plantable and autochthonous tumors, the differential effects
angiosarcoma, dedifferentiated liposarcoma, and leiomyo- of radiotherapy suggested that the immune-modulatory
sarcoma [34]. Generally speaking, genomic complex STS effect of radiotherapy also depends on the original TIME sta-
have a higher number of TAMs than translocation-associ- tus [39]. Currently, the SARC032 (NCT03092323) study is
ated STS in the TIME. Another possible resistant mecha- investigating the role of neoadjuvant radiotherapy plus pem-
nism for tumor to escape the phagocytic activity of brolizumab in localized high-grade STS. The outcome of
macrophages in the TIME is inducing the expression of this study will provide further evidence of the immunomodu-
CD47. CD47 is an ubiquitous marker on erythrocytes to latory effect of radiotherapy and combination with ICI
prevent macrophage phagocytosis. Tumor cell could upreg- in localized STS.
ulate CD47 and, when interacts with signal-regulatory pro- Although chemotherapy-induced immunogenic cell death
tein alpha (SIRPα) on macrophages, suppresses the has been proposed for a long period of time [40], the effect
anti-tumor activity of macrophages [36]. The distribution of cytotoxic chemotherapy such as doxorubicin and ifos-
of CD47 on tumor cells is bimodal, with most cases show- famide on transforming an immunosuppressive TIME
ing either 0% or higher than 90% of tumor cells positive remains uncertain. Studies examining post-chemotherapy
[34]. Sarcomas with a high percentage of CD47 tumor cell treatment STS TIME have not found either specific change
staining included chordoma, angiosarcoma, and pleomor- or even an increase in FoxP3+ TILs [12, 14]. One exception
phic liposarcoma [34]. is the cytotoxic drug trabectedin. Trabectedin is a cytotoxic
Because of the abundance of TAMs in the TIME, drugs agent that binds to the minor groove of DNA which could
targeting the specific marker are under intense investiga- inhibit transcription factor binding and forces DNA damage
tion. The leading target is colony-stimulating factor-1 [41]. In addition, it has been shown in sarcoma mouse mod-
receptor (CSF-1R), which is associated with M2 macro- els that trabectedin could apply the cytotoxic effect to TAMs
phage polarization [33]. The CSF-1R inhibitor pexidar- and subsequently decrease chemokine CCL2 secretion and
tinib has been approved by the US FDA for the treatment angiogenic processes [42]. This finding was validated in STS
of inoperable tenosynovial giant cell tumor, a rare neo- patients treated with trabectedin. After trabectedin treatment,
plasm that is associated with high levels of CSF1 and a reduction of the blood-circulating monocytes decreased, and
large amount of macrophage infiltration [37]. However, resected tumors also have decreased CD163+ macrophages
single-agent CSF-1R inhibitors have not been successful in the TIME [42, 43]. The second-generation lurbinectedin
in treating other cancer types. Combination with other also has similar effects limiting the production of inflamma-
systemic treatments such as chemotherapy or ICI is cur- tory factors that could subsequently decrease monocyte
rently under investigations [33]. adhesion and migration [44]. Targeted agents that focus on
anti-angiogenic molecules such as vascular endothelial
growth factor receptors (VEGFRs) could also facilitate the
61.4.4 The Effect of Anticancer Treatments trafficking of immune cells or maturation of dendritic cells.
on the Tumor Immune Combination of anti-angiogenic agents with ICI has shown
Microenvironment of Soft Tissue promising outcomes in advanced STS [45, 46], and biomark-
Sarcoma ers to elaborate the underlying mechanisms are eagerly
awaited.
Anticancer treatments, including conventional chemother- Lastly, immunotherapy also significantly influences the
apy, radiotherapy, and targeted therapies, not only target can- TIME dynamics. From the biomarker study of SARC 028, in
cer cells but may also assert their influence on the elements which sarcoma patients were treated with anti-PD-1 anti-
of TIME in boosting their treatment efficacy. Radiotherapy is body pembrolizumab, Keung et al. found that most of the
an integral part of the treatment for localized STS, and its immune cell type densities remained stable after pembroli-
positive impact on the TIME has been reported in various zumab treatment, except for effector cytotoxic T cells (CD3+
cancer types [38]. In STS studies, preoperative radiotherapy CD8+ CD45RO+) and regulatory T cells [47]. A phase 0
was associated with increased infiltration of tumor-­infiltrating study investigating the impact of systemic administration of
lymphocytes, CD4+ and CD8+ T cells, and FOXP3+ cells interferon-γ on the generally cold TIME of synovial sarcoma
[12]. However, some studies suggested radiotherapy did not and myxoid liposarcoma showed increased MHC class I
increase the T cell fraction or clonality as measured by T cell expression and increased T cell infiltrations after treatment
repertoire [14]. In a mouse immunocompetent sarcoma [48]. With limited efficacy of single-agent ICI in STS,
model, radiotherapy was able to ablate the myeloid immune-­ reshaping the TIME through other treatments or combina-
suppressive cells in the tumor microenvironment of trans- tion is a promising choice to go forward to improve the
plantable tumor, but not autochthonous primary tumor. efficacy.
644 T. W.-W. Chen et al.

61.5 The Tumor Immune subtypes, the TIME has been studied the most in WD/DD
Microenvironment in Specific liposarcoma. Some of the earlier descriptive studies estab-
Sarcoma Histologies lished the presence of B and T cell-rich TLS and also the
expression of PD-1 in the tumor-infiltrating T cells in WD/DD
61.5.1 Leiomyosarcoma [55]. The latter observation would have predicted favorable
response to ICI, and in fact, the initial results of the SARC028
Leiomyosarcoma (LMS) is one of the most common histolo- trial of anti-PD1 therapy showed a 20% objective response
gies of STS that could occur in extremities and truncal areas rate (ORR) for the DD cohort [50]. These data were limited by
as well as uterine or retroperitoneal areas. Nevertheless, the small number of patients enrolled with this subtype
studies have suggested that LMS from different organ sites (n = 10), and a subsequent expansion cohort study noted a
have different transcriptomic profiles and survival outcomes disappointingly lower ORR of 10% in DD (4 of 39 patients)
[49]. A high percentage of LMS has cold TIME [15] and is [56]. These results may be explained in part by the low muta-
less likely to respond to ICI treatment [50]. tional burden seen in WD/DD liposarcoma [57]. In support of
A primary focus of the TIME of LMS has been TAMs and low tumor immunogenicity, comprehensive gene expression
was pioneered by the Stanford groups led by Dr. van de Rijn. analysis in WD/DD liposarcoma shows that both TCR clonal-
They first identified the presence of a high level of TAMS in ity and PD-L1 expression are relatively low, despite an inflam-
LMS by gene expression profiling and confirmed by matory TIME [14]. Moreover, the TIME in WD/DD
IHC. The criteria for IHC identification of a macrophage liposarcoma is arguably more complex given that these tumors
were either CD68 or CD163 at the membranous, but not are naturally massive in size, that biologically distinct WD
nucleus, staining, and it needed to be a cell with dendritic (low grade) and DD (high grade) components may be juxta-
process. The first study showed that higher-dense infiltrates posed within the same tumor, and that treatment can signifi-
of either CD68+ or CD168+ macrophages were associated cantly alter the TIME—all of which contribute to significant
with a worse prognosis in non-uterine, but not uterine, LMS intratumoral and patient-to-patient heterogeneity [23, 58].
[51]. However, if a composite score (CD163, FCGR3a, and Myxoid round cell liposarcoma is defined by a FUS-­
CTSL1) was used to count the macrophages, then the pres- DDIT3 (or in a minority of cases, FUS-ESWR1) transloca-
ence of macrophages was associated with a poor prognosis tion. In contrast to the WD/DD subtype, relatively little is
in both uterine and non-uterine LMS [52]. Furthermore, known about the TIME in myxoid liposarcoma. The bulk of
TAMs also play a significant role in angiogenesis within the immunologic studies in this subtype have focused on the
tumor microenvironment of LMS. While the micro-vessel very high tumor expression of NY-ESO-1, a cancer-testis
density (MVD) was associated with a poor prognosis, it was antigen [59] . As a result, ongoing vaccine trials (e.g.,
the presence of TAMs, but not vascular endothelial growth CMB305) targeting this antigen have included a modest
factor (VEGF), that was correlated more closely with number of patients specifically with myxoid liposarcoma
MVD. In addition, the parallelism between TAM and MVD [60]. Recent preliminary results are promising with in fact
persisted from primary to metastatic tumors of LMS [53]. the longest progression-free survival (median 5.1 months) in
Treatments that aim to take advantage of the abundant the myxoid liposarcoma cohort.
TAMs in LMS have been explored. One method is to educate Importantly, the TIME is likely very different in one lipo-
the TAMs to perform anti-tumor functions by targeting the sarcoma subtype versus another, necessitating different
CD47-SIRP signaling pathway. LMS have been found to immunotherapeutic approaches. As an example of this,
have high levels of CD47, and upon treatment of anti-CD47 recent data have started to highlight a potential key role for
antibody in the in vivo mouse model, the primary tumor and macrophages in WD/DD liposarcoma pathogenesis and dis-
its metastatic tumors shrank dramatically as compared with ease progression [61]. An independent study of the TIME
the control treatment [54]. verified that among various sarcoma subtypes, DD liposar-
coma has one of the highest infiltrations of TAMs [34]. In
contrast, in the same study, TAMs were markedly sparse in
61.5.2 Liposarcoma myxoid liposarcoma.

Liposarcoma encompasses three subtypes (well differenti-

ated/dedifferentiated or WD/DD, myxoid round cell, and 61.5.3 E
 pigenetic Regulation-Associated Soft
pleomorphic—very rare) each with distinct genetic aberra- Tissue Sarcoma
tions and likely unique TIMEs.
WD/DD liposarcoma is characterized genetically by ampli- Epigenetic mechanisms have been appreciated as a driver
fication of chromosomal region 12q13-15, which includes model of tumorigenesis in mesenchymal tumors [62], given
important genes such as MDM2 and CDK4. Among the three the low mutation burden nature of STS [7]. Nacev et al.
61 Tumor Immune Microenvironment of Soft Tissue Sarcoma 645

addressed multiple levels of epigenetic dysregulation devel- apeutic intervention. The understanding of DNA methylation,
oped in different types of sarcomas through mechanisms of histone, and chromatin modifications on genes involved in
altered epigenetic writers, readers, and chromatin remodel- epigenetic pathways, as well as the potential effects on the
ers [62]. Dysfunction of enzymes involved in DNA methyla- immune status in the tumor microenvironment, may help the
tion and histone modifications, mutations in histone genes treatment of STS.
and remodeling and modifying complexes, and alterations of
chromatin pathway drive sarcomas predominantly epigene-
tic diseases. Open Questions
One of the effects on chromatin writers is loss-of-function • What are the modalities or methods currently used
mutations in the polycomb repressive complex 2 (PRC2) for the investigation of the tumor immune microen-
subunit suppressor of zeste 12 (SUZ12) or embryonic ecto- vironment? What are the pros and cons of each dif-
derm development (EED) that have been identified in ferent method?
approximately 80% of MPNST [63]. SUZ12 inactivation • What is the most important immune cell type or
causes a loss of repressive histone mark H3K27me3 and may structure that could contribute to a better anti-tumor
induce the recruitment of bromodomain proteins (BRD) and immunity in soft tissue sarcoma? Are there methods
associated transcription factors via an increase in H3K27 that could enhance the formation of these anti-­
acetylation, a transcription-activating signal [64]. The link- tumor structures?
age between loss of PRC2-associated H3K27me3 and TIME • What is the most abundant immune cell type in soft
regulation remains elusive. But studies that had examined tissue sarcoma immune microenvironment? How
the TIME differences between neurofibroma and MPNST could we transform the immunosuppressive tumor
have shown increased CD8+ T cells infiltration and tumor microenvironment by targeting these immunosup-
cell PD-L1 expression in MPNST but not neurofibroma, pressive immune cells?
which is in parallel of the loss of expression of H3K27me3 • How does the histology or their respective genomic
[65], providing indirect signal that chromatin writer muta- feature affect the tumor immune
tions could affect the TIME components through epigenetic microenvironment?
regulation. • What are the anticancer treatments in soft tissue
In contrast to MPNST, some STS histologies have hyper- sarcoma that could affect the dynamics of the
activated PRC2 function. However, the hyperactivation of immune microenvironment? How should they be
PRC2 is not the result of the gaining of activation mutation combined with other immunotherapies and under
but is because of the loss of function of PRC2 suppressors – which clinical scenario would fit the best?
the SWItch/Sucrose Non-Fermentable (SWI/SNF) com-
plexes [66]. SMARCB1 (INI-1), a core subunit of SWI/SNF,
for instance, is frequently inactivated in malignant rhabdoid
tumors (MRT) and epithelioid sarcoma (ES). The loss of References
INI-1 implies the utility of inhibitor of enhancer of zeste
1. Binnewies M, Roberts EW, Kersten K, et al. Understanding the
homolog 2 (EZH2), the catalytic subunit of PRC2, for epi-
tumor immune microenvironment (TIME) for effective therapy. Nat
genetic therapy. In 2020, an open-label, phase 2 basket study Med. 2018;24:541–50.
confirmed the clinical activity and safety of tazemetostat, a 2. Lee ATJ, Chew W, Wilding CP, et al. The adequacy of tissue micro-
first-in-class EZH2 inhibitor, in advanced ES with loss of arrays in the assessment of inter- and intra-tumoural heterogene-
ity of infiltrating lymphocyte burden in leiomyosarcoma. Sci Rep.
INI1/SMARCB1 [67]. Furthermore, recent work showed a
2019;9:14602.
correlation of SMARCB1 loss with global hypomethylation 3. Stack EC, Wang C, Roman KA, Hoyt CC. Multiplexed immuno-
and cytotoxic T cell infiltration in the immune microenviron- histochemistry, imaging, and quantitation: a review, with an assess-
ment, followed by the expression of immune checkpoint ment of Tyramide signal amplification, multispectral imaging and
multiplex analysis. Methods. 2014;70:46–58.
regulators [68]. Targeting EZH2 has shown to enhance anti-
4. Edin S, Kaprio T, Hagstrom J, et al. The prognostic importance of
gen presentation and anti-tumor immunity in some pre-­ CD20(+) B lymphocytes in colorectal cancer and the relation to
clinical models of cancers resistant to immunotherapy [69, other immune cell subsets. Sci Rep. 2019;9:19997.
70]. A recent case study of dedifferentiated chordoma with 5. Sturm G, Finotello F, Petitprez F, et al. Comprehensive evalua-
tion of transcriptome-based cell-type quantification methods for
INI-1 loss treated with tazemetostat showed increased T cell immuno-oncology. Bioinformatics. 2019;35:i436–45.
infiltration into the TIME after treatment [71], providing evi- 6. Finotello F, Rieder D, Hackl H, Trajanoski Z. Next-generation
dence that inhibition of the activated EZH2 indeed could computational tools for interrogating cancer immunity. Nat Rev
reverse the immunosuppressive TIME. Genet. 2019;20:724–46.
7. Cancer Genome Atlas Research Network. Electronic address edsc,
While targeting therapy including kinase inhibitors to Cancer Genome Atlas Research N. Comprehensive and integrated
selective genetic alteration fails in most STS, targeting epi- genomic characterization of adult soft tissue sarcomas. Cell.
genetic dysregulation provides another scientific-based ther- 2017;171:950–965.e928.
646 T. W.-W. Chen et al.

8. Rizvi NA, Hellmann MD, Snyder A, et al. Cancer immunology. 29. Pages F, Berger A, Camus M, et al. Effector memory T cells,
Mutational landscape determines sensitivity to PD-1 blockade in early metastasis, and survival in colorectal cancer. N Engl J Med.
non-small cell lung cancer. Science. 2015;348:124–8. 2005;353:2654–66.
9. Iglesia MD, Parker JS, Hoadley KA, et al. Genomic analysis of 30. Samji T, Khanna KM. Understanding memory CD8(+) T cells.
immune cell infiltrates across 11 tumor types. J Natl Cancer Inst. Immunol Lett. 2017;185:32–9.
2016;108:djw144. 31. Klaver Y, Rijnders M, Oostvogels A, et al. Differential quantities of
10. Davoli T, Uno H, Wooten EC, Elledge SJ. Tumor aneuploidy cor- immune checkpoint-expressing CD8 T cells in soft tissue sarcoma
relates with markers of immune evasion and with reduced response subtypes. J Immunother Cancer. 2020;8:e000271.
to immunotherapy. Science. 2017;355:eaaf8399. 32. Mantovani A, Marchesi F, Malesci A, et al. Tumour-associated
11. Chen L, Oke T, Siegel N, et al. The immunosuppressive niche of macrophages as treatment targets in oncology. Nat Rev Clin Oncol.
soft-tissue sarcomas is sustained by tumor-associated macrophages 2017;14:399–416.
and characterized by intratumoral tertiary lymphoid structures. Clin 33. Cassetta L, Pollard JW. Targeting macrophages: therapeutic
Cancer Res. 2020;26:4018–30. approaches in cancer. Nat Rev Drug Discov. 2018;17:887–904.
12. Dancsok AR, Setsu N, Gao D, et al. Expression of lymphocyte 34. Dancsok AR, Gao D, Lee AF, et al. Tumor-associated macrophages
immunoregulatory biomarkers in bone and soft-tissue sarcomas. and macrophage-related immune checkpoint expression in sarco-
Mod Pathol. 2019;32:1772–85. mas. Oncoimmunology. 2020;9:1747340.
13. Fujii H, Arakawa A, Utsumi D, et al. CD8(+) tumor-infiltrating 35. Shiraishi D, Fujiwara Y, Horlad H, et al. CD163 is required for pro-
lymphocytes at primary sites as a possible prognostic factor of cuta- tumoral activation of macrophages in human and murine sarcoma.
neous angiosarcoma. Int J Cancer. 2014;134:2393–402. Cancer Res. 2018;78:3255–66.
14. Pollack SM, He Q, Yearley JH, et al. T-cell infiltration and clonality 36. Feng M, Jiang W, Kim BYS, et al. Phagocytosis checkpoints
correlate with programmed cell death protein 1 and programmed as new targets for cancer immunotherapy. Nat Rev Cancer.
death-ligand 1 expression in patients with soft tissue sarcomas. 2019;19:568–86.
Cancer. 2017;123:3291–304. 37. Tap WD, Gelderblom H, Palmerini E, et al. Pexidartinib versus pla-
15. Petitprez F, de Reynies A, Keung EZ, et al. B cells are associated cebo for advanced tenosynovial giant cell tumour (ENLIVEN): a
with survival and immunotherapy response in sarcoma. Nature. randomised phase 3 trial. Lancet. 2019;394:478–87.
2020;577:556–60. 38. Weichselbaum RR, Liang H, Deng L, Fu YX. Radiotherapy
16. Becht E, Giraldo NA, Lacroix L, et al. Estimating the population and immunotherapy: a beneficial liaison? Nat Rev Clin Oncol.
abundance of tissue-infiltrating immune and stromal cell popula- 2017;14:365–79.
tions using gene expression. Genome Biol. 2016;17:218. 39. Wisdom AJ, Mowery YM, Hong CS, et al. Single cell analysis
17. Buckley CD, Barone F, Nayar S, et al. Stromal cells in chronic reveals distinct immune landscapes in transplant and primary sar-
inflammation and tertiary lymphoid organ formation. Annu Rev comas that determine response or resistance to immunotherapy. Nat
Immunol. 2015;33:715–45. Commun. 2020;11:6410.
18. Sautes-Fridman C, Verneau J, Sun CM, et al. Tertiary lymphoid 40. Galluzzi L, Humeau J, Buque A, et al. Immunostimulation with
structures and B cells: clinical impact and therapeutic modulation chemotherapy in the era of immune checkpoint inhibitors. Nat Rev
in cancer. Semin Immunol. 2020;48:101406. Clin Oncol. 2020;17:725–41.
19. Sorbye SW, Kilvaer T, Valkov A, et al. Prognostic impact of lym- 41. Ueda T, Kakunaga S, Ando M, et al. Phase I and pharmacokinetic
phocytes in soft tissue sarcomas. PLoS One. 2011;6:e14611. study of trabectedin, a DNA minor groove binder, administered as
20. Tsagozis P, Augsten M, Zhang Y, et al. An immunosuppressive a 24-h continuous infusion in Japanese patients with soft tissue sar-
macrophage profile attenuates the prognostic impact of CD20-­ coma. Investig New Drugs. 2014;32:691–9.
positive B cells in human soft tissue sarcoma. Cancer Immunol 42. Allavena P, Signorelli M, Chieppa M, et al. Anti-inflammatory
Immunother. 2019;68:927–36. properties of the novel antitumor agent yondelis (trabectedin):
21. Sautes-Fridman C, Petitprez F, Calderaro J, Fridman WH. Tertiary inhibition of macrophage differentiation and cytokine production.
lymphoid structures in the era of cancer immunotherapy. Nat Rev Cancer Res. 2005;65:2964–71.
Cancer. 2019;19:307–25. 43. Germano G, Frapolli R, Belgiovine C, et al. Role of macrophage
22. Lin Q, Tao P, Wang J, et al. Tumor-associated tertiary lym- targeting in the antitumor activity of trabectedin. Cancer Cell.
phoid structure predicts postoperative outcomes in patients with 2013;23:249–62.
primary gastrointestinal stromal tumors. Oncoimmunology. 44. Belgiovine C, Bello E, Liguori M, et al. Lurbinectedin
2020;9:1747339. reduces tumour-associated macrophages and the inflammatory
23. Yan L, Wang Z, Cui C, et al. Comprehensive immune characteriza- tumour microenvironment in preclinical models. Br J Cancer.
tion and T-cell receptor repertoire heterogeneity of retroperitoneal 2017;117:628–38.
liposarcoma. Cancer Sci. 2019;110(10):3038–48. 45. Wilky BA, Trucco MM, Subhawong TK, et al. Axitinib plus pem-
24. Posch F, Silina K, Leibl S, et al. Maturation of tertiary lymphoid brolizumab in patients with advanced sarcomas including alveolar
structures and recurrence of stage II and III colorectal cancer. soft-part sarcoma: a single-centre, single-arm, phase 2 trial. Lancet
Oncoimmunology. 2018;7:e1378844. Oncol. 2019;20:837–48.
25. Martin-Broto J, Moura DS, Van Tine BA. Facts and hopes 46. Martin-Broto J, Hindi N, Grignani G, et al. Nivolumab and suni-
in immunotherapy of soft-tissue sarcomas. Clin Cancer Res. tinib combination in advanced soft tissue sarcomas: a multicenter,
2020;26:5801–8. single-­arm, phase Ib/II trial. J Immunother Cancer. 2020;8:e001561.
26. D'Angelo SP, Shoushtari AN, Agaram NP, et al. Prevalence of 47. Keung EZ, Burgess M, Salazar R, et al. Correlative analyses of the
tumor-infiltrating lymphocytes and PD-L1 expression in the soft SARC028 trial reveal an association between sarcoma-associated
tissue sarcoma microenvironment. Hum Pathol. 2015;46:357–65. immune infiltrate and response to pembrolizumab. Clin Cancer
27. van Erp AEM, Versleijen-Jonkers YMH, Hillebrandt-Roeffen MHS, Res. 2020;26:1258–66.
et al. Expression and clinical association of programmed cell death- 48. Zhang S, Kohli K, Black RG, et al. Systemic interferon-gamma
­1, programmed death-ligand-1 and CD8(+) lymphocytes in primary increases MHC class I expression and T-cell infiltration in cold
sarcomas is subtype dependent. Oncotarget. 2017;8:71371–84. tumors: results of a phase 0 clinical trial. Cancer Immunol Res.
28. Toulmonde M, Penel N, Adam J, et al. Use of PD-1 targeting, mac- 2019; https://doi.org/10.1158/2326-­6066.CIR-­18-­0940.
rophage infiltration, and IDO pathway activation in sarcomas: a 49. Hemming ML, Fan C, Raut CP, et al. Oncogenic gene-expression
phase 2 clinical trial. JAMA Oncol. 2018;4:93–7. programs in leiomyosarcoma and characterization of conven-
61 Tumor Immune Microenvironment of Soft Tissue Sarcoma 647

tional, inflammatory, and uterogenic subtypes. Mol Cancer Res. locally advanced, relapsed, or metastatic cancer expressing
2020;18:1302–14. NY-ESO-1. Oncoimmunology. 2020;9:1847846.
50. Tawbi HA, Burgess M, Bolejack V, et al. Pembrolizumab in 61. Casadei L, Calore F, Braggio DA, et al. MDM2 derived from dedif-
advanced soft-tissue sarcoma and bone sarcoma (SARC028): a ferentiated liposarcoma extracellular vesicles induces MMP2 pro-
multicentre, two-cohort, single-arm, open-label, phase 2 trial. duction from preadipocytes. Cancer Res. 2019;79:4911–22.
Lancet Oncol. 2017;18:1493–501. 62. Nacev BA, Jones KB, Intlekofer AM, et al. The epigenomics of
51. Lee CH, Espinosa I, Vrijaldenhoven S, et al. Prognostic signifi- sarcoma. Nat Rev Cancer. 2020;20(10):608–23.
cance of macrophage infiltration in leiomyosarcomas. Clin Cancer 63. Lee W, Teckie S, Wiesner T, et al. PRC2 is recurrently inactivated
Res. 2008;14:1423–30. through EED or SUZ12 loss in malignant peripheral nerve sheath
52. Espinosa I, Beck AH, Lee CH, et al. Coordinate expression of tumors. Nat Genet. 2014;46:1227–32.
colony-stimulating factor-1 and colony-stimulating factor-1-related 64. Filippakopoulos P, Picaud S, Mangos M, et al. Histone recognition
proteins is associated with poor prognosis in gynecological and non- and large-scale structural analysis of the human bromodomain fam-
gynecological leiomyosarcoma. Am J Pathol. 2009;174:2347–56. ily. Cell. 2012;149:214–31.
53. Espinosa I, Edris B, Lee CH, et al. CSF1 expression in nongyneco- 65. Shurell E, Singh AS, Crompton JG, et al. Characterizing the
logical leiomyosarcoma is associated with increased tumor angio- immune microenvironment of malignant peripheral nerve sheath
genesis. Am J Pathol. 2011;179:2100–7. tumor by PD-L1 expression and presence of CD8+ tumor infiltrat-
54. Edris B, Weiskopf K, Volkmer AK, et al. Antibody therapy targeting ing lymphocytes. Oncotarget. 2016;7:64300–8.
the CD47 protein is effective in a model of aggressive metastatic 66. Wilson BG, Wang X, Shen X, et al. Epigenetic antagonism between
leiomyosarcoma. Proc Natl Acad Sci U S A. 2012;109:6656–61. polycomb and SWI/SNF complexes during oncogenic transforma-
55. Tseng WW, Malu S, Zhang M, et al. Analysis of the intratumoral tion. Cancer Cell. 2010;18:316–28.
adaptive immune response in well differentiated and dedifferenti- 67. Gounder M, Schoffski P, Jones RL, et al. Tazemetostat in advanced
ated retroperitoneal liposarcoma. Sarcoma. 2015;2015:547460. epithelioid sarcoma with loss of INI1/SMARCB1: an international,
56. Burgess MA, Bolejack V, Schuetze S, et al. Clinical activity of open-label, phase 2 basket study. Lancet Oncol. 2020;21:1423–32.
pembrolizumab (P) in undifferentiated pleomorphic sarcoma (UPS) 68. Chun HE, Johann PD, Milne K, et al. Identification and analy-
and dedifferentiated/pleomorphic liposarcoma (LPS): final results ses of extra-cranial and cranial rhabdoid tumor molecular sub-
of SARC028 expansion cohorts. J Clin Oncol. 2019;37:11015. groups reveal tumors with cytotoxic T cell infiltration. Cell Rep.
57. Beird HC, Wu CC, Ingram DR, et al. Genomic profiling of dediffer- 2019;29:2338–2354.e2337.
entiated liposarcoma compared to matched well-differentiated lipo- 69. Burr ML, Sparbier CE, Chan KL, et al. An evolutionarily conserved
sarcoma reveals higher genomic complexity and a common origin. function of polycomb silences the MHC class I antigen presenta-
Cold Spring Harb Mol Case Stud. 2018;4:a002386. tion pathway and enables immune evasion in cancer. Cancer Cell.
58. Snow H, Mitchell C, Hendry S, et al. Characterising the immune 2019;36:385–401.e388.
microenvironment in liposarcoma, its impact on prognosis and the 70. Zhou L, Mudianto T, Ma X, et al. Targeting EZH2 enhances antigen
impact of radiotherapy. J Surg Oncol. 2021;123:117–26. presentation, antitumor immunity, and circumvents Anti-PD-1 resis-
59. Pollack SM, Jungbluth AA, Hoch BL, et al. NY-ESO-1 is a ubiq- tance in head and neck cancer. Clin Cancer Res. 2020;26:290–300.
uitous immunotherapeutic target antigen for patients with myxoid/ 71. Gounder MM, Zhu G, Roshal L, et al. Immunologic correlates of
round cell liposarcoma. Cancer. 2012;118:4564–70. the abscopal effect in a SMARCB1/INI1-negative poorly differenti-
60. Somaiah N, Chawla SP, Block MS, et al. A phase 1b study evalu- ated chordoma after EZH2 inhibition and radiotherapy. Clin Cancer
ating the safety, tolerability, and immunogenicity of CMB305, a Res. 2019;25:2064–71.
lentiviral-­
based prime-boost vaccine regimen, in patients with
 Part XVI
The Role of Radiation in the Treatment of Primary
and Metastatic Cancer

Baldassarre Stea

Keywords Radiation sensitivity;Stereotactic radiosurgery;Stereotactic body radiation

therapy;Image-guided radiation;MR-Linac;Immune checkpoint inhibitors

Along with surgery and chemotherapy, radiation therapy is a widely utilized treatment
modality in the field of oncology, with an estimated 50–60% of all cancer patients usually
benefitting from its application during the course of their illness [1]. The feasibility of radiation
treatments for cancer can be dated back to the discovery of X-rays by Wilhelm Röntgen in
1885, although at that time the potential of X-rays was limited to diagnostic imaging due to
their poor penetration through tissues [2]. Today, radiation therapy is used for both palliation
of cancer symptoms, such as bleeding, spinal cord compression, or painful bony metastases,
and cure of many disease sites such as breast, prostate, lung, and head and neck. As such, radia-
tion therapy has become an integral part of combined multimodality treatment approaches to
many cancers.
Part XVI is devoted to multiple aspects of radiotherapy, including a chapter on basic prin-
ciples of radiobiology, one on technical factors of radiotherapy, another on the definitive treat-
ment of cancer, and three different chapters addressing the role of radiation for metastatic
disease. The chapter on radiobiology provides the biological basis for the modality. In this
chapter, Harryman and Cress review the basic principles underpinning the generation of radia-
tion damage. The generation and volume of base damage, single-strand DNA breaks (SSBs),
and double-strand breaks (DSBs) are dose dependent upon ionizing radiation and the kinetics
of repair and cellular outcomes are distinct if the damage is not repaired. The unit of radiation
dose is Gray (Gy). One Gray of radiation results in approximately 105 ionization events per
cell, which yield approximately 1000 SSBs and 40 DSBs. Radiation-induced DSBs are the
lethal events most relevant to clinical radiotherapy. The authors review the five “Rs” of radio-
therapy, which are responsible for the differential radiosensitivity of tumor tissue and normal
tissue and include Reassortment (cell cycle position), Reoxygenation (of hypoxic cells),
Repopulation (of tumor cells after each dose of radiation), Repair of DNA strand breaks, and
intrinsic Radiosensitivity of the tissue irradiated. They finally review the tumor microenviron-
ment, long recognized as a dose-limiting factor in the effectiveness of radiotherapy, and how
the microenvironment can influence radiation sensitivity [3]. Finally, they review the role of
hypoxia in the response to radiotherapy, which not only is a dose-limiting factor in the response
but will also select for clones of aggressive disease phenotype, a phenomenon that can be
detected in human tumors by multi-parametric MRI [4].
The delivery of radiation treatments relies on the use of targeted high-energy X-rays or
charged particles (protons, carbon ions) produced by linear accelerators (Linacs), or radioac-
650 The Role of Radiation in the Treatment of Primary and Metastatic Cancer

tive substances contained in receptacles or catheters (brachytherapy) to induce death of malignant cells.
When used with curative intent, the goal of radiation is to maximize the radiation dose delivery to the tumor
while minimizing the exposure of adjacent normal tissues. It is, thus, imperative to be able to accurately
target radiation delivery; this is accomplished by utilizing a variety of imaging technologies during the
planning process, including CT, MRI, and PET scans. It is, thus, not surprising that the success of modern
radiotherapy has paralleled that of medical imaging. The technology involved in modern treatment planning
and delivery and the successful application of radiation are covered in great details in the following chapters
of this section by Russell Hamilton. However, two major technical advances are worthy of mention here:
one involving treatment planning and the other image guidance. These technical innovations that evolved
over the past two decades in areas such as disease localization (CT-MRI-PET fusion), target motion man-
agement (4D-CT), treatment planning software, and dose delivery methodologies (stereotactic and inten-
sity-modulated radiation) have indeed revolutionized the field of radiation therapy resulting in better tumor
control with fewer complications, thus providing improved patient care. The treatment planning process as
well as the multiple imaging modalities available for accurate targeting of the tumor are elegantly reviewed
in more detail in the chapter by Russell Hamilton, which also includes a thorough presentation of all major
forms of radiation delivery methods (protons, heavy ions, and Flash radiotherapy) and the latest technical
innovations that have led to improved patient outcomes.
The second major advance in technology occurred with the introduction of image-guided radiation
(IGRT), whereby an image of the target (tumor) is obtained immediately before the delivery of the dose, to
ensure correct localization of the disease. In order to avoid a geographic miss, and to account for daily
internal motion of the target (variation in tumor position), the target volume (gross tumor volume or GTV)
is usually expanded by a variable margin of normal tissue to create a planning target volume (PTV). This
“safety” margin around a tumor differs with the degree of technology used and varies between 2 and
10 mm. Modern IGRT technology allows reduction in the margins used around a tumor, thus decreasing the
chance of damaging surrounding normal tissues [5]. As the treatment margins became smaller and tighter
around the GTV, it became imperative to ascertain the location of the target prior to treatment by taking a
snapshot of the tumor just prior to the delivery of radiation. Most modern Linacs are now equipped with
onboard imaging systems capable of obtaining a cone beam CT scan prior to each treatment; other forms of
image guidance are orthogonal X-ray pairs, optical surface monitoring systems, and more recently MR
guidance. This is also reviewed in much greater detail in one of the chapters in this section.
The constant need to continually reduce the PTV margins has led to the latest innovation in radiation
oncology, the integration of MR imaging with a linear accelerator (MR-Linacs). MR imaging has superior
soft-tissue contrast capabilities when compared to CT imaging. Therefore, MR-Linacs excel in targeting
lesions within fleshy organs such as liver metastases, or cancers in the abdominal cavity (e.g., pancreatic
cancer) which often blend with surrounding intestinal loops and other abdominal organs [6]. MR-Linacs are
equipped with automated beam gating for precise and accurate dosing; they can track a tumor motion in real
time and when the tumor moves (due to respiratory motion or patient positioning), the beam stops; this
gives radiation oncologists the confidence to shrink the PTV margins down to a few millimeters while
simultaneously escalating the dose. In addition to superior soft-tissue visualization, one of the most critical
innovations afforded by MR-Linacs is their ability for integrated adaptive treatments with daily on-table
re-optimization of the treatment plan. An adaptive treatment may be triggered by a change in tumor size or
by the repositioning of an organ at risk close to the tumor, necessitating a new plan to avoid treating the
normal tissue unnecessarily. By leveraging their superior imaging features and fast dose calculation algo-
rithms, it is possible to deliver more powerful and effective doses of radiation in fewer sessions (stereotactic
body radiation therapy or SBRT), a strategy which in turn paves the way to improved tumor control and
patient outcomes [7].
Typical SBRT treatments involve the delivery of 1–5 fractions of radiation, with dose fractions of
6–30 Gy per fraction [8]. The high-precision targeting allows for a higher ablative dose, which improves the
likelihood of achieving local control while also allowing for smaller margins for treatment setup, decreas-
ing the amount of normal tissue irradiated. This technique has been increasingly used to treat patients with
oligometastatic disease, as described in detail in the chapter by Ho and Hsu. In this setting, radiation
­therapy has been used to extend survival in patients with metastatic disease. In the recently published phase
The Role of Radiation in the Treatment of Primary and Metastatic Cancer 651

2 SABR-COMET study by Palma et al., patients with a controlled primary malignancy and 1–5 extracranial
metastatic lesions were randomized to either SBRT to all sites of metastatic disease or palliative standard of
care, including palliative doses of radiation at symptomatic sites. Remarkably, not only the median PFS was
improved in the SBRT arm compared to the control arm (11.6 vs. 5.4 months, p=0.001), but the overall
survival was also improved in the SBRT arm, with median survival of 50 months, compared to 28 months
in the standard-of-care arm (p = 0.006); furthermore, the 5-year OS was increased (42.3% vs. 17.7%,
respectively). This application of radiation in the metastatic setting can, therefore, influence not only the
quality but also the length of life of patients with limited number of metastases. For a more detailed refer-
ence on treating metastatic cancer, the reader is referred to a more comprehensive review [9].
One of the most common and painful presentations of metastatic cancer is in the form of spinal metas-
tases which are often treated with low-dose palliative radiation (8–20 Gy). Dr. Ryu and colleagues elegantly
review the evolution of spinal radiotherapy over the past three decades, presenting evidence of the efficacy
of large single fractions of stereotactic radiation (SRS) or a few large fractions (SBRT) in the treatment of
spinal metastases, providing evidence of improved pain control which is often the immediate goal of treat-
ment of spinal metastases. The authors review the radiobiological principles behind the use of single large
dose of radiation and critically report on two major clinical trials performed in the past decade, the NRG/
RTOG 0631 comparing two different doses of radiation delivered by either standard external beam therapy
(EBRT) or SRS and also the CCTG//TROG trial comparing a multi-fraction external beam with 2 fraction
SBRT. Although the NRG/RTOG trial showed equivalency of either technique in achieving high degree of
pain control (61.3% with SRS versus 43.7% with conventional EBRT at 12 months posttreatment, p = 0.20),
the second trial reported superior results in pain control when the SBRT technique (24 Gy in 2 fractions)
was utilized over the conventional EBRT technique (20 Gy in 5 fractions, p < 0.001). These results suggest
that the more biologically effective dose of 24 Gy in 2 fractions may be the preferred and more efficient way
to treat limited spinal metastases, although long-term results are still pending [10].
Recently, one of the most exciting developments in combined modality therapy of cancer has been seen
in the treatment of melanoma brain metastases. Stereotactic radiosurgery is a common technique used for
the treatment of small targets in the brain. On the other hand, immune checkpoint inhibitors have been
shown to be very effective in the treatment of systemic metastatic melanoma. In a recent retrospective
review [11], patients with melanoma metastatic to brain treated with SRS and ipilimumab had a favorable
median survival of almost 36 months (compared to historical controls of about 12 months). As for the tim-
ing and sequencing of the two modalities, patients receiving immunotherapy concurrently with and within
30 days of SRS had significantly improved local control and regional intracranial progression-free survival
compared with patients receiving sequential therapy with more than 30-day separation. This benefit appears
to be limited to high-dose radiation, however, suggesting synergy between immune checkpoint inhibitors
and radiosurgery.
Finally, one common use of radiation is in the postoperative setting where it is used as a prophylactic
strategy to sterilize the operative bed and regional nodes and prevent locoregional failures; this, in turn,
increases the cure rate of patients with high-risk features. This application of RT is widely diffused, being
recommended from resected brain tumors to head and neck cancers, breast cancer, prostate cancer, sarco-
mas, and even pediatric cancers (such as Wilms’ tumor). One example of the ability of radiotherapy to
prevent metastatic spread of cancer and thus increase survival comes from a meta-analysis of 10,801 women
in 17 randomized trials of radiotherapy versus no radiotherapy after breast-conserving surgery (BCS).
Overall, RT reduced the risk of any recurrence by nearly half; specifically, it reduced the 10-year risk of any
(i.e., locoregional or distant) first recurrence from 35.0% to 19.3% (absolute reduction 15.7%, p < 0.00001)
and reduced the 15-year risk of breast cancer death from 25.2% to 21.4% (absolute reduction 3.8%,
p = 0.00005). Therefore, with postoperative RT after BCS, one death was avoided at 15 years for every four
locoregional recurrences prevented at 10 years [12].
In summary, although the landscape of oncologic treatments continues to evolve with highly promising
discoveries in targeted small-molecule inhibitors and new immunotherapy drugs, the role of radiation
remains prominent in the management of numerous malignancies in a number of ways including definitive
radiotherapy alone, concurrent chemoradiation for locally advanced cancers, adjuvant radiotherapy follow-
ing surgical resection of primary tumors, and finally in combination with immune checkpoint inhibitors for
652 The Role of Radiation in the Treatment of Primary and Metastatic Cancer

brain metastases. Following a succinct review of the basic biophysical principles involved in the administra-
tion of radiation therapy, this section has separate chapters covering the current approach in the modern
treatment of the major malignancies afflicting mankind (breast, prostate, lung, cervical/endometrial cancer)
and the palliative role of radiation for metastatic cancer.

References
1. Delaney G, Jacob S, Featherstone C, Barton M. The role of radiotherapy in cancer treatment: estimat-
ing optimal utilization from a review of evidence-based clinical guidelines. Cancer. 2005;104(6):1129–
37. https://doi.org/10.1002/cncr.21324.
2. Gianfaldoni S, Gianfaldoni R, Wollina U, Lotti J, Tchernev G, Lotti T. An overview on radiotherapy:
from its history to its current applications in dermatology. Open Access Maced J Med Sci. 2017;5(4):521–
5. https://doi.org/10.3889/oamjms.2017.122.
3. Krisnawan VE, Stanley JA, Schwarz JK, DeNardo DG. Tumor microenvironment as a regulator of
radiation therapy: new insights into stromal-mediated Radioresistance. Cancers. 2020;12(10) https://
doi.org/10.3390/cancers12102916.
4. Gaustad JV, Hauge A, Wegner CS, Simonsen TG, Lund KV, Hansem LMK, Rofstad EK. DCE-MRI of
tumor hypoxia and hypoxia-associated aggressiveness. Cancers. 2020;12(7) https://doi.org/10.3390/
cancers12071979.
5. De Los SJ, Popple R, Agazaryan N, Bayouth JE, Bissonnette JP, Bucci MK, Dieterich S, Dong L,
Forster KM, Indelicato D, Langen K, Lehmann J, Mayr N, Parsai I, Salter W, Tomblyn M, Yuh WT,
Chetty IJ. Image guided radiation therapy (IGRT) technologies for radiation therapy localization and
delivery. Int J Radiat Oncol Biol Phys. 2013;87(1):33–45. https://doi.org/10.1016/j.ijrobp.2013.02.021.
6. Witt JS, Rosenberg SA, Bassetti MF. MRI-guided adaptive radiotherapy for liver tumours: visualising
the future. Lancet Oncol. 2020;21(2):e74–82. https://doi.org/10.1016/S1470-­2045(20)30034-­6.
7. Boldrini L, Cusumano D, Cellini F, Azario L, Mattiucci GC, Valentini V. Online adaptive magnetic
resonance guided radiotherapy for pancreatic cancer: state of the art, pearls and pitfalls Radiat
Oncol;14(1):71. 2019; https://doi.org/10.1186/s13014-­019-­1275-­3.
8. Palma DA, Olson R, Harrow S, Gaede S, Louie AV, Haasbeek C, Mulroy L, Lock M, Rodrigues GB,
Yaremko BP, Schellenberg D, Ahmad B, Griffioen G, Senthi S, Swaminath A, Kopek N, Liu M, Moore
K, Currie S, Bauman GS, Warner A, Senan S. Stereotactic ablative radiotherapy versus standard of care
palliative treatment in patients with oligometastatic cancers (SABR-COMET): a randomised, phase 2,
open-label trial. Lancet. 2019;393(10185):2051–8. https://doi.org/10.1016/S0140-­6736(18)32487-­5.
9. Stea B, Skrepnik T, Hsu CC, Abendroth R. The role of radiation therapy in the treatment of metastatic
cancer. Clin Exp Metastasis 2018 Aug;35(5-6):535-546. doi: https://doi.org/10.1007/
s10585-­018-­9926-­6.
10. Sahgal A, Myrehaug SD, Siva S, Masucci L, Foote MC, Brundage M, Butler J, Chow E, Fehlings MG,
Gabos Z, Greenspoon J, Kerba M, Lee YK, Liu MC, Maralani P, Thibault I, Wong R, Hum M, Ding K,
Parulekar W. CCTG SC.24/TROG 17.06: a randomized phase II/III study comparing 24Gy in 2 stereo-
tactic body radiotherapy (SBRT) fractions versus 20Gy in 5 conventional palliative radiotherapy (CRT)
fractions for patients with painful spinal metastases. Int J Radiat Oncol Biol Phys. 2020;108(5):1397–
8. https://doi.org/10.1016/j.ijrobp.2020.09.019.
11. Skrepnik T, Sundararajan S, Cui H, Stea B. Improved time to disease progression in the brain in patients
with melanoma brain metastases treated with concurrent delivery of radiosurgery and ipilimumab.
Onco Targets Ther. 2017;19;6(3):e1283461 https://doi.org/10.1080/2162402X.2017.1283461.
12. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG), Darby S, McGale P, Correa C, Taylor
C, Arriagada R, Clarke M, Cutter D, Davies C, Ewertz M, Godwin J, Gray R, Pierce L, Whelan T, Wang
Y, Peto R. Effect of radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year
breast cancer death: meta-analysis of individual patient data for 10,801 women in 17 randomised trials.
Lancet. 2011;378(9804):1707–16. https://doi.org/10.1016/S0140-­6736(11)61629-­2.
 Basic Principles of Radiobiology
and Cancer Metastasis Prevention 62
William L. Harryman and Anne E. Cress

Abstract
Learning Objectives
Basic biological principles of tumor cells and tissue • Fundamental understanding of the biological effects
responses to therapeutic ionizing radiation are presented in of therapeutic radiation on major cell biological
the context of metastasis prevention. Focuses include a functions.
description of cell survival, reassortment (cell cycle posi- • Understand five biological factors that influence the
tion), reoxygenation, repopulation, repair, radiosensitivity, radiation response in cells and tissues.
and the influence of the hypoxic tumor microenvironment. • Understand the dominant DNA damage repair path-
We emphasize understanding the confounding factors of ways induced by ionizing radiation.
the tumor microenvironment and tumor interaction with • Familiarity with the potential consequences of
ever-changing microenvironments during tumor progres- unrepaired DNA damage.
sion, including examples of how the tumor microenviron- • Understanding intrinsic and acquired radiation
ment influences tumor sensitivity to ionizing radiation. resistance.
Microenvironments change in response to tumor invasion, • Define the hypoxic tumor environment, and under-
and irradiation influences the microenvironment, suggest- stand why it influences radiation therapy response.
ing that integrated responses to radiation occur between the • Appreciate the complexity and dynamic nature of
tumor and the microenvironment. Future research will the cellular and tumor microenvironment.
require increased understanding of both tumor subtypes • Familiarity with new research areas that will address
and the spatial and temporal interactions of the microenvi- the integrative nature of tumor response to therapy.
ronment as tumors locally invade and metastasize. Prostate
cancer acts as an example in which extraprostatic extension
of disease into the periprostatic adipose tissue has unknown
influences on radiation therapy responses but may account, 62.1 Introduction
in part, for the reported limited therapeutic responses in
obese patients. Understanding the integrated responses of The biological effects of radiation principally involve dam-
tumor with the changing microenvironment during the age to deoxyribonucleic acid (DNA), which is the critical
early stages of metastasis holds potential for identifying target, damaged via the hydroxyl radical as described in
new targets to improve tumor eradication by ionizing radia- Radiobiology for the Radiologist [1]. For the clinical use of
tion and preventing tumor metastasis or recurrent disease. ionizing radiation, a major limitation is the consideration of
normal tissue tolerance and understanding the unique bal-
ance between tumor cell kill and surrounding normal tissue
W. L. Harryman toxicity. The understanding of radiation therapy tolerance
The University of Arizona Cancer Center, Tucson, AZ, USA
e-mail: [email protected] limits and the estimation of individual tolerance is a major
research area to avoid the severity of late connective tissue
A. E. Cress (*)
Department of Cellular and Molecular Medicine, The University of damage after therapy [2]. In this section, we will focus our
Arizona Cancer Center, Tucson, AZ, USA attention on the tumor-specific biological responses to IR
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 653
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_62
654 W. L. Harryman and A. E. Cress

therapy, including cell survival, DNA damage repair, cell age has occurred and an accumulation of unrepaired DNA
cycle arrest, and tumor biology responses. damage is expected. The linear portion of the survival curve
is used to calculate the slope (called D0 or mean lethal dose)
to provide a comparison between different tumor types.
62.2 Tumor Cell Survival to Radiation Tumor cells will die from radiation treatment by a variety of
mechanisms including mitotic catastrophe, apoptosis (in
Tumor cell survival to a radiation dose is determined using a hematopoietic derived cells), necroptosis, and proliferative
colony-forming assay that measures the clonogenic capacity death (epithelial derived cells).
of individual cells for up to 14 days after a radiation expo-
sure. This measure is chosen since metabolic assays of
­proliferation will underestimate cell killing since cells can 62.3  everal Biological Factors Influence
S
die immediately (called interphase death) or die after several Tumor Cell Survival to Radiation
cell divisions (called proliferative death) following exposure Exposure
to radiation. The colony-forming assay will estimate total
cell death from radiation exposure independent of metabo- These are referred to as the “5 Rs” and include reassortment
lism or the cell death pathway used. In effect, the ability to (cell cycle position), reoxygenation, repopulation, repair,
produce progeny determines cell survival. Typical cell sur- and radiosensitivity. We will take each of these in turn in the
vival curves are shown in Fig. 62.1. following sections.
At low doses of radiation, the shape of the survival curve
(black line, Fig. 62.1) contains a “shoulder” region, inter-
preted to mean that within the tumor cell population, effec- 62.3.1 Reassortment
tive DNA repair has occurred within 14 days post-irradiation,
also referred to as sublethal damage. In contrast, at higher In response to radiation, a tumor cell population will respond
radiation doses, the shape of the survival curve now becomes with up to three dominant cell cycle checkpoints. Arrest in
linear on the logarithmic scale, indicating more lethal dam- G1 (via cyclin D and Cdk4/6) occurs to allow repair of DNA
damage before DNA replication can occur. An arrest in S
phase (via cyclin E/CdK2) occurs to repair DNA damage
Sub-lethal damage early in the DNA replication phase. Arrest in G2 occurs (via
1.0 (“shoulder”)
cyclin B/CdK1) to repair DNA damage before mitosis.
Tumor cells are most sensitive to radiation therapy during
mitosis, G1, and G2 phases of the cell cycle and most resis-
0.1
tant during S phase. Within the same tumor cells, those cells
irradiated in late S phase will result in a shoulder region,
whereas cells irradiated within mitosis will display a linear
survival curve without a shoulder region. Since the duration
Cell Survival

of mitosis and S phase are very similar between cell types,

0.01 Lethal damage tumor cells that have longer cell cycle times most likely have
(linear portion)
longer G1 intervals. Longer G1 intervals also can result from
adverse microenvironments such as low pH, hypoxia, and
inflammation. We note that as tumors invade and metastasize
0.001 to distant sites, the tumor microenvironment will change [3].
The successful radiation treatment of oligometastatic disease
[4, 5] will likely require an understanding of the resulting
cell cycle distribution or change in radiation responsiveness
0.0001 due to a new and changing microenvironment.
0 1 2 3 4 5 6 7 8 9 10
Radiation dose (Gy)

Fig. 62.1 Idealized tumor cell survival data. Cell survival is measured 62.3.2 Reoxygenation
by determining the survival of cells (on a logarithmic scale) 14 days
after exposure to increasing doses of radiation (Gy). Typical sublethal Since most DNA damage by radiation depends upon the
damage occurs with low doses of radiation (shoulder), while lethal
damage is observed with higher doses (linear portion). In some
hydroxyl radical, a dramatic effect of oxygen on radiation
radiation-­sensitive cell types, sublethal damage is not observed (red sensitivity occurs. Most transplantable tumors in animals
survival data) have been shown to contain hypoxic cells that limit curabil-
62 Basic Principles of Radiobiology and Cancer Metastasis Prevention 655

ity by single doses of X-rays. Hypoxic fractions vary from repair. DNA damage occurs primarily by hydroxyl radical
0% to 50%, with a tendency to average about 15%. After damage to the phosphate-sugar backbone of DNA resulting
single doses of radiation, the initial tumor cell kill will be in base damage, single-stranded breaks (SSB), or double-­
followed by a reoxygenation of the residual tumor that was stranded breaks (DSB). Damaged non-helix-distorting bases
within the hypoxic regions. The oxygen enhancement of are repaired by base excision repair (BER), or bulky helix-­
radiation sensitivity can vary up to a threefold increase distorting lesions are removed by nucleotide excision repair
depending on the cell type. Although it is usually assumed (NER) operating throughout the cell cycle. BER is important
that the oxygen tension of most normal tissues is similar to for removing damaged bases that could otherwise cause
that of venous blood or lymph (20–40 mmHg), in fact, oxy- mutations by mispairing or lead to breaks in DNA during
gen probe measurements indicate that the oxygen tension replication. The generation and abundance of base damage,
may vary between different tissues over a wide range from 1 SSBs, and DSBs are dose dependent upon ionizing radiation,
to 100 mmHg. There is strong evidence that human tumors and the kinetics of repair and cellular outcomes are distinct if
contain hypoxic cells that would therefore be radiation resis- the damage is not repaired (Table 62.1). One gray (Gy) of
tant. This evidence includes histologic appearance, oxygen radiation results in approximately 105 ionizations per cell,
probe measurements, the binding of nitroimidazoles, PET which yields 1000 SSBs and 40 DSBs. Radiation-induced
and SPECT studies, and pretreatment hemoglobin levels. DSBs are the lethal events relevant to clinical radiotherapy.
Many tissues are therefore borderline hypoxic and contain a The two dominant pathways for DSB repair are non-­
small proportion of cells that are radio-biologically hypoxic. homologous end joining (NHEJ) and homologous recombi-
This is particularly true of, for example, normal liver and nation (HR). NHEJ is characterized by direct re-joining of
muscle or cancer tissue such as human prostate adenocarci- the two ends, which frequently generates small mutations at
noma. Hypoxic gradients within tumor populations can the repair site [6]. This form of DNA repair is active through-
affect treatment outcomes, and a major need is to identify out the cell cycle and repairs approximately 85% of the
new markers of these gradients for treatment planning. radiation-­induced DSBs. NHEJ uses a protein complex of
the Ku70/Ku80 proteins as a scaffold structure upon which
other enzymatic proteins (DNA PKCs, Artemis endonucle-
62.3.3 Repopulation ase, ligase IV/XRCC4 complex) are recruited to accomplish
repair. NHEJ is also used in normal processes such as the
Following single doses of radiation, there is a repopulation VDJ recombination of IgG or TCR genes in B and T cell
of the residual tumor, and this recapitulates the survival maturation.
kinetics of the original tumor. Fractionated radiation sche- HR initiates with resection to form single-stranded DNA
mas have been developed to increase the efficacy of the treat- (ssDNA), which invades “donor” homologous templates for
ment, in effect, reducing the shoulder region of the survival DNA synthesis and repair [7]. HR repairs DNA in the S and
curve. In model systems, the tumor growth delay induced by G2 phases of the cell cycle before the cell enters mitosis (M
the radiation is the gold standard for the measurement of phase) [8]. HR uses a protein complex of sensor and effector
radiation sensitization of an intervention [1]. The use of proteins (RPA, Rad51, BRCA2, and others) to accomplish
tumor growth delay is needed since it can reveal a pre-­clinical largely error-free repair.
sensitization of the tumor without the confounding factors of The cell cycle appears to largely determine whether HR
tumor necrosis or cell loss. In addition, it has proven to be a (in S and G2) or NHEJ (all phases, including G0) is used to
reliable estimate of sensitization that is reflective in the clini- repair double-stranded breaks, and since there is a potential
cal situation. for overlap, the determinants of pathway choice are an unre-
solved question. Given the potential for errors with the NHEJ
pathway, one of the remaining questions is how pathway
62.3.4 Repair choices are made by targeted cells [9]. If the damage is not
repaired or repaired by an error-prone process, the result is
We consider here the major types of DNA damage induced chromosome fragmentation, loss of DNA fragments (dele-
by ionizing radiation generated by gamma or X-rays for ther- tions), chromothripsis, insertion of DNA fragments in a
apeutic intent and the corresponding major types of DNA wrong location, or chromosome aberrations, which at sub-

Table 62.1 Comparison of dominant DNA repair pathways in response to ionizing radiation
Repair pathway DNA damage Incidence/cell (1 Gy X-ray) Repair mechanism Cell death or senescence
BER/NER Base/nucleotide 1000 Protein complex; easy/fast NO
HR/NHEJ SSB 1000 Protein complex; easy/fast NO
HR/NHEJ DSB 40 Protein complex; difficult/slow YES
656 W. L. Harryman and A. E. Cress

Fig. 62.2 Potential

biological consequences of DSB
unrepaired or error-prone
repair of DSB damage.
Different types of
chromosome damage have
been observed that can persist • Chromosome
in surviving tumor cell fragmentation
populations or result in cell • DNA Deletions
death by at least four different • Chromothripsis
pathways. Surviving cell • Copy number variation
populations with chromosome
Mitotic
damage may account for the
catastrophe
emergence of aggressive
tumors and recurrent disease

Survival
Cell death

Tumor
Evolution
Apoptosis Autophagy Senescence Necroptosis
Recurrent
Disease

lethal radiation doses may be a factor in tumor evolution or tion [16]. Current research efforts to utilize the emerging
recurrent disease. Figure 62.2 illustrates the potential conse- technology of single-cell analysis from biopsy specimens
quences of DSB on tumor biology. Chromatin and nuclear and the power of positional digital profiling now used to map
architecture are other emerging considerations in optimiza- the tumor microenvironment [17] will likely assist radiation
tion of DNA repair [10] and may reveal additional targeting sensitization efforts. In this strategy, it would be important to
candidates as adjuvants to radiation-induced tumor cell kill- apply an in situ analysis of DNA repair complex assembly in
ing. This is especially important since the basal cell popula- the biopsy analysis workflow since single markers are not
tion of epithelial-derived tumors contains heterochromatin, sufficient; markers such as γH2AX or RPA (replication pro-
and how this form of chromatin is repaired is largely tein A) will bind DNA independent of the DNA damage
unknown [11]. response [13, 18].
Markers of DNA damage repair include the phosphory-
lated form of histone H2A, called γH2AX, which leads to
chromatin decondensation after DNA double-stranded 62.3.5 Radiosensitivity
breaks to facilitate repair. The histone variant H2AX consti-
tutes about 10% of the H2A histones in human chromatin Radiosensitivity of different cell types is striking when con-
and can be detected as soon as 20s after the DSB is induced sidering that the hematopoietic stem cells are particularly
by ionizing radiation with half of the maximum production radiosensitive as compared to epithelial cells. Their survival
of γH2AX occurring in 1 min [12]. γH2AX is required for curves have little or no shoulder regions and have a mean
the accumulation of many sensor/effector proteins at the lethal dose (D0) of slightly less than 1 Gy as compared to
DSB site [13]. epithelial cells of 2Gy or higher. Hematopoietic tissues are
The clinical assessment of DNA repair capacity and located primarily in the bone marrow, with 60% located in
determining the most robust assay to complement the ongo- the pelvis and vertebrae and the remainder in the ribs, skull,
ing clinical trials targeting specific DNA repair proteins for sternum, scapula, and proximal sections of the femur and
radiation sensitization is a major unmet need [14]. Patient-­ humerus. Lymphocytes are very radiosensitive, largely
level DNA damage and repair pathway profiles have been because of apoptosis, and B cells are more radiosensitive
reported to provide a strong prognostic performance with than T cells. Total lymphoid irradiation to a dose of 30–40 Gy
long-term outcomes and may be useful for risk stratification is used for the treatment of lymphomas and leads to a long-­
of prostate cancer patients, a population where radiation lasting T cell lymphopenia.
therapy is a treatment modality [15]. In another example, Different cell types within the same organ can also dis-
loss of NHEJ will sensitize human prostate tumors to radia- play differences in radiosensitivity. For example, in the nor-
62 Basic Principles of Radiobiology and Cancer Metastasis Prevention 657

mal colon, the differentiated villus cells are radiation resistant HR-dependent DSB repair, may account for the ability of β1
as compared to the stem cells in the crypt. This explains the integrin (also known as CD29) antagonists to sensitize tumor
prodromal effects of radiation in which the loss of fluid and cells to the lethal effects of radiation treatment [32].
massive infections post-irradiation are due to loss of the vil- Alternatively, the repair of heterochromatin in response to
lus layer. Epithelial tumor cell populations of the lung are radiation treatment directly involves the cytoskeletal system,
another example in which squamous cell carcinoma and which has known linkages from the cell surface to the inner
adenocarcinoma are more radiation resistant as compared to nuclear membrane, where heterochromatin resides [11].
small-cell carcinoma, in part since one of the required NHEJ
DNA repair components, Ku80, is increased [19]. The use of
integrated omics profiling of treatment-naïve tumor is likely 62.5 The Hypoxic Tumor
to reveal more patterns of molecular subtyping that can be Microenvironment
exploited to individualize treatments, as shown recently in
prostate cancer where the basal cell subtype appears to be Tumors contain hypoxia gradients which can be observed
more responsive to radiation therapy [20]. [33, 34]. Hypoxia not only decreases the tumor response to
radiation therapy but also selects for aggressive disease as
detected by mpMRI studies in patients [35]. The investiga-
62.4 Acquired Radiation Resistance tion of 2000 patients with prostate cancer revealed that there
is an apparent selection for PTEN-minus aggressive disease
An enhancement of tumor survival to radiation damage by hypoxia [36]. Hypoxic suppression of HR is proposed as
occurs in vivo and in vitro when tumors are maintained under a “contextual-lethal” vulnerability in prostate tumors [37].
non-growth conditions after treatment. The repair of poten- There is evidence that patients with prostate cancer will have
tially lethal damage was first uncovered as a variable that areas of both chronic and acute hypoxia, both detectable by
was unaccounted for in standard assays of cytotoxic agents pimonidazole, a compound that is reductively activated in
[21]. Today this underscores the potential of dormant tumor hypoxic cells. An antibody reagent MAb1 binds to the result-
cells or tumor stem cells within the heterogeneous tumor cell ing adducts allowing their detection by immunochemical
population to be a source of recurrent disease following ther- means. The advent of the oral formulation may allow novel
apy. The acquired or adaptive response of the tumor popula- clinical trials that intensify cancer therapy to prostate cancer
tion and the ability to modify it are important aspects of patients whose tumors harbor aggressive genetic changes
fractionated radiotherapy [22]. and hypoxic subregions. Understanding and mapping the
Previously recognized as a dose-limiting factor for radia- hypoxic regions may help to stratify patients into risk cate-
tion therapy and a component of normal tissue tolerance, it is gories, aid with appropriate radiation treatment planning,
now recognized that the tumor microenvironment and its and result in the rational design of combination therapies to
response will influence the radiation sensitivity of recurrent prevent metastasis.
disease [23]. Detailed knowledge of these interactions will
likely be another biological avenue to exploit for increasing
the efficacy of radiation therapy beyond radiation dose frac- 62.6  he Complex Cellular and Tumor
T
tionation toward improving outcomes for the individual Microenvironment
patient [24].
Cell adhesion-mediated radioresistance (CAM-RR) and The tumor microenvironment contains many different cell
cell adhesion-mediated drug resistance (CAM-DR) are both types including immune effector cells, normal cell types
driven by integrin- and cytokeratin-dependent mechanisms associated with vessels or nerves, fibroblasts, and smooth
[25–27]. The blocking of integrin function can enhance muscle cells. For example, the multi-spatial mapping of
radiotherapy efficacy in pre-clinical models of breast cancer prostate cancer reveals a wide variety of myeloid and macro-
[28] and rare incurable cancers such as chordoma [29]. The phages as detected by multiplex tissue staining in close prox-
radiation sensitization of tumor clusters dependent on integ- imity to the tumor [38]. Thus, it would be expected that
rin function and adhesion complexes is an actionable step irradiation of these cells along with the tumor may impact
that avoids the problem of single cell heterogeneity within a tumor response to the radiation treatment. As stated earlier,
tumor cluster. The use of synthetic ligand mimetics or integ- while it is well known that irradiation can increase interstitial
rin blocking function antibodies can sensitize tumor clusters collagen deposition, it is now appreciated that in addition to
to ionizing radiation or increase reactive oxygen species the extracellular structural elements of the tissue, growth fac-
resulting in apoptosis and necrotic cell death [29–31]. Recent tors such as TGF-beta are radiation induced [39] which reor-
results suggest that the linkages between the tumor cell sur- ganize the extracellular matrix and induce tumors to alter
face and the coordination of nuclear events, like their phenotypes. Since many different cell types are present
658 W. L. Harryman and A. E. Cress

in the tumor microenvironment, it remains to be determined

how these various cell types in different combinations may Future Questions
affect tumor radiation responses as measured by overall • How can hypoxic tissues be sensitized to
patient survival. radiotherapy?
In the case of prostate cancer, when the tumor leaves the • How can DSB repair be blocked to enhance radia-
gland and extends past the capsule (known as extraprostatic tion sensitivity of tumor cells?
extension, EPE), patient outcomes drop from an overall sur- • What categories of cell types and functions in the
vival of 100% at 5 years to approximately 30%. EPE is most varied tumor microenvironments would aid in treat-
readily detected by entry of the tumor into the periprostatic ment selection?
adipose tissue (PPAT), since adipose tissue is not present • How does determining the radiosensitivity of the
within the prostate gland itself. Thus, EPE of the prostate can- variety of cell types found in the tumor microenvi-
cer changes the tumor microenvironment from that of a ten- ronment aid in designing better radiotherapy treat-
sile smooth muscle to adipose tissue [3]. Both the biophysical ment regimens?
changes and the biochemical change of the environment will • How might a multi-omics approach to radiotherapy
likely be important since adipocytes secrete an inflammatory increase its effectiveness?
cytokine called extracellular nicotinamide phosphoribosyl-
transferase (eNAMPT), and this is associated with prostate
cancer progression [40]. The influence of this microenviron-
ment change on prostate tumor radiation responses is References
unknown, but we note with interest that higher levels of body
fat and higher visceral fat are related to poorer outcomes in 1. Hall E, Giaccia A. Physics and chemistry of radiation absorption.
In: Radiobiology for the radiologist. 7th ed. Philadelphia, PA:
radiation therapy [41–43]. Future research to understand the Lippincott Williams & Wilkins; 2012. p. 3–11.
influences of the varied tumor microenvironments in combi- 2. PetersLJ.Radiationtherapytolerancelimits.Foroneorforall?–Janeway
nation with molecular subtyping of the tumor will be needed Lecture. Cancer. 1996;77(11):2379–85. https://doi.org/10.1002/
to fully realize the long-term goal of individualizing radiation (sici)1097-­0142(19960601)77:11<2379::Aid-­cncr29>3.0.Co;2-­t.
3. Harryman WL, Warfel NA, Nagle RB, Cress AE. The tumor
therapy for a maximum therapeutic response. microenvironments of lethal prostate cancer. Adv Exp Med Biol.
2019;1210:149–70. https://doi.org/10.1007/978-­3-­030-­32656-­2_8.
4. Hellman S, Weichselbaum RR. Oligometastases. J Clin Oncol.
1995;13(1):8–10. https://doi.org/10.1200/jco.1995.13.1.8.
Educational Scope of the Chapter 5. Rao A, Vapiwala N, Schaeffer EM, Ryan CJ. Oligometastatic pros-
• How can sublethal damage and lethal damage by tate cancer: a shrinking subset or an opportunity for cure? American
radiation be determined with data from a survival Society of Clinical Oncology educational book American Society
of Clinical Oncology Annual Meeting; 2019;39:309–320. https://
curve? doi.org/10.1200/edbk_239041.
• Name and define five biological factors that influ- 6. Chang HHY, Pannunzio NR, Adachi N, Lieber MR. Non-­
ence tumor cell survival to radiation exposure. homologous DNA end joining and alternative pathways to double-­
• What are the three most dominant DNA repair strand break repair. Nat Rev Mol Cell Biol. 2017;18(8):495–506.
https://doi.org/10.1038/nrm.2017.48.
responses to ionizing radiation damage? List at least 7. Kowalczykowski SC. An overview of the molecular mechanisms
three factors that distinguish them from each other. of recombinational DNA repair. Cold Spring Harb Perspect Biol.
• If tumors do not repair double-stranded DNA 2015;7(11) https://doi.org/10.1101/cshperspect.a016410.
breaks and survive it, name at least two potential 8. Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter
P. Molecular biology of the cell. 5th ed. New York: Garland
consequences that would impact radiation therapy. Science; 2008.
• Give two examples of intrinsic radiation sensitivity 9. Carballar R, Martínez-Láinez JM, Samper B, Bru S, Bállega E,
and two examples of acquired radiation resistance. Mirallas O, Ricco N, Clotet J, Jiménez J. CDK-mediated Yku80
• Define tumor hypoxia, and provide an example of phosphorylation regulates the balance between non-­homologous
end joining (NHEJ) and homologous directed recombination
how knowing this information will alter radiation (HDR). J Mol Biol. 2020;432(24):166715. https://doi.org/10.1016/j.
treatment planning and potentially prevent meta- jmb.2020.11.014.
static disease. 10. Mackenroth B, Alani E. Collaborations between chromatin and
• Explain why molecular subtyping of the tumor and nuclear architecture to optimize DNA repair fidelity. DNA Repair.
2020;97:103018. https://doi.org/10.1016/j.dnarep.2020.103018.
spatial phenotyping of the microenvironment will 11. Rawal CC, Caridi CP, Chiolo I. Actin’ between phase separated
assist in personalizing radiation therapy in the future. domains for heterochromatin repair. DNA Repair. 2019;81:102646.
https://doi.org/10.1016/j.dnarep.2019.102646.
62 Basic Principles of Radiobiology and Cancer Metastasis Prevention 659

12. Rogakou EP, Pilch DR, Orr AH, Ivanova VS, Bonner WM. DNA 27. Deville SS, Cordes N. The extracellular, cellular, and nuclear
double-stranded breaks induce histone H2AX phosphorylation stiffness, a trinity in the cancer resistome - a review. Front Oncol.
on serine 139. J Biol Chem. 1998;273(10):5858–68. https://doi. 2019;9:1376. https://doi.org/10.3389/fonc.2019.01376.
org/10.1074/jbc.273.10.5858. 28. Park CC, Zhang HJ, Yao ES, Park CJ, Bissell MJ. Beta1 integrin
13. Yuan J, Adamski R, Chen J. Focus on histone variant H2AX: to inhibition dramatically enhances radiotherapy efficacy in human
be or not to be. FEBS Lett. 2010;584(17):3717–24. https://doi. breast cancer xenografts. Cancer Res. 2008;68(11):4398–405.
org/10.1016/j.febslet.2010.05.021. https://doi.org/10.1158/0008-­5472.Can-­07-­6390.
14. Jalal S, Earley JN, Turchi JJ. DNA repair: from genome main- 29. Harryman WL, Gard JMC, Pond KW, Simpson SJ, Heppner
tenance to biomarker and therapeutic target. Clin Cancer Res. LH, Hernandez-Cortes D, Little AS, Eschbacher JM, Cress
2011;17(22):6973–84. https://doi.org/10.1158/1078-­0432. AE. Targeting the cohesive cluster phenotype in chordoma via
Ccr-­11-­0761. β1 integrin increases ionizing radiation efficacy. Neoplasia.
15. Evans JR, Zhao SG, Chang SL, Tomlins SA, Erho N, Sboner A, 2017;19(11):919–27. https://doi.org/10.1016/j.neo.2017.08.005.
Schiewer MJ, Spratt DE, Kothari V, Klein EA, Den RB, Dicker 30. Nair RR, Emmons MF, Cress AE, Argilagos RF, Lam K, Kerr WT,
AP, Karnes RJ, Yu X, Nguyen PL, Rubin MA, de Bono J, Knudsen Wang HG, Dalton WS, Hazlehurst LA. HYD1-induced increase in
KE, Davicioni E, Feng FY. Patient-level DNA damage and repair reactive oxygen species leads to autophagy and necrotic cell death
pathway profiles and prognosis after prostatectomy for high-risk in multiple myeloma cells. Mol Cancer Ther. 2009;8(8):2441–51.
prostate cancer. JAMA Oncol. 2016;2(4):471–80. https://doi. https://doi.org/10.1158/1535-­7163.Mct-­09-­0113.
org/10.1001/jamaoncol.2015.4955. 31. Pawar SC, Dougherty S, Pennington ME, Demetriou MC, Stea BD,
16. Tarish FL, Schultz N, Tanoglidi A, Hamberg H, Letocha H, Dorr RT, Cress AE. alpha6 integrin cleavage: sensitizing human
Karaszi K, Hamdy FC, Granfors T, Helleday T. Castration radio- prostate cancer to ionizing radiation. Int J Radiat Biol. 2007;83(11–
sensitizes prostate cancer tissue by impairing DNA double-strand 12):761–7. https://doi.org/10.1080/09553000701633135.
break repair. Sci Transl Med. 2015;7(312):312re311. https://doi. 32. Ahmed KM, Pandita RK, Singh DK, Hunt CR, Pandita TK.
org/10.1126/scitranslmed.aac5671. β1-integrin impacts Rad51 stability and DNA double-strand break
17. de Vries NL, Mahfouz A, Koning F, de Miranda N. Unraveling the repair by homologous recombination. Mol Cell Biol. 2018;38(9)
complexity of the cancer microenvironment with multidimensional https://doi.org/10.1128/mcb.00672-­17.
genomic and cytometric technologies. Front Oncol. 2020;10:1254. 33. Horsman MR, Mortensen LS, Petersen JB, Busk M, Overgaard
https://doi.org/10.3389/fonc.2020.01254. J. Imaging hypoxia to improve radiotherapy outcome. Nat
18. Cerrato A, Morra F, Celetti A. Use of poly ADP-ribose poly- Rev Clin Oncol. 2012;9(12):674–87. https://doi.org/10.1038/
merase [PARP] inhibitors in cancer cells bearing DDR defects: the nrclinonc.2012.171.
rationale for their inclusion in the clinic. J Exp Clin Cancer Res. 34. Horsman MR, Overgaard J. The impact of hypoxia and its modifi-
2016;35(1):179. https://doi.org/10.1186/s13046-­016-­0456-­2. cation of the outcome of radiotherapy. J Radiat Res. 2016;57 Suppl
19. Ye J, Ren Z, Gu Q, Wang L, Wang J. Ku80 is differentially 1(Suppl 1):i90–8. https://doi.org/10.1093/jrr/rrw007.
expressed in human lung carcinomas and upregulated in response 35. Gaustad JV, Hauge A, Wegner CS, Simonsen TG, Lund KV,
to irradiation in mice. DNA Cell Biol. 2011;30(12):987–94. https:// Hansem LMK, Rofstad EK. DCE-MRI of tumor hypoxia and
doi.org/10.1089/dna.2010.1196. hypoxia-associated aggressiveness. Cancers. 2020;12(7) https://
20. Ben-Salem S, Hu Q, Liu Y, Alshalalfa M, Zhao X, Wang I, doi.org/10.3390/cancers12071979.
Venkadakrishnan VB, Senapati D, Kumari S, Liu D, Sboner A, 36. Bhandari V, Hoey C, Liu LY, Lalonde E, Ray J, Livingstone J,
Barbieri CE, Feng F, Billaud JN, Davicioni E, Liu S, Heemers Lesurf R, Shiah YJ, Vujcic T, Huang X, Espiritu SMG, Heisler
HV. Diversity in androgen receptor action among treatment-naïve LE, Yousif F, Huang V, Yamaguchi TN, Yao CQ, Sabelnykova
prostate cancers is reflected in treatment response predictions and VY, Fraser M, Chua MLK, van der Kwast T, Liu SK, Boutros PC,
molecular subtypes. Eur Urol Open Sci. 2020;22:34–44. https://doi. Bristow RG. Molecular landmarks of tumor hypoxia across can-
org/10.1016/j.euros.2020.10.002. cer types. Nat Genet. 2019;51(2):308–18. https://doi.org/10.1038/
21. Richie JP, Weichselbaum RR, Little JB. Importance of repair of s41588-­018-­0318-­2.
potentially lethal damage in assays of cytotoxic agents. Surgery. 37. Ashton J, Bristow R. Bad neighbours: hypoxia and genomic insta-
1982;92(2):380–4. bility in prostate cancer. Br J Radiol. 2020;93(1115):20200087.
22. Raaphorst GP, Malone S, Szanto J, Gray R. Severe normal tissue https://doi.org/10.1259/bjr.20200087.
complication correlates with increased in vitro fibroblast radiosen- 38. Saylor J, Ma Z, Goodridge HS, Huang F, Cress AE, Pandol SJ,
sitivity in radical prostate radiotherapy: a case report. Int J Cancer. Shiao SL, Vidal AC, Wu L, Nickols NG, Gertych A, Knudsen
2000;90(6):336–42. BS. Spatial mapping of myeloid cells and macrophages by mul-
23. Krisnawan VE, Stanley JA, Schwarz JK, DeNardo DG. Tumor tiplexed tissue staining. Front Immunol. 2018;9:2925. https://doi.
microenvironment as a regulator of radiation therapy: new insights org/10.3389/fimmu.2018.02925.
into stromal-mediated radioresistance. Cancers. 2020;12(10) 39. Barcellos-Hoff MH. Radiation-induced transforming growth factor
https://doi.org/10.3390/cancers12102916. beta and subsequent extracellular matrix reorganization in murine
24. Schaue D, McBride WH. Opportunities and challenges of radio- mammary gland. Cancer Res. 1993;53(17):3880–6.
therapy for treating cancer. Nat Rev Clin Oncol. 2015;12(9):527– 40. Sun BL, Sun X, Casanova N, Garcia AN, Oita R, Algotar AM,
40. https://doi.org/10.1038/nrclinonc.2015.120. Camp SM, Hernon VR, Gregory T, Cress AE, Garcia JGN. Role
25. Bauman PA, Dalton WS, Anderson JM, Cress AE. Expression of of secreted extracellular nicotinamide phosphoribosyltransferase
cytokeratin confers multiple drug resistance. Proc Natl Acad Sci U (eNAMPT) in prostate cancer progression: Novel biomarker and
S A. 1994;91(12):5311–4. https://doi.org/10.1073/pnas.91.12.5311. therapeutic target. EBioMedicine. 2020;61:103059. https://doi.
26. Damiano JS, Cress AE, Hazlehurst LA, Shtil AA, Dalton WS. Cell org/10.1016/j.ebiom.2020.103059.
adhesion mediated drug resistance (CAM-DR): role of integrins 41. Di Bella CM, Howard LE, Oyekunle T, De Hoedt AM, Salama
and resistance to apoptosis in human myeloma cell lines. Blood. JK, Song H, Freedland SJ, Allott EH. Abdominal and pelvic adi-
1999;93(5):1658–67. pose tissue distribution and risk of prostate cancer recurrence after
660 W. L. Harryman and A. E. Cress

radiation therapy. Prostate. 2020;80(14):1244–52. https://doi. 43. Wang LS, Murphy CT, Ruth K, Zaorsky NG, Smaldone MC,
org/10.1002/pros.24054. Sobczak ML, Kutikov A, Viterbo R, Horwitz EM. Impact of
42. Strom SS, Kamat AM, Gruschkus SK, Gu Y, Wen S, Cheung MR, obesity on outcomes after definitive dose-escalated intensity-­
Pisters LL, Lee AK, Rosser CJ, Kuban DA. Influence of obesity on modulated radiotherapy for localized prostate cancer. Cancer.
biochemical and clinical failure after external-beam radiotherapy 2015;121(17):3010–7. https://doi.org/10.1002/cncr.29472.
for localized prostate cancer. Cancer. 2006;107(3):631–9. https://
doi.org/10.1002/cncr.22025.
 Technical Innovations in the Delivery
of Radiation Therapy 63
Russell J. Hamilton

Abstract
Learning Objectives
Radiation oncology is a technology-driven field quickly • Know the precision and accuracy of radiation dose
adopting technical innovations to improve treatment out- delivery achievable by the current generation of
comes. A few months after Rӧntgen’s discovery of X-rays, medical linear accelerators.
therapeutic X-ray use was proposed and applied to treat • Identify imaging systems employed during radio-
cancer. Since knowledge of the biological effects of radia- therapy delivery and their purposes.
tion increases with better capabilities to control radiation • Know techniques used for automated control of
delivery, therapeutic techniques continually evolve as radiation delivery.
radiotherapy incorporates new technologies. This chapter • Understand the principles of biology-guided radia-
describes innovations implemented in the current genera- tion therapy.
tion of radiotherapy machines. Significant improvements • Know there is evidence for the role of radiation
for imaging patients in the treatment room using com- therapy in metastatic disease through the abscopal
puted tomography, magnetic resonance imaging, and effect enhanced by immunotherapy.
optical or infrared surface detection combined with more • Know there is evidence for a radiobiological advan-
precise radiation delivery are altering practice patterns. tage of FLASH radiotherapy over present conven-
The current delivery systems enable treatment regimens tional radiotherapy.
having higher doses per fraction and fewer fractions with
similar or better outcomes than previous generation
machines. Patients’ internal and external anatomies are
monitored during radiation delivery. Any changes in 63.1 Introduction
patient’s anatomy or tumor position in excess of pre-
established parameters will trigger an automatic stoppage Radiation therapy is the delivery of radiation dose to a tar-
of the radiation delivery system by computer control geted volume in order to achieve a therapeutic outcome.
through analysis of imaging data. Emerging technologies Both malignant and benign diseases are treated with radia-
showing promise in preclinical and early clinical studies, tion. The dose, number of treatments, and frequency of treat-
including automated real-time treatment delivery to bio- ment depend on the clinical situation. Ionizing radiation may
logically functional regions identified by positron-emis- penetrate several centimeters into a patient to deliver a lethal
sion tomography, combining radiotherapy with dose to a tumor. However, normal tissues such as the skin,
immunotherapy to treat local and metastatic disease, and muscle, organs, and structures lying in the entrance and exit
ultrafast delivery of radiation at dose rates orders of mag- of the radiation path receive radiation dose too. This could
nitude greater than currently practiced, are described. lead to fibrosis, organ failure, or secondary cancers several
years after treatment. Thus, the technical goal of radiation
therapy is to deliver a therapeutic dose to the targeted volume
R. J. Hamilton (*)
while minimizing the deleterious effects. This is expressed
University of Arizona, Tucson, AZ, USA
e-mail: [email protected] by the therapeutic ratio which is the ratio of the probability

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 661
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_63
662 R. J. Hamilton

of treatment success (e.g., tumor control probability) to the ensure that the target is irradiated. Precise treatments require
likelihood of normal tissue damage. The higher the therapeu- accurate identification and geometric localization of targets
tic ratio, the greater the probability of a successful treatment prior to and during treatment delivery. Gross patient motions
without morbidity. or internal physiological processes may cause targets and
Shortly after the discovery of X-rays, pioneers of radia- normal structures to move during radiation therapy delivery.
tion therapy used the novel X-ray tube technology for treat- Anything that causes the target to move must be accounted
ing cancers [1]. The X-ray energies were sufficient to pass for in the beam aperture. A number of technologies that have
through the body to produce an image, but much less than been developed to enable accounting for these effects are
needed to treat deep-seated tumors. Treatments had limited discussed below.
efficacy because a toxic skin dose was reached before suffi- In this chapter, we describe the first section is devoted to
cient dose could be delivered to the targeted tissue within the current delivery technologies. Firstly, the state-of-the-art
body and thus a low therapeutic ratio. Microwave technology radiation therapy delivery system available, the medical lin-
tremendously advanced medical linear accelerator (linac) ear accelerator (linac), also the most widely used machine
design enabling production of X-ray energies >4 MV in a around the globe. This includes the entire process starting
short distance [2]. Entrance doses for these X-ray energies with pretreatment determination of the targets and radiation
are less than the dose at depth resulting in “skin sparing,” therapy treatment planning. The capabilities of linacs and
greater penetration, and hence a higher therapeutic ratio. their ancillary devices are discussed in detail. Examples of
Now, linacs typically provide X-ray energies between 6 and radiation treatments which highlight the technical sophisti-
18MV with entrance doses much less than 50% of the maxi- cation of radiation therapy delivery are presented. Next, a
mum dose (Dmax) which occurs at depths of 1.5–3 cm [3]. For machine which combines real-time magnetic resonance
tumors at depths greater than Dmax, the dose at Dmax is larger imaging with a linac, the MR-Linac, is described. A brief
than at the tumor, if radiation is delivered with a single X-ray description of proton therapy machines concludes the sec-
beam. Therefore, radiation beams from multiple directions tion. The second section describes emerging technologies.
are used so that their superimposed doses result in a high First, a machine that combines positron-­emission tomogra-
tumor dose and lower doses to normal tissues. Even lower phy with a linac is described. Next, FLASH radiotherapy,
entrance dose is achievable with heavy charged particle consisting of radiation dose delivery at rates that are orders
beams, such as produced by a medical proton accelerator. of magnitude higher than conventional therapies, is intro-
Heavy charged particles abruptly stop and deposit a signifi- duced. Then a brief mention of carbon ion therapy is given.
cant amount of their energy at the end of their range (Bragg The promising treatment combination of immunotherapy
peak) and therefore have Dmax at the end of their range and with radiotherapy ends the section. A summary of the impact
deposit most of their dose at the depth of the tumor. Thus, of the new technologies on radiotherapy concludes the
heavy charged particles have minimal exit dose, whereas chapter.
X-ray dose decreases by only 2–5% per cm with significant
exit dose [4].
Accurate radiation dosing of targets within a patient 63.2 Radiation Therapy Delivery
requires precise aiming of an ionizing radiation beam at the Technologies
target. The radiation beam passes through the patient and
necessarily irradiates normal tissues and organs along the 63.2.1 Treatment Planning
entrance and exit of its path. The therapeutic ratio is increased
by conforming the high-dose region as closely as possible to The treatment process begins by identifying a target, for
the target and maximizing the dose gradient which is the rate example, a tumor. All medical imaging technologies are uti-
at which the dose decreases away from the target. A straight- lized in the treatment planning process, including computed
forward geometric approach to improving dose conforming tomography (CT), magnetic resonance imaging (MR),
is accomplished by shaping the aperture of the radiation positron-­emission tomography (PET), single photon emis-
beam to coincide with that of the target. Using beams from sion computed tomography (SPECT), and ultrasound. The
several different directions around the patient permits treat- particular imaging studies used for a given case depend on
ment beams that avoid passing through critical normal tis- the disease site and stage. In addition to diagnostic exams, a
sues and organs as well as increases the dose gradient. radiation therapy patient undergoes a treatment planning
Reducing a beam aperture size to as small as possible imaging scan in a dedicated unit having ancillary equipment
without missing the target improves the therapeutic ratio. used in radiotherapy. This is typically performed using a CT
However, radiation delivery systems are comprised of simulator, but MR simulators are gaining in popularity. Each
mechanical systems with inherent positional uncertainties patient is fitted with an immobilization device for stabilizing
that must be accounted for in the beam aperture in order to and securing them during the treatment planning CT and
63 Technical Innovations in the Delivery of Radiation Therapy 663

each radiation therapy treatment. The CT simulation may shown to be superior to human planning [9, 10]. A goal of
include a single scan or a series of scans, for example, sev- future treatment planning systems is to employ artificial
eral CT series showing the respiratory cycle. The CT simula- intelligence software for treatment plan construction.
tion scans and diagnostic scans of interest are transferred to
a treatment planning computer.
Treatment planning computers utilize sophisticated soft- 63.2.2 Medical Linear Accelerator
ware for designing radiation therapy treatments. The result is
a series of radiation beams used to treat a patient. Firstly, the A state-of-the-art medical linear accelerator (linac), the
separate imaging data is registered (aka fused) with the treat- TrueBeam (Varian Medical Systems, Palo Alto), is pictured
ment planning CT so that the internal anatomy from each in Fig. 63.1. A multileaf collimator (MLC) which has a mini-
imaging set may be directly compared. A significant techno- mum leaf width of 2.5 mm and submillimeter positional
logical development in the past decade was deformable reg- accuracy shapes the radiation beam. During radiation deliv-
istration, permitting registration of data sets accounting for ery, the MLC leaves move in and out of the radiation beam
deviations in a patient’s position and/or internal displace- modulating the intensity providing intensity-modulated radi-
ments, for example, caused by edema or organ filling [5]. ation therapy (IMRT). The gantry can rotate around the
Next, a model of the patient is constructed by outlining patient enabling radiation beams to strike targets from all
the internal organs, for example, the lungs, heart, kidneys, angles (0–360°) with a mechanical precision better than
spinal cord, etc. The tumor and any clinical or subclinical 0.5 mm [11]. Volumetric modulated arc therapy (VMAT) is
disease, for example, nodes, must also be outlined (aka seg- treatment delivery using IMRT simultaneously with gantry
mented). This process may be carried out in the treatment rotation. Another degree of freedom is provided by the
planning software manually at a computer workstation using patient couch which can rotate ±90° permitting treatment
drawing tools. However, commercially available planning beams to enter in planes that are not coaxial. The couch has
systems offer auto-segmentation of anatomical structures three translational degrees of freedom as well as three rota-
based on customizable atlases [6]. An atlas is comprised of tional degrees of freedom (pitch, roll, and yaw), enabling
specific anatomical structures defined as ground truth on a precise patient alignment for treatment delivery.
series of similar cases by an expert. Treatment planning soft- Treatment planning systems provide dose calculations
ware is able to automatically segment structures of interest accurate to better than a few percent, depending on the algo-
based on an atlas selected by the radiation oncologist. rithm, and are accurate to 1% using a Monte Carlo algorithm
Artificial intelligence in medical imaging might soon reach [12]. In order to realize the therapeutic ratio of the prescribed
the point that a deep learning algorithm could perform the treatment plan, patients must be positioned in the treatment
segmentation for treatment planning [7]. room such that their radiation beams are directed as specified
Next, a radiation oncologist specifies therapeutic doses by the treatment plan. There are several options available to
for the targeted volumes and places dose constraints on the achieve this goal, including imaging systems mounted on the
normal tissues with the intent that meeting these dose values linac, in-room X-ray systems, and 3D cameras with com-
will yield a high probability of successful treatment with a puter vision.
low probability of morbidity. The goal is to find a treatment Two imaging systems are mounted on the gantry of the
plan, which is a set of radiation beams deliverable by the linac shown in Fig. 63.1. The imaging panel in the radiation
treatment machine, meeting the prescribed doses for targets beam line opposite (180°) the collimator is a megavoltage
and normal tissues. Treatment planning systems are not yet imaging panel known as an electronic portal imaging device
fully autonomous. Human planners must select the number (EPID). It enables capturing images of the patient taken with
of radiation beams, directions, and prioritize the dose con- the radiation beam (energies >4 MV). The radiation beam
straints. These choices are usually based on class solutions apertures may be viewed directly on the patient radiographs
for common disease sites but for unique or unusual situations and compared with those expected from the treatment plan-
requires creativity based on experience. Once these parame- ning system (so-called digitally reconstructed radiographs).
ters are set, the planning software performs an optimization With the EPID, it is possible to reconstruct the dose that the
process to find a solution. It is not known a priori if a plan patient actually received during a treatment session by
can be found meeting all constraints. Conversely, if a plan acquiring a sequence of portal images during radiation ther-
that meets all constraints is found, it is not known if a better apy delivery [13].
plan exists. Knowledge-based treatment planning was devel- The other gantry mounted imaging system is a kilovoltage
oped to address these uncertainties. Knowledge-based treat- X-ray system with an imaging panel mounted opposed to an
ment planning predicts the expected results, providing mean X-ray tube (Fig. 63.1). This system is comparable to a diag-
and standard deviation, for the target and normal tissue doses nostic X-ray system with X-ray energies ≤150 kV. In addi-
based on an atlas of similar cases [8]. This approach was tion to planar X-ray radiographs, this system may be used for
664 R. J. Hamilton

A E.2

E.1
D

Fig. 63.1 Medial linear accelerator (TrueBeam, Varian Medical captures megavoltage X-rays that pass through the patient to form digi-
Systems, Palo Alto, CA, USA). The C-arm gantry (a) support structure tal radiographs. A kilovoltage imaging system comprised of an X-ray
rotates around the patient couch (d). The therapeutic radiation beam tube (e1) and an X-ray detector (e2) is capable of acquiring radiographs,
exits the gantry by passing through the collimator (b) where the beam fluoroscopic images, and cone beam computed tomography (CBCT).
aperture is shaped by a multileaf collimator which can also modulate One of the three in-room 3D cameras which uses light and computer
the beam intensity. The electronic portal imaging device (EPID) (c) vision to locate the surface of patients (f)

fluoroscopy, and most importantly it can acquire a cone ful for monitoring treatment sites where the targeted volume
beam computed tomography (CBCT) scan. It is used to pre- can be visualized externally or is closely represented by the
cisely position the patient for treatment delivery. The CBCT skin surface such as in breast irradiation or the treatment of
scan may be obtained by either a full or half gantry rotation, superficial lesions. For monitoring and adjusting a patient
providing three-dimensional anatomical data comparable to based on viewing the external anatomy, this system may be
a diagnostic CT scan. The CBCT scan is volumetrically reg- used for any anatomic site. This imaging technology is called
istered with the treatment planning CT to determine if there surface guided radiation therapy (SGRT) [15–17].
are any translational or rotational differences between a Although the mechanical and geometric accuracies of a
patient’s actual position on the treatment couch and the linac are submillimeter under static conditions, high-­
patient’s position modelled in the treatment planning system. precision radiation therapy delivery requires accounting for
Differences are rectified by moving the patient by the deter- the internal motion of targets and organs. Most easily identi-
mined translations and rotations with the six degrees of free- fied is lung tumor motion with respiration. Protocols are
dom motion of the couch. The imaging systems permit available to account for respiratory motion by utilizing four-­
patient positional accuracy of less than 0.4 mm [14]. dimensional computed tomography (4DCT) for the treat-
An imaging system based on visible light, AlignRT ment planning simulation scan [18, 19]. During 4DCT
(Vision RT, London, UK), is also part of the treatment room acquisition, a method to monitor respiration is employed.
setup (Fig. 63.1). It utilizes 3D cameras and computer vision The acquired CT imaging detector data are tagged with the
algorithms to precisely locate the surface of a patient with current respiratory phase of the patient. This permits recon-
submillimeter accuracy [15]. The system may be used to struction of separate 3D CT imaging sets covering the range
position the patient based on the treatment planning CT. It is of respiratory motion observed during a 4DCT. Radiation
also used to monitor the patient during radiation delivery. It oncologists must then decide how to treat a given patient
interfaces with the linac and will turn off the radiation beam [20–22]. One method is to create an internal target volume
if the patient moves out of position. This system is most use- (ITV) consisting of the entire range of tumor motion within
63 Technical Innovations in the Delivery of Radiation Therapy 665

all phases of the 4DCT. The greater the distance traveled by hold it during radiation delivery. For these patients, DIBH is
the tumor during respiration, the larger the ITV is and the used for the treatment planning CT. During breast radiother-
worse the therapeutic ratio is. Another method is to only apy, significant reduction of lung and heart dose is achieved
deliver radiation during certain phases of the respiratory [32]. These systems are also used for the expiration breath
cycle, the gating window. A stable portion of the breathing hold (ExBH) technique where patients exhale and hold dur-
phase where there is little tumor motion, typically near end ing treatment planning CT and treatment delivery. ExBH is
of expiration, is used for the gating window. For this method, typically used in the treatment of lung and abdominal tumors
the ITV is significantly reduced compared to free breathing to minimize motion of the target volume, thus reducing the
since it is only defined in the gating window. Thus, much less amount of normal tissue receiving a high radiation dose [33].
normal tissue volume is irradiated for the same therapeutic A particularly powerful treatment technique for treating
tumor dose. patients having multiple brain metastases has emerged from
The treatment delivery system must also account for the new linac technology. Radiation dose delivery precision
internal patient respiratory motion during patient setup and of VMAT coupled with the accuracy of IGRT allows radia-
treatment delivery. For the linac of Fig. 63.1, this is accom- tion oncologists to treat several metastatic brain lesions
plished with either the respiratory management (RPM) simultaneously using a single isocenter [34]. Stereotactic
(Varian Medical Systems, Palo Alto, CA USA) or the radiosurgery (SRS) is a well-known established modality for
AlignRT system (Vision RT, London, UK). Either system treating brain lesions in one large highly conformal dose.
interfaces with the linac permitting 4D CBCT acquisition Historically, a SRS frame was rigidly attached to a patient by
[23] as well as imaging triggered at a specific respiratory screws penetrating into the skull. The emergence of high-­
phase [24]. RPM utilizes a fixture containing infrared reflec- precision IGRT led to frameless SRS eliminating the skull-­
tors which are monitored by a camera system. The fixture is fixed head frame. However, patients with multiple lesions
placed on the patient, typically in the chest or abdomen, and would still endure long treatment sessions with each lesion
the cameras measure the displacement of the fixture and treated separately. The new single isocenter multiple lesion
hence the patient. AlignRT uses light to detect the surface of SRS technique completes treatment of multiple lesions in the
the patient with a system of 3D cameras employing com- same time taken to treat a single lesion.
puter vision technology (SGRT). The process of using pre-
cise image guidance during CT simulation, treatment
planning, and treatment delivery is called image-guided 63.2.3 MR-Linac
radiation therapy (IGRT) [25].
Improvements in IGRT have led to a paradigm shift in Although the imaging systems on a linac have the submilli-
radiation therapy from treatment courses consisting of daily meter precision of an imaging pixel, they are incapable of
treatments for a period of several weeks to short radiation real-time three-dimensional imaging. Therefore, verification
courses of 3–5 days of highly conformal large dose per frac- of patient positioning and anatomical geometry are based on
tion regimens referred to as stereotactic radiation body ther- snapshots in time, careful monitoring, and reliance on repro-
apy (SBRT) or stereotactic ablative radiation therapy (SABR) ducibility. The ideal situation is having real-time 3D imaging
[26, 27]. These techniques employ treatment plans having during radiation therapy delivery. This is now possible using
high doses per fraction and steep dose gradients. Therefore, an MR-Linac which provides real-time 3D magnetic reso-
treatment accuracy is critical. SBRT treatment protocols nance imaging coupled with radiation therapy delivery.
require the presence of a radiation oncologist and medical There are now two commercial MR-Linac systems avail-
physicist at the treatment machine to review imaging for ana- able: MRIdian (ViewRay, Oakwood Village, Ohio, USA)
tomical alignment prior to radiation delivery [28]. Thus, [35] and Unity (Elekta, Stockholm, Sweden) [36]. The real-­
radiation oncologists are becoming more like surgeons with time imaging capabilities of the MR-Linac enable reduced
the linac room being the analog of the operating theater. radiation beam apertures and an increase in the therapeutic
SBRT has seen use for many disease sites, including to ratio. These systems are capable of monitoring an anatomi-
treat primary tumors in the lung, liver, prostate [29], pan- cal structure, for example, a tumor, and turning on or off the
creas [30, 31], metastatic tumors in the spine, liver [26], as radiation beam when the object moves into or out of posi-
well as to re-irradiate tumors in the head and neck and pelvis tion. The determination of the location of the structure must
[29]. SBRT offers a therapeutic option for inoperable patients be made quickly; therefore, this process operates under com-
and has become the standard of care for inoperable lung pri- puter control. Clinical results indicate that the improved pre-
mary tumors. cision of an MR-Linac offers advantages in the treatment of
Both RPM and SGRT technologies permit radiation treat- a variety of disease sites [37].
ment delivery using the deep inspiration breath hold (DIBH) Upon acquiring images to position a patient for conven-
technique. In this technique, patients take a deep breath and tionally fractionated radiation therapy or for SBRT treat-
666 R. J. Hamilton

ment, it is possible to observe significant anatomical changes therapy is particularly advantageous for pediatric patients,
from the planning CT that cannot be corrected by reposition- since following successful treatment, these patients are at
ing the patient. These could arise, for example, from patient increased risk for secondary cancers which are significantly
weight gain or loss, more or less edema present, and tumor less likely following proton therapy because of the overall
growth or shrinkage. The original plan therefore no longer lower total body dose for the same therapeutic dose. For the
applies, and a new plan based on the patient’s current anat- treatment of medulloblastoma, requiring treatment of the
omy is needed. Either the imaging acquired at the machine or craniospinal axis, protons have superior outcomes compared
a new planning CT is obtained for the new treatment plan. to photon therapy [40–42].
This adaptive planning process creates a delay in the treat-
ment course. The commercial MR-Linac systems permit
online adaptive planning in a matter of minutes with the 63.4 Radiation Therapy Emerging
patient remaining in the treatment room. Technologies
The real-time imaging and online adaptive planning fur-
ther alter the treatment role for radiation oncologists whose 63.4.1 Biology-Guided Radiotherapy
presence is essential during MR-Linac treatment delivery.
Positron-emission tomography (PET) imaging generates
volumetric images of functional and molecular biological
63.3 Proton Therapy processes using radiotracers designed to target specific pro-
cesses. The tracer 18F-FDG is used in clinical oncology for
Proton therapy is external beam radiation therapy using a detecting cancer metastasis. This PET scan is commonly
proton beam instead of photons (X-rays). Because of its used in radiation therapy treatment planning to identify tar-
mass and charge, interactions of protons with patient tissues gets. The PET and treatment planning scans are registered in
are significantly different than photons providing several the software, and a treatment plan is generated to treat the
advantages. Proton beams have a lower entrance (skin) dose tumor(s). For patient delivery, the radiation oncologist relies
and less scatter dose, deposit most of their dose in a narrow on the treatment planning CT to guide patient setup.
range of depths and have minimal exit dose. In fact, proton A new technology has recently been developed to com-
therapy delivers significantly less (50–60%) total body radia- bine PET and CT imaging with a 6MV medical linear accel-
tion dose than photons for the same tumor dose [38]. A dis- erator (RefleXion Medical, Hayward, CA, USA) and has
advantage of proton therapy is that the range of a proton been called biology-guided radiation therapy (BgRT) [43].
beam is strongly dependent on the tissue density in its path. The machine resembles a PET/CT scanner. A treatment plan-
Thus, it is more susceptible to patient anatomical variations ning process, similar to traditional radiotherapy planning, is
than photons. employed. A diagnostic PET is registered with a treatment
Protons were first suggested for therapy use in 1946 [39]. planning CT to designate the targets and subsequently con-
It was a major technological achievement over the course of struct a treatment plan. The departure from the conventional
decades to create a practical medical proton machine from radiotherapy is that PET is used to guide the radiation treat-
the realm of high-energy particle physics experiments. ment delivery. Prior to treatment, a patient is injected with a
Proton therapy machines are significantly more complex, PET radiotracer. The patient is placed on the couch which is
require a large footprint, and are more than an order of mag- advanced into the bore of the machine. A CT and PET scan
nitude more expensive than a photon linac. are acquired and compared with the treatment planning scans
The most precise method of proton therapy is intensity-­ to confirm patient anatomy. During treatment delivery, the
modulated proton therapy (IMPT). It is achieved using pen- gantry, containing PET detectors spanning symmetrically in
cil beam scanning where a narrow proton beam is swept opposed 90°arcs, a CT system, and a linear accelerator,
across the target using magnetic control without the need for rotates at 60 RPM around the patient. PET emissions are
beam aperture shaping collimators or compensators. Beams detected, and limited time-sampled PET images are continu-
from several directions are used to treat a target. This method ously reconstructed at 10 frames/s. The linear accelerator
permits close conforming of the dose to the target while lim- delivers radiation beams from 100 firing positions around the
iting doses to normal tissues. circle. The couch remains stationary at a so-called beam sta-
Not all cancer patients can be treated with protons, tion until the desired dose is delivered. It then moves 2.1 mm
because there are a limited number of proton centers. Proton to another beam station.
therapy is most useful when normal tissue doses cannot be BgRT is a promising technology. It removes difficulties
adequately reduced with photon therapy while still benefit- associated with tumor tracking since the radiation is deliv-
ing the patient. Proton therapy is also desirable when sparing ered to the PET avid volumes, so that the tumors serve as
normal tissues or organs is of primary importance. Proton their own fiducial markers. Additionally, PET tracers
63 Technical Innovations in the Delivery of Radiation Therapy 667

designed for a specific cancer could be utilized, for example, cantly more challenging than for protons due to the larger
PSMA-targeted tracers for prostate cancer or CA9-targeted mass and charge. The enormous costs of building a heavy
tracers for renal cancers. Possible variations in the PET sig- particle facility have thus far prevented constructing one in
nal due to inflammation, cancer cell death, or radiotracer the United States. However, Mayo Clinic announced plans to
administration over the course of treatment have to be build one in Florida [49]. Malouff [46] provides an extensive
accounted for and appear to be manageable [43]. review of the current state of the technology of this
modality.

63.4.2 FLASH Radiotherapy

63.4.4 I mmunotherapy Boost of the Radiation
FLASH radiotherapy (FLASH-RT) is the delivery of radia- Abscopal Effect
tion treatment at ultra-high dose rates (~100 Gy/s) that are
several orders of magnitude greater than used in current Although radiation is deposited locally during radiotherapy
practice (~0.1 Gy/s). Using this new technology, preliminary delivery, there are numerous reports of radiotherapy treat-
pilot and animal studies have shown that normal tissues are ment of a local site of disease leading to regression of meta-
protected when exposed to FLASH-RT compared to conven- static tumors not irradiated [50–52]. This phenomenon,
tional dose rate RT [44]. This so-called FLASH effect is known as the abscopal effect, is believed to be mediated by
robust since it is seen in several animal types and several lymphocytes and enhanced by the checkpoint blockade [53].
organs and by numerous investigators over a 40-year period. Patients receiving radiation therapy to control tumors pro-
Furthermore, FLASH-RT does not spare tumor cells as the gressing during checkpoint blockade immunotherapy have
anti-tumor efficacy of FLASH-RT has been found to be shown systemic responses indicating that radiation has an
equal to conventional RT. Another advantage of FLASH-RT immunostimulatory effect. There are several additional
is that the dose is delivered so rapidly, and organ and tumor mechanisms proposed to explain the abscopal effect [52, 54].
motion are not important during delivery. There is a growing consensus that radiotherapy combined
FLASH is believed to cause a rapid consumption of local with immunotherapy has a role in treating metastatic disease
oxygen at a rate much faster than tissue reoxygenation kinet- and that SBRT may provide an advantage over conventional
ics. This rapid oxygen depletion creates an induced hypoxia. radiotherapy. A comprehensive review of clinical trials
The depletion rate is optimized at dose rates greater than investigating this combination therapy concluded that it has
100 Gy/s. Additionally, FLASH deposits significantly more the potential to transform cancer treatment [55].
energy and ionizes more electrons than conventional Clinical trials combining immunotherapy with SBRT or
RT. Therefore, the redox reactions following FLASH propa- the emerging technologies of biology-guided radiotherapy or
gate in tissues through a series of biochemical and biophysi- FLASH-RT might prove valuable in the treatment of meta-
cal reactions that are kinetically different than following static or oligometastatic disease.
conventional RT. The result is that normal tissues have lower
pro-oxidant burdens during steady-state metabolism than
tumors and eliminate FLASH-mediated free radicals much 63.5 Conclusion
more effectively, maximizing differences in redox metabo-
lism between tumors and normal tissues [44]. Current technology provides high-precision radiation
FLASH-RT might be the next technological advancement treatment delivery to static and moving targets, enabling
to significantly improve the therapeutic ratio. The first case radiation oncologists to treat patients effectively with
report of human use of FLASH-RT has been published [45]. treatment regimens having only a few high dose fractions
Future randomized clinical trials are needed to prove its (SBRT), producing a paradigm shift in their practice which
efficacy. also necessitates their presence at the treatment console
for each treatment fraction. Additional advantages of
SBRT over conventional radiotherapy may also be found
63.4.3 Carbon Ion Therapy when used in combination with immunotherapy. Once
technology makes online adaptive planning, available
Radiation therapy using carbon ions or other heavy nuclei today with MR-Linacs, expeditious, SBRT treatment pre-
instead of protons is offered in only 13 centers worldwide cision will greatly improve.
[46]. Most notably are centers in Germany [47] and Japan There are several technologies on the verge of impacting
[48]. These ions offer reduced doses to normal tissues and radiotherapy practice. Incorporating artificial intelligence
increased radiobiological effect on tumors. The technologi- into treatment planning software should improve plan qual-
cal requirements for these particle accelerators are signifi- ity over current knowledge-­based planning. Biology-guided
668 R. J. Hamilton

radiotherapy may improve treatment efficacy of primary and 2019;9:239. https://doi.org/10.3389/fonc.2019.00239. PMID:
metastatic tumors, especially if novel functional radiotracers 31024843; PMCID: PMC6465886
7. Sahiner B, Pezeshk A, Hadjiiski LM, Wang X, Drukker K, Cha
are utilized. FLASH-RT-capable treatment delivery machines KH, Summers RM, Giger ML. Deep learning in medical imaging
will offer near-instantaneous radiation delivery as well as and radiation therapy. Med Phys. 2019;46(1):e1–e36. https://doi.
biological advantages greatly improving the therapeutic org/10.1002/mp.13264. Epub 2018 Nov 20. PMID: 30367497
ratio. It is reasonable to expect that one or several of these 8. Ge Y, Wu QJ. Knowledge-based planning for intensity-modulated
radiation therapy: a review of data-driven approaches. Med Phys.
technologies will be combined yielding even better patient 2019;46(6):2760–75. https://doi.org/10.1002/mp.13526. Epub
outcomes. 2019 Apr 24. PMID: 30963580; PMCID: PMC6561807
9. Chatterjee A, Serban M, Abdulkarim B, Panet-Raymond V,
Souhami L, Shenouda G, Sabri S, Jean-Claude B, Seuntjens
J. Performance of knowledge-based radiation therapy planning
Open Questions for the glioblastoma disease site. Int J Radiat Oncol Biol Phys.
• Will the clinical practice of radiation oncologists 2017;99(4):1021–8. https://doi.org/10.1016/j.ijrobp.2017.07.012.
morph into that of a surgeon with radiation oncolo- Epub 2017 Jul 14. PMID: 28870791
10. Foy JJ, Marsh R, Ten Haken RK, Younge KC, Schipper M, Sun
gists working at the treatment machine supervising Y, Owen D, Matuszak MM. An analysis of knowledge-based
computer-controlled radiation delivery in real time? planning for stereotactic body radiation therapy of the spine.
• Will treatment planning systems implement AI soft- Pract Radiat Oncol. 2017;7(5):e355–60. https://doi.org/10.1016/j.
ware that will take over anatomy contouring and prro.2017.02.007. Epub 2017 Mar 2. PMID: 28433524
11. Glide-Hurst C, Bellon M, Foster R, Altunbas C, Speiser M, Altman
treatment plan construction? M, Westerly D, Wen N, Zhao B, Miften M, Chetty IJ, Solberg
• If FLASH-RT is proven, will it be scalable to clini- T. Commissioning of the Varian TrueBeam linear accelerator: a
cal machines, significantly reducing treatment multi-institutional study. Med Phys. 2013;40(3):031719. https://
delivery times? doi.org/10.1118/1.4790563. PMID: 23464314
12. Papanikolaou N, Battista JJ, Boyer AL, Kappas C, Klein E,
• Will biology-guided radiation therapy be widely Mackie TR, Sharpe M, Van Dyk J. AAPM Report No. 85 Tissue
adopted? Inhomogeneity Corrections for Megavoltage Photon Beams Report
• Will carbon ion therapy prove beneficial enough to of Task Group No. 65 of the Radiation Therapy Committee of the
justify its cost? American Association of Physicists in Medicine. Medical Physics
Publishing. Madison, WI; 2004. ISBN: 978-1-888340-47-1.
13. van Elmpt W, McDermott L, Nijsten S, Wendling M, Lambin P,
Mijnheer B. A literature review of electronic portal imaging for
radiotherapy dosimetry. Radiother Oncol. 2008;88(3):289–309.
Patient Information and Guidelines https://doi.org/10.1016/j.radonc.2008.07.008. Epub 2008 Aug 14.
American Society for Radiation Oncology: https://www. PMID: 18706727
astro.org/ 14. Graulieres E, Kubler S, Martin E, Ferrand R. Positioning accuracy
American Associate of Physicist in Medicine: https:// of a single-isocenter multiple targets SRS treatment: a comparison
between Varian TrueBeam CBCT and Brainlab ExacTrac. Phys
www.aapm.org Med. 2020;80:267–73. https://doi.org/10.1016/j.ejmp.2020.10.022.
Epub ahead of print. PMID: 33221708
15. Hoisak JDP, Pawlicki T. The role of optical surface imaging sys-
tems in radiation therapy. Semin Radiat Oncol. 2018;28(3):185–
References 93. https://doi.org/10.1016/j.semradonc.2018.02.003. PMID:
29933878
1. Grubbe EH. X-rays in the treatment of cancer and other malignant 16. Batista V, Meyer J, Kügele M, Al-Hallaq H. Clinical paradigms
diseases. Med Rec. 1902;62(18):692–5. and challenges in surface guided radiation therapy: where do we
2. Thwaites DI, Tuohy JB. Back to the future: the history and go from here? Radiother Oncol. 2020;26:S0167-8140(20)30816-1.
development of the clinical linear accelerator. Phys Med Bio. https://doi.org/10.1016/j.radonc.2020.09.041. Epub ahead of print.
2006;51:R343–62. PMID: 32987044.
3. Olch AJ, Gerig L, Li H, Mihaylov I, Morgan A. Dosimetric effects 17. Freislederer P, Kügele M, Öllers M, Swinnen A, Sauer TO, Bert C,
caused by couch tops and immobilization devices: Report of Giantsoudi D, Corradini S, Batista V. Recent advanced in Surface
AAPM Task Group 176. Med Phys. 2014;41:061501. https://doi. Guided Radiation Therapy. Radiat Oncol. 2020;15(1):187.
org/10.1118/1.4876299. https://doi.org/10.1186/s13014-­020-­01629-­w. Erratum in: Radiat
4. Khan FM, Gibbons JP. Khan’s the physics of radiation therapy. Oncol. 2020 Oct 24;15(1):244. PMID: 32736570; PMCID:
Philadelphia, PA: Lippincott Williams & Wilkins; 2014. PMC7393906.
5. Rigaud B, Simon A, Castelli J, Lafond C, Acosta O, Haigron P, 18. Aznar MC, Warren S, Hoogeman M, Josipovic M. The impact of
Cazoulat G, de Crevoisier R. Deformable image registration for technology on the changing practice of lung SBRT. Phys Med.
radiation therapy: principle, methods, applications and evaluation. 2018;47:129–38.
Acta Oncol. 2019;58(9):1225–37. https://doi.org/10.1080/02841 19. Kwong Y, Mel AO, Wheeler G, Troupis JM. Four-dimensional
86X.2019.1620331. Epub 2019 Jun 3. PMID: 31155990 computed tomography (4DCT): a review of the current status and
6. Lee H, Lee E, Kim N, Kim JH, Park K, Lee H, Chun J, Shin JI, applications. J Med Imaging Radiat Oncol. 2015;59(5):545–54.
Chang JS, Kim JS. Clinical evaluation of commercial atlas-based 20. Keall PJ, Mageras GS, Balter JM, Emery RS, Forster KM, Jiang
auto-segmentation in the head and neck region. Front Oncol. SB, Kapatoes JM, Low DA, Murphy MJ, Murray BR, Ramsey CR,
63 Technical Innovations in the Delivery of Radiation Therapy 669

Van Herk MB, Vedam SS, Wong JW, Yorke E. The management upper-abdominal radiation therapy (BURDIE): addressing the chal-
of respiratory motion in radiation oncology report of AAPM Task lenge of a moving target. Radiat Oncol. 2020;15(1):1–7.
Group 76a. Medical Physics. 2006;33(10):3874–900. 34. Clark GM, Popple RA, Young PE, Fiveash JB. Feasibility of
21. Koksal C, Donmez Kesen N, Okutan M, Karaman S, Dagoglu single-isocenter volumetric modulated arc radiosurgery for treat-
Sakin N, Bilge H. Investigation of approaches for internal target ment of multiple brain metastases. Int J Radiat Oncol Biol Phys.
volume definition using 4-dimensional computed tomography 2010;76(1):296–302. https://doi.org/10.1016/j.ijrobp.2009.05.029.
in stereotactic body radiotherapy of lung cancer. Med Dosimy. 35. Mittauer KE, Hill PM, Bassetti MF, Bayouth JE. Validation of an
2021;46(2):136–42. MR-guided online adaptive radiotherapy (MRgoART) program:
22. Korreman S, Persson G, Nygaard D, Brink C, Juhler-Nøttrup deformation accuracy in a heterogeneous, deformable, anthropo-
T. Respiration-correlated image guidance is the most important morphic phantom. Radiother Oncol. 2020;146:97–109. Web.
radiotherapy motion management strategy for most lung cancer 36. Snyder JE, St-Aubin J, Yaddanapudi S, Boczkowski A, Dunkerley
patients. Int J Radiat Oncol Biol Phys. 2012;83(4):1338–43. DAP, Graves SA, Hyer DE. Commissioning of a 1.5T Elekta Unity
23. Kadoya N, Nemoto H, Kajikawa T, Nakajima Y, Kanai T, Ieko Y, MR-linac: a single institution experience. J Appl Clin Med Phys.
Ikeda R, Sato K, Dobashi S, Takeda K, Jingu K. Evaluation of four-­ 2020;21(7):160–72.
dimensional cone beam computed tomography ventilation images 37. Fischer-Valuck BW, Henke L, Green O, Kashani R, Acharya
acquired with two different linear accelerators at various gantry S, Bradley JD, Robinson CG, Thomas M, Zoberi I, Thorstad W,
speeds using a deformable lung phantom. Phys Med. 2020;77:75– Gay H, Huang J, Roach M, Rodriguez V, Santanam L, Li H, Li
83. https://doi.org/10.1016/j.ejmp.2020.07.030. Epub 2020 Aug 11. H, Contreras J, Mazur T, Hallahan D, Olsen JR, Parikh P, Mutic
PMID: 32795891 S, Michalski J. Two-and-a-half-year clinical experience with the
24. King RB, Agnew CE, O'Connell BF, Prise KM, Hounsell AR, world's first magnetic resonance image guided radiation therapy
McGarry CK. Time-resolved dosimetric verification of respiratory-­ system. Adv Radiat Oncol. 2017;2(3):485–93.
gated radiotherapy exposures using a high-resolution 2D ionisa- 38. Shravan K. Spot-scanning beam proton therapy vs intensity-­
tion chamber array. Phys Med Biol. 2016;61(15):5529–46. https:// modulated radiation therapy for ipsilateral head and neck
doi.org/10.1088/0031-­9155/61/15/5529. Epub 2016 Jul 6. PMID: malignancies: a treatment planning comparison. Med Dosim.
27384459 2013;38(4):390–4. https://doi.org/10.1016/j.meddos.2013.05.001.
25. Michalski J, Purdy JA, Gaspar L, Souhami L, Ballow M, Bradley PMID 23916884
J, Chao CKS, Crane C, Eisbruch A, Fallowil D, Forster K, Fowler 39. Wilson RR. Radiological use of fast protons. Radiology.
J, Gillin MT, Graham ML, Harms WB, Saiful Huq M, Kline RW, 1946;47:487–91.
Langer M, Mackie TR, Mukherji S, Podgorsak EB, Roach M, Ryu 40. Eaton BR. Endocrine outcomes with proton and photon radiotherapy
J, Sandler H, Schultz CJ, Schell M, Verhey LJ, Vicini F, Winter for standard risk medulloblastoma. Neuro Oncol. 2016;18(6):881–
KA. Image-guided radiation therapy committee. Int J Radiat Oncol 7. https://doi.org/10.1093/neuonc/nov302. PMC 4864263. PMID
Biol Phys. 2001;51(3):60–5. 26688075
26. Folkert MR, Timmerman RD. Stereotactic ablative body radio- 41. Hill-Kayser CE. Outcomes after proton therapy for treatment of
surgery (SABR) or Stereotactic body radiation therapy (SBRT). pediatric high-risk neuroblastoma. Int J Radiat Oncol Biol Phys.
Adv Drug Deliv Rev. 2017;109:3–14. https://doi.org/10.1016/j. 2019;104(2):401–8. https://doi.org/10.1016/j.ijrobp.2019.01.095.
addr.2016.11.005. Epub 2016 Dec 5. PMID: 27932046 PMID 30738983
27. Zindler JD, Thomas CR, Hahn SM, Hoffmann AL, Troost EGC, 42. Kahalley LS. Superior intellectual outcomes after proton radiother-
Lambin P. Increasing the therapeutic ratio of stereotactic ablative apy compared with photon radiotherapy for pediatric medulloblas-
radiotherapy by individualized isotoxic dose prescription. J Natl toma. J Clin Oncol. 2019;38(5):454–61. https://doi.org/10.1200/
Cancer Inst. 2016;108(2):Djv305. https://doi.org/10.1093/jnci/ JCO.19.01706. PMC 7007288. PMID 31774710
djv305. 43. Shirvani SM, Huntzinger CJ, Melcher T, Olcott PD, Voronenko
28. Halvorsen PH, Cirino E, Das IJ, Garrett JA, Yang J, Yin FF, Y, Bartlett-Roberto J, et al. Biology-guided radiotherapy: rede-
Fairobent LA. AAPM-RSS Medical Physics Practice Guideline 9.a. fining the role of radiotherapy in metastatic cancer. Br J Radiol.
for SRS-SBRT. J Appl Clin Med Phys. 2017;18(5):10–21. https:// 2020;93:20200873.
doi.org/10.1002/acm2.12146. Epub 2017 Aug 8. PMID: 28786239; 44. Vozenin MC, Hendry JH, Limoli CL. Biological benefits of
PMCID: PMC5874865 ultra-high dose rate FLASH radiotherapy: sleeping beauty awo-
29. Rubio C, Morera R, Hernando O, Leroy T, Lartigau SE. Extracranial ken. Clin Oncol (R Coll Radiol). 2019;31(7):407–15. https://
stereotactic body radiotherapy. Review of main SBRT features doi.org/10.1016/j.clon.2019.04.001. Epub 2019 Apr 19. PMID:
and indications in primary tumors. Rep Pract Oncol Radiother. 31010708; PMCID: PMC6850216
2013;18(6):387–96. 45. Bourhis J, Sozzi WJ, Jorge PG, Gaide O, Bailat C, Duclos F,
30. De Geus SWL, Eskander MF, Kasumova GG, Ng SC, Kent TS, Patin D, Ozsahin M, Bochud F, Germond JF, Moeckli R, Vozenin
Mancias JD, Callery MP, Mahadevan A, Tseng JF. Stereotactic MC. Treatment of a first patient with FLASH-radiotherapy.
body radiotherapy for unresected pancreatic cancer: a nationwide Radiother Oncol. 2019;139:18–22. https://doi.org/10.1016/j.
review. Cancer. 2017;123(21):4158–67. radonc.2019.06.019. Epub 2019 Jul 11. PMID: 31303340
31. Gkika E, Adebahr S, Schimek-Jasch T, Wiehle R, Zirlik K, Wittel 46. Malouff TD, Mahajan A, Krishnan S, Beltran C, Seneviratne DS,
U, Nestle U, Grosu A, Brunner TB. Stereotactic body radiotherapy Trifiletti DM. Carbon ion therapy: a modern review of an emerg-
(SBRT) in recurrent or oligometastatic pancreatic cancer: simulta- ing technology. Front Oncol. 2020;10:82. https://doi.org/10.3389/
neus intergrated protection (SIP) versus conventional SBRT a tox- fonc.2020.00082.
icity review of two different treatment approaches. J Clin Oncol. 47. Jäkel O, Krämer M, Schulz-Ertner D, Heeg P, Karger CP, Didinger
2016;34(15 suppl):E15692. B, Nikoghosyan A, Debus J. Treatment planning for carbon ion
32. Nissen HD, Appelt AL. Improved heart, lung and target dose with radiotherapy in Germany: review of clinical trials and treatment
deep inspiration breath hold in a large clinical series of breast can- planning studies. Radiother Oncol. 2004;73(2):S86–91.
cer patients. Radiother Oncol. 2013;106(1):28–32. 48. Tsujii H, Mizoe J, Kamada T, Baba M, Tsuji H, Kato H, Kato S,
33. Farrugia B, Khor R, Foroudi F, Chao M, Knight K, Wright et al. Clinical results of carbon ion radiotherapy at NIRS. J Radiat
C. Protocol of a study investigating breath-hold techniques for Res. 2007;48(Suppl A):A1–A13.
670 R. J. Hamilton

49. Mayo Clinic plans first carbon ion therapy facility in North 53. Buchwald ZS, Wynne J, Nasti TH, Zhu S, Mourad WF, Yan W,
America. Forefront Mayo Clinic Cancer Center’s Online Magazine; Gupta S, Khleif SN, Khan MK. Radiation, immune checkpoint
2020 March;9(1). blockade and the abscopal effect: a critical review on timing, dose
50. Ngwa W, Irabor OC, Schoenfeld JD, Hesser J, Demaria S, Formenti and fractionation. Front Oncol. 2018;8:612.
SC. Using immunotherapy to boost the abscopal effect. Nat Rev 54. Kang J, Demaria S, Formenti S. Current clinical trials testing the
Cancer. 2018;18(5):313–22. combination of immunotherapy with radiotherapy. J Immunother
51. Rodríguez-Ruiz ME, Vanpouille-Box C, Melero I, Formenti SC, of Cancer. 2016;4(1):51.
Demaria S. Immunological mechanisms responsible for radiation-­ 55. Lubas MJ, Kumar SS. The combined use of SBRT and immuno-
induced abscopal effect. Trends Immunol. 2018;39(8):644–55. therapy-­a literature review. Curr Oncol Rep. 2020;22(12):117.
52. Siva S, MacManus MP, Martin RF, Martin OA. Abscopal effects of
radiation therapy: a clinical review for the radiobiologist. Cancer
Lett. 2015;3569(1):82–90.
 Radiation Therapy in the Definitive
Treatment of Cancer 64
Alexander N. Garcia and Baldassarre Stea

Abstract
Learning Objectives
Radiation therapy is an important component of cancer • Understand the biology and physics underlying
treatment for a variety of malignancies both in the defin- radiation therapy.
itive and palliative setting. Radiation can be adminis- • Understand the treatment planning process and var-
tered by external means using linear accelerators or ious techniques for radiation therapy
internally by brachytherapy; core to the therapeutic administration.
effects of radiation is the induction of ionization events • Understand radiation therapy indications in the
within biological molecules which in turn results in management of common malignancies.
irreparable DNA damage and ultimately cell death.
Historically, the treatment effectiveness of radiotherapy
was limited in part by toxicity to normal tissues; how-
ever, recent advances in treatment planning and imaging
have allowed for highly conformal delivery of higher 64.1 Introduction
doses of radiation that subsequently have improved can-
cer-related outcomes. Radiation therapy is widely used Radiation therapy is a widely utilized modality in the treat-
as a definitive modality, concurrently with chemother- ment of cancer, with an estimated 50% of all patients poten-
apy for unresectable cancers, in the adjuvant setting tially benefitting from its utilization during the course of
after surgery or for consolidation after chemotherapy. their illness [1]. The successful application of radiation treat-
With the advent of increasingly specific small-molecule ments relies on the precise delivery of high-energy X-rays,
inhibitors and immunotherapy drugs, the oncologic particles (protons or carbon ions), or radioactive substances
landscape continues to evolve rapidly, creating new to induce death of malignant cells. The feasibility of radia-
opportunities for utilization of radiation therapy to tion treatments can be dated back to the discovery of X-rays
improve cancer-related outcomes. by Wilhelm Röntgen in 1885, although at that time the
potential of X-rays was limited to very superficial tumors
[1]. It is, thus, not surprising that the evolution of radiation
therapy as a cancer treatment modality has followed the
same path as that of medical imaging. The field of radiation
therapy has constantly evolved over the past century, aiming
for more accuracy and fewer side effects with each evolu-
tion. Indeed, precise targeting of a tumor while sparing
nearby normal tissues has increased steadily with the avail-
ability of image guidance before and during the delivery of
A. N. Garcia (*) radiation. Today, most linear accelerators (linacs) come
University of Radiation Oncology, Tucson, USA
e-mail: [email protected] equipped with the ability of obtaining a CT scan prior to
treatment to ascertain precision of targeting. A new emergent
B. Stea (*)
University of Arizona, Tucson, USA way to increase precision is by using MR-guided radiation;
e-mail: [email protected] the MRI-linac combines in one instrument a high-strength

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 671
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_64
672 A. N. Garcia and B. Stea

magnetic resonance imaging with a linear accelerator to pro- subsequent cell death. This can take place following direct
vide heightened precision, especially when targeting soft tis- ionization events in which photons and/or charged particles
sue tumors such as liver tumors. These new-generation (electrons, protons, alpha particles, heavy ions) directly
MRI-linacs are capable of monitoring the movement of a damage DNA to induce single- and double-­stranded breaks.
tumor in real time and continually adjust the radiation beam More commonly, however, indirect ionization occurs during
to compensate for any shift in its position. Aided by artificial which high-energy photons react with other molecules (such
intelligence, it is also possible to alter the daily treatment as water) to generate free radicals which in turn cause DNA
plan based on tumor volume and/or anatomical changes damage. The unit of radiation dose is gray (Gy), defined as
occurring in real time; this new technology gives a radiation the absorption of 1 J of radiation per kilogram of matter.
oncologist the ability to adapt the dose of radiation to a Numerous potential molecular and cellular outcomes occur
changing tumor during the course of radiation (adaptive after exposure to ionizing radiation, including chromosomal
radiotherapy). Radiation therapy is commonly used to treat aberrations, genetic mutations, disruption of cell cycle pro-
primary cancers, such as prostate cancer with the goal of not gression, and apoptosis. Theoretically, a sufficient increase
only controlling the primary but also preventing the insur- in the total radiation dose should improve the probability of
gence of metastatic disease (vide infra). In addition, radia- cell kill and tumor control to the point where any tumor
tion therapy plays an important role in controlling symptoms could be eradicated; however, in clinical practice this must
from metastatic disease (such as brain metastases, painful be balanced with the probability of normal tissue complica-
metastases, bleeding, spinal cord compression, airway tions. This concept is referred to as the therapeutic ratio and
obstruction) and indeed about fifty percent of patients treated is one of the most fundamental aspects of radiation therapy
with radiation are treated with palliative intent; this topic is in clinical practice [2]. Thus, maintenance of the therapeutic
covered in detail in another chapter. Finally, over the past ratio within a narrow range is the primary dose-limiting fac-
three decades, we have witnessed the benefit of combining tor in the delivery of radiation treatments. This limitation
chemotherapy with radiation in the treatment of major malig- necessitated fractionated radiation treatments, a practice of
nancies (e.g., lung, rectal, head and neck cancers). The past administering a smaller daily dose of radiation over a period
decade has also witnessed the emergence of immune check- of weeks to allow normal tissue to repair damage between
point inhibitors (ICI) in the treatment of many malignancies. treatments. In modern practice, advances in treatment plan-
There is now emergent evidence of synergistic interaction ning algorithms, dose conformality, and imaging before or
between high-dose per fraction radiation and ICI in the treat- during radiation treatments have dramatically improved our
ment of certain malignancies such as melanoma. Finally, as ability to spare normal tissues, thus ensuring targeted deliv-
the landscape of oncologic treatments continues to evolve ery of high-dose radiation to the tumor. These technological
with highly promising discoveries in targeted small-mole- advances have led to a change in practice patterns whereby
cule inhibitors and immunotherapy, the role of radiation most tumors are now being treated with hypofractionated
remains prominent in the management of numerous malig- radiation (fewer fractions of larger doses).
nancies in a number of ways including definitive radiother-
apy alone, concurrent chemoradiation, and adjuvant
radiotherapy following surgical resection of primary tumors. 64.3 Radiation Treatment Techniques
This chapter provides the reader with an overview of the
radiobiological and physics aspect of modern radiotherapy External beam radiotherapy (EBRT) from linear accelerators
in addition to highlighting the successful application of radi- (linacs) remains the most utilized technique for radiation
ation in the treatment of major cancers. treatments worldwide. With this modality a patient lies on a
flat surface (couch), while an external treatment machine
directs a beam of radiation to an area of interest. Early in the
64.2  hysics and Biology of Radiation
P history of radiation therapy, these units utilized radioisotope
Therapy cobalt-60 as the source of megavoltage radiation, a practice
that remains prevalent in developing countries. Modern-day
The physics and the biological effects of radiation are com- treatment machines, linear accelerators, instead produce
plex fields which are covered more extensively in other chap- high-energy X-rays as well as electrons that deliver radiation
ters; however, here, we review the basic principles to provide to a treatment volume conformal to the tumor (Fig. 64.1).
the reader with the knowledge to better understand the prac- Over the past 50 years, the evolution of EBRT has paral-
ticality of radiation treatments as well as potential future leled advances in imaging modalities. The development of
developments. Core to the generation of radiation-­induced computed tomography allowed for three-dimensional con-
damage is the formation of ionization events within atoms formal radiotherapy (3D CRT) in which the radiation field
which are then able to induce unrepairable DNA damage and conforms to a volume of interest in three dimensions [3].
64 Radiation Therapy in the Definitive Treatment of Cancer 673

image guidance, and treatment delivery have greatly facili-

tated its use [6]. In addition to providing improved patient
convenience by avoiding a protracted course of treatment,
these techniques provide a greater radiobiological dose of
radiation which leads to ablative effects on tumors. The
radiobiology underpinning this technology takes advantage
of the ability of high-dose radiation to cause small vessel
obliteration. This has translated into improved tumor control
rates not previously seen with fully fractionated radiotherapy
in a number of sites including the lung, liver, pancreas, kid-
ney, prostate, brain metastases, and spine [7] (Fig. 64.2).
Brachytherapy, in contrast to external beam radiotherapy,
is an interstitial or intracavitary form of treatment in which a
radioactive source is placed into or adjacent to the tumor of
interest. Brachytherapy requires the precise placement of the
radioactive source within the target tissue, thus allowing
delivery of extremely high doses of conformal radiation from
within. Furthermore, with brachytherapy there is a rapid
dose falloff with increasing distance from the radioactive
source (inverse square law) which allows for excellent
sparing of adjacent normal tissues to an extent that is not
possible with external beam radiotherapy. There are broad
applications of this technique, including intracavitary
brachytherapy in which the source is placed inside a body
cavity (e.g., cervical cancer), interstitial brachytherapy in
which sources are surgically inserted into the body tissue
(e.g., surgical beds for soft tissue sarcoma), or intraluminal
Fig. 64.1 Modern-day linear accelerators. Varian TrueBeam Linear
Accelerator and ViewRay MRIdian with MRI guidance
brachytherapy (radioactive plaques placed in the subtenon
space for treatment of choroidal melanoma). Furthermore,
depending on the disease site and radioisotope used,
This allows for more precise delivery of radiation that better treatment may further be differentiated as low dose rate
spares normal tissues using either physical blocks or more (LDR, <2 Gy/h) or high dose rate (HDR, >12 Gy/h).
commonly multileaf collimators (MLCs) within the treat- Treatment with LDR is generally accomplished over days to
ment machine. MLCs are mechanical lead-leaves that under months, such as in the case of permanent radioactive seed
computer control block the radiation beam to create a con- implant for early prostate cancer. HDR treatments generally
formal field of radiation. More recently, advances in treat- last minutes, as in the case of cervical cancer. Brachytherapy
ment planning software and mechanical engineering have is frequently utilized either alone or in conjunction with
facilitated sophisticated delivery techniques, such as inten- external beam radiotherapy in the management of many
sity-modulated radiotherapy (IMRT) in which the MLCs, malignancies, notably cervical, endometrial, prostate,
under computer control, are able to variably modulate the choroidal melanoma, recurrent brain tumors, and breast
intensity of the radiation beam to allow better conformality cancer.
to the treatment volume [4]. This technology has resulted in
profound advances in treatment delivery, particularly in head
and neck cancers [5], in which higher dose of radiation can 64.4 Modern-Day Applications
feasibly and more safely be administered to tumors while of Radiotherapy for the Most
better sparing critical structures, resulting in reductions of Common Cancers
toxicity and increased tumor control (improved therapeutic
ratio). 64.4.1 Prostate Cancer
Stereotactic body radiation therapy (SBRT) and stereo-
tactic radiosurgery (SRS) are two other increasingly utilized Prostate cancer is the most common non-cutaneous malig-
radiation techniques in which escalated doses of radiation nancy among men worldwide. Management of early-­stage
are administered over the course of one to five treatments. disease can involve active surveillance, prostatectomy, and
Advances in diagnostic imaging, patient immobilization, radiotherapy, while treatment options for high-risk disease
674 A. N. Garcia and B. Stea

Fig. 64.2 Examples of Stereotactic Body Radiation Treatment. Upper: ability to shift dose away from critical structures such as the stomach
A 67 year old man presented with hepatocellular carcinoma within the (yellow) and spinal cord (green) to minimize toxicity. Lower: 82 year
caudate lobe of the liver. Following CT simulation a SBRT plan was old female with early stage adenocarcinoma of the right upper lung
generated with a prescription of 50 Gy (Yellow) in 5 daily fractions, and lobe, treated with 50 Gy (blue) in 5 fractions with an internal boost to
an internal boost to gross tumor to 60 Gy (red). Note the conformality 57.5 Gy (red)
of the prescription isodose lines (Yellow: 50 Gy; Red: 60 Gy) and

are limited to either radical prostatectomy or radiotherapy, dose per fraction is given over a smaller number of frac-
with or without androgen deprivation therapy (ADT). tions or days (usually 4 weeks). This mild hypofraction-
Recently, the UK ProtecT trial randomized patients with ation has the advantage of not only being more convenient
localized prostate cancer to active surveillance, radical pros- for patients, but potentially more effective because pros-
tatectomy, or external beam radiotherapy with ­short-­term tate cancer is a slow-growing tumor which may inherently
ADT. At 10-year follow-up, approximately 50% of patients be more resistant to smaller fraction size. While these
on active surveillance necessitated definitive treatment. No studies have shorter follow-up compared to historic stud-
significant difference in prostate cancer-­specific survival was ies, the early 5-year results show non-inferiority to more
noted between the three groups, although a separate analysis protracted fractionated regimens. Additionally, low-risk
of patient-reported outcomes reported significant difference and favorable intermediate-­risk prostate cancer may be
in toxicity between the treatment arms. Radiotherapy was treated by brachytherapy alone. This is an attractive option
generally associated with worse acute GI toxicity in for some patients who prefer a single or a few invasive
comparison to radical prostatectomy, which instead was procedures over daily external beam treatments that could
associated with slightly worse urinary and sexual function. potentially last up to 8 weeks. The most common and best
Interestingly, despite equivalent survival outcomes patients studied brachytherapy technique is prostate seed implant
who did not receive upfront management with radiotherapy (PSI), in which numerous radioactive seeds are inserted in
or radical prostatectomy were at significantly increased risk the gland transperineally under ultrasound guidance. This
of development of metastatic disease (Metastatic disease per is a form of low-dose rate (LDR) brachytherapy in which
1000 person-years: 6.3 in active surveillance, 2.4 following the radioactive sources (commonly iodine-125 or palla-
radical prostatectomy, and 3.0 following definitive radiother- dium-103) are permanently placed within the prostate
apy). Based on these findings, the authors concluded that 33 (Fig. 64.3). High-dose rate (HDR) brachytherapy is
patients would require definitive management with radio- another modern technique which takes advantage of mod-
therapy to prevent 1 patient developing metastatic disease. ern imaging and treatment planning software.
Thus, low-risk and favorable intermediate-risk prostate Transperineal catheter insertion in the prostate is per-
cancer can be treated with definitive radiotherapy alone. formed under spinal anesthesia followed by application of
By far the most common treatment technique is EBRT to a high-activity radioisotope, iridium-192, over a period of
the prostate gland and some portion or all the seminal minutes. Literature evaluating prostate HDR brachyther-
vesicles. Dose escalation studies have identified the supe- apy continues to mature, although to date biochemical
riority of regimens treating to 78–80 Gy, administered in control rates are comparable to those of other radiation
daily fractions of 1.8–2 Gy, over 7–8 weeks. More techniques. Lastly, as an extrapolation to the theoretical
recently, however, studies have evaluated the utility and benefits of moderately hypofractionated radiotherapy,
safety of hypofractionated radiotherapy, in which a higher SBRT over 5 fractions is increasingly being utilized in the
64 Radiation Therapy in the Definitive Treatment of Cancer 675

Fig. 64.3 Prostate seed implantation of a patient with high-risk prostate dose (110 Gy). Orange line corresponds to 150% isodose line (165 Gy),
cancer following external beam radiation to the prostate and pelvic lym- and red line corresponds to 200% isodose line (220 Gy). The bladder and
phatics. A grid is superimposed over the patients imaging. The orange rectum are contoured (yellow and orange respectively) as avoidance
circles correspond to positions of needle insertion for seed placement. structures to limit dose exposure. Radio-opaque substance in between
Areas adjacent to the urethra (yellow circle) are generally avoided to prostate and rectum represents hydrogel spacer which serves to physi-
minimize toxicity. The green isodose line corresponds to prescription cally displace the rectum from the prostate to minimize dose exposure

management of early-stage prostate cancer. Given the an alternative is treatment of the pelvis with 46 Gy of EBRT
proximity of bladder and rectal tissues, extreme care must followed by a brachytherapy boost which may result in
be taken in treatment delivery, notably accurate daily improved cancer-related outcomes [8].
treatment imaging; however, SBRT is associated with sev-
eral promising endpoints such as tumor control and
patient-reported quality of life. 64.4.2 Breast Cancer
For patients with unfavorable intermediate- or high-risk
prostate cancer, active surveillance is generally not Breast cancer remains one of the most common malignancies,
recommended as nearly all patients may potentially benefit as well as the most common cause of cancer death in women.
from treatment. These patients are treated with some Treatment largely revolves around surgical resection of the
combination of external beam radiotherapy, brachytherapy, tumor, either with breast-conserving therapy or mastectomy,
ADT, or in select instances of all the above. The most followed by adjuvant radiation therapy in most cases [9].
frequently used agents for androgen deprivation are Numerous large-scale trials have shown the combination of
gonadotropin-releasing hormone agonists or oral breast-conserving therapy and adjuvant radiotherapy to be
antiandrogens. For unfavorable intermediate-risk prostate comparable in terms of local recurrence and overall survival.
cancer, a 4–6 month course of ADT is used in conjunction These trials have shown adjuvant radiotherapy to reduce local
with hypofractionated external beam radiotherapy to the recurrence rates in breast cancer patients by nearly two-­thirds,
prostate and seminal vesicles alone. In contrast, patients with and a meta-analysis performed by the Early Breast Cancer
high-risk prostate cancer have been shown to benefit from Trialists’ Collaborative Group (EBCTG) also reported a sur-
ADT for upwards of 2 years. This is combined with external vival benefit for postoperative breast radiotherapy; one breast
beam radiotherapy to both the prostate and regional cancer death was prevented at 15 years for every four local
lymphatics, although the benefits of nodal irradiation are a recurrence avoided [10]. Breast radiotherapy was adminis-
matter of controversy which is currently being evaluated by tered over a protracted course of 5–6 weeks to a dose of
ongoing clinical trials. External beam radiotherapy can be 50–60 Gy, with treatment volumes including the whole breast
administered to a definitive dose of approximately 78–80 Gy; with or without regional lymph nodes depending on risk fac-
676 A. N. Garcia and B. Stea

tors. More recently, based on radiobiologic analysis of breast 64.4.3 Lung Cancer
cancer cells in relation to normal tissue, it was theorized that
hypofractionated courses of radiation over 3–4 weeks may Lung cancer remains the most common cause of cancer
yield a more favorable therapeutic ratio. This has since been death in the United States [18]. The standard of care for
proven in numerous clinical trials [11] and remains the current early-stage non-small-cell carcinomas (NSCLC) remains
standard of care today. The hypofractionation paradigm is surgery; however, patients frequently present with extensive
being explored even further with 5-day courses of whole cardiopulmonary comorbidities which preclude surgical
breast or partial breast radiotherapy showing promising results resection. The standard of care for non-surgical patients is
in select patient populations [12]. Regional lymphatics includ- definitive radiotherapy, specifically SBRT which has been
ing axillary, supraclavicular, and internal mammary lymph shown to be superior to conventionally fractionated
nodes are frequently included in the treatment volume based radiotherapy [19]. There is limited data directly comparing
on the presence of risk factors such as node positivity and SBRT to surgical resection; however, local control rates
hormone-negative breast cancers. Studies such as MA.20 and compared across trials are generally similar.
EORTC 22922 have shown that inclusion of regional nodal Patients with locally advanced disease are often not surgi-
irradiation in these patients yields improved local regional cal candidates due to invasive primary tumors or extensive
control with a trend towards improved overall survival [13]. mediastinal nodal disease. The current standard of care for
Based on the observation that most of the in-breast tumor these patients remains conventionally fractionated radiother-
recurrences (IBTR) occur within one centimeter of the surgi- apy with concurrent platinum-based chemotherapy. The
cal bed, accelerated partial breast irradiation (APBI) has been standard dose of radiation remains 60 Gy/30 daily fractions
proposed, which dramatically reduces the overall treatment as a recent randomized clinical trial evaluating dose escala-
volume and time. To date, there have been multiple trials using tion yielded inferior survival outcomes [20]. Of note this trial
different radiation techniques (external beam radiation, inter- was performed prior to the widespread adoption of IMRT,
stitial brachytherapy, and balloon brachytherapy) to adminis- and thus dose escalation may be more feasible and beneficial
ter larger dose to a small volume which encompasses the in the modern era. Patients that undergo surgical resection do
lumpectomy cavity plus a margin. While there is limited long- not routinely undergo adjuvant radiation therapy; however, a
term follow-up for patients treated with APBI in comparison subset of patients with N2 disease or positive margins may
to conventional whole breast irradiation, early results show benefit from its use [21].
comparable rates of IBTR with excellent cosmetic outcomes Patients with small-cell lung cancer (SCLC) frequently
[14]. Similarly, intraoperative radiotherapy (IORT) has been present with unresectable disease, and definitive chemora-
studied on at least two clinical trials [15] in which a single diation remains the standard of care for patients with limited
high dose of radiotherapy is administered to the lumpectomy stage disease. In contrast to NSCLC, small-cell lung cancer
cavity at the time of surgery. While this offers improved patient is routinely treated with accelerated twice daily radiation
convenience and highly localized radiation dose, these clinical due the rapid cell division and inherently radiosensitive
trials have shown higher rates of IBTR which has largely pre- nature of the disease. This practice pattern is based on the
cluded its incorporation into treatment guidelines. landmark trial by Turrisi, which showed superiority of
Patients with advanced stage disease or high-risk genetics 45 Gy/30 fractions administered b.i.d compared to 45 Gy in
frequently undergo modified radical mastectomy. This proce- once daily fractions [22]. Criticisms of this trial focused pri-
dure removes the entire ipsilateral breast and level one and marily on insufficient dose in the once daily treatment arm,
two axillary lymphatics, and thus adjuvant radiotherapy may and indeed, further evaluation using escalated doses of
not be routinely recommended. However, patients with exten- 66 Gy showed comparable outcomes, although there was a
sive nodal burden, defined as four or more positive lymph trend towards superiority in the twice daily arm [23]. Given
nodes, have been shown to benefit from postmastectomy radi- the high propensity for CNS metastases in patients with
ation based on a meta-analysis performed by the Early Breast SCLC, multiple trials have evaluated the use of prophylactic
Cancer Trialists’ Collaborative Group. Postmastectomy radi- cranial radiation (PCI). A meta-analysis of these trials
ation in this patient population has resulted in a nearly 10% showed that patients with limited-stage disease who have a
absolute risk reduction in breast cancer mortality [16], good response to chemoradiation receive an approximate
although its benefits in patients with less than four positive 5% absolute improvement in 3-year overall survival from
lymph nodes or those that achieve a pathologic complete PCI [24]. However, due to the frequent occurrences of neu-
response to neoadjuvant chemotherapy remain areas of con- rocognitive deficits following PCI, a current clinical trial is
troversy that warrant further evaluation [17]. exploring the use of observation versus SRS in managing
these patients.
64 Radiation Therapy in the Definitive Treatment of Cancer 677

64.4.4 Gynecologic Cancers generally associated with heavy tobacco, alcohol use, and
HPV exposure. Management varies by subsite within the
Early-stage cervical cancers are managed with surgery alone, head and neck; however, early-stage disease can be managed
although patients with high-risk pathologic features including with either surgical resection or definitive radiotherapy with
lymph-vascular space invasion, deep stromal invasion, and or without chemotherapy [30]. The exception to this rule is
large tumors benefit from the addition of adjuvant radiotherapy malignancies located within the oral cavity which are best
[25]. Additionally, concurrent chemotherapy is included for managed by upfront surgical resection [31, 32], although
patients with positive margins, parametrial involvement, and adjuvant radiotherapy is frequently utilized for high-risk
positive lymph nodes [26]. For patients with advanced stage features including advanced stage disease, close or positive
disease or contraindications to surgery, definitive radiation is margins, deep invasion, or extracapsular extension (ECE).
recommended with concurrent platinum-based chemotherapy. Similarly, concurrent chemotherapy is frequently utilized in
Radiotherapy is generally administered in two phases; an initial instances of positive margins or ECE for any type of head
external beam radiation is administered to the primary tumor and neck cancer. Concurrent chemotherapy is generally
and regional lymphatics, with or without a boost to the parame- platinum-based which in addition to eradication of potential
trium, depending on disease stage. This is followed by an intra- distant disease serves as a radiosensitizer. Cetuximab has
cavitary HDR brachytherapy boost to the primary cervical additionally shown some benefit [33], albeit inferior to that
tumor. Local control rates and survival rates are promising with seen with cisplatin for patients with HPV-positive
definitive chemoradiation; however, prolonged treatment times oropharyngeal carcinoma [34]. Outcomes with definitive
(i.e., beyond 56 days) are associated with significantly reduced radiation or chemoradiation vary largely based on smoking
survival outcomes as accelerated repopulation of tumors occurs status, nodal burden, and HPV status. RTOG 0129 [35]
during protracted courses of radiation [27]. demonstrated the significance of these risk factors as low-­
In contrast to cervical cancer, endometrial cancer is primar- risk patients (HPV+, <10 pack years smoking, N0-2a) and
ily managed with surgery; however, the presence of high-risk high-risk patients (HPV-, >10 pack years smoking) resulting
features may necessitate adjuvant radiation, chemotherapy, or in 3-year overall survival rates of approximately 85% and
both. High-risk features include deep myometrial invasion, 55%, respectively. Based on these findings, current
high grade, or aggressive histology, such as clear cell and investigation is underway into potential radiation dose
LVSI. Approximately 75% of recurrences after primary sur- de-escalation in patients with HPV+ head and neck cancer,
gery occur within the vaginal vault [28]. Studies comparing results of which are expected to be reported in the coming
vaginal vault intracavitary brachytherapy with pelvic external years.
beam pelvic radiation showed no difference in vaginal recur-
rences or overall survival; as such, given the more favorable
toxicity profile of vaginal brachytherapy, this method has 64.5 Conclusion
become the standard of care to prevent a vaginal apex recur-
rence. However, patients with numerous high-risk features In addition to the aforementioned sites of disease, radiation
may still benefit from more aggressive management. The therapy has the potential to be impactful to some degree in
GOG 249 study evaluated patients with high intermediate-risk nearly every malignancy including bladder, gastrointestinal,
or early-stage high-risk endometrial cancer; in this trial, CNS, skin, lymphoma, and sarcomas. In addition to treat-
patients were randomized to pelvic radiotherapy or vaginal ment in the definitive setting, radiotherapy additionally has a
cuff brachytherapy plus chemotherapy. No differences were role in the management of metastatic disease where it is
detected in terms of survival outcomes; however, patients commonly utilized for the palliation of painful or symptom-
treated with vaginal cuff plus chemotherapy demonstrated atic sites in locations such as bone, lung, and brain. Although
greater acute toxicities suggesting that pelvic radiotherapy modest radiation doses with single or multifraction regimens
may be the preferred approach in this patient population [29]. have shown promising pain control, increased survival of
Conversely, the PORTEC 3 trial randomized patients with patients with metastatic disease has highlighted the potential
high-risk endometrial cancer to adjuvant pelvic radiation ver- benefit of aggressive SBRT regimens to maximize local con-
sus chemoradiation, demonstrating an overall survival benefit trol, particularly in the oligometastatic setting. The role of
in patients treated with chemotherapy and radiation. radiation therapy in the oligometastatic setting is a recent
area of interest, with preliminary studies showing promising
survival benefits in patients with minimal metastatic burden,
64.4.5 Head and Neck Cancer although this will be discussed in further detail elsewhere.
Common to nearly every application of radiotherapy is an
Head and neck squamous cell carcinomas represent an associated local-regional control benefit, although the ability
aggressive heterogeneous group of malignancies. They are for this to translate into a survival benefit varies based on
678 A. N. Garcia and B. Stea

tumor location and histology. As systemic therapy continues 9. Hughes KS, Schnaper LA, Bellon JR, Cirrincione CT, Berry DA,
McCormick B, Muss HB, Smith BL, Hudis CA, Winer EP, Wood
to improve rapidly for many cancers, it is likely that the role WC. Lumpectomy plus tamoxifen with or without irradiation in
of radiation therapy will become even more crucial to women age 70 years or older with early breast cancer: long-term
patients in the future as a form of consolidation after achiev- follow-up of CALGB 9343. J Clin Oncol. 2013;31(19):2382–7.
ing maximal response to chemotherapy or immunotherapy. https://doi.org/10.1200/jco.2012.45.2615.
10. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG).
Effects of chemotherapy and hormonal therapy for early breast
cancer on recurrence and 15-year survival: an overview of the
Open Questions randomised trials. Lancet. 2005;365(9472):1687–717. https://doi.
• Given ongoing advances in systemic therapy, will org/10.1016/s0140-­6736(05)66544-­0.
11. Whelan TJ, Pignol J-P, Levine MN, Julian JA, MacKenzie R, Parpia
indications for radiotherapy continue to evolve as S, Shelley W, Grimard L, Bowen J, Lukka H, Perera F, Fyles A,
well? What is the clinical advantage of integrating Schneider K, Gulavita S, Freeman C. Long-term results of hypo-
immunotherapy with stereotactic radiosurgery? fractionated radiation therapy for breast cancer. N Engl J Med.
• What will be the effect of more sophisticated imag- 2010;362(6):513–20. https://doi.org/10.1056/NEJMoa0906260.
12. Murray Brunt A, Haviland JS, Wheatley DA, Sydenham MA,
ing modalities on the delivery of radiation Alhasso A, Bloomfield DJ, Chan C, Churn M, Cleator S, Coles
treatments? CE, Goodman A, Harnett A, Hopwood P, Kirby AM, Kirwan CC,
• In addition to treatments with photons and elec- Morris C, Nabi Z, Sawyer E, Somaiah N, Stones L, Syndikus I,
trons, will protons and heavy ions see increased use Bliss JM, Yarnold JR. Hypofractionated breast radiotherapy for 1
week versus 3 weeks (FAST-Forward): 5-year efficacy and late nor-
in cancer treatments? Why? mal tissue effects results from a multicentre, non-inferiority, ran-
• Despite being correlated with improved patient out- domised, phase 3 trial. Lancet. 2020;395(10237):1613–26. https://
comes, the use of brachytherapy has decreased in doi.org/10.1016/s0140-­6736(20)30932-­6.
recent decades. What are some explanations for this 13. Whelan TJ, Olivotto IA, Parulekar WR, Ackerman I, Chua BH,
Nabid A, Vallis KA, White JR, Rousseau P, Fortin A, Pierce LJ,
trend? Manchul L, Chafe S, Nolan MC, Craighead P, Bowen J, McCready
DR, Pritchard KI, Gelmon K, Murray Y, Chapman JA, Chen BE,
Levine MN. Regional nodal irradiation in early-stage breast can-
cer. N Engl J Med. 2015;373(4):307–16. https://doi.org/10.1056/
NEJMoa1415340.
14. Livi L, Meattini I, Marrazzo L, Simontacchi G, Pallotta S, Saieva
References C, Paiar F, Scotti V, De Luca CC, Bastiani P, Orzalesi L, Casella
D, Sanchez L, Nori J, Fambrini M, Bianchi S. Accelerated partial
1. Gianfaldoni S, et al. An overview on radiotherapy: from its history to breast irradiation using intensity-modulated radiotherapy versus
its current applications in dermatology. Open Access Maced J Med whole breast irradiation: 5-year survival analysis of a phase 3 ran-
Sci. 2017;5(4):521–5. https://doi.org/10.3889/oamjms.2017.122. domised controlled trial. Eur J Cancer. 2015;51(4):451–63. https://
2. Hall EJ, Astor M, Bedford J, Borek C, Curtis SB, Fry M, doi.org/10.1016/j.ejca.2014.12.013.
Geard C, Hei T, Mitchell J, Oleinick N, et al. Basic radio- 15. Silverstein MJ, Fastner G, Maluta S, Reitsamer R, Goer DA, Vicini
biology. Am J Clin Oncol. 1988;11(3):220–52. https://doi. F, Wazer D. Intraoperative radiation therapy: a critical analysis of
org/10.1097/00000421-­198806000-­00003. the ELIOT and TARGIT trials. Part 2 – TARGIT. Ann Surg Oncol.
3. Tubiana M, Eschwège F. Conformal radiotherapy and intensity-­ 2014;21(12):3793–9. https://doi.org/10.1245/s10434-­014-­3999-­5.
modulated radiotherapy – clinical data. Acta Oncol. 2000;39(5):555– 16. McGale P, Taylor C, Correa C, Cutter D, Duane F, Ewertz M, Gray
67. https://doi.org/10.1080/028418600750013249. R, Mannu G, Peto R, Whelan T, Wang Y, Wang Z, Darby S. Effect
4. Taylor A, Powell ME. Intensity-modulated radiotherapy – of radiotherapy after mastectomy and axillary surgery on 10-year
what is it? Cancer Imaging. 2004;4(2):68–73. https://doi. recurrence and 20-year breast cancer mortality: meta-analysis
org/10.1102/1470-­7330.2004.0003. of individual patient data for 8135 women in 22 randomised tri-
5. Ghosh G, Tallari R, Malviya A. Toxicity profile of IMRT Vs. als. Lancet. 2014;383(9935):2127–35. https://doi.org/10.1016/
3D-CRT in head and neck cancer: a retrospective study. J Clin s0140-­6736(14)60488-­8.
Diagn Res. 2016;10(9):Xc01–xc03. https://doi.org/10.7860/ 17. Remick J, Amin NP. Postmastectomy breast cancer radiation ther-
jcdr/2016/21457.8583. apy. Treasure Island, FL: StatPearls; 2020.
6. Alongi F, Arcangeli S, Filippi AR, Ricardi U, Scorsetti M. Review 18. Dela Cruz CS, Tanoue LT, Matthay RA. Lung cancer: epidemiol-
and uses of stereotactic body radiation therapy for oligometa­ ogy, etiology, and prevention. Clin Chest Med. 2011;32(4):605–44.
stases. Oncologist. 2012;17(8):1100–7. https://doi.org/10.1634/ https://doi.org/10.1016/j.ccm.2011.09.001.
theoncologist.2012-­0092. 19. Nyman J, Hallqvist A, Lund J, Brustugun OT, Bergman B, Bergström
7. Chang BK, Timmerman RD. Stereotactic body radiation therapy: P, Friesland S, Lewensohn R, Holmberg E, Lax I. SPACE - A
a comprehensive review. Am J Clin Oncol. 2007;30(6):637–44. randomized study of SBRT vs conventional fractionated radio-
https://doi.org/10.1097/COC.0b013e3180ca7cb1. therapy in medically inoperable stage I NSCLC. Radiother Oncol.
8. Morris WJ, Tyldesley S, Rodda S, Halperin R, Pai H, McKenzie M, 2016;121(1):1–8. https://doi.org/10.1016/j.radonc.2016.08.015.
Duncan G, Morton G, Hamm J, Murray N. Androgen suppression 20. Bradley JD, Hu C, Komaki RR, Masters GA, Blumenschein GR,
combined with elective nodal and dose escalated radiation therapy Schild SE, Bogart JA, Forster KM, Magliocco AM, Kavadi VS,
(the ASCENDE-RT Trial): an analysis of survival endpoints for a Narayan S, Iyengar P, Robinson CG, Wynn RB, Koprowski CD,
randomized trial comparing a low-dose-rate brachytherapy boost to Olson MR, Meng J, Paulus R, Curran WJ Jr, Choy H. Long-term
a dose-escalated external beam boost for high- and intermediate-­ results of NRG oncology RTOG 0617: standard- versus high-dose
risk prostate cancer. Int J Radiat Oncol Biol Phys. 2017;98(2):275– chemoradiotherapy with or without cetuximab for unresectable
85. https://doi.org/10.1016/j.ijrobp.2016.11.026.
64 Radiation Therapy in the Definitive Treatment of Cancer 679

stage III non-small-cell lung cancer. J Clin Oncol. 2020;38(7):706– 29. Randall ME, Filiaci V, McMeekin DS, von Gruenigen V, Huang H,
14. https://doi.org/10.1200/jco.19.01162. Yashar CM, Mannel RS, Kim JW, Salani R, DiSilvestro PA, Burke
21. Burdett S, Stewart L. Postoperative radiotherapy in non-small-­ JJ, Rutherford T, Spirtos NM, Terada K, Anderson PR, Brewster
cell lung cancer: update of an individual patient data meta-­ WR, Small W, Aghajanian CA, Miller DS. Phase III trial: adjuvant
analysis. Lung Cancer. 2005;47(1):81–3. https://doi.org/10.1016/j. pelvic radiation therapy versus vaginal brachytherapy plus pacli-
lungcan.2004.09.010. taxel/carboplatin in high-intermediate and high-risk early stage
22. Turrisi AT 3rd, Kim K, Blum R, Sause WT, Livingston RB, endometrial cancer. J Clin Oncol. 2019;37(21):1810–8. https://doi.
Komaki R, Wagner H, Aisner S, Johnson DH. Twice-daily com- org/10.1200/jco.18.01575.
pared with once-daily thoracic radiotherapy in limited small-­cell 30. Kramer S, Gelber RD, Snow JB, Marcial VA, Lowry LD, Davis
lung cancer treated concurrently with cisplatin and etoposide. LW, Chandler R. Combined radiation therapy and surgery in the
N Engl J Med. 1999;340(4):265–71. https://doi.org/10.1056/ management of advanced head and neck cancer: final report of
nejm199901283400403. study 73-03 of the Radiation Therapy Oncology Group. Head Neck
23. Faivre-Finn C, Snee M, Ashcroft L, Appel W, Barlesi F, Bhatnagar Surg. 1987;10(1):19–30. https://doi.org/10.1002/hed.2890100105.
A, Bezjak A, Cardenal F, Fournel P, Harden S, Le Pechoux C, 31. Iyer NG, Tan DS, Tan VK, Wang W, Hwang J, Tan NC, Sivanandan
McMenemin R, Mohammed N, O'Brien M, Pantarotto J, Surmont V, R, Tan HK, Lim WT, Ang MK, Wee J, Soo KC, Tan EH. Randomized
Van Meerbeeck JP, Woll PJ, Lorigan P, Blackhall F. Concurrent once-­ trial comparing surgery and adjuvant radiotherapy versus concur-
daily versus twice-daily chemoradiotherapy in patients with limited-­ rent chemoradiotherapy in patients with advanced, nonmetastatic
stage small-cell lung cancer (CONVERT): an open-label, phase 3, squamous cell carcinoma of the head and neck: 10-year update
randomised, superiority trial. Lancet Oncol. 2017;18(8):1116–25. and subset analysis. Cancer. 2015;121(10):1599–607. https://doi.
https://doi.org/10.1016/s1470-­2045(17)30318-­2. org/10.1002/cncr.29251.
24. Aupérin A, Arriagada R, Pignon JP, Le Péchoux C, Gregor A, 32. Robertson AG, Soutar DS, Paul J, Webster M, Leonard AG,
Stephens RJ, Kristjansen PE, Johnson BE, Ueoka H, Wagner H, Moore KP, McManners J, Yosef HM, Canney P, Errington RD,
Aisner J. Prophylactic cranial irradiation for patients with small-cell Hammersley N, Singh R, Vaughan D. Early closure of a random-
lung cancer in complete remission. Prophylactic Cranial Irradiation ized trial: surgery and postoperative radiotherapy versus radio-
Overview Collaborative Group. N Engl J Med. 1999;341(7):476– therapy in the management of intra-oral tumours. Clin Oncol
84. https://doi.org/10.1056/nejm199908123410703. (R Coll Radiol). 1998;10(3):155–60. https://doi.org/10.1016/
25. Rotman M, Sedlis A, Piedmonte MR, Bundy B, Lentz SS, s0936-­6555(98)80055-­1.
Muderspach LI, Zaino RJ. A phase III randomized trial of postop- 33. Bonner JA, Harari PM, Giralt J, Cohen RB, Jones CU, Sur RK,
erative pelvic irradiation in Stage IB cervical carcinoma with poor Raben D, Baselga J, Spencer SA, Zhu J, Youssoufian H, Rowinsky
prognostic features: follow-up of a gynecologic oncology group EK, Ang KK. Radiotherapy plus cetuximab for locoregionally
study. Int J Radiat Oncol Biol Phys. 2006;65(1):169–76. https:// advanced head and neck cancer: 5-year survival data from a phase 3
doi.org/10.1016/j.ijrobp.2005.10.019. randomised trial, and relation between cetuximab-induced rash and
26. Monk BJ, Wang J, Im S, Stock RJ, Peters WA 3rd, Liu PY, Barrett survival. Lancet Oncol. 2010;11(1):21–8. https://doi.org/10.1016/
RJ 2nd, Berek JS, Souhami L, Grigsby PW, Gordon W Jr, Alberts s1470-­2045(09)70311-­0.
DS. Rethinking the use of radiation and chemotherapy after radi- 34. Gillison ML, Trotti AM, Harris J, Eisbruch A, Harari PM, Adelstein
cal hysterectomy: a clinical-pathologic analysis of a Gynecologic DJ, Jordan RCK, Zhao W, Sturgis EM, Burtness B, Ridge JA,
Oncology Group/Southwest Oncology Group/Radiation Therapy Ringash J, Galvin J, Yao M, Koyfman SA, Blakaj DM, Razaq
Oncology Group trial. Gynecol Oncol. 2005;96(3):721–8. https:// MA, Colevas AD, Beitler JJ, Jones CU, Dunlap NE, Seaward SA,
doi.org/10.1016/j.ygyno.2004.11.007. Spencer S, Galloway TJ, Phan J, Dignam JJ, Le QT. Radiotherapy
27. Vitzthum L, Yuan J, Jones D, Boldt A, Dusenbery K. Reducing plus cetuximab or cisplatin in human papillomavirus-positive oro-
prolonged chemoradiation treatment times for cervical cancer. pharyngeal cancer (NRG Oncology RTOG 1016): a randomised,
BMJ Open Qual. 2019;8(3):e000516. https://doi.org/10.1136/ multicentre, non-inferiority trial. Lancet. 2019;393(10166):40–50.
bmjoq-­2018-­000516. https://doi.org/10.1016/s0140-­6736(18)32779-­x.
28. Creutzberg CL, Nout RA, Lybeert ML, Wárlám-Rodenhuis CC, 35. Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tân
Jobsen JJ, Mens JW, Lutgens LC, Pras E, van de Poll-Franse LV, PF, Westra WH, Chung CH, Jordan RC, Lu C, Kim H, Axelrod R,
van Putten WL. Fifteen-year radiotherapy outcomes of the ran- Silverman CC, Redmond KP, Gillison ML. Human papillomavirus
domized PORTEC-1 trial for endometrial carcinoma. Int J Radiat and survival of patients with oropharyngeal cancer. N Engl J Med.
Oncol Biol Phys. 2011;81(4):e631–8. https://doi.org/10.1016/j. 2010;363(1):24–35. https://doi.org/10.1056/NEJMoa0912217.
ijrobp.2011.04.013.
 Radiation Therapy for Extracranial
Oligometastatic Disease 65
Q. A. Ho and C. C. Hsu

Abstract
Learning Objectives
Oligometastatic cancer is an intermediary state in which a • To define terminology and the characteristics of the
tumor has spread to a limited number (usually ≤5) of oligometastatic state.
metastatic sites rather than widespread (or polymeta- • To address the microscopic and biologic origins of
static). The concept first gained traction in 1995 and has oligometastases and their clinical implications.
served as a rationale for aggressive treatment to improve • To discuss the radiotherapeutic treatment options
survival within this select metastatic population. Although for oligometastases and future directions.
not readily distinguishable on presentation, these oligo-
metastatic cancers tend to progress much more slowly
than an aggressive cancer in its early polymetastatic stage.
The definition has now expanded to include four distinct 65.1 Introduction
categories: oligometastatic, oligorecurrent, oligoprogres-
sive, and oligoresidual. Although imaging and blood bio- Oligometastatic cancer, the intermediary between localized
markers are helpful, time and response to systemic and widely disseminated cancer, is a relatively new concept
therapy have been keys to identifying those patients with in oncology. Historically, the metastatic state of cancer has
oligometastatic disease who may benefit from aggressive been viewed as an inevitably progressive track that begins
local therapy. New studies on molecular profiling, partic- with a single detectable metastasis but in the presence of
ularly on specific micro-RNAs, have shown promise in widespread subclinical dissemination [1]. Subsequently, the
identifying such cases. Although surgical resection, or role of local therapy remained purely for palliation, regard-
metastectomy, was the primary treatment of oligometa- less of disease burden, whether encompassing a single site of
static disease initially, stereotactic body radiation ther- metastasis or widely disseminated disease. The concept of
apy (SBRT) has shown to be a safe and effective cancer metastasis consisting of a widely heterogeneous
alternative. Several phase 2 and 3 randomized trials have group has only recently been hypothesized. This has been
shown that radiation to the metastatic and/or the primary supported by increasingly robust evidence that aggressive
lesion in patients with limited metastatic burden can local treatment, such as radiation therapy, of patients with
improve progression-­free survival and even overall sur- limited metastatic burden provides a benefit to both
vival. As both radiation therapy and systemic therapy progression-­free and overall survival.
improve, so will their synergy and indications for their
continued use.
65.2 Background History

For an entire century, the Halsted hypothesis of cancer spread

was the predominant belief among oncologists [1]. The con-
cept called for a continuous spread of cancer, such that there
was a specific order of spread that began from the primary
Q. A. Ho · C. C. Hsu (*)
tumor, followed by the lymphatics and lymph nodes, and
Department of Radiation Oncology, The University of Arizona,
Tucson, AZ, USA finally to distant sites. Even among distant sites, there was a
e-mail: [email protected] belief that “closer” distant sites were more likely to become

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 681
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_65
682 Q. A. Ho and C. C. Hsu

involved. This hypothesis drove radical en bloc surgeries between studies but is generally considered to be 5 or less [2,
that, although were effective at local and regional control, 3]. The oligometastatic state differs from patients with
involved significant morbidity. The systemic (Fisher) hypoth- widely disseminated metastatic disease, who are classified as
esis of cancer behavior later became prominent as an alterna- having polymetastatic disease. In some studies, bone metas-
tive explanation of cancer spread. It stated that cancer begins tases are considered less severe and therefore more charac-
with systemic dissemination of microscopic metastases teristic of oligometastatic than visceral metastases [4]. This
(micrometastases), and that treatment should be focused on fits in with the theory of oligometastases as some organ sites
systemic control of micrometastases before they become have a greater likelihood of receiving metastases based on
macroscopic [1]. Figures 65.1 and 65.2 illustrate these two the location of the primary tumor and histology as well as the
concepts of metastasis. distant site’s microenvironment. Often, this limited meta-
In 1995, Hellman and Weichselbaum recognized that static state is due to a more indolent biology, and there is a
because cancers begin and can mutate into a wide variety of therapeutic window for localized therapy before widespread
different clonogens that behave differently, metastatic cancer dissemination [1]. For this reason, treatment of oligometas-
represents a spectrum of disease [1]. Thus, there may be a tases is becoming more common and supported by
select population among those with metastases whose dis- literature.
ease course may warrant more aggressive therapy. The oligo- Not all cases of limited metastases can truly be classified
metastatic disease state has been classified as having a as oligometastatic. As imaging represents only a single frame
limited number of visible metastases; this number differs in time, the visibility of a few metastatic lesions could be in

a b c

d e

Fig. 65.1 This diagram represents the Halsted hypothesis of cancer further spread hematogenously to distant (metastatic) sites that are
progression. (a) Cancer begins in a primary location (b) before spread- nearby the local tumor, (e) before widespread dissemination throughout
ing through the lymphatics (c) to the lymph nodes. (d) Cancer cells the body

a b c

Fig. 65.2 This diagram represents the Fisher hypothesis, also known therapy would not change the disease course. Although the “primary
as the systemic hypothesis of cancer. It states that cancer is a process tumor” is the first one to be visible, (a) there is already a background of
that begins with systemic dissemination. This hypothesis emphasizes tumor deposits that are not yet visible. (b) Metastatic lesions begin to
the importance of systemic therapy to target all the tumors, as local become visible, (c) until there is clear widespread dissemination
65 Radiation Therapy for Extracranial Oligometastatic Disease 683

the background of numerous microscopic lesions that may a b

inevitably present themselves despite systemic therapy.
These cases are not true oligometastases but rather an early
wide dissemination that carries a poorer prognosis with an
unlikely meaningful response to local therapy. Currently,
identifying the oligometastatic state incorporates imaging
and blood biomarkers, as well as follow-up and serial imag-
ing to determine systemic therapy response. This allows for
better identification of those with a less aggressively meta-
static course who may benefit from local therapy. However,
the advent of molecular profiling has shown potential to
expedite the identification of a true oligometastatic state, one
that will benefit tremendously from local consolidative ther-
apy (LCT) [5]. c d

65.2.1 Definitions and Characteristics

of Oligometastatic Disease

As limited metastases can exist without representing a “true”

oligometastatic state, it is plausible that some cancers will
bypass the oligometastatic intermediary. Two plausible expla-
nations for the oligometastatic state exist. The first explanation Fig. 65.3 This figure represents the more contemporary view of
metastasis, including oligometastasis. (a) Tumors may mutate such that
states that the progression of metastases is sequential, and that
they have multiple clonogens that will behave differently, as repre-
the oligometastatic state is simply a stepping stone to poly- sented by purple and lavender. (b) Some clonogens tend to be more
metastatic disease. The other explanation states that oligomet- indolent, and the first evidence of metastasis does not precede wide-
astatic and polymetastatic diseases are distinct processes spread dissemination of true oligometastatic disease. However, this is
indistinguishable from (c) limited metastasis that is in a background of
originating from distinct genetic phenotypes depending on the
microscopic metastases, which will inevitably become (d) widespread.
particular clonal populations involved (see Fig. 65.3) [6]. In true oligometastases, (b) clonogens tend to be limited and more indo-
It is likely that true metastatic progression is also a spec- lent in nature
trum of presentations within the two general explanations.
One could argue that any form of oligometastatic disease
will become widely disseminated if given a time. This could purposes, many major studies have specific cutoffs for what
be due to the inherent indolence of that particular clonogen is considered oligometastatic, with cutoffs varying between
itself, due to further mutations that occur if the lesion is left studies. However, five or less metastatic lesions are the com-
untreated, or selective pressures inducing new mutations due mon cutoff used by most existing major trials, with many
to systemic therapy. As both primary and metastatic lesions showing an overall survival benefit for treating these limited
can further mutate and disseminate additional clonogens, a metastases with focal therapy [2, 9, 10].
metastatic patient may have multiple lesions that will behave It is also important to consider the location of the meta-
differently from one another [7, 8]. It is important to define static disease. As cancer can spread through lymphatics or
the characteristics of a true oligometastatic state and also the vasculature, some organs more readily receive metastatic
variations that could take place. Better classification of the deposits than others. Others, such as malignant pleural effu-
oligometastatic state would help with identification of spe- sions or leptomeningeal disease, may not warrant radiation
cific cases where aggressive local therapy to the primary and/ dose escalation as ablation may not be feasible. For instance,
or metastatic lesion (s) would offer not just palliation but prostate metastases are more likely to spread to pelvic lymph
also benefits for progression-free and overall survival [1]. nodes and bones than visceral organs and very unlikely to
The first consideration is quantity, defined as the number spread to the central nervous system [11]. As it would require
of discrete lesions of metastatic disease. Regardless of size a biology that is more aggressive and has higher predilection
of the lesion, it is presumed that each metastatic lesion began to spread, a single metastasis in a less common location is
as a single deposit of tumor cells. A greater number of dis- more likely to represent underlying dissemination than a
tinct discrete lesions increase the likelihood of existing back- single metastasis in a common or “first-echelon” site. These
ground widespread dissemination, therefore decreasing the rare metastases are therefore less likely to benefit from local-
potential usefulness of localized treatment. For enrollment ized treatment and pertain a poorer survival. For instance,
684 Q. A. Ho and C. C. Hsu

brain metastases occur in less than 2% of prostate cancer The oligometastatic state is not necessarily a primary or
patients and carry a prognosis of about 2.8 months [12]. progressing presentation but can also be induced. Generally,
Furthermore, the site of metastases itself may portend to there are four distinct cohorts of oligometastatic patients.
prognosis based on potential complications of progression, The three better-known groups include (1) patients who
tolerance of treatment, and response to treatment, all of present as oligometastatic, (2) patients who present as non-­
which may be dependent or independent of histology and metastatic but progress to oligometastatic despite locore-
origin. gional treatment (oligorecurrence) after definitive therapy
The volume of tumor burden is also an important charac- and at least 6 month PFS, and (3) oligoprogressive occur-
teristic of oligometastatic disease. Larger lesions (at the time ring during adjuvant therapy or <6 months after definitive
of detection) may represent a more aggressive histology. therapy. The fourth group, oligoresidual, represents an
However, this could be limited to a locally aggressive biol- induced oligometastatic state when there is at least a partial
ogy and may not necessarily represent propensity to spread response in the disease at presentation with induction sys-
distantly. An isolated large tumor deposit could also repre- temic therapy. The first two scenarios represent an indolent
sent a true oligometastatic state, for which this single tumor progression that is more classically associated with an
has had a long period of time to grow without evidence of oligometastatic biology. Oligoprogressive is seemingly
other tumors detected radiographically. Size of lesion is also more aggressive based on a shorter PFS period but can still
important to consider for treatment tolerability and response. carry a reasonable therapeutic window. Oligoresidual, how-
Larger tumors generally have more areas of hypoxia cen- ever, occurs when a limited number of metastases that are
trally due to a poor blood supply [13], are therefore more resistant to a particular systemic therapy remain after the
resistant to radiation, and may not be reached by or respond majority of metastases are effectively treated with that sys-
to chemotherapy. Furthermore, it may be more difficult to temic therapy. Although these lesions are less likely to rep-
resect or irradiate large tumors due to associated toxicities of resent an indolent biology, their limited number allows them
treatment. Rusthoven et al. found that metastatic liver lesions to be addressed similar to truly indolent oligometastatic
smaller than 3 cm had much better 2-year control than lesions lesions [16]. Several studies have included oligometastatic
larger than 3 cm of 100% versus 77% (p = 0.015) [14]. This patients at presentation [3], while others have allowed for
study demonstrated not only the importance of size for con- oligorecurrent or oligoresidual patients also [10, 17].
trol but also consideration of toxicity as more healthy liver Table 65.1 and Fig. 65.4 define and summarize the different
parenchyma would receive spillover radiation when a larger forms of oligometastatic disease.
tumor is treated. In that study, however, treatment with radia-
tion was well tolerated, with only 1 of 47 patients having a
grade 3 toxicity to the chest wall requiring debridement and 65.2.2 Biological Basis of Oligometastases
hyperbaric oxygen and none with grade 4 or 5 toxicity [14].
Timing of presentation of metastatic disease can clini- The fundamental paradigm of tumor growth, invasion, and
cally give us clues as to the likelihood of a patient being metastases has evolved to incorporate numerous molecular
oligometastatic. Metastatic lesions may present as synchro- processes that each contribute to the malignant capability of
nous lesions along with the primary, or as metachronous tumors. Upstream to these molecular processes are geno-
lesions, at least 3 months after presentation of the primary types, resulting in various phenotypes. Each individual
tumor [15]. Metachronous lesions are more likely to repre- tumor cell has its own potential to grow, invade, and metas-
sent a true oligometastatic state, as it would suggest a tasize at its own rate and also to react differently to radiation
slower, step-wise progression of metastasis. This becomes and systemic therapy. The metastatic potential of a tumor is
even more likely if the latency between primary and meta- determined not only by the tumor itself but also by the micro-
static lesions becomes longer. Synchronous presentations, environment that will potentially host a metastatic lesion.
­however, are difficult to determine as it will remain unknown Pancreatic adenocarcinoma and its poor prognosis have
how quickly the existing lesions appeared relative to each been the subject of numerous studies at both the clinical and
other. Studies have shown that metachronous metastatic dis- molecular levels. It has been established that these tumors
ease portends a better prognosis than synchronous meta- contain geographically separate subclones that can present
static disease. Donskov et al. showed that in renal cell years before metastasis. Each subclone will differ by pos-
carcinoma, synchronous metastases carried an overall sur- sessing different mutations in different intermediate steps in
vival of 16.7 months, whereas metachronous metastases metastases [7, 8]. Different studies in the clear cell variant of
carried an OS between 23.8 and 41.7 months after the sec- renal carcinoma and also breast ductal carcinoma have sug-
ond lesion presents, depending on the duration between pre- gested that metastatic lesions are derived from different, and
sentations [15]. often unique, clones of the primary site [18, 19]. In short, it
65 Radiation Therapy for Extracranial Oligometastatic Disease 685

Table 65.1 Summary of the four types of oligometastatic disease, their defining features, and an example clinical scenario of their presentation.
The quantity of lesions in an “oligometastatic” state varies with some studies defining the number as ≤3 but will be defined here as up to five meta-
static lesions. Similarly, oligorecurrent versus oligoprogressive differs in the time course of presentation. Many definitions use “6 months or more”
of progression-free survival to become oligorecurrent as opposed to oligoprogressive. Finally, in some instances, the location of a metastatic lesion
may be considered a higher risk for wide dissemination (e.g., peritoneal carcinomatosis or malignant pleural effusions) and will no longer carry an
oligometastatic label despite a low number of lesions
Type of
oligometastatic
disease Defining features and time course Clinical presentation example
Oligometastatic Up to five metastatic lesions at presentation 67 M presents with prostate adenocarcinoma and three
vertebral body metastases
Oligorecurrent Presents as non-­metastatic. After definitive treatment and a 68F with breast ductal carcinoma who underwent
period of progression-­free survival (e.g., ≥6 months), there is definitive lumpectomy and radiation, found to have a
progression to having up to five metastatic lesions calvarial metastasis 2 years later
Oligoprogressive Presents as non-­metastatic. Up to five metastatic lesions appear 75 M found to have an adrenal metastases 2 months after
after definitive treatment during adjuvant treatment or within a completing chemoradiation and immunotherapy for stage
short period (e.g., ≤6 months) of progression-free survival III NSCLC
Oligoresidual Induced oligometastatic state. Presents as widely metastatic, 62 M presents with innumerable bone and visceral
but disease burden is cytoreduced with systemic therapy to metastases from NSCLC, which is reduced to three bony
having no more than five metastatic lesions lesions after six cycles of systemic therapy

a b

c d

Fig. 65.4 A visual representation of the different subtypes of oligo- free survival (PFS) in the case of oligorecurrence, (c) a non-metastatic
metastatic disease. (a) A case can present as oligometastatic with pri- case may progress during adjuvant treatment or after a short PFS period,
mary tumor and three sites of metastases (diamonds). (b) An initially in the case of oligoprogressive disease. (d) The oligometastatic state
non-metastatic cancer may progress to oligometastasis despite defini- may be derived from a widely metastatic state after induction systemic
tive treatment to the primary, and after a longer period of progression-­ therapy with oligoresidual disease

is the mutation of an original tumor that leads to different ondary sites, extravasation, micrometastases, and coloniza-
subclones within one primary site, with each subclone tion. Micro-RNAs are small RNAs that dictate regulation of
behaving differently and having various levels of metastatic gene expression [6]. These micro-RNAs have been shown to
potential. Each new lesion also has the same potential to be regulated by p53 and myc, leading to either the up- or
develop subclones, and “re-seed” other sites, resulting in fur- downregulation of tumor survival and proliferation, respec-
ther metastatic dissemination. tively [20]. The identification of specific micro-RNAs allow-
The cascade of metastatic development can be traced to ing for the classification of true oligometastatic disease that
the molecular level, with any mutation at each step poten- will benefit from aggressive local therapy is ideal, thus elimi-
tially leading to a bifurcation in behavior and, eventually, nating those with more aggressive phenotypes with shorter
metastatic potential. Individually, these steps assist tumor time to widespread dissemination that would not benefit
cells in invasion, escape, enter circulation, adhesion to sec- from aggressive local therapy.
686 Q. A. Ho and C. C. Hsu

As the original tumor increases its metastatic potential, it duction, oligoprogressive, or oligorecurrent. Those patients
may also create factors that will favor colonization at distant with a more aggressive cancer biology will declare them-
sites. This process is known as the pre-metastatic niche, and selves as polymetastatic and may not have an overall survival
upregulation of fibronectin and clustering of bone marrow-­ or disease-free survival benefit from aggressive local ther-
derived cellular infiltrates have been implicated in creating a apy, thus leading to potential unwarranted treatment-related
pre-metastatic niche, through the development of vascular toxicity from LCT.
endothelial growth factor receptor 1 (VEGFR1) [21]. After identifying a case of oligometastatic disease, the
A study by Lussier et al. evaluated the micro-RNAs of 63 assumption is that the patient is systemically controlled
patients with up to five lung metastases. It found 39 micro-­ except for the limited number of metastatic lesions. Although
RNAs that were expressed at different levels when compar- systemic therapy will continue, either sequentially or con-
ing the patients who had high rates of progression (HRP) and currently with local treatment, there is a window of time for
low rates of progression (LRP). Three of these micro-RNAs which LCT can be utilized on the known disease sites. Early
were downregulated in the LRP group compared to the HRP on, metastectomy, or the surgical resection of a metastatic
group, whereas the remainders were upregulated. These sus- lesion, was the primary “curative” approach for limited met-
pected micro-RNAs were validated using a separate dataset. astatic disease [1, 24–28]. Other local procedures such as
Patients in this dataset who were classified as LRP based on radiofrequency ablation, cryotherapy, and chemoemboliza-
the presence of the appropriate micro-RNAs were found to tion have also shown some benefit, particularly in the unre-
have an improved survival compared to the HRP group [22]. sectable scenario.
An unsupervised clustering analysis of micro-RNA expres- Local radiotherapy in this setting was initially limited to
sion found a divergence of HRP and LRP patients [22]. palliative doses that were much lower than doses used in the
Similarly, a study by Wong et al. found that the expression curative or definitive setting. As technology improved,
levels of three micro-RNAs (miR-23b, miR-449a, and miR-­ higher, ablative doses of radiation that provided local control
449b) predicted survival for the patients with such data avail- comparable to resection began to surface [14, 29]. Stereotactic
able [23]. body radiation therapy (SBRT), also known as stereotactic
ablative radiotherapy (SABR), was developed following ste-
reotactic radiosurgery (SRS) to treat extracranial sites of dis-
65.2.3 Treatment of Oligometastatic Cancer ease [29]. Stereotactic radiation’s purpose, in general, is for
higher-dose radiation to cause tumor death directly, causing
Systemic therapy remains the mainstay and usually the most DNA double-strand breaks, and indirectly through damage
important predictor of outcomes for metastatic patient, to the tumor vasculature as well as potential activation of the
regardless of disease burden. In cases of true oligometastatic immune cytokine cascade [30–32]. Extensive tumor cell
disease, systemic therapy, such as chemotherapy, immuno- injury and death induce a massive release of tumor-specific
therapy, or targeted agents, is often the first treatment. This antigens and pro-inflammatory and pro-oxidant cytokines.
serves two key purposes. The first is for cytoreduction to Furthermore, radiation promotes antigen presentation
reduce the overall quantity of metastatic disease. As cases of through increasing effector T-cell traffic to tumor, major his-
seemingly limited metastatic burden may truly be in a back- tocompatibility complex expression, adhesion molecules,
ground of widely disseminated disease, the most effective costimulatory molecules, heat shock proteins, and death
initial treatment would need to be systemic. This would receptors, as well as promoting antigen presentation through
allow for those patients who would not benefit from systemic activation and maturation of dendritic cells. These multiple
therapy and would progress to “declare” their more aggres- pathways ultimately lead to a downstream effect of a tumor-­
sive disease phenotype. A true oligometastatic state would specific heightened immune response, which may be of par-
often also benefit from systemic therapy as these visible ticular benefit for tumor control when used with immune
tumors would respond to systemic therapy and either remain checkpoint inhibitors [33–35]. These indirect mechanisms
stable or possibly shrink. However, there are instances where may enhance the efficacy of hypofractionated stereotactic
oligoprogressive patients may benefit from aggressive local treatment beyond its direct effects on irreparable DNA dam-
therapy, such as patients with non-small cell lung cancer age to the tumor.
(NSCLC) epidermal growth factor receptor (EGFR) muta- SBRT specifically localizes an extracranial tumor using a
tions with minimally progressive disease; in these instances, stereotactic coordinate system. Higher precision is obtained
the intent of local therapy is to eradicate resistant clonogens through three-dimensional planning, as well as image guid-
and allow for continuation of the same systemic therapy. In ance verification, such as a cone-beam CT, prior to delivery
these scenarios, the response to systemic therapy could be of radiation. Typical SBRT doses involve the delivery of one
used as a method to isolate truly oligometastatic disease, to five fractions of radiation, with dose fractions of 6–30 Gy
either primary oligometastatic, oligoresidual through cytore- per fraction [32]. The high precision targeting allows for a
65 Radiation Therapy for Extracranial Oligometastatic Disease 687

higher ablative dose that improves the likelihood of control The long-term data from one of these studies showed a
while also allowing for smaller margins for treatment setup, median survival for oligometastatic disease of 33 months
decreasing the amount of normal tissue irradiated, and thus [28]. For oligometastatic disease, particularly NSCLC, with
diminishing the chance of serious side effects. Subsequently, a primary site amendable to definitive therapy, the National
as SBRT is much less invasive than other forms of LCT, it Comprehensive Cancer Network (NCCN) guidelines allow
serves to achieve high control with low toxicity. More for either surgical resection or SBRT [38], although SBRT is
recently, NRG-BR001, a phase 1 study [36] looking at SBRT often preferred due to its noninvasive and less morbid
for two to four metastatic lesions from the breast, lung, or approach.
prostate, found that pre-specified doses for seven anatomic
sites could be safely delivered as there were no grade 3 or
higher toxicities. This protocol established the modern safety 65.2.5 Hepatic Metastases
for SBRT of metastatic lesions and is widely used by many
institutions. Although SBRT is often preferred due to its high Treatment of liver metastases from colorectal cancer has
rate of local control and short treatment duration, more frac- been widely studied for decades, given the unique venous
tionated regimens (less radiation per day through a longer drainage of the gastrointestinal organs. Although surgical
treatment duration) have remained useful in some settings, resection has remained a consistent option for oligometa-
particularly when SBRT does not meet toxicity constraints. static liver lesions from a colorectal primary, the use of a
Conventional radiation is defined as 1.8–2 Gy per day in the noninvasive option is safe and efficacious in this setting. A
definitive setting and is common in non-metastatic settings phase 2 trial by Scorsetti et al. [39] found that patients with
but is less common in oligometastatic settings given the inoperable liver metastases receiving 75 Gy in three consec-
long-expected duration of treatment. Hypofractionated radi- utive fractions had a 2-year local control rate of 91%, with-
ation (above 2 Gy per day but less than standard SBRT doses) out grade 3 or higher toxicity. As molecular analyses have
is often considered for larger primary site tumors in the matured, studies have identified molecular subtypes that may
oligometastatic setting, as it is considered safer than SBRT predict the better prognoses that might benefit from
in certain scenarios, but with a shorter duration than conven- metastasis-­direct therapy. Pitroda et al. noted that subtypes
tional radiation, usually 8–15 fractions instead of 25–35 that demonstrate VEGFA amplification along with stromal,
fractions. mesenchymal, and angiogenic signatures or exclusive
Recently, there is increasingly robust data that confirms NOTCH1 and PIK3C2B mutations with E2F/MYC activa-
both progression-free and overall survival benefits from local tion had a particularly poor survival. Their conclusion saw
therapy for oligometastatic cancer. These trials have looked that high-risk molecular profiles had a 19% 10-year OS,
at both treatment of the limited metastatic lesions and treat- whereas low-risk had a 94% survival [5].
ment of the primary and regional sites. Some trials look at
oligometastases from a specific primary site or histology,
whereas others combine several groups together. Table 65.2 65.2.6 Oligometastatic Lung Cancer
highlights the key findings from these studies.
Patients with oligometastatic NSCLC tend to progress at
sites of known disease as opposed to new sites [40].
65.2.4 Adrenal Metastases Subsequently, studies of LCT on oligometastatic NSCLC
have shown the most convincing evidence of this phenome-
Adrenal metastases are a common and well-studied site of non. An earlier single-arm phase 2 trial by Iyengar et al.
metastatic spread, occurring in a large portion of melanomas examined 24 patients with ≤6 extracranial sites of NSCLC
and breast, lung, renal, and gastrointestinal cancers [24, 37]. disease who failed at least first-line systemic therapy and
Adrenalectomy was the original modality of treatment, and a underwent concurrent erlotinib and SBRT to all sites of dis-
meta-analysis of non-small cell lung cancer (NSCLC) ease. In total 52 sites were treated, with improved median
patients with isolated adrenal metastases treated with adre- PFS (14.7 months) and median OS (20.4 months) compared
nalectomy showed a 5-year OS of 25% for metachronous to historically reported figures for second-line treatment of
tumors and 26% for synchronous tumors; however, there median PFS of 2–4 months and median OS of 9 months [17].
were associated complications of adrenalectomy, including None of the tested patients were positive for EGFR muta-
infections, cardiovascular events, ileus, and treatment-related tions; the majority of patients progressed in new distant sites,
mortality [25, 26]. Subsequently, several studies showed that and treatment was well tolerated with only 2 of 24 patients
SBRT to adrenal metastases is well tolerated, with two stud- having grade 3 toxicity [17]. This single-arm study sug-
ies showing no grade 3 or higher adverse effects [27, 28] and gested the benefit of local therapy in potentially changing a
another showing no grade 2 or higher adverse effects [24]. patient’s disease course. A subsequent phase 2 randomized
688 Q. A. Ho and C. C. Hsu

Table 65.2 Summary of prospective trials of oligometastatic cancer (unless otherwise stated) treated with LCT and their respective outcomes. All
trials had a majority of stage IV patients. Three trials were single-arm. Toxicity data (if available) listed at the end of each outcome section, as
represented by grading
Author of Local
prospective Site of Sample consolidative
trial primary Study design size Enrollment criteria Treatment Control arm therapy arm
Scorsetti [39] Colorectal Single-arm 42 Up to three liver 75 Gy/3 fx to liver N/A 1−/2−/3-year
phase 2 of liver patients; lesions. No lesion LC: 95/91/85%
RT 52 lesions extra-hepatic Median PFS:
progressive or 12 months
untreated gross Median OS:
disease 29 months
Acute G2/G3
toxicity: 78%/0%
Iyengar [17] NSCLC Single-arm 24 Up to six sites of 14–24 Gy/1 fx, N/A Median PFS:
phase 2 of patients; extracranial 22–33 Gy/3 fx, 14.7 months
SBRT and 52 lesions disease, early 30–40 Gy/5 fx Median OS:
erlotinib systemic 20.4 months G3
concurrently chemotherapy toxicity: 8%
failure
Iyengar [2] NSCLC Randomized 29 Up to six 16–27 Gy/1 fx, Median PFS: Median PFS:
non-EGFR/ phase 2 of patients; extracranial sites 24.5–33 Gy/3 fx, 3.5 months 9.7 months
ALK-­ maintenance 31 lesions including primary, 28–37.5 Gy/5 fx; Median OS (with (P = 0.01)
targetable chemotherapy in 14 no more than 3 in 45 Gy/15 fx (if SBRT crossover): Median OS: Not
+/− SBRT patients of the liver or lung, constraints not met) 17 months reached (p-value
treatment did not progress on G3/4/5 toxicity: not reported)
arm systemic therapy 2/1/0 G3/4/5 toxicity:
4/0/0
Gomez [3] NSCLC Randomized 49 patients ≤3 extra- and Resection or radiation PFS: 4.4 months PFS:
phase 2 of intracranial (SBRT, 15 fractions, OS: 17 months 14.2 months
consolidative metastatic sites or conventional Survival after (P = 0.022)
RT for NSCLC after systemic allowed) to metastatic progression: OS: 41.2 months
therapy, did not sites; chemoradiation 9.4 months (P = 0.017)
progress on allowed for primary Median time to Survival after
first-line systemic and regional site(s) new lesion: progression:
therapy 6 months 37.6 months
G3/4/5 toxicity: Median time to
2/0/0 new lesion:
14.2 months
G3/4/5 toxicity:
4/0/0
Wang [45] EGFR-­ Single-arm 26 patients 0–3 extra- and Conformal RT, IMRT, N/A Median TTP:
mutated study (21 with intracranial or gamma ray SBRT 6.3 months
stage stage IV) metastatic sites to thoracic disease Median PFS:
III–IV concurrently with TKI 10.2 months
NSCLC Median OS:
21.8 months
G3/4/5 toxicity:
12/3/0
RTOG 0937 ES-SCLC Randomized 86 patients 1–4 extracranial PCI (25 Gy/10 fx) 1-year OS: 60.1% 1-year OS:
[46] phase 2 of PCI lesions, had at least with randomization to Median OS: 50.8% (P = 0.21)
+/− partial response to consolidative RT to 15.8 months Median OS:
consolidative four to six cycles of the thorax and Median PFS: 13.8 months
extracranial RT platinum-based metastases 2.9 months Median PFS:
chemotherapy to (45 Gy/15 fx) or no G3+ toxicity: 4.9 months
one site, with no further treatment 23.8% (P = 0.0148)
progression at other G3+ toxicity:
sites 36.4%
(one person had
G5 pneumonitis)
65 Radiation Therapy for Extracranial Oligometastatic Disease 689

Table 65.2 (continued)

Author of Local
prospective Site of Sample consolidative
trial primary Study design size Enrollment criteria Treatment Control arm therapy arm
HORRAD Prostate Randomized 432 PSA >20 ng/ml Androgen deprivation Median OS: Median OS:
[49] study of ADT patients with no restrictions therapy randomized to 43 months median 45 months
+/− prostate RT on quantity of local radiotherapy to time to PSA (P = 0.4)
metastasis the prostate progression: Median time to
15 months PSA progression:
12 months
(P = 0.02)
Parker – Prostate Randomized 2061 Low and high Androgen deprivation PFS (all): PFS (all):
STAMPEDE phase 3 of patients disease burden per therapy and docetaxel 32.4 months 33.1 months
[4] systemic CHAARTED randomized to local FFS (all): (p = 0.468)
therapy definition radiotherapy to the 21.4 months FFS (all):
+/− prostate RT prostate 55 Gy/20 fx OS (all): 26.2 months
or 36 Gy/6 weekly fx 41.6 months (p < 0.0001)
PFS (low): OS (all):
39.4 months 42.5 months
FFS (low): (p = 0.266)
27.4 months PFS (low):
OS (low): 42.9 months
45.4 months (p = 0.033)
G3+ toxicity FFS (low):
6/12/24 months: 36.1 months
21/12/15% (p < 0.0001)
OS (low):
49.1 months
(p = 0.007)
G3+ toxicity
6/12/24 month:
22/13/13%
Ost [50] Prostate Phase 2 62 patients Three or fewer Not on systemic Median ADT-free Median ADT-free
randomized extracranial therapy, randomized survival: survival:
trial of metastases, to metastasis-directed 13 months 21 months
surveillance testosterone over therapy (SBRT Median PSA (P = 0.11)
versus MDT 50 ng/mL 30 Gy/3 fx or surgery) progression: Median PSA
6 months progression:
10 months
(P = 0.03)
G1/2+ toxicity:
17%/0%
Phillips – Prostate Phase 2 54 patients Recurrent 19.5–48 Gy/3–5 fx 6-month 6-month
ORIOLE randomized hormone-sensitive, SBRT progression: 61% progression: 19%
[51] trial of 1–3 metastatic 6-month new (P = 0.005)
observation sites, no ADT in lesions: 63% 6-month new
versus MDT last 6 months and G3+ toxicity: 0 lesions: 16%
SBRT less than 3 years (P = 0.006)
total G3+ toxicity: 0
Trovo [9] Breast Single-arm 54 patients Up to five SBRT 30–45 Gy/3 fx N/A 1−/2-year PFS:
phase 2 trial of with 92 extracranial lesions or IMRT 60 Gy/25 fx; 75/53%
MDT with lesions 89% received 2-year LC: 97%
primary concomitant systemic 2-year OS: 95%
controlled therapy G2/3+ toxicity:
2/0
(continued)
690 Q. A. Ho and C. C. Hsu

Table 65.2 (continued)

Author of Local
prospective Site of Sample consolidative
trial primary Study design size Enrollment criteria Treatment Control arm therapy arm
Palma – Breast, Phase 2 99 patients Up to five lesions, Control: 8 Gy/1 fx, Progression Progression
SABR-­ colorectal, randomized excluding one to 20 Gy/5 fx, events: 85% events: 59%
COMET [10, lung, trial of SBRT three intracranial-­ 30 Gy/10 fx Median PFS: Median PFS:
32] prostate, MDT only metastatic SBRT: 5.4 months 11.6 months
others patients, maximum 16–60 Gy/1–8 fx Median OS: (P = 0.001)
three lesions per depending on site as 28 months Median OS:
organ listed in protocol (see G2+ toxicity: 9% 50 months
citation supplement) Treatment-related (P = 0.006)
deaths: 0 G2+ toxicity:
29%
Treatment-­
related deaths:
4.5%
PFS progression-free survival, OS overall survival, LC local control, FFS failure-free survival, TTP time to progression

control trial (RCT) by Iyengar et al. randomized NSCLC retrospective review by Hu et al., patients with oligometa-
patients who did not possess EGFR- or ALK-targetable static EGFR-mutated NSCLC receiving TKI followed by
mutations with five or fewer metastatic sites, and who did LCT were found to have a significantly longer PFS of
not progress on systemic therapy, to SBRT to all sites of 15 months, compared to 10 months of TKI alone (P < 0.001).
gross disease followed by maintenance chemotherapy or OS was also longer in the combination group, at 34 months,
maintenance chemotherapy alone. Although only 29 patients compared to 21 months (P = 0.001) [43]. A different
were enrolled, this was due to the accrual terminating early propensity-­matched study by Yen et al. looked at EGFR-­
due to a significant improvement in outcomes during the mutated NSCLC for stages III–IV who responded to TKI
interim analysis which found PFS improving from 3.5 months and found that patients who received thoracic RT had an
to 9.7 months when adding SBRT (p = 0.01) [2]. A different adjusted hazard ratio for mortality of 0.72 for all patients
phase 2 RCT by Gomez et al. randomized patients with three (P = 0.0005), 0.76 for stage IIIB (P = 0.042), and 0.70 for
or fewer sites of metastatic disease, without progression after stage IV (P = 0.005) [44]. A single-arm prospective study
at least 3 months of first-line systemic therapy, to LCT ver- was reported by Wang et al., for which TKI and individual-
sus maintenance systemic therapy or observation. LCT was ized RT in the stage III–IV setting were well-tolerated, with
defined as ablation of all residual disease, with either surgery no mortalities and less than 10% grade 4 toxicity [45]. These
or radiation, with concurrent chemoradiation also allowed data strongly suggest that LCT/SBRT to oligometastatic dis-
for primary disease. Patients treated with LCT were permit- ease in the setting of NSCLC patients with driver mutations
ted to continue maintenance therapy or observed. Patients after response to targeted agents may result in improvements
receiving LCT had both a significant progression-free sur- in survival without significant toxicity.
vival benefit of median 14.2 months compared to 4.4 months However, data in the small-cell lung cancer (SCLC) has
(P = 0.022) and an overall survival benefit of median not quite developed to the same extent as its NSCLC coun-
41.2 months versus 17 months (P = 0.017) [3]. NRG-LU002 terpart. RTOG 0937 was a phase 2 RCT of patients with
is currently recruiting patients in a phase 2/3, national, multi-­ extensive-stage (ES) SCLC and one to four extracranial
institutional setting to determine if LCT does provide sur- metastases, with evidence of response to minimum of one
vival benefit beyond maintenance therapy alone for site, without evidence of progression to any site, comparing
oligometastatic/oligoresidual patients with up to three extra- prophylactic cranial radiation (PCI) alone versus PCI with
cranial metastatic sites, regardless of disease burden prior to consolidative radiation to residual intrathoracic and extracra-
systemic therapy [41]. nial metastatic sites. There was a 3- and 12- month
The above studies typically exclude EGFR-mutated progression-­free survival favoring PCI with LCT of 85.5%
NSCLC as it represents a different biology and subsequently and 25% versus 46.7% and 20.4% (P = 0.01), respectively.
a better prognosis. The effectiveness of tyrosine kinase However, the 1-year OS was not different, with 60.1% for
inhibitors (TKIs) has imparted a long natural history for PCI and 50.8% for PCI + LCT (P = 0.21) [46]. The lack of
these patients, and the majority of patients on erlotinib tend an OS benefit could be due to the limited options of systemic
to fail at the initial sites of disease (primary and original therapy. The very recent addition of atezolizumab to the
metastases) as opposed to new sites of metastases [42]. In a standard platinum-based doublet therapy has shown
65 Radiation Therapy for Extracranial Oligometastatic Disease 691

promising results, including an OS benefit [47]. This has ADT-free survival of 21 months versus 13 months in patients
prompted the NRG-LU007 RAPTOR study, which seeks to without MDT (P = 0.11) [50]. Although not statistically sig-
test the addition of LCT to atezolizumab and chemotherapy nificant, the sample size was only 62. The phase 2 ORIOLE
in ES-SCLC with ten or fewer sites of metastatic disease trial looked specifically at patients with castration-sensitive
after induction therapy [48]. prostate cancer with one to three asymptomatic metastases
up to 5 cm in diameter or 250 cm2. Patients were randomized
2:1 to SBRT or observation to the sites of metastases. Patients
65.2.7 Oligometastatic Prostate Cancer had previous definitive surgery or radiation to the prostate.
There was a benefit to PFS at 6 months (19% vs. 61%,
There is strong evidence in favor of treating the primary and P = 0.005), decrease risk of new lesion at 6 months, no grade
distant sites in oligometastatic prostate cancer. However, 3 or greater toxic effects, and an increase in clonotypic T-cell
careful patient selection remains the first step. Studies exam- expansion [51]. These data suggest that MDT for oligopro-
ining treatment of the primary site alone with RT include the gressive prostate cancer after prior definitive primary therapy
HORRAD Trial and STAMPEDE. The HORRAD Trial may improve PFS, and further studies are needed to deter-
examined men with a prostate-specific antigen (PSA) of over mine potential impacts on OS.
20 ng/ml and primary bone metastases and randomized
between androgen deprivation therapy (ADT) alone and
ADT with primary treatment of the prostate with external 65.2.8 Oligometastatic Breast Cancer
beam radiation (70 Gy/35 fx daily or 57.76 Gy/19 fx three
times weekly). It found that there was a median time to PSA Breast cancer is another malignancy for which an oligometa-
progression benefit of 15 versus 12 months (P = 0.02), but no static state can carry a favorable prognosis. A phase 2 study
survival benefit in the radiotherapy arm. However, 67% of by Trovo et al. that enrolled patients with up to five extracra-
patients had at least five metastatic lesions, and the median nial metastases (the primary tumor controlled) found that
PSA level was 142 ng/ml. Furthermore, on unplanned sub- treating metastatic lesions with SBRT conferred a 1- and
analysis, in men with less than five metastatic lesions, there 2-year PFS of 75% and 53%, respectively. The 2-year OS
was a trend toward a significant improvement in OS (HR, was 95% [9]. Following the initial safety trial BR-001, the
0.68; 95% CI, 0.42–1.10; P < 0.10) [49]. STAMPEDE, a NRG has been enrolling patients for BR-002, a randomized
larger multi-arm, multistage trial comparing EBRT phase 2/3 trial for women with up to four sites of metastatic
(55 Gy/20 fx daily or 36 Gy/6 fx weekly SBRT) to the pros- breast cancer, who received up to 12 months of first-line sys-
tate with ADT versus ADT alone, also found a failure-free temic therapy without progression; these patients are ran-
survival benefit (17 vs. 13 months, P = 0.0001) but not an domized to SBRT and/or surgical resection versus no LCT
overall survival benefit for all comers (48 vs. 46 months, [52]. There is also evidence for surgical resection of the pri-
P = 0.266). However, the trial stratified patients based on mary breast cancer even in the metastatic setting. An NCDB
metastatic burden for a preplanned analysis by using the study by Khan et al. found that margin-negative surgical
CHAARTED definition of high metastatic burden of four or resection with partial or total mastectomy had a superior
more bone metastases with one or more outside the pelvis or prognosis, with HR of 0.61 (95% CI, 0.58–0.65, P < 0.05),
vertebral body, visceral metastases, or both. Patients who did compared to patients not surgically resected in the metastatic
not meet these criteria were deemed as “low metastatic bur- setting [53]. A meta-analysis by Ly et al. found that patients
den” and had both a 3-year failure-free survival (HR, 0.59; with metastatic breast cancer who received local-regional
95% CI, 0.49–0.72; P < 0.0001) and a 3-year OS benefit therapy (surgery with or without radiation) had an overall
(HR, 0.68; 95% CI, 0.52–0.90; P = 0.007). The study also median survival of 25–42 months compared to 12.6–
pointed out the safety of local therapy, with only 1% more 28.3 months in those without surgery [54].
grade 3+ toxicity at 6 months and 2% less at 2 years com- Surgery has remained the key oncologic treatment of the
pared to systemic therapy alone, with an overall grade 3–4 primary site for breast cancer, and radiation has mainly served
toxicity rate of 4–5% [4]. These data suggest that treatment in an adjuvant role. However, with the success of SBRT, fur-
of the primary in the setting of metastatic prostate cancer ther technological advances have allowed for the develop-
with low metastatic burden may improve both PFS and OS. ment of GammaPod, which uses stereotactic image guidance,
Regarding metastasis-directed therapy (MDT) for pros- along with vacuum-assisted breast immobilization to deliver
tate cancer, there are also a few trials to show benefit. A high-intensity radiation using Cobalt-60 sources [55]. There
phase 2 RCT by Ost et al. found that patients previously is a potential for GammaPod radiotherapy to help with treat-
treated for localized disease who then present with three or ment to the primary site, which could prove useful in the
fewer extracranial metastases benefited from MDT with an metastatic setting, though there is minimal published data.
692 Q. A. Ho and C. C. Hsu

65.2.9 O
 ligometastatic Cancers with NRG-LU002. Additionally, there is a push to see how many
Controlled Primary sites of extracranial metastases should be treated for a PFS
and OS benefit. In the intracranial metastases setting, there is
The phase 2 SABR-COMET study by Palma et al. looked at evidence that treating up to ten intracranial lesions ablatively
patients with a controlled primary malignancy and one to five with stereotactic radiosurgery carries a benefit similar to
extracranial metastatic lesions. The majority of primary treating two to four lesions with SRS [60]. Ongoing trials
tumors were breast, colorectal, lung, and prostate. Patients such as SABR-COMET-10 are also looking at focally treat-
were randomized to either SBRT to all sites of metastatic dis- ing up to ten extracranial lesions, to see if there is a benefit
ease or palliative standard of care, including palliative doses similar to treating up to five extracranial lesions [56].
of radiation at symptomatic sites. The SBRT dosing was When compared to surgical metastasectomy, the toxicity
lesion- and site-specific per protocol and ranged from profile of radiation therapy is low; however, it is not without
16 Gy/1 fx for vertebral bodies to 60 Gy/8 fx for adrenal, risk. As discussed, serious side effects, including treatment-­
liver, or central lung lesions. Systemic therapy was at the dis- related mortality, can and have occurred, and careful plan-
cretion of the medical oncologist but had to be outside of ning at an experienced facility is always recommended.
SABR by 4 weeks, unless it was hormonal therapy. Median Newer technology brings a promise of safer delivery.
PFS was improved in the SBRT arm, at 11.6 months, com- Adaptive planning, through the use of fast algorithms and
pared to 5.4 months in the control arm (P = 0.001). Overall using magnetic resonance guidance, is particularly helpful in
survival was improved in the SBRT arm, with median sur- areas where the tumor shrinks or nearby normal organs at
vival of 50 months, compared to 28 months in the standard-­ risk may not be stationary. This new form of MR-guided
of-­care arm (P = 0.006) with 5-year OS of 42.3% versus radiation has been shown to be deliverable and safe [61].
17.7%, respectively. Grade 2 or higher adverse effects were Although the data for heavy particle beam radiotherapy has
29% in the SBRT arm, compared to 9% in the control arm. had mixed result, there remains optimism that less “exit
Furthermore, 4.5% of the deaths in the SBRT group were dose”, which is the dose of radiation past the tumor that exits
“possibly, probably, or definitely” related to treatment [10, the body, could lead to an improved side effect profile and
32]. Based on these phase 2 results, there are separate phase 3 potential dose escalation [62].
trials examining these questions among patients with extra- As systemic therapies improve, so do the number of cases
cranial metastatic sites numbering 1–3 (SABR-COMET-3) of oligometastatic/oligoresidual cancer, increasing the role
and those with four to ten sites (SABR-COMET-10) [56, 57]. of radiation therapy to extend the usefulness of the systemic
therapy, and subsequently improve OS. The advent of molec-
ular profiling will help improve our understanding of the
65.3 Synergy with Systemic Therapy biology of cancer and its response to radiation therapy, and
and Future Directions lead to the creation of more targeted agents. LCT, especially
radiation therapy, will continue to improve in efficacy and
Although mechanisms remain unclear, there is a potential safety, furthering the possibility of long-term survival (and
synergy between systemic therapy and radiation, as well as ideally cure in a select few) for oligometastatic cancer.
the potential abscopal effect of radiation. Radiation therapy
can create an abscopal effect, for which not only does the
radiation-targeted lesion regress but so do lesions at distant 65.4 Conclusion and Perspective
sites. It is hypothesized as a systemic immune-mediated
response for which high-dose radiation creates an immuno- • Oligometastatic cancer represents an intermediary
logical response that is further upregulated by systemic ther- between non-metastatic and widely metastatic cancer and
apy through antigen processing, as well as generation and may have improved progression-free and overall survival
trafficking of effector cells [58, 59]. Though enticing, the with aggressive metastasis-directed therapy.
exact mechanisms of the abscopal effect are not well-­ • Molecular analysis has given us clues on how cancer pro-
characterized, and reproducibility of this response has gresses to the oligometastatic state and may help us deter-
remained elusive. A greater understanding of these mecha- mine which patients benefit from metastasis-directed
nisms could enhance our delivery of treatment to further therapy.
improve outcomes. • Modern randomized clinical trials have largely shown a
As there is mounting phase 2 evidence of benefit for LCT benefit to primary/metastasis-directed therapy for oligo-
of up to five sites of extracranial metastatic disease, more metastatic cancer, and as both LCT and systemic thera-
mature phase 2/3 studies are still needed, including pies improve, so will their synergy to benefit survival.
NRG-BR002 for oligometastatic breast cancer as well as
65 Radiation Therapy for Extracranial Oligometastatic Disease 693

radiation therapy for oligometastatic breast cancer: results of a pro-

Open Questions spective phase II trial. Radiother Oncol. 2018;126:177–80.
10. Palma DA, Olson R, Harrow S, Gaede S, Louie AV, Haasbeek C,
• How do we identify and appropriately counsel Mulroy L, Lock M, Rodrigues GB, Yaremko BP, Schellenberg
patients who are at higher risk for SBRT toxicity, D. Stereotactic ablative radiotherapy versus standard of care
including treatment-related mortality? palliative treatment in patients with oligometastatic cancers
­
• Is there a combined threshold of tumor burden, (SABR-­COMET): a randomised, phase 2, open-label trial. Lancet.
2019;393:2051–8.
quantity, and/or location for which SBRT for oligo- 11. Zhao F, Wang J, Chen M, Chen D, Ye S, Li X, Chen X, Ren G,
metastatic disease would unlikely improve survival Yan S. Sites of synchronous distant metastases and prognosis in
(or even be detrimental)? prostate cancer patients with bone metastases at initial diagnosis:
• What would be the most appropriate treatment for a population-­ based study of 16,643 patients. Clin Transl Med.
2019;8:30.
polymetastatic disease with limited progressive 12. Hatzoglou V, Patel GV, Morris MJ, Curtis K, Zhang Z, Shi W,
disease? Huse J, Rosenblum M, Holodny AI, Young RJ. Brain metastases
• What would be the optimal sequencing of systemic from prostate cancer: an 11-year analysis in the MRI era with
therapy and local consolidative therapy based on the emphasis on imaging characteristics, incidence, and prognosis. J
Neuroimaging. 2014;24:161–6.
type of oligometastatic disease, burden, and response? 13. Wouters BG, Koritzinsky M, Brown JM, Van der Kogel AJ. The
• How can we use biomarkers to predict the quantity tumour microenvironment and cellular hypoxia responses. In:
of unidentified microscopic systemic disease and/or Joiner MC, Van der Kogel A, editors. Basic clinical radiobiology.
assess response to treatment of oligometastatic 5th ed. Boca Raton, FL: CRC Press; 2009. p. 206–16.
14. Rusthoven KE, Kavanagh BD, Cardenes H, Stieber VW, Burri SH,
disease? Feigenberg SJ, Chidel MA, Pugh TJ, Franklin W, Kane M, Gaspar
LE. Multi-institutional phase I/II trial of stereotactic body radiation
therapy for liver metastases. J Clin Oncol. 2009;27:1572–8.
15. Donskov F, Xie W, Overby A, Wells JC, Fraccon AP, Sacco CS,
Porta C, Stukalin I, Lee JL, Koutsoukos K, Yuasa T. Synchronous
References Versus Metachronous Metastatic Disease: Impact of Time to
Metastasis on Patient Outcome—Results From the International
1. Hellman S, Weichselbaum RR. Oligometastases. J Clin Oncol. Metastatic Renal Cell Carcinoma Database Consortium. European
1995;13:8–10. urology oncology. 2020;3(4):530–9.
2. Iyengar P, Wardak Z, Gerber DE, Tumati V, Ahn C, Hughes RS, 16. Niibe Y, Hayakawa K. Oligometastases and oligo-recurrence: the
Dowell JE, Cheedella N, Nedzi L, Westover KD, Pulipparacharuvil new era of cancer therapy. Jpn J Clin Oncol. 2010;40:107–11.
S. Consolidative radiotherapy for limited metastatic non-small-cell 17. Iyengar P, Kavanagh BD, Wardak Z, Smith I, Ahn C, Gerber
lung cancer: a phase 2 randomized clinical trial. JAMA Oncol. DE, Dowell J, Hughes R, Abdulrahman R, Camidge DR, Gaspar
2018 Jan 1;4:e173501. LE. Phase II trial of stereotactic body radiation therapy combined
3. Gomez DR, Tang C, Zhang J, Blumenschein GR Jr, Hernandez with erlotinib for patients with limited but progressive metastatic
M, Lee JJ, Ye R, Palma DA, Louie AV, Camidge DR, Doebele non–small-cell lung cancer. J Clin Oncol. 2014;32:3824–30.
RC. Local consolidative therapy vs. maintenance therapy or obser- 18. Gerlinger M, Rowan AJ, Horswell S, Larkin J, Endesfelder D,
vation for patients with oligometastatic non–small-cell lung can- Gronroos E, Martinez P, Matthews N, Stewart A, Tarpey P, Varela
cer: long-term results of a multi-institutional, phase II, randomized I. Intratumor heterogeneity and branched evolution revealed by
study. J Clin Oncol. 2019;37:1558. multiregion sequencing. N Engl J Med. 2012;366:883–92.
4. Parker CC, James ND, Brawley CD, Clarke NW, Hoyle AP, Ali 19. Klein CA. Selection and adaptation during metastatic cancer pro-
A, Ritchie AW, Attard G, Chowdhury S, Cross W, Dearnaley gression. Nature. 2013;501:365–72.
DP. Radiotherapy to the primary tumour for newly diagnosed, 20. Lujambio A, Lowe SW. The microcosmos of cancer. Nature.
metastatic prostate cancer (STAMPEDE): a randomised controlled 2012;482:347–55.
phase 3 trial. Lancet. 2018;392:2353–66. 21. Kaplan RN, Rafii S, Lyden D. Preparing the “soil”: the premeta-
5. Pitroda SP, Khodarev NN, Huang L, Uppal A, Wightman SC, Ganai static niche. Cancer Res. 2006;66:11089–93.
S, Joseph N, Pitt J, Brown M, Forde M, Mangold K. Integrated 22. Lussier YA, Xing HR, Salama JK, Khodarev NN, Huang Y,
molecular subtyping defines a curable oligometastatic state in Zhang Q, Khan SA, Yang X, Hasselle MD, Darga TE, Malik
colorectal liver metastasis. Nat Commun. 2018;9:1–8. R. MicroRNA expression characterizes oligometastasis (es). PLoS
6. Uppal A, Ferguson MK, Posner MC, Hellman S, Khodarev NN, One. 2011;6:e28650.
Weichselbaum RR. Towards a molecular basis of oligometastatic 23. Wong AC, Watson SP, Pitroda SP, Son CH, Das LC, Stack ME,
disease: potential role of micro-RNAs. Clin Exp Metastasis. Uppal A, Oshima G, Khodarev NN, Salama JK, Weichselbaum
2014;31:735–48. RR. Clinical and molecular markers of long-term survival after
7. Campbell PJ, Yachida S, Mudie LJ, Stephens PJ, Pleasance ED, oligometastasis-directed stereotactic body radiotherapy (SBRT).
Stebbings LA, Morsberger LA, Latimer C, McLaren S, Lin ML, Cancer. 2016;122:2242–50.
McBride DJ. The patterns and dynamics of genomic instability in 24. Chawla S, Chen Y, Katz AW, Muhs AG, Philip A, Okunieff P,
metastatic pancreatic cancer. Nature. 2010;467:1109–13. Milano MT. Stereotactic body radiotherapy for treatment of adrenal
8. Yachida S, Jones S, Bozic I, Antal T, Leary R, Fu B, Kamiyama metastases. Int J Radiat Oncol Biol Phys. 2009;75:71–5.
M, Hruban RH, Eshleman JR, Nowak MA, Velculescu VE. Distant 25. Lo CY, Van Heerden JA, Soreide JA, Grant CS, Thompson GB,
metastasis occurs late during the genetic evolution of pancreatic Lloyd RV, Harmsen WS. Adrenalectomy for metastatic disease to
cancer. Nature. 2010;467:1114–7. the adrenal glands. Br J Surg. 1996;83:528–31.
9. Trovo M, Furlan C, Polesel J, Fiorica F, Arcangeli S, Giaj-Levra N, 26. Tanvetyanon T, Robinson LA, Schell MJ, Strong VE, Kapoor R,
Alongi F, Del Conte A, Militello L, Muraro E, Martorelli D. Radical Coit DG, Bepler G. Outcomes of adrenalectomy for isolated syn-
694 Q. A. Ho and C. C. Hsu

chronous versus metachronous adrenal metastases in non-small-cell metastatic lung cancer? A patterns-of-failure analysis. Acta Oncol.
lung cancer: a systematic review and pooled analysis. In: Database 2009;48:578–83.
of abstracts of reviews of effects (DARE): quality-assessed reviews. 41. https://clinicaltrials.gov/ct2/show/NCT03137771
England: Centre for Reviews and Dissemination; 2008. 42. Patel SH, Rimner A, Foster A, Zhang Z, Woo KM, Helena AY,
27. Rudra S, Malik R, Ranck MC, Farrey K, Golden DW, Hasselle MD, Riely GJ, Wu AJ. Patterns of initial and intracranial failure in meta-
Weichselbaum RR, Salama JK. Stereotactic body radiation therapy static EGFR-mutant non-small cell lung cancer treated with erlo-
for curative treatment of adrenal metastases. Technol Cancer Res tinib. Lung Cancer. 2017;108:109–14.
Treat. 2013 Jun;12:217–24. 43. Hu F, Xu J, Zhang B, Li C, Nie W, Gu P, Hu P, Wang H, Zhang
28. Buergy D, Rabe L, Siebenlist K, Stieler F, Fleckenstein J, Giordano Y, Shen Y, Wang S. Efficacy of local consolidative therapy for
FA, Wenz F, Boda-Heggemann J. Treatment of adrenal metastases oligometastatic lung adenocarcinoma patients harboring epi-
with conventional or hypofractionated image-guided radiation ther- dermal growth factor receptor mutations. Clin Lung Cancer.
apy–patterns and outcomes. Anticancer Res. 2018;38:4789–96. 2019;20:e81–90.
29. Onishi H, Araki T, Shirato H, Nagata Y, Hiraoka M, Gomi 44. Yen YC, Hsu HL, Chang JH, Lin WC, Chang YC, Chang CL, Chow
K, Yamashita T, Niibe Y, Karasawa K, Hayakawa K, Takai JM, Yuan KS, Wu AT, Wu SY. Efficacy of thoracic radiotherapy
Y. Stereotactic hypofractionated high-dose irradiation for stage I in patients with stage IIIB–IV epidermal growth factor receptor-­
nonsmall cell lung carcinoma: clinical outcomes in 245 subjects in mutant lung adenocarcinomas who received and responded to
a Japanese multiinstitutional study. Cancer. 2004;101:1623–31. tyrosine kinase inhibitor treatment. Radiother Oncol. 2018;129:
30. Churilla TM, Chowdhury IH, Handorf E, Collette L, Collette S, 52–60.
Dong Y, Alexander BM, Kocher M, Soffietti R, Claus EB, Weiss 45. Wang J, Xia TY, Wang YJ, Li HQ, Li P, Wang JD, Chang DS, Liu
SE. Comparison of local control of brain metastases with stereo- LY, Di YP, Wang X, Wu WZ. Prospective study of epidermal growth
tactic radiosurgery vs surgical resection: a secondary analysis of a factor receptor tyrosine kinase inhibitors concurrent with individ-
randomized clinical trial. JAMA Oncol. 2019;5:243–7. ualized radiotherapy for patients with locally advanced or meta-
31. Kim MS, Kim W, Park IH, Kim HJ, Lee E, Jung JH, Cho LC, static non–small-cell lung cancer. Int J Radiat Oncol Biol Phys.
Song CW. Radiobiological mechanisms of stereotactic body radia- 2011;81:e59–65.
tion therapy and stereotactic radiation surgery. Radiat Oncol J. 46. Gore EM, Hu C, Sun AY, Grimm DF, Ramalingam SS, Dunlap
2015;33:265. NE, Higgins KA, Werner-Wasik M, Allen AM, Iyengar P, Videtic
32. Palma DA, Olson R, Harrow S, Gaede S, Louie AV, Haasbeek C, GM. Randomized phase II study comparing prophylactic cranial
Mulroy L, Lock M, Rodrigues GB, Yaremko BP, Schellenberg D. irradiation alone to prophylactic cranial irradiation and consolida-
Stereotactic ablative radiotherapy for the comprehensive treatment tive extracranial irradiation for extensive-disease small cell lung
of oligometastatic cancers: long-term results of the SABR-COMET cancer (ED SCLC): NRG oncology RTOG 0937. J Thorac Oncol.
phase II randomized trial. J Clin Oncol. 2020:1;38(25):2830. 2017;12:1561–70.
33. Finkelstein SE, Timmerman R, McBride WH, Schaue D, Hoffe SE, 47. Horn L, Mansfield AS, Szczęsna A, Havel L, Krzakowski M,
Mantz CA, Wilson GD. The confluence of stereotactic ablative radio- Hochmair MJ, Huemer F, Losonczy G, Johnson ML, Nishio M,
therapy and tumor immunology. Clin Dev Immunol. 2011;2011:7. Reck M. First-line atezolizumab plus chemotherapy in extensive-­
Article ID 439752. https://doi.org/10.1155/2011/439752. stage small-cell lung cancer. N Engl J Med. 2018;379:2220–9.
34. Blank C, Gajewski TF, Mackensen A. Interaction of PD-L1 on 48. https://clinicaltrials.gov/ct2/show/NCT04402788
tumor cells with PD-1 on tumor-specific T cells as a mechanism of 49. Boevé LM, Hulshof MC, Vis AN, Zwinderman AH, Twisk JW,
immune evasion: implications for tumor immunotherapy. Cancer Witjes WP, Delaere KP, van Moorselaar RJ, Verhagen PC, van
Immunol Immunother. 2005;54:307–14. Andel G. Effect on survival of androgen deprivation therapy alone
35. Seung SK, Curti BD, Crittenden M, Walker E, Coffey T, Siebert compared to androgen deprivation therapy combined with concur-
JC, Miller W, Payne R, Glenn L, Bageac A, Urba WJ. Phase 1 study rent radiation therapy to the prostate in patients with primary bone
of stereotactic body radiotherapy and interleukin-2—tumor and metastatic prostate cancer in a prospective randomised clinical trial:
immunological responses. Sci Transl Med. 2012;4:137ra74. data from the HORRAD trial. Eur Urol. 2019;75:410–8.
36. Chmura SJ, Winter K, Salama JK, Robinson CG, Pisansky TM, 50. Ost P, Reynders D, Decaestecker K, Fonteyne V, Lumen N,
Borges V, Al-Hallaq HA, Matuszak MM, Park SS, Gonzalez VJ, DeBruycker A, Lambert B, Delrue L, Bultijnck R, Claeys T,
Hasan Y. Phase I trial of stereotactic body radiation therapy (SBRT) Goetghebeur E. Surveillance or metastasis-directed therapy for
to multiple metastatic sites: a NRG oncology study. Int J Radiat oligometastatic prostate cancer recurrence: a prospective, random-
Oncol Biol Phys. 2018;102:S68–9. ized, multicenter phase II trial. J Clin Oncol. 2018;36(5):446–453.
37. Lam KY, Lo CY. Metastatic tumours of the adrenal glands: https://doi.org/10.1200/JCO.2017.75.4853.
a 30-year experience in a teaching hospital. Clin Endocrinol. 51. Phillips R, Shi WY, Deek M, Radwan N, Lim SJ, Antonarakis ES,
2002;56:95–101. Rowe SP, Ross AE, Gorin MA, Deville C, Greco SC. Outcomes
38. Ettinger DS, Wood DE, Aggarwal C, Aisner DL, Akerley W, of observation vs stereotactic ablative radiation for oligometastatic
Bauman JR, Bharat A, Bruno DS, Chang JY, Chirieac LR, D’Amico prostate cancer: the ORIOLE phase 2 randomized clinical trial.
TA. NCCN guidelines insights: non–small cell lung cancer, version JAMA Oncol. 2020;6:650–9.
1.2020: featured updates to the NCCN guidelines. J Natl Compr 52. https://clinicaltrials.gov/ct2/show/NCT02364557
Cancer Netw. 2019;17:1464–72. 53. Khan SA, Stewart AK, Morrow M. Does aggressive local ther-
39. Scorsetti M, Comito T, Tozzi A, Navarria P, Fogliata A, Clerici E, apy improve survival in metastatic breast cancer? Surgery.
Mancosu P, Reggiori G, Rimassa L, Torzilli G, Tomatis S. Final 2002;132:620–7.
results of a phase II trial for stereotactic body radiation therapy for 54. Ly BH, Nguyen NP, Vinh-Hung V, Rapiti E, Vlastos G. Loco-­
patients with inoperable liver metastases from colorectal cancer. J regional treatment in metastatic breast cancer patients: is there a
Cancer Res Clin Oncol. 2015;141:543–53. survival benefit? Breast Cancer Res Treat. 2010;119:537–45.
40. Rusthoven KE, Hammerman SF, Kavanagh BD, Birtwhistle MJ, 55. Yu CX, Shao X, Zhang J, Regine W, Zheng M, Yu YS, Deng J,
Stares M, Camidge DR. Is there a role for consolidative stereotac- Duan Z. GammaPod—a new device dedicated for stereotactic
tic body radiation therapy following first-line systemic therapy for radiotherapy of breast cancer. Med Phys. 2013;40:051703.
65 Radiation Therapy for Extracranial Oligometastatic Disease 695

56. Palma DA, Olson R, Harrow S, Correa RJ, Schneiders F, Haasbeek therapy for metastatic malignant melanoma. Am J Clin Oncol.
CJ, Rodrigues GB, Lock M, Yaremko BP, Bauman GS, Ahmad 2015;38:119–25.
B. Stereotactic ablative radiotherapy for the comprehensive treat- 60. Yamamoto M, Serizawa T, Shuto T, Akabane A, Higuchi Y,
ment of 4–10 oligometastatic tumors (SABR-COMET-10): Kawagishi J, Yamanaka K, Sato Y, Jokura H, Yomo S, Nagano
study protocol for a randomized phase III trial. BMC Cancer. O. Stereotactic radiosurgery for patients with multiple brain metas-
2019;19:1–5. tases (JLGK0901): a multi-institutional prospective observational
57. Olson R, Mathews L, Liu M, Schellenberg D, Mou B, Berrang study. Lancet Oncol. 2014;15:387–95.
T, Harrow S, Correa RJ, Bhat V, Pai H, Mohamed I. Stereotactic 61. Henke L, Kashani R, Robinson C, Curcuru A, DeWees T, Bradley
ablative radiotherapy for the comprehensive treatment of 1–3 J, Green O, Michalski J, Mutic S, Parikh P, Olsen J. Phase I trial of
Oligometastatic tumors (SABR-COMET-3): study protocol for a stereotactic MR-guided online adaptive radiation therapy (SMART)
randomized phase III trial. BMC Cancer. 2020;20:1–2. for the treatment of oligometastatic or unresectable primary malig-
58. Postow MA, Callahan MK, Barker CA, Yamada Y, Yuan J, Kitano nancies of the abdomen. Radiother Oncol. 2018;126:519–26.
S, Mu Z, Rasalan T, Adamow M, Ritter E, Sedrak C. Immunologic 62. Sulaiman NS, Fujii O, Demizu Y, Terashima K, Niwa Y, Akagi T,
correlates of the abscopal effect in a patient with melanoma. N Engl Daimon T, Murakami M, Sasaki R, Fuwa N. Particle beam radia-
J Med. 2012;366:925–31. tion therapy using carbon ions and protons for oligometastatic lung
59. Okwan-Duodu D, Pollack BP, Lawson D, Khan MK. Role of radia- tumors. Radiat Oncol. 2014;9:183.
tion therapy as immune activator in the era of modern immuno-
 Spine Radiotherapy
66
Mark Ashamalla, Alexander N. Slade, Kartik Mani,
and Samuel Ryu

Abstract
by reviewing both historical and contemporary clin-
A large portion of cancer patients will develop spinal ical trials.
metastases during the natural course of their disease. • To understand the role of SRS/SBRT in the treat-
While spine metastases may be asymptomatic in some, ment of spine metastases in multiple clinical sce-
they commonly lead to severe pain affecting one’s quality narios, including de novo spine metastases,
of life and may progress to cord compression. The treat- reirradiation, and postoperatively in cases of cord
ment paradigm for spine metastases has continued to compression.
evolve over time with overall advancements in imaging
which has allowed increasing the use of stereotactic body
radiotherapy (SBRT) and stereotactic radiosurgery
(SRS). These are highly conformal therapies delivering
high radiation doses in a single or two to five fraction 66.1 Introduction
regimens by using multiple beams or arcs to target the
metastasis all while sparing the spinal cord. This chapter Spinal metastases are a complicated problem affecting a
reviews the evolution of spinal stereotaxis, the radiobio- large portion of adults in the United States across multiple
logical basis behind radiosurgery, and several modern- cancer subtypes. These lesions can be asymptomatic, cause
day uses of radiosurgery including de novo spine bothersome to debilitating pain impacting performance sta-
metastases, reirradiation of spine metastases, radiosur- tus and quality of life, and in the worst case, progress to
gery in the postoperative setting, and the role of SRS in metastatic epidural spinal cord compression (MESCC). The
cases of neurologic compromise. latter occurs in 10% of these patients, with more than 20,000
new cases reported every year in the United States [1].
Spinal metastases are ideally managed in a multidisci-
Learning Objectives plinary fashion with pain control usually being the immedi-
• To understand the historical management of spine ate goal of treatment. External beam radiation therapy
metastases and review the evolution of treatment (EBRT) is widely used with doses of 30 Gy in ten fractions
paradigms with the advent of radiosurgery. directed to the involved spine including one to two vertebral
• To understand the radiobiological basis behind bodies superior and inferior to the affected spinal level. With
radiosurgery and its ability to improve the therapeu- the recent technical progress of radiation delivery and immo-
tic ratio. bilization, spine radiosurgery (SRS)/stereotactic body radia-
• To understand the evidence supporting the use of tion therapy (SBRT) has been used to deliver large doses in
radiosurgery in palliation and local disease control a single fraction, or over the course of a few fractions,
directed to the involved spine only. When there is mechani-
cal spine instability, a surgical approach with stabilization is
preferred, with kyphoplasty used for pain control of com-
pression fractures. Many patients ultimately receive some
M. Ashamalla (*) · A. N. Slade · K. Mani · S. Ryu combination of the above. The degree and type of interven-
Department of Radiation Oncology, Stony Brook University tion depend on several critical factors: performance status,
Hospital, Stony Brook, NY, USA the extent of disease both systemically and at the involved
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 697
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_66
698 M. Ashamalla et al.

spine level, the stability of the underlying spine, and the 8–10 Gy in a single dose as opposed to fractionated regimens
degree of symptoms the patient is experiencing. [9]. Together with the experimental data and experientially
Metastatic epidural spinal cord compression (MESCC) driven linear quadratic BED predictability, large fraction
with canal compromise is treated with decompressive sur- regimens such as SRS and SBRT were brought into clinical
gery followed by EBRT [2]. It is estimated that only 15% of use. In fact, such large single doses have been used exten-
these patients are favorable surgical candidates [1]; however, sively in radiosurgery of various brain lesions, and the clini-
such a comparison has not been repeated in the era of radia- cal effectiveness of controlling brain metastases is well
tion dose escalation with SRS/SBRT. For other patients established. However, such large doses could not be given to
showing radiographic compression without acute neurologi- the spine where the spinal cord lies within and thus came the
cal deficits, or those with non-compressive lesions deemed need to develop another new method of radiation delivery
high risk for potential neurologic compromise, there is no under accurate stereotactic localization to avoid potential
clear consensus on the optimal treatment strategy. Some damage to the spinal cord.
algorithms, such as the mechanical stability, neurological
risk, oncological parameters, and patient wishes (MNOP),
have been developed to guide practitioners to preferred 66.3 History of Spinal Stereotaxis
options based on many of the factors above [3]. These treat-
ment options also incorporate more advanced radiation tech- The advent of spinal stereotaxis dates back to the late nine-
niques. SRS and SBRT provide highly conformal, ablative teenth century when frames were first created by Dittmar and
radiation to spinal tumors and have been shown to be safe Woroschiloff in an effort to study spinal cord lesions in labo-
and effective [4–6], even in cases of epidural extension or ratory animals, which was then followed by Clarke in 1920
cord involvement. It should be noted that SRS/SBRT can be with the creation of the first stereotactic device for experi-
delivered either as a primary treatment or in the adjuvant set- mental use in the spine, requiring the fixation of clamps to
ting following decompressive surgery and stabilization. This exposed lamina [10]. Mullan and investigators tried clinical
chapter will focus on radiotherapeutic uses of SRS and SBRT applications of spinal stereotaxis for percutaneous cordot-
for spine metastasis and MESCC. omy, which subsequently lead to the development of the ste-
reotactic frame by Rand in 1965, which was later used by
Crue and Hitchcock in various procedures involving trigemi-
66.2 Radiobiology of SRS/SBRT nal and spinothalamic tractotomy [11]. More recently,
Hamilton et al. developed an anchoring device to the verte-
Oncological radiobiology has been mostly studied with bral bone by an open procedure in order to perform stereo-
regard to cell survival and the dynamics of tissue or tumor tactic delivery of high radiation doses using a linear
mass after daily fractionated doses of 1.8–2 Gy. To compare accelerator [12]. In the same time period, a noninvasive
various different radiation doses or fractionations, we calcu- method of table-attached or floor-mounted frame was devel-
late the dose delivered to different tissues, using what is oped for spine stereotaxis by Lax et al. [13]. This became the
called the biological effective dose (BED). BED incorpo- basis of frameless stereotaxis of spine radiosurgery under
rates not only the total dose given but also the dose delivered image guidance. Since then, the field of spine radiosurgery
per fraction [7]. The calculated BED then predicts the lethal- has welcomed a host of new linear accelerator-based radia-
ity of the delivered radiation, using the linear quadratic cell tion delivery systems.
survival formula, which contains surrogates for radiosensi-
tivity of tumor versus normal tissue to the doses used in dif-
ferent fractionation regimens. Putting all these radiobiological 66.4  elivery Methods of High Radiation
D
principles together leads us to infer that by delivering the Doses
same total dose, but in fewer fractions (i.e., SRS), we will
deliver a greater biologically effective dose to the tumor. While radiosurgery was first developed for intracranial
However, if the goal is to increase the therapeutic ratio, one lesions due to its precision, the technology of modern linear
may be best served by greater total dose fractionation. accelerators (LINAC) has allowed this treatment to be uti-
There is additional direct experimental evidence that a lized in the management of other body sites, particularly the
large dose in a single fraction results in significantly increased spine. In general, LINAC-based SRS uses multiple arcs
tumor cell kill in a transplanted fibrosarcoma tumor model. delivering photons to a specified target with intensity modu-
This increased tumor control was due to microvascular endo- lated radiation. Currently, dynamic LINAC systems are
thelial apoptosis by activation of the ceramide pathway by mostly used, which involve gantry rotation around the
acid sphingomyelinase [8]. The enhanced radiation response patient, while micromultileaf collimators are simultaneously
was seen more commonly with radiation doses higher than in motion, three-dimensionally conforming to the tumor, all
66 Spine Radiotherapy 699

while the treatment couch is being robotically controlled.

There are several currently commercially available LINAC
radiosurgery systems.
Another method of treatment delivery is charged particle
radiosurgery which can use either the Bragg peak method,
where charged particles come to stop within the target vol-
ume, or the plateau-beam method, where the target is cov-
ered by energy deposition of the charged particles spread
across it. Since it also has the potential to be used for deliv-
ery of high doses, it may be developed for a greater role in
SRS/SBRT.

66.5 SRS for Spine Metastases

EBRT has been used for palliation of pain, commonly utiliz-

ing a radiation regimen delivering 30 Gy in ten fractions.
Conventional EBRT as monotherapy was used with pallia-
tive intent, assuming generally that the spine metastasis was
in a patient at a terminal stage of cancer. Thus, EBRT is Fig. 66.1 SRS spine treatment plan for metastasis. A lung cancer
usually accomplished by a rather simple technique using patient presenting with a painful L4 metastasis causing radiculopathy is
two or more opposed beams depositing a high dose across depicted. The patient was treated with 16 Gy to the 100% isodose line
the entire width of the patient including the spinal cord and (red color) surrounding the target volume. Thin green line is 12 Gy and
pink line is 8 Gy isodose. At 2 months after radiosurgery, the patient
other internal organs within the radiation beam paths [3]. had complete resolution of the presenting pain symptoms and after
EBRT is given to the involved spine including one to two 1 year, still had a durable tumor response on follow-up imaging. The
uninvolved vertebral bodies superior and inferior to the patient remained fully ambulatory and active
affected vertebral level(s). This can be delivered as mono-
therapy, adjuvantly following surgery, or salvage in a reir- long-term tumor control of the involved spine and preventing
radiation setting [3]. potential neurological complications.
The treatment paradigm for spine metastases has contin- A typical treatment with SRS is shown in Fig. 66.1.
ued to evolve over time with overall advancements in imag- Common radiosurgical dose regimens are single-fraction
ing, most notably with MRI. Through the increasingly SRS delivering 15–20 Gy, ultra-high single-dose radiother-
routine use of spine MRI, we have improved our ability to apy of 24 Gy, or hypofractionated regimens such as
detect and diagnose early and more localized spine metasta- 3 × 9 Gy. There is some evidence that high single doses may
ses prior to the development of any symptoms or neurologi- improve local control of oligometastatic lesions over more
cal compromise. However, the true incidence of clinically hypofractionated regimens (3 × 9 Gy SBRT) [17]; however
occult spine metastases remains unknown. Advanced imag- this remains to be confirmed.
ing also helps visualize anatomical details of the spinal col- Since the initial reported clinical experience of spine
umn and the spinal cord when designing SBRT and SRS SRS, there has been an increasing use of SRS/SBRT for
spine plans. These are highly conformal therapies delivering spine metastases. Many clinical trials as well as institutions
high radiation doses in a single or two to five fraction regi- have presented their experiences with SRS/SBRT for spinal
mens by using multiple beams or arcs to target the metastasis metastases reporting rapid and durable pain relief in the
all while sparing the spinal cord. In contrast to EBRT, SRS range of 85–95%, in addition to demonstrating long-term
and SBRT target only the involved spine, and thus they are tumor control in the treated spine region with a median dura-
used for only limited or localized spine metastases. In terms tion of 13 months. Quality of life was also improved second-
of defining the target volume, a consensus target delineation ary to pain control. Based on the spine SRS experiences of
method has been reported considering the extent of involve- many single institutions, multiple randomized clinical trials
ment and anatomical boundaries of the vertebral body [14]. have been conducted [4, 6, 17–20].
Early studies as well as a more recent study in the setting of The first and earliest phase II/III randomized study was
oligometastases by Palma et al. have suggested improvement conducted by the NRG RTOG 0631 [21]. This study com-
of local spine tumor control and possible survival benefit pared a single dose of 16 Gy or 18 Gy with SRS to 8 Gy with
[15, 16]. Therefore, the primary goal when treating spine EBRT, in 352 enrolled patients. SRS was delivered to the
metastases is not only pain palliation but also achieving involved spine segment only, while EBRT was given to the
700 M. Ashamalla et al.

involved spine with one additional spine segment above and 30 Gy in ten fractions using 3DCRT in a sample of 60
below the treated spine. For SRS, the spinal cord dose con- patients, using a study design similar to that used in RTOG
straints were either 10 Gy to V10% of the spinal cord volume 0631. The investigators reported a faster pain response within
(defined from 6 mm superior to 6 mm inferior to the treated 3 months and a significant pain reduction at 6 months follow-
target spine) or 10 Gy to V0.35cc. The eligibility criteria ing SRS over 3DCRT with statistical significance [23].
were patients (age ≥ 18 years) with one to three treatment-­ Taken together, SRS appears to be the preferred treatment
naïve spine metastases: (1) a solitary spine metastasis, (2) for localized spine metastases to achieve short- and long-­
two contiguous spine levels involved, or (3) a maximum of term pain response and potential tumor cure of the treated
three separate sites. Each separate site could involve two spine. Localized limited spine metastases may also be a
contiguous vertebral bodies. Every patient was required to component of an oligometastatic state where consolidative
have a baseline pain score equal to or higher than 5 at presen- radiation with SRS/SBRT for metastatic disease has been
tation using a 0–10 numerical rating pain scale. Pain medica- shown to provide significant survival benefit [16, 24]. It is
tion including narcotics was allowed and recorded. Minimal important to consider the current oncological landscape
epidural lesions with at least a 3-mm gap between the tumor where clinically occult and asymptomatic solitary or limited
and the spinal cord and paraspinal masses (< 5 cm) directly spine metastases are more frequently detected as part of
attached to the vertebral body were eligible. Radioresistant oligometastases or longitudinally, during the course of sys-
tumors (melanoma, renal cell carcinoma, soft tissue sar- temic therapy, by a wide use of PET and/or MRI scans. SRS
coma) were also eligible. Vertebral compression fractures can achieve long-term tumor control that can also prevent
with more than 50% height loss and/or bony retropulsion neurological progression. Choice of treatment should be
were ineligible. individualized considering the systemic extent of metastases,
The results demonstrated a surprisingly lower pain tumor burden within the involved spine, presenting symp-
response rate in the SRS arm, unexpected from previously toms, palliative care measures, and general plan of oncologi-
reported outcomes. The primary endpoint of patient-reported cal care.
short-term pain response of the index spine was 41.3% in the
SRS arm compared to 60.5% in the cEBRT arm at 3 months
(p = 0.99). However, long-term pain response after 12 months 66.6 SRS for Reirradiation
post-treatment showed an improvement of pain response of
61.3% by SRS compared to 43.7% by cEBRT (p = 0.20). Of By some estimates, nearly half of patients who receive con-
note, complete pain relief at the index spine was reported as ventionally fractionated radiotherapy will be diagnosed with
41.9% in the SRS arm compared to 31.3% in the cEBRT arm persistent or progressive disease requiring further radiation
at 24 months. Multivariate analysis showed that the number [25]. This presents a unique and difficult scenario for clini-
of spine metastases (1 versus 2–3) was the most significant cians because reirradiation with further conventionally
predictor of pain response over time (p = 0.023). Pre-­ fractionated radiotherapy, many times just months following
treatment pain score did not impact pain response. the initial course, will require a dose lower than that given
Radioresistant tumors responded similarly to treatment com- during the first course. For this reason, these subsequent
pared to other histologies. courses often provide inferior local control, with overall
Another randomized phase II/III study conducted by the response rates of approximately 50% and complete response
Canadian Cancer Trials Group (CCTG) in collaboration with rates reported as less than 15% [26].
the Trans Tasman Radiation Oncology Group (TROG) Given the inferior efficacy with conventionally fraction-
recently compared 24 Gy in two SBRT fractions versus ated radiotherapy, clinicians turned to radiosurgery instead
20 Gy in five conventional palliative radiotherapy (CRT) in cases requiring reirradiation. By most estimates, reirradi-
fractions [22]. Preliminary results showed that patients in ating with radiosurgery provides significantly greater local
both treatment groups experienced reductions in pain from control (>75%) than that observed with conventionally frac-
spinal metastases. At 3 months, 35% of patients in the SBRT tionated radiotherapy, all while demonstrating a reasonable
arm of the trial, compared to 14% of those in the CRT arm, toxicity profile, which by most accounts is comparable to de
reported a complete response with no remaining pain from novo SBRT [27]. Thibault et al. evaluated the role of radio-
their lesions (p < 0.001). After 3 months, 92% of patients in surgery in the salvage setting for patients with in-field failure
the SBRT arm and 86% of those in the CRT arm were cancer-­ of radiotherapy and reported an 81% local control rate at
free at the treated site (p = 0.4); the rates at 6 months were 1 year with no observed radiation-induced vertebral com-
75% and 69% (p = 0.4). Long-term results are still pending pression fractures or myelopathies observed [28]. As we
at this time. continue to improve outcomes in cancer, a greater proportion
A smaller German randomized phase II trial compared of patients may require reirradiation in many instances.
pain response from 24 Gy single-fraction SRS/SBRT to Radiosurgery has been demonstrated as, and will certainly
66 Spine Radiotherapy 701

continue to be, explored as a viable and preferred treatment When MESCC or canal compromise is present, the usual
choice for recurrent tumors requiring multiple courses of first course of action is immediate administration of gluco-
radiation. corticoids, generally with dexamethasone 10 mg IV bolus
followed by 16 mg per day (4 mg every 6 h), tapered over
2 weeks [32], but judicious use of lower doses is recom-
66.7 SRS Following Surgery mended to avoid steroid complications.
The treatment goals for patients with spinal cord com-
Prospective randomized trials in the early 1990s demon- pression (or compromise) are decompression of the spinal
strated that surgery (decompressive corpectomy) followed cord, control of the epidural disease, and preservation or
by conventional EBRT (30 Gy in ten fractions) provided sig- improvement of neurological function. Surgical resection or
nificant improvement in overall ambulatory rates compared radiation is the mainstay of treatment and offers the best
to EBRT alone. The results of this era’s trials served as proof option for epidural tumor control and prevention of cord
of concept for future attempts to further improve on these compression with subsequent neurologic deficits. With sur-
results with SRS/SBRT, which came into practical use in the gical resection and postoperative EBRT, the resulting superi-
2000s. Initial experiences [29], while vitally important for ority of ambulation over EBRT alone was tested using an
the progression of the field, were limited by the difficulties endpoint of four steps with walking aid. However, this often
defining the tumor, surgical bed, and spinal cord when adja- requires corpectomy using an anterior and posterior approach
cent to neurosurgical hardware. Myelograms were initially with extensive instrumentation. Many oncological patients
used as a substitute for the poor quality of CT and MRI are not suitable candidates for such involved surgery, but still
images. Later, consensus guidelines were developed to aid require a spinal cord decompression, ideally using a nonin-
providers and better standardize postoperative target vol- vasive approach, to improve their neurological function.
umes when using radiosurgery [30]. These guidelines include With the progress of spine SRS, one such landmark phase II
recommendations on the extent of preoperative disease sites trial in the 2000s studied SRS (to a dose of 14–20 Gy in a
to be covered, coverage of postoperative gross and ­subclinical single fraction) in the management of epidural spinal cord
disease, and different planning target volume (PTV) expan- compression [33]. Ryu et al. demonstrated a 65% mean
sions according to the institution’s immobilization systems. reduction in epidural tumor volume seen on 2-month MRI
The use of separation surgery then grew from a need to scan with an 80% overall epidural tumor response rate. Of
protect the cord while simultaneously avoiding potential note, this study also showed that the majority (94%) of
“underdosing” of the target volume. The intent was to reduce patients who were intact prior to radiosurgery remained
the tumor burden and to provide a necessary gap of a few intact following treatment, with an improvement of neuro-
millimeters between the cord and tumor, thereby allowing a logic deficits in 63% of patients presenting with such defi-
sufficient radiation dose to be given to the tumor. This cits. Figure 66.2 shows an example of MR images before and
approach was studied by Memorial Sloan Kettering, with after radiosurgery in a patient with renal cell carcinoma,
patients receiving 24 Gy in one fraction or 24–30 Gy in three commonly regarded as a radioresistant tumor, causing
fractions, and demonstrated a low local tumor progression MESCC.
ranging from 4 to 9% at 1 year post-treatment [31]. While not possible to compare two distinct trials men-
tioned above in their entireties, one may be able to compare
the overall outcomes of preservation or improvement of neu-
66.8  adiation in MESCC or Canal
R rologic function as it relates to each treatment modality. A
Compromise and Cases of Neurologic comparison between subset results of these two trials [2, 33]
Compromise is seen in Table 66.1, with 94% of patients remaining ambu-
latory or neurologically intact after either therapy (SRS or
Not all epidural tumors causing canal compromise cause surgery) and 62–63% of patients presenting as non-­
neurologic deficits; however if left untreated, they most cer- ambulatory or with neurological deficits improving to ambu-
tainly can progress to the point of symptomatic and overt latory or intact following surgery or SRS.
spinal cord compression. These are complicated cases requir- Contrary to current practice, these studies suggest that
ing multidisciplinary care among neurosurgery, radiation SRS can be used as the primary treatment for MESCC for
oncology, and medical oncology. Together with pertinent patients who do not have neurological deficits, as SRS can
neurological examination, appropriate MR imaging is take effect gradually to reduce the epidural lesion, while
required for accurate evaluation. We recommend the patient upfront surgical resection is better indicated for patients with
to undergo complete spine imaging to rule out additional lev- acute neurological signs as it provides more immediate spi-
els of compression and aid in treatment decision-making. nal cord decompression.
702 M. Ashamalla et al.

a b

Fig. 66.2 SRS spine treatment plan for MESCC. A renal cell cancer with 18 Gy to the 90% isodose line surrounding the target volume of the
patient with a T12 spinal cord compression causing radiculopathy is epidural and paraspinal tumor and vertebral body. Panel B shows that at
depicted. Panel A shows significant epidural spinal cord compression 3 months after radiosurgery, epidural decompression was achieved and
with minimal thecal sac patency at presentation. The patient was treated thecal sac patency improved to 95%

Table 66.1 Comparison of clinical trial results of surgery versus single-fraction SRS has been reported by some groups as
radiosurgery cEBRT, conventionally fractionated external beam radia-
being associated with a higher incidence of pain flares [35].
tion therapy; SRS, stereotactic radiosurgery
Pain flares are transient and responsive to short courses of
Patchell’s phase III cEBRT Surgery + Ryu’s phase SRS
trial [2] alone cEBRT II trial [33] alone
low-dose glucocorticoids (e.g., dexamethasone 4 mg daily
Ambulatory patients 74% 94% Intact patients 94% for up to 5 days). However, prophylactic use of steroids is not
remain ambulatory (26/35) (32/34) remain intact (33/35) usually advised. The etiology of pain flares is not yet com-
Overall ambulatory 57% 84% Overall intact 81% pletely understood at this time, but practitioners of spine
rate (29/51) (42/50) rate (50/62) radiosurgery are generally advised to limit the radiation dose
Non-ambulatory 19% 62% Deficit 63%
to the spinal nerve root to 14 Gy.
patients improve to (3/16) (10/16) patients (17/27)
ambulatory improve to Other side effects from radiation have been reported on
intact and are related to varying doses to neighboring tissues,
including the larynx, pharynx, esophagus, bowel, lung, kid-
neys, etc., which are therefore contoured when planning a
66.9 Adverse Effects of Radiation radiosurgical treatment. Patients will typically present with
symptoms of dysphagia and odynophagia depending on the
When considering the potential adverse effects of radiother- location of the spinal target. Generally, these side effects are
apy for spinal metastases, one is best served by separating self-limited and will resolve within a few weeks following
effects in the acute setting from those more likely to be pres- treatment, but there have been reports of long-term develop-
ent long term. ment of tracheoesophageal fistulas [36]; for this reason, we
limit the esophageal dose to a total dose of 10–12 Gy in a
single fraction [37].
66.10 Acute Complications

One of the most well-reported adverse effects of radiosur- 66.11 Long-Term Complications
gery in the acute setting is what has been termed pain flare,
an acute exacerbation of pain usually occurring within It is of utmost importance to avoid any and all radiation-­
1–5 days following treatment. Pain flares following SBRT induced damage to the spinal cord as even minor injury may
have been reported with an incidence in the range of 20–60% severely and adversely affect a patient’s quality of life.
and may present anytime starting as early as 24 h following Several analyses and consensus of normal tissue dose
treatment to over 3 weeks later [34]. Compared to SBRT, constraints for stereotactic radiotherapy have been performed
66 Spine Radiotherapy 703

and published by Task Group 101 [38] and the UK groups References
[39], respectively. The NRG RTOG 0631 study used a partial
spinal cord volume with a tolerance dose of 10 Gy to the 1. Byrne TN, Borges LF, Loeffler JS. Metastatic epidural spinal cord
10% partial volume of the spinal cord defined as including compression: update on management. Semin Oncol. 2006;33:307–
6 mm above and below the target volume (calculated to be 11.
equivalent to 0.35 cc) [15], with the tolerance dose noted to 2. Patchell RA, et al. Direct decompressive surgical resection in the
treatment of spinal cord compression caused by metastatic cancer:
vary according to the dose fractionation being used. a randomised trial. Lancet. 2005;366:643–8.
Vertebral compression fractures have been reported in up to 3. Spratt DE, et al. An integrated multidisciplinary algorithm for the
15% of patients after spine SRS [40]. In order to better asses management of spinal metastases: an international spine oncology
which patients may be at highest risk for such an event, analy- consortium report. Lancet Oncol. 2017;18:e720–30.
4. Ryu S, et al. Pain control by image-guided radiosurgery for solitary
ses of predictive factors have been performed demonstrating spinal metastasis. J Pain Symptom Manag. 2008;35:292–8.
that the patients at highest risk of developing vertebral com- 5. Gerszten PC, et al. Single-fraction radiosurgery for the treatment of
pression fractures are those with prior kyphotic/scoliotic spinal breast metastases. Cancer. 2005;104:2244–54.
deformities, lytic lesions, and receiving higher doses (≥ 20 Gy). 6. Chang EL, et al. Phase I/II study of stereotactic body radiotherapy
for spinal metastasis and its pattern of failure. J Neurosurg Spine.
2007;7:151–60.
7. Fowler JF. 21 years of biologically effective dose. Br J Radiol.
66.12 Conclusions 2010;83:554–68.
8. Garcia-Barros M, Paris F, Cordon-Cardo C, Lyden D, Rafii S,
Haimovitz-Friedman A, Fuks Z, Kolesnick R. Tumor response
Stereotactic radiosurgery is the culmination of years of to radiotherapy regulated by endothelial cell apoptosis. Sci-
extensive research into the application of radiobiology, radi- ence 2003;300:1155–1159. doi: https://doi.org/10.1126/sci-
ation physics, and advanced imaging and technology in ence.1082504.: PMID: 12750523.
radiotherapy. This synergy has enabled radiation oncolo- 9. Fuks Z, Kolesnick R. Engaging the vascular component of the tumor
response. Cancer Cell. 2005;8:89–91. https://doi.org/10.1016/j.
gists to ablate metastatic lesions in the spine (whether in the ccr.2005.07.014. PMID: 16098459
initial or reirradiation setting). There are several practical 10. Gabriel EM, Nashold BS Jr. History of spinal cord stereotaxy. J
advantages of using spine SRS. For instance, since only the Neurosurg. 1996;85:725–31.
directly involved spine is treated, more functional bone mar- 11. Hitchcock E. Stereotactic cervical myelotomy. J Neurol Neurosurg
Psychiatry. 1970;33:224–30.
row can also be preserved. This may help facilitate the con- 12. Hamilton AJ, Lulu BA. A prototype device for linear accelerator-­
tinuation of systemic therapy, which is often required in based extracranial radiosurgery. Acta Neurochir Suppl. 1995;63:
these patients. A single session of SRS or a few sessions of 40–3.
SBRT will not interfere with an ongoing chemotherapy 13. Lax I, Blomgren H, Naslund I, Svanstrom R. Stereotactic radio-
therapy of malignancies in the abdomen. Methodological aspects.
schedule as opposed to the multiple visits for conventional Acta Oncol. 1994;33:677–83.
fractionated radiotherapy. Single or shorter fractionation is 14. Cox BW, Spratt DE, Lovelock M, Bilsky MH, Lis E, Ryu S, Shee-
indeed more convenient for both the patient and their care- han J, Gerszten PC, Chang E, Gibbs I, Soltys S, Sahgal A, Deasy J,
givers all while causing no increased toxicity to the treated Flickinger J, Quader M, Mindea S, Yamada Y. International Spine
Radiosurgery Consortium consensus guidelines for target volume
spine compared to conventional radiotherapy. definition in spinal stereotactic radiosurgery. Int J Radiat Oncol
Biol Phys. 2012;83:e597–605.
15. Ryu S, Jin JY, Jin R, et al. Partial volume tolerance of the spi-
Open Questions nal cord and complications of single-dose radiosurgery. Cancer.
2007;109:628–36.
• What is the role of charged particle therapy in spine 16. Palma DA, Olson R, Harrow S, Gaede S, Louie AV, Haasbeek C,
radiosurgery? Mulroy L, Lock M, Rodrigues GB, Yaremko BP, Schellenberg D,
• Is there a synergistic effect between SRS and Ahmad B, Senthi S, Swaminath A, Kopek N, Liu M, Moore K, Cur-
immunotherapy? rie S, Schlijper R, Bauman GS, Laba J, Qu XM, Warner A, Senan
S. Stereotactic ablative radiotherapy for the comprehensive treat-
• What is the appropriate sequence of SRS when ment of oligometastatic cancers: long-term results of the SABR-­
combined with separation surgery? COMET phase II randomized trial. J Clin Oncol. 2020;38:2830–8.
• What are the precise criteria to define which treat- 17. Zelefsky MJ, Yamada Y, Greco C, Lis E, Schöder H, Lobaugh S,
ment (surgery/SRS) a patient with spinal metasta- Zhang Z, Braunstein S, Bilsky MH, Powell SN, Kolesnick R, Fuks
Z, Phase III. Multi-center, prospective, randomized trial comparing
ses should receive? single dose 24 Gy radiotherapy to a 3-fraction SBRT regimen in the
• Will advancements in neuroimaging and radiomics/ treatment of Oligometastatic cancer. Inte J Radiat Oncol Biol Phys.
radiogenomics incorporating PET or MR spectros- 2021; https://doi.org/10.1016/j.ijrobp.2021.01.004)
copy help improve the therapeutic ratio and post-­ 18. Ryu S, Fang Yin F, Rock J, Zhu J, Chu A, Kagan E, Rogers L,
Ajlouni M, Rosenblum M, Kim JH. Image-guided and intensity-­
treatment surveillance? modulated radiosurgery for patients with spinal metastasis. Can-
cer. 2003 Apr 15;97:2013–8. https://doi.org/10.1002/cncr.11296.
PMID: 12673732
704 M. Ashamalla et al.

19. Yamada Y, Bilsky MH, Lovelock DM, et al. High-dose, single-­ 29. Rock JP, Ryu S, Shukairy MS, et al. Postoperative radiosurgery for
fraction image-guided intensity-modulated radiotherapy for malignant spinal tumors. Neurosurgery. 2006;58:891–8. https://doi.
metastatic spinal lesions. Int J Radiat Oncol Biol Phys. 2008;71: org/10.1227/01.NEU.0000209913.72761.4F.
484–90. 30. Redmond KJ, Robertson S, Lo SS, et al. Consensus contouring
20. Gerszten PC, Burton SA, Ozhasoglu C, et al. Radiosurgery for spi- guidelines for postoperative stereotactic body radiation therapy for
nal metastases: clinical experience in 500 cases from a single insti- metastatic solid tumor malignancies to the spine. Int J Radiat Oncol.
tution. Spine (Phila Pa 1976). 2007;32:193–9. 2017;97:64–74. https://doi.org/10.1016/j.ijrobp.2016.09.014.
21. Ryu S, Deshmukh S, Timmerman RD, Movsas B, Gerszten PC, 31. Moussazadeh N, Laufer I, Yamada Y, Bilsky MH. Separation sur-
Yin FF, et al. Radiosurgery compared to external beam radio- gery for spinal metastases: effect of spinal radiosurgery on surgical
therapy for localized spine metastasis: phase III results of NRG treatment goals. 2014. http://journals.sagepub.com/doi/pdf/10.117
oncology/RTOG 0631. Int J Radiat Oncol Biol Phys. 2019;105: 7/107327481402100210;21. Accessed October 2, 2018
S2–3. 32. Sørensen S, Helweg-Larsen S, Mouridsen H, Hansen HH. Effect
22. Sahgal A, Myrehaug SD, Siva S, Masucci L, Foote MC, Brundage of high-dose dexamethasone in carcinomatous metastatic spinal
M, et al. CCTG SC. 24/TROG 17.06: a randomized Phase II/III cord compression treated with radiotherapy: a randomised trial.
study comparing 24Gy in 2 stereotactic body radiotherapy (SBRT) Eur J Cancer. 1994;30A:22–7. http://www.ncbi.nlm.nih.gov/
fractions versus 20Gy in 5 conventional palliative radiotherapy pubmed/8142159. Accessed September 27, 2018
(CRT) fractions for patients with painful spinal metastases. Int J 33. Ryu S, Rock J, Jain R, et al. Radiosurgical decompression of meta-
Radiat Oncol Biol Phys. 2020;108:1397–8. static epidural compression. Cancer. 2010; https://doi.org/10.1002/
23. Sprave T, Verma V, Förster R, Schlampp I, Bruckner T, Bostel T, cncr.24993.
Welte SE, Tonndorf-Martini E, Nicolay NH, Debus J, Rief H. Ran- 34. Chiang A, Zeng L, Zhang L, et al. Pain flare is a common adverse
domized phase II trial evaluating pain response in patients with event in steroid-naïve patients after spine stereotactic body radiation
spinal metastases following stereotactic body radiotherapy versus therapy: a prospective clinical trial. Int J Radiat Oncol Biol Phys.
three-dimensional conformal radiotherapy. Radiother Oncol. 2018 2013;86:638–42. https://doi.org/10.1016/j.ijrobp.2013.03.022).
Aug;128:274–82. https://doi.org/10.1016/j.radonc.2018.04.030. 35. Pan HY, Allen PK, Wang XS, et al. Incidence and predictive factors
Epub 2018 May 26. PMID: 29843899 of pain flare after spine stereotactic body radiation therapy: second-
24. Mitchell KG, Farooqi A, Ludmir EB, Corsini EM, Zhang J, Sepesi ary analysis of Phase 1/2 trials. Int J Radiat Oncol. 2014;90:870–6.
B, Vaporciyan AA, Swisher SG, Heymach JV, Zhang J, Gomez DR, https://doi.org/10.1016/J.IJROBP.2014.07.037.
Antonoff MB. Improved overall survival with comprehensive local 36. Jin R, Rock J, Jin J-Y, et al. Single fraction spine radiosurgery for
consolidative therapy in synchronous oligometastatic non-small-­ myeloma epidural spinal cord compression. J Exp Ther Oncol.
cell lung cancer. Clin Lung Cancer. 2020;21:37–46.e7. https:// 2009;8:35–41. http://www.ncbi.nlm.nih.gov/pubmed/19827269.
doi.org/10.1016/j.cllc.2019.07.007. Epub 2019 Aug 1. PMID: Accessed December 5, 2018
31447303 37. Schipani S, Wen W, Jin J-Y, Kim JK, Ryu S. Spine radiosurgery:
25. Hartsell WF, Scott CB, Bruner DW, et al. Randomized trial of a dosimetric analysis in 124 patients who received 18 Gy. Int J
short- versus long-course radiotherapy for palliation of painful Radiat Oncol Biol Phys. 2012;84:e571–6. https://doi.org/10.1016/j.
bone metastases. J Natl Cancer Inst. 2005;97:798–804. https://doi. ijrobp.2012.06.049].
org/10.1093/jnci/dji139). 38. Hanna GG, Murray L, Patel R, et al. UK consensus on Nor-
26. Chow E, van der Linden YM, Roos D, et al. Single versus mul- mal tissue dose constraints for stereotactic radiotherapy. Clin
tiple fractions of repeat radiation for painful bone metastases: Oncol (R Coll Radiol). 2018;30:5–14. https://doi.org/10.1016/j.
a randomised, controlled, non-inferiority trial. Lancet Oncol. clon.2017.09.007).
2014;15:164–71. https://doi.org/10.1016/S1470-­2045(13)70556-­4. 39. Benedict SH, Yenice KM, Followill D, et al. Stereotactic body
27. Myrehaug S, Sahgal A, Hayashi M, et al. Reirradiation spine ste- radiation therapy: the report of AAPM task group 101 [published
reotactic body radiation therapy for spinal metastases: systematic correction appears in med Phys. 2012 Jan;39(1):563. Dosage
review. J Neurosurg Spine. 2017;27:428–35. https://doi.org/10.317 error in article text]. Med Phys. 2010;37:4078–101. https://doi.
1/2017.2.SPINE16976). org/10.1118/1.3438081.
28. Thibault I, Campbell M, Tseng C-L, et al. Salvage stereotactic body 40. Boyce-Fappiano D, Elibe E, Schultz L, et al. Analysis of the factors
radiotherapy (SBRT) following in-field failure of initial SBRT for contributing to vertebral compression fractures after spine stereo-
spinal metastases. Int J Radiat Oncol. 2015;93:353–60. https://doi. tactic radiosurgery. Int J Radiat Oncol. 2017;97:236–45. https://doi.
org/10.1016/j.ijrobp.2015.03.029). org/10.1016/j.ijrobp.2016.09.007.
 Part XVII
Immunotherapy of Solid Malignancies

Steven O’Day

The human immune system is characterized by power, specificity, and memory. Immune sur-
veillance is a primary defense against cancer. Immunosuppression induced by disease states,
exogenous medications, or natural aging increases the incidence of cancer in humans.
Extraordinary scientific progress in our understanding of the immune system in the last two
decades has led to transformative therapies in oncology. This has resulted in a rapidly increas-
ing proportion of cancer patients experiencing long-term survival.
At the turn of the twentieth century, Dr. William Coley treated cancer patients with a mix-
ture of heat-killed bacteria, later called “Coley’s toxins,” and reported dramatic tumor regres-
sions in select patients [1]. His rationale for this approach was based on the observation that
rare spontaneous cancer regressions occurred after severe bacterial infections. Later, additional
nonspecific immune approaches were used to treat bladder cancer and melanoma with the
tuberculosis vaccine, Bacillus Calmette-Guérin (BCG) [2]. Toward the end of the century, a
greater appreciation for the role of T cells in cancer immunity emerged, and T-cell growth fac-
tors (immune cytokines) were discovered [3]. This discovery led to the use of high-dose inter-
leukin-­2 (IL-2) and interferon alpha (INF-a) in the treatment of melanoma and renal cell
cancer. The twentieth century closed with seminal work from the NCI and Dr. Steven
Rosenberg’s lab where adoptive autologous tumor-infiltrating lymphocytes were used in con-
junction with high-dose IL-2 as a therapeutic strategy against advanced cancer [4]. Across all
these therapeutic efforts, only a small percentage of patients were observed to achieve tumor
eradication or prolonged tumor control; predictive biomarkers of response and survival were
elusive. The significant toxicity of high-dose cytokines and the labor-intensive preparation of
adoptive lymphocytes were formidable obstacles to broad clinical benefit and rigorous ran-
domized trials.
The current revolution of immuno-oncology (I-O) began in the first decade of the twenty-­
first century when the T-cell checkpoint, CTLA-4, was identified as a critical inhibitory T-cell
checkpoint which dampens the adaptive immune response, preventing prolonged activation
and, ultimately, autoimmunity [9]. Blocking CTLA-4 resulted in more potent T-cell priming,
activation, and memory. The seminal discovery of this checkpoint inhibitor was recently rec-
ognized with the Nobel Prize in Medicine [5]. Ipilimumab was the first therapeutic CLTA-4-­
blocking antibody to enter human studies in advanced melanoma, approximately 20 years ago.
Early studies with the single agent ipilimumab demonstrated remarkable tumor responses, or
long-term disease control, in approximately 25% of patients with widespread disease. In 2010,
a large randomized trial with ipilimumab demonstrated improved median and long-term sur-
vival in patients with advanced melanoma. This was followed by FDA approval of ipilimumab
in 2011 [6]. The revolution of checkpoint inhibitors had begun.
The success of ipilimumab represented a dramatic shift in oncology’s treatment paradigm.
Focus moved away from cancer cell-directed cytotoxic chemotherapy drugs, whose hallmarks
were rapid kinetics of responses, short durations of response, and significant off-target toxicity.
706 Immunotherapy of Solid Malignancies

The CTLA-4 treatment paradigm shift was notable for slower kinetics of response, more durable responses,
and immune-related toxicity that often correlated with clinical benefit. Immune-mediated toxicity manage-
ment was optimized; short courses of immunosuppressive medications were used and did not negatively
impact ongoing antitumor responses.
Shortly after the discovery of CTLA-4, rapid identification of additional T-cell checkpoints ensued. This
began in the last decade with the discovery of a second pivotal immune checkpoint, PD-1, and its ligand
PDL-1. Antibodies designed to block PD-1 and PDL-1 have transformed cancer therapy. PD-1 upregulation
on T cells indicates an exhausted T-cell phenotype, a much later stage of the adaptive T-cell response than
CTLA-4 upregulation. By blocking PD1 on the surface of T cells, the T cell is reactivated and cytolytic
capability is restored [7]. The first two FDA-approved anti-PD-1 antibodies were nivolumab and pembroli-
zumab. These agents demonstrated significant therapeutic benefits including durable responses and long-­
term survival for patients across a broad spectrum of solid tumors [8]. PD-1 antibodies activate T cells
predominately in the tumor microenvironment resulting in a more rapid kinetics of response and less
immune-related toxicity than CTLA-4 antibodies because they differ in T-cell biology. On the other hand,
CTLA-4’s biology may drive more durable responses and shorter treatment courses, due to its enhanced
T-cell priming and memory inducement. In the last 5 years, combinations of CTLA-4- and PD-1-blocking
antibodies have resulted in further advances in cancer I-O therapies. These combinations have improved
overall response, disease control, and durability of response across many solid tumor types [8]. Lower doses
of CTLA-4 inhibitors in combination with PD-1 inhibitors appear to maintain efficacy and reduce immune-­
related toxicity. Long-term survival is now a reality for distinct subsets of patients suffering from advanced
cancer.
The clinical validation of CTLA-4 and PD-1/PDL-1 as immune targets led to an explosion of combina-
tion immunotherapy treatment approaches leading to further advances in the clinic. These strategies include
both I-O/I-O and non-I-O/I-O combinations. Further efforts to optimize T effector function, such as addi-
tional agonistic and antagonist T-cell antibodies, as well as genetically engineered adoptive CAR-T cells are
ongoing. In addition, many new clinical trials focus on strategies to activate optimally and repolarize both
the non-T-cell immune compartment (NK cells, APCs, and suppressive myeloid cells) and tumor stromal
cells. Finally, optimal I-O treatments combined with cancer cell-directed cytotoxic therapies (chemother-
apy, antibody drug conjugates, molecular targeted agents) are generating significant successes in the clinic
and are an area of active clinical research [9].
The following chapters summarize the impact of the immuno-oncology revolution on lung, genitouri-
nary, and skin malignancies. The validated checkpoints, CTLA-4 and PD-1/PDL-1, have positively impacted
the treatment of each of these disease areas. However, the development of combination approaches varies
between and among each disease category. Melanoma has led to the checkpoint revolution and remains at
the forefront of its progress. IO/non-IO combinations in melanoma are focused on both immune check-
points and molecularly targeted agents (BRAF/MEK). Lung cancer has been at the forefront of the combi-
nation of immune checkpoints with chemotherapy. Genitourinary cancers are the leaders of combinations
of immune checkpoints with tyrosine kinase and VEGF inhibitors. Overall, significant strides are being
made in the development of innovative immunotherapy approaches for patients with all three of these can-
cer types. Sharing insights across these diseases will contribute tremendously to the progress of treatment
for years to come. Our continued drive toward improvement will lead to long-term disease control and may
help us eradicate these cancers in the coming decades.

References
1. McCarthy EF. The toxins of William B. Coley and the treatment of bone and soft-tissue sarcomas. Iowa
Orthop J. 2006; https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1888599/
2. Zheng YQ, Naguib YW, Dong Y, Shi YC, Bou S, Cui Z. Applications of bacillus Calmette-Guerin and
recombinant bacillus Calmette-Guerin in vaccine development and tumor immunotherapy. Expert Rev
Vaccines. 2015; https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4920355/
3. Berraondo P, Sanmamed MF, Ochoa MC, et al. Cytokines in clinical cancer immunotherapy. Br J Cancer.
2019; https://www.nature.com/articles/s41416-­018-­0328-­y
Immunotherapy of Solid Malignancies 707

4. Mayor P, Starbuck K, Zsiros E. Adoptive cell transfer using autologous tumor infiltrating lymphocytes
in gynecologic malignancies. Gynecol Oncol. 2018; https://www.ncbi.nlm.nih.gov/pmc/articles/
PMC6167746/
5. Press release: The Nobel Prize in Physiology or Medicine. 2018. https://www.nobelprize.org/prizes/
medicine/2018/press-­release/
6. Wolchok JD, Hodi FS, Weber JS, Allison JP, Urba WJ, Robert C, O’Day SJ, Hoos A, Humphrey R,
Berman DM, Lonberg N, Korman AJ. Development of ipilimumab: a novel immunotherapeutic approach
for the treatment of advanced melanoma. Ann N Y Acad Sci. 2013 Jul;1291(1):1–13. https://www.ncbi.
nlm.nih.gov/pmc/articles/PMC3910157/
7. Zuazo M, Gato-Canas M.. https://atm.amegroups.com/article/view/15376/html
8. Gong, et al. Development of PD-1 and PD-L1 inhibitors as a form of cancer immunotherapy: a compre-
hensive review of registration trials and future considerations. J Immunother Cancer. 2018; https://jitc.
bmj.com/content/jitc/6/1/8.full.pdf
9. Kamta J, Chaar M, Ande A, et al. Advancing cancer therapy with present and emerging immuno-­
oncology approaches. Front Oncol. 2017; h­ ttps://www.frontiersin.org/articles/10.3389/fonc.2017.00064/
full
 Immunotherapy in Melanoma
and Merkel Cell Cancer 67
Melissa Chow, Elizabeth A. Sangalang, Christine Chow,
and Adil I. Daud

Abstract 67.1 Introduction

Malignant melanoma is one of the most aggressive and
lethal forms of skin cancer, with incidence rates continu- Melanoma is a cancer characterized by the abnormal prolif-
ing to rise due to increased sun exposure. While primary eration of melanocytes, usually within the skin but also in the
cutaneous melanoma is often managed through surgery, mucous membranes and in the pigmented tissues of the eye.
this alone is not curative for advanced metastatic stages. Unlike many other more common skin cancers that remain
Within the past few decades, breakthrough discoveries in local or regional, melanoma has a propensity to spread
immunotherapy have led to successful treatment and a regionally and to distant organs. Although only accounting
better quality of life for patients living with metastatic for 1% of skin cancer diagnoses, melanoma is the most lethal
melanoma. In this chapter, we will explore the history, form of skin cancer, killing an estimated 6850 Americans
molecular biomarkers, and revolutionary studies that led in 2020 [1]. When detected early, melanoma is highly cur-
to the new era of immunotherapy and targeted therapy for able with surgery and has favorable outcomes. However,
melanoma. Furthermore, melanoma treatment in the adju- with more advanced stages of the disease, surgery alone is
vant and neoadjuvant setting is reviewed. Lastly, we not curative in many cases, and additional therapy is needed.
briefly examine therapy options for patients with Merkel While in the past, therapy for melanoma was limited to che-
cell carcinoma. motherapy and cytokine immunotherapy with only limited
success, recent advances in targeted therapy and immuno-
therapy have revolutionized treatments for melanoma, and
Learning Objectives
we will review these advances in the following chapter.
By the end of this chapter, learners will be able to:

• Discuss the history of immunotherapy and studies 67.2 Principles of Immunotherapy

leading to its use.
• Define the underlying mechanisms that drive Modern immunotherapy owes much to William B. Coley, a
immunotherapy. surgeon, who observed that a patient with lymphoid sarcoma
• Identify key biomarkers that indicate viable treat- developed a remission following a postoperative wound
ment options and tumor responsiveness. infection. In 1893, he treated his first cancer patient with a
• Differentiate between the pros and cons of various cocktail of heat-killed bacterial toxins called “Coley’s tox-
therapy options. ins” [2]. He subsequently went on to treat scores of cancer
• Appraise the capability of treatment in the adjuvant cases and published his experience with this approach. In
and neoadjuvant setting. subsequent decades, advances in chemotherapy and targeted
therapy drove significant advances in the treatment of cancer,
but in recent years, advances in immunotherapy have driven
gains in treatment.
In the 1970s, researchers recognized the role of T cells,
M. Chow · E. A. Sangalang · C. Chow · A. I. Daud (*) a type of white blood cell which helps the body target its
Department of Medicine, Division of Hematology Oncology, immune response to specific pathogens, in combating can-
University of California, San Francisco, San Francisco, CA, USA cer. An important cytokine driving the growth of T cells was
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 709
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_67
710 M. Chow et al.

interleukin-2 (IL-2) which was cloned in 1982. Between 67.3 Tumor Immunity
1985 and 1993, a series of clinical trials were completed
that assessed the efficacy of this newly discovered protein Immune checkpoint inhibitors (ICIs) have revolutionized the
as a way to fight melanoma [3]. A total of 270 patients were treatment of advanced melanoma. These agents block nega-
entered into eight clinical trials, and patients received high-­ tive regulators of T-cell priming or T-cell activity in the tumor
dose bolus IL-2. The combined analysis of all patients treated microenvironment [13]. Currently, it is thought that in many
showed the overall response rate (ORR) in patients with mel- cases, the presence of cancer in the body triggers the immune
anoma to be 16% (95% confidence interval [CI], 12–21%) system. Tumors release pathogen-associated molecular patterns
where 6% of patients obtained a complete response (CR) and (PAMPs) and danger-associated molecular patterns (DAMPs),
10% of patients had a partial response (PR). The median pro- which are sensed by and stimulate dendritic cells (DCs) which
gression-free survival (PFS) for all patients with a response then present tumor antigens to T cells usually within the drain-
was more than 1 year. High-dose IL-2 was associated with ing lymph node with co-stimulation. T cells encounter DCs
severe adverse events including capillary leak syndrome and and form an immunological synapse which leads to activation
the effects of secondary cytokine release [4]. Adverse effects and traffic into the tumor to release cytokines and granzymes
(AEs) seen in patients who received IL-2 included hypoten- and kill tumor cells. Tumor cells in some cases can produce
sion, nausea and vomiting, diarrhea, confusion, shortness programmed death ligand 1 (PD-L1) which can bind to pro-
of breath, pulmonary edema, abnormal liver function tests, grammed death 1 (PD-1) on the T cell and inhibit T-cell func-
renal failure, pancytopenia, rash, fever, chills, malaise, and tion, thus evading the body’s immune system.
infections [5]. Nonetheless, the durable CRs and PRs seen in The tumor microenvironment (TME) is used to designate
patients on IL-2 therapy lead to its FDA approval as the first the local environment surrounding a tumor which includes
immunotherapy treatment for metastatic melanoma in 1998. stromal cells, blood vessels, tumor cells, and a variety of
Another major advance in melanoma treatment was infiltrating and resident inflammatory cells. Immune cells
interferon-­alpha (IFN-α) which is a member of a class of found within the TME include tumor-infiltrating lympho-
proteins called cytokines which trigger the activation of the cytes (TILs), myeloid-derived suppressor cells, tumor-­
innate and adaptive immune system [6]. Its role in immune associated macrophages, T cells, B cells, and dendritic cells.
activation led to its utilization in various diseases including Tumors manipulate the microenvironment by promoting
viral infections and in cancer [7]. IFN-α therapy has been growth and immune tolerance, ultimately inducing tumor
tested in a variety of regimens including low dose, intermedi- progression [14–16]. Generally, tumor microenvironments
ate dose, and high dose as an adjuvant treatment for patients can be classified into “hot” or immune-infiltrated TME or
with stage II–III high-risk melanoma. Based on the Eastern “cold” immune desert TME [17]. More sophisticated clas-
Cooperative Oncology Group and intergroup trials El684 and sifications have been proposed such a three-category system:
El694, a significant positive impact on relapse-free survival infiltrated-excluded TME, infiltrated inflamed TME, and
and overall survival (OS) was seen in patients taking the infiltrated tertiary lymphoid structures [18]. The TME helps
IFN-α high-dose regimen [8]. However, the E1690 trial which elucidate how mutational burden, presence of TILs, host fac-
compared low-dose and high-dose interferon to placebo was tors such as obesity and gender, presence of liver metastasis,
negative for OS although some improvement in relapse-free and other factors affect patients’ immune responsiveness.
survival was seen. Based on meta-data analysis pooled from Tumor-infiltrating lymphocytes (TILs) are T cells that
12 randomized clinical trials, significantly longer relapse- invade the tumor environment and actively engage in tumor
free survival was recorded in patients with IFN-α compared destruction. TILs can directly migrate, interact, and prolifer-
with the observation group (hazard ratio [HR], 0.83; 95% CI, ate within the tumor. As a result, TILs have been explored
0.77–0.90; p = 0.000003); however, OS was not improved as a biomarker to predict immunotherapy response [19].
(HR, 0.93; 95% CI, 0.85–1.02; p = 0.1) [8]. Another meta- The presence of checkpoint-positive TIL appears to corre-
data analysis which included 13 trials also demonstrated an late with clinical response to PD-1 blockade. Treatment with
IFN-ɑ effect on recurrence (HR, 0.87; 95% CI, 0.81–0.93; anti-PD-1 resulted in increased activation of the exhausted
p < 0.0001) [9]. A third meta-analysis included 14 random- T-cell population [20]. Their data demonstrates a correlation
ized clinical trials that showed similar results where IFN-α between the subsets of patients who have an abundance of
patients had significant improvement in disease-free survival partially exhausted tumor-infiltrating CD8+ T cells to ICI
rates (HR, 0.82; 95% CI, 0.77–0.87; p < 0.001) [10]. The tox- therapy response. This suggests that patients with a lower
icity profile of high-dose IFN-α includes flu-like symptoms, population of exhausted T cells are less likely to respond
chronic fatigue, myelosuppression, elevated transaminases, to anti-PD-1 monotherapy, and the risk-benefit profile may
and neurologic symptoms [11, 12]. Ultimately, interferon fell point to other treatment options.
out of favor as the effects on survival were not conclusive and Tumor mutation burden (TMB) is a measure of the total
its toxicity was considerable. quantity of mutations within a tumor [21]. A high TMB in
67 Immunotherapy in Melanoma and Merkel Cell Cancer 711

tumors can result in neoantigens that are not subject to self-­ associated antigen 4 (CTLA-4). The discovery of ipilim-
tolerance and enhance immune cell recognition. Snyder et al. umab by James Allison in 1982 and subsequent clinical trials
initially showed that high mutational burden correlates with with ipilimumab started the current era of immunotherapy
a sustained clinical benefit from cytotoxic T-lymphocyte-­ [37]. CTLA-4 functions during the T-cell priming phase
associated antigen 4 (CTLA-4) blockade in patients with by binding B-7 at a higher affinity than CD-28, a T-cell
metastatic melanoma [22]. Rizvi et al. demonstrated that a co-­stimulatory protein, and this displacement of CD-28 by
higher mutation burden was associated with ORR, durable CTLA-4 reduces T-cell activation [38]. This mechanism
clinical benefit, and prolonged PFS in patients with non-­ reduces T-cell priming and limits response to antigens pre-
small cell lung cancer patients treated with anti-PD-1 [23]. sented on antigen-presenting cells. When CTLA-4 protein
The International Cancer Genome Consortium identified is blocked by ipilimumab from interacting with its ligands,
malignant melanoma as the tumor with the highest muta- B7–1 and B7–2, on antigen-presenting cells, the interruption
tion prevalence among 27 histotypes, indicating the geno- of the inhibitory signal leads to the activation of T cells [24].
toxic effects of UV-A exposure are comparable to smoking Phase I and II clinical trials of ipilimumab and tremeli-
signatures [24]. Melanoma patients not coincidentally also mumab (another CTLA-4 antibody) showed a low ORR but
had one of the highest response rates to single-agent PD-1/ durable response rates in patients with responses [39]. In a
PD-L1 inhibition [25, 26]. Nevertheless, additional large- pivotal phase III study comparing ipilimumab or ipilimumab
sized trials are warranted to confirm TMB assessment as a plus gp100 to gp100 vaccine alone, ipilimumab was found to
novel biomarker to aid in patient selection. improve survival in patients with melanoma [40]. The ORR
The intestinal commensal microbiota plays an important role was 10.9% in patients who received ipilimumab alone, com-
in helping digestion and benefitting the immune system. The pared to 5.7% in the ipilimumab plus gp100 regimen and
microbiota is composed of microorganisms, bacteria, fungi, 1.5% in the gp100 alone group. However, ipilimumab was
viruses, protozoa, and their collective genetic material present associated with a high incidence of adverse effects where
in the gastrointestinal tract. Recently, the intestinal commen- 12.9% of patients had severe or fatal AEs. The most common
sal microbiota has also been shown to influence the response immune-related AEs seen in patients taking ipilimumab are
in patients undergoing anti-PD-1 immunotherapy [15, 27–31]. colitis, hepatitis, dermatitis, neuropathy, and endocrinopa-
Wargo and colleagues studied the diversity of oral and gut thy. Based on these results, in 2011, the FDA approved ipili-
microbiomes in melanoma patients and showed a correlation mumab for the treatment of metastatic melanoma.
between diversity of gut flora and response to checkpoint block-
ade (n = 112) [27]. Differences were observed in the patient
gut microbiome of responders (R) versus non-responders (NR); 67.4.2 PD-1 Blockade
analysis of the fecal microbiome samples (n = 43, 30R, 13NR)
revealed significantly higher alpha diversity (P < 0.01) and
The programmed death 1 (PD-1) protein was initially identi-
an abundance of Ruminococcaceae (p < 0.01) in responding fied in 1992 by Honjo and colleagues at the University of
patients. Similar findings have been reported by other groups;
Kyoto [41]. In 2000, Honjo and colleagues identified the
however, this remains an area of active investigation. ligands that interacted with PD-1, programmed death ligand
PD-L1 expressions are a good surrogate for immune infil-
1 (PD-L1). In 2002, Freeman et al. determined that blocking
tration and have been used as a biomarker for PD-1 response
the PD-1 interaction with its ligand stimulated the immune
[32–35]. While PD-L1 expression is highly correlated with system and thus created a powerful anti-tumor response.
PD-1 inhibitor response in patients with melanoma, it has been
Based on these groundbreaking findings, a fully human anti-­
observed that a lack of PD-L1 expression does not ensure aPD-­1 antibody was developed in 2005 and studied in patients.
lack of response and vice versa high-level expression of PD-L1
PD-1 has emerged as a key immune regulator acting as a
does not ensure response [32]. When T cells attack tumors,checkpoint restraining T-cell-mediated autoimmunity [42].
interferon-𝜸 (IFN-𝜸) is also produced and secreted into the
When the PD-L1 ligand on tumor or infiltrating immune cells
TME and can be used as a marker for immune response [36]. engages with the PD-1 protein on activated T cells, apoptosis
in regulatory T cells is reduced and thus, continues to sup-
press immune response. Tumor cells are able to evade T-cell
67.4 Immune Checkpoint Therapy destruction through exploitation of this pathway.
in Melanoma Immune checkpoint inhibitors block the interaction
between cancer cells and the immune system with the use
67.4.1 CTLA-4 Blockade of PD-1 antibodies (Fig. 67.1). Other pathways targeted
for similar immune checkpoint inhibition include cytotoxic
Ipilimumab is a humanized monoclonal antibody to a T-cell T-lymphocyte-associated protein 4, lymphocyte-activation
surface checkpoint known as cytotoxic T-lymphocyte-­ gene 3, and killer-cell immunoglobulin-like receptor [43].
712 M. Chow et al.

a Activated T-Cell

Lymph nodes
Perforin
Granzyme
Tumor

Melanoma T-Cell Clonal

Expression

Tumor Antigen

Dendritic Cell Resting T-Cell

b Adaptive Resistance to T-Cell Killing

Inhibatory Signal

1
PD-L
PD-1

MHC
TCR

Tumor cell
Antigen
Inactive T Cell

Anti PD-1

Cytotoxic Granules

Tumor cell death

Active T Cell

Fig. 67.1 (a) Evasion of PD-1 immune checkpoint through upregu- mechanism of T-cells by antigen-presenting cell ligand CD-28, a T-cell
lated expression of PD-L1 receptor by tumor cell versus activation of co-stimulatory protein (Source: BioRender.com)
T-cell by immune checkpoint inhibitor PD-L1 antibody. (b) Activation
67 Immunotherapy in Melanoma and Merkel Cell Cancer 713

67.4.3 Pembrolizumab 67.4.5 Ipilimumab/Nivolumab

In 2015, pembrolizumab, an anti-PD-1 blocking antibody, The combination of ipilimumab and nivolumab has been
gained FDA approval for the treatment of patients with explored in several trials for melanoma. CTLA-4 functions
metastatic melanoma [44]. This approval was based on two during the priming phase of immune response, while in con-
studies comparing pembrolizumab with ipilimumab or inves- trast, PD-1 works to maintain peripheral tolerance [46]. Tar-
tigator’s choice chemotherapy. In the KEYNOTE-006 study, geting both immune system pathways using anti-CTLA-4
834 ipilimumab-naive metastatic melanoma patients were and anti-PD-1 antibodies to stimulate a more powerful
equally randomized to pembrolizumab 10 mg/kg intrave- immune response was the rationale for combination therapy.
nously every 2 or 3 weeks or up to four doses of ipilimumab In CheckMate 067, a phase III study, 945 treatment-naive
3 mg/kg intravenously every 3 weeks. In the KN002 study, patients were randomized (1:1:1) to receive nivolumab 3 mg/
540 metastatic melanoma patients previously treated with kg intravenously every 2 weeks; nivolumab 1 mg/kg intra-
ipilimumab were randomized to pembrolizumab 2 or 10 mg/ venously every 3 weeks plus ipilimumab 3 mg/kg intrave-
kg intravenously every 3 weeks or investigator’s choice of nously every 3 weeks for four doses, followed by nivolumab
chemotherapy: dacarbazine, temozolomide, carboplatin plus 3 mg/kg every 2 weeks; or ipilimumab 3 mg/kg every
paclitaxel, paclitaxel, or carboplatin. In KN006, patients 3 weeks for four doses [47]. ORRs were higher in the ipi-
receiving pembrolizumab every 2 weeks demonstrated an limumab plus nivolumab arm at 57.6% with 11.5% as a CR
ORR of 34% (HR, 0.63; 95% CI, 0.47–0.83; P < 0.001) and compared to a 19% ORR (2.2% CR) in the ipilimumab arm
33% (HR, 0.69; 95% CI, 0.52–0.90; P = 0.004) in the every and a 43.7% ORR (8.9% CR) in the nivolumab arm. Addi-
3-week dosage schedule compared to 12% in the ipilimumab tionally, at the 60-month follow-up time point, median OS
arm. Similarly improved ORR for pembrolizumab was seen was >60.0 months for the ipilimumab plus nivolumab arm
in the KN002 study where the overall response was 21% in versus 36.9 months and 19.9 months in the nivolumab alone
the pembrolizumab 2 mg/kg arm and 25% in the pembroli- and ipilimumab alone arms, respectively. PFS was also sig-
zumab 10 mg/kg arm versus only 5% in the chemotherapy nificantly improved in the ipilimumab plus nivolumab arm at
arm. The improved overall response and statistically signifi- 11.5 months compared to 2.9 months (99.5% CI, 0.31–0.57;
cant PFS seen in patients taking pembrolizumab led to its p < 0.001) in the ipilimumab arm and 6.9 months (95% CI,
rapid FDA approval. Pembrolizumab was relatively toler- 0.60–0.92) in the nivolumab arm. Unfortunately, more grade
able, with common adverse effects including fatigue, cough, 3+ treatment-­related AEs, 55.0%, were seen in the ipilim-
nausea, rash, itching, decreased appetite, shortness of breath, umab plus nivolumab arm compared to 27.3% in the ipili-
constipation, joint pain, and diarrhea. These AEs typically mumab arm and 16.3% in the nivolumab arm. In 2015, the
resolved when patients were treated with systemic cortico- promising PFS and ORR data led this to become the first
steroids. ever combination immunotherapy treatment approved by the
FDA for metastatic melanoma.

67.4.4 Nivolumab
67.5 Immunotherapy and Targeted
Nivolumab is a human IgG4 monoclonal antibody that Therapy Combinations
blocks PD-1 binding to PD-L1. It was active in phase I
clinical trial in a 2-weekly IV dose [45]. In a pivotal trial in The concept of immunotherapy and targeted therapy com-
treatment-­naïve melanoma, 418 patients were randomized bination has preclinical and some clinical support. This has
to nivolumab or dacarbazine; overall survival, PFS, and been explored in three recent trials.
ORR were all statistically superior for nivolumab. In the KEYNOTE-022, a randomized phase II study, examined
CheckMate 037 trial, the ORR was 32% (95% CI, 23.5– pembrolizumab and BRAF/MEK-targeted therapies dab-
40.8) in nivolumab-treated patients compared to the ORR rafenib and trametinib in patients with BRAF-mutant mela-
of 10.6% in the chemotherapy arm. The most common noma [48]. One hundred twenty patients were randomly
immune-related AEs were pneumonitis, colitis, hepatitis, assigned to either the pembrolizumab 2 mg/kg arm or pla-
hypothyroidism, hyperthyroidism, nephritis, rash, pruri- cebo intravenously every 3 weeks combined with dabrafenib
tus, dermatitis, and vitiligo. Based on this data, the FDA 150 mg orally twice a day and trametinib 2 mg orally once
granted accelerated approval for nivolumab for advanced a day. At 36.6 months of long-term follow-up, the median
melanoma. PFS was improved in the triple combination therapy arm at
714 M. Chow et al.

16.9 months (95% CI, 11.3–27.9) compared to 10.7 months ment-related AEs with 54.7% of patients experiencing AEs
(95% CI, 7.2–16.8) in patients that received placebo, dab- compared to 33.3% in the control arm.
rafenib, and trametinib. The OS data at 24 months was 63.0% Considering all three trials, the results of combination
(95% CI, 49.4–73.9%) for triple combination compared to immunotherapy and targeted therapy in melanoma are dis-
51.7% (95% CI, 38.4–63.4%) in patients taking dabrafenib appointing, and outside of clinical trials or certain specific
and trametinib alone. The ORR was 63% (95% CI, 24.8–8.3) scenarios, currently there is not a role for these combinations
in the triple combination arm, with over 20% achieving a in melanoma.
CR versus 72% (95% CI, 24.8–8.3) in the dabrafenib and
trametinib arm with 15% reaching a CR. The triple combi-
nation arm recorded significantly more grade 3–5 treatment- 67.6 Adjuvant Therapy for Melanoma
related AEs in a total of 35 patients (58%) compared to 15
patients (25%) in the dabrafenib and trametinib arm. Despite 67.6.1 Ipilimumab
the trends toward improved PFS, the trial did not meet its
primary endpoint. Given the success of CTLA-4 blockade in advanced mela-
IMspire150, a multicenter phase III study, randomized 514 noma, there have been extensive efforts to treat patients with
treatment-naive patients with metastatic melanoma to PD-L1 surgically resectable melanoma in the post-surgical (adju-
inhibitor atezolizumab, BRAF inhibitor vemurafenib, and vant) setting or in the pre-surgical (neoadjuvant) setting.
MEK inhibitor cobimetinib (triple regimen) or vemurafenib Given the effectiveness of checkpoint inhibitors in later-­stage
and cobimetinib (control group) [49]. Patients in the control melanoma, these agents have been tested extensively in these
arm received vemurafenib 960 mg orally twice a day with earlier stages as well. An important adjuvant trial with ipili-
cobimetinib 60 mg/d orally on days 1 to 21 of a 4-week cycle. mumab was EORTC 18071, a phase III trial that randomized
The triple regimen arm received atezolizumab 840 mg intra- 951 patients with stage III melanoma to receive ipilimumab
venously on days 1 and 15 of 4-week cycles plus the same 10 mg/kg intravenously (n-475) or placebo (n = 476), [51].
dose of vemurafenib and cobimetinib for cycle 1; during The median RFS was significantly prolonged in the adjuvant
cycle 2, the dose of vemurafenib was lowered to 720 mg. ipilimumab group at 26.1 months (95% CI, 19.3–39.3) com-
Longer median PFS was found in the triple combination arm pared with 17.1 months (95% CI, 13.4–21.6) in the placebo
at 15.1 months compared to 10.6 months in the control arm group (HR, 0.75; p = 0.0013). Additional follow-­up on this
(p = 0.025). The OS was slightly higher (28.8 months) in the trial corroborated the benefit of ipilimumab [52]. In the ipili-
triple combination arm versus 25.1 months in the control. mumab group, the most common grade 3–4 immune-related
Common treatment-related AEs in the triple combination AEs were gastrointestinal (16% vs. <1% in the placebo
arm versus the control were blood creatinine phosphokinase group), hepatic (11% vs. <1%), and endocrine (8% vs. none).
increase (51.3% vs 44.8%, respectively), diarrhea (42.2% AEs led to discontinuation of treatment in 245 (52%) of 471
vs. 46.6%), rash (40.9% in both arms), arthralgia (39.1% vs. patients who started ipilimumab (182 patients discontinued
28.1%), pyrexia (38.7% vs 26.0%), alanine aminotransferase during the initial four doses). Five patients (1%) died due to
increase (33.9% vs. 22.8%), and lipase (32.2% vs. 27.4%). drug-related AEs in the ipilimumab group: three from coli-
Serious treatment-­ related AEs were comparable for both tis, one from myocarditis, and one from multiorgan failure
groups, 33.5% in the triple combination arm versus 28.8% with Guillain-Barré syndrome. Based on the improved RFS,
in the control arm. DMFS, and OS seen in this study, adjuvant ipilimumab for
A third trial was also reported. In COMBI-i, phase III high-risk patients with stage III cutaneous melanoma after
study, 532 untreated stage IIIC or IV patients with melanoma having undergone a complete regional lymph node dissec-
received PD-1 inhibitor, spartalizumab 400 mg intravenously tion was FDA approved in 2015.
every 4 weeks in combination with dabrafenib 150 mg orally
twice a day and trametinib 2 mg orally once a day or placebo
in combination with dabrafenib and trametinib [50]. Cal- 67.6.2 Pembrolizumab
culated PFS for patients placed in the spartalizumab, dab-
rafenib, and trametinib arm was not significantly improved A recent phase III trial, KEYNOTE-054, randomized
compared to the control arm. At 27.2 months of follow-up, patients with complete lymph node dissection stage III cuta-
PFS was 16.2 months in the triple combination arm and neous melanoma to receive pembrolizumab (514 patients) or
12.0 months on the control arm (HR = 0.82; P = 0.042). placebo (505 patients) [53]. At the 3-year median follow-up,
Objective response rates were similar in both study arms, significantly prolonged RFS was seen in the pembrolizumab
68.5% responded in the triple combination arm with 19.9% group (190 RFS events) compared with placebo (283 RFS
CRs versus 64.2% in the control arm with 17.7% CRs. Addi- events) (HR, 0.56; 95% CI, 0.47 to 0.68; P < 0.001) [54].
tionally, the triple combination arm had more grade 3+ treat- Treatment-related AEs occurred in 398 (78.8%) patients in
67 Immunotherapy in Melanoma and Merkel Cell Cancer 715

the pembrolizumab group and in 333 (66.3%) in the placebo identified as a tolerable neoadjuvant dosing regimen. Results
group. Grade 3–5 treatment-related AEs were observed in 74 showed that 63% (95% CI, 44–80) of patients in group A,
(14.5%) patients in the pembrolizumab group and 17 (3.4%) 57% (95% CI, 37–75) of patients in group B, and 35% (95%
in the placebo group. These data lead to FDA approval in CI, 17–56) of patients in group C achieved radiological
2019 [55]. objective response. Based on response and the toxicity pro-
file, group B was determined to be the preferred regimen for
neoadjuvant ipilimumab/nivolumab. The PRADO study is an
67.6.3 Nivolumab extension cohort of the OpACIN-­neo trial and confirmed the
response rate and safety of neoadjuvant ipilimumab 1 mg/kg
In CheckMate 238, 906 patients with completely resected plus nivolumab 3 mg/kg (group B in the OpACIN study) [59].
metastatic melanoma were randomized to receive nivolumab For the total cohort of patients, 70 (71%) achieved a patho-
3 mg/kg intravenously every 2 weeks or ipilimumab 10 mg/ logic response in the index lymph node (95% CI, 0.61–0.79):
kg intravenously every 3 weeks for four doses and then every a CR in 49 patients (50%) of those patients and near patho-
12 weeks starting at week 24 [56]. Most notable, at 4 years, logic CR in 11 (11%). At week 12 of follow-up, there was
the patients receiving nivolumab had a RFS rate of 51.7% a 22% rate of grade 3 to 4 immune-related adverse events,
compared to 41.2% in the ipilimumab arm (HR, 0.71; 95% commonly liver enzyme elevation and colitis. Surgery-related
CI, 0.60–0.86; p = 0.0003). Treatment-related grade 3–4 AEs adverse events of any grade were observed in 54% patients.
were reported in 14.4% of patients receiving nivolumab and
45.9% of those receiving ipilimumab [57]. The results from
this trial led to the rapid FDA approval for nivolumab for 67.8 Merkel Cell Carcinoma
patients with metastatic melanoma in 2017.
Merkel cell carcinoma (MCC) is a rare, aggressive cutaneous
neuroendocrine neoplasm with a high risk of recurrence and
67.7 Neoadjuvant Therapy death. MCC can be caused by integration of the Merkel cell
polyomavirus (MCPyV) [64]. Recently, PD-1 blockade has
Neoadjuvant therapy is now being explored as a promising been shown to be extremely effective in MCC [65]. Infec-
treatment option for high-risk resectable melanoma [58]. tion by MCPyV can lead to MCC in immunosuppressed
It is possible that T-cell checkpoint blockade given neoad- individuals.
juvantly may induce a stronger and broader tumor-specific Prior to the advent of PD-1 blockade, cytotoxic chemo-
T-cell response as compared to adjuvant therapy due to the therapy regimens were used, but these rarely lead to durable
presence of tumor [59]. Additionally, neoadjuvant therapy responses [66]. Lipson et al. studied 67 MCC specimens
may assist in surgical resectability [60]. We summarize some and analyzed MCC PD-L1 expression and its association
recent neoadjuvant trials of ICI. with TILs and showed that PD-L1 expression was observed
Huang et al. conducted a single-arm trial of neoadju- in 49% of patients [67]. Significant associations were also
vant pembrolizumab in melanoma and showed that ≈30% observed between the presence of MCPyV DNA, a brisk
of patients experienced a complete or major pathologic inflammatory response, and tumor cell PD-L1 expression.
response [61]. Comparing neoadjuvant nivolumab to ipili-
mumab/nivolumab, Amaria et al. showed that the combina-
tion yielded high response rates (ORR, 73%; pCR, 45%) but 67.8.1 Avelumab
substantial toxicity (73% grade 3 treatment-related adverse
events (trAEs)), whereas nivolumab monotherapy yielded Avelumab is a PD-L1 antibody that was initially explored in
modest responses (ORR, 25%; pCR, 25%) and low toxicity MCC. In a single-group, phase II trial, 88 patients with stage
(8% grade 3 trAEs) [62]. In the OpACIN-neo phase II trial, IV chemotherapy refractory MCC were enrolled to receive
stage III melanoma patients with resectable lymph nodes avelumab 10 mg/kg intravenously every 2 weeks [68]. The
were randomly assigned to one of three neoadjuvant treat- estimated proportion of responders with ≥1 year duration of
ment arms (group A, ipilimumab plus nivolumab; group B, response was 74% (95% CI, 53–87). Estimated 1-year PFS
two cycles of ipilimumab 1 mg/kg plus nivolumab 3 mg/kg; rate was 30% (95% CI, 21–41), and 1 year OS rate was 52%
and group C, two cycles of ipilimumab 3 mg/kg once fol- (95% CI, 41–62). Median OS was 12.9 months (95% CI,
lowed by two cycles of nivolumab 3 mg/kg) [63]. Grade 3–4 7.5–not estimable). There were no grade 4 or 5 treatment-­
immune-related AEs were observed in 12 (40%) of 30 patients related AEs. MCC patients taking avelumab were seen to
in group A, six (20%) of 30 in group B, and 13 (50%) of 26 in have durable responses beyond 1 year, and the treatment was
group C. None of the grade 3–4 AEs were observed in more well tolerated. Based on these results, avelumab became the
than one patient in the group B treatment arm, which was first FDA-approved treatment for MCC in 2017.
716 M. Chow et al.

67.8.2 Pembrolizumab 4. Curti B, Daniels GA, McDermott DF, et al. Improved survival and
tumor control with Interleukin-2 is associated with the development
of immune-related adverse events: data from the PROCLAIMSM
In the single-arm phase II KEYNOTE-017 study, 50 patients registry. J Immunother Cancer. 2017;5:102.
with recurrent locally advanced metastatic MCC received 5. Pachella LA, Madsen LT, Dains JE. The toxicity and benefit of vari-
pembrolizumab 2 mg/kg intravenously every 3 weeks [69]. ous dosing strategies for Interleukin-2 in metastatic melanoma and
renal cell carcinoma. J Adv Pract Oncol. 2015;6:212–21.
The ORR was 56% (95% CI, 41–70) with 24% CRs. The 6. Ivashkiv LB, Donlin LT. Regulation of type I interferon responses.
median response duration was not reached; however, among Nat Rev Immunol. 2014;14:36–49.
the 28 patients with responses, 96% had response durations 7. Minn AJ, Wherry EJ. Combination cancer therapies with immune
of greater than 6 months. The efficacy and safety informa- checkpoint blockade: convergence on interferon signaling. Cell.
2016;165:272–5.
tion provided in this study led the FDA to grant accelerated 8. Tarhini AA, Gogas H, Kirkwood JM. IFN-α in the treatment of
approval to pembrolizumab for advanced MCC in 2018. melanoma. J Immunol. 2012;189:3789–93.
Based on the data above, PD-1 and PD-L1 blockade is the 9. Wheatley K, Ives N, Eggermont A, et al. Interferon-α as adjuvant
most commonly employed therapy for MCC. Adjuvant trials therapy for melanoma: an individual patient data meta-analysis of
randomised trials. J Clin Oncol. 2007;25:8526.
in MCC are ongoing. 10. Mocellin S, Pasquali S, Rossi CR, Nitti D. Interferon alpha adju-
vant therapy in patients with high-risk melanoma: a systematic
review and meta-analysis. J Natl Cancer Inst. 2010;102:493–501.
67.9 Conclusion 11. Sleijfer S, Bannink M, Van Gool AR, et al. Side effects of
interferon-α therapy. Pharm World Sci. 2005;27:423.
12. Kirkwood JM, Bender C, Agarwala S, et al. Mechanisms and man-
Over the past few decades, the emergence of melanoma and agement of toxicities associated with high-dose interferon alfa-2b
Merkel cell carcinoma treatments has been fruitful in achiev- therapy. J Clin Oncol. 2002;20:3703–18.
ing remission and quality of life for cancer patients. Although 13. Ribas A, Wolchok JD. Cancer immunotherapy using checkpoint
blockade. Science. 2018;359:1350–5.
many of these approaches spark optimism into the field as a 14. Gajewski TF, Corrales L, Williams J, Horton B, Sivan A, Spranger
whole, much remains undiscovered. Continuation of the S. Cancer immunotherapy targets based on understanding the T
development in melanoma and Merkel cell carcinoma treat- cell-inflamed versus non-T cell-inflamed tumor microenvironment.
ments offers a potentially rich area of future research. Adv Exp Med Biol. 2017;1036:19–31.
15. Spranger S, Sivan A, Corrales L, Gajewski TF. Tumor and host fac-
tors controlling antitumor immunity and efficacy of cancer immu-
notherapy. Adv Immunol. 2016;130:75–93.
Perspective 16. Strausberg RL. Tumor microenvironments, the immune system and
cancer survival. Genome Biol. 2005;6:211.
• When should immunotherapy be discontinued in a 17. Mujal AM, Krummel MF. Immunity as a continuum of archetypes.
patient who is having a partial or complete response? Science. 2019;364:28–9.
• Is neoadjuvant immunotherapy followed by surgery 18. Binnewies M, Roberts EW, Kersten K, et al. Understanding the
more advantageous than surgery followed by adju- tumor immune microenvironment (TIME) for effective therapy. Nat
Med. 2018;24:541–50.
vant therapy? 19. Tumeh PC, Harview CL, Yearley JH, et al. PD-1 blockade induces
• How should targeted therapy versus immunother- responses by inhibiting adaptive immune resistance. Nature.
apy be decided for each individual patient? 2014;515:568–71.
• In what cases should triple combination therapy be 20. Daud AI, Loo K, Pauli ML, et al. Tumor immune profiling predicts
response to anti-PD-1 therapy in human melanoma. J Clin Invest.
considered over combination immunotherapy in 2016;126:3447–52.
patients with adverse prognostic factors? 21. Havel JJ, Chowell D, Chan TA. The evolving landscape of bio-
• How might personalized immunotherapy change markers for checkpoint inhibitor immunotherapy. Nat Rev Cancer.
the direction of melanoma treatments? 2019;19:133–50.
22. Snyder A, Makarov V, Merghoub T, et al. Genetic basis for clini-
cal response to CTLA-4 blockade in melanoma. N Engl J Med.
2014;371:2189–99.
23. Rizvi NA, Hellmann MD, Snyder A, et al. Mutational landscape
References determines sensitivity to PD-1 blockade in non–small cell lung can-
cer. Science. 2015;348:124–8.
1. Cancer.org. Melanoma skin cancer statistics. https://www.cancer. 24. Alexandrov LB, Nik-Zainal S, Wedge DC, et al. Signatures of
org/cancer/melanoma-­skin-­cancer/about/key-­statistics.html. In: . mutational processes in human cancer. Nature. 2013;500:415–21.
Accessed 29 Jan 2021. 25. Weber JS, D’Angelo SP, Minor D, et al. Nivolumab versus che-
2. Carlson RD, Flickinger JC, Snook AE. Talkin’ toxins: from Coley’s motherapy in patients with advanced melanoma who progressed
to modern cancer immunotherapy. Toxins. 2020;12:241. after anti-CTLA-4 treatment (CheckMate 037): a randomised, con-
3. Atkins MB, Lotze MT, Dutcher JP, et al. High-dose recombinant trolled, open-label, phase 3 trial. Lancet Oncol. 2015;16:75–384.
interleukin 2 therapy for patients with metastatic melanoma: analy- 26. Robert C, Long GV, Brady B, et al. Nivolumab in previously
sis of 270 patients treated between 1985 and 1993. J Clin Oncol. untreated melanoma without BRAF mutation. N Engl J Med.
1999;17:2105. 2015;372:320–30.
67 Immunotherapy in Melanoma and Merkel Cell Cancer 717

27. Gopalakrishnan V, Spencer CN, Nezi L, et al. Gut microbiome 42. Okazaki T, Chikuma S, Iwai Y, Fagarasan S, Honjo T. A rheo-
modulates response to anti–PD-1 immunotherapy in melanoma stat for immune responses: the unique properties of PD-1
patients. Science. 2018;359:97–103. and their advantages for clinical application. Nat Immunol.
28. Sivan A, Corrales L, Hubert N, Williams JB, Aquino-Michaels 2013;14:1212–8.
K, Earley ZM, Benyamin FW, Lei YM, Jabri B, Alegre M-L, 43. O’Malley DP, Yang Y, Boisot S, et al. Immunohistochemical detec-
Chang EB, Gajewski TF. Commensal Bifidobacterium promotes tion of PD-L1 among diverse human neoplasms in a reference
antitumor immunity and facilitates anti-PD-L1 efficacy. Science. laboratory: observations based upon 62,896 cases. Mod Pathol.
2015;350:1084–9. 2019;32:929–42.
29. Vétizou M, Pitt JM, Daillère R, Lepage P, Waldschmitt N, Flament 44. Barone A, Hazarika M, Theoret MR, et al. FDA approval summary:
C, Rusakiewicz S, Routy B, Roberti MP, Duong CPM, Poirier-­ pembrolizumab for the treatment of patients with Unresectable or
Colame V, Roux A, Becharef S, Formenti S, Golden E, Cording S, metastatic melanoma. Clin Cancer Res. 2017;23:5661–5.
Eberl G, Schlitzer A, Ginhoux F, Mani S, Yamazaki T, Jacquelot N, 45. Hazarika M, Chuk MK, Theoret MR, et al. U.S. FDA approval
Enot DP, Bérard M, Nigou J, Opolon P, Eggermont A, Woerther P-L, summary: Nivolumab for treatment of Unresectable or metastatic
Chachaty E, Chaput N, Robert C, Mateus C, Kroemer G, Raoult D, melanoma following progression on Ipilimumab. Clin Cancer Res.
Boneca IG, Carbonnel F, Chamaillard M, Zitvogel L. Anticancer 2017;23:3484–8.
immunotherapy by CTLA-4 blockade relies on the gut microbiota. 46. Qin S, Xu L, Yi M, et al. Novel immune checkpoint targets: moving
Science. 2015;350:1079–84. beyond PD-1 and CTLA-4. Mol Cancer. 2019;18:155.
30. Wargo JA, Reddy SM, Reuben A, Sharma P. Monitoring immune 47. Hodi FS, Chiarion-Sileni V, Gonzalez R, et al. Nivolumab plus ipi-
responses in the tumor microenvironment. Curr Opin Immunol. limumab or nivolumab alone versus ipilimumab alone in advanced
2016;41:23–31. melanoma (CheckMate 067): 4-year outcomes of a multicentre,
31. Zitvogel L, Ma Y, Raoult D, Kroemer G, Gajewski TF. The micro- randomised, phase 3 trial. Lancet Oncol. 2018;19:1480–92.
biome in cancer immunotherapy: diagnostic tools and therapeutic 48. Ferrucci PF, Giacomo AMD, Vecchio MD, et al. KEYNOTE-022
strategies. Science. 2018;359:1366–70. part 3: a randomized, double-blind, phase 2 study of pembroli-
32. Daud AI, Wolchok JD, Robert C, Hwu W-J, Weber JS, Ribas A, zumab, dabrafenib, and trametinib in BRAF-mutant melanoma. J
Hodi FS, Joshua AM, Kefford R, Hersey P, Joseph R, Gangadhar Immunother Cancer. 2020;8:1789–801.
TC, Dronca R, Patnaik A, Zarour H, Roach C, Toland G, Lunceford 49. Goodman A. Triplet regimen improves progression-free survival in
JK, Li XN, Emancipator K, Dolled-Filhart M, Kang SP, Ebbinghaus advanced melanoma. In: The ASCO Post; 2020. https://ascopost.
S, Hamid O. Programmed death-ligand 1 expression and response com/issues/may-­25-­2020/triplet-­regimen-­improves-­progression-­
to the anti-programmed death 1 antibody Pembrolizumab in mela- free-­survival-­in-­advanced-­melanoma/ Accessed 30 Jan 2021.
noma. J Clin Oncol. 2016b;34:4102–9. 50. Helwick C. COMBI-i trial: no statistical benefit to adding
33. Lyford-Pike S, Peng S, Young GD, Taube JM, Westra WH, Akpeng Spartalizumab to Dabrafenib/Trametinib in melanoma. In: The
B, Bruno TC, Richmon JD, Wang H, Bishop JA, Chen L, Drake ASCO Post; 2020. https://ascopost.com/issues/october-­25-­2020/
CG, Topalian SL, Pardoll DM, Pai SI. Evidence for a role of the combi-­i-­trial-­no-­statistical-­benefit-­to-­adding-­spartalizumab-­to-­
PD-1:PD-L1 pathway in immune resistance of HPV-associated head dabrafenibtrametinib-­in-­melanoma/ Accessed 30 Jan 2021.
and neck squamous cell carcinoma. Cancer Res. 2013;73:1733–41. 51. Eggermont AMM, Chiarion-Sileni V, Grob J-J, et al. Adjuvant ipi-
34. Taube JM, Anders RA, Young GD, Xu H, Sharma R, McMiller TL, limumab versus placebo after complete resection of high-risk stage
Chen S, Klein AP, Pardoll DM, Topalian SL, Chen L. Colocalization III melanoma (EORTC 18071): a randomised, double-blind, phase
of inflammatory response with B7-h1 expression in human melano- 3 trial. Lancet Oncol. 2015;16:522–30.
cytic lesions supports an adaptive resistance mechanism of immune 52. Eggermont AMM, Chiarion-Sileni V, Grob JJ, et al. Ipilimumab
escape. Sci Transl Med. 2012;4:127ra37. versus placebo after complete resection of stage III melanoma:
35. Taube JM, Klein A, Brahmer JR, Xu H, Pan X, Kim JH, Chen Long-term follow-up results the EORTC 18071 double-blind phase
L, Pardoll DM, Topalian SL, Anders RA. Association of PD-1, 3 randomized trial. JCO. 2019;37:2512.
PD-1 ligands, and other features of the tumor immune microen- 53. Eggermont AMM, Blank CU, Mandala M, et al. Adjuvant
vironment with response to anti-PD-1 therapy. Clin Cancer Res. Pembrolizumab versus placebo in resected stage III melanoma. N
2014;20:5064–74. Engl J Med. 2018;378:1789–801.
36. Ayers M, Lunceford J, Nebozhyn M, Murphy E, Loboda A, 54. Eggermont AMM, Blank CU, Mandala M, et al. Longer fol-
Kaufman DR, Albright A, Cheng JD, Kang SP, Shankaran V, Piha-­ low-up confirms recurrence-free survival benefit of adjuvant
Paul SA, Yearley J, Seiwert TY, Ribas A, McClanahan TK. IFN-γ-­ Pembrolizumab in high-risk stage III melanoma: updated results
related mRNA profile predicts clinical response to PD-1 blockade. from the EORTC 1325-MG/KEYNOTE-054 trial. J Clin Oncol.
J Clin Invest. 2017;127(8):2930–40. 2020;38:3925–36.
37. Abhyankar D. Ipilimumab: a tale of twists and turns. South Asian J 55. FDA (2019) Approves pembrolizumab for adjuvant treatment of mel-
Cancer. 2018;7:193–4. anoma. https://www.fda.gov/drugs/drug-­approvals-­and-­databases/
38. Krummel MF, Allison JP. CD28 and CTLA-4 have oppos- fda-­a pproves-­p embrolizumab-­a djuvant-­t reatment-­m elanoma
ing effects on the response of T cells to stimulation. J Exp Med. Accessed 30 Jan 2021.
1995;182:459–65. 56. Ascierto PA, Vecchio MD, Mandalá M, et al. Adjuvant nivolumab
39. Wolchok JD, Hodi FS, Weber JS, Allison JP, Urba WJ, Robert C, versus ipilimumab in resected stage IIIB–C and stage IV mela-
O’Day SJ, Hoos A, Humphrey R, Berman DM, Lonberg N, Korman noma (CheckMate 238): 4-year results from a multicentre,
AJ. Development of ipilimumab: a novel immunotherapeutic double-­blind, randomised, controlled, phase 3 trial. Lancet Oncol.
approach for the treatment of advanced melanoma. Ann N Y Acad 2020;21:1465–77.
Sci. 2013;1291:1–13. 57. FDA (2017) Grants regular approval to nivolumab for adjuvant
40. Fellner C. Ipilimumab (Yervoy) prolongs survival in advanced mel- treatment of melanoma. https://www.fda.gov/drugs/resources-­
anoma. P T. 2012;37:503–30. information-­a pproved-­d rugs/fda-­g rants-­r egular-­a pproval-­
41. Kudo M. Scientific rationale for combination immunotherapy of nivolumab-­adjuvant-­treatment-­melanoma Accessed 2 Feb 2021.
hepatocellular carcinoma with anti-PD-1/PD-L1 and anti-CTLA-4 58. Kelly ZR, Gorantla VC, Davar D. The role of Neoadjuvant therapy
antibodies. Liver Cancer. 2019;8:413–26. in melanoma. Curr Oncol Rep. 2020;22:80.
718 M. Chow et al.

59. practiceupdate.com. ASCO 2020: neoadjuvant ipilimumab and 63. Rozeman EA, Menzies AM, van Akkooi ACJ, et al. Identification
nivolumab may spare therapeutic lymph node dissection in stage iii of the optimal combination dosing schedule of neoadjuvant ipi-
melanoma. In: Practice update; 2020. https://www.practiceupdate. limumab plus nivolumab in macroscopic stage III melanoma
com/content/asco-­2020-­neoadjuvant-­ipilimumab-­and-­nivolumab-­ (OpACIN-neo): a multicentre, phase 2, randomised, controlled
may-­s pare-­t herapeutic-­l ymph-­n ode-­d issection-­i n-­s tage-­i ii-­ trial. Lancet Oncol. 2019;20:948–60.
melanoma/101960 Accessed 30 Jan 2021. 64. Feng H, Shuda M, Chang Y, Moore PS. Clonal integration of
60. Khunger A, Buchwald ZS, Lowe M, et al. Neoadjuvant therapy a Polyomavirus in human Merkel cell carcinoma. Science.
of locally/regionally advanced melanoma. Ther Adv Med Oncol. 2008;319:1096–100.
2019;11:1–19. 65. Van Akkooi, Alexander Christopher, Paolo Antonio Ascierto, Paul
61. Huang AC, Orlowski RJ, Xu X, Mick R, George SM, Yan PK, D. Nathan, Paul Nghiem, Philip Reimann, Frank Hermann, and
Manne S, Kraya AA, Wubbenhorst B, Dorfman L, D’Andrea K, Jürgen C. Becker. “A Phase II, Open-Label Study to Investigate
Wenz BM, Liu S, Chilukuri L, Kozlov A, Carberry M, Giles L, Kier the Efficacy and Safety of Domatinostat in Combination with
MW, Quagliarello F, McGettigan S, Kreider K, Annamalai L, Zhao Avelumab in Patients with Advanced Unresectable/Metastatic
Q, Mogg R, Xu W, Blumenschein WM, Yearley JH, Linette GP, Merkel Cell Carcinoma Progressing on Anti-PD-(L)1 Antibody
Amaravadi RK, Schuchter LM, Herati RS, Bengsch B, Nathanson Therapy: The Merklin 2 Study.” J Clin Oncol. 2021;31:S767–S768.
KL, Farwell MD, Karakousis GC, Wherry EJ, Mitchell TC. A sin- 66. Kaufman Howard L., et al. “Abstract CT079: Durable Responses
gle dose of neoadjuvant PD-1 blockade predicts clinical outcomes to Avelumab (Anti-PD-L1) in Patients with Merkel Cell Carcinoma
in resectable melanoma. Nat Med. 2019;25:454–61. https://doi. Progressed after Chemotherapy: 1-Year Efficacy Update.” Clinical
org/10.1038/s41591-­019-­0357-­y. Trials, 2017;77:13.
62. Amaria RN, Reddy SM, Tawbi HA, Davies MA, Ross MI, Glitza 67. Lipson EJ, Vincent JG, Loyo M, et al. PD-L1 expression in the
IC, Cormier JN, Lewis C, Hwu W-J, Hanna E, Diab A, Wong Merkel cell carcinoma microenvironment: association with inflam-
MK, Royal R, Gross N, Weber R, Lai SY, Ehlers R, Blando mation, Merkel cell polyomavirus and overall survival. Cancer
J, Milton DR, Woodman S, Kageyama R, Wells DK, Hwu P, Immunol Res. 2013;1:54–63.
Patel SP, Lucci A, Hessel A, Lee JE, Gershenwald J, Simpson 68. cancer.gov (2017) Avelumab first approved drug for merkel cell
L, Burton EM, Posada L, Haydu L, Wang L, Zhang S, Lazar carcinoma National Cancer Institute. https://www.cancer.gov/
AJ, Hudgens CW, Gopalakrishnan V, Reuben A, Andrews MC, news-­events/cancer-­currents-­blog/2017/avelumab-­fda-­merkel-­cell.
Spencer CN, Prieto V, Sharma P, Allison J, Tetzlaff MT, Wargo Accessed 30 Jan 2021.
JA. Neoadjuvant immune checkpoint blockade in high-risk 69. Bradford D, Demko S, Jin S, et al. FDA accelerated approval of
resectable melanoma. Nat Med. 2018;24:1649–54. https://doi. Pembrolizumab for recurrent locally advanced or metastatic Merkel
org/10.1038/s41591-­018-­0197-­1. cell carcinoma. Oncologist. 2020;25:e1077–82.
 Landscape of Immunotherapy in
Lung Cancer 68
Nirali Sanghavi, Umme Farwa, Faisal Khurshid,
and Hatim Husain

Abstract
• The importance of biomarkers and different bio-
Lung cancer is the leading cause of cancer-related mortal- markers that have been studied so far in relation to
ity not only in the USA but also worldwide. There have NSCLC.
been many advances and research in treatment protocols • Strategies used when prescribing immunotherapeu-
for non-small cell lung cancer in the last few years. tic drugs as monotherapy or in combination
Screening with low-dose CT scan for smokers aged therapy.
≥65 years has offered a mortality benefit of 20% and a • Recent advances with combinations of immuno-
6.7% decrease in all-cause mortality. The changing land- therapies.
scape of treatment from platinum compounds to tyrosine
kinase inhibitors to immune checkpoint inhibitors has
dramatically helped in increasing the survival of those
high-risk individuals. Different trials that have been car-
ried out to study the efficacy and effectiveness of different 68.1 Introduction
drugs give us a different perspective in determining the
most appropriate treatment for an individual patient. The Lung cancer according to the World Health Organization
checkpoint inhibitors have become the first-line treatment (WHO) is the leading cause of cancer-related mortality
option for patients with NSCLC as monotherapy or in worldwide with 1.5 million deaths each year [1]. Industrial-
combination with chemotherapeutic drugs. The role of ized countries such as Europe and the USA have the high-
biomarkers such as PD-L1 in tailoring the treatment for est mortality rates. The most common type of lung cancer is
NSCLC has been very crucial. However, there is still a non-small cell lung cancer that accounts for 85% of all cases.
need to study different biomarkers and the correct combi- Non-small cell lung cancer (NSCLC) is the primary cause
nation and sequence of immune checkpoint inhibitors to of cancer death worldwide. Among men, lung cancer is the
be administered. leading cause of cancer death for those aged 40 years or older.
While cigarette smoking is the leading cause of lung cancer,
other risk factors include secondhand or passive smoking,
alcohol, crystalline silica, chrysotile asbestos, and uranium.
Learning Objectives
It is suggested that primary causes of recurrence and death
• Trials related to immune checkpoint inhibitors have
in patients with lung cancer are invasion and metastasis of
given promising results when compared to chemo-
tumors [2]. Previous approaches to use nonspecific cytotoxic
therapy drugs and a favorable toxicity profile.
chemotherapy for lung cancer may be associated with severe
• The mechanism of action of different immune
adverse effects and failure to resolve metastatic or subclini-
checkpoint inhibitors.
cal disease [3]. Almost 75% of all the patients with NSCLC
are diagnosed with the disease at an advanced stage [4]. His-
torically these patients typically have 4–5 months of median
survival time after diagnosis with only 10% surviving for
1 year [5].
New approvals for immunotherapy have improved out-
N. Sanghavi · U. Farwa · F. Khurshid · H. Husain (*) comes for patients with metastatic lung cancer. The checkpoints
University of California San Diego, La Jolla, CA, USA

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 719
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_68
720 N. Sanghavi et al.

of PD-1, PD-L1, and CTLA-4 have been particularly important T-lymphocyte-associated antigen 4 (CTLA-4, also known as
with several others being currently explored. Pembrolizumab, CD152) and programmed cell death protein 1 (PD-­1, also
atezolizumab, nivolumab, ipilimumab, and tremelimumab are known as CD279). They are both inhibitory receptors and
all inhibitory immune checkpoint inhibitors that are capable regulate immune responses at different levels and by differ-
of binding and inactivating the effects of the above-mentioned ent mechanisms. By altering this equilibrium, malignancies
proteins. This class of drugs has demonstrated positive results can escape immune surveillance and thrive.
in the treatment of many cancers and has been studied for use
in NSCLC. This chapter will focus on the role and evidence
of immune response in NSCLC pathology, the mechanisms of 68.3.1 PD-1/PD-L1 Pathway
immune checkpoint inhibitors, several clinical trials, and their
results with respect to efficacy and safety. PD-1 is expressed by activated T cells and is a key immune
checkpoint receptor that mediates immunosuppression. It
attenuates T-cell receptor signaling by binding to its ligands
68.2 Chemotherapy as a Historical PD-L1 (B7-H1) and PD-L2 (B7-DC). PD-L1 expression on
Standard tumor and tumor-infiltrating immune cells is thought to be pri-
marily induced by Th1 cytokines (in particular, interferon-γ)
In 1978, cisplatin was approved by the US Food and Drug released by infiltrating tumor immune cells. Clinical trials
Administration (FDA) as the first platinum compound for use. have produced promising results evaluating antagonist anti-
Since then several platinum drugs have entered clinical trials, bodies that target the coinhibitory immune checkpoint, PD-1
with carboplatin and cisplatin approved for chemotherapy expressed on activated antitumor T cells, and its primary
for lung cancer worldwide. Platinum drugs are used to treat ligand, PD-L1, expressed on tumor cells and immune cells
advanced non-small cell lung cancer (NSCLC) without muta- within the tumor microenvironment (Fig. 68.1). Targeting this
tions that were sensitive to targeted therapy for the past two pathway can be categorized as directed against PD-1 such as
decades as the standard of care and as first-line treatment [6]. nivolumab and pembrolizumab or targeted against PD-L1,
Platinum drugs work by targeting nuclear DNA, resulting such as atezolizumab and durvalumab [9–11].
in a broad spectrum of both anti-cancer and adverse effects.
These agents form adducts preferentially with the N7 atom on
guanine and adenosine bases. This binding stops DNA repli- 68.3.2 Anti-CTLA-4 Pathway
cation and transcription, which then initiates cellular apop-
tosis [7]. Platinum compounds attack indiscriminately on all CTLA-4 was the first immune checkpoint receptor to be clin-
rapidly dividing cells leading to severe side effects and induc- ically targeted and is expressed exclusively on T cells where it
ing the ability of cancers to develop drug resistance. Common primarily regulates the amplitude of the early stages of T-cell
side effects of cisplatin and carboplatin include nausea and activation. Primarily, CTLA-4 counteracts the activity of the
vomiting, myelosuppression, neuropathy, ototoxicity, hepato- T-cell co-stimulatory receptor, CD28. CTLA-4 expression
toxicity, and nephrotoxicity. Cisplatin is a highly emetogenic on the surface of T cells dampens the activation of T cells
agent, with considerable variability between individuals. by outcompeting CD28 in binding CD80 and CD86, as well
Another treatment option is the combination of the taxane as actively delivering inhibitory signals to the T cell [12].
docetaxel and ramucirumab, a recombinant human immuno- CTLA-4 has also been shown to regulate tumor immunity
globulin G1 monoclonal antibody inhibiting vascular endo- through regulatory T cells (Tregs), which express high lev-
thelial growth factor receptor-2 in patients with stage IV els of surface CTLA-4, and therefore suppress activation and
NSCLC as per the results of REVEL trial [8]. This has been expansion of effector cells. Tregs are upregulated on tumor
approved after first-line platinum-based chemotherapy and cells, and blocking Treg function through anti-­CTLA-­4 anti-
clinical trials with checkpoint inhibitors have compared with bodies may be an effective mechanism to enhance antitu-
docetaxel in the second line. mor activity. Ipilimumab is a fully human immunoglobulin
G1 anti-CTLA-4 monoclonal antibody, now approved for
melanoma, RCC, HCC, NSCLC, CRC, and mesothelioma.
68.3  echanism of Immune Checkpoint
M Tremelimumab is a fully human IgG2 anti-CTLA-4 antibody
Inhibitors also under investigation (Fig. 68.2).

Antigen recognition by the T-cell receptor (TCR) is regulated

by a balance between co-stimulatory and inhibitory signals 68.3.3 TIGIT Pathway
which are called immune checkpoints. The two immune
checkpoint receptors that have been most actively studied in TIGIT is an immune checkpoint that inhibits the activa-
the context of clinical cancer immunotherapy are cytotoxic tion of both T cells and NK cells. The TIGIT receptor has
68 Landscape of Immunotherapy in Lung Cancer 721

Activatory
co-stimulation

CD80 CD28

PD-1 PD-L1
pMHC TCR
Ag Tumor
APC T CELL
PD-1 PD-L1

PD-L1 PD-1
Inhibitory
interactions

Fig. 68.1 The PD-1/PD-L1pathway. PD-1 expressed on activated T found on antigen-presenting cells (dendritic cells) inhibits the T-cell
cells interacts with its ligands PD-L1 and PD-L2 when an immune function by binding to PD-1. MHC, major histocompatibility complex;
response is triggered. The above figure depicts that the PD-L1 ligand TCR, T-cell receptor

Inhibitory
interactions

B7 CTLA-4

MHC TCR
Ag
Ag
APC T CELL

B7 CD28

Activatory
co-stimulation

Fig. 68.2 The cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4) cell. CTLA-4 translocates to the plasma membrane of the T cells which
pathway. T cell is activated when there is co-stimulatory signal from subsequently inhibits activation of the T cells. MHC, major histocom-
interaction between B7 on antigen presentation cells and CD28 on a T patibility complex; TCR, T-cell receptor

a transmembrane IgV domain. It is upregulated in tumor rates with existing immunotherapies [14]. In a phase 2
antigen-­specific CD8+ T cells and CD8+ TILs involved in CITYSCAPE trial, tiragolumab, anti-TIGIT antibody, plus
adaptive immune tumor surveillance. TIGIT shares its two atezolizumab was compared to placebo plus atezolizumab
ligands, CD155 (PVR) and CD112 (PVRL2, nectin-2), with as first-line treatment in NSCLC. The combination of
the co-­stimulatory molecule CD226, and together they com- tiragolumab and atezolizumab had better ORR (31.3% vs
prise a pathway that strongly resembles the well-known B7/ 16.2%) and PFS (5.4 months vs 3.6 months) as compared to
CD28/CTLA-4 pathway [13]. Similar to CTLA-4, TIGIT can placebo with atezolizumab [15]. These data have supported
inhibit immune responses by delivering inhibitory signals to the initiation of a phase 3 trial in PD-L1 > 50% 1 L NSCLC
effector T and NK cells, by inducing tolerogenic dendritic (SKYSCRAPER-01).
cells, and by enhancing the suppressive capacity of Tregs. In a phase 1 trial, another anti-TIGIT Ab, vibostolimab,
Both TIGIT and PD-1 are also upregulated in a variety of in combination with pembrolizumab was used for meta-
different cancers. Anti-TIGIT antibodies have demonstrated static NSCLC who had not previously received anti-PD-1/
synergy with anti-PD-1/PD-L1 antibodies in preclinical PD-L1 therapy, but the majority of whom had received ≥1
models. Currently, there are multiple first-in-man phase 1 tri- prior lines of therapy (Fig. 68.3). This trial showed that in
als hoping to exploit this new pathway and improve response patients whose tumors expressed PD-L1 ≥ 1% (n = 13),
722 N. Sanghavi et al.

have a 67–100% response rate in contrast to PD-L1-nega-

TUMOR CELL KILLING
tive NSCLC, where the response rate is about 0–15% [21].
PD-L1 may be less useful when certain signaling pathways
are active such as EGFR, MAPK, and PI3K-AKT. These
TIGIT CD112 results were reported in OAK and CheckMate 017 [18]. The
T CELL
or Tumor use of PD-L1 is confounded by multiple unresolved issues:
NK CELL
CD226 CD155 variable detection antibodies, differing IHC cutoffs, tissue
preparation, processing variability, primary versus meta-
static biopsies, oncogenic versus induced PD-L1 expression,
and staining of tumor versus immune cells. Other biomark-
ers, such as gene expression signatures, TMB, and T-cell
receptor sequencing, are currently being evaluated as poten-
IMMUNE CELL ACTIVATION
tial markers of response.
Fig. 68.3 Tiragolumab binds to TIGIT, leaving the CD226 receptor
free to bind to CD155 and CD112 without competition. This blocks
activation of TIGIT’s immunosuppressive function and activates antitu- 68.4.2 TMB and Neoantigen Burden
mor immune cells
Tumor mutation burden is defined as the total number of
the ORR was 46% (95% CI, 19–75), and median PFS was mutations, including insertion/deletion and replacement,
8.4 months (95% CI, 3.9–10.2). In patients with PD-L1 per megabyte of exonic regions of the gene examined in
expression of <1% (n = 12), the ORR was 25% (95% CI, the tumor specimen (mut/Mb). Somatic cell mutations are
6–57), and median PFS was 4.1 months (95% CI, 1.9–not transcribed and expressed to RNA and proteins, respectively,
reached [NR]) [16]. producing neoantigens or peptides. These are called non-­
autoantigens which activate T cells and elicit an immune
response. These tumor cells are targeted by a large num-
68.4 Biomarkers in Immunotherapy ber of tumor-specific T cells in patients with a high TMB
(TMB). The response to anti-PD-1/PD-L1 therapy greatly
Immune surveillance is said to occur in lung cancer where depends on these tumor-specific T cells. Thus, tumors with
the tumor cells express antigens that are recognized by the TMB-H can be more responsive to immunotherapy indicat-
immune system and activate an immune response. These ing that TMB and neoantigens may serve as biomarkers [22].
antigens can either be specific to the neoplasm or antigens In addition, several studies demonstrated the relationship of
associated with the tumor that are also expressed on nor- the mutational landscape of tumors and its responsiveness to
mal cells but have an aberrant or deregulated expression on PD-1 blockade in NSCLC and anti-CTLA-4 therapy in mela-
tumors [17]. PD-L1 and tumor mutational burden (TMB) are noma. These two malignancies often have a high TMB, the
currently predictors of response for this class of immuno- former due to increased carcinogen exposure via cigarette
therapy in lung cancer. smoke and the latter due to exposure to DNA-damaging UV
light [23]. CheckMate 026 revealed that nivolumab resulted
in significantly better ORR (47 vs 28%) and PFS (9.7 vs
68.4.1 Expression of PD-L1 5.8 months) in the TMB-H arm (10 mut/MB cutoff) com-
pared with that of platinum-based chemotherapy [6]. The
PD-L1 expression on tumor cells has been a selection criteria results of CheckMate 227 showed that the PFS of patients
for monotherapy [18]. KEYNOTE-024 proved that patients with TMB-H (≥10 mut/Mb) treated with nivolumab and ipi-
with advanced NSCLC with high PD-L1 expression (Tumor limumab was superior to the chemotherapy. Surprisingly, in
Proportion Score by 22C3 ≥ 50%) had a superior overall sur- the same trial, PFS was improved regardless of PD-L1 levels.
vival (OS) with pembrolizumab compared to chemotherapy Similar results were found in CheckMate 012 and Check-
[19]. However, the results of KEYNOTE-042 suggested Mate 026, where no notable correlation was found between
that the efficacy of immunotherapy was similar to chemo- TMB and PD-L1 expression. This indicates that TMB may
therapy when PD-L1 expression was 1–49%. This indicates be an independent biomarker, but together with PD-L1, it
that if there is a higher expression of PD-L1, the effect of may predict the efficacy of immunotherapy [24]. Pembroli-
immunotherapy is also better [20]. Similarly, it has been zumab has recently received FDA approval for a TMB high
shown that patients with PD-L1 over-expressing NSCLC indication across tumor types.
68 Landscape of Immunotherapy in Lung Cancer 723

68.4.3 Tumor-Infiltrating Lymphocytes (TILs) trial compared the investigator’s choice of platinum doublet
against nivolumab. The primary efficacy analysis showed
Apart from TMB, TILs can also be used to predict the efficacy there was no advantage of treatment with nivolumab in RR
of anti-PD-1/PD-L1 therapy. The PD-1/PD-L1 inhibitors (26% vs 33%), PFS (4.2 vs 5.9 months), HR (1.15; 95% CI,
require lymphocytes near the tumor, i.e., TILs, particularly 0.91–1.45; p = 0.25) [6].
infiltration by CD8 + T cells for their efficacy and favorable Another phase 3 trial CheckMate 057 compared
outcome [25]. The proliferation of CD8 + T cells has been nivolumab against docetaxel in patients with advanced
directly associated with the shrinkage of tumors on imag- non-­ squamous NSCLC. Nivolumab showed a clinically
ing after immune checkpoint inhibitor treatment [26]. This is meaningful survival benefit, with an improved safety pro-
currently under investigation (Fig. 68.4). file compared to docetaxel (27% lower risk of death at a
minimum follow-up of 13.2 months), which persisted with
68.4.3.1 Immunotherapy as First-Line extended follow-up (28% lower risk of death at a minimum
Treatment follow-up of 17.2 months). The frequencies of all-causality,
any-grade adverse events were similar between nivolumab
Immunotherapy Monotherapy and docetaxel, but grade 3–4 adverse events were fewer with
Pembrolizumab was established as the standard of care after nivolumab than docetaxel [3].
the results of the phase 3 KEYNOTE-024 trial. The first-line Durvalumab is a selective high-affinity human IgG1
trial had compared the investigator’s choice of platinum-­ monoclonal antibody that blocks PD-L1 binding to PD-1 and
based chemotherapy with pembrolizumab in patients with CD80 which allows T cells to recognize and kill tumor cells.
NSCLC who had PD-L1 expression of >50% by 22C3 TPS. The PACIFIC trial reported a PFS benefit in patients treated
Pembrolizumab had shown increase in RR (45% vs 28%), with durvalumab versus the placebo arm (16.8 vs 5.6 months;
PFS (10.3 vs 6 months), HR (0.5 0; 95% CI, 0.37–0.68; P < 0.001) and a higher response rate (RR) (28.4% vs 16%;
P < 0.001), and OS (30 vs 14.2 months) [27]. However, P < 0.001). In addition, an OS benefit was observed (hazard
CheckMate 026, a phase 3 trial, studied nivolumab in patients ratio [HR] of death, 0.68; 95% CI, 0.54–0.86). The toler-
with NSCLC with PD-L1 expression of >1% by 28–8. The ability of durvalumab was similar to that in previous studies.
These findings led to the Food and Drug Administration’s
(FDA) approval of durvalumab for unresectable stage III
Immune NSCLC [28].
checkpoint
On the basis of the results of IMpower110, atezolizumab
inhibitors
was approved by FDA as a monotherapy in patients with
PD-L1 expression of ≥50% on TPS by SP142 [29]. The
study showed a statistically significant improvement in OS
Effective Less responsive for patients with high PD-L1 tumor expression receiving
atezolizumab compared to those treated with platinum-
based chemotherapy. The median OS was 20.2 with atezoli-
zumab versus 13.1 months with chemotherapy (HR, 0.59;
95% CI, 0.40–0.89; P = 0.0106) at a median follow-up of
PD-L1 15.7 months [30].
TMB
EGFR/ALK
Most patients with advanced NSCLC do not respond
TILS to PD-1 inhibitor monotherapy, but in a phase 1 study,
JAK1/2
MSI-H/dMMR
DDR
PTEN cemiplimab plus radiotherapy (RT) demonstrated antitumor
STK11/LKB1
KRAS mutation activity in pretreated pts. with NSCLC. Overall response
TP53 mutation rate (ORR; complete response [CR] + partial response
[PR]) was 18.2% (0 CR and 6 PRs) with a median dura-
tion of response of 14.9 months (95% CI, 5.5–14.9). Dis-
Fig. 68.4 Overview of biomarkers for immune checkpoint inhibition ease control rate (ORR + stable disease [SD]) was 72.7% (6
in non-small cell lung cancer. ICIs, immune checkpoint inhibitors;
PD-L1, programmed death ligand 1; TMB, tumor mutational burden;
PRs + 18 SDs) [31].
TILs, tumor-infiltrating lymphocytes; MSI-H, microsatellite instability, Also, the EMPOWER-Lung 1, a phase 3 trial, compared
high; dMMR, DNA mismatch repair; DDR, DNA damage repair; cemiplimab monotherapy against platinum-doublet chemo-
KRAS, Kirsten rat sarcoma viral oncogene homolog; TP53, tumor pro- therapy in patients with NSCLC with PD-1 ≥ 50%. The results
tein p53; EGFR/ALK, epidermal growth factor receptor/ALK tyrosine
kinase receptor; JAK1/JAK2, Janus kinase 1/ Janus kinase 2; PTEN,
showed that with median follow-up of 10.8 months, median
phosphate and tension homology deleted on chromosome ten; STK11, OS was not reached with cemiplimab (95% CI, 17.9–not
serine/threonine kinase 11; LKB1, Lkb1 kinase evaluable [NE]) versus 14.2 months (95% CI, 11.2–17.5) with
724 N. Sanghavi et al.

chemotherapy (HR, 0.57; 95% CI, 0.42–0.77; p = 0.0002). In CheckMate 227 two treatment groups were formed
Median PFS was 8.2 months (95% CI, 6.1–8.8) with depending on the expression of PD-L1 [39]. In patients
cemiplimab versus 5.7 months (95% CI, 4.5–6.2) with chemo- with PD-L1-negative tumors (PD-L1 < 1%), a combination
therapy (HR, 0.54; 95% CI, 0.43–0.68; P < 0.0001) [32]. of nivolumab and ipilimumab was compared with the com-
bination of nivolumab with chemotherapy and chemother-
Combination Chemoimmunotherapy apy alone. Overall survival rates at 2 years were 40.4% for
The KEYNOTE-189 for metastatic non-squamous NSCLC nivolumab in combination with ipilimumab and 23% with
suggested that the combination of pembrolizumab with plati- chemotherapy. The results showed that the progression-free
num doublet is more effective than chemotherapy alone [19, interval was significantly higher with the combination of
33]. The ORR (47.6% vs 18.9%) and PFS (8.8 months vs nivolumab with ipilimumab than chemotherapy alone. HR
4.9 months) were significantly improved with the addition for death was 0.62 (95% CI, 0.48 to 0.78). The median dura-
of pembrolizumab. Similarly, KEYNOTE-407 for metastatic tion of survival was significantly better in nivolumab with
squamous NSCLC proved that the combination of pembro- ipilimumab group (17.2 months) versus chemotherapy alone
lizumab and chemotherapy with carboplatin plus paclitaxel (12.2 months).
or nab-paclitaxel resulted in significantly longer overall sur- In patients with ≥1% PD-L1 expression, the combination
vival (15.9 months vs 11.3 months) and progression-free sur- of nivolumab with ipilimumab was compared with nivolumab
vival (6.4 months vs 4.8 months) than chemotherapy alone alone and chemotherapy alone. The trial increased overall
[34] with benefit observed across PDL1 expression levels. survival to 17.1 months in nivolumab with ipilimumab as
In a phase 3 trial, IMpower150, atezolizumab, anti-PD- compared to 14.9 months in the chemotherapy group alone.
­L1 antibody, was studied in patients with non-squamous The overall survival rates with nivolumab and ipilimumab at
NSCLC. The patients were randomized to carboplatin, 1 year 62.6% and 40% at 2 years, while with chemotherapy
paclitaxel, and bevacizumab or carboplatin, paclitaxel, and at 1 year 56.2% and 32.8% at 2 years. The objective response
atezolizumab or the same chemoimmunotherapy plus bevaci- rate was greater with immunotherapy combination (35.9%),
zumab. The results showed improvement in the OS, PFS, and while with chemotherapy it was 30.0%. The median dura-
ORR when compared to chemotherapy alone [35]. Atezoli- tion of response was 23.2 months with nivolumab and ipili-
zumab was approved by the FDA as the first-line agent in mumab and 6.2 months with chemotherapy.
combination with carboplatin, paclitaxel, and ­bevacizumab In addition, the CheckMate 227 also proved that the
in patients with metastatic non-squamous NSCLC with no median overall survival in patients with baseline brain
EGFR or ALK driver mutations. metastasis was improved to 17.4 months with nivolumab
For squamous NSCLC, the IMpower131 trial showed that and ipilimumab versus 13.7 months with chemotherapy
atezolizumab plus chemotherapy achieved a median overall alone [40].
survival of 14.2 months versus 13.5 months seen in patients The CheckMate 9LA, a phase 3 trial, in patients with stage
who received chemotherapy alone [36]. 4 NSCLC, compared nivolumab+ipilimumab+chemotherapy
of two cycles against four cycles of chemotherapy alone. The
Combination Immunotherapy combination of chemotherapy included four drugs: carbopla-
The phase 3 MYSTIC trial compared results of treatment tin, cisplatin, paclitaxel, and pemetrexed. The results showed
with durvalumab against durvalumab and tremelimumab a median OS was longer in the immunotherapy-­containing
combination against chemotherapy alone in patients with arm than in the chemotherapy arm (15.6 vs 10.9 months; HR,
metastatic NSCLC. It did not meet its primary endpoints 0.66). Combination immunotherapy with chemotherapy also
of improved OS with durvalumab versus chemotherapy or improved median PFS (6.7 vs 5.0 months; HR, 0.68) and the
improved OS or PFS with durvalumab plus tremelimumab median duration of response (11.3 vs 5.6 months). The over-
versus chemotherapy in patients with 25% of tumor cells all response rate also favored the immunotherapy-containing
expressing PD-L1. Exploratory analyses identified a TMB arm (38% vs 25%) [41]. In regard to toxicity, grade 3–4
threshold of 20 mutations per megabase for optimal OS ben- treatment-related adverse events were reported to be 47%
efit with durvalumab plus tremelimumab [37]. in nivolumab+ipilimumab+chemotherapy versus 38% in the
In phase 1 of CheckMate 012, chemotherapy-naïve chemotherapy alone arm [15].
patients were treated with different schedules and doses of
nivolumab and anti-CTLA4 antibody ipilimumab. The ORR
was 21% in patients with less than 1% tumoral PD-L1 expres- 68.5 Cellular Therapies
sion, 57% in PD-L1 at least 1%, and 92% in patients with
50% PD-L1. PFS also showed improvement in patients that Another advancement in this field is adoptive cellular ther-
were treated with combination rather than with nivolumab apy which relies on the ex vivo generation of tumor-reactive
alone [38]. immune cells and administration of the cells in large numbers
68 Landscape of Immunotherapy in Lung Cancer 725

in the autologous host. Different forms of natural tumor-­ 2 trial is studying the efficacy, safety, and toxicity of TIL
reactive immune cells have been generated for NSCLC in therapy in patients with NSCLC [41].
clinical trials such as TILs, cytokine-induced killer (CIK),
CD+ T cells, natural killer (NK) cells, and chimeric antigen
receptors (CAR) [42, 43]. 68.5.2 The CAR Therapy

It has a key advantage compared to TIL therapy because the

68.5.1 The TIL Therapy latter targets tumor cells only when their antigens are bound
by the major histocompatibility complex, while CAR therapy
Naturally occurring T cells that have already infiltrated has the ability to bind to cancer cells even if their antigens are
patients’ tumors are first harvested and subsequently acti- not presented on the surface via MHC [44]. The preclinical
vated and expanded. This is followed by re-infusing the large studies have shown that PD-L1 CAR-T cells can be activated
number of previously activated cells into patients, where and have significant killing effects on ­PD-L1-­positive tumor
they then seek out and destroy the tumor. The ongoing phase cells in vitro (Table 68.1). They inhibited tumor growth with

Table 68.1 Results of different trials comparing various chemotherapy and immunotherapy agents
Trial Subset n OS PFS ORR
KEYNOTE trial Carboplatin + pemetrexed + 410 Not reached 8.8 months 47.6%
189 pembrolizumab)
Placebo + pemetrexed-cisplatin/ 206 11.3 months 4.9 months 18.9%
carboplatin
HR, 0.49 (95% CI, 0.38–0.64) HR, 0.52 (95% CI, P < 0.001
P < 0.001 0.43–0.64)
P < 0.001
KEYNOTE-021 Pembrolizumab + pemetrexed-carboplatin Not reached 24.0 months 56.7%
Pemetrexed + carboplatin 21.1 months 9.3 months 30.2%
HR, 0.56 (95% CI, 0.32–0.95) HR, 0.53 (95% CI,
P = 0.0151 0.33–0.86)
P = 0.0049
KEYNOTE-407 Pembrolizumab + carboplatin + paclitaxel/ 278 15.9 months 6.4 months 57.9%
nab-paclitaxel
Carboplatin + paclitaxel/nab –paclitaxel 281 11.3 months 4.8 months 38.4%
HR, 0.64 (95% CI, 0.49–0.85) HR, 0.56 (95% CI,
P < 0.001 0.45–0.70)
P < 0.001
CheckMate 9LA Nivolumab + ipilimumab + Carboplatin, 361 14.13 months 6.83 months 37.7%
paclitaxel, pemetrexed, Cisplatin
Carboplatin, paclitaxel, pemetrexed, 358 10.74 months 4.96 months 25.1%
Cisplatin
HR, 0.69 (95% CI, 0.56–0.86) HR, 0.70 (95% CI,
P < 0.001 0.58–0.84)
P < 0.001
CheckMate 227 Nivolumab + ipilimumab 396 17.1 months 35.9%
PD-L1 >/+ 1%
Chemotherapy alone 397 14.9 months 30%
PD-L1 >/= 1%
Nivolumab + ipilimumab 187 17.2 months 40.4%
PD-L1 < 1%
Chemotherapy alone 186 12.2 months 23%
PD-L1 < 1%
CheckMate 057 Nivolumab 292 12.2 months (95% confidence 19 months (95% CI, 19% (95% CI,
interval [CI], 9.7 to 15.0 14–23) 15 to 24)
Docetaxel 290 9.4 months (95% CI, 8.1 to 8 months (95% CI, 12% (95% CI, 9
10.7) with docetaxel 5–12) to 17)
IMpower130 Atezolizumab + carboplatin + 483 18.6 months 7 months 49%
nab-paclitaxel
Carboplatin + nab-paclitaxel 240 13.9 months 5.5 months 32%
HR, 0.79 (95% CI, 0.64–0.98) HR, 0.64 (95%
P = 0.033 CI,0.54–0.77)
P < 0.0001
(continued)
726 N. Sanghavi et al.

Table 68.1 (continued)

Trial Subset n OS PFS ORR
IMpower131 Atezolizumab + carboplatin + 343 14.2 months 6.5 months
nab-paclitaxel
338 12.6 months 5.6 months
Carboplatin+ nab-paclitaxel 340 13.5 months 5.6 months
HR, 0.96 (95% CI, 0.78–1.18) HR, 0.60 (95% CI,
P = 0.6931 0.60–0.85)
P = 0.0001
IMpower132 Atezolizumab + pemetrexed + cisplatin/ 292 18.1 months 7.6 months 47%
carboplatin
Pemetrexed + cisplatin/carboplatin 286 13.6 months 5.2 months 32%
HR, 0.81 (95% CI, 0.64–1.03) HR, 0.60 (95% CI,
P = 0.0797 0.49–0.72)
P < 0.0001
IMpower133 Atezolizumab +carboplatin-etoposide 201 12.3 months 5.2 months 60.2%
Placebo + carboplatin-etoposide 202 10.3 months 4.3 months 64.4%
HR, 0.70 (95% CI, 0.54–0.91) HR, 0.77 (95% CI,
P = 0.007 0.62–0.96)
P = 0.02
IMpower150 Atezolizumab + carboplatin + paclitaxel 402 19.0 (15.7–21.5) months 8.3 months 49.3%
Atezolizumab + bevacizumab + paclitaxel 400 19.2 (17.0–23.8) months 8.5 (7.7–9.7) 63.5%
+ carboplatin months
Bevacizumab-carboplatin-paclitaxel 400 14.7 (12.9–17.1) months 6.8 (5.8–7.3) 48.0%
months
HR, 0.78 (95% CI, 0.64–0.96) HR, 0.62 (95% CI,
P = 0.02 0.52–0.74)
P < 0.001
OS overall survival, PFS progression-free survival, ORR objective response rate, HR hazard ratio, CI confidence interval

no obvious observed toxicity in mouse xenograft models. In

the ongoing phase 1 clinical trial, investigators are exploring Open Questions
the safety and efficacy of the new CAR-T cells in the treat- • What are the right predictive biomarkers to guide
ment of advanced NSCLC patients [45]. sequence and combination of these novel immune
checkpoint inhibitors?
• What are the evolving trials of targeted molecular
68.6 Conclusion therapies in combinations with immunotherapies?
• What are strategies to anticipate and treat toxicities
Considering all the aspects discussed in this chapter, it is evident with these agents?
that the evolution of immunotherapy has established a new era • How the use of immunotherapy affects the rate of
for the treatment of NSCLC. Recent trials and studies of check- recurrence of early-stage NSCLC?
point inhibitors have shown promising preliminary results in
attaining consequential and durable treatment responses with
manageable toxicity. Identifying patient populations that can
derive the greatest benefit to treatment with immunotherapeutic
References
agents and combinations of therapy remains one of the most 1. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal
important challenging questions to answer. Exploration of novel A. Global cancer statistics, 2012. CA Cancer J Clin. 2015;65(2):87–
checkpoints and cellular-based therapy has recently yielded 108.
encouraging responses that will need to be further tested in clin- 2. Ada G. The coming of age of tumour immunotherapy. Immunol
Cell Biol. 1999;77(2):180–5.
ical trials. Antigen-specific vaccines in combination with immu- 3. Borghaei H, Paz-Ares L, Horn L, Spigel DR, Steins M, Ready NE,
nomodulatory agents may have synergistic effects in augmenting et al. Nivolumab versus docetaxel in advanced nonsquamous non-­
the antitumor immune response. small-­cell lung cancer. N Engl J Med. 2015;373(17):1627–39.
68 Landscape of Immunotherapy in Lung Cancer 727

4. JM TR, Kmieciak M, Manjili MH, Knutson KL. Tumor immunoed- ade in patients with non-small-cell lung cancer profiled with targeted
iting and immunosculpting pathways to cancer progression. Sem next-generation sequencing. J Clin Oncol. 2018;36(7):633–41.
Cancer Biol. 2009;17(4):275–87. 25. Geng Y, Shao Y, He W, Hu W, Xu Y, Chen J, et al. Prognostic role
5. Page DB, Bourla AB, Daniyan A, Naidoo J, Smith E, Smith M, of tumor-infiltrating lymphocytes in lung cancer: a meta-analysis.
et al. Tumor immunology and cancer immunotherapy: summary of Cell Physiol Biochem. 2015;37(4):1560–71.
the 2014 SITC primer. J Immunother Cancer. 2015;3(1):1–10. 26. Tumeh PC, Harview CL, Yearley JH, Shintaku IP, Emma J, Taylor
6. Carbone DP, Reck M, Paz-Ares L, Creelan B, Horn L, Steins M, M, et al. Immune Resistance. Nature. 2015;515(7528):568–71.
et al. First-line nivolumab in stage IV or recurrent non-small-cell 27. Reck M, Rodriguez-Abreu D, Robinson AG, Hui R, Csöszi T, Fülöp
lung cancer. N Engl J Med. 2017;376(25):2415–26. A, et al. Pembrolizumab versus chemotherapy for PD-L1-positive
7. Rossi A, Di Maio M. Platinum-based chemotherapy in advanced non-small-cell lung cancer. N Engl J Med. 2016;375(19):1823–33.
non-small-cell lung cancer: optimal number of treatment cycles. 28. Antonia SJ, Villegas A, Daniel D, Vicente D, Murakami S, Hui R,
Expert Rev Anticancer Ther. 2016;16(6):653–60. et al. Durvalumab after chemoradiotherapy in stage III non–small-­
8. Garon EB, Cao D, Alexandris E, John WJ, Yurasov S, Perol M. A cell lung cancer. N Engl J Med. 2017;377(20):1919–29.
randomized, double-blind, phase III study of docetaxel and ramu- 29. fda.gov. https://www.fda.gov/drugs/resources-­information-­
cirumab versus docetaxel and placebo in the treatment of stage IV approved-­drugs/fda-­approves-­atezolizumab-­first-­line-­treatment-­
non-small-cell lung cancer after disease progression after 1 previ- metastatic-­nsclc-­high-­pd-­l1-­expression
ous platinum-based therapy (REVEL): treatment rationale an. Clin 30. Spigel D, de Marinis F, Giaccone G, Reinmuth N, Vergnenegre A,
Lung Cancer. 2012;13(6):505–9. Barrios CH, et al. IMpower110: interim overall survival (OS) anal-
9. Miyauchi E, Inoue A. Immune checkpoint therapy for non-small-­ ysis of a phase III study of atezolizumab (atezo) vs platinum-based
cell lung cancer. Japanese J Cancer Chemother. 2016;43(6): chemotherapy (chemo) as first-line (1L) treatment (tx) in PD-L1–
666–71. selected NSCLC. Ann Oncol. 2019;30(October):v915.
10. Postow MA, Callahan MK, Wolchok JD. Immune checkpoint 31. Moreno V, Gil-Martin M, Johnson ML, Aljumaily R, Lopez Criado
blockade in cancer therapy. J Clin Oncol. 2015;33(17):1974–82. P, Northfelt DW, et al. Cemiplimab, a human monoclonal anti-­
11. Xia B, Herbst R. Immune checkpoint therapy for non-small-cell PD-­1, plus radiotherapy (RT) in advanced non-small cell lung can-
lung cancer: an update. Immunotherapy. 2016;8(3):279–98. cer (NSCLC): results from a phase I expansion cohort (EC 2). Ann
12. Pardoll DM. The blockade of immune checkpoints in cancer immu- Oncol. 2018;29(December):x26.
notherapy. Nat Rev Cancer. 2012;12(4):252–64. 32. Sezer A, Kilickap S, Gümüş M, Bondarenko I, Özgüroğlu M,
13. Dixon KO, Schorer M, Nevin J, Etminan Y, Amoozgar Z, Kondo Gogishvili M, et al. LBA52 EMPOWER-lung 1: phase III first-­
T, et al. Functional anti-TIGIT antibodies regulate develop- line (1L) cemiplimab monotherapy vs platinum-doublet chemo-
ment of autoimmunity and antitumor immunity. J Immunol. therapy (chemo) in advanced non-small cell lung cancer (NSCLC)
2018;200(8):3000–7. with programmed cell death-ligand 1 (PD-L1) ≥50%. Ann Oncol.
14. Solomon BL, Garrido-Laguna I. TIGIT: a novel immunotherapy 2020;31:S1182–3.
target moving from bench to bedside. Cancer Immunol Immuno- 33. Patnaik A, Powell SF, Gentzler RD, Martins RG, Stevenson JP,
ther. 2018;67(11):1659–67. Jalal SI, et al. study. 2019;17(11):1497–1508.
15. ascopubs.org. https://ascopubs.org/doi/abs/10.1200/JCO.2020. 34. Paz-Ares L, Luft A, Vicente D, Tafreshi A, Gümüş M, Mazières J,
38.15_suppl.9503 et al. Pembrolizumab plus chemotherapy for squamous non-small-­
16. merck.com. https://www.merck.com/news/merck-­presents-­ cell lung cancer. N Engl J Med. 2018;379(21):2040–51.
promising-­new-­data-­for-­three-­investigational-­medicines-­from-­ 35. Socinski MA, Jotte RM, Cappuzzo F, Orlandi F, Stroyakovskiy D,
diverse-­a nd-­e xpansive-­o ncology-­p ipeline-­a t-­e smo-­v irtual-­ Nogami N, et al. Atezolizumab for first-line treatment of metastatic
congress-­2020/ nonsquamous NSCLC. N Engl J Med. 2018;378(24):2288–301.
17. Duma N, Santana-Davila R, Molina JR. Non–small cell lung can- 36. Stinchcombe TE. IMpower 131: the exception to the rule. J Thorac
cer: epidemiology, screening, diagnosis, and treatment. Mayo Clin Oncol. 2020;15(8):1258–60.
Proc. 2019;94(8):1623–40. 37. Rizvi NA, Cho BC, Reinmuth N, Lee KH, Luft A, Ahn MJ, et al.
18. Wang L, Yue HU, Wang S, Shen J, Wang X. Biomarkers of immu- Durvalumab with or without Tremelimumab vs standard chemo-
notherapy in non-small cell lung cancer (review). Oncol Lett. therapy in first-line treatment of metastatic non-small cell lung can-
2020;20(5):1–14. cer: the MYSTIC phase 3 randomized clinical trial. JAMA Oncol.
19. Gandhi L, Rodríguez-Abreu D, Gadgeel S, Esteban E, Felip E, De 2020;6(5):661–74.
Angelis F, et al. Pembrolizumab plus chemotherapy in metastatic 38. Med CC. Nivolumab plus ipilimumab as first-line treatment for
non-small-cell lung cancer. N Engl J Med. 2018;378(22):2078–92. advanced non-small-cell lung cancer (CheckMate 012): results of
20. Nosaki K, Saka H, Hosomi Y, Baas P, de Castro G, Reck M, et al. an open- label, phase 1, multicohort study. 2015;40(4):1291–1296.
Safety and efficacy of pembrolizumab monotherapy in elderly 39. Hellmann MD, Paz Ares L, Bernabe Caro R, Zurawski B, Kim SW,
patients with PD-L1–positive advanced non–small-cell lung can- Carcereny Costa E, et al. Nivolumab plus ipilimumab in advanced
cer: pooled analysis from the KEYNOTE-010, KEYNOTE-024, non-small-cell lung cancer. N Engl J Med. 2019;381(21):2020–31.
and KEYNOTE-042 studies. Lung Cancer. 2019;135(July):188– 40. ascopost.com. https://ascopost.com/issues/july-­25-­2020/
95. Available from: https://doi.org/10.1016/j.lungcan.2019.07.004 nivolumabipilimumab-shows-benefit-­in-­patients-­with-­non-­small-­
21. Patel SP, Kurzrock R. PD-L1 expression as a predictive biomarker cell-­lung-­cancer-­and-­brain-­metastases/
in cancer immunotherapy. Mol Cancer Ther. 2015;14(4):847–56. 41. clinicaltrials.gov. https://clinicaltrials.gov/ct2/show/results/NCT
22. Johnson DB, Frampton GM, Rioth MJ, Yusko E, Xu Y, Guo X, et al. 03215706
Targeted next generation sequencing identifies markers of response 42. Rohaan MW, Wilgenhof S, Haanen JBAG. Adoptive cellular thera-
to PD-1 blockade. Cancer Immunol Res. 2016;4(11):959–67. pies: the current landscape. Virchows Arch. 2019;474(4):449–61.
23. Tray N, Weber JS, Adams S. Predictive biomarkers for checkpoint 43. Wang J, Wang X. Beyond chemotherapy and targeted therapy:
immunotherapy: current status and challenges for clinical applica- adoptive cellular therapy in non-small cell lung cancer. Mol Biol
tion. Cancer Immunol Res. 2018;6(10):1122–8. Rep. 2014;41(9):6317–23.
24. Rizvi H, Sanchez-Vega F, La K, Chatila W, Jonsson P, Halpenny D, 44. cancerresearch.org. https://www.cancerresearch.org/immunother-
et al. Molecular determinants of response to anti-programmed cell apy/treatment-­types/adoptive-­cell-­therapy
death (PD)-1 and anti-programmed death-ligand 1 (PD-L1) block- 45. https://clinicaltrials.gov/ct2/show/NCT03330834
 Immuno-Oncologic Treatment
of Genitourinary Malignancies 69
Przemyslaw Twardowski

Abstract 69.1 Introduction

Tumors of the genitourinary tract represent almost 30% of
malignancies in the United States. Prostate cancer accounts Genitourinary malignancies encompass a diverse group of
for majority of cases followed by urothelial, kidney, and neoplasms originating from the urinary and reproductive
testicular cancers. Despite their anatomical proximity, the tracts. Prostate, urothelial (bladder, renal pelvis, ureter, ure-
biology of these tumors is quite heterogeneous. In the last thra), kidney, and testicular cancers constitute the majority of
10 years, immunotherapy became one of the main cancer genitourinary tumors; penile and adrenal cancers while
treatment modalities, and the field of genitourinary malig- included in that category represent a small minority in the
nancies benefitted tremendously from that development United States. Despite anatomical and functional connec-
although to a variable degree depending on the specific tions between genitourinary tissues, the biological character-
tumor type. Management of advanced kidney cancer istics of their neoplastic counterparts are extremely
already relies on checkpoint inhibitors as the main compo- heterogeneous and defy unified therapeutic strategies.
nent of treatment, and urothelial cancers to a large extent Historically kidney cancer represented one of the most prom-
are still dependent on platinum chemotherapies; however ising targets for immunotherapy and continues to exemplify
checkpoint blockade plays an increasingly critical role in this today. Indeed, current frontline standard of care for
their therapy. Significant progress remains to be accom- advanced kidney cancer is based on immunotherapeutic
plished to realize the promise of immunotherapy in the drugs. Urothelial cancers are also responsive to immunother-
area of prostate cancer, the most prevalent cancer type of apeutic strategies, but currently immunotherapy still remains
the group. complimentary to chemotherapy rather than the preeminent
treatment option. While significant advances were made in
the treatment of metastatic prostate cancer, they are based
Learning Objectives primarily on androgen hormone manipulation, and immuno-
• Understand the role of immunotherapy in the man- therapy is relegated to niche applications. Testicular cancer
agement of genitourinary malignancies based on is exquisitely sensitive to and curable with surgery and/or
the review of pivotal clinical trial results that estab- chemotherapy and radiation, and no role for immuno-­
lished current standard of care. oncology drugs has emerged as a potential treatment option
• Review the distinct differences in the application of thus far. In this chapter, we will review the current state of
various immunotherapeutic agents in the manage- immuno-oncology drugs in the treatment of genitourinary
ment of kidney, bladder, and prostate cancers. malignancies.
• Understand the directions of ongoing clinical
research aiming at future expansion of immuno-
therapy in the treatment of genitourinary cancers. 69.2 Kidney Cancer

Kidney cancer represents a tumor that historically was con-

sidered susceptible to recognition and elimination by the
P. Twardowski (*)
Medical Oncology and Urologic Oncology, Department of immune system [1]. Kidney cancer is very resistant to stan-
Urologic Oncology, John Wayne Cancer Institute, dard chemotherapeutic agents, and rare cases of spontaneous
Santa Monica, CA, USA regression of metastatic renal cell carcinoma (RCC) raised
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 729
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_69
730 P. Twardowski

the intriguing possibility of immune-mediated anticancer (3 mg per kilogram) plus ipilimumab (1 mg per kilogram)
effect and instigated interest in pharmacological modulation intravenously every 3 weeks for four doses, followed by
of the immune system in the therapy of RCC [2]. Immune nivolumab (3 mg per kilogram) every 2 weeks, or sunitinib
cytokines were investigated and provided a therapeutic (50 mg) orally once daily for 4 weeks (6-week cycle). The
option since the 1980s. Interferon alpha (IFNα) resulted in co-primary endpoints were OS (alpha level, 0.04), ORR
very modest response rates and improvement in overall sur- (alpha level, 0.001), and progression-free survival (PFS)
vival when compared to chemotherapy and progesterone [3]. (alpha level, 0.009) among patients with intermediate or
In 1992, another cytokine interleukin-2 (IL-2) gained FDA poor prognostic risk. The median OS was not reached with
approval for the treatment of RCC [4]. This agent when nivolumab plus ipilimumab versus 26.0 months with suni-
given intravenously at high doses produced an overall tinib (hazard ratio for death, 0.63; p < 0.001). The ORR was
response rate of about 20% with 8–10% complete responses. 42% versus 27% (p < 0.001), and the complete response
What was remarkable about these responses was that they (CR) rate was 9% with CPI combination versus 1% with
were frequently durable and resulted in a curative effect in a sunitinib among intermediate- and poor-risk patients with
small subset of patients with metastatic RCC. Unfortunately, previously untreated advanced renal cell carcinoma. The
high-dose IL-2 is associated with significant side effects results of this clinical trial established a combination of
requiring unique expertise and access to intensive care nivolumab and ipilimumab as a standard frontline regimen in
resources and is limited to patients with excellent perfor- patients with intermediate- and poor-risk RCC.
mance status and intact cardiopulmonary function. After 2 VEGF-targeted agents became incorporated into the
decades since the approval of IL-2 without major break- frontline management of kidney cancer since the approval of
throughs, the modern era of immunotherapy began with the sorafenib in 2005 [7]. Multiple agents in that category (suni-
development of checkpoint inhibitors (CPIs) such as anti-­ tinib, pazopanib, axitinib, cabozantinib, lenvatinib) received
programmed death receptor 1 (anti-PD-1), anti-programmed regulatory approval in various lines of treatment of kidney
death receptor ligand 1 (anti-PD-L1), and anti-CTLA-4 anti- cancer. These compounds work primarily by inhibition of
bodies. Early CPI clinical investigation efforts in kidney can- the VEGF axis which is central in the pathophysiology of
cer focused on pretreated patients who progressed on RCC, but they differ in their effect on other pathways affect-
standard vascular endothelial growth factor (VEGF) target- ing redundancies of angiogenic and cancer cell signaling [8].
ing therapy. The first Food and Drug Administration (FDA) In the context of the pivotal role assumed by checkpoint
approval was granted in 2015 for PD-1 inhibitor nivolumab inhibitors in the treatment of RCC, simultaneous targeting of
which demonstrated a survival advantage over mTOR inhibi- angiogenic pathways including VEGF, fibroblast growth fac-
tor everolimus (standard of care at that time) in patients with tor (FGF), and platelet-derived growth factor (PDGF) has
metastatic clear cell RCC who progressed on one to two anti- emerged as a strategy to modulate and enhance the immune
angiogenic regimens. The median overall survival (OS) was response. It turns out that in addition to stimulation of angio-
25.0 months (95% confidence interval [CI] 21.8–not esti- genesis, VEGF signaling results in the increase of myeloid-­
mable) with nivolumab and 19.6 months (95% CI 17.6–23.1) derived suppressor cells (MDSCs) and regulatory T cells
with everolimus. The hazard ratio for death with nivolumab (Tregs) which negatively impacts immune response [9].
versus everolimus was 0.73 (98.5% CI 0.57–0.93; p = 0.002). VEGF also inhibits the maturation of dendritic cells into
The objective response rate (ORR) was greater with mature antigen-presenting cells (APCs) and reduces lym-
nivolumab than with everolimus (25% vs. 5%; odds ratio, phocyte infiltration [9]. Direct and indirect evidence sug-
5.98 [95% CI 3.68–9.72]; p < 0.001) [5]. Once the efficacy gests that VEGF inhibitors and multikinase inhibitors that
of CPI was established in the second-line setting and beyond, include VEGF targeting can modulate the tumor microenvi-
they were rapidly tested and incorporated into the first-line ronment and reverse VEGF-mediated immune-inhibitory
regimens. Combination of nivolumab and CTLA-4 inhibitor mechanisms [10]. That hypothesis was validated in multiple
ipilimumab demonstrated broad therapeutic advantage over randomized clinical trials comparing this combination strat-
the previous standard-of-care VEGF receptor tyrosine kinase egy to sunitinib control (Table 69.1).
inhibitor (VEGFR TKI) sunitinib in a CheckMate 214 trial All three immuno-oncology (IO)/VEGF targeting combi-
[6]. Data was particularly strong in patients with intermedi- nations substantially outperformed sunitinib with respect to
ate- and poor-risk cancer based on the International key endpoints of OS, PFS, and ORR. Even though these com-
Metastatic Renal Cell Carcinoma Database Consortium binations have not been directly compared to single-­agent
(IMDC) criteria which are used to assess the prognosis of PD-1 or programmed death ligand 1 (PD-L1) inhibitors,
patients with metastatic RCC utilizing simple clinical and phase II data with single CPIs yields responses of < 40% [13],
laboratory criteria. One thousand ninety-six (of which 847 suggesting that combination strategies are more effective.
fell into the intermediate- or high-risk group) patients with There is now a broad consensus that combination therapy
clear cell RCC were randomized to receive either nivolumab incorporating checkpoint inhibitors represents a modern
69 Immuno-Oncologic Treatment of Genitourinary Malignancies 731

Table 69.1 First-line combination IO/VEGF targeting therapy (phase III trials in RCC)
Nivolumab/cabozantinib
Pembrolizumab/axitinib Avelumab/axitinib 9ER (Choueiri T, et al. ESMO 2020, Abstract
KN-426 [11] Javelin 101 [12] 6960)
OS HR 0.53 (95% CI 0.38–0.74; p < 0.0001) Immature HR 0.6 (95% CI 0.4–0.89; p = 0.001)
PFS HR 0.69 (95% CI 0.57–0.84; p < 0.001) HR 0.69 (95% CI 0.574–0.825; 0.51 (95% CI 0.41–0.64; p < 0.0001)
p < 0.0001)
ORR 59.3% 51.4% 55.7%
CR 5.8% 3.4% 8%

standard of care for frontline therapy of advanced [16]. BCG was originally developed as a vaccine against
RCC. Whether pure immunotherapy combo like nivolumab/ tuberculosis (TB) and is derived from attenuated
ipilimumab or CPI/VEGF inhibitor combination should be Mycobacterium bovis. Early autopsy observations of the
used is a subject of debate and depends on specific circum- potential decrease in cancer incidence in tuberculosis patients
stances and nuanced patient selection. In general, IO/IO ultimately led to the investigation of BCG for the treatment
combination is a strong contender in patients with intermedi- of cancer including non-muscle-invasive bladder cancer
ate- or poor-risk metastatic RCC by the International (NMIBC) [17]. Intravesical administration of BCG in the
Metastatic Renal Cell Carcinoma Database Consortium setting of superficial bladder cancer maximizes direct expo-
(IMDC); however its superiority over a single VEGF inhibi- sure of cancer cells to therapy and minimizes, although not
tor (sunitinib) in the favorable-risk subgroup is unclear [6, completely prevents, toxic systemic effects. NMIBC repre-
14]. IO/VEGF inhibitor combinations maintain their effec- sents around 70% of the 70,000 new bladder cancer cases
tiveness across a whole spectrum of IMDC risk groups diagnosed in the United States each year [18]. The develop-
(including favorable-risk group), but VEGF targeting com- ment of BCG as an immunotherapy for NMIBC and its
ponent results in a different adverse profile than IO/IO com- proven effects of reducing recurrence and progression have
bination, and that may influence the choice of therapy in an transformed the treatment of this malignancy, and its con-
individual patient depending on underlying comorbidities tinuous importance is demonstrated by a significant negative
and patient and physician preferences. The ongoing clinical impact of recent shortages of BCG on delivery of appropriate
trial COSMIC-313 will attempt to investigate whether the care for patients with NMIBC [19, 20]. The effects of BCG
triple combination of nivolumab with ipilimumab and cabo- on the immune function are very complex and result in acti-
zantinib (IO/IO/VEGF inhibitor) is superior to nivolumab vation of innate and adaptive immune systems and involve-
with ipilimumab (IO/IO). As of now, there is no accepted ment of multiple cytokines including IL-2, IL-6, IL-8, IL-10,
biomarker predicting responsiveness to immunotherapy in IL-12, IFN, and tumor necrosis factor-α (TNF-α). Prior to
RCC. Tumor mutational burden (TMB) is relatively low as the development of BCG immunotherapy, 50% of patients
compared to other immune-responsive tumors: T-cell infil- with carcinoma in situ (CIS) developed the muscle-invasive
tration and PD-L1 expression also fall short in their ability to disease. A large, randomized controlled trial by the Southwest
select patients for IO therapy in RCC. There is an ongoing Oncology Group (SWOG) comparing induction plus a
effort to address these shortcomings and to better understand 3-week maintenance course versus induction alone using
the genomic and immune landscape affecting response to intravesical BCG in CIS showed that the addition of a 3-week
immunotherapy in kidney cancer [15]. Another area of sig- maintenance course beginning at 3 months produced a com-
nificant interest focuses on the role of immunotherapy in the plete response rate of 84%, and, with ongoing 3-monthly
management of less common variants of kidney cancer (non-­ installations, over 70% of those who responded remained
clear cell RCC) which constitute approximately 25% of kid- disease-free for over 5 years [21]. There are some data sup-
ney tumors and were less intensely studied as compared to porting the need to continue maintenance for a period of at
the dominant clear cell variant with respect to the role of least 3 years, as evidenced by a study comparing a 1-year
immunotherapy. with a 3-year maintenance regimen for high-risk disease
only [22].
Frontline therapy for muscle-invasive and advanced/
69.3 Urothelial Cancer metastatic urothelial cancer (UC) has not changed for
decades and incorporates platinum (preferable cisplatin)
From the historical perspective, application of the Bacillus chemotherapy combinations either with gemcitabine or
Calmette–Guerin (BCG) vaccine in the therapy of superficial alternatively a combination of dose-dense methotrexate,
bladder cancer represents the earliest approved indication for vinblastine, adriamycin, and cisplatin (ddMVAC). This
the use of immunotherapy in genitourinary malignancies strategy results in response rates of 50–70% and achieves
732 P. Twardowski

occasional durable remissions [23]. In many ways, UC rep- nostic VENTANA PD-L1 SP142 assay] or patients who are
resents an attractive target for the development of IO drugs ineligible for any platinum chemotherapy. Ultimately
based on the biology of neoantigen-rich tumors (likely Imvigor130 did result in improvement in PFS in the chemo-­
related to high mutational load from tobacco exposure) and immunotherapy arm (A) versus chemotherapy alone (arm
a desperate need for improved treatments especially beyond C) (hazard ratio 0.82; 95% CI 0.7–0.96), but the OS data
first-line combinations where historically no effective ther- has not reached the level of significance at the time of latest
apies existed [24]. Initial clinical development focused analysis [28]. In addition, KEYNOTE-361 study with a
exactly on that scenario: patients who progressed after prior very similar design to Imvigor130 but utilizing pembroli-
platinum-based chemotherapy or who were not medically zumab has failed to meet the primary study endpoint of OS
fit for chemotherapy. Over a short period of time, five (Alva A, et al. Pembrolizumab combined with chemother-
checkpoint inhibitors (atezolizumab, pembrolizumab, apy versus chemotherapy alone as first-line therapy for
nivolumab, durvalumab, and avelumab) were approved in advanced urothelial carcinoma: KEYNOTE-361. ESMO
that setting [25]. The approval was based on a response rate 2020, Abstract LBA23). At a time of earlier interim analysis
(15–21%) that was considered favorable and durable as of KEYNOTE-361, the similar observation to Imvigor130
compared to historical controls of taxane chemotherapy. was made indicating inferior OS in patients with low PD-L1
Subsequently, pembrolizumab demonstrated overall sur- expression treated with single-arm pembrolizumab and
vival (OS) benefit in a randomized phase III study resulted in label update of pembrolizumab restricting its use
(KEYNOTE-045) comparing it to standard second-line tax- in first-line setting to any platinum-ineligible patients or
ane or vinflunine chemotherapy [26]. Median OS was cisplatin-ineligible patients whose tumors exhibit high
10.3 months (95% CI 8.0–11.8) in the pembrolizumab expression of PD-L1 (based on combined positive score
group, as compared to 7.4 months (95% CI 6.1–8.3) in the (CPS ≥ 10) by Dako PD-L1 IHC 22C3 assay). While the
chemotherapy group [hazard ratio (HR) for death, 0.73; results of these two large randomized clinical trials evaluat-
95% CI 0.59–0.91; p = 0.002]. A survival benefit was dem- ing frontline application of concurrent chemo-immunother-
onstrated regardless of PD-L1 expression [combined posi- apy failed to unequivocally change the standard of care, a
tive score (CPS) of < 10% or ≥ 10%). No difference in PFS different strategy of CPI maintenance after platinum che-
was observed, and toxicity profile favored pembrolizumab. motherapy induction did just that. JAVELIN Bladder 100
Atezolizumab was also evaluated in a randomized phase III Phase III switch-maintenance study randomized patients
study (IMvigor211) comparing it to t­axane or vinflunine with metastatic UC to standard four to six cycles of platin-
chemotherapy in the second-line setting [27]. The hierarchi- gemcitabine combination or the same chemotherapy fol-
cal design of the study had a primary endpoint that focused lowed by maintenance of PD-L1 inhibitor avelumab in
on OS in a subset of patients with high PD-L1 expression, those patients who did not progress after four to six cycles
and it was negative making additional statistical analyses of chemotherapy. Eighty-five percent of patients did not
only exploratory. It is unclear whether a positive OS out- progress after initial chemotherapy, and the study demon-
come with pembrolizumab versus less apparent benefit of strated significant improvement in OS (median 21.4 months
atezolizumab in these randomized phase III studies repre- in CPI maintenance arm versus 14.3 months in standard
sents a true difference in drug activity or is a result of dis- chemotherapy alone arm; the hazard ratio for death, 0.69
similar statistical design of respected trials. Despite high (95% CI 0.56–0.86; p = 0.001) favoring the experimental
hopes for quick expansion of CPI therapy into first-line and arm [29]. The results were positive in the entire intention to
adjuvant space, the emerging results of clinical trials are treat the study population, but the high PD-L1 expressors
mixed and have not resulted in the replacement of platin- appeared to experience the most robust benefit. On June 30,
based chemotherapy in the frontline setting. IMvigor130 2020, avelumab received FDA approval as a maintenance
study randomized patients with metastatic and locally treatment for patients with locally advanced or metastatic
advanced UC to three arms: (A) atezolizumab + platinum/ urothelial carcinoma that has not progressed with first-line
gemcitabine, (B) atezolizumab monotherapy, and (C) pla- platinum-containing chemotherapy. Therefore, while platin-­
cebo + platinum/gemcitabine combination [28]. Interim based combination chemotherapy remains a mainstay of
analysis of that study demonstrated that patients with low frontline treatment of metastatic UC, subsequent switch-­
PD-L1 expression treated on atezolizumab monotherapy maintenance avelumab represents a new modification of
arm (B) had inferior OS as compared to chemotherapy arms standard of care in that circumstance. Multiple clinical tri-
and resulted in restriction in accrual to that arm to high als are evaluating the role of CPI therapy in the periopera-
PD-L1 expressors only. It also resulted in an adjustment of tive setting in patients with muscle-invasive disease
FDA label for frontline use of atezolizumab to cisplatin- following cystectomy or nephroureterectomy. IMvigor010
ineligible patients whose tumors exhibit high expression of randomized patients to 1 year of adjuvant atezolizumab or
PD-L1 [PD-L1-stained tumor-infiltrating immune cells observation and was the first reported trial of that group.
(ICs) covering ≥ 5% of the tumor area by companion diag- Unfortunately, primary endpoint of disease-free survival
69 Immuno-Oncologic Treatment of Genitourinary Malignancies 733

Basal Squamous
Luminal
KRT5,6,14+, GATA3+, FOXA1+
KRT20+, GATA3+, FOXA1+

Luminal-papillary Luminal-infiltrated Females, squamous differentiation, High-

PD-L1 expression, immune infiltrates

FGFF3 mutation, fusion, EMT markers, (TWIST1, ZEB1) wild type

amplification, papillary histology p-53 miR-200 Moderate-high-PD-L1 Good response to CPI

Neuronal

Good response to CPI SOX2, DLX6, MSI1, PLEKHG4 E2F3/SOX4

Poor response to CPI, FGFR
amplification, high cell cycle
inhibitor if alteration present

Cispiatin-Etoposide therapy

Fig. 69.1 Proposed molecular subtyping of urothelial carcinoma into luminal, basal squamous, and neuronal

(DFS) did not differ between the arms, and OS data is frontline setting), it has not achieved discriminatory sig-
immature (Hussain M, et al. IMvigor010: Primary analysis nificance for patient selection in the majority of clinical
from a phase III randomized study of adjuvant atezolizumab scenarios. Classification of urothelial cancer by molecular
versus observation in high-risk muscle-invasive urothelial markers into luminal (divided into luminal-­papillary, lumi-
carcinoma. ASCO 2020, Abstract 5000). On the other hand, nal-infiltrated), basal/squamous, and neuronal subtypes
a press release from September 24, 2020, indicated that [30]. Figure 69.1 represents one of the latest efforts of
CheckMate 274, a pivotal phase III trial evaluating identifying unique subsets of urothelial carcinoma with
nivolumab after surgery in patients with high-risk, muscle- important therapeutic implications. Luminal infiltrated
invasive urothelial carcinoma, met its primary endpoints of (accounting for about 19% of cases) and basal squamous
improving disease-free survival (DFS) versus placebo in (accounting for 35% of cases) subtypes appear to respond
both all randomized patients and in patients whose tumor particularly well to CPI therapy, but these observations
cells express PD-L1 ≥ 1%. Details of the data have not require prospective validation.
been presented as of writing of this manuscript. Future
directions of immunotherapeutics in urothelial cancer
involve novel combinations and refinement of identifica- 69.4 Prostate Cancer
tion of appropriate patients for immunotherapy. There are
multiple new combinations in development including IO/ Prostate cancer has largely missed the recent checkpoint
IO, IO/VEGF TKIs, and IO with antibody-drug conjugates inhibitor revolution likely as a result of being a low muta-
(ADCs). Of particular interest is very impressive phase II tional burden, immunologically “cold” tumor, thus not pos-
data of enfortumab vedotin (ADC targeting nectin-4 and sessing immune infiltrate substrate upon which the CPI
already FDA approved as a single drug in third-line set- therapy can perform its function [31]. An exception is a small
ting) combined with pembrolizumab in the first-line UC subset (approximately 2%) of prostate cancers associated
demonstrating 73% response rate (Rosenberg JE, et al. GU with mismatch repair alterations in which pembrolizumab
Ca Symp 2020. Abstract 441). This combination is cur- demonstrated significant activity in a study combining mul-
rently being tested in a pivotal randomized phase III study. tiple tumor types with underlying mismatch repair genotype
Biomarker-driven selection for treatment with immuno- [32]. This garnered FDA approval which represented a unique
therapeutics (particularly CPIs) has been met with mixed circumstance where tissue agnostic but molecularly defined
success. While higher expression of PD-L1 on tumor cells malignancies were combined for drug approval [33]. Outside
and/or infiltrating immune cells (ICs) has been associated of this scenario studies of checkpoint inhibitors have failed to
with a trend of stronger clinical benefit (particularly in achieve results that would justify their approval in advanced
734 P. Twardowski

prostate cancer [34]. Despite these setbacks, there are ongo- therapeutic options. Several studies investigated the potential
ing research efforts to attempt to convert the “cold” prostate role of CPIs in that setting although theoretically testicular
cancer microenvironment to one that is immunologically cancer may not represent an ideal target for CPI treatment
more active and receptive to CPI therapy. One of the exam- [41]. Testicular cancer has a low tumor mutational burden,
ples that has advanced to phase III study is a combination of and mammalian testis represents an immunologically privi-
PD-L1 inhibitor atezolizumab and a broad-­spectrum VEGFR leged site with the immunological microenvironment that
TKI cabozantinib. While these agents have limited anti-pros- protects germ cells from autoimmune attack; it is associated
tate cancer activity on their own [35, 36], a combination of with the deficient immune response against antigens [42].
both demonstrated a very respectable response rate of 32% in Nevertheless, PD-L1 expression is significantly higher in tes-
heavily pretreated metastatic castration-­ resistant prostate tis cancer as compared to normal testicular tissue, and high
cancer (mCRPC) (Agarwal N, et al. Cabozantinib in combi- PD-L1 expression has been associated with poor prognosis
nation with atezolizumab in patients with metastatic castra- [43]. Even though case reports of isolated durable response
tion-resistant prostate cancer: results of cohort 6 of with CPI were published unfortunately, pembrolizumab,
COSMIC-021 study. ASCO, 2020 Abstract 5564). Multiple nivolumab, duralumin +/− tremelimumab, and avelumab did
other immunotherapeutic strategies in the early development not demonstrate activity in phase II studies that would war-
of prostate cancer include bispecific antibodies and chimeric rant their further development in this disease [44].
antigen receptor therapy (CART) platforms [31]. Historically
a different type of immunotherapeutic approach has been suc-
cessful: as one of the first cancer therapeutic vaccines sipu- 69.6 Conclusion
leucel-T was approved for the treatment of advanced mCRPC
in 2010. Sipuleucel-T is an active cellular immunotherapy Genitourinary malignancies represent anatomically connected
consisting of autologous peripheral blood mononuclear cells but biologically diverse group of tumors with very different
(PBMCs), including antigen-presenting cells (APCs), that molecular and immunological landscapes. Historically renal
have been activated in vivo with a recombinant fusion protein cell cancer was subject of particular interest for the immuno-
(PA2024) [37]. PA2024 consists of a prostate antigen, pros- therapeutic approaches, and it remains so today with CPIs pro-
tate acid phosphatase (PAP), that is fused to granulocyte- viding a cornerstone of effective treatment for advanced
macrophage colony-stimulating factor, an immune cell disease. Urothelial cancer also witnessed the significant appli-
activator [37]. In a phase III trial evaluating patients with cation of CPIs in many clinical scenarios although immuno-
asymptomatic or minimally symptomatic mCRPC, sipuleu- therapy has not displaced chemotherapy as a frontline
cel-T demonstrated 22% reduction in the risk of death as treatment strategy and arguably the benefits of CPIs in urothe-
compared with placebo group (hazard ratio 0.78; 95% CI lial cancers are less generalized than in kidney cancers.
0.61–0.98, p = 0.03) [38]. This reduction represented Immunotherapy plays only a niche role in the management of
4.1 months of improvement in median OS (25.8 months for prostate cancer with the availability of autologous vaccine
sipuleucel-T versus 21.7 months in the placebo group). No sipuleucel-T for low-volume castration-resistant prostate can-
significant improvement in PFS and objective or PSA cer, but no definitive role for CPIs has emerged so far in pros-
responses was observed. Sipuleucel-T is a viable treatment tate cancer outside of rare cases associated with mismatch
option especially in patients with low-­volume, asymptomatic repair mutations. No application for any immunotherapy has
mCRPC; however, lack of meaningful responses makes it less materialized yet for the management of germ cell tumors.
attractive for many patients with advanced, clinically symp- Ongoing and future research efforts aiming at enhancing
tomatic prostate cancer [39]. Sipuleucel-T is currently stud- the benefits of immunotherapy in the management of genito-
ied in a phase III randomized trial in patients with low-risk urinary malignancies will focus on the following areas:
and favorable intermediate-risk prostate cancer who are oth-
erwise managed with active surveillance. • Discovery and validation of biomarkers aiming at more
precise selection of patients for immunotherapy.
• Refinements of currently available treatments with respect
69.5 Testicular Cancer/Germ Cell Tumors to dosing and duration of immunotherapy.
• Strategies to minimize toxicities of immunotherapy.
Germ cell tumors (GCT) represent very unique malignancies • New combinations of immunotherapy drugs and
in which the majority of patients achieve a curative outcome immunotherapy-­ enhancing compounds aiming at con-
with cisplatin-based combination chemotherapy even in cases verting immunologically “cold” tumors to more immuno-
of extensive, metastatic disease [40]. Unfortunately, a frac- logically responsive ones.
tion of patients who fail first- and especially second-line sal- • Better understanding of mechanism of intrinsic and
vage therapy have a very poor prognosis and lack of effective acquired resistance to immunotherapy.
69 Immuno-Oncologic Treatment of Genitourinary Malignancies 735

References full dose and 1 year with 3 years of maintenance BCG. Eur Urol.
2014;65(1):69–76.
23. Bellmunt J, et al. Optimizing therapeutic strategies in advanced
1. Hammers H. Immunotherapy in kidney cancer: the past, present, bladder cancer: update on chemotherapy and the role of targeted
and future. Curr Opin Urol. 2016;26(6):543–7. agents. Crit Rev Oncol Hematol. 2009;69(3):211–22.
2. Muraoka K, Nishikawa Y, Isoyama T. Spontaneous regression of 24. Farina MS, Lundgren KT, Bellmunt J. Immunotherapy in uro-
metastatic renal cell carcinoma after cytoreductive nephrectomy: a thelial cancer: recent results and future perspectives. Drugs.
case report. Urol Case Rep. 2020;31:101179. 2017;77(10):1077–89.
3. Naitoh J, Belldegrun AS. Interferon and kidney cancer: a promise 25. Alhalabi O, et al. Current and future landscape of immune check-
unfulfilled? Cancer J Sci Am. 1998;4(3):155–6. point inhibitors in urothelial cancer. Oncology (Williston Park).
4. Dutcher JP. Interleukin-2 based therapy for kidney cancer. Cancer 2019;33(1):11–8.
Treat Res. 2003;116:155–2. 26. Bellmunt J, et al. Pembrolizumab as second-line ther-
5. Motzer RJ, et al. Nivolumab versus everolimus in advanced renal-­ apy for advanced urothelial carcinoma. N Engl J Med.
cell carcinoma. N Engl J Med. 2015;373(19):1803–13. 2017;376(11):1015–26.
6. Motzer RJ, et al. Nivolumab plus ipilimumab versus suni- 27. Powles T, et al. Atezolizumab versus chemotherapy in patients with
tinib in advanced renal-cell carcinoma. N Engl J Med. platinum-treated locally advanced or metastatic urothelial carci-
2018;378(14):1277–90. noma (IMvigor211): a multicentre, open-label, phase 3 randomised
7. Escudier B, et al. Sorafenib in advanced clear-cell renal-cell carci- controlled trial. Lancet. 2018;391(10122):748–57.
noma. N Engl J Med. 2007;356(2):125–4. 28. Galsky MD, et al. Atezolizumab with or without chemo-
8. Vachhani P, George S. VEGF inhibitors in renal cell carcinoma. therapy in metastatic urothelial cancer (IMvigor130): a multi-
Clin Adv Hematol Oncol. 2016;14(12):1016–28. centre, randomised, placebo-controlled phase 3 trial. Lancet.
9. Bergerot P, et al. Cabozantinib in combination with immuno- 2020;395(10236):1547–57.
therapy for advanced renal cell carcinoma and urothelial car- 29. Powles T, et al. Avelumab maintenance therapy for advanced or met-
cinoma: rationale and clinical evidence. Mol Cancer Ther. astatic urothelial carcinoma. N Engl J Med. 2020;383(13):1218–30.
2019;18(12):2185–93. 30. Robertson AG, et al. Comprehensive molecular characterization of
10. Incio J, et al. PlGF/VEGFR-1 signaling promotes macrophage muscle-invasive bladder cancer. Cell. 2018;174(4):540–56.
polarization and accelerated tumor progression in obesity. Clin 31. Rizzo A, et al. Is there a role for immunotherapy in prostate cancer?
Cancer Res. 2016;22(12):2993–04. Cells. 2020;9(9):2051.
11. Rini BI, et al. Pembrolizumab plus axitinib versus suni- 32. Marabelle A, et al. Efficacy of pembrolizumab in patients with non-
tinib for advanced renal-cell carcinoma. N Engl J Med. colorectal high microsatellite instability/mismatch repair-deficient
2019;380(12):1116–27. cancer: results from the phase II KEYNOTE-158 study. J Clin
12. Choueiri TK, et al. Updated efficacy results from the JAVELIN Oncol. 2020;38(1):1–10.
Renal 101 trial: first-line avelumab plus axitinib versus suni- 33. Prasad V, Kaestner V, Mailankody S. Cancer drugs approved based
tinib in patients with advanced renal cell carcinoma. Ann Oncol. on biomarkers and not tumor type—FDA approval of pembroli-
2020;31(8):1030–9. zumab for mismatch repair-deficient solid cancers. JAMA Oncol.
13. Xu W, Atkins MB, McDermott DF. Checkpoint inhibitor immuno- 2018;4(2):157–8.
therapy in kidney cancer. Nat Rev Urol. 2020;17(3):137–50. 34. Claps M, et al. Immune-checkpoint inhibitors and metastatic pros-
14. Heng DY, et al. External validation and comparison with other mod- tate cancer therapy: learning by making mistakes. Cancer Treat
els of the International Metastatic Renal-Cell Carcinoma Database Rev. 2020;88:102057.
Consortium prognostic model: a population-based study. Lancet 35. Smith MR, et al. Cabozantinib in chemotherapy-pretreated meta-
Oncol. 2013;14(2):141–8. static castration-resistant prostate cancer: results of a phase II non-
15. Lopez-Beltran A, et al. The identification of immunological bio- randomized expansion study. J Clin Oncol. 2014;32(30):3391–9.
markers in kidney cancers. Front Oncol. 2018;8:456. 36. Heidegger I, et al. A systematic review of the emerging role of
16. Fuge O, et al. Immunotherapy for bladder cancer. Res Rep Urol. immune checkpoint inhibitors in metastatic castration-resistant
2015;7:65–79. prostate cancer: will combination strategies improve efficacy? Eur
17. Morales A. Adjuvant immunotherapy in superficial bladder cancer. Urol Oncol. 2021;4(5):745–54.
Natl Cancer Inst Monogr. 1978;49:315–9. 37. Harzstark AL, Small EJ. Sipuleucel-T for the treatment of prostate
18. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer cancer. Drugs Today (Barc). 2008;44(4):271–8.
J Clin. 2020;70(1):7–30. 38. Kantoff PW, et al. Sipuleucel-T immunotherapy for castration-­
19. Sylvester RJ, van der Meijden AP, Lamm DL. Intravesical Bacillus resistant prostate cancer. N Engl J Med. 2010;363(5):411–22.
Calmette–Guerin reduces the risk of progression in patients with 39. Graff JN, Chamberlain ED. Sipuleucel-T in the treatment of pros-
superficial bladder cancer: a meta-analysis of the published results tate cancer: an evidence-based review of its place in therapy. Core
of randomized clinical trials. J Urol. 2002;168(5):1964–70. Evid. 2015;10:1–10.
20. Nikas CV, Smith AB. Goldilocks and the BCG: Bacillus 40. Hanna N, Einhorn LH. Testicular cancer: a reflection on 50 years of
Calmette–Guerin dose reduction in the age of shortage. Eur Urol. discovery. J Clin Oncol. 2014;32(28):3085–92.
2020;78(5):699–700. 41. Semaan A, et al. Immunotherapy: last bullet in platinum refractory
21. Lamm DL, et al. Maintenance Bacillus Calmette–Guerin immu- germ cell testicular cancer. Future Oncol. 2019;15(5):533–41.
notherapy for recurrent TA, T1 and carcinoma in situ transitional 42. Zhao S, et al. Testicular defense systems: immune privilege and
cell carcinoma of the bladder: a randomized Southwest Oncology innate immunity. Cell Mol Immunol. 2014;11(5):428–37.
Group Study. J Urol. 2000;163(4):1124–9. 43. Cierna Z, et al. Prognostic value of programmed-death-1 receptor
22. Brausi M, et al. Side effects of Bacillus Calmette–Guerin (BCG) in (PD-1) and its ligand 1 (PD-L1) in testicular germ cell tumors. Ann
the treatment of intermediate- and high-risk Ta, T1 papillary carci- Oncol. 2016;27(2):300–5.
noma of the bladder: results of the EORTC genito-urinary cancers 44. Kalavska K, et al. Immunotherapy in testicular germ cell tumors.
group randomised phase 3 study comparing one-third dose with Front Oncol. 2020;10:573977.
 Part XVIII
Current Topics

Stanley P. Leong

In this section of Current Topics, each topic has been uniquely chosen to highlight the future
directions of cancer metastasis. With promising biomarkers discovered in cancer, a newly
emerging field of personalized cancer therapy is taking hold in treating cancer patients and is
addressed by Drs. Kim and Kashani-Sabet. Cancer stem cells have been found to be associated
with cancer recurrence and metastasis. Drs. Ghadially, Kim, and Charruyer detail the origin of
stem cells and discuss their role in cancer metastasis. Certain cancers may be caused by viruses
such as hepatitis viruses for hepatocellular carcinoma of the liver and human papillomavirus
(HPV) causing anogenital and oropharyngeal cancer in ways distinctly different from other
spontaneous-arising cancers in most of the cases. Drs. Dickey, Binning, and Rathwell describe
the epidemiology and mechanisms of cancer induction by HPV, thus, illustrating the unique
virus-induced causes of cancer. Drs. White, Wargo, and McQuade summarize the evidence
regarding the significant association between microbiome and cancer progression. Artificial
intelligence (AI), a recent development in computer science consisting of a number of super-
vised and unsupervised computational techniques, is addressed by Dr. Benjamin Franc to illus-
trate AI’s potential application in cancer research. These important novel topics may become
important frontiers in the study of cancer metastasis.
 Personalized Systemic Cancer Therapy
70
Kevin B. Kim and Mohammed Kashani-Sabet

Abstract static setting, molecular targeted therapies and checkpoint

Over the past several decades, the landscape of systemic inhibitors have been revolutionizing cancer therapy, with
cancer therapy has changed drastically owing to advances improved tumor control and longer survival in an increasing
in our understanding of cancer biology, immunology, and number of advanced malignancies. Targeted drug therapy uti-
genetics. As a result, an increasing number of patients lizes drugs that inhibit molecules and proteins prominently
with advanced stages of cancer now live longer than ever. involved in growth and survival of tumor cells and thus lead
The ultimate goal of personalized cancer therapy is to to cell death and clinical regression of tumor and prolonga-
optimize clinical outcomes by selecting the “right” drug tion of patients’ survival. Early examples of molecular tar-
therapies for each patient on the basis of his or her tumor geted drugs are trastuzumab for advanced breast cancer and
genetics and/or epigenetics and by avoiding unnecessary imatinib for chronic myelogenous leukemia (CML), which
toxicity of ineffective therapy. Advances in genetic obtained Food and Drug Administration (FDA) approval in
sequencing and molecular cytogenetic techniques have 1998 and 2002, respectively, in the United States. For tar-
identified oncogenic driver mutations and epigenetic geted therapy to be effective, it is essential that tumor cells
abnormalities, which in turn have enabled the develop- contain relevant oncogenic molecules. For example, HER2
ment of effective molecular targeted drugs. In this chap- is a receptor tyrosine kinase, and HER2 gene amplification,
ter, the successful development of molecular targeted which occurs in ~ 15–20% of breast cancers, results in over-
therapies in malignant melanoma is described to exem- expression of a functional HER2 receptor [1]. The activated
plify the concept of personalized cancer therapy. HER2 signaling pathway, in turn, leads to enhanced cell
proliferation and invasion and suppression of apoptosis [2].
Accordingly, trastuzumab, a monoclonal antibody that binds
Learning Objectives to the extracellular domain of the HER2 receptor, is effec-
• To understand the principles of molecular targeted tive preferentially in cancer cells with HER2 overexpression
cancer therapy. and/or amplification and is not beneficial to patients with-
• To understand mechanisms of resistance to targeted out HER2-positive breast cancer. In addition, a hallmark of
therapy. chronic myelogenous leukemia (CML) is the translocation
involving the BCR and ABL genes, creating a constitutively
active tyrosine kinase that stimulates proliferative and sur-
vival signaling pathways [3]. Imatinib, which is an inhibitor
70.1 Introduction of BCR-ABL, as well as KIT and PDGF receptor-­alpha, is
very effective in treating this disease [4]. Similarly, vascular
Cancer therapy has evolved drastically over the past 2 decades endothelial growth factor- (VEGF) or VEGF receptor-tar-
as our knowledge of cancer genetics and immunology has geting drugs are indicated in advanced renal cell carcinoma
advanced. Although cytotoxic chemotherapy is still a main- (RCC) because of prevalent deficient von Hippel-Lindau
stay of treatment for a majority of cancer types in the meta- (VHL) gene functions, resulting in activation of the VEGFR
pathway in a vast majority of RCCs [5]. Recently, non-small
K. B. Kim (*) · M. Kashani-Sabet cell lung cancer, especially non-­squamous cell carcinoma of
California Pacific Medical Center Research Institute, the lung, has received well-­deserved attention in regard to
San Francisco, CA, USA the emergence of successful molecularly targeting therapies.
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 739
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_70
740 K. B. Kim and M. Kashani-Sabet

KRAS mutations are the most common driver genetic altera- rate for patients with advanced, metastatic melanoma was
tions, accounting for 20–30% among patients with non- approximately 10% in 2009 [10], whereas the 5-year sur-
squamous cell lung cancers [6]. EGFR mutations (12–15%), vival rates with the checkpoint inhibitor and/or inhibitors
ALK gene rearrangement (2–8%), BRAF mutation (1–5%), of BRAF and MEK kinases range between 34 and 52%
MET alterations (4–5%), ROS1 gene alterations (1–2%), and [11–13].
RET gene arrangements (1–2%) also comprise a substantial In terms of the genetics of melanoma, the most common
proportion of driver oncogenic alterations in this population mutations in cutaneous melanoma occur within the mitogen-­
[6]. Interestingly, the prevalence of these genetic alterations activated protein (MAP) kinase signal transduction pathway
vary depending on patients’ ethnicity and demographic or that leads to activation of the extracellular signal-regulated
pathological characteristics. For instance, EGFR mutations kinase (ERK), leading to proliferation, survival, invasion,
are more frequent among never-smokers, East Asian women, and metastasis of melanoma cells. Approximately 50–60%
and those with adenocarcinoma histology. Fortunately, there of cutaneous melanomas harbor a single-base substitution
are an increasing number of molecularly targeted drugs for mutation in the BRAF gene, and about 20% harbor a muta-
these alterations, which have shown a significant clinical tion in the NRAS gene, which are largely mutually exclu-
benefit. Examples of successful targeted therapies for non-­ sive [14–16]. Interestingly, non-cutaneous melanomas have
small cell lung cancers include erlotinib, gefitinib, afatinib, different genetic profiles. In acral lentiginous and mucosal
dacomitinib, and osimertinib for EGFR mutations; crizo- melanoma, BRAF and NRAS mutations occur much less fre-
tinib, ceritinib, alectinib, brigatinib, lorlatinib, and entrec- quently; multiple reports showed that 3–8% and 13–21%
tinib for ALK gene rearrangements; dabrafenib/trametinib of mucosal and acral lentiginous melanomas harbor BRAF
for BRAF mutations; crizotinib, capmatinib, and tepotinib for mutations, respectively, and 5–15% and 15–30% of mucosal
MET alterations (including exon 14 skipping mutations); and and acral melanomas have NRAS mutations [17–21]. Inter-
crizotinib, ceritinib, lorlatinib, entrectinib, repotrectinib, and estingly, oncogenic mutations in the KIT gene are more clin-
taletrectinib for ROS1 gene rearrangements [6]. Long con- ically relevant, as these mutations occur in 11–21% of cases
sidered non-targetable and non-actionable KRAS mutations [21]. In uveal melanoma, mutations occurring in GNAQ or
are now considered a likely target for novel tyrosine kinase GNA11 gene are found in approximately 80% of the cases,
inhibitors. Several small molecule inhibitors have shown and BRAF and NRAS gene mutations are rare [22, 23].
promising activity in patients with a G12C KRAS mutation These genetic alterations in melanoma are kinase-­activating
in early-phase trials and are currently in ­registration trials mutations, with a potential for developing molecularly tar-
with a hope for FDA approval in the near future [7]. geted therapeutic drugs. For BRAF mutations in patients
In this chapter, we will review advances in molecular tar- with cutaneous melanoma and KIT mutations in those with
geted therapies in malignant melanoma in detail to exem- acral or mucosal melanoma, effective targeted drugs are
plify the concept of personalized cancer therapy. now available.

70.2 Melanoma 70.2.1 BRAF-Mutant Melanoma

Skin cancer is the most common type of malignancy in the 70.2.1.1 BRAF Inhibitors
United States, and melanoma is the deadliest among the BRAF is a serine–threonine kinase protein substrate in the
skin cancers. The incidence of melanoma has been increas- MAP kinase pathway, with a critical function in melanoma
ing significantly and steadily over the past decades [8]. In cells. Oncogenic mutations in BRAF occur in nearly 50%
2020, approximately 100,350 patients are estimated to be of melanomas, and a vast majority of the mutations occur
diagnosed with melanoma, and 6859 patients are expected at codon 600, among which V600E and V600K mutations
to die from the disease in the United States [9]. Fortunately, account for 80 and 15%, respectively [16]. The V600 BRAF
the estimated number of deaths has decreased since 2016, in mutations result in constitutive activation of the MEK and
which more than 10,000 patients were expected to die. This ERK protein in the MAP kinase pathway, which in turn
is owing in a large part to the development of novel thera- leads to proliferation, invasion, and survival of melanoma
pies, such as checkpoint inhibitors and molecular targeted cells (Fig. 70.1a). Accordingly, inhibition of V600 BRAF
drugs. A number of large randomized clinical trials demon- kinase blocks the activation of MEK and ERK proteins and
strated that overall survival (OS) of patients with advanced, suppresses the MAP kinase pathway, causing cell growth
metastatic melanoma has improved with these novel drugs arrest [16] (Fig. 70.1b). An initial attempt to inhibit the
compared to conventional cytotoxic chemotherapy, and sig- BRAF kinase in melanoma included a trial of sorafenib, an
nificantly higher proportion of patients are alive at 5 years inhibitor of RAF kinases. However, in a phase II study of
in the metastatic setting. For example, the 5-year survival sorafenib, there was no clinical response among 37 patients
70 Personalized Systemic Cancer Therapy 741

Receptor Tyrosine
a b
Kinases

N-Ras N-Ras
BRAF
inhibitor
B-Raf C-RAF B-Raf C-RAF

MEK MEK

ERK1/2 ERK1/2

NUCLEUS NUCLEUS

c d

N-Ras N-Ras
BRAF BRAF
inhibitor inhibitor
B-Raf C-RAF B-Raf C-RAF
MEK
MEK inhibitor MEK

ERK1/2 ERK1/2

NUCLEUS NUCLEUS

Fig. 70.1 a V600 BRAF mutation constitutively activates the MAP BRAF inhibitors, through multiple mechanisms, many of which lead to
kinase pathway, leading to cell proliferation and survival. b Selective reactivation of the MAP kinase pathway. d Addition of MEK inhibitors
BRAF inhibitors suppress the MAP kinase pathway, leading to cell suppresses the MAP kinase pathway more completely, resulting in a
growth arrest and apoptosis. c Over time, cells acquire resistance to delay of resistance to BRAF inhibitor therapy

with metastatic melanoma [24]. Likewise, a combination of bind and inhibit V600E BRAF more specifically compared
sorafenib with carboplatin and paclitaxel in a large phase III to wild-type BRAF, using a scaffold-based crystallographic
randomized study failed to show improvement in progres- drug design approach, and as a consequence, it has greater
sion-free survival (PFS) and OS compared to chemotherapy potency for V600 BRAF than for wild-type BRAF (IC50
alone [25]. These results indicated that sorafenib cannot suf- 31 nM/L versus 100 nM/L) [26, 27]. However, despite prom-
ficiently inhibit the activation of the mutated BRAF-driven ising preclinical efficacy, the initial formulation of vemu-
MAP kinase pathway because it is more potent in inhibiting rafenib (crystalline formulation) was not effective in eliciting
wild-type BRAF and CRAF protein than V600 BRAF, caus- clinical responses because it was not able to achieve the ade-
ing significant toxicity before it inhibits the MAPK pathway quate serum concentration necessary for tumor cell killing.
in patients with V600 BRAF-mutant melanoma. Only after the drug was reformulated as a micro-­precipitated
With the failure of sorafenib as a treatment strategy bulk powder with a higher bioavailability, clinical activity
for melanoma, a search for more specific inhibitors of the was observed. In a phase I study, 32 patients with advanced
mutant BRAF kinase wad intensified. Vemurafenib was the melanoma harboring a V600E BRAF mutation were treated
first selective inhibitor of V600 BRAF-mutant melanoma with 960 mg of vemurafenib orally twice a day, and 2 and 24
to be investigated clinically. Vemurafenib was designed to patients had a complete and partial response, respectively,
742 K. B. Kim and M. Kashani-Sabet

with an overall response rate of 81%. In many patients, met- in the encorafenib arm was 0.68 (95% CI 0.52–0.88; two-­
abolic responses assessed by PET/CT scan were observed sided p = 0.0038) with a median PFS of 9.6 months, com-
within 2 weeks of treatment. In a phase III randomized study pared with 7.3 months for the vemurafenib arm [32]. The HR
that enrolled 675 treatment-naive patients with advanced for OS in the encorafenib arm was 0.76 (95% CI 0.58–0.98;
melanoma harboring a V600E BRAF mutation (including 19 two-sided p = 0.033) with a median OS of 23.5 months, com-
patients with V600K mutation), those who were treated with pared to 16.9 months with vemurafenib. Overall response
vemurafenib had significant progression-­free survival (PFS) rates assessed by a central radiology review panel were 52%
and OS advantage over those who were treated with dacar- (95% CI 44–59) and 41% (95% CI 34–41) in the encorafenib
bazine (HR 0.26 [95% CI 0.20–0.33, p < 0.001] for PFS; and the vemurafenib group, respectively.
HR 0.37 [95% CI 0.26–0.55, p < 0.001) for OS) [28]. The
median PFS was 5.3 months and 1.6 months in the vemu- 70.2.1.2 MEK Inhibitors
rafenib and dacarbazine arm, respectively. The response An alternative to suppressing the MAP kinase pathway by
rates were 48% and 5% in the vemurafenib and dacarbazine direct inhibition of mutant BRAF kinase is through inhibi-
arm, respectively. On the basis of the significant PFS and OS tion of the downstream substrate of the BRAF protein. MEK
benefit, vemurafenib was approved by the FDA in the United kinase is the only known downstream substrate of BRAF,
States in 2011. and thus there has been an increasing interest in the inves-
Likewise, dabrafenib, another selective oral inhibitor tigation of MEK inhibitors in the treatment of V600 BRAF-­
of V600 BRAF, was investigated in patients with unresect- mutant melanoma. However, the clinical development of
able metastatic melanoma. In a phase III randomized study earlier versions of MEK inhibitors, including, CI-1040 and
comparing the clinical benefit of dabrafenib with that of PD0325901, was not successful because of intolerable toxic-
dacarbazine that enrolled 250 treatment-naïve patients, those ity and/or a lack of clinical activity [33] until trametinib, a
who received dabrafenib 150 mg twice daily treatment had specific inhibitor of MEK1 and MEK2 proteins, was shown
a superior PFS (HR 0.30 [95% CI 0.18–0.51; p < 0.0001]), to have promising clinical activity in phase I and II stud-
the primary endpoint of the study, with median PFS dura- ies [34, 35]. In a randomized phase III study, 322 patients
tions of 5.1 months and 2.7 months in the dabrafenib arm and with metastatic V600E or V600K BRAF-mutant mela-
the dacarbazine arm, respectively [29]. The response rate of noma were assigned to receive either trametinib treatment
­dabrafenib was higher (50%) than that of dacarbazine (6%). or cytotoxic chemotherapy [36], with the primary and the
At a median follow-up duration of 16.9 months, OS was also secondary endpoint of PFS and OS, respectively. Patients
longer with dabrafenib treatment, with HR of 0.77 (95% CI in the trametinib arm had a longer PFS (HR 0.45 [95% CI
0.52–1.13) [30]. The median OS durations were 20.0 months 0.33–0.63] p < 0.001) and a higher response rate (22% ver-
and 15.6 months for dabrafenib and chemotherapy, respec- sus 8%) than those in the cytotoxic chemotherapy arm. The
tively, despite the fact that 59% of patients treated with che- median PFS duration was 4.8 months in the trametinib arm
motherapy crossed over to dabrafenib treatment. In 2013, and 1.5 months in the chemotherapy arm. Trametinib was
the FDA approved dabrafenib for treatment of advanced also associated with superior OS over chemotherapy, with
V600E/K BRAF-mutant melanoma. a HR of 0.54 (95% CI 0.32–0.92; p = 0.01). The OS rates
Encorafenib is another selective oral BRAF inhibitor with at 1 and 2 years were 60.9% and 32.0%, respectively, in the
a significantly longer dissociation half-life from V600E-­ trametinib arm, whereas they were 49.6% and 29.4% in the
mutant BRAF than the other two BRAF inhibitors (30 h for chemotherapy arm [37]. On the basis of the superior clinical
encorafenib versus 2 h for dabrafenib and 0.5 h for vemu- benefit over the then-standard chemotherapy, trametinib was
rafenib). This is the only BRAF inhibitor that has been com- approved by the FDA for treatment of V600E or K BRAF-
pared directly with another BRAF inhibitor, vemurafenib, in mutant advanced melanoma in 2013. It is important to note
a clinical trial. In the COLUMBUS trial, which randomized that trametinib has low clinical activity in patients with V600
577 patients with V600E/K BRAF-mutant melanoma at 1:1:1 BRAF-mutant melanoma whose disease had progressed on
ratio to receive vemurafenib, encorafenib, or a combination a previous BRAF inhibitor treatment. In a phase II study of
of encorafenib and binimetinib (a selective MEK) inhibitor, trametinib, none of 40 patients who were previously treated
the clinical benefit of encorafenib was compared to that of with a selective BRAF inhibitor responded to trametinib
vemurafenib, separately from the comparison with the com- treatment [35]. These data strongly suggest that for V600
bination treatment. The updated analysis, with a median BRAF-mutant melanoma, selective BRAF inhibitors are
follow-­up of 36.8 months, revealed statistically significantly favored over trametinib as initial targeted therapy despite
longer PFS and OS in the encorafenib arm (dose of 300 mg a lack of head-to-­head comparison of these two classes of
once daily) over the vemurafenib arm [31]. The HR for PFS drugs.
70 Personalized Systemic Cancer Therapy 743

70.2.1.3  ombination of BRAF Inhibitor

C despite the continued use of BRAF inhibitors. To date, no
and MEK Inhibitor new mutations in the BRAF gene have been reported in the
Although the selective BRAF inhibitors significantly setting of treatment resistance. These observations were cor-
improve clinical responses and overall survival, most patients roborated by preclinical studies which showed that addition
who are treated with BRAF inhibitors ultimately have dis- of a MEK inhibitor to a BRAF inhibitor delayed resistance
ease progression. There are multiple mechanisms by which to treatment in V600E BRAF-mutant melanoma cells [38,
melanoma cells escape BRAF inhibitor-induced cytotoxic- 40–42] (Fig. 70.1d). To date, three different combination
ity over time. These include development of new kinase-­ regimens of MEK inhibitor and BRAF inhibitor have been
activating mutations in the NRAS gene, alternate splicing investigated clinically in patients with V600 BRAF-mutant
of BRAF, amplification of BRAF and/or CRAF, acquisition advanced melanoma: dabrafenib (BRAF inhibitor) and tra-
of new MEK mutations, or overexpression of COT. Impor- metinib (MEK inhibitor), vemurafenib (BRAF inhibitor)
tantly, all of these molecular alterations can bypass the inhi- and ­cobimetinib (MEK inhibitor), and encorafenib (BRAF
bition of V600 BRAF kinase by BRAF inhibitors [38–42] inhibitor) and binimetinib (MEK inhibitor). In separate large
(Fig. 70.1c). Although the resistance to BRAF inhibitors can randomized phase III trials, these three regimens were com-
be induced by the activation of receptor tyrosine kinases, such pared to vemurafenib alone with primary endpoints of OS
as the MET receptor, and NRAS mutation, which also acti- [44–46]. Each of these combination regimens was associated
vates the alternate survival pathway, the PI3K/AKT/mTOR with higher response rates, as well as prolonged PFS and OS,
pathway [43], approximately 70% of the resistance mecha- than vemurafenib alone (Table 70.1). The HRs for PFS were
nisms involve the reactivation of the MAP kinase pathway 0.51–0.61 with the combination regimens, with median PFS

Table 70.1 Pivotal phase III studies of targeted therapies for advanced BRAF-mutant melanoma
Primary No.
Study Treatment endpoint pts OS PFS ORR
BRIM3 study Vemurafenib vs dacarbazine OS and 675 HR 0.70 [95% CI HR 0.38 [95% CI 57% (Vem) vs 9%
[28] PFS 0.57–0.87]; p = 0.0008 0.32–0.46]; p < 0.0001 (dacarbazine)
Median 13.6 mo (Vem) Median 6.9 mo (Vem)
vs 9.7 mo (dacarbazine) vs 1.6 mo (dacarbazine)
BREAK3 study Dabrafenib vs dacarbazine PFS 250 HR 0.82 [95% CI HR 0.30 [95% CI 50% (Dab) vs 6%
[29, 30] 0.57–1.18]; p, n/a 0.18–0.51]; p < 0·0001 (dacarbazine)
Median 20.0 mo (Dab) Median 5.1 mo (Dab)
vs 15.6 mo vs 2.7 mo (dacarbazine)
(dacarbazine)
COMBI-D Dabrafenib + trametinib vs PFS 423 HR 0.71 [95% CI HR 0.67 [95% CI 69% (Dab + Tram)
study [48] dabrafenib 0.55–0.92]; p = 0.0107 0.53–0.84]; p = 0.0004 vs 53% (Dab)
Median 25.1 mo Median 11 mo
(Dab + Tram) vs 18.7 (Dab + Tram) vs 8.8
mo (Dab) mo (Dab)
COMBI-V Dabrafenib + trametinib vs OS 704 HR 0.69 [95% CI HR 0.56 [95% CI 64% (Dab + Tram)
study [44] vemurafenib 0.53–0.89]; p = 0.005 0.46–0.69]; p < 0.001 vs 51% (Vem)
Median 26 mo Median 11.4 mo
(Dab + Tram) vs 16.8 (Dab + Tram) vs 7.4
mo (Vem) mo (Vem)
Co-BRIM study Vemurafenib + cobimetinib vs PFS 495 HR 0.70 [95% CI HR 0.58 [95% CI 69.6%
[45] vemurafenib 0.55–0.90]; p = 0.005 0.46–0.72]; p < 0.0001 (Vem + Cobi) vs
Median 22.3 mo Median 12.3 mo 50.0% (Vem)
(Vem + Cobi) vs 17.4 (Vem + Cobi) vs 7.2
mo (Vem) mo (Vem)
COLUMBUS Encorafenib + binimetinib vs PFS 577 For Enco + Bini vs For Enco + Bini vs 64% (Enco + Bini)
study [31, 46] encorafenib vs vemurafenib Vem: HR 0.61 [95% CI Vem: HR 0.51 [95% CI vs 41% (Enco) vs
0.47–0.79]; p < 0.0001 0.39–0.67]; p < 0.0001 40% (Vem)
Median 33.6 mo Median 14.9 mo
(Enco + Bini) vs 16.9 (Enco + Bini) vs 7.3
mo (Vem) mo (Vem)
No. Pts number of patients, OS overall survival, PFS progression-free survival, ORR objective response rate, mo months, HR hazard ratio, CI
confidence interval, p p-value, Vem vemurafenib, Dab dabrafenib, Tram trametinib, Cobi cobimetinib, Enco encorafenib, Bini binimetinib, n/a not
available
744 K. B. Kim and M. Kashani-Sabet

durations of 11–14.9 months, compared to 7.2–7.3 months in the adjuvant setting in patients with surgically resected
with vemurafenib treatment. The HRs for OS were 0.61– V600E/K BRAF-mutant stage III melanoma [50]. In a large
0.70 with the combination regimens, with median OS dura- phase III study with 870 patients who had resected stage III
tions of 22.3–33.6 months, compared to 16.9–18.0 months melanoma, the combination of dabrafenib and trametinib
with vemurafenib treatment [31]. Although the combination was compared to matching placebo treatments; at 5 years,
of encorafenib and binimetinib had numerically the longest 52% (95% CI 48–58) and 36% (95% CI 32–41) of patients
PFS and OS durations among the three combination regi- were alive without melanoma relapse in the BRAF/MEK
mens, it is difficult to compare the overall clinical benefit of inhibitor treatment group and placebo group, respectively.
the three regimens as they have not been compared head to As of April 2018, FDA granted approval to dabrafenib and
head within the same trial. trametinib in combination for adjuvant treatment of this
In general, the combination regimens are well tolerated. patient population.
Grades 3 or 4 adverse events occurred in 41–62% of patients,
and most of these events were reversible upon treatment dis-
continuation [44–46]. There were no appreciable differences 70.2.2 KIT-Mutant Melanoma
in frequency of severe adverse events between the combi-
nation regimens and single-agent BRAF inhibitor. However, Although BRAF mutations are the most frequent genetic
each of the combination regimens has somewhat different abnormalities found in cutaneous melanomas, different
safety profiles. For example, pyrexia and chills were more genetic mutations are more relevant in other melanoma sub-
the common adverse events for the combination of dab- types. In the case of acral lentiginous or mucosal melanomas,
rafenib and trametinib, whereas diarrhea, nausea, and skin mutations in the KIT receptor kinase are more common than
rash were the most common for the vemurafenib and cobi- BRAF mutations: KIT mutations are found in 15–21% and
metinib regimen [45]. For the encorafenib and binimetinib 11% of mucosal and acral lentiginous melanoma, respec-
regimen, diarrhea, nausea/vomiting, and fatigue were the tively [21, 51]. In comparison, V600 BRAF mutation occurs
most common toxicities [46]. only in approximately 3% of mucosal melanomas, with a
Interestingly, certain adverse events occur less frequently higher percentage (13%) of non-V600 BRAF mutations [51].
with the combination approach than with the BRAF inhibi- In the early 2000s, there was a great interest in investigating
tors alone. Development of keratoacanthomas and/or cuta- KIT inhibitors in patients with advanced melanoma because
neous squamous cell carcinomas occurred in 9–26% of of a high frequency of KIT overexpression in melanoma
patients during treatment with single-agent vemurafenib or [52]. There were three phase II studies of imatinib, a potent
dabrafenib [47, 48], whereas they occurred only in 1–3% oral KIT, BCR-ABL, and PDGFR inhibitor, conducted in the
of patients when they were treated with a BRAF inhibitor United States and in Europe, enrolling patients with unresect-
in combination with a MEK inhibitor [31, 44, 45, 48]. This able, metastatic melanoma. Among 62 patients treated with
observation is consistent with the hypothesis that BRAF imatinib in the three studies, there was only one patient with
inhibitors that are more specific for V600 BRAF kinase lead to a clinical response, concluding that imatinib was deemed
a paradoxical activation of the MAP kinase signal transduc- ineffective in melanoma [53–55]. However, these studies
tion pathway through a dimerization with other RAF kinases enrolled patients with melanoma regardless of KIT mutation
(such as CRAF) in keratinocytes that do not harbor mutant status, and it is likely that the vast majority of patients in
BRAF [49]. Accordingly, when a MEK inhibitor is added to these studies did not have a KIT-mutant melanoma.
a BRAF inhibitor, the MAP kinase pathway is inhibited at Since the reporting of the frequency of KIT mutations in
the level downstream of the RAF kinases, and this inhibition, acral and mucosal melanoma [21], enthusiasm for studying
in turn, suppresses the development of keratoacanthoma or imatinib in patients with KIT-mutant advanced melanoma
cutaneous squamous cell carcinoma. Fortunately, occurrence was revived. In three initial single-arm phase II trials of a
of other serious secondary malignancies is rare with either potent inhibitor, imatinib, response rates were 16–29%, with
single-agent BRAF treatment or the combination regimens. a median PFS duration of 3–3.7 months among 118 patients
On the basis of the significant PFS and OS benefits, each with advanced melanoma harboring a KIT mutation [56–58].
of the three BRAF/MEK inhibitor combination regimens Among the mutations occurring in the KIT kinase, clinical
was approved by the FDA for treatment of advanced V600 responses to imatinib treatment were observed mostly in
BRAF-mutant melanoma: dabrafenib and trametinib in 2013, patients with mutations occurring in exon 9 or 11. Because
vemurafenib and cobimetinib in 2015, and encorafenib and of the rather small numbers of patients in each of these
binimetinib in 2018. single-­arm trials, imatinib has not been filed for FDA review
In addition to the significant clinical benefit including OS for treatment of KIT-mutant melanoma. However, imatinib
benefits in patients with metastatic disease, a co-inhibition has been widely used by many oncologists for patients with
of BRAF and MEK kinases improved relapse-free survival a KIT-mutant melanoma.
70 Personalized Systemic Cancer Therapy 745

Nilotinib is a second-generation tyrosine kinase inhibitor References

of BCR-ABL, KIT, and PDGFR-alpha which was approved
1. King CR, Kraus MH, Aaronson SA. Amplification of a novel
for treatment of imatinib-resistant CLL. In a phase II study v-erbB-related gene in a human mammary carcinoma. Science.
of nilotinib, 42 patients with advanced melanoma harboring 1985;229(4717):974–6.
a KIT mutation were treated. While a response rate of 26.2% 2. Karunagaran D, Tzahar E, Beerli RR, Chen X, Graus-Porta D,
was observed, with a median PFS of 4.2 months, nilotinib Ratzkin BJ, et al. ErbB-2 is a common auxiliary subunit of NDF
and EGF receptors: implications for breast cancer. EMBO J.
did not appear to be more effective than imatinib [59]. 1996;15(2):254–64.
3. Sawyers CL. Signal transduction pathways involved in BCR-ABL
transformation. Baillieres Clin Haematol. 1997;10(2):223–31.
70.2.3 Future Direction 4. O'Brien SG, Guilhot F, Larson RA, Gathmann I, Baccarani M,
Cervantes F, et al. Imatinib compared with interferon and low-dose
cytarabine for newly diagnosed chronic-phase chronic myeloid leu-
The development of selective BRAF and MEK inhibitors epit- kemia. N Engl J Med. 2003;348(11):994–1004.
omizes the success of molecular targeted therapy for V600 5. Nickerson ML, Jaeger E, Shi Y, Durocher JA, Mahurkar S, Zaridze
BRAF-mutant melanoma and, at the same time, poses new D, et al. Improved identification of von Hippel-Lindau gene altera-
tions in clear cell renal tumors. Clin Cancer Res. 2008;14(15):
challenges. Despite the high response rates and prolonged 4726–34.
tumor control and survival of patients treated with this combi- 6. Konig D, Savic Prince S, Rothschild SI. Targeted therapy in
nation approach, a majority of patients ultimately experience advanced and metastatic non-small cell lung cancer. An update on
disease progression and succumb to the disease. Likewise, the treatment of the most important actionable oncogenic driver altera-
tions. Cancers (Basel). 2021;13(4):804.
KIT inhibitors render a meaningful clinical benefit in patients 7. Uprety D, Adjei AA. KRAS: From undruggable to a druggable can-
with KIT-mutant melanoma, but the degree of clinical benefit cer target. Cancer Treat Rev. 2020;89:102070.
is suboptimal. A key to making further advances in targeted 8. Guy GP Jr, Thomas CC, Thompson T, Watson M, Massetti GM,
therapy for melanoma lies in a better understanding of the Richardson LC, et al. Vital signs: melanoma incidence and mor-
tality trends and projections—United States, 1982–2030. MMWR
molecular biology of melanoma that would enable the devel- Morb Mortal Wkly Rep. 2015;64(21):591–6.
opment of more effective and less toxic drugs. Currently, 9. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer
many clinical trials incorporate tumor and/or blood collections J Clin. 2020;70(1):7–30.
for various biomarker analyses to discover novel mechanisms 10. Balch CM, Gershenwald JE, Soong SJ, Thompson JF, Atkins MB,
Byrd DR, et al. Final version of 2009 AJCC melanoma staging and
of drug resistance, and these efforts will likely lead to more classification. J Clin Oncol. 2009;27(36):6199–206.
prolonged disease control and longer survival in melanoma 11. Robert C, Ribas A, Schachter J, Arance A, Grob JJ, Mortier L,
patients. In addition, identification of new actionable molecu- et al. Pembrolizumab versus ipilimumab in advanced melanoma
lar targets through genetic and epigenetic analyses will result (KEYNOTE-006): post-hoc 5-year results from an open-label,
multicentre, randomised, controlled, phase 3 study. Lancet Oncol.
in development of novel classes of effective targeted therapy 2019;20(9):1239–51.
in patients whose diseases do not harbor a BRAF or KIT muta- 12. Larkin J, Chiarion-Sileni V, Gonzalez R, Grob JJ, Rutkowski
tion or become resistant to the currently available targeted P, Lao CD, et al. Five-year survival with combined nivolumab
therapy regimens. Furthermore, it will be essential to under- and ipilimumab in advanced melanoma. N Engl J Med.
2019;381(16):1535–46.
stand the effect of the molecular targeted drugs on activation 13. Robert C, Grob JJ, Stroyakovskiy D, Karaszewska B, Hauschild
of immune cells and to develop optimal sequences of targeted A, Levchenko E, et al. Five-year outcomes with dabrafenib
therapy and checkpoint immunotherapy in order to further plus trametinib in metastatic melanoma. N Engl J Med.
improve outcomes associated with melanoma therapy. 2019;381(7):626–36.
14. Sensi M, Nicolini G, Petti C, Bersani I, Lozupone F, Molla A,
et al. Mutually exclusive NRASQ61R and BRAFV600E mutations
at the single-cell level in the same human melanoma. Oncogene.
Conclusion and Perspective 2006;25(24):3357–64.
• This chapter has covered the general principle of 15. Reddy BY, Miller DM, Tsao H. Somatic driver mutations in mela-
noma. Cancer. 2017;123(S11):2104–17.
personalized cancer therapy, including the identifi- 16. Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg
cation of BRAF and KIT mutations in melanoma S, et al. Mutations of the BRAF gene in human cancer. Nature.
and the development of molecular targeted drugs 2002;417(6892):949–54.
for melanomas harboring these mutations. In addi- 17. Tacastacas JD, Bray J, Cohen YK, Arbesman J, Kim J, Koon HB,
et al. Update on primary mucosal melanoma. J Am Acad Dermatol.
tion, the development of various mechanisms of 2014;71(2):366–75.
drug resistance and strategies to overcome acquired 18. Cosgarea I, Ugurel S, Sucker A, Livingstone E, Zimmer L, Ziemer
drug resistance were discussed. The reader should M, et al. Targeted next generation sequencing of mucosal mela-
understand the concept of molecular targeted drugs nomas identifies frequent NF1 and RAS mutations. Oncotarget.
2017;8(25):40683–92.
and how these drugs can improve the clinical out- 19. Dika E, Veronesi G, Altimari A, Riefolo M, Ravaioli GM, Piraccini
come of patients with advanced cancer. BM, et al. BRAF, KIT, and NRAS mutations of acral melanoma in
white patients. Am J Clin Pathol. 2020;153(5):664–71.
746 K. B. Kim and M. Kashani-Sabet

20. Zebary A, Omholt K, Vassilaki I, Hoiom V, Linden D, Viberg L, with BRAF V600 E/K-mutant advanced or metastatic melanoma.
et al. KIT, NRAS, BRAF and PTEN mutations in a sample of Eur J Cancer. 2019;109:61–9.
Swedish patients with acral lentiginous melanoma. J Dermatol Sci. 38. Nazarian R, Shi H, Wang Q, Kong X, Koya RC, Lee H, et al.
2013;72(3):284–9. Melanomas acquire resistance to B-RAF(V600E) inhibition by
21. Curtin JA, Busam K, Pinkel D, Bastian BC. Somatic activa- RTK or N-RAS upregulation. Nature. 2010;468(7326):973–7.
tion of KIT in distinct subtypes of melanoma. J Clin Oncol. 39. Emery CM, Vijayendran KG, Zipser MC, Sawyer AM, Niu L, Kim
2006;24(26):4340–6. JJ, et al. MEK1 mutations confer resistance to MEK and B-RAF
22. Van Raamsdonk CD, Bezrookove V, Green G, Bauer J, Gaugler L, inhibition. Proc Natl Acad Sci USA. 2009;106(48):20411–6.
O'Brien JM, et al. Frequent somatic mutations of GNAQ in uveal 40. Johannessen CM, Boehm JS, Kim SY, Thomas SR, Wardwell L,
melanoma and blue naevi. Nature. 2009;457(7229):599–602. Johnson LA, et al. COT drives resistance to RAF inhibition through
23. Van Raamsdonk CD, Griewank KG, Crosby MB, Garrido MC, MAP kinase pathway reactivation. Nature. 2010;468(7326):
Vemula S, Wiesner T, et al. Mutations in GNA11 in uveal mela- 968–72.
noma. N Engl J Med. 2010;363(23):2191–9. 41. Poulikakos PI, Persaud Y, Janakiraman M, Kong X, Ng C, Moriceau
24. Eisen T, Ahmad T, Flaherty KT, Gore M, Kaye S, Marais R, et al. G, et al. RAF inhibitor resistance is mediated by dimerization
Sorafenib in advanced melanoma: a phase II randomised discon- of aberrantly spliced BRAF(V600E). Nature. 2011;480(7377):
tinuation trial analysis. Br J Cancer. 2006;95(5):581–6. 387–90.
25. Flaherty KT, Lee SJ, Zhao F, Schuchter LM, Flaherty L, Kefford 42. Trunzer K, Pavlick AC, Schuchter L, Gonzalez R, McArthur GA,
R, et al. Phase III trial of carboplatin and paclitaxel with or without Hutson TE, et al. Pharmacodynamic effects and mechanisms of
sorafenib in metastatic melanoma. J Clin Oncol. 2013;31(3):373–9. resistance to vemurafenib in patients with metastatic melanoma. J
26. Tsai J, Lee JT, Wang W, Zhang J, Cho H, Mamo S, et al. Discovery Clin Oncol. 2013;31(14):1767–74.
of a selective inhibitor of oncogenic B-Raf kinase with potent anti- 43. Straussman R, Morikawa T, Shee K, Barzily-Rokni M, Qian ZR, Du
melanoma activity. Proc Natl Acad Sci USA. 2008;105(8):3041–6. J, et al. Tumour micro-environment elicits innate resistance to RAF
27. Bollag G, Hirth P, Tsai J, Zhang J, Ibrahim PN, Cho H, et al. inhibitors through HGF secretion. Nature. 2012;487(7408):500–4.
Clinical efficacy of a RAF inhibitor needs broad target blockade in 44. Robert C, Karaszewska B, Schachter J, Rutkowski P, Mackiewicz
BRAF-mutant melanoma. Nature. 2010;467(7315):596–9. A, Stroiakovski D, et al. Improved overall survival in mela-
28. Chapman PB, Hauschild A, Robert C, Haanen JB, Ascierto P, Larkin noma with combined dabrafenib and trametinib. N Engl J Med.
J, et al. Improved survival with vemurafenib in melanoma with 2015;372(1):30–9.
BRAF V600E mutation. N Engl J Med. 2011;364(26):2507–16. 45. Larkin J, Ascierto PA, Dreno B, Atkinson V, Liszkay G, Maio M,
29. Hauschild A, Grob JJ, Demidov LV, Jouary T, Gutzmer R, Millward et al. Combined vemurafenib and cobimetinib in BRAF-mutated
M, et al. Dabrafenib in BRAF-mutated metastatic melanoma: melanoma. N Engl J Med. 2014;371(20):1867–76.
a multicentre, open-label, phase 3 randomised controlled trial. 46. Dummer R, Ascierto PA, Gogas HJ, Arance A, Mandala M,
Lancet. 2012;380(9839):358–65. Liszkay G, et al. Encorafenib plus binimetinib versus vemu-
30. Hauschild A, Grobb J, Demidov L, Jouary T, Gutzmer R, Millward rafenib or encorafenib in patients with BRAF-mutant melanoma
M, et al. An update on overall survival (OS) and follow-on ther- (COLUMBUS): a multicentre, open-label, randomised phase 3
apies in BREAK-3, a phase III, randomized trial: dabrafenib trial. Lancet Oncol. 2018;19(5):603–15.
(D) vs. dacarbazine (DTIC) in patients (pts) with BRAF V600E 47. Lacouture ME, Duvic M, Hauschild A, Prieto VG, Robert
mutation-­ positive metastatic melanoma (MM). Ann Oncol. C, Schadendorf D, et al. Analysis of dermatologic events in
2014;25(Supplement 4):4374–493. vemurafenib-treated patients with melanoma. Oncologist.
31. Dummer R, Ascierto PA, Gogas HJ, Arance A, Mandala M, Liszkay 2013;18(3):314–22.
G, et al. Overall survival in patients with BRAF-mutant mela- 48. Long GV, Stroyakovskiy D, Gogas H, Levchenko E, de Braud F,
noma receiving encorafenib plus binimetinib versus vemurafenib Larkin J, et al. Combined BRAF and MEK inhibition versus BRAF
or encorafenib (COLUMBUS): a multicentre, open-label, ran- inhibition alone in melanoma. N Engl J Med. 2014;371(20):1877–88.
domised, phase 3 trial. Lancet Oncol. 2018;19(10):1315–27. 49. Su F, Viros A, Milagre C, Trunzer K, Bollag G, Spleiss O, et al.
32. Dummer R, Ascierto PA, Gogas H, Arance AM, Mandalà M, RAS mutations in cutaneous squamous-cell carcinomas in patients
Liszkay G, et al. Overall survival in COLUMBUS: a phase 3 treated with BRAF inhibitors. N Engl J Med. 2012;366(3):207–15.
trial of encorafenib (ENCO) plus binimetinib (BINI) vs vemu- 50. Dummer R, Hauschild A, Santinami M, Atkinson V, Mandala
rafenib (VEM) or enco in BRAF-mutant melanoma. J Clin Oncol. M, Kirkwood JM, et al. Five-year analysis of adjuvant dab-
2018;36(15_suppl):9504. rafenib plus trametinib in stage III melanoma. N Engl J Med.
33. Salama AK, Kim KB. MEK inhibition in the treatment of advanced 2020;383(12):1139–48.
melanoma. Curr Oncol Rep. 2013;15(5):473–82. 51. Garrido MC, Bastian BC. KIT as a therapeutic target in melanoma.
34. Falchook GS, Lewis KD, Infante JR, Gordon MS, Vogelzang NJ, J Investig Dermatol. 2010;130(1):20–7.
DeMarini DJ, et al. Activity of the oral MEK inhibitor trametinib in 52. Shen SS, Zhang PS, Eton O, Prieto VG. Analysis of protein tyro-
patients with advanced melanoma: a phase 1 dose-escalation trial. sine kinase expression in melanocytic lesions by tissue array. J
Lancet Oncol. 2012;13(8):782–9. Cutan Pathol. 2003;30(9):539–47.
35. Kim KB, Kefford R, Pavlick AC, Infante JR, Ribas A, Sosman JA, 53. Ugurel S, Hildenbrand R, Zimpfer A, La Rosee P, Paschka P,
et al. Phase II study of the MEK1/MEK2 inhibitor trametinib in Sucker A, et al. Lack of clinical efficacy of imatinib in metastatic
patients with metastatic BRAF-mutant cutaneous melanoma pre- melanoma. Br J Cancer. 2005;92(8):1398–405.
viously treated with or without a BRAF inhibitor. J Clin Oncol. 54. Wyman K, Atkins MB, Prieto V, Eton O, McDermott DF, Hubbard
2013;31(4):482–9. F, et al. Multicenter phase II trial of high-dose imatinib mesylate in
36. Flaherty KT, Robert C, Hersey P, Nathan P, Garbe C, Milhem M, metastatic melanoma: significant toxicity with no clinical efficacy.
et al. Improved survival with MEK inhibition in BRAF-mutated Cancer. 2006;106(9):2005–11.
melanoma. N Engl J Med. 2012;367(2):107–14. 55. Kim KB, Eton O, Davis DW, Frazier ML, McConkey DJ, Diwan
37. Robert C, Flaherty K, Nathan P, Hersey P, Garbe C, Milhem M, AH, et al. Phase II trial of imatinib mesylate in patients with meta-
et al. Five-year outcomes from a phase 3 METRIC study in patients static melanoma. Br J Cancer. 2008;99(5):734–40.
70 Personalized Systemic Cancer Therapy 747

56. Carvajal RD, Antonescu CR, Wolchok JD, Chapman PB, Roman mutationally activated or amplified KIT arising on mucosal, acral,
RA, Teitcher J, et al. KIT as a therapeutic target in metastatic mela- and chronically sun-damaged skin. J Clin Oncol. 2013;31(26):
noma. JAMA. 2011;305(22):2327–34. 3182–90.
57. Guo J, Si L, Kong Y, Flaherty KT, Xu X, Zhu Y, et al. Phase II, 59. Guo J, Carvajal RD, Dummer R, Hauschild A, Daud A, Bastian BC,
open-label, single-arm trial of imatinib mesylate in patients with et al. Efficacy and safety of nilotinib in patients with KIT-mutated
metastatic melanoma harboring c-Kit mutation or amplification. J metastatic or inoperable melanoma: final results from the global,
Clin Oncol. 2011;29(21):2904–9. single-arm, phase II TEAM trial. Ann Oncol. 2017;28(6):1380–7.
58. Hodi FS, Corless CL, Giobbie-Hurder A, Fletcher JA, Zhu M,
Marino-Enriquez A, et al. Imatinib for melanomas harboring
 Cancer Stem Cells and Their Role
in Metastasis 71
Ruby Ghadially, Richard W. Kim,
and Alexandra Charruyer-Reinwald

Abstract
Learning Objectives
Cancer stem cells have been a major focus of study in • Understand the concept of a cancer stem cell.
recent years. However, the more we learn about them, the • Understand how markers can be used to identify
more challenges we seem to face. Cancer stem cells, cells cancer stem cells.
responsible for ongoing tumor propagation, have been iso- • Understand the signaling pathways used by cancer
lated from many tumors. In many tumors the cancer stem stem cells to produce metastases.
cell or a subset of cancer stem cells is the cell responsible • Understand the role of cancer stem cells in cancer
for metastasis. Findings make it clear that the cancer stem progression and metastasis.
cell is a key therapeutic target, but given clonal evolution
and phenotype instability of the cancer stem cell, the non-
cancer stem cells must be targeted simultaneously.
Targeting cancer stem cells can be difficult due to their rar-
ity and our inability to propagate them in culture. However, 71.1 Introduction
targeting the pathways that maintain stemness or targeting
microenvironmental controls of the stem cell state can cir- Not all cells in a cancer are equally proliferative and inva-
cumvent this problem. sive. In fact, most are not capable of producing a further
tumor if separated from the parent tumor. Those that are
capable of initiating further tumors in xenograft models have
been termed tumor-initiating cells or cancer stem cells
(CSCs). These cells usually constitute a small proportion of
Abbreviations the total cancer cell population.
It was initially believed that CSCs arise only from normal
SC Stem cell stem cells (SCs). However, it was soon shown that more dif-
CSC Cancer stem cell ferentiated progenitor cells could acquire mutations that
resulted in the abilities to proliferate indefinitely and to gen-
erate tumors when transplanted into a new host, qualifying
them as CSCs. The CSC state can be regulated by intrinsic
and extrinsic factors. Intrinsic regulators include genetic and
epigenetic factors. Extrinsic factors include the niche and the
immune system [42].
R. Ghadially (*) · A. Charruyer-Reinwald
Department of Dermatology, Eli and Edythe Broad Center of
Regeneration Medicine and Stem Cell Research, University of
California San Francisco, San Francisco, CA, USA 71.2 Markers of Cancer Stem Cells
e-mail: [email protected]
R. W. Kim Different types of markers are used to identify and/or isolate
School of Medicine, University of California San Francisco, CSCs and include cell surface molecules, functional mark-
San Francisco, CA, USA ers, transcription factors, and signaling pathway molecules.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 749
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_71
750 R. Ghadially et al.

Functional markers include ALDH1, ABCG2, and side implicated in the promotion of cell migration and invasion in
population cells. ALDH1 catalyzes the oxidation of retinal ovarian cancer [49]. Furthermore, knockdown of Sox2
(retinaldehyde), the retinol metabolite, to retinoic acid (RA), decreases the growth of prostate CSCs in anchorage-­
and ALDH1 plays a vital role as a marker of SCs and CSCs independent cultures [67].
[75]. CSCs from melanoma [8, 50, 83], breast cancer [21], Nanog is a transcription factor that helps maintain embry-
pancreatic adenocarcinoma [63], prostate cancer [31], onic SC properties. The level of Nanog expression is higher
colorectal cancer [33], lung cancer [38, 76], and endometri- in CSCs than in non-stem cancer cells in many cancer types,
oid adenocarcinoma [62] express high levels of ALDH1. including renal carcinoma [9], breast cancer [55], colorectal
ABCG2 is expressed in both normal SCs and CSCs and is cancer [35], prostate cancer [37], gliomas [58, 84], and endo-
a membrane transporter. ABCG2 is a CSC marker in breast metrial adenocarcinoma [88].
cancer [70, 74], hepatocellular carcinoma [85], and ovarian Klf4 belongs to the Kruppel-like transcription factor fam-
cancer [16]. ABCG2 has been shown to reduce the intracel- ily and is important in cell cycle regulation and pluripotency
lular concentrations of multiple anticancer chemotherapeutic [24]. Klf4 is expressed at high levels in colorectal CSCs, and
agents in vitro [3]. knockdown of Klf4 leads to a decrease in sphere formation,
Side population cells are cells that are able to exclude migration, invasion, and epithelial–mesenchymal transition,
Hoechst 33342 dye reflecting their high drug efflux capacity. demonstrating a role for Klr4 in maintaining colorectal CSCs
Side population cells constituted CSCs in neuroblastomas, [43]. Klf4 is also required for maintenance of breast CSCs.
breast cancer, lung cancer, and glioblastoma cell lines [28]. Altogether, studies of Klf4 indicate that Klf4 has an impor-
Side population cells also expressed high levels of ABCG2. tant role not only for self-renewal of CSCs but for cell inva-
Commonly used cell surface markers include CD44, sion and metastasis [24].
CD133, CD24, CD34, CD90, and CD38 [42]. Studies indi- C-Myc is a member of the Myc family of genes and is a
cate a role for CD133 in determining cellular fate and main- regulator of cell proliferation and key for maintaining SC
taining the SC-like state [6, 7, 51]. CD133 identifies CSC pluripotency and proliferation [41]. Myc is elevated in a
populations in multiple solid tumor types [22]. CD133 was large number of poorly differentiated aggressive malignan-
found to be a CSC marker in colon cancer [66], brain cancers cies [82]. c-Myc was expressed at high levels in gliomas [79]
[71, 72], hepatocellular carcinoma [73, 81], lung cancer [17], and in small-cell lung cancer [68]. c-Myc induced self-­
and pancreatic cancer cell lines [59]. renewal capacity in hepatocarcinoma cells [1].
CD44 appears to have a significant regulatory role in most In summary, pluripotency factors Oct4, Nanog, Sox2,
cancer types, although the signaling pathways in tumor cells Klf4, and c-Myc are markers of CSCs that appear to play
remain unresolved. CD24, being an epithelial marker, may crucial roles in cancer growth and progression. Further
not be expressed in all cancers but seems to have a significant investigation is needed to understand the roles of these tran-
role in those tumors expressing CD24 [36]. Breast [2] and scription factors in the CSC phenotype, tumorigenesis, and
prostate [34] CSCs are CD44+ CD24−, while pancreatic tumor progression/metastasis.
[44], ovarian [20, 86], and colorectal [80] CSCs are CD44+ Micro-RNAs (miRNAs), including let-7, miR-22, miR-­
CD24+. 34a, miR-128, the miR-200 family, and miR-451, are
Transcription factors, including Oct4, Sox2, Nanog, expressed in CSCs and are known to regulate the self-­
c-Myc, and Klf4, are part of the core network that regulates renewal, differentiation, epithelial–mesenchymal transition,
pluripotency of embryonic SCs [24]. Research has demon- tumorigenicity, migration, and invasion of CSCs, contribut-
strated the role of these pluripotency factors in CSC pheno- ing to metastatic capability. Thus, through these miRNAs,
types and provided new insights into tumorigenesis and tumor progression and metastasis can be promoted or inhib-
metastasis. ited [24].
Oct4 is essential for maintenance of embryonic SC prop-
erties [57], and studies indicate that Oct4 plays important
roles in tumorigenesis and tumor metastasis. In murine lung 71.3 Stem Cell Divisions and Cancer
carcinoma cells and human breast cancer MCF7 cells, abla-
tion of Oct4 expression leads to apoptosis of CSCs (through Understanding the patterns of SC divisions has been an area
Oct4/Tcl1/Akt1) and inhibition of tumor growth. Also, of extensive research in the past years. In many human
expression of Oct4 in prostate CSCs maintained CSC prop- organs, tissue maintenance appears to involve a balance of
erties [64]. These studies demonstrate a role for Oct4 in CSC SC self-renewal and differentiation (Fig. 71.1). Asymmetric
survival. SC divisions produce one SC and one more differentiated
Sox2 is thought to be critical for the maintenance of Oct4 committed progenitor. Symmetric SC divisions lead to two
expression and hence SC properties [52]. In melanoma SCs or to two differentiated committed progenitor cells.
CSCs, Sox2 is expressed at high levels [69], and Sox2 is Asymmetric SC divisions are identified by asymmetric
71 Cancer Stem Cells and Their Role in Metastasis 751

Fig. 71.1 Schematic of stem Self-Renewal Differentiation

cell self-renewal and
differentiation divisions

Asymmetric Symmetric Symmetric

Stem cell Committed progenitor

segregation of fate determinant proteins such as Numb. The (NICD) from the cell membrane. NICD translocates to the
existence of both symmetric and asymmetric SC divisions nucleus to join a complex that regulates transcription of its
allows for multiple functions. For example, SCs can be lost target genes [26].
by symmetric differentiation, replaced by asymmetric self-­ The role of Notch regulation in CSCs and its importance
renewal division [39], or expanded by symmetric ­self-­renewal for tumor growth has been shown in multiple cancers, includ-
division. In response to injury, symmetric divisions could be ing colon cancer, ovarian cancer, breast cancer, glioblastoma,
“switched on” in response to a sudden SC loss. Meanwhile and chronic myeloid leukemia (CML).
asymmetric division is operational during normal tissue Inhibition of the expression of Notch target genes in
homeostasis. human colon cancer xenografts, with the use of anti-human
A dysregulation in either symmetric or asymmetric divi- neutralizing DLL4 antibodies, was associated with reduced
sions may lead to the loss of homeostatic control and CSC frequency, as shown by flow cytometric and in vivo
unchecked cell growth. Alteration in the control of prolifera- tumorigenicity studies. Furthermore, in breast cancer xeno-
tion/differentiation divisions can lead to abnormal SC expan- grafts, Taxol increased CSCs threefold, while DLL4 antibod-
sion [65]. Similarly, disruption of asymmetric divisions can ies, either alone or in combination with Taxol, decreased
result in excess SC self-renewal divisions and subsequent CSCs [29].
increases in SC number and can be responsible for cancerous In ovarian cancer Notch3 overexpression resulted in
growth of undifferentiated cells [10]. expansion of CSCs and increased tumor resistance to plati-
num. In contrast, both γ-secretase inhibitor (Notch pathway
inhibitor) and Notch3 siRNA knockdown depleted CSCs, as
71.4  ignaling Pathways of Cancer
S shown by decreased side population cells and gene transcrip-
Stem Cells tion of SC markers. This was accompanied by increased
tumor sensitivity to platinum [53]. These studies indicate a
Self-renewal is required for maintenance of malignant role for Notch in maintaining CSCs and in tumor resistance
clones. Multiple signal transduction pathways are involved to platinum.
in the self-renewal of CSCs. These pathways include, but are In breast cancer, Notch4 signaling activity was eightfold
not limited to, Notch, Hedgehog, Wnt, FGF, and TGF-β [24]. higher in CSC-enriched cell populations compared with dif-
Interruption of these pathways provides a promise for thera- ferentiated tumor cells. Pharmacologic (γ-secretase inhibi-
peutics that can successfully kill the CSC [60]. tor) or genetic inhibition (siRNA) of Notch4 reduced ESA+/
Notch Signaling and Cancer Stem Cells: The biological CD44+/CD24low CSCs (flow cytometry), decreased mam-
consequences of Notch activation in SCs range from SC mosphere forming units, and reduced xenograft tumors
maintenance to SC expansion and to promotion of SC dif- in vivo in mice [25].
ferentiation [45]. Initiation of Notch signaling occurs when Notch signaling regulates normal SCs in the brain, and
ligands (DLL/JAG) and receptors (Notch1–Notch4) interact glioblastomas contain stem-like cells with high Notch activ-
on adjacent cells. This interaction leads to proteolytic cleav- ity. Notch blockade by γ-secretase inhibitors reduced neuro-
age of Notch and release of the Notch intracellular domain sphere growth and clonogenicity in vitro, whereas
752 R. Ghadially et al.

overexpression of an active form of Notch2 increased tumor ing hedgehog blockade were unable to form tumors in mice,
growth. The putative CSC markers CD133, Nestin, Bmi1, indicating a role for hedgehog in CSC growth [5].
and Olig2 were reduced following Notch blockade in glio- Human breast CSCs (CD44+CD24−/lowLin−) showed
blastoma neurosphere cultures. Injection of γ-secretase sub- increased expression of hedgehog pathway members, includ-
cutaneously in mice inhibited tumor formation. In vivo ing PTCH1, Gli1, and Gli2 compared with the non-CSC
delivery of γ-secretase by implantation of drug-impregnated population. Activation of hedgehog signaling increased
polymer beads also effectively blocked tumor growth and mammosphere-initiating cell numbers and mammosphere
significantly prolonged survival. Thus, Notch pathway inhi- size, whereas inhibition of hedgehog resulted in a reduction
bition appears to deplete stem-like cancer cells in glioblasto- of these effects [47].
mas, suggesting that γ-secretase inhibitors may be useful as In mouse models of chronic myeloid leukemia, loss of
chemotherapeutic reagents to target CSCs in malignant glio- hedgehog signaling by genetically disrupting Smoothened
mas [18]. resulted in the inhibition of BCR-ABL expressing leukemic
Retroviral expression of Hes1 (a notch pathway effector SCs and prolonged survival. Loss of Smoothened (compo-
protein) in common myeloid progenitors and granulocyte-­ nent of the hedgehog pathway) in mice impaired hematopoi-
macrophage progenitors in the presence of interleukin-3 etic SC self-renewal and decreased induction of chronic
immortalizes these cells in vitro. The protein coded for by the myelogenous leukemia (CML) by the BCR-ABL1 oncopro-
abnormal BCR-ABL1 fusion gene (Philadelphia chromo- tein. Loss of Smoothened caused depletion of chronic
some) is responsible for the uncontrolled growth of leukemic myeloid leukemia CSCs, whereas control hematopoietic
cells. The combination of Hes1 and BCR-ABL in common cells transduced with BCR-ABL1 developed chronic
myeloid progenitors and granulocyte-macrophage progeni- myeloid leukemia in 94% of recipients within 3 months,
tors caused acute leukemia resembling the blast crisis of similarly transduced Smoothened −/− cells developed
chronic myelogenous leukemia and resulted in rapid death of chronic myeloid leukemia in only 47%, and showed increased
the recipient mice. BCR-ABL alone caused chronic myeloid tumor latency. These data suggest that hedgehog is required
leukemia-like disease when expressed in (c-Kit-­positive, Sca- for the initiation and propagation of CML. Further, constitu-
1-positive, lineage-negative) hematopoietic SCs, but not in tively active Smoothened augments chronic myelogenous
committed progenitors (common myeloid progenitors or leukemia CSC number and accelerates disease. As a possible
granulocyte-macrophage progenitors). Leukemic cells mechanism for Smoothened-induced effects, the cell fate
derived from Hes1 and BCR-ABL-expressing common determinant Numb, which depletes chronic myelogenous
myeloid progenitors and granulocyte-macrophage progeni- leukemia CSCs, is increased in the absence of Smoothened
tors were more immature than those derived from BCR-ABL- activity [87].
expressing hematopoietic SCs. The authors thus believe Hes1 The subset of multiple myeloma cells with hedgehog
may be a key molecule in blast crisis transition [56]. pathway activity is located within the CSC compartment.
Hedgehog Signaling and Cancer Stem Cells: The hedge- Hedgehog ligand promotes expansion of multiple myeloma
hog signaling pathway influences the transcription of many CSCs without differentiation, whereas hedgehog pathway
target genes, both during development and in adults [32]. blockade inhibits clonal expansion and is accompanied by
Hedgehog has a major role in maintaining the self-renewing terminal differentiation of purified multiple myeloma CSCs.
capacity of adult somatic SCs and has been widely implicated Thus, hedgehog pathway activation is heterogeneous across
in CSC function and maintenance [11]. Activation of hedge- the spectrum of multiple myeloma CSCs and their more dif-
hog occurs in multiple cancers, and hedgehog s­ignaling has ferentiated progeny [61].
been shown to regulate CSCs in multiple tumors including The above studies validate hedgehog pathway inhibition
pancreatic adenocarcinoma, glioblastoma, breast cancer, mul- as a therapeutic strategy for cancers and their metastatic
tiple myeloma, and chronic myeloid leukemia (CML) [54]. spread. By targeting specific cellular subpopulations likely
In a pancreatic adenocarcinoma xenograft model, involved in tumor initiation at metastatic sites, hedgehog
cyclopamine-­induced blockade of hedgehog signaling pref- inhibitors provide a new paradigm for therapy of dissemi-
erentially reduced “ALDH-high” cells threefold. Metastatic nated malignancies, particularly when used in combination
spread was profoundly inhibited, with only one of seven with conventional antimetabolites that reduce “bulk” tumor
cyclopamine-treated mice developing pulmonary microme- size [19].
tastases versus seven of seven control mice developing mac- WNT/B-Catenin Signaling and CSCs: More than 30
rometastases [19]. extracellular Wnt ligands bind to the receptor complex of
Shh ligand was expressed in some primary glioblastomas Frizzled and LRP5/6 (member of the LDL receptor family)
and in glioblastoma-derived neurospheres. Cyclopropylamine [48]. Wnt signaling regulates the activity of somatic SCs,
completely inhibited glioblastoma-derived neurosphere for- including those in the skin, blood, brain, and the mammary
mation and glioblastoma cells injected intracranially follow- gland, whereas aberrant signaling results in neoplasia [30].
71 Cancer Stem Cells and Their Role in Metastasis 753

The Wnt pathway is activated in CD133(+) colon CSCs 71.6 Targeting CSCs for Cancer Therapy
[13, 14]. Also, in colon adenocarcinomas, high activity of the
Wnt pathway is observed preferentially in tumor cells located Many cancer treatments kill most cancer cells, but CSCs
close to stromal myofibroblasts, indicating that Wnt activity exhibit resistance to therapy, and this may lead to tumor
and cancer stemness may be regulated by extrinsic cues. In relapse and metastasis. Multiple mechanisms exist whereby
agreement with this notion, myofibroblast-secreted factors, CSCs can resist therapeutics.
specifically hepatocyte growth factor, activate beta-catenin-­ Induction of epithelial–mesenchymal transition can lead
dependent transcription and subsequently CSC clonogenic- to the acquisition of resistance to both chemotherapy and
ity. More significantly, myofibroblast-secreted factors also radiotherapy, as shown in ovarian cancer [40] and in breast
restore the CSC phenotype in more differentiated tumor cells cancer [23]. How does epithelial–mesenchymal transition
both in vitro and in vivo. Thus, microenvironmental Wnt sig- lead to chemo- and radio-resistance? Resistance through epi-
naling also has a role in CSC behavior and maintenance [77]. thelial–mesenchymal transition may be attributable to the
lower proliferative rate that results from the acquisition of
mesenchymal properties. Also resistance to chemo in normal
71.5  he Role of CSCs in Cancer
T SCs may be due to high levels of anti-apoptotic proteins as
Progression and Metastasis well as the ABC transporters [60].
It is challenging to perform chemical screens for therapy
CSCs can, by multiple paradigms, account for tumor hetero- of CSCs specifically, first because they are a rare population
geneity. A cancer may arise from a phenotypically uniform of cancer cells and second because if they are successfully
single CSC population. However, in some cases, more than isolated by flow cytometry a pure population is not able to be
one CSC population may exist within a tumor. In other cases, maintained in culture.
CSCs may stay dormant for months or years, often after sur- Since CSCs are so hard to target, targeting the tumor
viving cancer therapies that kill the non-CSCs and become microenvironment is an attractive approach. In colon cancer,
activated to form tumors at a later date. Also, CSC popula- high Wnt pathway activity is observed preferentially in
tions, distinct from the parent population, may develop as a tumor cells located close to stromal myofibroblasts, and
tumor progresses, due to genetic or epigenetic modifications myofibroblast-secreted factors activate β-catenin-dependent
and often associated with greater tumor aggression. Finally, transcription and subsequently CSC clonogenicity.
the CSC phenotype can be unstable, resulting in phenotypic Furthermore, myofibroblast-secreted factors also restore the
reversion and switching of the CSC phenotype. Thus, CSCs CSC phenotype in more differentiated tumor cells both
are dynamic in nature, and it is possible for clonal evolution in vitro and in vivo [77]. Thus, stroma cells appear to have a
and CSCs to exist and act together [78]. role in both inducing and maintaining a SC-like state in the
Metastasis is a multifaceted event requiring loss of cell– CSC.
cell adhesion, intravasation into blood vessels or lymphatic As CSCs and normal SCs share a number of properties,
system, extravasation, as well as a receptive tissue environ- signaling pathways such as Wnt, TGF-b, Notch, and
ment. Not all cells in a tumor are able to move out of the pri- hedgehog pathways are being explored as targets to elimi-
mary tumor and metastasize, and only a small subpopulation nate CSCs. Several inhibitors of the Wnt pathway have
of those that enter the circulation initiate a metastatic lesion been developed. They may be effective in preventing
[4]. This subpopulation is thought to be of the CSC pheno- secretion of active Wnt molecules by stromal cells, block-
type, also known as metastasis-initiating cells (MICs). Using ing paracrine signaling that induces and maintains CSCs
bioluminescence imaging, Liu et al. showed that breast can- [60]. Several molecules targeting the hedgehog and Notch
cer CD44+ CSCs are directly involved in metastasis [46]. pathways have also been developed. Sonidegib, a small
CSC subsets within primary tumors may harbor CSC sub- molecule that inhibits Smoothened (hedgehog pathway
sets with tumor-propagating and/or metastatic capacity. For activator), was approved for adults with locally advanced
example, in pancreatic cancer, both CD133+CXCR4+ cells basal cell carcinoma in 2015. Also, a small molecule
and CD133+CXCR4− cells were tumor-initiating cells, but gamma secretase inhibitor (prevents Notch cleavage),
only CD133+CXCR4+ cells produced metastases. Inhibition MK0752, has been developed, and phase 1 studies were
of CXCR4 signaling reduced the metastatic potential of the published in 2018) [12].
tumors [27]. Likewise, in human colon cancer, murine xeno- Another therapeutic approach is to cause CSCs to differ-
grafts showed that tumor initiation, self-renewal, and metas- entiate into mature non-CSCs. These agents rely on the
tasis formation are limited to particular subpopulations of premise that a defect in terminal differentiation is character-
CSCs [15]. Thus, substantial evidence shows that subpopula- istic of cancer. Finally, there is the idea of using the CSC-­
tions of CSCs within the tumor have specific capabilities in specific antigens for immunotherapy, activating adaptive
regard to tumor maintenance, growth, and metastasis. immune cells against specific antigens [60].
754 R. Ghadially et al.

9. Bussolati B, et al. Identification of a tumor-initiating stem cell pop-

Conclusion and Perspective ulation in human renal carcinomas. FASEB J. 2008;22(10):3696–
705. https://doi.org/10.1096/fj.08-­102590.
• Not all cells in a cancer are equally proliferative and 10. Clarke RB, et al. Regulation of human breast epithelial
invasive, and most are not capable of producing a stem cells. Cell Prolif. 2003;36(Suppl 1):45–8. https://doi.
further tumor if separated from the parent tumor. org/10.1046/j.1365-­2184.36.s.1.5.x.
• A small proportion of the total cancer cell popula- 11. Cochrane CR, et al. Hedgehog signaling in the maintenance
of cancer stem cells. Cancers. 2015;7(3):1554–85. https://doi.
tion is capable of initiating further tumors in xeno- org/10.3390/cancers7030851.
graft models, and these cells have been termed 12. Cook N, et al. A phase I trial of the γ-secretase inhibitor MK-0752 in
tumor-initiating cells or cancer stem cells (CSCs). combination with gemcitabine in patients with pancreatic ductal
• Many cancer treatments kill most cancer cells, but adenocarcinoma. Br J Cancer. 2018;118(6):793–01. https://doi.
org/10.1038/bjc.2017.495.
CSCs exhibit resistance to therapy, and this may 13. Corbo C, et al. Protein cross-talk in CD133+ colon cancer cells
lead to tumor relapse and metastasis. indicates activation of the Wnt pathway and upregulation of
• Since CSCs are so hard to target, targeting the SRp20 that is potentially involved in tumorigenicity. Proteomics.
tumor microenvironment is an attractive approach, 2012;12(12):2045–59. https://doi.org/10.1002/pmic.201100370.
14. Deng Y-H, et al. 5-Fluorouracil upregulates the activity of Wnt
as is targeting signaling pathways such as Wnt, signaling pathway in CD133-positive colon cancer stem-like
TGF-b, Notch, and hedgehog pathways that are key cells. Chin J Cancer. 2010;29(9):810–5. https://doi.org/10.5732/
to CSC survival and growth. cjc.010.10134.
• Can we improve our ability to isolate CSCs in order 15. Dieter SM, et al. Distinct types of tumor-initiating cells form
human colon cancer tumors and metastases. Cell Stem Cell.
to better study this small population of cells? 2011;9(4):357–65. https://doi.org/10.1016/j.stem.2011.08.010.
• Can we target CSC signaling to decrease metastatic 16. Dou J, et al. Using ABCG2-molecule-expressing side population
ability? cells to identify cancer stem-like cells in a human ovarian cell
• Can we combine CSC-targeted treatments with line. Cell Biol Int. 2011;35(3):227–34. https://doi.org/10.1042/
CBI20100347.
other treatment modalities to improve cancer 17. Eramo A, et al. Identification and expansion of the tumorigenic lung
survival? cancer stem cell population. Cell Death Differ. 2008;15(3):504–14.
https://doi.org/10.1038/sj.cdd.4402283.
18. Fan X, et al. NOTCH pathway blockade depletes CD133-positive
glioblastoma cells and inhibits growth of tumor neurospheres and
xenografts. Stem Cells (Dayton, Ohio). 2010;28(1):5–16. https://
doi.org/10.1002/stem.254.
References 19. Feldmann G, et al. Blockade of hedgehog signaling inhibits pan-
creatic cancer invasion and metastases: a new paradigm for combi-
1. Akita H, et al. MYC activates stem-like cell potential in hepa- nation therapy in solid cancers. Cancer Res. 2007;67(5):2187–96.
tocarcinoma by a p53-dependent mechanism. Cancer Res. https://doi.org/10.1158/0008-­5472.CAN-­06-­3281.
2014;74(20):5903–13. https://doi.org/10.1158/0008-­5472. 20. Gao M-Q, et al. CD24+ cells from hierarchically organized
CAN-­14-­0527. ovarian cancer are enriched in cancer stem cells. Oncogene.
2. Al-Hajj M, et al. Prospective identification of tumorigenic breast 2010;29(18):2672–80. https://doi.org/10.1038/onc.2010.35.
cancer cells. Proc Natl Acad Sci USA. 2003;100(7):3983–8. https:// 21. Ginestier C, et al. ALDH1 is a marker of normal and malignant
doi.org/10.1073/pnas.0530291100. human mammary stem cells and a predictor of poor clinical out-
3. An Y, Ongkeko WM. ABCG2: the key to chemoresistance in cancer come. Cell Stem Cell. 2007;1(5):555–67. https://doi.org/10.1016/j.
stem cells? Expert Opin Drug Metab Toxicol. 2009;5(12):1529–42. stem.2007.08.014.
https://doi.org/10.1517/17425250903228834. 22. Glumac PM, LeBeau AM. The role of CD133 in cancer: a concise
4. Baccelli I, et al. Identification of a population of blood circulating review. Clin Transl Med. 2018;7(1):18. https://doi.org/10.1186/
tumor cells from breast cancer patients that initiates metastasis in s40169-­018-­0198-­1.
a xenograft assay. Nat Biotechnol. 2013;31(6):539–44. https://doi. 23. Gupta PB, et al. Identification of selective inhibitors of cancer stem
org/10.1038/nbt.2576. cells by high-throughput screening. Cell. 2009;138(4):645–59.
5. Bar EE, et al. Cyclopamine-mediated hedgehog pathway https://doi.org/10.1016/j.cell.2009.06.034.
inhibition depletes stem-like cancer cells in glioblastoma. 24. Hadjimichael C, et al. Common stemness regulators of embryonic
Stem Cells. 2007;25(10):2524–33. https://doi.org/10.1634/ and cancer stem cells. World J Stem Cells. 2015;7(9):1150–84.
stemcells.2007-­0166. https://doi.org/10.4252/wjsc.v7.i9.1150.
6. Bauer N, et al. New insights into the cell biology of hematopoietic 25. Harrison H, et al. Regulation of breast cancer stem cell activity by
progenitors by studying prominin-1 (CD133). Cells Tissues Organs. signaling through the Notch4 receptor. Cancer Res. 2010;70(2):709–
2008;188(1–2):127–38. https://doi.org/10.1159/000112847. 18. https://doi.org/10.1158/0008-­5472.CAN-­09-­1681.
7. Bauer N, et al. Haematopoietic stem cell differentiation promotes 26. Henrique D, Schweisguth F. Mechanisms of Notch signal-
the release of prominin-1/CD133-containing membrane vesicles— ing: a simple logic deployed in time and space. Development.
a role of the endocytic–exocytic pathway. EMBO Mol Med. 2019;146(3):172148. https://doi.org/10.1242/dev.172148.
2011;3(7):398–09. https://doi.org/10.1002/emmm.201100147. 27. Hermann PC, et al. Distinct populations of cancer stem cells deter-
8. Boonyaratanakornkit JB, et al. Selection of tumorigenic melanoma mine tumor growth and metastatic activity in human pancreatic can-
cells using ALDH. J Investig Dermatol. 2010;130(12):2799–08. cer. Cell Stem Cell. 2007;1(3):313–23. https://doi.org/10.1016/j.
https://doi.org/10.1038/jid.2010.237. stem.2007.06.002.
71 Cancer Stem Cells and Their Role in Metastasis 755

28. Hirschmann-Jax C, et al. A distinct “side population” of cells 47. Liu S, et al. Hedgehog signaling and Bmi-1 regulate self-renewal
with high drug efflux capacity in human tumor cells. Proc Natl of normal and malignant human mammary stem cells. Cancer
Acad Sci USA. 2004;101(39):14228–33. https://doi.org/10.1073/ Res. 2006;66(12):6063–71. https://doi.org/10.1158/0008-­5472.
pnas.0400067101. CAN-­06-­0054.
29. Hoey T, et al. DLL4 blockade inhibits tumor growth and reduces 48. Logan CY, Nusse R. The Wnt signaling pathway in development
tumor-initiating cell frequency. Cell Stem Cell. 2009;5(2):168–77. and disease. Annu Rev Cell Dev Biol. 2004;20:781–10. https://doi.
https://doi.org/10.1016/j.stem.2009.05.019. org/10.1146/annurev.cellbio.20.010403.113126.
30. Holland JD, et al. Wnt signaling in stem and cancer stem cells. 49. Lou X, et al. SOX2 targets fibronectin 1 to promote cell migration
Curr Opin Cell Biol. 2013;25(2):254–64. https://doi.org/10.1016/j. and invasion in ovarian cancer: new molecular leads for therapeutic
ceb.2013.01.004. intervention. Omics J Integr Biol. 2013;17(10):510–8. https://doi.
31. van den Hoogen C, et al. High aldehyde dehydrogenase activity org/10.1089/omi.2013.0058.
identifies tumor-initiating and metastasis-initiating cells in human 50. Luo Y, et al. ALDH1A isozymes are markers of human mela-
prostate cancer. Cancer Res. 2010;70(12):5163–73. https://doi. noma stem cells and potential therapeutic targets. Stem Cells
org/10.1158/0008-­5472.CAN-­09-­3806. (Dayton, Ohio). 2012;30(10):2100–13. https://doi.org/10.1002/
32. Hooper JE, Scott MP. Communicating with Hedgehogs. Nat stem.1193.
Rev Mol Cell Biol. 2005;6(4):306–17. https://doi.org/10.1038/ 51. Marzesco A-M, et al. Release of extracellular membrane particles
nrm1622. carrying the stem cell marker prominin-1 (CD133) from neu-
33. Huang EH, et al. Aldehyde dehydrogenase 1 is a marker for normal ral progenitors and other epithelial cells. J Cell Sci. 2005;118(Pt
and malignant human colonic stem cells (SC) and tracks SC overpop- 13):2849–58. https://doi.org/10.1242/jcs.02439.
ulation during colon tumorigenesis. Cancer Res. 2009;69(8):3382– 52. Masui S, et al. Pluripotency governed by Sox2 via regulation of
9. https://doi.org/10.1158/0008-­5472.CAN-­08-­4418. Oct3/4 expression in mouse embryonic stem cells. Nat Cell Biol.
34. Hurt EM, et al. CD44 + CD24 − prostate cells are early cancer pro- 2007;9(6):625–35. https://doi.org/10.1038/ncb1589.
genitor/stem cells that provide a model for patients with poor prog- 53. McAuliffe SM, et al. Targeting Notch, a key pathway for ovarian
nosis. Br J Cancer. 2008;98(4):756–65. https://doi.org/10.1038/ cancer stem cells, sensitizes tumors to platinum therapy. Proc Natl
sj.bjc.6604242. Acad Sci USA. 2012;109(43):E2939–48. https://doi.org/10.1073/
35. Ibrahim EE, et al. Embryonic NANOG activity defines colorectal pnas.1206400109.
cancer stem cells and modulates through AP1- and TCF-dependent 54. Merchant AA, Matsui W. Targeting Hedgehog—a cancer stem
mechanisms. Stem Cells (Dayton, Ohio). 2012;30(10):2076–87. cell pathway. Clin Cancer Res. 2010;16(12):3130–40. https://doi.
https://doi.org/10.1002/stem.1182. org/10.1158/1078-­0432.CCR-­09-­2846.
36. Jaggupilli A, Elkord E. Significance of CD44 and CD24 as can- 55. Nagata T, et al. Prognostic significance of NANOG and KLF4 for
cer stem cell markers: an enduring ambiguity. Clin Dev Immunol. breast cancer. Breast Cancer (Tokyo, Japan). 2014;21(1):96–101.
2012;2012:708036. https://doi.org/10.1155/2012/708036. https://doi.org/10.1007/s12282-­012-­0357-­y.
37. Jeter CR, et al. NANOG promotes cancer stem cell characteristics 56. Nakahara F, et al. Hes1 immortalizes committed progenitors and
and prostate cancer resistance to androgen deprivation. Oncogene. plays a role in blast crisis transition in chronic myelogenous leu-
2011;30(36):3833–45. https://doi.org/10.1038/onc.2011.114. kemia. Blood. 2010;115(14):2872–81. https://doi.org/10.1182/
38. Jiang F, et al. Aldehyde dehydrogenase 1 is a tumor stem cell-­ blood-­2009-­05-­222836.
associated marker in lung cancer. Mol Cancer Res. 2009;7(3):330– 57. Nichols J, Evans EP, Smith AG. Establishment of germ-line-­
8. https://doi.org/10.1158/1541-­7786.MCR-­08-­0393. competent embryonic stem (ES) cells using differentiation inhibit-
39. Klein AM, Simons BD. Universal patterns of stem cell fate in ing activity. Development. 1990;110(4):1341–8.
cycling adult tissues. Development. 2011;138(15):3103–11. https:// 58. Niu C-S, et al. Expression of NANOG in human gliomas and
doi.org/10.1242/dev.060103. its relationship with undifferentiated glioma cells. Oncol Rep.
40. Kurrey NK, et al. Snail and slug mediate radioresistance and chemo- 2011;26(3):593–601. https://doi.org/10.3892/or.2011.1308.
resistance by antagonizing p53-mediated apoptosis and acquiring a 59. Olempska M, et al. Detection of tumor stem cell markers in pan-
stem-like phenotype in ovarian cancer cells. Stem Cells (Dayton, creatic carcinoma cell lines. Hepatobiliary Pancreat Dis Int.
Ohio). 2009;27(9):2059–68. https://doi.org/10.1002/stem.154. 2007;6(1):92–7.
41. Laurenti E, Wilson A, Trumpp A. Myc’s other life: stem cells 60. Pattabiraman DR, Weinberg RA. Tackling the cancer stem
and beyond. Curr Opin Cell Biol. 2009;21(6):844–54. https://doi. cells—what challenges do they pose? Nat Rev Drug Discov.
org/10.1016/j.ceb.2009.09.006. 2014;13(7):497–512. https://doi.org/10.1038/nrd4253.
42. Lee SY, et al. Induction of metastasis, cancer stem cell phe- 61. Peacock CD, et al. Hedgehog signaling maintains a tumor
notype, and oncogenic metabolism in cancer cells by ionizing stem cell compartment in multiple myeloma. Proc Natl Acad
radiation. Mol Cancer. 2017;16(1):10. https://doi.org/10.1186/ Sci USA. 2007;104(10):4048–53. https://doi.org/10.1073/
s12943-­016-­0577-­4. pnas.0611682104.
43. Leng Z, et al. Krüppel-like factor 4 acts as an oncogene in colon can- 62. Rahadiani N, et al. Expression of aldehyde dehydrogenase
cer stem cell-enriched spheroid cells. PLoS One. 2013;8(2):e56082. 1 (ALDH1) in endometrioid adenocarcinoma and its clini-
https://doi.org/10.1371/journal.pone.0056082. cal implications. Cancer Sci. 2011;102(4):903–8. https://doi.
44. Li C, et al. Identification of pancreatic cancer stem cells. Cancer org/10.1111/j.1349-­7006.2011.01864.x.
Res. 2007;67(3):1030–7. https://doi.org/10.1158/0008-­5472. 63. Rasheed ZA, et al. Prognostic significance of tumorigenic cells with
CAN-­06-­2030. mesenchymal features in pancreatic adenocarcinoma. J Natl Cancer
45. Liu H, et al. Cancer stem cells from human breast tumors are Inst. 2010;102(5):340–51. https://doi.org/10.1093/jnci/djp535.
involved in spontaneous metastases in orthotopic mouse mod- 64. Rentala S, Mangamoori LN. Oct-4 expression maintained stem
els. Proc Natl Acad Sci. 2010;107(42):18115–20. https://doi. cell properties in prostate cancer-derived CD133+MDR1+ cells.
org/10.1073/pnas.1006732107. Trop J Pharm Res. 2009;8(1):3–9. https://doi.org/10.4314/tjpr.
46. Liu J, et al. Notch signaling in the regulation of stem cell self-­ v8i1.14706.
renewal and differentiation. Curr Top Dev Biol. 2010;92:367–409. 65. Reya T, Clevers H. Wnt signalling in stem cells and cancer. Nature.
https://doi.org/10.1016/S0070-­2153(10)92012-­7. 2005;434(7035):843–50. https://doi.org/10.1038/nature03319.
756 R. Ghadially et al.

66. Ricci-Vitiani L, et al. Identification and expansion of human colon-­ 78. Visvader JE, Lindeman GJ. Cancer stem cells: current status and
cancer-­initiating cells. Nature. 2007;445(7123):111–5. https://doi. evolving complexities. Cell Stem Cell. 2012;10(6):717–28. https://
org/10.1038/nature05384. doi.org/10.1016/j.stem.2012.05.007.
67. Rybak AP, Tang D. SOX2 plays a critical role in EGFR-­ 79. Wang J, et al. c-Myc is required for maintenance of glioma cancer
mediated self-renewal of human prostate cancer stem-like cells. stem cells. PLoS One. 2008;3(11):e3769. https://doi.org/10.1371/
Cell Signal. 2013;25(12):2734–42. https://doi.org/10.1016/j. journal.pone.0003769.
cellsig.2013.08.041. 80. Yeung TM, et al. Cancer stem cells from colorectal cancer-derived
68. Salcido CD, et al. Molecular characterisation of side popula- cell lines. Proc Natl Acad Sci USA. 2010;107(8):3722–7. https://
tion cells with cancer stem cell-like characteristics in small-cell doi.org/10.1073/pnas.0915135107.
lung cancer. Br J Cancer. 2010;102(11):1636–44. https://doi. 81. Yin S, et al. CD133 positive hepatocellular carcinoma cells possess
org/10.1038/sj.bjc.6605668. high capacity for tumorigenicity. Int J Cancer. 2007;120(7):1444–
69. Santini R, et al. SOX2 regulates self-renewal and tumorigenicity of 50. https://doi.org/10.1002/ijc.22476.
human melanoma-initiating cells. Oncogene. 2014;33(38):4697– 82. Yin XY, et al. Inverse regulation of cyclin B1 by c-Myc and p53 and
708. https://doi.org/10.1038/onc.2014.71. induction of tetraploidy by cyclin B1 overexpression. Cancer Res.
70. Sicchieri RD, et al. ABCG2 is a potential marker of tumor-initiating 2001;61(17):6487–93.
cells in breast cancer. Tumour Biol. 2015;36(12):9233–43. https:// 83. Yue L, et al. Targeting ALDH1 to decrease tumorigenic-
doi.org/10.1007/s13277-­015-­3647-­0. ity, growth and metastasis of human melanoma. Melanoma
71. Singh SK, et al. Identification of a cancer stem cell in human brain Res. 2015;25(2):138–48. https://doi.org/10.1097/
tumors. Cancer Res. 2003;63(18):5821–8. CMR.0000000000000144.
72. Singh SK, et al. Identification of human brain tumour initiating 84. Zbinden M, et al. NANOG regulates glioma stem cells and is
cells. Nature. 2004;432(7015):396–401. https://doi.org/10.1038/ essential in vivo acting in a cross-functional network with GLI1
nature03128. and p53. EMBO J. 2010;29(15):2659–74. https://doi.org/10.1038/
73. Suetsugu A, et al. Characterization of CD133+ hepatocellular car- emboj.2010.137.
cinoma cells as cancer stem/progenitor cells. Biochem Biophys 85. Zhang G, et al. Expression of potential cancer stem cell marker
Res Commun. 2006;351(4):820–4. https://doi.org/10.1016/j. ABCG2 is associated with malignant behaviors of hepatocellular
bbrc.2006.10.128. carcinoma. Gastroenterol Res Pract. 2013;2013:782581. https://
74. Tiezzi DG, et al. ABCG2 as a potential cancer stem cell marker in doi.org/10.1155/2013/782581.
breast cancer. J Clin Oncol. 2013;31(15_suppl):e12007. https://doi. 86. Zhang S, et al. Identification and characterization of ovar-
org/10.1200/jco.2013.31.15_suppl.e12007. ian cancer-initiating cells from primary human tumors. Cancer
75. Tomita H, et al. Aldehyde dehydrogenase 1A1 in stem cells and can- Res. 2008;68(11):4311–20. https://doi.org/10.1158/0008-­5472.
cer. Oncotarget. 2016;7(10):11018–32. https://doi.org/10.18632/ CAN-­08-­0364.
oncotarget.6920. 87. Zhao C, et al. Hedgehog signalling is essential for main-
76. Ucar D, et al. Aldehyde dehydrogenase activity as a functional tenance of cancer stem cells in myeloid leukaemia.
marker for lung cancer. Chem Biol Interact. 2009;178(1):48–55. Nature. 2009;458(7239):776–9. https://doi.org/10.1038/
https://doi.org/10.1016/j.cbi.2008.09.029. nature07737.
77. Vermeulen L, et al. Wnt activity defines colon cancer stem 88. Zhou X, et al. Expression of the stem cell marker Nanog in human
cells and is regulated by the microenvironment. Nat Cell Biol. endometrial adenocarcinoma. Int J Gynecol Pathol. 2011;30(3):262–
2010;12(5):468–76. https://doi.org/10.1038/ncb2048. 70. https://doi.org/10.1097/PGP.0b013e3182055a1f.
 HPV-Induced Cancers
72
Brittney L. Dickey, Jennifer M. Binning, and Julie Rathwell

Abstract
Learning Objectives
Highly prevalent human papillomavirus (HPV) can lead • Understand the process by which HPV infects cells
to cancer in the anogenital and oropharyngeal regions. to cause cancer.
Oncogenic, or high-risk, HPV types are attributed to • Describe the epidemiology of HPV-associated
690,000 cancers worldwide. HPV infects basal epithe- cancers.
lial cells and is propagated solely through host cellular • Explore the patterns of metastases in the presence
differentiation. As viral amplification occurs, virions are of an HPV infection.
shed from infected cells and continues the transmission • Discuss how HPV-related cancers can be
process. While most HPV infections go unnoticed and eliminated.
clear within 2 years, persistent infections can lead to
precancerous and cancerous lesions. Malignant transfor-
mation is driven by viral oncoproteins E5, E6, and E7,
and these oncoproteins remain important for tumor
maintenance. Cervical cancer has been shown to metas- 72.1 Introduction
tasize via the lymphovascular system due to close prox-
imity of the lymphatic spaces, though direct local Human papillomaviruses (HPV) are epitheliotropic, double-­
extension is also common. Oropharyngeal cancer (OPC), stranded DNA viruses and are highly prevalent. The majority
though, has more diverse etiology with some cases of HPV infections are benign; however, infection with onco-
attributed to HPV, while others are due to tobacco and genic types that persist can lead to cancer at six anatomic
alcohol use. HPV-associated OPC cases are known to sites: cervix, vagina, vulva, penis, anus, and oropharynx. In
have better survival, and early studies indicate differ- this chapter, we will describe the types of HPV that can
ences in rate or site of metastases by HPV status may cause cancer, how the virus is transmitted, the cellular
exist, though much is still unknown. Compared to other changes caused by an HPV infection that leads to cancer,
oncoviruses, there are proven tools including vaccine metastasis of HPV-related cancers, and the effort to elimi-
and screening that can not only prevent HPV-associated nate these cancers.
cancers but potentially eliminate them.

72.2 HPV Infection

72.2.1 HPV Types

B. L. Dickey (*) · J. M. Binning · J. Rathwell Approximately 200 papillomavirus genotypes have been
Moffitt Cancer Center, Tampa, USA
e-mail: [email protected]; identified and are grouped into five genera: alpha, beta,
[email protected]; [email protected] gamma, mu, and nu [1]. Among these, alpha and beta types

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 757
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_72
758 B. L. Dickey et al.

Table 72.1 HPV types associated with disease virus does not possess its own replication machinery, it relies
Risk HPV type Genus Associated disease on cellular division and stratification of the epithelium to
High-risk 16, 18, 31, 33, Alpha Cervical, anal, vaginal, successfully propagate. This occurs in the basal layers and
HPV or 35, 39, 45, 51, vulvar, penile, and progresses toward the suprabasal layers, to maintain cell pro-
oncogenic 52, 56, 58, 59 oropharyngeal cancers and
associated precursor lesions liferation in highly differentiated suprabasal regions. This
Probable 68 Alpha Cervical cancer vertical maturation of infected keratinocytes completes the
carcinogen viral life cycle and allows viral genome amplification to
Possible 5, 8 Beta Squamous cell carcinoma of occur [4]. Viruses produced through this process are shed
carcinogen the skin in persons with
from infected cells and can begin the transmission process
epidermodysplasia
verruciformis anew (Fig. 72.1).
26, 30, 34, 53, Alpha Uncertain
66, 67, 69, 70,
73, 82, 85, 97 72.2.3 Epidemiology of HPV Infection
Low risk 6, 11 Alpha Genital warts, recurrent
respiratory papillomatosis
Anogenital HPV infection is the most common sexually trans-
HPV types are listed with their associated oncogenic risk, genus type,
mitted infection in the USA, with a prevalence of approxi-
and disease. Oncogenic risk is defined according to the criteria from the
International Agency for Research on Cancer Evaluation [3, 4] mately 70 million cases [7]. Due to limited reporting of genital
infection, HPV incidence rates and prevalence are likely
underestimated, especially globally. However, it is expected
have been the most extensively studied [2]. Alpha-HPVs that nearly all sexually active individuals will acquire HPV at
include the genotypes associated with cancer, while infec- some time [8]. The principal risk factor is sexual behavior,
tions with all other types are asymptomatic. Twelve types particularly early age at sexual exposure and multiple partners
(HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 59) that [7]. Due to such transmission patterns, anogenital HPV infec-
exclusively infect the inner tissue lining are considered high-­ tions are very common. Among young women, prevalence
risk HPVs and group 1 carcinogens by the International peaks at approximately 20–25 years. Variations in peak preva-
Agency for Research on Cancer (IARC), and other HPV lence globally are slight and are mostly attributed to a differ-
types are considered probably oncogenic (i.e., HPV68 for ence in average age at sexual initiation. However, a second
cervical cancer) or uncertain. In contrast, low-risk types peak after menopause is observed in some countries, though
(HPV6 and 11) cause genital warts and are considered to not the reasons for this peak are not well understood [3].
be a cause of cancer [3] (Table 72.1). Interestingly, anogenital HPV infection prevalence among
men does not vary as much by age, with relatively flat age-
specific prevalence curves, and varies by country, specific ana-
72.2.2 Viral Transmission tomical site, and risk group (i.e., high-­risk groups, such as
HIV-infected men, compared to low-risk groups) [9].
HPV infection occurs when virions infect basal epithelial Unlike anogenital infections, oral HPV infections are
cells at cutaneous or mucosal sites via direct access to the rare. In the USA, the prevalence of any oral HPV infection
basement membrane of the epithelium (reviewed in [4]). was estimated to be 11.5% among men and 3.3% among
Mucosal infections are mostly restricted to the anogenital women, whereas the prevalence of genital HPV was 45.2%
and oral cavities as well as laryngeal epithelium, whereas all among men and 39.9% among women [10]. Oral sex may
cutaneous tissues are susceptible to HPV infections. Natural lead to an HPV infection of the oral cavity, with men who
transmission of mucosal infections primarily involves direct perform oral sex on women being more susceptible. The key
genital mucosa contact [5] and is believed to occur following factor for an oral HPV infection is sexual behavior, though it
the mechanical wounding that arises during vaginal and anal appears to be more closely associated with the number of
sex [6]. The unique nature of the tonsillar crypt epithelium, lifetime oral sex partners, with an overall odds of infection of
with a disrupted basal cell layer and basement membrane, 5.7 for six or more lifetime partners [7].
likely allows viral infection of the basal cells without the pre- Most individuals with an HPV infection are unaware of
requisite wounding [6]. Natural transmission of cutaneous the infection and do not incur noticeable symptoms.
infections involves direct contact of an infected keratinocyte, Approximately half of all HPV infections clear within
such as a wart, with uninfected skin cells possibly via micro- 6 months, and more than 90% clear within a few years [3,
abrasions in the skin. 11]. Infections may present as asymptomatic infections,
Once the virus infects those cells most susceptible to small epithelial lesions, large cutaneous lesions, or cancer-
infection, usually keratinocytes, the viral life cycle for HPV ous lesions. The factors that determine whether an HPV
is strictly dependent upon cellular differentiation. As the infection will be cleared are poorly defined, yet the interplay
72 HPV-Induced Cancers 759

Normal cervix Squamous intraepithelial lesion Invasive cancer

Low grade High grade

Cervical intraepithelial neoplasia

Grade 1 Grade 2 Grade 3

Infectious viral particles

Squamous epithelium

Superficial zone

Midzone

Basal layer
Basement membrane
Dermis

Episome

LCR E6
E7
Integration
Nuclei with episomal Overexpression of
viral DNA E 6 and E 7 L1 L2 LCR E7 E2
L1 E1
Nuclei with intergrated Expression of early
viral DNA and late genes
Normal nuclei Host E6 E 1 Host
E2 DNA DNA
L2 E4
E5

Fig. 72.1 HPV viral integration and transformation to cervical cancer. until the virus is shed and continues this replication process in nearby
The viral integration and cellular changes to cancer process are similar cells. Low-grade intraepithelial lesions support productive viral replica-
for all sites of HPV-induced cancers. Through abrasions in the cervical tion and are considered low-risk cervical lesions. Infection with high-­
epithelium, HPV is able to access and infect basal cells. After infection, risk HPV types will progress to high-grade intraepithelial lesions or
expression of HPV genes E1, E2, E4, E5, E6, and E7 allows viral DNA precancer. Untreated lesions will progress to microinvasion and inva-
to replicate from episomal DNA. Replication continues through the sive cancer and is associated with integration of HPV into host chromo-
upper layers of the epithelium as late genes, L1 and L2, are expressed somes (red nuclei), with loss or disruption of E2, and upregulation of
and encapsulate the viral genome in order to reproduce in the nucleus E6 and E7 oncogene expression. Reproduced from Crosbie et al. [17]

between HPV defense mechanisms, tissue tropism, and the To better understand carcinogenesis following HPV
host immune system undoubtedly plays an important role in infection, it is essential to also understand several key HPV
this process. proteins that drive this process. HPV encodes for eight pro-
teins, of which HPVs E5, E6, and E7 are known viral onco-
proteins. E6 and E7 are major oncogenic drivers and together
72.3 HPV and Cancer are sufficient to immortalize primary human keratinocytes
[12, 13]. E6 primarily functions by interfering with cell sur-
72.3.1 Transformation from Infection vival pathways, and E7 primarily functions by promoting
to Cancer cellular proliferation. Arguably the most important cellular
interactions allow E6 and E7 to inhibit two well-established
The process of HPV-related carcinogenesis starts with infec- tumor suppressors, p53 and retinoblastoma protein (pRB),
tion and is followed by persistent infection, development of respectively. Specifically, E6 simultaneously binds to p53
a high-grade precursor or “precancerous” lesion, and inva- and E6-associated protein (E6AP), a ubiquitin ligase. Thus,
sion (Fig. 72.1). The most important viral factor for progres- p53 is targeted for degradation via the ubiquitin-proteasome
sion to precancerous lesion or invasion is the HPV type, pathway, thereby preventing it from inducing apoptosis [14].
specifically high-risk types. In total, this process from HPV Similarly, E7 also targets pRb for degradation via the
infection to cancer can take many years (~ 10–20) years to ubiquitin-­proteasome pathway. This activity of E7 is impor-
develop [11]. tant for driving cell cycle progression in a differentiating
760 B. L. Dickey et al.

epithelium, which, as mentioned above, is essential for HPV- Oropharyngeal cancer (OPC), a subset of head and neck
DNA replication. In addition to the pRb interaction, E7 also squamous cell cancers (HNSCCs), includes cancers arising
binds to the other pocket proteins (i.e., p107 and p130) using at the base of the tongue, tonsils, posterior pharyngeal wall,
its LXCXE motif. Collectively, E7 is able to regulate cellular and soft palate and represents the second most common
proliferation, differentiation, and apoptosis by interacting HPV-induced cancer site. While nearly all cervical cancers
with and deregulating these cellular proteins. are caused by HPV infection, OPC has two distinct etiolo-
In addition to E6 and E7, E5 plays an important role in gies: tobacco and alcohol use or HPV infection. Both may be
early carcinogenesis [15]. E5, a small membrane-bound pro- present, but HPV-induced OPC (HPV-OPC) tends to occur in
tein, itself cannot immortalize human primary cells but can younger males with less tobacco and alcohol use [7].
enhance the immortalization of keratinocytes by E6/E7 [16]. Interestingly, OPC associated with tobacco/alcohol has
E5 promotes tumorigenesis in part by enhancing the expres- decreased over the last decade, while HPV-OPC has
sion of epidermal growth factor receptor-1 (EGFR-1) and increased, particularly among men. Similar to cervical can-
increasing phosphorylation of protein kinase B (AKT). This cer, HPV-OPC is primarily associated with infection from
in turn activates the MEK/ERK1/2 pathway to induce HPV-­ HPV types 16 and 18 (90%), while the other high-risk HPV
associated carcinogenesis [16]. types are associated with the remaining cases of HPV-­
OPC. Unlike cervical cancer, HPV-OPC is increasing pri-
marily in high-income countries, with rates in LMICs
72.3.2 HPV and Associated Cancers remaining low [20]. When compared to HPV-negative OPC,
HPV-OPC is highly responsive to treatment and exhibits
In 2018 approximately 690,000 cancer cases were attributed superior treatment outcomes and survival. Because of this,
to infection with high-risk HPV types (age-standardized inci- the American Joint Committee on Cancer recently revised
dence rate of 8.0 per 100,000) [16]. It is second only to guidance indicating that OPCs should be staged, with differ-
Helicobacter pylori among cancer-causing infections. HPV is ent staging for HPV-positive and HPV-negative OPC [21].
associated most frequently with cervical cancer (570,000 However, treatment for OPC is the same regardless of HPV
attributable cases in 2018), oropharyngeal cancer (OPC) status. Ongoing studies are examining options for de-­
(42,000 attributable cases), and anal squamous cell carcinoma escalation among HPV-OPC with many omitting chemother-
(29,000 cases). Other sites associated with HPV infection apies in favor of RT alone when HPV-positive patients
include the oral cavity, larynx, penis, vagina, and vulva [18]. present with low risk of metastatic disease [7].
Cervical cancer represents 80% of all cases attributed to Immunosuppression is a key factor in the progression
HPV [18]. It is the fourth most common cancer and the from HPV infection to precancer or cancer, and, thus, people
fourth leading cause of cancer death among women world- living with HIV (PLWH) represent an important, high-risk
wide. Due to limited access to screening and treatment of group. Due to similar risk and transmission factors, many
precancerous lesions, cervical cancer disproportionately PLWH are also infected with HPV, and women living with
affects women in low- and middle-income countries HIV (WHIV) are more likely to be diagnosed with high-­
(LMICs), with age-standardized rates as high as 43 per grade precancerous lesions or invasive cervical cancer com-
100,000 in regions of Africa, but nearly 6 per 100,000 in pared to HIV-uninfected women. In fact, cervical cancer is
North America, Australia, and portions of Europe. In high-­ up to five times more likely to develop among WHIV. This is
resource countries, maximum incidence occurs at age 40, due to a multitude of factors including increased risk of
compared to a global average age at diagnosis of 53 years infection with HPV, reduced HPV clearance, simultaneous
[19]. Nearly all cervical cancer is caused by HPV and infection with multiple types of HPV, and compromised
­specifically high-risk HPV types 16 and 18. Together they immune function [7]. Because HIV infection increases the
cause an estimated 72% of all cervical cancer cases, while risk of progression to cancer, cervical cancer is considered
other high-risk types 31, 33, 45, 52, and 58 cause another an acquired immunodeficiency syndrome (AIDS)-defining
17% of cases [18]. Current treatment for cervical cancer condition [7].
depends on the stage at diagnosis. Early-stage treatment usu- The relationship between HIV status and HPV-related
ally involves surgical techniques and may include modified OPC is less clear. HNSCC, regardless of HPV status, is
radical hysterectomy with pelvic lymphadenectomy; how- threefold higher in PLWH in the USA. When stratified by
ever, HPV-­related cervical cancers are radiosensitive, and HPV status, the incidence of HPV-related HNSCCs
radiation therapy may be considered as well. Persistent, increased, while HPV-unrelated HNSCC decreased [22].
recurrent, or metastatic cervical cancer is treated with che- The risk of developing OPC is higher for PLWH, though it is
motherapy and vascular endothelial growth factor (VEGF) not yet understood whether the distribution of HPV types in
inhibitor. Options for immunotherapy, including therapeutic OPCs differs between PLWH and the general population
vaccines, are currently emerging and under investigation [7]. [23]. In the USA, recent estimates show a 1.6-fold increase
72 HPV-Induced Cancers 761

in HPV-related OPCs among PLWH during the period of IB–IIA disease and is an important prognostic factor, reduc-
1996–2012, though HPV-unrelated OPCs exhibited a similar ing 5-year survival rates from 80 to 53% when present.
increase (2.2-fold) [24]. Worth mentioning, though relatively In addition to lymphatic metastases, cervical cancer can
rare in the general population, anal cancer associated with also metastasize via vascular pathways. Several factors facil-
HPV infection is much higher among PLWH and is the most itate lymphovascular metastases from primary cervical can-
common non-AIDS-defining cancer among PLWH. As with cer. Lymphatic vessel density (LVD), for example, is
cervical cancer, immunosuppression facilitates the progres- positively associated with lymph node metastases in cases of
sion of HPV infection to cancer at the anal canal. In the USA cervical cancer. Specifically, increasing peritumoral LVD
anal cancer was 19 times more likely to occur in PLWH has been linked with growth factor expression, lymphovas-
compared to the general population [24]. cular space invasion, LNMs, and overall survival. Vascular
endothelial growth factor (VEGF), a key protein in cervical
cancer metastases, especially through the lymphovascular
72.4 HPV-Related Cancer Metastasis system, plays a role in the blood and lymph vessel formation.
Many studies have shown that VEGF is a promoter of cervi-
HPV-related cancers are particularly responsive to radiother- cal cancer progression through many oncogenic signaling
apy and, thus, show favorable survival outcomes. However, pathways. VEGF ligand C (VEGF-C) secreted by the pri-
as with all cancers, a proportion of HPV-related cancers mary tumor is known to induce expansion of the lymphatic
recur and metastasize, and diagnoses at later stages make this network within the primary sentinel lymph node. Tumors
outcome more likely. As previously mentioned, the HPV E6 expressing VEGF-C, such as cervical tumors, are more likely
and E7 proteins directly contribute to malignant transforma- to metastasize from the tumor-involved lymph node. Two
tion via their ability to impair p53 and pRB. In both oral and oncogenic signaling pathways associated with VEGF-C, for
cervical epithelial cells, E6 and E7 from high-risk HPVs lead example, include cyclooxygenase-2 (COX-2) and semapho-
to upregulation of proteins associated with the epithelial-to-­ rin 4D (Sema4D). VEGF-C is correlated with COX-2 expres-
mesenchymal transition (EMT) which modulates the pro- sion, lymphangiogenesis, and lymph node metastasis, and
gression of associated carcinomas [8]. Additionally, HPV VEGF-C/D is associated with angiogenesis inducing
type may influence the outcome. Data from The Cancer SEMA4D, which together are positively correlated with
Genome Atlas suggest improved survival for HPV-OPC may lymphatic invasion, lymph node metastases, and FIGO stage
be limited to cases caused by HPV16 [25]. Studies of cervi- [27].
cal cancer have found HPV 18 more closely associated with The relative simplicity of HPV compared to other onco-
cervical adenocarcinoma than squamous cell carcinoma, and genic viruses, such as Epstein-Barr virus, and its unique
adenocarcinoma can be more aggressive [4]. This is an area ability to elicit a host immune response limits its direct role
of ongoing study. in cancer progression. However, HPV E5, E6, and E7 pro-
teins are associated with many roles necessary to cancer pro-
gression including regulation of cell cycle, cellular
72.4.1 Cervical Cancer Metastasis senescence, cell motility, and invasion and compromised cel-
lular genomic stability. The cellular changes from these pro-
Nearly all cervical cancer is caused by HPV. In these can- teins during HPV integration are associated with the
cers, spread primarily occurs via direct local extension and malignant behavior of HPV-induced cancers. Metastasis is
lymphatic embolization. Approximately 5% of cervical can- often a late event in HPV-induced cancers that mostly relies
cer cases will have hematogenous spread via veins and on genomic instability and chromosome aberrations inversely
­lymphatic spaces that lie close to the basement membrane associated with the transforming capability of the various
with the most common sites of spread including the lungs, high-risk strains [8].
liver, and bone. However, hematogenous spread usually
occurs only in the advanced stage, poorly differentiated
tumors, and aggressive cell types [26]. 72.4.2 Metastases in OPC
Lymphatic spread is one of the major routes of cervical
cancer metastases due to the close proximity of lymphatic As previously described, patients with HPV-OPC have better
spaces to the basement membrane (1–2 mm) and, as such, is overall and disease-specific survival compared to HPV-­
even more prevalent with large-volume tumors. The most negative OPC [25]. However, HPV-OPC has not been stud-
common site of the first draining lymph node is the external ied as extensively as cervical cancer, and many questions
iliac, followed by the obturator, parametrial, and common remain to be answered, particularly regarding the differences
iliac nodes [27]. Pelvic lymph node metastasis (LNM) occurs and similarities between HPV-driven OPC and OPC caused
in approximately 15–25% of cervical cancer cases with stage by tobacco and alcohol use.
762 B. L. Dickey et al.

One area of study is whether HPV-positive and HPV-­ ligase complex that monoubiquitinates histone H2B at
negative OPCs have different rates and sites of distant metas- lysine 120. Decreased activity of the RNF20/RNF40 com-
tases. Patients with HPV-OPC are more likely to present with plex is associated with increased invasiveness and tumor
an enlarged lymph node as the primary symptom. progression, and expression of L2 upregulates several genes
Approximately half of all HPV-OPC cases have an enlarged in the EMT and tumor invasiveness pathways [34]. Though
lymph node, compared to only 18% of HPV-negative OPC L2 has not been shown to contribute to metastasis in vivo,
cases, and HPV-OPC is known to metastasize to lymph these data support a potential role for this protein in viral-
nodes earlier in the disease stage (with smaller primary mediated EMT that needs to be explored. In addition to viral
tumors) than HPV-negative OPC [7, 28]. proteins, numerous studies have implicated miRNA in all
Regardless of HPV status, studies have found no signifi- stages of cancer progression [35]. HPV oncoproteins can
cant difference in the site of first progression (recurrence), dysregulate the expression of several cellular miRNAs and
whether locoregional or distant, and distant metastases tend in doing so contribute to the development of HPV-positive
to occur within 3 years of the primary OPC. Some retrospec- malignancies.
tive analyses have suggested HPV-OPC is more likely to
have atypical sites for distant metastases, though this may be
due to longer survival among HPV-OPC patients, allowing 72.5  PV-Related Cancers: From Control
H
more time to develop metastases at sites not frequently to Elimination
observed among HPV-negative patients [25]. Common sites
for distant metastases of OPC, regardless of etiology, are the As the understanding of the mechanisms by which HPV
lungs, liver, and bone [29, 30], although a retrospective causes cancer has improved, tools to control and prevent
review of ~ 1500 HNSCC cases from a single institution those cancers have been developed, making it possible to
showed a higher rate of metastases to the lung in HPV-­ eliminate HPV-related cancers.
positive HNSCC (5.5%) compared to HPV-negative HNSCC
(1.8%) [29]. Recent data have also found that 36% of HPV-­
positive tumors metastasized to multiple sites, compared to 72.5.1 Vaccine
only 2% of HPV-negative tumors. In studies investigating
timing and location of distant progression, HPV-OPC had a Primary prevention of HPV-related cancer was made possi-
longer time to distant progression than HPV-negative OPC, ble with the advent of HPV vaccines. Three licensed, pro-
though not statistically different, and cases of HPV-OPC phylactic HPV vaccines have been developed, all providing
were significantly more likely to progress to nonregional protection against HPV types 16 and 18, the types responsi-
lymph nodes than HPV-negative OPC. Overall, the end-­ ble for ~ 70% of cervical cancers and ~ 90% of oropharyn-
organs of metastases were significantly different between geal cancers. The bivalent vaccine only provides protection
HPV-positive and HPV-negative OPCs, but when compared against HPV types 16 and 18, while the quadrivalent vaccine
by specific organs, only nonregional lymph nodes were sig- provides additional protection against HPV6 and 11 which
nificantly different. This suggests that HPV-OPCs may cause genital warts. Finally, the 9-valent vaccine includes
metastasize through the lymphatic system, whereas HPV-­ protection against five additional high-risk HPV types: 31,
negative OPCs may not to the same degree [30]. 33, 45, 52, and 58. Long-term follow-up has found that all
The molecular events governing this process are not three vaccines provide sustained immunogenicity and effec-
completely understood, but growing evidence supports the tiveness against HPV infection [36–38].
role of HPV viral proteins in promoting cellular invasion The HPV vaccines have also been shown to protect
and migration potential. For instance, HPV E6 and E7 regu- against other HPV-related diseases, including vulvar, vagi-
late EMT through direct (i.e., downregulating E-cadherin) nal, and anal cancers and genital warts [39]. Based on evi-
and indirect mechanisms (through Jagged1 and the PI3-K dence from ancillary studies to vaccine trials, in 2020 the US
pathway). Furthermore, HPV E7 leads to the upregulation Food and Drug Administration granted provisional licensure
of p16 via the loss of pRB with reduced negative feedback of the 9-valent vaccine against oropharyngeal and other head
[31, 32]. Overexpression of p16 has been shown to regulate and neck cancers caused by the HPV types targeted by the
vascular invasiveness and stimulate nodal spread. This vaccine [40–42]. At the time of this writing, a definitive,
occurs by promoting growth of tumor cells via inhibition of phase III international, multi-center randomized, double-­
angiogenesis and enhancement of lymphangiogenesis [33]. blind, placebo-controlled clinical trial to determine vaccine
In addition to E6 and E7, a recent HPV interactome study effectiveness against persistent oral HPV infections
discovered that HPV late capsid protein (L2) interacts with (NCT04199689) is underway and will be completed
RNF20 and RNF40, which form a heterodimeric ubiquitin December 31, 2025.
72 HPV-Induced Cancers 763

72.5.2 Screening and Treatment Control is the reduction of disease incidence, prevalence,
for Preneoplastic Lesions morbidity, and mortality to acceptable levels. In high-­
resource countries such as the USA, Australia, and the UK,
Currently, cervical cancer is the only HPV-related cancer for cervical cancer screening programs have been established
which there exist evidence-based screening methods. for decades, and cervical cancer is considered controlled.
Cervical intraepithelial neoplasia (CIN) is a precancerous The critical tools for eliminating HPV-related cancers are
lesion to cervical cancer that may be detected via screening vaccine, screening, and treatment for precancers and cancers.
and treated to prevent progression to cancer. Several meth- The vaccine will reduce incidence and transmission of HPV
ods exist to detect CIN. Cytology-based screening (Pap test- infection, which will in turn reduce HPV-related diseases.
ing) was implemented in high-income countries as early as Mathematical modeling has shown that we can eliminate cer-
the 1950s and has resulted in dramatic declines in cervical vical cancer in some countries by 2030 with other countries to
cancer incidence due to its ability to detect morphologic follow; however, the fastest declines in cervical cancer inci-
changes in the cervical epithelium and early HPV infections. dence will only be seen if there is a rapid scale-up of cervical
There is now a shift from cytology-based screening to testing cancer screening. HPV 16, 18, 6, and 11 can be eliminated
for high-risk HPV as it is more effective, particularly in pop- with gender-neutral vaccination of ≥ 80% by the quadrivalent
ulations with high rates of HPV vaccine uptake [7, 43]. or 9-valent vaccine, which would prevent the related diseases
The use of high-risk HPV testing will also benefit women that affect many men and women worldwide [46, 49, 50].
living in low- and middle-income countries (LMICs), where a
lack of national screening programs has contributed to high rates
of cervical cancer and associated morbidity and mortality. In 72.6 Conclusion
low-resource settings, the American Society of Clinical Oncology
(ASCO) recommends two-lifetime screenings with HPV testing Cervical and oropharyngeal cancer represent a unique circum-
[44]. Researchers are also studying the efficacy of HPV testing stance in cancer prevention and care. Diagnoses of either often
from self-collected swabs for cervical cancer screening. This accompanies severe mental and physical challenges, with late-
method could not only improve screening among women in low- stage diagnoses comes the potential loss of fertility (cervical
resource countries, it can help reduce the screening disparities cancer) and extreme physical transformations of the head and
that have persisted in high-resource countries. neck from treatment (OPC). In spite of these challenges, most
Screening is only effective for secondary prevention if cases of cervical cancer, half of OPC, and many other ano-
women receive treatment for precancerous and cancerous genital cancers can be prevented with the HPV vaccine, poten-
lesions. Effective ablative treatments exist, including meth- tially preventing the devastating consequences of these
ods that are relatively affordable in LMICS, and can extir- cancers. However, we have not yet reached universal vaccina-
pate and therefore cure 75–85% of high-grade cervical tion against HPV, and while treatable precancerous lesions
lesions before ever progressing to cancer [45]. However, dis- exist for cervical and anal cancer, many at-risk persons still do
parities in treatment persist, and efforts are needed to ensure not participate in screening. Further, there are no FDA-
all women with cervical lesions are connected to care. approved screening options for OPC or cancers at other ano-
genital sites. While we continue to make advancements in
HPV-related cancers, there is still much to be understood
72.5.3 Elimination of HPV-Related Cancers about the associated progression and management of disease.

As described above, metastatic and recurrent cancers associ-

ated with HPV result in significant morbidity and mortality
worldwide. Without further intervention, modelers predict Open Questions
44.4 million cases of cervical cancer will be diagnosed glob- • How do we improve vaccination and screening
ally between 2020 and 2069 [46]. In May 2018, the World rates for cervical cancer?
Health Organization (WHO), led by Director-General Tedros • Can we develop a screening test for OPC to improve
Adhanom Ghebreyesus, issued a Call to Action to eliminate early detection of disease?
cervical cancer with adoption by the member states in August • Determine differences in metastases of OPC based
2020 [47]. The WHO strategy for eliminating cervical cancer on HPV status?
consists of 90% of girls vaccinated with the HPV vaccine by • What molecular differences drive changes in metas-
the age of 15; 70% of women screened using a high-­ tasis for HPV-positive cancers?
performance test by the age of 35 and again by the age of 45; • Will de-escalation reduce undesirable physical
and 90% of women identified with cervical disease treated. challenges associated with OPC treatment while
Elimination of a disease is defined as a reduction to zero, maintaining treatment efficacy?
or near zero, incidence in a defined geographic area [48].
764 B. L. Dickey et al.

References American Joint Committee on cancer eighth edition cancer staging

manual. CA Cancer J Clin. 2017;67(2):122–37.
22. Beachler DC, Abraham AG, Silverberg MJ, Jing Y, Fakhry C, Gill
1. International Human Papillomavirus (HPV) Reference Center. MJ, et al. Incidence and risk factors of HPV-related and HPV-­
2018. https://www.hpvcenter.se/. Accessed 10 Dec 2020. unrelated head and neck squamous cell carcinoma in HIV-infected
2. Egawa N, Doorbar J. The low-risk papillomaviruses. Virus Res. individuals. Oral Oncol. 2014;50(12):1169–76.
2017;231:119–27. 23. Wang CC, Palefsky JM. Human papillomavirus-related oro-
3. de Sanjosé S, Brotons M, Pavón MA. The natural history of human pharyngeal cancer in the HIV-infected population. Oral Dis.
papillomavirus infection. Best Pract Res Clin Obstet Gynaecol. 2016;22:98–106.
2018;47:2–13. 24. Hernández-Ramírez RU, Shiels MS, Dubrow R, Engels
4. Doorbar J, Quint W, Banks L, Bravo IG, Stoler M, Broker TR, et al. EA. Cancer risk in HIV-infected people in the USA from 1996
The biology and life-cycle of human papillomaviruses. Vaccine. to 2012: a population-based, registry-linkage study. Lancet HIV.
2012;30:F55–70. 2017;4(11):e495–504.
5. Bosch FX, Castellsagué X, Muñoz N, De Sanjosé S, Ghaffari AM, 25. Faraji F, Eisele DW, Fakhry C. Emerging insights into recurrent and
González LC, et al. Male sexual behavior and human papilloma- metastatic human papillomavirus-related oropharyngeal squamous
virus DNA: key risk factors for cervical cancer in Spain. J Natl cell carcinoma. Laryngosc Investig Otolaryngol. 2017;2:10–8.
Cancer Inst. 1996;88(15):1060–7. 26. Gallup D. The spread and staging of cervical cancer. London: The
6. Howard JD, Chung CH. Biology of human papillomavirus-related Global Library of Women’s Medicine’s; 2009.
oropharyngeal cancer. Semin Radiat Oncol. 2012;22:187–93. 27. Ahmad A. Introduction to cancer metastasis. Cambridge: Academic
7. Berman TA, Schiller JT. Human papillomavirus in cervical can- Press; 2016. p. 1–403.
cer and oropharyngeal cancer: one cause, two diseases. Cancer. 28. Paver EC, Currie AM, Gupta R, Dahlstrom JE. Human papilloma
2017;123:2219–29. virus related squamous cell carcinomas of the head and neck:
8. Müller-Coan BG, Caetano BFR, Pagano JS, de Elgui OD. Cancer diagnosis, clinical implications and detection of HPV. Pathology.
progression goes viral: the role of oncoviruses in aggressiveness of 2020;52:179–91.
malignancies. Trends Cancer. 2018;4:485–98. 29. Sacks R, Law JY, Zhu H, Beg MS, Gerber DE, Sumer BD, et al.
9. Smith JS, Gilbert PA, Melendy A, Rana RK, Pimenta JM. Age-­ Unique patterns of distant metastases in HPV-positive head and
specific prevalence of human papillomavirus infection in males: a neck cancer. Oncology. 2020;98(3):179–85.
global review. J Adolesc Health. 2011;48:540–52. 30. Tiedemann D, Jakobsen KK, von Buchwald C, Grønhøj
10. Mcquillan G, Kruszon-Moran D, Markowitz LE, Unger ER, Paulose- C. Systematic review on location and timing of distant progres-
Ram R. United States, 2011–2014 National Health and Nutrition sion in human papillomavirus-positive and human papillomavirus-­
Examination Survey (NHANES) 2011. https://www.cdc.gov/nchs/ negative oropharyngeal squamous cell carcinomas. Head Neck.
data/databriefs/db280_table.pdf#1. Accessed 2020 Nov 24. 2019;41:793–8.
11. Schiffman M, Castle PE, Jeronimo J, Rodriguez AC, Wacholder 31. Lawrence MS, Sougnez C, Lichtenstein L, Cibulskis K, Lander
S. Human papillomavirus and cervical cancer. Lancet. E, Gabriel SB, et al. Comprehensive genomic characteriza-
2007;370:890–907. tion of head and neck squamous cell carcinomas. Nature.
12. Münger K, Phelps WC, Bubb V, Howley PM, Schlegel R. The E6 2015;517(7536):576–82.
and E7 genes of the human papillomavirus type 16 together are 32. Chung CH, Gillison ML. Human papillomavirus in head and neck
necessary and sufficient for transformation of primary human kera- cancer: its role in pathogenesis and clinical implications. Clin
tinocytes. J Virol. 1989;63(10):4417–21. Cancer Res. 2009;15:6758–62.
13. Hawley-Nelson P, Vousden KH, Hubbert NL, Lowy DR, Schiller 33. Dok R, Glorieux M, Holacka K, Bamps M, Nuyts S. Dual role for
JT. HPV16 E6 and E7 proteins cooperate to immortalize human p16 in the metastasis process of HPV positive head and neck can-
foreskin keratinocytes. EMBO J. 1989;8(12):3905–10. cers. Mol Cancer. 2017;16:113.
14. Pal A, Kundu R. Human papillomavirus E6 and E7: the cervi- 34. Eckhardt M, Zhang W, Gross AM, Von Dollen J, Johnson JR,
cal cancer hallmarks and targets for therapy. Front Microbiol. Franks-Skiba KE, et al. Multiple routes to oncogenesis are pro-
2020;10:3116. moted by the human papillomavirus-host protein network. Cancer
15. Estêvão D, Costa NR, da Gil Costa RM, Medeiros R. Hallmarks Discov. 2018;8(11):1474–89.
of HPV carcinogenesis: the role of E6, E7 and E5 oncoproteins 35. Santos JMO, da Silva SP, Costa NR, da Gil Costa RM, Medeiros
in cellular malignancy. Biochim Biophys Acta Gene Regul Mech. R. The role of microRNAs in the metastatic process of high-risk
2019;1862:153–62. HPV-induced cancers. Cancers. 2018;10:493.
16. Zhang L, Wu J, Ling MT, Zhao L, Zhao KN. The role of the PI3K/ 36. Olsson SE, Restrepo JA, Reina JC, Pitisuttithum P, Ulied A,
Akt/mTOR signalling pathway in human cancers induced by infec- Varman M, et al. Long-term immunogenicity, effectiveness, and
tion with human papillomaviruses. Mol Cancer. 2015;14:87. safety of nine-valent human papillomavirus vaccine in girls and
17. Crosbie EJ, Einstein MH, Franceschi S, Kitchener HC. Human pap- boys 9 to 15 years of age: Interim analysis after 8 years of follow-
illomavirus and cervical cancer. Lancet. 2013;382:889–99. ­up. Papillomavirus Res. 2020;10:100203.
18. de Martel C, Georges D, Bray F, Ferlay J, Clifford GM. Global 37. Kjaer SK, Nygård M, Dillner J, Marshall JB, Radley D, Li M, et al.
burden of cancer attributable to infections in 2018: a worldwide A 12-year follow-up on the long-term effectiveness of the quadri-
incidence analysis. Lancet Glob Health. 2020;8(2):e180–90. valent human papillomavirus vaccine in 4 Nordic countries. Clin
19. Arbyn M, Weiderpass E, Bruni L, de Sanjosé S, Saraiya M, Infect Dis. 2018;66(3):339–45.
Ferlay J, et al. Estimates of incidence and mortality of cervi- 38. Schwarz TF, Galaj A, Spaczynski M, Wysocki J, Kaufmann
cal cancer in 2018: a worldwide analysis. Lancet Glob Health. AM, Poncelet S, et al. Ten-year immune persistence and
2020;8(2):e191–203. safety of the HPV-16/18 AS04-adjuvanted vaccine in
20. Gillison ML, Chaturvedi AK, Anderson WF, Fakhry females vaccinated at 15–55 years of age. Cancer Med.
C. Epidemiology of human papillomavirus-positive head and neck 2017;6(11):2723–31.
squamous cell carcinoma. J Clin Oncol. 2015;33:3235–42. 39. Luostarinen T, Apter D, Dillner J, Eriksson T, Harjula K, Natunen
21. Lydiatt WM, Patel SG, O’Sullivan B, Brandwein MS, Ridge JA, K, et al. Vaccination protects against invasive HPV-associated can-
Migliacci JC, et al. Head and neck cancers-major changes in the cers. Int J Cancer. 2018;142:2186–7.
72 HPV-Induced Cancers 765

40. Lehtinen M, Apter D, Eriksson T, Harjula K, Hokkanen M, 46. Simms KT, Steinberg J, Caruana M, Smith MA, Lew JB,
Lehtinen T, et al. Effectiveness of the AS04-adjuvanted HPV-­ Soerjomataram I, et al. Impact of scaled up human pap-
16/18 vaccine in reducing oropharyngeal HPV infections in young illomavirus vaccination and cervical screening and the
females—Results from a community-randomized trial. Int J Cancer. potential for global elimination of cervical cancer in 181
2020;147(1):170–4. countries, 2020–99: a modelling study. Lancet Oncol.
41. Mehanna H, Bryant TS, Babrah J, Louie K, Bryant JL, Spruce 2019;20(3):394–407.
RJ, et al. Human papillomavirus (HPV) vaccine effectiveness and 47. World Health Organization. 73rd World Health Assembly Decisions.
potential herd immunity for reducing oncogenic oropharyngeal Newsroom. 2020. https://www.who.int/news-­room/feature-­stories/
HPV-16 prevalence in the United Kingdom: a cross-sectional study. detail/73rd-­world-­health-­assembly-­decisions. Accessed 23 Nov
Clin Infect Dis. 2019;69(8):1296–302. 2020.
42. Chaturvedi AK, Graubard BI, Broutian T, Pickard RKL, Tong ZY, 48. Dowdle WR. The principles of disease elimination and eradication.
Xiao W, et al. Effect of prophylactic human papillomavirus (HPV) Bull World Health Organ. 1998;76(Suppl. 2):22–5.
vaccination on oral HPV infections among young adults in the 49. Brisson M, Bénard É, Drolet M, Bogaards JA, Baussano
United States. J Clin Oncol. 2018;36(3):262–7. I, Vänskä S, et al. Population-level impact, herd immu-
43. Wentzensen N, Arbyn M. HPV-based cervical cancer screening-­ nity, and elimination after human papillomavirus vaccina-
facts, fiction, and misperceptions. Prev Med (Baltim). 2017;98:33–5. tion: a systematic review and meta-analysis of predictions
44. Ogilvie G, Nakisige C, Huh WK, Mehrotra R, Franco EL, from transmission-dynamic models. Lancet Public Health.
Jeronimo J. Optimizing secondary prevention of cervical can- 2016;1(1):e8–17.
cer: recent advances and future challenges. Int J Gynecol Obstet. 50. Hall MT, Simms KT, B L, Smith MA, Brotherton JM, Saville
2017;138:15–9. M, et al. The projected timeframe until cervical cancer elimi-
45. Castle PE, Murokora D, Perez C, Alvarez M, Quek SC, Campbell nation in Australia: a modelling study. Lancet Public Health.
C. Treatment of cervical intraepithelial lesions. Int J Gynecol 2019;4(1):e19–27.
Obstet. 2017;138:20–5.
 Microbiome and Cancer Metastasis
73
Michael G. White, Jennifer A. Wargo,
and Jennifer L. McQuade

Abstract 73.1 Introduction

The commensal microbiome is a unique set of bacteria,
viruses, and fungi that exist on and in a person’s body. As The commensal microbiome is composed of communities of
these organisms are increasingly understood as a neces- bacterial and viral flora constituting a complex ecosystem
sary and symbiotic part of human health, they have also that exists in various configurations at various sites through-
been associated with the development and progression of out the body. It is unique to each individual, varies over time,
a variety of cancers. Additionally, these organisms have and responds to physiologic stress as well as changes in a
been noted to aid malignancies in their ability to evade person’s environment. This system is multifaceted, particu-
host immune surveillance functions as well as inactivate larly when considering its interactions with the host immune
cytotoxic chemotherapies. As the role of the tumor micro- system in maintaining and modulating homeostasis [1, 2]. As
environment and immune system is increasingly recog- this symbiotic relationship between the gut microbiome and
nized as vital to the treatment of malignant disease, the host immune systems continues to be defined, the effect of
gut, tumoral, and circulating microbiome must be both the gut and tumor microbiome on cancer and cancer
included in this equation as well. therapies can be categorized as follows: (1) the direct effect
of the gut microbiome on cancer initiation and progression,
(2) the effect of the tumor microbiome on the progression of
disease, and (3) the effect of both the gut and tumor microbi-
Learning Objectives ome on the response to treatment (Fig. 73.1).
• Understand the role of the gut, circulating, and
tumoral microbiome in the development of malig-
nancy, response to treatment, and evasion of 73.1.1 The Microbiome and Carcinogenesis
therapeutics.
• Describe the role of preclinical models of the gut The gut microbiome is a community of trillions of bacteria
microbiome in understanding links between correl- with increasing complexity as one travels distally through
ative associations and causative effects. the gastrointestinal tract. These organisms play an essential
• Articulate future directions of study including fecal role in digestion, through fermentation, of various substrates
microbial transplant, consortia microbiome, and and in enzymatic metabolism. These processes not only aid
metabolomics. in nutrient absorption but also produce various metabolites
known to have local and systemic effects [3]. These metabo-
lites have been shown to act as systemic mediators of overall
health and may also have direct local effects on cancer initia-
M. G. White · J. A. Wargo tion and progression and immune modulation. Similarly, gut
Department of Surgical Oncology, The University of Texas MD bacteria have been associated with local inflammation and
Anderson Cancer Center, Houston, TX, USA
have been directly linked to carcinogenesis through a variety
e-mail: [email protected]; [email protected]
of mechanisms.
J. L. McQuade (*)
Currently, at least 20% of cancers are linked to microbes,
Department of Melanoma Medical Oncology, The University of
Texas MD Anderson Cancer Center, Houston, TX, USA and many more associations are being discovered given new
e-mail: [email protected] tools for analysis [4]. One of the most notable associations to

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 767
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_73
768 M. G. White et al.

Fig. 73.1 Mechanisms through which commensal microbial bacteria lead to carcinogenesis and metastasis. Created by BioRender.com

date includes the implication of Helicobacter pylori in gas- ancestry, it conferred substantial risk for those patients with
tric cancer carcinogenesis—for which Barry Marshall and Amerindian ancestry.
Robin Warren won the Nobel Prize in Medicine and Along with the germline risk, H. pylori can directly
Physiology in 2005. As a result of the dysregulated interface induce carcinogenesis through production of the CagA onco-
between an overactive host immune system, triggered by a gene. CagA induces cell proliferation through activation of
pathogenic bacterium, this example illustrates how host PI3K-AKT, MEK-ERK, and β-catenin-WNT pathways lead-
microbial interactions can contribute to malignancy. ing to proliferation and potential malignant degeneration
Initial correlative work studying the effect alterations in [7–9]. Similarly, CagA interferes with p53 and RUNX3,
the gut microbiome have on initiation and/or progression of decreasing apoptosis and further increasing the risk of malig-
malignancies demonstrated associations between a prepon- nancy [10, 11]. Alterations of the host immune system
derance or paucity of various bacterial taxa and malignancy. through induction of increased local IL8 by CagA have also
Ongoing work is elucidating causative associations, although been suggested as a factor leading to the development of gas-
these associations are largely preliminary and true causality tric cancer in these patients [12]. Finally, on a somatic muta-
in the clinical setting remains elusive. Causality is perhaps tional level, colonization by H. pylori has been shown to
best demonstrated in the association between H. pylori and induce genetic instability in cell line, murine, and clinical
the development of gastric cancer. This interaction involves specimens [5, 13]. The multifactorial nature of this onco-
the virulence factors of specific strains of H. pylori, germline genic process is indicative of the complex interplay between
risk, and environmental factors including diet and the micro- host and commensal flora. In the instance of H. pylori, carci-
biome inducing a local inflammatory response and carcino- nogenesis involves local immune modulation, increased pro-
genesis in a subset of individuals with H. pylori colonization liferation, and induction of somatic mutations leading to the
[5]. The role of germline risk has been well demonstrated by de-differentiation of terminally differentiated gastric cells,
Kodaman and colleagues. Their work demonstrated that subsequent hyperplasia, and carcinogenesis.
ancestral background correlated with gastric cancer risk. Beyond bacterial associations, the relationship between
Moreover, this risk is predicated on the strain of H. pylori pathogenic viral colonization and carcinogenesis has been
with which a patient is colonized [6]. While African H. pylori clearly defined in a number of malignancies. Infection with
ancestry was relatively benign in those patients with African hepatitis B or C viruses has been shown to induce a cascade
73 Microbiome and Cancer Metastasis 769

of changes resulting in alteration of DNA methylation, Fusobacterium within colorectal cancers in observational
miRNA expression, and activation of a number of signal and mechanistic studies [25]. Higher levels of Fusobacterium
transduction pathways and can lead to hepatocellular carci- have been found in colorectal tumors as compared to nor-
noma (HCC). Additionally, immune-mediated changes and mally matched colonic mucosa. Interestingly, intra-tumoral
effects of oxidation on the tumor immune microenvironment Fusobacterium is more commonly seen in right-sided can-
induced by hepatitis B and C viruses allow these tumors to cers and in BRAF-mutated tumors. Additionally, taxonomic
evade the host immune system [14]. associations were noted by canonical molecular subtype and
Anal squamous cell carcinoma, a variety of genitourinary micro-satellite instability [26]. Suggested mechanisms of
carcinomas, and head and neck squamous cell carcinoma carcinogenesis are multifactorial. Fusobacterium has been
have been linked to human papillomavirus (HPV). suggested as a chemoattractant for myeloid cells, has been
Mechanistically, this has been shown to be largely due to the shown to increase β-catenin signaling, inactivates NK cells,
over-expression of oncogenes E6 and E7 leading to the alter- acts as an immunomodulator, and alters the production of
ation of various signaling pathways including p53, PI3K, miRNAs [27]. The causality of these effects is emphasized
Akt, and Wnt/Notch. Meanwhile oncogene E5 has been by the ability to negate these risks in Fusobacterium-derived
implicated in evasion of the host immune system [15]. Unlike colorectal cancer murine model with treatment with antibiot-
their bacterial counterparts, these pathogenic viruses can be ics [28].
prevented by vaccination leading to a decreased disease inci- Pancreatic cancer also demonstrates a unique tumoral
dence, as has been demonstrated with HPV-related cervical microbial signature that effects the surrounding tumor micro-
cancer [16, 17]. Unfortunately, eradication of hepatitis C environment. As with the previously described malignancies,
virus with antivirals does not appear to significantly alter the these tumoral bacteria play a number of roles. This includes
risk of development of hepatocellular carcinoma after an ini- potentially beneficial alterations that assist in altering both
tial infection [18]. the innate and adaptive immune responses by reducing
In delineating mechanisms of carcinogenesis, the direct myeloid-derived suppressor cells, increasing M1 macro-
effect of the gut microbiome in the development of mucosal phage differentiation, and enhancing differentiation of CD4
tumors, including colorectal adenomas and colorectal can- and CD8 T cells [29]. Detrimental effects of tumoral pancre-
cers, has been of significant interest given their inherent atic bacteria and their inactivation of cytotoxic therapies are
proximity. While initial studies failed to demonstrate a dif- discussed later in the chapter. While associations with
ference in commensal flora between cases and controls [19] tumoral bacteria are numerous and exist across histologies,
gut dysbiosis as measured by either a preponderance of spe- the question of correlation versus causation persists and
cific taxa or lack of diversity has since been noted to be asso- requires systematic investigation [30, 31].
ciated with the generation of adenomas and cancers in further
studies [20–23].
Although the vast majority of the microbiome exists 73.1.2 T
 he Role of the Microbiome
within the gastrointestinal tract, mucosal surfaces through- in Metastatic Disease
out the body harbor bacterial and viral flora. In the case of
lung cancer, the diversity and specific taxa within lower The tumor microbiome may also play a role in metastatic
bronchial commensal flora have been associated with carci- disease. Interestingly, resected liver metastases from patients
nogenesis and higher-stage disease. Moreover, these findings with Fusobacterium-positive primary CRCs have been noted
were able to be recapitulated within a murine model and to have Fusobacterium within the distant metastases as well
demonstrated to be mediated by IL17’s effects on the tumor’s as their primary tumors [28]. These Fusobacterium-­
immune infiltrate [24]. associated colorectal cancers were found to behave more
The commensal microbiome has been shown to play a aggressively with worse overall survival.
role in the development of malignancies, once developed Mechanistically there are variable ways through which
tumors themselves host an array of viruses and bacteria asso- Fusobacterium can aid or induce metastates [26, 28, 32].
ciated with and living within tumors themselves. These While progression from a localized to metastatic lesion is
organisms are unique in their ability to thrive in an often-­ complex and multifactorial, requiring detachment from a pri-
hostile and hypoxic environment. Their role in the develop- mary lesion, proliferation, and entry into a metastatic site, a
ment of malignancy is an area of active study. Despite the number of bacteria have recently been demonstrated to
knowledge that bacteria exist within tumors, little work was potentially aid in this process [33, 34]. Both Fusobacterium
done in studying the potential role of these bacteria until the and Bacteroides have been shown to bind E-cadherin and
recent surge of interest in microbiome research. disrupt epithelial integrity leading to an increased ability of a
The relationship between tumors and their microbiome tumor to detach from a primary malignancy [35, 36]. In the
has been most robustly delineated in the case of case of Fusobacterium, this has been shown to be mediated
770 M. G. White et al.

by a virulence factor, FadA, and disruption of the Wnt path- more complex in a clinical setting, the ability of a single
way as well as upregulating CARD3 and promoting metasta- metabolite to alter p53 in this reductive model is
sis [34, 37]. While Bacteroides works via toxin secretion that impressive.
has been shown to cleave E-cadherin [38]. Once liberated Direct effects have also been noted in the case of the
from a primary tumor, E. coli and E. faecium have been metabolization of bile acids by the gut microbiome increas-
shown to promote organization of metastatic cells through ing the likelihood of hepatocellular carcinoma development.
cross talk of quorum-sensing peptides that promote invasion These bile acids have been shown to affect the recruitment of
and angiogenesis [39]. Finally, the gut microbiome has been natural killer T cells to cancerous liver cells [40]. Alterations
tied to immune evasion through the manipulation of the in the gut microbiome and subsequent alterations in the
hepatic immune microenvironment modulated by NK T cells immune cells within the liver have the potential to allow for
[40]. While certainly not the sole driver of metastases, this the development of primary liver tumors as well as metasta-
evidence increasingly suggests these microbes are much ses through impaired immunologic anticancer surveillance.
more than bystanders. Recently, distinct microbial signatures by cancer histol-
Microbial signatures for cancer histologies beyond GI ogy have also been noted in patients’ circulating plasma
malignancies exist as well. In the case of breast cancer, dis- [48]. While circulating bacterial DNA is an intriguing find-
tinct tumor microbial signatures exist based on the subtype ing on its own, it was also shown that these signals were
of breast cancer [41]. Moreover, although not in physical prognostic of cancer-specific outcomes across multiple his-
contact, the gut microbiome has a significant effect on the tologies [48]. The effects these bacteria have on local and
progression of murine-modeled breast cancer that exempli- distant tissues continue to be an area of study as they cer-
fies the global effect the gut microbiome has on not only car- tainly play critical roles in the tumoral immune microenvi-
cinogenesis but also disease progression. In a murine model ronment and in the evasion of therapies.
of hormone-responsive breast cancer, gut dysbiosis induced
by antibiotic administration was shown to increase peritu-
moral inflammation as well as metastases to both lymph 73.1.3 G
 ut Microbiome: Response
nodes and the lung [42]. Findings such as these call into to and Evasion of Treatments
question the potential sinister effects antibiotic administra-
tion during cancer treatment may have. An additional layer of complexity is added to the host-­
Beyond individual bacterial taxa, the gut microbiome has microbiome interaction with the addition of treatment,
long had a clearly defined role in mediating the levels of cir- whether immunotherapy or cytotoxic chemotherapy. To date,
culating metabolites, with variability induced by a patient’s significant work has begun to delineate these multifaceted
diet or antibiotics [43–45]. As part of its symbiotic relation- interfaces, leading to the initiation of a number of clinical
ship with a host, the gut microbiome breaks down gut trials studying the effect of therapeutic alterations to the
­contents to aid in nutrient bioavailability. By-products of microbiome in order to enhance existing therapies [49]. At
these reactions have local effects and are also released into the level of direct effect, tumoral microbial flora has been
the portal and then systemic circulation where they can have shown to induce resistance to cytotoxic chemotherapy. In the
distant effects. These metabolites have been shown to affect case of pancreatic adenocarcinoma, gemcitabine levels were
a host subject’s metabolism, body weight, and a variety of found to be lower in the tumor than the surrounding tissues.
health factors [45]. In the case of oncologic health, changes This decrease was then shown to be mediated by tumoral
in the gut microbiome can have global effects such as alter- Gammaproteobacteria that was able to enzymatically
ing the metabolization of estrogens, thereby increasing degrade gemcitabine, thereby decreasing its efficacy [50].
breast cancer risk [46]. Similarly, it has been observed that irinotecan dose variabil-
Recent work has suggested metabolites may also play a ity correlates with the variability of the gut microbiome
crucial role in the determination of the function of molecules between patients. Investigators have therefore been able to
such as p53 as alternatively a tumor suppressor or oncogene. define a “metabolizer” gut microbe phenotype [51]. While in
This has significant implications in the development of the case of Fusobacterium-associated colorectal cancer,
colorectal adenomas and adenocarcinoma as induced dysbi- miRNA-activating autophagy can lead to resistance to che-
osis in murine and organoid models demonstrated a switch motherapy [52]. These early clinical observations are
of p53 from a tumor suppressor to an oncogene. This mod- strengthened by associations in murine models demonstrat-
eled effect was shown to be driven by a single metabolite, ing an association with the gut microbiome having an impact
gallic acid, recapitulating the effect of a complex gut micro- on the efficacy of cyclophosphamide, a chemotherapeutic
biome switching the effect of p53 from a tumor suppressor to that also is an immunomodulator. Further complicating this
an oncogene [47]. Although these effects are likely much association is the fact that cyclophosphamide also effects the
73 Microbiome and Cancer Metastasis 771

gut microbiome itself [53]. These associations are demon- ever, makes its study difficult. Nevertheless, the functional
strative of the many degrees of freedom that must be basis of these interactions and their effect on carcinogenesis
accounted for in delineating these interactions. and treatment response is beginning to be elucidated. In
Variations in the gut microbiome were first associated order to optimally deliver care to their patients, it is increas-
with response to immunotherapy in murine models that were ingly important for medical and surgical oncologists to
later translated to clinical associations demonstrating the gut remain abreast of these developments.
microbiome-modulated response to anti-PD-1 therapy [54]. While induction of gut dysbiosis and downstream detri-
The gut flora has since been shown to shape response to mental effects can readily be obtained and modeled, the larg-
checkpoint inhibitors across multiple tumor types: epithelial est hurdle the investigators are facing has been the definition
cancers, pancreatic adenocarcinoma, and melanoma [54– and induction of the “ideal” microbiome with appropriate
57]. This is of particular clinical importance as the use of diversity and specific taxa to optimize both general and
antibiotics at time of checkpoint inhibitor initiation has been oncologic health. The question of what is the optimal micro-
associated with worse outcomes in multiple studies, again biome remains to be answered. In fact, there likely is no
suggesting dysbiosis-mediated effects on treatment response “ideal” microbiome, but one can optimize an outcome of
[58]. As antibiotic administration is not uncommon among interest at the expense of other outcomes. To date, interven-
patients receiving systemic therapy for their malignancy, its tions to improve a patient’s microbiome have largely treated
effect on induced gut dysbiosis and downstream outcomes benign disease such as Clostridium difficile infections or irri-
should be considered. In elucidating the mechanism of this table bowel disease. However, microbiome modulation via
effect, murine models of ICB have been invaluable in dem- fecal transplant is now being tested in clinical trials, while
onstrating improved results with fecal microbiota transplan- promising results have been noted in optimizing oncologic
tation (FMT) in immunotherapy-treated mice [59]. These care in a variety of murine models. Moving forward, the
findings have been recently translated to a phase 1 clinical effect diet may have on optimization of immunomodulatory
trial by Baruch and colleagues demonstrating FMT is able to cancer therapies has the potential to improve outcomes and
induce a response to ICB in otherwise immunotherapy-­ possibly affect the incidence of malignancies at a population
refractory melanoma patients. This work was able to find level. More directed alterations of the microbiome through
that in a group of ten heavily pretreated patients with meta- the development of fecal consortia, designer probiotics, or
static melanoma, FMT resulted in a more diverse microbi- individual bacterial metabolites may also lead to further
ome, improved immune infiltrate into tumors, and three advancements in the field and will likely be adapted to the
clinical responses to treatment including one complete optimization of oncology care to optimize cancer-specific
response [60]. outcomes as well as minimize adverse effects of treatments.
There are a number of potential mechanisms actively Significant questions exist studying the link between the
being investigated to delineate the effects the gut microbiome commensal microbiome and malignancy. The investigation
has on immunotherapies. These mechanisms include both of these questions and their answers will direct the future of
immunostimulatory and immunosuppressive effects. In the this field:
case of immunostimulatory effects, proposed mechanisms
involve altered antigenicity through T-cell activation. This has
been proposed to occur through antigenic, adjuvantic, or • What is the optimal composition of the gut microbi-
bystander effects via unactivated or conditioned dendritic ome to optimize oncologic outcomes?
cells and cytotoxic T lymphocytes. Similarly, immunosup- • How can the commensal gut microbial composition
pressive effects are thought to be mediated through local sup- and the associated metabolome be influenced to
pressor cells or through locally produced bacterial metabolites. optimize outcomes whether through dietary
As previously mentioned, cytokines or microbial-derived changes, bacterial consortia, or FMT?
metabolites can also have significant immunostimulatory or • Delineation of microbial mechanisms behind carci-
immunosuppressive effects with downstream effects on phar- nogenesis, progression of disease, evasion of host
macologically invoked immunomodulation [61]. immune surveillance, and cytotoxic therapies.
• How can the circulating or stool microbial signa-
tures be leveraged to augment or supplant existing
73.2 Conclusion screening or surveillance strategies for various
malignancies?
The diverse interactions between a patient and their microbi- • Can the tumoral virome offer insight correlations
ome clearly play an important role as a determinant of over- and potential further causative associations with
all and oncologic health. The complexity of a system that malignancy?
varies between patients, within patients, and over time, how-
772 M. G. White et al.

References reveal the presence of different bacterial and archaeal biomarkers. J

Gastroenterol. 2015;50(2):167–79.
22. Zackular JP, Rogers MA, Ruffin MTt, Schloss PD. The human gut
1. Sender R, Fuchs S, Milo R. Revised estimates for the num-
microbiome as a screening tool for colorectal cancer. Cancer Prev
ber of human and bacteria cells in the body. PLoS Biol.
Res (Phila). 2014;7(11):1112–21.
2016;14(8):e1002533.
23. Flemer B, Lynch DB, Brown JM, et al. Tumour-associated and
2. Nejman D, Livyatan I, Fuks G, et al. The human tumor microbiome
non-tumour-associated microbiota in colorectal cancer. Gut.
is composed of tumor type-specific intracellular bacteria. Science.
2017;66(4):633–43.
2020;368(6494):973–80.
24. Tsay JJ, Wu BG, Sulaiman I, et al. Lower airway dysbiosis affects
3. Valdes AM, Walter J, Segal E, Spector TD. Role of the gut micro-
lung cancer progression. Cancer Discov. 2021;11(2):293–307.
biota in nutrition and health. BMJ. 2018;361:k 2179.
25. Kostic AD, Gevers D, Pedamallu CS, et al. Genomic analysis
4. de Martel C, Ferlay J, Franceschi S, et al. Global burden of cancers
identifies association of fusobacterium with colorectal carcinoma.
attributable to infections in 2008: a review and synthetic analysis.
Genome Res. 2012;22(2):292–8.
Lancet Oncol. 2012;13(6):607–15.
26. Hamada T, Zhang X, Mima K, et al. Fusobacterium nucleatum
5. Amieva M, Peek RM Jr. Pathobiology of helicobacter pylori-­
in colorectal cancer relates to immune response differentially by
induced gastric cancer. Gastroenterology. 2016;150(1):64–78.
tumor microsatellite instability Status. Cancer Immunol Res.
6. Kodaman N, Pazos A, Schneider BG, et al. Human and helicobacter
2018;6(11):1327–36.
pylori coevolution shapes the risk of gastric disease. Proc Natl Acad
27. Hashemi Goradel N, Heidarzadeh S, Jahangiri S, et al.
Sci. 2014;111(4):1455–60.
Fusobacterium nucleatum and colorectal cancer: a mechanistic
7. Nagy TA, Frey MR, Yan F, Israel DA, Polk DB, Peek RM Jr.
overview. J Cell Physiol. 2019;234(3):2337–44.
Helicobacter pylori regulates cellular migration and apoptosis by
28. Bullman S, Pedamallu CS, Sicinska E, et al. Analysis of fusobac-
activation of phosphatidylinositol 3-kinase signaling. J Infect Dis.
terium persistence and antibiotic response in colorectal cancer.
2009;199(5):641–51.
Science. 2017;358(6369):1443–8.
8. Suzuki M, Mimuro H, Kiga K, et al. Helicobacter pylori Cag A
29. Pushalkar S, Hundeyin M, Daley D, et al. The pancreatic cancer
phosphorylation-independent function in epithelial proliferation
microbiome promotes oncogenesis by induction of innate and
and inflammation. Cell Host Microbe. 2009;5(1):23–34.
adaptive immune suppression. Cancer Discov. 2018;8(4):403–16.
9. Higashi H, Tsutsumi R, Fujita A, et al. Biological activity of the
30. Xavier JB, Young VB, Skufca J, et al. The cancer microbiome:
helicobacter pylori virulence factor Cag A is determined by varia-
distinguishing direct and indirect effects requires a systemic view.
tion in the tyrosine phosphorylation sites. Proc Natl Acad Sci.
Trends Cancer. 2020;6(3):192–04.
2002;99(22):14428–33.
31. Riquelme E, Zhang Y, Zhang L, et al. Tumor microbiome diver-
10. Tsang YH, Lamb A, Romero-Gallo J, et al. Helicobacter pylori
sity and composition influence pancreatic cancer outcomes. Cell.
Cag A targets gastric tumor suppressor RUNX3 for proteasome-­
2019;178(4):795–806 e712.
mediated degradation. Oncogene. 2010;29(41):5643–50.
32. Mima K, Nishihara R, Qian ZR, et al. Fusobacterium nuclea-
11. Buti L, Spooner E, Van der Veen AG, Rappuoli R, Covacci A,
tum in colorectal carcinoma tissue and patient prognosis. Gut.
Ploegh HL. Helicobacter pylori cytotoxin-associated gene A (Cag
2016;65(12):1973–80.
A) subverts the apoptosis-stimulating protein of p 53 (ASPP2)
33. Tahara T, Yamamoto E, Suzuki H, et al. Fusobacterium in colonic
tumor suppressor pathway of the host. Proc Natl Acad Sci.
flora and molecular features of colorectal carcinoma. Cancer Res.
2011;108(22):9238–43.
2014;74(5):1311–8.
12. Barrozo RM, Cooke CL, Hansen LM, et al. Functional plasticity in
34. Rubinstein MR, Wang X, Liu W, Hao Y, Cai G, Han
the type IV secretion system of Helicobacter pylori. PLoS Pathog.
YW. Fusobacterium nucleatum promotes colorectal carcinogen-
2013;9(2):e1003189.
esis by modulating E-cadherin/beta-catenin signaling via its Fad A
13. Machado AM, Figueiredo C, Touati E, et al. Helicobacter pylori
adhesin. Cell Host Microbe. 2013;14(2):195–06.
infection induces genetic instability of nuclear and mitochondrial
35. Purcell RV, Pearson J, Aitchison A, Dixon L, Frizelle FA,
DNA in gastric cells. Clin Cancer Res. 2009;15(9):2995–3002.
Keenan JI. Colonization with enterotoxigenic Bacteroides fragi-
14. Arzumanyan A, Reis HM, Feitelson MA. Pathogenic mechanisms
lis is associated with early-stage colorectal neoplasia. PLoS One.
in HBV- and HCV-associated hepatocellular carcinoma. Nat Rev
2017;12(2):e0171602.
Cancer. 2013;13(2):123–35.
36. Kostic AD, Chun E, Robertson L, et al. Fusobacterium nucleatum
15. Gupta S, Kumar P, Das BC. HPV: molecular pathways and targets.
potentiates intestinal tumorigenesis and modulates the tumor-­
Curr Probl Cancer. 2018;42(2):161–74.
immune microenvironment. Cell Host Microbe. 2013;14(2):207–15.
16. Guo F, Cofie LE, Berenson AB. Cervical cancer Incidence in young
37. Chen Y, Chen Y, Zhang J, et al. Fusobacterium nucleatum pro-
U.S. females after human papillomavirus vaccine introduction. Am
motes metastasis in colorectal cancer by activating autophagy sig-
J Prev Med. 2018;55(2):197–04.
naling via the upregulation of CARD3 expression. Theranostics.
17. Lei J, Ploner A, Elfstrom KM, et al. HPV vaccination and the risk
2020;10(1):323–39.
of invasive cervical cancer. N Engl J Med. 2020;383(14):1340–8.
38. Wu S, Lim KC, Huang J, Saidi RF, Sears CL. Bacteroides fragilis
18. Calvaruso V, Cabibbo G, Cacciola I, et al. Incidence of hepatocellu-
enterotoxin cleaves the Zonula Adherens protein, E-cadherin. Proc
lar carcinoma in patients with HCV-associated cirrhosis treated with
Natl Acad Sci. 1998;95(25):14979–84.
direct-acting antiviral agents. Gastroenterology. 2018;155(2):411–
39. Wynendaele E, Verbeke F, D'Hondt M, et al. Crosstalk between the
421 e414.
microbiome and cancer cells by quorum sensing peptides. Peptides.
19. Goedert JJ, Gong Y, Hua X, et al. Fecal microbiota characteris-
2015;64:40–8.
tics of patients with colorectal adenoma detected by screening: a
40. Ma C, Han M, Heinrich B, et al. Gut microbiome-mediated bile
population-­based study. EBioMedicine. 2015;2(6):597–03.
acid metabolism regulates liver cancer via NKT cells. Science.
20. Shen XJ, Rawls JF, Randall T, et al. Molecular characterization of
2018;360(6391).
mucosal adherent bacteria and associations with colorectal adeno-
41. Banerjee S, Tian T, Wei Z, et al. Distinct microbial signatures
mas. Gut Microbes. 2010;1(3):138–47.
associated with different breast cancer types. Front Microbiol.
21. Mira-Pascual L, Cabrera-Rubio R, Ocon S, et al. Microbial mucosal
2018;9:951.
colonic shifts associated with the development of colorectal cancer
73 Microbiome and Cancer Metastasis 773

42. Buchta Rosean C, Bostic RR, Ferey JCM, et al. Preexisting com- 52. Yu T, Guo F, Yu Y, et al. Fusobacterium nucleatum promotes che-
mensal dysbiosis is a host-intrinsic regulator of tissue inflammation moresistance to colorectal cancer by modulating autophagy. Cell.
and tumor cell dissemination in hormone receptor-positive breast 2017;170(3):548–563 e516.
cancer. Cancer Res. 2019;79(14):3662–75. 53. Viaud S, Saccheri F, Mignot G, et al. The intestinal microbiota
43. Arends J. Metabolism in cancer patients. Anticancer Res. modulates the anticancer immune effects of cyclophosphamide.
2010;30(5):1863–8. Science. 2013;342(6161):971–6.
44. Fujisaka S, Avila-Pacheco J, Soto M, et al. Diet, genetics, and the 54. Gopalakrishnan V, Spencer CN, Nezi L, et al. Gut microbiome
gut microbiome drive dynamic changes in plasma metabolites. Cell modulates response to anti-PD-1 immunotherapy in melanoma
Rep. 2018;22(11):3072–86. patients. Science. 2018;359(6371):97–103.
45. Schulfer AF, Schluter J, Zhang Y, et al. The impact of early-­ 55. Matson V, Fessler J, Bao R, et al. The commensal microbiome is
life sub-therapeutic antibiotic treatment (STAT) on excessive associated with anti-PD-1 efficacy in metastatic melanoma patients.
weight is robust despite transfer of intestinal microbes. ISME J. Science. 2018;359(6371):104–8.
2019;13(5):1280–92. 56. Temraz S, Nassar F, Nasr R, Charafeddine M, Mukherji D,
46. Kwa M, Plottel CS, Blaser MJ, Adams S. The intestinal microbi- Shamseddine A. Gut microbiome: a promising biomarker
ome and estrogen receptor-positive female breast cancer. J Natl for immunotherapy in colorectal cancer. Int J Mol Sci.
Cancer Inst. 2016;108(8):djw02. 2019;20(17):4155.
47. Kadosh E, Snir-Alkalay I, Venkatachalam A, et al. The gut microbi- 57. Routy B, Le Chatelier E, Derosa L, et al. Gut microbiome influ-
ome switches mutant p 53 from tumour-suppressive to oncogenic. ences efficacy of PD-1-based immunotherapy against epithelial
Nature. 2020;586(7827):133–138. tumors. Science. 2018;359(6371):91–7.
48. Poore GD, Kopylova E, Zhu Q, et al. Microbiome analyses of 58. Pinato DJ, Howlett S, Ottaviani D, et al. Association of prior anti-
blood and tissues suggest cancer diagnostic approach. Nature. biotic treatment with survival and response to immune check-
2020;579(7800):567–74. point inhibitor therapy in patients with cancer. JAMA Oncol.
49. Gopalakrishnan V, Helmink BA, Spencer CN, Reuben A, Wargo 2019;5(12):1774–78
JA. The influence of the gut microbiome on cancer, immunity, and 59. Vetizou M, Pitt JM, Daillere R, et al. Anticancer immunother-
cancer immunotherapy. Cancer Cell. 2018;33(4):570–80. apy by CTLA-4 blockade relies on the gut microbiota. Science.
50. Geller LT, Barzily-Rokni M, Danino T, et al. Potential 2015;350(6264):1079–84.
role of intratumor bacteria in mediating tumor resis- 60. Baruch EN, Youngster I, Ben-Betzalel G, et al. Fecal microbiota
tance to the chemotherapeutic drug gemcitabine. Science. transplant promotes response in immunotherapy-refractory mela-
2017;357(6356):1156–60. noma patients. Science. 2021;371(6529):602–609.
51. Guthrie L, Gupta S, Daily J, Kelly L. Human microbiome sig- 61. Fessler J, Matson V, Gajewski TF. Exploring the emerging role of
natures of differential colorectal cancer drug metabolism. NPJ the microbiome in cancer immunotherapy. J Immunother Cancer.
Biofilms Microbiomes. 2017;3:27. 2019;7(1):108.
 Application of Artificial Intelligence
in Research on Cancer and Its 74
Metastasis

Benjamin Franc

Abstract
Learning Objectives
Artificial intelligence (AI) is composed of a number of • To understand the basic tools of artificial intelli-
supervised and unsupervised computational learning gence and how their selection relates to the avail-
techniques. At present, cancer research utilizes some of able data sources.
these techniques to analyze and cluster large datasets, • To identify key areas where artificial intelligence
such as those resulting from genomic or proteomic techniques are already used in the analysis and clus-
analyses. AI is also beginning to enable automation of tering of data in cancer research.
research tasks, increasing efficiency and reproducibil- • To demonstrate areas in cancer research where arti-
ity in classifications, and extending capabilities in ficial intelligence techniques can contribute greater
basic science and clinical research beyond what is speed and reproducibility.
readily achievable by human investigators alone. The • To discuss the role of artificial intelligence in can-
complex and nonlinear interactions between cancer cer drug discovery and drug selection.
and its environment can be modeled with AI tech- • To highlight artificial intelligence research as an
niques, allowing new discoveries in the molecular emerging field within the broader field of cancer
pathways and signaling networks leading to cancer and research.
its metastasis to local and distant sites. Meanwhile, in
cancer drug discovery and clinical trials, AI models
predict the chemical, biologic, and pharmaceutical
properties of small-molecule candidates and new appli-
cations of existing drugs and identify characteristics of
74.1 Introduction
patients in whom higher drug efficacy might be
Artificial intelligence (AI) includes a wide array of compu-
expected in clinical trials. Through facilitating analysis
tational techniques, some of which have become embedded
of sources of big data and increasing the rate of discov-
in cancer research already. Others are only just beginning to
ery in cancer research, AI promises to drive the field of
be explored as tools to better understand the complex and
cancer research forward at an exponential rate over the
interdependent signals and events that contribute to cancer
upcoming decade.
genesis and growth. An improvement over traditional com-
putational modeling techniques, AI can address the systems
biology inherent in studying cancer as well as cancer’s con-
tinuing evolutionary path. Moreover, AI can accommodate
the complex and nonlinear interactions between cancer and
its environment that underlie metastasis [1]. Following a
B. Franc (*) brief overview of a sampling of AI tools, this chapter will
Department of Radiology, Division of Nuclear Medicine and
Molecular Imaging, Stanford University, Stanford, CA, USA explore a range of areas where AI can uniquely contribute to
e-mail: [email protected] research on cancer and its metastases.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 775
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4_74
776 B. Franc

74.1.1 The Tools of AI capable of discovering complex nonlinear relationships

between features and labels and have, for example, shown
AI denotes a process of observational learning of patterns by problem-solving capabilities similar to or exceeding those of
a computer. The distinction between modern AI and earlier humans [3–5].
model-driven computational algorithms is the current lack of
a requirement for a priori knowledge of an underlying pro-
cess. Machine learning (ML), a subfield of AI, focuses on 74.2  I as a Basic Tool to Increase
A
developing algorithms that learn to perform tasks or make the Capabilities and Efficiency
predictions from data. of Cancer Research
ML can be subdivided into supervised and unsupervised
learning techniques (Fig. 74.1). Supervised techniques are 74.2.1 D
 riving Data Analysis and Clustering
applied in learning tasks when data is organized as input in Cancer Research
examples (features) with corresponding outputs (labels), and
learning consists of building a specific statistical inference Genomics, proteomics, and many other “-omics” generate
model. Supervised learning techniques include regression extremely large datasets that AI routinely helps researchers in
and classification algorithms such as linear and logistic parsing. For example, in the field of proteomics, ML is used
regression, discriminant analysis, decision tree, random for- to predict and quantify proteins present within mass spec-
est, naïve Bayes, support vector machines (SVM), and neural trometry data and identify biomarkers of disease as well as in
networks. These types of ML models have been evolving for parameter estimation and sequence modeling in protein’s
decades and continue to be the most commonly use in health- multiple sequence alignment [6]. In analyzing microarray
care [3]. Preparing datasets with features and labels can be data, many ML-based methods have been developed to assist
time-consuming; however once accomplished, supervised in the selection of reliable gene predictors for disease and the
ML algorithms can be trained and tested rapidly. final complete set of genes for the classification task [7].
Unsupervised techniques find patterns in data without the AI has been key to the proliferation of the field of cancer
need for data annotation and produce classes within a dataset radiomics, a field focused on identifying imaging pheno-
through structural learning. Clustering analysis and principal types that can be used in diagnosis, characterization, and
component analysis (PCA) are techniques typically used for treatment management. These imaging phenotypes can be
these tasks. defined by pre-defined statistical imaging features, of
The availability of large digital datasets and modern com- which over 5000 have been described. Such radiomic fea-
putational speed, a subset of ML techniques, classified as tures can identify key components of tumor phenotype for
deep learning (DL), have rapidly evolved and matured. DL one or multiple lesions at a single or multiple time points
techniques are often supervised learning that uses artificial over the course of treatment [8], potentially allowing
neural networks (ANNs). ANNs use multiple layers of non- enhanced patient stratification, prognostication, and treat-
linear modules that accept features as inputs, progressively ment monitoring for targeted therapies. Radiogenomics,
extract higher-level features, and ultimately predict labeled the translation of intratumoral phenotypic features to geno-
outputs. Examples of ANN include convoluted neural net- types, requires the development of new methodologies to
works (CNNs), deep belief networks (DBN), and recurrent summarize phenotypes of large heterogeneous populations
neural networks (RNN). DL may also be used in unsuper- of cells within a single tumor and look for underlying geno-
vised or semi-supervised learning tasks. DL techniques are typic similarities [9].

Fig. 74.1 Machine learning

techniques. Machine learning Supervise learning: each
techniques include two major training example has a
classes: supervised and ground truth label. The model
unsupervised (From https:// learns a decision boundary
link.springer.com/ and replicates the labelling on
new data.
article/10.1007%2Fs00117-­
018-­0407-­3) [2]
Unsupervise learning:
training examples do not
have ground truth label. The
model identifies structure
such as clusters. New data
can be assigned to clusters.
Training date Resulting model Applied to new input
74 Application of Artificial Intelligence in Research on Cancer and Its Metastasis 777

By enabling automated extraction of phenotype informa- specifications of the imaging device and associated post-­
tion from electronic medical records, natural language pro- processing software, challenges to which various solutions
cessing (NLP), a type of machine learning, has evolved to have been proposed [19].
become an important tool in research seeking to find associa- By making the entire slide available for analysis, whole-­
tions between genomic data and the manifestations and time- slide imaging enables a computer to select the field to ana-
line of various cancer subtypes. Data discovery can be on lyze rather than leaving this task to a researcher or clinical
primary data or even data from the literature, for example, pathologist. For example, many DL convolutional neural
using NLP to identify every study in the literature about a network (CNN) approaches have been applied to the histo-
specified gene-cancer association and using that data to opti- logical classifications of gastric cancer (adenocarcinoma,
mize calculation of accurate risk estimates of developing adenoma, and non-neoplastic) from whole-slide images
cancer for a cancer susceptibility gene of interest [10]. Many (WSIs), some of which have focused on local details, while
NLP tools are open source [11]. others have analyzed the morphological features from the
whole image [16]. One of the most common ML model-­
based approaches to segmentation is nuclei segmentation to
74.2.2 Automation of Tasks in Cancer Research extract features for nuclear morphometrics, an area to which
deep learning techniques have recently been applied with
AI-driven automated identification of cancer on imaging is significant success [19]. When applying segmentation mod-
useful in both basic and clinical cancer research settings. The els utilized to automatically select a region of interest for
ability to use AI techniques to rapidly and objectively iden- subsequent classification of the nucleus by the diagnostic AI
tify and measure cancer and metastatic disease in animal model, such models may not be able to be generalized to sets
models on whole-body tomographic imaging, such as com- of WSIs when the training and test images belong to differ-
puted tomography (CT) or positron emission tomography ent organs. This challenge may be addressed by training
(PET) [12, 13], or detect and grade cancers on histologic sec- nuclei segmentation algorithms on images taken from differ-
tions opens the doors to higher research efficiency and repro- ent organs [19].
ducibility [14]. The power of AI over the experience of any In developing a CNN capable of identifying breast cancer
single researcher, imaging physician, or pathologist is the metastases in lymph nodes on WSIs, Litjens and colleagues
ability to cross-reference imaging or other data from indi- used small image regions around areas annotated by a pathol-
vidual tumors to databases of comparisons [15]. As with ogist to train the algorithm to provide per-pixel predictions of
many ML applications, finding sources of labeled data on the presence of metastases. The authors found that, using the
which to train models has posed a challenge. Creative trained algorithm, all micro- and macro-metastases of breast
approaches have been applied to overcome the challenge of cancer could be identified automatically. The authors also
a relatively small number of well-annotated slides, such as reported that, of those slides containing benign or normal tis-
an intermediate step where a separate DL model predicts and sue, 30–40% were excluded without any additional immuno-
segments the area of the slide where a cancer would most histochemical markers or human intervention [20]. This
likely be found, thus allowing the histologic classifier model points out the issue of false positives in the task of classifying
greater predictive accuracy [16]. regions in lymph nodes, often secondary to the presence of
AI solutions for pathology have been shown to make reactive lymphoid follicles. In order to address this complex-
diagnoses over tenfold faster than pathologists [17]. In a ity, a two-step process has been developed wherein a WSI is
study comparing the accuracy of various AI algorithms and analyzed by a deep learning algorithm to exclude regions
experienced pathologists in diagnosing metastatic disease on containing lymphoid follicles, followed by a second algo-
whole-slide images (WSIs) of lymph nodes from women rithm that detects malignant cells, resulting in an error reduc-
with breast cancer, the accuracies of some AI algorithms in tion of 89% in slides with reactive lymphoid follicles [21].
classifying WSIs as having metastatic disease or not were In addition to the task of differentiating malignant from
comparable to the accuracies of an experienced pathologist, non-malignant tissue on histopathology, ML methods have
and the mean area under the receiver operating characteristic been utilized to detect tubular formation, nuclear pleomor-
curve (AUC) of the five most accurate AI algorithms was phism, and other tasks of tumor grading as well as biomarker
comparable to that of a pathologist interpreting the slides assessment in immunohistochemically stained images [22].
without time constraints [18]. Some considerations must be An example of DL applied to grading prostate cancer tissue
made when contemplating the use of a pretrained model to microarray spot images is provided in Fig. 74.2. Automated
identify cancer on histopathological images including differ- AI-based cytologic analyses have also been developed for
ences in age of the sample, method of slide preparation, and various cancer and specimen types [24].
778 B. Franc

Fig. 74.2 Schematic of deep learning model training used by Arvaniti image, and a final output layer that produces the final classification
and colleagues to enable automated Gleason grading of prostate cancer decision for each input image patch by computing a probability distri-
tissue microarrays. (a) Pathologists annotate grades of prostate cancer bution over the four Gleason classes. (Bottom row) The trained patch-­
on tissue microarray (TMA) spot images (blue, Gleason pattern 3 level convolutional neural network is then applied to the entire TMA
region; yellow, Gleason pattern 4 region; red, Gleason pattern 5 region). spot image, generating pixel-level probability maps for each of the four
(b) (Top row) A deep neural network trains as a patch-level classifier Gleason classes. Finally, a TMA spot is assigned a Gleason score (the
using an architecture based on convolution blocks to extract increas- sum of the primary and secondary Gleason patterns detected in the
ingly complex features from the input image, a subsequent global aver- pixel-level maps) (From https://doi.org/10.1101/280024) [23]
age pooling layer that summarizes the local patterns across the entire

74.2.3 Enabling New Technologies also been employed to train an algorithm that can identify
for Cancer Research the cancer subtype (as well as healthy tissue) based upon
mRNA transcription profiles [27]. Finally, some initial work
In addition to increasing efficiency and reproducibility in suggests the ability of AI methods to differentiate between
classifications that can also be made by humans, AI can cancer cells and their radioresistant clones [28].
improve on and extend capabilities by performing tasks that Flow cytometry, a critical tool that uses anisotropic elastic
humans simply can’t. For example, Oei and colleagues have light scattering of individual cells and emission of fluores-
recently applied CNNs to classify the malignancy of cells cent labels to enable large-scale cell analysis in the labora-
based on intracellular actin patterns on actin-labeled fluores- tory, has the limitation of providing only a single value per
cence microscopy images [25]. Zheng and colleagues have detection channel and potential adverse effects of conjugated
applied DL to identify the origins of unknown primary can- fluorophores on cellular viability and cell signaling. Label-­
cers from their metastases and identify cancer types of circu- free cell assay techniques aim to remedy this, but sample
lating cancer cells based on DNA methylation [26]. DL has sizes are limited, and the assays often cannot sufficiently
74 Application of Artificial Intelligence in Research on Cancer and Its Metastasis 779

differentiate cells because they rely on a single feature. Using In the analysis of data collected from clinical trials, indi-
a high-throughput system that acquired quantitative optical cators of response of cancer to a therapy are not represented
phase and intensity images of individual cells and extracted by low-dimensional data that can be easily analyzed;
from them multiple biophysical measurements, Chen and instead, response is now measured through changes on
colleagues integrated feature extraction and supervised imaging, dose-response curves, or other high-dimensional
learning to perform accurate classification of cells. As the data for which analysis is much more complex [35]. Trained
authors themselves indicate, this type of label-free high-­ ML algorithms can determine tumor volume and its change
throughput cell analysis will be critical to the future of per- over time and tracking of multiple lesions [15]. Automated
sonalized genomics, cancer diagnostics, and pharmaceutical PET segmentation of nodules based on neural networks
development [29]. Similarly, DL techniques have advanced trained in the spatial and wavelet domains has been shown
single-cell analysis techniques based on morphological to be reproducible and volumetrically accurate and demon-
behavior on functionalized substrates [30] and 2D cell track- strate lower absolute relative error when compared to other
ing and whole-cell segmentation technology in cell cultures, automated techniques [36]. Other ML approaches have
enabling efficient studies of cellular dynamics characteriza- been useful in dealing with segmentation of larger and
tion and drug efficacy [31]. more complicated tumors of the head and neck, particularly
A very exciting set of new diagnostic tools may result in the setting of heterogeneous radiopharmaceutical uptake,
from AI-driven correlation of digital pathology and radiol- in segmenting brain tumors and classifying brain scans
ogy, also referred to as radiology-pathology correlation. [37–39]. Given the immense amount of data available in the
Registration of WSIs from adjacent tissue slices allows medical record, NLP makes it possible to parse out data
three-dimensional pathology datasets containing both tumors about the response of cancer to various treatments directly
and their environments [19]. Registration between tissue from patient records, allowing insight into drug efficacy in
slices enables colocalization of pathology data with radio- its “real-world” application, and enables pharmacovigi-
graphic imaging data. For example, in prostate cancer, ML lance and drug-safety surveillance studies after regulatory
techniques have enabled complex histopathologic spatial approval [40].
analysis with radiologic imaging, revealing that multipara-
metric magnetic resonance imaging (MRI) signal character-
istics are associated with tissue composition density of 74.3  lucidating the Origins and Molecular
E
stroma, epithelium, epithelial nuclei, and lumen, and these Signaling Pathways of Cancer and Its
densities can be used to delineate cancerous regions of vary- Metastasis Using AI
ing grades. In a similar analysis, epithelium and lumen den-
sity can be used to distinguish cancerous regions of varying Understanding the molecular pathways that lead to the gen-
grade [32]. Such technologies have the potential to enable esis and growth of cancers is key to uncovering biomarkers
in vivo monitoring and characterization of tissues at the his- useful in diagnosis and targets for new therapies. For exam-
tologic level via medical imaging techniques [19]. ple, ANNs have been used to predict subcellular localization
of protein sequences using the amino acid sequence and fea-
tures predicted from it, such as the N-terminal-sorting signal
74.2.4 C
 reating Efficiencies in Clinical [41]. In the study of the loss of differentiation associated
Research with cancers, ML techniques have uncovered novel dedif-
ferentiation mechanisms and novel associations of immune
AI is beginning to be employed in prospective clinical trial checkpoint expression and infiltrating immune cells with the
recruitment enriching – strategies to help define candidates level of cancer cell dedifferentiation [42].
most likely to benefit from an intervention. For example, ML Clues to molecular mechanisms of a cancer can be found in
models can elucidate the most predictive features associated its genome, but the relationship between the presence of spe-
with recurrence, such as in studies seeking population-­specific cific mutations and carcinogenesis is rarely linear in nature,
predictors using datasets of a large number of breast cancer making it particularly attractive to utilize current AI techniques
patients, and may then be used to define a set of subjects who to help unravel these relationships. In the task of differentiating
may benefit from a novel adjuvant therapy [33]. NLP is being driver mutations (those that directly contribute to a cancer’s
looked to as a means of identifying a more diverse set of clini- development or growth) from passenger mutations (those that
cal trial candidates and doing it more r­ apidly by using standard do not directly contribute), several ML approaches have been
clinical trial recruitment processes [34]. applied. For example, the cancer-­ specific high-throughput
780 B. Franc

annotation of somatic mutation (CHASM) approach trained an from the invasive tumor front and targeted proteomic analy-
RF model using known driver mutations from a curated data- sis of proteins in saliva yielded a peptide-based signature
base or literature indicating a biologic oncogenic role and capable of differentiating between patients who will have
known passenger mutations defined by background base lymph node metastases [47].
substitution rates. Models were trained based on the features of
amino acid substitution properties, alignment-based estimates
of conservation at the mutated position, predicted local struc- 74.4  ancer Drug Discovery, Development,
C
ture at the mutated position, and others [43]. and Selection with AI
In addition, while traditionally cancer was viewed from
the perspective of its organ of origin, it is now appreciated In drug development, AI models predict the chemical, bio-
that cancer of a single organ may consist of numerous logic, and pharmaceutical properties of small-molecule can-
subtypes, each expressing different biologic behavior based didates and new applications of existing drugs and identify
on its genetic profile. The problem of identifying these characteristics of patients in whom higher drug efficacy
molecular subtypes requires analysis and clustering of gene might be expected in clinical trials [48–51].
expression profiles, a task that has also benefited from appli- One of the greatest strengths of AI is being able to draw
cation of ML approaches. For example, instead of addressing upon multiple large datasets – such as databases of cancer
high-­dimensional gene expression data head-on using tradi- drugs, drug resistance, next-generation sequencing, genetic
tional clustering methods, an ANN model with a single layer variants, and structural biology – at once to solve a problem
of hidden nodes can be trained, and the observations at the (Fig. 74.3). AI in combination with sources of a large amount
hidden nodes can then be used for the clustering analysis of data in the biosciences has been used to build in silico
[44]. Likewise, AI has helped elucidate the role of the tumor models of cancer, to enable computer-aided design and test-
microenvironment surrounding cancers [45]. ing of potential therapeutic compounds [53]. For example,
Beyond its contributions to better understand carcinogen- an RF-based model has been trained to screen compounds
esis, AI has been used to elucidate the phenomenon of metas- for their ability to serve as substrates or inhibitors to proteins
tasis, a process involving both the tumor itself and its conferring multi-drug resistance [54].
surroundings, including its immune microenvironment [46]. One of the most rapidly advancing classes of cancer ther-
In oral squamous cell carcinoma, ML techniques combined apies includes drugs modulating the immune response to
with discovery proteomics on neoplastic islands and stroma cancer as well as antitumor vaccines. One key to design

Target identification Compound screening Preclinical Clinical

and validation and lead discovery development development

Successful applications in drug discovery

• Target identification and • Compound design with • Tissue-specific biomarker • Determination of drug
prioritization based on desirable properties identification response by cellular
gene–disease associations • Compound synthesis • Classification of cancer phenotyping in oncology
• Target druggability predictions reaction plans drug–response signatures • Precise measurements of the
• Identification of alternative • Ligand-based • Prediction of biomarkers tumour microenvironment
targets (splice variants) compound screening of clinical end points in immuno-oncology

Required data characteristics

• Current data are highly • Large amounts of • Biomarkers: • Pathology: well-curated
heterogeneous: need training data needed reproducibility of models expert annotations for
standardized high-dimensional • Models for compound based on gene broad-use cases (cancer
target–disease–drug reaction space and expression data versus normal cells)
association data sets rules • Dimension reduction of • Gold standard data sets to
• Comprehensive omics data • Gold standard ADME single-cell data for cell improve interpretability and
from disease and normal states data type and biomarker transparency of models
• High-confidence associations • Numerous protein identification • Sample size: high number of
from the literature structures • Proteomic and images per clinical trail
• Metadata from successful and transcriptomic data of
failed clinical trails high quality and quantity

Fig. 74.3 AI applications in cancer drug discovery and development. distribution, metabolism, and excretion) (From https://www-­nature-­
Applications of machine learning in various steps of the drug develop- com.stanford.idm.oclc.org/articles/s41573-­019-­0024-­5) [52]
ment process and associated data requirements. (ADME = absorption,
74 Application of Artificial Intelligence in Research on Cancer and Its Metastasis 781

immune-mediated strategies involves identifying tumor-­ 74.5  I as an Emerging Field in Cancer

A
associated antigens and understanding their role in eliciting Research
an immune response. Large immunoinformatics databases,
such as the publicly available Human Potential Tumor 74.5.1 Cancer Diagnosis
Associated Antigen.
(HPtaa) database (http://www.hptaa.org), have been Given its particularly complicated origins including genetic
assembled. ML can be helpful in the task of identifying the and environmental factors interacting with the immune sys-
small number of antigens expressed by specific cancer cells tem and other normal tissues in the body, cancer diagnosis is
that can be targeted to trigger an immune response to them. an area of AI application, using data from general health
In addition to techniques such as quantitative structure-­ screening and diagnostic tests, including blood testing, imag-
activity relationship (QSAR) analysis, homology modeling, ing, and pathology [60].
protein threading, and docking methodologies, machine
learning also is used to predict epitopes of B and T cells as
well as vaccine candidates. Relatively accurate ML-based 74.5.2 Cancer Treatment
prediction models have been trained for both continuous epi-
topes (those based on linear sequences of amino acids) and AI has become a pillar of personalized healthcare whereby
discontinuous ones. To advance T-cell epitope prediction, ML algorithms identify patterns of disease within a host of
several ANNs have been trained to predict the binding affin- data sources to inform therapeutic choices [61, 62]. For
ity of novel peptide sequences to specific major histocompat- example, using an ML algorithm trained on computed
ibility complex (MHC) class I and II molecules [55]. tomography (CT) radiomic features and dosimetric parame-
AI has also enabled an interesting approach to appropriate ters, the response of patients with esophageal cancer (EC)
drug selection that takes into account features of the drug and undergoing concurrent chemoradiation (CRT) can be pre-
the cell line. Using a combination of ML techniques, the dicted [63].
response of specific cancer cell lines to a therapy has been
predicted based on both the genomic features of the cell lines
and the chemical properties of the drugs, making possible in 74.5.3 D
 evelopment of Models to Predict
silico efficacy studies and patient-specific drug selection Cancer Outcomes and Prognosis
based on genomic characteristics [56]. In the testing of lead
drug candidates on actual cancer cell lines, AI assists in Outcomes such as overall survival, disease-free survival,
decoding the large amount of data generated from high-­ cancer recurrence, or cancer metastasis can be predicted
throughput cancer-on-a-chip studies [57]. AI models have after training ML models, sometimes more accurately than
also been trained to identify cancer patients likely to benefit when using more traditional methods, such as the tumor-­
from adjuvant chemotherapy based on features of the cancer node-­metastasis staging system defined by the American
cells’ nuclei and perinuclear regions on pathology slides [58]. Joint Committee on Cancer [16]. Examples of ML-based
In the era of precision medicine where therapies selected outcome prediction include the prediction of lung cancer
specifically for a single patient based upon tumor genetics, histopathology and metastases based on radiomic features
patient comorbidities, and many other clinical factors are from CT imaging [64], cervical nodal metastases in patients
envisaged, AI is being looked to in order to identify drugs with squamous cell carcinoma of the head and neck using
associated with decreased rates of toxicity and increased effi- RF models trained on texture analysis of multi-energy CT
cacy for any given patient in a matter of hours rather than images [65], invasiveness of pulmonary adenocarcinomas
months. Specifically, ANNs have been employed to identify on CT using 3D DL [66], metastasis of rectal cancer to the
genetic variants and mutations from genomic data, to per- liver based on the primary tumor’s MRI radiomic features
form virtual screening of drug candidates based on ligand-­ [67], and radioresistance of primary nasopharyngeal can-
receptor binding or on overall structure, and to increase the cers prior to IMRT using radiomic features from CT, MR,
efficiency of molecular dynamics simulations, for example, and PET [68].
by making it possible to model thousands of atoms simulta- The features found by ML to be most predictive of out-
neously, model solvation, classify chemical trajectory data, comes are often not intuitive and are independent of classical
model molecular properties based on excited state electrons, cancer predictors. For example, multimodality DL, combin-
rapidly screen for novel materials and catalysts, estimate ing clinical, pathological, and imaging information, increased
thermochemical properties of active molecules, and model the predictive power of clinical outcomes in glioblastoma
free energy changes during chemical reactions [59]. multiforme, where survival is poor and ranges from 1 to
782 B. Franc

2 years in most patients [69]. In another study, tissue micro- cancer patients preoperatively by training on conventional
array samples are used to classify breast cancer patients as ultrasound, shear wave elastography of the breast, and clini-
low-risk versus high-risk yielded groups with significant sur- copathologic features [74]. Finally, Rice and colleagues have
vival differences independent of multiple grades, tumor used ML (RF) to identify esophageal cancer features predic-
sizes, hormone receptor status, and the presence/absence of tive of nodal disease extent with the aim of separating patients
lymph nodes [70]. Pan-cancer prognostication may also be who require limited versus extensive lymphadenectomy [75].
possible using clinical data, mRNA expression data,
microRNA expression data, and whole-slide histopathology
images [71]. 74.6 Conclusion and Future Directions

AI is already an integral aspect of many facets of research on

74.5.4 Understanding and Predicting cancer and its metastasis, including tasks of classification,
Lymphatic-Based Metastasis regression, and clustering (Table 74.1). However, when
employed using large-scale computing facilities, AI tech-
The direct application of ML to identify clinical and patho- niques such as DL are on the cusp of transforming cancer
logic features associated with metastasis through the lym- research. Industry and academia are actively developing
phatic system and to predict nodal involvement in cancer has powerful DL frameworks that target a range of tasks, and the
been explored across many cancer types. Nodal involvement US Department of Energy (DOE) Exascale Computing
in gastric cancer patients, one of the key prognostic indicators Project’s Cancer Distributed Learning Environment
in that population, has been predicted using an ANN-­based (CANDLE) is building software infrastructure to support
algorithm using input features of depth of infiltration, size, scalable deep learning on DOE supercomputing resources.
location, and shape (polypoid, fungating, ulcerating, or infil- The CANDLE initiative has begun to address problems such
trating) [72]. With the aim of identifying breast cancer as understanding protein interactions in the RAS/RAF path-
patients who are at low risk of metastasis at presentation and way, predict response of cancer to drug candidates in a
who may not require a sentinel node biopsy as part of their patient-specific manner, and analyze millions of medical
surgery, Dihge and colleagues have recently developed records from cancer patients with varying lifestyles, environ-
machine learning–based prediction models of axillary nodal mental exposures, cancer types, and healthcare providers to
metastasis with predictors from RNA sequencing and/or clin- identify optimal cancer treatment strategies [76]. AI applied
icopathologic features [73]. Zheng and colleagues have at these scales promises to completely transform our under-
recently used DL to predict axillary nodal positivity in breast standing of cancer and its metastasis.

Table 74.1 Overview of ML tools for supervised and unsupervised learning tasks and examples of their application in research on cancer and its
metastases
Supervised learning Unsupervised learning
General task Classification Regression Clustering
Techniques Support vector machines (SVM) Linear regression K-means
Discriminant analysis Generalized linear model (GLM) K-medoids
Naïve Bayes Ground-penetrating radar (GPR) Fuzzy C means
Nearest neighbor Support vector regression (SVR) Hierarchical
Random forest Ensemble methods Gaussian mixture
Neural networksa Decision trees Hidden Markov model
Random forest Neural networksa
Neural networksa
Example applications Cell classification Elucidating molecular pathways of Development of tumor profiles:
in cancer research Tumor/metastasis detection and cancer genesis, growth, and genomics, transcriptomics,
segmentation on imaging dedifferentiation metabolomics
Biomarker signatures predicting lymph Identification of driver mutations Discovery of cancer subtypes
node metastases Models for prediction of cancer Prediction of second primary
Prediction of tumor-associated antigens survival, recurrence, and progression cancers
to target for immunotherapy
Screening therapeutic candidates for
tumor target binding and/or predicted
efficacy
a
Includes artificial neural networks for deep learning, such as convoluted neural networks (CNNs), deep belief networks (DBN), and recurrent
neural networks (RNN)
74 Application of Artificial Intelligence in Research on Cancer and Its Metastasis 783

12. Schwyzer M, Martini K, Benz DC, et al. Artificial intelligence

Questions Addressing the Future of the Field of Artificial for detecting small FDG-positive lung nodules in digital PET/CT:
impact of image reconstructions on diagnostic performance. Eur
Intelligence in Cancer Research Radiol. 2020;30:2031–40.
• Using observational data extracted from the medi- 13. Schwyzer M, Ferraro DA, Muehlematter UJ, et al. Automated
cal records and images of a large number of can- detection of lung cancer at ultralow dose PET/CT by deep neural
cer patients, can artificial intelligence techniques networks - initial results. Lung Cancer. 2018;126:170–3.
14. Sikpa D, Fouquet JP, Lebel R, Diamandis P, Richer M, Lepage
identify heretofore unknown cancer phenotypes M. Automated detection and quantification of breast cancer brain
that could benefit from alterations in therapeutic metastases in an animal model using democratized machine learn-
approach? ing tools. Sci Rep. 2019;9:17333.
• Will outcome models of cancer and metastases gen- 15. Bi WL, Hosny A, Schabath MB, et al. Artificial intelligence in
cancer imaging: clinical challenges and applications. CA Cancer J
erated by artificial intelligence techniques lead to Clin. 2019;69:127–57.
changes in our approach to cancer treatment? 16. Niu PH, Zhao LL, Wu HL, Zhao DB, Chen YT. Artificial intelli-
• Can artificial intelligence techniques improve the gence in gastric cancer: application and future perspectives. World
equity and future applicability of randomized clini- J Gastroenterol. 2020;26:5408–19.
17. Pokkalla H, Pethia K, Glass B, et al. Machine learning models
cal trials? accurately interpret liver histology in patients with nonalcoholic
• Will artificial intelligence techniques elucidate steatohepatitis (NASH). Hepatology. 2019;70:187.
mechanisms of interaction between the immune 18. Ehteshami Bejnordi B, Veta M, Johannes van Diest P, et al.
system and cancer that can lead to new approaches Diagnostic assessment of deep learning algorithms for detection
of lymph node metastases in women with breast cancer. JAMA.
to immune-based therapeutics? 2017;318:2199–210.
• How can artificial intelligence integrate clinical 19. Niazi MKK, Parwani AV, Gurcan MN. Digital pathology and artifi-
outcomes data, molecular structural data, pathol- cial intelligence. Lancet Oncol. 2019;20:e253–e61.
ogy, and medical imaging to drive drug 20. Litjens G, Sanchez CI, Timofeeva N, et al. Deep learning as a tool
for increased accuracy and efficiency of histopathological diagno-
discovery? sis. Sci Rep. 2016;6:26286.
21. Pham HHN, Futakuchi M, Bychkov A, Furukawa T, Kuroda K,
Fukuoka J. Detection of lung cancer lymph node metastases from
whole-slide histopathologic images using a two-step deep learning
approach. Am J Pathol. 2019;189:2428–39.
22. Robertson S, Azizpour H, Smith K, Hartman J. Digital image anal-
References ysis in breast pathology-from image processing techniques to artifi-
cial intelligence. Transl Res. 2018;194:19–35.
1. La Porta CAM, Zapperi S. Explaining the dynamics of tumor 23. Arvaniti E, Fricker KS, Moret M, et al. Automated Gleason grading
aggressiveness: at the crossroads between biology, artificial intel- of prostate cancer tissue microarrays via deep learning. Sci Rep.
ligence and complex systems. Semin Cancer Biol. 2018;53:42–7. 2018;8:12054.
2. Langs G, Rohrich S, Hofmanninger J, et al. Machine learning: from 24. Landau MS, Pantanowitz L. Artificial intelligence in cytopathol-
radiomics to discovery and routine. Radiologe. 2018;58:1–6. ogy: a review of the literature and overview of commercial land-
3. Jiang F, Jiang Y, Zhi H, et al. Artificial intelligence in healthcare: scape. J Am Soc Cytopathol. 2019;8:230–41.
past, present and future. Stroke Vasc Neurol. 2017;2:230–43. 25. Oei RW, Hou G, Liu F, et al. Convolutional neural network for cell
4. Esteva A, Kuprel B, Novoa RA, et al. Dermatologist-level clas- classification using microscope images of intracellular actin net-
sification of skin cancer with deep neural networks. Nature. works. PLoS One. 2019;14:e0213626.
2017;542:115–8. 26. Zheng C, Xu R. Predicting cancer origins with a DNA methylation-­
5. Gulshan V, Peng L, Coram M, et al. Development and validation of based deep neural network model. PLoS One. 2020;15:e0226461.
a deep learning algorithm for detection of diabetic retinopathy in 27. Azarkhalili B, Saberi A, Chitsaz H, Sharifi-Zarchi A. DeePathology:
retinal fundus photographs. JAMA. 2016;316:2402–10. deep multi-task learning for inferring molecular pathology from
6. Do CB, Katoh K. Protein multiple sequence alignment. Methods cancer transcriptome. Sci Rep. 2019;9:16526.
Mol Biol. 2008;484:379–413. 28. Toratani M, Konno M, Asai A, et al. A convolutional neural net-
7. Liu Q, Sung AH, Chen Z, et al. Gene selection and classification for work uses microscopic images to differentiate between mouse
cancer microarray data based on machine learning and similarity and human cell lines and their radioresistant clones. Cancer Res.
measures. BMC Genomics. 2011;12(Suppl 5):S1. 2018;78:6703–7.
8. Huang SY, Franc BL, Harnish RJ, et al. Exploration of PET and 29. Chen CL, Mahjoubfar A, Tai LC, et al. Deep learning in label-free
MRI radiomic features for decoding breast cancer phenotypes and cell classification. Sci Rep. 2016;6:21471.
prognosis. NPJ Breast Cancer. 2018;4:24. 30. Hasan MR, Hassan N, Khan R, Kim YT, Iqbal SM. Classification of
9. Chidester B, Do MN, Ma J. Discriminative bag-of-cells for cancer cells using computational analysis of dynamic morphology.
imaging-­genomics. Pac Symp Biocomput. 2018;23:319–30. Comput Methods Prog Biomed. 2018;156:105–12.
10. Hughes KS, Zhou J, Bao Y, Singh P, Wang J, Yin K. Natural lan- 31. Al-Kofahi Y, Zaltsman A, Graves R, Marshall W, Rusu M. A deep
guage processing to facilitate breast cancer research and manage- learning-based algorithm for 2-D cell segmentation in microscopy
ment. Breast J. 2020;26:92–9. images. BMC Bioinform. 2018;19:365.
11. Savova GK, Tseytlin E, Finan S, et al. DeepPhe: a natural language 32. Harmon SA, Tuncer S, Sanford T, Choyke PL, Turkbey B. Artificial
processing system for extracting cancer phenotypes from clinical intelligence at the intersection of pathology and radiology in pros-
records. Cancer Res. 2017;77:e115–e8. tate cancer. Diagn Interv Radiol. 2019;25:183–8.
784 B. Franc

33. Mosayebi A, Mojaradi B, Bonyadi Naeini A, Khodadad Hosseini 53. Bhattacharya T, Brettin T, Doroshow JH, et al. AI meets exascale
SH. Modeling and comparing data mining algorithms for prediction computing: advancing cancer research with large-scale high perfor-
of recurrence of breast cancer. PLoS One. 2020;15:e0237658. mance computing. Front Oncol. 2019;9:984.
34. Woo M. An AI boost for clinical trials. Nature. 2019;573:S100–S2. 54. Kadioglu O, Efferth T. A machine learning-based prediction plat-
35. Schneider P, Walters WP, Plowright AT, et al. Rethinking drug form for P-glycoprotein modulators and its validation by molec-
design in the artificial intelligence era. Nat Rev Drug Discov. ular docking. Cells. 2019;8(10):1286. https://doi.org/10.3390/
2020;19:353–64. cells8101286.
36. Sharif MS, Abbod M, Amira A, Zaidi H. Artificial neural etwork-­ 55. Tomar N, De RK. Immunoinformatics: a brief review. Methods Mol
based system for PET volume segmentation. Int J Biomed Imaging. Biol. 2014;1184:23–55.
2010;2010. https://doi.org/10.1155/2010/105610. 56. Menden MP, Iorio F, Garnett M, et al. Machine learning prediction
37. Belhassen S, Zaidi H. A novel fuzzy C-means algorithm for unsu- of cancer cell sensitivity to drugs based on genomic and chemical
pervised heterogeneous tumor quantification in PET. Med Phys. properties. PLoS One. 2013;8:e61318.
2010;37:1309–24. 57. Fetah KL, DiPardo BJ, Kongadzem EM, et al. Cancer modeling-­
38. Blanc-Durand P, Van Der Gucht A, Schaefer N, Itti E, Prior on-­a-chip with future artificial intelligence integration. Small.
JO. Automatic lesion detection and segmentation of 18F-FET PET 2019;15:e1901985.
in gliomas: a full 3D U-Net convolutional neural network study. 58. Bera K, Schalper KA, Rimm DL, Velcheti V, Madabhushi
PLoS One. 2018;13:e0195798. A. Artificial intelligence in digital pathology - new tools for diagno-
39. Nobashi T, Zacharias C, Ellis JK, et al. Performance compari- sis and precision oncology. Nat Rev Clin Oncol. 2019;16:703–15.
son of individual and ensemble CNN models for the classifica- 59. Nagarajan N, Yapp EKY, Le NQK, Kamaraj B, Al-Subaie AM, Yeh
tion of brain 18F-FDG-PET scans. J Digit Imaging. 2020;33: HY. Application of computational biology and artificial intelligence
447–55. technologies in cancer precision drug discovery. Biomed Res Int.
40. Savova GK, Danciu I, Alamudun F, et al. Use of natural language 2019;2019:8427042.
processing to extract clinical cancer phenotypes from electronic 60. Putcha G. Blood-based detection of early-stage colorectal can-
medical records. Cancer Res. 2019;79:5463–70. cer using multiomics and machine learning. Proceedings of the
41. Nair R, Rost B. Protein subcellular localization prediction American Society of Clinical Oncology Gastrointestinal Cancers
using artificial intelligence technology. Methods Mol Biol. Symposium; 2020.
2008;484:435–63. 61. Mamoshina P, Kochetov K, Cortese F, et al. Blood biochemistry
42. Malta TM, Sokolov A, Gentles AJ, et al. Machine learning identi- analysis to detect smoking status and quantify accelerated aging in
fies stemness features associated with oncogenic dedifferentiation. smokers. Sci Rep. 2019;9:142.
Cell. 2018;173(338–54):e15. 62. Ahmad MA, Eckert C, McKelvey G, Zolfagar K, Zahid A, Teredesai
43. Zhang J, Liu J, Sun J, Chen C, Foltz G, Lin B. Identifying driver A. Death vs. data science: predicting end of life. Proceedings of
mutations from sequencing data of heterogeneous tumors in the Association for the Advancement of Artificial Intelligence
the era of personalized genome sequencing. Brief Bioinform. Conference on Artificial Intelligence; 2018.
2014;15:244–55. 63. Jin X, Zheng X, Chen D, et al. Prediction of response after chemo-
44. Wang K, Duan X, Gao F, Wang W, Liu L, Wang X. Dissecting radiation for esophageal cancer using a combination of dosimetry
cancer heterogeneity based on dimension reduction of tran- and CT radiomics. Eur Radiol. 2019;29:6080–8.
scriptomic profiles using extreme learning machines. PLoS One. 64. Ferreira Junior JR, Koenigkam-Santos M, Cipriano FEG, Fabro
2018;13:e0203824. AT, Azevedo-Marques PM. Radiomics-based features for pattern
45. Wang S, Rong R, Yang DM, et al. Computational staining of pathol- recognition of lung cancer histopathology and metastases. Comput
ogy images to study the tumor microenvironment in lung cancer. Methods Prog Biomed. 2018;159:23–30.
Cancer Res. 2020;80:2056–66. 65. Forghani R, Chatterjee A, Reinhold C, et al. Head and neck squa-
46. Van den Eynde M, Mlecnik B, Bindea G, et al. The link between mous cell carcinoma: prediction of cervical lymph node metasta-
the multiverse of immune microenvironments in metasta- sis by dual-energy CT texture analysis with machine learning. Eur
ses and the survival of colorectal cancer patients. Cancer Cell. Radiol. 2019;29:6172–81.
2018;34(1012–26):e3. 66. Zhao W, Yang J, Sun Y, et al. 3D deep learning from CT scans pre-
47. Carnielli CM, Macedo CCS, De Rossi T, et al. Combining discov- dicts tumor invasiveness of subcentimeter pulmonary adenocarci-
ery and targeted proteomics reveals a prognostic signature in oral nomas. Cancer Res. 2018;78:6881–9.
cancer. Nat Commun. 2018;9:3598. 67. Shu Z, Fang S, Ding Z, et al. MRI-based radiomics nomogram to
48. Zhavoronkov A, Ivanenkov YA, Aliper A, et al. Deep learning detect primary rectal cancer with synchronous liver metastases. Sci
enables rapid identification of potent DDR1 kinase inhibitors. Nat Rep. 2019;9:3374.
Biotechnol. 2019;37:1038–40. 68. Li S, Wang K, Hou Z, et al. Use of radiomics combined with
49. Mamoshina P, Volosnikova M, Ozerov IV, et al. Machine learn- machine learning method in the recurrence patterns after intensity-­
ing on human muscle transcriptomic data for biomarker discov- modulated radiotherapy for nasopharyngeal carcinoma: a prelimi-
ery and tissue-specific drug target identification. Front Genet. nary study. Front Oncol. 2018;8:648.
2018;9:242. 69. Peeken JC, Goldberg T, Pyka T, et al. Combining multimodal imag-
50. Aliper A, Plis S, Artemov A, Ulloa A, Mamoshina P, Zhavoronkov ing and treatment features improves machine learning-based prog-
A. Deep learning applications for predicting pharmacological prop- nostic assessment in patients with glioblastoma multiforme. Cancer
erties of drugs and drug repurposing using transcriptomic data. Mol Med. 2019;8:128–36.
Pharm. 2016;13:2524–30. 70. Turkki R, Byckhov D, Lundin M, et al. Breast cancer outcome pre-
51. Fleming N. How artificial intelligence is changing drug discovery. diction with tumour tissue images and machine learning. Breast
Nature. 2018;557:S55–S7. Cancer Res Treat. 2019;177:41–52.
52. Vamathevan J, Clark D, Czodrowski P, et al. Applications of 71. Cheerla A, Gevaert O. Deep learning with multimodal repre-
machine learning in drug discovery and development. Nat Rev sentation for pancancer prognosis prediction. Bioinformatics.
Drug Discov. 2019;18:463–77. 2019;35:i446–i54.
74 Application of Artificial Intelligence in Research on Cancer and Its Metastasis 785

72. Bollschweiler EH, Monig SP, Hensler K, Baldus SE, Maruyama 74. Zheng X, Yao Z, Huang Y, et al. Deep learning radiomics can pre-
K, Holscher AH. Artificial neural network for prediction of lymph dict axillary lymph node status in early-stage breast cancer. Nat
node metastases in gastric cancer: a phase II diagnostic study. Ann Commun. 2020;11:1236.
Surg Oncol. 2004;11:506–11. 75. Rice TW, Ishwaran H, Hofstetter WL, et al. Esophageal can-
73. Dihge L, Vallon-Christersson J, Hegardt C, et al. Prediction cer: associations with (pN+) lymph node metastases. Ann Surg.
of lymph node metastasis in breast cancer by gene expres- 2017;265:122–9.
sion and clinicopathological models: development and vali- 76. Wozniak JM, Jain R, Balaprakash P, et al. CANDLE/Supervisor:
dation within a population-based cohort. Clin Cancer Res. a workflow framework for machine learning applied to cancer
2019;25:6368–81. research. BMC Bioinform. 2018;19:491.
 Glossary

Abscopal effect Radiotherapy treatment of a local site of linked to pericyte mimicry and extravascular migratory
disease leading to regression of metastatic tumors not metastasis.
irradiated. Antibody drug conjugates (ADC) A class of highly potent
Acoustic impedance A physical property of tissue which biological drugs built by attaching a small-­molecule anti-
describes how much resistance an ultrasound beam cancer drug or another therapeutic agent to an antibody,
encounters as it passes through the tissue. It is dependent with either a permanent or a labile linker. The antibody
on the density of the tissue (d, in kg/m3) and the speed targets a specific antigen only found on target cells.
of the sound wave (c, in m/s), with acoustic impedance Anti-resorptives Therapeutics that target bone resorption
(Z) = d x c. in order to preserve skeletal health and improve patient
Acral metastases Metastatic disease burden that travels quality of life.
distal to the elbows or the knees, involving forearm/hand APC Adenomatous polyposis coli gene.
or leg/ankle/foot anatomy. Androgen receptor (AR) Receptor protein that binds
Androgen deprivation therapy (ADT) Therapy to reduce androgen and regulates gene expression in the nucleus.
the levels of androgen hormones with drugs or surgery to Artificial intelligence Development of computer systems
treat prostate cancer. capable of performing classification or other tasks that
AJCC staging Established in 1959, the American Joint would otherwise require human intelligence.
Committee on Cancer (AJCC) is an organization that for- ATM Serine/threonine kinase involved in DNA repair.
mulates cancer classification systems, including cancer Autosomal dominant A pattern of inheritance character-
staging (TNM staging). Through collaborations with 20 istic of some genetic diseases. “Autosomal” means that
member organizations that are involved in cancer epide- the gene in question is located on one of the numbered,
miology, patient care, cancer control, cancer registration, or nonsex, chromosomes. “Dominant” means that a sin-
professional education, research, or biostatistics, AJCC gle copy of the disease-associated mutation is enough to
facilitates an evidence-based process to support classifi- cause the disease.
cations that are biologically and prognostically relevant: Autosomal recessive A genetic condition that appears
https://cancerstaging.org/Pages/default.aspx. only in individuals who have received two copies of an
AKT Is a serine/threonine protein kinase also known as autosomal gene, one copy from each parent. The gene is
protein kinase B (PKB). on an autosome, a nonsex chromosome. The parents are
Allograft-prosthetic composite This is a segmental bone- carriers who have only one copy of the gene and do not
replacing hybrid implant that includes human allograft exhibit the trait because the gene is recessive to its normal
bone that is coupled with a metallic implant. counterpart gene.
Anaplastic lymphoma kinase (ALK) The ALK receptor Axillary lymph node dissection Procedure in which the
has function on cellular communication, CNS develop- lymph nodes in levels I and II and sometimes level III
ment, and neuromuscular junction synapse development. nodes of the axilla are removed.
ALK mutations were also identified in lymphomas and BCBM Breast cancer brain metastasis (BCBM) is the type
lung cancer. of brain metastasis occurring in patients diagnosed with
Androgen Male sex hormone (e.g., testosterone). breast cancer.
Angiotropism In cancer literature, angiotropism defines Binding affinity It is widely used as one of the param-
tumor cells closely apposed to the external surfaces of the eters for quantitative evaluation of action and indicates
endothelial cells of blood vessels in a pericyte-like loca- the strength of interaction between a biomolecule and its
tion without evidence of intravasation. Angiotropism is ligand/binding partner.

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 787
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System,
https://doi.org/10.1007/978-3-030-93084-4
788 Glossary

Blood-tumor barrier The disrupted semipermeable Checkpoint inhibitors Drugs that are manufactured as
blood-brain barrier in the setting of tumor development. antibodies against CTLA4, or PD1 or PDL1 cell recep-
Also referred to as the blood-brain-tumor barrier. tors, which block checkpoint proteins expressed by cer-
Bone metastases The spread of cancer cells to bone from tain immune system cells, such as T cells, and also cancer
a different site. cells. These checkpoints help keep immune responses
Bone sarcoma Mesenchymal tumors of the skeleton. from being too strong and sometimes can keep T cells
Bone scan Indirect imaging for bone metastases using from killing cancer cells. When these checkpoints are
bone-seeking radiopharmaceuticals such as 99mTc-­MDP blocked, T cells can attack cancer cells more effectively.
and 18F-NaF PET/CT. Examples of checkpoint proteins found on T cells or can-
Brachytherapy Utilization of internalized radiotherapy cer cells include PD-1/PD-L1 and CTLA-4/B7-1/B7-2.
seeds deposited at the site of surgical resection to provide CDK 4/6 Cyclin-dependent kinase 4 and cyclin-­dependent
a constant, low dose of radiation directly to the tumor bed kinase 6 are proteins that regulate cell cycle. Many drugs
to improve local control after surgery. known as CDK 4/6 inhibitors can block those proteins,
BRAF A member of the Raf kinase family that is down- affecting multiple pathways, and halt cell cycle in G1
stream of RAS and activates its substrate, the MEK pro- phase.
tein, in the MAP kinase pathway within tumor cells. Some Cell-free DNA Circulating DNA, typically as nucleosome-­
mutations in BRAF gene can cause birth defects and associated fragments or exosomal, secreted, or released
other mutations, and cancers like lymphomas, colorec- by dying cells into plasma and other body fluids.
tal cancer, melanoma, lung cancer, and brain tumors. Cell-free tumor DNA It refers to free DNA that has been
Approximately, 50% of melanomas harbor a mutation spilled from the tumor and circulates in the bloodstream.
in exon 15, codon 600, which constitutively activates the Cell signaling It is the mechanism of information trans-
MAP kinase pathway. fer that governs cellular functions and activities, self, and
Brain metastasis Brain metastases are growths that spread cell-to-cell communication. Some components include,
to the brain from a cancer occurring in another part of but are not limited to, growth factors, receptors, transfer
the body. They are distinct from primary brain tumors of signals through a cascade of enzymatic activities, and
which initiate in the brain. Brain metastases are much proteins from an activated state to an inactivated state,
more common than primary brain tumors: an estimated controlled through an array of positive and negative sig-
200,000 people are diagnosed with brain metastasis each naling loops.
year (US alone), compared to about 17,000 diagnoses for Cement augmentation Filling of the vertebral body with
primary brain tumors. Furthermore, the number of brain cement, such as polymethylmethacrylate, through a per-
metastasis diagnoses has dramatically risen in recent cutaneous technique aimed at alleviating pain from patho-
years worldwide. Many oncologists believe that this is logic compression fractures or lytic lesions.
due to better early detection of brain metastasis, as well Cement osteoplasty Percutaneous injection of acrylic
as better treatments for primary cancers. As more patients cement (polymethylmethacrylate or PMMA) in a verte-
live longer with a primary cancer, the disease has more bral body with the goal of treating pain and preventing
time to spread to the brain. This has led to the renewed further collapse in the setting of a vertebral body fracture.
and important focus on treating brain metastasis. Chemical biliary sclerosis Inflammation and scarring of
BRCA1 Breast cancer gene 1, which is involved in DNA the biliary system secondary to chemical insults to the
repair, involved in hereditary breast, ovarian, and other liver including chemotherapy.
malignancies. Chimeric antigen receptor T cell Genetically engineered
BRCA2 Breast cancer gene 2, which is involved in DNA T cells to produce an artificial T-cell receptor for use in
repair, involved in hereditary breast, ovarian, and other immunotherapy.
malignancies. Circulating tumor cells (CTCs) These are viable and
Breslow The term associated with the measured thickness intact tumor cells that have shed into the vasculature
of a melanoma. It is distance between the epidermal gran- after detaching from primary and/or metastatic tumors,
ular layer and point at which the lowermost melanoma and circulate in blood. CTCs can extravasate and become
cell is seen. “seeds” for the subsequent growth of additional tumors in
Cancer-associated fibroblasts (CAFs) These are cells distant organs (metastases), a mechanism that is respon-
with an elongated morphology and are negative for epi- sible for the vast majority of cancer-related deaths.
thelial, endothelial, and leukocyte markers and lack the Circulating tumor DNA (ctDNA) DNA fragments
mutations found within cancer cells. released by tumor cells but not associated with cancer
Cell survival curve Cell survival curves are used to cells and detectable in the bloodstream of cancer patients.
describe the relationship between radiation dose and pro- Testing allows for personalized surveillance based on a
portion of cells that survive. cancer patient’s unique set of gene mutations.
Glossary 789

CTL4 Cytotoxic T-lymphocyte-associated protein or multisystem of distinct cell types. Differentiated lym-
CD152. This receptor member of the immunoglobulin phatic endothelial cells (LECs) retain their identity even
superfamily is expressed in T cells and participates in the if removed from the body and maintained in culture.
regulation of T cells, inhibiting their function. Drugs have Diffusion-weighted imaging A method of signal contrast
been manufactured as antagonist antibodies (anti-CTL4) generation based on the differences in Brownian motion.
to target this receptor, thus blocking its activity and stimu- Diagnostic leukapheresis (DLA) A short form of leuka-
lating T-cell response. pheresis that collects white blood cells from peripheral
Codman triangle An aggressive periosteal reaction due blood via density centrifugation and co-collects CTCs.
to rapid growth of a mass that pushes the periosteum off DNA damage repair DNA repair pathways that are active
faster than new periosteum can lay down new bone. throughout different stages of the cell cycle, allowing the
Color Doppler ultrasound A method of visually detecting cells to repair DNA damage.
and depicting motion, usually blood flow, using a color Doppler shift (Doppler effect) The change in frequency
map that is incorporated into a standard B-mode image. of a sound wave due to a reflector moving towards or
Doppler shifts of returning ultrasound waves within a away from an object, which in ultrasound is the trans-
defined region of interest are color-coded based on the ducer. Named after the Austrian physicist, Christian
average velocity and direction. Andreas Doppler (1803–1853).
Commensal Nonpathogenic, normally occurring micro- Downstaging The process by which a patient’s risk of
bial flora. developing recurrent disease is deemed to be lower than
Complete decongestive therapy (CDT) Gold standard for initially thought, due to either additional information
conservative lymphedema management used to decrease becoming available or a review of preexisting informa-
the limb volume (in phase I) and to maintain the obtained tion resulting in a change of parameters.
effect (in phase II). The components of CDT are as fol- Dysplasia A condition in which the arrangement and mor-
lows: skin care, manual lymphatic drainage, compression phology of cells in a tissue are disrupted. It is considered
therapy, and exercise. a preliminary stage of cancer.
Contrast resolution The ability to distinguish between Early metastatic A limited number of metastases that may
differences in image intensity. The contrast resolution of be indolent (true oligometastasis), or may precede wide-
MRI is superior to CT. spread dissemination.
Convolutional neural network A class of deep neural Edema Excess fluid collection in body parts and cavities
networks, the connectivity between which resembles the (e.g., ascites in abdominal cavity) resulting from high
organization of the visual cortex. or low flow imbalance between lymph formation and
CTNNB1 Catenin beta-1. absorption.
Cystoprostatectomy Surgical removal of the prostate and EGFR Epidermal growth factor receptor. It is a surface
urinary bladder. protein that stimulates cells to divide when epidermal
D2 lymphadenectomy Surgical removal of the N2 nodal growth factor and epidermal growth factor-like peptides
basin; nodes of the right and left gastric cardia, greater bind to it. EGFR mutations and amplifications are found in
and lesser curvature, supra- and infrapyloric areas, left multiple cancers such as lung cancers and glioblastomas.
gastric, common hepatic, celiac, and splenic arteries. Left Endocrine therapy Treatment that blocks the effect of or
paracardial nodes omitted for distal gastrectomy. inhibits the production of estrogen and/or progesterone.
DNA damage response (DDR) Denotes a network of cel- Endoprosthesis A modular metallic implant—composed
lular pathways that sense, signal, and repair DNA lesions. usually of cobalt, chromium, and molybdenum—that
Dedifferentiation Cells grow from a terminally differenti- may be assembled on-demand in the operating room to
ated cell to a less differentiated, less specialized state. fill large long-bone defects of any size.
Deep learning Supervised learning that uses multiple lay- Endosteal niche Site within bone that is rich in cells of
ers of nonlinear modules that accept features as inputs, hematopoietic stem cell and mesenchymal stem cell ori-
progressively extracting higher level features, and ulti- gin, including osteoblasts and osteoclasts. The endosteal
mately predicting labeled outputs. niche participates in recruitment, dormancy, and meta-
Dermatolymphangioadenitis Acute inflammation of static outgrowth.
skin, subcutaneous tissue, lymphatic vessels, and lymph Epigenetics A field that studies the heritable changes in
nodes. the expression of genes that do not involve alterations in
Diapedesis Passage of cells across an intact endothelium. the DNA sequence.
Differentiation Simply the process of becoming different. Epithelial-mesenchymal transition (EMT) An EMT is a
In embryology a single-celled zygote differentiates into a biologic process that allows an epithelial cell to undergo
multicellular embryo that develops into a more complex multiple biochemical, morphologic, and functional
790 Glossary

changes that enable it to assume a mesenchymal cell the whole liver volume, with exclusion of tumors, cysts,
phenotype conferring invasiveness, dissemination, and and ablation cavities.
chemo/immunotherapy resistance. This change allows Gantry Ring-shaped structure that contains the X-ray
cancer cells to escape from the primary site and invade tube, collimators, and detectors in CT equipment. Patients
surrounding tissue. are positioned on a table that moves through the central
Epithelioid sarcoma Epithelioid mesenchymal tumors opening of the gantry during image acquisition.
that commonly arise in the extremities or truncal area. The Gastric cancer lymph node stations Numerical classifi-
pathognomonic feature is the loss of INI-1 (SMARCB1) cation of the regional lymph nodes for gastric cancer, as
protein expression. defined by the Japanese Gastric Cancer Association.
Exosome Exosomes are extracellular vesicles that are shed Germline mutations A gene change in a body’s reproduc-
by the parent cells into ECM with double-layered lipid tive cell (egg or sperm) that becomes incorporated into
membrane containing molecules of proteins, lipids, and the DNA of every cell in the body of the offspring.
nucleic acids. Glycolysis The pathway leading from intracellular glucose
External beam radiation therapy Form of radiation directly to pyruvate, resulting in the generation of two
treatment in which a patient lies flat on a treatment table moles of pyruvate, ATP and NADH, from one mole of
as an external treatment machine directs a beam of radia- glucose.
tion to an area of interest within the patient. Graded prognostic assessment A contemporary set of
Extracellular vesicle Membranous vesicles produced by cells prognostic classification systems to predict outcomes of
that contain cellular cytosol and cellular genetic material. patients with brain metastases.
Extravascular migratory metastasis Extravascular Hallmarks of cancer Common traits, characteristics, and
migratory metastasis or EVMM is a non-­hematogenous, capabilities that normal cells acquire to become autono-
non-intravascular mode of tumor metastasis via angiot- mous through stepwise gain or loss of functions of genes
ropism and pericyte mimicry. When angiotropic cancer conferring survival advantage and dissemination to dis-
cells are engaged in pericyte mimicry and EVMM, they tant sites.
migrate away from the advancing front of the tumor to Halsted model of cancer Based on his understanding of
form regional and distant metastases, along blood vessels breast cancer, Dr. William Stewart Halsted postulated that
as a scaffolding to guide dissemination. cancer is an orderly disease that spreads from the primary
Extrinsic regulators Factors external to the cell, including site, by direct extension through the lymphatics to the
the niche and systemic factors, that can influence stem lymph nodes, and then to distant metastatic sites.
cell behavior. Haptotactic gradient Solid-phase chemokine gradient
FAP Familial adenomatous polyposis. immobilized in extracellular matrix.
FDA United States Federal Drug Administration. This HCC Hepatocellular carcinoma.
agency regulates drug approval and oversees the safety Hepatic artery infusion (HAI) pump A surgically
and efficacy of chemotherapies, immunotherapies, and implanted pump into the gastroduodenal artery, which
target-based drugs. allows for the delivery of chemotherapy to liver tumors
Fisher systemic theory of cancer Dr. Bernard Fisher via the hepatic artery. As the hepatic artery accounts for
articulated in a lecture in 1980 his theory that cancer is a the majority of the blood supply of tumors, HAI chemo-
systemic disease, and that systemic dissemination of can- therapy allows for higher drug concentrations with less
cer cells has frequently already occurred at the time of systemic effects.
diagnosis of localized disease. Hepatic artery infusion pump therapy Cancer regional
Fluoroscopy Imaging technique that utilizes real-time therapy consisting of a surgically placed subcutaneous
X-ray imaging for both diagnostic and interventional pro- pump with vascular access to the common hepatic artery
cedures. Fluoroscopy systems usually contain an X-ray designed to deliver chemotherapy.
tube and an image intensifier or a flat panel detector cou- HIFU High-intensity focused ultrasound.
pled with a monitor. Histopathology Microscopic evaluation of tissue, ranging
FLAIR A T2-weighted sequence (on MRI) where the sig- in scale from the overall tissue architecture to the qualita-
nal of CSF is subtracted, to highlight brain edema. tive features of chromatin in cell nuclei.
Future liver remnant (FLR) The volume of functional HPB Hepatopancreaticobiliary.
liver parenchyma remaining after resection, which is HER2 Receptor tyrosine-protein kinase erbB-2. It is also
associated with both the risk of postoperative liver failure known as HER2/neu. Human epidermal growth factor
and morbidity and mortality. Using standard radiographic receptor 2 (HER2) is a receptor of the EGFR family. Its
software, the proposed line of resection is drawn, and the mutation has been seen in breast, ovarian, uterine, gastro-
remaining liver volume is calculated as a percentage of intestinal, and other cancers.
Glossary 791

HU Hounsfield units are named after Sir Godfrey tumor invasive margin, with some studies showing a cor-
Hounsfield, the Nobel Prize winner and inventor of relation between increasing scores and improved patient
CT. HU is a dimensionless relative quantitative measuring survival.
unit of radio density unit used in CT scanning to express Immunotherapy A type of treatment that activates or sup-
CT numbers in a standardized and convenient form. presses a person’s own immune system with substances
Human papillomavirus (HPV) A double-stranded DNA that stimulate the immune response to fight diseases such
virus that can infect epithelium. There are hundreds of as cancer.
types of HPVs, some of which can cause cancer. International Metastatic Renal Cell Carcinoma Database
Iliopectineal line An anatomic and radiographic line that Consortium (IMDC) International Metastatic Renal
approximates the pelvic brim; is used to evaluate the Cell Carcinoma Database Consortium (IMDC) criteria
integrity of the anterior column of the acetabulum based are used to assess the prognosis of patients with meta-
upon its continuity. static RCC utilizing simple clinical and laboratory crite-
Ilioischial line A radiographic line that is used to evaluate ria: (1) KPS (<80%), (2) <1 year from time of diagnosis,
the integrity of the posterior column of the acetabulum (3) hemoglobin < LLN, (4) corrected calcium > ULN, (5)
based upon its integrity. neutrophils > ULN, and (6) platelets > ULN. Presence of
Image-guided radiation therapy (IGRT) The process 0 of these factors defines favorable risk group, 1–2 fac-
of using precise image guidance during CT simulation, tors intermediate risk group, and ≥3 factors poor risk
treatment planning, and treatment delivery. group.
Image segmentation Technique of dividing an image In-transit melanoma Melanoma involvement of the der-
into parts, often for the purpose of image compression or mis, subcutaneous tissue, and/or lymphatics that occurs
object recognition. between the site of the primary tumor and regional lymph
IMRT Intensity-modulated radiation therapy. Technique node basin.
of external beam radiation in which metal leaves within Intrathecal Introduced into or occurring in the space
the treatment machine variably modulate the intensity of under the arachnoid membrane of the brain or spinal cord.
the radiation beam to allow better conformality of dose to Intratumor heterogeneity A phenomenon in which mul-
a treatment volume. tiple subclones with different genomes exist in a single
Immediate reconstruction The placement of either a tumor.
prosthetic or an autologous tissue for breast reconstruc- Intrinsic regulators Stem cell autonomous factors,
tion at the time of oncologic surgery. This can be done including genetic factors, that are intrinsic to the cell and
as the first stage of a multistage reconstruction, or as a not influenced by extrinsic factors.
definitive single-stage reconstruction. Ionizing radiation Radiation, traveling as a particle or
Immune checkpoints Key regulators of the immune electromagnetic wave, that carries sufficient energy to
system that when stimulated can dampen the immune detach electrons from atoms or molecules, thereby ion-
response to an immunological stimulus. izing an atom or a molecule.
Immune checkpoint inhibitors A class of drugs that Karnofsky performance status (KPS) A scale from 0 to
blocks the proteins that regulate and keeps the immune 100 that assesses a patient’s functional ability, including
system in check. for self-care.
Immune-mediated therapy A type of targeted therapy KIT It is a receptor tyrosine kinase protein found on the
which optimizes the host immune system to recognize cell membrane of different cell types. Mutations in KIT
and eradicate invading tumor cells. gene, especially those occurring in exons 9 and 11, drive
Immunocytochemistry Immunocytochemistry (ICC) cell growth and survival in a number of cancer cells,
(also known as immunohistochemistry—IHC) is a tech- including melanoma and gastrointestinal stromal tumor.
nique for the detection and visualization of proteins, or KRAS/RAS It is a gene that encodes K-RAS, a protein
other antigens, in cells using antibodies specifically rec- part of the RAS/MAPK pathway. KRAS mutations and
ognizing the target of interest. amplifications are found in colorectal and lung cancers.
Immunogenicity The ability to elicit an immune response Laser interstitial thermal therapy (LITT) Minimally
in the human body to an antigen that enters the body. invasive approach to brain tumor surgery featuring ste-
Immunomodulation Defined for this chapter as three reotactic placement of a laser fiber and thermal ablation
major immune cells, i.e., TAMs, MDSCs, and regulatory of a brain lesion via twist-drill burr hole.
T (Treg) cells, in TME that help regulate host immune Lauren classification General classification of gastric
response and are “friendly” to tumor. adenocarcinoma into two histologic subtypes, intestinal
Immunoscore A score assigned based on the quantifi- and diffuse type, according to the tendency for the forma-
cation of T-cell infiltration within the tumor center and tion of recognizable gland structures.
792 Glossary

Leiomyosarcoma Soft-tissue sarcoma transformed involved nodes, but nothing is palpable; therapeutic
from mesenchymal cells with smooth muscle features. (TLND), where lymph nodes have been palpated clini-
Commonly occurs in extremities and truncal areas, retro- cally or detected on imaging; and completion (CLND),
peritoneum, and uterus. whereby, following the removal of an involved sentinel
Leptomeningeal disease A late complication of meta- node, the remaining nodes from that lymph node basin
static cancer in which tumor cells spread via the CSF to are removed, on the suspicion that other nodes may be
the arachnoid and pia of the brain and spinal cord. harboring micrometastatic disease.
Limited metastases A limited number of metastatic Lymphadenopathy An enlarged lymph node.
lesions. Lymphangion Lymphatic functional contractile unit
Liposarcoma The most common subtype of soft-­tissue sar- between two valves.
coma with varying degrees of adipocytic differentiation. Lymphangiogenesis Growth of new lymphatic vessels by
Based on the histology and molecular differences, liposar- budding and sprouting from existing vessels.
coma could be classified as well-­differentiated/dediffer- Lymphatic-venous bypass Known also as lymphati-
entiated liposarcoma (with MDM2 gene amplification), covenular anastomosis—surgical connection of lymph
myxoid round-cell liposarcoma (with FUS-DDIT3 gene vessel with vein to restore the flow of lymph back to the
rearrangement), and pleomorphic liposarcoma (com- venous system.
monly with TP53 mutation). Different types of liposar- Lymphedema Abnormal swelling that generally occurs in
coma have different clinical behavior, prognosis, and the arms or legs.
tumor immune microenvironment compositions. Lymphovascular invasion The identification of tumor
Liquid biopsy The compendium of tests performed on a cells within lymphatic vessels.
sample of blood to look for cancer cells that are circulat- Machine learning A subfield of AI, which focuses on
ing in the blood (CTCs), or for fragments of DNA from developing algorithms that learn to perform tasks or make
tumor cells that are in the blood (ctDNA), or any biologi- predictions from data.
cal fluid containing tumoral material. A liquid biopsy may Malignancy of unknown primary A tumor presenting as
be used to help find cancer at an early stage. It may also be a metastasis without a known site of origin.
used to help plan treatment or to understand out how well Mammosphere-forming cells In culture, many cell types,
treatment is working and/or if cancer has relapsed. Being benign and malignant, spontaneously grow in spherical
able to take multiple samples of blood over time may formations, a stem cell or CSC being at the origin. The
also help oncologists understand what type of molecular stem cell or sphere-forming cell in mammary cells is
changes are taking place in a tumor. known as the mammosphere-forming cell.
Local consolidative therapy Focal treatment (usually sur- MAP therapy Methotrexate, doxorubicin, cisplatin.
gery or radiation) to the primary and/or metastatic lesions. Margin The outside edge of a surgical specimen. A posi-
Local recurrence Recurrence of any cancer in the primary tive margin is one that is involved with cancer cells.
site after prior surgical resection. In breast cancer this Margin width The distance between the tumor cells and
refers to a regrowth in the ipsilateral breast. Also referred the edge of the surgical specimen.
to as in-breast tumor recurrence, or IBTR. Maturation Process by which the immature lymphatic
Locus (plural loci) The specific physical location of a plexus becomes hierarchically organized into lymphatic
gene or other DNA sequence on a chromosome, like a capillaries that collect lymph and collecting lymphatic
genetic street address. vessels that transport the fluid.
Loss of heterozygosity If there is one normal and one Metastatic castration-resistant prostate cancer
abnormal allele at a particular locus, as might be seen (mCRPC) Metastatic prostate cancer with rising serum
in an inherited autosomal dominant cancer susceptibil- levels of PSA or progression of disease despite androgen
ity disorder, loss of the normal allele produces a locus deprivation therapy (ADT).
with no normal function. When the loss of heterozygosity MD Anderson Bone Response Criteria A set of rules that
involves the normal allele, it creates a cell that is more are designed to standardize the interpretation of the imag-
likely to show malignant growth if the altered gene is a ing of bone metastases following therapy.
tumor-suppressor gene. Also called LOH. Medical linear accelerator A particle accelerator that
Lung cancer Malignancy that occurs in the parenchyma accelerates charged particles to megavoltage energies
of the lung. using a series of oscillating electric fields along a lin-
Lymphadenectomy A surgical procedure which involves ear beamline adopted which is adapted for medical
the complete removal of the lymph nodes from a defined use.
surgical area. This could be elective (ELND) where the MEK An essential kinase protein involved in the MAP
lymph nodes are removed where there is suspicion of kinase signaling pathway. MEK is typically activated by
Glossary 793

the upstream RAF kinases, and in turn activates the down- Multicentric disease Disease in more than one quadrant
stream ERK protein, resulting in increased cell growth. of the breast.
Mesenchymal tissue Tissue originating from the meso- Multikinase inhibitors Drugs that work by inhibiting
derm, which can develop into multiple tissue types multiple intracellular and cell surface enzymes called
including bone, cartilage, muscle, fat cells, and connec- kinases, some of which are implicated in tumor growth
tive tissue. and metastatic progression of cancer, thus decreasing
MET Process by which motile cells acquire cell junctions tumor growth and replication. Examples include suni-
to associate with each other and form stable clusters. tinib, sorafenib, cabozantinib, and lenvatinib.
MR-Linac A radiation therapy delivery machine that Natural language processing A subset of machine learn-
combines real-time 3D magnetic resonance imaging with ing techniques that help computers analyze, interpret, and
radiation treatment delivery. manipulate human language.
Metastasectomy Surgical resection of a metastatic focus Necrotic or apoptotic cells Accidental cell death or pro-
of a malignant tumor. grammed cell death.
Metastasis The migration and spread of cancer cells from Neoadjuvant therapy Treatment given as a first step to
a primary site to regional lymph nodes and/or distant shrink a tumor before the main treatment, which is usu-
organs in the body, resulting in the development of sec- ally surgery, is given.
ondary tumors and a major contributor to the deaths of Neoantigen A cancer antigen that has newly emerged due
cancer patients. to genetic mutations that occur in cancer cells.
Metastatic organotropism Predilection of tumors to Next-generation sequencing Next-generation sequenc-
metastasize to specific rather than random organs. ing (NGS) is a massively parallel sequencing technology
Microbiome The collection of microbes that exist in, on, that offers ultrahigh throughput, scalability, and speed.
and throughout the human body. The technology is used to determine the order of nucle-
Micrometastatic The presence of submillimeter meta- otides in entire genomes or targeted regions of DNA or
static disease. RNA.
Microsatellite instability A phenomenon in which a NOMS A decision-making framework which takes into
sequence shows a different number of repetitions in tumor account the neurologic, oncologic, mechanical, and sys-
tissues than in normal tissues due to an impaired ability temic status of a patient to guide surgical and/or radiation
to repair base sequence errors that occur during DNA treatment of metastasis disease of the spine.
replication. Non-small cell lung cancer The most common type of
Minimally invasive parafascicular surgery lung cancer.
(MIPS) Utilization of tubular cylindrical or ovoid NSAIDs Nonsteroidal anti-inflammatory drugs.
retractors, also known as port-based or channel-based Oligometastatic (at presentation) An indolent early
systems, to allow for brain cannulation using a sulcus or metastatic state with a limited number of (often up to 5)
small corticotomy to provide access to deep-seated sub- metastatic lesions.
cortical brain tumors with minimal disruption of the adja- Oligometastatic disease Distant disease that is limited in
cent cortex. number of metastases, usually between 1 and 5 different
Minimally invasive solutions Interventional techniques lesions.
for treating metastatic skeletal or soft-tissue disease that Oligometastatic theory of cancer The theory that inter-
shortens recovery time and improves both functionality mediate steps exist between initial tumor cell dispersion
and pain control. Such techniques are generally delivered and fulminant metastatic disease, and that patients who
through percutaneous on noninvasive routes, and can develop a small number of confined metastatic lesions
include cryoablation, radiofrequency ablation, cemento- may experience improved survival through local manage-
plasty, or MRI-guided high-­frequency ultrasound. ment of these metastases.
MGMT 0(6)-methylguanine-DNA-methyltransferase is a Oligoprogressive A nonmetastatic case that quickly
repair protein. Methylation of MGMT has been shown to progresses to an oligometastatic state after definitive
affect the efficacy of alkylating agents in brain tumors, therapy.
and the survival of patients with lung cancer. Oligorecurrent A nonmetastatic case that slowly pro-
mTOR Mammalian target of rapamycin is a protein gresses to an oligometastatic state after a longer (usually
kinase that regulates multiple cell functions including cell at least 6 months) period of progression-free survival.
growth and proliferation. Its activity is dysfunctional in Oligoresidual A state of oligometastatic disease formed
multiple cancers including breast, lung, melanoma, and after a more widely metastatic state is cytoreduced with
kidney cancers. systemic therapy.
794 Glossary

Ommaya Reservoir Intraventricular catheter system that Pericyte mimicry See “angiotropism.”
can be used for the aspiration of cerebrospinal fluid or for Permeative process Characterized by multiple endosteal
the delivery of drugs into the cerebrospinal fluid. lucencies with cortical sparing (“moth-eaten”).
Oncovirus A virus that can cause cancer. Pharmacodynamic endpoint An outcome measure to
Organoid Three-dimensional cell culture to grow similar capture the potential clinical benefit of a therapeutic
to human tissues, organs, or tumor. intervention.
Organotropism The predilection of circulating tumor PI3K Phosphoinositide 3-kinase.
cells to target specific secondary sites, often involving a PI3K/AKT pathway Signal transduction pathway, which
bidirectional interaction where the primary tumor influ- promotes cellular processes such as proliferation, cell sur-
ences changes at a distant site to create an environment vival, and growth.
conducive to metastatic proliferation. Polyclonal seeding Multiple cancer cells from different
Oropharyngeal cancer A type of head and neck cancer, subclones of a tumor tissue (primary or metastatic) dis-
occurring in the oropharynx (base of the tongue, tonsils, seminate to secondary sites and proliferate in parallel.
posterior pharyngeal wall, and soft palate). Polymetastatic A metastatic state with a diffuse number
Osteoarticular allograft This is a portion of a bone that (usually >5–10) of metastatic lesions.
is harvested from a human cadaver and includes the carti- Pre-analytical phase It is the phase between sampling
lage of the adjacent joint surface. a biological specimen (such as a tissue biopsy or a vein
Osteotropism The predilection of disseminated tumor puncture) in a patient and processing the sample in the
cells to metastasize to the skeleton compared to other laboratory before using a method analysis.
organs. Precision medicine Personalized medicine refers to the
Parenchyma The functional tissue of an organ as distin- practice of providing optimal treatment based on indi-
guished from the connective and supporting tissue. When vidual patient data.
used in the context of a sentinel node a tumor deposit is Pre-metastatic niche The microenvironment of a second-
within the central anatomical area of the node. ary organ site is modified by tumor-secreted factors with
Poly (ADP-ribose) polymerase (PARP) A family of pro- systemic effect, making that organ more conducive to the
teins mainly involved in DNA repair, but also other cel- proliferation of metastatic tumor cells at that site.
lular processes such as programmed cell death. Prostate A gland of the male reproductive system that
Pathologic fracture predictive models An assortment of secretes seminal fluid.
models developed through retrospective or prospective PSA Prostate-specific antigen is a protein produced by
collection of pathologic fracture events, image-­based pre- normal and cancer cells of the prostate gland.
dictive algorithms, or nomogram predictive models that PSMA Prostate-specific membrane antigen. A pro-
allows a cancer practitioner to predict the risk for fracture. tein expressed in all forms of prostate tissue, including
Newer models also help a practitioner balance the overall carcinoma.
prognosis with fracture risk in order to better guide treat- PTEN Phosphatase and tensin homolog. A protein that
ment solutions. acts as a tumor suppressor.
PD1/PDL1 Programmed cell death protein 1/programmed Quality of life In the context of metastatic skeletal disease,
cell death 1 ligand. The protein regulates apoptosis of T measures of treatment that enhance the body’s functional
cells in lymph nodes and regulatory T cells. The ligand condition, such as maintaining mobility and improving pain.
is highly expressed in multiple cancers and is thought to Radical mastectomy Surgical removal of the breast
be one of the mechanisms behind cancer evasion from together with the overlying skin, underlying pectoralis
the immune system. Anti-PD1/PDL1 are drugs that try to muscle, and axillary lymph nodes.
stimulate the immune system, in particular CD8 T cells, Radical prostatectomy Surgical removal of the prostate
to target cancer antigens. gland.
Pelvic lymph node dissection Surgical removal of the Radiobiology Radiobiology is a field of clinical and basic
pelvic lymph nodes, which may include regions such as medical sciences that involves the study of the action
obturator, external iliac, internal iliac, common iliac, pre- of ionizing radiation on living things, especially health
sacral, and periaortic. effects of radiation.
Percutaneous ablation Percutaneous chemical or thermal Radiomics A field focused on identifying imaging pheno-
mediated tissue destruction. types based on mathematical features within images.
Percutaneous cementoplasty Procedure by which bone is Radiosensitive A lesion which is likely to respond to
consolidated by percutaneous injection of cement. radiotherapy treatment.
Percutaneous osteosynthesis Percutaneous fixation of a RECIST criteria Response Evaluation Criteria in Solid
fracture or area in which there is an impending fracture. Tumors.
Glossary 795

Recursive partitioning analysis (RPA) A prognostic Somatic mutations Mutations in individual body cells,
classification for patients with brain metastasis consist- but not in the germ cells. The mutations only have effects
ing of three groups based on patients’ performance status, on the body cells concerned and are therefore not trans-
extracranial disease, and age. mitted to progeny.
Regional recurrence Recurrence of breast cancer in the Spatial resolution The ability to distinguish between
regional lymph node basin of the index cancer. objects or structures that differ in density. A high spatial
Reverse total shoulder arthroplasty An alternative resolution is important to discriminate between structures
design to conventional shoulder arthroplasty implants that that are located within a proximity to each other. The spa-
inverts the normal arthroplasty design, employing a con- tial resolution of CT is superior to MRI.
vex glenoid component and concave humeral component. Specification The labile phase of lymphatic endothe-
Sarcoma A malignant tumor of the cells of mesenchymal lial cell (LEC) development during which PROX1 is
origin. expressed in a subset of blood vessel endothelial cells
SBRT Stereotactic body radiation therapy. Form of exter- (BECs). These LEC progenitors could revert back to
nal beam radiation therapy in which relatively high doses BECs in the absence of signals such as VEGF-C.
of radiation are accurately delivered to a treatment vol- Spectrum theory of cancer First described in 1994 by Dr.
ume over 1–5 daily fractions. Samuel Hellman, who theorized that cancer biology is a
Sciatic buttress Caudal part of the iliac bone which con- heterogenous spectrum, with the growth and progression
nects the acetabulum to the sacrum/spine and is respon- of different tumors being a function of that specific can-
sible for transmitting force from the hip joint to the axial cer’s biology.
skeleton. SPECT Single-photon emission computed tomography.
Secretomes Denotes the collection of proteins released by Sporadic lymphatic anomaly Non-inherited diseases
a cell into the extracellular space as defined by TME. caused by errors in the development of the lymphatic
Separation surgery A strategy of open surgical decom- system.
pression that clears tumor away from the spinal cord Stage How extensive the amount of cancer is within the
and thecal sac, typically through a posterior approach, to body, which often is defined clinically and/or pathologi-
provide 2–3 mm margin of clearance, with the intent of cally by the size of the primary tumor in the organ of ori-
allowing adjunctive radiation therapy to be safely applied gin, the presence of regional lymph node metastasis, and
to the tumor without the risk of radiation toxicity to the the documentation of systemic metastases.
neural structures. Staged colorectal cancer metastasectomy For patients
Sentinel lymph node The first node or nodes drain- with synchronous colorectal cancer and hepatic metasta-
ing the region of a primary tumor, most likely to harbor ses, the primary tumor and hepatic metastases are resected
metastasis. in separate operations. This may be done by resection
Sentinel lymph node biopsy Procedure to identify and of the primary tumor first, followed by resection of the
remove the first node in which cancer cells are likely to hepatic metastases at a later date, or vice versa via a liver-
spread. first approach.
Sentinel lymph node dissection Surgical procedure in Stage migration Improvements in stage-specific survival
which a tumor drain is removed from the first lymph node for a type of cancer due to reclassification of staging
and histopathologically evaluated to check if cancer has groups and more accurate staging, but without demon-
spread beyond a primary site of tumor. stration of improvements in survival from the cancer as
Shannon index A statistical diversity index. a whole.
Shear stress Frictional force exerted on endothelial cells Stereotactic body radiotherapy A noninvasive treatment
due to the movement of fluids. Shear stress regulates vari- modality in which high-dose targeted radiation is deliv-
ous aspects of lymphatic vascular development. ered to a tumor with narrow margins in a hypofraction-
Siewert classification Classification of gastroesophageal ation schedule while minimizing radiation to adjacent
junction (GEJ) adenocarcinomas according to the epi- normal tissue.
center of the tumor. Type I: 1–5 cm above GEJ, type II: Stereotactic body radiation therapy (SBRT)/stereotactic
from 1 cm above to 2 cm below GEJ, and type III: 2–5 cm ablative radiotherapy (SABR) Highly focused and
below GEJ. precise radiation treatment delivered using stereotactic
Small cell lung cancer Highly malignant and rapidly targeting and image guidance, typically delivered over
dividing type of lung cancer characterized by early 1–5 treatments, with an ablative/near-ablative dose to
metastasis. tumor, while limiting dose to adjacent organs at risk.
Soft-tissue sarcoma (STS) Mesenchymal tumors devel- Stereotactic radiosurgery (SRS) A technique that deliv-
oping in the soft tissues. ers radiation in a highly conformal manner with high spa-
796 Glossary

tial precision typically delivered over a single dose, but Tolerogenic dendritic cells Heterogenous pool of den-
can also be delivered in a hypofractionated schedule of dritic cell with immunosuppressive properties, priming
3–5 doses. the immune system into tolerogenic state against different
STIR An example of a fat suppression technique with antigens through regulation of T cells such as inducing T
MRI classically used to differentiate fat from edema on cell anergy, T cell apoptosis, and induction of Tregs.
T2 sequences. TP53 A tumor-suppressor gene, which is the most fre-
Subcapsular Used in the context of a deposit of recurrent quently mutated gene in human cancer. Induces growth
disease located directly under the outer lining capsule of arrest or apoptosis in response to cellular stress.
a sentinel node. Transarterial embolization Occlusion of arterial blood
Supervised learning Machine learning techniques applied supply, typically that of a tumor.
in learning tasks when data is organized as input exam- Transendothelial migration Migration of cancer cells
ples (features) with corresponding outputs (labels). from circulation through endothelium of blood vessels
Synchronous colorectal cancer metastasectomy For in order to access distant organs and form metastatic
patients with synchronous colorectal cancer and hepatic colonies.
metastases, synchronous resection involves removal of Triple-negative breast cancer Biologic subtype of breast
the primary tumor and hepatic metastases in the same cancer which does not express estrogen receptor, proges-
operation. terone receptor, or HER2.
T1-weighted sequence A MRI sequence timed to show Tumor-associated macrophages Macrophages that are
fat, proteinaceous fluid, and contrast as hyperintense. commonly observed in the tumor microenvironment.
T2-weighted sequence A MRI sequence timed to show Most tumor-associated macrophages are M2 macrophages
water and fat as hyperintense. Commonly used with fat that are associated with tumor angiogenesis, matrix pro-
suppression techniques to highlight edema. tein degradation, and induction of an immunosuppressive
Targeted drug therapy A type of drug therapy that intends tumor microenvironment.
to inhibit abnormal molecules in cancer cells that activate Tumor biomarker A biological molecule found in body
the signaling transduction pathway to drive cell growth fluids or tissues indicating the possibility of a cancer
and survival. somewhere in the body.
Targeted molecular therapy A form of molecular medi- Tumor immune microenvironment The compositions
cine which functions to impede the growth and progres- of the immune cells and their associated chemokines or
sion of cancer cells through inhibition of specific targeted cytokines in the tumor microenvironment.
molecules necessary for cellular pathways driving tumor Tumor microenvironment (TME) Denotes the compart-
growth. ment that provides a space rich in lympho-­vasculature,
Tertiary lymphoid structure Tertiary lymphoid organs host cells, stroma, extracellular matrix and metabolic
that contain B cell and T cell zones and, when matured, collaborator of proliferating tumor cells, and a signaling
contain germinal centers. Antigen-presenting cells in ter- network for cellular and molecular cross talks, which is
tiary lymphoid structure are responsible for instructing T pivotal for tumor establishment, evasion of host immune
and B cells towards tumor-associated antigens. surveillance, growth, invasion, and metastasis.
Therapeutic ratio The ratio of the probability of treatment Tumor mutational burden (TMB) Tumor mutational
success (e.g., tumor control probability) to the probability burden, defined as the total number of somatic mutations
of normal tissue damage. per coding area of a tumor genome, is an emerging clini-
Thermal ablation A minimally invasive approach to the cal biomarker associated with the response to immune
treatment of tumors that involves the use of high-­intensity checkpoint inhibitor (ICI) therapy. TMB has been shown
hyperthermic procedures to apply energy to tissue and to vary markedly among tumor types as well as among
cause necrosis. Commonly used thermal ablation tech- patients within tumor types.
niques include radiofrequency ablation, microwave abla- Two-stage hepatectomy A treatment strategy for bilobar
tion, and high-frequency ultrasound ablation. disease with curative intent in which tumors of one hemi-
Thoracic duct Principal central lymphatic trunk collect- liver are resected during the first stage, followed by a
ing lymph from most of the body and emptying into the period of convalescence that allows for liver hypertrophy
central venous system in the neck. to occur, and subsequent resection of the tumors of the
TKI Tyrosine kinase inhibitors are medications that inhibit contralateral hemi-liver.
enzymes called tyrosine kinases that are involved in mul- Unsupervised learning Machine learning techniques that
tiple intracellular mechanisms. find patterns in data without the need for data annotation
TM-ILP Isolated limb perfusion with TNF-alpha and and produce classes within a dataset through structural
melphalan. learning.
Glossary 797

Variant This refers to the individual differences that are VEGFR Vascular endothelial growth factor receptors are
recognized when the structure of the DNA of each person tyrosine kinases that bind the protein vascular endothelial
is examined in detail. growth factor (VEGF) to stimulate the formation of new
Vascular co-option Cells (tumor, cancer, circulating blood vessels (angiogenesis and vasculogenesis).
tumor cells) interacting physically, cell-to-cell, with pre- Warburg effect Enhanced conversion of glucose into lac-
existing vessels in the absence of neoangiogenesis. May tate in cancer cells even in the presence of oxygen, a con-
precede neoangiogenesis. dition also called aerobic glycolysis and now recognized
Vascular niche Sites within bone that are rich in vas- to be a common phenotype of many cancer cells.
culature and endothelial cells. In the context of bone WGS Whole-genome sequencing by means of next-­
metastasis, the vascular niche can regulate dormancy and generation sequencing (NGS).
metastatic outgrowth via provision of oxygen, nutrients, Whole-brain radiation therapy (WBRT) A technique
and growth factors. that delivers radiation to the entire brain in a uniform
Vasogenic edema Edema secondary to venous obstruc- fashion for the treatment of brain metastases.
tion, in turn secondary to mass effect. To be distinguished Wide resection Resection of the tumor, pseudocapsule,
from cytotoxic edema, more commonly seen following and a surrounding cuff of normal tissue.
infarct in strokes.
Index

A B
Abdominoperineal resection (APR), 381 Bacillus Calmette–Guerin (BCG), 731
Ablation, 323, 533, 534, 536 Barcelona Clinic Liver Cancer (BCLC) staging system, 437
Abscopal effect, 667 Benzopyrones, 266
Accelerated partial breast irradiation (APBI), 676 Beta-catenin, 620, 624
Accreditation, 152 Bevacizumab, 474, 563, 599
Accuracy, 328 Biliary sclerosis, 471
Acquisition, 743 Billroth, T., 381
Acrometastases, 499 Binimetinib, 742
Active surveillance, 621–623, 625 Biological effective dose (BED), 698
Adaptive planning, 666 Biology-guided radiation therapy (BgRT), 666, 668
Adenoma-carcinoma, 437 Biomarkers, 172, 722–724
Adjuvant therapy, 433, 714 Biopsy, 323
Adrenal metastases, 687 Bisphosphonates, 490
Advanced/metastatic stage, 148 Bivalent vaccine, 762
Alkylating agents, 379 Bladder cancer, 587, 589–591, 731
Alliance A011202, 356 Bleomycin, 379
Alpha-HPVs, 758 Blood-brain barrier, 576, 579
Alveolar soft part sarcoma (ASPS), 629, 631 Blood-lymph separation, 201–202
American College of Surgeons Oncology Groups (ACOSOG), 354, Blood-tumor-barrier, 542
355, 399 Blue dye, 354, 357
American Joint Committee on Cancer (AJCC) staging system, 602 Bone metastasis, 295, 296, 298, 300, 301, 304, 307, 484
5-Aminolevulinic acid (5-ALA), 555 Bone sarcomas, 122
Anchoring filaments, 202 Brachytherapy, 556, 673
Aneuploidy, 16 BRAF, 740
Angiogenesis, 6, 21–26, 542, 544 Brain metastasis (BM), 157, 541–543, 545, 548
Angiosarcoma, 122, 607, 632–633 Brain Metastasis Selected Marker (BMSM), 158
Angiotropism, 74–79, 81 Breast cancer, 43–44, 47–49, 171, 777, 782
Anoikis, 203 subtypes, 396
Anthracyclines, 379 Breast cancer BM (BCBM), 161
Antibiotics, 770 Breast conservation therapy (BCT), 354, 355
Antibody drug conjugates, 733 Breast-conserving surgery, 397
Antimetabolite therapy, 379
Anti-resorptives, 490
Anti-tumor immunity, 112 C
Apolipoprotein B mRNA editing enzyme, cata-lytic polypeptide-like Cachexia, 488
(APOBEC), 17 Canadian Cancer Trials Group (CCTG), 700
Apparent diffusion coefficient (ADC), 300, 301, 303–305 Cancer, 275–280
Arthroplasty, 494 lymphangiogenesis, 209–212, 214, 216–223, 225
Articulation, 497 metastasis, 209–212, 214, 216–223, 225, 249–252, 254–257
Artificial neural networks (ANN), 776 staging, 369–371
Aseptic chemical meningitis, 580 therapeutic vaccine, 734
Assessment of response, 625 Cancer of unknown primary origin (CUP), 58
Associating liver partition and portal vein ligation for staged Cancer-survivor, 269
hepatectomy (ALPPS), 440 Carcinogenesis, 170
Asymmetric divisions, 751 CD44+/CD24− (stem-like) CTCs, 161
Atezolizumab, 723 CDH1, 422, 427
Avelumab, 715 CDK4/6 inhibitor, 561
Axillary lymph node dissection (ALND), 398 Cell cycling, 15
Axillary recurrence, 354–356 Cell-free DNA, 67–70, 415
Axillary staging, 353 CellSearchTM platform, 148, 158

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 799
S. P. Leong et al. (eds.), Cancer Metastasis Through the Lymphovascular System, https://doi.org/10.1007/978-3-030-93084-4
800 Index

Cellular therapy, 724 Cytoreductive surgery/hyperthermic intraperitoneal chemotherapy

Cement augmentation, 527–529, 531 (CRS/HIPEC), 413, 414
Cemented, 519, 520 Cytosine-phosphate-guanine (CpG), 18
Cementoplasty, 518, 533, 535, 536 Cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4), 380, 711,
Cerebrospinal fluid (CSF) analysis, 578 720, 730
Cervical carcinoma, 381
Cervical intraepithelial neoplasia (CIN), 763
CheckMate-214, 730 D
CheckMate 274, 733 D2 lymphadenectomy, 426
Checkpoint inhibitor, 561–563, 730, 732, 734, 771 D2 resection, 425–426
Chemokine, 109, 111, 112 Dabrafenib, 742
Chemotherapy, 98–101, 560–563, 579–581, 621, 623, 624 Dacarbazine, 742
Chimeric antigen receptors (CAR), 725 Decompressing the mass effect, 569
Cholangiocarcinoma, 432 Dedifferentiated (DDLPS) liposarcoma, 612, 613
Chondrosarcoma, 121, 605 Deep inspiration breath hold (DIBH), 665
Chromoplexy, 16 Deep learning (DL), 776, 777
Chromosomal alterations, 16 De-escalation, 760, 763
Chromosomal aneuploidy, 15–16, 18, 19 Definitions, 683–684
Chromosomal instability, 15 Dendritic cell (DC), 232–234
Chromosomal instability pathway, 439 Denosumab, 490
Chromothripsis, 16 DEPArray platform, 161
Chylous disorders, 269 Deregulating Cellular Energetics, 9–10
Circulating cancer cells (CTC), 31 Dermatologic Cooperative Oncology Group (DeCOG), 339, 345
Circulating cell-free DNA (cfDNA), 171 Desmoid tumors (DT), 619, 623, 624
Circulating nucleic acids (cNAs), 171 Dewar classification, 347
Circulating tumor cells (CTCs), 111, 121–126, 139–143, 158, 171, Diaphyseal femur, 508
175–179, 181, 578 Diaphyseal lesions, 510
Circulating tumor DNA (ctDNA), 125–126, 129–135, 139–143, 158, Diaspora theory, 407
162, 171, 578 Dielectrophoretic array, 122
Circulation, 31–32 Diffuse gastritis/duodenitis, 473
Classification of the oligometastatic state, 683 Diffuse-type, 422
Clear cell sarcoma, 596, 629, 632, 633 Diffusion-weighted imaging (DWI), 295, 300, 302, 304, 306, 307
Clinical risk score, 439 Disease-specific survival (DSS), 413
Clonogen, 682, 683, 686 Disseminated tumor cells (DTCs), 97–99, 101
Coaxial, 328 Distal cholangiocarcinoma, 433
Cobimetinib, 743 Distal femur, 508
Cognitive decline, 568 DMNTs, 18
Collagen, 89, 90, 94 DNA methylation, 17–18
Collecting lymphatic vessels, 202 DNA microarray, 382
Colon cancer, 369–373 Doppler, 286, 288
Colonization, 32–34 Dormancy, 108, 109, 111, 487
Colorectal cancer (CRC), 45, 48, 129, 134, 171, 381 Dormant, 97–101
Colorectal liver metastasis (CRLM), 411, 439, 469 Dormant cancer cells, 33
Columns, 516 Double stranded breaks (DSBs), 655
Combination therapy, 382 Downregulation of cell proliferation inhibitor, 7
Complete lymph node dissection (CLND), 380 Droplet digital PCR (ddPCR), 133, 134
Complications, 328, 520 Dual mobility, 520
Compression therapy, 265 Dual-tracer, 355, 357
Computed tomography (CT), 310–317, 319, 327, 596, 777, 781 Durvalumab, 723
Cone beam computed tomography (CBCT), 664, 665 Dutch D1D2, 381
Congruency, 516 Dye, 354
Constrained liners, 520 Dysplasia-carcinoma, 437
Contralateral axillary lymph nodes, 400
Contrast resolution, 283, 285, 289, 291
Conventional external beam radiation therapy (cEBRT), 526, 530 E
Convoluted neural networks (CNN), 776, 777 Early lymphoid aggregates (E-TLS), 642
CpG island methylation phenotype (CIMP) pathway, 18, 439 Early stage, 148
Craniotomy, 549–552 Early surgical treatment, 378
Cryoablation, 331, 518, 530, 534 Edema, 193
CSF ctDNA, 163 Electronic portal imaging device (EPID), 663
CTNNB1, 620, 621, 625, 626 Eloquent locations, 549–551
Cytarabine, 580 Embolization, 518, 533, 535
Cytogenetics, 15 Embryonic-like migration, 81
Cytology, 327, 578, 580 Embryonic migratory pathways, 81
Index 801

En bloc resection, 380, 494, 555 Gene rearrangements, 150

Encorafenib, 742 Gene signatures, 486
Endocrine therapy, 396 General approach to treatment, 686–687
Endometrial cancer, 44–46 Genetic alterations, 15, 17
Endoprosthetic, 520 Genitourinary cancer, 381
Endosteal niche, 487 Genitourinary malignancies, 587, 729, 731, 734
Endothelial cells, 25–26 Genome-wide methylation, 172
Enostoses, 324 Genomic instability, 9
Enucleations, 436 Germline mutation, 42–45
EpCAM− CTCs, 158 Germline risk, 768
Epidrugs, 170 Glycocalyx, 233–238
Epigenetics, 169 Gorham-Stout disease (GSD), 243, 244
Epithelial cell adhesion molecule (EpCAM), 158 G protein-coupled receptor (GPCR), 201
Epithelial-mesenchymal transition (EMT), 22, 29, 100, 109, 110, 203, Gp78, 599
753 Graded Prognostic Assessment (GPA), 568
Epithelioid and clear cell sarcoma, 632
Epithelioid sarcoma, 596, 629, 632, 633
Esophageal adenocarcinomas, 16 H
Esophageal cancer, 130, 132, 133, 135, 361, 365, 366 Hallmarks of cancer, 3, 5, 9, 11, 12
Ewing sarcoma, 606 Halsted, W.S., 378, 380
Exosome, 24–25, 34, 487, 790 Halsted hypothesis of cancer, 681
Expiration breath hold (ExBH), 665 Halstedian model of cancer, 378
Extended lymphadenectomy, 410, 435 Harrington reconstruction, 519, 520
External beam, 517 Hartmann, H., 381
External beam radiation therapy (EBRT), 697 Harvested nodes, 408
Extracellular matrix (ECM), 21, 107–110 Hedgehog signaling, 752
Extracellular vesicles (EVs), 613 Helicobacter pylori, 422, 768
Extravasate, 205 Hemiarthroplasty, 497
Extravasation, 32–34, 111 Hemipelvectomy, 520
Extravascular migratory metastasis (EVMM), 74–75, 80–81 Hepatic artery infusion (HAI), 466, 476
chemotherapy, 466
pump therapy, 470, 475
F Hepatic metastases, 411–413, 687
Fecal transplantation, 771 Hepatisis B, 769
Femur, 507 Hepatisis C, 769
Fibroblast activation protein, 23 Hepatocellular carcinoma, 436
Fibronectin, 89–92 Hepatocyte growth factor (HGF), 24
Fibrosarcoma, 595, 606 Hereditary breast and ovarian cancer (HBOC) syndrome, 38–44
Fine needle aspiration (FNA), 328 Hereditary cancer syndromes, 38–47
Fisher, B., 378 Heterogeneity, 22, 23, 27
Flare, 298, 299, 307 Hibernomas, 324
FLASH radiotherapy (FLASH-RT), 662, 667, 668 High endothelial venules (HEVs), 205
Flow cytometry, 578 Hilar cholangiocarcinoma, 433
Floxuridine (FUDR), 470 Hippocampal avoidance, 568, 570
Fluorescent imaging, 255, 257 Histone, 169
Fluorescent in situ hybridization (FISH), 150 HIV, 760
Fluorodeoxyglucose positron emission tomography (FDG-PET), 596 Homing, 486
Fluoroscopy, 327 Host defense, 10–11
Food and Drug Administration (FDA), 148 Human epidermal growth factor receptor 2 (HER2) positive, 356
Four-dimensional computed tomography (4DCT), 664 Human papillomaviruses (HPV), 757, 762, 769
Fraction, 517 vaccines, 762, 763
Fusobacterium, 769 Hyaluronan, 231–238
Future liver remnant (FLR), 411, 412, 464–466 Hyperdiploid karyotype, 16
Hypermethylation, 170
Hyperthermic intraperitoneal chemo therapy (HIPEC), 414
G Hypofractionated, 674
Gallbladder carcinoma, 431 Hypomethylation, 170
GammaPod, 691 Hypoxia, 31
Gastrectomy, 381 Hypoxia inducible factor 1 α (HIF1α), 599
Gastric cancer, 132, 133, 361–366, 777, 782 Hysterectomy, 381
Gastric cancer lymph node stations, 423
Gastroduodenal artery (GDA), 471
Gastroesophageal junction (GEJ) adenocarcinomas, 423, 425 I
Gastrointestinal stroma tumor (GIST), 595, 631 Ilioischial, 516
Gastrointestinal ulcers, 473 Iliopectineal, 516
Gene expression profiles (GEPs), 59, 382 Image guidance, 671
802 Index

Image-guided radiation therapy (IGRT), 665 Kidney cancer, 729–731, 734

Imaging, 275–280 KIT, 740, 744
Imatinib, 744 KN-426, 731
Imatinib mesylate, 379 Knowledge-based planning, 663, 667
Immune cells, 26–27
Immune checkpoint, 112, 779
Immune checkpoint inhibitors (ICIs), 151, 710 L
Immune cytokines, 730 Label-free cell assay, 778
Immune mediated therapies, 380 Laminar shear stress (LSS), 202
Immune modulatory cells, 27 Laminins, 77, 79
Immune system, 409 Laser ablation, 332
Immunocytochemistry, 150 Laser interstitial thermal therapy (LITT), 530, 553
Immunohistochemistry, 64–65 Lauren classification, 422
Immunologic responses, 410 Leiomyosarcoma (LMS), 121, 595, 629–631, 640, 642–644
Immunoscore, 415 Leptomeningeal carcinomatosis, 575, 580, 581
Immunotherapeutic drugs, 729 Leptomeningeal contrast enhancement, 577
Immunotherapy, 560–564, 598 Leptomeningeal disease (LMD), 555
IMvigor010, 732, 733 Leptomeningeal nodules, 577
IMvigor211, 732 Linac, 662–665
Incidental GBC, 432 Linear accelerators (LINAC), 698
Incubator hypothesis, 380 Linitis plastica, 422
Inflammatory breast cancer, 398 Liposarcoma, 121, 595, 612, 629, 640, 642–644
Innervation, 490 Liquid biopsy, 171, 177, 578, 580
Instability, 520 Liver-directed regional therapy, 474
Integrin β1 (ITGβ1), 159 Liver metastases, 411
Intensity modulated proton therapy (IMPT), 666 Long ncRNAs (lncRNAs), 170
Intensity modulated radiation therapy (IMRT), 663 Low-dose computed tomography, 151
Interferon-alpha (IFN-α), 710 Low-grade fibromyxoid sarcoma (LGFMS), 631, 632
Internal mammary nodes (IMN), 357 Lumbar puncture, 578, 579
Internal target volume (ITV), 664, 665 Lumpectomy, 380
International Association for the Study of Lung Cancer (IASLC), 148 Lung cancer, 172, 447, 448, 450–453, 781
Interstitial fluid pressure, 201 Lymphadenectomy, 380, 381, 410
Intestinal-type, 422 Lymphadenopathy, 319
Intra-arterial therapies, 440 Lymphangiogenesis, 199, 598
Intrahepatic cholangiocarcioma (ICC), 432, 476 Lymphangions, 189
Intralesional curettage, 506, 509, 510 Lymphangitic carcinomatosis, 313, 314
Intramedullary osteosarcoma, 604 LYMPHA technique, 267, 268
Intraoperative, 676 Lymphatic capillaries, 202
Intra-operative MRI, 554–555 Lymphatic development, 241–244
Intraoperative radiation therapy, 530 Lymphatic drainage, 587
Intrathecal, 575, 579, 580 Lymphatic endothelium, 236, 238
Intravasation, 31–32, 110, 205 Lymphatic imaging, 249–252, 254–257
Invasion, 6, 29–31 Lymphatic malformations (LMs), 241, 243
Ionization, 672 Lymphatic muscle cells (LMCs), 203
Intraoperative radiotherapy (IORT), 676 Lymphatic plexus, 202
Ipilimumab, 380, 711 Lymphatic system, 209–211, 214, 217, 219, 220, 223–225
ISET, 148 Lymphatic valves (LVs), 203
ISO 15189 norm, 152 Lymphatic-venous by-pass, 261, 269
Isosulfan blue, 354 Lymphatic vessel density (LVD), 761
Lymphatic vessel endothelial hyaluronan receptor 1 (LYVE-1),
232–238
J Lymphedema, 199, 241–245, 354, 356
JAVELIN Bladder 100, 732 Lymphedema prevention, 266–269
Javelin 101, 731 Lymph node (LN), 209–212, 214–223, 225, 276–279, 408, 448,
Judet, 516 450–453
Jugular lymph sac, 201 count, 408, 409
Juxtacortical osteosarcoma, 604 dissection, 262, 264, 267, 268, 588, 589, 591
harvests, 409
ratio, 432
K recovery, 409
Kaposiform lymphangiomatosis (KLA), 243, 244 retrieval, 409
Karnofsky Performance Status, 568 sampling, 408
Keratinocytes, 758, 760 yield, 409
KEYNOTE-045, 732 Lymph node metastases (LNM), 436, 629, 631–633, 635
KEYNOTE-361, 732 Lymphography, 249–251, 257
Ki67, 161 Lymphoscintigraphy, 340, 341
Index 803

Lymphovascular, 168 Mirel’s scoring system, 295

Lymphovascular niche, 203 Mismatch repair genotype, 733
Lymphovenous valves (LVVs), 201 Mitogen-activated protein kinase (MAPK/MEK) pathway, 379, 740
Lymphscintigraphy, 250–252, 257 Modular, 520
Lynch syndrome, 38, 44, 45 Molecular genetics, 620–621
Molecular profiling, 683
Molecular subtyping of urothelial carcinoma, 733
M Molecular targeted therapies, 739
Machine learning (ML), 776 Molecular variants, 16
Macrometastasis, 57 MRI-linacs, 672
Macrophage, 489 MR imaging-guided extracorporeal focused ultrasound (MRgFUS),
Magnetic resonance (MR), 252, 257, 310–312 331
Magnetic resonance imaging (MRI), 310, 315–317, 319 MR-linac, 662, 665–667
Malignant peripheral nerve sheath tumors, 595 Mucosal infection, 758
Map lymphatic system of the prostate, 588 Multicenter Selective Lymphadenectomy Trial-1 (MSLT-1), 339, 345
Marker hypothesis, 380 Multicenter Selective Lymphadenectomy Trial-2 (MSLT-2), 339, 345
Markers of cancer stem cells, 749–750 Multidisciplinary, 517
Mastectomy, 355, 397 Multiparametric flow cytometry, 159
Matrix metalloproteinase (MMPs), 23 Multiple brain metastases, 549
Maturation, 202 Multiple endocrine neoplasia (MEN), 46–47
MD Anderson bone metastases response criteria (MDA criteria), 296, Multiple endocrine neoplasia 2A (MEN2A), 46
307 Multiple endocrine neoplasia 2B (MEN2B), 46
MDM2, 614 Multiple metastases, 551–552
Mechanistic target of rapamycin (mTOR) inhibitor, 562 Mural cell, 202
Medical linear accelerator (linac), 662, 663 Mutational signatures, 15–17
Melanoma, 74–77, 79–81, 339–349, 380, 385–392, 709, 740 Myeloid-derived suppressor cells (MDSCs), 489
Memantine, 568, 570 Myxofibrosarcoma, 595, 633
Meningeal biopsy, 578 Myxoid liposarcoma (MLS), 613, 631, 634–635
Merkel cell carcinoma (MCC), 715 Myxoid-round cell liposarcomas, 615
Mesenchymal stem cells, 488
Mesenchymal to epithelial transition (MET), 203
Metabolic heterogeneity, 175–177, 181 N
Metabolic reprogramming, 175–181 Nasopharyngeal cancers, 781
Metabolism, 97, 99–101, 266 Natural language processing (NLP), 777, 779
Metachronous lesions, 684 Necrosis, 324
Metastasectomy, 412, 414, 448, 454 Neoadjuvant chemotherapy (NAC), 355
Metastasis, 6, 37–49, 167, 175–179, 181, 276, 278, 450–454, 587–591 Neoadjuvant systemic therapy, 398
Metastasis susceptibility, 47–48 Neoadjuvant therapy, 382, 715
Metastatectomy, 475 Neoplastic meningitis, 575, 576, 579
Metastatic, 515–517, 519, 520 Neural crest, 76, 80, 81
cascade, 108–110, 113 Neurocognitive decline, 568
colonization, 33 Neurologically intact, 701
organotropism, 32 Neurologic, Oncologic, Mechanical, and Systemic (NOMS), 524–526,
Metastatic castration-resistant PC (mCRCP), 139, 141, 142 531
Metastatic epidural spinal cord compression (MESCC), 697, 698, 701, Neurotropism, 545
702 Next-generation sequencing (NGS), 67, 150, 163
Metastectomy, 686 9ER, 731
Methotrexate, 579, 580 9-valent vaccine, 762, 763
Methylation, 169 Nivolumab, 380, 713
5-Methylcytosine, 18 Nodal yields, 408
Methylene blue, 354 Noncoding RNAs (NcRNAs), 170
Microarray, 776 Non-small cell lung carcinoma (NSCLC), 369–373, 719
Microbiota, 711 Nonunion, 498
Microenvironment, 541, 542, 544, 684, 794 Notch signaling, 751
Microfluidics, 122 Notch1, 162
Micrometastasis, 33, 57, 111, 399 NSABP B-51, 356
Micrometastatic, 339, 341, 343, 348 Nuclei segmentation, 777
MicroRNA (miRNA), 18–19, 170, 685
Microsatellite instability (MSI) pathway, 439
Microwave ablation (MWA), 331, 466 O
Milan criteria, 438 Observation, 436
Miliary nodules, 314 Oligodendrogliomas, 16
Minimally invasive approaches, 552–554 Oligometastatic, 329, 684
Minimally invasive parafascicular surgery (MIPS), 553 Oligometastatic breast cancer, 691
Minimally invasive solutions, 506 Oligometastatic cancer, 681
Minimal residual disease (MRD), 122, 129, 132–135, 147 Oligometastatic cancers with controlled primary, 692
804 Index

Oligometastatic disease, 400 Positron-emission tomography with fused CT (PET/CT), 283,

Oligometastatic lung cancer, 687–691 290–293, 295, 298–300, 304
Oligometastatic prostate cancer, 691 Positron-emission tomography with fused MRI (PET/MRI), 283,
Oligoprogressive, 684 290–293, 295, 298
Oligorecurrence, 684 Precancerous, 763
Oligoresidual, 684 Precancerous lesion, 759, 760, 763
Ommaya reservoir, 580 Precision medicine, 410
Oncogenic mutations, 740 Predictive biomarker, 150
Oncoproteins, 759 Pre-metastatic niche (PMN), 33, 486, 614, 686
Oral squamous cell carcinoma, 780 Primary follicle-like TLS (PFS-TLS), 642
Organotropism, 544–545, 794 Procollagen-lysine, 2 oxoglutarate 5-dioxygenase (PLOD2), 599
Oscillatory shear stress (OSS), 202 Programmed cell death protein 1 (PD-1), 380, 711, 720
Osteoblast, 484 Prophylactic cranial radiation (PCI), 676
Osteoblastic lesions, 488 Proportion, 409
Osteoclast, 484 Prostate cancer, 44, 47–49, 139–142, 171, 587–590, 733, 734, 779
Osteocytes, 484 Prostatectomy, 381
Osteolytic lesions, 488 Prostate seed implant (PSI), 674
Osteomyelitis, 324 Prosthetic infection, 520
Osteosarcoma, 602 Proton therapy, 666
Osteotropism, 485 Proximal femur, 507
Ovarian cancer, 44, 48 Proximal tibia, 510
Oxidative stress, 203 Pulmonary adenocarcinomas, 781
Oxygen hemostasis, 31 Pump related complication, 472
Pylorus-preserving pancreaticoduodenectomy, 381

P
p16, 762 Q
p38/MAPK, 487 Quadrivalent, 763
p53, 759, 761, 770 Quadrivalent vaccine, 762
Pain flare, 702 Quality of life, 505, 512
Palliative, 378 Quality of Life after Treatment for Brain Metastases (QUARTZ) trial,
PanCK+ (epithelial), 161 571
Pancreatic ductal adenocarcinoma, 434 Quiescence, 202
Pancreatic neuroendocrine tumors, 435
Pap testing, 763
Paradoxical activation of the MAP, 744 R
Pathologic complete response, 400 Radiation, 517
Pathologic fracture predictive models, 507 Radiation therapy, 530, 579, 625
Pathologic margins, 397 Radical hysterectomy, 760
Pathologic response, 382 Radical lymphadenectomy, 597
Patterning, 202–203 Radical mastectomy, 378, 380
Pelvic brim, 516 Radical resection, 432
Pelvic exenteration, 381 Radiobiology for the Radiologist, 653
Pelvic lymphadenectomy, 760 Radiocolloid, 354
Pelvic lymph node metastasis, 761 Radiofrequency ablation (RFA), 328, 466, 518, 529
Pelvis, 515–517, 520 Radiogenomics, 776
Pembrolizumab, 380, 598, 713, 723 Radiography, 283–284
Percutaneous, 518 Radiologic-pathology correlation, 779
Percutaneous cement augmentation, 527–529 Radiomic, 776, 781
Percutaneous pedicle screw fixation, 527 Radiopharmaceuticals, 490
Periacetabular, 519, 520 Radiosurgery for intact metastases, 570
Periampullary carcinoma, 381 Radiotherapy, 676
Pericyte mimicry (PM), 74–81 Radium-223, 298, 300, 307
Peritoneal metastases, 413–414 Ratio, 672
Pharmacogenetics, 266 Reconstruction, 520
Photoacoustics, 254–257 Reconstruction cage, 519
Photons, 494 Rectal carcinoma, 381
Piecemeal resection, 555 Recurrent brain metastasis, 549
PKCs, 655 Recurrent breast cancer, 357
Planning target volume (PTV), 701 Recursive Partitioning Analysis (RPA), 568
Plasticity, 22, 23, 26, 27 Regional nodal irradiation, 676
Platelets, 201 Regulatory T cells, 489
Pleomorphic liposarcomas, 613 Regulatory T lymphocytes (Tregs), 614
Polymethylmethacrylate, 495 Replication protein A (RPA), 656
Portal vein embolization (PVE), 465 Replicative immortality, 7–8
Positron-emission tomography (PET), 283, 288–292, 315 Resection, 436
Index 805

Resection arthroplasty, 520 Surgical margin, 434

Resistance to therapy, 30 Surgical oncology, 387
Retinoblastoma protein (pRB), 759, 761, 762 Surgical resection of the primary breast tumor is of benefit for patients
Reverse phase protein array, 19 with metastatic breast cancer, 401
Rhabdomyosarcoma, 596, 633 Surgical technique, 506
RNA, 18–19 Synchronous, 412
Rotterdam criteria, 345, 347 lesions, 684
resection, 411
staged, 411–412
S Synergy between systemic therapy and radiation, 692
Sacroiliac ligaments, 516 Synovial sarcoma (SS), 595, 596, 629, 633–635
Sarcomas, 121, 595, 625 Systemic chemotherapy, 379
Sciatic buttress, 516, 517, 520 Systemic (Fisher) hypothesis of cancer, 682
Secondary follicle-like TLS (SFL-TLS), 642 Systemic theory, 378, 380
Secretome, 23–24, 795 Systemic therapy, 379–380
Sedation, 327
Senescence-associated secretome, 24
Sentinel lymph node (SLN), 56, 210, 215–216, 218–220, 225 T
Sentinel lymph node biopsy (SLNB), 370, 373, 380, 398–399, 596 T1, 310–312, 315, 316, 319, 576
SENTInel NeoAdjuvant (SENTINA), 355 T2, 310–312, 315, 316, 319
Sentinel node, 340–345, 348, 349 Tantalum, 519
Sentinel node basin, 363, 365 Target-based therapy, 560–563
Sentinel Node Biopsy Following Neoadjuvant Chemotherapy (SN Targeted axillary dissection, 356
FNAC), 355, 356 Targeted therapy, 98–101
Separation surgery, 525, 526, 530, 531, 701 Taxanes, 379
SF3B1, 19 T-cell epitope, 781
Shear stress, 202 T-cells, 489, 709
Siewert classification, 423 Tenascin, 90
Single fraction, 702 Ten-eleven-translocation enzymes (TET), 18
Single-photon emission computed tomography (SPECT), 290 Tertiary lymphoid structure (TLS), 640–642
Single-photon emission computed tomography-computed tomography Tesla, 288
(SPECT-CT), 290 Testicular cancer, 729, 734
Site-specific metastasis, 108–110 TGF-beta (TGFβ), 23
Skeletal remodeling, 484 Therapeutic, 672
Society of Surgical Oncology “Choosing Wisely” guidelines, 357 Therapeutic ratio, 661–663, 665
Soft tissue, 511 Thiotepa, 580
Soft tissue sarcoma (STS), 121, 611, 629–635 Thoracic duct, 188
Solitary brain metastasis, 568 Thyroid cancer, 369, 370, 372, 373
Sourcil, 516 Tibia/fibula, 509–511
Spatial resolution, 289 TIGIT, 720
Specification, 200 Tissue microarray, 640–642
Spectrum theory, 378, 380 TNF-α and melphalan-based isolated limb perfusion
Spinal cord decompression, 701 (TM-ILP), 625
Spinal metastasis, 531, 699–700 Tolerance dose, 703
Spinal stereotaxis, 698 Topoisomerase inhibitors, 379
Spine radiosurgery (SRS), 570, 673, 697–703 Total hip arthroplasty, 519
Spleen preservation, 436 Total mesorectal excision (TME), 381
Squamous cell carcinoma of the head and neck, 781 Trabectedin, 614
Stage migration, 408 Trametinib, 742
Staged approach, 412 Transarterial embolization, 495
Staging, 275–280 Transcriptional profiling on CTCs, 161
Staging laparoscopy, 425 Transcription factor, 201, 750
Stem cell divisions, 750–751 Transendothelial migration, 32, 111
Stem-cell-like properties, 7–8 Treg cells, 27
Stereotactic, 518 Triple negative (TN), 356
Stereotactic ablative radiotherapy (SABR), 665, 686 Tumor-associated antigens, 781
Stereotactic body radiotherapy (SBRT), 466, 467, 665, 667, 673, 676, Tumor-associated macrophages (TAM), 21, 26, 642–644
686, 697–703 Tumor before surgical resection, 569
Stereotactic radiosurgery (SRS), 524, 530, 567 Tumor heterogeneity, 109
Stiffness, 22, 94 Tumor infiltrating lymphocytes (TILs), 409, 598, 710, 723, 725
Structural integrity, 516, 517 Tumor microenvironment (TME), 21, 30, 107–109, 111–113, 542,
Supervised learning techniques, 776 545, 613, 654, 656–658, 710
Surface guided radiation therapy (SGRT), 664, 665 Tumor mutation burden (TMB), 17, 151, 710, 722
Surgery, 620, 621, 623, 625 Tumor-node-metastasis (TNM) staging system, 56
Surgical cavity shrinkage, 569 Tumour metastasis, 231–233, 235–238
Surgical indications, 548–549 2D cell tracking, 779
806 Index

Two-staged hepatectomy (TSH), 439, 464, 465 Vascular niche, 488

Tyrosine kinase inhibitors (TKI), 379, 560, 623 Vemurafenib, 741, 742
Ventricular devices, 579
Vertebral compression fractures, 703
U Vicious cycle, 488
Ultra-mutator phenotype, 17 Vinca alkaloids, 379
Ultrasound, 327 Volumetric modulated arc therapy (VMAT), 663, 665
Uncontained defects, 518
Undifferentiated high-grade pleomorphic sarcoma, 606–607
Undifferentiated pleomorphic sarcoma (UPS), 122, 595, 629, 634 W
Unknown primary cancers, 58–59, 778 WBRT+ radiosurgery, 570
Unsupervised learning, 776 Well-differentiated (WDLPS), 612, 613
Urokinase plasminogen activator receptor (uPAR), 159 Whipple, A., 381
Urothelial cancer, 729, 731–734 Whole-body DWI, 300
Whole-body magnetic resonance with diffusion-weighted imaging
(WB-MRI-DWI), 304, 307
V Whole-body MRI (WBMRI), 295, 304
V600 BRAF, 740 Whole brain radiation therapy (WBRT), 567
Vaccine, 762, 763 Whole-slide images, 777
Vaginal cuff brachytherapy, 677 Will Rogers phenomenon, 408
Vascular endothelial growth factor (VEGF), 25, 598, 730, 731, 760, 761 WNT/B-catenin signaling, 752