Precision Medicine 

in
Cancers and
Non-Communicable
Diseases
Precision Medicine in
Cancers and
Non-Communicable
Diseases

Edited by
Debmalya Barh
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

© 2019 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Printed on acid-free paper

International Standard Book Number-13: 978-1-4987-7560-1 (Hardback)

This book contains information obtained from authentic and highly regarded sources. While all reasonable efforts have been
made to publish reliable data and information, neither the author[s] nor the publisher can accept any legal responsibility or
liability for any errors or omissions that may be made. The publishers wish to make clear that any views or opinions expressed
in this book by individual editors, authors or contributors are personal to them and do not necessarily reflect the views/
opinions of the publishers. The information or guidance contained in this book is intended for use by medical, scientific or
health-care professionals and is provided strictly as a supplement to the medical or other professional’s own judgement, their
knowledge of the patient’s medical history, relevant manufacturer’s instructions and the appropriate best practice guidelines.
Because of the rapid advances in medical science, any information or advice on dosages, procedures or diagnoses should be
independently verified. The reader is strongly urged to consult the relevant national drug formulary and the drug companies’
and device or material manufacturers’ printed instructions, and their websites, before administering or utilizing any of the
drugs, devices or materials mentioned in this book. This book does not indicate whether a particular treatment is appropriate
or suitable for a particular individual. Ultimately it is the sole responsibility of the medical professional to make his or her
own professional judgements, so as to advise and treat patients appropriately. The authors and publishers have also attempted
to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission
to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us
know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in
any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, micro-
filming, and recording, or in any information storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.
copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-
8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that
have been granted a photocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identifi-
cation and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data

Names: Barh, Debmalya, editor.

Title: Precision medicine in cancers and non-communicable diseases / [edited by] Debmalya Barh.
Description: Boca Raton, FL : CRC Press, 2019. | Includes bibliographical references and index.
Identifiers: LCCN 2018014610| ISBN 9781498775601 (hardback : alk. paper) | ISBN 9781315154749 (ebook)
Subjects: | MESH: Precision Medicine | Neoplasms--therapy | Noncommunicable Diseases--therapy
Classification: LCC RC263 | NLM WB 300 | DDC 616.99/4--dc23
LC record available at https://lccn.loc.gov/2018014610

Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Dedication

I dedicate this book to Dr. Candan Hizel, one of my best friends whose philosophy of life, humanity,
humbleness, and passion for science touches my heart.
Contents

Preface ix
About the editor xi
Contributors xiii

Part 1  PRECISION MEDICINE IN ONCOLOGY 1

1 Precision medicine in oncology: An overview 3

Fazilet Yılmaz, Sultan Ciftci Yılmaz, Esra Gunduz, and Mehmet Gunduz
2 Circulating tumor cells and circulating tumor DNA in precision medicine 17
Anjana Munshi and Satrupa Das
3 Oral squamous cell carcinoma and liquid biopsies: A new tool for precision oncology 29
Laura Álvarez Rodríguez, Laura Muinelo Romay, Abel García García, Rafael López-López,
Alicia Ábalo Piñeiro, and Mario Pérez-Sayáns
4 Precision medicine for brain gliomas 39
Yusuf Izci
5 Precision medicine for colorectal cancer 49
Candan Hızel, Şükrü Tüzmen, Arsalan Amirfallah, Gizem Çalıbaşı Koçal, Duygu Abbasoğlu,
Haluk Onat, Yeşim Yıldırım, and Yasemin Baskın
6 Precision medicine in prostate cancer 121
Nigel P. Murray
7 Breast cancer epigenetic targets for precision medicine 137
Ramona G. Dumitrescu
8 Precision medicine in ovarian carcinoma 145
Shailendra Dwivedi, Purvi Purohit, Radhieka Misra, Jeewan Ram Vishnoi, Apul Goel,
Puneet Pareek, Sanjay Khattri, Praveen Sharma, Sanjeev Misra, and Kamlesh Kumar Pant
9 Precision medicine in myelodysplastic syndromes 173
Ota Fuchs
10 Precision medicine in acute myeloid leukemia 195
Ota Fuchs

Part 2  PRECISION MEDICINE IN NCDs 215

11 Precision medicine in coronary artery disease 217

Melvin George, Luxitaa Goenka, and Sandhiya Selvarajan
12 Precision medicine in stroke and other related neurological diseases 235
Anjana Munshi, Vandana Sharma, and Sulena Singh

vii
viii Contents

13 Precision medicine in osteoporosis and bone diseases 243

Fatmanur Hacievliyagil Kazanci, Fatih Kazanci, M. Ramazan Yigitoglu, and Mehmet Gunduz
14 Precision medicine in diabetes mellitus 259
Sandhiya Selvarajan, Akila Srinivasan, Nishanthi Anandabaskar, Sadishkumar Kamalanathan,
and Melvin George
15 Precision medicine in multiple sclerosis 269
Shoaib Ahmad
16 Precision medicine in asthma and chronic obstructive pulmonary disease 279
Shoaib Ahmad
17 Customized DNA–directed precision nutrition to balance the brain reward circuitry and
reduce addictive behaviors 295
Kenneth Blum, Marcelo Febo, Eric R. Braverman, Mona Li, Lyle Fried, Roger Waite,
Zsolt Demotrovics, William B. Downs, Debmalya Barh, Bruce Steinberg,
Thomas McLaughlin, and Rajendra D. Badgaiyan

Index 311
Preface

Precision medicine is a holistic approach that con- of precision medicine in oncology and examin-
siders variability in genetic makeup, environment, ing circulating tumor cells (CTC) and tumor
and lifestyle of an individual for personalized dis- DNA (CTD) as emerging tools in cancer precision
ease prevention and treatment. Currently, the avail- medicine. In Chapter 2, Dr. Anjana Munshi and
ability of large-scale omics data, gene-environment colleagues have provided a brief account on how
and gene-lifestyle interaction information, as well as CTC and CTD can be characterized and how they
cutting-edge big data analytics and predictive algo- can be used in clinical decision making. Dr. Mario
rithms has enabled us to develop precision medicine Pérez-Sayáns and team have focused on the impor-
strategies on a case-by-case basis. However, it is the tance, challenges, and opportunities of CTC in
beginning and there is a long way to go! oral squamous cell carcinoma precision medicine
Although the precision medicine strategies in Chapter 3. In Chapter 4, Dr. Yusuf Izci has dis-
can be applied to improve every aspect of health cussed the clinical scenario of precision medicine
and wellness, currently the approach is mostly in brain gliomas. Dr. Candan Hizel and his col-
restricted to non-communicable diseases (NCDs), leagues have provided a detailed account of preci-
in which cancers are the primary focus. Few sion medicine approaches in colorectal cancer in
books are available thus far on this topic, and Chapter 5. Precision medicine in prostate cancer
those are mostly related to technical and ethi- with an emphasis on gene–environmental interac-
cal aspects. Precision Medicine in Cancers and tions is described by Dr. Nigel Murray in Chapter 6.
Non-Communicable Diseases is probably the first In Chapter 7, the role of epigenetics in breast cancer
book of its kind to address specific NCD-related precision medicine is discussed by Dr. Ramona G.
precision medicine opportunities. In general, the Dumitrescu. In Chapter 8, Dr. Shailendra Dwivedi
book describes the prevalence, incidence, mortal- and colleagues have given an update on ovarian
ity rate, currently used diagnosis and treatment cancer precision medicine. Finally, in Chapters
approaches, and explores how the precision medi- 9 and 10, Dr.  Ota Fuchs presents the application
cine approach can benefit in predicting, prevent- of the precision medicine concept in myelodys-
ing, managing, and treating cancers and NCDs. plastic syndromes and acute myeloid leukemia,
Further, case studies, challenges, and research respectively.
opportunities are also discussed in some chapters.
A total of 17 chapters, divided between two sec- Part 2
tions, are included in this volume. Part 1 is composed
Part 2 deals with NCDs other than cancers. The
of ten chapters on oncology, and the remaining
section starts with precision medicine in coronary
seven chapters comprising Part 2 describe precision
artery disease (Chapter 11) developed by Dr. Melvin
medicine aspects in a number of other NCDs.
George and team. In Chapter 12, Dr.  Anjana
Munshi’s group has highlighted the precision
Part 1
medicine aspects in stroke and other neurologi-
The book starts with a chapter (Chapter 1) by Dr. cal diseases. In Chapter 13, Dr. Fatmanur Kazanci
Mehmet Gunduz’s group describing the overview and his colleagues have discussed the importance

ix
x Preface

of lifestyle, environmental, and genetic factors in and his team have provided an overview on how the
preventing and treating osteoporosis. Dr. Sandhiya DNA-directed precision nutrition benefits in man-
Selvarajan and colleagues have demonstrated aging reward deficiency syndrome and addiction.
the application of precision medicine approach It is my hope that this book will be useful to the
in type-1 and type-2 diabetes in Chapter  14. readers in understanding the current status, chal-
In Chapters 15 and 16, Dr. Shoaib Ahmad has lenges, and future aspects of various cancers and
reviewed the precision medicine potential in mul- other NCD-specific precision medicine.
tiple sclerosis and asthma/COPD, respectively. In
the final chapter (Chapter 17), Dr. Kenneth Blum Debmalya Barh, PhD
About the editor

Debmalya Barh, MSc (Applied Genetics), MTech (Biotechnology),

MPhil (Biotechnology), PhD (Biotechnology), PhD (Bioinformatics),
Post-Doc (Bioinformatics), and PGDM (Post Graduate in Management),
has more than 15 years of academic, health care, molecular diagnostic,
and bioinformatics industry experience and is an expert in integrative
omics-based biomarker, targeted drug discovery, molecular diagnosis,
and precision medicine in cancers and various complex diseases and
human traits. Dr. Barh works with more than 400 scientists from more
than 100 top-ranked organizations across more than 40 countries. He
has over 150 journal publications and has authored more than 30 book
chapters. He is a branded editor for over 20 cutting-edge, omics-related
reference books published by top-ranked publishers including Taylor &
Francis. He frequently reviews articles for Nature publications, Elsevier,
AACR journals, NAR, BMC journals, PLOS ONE, and Frontiers, to name a few. He has been recognized
by Who’s Who in the World and Limca Book of Records for his significant contributions in managing
advance scientific research.

xi
Contributors

Duygu Abbasoğlu PhD Debmalya Barh MSc, MTech, MPhil, PhD, PGDM
Department of Biology Institute of Integrative Omics and Applied
Anadolu University Biotechnology (IIOAB)
Eskișehir, Turkey West Bengal, India

and
Shoaib Ahmad MPharm, PhD, DBME, PGDBA
University School of Pharmaceutical Sciences Division of Bioinformatics and Computational
Rayat-Bahra University Genomics
Punjab, India NITTE University Center for Science Education
and Research (NUCSER)
(NITTE Deemed to be University)
Arsalan Amirfallah PhD Mangalore, Karnataka, India
Pathology Department/Cell Biology Unit
Landspitali University Hospital Yasemin Baskın MD, PhD
BioMedical Center (BMC) Department of Basic Oncology
University of Iceland Institute of Oncology
Reykjavik, Iceland and
Personalized Medicine
Nishanthi Anandabaskar MD and Pharmacogenomics Research Center
Department of Pharmacology Dokuz Eylül University
Sri Manakula Vinayagar Medical College İzmir, Turkey
and Hospital
Kenneth Blum PhD
Puducherry, India
Western University Health Sciences
Graduate School of Biomedical Sciences
Rajendra D. Badgaiyan PhD Pomona, California
Department of Psychiatry and
Ichan School of Medicine
New York, New York Department of Clinical Neurology
PATH Foundation NY
and New York City, New York

Department of Psychiatry and

Wright State University Division of Neurogenetic Research & Addiction
and Therapy
Dayton Veterans Affairs Medical Center The Florida House Experience
Dayton, Ohio Deerfield Beach, Florida

xiii
xiv Contributors

Eric R. Braverman PhD Melvin George DM

Department of Clinical Neurology Department of Clinical Pharmacology
Path Foundation NY SRM Medical College Hospital & Research Centre
New York, New York Kancheepuram, India

Satrupa Das PhD Apul Goel MBBS, MS, MCh

Department of Molecular Biology Department of Urology
Institute of Genetics and Hospital for King George Medical University
Genetic Diseases Lucknow, India
Osmania University
Hyderabad, India Luxitaa Goenka MSc
Department of Clinical Pharmacology
and SRM Medical College Hospital & Research Centre
Dr. NTR University of Health Sciences Kancheepuram, India
Andhra Pradesh, India
Esra Gunduz dmd, PhD
Department of Medical Genetics
Zsolt Demotrovics PhD
Turgut Ozal University
Institute of Psychology
Ankara, Turkey
Eötvös Loránd University
Budapest, Hungary Mehmet Gunduz md, PhD
Department of Medical Genetics
William B. Downs PhD
Turgut Ozal University
Victory Nutrition International, LLC
Ankara, Turkey
Lederach, Pennsylvania
Candan Hızel PhD
Ramona G. Dumitrescu PhD, MPH Opti-Thera Inc.
Kelly Government Solutions Montreal, Quebec, Canada
Bethesda, Maryland
Yusuf Izci MD
Shailendra Dwivedi MSc, PhD Department of Neurosurgery
Department of Biochemistry Gulhane Education and Research Hospital
All India Institute of Medical Sciences University of Health Sciences
Jodhpur, India Ankara, Turkey

Marcelo Febo PhD Sadishkumar Kamalanathan dm

Department of Psychiatry Department of Endocrinology
McKnight Brain Institute Jawaharlal Institute of Postgraduate Medical
University of Florida College of Medicine Education and Research (JIPMER)
Gainesville, Florida Puducherry, India

Fatih Kazanci md, PhD

Lyle Fried PhD
Department of Medical Genetics
The Shores Treatment & Recovery
Turgut Ozal University
Port St. Lucie, Florida
Ankara, Turkey
Ota Fuchs PhD Fatmanur Hacievliyagil Kazanci md, PhD
Department of Genomics Department of Medical Biochemistry
Institute of Hematology and Blood Transfusion Turgut Ozal University
Prague, Czech Republic Ankara, Turkey

Abel García García PhD, MD Sanjay Khattri MBBS, MD

Oral Medicine, Oral Surgery and Implantology Unit Department of Pharmacology and Therapeutics
Health Research Institute of Santiago (IDIS) King George Medical University
Santiago de Compostela, Spain Lucknow, India
 Contributors xv

Gizem Çalıbaşı Koçal PhD Mario Pérez-Sayáns PhD, DDS

Department of Basic Oncology Oral Medicine, Oral Surgery and
Institute of Oncology Implantology Unit
Dokuz Eylül University Health Research Institute of Santiago (IDIS)
İzmir, Turkey Santiago de Compostela, Spain

Mona Li PhD
Alicia Ábalo Piñeiro PhD
Division of Neurogenetics & Addiction Therapy
Liquid Biopsy Analysis Unit
Florida House
Health Research Institute of Santiago (IDIS)
Deerfied Beach, Florida
Santiago de Compostela, Spain
Rafael López-López MD, PhD
Translational Medical Oncology Laboratory Purvi Purohit MSc, PhD
Health Research Institute of Santiago (IDIS) Department of Biochemistry
Santiago de Compostela, Spain All India Institute of Medical Sciences
Jodhpur, India
Thomas McLaughlin PhD
Center for Psychiatric Medicine
Laura Álvarez Rodríguez DDS
North Andover, Massachusetts
Oral Medicine, Oral Surgery
Radhieka Misra MBBS and Implantology Unit
Era’s Lucknow Medical College and Hospital Health Research Institute of Santiago (IDIS)
Lucknow, India Santiago de Compostela, Spain

Sanjeev Misra MBBS, MS, MCh

Laura Muinelo Romay PhD
Department of Onco-Surgery
Liquid Biopsy Analysis Unit
All India Institute of Medical Sciences
Health Research Institute of Santiago (IDIS)
Jodhpur, India
Santiago de Compostela, Spain
Anjana Munshi MPhil, PhD
Department of Human Genetics and Molecular Sandhiya Selvarajan md, dnb, dm
Medicine Department of Clinical Pharmacology
Central University of Punjab Jawaharlal Institute of Postgraduate Medical
Bathinda, India Education and Research (JIPMER)
Nigel P. Murray mrcp BSc (Hons), MB BCh Puducherry, India
Department of Medicine
Hospital de Carabineros de Chile Praveen Sharma MSc, PhD
and Department of Biochemistry
Faculty of Medicine All India Institute of Medical Sciences
University Finis Terrae Jodhpur, India
Santiago, Chile
Vandana Sharma PhD
Haluk Onat MD Dr. NTR University of Health Sciences
İstanbul Onkoloji Hospital Maltepe Andhra Pradesh, India
İstanbul, Turkey
and
Kamlesh Kumar Pant MBBS, MD Department of Molecular Biology
Department of Pharmacology and Therapeutics Indraprastha Apollo Hospital
King George Medical University New Delhi, India
Lucknow, India

Puneet Pareek MBBS, MD Sulena Singh DM

Department of Radio-Therapy Department of Neurology
All India Institute of Medical Sciences GGS Medical College
Jodhpur, India Faridkot, Punjab
xvi Contributors

Akila Srinivasan MBBS Roger Waite PhD

Department of Pharmacology Division of Precision Behavioral Management
Jawaharlal Institute of Postgraduate Medical Geneus Health, LLC
Education and Research (JIPMER) San Antonio, Texas
Puducherry, India
M. Ramazan Yigitoglu MD, PhD
Bruce Steinberg PhD Department of Medical Biochemistry
Department of Psychology Turgut Ozal University
Curry College Ankara, Turkey
Milton, Massachusetts
Yeşim Yıldırım MD
Şükrü Tüzmen PhD Department of Internal Medicine and
Molecular Biology and Genetics Program Medical Oncology
Department of Biological Sciences Anadolu Medical Center Hospital
Eastern Mediterranean University (EMU) Gebze, Turkey
Famagusta, North Cyprus
Fazilet Yılmaz MD
and Department of Medical Genetics
Arizona State University (ASU) Turgut Ozal University
Phoenix, Arizona Ankara, Turkey

Jeewan Ram Vishnoi MBBS, MS, MCh Sultan Ciftci Yılmaz PhD
Department of Onco-Surgery Department of Medical Genetics
All India Institute of Medical Sciences Turgut Ozal University
Jodhpur, India Ankara, Turkey
 Part     1
Precision medicine in oncology
 1
Precision medicine in oncology:
An overview

FAZILET YILMAZ, SULTAN CIFTCI YILMAZ, ESRA GUNDUZ,

AND MEHMET GUNDUZ

Epidemiology of oncology 3 Future of precision medicine in oncology 10

Genetic mechanism of cancer 3 Why future developments are extremely
Current clinical approaches 4 important in personalized oncology 10
Personalized oncology 5 Next-generation sequencing (massively
Predicting the cancer risk 7 parallel sequencing) 11
Personalized prophylactic treatment in high- Single-cell genome sequencing 11
risk populations 8 Patient-derived tumor xenograft 12
Choosing personalized treatment 9 Mixture of barcodes tumors PRISM 12
Molecular features of the cancer 9 Mutanomes 12
Personal metabolic features 9 Conclusion 13
Predicting drug sensitivity and resistance 10 References 13

EPIDEMIOLOGY OF ONCOLOGY GENETIC MECHANISM OF CANCER

According to the World Health Organization Cancer, the cell population with an altered genetic
(WHO), cancer is a leading cause of death world- profile, does not fit to the normal tissue structure
wide. In 2012, 8.2 million people died of cancer and proliferates without control. Morphology of
(Stewart and Wild, 2014). The most common can- the cancer cells transit from dysplasia to neoplasia
cers in the United States are breast, lung, pros- (Weinberg, 2014). The genetic heterogeneity is the
tate, and colorectal carcinoma (American Cancer most common feature of the cancer cell groups.
Society, 2016). The heterogeneity increases with the cell divi-
When we rank cancer types according to num- sions and can be due to genetic and/or epigenetic
ber of deaths, lung cancer (1.59 million deaths) changes (Swanton, 2012).
is the most common, followed by liver, stomach, One nucleotide change in the genome is called
colorectal and breast cancer. single nucleotide variation (SNV). Phenotypical
Beyond that, every year 14 million people receive alterations can be observed based on the location
a cancer diagnosis (Stewart and Wild, 2014). Even of the SNV. Generally, a change occurring in an
if it does not cause death in the short term, it causes exon is more dangerous than a change occurring
lifestyle changes and morbidity with a heavy emo- in an intron. In the case of a silent mutation, no
tional and financial burden. change can be observed, but missense, nonsense,

3
4  Precision medicine in oncology: An overview

or frameshift mutations usually cause dramatic Mutations in tumor suppressor genes can be inher-
changes in the protein structure. Typically, cancer ited. However, since the mutations in oncogenes are
is coupled with mutation in an oncogene and/or inherited in a dominant manner and most of them
tumor suppressor gene (Klug et al., 2012). are incompatible with life, they cannot be inher-
Both proto-oncogenes and the tumor sup- ited. Hence, most of the familial mutations associ-
pressor genes normally exist in a genome. Proto- ated with the cancer formation are related with the
oncogenes mostly code proteins related with the TSGs (Lodish and Zipursky, 2000).
activation of cell division and differentiation (Todd Upon the mutation of the proto-oncogenes and/
and Wong, 1999). Some of the well-studied proto- or TSGs playing critical roles in the cell such as cell
oncogenes are RAS, WNT, MYC, ERK, and TRK. cycle, proliferation, differentiation, and apoptosis,
These proto-oncogenes turn into oncogenes upon cancerous formation begins. Anomalies accumu-
a mutation or translocation to a transcriptionally late in the course of time by addition of new muta-
active site in the genome. Oncogenes lead to forma- tions and a heterogenic tumor develops (American
tion of cancer cells by promoting cell proliferation Cancer Society, 2014). Tumor tissue inhibits the
and undifferentiation (Weinstein and Joe, 2006). functioning of the normal tissue and clinical
One example of a proto-oncogene turning into symptoms arise depending on the location.
an oncogene is formation of a BCR-ABL fusion
gene observed in chronic myeloid leukemia (CML) CURRENT CLINICAL APPROACHES
after reciprocal translocation between chromo-
some 9 and 22. Normally, an ABL gene codes a Conventionally, cancers are classified according to
tyrosine kinase and the expression of the gene is their histopathological features (Weinberg, 2014).
tightly regulated. After the translocation, the newly Before we had an understanding of cancer’s molec-
formed fusion gene BCR-ABL starts to express ular profile, the main progression factors were can-
tyrosine kinase independent of regulation mecha- cer localization and the cancer’s tissue and cell type
nism. Excess formation of tyrosine kinase activates of origin. In the past, the only curative treatment
proteins related to cell cycle and cell division, then was surgery. Therefore, classification of the can-
in turn the cell starts to proliferate without control cer was based on surgical treatment options and
(Druker, 2002). tissue  localization (Brauer, 1964). Later on, sup-
Another important gene group related to can- portive methods such as chemotherapy and radio-
cer formation is tumor suppressor genes (TSGs). therapy are added to cancer treatment (“Cancer
TSGs are normally responsible for inhibiting research,” 1932).
uncontrolled cell proliferation and inducing apop- Nowadays, besides the conventional chemo-
tosis when necessary. Genes like Rb, p53, and therapeutic drugs, agents targeting the precise
PTEN mutated in many cancers belong to this molecules in specific cancers are being devel-
group. Different than oncogenes, in most cases oped. These new agents produced according to
a single allele mutation will not drive the cells to molecular characterization are named as targeted
the tumor formation. Mutations in both alleles are therapy. Even though these new agents are basi-
required for the cancer development (Lodish and cally the same as chemotherapeutics, their act-
Zipursky, 2000). ing mechanisms differ from them. These agents
One of the most common examples of cancer are either in a form of monoclonal antibody or a
formation due to TSG mutation is BRCA1 (breast small molecule (National Cancer Institute, 2014).
cancer 1) gene mutation, which causes familial Some of the agents developed for the targeted
breast and ovarian cancer. This gene is responsible therapy are approved for the cancer treatment.
for recognition of damaged DNA and can lead The following cancers are mostly studied for the
to cell destruction in case the damage cannot be targeted therapy for the indicated reasons: breast
repaired. It is mostly expressed in the breast tissue. cancer for being the most common, lung cancer
When a mutation occurs in this gene, in spite of for having the highest mortality rate (Stewart and
the DNA damage, the cell continues to divide and Wild, 2014), and hematological cancers for having
accumulate DNA damage. As a result of this pro- no surgical treatment alternatives. In this section,
cess, an abnormal population of cells will be assem- general treatment aspects of these three cancer
bled with a potential to have cancerous formation. types will be elaborated.
 Personalized oncology  5

The most common cancer type worldwide is directed treatment options are available for epider-
breast cancer (Stewart and Wild, 2014). Breast can- mal growth factor receptor (EGFR) and anaplastic
cer can exhibit different clinical symptoms and lymphoma kinase (ALK), aberrantly expressed in
sometimes can be silent until it has metastasized. tumors (Popat and Yap, 2014).
It also shows a varying progress pattern accord- Hematologic malignancies do not form solid
ing to histopathological classification. Infiltrating tumors as breast and lung cancers. These malig-
ductal carcinoma is the most common type of nancies originate from the circulatory and immune
breast cancer. Risk factors for breast cancer include system–related cells. Initially, they are classified as
BRCA1 and BRCA2 mutations, positive family his- leukemia, lymphoma, and myeloma. Subsequently,
tory (especially positive in mother and sisters), age, they are subclassified based on the differentiation
age of first menstrual cycle, age of first pregnancy stage of the hematopoietic cells they originate from
and parity, age of menopause, usage of oral con- (Harris et al., 2000).
traceptives, alcohol consumption, breast tissue In hematological malignancies, molecular clas-
density, bone density, previous breast mass, and sification and genetic-based diagnosis is preferred
previous atypical biopsies (Collaborative Group (Tefferi et al., 2009). Genetic-based molecular diag-
on Hormonal Factors in Breast Cancer, 1997, 2001; nosis allows us to differentiate cells that cannot
Modan et al., 1998). be differentiated with histopathological findings.
In breast cancer, diagnosis is confirmed by In hematological malignancies, chemotherapeu-
histopathological existence of malignant cells. tics and the targeted therapies are the treatment
The mass is surgically removed and the hormone options. Imatinib, used for chronic myeloid leu-
receptor (ER and HER2) profile is identified. kemia, is one example of these treatment options
Furthermore, the sample should be examined (Zarin et al., 2015).
to determine whether metastasis has occurred. General information of common cancer types
Considering receptor situation and metastases, are summarized in Table 1.1.
chemotherapeutics are given as a single agent or Unfortunately, the current clinical approaches
combined. If there is any nodal or distance metas- for the treatment of many cancer types are insuffi-
tases with clear resection margins, surgical resec- cient and the chosen chemotherapeutical methods
tion is considered as the total cure. Genetic profile are not specific enough for a complete treatment of
(ER+, HER2+, and triple negative) is especially the tumors. Herewith, precision treatment meth-
important for choice of chemotherapeutic agent ods can have a higher potential in completely cur-
and personalized treatment (Hutchinson, 2010). ing the cancer.
When HER2 is positive, HER2 inhibitors such as
trastuzumab (Herceptin) (Haq and Gulasingam, PERSONALIZED ONCOLOGY
2016; Rugo et al., 2016) and pertuzumab (Perjeta)
(Gollamudi et al., 2016) can be added to the treat- The intensity of a disease varies from one individ-
ment. Since there is no targetable receptor by a ual to another. One of the most important reasons
drug, the triple negative is the most aggressive for these personal differences is single nucleo-
form of breast cancer. tide polymorphisms (SNPs), which cause genetic
Among the many types of cancer, lung cancer variations in populations. Even a genetic disease
is the deadliest worldwide with 1.8 million diag- with the same mutation exerts itself at different
noses and 1.6 million deaths in 2012. Because of severities due to different penetrance and the dif-
these clinical differences, lung cancer is classi- ferent gene expression profile of a person (Childs
fied as small cell and non-small cell lung cancer et  al., 2015; Ustinova et  al., 2015; Zahary et  al.,
(NSCLC) (Stewart and Wild, 2014). Small cell lung 2015). Also, despite the same severity of a disease,
cancer is the most aggressive type of lung cancer response to a drug varies from person to person.
and has the worst prognosis. NSCLCs include Environmental factors are other reasons for
lung adenocarcinoma, squamous cell carcinoma, diversity in disease and drug response. Consuming
and the large cell carcinoma. The best treatment certain foods at excess rates in some communities
is surgical resection. If there is no benefit from results in altered liver enzyme activities, which in
surgery, other options are radiotherapy, radiofre- turn cause rapid drug degradation, and inhibit
quency ablation (RFA), or cryoablation. Currently, the drug from reaching the intended blood level
6  Precision medicine in oncology: An overview

Table 1.1  General cancer information

Approved screening and Targeted therapy

prevention methods Current treatments molecule
Oral cavity and Visual inspection Radiation therapy, surgery,
pharynx chemotherapy in advanced
disease
Digestive system Colorectal, colonoscopy Surgery, chemotherapy, radiation Colorectal: EGFR
Respiratory Low dose spiral computed Surgery, radiation therapy, Lung: EGFR, ALK
system tomography (LDCT) chemotherapy
Bones and joints
Soft tissue
Skin Regular examination with Surgical excision, electrodissection Melanoma: BRAF
inspection and curettage, radiation therapy,
topical agents, immunotherapy
Breast Mammography Surgery, radiation, hormonal HER2
therapy, targeted therapy
Genital system Cervical: Pap smear and Ovary: Surgery, chemotherapy
HPV vaccine Uterine cervix: Electrosurgical
excision, cryotherapy, laser
ablation, conization, surgery,
radiotherapy, chemotherapy
Uterine corpus: Surgery, radiation,
hormone therapy, chemotherapy
Urinary system — Kidney: Surgery Kidney: Targeted
Bladder: Surgery, immunotherapy, therapy studies
chemotherapy
Prostate: Surgery, radiation,
radioactive seed implant,
hormonal therapy
Eye and orbit
Brain and other
nervous system
Endocrine system — Thyroid: Surgery, radioactive iodine
Lymphoma
Myeloma
Leukemia — Chemotherapy CML: BCR-ABL
Other and
unspecified
primary sites
Source: American Cancer Society, Cancer Facts & Figures 2016, American Cancer Society, Atlanta, 2016.

(National Consumers League and the U.S. Food knowledge, specific screening and treatment for
and Drug Administration [FDA], 2016). each patient are becoming available.
Since the response of a person both to a disease Wide molecular diversity in cancer emphasizes
and drug treatment varies, prescribing the same the importance of personalized medicine in this
treatment for all patient approaches is changing. field (Boisguérin et al., 2014). Some cancer types do
Owing to improved technologies and cumulative not have curative treatment options and urgently
 Personalized oncology  7

need personalized molecular treatment alterna- Predicting the cancer risk

tives. Taking into account all of these, personal-
ized oncology is really important and open for Prediction of the cancer risk is a relatively under-
improvement (Sicklick et al., 2016). developed part of precision medicine. Even
Currently, most developments in personalized though some tests are available, risk assessment
medicine are related to whole genome sequencing. is extremely limited to only a couple of the can-
At the beginning, a whole genome sequence cost cer types (Hamilton et  al., 2013). One reason for
about $2.7 billion (Weinberg, 2014). Today, whole this can be that with cancer being a very geneti-
genome/exome sequencing is easily achievable cally versatile disease, it becomes statistically hard
with next generation sequencing (NGS), within to determine the course of the disease. A second
hours for approximately $1000. It provides high- reason can be the low level of the genetic screen
throughput and totally personalized data in a performed and the absence of a well-defined data
short time with a lower price. Nevertheless, merely pool available to all researchers and the clinicians.
knowing the patients’ whole sequence cannot be Another reason can be the lack of sufficient bioin-
considered as having entire knowledge about the formatics tools and databases to correctly interpret
patient. To have precise information extracted the obtained data.
from the sequencing for treatment purposes, the Cancer risk can be predicted with different
data first should be analyzed. During the analysis, genetic tests for some cancer types (Heald et  al.,
other factors such as epigenetic and environmental 2012). However, performing merely a test is not
factors such as exposomics (exposure of infections, enough for determining the risk and giving correct
toxic agents) should be evaluated with the obtained information to the patient. Pre- and posttest con-
data. sultation, including epigenetic and environmental
Even though the most widely used current factors, are fundamental for correct evaluation and
genetic analysis for tumor profiling is NGS, some processing. Also choosing the appropriate genetic
other genetic screenings are also available. Since test influences accuracy and avoidance of false-
some of the cancer types have been well identified, negative and false-positive results. As a necessity,
instead of sequencing the entire genome, specific the results should be explained to the patient in a
genes or loci related to the tumor formation can be manner revealing the potential expectation and
screened. Some of the currently ­available tests for the potential outcomes clearly.
tumor profiling are listed in Table 1.2. Recently, some companies’ permission for com-
Precision medicine in oncology can be evalu- mercial genetic tests for BRCA gene mutation in
ated in four main sections. The first one is predict- the United States has been removed. The main
ing the cancer risk in whole life and predicting the reason was the result obtained without expert con-
recurrence. Second is personalized prophylactic sultation can be misleading because breast cancer
treatment in the high-risk population. The third does not solely depend on the BRCA mutation
one is personalizing the treatment according to (Cook-Deegan and Niehaus, 2014).
molecular features of the cancer and the metabolic Risk assessment for breast cancer is commonly
features of the patient. Fourth is predicting the practiced. There are many tools for risk assess-
drug resistance (Figure 1.1). ment according to BRCA mutations, family his-
tory, and the personal risk status. Some quantitative
risk assessment tools are BOADICEA (Lee et  al.,
Table 1.2  Available genetic screens 2014), BRCAPRO (BRCAPRO, 2015), Tyrer–Cuzick
for disease profiling (Boughey et al., 2010), and BCRAT (Park et al., 2011).
Specific single gene tests BRCA mutation analysis was patented from a
Specific gene panels commercial company. But, the U.S. Supreme Court
Genotyping panels invalidated the patent in 2013 (Marshall and Price,
Whole genome or exome sequencing
2013). Now, there are alternative companies for
BRCA mutation tests with extended test content.
Epigenetics arrays
More precise cancer risk prediction availability
Pharmacogenomics testing
and taking the necessary precautions in the future
8  Precision medicine in oncology: An overview

Being at
cancer risk

Drug Personalized Prophylactic

resistance oncology treatment

Personalized
treatment

Metabolic Molecular
features of features of
patient cancer

Figure 1.1  Precision medicine in oncology.

can greatly contribute to the welfare of societies. The first idea about aspirin usage in colorec-
For these reasons, more platforms and agencies tal cancer prevention emerged in a case control
should promote and support research and investi- study from 1988. In 1998, another study reported
gations in the cancer risk prediction area. no relationship with CRC and aspirin (Sturmer
et al., 1998). In 2007, a study showed that aspirin is
Personalized prophylactic treatment effective on PTGS2 overexpressing tumors (Harris
in high-risk populations et  al., 2000). Because of some discordant results,
consecutive studies investigated the molecular
Prophylaxis means preventing diseases before relation between aspirin treatment and cancer pre-
having the disease. Some trials for prophylaxis of vention. Some SNPs related with the effect of the
high-risk populations were conducted for some of aspirin have been identified.
the most common cancer types. Daily usage of aspirin causes many side effects,
An example of personalized prophylactic especially on the gastrointestinal system. Besides,
treatment in high-risk populations is daily aspi- aspirin does not exert its effect on every individual.
rin usage in colorectal cancer (CRC) prevention. Under certain circumstances, aspirin usage should
Aspirin (acetyl salicylic acid) is used as an anal- be considered on a personal basis to obtain the
gesic, antipyretic, and antithrombolytic agent. most effective results with the lowest side effects.
Currently, aspirin is being used as a prophylac- The U.S. Preventive Services Task Force (USPSTF)
tic treatment for colorectal cancer in high-risk recommends a low dose of aspirin to only people
populations. 50–69 years old with a high-risk profile.
 Personalized oncology  9

Familial adenomatous polyposis (FAP) is char- The first drug used based on molecular features
acterized with a vast number of adenomatous of the tumor is imatinib targeting the BCR-ABL
polyp formation in the colon and mutation in the fusion gene coding a tyrosine kinase always on due
APC gene. Individuals with FAP develop colon to translocation of the gene. Imatinib blocks the
cancer after 40 years of age. Prophylactic colec- BCR-ABL protein leading to inactivation of the sig-
tomy is recommended to these individuals around naling pathway. Eventually, cell proliferation stops.
25 years of age (Hagen et al., 2015). An FDA-approved companion diagnostic test for
Another highly practiced preventive cancer imatinib is commercially available (FDA, 2016b).
treatment is prophylactic mastectomy. It is rec- Breast cancer drugs Herceptin (trastuzumab),
ommended to people with a strong positive fam- Perjeta (pertuzumab), and Kadcyla (ado-trastu-
ily history and/or high mutation risk in the BRCA zumab emtansine) also work with the same
gene (National Cancer Institute, 2016). Bilateral approach. They target HER2, an oncogene acti-
prophylactic mastectomy has been shown to vated in some breast cancer (FDA, 2016b).
reduce the risk of individuals with strong fam- Another example of drugs developed based on
ily history (Rebbeck et  al., 2004; Domchek et  al., molecular characteristics of the tumor are epi-
2010). Two FDA-approved drugs—tamoxifen and dermal growth factor receptor (EGFR) blockers
raloxifene—have also been administered to indi- that have been used for non-small cell lung cancer
viduals with a risk of 1.67% or greater of developing (NSCLC). EGFR belongs to the HER receptor fam-
breast cancer in the next 5 years in order to reduce ily and is related to several pathways, such as RAS,
the risk (Cummings et al., 1999; Fisher et al., 2005). Akt, and mTOR, playing a role in cell proliferation.
The first generation of drugs developed for target-
Choosing personalized treatment ing EGFR are erlotinib and Iressa (gefitinib). Ten
months after treatment, many patients developed
In precision medicine, the decision of drug selec- resistance to these drugs, so a second-generation
tion for an optimal treatment should be based on drug targeting pan-HER receptors, afatinib, has
the molecular features of the cancer and the meta- been developed. AZD9291 is the third-generation
bolic features of the patients. Both aspects are elab- drug developed specifically for targeting mutant
orated upon in the next sections. EGFR (Popat and Yap, 2014).
Until now, most of the targeted therapies for
MOLECULAR FEATURES OF THE CANCER cancer treatment have been intended to affect one
Cancer’s molecular features were first identified in mutation or one pathway. Even though in most
hematological malignancies. The reason for such cases a driver mutation plays an essential role in
discovery was the search for alternative treatment cancer formation initially, in the later stages of the
options due to no possibility of surgical interven- tumor development, cells accumulate more muta-
tions. Another reason was the cases showing dif- tions, so usually targeting a single aberration does
ferent prognoses with no histological distinction. not completely cure the malignant formation.
Additionally, thanks to some of the solid tumors Today with the developing sequencing techniques,
with no treatment, researchers have been directed it is possible to analyze various mutations in the
to investigate the molecular mechanism of the tumors. Treatment approaches should be radically
malignancies for developing potential drugs for revised based on mutations, SNPs, metabolic dif-
targeting the precise mechanisms. ferences, and environmental factors (Lawrence
Several FDA-approved oncology drugs with et al., 2003).
companion diagnostic tests are currently commer-
cially available (Jørgensen, 2015). Drugs targeting PERSONAL METABOLIC FEATURES
oncogenes such as HER2, KRAS, BRAF, ALK, and Each individual possesses a different metabolism.
cKi have been developed (Marrone et  al., 2014). Since drug pharmacokinetics depends on metabo-
In most cancer cases, expressions of these genes lism, the drugs cannot show the same effect in all
are upregulated in tumors compared to normal individuals. Some of the drugs can be absorbed
tissues and suppressing the expression results in less than desired by the patients, and some of them
inhibition of cancer progression while relatively can be degraded very rapidly based on the meta-
unharming the normal tissue. bolic differences so they cannot reach the optimal
10  Precision medicine in oncology: An overview

level and cannot show the intended affect. All of Based on the pharmacokinetics of cisplatin,
these drug response differences due to the different resistance can develop due to quick inhibition of
genetic backgrounds of the individuals are named cisplatin by glutathione s-transferase or due to low
pharmacogenomics (Scott, 2011). Precision medi- levels of the drug inside the cells caused by trans-
cine in oncology also aims to oversee these poten- port of the cisplatin out of the cell. Different com-
tial differences. binations of SNP in each molecule for each patient
The first pharmacogenomics studies were con- at this step defines the tumor’s drug resistance.
ducted on cytochrome P450 enzymes. Cytochrome SNP variations among the individuals are one of
P450 enzymes (CYPs) are responsible for drug the primary reasons for the development of drug
metabolism in the liver. Fifty-eight identified CYPs resistance.
exist in the human. Some of the SNPs have been
determined for the CYPs. FDA-approved pharma- FUTURE OF PRECISION MEDICINE
cogenomic biomarkers have been used, mostly for IN ONCOLOGY
drugs used for pulmonary, cardiac, rheumatologic,
and infectious diseases (FDA, 2016a). Even though Why future developments are
some examples exist in the oncology area, more extremely important in personalized
extensive studies are needed in the oncology phar- oncology
macogenomics field.
Cancer is an aggressive disease with a high inci-
Predicting drug sensitivity dence of mortality. Besides being highly deadly,
and resistance cancer seriously decreases the quality of life with
high emotional and financial burden. In most
Chemotherapeutics are either used alone or com- cases, cancers can metastasize or relapse after a
bined with a surgical treatment, depending on the treatment and become harder to treat due to accu-
cancer type. However, due to accumulated muta- mulated mutations. There are also several aggres-
tions over time, some of the tumors develop resis- sive cancer types without any precise treatment.
tance to the chemical treatment (Szakacs et  al., One example is anaplastic thyroid cancer
2004; Restifo et  al., 2016). Prediction medicine (ATC). It is an extremely aggressive cancer type in
investigates the potential of the tumors to develop which most patients die within 6 months of diag-
resistance to specific drugs. nosis. ATC quickly invades the surrounding tissue
One of the most used anticancer drugs is cis-­ and metastasizes, which makes surgical interven-
diamminedichloridoplatinum (cisplatin). It was tion impossible. Being also highly resistant to the
also the first anticancer drug containing platinum. known chemotherapeutics, personalized medicine
Platinum complex binds to DNA and forms cross can be the solution for ATC treatment.
bands in the DNA strand. Subsequently, cells undergo Gallbladder cancer is another example of a can-
apoptosis. Some of the tumor cells gain resistance cer with a very low treatment success rate. Because
after a course of the treatment. Cisplatin metabolism of the gallbladder’s anatomic location, radical sur-
is explained in the following mechanism: gical resection is impractical. Currently, research-
ers are focusing on molecular characterization
1. Cells take the cisplatin inside the cell by either and personalized medicine for gallbladder cancer
passive transport or by an identified trans- treatment, thanks to genetics methods such as
porter such as CTR1, CTR2, and CTR3. next-generation sequencing. Thus far, ARID1A,
2 . When it enters the cell, it binds to the DNA. BRAF, CDKN2A/B, EGFR, ERBB2-4, HKN-RAS,
3. The damage caused by the cisplatin is fixed PIK3CA, PBRM1, and TP53 mutations have been
by the members of nucleotide excision repair associated with this cancer type (Sicklick et  al.,
(NER) pathway. 2016). But still, no approved treatment exists.
4. Cisplatin is inhibited by the glutathione As described in both examples and when evalu-
s-transferase. ated as a whole, the success rate in treating cancer
5. Conjugated cisplatin is exported out of the cell is still extremely low. At present, cancer is the one
by either MPR2 or ATP7A/ATP7B transporters of leading causes of disease-related deaths world-
(Amable, 2016). wide. Given the fact that each individual with a
 Future of precision medicine in oncology  11

Which treatment Sample from healthy people

is optimum for and
my cancer? Am I going to get biopsy from cancer
cancer?

2 3
1
4 Genomic
material
Doctor and pre-test
consultation

7 6 5
ACACACACACACACAC ACACGAAGG
ACACGAAGGTGTAGCT
TGAAGCGCCTGCTTTCA
8 GTCTTTCCCAGTCTTTC
CAAGCATCTCTCAGGCT
GAAAGCAAGCG
Risks and doctor and post-test
Bioinformatics Reference sequence Raw data
consultation

Figure 1.2  Application of precision medicine in oncology.

different genetic background responds differently three main manufacturers in the market today
to both the disease and treatment, from a cancer with several machines available that are based on
aspect, precise medicine with targeted molecular different mechanisms and which are continuously
therapy seems extremely promising for finding an being further developed.
ultimate cure for cancer (Figure 1.2). Without a While analysis was previously carried out on
doubt, some of the developing fields described next a specific mutation within a specific exon found
bear great importance for carrying precision medi- within a particular gene, today in a single reaction,
cine in oncology to the next level. all of the mutations contained within a sample
can be sequenced. Such analyses are referred to
Next-generation sequencing with terminology such as whole genome sequenc-
ing (WGS) or whole exome sequencing (WES).
(massively parallel sequencing)
However, in this field one of the greatest current
The human genome sequencing project was started problems is the lack of sufficient bioinformatics
using Sanger sequencing technology. However, the software to handle the amount of data generated
disadvantages of this method include high cost, by this method as well as the lack of databases for
labor intensiveness, and time. Later, the massively analysis of incidental mutations.
parallel sequencing method known as next-gen- When looking at NGS from the perspective of
eration sequencing (NGS) was developed and is precision oncology, in the future when databases
commonly used today. With this method, genomic with sufficient information have been made avail-
material is digested randomly into fragments of a able the following could be possible: (1) a single
particular size. The method is based on the reading WGS analysis for an individual can determine
of many regions repeatedly (ranging from 100 to genetic risk classification for all cancer types or (2)
10,000 times). This new method made it possible mutation analysis of an existing tumor can provide
for the sequencing of the entire human genome instruction regarding personalized treatment.
(once sample prep is complete) in hours, something
not possible using the Sanger sequencing method. Single-cell genome sequencing
Also, the cost is extremely low and with the spread
of this technology throughout laboratories, its use Cancer cells possess a wide range of intratumoral
has increased quite significantly. There are two or heterogeneity in terms of morphology, genotype,
12  Precision medicine in oncology: An overview

gene expression, metabolism, proliferative, and Mixture of barcodes tumors PRISM

metastatic potential. Such heterogeneity creates
a need for analysis at the single-cell level. Single- The PRISM method allows the screening of many
cell analysis is also important for tumors that drugs by treating several cancer cell lines con-
cannot easily be cultured in an in vitro environ- tained within a single pool. Initially, cells are
ment. Previously, sequencing was only able to labeled with a 24-nucleotide barcode by using len-
target some specific regions. Now, whole genome/ tivirus as a carrier. After the drug administration,
exome sequencing has become more available and these barcodes make it possible to identify the dead
single-cell analysis is being performed successfully cell lines that are considered as responsive to the
and becoming more accessible (Van Loo and Voet, drug treatment.
2014). Sequencing/cell resolution is increasing each To test this assay’s sensitivity and reliability, it
day. was compared with currently used methods. Once
For single-cell sequencing, cells should be the sensitivity of this method was shown, many
selected individually with methods such as fluores- studies were carried out using this assay. Thanks
cence-activated cell sorting (FACS), microfluidics- to this method, it is possible to test a host of dif-
based cell sorting, laser capture microdissection ferent molecules/drugs/materials of different cell
or microdroplet technologies (Nakamura et  al., lines at once. In short, the effect of drugs on a
2016; Zhang et al., 2016). Normally, DNA material host of different cell lines can be studied simul-
of a single cell is around 6 pg and not enough for taneously. At the same time, by screening many
analysis (Van Loo and Voet, 2014). However, this molecules it affords the opportunity to identify
material can be amplified by commercially avail- molecules with therapeutic potential. It is thought
able genome amplification kits and the sequenc- that this method, used together with single-cell
ing is performed thereafter (Qiagen, 2016). The next-generation sequencing technology, can offer
processes that can be done at single cell resolution an even more sensitive level of resolution for the
are whole genome, exome, transcriptome, (scRNA- understanding of cellular mutations and cel-
seq), and epigenomic (methylation and chroma- lular behavior. This high-resolution technology
tin structure) analyses. Since it can give detailed not only further opens the door to predicting a
information about cancer’s behavior and reveal the patient’s overall tumor response to a drug but even
potential treatment options, single-cell analysis presents the possibility of identifying the response
has great importance from the aspect of precision of individual cells within that patient’s tumor
medicine in oncology. based on each cell’s specific mutation profile (Yu
et al., 2016).
Patient-derived tumor xenograft With time, bioinformatics and data mining are
becoming ever more important in personalized
A xenograft is the transplant of one tissue to medicine.
another species. Patient-derived xenograft (PDX)
is transplanting cancer tissue of one individual Mutanomes
to another species and aims to investigate the
molecular characteristics, drug response, and the A mutanome is used to describe total mutations in
aggressiveness of the tumor formation. The most tumor cells and is also known as patient centered
common application of PDX is transplanting the tumor vaccination. It is a new treatment approach
cancer tissue from a patient to a nude mouse with using the body’s own defense system to eliminate
a suppressed immune system and obtaining infor- tumor cells and it is specifically designed for each
mation about histology, genomic structure, cellu- patient’s tumors (Kuhn et al., 2011).
lar heterogeneity, and drug responsiveness of the Via immune informatics, these mutations are
tumor. A treatment approach can be tested on this analyzed and the correct mutation for using them
mouse model and the potential outcome of the as a vaccine is decided (Patronov and Doytchinova,
treatment can be anticipated. It is an important 2013). Choosing the appropriate mutation is criti-
development regarding personalized medicine for cal, since some of the mutations can be in a non-
being an in vivo study at the single-patient level protein coding region, some of them can be shared
(Cho et al., 2016). with the patient’s germ-line genome, and some of
 References 13

them can be nonimmunogenic (Boisguérin et al., deciding how a patient should be informed regard-
2014). Another consideration is choosing an epit- ing the results are other concerns that should be
ope not targeting the other tissues enabled by bio- resolved.
informatics software. In early attempts, peptides Another major problem is the need of the bioin-
were given directly. Nowadays, modified mRNA formatics tools to evaluate high-throughput data.
(modRNA) technologies are improved and they Storage and the software developments should be
are more effective than peptide vaccines. When synchronized with the data production.
compared to peptides, modRNAs are more stable, Another major concern is the ethical aspect of
have a long-term effect, and are less immunogenic. the obtained data. The rules for which part of the
Also, synthesized mRNA can be engineered as a patient’s information can used and which part can
­poly-epitopic nucleotide sequence. be shared with insurance companies, relatives, or
One example is autologous patient-specific vac- even the patients themselves should be well defined
cines used for advanced melanoma. Prophylactic (Rothstein, 2012; Allen et al., 2014).
immunization with 27mer peptides encoding Even though precision medicine appears to be
specific mutations resulted in complete protec- extremely promising for the oncology field, it still
tion from the tumor and 40% survival of the mice, has a long journey of revision and development in
whereas all mice in the control group died within the aspects of defining the targets, drug designing,
44 days (Kreiter et al., 2015). improving the appropriate methods, and making
One major handicap of personalized vaccina- new policies.
tion is that there is a time-consuming process of
approval for each vaccine. Theoretically, all new REFERENCES
vaccines should pass the preclinical and the phase
studies. Acceleration of the approval stage for per- Allen NL et al. Biobank participants’ preferences
sonalized vaccination will greatly improve the field for disclosure of genetic research results:
and can be very promising in the treatment of the Perspectives from the OurGenes, OurHealth,
cancer. OurCommunity project. Mayo Clin Proc
89(6);2014:738–46.
CONCLUSION Amable L. Cisplatin resistance and opportuni-
ties for precision medicine. Pharmacol Res
The potential of precision medicine carries great 106;2016:27–36.
importance for several different areas of medicine, American Cancer Society. Cancer Facts & Figures
and it bears an extra significance for the oncology 2016. American Cancer Society, Atlanta, 2016.
field for many reasons. First, cancer is a heterogenic American Cancer Society. Genes and cancer.
disease that genetically varies not only among 2014. https://www.cancer.org/cancer/cancer-
the patients but also among the cells of the same basics/cancer-prevalence.htm
tumor requiring different clinical courses. Second Boisguérin V et al. Translation of genomics-
is the need of precognition of the disease and tak- guided RNA-based personalised cancer
ing precautions because of the lack of a definitive vaccines: Towards the bedside. Br J Cancer
treatment. Third is the failure of many treatment 111(8);2014:1469–75.
methods due to personal differences of drug effect Boughey JC et al. Evaluation of the Tyrer-Cuzick
and resistance mechanisms. Owing to these neces- (International Breast Cancer Intervention
sities, precision medicine carries a great potential Study) model for breast cancer risk prediction
in the oncology field and some major problems in women with atypical hyperplasia. J Clin
should be addressed to get to a promising point for Oncol 28(22);2010:3591–6.
precision medicine in oncology. Brauer W. [The TNM System?]. Radiobiol
Thanks to high throughput analysis, a vast Radiother (Berl) 5;1964:761–5.
amount of data is produced on a daily basis. The BRCAPRO. 2015. Available from http://bcb.dfci.
major problem is interpretation of the data and harvard.edu/bayesmendel/brcapro.php.
reflecting the outcome to the clinics. Even on Accessed 2016.
a personal basis, analysis of the data obtained, Cancer research: Chemotherapy and radium
determining the importance of the outcome, and therapy. Br Med J 2(3746);1932:766–8.
14  Precision medicine in oncology: An overview

Childs EJ et al. Common variation at 2p13.3, Hamilton W et al. Evaluation of risk assess-
3q29, 7p13 and 17q25.1 associated with ment tools for suspected cancer in general
susceptibility to pancreatic cancer. Nat Genet practice: A cohort study. Br J Gen Pract
47(8);2015:911–6. 63(606);2013:e30–6.
Cho SY et al. An integrative approach to preci- Haq R, and Gulasingam P. Duration of trastu-
sion cancer medicine using patient-derived zumab in patients with HER2-positive meta-
xenografts. Mol Cells 39(2);2016:77–86. static breast cancer in prolonged remission.
Collaborative Group on Hormonal Factors in Curr Oncol 23(2);2016:91–5.
Breast Cancer. Breast cancer and hormone Harris NL et al. The World Health Organization
replacement therapy: Collaborative reanaly- classification of hematological malignancies
sis of data from 51 epidemiological studies report of the Clinical Advisory Committee
of 52,705 women with breast cancer and Meeting, Airlie House, Virginia, November
108,411 women without breast cancer. Lancet 1997. Mod Pathol 13(2);2000:193–207.
350(9084);1997:1047–59. Heald B, Edelman E, and Eng C. Prospective
Collaborative Group on Hormonal Factors comparison of family medical history with
in Breast Cancer. Familial breast cancer: personal genome screening for risk assess-
Collaborative reanalysis of individual data ment of common cancers. Eur J Hum Genet
from 52 epidemiological studies includ- 20(5);2012:547–51.
ing 58,209 women with breast cancer and Hutchinson L. Breast cancer: Challenges, contro-
101,986 women without the disease. Lancet versies, breakthroughs. Nat Rev Clin Oncol
358(9291);2001:1389–99. 7(12);2010:669–70.
Cook-Deegan R, and Niehaus A. After Myriad: Jørgensen JT. Companion diagnostics: The key
Genetic testing in the wake of recent to personalized medicine. Expert Rev Mol
Supreme Court decisions about gene patents. Diagn 15(2);2015:153–6.
Curr Genet Med Rep 2(4);2014:223–41. Klug WS et al. Concepts of Genetics. 10th edi-
Cummings SR et al. The effect of raloxifene tion. Pearson, Boston, 2012.
on risk of breast cancer in postmenopausal Kreiter S et al. Mutant MHC class II epitopes
women: Results from the MORE r­ andomized drive therapeutic immune responses to can-
trial. Multiple Outcomes of Raloxifene cer. Nature 520(7549);2015:692–6.
Evaluation. JAMA 281(23);1999:2189–97. Kuhn AN et al. Determinants of intracellular
Domchek SM et al. Association of risk-reducing RNA pharmacokinetics: Implications for
surgery in BRCA1 or BRCA2 mutation car- RNA-based immunotherapeutics. RNA Biol
riers with cancer risk and mortality. JAMA 8(1);2011:35–43.
304(9);2010:967–75. Lawrence A, Loeb KRL, and Anderson JP.
Druker BJ. Perspectives on the development of Multiple mutations and cancer. PNAS
a molecularly targeted agent. Cancer Cell 100;2003:776–81.
1(1);2002:31–6. Lee AJ et al. BOADICEA breast cancer risk pre-
Fisher B et al. Tamoxifen for the prevention diction model: Updates to cancer incidences,
of breast cancer: Current status of the tumour pathology and web interface. Br J
National Surgical Adjuvant Breast and Cancer 110(2);2014:535–45.
Bowel Project P-1 study. J Natl Cancer Inst Lodish H et al. Section 24.2, Proto-oncogenes
97(22);2005:1652–62. and tumor-suppressor genes. In: Molecular
Gollamudi J et al. Neoadjuvant therapy for Cell Biology. 4th edition. W. H. Freeman, New
early-stage breast cancer: The clinical York, 2000. Available from: https://www.ncbi.
utility of pertuzumab. Cancer Manag Res nlm.nih.gov/books/NBK21662/.
8;2016:21–31. Marrone M et al. Multi-marker solid tumor panels
Hagen CE, Setia N, and Lauwers GY. Familial using next-generation sequencing to direct
adenomatous polyposis: A review of gastro- molecularly targeted therapies. PLoS Curr 2014.
intestinal manifestations. Diagn Histopathol http://currents.plos.org/genomictests/article/
21(4);2015:152–60. multi-marker-solid-tumor-panels-using-next-
 References 15

generation-sequencing-to-direct-molecularly- Rothstein MA. Disclosing decedents’ research

targeted-therapies/. results to relatives violates the HIPAA Privacy
Marshall E, and Price M. U.S. Supreme Court Rule. Am J Bioeth 12(10);2012:16–7.
strikes down human gene patents. Science. Rugo HS et al. A clinician’s guide to biosimilars in
June 13, 2013. oncology. Cancer Treat Rev 46;2016:73–9.
Modan B et al. Cancer incidence in a cohort Scott SA. Personalizing medicine with
of infertile women. Am J Epidemiol clinical pharmacogenetics. Genet Med
147(11);1998:1038–42. 13(12);2011:987–95.
Nakamura K et al. Culture-independent method Sicklick JK et al. Genomics of gallbladder
for identification of microbial enzyme- cancer: The case for biomarker-driven
encoding genes by activity-based single-cell clinical trial design. Cancer Metastasis Rev
sequencing using a water-in-oil microdroplet 35(2);2016:263–75.
platform. Sci Rep 6;2016:22259. Stewart BW, and Wild CP. (Eds). World Cancer
National Cancer Institute. Surgery to reduce the Report 2014. World Health Organization,
risk of breast cancer. https://www.cancer. Geneva, 2014.
gov/types/breast/risk-reducing-surgery-fact- Sturmer T et al. Aspirin use and colorectal
sheet. Accessed April 8, 2016. cancer: Post-trial follow-up data from the
National Cancer Institute. Targeted cancer thera- Physicians’ Health Study. Ann Intern Med
pies. April 25, 2014. Available from http:// 128(9);1998:713–20.
www.cancer.gov/about-cancer/treatment/ Swanton C. Intratumor heterogeneity: Evolution
types/targeted-therapies/targeted-therapies- through space and time. Cancer Res
fact-sheet#q1. Accessed 2016. 72(19);2012:4875–82.
National Consumers League and the U.S. Food Szakacs G et al. Predicting drug sensitivity and
and Drug Administration (FDA). Avoid food- resistance: Profiling ABC transporter genes in
drug interactions. 2016. Publication no. CDER cancer cells. Cancer Cell 6(2);2004:129–37.
10-1933. Tefferi A, Thiele J, and Vardiman JW. The
Park S et al. Comparison of breast cancer risk 2008 World Health Organization clas-
estimations, risk perception, and screen- sification system for myeloproliferative
ing behaviors in obese rural Korean women. neoplasms: Order out of chaos. Cancer
Oncol Nurs Forum 38(6);2011:E394–401. 115(17);2009:3842–7.
Patronov A, and Doytchinova I. T-cell epitope Todd R, and Wong DT. Oncogenes. Anticancer
vaccine design by immunoinformatics. Open Res 19(6a);1999:4729–46.
Biol 3(1);2013:120139. http://doi.org/10.1098/ Ustinova M et al. Impact of KRAS variant
rsob.120139 rs61764370 on breast cancer morbidity. Exp
Popat S, and Yap T. Toward precision medicine Oncol 37(4);2015:292–4.
with next-generation EGFR inhibitors in non- U.S. Food and Drug Administration (FDA).
small-cell lung cancer. Pharmacogenomics Homepage. 2016a. Available from http://
Pers Med 7;2014:285–92. www.fda.gov.
Qiagen. Overview of whole genome amplifica- U.S. Food and Drug Administration (FDA).
tion. Available from: https://www.qiagen.com/ List of cleared or approved compan-
de/resources/technologies/wga/overview-on- ion diagnostic devices (in nitro and
wga/. Accessed 2016. imaging tools). 2016b. Available from
Rebbeck TR et al. Bilateral prophylactic mastec- http://www.fda.gov/MedicalDevices/
tomy reduces breast cancer risk in BRCA1 and ProductsandMedicalProcedures/
BRCA2 mutation carriers: The PROSE Study InVitroDiagnostics/ucm301431.htm.
Group. J Clin Oncol 22(6);2004:1055–62. Van Loo P, and Voet T. Single cell analysis of
Restifo NP, Smyth MJ, and Snyder A. cancer genomes. Curr Opin Genet Dev
Acquired resistance to immunotherapy 24;2014:82–91.
and future challenges. Nat Rev Cancer Weinberg RA. The Biology of Cancer. Second
16(2);2016:121–6. edition. Garland Science, New York, 2014.
16  Precision medicine in oncology: An overview

Weinstein IB, and Joe AK. Mechanisms of dis- Association with colorectal cancer susceptibil-
ease: Oncogene addiction—A rationale for ity risk in a Malaysian population. Oncol Lett
molecular targeting in cancer therapy. Nat 10(5);2015:3216–22.
Clin Pract Oncol 3(8);2006:448–57. Zarin DA, Tse T, and Ross JS. Trial-results report-
Yu C et al. High-throughput identification of ing and academic medical centers. N Engl J
genotype-specific cancer vulnerabilities in Med 372(24);2015:2371–2.
mixtures of barcoded tumor cell lines. Nat Zhang X et al. Single-cell sequencing for precise
Biotechnol 34(3);2016:419–23. cancer research: Progress and prospects.
Zahary MN et al. Polymorphisms of cell cycle Cancer Res 76(6);2016:1305–12.
regulator genes G870A and C215G:
 2
Circulating tumor cells and circulating
tumor DNA in precision medicine

ANJANA MUNSHI AND SATRUPA DAS

Introduction 17 General features of ctDNA 21

Characteristics and utility of circulating ctDNA selection approaches 21
tumor cells (CTCs) 18 ctDNA molecular characterization and cell
General features of CTCs 18 analysis 21
CTC selection approaches 18 ctDNA clinical utility 22
CTC molecular characterization and cell analysis 19 Discussion and conclusion 22
CTC clinical and research utility 20 References 23
Characteristics and utility of circulating
tumor DNA (ctDNA) 21

INTRODUCTION This has led researchers to look into alterna-

tive biological samples, one being “liquid biopsy,”
Oncology is a unique area of therapeutics owing which involves screening of peripheral blood by
to the plethora of cancers that have emerged over collecting and enriching the circulating tumor
several years. It gets more interesting because cells (CTCs) and circulating tumor DNA (ctDNA).
the mutations that drive progression of tumors This is increasingly being considered due to its
also serve as the specific targets for treating can- higher sensitivity, noninvasive nature, early tumor
cer. Significant progress has been made in cancer diagnosis ability, better recurrence monitoring,
research, to identify critical targets for its treat- and therapeutic guidance (Sestini et al., 2015). The
ments, but unfortunately there is no universal primary cause for cancer mortality is its ability of
answer that can help reduce or eliminate the bur- tissue invasion and metastasis, and thus the most
den of cancer. Several modes of treatment (che- effective therapy would be to delay or prevent this
motherapy, surgery, radiation therapy) have been process. Therefore, an early and accurate detection
developed, but procedures get complicated with of tumor status is highly advantageous, and liquid
advancement in stage and development of resistant biopsy offers the potential “real-time” scenario
tumor cells due to genetic heterogeneity. Although diagnosis (Batth et al., 2017; Zhang et al., 2017).
tumor biopsies (solid biopsy), traditional biomark- However, despite the numerous benefits as a
ers, and imaging techniques have yielded valuable potential diagnostic resource, clinical utility of
information for years, they are backed by several liquid biopsy suffers due to certain limitations.
drawbacks due to their invasiveness and inabil- In general there is a lack of consensus on detec-
ity to capture variable gene expressions (Batth tion methods and also difficulty in analysis of
et al., 2017). overwhelming sequencing data and lack of strong
17
18  Circulating tumor cells and circulating tumor DNA in precision medicine

proof for application in evidence-based medicine. defined by Ki67 staining that is highly variable
Discovery of this biological specimen has revealed among different patients (Fidler, 1973; Liotta et al.,
the presence of a unique cancer cell protein in 1976; Duda et al., 2010; Cho et al., 2012). Similarly,
CTCs, oncogenic mutations, and also epigenetic single-cell analyses have revealed heterogeneity
changes in CTCs and ctDNA that cause “genetic in signaling pathways among CTCs derived from
evolution of cells” to adapt to various treatment individual patients (Miyamoto et al., 2012; Powell
strategies (Batth et  al., 2017; Zhang et  al., 2017). et  al., 2012; Heitzer et  al., 2013a; Ozkumur et  al.,
Therefore, it is necessary to understand the role of 2013). On average in a metastatic carcinoma patient
CTCs and ctDNA to further knowledge of oncol- approximately 5 to 50 CTCs for every 7.5 mL of
ogy precision medicine for diagnosis, recurrence blood can be detected, which undoubtedly leads
monitoring, prognosis assessment, and medication to grave technical difficulties (Allard et  al., 2004;
planning of various kinds of cancers. Cristofanilli et  al., 2007; Ross and Slodkowska,
2009; Hou et al., 2013).
CHARACTERISTICS AND UTILITY OF
CIRCULATING TUMOR CELLS (CTCs) CTC selection approaches

General features of CTCs The primary challenge in clinical use of CTCs is

their detection due to rarity (1 cell per 1 × 109 nor-
Circulating tumor cells were discovered in the mal blood cells in patients with metastatic cancer),
1860s by Thomas Ashworth during his micro- which makes them difficult to identify and iso-
scopic examination of peripheral blood. Later, late (Pantel et  al., 2008). The simplest method of
studies suggested that solid tumor cells could isolation is based on size-based membrane filters,
break and enter the bloodstream by passive and but studies have revealed that there is consider-
active approaches (Thiery, 2002; Tsuji et al., 2008, able overlap between CTCs and leukocytes, so this
2009; Joosse and Pantel, 2013; Thiery and Lim, method would miss a large proportion of CTCs.
2013). CTCs serve as a significant prognostic fac- As a result there are multiple selection and capture
tor and it has been estimated that approximately methods that have been recently developed for
106 separating tumor cells/gm of tumor mass make their enrichment. Depending on detection prin-
contact with blood, and this strongly correlates ciple, the methods have been categorized into two
with metastasis and secondary tumor foci (Chang types: cell surface marker-dependent and marker-
et al., 2000). Thus, it is rightly said that CTCs act as independent approaches.
ambassadors of a localized disease that may remain The cell surface marker-dependent approach
dormant or become activated at a later time caus- utilizes positive selection that relies on epithelial
ing a relapse. As such thousands of tumor cells leak cell adhesion markers like EpCAM and cytokeratin
into vasculature but most get eliminated from the (CK) (Parkinson et al., 2012). The CellSearch sys-
bloodstream and have obstacles to their survival tem (Veridex, New Jersey), a technique approved
due to shearing forces of blood flow, immune cell by the U.S. Food and Drug Administration (FDA),
attack, and anoikis (Douma et al., 2004; Mitchell is the only method that has demonstrated the prog-
and King, 2013; Steinert et al., 2014). Most of them nostic value of determining CTC numbers and has
are accidental CTCs, as they are passively pushed set the benchmark for CTC isolation technolo-
by external forces like tumor growth, mechani- gies. Further, CTC-chips, having a microfluidic
cal force due to surgical operation, or other kinds platform containing antibody-coated microposts
of friction (McDonald and Baluk, 2002). Once in (against EpCAM or MUC1) have been used to
circulation they persist for a short time and their improve the capturing ability. This technology has
morphology is highly heterogeneous. Many may been regarded as highly sensitive, with good yield
be apoptotic or damaged following simple isola- accompanied by simplification of prelabeling pro-
tion techniques, whereas others appear similar in cesses (Nagrath et  al., 2007; Maheswaran et  al.,
appearance to cells from matched tumor biopsies. 2008; Thege et  al., 2014). However, the one limi-
They may also travel in clusters, ranging from two tation of this technology is in subsequent single-
CTCs undergoing mitosis or a large microemboli CTC analysis, which was overcome by replacing
having >50 cells. Proliferative index of CTCs is the microposts with an advanced chip that uses
 Characteristics and utility of circulating tumor cells (CTCs)  19

surface ridge or herringbone grooves in the ceiling the molecular properties of CTCs by analyzing the
of the channel. This was called the herringbone or protein, and RNA and DNA contents (Pantel et al.,
HBCTC-chip (Stott et al., 2010a). 2008; Yu et al., 2011). Traditional methods include
Similarly, there are tumors that lack epithelial use of fluorescent-conjugated antibodies for iden-
markers that can be enriched by use of negative tifying cells based on positive results for pan-cyto-
selection (CD45) and size-based methods like the keratin and negative for the common leukocyte
ISET system or the ScreenCell approach based antigen CD45 (Allard et al., 2004). More recently,
on density centrifugation or filtration (Powell the use of microfluidic isolation technologies has
et  al., 2012; Wu et  al., 2014; Kallergi et  al., 2016). enabled the use of high-resolution light micro-
However, in cases of CTC metastasis there is a need scopy to study various cytopathological protocols
of the EMT (epithelial to mesenchymal transition) along with standardized immunohistochemis-
process. In this, transformed tumor cells can lose try applications (Ozkumur et  al., 2013). Recently,
epithelial markers like EpCAM and CKs, and those an antibody-based method for quantification of
with epithelial markers may essentially not be the live CTCs is EPISPOT (Epithelial ImmunoSpot)
reason for cancers. Further, it is also suggested assay that captures the secreting proteins such
that rare primary tumor cells simultaneously as cathepsin D, MUC1, and CK19. Isolated CTCs
expressed mesenchymal and epithelial mark- via negative selection using anti-CD45 immu-
ers, but mesenchymal cells were highly enriched nomagnetic beads are cultured in tissue culture
in CTCs. Therefore, it is essential that CTCs be plates precoated with antibodies that capture the
detected and allocated for different phenotypes, secreted protein of interest. It can be rightly called
for example, epithelial (epithelial+/mesenchymal), “CTC protein fingerprinting,” because it differ-
complete EMT (epithelial/mesenchymal+), and entiates between apoptotic and viable CTCs, and
intermediate EMT (epithelial+/mesenchymal+) also identifies and differentiates between differ-
(Yu et al., 2013; Zhang et al., 2017). Additionally, it ent proteins within CTCs (Alix-Panabieres, 2012;
is also found that there is heterogeneity of EpCAM Alix-Panabieres and Pantel, 2013). The CTCs are
expression on the surface of CTCs that can con- also characterized for their expression of protein
tribute to variation in detection, and EpCAM markers by fluorescence microscopy (immunoflu-
methods cannot detect nonepithelial cancers such orescence to simultaneously visualize differently
as sarcomas (Allard et al., 2004). labeled targets within CTCs). Promising protein-
To distinguish epithelial from mesenchymal based identification involves dual Ki67/PSA stain-
cancer cells, a sophisticated method is RNA in ing in prostate cancer CTCs (Stott et  al., 2010b).
situ hybridization (RNA-ISH), which differentially Also dual staining for androgen-induced PSA and
stains cells according to the expression levels of androgen-suppressed prostate-specific membrane
epithelial and mesenchymal genes (Yu et al., 2013). antigen (PSMA) markers quantitate heterogene-
Yet another capture platform called CTC-iChip is ity in androgen signaling status of prostate CTCs
virtually applicable to all cancers because it helps before and after hormonal therapy (Miyamoto
in isolation of EpCAM+ and EpCAM− CTCs using et al., 2012).
a series of steps. The only limitation in its wide- Another technology is RNA-based expression
spread application is its lack of validation as com- monitoring of CTCs (successful where isolation
pared to the CellSearch system with regard to its techniques do not involve formaldehyde fixation).
specificity, reproducibility, and clinical relevance. Studies have demonstrated successful reverse-
transcription polymerase chain reaction (RT-PCR)
CTC molecular characterization amplification of lineage-specific transcripts in
and cell analysis CTC-enriched cell populations, with readily
detectable tumor-specific translocations too (e.g.,
After the capturing of CTCs they need to be char- TMPRSS2-ERG in prostate cancer and EML4-
acterized and analyzed for molecular properties. ALK in non-small cell lung cancer) (Seiden et al.,
Baseline enumeration involves counting of cells, 1994; Xi et al., 2007; Attard et al., 2009; Stott et al.,
which barely makes use of potential information 2010a; Ozkumur et al., 2013). Apart from all these
that exists within these cells and is in no way useful methods, the most recent application of the next-
for oncologists. Therefore, we need methods to study generation sequencing (NGS) in whole-genome
20  Circulating tumor cells and circulating tumor DNA in precision medicine

expression profiling has been used (Yu et al., 2012, develop resistance against a therapy and may recur
2013). Isolation of single CTCs and their tran- or spread so there is a need to identify secondary
scriptome profile has shed light on heterogeneity mutations) in cancer patients (Allard et al., 2004;
of CTCs. Further, the development of highly sen- Cristofanilli et al., 2004, 2005; de Bono et al., 2008;
sitive, quantifiable dual-colorimetric RNA-ISH Olmos et  al., 2009; Danila et  al., 2011). This can
has also contributed in understanding the epi- help patients who are unlikely to benefit from ini-
thelial versus mesenchymal transcripts (by direct tial therapy to spare them from side effects and loss
visualization of the hybridization pattern within of time (Heitzer et al., 2013b).
cells) within individual CTCs (Yu et  al., 2012, The use of a CTC-chip in patients with meta-
2013). Genotyping of CTCs is yet another explored static small-cell lung cancer to detect for EGFR
avenue where allele-specific PCR-based assays for mutation states has also been reported. Monitoring
CTC-enriched cell populations have been dem- of CTC cells revealed that the attainment of recur-
onstrated, for example, EGFR-mutant non-small rent T790M-EGFR drug resistance mutation coin-
cell lung cancer having high concordance between cided with development of clinically refractory
tumor biopsies at presentation and CTC-derived disease (Maheswaran et al., 2008). Similarly, use of
genotypes (Maheswaran et al., 2008; Attard et al., a CTC-iChip for RNA sequencing on a breast CTC
2009; Heitzer et al., 2013a; Ozkumur et al., 2013). cluster led to the discovery of plakoglobin that
The cytogenetic composition of CTCs can also helps in the maintenance of CTC clusters (Aceto
be assessed with interphase fluorescence in situ et al., 2014). Now with the availability of genome-
hybridization (FISH) that detects copy number wide analysis strategies, use of array-CGH and
changes only for genomic regions. For a more NGS can reveal all possible mechanisms of resis-
genome-wide level, whole genome amplification by tance instead of analyzing only specific previously
array-comparative genomic hybridization (array- known mutations (Heitzer et al., 2013a). In a whole
CGH) for single or pooled CTCs can be performed exome sequencing of two prostate cancer patients,
(Yu et  al., 2012; Heitzer et  al., 2013a; Magbanua CTCs revealed 70% overlap with mutations found
et  al., 2012, 2013). Recent studies also reveal that in lymph node metastasis and primary tissue (Lohr
CTC lines can be developed and kept for long- et al., 2014). CTCs also showed heterogeneity in the
term culture; these can be used to study functional TP53, PIK3CA, ESR1, and KRAS genes. Further, it is
properties for invasiveness or metastases when also suggested that we can study other relevant fea-
xenografted into nude mice (Ameri et  al., 2010; tures of the tumor genome not present or observed
Zhang et al., 2013). during earlier initial diagnosis. This was observed
in a patient where CTC examination at 34 and 24
CTC clinical and research utility months after diagnosis of primary tumor and liver
metastasis revealed high amplification of CDK8
Clinical trials for CTC count as a predictor of not found earlier (Heitzer et al., 2013a). Such criti-
overall survival (OS) and progression-free sur- cal identification has led to identification of viable
vival (PFS) have been conducted in various tis- targets for therapy (in the aforementioned case use
sues. Although studies do support the prognostic of CDK inhibitors) (Dickson et al., 2010; Wang and
potential, there is no substantial evidence of cor- Ren, 2010; Ramaswamy et al., 2012).
relation found between CTC count and patho- Another use of CTCs is to understand the pro-
logical status. However, one study suggests CTC cess of metastasis and understanding EMT pro-
counts to be viable predictors of disease relapse cess in tumor metastasis. Using an endogenous
alone and not as potential indicators of tumor mouse pancreatic cancer model, single-molecule
characteristics, lymph node positivity, therapeutic RNA sequencing from CTCs revealed enriched
response, or OS (Pierga et al., 2008). In contrast to expression of Wnt2. WNT2 resulted in increased
this, the SWOG S0500 study confirms CTC counts metastatic propensity also in human pancreatic
to have strong prognostic value for OS (Raimondi cancer cells (Yu et al., 2012). Another study using
et al., 2014; Smerage et al., 2014). Moving further xenograft assay demonstrated that primary human
from simple counting of cells, presence of CTC in luminal breast cancer CTCs contain cells that give
peripheral blood can also be used as a measure of rise to metastasis in mice in various organs. The
tracking therapeutic response (cancer cells may other interesting study is that of measuring the
 Characteristics and utility of circulating tumor DNA (ctDNA)  21

expression of mesenchymal and epithelial markers et al., 2011). Somatic genetic mutations and tumor-
in CTCs from breast cancer patients. Monitoring of specific alterations of cancers can be detected in
serial CTCs suggested an association of mesenchy- ctDNA. Therefore, ctDNA carries genomic and
mal CTCs with disease progression. Additionally, epigenomic alterations concordant to tumor muta-
in one patient reversible shifts between cell fates for tional spectrum, such as point mutations, degree
mesenchymal and epithelial cells were associated of integrity, rearranged genomic sequences, copy
with response to therapy and disease progression number variation (CNV), microsatellite instabil-
(Baccelli et al., 2013). ity (MSI), loss of heterozygosity (LOH), and DNA
methylation (Marzese et al., 2013). These biological
CHARACTERISTICS AND UTILITY characteristics differentiate ctDNA from cfDNA
OF CIRCULATING TUMOR DNA and qualify ctDNA as a biomarker that can help
(ctDNA) in precision medicine to detect residual disease or
help in monitoring of tumor progression during
General features of ctDNA therapy.

Cell-free DNA (cfDNA) is released into circulation ctDNA selection approaches

through various pathologic and normal physiologic
mechanisms. Fragments of DNA are released into Unlike CTC capture, ctDNA isolation does not
the bloodstream from dying cells during cellular depend on specialized equipment and is analyzed
turnover or from apoptotic or necrotic cells (Jahr from plasma. Plasma is preferable to serum, and
et  al., 2001; Stroun et  al., 2001b). Normal physi- should be processed and stored promptly after
ological pathways clear these cells through infil- whole blood collection in EDTA tubes to prevent
trating phagocytes and this leads to low levels of increase in cfDNA levels due to cell lysis of nor-
cfDNA. Similarly, from solid tumors, cfDNA can mal blood cells. cfDNA can then be extracted from
be released through necrosis, autophagy, and other plasma using commercially available kits, and
physiologic events induced by microenvironmen- the analysis of ctDNA can proceed using assays
tal stress and treatment pressure. However, unlike designed to detect somatic genomic aberrations.
apoptosis, necrosis generates larger DNA frag- ctDNA analysis is independent of EpCAM mark-
ments due to an incomplete and random digestion ers and reflects an average of all tumor cells releas-
of genomic DNA (Wang et al., 2003). Nevertheless, ing DNA into circulation (Heitzer et al., 2013).
not all cfDNA originate from cell death and it is
observed that live cells spontaneously release newly ctDNA molecular characterization
synthesized DNA as part of a homeostatically regu- and cell analysis
lated system (Anker et al., 1975; Stroun et al., 2000,
2001a, 2001b). Further, stimulation of lympho- Recent advances in genomics technologies are
cytes also results in the release of large amounts of paving the way for the analysis of ctDNA. NGS
cfDNA in the absence of cell death (Rogers et al., technologies are now being used for plasma DNA
1972; Rogers, 1976; Stroun et al., 2000). analysis to allow more comprehensive detection of
In cancer patients, a fraction of cfDNA is tumor mutations across wider genomic regions. Types of
derived and is termed ctDNA (derived from a pri- tumor-specific aberrations that have been detected
mary tumor, metastatic lesions or CTCs) (Jen et al., in ctDNA include somatic single nucleotide vari-
2000). Cancer patients normally have higher levels ants (SNVs), chromosomal rearrangements, and
of ctDNA than healthy individuals, but the levels epigenetic alterations (Leon et  al., 1977; Esteller
vary widely, from 0.01% to more than 90% (Diehl et  al., 1999; Silva et  al., 1999; Wong et  al., 1999;
et  al., 2008; Delgado et  al., 2013). It is found in Lecomte et al., 2002; Hanley et al., 2006; Umetani
fragment lengths in range of 140–170 bp due to its et al., 2006; Warton and Samimi, 2015). The detec-
original format as a histone-packaged nucleosome tion of SNVs in plasma DNA has been achieved by
(Pixberg et  al., 2015). The variability of ctDNA the use of a variety of PCR approaches (Nawroz
levels in cancer patients can be associated with et al., 1996; Tomita et al., 2007; Zhai et al., 2012),
tumor burden, stage, vascularity, cellular turnover, but digital PCR has now emerged as a sensitive
and response to therapy (Diehl et al., 2008; Kohler analytical tool for the detection of mutations at low
22  Circulating tumor cells and circulating tumor DNA in precision medicine

allele fractions (Diaz and Bardelli, 2014). Methods to concerns over false-negative tests (Batth et  al.,
involving the use of digital PCR include droplet- 2017). However, efforts are underway to make
based systems; microfluidic platforms; and the use use of ctDNA testing for reliable results due to its
of beads, emulsions, amplification, and magnetics higher turnaround time and specimen availability.
(BEAMing) (Yoshimasu et al., 1999; Ito et al., 2002; Tumor-specific DNA levels can be used as a sur-
Hanley et  al., 2006; Heitzer et  al., 2013b; Lipson rogate of treatment response. Therefore, ctDNA
et al., 2014); Targeted deep sequencing using PCR- analysis can act as a biomarker to reveal the tumor
based (e.g., TAm-Seq, Safe-Seq, Ion AmpliSeq™) or burden in a patient by measuring the amount of
capture-based (e.g., CAPP-seq) approaches have DNA released, as dying tumor cells will increase
been used to sequence specified genomic regions in the DNA release into circulation during treat-
plasma DNA (Leon et al., 1977; Vasioukhin et al., ment (McBride et al., 2010; Haber and Velculescu,
1994; Shinozaki et  al., 2007; Heyn and Esteller, 2014). Additionally, it can also assist in detection
2012; Schwarzenbach et  al., 2012). Additionally, of early relapse, that is, the primary tumor can be
whole-exome analysis of plasma DNA has opened sequenced for identifying tumor-specific “driver”
up new opportunities to carefully characterize or “passenger” translocations that can help in
mutation profiles, without the need to focus on sensitive monitoring for early tumor recurrence
predefined or existing mutations (Esteller et  al., (Leary et al., 2010, 2012; Sausen et al., 2013).
1999). Chromosomal rearrangements such as
translocations or gains/losses of chromosomal DISCUSSION AND CONCLUSION
regions can also be detected in ctDNA, provid-
ing excellent sensitivity and specificity as tumor CTCs and ctDNA are capable of providing snap-
biomarkers (Umetani et  al., 2006; Bidard et  al., shots of genomic alterations of tumors. Both
2014). Personalized analysis of rearrangement ends specimens have revolutionized the blood-based
(PARE) is a method that involves the identification diagnostics in clinical oncology owing to their
of specific somatic rearrangements in tumor tissue multiple applications through different stages of
and the subsequent design of PCR-based assays cancers. CTCs use highly selective approaches and
to detect these alterations in plasma DNA (Leon represent pure tumor cell population and com-
et  al., 1977). Moreover, in selected cases, whole- bined with WGA and NGS provide unique insights
genome sequencing has now been directly applied into tumors and their evolution at various stages of
to plasma DNA analysis to view somatic chromo- progressive cancers. Similarly, ctDNA offers a more
somal alterations and copy number aberrations easily obtainable specimen and analysis of real-time
in ctDNA genome-wide (Silva et  al., 1999; Wong DNA content from tumors. Therefore, it can be said
et al., 1999; Kawakami et al., 2000). that liquid biopsy has an important contribution in
the field of precision medicine by offering a non-
ctDNA clinical utility invasive cancer monitoring technology. Although
with certain limitations in clinical care as of now,
ctDNA analysis involves ease of collection, higher it is more sensitive, accurate, and provides substan-
levels as compared to CTCs, and better medium tially more information when compared with tradi-
for high-throughput analysis. Owing to this, geno- tional detection and monitoring approaches.
typing of ctDNA can be said to be rapid, economi- Focus now should be on using the combined
cal and reliable for clinical applications (Sausen knowledge from these specimens and the need
et al., 2013; Bettegowda et al., 2014). Use of ctDNA for larger research studies. Although many stud-
in clinical care has been tested in cases where the ies have been carried out to examine their clinical
presence of mutations is associated with response utility most of them were retrospective. Therefore,
to targeted therapy. In Europe, regulatory labeling to validate their potential, there is a need for rigor-
of Gefitinib was updated to allow use of ctDNA ous clinical research. As these technologies mature
for assessment of activating EGFR mutations in over the years, there is hope of better diagnosis,
patients where a tumor sample is not available treatment, and management of cancers.
for tests. Similarly, ctDNA use in KRAS or NRAS In conclusion we can say that both biological spec-
mutations for EGFR inhibitors in colorectal can- imens need to be studied independently and simulta-
cer proved slower to be adopted in clinical care due neously in their applications to clinical oncology and
 References 23

precision medicine. Tumors evolve with time, and, Bettegowda C, Sausen M, Leary RJ et al.
hence, there is a need for repeated sampling to view Detection of circulating tumor DNA in early-
the real-time tumor status for patient management. and late-stage human malignancies. Sci Transl
Both specimens offer their particular capabilities Med 6;2014:224ra24.
and limitations and with time they might become Bidard FC, Madic J, Mariani P et al. Detection
essential tools for better cancer management. rate and prognostic value of circulating
tumor cells and circulating tumor DNA in
REFERENCES metastatic uveal melanoma. Int J Cancer
134;2014:1207–13.
Aceto N, Bardia A, Miyamoto DT et al. Chang YS, di Tomaso E, McDonald DM, Jones R,
Circulating tumor cell clusters are oligoclonal Jain RK, and Munn LL. Mosaic blood vessels
precursors of breast cancer metastasis. Cell in tumors: Frequency of cancer cells in con-
158;2014:1110–22. tact with flowing blood. Proc Natl Acad Sci
Alix-Panabieres C. EPISPOT assay: Detection of USA 97;2000:14608–13.
viable DTCs/CTCs in solid tumor patients. Cho EH, Wendel M, Luttgen M et al.
Recent Results Cancer Res 195;2012:69–76. Characterization of circulating tumor cell
Alix-Panabieres C, and Pantel K. Circulating aggregates identified in patients with epithe-
tumor cells: Liquid biopsy of cancer. Clin lial tumors. Phys Biol 9;2012:016001.
Chem 59;2013:110–8. Cristofanilli M, Broglio KR, Guarneri V et al.
Allard WJ, Matera J, Miller MC, Repollet M, Circulating tumor cells in metastatic breast
Connelly MC, Rao C, Tibbe AG, Uhr JW, and cancer: Biologic staging beyond tumor bur-
Terstappen LW. Tumor cells circulate in the den. Clin Breast Cancer 7;2007:471–9.
peripheral blood of all major carcinomas Cristofanilli M, Budd GT, Ellis MJ et al. Circulating
but not in healthy subjects or patients with tumor cells, disease progression, and survival
nonmalignant diseases. Clin Cancer Res in metastatic breast cancer. N Engl J Med
10;2004:6897–904. 351;2004:781–91.
Ameri K, Luong R, Zhang H, Powell AA, Cristofanilli M, Hayes DF, Budd GT et al.
Montgomery KD, Espinosa I, Bouley DM, Circulating tumor cells: A novel prognostic
Harris AL, and Jeffrey SS. Circulating tumour factor for newly diagnosed metastatic breast
cells demonstrate an altered response to cancer. J Clin Oncol 23;2005:1420–30.
hypoxia and an aggressive phenotype. Br J Danila DC, Fleisher M, and Scher HI. Circulating
Cancer 102;2010:561–9. tumor cells as biomarkers in prostate cancer.
Anker P, Stroun M, and Maurice PA. Spontaneous Clin Cancer Res 17;2011:3903–12.
release of DNA by human blood lymphocytes de Bono JS, Scher HI, Montgomery RB, Parker C,
as shown in an in vitro system. Cancer Res Miller MC, Tissing H, Doyle GV, Terstappen LW,
35;1975:2375–82. Pienta KJ, and Raghavan D. Circulating tumor
Attard G, Swennenhuis JF, Olmos D et al. cells predict survival benefit from t­ reatment in
Characterization of ERG, AR and PTEN gene metastatic castration-resistant prostate cancer.
status in circulating tumor cells from patients Clin Cancer Res 14;2008:6302–9.
with castration-resistant prostate cancer. Delgado PO, Alves BC, Gehrke Fde S et al.
Cancer Res 69;2009:2912–8. Characterization of cell-free circulating DNA
Baccelli I, Schneeweiss A, Riethdorf S et al. in plasma in patients with prostate cancer.
Identification of a population of blood circu- Tumour Biol 34;2013:983–6.
lating tumor cells from breast cancer patients Diaz LA Jr, and Bardelli A. Liquid biopsies:
that initiates metastasis in a xenograft assay. Genotyping circulating tumor DNA. J Clin
Nat Biotechnol 31;2013:539–44. Oncol 32;2014:579–86.
Batth IS, Mitra A, Manier S, Ghobrial IM, Menter Dickson MA, Shah MA, Rathkopf D et al. A phase
D, Kopetz S, and Li S. Circulating tumor mark- I clinical trial of FOLFIRI in combination with
ers: Harmonizing the Yin and Yang of CTCs the pan-cyclin-dependent kinase (CDK) inhibi-
and ctDNA for precision medicine. Ann Oncol tor flavopiridol. Cancer Chemother Pharmacol
28;2017:468–77. 66;2010:1113–21.
24  Circulating tumor cells and circulating tumor DNA in precision medicine

Diehl F, Schmidt K, Choti MA et al. Circulating for overlooked synchronous metastases

mutant DNA to assess tumor dynamics. Nat in colorectal carcinoma. Jpn J Clin Oncol
Med 14;2008:985–90. 32;2002:54–8.
Douma S, Van Laar T, Zevenhoven J, Meuwissen Jahr S, Hentze H, Englisch S et al. DNA frag-
R, Van Garderen E, and Peeper DS. ments in the blood plasma of cancer patients:
Suppression of anoikis and induction of Quantitations and evidence for their origin
metastasis by the neurotrophic receptor trkb. from apoptotic and necrotic cells. Cancer Res
Nature 430;2004:1034–9. 61;2001:1659–65.
Duda DG, Duyverman AM, Kohno M et al. Jen J, Wu L, and Sidransky D. An overview on
Malignant cells facilitate lung metastasis by the isolation and analysis of circulating tumor
bringing their own soil. Proc Natl Acad Sci DNA in plasma and serum. Ann NY Acad Sci
USA 107;2010:21677–82. 906;2000:8–12.
Esteller M, Sanchez-Cespedes M, Rosell R, Joosse SA, and Pantel K. Biologic challenges
Sidransky D, Baylin SB, and Herman JG. in the detection of circulating tumor cells.
Detection of aberrant promoter hypermeth- Cancer Res 73;2013:8–11.
ylation of tumor suppressor genes in serum Kallergi G, Politaki E, Alkahtani S, Stournaras C,
DNA from non-small cell lung cancer patients. and Georgoulias V. Evaluation of isolation
Cancer Res 59;1999:67–70. methods for circulating tumor cells (ctcs). Cell
Fidler IJ. The relationship of embolic homogene- Physiol Biochem 40;2016:411–9.
ity, number, size and viability to the incidence Kawakami K, Brabender J, Lord RV et al.
of experimental metastasis. Eur J Cancer Hypermethylated APC DNA in plasma
9;1973:223–7. and prognosis of patients with esopha-
Haber DA, and Velculescu VE. Blood-based geal adenocarcinoma. J Natl Cancer Inst
analysis of cancer: Circulating tumor cells and 92;2000:1805–11.
circulating tumor DNA. Cancer Discovery Kohler C, Barekati Z, Radpour R, and Zhong
4;2014:650–61. XY. Cell-free DNA in the circulation as a
Hanley R, Rieger-Christ KM, Canes D et al. DNA potential cancer biomarker. Anticancer Res
integrity assay: A plasma-based screening 31;2011:2623–8.
tool for the detection of prostate cancer. Clin Leary RJ, Kinde I, Diehl F et al. Development of
Cancer Res 12;2006:4569–74. personalized tumor biomarkers using mas-
Heitzer E, Auer M, Gasch C et al. Complex tumor sively parallel sequencing. Sci Transl Med
genomes inferred from single circulating 2;2010:20ra14.
tumor cells by array-CGH and next-generation Leary RJ, Sausen M, Kinde I et al. Detection of
sequencing. Cancer Res 73;2013a:2965–75. chromosomal alterations in the circulation of
Heitzer E, Auer M, Ulz P, Geigl JB, and Speicher cancer patients with whole-genome sequenc-
MR. Circulating tumor cells and DNA as liquid ing. Sci Transl Med 4;2012:162ra54.
biopsies. Genome Med 5;2013:73. Lecomte T, Berger A, Zinzindohoue F et al.
Heitzer E, Ulz P, Belic J et al. Tumor-associated Detection of free-circulating tumor-asso-
copy number changes in the circulation of ciated DNA in plasma of colorectal cancer
patients with prostate cancer identified patients and its association with prognosis.
through whole-genome sequencing. Genome Int J Cancer 100;2002:542–8.
Med 5;2013b:30. Leon SA, Shapiro B, Sklaroff DM, and Yaros MJ.
Heyn H, and Esteller M. DNA methylation profil- Free DNA in the serum of cancer patients
ing in the clinic: Applications and challenges. and the effect of therapy. Cancer Res
Nat Rev Genet 13;2012:679–92. 37;1977:646–50.
Hou HW, Warkiani ME, Khoo BL, Li ZR, Soo RA, Liotta LA, Saidel MG, and Kleinerman J. The
Tan DS, Lim WT, Han J, Bhagat AA, and Lim CT. significance of hematogenous tumor cell
Isolation and retrieval of circulating tumor cells clumps in the metastatic process. Cancer Res
using centrifugal forces. Sci Rep 3;2013:1259. 36;1976:889–94.
Ito K, Hibi K, Ando H et al. Usefulness of ana- Lipson EJ, Velculescu VE, Pritchard TS et al.
lytical CEA doubling time and half-life time Circulating tumor DNA analysis as a real-time
 References 25

method for monitoring tumor burden in mela- Olmos D, Arkenau HT, Ang JE et al. Circulating
noma patients undergoing treatment with tumour cell (CTC) counts as intermediate end
immune checkpoint blockade. J Immunother points in castration-resistant prostate cancer
Cancer 2;2014:42. (CRPC): A single-centre experience. Ann Oncol
Lohr JG, Adalsteinsson VA, Cibulskis K et al. 20;2009:27–33.
Whole-exome sequencing of circulat- Ozkumur E, Shah AM, Ciciliano JC et al. Inertial
ing tumor cells provides a window into focusing for tumor antigen-dependent and
metastatic prostate cancer. Nat Biotechnol -independent sorting of rare circulating tumor
32;2014:479–84. cells. Sci Transl Med 5;2013:179ra47.
Magbanua MJ, Sosa EV, Roy R, Eisenbud LE, Pantel K, Brakenhoff RH, and Brandt B.
Scott JH, Olshen A, Pinkel D, Rugo HS, Detection, clinical relevance and specific
and Park JW. Genomic profiling of isolated biological properties of disseminating tumour
circulating tumor cells from metastatic breast cells. Nat Rev Cancer 8;2008:329–40.
cancer patients. Cancer Res 73;2013:30–40. Parkinson DR, Dracopoli N, Petty BG et al.
Magbanua MJ, Sosa EV, Scott JH, Simko J, Considerations in the development of circu-
Collins C, Pinkel D, Ryan CJ, and Park JW. lating tumor cell technology for clinical use.
Isolation and genomic analysis of circulat- J Transl Med 10;2012:138.
ing tumor cells from castration resistant Pierga J-Y, Bidard F-C, Mathiot C, Brain E,
metastatic prostate cancer. BMC Cancer Delaloge S, Giachetti S, de Cremoux P,
12;2012:78. Salmon R, Vincent-Salomon A, and Marty
Maheswaran S, Sequis LV, Nagrath S et al. M. Circulating tumor cell detection predicts
Detection of mutations in EGFR in circu- early metastatic relapse after neoadjuvant
lating lung-cancer cells. N Engl J Med chemotherapy in large operable and locally
359;2008:366–77. advanced breast cancer in a phase II random-
Marzese DM, Hirose H, and Hoon DS. Diagnostic ized trial. Clin Cancer Res 14;2008:7004–10.
and prognostic value of circulating tumor- Pixberg C, Schulz W, Stoecklein N, and Neves R.
related DNA in cancer patients. Expert Rev Characterization of DNA methylation in circulat-
Mol Diagn 13;2013:827–44. ing tumor cells. Genes (Basel) 6;2015:1053–75.
McBride DJ, Orpana AK, Sotiriou C et al. Use Powell AA, Talasaz AH, Zhang H et al. Single
of cancer-specific genomic rearrange- cell profiling of circulating tumor cells:
ments to quantify disease burden in plasma Transcriptional heterogeneity and diversity
from patients with solid tumors. Genes from breast cancer cell lines. PLOS ONE
Chromosomes Cancer 49;2010:1062–9. 7;2012:e33788.
McDonald DM, and Baluk P. Significance of Raimondi C, Gradilone A, Naso G et al. Clinical
blood vessel leakiness in cancer. Cancer Res utility of circulating tumor cell counting
62;2002:5381–5. through CellSearch (VR): The dilemma of a
Mitchell MJ, and King MR. Computational and concept suspended in limbo. Onco Targets
experimental models of cancer cell response Ther 7;2014:619–25.
to fluid shear stress. Front Oncol 3;2013:44. Ramaswamy B, Phelps MA, Baiocchi R et al.
Miyamoto DT, Lee RJ, Stott SL et al. Androgen A dose-finding, pharmacokinetic and phar-
receptor signaling in circulating tumor cells as macodynamic study of a novel schedule of
a marker of hormonally responsive prostate flavopiridol in patients with advanced solid
cancer. Cancer Discov 2;2012:995–1003. tumors. Invest New Drugs 30;2012:629–38.
Nagrath S, Sequist LV, Maheswaran S et al. Rogers JC. Identification of an intracellular pre-
Isolation of rare circulating tumour cells in cursor to DNA excreted by human lympho-
cancer patients by microchip technology. cytes. Proc Natl Acad Sci USA 73;1976:3211.
Nature 450;2007:1235–9. Rogers JC, Boldt D, Kornfeld S, Skinner A, and
Nawroz H, Koch W, Anker P, Stroun M, and Valeri CR. Excretion of deoxyribonucleic acid
Sidransky D. Microsatellite alterations in by lymphocytes stimulated with phytohemag-
serum DNA of head and neck cancer patients. glutinin or antigen. Proc Natl Acad Sci USA
Nat Med 2;1996:1035–7. 69;1972:1685–9.
26  Circulating tumor cells and circulating tumor DNA in precision medicine

Ross JS, and Slodkowska EA. Circulating and Stroun M, Lyautey J, Lederrey C, Mulcahy
disseminated tumor cells in the manage- HE, and Anker P. Alu repeat sequences
ment of breast cancer. Am J Clin Pathol are ­present in increased proportions
132;2009:237–45. compared to a unique gene in plasma/
Sausen M, Leary RJ, Jones S et al. Integrated serum DNA: Evidence for a preferential
genomic analyses identify ARID1A and release from ­viable cells? Ann NY Acad Sci
ARID1B alterations in the childhood cancer 945;2001a:258–64.
neuroblastoma. Nat Genet 45;2013:12–7. Stroun M, Lyautey J, Lederrey C, Olson-Sand
Schwarzenbach H, Eichelser C, Kropidlowski J, A, and Anker P. About the possible origin
Janni W, Rack B, and Pantel K. Loss of hetero- and mechanism of circulating DNA apopto-
zygosity at tumor suppressor genes detect- sis and active DNA release. Clin Chim Acta
able on fractionated circulating cell-free 313;2001b:139–42.
tumor DNA as indicator of breast cancer pro- Stroun M, Maurice P, Vasioukhin V et al. The ori-
gression. Clin Cancer Res 18;2012:5719–30. gin and mechanism of circulating DNA. Ann
Seiden MV, Kantoff PW, Krithivas K et al. NY Acad Sci 906;2000:161–8.
Detection of circulating tumor cells in men Thege FI, Lannin TB, Saha TN, Tsai S, Kochman
with localized prostate cancer. J Clin Oncol ML, Hollingsworth MA, Rhim AD, and Kirby
12;1994:2634–9. BJ. Microfluidic immunocapture of circulat-
Sestini S, Boeri M, Marchiano A et al. Circulating ing pancreatic cells using parallel epcam and
micro RNA signature as liquid-biopsy to muc1 capture: Characterization, optimiza-
monitor lung cancer in low-dose com- tion and downstream analysis. Lab Chip
puted tomography screening. Oncotarget 14;2014:1775–84.
6;2015:32868–77. Thiery JP. Epithelial-mesenchymal transitions
Shinozaki M, O’Day SJ, Kitago M, Amersi F, Kuo in tumour progression. Nat Rev Cancer
C, and Kim J. Utility of circulating B-RAF DNA 2;2002:442–54.
mutation in serum for monitoring melanoma Thiery JP, and Lim CT. Tumor dissemination: An
patients receiving biochemotherapy. Clin EMT affair. Cancer Cell 23;2013:272–3.
Cancer Res 13;2007:2068–74. Tomita H, Ichikawa D, Ikoma D et al.
Silva JM, Dominguez G, Villanueva MJ, Gonzalez Quantification of circulating plasma DNA
R, Garcia JM, Corbacho C, Provencio M, fragments as tumor markers in patients
España P, and Bonilla F. Aberrant DNA with esophageal cancer. Anticancer Res
methylation of the p16INK4a gene in plasma 27;2007:2737–41.
DNA of breast cancer patients. Br J Cancer Tsuji T, Ibaragi S, and Hu GF. Epithelial-
80;1999:1262–4. mesenchymal transition and cell cooperativity
Smerage JB, Barlow WE, Hortobagyi GN et al. in metastasis. Cancer Res 69;2009:7135–9.
Circulating tumor cells and response to che- Tsuji T, Ibaragi S, Shima K, Hu MG, Katsurano M,
motherapy in metastatic breast cancer: SWOG Sasaki A, and Hu GF. Epithelial-mesenchymal
S0500. J Clin Oncol 32(31);2014:1–8. transition induced by growth suppressor
Steinert G, Scholch S, Niemietz T et al. Immune p12cdk2-ap1 promotes tumor cell local inva-
escape and survival mechanisms in circulating sion but suppresses distant colony growth.
tumor cells of colorectal cancer. Cancer Res Cancer Res 68;2008:10377–86.
74;2014:1694–704. Umetani N, Giuliano AE, Hiramatsu SH et al.
Stott SL, Hsu CH, Tsukrov DI et al. Isolation of Prediction of breast tumor progression by
circulating tumor cells using a microvortex- integrity of free circulating DNA in serum.
generating herringbone-chip. Proc Natl Acad J Clin Oncol 24;2006:4270–6.
Sci USA 107;2010a:18392–7. Vasioukhin V, Anker P, Maurice P, Lyautey J,
Stott SL, Lee RJ, Nagrath S et al. Isolation Lederrey C, and Stroun M. Point mutations of
and c­ haracterization of circulating tumor the N-ras gene in the blood plasma DNA of
cells from patients with localized and patients with myelodysplastic syndrome or
metastatic prostate cancer. Sci Transl Med acute myelogenous leukaemia. Br J Haematol
2;2010b:25ra23. 86;1994:774–9.
 References 27

Wang BG, Huang HY, Chen YC et al. Increased Yu M, Bardia A, Wittner BS et al. Circulating
plasma DNA integrity in cancer patients. breast tumor cells exhibit dynamic changes
Cancer Res 63;2003:3966–8. in epithelial and mesenchymal composition.
Wang LM, and Ren DM. Flavopiridol, the first Science 339;2013:580–4.
cyclin-dependent kinase inhibitor: Recent Yu M, Stott S, Toner M, Maheswaran S,
advances in combination chemotherapy. Mini and Haber DA. Circulating tumor cells:
Rev Med Chem 10;2010:1058–70. Approaches to isolation and characterization.
Warton K, and Samimi G. Methylation of cell-free J Cell Biol 192;2011:373–82.
circulating DNA in the diagnosis of cancer. Yu M, Ting DT, Stott SL et al. RNA sequencing
Front Mol Biosci 2;2015:13. of pancreatic circulating tumour cells impli-
Wong IH, Lo YM, Zhang J et al. Detection of aber- cates WNT signalling in metastasis. Nature
rant p16 methylation in the plasma and serum of 487;2012:510–3.
liver cancer patients. Cancer Res 59;1999:71–3. Zhai R, Zhao Y, Su L, Cassidy L, Liu G, and
Wu S, Liu Z, Liu S, Lin L, Yang W, and Xu J. Christiani DC. Genome-wide DNA
Enrichment and enumeration of circulating ­methylation profiling of cell-free serum
tumor cells by efficient depletion of leukocyte DNA in esophageal adenocarcinoma
fractions. Clin Chem Lab Med 52;2014:243–51. and Barrett esophagus. Neoplasia
Xi L, Nicastri DG, El-Hefnawy T, Hughes SJ, 14;2012:29–33.
Luketich JD, and Godfrey TE. Optimal mark- Zhang L, Ridgway LD, Wetzel MD, Ngo J, Yin
ers for real-time quantitative reverse tran- W, Kumar D, Goodman JC, Groves MD, and
scription PCR detection of circulating tumor Marchetti D. The identification and char-
cells from melanoma, breast, colon, esopha- acterization of breast cancer CTCs com-
geal, head and neck, and lung cancers. Clin petent for brain metastasis. Sci Transl Med
Chem 53;2007:1206–15. 5;2013:180ra148.
Yoshimasu T, Maebeya S, Suzuma T et al. Zhang W, Xia W, Lv Z, Xin Y, Ni C, and Yang L.
Disappearance curves for tumor markers after Liquid biopsy for cancer: Circulating tumor
resection of intrathoracic malignancies. Int J cells, circulating free DNA or exosomes? Cell
Biol Markers 14;1999:99–105. Physiol Biochem 41;2017:755–68.
 3
Oral squamous cell carcinoma
and liquid biopsies: A new tool for
precision oncology

LAURA ÁLVAREZ RODRÍGUEZ, LAURA MUINELO ROMAY,

ABEL GARCÍA GARCÍA, RAFAEL LÓPEZ-LÓPEZ, ALICIA ÁBALO PIÑEIRO,
AND MARIO PÉREZ-SAYÁNS

Oral squamous cell carcinoma (OSCC) 29 have the potential to improve the health
Circulating tumor cells (CTCs): Biological of OSCC patients? 34
relevance and technological development 31 Challenges and research opportunities for
Why should the study of CTCs be included the clinical inclusion of CTC analysis to
in precision medicine for OSCC patients? 33 manage OSCC 35
Why is personalized and precision medicine Conclusion 36
through CTC study necessary? Does it References 36

ORAL SQUAMOUS CELL advanced-aged adults in a higher frequency (from

CARCINOMA (OSCC) the fourth decade onward, with a maximum peek in
the 60s). However, a preoccupying number of neo-
Head and neck cancers are those malignant tumors plasms have been detected in young adults (under
located in the area of the paranasal sinus, the naso- 45 years). This disease is more prevalent in men, the
pharynx, oropharynx (tonsils, soft palate, base of proportion being 2:1 (men:women), although these
the tongue), hypopharynx, larynx, oral cavity (oral figures are matching up since women are acquiring
mucosa, gums, hard palate, tongue, and floor of toxic habits similar to men (Ragin et al., 2007).
the mouth), and salivary glands (Barnes, 2005). The etiology of this type of cancer is multifactor;
The term oral cancer is used as a synonym for the main risk factors are smoking and alcohol. In
oral squamous cell carcinoma (OSCC), which some cases, vitamin-poor diets (A and C, mainly),
accounts for 90% of all head and neck cancers and poor oral hygiene, infections such as Epstein-Barr
3%–4% of malignancies (Bagan et  al., 2010). It is or HPV 16 and 18, radiation, and the presence of
the sixth cancer in the world in terms of incidence genetic factors have proven to be relevant (Lifelong
(approximately half a million people per year). The Learning Programme, 2015).
disease is located among the main types of solid The most commonly used system to discover
cancers and is responsible for over 65,000 deaths in the extension of this type of head and neck can-
Europe every year (Wikner et al., 2014). The inci- cer is TNM, which refers to the tumor size (T), the
dence increases with age and affects middle- to extension of the propagation to the regional lymph

29
30  Oral squamous cell carcinoma and liquid biopsies: A new tool for precision oncology

nodes (N), and the development of metastasis (M) patients suffer local recurrence and 25% develop
(Barnes, 2005). long-distance metastasis (Figure 3.1a–d).
Patients diagnosed with this type of cancer are However, prognosis of these patients still
treated by surgery, radiotherapy, chemotherapy, or a remains poor (Zbären and Lehmann, 1987;
combination of these three treatments. The choice of Slootweg et al., 1992; Wheeler et al., 2014). Although
treatment depends on the location of the tumor, its the most important prognostic indicator of recur-
extension, histological subtype, tumoral stage, and rence is metastasis of the lymph glands in the neck,
the patient’s general condition. Surgery and radio- the incidence of distant metastasis has dramati-
therapy are used independently to treat cases of cally increased (Shah, 1990; Forastiere et al., 2001).
nonmetastatic disease (stages I and II), but cancers Despite the significant advances in diagnostics
in advanced stages (III and IV) require concomi- (Yongkui et al., 2013) and the therapeutic options
tant radiotherapy and chemotherapy. One third of (Teymoortash and Werner, 2012), the 5-year sur-
patients have tumors in their initial phases (stages vival rate has remained stable at approximately 50%
I and II), while the other two thirds show advanced throughout the last decades and in all tumor phases
stage tumors (stages III and IV). Over 50% of cancer (Cohen et al., 2004; Cooper et al., 2004). Although

(a) (b)

9.7 mm (2D)
8.9 mm (2D)
46.5 mm (2D)

(c) (d)

Figure 3.1  (a) Woman, 60 years old, with an OSCC located on the base of the tongue. (b) Same
woman with cervical metastasis some months later. (c and d) Male patients (71 and 97 years, respec-
tively) who, after presenting an OSCC in the floor of the mouth and lower gum, developed metastasis,
cervical for the first one and thoracic for the second one.
 Circulating tumor cells (CTCs): Biological relevance and technological development  31

MET
Basal
lamina

CTCs

Micrometastasis
Primary
tumor
EMT
Macrometastasis

Figure 3.2  Process by which tumor cells leave the primary tumor, invade blood vessels, and are then able
to extravasate to target organs and form metastases. In this process the cells begin to show the phenom-
enon of epithelial–mesenchymal transition (EMT) by which they lose their adhesion properties and acquire
a migratory phenotype. Once they are in the blood, these cells must be able to survive and be implanted
into a new tissue, so a process called mesenchymal–epithelial transition (MET) is believed necessary.

currently fewer patients suffer from locoregional 2004) (Figure 3.2). But the appearance of the dis-
recurrence, several have developed distant disease seminated disease has dramatic consequences in
due to hidden cervical metastasis. These microme- cancer evolution.
tastasis seem to contribute to an increase in mor- The presence of CTCs was first described by
tality and morbidity (Partridge, 1999; Psychogios Ashworth in 1869 (Ashworth, 1869; Cristofanilli
et al., 2013). Therefore, better methods are needed et al., 2005) for women with breast cancer, but due
for early detection of metastatic spreading and resid- to the lack of sensitive cytology techniques to detect
ual tumors, and for the decision-making process in these cells, their analysis had not shown clinical
terms of individual therapeutic interventions. impact until the past two decades. One of the main
problems for CTC analysis in blood is the low num-
CIRCULATING TUMOR CELLS bers compared to other cell types. The frequency is
(CTCs): BIOLOGICAL RELEVANCE estimated in approximately 1 CTC per 106 nucleated
AND TECHNOLOGICAL cells in the blood of a metastatic patient; therefore
DEVELOPMENT we require highly sensitive isolation techniques for
their analysis (Alix-Panabières, 2012). To solve this
Distant metastasis is mainly the result of the hema- problem, different enrichment strategies have been
togenous spreading of tumor cells from a primary developed to recover the CTCs present in the blood,
tumor. It is known that primary tumor cells spread based on the differential properties of these cells
to other areas of the organism through the blood- compared to the other cells present in the blood-
stream and the lymphatic system, which are called stream (Figure 3.3). To date, there are more than
disseminated tumor cells (DTCs), and more spe- 40 published technologies for the isolation of these
cifically those traveling through the bloodstream, cells, and numerous research groups are working
circulating tumor cells (CTCs). When these cells to increase the list of devices every day (Parkinson
reach other organs they may settle and develop a et al., 2012). The techniques for CTC isolation can
new tumor, depending on the conditions and the be divided into two groups: those based on physical
microenvironment of these new areas. parameters such as cell size, density, deformability,
This process is considered very inefficient. It or electric properties; and those based on biologi-
is estimated that only 1 of each 105–106 CTCs in cal characteristics, such as invasion capacity or the
peripheral blood enter into tissues at a distance expression of surface antigens (Partridge, 1999).
from the primary tumor and only a small percent- From among the methods based on bio-
age of these cells develop a metastasis (Ring et al., logic properties, CellSearch System (Janssen
32  Oral squamous cell carcinoma and liquid biopsies: A new tool for precision oncology

Blood
vessel
Venipuncture
Inmunohistochemistry
Enrichment strategies inmunofluorescence
flow cytometry

Physical parameters Biological characteristics

Filtration Density Invasive properties Magnetic
nanoparticles

RT-PCR
ISET TM R Ficoll-hypaque Vita assay Cell search system oRT-PCR

Figure 3.3  Different preanalytic enrichment strategies for further analysis of CTCs.

Diagnostics) is the most used. This system captures based on the results obtained in metastatic patients
CTCs using magnetic nanoparticles conjugated with colon, breast, and prostate cancer (Cristofanilli
with antibodies targeting the EpCAM surface et al., 2004; de Bono et al., 2008; Cohen et al., 2009).
antigen (immunomagnetic separation of EpCAM- The highest limitation to this technology is the low
positive tumor cells). Upon isolation, immunoflu- efficiency in certain tumors that do not express
orescence is developed to detect the expression of sufficient EpCAM levels and the low purity of the
epithelial cytokeratins (CK8, 18 and 19) and CD45 obtained CTCs, which condition the performance
(which is only expressed in leukocytes) (Villegas of postisolation molecular studies.
et  al., 2011). Following this step, an automated Another commercial system based on immuno-
system analyzes the data through florescence isolation is CellCollector (GILUPI). This medical
microscopy. Nucleotic cells >4 μm in size, with device has shown high specificity and sensitivity for
positive marking for cytokeratins and negative for in vivo isolation of CTCs from patients with breast
CD45, are considered as CTCs, always taking into and microcytic lung cancer. CellCollector con-
account that the detection limit of the CellSearch is sists of a medical wire, functionalized with anti-
about 1 CTC for every 105–107 mononuclear cells EpCAM antibodies, that is inserted for 30 minutes
(Figure 3.4). into a peripheral vein, allowing the isolation of
This is the only system approved for clinical use CTCs from a higher volume of blood than other
by the U.S. Food and Drug Administration (FDA), conventional methods (Saucedo-Zeni et al., 2012).

CellSearch System DAPI/CK-PE CK-PE DAPI

Figure 3.4  Images of the CellSearch System and the CTCs isolated from a metastatic cancer patient
using this technology.
 Why should the study of CTCs be included in precision medicine for OSCC patients?  33

The advantage of the methods based on physi- especially prognostic and therapy-monitoring mark-
cal properties to isolate CTCs compared to immu- ers that allow the development of a personalized and
nological techniques is that these may be applied more efficient treatment. Quantification and charac-
regardless of the antigenic tumor phenotype. terization of CTCs in patients with this tumor could
Different systems are based on the size of the CTCs be an invaluable tool to evaluate the state of the dis-
for their isolation, such as ISET, ScreenCell, or ease in a dynamic and noninvasive way.
Metacell (Vona et al., 2000; Paterlini-Brechot and CTC analysis, with its potential to inform us of
Benali, 2007; Kolostova et al., 2015). These systems the tumor burden, is a prognostic factor that has
are based on the use of porous membranes (pores been validated for multiple tumors (Cristofanilli
normally <8 µm) in which tumor cells larger than et al., 2004; de Bono et al., 2008; Cohen et al., 2008;
10 µm are retained. The downside of this strategy, Hou et al., 2009). In advanced head and neck tumors,
once again, is the low purity of the CTCs after their a study developed using the CellSearch system by
isolation. There are also methods based on centrif- Bozec et al. (2013) detected CTCs in 16% of patients.
ugation with density gradients such as Oncoquick, In locally advanced head and neck squamous
using Ficoll to retain cells because of their high carcinomas, Tinhofer et al. (2014) detected CTCs
density (Rosenberg et al., 2002). in 29% of the 144 patients under study. The num-
More recently, numerous systems based on ber of CTCs was higher in cases with lymph node
microfluids have been developed. Such systems affectation and location in the tonsils or the base
can combine both physical and biological strate- of the tongue, regardless of the papillomavirus sta-
gies. These systems improve the purity of the iso- tus or smoking habits. The presence of lymphatic
lation of systems such as CellSearch. CTC-chip, metastasis is currently the most important prog-
for example, isolated EpCAM-positive cells with nostic factor affecting the survival of head and
higher efficiency than CellSearch passing the sam- neck cancer patients (survival can be reduced by
ple through a chip with microcolumns functional- 50% in the N1 stages) (Snow et al., 1982; Leemans
ized with anti-EpCAM antibodies (Nagrath et al., et  al., 1993; Jones et  al., 1994); a significant per-
2007; Maheswaran et al., 2008). centage develop distant disease due to hidden
Once these cells are isolated, they can be cervical metastasis. There are differing opinions
detected with two strategies: using direct methods, in regard to the impact of CTC presence on the
based on immunocytochemistry, immunofloures- development of regional or distant metastasis.
cense (as those developed by CellSearch) and flow The lymph node affectation (the most determin-
cytometry; and indirect methods, mainly based ing factor in the prognosis and survival of head
on the study of nucleotic acid. The most com- and neck cancer patients) has been related to the
monly used methods for indirect detection are the presence of CTCs in peripheral blood by some
quantitative polymerase chain reaction (PCR) and authors. Hristozova et al. (2011) defend that there
the quantitative reverse transcription (qRT-PCR). is a significant relationship between the number
These methods depend on the expression patterns of CTCs and the development of locorregional
of certain genes, the detection of known genetic metastasis in head and neck carcinomas, despite
mutations, amplifications, genomic aberrations, CTC-positive cases observed in patients with N0/
or methylation patterns in tumor cells (Lacroix, N1. Their frequency was significantly higher in
2006). The main problem for the detection of CTCs patients with stage N2b or higher. However, stud-
is that there is no universally applicable marker for ies by Grobe et al. (2014) and Jatana et al. (2010)
all tumors and all patients due to tumoral intra- found no association between the presence of
and interheterogeneity. CTCs and locorregional metastasis in OSCC. In
addition, Grobe et  al. (2014) and Nichols et  al.
WHY SHOULD THE STUDY OF (2012) established that the presence of distant
CTCs BE INCLUDED IN PRECISION metastasis is directly correlated to the presence
MEDICINE FOR OSCC PATIENTS? of CTCs in OSCC and head and neck carcinomas.
In Nichols’s study, 40% of patients with pulmo-
As in most tumors, in OSCCs there is an impor- nary nodes affected had CTCs. Supporting the
tant clinical need to be covered, which consists of value of monitoring the CTC levels as a prognostic
the identification of early screening markers and marker, Jatana et al. (2010) determined in a study
34  Oral squamous cell carcinoma and liquid biopsies: A new tool for precision oncology

with 48 head and neck squamous cell carcinoma primary tumor and generate metastasis that can-
(HNSCC) patients that those without CTCs pre- not be biopsied in many cases. In this sense, the
sented higher progression-free survival (PFS). group of Tinhofer et  al. (2012) determined using
In a recent study using CellSearch, Bozec et  al. flow cytometry that the number of CTCs of head
(2013) found that CTCs were detected in 3/9 patients and neck cancer patients increased after radio-
(33%) and 8/49 (16%), respectively, in advanced-stage therapy, although they were unable to determine if
HNSCC patients. He et al. (2013) reported that CTCs this increase had clinical impact. In CTC-positive
were present in patients presenting with nodal stage patients, the presence of epidermal growth factor
2 and 3 rather than 0 or 1. In addition, Nichols et al. receptor (EGFR) was detected in 100% of cases and
(2012), reported that CTCs were detected in 6/15 the p-EGFR form in 55%. Additionally, combined
(40%) of HNSCC patients with suspected metasta- treatment of the anti-EGFR antibody Cetuximab
sis. In regard to the value of these cells as a prog- in combination with radiotherapy reduced the
nostic marker, in the study of oral tumors developed receptor-active CTC levels, in comparison with the
by Tinhofer et al. (2014) they found no association treatment based on cisplatin/5-fluorouracil. This
between CTCs and PFS, but they found this associa- study showed the potential to monitor the molecu-
tion in tumors not located in the oral cavity, being lar alterations in CTCs to guide OSCC treatments.
the rate of PFS was lower in CTC-positive cases. In For example, in monitoring p-EGFR levels in
another study analyzing head and neck tumors, CTCs, we could say that Cetuximab is reaching its
the correlation between the presence of CTCs and therapeutic target efficiently. Although currently
the risk of local and distant recurrence and even the Cetuximab is the only target therapy approved for
patients’ response to treatment has been described use on head and neck cancer patients, other thera-
(Hristozova et al., 2011; Buglione et al., 2012). pies are emerging as good alternatives. Such is the
All of these results show the potential of CTC case of Rapamycin to block the PI3 K pathway. The
quantification to help oncologists detect tumor status of this gene has been analyzed in the CTC
dissemination and predict a patient’s prognosis in population in other tumors (Schneck et al., 2013)
a more realistic way. Despite this, more research and this strategy could also be applied to OSCC.
is needed to clarify the entire clinical potential of Another example of how molecular charac-
CTCs in these types of tumors. terization of a CTC can help in the diagnosis and
therapies selection is the possibility of analyzing
the viral origin of oral carcinomas. The group of
WHY IS PERSONALIZED AND Weismann et  al. (2009) determined the presence
PRECISION MEDICINE THROUGH of E6/E7 HR-HPV 16 or HR-HPV 18 variants in
CTC STUDY NECESSARY? DOES IT isolated CTCs of patients with cervical carcinoma.
HAVE THE POTENTIAL TO IMPROVE HPV-negative head and neck squamous carcino-
THE HEALTH OF OSCC PATIENTS? mas are related to p53 mutations and drops in p16
levels, whereas those virus-induced tumors are
The ability to detect CTCs has a clear impact on the characterized by a p53 wild type with an increase
prognosis and treatment of cancer patients, provid- of both p16 and pRb levels, which makes these
ing clear evidence of the dissemination of the primary tumors two different entities from a molecular
cancer and being a risk factor for future development point of view (zur Hausen, 2000; Beachler et  al.,
of metastasis in addition to a peripheral marker to pre- 2012). There are also multiple studies demonstrat-
dict treatment susceptibility and disease surveillance. ing that HPV-induced tumors have a better prog-
The chance to isolate OSCC patients’ CTCs at nosis than those that are not HPV-induced being
different moments throughout the evolution of in a similar disease stage (Vermorken et al., 2014).
the disease allows for the perspective of the tumor One of the main characteristics of high aggres-
in real time that would avoid inefficient over- sive head and neck carcinomas are the expression of
treatments and secondary effects as well as the mesenchymal markers. For epithelial tumor cells,
expenses derived from these treatments. In this for migration through the basal lamina and inva-
sense, the possibility to molecularly characterize sion into other tissues, an epithelial to mesenchy-
OSCC tumors is especially interesting using CTCs, mal transition or EMT is necessary. This process
especially when patients undergo surgery for the is a dynamic mechanism through which epithelial
 Challenges and research opportunities for the clinical inclusion of CTC analysis to manage OSCC  35

cells increase their mobility and cell plasticity due on liquid biopsy study. Despite this evidence, we
to the loss of epithelial markers such as E-cadherin must face several challenges before CTC analy-
or EpCAM and the increase of mesenchymal mark- sis could be a clinical reality for OSCC patient
ers’ expression, such as N-cadherin and vimentin. management.
It is important to identify those tumors with a more First, it is necessary to validate the results
mesenchymal phenotype, since these will be less obtained with systems such as CellSearch in larger
sensitive to different therapies and will have higher and more homogeneous groups of patients, pro-
chances of relapse after surgery. In this sense, dif- viding more robust results in terms of prognostic
ferent studies have characterized the EMT phe- value of the CTC monitoring. Furthermore, there
nomenon in CTCs (Alonso-Alconada et  al., 2014; is a scarce amount of studies that monitor the CTC
Barriere et al., 2014; Lim et al., 2014) and described levels through the different stages of the disease,
the expression of mesenchymal markers in CTCs. at a basal level, after surgery or after the treatment
The only study developed on OSCC to charac- with chemo- or radiotherapy. Undoubtedly, in
terize the EMT process in CTCs was developed on addition to the possibility of determining tumor
a cohort of 20 patients diagnosed with oral squa- dissemination in a more precise way than imag-
mous cell carcinoma by the Maxillofacial Surgery ing tests, the option of using CTC levels to monitor
Service of the Complejo Hospitalario Universitario the disease would allow for a qualitative leap in the
de Santiago de Compostela. An amount of 7.5 mL management of metastatic OSCC.
of peripheral blood was extracted from each patient On the other hand, as mentioned earlier, there
before surgery to extract the tumor. Collected is a technical limitation to studying CTCs in OSCC
samples were processed using the CELLectionTM residing in the heterogeneous expression of epithe-
Epithelial Enrich kit (Invitrogen, Dynal, Oslo, lial markers, such as EpCAM, which are not always
Norway) to isolate CTCs based on EpCAM expres- sufficient to allow the CTCs isolation in an efficient
sion. The CTC population was analyzed for the manner. Therefore, it is very important to validate
expression of 12 markers related to the process and develop new isolation and detection tech-
of EMT (E-cadherin, HIFT1A, KRT 8, KRT 19, niques that provide higher detection sensitivity.
KRT 20, SNAIL1, SNAIL2, TGF-β1, TIMP1, VIM, In this sense, new microfluidics-based technolo-
ZEB1 and ZEB 2). Our study’s data, which have gies for the enrichment of CTCs and new detec-
not been previously published, demonstrated that tion methods, such as biosensors, offer a promising
a combination of immunoisolation based on the opportunity to make CTC detection systems of
presence of EpCAM, and a gene expression anal- small dimensions and provide ease of use that can
ysis by RT-qPCR is a valid strategy for the detec- be applied at health centers and hospitals without
tion and molecular characterization of CTCs in requiring large facilities and significant financial
OSCC patients. Additionally, TGFβ 1, VIM, and costs.
ZEB 1 genes (related to the EMT) were identified In parallel, it is important to start applying high-
as potential CTC markers in OSCC patients, being resolution molecular techniques such as massive
their expression levels are statistically different sequencing of genomes and exomes to characterize
than ones obtained in healthy controls. In regard the CTCs of OSCC patients. These techniques are
to this result, Buglione et al. (2012) related the low being used in other types of tumors to identify the
CTC frequency in head and neck tumors located molecular bases that are behind resistance to ther-
in the oral cavity to the highest expression of mes- apies and are enabling the development of single-
enchymal markers in these locations, making CTC cell studies that provide a better understanding of
isolation based on EpCAM inefficient. the phenomenon of tumor heterogeneity.
Last, the greatest challenge in the field of CTCs
CHALLENGES AND RESEARCH consists of their isolation and ex vivo culture. Due
OPPORTUNITIES FOR THE CLINICAL to the reduced number of CTCs in circulation and
INCLUSION OF CTC ANALYSIS TO the mechanical damage suffered during the isola-
MANAGE OSCC tion process and their limited proliferation capac-
ity, there are very few groups that have been able
Nowadays it is completely accepted that the most to obtain stable in vitro cultures. The generation
direct path to develop precision medicine is based of these cultures from OSCC patients would allow
36  Oral squamous cell carcinoma and liquid biopsies: A new tool for precision oncology

us to develop functional studies to identify and markers: Characterization of cell subpopula-

validate new therapeutic targets making in vitro tions. Ann Transl Med 2(11);2014:109.
screening of drugs the most efficient therapy for Beachler DC, Weber KM, Margolick JB et al. Risk
each patient in each stage of the disease. factors for oral HPV infection among a high
prevalence population of HIV-positive and
CONCLUSION at-risk HIV-negative adults. Cancer Epidemiol
Biomarkers Prev 21(1);2012:122–33.
Despite the great advances in terms of therapeu- Bozec A, Peyrade F, Hebert C et al. Significance
tic alternatives and the surge of new monitoring of circulating tumor cell detection using
and predictive markers in the past two decades, the CellSearch system in patients with
OSCC tumors still have an important void to be locally advanced head and neck squamous
filled in terms of clinical tools to develop preci- cell carcinoma. Eur Arch Otorhinolaryngol
sion oncology. The analysis and characterization 270(10);2013:2745–9.
of the CTC population that is escaping from the Buglione M, Grisanti S, Almici C et al. Circulating
primary tumor and metastasis are a true oppor- tumour cells in locally advanced head and
tunity to learn more about the molecular events neck cancer: Preliminary report about their
behind carcinogenesis and aggressiveness of possible role in predicting response to non-
OSCC tumors, as well as the best way to monitor surgical treatment and survival. Eur J Cancer
patients with systemic disease. Although different 48(16);2012:3019–26.
studies support the clinical usefulness of the CTC Cohen EE, Lingen MW, and Vokes EE. The
study to reach precision oncology in OSCC, there expanding role of systemic therapy
is a higher commitment to the development of a in head and neck cancer. J Clin Oncol
promising work field that may provide knowledge 22(9);2004:1743–52.
and technology to bring clinicians a real tool for Cohen SJ, Punt CJ, Iannotti N et al. Relationship
personalized medicine. of circulating tumor cells to tumor response,
progression-free survival, and overall survival
REFERENCES in patients with metastatic colorectal cancer.
J Clin Oncol 26(19);2008:3213–21.
Alix-Panabières C. EPISPOT assay: Detection of Cohen SJ, Punt CJ, Iannotti N et al. Prognostic
viable DTCs/CTCs in solid tumor patients. In significance of circulating tumor cells in
M Ignatious, C Sotiriou, and K Pantel (Eds), patients with metastatic colorectal cancer.
Minimal Residual Disease and Circulating Ann Oncol 20(7);2009:1223–9.
Tumor Cells in Breast Cancer, pp. 69–76. Cooper JS, Pajak TF, Forastiere AA et al.
Berlin, Springer; 2012. Postoperative concurrent radiotherapy and
Alonso-Alconada L, Muinelo-Romay L, Madissoo chemotherapy for high-risk squamous-cell
K et al. Molecular profiling of circulating carcinoma of the head and neck. N Engl J
tumor cells links plasticity to the metastatic Med 350(19);2004:1937–44.
process in endometrial cancer. Mol Cancer Cristofanilli M, Budd GT, Ellis MJ et al. Circulating
13;2014:223. tumor cells, disease progression, and survival
Ashworth T. A case of cancer in which cells in metastatic breast cancer. N Engl J Med
similar to those in the tumours were seen 351(8);2004:781–91.
in the blood after death. Aust Med J Cristofanilli M, Hayes DF, Budd GT et al.
14(3);1869:146–9. Circulating tumor cells: A novel prog-
Bagan J, Sarrion G, and Jimenez Y. Oral nostic factor for newly diagnosed
cancer: Clinical features. Oral Oncol metastatic breast cancer. J Clin Oncol
46(6);2010:414–7. 23(7);2005:1420–30.
Barnes L. Pathology and Genetics of Head and de Bono JS, Scher HI, Montgomery RB et al.
Neck Tumours. Lyon, IARC Press; 2005. Circulating tumor cells predict survival benefit
Barriere G, Fici P, Gallerani G, Fabbri F, Zoli W, from treatment in metastatic castration-
and Rigaud M. Circulating tumor cells and resistant prostate cancer. Clin Cancer Res
epithelial, mesenchymal and stemness 14(19);2008:6302–9.
 References 37

Forastiere A, Koch W, Trotti A, and Sidransky Available at http://oralcancerldv.org/es/.

D. Head and neck cancer. N Engl J Med Accessed October 6, 2015.
345(26);2001:1890–1900. Lim SH, Becker TM, Chua W, Ng WL, de Souza
Grobe A, Blessmann M, Hanken H et al. Prognostic P, and Spring KJ. Circulating tumour cells
relevance of circulating tumor cells in blood and the epithelial mesenchymal transi-
and disseminated tumor cells in bone marrow tion in colorectal cancer. J Clin Pathol
of patients with squamous cell carcinoma of the 67(10);2014:848–53.
oral cavity. Clin Cancer Res 20(2);2014:425–33. Maheswaran S, Sequist LV, Nagrath S et al.
He S, Li P, Long T et al. Detection of circulating Detection of mutations in EGFR in circu-
tumour cells with the CellSearch system in lating lung-cancer cells. N Engl J Med
patients with advanced-stage head and neck 359(4);2008:366–77.
cancer: Preliminary results. J Laryngol Otol Nagrath S, Sequist LV, Maheswaran S et al.
127;2013:788. Isolation of rare circulating tumour cells in
Hou J, Greystoke A, Lancashire L et al. Evaluation cancer patients by microchip technology.
of circulating tumor cells and serological cell Nature 450(7173);2007:1235–9.
death biomarkers in small cell lung cancer Nichols AC, Hammond A, Palma DA et al.
patients undergoing chemotherapy. Am J Detection of circulating tumor cells in advanced
Pathol 175(2);2009:808–16. head and neck cancer using the CellSearch
Hristozova T, Konschak R, Stromberger C et al. system. Head Neck 34(10);2012:1440–4.
The presence of circulating tumor cells (CTCs) Parkinson DR, Dracopoli N, Petty BG et al.
correlates with lymph node metastasis in non- Considerations in the development of cir-
resectable squamous cell carcinoma of the culating tumor cell technology for clinical
head and neck region (SCCHN). Ann Oncol use. J Transl Med 10(1);2012:138.
22(8);2011:1878–85. Partridge M. Head and neck cancer and precan-
Jatana KR, Balasubramanian P, Lang JC cer: Can we use molecular genetics to make
et al. Significance of circulating tumor better predictions? Ann R Coll Surg Engl
cells in patients with squamous cell car- 81(1);1999:1–11.
cinoma of the head and neck: Initial Paterlini-Brechot P, and Benali NL. Circulating
results. Arch Otolaryngol Head Neck Surg tumor cells (CTC) detection: Clinical
136(12);2010:1274–9. impact and future directions. Cancer Lett
Jones A, Roland N, Field J, and Phillips D. The 253(2);2007:180–204.
level of cervical lymph node metastases: Psychogios G, Mantsopoulos K, Bohr C, Koch
Their prognostic relevance and relationship M, Zenk J, and Iro H. Incidence of occult
with head and neck squamous carcinoma cervical metastasis in head and neck carcino-
primary sites. Clin Otolaryngol Allied Sci mas: Development over time. J Surg Oncol
19(1);1994:63–69. 107(4);2013:384–7.
Kolostova K, Spicka J, Matkowski R, and Bobek V. Ragin CC, Modugno F, and Gollin SM. The epi-
Isolation, primary culture, morphological demiology and risk factors of head and neck
and molecular characterization of circulating cancer: A focus on human papillomavirus.
tumor cells in gynecological cancers. Am J J Dent Res 86(2);2007:104–14.
Transl Res 7(7);2015:1203. Ring A, Smith IE, and Dowsett M. Circulating
Lacroix M. Significance, detection and markers tumour cells in breast cancer. The Lancet
of disseminated breast cancer cells. Endocr Oncology 5(2);2004:79–88.
Relat Cancer 13(4);2006:1033–67. Rosenberg R, Gertler R, Friederichs J et al.
Leemans CR, Tiwari R, Nauta J, Van der Waal I, Comparison of two density gradient cen-
and Snow GB. Regional lymph node involve- trifugation systems for the enrichment of
ment and its significance in the development disseminated tumor cells in blood. Cytometry
of distant metastases in head and neck carci- 49(4);2002:150–8.
noma. Cancer 71(2);1993:452–6. Saucedo-Zeni N, Mewes S, Niestroj R et al.
Lifelong Learning Programme. Cáncer Oral. A novel method for the in vivo isolation of
Prevención y diagnóstico precoz. 2015. circulating tumor cells from peripheral blood
38  Oral squamous cell carcinoma and liquid biopsies: A new tool for precision oncology

of cancer patients using a functionalized Villegas LFS, Pérez NMS, Lázaro MV, de las
and structured medical wire. Int J Oncol C, María Luisa M, Fernández SR, and
41(4);2012:1241–50. de Castro SV. Actualidad y futuro en las téc-
Schneck H, Blassl C, Meier-Stiegen F et al. nicas de cuantificiación de células tumorales
Analysing the mutational status of PIK3CA in circulantes: su importancia en tumores sólidos
circulating tumor cells from metastatic breast epiteliales [Current and future techniques in
cancer patients. Molr Oncol 7(5);2013:976–86. the quantification of circulating tumor cells:
Shah JP. Patterns of cervical lymph node Their importance in solid epithelial tumors].
metastasis from squamous carcinomas of Revista del Laboratorio Clínico [Journal of the
the upper aerodigestive tract. Am J Surg Clinical Laboratory] 4(3);2011:163–9.
160(4);1990:405–9. Vona G, Sabile A, Louha M et al. Isolation by size
Slootweg PJ, Bolle CW, Koole R, and Hordijk G. of epithelial tumor cells: A new method for
Cause of death in squamous cell carcinoma the immunomorphological and molecular
of the head and neck: An autopsy study characterization of circulating tumor cells. Am
on 31 patients. J Craniomaxillofac Surg J Pathol 156(1);2000:57–63.
20(5);1992:225–7. Weismann P, Weismanova E, Masak L et al.
Snow G, Annyas A, Ev S, Bartelink H, and Hart A. The detection of circulating tumor cells
Prognostic factors of neck node metastasis. expressing E6/E7 HR-HPV oncogenes in
Clin Otolaryngol Allied Sci 7(3);1982:185–92. peripheral blood in cervical cancer patients
Teymoortash A, and Werner J. Current advances after radical hysterectomy. Neoplasma
in diagnosis and surgical treatment of lymph 56(3);2009:230–8.
node metastasis in head and neck cancer. Wheeler SE, Shi H, Lin F et al. Enhancement of
GMS Curr Top Otorhinolaryngol Head Neck head and neck squamous cell carcinoma pro-
Surg 11;2012:Doc04. liferation, invasion, and metastasis by tumor-
Tinhofer I, Hristozova T, Stromberger C, KeilhoIz associated fibroblasts in preclinical models.
U, and Budach V. Monitoring of circulating Head Neck 36(3);2014:385–92.
tumor cells and their expression of EGFR/ Wikner J, Gröbe A, Pantel K, and Riethdorf S.
phospho-EGFR during combined radiother- Squamous cell carcinoma of the oral cavity
apy regimens in locally advanced squamous and circulating tumour cells. World J Clin
cell carcinoma of the head and neck. Int J Oncol 52;2014:114.
Radiat Oncol Biol Phys 83(5);2012:e685–90. Yongkui L, Jian L, and Jingui L. 18 FDG-PET/CT
Tinhofer I, Konschak R, Stromberger C et al. for the detection of regional nodal metastasis
Detection of circulating tumor cells for pre- in patients with primary head and neck cancer
diction of recurrence after adjuvant chemo- before treatment: A meta-analysis. Surg
radiation in locally advanced squamous cell Oncol 22(2);2013:e11–16.
carcinoma of the head and neck. Ann Oncol Zbären P, and Lehmann W. Frequency and sites
25(10);2014:2042–7. of distant metastases in head and neck squa-
Vermorken JB, Psyrri A, Mesia R et al. Impact mous cell carcinoma: An analysis of 101 cases
of tumor HPV status on outcome in patients at autopsy. Arch Otolaryngol Head Neck Surg
with recurrent and/or metastatic squamous 113(7);1987:762–4.
cell carcinoma of the head and neck receiv- zur Hausen H. Papillomaviruses causing can-
ing chemotherapy with or without cetuximab: cer: Evasion from host-cell control in early
Retrospective analysis of the phase III EXTREME events in carcinogenesis. J Natl Cancer Inst
trial. Ann Oncol 25(4);2014:801–7. 92(9);2000:690–8.
 4
Precision medicine for brain gliomas

YUSUF IZCI

Background 39 Why precision medicine is needed in brain

Introduction 39 gliomas 45
Prevalence 39 Clinical scenario for precision medicine in
Incidence 41 brain gliomas 45
Survival and mortality 41 Future directions 46
Current approaches for brain gliomas 42 Conclusion 46
Impact of precision medicine on the References 46
treatment of brain gliomas 43

BACKGROUND Glioblastoma is the most common and malignant

histological type of glioma (approximately 45% of
Introduction all gliomas) (Akay et al., 2002; Zong et al., 2012). It
may be solitary or multicentric in the brain (Izci
Gliomas are the most common primary intracra- et  al., 2005). Glioblastoma may present with dif-
nial tumors, representing about 80% of malig- ferent clinical and radiological characteristics and
nant brain tumors (Zong et  al., 2012; Rajesh has a 5-year relative survival of about 5% (Akay
et al., 2017). Three different types of gliomas have et al., 2002; Izci et al., 2005; Bondy et al., 2008; de
been described in the brain as astrocytoma, oli- Robles et al., 2015).
godendroglioma, and ependymoma (Izci, 2014).
Astrocytoma is the most frequent histological type Prevalence
of glioma and arises from the astrocytes (Figures
4.1 and 4.2) (Zong et al., 2012). About 10% of the gli- Complete prevalence proportions of brain gliomas
omas are oligodendrogliomas (Figure 4.3). Mixed estimate the total number of people living with
gliomas, primarily oligoastrocytomas, account for glioma and include both newly diagnosed cases
about 5%–10% of all gliomas. Ependymomas arise within the year and patients surviving during pre-
from the ependymal cells, which lie in the ven- vious years (Porter et  al., 2010). Complete preva-
tricular system and the central canal of the spinal lence of glioma reflects the complex relationships
cord. It is relatively rare in adults, accounting for among incidence, survival, and population demo-
2%–3% of primary brain tumors, but it is frequent graphics of this disease (Ostrom et al., 2016; Zhang
in children. Today, biological markers help pathol- et al., 2016). Therefore, it provides important infor-
ogists separate oligodendrogliomas from other mation about the total burden of brain gliomas
types of gliomas. Glioma has a poor prognosis in a population. So, the estimation of complete
(Perry and Wesseling, 2016). Although relatively prevalence provides critical information about gli-
rare, it causes significant mortality and morbidity. omas for public health and health care planning,

39
40  Precision medicine for brain gliomas

(a) (b)

Figure 4.1  (a) Axial and (b) coronal magnetic resonance images of a patient with grade II a
­ strocytoma
(low grade).

(a) (b)

Figure 4.2  (a) Axial and (b) coronal magnetic resonance images of a patient with grade IV
­astrocytoma (glioblastoma).

including treatment protocols, decision making, also critical that glioma registries continue to col-
and funding research for gliomas (de Robles et al., lect established and emerging prognostic and pre-
2015). With increasing focus on precision medicine dictive factors for this important disease (Ostrom
and the incorporation of molecular parameters et al., 2016).
into the World Health Organization (WHO) 2016 The most prevalent types of brain glioma var-
central nervous system (CNS) tumor classification, ied by age at prevalence. The most prevalent histo-
it is critical that the data are complete, and factors logical types are glioblastoma (6.46 per 100,000),
collected at the population level are fully integrated diffuse astrocytoma (3.76 per 100,000), and pilo-
into brain glioma reporting (Louis et al., 2016). It is cytic astrocytoma (3.71 per 100,000). The most
 Background 41

(a) (b) (c)

Figure 4.3  (a) Axial, (b) coronal, and (c) sagittal magnetic resonance images of a patient with
­oligodendroglioma grade II.

prevalent tumor types by age are childhood pilo- 100,000 persons. Age-adjusted incidence of glio-
cytic astrocytoma (30.6% of cases, prevalence 6.82 blastoma (WHO grade IV), the most common and
per 100,000), adolescent and young adult pilocytic most deadly glioma subtype in adults, ranges from
astrocytoma (12.2% of cases, prevalence 5.92 per 0.59 to 3.69 per 100,000 persons (Porter et al., 2010;
100,000), and adult glioblastoma (22.1% of cases, Ostrom et al., 2014).
prevalence 12.76 per 100,000) (Ostrom et al., 2016). Low-grade glioma accounts for 20% of all glio-
Even though brain tumors are not the most preva- mas, with most patients presenting in the fourth
lent cancer in all age groups, these tumors are con- decade (Hayhurst, 2017). Men are more often
sistently prevalent throughout life and therefore affected than women, and approximately 80%
merit documentation (Ostrom et al., 2014). of patients present with seizures (Pouratian and
Schiff, 2010; Perry and Wesseling, 2016; Bush
Incidence and Butowski, 2017). The incidence of low-grade
glioma is uncertain as International Statistical
The incidence rate of a disease can be defined as Classification of Diseases codes do not differen-
the number of at-risk cases per population in a spe- tiate grade of tumor. However, the RARECARE
cific time period. The average annual age-adjusted European registry project estimates the overall
incidence rate of brain tumors in the United States incidence across Europe of astrocytoma to be 4.8
between 2007 and 2011 was 7.25 per 100,000 per- per 100,000 and of oligodendroglioma to be 0.4 per
sons (Ostrom et  al., 2014). Many different orga- 100,000 (Crocetti et al., 2012).
nizations track the incidence of brain gliomas.
This can be found using data collected through Survival and mortality
government cancer surveillance (i.e., statewide or
countrywide cancer registries) or through the use Standard treatment (surgery, radiotherapy, and
of health system records. Incidence rates of brain chemotherapy) in glioblastoma has very limited
glioma vary significantly by histological type, age effectiveness, with median overall survival of
at diagnosis, gender, and country (Bondy et  al., patients no longer than 15 months (Akay et  al.,
2008; Porter et al., 2010). It is difficult to compare 2002; Izci, 2014). The most important prognostic
the incidence rates of brain gliomas from different factors for glioblastoma are extent of tumor resec-
sources because of the lack of consistent definition tion, age at diagnosis, and Karnofsky performance
of glioma and various glioma histological types as status of the patient (Walid, 2008). Survival also
well as differences in data collection techniques. varies significantly by grade across all glioma sub-
Overall age-adjusted incidence rates (adjusted to types. Many groups that track the incidence of
the national population of each respective study) glioma also track the proportion of persons who
for all types of gliomas range from 4.67 to 5.73 per survive set periods of time after their diagnoses.
42  Precision medicine for brain gliomas

Pilocytic astrocytoma (grade I) has the highest disease risk are the basis of precision medicine for
relative survival. Glioblastoma has the poorest brain gliomas (Izci, 2014).
overall survival, with only 0.05%–4.7% of patients
surviving 5 years after the diagnosis. In general, CURRENT APPROACHES FOR BRAIN
gliomas with an oligodendroglial component have GLIOMAS
increased survival, as opposed to those with an
astrocytic component. Age is significantly associ- Current management techniques for low- and
ated with survival after diagnosis for all types of high-grade brain gliomas are different. The tra-
gliomas, but the effect is most pronounced for glio- ditional approach in low-grade gliomas was to
blastoma. Recently, it was shown in population- “watch, wait and rescan,” with intervention at the
based parallel cohorts of diffuse low-grade gliomas time of tumor progression, which often heralds
that early surgical resection was associated with change to a malignant glioma (Hayhurst, 2017).
better overall survival than were biopsy and watch- However, recent advances in our knowledge of the
ful waiting. The 22981/26981 trial by the European biology of low-grade gliomas, advances in surgi-
Organisation for Research and Treatment of cal techniques, and learning the long-term impact
Cancer/National Cancer Institute of Canada of adjuvant treatments have caused a shift toward
demonstrated a survival benefit for glioblastoma aggressive early treatment of low-grade glioma,
patients who received concurrent temozolomide aiming to delay the time to malignant transforma-
with postoperative radiation, with median sur- tion, to increase overall survival, and to improve
vival of 14.6 months for those receiving concurrent neurological functional outcome.
therapy versus 12.1 months for those who received The role of surgery in the management of low-
radiotherapy alone (Stupp et al., 2005). This treat- grade glioma is still controversial. Because they
ment regimen, known as the “Stupp protocol,” was have a predilection for eloquent regions, the risk of
the result of this trial and was first presented in developing new neurological deficits had tempered
2004. In the years since this trial was completed, enthusiasm for radical resection of these tumors
it has been established as the standard of care (Pouratian and Schiff, 2010). However, there is
for primary glioblastoma. For various r­easons— increasing evidence that resection of low-grade
including tolerance of chemotherapy, access to glioma improves overall survival of the patient and
chemotherapeutic agents, and overall performance importantly delays the time to malignant progres-
status—not all persons with glioblastoma receive sion of the disease. Contemporary data support-
this regimen. This result was then confirmed in ing early surgery and the widespread adoption of
a large study of glioblastoma patients, and sev- awake craniotomy with intraoperative functional
eral analyses found statistically significant trends imaging and mapping have significantly changed
in increasing survival for glioblastoma after this the management of low-grade gliomas.
development, especially in those who received The standard treatment for high-grade gliomas
surgery followed by radiation. There has been an starts with surgical procedures aiming to remove
increasing trend in survival from oligodendrogli- the tumor as much as possible without injuring
oma, which is also attributed to improvements in the normal brain tissue (Akay et al., 2002). Then,
diagnosis and treatment. radiotherapy and chemotherapy protocols includ-
Precision medicine for brain gliomas means ing temozolomide follow the surgical treatment
matching the right patient to the right treatment at (Stupp et al., 2005). Although this protocol is effec-
the right time (Halkin, 2015). This requires gaining tive at shrinking the tumors, it may cause many
a deeper understanding of the specific molecular side effects that significantly affect the quality of
characteristics that are driving a patient’s tumor life of these patients. At this point, it is critical
growth and finding the right treatments to target to identify genetic and molecular markers of the
those specific molecular abnormalities that are tumor in order to determine the targets for per-
responsible for the disease. So, sequencing the sonalized treatments (Ataman et  al., 2004; Tan
patient’s genome and tracking the changes in a bil- et al., 2016). Despite various treatment modalities
lion biomarkers, including RNA, proteins, metab- for high-grade gliomas, the patients have a dismal
olites, and antibodies, to predict and validate prognosis (Walid, 2008).
 Impact of precision medicine on the treatment of brain gliomas  43

IMPACT OF PRECISION MEDICINE 2014). The ultimate goal of targeted therapy should
ON THE TREATMENT OF BRAIN be “selectivity” inhibiting only tumor cells (Adair
GLIOMAS et al., 2014; Oh et al., 2014; Morokoff et al., 2015).
The targeted approaches currently in clinical tri-
The synonyms of precision medicine are indi- als or in laboratory development include drugs,
vidualized medicine, personalized medicine, and monoclonal antibodies, immunotherapy, small
targeted therapy. This approach is particularly per- molecules inhibiting specific proteins, and spe-
tinent in the area of brain tumors, especially with cific targeting of glioma stem cells (Oh et al., 2014;
the most common and deadliest form of brain can- Prados et al., 2015; Miller and Wen, 2016). Some of
cer, glioblastoma (Olar and Aldape, 2014; Duffau these targets for glial tumors are summarized in
and Taillandier, 2015; Chen et al., 2016). DNA and Table 4.1. Thus, there is a need for unconventional
RNA sequencing technology and the sprouting treatment strategies to curb glioma. Strategies such
fields of proteomics and metabolomics are trans- as gene therapy, microRNA (miRNA) therapy, stem
forming our understanding of glioma biology. cells, and immunotherapy may potentially lead to
Researchers are now rapidly discovering biomark- effective treatments (Kouri et  al., 2015; Morokoff
ers to probe the underlying pathology and define et  al., 2015; Hodges et  al., 2016; Chandran et  al.,
the dynamic changes that occur during develop- 2017).
ment of glioma (Izci, 2014). Integrative personal omics profiles (iPOPs) is
Progress in glioma therapy could be attained by the analysis of personal transcriptomes, proteomes,
improved comprehension of glioma biology, iden- cytokines, metabolomes, and autoantibodies from
tification of relevant targets and signaling path- a single individual over a 14-month period (Chen
ways for treatment interventions, development of et al., 2012). It is important to learn how iPOPs can
personalized medicine, optimization of surgery provide critical insight into the causes and devel-
and radiotherapy, and innovative neuroimaging opment of disease. By learning this profile, it will
modalities (Grant et al., 2014; Rajesh et al., 2017). be easy to understand
Proteogenomic characterization is a potential
strategy that could lead to identification of molec- ●● How “omics” profiling of transcriptomes, pro-
ular drivers, molecular classification of disease teomes, cytokines, metabolomes, and autoanti-
subgroups, and glioma treatment (Grant et  al., bodies has revealed dynamic and wide-ranging

Table 4.1  Some of the potential targets for different types of brain gliomas

Tumor type Target

Low-grade astrocytoma RAS/RAF/mitogen-activated protein kinase pathway,
(Grades I and II) phosphatidylinositol 3-kinase/mammalian target of rapamycin
pathway, tyrosine kinase receptor, EGFR, PDGFR, BRAF/MAPK
signaling pathway, PI3 K/mTOR signaling pathway
Oligodendroglioma (Grades 1p19q deletion, p130Cas, VEGF receptor, tyrosine kinase
II and III)
Oligoastrocytoma (Grades II VEGF receptor, tyrosine kinase
and III)
Glioblastoma (Grade IV) mTOR/PI3 K/Akt pathway, e-MET pathway, ZEB1 pathway, Wnt
family, Sonic hedgehog, Notch2, transforming growth factor beta
(TGF-β), bone morphogenetic protein (BMP) signaling, 27
Homeobox family, B lymphoma Mo-MLV insertion region 1
homolog (Bmi-1), PTEN (phosphatase and tensin homolog),
telomerase, efflux transporters, epidermal growth factor (EGF),
microRNA, and VEGF receptors, L1 CAM, p38MAPK signaling
Ependymoma (Grades I, II, III) ErbB2, farnesyltransferase, p53, histone deacetylase
44  Precision medicine for brain gliomas

variation in the different molecular compo- not receive aggressive treatment. This is important
nents that occur during the progression of a for precision medicine in glioblastoma and clini-
glioma. cians may be able to use this information in the
●● How analysis of heteroallelic expression sug- future to avoid unnecessary treatment regimens
gests that significant changes in differential for patients in the proneural glioblastomas (Verma
allele expression occur in healthy and diseased et al., 2016).
states. The mesenchymal glioblastoma subgroup con-
●● How integrated analysis of multiomic data tains the most frequent number of mutations in
on coding and gene regulation can lead to the NF1 tumor suppressor gene (37%) (Verhaak
enhanced prediction of glioma risk. et al., 2010). Frequent mutations in the PTEN and
TP53 tumor suppressor genes also occurred in the
Glioblastoma has four well-characterized sub- group. Patients in the mesenchymal group had
types, which can all respond differently to different significant increases in survival after aggressive
treatments. These subtypes are proneural, neural, treatment, unlike those in the proneural, and, to
classical, and mesenchymal. The subtypes were an extent, in the neural subgroups (Verhaak et al.,
found to differ by the type of genetic abnormalities 2010; Evans, 2011).
they carried and by the patients’ clinical character- Although the neural subgroup had mutations in
istics (Verhaak et al., 2010). many of the same genes as the other groups, the
Classical glioblastomas are characterized by group did not stand out from the others as having
abnormally high levels of epidermal growth fac- significantly higher or lower rates of mutations.
tor receptor (EGFR). EGFR is a protein found on The neural group was characterized by the expres-
the surface of some cells that, when bound by epi- sion of several gene types that are also typical of
dermal growth factor, sends signals for the cell to the brain’s normal, noncancerous nerve cells, or
keep growing. The EGFR abnormalities occur at neurons. Patients in the neural group were the old-
a much lower rate in the three other glioblastoma est, on average. They also had some improvement
subtypes. However, TP53, the most frequently in survival after aggressive treatment, but not as
mutated gene in glioblastoma, is not mutated in much as the classical and mesenchymal groups
any of the classical glioblastoma. TP53 is the gene (Verhaak et al., 2010; Evans, 2011).
for a protein that normally suppresses tumor Furthermore, glioblastoma is one of the most het-
growth. Clinically, the classical group survived the erogeneous tumors—meaning individual tumors
longest of the subgroups in response to aggressive might be made up of cancer cells, which have a wide
treatment (Verhaak et al., 2010). variety of molecular alterations and/or ­mutations
Different than classical glioblastomas, TP53 is (Halkin, 2015). As such, it is not likely that one treat-
significantly mutated in proneural glioblastomas ment will be able to serve as a silver bullet for all
(54%). These glioblastomas are also characterized glioblastoma patients (Halkin, 2015). This also helps
by having the most frequent mutations in the IDH1 explain why chemotherapy and radiotherapy have
gene. IDH1, when mutated, codes for a protein that been largely ineffective in the treatment of glio-
can contribute to abnormal cell growth (Verhaak blastoma patients. What is more likely is that treat-
et al., 2010). Another gene, platelet derived growth ments, or combinations of treatments, will need
factor receptor alpha (PDGFRA), was mutated and to be administered to distinct subgroups (or sub-
expressed in abnormally high amounts only in the populations) of patients whose tumors harbor the
proneural glioblastomas and not in any other sub- specific molecular alteration(s) and/or mutation(s)
groups. When PDGFRA is altered, too much of its that the medicines can specifically affect (Lathia
protein can be produced, leading to uncontrolled et al., 2015).
tumor growth (Müller et  al., 2016). Unlike the Defeat Glioblastoma Research Collaborative
other groups, whose patients were similar in age (www.defeatgbm.org) is a research group that was
on average, the proneural subgroup was signifi- founded in 2013 for the development of targeted
cantly younger. They also tended to survive lon- treatment of glioblastoma. This group is made up of
ger. However, patients in the proneural group who four “synergistic research cores,” or project teams.
received aggressive treatment did not survive sig- Three of these cores are working simultaneously
nificantly longer than proneural patients who did and the fourth core is focusing on adaptive clinical
 Clinical scenario for precision medicine in brain gliomas  45

trial designs. The aim of Core 1 (Target Discovery) challenges facing precision medicine. In particu-
is to identify targets (the molecular alterations and lar, it is necessary to understand the significance of
mutations) within glioblastoma cells that are driv- whether targeting specific molecular alterations by
ing the growth of these tumors. The aim of Core 2 therapies will be sufficient to have clinical benefit
(Drug Discovery) is to find specific drugs, or com- as not all molecular alterations are closely tied to
binations of drugs, that attack the targets that are tumor growth. Moreover, because of the histologi-
being discovered. The aim of Core 3 (Biomarker cal and molecular complexity of brain gliomas, it
Development) is to identify specific biological is likely that combinations of targeted therapies
processes that occur in glioblastoma cells that can will be required to affect the course of the disease.
predict whether a particular drug, or combination Therapies themselves can produce changes in the
of drugs, will be effective for a patient or group(s) molecular profile of tumors and cause drug resis-
of patients. After researchers identify a promising tance, so we need to be able to identify and block
target for treatment (for example, a mutation in a resistance pathways. In our field, we need to build
specific gene), they find a drug(s) that is designed a solid base of clinical evidence through prospec-
to attack the mutation, and finally identify the tive clinical trials to define the best therapies for
predictive markers that let scientists and doctors specific types of molecular “signatures” found in
know how patients may benefit from the potential brain glioma cells.
new drug. The next step, before a new treatment
regimen is approved for use by patients, is to test CLINICAL SCENARIO FOR
it through the clinical trial process (www.defeat- PRECISION MEDICINE IN BRAIN
gbm.org). GLIOMAS
Johnston et al. (2013) developed a patient-­specific
mathematical model for glioblastoma, yet the inci- A 10-year-old female patient presented with epi-
dence is low, posing difficulty in creating statisti- lepsy and weakness in her left arm and leg lasting
cally powered clinical trials, delaying advances in for 2 days. Magnetic resonance imaging revealed
treatment. This model for disease progression cou- a right parietal contrast enhancing mass lesion at
pled with clinical imaging data may provide a tool the right parietal lobe with mild brain edema and
for quantifying each individual tumor’s response midline shift. In her ophthalmological examina-
relative to its inherent growth kinetics rather than tion, grade 2 papilledema was present. She under-
a universal approach, and, potentially, predict- went surgery and the mass lesion was subtotally
ing treatment efficacy. They used this model to removed. The histopathological diagnosis was
evaluate therapeutic effect of a novel treatment for glioblastoma, which is a rare malignant brain
newly diagnosed glioblastoma compared to stan- glioma in childhood. She received chemotherapy
dard care treatment. Patient-specific mathematical and whole brain radiotherapy after surgery. But the
modeling can be used to evaluate treatment gains tumor reoccurred in the same site 2 years after sur-
in small clinical-trial design. The addition of O6- gery and the neurological condition of the patient
benzylguanine to temozolomide in the postradio- deteriorated. She was reoperated for the tumor and
therapy milieu can provide greater treatment gains the histopathological diagnosis was again glio-
per milligram of temozolomide dose than temo- blastoma. But the molecular analysis of the tumor
zolomide alone (Adair et al., 2014). by PCR amplification and subsequent sequencing
revealed a BRAF V600E mutation and she received
WHY PRECISION MEDICINE IS a BRAF inhibitor “vemurafenib.” The vemurafenib
NEEDED IN BRAIN GLIOMAS therapy was started at 720 mg orally twice daily for
28 days (about 600 mg/m2 per dose). A complete
Repeated surgeries, followed by many rounds of response was observed after 4 months of therapy,
chemo-/radiotherapy in different doses make the and there was no residual or recurrent tumor in the
glioma patients unhappy and depressed. New ther- MRI of the patient 1 year after the second surgery.
apeutic approaches are needed to overcome this This patient was treated with targeted BRAF
problem. As we enter an era that anticipates more inhibitor therapy after the second surgery. If the
tailored therapies for brain gliomas, many brain molecular analysis of the tumor was performed
tumor research groups seek to address several after the first surgery, she would be better and
46  Precision medicine for brain gliomas

might not undergo a second surgery for recur- gliomagenesis genes, targeting miRNA oncogenic
rence of glioblastoma. In addition, this mutation is activity, sensitizing cancer stem cells by HedgehogGli/
not frequent in glioblastoma, but it is well known Akt inhibitors+radiation, and employing tumor
that BRAF inhibitors extend survival and improve lysates as antigen sources for efficient depletion of
the quality of life in patients with BRAF V600E- tumor-specific cancer stem cells by cytotoxic T lym-
mutated melanoma. Vemurafenib is a BRAF phocytes along with conventional strategies will pro-
inhibitor. It was recently approved by the United vide a new paradigm in glioma therapeutics with a
States Food and Drug Administration (FDA) for focus on early diagnosis and successful management
the treatment of melanoma with V600E mutation. (Rajesh et al., 2017).
But, it should not be forgotten that this response Finally, a number of elements should be blended
is unlikely to be uniform for all brain gliomas. for big data to deliver on precision medicine. First,
This case also showed that the identification of the appropriate data sources must be available at scale,
potential responders through careful histologi- whether it is from research cohorts, health care sys-
cal and mutational analysis is critical for targeted tems, or dedicated biobanks. Second, the method
therapy in brain gliomas. of data collection must be robust, with technolo-
gies that extract viable measurements with suffi-
FUTURE DIRECTIONS cient precision and accuracy at high throughput.
Third, a multidisciplinary team with comple-
Despite the continuous development of new che- mentary expertise in bioinformatics, statistics,
motherapeutic agents, the ability to bypass the software engineering, quality control, biological
blood-brain barrier, and acquired drug resis- sample processing, and clinical medicine must be
tance are still constraints (Ataman et  al., 2004; engaged in a collaborative framework, preferably
Chen et al., 2012; Verma et al., 2016). Rather than across jurisdictional and institutional boundaries.
attempting to control the migration of diffuse Fourth, expert panels that collect and assimilate
glioma, interventions that specifically target the the information generated by these efforts must
invasive phenotype should be developed. However, be able to distill it into actionable interventions
recent advancements in neuroimaging and genet- for patient care. Finally, a body of educators must
ics will contribute to early diagnosis and initiation be convened to train both their peers and the next
of glioma management. Advancements in nonin- generation of health care practitioners to dissemi-
vasive imaging protocols and a better understand- nate and implement these interventions, which
ing of glioma biology will enable neuro-oncologists can then be iteratively evaluated for public health
to decipher the molecular, cellular, genetic, and impact.
epigenetic makeup of tumors. This information
might pave the way to personalized glioma thera- CONCLUSION
pies (Levin, 2016).
Stem cell transplantation, nanotechnology, Precision medicine for brain glioma is still under
drug dosing or timing variations, moderation of construction. Many studies on molecular and
immunosuppressive mechanisms, and a better genetic basis of gliomas are needed to better under-
understanding of miRNA and mRNA interactions stand the development of these malignant tumors.
are some of the strategies that may facilitate the In order to perform personalized treatment for
stratification of brain gliomas (Lathia et al., 2015). these patients, the etiology and pathophysiology
Proteogenomic characterization may identify should be well known. In addition, definitive and
molecular drivers and lead to molecular classifica- clear targets should be determined for precise
tion of glioma subgroups. The insights provided by treatment.
omics studies will facilitate identification of glioma
protein signatures and biomarkers as well as the REFERENCES
design of potential treatment regimens for brain
gliomas (Izci, 2014). Adair JE, Johnston SK, Mrugala MM et al. Gene
Future therapies comprising progressive neu- therapy enhances chemotherapy tolerance
rointervention techniques, termination of DNA and efficacy in glioblastoma patients. J Clin
synthesis by prodrug-activating genes, silencing Invest 124(9);2014:4082–92.
 References 47

Akay KM, Baysefer A, Kayali H, Izci Y, and Evans C. TCGA scientists discover four distinct
Timurkaynak E. Glioblastoma multiforme: subtypes of glioblastoma distinguished by gene
Correlation of radiological findings sur- expression patterns and clinical characteristics.
gery and prognosis. Gulhane Med J The Cancer Genome Atlas. https://cancerge-
44(2);2002:142–8. nome.nih.gov/researchhighlights/research-
Ataman F, Poortmans P, Stupp R, Fisher B, and briefs/foursubtypes. Posted February 1, 2011.
Mirimanoff RO. Quality assurance of the Accessed online March 31, 2017.
EORTC 26981/22981; NCIC CE3 intergroup Grant R, Kolb L, and Moliterno J. Molecular
trial on radiotherapy with or without temo- and genetic pathways in gliomas: The future
zolomide for newly-diagnosed glioblastoma of personalized therapeutics. CNS Oncol
multiforme: The individual case review. Eur J 3(2);2014:123–36.
Cancer 40(11);2004:1724–30. Halkin T. Precision medicine, adaptive clinical
Bondy ML, Scheurer ME, Malmer B et al. Brain trials, and defeat GBM. National Brain Tumor
tumor epidemiology: Consensus from the Society. http://blog.braintumor.org/precision-
brain tumor epidemiology consortium (BTEC). medicine-adaptive-clinical-trials-defeat-gbm.
Cancer 113(7 Suppl.);2008:1953–68. Published online February 9, 2015. Accessed
Bush NA, and Butowski N. The effect of molecu- on March 31, 2017.
lar diagnostics on the treatment of glioma. Hayhurst C. Contemporary management of low
Curr Oncol Rep 19(4);2017:26. grade glioma: A paradigm shift in neuro-
Chandran M, Candolfi M, Shah D, Mineharu oncology. Pract Neurol 17(3);2017:183–90.
Y, Yadav VN, Koschmann C, Asad AS, Hodges TR, Ferguson SD, and Heimberger AB.
Lowenstein PR, and Castro MG. Single vs. Immunotherapy in glioblastoma: Emerging
combination immunotherapeutic strate- options in precision medicine. CNS Oncol
gies for glioma. Expert Opin Biol Ther 5(3);2016:175–86.
17(5);2017:543–54. Izci Y. Biomarkers for brain gliomas. In: Cancer
Chen R, Mias GI, Li-Pook-Than J et al. Biomarkers: Minimal and Noninvasive Early
Personal omics profiling reveals dynamic Diagnosis and Prognosis. Barh D, Carpi A,
molecular and medical phenotypes. Cell Verma M, and Gunduz M (Eds), CRC Press/
148(6);2012:1293–307. Taylor & Francis Group, Boca Raton, FL, 2014,
Chen Y, Gao Z, Wang B, and Xu R. Towards preci- pp. 199–218.
sion medicine-based therapies for glioblas- Izci Y, Gurkanlar D, and Timurkaynak E.
toma: Interrogating human disease genomics Multicentric gliomas: Still remains a controver-
and mouse phenotypes. BMC Genomics sial issue. Report of three cases and literature
17(Suppl. 7);2016:516. review. Turk Neurosurg 15(2);2005:71–5.
Crocetti E, Trama A, Stiller C, Caldarella A, Johnston SK, Adair JE, Bridge C et al. Patient-
Soffietti R, Jaal J, Weber DC, Ricardi U, specific mathematical modeling as a
Slowinski J, and Brandes A, RARECARE precision-medicine approach to evaluat-
Working Group. Epidemiology of glial and ing therapeutic gains of a novel chemo-
non-glial brain tumours in Europe. Eur J protection treatment in newly diagnosed
Cancer 48(10);2012:1532–42. glioblastoma. Abstract. J Clin Oncol 31(15;
de Robles P, Fiest KM, Frolkis AD, Pringsheim T, suppl);2013. doi:10.1200/jco.2013.31.15_
Atta C, St Germaine-Smith C, Day L, Lam D, suppl.e13041.
and Jette N. The worldwide incidence and Kouri FM, Ritner C, and Stegh AH. miRNA-182
prevalence of primary brain tumors: A system- and the regulation of the glioblastoma pheno-
atic review and meta-analysis. Neuro Oncol type-toward miRNA-based precision thera-
17(6);2015:776–83. peutics. Cell Cycle 14(24);2015:3794–800.
Duffau H, and Taillandier L. New concepts in the Lathia JD, Mack SC, Mulkearns-Hubert EE,
management of diffuse low-grade glioma: Valentim CL, and Rich JN. Cancer stem cells in
Proposal of a multistage and individual- glioblastoma. Genes Dev 29(12);2015:1203–17.
ized therapeutic approach. Neuro Oncol Levin VA. Personalized medicine in neuro-oncol-
17(3);2015:332–42. ogy. CNS Oncol 5(2);2016:55–8.
48  Precision medicine for brain gliomas

Louis DN, Perry A, Reifenberger G, von Deimling Prados MD, Byron SA, Tran NL et al. Toward
A, Figarella-Branger D, Cavenee WK, Ohgaki precision medicine in glioblastoma: The
H, Wiestler OD, Kleihues P, and Ellison DW. The promise and the challenges. Neuro Oncol
2016 World Health Organization classification 17(8);2015:1051–63.
of tumors of the central nervous system: A sum- Rajesh Y, Pal I, Banik P, Chakraborty S, Borkar SA,
mary. Acta Neuropathol 131(6);2016:803–20. Dey G, Mukherjee A, and Mandal M. Insights
Miller JJ, and Wen PY. Emerging targeted thera- into molecular therapy of glioma: Current
pies for glioma. Expert Opin Emerg Drugs challenges and next generation blueprint.
21(4);2016:441–52. Acta Pharmacol Sin 38(5);2017 Mar 20:591–
Morokoff A, Ng W, Gogos A, and Kaye AH. 613, doi: 10.1038/aps.2016.167, Epub ahead of
Molecular subtypes, stem cells and heteroge- print.
neity: Implications for personalised therapy in Stupp R, Mason WP, van den Bent MJ et al.
glioma. J Clin Neurosci 22(8);2015:1219–26. Radiotherapy plus concomitant and adjuvant
Müller S, Liu SJ, Di Lullo E et al. Single-cell temozolomide for glioblastoma. N Engl J Med
sequencing maps gene expression to muta- 352(10);2005:987–96.
tional phylogenies in PDGF- and EGF-driven Tan SY, Sandanaraj E, Tang C, and Ang BT.
gliomas. Mol Syst Biol 12(11);2016:889. Biobanking: An important resource for preci-
Oh YT, Cho HJ, Kim J et al. Translational valida- sion medicine in glioblastoma. Adv Exp Med
tion of personalized treatment strategy based Biol 951;2016:47–56.
on genetic characteristics of glioblastoma. Verhaak RG, Hoadley KA, Purdom E et al.
PLOS ONE 9(8);2014:e103327. Integrated genomic analysis identifies
Olar A, and Aldape KD. Using the molecular clas- clinically relevant subtypes of glioblas-
sification of glioblastoma to inform personal- toma characterized by abnormalities in
ized treatment. J Pathol 232(2);2014:165–77. PDGFRA, IDH1, EGFR, and NF1. Cancer Cell
Ostrom QT, Bauchet L, Davis FG et al. The epide- 17(1);2010:98–110.
miology of glioma in adults: A “state of the sci- Verma A, Gunasekar S, Goel V et al. A molecular
ence” review. Neuro Oncol 16(7);2014:896–913. approach to glioblastoma multiforme. Int J
Ostrom QT, Gittleman H, Kruchko C, Louis DN, Mol & Immuno Oncology 1(1);2016:35–44.
Brat DJ, Gilbert MR, Petkov VI, and Barnholtz- Walid MS. Prognostic factors for long-term
Sloan JS. Completeness of required site-spe- survival after glioblastoma. Perm J 12(4);2008
cific factors for brain and CNS tumors in the Fall:45–8.
Surveillance, Epidemiology and End Results www.defeatgbm.org. Accessed on March 21,
(SEER) 18 database (2004-2012, varying). 2017.
J Neurooncol 130(1);2016:31–42. Zhang AS, Ostrom QT, Kruchko C, Rogers
Perry A, and Wesseling P. Histologic classification L, Peereboom DM, and Barnholtz-Sloan
of gliomas. Handb Clin Neurol 134;2016:71–95. JS. Complete prevalence of malignant
Porter KR, McCarthy BJ, Freels S, Kim Y, and primary brain tumors registry data in
Davis FG. Prevalence estimates for primary the United States compared with other
brain tumors in the United States by age, common cancers, 2010. Neuro Oncol
gender, behavior, and histology. Neuro Oncol 19(5);2016:726–35.
12(6);2010:520–7. Zong H, Verhaak RG, and Canoll P. The cellular
Pouratian N, and Schiff D. Management of low- origin for malignant glioma and prospects for
grade glioma. Curr Neurol Neurosci Rep clinical advancements. Expert Rev Mol Diagn
10(3);2010:224–31. 12(4);2012:383–94.
 5
Precision medicine for colorectal cancer

CANDAN HIZEL, ŞÜKRÜ TÜZMEN, ARSALAN AMIRFALLAH,

GIZEM ÇALIBAŞI KOÇAL, DUYGU ABBASOĞLU, HALUK ONAT,
YEŞIM YILDIRIM, AND YASEMIN BASKIN

Introduction: Overview of epidemiology/ 5-FU PGx 71

incidence, etiology, and precision Targeted therapies: PGx for somatic
medicine for colorectal cancer (CRC) mutations in anticancer agents 74
genomics in “postgenomic” era 50 Targeting genes 74
CRC epidemiology 50 Targeting the pathways 77
Genetic and nongenetic paradigms of CRC 50 From genome-wide association study
Implementation of biomarker-driven colon (GWAS) to genetic risk score (GRS)
cancer therapy: Prognostic and predictive for CRC 80
biomarkers 52 Precision/personalized medicine
Microsatellite instability (MI) and approaches for colon cancer driven by
chromosomal instability (CIN) 53 systems biology 81
18q loss of heterozygosity (LOH) 54 Microbiome-driven carcinogenesis in
TP53 in CRCs 55 colorectal cancer (metagenomics) 81
PIK3CA and PTEN in CRC 56 Metabolomics of colon cancer tissues
BRAF in CRC 57 reflects “metabolic codes” of patients 83
KRAS in CRC 59 Glycomics, as part of an array of metastatic
Immune-related markers in CRC 60 codes 83
Protease related markers in CRC: Focus on Interactomics in colon cancer tissues: Where
cathepsin B (CATSB) 62 are we? 84
Telomere length in CRC 63 High-throughput genomic technologies for
CRC epigenetics 63 detecting potential “driver” genomic
DNA methylation 63 alterations as a risk assessment tool and
Histone modifications 64 for treatment selection in CRC 85
microRNAs 65 Next-generation analysis of CRC genomes
Pharmacogenetics (PGx) for precision for precision medicine 85
medicine in CRC chemotherapy 66 Microarrays in diagnostics and treatment
Toxicity-related biomarkers: PGx for germ-line selection in CRC 86
variations in anticancer agent 68 Conclusion and perspective 86
Irinotecan PGx (irinogenetics): Irinotecan References 87
and UGT1A1 68

49
50  Precision medicine for colorectal cancer

INTRODUCTION: OVERVIEW OF heterogeneous molecular profile in the distal colon

EPIDEMIOLOGY/INCIDENCE, and rectum (Ballester et  al., 2016; Connell et  al.,
ETIOLOGY, AND PRECISION 2017). It is anticipated that increased screening for
MEDICINE FOR COLORECTAL CRC has been associated with a decreased inci-
dence in the past two decades (Stracci et al., 2014),
CANCER (CRC) GENOMICS IN
albeit so far no evidence supports that screening
“POSTGENOMIC” ERA of average-risk individuals less than 50 years old
will lead to early detection and increased survival.
CRC epidemiology
Hence, together with clinicopathologic features,
Colorectal cancer (CRC) is a multifactorial understanding of the distinctive heterogeneous
polygenic complex disorder that represents the and poorly clarified molecular profile, including
supremacy of health care leading to the third most genetic and epigenetic mechanism of young-onset
commonly diagnosed malignancy among both CRC, will be crucial to tailor specific screening
gender groups (behind lung and breast cancer) and management strategies (Stigliano et al., 2014;
with 1.4 million cases in 2012 (Ferlay et al., 2015) Ballester et al., 2016; Tezcan et al., 2016).
and the fourth leading cause of cancer deaths Although several screening tests are now available,
worldwide. It accounts for approximately 700,000 at present none of them has been proven the best one
deaths per year, affecting men and women almost (Pox, 2014). Although colonoscopy is usually used in
equally (third most commonly diagnosed malig- high-risk subjects with family history of either CRC
nancy for men behind lung and prostate cancer, or adenomas, fecal occult blood test (FOBT) is pre-
and the second most commonly diagnosed malig- ferred (Health Quality Ontario, 2009) in interme-
nancy for women behind breast cancer) (Ferlay diate-risk subjects. Colonoscopy is performed only
et al., 2015; Siegel et al., 2018) with the highest inci- when FOBT results are positive (Stracci et al., 2014).
dence rates in high-income countries (Ferlay et al., For comprehensive reviews of CRC screening, the
2015; Arnold et  al., 2017; Bhandari et  al., 2017). reader is referred to articles by Issa and Noureddine
However, CRC incidence is also rapidly increas- (2017) and Bhurgri and Samiullah (2017).
ing in low- and middle-income countries (LMICs) Although CRC occurs in all ethnic groups and
(Bhandari et al., 2017). It is expected that its bur- races, variation in occurrence and survival rate
den will be increased by 60% with more than 2.2 (stage-specific survival) in different ethnic groups
million new cases and 1.1 million cancer deaths by and races is underlined (Wallace et  al., 2016;
2030 (Arnold et al., 2017). Survival of CRC patients Fedewa et al., 2017). The American Cancer Society
mostly depends on the disease stage at diagnosis, (ACS) documented that racial and ethnic minori-
and approximately 20%–25% of patients with CRC ties are more likely to develop cancer, particularly
already have liver metastases at the time of diagno- CRC and die from it when compared to the gen-
sis and 20%–30% of patients will develop metasta- eral population of the United States (Jackson et al.,
ses subsequently during disease progression with 2016). Accordingly, the high incidence and mortal-
stage II/III (Loree and Kopetz, 2017). ity rate in African American compared to white
Despite these bleak statistics, due to progress Americans in the United States is reported (Jackson
in high-throughput techniques in molecular biol- et al., 2016; Fedewa et al., 2017). However, Hispanic
ogy from genetic sequencing to targeted therapies Americans and Native Americans (Alaska Natives)
has led to the availability of more precise and per- have the lowest incidence rate.
sonalized screening practices leading to a gradual
decrease in both CRC incidence and mortality over Genetic and nongenetic paradigms
the past few decades particularly for late-onset CRC of CRC
in adults aged between 50 and 75 (Siegel et al., 2018).
On the flip side of this, most research publications The vast majority of tumors such as CRC encoun-
have reported a steady increase in incidence rates tered in human pathology are carcinomas, that
in both hereditary and sporadic forms of young- is, tumors of epithelial origin that generally
onset CRC (diagnosed at ≤40 years old) diagnosed develop in several phases (Vogelstein et al., 2013).
with advanced-stage tumors having distinctive Additionally, cancer such as CRC is a genetic
 Introduction 51

disease under the influence of the environmen- associated with approximately 15%–20% of all
tal/lifestyle factors that is developed in a multi- sporadic CRC cases (Issa, 2008; Wiesner et  al.,
step process resulting from an accumulation of 2010). As for CIMP, besides genomic instability
a series of genetic changes leading to gain/loss of CIN and MSI, a broad spectrum of studies under-
gene function, such as mutations in DNA repairing line importance of a crosstalk between genetic
genes, in oncogenes and tumor suppressor genes, and epigenetic mechanisms in cancer formation,
which regulate normal cellular function in the including CRC, indicating that genetic aberran-
genome of cells (Vogelstein et al., 2013). Moreover, cies such as CIMP that are inherited or acquired
once a cancer has begun its anarchic growth it during a lifetime have the potential of disrupt-
will pick up more and more mutations (Hanahan ing epigenetic patterns, which in turn epigen-
and Weinberg, 2011; De Palma and Hanahan, etic modifications can drive genome instability
2012). Thus, understanding the molecular basis of and mutagenesis (Pancione et  al., 2012; You and
tumor-associated genetic and epigenetic modifi- Jones, 2012; Hughes et al., 2013; Nazemalhosseini
cations that alter gene expression patterns is cru- Mojarad et  al., 2013). Additionally, other numer-
cial (De Palma and Hanahan, 2012; Pierotti, 2017; ous factors are important in the development of
Inamura, 2018). CRC, including colonocyte metabolism, high-risk
CRC is one of the best molecularly character- luminal environment, inflammation, and envi-
ized neoplasms and it is important to note that ronmental and lifestyle factors such as such as diet
most of the important theoretical frameworks for (Aguirre-Portolés et  al., 2017), physical inactivity
understanding the basis of carcinogenesis have (Namasivayam and Lim, 2017), cigarette smok-
grown from genomic research in CRC (Vogelstein ing (Nishihara et al., 2013; Drew et al., 2017), and
et  al., 2013). As formulated in the 1990s by alcohol consumption (may be relative to MTHFR
Vogelstein and Fearon (Vogelstein et  al., 1988, genotype status) (Kim et al., 2012; Svensson et al.,
1989; Fearon and Vogelstein, 1990), it is widely 2016). All these potentially modifiable/behavioral
accepted that in most cases carcinoma arises from risk factors, which are a valuable tool for setting
preexisting adenomas and every step from normal priorities for CRC prevention and control, could
mucosa toward the carcinoma involves specific affect epigenetic process (Aleksandrova et  al.,
well-known genetic (conversion of proto-oncogene 2014; Rattray et  al., 2017). Likewise, disease con-
to oncogenes, and/or inactivation of tumor sup- ditions, such as inflammatory bowel disease (IBD)
pressor genes; Wnt, P13K, and RAS-MAPK path- (Du et al., 2017), existing of previous polyps (Jelsig,
ways with KRAS, NRAS, and BRAF genes) (Smith 2016), diabetes (González et  al., 2017; Yang et  al.,
et al., 2002) and epigenetic alterations (DNA meth- 2017), obesity (Heo et al., 2004), and nonalcoholic
ylation, histone modifications, noncoding RNAs/ fatty liver disease (NAFLD) (Mikolasevic et  al.,
miRNAs) (Bardhan and Liu, 2013; Danese and 2017). On the other hand, nonmodifiable risk fac-
Montagnana, 2017). Hence, all these alterations tors, such as gender, increasing age, personal his-
are driving forces of CRC carcinogenesis, which tory of adenomatous polyps and inflammatory
are associated with the stages of CRC progression. bowel disease, ethnicity/race, and genetic inheri-
Moreover, the connection between inflammation tance, are associated with the incidence of colorec-
(through immune cells, cytokines) and colorectal tal cancer. Contrary to modifiable risk factors that
carcinogenesis, including initiation, promotion, could theoretically be avoided, nonmodifiable risk
progression, and metastasis, has been established factors are not considered part of the “environ-
(Terzić et al., 2010). mental nature” of this disease. Therefore, they are
It is anticipated that CRC can arise from one not controllable; however, they play a crucial role
or a combination of genetic alterations of chro- in screening and identifying susceptible individu-
mosomal instability (CIN), DNA microsatellite als for CRC (Thélin and Sikka, 2015).
instability (MSI), and CpG island methylator phe- The majority of CRC cases is sporadic (approxi-
notype (CIMP). Accordingly, a majority of CRCs mately 70%–75% of all CRCs), with no known
develop via CIN, which accounts for about 80%– genetic predisposing factors for disease and about
85% of sporadic CRC and it is also associated with 80% are caused by somatic defects in the CIN
familial adenomatous polyposis (Wiesner et  al., pathway through the KRAS, APC, and TP53 muta-
2010; Roper and Hung, 2013). MSI phenotype is tions (Wiesner et al., 2010; Roper and Hung, 2013).
52  Precision medicine for colorectal cancer

The microsatellite instability (MSI) pathway, which susceptibility to common multifactorial complex

is caused by DNA mismatch repair (MMR), repre- diseases such as cancer as well as in drug effec-
sents a major pathway for inherited and familial tiveness and safety for more precise personalized
form (Boland and Goel, 2010; Pino and Chung, health care (Hamet, 2012; Kittles, 2012; Hizel
2011). The inherited form, which represents et  al., 2017). To this end, the interplay between
approximately 15%–20% of all CRCs, is more likely genetic and environmental factors in CRC pathol-
to be the result of low-penetrance alleles (with the ogy should be taken into consideration. Together
possibility of some undiscovered high- and moder- with recent knowledge of the human genome,
ate-penetrance gene) at several genetic loci (Mishra conventional trio therapies—radiotherapy, che-
and Hall, 2012; Lynch and Shaw, 2013; Hahn et al., motherapy, surgery—in CRC and all other types
2016). Well-described forms of hereditary CRC of cancer are all still improving and have become
include familial adenomatous polyposis (FAP) more efficient in recent years. Accordingly, the
(Jaiswal et  al., 2005; Half et  al., 2009; Lynch and tools of molecular biology have moved out of the
Shaw, 2013; Hahn et  al., 2016) and another auto- lab and into the clinic, and recent progress in high-
somal dominant disorder, hereditary nonpolypo- throughput technologies such as next-generation
sis colorectal cancer (HNPCC) or Lynch syndrome sequencing, as well as “omics” technologies, such
(approximately 5%–10% of the total CRC), which is as transcriptomics, proteomics, metabolomics,
highly penetrant with an 80% lifetime risk for CRC epigenomics, and metagenomics, have untangled
(Lynch and Shaw, 2013). Certain other rare polyp- the genomic architecture and the molecular het-
osis syndromes such as hamartomatous polyposis erogeneity of CRC and a new branch of therapy
syndromes comprise less than 1% of all hereditary has sprung up for more personalized and preci-
colorectal cancer (Campos et al., 2015; Jelsig, 2016). sion medicine in cancer (Bhati et al., 2012).
Being located predominantly in the right colon
(Iacopetta, 2002), FAB is an autosomal dominant IMPLEMENTATION OF BIOMARKER-
disorder due to the presence of a mutated copy of DRIVEN COLON CANCER THERAPY:
the adenomatous polyposis gene (APC) located on PROGNOSTIC AND PREDICTIVE
chromosome 5 (Nishisho et al., 1991). Tumor sup- BIOMARKERS
pressor APC is considered a “gatekeeper” gene that
is responsible for controlling, or inhibiting, cell In the era of precision medicine, with the prog-
growth (Deininger, 1999; Eshghifar et  al., 2017). ress of molecular genomics and high-throughput
Whereas, being an autosomal dominant disor- omics-based technologies, biomarker studies hold
der, HNPCC, which occurs predominantly in the promise for early and effective prediction of sus-
right colon and cannot be differentiated endo- ceptibility of disease and treatment efficacy/toxic-
scopically from sporadic colon polyps (Iacopetta, ity, diagnosis, and prognosis for a tailored health
2002; Rijcken et al., 2002; Jang and Chung, 2010), is care approach to an individual (Barh and Azevedo,
caused by inheritance of defective DNA mismatch 2017; Kamel and Al-Amodi, 2017; Wilson and
repair genes MLH1, MSH2, PMS2, and MSH6 Altman, 2017).
(Jaiswal et al., 2005; Lynch and Shaw, 2013; Hahn The term biomarker relates to a measurement
et al., 2016). They are “caretaker” genes involved in variable of biological molecules that is associated
the maintenance of the genome stability including with disease and treatment outcome (Ballman,
mismatch repair of DNA (Deininger, 1999). 2015). Respectively, exploitation of precision
Taken as a whole, even if genomic studies are medicine in oncology demands a precise under-
one of the principal components for identifying standing of disease progression as well as proper
the “Achilles’ heel” for the subject affected by the application of molecular targeting of therapies
disease, it is only one piece of the puzzle (Hizel (Hizel et  al., 2017). However, since not all bio-
et  al., 2017). Thereby, the creation of robust data markers are the same, there is noteworthy con-
from genomic risk to the “exposome,” together fusion about the distinction between a predictive
with medical histories, social factors, and lifestyle biomarker and a prognostic biomarker (Ballman,
factors, is pivotal and they cannot act in isola- 2015). To this end, while prognostic biomark-
tion. But they can act in concert in order to ful- ers are associated with the clinical disease out-
fill the prior engagement of precision medicine in come that is independent of treatment, predictive
 Implementation of biomarker-driven colon cancer therapy: Prognostic and predictive biomarkers  53

biomarkers are measures of the possibility of Having few or no somatic copy number altera-
response (responder/nonresponder) of a partic- tions (CNAs) tumors with dMMR/MSI pheno-
ular therapy as well as identification of patients type occur in approximately 15%–20% of CRC
most likely to benefit from a given treatment patients of which 12% in sporadic CRC tumors and
without toxicity (Oldenhuis et al., 2008; Carethers 3% are associated with Lynch syndrome (Boland
and Jung, 2015; Verdaguer et al., 2017). It is also and Goel, 2010). Contrary to microsatellite stable
anticipated that measurements coming from the MSS tumors (also termed chromosomal instabil-
signature of combined multiple biomarkers have ity), MSI is attributable to DNA mismatch repair
stronger predictive power compared to a single genes in sporadic and germ line tumors (Boland
measurement (Ballman, 2015). and Goel, 2010; Lee and Chan, 2011; Sinicrope and
The following will review recent knowledge in Sargent, 2012; Kawakami et al., 2015).
the application of molecular biomarker in CRC Most of the sporadic CRCs with MSI tumors
pathogenesis and treatment. occur generally in older individuals who carry
somatic mutations in the BRAF oncogene (V600E)
Microsatellite instability (MI) and in approximately half of cases (Boland and Goel,
chromosomal instability (CIN) 2010; Kawakami et al., 2015). Whereas, while MSI
tumors in sporadic CRCs are due to transcrip-
Being important in the carcinogenesis pathway in tional silencing/epigenetic inactivation (hyper-
many cancers including CRC, microsatellite insta- methylation) of the DNA mismatch repair MLH1
bility (MSI) is the state of genetic hypermutability gene promoter together with the CpG island meth-
(highly predisposed to mutation) (Strauss, 1998) ylator phenotype (Issa, 2004; Grady and Carethers,
due to impaired DNA base MMR error such as 2008; Barault et  al., 2008a; Nosho et  al., 2008),
frameshift mutations and base-pair substitutions MSI tumors in Lynch syndrome, which occurs in
during DNA replication (S phase of the cell cycle) younger individuals carrying KRAS mutations (but
in the DNA MMR allowing accumulation of muta- never BRAF mutations), are caused by germ line
tions in the DNA (Popat et al., 2005). The presence mutations in DNA mismatch repair genes: MLH1,
of MSI is an indication of phenotypic evidence that MSH2, MSH6, and PMS2 (Boland and Goel, 2010;
MMR is not functioning properly. Under normal Lee and Chan, 2011; Sinicrope and Sargent, 2012;
conditions, dMMR corrects spontaneously occur- Kawakami et al., 2015). Respectively, this distinc-
ring DNA polymerase errors during DNA repli- tion concerning methylation status suggests that
cation; however, cells having an impaired dMMR most of the CRCs with sporadic MSI tumors come
system are unable to correct errors leading to from CIMP status creating an important distinc-
accumulation of blunders resulting in creation of tion from Lynch syndrome (Barault et al., 2008a;
novel microsatellite fragments that are repeated Boland and Goel, 2010); however, inheritance of
sequences of DNA (ranging in length from 1 to a cancer-associated MLH1 germ line epimuta-
6 or more base pairs) (Boland and Goel, 2010; tion (abnormal gene silencing) is demonstrated
Kawakami et al., 2015; Inamura, 2018). (Hitchins et  al., 2007). Consequently, presence of
The MMR and CIN (accounting for about 80%– the BRAF V600E mutation in sporadic CRCs with
85% of sporadic CRCs) have little overlap in CRC MSI tumors essentially excludes Lynch syndrome
tumorigenesis (Kawakami et al., 2015). CIN results (Funkhouser et  al., 2012), with the exception of
from defects in the chromosomal segregation rare cases related to germ line mutation in the
process characterized by an imbalance in chro- PMS2 gene that encodes mismatch repair endonu-
mosomal number (aneuploidy), subchromosomal clease PMS2 (Nicolaides et al., 1994; Senter et al.,
genomic amplifications, and a high frequency of 2008; Gatalica et al., 2016).
loss of heterozygosity (LOH) and changes in chro- It anticipated that in CRC, tumor location to the
mosomal copy number through classic adenoma- splenic flexure, left side (distal) (LC) versus right
to-carcinoma progression (alterations in the Wnt side (proximal) (RC) of the colon affects colorectal
and TGF-β signaling pathways, activation of KRAS, carcinogenesis (Sugai et  al., 2006), and incidence
and inactivation of APC and TP53) (Vogelstein of location change according to geographic region,
et al., 1989; Fearon and Vogelstein, 1990; Lengauer age, and gender (Iacopetta, 2002). Accordingly,
et al., 1997; Dienstmann et al., 2017). recent studies have revealed that CRCs with an
54  Precision medicine for colorectal cancer

MSI phenotype with frequent BRAF mutations, et al., 2013; Webber et al., 2015; Kawakami et al.,
infrequent TP53 mutation, poorly differentiated 2015; Fujiyoshi et  al., 2017) but may be sensitive
histology, abundant tumor-infiltrating lympho- to irinotecan (Fallik et al., 2003; Pino and Chung,
cytes (TILs), and CIMP-high status preferentially 2011) and mitomycin C (Devaud and Gallinger,
occur on the RC (proximal portion) of the colon. 2013). Also, there is evidence to suggest that MSI-H
Whereas CRC with an MSS phenotype character- patients respond to surgery alone better than com-
ized by frequent copy number alterations, CIMP- bined surgery and chemotherapy and, therefore
low status, TP53 mutation, and is commonly found preventing patients from needless chemotherapy
on the LC (distal portion) of the colon (Sugai et al., exposure (Buecher et  al., 2013; Kawakami et  al.,
2006; Takahashi et al., 2016), suggesting that tumor 2015). Additionally, according to recent stud-
location may be associated with the development ies, colorectal tumors with MSI-H have increased
of CRC. Moreover, compelling evidence from pub- immunogenicity leading to positive response to
lished studies suggest that besides poorer survival immunotherapy compared to MSS (Xiao and
characteristics of RC compared to LC, patients Freeman, 2015; Bashir and Snook, 2017).
with an RC tumor may respond poorly to antiepi- In summary, the MSI-H phenotype is an
dermal growth factor receptor (EGFR) antibod- important feature of the molecular pathogenesis
ies, such as cetuximab or panitumumab (Meguid of cancer that may provide an important basis
et al., 2008; Brule et al., 2015; Kim et al., 2017). for novel diagnostic and personalized therapeutic
Though not all studies have confirmed the prog- approaches (Zhang et al., 2016).
nostic value of MSI, most studies have emphasized
that CRC patients with an MSI tumor with greater 18q loss of heterozygosity (LOH)
numbers of cytotoxic TILs (Michael-Robinson
et al., 2001; Phillips et al., 2004) have been strongly Loss of heterozygosity (LOH) is a cross-chromo-
associated to longer survival as compared with somal event due to loss of one parent’s contribution
proficient-MMR tumors including microsatellite to the cell, which may be caused through a variety
stable (MSS) and MSI-low (MSI-L) CRCs (Sankila of pathways, such as mitotic recombination, inap-
et  al., 1996; Popat et  al., 2005; Sinicrope and propriate repair of DNA, deletion, gene conversion,
Sargent, 2012; Cortes-Ciriano et al., 2017) particu- or chromosomal loss leading to loss of the entire
larly among young patients. gene along with the surrounding chromosomal
According to an international consensus meet- region (Ryland et al., 2015). During LOH the loss of
ing in 1997 (Boland et  al., 1998) the definition one allele leaves the other allele hemizygote (pos-
of MSI was standardized as MSI-high (MSI-H) session of only one allele at a given locus) (Ryland
(resulting from MSI of greater than 30% of unsta- et  al., 2015). Clinical manifestations of LOH and
ble MSI biomarkers) and MSI-low (MSI-L) or mic- MSI are different in colorectal cancer patients.
rosatellite stable (MSS) (resulting from less than High-frequency LOH is associated with high met-
30% of unstable MSI biomarkers) (Boland et  al., astatic potential of colorectal cancers (Roper and
1998; Pawlik et  al., 2004). In CRC, recent studies Hung, 2013). As it was first reported in 1989 by
revealed another type of MSI called “elevated mic- Vogelstein and his colleagues, LOH of tumor sup-
rosatellite alterations at selected tetranucleotide pressor genes (TSGs) is believed to be one of the
repeats” (EMAST) (Lee et  al., 2012; Venderbosch crucial steps to the carcinogenesis of colorectal
et al., 2015; Watson et al., 2016). Although scientific cancer. Respectively, the interrelationship between
data from the available literature is not consistent LOH, loss of TSG initiation, and progression
and still, additional work needs to be done, MSI-H of CRC is demonstrated (Ozaslan and Aytekin,
status is also associated with a different response 2009). As proven in several studies, mitotic cross-
to classic chemotherapeutic treatment modalities over occurs more frequently in tumor cells and an
(Devaud and Gallinger, 2013; Kawakami et  al., increased mitotic crossover index in colorectal car-
2015; Fujiyoshi et  al., 2017). Several studies high- cinoma cells have been also confirmed (Rovcanin
lighted MSI-H status as a strong predictive factor et al., 2014). Thus, mitotic crossover is underlined
for nonresponse (resistant) to adjuvant 5-fluoro- as one of the principal sources of increased LOH in
uracil (5-FU) chemotherapy (Ribic et  al., 2003; the tumor cells of patients with CRC (LaFave and
Des Guetz et al., 2009; Jensen et al., 2009; Buecher Sekelsky, 2009).
 Implementation of biomarker-driven colon cancer therapy: Prognostic and predictive biomarkers  55

As is well known, inactivation of TSGs is one of of studies are controversial, deletion of 18q has
the most crucial steps in both sporadic and famil- been demonstrated in most studies as an impor-
ial cancer development (Payne and Kemp, 2005). tant step in the development and the progression
Equally, LOH, which is the most common molecu- of CRC tumorigenesis (Jen et al., 1994; Wang et al.,
lar genetic alteration in a variety of human cancers 2010). For example, among the genes located on
including CRC, indicates the presence of a TSG. 18q, of particular importance is for DCC (deleted
Since TSGs are recessive genes (an exception is the in colorectal carcinoma) encoding for a neutrin-1
X-linked tumor suppressor such as Wilms tumor receptor that is important in apoptosis, cell adhe-
gene on the X chromosome), two alleles must be sion and tumor suppression (Mazelin et al., 2004;
inactivated in order for a TSG to lose its inhibi- Kefeli et  al., 2017). The other one is SMAD-4,
tor effect on uncontrolled cell proliferation (Payne which encodes for a nuclear transcription fac-
and Kemp, 2005). Therefore, a “first hit” such as tor in transforming growth factor-β1 (TGF-beta)
physical loss of one copy of an allele is not a trigger signaling involved in tumor suppression (Zauber
for the loss of the second allele, yet it renders one et al., 2008). Given the higher prevalence of Smad4
allele inactive. Thus, in most of the cancer as pro- mutations in CRCs, specifically in those with dis-
posed in Knudson’s “two-hit” hypothesis (Devilee tant metastases compared to locally advanced
et al., 2001), the second copy (the remaining wild tumors, a strong prognostic value of chromosome
type allele) of a tumor suppressor gene has been 18q deletion and SMAD-4 protein inactivation is
inactivated by a “second hit” such as point muta- underscored (Tanaka et  al., 2008; Zauber et  al.,
tion or hypermethylation (Paige, 2003; Payne and 2008; Wang et al., 2010; Jia et al., 2017).
Kemp, 2005). To this end, in cancer, LOH analysis Taken together although evidence from most
is the common method to identify genomic poten- published studies suggest that high-frequency 18q
tial regions harboring tumor suppressor genes by LOH with high metastatic potential of CRC could
noting the presence of heterozygosity at a genetic be prognostic genetic markers, large-scale pro-
locus in an organism’s germ line DNA, as well as spective randomized control trials are needed to
to characterize different tumor types, pathological confirm the final verdict.
stages, and progression (Zheng et al., 2005; Cacev
et al., 2006). TP53 in CRCs
Since LOH is a pivotal characteristic of CIN-
positive tumors, determination of LOH distin- The tumor suppressor TP53 gene is located on the
guishes the tumors arising from the CIN pathway short arm of chromosome 17 (17p13.1). It is com-
from tumors with the dMMR/MSI. Additionally, posed of 11 exons and encodes a 53 kD nuclear
in CRC the association of a worse disease prog- phosphoprotein p53, which is a transcription fac-
nosis with small and large allelic loss is indicated tor interacting with as many as 106 other proteins
(Brosens et al., 2010; Roper and Hung, 2013). The within the cell (Isobe et  al., 1986; McBride et  al.,
interrelationship between CRC tumorigenesis and 1986). Even though TP53 gene germ-line muta-
LOH at various loci on different chromosomes, tions cause inherited syndrome de Li–Fraumeni
such as 1p, 1q, 4q, 5q, 8p, 9q, 10p, 11q, 12p, 14q, (LFS) (OMIM #151623) (60%–80% of “classic”
15q, 17p (loss of p53), 17q, 18p, 18q, and 22q, is LFS families) (Varley, 2003; Malkin, 2011; Merino
emphasized in a large number of such studies sug- and Malkin, 2014), acquired mutations occur in
gesting that LOH on those chromosome regions a wide variety of cancer types including CRC for
anchoring TSGs is essential for colorectal tumori- which acquired TP53 mutations is significant as
genesis mechanisms (Shima et al., 2005; Wan et al., a late event from adenoma to carcinoma (Fearon
2006; Ozaslan and Aytekin, 2009; Mayrhofer et al., and Vogelstein, 1990). The p53 protein plays a cru-
2014). In particular, several studies have confirmed cial role in tumor suppression and the loss of its
that chromosome 18q is commonly deleted in function is required for cancer progression (Surget
most cancers, such as prostate, esophageal, blad- et  al., 2013). Though the p53 protein is encoded
der carcinomas, and sporadic CRC (Jen et  al., by wild-type TP53 tumor suppressor gene, which
1994; Wang et  al., 2010). Since the chromosome is in an inert and unstable state in normal cells in
18q contains several potentially important genes the absence of stresses, its transcriptional activity
implicated in CRCs, and although the outcomes and stability are significantly induced in response
56  Precision medicine for colorectal cancer

to a variety of signals, such as oncogenic stresses and inconclusive results related to its role in clini-
and genotoxicity (Ko and Prives, 1996; Vogelstein cal practice (Lee and Chan, 2011), the importance
et al., 2000), leading to control of transcription of of abnormal p53 protein expression as well as
many genes through senescence of DNA damage, somatic mutation with poor survival prognosis in
and G1/S cell cycle arrest by stimulating p21Waf1 late stage CRC or lack of response to therapy are
(cyclin-dependent kinase inhibitor 1) protein underlined by several studies (Russo et al., 2005;
(Taylor and Stark, 2001). Accordingly, once DNA Iacopetta et al., 2006; Robles and Harris, 2010; Li
repair is successfully complete, the p53 reactivates XL et al., 2015). A recent meta-analysis has under-
the cell cycle; yet if repair fails, programmed cell lined biologically different molecular pathways
death apoptosis is triggered to eliminate damaged between bowel disease-associated colorectal can-
cells (Amelio et al., 2016). The TP53 gene is among cer (IBD-CRC) and sporadic colorectal cancer
the most frequently mutated in cancer having (S-CRC), with IBD patients having TP53 muta-
mostly missense mutations (>75% of all p53 altera- tion more likely to develop IBD-CRC (Du et  al.,
tions) including the hot spot mutations (R175H, 2017). It has also been demonstrated that TP53
R248W and R273H) (Muller and Vousden, 2013; mutation was associated with more advanced
Leroy et al., 2014) which cause loss of p53-depen- stage cancer (Dukes class A, B, C) in patients with
dent tumor suppressor function, but also gains IBD-CRC suggesting that TP53 status may be an
new oncogenic function (gain-of-function muta- important biomarker to improve cancer and dys-
tions) in order to promote cancer as well as drug plasia screening among patients with IBD (Du
resistance suggesting an important implication on et al., 2017). It is anticipated that the standardized
cancer therapy (Xu, 2008; Liu et al., 2010; Liu et al., immunohistochemical (IHC) procedure to evalu-
2014; Alexandrova et al., 2017). ate p53 overexpression could be beneficial for CRC
Tp53 is an enigmatic cancer gene due to patients under a standard chemotherapy regime to
its caretaker/gatekeeper character (Rubbi and get optimal response (Iacopetta, 2003; Akshatha
Milner, 2007). Its caretaker function is related to et al., 2016; Vandana et al., 2017).
its capacity to arrest cell cycle and/or apoptosis In summary, although TP53 targeted therapies
in response to DNA damage that points to P53 are still in the early phases of testing, research in
as the “guardian of the genome” for maintaining this field is expanding rapidly. Respectively, Tp53
genomic stability, and its gatekeeper function is having dual caretaker–gatekeeper functions,
related to its role in controlling cell proliferation deserves additional prospective investigations in
during which mutations in the Tp53 gene result CRC for the development of personalized thera-
in disruption of cell cycle control mechanisms peutics (Surget et al., 2013; Sorrell et al., 2013).
(Levine, 1997; Deininger, 1999; Rubbi and Milner,
2007). Moreover, a large number of such studies PIK3CA and PTEN in CRC
reported that a guanine (G)/cytosine (C) common
single SNP (rs1042522) at the second position of Phosphatidylinositol-3-kinase (PI3K) (also called
codon 72 in exon 4 of TP53 gene (Arg72Pro amino phosphatidylinositol-4,5-bisphosphate 3-kinase)
acidic substitution) may modulate individual can- is one of the pivotal kinase enzymes in the PI3K/
cer susceptibility. AKT1/MTOR intracellular signaling pathway,
Since mutation of the TP53 gene is thought playing a crucial role in cell cycle regulation, such
to increase the protein half-life that is associ- as cellular growth, and proliferation as well as sur-
ated with overexpression in the nucleus and vival of various solid tumors (Shaw and Cantley,
cytoplasm, most translational studies directed 2006; Manning and Cantley, 2007). It has regu-
their efforts at determining whether p53 muta- latory (p85) and a catalytic (p110) subunit PI3K
tion and overexpression have prognostic value phosphorylate phosphatidylinositol, which is a
in CRC (Remvikos et  al., 1990). The majority of phospholipid and an important cell membrane
translational studies carried out in the 1990s were constituent in cell signaling as second messenger
aimed at determining whether p53 mutation and (Manning and Cantley, 2007). The catalytic sub-
overexpression have prognostic value in CRC. unit of PI3K p110α is encoded by a PICK3CA gene
Though p53 research is one of the controversial that is localized on chromosome 3q26.32 (Volinia
areas and still remains unclear with inconsistent et al., 1994).
 Implementation of biomarker-driven colon cancer therapy: Prognostic and predictive biomarkers  57

Having oncogenic character, the PIK3CA gene biomarkers’ clinical benefit from anti-EGFR anti-
is mutated in many different tumors, includ- bodies in metastatic colorectal cancer (mCRC)
ing CRC (Samuels et  al., 2004; Cathomas, 2014). (Sood et  al., 2012; Yang et  al., 2013; Therkildsen
PIK3CA is mutated in approximately 10%–30% et al., 2014).
of CRCs and 80% of mutations found in two hot
spots in exon 9 (E532K, E545K) and exon 20 BRAF in CRC
(H1047R) (Samuels et al., 2004; Velho et al., 2005;
Velho et al., 2008; Cathomas, 2014), which makes The BRAF gene is located on chromosome 7
PIK3CA one of the most frequently mutated genes (7q34) and encodes the BRAF protein, which
in CRC. Additionally, in the majority of studies, mainly functions in the MAP kinase/ERK signal-
PIK3CA mutations have shown to be significantly ing pathway. This intracellular pathway located
associated with KRAS mutation and proximal downstream to the EGFR signaling cascade
tumor location (Barault et al., 2008a; Rosty et al., regulates cellular growth, differentiation, pro-
2013). Though it has not been consistently reported liferation, senescence, and apoptosis of tumor
(Mei et  al., 2016), association of PIK3CA muta- cells (Peyssonnoux and Eychene, 2001). BRAF
tions (particularly for exon 20 mutation) with a mutations have been described in approximately
worse clinical outcome and with a negative pre- 9%–14% of CRC with wild-type KRAS mutation
diction of a response to targeted therapy by anti- (Di Nicolantanio et  al., 2008; Bokemeyer et  al.,
EGFR monoclonal antibodies (mAbs), such as 2012; Therkildsen et al., 2014). A point mutation,
cetuximab and panitumumab, have been demon- which leads to a valine substitution by glutamic
strated (Ogino et al., 2009a; Souglakos et al., 2009; amino acid, is the most common mutation in the
Whitehall et al., 2012; Mao C et al., 2012; Day et al., BRAF gene (Ikenoue et al., 2003). It has been well
2013; Cathomas, 2014). Besides lack of response to demonstrated that KRAS and BRAF mutations are
anti-EGFR mAbs, particularly in KRAS wild-type almost mutually exclusive in CRC (Rajagopalan
patients, a promising predictive value of PIK3CA et  al., 2002; Fransen et  al., 2004). In clinical
mutations for poor survival in metastatic colorec- aspects, BRAF mutations have been observed
tal cancer (mCRC) patients treated with anti- more frequently in colon cancer versus rectal
EGFR mAbs has been suggested (Wu et al., 2013). cancer and also in proximal tumors compared to
Furthermore, some in vitro and in vivo studies distal tumors (Domingo et al., 2005; Deng et al.,
demonstrated the benefit of aspirin in CRC patho- 2008). Moreover it has been stated that BRAF
genesis harboring PIK3CA mutation, suggesting mutation can be specific for sporadic cancer rather
that PIK3CA mutation may predict response to than Lynch syndrome–associated hereditary non-
aspirin therapy for colorectal cancer (Liao et  al., poliposis colorectal cancer (Domingo et al., 2005;
2012; Paleari et al., 2016; Gu et al., 2017; Zumwalt Funkhouser et al., 2012). In a recent study evalu-
et al., 2017). ating the correlation between KRAS/BRAF muta-
PIK3CA gene mutation, in many instances, tional status and clinicopathological features in
occurs simultaneously with PTEN mutation advanced and recurrent CRC found that in nearly
(Chalhoub and Baker, 2009; Day et  al., 2013). 60% of patients with CRCs with BRAF mutations,
Acting as tumor suppressor gene, PTEN, which the tumor metastasized to the peritoneum com-
is localized on chromosome 10q23, encodes for a pared with nearly 15% of patients with other sub-
protein phosphatase and tensin homolog (PTEN) types (Souglakos et al., 2009). Furthermore it has
implicated in cell cycle regulation to prevent cells been demonstrated that at least two-thirds of the
from growing and dividing too rapidly (Steck et al., poorly differentiated or mucinous carcinomas had
1997). Despite the significance of PTEN mutations BRAF mutation (Yokota et al., 2011). In mucinous
in sporadic CRC remains still unclear some recent adenocarcinoma that accounts for 5%–15% of all
studies emphasized that PTEN loss together with primary colorectal cancer (Symonds and Vickery,
BRAF mutations, PIK3CA exon 20 mutations 1976; Lungulescu et  al., 2017), a poor progno-
could be predictive of worse outcomes in KRAS sis due to a poor response to oxaliplatin and/or
wild-type mCRC treated with anti-EGFR MAbs irinotecan-based chemotherapy was reported
(Sood et al., 2012; Yang et al., 2013). Accordingly, (Catalano et  al., 2009). Poor prognosis of BRAF
this suggests the power of combined multiple mutant tumors may be partially explained with
58  Precision medicine for colorectal cancer

these histological properties. In a recent study better survival compared to patients with BRAF
including 2328 patients, BRAF mutation status mutant/MSS tumors. Moreover, it was stated that
and patient and tumor characteristics have been V600E BRAF mutations were a negative prog-
evaluated (Gonsalves et  al., 2014). In this study nostic feature in MSS early stage resected CRC
BRAF mutated tumors were more likely found (Roth, 2010; Popovici et al., 2013). Patients with
in patients aged 70 years or more, women, and advanced and recurrent CRC treated with sys-
non-Hispanic white people. Additionally, these temic chemotherapy (n = 229) were analyzed for
tumors were also more likely to have four or more KRAS/BRAF genotypes. The median overall sur-
positive lymph nodes and T4 stages, which lead vival (OS) for BRAF mutation-positive was 11.0
to more aggressive phenotype and poor progno- months, which was significantly worse than that
sis (Gonsalves et  al., 2014). BRAF-mutated and for patients with wild-type BRAF (40.6 months,
wild-type tumors can be differentiated by means P < 0.001). This finding is consistent with other
of some molecular features, namely, decreased studies including both early stages and advanced
expression of CDX2, loss of P16, and positivity for stages of CRC (Roth, 2010; Ogino et  al., 2009b;
DNA methyltransferase-3B, a marker of de novo Fariña-Sarasqueta, 2010). Many reports indi-
CpG island methylation phenotype (CIMP) or cated that BRAF mutations were associated with
SIRT1 histone deacylase expression (Baba et  al., lack of response to anti-EGFR targeting agents
2009). Molecularly BRAF mutations have been (Di Nicolantanio et al., 2008; Laurent-Puig et al.,
linked to high levels of MSI (MSI-H), MLH1 hyper- 2009; Zlobec et  al., 2009; Loupakis et  al., 2009;
metilation, and CIMP-high status (Ahlquist et al., Bokemeyer et al., 2012). One of these studies was
2008; Barault et  al., 2008a, 2008b; Nosho et  al., conducted by Di Nicolantanio et  al. (2008) and
2008; Funkhouser et al., 2012). Colon cancers that in this trial patients with BRAF mutation had
arise from serrated polyps are typically proximal significantly shorter overall and progression-free
tumors, dMMR status, and high CIMP and BRAF survival when treated with cetuximab- or pani-
mutant phenotypes (Leggett and Whitehall, 2010). tumumab-based chemotherapy, compared with
Velho et al. (2008) postulated that BRAF mutations wild-type patients. Another study addressed
are likely to precede MSI carcinomas because the the predictive value of BRAF in patients treated
frequency of mutation in serrated polyps is simi- with cetuximab in first line, second line or fur-
lar to that of MSI cancers but statistically different ther setting in combination with chemotherapy
from MSS tumors. All these findings support that (Souglakos et al., 2009). No patients with BRAF
BRAF mutation may be an early event in colorec- mutation responded to cetuximab. The response
tal carcinogenesis. Zlobec et al. (2009) by using a rate to cetuximab-based chemotherapy was sig-
representative cohort of 404 sporadic colorectal nificantly lower in case of BRAF mutation than
cancers, described BRAF mutation as a significant wild type (8.3% versus 38%) (De Rook et  al.,
molecular biomarker of poor outcome in patients 2010). An analysis of CRYSTAL and OPUS studies
with right-sided disease. This poor outcome was revealed that BRAF mutation was detected in 9%
independent from tumor size, lymph node status, of KRAS wild-type tumors. Although some ben-
and MSI status. efit was observed in KRAS-wild/BRAF-mutant
The prognostic role of BRAF was tested in patients after adding cetuximab to the chemo-
patients with resected stage II and stage II CRC. therapy, this was not statistically significant
In this translational study, including three phase (Tejpar et al., 2014). Overall findings may suggest
III trials (PETTAC-3, EORTC 40993, SAKK), that BRAF mutation appears not to be predictive
1564 paraffin blocks were collected and tested for resistance to cetuximab plus chemotherapy
(Roth, 2010). It was noticed that BRAF tumor (Van Cutsem et  al., 2009). In the PRIME study,
mutation status was prognostic for OS in all mCRC patients were treated with panitumumab
patients (both stages II and III combined) and plus FOLFOX4 and BRAF mutation was detected
in stage III alone. This is especially promi- in 8% of the patients (Di Nicolantanio et  al.,
nent in microsatellite stable (MSS) CRC with 2008). In patients without RAS and BRAF muta-
BRAF mutation (Roth, 2010). Lochhead et  al. tions, 28.3 months overall survival was observed
(2013) recently demonstrated that patients with in the panitumumab–FOLFOX4 group. However,
BRAF wild/MSI high tumor had significantly OS in patients with RAS wild-type BRAF mutant
 Implementation of biomarker-driven colon cancer therapy: Prognostic and predictive biomarkers  59

was 10.5 months. This study also confirmed that on RAS GTPase activity, whereas exon 4 muta-
BRAF mutations appeared to confer a poor prog- tions have been suggested to increase GDP to GTP
nosis, regardless of the treatment group. exchange (Malumbres and Barbacid, 2003).
Recently, the predictive role of BRAF muta- KRAS mutations are believed to be an early
tions in randomized trials that compared cetux- event in colorectal tumorigenesis. In fact, it is not
imab or panitumumab plus chemotherapy (or required to initiate adenomas but mainly con-
monotherapy) with standard therapy or best sup- tributes to their progression. Mutations of the
portive care (BSC) were included. Cetuximab and KRAS gene have been identified in tissues from
panitumumab were assessed in a meta-analysis both adenoma and carcinoma cases, but at much
(Doullard et  al., 2013). Nine phase III trials and lower frequencies in colon adenoma tissues than in
one phase II trial (six first-line and two second-line carcinoma tissues. Many have hypothesized that
trials, plus two trials involving chemorefractory development of KRAS mutation is an important
patients), including 463 RAS-wild/BRAF-mutant role in the multistep process early in the carci-
CRC patients, were analyzed. The addition of anti- nogenesis. Nearly 30%–40% of CRCs have KRAS
EFGR treatment in the BRAF-mutant subgroup mutations and the point mutations are the most
did not show significantly improved PFS (HR, common mutation in KRAS gene (Capella et  al.,
0.88; p = 0.33), OS (HR, 0.91; p = 0.63), and ORR 1991; Oudejans et al., 1991). Almost 90% of KRAS
(relative risk, 1.31; p = 0.25) compared with con- mutations occur in codons 12 and 13 in the phos-
trol regimens (Pietrantonio et al., 2015). phate-binding loop, and mutations in either codon
Taken together, CRC with BRAF mutation have transforming capacity (Andreyev et al., 1998).
seems to originate from the serrated pathway of In a recent study, seven of the most common KRAS
carcinogenesis and has aggressive features like mutations in codon 12 and codon 13 were exam-
poor differentiated or mucinous histology, four or ined in 2478 colorectal tumor samples (Yoon et al.,
more lymph nodes involvement, and have a ten- 2014). Mutations in these codons were detected in
dency to peritoneal spread. Moreover, association 35.4% of tumors (27.6% in codon 12 and 7.8% in
of BRAF mutations with the efficacy of anti-EGFR codon 13). Within codon 12, 82% of mutations were
therapy remains controversial (van Brummelen found in the second base position (G12D). In vitro
et  al., 2017), but BRAF mutation appears to be a studies illustrated that when compared to codon 13
strong negative prognostic marker for patients mutations, Kras codon 12 mutations have greater
with KRAS wild-type colorectal cancer treated transforming ability characterized by inhibition of
with anti-EGFR therapies. All these findings sup- apoptosis, enhanced loss of contact inhibition, and
port the idea that BRAF mutations are negative increased predisposition to anchorage-indepen-
prognostic biomarkers. dent growth (Moerkerk et al., 1994).
Prognostic value of KRAS has been investi-
KRAS in CRC gated in several studies in both early and advanced
stages; however, inconsistent results had been
Ras is a family of genes that encode a class of released in early trials (Oudejans et al., 1991; Dix
21 kD membrane-bound proteins that bind gua- et al., 1994; Tanaka et al., 1994). Therefore RASCAL
nine nucleotides and have intrinsic GTPase (Kirsten Ras In Colorectal Cancer Collaborative
activity. This family consists of three functional Group) investigators aimed to definitively deter-
genes—H-ras, K-ras, and N-ras—which encode mine whether the presence of a KRAS mutation
highly similar proteins. These are the most com- is of prognostic significance. In this study 2721
mon oncogenes in human cancer. KRAS is the patients from 22 groups in 13 countries were
most frequently altered gene and is located on the recruited (Andreyev et al., 1998). The study group
short arm of chromosome 12. The action of this included patients of all stages but only 10% were
gene manifests in the RAS-RAF-MAPK pathway, Dukes stage D. KRAS codon 12 or codon 13 was
downstream of epidermal growth factor recep- detected in 37.7% of the tumors. Mutations were
tor (EGFR) (Moerkerk et al., 1994). Of the KRAS not associated with sex, age, tumor site or Dukes
mutations, 90% are located in exon 2 and 10% in stage. Multivariate analysis suggested that the
exons 3 and 4. Activating mutations in exons 2 presence of a mutation increased the risk of recur-
and 3 have been suggested to have similar effects rence (p < 0.001) and death (p = 0.004). When
60  Precision medicine for colorectal cancer

comparing all of the specific point mutations, codon 12 mutations were detected in 79% of
overall survival was adversely affected by the pres- mutated cases and significantly associated shorter
ence of a glycine to valine amino acid substitution TTR. A similar trend was also noted for the pG13D
on codon 12 (C12V). mutations suggesting that KRAS exon 2 mutations
In a similar study conducted in 404 sporadic and are related to a worse prognosis.
94 hereditary CRC patients, KRAS G12D muta- Both RASCAL and PETTAC8 showed that a
tion had poorer prognosis compared to patients tumor-carrying codon 12 mutation had a statisti-
with other KRAS mutations (Zlobec et  al., 2010). cally significant impact on worse PFS (Oudejans
In the RASCAL II study, association of Kras muta- et al., 1991; Malumbres and Barbacid, 2003). KRAS
tions with different stages of colorectal cancer was 12 mutations may confer a more aggressive behav-
evaluated. The results confirmed that C12V muta- ior than KRAS 13 mutations (Guerrero et  al.,
tion of KRAS was associated with a 50% increase 2000). However, a similar trend was shown also for
of relapse or death in patients with Dukes stage C codon 13 mutations in different studies. In 2002
cancer but not the other stages of CRC (Andreyev Bazan et  al. found that codon 13 mutations had
et al., 2001). a stronger predictive value than any K-ras muta-
The prognostic role of KRAS/BRAF was tested tions. Multivariate analysis showed that this spe-
in patients with resected stage II and stage II cific codon 13 K-ras mutation was an independent
CRC. In this translational study, paraffin blocks prognostic factor for DFS and OS in 160 untreated
from PETTAC-3, EORTC 40993, and SAKK trials patients who underwent surgery for a primary
were evaluated (Ogino et  al., 2009b). KRAS and tumor. The prognostic role of KRAS/BRAF muta-
BRAF tumor mutation rates were 37% and 7.9%, tion was analyzed in the CRYSTAL and OPUS
respectively, and were not significantly different studies including mCRC patients who received
according to tumor stage. In a multivariate analy- FOLFIRI/FOLFOX with or without cetuximab
sis, KRAS mutation was associated with grade of as the first-line treatment (Bokemeyer et  al.,
the tumor. In univariate and multivariate analy- 2012). Subgroup analyses in this study showed
sis, KRAS mutations did not have a major prog- that patients with KRAS codon 13 D mutations
nostic value regarding relapse-free survival (RFS) showed poor prognosis. Another study conducted
or OS. Other recent trials, NCCTG N0147 and by Yokota et al. (2011) also supports the idea that
PETACC-8, have evaluated chemotherapy with OS for patients with KRAS codon 13 mutation was
and without cetuximab for patients with stage III significantly worse than for those with wild-type
disease (Blons et al., 2014; Gonsalves et al., 2014). tumors.
In the first trial, primary tumors were assessed for KRAS mutation status was assessed in a series
KRAS and BRAF mutations and defective MMR of resected or sampled CRC liver metastases (Nash
status (Gonsalves et al., 2014). The results showed et al., 2010). Nash et al. (2010) observed that KRAS
that three-year disease-free survival in patients mutations have also been associated with more
with wild-type KRAS was significantly better than rapid and aggressive metastatic behavior of CRC
that in patients with KRAS mutants (72.3% versus liver metastases. KRAS mutation was indepen-
64.2% HR = 0.7 p = 0.004). It is suggested that dently associated with poor prognosis after liver
KRAS mutations are independent prognostic fac- resection. Both codon 12 and codon 13 mutations
tors. Furthermore, tumors with KRAS mutations in the KRAS gene may result in different biologi-
were less likely to have dMMR and high-grade cal properties that might lead to poor prognosis in
histology but more often right-sided. Consistent colorectal cancer.
with this study, in the PETTAC8 trial, the prog- In conclusion, presence of KRAS mutations
nostic impact of KRAS exon 2 mutations in stage both in codon 12 and 13 are associated with inferior
III colon cancer were analyzed (Blons et al., 2014). survival both in early stage and mCRC patients.
Frequency KRAS mutations were 38.5% and pres-
ence of the mutation was linked to shorter TTR (p Immune-related markers in CRC
< 0.001). In this study it was also indicated that
KRAS mutation was more frequent in women, in In the last few years years increasing proof from
proximal tumors, and was associated with age, and genetically engineered mice and clinical epidemi-
no vascular and lymphatic invasion. Furthermore, ology have shown the significance of the immune
 Implementation of biomarker-driven colon cancer therapy: Prognostic and predictive biomarkers  61

barrier in formation and progression of tumors (McDermott and Atkins, 2013). PD-L1 is highly
(Hanahan and Weinberg, 2011). The development expressed in several cancers, including CR. It’s
of cancer in immunodeficient mice and the exis- well established that this leads to cancer immune
tence of antitumor immune responses in colon evasion, and consequently, this is often associ-
and ovarian cancers has been supported through ated with a worse prognosis (McDermott and
clinical epidemiology studies (Pagès et  al., 2010). Atkins, 2013). Accordingly, interaction of PD-1
During growth of cancer cells, the major his- with its ligand PD-L1 which promotes the exhaus-
tocompatibility complex (MHC) interacts with tion of peripheral T-effector cells (CD4+, CD8+,
T-cell receptor (TCR) and expresses CD4+ or Treg cells), induction of regulatory T cells (Treg)
CD8+ lineage markers (Sukari et al., 2016). Both (Zhang X et al., 2015; Francisco et al., 2009), have
innate immunity (natural killer [NK] cells, uncon- been shown to associate with poor prognosis in
ventional T lymphocytes, and tumor infiltrat- CRC. Furthermore, the PD-1/PD-L1 interac-
ing macrophages [TIM]) and adoptive immunity tion directly inhibits apoptosis of the tumor cell
(intratumoural memory CD8T cells, and tumor- (Francisco et al., 2009). To this end, large pharma-
infiltrating CD45RO+ memory T lymphocytes ceutical companies have developed drugs targeting
[CD45RO+ T cells]) have an important role in the cytotoxic T cell-associated antigen 4 (CTLA-4),
development as well as in prevention and recur- programmed cell death receptor 1(PD-1), and its
rence of CRC tumors leading to better prognosis of ligands PD-L1 and PD-L2 in mCRC (Sanchez-
colorectal cancers (Sandel et al., 2005; Galon et al., Castañón et al., 2016). The association between the
2006; Sanchez-Castañón et  al., 2016). Expression MSI status and the efficacy of the programmed cell
of tumor-associated antigens (TAAs), such as death 1 (PD-1) receptor inhibitor pembrolizumab
Wilms’ tumor gene 1 (WT1), carcinoembryonic has been also demonstrated (Pardoll, 2012; Boland
antigen (CEA), mucin 1 (MUC1), and melanoma- and Ma, 2017; Gong et al., 2017). Respectively, the
associated antigen gene (MAGE) in human CRC U.S. Food and Drug Administration (FDA) has
cells make it an appropriate and potential tar- granted accelerated approval to checkpoint inhibi-
get for immunological-based therapies (Kajihara tor pembrolizumab (Keytruda) for pediatric and
et  al., 2016). In CRC genetic and epigenetic aber- adult patients with unrespectable, metastatic, or
rations due to CIN pathway, MSI phenotype, and MSI-H that has progressed following chemother-
CIMP could lead to manifold oncogenes muta- apy (Bilgin et al., 2017; Chang et al., 2017; Lemery
tions, resulting in immunogenic CRC (i.e., capable et al., 2017).
of inducing an immune response). Thus, since den- Cytokines are small, secreted proteins provid-
dritic cell–based cancer immunotherapy induces ing interactions and communications between
TAAs (CEA, WT1, MAGE, and MUC1) in CRC cells through cell signaling. They are other
cells, it is anticipated that a combination of immu- immune-related markers that have an important
notherapy with conventional chemotherapy and/ role in innate and specific immune response (Van
or radiotherapy, surgery, and even with immune- der Meide and Schellekens, 1996). Despite some
checkpoint inhibitors could mediate a potent anti- cytokines approved in cancer therapy, such as
tumor effect (Koido et  al., 2012; Kajihara et  al., IL-2 in the treatment of melanoma and renal cell
2016). Furthermore, since TILs interact most carcinoma and IFN-α (also called Multiferon) in
closely with the tumor cells and reflect tumor host the treatment of multiple hematological malignan-
interactions more accurately (Holmes, 1985), it is cies, cervical cancer, carcinoid syndrome, medul-
demonstrated that CRC patients with MSI tumors lary thyroid cancer, basal cell, and squamous cell
owing to high levels of TILs have better p ­ rognosis carcinoma, studies involving cytokine therapies
(Michael-Robinson et  al., 2001; Phillips et  al., in CRC are limited so far (Lynch and Murphy,
2004). 2016). However, a phase I study of 33 patients
PD-L1 is the ligand for transmembrane pro- with advanced solid tumors, including four CRC
grammed cell death protein 1 (PD1) (also known patients, demonstrated that daily subcutane-
as cluster of differentiation 279-CD279), which ous administration recombinant IL-10 (AM0010)
is expressed on T cells and B cells and is a criti- resulted in stable systemic Th1 immune stimula-
cal immune checkpoint pathway responsible for tion. Yet, even though cytokine therapy represents
mediating tumor-induced immune suppression an important approach, it needs more investigation
62  Precision medicine for colorectal cancer

for being used in CRC as a therapy target (Lynch demonstrated (Kruszewski et  al., 2004), which
and Murphy, 2016). Major histocompatibility com- could lead to a poor prognosis (Khan et al., 2004).
plex class I (MHC-I) low expression is associated Moreover, in human tumors, such as melanoma,
with poor prognosis in colorectal patients. breast cancer, and colon cancer, absence of exon
2 and exon 3 are also demonstrated (Gong et  al.,
Protease related markers in CRC: 1993). Likewise, it is reported that CATSB enzyme
Focus on cathepsin B (CATSB) activity levels are inversely correlated with the
Dukes stages; levels of expression of CATSB
An important property of metastatic cells is their mRNA in colon carcinomas were higher in tumors
ability to dissolve the extracellular matrix (Liotta belonging to the Dukes A1 and B1 stages (Dukes,
et al., 1986; Hart and Saini, 1992). In fact, an impor- 1932) than in Duke stage B3 carcinomas and in the
tant level of enzymes, degrading the extracellular normal mucosa (Murnane et  al., 1991). In addi-
matrix, is often found in malignant tumors. Most tion, the increase in CATSB activity is associated
of these enzymes may be involved in the activation, with increased CATSB mRNAs in the earliest
inhibition, or regulation of other enzymes released stages, suggesting that CATSB elevation may be
by normal or tumor cells. The degradation of the important in tumors that are about to invade the
extracellular matrix depends on a series of extra- intestinal wall (Tan et al., 2013). In agreement with
cellular neutral proteases; a protease (also called other studies, a comparative study of CATSB gene
peptidase, proteinase or peptide hydrolase) is an transcripts in colon tumors and the corresponding
enzyme that participates in the catabolism (deg- normal tissues established a relationship between
radation) of proteins by cleaving peptide bonds. the spliced form for exon 2 and the complete form
Intracellular proteolysis is an extremely controlled of the mRNA of cathepsin B comprising exons
process that takes place in all cell compartments. 1–5. Accordingly, authors demonstrated that the
Proteolytic activity is involved in different pro- ratio form spliced for exon 2/complete form was
cesses: protein renewal especially during cell significantly increased in colorectal tumors com-
growth as well as protumorigenic processes, such pared with matched normal samples, suggesting
as cell transformation, proliferation, invasion, and that alternative splicing of CATSB mRNA at the
malignant progression (Sloane and Berquin, 1993; 5′ UTR may be an indicator of cellular transfor-
Fennelly and Amaravadi, 2017). mation during colorectal carcinogenesis (Hizel
Several independent studies underlined the et al., 1998). Furthermore, a recent study in Middle
important role of lysosomal cysteine proteases East populations diagnosed with CRC confirmed
cathepsin B (CATSB) (Herszènyi et  al., 1999; that there was a significant increase in the CATSB
Tan et al., 2013; Bian et al., 2016) and L (CATSL) expression at the mRNA and protein levels in
(Herszènyi et  al., 1999; Tan et  al., 2013; Sudhan tumor tissue compared to normal adjacent mucosa
and Siemann, 2015), and the serine protease uro- in late stage CRC suggesting that CATSB may be
kinase type plasminogen activator (UPA) with its an important prognostic biomarker for late stage
inhibitor type-1 (PAI-1) (Herszènyi et  al., 1999; CRC with lymph node metastases in this popula-
Sakakibara et al., 2005) in colorectal cancer inva- tion (Abdulla et al., 2017). Hence, characterization
sion and metastasis, and they appear to be synthe- of the aberrant splice variants may improve under-
sized in an increased amount and exaggeratedly standing of malignant transformation in CRC.
secreted (Sloane and Berquin, 1993). The human In light of the aforementioned studies, since
cathepsin B gene is a “housekeeping” gene hav- proteolytic enzymes play a crucial role in tumor
ing 12 exons and is located on chromosome 8 at development and progression due to their abili-
p 22. Being very variable in 5′ untranslated region ties to degrade extracellular matrix, validated pro-
(UTR) the expression of human cathepsin B is reg- teinase markers may be of interest as independent,
ulated through alternative splicing of the mRNAs predictive, and prognostic factors of malignant
producing several transcript species (Gong et  al., progression of neoplastic disease. Furthermore,
1993; Berquin et al., 1995). high expression of proteinases such as CATSB in
Close relationship between the intensity of advanced stages of cancer can offer significant ther-
angiogenesis and overexpression of the CATSB apeutic benefit when using proteinase inhibitors
mRNA in resected colon adenocarcinoma is in the treatment of CRC leading to new methods
 CRC epigenetics  63

for individualized cancer therapy in the precision the available research publication is controversial
medicine approach (Guzińska-Ustymowicz, 2006; (Wang W et al., 2017), a broad spectrum of stud-
Duffy and Crown, 2008; Herszényi et al., 2014). ies emphasize the critical association between
TL and CRC for improved molecular diagnosis.
Telomere length in CRC Accordingly, TL was smaller in CRC tumors than
in the healthy adjacent mucosa, suggesting that
Telomeres are DNA sequences containing noncod- longer telomeres could be indicative of poor prog-
ing arrays of TG-rich/TTAGGG tandem repetitive nosis in CRC (Mzahma et al., 2015; Zhang C et al.,
nucleotide sequences (approximately 1000–2000 2015; Fernández-Marcelo et al., 2016; Balc’h et al.,
TTAGGG tandem base pair) at each end of a chro- 2017). A correlation between TL and tumor stage
mosome (Moyzis et al., 1988) playing a crucial role is also reported, demonstrating the presence of the
in facilitating chromosome replication and protect- longer TL in advanced CRC tumors (Engelhardt
ing the end of the chromosome from fusion with et  al., 1997; Gertler et  al., 2004). A significant
neighboring chromosomes or from DNA repairing association between length and KRAS mutation
activities that reseal chromosome internal DNA status has also been reported (Balc’h et al., 2017).
breaks occurring during DNA damage (De Lange, Furthermore, recently, the clinical utility of TL
2009). During DNA replication, telomere shortening has been demonstrated as a potential biomarker
occurs due to incomplete replication of telomere, and for mCRC patients with nonmutated KRAS sta-
continuous shortening leads to chromosomal deg- tus treated with anti-EGFR indicating an elevated
radation and cell death (O’Sullivan and Karlseder, telomere length and a good response (inhibition of
2010; Shammas, 2011). In order to maintain telomere proliferation) to anti-EGFR (Augustine et al., 2015;
elongation, at the end of each cell division telomere is Augustine et al., 2017).
replenished by “ribonucleoprotein complex,” which Though some current evidence supports the use
is reverse transcriptase enzyme telomerase, and it is of telomere status as a prognostic factor for overall
actively expressed in more than 80% of tumors, par- survival and treatment in CRC patients, additional
ticularly in tumors with metastatic potential allow- large well-designed prospective cohort studies are
ing to escape from the inhibition of cell proliferation needed to further assess the telomere status as an
because of shortened telomeres (Aschacher et  al., independent prognostic factor for personalized
2016; Ivancich et al., 2017). In humans, while detect- approach in CRC patients.
able levels of telomerase activity is limited to stem
cells, germ line cells, and also cardiovascular cells, its CRC EPIGENETICS
activity is absent in most normal cells (Collins and
Mitchell, 2002; Shay and Wright, 2010; Pech et  al., Epigenetic alterations in cancer include abnor-
2015; Zurek et al., 2016; Ivancich et al., 2017). Since mal DNA methylations, histone modifications,
cancer cells necessitate a mechanism to maintain and expressions of miRNAs. According to the lit-
their telomeric DNA in order to continue dividing erature, these alterations in colorectal cancers are
indefinitely, shortening of telomeres and the induc- more frequent than genetic alterations (Puccini
tion of cell senescence could be a potential cancer et  al., 2017). Epigenetic changes in cancers were
biomarker and therapeutic tool as anticancer drugs unfamiliar to scientists until 30 years ago when
(Philippi et  al., 2010; Ruden and Puri, 2013; Piñol- the first results were published about DNA meth-
Felis et al., 2017). ylation alteration in human cancers. Accordingly,
It is widely acknowledged that telomere dys- these results comparing human colorectal tumors
function and telomere length (TL) is at the origin of with normal mucosa of the same patient demon-
various degenerative disorders as well as predispo- strated the existence of high hypomethylation of
sition to cancer (Shin et al., 2006; Shammas, 2011; almost one-third of single copy genes (Feinberg
Shay and Wright, 2011; Armanios and Blackburn, and Vogelstein, 1983).
2012). Although telomerase activity is associated
with poor prognosis (Garcia-Aranda et  al., 2006; DNA methylation
Bertorelle et al., 2013), the prognostic role of telo-
mere length in CRCs is still unclear (Wang  W In human cancers, DNA methylation occurs at
et  al., 2017). Although some scientific data from cytosine residues in regions with a high frequency
64  Precision medicine for colorectal cancer

of CpG sites (CpG dinucleotides), which consti- regulation of gene expression, posttranslational
tute CpG islands. Almost 70% of the human gene modifications at N-terminal tails of the histones
promoter region harbors CpG islands with high control chromatin structure as well as gene expres-
CpG content (Saxonov et  al., 2006), which are sion (Das and Tyler, 2013). Histone proteins expose
methylated by DNA methyltransferases (DNMTs) various posttranslational modifications, which
(Puccini et al., 2017). Although hypomethylation include methylation, acetylation, phosphoryla-
of CpG islands in the promoter region results in tion, deamination, and ribosylation. These modi-
overexpression of the genes or set of genes (Ehrlich, fications have a regulatory role in gene expression
2009), hypermethylation could be an important due to determination of chromatin structure as
cause of loss of expression of genes (Jin et al., 2011). tight or loose (Sachan and Kaur, 2015). The type
DNA hypermethylation in colorectal cancers is of modification and the specific amino acid that is
due to CIMP, which strongly correlates with clini- involved in modification control the effect of his-
copathological and molecular characteristics of tone modification. Acetylation/deacetylation and
CRC (Issa, 2004; Nazemalhosseini Mojarad et al., methylation/demethylation of lysine and arginine
2013). Besides MLH1, MSH2, MSH6, and PMS2 residues within histone tails are the most exten-
gene mutations, DNA MMR system hypermethyl- sively characterized modifications in CRC. Most of
ation could be another source for MSI of colorectal the acetylation locations were determined on the
tumors (Grady and Carethers, 2008) On the other N-terminal tail of histones due to the easy access
hand, global DNA hypomethylation has a cru- for modifications (Goel and Boland, 2012).
cial role in tumor formation (Antelo et  al., 2012; The acetylation reactions of lysine amino acids
Pavicic et al., 2012). LINE-1 or short interspersed on histone tails are reversible modifications and
nucleotide elements (SINE, or Alu) are mostly serve as activators or repressors of transcription
hypomethylation sites in DNA in most cancers function. For example, hypoacetylation silences
including colorectal cancers, and some stud- gene expression, whereas hyperacetylation permits
ies showed LINE-1 hypomethylation correlated active gene transcription due to the destabilization
with worse survival in colorectal patients (Antelo of chromatin fibers and increasing the mobility of
et al., 2012; Pavicic et al., 2012). Overall, it is obvi- nucleosomes (Das and Tyler, 2013). These acetyla-
ous that epigenetic deregulations caused by DNA tion reactions are carried out by removing acetyl
hypermethylation and hypomethylation contrib- groups from lysine amino acids by histone acet-
ute separately to the process of CRC carcinogene- yltransferases (HATs) and histone deacetylases
sis Hence, acknowledging epigenetic data–related (HDACs), as coactivators and corepressors of tran-
methylation status might represent a promising scription. The balance between HATs and HDACs
tool for a more precise diagnostic, prognostic, and controls the transcriptional inhibition of tumor
personalized therapeutic approach (Lam et  al., suppressor genes. The opposite of acetylation asso-
2016; Puccini et al., 2017; Werner et al., 2017). ciates with transcriptional repression of genes
(Bardhan and Liu, 2013). HDACs play a crucial role
Histone modifications in CRC development. Up to now, 18 HDACs have
been identified as corepressor multiprotein com-
The chromatin in the mammalian genome includes plexes and they are divided into four classes: Class
tightly packed DNA inside the nucleus. The small- I (HDAC1, HDAC2, HDAC3, and HDAC8), Class
est unit of chromatin is nucleosomes formed by II (HDAC4, HDAC5, HDAC6, HDAC7, HDAC9,
wrapping of 147 bp of DNA around eight his- and HDAC10), Class III (Sirt1, Sirt2, Sirt3, Sirt4,
tone proteins, (including H3/H4 tetramer and Sirt5, Sirt6, and Sirt7), and Class IV HDACs (only
two H2A/H2B dimers). Due to the numerous HDAC11). Classes I, II, and IV HDACs have simi-
direct protein–DNA interactions, nucleosome lar features due to their structure and function,
has a highly stable unit. During the conducting but class III does not show similarity (Bolden et al.,
of genomic functions, processive enzymes (with 2006). Increased expression of several HDACs has
the ability to catalyze consecutive reactions with- been determined in CRC. In addition, upregula-
out dissociating from its substrate) reach the cod- tion of Class I HDACs (HDAC1, HDAC2, HDAC3)
ing regions of genes (Das and Tyler, 2013). Except has been associated with reduced patient survival
for DNA methylation-based transcriptional in CRC (Ashktorab et al., 2009). The elevated levels
 CRC epigenetics  65

of HDAC2 are accompanied with the hypoacety- RNA-induced silencing complex to induce the
lation of H4K12 and H3K18 histones on the mul- degradation and suppression of the target mRNA
tistep carcinogenesis process of CRC (Ishihama (Corté et  al., 2012). The initial data, obtained by
et  al., 2007). The overexpression of HDAC3 was Michael et  al.’s research, suggested that miRNA-
recorded in duodenal adenomas of Apc1638N/+ 143 and miRNA-145 were steadily downregulated
mice and human colon cancers. Silencing of at different stages of CRC (Michael et  al., 2003).
HDAC3 in colon cancer cell lines induced growth Then, accumulation of information about CRC-
inhibition, and shortened survival and apoptosis associated miRNAs supported that miRNAs have
(Wilson et al., 2006). Silencing of HDAC4 expres- a role on the initiation, progression, and develop-
sion in HCT116 colorectal cancer cells resulted in ment of CRC (Strubberg and Madison, 2017). They
growth inhibition and increased apoptosis and p21 can have tumor suppressors or oncogene function
transcription (Wilson et al., 2008). according to the target gene (Zhu and Fang, 2016).
Like histone acetylation, methylation of lysine Different miRNA profiles have been assessed in
and arginine amino acids on N-terminal histone several studies: hsa-miR-17-3p, hsa-miR-20a, hsa-
tails is a reversible posttranslational modifica- miR-21, hsa-miR-25, hsa-miR-31, hsa-miR-92, hsa-
tion. It is regulated by histone methyltransferases miR-93, hsa-miR-96, hsa-miR-106 (includes 106a
(HMTs) and histone demethylases (HDMs) (Klose and 106b), hsa-miR-135b, hsa-miR-183, and hsa-
and Zhang, 2007). Each of the HMTs have a role miR-203 were detected as upregulated in CRC; and
either alone or in a complex with other HMTs. hsa-miR-1, hsa-miR-30a, hsa-miR-126, hsa-miR-
Histone H3 lysine 4 (H3K4) is methylated by SET1 133b, hsa-miR-143, hsa-miR-145, hsa-miR-191,
and MLL, and demethylated by the LSD1 and and hsa-miR-192 were detected as downregulated
JARID1 family of HDMs (Barski et  al., 2007). In in CRCs when compared with healthy individuals
histone H3, multiple-methylation can be observed (Corté et al., 2012; Ng et al., 2009; Bandrés et al.,
at multiple lysine sites (including histone H3 lysine 2006).
4 [H3K4], histone H3 lysine 9 [H3K9], histone H3 Circulating miRNAs have been found in
lysine 27 [H3K27], histone H3 lysine 36 [H3K36], the plasma or serum samples of CRC patients.
and histone H3 lysine 79 [H3K79]) or at a single Therefore plasma miRNAs can be used as nonin-
lysine (up to three methyl groups) (Bardhan and vasive early detection biomarkers. hsa-miR-17-3p
Liu, 2013). H3K27 trimethylation (H3K27me3), and hsa-miR-92a-1 have been found to be associ-
controlled by the RAS pathway, is frequently cor- ated with CRC in Ng et al.’s study (Ng et al., 2009).
related with gene silencing and affects cyclin D1 Especially, the function of hsa-miRNA-92a-1 was
and E-cadherin expression. Increased expression found and it was validated that it is associated with
of RAS oncoprotein could influence gene-specific the discrimination of advanced adenoma patients
histone modification during the epithelial mes- from healthy individuals. Compared to healthy
enchymal transition of the CRC cell line, Caco2 individuals, hsa-miR-29a and hsa-miR-221 have
(Peláez et al., 2010). been detected as overexpressed in the plasma
samples of all stages of CRC patients (Huang et al.,
microRNAs 2010; Pu et  al., 2010). As screening biomarkers,
miRNAs from fecal specimens of CRC patients
MicroRNAs (miRNAs), small noncoding RNAs have been evaluated. In a study, it was found that
18–25 nucleotides long, are involved in the reg- hsa-miR-17 cluster and hsa-miR-135 expression
ulation of proliferation, cellular growth, dif- levels were significantly elevated in CRC patients
ferentiation, and apoptosis by controlling the than healthy individuals (Koga et  al., 2010). In
translation of target mRNAs (Huang et al., 2011). another study from Link and his colleagues hsa-
miRNA biosynthesis includes a complex process. miR-21 and hsa-miR-106a have been evaluated as
Initially, primary miRNAs (pri- miRNAs, up to diagnostic biomarkers due to their overexpression
1 kb) are transcribed by the RNA polymerase II, in CRC patients compared with healthy controls
second precursor miRNAs (pre-miRNAs, 60–70 (Link et al., 2010).
base pairs long) are generated from pri-miRNAs Dysregulation of miRNA synthesis and func-
by the cleavage of RNase III Drosha. Finally, tion have a role in the development and progres-
formed mature miRNA is integrated into the sion of CRC due to their effect on the translation
66  Precision medicine for colorectal cancer

of oncogene or tumor suppressor genes. In two responses. Recent breakthroughs in genomics and
independent studies, over-expression of hsa- improvements in bioinformatics have uncovered
miR-21 was associated with poor prognosis (Corté the leading role of genes involved in drug metabo-
et al., 2012). In a group comprised of stage II CRC lism and disposition. Variation in drug-response
patients, low expression levels of hsa-miR-320 or phenotypes, in the forms of resistance and ADRs
hsa-miR-498 were correlated with poor progres- as well as the motivation to administer medica-
sion free survival compared to tumors with high tions more efficiently, are one of important major
expression (Schepeler et al., 2008). In another study components of precision medicine for personal-
that was conducted by Cheng and his colleagues, ized treatment (Aydin Son et al., 2013).
hsa-miR-141 was found to be related with stage IV Accordingly, pharmacogenetics (PGx), which
colon cancer and poor survival (Cheng et al., 2011). studies genetic variations among individuals to
High plasma levels of hsa-miR-221 were associated predict responses to therapeutic agents, has gained
with poor survival by Pu and his colleagues (Pu considerable momentum for both prescribed and
et al., 2010). over-the-counter medications. Dating back to the
In literature it was reported that miRNAs have 1950s (Kalow, 1962), PGx, which may uncover the
a role in drug resistance or sensitivity. Therefore associations between genetic variation and ADRs,
miRNAs can have the potential to predict anti- can be regarded as the 21st century’s answer for
cancer agents’ response. In a study conducted by rational drug use by selecting “the right drug” to
Nakajima et  al. (2006), hsa-let-7g and hsa-miR- “the right patient” at “the right dosage” to perform
181b were found related with 5-fluorouracil-based personalized drug dosing in clinic (Ingelman-
treatment response in colon cancer. The miRNA Sundberg, 2001; Becquemont, 2009; Ferraldeschi
let-7 family controls the translation of KRAS and Newman, 2011; Aydin Son et  al., 2013). In
protein and has a role in the resistance to EGFR- contrast to the current “one size fits all” approach
targeting monoclonal antibodies. A polymor- for drug prescription and dosing, pharmacogenet-
phism on the let-7 complementary site of KRAS ics/genomics being among one of the first clini-
was correlated with increased KRAS expression, cal applications in the field of genomics medicine
which is associated with resistance to anti-EGFR represents the near-term payoff for the pioneering
treatment (Johnson et al., 2005; Chin et al., 2008). area of precision medicine for personalized health
Different studies confirmed that miRNA profiling care (Ingelman-Sundberg, 2001; Becquemont,
plays an important role in the understanding of 2009; Ferraldeschi and Newman, 2011; Aydin Son
mechanisms underlying colorectal carcinogenesis, et al., 2013).
cancer progression, and metastases. Application of Since the completion of the Human Genome
miRNAs as cancer biomarkers for early diagnosis, Project in 2003 our world is today going through
prognosis, and treatment response prediction are a new era that is commonly referred to as the
promising, but profiling studies need to be vali- “postgenomic era.” Meanwhile pharmacogenet-
dated due to various differences between studies ics has gradually evolved into pharmacogenomics
that can be explained by several components such through the intensive application of genome-wide
as sample type, tumor location, stage of disease, association studies (GWAS), and today these two
genetic background, and methodology (Slattery terms are used interchangeably. (Pirmohamed,
et al., 2011). 2001; Lindpaintner, 2002; Moraes and Góes,
2016). However, while pharmacogenetics aims to
PHARMACOGENETICS (PGX) FOR distinguish responders from nonresponders (i.e.,
PRECISION MEDICINE IN CRC starts with an unexpected drug response result
CHEMOTHERAPY and looks for a genetic cause), pharmacogenom-
ics as the whole-genome application takes a much
In many countries, lack of response, that is, resis- more global approach to the impact of variation
tance and adverse drug reactions (ADRs) are a in genetic information on drug response, by look-
significant public health problem. Among indi- ing for genetic differences within a population that
viduals, genetic differences in pharmacogenes, explains certain observed responses to a drug or
that is, any gene that interacts with a pharmaceuti- susceptibility to a health problem (Pirmohamed,
cal drug, significantly affects drug clearance and 2001). Moreover, much of the work in the area of
 Pharmacogenetics (PGx) for precision medicine in CRC chemotherapy  67

pharmacogenomics is focused on identification of In cancer PGx predictive biomarkers for drug

disease susceptibility genes representing potential responses due to genetic variations can originate
new molecular drug targets for therapeutic inter- from the host (individual’s germ-line mutations)
vention by correlating gene expression or single and tumor cells (somatic/de novo mutations)
nucleotide polymorphisms (SNPs) with a drug’s (Pesenti et  al., 2015). The former affects phar-
efficacy or toxicity (Pirmohamed, 2001; Sheffield macokinetic (PK) (i.e., absorption, distribution,
and Phillimore, 2009; Milos and Seymour, 2004). metabolism, and excretion [ADME]) profile of the
However, this article does not distinguish between drug and are more likely to be related to predic-
those terms, but the abbreviation “PGx” will be tion of ADRs through alteration of drug plasma
used to cover both pharmacogenetics and phar- levels such as UGT1A1 variants and irinotecan-
macogenomics throughout this chapter. Though induced neutropenia (Rodríguez-Antona and
application of PGx testing is not widely accepted Taron, 2015; Bertholee et  al., 2017), whereas the
in routine medical practice, the scenario is likely latter affects the selection of chemotherapeutics
to be changing in the future for pharmacogenet- related to effectiveness of the response of a tumor
ics based diagnostics (PGDx) (Cecchin et al., 2017; to the treatment, such as cetuximab and EGFR/
Lemay et  al., 2017; Manson et  al., 2017). In PGx KRAS status (Deenen et  al., 2011b; Rodríguez-
research, numerous studies have used candidate Antona and Taron, 2015). Polymorphisms in
gene approaches to identify toxicity biomarkers. the human genome have been implicated in the
Since the mechanisms of action of cancer drugs expression and functioning of enzymes that is
through which a drug molecule produces its phar- involved in the distribution and metabolism
macological effect are well demonstrated, these of anticancer drugs affecting drug efficacy and
candidate gene studies are relatively prosperous toxicity, and consequently the outcome of treat-
compared to the identification of cancer suscepti- ment of patients (risk/benefit) (Weng et al., 2013).
bility genes (Rodríguez-Antona and Taron, 2015). Surgical resection, particularly for non-mCRC as
Furthermore, starting with the determination of the first choice treatment, and chemotherapy are
phenotypes (in the case of PGx studies, patients extensively used in the treatment of CRC patients
responding or developing drug-induced ADRs/ (Pozzo et  al., 2008; Chua and Morris, 2012). In
toxicities, and patients who respond to the treat- spite of complete resection, approximately half
ment without ADRs as controls), GWAS, which is a of CRC patients could still develop metasta-
very useful method for identifying common SNPs ses (Nordlinger et  al., 2009; Power and Kemeny,
with a minor allele frequency (MAF) >5%, became 2010). Fluoropyrimidine (e.g., 5-fluorouracil,
a very powerful tool for PGx study to identify SNPs capecitabine), irinotecan, and oxaliplatin are the
associated with ADRs/toxicities or drug resistance most common anticancer drugs used to treat
phenotype. (Daly, 2010; Ritchie, 2012). colorectal cancers (Aparicio et  al., 2005; Braun
Being an important backbone treatment of and Seymour, 2011). Moreover, for more effec-
cancer patients with lymph node positive and met- tive therapy in conjunction with chemotherapy
astatic disease, cytotoxic chemotherapy is becom- and the addition of the monoclonal antibodies,
ing increasingly personalized because of large such as cetuximab, bevacizumab, panitumumab,
differences among treatment outcomes (resis- or vascular endothelial growth factor aflibercept,
tance/toxicity) (Hammond et al., 2016). Together have ameliorated the median survival of mCRC
with nonmodifiable genetic factors other non- approximately from 8 to 24 months (Gallagher and
modifiable factors (ethnicity, age, gender, organ Kemeny, 2010; Van Cutsem et al., 2012). However,
function) and modifiable factors (smoking, alco- important interpatient variability in drug resis-
hol consumption, comorbidity, pregnancy, drug– tance and toxicity has been reported in many
drug interactions, and drug–food interactions) mCRC patients receiving combination therapy
have a critical impact on drug efficacy and toxicity FOLFIRI (5-fluorouracil infusion leucovorin, and
(Marques and Ikediobi, 2010). While chemothera- irinotecan) or FOLFOX (fluorouracil, leucovorin,
peutic agents act through cytotoxic effects, tar- and oxaliplatin). Chemotherapy side effects ham-
geted drugs are directed toward a tumor-specific per the efficient treatment of mCRC due to somatic
alteration exhibiting cytostatic effects (Kummar (acquired) mutation in tumors and/or individual’s
et al., 2006). germ-line DNA (intrinsic) mutation affecting the
68  Precision medicine for colorectal cancer

activity of metabolic enzymes, transporters, and and Ratain, 2004; Guillemette et al., 2014), which is
receptors (Braun and Seymour, 2011). To this end, then excreted via bile into the intestine (Innocenti
since drug resistance and toxicity have become and Ratain, 2003; Ferraldeschi, 2010). However,
more significant public health problems, in the in the intestine, SN-38G is deconjucated back to
context of precision medicine there is an unmet active SN-38 by bacterial β-glucuronidases and
and urgent need to develop predictive biomarkers then SN-38 is reabsorbed in the systemic circu-
for a personalized approach in cancer pharmaco- lation called enterohepatic circulation of SN-38
therapy (Derks and Diosdado, 2015). (Ferraldeschi, 2010). Consequently, glucurono-
syltransferases expressed in the gut play a role in
Toxicity-related biomarkers: PGx for the reglucuronidation and detoxification of SN-38
germ-line variations in anticancer in the intestine (Chen et  al., 2013). Most UGTs,
agent including UGT1A1 (also small intestine and colon,
but not in the kidney) are predominantly expressed
This section focuses on the anticancer drugs 5-FU in the liver (Strassburg et al., 1998). Human UGTs
and irinotecan as well as related genetic poly- show tissue-specific expression with UGT1A7,
morphisms, including UGT1A1, DPYD, TS, and UGT1A8, and UGT1A10, which are extrahepatic
MTHF that alter significantly the pharmacokinet- and utterly expressed in the gastrointestinal tract
ics of these drugs. (Strassburg et al., 1997; Hazama et al., 2013). Being
essential for glucuronidation of hydrophobic
IRINOTECAN PGX (IRINOGENETICS): endobiotic and xenobiotic compounds (Tukey and
IRINOTECAN AND UGT1A1 Strassburg, 2000), glucuronidation leads to the for-
The prodrug irinotecan (Camptosar, CPT-11), mation of water-soluble metabolites for different
which is a semisynthetic analog of the natural therapeutic drugs, including irinotecan, morphine
alkaloid camptothecin, was introduced in the (Coffman et al., 1997), acetaminophen (Court and
early 1990s into clinical applications as an impor- Greenblatt, 2000), immunosuppressants (tacroli-
tant DNA topoisomerase I inhibitor leading to mus, cyclosporine) (Strassburg et  al., 2001), anti-
cell death (Xu and Villalona-Calero, 2002). It has HIV drugs (raltegravir), chloramphenicol, and oral
been established as an effective treatment in sev- contraceptive agents (a-ethinylestradiol), as well
eral solid tumor malignancies such as lung cancer as for an array of compounds, such as bilirubin,
(Sevinc et al., 2011) and CRC either as a single agent bile acids, and steroid hormones (de Wildt et  al.,
or in combination with 5-FU (Stucky-Marshall, 1999). Glucuronidation of all these compounds is
1999; Pizzolato and Saltz, 2003; Ferraldeschi, 2010; mediated primarily by phase II biotransformation
Marcuello et al., 2011; Fujita et al., 2015). enzymes UGT1A isoforms (1A1, 1A7 and 1A9)
Having a complex metabolism, irinotecan is (Innocenti and Ratain, 2004; Guillemette et  al.,
metabolized by carboxylesterases principally in 2014), which is then excreted in the bile through
the liver, and also in various tissues, into an active the small intestine (Innocenti and Ratain, 2003;
cytotoxic metabolite SN-38 (7-ethyl-10-hydroxy- Ferraldeschi, 2010). UGT1A is encoded by the
camptothecin). SN-38, with poor aqueous solu- UGT1A gene located on chromosome 2q37 (van Es
bility that is at least 100-fold more potent than et  al., 1993). Irinotecan-related toxicities, such as
irinotecan (Kawato et al., 1991; Gupta et al., 1994; diarrhea and myelosuppression, which have been
Khanna et al., 2000; Mathijssen et al., 2001; Gagne associated with an increased level of SN-38, are the
et  al., 2002), binds the nuclear enzyme topoi- most important dose-limiting toxicities interfering
somerase I, which is essential for DNA synthe- with its optimal utilization (Innocenti and Ratain,
sis and replication (Innocenti and Ratain, 2003; 2006; Ramchandani et al., 2007; Etienne-Grimaldi
Ferraldeschi, 2010). SN-38 is further catabolized et al., 2015).
(glucuronidated) principally by hepatic phase II Increased or decreased UGT1A1 enzyme
biotransformation enzymes uridine 5′-diphospho- expression affects the glucuronidation activity and
glucuronosyltransferase (UDP-glucuronosyl trans- also the metabolic clearance of drugs (Sugatani,
ferase)—UGT 1A enzymes (1A1, UGT1A7, and 2013; Guillemette et  al., 2014). Respectively, it is
UGT1A9) to the inactive compound SN-38 gluc- well known that mutations in the UGT1A1 gene
uronide (SN-38G) (Gagne et  al., 2002; Innocenti have been implicated in Gilbert’s syndrome and
 Pharmacogenetics (PGx) for precision medicine in CRC chemotherapy  69

Crigler–Najjar type 2 syndrome leading to mild et  al., 2017) and neutropenia, which can be life-
hyperbilirubinemia as well as the more aggres- threatening (Marcuello et  al., 2004; Toffoli et  al.,
sive childhood subtype, Crigler–Najjar type 1 2006; Stewart et  al., 2007; Ramchandani et  al.,
(Bosma et al., 1995; Costa, 2006; Strassburg, 2008). 2007; Hoskins et  al., 2007; Kweekel et  al., 2008;
TATA box polymorphism in the promoter region Ruzzo et al., 2008; Palomaki et al., 2009; Hu et al.,
of UGT1A1, which can vary between 5 and 8 thy- 2010; Toffoli et  al., 2006; Glimelius et  al., 2011;
mine–adenine (TA) repeats, is the most thoroughly Tziotou et  al., 2014; Emami et  al., 2017; Liu  XH
studied polymorphism for irinotecan metabolism et  al., 2017), of which the incidence depends also
(Lankisch et  al., 2005). Respectively, UGT1A1*28 in part on patient characteristics such as age and
or TA7 (rs8175347) polymorphism is character- lifestyle factors (Marques and Ikediobi, 2010).
ized by seven TA repeats in the promoter region of A meta-analysis study of 821 patients assess-
UGT1A1) as opposed to six TA repeats representing ing the association between irinotecan dose and
the wild-type allele UGT1A1*1 or TA6 (Guillemette, risk of irinotecan-related hematologic toxici-
2003; Lankisch et al., 2005; Barbarino et al., 2014). ties (grade III–IV neutropenia) in patients with a
The length of this TA repeat sequence is inversely UGT1A1*28/*28 genotype demonstrated that the
correlated with UGT1A1 enzyme activity; thus, risk of experiencing irinotecan-induced hemato-
the *28 polymorphism results in a decreased rate logic toxicity in homozygous UGT1A1*28/*28 car-
of transcription initiation/expression of UGT1A1 rier patients was the function of the administered
affecting the elimination of substrate drugs. Extra irinotecan dose, but not only UGT1A1*28/*28
TA repeats impair proper UGT1A1 gene transcrip- status. Respectively, UGT1A1*28/*28 patients
tion leading to a decrease in the transcriptional had a greater risk of hematologic toxicities at high
activity of the gene by ≈70% in homozygous (>250 mg/m2) and moderate (150–250 mg/m2)
UGT1A1*28 (TA7/7) and ≈25% in heterozygous irinotecan dosage. However, no association was
UGT1A1*28 (TA6/7) (Bosma et  al., 1995; Tukey established in UGT1A1*28/*28 carrier patients at
et  al., 2002; Barbarino et  al., 2014). Furthermore, low irinotecan dosage (80–125 mg/m2) (Hoskins
individuals who are carriers of homozygous et  al., 2007). Moreover, the same study found
UGT1A1*28 genotype (TA7/7) have higher levels no  association between UGT1A1*28 status and
of serum bilirubin compared with those who have irinotecan-induced diarrhea. If these aforemen-
heterozygous UGT1A1*28 (TA7/6) or the wild- tioned results are bona fide, then the susceptibility
type allele (TA6/6) genotype (Barbarino et  al., of UGT1A1*28/*28 carrier patients to irinotecan
2014; Guillemette et al., 2014). UGT1A1*28 allelic toxicity (severe neutropenia) is dose-dependent
frequencies vary among ethnic groups, which are and increment or reduction could be indicative of
commonly found in African-American (≈43%) the possible involvement of additional novel target
and Caucasian (≈36%) populations; yet is the low- genes and pathways (Lévesque et al., 2013; Makondi
est in Asian populations (≈16%) (Ando et al., 1998; et al., 2017) Additionally, nongenetic lifestyle fac-
Beutler et  al., 1998; Premawardhena et  al., 2003). tors (de Jong et al., 2008b; Marques and Ikediobi,
Although some studies are less compelling based 2010), such as smoking habits (Dumez et al., 2005),
on several scientific findings, it has been acknowl- medical conditions, including renal function (de
edged that patients with homozygotes (TA7/7) or Jong et al., 2008a), and comedication (Kiang et al.,
heterozygous (TA6/7) for UGT1A1*28 have a lower 2005) may lead to risk of toxicity. Additionally,
irinotecan glucuronidation capacity together with the frequencies of the genetic variations in genes
increased accumulation of cytotoxic active metab- encoding drug metabolizing enzymes (DMEs) as
olite SN-38 compared to wild-type UGT1A1*1 well as drug targets associated with therapeutic
(TA6/6) (Innocenti et al., 2004; Toffoli et al., 2006; phenotypes with known toxicity within a racially
Côté et al., 2007; Parodi et al., 2008; Braun et al., and ethnically defined population is important
2009; Takano and Sugiyama, 2017). Patients with to investigate and to take into consideration for
TA7/7 genotype are exposed to a higher risk of tox- dosing guidelines (Innocenti et al., 2002; Nguyen
icity including diarrhea (though there are few data et al., 2007; Shimoyama, 2010). Ethnic differences
regarding the relationship with diarrhea) (Gupta related to UGT enzymes should be considered in
et  al., 1994; Iyer et  al., 2002; Carlini et  al., 2005; order to determine gene polymorphisms as a pre-
Massacesi et  al., 2006; Tziotou et  al., 2014; Peng dictor of treatment outcome in patients receiving
70  Precision medicine for colorectal cancer

irinotecan-based chemotherapy (Nguyen et  al., and Prevention (EGAPP™) Working Group did
2007; Shimoyama et  al., 2010). Respectively, in not find enough evidence indicating benefit of
addition to UGT1A1*28 polymorphism in Asians, routine UGT1A1 genotyping for metastatic CRC
UGT1A1*6 polymorphisms are an important pre- cases being treated with irinotecan, whereas it
dictor of irinotecan-induced hematologic toxic- did not either conclude that it should never be
ity and severe diarrhea compared with that in used (EGAPP, 2009; https://www.cdc.gov/genom-
Caucasians (Gao et al., 2013; Hazama et al., 2013; ics/gtesting/egapp/recommend/ugt1a1.htm).
Han et  al., 2014; Yang et  al., 2015; Cheng et  al., Considering different recommendations from dif-
2014; Zhang et al., 2017). ferent organizations related to UGT1A1*28 and
Based on the aforementioned studies related irinotecan response there is obvious and urgent
to UGT1A1*28 status and irinotecan-based toxic- need to find additional genetic variants contribut-
ity, in 2005 the FDA (https://www.fda.gov/Drugs/ ing to variability in irinotecan pharmacokinetics
ScienceResearch/ucm572698.htm) approved a (Rosner et al., 2008).
UGT1A1*28 genotyping test based on individual Even though interindividual variability of iri-
patient tolerance to treatment without any empha- notecan response (dose/toxicity) has been attrib-
sis or clear recommendation on the necessity uted essentially to inherited UGT1A1*28 genetic
of screening patients for UGT1A1*28 mutation variation, mounting of scientific evidence under-
before the administration of irinotecan (preemp- lined the critical role of different UGT1 haplo-
tive PGx test) (Innocenti and Ratain, 2006; Perera types in the 3′ UTR and central region of the gene
et  al., 2008; Vivot et  al., 2015). However, leaving (Biason et al., 2008; Lévesque et al., 2013) in dif-
this preemptive PGx test initiative for UGT1A1*28 ferent ethnic populations other than UGT1A1*28.
to the discretion of the treating physician brings up Accordingly, UGT1A1*6 (211G > A or G71R,
the  issue of the importance of PGx education for rs4148323), particularly prevalent in the Asian pop-
the physicians (Frueh and Gurwitz, 2004; Perera ulation with frequency of 0.13 to 0.25, UGT1A1*60
et al., 2008; Zgheib et al., 2011). (–3279T >  G, rs4124874), UGT1A1*93
Given that irinotecan doses are usually lower (–3156G > A, rs10929302), UGT1A11A7*2 and*3
in Europe (180 mg/m2, biweekly, combination) (387T > G, 622T > C), UGT1A9*22 contribute
and Japan (150 mg/m2, biweekly, combination) to irinotecan response (Innocenti et  al., 2002;
than in the United States (350 mg/m2, triweekly, Carlini et  al., 2005; Saito et  al., 2009; Cecchin
monotherapy) (Takano and Sugiyama, 2017), a et  al., 2009; Xu et  al., 2013; Hazama et  al., 2013;
guideline from the European Society for Medical Crona et al., 2016; Cui et al., 2016; Bai et al., 2017;
Oncology (ESMO) considers UGT1A1 testing only Campbell et  al., 2017; Takano and Sugiyama,
if potentially severe toxicity related to irinotecan 2017). Additionally, genes encoding drug DMEs
treatment occurs underlying the importance of and transporters, such as CYP3A4/3A5 (cyto-
UGT1A1 PGx testing when irinotecan is used at chrome P450) (Santos et al., 2000; Sai et al., 2008;
high doses (300–350 mg/m2) (Schmoll et al., 2012). van der Bol, 2010), P-glycoprotein or multidrug
A guideline from the Dutch Pharmacogenetics resistance proteins MDR1 (ATP-binding cassette
Working Group (KNMP) and Japanese Society transporter ABCB1) (Smith et al., 2006; Li W et al.,
for Cancer of the Colon and Rectum (JSCCR) 2016), ABC transporter (ABCC5 ABCG1) (Di
both consider UGT1A1 PGx testing. However, Martino et al., 2011; Chen et al., 2015), SLCO1B1
while KNMP considers a reduction by 30% in the (solute carrier organic anion transporter fam-
initial irinotecan dose for TA7/7 carriers if the ily member 1B1) (Xiang et  al., 2006; Teft et  al.,
regimen contains >250 mg/m2 of irinotecan but 2015; Crona et al., 2016) as well as gene encoding
not for heterozygous (TA6/7) carriers that is in carboxylesterase 2 (activates the prodrug irino-
agreement with the FDA (Swen et  al., 2011), the tecan into SN-38) (CES2) (Khanna et  al., 2000;
guideline from JSCCR considers UGT1A1 PGx Cecchin et  al., 2005; Capello et  al., 2015) are all
testing as preemptive, before administering irino- implicated in irinotecan metabolism, and genetic
tecan to patients, particularly if the patient has a variations in all these genes could affect irinote-
high serum bilirubin level (Watanabe et al., 2015). can response. Related to inhibitor/inducers effect,
On the other hand, in 2009 a guideline from the CYP3A4 enzyme could be particularly important
Evaluation of Genomic Applications in Practice in drug–drug (e.g., ketoconazole) (Haaz et  al.,
 Pharmacogenetics (PGx) for precision medicine in CRC chemotherapy  71

1998; Charasson et  al., 2002; Kehrer et  al., 2002; UGT1A1, which could influence treatment toxicity
Sasaki et  al., 2013) and drug–lifestyle/herbal and efficacy (Yasar et al., 2013).
interactions (Goey et al., 2013). Respectively, clini-
cal outcome of irinotecan could probably be due 5-FU PGX
to the result of complex interplay between gene
encoding DMEs in metabolic detoxification path- 5-FU and dihydropyrimidine
ways phase I (CYP3A4, CYP3A5) and in phase II dehydrogenase (DPD)
(UGT) as well as transporters, suggesting that a The fluoropyrimidine anticancer drug 5-fluoro-
combined signature of the UGT1 haplotypes and uracil (5-FU) and its oral prodrug capecitabine
other genes might provide more precise informa- are the backbone of chemotherapy for colorectal
tion about irinotecan pharmacokinetics, pharma- cancer (Tanaka et  al., 2000). Even though 5-FU
codynamics affecting increased systemic exposure was reported in the 1950s to have anticancer activ-
to SN-38 on a cellular level that lead to increased ity (Heidelberger et al., 1957), its efficacy was first
irinotecan-based toxicity, such as neutropenia recognized in 1990 (Moertel et  al., 1990) before
and diarrhea, as well as efficacy (Mathijssen et al., it became the mainstay of therapy for advanced
2001; Innocenti et al., 2009; Crona et al., 2016). To or mCRC. There are three methods using FU:
this end, cognizance of these genomic variations FOLFOX contains infusional fluorouracil, leu-
probably improve the personalized approach for covorin, and oxaliplatin; FOLFIRI contains fluo-
irinotecan treatment. Though genotyping for the rouracil, leucovorin, and irinotecan; and XELOX
UGT1A1 polymorphisms could be important to contains capecitabine plus oxaliplatin (Aparicio
prevent severe toxicity such as neutropenia, pre- et al., 2005).
diction for resistance (high levels of UGT activ- Some of the patients (10%–30%) using these
ity) to irinotecan is not clear (Panczyk, 2014). regimens have grade 3 toxicity such as diarrhea,
Since the intratumoral SN-38 level can be altered nausea, mucositis/stomatitis myelosuppression,
by increased efflux through active transport out hand-foot syndrome, and occasionally cardiac
of cells, some in vivo and in vitro studies have toxicity. Along with this, FU leads to 0.5%–1.0%
been investigated with the role of polymorphisms mortality (grade 5). Therefore, it is important to
of ATP-binding cassette (ABC) transporter pro- identify biomarkers that predict 5-FU toxicity
teins, such as ABCC1/MRP1, ABCC2/MRP2, and (Rosmarin et  al., 2014). 5-FU metabolism com-
ABCG2/BCRP (Sun et al., 2012; Li W et al., 2016; prises a plurality of enzyme reactions. After the use
Tuy et al., 2016; Nielsen et al., 2017) as well as poly- of parenteral 5-FU, 70%–90% of the drug is metab-
morphisms of CYP3A4 and CYP3A5 on treatment olized with dihydropyrimidine dehydrogenase
resistance (Hammond et al., 2016). However, they (DPD). The dihydropyrimidine dehydrogenase
are inconclusive and need to be discovered more. DPD, which plays a key role in the metabolism of
Additionally, epigenetic changes, essentially the fluoropyrimidines, is encoded by the DPYD
hypermethylation, have also been implicated in gene. DPD  is the rate-controlling enzyme for
development of resistance processes (Hammond inactivation of 5-FU and more than 80% of 5-FU
et al., 2016). is metabolized by DPD in the liver to the inac-
Even if published clinical data are limited in tive metabolite 5,6-dihydro-5-fluorouracil (van
relation to the role of epigenetics in the regulation Staveren et al., 2013).
of UGTs, and for the DMEs in general, preclini- The DPYD gene contains 23 exons and is located
cal studies in colon cancer cell lines and clinical on chromosome 1p22. More than 30 genetic poly-
studies in colon tumors underlined the crucial role morphisms in DPYD caused reduced function or
of DNA methylation at CpG sites in the promoter nonfunctional DPD enzyme leading to reduced
region of the UGT1A1 gene in the silencing of clearance of 5-FU resulting in increased 5-FU
UGT1A1 expression in colon cancer and on cellular toxicity in CRC patients (Del Re et  al., 2010).
SN-38 detoxification influencing clinical response DPD deficiency occurs in 4%–5% of the popula-
to irinotecan (Gagnon et al., 2006; Bélanger et al., tion (Deenen et  al., 2011b). The most common
2010; Xie et  al., 2014) suggesting that differential genetic variant of the DPYD gene with partial or
methylation of the CpG site may explain interin- complete DPD deficiency is due to a G to A point
dividual variability in hepatic glucuronidation by mutation within the 5′-splicing site of intron 14
72  Precision medicine for colorectal cancer

(IVS14 + 1G > A) called DPYD*2A (rs3918290) to optimize and to apply quick DPYD PGx tests
polymorphism. DPYD*2A leads to catalytically for screening patients with DPD deficiency before
inactive enzyme with a frequency of approximately being treated with fluoropyrimidines for the first
1% in Caucasians. DPD activity is reduced by 50% time (van Staveren et  al., 2013; Etienne-Grimaldi
in heterozygous genotype resulting in increased et al., 2017). Hence, the recommended 5-FU start-
5-FU exposure. However, DPD activity in patients ing doses should be adjusted according to the
with homozygous DPYD*2A is about 0% (van enzyme activity of DPD (Henricks et  al., 2015;
Kuilenburg et al., 2001; Meulendijks et al., 2016a, Etienne-Grimaldi et al., 2017).
2016b; Deenen et al., 2016). In patients with com-
plete DPD, nonfloropyrimidin-based treatment 5-FU and thymidylate synthase (TS)
is recommended instead of fluoropyrimidines, or Thymidylate synthase (TYMS) is crucial enzyme
if fluoropyrimidines treatment is imperative for for providing essential nucleotide precursors (de
attentive monitoring, a starting dose reduction novo pyrimidine synthesis) in order to maintain
of approximately 10% is recommended (Caudle DNA synthesis and repair. Respectively, TYMS
et al., 2013). However, some variants do not lead to is implicated in the conversion of deoxyuridine
completely inactive enzymes; the variant c. 1129- monophosphate (dUMP) to deoxythymidine
5923C > G (rs75017182), known as haplotype B3, monophosphate (dTMP) (Carreras and Santi,
is a deeply intronic variant encoding partially 1995). Hence, TYMS expression is a rate-limiting
nonfunctional protein expression with enzyme step for cell proliferation as well as for cancer
activity 50% lower in homozygous individuals growth (Rustum, 2004). Compelling evidence
(Meulendijks et al., 2016a, 2016b). from several clinical studies have demonstrated
Although enzyme activities can be measured that TYMS protein and mRNA levels were higher
for toxicity estimation, these tests can be cumber- in different cancers (Berger and Berger, 2006),
some and expensive for routine use. Following ini- including breast cancer, lung cancer, gastric cancer,
tial reports linked to the use of lethal FU in the and CRC (Yamada et  al., 2001; Kamoshida et  al.,
severe DPD deficiency, several common genetic 2004; Popat et al., 2006), and this increase in TYMS
polymorphisms in genes involved in FU metabo- levels has been associated with poor clinical out-
lism and rare variants have been reported to affect come in these cancers. It is also shown that TYMS
the risk of adverse events (Henricks et al., 2015). expression correlates closely with transcription
Theoretically, prior to dosage adjustments, FU factor E2F1 expression in colon cancer specimens.
toxicity can be estimated by testing a panel of poly- In accordance with this TYMS—E2F1 correlation
morphisms. However, the important limitations TYMS could be considered as an E2F1-regulated
are the presence of different polymorphisms in the enzyme, which is essential for DNA synthesis and
same gene, as well as the involvement of patients in repair (Kasahara et al., 2000; Rahman et al., 2004).
different FU programs. At the same time, several Additionally, an in vitro study in immortalized
polymorphisms with lack of validation may have NIH/3T3 mouse fibroblast cells underlined onco-
been included in FU toxicity kits (Boisdron-Celle gene-like activity of TYMS that suggests a connec-
et al., 2007; Afzal et al., 2011). tion between TYMS-regulated DNA synthesis and
In order to accelerate clinical uptake of DPD the induction of a neoplastic phenotype (Rahman
testing, the following points are recommended: (a) et al., 2004; Bertino and Banerjee, 2004).
a consensus definition of DPD deficiency should TYMS has a potential therapeutic interest as
be decided internationally to determine the inci- critical target of chemotherapeutic agents 5-FUl,
dence of DPD deficiency by comparing different and its prodrug uracil-tegafur with leucovorin
study results; (b) sensitivity and specificity for the (UFT/LV), capecitabine, and methotrexate, which
prevention of fluoropyrimidine-induced toxicity exert their anticancer effects by inhibiting TYMS
should be determined in each test; (c) optimization (Longley et  al., 2003; Rustum, 2004). Moreover,
of a genotyping and phenotyping strategy should overexpression of TYMS is linked to chemotherapy
be explored to determine the most appropriate test resistance (Panczyk, 2014; Hammond et al., 2016).
for detection of patients with DPD deficiency; and TYMS is encoded by the TYMS gene located
finally (d) the test should focus specifically on cost on human chromosome 18 TYMS gene, which
effectiveness. To this end, there is an obvious need contains seven exons, and various types of
 Pharmacogenetics (PGx) for precision medicine in CRC chemotherapy  73

polymorphisms of the TYMS gene have been linked (Mandola et  al., 2003). The presence of this poly-
to variable TYMS protein levels and a therapeutic morphism leads to the tri-allelic locus of 2R,
outcome related to 5-FU treatment (Ferraldeschi, 3RG, and 3RC with two different levels of TYMS
2010). A common 28-bp variable number tandem activity such as a high expression group (carri-
repeat (VNTR) polymorphism (rs34743033) is ers of 2R/3G, 3C/3G, and 3G/3G genotype) and a
found in the 5′ untranslated regions (5′UTR) of low expression group (carriers of 2R/2R, 2R/3C,
TYMS (Horie et al., 1995) that occurs in a different and 3C/3C genotypes). This SNP occurs within
population with a variable number of reiterations upstream stimulatory factor 1 (USF-1), leading to a
leading to occurrence of two alleles, two tandem decreased transcriptional activity of TYMS gene by
repeats (2R), and three tandem repeats (3R). While converting the transcriptional activity from a 3R to
3R represents the wild-type form, 2R represents a 2R, thereby decreasing TS levels (approximately
the variant form with three different genotypes: three- to four-fold) of the wild-type triple repeat
2R/2R, 2R/3R, and 3R/3R (Marsh et  al., 2001; (3RG/3RG) variant, which is comparable with
Kawakami and Watanabe, 2003). The number of the 2R/2R or 2R/3RG genotypes (Kawakami and
tandem repeats affects TYMS activity levels medi- Watanabe, 2003; Gusella and Padrini, 2007). The
ated through effects of the repeats on translation presence of this SNP (rs2853542) and double poly-
efficiency. Subjects with two tandem repeats, three morphism in the TYSM gene could explain low
tandem repeats, or a heterozygous genotype were TS expression level together with good response
observed (Marsh and McLeod, 2001; Ferraldeschi, to 5-FU chemotherapy (Panczyk, 2014). In other
2010). A triple repeat presence (TSER*3) results words, testing for all these polymorphisms may
in 2.6 times more mRNA expression than dou- help stratify patients at risk for 5-FU treatment.
ble repeats (TSER*2) (Marsh and McLeod, 2001; It is underscored that patients with TSER*3 and
Ferraldeschi, 2010; Lima et  al., 2013). TSER*3, TSER*3 G > C SNPs may be at greatest risk for
which leads to higher TYMS mRNA expression, is toxicity and decreased response. Furthermore,
significantly higher in the Asian population com- an important ins/del polymorphism of the hexa-
pared to other ethnic groups (67% in Chinese and nucleotide TTAAAG sequence (6-bp insertion/
about 40% in Caucasians) (Marsh et al., 1999). On deletion) at 1494 position (1494del6, rs34489327,
the other hand, according to world population also referred to as rs16430) within the untranslated
studies, alleles with 4, 5, and 9 TSER repeats have transcriptional region (3′UTR) of the TYMS gene
been observed mostly in Asian and African popu- is described (Ulrich et al., 2000). Consequently, this
lations (Marsh et al., 1999; Marsh et al., 2000) and ins/del polymorphism yields three different geno-
in particular TSER*5 is found mostly in Asians types: ins/ins (homozygous for insertion of 6 bp),
(0.18%). Although the outcome of published stud- del/del (homozygous for deletion of the 6 bp), and
ies are conflicting, it is anticipated that the TSER ins/del (heterozygous). Hence, it was demonstrated
genotype is associated with 5-fluorouracil toxicity that individuals with homozygous deletion (del/
and efficacy (Ferraldeschi, 2010; Schwarzenbach, del) had significantly lower mRNA levels of the
2010; Panczyk, 2014). According to the results of TYMS gene, which was also associated with greater
different studies the 3R allele is responsible for sensitivity to 5-FU–based therapy as compared
an approximately four times higher mRNA level to individuals with homozygous insertion (ins/
of the TYMS gene, suggesting that individuals ins) (P = 0.017) (Kawakami and Watanabe, 2003;
homozygous for TSER*3 (TSER*3/TSER*3) have Mandola et al., 2003; Stoehlmacher et al., 2008).
much less favorable response to 5-FU treatment Related to 5-FU chemotherapy resistance and
as compared to those who had the 2R/2R with TS, higher TS expression in MSI-H is both sporadic
low TYMS mRNA (Marsh et  al., 2001; Pullarkat (86%) and hereditary (100%). Tumors compared to
et  al., 2001; Gosens et  al., 2008). However, the MSI-negative tumors had been demonstrated as a
2R/2R genotype is associated with an increased strong predictive factor for nonresponse (resistant)
risk of 5-FU toxicity (Marsh et al., 2001; Pullarkat to adjuvant 5-FU chemotherapy (Gatalica, 2014;
et  al., 2001; Gosens et  al., 2008). An additional Gatalica et al., 2015, 2016). In a poorly differenti-
SNP (rs2853542) of guanine instead of cytosine ated gastric cancer cell line MKN45, the important
(G > C) at the 12th nucleotide of 3R alleles has role of decreased activity of orotate phosphori-
been described as two different alleles (3RC/3G) bosyltransferase (OPRT), which is involved in
74  Precision medicine for colorectal cancer

phosphoribosylation of 5-FU leading to tumor cytotoxic activity of 5-FU by increasing the for-
growth inhibition, is also suggested (Tsutani et al., mation and stability of the triple inhibitory com-
2008). However, a recent in vitro study in the same plex by increasing intracellular concentrations of
cell line (MKN45) hypothesized that irrespective CH2THF (Longley et al., 2003). While some stud-
of decreased OPRT levels, a decrease in the intra- ies demonstrated only the significant association
cellular FdUMP level could be a probable mecha- of MTHFR C677T polymorphism with increased
nism involved in the resistance to 5-FU (Mori 5-FU response but not for toxicity prediction
et  al., 2017). Though most retrospective clinical (Jakobsen et  al., 2005; Etienne-Grimaldi et  al.,
trials have shown that TYMS genotyping may 2010), the others also demonstrated the increased
help predict the 5-FU response, insufficiency of risk of undesirable side effects together with
TYMS genotyping alone in accurate prediction of increased 5-FU sensitivity survival in stage III and
outcome response to 5-FU is underlined (Vignoli stage IV CRC (Derwinger et al., 2009). However, in
et  al., 2011). To this end, a comprehensive evalu- some studies of advanced CRC, only populations
ation of each TYMS polymorphism in large-scale with the 1298CC genotype were associated with
prospective randomized control trials is needed to the risk of developing serious adverse events after
guarantee the efficacy of PGx testing for TYMS in 5-FU–based chemotherapy.
patients prior to 5-FU chemotherapy treatment. In conclusion, considering contradictory results
on MTHFR polymorphisms and 5-FU response
5-FU and MTHFR and toxicity, it still remains controversial whether
MTHFR (5,10-methylenetetrahydrofolate reduc- MTHFR polymorphisms can possibly predict tox-
tase) is a pivotal enzyme in folate metabolism icity or 5-FU response in patients treated with
catalyzing irreversible conversion of 5,10-methy- 5-FU (Panczyk, 2014; Ab Mutalib et al., 2017).
lenetetrahydrofolate (CH2THF) to 5-methyl-
tetrahydrofolate (CH3THF), a cosubstrate for Targeted therapies: PGx for somatic
homocysteine remethylation to methionine
(Goyette et  al., 1994). This enzyme is encoded by mutations in anticancer agents
MTHFR gene and is located on human chromo- While most of the standard cytotoxic chemo-
some 1 (Goyette et al., 1994). Two nonsynonymous therapies act on cancerous and all rapidly dividing
variants, C677T in exon 4 (Ala222Val, rs1801133) normal cells, cytostatic targeted therapies act on
and A1298C in exon 7 (Glu 429Ala, rs1801131), specific molecular targets that are associated with
have been identified for the MTHFR gene and have cancer (Calvo et al., 2016). To this end, in the era
been associated with decreased enzymatic activity of personalized medicine targeted therapies (also
and altered intracellular folate distribution (Frosst referred to as biologic treatments/molecularly tar-
et  al., 1995). Despite contradictory data, several geted drugs) involving tumor growth and progres-
clinical studies have demonstrated the potential sion is an attractive topic of oncology in patient
predictive role of MTHFR genetic variants in tox- stratification (Papadatos-Pastos et al., 2015), hence
icity and efficacy of 5-FU (De Mattia and Toffoli, the introduction of EGFR inhibitors has provided
2009). Since a reduction in MTHFR enzymatic new treatment options for metastatic CRC patients
activity could lead to a decrease in intracellular (Haddad et al., 2017).
CH3THF concentrations, it is hypothesized that
tumors exhibiting the rare MTHFR variants may TARGETING GENES
be more sensitive to 5-FU cytotoxicity compared
to patients with a wild-type genotype (Etienne- KRAS
Grimaldi et al., 2007). RAS mutational status is a pivotal factor when
The critical point of 5-FU activity is the for- using these targeted therapies. EGFR activa-
mation of an inhibitor triple complex; the active tion shows its effect via the RAS-RAF-MAPK
metabolite consists of 5-fluoro-2-deoxyuridine- and PI3K-AKT-mTOR pathways. Mutations in
5-monophosphate (5-FdUMP), TS, and 5,10 these signaling pathways may result in receptor-
methylenetetrahydrofolate, hence inhibition of independent continuous activation that results in
TS activity. Accordingly, it has been hypothesized unresponsiveness to the therapy. The EGFR anti-
that MTHFR polymorphisms may increase the bodies cetuximab, a chimeric IgG1 antibody, and
 Pharmacogenetics (PGx) for precision medicine in CRC chemotherapy  75

panitumumab, a humanized IgG2 antibody, both wild-type tumors, the addition of cetuximab to
bind to and block the EGFR and have proven effec- FOLFOX was associated with an increased objec-
tive in all lines of mCRC treatment (Siddiqui and tive response rate (61% versus 37%; p = 0.011) and
Piperdi, 2010). improved PFS (7.7 versus 7.2 months HR, 0.57; 95%
The association of KRAS gene mutation and CI, 0.36–0.91; p = 0.016) compared with chemo-
response to therapy was first reported in 2006 in therapy alone (Bokemeyer et al., 2009). New results
patients with metastatic colorectal cancer who from the OPUS study revealed that patients with
were treated with an EGFR agent (Lievre et  al., any activating mutation of KRAS or NRAS are
2006). In this study it was also noted that EGFR unlikely to benefit from the addition of cetuximab
overexpression has also been linked to poor prog- to FOLFOX4 (Tejpar et al., 2014).
nosis and increased risk of metastasis in colorec- Pooled individual patient data from OPUS and
tal cancer (Lievre et al., 2006). Studies have shown CRYSTAL studies were analyzed and 845 patients
that tumors with a mutation in exon 2 (codon 12 with KRAS wild-type tumors with the addition of
or 13 mutation) of the KRAS gene are unlikely cetuximab to chemotherapy improved OS (haz-
to get benefit from cetuximab or panitumumab ard ratio [HR] 0.81; p = 0.0062), PFS (HR 0.66;
treatment (Peeters et  al., 2013). Recent evidence p < 0.001) and ORR (odds ratio 2.16; p < 0.0001).
has suggested that other RAS mutations in exons Consequently, prognosis was worse in each treat-
3 and 4 of KRAS and NRAS genes may also be ment arm for patients with BRAF tumor muta-
predictive of unresponsiveness to anti-EGFR tions compared to those with BRAF wild-type
therapies (Doullard et al., 2013). A meta-analysis tumors (Bokemeyer et  al., 2012). However, in
including nine randomized controlled trials and contrast to other studies in phase III MRC COIN
a total of 5948 patients demonstrated tumors with trial, no benefit in OS or PFS was observed with
all RAS wild-type had significantly superior sur- the addition of anti-EGFR agents to FOLFOX or
vival with the anti-EGFR therapy compared to CapeOx as first-line treatments for patients with
tumors with KRAS exon 2 mutation along with locally advanced or metastatic CRC with wild-
other mutations, including KRAS exons 3 and 4 type KRAS exon 2 (Maughan et  al., 2011). Even
and NRAS exons 2, 3, and 4. Accordingly, these patients with wild-type tumors for all three genes
results suggest that extended RAS mutation test- (KRAS, BRAF, NRAS) did not show any evidence
ing in addition to KRAS exon 2 (KRAS exon 3 and of a benefit from the addition of cetuximab. This
4 and NRAS exon 2, 3, and 4) should be under- was explained by a more advanced stage of disease
taken before the administration of an anti-EGFR at the first presentation in the COIN trial than
mAb. A study evaluating addition of cetuximab other similar trials.
to chemotherapy (FOLFIRI) for metastatic CRC Panitumumab was also tested in the first-line
in the CRYSTAL trial indicated that first-line setting in metastatic CRC patients. In the PRIME
treatment with cetuximab plus FOLFIRI reduced study, RAS mutations were assessed in patients
the risk of progression of metastatic colorectal treated with FOLFOX4 with and without panitu-
cancer compared to FOLFIRI alone. Additionally, mumab. Patients with additional RAS mutation
this benefit of cetuximab was limited to KRAS other than at KRAS exon 2 showed lack of response
wild-type tumors (Van Cutsem et  al., 2009). to treatment with FOLFOX4/panitumumab versus
Follow-up and post hoc analysis showed signifi- FOLFOX4 alone. This particular group showed a
cant improvements in RR, PFS, and OS when significantly shorter median PFS and OS with the
treated with FOLFIRI/cetuximab compared with addition of anti-EGFR therapy. Patients with no
FOLFIRI alone in the RAS wild-type population RAS mutations treated with FOLFOX4/panitu-
compared to RAS mutations (KRAS exons 2, 3, mumab conferred a longer median PFS and OS
and 4, and NRAS exons 2, 3, and 4) (Van Cutsem compared with FOLFOX4 alone. Addition of pani-
et al., 2011, 2015). tumumab to FOLFIRI in the second-line setting
Moreover, efficacy of cetuximab in combination was assessed by Peeters et  al. (2015). Consistent
with FOLFOX as a first-line treatment for mCRC with the PRIME study, among all RAS wild-type
was tested in the OPUS study (Bokemeyer et  al., (KRAS at exons 2, 3, and 4, and NRAS at exons 2,
2011). A retrospective evaluation of the OPUS 3, and 4) patients, improvements in outcome were
study showed that in patients with KRAS exon 2 observed in PFS when panitumumab was added to
76  Precision medicine for colorectal cancer

chemotherapy as a second-line treatment (Peeters response (Corcoran et al., 2015). Although com-
et al., 2014). bined BRAF/MEK inhibition did show more
activity than single-agent BRAF inhibition, the
BRAF effectiveness of combined BRAF/MEK inhibi-
BRAF is a targetable mutation in many tumors tion was still limited. Analysis of tumor biopsies
including malign melanoma. Recently, improve- demonstrated that the level of phosphorylated
ment in both overall survival and response rate ERK, which could be a marker of inhibition, was
was achieved in metastatic malignant melanoma decreased in the posttreatment biopsies analyzed
with the use of new BRAF inhibitors, namely, (Corcoran et al., 2015). Triplet therapy that inhib-
vemurafenib and dabrafenib. Unlike the results its the BRAF, MEK, and EGFR pathways appears
in melanoma, the response of BRAF inhibi- promising in BRAF-mutated colorectal cancer
tors was not successful in metastatic CRC. In a that does not respond to BRAF inhibition alone.
small phase I study in patients with BRAFV600E A phase I/II trial evaluated the combination of
mutant metastatic disease, only 1 of 19 patients dabrafenib (150 mg twice daily) plus panitu-
had a partial response with single-agent vemu- mumab (6 mg/kg every 2 weeks) in 20 patients
rafenib (Kopetz et al., 2015). Similarly, a phase I with BRAF V600E–mutated metastatic colorectal
trial of the BRAF inhibitor dabrafenib also failed cancer (Atreya et al., 2015). The study included 35
in BRAF-mutated colorectal cancer (Kefford patients on the triplet arm, with a dose-escalation
et al., 2010). scheme. Of these patients, 24 received the phase
One reason for unresponsiveness of BRAF inhi- II regimen of dabrafenib (150 mg twice daily)
bition in BRAF-mutated colorectal cancer could plus panitumumab (6 mg/kg every 2 weeks) and
be explained by Prahallad et  al.’s study that sug- trametinib (2 mg daily). The primary endpoint
gested that BRAF (V600E) inhibition causes a was overall response rate, and progression-free
rapid feedback activation of EGFR and this results survival was a secondary endpoint. A total of 2
in continuous proliferation in the presence of of 20 patients (10%) responded to the doublet of
BRAF (V600E) inhibition. Because of low levels dabrafenib and panitumumab; 1 had a confirmed
of EGFR expression in melanoma cells, they are complete response. With triplet therapy, how-
not subject to this feedback activation (Prahallad ever, 9 of 35 patients (26%) responded, including
et  al., 2012). In light of this data, Prahallad et  al. 1 complete response. The unconfirmed response
could show that the combination of vemurafenib rate was 34%, and the stable disease rate was 60%.
and the EGFR-blocking antibody cetuximab was Median progression-free survival was 3.4 months
significantly more effective in mice bearing BRAF with the doublet and 4.1 months with the trip-
V600E mutated colorectal cancer than each of the let. Median duration of response for all patients
drugs alone. Connolly et al. (2014) presented a case receiving the triplet was 5.4 months. Tumor biop-
report of cetuximab and vemurafenib combination sies before and after 2 weeks of treatment revealed
for a refractory patient with BRAF mutant meta- that phosphorylated ERK staining intensity was
static colon cancer. consistently reduced by triplet therapy, but not
The next step in developing targeted therapy doublet therapy (Atreya et  al., 2015). The other
for BRAF-mutated colorectal cancer was based reason for unresponsiveness of BRAF inhibitors
on the hypotheses that inhibiting BRAF and its in colon cancer could be explained by PI3K/AKT
downstream signaling partner MEK would be pathway activation in BRAF-mutated colorectal
more effective by increasing the level of inhibition cancers. It was shown that the phosphoinositide
of the MAPK pathway and blocking some mecha- 3-kinase (PI3K)/AKT pathway was activated to a
nisms of acquired resistance. Recently tested was greater extent in BRAF-mutated colorectal can-
the strategy of combined BRAF and MEK inhi- cers than in melanoma. As a result of PI3K/AKT
bition with dabrafenib and trametinib in BRAF pathway activation de novo an acquired resistance
V600–mutant CRC, which showed that 5 out of to BRAF inhibition was observed in cell lines and
43 patients (12%) achieved a partial response or murine models. This activation occurs through
better, including one complete response, with PIK3CA mutation or PTEN loss and is associated
duration of response >36 months; 24 patients with the inherent CpG island phenotypes associ-
achieved stable disease as best confirmed ated with BRAF-mutated CRC through epigenetic
 Pharmacogenetics (PGx) for precision medicine in CRC chemotherapy  77

silencing. In vivo, vemurafenib combined with Furthermore, dominant-negative mutants of RAF

either inhibitors of AKT or methyltransferase can impair RAS transforming activity, confirm-
showed greater tumor growth inhibition than ing that inhibition of RAF is a viable therapeutic
vemurafenib alone. Clones with acquired resis- approach.
tance to vemurafenib in vitro showed PI3K/AKT Sorafenib is one of the most promising agents of
activation with EGFR or KRAS amplification RAF kinase inhibitors. It targets the ERK pathway
(Mao M et al., 2012). and inhibits angiogenesis through VEGFR-2 and
As a conclusion, the existence of the specific PDGFR tyrosine kinases and their associated sig-
mutation in colon cancer does not confer sufficient naling cascades. In a preclinical model, sorafenib
sensitivity to a single agent targeting the particular demonstrated inhibition of the MAPK pathway in
mutation. Combination of inhibitors in the RAS/ colon cancer cell lines expressing mutant KRAS or
RAF/MEK/ERK pathway seem to be promising in wild-type or mutant BRAF (Wilhelm et al., 2004;
BRAF-mutated colorectal cancers. Samalin et al., 2014).
Vemurafenib is a specific BRAF inhibitor and
has proven effects in BRAF-mutated malignant
TARGETING THE PATHWAYS
melanoma; however, it is ineffective in BRAF-
RAF kinases mutant colorectal cancer (Prahallad et  al., 2012).
RAF is the most characterized downstream effec- Another BRAF inhibitor dabrafenib also failed in
tor kinase of RAS. The Raf serine threonine kinase BRAF-mutant colorectal cancer (Holderfield et al.,
family consists of three isoforms: RAF-1 (C-RAF), 2014). One possible explanation of resistance of
A-RAF, and B-RAF. Mutations in RAF-1 (C-RAF) BRAF inhibitor therapy in colon cancer could be
or A-RAF have not been detected in human can- EGFR-mediated reactivation of MAPK signaling
cers (Chong et al., 2003). (Samalin et al., 2014). A blockade of BRAF causes
B-RAF is the strongest RAF kinase that acti- rapid feedback activation of EGFR and triggers
vates MEK. Due to many missense mutations in the sustained MAPK signaling and cell proliferation
BRAF gene, it could be a potential target for vari- via activation of RAS and CRAF. Besides this,
ous types of cancers. Activation of the upstream transactivation of BRAF–CRAF heterodimers in
signaling pathway of RAS results in RAF activa- the presence of vemurafenib may result in resis-
tion, which induces a downstream signal transduc- tance (Corcoran et al., 2015).
tion cascade beginning with MEK (Pearson et al., To overcome EGFR-activated resistance of Braf
2001). Increased MEK activity promotes activation inhibitors, Connolly et al. (2014) presented a case
of two kinases, namely, ERK1 and Erk2. Signaling report of cetuximab and vemurafenib combina-
through the ERK pathway may lead to increased tion for a refractory patient with BRAF-mutant
growth factors and cytokines expression so that metastatic colon cancer. Recently Hong et  al.
this pathway is further stimulated in an autocrine (2014) presented a phase 1B study of vemurafenib
fashion (Steelman et  al., 2004). Through interac- in combination with irinotecan and cetuximab
tion with the RB gene p27kip, ERK pathways may in patients with BRAF-mutated advanced can-
also increase cyclin D and E expression in the cell cers and metastatic colorectal cancer. The results
cycle (Steelman et al., 2004). Besides this, ERK sig- showed that the combination of vemurafenib with
naling regulates cell motility, extracellular matrix irinotecan and cetuximab was well tolerated in
remodeling, and induce the production of VEGF patients with BRAF-mutated mCRC. Even with
and other antiangiogenic factors together with a low vemurafenib dose, PRs were seen in 4 of 5
inactivation of caspase and BAD. ERK also down- evaluable mCRC patients in the first cohort (Hong
regulates p53, which is a pivotal tumor suppressor et al., 2014).
protein. ERK could be a potential target in many Combined BRAF and MEK inhibition with
cancer types. Because RAF is the only activator of dabrafenib and trametinib in BRAF V600–mutant
ERK, drugs targeting the ERK pathway at the level colorectal cancer was recently tested (Corcoran
of RAF may be particularly useful to control tumor et  al., 2015). The results showed that 5 out of 43
development and progression. In colon cancer, patients (12%) achieved a partial response or bet-
mutations of B-RAF and K-RAS are often found ter, including 1 (2%) complete response, with dura-
in a mutually exclusive fashion in the same tumor. tion of response >36 months; 24 patients (56%)
78  Precision medicine for colorectal cancer

achieved stable disease as the best confirmed MEK

response. Although combined BRAF/MEK inhi- Mitogen-activated protein kinase (MAPK) cas-
bition did show more activity than single-agent cades are key signaling pathways involved in the
BRAF inhibition, the effectiveness of combined regulation of normal cell proliferation, survival,
BRAF/MEK inhibition was still limited. and differentiation. Dysregulation of MAPK cas-
Another reason for unresponsiveness was cades contribute to cancer development. Four dis-
shown by Mao et  al. that the phosphoinositide tinct MAPK cascades have been identified. They
3-kinase (PI3K)/AKT pathway in colon cell lines is are called extracellular signal-regulated kinase
activated to a greater extent in BRAF-mutated CRC (ERK1/2), c-Jun N-terminal kinase (JNK), p38,
tumors than in melanoma (Mao M et  al., 2012) and ERK5. Each of them is composed of three
PI3K/AKT pathway activation results in both de sequentially acting kinases, activating one after the
novo and acquired resistance to BRAF inhibition. other (MAPKKK/MAP3K, MAPKK/MAP2K, and
Combination therapy that inhibits the BRAF, MAPK) (Thompson and Lyons, 2005).
MEK, and EGFR pathways appears promising in MEK proteins belong to a family of enzymes
BRAF-mutated colorectal cancer. A phase I/II trial that are involved in the four MAP kinase signal-
evaluated the combination of dabrafenib (150 mg ing pathways (Thompson and Lyons, 2005). Seven
twice daily) plus panitumumab (6 mg/kg every 2 MEK enzymes have been described each of which
weeks) in 20 patients with BRAF-mutated meta- selective phosphorylate serine/threonine and tyro-
static colorectal cancer (Atreya et al., 2015). With sine residues within their specific MAP kinase
this triple therapy, the unconfirmed response rate substrate. MEK1 and MEK2 are the most com-
was 34%, and the stable disease rate was 60%. One mon ones and they are involved in the RAS/RAF/
patient had complete response. Median progres- MEK/ERK pathway (Cargnello and Roux, 2011).
sion-free survival was 4.1 months with the triple This pathway is activated after ligand binding,
therapy. Median duration of response was 5.4 which results in membrane-bound GTPase activa-
months. tion of RAS. After that, Ras recruits and activates
Regorafenib is a novel oral multikinase inhibi- RAF kinases (Roskoski, 2010). The activated RAF
tor that targets protein kinases involved in tumor kinases interact and activate MEK1/2, which in
angiogenesis (VEGFR1–3 and tyrosine kinase turn catalyze the phosphorylation of particular
with immunoglobulin and epidermal growth residues in the activation of ERK1/2 (Sacks, 2006).
factor homology domain 2 [TIE2]), oncogen- Unlike RAF and MEK1/2 kinases, ERK1 and ERK2
esis (KIT, RET, and RAF) and the tumor micro- have a wide variety of cytosolic and nuclear sub-
environment (platelet-derived growth factor strates (Cargnello and Roux, 2011). Activated ERKs
receptor-b and fibroblast growth factor receptor function in a diverse cellular process like prolif-
1 [FGFR1]). Recently, in the CORRECT trial, eration, survival, differentiation, motility, and
regorafenib demonstrated a significant improve- angiogenesis. The RAS/RAF/MEK/ERK pathway
ment in overall survival (6.4 months in the is activated in human cancers via several different
regorafenib group versus 5 months in the pla- mechanisms. Increased ERK1/2 signaling is often
cebo group (p = 0.0052)) in a phase III study in due to direct mutational activation of the RAS and
patients with refractory metastatic CRC (Grothey B-RAF genes. This results in ERK activation and
et  al., 2013). Regorafenib was also tested in an ERK-dependent growth transformation (Roskoski,
Asian pop­u lation in the phase III CONCUR trial, 2010). Furthermore, the only known substrates of
which also supports overall survival benefit in RAF are MEK1 and MEK2, and no substrates for
previously treated metastatic colorectal cancer MEK have been identified other than ERK1 and
patients (Li J et  al., 2015). Although regorafenib ERK2, which means MEK inhibitors would be a
is an RAF inhibitor, its antitumor effect mostly potent inhibitor of RAS- and RAF-mediated acti-
depends on inhibition of angiogenesis in tumor vation of ERK. This hypothesis promotes treatment
cells (Tampellini et al., 2016). Preclinical studies by MEK inhibitors in colorectal cancer because
demonstrate that regorafenib acts independently KRAS and BRAF mutation–positive colorectal
of the mutational status of KRAS and BRAF cancer tumor cells are expected to exhibit elevated
(Grothey et al., 2013). ERK activation. Therefore ERK1/2 levels could be
 Pharmacogenetics (PGx) for precision medicine in CRC chemotherapy  79

used as a biomarker of response to MEK inhibition. S6K1 and 4EBP1, which are involved in mRNA
In the study of Yeh et al. (2009), they found that the translation. mTORC2 activates PKC-α and AKT,
majority of colorectal cancer cell lines show growth and regulates the actin cytoskeleton (Populo et  al.,
inhibition using MEK inhibitors, specifically those 2012). mTORC1 signaling occurs as an early event in
that are BRAF or KRAS mutation positive; how- the process of tumorigenesis. It has been described in
ever, ERK activation did not correlate reliably with a mouse model of adenomatous polyposis (FAP) and
BRAF and KRAS mutation status. It was noted that as an intestinal polyp formation. Therefore mTORC1
ERK is not differentially activated in tumor tissue. could be a target for drug development of colon pol-
These results show complexities of MEK inhibitors yps and cancer (Crunkhorn, 2015).
as anti-Ras therapy. Rapamycin, a macrolide antibiotic, was the
Currently 13 MEK inhibitors have been clini- first mTOR inhibitor discovered. Several deriva-
cally tested, but only trametinib as a selective tives of rapamycin (e.g., sirolimus) or with more
inhibitor of MEK 1 and 2 has emerged as the first favorable pharmacokinetic and solubility proper-
MEK inhibitor to show favorable clinical efficacy ties (e.g., temsirolumus and everolimus) have been
(Akinleye et  al., 2013). Initial clinical results of synthesized. The mTORC1 complex is sensitive to
MEK inhibitors have yielded limited single-agent rapamycin; however, mTORC2 is considered resis-
activity in colorectal cancer (Adjei et  al., 2008; tant to rapamycin (Klümpen et al., 2010).
Balmanno et  al., 2009; Bennouna et  al., 2011). Combinational therapies consisting of mTOR
Recently Spreafico et  al. (2013) designed a study inhibitors and other agents such as antiangio-
to test the combination of a MEK Inhibitor, selu- genic molecular or anti-EGFR agents have shown
metinib, and the Wnt/calcium pathway modula- promising results. One example of this combina-
tor cyclosporin A in preclinical models of CRC tion could be the trial of Gulhati et  al. (2012). In
to overcome resistance to MEK inhibition. It was this trial, sorafenib, a multikinase inhibitor RAF,
suggested that the combination of MEK blockade VEGFR, and PDGFR, was used in combination
and Wnt pathway modulation has shown synergis- with rapamycin in a preclinical setting. A com-
tic antiproliferative effects in preclinical colorectal bination of rapamycin with sorafenib synergisti-
cancer models (Spreafico et al., 2013). cally inhibits proliferation of CRC cells. CRCs
MEK inhibitors show promise in colorectal with KRAS and PIK3CA mutations are partially
cancer. Studies that promote the combination of sensitive to either rapamycin or sorafenib mono-
BRAF/MEK/EGFR inhibitors are still recruiting therapy, but highly sensitive to combination treat-
patients and results are awaiting (ClinicalTrials. ment. It was concluded that the combination with
gov Identifier:NCT01750918). sorafenib enhances therapeutic efficacy of rapamy-
cin on induction of apoptosis and inhibition of
MTOR cell-cycle progression, migration, and invasion
Mammalian target of rapamycin (mTOR), a serine/ of CRCs (Gulhati et  al., 2012). The phase I study
threonine tyrosine kinase, regulates cell division evaluating efficacy and tolerability of the antian-
and growth, largely by promoting key anabolic giogenic agent tivozanib (an oral VEGF receptor-1,
processes. Upon activation, mTOR relays the cellu- -2, -3 inhibitor) plus everolimus in metastatic CRC
lar signal to downstream effectors to stimulate cell indicated that the oral combination of tivozanib
growth, proliferation, and angiogenesis (Laplante and everolimus was well tolerated, with stable dis-
and Sabatini, 2012). The PI3K/Akt/mTOR pathway ease achieved in 50% of patients with refractory,
is activated by PI3K gene mutation and amplifica- metastatic colorectal cancer (Wolpin et al., 2013).
tion, AKT mutation and amplification, or loss of mTOR inhibitors have been evaluated in combina-
the PTEN tumor suppressor. Dysregulation of this tion with antiEGFR mAbs such as cetuximab and
pathway is seen in 40%–60% of patients with colon panitumumab (Cuinci et  al., 2014; McRee, 2014).
cancer (Laplante and Sabatini, 2012). In a phase 1 trial of everolimus a combination with
mTOR forms two distinct complexes, namely, cetuximab was safely administered in patients
mTORC1 and mTORC2 (Kim et al., 2002). The com- with refractory CRC (Cuinci et al., 2014). Another
plexes are constituted by different proteins and have phase I trial of a regimen of everolimus in addi-
distinct functions in the cell cycle. mTORC1 activates tion to 5-FU/LV and mFOLFOX6 appears safe and
80  Precision medicine for colorectal cancer

tolerable, but the further addition of panitumumab of predictive and prognostic markers for disease
resulted in an unacceptable level of toxicity (Cuinci susceptibility and PGx (Daly, 2010). In order to be
et al., 2014; McRee, 2014). significantly correlated with complex disease traits,
In conclusion, the mTOR pathway is another in GWAS, the requisite is that SNPs should meet
potential target of therapy in metastatic colorectal the stringent genome-wide threshold that is usually
cancer patients and the combinational approach set to P < 5 × 10−8 (Zeng et al., 2015).
may further increase the efficacy of treatment. Although common genetic variants are largely
responsible for complex diseases such as CRC, sev-
FROM GENOME-WIDE eral strong lines of evidence indicate that rare vari-
ASSOCIATION STUDY (GWAS) ants, which are MAF frequencies <1%–5%, play a
TO GENETIC RISK SCORE (GRS) crucial role in complex disease etiology with possible
FOR CRC larger genetic effects than common variants (Lettre,
2014; Bomba et al., 2017). Even if GWAS have been
Since the completion of both the Human Genome successful in identifying common genetic varia-
Project (HGP) (Venter et  al., 2001) and the tions with well-established “risk” loci associated to
International HapMap project (Frazer et al., 2007), various complex diseases, because of the small effect
recent progress in genotyping technology has facil- of these individual SNPs on complex common dis-
itated the use of GWAS as a “gene-hunting study” ease, the clinical utility of these genetic variations is
for linking specific genetic variants with human skeptical in personalized risk prediction related to
disease by many thousand loci simultaneously lifestyle, demographic, and clinical factors for the
(Hindorff et al., 2009; Visscher et al., 2017). disease in question (Visscher et al., 2017). Moreover,
One of the essential ideas underpinning pre- detecting rare variants from GWAS data is difficult
cision medicine is to determine the information and often explains only a small proportion of trait
about the genetic risk factors for different diseases heritability, in explaining the “missing heritability”
in order to inform patients and to introduce pro- of complex human trait (Blanco-Gómez et al., 2016).
active changes in behavior such as environmental/ The advance of GWAS has identified multiple com-
lifestyle factors. Thus, understanding gene–envi- mon genetic variants influencing susceptibility to
ronment interactions (G × E) relevant to genetic different types of cancers, including CRC, which has
variations for common and complex human dis- a strong heritable background with an increasing
eases is an important challenge for precision medi- number of susceptibility loci. To date, GWAS studies
cine (Simon et  al., 2016). Accordingly, starting on CRC have identified common SNP variants con-
with the determination of phenotype susceptibili- ferring CRC susceptibility risk at more than 40 loci
ties, disease conditions, and drug efficacy or side (Whiffin et  al., 2014; Abulí et  al., 2016; Frampton
effects, GWAS, which allow interrogation of more and Houlston, 2017). However, most individual
than a million SNPs in the genome, aims to dis- genetic risk variants at these loci confer a modest
cover genomic level variance among individuals effect in explaining only a small fraction of the CRC
or different case-control groups within a popula- risk (relative risk-RR) as well as small fraction of the
tion in a holistic and agnostic manner (Hindorff heritable component leading to limited ability of
et  al., 2009; Stranger et  al., 2011). That is to say, genetic testing for CRC prediction. Hence, since it is
GWAS are a type of case-control study employing intimidating to consider any single SNP as a clinical
whole genome comparisons of allele frequencies test for CRC risk susceptibility, there is no clinical
in which individuals with the disease condition utility for public health. Likewise, the value of these
being studied are compared to similar individuals SNPs as prognosticators has been explored only in a
without the disease condition under study (i.e., to few studies (Dai et al., 2012; Noci et al., 2016).
find correlations between specific SNPs and a phe- Like most of the complex diseases, genetic
notype). Furthermore, GWAS is a very useful pre- roots of CRC are multifactorial, nondeterministic,
eminent tool for identifying common SNPs with a and genetic variants scattered across the genome,
minor allele frequency (MAF) >5% as a risk factor contributing small risks for CRC. That is to say,
associated with multifactorial complex disorder information from multiple SNPs is needed to char-
(Panagiotou et al., 2010; De La Vega, 2011; Lowe and acterize genetic susceptibility to CRC. Respectively,
Reddy, 2015), which accelerated the identification in the context of precision medicine, since the risk
 Precision/personalized medicine approaches for colon cancer driven by systems biology  81

estimation, risk stratification, and appropriate human microbiota that could be considered an
treatment for CRC events is crucial for personalized environmental factor modulating the host metab-
prevention counseling and therapy, it has been sug- olism comprehends the populations of more than
gested that combining multiple single genetic vari- 400 microbial species (the commensal bacteria as
ants with minor effects into polygenic risk scores symbiotic and pathogenic, viruses and fungi, sin-
(PRS) (also called genetic risk score, or genome-wide gle-celled animals, such as protists, and archaea)
score) from GWAS along with conventional nonge- with more than 100 trillion cells (approximately 10
netic factors might reduce bias and might improve times the total number of cells in the human body)
power disease risk prediction for complex disorders encoding 100-fold more unique genes than our
including CRC (Garcia-Closas et  al., 2014; Cooke own genome (Ley et al., 2006) living in or on the
and Igo, 2016; Chen et  al., 2016; Frampton et  al., human body as important determinants of health
2016; Läll et al., 2017). PRS/GRS, which is a numeric and disease, the genetic constitutes of microbial
summary measure of genetic risk, does not depend cells are called microbiome (“second genome”). It
on single genetic variants (a single common SNP), is also important to note that the human micro-
but is a number based on variation in multiple sets biome contains 150 times more genes than the
of SNPs estimated from GWAS to serve as the best human genome (Qin et al., 2010). The composition
prediction for the trait (Dudbridge, 2013; Krapohl of the gut microbiome was originally coined by
et al., 2017). Accordingly, since the PRS/GRS aggre- Joshua Lederberg as “the ecological community of
gates the effects of thousands of SNPs from GWAS, symbiotic, and pathogenic microorganisms shar-
it appears to be a more realistic tool and so there is ing our body space” and describing the collective
growing interest in constructing whether the PRS/ genome of our indigenous microbes (microflora),
GRS is associated with the CRC risk. Personalized and so it is equivalent to the gut metagenome
screening using PRS/GRS has potential implication (Lederberg and McCray, 2001). Despite the impor-
in optimizing population screening by stratify- tance of genetic polymorphisms in human disease
ing for CRC patients according to their PRS/GRS risk management and in therapeutic outcomes, it is
as to low, medium, or high risk leading to an ear- well recognized the role of the microbiome which
lier screening for high-risk individuals as well as a has been, more often than not, overlooked (Grice
decrease in the number needed to treat (NNT) for and Segre, 2012). Therefore, manipulation of the
chemoprevention (Pharoah et al., 2008; Shaik et al., gut microbiota may be an important therapeutic
2015; Frampton et al., 2016). Additionally, an indi- strategy in order to dissect and characterize the
vidual with a high gene risk score for CRC but who association of the modulation and perturbation of
has not yet developed the disease, could be advised the human microbiome with disease as well as to
to change his or her lifestyle factors (Maher, 2015). characterize a strategy regulating energy balance
Taken together, since aggregate multiple genetic (harvested from food) in the body (Kim B-S et al.,
loci identified by GWAS will influence the CRC 2013; Manzat-Saplacan et al., 2015). Metagenomics,
risk, the combination of genetic factors and con- which is one of the fastest burgeoning fields of the
ventional nongenetic factors will lead to more important “-omics” languages (Maccaferri et  al.,
precise and personalized risk prevention and a pre- 2011; Banerjee et  al., 2015; Proctor, 2016; Hizel,
dictive model for colorectal cancer (Iwasaki et al., 2017), has since its advent gained importance in
2017; Frampton and Houlston, 2017). bringing out the crucial link between sporadic
CRC (Banerjee et al., 2015; Yu et al., 2017) as well
PRECISION/PERSONALIZED as metabolic disorders, such as obesity and type 2
MEDICINE APPROACHES FOR diabetes (Baothman et  al., 2016). Metagenomics
COLON CANCER DRIVEN BY analyzes the structure and functional potential
SYSTEMS BIOLOGY of microbial communities in their native habitats
through the sequencing of PCR amplicons from
Microbiome-driven carcinogenesis in the ribosomal 16S (NGS-based 16S rRNA sequenc-
colorectal cancer (metagenomics) ing) and whole metagenome shotgun (WMS)
sequencing (Thomas et al., 2012; Jovel et al., 2016).
The terms microbiome and microbiota are used Respectively, accumulated metagenomic studies
often interchangeably; however, whereas the point out the important crucial role of gut microbial
82  Precision medicine for colorectal cancer

dysbiosis (i.e., microbial imbalance on or inside health and disease (Vital et al., 2014). Furthermore,
the body), which fosters the discharge of bacterial the effect of butyrate on epigenetic gene regulation
genotoxins, metabolites leading to chronic inflam- is also demonstrated. Hence, different composi-
mation and oxidative DNA damage in the etiology, tions of gut microbiota in cancer as well as in obe-
and progress of sporadic CRC and also particu- sity and type 2 diabetes could affect the epigenetic
lar enriched bacterial species identified in fecal regulation of genes, such as genes encoding SCFA
microbiomes from CRC compared to healthy fecal receptors FFAR2 and FFAR3 causing changes in
microbiomes (Sobhani et al., 2013; Gao et al., 2015; gene expression and signaling of FFARs by histone
Ray and Kidane, 2016). Accordingly, the presence deacetylases (HDACs) inhibition and hyperacety-
of several oral pathogens, particularly enriched lation (Davie, 2003).
abundance of Fusobacterium, positively correlated Knowledge of the composition of the gut micro-
with lymph node metastasis, have been reported biome ecosystem and gut microbial marker, such
underlying the importance of its investigation as as the ratio of SCFA, are future important key ele-
a marker for colorectal cancer presence, risk, or ments of precision medicine, though much more
prognosis (Castellarin et al., 2012; Flanagan et al., research is needed to achieve more precise diag-
2014; Purcell et al., 2017). nosis and therapeutic intervention and might
Diet is a major external force shaping gut com- eventually aid in lifestyle interventions for disease
munities and the clear association related to con- prevention and/or modulation (Kasubuchi et  al.,
tribution of diet (meat, fruits, vegetables, and fiber) 2015). Recent metagenomic analysis underscores
to microbiota (Read and Holmes, 2017; Akin and the acetyl-CoA pathway as the main pathway
Tözün, 2014) and to pathogenesis of CRC (Ryan- for butyrate production in healthy individuals.
Harshman and Aldoori, 2007) is increasingly well Accordingly, dietary interventions such as high
acknowledged. Respectively, since microbiota dietary fiber intake, which are the important
has a crucial role in maintaining gut homeosta- source of SCFAs influencing microbial composi-
sis and host health, diet–microbiota–microbiome tion, could be considered as an option in personal
interactions cannot be defined without reflection nutrition in order to increase, for example, butyr-
on host–diet interactions. A recent large cohort ate concentrations and reduce insulin resistance
GWAS identified a significant interaction between for the engagement against metabolic syndrome
rs4143094 SNP on chromosome 10p14 near the such as diabetes and also it could be an important
gene GATA3 (GATA binding protein 3), though avenue in drug development (example NaB might
interaction of dietary factors with genetic variants be a promising ­molecule) for the prevention and
to modify risk of colorectal cancer is still under treatment of CRC (Davie, 2003; Alenghat, 2015;
challenging consideration (Figueiredo et al., 2014). Kasubuchi et al., 2015).
There is evidence to suggest an inverse rela- Taken together, since every individual is dif-
tionship between the level of dietary fibers, which ferent and because of the potentially modifiable
is a source of short-chain fatty acid (SCFA), and nature of gut microbiota identifying the role of
the incidence of human CRC (Park et  al., 2005; particular bacterial species in CRC development,
Donohoe et  al., 2011; Vital et  al., 2014). As the the increased understanding of particular bacte-
main energy sources for the host, SCFAs can rial species in sporadic CRC pathogenesis through
be used for de novo synthesis of lipids and glu- metagenomic systems biology approach will cer-
cose (den Besten et  al., 2013). Distinct composi- tainly help to improve CRC predictions and pre-
tion of gut microbiota produces different SCFAs, vention (Eklöf et  al., 2017; Purcell et  al., 2017).
and butyrate is one of the SCFA end products of Hence, besides genomic information, better com-
colonic fermentation known to modulate several prehension of the impact of gut microbial dys-
cellular processes such as cell differentiation and biosis together with novel gut microbiota genetic
inhibition of cell proliferation in different tumor biomarkers could provide encouraging direction
cell lines (Hague et  al., 1996; Hizel et  al., 1999; in early diagnosis, prognosis, prevention, and
Comalada et al., 2006). treatment of CRC (Sobhani et al., 2013; Cho et al.,
Being the main energy source for colonocytes, 2014; Gao et al., 2015; Manzat-Saplacan et al., 2015;
production of butyrate plays a critical role in Purcell et al., 2017; Yu et al., 2017).
 Precision/personalized medicine approaches for colon cancer driven by systems biology  83

Metabolomics of colon cancer potential hits can be assessed further in primary

tissues reflects “metabolic codes” cells, biofluids, and tissue samples as biosignatures
for CRC (Halama et al., 2015).
of patients
CRC is linked to mutations in DNA, amino acids, Glycomics, as part of an array
pentose-phosphate pathway carbohydrates, and of metastatic codes
glycolytic, gluconeogenic, and tricarboxylic acid
intermediates (Brown et  al., 2016). In the year Although carbohydrates are considered to be the
2016, Brown et  al. (2016) utilized a nontargeted most abundant macromolecules in nature, they have
global metabolome perspective to investigate been underestimated and considered less function-
human CRC, adjacent mucosa, and stool. In this ally significant than nucleic acids and proteins. Due
research, they identified metabolite profile varia- to their large and complex heterogeneity, carbohy-
tions between CRC and adjacent mucosa from drates and glycoconjugates have been difficult to
patients who had parts of their colon removed. isolate from their natural sources owing to the lack
Furthermore, the analyses of the metabolic path- of efficient technologies. Consequently, this dimin-
ways unveiled connections among complex net- ished the recognition of their significance in the
works of metabolites (Halama et al., 2015; Brown most basic biological practices, leading to the lack
et  al., 2016; Farshidfar et  al., 2016). The extensive of exploration of the carbohydrates in the biological
and thorough characterization of tumor phe- arena. Nevertheless, in the past few decades complex
notypes facilitated the therapeutic approaches carbohydrate expression has been acknowledged to
(Halama et al., 2015). Lately, omics strategies have be crucial in the establishment of living networks,
shed light upon tumor biology. Such applications including cellular signaling, cellular differentiation,
have been to a large degree executed to avail biosig- and the immune system (Adamczyk et  al., 2012;
natures to examine the disease and ameliorate ther- Drake, 2015; Siobhan et  al., 2015). The recogni-
apeutic results. The orchestration of meta­bolomics tion of this structure and functional connection of
for studying tumorigenesis is particularly instru- carbohydrates facilitated chemical and biological
mental, since it demonstrates the biochemical protocols to unveil new areas in molecular biology,
outcome of a number of tumor-specific functional coining the term glycobiology in the early 1980s.
changes associated with the disease (Halama et al., State-of-the-art protocols have thrived ever since,
2015; Farshidfar et al., 2016). In 2015 Halama et al. leading to the examination of “glycomaterials” for
experimented utilizing nontargeted metabolo- their glycan function (Adamczyk et al., 2012; Drake,
mics-based mass spectroscopy together with ultra- 2015; Siobhan et al., 2015). The normal activity of gly-
high-performance liquid chromatography and gas cosylation is interfered during cancer cell formation
chromatography in order to perform metabolic (Kim and Varki, 1997; Kannagi et al., 2004). These
phenotyping of four cancer cell lines: two colon alterations in the tumor cells lead to the restructur-
cancer (HCT15, HCT116) and two ovarian cancer ing of the cell surface glycans of the transformed
(OVCAR3, SKOV3). They then applied the MetaP tumors regulating metastatic potential of the cancer
server to statistically analyze their data (Halama cells (Shriver et al., 2004). Additionally, parallel to
et  al., 2015). According to Halama et  al.’s study, the alterations in the glycosylation status of the can-
where a total of 225 metabolites were detected in cer cell surfaces, gene expression profiles of the car-
all colon cancer cell lines mentioned herein, 67 of bohydrate binding proteins get modified during this
these molecules exceptionally discriminated colon neoplastic transformation causing a comprehensive
cancer cells from ovarian cancer cells. Metabolic alteration between the glycans and their associated
biomarkers identified in this study suggested receptors (Kannagi et al., 2004; Shriver et al., 2004).
elevation of β-oxidation and urea cycle metabo- The mechanism via which glycosylation modifica-
lism in colon cancer cell lines. As a conclusion tion triggers cancer cell metastasis and invasion are
Halama et  al.’s study provided a panel of specific yet to be determined, but the functions of the par-
metabolic identifiers between colon and ovarian ticular cell surface bound glycoproteins and their
cancer cell lines. The novel findings can be consid- carbohydrate motifs have been recently identified
ered as potential drug targets. Furthermore, these (Dall’Olio et al., 2012; Geng et al., 2012). Due to the
84  Precision medicine for colorectal cancer

recent developments in the glycoanalytical method- databases are being utilized to elucidate and build
ologies, a better comprehension of the carbohydrate libraries between proteins and other macromol-
connection to the cell surface lipids and proteins ecules. These efforts help to reveal large network
has emerged (Siobhan et  al., 2015). Learning how graphs of interactoms exhibiting the complexity
the glycan determinants work on cancer-associated of association of differentially expressed proteins
proteins helped to unveil a new level of complexity, or genes in a disease status (Sjöblom et al., 2006).
facilitating better comprehension of how neoplasia According to the Nibbe et al. (2009) study investi-
changes the normal process in cells. Additionally, gating the interactom graph, which portrays 70 dif-
the orchestration of glycan determinants on can- ferent proteins known to interact with the so-called
cer can have connections with cancer cell metas- gatekeeper gene APC, reveals mutations in over 90%
tasis, proliferation, survival, and immune escape. of CRC tumors (Sjöblom et al., 2006). Conversely,
This extensive knowledge about the cancer glycome in another well-studied WNT-signaling pathway,
has implications to be very effective in the cancer also known to be dysregulated in CRC, there were
field (Siobhan et  al., 2015). However, much still relatively few interactions involving APC (Sjöblom
remains to be unveiled pertaining to the functional et al., 2006). Evidently, the variations in the activity
interactions of cell surface glycans with the regula- of these interactions may explain the heterogene-
tion of the metastatic potential of the cancer cells. ity of tumors in patients at large. The dissimilarity
Thorough knowledge of carbohydrate remodeling in the aggressiveness of the patient tumors may be
in cancer will require complete profiling of glyco- the origin of the differential outcome of treatments
sylation patterns in the tumor microenvironment repeatedly observed in the clinic (Yu et al., 2005).
(Siobhan et al., 2015). Improved biosignatures of CRC will be essential
to elucidate the coordinated, differential expres-
Interactomics in colon cancer tissues: sion of numerous proteins and/or genes that syn-
Where are we? ergistically expedite or delay the orchestration of
networks responsible for CRC (De Las Rivas and
Chromatin looping can make it possible for distant de Luis, 2004; Mathivanan et al., 2006; McMurray
regulatory elements to affect the gene expression et  al., 2008). Previously, Mathivanan et  al. (2006)
of their target genes located far away upstream or evaluated this in the human protein–protein inter-
downstream from where these cis- and transregula- action (PPI) data set in the public domain. Human
tory elements are located. Chromosome conforma- PPIs mark a new chapter in biomarker discovery
tion capture (3C2)–based methods can be utilized in neoplasia as they present an operative setting
to determine physical interactions between enhanc- in which the PPIs examine the mechanistic role of
ers and promoters (Spilianakis et al., 2005; Simonis proteins and genes found to be notably differen-
et al., 2006; Jäger et al., 2015). GWAS helped deter- tially expressed by traditional means (Mathivanan
mine SNPs that are linked to complex diseases. It et  al., 2006). Interactomics provide a database of
has been determined that there may exist many protein–protein networks, which may be exam-
SNPs that are located within regulatory elements ined for differentially active interactions between
and influence through long-range regulation of tumors and control samples. This in turn provides
gene expression (Pomerantz et al., 2009; Ahmadiyeh a strong basis for integrative strategy, facilitating
et al., 2010; Zhang et al., 2012). Consequently, pro- enhanced and functional biosignatures of cancer.
tein–DNA interaction can be facilitated upon A proteomics-first approach furnishes an uncon-
dynamic modification of chromatin architecture. ventional method to explore the functional subsets
Functionality of proteins in a cell depends on of proteins discriminative of CRC. Consequently,
many parameters. Protein–protein interactions to put all the perspectives mentioned herein in a
and the interactions of the proteins with drugs and nutshell, systems biology–based strategies combine
other macromolecules in the cell contribute to the all the details about the cell, and create a potential
formation of cellular phenotypes. Thus far it is not to convey more powerful and precise classifiers
very clear as to how these interactions are regulated. of diseases, in contrast with traditional protocols
Nevertheless, there is a wide range of evidence that (Mathivanan et al., 2006; Jonsson and Bates, 2007;
neoplasia is caused by impairments of networks Nibbe et al., 2009; Hudler et al., 2014; Tanase et al.,
operated by dysfunctional proteins. Computer 2016; Brandi et al., 2017).
 High-throughput genomic technologies for detecting Potential “driver” genomic alterations in CRC  85

HIGH-THROUGHPUT GENOMIC (NGS) has transformed the speed and throughput

TECHNOLOGIES FOR DETECTING of classifying cancer-related genomic aberrations
POTENTIAL “DRIVER” GENOMIC (Tran et al., 2012; Kim TM et al., 2013; Pellicer et al.,
ALTERATIONS AS A RISK 2017; Wang S et al., 2017). The challenge is how to
take advantage of this state-of-the-art technology
ASSESSMENT TOOL AND FOR
to better comprehend the fundamental molecular
TREATMENT SELECTION IN CRC mechanism of colorectal cancer and to better rec-
The application of personalized cancer medicine ognize clinically applicable genetic/genomic biosig-
(PCM) relies upon the following arguments: natures for diagnosis and precision medicine (Kim
TM et al., 2013). Lately, the development of NGS has
1. Human diseases involve numerous genetic dramatically altered the speed and throughput of
anomalies. DNA sequencing (Ajay et al., 2011). Conventionally,
2 . Oncogenes and tumor suppressor genes genome sequencing has been expensive and labo-
are responsible for a certain subset of these rious, but novel methodologies have reduced both
anomalies. of these restraints (Metzker, 2009; Shendure, 2011).
3. Individualized anticancer agents are available The growing number of tailored therapeutics
that effectively regulate these targets (Tran has generated an increased demand for genetic/
et al., 2012). genomic anomalies in clinical samples in a more
convenient (cheaper and timely) manner (Tran
This section focuses on the technologies under- et  al., 2012; Armengol et  al., 2016). The collection
lying high-throughput genomics for detecting of repetitive mutations is on the rise, and these
potential genomic alterations as risk evaluation mutations have been attractive in testing to assess
for CRC. Additionally, it inspects the preliminary their part in predicting mutations for a number
results of next-generation genome sequencing of molecularly targeted factors in development.
analysis of CRC genomes for personalized medi- This demands a necessity for higher-throughput
cine and microarrays for diagnostic and treatment genotyping technologies (D’Haene et  al., 2015;
selection in CRC, and reviews the challenges met Armengol et al., 2016).
in the discovery phase of new genetic aberra- High-throughput genotyping technologies,
tions. Since the completion of the Human Genome composed of microarrays and multiplexed tests,
Project, the appearance of the scientific age of have been efficiently utilized for genotyping clini-
“omics” has transformed the cancer research. cal CRC specimens (Dias-Santagata et  al., 2010;
The International Cancer Genome Consortium D’Haene et  al., 2015). CRC can be regarded as a
(ICGC) is synchronizing scientific research tar- complex disease considering the common variant
geted at discovering genomic modifications that hypothesis with a polygenic model of inheritance.
are linked to cancer (Tran et al., 2012; Sikdar et al., The genetic constituents of common complex dis-
2016; ICGC International Data Access Committee, eases correlate with the variants of moderate out-
2016). Discoveries along these lines greatly depend come (Esteban-Jurado et  al., 2014). Thus far, 30
on the development of novel tools in molecular common, low-penetrance susceptibility variants
diagnostics, especially in the area of whole genome have been discovered for CRC (Esteban-Jurado
sequencing. The enhanced timelines and cost- et  al., 2014). Lately, novel sequencing methodolo-
effective technologies have expedited discovery in gies including whole genome and exome sequenc-
cancer genomics as well as in translational medi- ing enable a new approach to expedite the discovery
cine (Tran et  al., 2012). Nevertheless, challenges of previously undescribed susceptibility genes
still exist prior to PCM becoming available for the responsible for human diseases. Through utilizing
population at large on an everyday bases. whole genome sequencing, germ-line mutations in
the POLE and POLD1 genes have been identified
Next-generation analysis of CRC to be accountable for a novel form of CRC genetic
genomes for precision medicine susceptibility named polymerase proofreading-
associated polyposis (Esteban-Jurado et al., 2014).
The advancement in the new DNA sequencing tech- Consequently, simultaneous sequencing of gene
nology recognized as next-generation sequencing groups via NGS will help analyze all applicable
86  Precision medicine for colorectal cancer

driver mutations of CRC in parallel, which will development of diagnostic technologies will also
be required in routine clinical applications in the facilitate screening protocols of CRC in a multi-
coming years (Huth et al., 2014). step process that results from genetic or epigenetic
alterations. Cancers with high degrees of methyla-
Microarrays in diagnostics and tion (the CpG island methylator phenotype, CIMP)
treatment selection in CRC display a clinically discrete group that is classified
by epigenetic variability (Issa, 2004, 2008). Thus, a
Administration of new molecular diagnostic methylome signature in CRC should be described
assays may facilitate a better survival rate for CRC employing a methylation microarray analysis (i.e.,
patients (Huth et al., 2014; Mahasneh et al., 2017). Illumina HumanMethylation27 array) (Issa, 2004).
Protocols for molecular evaluation of single genes This will lead to identification of a defined list for
(i.e., TP53 and/or KRAS) in addition to microar- methylome specific genes that will eventually pro-
ray-based methodologies are inevitable. Innovative vide better clinical management of CRC patients
screening applications are needed to be able to in the future (Issa, 2004; Ashktorab et  al., 2013;
detect precancerous state and early-stage malig- Kim et al., 2014). Nevertheless, additional studies
nancies in healthy populations to facilitate correc- are necessary to improve specificity and sensitivity
tive therapeutic interventions (Huth et  al., 2014; prior to initiating clinical trials.
Mahasneh et al., 2017). A recently developed non- CRC is a multifactorial condition that emerges
invasive multitarget stool DNA kit (Cologuard™) due to genomics, genetics, and epigenetic changes
for CRC screening involves KRAS mutation analy- in a lot of tumor suppressor genes, oncogenes, mis-
sis, as well as testing of the methylation condition match repair genes, and cell cycle regulatory ele-
of the NDRG4 and BMP3 genes (Huth et al., 2014; ments. These molecular aberrations are considered
Mahasneh et  al., 2017). Additionally, besides the as candidate CRC biosignatures. They can present
fecal occult blood testing (FOBT) and colonos- the physicians with diagnostic, prognostic, and
copy, other novel assays also offer approaches for treatment response details. The overall aim is to
early detection of CRC such as the blood-based determine applicable, affordable, and appropri-
Septin 9 DNA methylation test (Huth et al., 2014). ate biosignatures, which will be instrumental in
Liquid biopsy analysis also presents diagnostic patient care decisions, resulting in direct benefits
potential for screening CRC developing resistance to patients (Mahasneh et al., 2017).
to treatment. Determination of additional novel It is essential that the scientific methods con-
molecular biomarkers and diagnostic methods in tinue to be tested prior to administering the novel
CRC will facilitate early detection and targeted omics-based technologies toward the management
therapy of colorectal cancer (Huth et  al., 2014; of patient care. Nevertheless, these novel applica-
Mahasneh et al., 2017). tions provide substantial capacity for improved
The MS status of CRC could also be predicted care of cancer patients notably by enabling novel
depending on miRNA expression profiles (Schepeler trial designs, which will provide us with the
et al., 2008; Carvalho et al., 2017; Liu D et al., 2017; answers to rapidly translate these findings from
Strubberg and Madison, 2017). Spotted locked bench to bedside.
nucleic acid (LNA)–based oligonucleotide microar-
rays can be utilized to screen the expression profiles CONCLUSION AND PERSPECTIVE
of miRNAs. Hence, miRNAs can also be presented
as promising tools to classify colon cancers as either In this postgenomic era, fast-tracked development
MSI or MSS (T Schepeler et  al., 2008; Carvalho of omics languages are on the verge of transform-
et al., 2017; Liu J et al., 2017; Strubberg and Madison, ing health care leading to better understanding
2017). Mutations within the TP53 gene are the most of the molecular underpinnings of different poly-
common genetic/genomic aberrations in human genic multifactorial chronic complex diseases
cancers such as CRC. The GeneChip p53 assay is including cancer, which represents the prepon-
based on the recently designed oligonucleotide derance of health care. Hence, together with envi-
microarray technology (Takahashi et al., 2003). ronmental factors, the deployment of precision
Considering the presence of epigenetic changes medicine relies on the ability to identify accurately
and posttranslational modifications in CRC, relevant patient subgroups from extensive clinical
 References 87

and precise genomic data sets in order to create a Abdulla MH, Valli-Mohammed MA, Al-Khayal K
“bespoke” health care approach to the detection, et al. Cathepsin B expression in colorectal
prevention, and treatment of diseases (Hizel, 2017). cancer in a Middle East population: Potential
Taking into consideration the genetic environ- value as a tumor biomarker for late disease
ment of diseases in individual patients together stages. Oncol Rep 37(6);2017:3175–80.
with the nongenetic exposures, personalized Abulí A, Castells A, Bujanda L et al. Genetic vari-
genetic screening could answer some important ants associated with colorectal adenoma sus-
pressing questions related to the existence of dis- ceptibility. PLOS ONE 11(4);2016:e0153084.
ease, prediction of disease occurrence, treatment Adamczyk B, Tharmalingam T, and Rudd PM.
choice through PGx testing (Hizel, 2017), and strat- Glycans as cancer biomarkers. Biochim
ification of subjects according to their polygenic Biophys Acta 1820(9);2012:1347–53.
risk score (PRS) (low, medium or high) leading to a Adams JM, and Cory S. The Bcl-2 apoptotic
decrease in the number needed to treat (NNT) for switch in cancer development and therapy.
chemoprevention (Frampton et al., 2016). However, Oncogene 26(9);2007:1324–37.
in spite of the quickening pace of omics technolo- Adjei AA, Cohen RB, Franklin W et al. Phase
gies and a host of available literature on biomarkers I pharmacokinetic and pharmacodynamic
in CRC, currently only a few biomarkers are used study of the oral, small-molecule mitogen-
in daily clinical practice such as KRAS, BRAF, and activated protein kinase kinase 1/2 inhibi-
MSI (Ballman, 2015; Carethers and Jung, 2015; tor AZD6244 (ARRY-142886) in patients
Gatalica et al., 2015; Mahasneh et al., 2017). Even if with advanced cancers. J Clin Oncol
UGT1A1*28 PGx testing is approved by the FDA, it 26;2008:2139–46.
is not routinely used for prediction of toxicity of iri- Afzal S, Gusella M, Vainer B et al. Combinations
notecan in clinical practice resulting in the dearth of polymorphisms in genes involved in
of alternative treatments for risk genotype carrier the 5-Fluorouracil metabolism pathway
patients (Innocenti and Ratain, 2006; Vivot et al., are ­associated with gastrointestinal tox-
2015). To this end, considering the high incidence icity in chemotherapy treated colorec-
of CRC with high significant cost for society, there tal cancer patients. Clin Cancer Res
is a growing need for validation of efficient prog- 17(11);2011:3822–9.
nostic or predictive biomarkers in large prospec- Aguirre-Portolés C, Fernández LP, and Ramírez
tive cohorts (Duffy and Crown, 2008; Danese and de Molina A. Precision nutrition for target-
Montagnana, 2017; Kamel and Al-Amodi, 2017; ing lipid metabolism in colorectal cancer.
Lemery et  al., 2017). Moreover, since science is Nutrients 9(10);2017, pii: E1076.
moving toward more interdisciplinary involvement Ahlquist T, Bottillo I, Danielsen SA et al. RAS
of multiple disciplines and because of the multi- signaling in colorectal carcinomas through
disciplinary nature of clinical cancer genomics, alteration of RAS, RAF, NF1, and/or RASSF1A.
an “integrated knowledge ecosystem” with con- Neoplasia 10(7);2008:680–6.
tributions from all sectors of science and society Ahmadiyeh N, Pomerantz MM, Grisanzio C et al.
is needed for successful application of precision 8q24 prostate, breast, and colon cancer
genomic medicine (Lemay et al., 2017). risk loci show tissue-specific long range
Last, precision medicine could be desirable for ­interaction with MYC. Proc Natl Acad Sci USA
the society inasmuch as it will be consistent with 107(21);2010:9742–6.
social, ethical, local technological, and local dis- Ajay SS, Parker SC, Abaan HO, Fajardo KV, and
ease burden realities of the society in countries Margulies EH. Accurate and comprehensive
with different income levels (Hizel, 2017). sequencing of personal genomes. Genome
Res 21(9);2011:1498–5.
REFERENCES Akin H, and Tözün N. Diet, microbiota, and
colorectal cancer. J Clin Gastroenterol
Ab Mutalib NS, Md Yusof NF, Abdul SN, and 48(Suppl. 1);2014:S67–9.
Jamal R. Pharmacogenomics DNA biomarkers Akinleye A, Furqan M, Mukhi N, Ravella P, and
in colorectal cancer: Current update. Front Liu D. MEK and the inhibitors: From bench to
Pharmacol 8;2017:736. bedside. J Hematol Oncol 6;2013:27.
88  Precision medicine for colorectal cancer

Akshatha C, Mysorekar V, Arundhathi S, colorectal ­cancer. Gastrointest Cancer Res

Arul P, Raj A, and Shetty S. Correlation 5(1);2012:19–27.
of p53 Overexpression with the clinico- Armanios M, and Blackburn EH. The telomere
pathological prognostic factors in colorec- syndromes. Nat Rev Genet 13;2012:693–704.
tal ­adenocarcinoma. J Clin Diagn Res Armengol G, Sarhadi VK, Ghanbari R et al.
10(12);2016:EC05–8. Driver gene mutations in stools of colorectal
Aleksandrova K, Pischon T, Jenab M et al. carcinoma patients detected by targeted
Combined impact of healthy lifestyle factors next-generation sequencing. J Mol Diagn
on colorectal cancer: A large European cohort 18(4);2016:471–9.
study. BMC Med 12;2014:168. Arnold M, Sierra MS, Laversanne M,
Alenghat T. Epigenomics and the microbiota. Soerjomataram I, Jemal A, and Bray F. Global
Toxicol Pathol 43(1);2015:101–6. patterns and trends in colorectal cancer inci-
Alexandrova EM, Mirza SA, Xu S, Schulz- dence and mortality. Gut 66(4);2017:683–691.
Heddergott R, Marchenko ND, and Moll UM. Aschacher T, Wolf B, Enzmann F et al. LINE-1
p53 loss-of-heterozygosity is a necessary induces hTERT and ensures telomere main-
prerequisite for mutant p53 stabilization tenance in tumour cell lines. Oncogene
and gain-of-function in vivo. Cell Death Dis 35(1);2016:94–104.
8(3);2017:e2661. Ashkenazi A. Directing cancer cells to
Amado RG, Wolf M, Peeters M et al. Wild-type self-destruct with pro-apoptotic recep-
KRAS is required for panitumumab efficacy tor agonists. Nat Rev Drug Discov
in patients with metastatic colorectal cancer. 7(12);2008:1001–12.
ICE 26(10);2008:1626–34. Ashktorab H, Belgrave K, Hosseinkhah
Amelio I, Knight RA, Lisitsa A, Melino G, and F et al. Global histone H4 acetyla-
Antonov AV. p53MutaGene: An online tool to tion and HDAC2 expression in colon
estimate the effect of p53 mutational status adenoma and ­carcinoma. Dig Dis Sci
on gene regulation in cancer. Cell Death Dis 54(10);2009:2109–17.
7;2016:e2148. Ashktorab H, Rahi H, Wansley D et al. Toward a
Ando Y, Chida M, Nakayama K, Saka H, and comprehensive and systematic methylome
Kamataki T. The UGT1A1*28 allele is signature in colorectal cancers. Epigenetics
relatively rare in a Japanese population. 8;2013:807–15.
Pharmacogenetics 8(4);1998:357–60. Atreya CE, Cutsem V, Bendell J et al. Updated
Andreyev HJ, Norman AR, Cunningham D, efficacy of the MEK inhibitor trametinib (T),
Oates JR, and Clarke PA. Kirsten ras muta- BRAF inhibitor dabrafenib (D), and anti-EGFR
tions in patients with colorectal cancer: The antibody panitumumab (P) in patients (pts)
multicenter “RASCAL” study. Natl Cancer Inst with BRAF V600E mutated (BRAFm) meta-
90(9);1998:675–84. static colorectal cancer (mCRC). J Clin Oncol
Andreyev HJ, Norman AR, Cunningham D et al. 33;2015 (suppl; abstr 103).
Kirsten ras mutations in patients with colorec- Augustine T, Maitra R, and Goel S. Telomere
tal cancer: The “RASCAL II” study. Br J length regulation through epidermal growth
Cancer 85;2001:692–6. factor receptor signaling in cancer. Genes
Antelo M, Balaguer F, Shia J et al. A high degree Cancer 8(5–6);2017:550–8.
of LINE-1 hypomethylation is a unique feature Augustine TA, Baig M, Sood A et al. Telomere
of early-onset colorectal cancer. PLOS ONE length is a novel predictive biomarker
7(9);2012:e45357. of sensitivity to anti-EGFR therapy in
Aparicio J, Fernandez-Martos C, Vincent JM metastatic colorectal cancer. Br J Cancer
et al. FOLFOX alternated with FOLFIRI as 112(2);2015:313–8.
first-line chemotherapy for metastatic Aydin Son Y, Tüzmen Ş, Hızel C. Designing and
colorectal cancer. Clin Colorectal Cancer implementing pharmacogenomics study. In:
5(4);2005:263–7. Omics for Personalized Medicine. Barh D,
Armaghany T, Wilson JD, Chu Q, and Dahawan D, Ganguly NK (Eds), New Delhi,
Mills G. Genetic alterations in Springer, 2013, pp. 97–117.
 References 89

Baba Y, Nosho K, Shima K et al. Relationship of with poor survival in a population-based

CDX2 loss with molecular features and prog- series of colon cancers. Int J Cancer
nosis in colorectal cancer. Clin Cancer Res 122(10);2008b:2255–9.
15(4);2009:4665–73. Barbarino JM, Haidar CE, Klein TE, and Altman
Babaei H, Zeinalian M, Emami MH, RB. PharmGKB summary: Very important
Hashemzadeh M, Farahani N, and Salehi R. pharmacogene information for UGT1A1.
Simplified microsatellite instability detec- Pharmacogenet Genomics 24(3);2014:177–83.
tion protocol provides equivalent sen- Bardhan K, and Liu K. Epigenetics and colorec-
sitivity to robust detection strategies in tal cancer pathogenesis. Cancers (Basel)
Lynch syndrome patients. Cancer Biol Med 5(2);2013:676–713.
14(2);2017:142–50. Barh D, Azevedo V. Omics Technologies and Bio-
Bai Y, Wu HW, Ma X, Liu Y, and Zhang YH. Engineering, Volume 1: Towards Improving
Relationship between UGT1A1*6/*28 gene Quality of Life. London, Elsevier, 2017.
polymorphisms and the efficacy and toxic- Barski A, Cuddapah S, Cui K et al. High-
ity of irinotecan-based chemotherapy. Onco resolution profiling of histone methylations in
Targets Ther 10;2017:3071–81. the human genome. Cell 129(4);2007:823–37.
Balc’h EL, Grandin N, Demattei MV et al. Bashir B, and Snook AE. Immunotherapy regi-
Measurement of telomere length in colorectal mens for metastatic colorectal carcinomas.
cancers for improved molecular diagnosis. Int Hum Vaccin Immunother 2017:1–5.
J Mol Sci 18(9);2017, pii: E1871. Bazan V, Migliavacca M, Zanna I et al. Specific
Ballester V, Rashtak S, and Boardman L. Clinical codon 13 K-ras mutations are predictive of
and molecular features of young-onset clinical outcome in colorectal cancer patients,
colorectal cancer. World J Gastroenterol whereas codon 12 K-ras mutations are associ-
22(5);2016:1736–44. ated with mucinous histotype. Ann Oncol
Ballman KV. Biomarker: Predictive or prognostic? 13;2002:1438–46.
J Clin Oncol 33(33);2015:3968–71. Becquemont L. Pharmacogenomics of adverse
Balmanno K, Chell SD, Gillings AS, Hayat S, and drug reactions: Practical applications and
Cook SJ. Intrinsic resistance to the MEK1/2 ­perspectives. Pharmacogenomics 10(6);2009:96.
inhibitor AZD6244 (ARRY-142886) is associ- Bélanger A, Tojcic J, Harvey M, and
ated with weak ERK1/2 signalling and/or Guillemette C. Regulation of UGT1A1 and
strong PI3K signalling in colorectal cancer cell HNF1 ­transcription factor gene expression by
lines. Int J Cancer 125;2009:2332–41. DNA methylation in colon cancer cells. BMC
Bandrés E, Cubedo E, Agirre X et al. Mol Biol 11;2010:9.
Identification by Real-time PCR of 13 mature Bennouna J, Lang I, Valladares-Ayerbes M et al.
microRNAs differentially expressed in A phase II, open-label, randomised study to
colorectal cancer and non-tumoral tissues. assess the efficacy and safety of the MEK1/2
Mol Cancer 5;2006:29. inhibitor AZD6244 (ARRY-142886) versus
Banerjee J, Mishra N, and Dhas Y. capecitabine monotherapy in patients with
Metagenomics: A new horizon in cancer colorectal cancer who have failed one or two
research. Meta Gene 5;2015:84–9. prior chemotherapeutic regimens. Invest New
Baothman OA, Zamzami MA, Taher I, Abubaker J, Drugs 29;2011:1021–8.
and Abu-Farha M. The role of gut microbiota Berger FG, and Berger SH. Thymidylate synthase
in the development of obesity and diabetes. as a chemotherapeutic drug target: Where
Lipids Health Dis 15;2016:108. are we after fifty years? Cancer Biol Ther
Barault L, Charon-Barra C, Jooste V et al. 5(9);2006:1238–41.
Hypermethylator phenotype in sporadic colon Berquin IM, Cao L, Fong D, and Sloane BF.
cancer: Study on a population-based series of Identification of two new exons and mul-
582 cases. Cancer Res 68(20);2008a:8541–6. tiple transcription start points in the
Barault L, Veyries N, Jooste V et al. Mutations in 5′-untranslated region of the human
the RAS-MAPK, PI(3)K (phosphatidylinositol- ­cathepsin-B-encoding gene. Gene
3-OH kinase) signaling network correlate 159(2);1995:143–9.
90  Precision medicine for colorectal cancer

Bertholee D, Maring JG, and van Kuilenburg AB. Boisdron-Celle M, Remaud G, Traore S,

Genotypes affecting the pharmacokinet- Poirier AL, Gamelin L, Morel A, and Gamelin
ics of anticancer drugs. Clin Pharmacokinet E. 5-Fluorouracil-related severe toxicity: A
56(4);2017:317–37. comparison of different methods for the
Bertino JR, and Banerjee D. Thymidylate synthase pretherapeutic detection of dihydropyrimi-
as an oncogene? Cancer Cell 5(4);2004:301–2. dine dehydrogenase deficiency. Cancer Lett
Bertorelle R, Briarava M, Rampazzo E et al. 249;2007:271–82.
Telomerase is an independent prognostic Bokemeyer C, Bondarenko I, Hartmann JT et al.
marker of overall survival in patients with Efficacy according to biomarker status of
colorectal cancer. Br J Cancer 108(2);2013: cetuximab plus FOLFOX-4 as first-line treat-
278–84. ment for metastatic colorectal cancer: The
Beutler E, Gelbart T, and Demina A. Racial vari- OPUS study. Ann Oncol 22(7);2011:1535–46.
ability in the UDP-glucuronosyltransferase Bokemeyer C, Bondarenko I, Makhson A et al.
1 (UGT1A1) promoter: A balanced polymor- Fluorouracil, leucovorin, and oxaliplatin with
phism for regulation of bilirubin metabolism? and without cetuximab in the first-line treat-
Proc Natl Acad Sci USA 95(14);1998:8170–4. ment of metastatic colorectal cancer. J Clin
Bhandari A, Woodhouse M, and Gupta S. Oncol 27(5);2009:663–71.
Colorectal cancer is a leading cause of cancer Bokemeyer C, Van Cutsem E, Rougier P et al.
incidence and mortality among adults younger Addition of cetuximab to chemotherapy as
than 50 years in the USA: A SEER-based first-line treatment for KRAS wild-type meta-
analysis with comparison to other young-onset static colorectal cancer: Pooled analysis of the
cancers. J Investig Med 65(2);2017:311–5. CRYSTAL and OPUS randomised clinical trials.
Bhati A, Garg H, Gupta A, Chhabra H, Kumari A, Eur J Cancer 48(10);2012:1466–75.
and Patel T. Omics of cancer. Asian Pac J Boland CR, and Goel A. Microsatellite instabil-
Cancer Prev 13(9);2012:4229–33. ity in colorectal cancer. Gastroenterology
Bhurgri H, and Samiullah S. Colon cancer screen- 138(6);2010:2073–2087.e3.
ing—Is it time yet? J Coll Physicians Surg Pak Boland CR, Thibodeau SN, Hamilton SR et al.
27(6);2017:327–8. A National Cancer Institute Workshop on
Bian B, Mongrain S, Cagnol S et al. Cathepsin Microsatellite Instability for cancer detection
B promotes colorectal tumorigenesis, cell and familial predisposition: Development of
invasion, and metastasis. Mol Carcinog international criteria for the determination of
55(5);2016:671–87. microsatellite instability in colorectal cancer.
Biason P, Masier S, and Toffoli G. UGT1A1*28 Cancer Res 58(22);1998:5248–57.
and other UGT1A polymorphisms as deter- Boland PM, and Ma WW. Immunotherapy for
minants of irinotecan toxicity. J Chemother colorectal cancer. Cancers (Basel) 9(5);2017,
20(2);2008:158–65. pii: E50.
Bilgin B, Sendur MA, Bülent Akıncı M, Bolden JE, Peart MJ, and Johnstone RW.
Şener Dede D, and Yalçın B. Targeting the Anticancer activities of histone deacety-
PD-1 pathway: A new hope for gastrointestinal lase inhibitors. Nat Rev Drug Discov
­cancers. Curr Med Res Opin 33(4);2017:749–59. 5(9);2006:769–84.
Blanco-Gómez A, Castillo-Lluva S, Del Mar Bomba L, Walter K, and Soranzo N. The impact
Sáez-Freire M, Hontecillas-Prieto L, Mao of rare and low-frequency genetic vari-
JH, Castellanos-Martín A, and Pérez-Losada ants in common disease. Genome Biol
J. Missing heritability of complex dis- 18(1);2017:77.
eases: Enlightenment by genetic variants Bond CE, Nancarrow DJ, Wockner LF, Wallace
from intermediate phenotypes. BioEssays L, Montgomery GW, Leggett BA, and
38(7);2016:664–73. Whitehall VL. Microsatellite stable colorec-
Blons H, Emile JF, Le Malicot K et al. Prognostic tal cancers stratified by the BRAF V600E
value of KRAS mutations in stage III colon mutation show distinct patterns of chromo-
cancer: Post-hoc analysis of the PETACC8 somal instability. PLOS ONE 9(3);
phase III trial dataset. Ann Oncol 25(12);2014. 2014:e91739.
 References 91

Bosma PJ, Chowdhury JR, Bakker C et al. The Calvo E, Walko C, Dees EC, and Valenzuela B.
genetic basis of the reduced expression Pharmacogenomics, pharmacokinetics, and
of bilirubin UDP-glucuronosyltransferase pharmacodynamics in the era of targeted
1 in Gilbert’s syndrome. N Engl J Med therapies. Am Soc Clin Oncol Educ Book
333(18);1995:1171–5. 35;2016:e175–84.
Bourlioux P, Koletzko B, Guarner F, and Braesco Campbell JM, Stephenson MD, Bateman E, Peters
V. The intestine and its microflora are part- MD, Keefe DM, and Bowen JM. Irinotecan-
ners for the protection of the host: Report induced toxicity pharmacogenetics: An umbrella
on the Danone Symposium “The Intelligent review of systematic reviews and meta-analyses.
Intestine,” held in Paris, June 14, 2002. Am J Pharmacogenomics J 17(1);2017:21–8.
Clin Nutr 78(4);2003:675–83. Campos FG, Figueiredo MN, and Martinez CA.
Brandi J, Manfredi M, Speziali G, Gosetti F, Colorectal cancer risk in hamartomatous poly­
Marengo E, and Cecconi D. Proteomic posis syndromes. World J Gastrointest Surg
approaches to decipher cancer cell sec- 7(3);2015:25–32.
retome. Semin Cell Dev Biol 2017, pii: Capella G, Cronauer-Mitra S, Pienado MA, and
S1084-9521(17)30301-30304. doi: 10.1016/j. Perucho M. Frequency and spectrum of muta-
semcdb.2017.06.030. tions at codons 12 and 13 of the c-K-ras gene
Braun MS, Richman SD, Thompson L et al. in human tumors. Environ Health Perspect
Association of molecular markers with 93;1991:125–31.
toxicity outcomes in a randomized trial Capello M, Lee M, Wang H et al. Carboxylesterase
of chemotherapy for advanced colorec- 2 as a determinant of response to irinotecan
tal cancer: The FOCUS trial. J Clin Oncol and neoadjuvant FOLFIRINOX therapy in pan-
27(33);2009:5519–28. creatic ductal adenocarcinoma. J Natl Cancer
Braun MS, and Seymour MT. Balancing the Inst 107(8);2015, pii: djv132.
efficacy and toxicity of chemotherapy in Capitain O, Boisdron-Celle M, Poirier AL,
colorectal cancer. Ther Adv Med Oncol Abadie-Lacourtoisie S, Morel A, and
3(1);2011:43–52. Gamelin E. The influence of fluorouracil out-
Brosens RP, Belt EJ, Haan JC et al. Deletion of come parameters on tolerance and efficacy
chromosome 4q predicts outcome in stage II in patients with advanced colorectal cancer.
colon cancer patients. Anal Cell Pathol (Amst) Pharmacogenomics J 8(4);2008:256–67.
33(2);2010:95–104. Carethers JM, and Jung BH. Genetics and
Brown DG, Rao S, Weir TL, O’Malia J, Bazan M, genetic biomarkers in sporadic colorectal
Brown RJ, and Ryan EP. Metabolomics and ­cancer. Gastroenterology 149(5);2015:1177–90.
metabolic pathway networks from human Cargnello M, and Roux PP. Activation and func-
colorectal cancers, adjacent mucosa, and tion of the MAPKs and their substrates, the
stool. Cancer Metab 4;2016:11. MAPK-activated protein kinases. Microbiol
Brule SY, Jonker DJ, Karapetis CS et al. Location Mol Biol Rev 75(1);2011:50–83.
of colon cancer (right-sided versus left-sided) Carlini LE, Meropol NJ, Bever J et al. UGT1A7
as a prognostic factor and a predictor of and UGT1A9 polymorphisms predict response
benefit from cetuximab in NCIC CO.17. Eur J and toxicity in colorectal cancer patients
Cancer 51(11);2015:1405–14. treated with capecitabine/irinotecan. Clin
Buecher B, Cacheux W, Rouleau E, Cancer Res 11;2005:1226–36.
Dieumegard B, Mitry E, and Lièvre A. Role Carreras CW, and Santi DV. The catalytic mecha-
of microsatellite instability in the manage- nism and structure of thymidylate synthase.
ment of colorectal cancers. Dig Liver Dis Annu Rev Biochem 64;1995:721–62.
45(6);2013:441–9. Carvalho TI, Novais PC, Lizarte FSN et al.
Cacev T, Jokić M, Spaventi R, Pavelić K, and Analysis of gene expression EGFR and KRAS,
Kapitanović S. Loss of heterozygosity testing microRNA-21 and microRNA-203 in patients
using real-time PCR analysis of single nucleo- with colon and rectal cancer and correlation
tide polymorphisms. J Cancer Res Clin Oncol with clinical outcome and prognostic factors.
132(3);2006:200–4. Acta Cir Bras 32(3);2017:243–50.
92  Precision medicine for colorectal cancer

Castellarin M, Warren RL, Freeman JD et al. Charasson V, Haaz MC, and Robert J.
Fusobacterium nucleatum infection is preva- Determination of drug interactions occurring
lent in human colorectal carcinoma. Genome with the metabolic pathways of irinotecan.
Res 22(2);2012:299–306. Drug Metab Dispos 30(6);2002:731–3.
Catalano V, Loupakis F, Graziano F et al. Mucinous Chen H, Liu X, Brendler CB et al. Adding genetic
histology predicts for poor response rate and risk score to family history identifies twice
overall survival of patients with colorectal as many high-risk men for prostate cancer:
cancer and treated with first-line oxaliplatin- Results from the prostate cancer prevention
and/or irinotecan-based chemotherapy. Br J trial. Prostate 76(12);2016:1120–9.
Cancer 100(6);2009:881–7. Chen S, Villeneuve L, Jonker D et al. ABCC5 and
Cathomas G. PIK3CA in colorectal cancer. Front ABCG1 polymorphisms predict irinotecan-
Oncol 4;2014:35. induced severe toxicity in metastatic colorec-
Caudle KE, Thorn CF, Klein TE, Swen JJ, tal cancer patients. Pharmacogenet Genomics
McLeod HL, Diasio RB, and Schwab M. 25(12);2015:573–83.
Clinical Pharmacogenetics Implementation Chen S, Yueh MF, Bigo C et al. Intestinal gluc-
Consortium guidelines for dihydropyrimi- uronidation protects against chemotherapy-
dine dehydrogenase genotype and fluo- induced toxicity by irinotecan (CPT-11). Proc
ropyrimidine dosing. Clin Pharmacol Ther Natl Acad Sci USA 110(47);2013:19143–8.
94(6);2013:640–5. Cheng H, Zhang L, Cogdell DE et al. Circulating
Cecchin E, Corona G, Masier S et al. plasma MiR-141 is a novel biomarker for
Carboxylesterase isoform 2 mRNA expres- ­metastatic colon cancer and predicts
sion in peripheral blood mononuclear cells is poor prognosis. PLOS ONE 6(3);
a predictive marker of the irinotecan to SN38 2011:e17745.
activation step in colorectal cancer patients. Cheng L, Li M, Hu J et al. UGT1A1*6 polymor-
Clin Cancer Res 11(19 Pt 1);2005:6901–7. phisms are correlated with irinotecan-induced
Cecchin E, Innocenti F, D’Andrea M et al. toxicity: A system review and meta-analysis
Predictive role of the UGT1A1, UGT1A7, and in Asians. Cancer Chemother Pharmacol
UGT1A9 genetic variants and their haplotypes 73(3);2014:551–60.
on the outcome of metastatic colorectal Chin LJ, Ratner E, Leng S et al. A SNP in a
cancer patients treated with fluorouracil, let-7 microRNA complementary site in the
leucovorin, and irinotecan. J Clin Oncol KRAS 3′ untranslated region increases
27(15);2009:2457–65. non-small cell lung cancer risk. Cancer Res
Cecchin E, Roncato R, Guchelaar HJ, Toffoli G, 68(20);2008:8535–40.
for the Ubiquitous Pharmacogenomics Cho M, Carter J, Harari S, and Pei Z. The inter-
Consortium. Ubiquitous pharmacogenomics relationships of the gut microbiome and
(U-PGx): The time for implementation is now. ­inflammation in colorectal carcinogenesis.
An Horizon2020 program to drive pharma- Clin Lab Med 34(4);2014:699–710.
cogenomics into clinical practice. Curr Pharm Chong H, Vikis HG, and Guan KL. Mechanisms of
Biotechnol 18(3);2017:204–9. regulating the Raf kinase family. Cell Signal
Chalhoub N, and Baker SJ. PTEN and the PI3- 15;2003:463–9.
kinase pathway in cancer. Annu Rev Pathol Chua TC, and Morris DL. Resectable colorec-
4;2009:127–50. tal liver metastases: Optimal sequencing
Chang L, Chang M, Chang HM, and Chang F. of chemotherapy. J Gastrointest Cancer
Microsatellite instability: A predictive bio- 43(3);2012:496–8.
marker for cancer immunotherapy. Appl Coffman BL, Rios GR, King CD, and Tephly
Immunohistochem Mol Morphol 26(2);2017 TR. Human UGT2B7 catalyzes morphine
Sep 4:e15–21. ­glucuronidation. Drug Metab Dispos
Chang SC, Lin JK, Lin TC, and Liang WY. Loss 25;1997:1–4.
of heterozygosity: An independent prog- Collins K, and Mitchell JR. Telomerase
nostic factor of colorectal cancer. World J in the human organism. Oncogene
Gastroenterol 11(6);2005:778–84. 21(4);2002:564–79.
 References 93

Comalada M, Bailon E, de Haro O et al. The with metastatic gastrointestinal cancer

effects of short-chain fatty acids on colon treated with irinotecan-based regimens.
epithelial proliferation and survival depend Oncol Lett 12(5);2016:4231–7.
on the cellular phenotype. J Cancer Res Clin Cuinci C, Perini RF, Avadhani AN et al. Phase 1
Oncol 132(8);2006:487–97. and pharmacodynamic trial of everolimus in
Connell LC, Mota JM, Braghiroli MI, and Hoff PM. combination with cetuximab in patients with
The rising incidence of younger patients with advanced cancer. Cancer 120(1);2014:77–85.
colorectal cancer: Questions about screening, Cunningham D, Atkin W, Lenz HJ, Lynch HT,
biology, and treatment. Curr Treat Options Minsky B, Nordlinger B, and Starling N.
Oncol 18(4);2017:23. Colorectal cancer. Lancet 375;2010:1030–47.
Connolly K, Brungs D, Szeto E, and Epstein RJ. Côté JF, Kirzin S, Kramar A et al. UGT1A1 poly-
Anticancer activity of combination targeted morphism can predict hematologic toxicity in
therapy using cetuximab plus vemurafenib patients treated with irinotecan. Clin Cancer
for refractory BRAF (V600E)-mutant meta- Res 13(11);2007:3269–75.
static colorectal carcinoma. Curr Oncol Dai J, Gu J, Huang M, Eng C, Kopetz ES, Ellis
21(1);2014:e151–1. LM, Hawk E, and Wu X. GWAS-identified
Cooke Bailey JN, and Igo RP Jr. Genetic colorectal cancer susceptibility loci associ-
risk scores. Curr Protoc Hum Genet ated with clinical outcomes. Carcinogenesis
91;2016:1.29.1–9. 33(7);2012:1327–31.
Corcoran RB, Atreya CE, Falchook GS et al. Dall’Olio F, Malagolini N, Trinchera M, and
Combined BRAF and MEK inhibition with Chiricolo M. Mechanisms of cancer-asso-
dabrafenib and trametinib in BRAF V600– ciated glycosylation changes. Front Biosci
mutant colorectal cancer. J Clin Oncol 17;2012:670–99.
33(34);2015:4023–31. Daly AK. Genome-wide association stud-
Cortes-Ciriano I, Lee S, Park WY, Kim TM, and ies in pharmacogenomics. Nat Rev Genet
Park PJ. A molecular portrait of microsatel- 11(4);2010:241–6.
lite instability across multiple cancers. Nat Danese E, and Montagnana M. Epigenetics of
Commun 8;2017:15180. colorectal cancer: Emerging circulating diag-
Corté H, Manceau G, Blons H, and Laurent-Puig P. nostic and prognostic biomarkers. Ann Transl
MicroRNA and colorectal cancer. Dig Liver Dis Med 5(13);2017:279.
44(3);2012:195–200. Das C, and Tyler JK. Histone exchange and
Costa E. Hematologically important mutations: histone modifications during transcrip-
Bilirubin UDPglucuronosyltransferase gene tion and aging. Biochim Biophys Acta
mutations in Gilbert and Crigler-Najjar syn- 1819(3–4);2013:332–42.
dromes. Blood Cells Mol Dis 36;2006:77–80. Davie JR. Inhibition of histone deacety-
Court MH, and Greenblatt DJ. Molecular genetic lase activity by butyrate. J Nutr 133(7
basis for deficient acetaminophen glucuroni- Suppl.);2003:2485S–93S.
dation by cats: UGT1A6 is a pseudogene, and Day FL, Jorissen RN, Lipton L et al. PIK3CA
evidence for reduced diversity of expressed and PTEN gene and exon mutation-specific
hepatic UGT1A isoforms. Pharmacogenetics clinicopathologic and molecular associa-
10(4);2000:355–69. tions in colorectal cancer. Clin Cancer Res
Crona DJ, Ramirez J, Qiao W et al. Clinical 19(12);2013:3285–96.
­validity of new genetic biomarkers of de Jong FA, van der Bol JM, Mathijssen RH, van
­irinotecan neutropenia: An independent Gelder T, Wiemer EA, Sparreboom A, and
replication study. Pharmacogenomics J Verweij J. Renal function as a predictor of iri-
16(1);2016:54–9. notecan-induced neutropenia. Clin Pharmacol
Crunkhorn S. mTOR inhibition curbs colorectal Ther 84;2008a:254–62.
cancer. Nat Rev Drug Discov 14;2015:14–5. de Jong FA, Sparreboom A, Verweij J, and
Cui C, Shu C, Cao D et al. UGT1A1*6, UGT1A7*3 Mathijssen RH. Lifestyle habits as a contribu-
and UGT1A9*1b polymorphisms are predic- tor to anti-cancer treatment failure. Eur J
tive markers for severe toxicity in patients Cancer 44;2008b:374–82.
94  Precision medicine for colorectal cancer

De La Vega FM, Bustamante CD, and Leal SM. R. Dihydropyrimidine dehydrogenase

Genome-wide association mapping and rare ­polymorphisms and fluoropyrimidine toxic-
alleles: From population genomics to person- ity: Ready for routine clinical application
alized medicine—Session introduction. Pac within personalized medicine? EPMA J
Symp Biocomput 2011:74–5. 1(3);2010:495–502.
De Lange T. How telomeres solve the end-pro- Deng G, Kakar S, Tanaka H, Matsuzaki K, Miura S,
tection problem. Science 326;2009:948–52. Sleisenger MH, and Kim YS. Proximal and
De Las Rivas J, and de Luis A. Interactome data distal colorectal cancers show distinct gene-
and databases: Different types of protein inter- specific methylation profiles and clinical
action. Comp Funct Genomics 5;2004:173–8. and molecular characteristics. Eur J Cancer
De Mattia E, and Toffoli G. C677 T and A1298C 44;2008:1290–301.
MTHFR polymorphisms, a challenge for Derks S, and Diosdado B. Personalized cancer
antifolate and fluoropyrimidine-based medicine: Next steps in the genomic era. Cell
therapy personalisation. Eur J Cancer Oncol (Dordr) 38(1);2015:1–2.
45(8);2009:1333–51. Derwinger K, Wettergren Y, Odin E, Carlsson
den Besten G, van Eunen K, Groen AK, G, and Gustavsson B. A study of the
Venema K, Reijingoud DJ, and Bakker BM. MTHFR gene polymorphism C677T in colorec-
The role of short-chain fatty acids in the tal cancer. Clin Colorectal Cancer. 8(1);2009:
interplay between diet, gut microbiota, 43–8.
and host energy metabolism. J Lipid Res Des Guetz G, Schischmanoff O, Nicolas P,
54;2013:2325–40. Perret GY, Morere JF, and Uzzan B. Does mic-
De Palma M, and Hanahan D. The biology of per- rosatellite instability predict the efficacy of
sonalized cancer medicine: Facing individual adjuvant chemotherapy in colorectal cancer?
complexities underlying hallmark capabilities. A systematic review with meta-analysis. Eur J
Mol Oncol 6(2);2012:111–27. Cancer 45(10);2009:1890–6.
De Rook W, Claes B, Bernasconi D et al. Effects Devaud N, and Gallinger S. Chemotherapy of
of KRAS, BRAF, NRAS, and PIK3CA muta- MMR-deficient colorectal cancer. Fam Cancer
tions on the efficacy of cetuximab plus 12(2);2013:301–6.
­chemotherapy in chemotherapy-refractory Devilee P, Cleton-Jansen AM, and Cornelisse CJ.
metastatic colorectal cancer: A retrospec- Ever since Knudson. Trends Genet
tive consortium analysis. Lancet Oncol 17(10);2001:569–73.
11(8);2010:753–62. de Wildt SN, Kearns GL, Leeder JS, and van
Deenen MJ, Cats A, Beijnen JH, and Schellens den Anker JN. Glucuronidation in humans.
JH. Part 4: Pharmacogenetic variability in Pharmacogenetic and developmental
anticancer pharmacodynamic drug effects. aspects. Clin Pharmacokinet 36;1999:439–52.
Oncologist 16(7);2011a:1006–20. D’Haene N, Le Mercier M, De N et al. Clinical
Deenen MJ, Meulendijks D, Cats A et al. Upfront validation of targeted next generation
genotyping of DPYD⁄2A to individualize sequencing for colon and lung cancers. PLOS
fluoropyrimidine therapy: A safety and cost ONE 10(9);2015:e0138245.
analysis. J Clin Oncol 34(3);2016:227–3. Di Martino MT, Arbitrio M, Leone E et al. Single
Deenen MJ, Tol J, Burylo AM et al. Relationship nucleotide polymorphisms of ABCC5 and
between single nucleotide polymor- ABCG1 transporter genes correlate to
phisms and haplotypes in DPYD and irinotecan-associated gastrointestinal toxic-
toxicity and ­efficacy of capecitabine in ity in colorectal cancer patients: A DMET
advanced colorectal cancer. Clin Cancer Res microarray profiling study. Cancer Biol Ther
17(10);2011b:3455–68. 12(9);2011:780–7.
Deininger P. Genetic instability in cancer: Di Nicolantanio F, Martini M, Molinari F et al.
Caretaker and gatekeeper genes. Ochsner J Wild-type BRAF is required for response
1(4);1999:206–9. to panitumumab or cetuximab in meta-
Del Re M, Di Paolo A, van Schaik RH, static colorectal cancer. J Clin Oncol
Bocci G, Simi P, Falcone A, and Danesi 26(35);2008:5705–12.
 References 95

Dias-Santagata D, Akhavanfard S, David SS et al. Dumez H, Guetens G, De Boeck G, Highley

Rapid targeted mutational analysis of human MS, de Bruijn EA, van Oosterom AT, and
tumours: A clinical platform to guide per- Maes RA. In vitro partition of irinotecan
sonalized cancer medicine. EMBO Mol Med (CPT-11) in human volunteer blood: The influ-
2;2010:146–58. ence of ­concentration, gender and smoking.
Dienstmann R, Vermeulen L, Guinney J, Kopetz S, Anticancer Drugs 16;2005:893–5.
Tejpar S, and Tabernero J. Consensus molecu- Ehrlich M. DNA hypomethylation in cancer cells.
lar subtypes and the evolution of precision Epigenomics 1(2);2009:239–59.
medicine in colorectal cancer. Nat Rev Cancer Eklöf V, Löfgren-Burström A, Zingmark C et al.
17(2);2017:79–92. Cancer-associated fecal microbial markers
Dix BR, Robbins P, Soong R, Jenner D, House AK, in colorectal cancer detection. Int J Cancer
and Iacopetta BJ. The common molecu- 141(12);2017:2528–36.
lar genetic alterations in Dukes’ B and C Emami AH, Sadighi S, Shirkoohi R, and
colorectal carcinomas are not short-term Mohagheghi MA. Prediction of response to
prognostic indicators of survival. Int J Cancer irinotecan and drug toxicity based on phar-
59(6);1994:747–51. macogenomics test: A prospective case study
Domingo E, Niessen RC, Oliveira C et al. BRAF- in advanced colorectal cancer. Asian Pac J
V600E is not involved in the colorectal Cancer Prev 18(10);2017:2803–7.
tumorigenesis of HNPCC in patients with Engelhardt M, Drullinsky P, Guillem J, and
functional MLH1 and MSH2 genes. Oncogene Moore MA. Telomerase and telomere length
24(24);2005:3995–8. in the development and progression of pre-
Donohoe DR, Garge N, Zhang X, Sun W, malignant lesions to colorectal cancer. Clin
O’Connell TM, Bunger MK, and Bultman Cancer Res 3(11);1997:1931–41.
SJ. The microbiome and butyrate regu- Eshghifar N, Farrokhi N, Naji T, and Zali M. Tumor
late energy metabolism and autophagy suppressor genes in familial adenomatous
in the mammalian colon. Cell Metab polyposis. Gastroenterol Hepatol Bed Bench
13(15);2011:517–26. 10(1);2017:3–13.
Doullard JY, Oliner KS, Siena S et al. Esteban-Jurado C, Garre P, Vila M et al. New
Panitumumab-FOLFOX4 treatment and genes emerging for colorectal cancer
RAS mutations in colorectal cancer. NEJM predisposition. World J Gastroenterol
369(11);2013:1023–34. 20(8);2014:1961–71.
Drake RR. Glycosylation and cancer: Moving Etienne-Grimaldi MC, Boyer JC, Beroud C et al.
glycomics to the forefront. Adv Cancer Res New advances in DPYD genotype and risk
126;2015:1–10. of severe toxicity under capecitabine. PLOS
Drew DA, Nishihara R, Lochhead P et al. A pro- ONE 12(5);2017:e0175998.
spective study of smoking and risk of synchro- Etienne-Grimaldi MC, Boyer JC, Thomas F et al.
nous colorectal cancers. Am J Gastroenterol UGT1A1 genotype and irinotecan therapy:
112(3);2017:493–501. General review and implementation in
Du L, Kim JJ, Shen J, Chen B, and Dai N. routine practice. Fundam Clin Pharmacol
KRAS and TP53 mutations in inflamma- 29(3);2015:219–37.
tory bowel disease-associated colorec- Etienne-Grimaldi MC, Francoual M, Formento JL,
tal cancer: A m
­ eta-analysis. Oncotarget and Milano G. Methylenetetrahydrofolate
8(13);2017:22175–86. reductase (MTHFR) variants and fluoroura-
Dudbridge F. Power and predictive accu- cil-based treatments in colorectal cancer.
racy of polygenic risk scores. PLoS Genet Pharmacogenomics 8(11);2007:1561–6.
9(3);2013:e1003348. Etienne-Grimaldi MC, Milano G, Maindrault-
Duffy MJ, and Crown J. A personalized approach Goebel F et al. Methylenetetrahydrofolate
to cancer treatment: How ­biomarkers can reductase (MTHFR) gene polymorphisms
help. Clin Chem 54(11);2008:1770–9. and FOLFOX response in colorectal
Dukes CE. The classification of cancer of the cancer patients. Br J Clin Pharmacol
­rectum. J Pathol Bactériol 35;1932:323–32. 69;2010:58–66.
96  Precision medicine for colorectal cancer

Evaluation of Genomic Applications in Practice Ferraldeschi R, and Newman WG.

and Prevention (EGAPP) Working Group. Pharmacogenetics and pharmacogenom-
Recommendations from the EGAPP Working ics: A clinical reality. Ann Clin Biochem 48(Pt
Group: Can UGT1A1 genotyping reduce 5);2011:410–7.
morbidity and mortality in patients with meta- Figueiredo JC, Hsu L, Hutter CM et al. Genome-
static colorectal cancer treated with irinote- wide diet-gene interaction analyses for
can? Genet Med 11(1);2009:15–20. risk of colorectal cancer. PLoS Genet
Fallik D, Borrini F, Boige V et al. Microsatellite 10(4);2014:e1004228.
instability is a predictive factor of the tumor Flanagan L, Schmid J, Ebert M et al.
response to irinotecan in patients with Fusobacterium nucleatum associates with
advanced colorectal cancer. Cancer Res stages of colorectal neoplasia development,
63(18);2003:5738–44. colorectal cancer and disease outcome. Eur J
Fariña-Sarasqueta A. The BRAF V600E muta- Clin Microbiol Infect Dis 33(8);2014:1381–90.
tion is an independent prognostic factor for Frampton M, and Houlston RS. Modeling the
survival in stage II and stage III colon cancer prevention of colorectal cancer from the
patients. Ann Oncol 21;2010:2396–402. combined impact of host and behavioral risk
Farshidfar F, Weljie AM, Kopciuk KA et al. factors. Genet Med 19(3);2017:314–21.
A validated metabolomic signature for Frampton MJ, Law P, Litchfield K et al.
colorectal cancer: Exploration of the clini- Implications of polygenic risk for person-
cal value of metabolomics. Br J Cancer alised colorectal cancer screening. Ann Oncol
115(7);2016:848–57. 27(3);2016:429–34.
Fearon ER, and Vogelstein B. A genetic model for Francisco LM, Salinas VH, Brown KE, Vanguri VK,
colorectal tumorigenesis. Cell 61;1990:759–67. Freeman GJ, Kuchroo VK, and Sharpe AH.
Fedewa SA, Flanders WD, Ward KC, Lin CC, PD-L1 regulates the development, mainte-
Jemal A, Goding Sauer A, Doubeni CA, and nance, and function of induced regulatory T
Goodman M. Racial and ethnic disparities cells. J Exp Med 206(13);2009:3015–29.
in interval colorectal cancer incidence: A Fransen K, Klintenas M, Osterstrom A,
population-based cohort study. Ann Intern Dimberg J, Monstein HJ, and Söderkvist P.
166;2017:857–66. Mutation analysis of the BRAF, ARAF and
Feinberg AP, and Vogelstein B. Hypomethylation RAF-1 genes in human colorectal adenocarci-
distinguishes genes of some human can- nomas. Carcinogenesis 25(4);2004:527–33.
cers from their normal counterparts. Nature Frazer KA, Ballinger DG, Cox DR et al. A second
301(5895);1983:89–92. generation human haplotype map of over 3.1
Fennelly C, and Amaravadi RK. Lysosomal million SNPs. Nature 449(7164);2007:851–61.
biology in cancer. Methods Mol Biol Frosst P, Blom HJ, Milos R et al. A candidate
1594;2017:293–308. genetic risk factor for vascular disease: A
Ferlay J, Soerjomataram I, Dikshit R et al. common mutation in methylenetetrahydrofo-
Cancer incidence and mortality world- late reductase. Nat Genet 10;1995:111–3.
wide: Sources, methods and major pat- Frueh FW, and Gurwitz D. From pharmacogenet-
terns in GLOBOCAN 2012. Int J Cancer ics to personalized medicine: A vital need for
136(5);2015:E359–86. educating health professionals and the com-
Fernández-Marcelo T, Sánchez-Pernaute A, munity. Pharmacogenomics 5(5);2004:571–9.
Pascua I, De Juan C, Head J, Torres-García AJ, Fujita K, Kubota Y, Ishida H, and Sasaki Y.
and Iniesta P. Clinical relevance of telomere Irinotecan, a key chemotherapeutic drug
status and telomerase activity in colorectal for metastatic colorectal cancer. World J
cancer. PLOS ONE 11(2);2016:e0149626. Gastroenterol 21(43);2015:12234–48.
Ferraldeschi R. Pharmacogenetics in colorec- Fujiyoshi K, Yamaguchi T, Kakuta M et al.
tal cancer. In: Pharmacogenetics: Making Predictive model for high-frequency mic-
Cancer Treatment Safer and More Effective. rosatellite instability in colorectal cancer
Neman WG (Ed.), Springer, New York, 2010, patients over 50 years of age. Cancer Med
pp. 61–86. 6(6);2017:1255–63.
 References 97

Funkhouser WK Jr, Lubin IM, Monzon FA, ­ iomarker differences between MSI-H
b
Zehnbauer BA, Evans JP, Ogino S, and and MSS colorectal cancers. J Clin Oncol
Nowak JA. Relevance, pathogenesis, 33(Suppl; abstr 3597);2015:3597.
and ­testing algorithm for mismatch repair- Gatalica Z, Vranic S, Xiu J, Swensen J, and Reddy
defective colorectal carcinomas: A report S. High microsatellite instability (MSI-H)
of the association for molecular pathology. colorectal carcinoma: A brief review of pre-
J Mol Diagn 14(2);2012:91–103. dictive biomarkers in the era of personalized
Gagne JF, Montminy V, Belanger P, Journault medicine. Fam Cancer 15(3);2016:405–12.
K, Gaucher G, and Guillemette C. Common Geng Y, Marshall JR, and King MR.
human UGT1A polymorphisms and the Glycomechanics of the metastatic cas-
altered metabolism of irinotecan active cade: Tumor cell-endothelial cell interac-
metabolite 7-ethyl-10-hydroxycamptothecin tions in the circulation. Ann Biomed Eng
(SN-38). Mol Pharmacol 62;2002:608–17. 40;2012:790–805.
Gagnon J, Bernard O, Villeneuve L, Têtu B, Gertler R, Rosenberg R, Stricker D, Friederichs
and Guillemette C. Irinotecan inactivation J, Hoos A, Werner M, Ulm K, Holzmann B,
is modulated by epigenetic silencing of Nekarda H, and Siewert J-R. Telomere length
UGT1A1 in colon cancer. Clin Cancer Res and human telomerase reverse transcriptase
12;2006:1850–8. expression as markers for progression and
Gallagher DJ, and Kemeny N. Metastatic prognosis of colorectal carcinoma. J Clin
­colorectal cancer: From improved survival to Oncol 22(10);2004:1807–14.
potential cure. Oncology 78;2010:237–48. Glimelius B, Garmo H, Berglund A, Fredriksson LA,
Galon J, Costes A, Sanchez-Cabo F et al. Berglund M, Kohnke H, Byström P, Sørbye H,
Type, density, and location of immune and Wadelius M. Prediction of irinotecan
cells within human colorectal tumors and 5-fluorouracil toxicity and response in
predict ­clinical outcome. Science patients with advanced colorectal cancer.
313(5795);2006:1960–4. Pharmacogenomics J 11;2011:61–71.
Gao J, Zhou J, Li Y, Lu M, Jia R, and Shen L. Goel A, and Boland CR. Epigenetics of
UGT1A1 6/28 polymorphisms could ­predict colorectal cancer. Gastroenterology
irinotecan-induced severe neutropenia 143(6);2012:1442–60.
not diarrhea in Chinese colorectal cancer Goey AK, Mooiman KD, Beijnen JH, Schellens JH,
patients. Med Oncol 30(3);2013:604. and Meijerman I. Relevance of in vitro
Gao Z, Guo B, Gao R, Zhu Q, and Qin H. and clinical data for predicting CYP3A4-
Microbiota disbiosis is associated mediated herb-drug interactions in can-
with colorectal cancer. Front Microbiol cer patients. Cancer Treat Rev 39(7);2013
6;2015:20. 8:773–83.
Garcia-Aranda C, de Juan C, Diaz-Lopez A et al. Gong J, Wang C, Lee PP, Chu P, and Fakih M.
Correlations of telomere length, telomerase Response to PD-1 blockade in microsatellite
activity, and telomeric-repeat binding factor stable metastatic colorectal cancer harboring
1 expression in colorectal carcinoma. Cancer a POLE mutation. J Natl Compr Canc Netw
106(3);2006:541–51. 15(2);2017:142–7.
Garcia-Closas M, Gunsoy NB, and Chatterjee Gong Q, Chan SJ, Bajkowski AS, Steiner DF,
N. Combined associations of genetic and and Frankfater A. Characterization of the
environmental risk factors: Implications for cathepsin B gene and multiple mRNAs in
prevention of breast cancer. J Natl Cancer Inst human tissues: Evidence for alternative splic-
106(11);2014. ing of cathepsin B pre-mRNA. DNA Cell Biol
Gatalica Z. Thymidylate synthase over-expres- 12;1993:299–309.
sion underlies the observed lack of 5-FU Gonsalves WI, Mahoney MR, Sargent DJ et al.
therapy benefit for MSI-H colorectal c­ ancers. Patient and tumor characteristics and BRAF
Ann Oncol 25(Suppl. 4);2014:iv167–209. and KRAS mutations in colon cancer, NCCTG/
Gatalica Z, Vijayvergia N, Vranic S, Xiu J, Reddy alliance N0147. J Natl Cancer Inst 106(7);2014,
S, Lynch HT, and El-Deiry WS. Therapeutic pii: dju106.
98  Precision medicine for colorectal cancer

González N, Prieto I, Del Puerto-Nevado L et al. Gulhati P, Zaytseva YY, Valentino JD et al.
2017 update on the relationship between dia- Sorafenib enhances the therapeutic efficacy
betes and colorectal cancer: Epidemiology, of rapamycin in colorectal cancers harboring
potential molecular mechanisms and oncogenic KRAS and PIK3CA. Carcinogenesis
therapeutic implications. Oncotarget 33;2012:1782–90.
8(11);2017:18456–85. Gupta E, Lestingi TM, Mick R, Ramirez J, Vokes EE,
Gosens MJ, Moerland E, Lemmens VP, Rutten HT, and Ratain MJ. Metabolic fate of irinotecan
Tan-Go I, and van den Brule AJ. Thymidylate in humans: Correlation of glucuronidation
synthase genotyping is more predictive for with diarrhea. Cancer Res 54(14);1994:3723–5.
therapy response than immunohistochemistry Gusella M, and Padrini R. G > C SNP of thymi-
in patients with colon cancer. Int J Cancer dylate synthase with respect to colorectal
123(8);2008:1941–9. cancer. Pharmacogenomics 8(8);2007:985–96.
Goyette P, Sumner JS, Milos R, Duncan AM, Guzińska-Ustymowicz K. MMP-9 and cathep-
Rosenblatt DS, Matthews RG, and Rozen sin B expression in tumor budding as an
R. Human methylenetetrahydrofolate indicator of a more aggressive phenotype
reductase: Isolation of cDNA, mapping of colorectal cancer (CRC). Anticancer Res
and mutation identification. Nat Genet 26;2006:1589–94.
7(2);1994:195–200. Haaz MC, Rivory L, Riché C, Vernillet L, and
Grady WM, and Carethers JM. Genomic Robert J. Metabolism of irinotecan (CPT-11)
and epigenetic instability in colorectal by human hepatic microsomes: Participation
cancer pathogenesis. Gastroenterology of cytochrome P-450 3A and drug interac-
135(4);2008:1079–99. tions. Cancer Res 58(3);1998:468–72.
Grice EA, and Segre JA. The human microbiome: Haddad J, Slika S, and Mahfouz R. Epidermal
Our second genome. Annu Rev Genomics growth factor receptor (EGFR) in the era of
Hum Genet 13;2012:151–70. precision medicine: The tale of a perfect
Grothey A, Van Cutsem E, Sobrero A, Siena S, example of targeted therapy. A review. Meta
Falcone A, and Ychou M. Regorafenib Gene (Elsevier) 11;2017:157–63.
monotherapy for previously treated meta- Hague A, Butt AJ, and Paraskeva C. The role of
static colorectal cancer (CORRECT): An butyrate in human colonic epithelial cells:
international, multicentre, randomised, An energy source or inducer of differentiation
placebo-controlled, phase 3 trial. Lancet and apoptosis? Proc Nutr Soc 55;1996:937–43.
381(9863);2013:303–12. Hahn MM, de Voer RM, Hoogerbrugge N,
Gu M, Nishihara R, Chen Y et al. Aspirin Ligtenberg MJ, Kuiper RP, and van Kessel
exerts high anti-cancer activity in PIK3CA- AG. The genetic heterogeneity of colorec-
mutant colon cancer cells. Oncotarget tal c­ ancer predisposition—Guidelines
8(50);2017:87379–89. for gene discovery. Cell Oncol (Dordr)
Guerrero S, Casanova I, Farré L, Mazo A, Capellà 39(6);2016:491–510.
G, and Mangues R. K-ras codon 12 muta- Halama A, Guerrouahen BS, Pasquier J, Diboun I,
tion induces higher level of resistance to Karoly ED, Suhre K, and Rafii A. Metabolic
­apoptosis and predisposition to anchorage- ­signatures differentiate ovarian from colon
independent growth than codon 13 mutation cancer cell lines. J Transl Med 13;2015:223.
or proto-oncogene overexpression. Cancer Half E, Bercovich D, and Rozen P. Familial
Res 60(23);2000:6750–6. adenomatous polyposis. Orphanet J Rare Dis
Guillemette C. Pharmacogenomics of human 4;2009:22.
UDP glucuronosyltransferase enzymes. Hamet P. Future needs in exploration of gene-
Pharmacogenomics J 3(3);2003:136–58. environment interactions. J Hypertens
Guillemette C, Lévesque É, and Rouleau M. 30;2012:1915–6.
Pharmacogenomics of human uridine Hammond WA, Swaika A, and Mody K.
diphospho-glucuronosyltransferases and Pharmacologic resistance in colorectal
clinical implications. Clin Pharmacol Ther cancer: A review. Ther Adv Med Oncol
96(3);2014:324–39. 8(1);2016:57–84.
 References 99

Han FF, Guo CL, Yu D, Zhu J, Gong LL, Li GR, and germ-line epimutation. N Engl J Med
Lv YL. Associations between UGT1A1*6 or 356(7);2007:697–705.
UGT1A1*6/*28 polymorphisms and irinote- Hizel C. Omics’ language metagenomics is in
can-induced neutropenia in Asian cancer wait for precision medicine as a new clinical
patients. Cancer Chemother Pharmacol frontier. Focus on Colorectal Cancer Precis
73(4);2014:779–88. Med J 2(1);2017:1–5.
Hanahan D, and Weinberg RA. Hallmarks Hizel C, Ferrara M, Cure H et al. Evaluation of the
of cancer: The next generation. Cell 5’ spliced form of human cathepsin B mRNA
144(5);2011:646–74. in colorectal mucosa and tumors. Oncol Rep
Hart IR, and Saini A. Biology of tumour metasta- 5(1);1998:31–4.
sis. Lancet 339;1992:1453–7. Hizel C, Maurizis JC, Rio P, Communal Y,
Hazama S, Mishima H, Tsunedomi R et al. Chassagne J, Favy D, Bignon YJ, and
UGT1A1*6, 1A7*3, and 1A9*22 genotypes Bernard-Gallon DJ. Isolation, purification
predict severe neutropenia in FOLFIRI- and quantification of BRCA1 protein from
treated metastatic colorectal cancer in two tumour cells by affinity perfusion chroma-
prospective studies in Japan. Cancer Sci tography. J Chromatogr B Biomed Sci Appl
104(12);2013:1662–9. 721(2);1999:163–70.
Health Quality Ontario. Fecal occult blood test Hizel C, Tremblay J, Bartlett G, Hamet P.
for colorectal cancer screening: An evidence- Introduction: Every individual is different and
based analysis. Onta Health Technol Assess precision medicine offers options for dis-
Ser 9(10);2009:1–40. ease control and treatment. In: Progress and
Heidelberger C, Chaudhuri N, Danneberg P Challenges in Precision Medicine. Verma M,
et al. Fluorinated pyrimidines, a new class and Barh D (Eds), Elsevier Press, San Diego,
of tumour-inhibitory compounds. Nature 2017, pp. 1–33.
179;1957:663–6. Holderfield M, Deuker MM, McCormick F, and
Henricks LM, Lunenburg CA, Meulendijks D McMahon M. Targeting RAF kinases for can-
et al. Translating DPYD genotype into cer therapy: BRAF-mutated melanoma and
DPD p ­ henotype: Using the DPYD gene beyond. Nat Rev Cancer 14(7);2014:455–67.
activity score. Pharmacogenomics Holmes EC. Immunology of tumor infiltrating
16(11);2015:1277–86. lymphocytes. Ann Surg 201(2);1985:158–63.
Heo M, Allison DB, and Fontaine KR. Overweight, Hong DS, Morris VK, Fu S et al. Phase 1B study of
obesity, and colorectal cancer screening: vemurafenib in combination with irinotecan
Disparity between men and women. BMC and cetuximab in patients with BRAF-mutated
Public Health 4;2004:53. advanced cancers and metastatic colorectal
Herszènyi L, Plebani M, Carraro P et al. The role cancer. J Clin Oncol 32;2014: 3516, 5s, (suppl;
of cysteine and serine proteases in colorectal abstr 3516).
carcinoma. Cancer 86(7);1999:1135–42. Horie N, Aiba H, Oguro K, Hojo H, and
Herszényi L, Barabás L, Hritz I, István G, and Takeishi K. Functional analysis and DNA
Tulassay Z. Impact of proteolytic enzymes polymorphism of the tandemly repeated
in colorectal cancer development and sequences in the 5′-terminal regulatory region
progression. World J Gastroenterol of the human gene for thymidylate synthase.
20(37);2014:13246–57. Cell Struct Funct 20(3);1995:191–7.
Hindorff LA, Sethupathy P, Junkins HA, Ramos Hoskins JM, Goldberg RM, Qu P, Ibrahim JG,
EM, Mehta JP, Collins FS, and Manolio TA. and McLeod HL. UGT1A1*28 genotype and
Potential etiologic and functional implications ­irinotecan-induced neutropenia: Dose mat-
of genome-wide association loci for human ters. J Natl Cancer Inst 99(17);2007:1290–5.
diseases and traits. Proc Natl Acad Sci USA Hu ZY, Yu Q, Pei Q, and Guo C. Dose-dependent
106(23);2009:9362–7. association between UGT1A1*28 genotype
Hitchins MP, Wong JJ, Suthers G, Suter CM, and irinotecan-induced neutropenia: Low
Martin DI, Hawkins NJ, and Ward RL. doses also increase risk. Clin Cancer Res
Inheritance of a cancer-associated MLH1 16(15);2010:3832–42.
100  Precision medicine for colorectal cancer

Huang Y, Shen XJ, Zou Q, Wang SP, Tang SM, 1A1 promoter in different ethnic groups.
and Zhang GZ. Biological functions of Pharmacogenetics 12;2002:725–33.
microRNAs: A review. J Physiol Biochem Innocenti F, Kroetz DL, Schuetz E et al.
67(1);2011:129–39. Comprehensive pharmacogenetic analysis of
Huang Z, Huang D, Ni S, Peng Z, Sheng W, and irinotecan neutropenia and pharmacokinetics.
Du X. Plasma microRNAs are promising novel J Clin Oncol 27(16);2009:2604–14.
biomarkers for early detection of colorectal Innocenti F, and Ratain MJ. Irinotecan treatment
cancer. Int J Cancer 127(1);2010:118–26. in cancer patients with UGT1A1 polymor-
Hudler P, Kocevar N, and Komel R. Proteomics phisms. Oncology (Williston Park) 17(5 Suppl.
approaches in biomarker discovery: New 5);2003:52–5.
perspectives in cancer diagnostics. Sci World Innocenti F, and Ratain MJ. “Irinogenetics” and
J 2014;2014:260348. UGT1A: From genotypes to haplotypes. Clin
Hughes LA, Melotte V, de Schrijver J et al. The Pharmacol Ther 75(6);2004:495–500.
CpG island methylator phenotype: What’s in a Innocenti F, and Ratain MJ. Pharmacogenetics
name? Cancer Res 73(19);2013:5858–68. of irinotecan: Clinical perspectives on the
Huth L, Jäkel J, and Dahl E. Molecular diagnostic utility of genotyping. Pharmacogenomics
applications in colorectal cancer. Microarrays 7(8);2006:1211–21.
3(3);2014:168–79. Innocenti F, Undevia SD, Iyer L et al. Genetic
Iacopetta B. Are there two sides to colorectal ­variants in the UDP-glucuronosyltransferase
cancer? Int J Cancer 101(5);2002:403–8. 1A1 gene predict the risk ofsevere neutrope-
Iacopetta B. TP53 mutation in colorectal cancer. nia of irinotecan. J Clin Oncol 22(8);2004:
Hum Mutat 21(3);2003:271–6. 1382–8.
Iacopetta B, Russo A, Bazan V et al. Functional Ishihama K, Yamakawa M, Semba S, Takeda H,
categories of TP53 mutation in colorectal can- Kawata S, Kimura S, and Kimura W.
cer: Results of an international collaborative Expression of HDAC1 and CBP/p300 in
study. Ann Oncol 17(5);2006:842–7. human colorectal carcinomas. J Clin Pathol
ICGC Data Access Compliance Office, and 60(11);2007:1205–10.
ICGC International Data Access Committee. Isobe M, Emanuel BS, Givol D, Oren M, and
Analysis of five years of controlled access and Croce CM. Localization of gene for human
data sharing compliance at the International p53 tumour antigen to band 17p13. Nature
Cancer Genome Consortium. Nat Genet 320(6057);1986:84–5.
48(3);2016:224–55. Issa IA, and Noureddine M. Colorectal can-
Ide T, Kitajima Y, Ohtaka K, Mitsuno M, cer screening: An updated review of the
Nakafusa Y, and Miyazaki K. Expression of available options. World J Gastroenterol
the hMLH1 gene is a possible predictor for 23(28);2017:5086–96.
the clinical response to 5-fluorouracil after a Issa JP. CpG island methylator phenotype in
surgical resection in colorectal cancer. Oncol ­cancer. Nat Rev Cancer 4(12);2004:988–93.
Rep 19;2008:1571–6. Issa JP. Colon cancer: It’s CIN or CIMP. Clin
Ikenoue T, Hikiba Y, Kanai F et al. Functional anal- Cancer Res 14;2008:5939–40.
ysis of mutations within the kinase ­activation Ivancich M, Schrank Z, Wojdyla L, Leviskas
segment of B-Raf in human c­ olorectal tumors. B, Kuckovic A, Sanjali A, and Puri N.
Cancer Res 63(23);2003:8132–7. Treating cancer by targeting telomeres and
Inamura K. Colorectal cancers: An update on ­telomerase. Antioxidants (Basel) 6(1);2017,
their molecular pathology. Cancers (Basel) pii: E15.
10(1);2018, pii: E26. Iwasaki M, Tanaka-Mizuno S, Kuchiba A et al.
Ingelman-Sundberg M. Pharmacogenetics: Inclusion of a genetic risk score into a
An opportunity for a safer and more effi- validated risk prediction model for colorec-
cient pharmacotherapy. J Intern Med tal cancer in Japanese men improves
250(3);2001:186–200. ­performance. Cancer Prev Res (Phila)
Innocenti F, Grimsley C, Das S et al. Haplotype 10(9);2017 Jul 20:535–41, doi: 10.1158/1940-
structure of the UDP-glucuronosyltransferase 6207.CAPR-17-0141 [Epub ahead of print].
 References 101

Iyer L, Das S, Janisch L et al. UGT1A1*28 poly- using 16S or shotgun metagenomics. Front
morphism as a determinant of irinotecan Microbiol 7;2016:459.
disposition and toxicity. Pharmacogenomics J Kajihara M, Takakura K, Kanai T et al. Dendritic
2(1);2002:43–7. cell-based cancer immunotherapy for
Jackson CS, Oman M, Patel AM, and Vega KJ. colorectal cancer. World J Gastroenterol
Health disparities in colorectal cancer among 22(17);2016:4275–86.
racial and ethnic minorities in the United States. Kalow W. Pharmacogenetics: Heredity and the
J Gastrointest Oncol 7(Suppl. 1);2016:S32–43. Response to Drugs. W.B. Saunders Co.,
Jäger R, Migliorini G, Henrion M et al. Capture Philadelphia, 1962.
Hi-C identifies the chromatin interactome Kamel HFM, and Al-Amodi HSAB. Exploitation
of colorectal cancer risk loci. Nat Commun of gene expression and cancer biomarkers in
6;2015:6178. paving the path to era of personalized medi-
Jaiswal AS, Balusu R, and Narayan S. Involvement cine. Genomics Proteomics Bioinformatics
of adenomatous polyposis coli in colorectal 15(4);2017:220–35.
tumorigenesis. Front Biosci 10;2005:1118–34. Kamoshida S, Matsuoka H, Ishikawa T, Maeda K,
Jakobsen A, Nielsen NJ, Gyldenkerne N, and Shimomura R, Inada K, and Tsutsumi Y.
Lindeberg J. Thymidylate synthase and Immunohistochemical evaluation of thy-
methylenetetrahydrofolate reductase gene midylate synthase (TS) and p16INK4a in
polymorphism in normal tissue as predic- advanced colorectal cancer: Implication
tors of fluorouracil sensitivity. J Clin Oncol of TS expression in 5-FU-based ­adjuvant
23;2005:1365–9. chemotherapy. Jpn J Clin Oncol
Jang E, and Chung DC. Hereditary colon cancer: 34(10);2004:594–601.
Lynch syndrome. Gut Liver 4(2);2010:151–60. Kannagi R, Izawa M, Koike T, Miyazaki K, and
Jelsig AM. Hamartomatous polyps—A clinical Kimura N. Carbohydrate-mediated cell adhe-
and molecular genetic study. Dan Med J sion in cancer metastasis and angiogenesis.
63(8);2016, pii: B5280. Cancer Sci 95;2004:377–84.
Jen J, Kim H, Piantadosi S et al. Allelic loss of Kasahara M, Takahashi Y, Nagata T et al.
chromosome 18q and prognosis in colorectal Thymidylate synthase expression
cancer. N Engl J Med 331(4);1994:213–21. ­correlates closely with E2F1 expres-
Jensen SA, Vainer B, Kruhoffer M, and sion in colon cancer. Clin Cancer Res
Sorensen JB. Microsatellite instability in 6(7);2000:2707–11.
colorectal cancer and association with thymi- Kasubuchi M, Hasegawa S, Hiramatsu T,
dylate synthase and dihydropyrimidine dehy- Ichimura A, and Kimura I. Dietary gut micro-
drogenase expression. BMC Cancer 9;2009:25. bial ­metabolites, short-chain fatty acids,
Jia X, Shanmugam C, Paluri RK et al. Prognostic and host metabolic regulation. Nutrients
value of loss of heterozygosity and sub- 7(4);2015:2839–49.
cellular localization of SMAD4 varies with Kawakami H, Zaanan A, and Sinicrope FA.
tumor stage in colorectal cancer. Oncotarget Microsatellite instability testing and its
8(12);2017:20198–212. role in the management of c­ olorectal
Jin B, Li Y, and Robertson KD. DNA methylation: ­cancer. Curr Treat Options Oncol
Superior or subordinate in the epigenetic 16(7);2015:30.
hierarchy? Genes Cancer 2(6);2011:607–17. Kawakami K, and Watanabe G. Identification and
Johnson SM, Grosshans H, Shingara J et al. RAS functional analysis of single nucleotide poly-
is regulated by the let-7 microRNA family. Cell morphism in the tandem repeat sequence
120(5);2005:635–47. of thymidylate synthase gene. Cancer Res
Jonsson PF, and Bates PA. Global topological 63(18);2003:6004–7.
features of cancer proteins in the human Kawato Y, Aonuma M, Hirota Y, Kuga H, and
­interactome. Bioinformatics 22(18);2007: Sato K. Intracellular roles of SN-38, a metabo-
2291–7. lite of the camptothecin derivative CPT-11, in
Jovel J, Patterson J, Wang W et al. the antitumor effect of CPT-11. Cancer Res
Characterization of the gut microbiome 51;1991:4187–91.
102  Precision medicine for colorectal cancer

Kefeli U, Ucuncu Kefeli A, Cabuk D et al. Kim TM, Lee SH, and Chung YJ. Clinical appli-
Netrin-1 in cancer: Potential biomarker cations of next-generation sequencing in
and therapeutic target? Tumour Biol colorectal cancers. World J Gastroenterol
39(4);2017:1010428317698388. 19(40);2013:6784–93.
Kefford R, Arkenau H, Brown MP et al. Phase I/ Kim YJ, and Varki A. Perspectives on the
II study of GSK2118436, a selective inhibi- ­significance of altered glycosylation of
tor of oncogenic mutant BRAF kinase, in glyco-proteins in cancer. Glycoconj J
patients with metastatic melanoma and other 14;1997:569–76.
solid tumours. J Clin Oncol 28(Suppl.);2010, Kittles R. Genes and environments: Moving
abstract 10523. toward personalized medicine in the
Kehrer DF, Mathijssen RH, Verweij J, de Bruijn P, context of health disparities. Ethn Dis
and Sparreboom A. Modulation of irinotecan 22;2012:S1-43–6.
metabolism by ketoconazole. J Clin Oncol Klose RJ, and Zhang Y. Regulation of his-
20;2002:3122–9. tone methylation by demethylimination
Khan A, Krishna M, Baker S, and Banner B. and demethylation. Nat Rev Mol Cell Biol
Cathepsin B and tumor-associated laminin 8(4);2007:307–18.
expression in the progression of colorec- Klümpen HJ, Beijnen JH, Gurney H, and
tal adenoma to carcinoma. Mod Pathol Schellens JH. Inhibitors of mTOR. Oncologist
11;2004:704–8. 15(12);2010:1262–9.
Khanna R, Morton CL, Danks MK, and Potter PM. Knudson AG. Mutation and cancer: Statistical
Proficient metabolism of irinotecan by a study of retinoblastoma. Proc Natl Acad Sci
human intestinal carboxylesterase. Cancer USA 68(4);1971:820–3.
Res 60;2000:4725–8. Ko LJ, and Prives C. p53: Puzzle and paradigm.
Kiang TK, Ensom MH, and Chang TK. UDP- Genes Dev 10(9);1996:1054–72.
glucuronosyl-transferases and clinical Koga Y, Yasunaga M, Takahashi A et al.
drug-drug interactions. Pharmacol Ther MicroRNA expression profiling of exfoliated
106;2005:97–132. colonocytes isolated from feces for colorec-
Kim B, Kang S, Jeong G, Park SB, and Kim SJ. tal cancer screening. Cancer Prev Res (Phila)
Identification and comparison of aberrant 3(11);2010:1435–42.
key regulatory networks in breast, colon, Koido S, Homma S, Takahara A et al.
liver, lung, and stomach cancers through Immunotherapy synergizes with che-
methylome database analysis. PLOS ONE motherapy targeting pancreatic cancer.
9(5);2014:e97818. Immunotherapy 4(1);2012:5–7.
Kim B-S, Jeon Y-S, and Chun J. Current status Kopetz S, Desai J, Chan E et al. Phase II pilot
and future promise of the human microbi- study of vemurafenib in patients with meta-
ome. Pediatr Gastroenterol Hepatol Nutr static BRAF-mutated colorectal cancer. J Clin
16;2013:71–7. Oncol 33(34);2015:4032–8.
Kim D, Kim SY, Lee JS et al. Primary tumor Krapohl E, Patel H, Newhouse S et al. Multi-
location predicts poor clinical outcome polygenic score approach to trait prediction.
with cetuximab in RAS wild-type meta- Mol Psychiatry 2017 Aug 8, doi: 10.1038/
static colorectal cancer. BMC Gastroenterol mp.2017.163 [Epub ahead of print].
17(1);2017:121. Kruszewski WJ, Rzepko R, Wojtacki J,
Kim DH, Sarbassov DD, Ali SM et al. mTOR Skokowski J, Kopacz A, Jaśkiewicz K, and
­interacts with raptor to form a nutrient Drucis K. Overexpression of cathepsin B cor-
sensitive complex that signals to cell growth relates with angiogenesis in colon adenocarci-
machinery. Cell 110;2002:163–75. noma. Neoplasma 51;2004:38–43.
Kim J, Cho YA, Kim DH et al. Dietary Kummar S, Gutierrez M, Doroshow JH, and
intake of folate and alcohol, MTHFR Murgo AJ. Drug development in oncol-
C677 T ­polymorphism, and colorec- ogy: Classical cytotoxics and molecularly
tal cancer risk in Korea. Am J Clin Nutr ­t argeted agents. Br J Clin Pharmacol
95(2);2012:405–12. 62(1);2006:15–26.
 References 103

Kweekel DM, Gelderblom H, Van der Straaten T, Interdisciplinary perspective, world tidal
Antonini NF, Punt CJ, Guchelaar HJ, Dutch wave and potential growth for the emerg-
Colorectal Cancer Group study. UGT1A1*28 ing countries. In: Progress and Challenges
genotype and irinotecan dosage in patients in Precision Medicine. Verma M, and Barh
with metastatic colorectal cancer: A Dutch D (Eds), Elsevier Press, San Diego, 2017, pp.
Colorectal Cancer Group study. Br J Cancer 301–14.
99;2008:275–82. Lemery S, Keegan P, and Pazdur R. First FDA
LaFave MC, and Sekelsky J. Mitotic recombina- approval agnostic of cancer site—When a
tion: Why? When? How? Where? PLoS Genet biomarker defines the indication. N Engl J
5(3);2009:e1000411. Med 377(15);2017:1409–12.
Läll K, Mägi R, Morris A, Metspalu A, and Lengauer C, Kinzler KW, and Vogelstein B.
Fischer K. Personalized risk prediction for Genetic instability in colorectal cancers.
type 2 diabetes: The potential of genetic risk Nature 386(6625);1997:623–7.
scores. Genet Med 201719(3);2017:322–9. Leroy B, Anderson M, and Soussi T. TP53 muta-
Lam K, Pan K, Linnekamp JF, Medema JP, and tions in human cancer: Database reassess-
Kandimalla R. DNA methylation based ment and prospects for the next decade.
biomarkers in colorectal cancer: A sys- Hum Mutat 35(6);2014:672–88.
tematic review. Biochim Biophys Acta Lettre G. Rare and low-frequency variants in
1866(1);2016:106–20. human common diseases and other complex
Lane DP. p53, guardian of the genome. Nature traits. J Med Genet 51(11);2014:705–14.
358;1992:15–6. Levine AJ. p53, the cellular gatekeeper for
Lankisch TO, Vogel A, Eilermann S et al. growth and division. Cell 88;1997:323–31.
Identification and characterization of a func- Ley RE, Peterson DA, and Gordon JI. Ecological
tional TATA box polymorphism of the UDP gluc- and evolutionary forces shaping micro-
uronosyltransferase 1A7 gene. Mol Pharmacol bial diversity in the human intestine. Cell
67(5);2005:1732–9. Epub 2005 Feb 16. 124;2006:837–48.
Laplante M, and Sabatini DM. mTor signal- Li J, Qin S, Xu R et al. Regorafenib plus best
ing in growth control and disease. Cell supportive care versus placebo plus best
149(2);2012:274–93. supportive care in Asian patients with previ-
Laurent-Puig P, Cayre A, Manceau G et al. ously treated metastatic colorectal cancer
Analysis of PTEN, BRAF, and EGFR status in (CONCUR): A randomised, double-blind,
determining benefit from cetuximab therapy placebo-controlled, phase 3 trial. The Lancet
in wild-type KRAS metastatic colon cancer. Oncol 16(6);2015:619–29.
J Clin Oncol 27(35);2009:5924–30. Li W, Zhang H, Assaraf YG et al. Overcoming
Lederberg J, and McCray AT. “Ome Sweet” ABC transporter-mediated multidrug resis-
Omics—A genealogical treasury of words. tance: Molecular mechanisms and novel
Scientist 15;2001:8. therapeutic drug strategies. Drug Resist
Lee JK, and Chan AT. Molecular prognostic Updat 27;2016:14–29.
and predictive markers in colorectal cancer: Li XL, Zhou J, Chen ZR, and Chng WJ.
Current status. Curr Colorectal Cancer Rep P53 mutations in colorectal cancer—
7(2);2011:136–44. Molecular pathogenesis and pharmaco-
Lee SY, Miyai K, Han HS et al. Microsatellite logical reactivation. World J Gastroenterol
­instability, EMAST, and morphol- 21(1);2015:84–93.
ogy ­associations with T cell infiltra- Liao X, Lochhead P, Nishihara R et al. Aspirin
tion in c­ olorectal neoplasia. Dig Dis Sci use, tumor PIK3CA mutation, and
57(1);2012:72–8. colorectal-cancer survival. N Engl J Med
Leggett B, and Whitehall V. Role of the ­serrated 367(17);2012:1596–606.
pathway in colorectal cancer pathogenesis. Lievre A, Bachet J, Le Corre D et al. KRAS muta-
Gastroenterology 138(6);2010:2088–100. tion status is predictive of response to cetux-
Lemay V, Hamet P, Hizel C, Lemaire E, imab therapy in colorectal cancer. Cancer Res
Tremblay Y. Personalized medicine: 66(8);2006:3992–5.
104  Precision medicine for colorectal cancer

Lima A, Azevedo R, Sousa H, Seabra V, and Loupakis F, Ruzzo A, Cremolini C et al. KRAS
Medeiros R. Current approaches for TYMS codon 61, 146 and BRAF mutations predict
polymorphisms and their importance in resistance to cetuximab plus irinotecan in KRAS
molecular epidemiology and pharmacogenet- codon 12 and 13 wild-type metastatic colorec-
ics. Pharmacogenomics 14(11);2013:1337–51. tal cancer. Br J Cancer 101(4);2009:715–21.
Lindpaintner K. The impact of pharmacogenetics Low SK, Takahashi A, Mushiroda T, and Kubo M.
and pharmacogenomics on drug discovery. Genome-wide association study: A useful
Nat Rev Drug Discov 1;2002:463–9. tool to identify common genetic variants
Link A, Balaguer F, Shen Y, Nagasaka T, associated with drug toxicity and efficacy in
Lozano JJ, Boland CR, and Goel A. Fecal cancer pharmacogenomics. Clin Cancer Res
MicroRNAs as novel biomarkers for colon can- 20(10);2014:2541–52.
cer screening. Cancer Epidemiol Biomarkers Lowe WL Jr, and Reddy TE. Genomic approaches
Prev 19(7);2010:1766–74. for understanding the genetics of complex
Liotta LA, Nageswara RC, and Wewer UM. disease. Genome Res. 25;2015:1432–41.
Biochemical interactions of tumor cells with Lungulescu CV, Răileanu S, Afrem G et al.
the basement membrane. Ann Rev Biochem Histochemical and immunohistochemical
55;1986:1037–57. study of mucinous rectal carcinoma. J Med
Liu D, Li J, Gao J, Li Y, Yang R, and Shen L. Life 10(2);2017:139–43.
Examination of multiple UGT1A and DPYD Lynch D, and Murphy A. The emerging role of
polymorphisms has limited ability to pre- immunotherapy in colorectal cancer. Ann
dict the toxicity and efficacy of metastatic Transl Med 4(16);2016:305.
colorectal cancer treated with irinotecan- Lynch HT, and Shaw TG. Practical genetics of
based chemotherapy: A retrospective analy- colorectal cancer. Chin Clin Oncol 2(2);2013:12.
sis. BMC Cancer 17(1);2017:437. Lévesque E, Bélanger AS, Harvey M, Couture F,
Liu DP, Song H, and Xu Y. A common gain Jonker D, and Innocenti F. Refining the
of function of p53 cancer mutants in UGT1A haplotype associated with irinotecan-
inducing genetic instability. Oncogene induced hematological toxicity in metastatic
29(7);2010:949–56. colorectal cancer patients treated with 5
Liu J, Liu F, Li X, Song X, Zhou L, and fluorouracil/irinotecan-based regimens.
Jie J. Screening key genes and miR- J Pharmacol Exp Ther 345(1);2013:95–101.
NAs in early-stage colon adenocarci- Maccaferri S, Biagi E, and Brigidi P.
noma by RNA-sequencing. Tumour Biol Metagenomics: Key to human gut microbiota.
39(7);2017:1010428317714899. Dig Dis 29;2011:525–30.
Liu J, Zhang C, and Feng Z. Tumor suppres- Mahasneh A, Al-Shaheri F, and Jamal E.
sor p53 and its gain-of-function mutants in Molecular biomarkers for an early diagnosis,
cancer. Acta Biochim Biophys Sin (Shanghai) effective treatment and prognosis of colorec-
46(3);2014:170–9. tal cancer. Current updates. Exp Mol Pathol
Liu XH, Lu J, Duan W et al. Predictive value of 102(3);2017:475–83.
UGT1A1*28 polymorphism in irinotecan-based Maher BS. Polygenic scores in epidemiology: Risk
chemotherapy. J Cancer 8(4);2017:691–703. prediction, etiology, and clinical utility. Curr
Lochhead P, Kuchiba A, Imamura Y et al. Epidemiol Rep 2(4);2015:239–44.
Microsatellite instability and BRAF mutation Makondi PT, Chu CM, Wei PL, and Chang YJ.
testing in colorectal cancer prognostication. Prediction of novel target genes and ­pathways
J Natl Cancer Inst 105(15);2013:1151–6. involved in irinotecan-resistant colorectal can-
Longley DB, Harkin DP, and Johnston PG. 5-fluo- cer. PLOS ONE 12(7);2017:e0180616.
rouracil: Mechanisms of action and clinical Malkin D. Li-fraumeni syndrome. Genes Cancer
strategies. Nat Rev Cancer 3(5);2003:330–8. 2(4);2011:475–84.
Loree JM, and Kopetz S. Recent developments in Malumbres M, and Barbacid M. RAS onco-
the treatment of metastatic colorectal cancer. genes: The first 30 years. Nat Rev Cancer
Ther Adv Med Oncol 9(8);2017:551–64. 3(6);2003:459–65.
 References 105

Mandola MV, Stoehlmacher J, Muller-Weeks Marsh S, Ameyaw MM, Githang’a J, Indalo A,

S, Cesarone G, Yu MC, Lenz HJ, and Ofori-Adjei D, and McLeod HL. Novel
Ladner RD. A novel single nucleotide thymidylate synthase enhancer region
polymorphism within the 5′ tandem repeat alleles in African populations. Hum Mutat
polymorphism of the thymidylate syn- 16;2000:528.
thase gene abolishes USF-1 binding and Marsh S, Collie-Duguid ES, Li T, Liu X, and
alters transcriptional activity. Cancer Res McLeod HL. Ethnic variation in the thy-
63;2003:2898–904. midylate synthase enhancer region poly-
Manning BD, and Cantley LC. AKT/PKB signaling: morphism among Caucasian and Asian
Navigating downstream. Cell 129(7);2007: populations. Genomics 58;1999:310–2.
1261–74. Marsh S, McKay JA, Cassidy J, and McLeod HL.
Manson LE, van der Wouden CH, Swen JJ, Polymorphism in the thymidylate synthase
and Guchelaar HJ. The Ubiquitous promoter enhancer region in colorectal
Pharmacogenomics consortium: Making ­cancer. Int J Oncol 19(2);2001:383–6.
effective treatment optimization accessible to Marsh S, and McLeod HL. Thymidylate synthase
every European citizen. Pharmacogenomics pharmacogenetics in colorectal cancer. Clin
18(11);2017:1041–5. Colorectal Cancer 1(3);2001:175–8.
Manzat-Saplacan RM, Mircea PA, Balacescu Massacesi C, Terrazzino S, Marcucci F et al.
L, Chira RI, Berindan-Neagoe I, and Uridine diphosphate glucuronosyl transferase
Balacescu O. Can we change our microbiome 1A1 promoter polymorphism predicts the
to prevent colorectal cancer development? risk of gastrointestinal toxicity and fatigue
Acta Oncol 54(8);2015:1085–95. induced by irinotecan-based chemotherapy.
Mao C, Yang ZY, Hu XF, Chen Q, and Tang Cancer 106(5);2006:1007–16.
JL. PIK3CA exon 20 mutations as a poten- Mathijssen RH, van Alphen RJ, Verweij
tial biomarker for resistance to anti-EGFR J, Loos WJ, Nooter K, Stoter G, and
monoclonal antibodies in KRAS wild-type Sparreboom A. Clinical pharmacokinetics and
metastatic colorectal cancer: A system- metabolism of irinotecan (CPT-11). Clin Cancer
atic review and meta-analysis. Ann Oncol Res 7(8);2001:2182–94.
23(6);2012:1518–25. Mathivanan S, Periaswamy B, Gandhi TK et al.
Mao M, Tian F, Mariadason JM et al. Resistance An evaluation of human protein-protein
to BRAF inhibition in BRAF-mutant colon interaction data in the public domain. BMC
cancer can be overcome with PI3K inhibition Bioinformatics 7(Suppl 5);2006:S19.
or demethylating agents. Clin Cancer Res Maughan TS, Adams RA, Smith CG et al.
19(3);2012:657–67. Addition of cetuximab to oxaliplatin-based
Marcuello E, Altés A, Menoyo A, Del Rio E, first-line combination chemotherapy for
Gómez-Pardo M, and Baiget M. UGT1A1 gene ­treatment of advanced colorectal cancer:
variations and irinotecan treatment in patients Results of the randomised phase 3 MRC
with metastatic colorectal cancer. Br J Cancer COIN trial. Lancet 377(9783);2011:2103–14.
91(4);2004:678–82. Mayrhofer M, Kultima HG, Birgisson H et al. 1p36
Marcuello E, Páez D, Paré L, Salazar J, Sebio deletion is a marker for tumour d ­ issemination
A, del Rio E, and Baiget M. A genotype- in microsatellite stable stage II-III colon can-
directed phase I-IV dose-finding study of cer. BMC Cancer 14;2014:872.
irinotecan in combination with fluorouracil/ Mazelin L, Bernet A, Bonod-Bidaud C et al.
leucovorin as first-line treatment in advanced Netrin-1 controls colorectal tumorigen-
colorectal cancer. Br J Cancer 105(1);2011 esis by regulating apoptosis. Nature
Jun 28:53–7. 431(7004);2004:80–4.
Marques SC, and Ikediobi ON. The clinical McBride OW, Merry D, and Givol D. The gene for
application of UGT1A1 pharmacogenetic human p53 cellular tumor antigen is located
testing: Gene-environment interactions. Hum on chromosome 17 short arm (17p13). Proc
Genomics 4(4);2010:238–49. Natl Acad Sci USA 83(1);1986:130–4.
106  Precision medicine for colorectal cancer

McDermott DF, and Atkins MB. PD-1 as a poten- liver disease and colorectal cancer. Postgrad
tial target in cancer therapy. Cancer Med Med J 93(1097);2017:153–8.
2(5);2013:662–73. Milos PM, and Seymour AB. Emerging strategies
McMurray HR, Sampson ER, Compitello G et al. and applications of pharmacogenomics. Hum
Synergistic response to oncogenic mutations Genomics 1;2004:444–55.
defines gene class critical to cancer pheno- Mishra N, and Hall J. Identification of patients
type. Nature 453(7198);2008:1112. at risk for hereditary colorectal cancer. Clin
McRee AJ, Davies JM, Sanoff HG et al. A phase Colon Rectal Surg 25(2);2012:67–82.
I trial of everolimus in combination with Moerkerk P, Arends JW, Van Driel M, de
5-FU/LV, mFOLFOX6 and mFOLFOX6 plus Bruïne A, de Goeij A, and ten Kate J. Type
panitumumab in patients with refractory and number of Ki-ras point mutations relate
solid tumors. Cancer Chemother Pharmacol to stage of human colorectal cancer. Cancer
74(1);2014:117–23. Res 54(13);1994:3376–8.
Meguid RA, Slidell MB, Wolfgang CL, Chang DC, Moertel C, Fleming T, Macdonald J et al.
and Ahuja N. Is there a difference in survival Levamisole and fluorouracil for adjuvant
between right- versus left-sided colon can- therapy of resected colon carcinoma. N Engl
cers? Ann Surg Oncol 15(9);2008:2388–94. J Med 322;1990:352–8.
Mei ZB, Duan CY, Li CB, Cui L, and Ogino S. Moraes F, and Góes A. A decade of human
Prognostic role of tumor PIK3CA muta- genome project conclusion: Scientific diffu-
tion in colorectal cancer: A systematic sion about our genome knowledge. Biochem
review and meta-analysis. Ann Oncol Mol Biol Educ 44(3);2016:215–23.
27(10);2016:1836–48. Mori R, Futamura M, Tanahashi T, Tanaka Y,
Merino D, and Malkin D. p53 and hereditary can- Matsuhashi N, Yamaguchi K, and Yoshida K.
cer. Subcell Biochem 85;2014:1–16. 5FU resistance caused by reduced fluoro-
Metzker ML. Sequencing technologies: The next deoxyuridine monophosphate and its
generation. Nat Rev Genet 11(1);2009:31–46. reversal using deoxyuridine. Oncol Lett
Meulendijks D, Cats A, Beijnen JH, and Schellens 14(3);2017:3162–8.
JH. Improving safety of fluoropyrimidine Moyzis RK, Buckingham JM, Cram LS, Dani M,
chemotherapy by individualizing treatment Deaven LL, Jones MD, Meyne J, Ratliff RL,
based on dihydropyrimidine dehydrogenase and Wu JR. A highly conserved repetitive
activity—Ready for clinical practice? Cancer DNA sequence, (TTAGGG)n, present at the
Treat Rev 50;2016a:23–34. telomeres of human chromosomes. Proc Natl
Meulendijks D, Henricks LM, van Kuilenburg AB Acad Sci USA 85(18);1988:6622–6.
et al. Patients homozygous for DPYD c.1129- Muller PA, and Vousden KH. p53 mutations in
5923C > G/haplotype B3 have partial DPD cancer. Nat Cell Biol 15(1);2013:2–8.
deficiency and require a dose reduction Murnane MJ, Sheahan K, Ozdemirli M, and
when treated with fluoropyrimidines. Cancer Shuja S. Stage-specific increases in cathepsin
Chemother Pharmacol 78(4);2016b:875–80. B messenger RNA content in human colorec-
Michael MZ, O’Connor SM, van Holst Pellekaan tal carcinoma. Cancer Res 51;1991:1137–42.
NG, Young GP, and James RJ. Reduced Mzahma R, Kharrat M, Fetiriche F et al. The
­accumulation of specific microRNAs in relationship between telomere length and
colorectal neoplasia. Mol Cancer Res clinicopathologic characteristics in colorectal
1(12);2003:882–91. cancers among Tunisian patients. Tumour Biol
Michael-Robinson JM, Biemer-Hüttmann A, 36(11);2015:8703–13.
Purdie DM et al. Tumour infiltrating lym- Nakajima G, Hayashi K, Xi Y et al. Non-coding
phocytes and apoptosis are independent MicroRNAs hsa-let-7 g and hsa-miR-181b are
features in colorectal cancer stratified accord- associated with chemoresponse to S-1 in
ing to microsatellite instability status. Gut colon cancer. Cancer Genomics Proteomics
48(3);2001:360–6. 3(5);2006:317–24.
Mikolasevic I, Orlic L, Stimac D, Hrstic I, Namasivayam V, and Lim S. Recent advances in
Jakopcic I, and Milic S. Non-alcoholic fatty the link between physical activity, sedentary
 References 107

behavior, physical fitness, and colorectal can- and the role of targeted agents in the treat-
cer. F1000Res 6;2017:199. ment of patients with colorectal liver metasta-
Nash GM, Gimbel M, Shia J et al. KRAS muta- ses: Recommendations from an expert panel.
tion correlates with accelerated metastatic Ann Oncol 20;2009:985–92.
progression in patients with colorectal liver Nosho K, Irahara N, Shima K et al.
metastases. Ann Surg Oncol 17;2010:572–8. Comprehensive biostatistical analysis of CpG
Nazemalhosseini Mojarad E, Kuppen PJ, island methylator phenotype in colorec-
Aghdaei HA, and Zali MR. The CpG island tal cancer using a large population-based
methylator phenotype (CIMP) in colorectal sample. PLOS ONE 3(11);2008:e3698.
cancer. Gastroenterol Hepatol Bed Bench Ogino S, Nosho K, Kirkner GJ et al. CpG island
6(3);2013:120–8. methylator phenotype, microsatellite instabil-
Ng EK, Chong WW, Jin H et al. Differential ity, BRAF mutation and clinical outcome in
expression of microRNAs in plasma of colon cancer. Gut 58(1);2009b:90–6.
patients with colorectal cancer: A potential Ogino S, Nosho K, Kirkner GJ et al. PIK3CA
marker for colorectal cancer screening. Gut mutation is associated with poor prognosis
58(10);2009:1375–81. among patients with curatively resected colon
Nguyen A, Desta Z, and Flockhart DA. Enhancing cancer. J Clin Oncol 27(9);2009a:1477–84.
race-based prescribing precision with Oldenhuis CN, Oosting SF, Gietema JA, and de
pharmacogenomics. Clin Pharmacol Ther Vries EG. Prognostic versus predictive value
81(3);2007:323–5. of biomarkers in oncology. Eur J Cancer
Nibbe RK, Markowitz S, Myeroff L, Ewing R, and 44(7);2008:946–53.
Chance MR. Discovery and scoring of pro- O’Sullivan RJ, and Karlseder J. Telomeres:
tein interaction subnetworks discriminative Protecting chromosomes against
of late stage human colon cancer. Mol Cell genome instability. Nat Rev Mol Cell Biol
Proteomics 8(4);2009:827–45. 11;2010:171–81.
Nicolaides NC, Papadopoulos N, Liu B et al. Oudejans JJ, Slebos RJC, Zoetmulder FAN,
Mutations of two PMS homologues in Mooi WJ, and Rodenhuis S. Differential
hereditary nonpolyposis colon cancer. Nature activation of ras genes by point muta-
371(6492);1994:75–80. tion in human colon cancer with metas-
Nielsen DL, Palshof JA, Brünner N, Stenvang J, tases to either lung or liver. Int J Cancer
and Viuff BM. Implications of ABCG2 expres- 49(6);1991:875–9.
sion on irinotecan treatment of colorectal Ozaslan M, and Aytekin T. Loss of heterozygos-
cancer patients: A review. Int J Mol Sci ity in colorectal cancer. Afr J Biotechnol
18(9);2017, pii: E1926. 8;2009:7308–12.
Nishihara R, Morikawa T, Kuchiba A et al. A Pagès F, Galon J, Dieu-Nosjean MC, Tartour E,
prospective study of duration of smok- Sautès-Fridman C, and Fridman WH. Immune
ing cessation and colorectal cancer risk by infiltration in human tumors: A prognostic
epigenetics-related tumor classification. Am J factor that should not be ignored. Oncogene
Epidemiol 178(1);2013:84–100. 29(8);2010:1093–102.
Nishisho I, Nakamura Y, Miyoshi Y, Miki Y, Paige AJ. Redefining tumour suppressor genes:
Ando H, Horii A, Koyama K, Utsunomiya J, Exceptions to the two-hit hypothesis. Cell Mol
Baba S, and Hedge P. Mutations of chromo- Life Sci Oct 60(10);2003:2147–63.
some 5q21 genes in FAP and colorectal can- Paleari L, Puntoni M, Clavarezza M, DeCensi M,
cer patients. Science 253(5020);1991:665–9. Cuzick J, and DeCensi A. PIK3CA mutation,
Noci S, Dugo M, Bertola F, Melotti F, Vannelli A, aspirin use after diagnosis and survival of
Dragani TA, and Galvan A. A subset of colorectal cancer. A systematic review and
genetic susceptibility variants for colorec- meta-analysis of epidemiological studies. Clin
tal cancer also has prognostic value. Oncol (R Coll Radiol) 28(5);2016:317–26.
Pharmacogenomics J 16(2);2016:173–9. Palomaki GE, Bradley LA, Douglas MP, Kolor K,
Nordlinger B, Van Cutsem E, Gruenberger T et al. and Dotson WD. Can UGT1A1 genotyping
Combination of surgery and chemotherapy reduce morbidity and mortality in patients
108  Precision medicine for colorectal cancer

with metastatic colorectal cancer treated with Pearson G, Robinson F, Beers Gibson T, Xu BE,
irinotecan? An evidence-based review. Genet Karandikar M, Berman K, and Cobb MH.
Med 11(1);2009:21–34. Mitogen-activated protein (MAP) kinase path-
Panagiotou OA, Evangelou E, and Ioannidis JP. ways: Regulation and physiological functions.
Genome-wide significant associations for Endocr Rev 22(2);2001:153–83.
variants with minor allele frequency of 5% Pech MF, Garbuzov A, Hasegawa K et al. High
or less—An overview: A HuGE review. Am J telomerase is a hallmark of undifferentiated
Epidemiol 172(8);2010:869–89. spermatogonia and is required for mainte-
Pancione M, Remo A, and Colantuoni V. Genetic nance of male germline stem cells. Genes Dev
and epigenetic events generate multiple 29(23);2015:2420–34.
pathways in colorectal cancer progression. Peeters M, Douillard JY, Van Cutsem E, Siena S,
Patholog Res Int 2012;2012:509348. Zhang K, Williams R, and Wiezorek J. Mutant
Panczyk M. Pharmacogenetics research on KRAS codon 12 and 13 alleles in patients with
chemotherapy resistance in colorectal cancer metastatic colorectal cancer: Assessment
over the last 20 years. World J Gastroenterol as prognostic and predictive biomarkers
20(29);2014:9775–827. of response to panitumumab. J Clin Oncol
Papadatos-Pastos D, De Miguel Luken MJ, and 31(6);2013:759–65.
Yap TA. Combining targeted therapeutics Peeters M, Oliner KS, Price TJ et al. Analysis of
in the era of precision medicine. Br J Cancer KRAS/NRAS mutations in a phase III study of
112(1);2015:1–3. panitumumab with FOLFIRI compared with
Pardoll DM. The blockade of immune check- FOLFIRI alone as second-line treatment for
points in cancer immunotherapy. Nat Rev metastatic colorectal cancer. Clin Cancer Res
Cancer 12(4);2012:252–64. 21(24);2015:5469–79.
Park Y, Hunter DJ, Spiegelman D et al. Dietary Peeters M, Price TJ, Cervantes A et al. Final
fiber intake and risk of colorectal cancer: A results from a randomized phase 3 study of
pooled analysis of prospective cohort studies. FOLFIRI {+/−} panitumumab for second-line
JAMA 294(22);2005:2849–57. treatment of metastatic colorectal cancer.
Parodi L, Pickering E, Cisar LA, Lee D, and Soufi- Ann Oncol 25(1);2014:107–16.
Mahjoubi R. Utility of pretreatment bilirubin Pellicer M, García-González X, García MI et al.
level and UGT1A1 polymorphisms in multivari- Use of exome sequencing to d ­ etermine
ate predictive models of neutropenia associ- the full profile of genetic variants in the
ated with irinotecan treatment in previously fluoropyrimidine pathway in ­colorectal can-
untreated patients with colorectal cancer. cer patients affected by severe toxicity.
Arch Drug Inf 1(3);2008:97–106. Pharmacogenomics 18(13);2017:1215–23.
Parsons MT, Buchanan DD, Thompson B, Young Peláez IM, Kalogeropoulou M, Ferraro A,
JP, and Spurdle AB. Correlation of tumour Voulgari A, Pankotai T, Boros I, and Pintzas A.
BRAF mutations and MLH1 methylation Oncogenic RAS alters the global and gene-
with germline mismatch repair (MMR) gene specific histone modification pattern during
mutation status: A literature review assess- epithelial-mesenchymal transition in colorec-
ing utility of tumour features for MMR variant tal carcinoma cells. Int J Biochem Cell Biol
classification. J Med Genet 49(3);2012:151–7. 42(6);2010:911–20.
Pavicic W, Joensuu EI, Nieminen T, and Peng H, Duan Z, Pan D, Wen J, and Wei X.
Peltomäki P. LINE-1 hypomethylation in UGT1A1 gene polymorphism predicts
familial and sporadic cancer. J Mol Med (Berl) irinotecan-induced severe neutropenia and
90(7);2012:827–35. diarrhea in Chinese cancer patients. Clin Lab
Pawlik TM, Raut CP, and Rodriguez-Bigas MA. 63(9);2017:1339–46.
Colorectal carcinogenesis: MSI-H versus Perera MA, Innocenti F, and Ratain MJ.
­MSI-L. Dis Markers 20(4–5);2004:199–206. Pharmacogenetic testing for uridine diphos-
Payne SR, and Kemp CJ. Tumor sup- phate glucuronosyltransferase 1A1 polymor-
pressor genetics. Carcinogenesis phisms: Are we there yet? Pharmacotherapy
26(12);2005:2031–45. 28;2008:755–68.
 References 109

Pesenti C, Gusella M, Sirchia SM, and Miozzo M. Popat S, Hubner R, and Houlston RS. Systematic
Germline oncopharmacogenetics, a promis- review of microsatellite instability and
ing field in cancer therapy. Cell Oncol (Dordr) colorectal cancer prognosis. J Clin Oncol
38(1);2015:65–89. 23(3);2005:609–18.
Peyssonnoux C, and Eychene A. The Raf/MEK/ Popovici V, Budinska E, Bosman F, Tejpar S,
ERK pathway: New concepts of activation. Roth A, and Delorenzi M. Context depen-
Biol Cell 93(1–2);2001:53–62. dent interpretation of the prognostic value
Pharoah PD, Antoniou AC, Easton DF, and of BRAF and KRAS mutations in colorectal
Ponder BA. Polygenes, risk prediction, and cancer. BMC Cancer 13(1);2013:439.
targeted prevention of breast cancer. N Engl Populo H et al. mTOR signaling pathway in human
J Med 358(26);2008:2796–803. cancer. Int. J Mol Sci 13;2012:1886–191.
Philippi C, Loretz B, Schaefer UF, and Lehr CM. Power DG, and Kemeny NE. Role of adju-
Telomerase as an emerging target to vant therapy after resection of colorec-
fight cancer—Opportunities and chal- tal cancer liver metastases. J Clin Oncol
lenges for nanomedicine. J Control Release 28(13);2010:2300–9.
146(2);2010:228–40. Pox CP. Controversies in colorectal cancer
Phillips SM, Banerjea A, Feakins R et al. Tumour- screening. Digestion 89(4);2014:274–81.
infiltrating lymphocytes in colorectal cancer Pozzo C, Barone C, and Kemeny NE. Advances
with microsatellite instability are activated in neoadjuvant therapy for colorectal can-
and cytotoxic. Br J Surg 91(4);2004:469–75. cer with liver metastases. Cancer Treat Rev
Pierotti MA. The molecular understanding of can- 34(4);2008:293–301.
cer: From the unspeakable illness to a curable Prahallad A, Sun C, Huang S et al.
disease. Ecancermedicalscience 11;2017:747. Unresponsiveness of colon cancer to
Pietrantonio F, Petrelli F, Coinu A et al. Predictive BRAF (V600E) inhibition through feedback
role of BRAF mutations in patients with ­activation of EGFR. Nature 483(7387);
advanced colorectal cancer receiving cetux- 2012:100–3.
imab and panitumumab: A meta-analysis. Eur Premawardhena A, Fisher CA, Liu YT et al. The
J Cancer 51(5);2015:587–94. global distribution of length polymorphisms
Pino MS, and Chung DC. Microsatellite insta- of the promoters of the glucuronosyltrans-
bility in the management of colorectal ferase 1 gene (UGT1A1): Hematologic and
cancer. Expert Rev Gastroenterol Hepatol evolutionary implications. Blood Cells Mol Dis
5(3);2011:385–99. 31(1);2003:98–101.
Pirmohamed M. Pharmacogenetics and Proctor LM. The National Institutes of
pharmacogenomics. Br J Clin Pharmacol Health Human Microbiome Project.
52(4);2001:345–7. Semin Fetal Neonatal Med 2016, pii:
Pizzolato JF, and Saltz LB. The camptothecins. S1744-165X(16)30016-6.
Lancet 361(9376);2003:2235–42. Pu XX, Huang GL, Guo HQ et al. Circulating
Piñol-Felis C, Fernández-Marcelo T, Viñas- miR-221 directly amplified from plasma is a
Salas J, and Valls-Bautista C. Telomeres potential diagnostic and prognostic marker
and telomerase in the clinical manage- of colorectal cancer and is correlated with
ment of colorectal cancer. Clin Transl Oncol p53 expression. J Gastroenterol Hepatol
19(4);2017:399–408. 25(10);2010:1674–80.
Pomerantz MM, Ahmadiyeh N, Jia L et al. The Puccini A, Berger MD, Naseem M et al.
8q24 cancer risk variant rs6983267 shows Colorectal cancer: Epigenetic alterations and
long-range interaction with MYC in colorectal their clinical implications. Biochim Biophys
cancer. Nat Genet 41(8);2009:882–4. Acta 1868(2);2017:439–48.
Popat S, Chen Z, Zhao D et al. A prospective, Pullarkat ST, Stoehlmacher J, Ghaderi V et al.
blinded analysis of thymidylate synthase and Thymidylate synthase gene polymorphism
p53 expression as prognostic markers in the determines response and toxicity of 5-FU
adjuvant treatment of colorectal cancer. Ann chemotherapy. Pharmacogenomics J
Oncol 17(12);2006:1810–7. 1;2001:65–70.
110  Precision medicine for colorectal cancer

Purcell RV, Visnovska M, Biggs PJ, Schmeier S, and to rapid malignant transformation. Gut
Frizelle FA. Distinct gut microbiome patterns 50(3);2002:382–6.
associate with consensus molecular subtypes Ritchie MD. The success of pharmacogenom-
of colorectal cancer. Sci Rep 7(1);2017:11590. ics in moving genetic association stud-
Qin J, Li R, Raes J et al. A human gut microbial ies from bench to bedside: Study design
gene catalogue established by metagenom- and implementation of precision medi-
ics sequencing. Nature 464(7285);2010:59–65. cine in the p
­ ost-GWAS era. Hum Genet
Rahman L, Voeller D, Rahman M et al. 131(10);2012:1615–26.
Thymidylate synthase as an oncogene: A Robles AI, and Harris CC. Clinical outcomes and
novel role for an essential DNA synthesis correlates of TP53 mutations and cancer.
enzyme. Cancer Cell 5(4);2004:341–51. Cold Spring Harb Perspect Biol 2(3);2010:
Rajagopalan H, Bardelli A, Lengauer C, a001016.
Kinzler KW, Vogelstein B, and Velculescu Rodríguez-Antona C, and Taron M.
VE. Tumorigenesis: RAF/RAS oncogenes Pharmacogenomic biomarkers for per-
and mismatch-repair status. Nature sonalized cancer treatment. J Intern Med
418(6901);2002:934. 277(2);2015:201–17.
Ramchandani RP, Wang Y, Booth BP, Ibrahim A, Roper J and Hung KE. Molecular mechanisms
Johnson JR, Rahman A, Mehta M, Innocenti F, of colorectal carcinogenesis. In: Molecular
Ratain MJ, and Gobburu JV. The role of SN-38 Pathogenesis of Colorectal Cancer. Haigis KM
exposure, UGT1A1*28 polymorphism, and base- (Ed), Springer Science+Business Media, New
line bilirubinlevel in predicting severe irinotecan York, 2013, pp. 25–65.
toxicity. J Clin Pharmacol 47(1);2007:78–86. Roskoski R Jr. RAF protein-serine/threonine
Raskov H, Pommergaard HC, Burcharth J, kinases: Structure and regulation. Biochem
and Rosenberg J. Colorectal carcinogen- Biophys Res Commun 399(3);2010:313–7.
esis—Update and perspectives. World J Rosmarin D, Palles C, Church D et al. Genetic
Gastroenterol 20(48);2014:18151–64. markers of toxicity from capecitabine and other
Rattray NJW, Charkoftaki G, Rattray Z, fluorouracil-based regimens: Investigation in
Hansen JE, Vasiliou V, and Johnson CH. the QUASAR2 study, systematic review, and
Environmental influences in the etiology of meta-analysis. J Clin Oncol 32(10);2014:1031–9.
colorectal cancer: The premise of metabolo- Rosner GL, Panetta JC, Innocenti F, and Ratain MJ.
mics. Curr Pharmacol Rep 3(3);2017:114–25. Pharmacogenetic pathway analysis of irinote-
Ray D, and Kidane D. Gut microbiota imbalance can. Clin Pharmacol Ther 84(3);2008:393–402.
and base excision repair dynamics in colon Rosty C, Young JP, Walsh MD et al. PIK3CA
cancer. J Cancer 7(11);2016:1421–30. activating mutation in colorectal carcinoma:
Read MN, and Holmes AJ. Towards an ­integrative Associations with molecular features and
understanding of diet-host-gut microbiome survival. PLOS ONE 8(6);2013:e65479.
interactions. Front Immunol 8;2017:538. Roth AD. Prognostic role of KRAS and BRAF in
Remvikos Y, Laurent-Puig P, Salmon RJ, Frelat G, stage II and III resected colon cancer: Results
Dutrillaux B, and Thomas G. Simultaneous of the translational study on the PETACC-3,
monitoring of p53 protein and DNA content EORTC 40993, SAKK 60-00 trial. J Clin Oncol
of colorectal adenocarcinomas by flow cytom- 28(3);2010:466–74.
etry. Int J Cancer 45;1990:450–6. Rovcanin B, Ivanovski I, Djuric O, Nikolic D,
Ribic CM, Sargent DJ, Moore MJ et al. Tumor Petrovic J, and Ivanovski P. Mitotic cross-
microsatellite-instability status as a predictor over—An evolutionary rudiment which pro-
of benefit from fluorouracil-based adjuvant motes carcinogenesis of colorectal carcinoma.
chemotherapy for colon cancer. N Engl J Med World J Gastroenterol 20(35);2014:12522–5.
349(3);2003:247–57. Rubbi CP, and Milner J. P53: Gatekeeper, care-
Rijcken FE, Hollema H, and Kleibeuker JH. taker or both?. In: 25 Years of p53 Research.
Proximal adenomas in hereditary non-­ Hainaut P, Wiman KG (Eds). Springer,
polyposis colorectal cancer are prone Dordrecht, 2007, pp. 233–253.
 References 111

Ruden M, and Puri N. Novel anticancer therapeu- colorectal cancer. Irinotecan Study Group. N
tics targeting telomerase. Cancer Treat Rev Engl J Med 343(13);2000:905–14.
39;2013:444–56. Samalin E, Bouché O, Thézenas S et al. Sorafenib
Russo A, Bazan V, Iacopetta B, Kerr D, Soussi T, and irinotecan (NEXIRI) as second- or later-
Gebbia N; TP53-CRC Collaborative Study line treatment for patients with metastatic
Group. The TP53 colorectal cancer interna- colorectal cancer and KRAS-mutated
tional collaborative study on the prognostic tumours: A multicentre Phase I/II trial. British
and predictive significance of p53 muta- J Cancer 110(5);2014:1148–54.
tion: Influence of tumor site, type of muta- Samuels Y, Diaz LA Jr, Schmidt-Kittler O et al.
tion, and adjuvant treatment. J Clin Oncol Mutant PIK3CA promotes cell growth and
23(30);2005:7518–28. invasion of human cancer cells. Cancer Cell
Rustum YM. Thymidylate synthase: A criti- 7(6);2005:561–73.
cal target in cancer therapy? Front Biosci Samuels Y, Wang Z, Bardelli A et al. High fre-
9;2004:2467–73. quency of mutations of the PIK3CA gene in
Ruzzo A, Graziano F, Loupakis F et al. human cancers. Science 304(5670);2004:554.
Pharmacogenetic profiling in patients Sanchez-Castañón M, Er TK, Bujanda L, and
with advanced colorectal cancer treated Herreros-Villanueva M. Immunotherapy in
with first-line FOLFIRI chemotherapy. colorectal cancer: What have we learned so
Pharmacogenomics J 8;2008:278–88. far? Clin Chim Acta 460;2016:78–87.
Ryan-Harshman M, and Aldoori W. Diet and Sandel MH, Speetjens FM, Menon AG et al.
colorectal cancer: Review of the evidence. Natural killer cells infiltrating colorectal can-
Can Fam Physician 53(11);2007:1913–20. cer and MHC class I expression. Mol Immunol
Ryland GL, Doyle MA, Goode D, Boyle SE, 42(4);2005:541–6.
Choong DY, and Rowley SM. Loss of het- Sankila R, Aaltonen LA, Jarvinen HJ, and
erozygosity: What is it good for? BMC Med Mecklin JP. Better survival rates in
Genomics 8;2015:45. patients with MLH1-associated heredi-
Sachan M, and Kaur M. Epigenetic modifications. tary colorectal cancer. Gastroenterology
Therapeutic potential in cancer. Braz Arch Biol 110(3);1996:682–7.
Technol 58(4);2015:526–39. Santos A, Zanetta S, Cresteil T et al. Metabolism
Sacks DB. The role of scaffold proteins in MEK/ of irinotecan (CPT-11) by CYP3A4 and
ERK signalling. Biochem Soc Trans 34(Pt 5); CYP3A5 in humans. Clin Cancer Res
2006:833–6. 6(5);2000:2012–20.
Sai K, Saito Y, Fukushima-Uesaka H et al. Sasaki T, Fujita K, Sunakawa Y et al.
Impact of CYP3A4 haplotypes on irinote- Concomitant polypharmacy is associated
can pharmacokinetics in Japanese cancer with irinotecan related adverse drug reac-
patients. Cancer Chemother Pharmacol tions in patients with cancer. Int J Clin Oncol
62(3);2008:529–37. 18(4);2013:735–42.
Saito Y, Sai K, Maekawa K et al. Close association Saxonov S, Berg P, and Brutlag DL. A genome-
of UGT1A9 IVS1 + 399C > T with UGT1A1*28, wide analysis of CpG dinucleotides in the
*6, or *60 haplotype and its apparent influ- human genome distinguishes two distinct
ence on 7-ethyl-10-hydroxycamptothecin (SN- classes of promoters. Proc Natl Acad Sci USA
38) glucuronidation in Japanese. Drug Metab 103(5);2006:1412–7.
Dispos 37(2);2009:272–6. Schepeler T, Reinert JT, Ostenfeld MS et al.
Sakakibara T, Hibi K, Koike M, Fujiwara M, Diagnostic and prognostic microRNAs
Kodera Y, Ito K, and Nakao A. Plasminogen in stage II colon cancer. Cancer Res
activator inhibitor-1 as a potential marker 68(15);2008:6416–24.
for the malignancy of colorectal cancer. Br J Schmoll HJ, Van Cutsem E, Stein A et al.
Cancer 93(7);2005:799–803. ESMO consensus guidelines for man-
Saltz LB, Cox JV, Blanke C et al. Irinotecan plus agement of patients with colon and
fluorouracil and leucovorin for metastatic rectal cancer. A personalized approach
112  Precision medicine for colorectal cancer

to clinical decision making. Ann Oncol Shin JS, Hong A, Solomon MJ, and Lee CS.
23(10);2012:2479–516. The role of telomeres and telomerase in
Schork N, Murray SS, Frazer KA, and Topol the pathology of human cancer and aging.
EJ. Common vs. rare allele hypotheses for Pathology 38;2006:103–13.
complex diseases. Curr Opin Genet Dev Shriver Z, Raguram S, and Sasisekharan R.
19;2009:212–9. Glycomics: A pathway to a class of new and
Schwarzenbach H. Predictive diagnostics in improved therapeutics. Nat Rev Drug Discov
colorectal cancer: Impact of genetic polymor- 3;2004:863–73.
phisms on individual outcomes and treatment Siddiqui AD, and Piperdi B. KRAS muta-
with fluoropyrimidine-based chemotherapy. tion in colon cancer: A marker of resis-
EPMA J 1(3);2010:485–94. tance to EGFR-I Therapy. Ann Surg Oncol
Senter L, Clendenning M, Sotamaa K et al. The 17(4);2010:1168–76.
clinical phenotype of Lynch syndrome due to Siegel RL, Miller KD, and Jemal A. Cancer
germ-line PMS2 mutations. Gastroenterology ­statistics, 2018. CA Cancer J Clin
135(2);2008:419–28. 68(1);2018:7–30.
Sevinc A, Kalender ME, Altinbas M, Ozkan Sikdar S, Datta S, and Datta S. Exploring the
M, Dikilitas M, Camci C; Anatolian Society importance of cancer pathway by meta-anal-
of Medical Oncology (ASMO). Irinotecan ysis of differential protein expression net-
as a second-line monotherapy for small works in three different cancers. Biol Direct
cell lung cancer. Asian Pac J Cancer Prev 11(1);2016:65.
2(4);2011:1055–9. Simon PH, Sylvestre MP, Tremblay J, and Hamet P.
Shaik AP, Shaik AS, and Al-Sheikh YA. Colorectal Key considerations and methods in the study
cancer: A review of the genome-wide of gene-environment interactions. Am J
­association studies in the kingdom of Saudi Hypertens 29(8);2016 Apr 1:891–9, pii: hpw021.
Arabia. Saudi J Gastroenterol 21(3);2015:123–8. Simonis M, Klous P, Splinter E et al. Nuclear
Shammas MA. Telomeres, lifestyle, cancer, organization of active and inactive chromatin
and aging. Curr Opin Clin Nutr Metab Care domains uncovered by chromosome con-
14(1);2011:28–34. formation capture-on-chip (4C). Nat Genet
Shaw RJ, and Cantley LC. Ras, PI(3)K and mTOR 38(11);2006:1348–54.
signalling controls tumour cell growth. Nature Sinicrope FA, and Sargent DJ. Molecular path-
441(7092);2006:424–30. ways: Microsatellite instability in colorec-
Shay JW, and Wright WE. Telomeres and telom- tal cancer: Prognostic, predictive, and
erase in normal and cancer stem cells. FEBS therapeutic implications. Clin Cancer Res
Lett 584(17);2010:3819–25. 18;2012:1506–12.
Shay JW, and Wright WE. Role of telomeres Siobhan V, Glavey SV, Huynh D, Reagan MR
and telomerase in cancer. Semin Cancer Biol et al. The cancer glycome: Carbohydrates
21(6);2011:349–53. as mediators of metastasis. Blood Rev
Sheffield LJ, and Phillimore HE. Clinical use of 29(4);2015:269–79.
pharmacogenomic tests. Clin Biochem Rev Sjöblom T, Jones S, Wood LD et al. The
30;2009:55–65. consensus coding sequences of human
Shendure J. Next-generation human genetics. breast and colorectal cancers. Science
Genome Biol 12(9);2011:408. 314(5797);2006:268–74.
Shima H, Hiyama T, Tanaka S et al. Loss of Slattery ML, Wolff E, Hoffman MD, Pellatt
heterozygosity on chromosome 10p14- DF, Milash B, and Wolff RK. MicroRNAs
p15 in colorectal carcinoma. Pathobiology and colon and rectal cancer: Differential
72(4);2005:220–4. ­expression by tumor location and sub-
Shimoyama S. Pharmacogenetics of irinotecan: type. Genes Chromosomes Cancer
An ethnicity-based prediction of irinotecan 50(3);2011:196–206.
adverse events. World J Gastrointest Surg Sloane BF, Berquin IM. Proteases and cancer:
2(1);2010:14–21. An introduction. In: Proteolysis and Protein
 References 113

Turnover. Bond JS and Barrett AJ (Eds), models of colorectal cancer. Clin Cancer Res
Portland Press (Ed.), London, 1993, pp. 19(15);2013:4149–62.
225–31. Steck PA, Pershouse MA, Jasser SA et al.
Smith G, Carey FA, Beattie J et al. Mutations Identification of a candidate tumor suppres-
in APC, Kirsten-ras, and p53—Alternative sor gene, MMAC1, at chromosome 10q23.3
genetic pathways to colorectal cancer. Proc that is mutated in multiple advanced cancers.
Natl Acad Sci USA 99;2002:9433–8. Nat Genet 15(4);1997:356–62.
Smith NF, Figg WD, and Sparreboom A. Steelman LS, Pohnert SC, Shelton JG, Franklin
Pharmacogenetics of irinotecan metabolism RA, Bertrand FE, and McCubrey JA. JAK/­
and transport: An update. Toxicol In Vitro STAT, Raf/MEK/ ERK, PI3K/Akt and BCR-ABL
20(2);2006:163–75. in cell cycle progression and leukemogenesis.
Sobhani I, Amiot A, Le Baleur Y, Levy M, Auriault Leukemia 18(2);2004:189–218.
ML, Van Nhieu JT, and Delchier JC. Microbial Stewart CF, Panetta JC, O’Shaughnessy MA et al.
dysbiosis and colon carcinogenesis: Could UGT1A1 promoter genotype correlates with
colon cancer be considered a bacteria- SN-38 pharmacokinetics, but not severe toxic-
related disease? Therap Adv Gastroenterol ity in patients receiving low-dose irinotecan.
6(3);2013:215–29. J Clin Oncol 25(18);2007:2594–600.
Sood A, McClain D, Maitra R et al. PTEN gene Stigliano V, Sanchez-Mete L, Martayan A, and
expression and mutations in the PIK3CA gene Anti M. Early-onset colorectal cancer: A
as predictors of clinical benefit to anti-epider- sporadic or inherited disease? World J
mal growth factor receptor antibody therapy Gastroenterol. 20(35);2014:12420–30.
in patients with KRAS wild-type metastatic Stoehlmacher J, Goekkurt E, Mogck U et al.
colorectal cancer. Clin Colorectal Cancer Thymidylate synthase genotypes and tumour
11(2);2012:143–50. regression in stage II/III rectal cancer patients
Sorich MJ, Wiese MD, Rowland A, after neoadjuvant fluorouracil-based chemo-
Kichenadasse G, McKinnon RA, and radiation. Cancer Lett 272(2);2008:221–5.
Karapetis CS. Extended RAS mutations and Stracci F, Zorzi M, and Grazzini G. Colorectal can-
anti-EGFR monoclonal antibody survival ben- cer screening: Tests, strategies, and perspec-
efit in metastatic colorectal cancer: A meta- tives. Front Public Health 2;2014:210.
analysis of randomized controlled trials. Ann Stranger BE, Stahl EA, and Raj T. Progress and
Oncol 26(1);2015:13–21. promise of genome-wide association studies
Sorrell AD, Espenschied CR, Culver JO, and for human complex trait genetics. Genetics
Weitzel JN. Tumor protein p53 (TP53) testing 187(2);2011:367–83.
and Li-Fraumeni syndrome: Current status of Strassburg CP. Pharmacogenetics of
clinical applications and future directions. Mol Gilbert’s syndrome. Pharmacogenomics
Diagn Ther 17(1);2013:31–47. 9(6);2008:703–15.
Souglakos J, Philips J, Wang R et al. Prognostic Strassburg CP, Barut A, Obermayer-Straub P,
and predictive value of common mutations for Li Q, Nguyen N, Tukey RH, and Manns MP.
treatment response and survival in patients Identification of cyclosporine A and tacro-
with metastatic colorectal cancer. Br J Cancer limus glucuronidation in human liver and
101(3);2009:465–72. the gastrointestinal tract by a differentially
Spilianakis CG, Lalioti MD, Town T, Lee GR, and expressed UDP-glucuronosyltransferase:
Flavell RA. Interchromosomal associations UGT2B7J. Hepatol 34(6);2001:865–72.
between alternatively expressed loci. Nature Strassburg CP, Manns MP, and
435(7042);2005:637–45. Tukey RH. Expression of the UDP-
Spreafico A, Tentler JJ, Pitts TM et al. Rational glucuronosyltransferase 1A locus in human
combination of a MEK inhibitor, selu- colon. Identification a characterization of
metinib, and the Wnt/calcium pathway the novel extrahepatic UGT1A8. J Biol Chem
modulator, cyclosporin A, in preclinical 273;1998:8719–26.
114  Precision medicine for colorectal cancer

Strassburg CP, Oldhafer K, Manns MP, and Tukey folate metabolic pathways and colorec-
RH. Differential expression of the UGT1A tal cancer risk: The JPHC Study. Sci Rep
locus in human liver, biliary, and gastric tissue: 6;2016:36607.
Identification of UGT1A7 and UGT1A10 tran- Swen JJ, Nijenhuis M, de Boer A et al.
scripts in extrahepatic tissue. Mol Pharmacol Pharmacogenetics: From bench to byte—An
52;1997:212–20. update of guidelines. Clin Pharmacol Ther
Strauss BS. Hypermutability in carcinogenesis. 89(5);2011:662–73.
Genetics 148(4);1998:1619–26. Symonds DA, and Vickery AL. Mucinous car-
Strubberg AM, and Madison BB. MicroRNAs in cinoma of the colon and rectum. Cancer
the etiology of colorectal cancer: Pathways 37(4);1976:1891–900.
and clinical implications. Dis Model Mech Takahashi Y, Ishii Y, Nagata T, Ikarashi M,
10(3);2017:197–214. Ishikawa K, and Asai S. Clinical application
Stucky-Marshall L. New agents in gastrointesti- of oligonucleotide probe array for full-length
nal malignancies: Part 1: Irinotecan in clinical gene sequencing of TP53 in colon cancer.
practice. Cancer Nurs 22(3);1999:212–9. Oncology 64;2003:54–60.
Sud A, Kinnersley B, and Houlston RS. Genome- Takahashi Y, Sugai T, Habano W et al. Molecular
wide association studies of cancer: Current differences in the microsatellite stable
insights and future perspectives. Nat Rev phenotype between left-sided and right-
Cancer 17(11);2017:692–704. sided colorectal cancer. Int J Cancer
Sudhan DR, and Siemann DW. Cathepsin L tar- 139(1);2016:2493–501.
geting in cancer treatment. Pharmacol Ther Takano M, and Sugiyama T. UGT1A1 poly-
155;2015:105–16. morphisms in cancer: Impact on irinote-
Sugai T, Habano W, Jiao Y-F et al. Analysis of can treatment. Pharmgenomics Pers Med
molecular alterations in left- and right-sided 10;2017:61–8.
colorectal carcinomas reveals distinct path- Tampellini M, Sonetto C, and Scagliotti GV.
ways of carcinogenesis: Proposal for new Novel anti-angiogenic therapeutic strategies
molecular profile of colorectal carcinomas. in colorectal cancer. Expert Opin Investig
J Mol Diagn 8(2);2006:193–201. Drugs 25(5);2016:507–20.
Sugai T, Yoshida M, Eizuka M et al. Analysis of Tan GJ, Peng ZK, Lu JP, and Tang FQ. Cathepsins
the DNA methylation level of cancer-related mediate tumor metastasis. World J Biol Chem
genes in colorectal cancer and the sur- 4(4);2013:91–101.
rounding normal mucosa. Clin Epigenetics Tanaka F, Fukuse T, Wada H, and Fukushima M.
9;2017:55. The history, mechanism and clinical use
Sugatani J. Function, genetic polymorphism, and of oral 5-fluorouracil derivative chemo-
transcriptional regulation of human UDP- therapeutic agents. Curr Pharm Biotechnol
glucuronosyltransferase (UGT) 1A1. Drug 1(2);2000:137–64.
Metab Pharmacokinet 28(2);2013:83–92. Epub Tanaka M, Omura K, Watanabe Y, Oda Y, and
2012 Oct 23. Nakanishi I. Prognostic factors of colorec-
Sukari A, Nagasaka M, Al-Hadidi A, and Lum LG. tal cancer: K-ras mutation, overexpression
Cancer Immunology and Immunotherapy. of the P53-protein, and cell proliferative
Anticancer Res 36(11);2016:5593–606. ­activity. J Surg Oncol 57(1);1994:57–64.
Sun YL, Patel A, Kumar P, and Chen ZS. Role of Tanaka T, Watanabe T, Kazama Y, Tanaka J,
ABC transporters in cancer chemotherapy. Kanazawa T, Kazama S, and Nagawa H. Loss
Chin J Cancer 31(2);2012:51–7. of Smad4 protein expression and 18qLOH
Surget S, Khoury MP, and Bourdon JC. as molecular markers indicating lymph node
Uncovering the role of p53 splice variants in metastasis in colorectal cancer—A study
human malignancy: A clinical perspective. matched for tumor depth and pathology.
Onco Targets Ther 7;2013:57–68. J Surg Oncol 97(1);2008:69–73.
Svensson T, Yamaji T, Budhathoki S, Hidaka A, Tanase C, Albulescu R, and Neagu M. Proteomics
Iwasaki M, and Sawada N. Alcohol consump- approaches for biomarker panels in cancer. J
tion, genetic variants in the alcohol- and Immunoassay Immunochem 37(1);2016:1–15.
 References 115

Taylor WR, and Stark GR. Regulation of Tran B, Dancey JE, Kamel-Reid S et al. Cancer
the G2/M transition by p53. Oncogene genomics: Technology, discovery, and transla-
20(15);2001:1803–15. tion. J Clin Oncol 30(6);2012:647–60.
Teft WA, Welch S, Lenehan J et al. OATP1B1 Tsutani Y, Yoshida K, Sanada Y, Wada Y,
and tumour OATP1B3 modulate expo- Konishi K, Fukushima M, and Okada M.
sure, toxicity, and survival after irinote- Decreased orotate phosphoribosyltransferase
can-based chemotherapy. Br J Cancer activity produces 5-fluorouracil resistance in
112(5);2015:857–65. a human gastric cancer cell line. Oncol Rep
Tejpar S, Lenz H-J, Köhne C-H et al. Effect of 20(6);2008:1545–51.
KRAS and NRAS mutations on treatment out- Tukey RH, and Strassburg CP. Human UDP-
comes in patients with metastatic colorectal glucuronosyltransferases: Metabolism,
cancer (mCRC) treated first-line with cetux- expression, and disease. Annu Rev Pharmacol
imab plus FOLFOX-4: New results from OPUS Toxicol 40;2000:581–616.
study. JCO 32(suppl3 abstr LBA444);2014. Tukey RH, Strassburg CP, and Mackenzie PI.
Terzić J, Grivennikov S, Karin E, and Pharmacogenomics of human UDP-
Karin M. Inflammation and colon cancer. glucuronosyltransferases and irinotecan toxic-
Gastroenterology 138;2010:2101–14.e5. ity. Mol Pharmacol 62(3);2002:446–50.
Tezcan G, Tunca B, Ak S, Cecener G, and Egeli U. Tuy HD, Shiomi H, Mukaisho KI et al. ABCG2
Molecular approach to genetic and epigenetic expression in colorectal adenocarcinomas
pathogenesis of early-onset colorectal cancer. may predict resistance to irinotecan. Oncol
World J Gastrointest Oncol 8(1);2016:83–98. Lett 12(4);2016 Oct:2752–60.
Therkildsen C, Bergmann TK, Henrichsen- Tziotou M, Kalotychou V, Ntokou A et al.
Schnack T, Ladelund S, and Nilbert M. The Polymorphisms of uridine glucuronosyl-
predictive value of KRAS, NRAS, BRAF, transferase gene and irinotecan toxicity:
PIK3CA and PTEN for anti-EGFR treatment Low dose does not protect from toxicity.
in metastatic colorectal cancer: A system- Ecancermedicalscience 8;2014:428.
atic review and meta-analysis. Acta Oncol Ulrich CM, Bigler J, Velicer CM, Greene EA, Farin
53(7);2014:852–64. FM, and Potter JD. Searching expressed
Thomas T, Gilbert J, and Meyer F. sequence tag databases: Discovery and con-
Metagenomics—A guide from sampling to firmation of a common polymorphism in the
data analysis. Microb Inform Exp 2(1);2012:3. thymidylate synthase gene. Cancer Epidemiol
Thompson N, and Lyons J. Recent progress in Biomarkers Prev 9(12);2000:1381–5.
targeting the Raf/MEK/ERK pathway with van Brummelen EMJ, de Boer A, Beijnen JH, and
inhibitors in cancer drug discovery. Curr Opin Schellens JHM. BRAF mutations as predictive
Pharmacol 5(4);2005:350–6. biomarker for response to anti-EGFR monoclo-
Thélin C, and Sikka S. Epidemiology of nal antibodies. Oncologist 22(7);2017:864–72.
colorectal cancer—Incidence, lifetime risk Van Cutsem E, Köhne CH, Hitre E et al.
factors statistics and temporal trends. Cetuximab and chemotherapy as initial treat-
In: Screening for Colorectal Cancer with ment for metastatic colorectal cancer. N Engl
Colonoscopy, Ettarh R (Ed), InTech, 2015, J Med 360(14);2009:1408–17.
doi: 10.5772/61945. Available from https:// Van Cutsem E, Köhne CH, Láng I et al. Cetuximab
www.intechopen.com/books/screening- plus irinotecan, fluorouracil, and leucovorin as
for-colorectal-cancer-with-colonoscopy/ first-line treatment for metastatic colorectal
epidemiology-of-colorectal-cancer-incidence- cancer: Updated analysis of overall survival
lifetime-risk-factors-statistics-and-temporal- according to tumor KRAS and BRAF mutation
trends. status. J Clin Oncol 29(15);2011:2011–9.
Toffoli G, Cecchin E, Corona G et al. The role of Van Cutsem E, Lenz H-J, Köhne CH et al.
UGT1A1*28 polymorphism in the pharmaco- Fluorouracil, leucovorin, and irinotecan
dynamics and pharmacokinetics of irinotecan plus cetuximab treatment and RAS muta-
in patients with metastatic colorectal cancer. tions in colorectal cancer. J Clin Oncol
J Clin Oncol 24(19);2006:3061–8. 33(7);2015:692–700.
116  Precision medicine for colorectal cancer

Van Cutsem E, Tabernero J, Lakomy R et al. metastatic colorectal cancer. PLOS ONE
Addition of aflibercept to fluorouracil, 10;2015:e0124538.
leucovorin, and irinotecan improves survival Venter JC, Adams MD, Myers EW et al. The
in a phase III randomized trial in patients sequence of the human genome. Science
with metastatic colorectal cancer previously 291(5507);2001:1304–51.
treated with an oxaliplatin-based regimen. Verdaguer H, Saurí T, and Macarulla T. Predictive
J Clin Oncol 30;2012:3499–506. and prognostic biomarkers in personal-
van der Bol JM, Mathijssen RH, Creemers GJ ized gastrointestinal cancer treatment.
et al. A CYP3A4 phenotype-based dosing J Gastrointest Oncol 8(3);2017:405–17.
algorithm for individualized treatment of iri- Vignoli M, Nobili S, Napoli C et al. Thymidylate
notecan. Clin Cancer Res 16(2);2010:736–42. synthase expression and g Thymidylate syn-
Van der Meide PH, and Schellekens H. Cytokines thase expression and genotype have no major
and the immune response. Biotherapy impact on the clinical outcome of colorectal
8(3–4);1996:243–9. cancer patients treated with 5-fluorouracil.
van Es HH, Bout A, Liu J et al. Assignment of Pharmacol Res 64(3);2011:242–8.
the human UDP glucuronosyltransferase Visscher PM, Wray NR, Zhang Q, Sklar P,
gene (UGT1A1) to chromosome region 2q37. McCarthy MI, Brown MA, and Yang J.
Cytogenet Cell Genet 63(2);1993:114–6. 10 years of GWAS discovery: Biology, func-
van Kuilenburg AB, Muller EW, Haasjes J et al. tion, and translation. Am J Hum Genet
Lethal outcome of a patient with a complete 101(1);2017:5–22.
dihydropyrimidine dehydrogenase (DPD) Vital M, Howe AC, and Tiedje JM. Revealing
deficiency after administration of 5-fluoroura- the bacterial butyrate synthesis pathways
cil: Frequency of the common IVS14 + 1G > A by analyzing (meta)genomic data. MBio
mutation causing DPD deficiency. Clin Cancer 5(2);2014:e00889.
Res 7(5);2001:1149–53. Vivot A, Boutron I, Ravaud P, and Porcher R.
van Staveren MC, Guchelaar HJ, van Guidance for pharmacogenomic biomarker
Kuilenburg AB, Gelderblom H, and Maring testing in labels of FDA-approved drugs.
JG. Evaluation of predictive tests for Genet Med 17(9);2015:733–8.
screening for dihydropyrimidine dehydro- Vogelstein B, Fearon ER, Hamilton SR et al.
genase ­deficiency. Pharmacogenomics J Genetic alterations during colorectal-
13;2013:389–95. tumor development. N Engl J Med
Vandana G, Lokesh Rao Magar S, Sweth D, 319(9);1988:525–32.
and Sandhya S. An expression of p53 Vogelstein B, Fearon ER, Kern SE, Hamilton SR,
marker in colorectal cancer with histo- Preisinger AC, Nakamura Y, and White R.
pathological correlation. Int Arch Integr Med Allelotype of colorectal carcinomas. Science
4(12);2017:168–84. 244(4901);1989:207–11.
Varley JM. Germline TP53 mutations and Vogelstein B, Lane D, and Levine AJ. Surfing the
Li-Fraumeni syndrome. Hum Mutat p53 network. Nature 408(6810);2000:307–10.
21(3);2003:313–20. Vogelstein B, Papadopoulos N, Velculescu VE,
Velho S, Moutinho C, Cirnes L et al. BRAF, KRAS Zhou S, Diaz LA Jr, and Kinzler KW.
and PIK3CA mutations in colorectal serrated Cancer genome landscapes. Science
polyps and cancer: Primary or secondary 339(6127);2013:1546–58.
genetic events in colorectal carcinogenesis? Volinia S, Hiles I, Ormondroyd E, Nizetic D,
BMC Cancer 8;2008:255. Antonacci R, Rocchi M, and Waterfield MD.
Velho S, Oliveira C, Ferreira A et al. The Molecular cloning, cDNA sequence and chro-
prevalence of PIK3CA mutations in gas- mosomal localization of the human phosphati-
tric and colon cancer. Eur J Cancer dylinositol 3-kinase p110 alpha (PIK3CA) gene.
41(6);2005:1649–54. Genomics 24(3);1994:472–7.
Venderbosch S, van Lent-van Vliet S, de Haan Wallace K, DeToma A, Lewin DN, Sun
AF et al. EMAST is associated with a S, Rockey D, Britten CD, Wu JD, Ba
poor prognosis in microsatellite instable A, Alberg AJ, and Hill EG. Racial
 References 117

differences in stage IV colorectal cancer Whitehall VL, Rickman C, Bond CE et al.

survival in younger and older patients. Clin Oncogenic PIK3CA mutations in colorec-
Colorectal Cancer 16(3);2016:178–86, pii: tal cancers and polyps. Int J Cancer
S1533-0028(16)30260-2. 131(4);2012:813–20.
Wan J, Li H, Li Y, Zhu ML, and Zhao P. Loss of Wiesner GL, Slavin TP, and Barnholtz-Sloan JS.
heterozygosity of Kras2 gene on 12p12-13 in Colorectal cancer. In Genomics and Personalized
Chinese colon carcinoma patients. World J Medicine. 1st edition. Ginsbur GS, Williard HF
Gastroenterol 12(7);2006:1033–7. (Eds), Academic Press, 2010, pp. 457–476.
Wang S, Zhang C, Zhang Z, Qian W, Sun Y, and Ji Wilhelm SM, Carter C, Tang L, Wilkie D,
B. Transcriptome analysis in primary colorec- McNabola A, and Rong H. BAY 43-9006
tal cancer tissues from patients with and with- exhibits broad spectrum oral antitumor activ-
out liver metastases using next-generation ity and targets the RAF/MEK/ERK pathway
sequencing. Cancer Med 6(8);2017:1976–87. and receptor tyrosine kinases involved in
Wang W, Wang GQ, Sun XW et al. Prognostic tumor progression and angiogenesis. Cancer
values of chromosome 18q microsatellite Res 64(19);2004:7099–109.
alterations in stage II colonic carcinoma. Wilson AJ, Byun DS, Nasser S et al. HDAC4
World J Gastroenterol 16(47);2010:6026–34. promotes growth of colon cancer cells
Wang W, Zheng L, Zhou N et al. Meta-analysis via repression of p21. Mol Biol Cell
of associations between telomere length and 19(10);2008:4062–75.
colorectal cancer survival from observational Wilson AJ, Byun DS, Popova N et al. Histone
studies. Oncotarget 8(37);2017:62500–7. deacetylase 3 (HDAC3) and other class I
Watanabe T, Itabashi M, Shimada Y et al. HDACs regulate colon cell maturation and
Japanese society for cancer of the colon and p21 expression and are deregulated in human
rectum (JSCCR) guidelines 2014 for treat- colon cancer. J Biol Chem (19);2006:13548–58.
ment of colorectal cancer. Int J Clin Oncol Wilson JL, and Altman RB. Biomarkers: Delivering
20(2);2015:207–39. on the expectation of molecularly driven,
Watson MM, Lea D, Rewcastle E, Hagland HR, and quantitative health. Exp Biol Med (Maywood)
Søreide K. Elevated microsatellite alterations 243(3);2017:313–22.
at selected tetranucleotides in early-stage- Wolpin B, Ng K, Zhu AX et al. Multicenter phase
colorectal cancers with and without high-fre- II study of tivozanib (AV-951) and everoli-
quency microsatellite instability: Same, same mus (RAD001) for patients with refractory,
but different? Cancer Med 5(7);2016:1580–7. metastatic colorectal cancer. Oncologist
Webber EM, Kauffman TL, O’Connor E, and 18;2013:377–8.
Goddard KA. Systematic review of the predic- Wong RS. Apoptosis in cancer: From pathogenesis
tive effect of MSI status in colorectal cancer to treatment. J Exp Clin Cancer Res 30;2011:87.
patients undergoing 5FU-based chemother- Wu S, Gan Y, Wang X, Liu J, Li M, and Tang Y.
apy. BMC Cancer 15;2015:156. PIK3CA mutation is associated with poor sur-
Weinshilboum RM, and Wang L. vival among patients with metastatic colorec-
Pharmacogenomics: Precision medi- tal cancer following anti-EGFR monoclonal
cine and drug response. Mayo Clin Proc antibody therapy: A meta-analysis. J Cancer
92(11);2017:1711–22. Res Clin Oncol 139(5);2013:891–900.
Weng L, Zhang L, Peng Y, and Huang RS. Xiang X, Jada SR, Li HH et al. Pharmacogenetics
Pharmacogenetics and pharmacogenomics: of SLCO1B1 gene and the impact of *1b and
A bridge to individualized cancer therapy. *15 haplotypes on irinotecan disposition
Pharmacogenomics 14(3);2013:315–24. in Asian cancer patients. Pharmacogenet
Werner RJ, Kelly AD, and Issa JJ. Epigenetics and Genomics 16;2006:683–91.
precision oncology. Cancer J 23(5);2017:262–9. Xiao Y, and Freeman GJ. The microsatellite
Whiffin N, Hosking FJ, Farrington SM et al. instable subset of colorectal cancer is a
Identification of susceptibility loci for colorec- particularly good candidate for checkpoint
tal cancer in a genome-wide meta-analysis. blockade immunotherapy. Cancer Discov
Hum Mol Genet 23(17);2014:4729–37. 5(1);2015:16–8.
118  Precision medicine for colorectal cancer

Xie FW, Peng YH, Wang WW, Chen X, Chen X, expression and activity via DNA methylation
Li J, Yu ZY, and Ouyang XN. Influence of in healthy human livers. J Pharm Pharmacol
UGT1A1 gene methylation level in colorectal 65;2013:874–83.
cancer cells on the sensitivity of the chemo- Yeh JJ, Routh ED, Rubinas T et al. KRAS/
therapy drug CPT-11. Biomed Pharmacother BRAF mutation status and ERK1/2 activa-
68(7);2014:825–31. tion as biomarkers for MEK1/2 inhibitor
Xu JM, Wang Y, Ge FJ, Lin L, Liu ZY, and therapy in colorectal cancer. Mol Cancer Ther
Sharma MR. Severe irinotecan-induced toxic- 8(4);2009:834–43.
ity in a patient with UGT1A1 28 and UGT1A1 Yokota T, Ura T, Shibata N et al. BRAF mutation
6 polymorphisms. World J Gastroenterol is a powerful prognostic factor in advanced
19(24);2013:3899–903. and recurrent colorectal cancer. Br J Cancer
Xu Y. Induction of genetic instability by gain- 104(5);2011:856–62.
of-function p53 cancer mutants. Oncogene Yoon HH, Tougeron D, Shi Q et al. KRAS codon
27(25);2008:3501–7. 12 and 13 mutations in relation to disease-
Xu Y, and Villalona-Calero M. Irinotecan: free survival in BRAF–Wild-Type stage III
Mechanisms of tumor resistance and novel colon cancers from an adjuvant chemo-
strategies for modulating its activity. Ann therapy trial (N0147 Alliance). Clin Cancer Res
Oncol 13;2002:1841–51. 20(11);2014:3033–43.
Yamada H, Ichikawa W, Uetake H, Shirota Y, You JS, and Jones PA. Cancer genetics and epi-
Nihei Z, Sugihara K, and Hirayama R. genetics: Two sides of the same coin? Cancer
Thymidylate synthase gene expres- Cell 22(1);2012:9–20.
sion in ­primary colorectal cancer and Yu J, Feng Q, Wong SH et al. Metagenomic anal-
metastatic sites. Clin Colorectal Cancer ysis of faecal microbiome as a tool towards
1(3);2001:169–73. targeted non-invasive biomarkers for colorec-
Yamauchi M, Morikawa T, Kuchiba A et al. tal cancer. Gut 66(1);2017:70–8.
Assessment of colorectal cancer molecular Yu J, Shannon WD, Watson MA, and McLeod HL.
features along bowel subsites challenges Gene expression profiling of the irinotecan
the conception of distinct dichotomy of pathway in colorectal cancer. Clin Cancer Res
proximal versus distal colorectum. Gut 11(5);2005:2053–62.
61(6);2012:847–54. Zauber P, Sabbath-Solitare M, Marotta SP, and
Yang C, Liu Y, Xi WQ et al. Relationship between Bishop T. Loss of heterozygosity for chro-
UGT1A1*6/*28 polymorphisms and severe mosome 18q and microsatellite instability
toxicities in Chinese patients with pancreatic are highly consistent across the region of
or biliary tract cancer treated with irinotecan- the DCC and SMAD4 genes in colorec-
containing regimens. Drug Des Devel Ther tal carcinomas and adenomas. J Appl Res
9;2015:3677–83. 8(1);2008:14–23.
Yang J, Nishihara R, Zhang X, Ogino S, and Zeng P, Zhao Y, Qian C et al. Statistical analysis
Qian ZR. Energy sensing pathways: Bridging for genome-wide association study. J Biomed
type 2 diabetes and colorectal cancer? J Res 29(4);2015:285–97.
Diabetes Complications 31(7);2017:1228–36. Zgheib NK, Arawi T, Mahfouz RA, and Sabra R.
Yang ZY, Wu XY, Huang YF et al. Promising Attitudes of health care professionals toward
biomarkers for predicting the outcomes of pharmacogenetic testing. Mol Diagn Ther
patients with KRAS wild-type metastatic 15;2011:115–22.
colorectal cancer treated with anti-epidermal Zhang C, Chen X, Li L, Zhou Y, Wang C,
growth factor receptor monoclonal antibod- and Hou S. The association between
ies: A systematic review with meta-analysis. telomere length and cancer prognosis:
Int J Cancer 133(8);2013:1914–25. Evidence from a ­meta-analysis. PLOS ONE
Yasar U, Greenblatt DJ, Guillemette C, and 10(7);2015:e0133174.
Court MH. Evidence for regulation of UDP- Zhang CM, Lv JF, Gong L, Yu LY, Chen XP, Zhou
glucuronosyltransferase (UGT)1A1 protein HH, and Fan L. Role of deficient mismatch
 References 119

repair in the personalized management of Zhu L, and Fang J. The structure and clini-
colorectal cancer. Int J Environ Res Public cal roles of MicroRNA in colorectal cancer.
Health 13(9);2016, pii: E892. Gastroenterol Res Pract 2016;2016:1360348.
Zhang X, Cowper-Sal lari R, Bailey SD, Moore Zlobec I, Bihl MP, Schwarb H, Terracciano L,
JH, and Lupien M. Integrative functional and Lugli A. Clinicopathological and protein
genomics identifies an enhancer looping to characterization of BRAF- and K-RAS-mutated
the SOX9 gene disrupted by the 17q24.3 colorectal cancer and implications for prog-
prostate cancer risk locus. Genome Res nosis. Int J Cancer 127(2);2009:367–80.
22(8);2012:1437–46. Zlobec I, Kovac M, Erzberger P et al. Combined
Zhang X, Kelaria S, Kerstetter J, and Wang J. analysis of specific KRAS mutation, BRAF
The functional and prognostic implications and microsatellite instability identifies
of regulatory T cells in colorectal carcinoma. prognostic subgroups of sporadic and
J Gastrointest Oncol 6(3);2015:307–13. hereditary colorectal cancer. Int J Cancer
Zhang X, Yin JF, Zhang J, Kong SJ, Zhang HY, 127;2010:2569–75.
and Chen XM. UGT1A1*6 polymorphisms Zumwalt TJ, Wodarz D, Komarova NL et al.
are correlated with irinotecan-induced Aspirin induced chemoprevention and
neutropenia: A systematic review and meta- response kinetics are enhanced by PIK3CA
analysis. Cancer Chemother Pharmacol mutations in colorectal cancer cells. Cancer
80(1);2017:135–49. Prev Res (Phila) 10(3);2017:208–18.
Zheng HT, Peng ZH, Li S, and He L. Loss of Zurek M, Altschmied J, Kohlgruber S, Ale-
heterozygosity analyzed by single nucleo- Agha N, and Haendeler J. Role of telomerase
tide polymorphism array in cancer. World J in the cardiovascular system. Genes (Basel)
Gastroenterol 11(43);2005:6740–4. 7(6);2016:29.
 6
Precision medicine in prostate cancer

NIGEL P. MURRAY

Introduction 121 Clinical application of SNPs in prostate cancer 129

The prostate-specific antigen (PSA) Environmental effects on SNPs 130
controversy for prostate cancer screening 123 Conclusions 130
Genetic screening and personalized medicine 124 Acknowledgments 131
Rare high penetrance prostate cancer Conflicts of interest 131
susceptibility genes 124 References 131
Common and low-penetrance risk-associated
single nucleotide polymorphisms 126

INTRODUCTION it concluded that the evidence was insufficient to

assess the balance of benefits and harms of prostate
With the changing demographics of the world cancer screening using the serum prostate-specific
population and increasing life expectancy, prostate antigen (PSA) in men younger than 75 years, and
cancer has become the most common nonskin can- recommended against screening in men older than
cer in developed countries. In the United Kingdom, 75 years (Lin et al., 2011).
it is the most common cancer in men, representing Worldwide, an estimated 899,000 men were
24% of all new cancer cases. For example, 37,051 diagnosed with prostate cancer in 2008, and more
cases were registered in 2008 with a lifetime risk than two-thirds are diagnosed in developed coun-
of 1 in 9 (Office for National Statistics, Cancer tries (Ferlay et al., 2008). The highest rates are in
Statistics Registrations, UK, 2008) and this figure Australia/New Zealand, Western and Northern
increased to 40,331 in 2015 (Office for National Europe, and North America, largely because the
Statistics, Cancer Statistics Registrations, UK, prostate specific antigen (PSA) testing and sub-
2017). In the United States an estimated 218,890 sequent biopsy has become widespread in these
men were newly diagnosed with prostate cancer regions (Table 6.1).
in 2007 with a lifetime risk of 1 in 6 (National As can be seen in Table 6.1, the number of cases
Cancer Institute Surveillance Epidemiology and of prostate cancer reported to the World Health
End Results Program, Cancer Stat Facts: Cancer of Organization (WHO) has increased in all world
Prostate, 2006), however in 2016 the number of new regions.
cases registered had decreased to 180,890 (National Within the United States there are signifi-
Cancer Institute Surveillance Epidemiology and cant differences between racial groups, with the
End Results Program, Cancer Stat Facts: Cancer incidence of prostate cancer being 50% higher in
of Prostate, 2016). This decrease may be the result African Americans than for white Americans, while
of the United States Preventive Services Task Force rates for Asian Americans are 40% lower than for
(USPSTF) recommendations of 2012, whereby white Americans. The 2001–2005 age standardized
121
122  Precision medicine in prostate cancer

Table 6.1  Number of cases and deaths reported to the WHO by region for 2008 and 2012

Cases Deaths Cases Deaths

Estimated number 2008 2008 2012 2012
World 899,000 258,000 1,095,000 307,000
More developed regions 644,000 136,000 742,000 142,000
Less developed regions 255,000 121,000 353,000 165,000
WHO Africa region (AFRO) 34,000 24,000 52,000 37,000
WHO Americas region (PAHO) 334,000 76,000 413,000 85,000
WHO East Mediterranean region (EMRO) 12,000 9,000 19,000 12,000
WHO Europe region (EURO) 379,000 94,000 420,000 101,000
WHO Southeast Asia region (SEARO) 28,000 19,000 39,000 25,000
WHO Western Pacific region (WPRO) 109,000 33,000 153,000 46,000
IARC membership (22 countries) 611,000 128,000 791,000 157,000
United States of America 186,000 28,000 233,000 30,000
China 33,000 14,000 47,000 23,000
India 14,000 10,000 19,000 12,000
European Union 323,000 71,000 345,000 72,000

incidences were 249/100,000, 157/100,000, and 94/­ Statistics registrations, UK, 2010). In the United
100,000 for African Americans, white Americans, States the median age at diagnosis for prostate can-
and Asian Americans, respectively (Ries, 2005). cer is 67 years, with 0.6% of cases between 35 and 44
The risk of prostate cancer rises steeply with age, years, 9.1% between 45 and 54 years, 30.7% between
with the highest rates occurring in the 75–79 year 55 and 64 years, 35.3% between 65 and 74 years,
old age group. In the United Kingdom, the incidence 19.9% between 75 and 84 years, and 4.4% for 85-plus
is 155/100,000 men aged 55–59 years, 510/100,000 years of age (National Cancer Institute Surveillance
for the group 65–69 years, and 751/100,000 by Epidemiology and End Results Program, Cancer
75–79 years (Office for National Statistics, Cancer Stat Facts: Cancer of Prostate, 2006) (Figure 6.1).

Male rates Male cases

8,000 800
Average number of cases per year

6,000 600
Rate per 100,000

4,000 400

2,000 200

0 0
0 to 05 to 10 to 15 to 20 to 25 to 30 to 35 to 40 to 45 to 50 to 55 to 60 to 65 to 70 to 75 to 80 to 85+
04 09 14 19 24 29 34 39 44 49 54 59 64 69 74 79 84
Age at diagnosis

Figure 6.1  Age at diagnosis. (From Office for National Statistics, Cancer Statistics Registrations:
Registrations of cancer diagnosed in 2010, England. Series MB1 London; National Statistics.)
 The prostate-specific antigen (PSA) controversy for prostate cancer screening  123

However, studies published using postmortem (DRE) and a group of usual care that sometimes
data have shown that approximately half of all men included screening. In this control group, at the
in their 50s have histological evidence of prostate sixth year 52% of patients underwent screening
cancer, which in men over 80 rises to 80% (Sakr tests. They concluded that there was a 22% increase
et al., 1996; Burford et al., 2008), but only 1 in 26 in the rate of prostate cancer diagnosis in the
(3.8%) of these men will ultimately die from their screening group as compared to the control group;
cancer. In other words, more men will die with however there was no reduction in prostate cancer
their prostate cancer than from it; this is impor- mortality during the first seven years of the trial.
tant when considering population screening of In contrast the European study showed a reduction
asymptomatic men (Frankel et  al., 2003; Selley of 20% in the rate of death from prostate cancer
et al., 1997). among men between the ages of 55 and 69 years at
study entry (Schroder et al., 2009).
THE PROSTATE-SPECIFIC ANTIGEN Serum total PSA is prostate specific; how-
(PSA) CONTROVERSY FOR ever, it is increased in benign diseases, such as
PROSTATE CANCER SCREENING hyperplasia and prostatitis (Bozeman et al., 2002;
Punglia et al., 2006). As such, 10% to 20% of men
PSA is a glycoprotein produced almost exclusively aged between 50 and 70 years will have a raised
by the epithelial of the prostate gland. Serum lev- PSA. However, only 25% of those with a serum
els may be elevated due to increased PSA produc- total PSA of 4–10 ng/mL will be found to have
tion or architectural distortions in the prostate a biopsy positive for cancer (Smith et  al., 1997).
gland that allow greater PSA access to the circu- Furthermore the Prostate Cancer Prevention
lation. Only approximately 30% of men with a Trial (Thompson et al., 2003) reported that 39.2%
serum PSA >4.0 ng/mL, a standard cutoff point of men with a PSA of 2.1–3.0 ng/mL; 27.7% of
to determine the need for a prostate biopsy, have men with a PSA of 1.1–2.0 ng/mL; and 1.3% of
cancer confirmed on biopsy. Many false positives men with a PSA of <1.0 ng/mL had an end-of-
are attributed to benign prostatic hyperplasia or to trial prostate biopsy with foci of adenocarcinoma.
subclinical prostatic inflammation (Nadler et  al., In other words, 38% of men with prostate cancer
1995). It has been published that the cutoff value have a PSA <4.0 ng/mL and 70% of men with a PSA
of 4.0 ng/mL has a sensitivity of 46% with respect >4.0 ng/mL do not have cancer.
to the identification of cases of prostate cancer than In addition, the frequency of men having an ele-
would occur over a 10-year period. However, the vated PSA and benign biopsy is country dependent
specificity of 91% in this population with a mean (Belbase et al., 2013) and may be significantly dif-
age of 63 years fell to 54% in men over 80 years as a ferent between rural and metropolitan populations
result of the increase of benign prostatic hyperpla- in the same country (Lalitha et al., 2012).
sia (Gann et al., 1995). Moreover, two studies have Not all prostate cancers need treatment. It has
shown that biopsies taken from men with serum been estimated that of screen-detected prostate
PSA values of 2.5–4.0 ng/mL detected cancer in cancers, 23%–42% are overtreated (Draisma et al.,
12%–23% of cases (Babaian et  al., 2001; Catalona 2009). Active surveillance (AS) is a recognized ini-
et al., 1997). Thus there is no cutoff point for PSA tial treatment option for men with early stage low-
that determines if there is or is not prostate cancer. grade prostate cancer. The option to delay or avoid
To confound matters, the benefits of screening definitive therapy avoids or minimizes patient
are controversial. A screening test should ideally morbidity without compromising long-term out-
detect clinically significant cancer, which if not comes in appropriately selected patients (Dall’Era
treated would increase mortality and/or morbid- et al., 2008; Warlick et al., 2006). However, active
ity, and not detect those cancers that clinically surveillance requires repeat biopsies, often yearly,
would not cause harm to the patient. Two large- with the patient thus assuming the risks of biopsy.
scale studies designed to address this question Men with clinically insignificant prostate cancers
from the United States and Europe reported con- that were never destined to have symptoms or to
flicting results. The United States study reported by affect their life expectancy may not benefit from
Andriole et al. (2009), compared a screening group knowing that they have the “disease.” The detec-
with annual serum and digital rectal examination tion of clinically insignificant prostate cancer could
124  Precision medicine in prostate cancer

be considered as an adverse effect of the prostate RARE HIGH PENETRANCE

biopsy. As such, there is considerable anxiety and PROSTATE CANCER SUSCEPTIBILITY
distress found in men undergoing active surveil- GENES
lance (van den Bergh et al., 2009).
Thus for every 100 men with an elevated PSA To identify chromosomal regions that may contain
between 4–10  ng/mL, only approximately 14 major prostate cancer susceptibility genes, linkage
will have a clinically significant prostate cancer studies are used in families with multiple affected
detected, or 86 men underwent a biopsy with the members. In most inherited cancer syndromes,
associated risks for a benign disease. A prostate cases tend to be diagnosed significantly earlier
biopsy is not without risks; infection and hem- than sporadic cases. In prostate cancer the results
orrhage being the main potentially serious side from segregation analysis suggests that heredi-
effects, with a 30-day complication rate of 3.7%, tary cancers are diagnosed at an earlier age, some
especially in older patients (Anastasiadis et  al., 6–8 years younger than in sporadic cases (Carter
2014). Thus avoiding unnecessary biopsies is a et al., 1992). Linkage studies are hindered by sev-
worthwhile aim if the number of clinically signifi- eral related problems, the first being a high lifetime
cant cancers detected is not prejudiced. incidence of prostate cancer (approximately 15%)
Therefore, any new screening test developed to and second an estimated 90% of cases are sporadic
detect prostate cancer has to improve on the stan- in nature. This problem is especially relevant with
dard serum PSA test; in other words detect clini- the advent of serum PSA screening in which the
cally significant prostate cancer and not indolent incidence rates have increased more than threefold
cancer. as a result of increased detection.
Polymorphic markers (microsatellite repeats or
GENETIC SCREENING AND single nucleotide polymorphisms [SNPs]) across
PERSONALIZED MEDICINE the genome are genotyped in family members, and
their cosegregation (linkage) with the presence of
That there is a genetic component to prostate cancer prostate cancer is tested using a logarithm of odds
risk has been shown by a family history of prostate score (LOD). A score of over 3 is considered as
cancer. Evidence for this genetic link was reported evidence that the linkage is significant. With the
in the 1960s (Woolf, 1960). Later, two large meta- chromosome region identified, sequencing meth-
analyses of studies published between 1966 and ods are used to pinpoint the susceptibility gene.
2002 (Goh et al., 2012; Johns and Houlson, 2003) In the early years of this century, three genes
concluded that a positive family history was signif- were identified: Hereditary Prostate Cancer 1
icantly associated with an increased risk of devel- (HCP1) on chromosome 1q24 was reported from
oping prostate cancer. The estimated relative risk families in the United States and Sweden (Carter
varied between studies, being between 1.93 and et al., 1992). The RNASEL gene in this region was
2.50. This risk is higher in men with a positive fam- identified in 2002 as a susceptibility gene using a
ily history in first-degree relatives than second- combination of fine mapping and direct sequenc-
degree relatives, and is higher still in those with ing (Carpten et  al., 2002). RNASEL is a constitu-
affected brothers rather than fathers, and in those tively expressed latent endonuclease, of which
whose relatives were diagnosed with prostate can- two ­ mutations—Met1Ile and Glu265X—have
cer before the age of 60 years and decreased with been reported to segregate with prostate cancer
age. From monozygotic and dizygotic twin studies (Carpten et al., 2002). The Glu265X mutation has
(Ahlbom et al., 1997; Page et al., 1997), the concor- been reported to be associated with prostate can-
dance rates varied from 20%–27% for monozygotic cer in both familial and sporadic forms of prostate
twins to between 4%–7% for dizygotic twins. cancer (Rokman et  al., 2002). One of these vari-
Thus men with a positive family history of pros- ants, the Arg462Gln, has been found in up to 13%
tate cancer have a 1.5- to 2.5-fold increased risk of of prostate cancer cases in a Finnish family–based
developing prostate cancer (Johns and Houlson, case-control study but no association in patients
2003); however, less than 10% of men with prostate without a family history (Casey et al., 2002). This
cancer have a family history (Goh et al., 2012). variant has shown a strong association with disease
 Rare high penetrance prostate cancer susceptibility genes  125

severity; however, this effect is modified by family A third gene was identified in the region p22–23
history and racial group. In European Americans of chromosome 8, MSR1 (Xu et al., 2012), and is a
without a family history, the allele is associated member of a family of membrane receptors called
with early stage low-grade disease, differing from scavenger receptors. MSR1 can bind to molecules
those patients with a family history where the allele ranging from bacteria to modified lipoproteins
was associated with aggressive disease (Rennert (Platt and Gordon, 2001) and is reported to be com-
et al., 2005). Inversely Wang et al. (2002) found an monly deleted in prostate cancer (Lieberfarb et al.,
inverse relationship with familial prostate cancer, 2003). In families with hereditary prostate can-
associated with less aggressive cancer in younger cer six missense and one nonsense mutation have
patients and no association with sporadic cancer. been described, and the prevalence of MSR1 muta-
In 2001, using positional cloning and muta- tions is higher in European and African-American
tion screening of the chromosome 17p ELAC2 was members than in case controls (Xu et  al., 2002a,
identified in Utah families with hereditary prostate 2002b). The most common mutations found in
cancer (Tavtigian et  al., 2001). A possible role in these men were the Arg293X and Ser41Tyr variants
the control of the cell cycle has been hypothesized (Xu et  al., 2002a). The IVS7delTTA homozygous
as it encodes for a 3′ processing endoribonuclease, genotype is significantly associated with low-grade
which interacts with gamma-tubulin, a component sporadic-type prostate cancer, which appears at
of the mitotic cycle (Korver et  al., 2003; Takaku a later age in European racial groups. This differs
et  al., 2003). However, mutations in the gene are from the Arg293X variant, which is associated with
rare, but sequence analysis has identified two mis- high-grade sporadic disease in men younger than
sense changes—Ser217-Leu and Ala541Thr—that 60  years. Rennert et  al. (2005) found no signifi-
have been reported to be associated with prostate cant association of MSR1 variants with familial or
cancer (Rebbeck et al., 2000; Tavtigian et al., 2001). sporadic prostate cancer in European or African-
More recently however, there was no reported American men. MSR1 mutations were found in
association between these two variants and pros- 4.4% of white men with nonhereditary prostate
tate cancer (Rennert et  al., 2005). Furthermore cancer as opposed to 0.8% of unaffected men. In
three meta-analyses showed conflicting results. Afro-American men these frequencies were 12.5%
Severi et al. (2003) found no association with pros- and 1.8%, respectively (Xu et al., 2002a, 2002b).
tate cancer risk in a study of 1,557 patients and While the evidence for these three genes is
data obtained from seven previous studies. Using conflicting, the more recent discovery of the pros-
new data and data from seven previously published tate cancer susceptibility gene HOXB13 seems to
reports, Meitz et al. (2002) showed only a moderate be more clear-cut. More than 200 genes from the
risk with an odds ratio of 1.27. In Canadian patients chromosomal region 17q21–22 were sequenced
the number of men homozygotes for Ser217-Leu from families with hereditary prostate cancer.
did not differ between cases and controls (8.6% Probands from four families were found to have a
versus 8.5%), although heterozygotic expression rare but recurrent mutation (G84E) in HOXB13, a
was found in 61.8% of cancer cases versus 50.3% of homeobox transcription factor important in pros-
controls and was more common in men with pros- tate development (Ewing et al., 2012). The mutation
tatic intraepithelial neoplasia (PIN), 42.3% versus was found in all men with prostate cancer within
26.7% (Alder et al., 2003). In Ala541Thr heterozy- these four families. In population studies, the het-
gotes there appeared to be an increase in late onset erozygous carrier state in sporadic prostate cancer
prostate cancer and PIN (Alder et  al., 2003). In a is increased by a factor of 20. The G84E mutation
larger study of U.S. males, the Ser217-Leu muta- in a prescreened white Canadian population was
tion was found in 32% of cases and 29% of controls, more frequent in men with a prostate biopsy posi-
whereas the Ala541Thr variant was found in 4% of tive for cancer, 0.7% versus those with a biopsy
both cases and controls (Stanford et al., 2003). They negative for cancer 0.1% (Akbari et al., 2012). This
reported that the heterozygote state was associated signifies that the mutation cosegregates with pros-
with less aggressive localized cancer, whereas the tate cancer in hereditary prostate cancer fami-
homozygote state was associated with Gleason lies and is associated with prostate cancer risk in
scores ≥7. unrelated cases and controls. In the Reduction by
126  Precision medicine in prostate cancer

Dutasteride of Prostate Cancer Events (REDUCE) era, the trigger for prostate biopsy is a total serum
Trial, all 3508 men had an initial negative prostate PSA of 4–10 ng/mL. This Danish study reported
biopsy and were biopsied after 2 and 4 years of an association with higher-grade prostate cancer
treatment or placebo. The G84E mutation was only defined as a Gleason score of ≥7 (83.3% versus
detected in Caucasians, with the highest frequency 60.9%) and positive surgical margins 56.0% versus
in Northern Europe (1.06%), followed by Western 28.5%. Risk allele carriers were also more likely
Europe (0.60%) and North America (0.31%). No to have aggressive disease, defined as a preopera-
mutation carrier was observed in Southern or tive total PSA ≥20 ng/mL, a Gleason score of ≥7,
Eastern Europe, Latin America, Australia, and and/or the presence of regional or distant disease.
South Africa, which highlights the importance of However, there was no significant association
differences in the genetic load of differing popu- with biochemical failure after primary treatment
lations. In Caucasians the detected mutation fre- (Storebjerg et al., 2016).
quency was 0.99% and 0.24% in men with a biopsy
positive and negative for cancer, respectively. In COMMON AND LOW-PENETRANCE
those men with a biopsy positive for cancer, the RISK-ASSOCIATED SINGLE
detection frequency was higher in those with a NUCLEOTIDE POLYMORPHISMS
family history of prostate cancer, 4.31% versus
0.34% in those without a family history. Genetic association studies are designed to iden-
After 4 years of follow-up the prostate cancer tify common and low-penetrance genes in the
detection rate was 53.8% in the 13 men heterozy- general population, comparing the allele/genotype
gotic for G84E and 22% among the 3186 noncar- frequencies of markers, typically single nucleotide
riers for a relative risk of 2.45 (Chen et al., 2013). polymorphisms (SNPs), between healthy controls
A Swedish-based study of 4693 controls and 5003 and cancer patients. With the development of high
prostate cancer cases reported that the G84E muta- throughput and low-cost genotyping arrays it has
tion was present in 1.3% of the population controls become feasible to systematically screen hundreds
and strongly associated with prostate cancer risk of SNPs in the genome for their association with
with an odds ratio of 3.4. The strongest associated disease risk without the need to limit the search
was reported for young onset prostate cancer (odds to specific genes or chromosomal regions. The
ratio 8.6) and hereditary prostate cancer (odds Genome-Wide Association Study (GWAS) (http://
ratio 6.6). As in the REDUCE study, haplotype www.genome.gov/gwastudies/) has identified
analysis supported the idea that the G84E muta- thousands of SNPs in independent population
tion is a founder mutation. Carriers for the muta- groups, which are consistently associated with the
tion had an estimated cumulative risk of 33% of risk for many complex diseases including pros-
developing prostate cancer up to the age of 80 years tate cancer. Many of these SNPs are not located in
as compared to 12% in noncarriers (Karlsson et al., apparent candidate genes or pathways and others
2014). However, although carriers of this mutation are not within genes (Hindorff et al., 2009). Since
were younger at the time of diagnosis and more 2007 when the first two GWASs of prostate cancer
likely to have a family history of prostate cancer, were reported, more than 50 prostate cancer risk
there was no association with the Gleason score SNPs have been consistently associated with pros-
in the surgical piece or pathological stage of the tate cancer risk in Caucasians, African Americans,
cancer (Beebe-Dimmer et  al., 2014). In Danish Japanese, and Chinese men from the GWAS and
men undergoing radical prostatectomy, 2.51% fine mapping of implicated regions. Table 6.2 shows
were positive for G84E while in a healthy control the SNPs associated with prostate cancer risk.
population the mutation frequency was 0.49%. From these studies several important observa-
Differing from the study of Beebe-Dimmer et  al. tions can be noted, such as, most of these mark-
(2014), the authors found that carriers were more ers can be consistently reproduced in independent
likely to have a higher serum PSA level at diagno- study populations, and few of the SNPs were found
sis, mean PSA of 19.9 ng/mL versus 13.6 ng/mL. In in well-known candidate genes and pathways and
this context it must be mentioned that the study many are in intergenic regions. Whereas some SNPs
was conducted between 1997 and 2011. In men indicate an increased risk for prostate cancer in
with prostate cancer detected in the PSA screening multiple races, some are race specific. The majority
Table 6.2  SNPs associated with prostate cancer risk

Location Risk
SNP Cytoband (build36) allele OR (95%CI) Population Reference # Cases/controls
rs10187424 2p11 85.647.808 A 1.19(1.09–1.32) Caucasian Kote-Jarai et al. (2011) 37,250/36,359
rs721048 2p15 62.985.235 A 1.15(1.10–1.21) Caucasian Gudmundsson et al. (2008) 10,054/28,879
rs1465618 2p21 43.407.453 A 1.27(1.14–1.43) Caucasian Eeles et al. (2009) 21,733/20,655
rs13385191 2p24 20.751.746 G 1.15(1.10–1.21) Japanese Takata et al. (2010) 4,584/8,801
rs12621278 2q31 173.019.799 A 1.35(1.10–1.64) Caucasian Eeles et al. (2009) 21,733/20,655
rs2292884 2q37.3 238.107.965 G 1.14(1.09–1.19) Caucasian Schumacher et al. (2011) 10,140/11,190
rs2055109 3p11.2 87.550.022 C 1.20(1.13–1.29) Japanese Akamatsu et al. (2012a,b) 7,141/11,804
rs2660753 3p12 87.193.364 T 1.18(1.06–1.31) Caucasian Eeles et al. (2008) 5,122/5,260
rs10934853 3q21 129.521.063 A 1.12(1.08–1.16) Caucasian Gudmundsson et al. (2009) 13,774/47,614
rs6763931 3q23 142.585.523 T 1.18(1.07–1.29) Caucasian Kote-Jarai et al. (2011) 37,250/36,359
rs10936632 3q26 171.612.796 A 1.14(1.02–1.25) Caucasian Kote-Jarai et al. (2011) 37,250/36,359
rs17021918 4q22 95.781.900 C 1.19(1.08–1.32) Caucasian Eeles et al. (2009) 21,733/20,655
rs7679673 4q24 106.280.983 C 1.19(1.14–1.37) Caucasian Eeles et al. (2009) 21,733/20,655
rs2121875 5p12 44.401.302 G 1.09(0.99–1.21) Caucasian Kote-Jarai et al. (2011) 37,250/36,359
rs12653946 5p15 1.948.829 T 1.26(1.20–1.33) Japanese Takata et al. (2010) 4,584/8,801
rs130067 6p21 31.226.490 G 1.20(1.07–1.34) Caucasian Kote-Jarai et al. (2011) 37,250/36,359
rs1983891 6p21 41.644.405 T 1.15(1.09–1.21) Japanese Takata et al. (2010) 4,584/8,801
rs339331 6q22 117.316.745 T 1.22(1.15–1.28) Japanese Takata et al. (2010) 4,584/8,801
rs9364554 6q25 160.753.654 T 1.17(1.08–1.26) Caucasian Eeles et al. (2008) 5,122/5,260
rs10486567 7p15 27.943.088 G 1.19(0.95–1.49) Caucasian Thomas et al. (2008) 5,113/5,121
rs6465657 7q21 97.654.263 C 1.12(1.05–1.20) Caucasian Eeles et al. (2008) 5,122/5,260
rs2928679 8p21 23.494.920 A 1.53(1.27–1.83) Caucasian Eeles et al. (2009) 21,733/20,655
rs1512268 8p21 23.582.408 A 1.23(1.12–1.35) Caucasian Eeles et al. (2009) 21,733/20,655
rs1447295 8q24 (Region 1) 128.554.220 A 2.23(1.58–3.14) Caucasian Yeager et al. (2007) 4,296/4,299
rs16901979 8q24 (Region 2) 128.194.098 A 1.79(1.53–2.11) Caucasian Gudmundsson et al. (2007) 2,663/5,509
rs6983267 8q24 (Region 3) 128.482.487 G 1.58(1.40–1.78) Caucasian Yeager et al. (2007) 4,296/4,299
rs16902094 8q24 (Region 4) 128.389.528 G 1.21(1.15–1.26) Caucasian Gudmundsson et al. (2009) 13,774/47,614
(Continued )
Common and low-penetrance risk-associated single nucleotide polymorphisms  127
Table 6.2 (Continued)  SNPs associated with prostate cancer risk

Location Risk
SNP Cytoband (build36) allele OR (95%CI) Population Reference # Cases/controls
rs620861 8q24 (Region 4) 128.404.855 C 1.28(1.16–1.43) Caucasian Al Olama et al. (2013) 5,504/5,834
rs10086908 8q24 (Region 5) 128.081.119 T 1.25(1.12–1.37) Caucasian Al Olama et al. (2009) 5,504/5,834
rs817826 9q31.2 109.196.121 C 1.41(1.29–1.54) Chinese Xu et al. (2012) 4,484/8,934
rs1571801 9q33 123.467.194 T 1.36(1.13–1.63) Caucasian Duggan et al. (2007) 1,235/1,599
rs10993994 10q11 51.219.502 T 1.57(1.36–1.81) Caucasian Thomas et al. (2008) 5,113/5,121
rs2252004 10q26 122.834.699 G 1.16(1.10–1.22) Japanese Akamatsu et al. (2012a,b) 7,141/11,804
rs4962416 10q26 126.686.862 C 1.46(1.22–1.76) Caucasian Thomas et al. (2008) 5,113/5,121
rs7127900 11p15 2.190.150 A 1.28(1.14–1.44) Caucasian Eeles et al. (2009) 21,733/20,655
128  Precision medicine in prostate cancer

rs1938781 11q12 58.671.686 C 1.16(1.11–1.21) Japanese Akamatsu et al. (2012a,b) 7,141/11,804

rs12418451 11q13 68.691.995 A 1.36(1.18–1.56) Caucasian Zheng et al. (2008) 7,012/4,775
rs10896449 11q13 68.751.243 G 1.41(1.22–1.61) Caucasian Thomas et al. (2008) 5,113/5,121
rs10875943 12q13 47.962.277 C 1.18(1.06–1.31) Caucasian Kote-Jarai et al. (2011) 37,250/36,359
rs902774 12q13 51.560.171 A 1.17(1.11–1.24) Caucasian Schumacher et al. (2011) 10,140/11,190
rs9600079 13q22 72.626.140 T 1.18(1.12–1.24) Japanese Takata et al. (2010) 4,584/8,801
rs11649743 17q12 33.149.092 G 1.50(1.26–1.77) Caucasian Sun et al. (2009) 9,626/7,337
rs4430796 17q12 33.172.153 A 1.22(1.15–1.30) Caucasian Gudmundsson et al. (2007) 3,490/14,345
rs7210100 17q21.32 44.791.748 A 1.51(1.35–1.69) African American Haiman et al. (2011) 5,262/6554
rs1859962 17q24 66.620.348 G 1.20(1.14–1.27) Caucasian Gudmundsson et al. (2007) 3,490/14,345
rs8102476 19q13 43.427.453 C 1.12(1.08–1.15) Caucasian Gudmundsson et al. (2009) 13,774/47,614
rs887391 19q13 46.677.464 T 1.15(1.09–1.21) Caucasian Hsu et al. (2009) 9,516/7,252
rs2735839 19q13 56.056.435 G 1.20(1.10–1.33) Caucasian Eeles et al. (2008) 5,122/5,260
rs103294 19q13.4 59.489.660 C 1.28(1.21–1.36) Chinese Xu et al. (2012) 4,484/8,934
rs9623117 22q13 38.782.065 C 1.18(1.11–1.26) Caucasian Sun et al. (2009) 9,626/7,337
rs5759167 22q13 41.830.156 G 1.14(1.04–1.25) Caucasian Eeles et al. (2009) 21,733/20,655
rs5945619 Xp11 51.258.412 C 1.19(1.07–1.31) Caucasian Eeles et al. (2008), 5,122/5,260
Gudmundsson et al. (2008) (10,054/28,879)
rs5919432 Xq12 66.938.275 A 1.09(1.00–1.20) Caucasian Kote-Jarai et al. (2011) 37,250/36,359
 Clinical application of SNPs in prostate cancer  129

of these SNPs are present in the general popula- cancer was predicted by the number of genetic
tion with a frequency of 5% or higher, conferring variants (Kashyap et  al., 2014). Of the 4548 men
a modest risk increase, commonly a relative risk of who underwent biopsy, 1834 (40.3%) had prostate
1.1–1.2. However, when considered together they cancer detected. In men with no variant pres-
confer a stronger genetic risk for the development ent, the detection rate was 29% increasing to 63%
of prostate cancer. With regard to ethnic popula- in men with at least seven risk alleles. To assess
tions and potential race-specific differences, Han genetic risk, a population standardized GRS has
et  al. (2015) reported that the majority of GWAS been developed. This score assesses prostate can-
identified loci harbor risk alleles that are common cer risk using both the odds ratio and allele fre-
and shared across populations. To support this quency of each SNP to assign a risk score, where
concept, the predictive value of 105 known prostate 1.0 represents the average population risk, <1.0
cancer risk SNPs were used to predict prostate can- a lower risk, and >1.0 a higher risk. Studies have
cer. The genetic risk score (GRS) performed better classified patients into three risk groups based on
in the Caucasian and Latin groups than in African- the GRS: <0.5 as low risk, 0.5–1.5 as intermediate
Americans and those from East Asia (Hoffmann risk, and >1.5 as high risk. The first prostate can-
et al., 2015). It has been further suggested that race-/ cer risk–associated SNPs identified were associated
ethnic-specific GRS should be used (Na et al., 2013). with a relatively high ability to predict the risk of
Caution is further warranted. Klein et al. (2012), prostate cancer, with estimated odds ratios rang-
in a study of 891 cases and 2521 controls, were ing from 1.20 to 1.79. With increasing information,
unable to replicate many SNPs and found a signifi- the predictive effect of newly determined SNPs is
cant association with prostate cancer at only 14 of much lower when compared with the original dis-
37 SNPs, with a poor predictive value for the detec- coveries, with odds ratios in the order of 1.06–1.15.
tion of prostate cancer. They also commented that Ren et  al. (2013) described a plateau effect, and
only 25%–60% of reported SNPs were associated that incorporating new SNPs had little additional
with prostate cancer risk and some associations predictive value. However, comparing SNPs used
were weak, especially with regard to aggressive dis- in 2007 with those used in 2013, Krier et al. (2016)
ease (Klein et al., 2012). reported a 50% change in classification of patients.
As such there is no consensus at present of how
CLINICAL APPLICATION OF SNPs many SNPs should be incorporated into the GRS.
IN PROSTATE CANCER To date, a family history has been used as a risk
factor for prostate cancer detection. However, the
Current screening methods, digital rectal exami- family history may change with time and is strongly
nation, total PSA, free percent PSA, PSA density, dependent on other factors such as family size, age
Prostate Health Index, and differing nomograms of relatives, health care access within the family,
have been used to identify men with possible pros- and the level of family communication. This is seen
tate cancer. These tests are used to select men to in that family history is a relatively poor indicator
undergo a prostate biopsy, which is the gold stan- of prostate cancer risk as measured by area under
dard for the diagnosis of prostate cancer. As previ- the receiver operating curve (AUC), reported as
ously mentioned, only about one third of men who being as low as 0.52 (little more than tossing a
undergo a prostate cancer biopsy will have cancer coin) (Liss et al., 2015). For the GRS to be clinically
diagnosed and not all will need treatment. useful it must provide additional information that
In this context the use of precision medicine cannot be assessed using currently available meth-
has to be applied, in which the additive effect of ods, must be applicable to all men, and outperform
numerous genes along with environmental deter- current methods of risk assessment.
minants create a normal distribution of disease In the Prostate Cancer Prevention Trial both
risk in the general population. The risk alleles family history and GRS were used to identify men at
identified by GWAS studies individually confer a higher risk for prostate cancer and the results of the
modest increase in the risk of prostate cancer. In prostate biopsy. Of the participating men, 17% had
men undergoing prostate biopsy for an elevated a positive family history, of whom 29% had prostate
total PSA, defined as >4.0 ng/mL, or an abnormal cancer diagnosed at biopsy. This was significantly
digital rectal examination, the presence of prostate higher than in men without a family history, in
130  Precision medicine in prostate cancer

whom 23.4% had a cancer detected. When a posi- prostate cancer risk. However, it has been shown
tive family and/or a GRS of >1.4 were used as selec- that some pesticides modify the association
tion criteria, 36% of men were classified as high risk between genetic variants on chromosome 8q24
and the detection of prostate cancer was reported and prostate cancer risk (Koutros et  al., 2010).
to be 30.98%. The authors also reported that when Obesity has been linked to prostate cancer, both
the GRS plus family history were used nearly 50% for increased risk of developing prostate cancer
of cases with prostate cancer would be detected, and a worse prognosis. Two recent studies com-
including 45% of high-grade cancers (Chen 2016). bined body mass index with the GRS in Chinese
In the REDUCE study (Kader et al., 2012), follow- patients. Zhang et  al. (2015) reported that the
ing an initial negative biopsy (which is not a general predictive value of body mass index was strongly
screening population) the AUC increased from 0.52 modified by the GRS. In those patients defined
(family history alone) to 0.59 (GRS). Although sta- as high GRS (>1.4), the body mass index was an
tistically significant, both AUCs are rated as poor independent predictor of prostate cancer, whereas
in terms of predictive value. A model combining in those with a low GRS (<0.5) the body mass
clinical parameters had an AUC of 0.64, which index had no predictive value. Liu et  al. (2016)
increased to 0.67 when combined with the GRS, reported that in men with a body mass index of
both prediction values being considered as accept- >28 kg/m2 the presence of the alleles rs6983561 CC
able. In the REDUCE trial the authors concluded and rs16901966 GG increased the odds ratio to 7.66
that among men classified as low or high risk by and 5.33, respectively. In smokers, the presence of
clinical parameters, the GRS had limited clinical the alleles rs7679673 CC + CA and rs12653946 TT
impact in men with a previous negative for cancer increased the odds ratio to 2.77 and 3.11, respec-
prostate biopsy. In those with an intermediate risk, tively, whereas in men who consumed alcohol the
defined as 23% for any prostate cancer and 6% for presence of the allele rs7679673 increased the odds
high-risk prostate cancer, the addition of the GRS ratio to 4.37. However, apart from a healthy diet
had the following affect: In men with a low GRS the and avoiding being overweight, not smoking, lim-
detection rate of prostate cancer was 14% with 4% iting alcohol consumption, and the general advice
being high-grade cancers. Those men with a high of healthy living, there is no evidence at present of
GRS had a detection rate of 35% for any prostate the direct effect on prostate cancer risk.
cancer and 14% had a high-grade cancer.
In terms of predicting prostate cancer, the SNPs CONCLUSIONS
have a poor predictive value on their own with
AUC of between 0.57 and 0.64 (Aly et al., 2011; Hsu The idea that genetic mapping could improve the
et al., 2010; Klein et al., 2012). Nam et al. (2009) and identification of men destined to develop prostate
Johansson et  al. (2013) found only small increases cancer would appear to be the hallmark of preci-
in the AUC when incorporating SNPs in a general sion medicine. However, in a cancer where the
screening population. Using the GRS the AUC was familial or genetic component is of the order of
0.64, whereas for total PSA and percent free PSA the 10%, it would not be surprising that as a popula-
AUC was 0.86. Combining the GRS with serum PSA tion screening test it may not be ideal. To date, the
produced an AUC difference that was statistically use of GRS based on SNP variants has had limited
significant; the combined AUC was only 0.87 and success on an individual basis. The genetic profile
thus not clinically significant. More recently Gilbert cannot be used to recommend prophylactic treat-
et  al. (2015) reported that incorporating known ment nor a prostate biopsy. In the absence of life-
genetic variants did not improve the accuracy of style interventions or chemoprevention for prostate
PSA testing to predict prostate cancer at biopsy or cancer, the aim of population health care is early
to identify patients with high-risk prostate cancer. detection and adequate treatment. Genetic profil-
ing to date has not found its place in this aspect of
ENVIRONMENTAL EFFECTS ON cancer detection.
SNPs In the clinical situation, an ideal biomarker would
identify those men who would need a prostate biopsy
In prostate cancer, there is little evidence to link to confirm a clinically significant prostate cancer
environmental factors on genetic variants and that would need treatment and thus decrease the
 References 131

mortality or morbidity of the disease. It would have ACKNOWLEDGMENTS

to be cost effective and in the real world not require
expensive setup costs in terms of equipment and I would like to thank Mrs. Ana Maria Palazuelos
highly trained staff. Thus, it could be implemented for her help in the writing of this manuscript.
in a general district hospital of any country.
For low-frequency, high-penetrance genes, in CONFLICTS OF INTEREST
terms of population screening the cost-benefit
ratio does not warrant general testing. However, There were no conflicts of interest.
in high-risk families, defining the genetic marker
may be beneficial to identify individual family REFERENCES
members who may benefit from PSA testing.
With common low-penetrance SNPs, the fun- Ahlbom A, Lichtenstein P, Malmstrom H et al.,
damental question is how to use the information, Cancer in twins: Genetic and non-genetic
and to differentiate between statistical significance familial risk factors. J Natl Cancer Inst
and clinical significance. SNP testing does not 89;1997:287–93.
determine if a patient has prostate cancer, nor does Akamatsu S, Takata R, Haiman CA et al.,
it signify the need for a prostate biopsy. The GRS is Common variants at 11q12, 10q26 and 3p11.2
not intended to be used as a diagnostic marker for are associated with prostate cancer suscepti-
discriminating cancer versus noncancer patients. bility in Japanese. Nat Genet 44;2012a:426–9.
The GRS should be used as a measurement of Akamatsu S, Takahashi A, Takata R et al.,
genetic susceptibility, similar to the family history. Reproducibility, performance and clinical utility
The family history does not require a laboratory of a genetic risk prediction model for prostate
test, but neither does it provide a direct measure- cancer in Japanese. PLOS ONE 2012b;e46454.
ment of the patient’s inherited risk. A genetic score Akbari MR, Trachtenberg J, Lee J et al.,
of 1.0 indicates an average risk of developing pros- Association between germline HOXB13 G84E
tate cancer and is a continuous variable. In the U.S. mutation and risk of prostate cancer. J Natl
population, 8% have a twofold higher risk and 2% a Cancer Inst 104;2012:1260–2.
threefold higher risk of developing prostate cancer Alder D, Kanji N, Trpkov K et al., HPC2/ELAC2
based on SNP analysis. gene variants associated with incident pros-
However, the GRS could be used to stratify men tate cancer. J Hum Genet 48;2003:634–8.
into low-, intermediate-, and high-risk groups. Al Olama A, Kote-Jarai Z, Scummacher FR et al.,
This would then determine at what age PSA screen- A meta-analysis of GWAS to identify pros-
ing would be more appropriate. Men with a lower tate cancer susceptibility loci associated with
GRS risk group could be screened less frequently aggressive and non-aggressive disease. Hum
or screening started at an older age. Mol Genet 22;2013:408–15.
A concern over SNP profiling is that against Aly M, Wiklund F, Xu J et al., Polygenic risk score
environmental or lifestyle risk factors it may have improves prostate cancer risk prediction:
little influence on the development of a cancer, or Results from the Stockholm-1 cohort study.
if the modification of environmental or lifestyle Eur Urol 60;2011:21–8.
factors can change the risk in developing a cancer. Anastasiadis E, van der Meulin J, Emberton M.
The presence of a genetic variant does not signify Hospital admissions after TRUS biopsy of
it will be expressed, nor is there evidence on how the prostate in men diagnosed with prostate
other modifiable risk factors may influence its cancer: A database analysis in England. Int J
expression. Urol 22;2014:181–6.
Furthermore, in identifying patients at risk of Andriole GL, Grubb RL, Buys SS et al., Mortality
aggressive prostate cancer, the use of SNP profiling results from a randomized prostate cancer
has not been shown to improve predictive values. screening trial. N Eng J Med 360;2009:1310–9.
In terms of practical prostate cancer screening, Babaian RJ, Johnston DA, Naccarato W et al.,
the use of GRS may be statistically significant, but The incidence of prostate cancer in a screen-
in the day-to-day clinical management of patients ing population with a serum PSA between 2.5
it has not proved its potential. and 4.0 ng/ml. J Urol 165;2001:757–60.
132  Precision medicine in prostate cancer

Beebe-Dimmer JL, Isaacs WB, Zuhlke KA et al., Draisma G, Etzioni R, Tsodikov A et al., Lead
Prevalence of the HOXB13 G84E prostate time and overdiagnosis in PSA screening:
cancer risk allele in men treated with radical Importance of methods and context. J Natl
prostatectomy. BJU Int 113;2014:830–5. Cancer Inst 101;2009:374–88.
Belbase NP, Agrawal CS, Pokharel PK et al., Duggan D, Zheng SL, Knowlton M et al., Two
Prostate cancer screening in a healthy popula- GWAS of aggressive prostate cancer impli-
tion cohort in Eastern Nepal: An explana- cate putative prostate suppressor gene
tory trial study. Asian Pac J Cancer Prev DAB2IP. J Natl Cancer Inst 99;2007:1836–44.
14;2013:2835–8. Eeles RA, Kote-Jarai Z, Al Olama AA et al.,
Bozeman CB, Carver BS, Eastham JA, Identification of seven new prostate cancer
Venable DD. Treatment of chronic susceptibility loci through a GWAS. Nat
­prostatitis lowers serum PSA. J Urol Genet 41;2009:1116–21.
167;2002:1723–6. Eeles RA, Kote-Jarai Z, Giles GG et al., Multiple
Burford DC, Kirby M, Austoker J. Prostate Cancer newly identified loci associated with pros-
Risk Management Programme information tate cancer susceptibility. Nat Genet
for Primary Care; PSA testing for asymp- 40;2008:316–21.
tomatic men. 2008, NHS Cancer Screening Ewing CM, Ray AM, Lange EM et al., Germline
Programmes: Sheffield. mutations in HOXB13 and prostate cancer
Carpten J, Nupponen N, Issacs S et al., Germline risk. N Eng J Med 366;2012:141–9.
mutations in the ribonuclease L gene in fami- Ferlay J, Shin HR, Bray F et al., GLOBOCAN 2008
lies showing linkage with HPC-1. Nat Genet v1.2. Cancer incidence and mortality world-
30;2002:181–4. wide. Lyon, France: International Agency for
Carter BS, Beaty TS, Steinberg GD et al., Research on Cancer, 2010. http://globocan.
Mendelian inheritance of familial pros- iarc.fr.
tate cancer. Proc Natl Acad Sci USA Frankel S, Smith GD, Donovan J, Neal D et al.,
89;1992:3367–71. Screening for prostate cancer. Lancet
Casey G, Neville PJ, Plummer SJ et al., 361(9363);2003:1122–8.
Arg462Gln variant is implicated in up to Gann PH, Hennekens CH, Stampfer MJ. A pro-
13% of prostate cancer cases. Nat Genet spective evaluation of plasma PSA for detec-
32;2002:581–3. tion of prostate cancer. JAMA 273;1995:289–94.
Catalona WJ, Smith DS, Ornstein DK. Gilbert R, Martin RM, Evans DM et al.,
Prostate cancer detection in men with serum Incorporating known genetic variants does not
PSA concentrations of 2.6–4.0 ng/ml and improve the accuracy of PSA testing to identify
benign prostate examination: Enhancement high risk prostate cancer on biopsy. PLOS ONE
with free PSA measurements. JAMA 2015; doi: 10.1371/journal.pone.0136735.
277;1997:1452–5. Goh CI, Schumacher FR, Easton D et al., Genetic
Chen H, Liu X, Brendier CB et al., Adding variants associated with predisposition to
genetic risk score to family history prostate cancer and potential clinical implica-
identifies twice as many high risk men tions. J Intern Med 271;2012:353–65.
for prostate cancer. Results from the Gudmundsson J, Sulem P, Gudbjartsson DF
Prostate Cancer Prevention Trial. Prostate et al., GWA and replication studies identify
76;2016:1120–9. four variants associated with prostate cancer
Chen Z, Greenwood C, Isaacs WB et al., The susceptibility. Nat Genet 41;2009:1122–6.
G84E mutation of HOXB13 is associated with Gudmundsson J, Sulem P, Manolescu A et al.,
increased risk for prostate cancer: Results GWAS identifies a second prostate cancer
from the REDUCE trial. Carcinogenesis susceptibility variant at 8q24. Nat Genet
34;2013:1260–4. 39;2007:631–7.
Dall’Era MA, Cooperberg MR, Chan JM et al., Gudmundsson J, Sulem P, Rafnar T et al.,
2008. Active surveillance for early stage pros- Common sequence variants on 2p15 and
tate cancer: Review of the current literature. Xp11.22 confer susceptibility to prostate can-
Cancer 112;2008:1650–9. cer. Nat Genet 40;2008:281–3.
 References 133

Haiman CA, Chen GK, Blot WJ et al., GWAS of PSA at baseline to predict prostate cáncer in
prostate cancer in men of African ancestry unscreened men. Eur Urol 61;2012: 471–7.
identifies a susceptibility locus a 17q21. Nat Korver W, Guevara C, Chen Y et al., The
Genet 43;2011:570–3. product of the candidate prostate cancer
Han Y, Signorello LB, Strom SS et al., susceptibility gene ELAC2 interacts with
Generalizability of established prostate can- the gamma-tubulin complex. Int J Cancer
cer risk variants in men of African ancestry. Int 104;2003:283–8.
J Cancer 136;2015:1210–7. Kote-Jarai Z, Olama AA, Giles GG et al., Seven
Hindorff LA, Sethupathy P, Junkins HA et al., prostate cancer susceptibility loci identi-
Potential etiologic and functional implications fied by a multistage GWAS. Nat Genet
of genome wide association loci for human 43;2011:785–91.
diseases and traits. Proc Natl Acad Sci USA Koutros S, Beane-Freeman LE, Berndt SI et al.,
106;2009:9362–7. Pesticide use modifies the association between
Hoffmann TJ, Van Den Eeden SK, Sakoda genetic variants on chromosome 8q24 and
LC et al., A large multiethnic genome prostate cancer. Cancer Res 70;2010:9224–33.
wide association study of prostate can- Krier J, Barfield R, Green RC et al.,
cer i­dentifies novel risk variants and sub- Reclassification of genetic based risk predic-
stantial ethnic differences. Cancer Discov tions as GWAS data accumulate. Genome
5;2015:878–891. Med 8;2016:20.
Hsu FC, Sun J, Wiklund F et al., A novel prostate Lalitha K, Suman G, Pruthvish S, Mathew A,
cancer susceptibility locus at 19q13. Cancer Murthy NS. Estimation of time trends of inci-
Res 69;2009:2720–3. dence of prostate cancer: An Indian scenario.
Hsu FC, Sun J, Zhu Y et al., Comparison of two Asian Pac J Cancer Prev 13;2012:6245–50.
methods for estimating absolute risk of Lieberfarb ME, Lin M, Lechpammer M et al.,
prostate cancer based on SNPs and family Genome wide loss of heterozygosity analysis
history. Cancer Epidemiol Biomarkers Prev from laser capture microdissected prostate
19;2010:1083–8. cancer using single nucleotide polymor-
Johansson M, Holmstrom B, Hinchcliffe SR et al., phic allele (SNP) arrays and a novel bioin-
Combining 33 genetic variants with prostate formatics platform of dChipSNP. Cancer
specific antigen for prediction of prostate 63;2003:4781–5.
cancer: Longitudinal study. Int J Cancer Lin K, Croswell JM, Koenig H et al., PSA based
130;2012:129–37. screening for prostate cancer. An evidence
Johns LE, Houlston RS. A systemic review and update for the US Preventive Services Task
meta-analysis of familial prostate cancer risk. Force Rockville (MD): Agency for Healthcare
BJU Int 91;2003:789–94. Research and Quality (US) 2011 Oct. Report
Kader AK, Sun J, Reck BH et al., Potential impact N: 12-05160-EF-1.
of adding genetic markers to clinical param- Liss MA, Xu J, Chen H et al., Prostate genetic
eters in predicting prostate biopsy outcomes score (PGS-33) is independently associated
in men following an initial negative biopsy: with risk of prostate cancer in the PCLO Trial.
Findings from the REDUCE Trial. Eur Urol Prostate 75;2015:1322–8.
62;2012:953–61. Liu M, Shi X, Yang F et al., The cumulative effect
Karlsson R, Aly M, Clements M et al., A popula- of gene-gene and gene-environment inter-
tion based assessment of germline HOXB13 actions on the risk of prostate cancer in
G84E mutation and prostate cancer risk. Chinese men. Int J Environ Res Public Health
Eur Urol 65;2014:169–76. 13;2016:162.
Kashyap A, Kluzniak W, Wokolorczyk D et al., Meitz JC, Edwards SM, Easton DF et al., HPC2/
The presence of prostate cancer at biopsy is ELAC2 polymorphisms and prostate cancer
predicted by a number of genetic variants. Int risk: Analysis by age of onset of disease. Br J
J Cancer 134;2014:1139–46. Cancer 87;2002:905–8.
Klein RJ, Hallden C, Gupta A et al., Evaluation Nadler RB, Humphrey PA, Smith DS et al.,
of multiple risk-associated SNPSs versus Effect of inflammation and benign prostatic
134  Precision medicine in prostate cancer

hyperplasia on elevated serum PSA levels. J ELAC2/HPC2, RNASEL/HPC1 and MSR1

Urol 154;1995:407–13. with prostate cancer severity in European
Nam RK, Zhang WW, Trachtenberg J et al., Utility American and African American men. Cancer
of incorporating genetic variants for early Epidemiol Biomarkers Prev 14;2005:949–57.
detection of prostate cancer. Clin Cancer Res Ries LAG. SEER Cancer Statistics Review 1975–
15;2009:1787–93. 2002. 2005. NCI, Bethesda, MD.
Na R, Liu F, Zhang P et al., Evaluation of reported Rokman A, Ikonen T, Seppala EH et al., Germline
prostate cancer risk associated SNPs from alterations of the RNASEL gene, a candidate
GWAS of various racial populations in Chinese HPC1 gene at 1q25 in patients and fami-
men. Prostate 73;2013:1623–35. lies with prostate cancer. Am J Hum Genet
National Cancer Institute Surveillance 70;2002:1299–304.
Epidemiology and End Results Program, Sakr WA, Grigon DJ, Hass GP et al., Age and
Cancer Stat Facts: Cancer of Prostate, 2006. racial distribution of prostatic intraepithelial
www.seer.cancer.gov/statfacts/html/prost. neoplasia. Eur Urol 30(2);1996:138–44.
html, Accessed 2017. Schroder FH, Hugosson J, Roobol MJ et al.,
National Cancer Institute Surveillance Screening and prostate cancer mortality in
Epidemiology and End Results Program, a randomized European study. N Eng J Med
Cancer Stat Facts: Cancer of Prostate, 2016. 360;2009:1320–8.
http://seer.cancer.gov, Accessed 2017. Schumacher FR, Berndt SI, Siddiq A et al.,
Office for National Statistics, Cancer Statistics Genome-wide association study identifies
Registrations: Registrations of cancer diag- new prostate cancer susceptibility loci. Hum
nosed in 2008, England. Series MB1 No.39 Mol Genet 20;2011:3867–75.
2010 London; National Statistics. Selley, S, Donovan J, Faulkner A et al., Diagnosis,
Office for National Statistics, Cancer Statistics management and screening of early local-
Registrations: Registrations of cancer diag- ized prostate cancer. Health Technol Assess
nosed in 2010, England. Series MB1 London; 1(2);1997.
National Statistics. Severi G, Giles GG, Southey MC et al.,
Office for National Statistics, Cancer Statistics ELAC2/HPC2 polymorphisms, PSA lev-
Registrations: Registrations of cancer diag- els and ­prostate cancer. J Natl Cancer Inst
nosed in 2015, England. Series MB1 London; 95;2003:818–24.
National Statistics, 2017. Smith DS, Humphrey PA, Catalona WJ. The early
Page WF, Braun MM, Partin AW et al., Heredity detection of prostate carcinoma with PSA.
and prostate cancer: A study of World War II Cancer 80;1997:1852–6.
veteran twins. Prostate 33;1997: 240–250. Stanford JL, Sabacan LP, Noonan EA et al.,
Platt N, Gordon S. Is the class A macrophage Association of HPC2/ELAC2 polymorphisms
scavenger receptor (SR-A) multifunctional? J with risk of prostate cancer in a population
Clin Invest 108;2001:649–54. based study. Cancer Epidemiol Biomarkers
Punglia RS, D’Amico AV, Catalona WJ et al., Prev 12;2003:876–81.
Impact of age, benign prostatic hyper- Storebjerg TM, Hoyer S, Kirkegaard P et al.,
plasia and cancer on PSA level. Cancer Prevalence of the HOXB13 G84E mutation
106;2006:1507–13. in Danish men undergoing radical prosta-
Rebbeck TR, Walker AH, Zeigler-Johnson C et al., tectomy and its correlations with prostate
Association of the HPC2/ELAC2 genotypes cancer risk and aggressiveness. BJU Int
and prostate cancer. Am J Hum Genet 118;2016:646–53.
67;2000:1014–9. Sun J, Zheng SL, Wiklund F et al., Sequence
Ren S, Xu J, Zhou T et al., Plateau effect of pros- ­variants at 22q13 are associated with
tate cancer risk associated SNPs in discrimi- ­prostate cancer risk. Cancer Res
nating prostate biopsy outcomes. Prostate 69;2009:10–15.
73;2013:1824–35. Takaku H, Minagawa A, Taskagi M et al.,
Rennert H, Zeigler-Johnson CM, Addya K et al., A ­candidate prostate cancer susceptibil-
Association of the susceptibility alleles in ity gene encodes tRNA 3′ processing
 References 135

­endoribonuclease. Nucleic Acids Res Woolf CM. An investigation of the familial

31;2003:2272–8. aspects of carcinoma of the prostate. Cancer
Takata R, Akamatsu S, Kubo M et al., GWAS iden- 13;1960:739–44.
tifies five new susceptibility loci for prostate Xu J, Mo Z, Ye D et al., GWAS in Chinese
cancer in the Japanese population. Nat Genet men identifies two new prostate cancer
42;2010:751–4. risk loci at 9q31.2 and 19q13.4. Nat Genet
Tavtigian SV, Simard J, Teng DH et al., A 44;2012:1231–5.
­candidate prostate cancer susceptibil- Xu J, Zheng SL, Komiya A et al., Common
ity gene at chromosome 17p. Nat Genet sequence variants of the macrophage
27;2001:321–5. scavenger receptor 1 gene are associated
Thomas G, Jacobs KB, Yeager M et al., Multiple with prostate cancer. Am J Hum Genet
loci identified in a GWAS of prostate cancer. 72;2002a:208–12.
Nat Genet 40;2008:310–5. Xu J, Zheng SL, Komiya A et al., Germline
Thompson IM, Goodman PJ, Tangen CM et al., mutations and sequence variants of the
The influence of finisteride on the devel- macrophage scavenger receptor 1 gene are
opment of prostate cancer. N Eng J Med associated with prostate cancer risk. Nat
349;2003:215–24. Genet 32;2002b:321–5.
van den Bergh RC, Essink-Bot M, Roobol MJ Yeager M, Orr N, Hayes RB et al., GWAS of
et al., Anxiety and distress during active ­prostate cancer identifies a second risk
surveillance for early prostate cancer. Cancer locus at 8q24. Nat Genet 39;2007:645–9.
115;2009:3868–78. Zhang GM, Zhu Y, Chen HT et al., Association
Wang L, McDonnell SK, Elkins DA et al., Analysis between the body mass index and prostate
of the RNASEL gene in familial and spo- cancer at biopsy is modified by genetic risk:
radic prostate cancer. Am J Hum Genet A cross sectional analysis in China. Medicine
71;2002:116–23. (Baltimore) 94;2015:e1603.
Warlick C, Trock BJ, Landis P et al., Delayed Zheng SL, Sun J, Wiklund F et al., Cumulative
versus immediate surgical intervention and association of five genetic variants
prostate cancer outcome. J Natl Cancer Inst with ­prostate cancer. N Eng J Med
98;2006:355–7. 58;2008:910–9.
 7
Breast cancer epigenetic targets
for precision medicine

RAMONA G. DUMITRESCU

Introduction 137 The role of estrogen receptor (ER)

Epigenetic targets for breast cancer prevention 137 methylation in breast cancer treatment 139
Epigenetic targets for breast cancer Several other epigenetic changes and
treatments 139 epigenetic modulation in breast
The role of BRCA1 methylation in cancer treatment 140
predicting the treatment response 139 References 142

INTRODUCTION Thus, these epigenetic biomarkers could play an

essential role for primary, secondary, and tertiary
Breast cancer is the most common cancer in prevention.
women and the second leading cause of cancer Unfortunately, acquired or intrinsic resistance
death among women in the United States. Due to to targeted therapy has been reported in breast
the improved understanding of molecular defects cancer treatment. It has been observed that tar-
involved in the initiation and progression of dif- geted therapy resistance in breast cancer can be
ferent subtypes of breast cancer (Polyak, 2011), tar- mediated by epigenetic changes (Liu et  al., 2016;
geted therapeutics have been developed, leading to Walsh et al., 2016). Epigenetic modifications could
great advancements for individualized breast can- be important targets for both the prevention and
cer patients’ treatment. treatment of primary tumors as well as resistant
Among these defects, epigenetic changes tumors due to the reversible nature of the epigen-
involved in altering gene expression play an essen- etic changes. Thus, the study of cancer epigenetic
tial role in cancer development and progression. changes could lead to expanding the range of ther-
Recent studies described the existence of breast apeutic options in precision medicine.
cancer–specific DNA methylation profiles (Cancer In this chapter, several epigenetic targets for
Genome Atlas Network, 2012) that can help in breast cancer prevention and treatment with
identifying breast cancer specific pathways, which potential important implications for the disease
could be modified by targeted interventions, lead- prognosis and survival will be discussed.
ing to reduced breast cancer incidence. These
methylation signatures can also help in refining Epigenetic targets for breast cancer
breast cancer screening to achieve early detection prevention
and can help in clinical care by determining per-
sonalized cancer treatments and predicting dis- Dysregulation of the epigenome in cancer is well
ease-free and overall survival (Terry et  al., 2016). recognized to play a crucial role in development

137
138  Breast cancer epigenetic targets for precision medicine

of breast cancer (Baylin and Jones, 2011) but it is the hypothesis that breast cancer initiates when
not well known how early these epigenetic changes there is epigenetic disruption of the transcription
take place. Although most studies examined breast factor binding that could lead to dysregulation
tumors to look for these epigenetic events, the of numerous networks involved in cancer develop-
early changes in the epigenome are very difficult ment (Locke et al., 2015).
to be examined and detected. One model used In the last few years, it has been reported that
to study the epigenetic remodeling during breast blood-based DNA methylation markers could be
cells’ malignant transformation is represented by used to evaluate breast cancer risk (Severi et  al.,
the human mammary epithelial cells (HMECs) 2014; Tang et al., 2016).
(Hinshelwood and Clark, 2008; Locke and Clark, It has been shown that a new epigenetics-based
2012). These cells derive from breast tissue surgi- system was able to differentiate healthy volunteers
cally removed during the reduction mammoplasty from breast cancer patients, with high accuracy
procedure in healthy women. (Uehiro et al., 2016). This system is based on 12 novel
Another related model is represented by the epigenetic markers that were identified after a meth-
variant HMECs (vHMECs), which are cells under- ylation array analysis was conducted. The authors
going early carcinogenic transformation by going suggested that early breast cancer detection based on
through selection or overcoming senescence barri- this system is similar to the mammography screen-
ers. These cells are believed to be a model of basal- ing detection (Uehiro et al., 2016).
like breast carcinogenesis (Hinshelwood et  al., Similar results were shown by a different group,
2008; Locke and Clark, 2012). The changes observed when a six-gene methylation panel had a high sen-
in these cells include the promoter hypermethyl- sitivity and specificity in breast cancer diagnosis
ation of p16INK4a tumor suppressor gene, silenc- when compared with healthy and benign disease
ing of the transforming growth factor beta gene controls. This further suggests that the detection
(TGFB), and increased expression of the chromatin of aberrant methylation of several genes in serum
methyltransferase EZH2 to name few early events DNA could be a potential diagnosis tool for breast
(Hinshelwood et al., 2007). Thus, the HMECs could cancer (Shan et al., 2016).
be a good model to look at for specific changes from Recent studies linked the “epigenetic clock,”
normal cells to premalignant cells. or the DNA methylation-based markers of age-
Looking at the gene expression between HMECs ing, to the risk of cancer development. In order to
and vHMECs, it has been observed that genes dif- identify possible epigenetics biomarkers for breast
ferentially expressed in vHMECs have critical cancer susceptibility and risk stratification, a new
roles in cellular growth and proliferation, cellular study conducted a DNA methylome analysis in the
movement, cell cycle and DNA replication, recom- European Prospective Investigation into Cancer
bination and DNA repair, and the most significant and Nutrition (EPIC) cohort using the Illumina
pathways identified, including the oncogenes MYC HumanMethylation 450K BeadChip arrays. It was
and EZH2 (Locke et al., 2015). The same analysis found that higher CpG islands’ DNA methylation
looked at the activity of transcriptions factors and is significantly associated with postmenopausal
found that NF-kB and E2F as well as stem cell/can- breast cancer susceptibility, suggesting that accel-
cer-associated transcription factor NANOG and erated epigenetic ageing increases risk of breast
ligand-activated transcription factor AHR signal- cancer development (Ambatipudi et al., 2017).
ing pathways were activated. The transcription fac- Furthermore, when the advanced DNA meth-
tors that were found by this analysis to be inhibited ylation (DNAm) age in breast tissue was com-
were p53 and RB (Locke et al., 2015). pared with the DNAm age of peripheral blood
Furthermore, differentially methylated regions tissue from the same individuals, it was observed
(DMRs) were identified in early passages of that DNAm age was highly correlated with chron-
vHMECs. The coordinated DNA hypermethyl- ological age in both breast tissues and peripheral
ation of target loci were regulated by key tran- blood. The study showed that breast tissue has a
scriptional factors like p53, AHR, and E2F family higher epigenetic age relative to peripheral blood,
members. This together with long-range epigenetic but the difference between the DNAm age in
dysregulation is believed to represent the earliest breast and blood is decreasing with advancing age
stage of breast malignancy. This finding supports (Sehl et al., 2017).
 Introduction 139

In addition, several studies showed that to genetic mutations, so that more women other
another DNA methylation marker, the global than mutation carries could benefit from targeted
hypomethylation is associated with increased interventions.
breast cancer risk (Choi et  al., 2009; Tang et  al., Anticancer drugs leading to the formation of
2016). double-stranded DNA breaks (DSBs) have been
When the epigenome-wide methylation analy- experienced in breast cancer, because this therapy
sis was conducted in three independent prospec- is predicted to be effective in killing BRCA1- or
tive nested case-control studies in relationship to BRCA2-defective cells. For example, PARP inhibi-
breast cancer risk, it was observed that epigenome- tors have been reported to be effective in targeting
wide hypomethylation is associated with breast cancer cells exhibiting errors in the DNA repair
cancer risk (van Veldhoven et  al., 2015). In addi- mechanisms (Bryant et al., 2005).
tion, decreased average methylation levels were The determination of methylation status of
detected in blood samples, years before breast BRCA1 and other genes of the BRCA/homologous
cancer diagnosis. This finding suggests that this recombination (HR) pathway could be used as a
genome-wide epigenetic change could be used as a predictor of the response of breast tumors to PARP
clinical biomarker, with predictive value for breast inhibitors and cisplatin therapy (Stefansson and
cancer risk (van Veldhoven et al., 2015). Esteller, 2013; Cai et al., 2014).
Therefore, it was suggested that in triple-nega-
Epigenetic targets for breast cancer tive breast cancers (TNBCs) patients with acquired
treatments or inherited defects in the BRCA1 gene, the use
of platinum-based drugs, or PARP inhibitors, for
Personalized medicine in breast cancer treat- the treatment of these aggressive tumors could
ment would involve tailored drug usage based have a significant impact on the disease outcome
on the genetic defects identified in the primary (Stefansson and Esteller, 2013).
tumor (Cho et al., 2012). Some genetic defects are BRCA1 loss associated with impaired DNA
predictive of the response to specific drugs, as in repair has been observed in triple-negative breast
the case of BRCA1 and BRCA2 mutations in rela- cancers. In fact, when DNA methylation was ana-
tion to PARP inhibitors and PIK3CA mutations lyzed in these tumors, BRCA1 promoter meth-
in relation to PIK3 inhibitors. However, in many ylation was found to be associated with reduced
sporadic breast cancers, several genes are inac- BRCA1 expression, aggressive phenotype, and
tivated through DNA methylation changes, and shorter overall survival. Thus, BRCA1 promoter
these changes play a role in the tumors’ response to methylation is an important mechanism that leads
breast cancer therapy. Several of these methylation to functional loss of BRCA1 (Yamashita et  al.,
changes with important roles in the personalized 2015) and could be a promising biomarker for the
breast cancer treatment will be discussed. prognosis of triple-negative breast cancers (Zhu
et al., 2015).
THE ROLE OF BRCA1 METHYLATION IN When BRCA 1 methylation and TP53 mutations
PREDICTING THE TREATMENT RESPONSE were analyzed in triple-negative breast cancer
Breast cancers develop in about 5% of BRCA1 and patients without pathological complete response
BRCA2 mutation carriers, however, somatic muta- to taxane-based neoadjuvant chemotherapy, it was
tions of BRCA1 rarely occur in sporadic breast observed that BRCA1 methylation was associated
cancer. Lower than normal rates of BRCA1 expres- with a significant decrease in disease-free survival,
sion or loss of BRCA1 expression through epigen- providing important information for breast cancer
etic inactivation was observed in sporadic breast treatment and prognosis (Foedermayr et al., 2014).
tumors. Thus, epigenetic silencing of BRCA1 is
believed to be an important factor that contrib- THE ROLE OF ESTROGEN RECEPTOR
utes to breast cancer development. Considering (ER) METHYLATION IN BREAST CANCER
that BRCA1 methylation is present in about TREATMENT
10%–15% of breast cancer patients (Birgisdottir The expression of estrogen receptor (ER) deter-
et al., 2006), it would be extremely valuable to take mines if a breast cancer patient receives endocrine
into account the epigenetic changes, in addition therapy but does not assure a therapeutic response.
140  Breast cancer epigenetic targets for precision medicine

More specifically, the loss of ERα expression leads methylation. More specifically, ERα may direct
to lack of response to antihormone treatment. In DNA methylation–associated silencing of a sub-
a significant fraction of ER+/PR– and ER–/PR– population of basal markers, cancer stem cells,
breast tumors, the loss of expression is a result of epithelial–mesenchymal transition, and inflamma-
epigenetic changes within the ERα gene promoter, tory and tumor suppressor genes with a potential
such as DNA methylation and histone deacety- role in enforcing a phenotype with luminal char-
lation (Stearns et al., 2007). acteristics (Ariazi et al., 2017). These findings show
Patients expressing ER and PR proteins respond how important the ER-associated methylation is
to hormonal treatment. Therefore, it was hypoth- for the breast cancer phenotype and treatment.
esized that reversing the ER promoter methylation
may help tumors become responsive to tamoxifen SEVERAL OTHER EPIGENETIC CHANGES
therapy and this was amply described in a review AND EPIGENETIC MODULATION IN BREAST
by Stearns and his collaborators (Stearns et  al., CANCER TREATMENT
2007). When ER-negative breast cancer cell lines As mentioned before, promoter methylation of sev-
were treated with HDAC inhibitors and/or demeth- eral other key genes has been shown to be critical
ylating agents like trichostatin A (TSA) and 5-aza- in breast cancer initiation and progression (Baylin
2-deoxycytidine, the reexpression of a functional and Jones, 2011).
ER was observed. Similar findings were reported in Recently, Mathe and his group looked if the
ER-negative breast cancer cell lines after the admin- DNA methylation contributes to the altered gene
istration of an HDAC inhibitor, scriptaid. Also, expression in TNBCs as well as in disease pro-
scriptaid growth inhibition was observed together gression from primary breast tumor to lymph
with increased acetylation of histone tails. Similar node metastasis (Mathe et  al., 2016). The authors
to trichostatin A combination treatment, the scrip- reported that this whole genome DNA methyla-
taid and 5-aza-2-deoxycytidine were more effective tion analysis in a TNBC cohort showed an altered
in reexpressing ER than either agent alone (Stearns methylation of 18 genes associated with lymph
et al., 2007). These findings showed that the admin- node metastasis that contributes to the dysregula-
istration of HDAC inhibitors can reactivate the ER tion of gene expression changes and is associated
in ER– breast cancer cell lines and that restores the with overall survival (Mathe et al., 2016).
sensitivity to tamoxifen and then led to cell growth In the ER+ breast cancers, the epigenetic-asso-
inhibition. Therefore, targeting to reactivate genes ciated overexpression of histone H2B variants was
that are silenced by promoter hypermethylation in shown to play an important role in the endocrine
breast cancer with demethylating agents could be response of these tumors to treatment (Nayak
an important therapeutic approach. et al., 2015).
Furthermore, when a new study looked at In addition, in the ER+ breast cancers, the
the DNA methylation role in the breast cancers RUNX3 hypermethylation was frequently observed
response to endocrine therapy, it was found that and a higher RUNX3 mRNA expression was asso-
DNA hypermethylation of the estrogen-respon- ciated with better relapse-free survival (Lu et  al.,
sive enhancers leads to reduced ER binding and 2017). Thus, RUNX3 methylation could be a thera-
reduced expression of genes regulated by the ER peutic target for the development of personalized
(Stone et al., 2015). The study also showed that ER therapy.
enhancers’ hypomethylation plays a critical role Some HDAC inhibitors and DNMT inhibi-
in the transition from normal cells to ER-positive tors have been investigated in preclinical models
cancerous cells that respond to endocrine therapy. in breast cancer, and it was shown that drugs that
These findings highlight the important role of modulate epigenetic changes may be used as single
estrogen-regulated enhancers’ DNA methylation agent or in combination with standard treatments.
for the endocrine sensitivity in breast cancer and Some of these preclinical studies have shown that
also for the potential prediction of ER-positive sta- HDAC inhibitors’ effect may not be complete or can
tus (Stone et al., 2015). be reversible if the treatment is stopped. Therefore,
Recently, a new role for the ERα-associated the combination therapy of HDAC inhibitors
gene expression regulation was proposed. It has and tamoxifen or retinoic acid could be a better
been reported that ERα can silence genes via DNA approach. Another combination that was suggested
 Introduction 141

was HDAC inhibitors and DNMT inhibitors or Moreover, it has previously been observed that
drugs like trastuzumab. Others suggested HDAC triple-negative breast cancer cell lines show het-
inhibitors and chemotherapeutic agents, such as erogeneous responses to the PARP and HDAC
docetaxel and gemcitabine, so that growth inhibi- inhibitors. When the coadministration of olaparib
tion and apoptosis are achieved (Stearns et al., 2007). and SAHA were studied to examine their effect
For the human epidermal growth factor (HER) on the growth of TNBC, it was found that cells
2-positive breast cancers, the standard therapy is that expressed functional phosphatase and tensin
targeting the HER2. This therapeutic approach homolog (PTEN) would favorably respond to the
was associated with disease-free survival if a combination therapy (Min et al., 2015).
pathological complete response to the therapy Recently, carnitine palmitoyl transferase-1A
is achieved. A  new study conducted a genome- (CPT1A) was identified as a new tumor specific
wide DNA methylation analysis and identified target in human breast cancer (Pucci et al., 2016).
eight genomic regions differentially methylated It has been suggested before that CPT1A variant
in patients with pathological complete response. 2 product is involved in the epigenetic regulation
Among those regions, HSD17B4 encoding type 4 of cancer survival, cell death escaping, and metas-
17β-hydroxysteroid dehydrogenase was most sig- tasis pathways. Indeed, the inactivation of CPT1A
nificantly differentially methylated (Fujii et  al., variant 2 by using small interfering RNAs (siRNAs)
2017). The authors concluded that the pathological led to apoptosis in several breast cancer cell lines.
complete response of HER2-positive breast cancers CPT1A silencing was associated with reduction of
to trastuzumab and chemotherapy can be predicted HDAC activity and histone hyperacetylation, lead-
by HSD17B4 methylation, and this can have impor- ing to upregulated transcription of proapoptotic
tant treatment implications for this group of breast genes (BAD, CASP9, COL18A1) and downmodu-
cancer patients (Fujii et al., 2017). lation of invasion and metastasis genes (TIMP-1,
Another tumor suppressor gene frequently PDGF-A, SERPINB2). Thus, CPT1A has been pro-
hypermethylated in breast cancer is RARβ2 and posed as a new tumor-specific target, more effec-
the silencing of this gene is reinforced by the his- tive than the well-known aforementioned HDAC
tone deacetylation at the promoter region. Thus, inhibitors (Pucci et al., 2016).
similarly to ER methylation, the reactivation of Aberrant DNA hypermethylation has been
RARβ2 could restore RARβ effects in breast can- shown to play a critical role in the regulation of
cer. In fact, the treatment with 5-aza-2deoxycyti- renewal and maintenance of cancer stem cells
dine and TSA has been reported to restore RARβ2 (CSCs), which are targets for cancer initiation by
activity, suggesting that these drugs with effect on chemical and environmental factors. The admin-
chromatin remodeling can reactivate the RARβ2 istration of a DNA hypermethylation inhibitor like
gene and overcome the retinoic acid resistance in decitabine (DAC) can have an important effect on
breast cancer (Stearns et al., 2007). the chemotherapeutic response and the acquired
In a recent study, it was found that the combina- drug resistance of the CSCs (Li et al., 2015). Based
tion of entinostat, ATRA, and doxorubicin (EAD) on this hypothesis, in vitro studies were conducted
restored the epigenetically silenced RAR-β expres- to look at the effect of the treatment with nanopar-
sion and resulted in significant tumor regression. ticles loaded with low-dose DAC (NPDAC) com-
The findings in this study suggest that Entinostat bined with nanoparticles loaded with doxorubicin
mediates Doxorubicin’s action on cytotoxicity and (NPDOX) on the proportion of CSCs with high
differentiation driven by retinoids in order to achieve aldehyde dehydrogenase activity (ALDHhi). It
tumor regression in TNBC (Merino et al., 2016). has been observed that this therapeutic approach
Furthermore, several inhibitors of histone achieved the highest proportion of apoptotic tumor
deacetylases were considered of a great therapeu- cells and the lowest proportion of ALDHhi CSCs.
tic value (Stearns et al., 2007), especially for TNBC This finding suggests that the combined treatment
patients. A recent study examining the effects of NPDAC and NPDOX has a critical role in the
of suberoyl anilide hydroxamic acid (SAHA) on inhibition of breast cancer growth (Li et al., 2015).
TNBCs has revealed that SAHA has great potential Unfortunately, many cancer cells acquire multi-
to be used as an anticancer agent in the treatment drug resistance (MDR) that was associated with the
of these type of tumors (Wu et al., 2016). altered expression of glucosylceramide synthase
142  Breast cancer epigenetic targets for precision medicine

(GCS). Liu and his group analyzed the associa- Cho SH, Jeon J, and Kim SI. Personalized medi-
tion of methylation at the GCS gene promoter with cine in breast cancer: A systematic review.
its expression and MDR in invasive ductal breast J Breast Cancer 15(3);2012:265–72.
cancer and found that the GCS promoter methyla- Choi JY, James SR, Link PA, McCann SE, Hong CC,
tion changes correlate with multidrug resistance in Davis W, Nesline MK, Ambrosone CB,
breast cancer (Liu et al., 2016). and Karpf AR. Association between global
In conclusion, the epigenetic changes described DNA hypomethylation in leukocytes and
play a crucial role in breast cancer initiation, pro- risk of breast cancer. Carcinogenesis
gression, and response to targeted therapies. These 30(11);2009:1889–97.
epigenetic biomarkers, once validated in clini- Foedermayr M, Sebesta M, Rudas M et al.
cal studies, can provide essential information for BRCA-1 methylation and TP53 mutation in
breast cancer risk assessment for predicting the triple-negative breast cancer patients without
type of breast cancer and the most effective treat- pathological complete response to taxane-
ment for that tumor type. This is moving the field based neoadjuvant chemotherapy. Cancer
of precision oncology forward, with a great impact Chemother Pharmacol 73(4);2014:771–8.
in breast cancer patients’ lives. Fujii S, Yamashita S, Yamaguchi T, Takahashi
M, Hozumi Y, Ushijima T, and Mukai H.
REFERENCES Pathological complete response of HER2-
positive breast cancer to trastuzumab and
Ambatipudi S, Horvath S, Perrier F et al. DNA chemotherapy can be predicted by HSD17B4
methylome analysis identifies accelerated methylation. Oncotarget 8(12);2017:19039–48.
epigenetic ageing a ­ ssociated with postmeno- Hinshelwood RA, and Clark SJ. Breast cancer
pausal breast cancer susceptibility. Eur J epigenetics: Normal human mammary epithe-
Cancer 75;2017:299–307. lial cells as a model system. J Mol Med (Berl)
Ariazi EA, Taylor JC, Black MA, Nicolas E, 86(12);2008:1315–28.
Slifker MJ, Azzam DJ, and Boyd J. A new role Hinshelwood RA, Huschtscha LI, Melki J et al.
for ERα: Silencing via DNA methylation of Concordant epigenetic silencing of transform-
basal, stem cell, and EMT genes. Mol Cancer ing growth factor-beta signaling pathway
Res 15(2);2017:152–64. genes occurs early in breast carcinogenesis.
Baylin SB, and Jones PA. A decade of explor- Cancer Res 67(24);2007:11517–27.
ing the cancer epigenome—biological and Li SY, Sun R, Wang HX, Shen S, Liu Y, Du XJ, Zhu
translational implications. Nat Rev Cancer YH, and Jun W. Combination therapy with
11(10);2011:726–34. epigenetic-targeted and chemotherapeutic
Birgisdottir V, Stefansson OA, Bodvarsdottir SK, drugs delivered by nanoparticles to enhance
Hilmarsdottir H, Jonasson JG, and Eyfjord JE. the chemotherapy response and overcome
Epigenetic silencing and deletion of the resistance by breast cancer stem cells.
BRCA1 gene in sporadic breast cancer. Breast J Control Release 205;2015:7–14.
Cancer Res 8(4);2006:R38. Liu J, Zhang X, Liu A et al. Altered methyla-
Bryant HE, Schultz N, Thomas HD, Parker KM, tion of glucosylceramide synthase promoter
Flower D, Lopez E, Kyle S, Meuth M, regulates its expression and associates
Curtin NJ, and Helleday T. Specific killing with acquired multidrug resistance in inva-
of BRCA2-deficient tumours with inhibitors sive ductal breast cancer. Oncotarget
of poly(ADP-ribose) polymerase. Nature 7(24);2016:36755–66.
434(7035);2005:913–7. Locke WJ, and Clark SJ. Epigenome remodeling
Cai F, Ge I, Wang M, Biskup E, Lin X, and Zhong in breast cancer: Insights from an early in vitro
X. Pyrosequencing analysis of BRCA1 methyl- model of carcinogenesis. Breast Cancer Res
ation level in breast cancer cells. Tumour Biol 14(6);2012:215.
35(4);2014:3839–44. Locke WJ, Zotenko E, Stirzaker C, Robinson MD,
Cancer Genome Atlas Network. Comprehensive Hinshelwood RA, Stone A, Reddel RR,
molecular portraits of human breast tumours. Huschtscha LI, and Clark SJ. Coordinated
Nature 490(7418);2012:61–70. epigenetic remodelling of transcriptional
 References 143

networks occurs during early breast carcino- Stearns V, Zhou Q, and Davidson NE. Epigenetic
genesis. Clin Epigenetics 7(1);2015:52. regulation as a new target for breast cancer
Lu DG, Ma YM, Zhu AJ, and Han YW. An early therapy. Cancer Invest 25(8);2007:659–65.
biomarker and potential therapeutic target of Stefansson OA, and Esteller M. Epigenetic
RUNX 3 hypermethylation in breast cancer, a modifications in breast cancer and their
system review and meta-analysis. Oncotarget role in personalized medicine. Am J Pathol
8(13);2017:22166–74. 183(4);2013:1052–63.
Mathe A, Wong-Brown M, Locke WJ, Stirzaker C, Stone A, Zotenko E, Locke WJ et al. DNA meth-
Braye SG, Forbes JF, Clark SJ, Avery-Kiejda ylation of oestrogen-regulated enhancers
KA, and Scott RJ. DNA methylation profile of defines endocrine sensitivity in breast cancer.
triple negative breast cancer-specific genes Nat Commun 6;2015:7758.
comparing lymph node positive patients Tang Q, Cheng J, Cao X, Surowy H, and
to lymph node negative patients. Sci Rep Burwinkel B. Blood-based DNA methylation
6;2016:33435. as biomarker for breast cancer: A systematic
Merino VF, Nguyen N, Jin K et al. Combined review. Clin Epigenetics 8;2016:115.
treatment with epigenetic, differentiating, Terry MB, McDonald JA, Wu HC, Eng S, and
and chemotherapeutic agents cooperatively Santella RM. Epigenetic biomarkers of
targets tumor-initiating cells in triple negative breast cancer risk: Across the breast cancer
breast cancer. Cancer Res 76(7);2016:2013–24. prevention continuum. Adv Exp Med Biol
Min A, Im S, Kim DK et al. Histone deacetylase 882;2016:33–68.
inhibitor, suberoylanilide hydroxamic acid Uehiro N, Sato F, Pu F, Tanaka S, Kawashima M,
(SAHA), enhances anti-tumor effects of the Kawaguchi K, Sugimoto M, Saji S, and Toi M.
poly (ADP-ribose) polymerase (PARP) inhibitor Circulating cell-free DNA-based epigenetic
olaparib in triple-negative breast cancer cells. assay can detect early breast cancer. Breast
Breast Cancer Res 17;2015:33. Cancer Res 18(1);2016:129.
Nayak SR, Harrington E, Boone D et al. A role van Veldhoven K, Polidoro S, Baglietto L et al.
for histone H2B variants in endocrine- Epigenome-wide association study reveals
resistant breast cancer. Horm Cancer decreased average methylation levels
6(5–6);2015:214–24. years before breast cancer diagnosis. Clin
Polyak K. Heterogeneity in breast cancer. J Clin Epigenetics 7;2015:67.
Invest 121(10);2011:3786–8. Walsh L, Gallagher WM, O’Connor DP, and
Pucci S, Zonetti MJ, Fisco T, Polidoro C, Ní Chonghaile T. Diagnostic and therapeutic
Bocchinfuso G, Palleschi A, Novelli G, implications of histone epigenetic modula-
Spagnoli LG, and Mazzarelli P. Carnitine tors in breast cancer. Expert Rev Mol Diagn
palmitoyl transferase-1A (CPT1A): A new 16(5);2016:541–51.
tumor specific target in human breast cancer. Wu S, Luo Z, Yu PJ, Xie H, and He YW.
Oncotarget 7(15);2016:19982–96. Suberoylanilide hydroxamic acid (SAHA)
Sehl ME, Henry JE, Storniolo AM, Ganz PA, and promotes the epithelial mesenchymal
Horvath S. DNA methylation age is elevated in transition of triple negative breast cancer
breast tissue of healthy women. Breast Cancer cells via HDAC8/FOXA1 signals. Biol Chem
Res Treat 2017 Mar 31 (Epub ahead of print). 397(1);2016:75–83.
Severi G, Southey MC, English DR, Jung CH, Yamashita N, Tokunaga E, Kitao H et al.
Lonie A, McLean C, Tsimiklis H, Hopper JL, Epigenetic inactivation of BRCA1 through
Giles GG, and Baglietto L. Epigenome-wide promoter hypermethylation and its clinical
methylation in DNA from peripheral blood importance in triple-negative breast cancer.
as a marker of risk for breast cancer. Breast Clin Breast Cancer 15(6);2015:498–504.
Cancer Res Treat 148(3);2014:665–73. Zhu X, Shan L, Wang F et al. Hypermethylation
Shan M, Yin H, Li J et al. Detection of aberrant of BRCA1 gene: Implication for prognostic
methylation of a six-gene panel in serum DNA biomarker and therapeutic target in sporadic
for diagnosis of breast cancer. Oncotarget primary triple-negative breast cancer. Breast
7(14);2016:18485–94. Cancer Res Treat 150(3);2015:479–86.
 8
Precision medicine in ovarian
carcinoma

SHAILENDRA DWIVEDI, PURVI PUROHIT, RADHIEKA MISRA, JEEWAN

RAM VISHNOI, APUL GOEL, PUNEET PAREEK, SANJAY KHATTRI, PRAVEEN
SHARMA, SANJEEV MISRA, AND KAMLESH KUMAR PANT

Introduction 145 Micro-RNA expression in ovarian cancer 155

Advancement of cutting edge technology Proteomics 156
for accurate diagnosis 146 Current scenario in ovarian cancer therapy 158
Genomics 146 Cytoreductive surgery and interval
Genome and genetic characterization debulking surgery (IDS) 158
tools and techniques 147 Chemotherapy in ovarian cancer 159
High-throughput sequencing technologies 150 Precision medicine in ovarian cancer 159
Somatic and germ-line mutation/variance Antiangiogenic therapy 159
studied in ovarian carcinoma 151 Drugs targeting mutations in ovarian
Transcriptomics 153 cancer precision medicine 162
Epigenetic (DNA methylation analysis) 154 Future prospects 165
Epigenetic alterations in cancer 154 References 166
Epigenetic role in ovarian cancer 155

INTRODUCTION preliminary staging, along with surgical and

chemotherapeutic interventions, has enriched
Advancement in preventing, screening, and ther- the short-term course of ovarian carcinoma.
apy of ovarian carcinoma has been affected by the Yet, despite such developments, most patients
fact that ovarian cancer is neither a frequent nor revert after primary treatment and succumb to
a rare disease. The risk of developing this disease disease advancement. Surgical management of
is 1 in 70 and the occurrence is 1 in 2500 for post- advanced-stage ovarian cancer has long been
menopausal women, patients mostly diagnosed a central view of genuine management of the
after crossing 50 years of age. It is fifth most com- patients suffering from this disease, although the
mon reason of cancer-linked death in women. The complete elimination is not possible. Though the
probable yearly incidence of this cancer globally order of chemotherapy and surgical intercession
is over 200,000 individuals, with about 125,000 is disputed, there is the comprehensive attitude
deaths (Sankaranarayanan and Ferlay, 2006). that integration of the two modalities signifies
Relentless progress in the knowledge of the the best primary strategy for women with meta-
natural history of the disease and thorough static disease.

145
146  Precision medicine in ovarian carcinoma

Precision medicine focuses on the designing but also in quality control of the tests and have
of the medical treatment to the individual char- added calibration of traditional biomarkers. It
acteristics of each patient. Precision medicine is documented since Gregor Mendel that factors
depends on the broad exploration of individual are responsible for expression of characters and
molecular database of genomics, epigenetics, after the launching of human genome, the pic-
proteomics, and metabolomics, as well as in ture of genotype to phenotype connection seems
silico tactics to acquire a thorough knowledge more bright and dynamic. The purposes of genes
of the association between the control of gene(s) and proteins have been reflected by postgenomics
(functional protein) and disease status. This advanced technologies such as expression profiling
advancement relies on modern cutting-edge by DNA microarray, proteomics, and genetic vari-
technologies of real-time PCR, microarray, and ance analysis, coupled with bioinformatics. Now,
next-generation sequencing (NGS) having capa- genetics has become the dynamic force in medical
bilities to explore the molecular characteristics research and is now ready for amalgamation into
in short time. medical practice.
As we are aware that any drug has its benefi-
cial effects to some individuals while others fail to Genomics
show its response, now it is established that this fact
depends on molecular characteristics of individu- Genomics is an interdisciplinary arena of science
als. Each individual or group having an exclusive concentrating on genomes. A genome is actually
phenotype (like disease) is mainly due to manifes- the study of a whole set of DNA within a single
tation of its molecular characteristics, that is, vari- cell of an organism, and as such genomics is a
ations (mutations) in gene make up or expression division of molecular biology related with the
profile, so for targeting any disease, our first goal is structure, function, evolution, and mapping of
to identify such changes and then we can observe genomes. The genetics of humans fundamentally
the suitability of any drug. Thus we will be able to explains their individual characteristics. Delicate
sub-group them into beneficial or null or negative variances found in the genetic makeup of a popu-
response groups. The control of gene expression lation generate the diverse characters perceived in
can also take place at the phases of translation and community.
posttranslation modification. Recent cutting-edge Further, differences in a genetic frame among a
in genomics, transcriptome profiling, and epi- population result into different gene isoforms that
genetic fingerprinting have been useful to cancer may consequently alter gene function producing
research and can probably be incorporated into a phenotypic variations in the form of characteris-
cancer systems biology approach for cultivating tics either protective or more susceptible for any
cancer medicine. disease. These are normally denoted as mutated
genes. Similarly like other several diseases and
ADVANCEMENT OF CUTTING EDGE cancer in ovarian cancer, mutated genes inherited
TECHNOLOGY FOR ACCURATE from a parent can make a person more risk or sus-
DIAGNOSIS ceptible of developing the disease in their lifespan.
Though, inherited mutations only interpret for a
The arena of oncology is relentlessly expanding small fraction of (20%–25%) all diagnosed cases in
mainly due to the utilization of enormous funds ovarian cancer.
and improvements in the basic sciences. Modern According to current predictions of 2017 by The
molecular diagnostic techniques and tools of American Cancer Society, about 22,440 women
genomics, transcriptomics and proteomics have will receive a new diagnosis of ovarian cancer and
been widely exploited in diagnosis of various can- about 14,080 women will die from ovarian cancer.
cers. Now automated and more advanced technol- Thus, it ranks fifth in cancer deaths among women,
ogy like PCR, real-time PCR, etc. have improved responsible for more deaths than any other cancer
the amplification and detection of nucleic acid of the female reproductive system. So sporadic,
sequences. scattered unpredictable gene alterations or deregu-
These automated and robotic platforms have lated gene expression would be the main sources
delivered precision of not only in assay efficiency for maximum ovarian cancer patients.
 Advancement of cutting edge technology for accurate diagnosis  147

GENOME AND GENETIC loss of restriction events. Later, several enzy-

CHARACTERIZATION TOOLS matic methods for mutation screening have
AND TECHNIQUES been developed. These procedures employ the
In 2003 with the publication of human genome activity of resolvase enzymes T4 endonuclease
draft, a new revolutionary phase of molecular VII and, in recent times, T7 endonuclease I to
biotechnology comes into existence, but still the digest heteroduplex DNA made by annealing
value of conventional tools and techniques is wild-type and mutant DNA. Digested frag-
not outdated. Validated mutations still screened ments specify the presence and the position of
by RFLP, probe based method or SYBR green- any mutations. One more enzymatic method
based approaches by utilizing real time PCR, or for mutation exploration is the oligonucleotide
by micro-array. As discussed earlier, cancer has ligation assay. In this method, two oligonucle-
a tremendously multifaceted etiology and in the otides are hybridized to complementary DNA
course of the development of the disease several fragments at locations of probable mutations.
number of mutations can occur. Thus, genomic The oligonucleotide primers are custom-
sequencing has a great value but due to high cost ized such that the 3′-end of the one primer
still few laboratories process the sample by real- is instantly contiguous to the 5′-end of the
time polymerase chain reaction (PCR), microarray other primer. Thus, if the first primer matches
or by sanger sequencing. Genomic sequencing has entirely with the target DNA, then the prim-
made it possible to screen any mutation or varia- ers can be ligated by DNA ligase. On the other
tion if it has any preexistence that would make it hand, if a disparity happens at the 3′-end of
more susceptible to develop cancer in their life- the first primer, then no ligation products will
time. For instance, mutations found in the BRCA1 be found.
and BRCA2 genes upsurge a woman’s susceptibil- 2 . Electrophoretic-based techniques—This class
ity for developing ovarian cancer in her lifetime to is well known by several diverse methods
approximately 40% and 18%, respectively. Further, planned for detection of known or unknown
these high throughput sequencing platforms have mutations, centered on the different electro-
the capacity to screen novel sporadic mutations in phoretic mobility of the mutant alleles, under
a very short time and with more precision. denaturing or nondenaturing conditions.
Single-strand conformation polymorphism
Genotyping or mutation analysis (SSCP) and heteroduplex analysis (HDA) were
Restriction fragment length polymorphism (RFLP) among the main approaches considered to
is a technique able to detect variation in homolo- detect molecular defects in genomic loci. In
gous DNA sequences that can be screened by the arrangement with capillary electrophoresis,
presence of fragments of variable lengths after SSCP and HDA investigation now offer an out-
digestion of the DNA by using precise restriction standing, modest, and fast mutation-finding
endonucleases. These fragments of varied length platform with low processing costs and, most
could be separated on the basis of molecular size in interestingly, the capability of easily being
agarose gel electrophoresis. automated, thus providing high-throughput
Further, these fragments could be utilized analysis of a patient’s DNA. Likewise, denatur-
in characterizing unique blotting patterns after ing and temperature gradient gel electropho-
hybridization with a complementary probe labeled resis (DGGE and TGGE, respectively) can be
with some fluorescent dye. RFLP probes are com- utilized equally well for mutation screening. In
monly utilized in genome sequence mapping and this method, electrophoretic mobility differ-
in variation analysis (genotyping, hereditary dis- ences between a wild type and mutant allele
ease diagnostics, etc.). These techniques can be can be visualized in a gradient of denaturing
generally categorized into three subgroups: agents, such as urea and formamide, or of
increasing temperature. Finally, a progres-
1. RFLP, enzymatic methods—RFLP analysis sively utilized mutation detection technique is
was traditionally the first technique broadly the two-dimensional gene scanning, based on
utilized in investigating the variations in two-dimensional electrophoretic separation of
restriction enzyme sites, leading to the gain or amplified DNA fragments, as per their size and
148  Precision medicine in ovarian carcinoma

base pair sequence. The latter involves DGGE, DNA sequence. This technique uses fluorescent
following the size separation step. labeled probes to screen, approve, and quantify the
3. Solid phase-based techniques/hybridization PCR yields as they are being produced in real time.
or blotting techniques—These established In real-time PCR, which is poised on three novel
methods involve the foundation for most of the characteristics, as temperature cycling happens
contemporary mutation screening technolo- substantially faster than in standard PCR assays,
gies, because they do not require any extra hybridization of specific DNA probes take place
effort as of being fully automated and there- continuously throughout the amplification reac-
fore are extremely praised for high through- tion and a fluorescent dye is attached to the probe
put mutation detection or screening. In the and fluoresces only when hybridization occurs. No
1970s there was advancement in nucleic acid post-PCR handling of amplified products is needed.
hybridization techniques, which is centered The generation of amplified products is perceived
on the pairing of two complementary nucleo- automatically by continuous monitoring of fluo-
tide strands. This pairing is primarily due rescence. In recent years, several commercial auto-
to envelopment of hydrogen, thus duplex or mated real-time PCR systems have been available
hybrid results. The hybrids may be consequent (Light Cycler & TaqMan). In these systems, such as
of DNA-DNA, RNA-RNA, or DNA-RNA, thus the Light CyclerTM and the Smart Cycler®, the real-
the single-stranded molecule may be DNA time fluorescence monitoring is accomplished by
or RNA in which one nucleic acid strand (the using fluorescent dyes such as SYBR-Green I, which
probe) originates from an organism of recog- binds nonspecifically to double-stranded DNA pro-
nized identity and the other strand (the target) duced during the PCR amplification. Others, such
originates from an unidentified organism to be as the TaqMan, use florescent probes that bind
detected or identified. exactly to amplification target sequences.

A quick, precise, and appropriate method for Microarrays

the discovery of known mutations is reverse dot- A microarray is an assembly of enriched charac-
blot, originally established by Saiki et  al. (1989), teristics of a microscopic system, in which nor-
and executed for the screening of b-thalassemia mally DNA is hybridized with target molecules for
mutations. The crux of this technique is the appli- quantitative (gene expression) or qualitative (diag-
cation of oligonucleotides, bound to a membrane, nostic) test of enormous quantities of genes con-
as hybridization targets for amplified DNA. The currently or to genotype multiple loci of a genome.
extra benefit of this technique is that one mem- Each DNA spot contains approximately picomoles
brane strip can be applied to screen many vari- (10−12 moles) of a specific DNA sequence, known
ous known mutations in a single individual (a one as probes (or reporters). Broadly due to improve-
strip-one patient type of assay), the potential of ments in fabrication, robotics, and bioinformatics,
automation, and the ease of analysis of the results, microarray technology has improved in terms of
using a classical avidin-biotin system. Though this efficiency, resolution power, robustness, sensitiv-
technique cannot be utilized for the uncovering of ity, and specificity. These enhancements have per-
unknown mutations, constant expansion has given mitted the transition of microarrays from strictly
rise to allele-specific hybridization of amplified research bench site to bed site in clinical diagnos-
DNA (PCR-ASO) on filters and newly extended tic applications. Microarrays can be distinguished
on DNA oligonucleotide microarrays for high on the basis of features such as the nature of the
throughput mutation analysis (Gemignani et  al., probe, the solid-surface support applied, and the
2002; Chan et al., 2004). specific method utilized for probe addressing and/
or target detection. An effective hybridization epi-
Real-time polymerase chain reaction sode between the labeled target and the immobi-
(PCR): Semiquantitative approach lized probe will consequently lead to an increase
Higuchi first demonstrated methodological of fluorescence intensity over a background level,
improvement in the form of “real-time PCR” dis- which can be studied by using a fluorescent scan-
covered approximately two decades earlier, which is ner (Miller, 2009). The entire strength of the sig-
modest, quantitative evaluation for any amplifiable nal, from a spot (feature), can be determined by the
 Advancement of cutting edge technology for accurate diagnosis  149

amount of target sample binding to the probes that favor due to its practical complexity barring its uti-
exist on that spot. Microarrays use relative quan- lization in standard molecular biology kits, wide use
tization in which the intensity of characteristics is of harmful chemicals, and complications with scale-
compared to the intensity of the same characteris- up. Each of four reactions (G, A + G, C, and C + T)
tics under a different condition, and the identity of is separated into different bands. Thus a series of
the feature is known by its position. In situ synthe- labeled fragments are produced, from the radiola-
sized arrays are exceedingly high-density micro- beled end to the first cut site in each molecule. The
arrays that use oligonucleotide probes, of which fragments in the four reactions are electrophoresed
Gene Chips (Affymetrix, Santa Clara, California) side by side in denaturing acrylamide gels for size
are the most commonly known. In addition to the separation. To see the fragments, the gel is visual-
printed oligonucleotide arrays discussed earlier, ized onto x-ray film for autoradiography, yielding a
the oligonucleotide probes are blended directly on series of dark bands each corresponding to a radio-
the surface of the microarray, which is typically a labeled DNA fragment, from which the sequence
1.2 cm2 quartz wafer. As in situ synthesized probes may be inferred, also sometimes known as the
are usually short (20–25 bp), multiple probes per “chemical sequencing” method.
target are involved to improve sensitivity, speci- Another method, known as the Sanger sequenc-
ficity, and statistical accuracy. As with in situ ing method, is centered on the principle that sin-
hybridized microarrays, Bead Arrays (Illumina, gle-stranded DNA molecules that vary in length
San Diego, California) offer a spotted substrate for by just a single nucleotide can be separated from
the high-density detection of target nucleic acids. one another by using polyacrylamide gel electro-
Though, instead of glass slides or silicon wafers as phoresis. The stable laser beam excites the fluores-
direct substrates, Bead Arrays rely on 3 µm silica cently labeled DNA bands and the light radiated
beads that arbitrarily self-assemble onto one of is noticed by sensitive photodetectors. Automated
two available substrates: the Sentrix Array Matrix DNA sequencing is freely automated by a variation
(SAM) or the Sentrix Bead Chip (Oliphant et  al., of Sanger’s sequencing method in which dideoxy-
2002; Fan et  al., 2006). Nothing like the other nucleosides castoff for each reaction is labeled with
array, the special characteristics of Bead Arrays a differently colored fluorescent tag. Several com-
depend on passive transport for the hybridiza- mercial and noncommercial software packages
tion of nucleic acids. One additional type of array, can trim low-quality DNA traces automatically.
electronic microarrays, exploits active hybridiza- DNA sequencing is still painstakingly the golden
tion via electric fields to control nucleic acid trans- standard and the final experimental procedure for
port. Microelectronic cartridges (NanoChip 400; mutation detection. However, the costs for the ini-
Nanogen, San Diego, California) modify comple- tial investment and the difficulties for standardiza-
mentary metal-oxide semiconductor technology tion and interpretation of ambiguous results have
for the electronic addressing of nucleic acids. restricted its use to basic research laboratories.
Because the chain-terminator tactics or Sanger
Traditional methods and approach is more efficient and this procedure
next-generation sequencing involves fewer toxic chemicals and lower amounts
Two dissimilar approaches for sequencing DNA of radioactivity than the method of Maxam and
were established in 1977: the chain termination Gilbert, it swiftly became the method of choice.
method and the chemical degradation method. In The declaration of the first draft sequence of the
1976–1977, A. Maxam and W. Gilbert developed human genome in February 2001 (International
a DNA sequencing technique based on chemi­ Human Genome Sequencing Consortium, 2001;
cal  alteration of DNA and successive cleavage at Venter et  al., 2001) and then with the genomic
specific bases. Maxam–Gilbert sequencing swiftly sequence of other organisms, molecular biology has
became more recognized, as purified DNA could moved into a new era with exclusive opportunities
be utilized directly, though the initial Sanger tech- and challenges. Technology has improved rapidly
nique required that every read initiated be cloned in the last two decades and new mutation-detection
for making of single-stranded DNA. Though, with techniques have been claimed, whereas old meth-
the enhancement of the chain-termination method, odologies have advanced to fit the increasing need
Maxam–Gilbert sequencing has dropped out of for automated and high-throughput screening.
150  Precision medicine in ovarian carcinoma

The chromatographic screening of polymorphic Genuine sequencing is accomplished by hybrid-

changes of disease-causing mutations by utilizing izing a primer complementary to the adapter
denaturing high-performance liquid chromatog- sequence, then cyclically adding DNA polymerase
raphy (DHPLC) is one of the novel technologies and a mixture of four differently colored fluores-
that occurred. DHPLC discloses the existence of cent reversible dye terminators to the captured
a genetic variation by the differential retention of DNA in the flow cell.
homo- and heteroduplex DNA on reversed-phase By using this technique, nonaltered DNA
chromatography under partial denaturation. fragments and unincorporated nucleotides are
Single-base substitutions, deletions, and insertions washed away, while apprehended DNA fragments
can be identified effectively by ultraviolet (UV) or are extended one nucleotide at a time. After each
fluorescence monitoring within 2 to 3 minutes in nucleotide-coupling cycle takes place, the flow cell
unpurified PCR products as large as 1.5 kilobases. is scanned, and digital images are developed to
These characteristics, together with its low cost, record the locations of fluorescently labeled nucle-
make DHPLC one of the most potent techniques otide amalgamations. Next to imaging, the fluores-
for mutational analysis. cent dye and the terminal 3′ blocker are chemically
detached from the DNA before the next nucleotide
HIGH-THROUGHPUT SEQUENCING coupling cycle.
TECHNOLOGIES The Illumina method is the most extensively
Sanger sequencing and added sequence analysis used NGS platform, but it is restricted by a rela-
methods have augmented sequencing outputs by tively low multiplexing capability (Erlich et  al.,
orders of magnitude and driven down per-base 2008). The Illumina system has been applied in
sequencing cost considerably (Church, 2006; programs for gene innovation, whole exome analy-
Roukos, 2010). It is now normally estimated that sis, and SNP detection by resequencing (Margulies
NGS will permit the in-depth characterization of et al., 2005).
the cancer cell genome and further improvement
in the fields of molecular pathology and personal- Roche second-generation sequencing
ized medicine for patients with cancer. First presented in 2005, the 454 Genome Sequencer
FLX Titanium System (Roche, Branford,
Illumina’s Genome Analyzer Connecticut) NGS platform regulated by highly
The massively parallel signature sequencing parallel PCR reactions happening in minute emul-
advanced by Lynx Therapeutics was the second sions consists of a primer-coated bead with a single
or next-generation approach to DNA sequencing. captured DNA template covered with the DNA
The elementary Lynx Therapeutics platform was polymerase, oligonucleotide ligation after PCR
a microsphere (bead)-based system that discovers amplification in an emulsion primers, and nucleo-
nucleotides in groups of four via an adapter liga- side triphosphates (NTPs) vital for PCR in an oil
tion and adapter decoding strategy using reversible droplet. PCR amplification results in each bead
dye terminators (Mardis, 2008). Lynx Therapeutics becoming covered with a single DNA amplicon. The
(Hayward, California) merged with Solexa, which emulsions are cracked, and the DNA-coated beads
was later acquired by Illumina. This short-read are loaded onto an array of picoliter wells for the
sequencing technique is today incorporated into sequencing reaction (Tawfik and Griffiths, 1998).
a fluidic flow cell design (HiSeq and Genome Pyrosequencing is accomplished over the pico-
Analyzer systems, Illumina, San Diego, California) liter well array and the nucleotide additions are
with eight individual lanes. The flow cell surface is visualized and located by a fiberoptic-coupled
established with capture oligonucleotide anchors, imaging camera. The system provides longer read
which hybridize the properly modified DNA seg- lengths than other NGS technologies, a strength of
ments of a sequencing library generated from a this system (Margulies et al., 2005).
genomic DNA sample.
By a process called bridge amplification, Helicos HeliScope: Single-molecule
engaged DNA templates are amplified in the flow sequencing
cell by bending over and hybridizing to a contigu- The HeliScope platform includes fragmenting the
ous anchor oligonucleotide primer (Mardis, 2008). sample DNA and performing polyadenylation at
 Advancement of cutting edge technology for accurate diagnosis  151

the 3′ ends of the fragments. Denatured polyad- associated with the four-color optical detection,
enylated strands are occupied by hybridization to currently used in all other NGS platforms. The
poly (dT) oligonucleotides immobilized on a flow- Ion Torrent method relies on standard DNA poly-
cell surface. This technique was the first process merase sequencing with unchanged dNTPs but
to effectively perform single-molecule sequenc- uses semiconductor-based screening of hydrogen
ing (Braslavsky et al., 2003). The flow cell is cycli- ions liberated during every cycle of DNA polym-
cally swamped up with the fluorescently labeled erization. Every nucleotide incorporated into the
deoxynucleoside triphosphates (dNTPs) in the budding complementary DNA strand causes the
existence of DNA polymerase, which incorporates release of a hydrogen ion that is detected by a hyper-
nucleotides from the oligo-dT primer. The flow sensitive ion sensor. The initial Ion Torrent system
cell is imaged at each cycle using a CCD camera, has relatively low parallelism, so it has a tendency
allowing the documentation of the location of to be concentrated on short sequence determina-
each nucleotide incorporation event. As in other tion of mutation hot spots throughout the genome.
systems, the fluorescent label is cut and washed All NGS platforms have high entry costs, but all
away before each succeeding sequencing cycle. also have the probability to dramatically reduce the
(Braslavsky et al., 2003; Margulies et al., 2005). The cost of comprehensive genomic profiling of cancer
HeliScope system is precise enough to provide the cells in the forthcoming years. Some techniques
most nonbiased DNA sequence, which is its power, suggest speed, such as the 454 Pyro sequencing
although relative to competing NGS platforms, it and Ion Torrent platforms, but, compared with the
has relatively high NTP incorporation error rates Illumina and SOLiD platforms, may not be as well
(Margulies et al., 2005). suited for clinical somatic tumor DNA sequencing
due to their relatively restricted capacities for sup-
SOLiD sequencing: porting highly parallel, deep sequencing.
Sequencing-by-ligation approach
SOLiD sequencing (Supported Oligonucleotide Somatic and germ-line mutation/
Ligation and Detection) is based on DNA ligase– variance studied in ovarian carcinoma
mediated oligonucleotide ligation after PCR
amplification in an emulsion format. The prim- Liede et  al. (1998) acknowledged in 1 out of 8 of
ers in SOLiD NGS are progressively offset to allow the ovarian cancer cases, the 185delAG mutation
the adapter bases to be sequenced when utilized in the BRCA1 gene (113705.0003) segregated with
in conjunction with the color-space coding for the cancer. Liede et al., established that site-specific
defining the template sequence by deconvolution. ovarian cancer families perhaps signify a variant of
Fluorescent signals are taken by CCD camera the breast-ovarian cancer syndrome and have char-
imaging before enzymatic cleavage of the ligated acteristic mutation in either BRCA1 or BRCA2.
probes and, after washing, repeating the sequenc- Stratton et al. (1999) showed a population-based
ing process. The SOLiD tactic has been used in study to conclude the involvement of germ-line
applications similar to the Illumina NGS, including mutations, they reported 2 out of 101 women with
whole genome sequencing, whole exome capture, invasive ovarian cancer had germ-line mutations in
and sequencing and SNP finding. Strengths of the the MLH1 gene (120436), and no germ-line muta-
SOLiD approach include reduction in sequencing tions were recognized in any of the other genes
error rates relative to the Illumina NGS by using explored, including BRCA1, the ovarian cancer
two-base encoding. A drawback of the SOLiD sys- cluster region (nucleotides 3139-7069) of BRCA2,
tem has been its relatively long run times and com- and MSH2. This study concluded that germ-line
plex analysis requirements (Margulies et al., 2005). mutations in BRCA1, BRCA2, MSH2, and MLH1
add to only a lesser of cases of early-onset epithelial
Ion Torrent sequencing: Ion ovarian cancer (Stratton et al., 1999).
semiconductor sequencing Cesari et  al. (2003) recognized the whole
This platform was picked up by Life Technologies, PARK2 gene (602544) within a Loss of Heterozy­
which suggests that this postlight sequencing tech- gosity (LOH) region on chromosome 6q25-q27.
nology has the utmost significant benefit of being LOH investigation of 40 malignant breast and
the first platform to eliminate the cost and effort ovarian tumors showed a shared minimal region
152  Precision medicine in ovarian carcinoma

of loss, including the markers D6S305 (50%) and CTNNB1, ARID1A, and PIK3CA mutations were
D6S1599 (32%), both of which are located within commonly seen in endometrioid ovarian can-
the PARK2 gene. Further, they found two somatic cer histology, KRAS mutations were prevalent in
truncating deletions in the PARK2 gene (see, e.g., mucinous types (Wiegand et al., 2010).
602544.0016) in 3 of 20 ovarian cancers. This Bell et al. (2011) completed a study based on 316
report proved that PARK2 may work as a tumor HGS-OvCa samples and compared with normal
suppressor gene (Cesari et al., 2003). samples for each individual. The study was based
Sellar et al. (2003) demonstrated that D11S4085 on captured 180,000 exons from 18,500 genes
on 11q25 is positioned in the second intron of the totaling 33 megabases of nonredundant sequence.
OPCML gene (600632). OPCML was commonly Enormously parallel sequencing by using the
somatically disabled in epithelial ovarian cancer Illumina GAIIx platform or ABI SOLiD 3 plat-
tissue by allele loss and by CpG island methylation. form (80 sample pairs) obtained 14 gigabases per
OPCML behaves like a tumor suppressor gene as sample bases in total. TP53 was mutated in 303 of
proved both in vitro and in vivo. A somatic mis- 316 samples. BRCA1 and BRCA2 had germ-line
sense mutation from an individual with epithelial mutations in 9% and 8% of cases, respectively, and
ovarian cancer represents perfect proof of loss of showed somatic mutations in a further 3% of cases.
function (600632.0001) (Sellar et al., 2003). Further, this group also demonstrated the presence
The Cancer Genome Atlas Project has com- of other mutated genes: RB1, NF1, FAT3, CSMD3,
pleted whole exome sequencing on ovarian can- GABRA6, and CDK12.
cer (Cancer Genome Atlas Research Network Recently a study reported 11,479 somatic muta-
2011). Screening DNA from 316 high-grade serous tions in the 142 fresh TCGA cases. These muta-
ovarian cancer patients and compared with nor- tions were manually reviewed, resulting in a total
mal for 19,356 somatic mutations (about 61 per of 27,280 mutations in 429 cases. TP53, NF1, RB1,
tumor) were interpreted. High-grade serous ovar- CDK12(CRKRS), and BRCA14, as well as the novel
ian cancer was recognized by TP53 mutations in SMG, KRAS. BRCA2 and RB1CC1 were reported
almost all tissues (96%). BRCA1 and BRCA2 were significantly associated. This group also identified
mutated in 22% of tumors, as the study considered 4 NRAS mutations; 3 NF2 mutations; and 3, 8, and
a mixed type of germ-line and somatic mutations. 10 mutations in the identified tumor suppressor
Additional significant mutated genes, including genes: ATR, ATM, and APC, respectively. Somatic
NF1, RB1, FAT3, CSMD3, GABRA6, and CDK12, truncation mutations were also detected in his-
occurred in 2%–6% of cases. Mutational analysis tone modifier genes including ARID1A, ARID1B,
also presented that mutations in BRAF, PIK3CA, ARID2, SETD2, SETD4, SETD6, JARID1C, MLL,
KRAS, and NRAS may be central forces in high- MLL2, and MLL3 as well as the DNA excision
grade serous carcinoma. For instance, clear cell repair gene ERCC6 (Kanchi et al., 2014). Tables 8.1
types have few TP53 mutations but have recur- and 8.2 have shown the various mutations as char-
rent ARID1A and PIK3CA mutations. Although acterized by various researchers.

Table 8.1  Mutations predicted and verified by various studies

Number of Number of
Gene mutations proposed mutations verified Reference
TP53 302 294 Bell et al. (2011), Nick et al. (2015)
BRACA1 11 10 Bell et al. (2011), Couch et al. (2013)
BRACA2 10 10 Bell et al. (2011), Kanchi et al. (2014)
CSMD3 19 19 Bell et al. (2011), Kanchi et al. (2014)
NF1 13 13 Bell et al. (2011), Sangha et al. (2008)
CDK12 9 9 Bell et al. (2011), Dong, Lu, and Lu (2016)
FAT3 19 18 Bell et al. (2011), Kanchi et al. (2014)
GABRA6 6 6 Bell et al. (2011), Kanchi et al. (2014)
RB1 6 6 Bell et al. (2011), Kanchi et al. (2014)
 Advancement of cutting edge technology for accurate diagnosis  153

Table 8.2  Prevalence of somatic mutations in epithelial ovarian cancer

Epithelial High grade

Gene ovarian serous Low grade Clear cell Endometrioid Mucinous
mutation cancer overall (type II) serous (type I) (type I) (type I) (type I)
BRAF 11% (Kurman 24%–33%
<1% (Bell 1% (Kuo 24% (Singer 50%–75%
and Shih, et al., (Nakayama et al., 2009; et al., 2003) (Gemignani
2011) 2011) et al., 2006; Zannoni et al.,
Singer et al., et al., 2016) 2003)
2003)
KRAS 11% Kurman <1% (Bell 33% <1%–7% <1% (Singer 50%–75%
and Shih, et al., (Nakayama (Kuo et al., et al., 2003) (Gemignani
(2011) 2011) et al., 2006; 2009; et al.,
Singer et al., Singer 2003)
2003) et al.,
2003)
PIK3CA 6.7% (Levine <1% (Bell 5% (Nakayama 20%–33% 20%–33% (Cho Rare
et al., 2005; et al., et al., 2006; (Kuo et al., and Shih,
Campbell 2011) Wu et al., 2009) 2009; Samuels
et al., 2005) 2013) and Waldman,
2010)
PTEN​ 20% (Kurman <1% 20% (Landen <1%–5% 20%–31% Rare
and Shih, mutation et al., 2008) (Kuo et al., (Kurman and (Obata
2011) (Bell 2009; Shih, 2011; et al., 1998)
et al., Willner Willner et al.,
2011) et al., 2007;
2007) McConechy
et al., 2014)

Transcriptomics ultimately could be very fruitful in management of

these patients (Dwivedi et al., 2013). Currently, one
For proper functioning of our biological system, tactic broadly utilized to assess the gene expression
genes expression is needed in precise quality and in terms of copy number, that is, a semiquantita-
quantity so that smooth processing of all activities tive approach by the real-time PCR. In this process,
could be maintained. Modification or change may first, isolation of total RNA is required from any
happen in the controlling region of the particular sample like blood or tissue and then converted into
gene so the product of that gene may be abnormal cDNA by reverse transcriptase. Thus the converted
in quantity, thus deregulation of governing func- cDNA is quantified by using a fluorescently labeled
tion may occur. Commonly this abrupt regulation probe or the SYBR green-based method.
may manifest in two ways—nonsense-mediated This method permits researchers to evaluate
decay (NMD) or ubiquitin-mediated decay—and expression of genes related to pathways or patho-
may cause carcinogenesis. A few studies suggested genesis of patients, and further help to compare the
that this upregulation of gene expression may data with healthy individuals in a very short time.
advance the stage of the cancer (Dwivedi, Goel, For example, one study looked at gene expression
Khattri, et  al., 2015; Dwivedi, Goel, Mandhani, by quantifying the mRNAs of the genes that have
et al., 2015; Dwivedi, Singh, et al., 2015) a role in the developmental and hormonal regula-
Exploring a large sample size of cancer specimen tion of the transmembrane TJ protein, occludin
can establish a gene expression database that can be (OCLN), and the cytoplasmic TJ proteins, TJ pro-
used not only for molecular subtyping but also for tein 1 (TJP1) and cingulin (CGN) in bovine gran-
screening the advancement of the disease, which ulosa cells (GC) and theca cells (TC) and found
154  Precision medicine in ovarian carcinoma

altered gene expression of these genes during ovar- progression of ovarian carcinoma, but these factors
ian cancer (Zhang et al., 2017). alone are not responsible for carcinogenesis mecha-
An additional standard technique for transcrip- nism. So, it is now believed that epigenetic modifica-
tome profiling is microarray assays. Microarray tions have a role in carcinogenesis. Epigenetics can
exploration means extracting the total RNA from be defined as the potent permanent and inheritable
samples and transforming it into cDNA as real- alteration in gene expression that does not affect
time PCR. The microarrays are dependent on in DNA sequence but alter the morphology of the gene
situ hybridization of complementary nucleotide or chromosome. Epigenetic modifications to the
strands unlike gene polymerization in real-time genome generally take place during normal cell cycle
PCR. DNA spots are fabricated on the microarray regulation. More notably, a study has revealed these
surface and every spot includes a custom deliber- modifications tend to occur more commonly than
ated DNA sequence that works as a probe for spe- mutations. Epigenetic modifications among can-
cific gene detection. The sample has fluorescently cer cells effect abnormal gene expression via three
labeled cDNA and when this is combined with their process: (1) DNA methylation, (2) histone modifica-
corresponding spots on the microarray, a fluorescent tions, and (3) noncoding microRNAs. These modi-
signal is radiated based on the quantity of cDNA fications have shown a significant association with
bound to the probe DNA. Microarray assays can initiation and progression of ovarian cancers.
be accomplished in a swift and economical manner Two major types of epigenetic controls gener-
that proves it a potent tool for medical transcrip- ally exist: the first ensues at the gene level and a
tome profiling of patient tumor biopsies and can also second that arises at the chromosomal level. The
detect patient samples for distinctive subtype-spe- first is recognized as genomic methylation in
cific gene subtypes that can assist in forecast therapy which “methyl” groups is added to definite locus of
response, tumor progression, and patient prognosis. genes, which can either have active or inactive gene
Recently, a study by Carrarelli et  al. (2017) on expression. DNA methylation happens among
myostatin expression compared endometrium cytosine residues in cytosine–guanine (CpG)
in benign (endometriosis, polyps) and malignant dinucleotides, which are frequently dispersed in
(endometrial adenocarcinoma) patients and with the CpG-rich regions mentioned as “CpG islands.”
healthy women during the menstrual cycle. As This type of methylation is accomplished by DNA
myostatin is a growth factor member of the trans- methyltransferases (DNMTs), which are a family of
forming growth factor β superfamily, which is rec- enzymes that assist to transfer methyl groups onto
ognized to play main roles in cell proliferation and DNA. In humans, DNMTs are classified into two
differentiation. The current data demonstrated for groups: DNMT1 and DNMT3. Alterations in DNA
the first time the expression of myostatin in healthy methylation regulating gene expression are very
endometrium and aggravated copy number common and appear in both normal and cancer-
expression in endometriosis and endometrial can- ous cells. As per assumptions, about 80% of CpG
cer, proposing myostatin involvement in human dinucleotides in the human genome get methyl-
endometrial physiology and related pathologies. ated during the lifespan (Nguyen et al., 2014).
Similarly, Wei et al. (2017) reported high expres-
sion of ITGA6 (Integrin subunit alpha 6) in 287 ovar- EPIGENETIC ALTERATIONS IN CANCER
ian cancer patients of the TCGA cohort, which was DNA hypermethylation has a role in gene silencing,
significantly associated with poorer progression- whereas DNA hypomethylation is known for alter-
free survival. This study provided the basis of drug ation in gene expression. Both have been reported
resistance in ovarian cancer, and integrins could be in malignancy and cancer cells. Commonly,
a probable biomarker for prognosis of ovarian can- hypermethylated CpG regions within the DNA
cer. Table 8.3 shows the recent studies that proved cause silencing of tumor suppressor genes, demol-
significant gene expression in ovarian carcinoma. ishing the cell’s capability to repair DNA dam-
age, control cell growth, and inhibit proliferation.
Epigenetic (DNA methylation analysis) Whereas, DNA hypomethylation participates in
oncogenesis when formerly silenced oncogenes
As mentioned earlier, mutations, copy number, become transcriptionally activated. Furthermore,
and gene expression play a role in development and DNA hypomethylation has a role in triggering
 Advancement of cutting edge technology for accurate diagnosis  155

Table 8.3  Recent study on exploration of gene expression in ovarian carcinoma

Diagnostic/prognostic/ Types of
Gene Expression therapeutics study Reference
E74-like factor 5 Decreased Therapeutics (Gene Human sample Yan et al. (2017)
(ELF5) Therapy) study
Receptor for advanced Increased Diagnostic Human sample Rahimi et al.
glycation end study (2017)
products (RAGE)
RAP80 (HR-pathway- Decreased Poor prognosis Human sample Romeo et al.
related gene) study (2017)
Musashi-2 (MSI2) Increased Therapeutics (paclitaxel Human sample Lee et al. (2016)
resistance), poor study
prognosis
DNA-PKcs, Akt3, and Increased Poor prognosis Human sample Shin et al. (2016)
p53 study
Aberrant ALK Present in Prognosis Human sample Tang et al. (2016)
(anaplastic ovarian serous study
lymphoma kinase) carcinoma
YAP (autophagy Present Therapeutics (cisplatin- C13 K cell line L. Xiao et al.
related) resistant; protective) (2016)
Urothelial carcinoma Increased Poor prognosis and Human sample Zhang et al.
associated 1 (UCA1) therapeutics study (2016)
Cyclin Y (CCNY) Increased Therapeutics Cell line H. Liu et al. (2016)
Spalt-like transcription Increased Poor prognosis Cell line Yang, Xie, and
factor 4 (SALL4) Ding (2016)
Carnitine Increased Poor prognosis, Cell line Shao et al. (2016)
palmitoyltransferase therapeutics
1A (CPT1)

latent transposons and thus chromosomal insta- while working on lin-14 gene, which regulates the
bility take place in specific pericentromeric satel- development of nematodes. Mature miRNA is a
lite regions (Sapiezynski et al., 2016). class of endogenous, noncoding, single-stranded
small RNA, which is composed of about 20–22
EPIGENETIC ROLE IN OVARIAN CANCER nucleotides (Lehrbach and Miska, 2008). miRNA
Epigenetic modifications have shown their potent is involved in several physiological processes and
promise to be utilizing as biomarkers not for early is expressed aberrantly in many pathological con-
diagnosis but also to screen the advancement of ditions. These aberrant expressions of miRNA are
the disease and prognosis. Epigenetic databases meticulously linked to the manifestation, devel-
derived from the patient samples along with muta- opment, progression, diagnosis, and prognosis
tions and gene expression could be used in better of the human disease. The role for miRNAs in
management of the cancer. Table 8.4 summarizes cancer was initially reported from the labora-
the recent studies on epigenetic changes associated tory of Carlo Croce. A bicistronic miRNA cluster
with ovarian cancer. comprising miR-15a and miR-16 at chromosome
13q14 was detected to be mutated, deleted, or have
Micro-RNA expression in ovarian reduced expression in chronic lymphocytic leu-
cancer kemia. Afterward, germ-line mutations in miR-
15a/-16 were screened and it was proved that both
The exclusive biomolecule micro-RNA (miRNA) of these target antiapoptotic BCL-2 mRNA (Calin
was revealed by Victor Ambros et  al., in 2011 et al., 2006).
156  Precision medicine in ovarian carcinoma

Table 8.4  Recent study on exploration of epigenetic changes in ovarian carcinoma

Methylation/ Diagnostic/prognostic/
Gene acetylation therapeutics Reference
GATA Hypermethylation Better prognosis Bubancova et al. (2017)
miR-128-2 Methylation Progression of the Jang et al. (2017)
cancer
RGS2 (regulator of Methylated Therapeutics Cacan (2017)
G-protein signaling 2) (chemoresistance)
4-1BB ligand (4-1BBL/ Histone deacetylation Therapeutics Cacan (2017)
CD157) and OX-40 and DNA (chemoresistance)
ligand (OX-40L/CD252) methylation
Solute carrier family 6, Methylation Poor prognosis and Sung et al. (2017)
member 12 (SLC6A12) advancement of the
cancer
HNF1B Methylation Poor prognosis and Ross-Adams et al.
advancement of the (2016)
cancer
TBX15 Promoter methylation Therapeutics Gozzi et al. (2016)
P16INK4a Methylation Risk, diagnosis Xiao et al. (2016)
DAPK1 and SOX1 Promoter methylation Risk, diagnosis Kaur et al. (2016)
Cadherin 13 (CDH13), Promoter Diagnostic and Y. Xu et al. (2016)
Dickkopf WNT signaling hypermethylation advancement of the
pathway inhibitor 3 disease
(DKK3) and Forkhead
box L2 (FOXL2)

The alterations in miRNA expression can ensue Reversed-phase protein array (RPPA) is a high-
at both the DNA and RNA level. Modifications throughput antibody-based method that offers
in Dicer1/Ago2 expression effects global changes higher sensitivity, quantification, and multiplex-
in miRNA expression, though change of specific ing abilities than traditional immunoassay. The
miRNAs in cancer is also very frequent. This may Cancer Genome Atlas (TCGA) utilized the RPPA
happen via one of many mechanisms, like by germ- practice in many tumor types with numerous pro-
line mutation, deletion, or promoter methylation teins screened by RPPA method, but there were
(Calin et  al., 2005). In recent years, several novel restrictions due to unavailability of antibodies to
micro-RNA, as shown in Table 8.5, have been deci- detect isoform or phosphorylated antibodies with
phered and added into database, suggesting their various distinct functions. So, mass spectrometry
immense potential in ovarian cancer diagnosis, (MS) is now emerging as a technology of prefer-
prognosis, and therapeutics. ence for screening various proteins. Currently,
electrospray ionization-MS and matrix-assisted
Proteomics laser desorption/ionization (MALDI)-MS are the
main systems utilized in exploring protein profil-
The proteome is defined as the array of proteins ing, screening of posttranslational modifications,
expressed in a particular cell at a fixed set of condi- and meticulous quantification. Nowadays, mass
tions. In the human proteome more than a million spectrometry–based proteomics created many
proteins are produced. Initially, proteomic equip- milestones in terms of sensitivity, robustness, and
ment has progressed from gel-based techniques consistency.
(one- and two-dimensional SDS PAGE) to now gel- Additionally, the advancement of quantita-
free techniques of mass spectrometry. tive tactics has opened new vistas for exploring
 Advancement of cutting edge technology for accurate diagnosis  157

Table 8.5  Recent study on exploration of miRNA expression and their role in ovarian carcinoma

Increased/ Diagnostic/
decreased Involved mechanisms/ prognostic/
miRNA expression pathway therapeutics Reference
miR-125b Repression Cell migration by TGF-β Therapeutic target Yang et al. (2017)
let-7a and Deregulation Increased expression of High- Therapeutic target Agostini et al.
miR-30c mobility group AT-hook 2 (2017)
protein (HMGA2) i
miR-130b-3p Increased Cytidine monophosphate kinase Therapeutic target Zhou et al.
inhibition by the TGF-β (2017)
signaling pathway
miR-27b Increased Suppressed ovarian cancer cell Therapeutic target Liu et al. (2017)
migration and invasion by
binding with VE-cadherin
miR-196b Increased Invasion activities in recurrent Prognosis/ Chong et al.
epithelial ovarian carcinoma by therapeutic (2017)
regulating the HOXA9 gene target
miR-429 Increased Suppress zinc finger E-box Prognosis/ Zou et al. (2017)
binding homeobox 1 (ZEB1) therapeutic
and increased cisplatin target
sensitivity
miR-490-3p Increased Increased cisplatin (CDDP) Prognosis/ J. Tian et al.
sensitivity by downregulating therapeutic (2017)
ABCC2 expression target
let-7e Decreased Activation of BRCA1 and Rad51 Therapeutic target Xiao et al. (2017)
expression and enhancement
of DSB repair, which in turn
results in cisplatin resistance
miR-221 Increased Promotes cell proliferation by Diagnostic/ Li et al. (2017)
targeting the apoptotic prognostic/
protease activating factor-1 therapeutics
miR-28-5p Increased Progression of ovarian cancer Diagnostic/ J. Xu et al. (2017)
cell cycle, proliferation, prognostic/
migration and invasion, therapeutics
inhibited apoptosis, and
induced the process of EMT
through inhibition of N4BP1

the protein differential expression, and posttran- several areas of ovarian cancer research not only
scriptional and posttranslational modifications in deciphering mechanism and characterization of
in diverse conditions in an endeavor to compre- the biomarkers (diagnostics and prognostics) but
hend the functional significances of modified gene also in searching biomolecules involved in resis-
expression. Quantitative proteomics has perceived tance of the therapy (Sapiezynski et al., 2016).
major revolutions in precision and relative quan- The ovarian cancer patients do not exhibit any
tification methods, including spectral counting, specific symptoms during disease initiation and so
stable isotope labeling by amino acid in cell cul- when they are diagnosed the disease has progressed
ture, isotope-coded affinity tags, and isobaric tags into advanced stage. Currently for investigation
for relative and absolute quantification (iTRAQ). in the early stage, a biomarker with higher sensi-
Proteomic methods are now exclusively applied in tivity and specificity is warranted. Any standard
158  Precision medicine in ovarian carcinoma

Table 8.6  Current progress in proteomics-based biomarker

Subtype Markers Method Reference

Serous Wilm’s tumor 1 (WT-1) MALDI-TOF Zhu et al. (2006)
Ras-related protein (Rab-3D) MALDI-TOF Zhu et al. (2006)
Mesothelin LC-MS/MS Tian et al. (2011)
ICAM3, CTAG2, p53, STYXL1, Protein microarrays Katchman et al. (2017)
PVR, POMC, NUDT11,
TRIM39, UHMK1, KSR1, and
NXF3
Endometroid Estrogen receptor α (ERα) RPPA Sereni et al. (2015)
Clear cell Annexin-A4 (ANXA4) 2-DE, MALDI-TOF Toyama et al. (2012),
Zhu et al. (2006)
Napsin A Immuno-histochemistry Alshenawy and Radi
and tissue microarrays (2017), Skirnisdottir
et al. (2013)
Phosphoserine aminotransferase 2-DE Toyama et al. (2012)
(PSAT1)
Mucinous Serpin B5 (SPB5) 2-DE Toyama et al. (2012)
FOXA1 Immunohistochemistry Karpathiou et al. (2017)
REG4 Immunohistochemistry Lehtinen et al. (2016)
tissue microarrays
CEA5 LC-MS/MS Tian et al. (2011)
CEA6 LC-MS/MS Tian et al. (2011)

biomarker can be any DNA, RNA, or protein that with a higher specificity and sensitivity (Elzek and
may show its presence in body fluids like blood, Rodland, 2015). Table 8.6 summarizes the cur-
urine, and saliva. Several biomarkers have been rent progress in proteomics-based biomarkers that
characterized for a few types of cancer, but ovarian showed some potency to be a better biomarker.
cancer fails to show robustness that correlate with
ovarian tumor formation and progression. CURRENT SCENARIO IN OVARIAN
One such well-known biomarker is the CA-125 CANCER THERAPY
glycoprotein. Several researchers have claimed
that several patients with ovarian cancer showed Ovarian cancer remains one of the most lethal
increased levels of CA-125 (Sapiezynski et al., 2016). gynecological cancers with an estimated 225,000
In fact, various results showed that approximately women being diagnosed with it every year (Ferlay
78% (70%–90%) of patients with ovarian cancer et  al., 2015). The survival rate in ovarian cancer,
have raised levels of this glycoprotein. Thus due to however, is 40% over a period of five years. This is
its higher specificity in comparison to other known because most of the patients present at a later stage
biomarkers, this biomarker is recommended initially of the disease. The current treatment modality for
to presymptomatic women, though the strength of epithelial ovarian cancer is cytoreductive/interval
CA-125 as a biomarker for screening ovarian cancer debulking surgery (IDS) followed by platinum-
is disputed. There are several other conditions where based chemotherapy (Banerjee and Kaye, 2013).
the levels of CA-125 shoots up like in all inflammatory
diseases, cirrhosis, and liver diseases; diabetes mel- Cytoreductive surgery and interval
litus; and in cancers of endometrial, fallopian, lung debulking surgery (IDS)
cancers. False-positive outcomes weaken CA-125 to
be used as a standard biomarker. Therefore, there Cytoreductive surgery/tumor debulking surgery is
is need of some other potent biomarker that could done with the idea that removal of as much cancer
differentiate ovarian cancer from normal conditions as possible shall be useful for the patient, even when
 Current scenario in ovarian cancer therapy  159

complete resection is not possible. The primary the peritoneal cavity and remain confined to the
debulking surgery can be challenging in patients peritoneum and intra-abdominal viscera, HIPEC
of advanced (stage III and IV) ovarian cancer. It is can be useful in these patients. Improved long-
often a high-risk preposition, with extensive bowel term results can be achieved in highly selected
resection and major blood loss and even risk of patients using cytoreductive surgery (CRS), in
morbidity as reported in 2015 Cochrane reviews combination with intraoperative hyperthermic
comparing conventional, primary surgery with intraperitoneal chemotherapy (HIPEC) (Martín-
secondary IDS in ovarian cancer patients. Cameán et  al., 2016). Currently intraperitoneal
IDS is a secondary surgery that is performed cisplatin is validated in phase III trials but with
after two to four cycles of neoadjuvant chemother- increased nonhematologic toxicity; intraperito-
apy (NAC) or induction chemotherapy to remove neal carboplatin therapy phase III trial is under-
the bulk of the tumor, and followed by adjuvant way (Teo, 2014).
chemotherapy of the same type. Since most patients However, the advances in chemotherapy are
of ovarian cancer present in advance stages, IDS is marred with a significant risk of recurrence and
the surgery of choice. The term neoadjuvant che- resistance to therapy making ovarian cancer dif-
motherapy (NAC) describes the administration of ficult to cure. Hence, there is an urgent need to
chemotherapy when primary debulking surgery is develop smarter treatment options.
not feasible, and only a biopsy is done for histologic
diagnosis. If there is some tumor response after few Precision medicine in ovarian cancer
cycles of chemotherapy, then the secondary sur-
gery can be done before proceeding with further To improve outcomes of ovarian cancer, transla-
chemotherapy cycles (Tangjitgamol et al., 2010). tional research must be prioritized and acceler-
ated. For this it is important to identify multiple
Chemotherapy in ovarian cancer molecular abnormalities in human ovarian cancer
that can be used for earlier diagnosis or used as
Ovarian cancer although not chemocurable is therapeutic targets. Besides, the need of the hour
chemoresponsive. The 1960s and ’70s saw the is the enhancement of predictive models and bio-
use of single alkylating agents with little success. markers, developing therapeutic agents, and assays
Gradually there was the development of combina- targeting molecular abnormalities. Targeted/per-
tion chemotherapy, where initially a combination sonalized therapy is the latest approach for treat-
of cisplatin and cyclophosphamide was used, and ment of ovarian cancer. It utilizes the molecular
currently platinum compounds with paclitaxel are profile of a patient’s cancer for designing an effi-
the treatment of choice. Platinum compounds are cacious treatment plan. A variety of targeted
the most active cytotoxic agents currently used therapeutic approaches have been utilized for the
for the treatment of ovarian cancer. According management of ovarian cancer. The significant
to the American Cancer Society (ACS), the stan- ones are discussed in the subsequent sections.
dard approach is the combination of a platinum
compound, such as cisplatin or carboplatin, and ANTIANGIOGENIC THERAPY
a taxane, such as paclitaxel (Taxol®) or docetaxel Angiogenesis is crucial for tumor development
(Taxotere®). For IV chemotherapy, carboplatin is and progression. Clinical outcomes in ovarian
the choice of drug over cisplatin due to its fewer cancer are worsened with increased angiogen-
side effects and being as effective. Several other esis. Angiogenesis inhibition has been the choic-
drugs can produce regression of epithelial ovarian est therapeutic target for tumor control since the
cancers, including pegylated liposomal doxoru- 1970s. Vascular endothelial growth factor (VEGF)
bicin (PLD), gemcitabine, and topotecanin com- is the major component involved in angiogenesis
bined with paclitaxel and/or carboplatin with two (Sapiezynski et  al., 2016). VEGF overexpression
or three drugs in combination (Bast, 2011). leads to increased angiogenesis, and its association
Another important arm of chemotherapy in with ovarian cancer development and peritoneal
ovarian cancer is hyperthermic intraperitoneal dissemination is well established.
chemotherapy (HIPEC). Since epithelial ovar- The VEGF/VEGF receptor (VEGFR) signaling
ian cancer has the tendency to disseminate into pathway is the most promising angiogenic target.
160  Precision medicine in ovarian carcinoma

There are two major strategies for inhibition of group (39.7 vs. 30.3 months; HR = 0.78 (95%
angiogenesis in ovarian cancer tumor treatment. CI = 0.63–0.97); p = 0.03).
The ICON-7 trial however also reported certain
Inhibition of vascular endothelial growth side effects that need to be considered before mak-
factor (VEGF) ligand with antibodies or ing a decision on a combination of bevacizumab
soluble receptors with cytotoxic drugs. The treatment with bevaci-
Inhibition of VEGF receptor with tyrosine zumab was associated with an increase in grade 1
kinase inhibitors mucocutaneous bleeding, grade 2 or higher acute
hypertension (18% vs. 2%), grade 3 or higher
Elevated expression of the VEGF ligands and
thromboembolic events (7% vs. 3%), and gastro-
receptors promotes malignant progression and
intestinal perforation (10 cases vs. 3 cases). The
correlates with poor prognosis in EOC. There are
questionnaire-based quality of life scores showed
many drugs directed toward the VEGF/VEGFR
that continuation of treatment with bevacizumab
signaling pathway like bevacizumab, trebananib,
led to a small but clinically significant decline in
pazopanib, nintedanib, cediranib, and aflibercept.
quality of life compared to standard chemotherapy
Inhibition of VEGF ligand with (Bermejo et al., 2016).
The second trial was the GOG-218 trial which
antibodies or soluble receptors
randomly recruited 1873 women patients at stage
Bevacizumab III or IV of ovarian cancer. The study participants
Bevacizumab was developed by Roche and is a had undergone cytoreduction surgery and were
recombinant humanized monoclonal antibody divided into three groups according to their drug
capable of binding to all VEGF isoforms, thus combination. All patients received carboplatin
inhibiting the binding of VEGF to its receptor and AUC 6 and paclitaxel 175 mg/m2 every 3 weeks for
the downstream signaling (Conteduca et al., 2014). 6 cycles and the study treatment. The three study
It is one of the best antiangiogenic agents since its groups then received drugs as follows: Group  I
action leads to the reduction in vascularization received placebo every 3 weeks from cycle  2 to
and normalization of the tumor microenviorn- cycle 22; group II received bevacizumab every
ment without any cytotoxicity. Bevacizumab has 21 days from cycle 2 to cycle 6 followed by placebo
been used in combination with carboplatin and from cycle 7 to cycle 22; and group III received
paclitaxel (Conteduca et al., 2014). bevacizumab at the same dose from cycle 2 to cycle
There have been two major clinical trials on 22. The objectives were the same as ICON-7 trial.
bevacizumab, namely, international collaboration The results showed the arm with bevacizumab
on ovarian neoplasms 7 (ICON-7) and gynecologic (bevacizumab throughout) compared with the
oncology group study 218 (GOG-218). Both these standard chemotherapy arm (placebo arm) showed
trials are in phase III. The ICON-7 trial enrolled a statistically significant increase in PFS. However,
1528 women randomly to receive carboplatin and in the bevacizumab initiation group there was no
paclitaxel every 3 weeks for 6 cycles, or the same increase in PFS. OS was similar in the three groups:
regimen plus bevacizumab every 3 weeks dur- 39.3, 38.7, and 39.7 months, respectively, with no
ing chemotherapy, followed by 12 cycles or until statistically significant differences (Bermejo et al.,
­unacceptable toxicity or disease progression. The 2016). Compared to the ICON-7 trial, the GOG-
objectives were to measure progression-free sur- 218 trial had just one significant side effect, that is,
vival (PFS), overall survival (OS), response to grade 2 hypertension in bevacizumab as compared
treatment, toxicity, and quality of life. Final analy- to the placebo group.
sis with a 49-month median follow-up, showed an Avastin use was studied in the platinum-
increase in PFS in the high-risk-of-progression resistant epithelial ovarian cancer study
group, with an increase of 5.5 months (16.0 vs. (AURELIA). It is a phase III trial where the
10.5 months; HR = 0.73 (95% CI = 0.61–0.88); combination therapy of paclitaxel, topotecan, or
p = 0.001) compared to placebo group without liposomal doxorubicin is being used with bevaci-
bevacizumab. Further there was an increase of zumab. The addition of bevacizumab showed an
9.4 months in OS in the high-risk-of-progression improved median PFS as well as OS (16.6 months)
 Current scenario in ovarian cancer therapy  161

(Sapiezynski et  al., 2016). Bevacizumab is yet use of pazopanib is not currently recommended in
to be approved by the U.S. Food and Drug ovarian cancer management.
Administration (FDA) as the first line of treat- Some other significant antiangiogenic thera-
ment in ovarian cancer, although it is approved peutics include trebananib (a peptibody that
as the first line treatment in combination therapy. inhibits angiopoietin 1 and 2), nintedanib (inhib-
VEGF signaling involves the tyrosine kinases its VEGFR-1, VEGFR-2, and VEGFR-3; PDGFR α
and inhibition of these tyrosine kinases have and β; and FGFR-1, FGFR-2, and FGFR-3), cedira-
proven to be another promising area of ovarian nib (inhibits (VEGFR 1, VEGFR2, and VEGFR3),
cancer treatment. One such inhibitor is pazo- and c-Kit and aflibercept/VEGF trap (binds to and
panib, an orally administered multikinase inhibi- neutralizes all forms of VEGF-A and VEGF-B and
tor of VEGFR-1/-2/-3 and of platelet-derived inhibits placental growth factor [PGF] activation)
growth factor receptor (PDGFR)-α/-β and of (Friedlander et al., 2010).
c-Kit (Sapiezynski et al., 2016). Invasive ovarian cancer tumors cover a spec-
Pazopanib has been studied in combination trum from low-grade to high-grade malignancy,
therapy in phase I/II trials to evaluate the safety with differences in their morphologic, histologic,
and efficacy of paclitaxel (175 mg/m2) plus carbo- and clinical features. Further genomic analy-
platin (AUC 5 [group A] or AUC 6 [group B]) once sis has revealed that low-grade tumors and low-
every 3 weeks for up to 6 cycles, with either 800 malignancy potential (LMP) tumors are a separate
or 400 mg per day of pazopanib. However, drug class of tumors as compared to high-grade malig-
limiting toxicities were encountered in 4 out of 6 nancy (HGM) tumors due to a difference in their
patients receiving pazopanib in combination ther- mutated genes (high TP53, its downstream effector
apy at low dose (400 mg) and higher dose (800 mg), CDKN1A, activators of p53, such as PPM1A, and
which included gastrointestinal p ­ erforations and decreased levels of inhibitors of p53, UBE2D1 and
myelotoxicities (Friedlander et al., 2010). ADNP in LMP tumors versus high expression of
Further pazopanib was also studied as mono- PDCD4, E2F3, MCM4, CDC20, and PCNA in HGM
therapy in patients with recurrent epithelial ovar- tumors (Bast and Mills 2010).
ian/fallopian tube/primary peritoneal cancer. The Based on the gene expression profile, two types of
study was a nonrandomized, multicentric phase ovarian cancer can be defined (Cerrato et al., 2016):
II trial (VEG104450; NCT00281632). The primary
objective of the study was to study the response Type I—Low-grade and borderline serous cancers,
rate (determined by normalization of CA-125 lev- endometrioid, mucinous and clear-cell tumors
els) and secondary objectives were overall response with frequent mutations in PTEN, PI3 K cata-
and PFS. These patients were treated with pazo- lytic subunit-α (PIK3CA), KRAS, BRAF, and
panib (800 mg OD) until progression or toxicity. b-catenin (CTNNB1) genes
At the end of 113 days the overall response rate (as Type II—High-grade serous carcinomas, mixed
per CA 125) was seen in 18% patients (Friedlander malignant mesodermal tumors, carcinosarco-
et  al., 2010). The overall response rate in mono- mas, and undifferentiated cancers with high
therapy paved the way for maintenance therapy of genomic instability and up to 80% of patients’
epithelial ovarian, fallopian tube, or primary peri- TP53 is mutated and is characteristic of BRCA1
toneal cancer, stages II to IV with no progression of and BRCA2 mutations (fallopian tubes and the
disease postsurgery. This increased PFS compared peritoneum cancer)
with the placebo (17.9 vs. 12.3 months, HR = 0.77;
95% CI = 0.64–0.91; p = 0.0021). Grades 3 and 4 Unfortunately, these differences in the tumor
adverse events were observed, including hyperten- heterogeneity do not make any change in the treat-
sion (30.8%), neutropenia (9.9%), transaminase ment regimen of the veritable types of tumors.
elevation (9.4%), diarrhea (8.2%), fatigue (2.7%), Thus currently, with advances in the knowledge on
thrombocytopenia (2.5%), and palmoplantar the molecular basis of tumor genesis, new drugs
erythrodysesthesia (1.9%) in the pazopanib arm. are under trial targeting various mutated genes.
Although the maintenance therapy improved Some of the significant ones are discussed in the
median PFS, OS has not yet been shown. Thus the text to follow.
162  Precision medicine in ovarian carcinoma

DRUGS TARGETING MUTATIONS IN is a potent enhancer of D269H/E195R-induced

OVARIAN CANCER PRECISION MEDICINE apoptosis in wild-type p53-expressing cancer cells.
Poly(ADP ribose) polymerase (PARP) These results were replicated by Crane et al. (2015)
inhibitors who further showed it in 15 ovarian cancer cell
lines of varying histologic subtypes and demon-
Poly(ADP-ribose) polymerase (PARP) is a crucial strated apoptosis in sensitive cell lines treated with
enzyme involved in the base excision repair path- Nutlin-3a.
way of DNA repair mechanism. The PARP fam- Recently Zanjirband et  al. (2016) analyzed
ily has 17 structurally similar proteins. Cancer combination therapy of Nutlin-3 with cisplatin.
research has been mainly focusing on the PARP1. They demonstrated in ovarian cancer cell lines
PARP1s like SSBs detect DNA strand interruptions that Nutlin-3 or RG7388 effect in combination
and promote the NAD+-dependent synthesis of with cisplatin was additive to or synergistic in
poly(ADP-ribose) (PAR). Poly(ADP-ribosylation) a p53-dependent manner, resulting in increased
of histones cause their release from DNA and p53 activation, cell cycle arrest, and apoptosis,
allow access of chromatin more repair compo- associated with increased p21WAF1 protein and/
nents (Cerrato et al., 2016). PARP inhibition causes or caspase-3/7 activity compared to cisplatin
single-stranded breaks in the DNA that eventually alone.
collapse and form double-stranded breaks during
replication. Double-strand break repairs are car- Folate and folate receptor antagonists
ried out by homologous recombination mediated Folate is an essential component required by
by BRCA or by error susceptible nonhomologous rapidly dividing cells since it is important for
end joining, causing genetic instability and strand DNA synthesis and helps promote cell division.
breakage. PARP inhibition is based upon this very Targeting folate and folate receptors can prove to
theory of causing genetic instability and shall be potentially useful in tumor growth regression.
be very effective in ovarian cancer patients with The folate dependent drugs can be categorized as
BRCA mutations. This is because in individuals
with BRCA deficiencies, double-stranded breaks ●● Folate antagonists
shall be repaired by nonhomologous end joining, ●● Folate receptor inhibitors
which usually results in cell death. Some of the ●● Folate-conjugated therapeutic agents
significant PARP inhibitors utilized with positive
results and involved in further trials are summa- Folate antagonists
rized in Table 8.7. The results so far received in dif- Many folate antagonists have been studied to
ferent clinical trials suggest them to be potentially date. The most prominent is aminopterin and its
useful in treating patients with BRCA1 and BRCA2 successor molecule methotrexate. Both of these
mutations. competitively inhibit the enzyme folate reductase.
Besides folate reductase, another enzymatic tar-
MDM–TP53 inhibitor: Nutlin-3 alpha get is thymidylate synthase, which is regulated by
The TP53 tumor suppressor gene is the most fre- suicidal inhibition of fluorouracil-5, but has had
quently altered gene in human cancers including limited success partly due to toxicity and patient
most high- and low-grade ovarian cancers and tolerance issues. Recently drugs like pemetrexed
MDM2 is the main negative regulator of p53, regu- and raltitrexed inhibiting thymidylate synthase
lating p53 through ubiquitin-dependent degrada- have been used in ovarian cancer combination
tion (Zanjirband et al., 2016). Another important therapy. Pemetrexed in combination with beva-
strategy to treat low-grade ovarian cancer could cizumab showed median patient PFS improved
be to exploit the persistence of wild-type TP53 to 7.9 months and median OS increased to 25.7
in most low-grade cancers. Meijer et  al. (2013) months (Sapiezynski et al., 2016).
reported that in wild-type p53-expressing ovar-
ian, colon, and lung cancer cell lines and in an Folate receptor inhibitors
ex vivo model of human ovarian cancer, Nutlin-3 The reports of overexpression of folate receptors
enhanced p53, p21, MDM2, and DR5 surface (FR-α) in epithelial ovarian cancer (80% high)
expression. They further concluded that Nutlin-3 and its correlation with staging of ovarian cancer,
 Current scenario in ovarian cancer therapy  163

Table 8.7  PARPP inhibitors and their trials

PARP Trial
inhibitor Target Therapy type completed Results Ongoing trials
Olaparib BRCA-deficient Monotherapy Phase II Patients with Phase III trials
(Sapiezynski ovarian tumors and BRCA mutation SOLO
et al., 2016) combination had longer PFS
Rucaparib BRCA-mutated cells, Monotherapy Ongoing Results awaited Phase II trials
(Sapiezynski cells with low-level in 2017 ARIEL2 and 3
et al., 2016) expression of
HRR-associated
genes, platinum-
sensitive, relapsed
high-grade
epithelial ovarian
cancer, fallopian
tube cancer,
peritoneal cancer
Veliparib Patients with Monotherapy Phase II Effective as Phase II trials
(Sapiezynski mutation in the and trial monotherapy ongoing for
et al., 2016) BRCA 1 and/or 2 combination and has efficacy
gene acceptable
side effects
Niraparib BRCA mutation Monotherapy Phase I Phase I trial Phase II and III
(Sehouli carriers and trial showed trials (NOVA)
et al., 2016) patients with 300 mg as for efficacy and
sporadic cancer maximum tolerability in
tolerable limit monotherapy
for platinum-
sensitive
recurrent
ovarian
carcinoma
Talazoparib 1. In vitro studies Monotherapy Phase I Tumor response Phase II trials for
(Sehouli HRR-deficient (RECIST and/or efficacy in
et al., 2016) cells (BRCA1, CA-125) was ovarian cancer
BRCA2, PTEN shown at BMN ongoing
gene defects) 673 doses of
2. BRCA-associated 100–1100 µg
ovarian and daily in 11 out
peritoneal of 17 patients
carcinoma

makes FR-α as potential drug targets. Clinical tri- The results of a phase II trial showed that the
als have been conducted using farletuzumab as combinational therapy improved the overall tumor
response rate and farletuzumab was well tolerated
●● Combination with cisplatin and taxanes by patients. However, phase III trial results failed
(phase II) to reach the primary PFS endpoint and raised con-
●● Combination with carboplatin and taxane in cerns about its clinical efficacy in patients with
platinum-sensitive ovarian cancer (phase III) recurrent ovarian cancer (Sapiezynski et al., 2016).
164  Precision medicine in ovarian carcinoma

Module
Module 1 Module 2 Module 3 4

Figure 8.1  Structure of a typical folate-conjugated drug.

Folate-conjugated therapeutic agents HER-mediated pathways. It has been tolerated well

A typical folate drug conjugate contains four mod- in two platinum-resistant ovarian cancer cohorts
ules (Figure 8.1), namely, (61 and 62 patients) in a phase II trial. It was well
tolerated with a RR of 4.3% in heavily pretreated
Module 1—Pteroic acid OC patients (Gordon et al., 2006). A phase III trial
Module 2—A glutamic acid residue juxtaposed to is under way in combination with gemcitabine,
module 1 paclitaxel, or gemcitabine, and the other arm a pla-
Module 3—Has cleavable bonds cebo in combination with gemcitabine, paclitaxel,
Module 4—Has drug moiety or gemcitabine (Sapiezynski et al., 2016).

Vintafolide, a folate molecule conjugated with Estrogen receptor antagonists

desacetylvinlastine hydrazine, has entered phase II Estrogen receptors are a potential therapeutic tar-
and phase III trials. Vintafolide can deliver che- get in ovarian cancer owing to their expression in
motherapy to FR expressing cells, as it is a con- almost 60% of these patients. Besides, hormone
jugate of folate and the chemotherapeutic agent replacement therapy has lesser side effects com-
desacetylvinblastine monohydrazide. It has been pared to the cytotoxic therapy, making it an attrac-
studied in combination therapy of patients who have tive treatment modality to explore. Both estrogen
undergone less than three chemotherapeutic cycles receptor α (ERα) and ERβ get expressed in ovar-
in the phase II PRECEDENT study. It is given in ian cancer cells. In clinical trials with tamoxifen,
combination with PEGylated liposomal doxorubi- the overall response was only 13% but with fulves-
cin (PLD) for recurrent platinum-resistant ovarian trant it was 38%. Tamoxifen, however, was used
tumors. The combination therapy was clearly effec- in a limited fashion, that is, it was given only with
tive with median PFS 5 months versus 2.7 months chemotherapy for only 36 weeks (shorter duration
in monotherapy (Sapiezynski et al., 2016). than that recommended for breast cancer use in
this setting) and most patients had residual disease.
Human epidermal growth factor Further phase III trial with tamoxifen was also con-
receptor (HER) antagonists ducted versus thalidomide. The thalidomide group
Most of the high-grade ovarian cancer tumors show had a worse median and overall survival with a
defects with the signaling pathways, 67% showed higher risk of death (HR = 1.76, 95% CI = 1.16–
RB1 pathway defects, 45% had defects in PI3K and/ 2.68) and had higher toxicity. But the study did not
or Ras signaling, and 22% had defects with notch have a control arm (Simpkins et al., 2013). Recently
signaling (Bast and Mills 2010). Similarly ERBB2 Chan et al. (2014) reported that targeting ER sub-
is a proto-oncogene coding for human epidermal types may enhance the response to hormonal treat-
growth factor receptor (HER). It is associated with ment in women with ovarian cancer. They used
high expression in advanced epithelial ovarian highly modulated 1, 3-bis (4-hydroxyphenyl)-4-
cancer in some rare cases. Currently there are two methyl-5-[4-(2 -piperidinylethoxy)phenol]-1H-pyr-
HER antagonists being used in cancer therapeu- azole dihydrochloride (MPP) (ERα antagonist) or
tics: trastuzumab (Herceptin®) and pertuzumab 2,3-bis(4-hydroxy-phenyl)-propionitrile (DPN)
humanized monoclonal antibodies. Pertuzumab (ERβ agonist) significantly suppressed cell growth
prevents HER2 dimerization and inhibits multiple in ovarian cancer cell lines SKOV3 and OV2008.
 Future prospects  165

Table 8.8  Aromatase inhibitors and their clinical trials

Aromatase Therapy Trial

inhibitor Target type completed Results
Anastrozole Patients with ovarian, Monotherapy Phase II Well-tolerated oral
(Sapiezynski peritoneal, or fallopian agent but with
et al., 2016; tube carcinoma minimal tumoricidal
del Carmen activity in women with
et al., 2003) recurrent/persistent
Müllerian cancer
Letrozole Previously treated Monotherapy Phase II, 26% patients had a PFS
(Sapiezynski ER-positive ovarian Phase III trial of >6 months, and 5%
et al., 2016; cancer patients with a in CA breast patients had a PFS of
Smyth rising CA125 not in OC ≥2 years; minimal
et al., 2007) toxicity from letrozole
in this patient group
Exemestane Refractory ovarian Monotherapy Phase II Stable disease 36% and
(Sapiezynski cancer patients with well tolerated with
et al., 2016) prior chemotherapy of low toxicity
platinum and taxane

To date there has been low success with ER been discussed by (Sapiezynski et al., 2016). They
antagonists in ovarian cancer, however, it still used nanotechnology-based targeted delivery sys-
remains a lucrative target of research. tems (Figure 8.2). This is true personalized treat-
ment, where tumor proteins of a particular patient
Aromatase inhibitors are identified and targeted using nanotechnology-
The enzyme aromatase is an estrogen synthase and based targeted delivery systems. The systems con-
can be a potential target in ER-positive tumors. Some tain only one protein inhibitor (siRNA) to suppress
of the significant aromatase inhibitors used in ovar- one targeted protein or an anticancer drug. These
ian cancer are summarized in Table 8.8. Aromatase nanotechnology-based targeted delivery systems
inhibitors remain an active area of research in ovar- can be used in any combination with each other.
ian cancer, owing to fewer large prospective studies. The authors report encouraging in vitro and in
Although the results of the existing clinical trials vivo results at the preliminary stage, since their
are not clinically as much valuable, the use of these data showed that such a personalized treatment
drugs in combination therapy shall be an interest- approach is much more effective when compared
ing preposition to watch out for. with a standard treatment protocol. In the mice
model they selected five genes that were overex-
FUTURE PROSPECTS pressing (BCL2, MDR1, CD44, MMP9, PGR) and
caused metastases. The nanotechnology-based
The future of precision medicine in ovarian can- targeted delivery systems were designed accord-
cer treatment looks exciting. With the advances ingly. The systems (siRNAs) were delivered along
in the field of genomics and integration of basic with paclitaxel using a dendrimer-based deliv-
science and clinical databases, novel therapeutics ery system. The combination therapy reduced
can be developed targeted to specific oncopro- the expression of the target genes. This is because
teins responsible for multidrug resistance, tumor of significantly enhanced cell death induction,
progression and antiapoptotic cellular defense in imposed tumor shrinkage, and preventing the
ovarian cancer cells, or getting overexpressed in development of intraperitoneal metastases.
tumor cells compared to surrounding tissue. One Another nanotechnology-based treatment
such upcoming field is the nanotechnology-based regimen for ovarian cancer has been designed in
delivery of therapeutics. One such method has mice models for delivery of folate conjugates. It
166  Precision medicine in ovarian carcinoma

Tumour
tissue

Total
RNA
extracted

NTDs with
Rt PCR with
Overexpressed specific siRNA
selected
protein selected designed and
primers
used

Figure 8.2  Protocol for personalized treatment using nanotechnology-based targeted delivery
systems.

has been reported that nanoceria (NCe), nanopar- Ambros V. MicroRNAs and developmental
ticles of cerium oxide when conjugated to folic acid ­timing. Curr Opin Genet Dev 21;2011:511–7.
(NCe-FA), increased the cellular NCe internaliza- Banerjee S, and Kaye SB. New strategies in the
tion and inhibited cell proliferation. A combina- treatment of ovarian cancer: Current clinical
tion therapy of NCe-FA and cisplatinum lowered perspectives and future potential. Clin Cancer
the tumor burden significantly, and NCe-FA sig- Res 19(5);2013:961–68.
nificantly reduced proliferation and angiogenesis Bast RC. Molecular approaches to personalizing
in the xenograft mouse model (Hijaz et al., 2016). management of ovarian cancer. Ann Oncol
In times to come, the exciting field of precision 22(Suppl 8);2011:viii5–viii15.
medicine shall become a useful arena not just for Bast RC, and Mills GB. Personalizing therapy for
research but the advances shall translate into clini- ovarian cancer: BRCAness and beyond. J Clin
cal practice benefitting hundreds of ovarian cancer Oncol 28(22);2010:3545–48.
patients with early diagnosis and effective treatment. Bell D, Berchuck A, Birrer M et al. Integrated
genomic analyses of ovarian carcinoma.
REFERENCES Nature 474(7353);2011:609–15.
Bermejo MA, Alonso L, Rey Iglesias ME, Pérez
Agostini A, Brunetti M, Davidson B et al. López A, Montes F, and García Mata J.
Genomic imbalances are involved in miR- Antiangiogenic therapy in epithelial ovarian
30c and let-7a deregulation in ovarian cancer. In: Gynecologic Cancers: Basic Sciences,
tumors: Implications for HMGA2 expression. Clinical and Therapeutic Perspectives. Farghaly
Oncotarget 8(13);2017:21554–60. SA (Ed). InTech. 2016. doi:10.5772/60822.
Alshenawy HA, and Radi DA. Napsin-A, a possi- Braslavsky I, Hebert B, Kartalov E, and Quake SR.
ble diagnostic marker for differentiating clear Sequence information can be obtained from
cell ovarian carcinoma from other high-grade single DNA molecules. Proc Natl Acad Sci
ovarian carcinomas. Appl Immunohistochem USA 100;2003:3960–4.
Mol Morphol 2017 March. doi:10.1097/ Bubancova I, Kovarikova H, Laco J, Ruszova E,
PAI.0000000000000510. Dvorak O, Palicka V, and Chmelarova M.
 References 167

Next-generation sequencing approach in Chong GO, Jeon H-S, Han HS, Son JW, Lee
methylation analysis of HNF1B and GATA4 YH, Hong DG, Park HJ, Lee YS, and Cho YL.
Genes: Searching for biomarkers in ovarian Overexpression of microRNA-196b acceler-
cancer. Int J Mol Sci 18(2);2017:pii. ates invasiveness of cancer cells in recurrent
Cacan E. Epigenetic regulation of RGS2 epithelial ovarian cancer through regulation of
(Regulator of G-protein signaling 2) in chemo- homeobox A9. Cancer Genomics Proteomics
resistant ovarian cancer cells. J Chemother 14(2);2017:137–41.
29(3);2017:173–8. Church GM. The personal genome project. Mol
Calin GA, Ferracin M, Cimmino A et al. A Syst Biol 2005; 1:2005.0030.
microRNA signature associated with Conteduca V, Kopf B, Burgio SL, Bianchi E,
prognosis and progression in chronic Amadori D, and De Giorgi U. The emerging
lymphocytic leukemia. N Engl J Med role of anti-angiogenic therapy in ovarian can-
353(17);2005:1793–801. cer (review). Int J Oncol 44(5);2014:1417–24.
Calin GA, Garzon R, Cimmino A, Fabbri M, Couch FJ, Wang X, McGuffog L et al. Genome-
and Croce CM. MicroRNAs and leukemias: wide association study in BRCA1 mutation
How strong is the connection? Leuk Res carriers identifies novel loci associated with
30(6);2006:653–55. breast and ovarian cancer risk. PLoS Genet
Campbell IG, Russell SE, and Phillips WA. PIK3CA 9(3);2013:e1003212.
mutations in ovarian cancer. Clin Cancer Res Crane EK, Kwan S-Y, Izaguirre DI, Tsang YTM,
11(19 Pt 1);2005:7042; author reply 7042–3. Mullany LK, Zu Z, Richards JS, Gershenson
Cancer Genome Atlas Research Network. DM, and Wong K-K. Nutlin-3a: A poten-
Integrated genomic analyses of ovarian carci- tial therapeutic opportunity for TP53
noma. Nature 474(7353);2011:609–15. wild-type ovarian carcinomas. PLOS ONE
Carrarelli P, Funghi L, Ciarmela P, Centini G, 10(8);2015:e0135101.
Reis FM, Cruz CD, Mattei A, Vannuccini S, del Carmen MG, Fuller AF, Matulonis U, Horick
and Petraglia F. Deep infiltrating endome- NK, Goodman A, Duska LR., Penson R,
triosis and endometrial adenocarcinoma Campos S, Roche M, and Seiden MV. Phase
express high levels of myostatin and its II trial of anastrozole in women with asymp-
receptors messenger RNAs. Reprod Sci tomatic müllerian cancer. Gynecol Oncol
24(12);2017:1577–82. 91(3);2003:596–602.
Cerrato A, Morra F, and Celetti A. Use of poly Dong A, Lu Y, and Lu B. Genomic/epigenomic
ADP-ribose polymerase [PARP] inhibitors in alterations in ovarian carcinoma: Translational
cancer cells bearing DDR defects: The ratio- insight into clinical practice. J Cancer
nale for their inclusion in the clinic. J Exp Clin 7(11);2016:1441–51.
Cancer Res 35(1);2016:179. Dwivedi S, Goel A, Khattri S, Mandhani A,
Cesari R, Martin ES, Calin GA et al. Parkin, a gene Sharma P, Misra S, and Pant KK. Genetic vari-
implicated in autosomal recessive juvenile ability at promoters of IL-18 (pro-) and IL-10
parkinsonism, is a candidate tumor suppres- (anti-) inflammatory gene affects suscepti-
sor gene on chromosome 6q25-q27. Proc Natl bility and their circulating serum levels: An
Acad Sci USA 100(10);2003:5956–61. explorative study of prostate cancer patients
Chan K, Wong MS, Chan TK, and Chan V. A in North Indian populations. Cytokine
thalassaemia array for Southeast Asia. Br J 74(1);2015:117–22.
Haematol 124(2);2004:232–9. Dwivedi S, Goel A, Mandhani A, Khattri S,
Chan KK-L, Leung TH-Y, Chan DW, Wei N, Sharma P, Misra S, and Pant KK. Functional
Lau GT-Y, Liu SS, Siu MK-Y, and Ngan genetic variability at promoters of pro-(IL-18)
HY-S. Targeting estrogen receptor sub- and anti-(IL-10) inflammatory affects their
types (ERα and ERβ) with selective ER mRNA expression and survival in prostate
modulators in ovarian cancer. J Endocrinol carcinoma patients: Five year follow-up study.
221(2);2014:325–36. Prostate 75(15);2015:1737–46.
Cho KR, and Shih IM. Ovarian cancer. Annu Rev Dwivedi S, Shukla KK, Gupta G, and Sharma
Pathol 4;2009:287–313. P. Non-invasive biomarker in prostate
168  Precision medicine in ovarian carcinoma

carcinoma: A novel approach. Indian J Clin Hijaz M, Das S, Mert I et al. Folic acid
Biochem 28(2);2013:107–9. tagged nanoceria as a novel therapeu-
Dwivedi S, Singh S, Goel A, Khattri S, tic agent in ovarian cancer. BMC Cancer
Mandhani A, Sharma P, Misra S, and Pant 16(March);2016:220.
KK. Pro-(IL-18) and anti-(IL-10) inflamma- International Human Genome Sequencing
tory p­ romoter genetic variants (intrinsic Consortium. Initial sequencing and analysis of
factors) with tobacco exposure (extrinsic the human genome. Nature 409;2001:860–921.
factors) may influence susceptibility and sever- Jang K, Kim M, Gilbert CA, Simpkins F, Ince
ity of prostate carcinoma: A prospective study. TA, and Slingerland JM. VEGFA activates
Asian Pac J Cancer Prev 16(8);2015:3173–81. an epigenetic pathway upregulating ovar-
Elzek MA, and Rodland KD. Proteomics of ian cancer-initiating cells. EMBO Mol Med
ovarian cancer: Functional insights and 9(3);2017:304–18.
clinical applications. Cancer Metastasis Rev Kanchi KL, Johnson KJ, Lu C et al. Integrated
34(1);2015:83–96. analysis of germline and somatic variants in
Erlich H, Valdes AM, Noble J et al. HLA DR-DQ ovarian cancer. Nat Commun 5;2014:3156.
haplotypes and genotypes and type 1 Karpathiou G, Venet M, Mobarki M, Forest
diabetes risk: Analysis of the type 1 diabe- F, Chauleur C, and Peoc’h M. FOXA1 is
tes genetics consortium families. Diabetes expressed in ovarian mucinous neoplasms.
57(4);2008:1084–92. doi: 10.2337/db07-1331. Pathol 49(3);2017:271–76.
Fan J-B, Gunderson KL, Bibikova M et al. Illumina Katchman BA, Chowell D, Wallstrom G, Vitonis
universal bead arrays. Methods Enzymol AF, LaBaer J, Cramer DW, and Anderson
410;2006:57–73. KS. Autoantibody biomarkers for the detec-
Ferlay J, Soerjomataram I, Dikshit R, Eser S, tion of serous ovarian cancer. Gynecol Oncol
Mathers C, Rebelo M, Parkin DM, Forman 146(1);2017:129–36.
D, and Bray F. Cancer incidence and mortal- Kaur M, Singh A, Singh K, Gupta S, and Sachan
ity worldwide: Sources, methods and major M. Development of a multiplex MethyLight
patterns in GLOBOCAN 2012. Int J Cancer assay for the detection of DAPK1 and SOX1
136(5);2015:E359–86. methylation in epithelial ovarian cancer in a
Friedlander M, Hancock KC, Rischin D, Messing North Indian population. Genes Genetic Syst
MJ, Stringer CA, Matthys GM, Ma B, Hodge 91(3);2016:175–81.
JP, and Lager JJ. A phase II, open-label Kuo K-T, Mao T-L, Jones S et al. Frequent activat-
study evaluating pazopanib in patients with ing mutations of PIK3CA in ovarian clear cell
recurrent ovarian cancer. Gynecol Oncol carcinoma. Am J Pathol 174(5);2009:1597–601.
119(1);2010:32–7. Kurman RJ, and Shih IM. Molecular pathogenesis
Gemignani F, Perra C, Landi S, Canzian F, Kurg A, and extraovarian origin of epithelial ovarian
Tonisson N, Galanello R, Cao A, Metspalu A, cancer—Shifting the paradigm. Hum Pathol
and Romeo G. Reliable detection of beta- 42(7);2011:918–31.
thalassemia and G6PD mutations by a DNA Landen CN, Birrer MJ, and Sood AK. Early events
microarray. Clin. Chem. 48;2002:2051–4. in the pathogenesis of epithelial ovarian can-
Gordon MS, Matei D, Aghajanian C et al. Clinical cer. J Clin Oncol 26(6);2008:995–1005.
activity of pertuzumab (rhuMAb 2C4), a HER Lee J, An S, Choi YM et al. Musashi-2 is a novel
dimerization inhibitor, in advanced ovarian regulator of paclitaxel sensitivity in ovarian
cancer: Potential predictive relationship with cancer cells. Int J Oncol 49(5);2016:1945–52.
tumor HER2 activation status. J Clin Oncol Lehrbach NJ, and Miska EA. Functional genomic,
24(26);2006:4324–32. computational and proteomic analysis of C.
Gozzi G, Chelbi ST, Manni P, Alberti L, Fonda S, Elegans microRNAs. Brief Funct Genomics
Saponaro S, Fabbiani L, Rivasi F, Benhattar Proteomic 7(3);2008:228–35.
J, and Losi L. Promoter methylation and Lehtinen L, Vesterkvist P, Roering P et al. REG4 is
downregulated expression of the TBX15 highly expressed in mucinous ovarian cancer:
gene in ovarian carcinoma. Oncol Lett A potential novel serum biomarker. PLOS
12(4);2016:2811–19. ONE 11(3);2016:e0151590.
 References 169

Levine DA, Bogomolniy F, Yee CJ, Lash A, Nakayama K, Nakayama N, Kurman RJ, Cope L,
Barakat RR, Borgen PI, and Boyd J. Frequent Pohl G, and Samuels Y. Sequence mutations
mutation of the PIK3CA gene in ovar- and amplification of PIK3CA and AKT2 genes
ian and breast cancers. Clin Cancer Res in purified ovarian serous neoplasms. Cancer
11(8);2005:2875–8. Biol Ther 5(7);2006:779–85.
Li J, Li Q, Huang H, Li Y, Li L, Hou W, and You Z. Nguyen HT, Tian G, and Murph MM. Molecular
Overexpression of miRNA-221 promotes cell epigenetics in the management of ovarian
proliferation by targeting the apoptotic pro- cancer: Are we investigating a rational clinical
tease activating factor-1 and indicates a poor promise? Front Oncol 4;2014:71.
prognosis in ovarian cancer. Int J Oncol 2017 Nick AM, Coleman RL, Ramirez PT, and Sood
March. doi:10.3892/ijo.2017.3898. AK. A framework for a personalized surgi-
Liede A, Tonin PN, Sun CC, Serruya C, Daly MB, cal approach to ovarian cancer. Nat Rev Clin
Narod SA, and Foulkes WD. Is hereditary Oncol 12(4);2015:239–45.
site-specific ovarian cancer a distinct genetic Obata K, Morland SJ, Watson RH, Hitchcock A,
condition? Am J Med Genet 75(1);1998:55–8. Chenevix-Trench G, Thomas EJ, and Campbell
Liu H, Shi H, Fan Q, and Sun X. Cyclin Y regulates IG. Frequent PTEN/MMAC mutations in endo-
the proliferation, migration, and invasion of metrioid but not serous or mucinous epithelial
ovarian cancer cells via Wnt signaling path- ovarian tumors. Cancer Res 58(10);1998:2095–7.
way. Tumour Biol 37(8);2016:10161–75. Oliphant A, Barker DL, Stuelpnagel JR, and
Liu W, Lv C, Zhang B, Zhou Q, and Cao Z. Chee MS. BeadArray technology: Enabling
MicroRNA-27b functions as a new inhibitor an accurate, cost-effective approach to
of ovarian cancer-mediated vasculogenic high-throughput genotyping. Biotechniques
mimicry through suppression of VE-cadherin 32;2002:S56.
expression. RNA 23(7);2017:1019–27. Rahimi F, Karimi J, Goodarzi MT, Saidijam M,
Mardis ER. Next-generation DNA sequencing Khodadadi I, Razavi ANE, and Nankali M.
methods. Annu Rev Genomics Hum Genet. Overexpression of receptor for advanced gly-
9;2008:387–402. doi: 10.1146/annurev. cation end products (RAGE) in ovarian cancer.
genom.9.081307.164359. Cancer Biomark 18(1);2017:61–8.
Margulies M, Egholm M, Altman WE et al. Romeo M, Karachaliou N, Chaid I, Queralt C, De
Genome sequencing in microfabricated Aguirre I, Del Carmen Gómez M, Sanchez-
high-density picolitre reactors. Nature. Ronco M, Radua J, Ramírez JL, and Rosell
437(7057);2005:376–80. R. Low RAP80 mRNA expression correlates
Martín-Cameán M, Delgado-Sánchez E, Piñera with shorter survival in sporadic high-grade
A, Diestro MD, De Santiago J, and Zapardiel serous ovarian carcinoma. Int J Biol Markers
I. The role of surgery in advanced epithe- 32(1);2017:e90–95.
lial ovarian cancer. Ecancermedicalscience Ross-Adams H, Ball S, Lawrenson K, Halim S,
10;2016;666. Russell R, and Wells C. HNF1B variants associ-
McConechy MK, Ding J, Senz J et al. Ovarian and ate with promoter methylation and regulate
endometrial endometrioid carcinomas have gene networks activated in prostate and ovar-
distinct CTNNB1 and PTEN mutation profiles. ian cancer. Oncotarget 7(46);2016:74734–46.
Mod Pathol 27(1);2014:128–34. Roukos DH. Personalized cancer diagnostics
Meijer A, Kruyt FAE, van der Zee AGJ, Hollema H, and therapeutics. Expert Rev Mol Diagn.
Le P, ten Hoor KA, Groothuis GMM, Quax WJ, 9(3);2009:227–9.
de Vries EGE, and de Jong S. Nutlin-3 prefer- Saiki RK, Walsh PS, Levenson CH, and Erlich
entially sensitises wild-type p53-expressing HA. Genetic analysis of amplified DNA
cancer cells to DR5-selective TRAIL over with immobilized sequence-specific oligo-
rhTRAIL. Br J Cancer 109(10);2013:2685–95. nucleotide probes. Proc Natl Acad Sci USA
Miller MB, and Tang YW. Basic concepts of 86(16);1989:6230–4.
microarrays and potential applications in Samuels Y, and Waldman T. Oncogenic muta-
clinical microbiology. Clinical Microbiology tions of PIK3CA in human cancers. Curr Top
Reviews 22(4);2009:611–33. Microbiol Immunol 347;2010:21–41.
170  Precision medicine in ovarian carcinoma

Sangha N, Wu R, Kuick R et al. Neurofibromin 1 Smyth JF, Gourley C, Walker G et al.

(NF1) defects are common in human ovar- Antiestrogen therapy is active in selected
ian serous carcinomas and co-occur with ovarian cancer cases: The use of letrozole
TP53 mutations. Neoplasia (New York, N.Y.) in estrogen receptor-positive patients. Clin
10(12);2008:1362–72, following 1372. Cancer Res 13(12);2007:3617–22.
Sankaranarayanan R, and Ferlay J. Worldwide Stratton JF, Thompson D, Bobrow L et al.
burden of gynaecological cancer: The size The genetic epidemiology of early-
of the problem. Best Pract Res Clin Obstet onset ­epithelial ovarian cancer: A pop-
Gynaecol 20(2);2006:207–25. ulation-based study. Am J Hum Genet
Sapiezynski J, Taratula O, Rodriguez-Rodriguez 65(6);1999:1725–32.
L, and Minko T. Precision targeted ther- Sung HY, Yang SD, Park AK, Ju W, and Ahn JH.
apy of ovarian cancer. J Control Release Aberrant hypomethylation of solute car-
243(December);2016;250–68. rier family 6 member 12 promoter induces
Sehouli J, Braicu EI, and Chekerov R. PARP metastasis of ovarian cancer. Yonsei Med J
inhibitors for recurrent ovarian carcinoma: 58(1);2017:27–34.
Current treatment options and future per- Tang S, Yang F, Du X, Lu Y, Zhang L, and Zhou X.
spectives. Geburtshilfe Und Frauenheilkunde Aberrant expression of anaplastic lymphoma
76(2);2016:164–69. kinase in ovarian carcinoma independent of
Sellar GC, Watt KP, Rabiasz GJ et al. OPCML gene rearrangement. Int J Gynecol Pathol
at 11q25 is epigenetically inactivated 35(4);2016:337–47.
and has tumor-suppressor function in Tangjitgamol S, Manusirivithaya S,
epithelial o
­ varian cancer. Nat Genet Srijaipracharoen S et al. Endometrial cancer
34(3);2003:337–43. in Thai women: Clinicopathological presen-
Sereni MI, Baldelli E, Gambara G et al. Functional tation and survival. Asian Pac J Cancer Prev
characterization of epithelial ovarian cancer 11;2010:1267–72.
histotypes by drug target based protein Tawfik DS, Griffiths AD. Man-made cell-like
signaling activation mapping: Implications compartments for molecular evolution. Nat
for personalized cancer therapy. Proteomics Biotechnol 16;1998:652–6.
15(2–3);2015:365–73. Teo MCC. Update on the management and the
Shao H, Mohamed EM, Xu GG, Waters M, Jing K, role of intraperitoneal chemotherapy for
and Ma Y. Carnitine palmitoyltransferase 1A ovarian cancer. Curr Opin Obstet Gynecol
functions to repress FoxO transcription f­ actors 26(1);2014:3–8.
to allow cell cycle progression in ­ovarian can- Tian J, Xu Y-Y, Li L, and Hao Q. MiR-490-3p
cer. Oncotarget 7(4);2016:3832–46. sensitizes ovarian cancer cells to cisplatin by
Shin K, Kim KH, Yoon MS, Suh DS, Lee JY, Kim A, directly targeting ABCC2. Am J Transl Res
and Eo W. Expression of interactive genes 9(3);2017:1127–38.
associated with apoptosis and their prognos- Tian Y, Yao Z, Roden RBS, and Zhang H.
tic value for ovarian serous adenocarcinoma. Identification of glycoproteins associated
Adv Clin Exp Med 25(3);2016:513–21. with different histological subtypes of ovarian
Simpkins F, Garcia-Soto A, and Slingerland J. tumors using quantitative glycoproteomics.
New insights on the role of hormonal therapy Proteomics 11(24);2011:4677–87.
in ovarian cancer. Steroids 78(6);2013:530–37. Toyama A, Suzuki A, Shimada T et al.
Singer G, Oldt R, Cohen Y, Wang BG, Sidransky Proteomic characterization of ovarian
D, Kurman RJ, and Shih I-M. Mutations in cancers i­dentifying annexin-A4, phospho-
BRAF and KRAS characterize the develop- serine ­aminotransferase, cellular retinoic
ment of low-grade ovarian serous carcinoma. acid-binding protein 2, and serpin B5 as
J Natl Cancer Inst 95(6);2003;484–86. histology-specific biomarkers. Cancer Sci
Skirnisdottir I, Bjersand K, Akerud H, and 103(4);2012:747–55.
Seidal T. Napsin A as a marker of clear Venter JC, Adams MD, Myers EW et al. The
cell ovarian carcinoma. BMC Cancer sequence of the human genome. Science
13(November);2013:524. 291;2001:1304–51.
 References 171

Wei L, Yin F, Zhang W, and Li L. STROBE- Yang L, Zhang X, Ma Y, Zhao X, Li B, and Wang
compliant integrin through focal adhesion H. Ascites promotes cell migration through
involve in cancer stem cell and multidrug the repression of miR-125b in ovarian cancer.
resistance of ovarian cancer. Medicine Oncotarget 8(31);2017:51008–15.
96(12);2017:e6345. Yang M, Xie X, and Ding Y. SALL4 is a marker of
Wiegand KC, Shah SP, Al-Agha OM et al. poor prognosis in serous ovarian carcinoma
ARID1A mutations in endometriosis-asso- promoting invasion and metastasis. Oncol
ciated ovarian carcinomas. N Engl J Med Rep 35(3);2016:1796–806.
363;2010:1532–43. Zanjirband M, Edmondson RJ, and Lunec J.
Willner J, Wurz K, Allison KH, Galic V, Garcia RL, Pre-clinical efficacy and synergistic potential
Goff BA, and Swisher EM. 2007. Alternate of the MDM2-p53 antagonists, nutlin-3 and
molecular genetic pathways in ovarian carci- RG7388, as single agents and in combined
nomas of common histological types. Hum treatment with cisplatin in ovarian cancer.
Pathol 38(4);607–13. Oncotarget 7(26);2016:40115–34.
Wu R, Baker SJ, Hu TC, Norman KM, Fearon ER, Zannoni GF, Improta G, Pettinato A, Brunelli C,
and Cho KR. Type I to type II ovarian carci- Troncone G, Scambia G, and Fraggetta F.
noma progression: Mutant Trp53 or Pik3ca Molecular status of PI3KCA, KRAS and BRAF
confers a more aggressive tumor phenotype in ovarian clear cell carcinoma: An analysis of
in a mouse model of ovarian cancer. Am J 63 patients. J Clin Pathol 69(12);2016:1088–92.
Pathol 182(4);2013:1391–99. Zhang L, Cao X, Zhang L, Zhang X, Sheng H, and
Xiao L, Shi X-Y, Zhang Y, Zhu Y, Zhu L, Tian W, Tao K. UCA1 overexpression predicts clini-
Zhu B-K, and Wei Z-L. YAP induces cisplatin cal outcome of patients with ovarian cancer
resistance through activation of autophagy in receiving adjuvant chemotherapy. Cancer
human ovarian carcinoma cells. OncoTargets Chemother Pharmacol 77(3);2016:629–34.
and Therapy 9;2016:1105–14. Zhang T, Xu J, Deng S et al. Core signaling path-
Xiao M, Cai J, Cai L, Jia J, Xie L, Zhu Y, Huang B, ways in ovarian cancer stem cell revealed by
Jin D, and Wang Z. Let-7e sensitizes epithelial integrative analysis of multi-marker genomics
ovarian cancer to cisplatin through repressing data. PLoS ONE 13(5);2018:e0196351. https://
DNA double strand break repair. J Ovarian doi.org/10.1371/journal.pone.0196351
Res 10(1);2017:24. Zhou D, Zhang L, Sun W, Guan W, Lin Q,
Xu J, Jiang N, Shi H, Zhao S, Yao S, and Shen Ren W, Zhang J, and Xu G. Cytidine
H. miR-28-5p promotes the development ­monophosphate kinase is inhibited by
and progression of ovarian cancer through the TGF-β signalling pathway through the
inhibition of N4BP1. Int J Oncol 2017 March. ­upregulation of miR-130b-3p in human epithe-
doi:10.3892/ijo.2017.3915. lial ovarian cancer. Cell Signal 35;2017:197–207.
Xu Y, Li X, Wang H, Xie P, Yan X, Bai Y, and Zhang Zhu Y, Wu R, Sangha N, Yoo C, Cho KR,
T. Hypermethylation of CDH13, DKK3 and Shedden KA, Katabuchi H, and Lubman
FOXL2 promoters and the expression of EZH2 DM. Classifications of ovarian cancer tis-
in ovary granulosa cell tumors. Mol Med Rep sues by proteomic patterns. Proteomics
14(3);2016:2739–45. 6(21);2006:5846–56.
Yan H, Qiu L, Xie X, Yang H, Liu Y, Lin X, and Zou J, Liu L, Wang Q, Yin F, Yang Z, Zhang W,
Huang H. ELF5 in epithelial ovarian carcinoma and Li L. Downregulation of miR-429 contrib-
tissues and biological behavior in ovarian utes to the development of drug resistance in
carcinoma cells. Oncol Rep 37(3);2017:1412–8. epithelial ovarian cancer by targeting ZEB1.
doi: 10.3892/or.2017.5418. Am J Transl Res 9(3);2017:1357–68.
 9
Precision medicine in myelodysplastic
syndromes

OTA FUCHS

Introduction 173 Therapies for MDS 180

Methods for the diagnosis and treatment of New knowledge about molecular and
myelodysplastic syndromes (MDS) 174 cellular mechanisms of MDS as basis for
Cytogenetic analysis 174 new therapy 182
Importance of the morphological evaluation The concept of personalized or precision
of bone marrow slides and peripheral medicine 184
blood films for the diagnosis of MDS 174 Future opportunities 185
Histopathology 176 Acknowledgments 186
Classification of MDS patients 176 Conflict of Interest 186
Somatic mutations of individual genes as References 187
clinically relevant biomarkers in MDS 178

INTRODUCTION age decade above 70 years. Approximately 10,000

to 15,000 new cases are diagnosed annually in the
Myelodysplastic syndromes (MDS) are a family of United States of America. The overall incidence of
clonal hematopoietic stem cell malignancies char- MDS is higher in males than in females (1.5–2:1).
acterized by ineffective hematopoiesis, peripheral MDS is rare in children, where it is more fre-
cytopenias, frequent karyotypic abnormalities, quently associated with monosomy of chromo-
and risk of transformation to acute myeloid leuke- some  7. Adolescents with monosomy 7 frequently
mia (AML) (Nimer, 2008; Tefferi and Vardiman, have (72%) GATA2 mutations and the high risk for
2009; Sekeres, 2010; Adés et  al., 2014; Bejar and progression to advanced disease (Wlodarski et al.,
Steensma, 2014; Pellagatti and Boultwood, 2015; 2016). Therefore, timely hematopoietic stem cell
Kennedy and Ebert, 2017; Shastri et al., 2017). The transplantation must be considered in these cases.
incidence of MDS is approximately 3–5 cases per MDS in children and young and middle aged adults
100,000 population per year with 30 cases per is frequently associated with underlying genetic
100,000 population per year in patients after the age predisposition syndromes (bone marrow failure
of 70 (Phekoo et al., 2006; Ma et al., 2007; Germing such as Fanconi anemia and dyskeratosis congenita)
et al., 2008; Rollison et al., 2008; Cogle et al., 2011; and nonsyndromic familial MDS/AML predis-
Neukirchen et al., 2011; Rodger and Morison, 2012; position caused by mutations in GATA2, RUNX1,
Avgerinou et al., 2013; Otrock et al., 2015; Lubeck CEBPA, TERT/TERC (telomerase subunit genes),
et al., 2016). Thus, MDS are mainly a disease of the ANKRD26 (ankyrin repeat domain 26), ETV6 (gene
elderly with a sharp increase in incidence in the for ETS family transcriptional repressor, variant 6),

173
174  Precision medicine in myelodysplastic syndromes

and SRP72 (signal recognition particle 72) genes 2015). Conventional karyotyping is performed on
(Owen et al., 2008; Godley, 2014; West et al., 2014; metaphase cells (a stage in cell division when chro-
Babushok and Bessler, 2015; Churpek et  al., 2015; mosomes are condensed) with Giemsa stain result-
Babushok et al., 2016; Bannon and DiNardo, 2016; ing in a bending pattern of light and dark stripes,
Koeffler and Leong, 2017). known as G-banding. The patterns are specific,
The majority of MDS cases occur de novo (80%– allowing to identify each chromosome. Clonal chro-
90%), whereas 10%–20% of MDS cases are second- mosome abnormalities can be found in roughly 50%
ary, therapy-related MDS. The etiology of de novo of de novo MDS patients and in up to 90% of MDS
MDS is unclear. Cumulative exposure to environ- patients with therapy-related or secondary disease.
mental agents (benzene, ionizing radiation, tobacco The remaining MDS patients with normal karyotype
smoke pesticides, insecticides, and possibly hair can have molecular abnormalities (Bejar et al., 2011;
dyes or extremely low frequency magnetic fields) Haferlach, 2012; Itzykson et al., 2013; Nybakken and
and chronic inflammation, genetic differences in Bagg, 2014; Bejar, 2015; Kennedy and Ebert, 2017).
leukemogen susceptibility and metabolism, and The most common MDS abnormality is the
hematopoietic stem cell genomic senescence may interstitial deletion of chromosome 5/5q– or
contribute to disease pathogenesis in de novo cases del(5q)/, del(7q)/monosomy 7 and trisomy 8 (Table
(Bowen, 2013). Therapy-related MDS arise as a 9.1), but a number of other chromosome abnormal-
result of chemotherapy, radiation therapy, or a com- ities occur in MDS (Table 9.1). Karyotyping is criti-
bination of these therapies. Therapy-related MDS cal, not only for MDS patients but also for patients
have worse outcomes than de novo MDS (Zeidan with unexplained cytopenias in the absence of
et  al., 2017). Therapy-related MDS have more morphologically identifiable dysplasia in the dif-
complex karyotypes and mutations in p53. The ferential diagnosis. Trisomy 8, del(20q), and loss
revised version of the World Health Organization of chromosome Y without other alterations are not
(WHO) classification combines therapy-related presumptive evidence of MDS, because these cyto-
MDS (t-MDS) and therapy-related AML (t-AML) genetic abnormalities also occur in other myeloid
in the one entity of therapy-related myeloid neo- neoplasms. The presence of three or more cytoge-
plasms (t-MNs) (Heuser, 2016; Ganser and Heuser, netic abnormalities are associated with poor prog-
2017). Treatment options include best supportive nosis (Table 9.2).
care, azacitidine, or intensive chemotherapy/allo- In comparison to chromosome banding anal-
geneic transplant according to genetic risk profile ysis, fluorescence in situ hybridization (FISH)
of patients with t-MNs (Ganser and Heuser, 2017). has some advantages. FISH can be performed on
Median overall survival (OS) was 14 months and nondividing cells but is restricted according the
5-year OS was 13.8%. Improved outcomes were respective FISH probe used in analysis. Therefore,
seen with allogeneic transplantation. Clonal hemo- FISH is not suitable for an initial analysis where
poiesis of indeterminate potential (CHIP) is an age- we need comprehensive view on the karyotype. On
associated genetic event linked to increased risk the other hand, FISH is valuable for the verification
of primary hematological malignancies. Patients of a suspected anomaly like del(5q).
with cancer who have CHIP are at increased risk MDS is a dynamic disease with a high prob-
of developing therapy-related myeloid neoplasms ability for a karyotype evolution. Thus, the karyo-
(Gillis et al., 2017; Takahashi et al., 2017). type should be monitored repeatedly during the
course of the disease, mainly from peripheral
METHODS FOR THE DIAGNOSIS AND blood (immunomagnetically enriched circulating
TREATMENT OF MYELODYSPLASTIC CD34+ cells) due to ethical reasons.
SYNDROMES (MDS) Importance of the morphological
Cytogenetic analysis evaluation of bone marrow slides
and peripheral blood films for the
The most important diagnostic and prognostic diagnosis of MDS
method in patients with MDS is conventional karyo-
typing (Haase et  al., 2007; Haase, 2008; Tiu et al., The morphological evaluation of peripheral blood
2011; Giagounidis and Haase, 2013; Bacher et  al., films and bone marrow slides remains the essential
 Methods for the diagnosis and treatment of myelodysplastic syndromes (MDS)  175

Table 9.1  Frequency of chromosome abnormalities in MDS

% of the individual abnormality

With one As part of
Total (% of all additional complex
Abnormality MDS cases) Isolated abnormality abnormalities
5q− 15.1 47.0 17.0 36.0
−7/7q− 11.1 37.5 13.5 49.0
+8 8.4 46.8 21.4 31.8
−18/18q− 3.8 3.8 2.6 93.6
20q– 3.6 48.6 13.5 37.8
−5 3.3 1.4 5.8 92.8
−Y 2.8 70.7 8.6 20.7
+21 2.2 11.1 40.0 48.9
−17/17p− 2.0 2.4 2.4 95.2
inv/t(3q) 2.0 39.0 19.5 41.5
−13/13q− 1.9 12.5 15.0 72.5
+1/+1q 1.8 8.1 16.2 75.7
−21 1.6 9.1 12.1 78.8
+11 1.4 21.4 14.3 64.3
−12 1.3 0 7.7 92.3
12p− 1.2 28.0 24.0 48.0
t(5q) 1.2 25.0 12.5 62.5
11q− 1.1 34.8 17.4 47.8
9q− 1.1 34.8 13.0 52.2
t(7q) 1.1 27.3 27.3 45.0
−20 1.1 0 0 100.0
Source: Adapted from Haase D, Ann Hematol 87;2008:515–26.
Note: Total study cohort consisted of 2072 MDS patients.

Table 9.2  Cytogenetic abnormalities in de novo MDS with prognostic implications

Hazard
Median Progression
ratio
Prognostic Frequency survival to AML
subgroup Cytogenetic abnormality (%) (years) (years) OS AML
Very good −Y, del(11q) 4 5.4 NR 0.7 0.4
Good Normal, del(5q), del(12q), 72 4.8 9.4 1.0 1.0
del(20q); double abnormality
including del(5q)
Intermediate del(7q), +8, +19, i(17q); any 13 2.7 2.5 1.5 1.8
other single or double
independent clones
Poor −7, inv(3), t(3q) or del(3q); 4 1.5 1.7 2.3 2.3
double abnormality including
–7 or del(7q); complex (3
abnormalities)
Very poor Complex (>3 abnormalities) 7 0.7 0.7 3.8 3.6
Source: Adapted from Nybakken GE, and Bagg A. J Mol Diagn 16;2014:145–58.
Note: NR, not reached; OS, overall survival.
176  Precision medicine in myelodysplastic syndromes

method of MDS diagnosis, and it needs a consider- integrated marrow cytogenetic subset, marrow
able amount of expertise and training. In addition, blast percentage, and depth of cytopenias (hemo-
this method is cheap, rapid, and universally avail- globin, platelet and absolute neutrophil count)
able. A number of cytological abnormalities (see and was published in 2012 (Greenberg et  al.,
Table 9.3) need to be considered. 2012). Validation of WPSS for MDS and com-
parison with IPSS-R has been recently described
Histopathology (Della Porta et  al., 2015). Two other prognostic
systems for MDS subgroups (MD Anderson lower
Bone marrow trephine biopsies, examined by expe- risk MDS prognostic scoring system; chronic
rienced hematopathologists are important mainly myelomonocytic leukemia (CMML) prognos-
in cases of hypoplastic MDS or fibrotic MDS, where tic scoring system) exist (Jonas and Greenberg,
cytomorphology is of limited value for the diag- 2015; Zeidan et al., 2015). The 2016 revision to the
nosis and prognosis of MDS. Hematopathologists WHO classification of myeloid neoplasms (Arber
use immunohistochemistry for CD34+ cells and et al., 2016; Lichtman, 2016; Table 9.5) introduced
in del(5q) MDS as well as for TP53 overexpression refinements in morphological interpretation and
caused by TP53 mutation. cytopenia assessment and addressed the influ-
ence of recent developments in the molecular
Classification of MDS patients pathogenesis of MDS.
Mandatory for the diagnosis of MDS is the
The initial French-British-American (FAB) clas- presence of dysplasia in >10% of cells within one
sification system of MDS was published in 1982 or more cell lineages or presence of >15% ring sid-
(Bennett et al., 1982) and was later refined to the eroblasts or presence of MDS-associated cytoge-
International Prognostic Scoring System (IPSS) netic abnormalities. The WHO proposal has raised
(Greenberg et  al., 1997) and to the WHO prog- some concern regarding minimal diagnostic cri-
nostic scoring system (WPSS) (Malcovati et  al., teria particularly in patients with normal karyo-
2005, 2007; Vardiman et  al., 2009; Senent et  al., type without robust morphological markers of
2013; Table 9.4). The new revised IPSS (IPSS-R) dysplasia (such as sideroblasts or excess of blasts).

Table 9.3  Cytological abnormalities in MDS

Peripheral blood
Granulopoiesis: pseudo-Pelger cells, abnormal chromatin clumping, hypo-/degranulation, an increase
in the number of immature leukocytes (left shift)
Platelets: giant platelets, anisometry of platelets
Red cells: anisocytosis, poikilocytosis, dimorphic erythrocytes, polychromasia, hypochromasia,
megalocytes (large nonnucleated red blood cells), basophilic stippling, presence of nucleated
erythroid precursors, tear drop cells, ovalocytes (red blood cells with pale centers), fragmentocytes
(fragmented or split red cells)
Bone marrow
Cellularity of the marrow
Erythropoiesis: megaloblastoid changes, multinuclearity, nuclear budding, nuclear bridges, atypical
mitoses, sideroblastosis, ring sideroblasts, periodic acid-Schiff (PAS) positive red cell precursors
Megakaryocytes: micromegakaryocytes, mononuclear megakaryocytes, dumbbell-shaped nuclei,
hypersegmentation, multinuclearity with multiple isolated nuclei
Granulopoiesis: left shift, increased medullary blast count, Auer rods or Auer bodies, hypo-/
degranulation, pseudo-Pelger cells, nuclear anomalies (e.g., hypersegmentation, abnormal
chromatin clumping), deficiency of myeloperoxidase, increase and morphological abnormality of
monocytes
Source: Adapted from Giagounidis A, and Haase D. Best Pract Res Clin Hematol 26;2013:337–53.
 Methods for the diagnosis and treatment of myelodysplastic syndromes (MDS)  177

Table 9.4  World Health Organization MDS classification and criteria, 2008

Classification Blood findings Bone marrow findings

Refractory cytopenia with Unicytopenia or Unilineage dysplasia: >10% of cells in
unilineage dysplasia bicytopenia one myeloid lineage
(RCUD) No or rare blasts (<1%)
Refractory anemia (RA), Anemia <5% blasts; <15% of erythroid precursors
refractory neutropenia Neutropenia are ring sideroblasts
(RN), refractory Thrombocytopenia
thrombocytopenia (RT)
Refractory anemia with Anemia >15% of erythroid precursors are ring
ring sideroblasts (RARS) No blasts sideroblasts
Erythroid dysplasia only; <5% blasts
Refractory cytopenia with Cytopenia(s) Dysplasia in >10% of the cells in two or
multilineage dysplasia No or rare blasts (<1%) more myeloid lineages
(RCMD)
No Auer rods <5% blasts in marrow
<1 × 109/L monocytes No Auer rods
±15% ring sideroblasts
Refractory anemia with Cytopenia(s) Unilineage or multilineage dysplasia;
excess blasts-1 (RAEB-1) <5% blasts 5%–9% blasts
No Auer rods No Auer rods
<1 × 109/L monocytes
Refractory anemia with Cytopenia(s) Unilineage or multilineage dysplasia;
excess blasts-2 (RAEB-2) 5%–19% blasts 10%–19% blasts
Auer rods± Auer rods±
<1 × 109/L monocytes
Myelodysplastic Cytopenias Unequivocal dysplasia in <10% of cells in
syndrome–unclassified <1% blasts one or more myeloid cell lines when
(MDS-U) accompanied by a cytogenetic
abnormality considered as presumptive
evidence for a diagnosis of MDS
<5% blasts
MDS associated with Anemia Normal to increased megakaryocytes with
isolated del(5q) Usually normal or increased hypolobated nuclei
platelet count
No or rare blasts (<1%) <5% blasts
Isolated del(5q) cytogenetic abnormality
No Auer rods
Source: Adapted from Nybakken GE, and Bagg A. J Mol Diagn 16;2014:145–58.

Flow cytometry (FCM) immunophenotyping has of MDS, combinations of such parameters into
been proposed as a tool to improve the evaluation scoring systems have been shown to discriminate
of marrow dysplasia. Immunophenotyping is an MDS from other cytopenias with high sensitivity
accurate method for quantitative and qualitative and acceptable specificity. When morphology and
evaluation of hematopoietic cells. MDS have been cytogenetics are indeterminate, an abnormal phe-
found to have abnormal expression of several cel- notype determined by FCM can help to establish a
lular antigens. Although no single immunopheno- definitive diagnosis of MDS (Cremers et al., 2016;
typic parameter has been proven to be diagnostic Della Porta and Picone, 2017).
178  Precision medicine in myelodysplastic syndromes

Table 9.5  World Health Organization MDS classification and criteria, 2016 revision

Bone marrow
Classification Blood findings findings
MDS with single lineage dysplasia (MDS-SLD) Peripheral blood (PB) Bone marrow (BM)
blasts <1% blasts <5%
MDS with multilineage dysplasia (MDS-MLD) PB blasts <1% BM blasts <5%
MDS with ring sideroblasts and with single PB blasts <1% BM blasts <5%
lineage dysplasia (MDS-RS-SLD)
MDS with ring sideroblasts and with multilineage PB blasts <1% BM blasts <5%
dysplasia (MDS-RS-LD)
MDS with isolated del(5q) PB blasts <1% BM blasts <5%
MDS with excess blasts (MDS-EB-1) PB blasts 2%–4% BM blasts 5%–9%
MDS with excess blasts (MDS-EB-2) PB blasts 5%–19% BM blasts 10%–19%
MDS, unclassifiable (MDS-U) with 1% blood blasts PB blasts = 1% BM blasts <5%
MDS, unclassifiable (MDS-U) with single lineage PB blasts <1% BM blasts <5%
dysplasia and pancytopenia
MDS, unclassifiable (MDS-U) based on defining PB blasts <1% BM blasts <5%
cytogenetic abnormality
Refractory cytopenia of childhood PB blasts <2% BM blasts <5%
Source: Adapted from Arber DA et al., Blood 127;2016:2391–405.
Note: Cytopenia is defined as hemoglobin <10 g/dL, platelet count <100 × 109/L, and absolute neutrophil
count <1.8 × 109/L. Rarely, MDS may be present with mild anemia or thrombocytopenia above these
levels. PB monocytes must be <1 × 109/L. Isolated del(5q) is defined as del(5q) alone or del(5q) with 1
additional abnormality except −7 or del(7q). Cases with >15% ring sideroblasts by definition have signifi-
cant erythroid dysplasia and are classified as MDS-RS-LD.

Somatic mutations of individual clonal or subclonal, and leukemia-free survival

genes as clinically relevant deteriorated steadily as numbers of driver muta-
tions increased. Thus, analysis of oncogenic muta-
biomarkers in MDS
tions in a large cohort of patients illustrates the
Great developments in molecular technologies interconnection between the cancer genome and
(high-throughput next-generation sequencing) disease biology.
helped to elucidate differential roles of mutations A passenger mutation has not been selected, has
in MDS. It is estimated that one or more mutated not conferred clonal growth advantage, and has
genes can be identified in over 90% of MDS cases therefore not contributed to cancer development.
(Bejar et al., 2011; Haferlach, 2012; Itzykson et al., Passenger mutations are found within cancer
2013; Papaemmanuil et al., 2013; Haferlach et al., genomes because somatic mutations without func-
2014; Nybakken and Bagg, 2014; Bejar, 2015; tional consequences often occur during cell divi-
Kennedy and Ebert, 2017). The terms “driver” and sion. Thus, a cell that acquires a driver mutation
“passenger” mutation have been used. A driver will already have biologically inert somatic muta-
mutation is causally implicated in oncogenesis. It tions within its genome. These will be carried along
has conferred growth advantage on the cancer cell in the clonal expansion that follows and therefore
and has been positively selected in the microenvi- will be present in all cells of the final cancer.
ronment of the tissue in which the cancer arises. A Frequency and impact of chosen genetic muta-
driver mutation need not be required for mainte- tions identified in MDS are shown in Table 9.6. The
nance of the final cancer (although it often is), but most frequently mutated MDS genes are SF3B1,
it must have been selected at some point along the TET2, ASXL1, SRSF2, DNMT3A, RUNX1, U2AF1,
lineage of cancer development. Driver mutations ZRSR2, STAG2, TP53, NRAS, and EZH2. No single
have equivalent prognostic significance, whether gene is mutated in the majority of MDS cases and
 Methods for the diagnosis and treatment of myelodysplastic syndromes (MDS)  179

Table 9.6  Frequency and impact of chosen genetic mutations identified in MDS

Gene Prognostic Estimated

mutation impact frequency in MDS Additional findings
SF3B1 Favorable 20%–25% Highly associated with ring sideroblasts
TP53 Poor 8% (found in 20% Decreased response to lenalidomide in 5q– MDS,
of 5q– MDS) poor outcome post alloSCT, associated with
complex karyotype, higher BM blasts, and
thrombocytopenia
EZH2 Poor 5% Identifies lower risk MDS with aggressive course
ETV6 Poor 2%–3%
RUNX1 Poor 9%–10%
ASXL1 Poor 15% (found in
40% of CMML)
NRAS Poor 4%–8%
DNMT3A Poor 10%–15% Poor outcome post alloSCT
SRSF2 Poor 12% Occurrence with TET2 mutation is associated with
monocytosis
PTPN11 Poor 1%
U2AF1 Poor 7%–10%
TET2 Unclear 20%–30% Improved response to azacitidine, poor outcome post
alloSCT, occurrence with SRSF2 mutation is
associated with monocytosis
Source: Adapted from Lee EJ et al., Blood Rev 30;2016:1–10.

most genes are mutated in fewer than 5% of cases. were strongly enriched in low-risk MDS, com-
Nearly two-thirds of MDS patients carry a splicing pared to high-risk MDS. MDS patients with SF3B1
factor mutation and more than half have a muta- mutations are less likely to progress to sAML.
tion in an epigenetic regulator. Several additional SF3B1 mutations may define a distinct subset of
pathways are frequently affected including hema- MDS with ring sideroblasts (Malcovati et al., 2015;
topoietic transcription factors, tyrosine kinase sig- Malcovati and Cazzola, 2016).
naling, cohesin genes, and DNA damage response Mutations in other genes (e.g., DNMT3A and
and repair enzymes. U2AF1) are not enriched in specific MDS subtypes.
The clinical impact of selected genetic lesions These mutations represent founder or ancestral
has been studied but only in a limited number of mutations that initiate the early stage of MDS,
MDS patients in serially collected samples (Walter rather than secondary mutations involved in MDS
et  al., 2013). To better understand MDS progres- progression. Mutations of TP53, EZH2, RUNX1,
sion in terms of gene mutations, a large number ETV6, and ASXL1 are associated with greater risk
of serially collected samples is necessary. Clonal than predicted by the IPSS and the IPSS-R (Bejar
architecture is derived from the allelic burden of et  al., 2011). Further studies found additional
driver mutations. During MDS progression, the mutated genes (NRAS, CBL, PRPF8, PTPN11, and
number of mutations, their diversity, and clone NF1) with adverse MDS prognosis independent of
sizes increase (Makishima et al., 2017). Makishima the IPSS-R (Papaemmanuil et al., 2013; Haferlach
et al. (2017) compared high-risk and low-risk MDS et al., 2014). One or more of these adverse mutations
samples. Mutations in 10 genes were enriched in can be found in over one-third of MDS patients.
high-risk MDS, including GATA2, NRAS, KRAS, Therefore, we routinely underestimate disease risk
IDH2, TP53, RUNX1, STAG2, ASXL1, ZRSR2, using conventional analysis (Bejar, 2017). Complex
and TET2. MDS patients with these mutations karyotype MDS patients have about 50% TP53
might be simply associated with secondary AML mutations and have the worst overall outcomes,
(sAML) evolution from MDS. SF3B1 mutations even after treatment.
180  Precision medicine in myelodysplastic syndromes

Table 9.7  Clinical criteria for ICUS, CHIP, and MDS

Nonclonal Lower Higher

ICUS CHIP Risk MDS Risk MDS
Clonality − + + + +
Dysplasia − − − + +
Cytopenia + − + (CCUS) + +
BM blast % <5% <5% <5% <5% <19%
Overall risk Very low Very low Low (?) Low High
Source: Adapted from Steensma DP et al. Blood 126;2015:9–16.
Abbreviation: ICUS, idiopathic cytopenia of undetermined significance; CHIP, clonal hematopoiesis of indeterminate
potential; MDS, myelodysplastic syndrome; CCUS, clonal cytopenia of undetermined significance.

MDS patients with complex karyotype who do significance (ICUS). ICUS is a heterogeneous group
not carry TP53 mutations have better outcomes that includes a clonal subset. Clinical criteria for
approaching patients without complex karyo- ICUS, CHIP, and MDS are shown in Table 9.7.
type (Bejar and Haferlach, 2014). Furthermore,
TP53 mutations are present in approximately 20% Therapies for MDS
of MDS patients with isolated del(5q) and disturb
treatment with immunomodulatory or cereblon- Guidelines for the diagnosis and management of
binding drug lenalidomide and have worse survival adult myelodysplastic syndromes were published
(Kulasekararaj et  al., 2013; Saft et  al., 2014; Zhang (Killick et  al., 2014; Tefferi and Garcia-Manero,
et  al., 2017). The DNA hypomethylating agents 2015; Greenberg et  al., 2017). Erythropoietic
(HMAs) azacitidine and decitabine (Takahashi stimulating agents (ESAs) remain the first-line
et  al., 2016) or allogeneic hematopoietic stem cell treatment in most cases of non-del(5q) low-risk
transplantation (Bejar et  al., 2014; Brierley and MDS, where anemia is the major clinical problem
Steensma, 2016; de Witte et al., 2017) do not abrogate (Santini, 2016; Almeida et al., 2017). Lenalidomide
the negative prognostic effect of TP53 mutations. is used in low-risk MDS with del(5q) (List et  al.,
The overall survival of MDS patients was signifi- 2014; Komrokji and List, 2016). Lenalidomide was
cantly lower in the presence of TP53 mutations, also used in lower-risk patients with MDS without
especially when these mutations are combined with 5q deletion after failure of ESAs (Raza et al., 2008;
poor cytogenetics. Next-generation sequencing is a Sibon et  al., 2012; Santini et  al., 2016). Clinical
valuable technology, but may not be economically and genetic markers with potential for predicting
feasible for routine use. Alternatively, immunohis- responsiveness to lenalidomide are shown in Table
tochemistry is fast, reproducible, and cost effective 9.8 (Stahl and Zeidan, 2017b).
for routine laboratory use. The relationship between Best supportive care (BSC) was considered the
p53 expression and TP53 mutation is known and the primary standard treatment for high-risk MDS,
immunohistochemical pattern of p53 can be used as except for patients younger than 65 years of age
a measure of TP53 mutation burden and adverse with a compatible donor for allogeneic HSCT fol-
clinical outcome in MDS and sAML (Saft et al., lowing myeloablative or nonmyeloablative condi-
2014; McGraw et al., 2016). tioning regimens. Recently, however azacitidine
Several studies have demonstrated that somatic and decitabine as hypomethylating agents have
mutations in the blood of unselected individuals become the standard approach for older high-risk
in the premalignant state with a frequency that MDS patients who are not amenable to allogeneic
increases with age (Steensma et al., 2015; Jan et al., HSCT and have largely replaced chemotherapy.
2017; Makishima et  al., 2017). These individuals Azacitidine provides a survival advantage in high-
have CHIP and the majority of these persons with risk MDS patients, but decitabine does not for a
CHIP will never develop a hematologic disorder. variety of reasons. Guadecitibine (SGI-110) is a
Patients who do not meet the diagnostic criteria for second-generation HMA and a dinucleotide of
MDS have idiopathic cytopenias of undetermined decitabine and deoxyguanosine (Nebbioso et  al.,
 Methods for the diagnosis and treatment of myelodysplastic syndromes (MDS)  181

Table 9.8  Clinical and genetic markers with potential to predict responsiveness to lenalidomide

Marker Prediction of response

Clinical markers (Negoro et al., 2016; Stahl and
Zeidan, 2017b)
Sex Negative association with male sex
Platelet and neutrophil count Negative associations with thrombocytopenia
and neutropenia
IPSS-R Negative association with high/very high risk
Erythropoietic gene expression signature A total of 47 genes involved in erythroid
(Ebert et al., 2008a) differentiation are more highly expressed in
nonresponders to lenalidomide in MDS without
del(5q); patients who respond to lenalidomide
have evidence of defective erythroid
differentiation, but it failed to predict
responders in the larger MDS-005 trial
Mutations (Negoro et al., 2016; Stahl and Zeidan,
2017b)
U2AF1 mutations Negatively associated in del(5q) and non-del(5q)
cohort
TP53 mutations Negatively associated in del(5q) cohort
DEAD box RNA helicase mutations Positively associated in non-del(5q) cohort
High level of full length cereblon mRNA Positively associated in del(5q) cohort with
(Jonasova et al., 2015) response to lenalidomide
Polymorphism A/G located at −29 nucleotides A higher distribution of G alleles is a biomarker of
upstream of the transcription start site of the response to lenalidomide in non-del(5q) cohort
CRBN gene located on chromosome 3 at
nucleotide 3179748; NC_000003.12 (Sardnal
et al., 2013)
Low expression of NPM1 before treatment with Negatively associated with response to
lenalidomide (Chesnais et al., 2014) lenalidomide in non-del(5q) cohort
Immune environment (Epling-Burnette et al., Higher percentage of CD28 negative T cells
2013) negatively associated with response to
lenalidomide in non-del(5q) in non-del(5q) MDS
CD28 negative T cells
Source: Adapted from Stahl M, Zeidan AM, Cancer 2017b, doi: 10.1002/cncr.30585.

2015; Navada and Silverman, 2016). TET2-mutated processes) are also in clinical trials (Ornstein et al.,
MDS have a higher probability of responding to 2015; Garcia-Manero et al., 2017; Stahl and Zeidan,
these hypomethylating agents (Itzykson et  al., 2017a). However, there is a problem due to higher
2011). U2AF1 mutation was significantly associated toxicity in some cases. These agents have improved
with nonresponse to azacitidine, which was consis- both survival and quality of life, but results overall
tent in multivariate analysis. DNMT1, DNMT3A, remain poor (Zeidan et al., 2016; Bhatt and Blum,
RAS, and TP53 were independent predicting fac- 2017). Patients in the highest risk groups had a
tors of shorter overall survival (Jung et al., 2016). median OS from diagnosis of 11–16 months and
Combinations of azacitidine and lenalidomide should be considered for experimental approaches.
(DiNardo et  al., 2015) or azacitidine and histone Combination of HMA with Bcl-2 (B-cell lym-
deacetylase inhibitors vorinostat, entinostat, or phoma-2 inhibitors, e.g., venetoclax/ABT-199,
pracinostat (dual targeting of aberrant epigenetic GDC-0199/ or ABT-737) (Bogenberger et al., 2014;
182  Precision medicine in myelodysplastic syndromes

Jilg et  al., 2016), IDH2 inhibitor (AG-221) (Dang functions as a ligand trap for GDF11 and other
et al., 2017; Medeiros et al., 2017), CD33 antibody- TGF-β family ligands to suppress Smad2/3 activa-
drug conjugates (Daver et al., 2016), Fc-engineered tion and to increase hemoglobin synthesis (Mies
CD33 antibody BI 836858 (Eksioglu et  al., 2017), et al., 2016; Almeida et al., 2017).
CD16xCD33 bispecific killer cell engager (BiKE) Thrombocytopenia is commonly seen in
(Gleason et al., 2014), and hedgehog inhibitors like MDS patients, and bleeding complications
sonidegib (erismodegib, LDE255) and glasdegib are a major cause of morbidity and mortality.
(PF-04449913) (Medler et al., 2015; Zou et al., 2015; Thrombocytopenia is an independent factor for
Hanna and Shevde, 2016) are in ongoing trials. decreased survival and has been incorporated in
It will be a better way to develop new therapies newer prognostic scoring systems. The mechanisms
for MDS to move promising new agents into an of thrombocytopenia are multifactorial and involve
earlier treatment in the course of disease, instead a differentiation block of megakaryocytic progenitor
of conducting randomized trials in end-stage cells, leading to dysplastic, hypolobated, and micro-
patients (Blum, 2016; Zahr et al., 2016). Allogeneic scopic appearing megakaryocytes or increased
transplantation (alloHSCT) remains the only apoptosis of megakaryocytes and their precursors.
curative therapy for MDS patients (de Witte et al., Dysregulated thrombopoietin (TPO) signaling and
2017). Recent findings of a phase III randomized increased platelet destruction through immune or
trial comparing reduced-intensity conditioning nonimmune mechanisms are frequently observed
(RIC) regimens with less toxicity and high-inten- in MDS. The clinical management of patients
sity preparative regimens (myeloablastive condi- with low platelet counts remains challenging and
tioning [MAC]) have shown that overall survival approved chemotherapeutic agents, such as lenalid-
was higher with MAC, but this was not statistically omide and azacytidine, can also lead to a transient
significant. RIC resulted in lower treatment-related worsening of thrombocytopenia. Platelet transfu-
toxicity but higher relapse rates compared with sion was the only supportive treatment option for
MAC (Scott et al., 2017). MDS patients older than clinically significant thrombocytopenia. The TPO
65 years are today more frequently transplanted receptor agonists romiplostim and eltrombopag
and this treatment is covered as participation in have shown clinical activity in clinical trials in MDS
research. All fit patients without comorbidities (Li et al., 2016; Oliva et al., 2017).
should be considered for HSCT when the disease- Rigosertib (ON01910.Na) is an inhibitor of the
related factors allow the recommendation. Somatic phosphoinositide 3-kinase and polo-like kinase
mutations in ASXL1, RUNX1, RAS, JAK2, and pathways (a multikinase inhibitor). Rigosertib
TP53 were associated with unfavorable outcomes has demonstrated therapeutic activity for patients
and shorter survival after allogeneic HSCT (Della with high-risk MDS in clinical trials. Rigosertib-
Porta et al., 2016; Lindsley et al., 2017). induced apoptosis blocked the cell cycle at the
G2/M phase and subsequently inhibited the prolif-
New knowledge about molecular eration of CD34+ cells from MDS, while it mini-
and cellular mechanisms of MDS mally affected the normal CD34+ cells. Rigosertib
as basis for new therapy acted also via the activation of the p53 signaling
pathway. Bioinformatics analysis based on gene
Advances in supportive care are connected with expression profile and flow cytometry analysis
advances in MDS molecular biology and patho- revealed the abnormal activation of the Akt-PI3K,
genesis research (Gill et al., 2016). A new target has Jak-STAT, and Wnt pathways in high-grade MDS,
emerged in treating MDS-related anemia. Defects while the p38 MAPK, SAPK/JNK, and p53 path-
in erythroid differentiation lead to increased ways were abnormally activated in low-grade
production of erythropoietin without effective
­ MDS. Rigosertib could markedly inhibit the acti-
hemoglobin synthesis. Growth differentiation fac- vation of the Akt-PI3K and Wnt pathways, whereas
tor GDF11, an important erythropoiesis regulator, it activated the SAPK/JNK and p53 pathways in
accumulates and can be inhibited by transforming high-grade MDS (Chung, 2016; Garcia-Manero
growth factor β (TGF-β) superfamily ligand trap et al., 2016).
strategies (Dussiot et al., 2014; Suragani et al., 2014). In recent years, much research has been aimed
Luspatercept is an activin receptor antagonist that at the pathogenetic role of haploinsufficient genes
 Methods for the diagnosis and treatment of myelodysplastic syndromes (MDS)  183

located on the deleted region of the long arm the Toll-like receptor pathway, IL-6 induction, and
of chromosome 5 (Pellagatti and Boultwood, regulation of megakaryopoiesis (Starczynowski
2015). Ebert et  al. (2008b) and Dutt et al. (2011) et  al., 2010). Ito et  al. (2010) discovered that tha-
described impaired ribosome biosynthesis due lidomide (founding member of IMiDs) binds
to RPS14 (ribosomal protein 14 of the small ribo- CRBN in the terminal C-region (parts of exons 10
some subunit) gene haploinsufficiency leads to the and 11 of the CRBN gene code in this IMiD bind-
E3 ubiquitin ligase HDM2 (human homologue to ing region). Several researchers confirmed CRBN
mouse double minute 2, major negative regulator as the target of lenalidomide in multiple myeloma
of p53) inactivation by free ribosomal proteins, (MM), lymphoma, chronic lympocytic leukemia,
particularly RPL11. HDM2 degradation drives and del(5q) MDS (Ito and Handa, 2016). CRBN is
p53-mediated apoptosis of erythroid cells carrying an ubiquitously expressed 51kDa protein with a
the del(5q) aberration. This p53-mediated apopto- putative role in cerebral development, especially
sis of erythroid cells is a key effector of hypoplastic memory and learning (Chang and Stewart, 2011;
anemia in MDS patients with del(5q) (Dutt et al., Kim et al., 2016).
2011). RPS14 haploinsufficiency causes a block in We (Jonasova et  al., 2015) have found that
erythroid differentiation mediated by calprotec- del(5q) MDS patients (the so-called 5q minus syn-
tin (the heterodimeric S100 calcium-binding pro- drome) have higher levels of full-length CRBN
teins S100A8 and S100A9) (Schneider et al., 2016). mRNA than other patients with lower risk MDS,
Diminutive somatic deletions in the 5q region linking higher levels of a known lenalidomide
were also found in some of these patients. These target cereblon (CRBN) and an MDS subgroup
deletions were not identified by fluorescence in known to be especially sensitive to lenalidomide.
situ hybridization or cytogenetic testing but by CRBN is a member and substrate receptor of
single nucleotide polymorphism array genotyping the cullin 4 RING E3 ubiquitin ligase complex
(Vlachos et al., 2013). Therefore, some patients orig- (CRL4). CRBN recruits substrate proteins to the
inally considered as MDS patients without del(5q) CRL4 complex for ubiquitination and the subse-
can have a phenotype of atypical 5q– syndrome quent degradation in proteasomes. IMiDs binds
and can be sensitive to lenalidomide therapy. Low to CRBN in CRL4 complex and block normal
RPS14 expression in 50%–70% MDS patients with- endogenous substrates (CRBN and the homeo-
out del(5q) confers a higher apoptosis rate of nucle- box transcription factor MEIS2 in MM) from
ated erythrocytes and predicts prolonged survival binding to CRL4 for ubiquitination and degrada-
(Czibere et al., 2009; Wu et al., 2013). tion (Fischer et  al., 2014). After IMID binding to
What is the mechanism of lenalidomide in CRBN, CRL4 complex is recruiting transcrip-
del(5q)MDS based on what has been achieved and tion factors Ikaros (IKZF1) and Aiolos (IKZF3)
elucidated to date? Lenalidomide stabilizes E3 for ubiquitination and degradation in MM cells
ubiquitin ligase HDM2, thereby accelerating p53 (Krönke et  al., 2014). Degradation of these tran-
degradation (Wei et al., 2009, 2013). Lenalidomide scription factors explains lenalidomide’s growth
inhibits phosphatases PP2a and Cdc25c (coregu- inhibition of MM cells and increased interleukin-2
lators of cell cycle in which genes are very com- (IL-2) release from T cells. However, it is unlikely
monly deleted in del(5q) MDS) with consequent that degradation of IKZF1 and IKZF3 accounts
G2 arrest of del(5q) MDS progenitors and their for lenalidomide’s activity in MDS with del(5q).
apoptosis. PP2a and Cdc25c inhibition by lenalid- Krönke et al. (2015) identified a novel target casein
omide suppress HDM2 autoubiquitination and kinase1A1 (CSNK1A1) by quantitative proteomics
subsequent degradation. Thus, lenalidomide has in the myeloid cell line KG-1. CSNK1A1 is encoded
been shown to not only reverse apoptosis within in the del(5q) commonly deleted region and the
the erythroid compartment, but also directly gene is haploinsufficient. Lenalidomide treatment
induce apoptosis of the myeloid clone in del(5q) leads to increased ubiquitination of the remain-
MDS (Gandhi et al., 2006; Matsuoka et al., 2010). ing CSNK1A1 and decreased protein abundance.
Lenalidomide upregulates expression of other two CSNK1A1 negatively regulates β-catenin, which
haploinsufficient genes located on chromosome drives stem cell self-renewal, and CSNK1A1 haplo-
5q, genes for microRNAs (miR-145 and miR-146a) insufficiency causes the initial clonal expansion in
(Venner et  al., 2013). These miRs are involved in patients with the del(5q) MDS and contributes to
184  Precision medicine in myelodysplastic syndromes

the pathogenesis of del(5q) MDS. The further inhi- plasma. Further, like somatic gene mutations,
bition of CSNK1A1 in del(5q) MDS is associated S100A9-induced signaling activates NADPH oxi-
with del(5q) failure and p53 activation. The inhibi- dase (NOX) and increasing levels of reactive oxy-
tion of CSNK1A1 reduced RPS6 phosphorylation, gen species (ROS). ROS initiates cation influx, cell
induced p53 expression and growth inhibition, swelling, and β-catenin activation. Knockdown
and triggered the myeloid differentiation program. of NLRP3 or caspase-1, neutralization of S100A9,
TP53-null leukemia did not respond to CSNK1A1 and pharmacologic inhibition of NLRP3 or NOX
inhibition, strongly supporting the importance of suppress pyroptosis, ROS generation, and nuclear
the p53 expression for the yield of CSNK1A1 inhi- β-catenin in MDSs and are sufficient to restore
bition. CSNK1A1 mutations have been recently effective hematopoiesis. Thus, alarmins and
found in 5%–18% of MDS patients with del(5q) founder gene mutations in MDSs cause a common
(Schneider et al., 2014). These mutations are associ- redox-sensitive inflammasome circuit. They are
ated similarly to the effect of TP53 mutations with new candidates for therapeutic intervention.
rise to a poor prognosis in del(5q) MDS (Smith The effects of lenalidomide in non-del(5q) are
et al., 2015). Other studies did not find impact of thought to be secondary to modulation of the
CSNK1A1 mutations on lenalidomide treatment immune system. Hyperactivated T cells inhibit
in del(5q) MDS (Heuser et al., 2015; Negoro et al., hematopoiesis. Immunosuppressive therapies with
2016). antithymocyte globulin alone and in combination
While CSNK1A1 is the CRL4CRBN target in with prednisone or cyclosporine show response
del(5q) MDS, CRL4CRBN targets in lower risk non- rates between 25% and 40% (Haider et  al., 2016;
del(5q) remain to be determined. The mechanism Stahl and Zeidan, 2017b). Several trials with vari-
of action of lenalidomide is still unclear in non- ous combinations of lenalidomide are shown in
del(5q) MDS cells. Recent evidence shows that Table 9.9.
lenalidomide directly improves erythropoietin
receptor (EPOR) signaling by EPOR upregula- THE CONCEPT OF PERSONALIZED
tion mediated by a posttranscriptional mechanism OR PRECISION MEDICINE
(Basiorka et al., 2016a). Lenalidomide stabilizes the
EPOR protein by inhibition of the E3 ubiquitin The concept of personalized or precision medicine
ligase RNF41 (ring finger protein 41, also known as was introduced in the State of the Union speech
neuregulin receptor degradation protein-1 [Nrdp1] given by U.S. President Barack Obama in 2015.
and fetal liver ring finger [FLRF]) (Basiorka et al., He announced the precision medicine initiatives
2016a) and induces lipid raft assembly to enhance that would enable development of individualized
EPOR signaling in MDS erythroid progenitors patient care in order to improve patient outcomes.
(McGraw et al., 2012, 2014). Initial attempts to analyze the human genome
After failure of ESAs, lenalidomide yields red started with cytogenetic analyses and chromosome
blood cell transfusion independence in 20%–30% banding. Chromosomal analysis can only visualize
of lower-risk non-del(5q) MDS. Indeed, several the human genome at 0.5%–1% resolution. To fur-
observations suggest an additive effect of ESA and ther improve the detection of genomic abnormali-
lenalidomide in this situation (Komrokji et  al., ties, chip and array techniques were developed.
2012; Toma et  al., 2016). Basiorka et  al. reported Single nucleotide polymorphisms, frequently
activation of the NLRP3 inflammasome in MDS called SNPs (pronounced “snips”), are the most
(Basiorka et  al., 2016b; Sallman et  al., 2016). common type of genetic variation among people.
NRLP3 drives clonal expansion and pyroptotic Each SNP represents a difference in a single DNA
cell death. Independent of genotype, MDS hema- building block, called a nucleotide. For example, a
topoietic stem and progenitor cells (HSPCs) over- SNP may replace the nucleotide cytosine (C) with
express inflammasome proteins. Activated NLRP3 the nucleotide thymine (T) in a certain stretch of
complexes direct then activation of caspase-1, DNA. SNPs occur normally throughout a person’s
generation of interleukin-1β (IL-1β) and IL-18, DNA. They occur once in every 300 nucleotides on
and pyroptotic cell death. Mechanistically, pyrop- average, which means there are roughly 10 million
tosis is triggered by the alarmin S100A9 that is SNPs in the human genome. Most commonly, these
found in excess in MDS HSPCs and bone marrow variations are found in the DNA between genes.
 Future opportunities  185

Table 9.9  List of clinical trials investigating combinations of lenalidomide

Combination Clinical trial

partner Mechanism of action number Phase MDS population
Cyclosporine-A Immune-suppressive drug NCT00840827 Phase 2 Low, intermediate-1;
non-del(5q)
Prednisone Immune-suppressive drug NCT01133275 Phase 2 Low, intermediate-1;
non-del(5q)
Romiplostim TPO agonist NCT00418665 Phase 2 All MDS
Eltrombopag TPO agonist NCT02928419 Phase 2 Low, intermediate-1;
NCT01772420 with and without
del(5q)
Ezatiostat Small molecule glutathione NCT01062152 Phase 1–2 Low, intermediate-1;
analog inhibitor of non-del(5q)
glutathione S-transferase
leads to activation of Jun
kinase
Clofarabine Chemotherapy NCT01629082 Phase 1–2 Higher-risk MDS
Melphalan Chemotherapy NCT00744536 Phase 2 Higher-risk MDS
Bortezomib Proteasome inhibitor NCT02312102 Phase 1 Higher-risk MDS
NCT00580242
Lintuzumab Anti-CD33 monoclonal NCT00502112 Phase 1 All MDS
antibody
Source: Adapted from Stahl M, Zeidan AM, Cancer 2017b, doi: 10.1002/cncr.30585.

They can act as biological markers, helping scien- of the Human Genome Project, a 2.91 billion base
tists locate genes that are associated with disease. pair (bp) consensus sequence of the euchromatic
When SNPs occur within a gene or in a regulatory portion of the human genome was generated by
region near a gene, they may play a more direct role the whole genome shotgun sequencing method
in disease by affecting the gene’s function. Most and was published in 2001 (Venter et  al., 2001).
SNPs have no effect on health or development. The genomic landscape of MDS was analyzed by
Some of these genetic differences, however, have these methods (Bejar et al., 2011; Haferlach, 2012;
proven to be very important in the study of human Itzykson et  al., 2013; Papaemmanuil et  al., 2013;
health. Researchers have found SNPs that may help Haferlach et  al., 2014; Nybakken and Bagg, 2014;
predict an individual’s response to certain drugs, Bejar, 2015; Kennedy and Ebert, 2017). The impact
susceptibility to environmental factors such as of genomic data on diagnosis, prognosis, and ther-
toxins, and risk of developing particular diseases. apy of individual MDS patients was described in
SNPs can also be used to track the inheritance of this chapter.
disease genes within families.
The first DNA sequencing effort was developed FUTURE OPPORTUNITIES
in the 1970s when fragments of DNA were cloned
and amplified by polymerase chain reaction (PCR). Genomic and clinical data can be used to fur-
This method enabled reading of sequences about ther understand the biology and pathophysiology
1000 bp in length. New genomic technologies of MDS, a group of heterogeneous and complex
and mainly advances in bioinformatics helped to hematologic disorders primarily found within
develop next-generation sequencing (NGS), whole the older population. A risk-adapted treatment
genome sequencing, whole exome sequencing, and strategy is necessary in this disease with a highly
targeted deep sequencing of specific mutations variable clinical course. However, this personal-
(Platzbecker and Fenaux, 2015; Nazha and Sekeres, ized approach, including performance status, EPO
2017; Sheikine et  al., 2017). Since the completion level, ferritin level, prognostic scoring systems
186  Precision medicine in myelodysplastic syndromes

(good risk or poor risk), cytogenetics (especially Table 9.10  Frequent alterations of mesenchymal
presence of del(5q) for targeted treatment with stem and progenitor cells (MSPCs) in MDS
lenalidomide), and mutations, is not supported
Alterations
by prospective randomized trials. Although limi-
tations exist in the application of precision medi- Cytogenetic aberrations in MDS-MSPCs
cine in MDS, mutation profiling analyses will (chromosomes 1 and 7 are more frequently
further improve the classification of MDS and involved)
discover new therapeutic biomarkers. Recently, Lower expression of Dicer1, DROSHA, AURKA,
great progress has been done in the study of AURKB genes in MDS-MSPCs
bone marrow (BM) microenvironment in MDS Altered immunophenotype in MDS-MSPCs;
(Bulycheva et al., 2015; Cogle et al., 2015; Rankin decreased CD44 and CD49e, CD90, CD104,
et al., 2015; Balderman et al., 2016; Glenthɵj et al., and CD105 expression, increased CXCL12
2016; Pleyer et al., 2016; Korn and Méndez-Ferrer, expression
2017). This development will extend our current Impaired proliferation and differentiation
limited therapeutic portfolio by detection of new capacity of MDS-MSPCs
therapeutic targets for targeted therapies. These Impaired cytokine production, including IL-32
therapies will offer in the near future truly per- by MDS-MSPCs
sonalized approaches (Platzbecker and Fenaux, Dysregulation of Wnt signaling pathway in
2015). Accumulating evidence suggests that the MDS-MSPCs
altered BM microenvironment in general, and in Impaired hematopoietic stem and progenitor
particular the components of the stem cell niche, cells (HSPC) support by MDS-MSPCs
including mesenchymal stem and progenitor cells
(MSPCs) and their progeny, play a pivotal role in Source: Adapted from Bulycheva E et  al., Leukemia
29;2015:259–68.
the evolution and propagation of MDS. MSPCs
have potent immunomodulatory capacities and been identified as a potent inhibitor of the Wnt/β-
communicate with diverse immune cells, but also catenin pathway. It stimulates casein kinase 1α
interact with various other cellular components (CK1α), which is part of the β-catenin destruction
of the microenvironment as well as with normal complex where it phosphorylates β-catenin at ser-
and leukemic stem and progenitor cells. Moreover, ine 45, priming the subsequent phosphorylation of
compared to normal MSPCs, MSPCs in MDS and β-catenin by glycogen synthase kinase 3β (GSK3β).
AML often exhibit altered gene expression profiles, These phosphorylations mark β-catenin for pro-
an aberrant phenotype, and abnormal functional teasomal degradation. This is one of the central
properties (Table 9.10). These alterations contribute regulating events controlling the Wnt signal-
to the “reprogramming” of the stem cell niche into ing pathway. Furthermore, CK1 has been shown
a disease-permissive microenvironment where an to additionally regulate Wnt signaling through
altered immune system, abnormal stem cell niche phosphorylation of lymphocyte enhancer factor-1
interactions, and an impaired growth control lead (LEF-1) and β-catenin leading to disruption of the
to disease progression. Common signaling path- LEF-1/β-catenin transcription complex.
ways, such as PI3K/AKT and WNT/β-catenin,
regulate multiple MSPC properties, including pro- ACKNOWLEDGMENTS
liferation, differentiation, and cell-cell interaction.
Impairment of these signaling pathways was found This work was supported by a research project for
in stromal cells of BM microenvironment in MDS. conceptual development of research organization
This finding highlights a potential new strategy for (00023736; Institute of Hematology and Blood
treating some myeloid disorders. Inhibition of the Transfusion, Prague) from the Ministry of Health
Wnt/β-catenin pathway by pyrvinium in the bone of the Czech Republic.
marrow niche prevented the development of MDS
in the Apcdel/+ MDS mouse model (Falconi et al., CONFLICT OF INTEREST
2016). Pyrvinium, an anthelmintic (against infec-
tions with parasitic worms) drug approved by the There are no commercial or other associations in
U.S. Food and Drug Administration (FDA), has connection with this review article.
 References 187

REFERENCES Bejar R. Implications of molecular genetic diver-

sity in myelodysplastic syndromes. Curr Opin
Adés L, Itzkson R, and Fenaux P. Myelodysplastic Hematol 24;2017:73–8.
syndromes. Lancet 383;2014:2239–52. Bejar R, and Haferlach T. TP53 mutation status
Almeida A, Fenaux P, List AF, Raza A, divides MDS patients with complex karyo-
Platzbecker U, and Santini V. Recent types into distinct prognostic risk groups:
advances in the treatment of lower-risk non- Analysis of combined datasets from the
del(5q) myelodysplastic syndromes (MDS). International Working Group for MDS-
Leuk Res 52;2017:50–7. Molecular Prognosis Committee. Blood
Arber DA, Orazi A, Hasserjian R, Thiele J, 124;2014:532a.
Borowitz MJ, Le Beau MM, Bromfield CD, Bejar R, and Steensma DP. Recent develop-
Cazzola M, and Vardiman JW. The 2016 ments in myelodysplastic syndromes. Blood
revision to the World Health Organization 124;2014:2793–803.
classification of myeloid neoplasms and acute Bejar R, Stevenson K, Abdel-Wahab O et al.
leukemia. Blood 127;2016:2391–405. Clinical effect of point mutations in myelo-
Avgerinou C, Alamanos Y, Zikos P et al. The dysplastic syndromes. N Engl J Med
incidence of myelodysplastic syndrome in 364;2011:2496–506.
Western Greece is increasing. Ann Hematol Bejar R, Stevenson K, Caughey B et al. Somatic
92;2013:877–87. mutations predict poor outcome in patients
Babushok DV, and Bessler M. Genetic predis- with myelodysplastic syndrome after hemato-
position syndromes: When should they be poietic stem-cell transplantation. J Clin Oncol
considered in the work-up of MDS? Best Pract 32;2014:2691–8.
Res Clin Hematol 28;2015:55–68. Bennett JM, Catovsky D, Daniel MT, Flandrin
Babushok DV, Bessler M, and Olson TS. G, Galton DA, Gralnick HR, and Sultan
Genetic predisposition to myelodysplastic C. Proposals for the classification of the
syndrome and acute myeloid leukemia in myelodysplastic syndromes. Br J Haematol
children and young adults. Leuk Lymphoma 51;1982:189–99.
57;2016:520–36. Bhatt G, and Blum W. Making the most of
Bacher U, Schanz J, Braulke F, and Haase D. Rare hypomethylating agents in myelodysplastic
cytogenetic abnormalities in myelodysplastic syndrome. Curr Opin Hematol 24;2017:79–88.
syndromes. Mediterr J Hematol Infect Dis Blum W. Why the stagnation in effective ther-
7;2015:e2015034. apy for MDS? Best Pract Res Clin Hematol
Balderman SE, Li AJ, Hoffman CM et al. 29;2016:334–40.
Targeting of the bone marrow misroenvi- Bogenberger JM, Kornblau SM, Pierceall WE
ronment improves outcome in a murine et al. BCL-2 family proteins as 5-azacytidine
model of myelodysplastic syndrome. Blood sensitizing targets and determinants of
127;2016:616–25. response in myeloid malignancies. Leukemia
Bannon SA, and DiNardo CD. Hereditary predis- 28;2014:1657–65.
position to myelodysplastic syndrome. Int J Bowen DT. Occupational and environmental
Mol Sci 17;2016:838. etiology of MDS. Best Pract Res Clin Hematol
Basiorka AA, McGraw KL, De Ceuninck L et al. 26;2013:319–26.
Lenalidomide stabilizes the erythropoietin Brierley CK, and Steensma DP. Allogeneic
receptor by inhibiting the E3 ubiquitin ligase stem cell transplantation in myelodysplastic
RNF41. Cancer Res 76;2016a:3531–40. syndromes: Does pretransplant clonal bur-
Basiorka AA, McGraw KL, Eksioglu EA et al. The den matter? Curr Opin Hematol 23;2016:
NLRP3 inflammasome functions as a driver 167–74.
of the myelodysplastic syndrome phenotype. Bulycheva E, Rauner M, Medyouf H, Theurl I,
Blood 128;2016b:2960–75. Bornhäuser M, Hofbauer LC, and Platzbecker
Bejar R. Myelodysplastic syndromes diagnosis: U. Myelodysplasia is in the niche: Novel
What is the role of molecular testing? Curr concepts and emerging therapies. Leukemia
Hematol Malig Rep 10;2015:282–91. 29;2015:259–68.
188  Precision medicine in myelodysplastic syndromes

Chang XB, and Stewart AK. What is the func- on the outcomes of patients with myelodys-
tional role of the thalidomide binding plastic syndromes treated with allogeneic
protein cereblon? Int J Biochem Mol Biol hematopoietic stem-cell transplantation.
2;2011:287–94. J Clin Oncol 34;2016:3627–37.
Chesnais V, Renneville A, Toma A et al. NPM1 Della Porta MG, and Picone C. Diagnostic util-
expression level and a CRBN polymorphism ity of flow cytometry in myelodysplastic
are able to predict the rate of response to syndromes. Mediterr J Hematol Infect Dis
lenalidomide in non del(5q) lower risk MDS 9;2017:e2017017.
patients resistant to erythropoiesis-stimu- Della Porta MG, Tuechler H, Malcovati L et al.
lating agents: The GFM experience. Blood Valadation of WHO classification-based
124;2014:533. Prognostic Scoring System (WPSS) for
Chung SS. Novel therapeutic strategies in MDS: myelodysplastic syndromes and comparison
Do molecular genetics help. Curr Opin with the revised International Prognostic
Hematol 23;2016:79–87. Scoring System (IPSS-R). A study of the
Churpek JE, Pyrtel K, Kanchi KL et al. Genomic International Working Group for Prognosis
analysis of germ line and somatic variants in in Myelodysplasia (IWG-PM). Leukemia
familial myelodysplasia/acute myeloid leuke- 29;2015:1502–13.
mia. Blood 126;2015:2484–90. de Witte T, Bowen D, Robin M et al. Allogeneic
Cogle CR, Craig BM, Rollison DE, and List AF. hematopoietic stem cell transplantation
Incidence of myelodysplastic syndromes using for MDS and CMML: Recommendations
a novel claims-based algorithm: High number from an international expert panel. Blood
of uncaptured cases by cancer registries. 129;2017:1753–62.
Blood 117;2011:7121–5. DiNardo CD, Daver N, Jabbour E et al. 2015.
Cogle CR, Saki N, Khodadi E, Li J, Shahjahani Sequential azacitidine and lenalidomide
M, and Azizidoost S. Bone marrow niche in in patients with high-risk myelodysplastic
the myelodysplastic syndromes. Leuk Res syndromes and acute myeloid leukaemia: A
39;2015:1020–7. single-arm, phase 1/2 study. Lancet Haematol
Cremers EMP, Westers TM, Alhan C, Cali C, 2: e12–e20.
Wondergem MJ, Poddighe PJ, Ossenkoppele Dussiot M, Maciel TT, Fricot A et al. An activin
GJ, and van de Loosdrecht AA. receptor IIA ligand trap corrects ineffective
Multiparameter flow cytometry is instrumen- erythropoiesis in beta-thalassemia. Nat Med
tal to distinguish myelodysplastic syndromes 20;2014:398–407.
from non-neoplastic cytopenias. Eur J Cancer Dutt S, Narla A, Lin K et al. Haploinsufficiency
54;2016:49–56. for ribosomal protein genes causes selective
Czibere A, Bruns I, Junge B, Singh R, Kobbe G, activation of p53 in human erythroid progeni-
Haas R, and Germing U. Low RPS14 expres- tor cells. Blood 117;2011:2567–76.
sion is common in myelodysplastic syndromes Ebert BL, Galili N, Tamayo P et al. An erythroid
without 5q- aberration and defines a sub- differentiation signature predicts response to
group of patients with prolonged survival. lenalidomide in myelodysplastic syndrome.
Haematologica 94;2009:1453–5. PLOS Med 5;2008a:e35.
Dang L, Yen K, and Attar EC. IDH mutations Ebert BL, Pretz J, Bosco J et al. Identification of
in cancer and progress toward develop- RPS14 as a 5q- syndrome gene by RNA inter-
ment of targeted therapeutics. Ann Oncol ference screen. Nature 451;2008b:335–9.
27;2017:599–608. Eksioglu EA, Chen X, Heider K-H et al. Leukemia
Daver N, Kantarjian H, Ravandi F et al. A phase II 2017, doi: 10.1038/leu.2017.21.
study of decitabine and gemtuzumab ozogami- Epling-Burnette PK, Han Y, Mailloux AW et al.
cin in newly diagnosed and relapsed acute Novel predictor of lenalidomide response in
myeloid leukemia and high-risk myelodysplastic non-del(5q) MDS reveals linkage to molecular
syndrome. Leukemia 30;2016:268–73. mechanism: First characterization of T-cell
Della Porta MG, Galli A, Bacigalupo A et al. function in cereblon homozygous deficient
Clinical effects of driver somatic mutations mice. Blood 122;2013:748.
 References 189

Falconi G, Fabiani E, Fianchi L et al. Impairment Glenthɵj A, Ørskov AD, Hansen JW, Hadrup SR,
of PI3K/AKT and WNT/β-catenin pathways O’Connel C, and Grɵnbaek K. 2016. Immune
in bone marrow mesenchymal stem cells mechanisms in myelodysplastic syndrome.
isolated from patients with myelodysplastic Int J Mol Sci 17: 944.
syndromes. Exp Hematol 44;2016:75–83. Godley LA. Inherited predisposition to
Fischer ES, Bӧhm K, Lydeard JR et al. Structure of acute myeloid leukemia. Semin Hematol
the DDB1-CRBN E3 ubiquitin ligase in com- 51;2014:306–21.
plex with thalidomide. Nature 512;2014:49–53. Greenberg P, Cox C, LeBeau MM et al.
Gandhi AK, Kang J, Naziruddin S, Parton A, International scoring system for evaluating
Schafer PH, and Stirling D. Lenalidomide prognosis in myelodysplastic syndromes.
inhibits proliferation of Namalwa CSN.70 cells Blood 89;1997:2079–88.
and interferes with Gab1 phosphorylation and Greenberg PL, Stone RM, Al-Kali A et al.
adaptor protein complex assembly. Leuk Res Myelodysplastic syndromes, Version 2.
30;2006:849–58. 2017, NCCN Clinical Practice Guidelines
Ganser A, and Heuser M. Therapy-related in Oncology. J Natl Compr Canc Netw
myeloid neoplasms. Curr Opin Hematol 15;2017:60–87.
24;2017:152–8. Greenberg PL, Tuechler H, Schanz J et al.
Garcia-Manero G, Fenaux P, Al-Kali A et al. Revised international prognostic scoring
Rigosertib versus best supportive care for system for myelodysplastic syndromes. Blood
patients with high-risk myelodysplastic syn- 120;2012:2454–65.
dromes after failure of hypomethylating drugs Haase D. Cytogenetic features in myelodysplastic
(ONTIME): A randomised, controlled, phase 3 syndromes. Ann Hematol 87;2008:515–26.
trial. Lancet Oncol 17;2016:496–508. Haase D, Germing U, Schanz J et al. New insights
Garcia-Manero G, Montalban-Bravo G, Berdeja into the prognostic impact of the karyotype in
JG et al. Phase 2, randomized, double-blind MDS and correlation with subtypes: Evidence
study of pracinostat in combination with from a core dataset of 2124 patients. Blood
azacitidine in patients with untreated, higher- 110;2007:4385–95.
risk myelodysplastic syndromes. Cancer Haferlach T. Molecular genetics in myelodysplas-
123;2017:994–1002. tic syndromes. Leuk Res 36;2012:1459–62.
Germing U, Aul C, Niemeyer CM, Haas R, and Haferlach T, Nagata Y, Grossmann V et al.
Bennett JM. Epidemiology, classification Landscape of genetic lesions in 944 patients
and prognosis of adults and children with with myelodysplastic syndromes. Leukemia
myelodysplastic syndromes. Annu Hematol 28;2014:241–7.
87;2008:691–9. Haider M, Al Ali N, Padron E, Epling-Burnette
Giagounidis A, and Haase D. Morphology, cyto- P, Lancet J, List A, and Komrokji R. 2016.
genetics and classification of MDS. Best Pract Immunosuppressive therapy: Exploring an
Res Clin Hematol 26;2013:337–53. underutilized treatment option for myelodys-
Gill H, Leung AYH, and Kwong Y-L. Molecular plastic syndrome. Clin Lymphoma Myeloma
and cellular mechanisms of myelodysplastic Leuk 16(Suppl): S44–8.
syndrome: Implications on targeted therapy. Hanna A, and Shevde LA. Hedgehog signaling:
Int J Mol Sci 17;2016:440. Modulation of cancer properties and tumor
Gillis NK, Ball M, Zhang Q et al. Clonal hae- microenvironment. Mol Cancer 15;2016:24.
mopoiesis and therapy-related myeloid Heuser M. Therapy-related myeloid neoplasms:
malignancies in elderly patients: A proof-of- Does knowing the origin help to guide treat-
concept, case-control study. Lancet Oncol ment? Hematology Am Soc Hematol Educ
18;2017:112–21. Program 2016(1);2016:24–32.
Gleason MK, Ross JA, Warlick ED et al. Heuser M, Meggendorfer M, Cruz MM et al.
CD16xCD33 bispecific killer cell engager Frequency and prognostic impact of casein
(BiKE) activates NK cells against primary kinase 1A1 mutations in MDS patients with
MDS and MDSC CD33+ targets. Blood deletion of chromosome 5q. Leukemia
123;2014:3016–26. 29;2015:1942–5.
190  Precision medicine in myelodysplastic syndromes

Ito T, Ando H, Suzuki T, Ogura T, Hotta K, Komrokji RS, Lancet JE, Swern AS, Chen N,
Imamura Y, Yamaguchi Y, and Handa Paleveda J, Lush R, Saba HI, and List AF.
H. Identification of a primary target of Combined treatment with lenalidomide
thalidomide teratogenicity. Science and epoetin alfa in lower-risk patients
327;2010:1345–50. with myelodysplastic syndrome. Blood
Ito T, and Handa H. Cereblon and its down- 120;2012:3419–24.
stream substrates as molecular targets of Komrokji RS, and List AF. Short- and long-term
immunomodulatory drugs. Int J Hematol benefits of lenalidomide treatment in patients
104;2016:293–9. with lower-risk del(5q) myelodysplastic syn-
Itzykson R, Kosmider O, Cluzeau T et al. Impact dromes. Annals Oncol 27;2016:62–8.
of TET2 mutations on response rate to Korn C, and Méndez-Ferrer S. Myeloid malig-
azacitidine in myelodysplastic syndromes nancies and the microenvironment. Blood
and low blast count acute myeloid leukemias. 129;2017:811–22.
Leukemia 25;2011:1147–52. Krӧnke J, Fink EC, Hollenbach PW et al.
Itzykson R, Kosmider O, and Fenaux MD. Somatic Lenalidomide induces ubiquitination and
mutations and epigenetic abnormalities in degradation of CK1α in del(5q) MDS. Nature
myelodysplastic syndromes. Best Pract Res 523;2015:183–8.
Clin Hematol 26;2013:355–64. Krӧnke J, Udeshi ND, Narla A et al. Lenalidomide
Jan M, Ebert BL, and Jaiswal S. Clonal hemato- causes selective degradation of IKZF1 and
poiesis. Semin Hematol 54;2017:43–50. IKZF3 in multiple myeloma cells. Science
Jilg S, Reidel V, Müller-Thomas C et al. Blockade 343;2014:301–5.
of BCL-2 proteins efficiently induces apop- Kulasekararaj AG, Smith AE, Mian SA et al. 2013.
tosis in progenitor cells of high-risk myelo- TP53 mutations in myelodysplastic syndrome
dysplastic syndromes patients. Leukemia are strongly correlated with aberrations of
30;2016:112–23. chromosome 5, and correlate with adverse
Jonas BA, and Greenberg PL. MDS prognostic prognosis. Br J Haematol 160: 660–72.
scoring systems—Past, present, and future. Lee EJ, Podoltsev N, Gore SD, and Zeidan AM.
Best Pract Res Clin Haematol 28;2015:3–13. The evolving field of prognostification and
Jonasova A, Bokorova R, Polak J et al. High risk stratification in MDS: Recent develop-
level of full-length cereblon mRNA in ments and future directions. Blood Rev
lower risk myelodysplastic syndrome with 30;2016:1–10.
isolated 5q deletion is implicated in the Li W, Morrone K, Kambhampati S, Will B, Steidl
efficacy of lenalidomide. Eur J Haematol U, and Verma A. Thrombocytopenia in MDS:
95;2015:27–34. Epidemiology, mechanisms, clinical conse-
Jung SH, Kim JY, Yim SH et al. Somatic mutations quences and novel therapeutic strategies.
predict outcomes of hypomethylating therapy Leukemia 30;2016:536–44.
in patients with myelodysplastic syndrome. Lichtman MA. The World Health Organization
Oncotarget 23;2016:55264–75. revisits the classification of the myelodysplas-
Kennedy JA, and Ebert BL. Clinical implications tic syndromes: Improvement and insufficien-
of genetic mutations in myelodysplastic syn- cies. Blood Cells Mol Dis 60;2016:12–5.
drome. J Clin Oncol 35;2017:968–74. Lindsley RC, Saber W, Mar BG et al. Prognostic
Killick SB, Carter C, Culligan D et al. Guidelines mutations in myelodysplastic syndrome after
for the diagnosis and management of adult stem-cell transplantation. N Engl J Med
myelodysplastic syndromes. Br J Haematol 376;2017:536–47.
164;2014:503–25. List AF, Bennett JM, Sekeres MA, Skikne B,
Kim HK, Ko TH, Nyamaa B et al. Cereblon Fu T, Shammo JM, Nimer SD, Knight RD,
in health and disease. Pflugers Arch and Giagounidis A. Extended survival and
468;2016:1299–309. reduced risk of AML progression in erythroid-
Koeffler HP, and Leong G. Preleukemia: responsive lenalidomide-treated patients
One name, many meanings. Leukemia with lower-risk del(5q) MDS. Leukemia
31;2017:534–42. 28;2014:1033–40.
 References 191

Lubeck DP, Danese M, Jennifer D, Miller K, Medeiros BC, Fathi AT, DiNardo CD, Pollyea DA,
Richhariya A, and Garfin PM. Systematic Chan SM, and Swords R. Isocitrate dehydro-
literature review of the global incidence and genase mutations in myeloid malignancies.
prevalence of myelodysplastic syndrome and Leukemia 31;2017:272–81.
acute myeloid leukemia. Blood 128;2016:5930. Medler CJ, Waddell JA, and Solimando DA.
Ma X, Does M, Raza A, and Mayne ST. Cancer chemotherapy update. Gefitinib and
Myelodysplastic syndromes: Incidence sonidegib. Hosp Pharm 50;2015:868–72.
and survival in the United States. Cancer Mies A, Hermine O, and Platzbecker U. Activin
109;2007:1536–42. receptor II ligand traps and their therapeutic
Makishima H, Yoshizato T, Yoshida K et al. potential in myelodysplastic syndromes with
Dynamics of clonal evolution in myelodysplas- ring sideroblasts. Curr Hematol Malig Rep
tic syndromes. Nature Genet 49;2017:204–12. 11;2016:416–24.
Malcovati L, and Cazzola M. Recent advances Navada SC, and Silverman LR. The safety and
in understanding of myelodysplastic syn- efficacy of rigosertib in the treatment of
dromes with ring sideroblasts. Br J Haematol myelodysplastic syndromes. Expert Rev
174;2016:847–58. Anticancer Ther 16;2016:805–10.
Malcovati L, Germing U, Kuendgen A et al. Nazha A, and Sekeres MA. Precision medicine in
Time-dependent prognostic scoring system myelodysplastic syndromes and leukemias:
for predicting survival and leukemic evolution Lessons from sequential mutations. Annu Rev
in myelodysplastic syndromes. J Clin Oncol Med 68;2017:127–37.
25;2007:3503–510. Nebbioso A, Benedetti R, Conte M, Iside C, and
Malcovati L, Karimi M, Papaemmanuil E et al. Altucci L. Genetic mutations in epigenetic
SF3B1 mutation identifies a distinct subset of modifiers as therapeutic targets in acute
myelodysplastic syndrome with ring sidero- myeloid leukemia. Expert Opin Ther Targets
blasts. Blood 126;2015:233–41. 19;2015:1187–202.
Malcovati L, Porta MG, Pascutto C et al. Negoro E, Radiovoyevitch T, Polprasert C et al.
Prognostic factors and life expectancy in Molecular predictors of response in patients
myelodysplastic syndromes classified accord- with myeloid neoplasms treated with lenalido-
ing to WHO criteria: A basis for clinical deci- mide. Leukemia 30;2016:2405–9.
sion making. J Clin Oncol 23;2005:7594–603. Neukirchen J, Schoonen WM, Strupp C,
Matsuoka A, Tochigi A, Kishimoto M et al. Gattermann N, Aul C, Haas R, and Germing U.
Lenalidomide induces cell death in an MDS- Incidence and prevalence of myelodysplastic
derived cell line with deletion of chromosome syndromes: Data from the Düsseldorf MDS-
5q by inhibition of cytokinesis. Leukemia registry. Leuk Res 35;2011:1591–6.
24;2010:748–55. Nimer SD. Myelodysplastic syndromes. Blood
McGraw KL, Basiorka AA, Johnson JO et al. 111;2008:4841–51.
Lenalidomide induces lipid raft assembly to Nybakken GE, and Bagg A. The genetic basis
enhance erythropoietin receptor signaling in and expanding role of molecular analysis in
myelodysplastic syndrome progenitors. PLOS the diagnosis, prognosis, and therapeutic
ONE 9;2014:e114249. design for myelodysplastic syndromes. J Mol
McGraw KL, Fuhler GM, Johnson JO, Clark Diagn 16;2014:145–58.
JA, Caceres GC, Sokol L, and List AF. Oliva EN, Alati C, Santini V et al. Eltrombopag
Erythropoietin receptor signaling is versus placebo for low-risk myelodysplastic
membrane raft dependent. PLOS ONE syndromes with thrombocytopenia (EQoL-
7;2012:e34477. MDS): Phase 1 results of a single-blind, ran-
McGraw KL, Nguyen J, Komrokji RS et al. domised, controlled, phase 2 superiority trial.
Immunohistochemical pattern of p53 is Lancet Haematol 4;2017:e127–36.
a measure of TP53 mutation burden and Ornstein MC, Mukherjee S, and Sekeres MA.
adverse clinical outcome in myelodysplastic More is better: Combination therapies for
syndromes and secondary acute myeloid leu- myelodysplastic syndromes. Best Pract Res
kemia. Haematologica 101;2016:e320–3. Clin Hematol 28;2015:22–31.
192  Precision medicine in myelodysplastic syndromes

Otrock ZK, Chamseddine N, Salem ZM et al. A Sallman DA, Cluzeau T, Basiorka AA, and List
nationwide non-interventional epidemiological A. Unraveling the pathogenesis of MDS:
data registry on myelodysplastic syndromes in The NLRP3 inflammasome and pyroptosis
Lebanon. Am J Blood Res 5;2015:86–90. drive the MDS phenotype. Front Oncol
Owen C, Barnett M, and Fitzgibbon J. Familial 6;2016:151.
myelodysplasia and acute myeloid leukaemia— Santini V. Treatment of low-risk myelodysplastic
A review. Br J Haematol 140;2008:123–32. syndromes. Hematology Am Soc Hematol
Papaemmanuil E, Gerstung M, Malcovati L et al. Educ Program 2016;2016:462–9.
Clinical and biological implications of driver Santini V, Almeida A, Giagounidis A et al.
mutations in myelodysplastic syndromes. Randomized phase III study of lenalido-
Blood 122;2013:3616–27. mide versus placebo in RBC transfusion-­
Pellagatti A, and Boultwood J. The molecular dependent patients with lower-risk
pathogenesis of the myelodysplastic syn- non-del(5q) myelodysplastic syndromes
dromes. Eur J Haematol 95;2015:3–15. and ineligible for or refractory to erythro-
Phekoo KJ, Richards MA, Mϕller H, and Schey poiesis-stimulating agents. J Clin Oncol
SA. The incidence and outcome of myeloid 34;2016:2988–96.
malignancies in 2,112 adult patients in south- Sardnal V, Rouquette A, Kaltenbach S et al.
east England. Haematologica 91;2006:1400–4. A G polymorphism in the CRBN gene acts as
Platzbecker U, and Fenaux P. Personalized a biomarker of response to lenalidomide in
medicine in myelodysplastic syndromes: low/int-1 risk MDS without del(5q). Leukemia
Wishful thinking or already clinical reality? 27;2013:1610–3.
Haematologica 100;2015:568–71. Schneider RK, Ademá V, Heckl D et al. Role of
Pleyer L, Valent P, and Greil R. Mesenchymal casein kinase 1A1 in the biology and tar-
stem and progenitor cells in normal and dys- geted therapy of del(5q) MDS. Cancer Cell
plastic hematopoiesis-masters of survival and 26;2014:509–20.
clonality? Int J Mol Sci 17;2016:1009. Schneider RK, Schenone M, Ferreira MV
Rankin EB, Narla A, Park JK, Lin S, and Sakamoto et al. Rps14 haploinsufficiency causes a
KM. Biology of the bone marrow microenvi- block in erythroid differentiation medi-
ronment and myelodysplastic syndromes. Mol ated by S100A8 and S100A9. Nat Med
Genet Metab 116;2015:24–8. 22;2016:288–97.
Raza A, Reeves JA, Feldman EJ et al. Phase 2 Scott BL, Pasquini MC, Logan BR et al.
study of lenalidomide in transfusion-depen- Myeloablative versus reduced-intensity
dent, low-risk and intermediate-1 risk myelo- hematopoietic cell transplantation for acute
dysplastic syndromes with karyotypes other myeloid leukemia and myelodysplastic syn-
than deletion 5q. Blood 111;2008:86–93. dromes. J Clin Oncol 35;2017:1154–61.
Rodger EJ, and Morison IM. Myelodysplastic syn- Sekeres MA. Are we nearer to curing patients
drome in New Zealand and Australia. Intern with MDS? Best Pract Res Clin Haematol
Med J 42;2012:1235–42. 23;2010:481–7.
Rollison DE, Howlader N, Smith MT, Strom SS, Senent L, Arenillas L, Luño E, Ruiz JC, Sanz
Merritt WD, Ries LA, Edwards BK, and List AF. G, and Florensa L. Reproducibility of the
Epidemiology of myelodysplastic syndromes World Health Organization 2008 criteria for
and chronic myeloproliferative disorders in myelodysplastic syndrromes. Haematologica
the United States, 2001-2004, using data 98;2013:568–75.
from the NAACCR and SEER programs. Blood Shastri A, Will B, Steidl U, and Verma A. Stem and
112;2008:45–52. progenitor cell alterations in myelodysplastic
Saft L, Karimi M, Ghaderi M et al. p53 protein syndromes. Blood 129;2017:1586–94.
expression independently predicts outcome Sheikine Y, Kuo FC, and Lindeman NI. Clinical
in patients with lower-risk myelodysplastic and technical aspects of genomic diagnos-
syndromes with del(5q). Haematologica tics for precision oncology. J Clin Oncol
99;2014:1041–9. 35;2017:929–33.
 References 193

Sibon D, Cannas G, Baracco P et al. Tiu RV, Visconte V, Traina F, Schwandt A, and
Lenalidomide in lower-risk myelodysplas- Maciejewski JP. Updates in cytogenetics and
tic syndromes with karyotypes other than molecular markers in MDS. Curr Hematol
deletion 5q and refractory to erythropoi- Malig Rep 6;2011:126–35.
esis-stimulating agents. Br J Haematol Toma A, Kosmider O, Chevret S et al.
156;2012:619–25. Lenalidomide with or without erythropoietin
Smith AE, Kulasekararaj AG, Jiang J et al. in transfusion-dependent erythropoiesis-stim-
CSNK1A1 mutations and isolated del(5q) ulating agent-refractory lower-risk MDS with-
abnormality in myelodysplastic syndrome: out 5q deletion. Leukemia 30;2016:897–905.
A retrospective mutational analysis. Lancet Vardiman JW, Thiele J, Arber DA et al. The 2008
Haematol 2;2015:e212–221. revision of the World Health Organization
Stahl M, and Zeidan AM. Hypomethylating (WHO) classification of myeloid neoplasms
agents in combination with histone deacety- and acute myeloid leukemia: Rationale and
lase inhibitors in higher risk myelodysplastic important changes. Blood 114;2009:937–51.
syndromes: Is there a light at the end of the Venner CP, Woltosz JW, Nevill TJ et al.
tunnel. Cancer 123;2017a:911–4. Correlation of clinical response and
Stahl M, Zeidan AM. Lenalidomide use in response duration with miR-145 induction by
myelodysplastic syndromes: Insights into the lenalidomide in CD34+ cells from patients
biologic mechanisms and clinical applications. with del(5q) myelodysplastic syndrome.
Cancer 2017b. doi: 10.1002/cncr.30585. Haematologica 98;2013:409–13.
Starczynowski DT, Kuchenbauer F, Argiropoulos Venter JC, Adams MD, Myers EW et al. The
B et al. Identification of miR-145 and miR-146a sequence of the human genome. Science
as mediators of the 5q- syndrome phenotype. 291;2001:1304–51.
Nat Med 16;2010:49–58. Vlachos A, Farrar JE, Atsidaftos E et al.
Steensma DP, Bejar R, Jaiswal S, Lindsley RC, Diminutive somatic deletions in the 5q region
Sekeres MA, Hasserjian RP, and Ebert BL. lead to a phenotype atypical of classical 5q-
Clonal hematopoiesis of indeterminate poten- syndrome. Blood 122;2013:2487–90.
tial and its distinction from myelodysplastic Walter MJ, Shen D, Shao J et al. Clonal
syndromes. Blood 126;2015:9–16. diversity of recurrently mutated genes
Suragani RN, Cadena SM, Cawley SM et al. in myelodysplastic syndromes. Leukemia
Transforming growth factor-β superfamily 27;2013:1275–82.
ligand trap ACE-536 corrects anemia by pro- Wei S, Chen X, McGraw K, Zhang L, Komrokji
moting late-stage erythropoiesis. Nat Med R, Clark J, and Caceres G. Lenalidomide
20;2014:408–14. promotes p53 degradation by inhibiting
Takahashi K, Patel K, Bueso-Ramos C et al. MDM2 auto-ubiquitination in myelodysplas-
Clinical implications of TP53 mutations in tic syndrome with chromosome 5q deletion.
myelodysplastic syndromes treated with Oncogene 32;2013:1110–20.
hypomethylating agents. Oncotarget Wei S, Chen X, Rocha K et al. A critical role
7;2016:1472–87. for phosphatase haplodeficiency in the
Takahashi K, Wang F, Kantarjian H, Doss selective suppression of deletion 5q MDS
D, Khanna K, Thompson E, and Zhao L. by lenalidomide. Proc Natl Acad Sci USA
Prelekaemic clonal haemopoiesis and risk of 106;2009:12974–9.
therapy-related neoplasms: A case-control West AH, Godley LA, and Churpek JE. Familial
study. Lancet Oncol 18;2017:100–11. myelodysplastic syndrome/acute myeloid
Tefferi A, and Garcia-Manero G. CME leukemia syndromes: A review and utility for
Information: Annual clinical updates 2015: translational investigations. Annals NY Acad
Myelodysplastic syndromes. Am J Hematol Sci 1310;2014:111–8.
90;2015:831–41. Wlodarski MW, Hirabayashi S, Pastor V et al.
Tefferi A, and Vardiman JW. Myelodysplastic syn- Prevalence, clinical characteristics, and
dromes. N Engl J Med 361;2009:1872–85. prognosis of GATA2-related myelodysplastic
194  Precision medicine in myelodysplastic syndromes

syndromes in children and adolescents. Blood Zeidan AM, Sekeres MA, Garcia-Manero G et al.
127;2016:1387–97. Comparison of risk stratification tools in
Wu L, Li X, Xu F, Zhang Z, Chang C, and He predicting outcomes of patients with higher-
Q. Low RPS14 expression in MDS without risk myelodysplastic syndromes treated with
5q– aberration confers higher apoptosis azanucleotides. Leukemia 30;2016:649–57.
rate of nucleated erythrocytes and predicts Zeidan AM, Sekeres MA, Wang XF et al.
prolonged survival and possible response to Comparing the prognostic value of risk
lenalidomide in lower risk non-5q– patients. stratifying models for patients with lower-risk
Eur J Haematol 90;2013:486–93. myelodysplastic syndromes: Is one model
Zahr AA, Ramirez CB, Wozney J, Prebet T, and better? Am J Hematol 90;2015:1036–40.
Zeidan AM. New insights into the pathogene- Zhang L, McGraw KL, Sallman DA, and List AF.
sis of MDS and the rational therapeutic oppor- The role of p53 in myelodysplastic syndromes
tunities. Expert Rev Hematol 9;2016:377–88. and acute myeloid leukemia: Molecular
Zeidan AM, Al Ali N, Barnard J et al. Comparison aspects and clinical implications. Leuk
of clinical outcomes and prognostic utility Lymphoma 58;2017:1777–90.
of risk stratification tools in patients with Zou J, Zhou Z, Wan L, Tong Y, Qin Y, Wang C,
therapy-related vs de novo myelodysplastic and Zhou K. Targeting the sonic Hedgehog-
syndromes: Report on behalf of the MDS Gli1 pathway as a potential new therapeutic
Clinical Research Consortium. Leukemia 2017. strategy for myelodysplastic syndromes. PLOS
doi: 10.1038/leu. 2017.33. One 10;2015:e0136843.
 10
Precision medicine in acute myeloid
leukemia

OTA FUCHS

Introduction 195 Current therapy of AML and experimental

Methods for the diagnosis and treatment of therapies 200
acute myeloid leukemia (AML) 196 Therapeutic targeting of AML stem cells 206
Cytogenetic analysis 196 Familial predisposition to AML 206
Morphology 198 First Models For Predicting Outcome For
Flow cytometry 198 Individual Aml Patients 206
Classification of AML patients 198 Conclusion 207
Prognostic stratification significance Acknowledgments 207
of gene mutations in intermediate- Conflict of interest 208
risk AML 200 References 208

INTRODUCTION those over 65 years. Prognosis in the older patients

who account for the majority of new cases remains
Acute myeloid leukemia (AML) is a genetically poor (Shah et  al., 2013). Even with current treat-
heterogeneous, malignant clonal disease and the ments, as much as 70% of patients 65 years or older
most common acute leukemia in adults (about 80% will die of their disease within 1 year of diagno-
of cases in this group) (Yamamoto and Goodman, sis (Meyers et  al., 2013). AML is characterized by
2008). GLOBOCAN estimates the worldwide total clonal expansion of abnormal hematopoietic pro-
leukemia incidence of AML for 2012 to be 351,965 genitor cells and impaired production of normal
with an age-standardized rate (ASR) per 100,000 of blood cells (Estey and Dohner, 2006; Jan et al., 2017;
4.7, a 5-year prevalence of 1.5%, and a M (male):F Prada-Arismendy et al., 2017). A high relapse rate
(female) ratio of about 1.4 (GLOBOCAN 2014). for patients with AML is still a major barrier to the
The incidence of AML ranges from 3 to 5 cases per long-term survival of these patients (Rowe, 2016).
100,000 population in the United States. In 2015, The first reports of an unknown condition with
20,830 new cases of AML were expected to occur a “milky blood” were from Scottish physician Peter
in the United States, including 12,370 in males Cullen (1811). French physician Alfred Velpeau
and 8,100 in females, and over 10,000 patients died defined the leukemia-associated symptoms (1825)
from this disease (De Kouchkovsky and Abdul- followed by French physician Alfred Donné who
Hay, 2016; Kansal, 2016; Siegel et  al., 2015). The detected amaturation arrest of the white blood
incidence of AML increases with age, from about cells (1844). British physician John Hughes Bennett
1.3 per 100,000 population in patients less than 65 named the disease “leucocythemia,” based on the
years old, to 12.2 cases per 100,000 population in microscopic accumulation of purulent leucocytes

195
196  Precision medicine in acute myeloid leukemia

(1845). The same year, a famous German physician Table 10.1  Cytogenetic abnormalities used in
Rudolf Virchow defined a reversed white and red the WHO classification (2008) of AML
blood cell balance. He introduced the disease as
Cytogenetic abnormalities used to define
“leukemia” in 1847. British physician Henry Fuller
entities within the WHO category of AML
performed and described the first microscopic
with recurrent genetic abnormalities
diagnosis of a leukemia patient during life (1846)
t(8;21)(q22;q22); RUNX1-RUNX1T1
(Kampen, 2012; Thomas, 2013).
The purpose of this review is to describe some inv(16)(p13.1q22) or t(16;16)(p13;q22);
of the recent major advances that occurred due to CBFB-MYH11
the impact of genomics, mainly through the new t(15;17)(q22;q12-21); PML-RARA
genomic technologies, in the diagnostic classifi- t(9;11)(p22;q23); MLLT3-MLL
cation, risk stratification, and treatment of AML. t(6;9)(p23;q34); DEK-NUP214
New genomic technologies and mainly advances inv(3)(q21q26.2) or t(3;3)(q21;q26.2); RPN1-EVI1
in bioinformatics helped to develop next-gener- t(1;22)(p13;q13); RBM15-MKL1
ation sequencing (NGS), whole genome sequenc- Cytogenetic abnormalities sufficient to diagnose
ing, whole exome sequencing, and targeted deep the WHO category of AML with
sequencing of specific mutations (Eisfeld et  al., myelodysplasia-related changes
2017; Kohlmann et al., 2014; Lin et al., 2017; Nazha Complex karyotype
et al., 2016; Wang et al., 2016; White and DiPersio, (defined as 3 or more unrelated abnormalities,
2014). none of which can be a translocation or
inversion associated with AML with recurrent
METHODS FOR THE DIAGNOSIS genetic abnormalities)
AND TREATMENT OF ACUTE Unbalanced abnormalities
MYELOID LEUKEMIA (AML)   −7 or del(7q)
  −5 or del(5q)
Cytogenetic analysis   i(17q), an isochromosome for long arm of
chromosome 17 or t(17p)
Cytogenetic analysis is an important and manda-
tory component of AML diagnosis. Cytogenetics   −13 or del(13q)
remains the most important disease-related prog-  del(11q)
nostic factor and powerful independent prognostic   del(12p) or t(12p)
indicator in AML (Ferrara et al., 2008; Foran, 2010;  del(9q)
Grimwade and Mrózek, 2011; Röllig et  al., 2011).   idic(X)(q13), isodicentic X chromosome
Until now, more than 100 balanced chromosomal Balanced abnormalities
rearrangements (translocations, insertions, and  t(11;16)(q23;p13.3)
inversions) have been identified and cloned. These  t(3;21)(q26.2;q22.1)
chromosomal rearrangements are critical initiat-  t(1;3)(p36.3;q21.1)
ing events in the pathogenesis of AML. Karyotype  t(2;11)(p21;q23)
analysis identifies biologically distinct subsets of  t(5;12)(q33;p12)
AML that differ in their response to therapy and
 t(5;7)(q33;q11.2)
treatment outcome (Mrózek et  al., 2004; Smith
 t(5;17)(q33;p13)
et al., 2011).
 t(5;10)(q33;q21)
Cytogenetic abnormalities used in the WHO
classification (Vardiman et al., 2009) are shown in  t(3;5)(q25;q34)
Table 10.1. Approximately 10% of adult AML and
20% of childhood AML are classified as having Metzeler and Bloomfield, 2017; Mrózek et al., 2008;
core binding factor (CBF) leukemia with balanced Sangle and Perkins, 2011; Utsun and Marcucci,
chromosomal rearrangements that disrupt gene 2015; Yamagata et  al., 2005). Patients with two
RUNX1 (also known as CBFA2 or AML1), which specific, clonal, recurring cytogenetic abnormali-
plays a critical role in hematopoiesis and leukemo- ties t(8;21)(q22;q22), inv(16)(p13.1q22) or t(16;16)
genesis (Lam and Zhang, 2012; Marcucci, 2006; (p13.1q22) are called CBF AML. Compared to
 Methods for the diagnosis and treatment of acute myeloid leukemia (AML)  197

Table 10.2  Variations in cytogenetic risk in patients with acute myeloid leukemia. (United Kingdom
Medical Research Council cytogenetic group assignment 2010)

Risk group assignment Cytogenetic abnormality Comments

Favorable t(15;17)(q22;q12-21) Irrespective of additional
cytogenetic abnormalities
t(8;21)(q22;q22)
inv(16)(p13.1q22) or t(16;16)(p13.1q22)
Intermediate Other noncomplex entities,
not classified as favorable
or adverse
Normal karyotype
Adverse abn(3q) [excluding t(3;5)(q21–25;q31–35)],
inv(3)(q21q26) or t(16;16)(p13;q22),
add(5q), del(5q), –5,
–7, add(7q), del(7q),
t(6;11)(q27;q23),
t(10;11)(p11–13;q23),
other t(11;q23) [excluding t(9;11)(p21–22;q23)
and t(11;19)(q23;p13)],
t(9;22)(q34;q11),
–17, abn(17p)
Complex (>4 unrelated abnormalities)
Note: Table is based on multivariable analysis conducted in 5876 adults (16–59 years) treated in the United Kingdom
Medical Research Council AML10, 12, and 15 trials (Grimwade et al., 2010).

other cytogenetic AML subgroups, CBF AML karyotype (monosomies of chromosomes 5 and/
is considered a more favorable subset of AML or 7(-5/-7) (Brands-Nijenhuis et  al., 2016; Kayser
(Table 10.2). CBF AML results in the formation of et  al., 2012; Perrot et  al., 2011), abnormalities
hybrid fusion genes called RUNX1-RUNX1T1 (also of 3q/abn(3q) (Hinai and Valk, 2016; Lugthart
known as AML1-ETO) and CBFB-MYH11, which et  al., 2010), or a complex karyotype often asso-
can be quantified in patients before, during, and ciated with TP53 alterations (Rücker et al., 2012).
after the therapy, including stem cell transplanta- Complex karyotype anomalies (clinically defined
tion (Yin et al., 2012). Favorable karyotype patients as having an adverse outcome) have at least three
have a good prognosis with complete remission to five anomalies and do not contain the balanced
rates exceeding 90%, a five-year survival of at least chromosomal translocations/inversions defining
65%, and relapse rates too low and salvage rates too a favorable prognosis, such as t(15;17) or t(8;21)
high to benefit from routine use of allograft in first (Blum et  al., 2017). The monosomal karyotype
complete remission (Smith et al., 2011). entity is defined as two or more autosomal mono-
Patients with CBF AML and mutations in somies or combination of one monosomy with
the KIT gene (exon 17) have a higher relapse structural abnormalities. These patients are usu-
risk (Ayatollahi et  al., 2017; Paschka et  al., 2006; ally older and had MDS or were treated by che-
Shimada et al., 2006). The mechanism underlying motherapy or radiation. They have therapy-related
KIT gene mutations that adversely affects the prog- AML (tAML) (Heuser, 2016). Complete remission
nosis involves phosphorylation of the KIT receptor rates are around 60% and a five-year survival about
after physiologic binding of KIT ligand, which acti- 10%. Due to this very poor prognosis with current
vates downstream pathways supporting cell prolif- therapies, an allograft in first complete remis-
eration and survival (Lennartsson et al., 2005). sion or an experimental treatment approach may
About 10%–20% of AML patients have be justified (Fang et  al., 2011; Paun and Lazarus,
adverse cytogenetics, which includes monosomal 2012). However, patients with single monosomy,
198  Precision medicine in acute myeloid leukemia

more than 10% of cells with normal metaphase, the and moderate cytoplasm, generally with abundant
absence of del(17p), and achievement of complex cytoplasmic granules and Auer rods that may be in
remission after induction therapy were prognostic bundles (“faggot cells”).
factors for better overall survival (Jang et al., 2015). It is now necessary to use a combination of
The mixed lineage leukemia gene (MLL, also morphology with other techniques as immuno-
known as ALL-1 or HRX), located on chromo- cytochemistry or flow cytometry to confirm a
some 11q23, encodes a histone methyltransferase diagnosis. These latter techniques are essential
and is frequently rearranged in AML. Recurrent for the diagnosis of the newer described entities
translocations of MLL are generally considered of AML with minimal differentiation (FAB-M0)
to confer a poor prognosis (Meyer et  al., 2009; and acute megakaryoblastic leukemia (FAB-M7).
Pigneux et  al., 2015). Fusion proteins with more Cytogenetics and fluorescent in situ hybridization
than 60 partners were characterized and these techniques are also very important for diagnosis as
partners also have a significant effect on disease in FAB-M3 (promyelocytic leukemia) but also for
outcome (Coenen et al., 2011). Other entities that detecting those myeloid leukemias that are asso-
have been associated with poor prognosis are ciated with a favorable or unfavorable response
t(6;9)(p23;q34), t(9;22)(q34;q11), and 17p deletions (Bacher et al., 2005; Krause 2000).
(Smith et al., 2011).
AML with isolated trisomies for chromosomes Flow cytometry
4, 8, 11, and 21 should, with exception of trisomy
for chromosome 13, be classified as intermediate Flow cytometric immunophenotyping is an accu-
risk (Lazarevic et al., 2017). rate and objective method for quantitative and
Since cytogenetics provides the framework for qualitative evaluation of hematopoietic cells.
current risk stratification schemes in AML with Compared with AML-not otherwise specified
a major role in determining whether patients are (AML-NOS), granulocytic cells in AML with
candidates for allogenic transplant in first remis- myelodysplasia-related changes (AML-MRC)
sion, it is important to have clearly defined risk had higher CD33 expression but lower CD45,
groups not only for direction of patient manage- CD11b, and CD15. Monocytes in AML-MRC
ment but also for the comparison of data between had lower expression of CD14, CD56, and CD45.
different clinical results (Smith et al., 2011). Morphologic dysplasia was associated with sig-
nificantly lower granulocytic forward scatter,
Morphology side scatter, and CD10 but higher CD33 expres-
sion. Our results suggest that the workup of AML
The categories of AML with t(8;21)(q22:q22), cases should include flow cytometric assessment of
inv(16)(p13.1:q22) or t(16:16)(p13.1:q22), and granulocytes and monocytes. This analysis can aid
t(15:17)(q22:q12) are considered AML without a morphologic impression of multilineage dyspla-
regard to blast count. Identification of 20% or more sia in distinguishing AML-MRC from AML-NOS,
blasts in the blood or bone marrow is required in especially in cases with limited maturing myeloid
most categories of AML. The morphology of the cells (Weinberg et al., 2017a).
blasts often correlates with cytogenetic abnor-
malities. For example, AML with t(8:21) often has Classification of AML patients
large blasts with numerous azurophilic granules,
which may occasionally be very large (pseudo- To assist with patient diagnosis and management,
Chediak-Higashi type), and paranuclear clearing classifications, such as the French-American-
or hofs. Long, thin Auer rods are frequently seen British (FAB) (Bennett et  al., 1976, 1985) and the
and eosinophils may be increased. AML with World Health Organization (WHO) classification
inv(16) has blasts with myelomonocytic features (Arber et al., 2016; Döhner et al., 2010; Vardiman
and increased number of eosinophils at all stages et  al., 2009; Table 10.3), were developed to define
of maturation (Weinberg et al., 2017b). clinically relevant disease subtypes. FAB classi-
AML with t(15;17), also known as acute promy- fication is still widely used in clinical setting that
elocytic leukemia (APL), is characterized by abnor- groups AML into eight subgroups (M0-M7) based
mal promyelocytes with bilobed or indented nuclei on its degree of differentiation and morphology.
 Methods for the diagnosis and treatment of acute myeloid leukemia (AML)  199

Table 10.3  WHO classification (2016) of acute gene PML on chromosome 15 with retinoic acid
myeloid leukemia and related neoplasms receptor α (RARA) located on chromosome 17.
APL with fusion gene PML-RARA is treated with
Acute myeloid leukemia with recurrent genetic
molecularly targeted therapy such as the differ-
abnormalities
entiation-inducing agent all-trans retinoic acid
  AML with t(8;21)(q22;q22); RUNX1-RUNX1T1
(ATRA) and arsenic trioxide (ATO), especially in
  AML with inv(16)(p13.1q22) or t(16;16) cases where ATRA plus anthracycline-based che-
(p13;q22); CBFB-MYH11 motherapy cannot be used (Baljevic et  al., 2011;
  APL with t(15;17)(q22;q12); PML-RARA Lo-Coco et al., 2008).
  AML with t(9;11)(p22;q23); MLLT3-MLL A new provisional entity AML with mutated
  AML with t(6;9)(p23;q34); DEK-NUP214 RUNX1 (excluding cases with myelodysplasia-
  AML with inv(3)(q21q26.2) or t(3;3) related changes) was added (Gadzik et  al., 2016;
(q21;q26.2); GATA2, MECOM Haferlach et  al., 2016). Myeloid neoplasms with
  AML (megakaryoblastic) with t(1;22)(p13;q13); germ line predisposition is also a new category that
RBM15-MKL1 was added to the WHO classification (Babushok
  Provisional entity: AML with BCR-ABL1 et al., 2016; Churpek and Godley, 2016; Table 10.4).
  AML with mutated NPM1 The molecular basis of AML with inv(3)(q21q26.2)
  AML with biallelic mutations of CEBPA or t(3;3)(q21;q26.2) was revisited showing that
  Provisional entity: AML with mutated RUNX1 repositioning of a GATA2 enhancer element leads
Acute myeloid leukemia with myelodysplasia- to overexpression of the MECOM (EVI1) gene and
related changes to haploinsufficiency of GATA2 (Gröschel et  al.,
Therapy-related myeloid neoplasma
2014).
Acute myeloid leukemia, not otherwise
specified
  AML with minimal differentiation Table 10.4  WHO classification of AML with
  AML without maturation germ-line predisposition and guide for molecular
  AML with maturation genetic diagnostics
  Acute myelomonocytic leukemia Classification
  Acute monoblastic/monocytic leukemia Myeloid neoplasms with germ-line
   Pure erythroid leukemia predisposition without a preexisting disorder
   Acute megakaryoblastic leukemia or organ dysfunction
  Acute basophilic leukemia   AML with germ-line CEBPA mutation
  Acute panmyelosis with myelofibrosis   Myeloid neoplasms with germ line DDX41
Myeloid sarcoma mutation
Myeloid proliferations related to Down Myeloid neoplasms with germ-line
syndrome predisposition and preexisting platelet
  Transient abnormal myelopoiesis (TAM) disorders
  Myeloid leukemia associated with Down   Myeloid neoplasms with germ line RUNX1
syndrome mutation
Blastic plasmacytoid dendritic cell neoplasm   Myeloid neoplasms with germ line ANKRD26
Acute leukemias of ambiguous lineage mutation
  Myeloid neoplasms with germ line ETV6
mutation
Treatment for all subtypes of AML, except the
Myeloid neoplasms with germ-line
M3 subtype, acute promyelocytic leukemia (APL),
predisposition and other organ dysfunction
involves combination chemotherapy and a possible
  Myeloid neoplasms with germ line GATA2
hematopoietic stem cell transplantation as a part
mutation
of consolidation therapy (Burnett, 2012; Paun and
  Myeloid neoplasms associated with bone
Lazarus, 2012). In APL, the balanced t(15;17) trans-
marrow failure syndromes
location rearranges the promyelocytic leukemia
200  Precision medicine in acute myeloid leukemia

Prognostic stratification significance Valk, 2013). Functional categories of genes com-

of gene mutations in intermediate- monly affected in AML are presented in Table 10.5.
Genetic abnormalities were correlated with clinical
risk AML
characteristics and outcome of AML patients. The
Approximately 45% of adult patients with AML European Leukemia Net (ELN) published a recom-
have normal karyotype (cytogenetically normal mendation where a three-group classification sys-
AML; CN-AML patients) and are usually classified tem (favorable, intermediate, and adverse) was used
as an intermediate risk group. Cytogenetic analy- for ELN genetic risk stratification (Bullinger et al.,
sis is uninformative for these patients because they 2017; Döhner et al., 2017; Table 10.6).
have vastly different clinical outcomes within this
intermediate risk group. Analysis of specific genetic Current therapy of AML
abnormalities, not detectable by cytogenetics, and and experimental therapies
quantification of expression of specific genes with
prognostic significance are valuable tools for more Current therapy has not changed substantially in
accurate risk stratification and clinical outcome recent years. For decades, the standard therapy
prediction of these CN-AML patients (Al-Issa and has been initial treatment with cytarabine and
Nazha, 2016; Bullinger et  al., 2017; Döhner et  al., an anthacycline, the so-called 7 + 3 induction
2017; Papaemmanuil et  al., 2016; Sanders and regimen (a seven-day continuous intravenous

Table 10.5  Functional categories of genes commonly affected in AML

Functional category Selected gene members Role in leukemogenesis

Signaling genes Kinases (e.g., FLT3, KIT), Activated signaling confers a proliferative
phosphatases (e.g., PTPN1), advantage through RAS/RAF, JAK/STAT,
or RAS family members and PI3K/AKT signaling pathways
(e.g., KRAS, NRAS)
DNA methylation- DNMT3A, TET2, IDH1, IDH2 Deregulated DNA methylation and
associated genes transcriptional deregulation of leukemia-
relevant genes
Myeloid transcription TF fusions [t(8;21), Aberrant TF function results in
factor (TF) gene inv(16)/t(16;16)] transcriptional deregulation and impaired
fusion mutations hematopoietic differentiation
TF mutations (RUNX1, CEBPA) Aberrant TF function results in
transcriptional deregulation and impaired
hematopoietic differentiation
Chromatin-modifying Mutations (e.g., ASXL1, EZH2) Deregulation of chromatin modification
genes or KMT2A fusions (e.g., methylation of histones H3 and
H2A) as well as KMT2A fusion-driven
impairment of methyltransferases lead to
transcriptional deregulation
Nucleophosmin NPM1 Mutations of NPM1 result in aberrant
(NPM1) gene cytoplasmic localization of NPM1 and
NPM1-interacting proteins
Tumor-suppressor TP53, WT1, PHF6 Mutations lead to transcriptional
genes deregulation
Spliceosome-complex SRSF2, SF3B1, U2AF1, ZRSR2 Mutations lead to impaired spliceosome
genes function and deregulated RNA processing
Cohesin-complex STAG2, RAD21 Mutations may lead to impaired
genes chromosome segregation and impact
transcriptional regulation
 Methods for the diagnosis and treatment of acute myeloid leukemia (AML)  201

Table 10.6  2017 ELN risk stratification by et al., 2006) and led to the development of a liposo-
genetics mal formulation to these drugs in this ideal molar
ratio within 100 nanometer liposomes (Tardi et al.,
Risk category Genetic lesion
2009). This product named CPX-351 was well toler-
Favorable t(8,21)(q22;q22.1); ated and had better efficacy and improved response
RUNX1-RUNX1T1 rates compared to standard administration but with-
inv(16)(p13.1q22) or t(16;16) out an improvement in event-free survival or overall
(p13.1;q22); CBFB-MYH11 survival. However, complete remission was nearly
Mutated NPM1 without FLT3-ITD doubled in patients with secondary AML (Sallman
or with FLT3-ITDlow and Lancet, 2017). These encouraging results led to
Biallelic mutated CEBPA phase 3 study (NCT01696084) in high-risk AML
Intermediate Mutated NPM1 and FLT3-ITDhigh patients (patients with secondary AML, therapy-
Wild type NPM1 without related AML, but also AML with WHO-defined
FLT3-ITD or with FLT3-ITDlow MDS cytogenetic abnormalities). The median of
(without adverse-risk gene overall survival was 9.56 months in patients treated
mutations) with CPX-351 compared to 5.95 months with 7 + 3.
t(9;11)(p21.3;q23.3); These data supported CPX-351 as a new standard
MLLT–KMT2A therapy for older patients with secondary AML.
Cytogenetic abnormalities not A randomized trial found that fludarabine and
classified as favorable high dose cytarabine plus granulocyte colony-stim-
Adverse t(6;9)(p23;q34.1); DEK-NUP214 ulating factor (G-CSF; FLAG) plus idarubicin not
t(v;11q23.3); KMT2A rearranged only produced a lower relapse rate than daunorubi-
t(9;22)(q34.1; q11.2); BCR-ABL1 cin-cytarabine with or without etoposide, but was
inv(3)(q21q26.2) or t(3;3) associated with more deaths in remission resulting
(q21;q26.2); GATA2, MECOM in similar overall survival (Burnett et al., 2013).
(EVI1) New therapies are needed that can target and
–5 or del(5q); –7; –17/abn(17p) eradicate resistant subclones early in the disease
Complex karyotype, monosomal course. An example is the multikinase inhibitor
karyotype midostaurin, which when added to 7 + 3 induc-
Wild-type NPM1 and FLT3-ITDhigh tion has been found to benefit in patients with
Mutated RUNX1 FLT3-mutated AML (Gallogly and Lazarus, 2016).
Mutated ASXL1 Sorafenib (a tyrosine kinase inhibitor), crenolanib
Mutated TP53 (a FLT3 inhibitor with activity against FLT3-D835
TKD mutation), gilteritinib (ASP-2215, a potent
inhibitor of both FLT3-ITD and FLT3-TKD muta-
cytarabine infusion and 3 daily doses of dauno- tions), and a second-generation FLT3-ITD inhibi-
rubicin), followed by higher doses of cytarabine as tor quizaritinib (formerly known as AC220) were
consolidation. This approach is curative in about also used in combination with chemotherapy (De
40% of younger patients (age range between 18 and Kouchkovsky and Abdul-Hay, 2016; Grunwald and
60 years) overall, with patients in some of the more Levis, 2015). Terminal myeloid differentiation in
favorable subgroups enjoying cure rates of approx- vivo is induced by FLT3 inhibition in FLT3/ITD
imately 60%. The outcomes are worst in older AML AML (Sexauer et al., 2012). Unfortunately, owing to
patients, who represent the majority of individuals the resistance, the responses to FLT3 inhibitors are
with AML and in patients with AML after a prior not durable and last just 3–6 months. For patients
diagnosis of MDS (secondary AML). harboring FLT3-ITD mutations, allogeneic hema-
As a combination of cytarabine and daunorubicin topoietic stem cell transplantation (alloHSCT) is
represented the long-standing but inadequate stan- preferable, and autologous HSCT is used mainly for
dard of care for induction chemotherapy, in vitro patients with favorable or intermediate-risk cyto-
studies analyzed the optimal synergistic ratio of both genetics in first remission (Tamamyan et al., 2017).
compounds. These studies identified a 5:1 molar In AML, targeted inhibitors of both IDH1
ratio of cytarabine:daunorubicin to be ideal (Mayer and IDH2 have been characterized in preclinical
202  Precision medicine in acute myeloid leukemia

models (Chaturvedi et  al., 2013; Losman et  al., HSCT for poor-risk but not for good-risk AML.
2013; Wang et al., 2013). The conversion of gluta- Which patients should be offered allogeneic HSCT
mine to the α-ketoglutarate (α-KG) in cells occurs and whether this should be in first complete remis-
in two steps. Glutamine is first converted to glu- sion or after relapse is a major therapeutic dilemma
tamate and ammonia by glutaminase. Glutamate (Gale et al., 2014). The quantification of measurable
is subsequently converted to α-KG by either gluta- minimal residual disease (MRD) in first complete
mate pyruvate transaminase or glutamate oxaloac- remission is very important (Buccisano and Walter,
etate transaminase (Fathi et al., 2015). Glutamine 2017). A recent single-center analysis showed that
is the main source of α-KG in IDH-mutant leuke- patients with positive MRD at the time of transplan-
mic blasts. Therefore, the depletion of glutamine or tation have poor outcomes even if myeloablative
interruption of its metabolism are both selectively conditioning is able to reduce residual amounts of
harmful for AML cells with mutated IDH (Emadi leukemia cells (Zhou et al., 2016). Allogeneic HSCT,
et al., 2014). The promising results of these preclin- autologous transplantation, and consolidation che-
ical studies warranted early-phase clinical trials motherapy are considered of equivalent benefit for
(Dang et al., 2017; Fathi et al., 2015; Medeiros et al., intermediate-risk AML.
2017; Table 10.7). In patients not eligible for intensive chemother-
Both FLT3 and IDH inhibitors can be used as apy treatment, hypomethylating agents (HMAs)
a bridge to transplant. Current AML management have become a frequent alternative treatment
relies largely on intensive chemotherapy and allo- choice. Dysregulated epigenetic mechanisms play
geneic hematopoietic stem cell transplantation an important role in the pathogenesis of AML. The
(HSCT), at least in younger patients who can toler- term epigenetics refers to changes in gene expression
ate such intensive treatments. Safer allogeneic HSCT that are not caused by changes in DNA sequence.
procedures allow a larger proportion of patients to Cytosine methylation in DNA, modifications of
achieve durable remission. Improved identification histone proteins, or RNA-associated gene silencing
of patients at relatively low risk of relapse should play a role in epigenetic mechanisms (Wouters and
limit their undue exposure to the risk of HSCT in Delwel, 2016). Epigenetic dysregulation in AML is
first remission. The optimal treatment of AML in widespread and cannot be explained by recurrent
first complete remission is uncertain (Buccisano and somatic mutations alone. Two HMAs currently
Walter, 2017). Current consensus, based on cytoge- approved for clinical use in AML are 5-azacyti-
netic risk, recommends myeloablative allogeneic dine (azacitidine) and 5-aza-2′-deoxycytidine

Table 10.7  Clinical trials targeting IDH mutations in AML

Inhibitor Target Phase Identifier Patient population

AG-120 IDH1 1 NCT02074839 >18 years of age with IDH1 gene-mutated,
advanced hematologic malignancy including
relapsed and/or primary refractory AML
  >60 years of age with IDH1 gene-mutated,
untreated AML who are not candidates for
standard therapy due to age, performance
status, and/or adverse risk factors
AG-221 IDH2 1 NCT1915498 >18 years of age with IDH2 gene-mutated,
advanced hematologic malignancy
CB-839 Glutaminase 1 NCT02071927 > 18 years of age with relapsed/refractory
AML after at least one prior treatment
regimen
Erwinaze Serum glutamine 1 NCT02283190 > 18 years of age with IDH1 or IDH2 gene-
mutated relapsed/refractory AML to
first-line therapy, with or without additional
subsequent therapy
 Methods for the diagnosis and treatment of acute myeloid leukemia (AML)  203

(decitabine). Both are pyrimidine analogs that is incorporated into DNA, whereas azacitidine is
function as inhibitors of DNA methyltransfer- incorporated into RNA.
ases (Table 10.8). Both agents reverse aberrant Guadecitibine (SGI-110) is a second-generation
DNA hypermethylation and restore expression HMA and a dinucleotide of decitabine and deoxy-
of critical tumor-suppressor genes. Decitabine guanosine (Nebbioso et  al., 2015). A number of

Table 10.8  Examples of epigenetic targeted therapy in AML

Class of epigenetic
regulator Target Compound Preclinical or clinical use
DNA DNMTs Azacitidine Approved for clinical use
methyltransferase Decitabine Approved for clinical use
(DNMT) Novel inhibitors Preclinical and clinical use
Regulator of IDH1, IDH2 Inhibitors of See Table 10.7
methylation mutant IDH1/2
Histone lysine CREBBP (CBP) CREBBP inhibitor, Preclinical use
acetyltransferase EP300 (p300) EP300 inhibitor Preclinical use
Histone deacetylase HDACs HDAC inhibitors Several clinical trials ongoing, often in
(HDAC) combination with other treatment
modalities, e.g., with DNMT
inhibitors, Clinical Trials.gov
identifiers NCT01617226 and
NCT00867672; with conventional
chemotherapy NCT01802333 or in
conjuction with allogeneic stem cell
transplantation NCT01451268
Histone acetyl Bromodomain- BET inhibitors Several clinical trials ongoing with
reader containing proteins compounds, including OTX-015
(BET proteins) (NCT01713582), CPI-0610
(NCT02308761), GSK525762
(NCT01943851)
Histone lysine EZH2 EZH2 inhibitors Preclinical
Methyltransferase MLL complexes DOT1L inhibitors Clinical trial with compound EPZ-5676
(NCT01684150); inhibitors of
MLL-menin preclinical interface
Inhibitors of MLL-LEDGF preclinical
interface
Histone lysine LSD1 LSD1 inhibitors Clinical trial with compound
demethylase GSK2879552 (NCT02177612) and
tranylcypromine in combination with
tretinoine (NCT02261779)
Jumonji family of Small molecules Preclinical
histone targeting and
demethylases selectively
(KDMs) inhibiting
Jumonji histone
demethylase
activity
Histone arginine PRMTs PRMT inhibitors Preclinical
methyltransferase
204  Precision medicine in acute myeloid leukemia

DNMT inhibitors have been reported, but most kinases inhibitors, beta-catenin inhibitor (Fiskus
of them are nucleoside analogs that can lead to et al., 2015), G/M checkpoint abrogators, and che-
toxic side effects and lack specificity. By combin- motherapy (Bose and Grant, 2015).
ing docking-based virtual screening with bio- Mutations of the spliceosome machinery range
chemical analyses, a novel compound, DC_05, from <1% to 90% in secondary AML compared
was identified (Chen et al., 2014). DC_05 is a non- to <1% to 10.5% in de novo AML (Mohamed
nucleoside DNMT1 inhibitor with low micromo- et al., 2014; Saez et al., 2017; Yoshida et al., 2011).
lar IC50 values and significant selectivity toward Mutations in splicing factor 3 subunit b1 (SF3b1),
other AdoMet-dependent protein methyltransfer- U2 small nuclear RNA auxiliary factor 1 (U2AF1),
ases. Through a process of similarity-based analog serine arginine-rich splicing factor 2 (SRSF2), and
searching, compounds DC_501 and DC_517 were U2 small nuclear RNA auxiliary factor with zinc
found to be more potent than DC_05. These three finger CCCH-type (ZRSR2) occur in AML. A num-
potent compounds significantly inhibited cancer ber of natural products derived from distinct spe-
cell proliferation. cies of bacteria have been found to target the SF3B
Histone methylome (HMT) is another major component of the spliceosome and demonstrate
epigenetic determinant in gene expression and is potent antitumor activities. One of the first to be
frequently deregulated in AML, especially in MLL- identified was FR901464, a fermentation prod-
rearranged leukemia (Tsai and So, 2017). The first uct from Pseudomonas, which has been used as a
HMT inhibitor targeting DOTL1, EPZ4777, and its structural model for the synthesis of several stable
second-generation derivative, EPZ5676, have been chemical analogs of similar or greater potency.
developed and tested for suppressing MLL leuke- In particular, meayamycin B, a soluble synthetic
mia (Daigle et  al., 2011, 2013). Both compounds derivative, demonstrates important potential for
showed selective inhibitory effects on H3K79 development as a novel therapy for AML treat-
methylation and cells bearing MLL fusions, lead- ment (Wojtuszkiewicz et al., 2014). Meayamycin B
ing to the first clinical trial of HMT inhibitors in inhibits the SF3B1 subunit and can shift alterna-
AML. Protein arginine N-methyltransferase 1 tive splicing of MCL-1 to promote expression of the
(PRMT1) inhibitor AMI-408 targets H4R3 meth- proapoptotic MCL-1s isoform. Although SF3B1 is
yltransferase, leading to repression of MLL fusion among the most commonly mutated splice factors
targets (Cheung et  al., 2016). Enhancer of zeste 2 both in MDS and AML, these mutations are not
polycomb repressive complex 2 subunit (EZH2) located in the putative binding region of spliceo-
inhibitors DZNep and UNC1999 target H3K27 some inhibitors like meayamycin B, suggesting
methyltransferase, leading to derepression of poly- that they could be effective therapeutic options
comb targets (Xu et al., 2015; Zhou et al., 2011). for patients with these mutations (de Necochea-
Bromodomain and extraterminal (BET) proteins Campion et al., 2016). Further inhibitors (17S-FD-
bind to acetylated lysines in histones and control 895, pladienolide B, and sudemycins) were also
gene expression. ABBV-075 is a potent and selective tried (Shirai et al., 2017; Zhou and Chng 2017).
BET family bromodomain inhibitor that recently CD33 antibody-drug conjugates such as gem-
entered Phase I clinical trials. Apoptosis induced by tuzumab ozogamicin and vadastuximab talirine
ABBV-075 was mediated in part by modulation of (SGN-CD33A) (Burnett et  al., 2012; Daver et  al.,
the intrinsic apoptotic pathway, exhibiting synergy 2016; Minagawa et al., 2016; Table 10.9); Hedgehog
with the BCL-2 inhibitor venetoclax in preclinical inhibitors like sonidebig (erismodegib, LDE255),
models of AML (Bui et al., 2017). glasdegib (PF-04449913), and vismodegib; ned-
Histone deacetylase inhibitors (HDACIs) have dylation and subsequent ubiquitination inhibitors
been studied in clinical trials (Table 10.8). The (pevonedistat); and Wnt signaling pathway inhibi-
response rates for single-agent HDACIs in AML tors (Fukushima et  al., 2016; Hanna and Shevde,
have been relatively low. Therefore, most investiga- 2016; Ma et al., 2015; Medler et al., 2015; Swords et al.,
tions of HDACIs have involved combination with 2017) together with standard chemotherapy are in
other agents, particularly the hypomethylating ongoing trials. Further therapeutic approaches
agents decitabine and azacitidine (Bose and Grant, including NF-κB signaling pathway inhibition
2015; Shafer and Grant, 2016), proteasome inhibi- (Bosman et al., 2016; Fuchs 2010; Siveen et al., 2017;
tors, cyclin-dependent kinase inhibitors, tyrosine Zhou et  al., 2015), PI3 K/AKT/mTOR signaling
 Methods for the diagnosis and treatment of acute myeloid leukemia (AML)  205

Table 10.9  Trials for antibody-based targeting AML stem cells

AML stem cell antigen Description Antibody Trials

CD123 IL-3 receptor α chain SGN-123A, antibody-drug NCT02848248
conjugate
JNJ-56022473 dual-specific NCT02472145
Antibody with CD16
SL-401, antibody toxin chimeric NCT02113982
Protein using diphtheria toxin NCT02270463
XmAb14045, dual-specific NCT02730312
Antibody with CD3
MGD006, dual specific NCT02152956
antibody with CD3
KHK2823, IgG 1 antibody NCT02181699
CLL-1 C-type lectin-like
molecule
CD33 Receptor of myeloid SGN-CD33A, antibody-drug NCT02326584
cells, member of the conjugate NCT02785900
Ig superfamily NCT02312037
NCT02614560
IMGN779, antibody drug NCT02674763
conjugate
Gemtuzumab ozogamicin, NCT01869803
antibody-drug conjugate
CD44 Receptor for hyaluronic
acid
CD47 Receptor for signal Hu5F9-G4, anti-CD47 NCT02678338
Regulatory protein α Monoclonal antibody
(SIRP α)
TTI-621, antibody against NCT02663518
CD47
Binding domain of SIRP α
INBRX-103 (CC-90002), NCT02367196
Anti-CD47 monoclonal NCT02641002
antibody
CD96 Tactile, member of
Ig superfamily
CD93 Marker of a
nonquiescent
AML stem cell
population
In MLL-rearranged AML
CD25 IL-2 receptor α chain NCT02588092

pathway inhibition, (Brenner et  al., 2016; Fuchs inhibition had antiproliferative effects on primary
2011; Hauge et al., 2016) and targeting the S100A8/ human AML cells for a subset of patients identified
S100A9-TLR4-ERK/JNK/p38 pathway (Laouedj by gene expression profiling (Brenner et al., 2017).
et  al., 2017; Tamburini, 2017) were studied. Cell Immunotherapies, which have shown prom-
division cycle 25 (CDC25) protein phosphatases ise in the treatment of hematologic malignancies,
206  Precision medicine in acute myeloid leukemia

have the potential to target AML through path- cells–specific antigen was interleukin-3 receptor
ways that are distinct and complementary to α (CD123) within CD34+/CD38− compartment
current approaches (Lichtenegger et  al., 2015). A (Jordan et al., 2000). Multiple additional immuno-
second-generation CD33-specific chimeric antigen phenotypic differences that may distinguish AML
receptor capable of redirecting cytolytic effector T stem cells from normal hematopoietic stem cells
cells against leukemic cells was prepared. CD33 is were subsequently identified and can be used for
expressed in approximately 90% of AML cases and therapeutic intervention using antibody-based tar-
has demonstrated utility as a target of therapeu- geting strategies (Table 10.9) (Pollyea and Jordan,
tic antibodies. Chimeric antigen receptor (CAR)- 2017).
modified T cells efficiently killed leukemia cell
lines and primary tumor cells in vitro. The antileu- FAMILIAL PREDISPOSITION TO AML
kemia effect was CD33-specific, mediated through
T-cell effector functions, and displayed tumor lysis AML is not only sporadic but also familial MDS/
at effector:target ratios as low as 1:20. Furthermore, AML predisposition caused by specific mutations
the CD33-redirected T cells were effective in vivo, has been recognized. The myeloid neoplasms with
preventing the development of leukemia after pro- genetic predisposition represent a new category
phylactic administration and delaying the pro- in the revised 2016 World Health Organization
gression of established disease in mice. These data classification. According to the new classifica-
provide preclinical validation of the effectiveness tion, these disorders are subdivided based on the
of a second-generation anti-CD33 chimeric antigen clinical and genetic features, including myeloid
receptor therapy for AML, and support its contin- neoplasms with germ-line predisposition alone,
ued development as a clinical therapeutic (O’Hear or with preexisting platelet disorder, cytopenias,
et al., 2015; Minagawa et al., 2016). An alternative or other organ failures. The predisposing genetic
antibody-based immunotherapeutic strategy is a factors include mutations in RUNX1, CEBPA,
novel class of bispecific T-cell-engaging antibodies GATA2, ANKRD26 (ankyrin repeat domain 26),
(BiTEs) targeting CD33 antigen on AML cells and ETV6, (gene for ETS family transcriptional repres-
the CD3e component of the T-cell receptor com- sor, variant 6), DDX41, TERC, or TERT (telomer-
plex (AMG 330) (Lichtenegger et al., 2015). ase subunit genes), and SRP72 (signal recognition
particle 72 gene) (Babushok et al., 2016; Churpek
Therapeutic targeting of AML stem et al., 2015; Godley, 2014; Owen et al., 2008; West
cells et  al., 2014). The genes affected in familial AML
are important regulators of hemopoiesis and are
The leukemia stem cells (LSCs) are characterized by frequently implicated in leukemogenesis, pro-
their unlimited self-renewal, repopulating poten- viding deeper insight into the understanding
tial and long residence in a quiescent state of G0/G1 of normal and malignant hemopoiesis. Despite
phase. LSCs are considered to have a pivotal role in the growing knowledge of germ-line predispos-
the relapse and refractory of AML. Therefore, new ing events in the background of familial myeloid
therapeutic strategies to target LSCs with limited malignancies, the germ-line genetic component is
toxicity toward the normal hematopoietic popu- still unknown in a subset of these pedigrees, sug-
lation is critical for the ultimate curing of AML. gesting the presence of so-far-unidentified pre-
Ongoing research works with natural products disposing mutations.
like parthenolide (a natural plant extract derived
compound) and its derivatives that have the abil- FIRST MODELS FOR PREDICTING
ity to target multiple pathways that regulate the OUTCOME FOR INDIVIDUAL AML
self-renewal, growth, and survival of LSCs point PATIENTS
to ways for a possible complete remission in AML
(Siveen et al., 2017). AML stem cells are not only Papaemmanuil et  al. (2016) studied whole exome
in the quiescent phase of the cell cycle (G0) and sequences from 111 myeloid cancer genes in
can be also found in cycling cells. Their frequency diagnostic leukemia samples from 1540 patients
increases as a function of chemotherapy and sub- with AML who were undergoing intensive treat-
sequent relapse. The first clearly defined AML stem ment in three prospective clinical trials of the
 Acknowledgments 207

German-Austrian AML Study Group. They iden- application of genomic techniques have led to
tified driver point mutations and combined these major advances in our knowledge of the patho-
data with the clinical database to generate a com- genesis of AML and should finally lead to more
prehensive knowledge bank. The 231 predictor subset-specific AML therapies. These therapies
variables were derived from these data and they will be ideally tailored to each patient with AML
include fusion genes, copy number alterations, (Dombret and Gardin, 2016). Current advances
point mutations, gene–gene interactions, demo- in nonacute promyelocytic leukemia lack innova-
graphic features, clinical risk factors, and treat- tion. However, modification and intensification of
ment. Various validation strategies were used doses and schedules of standard cytotoxic drugs
to assess the accuracy of predictions. The model were improved and great progress in HSCT tech-
built from these three prospective clinical trials niques was achieved. Distinct AML subsets were
was tested on an independent AML cohort from characterized, but therapy remained uniform.
the United States (TCGA) (Cancer Genome Atlas New effective agents and their combination and
Research Network, 2013). The model for predict- also combined with standard chemotherapy are
ing outcomes developed from these trials is consid- still under investigation and under clinical trials.
erable more complex (Gerstung et  al., 2017) than Calculations show that databases require infor-
the ELN genetic scoring system (Döhner et  al., mation from thousands of patients for accurate
2010). The individual risk in this AML cohort was decision support. Clinical and genomic knowl-
continuous, with no cutoff points for stratifica- edge banks and models for predicting outcome of
tion. More detailed survival estimates confirmed AML patients will be in the future considerably
known ELN risk groups. However, the predicted more complex than those currently used in clinical
survival of one-third of patients deviated more practice.
than 20% from the ELN stratification. Gerstung Routine clinical use of MRD testing requires
et  al. (2017) found that clinical and demographic further refinements and standardization of assays
factors (patient age, performance status and blood and reported results. MRD results will be used not
counts) substantially influenced rates of early only as tool for therapy decision making but also
death, including death in remission. Genomic fea- as an alternate end point for drug development
tures, such as copy number changes, fusion genes, and regulatory approval (Hourigan et  al., 2017).
or driver point mutations, most strongly impacted Currently, survival is the end point used as evi-
the dynamics of disease remission and relapse. The dence for the clinical benefit of new drugs or new
model for predicting outcome was also used for the drug combinations for the therapy of AML.
estimation for individual AML patients whether to Research in the last few years revealed altera-
choose HSCT (allograft) or chemotherapy in first tions in the microenvironment of the hematopoi-
remission. This choice is very important because a etic stem cells (the niche compartment) in AML.
transplantation-associated mortality rate is 20%– Leukemic cells can remodel the niche into a per-
25% and the risk of graft versus host disease is missive environment favoring the expansion of
nonnegligible. Gerstung et al., estimated that 12% leukemic stem cells and progression of disease
(124/995) of patients aged 18–60 years would have (Korn and Méndez-Ferrer, 2017). The microenvi-
an improvement in survival of three years after ronment also mediates chemotherapy-resistance
HSCT in first complete remission as compared to pathways. Therefore, the understanding of niche-
standard chemotherapy. Only 28 from these 124 controlled resistance is important for combined
patients were identified as having “adverse risk“ by treatment targeting not only the mutated cells but
current criteria. Personally tailored management also the microenvironment of leukemic cells.
decisions could reduce the number of HSCT in
AML patients by 20%–25% while maintaining the ACKNOWLEDGMENTS
same survival rates.
This work was supported by a research project for
CONCLUSION conceptual development of research organization
(00023736; Institute of Hematology and Blood
Recent advances in AML biology and its genetic Transfusion, Prague) from the Ministry of Health
landscape due to the continuously increasing of the Czech Republic.
208  Precision medicine in acute myeloid leukemia

CONFLICT OF INTEREST Bose P, Grant S. Rational combinations of

targeted agents in AML. J Clin Med
There are no commercial or other associations in 4;2015:634–64.
connection with this review. Bosman MC, Schuringa JJ, Vellenga E.
Constitutive NF-κ activation in AML: Causes
REFERENCES and treatment strategies. Crit Rev Oncol
Hematol 98;2016:35–44.
Al-Issa K, Nazha A. Molecular landscape in acute Brands-Nijenhuis AV, Labopin M, Schouten HC
myeloid leukemia: where do we stand in 2016. et al. Monosomal karyotype as an adverse
Cancer Biol Med 13;2016:474–82. prognostic factor in patients with acute
Arber DA, Orazi A, Hasserjian R et al. The 2016 myeloid leukemia treated with allogeneic
revision to the World Health Organization hematopoietic stem-cell transplantation in
classification of myeloid neoplasms and acute first complete remission: A retrospective
leukemia. Blood 127;2016:2391–405. survey on behalf of the ALWP of the EBMT.
Ayatollahi H, Shajiei A, Sadeghian MH, Haematologica 101;2016:248–55.
Sheikhi M, Yazdandoust E, Ghazanfarpour Brenner AK, Andersson Tvedt TH, Bruserud Ø.
M, Shams SF, Shakeri S. Prognostic impor- The compexity of targeting PI3K-AKT-mTOR
tance of C-KIT mutations in core binding signalling in human acute myeloid leukaemia:
factor acute myeloid leukemia: A system- The importance of leukemic cell heterogene-
atic review. Hematol Oncol Stem Cell Ther ity, neighbouring mesenchymal stem cells
10;2017:1–7. and immunocompotent cells. Molecules
Babushok DV, Bessler M, Olson TS. Genetic 21;2016:E1512.
predisposition to myelodysplastic syn- Brenner AK, Reikvam H, Rye KP, Hagen KM,
drome and acute myeloid leukemia in Lavecchia A, Bruserud Ø. CDC25 inhibition
children and young adults. Leuk Lymphoma in acute myeloid leukemia: A study of patient
57;2016:520–36. heterogeneity and the effects of different
Bacher U, Kern W, Schnittger S, Hiddemann W, inhibitors. Molecules 22;2017:E446.
Schoch C, Haferlach T. Further correlations Buccisano F, Walter RB. Should patients with acute
of morphology according to FAB and WHO myeloid leukemia and measurable residual dis-
classification to cytogenetics in de novo acute ease be transplanted in first complete remis-
myeloid leukemia: A study on 2,235 patients. sion? Curr Opin Hematol 24;2017:132–8.
Ann Hematol 84;2005:785–91. Bui MH, Lin X, Albert DH S et al. Preclinical
Baljevic M, Park JH, Stein E, Douer D, Altman JK, characterization of BET family bromodomain
Tallman MS. Curing all patients with acute inhibitor ABBV-075 suggests combination
promyelocytic leukemia: Are we there therapeutic strategies. Cancer Res 2017. doi:
yet? Hematol Oncol Clin North Amer 10.1158/0008-5472.CAN-16-1793.
25;2011:1215–33. Bullinger L, Döhner K, Döhner H. Genomics
Bennett JM, Catovsky D, Daniel MT, Flandrin G, of acute myeloid leukemia diagnosis and
Galton DA, Gralnick HR, Sultan C. Proposals pathways. J Clin Oncol 2017. doi: 10.1200/
for the classification of the acute leukemias. JCO.2016.71.2208.
French-American-British (FAB) co-operative Burnett AK. New induction and postinduction
group. Brit J Haematol 33;1976:451–58. strategies in acute myeloid leukemia. Curr
Bennett JM, Catovsky D, Daniel MT, Flandrin G, Opin Hematol 19;2012:76–81.
Galton DA, Gralnick HR, Sultan C. Proposed Burnett AK, Russell NH, Hills RK et al.
revised criteria for the classification of acute Optimization of chemotherapy for younger
myeloid leukemia. A report of the French- patients with acute myeloid leukemia: Results
American-British cooperative group. Ann Int of the medical research council AML15 trial.
Med 10;1985:620–25. J Clin Oncol 31;2013:3360–8.
Blum S, Greve G, Lübert M. Innovative strategies Burnett AK, Russell NH, Hills RK et al. Addition
for adverse karyotype acute myeloid leuke- of gemtuzumab ozogamicin to induction che-
mia. Curr Opin Hematol 24;2017:89–98. motherapy improves survival in older patients
 References 209

with acute myeloid leukemia. J Clin Oncol de Necochea-Campion R, Shouse GP, Zhou Q,
30;2012:3924–31. Mirshahidi S, Chen CS. Aberrant splicing and
Cancer Genome Atlas Research Network. drug resistance in AML. J Hematol Oncol
Genomic and epigenomic landscapes of adult 9;2016:85.
de novo acute myeloid leukemia. N Engl J Döhner H, Estey EH, Amadori S et al. Diagnosis
Med 368;2013:2059–74. and management of acute myeloid leukemia
Chaturvedi A, Araujo Cruz MM, Jyotsana N et al. in adults: recommendations from an interna-
Mutant IDH1 promotes leukemogenesis in tional expert panel, on behalf of the European
vivo and can be specifically targeted in human LeukemiaNet. Blood 115;2010:453–74.
AML. Blood 122;2013:2877–87. Döhner H, Estey EH, Grimwade D et al.
Chen S, Wang Y, Zhou W et al. Identifying novel Diagnosis and management of AML in
selective non-nucleoside DNA methyltransfer- adults: 2017 ELN recommendations from
ase 1 inhibitors through docking-based virtual an international expert panel. Blood
screening. J Med Chem 57;2014:9028–41. 129;2017:424–47.
Cheung N, Fung TK, Zeisig BB et al. Targeting Dombret H, Gardin C. An update of current
aberrant epigenetic networks mediated by treatments for adult acute myeloid leukemia.
PRMT1 and KDM4C in acute myeloid leuke- Blood 127;2016:53–61.
mia. Cancer Cell 29;2016:32–48. Eisfeld AK, Mrózek K, Kohlschmidt J et al. The
Churpek JE, Godley LA. How I diagnose and man- mutational oncoprint of recurrent cytogenetic
age individuals at risk for inherited myeloid abnormalities in adult patients with de novo
malignancies. Blood 128;2016:1800–13. acute myeloid leukemia. Leukemia 2017. doi:
Churpek JE, Pyrtel K, Kanchi KL et al. Genomic 10.1038/leu.2017.86.
analysis of germ line and somatic variants in Emadi A, Jun SA, Tsukamoto T, Fathi AT,
familial myelodysplasia/acute myeloid leuke- Minden MD, Dang CV. Inhibition of
mia. Blood 126;2015:2484–90. ­glutaminase selectively suppresses the
Coenen EA, Zwaan CM, Meyer C, Marschalek growth of primary acute myeloid leuke-
R, Pieters R, van der Veken LT, Beverloo HB, mia cells with IDH mutation. Exp Hematol
van den Heuvel-Eibrink MM. KIAA1524: A 42;2014:247–51.
novel MLL translocation partner in acute Estey E, Dohner H. Acute myeloid leukemia.
myeloid leukemia. Leuk Res 35;2011:133–5. Lancet 368;2006:1894–907.
Daigle SR, Olhava EJ, Therkelsen CA et al. Fang M, Storer B, Estey E, Othus M, Zhang L,
Potent inhibition of DOTL1 as treat- Sandmaier BM, Appelbaum FR. Outcome
ment of MLL-fusion leukemia. Blood of patients with acute myeloid leukemia
122;2013:1017–25. with monosomal karyotype who undergo
Daigle SR, Olhava EJ, Therkelsen CA et al. haematopoietic cell transplantation. Blood
Selective killing of mixed lineage leukemia 118;2011:1490–4.
cells by a potent small-molecule DOT1L Fathi AT, Wander SA, Faramand R, Emadi A.
inhibitor. Cancer Cell 20;2011:53–65. Biochemical, epigenetic, and metabolic
Dang L, Yen K, Attar EC. IDH mutations in approaches to target IDH mutations in
cancer and progress toward develop- acute myeloid leukemia. Semin Hematol
ment of targeted therapeutics. Ann Oncol 52;2015:165–71.
27;2017:599–608. Ferrara F, Palmieri S, Leoni F. Clinically useful
Daver N, Kantarjian H, Ravandi F et al. A prognostic factors in acute myeloid leukemia.
phase II study of decitabine and gemtu- Crit Rev Oncol Hematol 66;2008:181–93.
zumab ozogamicin in newly diagnosed and Fiskus W, Sharma S, Saha S et al. Pre-clinical
relapsed acute myeloid leukemia and high- efficacy of combined therapy with novel
risk myelodysplastic syndrome. Leukemia beta-catenin antagonist BC2059 and his-
30;2016:268–73. tone deacetylase inhibitor against AML cells.
De Kouchkovsky I, Abdul-Hay M. Acute myeloid Leukemia 29;2015:1267–78.
leukemia: a comprehensive review and 2016 Foran JM. New prognostic markers in acute
update. Blood Cancer J 6;2016:e441. myeloid leukemia: Perspective from the clinic.
210  Precision medicine in acute myeloid leukemia

Hematology Am Soc Hematol Educ Prog leukemia. Hematol Oncol Clin North America
2010;2010:47–55. 25;2011:1135–61.
Fuchs O. Transcription factor NF-κB inhibi- Gröschel S, Sanders M, Hoogenboezem R
tors as single therapeutic agents or in et al. A single oncogenic enhancer rear-
­combination with classical chemothera- rangement causes concomitant EVI1 and
peutic agents for the treatment of hema- GATA2 ­deregulation in leukemia. Cell
tologic malignancies. Curr Mol Pharmacol 157;2014:369–81.
3;2010:98–122. Grunwald MR, Levis MJ. FLT3 tyrosine kinase
Fuchs O. Promising activity of mammalian tar- inhibition as a paradigm for targeted drug
get of rapamycin inhibitors in hematologic development in acute myeloid leukemia.
malignancies therapy. Curr Sign Transd Ther Semin Hematol 52;2015:193–9.
61;2011:44–54. Haferlach T, Stengel A, Eckstein C, Perglerová K,
Fukushima N, Minami Y, Kakiuchi S, Kuwatsuka Y, Alpermann T, Kern W, Haferlach C,
Hayakawa F, Jamieson C, Kiyoi H, Naoe T. Meggendorfer M. The new provisional WHO
Small-molecule Hedgehog inhibitor attenu- entity “RUNX1 mutated AML” shows specific
ates the leukemia-initiation potential of genetics but no prognostic influence of dys-
acute myeloid leukemia cells. Cancer Sci plasia. Leukemia 30;2016:2109–12.
107;2016:1422–9. Hanna A, Shevde LA. Hedgehog signaling:
Gadzik VI, Teleanu V, Papaemmanuil E et al. modulation of cancer properties and tumor
RUNX1 mutations in acute myeloid leu- microenvironment. Mol Cancer 15;2016:24.
kemia are associated with distinct clinic- Hauge M, Bruserud Ø, Hatfield KJ. Targeting
pathologic and genetic features. Leukemia of cell metabolism in human acute myeloid
30;2016:2160–8. leukemia-more than targeting of isoci-
Gale RP, Wiernik PH, Lazarus HM. Should trate dehydrogenase mutations and PI3 K/
persons with acute myeloid leukemia have AKT/mTOR signaling? Eur J Haematol
a transplant in first remission? Leukemia 96;2016:211–21.
28;2014:1949–52. Heuser M. Therapy-related myeloid neoplasms:
Gallogly MM, Lazarus HM. Midostaurin: An Does knowing the origin help to guide treat-
emerging treatment for acute myeloid leuke- ment? Hematology Am Soc Hematol Educ
mia patients. J Blood Med 7;2016:73–83. Program 2016(1);2016:24–32.
Gerstung M, Papaemmanuil E, Martincorena I Hinai AA, Valk PJ. Aberrant EVI1 expression
et al. Precision oncology for acute myeloid in acute mzeloid leukemia. Brit J Haematol
leukemia using a knowledge bank approach. 172;2016:870–8.
Nat Genet 49;2017:332–40. Hourigan CS, Gale RP, Gormley NJ,
GLOBOCAN 2014: Estimated Cancer Incidence, Ossenkoppele Gj, Walter RB. Measurable
Mortality and Prevalence Worldwide, 2012. residual disease testing in acute myeloid
www.globocan.iarc.fr. Accessed September leukaemia. Leukemia 2017. doi: 10.1038/
21, 2014. leu.2017.113.
Godley LA. Inherited predisposition to Jang JE, Min YH, Yoon J et al. Single monosomy
acute myeloid leukemia. Semin Hematol as a relatively better survival factor in acute
51;2014:306–21. myeloid leukemia patients with monosomal
Grimwade D, Hills RK, Moorman AV et al. karyotype. Blood Cancer J 5;2015:e358.
Refinement of cytogenetic classification in Jan M, Ebert BL, Jaiswal S. Clonal hematopoiesis.
acute myeloid leukemia: Determination of Semin Hematol 54;2017:43–50.
prognostic significance of rare recurring chro- Jordan CT, Upchurch D, Szilvassy SJ et al.
mosomal abnormalities among 5876 younger The interleukin-3 receptor alpha chain is
adult patients treated in the United Kingdom a unique marker for human acute myelog-
Medical Research Council trials. Blood enous l­eukemia stem cells. Leukemia
116;2010:354–65. 14;2000:1777–84.
Grimwade D, Mrózek K. Diagnostic and prog- Kampen KR. The discovery and early understand-
nostic value of cytogenetics in acute myeloid ing of leukemia. Leuk Res 36;2012:6–13.
 References 211

Kansal R. Acute myeloid leukemia in the era leukemogenesis and its effect is reversible.
of precision medicine: recent advances in Science 339;2013:1621–5.
diagnostic classification and risk stratification. Lugthart S, Gröschel S, Baverloo HB et al.
Cancer Biol Med 13;2016:41–54. Clinical, molecular, and prognostic signifi-
Kayser S, Zucknick M, Döhner K et al. Monosomal cance of WHO type inv(3)(q21q26.2)/t(3;3)
karyotype in adult acute myeloid leu- (q21;q26.2) and various other 3q abnormali-
kemia: prognostic impact and outcome ties in acute myeloid leukemia. J Clin Oncol
after different treatment strategies. Blood 28;2010:3890–98.
118;2012:551–8. Ma S, Yang LL, Niu T et al. SKLB-677, an FLT3 and
Kohlmann AA, Bacher U, Schnittger S, Haferlach T. Wnt/beta-catenin signaling inhibitor, displays
Perspective on how to approach molecular potent activity in models of FLT3-driven AML.
diagnostics in acute myeloid leukemia and Sci Rep 5;2015:15646.
myelodysplastic syndromes in the era of next Marcucci G. Core binding factor acute myeloid leu-
generation sequencing. Leuk Lymphoma kemia. Clin Adv Hematol Oncol 4;2006:339–41.
55;2014:1725–34. Mayer LD, Harasym TO, Tardi PG et al. Ratiometric
Korn C, Méndez-Ferrer S. Myeloid malignan- dosing of anticancer drug combinations:
cies and the microenvironment. Blood Controlling drug ratios after systemic adminis-
129;2017:811–22. tration regulates therapeutic activity in tumor-
Krause JR. Morphology and classification of acute bearing mice. Mol Cancer Ther 5;2006:1854–63.
myeloid leukemias. Clin Lab Med 20;2000:1–16. Medeiros BC, Fathi AT, DiNardo CD, Pollyea DA,
Lam K, Zhang DE. RUNX1 and RUNX1-ETO: Roles Chan SM, Swords R. Isocitrate dehydrogenase
in hematopoiesis and leukemogenesis. Front mutations in myeloid malignancies. Leukemia
Biosci 17;2012:1120–39. 31;2017:272–81.
Laouedj M, Tardif MR, Gil L, Raquil MA, Medler CJ, Waddell JA, Solimando DA. Cancer
Lachhab A, Pelletier M, Tessier PA, Barabé chemotherapy update. Gefitinib and
F. S100A9 induces differentiation of acute sonidegib. Hosp Pharm 50;2015:868–72.
myeloid leukemia cells through TLR4. Blood Metzeler KH, Bloomfield CD. Clinical relevance of
129;2017:1980–90. RUNX1 and CBFB alterations in acute myeloid
Lazarevic V, Rosso A, Juliusson G et al. Incidence leukemia and other hematological disorders.
and prognostic significance of isolated triso- Adv Exp Med Biol 962;2017:175–99.
mies in adult acute myeloid leukemia: A pop- Meyer C, Kowarz E, Hofmann J et al. New
ulation-based study from the Swedish AML insights to the MLL recombinome of acute
registry. Eur J Haematol 98;2017:493–500. leukemias. Leukemia 23;2009:1490–9.
Lennartsson J, Jelacic T, Linnekin D, Shivakrupa R. Meyers J, Yu Y, Kaye JA, Davis KL. Medicare
Normal and oncogenic forms of the recep- fee-for-service enrollees with primary acute
tor tyrosine kinase kit. Stem Cells 23;2005:​ myeloid leukemia: An analysis of treatment
16–43. patterns, survival, and healthcare resource
Lichtenegger FS, Krupka C, Köhnke T, Subklewe M. utilization and costs. Appl Health Econ Health
Immunotherapy for acute myeloid leukemia. Policy 11;2013:275–86.
Semin Hematol 52;2015:207–14. Minagawa K, Jamil MO, Al-Obaidi M et al. In
Lin PH, Li HY, Fan SC et al. A targeted next- vitro pre-clinical validation of suicide gene
generation sequencing in the molecular risk modified anti-CD33 redirected chimeric
stratification of adult acute myeloid leukemia: antigen receptor T-cells for acute myeloid
Implications for clinical practice. Cancer Med leukemia. PLOS ONE 11;2016:e0166891.
6;2017:349–60. Mohamed AM, Thénoz M, Solly F, Balsat M,
Lo-Coco F, Ammatuna E, Montesinos P, Sanz MA. Mortreux F, Wattel E. How mRNA is mis-
Acute promyelocytic leukemia: Recent spliced in acute myelogenous leukemia
advances in diagnosis and management. (AML)? Oncotarget 5;2014:9534–45.
Semin Oncol 35;2008:401–9. Mrózek K, Heerema N, Bloomfield CD.
Losman JA, Looper RE, Koivunen P et al. (R)-2- Cytogenetics in acute leukemia. Blood Rev
hydroxyglutarate is sufficient to promote 18;2004:115–36.
212  Precision medicine in acute myeloid leukemia

Mrózek K, Marcucci G, Paschka P, Bloomfield CD. Röllig C, Bornhäuser M, Thiede C et al. Long-term
Advances in molecular genetics and treat- prognosis of acute myeloid leukemia accord-
ment of core-binding factor acute myeloid ing to the new genetic risk classification of the
leukemia. Currt Opin Oncol 20;2008:711–8. European LeukemiaNet recommendations:
Nazha A, Zarzour A, Al-Issa K et al. The complex- Evaluation of the proposed reporting system. J
ity of interpreting genomic data in patients Clin Oncol 29;2011:2758–65.
with acute myeloid leukemia. Blood Cancer J Rowe JM. AML in 2016: Where we are now? Best
6;2016:e510. Pract Res Clin Hematol 29;2016:315–9.
Nebbioso A, Benedetti R, Conte M, Iside C, Rücker FG, Schlenk RF, Bullinger L et al. TP53
Altucci L. Genetic mutations in epigenetic alterations in acute myeloid leukemiawith
modifiers as therapeutic targets in acute complex karyotype correlate with specific copy
myeloid leukemia. Expert Opin Ther Targets number alterations, monosomal karyotype, and
19;2015:1187–1202. dismal outcome. Blood 119;2012:2114–21.
O’Hear C, Heiber JF, Schubert I, Fey G, Geiger Saez B, Walter MJ, Graubert TA. Splicing factor
TL. Anti-CD33 chimeric antigen recep- gene mutations in hematologic malignancies.
tor targeting of acute myeloid leukemia. Blood 129;2017:1260–9.
Haematologica 100;2015:336–44. Sallman DA, Lancet JE. What are the most prom-
Owen C, Barnett M, Fitzgibbon J. Familial ising new agents in acute myeloid leukemia?
­myelodysplasia and acute myeloid Curr Opin Hematol 24;2017:99–107.
leukaemia: A review. Br J Haematol Sanders MA, Valk PJM. The evolving molecular
140;2008:123–32. genetic landscape in acute myeloid leukae-
Papaemmanuil E, Gerstung M, Bullinger L mia. Curr Opin Hematol 20;2013:79–85.
et al. Genomic classification and prognosis Sangle NA, Perkins SL. Core-binding factor
in acute myeloid leukemia. N Engl J Med acute myeloid leukemia. Arch Pathol Lab Med
374;2016:2209–21. 135;2011:1504–09.
Paschka P, Marcucci G, Ruppert AS et al. Adverse Sexauer A, Perl A, Yang X et al. Terminal
prognostic significance of KIT mutations in myeloid differentiation in vivo is induced
adult acute myeloid leukemia with inv(16) by FLT3 inhibition in FLT3/ITD AML. Blood
and t(8;21): A Cancer and Leukemia Group B 120;2012:4205–14.
Study. J Clin Oncol 24;2006:3904–11. Shafer D, Grant S. Update on rational targeted
Perrot A, Luquet I, Pigneux A et al. Dismal therapy in AML. Blood Rev 30;2016:273–83.
prognostic value of monosomal karyotype in Shah A, Andersson TM, Rachet B, Bjorkholm M,
elderly patients with acute myeloid leukemia: Lambert PC. Survival and cure of acute
A GOELAMS study of 186 patients with unfa- myeloid leukaemia in England, 1971–2006:
vorable cytogenetic cytogenetic abnormali- A population-based study. Br J Haematol
ties. Blood 118;2011:679–85. 162;2013:509–516.
Paun O, Lazarus HM. Allogeneic hematopoietic Shimada A, Taki T, Tabuchi K et al. KIT mutations,
cell transplation for acute myeloid leukemia in and not FLT3 internal tandem duplication,
first complete remission: Have the indications are strongly associated with a poor progno-
changed? Curr Opin Hematol 19;2012: 95–101. sis in pediatric acute myeloid leukemia with
Pigneux A, Labopin M, Maertens J et al. t(8;21): A study of the Japanese Childhood
Outcome of allogeneic hematopoietic stem- AML Cooperative Study Group. Blood
cell transplantation for adult patients with 107;2006:1806–9.
AML and 11q23/MLL rearrangement (MLL-r Shirai CL, White BS, Tripathi M et al. Mutant
AML). Leukemia 29;2015:2375–81. U2AF1-expressing cells are sensitive to phar-
Pollyea DA, Jordan CT. Therapeutic targeting macological modulation of the spliceosome.
of acute myeloid leukemia stem cells. Blood Nat Commun 8;2017:14060.
129;2017:1627–35. Siegel RL, Miller KD, Jemal A. Cancer statistics,
Prada-Arismendy J, Arroyave JC, Röthlisberger S. 2015. CA Cancer J Clin 65;2015:5–29.
Molecular biomarkers in acute myeloid leuke- Siveen KS, Uddin S, Mohammad RM. Targeting
mia. Blood Rev 31;2017:63–76. acute myeloid leukemia stem cell signaling by
 References 213

natural products. Mol Cancer 16;2017:13. helps define an AML with myelodysplasia-
Smith ML, Hills RK, Grimwade D. Independent related changes category. J Clin Pathol
prognostic variables in acute myeloid leukae- 70;2017a:109–115.
mia. Blood Rev 25;2011:39–51. Weinberg OK, Sohani AR, Bhargava P, Nardi V.
Swords RT, Watts J, Erba HP et al. Expanded Diagnostic work-up of acute myeloid leuke-
safety analysis of pevonedistat, a first-in- mia. Am J Hematol 92;2017b:317–21.
class NEDD8-activating enzyme inhibitor, West AH, Godley LA, Churpek JE. Familial
in patients with acute nyeloid leukemia and myelodysplastic syndrome/acute myeloid
myelodysplastic syndromes. Blood Cancer J leukemia syndromes: A review and utility for
7;2017:e520. translational investigations. Annals NY Acad
Tamamyan G, Kadia T, Ravandi F et al. Sci 1310;2014:111–8.
Frontline treatment of acute myeloid leu- White BS, DiPersio JF. Genomic tools in acute
kemia in adults. Crit Rev Oncol Hematol myeloid leukemia: From the bench to the
110;2017:20–34. bedside. Cancer 120;2014:1134–44.
Tamburini J. S100 proteins in AML: Wojtuszkiewicz A, Assaraf YG, Jansen G, Koide K,
Differentiation and beyond. Blood Bressin RK, Basu U, Sonneveld E. Spliceosome
129;2017:1893–4. inhibitor meayamycin B as a novel potential
Tardi P, Johnstone S, Harasym N et al. chemotherapeutic agent in ALL and AML.
In vivo maintenance of synergistic Blood 124;2014:924.
cytarabine:daunorubicin ratios greatly Wouters BJ, Delwel R. Epigenetics and
enhances therapeutic efficacy. Leuk Res approaches to targeted epigenetic ther-
33;2009:129–39. apy in acute myeloid leukemia. Blood
Thomas X. First contributors in the history of 127;2016:42–52.
leukemia. World J Hematol 2;2013:62–70. Xu B, On DM, Ma A et al. Selective inhibition of
Tsai CT, So CWE. Epigenetic therapies by EZH2 and EZH1 enzymatic activity by a small
targeting aberrant histone methzlome in molecule suppresses MLL-rearranged leuke-
AML: Molecular mechanism, current pre- mia. Blood 125;2015:346–57.
clinical and clinical development. Oncogene Yamagata T, Maki K, Mitani K. RUNX1/AML1 in
36;2017:1753–9. normal and abnormal hematopoiesis. Inter J
Utsun C, Marcucci G. Emerging diagnostic and Hematol 82;2005:1–8.
therapeutic approaches in core binding factor Yamamoto JF, Goodman MT. Patterns of
acute myeloid leukaemia. Curr Opin Hematol leukemia incidence in the United States
22;2015:85–91. by subtype and demographic character-
Vardiman JW, Thiele J, Arber DA et al. istics, 1997–2002. Cancer Causes Control
The 2008 revision of the World Health 19;2008:379–90.
Organization (WHO) classification of Yin JAL, O’Brien MA, Hills RK, Daly SB,
myeloid neoplasms and acute leukemia: Wheatley K, Burnett AK. Minimal residual
Rationale and important changes. Blood disease monitoring by quantitative RT-PCR in
114;2009:937–51. core binding factor AML allows risk stratifi-
Wang B, Liu Y, Hou G et al. Mutational spectrum cation and predicts relapse: Results of the
and risk stratification of intermediate-risk United Kingdom MRC AML-15 trial. Blood
acute myeloid leukemia patients based on 120;2012:2826–35.
next-generation sequencing. Oncotarget Yoshida K, Sanada M, Shiraishi Y et al.
7;2016:32065–78. Frequent pathway mutations of splic-
Wang F, Travins J, DeLaBarre B et al. Targeted ing machinery in myelodysplasia. Nature
inhibition of mutant IDH2 in leukemia cells 478;2011:64–9.
induces cellular differentiation. Science Zhou J, Bi C, Cheong LL et al. The histone
2013;340:622–6. ­methyltransferase inhibitor, DZNep, up-
Weinberg OK, Hasserjian RP, Li B, Pozdnyakova O. regulates TXNIP, increases ROS production,
Assessment of myeloid and monocytic and targets leukemia cells in AML. Blood
dysplasia by flow cytometry in de novo AML 118;2011:2830–9.
214  Precision medicine in acute myeloid leukemia

Zhou J, Ching YQ, Chng W-J. Aberrant nuclear acute myeloid leukemia. Stem Cell Investig
factor-kappa B activity in acute myeloid 4;2017:6.
­leukemia: From molecular pathogen- Zhou Y, Othus M, Araki D et al. Pre- and post-
esis to therapeutic target. Oncotarget transplant quantification of measurable
6;2015:5490–500. (“minimal”) residual disease via multiparam-
Zhou J, Chng W-J. Aberrant RNA splicing eter flow cytometry in adult acute myeloid
and mutations in spliceosome complex in leukemia. Leukemia 30;2016:1456–64.
 2
Part    

Precision medicine in NCDs

 11
Precision medicine in coronary artery
disease

MELVIN GEORGE, LUXITAA GOENKA, AND SANDHIYA SELVARAJAN

Background 217 Development of advanced medications 226

Current diagnostic and therapeutic Development of improved and safer drugs 226
approaches for coronary artery disease Usage of appropriate doses of drug 226
(CAD) 217 Precocious screening for disease 226
Role of genetic, environmental, and lifestyle Reduction in the overall cost of health care 226
risk factors in precision medicine 218 Disadvantages 227
Genetic risk factors 219 Limited alternative therapy 227
Lifestyle risk factors 220 Cost concerns 227
Environmental and societal influences 220 Education of health care providers 227
Personalized medicine in CAD 220 Proteomics 227
Genomic-based gene expression score (GES) 221 Proteomic studies 227
Studies evaluating the gene expression score 222 Limitations of proteomics 228
Pharmacogenomics and pharmacogenetics 223 Challenges in the implementation of
Antiplatelet pharmacogenomics 223 precision medicine 228
Statin pharmacogenetics 224 Future considerations 229
Warfarin pharmacogenetics 225 Conclusion 229
Other examples 226 References 229

BACKGROUND contributes to several comorbidities. Moreover,

treatment of CAD exceeds $100 billion each year
Coronary artery disease (CAD) continues to be a (Heidenreich et al., 2011). As the treatment of CAD
leading cause of mortality and morbidity world- is becoming expensive each year, it is necessary to
wide, despite the significant improvement and place greater emphasis on cost containment and
development in its diagnosis and treatment. CAD improve outcome measures.
is a multifactorial disease that can have differing
clinical presentations (Figure 11.1). Around 17.5 CURRENT DIAGNOSTIC AND
million people have died due to CAD worldwide, THERAPEUTIC APPROACHES FOR
which represents 31% of the overall global deaths. CORONARY ARTERY DISEASE (CAD)
Among these deaths, 7.4 million were due to coro-
nary heart disease (CHD) and 6.7 million were Presently there are several diagnostic and thera-
due to stroke (Krishnan, 2012). CAD decreases peutic strategies available for the detection and
the quality of life (QOL) among individuals and treatment of CAD. The assessment of CAD

217
218  Precision medicine in coronary artery disease

Risk factors

High blood pressure

High cholesterol
Diabetes
Smoking
Obesity
Sedentary lifestyle
Genetic factors
Environmental stress

Congestive heart Atherosclerosis

failure

Heart attack and Coronary artery

stroke disease

Figure 11.1  Risk factors for coronary artery disease.

includes both invasive and noninvasive coronary shown a significant improvement in the treat-
assessment. Invasive coronary angiography (CAG) ment of CTO (Loh and Waksman, 2012; Gao and
has been the gold standard diagnostic approach Chen, 2016). Further, the diagnostic approaches
for establishing the presence, location, and sever- are associated with risks of exposure to radia-
ity of CAD. Invasive CAG is recommended for tion, contrast-induced anaphylaxis, and acute
patients with high risk of CAD according to kidney injury (Einstein et  al., 2007, 2010; Rhee
patient history, electrocardiogram, and biomark- et  al., 2012). Even though the current diagnostic
ers of cardiac injury. However, this technique and therapeutic approaches have been successful
is invasive and expensive and is associated with in the detection and treatment of CAD, the pres-
small but definite risk of harm. The diagnostic ent approaches have their own pros and cons. The
approaches utilized for patients with low to inter- present strategies pose an economic burden on
mediate risk of CAD include noninvasive diagnos- health care systems worldwide. The diagnosis of
tic approaches, such as exercise tests, radioisotope CAD is highly variable among patients present-
myocardial perfusion scanning, coronary calci- ing with symptoms of suspected CAD. Only 10%
fication score, multidetector cardiac computed of the chest pain cases evaluated by physicians
tomography (MDCT), magnetic resonance (MR) results in the diagnosis of CAD, despite the aug-
imaging, and stress echocardiography (Kohsaka mentation of the expensive diagnostic approaches
and Makaryus, 2008). The advances in the current (Klinkman et  al., 1994). Hence, it has become
therapeutic strategies for CAD have shown a tre- necessary to improvise and develop precise newer
mendous decrease in the mortality and morbid- diagnostic and therapeutic strategies for the diag-
ity rate among CAD patients. Basic research and nosis, treatment, and management of CAD.
clinical trials have evaluated the safety and effi-
cacy of statins in reducing the cholesterol levels ROLE OF GENETIC,
among CAD patients, which is one of the major ENVIRONMENTAL, AND LIFESTYLE
risk factors for the disease (Shepherd et al., 1995; RISK FACTORS IN PRECISION
Downs et  al., 1998; Emami et  al., 2017). Further MEDICINE
chronic total occlusions (CTO) are present among
some CAD patients; thereby, advances in percuta- The three primary components of precision
neous coronary intervention (PCI) strategies have medicine are genetic information, lifestyle, and
 Role of genetic, environmental, and lifestyle risk factors in precision medicine  219

Avoidable risk factors

Sedentary lifestyle

Alcohol consumption Combined hormone

replacement therapy

Lifestyle and
Genomics Phenotype environmental risk
factors

Figure 11.2  Association between genomics, lifestyle, and environmental risk factors with precision
medicine.

environmental factors (Figure 11.2). Further, CAD individualized treatment strategies (Kessler et al.,
is a known complex disease that is influenced by 2016). Several GWAS have been conducted to iden-
inherited genetic risk, and environmental and life- tify the single nucleotide polymorphisms (SNPs)
style factors, such as diets, saturated fats, smok- on chromosomes that have an association with the
ing, and sedentary lifestyle (Tenconi et  al., 1992; susceptibility to CAD. The discovery of pro-protein
Alizadeh et  al., 2017). These factors play a key convertase subtilisin/kexin type 9 (PCSK9) inhibi-
role in the pathogenesis and treatment of disease. tion, which is involved in LDL-C receptor recy-
Hence, identifying the link between these factors cling, has led to the discovery and development of
and the disease will guide physicians in under- alirocumab and evolocumab. These molecules are
standing the disease and develop precise diagnos- currently being studied for the treatment of hyper-
tic, therapeutic, and preventive approaches. cholesterolemia along with statin therapy, which
is one of the major risk factors for CAD (Postmus
Genetic risk factors et al., 2013).
The Coriell Personalized Medicine Collaborative
Genetic factors contribute to the risk of CAD and (CPMC) research study was done to investigate the
several developments have occurred in the past potential contributions of common genetic factors
decade. The evolution of genome-wide association to the screening, management, and prevention of
studies (GWAS) over the last 10 years has played complex diseases including CAD. The impact of
a key role in investigating the genetic architecture personalized risk reports including genetic and
of complex human disease. Evidence shows that nongenetic risk factors for CAD on heart health
the GWAS represent a powerful approach in the behaviors was assessed. A total of 683 CPMC par-
identification of genes involved in complex human ticipants who received personalized CAD risk
diseases such as CAD, cancer, and others (Sayols- reports, including genetic risk, family history
Baixeras et al., 2014). GWAS have so far been able risk, and self-reported nongenetic risks based on
to identify approximately 50 genetic variants asso- smoking and diabetes status, were evaluated in the
ciated with CAD. Although the application of this study. It was observed that subjects with increased
strategy in clinical practice is challenging, genetic genetic risk of CAD were significantly more likely
variants can be used to determine the new thera- to report increases in heart health behaviors after
peutic targets and personal genetic information viewing their personalized risk report. Thus, the
in order to improve the prognosis and treatment study indicated that subjects who were aware
of disease. The development in this field so far of  their genetic risk were highly motivated to
has been able to provide insight into the patho- increase their heart health behaviors (Scheinfeldt
physiology of CAD, which is the beginning of the et al., 2016).
220  Precision medicine in coronary artery disease

Thus, the genetic data of patients can provide disease process, in order to provide custom-tai-
a good platform for the diagnosis, prognosis, and lored therapeutic solutions.
treatment of CAD.
Environmental and societal influences
Lifestyle risk factors
Environmental risk factors and socioeconomic
In the recent decade, lifestyle risk factors have been status have an influence on CAD. The environ-
associated with increasing risk of CAD and have mental risk factor is vital but is considered as
increasingly become the center point of research an underestimated risk factor that contributes
interest worldwide (Brotman et al., 2007). In previ- to the development and severity of CAD (Shah
ous studies, it had been demonstrated that healthy et  al., 2016). It is known that the heart and vas-
lifestyle habits have reduced the disease and mor- cular system are highly susceptible to a variety of
tality rates, and the sociodemographic parameters environmental agents, such as ambient air pollu-
such as gender, age, marital status, economic level, tion and the metals arsenic, cadmium, and lead.
and paid employment correlates with healthy life- The exposure to these environmental risk factors
style (Kromhout et al., 2002). contributes to advancement in diseases and mor-
A meta-analysis was conducted to examine the tality through initiation of the pathophysiological
combined association of job strain and lifestyle processes associated with CAD inclusive of blood-
with the risk of CAD. The population attributable pressure control, carbohydrate and lipid metabo-
risk was calculated for three exposures: job strain, lism, vascular function, and atherogenesis (Chow
unhealthy lifestyle, and their combination. The et al., 2009). A study was conducted to analyze the
results of the meta-analysis showed that the risk interaction of a genetic variant of the CDKN2A/
of CAD was high among participants who had an B-rs10811661 gene locus with cardiovascular risk
unhealthy lifestyle and who reported job strain. factors and environmental exposures like (e.g.,
However, the risk of CAD was half among partici- diet and physical activity) among 1165 individu-
pants with job strain and healthy lifestyle. Further, als. Genotyping was carried out using the TaqMan
the 10-year incidence rate of CAD subjects with job real-time PCR-based method. There was a signifi-
strain and a healthy lifestyle was 53% lower than cant association between CDKN2A-rs10811661
the incidence among subjects with job strain and polymorphism with cardiovascular risk factors
an unhealthy lifestyle (Kivimäki et  al., 2013). A and dyslipidemia in a nondiabetic population.
retrospective study was conducted to evaluate the Thus, low energy diet and good physical activity
impact of various daily lifestyle indicators, includ- may relieve the unfavorable effects of T allele of
ing the dinner satiety rate, tobacco use, heavy alco- CDKN2A/B locus (Mehramiz et al., 2016). Hence,
hol use, sleep pattern, anxiety, and exercise, among the environmental risk factors contribute to the
CHD patients who had undergone PCI. The multi- development of CAD among the population. The
variate logistic analysis showed that dinner satiety, analysis of the environmental factors will enable
tobacco use, heavy alcohol use, and exercise sig- physicians to identify the cause of CAD and pro-
nificantly impacted CHD (p < 0.05). The clinical vide better therapeutics for CAD patients.
composite end point of target lesion revasculariza-
tion, defined as repeated PCI and coronary artery PERSONALIZED MEDICINE IN CAD
bypass grafting  (CABG), was observed in 55% of
the cases between 3 and 5 years of the follow-up The 21st century is poised to develop more targeted
period. Hence, it was concluded that poor lifestyle and personalized medicines for disease detection
may increase the in-stent restenosis rate, increase and treatment. Available therapies may be effective
the progress of original lesion, and enhance the for some individuals but be associated with treat-
development of new lesions among patients with ment failures or toxicities in others. “PM (precision
CHD (Wan et al., 2015). medicine) is defined as a novel approach that can
In conclusion, lifestyle factors can play a vital be characterized as molecular, immunologic, and
role in the prognosis and diagnosis of CAD. functional endotyping of the disease, leading to
Precision medicine has the ability to integrate personalized care, with the patient’s engagement
patient’s lifestyle to diagnose and classify the in decision making as to the treatment process
 Personalized medicine in CAD  221

“One-size does not-fit-all”

Patients with drug

toxicity

Genotyping Patients with no

response to drug
therapy

Toxic responders

Non-responders Patients with

Patient population
normal response to
with same disease Responders drug therapy

Figure 11.3  Differential response to drug therapy.

and consideration of predictive and preventive precision medicine is to integrate the genetic, life-
aspects of treatment” (Śliwczynski and Orlewska, style, and environmental information in order to
2016). Precision medicine is a newly emerging classify the populations into subgroups based on
approach, which will evaluate patient’s variability their susceptibility and response to the disease
in their environmental, lifestyle, and genetic risk (Duarte and Spencer, 2016). This will enable physi-
factors individually for the prevention and treat- cians to optimize the therapy for each individual
ment of CAD. CAD is a heritable complex disease based on their respective diagnostic testing. The
caused due to the manifold interactions of dif- difference between traditional therapy and preci-
ferent lifestyle, genetic, and environmental risk sion medicine is illustrated in Figure 11.4.
factors, leading to the change in the molecular
outlook of vascular and metabolic tissues to accel- Genomic-based gene expression
erate atherosclerosis (McGrath and Ghersi, 2016). score (GES)
Advancements in technologies, such as genome
sequencing, big data analysis, and electronic health The utility of genomic-based, personalized medi-
records, have enabled the health care to focus on cine may assist physicians’ medical decision mak-
the individual patient approach than the one-size- ing, including the need for further cardiac testing.
fits-all approach (Figure 11.3). Hence, the goal of The genomic-based gene expression score (GES)

Traditional Precision
medicine medicine

DNA testing

Same therapy
Tailored therapy

Each person will

Adverse benefit
Benefit No benefit
effects

Figure 11.4  Traditional versus precision medicine.

222  Precision medicine in coronary artery disease

will provide an assessment of the present state of were recorded during the 30-day follow-up. During
obstructive CAD by quantifying the gene expression the follow-up, only 0.4% of MACE were detected
changes associated with arthrosclerosis. The GES is (Herman et al., 2014).
calculated by the commercially available, Medicare- A multicenter study was conducted to vali-
covered, validated, quantitative diagnostic test, date, the diagnostic accuracy of a 23-gene
which measures the expression levels of 23 genes expression-based classifier for the assessment of
from the peripheral blood to determine the likeli- obstructive CAD in nondiabetic patients. The
hood of obstructive CAD, as diagnosed by CAG (at Diamond–Forrester (DF) risk score comprising
least one vessel with ≥50% angiographic coronary age, sex, and chest pain was selected to assess the
artery stenosis). The raw algorithm scores are cal- added value of GES to other clinical factors. The
culated from the median expression values for the primary outcome measure was the receiver-operat-
23 algorithm genes, grouped in the six terms, four ing characteristic (ROC) curve area for algorithm
sex-independent and two sex-specific. The scores score prediction of disease status. The ROC curves
are then converted to a 0–40 scale and patients with were estimated for the gene expression algorithm
lower GES are at a lower risk of underlying obstruc- score, the DF risk score, a combined model of algo-
tive coronary disease (Elashoff et al., 2011). rithm score and DF risk score, myocardial perfu-
sion imaging, and a combined model of algorithm
Studies evaluating the gene score and imaging. The primary outcome measure
expression score AUC was 0.70 ± 0.02, (P < 0.001), being indepen-
dently significant in both male (P < 0.001) and
A prospective trial (Investigation of a Molecular female (P < 0.001) subsets. The ROC also showed
Personalized Coronary Gene Expression Test higher AUC for the algorithm score and DF risk
on Primary Care Practice Pattern, clinical trial) score combination when compared to the DF risk
(PREDICT) was conducted to see if the GES score alone (AUC 0.72 versus 0.66, P = 0.003).
results  would reduce the diagnostic uncertainty Thus, the noninvasive whole blood test, based on
among patients presenting with symptoms sug- gene expression and demographics, might be use-
gestive of obstructive CAD and would lead to a ful for the assessment of obstructive CAD in non-
change in making medical decisions, including diabetic patients without known CAD (Rosenberg
the need for further cardiac testing. The primary et al., 2014).
outcome measure was the change in the diagnos- However, one of the limitations of the
tic plan, which was classified into the following PREDICT study was that the study population
­categories (in hierarchical order). was angiographically referred and the accuracy
of GES among patients with lower prevalence of
1. No further cardiac testing or treatment CAD population was unknown. The Coronary
2 . Change in lifestyle or medical therapy Obstruction Detection by Molecular Personalized
3. Stress testing (with or without imaging) or Gene Expression (COMPASS) study was designed
computed tomography/coronary angiography in such a way, where symptomatic nondiabetic
4. Invasive CAG patients referred for myocardial perfusion imaging
(MPI) were included. Thus, the COMPASS study
The investigators observed a change in diagnos- assessed the GES and MPI performance among a
tic plan following GES testing in 58% (n = 145) of low-risk population by minimizing the selection
patients, between the preliminary and final deci- bias. The prespecified primary end point was the
sion. The secondary outcome was to analyze the GES ROC characteristics to segregate ≥50% steno-
change in the diagnostic pattern for each patient sis, in order to identify patients with obstructive
independently. Out of the 145 patients, on whom CAD. It was observed that the area under the ROC
the diagnostic plan was altered, it was observed curve for GES was 0.79 (95% confidence interval,
that most patients demonstrated a decrease 0.73–0.84; P < 0.001). The sensitivity and specific-
(n = 93) than an increase (n = 52) in the inten- ity of the GES were 89% and 52%, respectively; the
sity of testing. In order to observe if the change in negative and positive predictive values were 96%
the diagnostic pattern had an effect on the patient and 24%, respectively. Thus, GES was found to be
outcomes, major adverse cardiac events (MACE) a highly significant predictor of obstructive CAD
 Pharmacogenomics and pharmacogenetics  223

by ROC analysis. Moreover, during the 6-month clinical practice if it continues to show good results
follow-up it was observed that the majority of among larger and more diverse cohorts (Vargas
the cardiovascular events and revascularizations et al., 2013).
(27 of 28%, 96%) occurred among patients with
GES > 15 (Thomas et al., 2013). PHARMACOGENOMICS AND
The two multicentric trials, the PREDICT and PHARMACOGENETICS
COMPASS, validated that GES could be used to
discriminate patients with and without CAD. There is an extensive variability in the response to
However, it is important to look at the reproduc- drugs between the patients in terms of both ben-
ibility of the technology, dependence of test per- efits and adverse effects leading to treatment fail-
formance on ethnicity, and the schedule of testing. ure or adverse reactions in most of the cases. The
Hence, a study was conducted where the testing massive development in the field of pharmacoge-
of more than 1500 patients from the PREDICT netics and pharmacogenomics has led to a greater
study was done, inclusive of their ethnicities and understanding of genetics for the prognosis and
the variation observed in the GES serial testing for treatment of CAD, but the implementation of this
one year among 192 patients from the COMPASS in daily clinical practice is limited. The genetic
study. From the PREDICT study, subjects who mutations that influence the drug metabolisms
completed the discovery and development phases can be explored in a targeted single-gene manner
were included in the present study. To determine (pharmacogenetics) or in a whole-genome manner
the sample and process stability, GES was mea- (pharmacogenomics). Pharmacogenomics aims to
sured for the validation set (n = 648) from the develop a rationale for the optimization of drug
PREDCIT study, whose samples were stored at therapy, with respect to patients’ genotype, so as
−20°C for approximately 5 years. There was no sig- to ensure the maximum efficacy with minimal side
nificant change in the test performance measured effects (Figure 11.5). Pharmacogenetics is defined
by ROC analysis, between the PREDICT and the as the science of identifying the genetic differences
present study (area under the curve [AUC] = 0.70 on the metabolic pathways that can affect indi-
for both, N = 501). The analysis of the 138 non- vidual’s response to drugs both in terms of both
Caucasian and 1364 Caucasian patients also therapy and adverse effects. This field combines
showed a similar performance (AUCs = 0.72 vs. pharmacology and genomics, so as to develop
0.70). Thus, the results demonstrated the sample medications that are designed to react with the
and GES analytical stability, but the extent of bio- patient’s genetic makeup. The factors that influence
logical variation over time on a per patient basis the response of patients to each drug include drug
was not analyzed. Hence, subjects who either had dose, adherence to a dosing regimen, drug–drug
invasive angiograms or research CT-angiograms interactions, genetic variation in genes encoding
to determine the coronary anatomy were included for drug metabolizing enzymes, and transporter
from the COMPASS study. The top four sites in the proteins. The genetic variation will have an impact
COMPASS study was approached and blood was on the pharmacokinetics of the drug through
drawn from these subjects approximately after one the regulation of drug absorption, distribution,
year of the study entry. A slight increase in the metabolism, or elimination. Since the drugs will
mean score from 15.9 to 17.3, corresponding to a have standard doses, individuals with this type of
2.5% increase in obstructive CAD likelihood, was genetic variation would experience adverse drug
observed. In conclusion, the changes in the car- reactions due to the concentration of drug that will
diovascular medications did not show significant either be toxic or nonefficacious (Chambliss and
change in GES score (Daniels et al., 2014). Chan, 2016).
The CAD test has been extensively evaluated in
the two multicentric trials. It is known that CAD Antiplatelet pharmacogenomics
has a destructive effect on the QOL and the cost for
diagnosing CAD is expensive. Thus, a simple GES The adjuvant therapy of aspirin and clopidogrel has
blood test might help and can be a valuable tool substantially reduced the risk of myocardial infarc-
to discriminate patients with obstructive CAD. tion, stent thrombosis, and death among patients
Hence, this test could be implemented in daily with acute coronary syndromes and those receiving
224  Precision medicine in coronary artery disease

Drug not
toxic and not
beneficial

Patient group
with same Drug not
Drug toxic
diagnosis and toxic and
but beneficial same beneficial
prescription

Drug toxic
and not
beneficial

Figure 11.5  Pharmacogenomics in precision medicine.

coronary stents (Yusuf et al., 2001; Steinhubl et al., and cilostazol therapy. The studies demonstrated
2002). However, in several studies it was observed that alternative therapies reduced the rate of HPPR
that thousands of patients inclusive of acute coro- among these patients (Jeong et al., 2009; Angiolillo
nary syndrome patients experienced genetic resis- et al., 2010). A systematic review and meta-analysis
tance to clopidogrel. The presence of the genetic was conducted to investigate the role of CYP2C19*2
mutation resulted in reduced clopidogrel active polymorphism in relation to adverse clinical out-
metabolite formation, attenuated antiplatelet effect, come in CAD patients who were undergoing clopi-
and there was a three-fold increase in the risk of dogrel treatment. The outcome measure MACE
stent thrombosis, myocardial infarction, and death was defined as any cardiovascular event (fatal and
(Mehta et al., 2001; Collet et al., 2009). nonfatal myocardial infarction, stroke, unstable
A GWAS was done to confirm the gene variant angina), death from cardiovascular causes, ischemic
CYP2C19*2 genotype to be associated with dimin- recurrences, stent thrombosis, and death from any
ished platelet response to clopidogrel treatment causes occurred during the follow-up. The results of
and poorer cardiovascular outcomes. The platelet the meta-analysis demonstrated that the CYP2C19*2
aggregation was measured at baseline and within polymorphism was associated with increased risk of
1 h after the last dose of clopidogrel on day 7 and MACE and stent thrombosis (Sofi et al., 2011).
about 400,000 SNPs was simultaneously assessed
using GWAS. The CYP2C19 variant demonstrated Statin pharmacogenetics
an increased risk (345%) for stent thrombosis,
along with risk of myocardial infarction and death The adherence of statin is often limited due to cer-
(Shuldiner et  al., 2009). Clopidogrel is one of the tain side effects. The SLCO1B1*5 variant is a major
highly prescribed drugs to prevent stent thrombo- risk factor for statin-specific side effects. However,
sis, myocardial infarction, and death among acute the effect of SLCO1B1*5 genotype-guided statin
coronary syndrome patients; a number of alternative therapy (GGST) is unknown. A pilot study was
therapies are available for individuals who are resis- done to see if SLCO1B1*5 therapy may improve
tant to the drug. The alternative therapies include perceptions and adherence to statins. The pri-
P2Y12 receptor blockade, prasugrel and cilostazol. mary outcome was the change in the Beliefs about
Studies have been conducted to evaluate the phar- Medications Questionnaire (BMQ) that measured
macodynamic response of switching acute coronary patients’ perceptions about needs for statins and
syndrome patients with high posttreatment plate- concerns about adverse effects. The GGST patients
let reactivity (HPPR) from clopidogrel to prasugrel demonstrated a higher “necessity” (pre: 15.1 ± 3.8
 Pharmacogenomics and pharmacogenetics  225

versus post: 15.6 ± 4.0, p = 0.24) and lower “con- A national, prospective, comparative effectiveness
cern” (pre: 15.2 ± 4.2 versus post: 14.7 ± 4.3, study was done to compare the 6-month incidence
p = 0.24) at four months compared to baseline. of hospitalization, including those due to bleeding
On examining the individual components of the or thromboembolism among patients receiving
necessity and concerns domains, the greatest warfarin genotyping versus a matched historical
change in BMQ was observed in the “need for statin control group. During the 6-month follow-up, the
to prevent sickness” (pre: 2.9 ± 0.9 versus post: patients who received the warfarin genotyping had
3.3 ± 0.9, p < 0.001) and the “concern for statin to 31% fewer hospitalizations overall and 28% fewer
disrupt life” (pre: 3.2 ± 1.4 versus post: 2.8 ± 1.2, hospitalizations for bleeding or thromboembo-
p = 0.006). The secondary outcome measure was lism compared to the control group (Epstein et al.,
to see if the GGST was associated with improved 2010). Another randomized and clinical effective-
statin adherence, statin prescribing behavior, and ness trial was conducted to compare the two phar-
LDL-c, compared to concurrent control subjects macogenetic algorithms and standard care for
receiving standard of care. It was observed that individualizing warfarin dosing (CoumaGen-II).
the GGST patients had more statin prescriptions The primary outcome measure was percentage
(p < 0.001), higher statin use (p < 0.001), and of out-of-range international normalized ratios
greater decrease in LDL-c (p = 0.059) during fol- (INR) at 1 and 3 months and percentage of time in
low-up (Li et al., 2014). therapeutic range. The study results demonstrated
that the genotype-guided therapy, regardless of
Warfarin pharmacogenetics the algorithm used, was superior to the standard
dosing. This was in terms of reduction in the per-
Warfarin still remains to be the mainstay for oral cent of out-of-range INRs and the percent of INRs
anticoagulation despite the recent approval of ≥4 or ≤1.5 at 3 months (Anderson et al., 2012). A
newer agents. However, warfarin also remains to recent meta-analysis was conducted to assess the
be one of the challenging medications to be man- effect of pharmacogenetics-based warfarin dos-
aged. Presently, warfarin ranks as one of the lead- ing in patients initiating warfarin therapy. The
ing drug-related causes of serious adverse events primary outcome measure was the time within
leading to hospitalizations. The challenges associ- the therapeutic range (TTR) and the secondary
ated with the usage of warfarin includes its narrow outcome measures were the time to maintenance
therapeutic index and wide interpatient variability dose and time to first INR, an INR greater than
in its dosage in order to achieve the required opti- 4, adverse events, major bleeding, thromboembo-
mal anticoagulation. The optimal dosage to achieve lism, and death from any cause. The results of the
therapeutic anticoagulation varies among patients study showed that the pharmacogenetics-based
from 0.5 to 11 mg/day or higher. The inappropriate dosing (PD) did not improve the TTR compared
dosage of warfarin leads to increased risk of bleed- to conventional dosing (CD). However, PD signifi-
ing and thromboembolic complications (Johnson cantly shortened the time to maintenance dose and
and Cavallari, 2013). Several GWAS and candi- the time to first therapeutic INR. Further, the PD
date gene data has shown that genotype contrib- significantly reduced the risk of adverse events and
utes to the interpatient variability in warfarin dose major bleeding, but there was no change in the per-
requirements (Aithal et al., 1999; Parra et al., 2015). centage of INR greater than 4, the risk of throm-
CYP2C9 and VKORC1, which encode for vitamin boembolic events and death from any cause (Shi
K epoxide reductase, are major genes that influ- et al., 2015).
ence the warfarin pharmacokinetics and phar- Presently the website WarfarinDosing.org is
macodynamics, respectively (Johnson et al., 2011; available to help clinicians to begin warfarin ther-
Giri et  al., 2014). Clinical trials were conducted apy by evaluating the required therapeutic dose
to observe the effect of genotype-guided warfarin among patients who are new to warfarin therapy.
dosing among patients, but the trial results were The evaluation is based on the patient’s clinical
inconsistent (Anderson et al., 2007; Caraco et al., and demographic factors. The genotypes of two
2008; Kimmel et al., 2013). Two large multicentric genes—cytochrome P450 2C9 (CYP2C9) and vita-
comparative studies were conducted to support the min K epoxide reductase (VKORC1)—are also
effectiveness of genotype-guided warfarin therapy. evaluated.
226  Precision medicine in coronary artery disease

OTHER EXAMPLES RNA molecules, which are associated with genes

and different diseases. Thus, drugs that can target
Currently, angiotensin-converting enzyme (ACE) the more specific disease can be manufactured.
inhibitors are recommended in clinical guidelines This will not only maximize the effectiveness of the
for the secondary prevention among patients with drug but can also minimize the damage caused to
stable CAD. However, the CAD population is com- the nearby cells.
prised of a heterogeneous group and the absolute
risk of cardiovascular complications varies among
the individuals. A study was done to integrate clini- Development of improved and safer
cal and pharmacogenetic determinants in an ulti- drugs
mate combined risk prediction model. The clinical,
More efficient and safer drugs can be prescribed
genetic, and outcome data were used from the 8726
for the patients. The physicians would be able to
stable CAD patients participating in the EUROPA/
analyze the patient’s genetic profile and prescribe
PERGENE trial of perindopril versus placebo.
the appropriate drugs from the onset itself. Thus,
Three SNPs (rs275651 and rs5182 in the angioten-
unnecessary therapy and adverse effects can be
sin II type I receptor gene and rs12050217 in the
avoided for the patient.
bradykinin type I receptor gene) were utilized to
construct the pharmacogenetic risk score (rang-
ing between 0 and 6 points). The primary end point Usage of appropriate doses of drug
was cardiovascular mortality, nonfatal myocardial
infarction, or resuscitated cardiac arrest. There The appurtenant dosage of drug can be prescribed
was a reduction in the absolute risk from 1.2% to to patients based on their individual genetic makeup
7.5% in the 73.5% of patients with PGXscore of 0 rather than the traditional method of treatment.
to 2. Thereby, the estimated annual numbers to be Thus, this approach would maximize the therapy
treated ranged from as low as 29 (clinical risk score value and decrease the possibility of drug overdose.
≥10 and PGXscore of 0) to 521 (clinical risk score
≤6 and PGXscore of 2). Further, the data recom- Precocious screening for disease
mended that long-term perindopril prescription
in patients with a PGX score of 0 to 2 is cost-effec- Having a prior knowledge about each person’s
tive. Hence, the combination of phenotypical and genetic code will enable the subjects to make the
genetic information demonstrated a very wide required changes in their lifestyle and environ-
range of gradients of absolute treatment benefit ment. This would avoid or cut short the severity
(Oemrawsingh et al., 2016). Studies were also done of genetic disease. Moreover, having an advanced
to determine whether ACE insertion/deletion (I/D) knowledge about the disease susceptibility will
polymorphism was associated with the severity provide careful monitoring and the treatment
CAD and its progression/regression in response could be provided at the appropriate stage to obtain
to fluvastatin therapy among the Lipoprotein and the maximum benefits of the therapy.
Coronary Atherosclerosis Study (LCAS) popula-
tion. The study results demonstrated that the ACE Reduction in the overall cost of
(I/D) polymorphism is associated with the response
health care
of plasma lipids and coronary atherosclerosis to
treatment with fluvastatin. The subjects with (D/D) The overall decrease in the following domains leads
mutation experienced a greater reduction in LDL- to an overall decrease in the cost of health care.
C, a higher rate of regression and a lower rate of
progression of CAD (Marian et al., 2000). 1. Decrease in the number of adverse drug
reactions.
Development of advanced 2 . Decrease in the number of failed drug trials.
medications 3. Decrease in the time taken for the approval of
the drug.
In the future, the pharmaceutical companies can 4. Decrease in the length of time the patients are
design drugs based on proteins, enzymes, and on medications.
 Proteomics 227

5. Decrease in the number of medications The advancements in the high throughput mass
the patient has to take to provide improved spectrometry for analyzing the clinical samples
therapy. have lead to the discovery of an outline of human
proteome. The collection of this data forms the base
Disadvantages for the comparative analysis of clinical proteomics.
The comparison of the patient sample with the
LIMITED ALTERNATIVE THERAPY existing database allows scientists to associate the
The few therapies that are approved can only be changes in proteome with particular disease states.
utilized for the treatment of a particular condition. Hence, the translation of the aforementioned find-
If the identified genetic mutations will prevent the ings in the clinic and hospital settings will allow
usage of drug, the patient might be left with no mass spectral analysis of patient proteomes. This
other substitute treatment. personalized proteomic analysis has the potential
to stimulate the diagnostic accuracy by reducing
COST CONCERNS the cost and time. Thus, this proteomic screening
The discovery and development of pharmacoge- will allow early detection of the presence and sever-
nomic drugs will be expensive and thus only a ity of disease, which allows for a quick response
small proportion of population might get the from medical personnel and improved patient
benefit. outcome (Duarte and Spencer, 2016). Many stud-
ies have been conducted to evaluate the levels of
EDUCATION OF HEALTH CARE PROVIDERS proteins of both normal and disease state for a suc-
Health care providers must have a better under- cessful diagnosis, prognosis, and therapy (Sabatine
standing regarding genomics. Hence, all the et al., 2005; Parguiña et al., 2010).
health care providers must be properly trained and
educated. Proteomic studies
In conclusion, the field of pharmacogenom-
ics is still scanty and its use is currently limited. A study was conducted to evaluate the modifica-
However, investigations are underway for new tions in the plasma protein map during unstable
approaches. Advancement in the field of genomic angina (UA) and AMI using proteomics. The pro-
science has assisted clinicians in prescribing tein plasma levels were quantified among patients
medications to patients based on their individual with AMI (n = 11) and UA (n = 9). The control
genetic data. Thus, the overall screening, thera- group was comprised of age-matched volunteers.
peutic, and management cost for patients can be The proteins that were analyzed included alpha-
reduced (Patil, 2015). 1-antitrypsin (AAT), apolipoprotein A-I, fibrino-
gen gamma chain, immunoglobulin gamma heavy
PROTEOMICS chain, and albumin. The main finding of this study
was that different AAT isoforms change in plasma
Proteomics is defined as a new biological approach during an acute coronary syndrome (ACS). Seven
to study the large-scale expression, function, and different AAT isoforms were seen in the plasma
interaction of the complement of proteins on a of control samples. However, the AAT isoform 1
large scale as a result of biological processes and was absent in the ACS samples. In the study, there
perturbations. In the recent years, novel biomark- was also a significant reduction of isoforms 5, 6,
ers have been evaluated for several cardiovascu- and 7 in the AMI samples when compared with
lar diseases such as CAD and heart failure. These UA samples. The fibrinogen gamma chain 1 and 2
biomarkers have shown promising results for the were increased in AMI patients compared to UA
prediction of cardiovascular outcomes. For these patients. Five apolipoprotein A-I isoforms were
biomarkers to become useful and be available for also identified, but these were reduced in plasma
personalized medicine, the following criteria need from AMI patients with respect to UA patients.
to be fulfilled. The biomarkers must be easy to The γ-immunoglobulin heavy chains were identi-
measure, must provide new information, and must fied and were found to be increased in the plasma
help clinicians for the management of patients among ACS patients. Thus, the proteomic analysis
(Stratz et al., 2012). will help in the mapping of the protein isoforms
228  Precision medicine in coronary artery disease

that are expressed in plasma during an ACS the treatment of each individual patient based on
(Mateos-Cáceres et al., 2004). their individual data. Even though studies have
Another study was done that combined phar- demonstrated that the application of personal-
macogenetics and iTRAQ-coupled LC-MS/MS ized medicine has provided positive results, its
pharmacoproteomics to analyze the plasma pro- usage in daily clinical practice comes with many
tein profiles in 53 patients. This approach identified challenges. Being a multidisciplinary approach to
a significantly higher level of transthyretin (TTR) science, it requires compliance with many com-
precursor among patients receiving low dose of ponents. The components are advancements in
warfarin but not in those on a high dose of war- technology, modifications in the health care infra-
farin. For the validation of the results, real-time structure and medical practice, improvements
RT-PCR, western blotting, and human IL-6 ELISA in the efficiency and quality of health care, and
assay were done. Thus, the study demonstrated ethical and legal issues regarding the storage of
that variations in TTR levels were associated with genomic information in medical records for a long
specific dosing regimens of warfarin. Moreover, in time (Lee et al., 2012).
addition to pharmacogenomic biomarkers pres- Another limitation of personalized medi-
ently emerging as an aid in the dosing of warfarin, cine is the translation of molecular and genomic
the pharmacoproteomics biomarkers such as TTR advances into clinical available means. The
may also be investigated for the dosing regimens of Human Genome Strategy Group (HGSG)
warfarin. Hence, the combined approach of phar- report and the whole genome sequencing by the
macogenomics and pharmacoproteomics can be Foundation for Genomics and Population Health
applied for other target-based therapies in match- (PHG Foundation) outlined many of the chal-
ing a particular marker in a subgroup of patients lenges involved in the implementation of genomics
(Saminathan et al., 2010). to test the health care system as a whole (Burton
et al., 2012). Although the research has been suc-
Limitations of proteomics cessful in the identification of the genetics that has
the ability to predict the susceptibility of the dis-
Even though the field of proteomics is important ease like the interaction of risk factors, the method
in cardiovascular medicine, there are several chal- utilized is not simple and small. It is necessary
lenges with respect to this field. The first being to develop basic scientific evidence that supports
the advantages and disadvantages of the differ- clinical interpretation of the genome based data.
ent technological platforms being utilized for the This in turn requires the standardization of the
evaluation of cardiovascular proteome. The second databases of the normal and pathogenic variations
challenge is to identify whether a combination or a linked to analytical tools (Lee et al., 2012).
one-platform approach is better. The third is with An appropriate development of laboratory setup
respect to the protocol, that is, whether a common is vital (i.e., biobanking, coordinated efforts, opera-
protocol should be for blood sample handling and tional and informatics support, standards, genomic
processing. The fourth challenge is the influence of technologies, core laboratories, economies of scale)
standard cardiovascular drugs on the blood pro- and the bioinformatical infrastructures is also nec-
teome. The fifth challenge is the method used to essary (Lee et al., 2012). Good clinical and practice
integrate current protein data with biomarker data guidelines are a must, as there must be a proper
from different investigators (Arab et al., 2006). understanding and transparency between the
physicians and patients. The information regard-
ing what and what will not be analyzed and why
CHALLENGES IN THE testing has to be done should be communicated to
IMPLEMENTATION OF PRECISION the patients. The confidentiality of the patient data
MEDICINE has to be protected during the usage of genetics
to predict the risks for the patients and relatives.
The main concept of precision medicine is to The clinicians should also be aware of the fact that
emphasize the scientific and technological inno- although some of the patients might be interested
vations so as to enable the physicians to alter or to know about their genetic inheritance, others may
modify the prognosis of disease, its diagnosis, and not be willing to know (Burton et al., 2012).
 References 229

One of the other challenges of the genomic CONCLUSION

approach is the readiness and willingness of phy-
sicians to the application of genomic technologies Precision medicine appears to have tremendous
for maximum effect so as to incorporate genomic potential in the field of cardiovascular medicine.
medicine into mainstream clinical practice. The The emergence and advances in precision medicine
clinicians who are working in the field of clinical will make it simpler for clinicians to tailor the drug
research should be aware of the advancements in treatment in accordance with the patient’s genetic,
the field of genomic medicine. Some of the clini- environmental, and lifestyle factors. The approaches
cians do not consider or recognize the field of that are currently being studied in precision medi-
genetics in clinical practice, as they perceive that cine include pharmacogenomics, pharmacogenet-
genetic conditions are rare and untreatable. Hence, ics, proteomics, and GES. The studies have shown
education and training for all physicians are inconsistent results leading to the requirement of
important. The collaboration between physicians more multicentric and well-powered studies. Thus,
and researchers will help in the understanding of precision medicine is an emerging field and further
inherited disorders within various areas of clinical research and efforts are required for it to be trans-
medicine (Burton et al., 2012). lated into daily clinical practice.

FUTURE CONSIDERATIONS REFERENCES

In conclusion, the large-scale genomic association Aithal GP, Day CP, Kesteven PJ, and Daly
studies have the ability to identify the genes and AK. Association of polymorphisms in the
genetic regions that can screen the risk factors for cytochrome P450 CYP2C9 with warfarin
disease. However, further research and efforts have dose requirement and risk of bleeding
to be taken to avoid the ascertainment bias among complications. Lancet (London, England)
the clinical research participants with wider inclu- 353(9154);1999:717–19.
sion criteria and follow-up to find the functional Alizadeh S, Djafarian K, Alizadeh H, Mohseni
variants in order to eliminate the need for tag SNPs R, and Shab-Bidar S. Common variants of
(Scheinfeldt et al., 2016). Although, there is incred- vitamin D receptor gene polymorphisms and
ible developments in the field of genomic medicine, susceptibility to coronary artery disease:
efforts are still required in the field of precision A systematic review and meta-analysis. J
medicine. Some of the technical challenges are Nutrigenet Nutrigenomics 10(1–2);9–2017:18.
data collection, processing, storing, and sharing. Anderson JL, Benjamin DH, Stevens SM et al.
Some are related to the broad scientific, clinical, Randomized trial of genotype-guided
social, and ethical ramifications (Chambliss and versus standard warfarin dosing in patients
Chan, 2016). initiating oral anticoagulation. Circulation
Although, data collection is taking place 116(22);2007:2563–70.
across the hospitals and communities in tradi- Anderson JL, Horne BD, Stevens SM et al.
tional ways; the data collection is also being done A randomized and clinical effectiveness trial
via social media analytics and wearable devices. comparing two pharmacogenetic algo-
The procedure for data collection has improved rithms and standard care for individualizing
through the adoption of systematic formats for warfarin dosing(CoumaGen-II). Circulation
electronic health records. However, improve- 125(16);2012:1997–2005.
ments need to be made in data processing and Angiolillo DJ, Saucedo JF, Deraad R,
integration. The storage and access to high vol- Frelinger AL, Gurbel PA, Costigan TM,
ume data is difficult, and efforts have to be taken Jakubowski JA, Ojeh CK, Effron MB, and
to enable the physicians, policy makers, and aca- SWAP Investigators. Increased platelet
demics to access the databases. During the devel- inhibition after switching from maintenance
opment of the precision approach, efforts have to clopidogrel to prasugrel in patients with
be taken to accommodate the ethical and trans- acute coronary syndromes: Results of the
parent use of data (Chambliss and Chan, 2016; SWAP (SWitching Anti Platelet) study. J Am
Laper et al., 2016). Coll Cardiol 56(13);2010:1017–23.
230  Precision medicine in coronary artery disease

Arab S, Gramolini AO, Ping P, Kislinger T, Stanley Einstein AJ, Moser KW, Thompson RC, Cerqueira
B, van Eyk J, Ouzounian M, MacLennan DH, MD, and Henzlova MJ. Radiation dose to
Emili A, and Liu PP. Cardiovascular pro- patients from cardiac diagnostic imaging.
teomics: Tools to develop novel biomarkers Circulation 116(11);2007:1290–1305.
and potential applications. J Am Coll Cardiol Einstein AJ, Weiner SD, Bernheim A, Kulon M,
48(9);2006:1733–41. Bokhari S, Johnson LL, Moses JW, and Balter
Brotman DJ, Golden SH, and Wittstein IS. The S. Multiple testing, cumulative radiation dose,
cardiovascular toll of stress. Lancet (London, and clinical indications in patients undergo-
England) 370(9592);2007:1089–1100. ing myocardial perfusion imaging. JAMA
Burton H, Cole T, and Lucassen AM. Genomic 304(19);2010:2137–44.
medicine: Challenges and opportunities Elashoff MR, Wingrove JA, Beineke P et al.
for physicians. Clin Med (Lond, England) Development of a blood-based gene
12(5);2012:416–19. expression algorithm for assessment of
Caraco Y, Blotnick S, and Muszkat M. CYP2C9 obstructive coronary artery disease in
Genotype-guided warfarin prescribing non-diabetic patients. BMC Med Genomics
enhances the efficacy and safety of anticoag- 4(March);2011:26.
ulation: A prospective randomized controlled Emami H, Takx RAP, Mayrhofer T, Janjua S, Park
study. Clin Pharm Ther 83(3);2008:460–70. J, Pursnani A, Tawakol A, Lu MT, Ferencik M,
Chambliss, AB, and Chan DW. Precision and Hoffmann U. Nonobstructive coronary
­medicine: From pharmacogenomics to artery disease by coronary CT angiography
­pharmacoproteomics. Clin Proteomics improves risk stratification and allocation of
13;2016:25. statin therapy. JACC Cardiovascular Imaging
Chow CK, Lock K, Teo K, Subramanian SV, McKee 10(9);2017:1031–8.
M, and Yusuf S. Environmental and soci- Epstein RS, Moyer TP, Aubert RE, O Kane DJ, Xia
etal influences acting on cardiovascular risk F, Verbrugge RR, Gage BF, and Teagarden JR.
factors and disease at a population level: A Warfarin genotyping reduces hospitalization
review. Int J Epidemiol 38(6);2009:1580–94. rates results from the MM-WES (Medco-Mayo
Collet J-P, Hulot J-S, Pena A et al. Cytochrome Warfarin Effectiveness Study). J Am Coll
P450 2C19 Polymorphism in young patients Cardiol 55(25);2010:2804–12.
treated with clopidogrel after myocardial Gao L, and Chen Y-D. Application of drug-
infarction: A cohort study. Lancet (London, coated balloon in coronary artery interven-
England) 373(9660);2009:309–17. tion: Challenges and opportunities. J Geriatr
Daniels SE, Beineke P, Rhees B, McPherson JA, Cardiol 13(11);2016:906–13.
Kraus WE, Thomas GS, and Rosenberg S. Giri AK, Khan NM, Grover S et al. Genetic
Biological and analytical stability of a periph- epidemiology of pharmacogenetic varia-
eral blood gene expression score for obstruc- tions in CYP2C9, CYP4F2 and VKORC1
tive coronary artery disease in the PREDICT genes associated with warfarin dosage in
and COMPASS studies. J Cardiovasc Transl the Indian population. Pharmacogenomics
Res 7(7);2014:615–22. 15(10);2014:1337–54.
Downs JR, Clearfield M, Weis S, Whitney E, Heidenreich PA, Trogdon JG, Khavjou OA et al.
Shapiro DR, Beere PA, Langendorfer A, Forecasting the future of cardiovascular
Stein EA, Kruyer W, and Gotto AM. Primary disease in the United States: A policy state-
prevention of acute coronary events with ment from the American Heart Association.
lovastatin in men and women with aver- Circulation 123(8);2011:933–44.
age cholesterol levels: Results of AFCAPS/ Herman L, Froelich J, Kanelos D, St Amant R,
TexCAPS. Air Force/Texas Coronary Yau M, Rhees B, Monane M, and McPherson
Atherosclerosis Prevention Study. JAMA J. Utility of a genomic-based, personal-
279(20);1998:1615–22. ized m
­ edicine test in patients present-
Duarte TT, and Spencer CT. Personalized pro- ing with symptoms suggesting coronary
teomics: The future of precision medicine. artery disease. J Am Board Fam Med
Proteomes 4(4);2016. 27(2);2014:258–67.
 References 231

Jeong Y-H, Lee S-W, Choi B-R, Kim I-S, Seo medicine research. Pac Symp Biocomput
M-K, Kwak CH, Hwang J-Y, and Park S-W. 21;2016:369–80.
Randomized comparison of adjunctive Lee M-S, Flammer AJ, Lerman LO, and Lerman
Cilostazol versus high maintenance dose clopi- A. Personalized medicine in cardiovascular
dogrel in patients with high post-treatment diseases. Korean Circ J 42(9);2012:583–91.
platelet reactivity: Results of the ACCEL- Li JH, Joy SV, Haga SB, Orlando LA, Kraus WE,
RESISTANCE (adjunctive Cilostazol versus Ginsburg GS, and Voora D. Genetically
high maintenance dose clopidogrel in patients guided statin therapy on statin perceptions,
with clopidogrel resistance) randomized study. adherence, and cholesterol lowering: A
J Am Coll Cardiol 53(13);2009:1101–9. pilot implementation study in primary care
Johnson JA, and Cavallari LH. Pharmacogenetics patients. Person Med J 4(2);2014:147–62.
and cardiovascular disease—Implications Loh JP, and Waksman R. Paclitaxel drug-coated
for personalized medicine. Pharmacol Rev balloons: A review of current status and
65;2013:987–1009. emerging applications in native coronary
Johnson JA, Gong L, Whirl-Carrillo M et al. artery De Novo lesions. JACC. Cardiovasc
Clinical pharmacogenetics implementa- Interv 5(10);2012:1001–12.
tion consortium guidelines for CYP2C9 and Marian AJ, Safavi F, Ferlic L, Dunn JK, Gotto AM,
VKORC1 genotypes and warfarin dosing. Clin and Ballantyne CM. Interactions between
Pharm Ther 90(4);2011:625–29. angiotensin-I converting enzyme insertion/
Kessler T, Vilne B, and Schunkert H. The impact of deletion polymorphism and response of
genome-wide association studies on the plasma lipids and coronary atherosclerosis
pathophysiology and therapy of cardiovascular to treatment with fluvastatin: The lipoprotein
disease. EMBO Mol Med 8(7);2016:688–701. and coronary atherosclerosis study. J Am Coll
Kimmel SE, French B, Kasner SE et al. A Cardiol 35(1);2000:89–95.
pharmacogenetic versus a clinical algo- Mateos-Cáceres PJ, García-Méndez A, Farré
rithm for warfarin dosing. N Engl J Med AL et al. Proteomic analysis of plasma from
369(24);2013:2283–93. patients during an acute coronary syndrome.
Kivimäki M, Nyberg ST, Fransson EI et al. J Am Coll Cardiol 44(8);2004:1578–83.
Associations of job strain and lifestyle risk McGrath S, and Ghersi D. Building towards preci-
factors with risk of coronary artery disease: A sion medicine: Empowering medical profes-
meta-analysis of individual participant data. sionals for the next revolution. BMC Medical
CMAJ 185(9);2013:763–69. Genomics 9(1);2016:23.
Klinkman MS, Stevens D, and Gorenflo DW. Mehramiz M, Ghasemi F, Esmaily H et al.
Episodes of care for chest pain: A preliminary Interaction between a variant of CDKN2A/B-
report from MIRNET. Michigan research net- Gene with lifestyle factors in determining dys-
work. J Fam Pract 38(4);1994:345–52. lipidemia and estimated cardiovascular risk: A
Kohsaka S, and Makaryus AN. Coronary angiog- step toward personalized nutrition. Clin Nutr
raphy using noninvasive imaging techniques (Edinburgh, Scotland) December 2016.
of cardiac CT and MRI. Curr Cardiol Rev Mehta SR, Yusuf S, Peters RJ et al. Effects of
4(4);2008:323–30. pretreatment with clopidogrel and aspirin
Krishnan MN. Coronary heart disease and risk followed by long-term therapy in patients
factors in India—On the brink of an epidemic? undergoing percutaneous coronary interven-
Indian Heart J 64(4);2012:364–67. tion: The PCI-CURE study. Lancet (London,
Kromhout D, Menotti A, Kesteloot H, and Sans S. England) 358(9281);2001:527–33.
Prevention of coronary heart disease by diet Oemrawsingh RM, Akkerhuis KM, Van Vark LC
and lifestyle: Evidence from prospective et al. Individualized angiotensin-converting
cross-cultural, cohort, and intervention stud- enzyme (ACE)-inhibitor therapy in stable
ies. Circulation 105(7);2002:893–98. coronary artery disease based on clinical
Laper SM, Restrepo NA, and Crawford DC. and pharmacogenetic determinants: The
The challenges in using electronic health PERindopril GENEtic(PERGENE) risk model.
records for pharmacogenomics and precision J Am Heart Assoc 5(3);2016:e002688.
232  Precision medicine in coronary artery disease

Parguiña AF, Parguiña AF, Grigorian-Shamajian L translating GWAS results to clinical care. Int J
et al. Proteins involved in platelet signaling Mol Sci 17(8);2016:1267.
are differentially regulated in acute coronary Shah SH, Arnett D, Houser SR, Ginsburg GS,
syndrome: A proteomic study. PLOS ONE MacRae C, Mital S, Loscalzo J, and Hall JL.
5(10);2010:e13404. Opportunities for the cardiovascular com-
Parra EJ, Botton MR, Perini JA et al. Genome- munity in the precision medicine initiative.
wide association study of warfarin main- Circulation 133(2);2016:226–31.
tenance dose in a Brazilian sample. Shepherd J, Cobbe SM, Ford I, Isles CG,
Pharmacogenomics 16(11);2015:1253–63. Lorimer AR, MacFarlane PW, McKillop
Patil J. Pharmacogenetics and pharma- JH, and Packard CJ. Prevention of coro-
cogenomics: A brief introduction. nary heart disease with pravastatin in men
J Pharmacovigilance 3(3);2015:e139. with hypercholesterolemia. N Engl J Med
doi:10.4172/2329-6887.1000e139. 333(20);1995:1301–7.
Postmus I, Trompet S, de Craen AJM, Buckley Shi C, Yan W, Wang G, Wang F, Li Q, and Lin
BM, Ford I, Stott DJ, Sattar N, Slagboom PE, N. Pharmacogenetics-based versus conven-
Westendorp RGJ, and Jukema JW. CSK9 SNP tional dosing of warfarin: A meta-analysis
rs11591147 is associated with low cholesterol of randomized controlled trials. PLOS ONE
levels but not with cognitive performance or 10(12);2015:e0144511.
noncardiovascular clinical events in an elderly Shuldiner AR, O’Connell JR, Bliden KP et al.
population. J Lipid Res 54(2);2013:561–66. Association of cytochrome P450 2C19
Rhee CM, Bhan I, Alexander EK, and Brunelli genotype with the antiplatelet effect and
SM. Association between iodinated contrast clinical efficacy of clopidogrel therapy. JAMA
media exposure and incident hyperthyroid- 302(8);2009:849–57.
ism and hypothyroidism. Arch Intern Med Śliwczynski A, and Orlewska E. Precision
172(2);2012:153–59. medicine for managing chronic diseases.
Rosenberg S, Froelich J, Kanelos D, St Amant R, Polskie Archiwum Medycyny Wewnetrznej
Yau M, Rhees B, Monane M, and McPherson J. 126(9);2016:681–87.
Multi-center validation of the diagnostic accu- Sofi F, Giusti B, Marcucci R, Gori AM, Abbate
racy of a blood-based gene expression test R, and Gensini GF. Cytochrome P450
for assessment of obstructive coronary artery 2C19*2 polymorphism and cardiovascular
disease in non-diabetic patients. J Am Board recurrences in patients taking clopidogrel:
Fam Med 27(2);2014:258–67. A meta-analysis. Pharmacogenomics J
Sabatine MS, Liu E, Morrow DA, Heller E, 11(3);2011:199–206.
McCarroll R, Wiegand R, Berriz G, Roth FP, Steinhubl SR, Berger PB, Mann JT, Fry ETA,
and Gerszten RE. Metabolomic identification DeLago A, Wilmer C, Topol EJ, and
of novel biomarkers of myocardial ischemia. CREDO Investigators. Early and sustained
Circulation 112(25);2005:3868–75. dual oral antiplatelet therapy follow-
Saminathan R, Bai J, Sadrolodabaee L, Karthik ing percutaneous coronary intervention:
GM, Singh O, Subramaniyan K, Ching CB, A randomized controlled trial. JAMA
Chen WN, and Chowbay B. VKORC1 phar- 288(19);2002:2411–20.
macogenetics and pharmacoproteomics in Stratz C, Amann M, Berg DD, Morrow DA,
patients on warfarin anticoagulant therapy: Neumann F-J, and Hochholzer W. Novel bio-
Transthyretin precursor as a potential bio- markers in cardiovascular disease: Research
marker. PLOS ONE 5(12);2010:e15064. tools or ready for personalized medicine?
Sayols-Baixeras S, Lluís-Ganella C, Lucas G, and Cardiol Rev 20(3); 2012:111–17.
Elosua R. Pathogenesis of coronary artery Tenconi MT, Romanelli C, Gigli F, Sottocornola
­disease: Focus on genetic risk factors and F, Laddomada MS, Roggi C, Devoti G,
identification of genetic variants. Appl Clin Gardinali P, and Research Group ATS-OB43
Genet 7;2014:15–32. of CNR. The relationship between educa-
Scheinfeldt LB, Schmidlen TJ, Gerry NP, tion and risk factors for coronary heart
and Christman MF. Challenges in disease: Epidemiological analysis from the
 References 233

nine communities study. Eur J Epidemiol symptomatic nondiabetic patients. PLoS Curr
8(6);1992:763–69. 5;2013.
Thomas GS, Voros S, McPherson JA et al. Wan Y-F, Ma X-L, Yuan C, Fei L, Yang J, and
A Blood-based gene expression test for Zhang J. Impact of daily lifestyle on
obstructive coronary artery disease tested coronary heart disease. Exp Ther Med
in symptomatic nondiabetic patients 10(3);2015:1115–20.
referred for myocardial perfusion imaging Yusuf S, Zhao F, Mehta SR, Chrolavicius S,
the COMPASS study. Circ Cardiovasc Genet Tognoni G, Fox KK, and Clopidogrel in
6(2);2013:154–62. Unstable Angina to Prevent Recurrent Events
Vargas J, Lima JA, Kraus WE, Douglas PS, and Trial Investigators. Effects of clopidogrel in
Rosenberg S. Use of the Corus® CAD gene addition to aspirin in patients with acute coro-
expression test for assessment of obstruc- nary syndromes without ST-segment eleva-
tive coronary artery disease likelihood in tion. N Engl J Med 345(7);2001:494–502.
 12
Precision medicine in stroke and other
related neurological diseases

ANJANA MUNSHI, VANDANA SHARMA, AND SULENA SINGH

Introduction 235 Epilepsy 237

Precision genetics for precision medicine Alzheimer’s disease 238
in stroke 236 Parkinson’s disease 238
Phenotype-based precision medicine in stroke 236 Harnessing the power of precision
Acute stroke imaging 237 medicine in stroke and other neurological
Precision medicine in other neurological diseases/conclusion 239
diseases 237 References 240

INTRODUCTION 2015). In the epileptic encephalopathies, trio

exome sequencing has identified that genes UDP-
The understanding of mechanisms of neuronal N-acetylglucosaminyltransferase subunit (ALG),
alteration and maintenance of their molecu- gamma-aminobutyric acid type a receptor β3 gene
lar signatures during disease progression is a (GABRB3), dynamin 1 (DNM1), hyperpolariza-
major requirement for clinically correct diagno- tion activated cyclic nucleotide gated potassium
sis of neurological disease. Numerous diagnostic channel 1 (HCN1), glutamate ionotropic recep-
investigations, including imaging techniques, are tor NMDA type subunit 2A (GRIN2A), gamma-
opted by concerned clinicians for prediction and aminobutyric acid type A receptor alpha1 subunit
analysis of the disease. Apart from these diagnos- (GABRA1), G protein subunit alpha O1 (GNAO1),
tic measures, genomic profiling is one of the cor- potassium sodium-activated channel subfamily T
nerstones of precision or personalized therapy, member 1 (KCNT1), sodium voltage-gated chan-
which not only forecasts the susceptibility to dis- nel alpha subunit 2 (SCN2A), sodium voltage-
ease but also predicts the best possible treatment gated channel alpha subunit 8 (SCN8A), and solute
for the individual patient. Many genes, includ- carrier family 35 member A2 (SLC35A2) are asso-
ing ATP binding cassette subfamily A member 7 ciated with epileptogenesis. Many of the proteins
(ABCA7), bridging integrator 1 (BIN1), comple- encoded by these genes have been found to be asso-
ment receptor 1 (CR1), phospholipase D3 gene ciated with synaptic transmission (Epi, 2015).
(PLD3), and phosphatidylinositol-binding clath- RNA sequencing (RNA-seq) provides the infor-
rin assembly protein gene (PICALM), have been mation regarding transcript structure, including
revealed to contribute toward the excess burden single base-pair resolution of transcript boundar-
of deleterious coding mutations in Alzheimer’s ies and boundaries between exons. The differen-
disease (Ma et  al., 2014; Jiang et  al., 2014; Tan tial transcription is examined using RNA-seq in
et al., 2014b; Cacace et al., 2015; Vardarajan et al., Alzheimer’s disease in brain tissue, which provides

235
236  Precision medicine in stroke and other related neurological diseases

the information about the protein coding genes, For replication, 47 affected sib pairs concordant
non-coding RNAs, and splicing. Networks pro- for stroke subtype and case-control series of an
duced may contribute to underlying neurological African American ancestry were sequenced. This
disease mechanisms. Several studies using animal was the first large-scale exome wide sequencing
models have observed the participation of the non- study which has identified protein-coding vari-
coding in the progression of Alzheimer’s disease ant in PDE4DIP phosphodiesterase 4D interacting
(Tan et  al., 2014a). Next-generation sequencing protein (PDE4DIP) (rs1778155) associated with an
can explain not only the complete molecular sig- intracellular signal transduction mechanism and
natures of cells by transcriptome analyses but also in acyl-CoA thioesterase (ACOT4) (rs35724886)
the cascade of events that induces or maintains with a fatty acid metabolism. Further confirmation
such signatures by epigenetic analyses. of PDE4DIP was found in affected sib-pair fami-
In order to investigate the gene–environment lies with large-vessel stroke subtype and in African
interaction, the science of epigenetics is preferred, Americans (Auer et al., 2015).
which also provides the information of disease Although this area in ischemic stroke is in its
pathogenesis. Epigenetic changes are a way in conceiving stage, once established in clinical prac-
which the environment can influence genetics, tice it may provide more precise treatment targets
bringing about potentially pathogenic alterations based on identification of various stroke-related
in the way genes are expressed. Epigenetics refers to protein-coding variants. Genotyping has already
changes in gene regulation brought about through begun to influence clinical practice with asso-
modifications to the DNA’s packaging proteins or ciation between drug and genetic variants being
the DNA molecules themselves without changing included in prescription guidelines for various
the underlying sequence (Mazzio and Soliman, medicines such as clopidogrel, metoprolol, and
2012). Several epigenetic studies have provided warfarin (Kaufman et  al., 2015). Clinical trials
compelling evidence for genome-wide significant involving gene expression profiling are underway
associations between differentially methylated to implement all these research-related develop-
regions and neurological diseases, and suggest that ments (NCT02014896). Studies based on animals
the correlation relationship could reflect a caus- to identify the role of a single gene that can control
ative one. Synergistically, all these findings could 80% variation in arterial collaterals in stroke are
lead researchers to focus on rare variants in neu- underway (Sealock et al., 2014).
rological diseases precisely, and as a consequence
of this treatment, regimens for individual patients PHENOTYPE-BASED PRECISION
may now be tailored and customized. MEDICINE IN STROKE
PRECISION GENETICS FOR Health care decisions are made on the basis of
PRECISION MEDICINE IN STROKE information provided by big data in an attempt
to use phenotypic precision to determine tar-
Genome-wide association studies have identified geted treatments for individuals from large data
many subtypes of specific single nucleotide poly- sets, such as administrative claims. For exam-
morphisms (SNPs) related with stroke (Traylor ple, the Observational Health Data Sciences and
et al., 2012; Holliday et al., 2012). Subsequent stud- Informatics collaborative is an international, mul-
ies have found genetic commonalities between tistakeholder partnership that uses large-scale ana-
stroke subtypes (Kaul and Munshi, 2012; Kilarski lytics to answer clinical questions from electronic
et  al., 2014; Rostanski and Marshall, 2016). health records and health insurance claims data.
Whole exome sequencing analysis carried out by Combining the data from various sources limits
the National Heart, Lung, and Blood Institute biases that are inherent in homogeneously sourced
Exome Sequencing Project analyzed  approxi- data (Rostanski and Marshall, 2016).
mately 6000 participants from different cohorts Electronic health records contain huge amounts
of European and African ancestry. In this study, of disease-related useful information that could
365 cases of ischemic stroke (small-vessel and potentially be used for building models for predict-
large-vessel subtypes) and around 809 controls ing onset of diseases and its models for predicting
belonging to European ancestry were sequenced. treatment. In a study, a stroke-prediction model
 Precision medicine in other neurological diseases  237

applied advanced machine-learning techniques to personalized patient-centered regime leading to

aggregated health records in patients with atrial precise treatment (de Villiers et al., 2015).
fibrillation. The model outperformed the CHADS2 The complete application of knowledge pro-
score (a method or rule used for clinical evalua- vided by exome sequencing, epigenetics, and other
tion of stroke patients with nonrheumatic atrial genetic advances may offer new insights into the
fibrillation) and helped in predicting stroke, thus complex regulation of gene expression in stroke.
providing the useful information for clinical inter- A higher rate of precision in phenotype descrip-
pretation (Shahn et al., 2015). Therefore, both the tion through the use of big data and the addition
medical claims and large health record data keep of advanced imaging state data promise to comple-
the potential to improve prediction accuracy when ment advances in genomics to provide precision
compared to individual patient-level prediction medicine for acute stroke treatment, ischemic
models. stroke and its subtypes, cryptogenic stroke, and
Meta-analyses have been used to convert diver- secondary stroke prevention.
gent clinical trial results into cohesive clinical
algorithms. New and updated statistical meth- PRECISION MEDICINE IN OTHER
ods used by the National Institutes of Health for NEUROLOGICAL DISEASES
the standardization of clinical data elements now
allows large-scale data analysis across many tri- Epilepsy
als using patient-level data pooling (Saver et  al.,
2012). The increased number of patients with Epilepsy is a neurological disorder characterized
consistent data creates greater power to examine by fits and seizures due to abnormal activity and
subgroups not possible in individual trials and discharge of neurons in brain. The management
enables focus on patient-centered outcome vari- of epilepsy needs comprehensive care in order to
ables. In an exploratory study, clinical and infarct address the efficacy, side effects and quality of life
location data was used from 131 patients with issues (Jin et  al., 2016). The mainstay of epilepsy
cryptogenic stroke combined with genetic profiles is drug therapy with antiepileptic drugs (AEDs).
derived from patients with a known stroke subtype However, there are almost two-thirds of patients
(Jickling et al., 2012). In this study the patients pre- responding to monotherapy or rational polyther-
dicted with cardio embolic etiology were found apy who achieve complete seizure control without
to have more myocardial infarctions and higher major side effects (Jin et al., 2016). The therapeutic
CHA2DS2-VASc scores when compared to those area in epilepsy is affected mainly because one-
predicted with a large or small vessel etiology. third of the patients with epilepsy are refractory
even in the context of introduction of new phar-
Acute stroke imaging macological molecules/medicines, also known as
pharmacoresistant epilepsy. Therefore, surgery in
Radiological neuroimaging is a noninvasive a subset of patients is the only option left.
approach used to observe the physical manifes- The role of precision medicine comes in diag-
tation of the disease or phenotype in order to nosis, treatment, and presurgical evaluation,
understand the neurological diseases. In cases which requires a multimodality approach where
of acute stroke, the use of advanced perfusion each provides unique, accurate, and complemen-
and collateral imaging has added another level tary information about the individual patient.
of phenotypic precision medicine. Newer clini- Voxel-based morphometric magnetic resonance
cal trials are underway that involve unknown imaging analysis, multimodality techniques, com-
time of stroke onset and delayed initiation of puter-aided subtraction ictal single-photon emis-
medical treatment and are using perfusion and sion computed tomography (SPECT), and many
collateral imaging to provide thrombolytic treat- other novel techniques have improved our ability
ment to a targeted subset of stroke patients. to identify subtle structural and metabolic lesions
Moreover, a combination of viscoelastic meth- causing focal seizure (Jin et al., 2016). Advances in
odologies, including thromboelastography and the field of epilepsy surgery and precision medi-
scanning electron microscopy, have been vali- cine have been reviewed by Jin et al. (2016). Using
dated as useful techniques for stroke patients in a all the updated knowledge and various modalities,
238  Precision medicine in stroke and other related neurological diseases

the clinical outcome of epileptic patients will defi- in Asymptomatic Alzheimer’s Disease (A4) trial
nitely improve reducing disease-associated mor- are now incorporating information of underlying
bidity and mortality. We are on the edge of using pathological mechanisms, selecting patients con-
the best surgical diagnosis and treatment of medi- sidered most likely to respond to the anticipated
cal refractory epilepsy for individual patients thus action of the therapeutic tested. The results of tri-
promising a better health care outcome. als are still awaited (Sperling et al., 2011; Moulder
et al., 2013).
Alzheimer’s disease Used for decades in the management of a few
rare diseases and now broadly applied in cancer
Alzheimer’s disease is a neurodegenerative dis- care, the precision medicine approach is now step-
order and is the most common form of demen- ping toward Alzheimer’s disease with the use of
tia or cognitive impairment in ageing societies. genetic variants and neuroimaging techniques.
Currently available drugs have slight influence on Although successful application of preci-
disease severity and progression. Interventions sion medicine into diagnosis and treatment of
used to prevent or decelerate disease progres- Alzheimer’s disease needs an extensive frame-
sion would reduce not only the individual suffer- work to identify risk groups, and the underlying
ing and would significantly relieve public health pathophysiological and molecular processes. This
burden (Reitz, 2016). According to an estimate if will develop new interventions, with the signifi-
we delay the onset of Alzheimer’s by 5 years this cant involvement of a huge network of biologists,
would reduce the prevalence by ∼50% (Reitz, physicians, technology developers, data scientists,
2016). Alzheimer’s is a pathologically and clinically geneticists, patient groups, stakeholders, and pol-
heterogeneous neurodegenerative disease. The dis- icy makers. It is anticipated that this is only the
ease-associated key pathological manifestations in initiation of a broad precision medicine approach
the brain involve intracellular deposits of hyper- targeting the clinical, biological, and molecular
phosphorylated tau protein in the form of neuro- complexity of Alzheimer’s disease. With the use of
fibrillary tangles and extracellular β-amyloid (Aβ) precision medicine in the treatment of Alzheimer’s
protein in diffuse and neurotic plaques, which are disease, we can hope for better diagnosis, and
generated via sequential cleavage of the amyloid development of safe, effective drugs or drug com-
precursor protein (APP) through β- (BACE1) and bination cocktails to improve clinical outcome of
ϒ-secretase (Reitz, 2016). The clinical complexity the disease.
has been reflected in several genetic studies where
significant progress has been made in identifying Parkinson’s disease
the underlying genetic variants, with mapping of
27 susceptibility loci (Ertekin-Taner, 2007; Harold Parkinson’s disease is a neurodegenerative disease,
et al., 2009; Kim et al., 2009; Seshadri et al., 2010; with a complex pathogenesis involving an interac-
Hollingworth et al., 2009; Naj et al., 2011; Cruchaga tion of several genetic variants, each with specific,
et  al., 2012; Guerreiro et  al., 2013; Jonsson et  al., however, minor impact, and a number of environ-
2013; Lambert et al., 2013; Logue et al., 2014; Tosto mental factors (Kieburtz and Wunderle, 2013).
et al., 2015). This disease is associated with devastating chronic
The susceptibility risk and molecular profiles complications involving motor symptoms such
of patients with Alzheimer’s disease have shown as slowness of movement, muscle rigidity, rest-
vast variation, and therefore, categorizing patients ing tremor, and, in some cases, posture and gait
on the basis of different risk or molecular profiles problems. Nonmotor symptoms include sleep dis-
into a single entity might lead to hiding small sub- orders, cognitive dysfunction, and digestive prob-
groups potentially responsive to a certain treat- lems. Current treatment of Parkinson’s is focused
ment regime and nonresponsive to another (Reitz, mainly on restoring movement by replacing lost
2016). The results obtained from many clinical tri- neurotransmitter dopamine, with the help of med-
als to date are not satisfactory and a first round of icines. However, these medicines have become less
ongoing trials including the Dominantly Inherited effective over time and only target a small aspect
Alzheimer Network (DIAN), the Alzheimer’s of the underlying disease pathology with reduced
Prevention Initiative, and Anti-Amyloid Treatment therapeutic efficacy. Therefore, a convergence of
 Harnessing the power of precision medicine in stroke and other neurological diseases/conclusion  239

applications of genomics, epigenomics, neuroim- provide more benefit to patients in the disease
aging, and metabolomics, and their accurate inter- course of time with better clinical outcome.
pretation is required to understand the underlying
disease pathology and to manage the disease in a HARNESSING THE POWER OF
best possible way. PRECISION MEDICINE IN STROKE
Numerous genetic variants and biomark- AND OTHER NEUROLOGICAL
ers have been identified to be associated with the DISEASES/CONCLUSION
development of Parkinson’s disease (Korczyn and
Hassin-Baer, 2015). The causative mutation in the The development of cost-effective advanced
α-synuclein was identified almost two decades genomic technologies and the increasing number
back, the gene encoding a major component of of targeted therapies has placed clinical practice in
Lewy bodies and Lewy neurites, supposed to be a unique and enviable position. Health care provid-
the hallmarks of Parkinson’s disease pathology ers have to deeply analyze the patient’s sample to
(Polymeropoulos et  al., 1997; Bu et  al., 2016). To know the identified genetic variants, altered protein
date, more than 16 loci and 11 associated genes have structure fitting/unfitting the drug molecule and
been recognized, which help in our understanding then select a targeted therapy to treat the patient’s
of the pathogenesis and clinical heterogeneity of disease individually rather than only going for his-
disease (Corti et al., 2011). Besides the genomic pro- tological, anatomical, and routine clinical exami-
filing, molecular imaging allows a window into the nation. However, there are many issues in the
pathophysiology of Parkinson’s in vivo, and mea- implementation of precision medicine for both the
sures the severity and progression of the disease patient and clinician, for example, interpretation of
(Bu et  al., 2016). Positron emission tomography genomic analysis by geneticists, designing targeted
or single-photon emission computed tomography therapies for patients with refractory disease, and
imaging, integrated with other individual informa- the prediction of clinical outcomes with the help
tion such as genomic profiling, can help the clini- of broad application of omics in medicine, which
cian to design new and tailored interventions for are just in infancy. In addition, the complexities of
individual patients in contrast to traditional meth- heterogeneous neurological disorders (including
ods of prescribing medicine. For instance, molecu- stroke) have to be taken into consideration before
lar imaging in the asymptomatic carriers with gene providing precision medicine.
mutation of Parkinson’s disease can inform about In spite of these issues, and in light of the breath-
the disease progression at its preclinical stage, taking responses that can have major impact on
therefore, predicting the response of neuroprotec- patients suffering from neurological diseases, a
tive treatments at the right time in right individuals tidal wave in the form of precision medicine has
in the future at the right time (Bu et al., 2016). been initiated and supported by many countries
New tracers used in the radio imaging of (in the United States by the National Institutes of
α-synuclein are under development and are of Health). Likewise, pharmaceutical companies and
potential importance in terms of diagnostic, prog- other industry stakeholders are developing novel
nostic, and monitoring biomarkers of Parkinson’s precise and tailored targeted therapies and diag-
disease. Other measures that can be of ultimate nostic kits to support precision medicine. This will
importance include deep brain stimulation in bypass the major adverse drug reactions, thereby
order to block abnormal neural signals that may informing about the responsiveness or nonrespon-
lead to Parkinson’s disease. The precision medicine siveness of the treatment. Moreover a customized
in Parkinson’s disease has been reviewed in detail treatment will be available for individual patients.
by Bu et al. (2016). Numerous clinical trials are underway that will
The concept of precision medicine is based provide necessary information regarding metrics,
on individualized evaluation and multidimen- such as survival and health care costs, efficiency
sional information about omics, molecular imag- of techniques used for diagnosis associated with
ing, deep brain stimulation surgery, and wearable genomics/proteomics, and transcriptomics for
sensors leading to tailored therapeutic strategies designing precision medicine. The near future will
in Parkinson’s disease. Precision medicine will accommodate the precision medicine paradigm
become a promising and feasible strategy and will in the standard of clinical care for patients with
240  Precision medicine in stroke and other related neurological diseases

neurological diseases including stroke as it has Hollingworth P, Harold D, Sims R et al. Common
already been initiated in oncology practice. variants at ABCA7, MS4A6A/MS4A4E,
EPHA1, CD33 and CD2AP are associ-
REFERENCES ated with Alzheimer’s disease. Nat Genet
43;2009:429–35.
Auer PL, Nalls M, Meschia JF et al. Rare and Jiang T, Yu JT, Tan MS et al. Genetic varia-
coding region genetic variants associ- tion in PICALM and Alzheimer’s disease
ated with risk of ischemic stroke: The risk in Han  Chinese. Neurobiol Aging
NHLBIExome Sequence Project. JAMA Neurol 35;2014:934.e1–3.
72(7);2015:781–8. Jickling GC, Stamova B, Ander BP et al.
Bu LL, Yang K, Xiong WX, Liu FT, Anderson B, Prediction of cardioembolic, arterial, and
Wang Y, Wang J. Toward precision medi- lacunar causes of cryptogenic stroke by
cine in Parkinson’s disease. Ann Transl Med gene expression and infarct location. Stroke
4(2);2016:26. 43(8);2012:2036–41.
Cacace R, Bossche VT, Engelborghs S et al. Jin P, Wu D, Li X, Ren L, Wang Y. Towards preci-
Rare variants in PLD3 do not affect risk sion medicine in epilepsy surgery. Ann Transl
for early-onset Alzheimer disease in a Med 4(6);2016:104.
European consortium cohort. Hum Mutat Jonsson T, Stefansson H, Steinberg S et al.
36;2015:1226–35. Variant of TREM2 associated with the
Corti O, Lesage S, Brice A. What genet- risk of Alzheimer’s disease. N Engl J Med
ics tells us about the causes and mecha- 368;2013:107–16.
nisms of Parkinson’s disease. Physiol Rev Kaufman AL, Spitz J, Jacobs M et al. Evidence
91;2011:1161–218. for clinical implementation of pharmacoge-
Cruchaga C, Haller G, Chakraverty S et al. Rare nomics in cardiac drugs. Mayo Clin Proc
variants in APP, PSEN1 and PSEN2 increase 90(6);2015:716–29.
risk for AD in late-onset Alzheimer’s disease Kaul S, Munshi A. Genetic basis of stroke:
families. PLOS ONE 7;2012:e31039. An Indian perspective. Neurol India
de Villiers S, Swanepoel A, Bester J et al. Novel 60;2012:498–03.
diagnostic and monitoring tools in stroke: Kieburtz K, Wunderle KB. Parkinson’s disease:
An individualized patient-centered precision Evidence for environmental risk factors. Mov
medicine approach. J Atheroscler Thromb Disord 28;2013:8–13.
23(5);2015. [Epub ahead of print]. doi:10.5551/ Kilarski LL, Achterberg S, Devan WJ et al. Meta-
jat.32748. analysis in more than 17,900 cases of ischemic
Epi PM Consortium. A roadmap for precision stroke reveals an ovel association at 12q24.12.
medicine in the epilepsies. Lancet Neurol Neurology 83(8);2014:678–85.
14;2015:1219–28. Kim M, Suh J, Romano D et al. Potential late-
Ertekin-Taner N. Genetics of Alzheimer’s dis- onset Alzheimer’s disease-associated
ease: A centennial review. Neurol Clin mutations in the ADAM10 gene attenuate
25;2007:611–67. {alpha}-secretase activity. Hum Mol Genet
Guerreiro R, Wojtas A, Bras J et al. TREM2 vari- 18;2009:3987–96.
ants in Alzheimer’s disease. N Engl J Med Korczyn AD, Hassin-Baer S. Can the disease
368;2013:117–27. course in Parkinson’s disease be slowed?
Harold D, Abraham R, Hollingworth P et al. BMC Medicine 13;2015:295.
Genome-wide association study identi- Lambert JC, Ibrahim-Verbaas CA, Harold D et al.
fies variants at CLU and PICALM associ- Meta-analysis of 74,046 individuals identi-
ated with Alzheimer’s disease. Nat Genet fies 11 new susceptibility loci for Alzheimer’s
41;2009:1088–93. disease. Nat Genet 45;2013:1452–8.
Holliday EG, Maguire JM, Evans TJ et al. Logue MW, Schu M, Vardarajan BN et al. Two
Common variants at 6p21.1 are associated rare AKAP9 variants are associated with
with large artery atherosclerotic stroke. Nat Alzheimer’s disease in African Americans.
Genet 44(10);2012:1147–51. Alzheimers Dement 10;2014:609–618.e11.
 References 241

Ma XY, Yu JT, Tan MS et al. Missense variants Seshadri S, Fitzpatrick AL, Ikram MA et al.
in CR1 are associated with increased risk of Genome-wide analysis of genetic loci
Alzheimer’ disease in Han Chinese. Neurobiol associated with Alzheimer disease. JAMA
Aging 35;2014:443.e17–21. 303;2010:1832–40.
Mazzio EA, Soliman KF. Basic concepts of epi- Shahn Z, Ryan P, Madigan D. Predicting health
genetics: Impact of environmental signals on outcomes from high-dimensional longitudinal
gene expression. Epigenetics 7;2012:119–30. health histories using relational random for-
Moulder KL, Snider BJ, Mills SL et al. Dominantly ests. Stat Analysis Data Mining ASA Data Sci J
inherited Alzheimer network: Facilitating 8(2);2015:128–36.
research and clinical trials. Alzheimers Res Sperling RA, Aisen PS, Beckett LA et al. Toward
Ther 5;2013:48. defining the preclinical stages of Alzheimer’s
Naj AC, Jun G, Beecham GW et al. Common disease: Recommendations from the National
variants at MS4A4/MS4A6E, CD2AP, Institute on Aging Alzheimer’s Association
CD33 and EPHA1 are associated with late- workgroups on diagnostic guidelines for
onset Alzheimer’s disease. Nat Genet Alzheimer’s disease. Alzheimer’s Dement
43;2011:436–41. 7;2011:280–92.
Polymeropoulos MH, Lavedan C, Leroy E et al. Tan L, Yu JT, Tan MS et al. xGenome-wide serum
Mutation in the alpha-synuclein gene iden- microRNA expression profiling identifies
tified in families with Parkinson’s disease. serum biomarkers for Alzheimer’s disease.
Science 276;1997:2045–7. J Alzheimers Dis 40;2014a:1017–27.
Reitz C. Toward precision medicine in Alzheimer’s Tan MS, Yu JT, Jiang T et al. Genetic variation
disease. Ann Transl Med 4(6);2016:107. in BIN1 gene and Alzheimer’s disease risk in
Rostanski SK, Marshall RS. Precision medi- Han Chinese individuals. Neurobiol Aging
cine for ischemic stroke. JAMA Neurol 35;2014b:1781.e1–8.
73(7);2016:773–4. Tosto G, Fu H, Vardarajan BN et al. F-box/LRR-
Saver JL, Warach S, Janis S et al. Standardizing repeat protein 7 is genetically associated with
the structure of stroke clinical and epidemio- Alzheimer’s disease. Ann Clin Transl Neurol
logic research data: The National Institute of 2;2015:810–20.
Neurological Disorders and Stroke (NINDS) Traylor M, Farrall M, Holliday EG et al. Genetic
Stroke Common Data Element (CDE) project. risk factors for ischaemic stroke and its
Stroke 43(4);2012:967–73. subtypes (the METASTROKE collaboration):
Sealock R, Zhang H, Lucitti JL, Moore SM, A meta-analysis of genome-wide association
Faber JE. Congenic fine-mapping identi- studies. Lancet Neurol 11(11);2012:951–62.
fies a major causal locus for variation in the Vardarajan BN, Ghani M, Kahn A et al. Rare cod-
native collateral circulation and ischemic ing mutations identified by sequencing of
injury in brain and lower extremity. Circ Res Alzheimer disease genome-wide association
114(4);2014:660–71. studies loci. Ann Neurol 78(3);2015:487–98.
 13
Precision medicine in osteoporosis
and bone diseases

FATMANUR HACIEVLIYAGIL KAZANCI, FATIH KAZANCI,

M. RAMAZAN YIGITOGLU, AND MEHMET GUNDUZ

Introduction 243 Importance of lifestyle, environmental,

Bone 243 and genetic factors in osteoporosis 247
Osteoporosis 243 Genetics of osteoporosis 247
Epidemiology 244 Epigenetic mechanisms 248
Morbidity and mortality rate 244 Single-gene disorders of bone 248
Diagnosis 245 Monogenic diseases with high BMD 251
Prevention and treatment 245 Monogenic diseases with decreased BMD 254
Emerging treatments for osteoporosis 246 Conclusion, challenges, and future directions 254
References 255

INTRODUCTION resorption exceeds neoformation contributes to a

decrease in bone volume and mineralization, loss
Bone of trabeculae, deterioration of trabecular connec-
tivity, and the formation of cavities and perfora-
Bone is a mineralized connective tissue that exerts tions, all of which leads to osteoporosis (Hendrickx
important functions in the body. Despite its inert et al., 2015). The rate of bone loss throughout life is
appearance, bone tissue is continuously remod- dependent on a person’s lifestyle and genetic back-
eled through the coordinated actions of bone cells ground, which influence the geometric architec-
(Datta et al., 2008). Osteoblasts, bone lining cells, ture of the bone.
osteocytes, and osteoclasts are the cells of the bone.
After bones have completed their longitudinal
growth, almost 10% of the bone is remodeled each Osteoporosis
year for mineral homeostasis and adaptation to
mechanical load (Divieti Pajevic, 2013; Hendrickx Osteoporosis is the most common metabolic disor-
et al., 2015). der of the elderly (Mitchell and Streeten, 2013). By
The remodeling process is under the control of the definition of the World Health Organization
local and systemic factors (Florencio-Silva et  al., (WHO), osteoporosis is characterized by reduced
2015). A disequilibrium between bone resorption bone mass, impaired bone quality, and predispo-
and formation (turnover) leads to a deterioration sition to fractures, which is diagnosed if the bone
in bone health. An increase in bone turnover where mineral density (BMD) measured by dual x-ray

243
244  Precision medicine in osteoporosis and bone diseases

absorptiometry (DXA) is more than 2.5 standard et  al., 2014), and osteoporosis-related fracture is
deviations (SD) below the sex-matched young estimated to happen every 3 seconds (International
adult mean and/or on the base of fractures occur- Osteoporosis Foundation, 2014).
ring without significant trauma (NIH Consensus The global life expectancy is increasing steadily
Development Panel on Osteoporosis Prevention, and the number of elderly individuals is rising
Diagnosis, and Therapy, 2001). Osteoporosis in every geographic region. By the year 2050,
may be classified as either primary or secondary. the population of individuals aged 65 years and
Primary osteoporosis is bone loss associated with over is expected to increase from 323 million to
the aging process and secondary osteoporosis is 1.555 billion (Harvey et  al., 2010). These demo-
bone loss caused by a variety of chronic diseases, graphic changes could lead to an increase in the
medications, and nutritional deficiencies. number of osteoporotic fractures occurring world-
wide among elderly individuals. The numbers of
EPIDEMIOLOGY hip fractures worldwide are projected to increase
Osteoporosis affects one-third of women and one from 1.7 million in 1990 to 6.3 million in 2050 and
out of eight men over the age of 50 (Li et al., 2010). the annual cost of hip fractures is forecast to rise
Most patients with osteoporosis are asymptom- to $131.5 billion by 2050 (at a cost of $21,000 per
atic, which makes epidemiological research espe- patient) (Johnell, 1997).
cially difficult. However, it is estimated that over Although hip fractures are a small proportion
200 million people worldwide have osteoporosis of all osteoporosis-related fractures, they are the
and the prevalence of osteoporosis is continuing most serious of all fractures, causing a high bur-
to expand with the increase in life expectancy den of morbidity, mortality, and health care costs.
(Reginster and Burlet, 2006). Postmenopausal Approximately 1.6 million hip fractures occur
women are associated with the highest risk of worldwide each year and 5%–10% of patients
morbidity. In the United States and the European experience a recurrent hip fracture (International
Union, about 30% of all postmenopausal women Osteoporosis Foundation, 2014). In a systematic
have osteoporosis, and it has been predicted that review of 22 studies, excess mortality during the
more than 40% of them will suffer one or more first year after a hip fracture ranged from 8.4%
fragility fractures during their remaining lifetime to 36.0%, and the risk of mortality following hip
(Melton et al., 1992). fracture was estimated to be at least twice as high
as that for age-matched control individuals from
MORBIDITY AND MORTALITY RATE the general population (Sattui and Saag, 2014). The
The main clinical outcome of the disease is bone risk of death related to hip fracture of a 50-year-
fracture. Hip, spine, and distal forearm fractures old woman is equivalent to her risk of death from
are the most serious fractures associated with breast cancer and 4 times higher than that from
mortality, morbidity, and economic cost (Harvey endometrial cancer (International Osteoporosis
et al., 2010). In the European Union, in 2000, the Foundation, 2014). In 2010, the number of deaths
number of osteoporotic fractures was estimated causally related to osteoporotic fractures in Europe
at 3.79 million and the direct costs of osteoporotic was 43,000 and almost 80% of these were due to
fractures to the health services were estimated at hip or vertebral fractures (Kanis et al., 2012).
€32 billion (Reginster and Burlet, 2006). In 2004, it Vertebral fractures are one of the most fre-
was forecast that there were approximately 10 mil- quent complications of osteoporosis; however,
lion Americans over the age of 50 years with osteo- only one-third to one-quarter of these fractures
porosis, and an additional 34 million Americans are clinically diagnosed and hospitalization rates
were at risk of the disease. Osteoporosis is directly are low (Sattui and Saag, 2014). In the European
responsible for ∼1.5 million fractures annually, Prospective Osteoporosis Study (EPOS), age-
with an estimated health care cost of $17 billion in standardized incidence of vertebral fracture was
the United States alone (Gass and Dawson-Hughes, reported as 10.7 per 1,000 person-years in women
2006). In 2014, it is reported that up to 49 million and 5.7 per 1,000 person-years in men (Felsenberg
individuals met the WHO osteoporosis criteria et al., 2002). The excess mortality 1 year after ver-
in a number of industrialized countries in North tebral fractures varies considerably, ranging from
America, Europe, Japan, and Australia (Wade 1.9% to 42.0% (Sattui and Saag, 2014).
 Introduction 245

DIAGNOSIS dose response effects for several of the clinical risk

The diagnosis of osteoporosis relies on the quan- factors and the exclusion of some common nonhip
titative assessment of BMD. The BMD value is the nonvertebral fractures from its output.
amount of bone mass per unit volume (volumet- Biochemical markers of bone turnover have
ric density), or per unit area (areal density), and is the potential to serve as useful tools in osteopo-
currently considered the best predictor of osteo- rosis management, but biological and analytical
porotic fractures. For every standard deviation variability of them hinders their use (Hlaing and
decrease in BMD, the fracture risk increases about Compston, 2014). They are usually used in the
twofold. BMD is expressed in grams per square monitoring of osteoporosis therapy (Wei et  al.,
centimeter (g/cm2), but values differ between 2016), because changes in BMD take longer to
machines and manufacturers. Therefore, T-scores occur than bone turnover markers (Hlaing and
and Z-scores are used. T-scores can be defined Compston, 2014).
as the number of standard deviations from the
mean bone density values in normal sex-matched PREVENTION AND TREATMENT
young adults (Hlaing and Compston, 2014). The Resistance and weight-bearing exercises can
T-score is used to make a diagnosis of osteopo- increase muscle mass and BMD temporarily
rosis in postmenopausal women and in men age (Mauck and Clarke, 2006). Some osteoporosis-
50 years and older. Z-scores represent the num- related fractures are caused by a minor fall, so
ber of standard deviations from the normal mean patients may benefit from balance programs,
value for age- and sex-matched control subjects which may reduce the risk of falls.
(Mauck and Clarke, 2006). Z-scores are used pref- Serum vitamin D levels decrease with advanc-
erentially to assess bone loss in premenopausal ing age because individuals are exposed to less
females and in men younger than age 50 years sunlight and the skin capacity for vitamin D pro-
(Mauck and Clarke, 2006). A wide variety of tech- duction is reduced (Mauck and Clarke, 2006).
niques is available to assess BMD, including quan- Consequently, calcium absorption decreases.
titative ultrasound (QUS), quantitative computed Therefore, calcium (1000 mg/day) and vitamin
tomography (QCT), digital x-ray radiogramme- D supplementation should be ordered to elderly
try, dual-photon radionuclide absorptiometry, patients if dietary intake and sun exposure are not
magnetic resonance imaging, and DXA (Kanis sufficient (Mauck and Clarke, 2006). A total cal-
et al., 2008). Among them, DXA is the most com- cium intake of 1500 mg per day (through diet and/
monly used and considered as the gold standard or supplements) and a total vitamin D intake of
for BMD evaluation (Karjalainen et  al., 2016). 600–800 IU per day is recommended for osteopo-
Ionizing radiation, sophisticated instrumentation, rotic patients (Black and Rosen, 2016). For patients
and higher costs are the major handicaps of this with osteoporosis, calcium supplementation
method. Most guidelines suggest a single BMD should be used as an adjunct to other pharmaco-
assessment at or around 65 years of age (Solomon logical interventions rather than as monotherapy.
et  al., 2016). However osteoporotic fractures are Vitamin D supplementation alone has not been
not always associated with low BMD (Compston, shown to reduce the risk of fractures or increase
2015). Some clinical risk factors other than low BMD except in vitamin D–deficient individuals
BMD contribute independently to fracture risk. (Mauck and Clarke, 2006; Black and Rosen, 2016).
Therefore, a need has risen for a more comprehen- The main purpose of pharmacological therapy
sive and effective tool in risk assessment of osteo- in osteoporosis is to reduce the risk of fracture,
porotic fractures. For this purpose, the Fracture not just increase BMD. Pharmacologic agents for
Risk Assessment Tool (FRAX) developed by the the treatment of osteoporosis can be classified as
WHO on the basis of data from several interna- either antiresorptive or anabolic. The antiresorp-
tional cohorts, incorporates established risk fac- tive agents are bisphosphonates, calcitonin, recep-
tors and BMD to predict individual 10-year risk of tor activator of nuclear factor kappa-B ligand
hip or major osteoporotic fracture. FRAX is freely (RANKL) antibody, and selective estrogen receptor
available on the Internet and it can be used either modulators (SERM). These agents suppress bone
with or without BMD values. FRAX has some turnover by reducing bone resorption. Teriparatide
limitations such as inability to take into account is the only anabolic drug, which stimulates bone
246  Precision medicine in osteoporosis and bone diseases

Table 13.1  Drugs and drug targets for estrogen receptors. In some tissues, such as bone,
osteoporosis it acts an estrogen agonist while in others, such
as breast and uterus, it has antiestrogen effects.
Drug Drug target
Raloxifene is approved for both the prevention and
Bisphosphonates Farnesyl pyrophosphate the treatment of postmenopausal osteoporosis at a
Teriparetide PTH receptor dose of 60 mg/d (Black and Rosen, 2016). Recently,
Denosumab RANKL the combination of another SERM, bazedoxifene,
SERMs Estrogen receptor with estrogen was approved by the FDA for the
Calcitonin Calcitonin receptor treatment of menopausal symptoms and the pre-
Odanacatib Cathepsin K vention of osteoporosis (Cosman et al., 2014).
Abaloparatide PTH receptor Recombinant human parathyroid hormone ana-
Romosozumab Sclerostin
logues are potent bone anabolic agents (Bhandari
et al., 2016) and teriparetide is FDA-approved for
osteoporosis treatment in patients at high risk for
formation. Drugs and drug targets for osteoporosis fracture. Teriparetide initially increases bone for-
are summarized in Table 13.1. mation but subsequently also resorption, so its
Bisphosphonates bind at the bone ­mineral sur- benefits are quickly lost (Harslof and Langdahl,
face, where they potently inhibit osteoclast-mediated 2016). Therefore, it is recommended to use an anti-
bone resorption and are used orally or intravenously. resorptive agent following teriparetide treatment
Bisphosphonates are the most potent and the most (Black et al., 2005).
widely prescribed a­ntiresorptive agents for both A number of combinations of therapies are pos-
osteoporosis ­prevention and ­treatment (Mauck and sible, but to date no studies regarding combination
Clarke, 2006; Maraka and Kennel, 2015). Atypical therapy have had sufficient power to demonstrate
femoral fractures and osteonecrosis of the jaw are additive or synergistic effects of different combina-
rare but are more s­ erious adverse effects of bisphos- tions on fracture outcomes.
phonates (Maraka and Kennel, 2015).
Calcitonin nasal spray is approved for the EMERGING TREATMENTS FOR
treatment of osteoporosis by the Food and Drug OSTEOPOROSIS
Administration (FDA). Calcitonin inhibits bone Osteoclasts secrete a lot of proteases during bone
resorption, but it is not considered as first-line resorption, and cathepsin K is one of these pro-
treatment for osteoporosis because its efficacy in teases (Harslof and Langdahl, 2016). Odanacatib
fracture reduction is lower than other medications is a reversible inhibitor of cathepsin K that inter-
(Chesnut et al., 2000). feres only with osteoclast activity (Gauthier et al.,
Denosumab is a human monoclonal antibody 2008). Studies showed that its antifracture efficacy
that decreases osteoclasts differentiation by bind- is comparable to current treatments (Nakamura
ing to the receptor activator of nuclear factor-κβ et  al., 2013; McClung et  al., 2014), but it has not
ligand (RANKL). RANKL is upregulated in post- been approved by medical authorities yet.
menopausal women as a result of estrogen decline The osteocytes inhibit bone formation by pro-
(Laskowski et al., 2016). As with bisphosphonates, ducing sclerostin (van Bezooijen et  al., 2004).
atypical femur fractures and osteonecrosis of the Romosozumab is an antisclerostin antibody that
jaw have been observed with denosumab treatment has been reported to stimulate bone formation as
(Black and Rosen, 2016). well as inhibit bone resorption (Chapurlat, 2016).
Estrogen has direct effects on bone tissue lead- However, the antifracture efficacy of this drug
ing to inhibition of bone resorption and mainte- remains to be determined in ongoing trials.
nance of bone formation. The FDA has withdrawn Abaloparatide is a novel synthetic analog of
approval of estrogen or hormone therapy for human PTH-related protein (Hattersley et  al.,
treatment of osteoporosis but has continued 2016). Preclinical trials showed that it induces
approval of their use for prevention of the disease bone formation without stimulating resorption
(Cosman et al., 2014). (Leder et al., 2015). Its effect seems to be superior
Raloxifene is a selective estrogen reception to teriparatide in preventing fracture (Hattersley
modulator (SERM) that selectively interacts with et al., 2016).
 Introduction 247

IMPORTANCE OF LIFESTYLE, Uitterlinden, 2010). Muscle strength, body mass

ENVIRONMENTAL, AND GENETIC FACTORS index, levels of calciotropic hormones in circula-
IN OSTEOPOROSIS tion, and biochemical markers of bone turnover
Osteoporosis is a multifactorial disease. Many are other heritable traits related to osteoporosis
genetic variations, and lifestyle and environmental (Ralston and Uitterlinden, 2010). Estrogen defi-
factors influence the susceptibility to osteoporosis ciency is one of the most important factors in the
and its complications. Dietary intake of calcium reduction of BMD in postmenopausal women
and vitamin D, physical activity, low body mass (Karasik and Cohen-Zinder, 2012) and twin stud-
index, excessive alcohol use, and smoking are mod- ies showed that age at menopause is genetically
ifiable risk factors of the disease (Ferrari, 2005). determined (Snieder et al., 1998). This information
Increased age, female gender, white race, and posi- provides further support for the role of genetic fac-
tive family history are the other risk factors that tors in determining bone loss in women. In order
cannot be modified (Evenson and Sanders, 2016). to expose the genetic determinants of osteoporo-
Improved understanding of all these hereditary sis, a number of linkage and candidate gene studies
and nonhereditary factors has directed the medi- have been conducted (Mafi Golchin et al., 2016).
cal community to an individualized approach to Linkage studies have identified loci for BMD,
prevention, diagnosis, and treatment of osteoporo- but these findings have not been replicated between
sis. The implementation of preventive strategies for studies, and meta-analysis indicated no such loci
osteoporosis and targeted education of genetically (Ioannidis et  al., 2007). This led investigators to
susceptible individuals in order to reduce risks focus on candidate gene studies, but again most of
through lifestyle modification would be done only these studies failed to meet expectations (nonrep-
by recognition of the genetic and environmental licative results) (Clark and Duncan, 2015). Some of
basis of the disease and the interactions between the osteoporosis-related genes investigated in can-
them. If individuals genetically susceptible to dis- didate gene studies are shown in Table 13.2. Today,
ease with potential to interact with modifiable risk genome-wide association studies (GWAS) have
factors can be detected and avoidance of those identified common genetic variants of small effect
risk factors can be achieved, the risk of osteopo- size that contribute to regulation of BMD and frac-
rosis and its complications can be minimized. ture risk in the general population (Richards et al.,
Furthermore, individual genetic and environmen- 2012). GWAS scan the entire genome to identify
tal factors could influence response to therapeutic novel genes/genome regions with modest effects on
agents used in osteoporosis prevention and treat- complex diseases such as osteoporosis (Hirschhorn
ment. A number of studies showed that efficacy and Daly, 2005). More than 66 loci influencing
and safety of the antiosteoporotic drugs are highly BMD have been reported after GWASs, some
variable among treated patients, even with a given of which are also associated with fracture risk.
drug and a given dosage (Marini and Brandi, 2012). Postulated bone mineral density genes identified
All these reasons encourage us to understand the in GWAS and meta-analyses of GWAS are shown
genetic basis of osteoporosis for both identifying in Table 13.3.
novel diagnostic and therapeutic targets and man- Many of these variants map to genes in known
aging osteoporosis more effectively. pathways related to bone biology, including the
focal adhesion signaling pathway, TGF-β signaling
GENETICS OF OSTEOPOROSIS pathway, osteogenic differentiation of human mes-
Today it is widely accepted that genetic factors have enchymal stem cells, endochondral ossification
a great role in etiopathogenesis of osteoporosis. pathway, vitamin D endocrine pathway, RANK/
Many different genetic variants and their inter- RANKL/OPG, and Wnt signaling (Mafi Golchin
action with environmental factors (diet, exercise, et al., 2016). RANK/RANKL/OPG and Wnt signal-
etc.) influence the susceptibility to osteoporosis. ing pathways are illustrated in Figures 13.1 and
According to twin and family studies, 50%–85% 13.2, respectively.
of BMD variance is genetically determined (Park The other loci have no apparently known bio-
et  al., 2012; Wagner et  al., 2013) and the risk of logical function related to bone tissue. Despite the
fragility fractures is heritable by mechanisms unarguably huge breakthroughs of GWAS, only 6%
partly independent from BMD (Ralston and of the heritability of BMD has been explained to
248  Precision medicine in osteoporosis and bone diseases

Table 13.2  Some of the osteoporosis-related genes investigated in candidate gene studies

Gene symbol Gene name Function Pathway

COL1A1 Collagen, type I, alpha 1 Synthesis of type I collagen
CYP19A1 Cytochrome P450, family 19, The key enzyme catalyzing Estrogen
subfamily A, polypeptide 1 the conversion of
androgens to estrogens
(Aromatase)
DBP D site of albumin promoter Vitamin D carrier Vitamin D
(albumin D-box) binding protein
ESR1 Estrogen receptor 1 Estrogen receptor Estrogen
ESR2 Estrogen receptor 2 Estrogen receptor Estrogen
LRP5 LDL receptor-related protein 5 Wnt coreceptor Wnt/β-catenin
signaling
SFRP1 Secreted frizzled-related protein 1 Modulation of Wnt signaling Wnt/β-catenin
signaling
SOST Sclerosteosis Inhibition of Wnt signaling Wnt/β-catenin
signaling
TNFSF11 Tumor necrosis factor ligand Regulation of bone turnover RANKL/RANK/
superfamily, member 11 (RANKL) OPG
TNFRSF11A Tumor necrosis factor receptor Regulation of bone turnover RANKL/RANK/
superfamily, member OPG
11a, NFκB activator (RANK)
TNFRSF11B Tumor necrosis factor receptor Regulation of bone turnover RANKL/RANK/
superfamily, member 11b (OPG) OPG
VDR Vitamin D receptor Vitamin D receptor Vitamin D
WNT10B Wingless-type MMTv integration Regulation of bone turnover Wnt/β-Catenin
site family, member 10B signaling

date (Richards et al., 2012). Genetic architecture of Single-gene disorders of bone

osteoporosis is affected from several genes that have
small effects, but a few of which have large effects. Several rare bone diseases have been identified that
occur as the result of mutations in single genes.
EPIGENETIC MECHANISMS These diseases have provided important and
Epigenetics refers to stable and heritable changes novel insights into genes and pathways involved in
in gene expression other than changes in DNA regulation of bone mass, bone cell function, and
sequence. Many environmental factors interact bone quality. Depending on the function of the
with genes through epigenetic mechanisms, and, genes involved and the nature of the genetic varia-
in some cases, epigenetic alterations are transmis- tion, a full spectrum of phenotypes is observed
sible beyond a single generation (Holroyd et  al., ranging from severe bone loss to extreme bone
2012). There are a few studies investigating the mass. The diseases mentioned in this section and
role of epigenetic mechanisms on differentiation their phenotypic features are summarized in Table
of bone cells and bone turnover (Boudin et  al., 13.4.
2016) and the results of these studies have shown While discussing these monogenic disorders,
the importance of epigenetic mechanisms on bone we will divide them into two subgroups according
biology (Holroyd et al., 2012). to BMD values.
 Introduction 249

Table 13.3  Postulated bone mineral density genes identified in genome-wide association studies and
meta-analyses

Also associated
with osteoporotic
Gene Full name Study design fracture?
ABCF2 ATP-binding cassette, subfamily F Meta-analysis
ADAMTS18 ADAM metallopeptidase with GWAS Y
thrombospondin type 1 motif, 18
ALDH7A1 Aldehyde dehydrogenase 7 family, GWAS Y
member A1
ANAPC1 Anaphase promoting complex subunit 1 Meta-analysis
ARHGAP1 Rho GTPase activating protein 1 Meta-analysis
AXIN1 Axin 1 Meta-analysis
C7ORF58 Chromosome 7 open reading frame 58 Meta-analysis
C12ORF23 Chromosome 12 open reading frame 23 Meta-analysis
C18ORF19 Chromosome 18 open reading frame 19 Meta-analysis
CDKAL1 CDK5 regulatory subunit associated Meta-analysis
protein 1-like 1
CLCN7 Chloride channel, voltage-sensitive 7 GWAS
CLDN14 Claudin 14 Meta-analysis
CPN1 Carboxypeptidase N, polypeptide 1 Meta-analysis
CRHR1 Corticotropin releasing hormone Meta-analysis
receptor 1
CTNNB1 Catenin (cadherin-associated protein), Meta-analysis Y
beta 1
CYLD Cylindromatosis (turban tumor syndrome) Meta-analysis
DCDC5 Doublecortin domain-containing 5 Meta-analysis Y
DHH Desert hedgehog Meta-analysis
DNM3 Dynamin 3 Meta-analysis
DOK6 Docking protein 6 GWAS
ERC1 ELKS/RAB6 interacting/CAST family Meta-analysis
member 1
ESR1 Estrogen receptor 1 GWAS
FAM9B Family with sequence similarity 9, Meta-analysis
member B
FLJ42280 Putative uncharacterized protein Meta-analysis
FMN2/GREM2 Formin 2/gremlin 2 GWAS
FOXL1 Forkhead box L1 Meta-analysis
FUBP3 Far upstream element (FUSE) binding Meta-analysis Y
protein 3
GALNT3 UDP-N-acetyl-alpha-D- GWAS
galactosamine:polypeptide
N-acetylgalactosaminyltransferase 3
GPATCH1 G patch domain containing 1 Meta-analysis
GPR177/WLS Wntless homolog (Drosophila) Meta-analysis
HDAC5 Histone deacetylase 5 Meta-analysis
(Continued)
250  Precision medicine in osteoporosis and bone diseases

Table 13.3 (Continued)  Postulated bone mineral density genes identified in genome-wide association
studies and meta-analyses

Also associated
with osteoporotic
Gene Full name Study design fracture?
IBSP Integrin-binding sialoprotein GWAS
IDUA Meta-analysis
IL21R Interleukin 21 receptor GWAS
INSIG2 Alpha L-iduronidase Meta-analysis
JAG1 Jagged 1 GWAS
KCNMA1 Potassium large conductance calcium- Meta-analysis
activated channel, subfamily M, alpha
member 1
LACTB2 Lactamase, beta 2 Meta-analysis
LEKR1 Leucine, glutamate and lysine rich 1 Meta-analysis
LIN7C Lin 7 homolog C Meta-analysis
LRP4 Low-density lipoprotein receptor-related Meta-analysis
protein 4
LRP5 Low-density lipoprotein receptor-related GWAS Y
protein 5
LRRC4C Leucine-rich repeat-containing 4C GWAS
MARK3 MAP/microtubule affinity-regulating GWAS
kinase 3
MBL2 Mannose-binding lectin (protein C) 2, Meta-analysis Y
soluble
MEF2C Myocyte enhancer factor 2C GWAS
MEPE Matrix extracellular phosphoglycoprotein Meta-analysis Y
MPP7 Membrane protein, palmitoylated 7 Meta-analysis
(MAGUK p55 subfamily member 7)
OSBPL1A Oxysterol binding protein-like 1A GWAS
PTHLH Parathyroid hormone-like hormone Meta-analysis
RAP1A RAS-related protein RAP1A GWAS
PKDCC Protein kinase domain containing, Meta-analysis
cytoplasmic homolog
NTAN1 Ribosomal protein S6 kinase, 90 kda, Meta-analysis
polypeptide 5
RSPO3 R-spondin 3 GWAS
RUNX2 Runt-related transcription factor 2 Meta-analysis
SALL1/CYLD Sal-like 1 (Drosophila) Meta-analysis
SLC25A13 Solute carrier family 25 Meta-analysis Y
SMG6 Smg-6 homolog, nonsense mediated Meta-analysis
MRNA decay factor (C. elegans)
SMOC1 SPARC-related modular calcium binding 1 Meta-analysis
SOST Sclerostin GWAS Y
SOX4 SRY (sex determining region Y)-box 4 GWAS
SOX6 SRY (sex determining region Y)-box 6 Meta-analysis
(Continued)
 Introduction 251

Table 13.3 (Continued)  Postulated bone mineral density genes identified in genome-wide association
studies and meta-analyses

Also associated
with osteoporotic
Gene Full name Study design fracture?
SOX9 SRY-box containing gene 9 Meta-analysis
SP7 Sp7 transcription factor 7 GWAS
SPP1 Secreted phosphoprotein 1 GWAS Y
SPP2 Secreted phosphoprotein 2, 24 kda GWAS
SPTBN1 Spectrin, beta, non-erythrocytic 1 Meta-analysis Y
STARD3NL STARD3 N-terminal like Meta-analysis Y
TBC1D8 TBC1 domain family, member 8 (with GWAS
GRAM domain)
TGFBR3 Transforming growth factor, beta GWAS Y
receptor III
TNFSF11/ Tumor necrosis factor (ligand) superfamily, GWAS
RANKL member 11
TNFRS-F11A/ Tumor necrosis factor receptor GWAS Y
RANK superfamily, member 11a, NFKB
activator
TNFRS-F11B/ Tumor necrosis factor receptor GWAS
OPG superfamily, member 11b
WNT16 Wingless-type MMTV integration site Meta-analysis Y
family, member 16
WNT4 Wingless-type MMTV integration site Meta-analysis Y
family, member 4
XKR9 XK, Kell blood group complex subunit- Meta-analysis
related family, member 9
ZBTB40 Zinc finger and BTB domain containing 40 Meta-analysis Y
Source: This table is modified from Liu YJ et al., J Bone Metab 21(2);2014:99–116.

MONOGENIC DISEASES WITH HIGH BMD the TNFRSF11A gene encoding RANK or in the
This group of diseases can occur as a result of TNFSF11 gene encoding RANKL cause osteo-
impaired bone resorption, increased bone forma- clast-poor osteopetrosis in many cases (Delgado-
tion, or increased bone turnover. Calle et  al., 2012). Due to the importance of the
RANK–RANKL interactions in bone resorption,
Impaired bone resorption this interplay was accepted as a potential target
Osteopetrosis is characterized by failure of for osteoporosis treatment. Currently, a RANKL-
osteoclastic bone resorption. The most common inhibitor, denosumab, is approved for the treat-
type of the disease is called osteoclast-rich osteo- ment of osteoporosis (Makras et al., 2015).
petrosis that occurs as the result of defects in Pycnodysostosis is an autosomal recessive bone
osteoclast function. Many different gene muta- disorder caused by disruption of the osteoclastic
tions that impair the ability of osteoclasts to function. Mutations resulting in loss of function
resorb bone have been reported in this type of in the cathepsin K (CTSK) gene have been shown
osteopetrosis. in this disease. CTSK is the main protease of the
Failure in osteoclast differentiation can cause bone resorption, and today odanacatib (CTSK
another type of the disease called osteoclast- inhibitor) is developed for osteoporosis treatment
poor osteopetrosis. Inactivating mutations in (McClung et al., 2014).
252  Precision medicine in osteoporosis and bone diseases

Macrophage Prefusion osteoclast

Multinucleated
osteoclast
RANKL

OPG

RANK Active
osteoclasts

Osteocytes and
osteoblasts

Bone
Bone matrix
resorption

Figure 13.1  RANK/RANKL/OPG pathway in bone turnover. Osteoblasts and osteocytes secrete
both RANKL and OPG, which is a soluble decoy receptor for RANKL. In the absence of OPG, RANKL
engages RANK present on prefusion osteoclasts. Binding of RANK/RANKL leads to osteoclast matu-
ration and activation, resulting in increased bone resorption.

WNT
WNT
Frizzled LRP5/6
LRP5/6

Beta-catenin Proteasome Beta-catenin

ub
ub
ub
Transcription of
target genes

Osteoblast
Nucleus

Figure 13.2  Wnt/β-catenin pathway in osteoblasts. In the liganded state on the right side of the
figure, Wnt protein binds coreceptors LRP5/6 and Frizzled. This engagement leads to stabilization
and accumulation of β-catenin in the cytoplasm and then translocation to the nucleus for transcription
of target genes. β-catenin promotes osteoblastic differentiation. In the unliganded state on the left
side of the figure, the absence of Wnt ligand binding to the Frizzled receptor and LRP 5/6 coreceptor
results in ubiquitination of β-catenin and subsequent degradation in proteasome. Ub, ubiquitin.
 Introduction 253

Table 13.4  Monogenic bone diseases associated with abnormal bone mass

Disease Phenotype Genes

Osteopetrosis High bone mass, CLCN7
fractures TCIRG1
CATK
OSTM1
RANKL (TNFSF11)
RANK (TNFRSF11A)
Pycnodysostosis High bone mass CTSK
Sclerosteosis High bone mass SOST
Van Buchem disease High bone mass SOST
Craniodiaphyseal dysplasia High bone mass SOST
Osteopathia striata High bone mass AMER1
Raine syndrome High bone mass FAM20C
Camurati–Engelmann disease High BMD TGFβ1
Paget’s disease of bone High BMD SQSTM1
VCP
Juvenile Paget’s disease High BMD OPG (TNFRSF11B)
Osteogenesis imperfecta Low BMD, fractures COL1A1
COL1A2
CRTAP
LEPRE
PPIB
Juvenile osteoporosis Low BMD WNT1
LRP5
PLS3
Osteoporosis-pseudoglioma syndrome Low bone mass, fractures LRP5

Increased bone formation or fetal stage. Inactivating mutations in the gene

Sclerosteosis occurs due to loss of function muta- coding APC membrane recruitment protein 1
tions in the SOST gene, encoding for sclerostin, an (AMER1), an inhibitor of Wnt signaling, have been
inhibitor of WNT/β-catenin signaling (van Lierop shown to give rise to disease (Jenkins et al., 2009).
et al., 2011). Van Buchem disease is another auto- Raine syndrome develops when inactivating muta-
somal recessive disorder caused by insufficiency of tions occur in the family with sequence similarity
sclerostin. It is caused by a deletion of an enhancer 20, member C (FAM20C) gene. This gene encodes
region downstream of SOST, required for adequate a casein kinase that is essential for the osteoblastic
sclerostin expression (van Lierop et  al., 2013). differentiation (Ababneh et al., 2013).
Another sclerostin-related disorder is cranio-
diaphyseal dysplasia. It is caused by mutations Increased bone turnover
impairing sclerostin secretion (Kim et  al., 2011). Camurati–Engelmann disease is a rare disor-
After the discovery of molecular pathways of these der characterized by increased bone turnover.
diseases, sclerostin gained attention as a target to Function mutations in transforming growth fac-
increase BMD in patients with osteoporosis. A tor, beta 1 (TGFb1) have been identified as the
monoclonal antibody targeting serum sclerostin, genetic cause of the disease (Kinoshita et al., 2000).
romosozumab, has been developed for osteoporo- Paget’s disease of bone (PDB) is caused by an
sis treatment (Makras et al., 2015). elevated bone turnover rate, leading to excessive
Osteopathia striata is a rare, X-linked disorder production of bone with impaired and disorga-
that is usually mortal for males in the neonatal nized structure.
254  Precision medicine in osteoporosis and bone diseases

Mutations in the genes (sequestosome 1 have some limitations. Major morbidity compli-
[SQSTM1] and valosin-containing protein [VCP]) cations of the disease are fractures, especially hip
encoding for proteins involved in osteoclast dif- fractures. The osteoporotic fractures are associated
ferentiation have been found to cause this dis- with at least a twofold increase in mortality. The
order. It presents with an autosomal dominant risk of death related to hip fracture of a 50-year-
mode of inheritance (Delgado-Calle et  al., 2012). old woman is equivalent to her risk of death from
Inactivating mutations in tumor necrosis factor breast cancer and 4 times higher than that from
receptor superfamily member 11b (TNFRSF11B) endometrial cancer. The number of elderly individ-
have been identified as the cause of juvenile Paget’s uals is rising in every geographic region, and these
disease (Whyte et  al., 2002). This gene encodes a demographic changes could lead to an increase in
decoy receptor for RANKL, namely, osteoprote- the number of osteoporotic fractures occurring
gerin (OPG). worldwide. The number of hip fractures worldwide
is forecast to be 6.3 million in 2050 with the annual
MONOGENIC DISEASES WITH DECREASED cost of $131.5 billion. Bone homeostasis involves
BMD numerous molecular pathways, and osteoporosis
Osteogenesis imperfecta is a clinically and geneti- is a complex disease caused by the synergic contri-
cally heterogeneous group of connective tissue bution of several genes. The effect of a single gene
disorders. As well as low BMD and fractures, on BMD value is relatively small except for rare
extraosseus connective tissue symptoms are seen cases of osteoporosis as explained earlier. Up to
in patients with osteogenesis imperfecta. The 50%–85% of bone loss acceleration is due to genetic
International Working Group on Constitutional factors. Furthermore, environmental factors and
Disorders of Bone divided the known osteogenesis lifestyle have important influence on susceptibil-
imperfecta types into five main groups on the basis ity of osteoporosis probably through epigenetic
of the specific clinical features and severity of the mechanisms. Several different methods have been
disease. Despite the increasing number of known applied to explore the genetic basis of osteoporo-
disease-causing genes, the large majority of the sis. More than 66 loci have been reported to be
cases (∼90%) is due to mutations in collagen, type associated with BMD in GWAS, but they explain
I, alpha 1 (COL1A1); and collagen, type I, alpha 2 only a small fraction of the total genetic vari-
(COL1A2) genes (van Dijk et al., 2012). ance in BMD. With the implementation of high-
Juvenile osteoporosis is a group of heritable dis- throughput, next-generation DNA sequencing
orders that present during childhood. Decreased technologies, identification of additional genes can
WNT signaling activity because of mutations be expected in the near future. For the moment,
in WNT1 itself and mutations in LRP5 (which a number of risk variants within disease-associ-
encodes a coreceptor involved in Wnt/β-catenin ated genes have been identified as a result of great
signaling) have been reported to cause juvenile efforts, and these data allowed novel insights into
osteoporosis. Recently, mutations in the plastin 3 the pathophysiology of osteoporosis as well as new
(PLS3) gene were identified in patients suffering therapeutic approaches. However, existing knowl-
from an X-linked variant of the disease (van Dijk, edge is not enough for providing an explanation
2015). for the diagnosis or a tool to assess the genetic risk
Osteoporosis-pseudoglioma syndrome is an of osteoporosis in an individual. The undiscovered
autosomal recessive disorder that is characterized genetic component likely consists of a combination
by severe bone loss and extreme fragility of bones. of many more common variants with increasingly
Loss of function mutations in the LRP5 gene have smaller effects and the contributions of rare vari-
been reported to cause the disease. ants. In the future, some of the missing variance
might be revealed by sequencing the exomes (or
CONCLUSION, CHALLENGES, even whole genomes) of rare variants that have
AND FUTURE DIRECTIONS larger effects on BMD than do any of the common
variants. Rare variant studies will be potentially
Osteoporosis is a highly prevalent bone disorder useful for two reasons. First, GWAS SNP arrays do
recognized by low BMD. DXA and FRAX are use- not tag variants with minor allele frequencies less
ful tools for diagnosis of the disease, although they than 5%. To assess the low frequency spectrum of
 References 255

allelic variation, direct genotyping of these SNPs is Chapurlat R. Cathepsin K inhibitors and antiscleros-
absolutely necessary. Second, rare variants within tin antibodies. The next treatments for osteo-
the exons may negatively influence gene function. porosis? Joint Bone Spine 83(3);2016:254–6.
In addition, functional studies must also be carried Chesnut CH, 3rd et al. A randomized trial of nasal
out to explore the likely functional significance of spray salmon calcitonin in postmenopausal
variants identified. Another point to keep in mind women with established osteoporosis: The
is inherited epigenetic modifications and gene-by- prevent recurrence of osteoporotic frac-
gene and gene-by-environmental interactions are tures study. PROOF Study Group. Am J Med
significant sources of variation. Although it is clear 109(4);2000:267–76.
that epigenetic regulation is involved in the regula- Clark GR, and Duncan EL. The genetics of osteo-
tion of bone mass, it is unclear which part of the porosis. Br Med Bull 113(1);2015:73–81.
missing heritability can be explained by epigenetic Compston J. FRAX—Where are we now?
mechanisms. There has been very limited research Maturitas 82(3);2015:284–7.
to date into the genetic determinants of response Cosman F et al. Clinician’s guide to prevention
to treatment with antiosteoporotic agents. With and treatment of osteoporosis. Osteoporos
advanced knowledge of the molecules and path- Int 25(10);2014:2359–81.
ways in bone biology, drug treatments could be Datta HK et al. The cell biology of bone metabo-
optimized according to the patient’s individual lism. J Clin Pathol 61(5);2008:577–87.
genetic profile to allow the most beneficial treat- Delgado-Calle J et al. DNA methylation contrib-
ment for each patient. utes to the regulation of sclerostin expres-
In conclusion, when planning a study related to sion in human osteocytes. J Bone Miner Res
osteoporosis, sample size should be increased to 27(4);2012:926–937.
identify further common variants, next-generation Divieti Pajevic P. Recent progress in osteo-
DNA sequencing technologies should be applied to cyte research. Endocrinol Metab (Seoul)
identify rare variants of larger effect size, and func- 28(4);2013:255–61.
tional studies should be done to explore the pos- Evenson AL, and Sanders GF. Educational inter-
sible functional significance of variants on BMD. vention impact on osteoporosis knowledge,
health beliefs, self-efficacy, dietary calcium,
REFERENCES and vitamin D intakes in young adults. Orthop
Nurs 35(1);2016:30–6.
Ababneh FK et al. Hereditary deletion of the Felsenberg D et al. Incidence of vertebral
entire FAM20C gene in a patient with ­fracture in Europe: Results from the
Raine syndrome. Am J Med Genet A European Prospective Osteoporosis
161(12);2013:3155–60. Study (EPOS). J Bone Miner Res
van Bezooijen RL et al. Sclerostin is an osteocyte- 17(4);2002:716–24.
expressed negative regulator of bone forma- Ferrari SL. Osteoporosis: A complex disorder
tion, but not a classical BMP antagonist. J Exp of aging with multiple genetic and environ-
Med 199(6);2004:805–14. mental determinants. World Rev Nutr Diet
Bhandari M et al. Does teriparatide improve 95;2005:35–51.
femoral neck fracture healing: Results from Florencio-Silva R et al. Biology of bone tissue:
a randomized placebo-controlled trial. Clin Structure, function, and factors that influence
Orthop Relat Res 474(5);2016:1234–44. bone cells. Biomed Res Int 2015;2015:421746.
Black DM et al. One year of alendronate Gass M, and Dawson-Hughes B. Preventing
after one year of parathyroid hormone osteoporosis-related fractures: An overview.
(1–84) for osteoporosis. N Engl J Med Am J Med 119(4 Suppl 1);2006:S3–11.
353(6);2005:555–65. Gauthier JY et al. The discovery of odanacatib
Black DM, and Rosen CJ. Clinical practice. (MK-0822), a selective inhibitor of cathepsin
Postmenopausal osteoporosis. N Engl J Med K. Bioorg Med Chem Lett 18(3);2008:923–8.
374(3);2016:254–62. Harslof T, and Langdahl BL. New horizons in
Boudin E et al. Genetic control of bone mass. osteoporosis therapies. Curr Opin Pharmacol
Mol Cell Endocrinol 432;2016:3–13. 28;2016:38–42.
256  Precision medicine in osteoporosis and bone diseases

Harvey N, Dennison E, and Cooper C. Kinoshita A et al. Domain-specific mutations in

Osteoporosis: Impact on health and econom- TGFB1 result in Camurati-Engelmann disease.
ics. Nat Rev Rheumatol 6(2);2010:99–105. Nat Genet 26(1);2000:19–20.
Hattersley G et al. Binding selectivity of abalo- Laskowski LK et al. A RANKL wrinkle:
paratide for PTH-Type-1–receptor conforma- Denosumab-induced hypocalcemia. J Med
tions and effects on downstream signaling. Toxicol 12(3);2016:305–8.
Endocrinology 157(1);2016:141–9. Leder BZ et al. Effects of abaloparatide, a human
Hendrickx G, Boudin E, and Van Hul W. A look parathyroid hormone-related peptide analog,
behind the scenes: The risk and pathogenesis on bone mineral density in postmenopausal
of primary osteoporosis. Nat Rev Rheumatol women with osteoporosis. J Clin Endocrinol
11(8);2015:462–74. Metab 100(2);2015:697–706.
Hirschhorn JN, and Daly MJ. Genome-wide asso- Li WF et al. Genetics of osteoporosis: Accelerating
ciation studies for common diseases and com- pace in gene identification and validation.
plex traits. Nat Rev Genet 6(2);2005:95–108. Hum Genet 127(3);2010:249–85.
Hlaing TT, and Compston JE. Biochemical mark- Liu YJ, Zhang L, Papasian CJ, and Deng HW.
ers of bone turnover—uses and limitations. Genome-wide Association Studies for
Ann Clin Biochem 51(Pt 2);2014:189–202. Osteoporosis: A 2013 Update. J Bone Metab
Holroyd C et al. Epigenetic influences in the 21(2);2014:99–116.
developmental origins of osteoporosis. Mafi Golchin M et al. Osteoporosis: A silent
Osteoporos Int 23(2);2012:401–10. disease with complex genetic contribution.
International Osteoporosis Foundation. 2014. J Genet Genomics 43(2);2016:49–61.
Available from http://www.iofbonehealth.org. Makras P, Delaroudis S, and Anastasilakis AD.
Accessed March 20, 2016. Novel therapies for osteoporosis. Metabolism
Ioannidis JP et al. Meta-analysis of genome-wide 64(10);2015:1199–214.
scans provides evidence for sex- and site-spe- Maraka S, and Kennel KA. Bisphosphonates for
cific regulation of bone mass. J Bone Miner the prevention and treatment of osteoporo-
Res 22(2);2007:173–83. sis. BMJ 351;2015:h3783.
Jenkins ZA et al. Germline mutations in WTX Marini F, and Brandi ML. Pharmacogenetics of
cause a sclerosing skeletal dysplasia but do osteoporosis: What is the evidence? Curr
not predispose to tumorigenesis. Nat Genet Osteoporos Rep 10(3);2012:221–7.
41(1);2009:95–100. Mauck KF, and Clarke BL. Diagnosis, screening,
Johnell O. The socioeconomic burden of frac- prevention, and treatment of osteoporosis.
tures: Today and in the 21st century. Am Mayo Clin Proc 81(5);2006:662–72.
J Med 103(2a);1997:20S–25S, discussion McClung MR et al. Odanacatib anti-fracture
25S–26S. efficacy and safety in postmenopausal women
Kanis J, Compston J, and Cooper C. The burden with osteoporosis: Results from the phase III
of fractures in the European Union in 2010. long-term odanacatib fracture trial. Arthritis
Osteoporos Int 23(Suppl 2);2012:S57. Rheum 66(11);2014:S987.
Kanis JA et al. A reference standard for the descrip- Melton LJ, 3rd et al. Perspective. How many
tion of osteoporosis. Bone 42(3);2008:467–75. women have osteoporosis? J Bone Miner Res
Karasik D, and Cohen-Zinder M. Osteoporosis 7(9);1992:1005–10.
genetics: Year 2011 in review. Bonekey Rep Mitchell BD, and Streeten EA. Clinical impact of
1;2012:114. recent genetic discoveries in osteoporosis.
Karjalainen JP et al. New method for point-of- Appl Clin Genet 6;2013:75–85.
care osteoporosis screening and diagnostics. Nakamura T et al. Effect of the cathepsin K
Osteoporos Int 27(3);2016:971–7. inhibitor odanacatib administered once
Kim SJ et al. Identification of signal peptide weekly on bone mineral density in Japanese
domain SOST mutations in autosomal patients with osteoporosis—A double-blind,
dominant craniodiaphyseal dysplasia. Human randomized, dose-finding study. Osteoporos
Genetics 129(5);2011:497–502. Int 25(1);2013:367–76.
 References 257

NIH Consensus Development Panel on van Dijk FS et al. EMQN best practice guide-
Osteoporosis Prevention, Diagnosis, and lines for the laboratory diagnosis of osteo-
Therapy. Osteoporosis prevention, diagnosis, genesis imperfecta. Eur J Hum Genet
and therapy. Jama 285(6);2001:785–95. 20(1);2012:11–9.
Park JH et al. Genetic influence on bone min- van Dijk FS. Genetics of osteoporosis in children.
eral density in Korean twins and families: Endocr Dev 28;2015:196–209.
The healthy twin study. Osteoporos Int van Lierop AH et al. Patients with sclerosteosis
23(4);2012:1343–9. and disease carriers: Human models of the
Ralston SH, and Uitterlinden AG. effect of sclerostin on bone turnover. J Bone
Genetics of osteoporosis. Endocr Rev Miner Res 26(12);2011:2804–11.
31(5);2010:629–62. van Lierop AH et al. Van Buchem disease:
Reginster J-Y, and Burlet N. Osteoporosis: A Clinical, biochemical, and densitometric fea-
still increasing prevalence. Bone 38(2, Suppl tures of patients and disease carriers. J Bone
1);2006:4–9. Miner Res 28(4);2013:848–54.
Richards JB, Zheng HF, and Spector TD. Genetics Wade SW et al. Estimating prevalence of osteo-
of osteoporosis from genome-wide associa- porosis: Examples from industrialized coun-
tion studies: Advances and challenges. Nat tries. Arch Osteoporos 9;2014:182.
Rev Genet 13(8);2012:576–88. Wagner H et al. Genetic influence on bone
Sattui SE, and Saag KG. Fracture mortality: phenotypes and body composition: A
Associations with epidemiology and osteo- Swedish twin study. J Bone Miner Metab
porosis treatment. Nat Rev Endocrinol 31(6);2013:681–9.
10(10);2014:592–602. Wei QS et al. Serum osteopontin levels in relation
Snieder H, MacGregor AJ, and Spector to bone mineral density and bone turnover
TD. Genes control the cessation of markers in postmenopausal women. Scand J
a woman’s reproductive life: A twin Clin Lab Invest 76(1);2016:33–9.
study of ­hysterectomy and age at Whyte MP et al. Osteoprotegerin deficiency
menopause. J Clin Endocrinol Metab and juvenile Paget’s disease. N Engl J Med
83(6);1998:1875–80. 347(3);2002:175–84.
Solomon CG, Black DM, and Rosen CJ.
Postmenopausal osteoporosis. N Engl J Med
374(3);2016:254–62.
 14
Precision medicine in diabetes mellitus

SANDHIYA SELVARAJAN, AKILA SRINIVASAN, NISHANTHI ANANDABASKAR,

SADISHKUMAR KAMALANATHAN, AND MELVIN GEORGE

Introduction 259 Gestational diabetes mellitus 264

Application of precision medicine in type 1 Advantages and limitations of precision
diabetes mellitus 260 medicine 264
Application of precision medicine in type 2 Conclusion 265
diabetes mellitus 262 References 265

INTRODUCTION noncommunicable disease has been mounting

with high prevalence in both developed and devel-
Precision medicine, one of the budding endeav- oping countries. The global prevalence of diabetes
ors of modern medicine, is an innovative concept mellitus among adults has increased from 108 mil-
endorsed since 2011 to advocate personalized ther- lion in 1980 to 422 million in 2014. In 2012, around
apy based on the distinct features of an individual 2.2 million deaths due to cardiovascular and other
patient (Klonoff, 2015). It varies from personalized complications were attributed to higher blood
medicine by focusing on interindividual genomic glucose levels. Likewise, nearly 1.6 million deaths
characteristics to predict treatment strategies for were reported to be due to diabetes mellitus during
each patient with the help of tools like genomics, the year 2015 (World Health Organization, 2016).
metabolomics, proteomics, and bioinformatics The prevalence of diabetes mellitus in India is 69.2
(Collins and Varmus, 2015). Precision medicine million (8.7%) as per the data published in 2015
embraces three salient components, namely, a (World Health Organization, India, 2016).
specific biomarker, methods of analyzing the bio- Diabetes mellitus is classified into various
marker, and treatment based on the biomarker subtypes based on the underlying pathogen-
(Figure 14.1). The prime intent of precision medi- esis as listed in Table 14.1 (American Diabetes
cine is to enhance the success rate of treatment apart Association, 2011). Of these subtypes, type 1 dia-
from lessening morbidity and mortality (Butz et al., betes mellitus (T1D) is an autoimmune disorder
2013). Succeeding the Precision Medicine Initiative resulting in pancreatic beta-cell destruction and
(PMI) by the National Institutes of Health (NIH) insulin deficiency. Type 2 diabetes mellitus (T2D)
to promote the application of precision medicine in is highly influenced by genetics, environment, and
various diseases, there has been a sudden upsurge lifestyle, and is associated with a combination of
in its application to various diseases including that dysfunctional pancreatic beta cells and or insulin
of diabetes mellitus (Fradkin et al., 2016). resistance. Gestational diabetes mellitus (GDM)
Diabetes mellitus, a metabolic disorder, is char- associated with glucose intolerance is seen in preg-
acterized by hyperglycemia owing to either insu- nant women, as the condition is diagnosed for the
lin resistance or insulin deficiency. This chronic first time during pregnancy.
259
260  Precision medicine in diabetes mellitus

Specific biomarker to
diagnose a disease and its
progression or response to
therapy

Qualitative or quantitative
test to identify such a
biomarker

Treatment strategy tailored

as per biomarker results to
minimize toxicity and
maximize efficacy

Figure 14.1  Components of precision medicine.

The role of genetics and the genes signifying APPLICATION OF PRECISION

augmented threat to the occurrence of various MEDICINE IN TYPE 1 DIABETES
types of diabetes mellitus have been identified. MELLITUS
GWAS (genome-wide association studies) have
identified a wide variety of genetic susceptibility Type 1 diabetes mellitus is an autoimmune disor-
loci for development of type 1 and type 2 diabetes der contributing to 5%–10% of the cases of diabe-
mellitus. The risk of type 1 diabetes is increased tes worldwide (Maahs et al., 2010). It is associated
from 1 in 300 to 1 in 15 in patients possessing with destruction of pancreatic beta cells resulting
DR3-DQ2/ DR4-DQ8 haplotypes together in in decreased insulin secretion and with course of
their MHC (major histocompatibility complex) time this condition progresses to a state of com-
region on chromosome 6 (Hartley, 2014). Likewise prehensive insulin deficiency (Gillespie, 2006).
numerous candidate genes, including peroxisome Type 1 diabetes mellitus is a genetic disorder and
proliferator-activated receptor gamma (PPARγ2), has been found to be associated with nearly 20
associated with increased risk of developing type human ­leukocyte antigens (HLA). The first HLA
2 diabetes have also been identified (Abbas et al., to be identified as the major contributor to the
2013). Further, the American Diabetes Association occurrence of familial type 1 diabetes mellitus
(ADA) states that the children diagnosed with dia- was found to be located in chromosome 6. The
betes during the initial 6 months of their life need key HLA found to be associated with the predic-
genetic testing for precise diagnosis of neonatal tion of occurrence of type 1 diabetes mellitus are
diabetes, as it is imperative to treat them with sul- DR4-DQ8, DR3-DQ2 and DR15-DQ (Pociot and
fonylureas (American Diabetes Association, 2016). McDermott, 2002). On the contrary, it has been
Although a wide array of genetic mutations is found that the presence of HLA like DR15-DQ6
implicated in development of diabetes mellitus, haplotype is highly protective against development
none of these genetic profiling tests are currently of type 1 diabetes m
­ ellitus (Price et al., 2001).
used in routine screening tests for diagnosis of Apart from HLA, polymorphisms of insulin
diabetes mellitus (Gillespie, 2006). In view of the genes in chromosome 11 have also been found to
significance of genes involved in instigating this confer genetic susceptibility to the development of
progressive disease, this chapter aims to focus on type 1 diabetes mellitus (Ramos-Lopez et al., 2008).
the applications of precision medicine in various It is also found that polymorphisms in the CTLA4
types of diabetes mellitus. (cytotoxic T-lymphocyte antigen-4) gene located
 Application of precision medicine in type 1 diabetes mellitus  261

Table 14.1  Classification of diabetes mellitus

Classification Underlying mechanisms

Type 1 diabetes mellitus Beta cell destruction resulting in insulin deficiency
Type 2 diabetes mellitus Can vary from predominant insulin resistance with relative insulin
deficiency to predominant defect in insulin secretion along with
insulin resistance
Other specific types of • Genetic defects of beta cell function
diabetes • Genetic defects in insulin action
• Diseases of exocrine pancreas:
• Pancreatitis
• Pancreatectomy
• Neoplasia
• Cystic fibrosis
• Hemochromatosis
• Fibrocalculous pancreatopathy
• Mutations in carboxyl ester lipase
• Endocrinopathies
• Acromegaly
• Cushing’s syndrome
• Glucagonoma
• Pheochromocytoma
• Hyperthyroidism
• Somatostatinoma
• Aldosteronoma
• Drug or chemical induced
• Glucocorticoids
• Vacor (a rodenticide)
• Pentamidine
• Nicotinic acid
• Diazoxide
• Beta adrenergic agonists
• Thiazides
• Hydantoins
• Asparaginase
• Alpha interferon
• Protease inhibitors
• Antipsychotics (atypical and others)
• Epinephrine
• Infections
• Congenital rubella
• Cytomegalo virus
• Coxsackie virus
• Uncommon forms of immune mediated diabetes
• Stiff person syndrome
• Anti-insulin receptor antibodies
• Other genetic syndromes sometimes associated with diabetes
• Wolfram’s syndrome
• Down’s syndrome
(Continued )
262  Precision medicine in diabetes mellitus

Table 14.1 (Continued )  Classification of diabetes mellitus

Classification Underlying mechanisms

• Klinefelter syndrome
• Turner’s syndrome
• Friedrich’s ataxia
• Huntington’s chorea
• Laurence–Moon–Biedl syndrome
• Mytonic dystrophy
• Porphyria
• Prader–Willi syndrome
Gestational diabetes Occurs for the first time during pregnancy or is diagnosed during
mellitus pregnancy
Source: American Diabetes Association, Diabetes Care 34(Suppl 1);2011:S62–69.

in chromosome 2q33 are associated with increased genetic basis. Some of the candidate genes associ-
risk of type 1 diabetes mellitus (Ueda et al., 2003). ated with insulin secretion or resistance resulting
Another gene associated with increased risk of type in type 2 diabetes are shown in Table 14.2 (Sacks
1 diabetes mellitus is the protein tyrosine phospha- and McDonald, 1996; Hani et  al., 1999; Barroso
tase, a nonreceptor type 22 (PTPN22) gene on chro- et  al., 2003; Komurcu-Bayrak, 2012; Brown and
mosome 1p13 that encodes for lymphoid-specific Walker, 2016; Magaña-Cerino et al., 2017).
phosphatase (LYP), which prevents spontaneous Numerous studies have been done to evaluate the
T cell activation. Similarly polymorphism in the association of various genes with the occurrence of
PTPN22 gene leads to development of type 1 dia- diabetes. It has been found that the KATP channels
betes mellitus (Steck et al., 2006). In addition, poly- present in β-cells have a SUR1 subunit encoded by
morphisms in IL2RA (CD25), which encodes for the the ABCC8 gene and a mutation in this gene has
subunit IL-2Rα of the interleukin-2 (IL-2) receptor been associated with a twofold higher risk for diabe-
complex is also implicated in the occurrence of type tes mellitus with an earlier age of onset in Southwest
1 diabetes mellitus (Vella et  al., 2005). Knowledge American Indians (Baier et  al., 2015). Apart from
of these genes and polymorphisms in the occur- KATP channels, studies have been done to explore
rence of type 1 diabetes mellitus would pave the way the role of PPAR-gamma mutations as a risk factor
for the early identification of individuals with an for the onset of type 2 diabetes mellitus, however,
increased risk of developing type 1 diabetes melli- with contradictory results. It has been observed that
tus (i.e., even before metabolic derangement occurs). PPAR-gamma, germ line loss-of-function muta-
This may help in establishing measures to prevent tions are associated with an early onset of insulin
the occurrence of type 1 diabetes mellitus. resistance type 2 diabetes mellitus and hyperten-
sion (Barroso et  al., 1999). Similarly the presence
APPLICATION OF PRECISION of proline allele of PPAR-gamma has been shown
MEDICINE IN TYPE 2 DIABETES to be associated with a greater risk for type 2 dia-
MELLITUS betes. On the contrary, the presence of Pro12Ala
polymorphism of PPAR-gamma has been found to
There has been an escalating trend observed in be protective against the development of type 2 dia-
the global prevalence of type 2 diabetes mellitus betes mellitus in Caucasians (Altshuler et al., 2000).
in the last decade that could possibly be attrib- However, a similar study done by Radha et al. (2006)
uted to the increase in average life span of people among Caucasians and South Asians in Texas and
(Guariguataa, 2014; World Health Organization, South Asians in Chennai showed that the PPAR-
2017). In recent times considerable measures have gamma Pro12Ala polymorphism provides protec-
been taken to detect all the genes involved in the tion against the occurrence of type 2 diabetes only
pathogenesis of type 2 diabetes mellitus. Yet at in Caucasians but not so in South Asians. In addi-
present only about 5% of patients with type 2 diabe- tion, the transcription factor 7-like 2 (TCF7L2) gene
tes mellitus have been found to have an identifiable present in most of the tissues including beta cells of
 Application of precision medicine in type 2 diabetes mellitus  263

Table 14.2  Candidate genes associated with type 2 diabetes mellitus

Candidate genes affecting Candidate genes affecting Candidate genes causing

insulin secretion insulin action insulin resistance
• SLC2A2 (GLUT2) • INSR (gene coding insulin • PPARG
• MicroRNA (miRNA)- receptor) • KLF14
463-3p and ATP-binding • PIK3R1 (gene encoding p85a • IRS1
cassette A4 (ABCG4) regulatory subunit of • GCKR
• Insulin promoter phosphatidylinositol 3-kinase) • FTO
factor-1 (IPF-1) gene • Intronic variant of SNP42 • IGF1
• SOS1 (son of sevenless • TCF7L2
homolog 1 in Drosophila) • NAT2
• ABCC8 (sulphonylurea • ELOVL6
receptor) • FABP
• KCNJ11 (KIR6.2) • RETN
• HNF4A • RBP4
• HNF1A • LEP and LEPR
• D2 and TSHR
Source: Sacks DB, and McDonald JM, Am J Clin Pathol 105(2);1996:149–56; Barroso I et al., PLoS Biol 1(1);2003:e20;
Brown AE, and Walker M, Curr Cardiol Rep 18;2016:75; Magaña-Cerino JM et al., Mol Genet Genomic Med
5(1);2017:50–65; Komurcu-Bayrak E., in Insulin Resistance, Chap 4, Sarika Arora (Ed.), InTech, 2012, http://
cdn.intechopen.com/pdfs/41435/InTech-Impact_of_genetic_polymorphisms_on_insulin_resistance.pdf;
Hani EH et al., J Clin Invest 104(9);1999:R41–8.

the pancreas has been found to have reduced expres- response as well as adverse effects depending on the
sion in fat deposits of obese type 2 diabetic patients genetic variations in drug metabolizing enzymes.
compared to people with normal blood glucose A study that performed an oral glucose tolerance
(Cauchi et al., 2006). A study has identified 6 SNPs test (OGTT) before and after metformin adminis-
(single nucleotide polymorphisms) that are associ- tration in healthy volunteers to appraise the effect
ated with elevated risk for type 2 diabetes mellitus, of metformin in the presence of variants of a drug
as shown in Table 14.3 (Zyriax et  al., 2013). These transporter OCT-1 such as OCT1-R61C, -G401S,
findings suggest a robust contribution of genes in and -G465R, found that the carriers of decreased
the occurrence of type 2 diabetes. function polymorphisms had significantly higher
Apart from causing the disease, genes are also plasma glucose levels compared to those with refer-
shown to have a vital contribution in determining ence OCT-1 alleles (Shu et al., 2007).
the response to various drugs used in the treatment A cohort study carried out on patients with type
of type 2 diabetes mellitus. Metformin, the first-line 2 diabetes found that those on metformin and other
drug used in the treatment for type 2 diabetes melli- drugs that inhibit the function of OCT1 transport
tus, is known to depict interindividual variations in were found to be intolerant to metformin. Similarly,

Table 14.3  SNPs associated with elevated risk for type 2 diabetes mellitus

Cytogenetic
Gene locus Gene name location SNP
KLF14 Kruppel-like factor 14 7q32.3 rs972283
NOTCH2 Notch 2 1p13-p11 rs10923931
DUSP9 Dual specificity phosphatase 9 Xq28 rs5945326
HHEX Hematopoietically expressed homeobox 10q24 rs1111875
SLC30A8 Solute carrier family 30 (zinc transporter), member 8 8q24.11 rs13266634
ZBED3 ZBED3 antisense RNA 1 5q13.3 rs4457053
Source: Zyriax B-C, Salazar R, Hoeppner W et al., PLoS ONE 8(9);2013:e75807.
264  Precision medicine in diabetes mellitus

in the same study, intolerance to metformin was pregnancy is known as gestational diabetes melli-
observed among those patients with reduced func- tus (GDM). The risk factors for GDM include obe-
tion OCT1 alleles compared to patients with nor- sity, family history of diabetes and past history of
mal or one deficient allele with an odds ratio of 2.4 GDM (American Diabetes Association, 2003). The
(CI 1.48–3.93, p < 0.001) (Dujic et  al., 2015). The treatment for this condition can vary from dietary
same study found that those diabetic patients with modifications to insulin (Seshiah et al., 2004). One
both reduced function OCT1 alleles and on OCT1 of the most common genes implicated in the onset
inhibitors were having metformin intolerance of GDM is glucokinase, located on chromosome 7,
4 times higher than other patients without these which regulates a key enzyme of glucose metabo-
risks (Dujic et al., 2015). This clearly envisages the lism. Mutations in this gene have been shown to
significant role of this drug transporter toward the be associated with GDM in around 5%–10% of
prediction of therapeutic response to metformin. Hispanic and Caucasian women (Stoffel et  al.,
Similarly, a study done to evaluate the role of 1993). Moreover, this gene is of key interest, as
genetic polymorphisms on therapeutic response mutations in this can lead to maturity onset diabe-
to sulfonylureas identified the presence of TT in tes of the young (MODY), which is a form of non-
rs12255372 and rs7903146 of the TCF7L2 gene is insulin-dependent diabetes mellitus (NIDDM)
futile in achieving the HbA1C target (Pearson et al., characterized by mild hyperglycemia appearing
2007). Likewise in a study done for 10 weeks with in childhood and a milder form of diabetes that in
pioglitazone at a dose of 30 mg on type 2 diabetic most patients is manageable by dietary modifica-
patients, it has been observed that the presence of tions alone (Froguel et al., 1993). Yet another gene
S447X variant in lipoprotein lipase (LPL) gene was shown to be associated with GDM is hepatocyte
associated with less response to pioglitazone com- nuclear factor 1-α (HNF1-α). A study conducted in
pared to carriers of S447S genotype with favorable Scandinavian women found that the frequency of L
effects on the lipid profile as well as blood pressure alleles was higher in patients with GDM when com-
(Wang et al., 2007). Moreover the genetic variations pared to controls. Moreover, the dominant model
have been found to be associated with the occur- of LL + lL versus II showed a higher odds ratio sig-
rence of adverse drug reactions following treat- nifying the presence of HNF1-α I27L (rs1169288)
ment with pioglitazone. A study instigating the role polymorphism associated with elevated risk of
of genetic variations responsible for the incidence GDM in Scandinavian females (Shaat et al., 2006).
of thiazolidinedione-induced sodium and water Similarly Thr130Ile polymorphism of the tran-
retention in type 2 diabetic patients has shown scription factor HNF4α has been found to be play-
that the AQP2 (aquaporin2) rs296766 T allele and ing an important role in GDM, as it was observed
SLC12A1 rs12904216 were associated with occur- that the frequency of minor alleles of this polymor-
rences of thiazolidinedione-induced edema (Chang phism was higher in women with GDM compared
et  al., 2011). Apart from the conventional drugs to healthy postpartum females in a study done in a
used in diabetes mellitus recently, it has been dem- Mexican population (Monroy et al., 2014). The fact
onstrated that a variant of α2A adrenergic-receptor that HNF4α plays a key role in regulating several
(α2AAR) encoding gene resulting in receptor over- transcriptomes of liver and pancreatic islets may be
expression and decreased insulin secretion among a reason for increasing the risk of diabetes by caus-
type 2 diabetes patients has been shown to have ing dysfunctioning of beta cells (Odom et al., 2004).
improved insulin secretion following administra-
tion of yohimbine, a α2 adrenergic-receptor blocker ADVANTAGES AND LIMITATIONS
(Tang et al., 2014). However, these findings need to OF PRECISION MEDICINE
be proven in clinical trials and the information can
be used for further research in drug discovery and Whole genome sequencing may result in the detec-
development to bring out new molecules. tion of genetic variants that are either protective
or offensive against the development of a specific
GESTATIONAL DIABETES MELLITUS disease (Collins and Varmus, 2015). Hence identi-
fying specific genomic biomarkers may help in pre-
Intolerance to glucose occurring for the first dicting the risk factor for the occurrence of various
time or being detected for the first time during diseases including that of diabetes mellitus. The
 References 265

knowledge obtained from genomic research can on the knowledge gained using genetic research,
be used to delay the onset as well as progression of precision medicine provides the scope for further
the disease through early intervention (Franks and research to develop new molecules catering to the
Poveda, 2017). needs of personalized therapy in diabetes mellitus.
However, we should remember that in multifac-
torial diseases like diabetes and hypertension it is REFERENCES
highly challenging to establish the causative genes.
One more limitation with genomic studies are that Abbas S, Raza ST, Ahmed F et al. Association
the majority of the SNPs associated with disease of genetic polymorphism of PPARγ-2, ACE,
are not located in protein coding sites (Shastry, MTHFR, FABP-2 and FTO genes in risk predic-
2002). Moreover, the genetic make upon its own tion of type 2 diabetes mellitus. J Biomed Sci
may not be adequate to predict susceptibility to a 20(1);2013:80.
particular disease, as the disease is predicted by a Altshuler D, Hirschhorn JN, Klannemark M et al.
wide range of environmental factors (Martin et al., The common PPAR gamma Pro12Ala poly-
1997). The genotype may not necessarily correlate morphism is associated with decreased risk of
with the phenotype in many medical conditions, type 2 diabetes. Nat Genet 26;2000:76–80.
as they do not conform to simple Mendelian inher- American Diabetes Association. Gestational
itance, as these conditions such as susceptibility to diabetes mellitus. Diabetes Care 26(Suppl
diabetes, hypertension, and cancer are known as 1);2003:s103–5.
“complex” traits (Lander and Schork, 1994; Weeks American diabetes Association. Diagnosis and
and Lathrop, 1995). Further, most of the genetic classification of diabetes mellitus. Diabetes
studies are carried out in homogeneous popula- Care 34(Suppl 1);2011:S62–69.
tions and thus the external validity of the data gen- American Diabetes Association. Standards of
erated comes down in real patients who may have medical care in diabetes. Diabetes Care
other accompanying comorbidities (Beigy, 2015). 39(Suppl);2016:S1–12.
Another aspect to be considered is affordabil- Aronson N. Making personalized medi-
ity of precision medicine, especially in resource- cine more affordable. Ann N Y Acad Sci
poor settings. Since specific treatment modalities 1346(1);2015:81–9.
are targeted at a more specific group of audience Baier LJ, Muller YL, Remedi MS et al. ABCC8
based on presence of particular genes or biomark- R1420H loss-of-function variant in a
ers, the market for newer drugs becomes smaller as Southwest American Indian community:
the target population comes down, thereby making Association with increased birth weight and
development costs of newer drugs unsustainable doubled risk of type 2 diabetes. Diabetes
from such small populations (Aronson, 2015). The 64(12);2015:4322–32.
accomplishment of precision medicine in the future Barroso I, Gurnell M, Crowley VE et al. Dominant
will be decided upon the success of therapy using negative mutations in human PPAR gamma
the concept and the associated cost (Klonoff, 2015). associated with severe insulin resistance,
diabetes mellitus and hypertension. Nature
CONCLUSION 402(6764);1999:880–3.
Barroso I, Luan JA, Middelberg RP et al.
Precision medicine aims at providing custom- Candidate gene association study in type 2
made, personalized therapy to an individual patient diabetes indicates a role for genes involved in
based on their genetic changes. Diabetes mellitus β-cell function as well as insulin action. PLoS
being one of the progressively increasing noncom- Biol 1(1);2003:e20.
municable diseases globally with multiple risk Beigy M. Generalized resemblance theory of
factors, measures need to be taken to incorporate evidence: A proposal for precision/per-
the latest advances in medical research to combat sonalized evidence-based medicine. arXiv
the disease at an earlier stage. Precision medi- ­preprint:1512.03825. December 13, 2015.
cine seems to be one of the promising approaches Brown AE, and Walker M. Genetics of insulin
in the near future to evaluate the multigenic risk resistance and the metabolic syndrome. Curr
factors associated with diabetes mellitus. Based Cardiol Rep 18;2016:75.
266  Precision medicine in diabetes mellitus

Butz K, Combest A, Steele S. Using precision from http://www.phgfoundation.org/docu-

medicine to improve patient care. 2013 ments/353_1393402422.pdf.
December. Available at: https://www.ppdi. Klonoff DC. Precision medicine for manag-
com/∼/media/Files/PPDI%20Files/Expert%20 ing diabetes. J Diabetes Sci Technol
Community/Whitepapers/Precision-Medicine- 9(1);2015:3–7.
Improve-Patient-Care-Part1.ashx. Accessed Komurcu-Bayrak E. Impact of genetic poly-
November 11, 2017. morphisms on insulin resistance. In: Insulin
Cauchi S, Meyre D, Dina C et al. Transcription Resistance, Chap 4, Sarika Arora (Ed.), InTech,
factor TCF7L2 genetic study in the French 2012. Available from http://cdn.intechopen.
population: Expression in human beta- com/pdfs/41435/InTech-Impact_of_genetic_
cells and adipose tissue and strong asso- polymorphisms_on_insulin_resistance.pdf.
ciation with type 2 diabetes. Diabetes Lander ES, and Schork NJ. Genetic dis-
55(10);2006:2903–8. section of complex traits. Science
Chang TJ, Liu PH, Liang YC et al. Genetic predis- 265(5181);1994:2037–48.
position and nongenetic risk factors of thia- Maahs DM, West NA, Lawrence JM et al.
zolidinedione-related edema in patients with Chapter 1: Epidemiology of type 1 dia-
type 2 diabetes. Pharmacogenet Genomics betes. Endocrinol Metab Clin North Am
21(12);2011:829–36. 39(3);2010:481–97.
Collins FS, and Varmus H. A new initiative Magaña-Cerino JM, Luna-Arias JP, Labra-Barrios
on precision medicine. N Engl J Med ML et al. Identification and functional analysis
372(9);2015:793–5. of c. 422_423InsT, a novel mutation of the
Dujic T, Zhou K, Donnelly LA et al. Association of HNF1A gene in a patient with diabetes. Mol
organic cation transporter 1 with intolerance Genet Genomic Med 5(1);2017:50–65.
to metformin in type 2 diabetes: A GoDARTS Martin N, Boomsma D, and Machin G. A twin-
study. Diabetes 64;2015:1786–93. pronged attack on complex traits. Nat Genet
Fradkin JE, Hanlon MC, and Rodgers GP. NIH 17(4);1997:387–92.
Precision medicine initiative: Implications Monroy VS, Díaz CA, Trenado LM et al. Thr130Ile
for diabetes research. Diabetes Care polymorphism of HNF4A gene is associated
39(7);2016:1080–4. with gestational diabetes mellitus in Mexican
Franks PW, and Poveda A. Lifestyle and preci- population. J Investig Med 62(3);2014:632–4.
sion diabetes medicine: Will genomics help Odom DT, Zizlsperger N, Gordon DB et al.
optimise the prediction, prevention and Control of pancreas and liver gene expres-
treatment of type 2 diabetes through lifestyle sion by HNF transcription factors. Science
therapy? Diabetologia 60(5);2017:784–792. 303(5662);2004:1378–81.
Froguel P, Zouali H, Vionnet N et al. Familial Pearson ER, Donnelly LA, Kimber C et al.
hyperglycemia due to mutations in glucoki- Variation in TCF7L2 influences therapeu-
nase: Definition of a subtype of diabetes mel- tic response to sulfonylureas. Diabetes
litus. N Engl J Med 328(10);1993:697–702. 56;2007:2178–82.
Gillespie KM. Type 1 diabetes: Pathogenesis and Pociot F, and McDermott MF. Genetics of
prevention. CMAJ 175(2);2006:165–70. type 1 diabetes mellitus. Genes Immun
Guariguataa L. Global estimates of diabetes 3(5);2002:235–49.
prevalence for 2013 and projections for 2035. Price P, Cheong KYM, Boodhoo A et al. Can
Diabetes Res Clin Pract 103(2);2014:137–49. MHC class II genes mediate resistance to type
Hani EH, Stoffers DA, Chèvre JC et al. Defective 1 diabetes? Immunol Cell Biol 79;2001:602–6.
mutations in the insulin promoter factor-1 Radha V, Vimaleswaran KS, Babu HN et al. Role of
(IPF-1) gene in late-onset type 2 diabetes mel- genetic polymorphism peroxisome prolifera-
litus. J Clin Invest 104(9);1999:R41–8. tor-activated receptor-gamma2 Pro12Ala on
Hartley K. The genomic contribution to diabetes. ethnic susceptibility to diabetes in South-Asian
Briefing note: Diabetes, genomics & public and Caucasian subjects: Evidence for hetero-
health. PHG Foundation. 2014. Available geneity. Diabetes Care 29;2006:1046–51.
 References 267

Ramos-Lopez E, Lange B, Kahles H et al. Insulin CTLA4 with susceptibility to autoimmune

gene polymorphisms in type 1 diabetes, disease. Nature 423(6939);2003:506–11.
Addison’s disease and the polyglandular Vella A, Cooper JD, Lowe CE et al. Localization
autoimmune syndrome type II. BMC Medical of a type 1 diabetes locus in the IL2RA/
Genetics 9;2008:65. CD25 region by use of tag single-nucle-
Sacks DB, and McDonald JM. The pathogenesis otide polymorphisms. Am J Hum Genet
of type II diabetes mellitus: A polygenic dis- 76(5);2005:773–9.
ease. Am J Clin Pathol 105(2);1996:149–56. Wang G, Wang X, Zhang Q et al. Response to
Seshiah V, Balaji V, Balaji MS et al. Gestational pioglitazone treatment is associated with the
diabetes mellitus in India. J Assoc Physicians lipoprotein lipase S447X variant in subjects
India 52;2004:707–11. with type 2 diabetes mellitus. Int Journal Clin
Shaat N, Karlsson E, Lernmark Å et al. Common Pract 61;2007:552–7.
variants in MODY genes increase the risk of Weeks DE, and Lathrop GM. Polygenic diseases:
gestational diabetes mellitus. Diabetologia Methods for mapping complex disease traits.
49(7);2006:1545–51. Trends Genet 11;1995:513–9.
Shastry BS. SNP alleles in human disease and World Health Organization, Global
evolution. J Hum Genet 47(11);2002:561–6. report on diabetes, 2016. Available
Shu Y, Sheardown SA, Brown C et al. Effect of at http://apps.who.int/iris/bitstr
genetic variation in the organic cation trans- eam/10665/204871/1/9789241565257_eng.
porter 1 (OCT1) on metformin action. J Clin pdf. Accessed October 4, 2017.
Invest 117;2007:1422–31. World Health Organization, Diabetes fact sheet,
Steck AK, Liu SY, McFann K et al. Association of 2017. Available at http://www.who.int/
the PTPN22/LYP gene with type 1 diabetes. mediacentre/factsheets/fs312/en/. Accessed
Pediatr Diabetes 7;2006:274–8. October 4, 2017.
Stoffel M, Bell KL, Blackburn CL et al. World Health Organization, India, 2016. World
Identification of glucokinase mutations in Health Day. Available at http://www.searo.
subjects with gestational diabetes mellitus. who.int/india/mediacentre/events/2016/en/.
Diabetes 42;1993:937–40. Accessed November 13, 2017.
Tang Y, Axelsson AS, Spégel P et al. Genotype- Zyriax B-C, Salazar R, Hoeppner W et al. The
based treatment of type 2 diabetes with an association of genetic markers for type 2
α2A-adrenergic receptor antagonist. Sci diabetes with prediabetic status—Cross-
Transl Med 6;2014:257ra139. sectional data of a diabetes prevention Trial.
Ueda H, Howson JMM, Esposito L et al. PLoS ONE 8(9);2013:e75807.
Association of the T-cell regulatory gene
 15
Precision medicine in multiple sclerosis

SHOAIB AHMAD

Background: Disease prevalence, incidence, Treatment options for MS 271

and mortality rate (preferably worldwide) 269 Why genetic, environmental, lifestyle, and
Introduction 269 other factors should be considered in
Incidence 270 precision medicine and how they impact
Age considerations 270 on predicting, preventing, managing,
Gender considerations 270 and treating the disease 273
Ethnic considerations 270 Lifestyle modification strategy 273
Genetics 270 Why personalized and precision medicine
Pathological basis 270 are needed and have potential in
Approaches currently used for the disease improving health 273
diagnosis and treatment (markers, Precision medicine approach 273
genomics, proteomics, epigenomics, General consideration with precision medicine 273
imaging, or any other relevant Challenges and research opportunities
technologies/approaches) 270 (future directions) 274
Diagnosis 270 Conclusion 275
Biomarkers 271 Acknowledgments 275
Biomarkers for diagnosis 271 References 275

BACKGROUND: DISEASE normal circumstances, myelin protects the nerve

PREVALENCE, INCIDENCE, AND fibers of the CNS involved in the conduction of
MORTALITY RATE (PREFERABLY electrical impulses. In MS, myelin covering is lost
WORLDWIDE) or damaged leaving behind the scar tissue (plaques
or lesions). In the pathological condition, the
Introduction impaired or disrupted nerve conduction then pro-
duces the characteristic symptoms of MS (Frankel
Multiple sclerosis is a disease affecting the central and Jones, 2017).
nervous system. This disease involves the scarring MS has been classified as a highly heteroge-
or hardening of patches on nervous tissue of the neous disease (Ziemssen et  al., 2016). MS is a
brain and spinal cord at multiple locations in the chronic pathophysiological condition. The general
body (Multiple Sclerosis Trust [MS Trust], 2017). perception of MS being a fatal, infectious, or con-
Multiple sclerosis (MS) is often described as a tagious disease is false (MS Trust, 2017). MS has
chronic autoimmune disease. It affects the central also been debated/discussed in books (Fletcher,
nervous system (Lugaresi et al., 2013), including 2000) and challenges in MS have been reviewed
the brain, spinal cord, and the optic nerves. Under (Hohlfeld, 2010).

269
270  Precision medicine in multiple sclerosis

Incidence Table 15.2  Genes doubted to be associated with

MS
More than 2.3 million people across the globe are
affected by MS (Miller and Leary, 2007; Frankel Doubtful genes
and Jones, 2017). Nearly 400,000 people with MS APOE CNTF HLA-DRB1*1501 MEFV
reside in the United States alone (Markowitz, CCR5 CRYAB IFNg OPN
2013). Another study has pointed out that MS CD24 ESR1 IL1B PD-1
affects approximately 300,000 U.S. citizens (Leist CD59 GSTM IL4 TGFB1
et  al., 2014). Over 100,000 people in the United
Kingdom are affected by MS (MS Trust, 2017).
Brain atrophy is involved in MS (Chard et al., 2004;
Pagani et  al., 2005; Tedeschi et  al., 2005; Bieniek
Age considerations et al., 2006; Charil et al., 2007; Grassiot et al., 2009;
MS is detected in most people when they are in De Stefano et al., 2010; Calabrese et al., 2007; Jones
their 20s or 30s (MS Trust, 2017). et  al., 2013; Pérez-Miralles et  al., 2013; Popescu
et al., 2013).
Meningeal inflammation has been shown to play
Gender considerations an important role in primary progressive multiple
It is known fact that 75% of MS patients are females sclerosis (Choi et  al., 2012). Meningeal B-cell fol-
and 25% are males (MS Trust, 2017). licles are associated with early onset of MS and its
related progression (secondary progressive multiple
sclerosis) as reported by Magliozzi et al. (2007).
Ethnic considerations Invasion of autoreactive T-cells and alterations
MS commonly affects persons from ethnic groups of the blood-brain barrier (BBB) represent early
(Frankel and Jones, 2017). These groups are listed pathological stages that are characterized by the
in Table 15.1. following events (Smorodchenko et al., 2007):

GENETICS 1. Autoreactive T-cells invasions

The forkhead/winged helix gene (FOXP3) is involved 2 . Changes in blood-brain barrier
in MS (Gholami et al., 2017). Interferon regulatory
factor 5 (IRF5) has been validated as the multiple MS patients are affected by aberrant activation
sclerosis risk gene (Vandenbroeck et al., 2011). The of helper T lymphocytes. It leads to the demy-
SLC9A9 gene is implicated in MS (Esposito et  al., elination of neurons and manifests in the form
2015). Several genes (Table 15.2) have been associ- of impaired CNS function as well as mobility.
ated with MS (Oksenberg and Hauser, 2008). Neuronal scarring or sclerosis appears as plaques
in MRI scans (Javan et al., 2017).
PATHOLOGICAL BASIS
Neuropathological basis of disease progression APPROACHES CURRENTLY USED
has been described in detail (Reynolds et al., 2011). FOR THE DISEASE DIAGNOSIS
Table 15.1  Ethnic communities in the United
AND TREATMENT (MARKERS,
States prone to the risk of MS GENOMICS, PROTEOMICS,
EPIGENOMICS, IMAGING, OR ANY
Ethnic community Remarks OTHER RELEVANT TECHNOLOGIES/
African-American More severe form of MS APPROACHES)
has been recorded
Asian — Diagnosis
Hispanic —
There is no single test that can detect MS. Red
Latino —
blood cell distribution width is an easy parameter
Caucasian Commonly originating for diagnosis of MS (Hu, 2016). The correlation
from Northern Europe of the clinical conditions and the test reports can
 Approaches currently used for the disease diagnosis and treatment  271

Table 15.3  Diagnostic techniques for MS

Simple diagnosis Specialist diagnosis

Medical history (focus on neurological aspects Lumbar puncture (or spinal tap)
and familial history)
Blood tests Visual evoked potentials (VEP)
Thorough neurological exam Somatosensory evoked potentials (SEPs)
Magnetic resonance imaging

often help in early detection of MS usually with the 1. Genetic/immunogenetic biomarkers

involvement of specialist physicians (Frankel and 2 . Laboratorial biomarkers in body fluids
Jones, 2017). The diagnostic techniques have been 3. Imaging: biomarkers
depicted in Table 15.3.
Magnetic resonance imaging (MRI) can be used Development of biomarkers for MS has been
for diagnosis of MS. MRI can be accomplished by discussed in detail (Bielekova and Martin, 2004).
two methods (Fisniku et al., 2008): Some of the important biomarkers for MS are
listed in Table 15.4.
1. Gadolinium contrast enhancement BIOMARKERS FOR DIAGNOSIS
2 . T2-hyperintense lesion load
Established biomarkers and potential markers
have been elaborately described by Derfuss (2012)
As per the National Institute for Health and Care
as described in Tables 15.5 and 15.6.
Excellence (NICE) guidelines, diagnosis of MS can
be based on the following blood tests (National
Institute for Health and Care Excellence, 2017): Treatment options for MS

1. Complete blood count (Hb, TLC, DLC, etc.) Therapeutic modalities for MS can be typically
2 . HIV serology (ELISA, RIA) categorized into two general groups (Frankel and
3. Inflammatory markers (ESR, C-reactive Jones, 2017; also see Table 15.7):
protein) The first group—The disease-modifying medi-
4. Liver function test (SGOT, SGPT, alkaline cations approved by the Food and Drug
phosphatase) Administration (FDA) are used to slow or
5. Renal function test hamper the disease progression.
6. Serum calcium
7. Serum glucose
Table 15.4  Biomarkers for MS
8. Thyroid function tests (T3, T4, TSH)
9. Vitamin B12 Biomarker Reference
Neurofilament light Malmeström et al.
Optic coherent tomography (OCT) can be used
protein (2003)
for detecting MS (Villoslada, 2010). OCT can help
Glial fibrillary acidic Malmeström et al.
in determining microcystic macular edema in the
protein (2003)
retina (Lang et al., 2015). Noninvasive OCT is used
N-acetylaspartate Trentini et al. (2014)
in detecting ocular diseases such as nystagmus
(Antony et  al., 2016). Volumetric assessment of Neurofilaments Trentini et al. (2014)
retinal ganglion cell layer by OCT can be used for Circulating T cell Pender et al. (2003)
diagnosing MS (Davies et al., 2011). Myelin basic protein Belogurov et al. (2008)
peptides
BIOMARKERS Circulating microRNAs Vistbakka et al. (2016)
Sphingosine-1- Brana et al. (2014)
Katsavos and Anagnostouli (2013) have given an
phosphate
elaborate description of biomarkers for MS and
receptors
have classified them into three groups:
272  Precision medicine in multiple sclerosis

Table 15.5  Established biomarkers for diagnosis Table 15.7  Drugs used in treatment of MS

Biomarkers Category Drug

(without clear
Corticosteroids Methylprednisolone
Valid biomarkers evidence)
Prednisone
1. Cerebral spinal fluid- 1. Neutralizing Muscle relaxants Baclofen
specific oligoclonal antibodies Tizanidine
bands against Dantrolene
2. Intrathecal beta- Onabotulinumtoxin A
immunoglobulin interferon Nerve conduction Fampridine
production 2. Neutralizing enhancers
3. Intrathecal antiviral antibodies Fatigue reducers Amantadine
immunoglobulin against Modafinil
production natalizumab Armodafinil
4. Magnetic resonance 3. Antibodies Bladder medications Tolterodine tartrate
imaging against JC Oxybutynin
5. Aquaporin 4 antibodies virus Mirabegron
Trospium chloride
Fesoterodine
Table 15.6  Potential biomarkers
Solifenacin succinate
Potential Medications for Carbamazepine
biomarkers Potential biomarkers paresthesias Amitriptyline
(valid) (without clear evidence) Gabapentin
1. CD56 bright 1. Cytokines/chemokines Pregabalin
natural killer 2. Myelin oligodendrocyte Duloxetine hydrochloride
cells glycoprotein antibodies Mood stabilizing Divalproex sodium
2. Genetics 3. Intrathecal/oligoclonal medications
immunoglobulin M Medication for Combination of
production pseudobulbar dextromethorphan and
4. Transcriptomics affect quinidine
Source: Based on Frankel D, and Jones H, Living with
multiple sclerosis [brochure], available at
nationalmssociety.org. Accessed on March 11,
The second group—Includes drugs and tech- 2017.
niques that contribute toward improvement
or alleviation of symptoms. In fact, symptom Established biomarkers (with valid evidence)
management is crucial for living well with MS. 1. Neutralizing antibodies against
natalizumab
Differences in ethnicity also have the potential to 2. Antibodies against JC virus
affect the pharmacokinetic profile of the drugs (Wu Established biomarkers (with some experimental/
et  al., 2012). Interferon regulatory factor 5 (IRF5) clinical evidence)
gene can affect IFNβ therapy in MS (Vandenbroeck 1. Magnetic resonance imaging
et  al., 2011). IRF5 variants affect clinical out- 2. Neutralizing antibodies against
comes of IFNβ therapy (Vosslamber et  al., 2011). beta-interferon
Pharmacogenomic analysis of response to IFNβ 3. Aquaporin 4 antibodies
therapy in MS has been reported (Byun et al., 2008). Potential biomarkers (with some experimental/
Anti-John Cunningham (anti-JC) virus antibodies clinical evidence)
are utilized in the safety monitoring of natalizumab 1. CD56 bright natural killer cells
therapy (Bloomgren et al., 2012). Several biomark- 2. Cytokines/chemokines
ers have been identified for MS treatment (Derfuss, 3. Transcriptomics
2012) and are classified as follows: 4. Genetics
 Why personalized and precision medicine are needed and have potential in improving health  273

WHY GENETIC, ENVIRONMENTAL, attention, and so forth. Mood changes need

LIFESTYLE, AND OTHER FACTORS talk therapy.
SHOULD BE CONSIDERED IN
PRECISION MEDICINE AND HOW Physical therapies can be used for exercises,
THEY IMPACT ON PREDICTING, endurance, and stamina. Occupational therapies
PREVENTING, MANAGING, AND help people achieve maximum independence and
optimal functioning. Speech therapy can help in
TREATING THE DISEASE
overcoming problems in speech or chewing and
Studies have indicated that obesity, serum vita- swallowing.
min D levels, exposure to sunlight, and smoking
may impact the disease progression (Ascherio WHY PERSONALIZED AND
and Munger, 2007). MRI is often used in the MS
PRECISION MEDICINE ARE
diagnosis. Patient’s age and gender can affect MRI
measures (Tintore et al., 2015).
NEEDED AND HAVE POTENTIAL IN
IMPROVING HEALTH
Lifestyle modification strategy Precision medicine approach
The patients need to develop healthy living hab- Personalized treatment has been discussed in MS
its and take diets that do not cause constipation. (Correale, 2011). Personalized treatment for MS is
Dietary changes, lifestyle modifications, and the need of the hour (Derfuss, 2012). Personalized
replenishment of fluid are proven methods to treatment of MS is in focus (Giovannoni and
­overcome constipation. One needs to undertake Rhoades, 2012). A combination of clinical fea-
exercises for the proper stress management. A bent tures and MRI parameters can help in finding the
toward spiritual life may also help. The necessary current state of disease and its progression or its
psychological support (from all family members, response to medication (Derfuss, 2012). Precision
friends, and peer group) can also play an important medicine for MS has witnessed a paradigm shift
part in the betterment of living with MS (Frankel (Dobrota et  al., 2014). Shifting paradigms in MS
and Jones, 2017). have been debated (Golan et al., 2016).
Major recommendations in lifestyle modifica-
tion are as follows:
General consideration with precision
1. Adequate intake of fluids helps in prevention of medicine
bladder complications.
2 . One needs to avoid unsafe exposure to harmful Multidisciplinary care interventions and reha-
UV rays. bilitation of MS patients has assumed a greater sig-
3. One is required to ensure proper and adequate nificance (Skovgaard et  al., 2012). Nintendo Wii (a
amounts of vitamin D. physiotherapy modality) can be used for manag-
4. Smoking is to be avoided at all costs. Active ing MS (Thomas et  al., 2014). Personalized train-
as well as passive smoking is to be reduced or ing games can be used in the rehabilitation of
eliminated to the best possible extent. MS patients (Octavia and Coninx, 2014). Patient-
5. Stretching exercises can help reduce spas- reported outcomes are an integral part of precision
ticity or muscle stiffness. Aerobic exercise medicine for MS (Lejbkowicz et  al., 2012). Self-
and energy-management strategies help in management and satisfaction play an important role
overcoming fatigue, which is one of the most in MS (Giovannoni and Rhoades, 2012). Multiple
important features of MS. Pelvic floor physi- sclerosis-associated retroviruses specifically need
cal therapy (popularly known as Kegel exer- personalized attention (Curtin et al., 2015).
cises) can help in overcoming genitourinary During recent years, precision medicine has
problems. become the focus of medical advancement (Willis
6. Cognitive problems can be managed with reha- and Lord, 2015). It has expedited the development
bilitation/training. Cognitive rehabilitation of novel therapeutic agents (D’Amico et al., 2016)
aids in the matters related to improve memory, as listed in Table 15.8.
274  Precision medicine in multiple sclerosis

Table 15.8  Therapeutic agents under investigation

Therapeutic agents Potential benefit(s)

Adrenocorticotropic hormone Anti-inflammatory effects
Biotin Reducing disease progression
Amiloride Neuroprotective effects
Fluoxetine Suppression of microglia activation
Riluzole Antiglutamatergic effects
LINGO-1 Inhibition of oligodendrocyte precursor cells differentiation
Domperidone Increase in prolactin secretion
Erythropoietin Neuroprotective abilities
Hematopoetic stem cell transplantation Resetting of immune system
Ibudilast Reduction in tumor necrosis factor levels
Idebenone Antioxidant properties
Lipoic acid Antioxidant properties
Lithium Regulation of inflammatory processes
Masitinib Degranulation of mast cells
MIS416 Modulation of t-cell-mediated autoimmune responses
NeuroVax Vaccine strategies
Oxcarbazepine Neuroprotective properties
Simvastatin Immunomodulatory and neuroprotective effects
Siponimod (BAF312) Antitumor, antioxidant, and anti-inflammatory benefits
Sunphenon epigallocatechin-3-gallate Antioxidant, and anti-inflammatory effects
(EGCg)
Source: D’Amico E et al., Int J Mol Sci 17;2016:1725, doi:10.3390/ijms17101725.

Antibodies also need to be explored for thera- imaging, or immunological nature need to be
peutic potential (D’Amico et  al., 2016). Presently developed. Academic and industry cooperation is
two types of monoclonal antibodies are under now needed to bring these markers into ­clinical
investigation: practice (Derfuss, 2012).
Omics technologies in MS (Table 15.9)
1. Monoclonal antibodies (for natalizumab and include the following techniques (Katsavos and
alemtuzumab) Anagnostouli, 2013):
2 . Anti-CD20 monoclonal antibodies (for ritux-
imab and ocrelizumab) 1. Genomics is used for whole DNA sequencing.
2 . Transcriptomics is used for studying RNA
sequences. It is commonly exemplified by
Challenges and research microarrays and next-generation sequencing.
opportunities (future directions) 3. Proteomics is used for investigating distribu-
tion of specific proteins.
Three challenges have been identified for personal- 4. Lipidomics is concerned with recognition of
ized medicine of MS (Gafson et al., 2017): cellular lipid pathways.
5. Metabolomics can be utilized for gaining insight
1. Precision diagnosis into metabolic pathways implicated in MS.
2 . Prediction of therapeutic response 6. Epigenomics is focused on epigenetic modifica-
3. Personalized monitoring of prediction tions that bring MS susceptibility.

Personalized medicine for MS is an uphill and Pharmacogenomics needs to be explored

challenging task. Biomarkers of clinical, genetic, beyond routine genotypic profiling. It can be used
 References 275

Table 15.9  Applications of omics in MS edema in SDOCT volumes. Proc SPIE Int Soc
Opt Eng 2016:9784, doi: 10.1117/12.2214676.
Technology Specific application in MS
Ascherio A, and Munger KL. Environmental risk fac-
Genomics Recognition of vitamin D tors for multiple sclerosis. Part II: Noninfectious
receptor element factors. Ann Neurol 61;2007:504–13.
Transcriptomics TOB-1 gene downregulation Belogurov AA, Kurkova IN, Friboulet A et al.
Proteomics Ninjurin-1 Recognition and degradation of myelin basic
Lipidomics CNS lipid epitopes, IgM protein peptides by serum autoantibod-
against myelin lipids ies: Novel biomarker for multiple sclerosis.
Metabolomics N-acetyloaspartate in CSF J. Immunol 180;2008:1258–67.
Epigenomics Chromatin architecture in MS Bielekova B, and Martin R. Development of
biomarkers in multiple sclerosis. Brain
127;2004:1463–78.
to determine metabolism of drugs such as inter- Bieniek M, Altmann DR, and Davies GR. Cord
ferons (IFNs). It also implies that this approach atrophy separates early primary progres-
can be used to find the drugs of choice when MS sive and relapsing remitting multiple
patients present with comorbidities (Hegen et al., sclerosis. J Neurol Neurosurg Psychiatry
2015). 77;2006:1036–9.
Bloomgren G, Richman S, Hotermans C et al. Risk
CONCLUSION of natalizumab-associated progressive mul-
tifocal leukoencephalopathy. N Engl J Med
Multiple sclerosis is a heterogeneous disease of 366;2012:1870–80.
the central nervous system resulting from demy- Brana C, Frossard MJ, Pescini Gobert R,
elination of the nervous tissue. MS is thought to Martinier N, Boschert U, and Seabrook TJ.
result from a faulty FOXP3 gene. No single diag- Immunohistochemical detection of sphingo-
nostic test exists for MS. Physicians often need to sine-1-phosphate receptor 1 and 5 in human
rely on correlation between the clinical features multiple sclerosis lesions. Neuropathol Appl
and findings from magnetic resonance imaging. Neurobiol 40;2014:564–78.
Corticosteroids, interferons, and other immuno- Byun E, Caillier SJ, Montalban X et al.
therapies need to be supplemented with stem cell Genomewide pharmacogenomic analy-
therapy/gene therapy. Omics can definitely play a sis of the response to interferon beta
role in the diagnosis and treatment of MS. Better therapy in multiple sclerosis. Arch Neurol
biomarkers can help in early detection of the dis- 65;2008:337–44.
ease. MS will undergo a paradigm shift with the Calabrese M, Atzori M, Bernardi V et al.
introduction of personalized or individualized Cortical atrophy is relevant in mul-
medicine in MS. tiple sclerosis at clinical onset. J Neurol
254;2007:1212–20.
ACKNOWLEDGMENTS Chard DT, Griffin CM, and Rashid W. Progressive
grey matter atrophy in clinically early relaps-
The author acknowledges the continuous encour- ing-remitting multiple sclerosis. Mult Scler
agement and moral support from his teachers at 10;2004:387–91.
Jamia Hamdard, New Delhi (Dr. Shibli Jameel and Charil A, Dagher A, Lerch JP. Focal corti-
Dr. R. Zafar) and his family members (particularly cal atrophy in multiple sclerosis: Relation
Mrs. Zeba and Dr. Sultan Anjum). to lesion load and disability. NeuroImage
34;2007:509–17.
REFERENCES Choi SR, Howell OW, Carassiti D, Magliozzi R,
Gveric D, Muraro PA, Nicholas R, Roncaroli
Antony BJ, Lang A, Swingle EK, Al-Louzi O, F, and Reynolds R. Meningeal inflamma-
Carass A, Solomon S, Calabresi PA, Saidha tion plays a role in the pathology of pri-
S, and Prince JL. Simultaneous segmentation mary progressive multiple sclerosis. Brain
of retinal surfaces and microcystic macular 135;2012:2925–37.
276  Precision medicine in multiple sclerosis

Correale J. Personalized treatment in mul- Giovannoni G, and Rhoades RW. Individualizing

tiple sclerosis. Curr Neurol Neurosci Rep treatment goals and interventions for people
11(6);2011:523–5. with MS. Curr Opin Neurol 25(Suppl);2012:
Curtin F, Perron H, Faucard R, Porchet H, and S20–7.
Lang AB. Treatment against human endog- Golan D, Staun-Ram E, and Miller A. Shifting
enous retrovirus: A possible personalized paradigms in multiple sclerosis: From dis-
medicine approach for multiple sclerosis. Mol ease-specific, through population-specific
Diagn Ther 19(5);2015:255–65. toward patient-specific. Curr Opin Neurol
D’Amico E, Patti F, Zanghì A, and Zappia M. A 29(3);2016:354–61.
personalized approach in progressive multiple Grassiot B, Desgranges B, Eustache F, and Defer
sclerosis: The current status of disease modi- G. Quantification and clinical relevance of
fying therapies (DMTs) and future perspec- brain atrophy in multiple sclerosis: A review.
tives. Int J Mol Sci 17;2016:1725, doi: 10.3390/ J Neurol 256;2009:1397–412.
ijms17101725. Hegen H, Auer M, and Deisenhammer F.
Davies EC, Galetta KM, Sackel DJ, Talman LS, Pharmacokinetic considerations in the treat-
Frohman EM, Calabresi PA, Galetta SL, and ment of multiple sclerosis with interferon-
Balcer LJ. Retinal ganglion cell layer volumet- beta. Expert Opin Drug Metab Toxicol
ric assessment by spectral-domain optical 11;2015:1803–19.
coherence tomography in multiple sclero- Hohlfeld R. “Gimme five”: Future challenges in
sis: Application of a high-precision manual multiple sclerosis. ECTRIMS Lecture 2009.
estimation technique. J Neuroophthalmol Mult Scler 16(1);2010:3–14.
31(3);2011:260–4. Hu Z. Red blood cell distribution width: A promis-
Derfuss T. Personalized medicine in multiple ing index for estimating activity of autoim-
sclerosis: Hope or reality? BMC Medicine mune disease. J Lab Precis Med 1;2016:4–9.
10;2012:116–20. Javan MR, Jalali Nezhad A, Safa A, Mohammadi
Dobrota R, Mihai C, and Distler O. Personalized MH, and Jamebozorgi K. Personalized medi-
medicine in systemic sclerosis: Facts and cine toward multiple sclerosis; current chal-
promises. Curr Rheumatol Rep 16(6);2014:425. lenges and future prospects. Int J Basic Sci
Esposito F, Sorosina M, Ottoboni L et al. A Med 2(1);2017:11–15.
pharmacogenetic study implicates SLC9a9 in Jones BC, Nair G, Shea CD, Crainiceanu CM,
multiple sclerosis disease activity. Ann Neurol Cortese IC, and Reich DS. Quantification
78;2015:115–27. of multiple-sclerosis-related brain atrophy
Fisniku LK, Brex PA, Altmann DR et al. Disability in two heterogeneous MRI datasets using
and T2 MRI lesions: A 20-year follow-up of mixed-effects modeling. NeuroImage Clin
patients with relapse onset of multiple sclero- 3;2013:171–9.
sis. Brain 131;2008:808–17. Katsavos S, and Anagnostouli M. Biomarkers in
Fletcher P. Do NICE and CHI have no interest in multiple sclerosis: An up-to-date overview.
safety? Opinion of the book NICE, CHI and Multiple Sclerosis Int 2013, article ID 340508.
the NHS reforms. Enabling excellence or http://dx.doi.org/10.1155/2013/340508.
imposing control? Adverse Drug React Toxicol Lang A, Carass A, Swingle EK, Al-Louzi O,
Rev 19(3);2000:167–76. Bhargava P, Saidha S, Ying HS, Calabresi PA,
Frankel D, and Jones H. Living with multiple and Prince JL. Automatic segmentation of
sclerosis [brochure]. Available at nationalms- microcystic macular edema in OCT. Biomed
society.org. Accessed on March 11, 2017. Opt Express 6(1);2015:155–69.
Gafson A, Craner MJ, and Matthews PM. Leist T, Hunter SF, Kantor D, and Markowitz
Personalised medicine for multiple sclerosis C. Novel therapeutics in multiple sclerosis
care. Multiple Sclerosis J 23(3);2017:362–9. management: Clinical applications. Am J Med
Gholami M, Darvish  H, Ahmadi H, Rahimi- 127(1);2014:S2.
Aliabadi S, Emamalizadeh B, Amirabadi Lejbkowicz I, Caspi O, and Miller A.
MRE, Jamshidi J, and Movafagh A. Iran Red Participatory medicine and patient empow-
Crescent Med J 19(1);2017:e34597. erment towards personalized healthcare
 References 277

in multiple sclerosis. Expert Rev Neurother atrophy and lesion load predict long term
12(3);2012:343–52. disability in multiple sclerosis. J Neurol
Lugaresi A, di Ioia M, Travaglini D, Pietrolongo E, Neurosurg Psychiatry 84;2013:1082–91.
Pucci E, and Onofrj M. Risk-benefit consider- Pérez-Miralles F, Sastre-Garriga J, Tintoré M
ations in the treatment of relapsing-remitting et al. Clinical impact of early brain atrophy
multiple sclerosis. Neuropsychiatr Dis Treat in clinically isolated syndromes. Mult Scler
9;2013:893–914. 19;2013:1878–86.
Magliozzi R, Howell O, Vora A, Serafini B, Reynolds R, Roncaroli F, Nicholas R, Radotra B,
Nicholas R, Puopolo M, Reynolds R, and Aloisi Gveric D, and Howell O. The neuropathologi-
F. Meningeal B-cell follicles in secondary pro- cal basis of clinical progression in multiple
gressive multiple sclerosis associate with early sclerosis. Acta Neuropathol 122;2011:155–70.
onset of disease and severe cortical pathol- Skovgaard L, Bjerre L, Haahr N et al. An inves-
ogy. Brain 130;2007:1089–104. tigation of multidisciplinary complex health
Malmeström C, Haghighi S, Rosengren L, care interventions—steps towards an inte-
Andersen O, and Lycke J. Neurofilament grative treatment model in the rehabilita-
light protein and glial fibrillary acidic pro- tion of people with multiple sclerosis. BMC
tein as biological markers in MS. Neurology Complement Altern Med 12;2012:50.
61;2003:1720–5. Smorodchenko A,Wuerfel J, Pohl EE, Vogt J,
Markowitz CE. Multiple sclerosis update. Am J Tysiak E, Glumm R, Hendrix S, Nitsch R, Zipp
Manag Care 19(16 Suppl);2013:s294–300. F, and Infante-Duarte C. CNS-irrelevant
Miller DH, and Leary SM. Primary-progressive T-cells enter the brain, cause blood-brain bar-
multiple sclerosis. Lancet Neurol rier disruption but no glial pathology. Eur J
6;2007:903–12. Neurosci 2007;26(6):1387–98.
Multiple Sclerosis Trust (MS Trust). 2017. What is De Stefano N, Giorgio A, and Battaglini M.
multiple sclerosis? Poster, available at mstrust. Assessing brain atrophy rates in a large
org.uk. Accessed on March 11, 2017. population of untreated multiple sclerosis
National Institute for Health and Care Excellence subtypes. Neurology 74;2010:1868–76.
(NICE). 2017. Diagnosing multiple sclerosis. Tedeschi G, Lavorgna L, and Russo P. Brain
Available at https://pathways.nice.org.uk/ atrophy and lesion load in a large population
pathways/multiple-sclerosis. Accessed on of patients with multiple sclerosis. Neurology
March 11, 2017. 65;2005:280–5.
Octavia JR, and Coninx K. Adaptive person- Thomas S, Fazakarley L, Thomas PW, Brenton
alized training games for individual and S, Collyer S, Perring S, Scott R, Galvin K,
collaborative rehabilitation of people and Hillier C. Testing the feasibility and
with multiple sclerosis. Biomed Res Int acceptability of using the Nintendo Wii in
2014;2014:345728. the home to increase activity levels, vital-
Oksenberg J, and Hauser SL. Genetics of mul- ity and well-being in people with multiple
tiple sclerosis. In: Multiple Sclerosis. Raine sclerosis (Mii-vitaliSe): Protocol for a pilot
CS, McFarland H, Hohlfeld R (Eds). Saunders randomised controlled study. BMJ Open
Elsevier, London; 2008, pp. 214–25. 4(5);2014:e005172.
Pagani E, Rocca MA, and Gallo A. Regional brain Tintore M, Rovira A, Rio J et al. Defining high,
atrophy evolves differently in patients with medium and low impact prognostic fac-
multiple sclerosis according to clinical pheno- tors for developing multiple sclerosis. Brain
type. AJNR 26;2005:341–6. 138;2015:1863–74.
Pender MP, Csurhes PA, Wolfe NP, Hooper KD, Trentini A, Comabella M, Tintoré M et al.
Good MF, McCombe PA, and Greer JM. N-acetylaspartate and neurofilaments as
Increased circulating T cell reactivity to GM3 biomarkers of axonal damage in patients
and GQ1b gangliosides in primary pro- with progressive forms of multiple sclerosis.
gressive multiple sclerosis. J Clin Neurosci J Neurol 261;2014:2338–43.
10;2003:63–6. Vandenbroeck K, Alloza I, Swaminathan B
Popescu V, Agosta F, Hulst HE et al. Brain et al. Validation of IRF5 as multiple sclerosis
278  Precision medicine in multiple sclerosis

risk gene: Putative role in interferon beta sclerosis. Genes Immun 12;2011:466–72.
therapy and human herpes virus-6 infection. Willis JC, and Lord GM. Immune biomarkers: The
Genes Immun 12;2011:40–45. promises and pitfalls of personalized medi-
Villoslada P. Biomarkers for multiple sclerosis. cine. Nat Rev Immunol 15;2015:323–29.
Drug News Perspect 23(9);2010:585–95. Wu K, Mercier F, David OJ et al. Population
Vistbakka J, Elovaara I, Lehtimäki T, and Hagman pharmacokinetics of fingolimod phosphate
S. Circulating microRNAs as biomarkers in in healthy participants. J Clin Pharmacol
progressive multiple sclerosis. Mult Scler J 52;2012:1054–68.
23(3);2016:403–12. Ziemssen T, Kern R, and Thomas K. Multiple
Vosslamber S, van der Voort LF, van den Elskamp sclerosis: Clinical profiling and data collec-
IJ et al. Interferon regulatory factor 5 gene tion as prerequisite for personalized medicine
variants and pharmacological and clinical approach. BMC Neurol 16;2016:124.
outcome of Interferonbeta therapy in multiple
 16
Precision medicine in asthma and
chronic obstructive pulmonary disease

SHOAIB AHMAD

Introduction 279 Approaches currently used for the disease

Precision medicine in asthma 280 diagnosis and treatment 285
Introduction 280 Pathogenesis of COPD 285
Epidemiology 280 Diagnosis 285
Molecular pathogenesis 280 Classification of diagnostic techniques 285
Approaches currently used for asthma COPD phenotypes 285
diagnosis and treatment 281 Therapeutic options for the treatment
Classification of asthma phenotypes 281 of COPD 285
New approaches in asthma Pulmonary rehabilitation 286
phenotyping 281 Vaccinations 286
Asthma treatment 281 Oxygen therapy 286
Asthma medication 281 Pharmacotherapy 286
Biological agents 282 Exacerbation management 286
Novel pharmacotherapeutic approaches 282 Exacerbation pharmacotherapy 286
Bronchial thermoplasty 282 Monitoring and follow-up 287
Genetics of asthma 282 Genetic linkage 287
Why personalized and precision medicine Novel approaches in the management
are needed and have potential in of COPD 287
improving health 283 Biomarkers 288
Biomarkers 283 Epigenetic biomarkers 288
Challenges and research opportunities 284 Challenges and research opportunities 288
Precision medicine for chronic obstructive Gene expression analysis 288
pulmonary disease (COPD) 284 Conclusions 288
Introduction 284 Acknowledgments 289
Occurrence 284 References 289

INTRODUCTION These two diseases are growing at an alarm-

ing level. Though the number of hospitalizations
Asthma and chronic obstructive pulmonary dis- and mortalities are on a downward trend, the loss
ease (COPD) are respiratory diseases. They are of human life and/or functionality and economic
more common in developed countries (Agusti burden compels us to look for better vistas for the
et al., 2016a, 2016b). clinical management of asthma and COPD.

279
280  Precision medicine in asthma and chronic obstructive pulmonary disease

Clinical and immunological biomarkers can be Table 16.1  Prevalence of asthma in United
used for prognosis, diagnosis, and prediction of States and Europe
therapeutic response for asthma as well as COPD.
Country Affected population
Biomarkers can also be used for drug targeting and
identifying the patients who can respond to novel United States 25 million
therapeutics (Heaney and McGarvey, 2017). United Kingdom 4.67 million
Asthma and COPD may share biological Germany 3.27 million
mechanisms. There are similarities in clinical, France 2.28 million
functional, and diagnostic features. A precision Italy 2.26 million
medicine strategy for asthma and COPD has been Spain 1.58 million
proposed (Agusti et al., 2016a, 2016b).
Source: National Heart, Lung, and Blood Institute
(NHLBI), August 4, 2014, http://www.nhlbi.nih.
PRECISION MEDICINE IN ASTHMA gov/health/health-topics/topics/asthma/; Aubier
M et  al., The Uncovering Asthma Report,
Marlborough, MA: Boston Scientific, 2015.
Introduction
Asthma is a common disease characterized by Efforts have been made to locate sex-specific evi-
chronic inflammation of the airway. It mainly dence in asthma (Moerman et al., 2009). Asthma in
affects the respiratory system and a number of the elderly population has been discussed in detail
symptoms are produced including wheezing, (Gillman and Douglass, 2012). The Behavioral Risk
breath shortness, chest tightness, and coughing. Factor Surveillance System (BRFSS) can be used
These symptoms vary over time. The intensity for estimating the prevalence of asthma amongst
of the symptoms also undergoes a change as the specific races (Goodman, 2010). Strategies have
expiratory airflow changes (Global Initiative for been devised for multiple imputation in longitu-
Asthma [GINA], 2015). dinal epidemiological studies (Spratt et al., 2010).
Incidence of asthma and COPD is increasing
EPIDEMIOLOGY in societies with Western lifestyles (Braun and
Epidemiology of adult asthma is a challenge Tschernig, 2006).
(Toren, 1999). Asthma has been reported to affect
1%–18% of the people in different countries (GINA, MOLECULAR PATHOGENESIS
2015). Asthma affects 300 million people in the Bronchoconstriction is a characteristic symp-
world, and another 100 million people may be tom of COPD and asthma (Seehase et  al., 2011).
added by 2025 (American Thoracic Society [ATS], Erythrocyte glutathione peroxidase activity lev-
2015). Annually, 250,000 deaths across the globe are els are elevated in the case of asthma (Tho and
attributed to asthma (Bateman et al., 2008). Asthma Candlish, 1987). IL-16 is a proinflammatory cyto-
affects 8% (18.7 million) of the adult population in kine involved in the development of asthma. It is
the United States (Blackwell et  al., 2014). As per responsible for recruitment of CD4+ T cells to
the estimates of 2014, the United States has  more sites of asthma (Nicoll et al., 1999). IL-13 is a Th2
than 25 million people affected with asthma. Out of cytokine acting as an important mediator of air-
this, 7 million are children (National Heart, Lung, way inflammation and leading to asthma lesions.
and Blood Institute [NHLBI], 2014). Asthma costs IL-13 can be estimated using the Erenna immuno-
$56 billion on an annual basis. This burden includes assay system (St Ledger et al., 2009). TNFalpha and
medical costs, lost schooldays and workdays, and IL-13 are involved in activation of oxytocin recep-
early deaths (ATS, 2015). The total number of peo- tors. This activity in airway smooth muscle cells is
ple in Europe affected with severe asthma in 2015 implicated in asthma development (Amrani et al.,
was estimated to be 1.5 million (Chung et al., 2014). 2010). 20-HETE has been reported to play a pivotal
The economic burden of asthma in Europe was pre- role in mediation of acute ozone-induced airway
dicted at €19.3 billion in 2015 (Demoly et al., 2009; hyperresponsiveness (AHR; Cooper et al., 2010).
Aubier et  al., 2015). Asthma has become a chal- Phosphoinositide 3-kinase gamma (PI3Kgamma)
lenge for the Western world, particularly the United is involved in the pathogenesis of asthma.
States and Europe (Table 16.1). PI3Kgamma modulates calcium oscillations and
 Precision medicine in asthma  281

hence regulates contraction of the airway smooth Table 16.3  Well-recognized asthma phenotypes
muscles. PI3Kgamma inhibitors can help overcome
Phenotype Linkage
AHR (Jiang et  al., 2010). Prostaglandin D(2) is a
noninvasive biomarker of asthma and it has been Allergic asthma Allergies
found to cause contractions in peripheral lung tis- Nonallergic asthma No allergies
sue (Larsson et al., 2011). Trichostatin A (TSA) is Late-onset asthma Adult life
an inhibitor of histone deacetylase (HDAC), which Asthma with fixed Long-standing asthma
produces an inhibitory effect on airway hyperre- airflow limitation
sponsiveness (Banerjee et al., 2012). Asthma with obesity Obesity
APPROACHES CURRENTLY USED FOR Source: Global Initiative for Asthma (GINA), Pocket
ASTHMA DIAGNOSIS AND TREATMENT guide for asthma management and preven-
tion, 2015, www.ginasthma.org.
Guidelines on grading quality of evidence and sub-
sequent recommendations in clinical practice have
been reported (Brozek et al., 2009). The NHLBI has NEW APPROACHES IN ASTHMA
published Expert Panel Report 3 titled “Guidelines PHENOTYPING
for the Diagnosis and Management of Asthma” Phenotyping has been a much debated topic in asthma
(Hahn, 2009). Concerns have been raised on over- research. Several clinical phenotypes of asthma were
reliance on instrumental approach for disease reported (Bel, 2004). The need to shift to endotyping
identification (Baser, 2009). A scheme for diagno- instead of phenotyping has been suggested (Agache,
sis (Table 16.2) has been suggested (GINA, 2015). 2013). Severe asthma phenotypes have been identi-
Generalization of longitudinal annual FEV1 fied (Moore et al., 2010). A cluster analysis approach
decline (LLD) has been suggested. LLD is actu- has been used for asthma phenotyping (Haldar et al.,
ally based on spirometry precision (Hnizdo et al., 2008). Wheezing phenotypes have been associated
2007). The American Thoracic Society (ATS) with childhood lung function and airway responsive-
curves are considered sufficient for testing spirom- ness (Henderson et  al., 2008). Latent class analysis
eter efficacy (Lefebvre et al., 2014). The peak flow has been used for distinction between phenotypes
meters are handy diagnostic tools with some vari- of childhood wheezing and coughing (Spycher et al.,
ability in comparison to the Fleisch pneumotacho- 2008). Wheezing phenotypes in young preschool
graph. It clearly indicated that value scales of such children have been reported (Just et al., 2013).
flow meters required adjustments (van Schayck
et al., 1990).
Asthma treatment
CLASSIFICATION OF ASTHMA PHENOTYPES
Asthma is always in discussion for internal medi-
Recognizable patterns of demographic, clinical, cine specialists (John et  al., 2013). Atopic march
and pathological characteristics found in asthma also affects asthma (Ker and Hartert, 2009).
patients are often termed as “asthma phenotypes”
(Table 16.3) (GINA, 2015). ASTHMA MEDICATION
Asthma cannot be completely cured. It can very
well be controlled using medication and lifestyle
Table 16.2  Scheme for diagnosis of asthma
modification. Currently, there are two types of
Diagnosis steps treatments available for asthma. The first treat-
ment is concerned with regular medication and it
History and family history
is termed maintenance treatment. The second one
Physical examination
is concerned with emergency management and
Radiological examination
termed as rescue treatment (ATS, 2015). Most asth-
Lung function testing
matic patients need medicine that can effectively
Bronchial provocation tests
reduce airway inflammation and also be helpful
Allergy tests
in preventing the onset of severe asthma. Most of
Exhaled nitric oxide
these medications are, however, ineffective during
Differential diagnosis
an asthma attack (ATS, 2015).
282  Precision medicine in asthma and chronic obstructive pulmonary disease

The maintenance medicines include the follow- hyperresponsiveness, eosinophilia, and

ing treatments: T-helper type 2 cytokines in allergic asthma
(Erpenbeck et al., 2006).
1. Inhaled corticosteroids (most common and b. PI3Kgamma inhibitors—Treatment with a
nonhabit-forming substances) PI3Kgamma inhibitor can reduce AHR (Jiang
2 . Leukotriene modifiers et al., 2012).
3. Long-acting beta agonists c. Histamine and serotonin antagonists—A new
4. Long-acting anticholinergics compound (C-027) has been found to inhibit
5. Theophylline IgE-mediated bronchoconstriction. C-027 has
6. Shots been found to have action as a histamine and
serotonin antagonist in airways (Cooper et al.,
The rescue medicines include the following 2011).
treatments: d
. Potassium channel activators—KCNQ (Kv7)
potassium channel activators are nowadays
1. Short-acting beta2 agonists (help in opening the
debated as bronchodilators (Brueggemann
airway by respiratory muscle relaxant action)
et al., 2014).
2 . Ipratropium
e. Precision-guided immunotherapy—Cytotoxic
Eosinophils can be used as biomarkers of thera- T lymphocyte antigen 4 (CTLA4) Ig-modified
peutic response (Bagnasco et al., 2016). dendritic cells have beneficial effects on asthma.
CCR7 helps in dendritic cells homing. Both
BIOLOGICAL AGENTS molecules may be combined together for
Biologics are expensive therapeutic options. precision-guided immunotherapy (Chen et al.,
Presently, 21 biologics are under development for 2014).
use by patients with respiratory diseases including f. Phosphodiesterase-4 inhibitors—Roflumilast
asthma. Only one biologic agent—­omalizumab— is an upcoming drug for curing asthma (Ilango
is approved by the U.S. Food and Drug et al., 2013).
Administration (FDA) for asthma. Omalizumab
is indicated for severe, persistent allergic asthma, Nowadays, anti-IL-13 strategies make use of
which is ordinarily not con­trolled by inhalational anrukinzumab, lebrikizunab, and tralokinumab,
corticosteroids. Another biologic, mepolizumab, is whereas anti-IL-4 strategies include pascolizumab,
under consideration for FDA approval for use in pitakinra, and dupilumab (Bagnasco et al., 2016).
patients with severe eosinophilic asthma. Another
BRONCHIAL THERMOPLASTY
third biologic, lebrikizumab, is in clinical trials for
reducing airway inflammation (ATS, 2015). Bronchial thermoplasty (BT) is only intended for
severe asthma adult patients within a research
NOVEL PHARMACOTHERAPEUTIC environment (ATS, 2015). BT has proved to be of
APPROACHES use in reduction in airway smooth muscle (ASM).
New pharmacological targets are being identified BT uses radiofrequency energy to reduce the mass
for asthma and COPD (Braun and Tschernig, 2006). of ASM. The BT modality relies on reducing the
Several animal models have been proposed for the mass of ASM by thermal injury and helps in reduc-
study of asthma (Mullane and Williams, 2014). ing the severity of asthma. The AlairTM Bronchial
Steroids have been found to cause complete rever- Thermoplasty System approved by the FDA in 2010
sal of albuterol-induced beta(2)-adrenergic recep- for conditional use in asthma has been used for BT
tor tolerance in human subjects (Cooper and (BlueCross BlueShield Association, 2014).
Panettieri, 2008).
Novel pharmacotherapeutic approaches include: Genetics of asthma
a. Surfactant proteins—Surfactant protein Asthma is linked to markers on chromosome 12
D (SP-D) has been found to inhibit early (Wilkinson et  al., 1998). Fetal growth and the
airway response in Aspergillus fumigatus- development of childhood asthma have a clear-cut
treated mice. SP-D also reduces bronchial relationship (Leadbitter et al., 1999). Several genes
 Precision medicine in asthma  283

and loci have been identified for asthma suscepti- 3. Ability to identify nonresponders asthma
bility (Guilleminault et al., 2017). phenotypes
Genes for asthma susceptibility are as follows: 4. Ability to characterize clinical asthma
phenotypes
1. ORMDL3
2 . GSDMB Multiple markers (Table 16.5) for type 2 inflam-
mation in asthma have been identified (Fricker
Identified susceptibility loci are as follows: et al., 2017).
Blood eosinophils as biomarkers have accept-
1. HLA-DRA on 6p21
able accuracy (Bayes and Cowan, 2016). Eosinophil
2 . HLA-DQB1 on 6p21
counts and serum IgE are considered most reliable
3. IL13 on 5q31
biomarkers. Lung function and nasal cytology can
4. IL1RL1/IL18RL1 on 2q12
be used as ancillary biomarkers (Ciprandi et  al.,
5. IL33 on chromosome 9p24
2017). Blood eosinophils are good biomarkers
6. WDR36/TSLP on 5q22
for T-helper type 2 cell-high asthma therapies. It
Efforts have been made to utilize genomics in can also be used to identify greater resistance to
research on health disparities (Fullerton et  al., steroids (Stokes and Casale, 2016). Blood eosino-
2012). Asthma endotyping is now in discussion phils and periostin are important biomarkers in
(Agache, 2013). Ancestry may also have been a clinical trials of type-2 cytokine inhibitors (Peters
factor in the development of asthma (Ortega and et  al., 2016). Periostin has potential to be used as
Meyers, 2014). Epigenetics is now considered to a biomarker for T-helper-2 (Th2)-driven asthma
explore a link between prenatal/early-life exposure (Jeanblanc et  al., 2017). Nineteen endogenous
and susceptibility for asthma (de Planell-Saguer urinary biomarkers have been identified for pos-
et  al., 2014). Six new loci have been found to be sible use in asthma/COPD (Khamis et  al., 2017).
associated with forced vital capacity. These loci are Genotyping has led to establishing rs572527200
listed in Table 16.4 (Loth et al., 2014). as a SNP marker for asthma (Ponomarenko et al.,
2015). Allergic sensitization and blood eosinophil
counts may find use in managing early asthma in
Why personalized and precision young children (Anderson et  al., 2017). It is very
medicine are needed and have unfortunate that biomarkers are not utilized in
potential in improving health clinical practice. They are also not included in cur-
rent guidelines on asthma (Agache and Rogozea,
BIOMARKERS 2017).
Zedan et  al. (2016) have established criteria for Asthma pharmacogenetics and individual
asthma biomarkers: genetic profiling have been suggested for personal-
ized medicine. It can help in creating tailor-made
1. Ability to identify poor symptoms perceivers treatment options for individualized patients. This
2 . Ability to predict disease activity and outcomes approach will also ensure that better therapeutic
outcomes are achieved and the heath burden is
Table 16.4  Six new loci associated relieved or minimized (Ortega et al., 2015).
with forced vital capacity

Position of locus (in or near) Table 16.5  Markers for type 2

inflammation in asthma
EFEMP1
BMP6 Markers
MIR129-2-HSD17B12
Eosinophils in sputum
PRDM11
Eosinophils in blood
WWOX
Exhaled NO
KCNJ2
Serum IgE
Source: Loth DW et  al., Nat Genet
Periostin
46(7);2014:669–77.
284  Precision medicine in asthma and chronic obstructive pulmonary disease

There is an important and immediate need for endotype based and mechanism-specific therapy is
identifying medicinally important pharmacoge- needed for severe persistent asthma (Lang, 2015).
nomics loci for asthma treatment. We are far away
from making this reality directly turn relevant and PRECISION MEDICINE FOR
work for a majority of patients. An effective combi- CHRONIC OBSTRUCTIVE
nation of network modeling, functional validation, PULMONARY DISEASE (COPD)
and integrative omics can make asthma pharma-
cogenomics affordable and clinically relevant (Park Introduction
et  al., 2015). CCR7 helps in CTLA4 Ig-modified
dendritic cells homing. CTLA4 Ig and CCR7 com- Chronic obstructive pulmonary disease (COPD)
bined together have potential for precision-guided is a progressive respiratory disease. COPD results
immunotherapy (Chen et al., 2014). in worsening obstruction to airflow in the lungs
over a considerable period of time. COPD also
Challenges and research includes chronic bronchitis and emphysema. In
case of chronic bronchitis, airway linings face con-
opportunities
stant irritation and inflammation. As a result, the
As per Gauthier et al. (2015) findings, areas of fur- lining starts getting thickened. This thickening
ther research in respiratory precision medicine is followed by development of thick mucus in the
include: airways. Both of these pathophysiological devel-
opments lead to difficulty in breathing. In cases
1. Identification of additional clinical and of emphysema, the walls between the air sacs of
molecular phenotypes the lungs are damaged, and, consequently, surface
2 . Identification of additional clinical and area available for the gaseous exchange is reduced.
molecular phenotypes Clinically, it has been observed that most COPD
3. Validation of predictive biomarkers patients suffer from both chronic bronchitis as well
4. Identification of new areas for possible thera- as emphysema (NHLBI, 2015a, 2015c; Sunovion
peutic interventions Pharmaceuticals Inc., 2015). COPD has been iden-
tified as the disease requiring high precision in
It is hoped that epigenetics will help in regulat- family practice for successful management (van
ing genes responsible for inflammation in aspi- Doorn-Klomberg et al., 2013).
rin-exacerbated asthma. DNA methylation and
histone protein modification methods can be used OCCURRENCE
for this purpose. Developing noninvasive clinical According to the estimates of World Health
biomarkers for this type of asthma is a challenge Organization (WHO), 64 million people across the
(Dahlin and Weiss, 2016). world suffer from COPD (Woolcock Institute of
Several markers (such as eosinophils, IgE, Medical Research, 2015). Fore example, around 1
FENO, and periostin) have potential to predict in every 13 Australians over the age of 40 years has
anti-IL-13 and anti-IL-4 therapeutic strategies COPD (Woolcock Institute of Medical Research,
(Bagnasco et  al., 2016). Molecular endotyping 2015) and India has been warned of COPD emer-
using airway gene expression profiling can help gence (Jindal, 2006; Ladhani, 2016). COPD kills
in understanding disease mechanisms. It can also more than 3 million people annually (Salvi and
aid in determination of potential targets for novel Agrawal, 2012; Suthar et al., 2015). COPD by 2020
treatments (Wesolowska-Andersen and Seibold, is expected to be the third leading cause of death
2015). (Jindal et al., 2006).
Transcriptomics have a potential for endotyp- COPD has affected the American population to
ing. A combination of FENO/urinary bromoty- a significant extent, where 15.7 million adults have
rosine can be used in predicting response to the been diagnosed with COPD (Wheaton et al., 2015),
steroids (Bayes and Cowan, 2016). Another chal- and an estimated 12 million adults remain undiag-
lenging task is personalized medicine and a bio- nosed (NHLBI, 2015c). In the United States, COPD
marker-based approach children suffering from causes over 120,000 deaths per year and is the third
severe asthma (Bozzetto et  al., 2015). Phenotype/ leading cause of death (Table 16.6; NHLBI, 2015c).
 Precision medicine for chronic obstructive pulmonary disease (COPD)  285

Table 16.6  COPD in United States of history, physical examination, and confirmed
airflow obstruction using postbronchodilator spi-
Parameter Result
rometry. Patients over the age of 35 presenting
Affected world population 67 million with complaints of exertion-related breathless-
Affected U.S. population 15.7 million ness, chronic cough, sputum production, frequent
Undiagnosed U.S. population 12 million winter “bronchitis,” or wheezing need to be prop-
Mortality in United States 0.12 million erly diagnosed for COPD (Derbyshire Joint Area
Economic burden in United States $29.5 billion Prescribing Committee, 2015). Diagnostic testing
may require a chest x-ray or CT (CAT scan), blood
tests, or sputum sample testing. In some cases, car-
The U.S. Centers for Disease Control and Preven­ diac ultrasound and overnight sleep studies may
tion (CDC) has issued Public Health Strategies for also be needed for accurate diagnosis (Woolcock
preventing COPD (CDC, 2011). Institute of Medical Research, 2015).
The diagnosis of COPD is primarily based on
Approaches currently used for the a physical examination by a trained general prac-
titioner or a specialist with appropriate training.
disease diagnosis and treatment
Many times, the patient’s individual medical and
PATHOGENESIS OF COPD family history is also helpful. Lung function test-
ing constitutes the main diagnostic criteria for
Pathogenesis of COPD includes the following the-
COPD. Spirometer conducted by a physician is
ories (Vijayan, 2013):
often the primary diagnostic criteria. Spirometry
1. Proteinase–antiproteinase hypothesis reliably measures the total amount of exhaled air.
2 . Immunological mechanisms Spirometry can also be used in detecting COPD in
3. Oxidant–antioxidant balance early stages much before the onset of characteristic
4. Systemic inflammation symptoms (NHLBI, 2015b).
5. Apoptosis and ineffective repair
CLASSIFICATION OF DIAGNOSTIC
Natural innate cytokine responds to TECHNIQUES
immune-modulators (Switalla et  al., 2010). Diagnostic techniques used for COPD are summa-
Lipopolysaccharides (LPS) elicit a range of immu- rized in Table 16.7.
nological effects and are involved in COPD. Single
nucleotide polymorphisms have been linked with COPD PHENOTYPES
susceptibility to COPD. These polymorphisms take Phenotyping is also relevant in COPD research.
place in the TSPYL-4 and NT5DC1 genes (Guo COPD phenotypes have been identified (Marsh
et  al., 2012). Desmosine (DES) and isodesmosine et al., 2008). The need for treating COPD by clini-
have the potential to be used as putative biomark- cal phenotypes has been highlighted (Miravitlles
ers for COPD (Viglio et al., 2014). Afamin, a mem- et al., 2013). Asthma has been identified as a risk
ber of the albumin gene family, cannot be used factor for COPD (Silva et al., 2004). Clinical phe-
for determining COPD progression (Dieplinger notypes of COPD and asthma have been discussed
et al., 2013). Six new loci are associated with forced (Carolan and Sutherland, 2013). Table 16.8 shows
vital capacity. These loci have been identified as the COPD phenotypes (Turner et al., 2015).
EFEMP1, BMP6, MIR129-2-HSD17B12, PRDM11,
WWOX, and KCNJ2. These loci were found in Therapeutic options for the
African-American, Korean, Chinese, and Hispanic
treatment of COPD
subjects (Loth et al., 2014).
Therapeutic options are classified into the follow-
DIAGNOSIS
ing broad categories (Delmer, 2015):
There is no single specific or confirmatory diag-
nostic test for COPD. To be honest, the accurate 1. Pulmonary rehabilitation
diagnosis depends on the clinical judgment of the 2 . Vaccinations
physicians. This judgment relies on a combination 3. Oxygen therapy
286  Precision medicine in asthma and chronic obstructive pulmonary disease

Table 16.7  Diagnostic techniques for COPD varenicline, bupropion, and nortriptyline (Delmer,
2015). Pulmonary rehabilitation for COPD is an
Example of diagnostic
accepted practice (Stein et al., 2009).
Type technique
Radiological Chest x-ray, nuclear medicine, VACCINATIONS
techniques real-time sonography, color Vaccinations such as influenza and pneumococ-
Doppler ultrasonography (USG), cal polysaccharide are also useful in treatment of
CT and MRI COPD. The influenza vaccine contains killed or
Traditional Cardiogenic oscillation, attenuated viruses. It is administered once a year.
pulmonary negative expiratory pressure, Pneumococcal polysaccharide vaccines are suit-
function exhaled temperature and able for adults aged 65 years or more. It can also
tests exhaled breath condensate, be given to younger adults having comorbid condi-
fractional exhaled nitric oxide tions (Delmer, 2015).
Newer Pulmonary arterial pressure,
Pulmonary inductive plethysmography, OXYGEN THERAPY
function oximetry-cutaneous Oxygen therapy involves long-term administra-
tests capnography, pulse oximetry, tion of oxygen. If administered for usually more
computed tomography than 15 hours a day, it can increase the survival
Sophisticated B-type natriuretic peptide rate in patients who have faced or are expected to
assays assay, presage st2 assay have chronic respiratory failure or severe hypox-
Invasive RBM biopsy, inflammatory cell emia associated with resting stage. This therapy
techniques counts also keeps the progressive hypoxemia under check
(Delmer, 2015).
4. Pharmacotherapy PHARMACOTHERAPY
5. Exacerbation management
The pharmacotherapy is aimed at achieving one or
6. Exacerbation pharmacotherapy
more of the following targets (Delmer, 2015):
7. Monitoring and follow-up
8. Overcoming the barriers associated with
1. Reduction of symptoms
COPD treatment
2 . Frequency of exacerbations
9. Educational strategies
3. Intensity of exacerbations
The details of these modalities or approaches 4. Prevent disease progression
are discussed next. 5. Increase exercise tolerance
6. Improvement of general health
PULMONARY REHABILITATION 7. Reduction in mortality rates
Studies have indicated the fruitfulness of nico- Common drugs used in the treatment of COPD
tine replacement therapy and use of drugs such as are summarized in Table 16.9 (Adil et al., 2015).

Table 16.8  Phenotypes of COPD EXACERBATION MANAGEMENT

COPD phenotype Exacerbation management is aimed at minimiz-
ing the impact as well occurrence of the exacerba-
High IgE phenotype tions. Bronchodilators, systemic corticosteroids,
Reversibility to β2-agonists phenotype and antibiotics may also be of great help. Adjunct
Type 1 respiratory failure phenotype therapy and adequate respiratory support may also
Type 2 respiratory failure phenotype be desired in certain cases (Delmer, 2015).
Upper zone emphysema phenotype
α1-antitrypsin deficiency phenotype EXACERBATION PHARMACOTHERAPY
Source: Turner AM et  al., Eur Respir Rev Short-acting β-agonist alone or in combination
24;2015:283–98. with a short-acting anticholinergic agent can
 Precision medicine for chronic obstructive pulmonary disease (COPD)  287

Table 16.9  Common drugs used in the Genetic linkage

treatment of COPD
Several genes (Table 16.10) have been identified
Category Drug in smokers that make them susceptible to COPD
β2-agonists Salbutamol (Agusti et al., 2016a, 2016b).
Salmetrol
Antibiotics Penicillin Novel approaches in the
Cephalosporin management of COPD
Fluoroquinolone
Nitro-imidazole 1. Dietary changes—Inflammatory processes in
Macrolide patients with COPD is involved in proteolysis
Aminoglycoside and subsequent destruction of elastic fibers. A
Inhaled Budesonide study has confirmed that foods with high levels
corticosteroids of antioxidants may have a protective effect in
Anticholinergics Ipratropium bromide the lung (Varraso et al., 2015).
Scopolamine 2 . Exercise—It has been experimentally shown
Xanthine derivatives Theophylline/etophylline in the clinical studies that a treatment method
Mucolytics Ambroxol involving deep-breathing exercise with oxygen
Acetyl cysteine inhalation led to improvement of lung func-
Bromhexine/terbutaline/ tion in COPD patients. The superiority of
guaifenesin the results was confirmed when the results of
Oral corticosteroids Hydrocortisone this particular group were compared with the
Methyl prednisolone results of the patients on deep-breathing exer-
Antihistamines Levocetirizine cise or oxygen inhalation alone. This approach
Chlorpheniramine has ample scope for use in clinical situations
Loratidine (Liu et al., 2015).
Leukotriene Montelukast
3. Prednisone treatment—Prednisone 30–50 mg
antagonist
daily for 5 days is prescribed for acute moder-
ate or severe exacerbations in COPD patients.
Source: Adil MS et  al., J Clin Diagn Res 9(11);2015:​ Previously, 10 to 14 days of oral prednisone treat-
FC05–8.
ment was recommended. It has been clinically
observed that 5 days of prednisone treatment has
be used for exacerbation management in COPD a dual advantage: It effectively treats the current
patients. For example, 40 mg of prednisone per exacerbations and prevents subsequent exacer-
day for 5 days can be given by oral route. In case bations (Leuppi et al., 2013; Walters et al., 2014;
of bacterial infections in COPD patients con- Abramson et al., 2014).
firmed by sputum cultures, aminopenicillin with 4. Antibiotic therapy—Long-term antibiotics are
or without clavulanic acid, macrolide, or tetra- effective for treating frequent exacerbations
cycline treatment for 5 to 10 days can be used
(Delmer, 2015).
Table 16.10  COPD susceptibility
MONITORING AND FOLLOW-UP genes in smokers
Once improvement results from pharmacologic COPD susceptibility genes
or nonpharmacological interventions, spirometry
FAM13A
can be done on an annual basis. Smoking habits
HHIP
and cessation needs constant monitoring. Indirect
CHRNA3
exposure to smoking also needs to be eliminated or
CHRNA5
minimized to the lowest level possible. Therapeutic
IREB2
regimen, patient compliance, and exacerbation
Region on chromosome 19
need close monitoring (Delmer, 2015).
288  Precision medicine in asthma and chronic obstructive pulmonary disease

of COPD. However, they should be prescribed miRNAs (Agusti et al., 2016a, 2016b). Eosinophils
only after evaluating the risk of antibiotic and type 2 inflammatory markers can be used to
resistance. Long-term macrolides prescribing identify asthma-COPD overlap syndrome (ACOS)
should not be used as a mere rule of thumb. molecular phenotype (Christenson, 2016).
Rather, it should be decided on a case-to-case
basis by a respiratory physician (Bryant, 2015). Challenges and research
5. Plant-based therapy—Peucedani Radix
(Qian-hu) used in COPD in Chinese medicine
opportunities
contains angular-type pyranocoumarins (APs), GENE EXPRESSION ANALYSIS
which can be measured by solid phase extrac-
Currently none of the available therapies can help in
tion-chiral LC-MS/MS (Song et al., 2014).
restricting the progression of COPD. Therefore, an
6. Rehabilitation with music therapy—Music
urgent need exists for the development of new and
therapy has been tried in patients with COPD
effective treatments for COPD. There is an urgency
at the Louis Armstrong Center of Music
to identify molecular targets for COPD treatment.
and Medicine—Mount Sinai Beth Israel
Here, techniques such as gene expression profil-
(MSBI). Patients receiving music therapy
ing, serial analysis of gene expression, or microar-
along with standard rehabilitation exhibited
rays may be utilized. Several research teams have
better improvement in symptoms as com-
mapped comparative gene expression in lung tissues
pared to the patients on standard rehabilita-
of COPD patients, and it has been found that differ-
tion alone (Canga et al., 2015; Mount Sinai
entially regulated genes are associated with disease
Hospital, 2015).
progression. Such studies have identified molecular
7. Cognitive-behavioral therapy—It has been
mechanisms of COPD and also suggested new tar-
clinically proven that cognitive-behavioral
gets for COPD treatments (Chen et al., 2008).
therapy has no significant effect on symptoms
of anxiety or depression in COPD patients
(Health Quality Ontario, 2015). CONCLUSIONS
BIOMARKERS Asthma and chronic obstructive pulmonary
Several physiological, cellular, proteomic, and disease (COPD) have a link with the so-called
genetic markers are available for COPD (Rooney Westernization and change in lifestyle. Both
and Sethi, 2015). Agusti et al. (2016a, 2016b) have asthma and COPD involve changes in the airways
classified a number of biomarkers (Table 16.11) for and lungs leading to loss of quality of life, thus
COPD. necessitating the standard of medical care. There
is an urgent need to include the omics technologies
EPIGENETIC BIOMARKERS in the diagnosis, molecular pathogenesis, pheno-
Epigenetic biomarkers can also be used in COPD. typing, management, and evolving guided-preci-
Such markers include histone methylation and sion therapy for asthma and COPD.

Table 16.11  Classification of biomarkers for COPD

Cellular Blood protein Sputum Serum

biomarkers biomarkers Gene studies transcriptomics metabolomics
Sputum neutrophils Fibrinogen Smoking history Airflow limitation Serum profile
Circulating WBC CC16 COPD susceptibility Blood biomarkers Exhaled breath
SP-D HHIP Emphysema condensate
CCL18 (PARC) COPD subtypes COPD susceptibility pH
sRAGE Emphysema Adenosine/
Inflammome Cachexia purines
Adipokines Blood biomarkers
Vitamin D
 References 289

ACKNOWLEDGMENTS Bagnasco D, Ferrando M, Varricchi G et al. A

critical evaluation of anti-IL-13 and anti-IL-4
The author wishes to thank all his teachers at Jamia strategies in severe asthma. Int Arch Allergy
Hamdard, New Delhi and also his family mem- Immunol 170(2);2016:122–31.
bers. Special thanks are also due to Dr. D. Barh for Banerjee A, Trivedi CM, Damera G, Jiang M,
suggestions and constructive criticism. The author Jester W, Hoshi T, Epstein JA, and Panettieri
also acknowledges the support from his wife (Mrs. Jr RA. Trichostatin A abrogates airway con-
Zeba) and daughters (Shifa and Azka) who sacri- striction, but not inflammation, in murine and
ficed their quality time. human asthma models. Am J Respir Cell Mol
Biol 46(2);2012:132–8.
REFERENCES Baser, O. Too much ado about instrumental
variable approach: Is the cure worse than the
Abramson M, Frith P, Yang I et al. COPD-X con- disease? Value Health 12(8);2009:1201–9.
cise guide for primary care. 2014. Available at Bateman ED, Hurd SS, Barnes PJ et al. Global
www.copdx.org.au. Accessed January 2015. strategy for asthma management and preven-
Adil MS, Khan MA, Khan MN, Sultan I, Khan tion: GINA executive summary. Eur Respir J
MA, Ali SA, and Farooqui A. EMPADE study: 31;2008:143–78.
Evaluation of medical prescriptions and Bayes HK, and Cowan DC. Biomarkers and
adverse drug events in COPD patients admit- asthma management: An update. Curr Opin
ted to intensive care unit. J Clin Diagn Res Allergy Clin Immunol 16(3);2016:210–7.
9(11);2015:FC05–8. Bel EH. Clinical phenotypes of asthma. Curr Opin
Agache, IO. From phenotypes to endotypes to Pulm Med 10;2004:44–50.
asthma treatment. Curr Opin Allergy Clin Blackwell DL, Lucas JW, and Clarke TC. Summary
Immunol 13(3);2013:249–56. health statistics for U.S. adults: National
Agache I, and Rogozea L. Asthma biomarkers: Health Interview Survey, 2012. Vital Health
Do they bring precision medicine closer Stat 10(260);Feb 2014:1–61.
to the clinic? Allergy Asthma Immunol Res BlueCross BlueShield Association. Bronchial
9(6);2017:466–76. thermoplasty for treatment of inadequately
Agusti A, Bel E, Thomas M et al. Treatable controlled severe asthma. TEC Assessments
traits: Toward precision medicine of 29(12);2014:1–32.
chronic airway diseases. Eur Respir J Bozzetto S, Carraro S, Zanconato S et al. Severe
47(2);2016a:410–9. asthma in childhood: Diagnostic and man-
Agusti A, Gea J, and Faner R. Biomarkers, the agement challenges. Curr Opin Pulm Med
control panel and personalized COPD medi- 21(1);2015:16–21.
cine. Respirology 21;2016b:24–33. Braun A, and Tschernig T. Animal models of
Amrani Y, Syed F, Huang C, Li K, Liu V, Jain D, asthma: Innovative methods of lung research
Keslacy S. Expression and activation of the and new pharmacological targets. Exp Toxicol
oxytocin receptor in airway smooth muscle Pathol 57(Suppl 2);2006:3–4.
cells: Regulation by TNFalpha and IL-13. Brozek JL, Akl EA, Alonso-Coello P et al. Grading
Respir Res 11;2010:104. quality of evidence and strength of recommen-
Anderson WC 3rd, Apter AJ, Dutmer CM et al. dations in clinical practice guidelines. Part 1 of
Advances in asthma in 2016: Designing 3. An overview of the GRADE approach and
individualized approaches to management. grading quality of evidence about interven-
J Allergy Clin Immunol 140(3);2017:671–80. tions. Allergy 64(5);2009:669–77.
American Thoracic Society (ATS). Patient Brueggemann LI, Haick JM, Neuburg S, Tate
resources for asthma. thoracic.org/patients/ S, Randhawa D, Cribbs LL, and Byron KL.
patient-resources/topic-specific/asthma.php, KCNQ (Kv7) potassium channel activators as
2015. bronchodilators: Combination with a beta2-
Aubier M, Herth FJ, and Niven R. The Uncovering adrenergic agonist enhances relaxation of rat
Asthma Report. Marlbourough, MA: Boston airways. Am J Physiol Lung Cell Mol Physiol
Scientific; 2015. 306(6);2014:L476–86.
290  Precision medicine in asthma and chronic obstructive pulmonary disease

Bryant L. Long-term antibiotics to reduce COPD bronchoconstriction and acts as a histamine

exacerbations: Pros and cons. Prim Health and serotonin antagonist in human airways.
Care 7(1);2015:78–80. Allergy Asthma Proc 32(5);2011:359–65.
Canga B, Azoulay R, Raskin J et al. AIR: Advances Dahlin A, and Weiss ST. Genetic and epigenetic
in Respiration—Music therapy in the treat- components of aspirin-exacerbated respira-
ment of chronic pulmonary disease. Respir tory disease. Immunol Allergy Clin North Am
Med 109(12);2015:1532–9. 36(4);2016:765–89.
Carolan BJ, and Sutherland ER. Clinical pheno- de Planell-Saguer M, Lovinsky-Desir S, and Miller
types of chronic obstructive pulmonary dis- RL. Epigenetic regulation: The interface
ease and asthma: Recent advances. J Allergy between prenatal and early-life exposure and
Clin Immunol 131;2013:627–34. asthma susceptibility. Environ Mol Mutagen
Centers for Disease Control and Prevention 55(3);2014:231–43.
(CDC). Public Health Strategic Framework for Derbyshire Joint Area Prescribing Committee.
COPD Prevention. Atlanta, GA: Centers for Chronic obstructive pulmonary disease
Disease Control and Prevention; 2011. (COPD) management, June 2015.
Chen ZH, Kim HP, Ryter SW, and Choi AMK. Delmer TL. The gold standard—Understanding
Identifying targets for COPD treatment and treating COPD, Universal Activity No.:
through gene expression analyses. Int J Chron 0798-0000-15-017-L01-PandT, released May
Obstruct Pulmon Dis 3(3);2008:359–70. 14, 2015.
Chen Y, Wang Y, and Fu Z. T lymphocyte Demoly P, Paggiaro P, Plaza V et al. Prevalence
antigen 4-modified dendritic cell ther- of asthma control among adults in France,
apy for ­asthmatic mice guided by the Germany, Italy, Spain and the UK. Eur Respir
CCR7 chemokine receptor. Int J Mol Sci Rev 18;2009:112, 105–12.
15(9);2014:15304–19. Dieplinger B, Egger M, Gabriel C, Poelz W,
Ciprandi G, Tosca MA, Silvestri M et al. Morandell E, Seeber B, Kronenberg F,
Inflammatory biomarkers for asthma Haltmayer M, Mueller T, and Dieplinger
endotyping and consequent personal- H. Analytical characterization and clini-
ized therapy. Expert Rev Clin Immunol cal ­evaluation of an enzyme-linked immu-
13(7);2017:715–21. nosorbent assay for measurement of
Christenson SA. The reemergence of the asthma- afamin in human plasma. Clin Chim Acta
COPD overlap syndrome: Characterizing a 425;2013:236–41.
syndrome in the precision medicine era. Curr Erpenbeck VJ, Ziegert M, Cavalet-Blanco D
Allergy Asthma Rep 16(11);2016:81. et al. Surfactant protein D inhibits early
Chung KF, Wensel SE, Brozek JL et al. airway response in Aspergillus fumigatus-­
International ERS/ATS guidelines on defini- sensitized mice. Clin Exp Allergy
tion, evaluation and treatment of severe 36(7);2006:930–40.
asthma. Eur Respir J 43;2014:343–73. Fricker M, Heaney LG, and Upham JW. Can bio-
Cooper PR, Mesaros AC, Zhang J, Christmas markers help us hit targets in difficult-to-treat
P, Stark CM, Douaidy K, Mittelman MA, asthma? Respirology 22(3);2017:430–42.
Soberman RJ, Blair IA, and Panettieri RA. 20- Fullerton SM, Knerr S, and Burke W. Finding
HETE mediates ozone-induced, neutrophil- a place for genomics in health dispari-
independent airway hyper-responsiveness in ties research. Public Health Genomics
mice. PLOS ONE 5(4);2010:e10235. 15(3–4);2012:156–63.
Cooper PR, and Panettieri Jr RA. Steroids Gauthier M, Ray A, and Wenzel SE. Evolving
completely reverse albuterol-induced ­concepts of asthma. Am J Respir Crit Care
beta(2)-adrenergic receptor tolerance in Med 192(6);2015:660–8.
human small airways. J Allergy Clin Immunol Gillman A, and Douglass JA. Asthma in the
122(4);2008:734–40. elderly. Asia Pac Allergy 2(2);2012:101–8.
Cooper PR, Zhang J, Damera G, Hoshi T, Zopf Global Initiative for Asthma (GINA). Pocket guide
DA, and Panettieri Jr RA. C-027 inhib- for asthma management and prevention.
its IgE-mediated passive sensitization 2015. Available at www.ginasthma.org.
 References 291

Goodman MS. Comparison of small-area analysis muscle contraction by modulating cal-

techniques for estimating prevalence by race. cium oscillations. J Pharmacol Exp Ther
Prev Chronic Dis 7(2);2010:A33. 334(3);2010:703–9.
Guilleminault L, Ouksel H, Belleguic C et al. Jiang H, Xie Y, Abel PW, Toews ML, Townley RG,
Personalised medicine in asthma: From cura- Casale TB, and Tu Y. Targeting phosphoinosit-
tive to preventive medicine. Eur Respir Rev ide 3-kinase gamma in airway smooth
26;2017:160010. muscle cells to suppress interleukin-13-­
Guo Y, Gong Y, Shi G, Yang K, Pan C, Li M, induced mouse airway hyperresponsiveness.
Li Q, Cheng Q, Dai R, Fan L, and Wan H. J Pharmacol Exp Ther 342(2);2012:305–11.
Single-nucleotide polymorphisms in the Jindal SK. Emergence of chronic obstructive
TSPYL-4 and NT5DC1 genes are associ- pulmonary disease as an epidemic in India.
ated with s­ usceptibility to chronic obstruc- Indian J Med Res 124;2006a:619–30.
tive ­pulmonary disease. Mol Med Rep Jindal SK, Aggarwal AN, Chaudhry K et al.
6(3);2012:631–8. A multicentric study on epidemiology of
Hahn DL. Importance of evidence grading for chronic obstructive pulmonary disease and its
guideline implementation: The example of relationship with tobacco smoking and envi-
asthma. Ann Fam Med 7(4);2009:364–9. ronmental tobacco smoke exposure. Indian
Haldar P, Pavord ID, Shaw DE et al. Cluster J Chest Dis Allied Sci 48;2006b:23–9.
analysis and clinical asthma phenotypes. Am John G, Darbellay P, Drepper M, Spechbach H,
J Respir Crit Care Med 178;2008:218–24. Fosenbauer MB, Perrier A, and Carballo S.
Health Quality Ontario. Cognitive-Behavioural [Hospital based internal medicine in 2012].
Therapy for Anxiety and Depression in Rev Med Suisse 9(370);2013:193–6, 198.
Patients with COPD: A Rapid Review. Just J, Saint-Pierre P, Gouvis-Echraghi R et al.
Toronto: Health Quality Ontario; 2015 Wheeze phenotypes in young children have
February. 21 p. Available at: http://www. different courses during the preschool period.
hqontario.ca/evidence/evidence-process/ Ann Allergy Asthma Immunol 111;2013:256–
episodes-of-care#community-copd. 61.e1.
Heaney LG, and McGarvey LP. Personalised Ker J, and Hartert TV. The atopic march: What’s
medicine for asthma and chronic obstruc- the evidence? Ann Allergy Asthma Immunol
tive pulmonary disease. Respiration 103(4);2009:282–9.
93(3);2017:153–61. Khamis MM, Adamko DJ, and El-Aneed A.
Henderson J, Granell R, Heron J et al. Associations Development of a validated LC- MS/MS
of wheezing phenotypes in the first 6 years method for the quantification of 19 endoge-
of life with atopy, lung function and airway nous asthma/copd potential urinary biomark-
responsiveness in mid-childhood. Thorax ers. Anal Chim Acta 989;2017:45–58.
63;2008:974–80. Ladhani S. COPD—An Indian perspective, J Lung
Hnizdo E, Sircar K, Yan T, Harber P, Fleming J, Pulm Respir Res 3(1);2016:00070.
and Glindmeyer HW. Limits of longitudi- Lang DM. Severe asthma: Epidemiology, burden
nal decline for the interpretation of annual of illness, and heterogeneity. Allergy Asthma
changes in FEV1 in individuals. Occup Environ Proc 36(6);2015:418–24.
Med 64(10);2007:701–7. Larsson AK, Hagfjard A, Dahlen SE, and Adner
Ilango K, Rajanandh MG, and Nageswari AD. M. Prostaglandin D(2) induces contractions
Roflumilast: An upcoming drug for curing through activation of TP receptors in periph-
asthma and COPD. IJPRT 5(2);2013:130–5. eral lung tissue from the guinea pig. Eur J
Jeanblanc NM, Hemken PM, Datwyler MJ et al. Pharmacol 669(1–3);2011:136–42.
Development of a new ARCHITECT auto- Leadbitter P, Pearce N, Cheng S, Sears MR,
mated periostin immunoassay. Clin Chim Acta Holdaway MD, Flannery EM, Herbison
464;2017:228–35. GP, and Beasley R. Relationship between
Jiang H, Abel PW, Toews ML, Deng C, Casale fetal growth and the development of
TB, Xie Y, and Tu Y. Phosphoinositide asthma and atopy in childhood. Thorax
3-kinase gamma regulates airway smooth 54(10);1999:905–10.
292  Precision medicine in asthma and chronic obstructive pulmonary disease

Lefebvre Q, Vandergoten T, Derom E, messages for managing patients with asthma.

Marchandise E, and Liistro G. Testing spirom- 2015a. Available at http://www.nhlbi.nih.gov/
eters: Are the standard curves of the ameri- files/docs/guidelines/asthma_qrg.pdf.
can thoracic society sufficient? Respir Care National Heart, Lung, and Blood Institute. How is
59(12);2014:1895–904. COPD diagnosed? 2015b. Available at http://
Leuppi JD, Schuetz P, Bingisser R et al. Short- www.nhlbi.nih.gov/health/health-topics/­
term vs conventional glucocorticoid therapy topics/copd/diagnosis.html. Accessed April
in acute exacerbations of chronic obstructive 20, 2015.
pulmonary disease: The REDUCE randomized National Heart, Lung and Blood Institute (NHLBI).
clinical trial. JAMA 309;2013:2223–31. What is COPD? 2015c. Available at http://
Liu YQ, Yan LX, Zhang LY, Song QH, and Xu RM. www.nhlbi.nih.gov/health/health-topics/­
Conspicuous effect on treatment of mild-to- topics/copd/. Accessed April 20, 2015.
moderate COPD by combining deep-breath- Nicoll J, Cruikshank WW, Brazer W, Liu Y,
ing exercise with oxygen inhalation. Int J Clin Center DM, and Kornfeld H. Identification of
Exp Med 8(6);2015:9918–24. domains in IL-16 critical for biological activity.
Loth DW, Artigas MS, Gharib SA et al. Genome- J Immunol 163(4);1999:1827–32.
wide association analysis identifies six new Ortega VE, and Meyers DA. Implications of
loci associated with forced vital capacity. population structure and ancestry on asthma
Nat Genet 46(7);2014:669–77. genetic studies. Curr Opin Allergy Clin
Marsh SE, Travers J, Weatherall M et al. Immunol 14(5);2014:381–9.
Proportional classifications of COPD pheno- Ortega VE, Meyers DA, and Bleecker ER.
types. Thorax 63;2008:761–7. Asthma pharmacogenetics and the devel-
Miravitlles M, Soler-Cataluna JJ, Calle M, and opment of genetic profiles for personal-
Soriano JB. Treatment of COPD by clinical ized ­medicine. Pharmgenomics Pers Med
phenotypes: Putting old evidence into clinical 8;2015:9–22.
practice. Eur Respir J 41;2013:1252–6. Park HW, Tantisira KG, and Weiss ST.
Moerman CJ, Deurenberg R, and Haafkens Pharmacogenomics in asthma therapy: Where
JA. Locating sex-specific evidence on clini- are we and where do we go? Annu Rev
cal questions in MEDLINE: A search filter Pharmacol Toxicol 55;2015:129–47.
for use on OvidSP. BMC Med Res Methodol Peters MC, Nguyen ML, and Dunican EM.
9;2009:25. Biomarkers of airway type-2 inflammation
Moore WC, Meyers DA, Wenzel SE et al. and integrating complex phenotypes to
Identification of asthma phenotypes using endotypes in asthma. Curr Allergy Asthma
cluster analysis in the Severe Asthma Rep 16(10);2016:71.
Research Program. Am J Respir Crit Care Med Ponomarenko M, Rasskazov D, Arkova O,
181;2010:315–23. Ponomarenko P, Suslov V, Savinkova L, and
Mount Sinai Hospital. Music therapy increases Kolchanov N. How to use SNP_TATA_com-
effectiveness of pulmonary rehabilitation parator to find a significant change in gene
for COPD patients. 2015, December 23. expression caused by the regulatory SNP of
Available at https://medicalxpress.com/ this gene’s promoter via a change in affin-
news/2015-12-musictherapy-effectiveness-pul- ity of the tata-binding protein for this pro-
monary-copd.html. Accessed February 3, 2016. moter. Biomed Res Int 2015;2015: article ID
Mullane K, and Williams M. Animal models 359835.
of asthma: Reprise or reboot? Biochem Rooney C, and Sethi T. Biomarkers for precision
Pharmacol 87(1);2014:131–9. medicine in airways disease. Ann NY Acad Sci
National Heart, Lung, and Blood Institute 1346(1);2015:18–32.
(NHLBI). 2014 August 4. Available at http:// Salvi S, and Agrawal A. India needs a national
www.nhlbi.nih.gov/health/health-topics/ COPD prevention and control programme.
topics/asthma/. JAPI 60;2012:4–5.
National Heart, Lung, and Blood Institute Seehase S, Schleputz M, Switalla S et al.
(NHLBI). Best practices in care: Six key Bronchoconstriction in nonhuman primates:
 References 293

A species comparison. J Appl Physiol (1985) Tho LL, and Candlish JK. Superoxide dismutase
111(3);2011:791–8. and glutathione peroxidase activities in
Silva GE, Sherrill DL, Guerra S, and Barbee RA. erythrocytes as indices of oxygen loading
Asthma as a risk factor for COPD in a longitu- in disease: A survey of one hundred cases.
dinal study. Chest 126;2004:59–65. Biochem Med Metab Biol 38(1);1987:74–80.
Song Y, Jing W, Yang F, Shi Z, Yao M, Yan R, Toren K. Challenges for the new century in the
and Wang Y. Simultaneously enantiospe- epidemiology of adult asthma. Scand J Work
cific determination of (+)-trans-khellactone, Environ Health 25(6);1999:558–63.
(+/–)-praeruptorin A, (+/–)-praeruptorin B, Turner AM, Tamasi L, Schleich F, Hoxha M,
(+)-praeruptorin E, and their metabolites, Horvath I, Louis R, and Barnes N. Clinically
(+/–)-cis-khellactone, in rat plasma using relevant subgroups in COPD and asthma.
online solid phase extraction-chiral LC-MS/ Eur Respir Rev 24;2015:283–98.
MS. J Pharm Biomed Anal 88;2014:269–77. van Doorn-Klomberg AL, Braspenning JC,
Spratt M, Carpenter J, Sterne JA, Carlin JB, Feskens RC, Bouma M, Campbell SM, and
Heron J, Henderson J, and Tilling K. Strategies Reeves D. Precision of individual and com-
for multiple imputation in longitudinal studies. posite performance scores: The ideal number
Am J Epidemiol 172(4);2010:478–87. of indicators in an indicator set. Med Care
Spycher BD, Silverman M, Brooke AM, Minder 51(1);2013:115–21.
CE, and Kuehni CE. Distinguishing phe- van Schayck CP, Dompeling E, van Weel
notypes of childhood wheeze and cough C, Folgering H, and van den Hoogen
using latent class analysis. Eur Respir J HJ. Accuracy and reproducibility of the
31;2008:974–81. assess peak flow meter. Eur Respir J
St Ledger K, Agee SJ, Kasaian MT, Forlow 3(3);1990:338–41.
SB, Durn BL, Minyard J, Lu QA, Todd J, Varraso R, Chiuve SE, Fung TT et al. Alternate
Vesterqvist O, and Burczynski ME. Analytical Healthy Eating Index 2010 and risk of chronic
validation of a highly sensitive microparticle- obstructive pulmonary disease among US
based immunoassay for the quantitation women and men: Prospective study. BMJ
of IL-13 in human serum using the Erenna 350;2015:h286.
immunoassay system. J Immunol Methods Viglio S, Stolk J, Luisetti M, Ferrari F, Piccinini
350(1–2);2009:161–70. P, and Iadarola P. From micellar electroki-
Stein K, Dyer M, Milne R, Round A, Ratcliffe J, netic chromatography to liquid chroma-
and Brazier J. The precision of health state tography-mass spectrometry: Revisiting
valuation by members of the general public the way of analyzing human fluids for the
using the standard gamble. Qual Life Res search of desmosines, putative biomarkers
18(4);2009:509–18. of chronic obstructive pulmonary disease.
Stokes JR, and Casale TB. Characterization Electrophoresis 35(1);2014:109–18.
of asthma endotypes: Implications for Vijayan VK. Chronic obstructive pulmonary dis-
therapy. Ann Allergy Asthma Immunol ease, Indian J Med Res 137;2013:251–69.
117(2);2016:121–5. Walters JA, Tan DJ, White CJ et al. Different
Sunovion Pharmaceuticals Inc Chronic obstruc- durations of corticosteroid therapy for
tive pulmonary disease (COPD) fact sheet. exacerbations of chronic obstructive pulmo-
2015. (BRO058-15 7/15). nary disease. Cochrane Database Syst Rev
Suthar NN, Patel KL, and Shah J. A study on 12;2014:CD006897.
awareness of chronic obstructive pulmonary Wesolowska-Andersen A, and Seibold MA.
disease (COPD) among Smokers. Ntl J of Airway molecular endotypes of asthma:
Community Med 6(4);2015:547–53. Dissecting the heterogeneity. Curr Opin
Switalla S, Lauenstein L, Prenzler F et al. Natural Allergy Clin Immunol 15(2);2015:163–8.
innate cytokine response to immunomodu- Wheaton, AG, Cunnningham, TJ, Ford, ES, and
lators and adjuvants in human precision- Croft, JB. Employment and activity limitations
cut lung slices. Toxicol Appl Pharmacol among adults with chronic obstructive pulmo-
246(3);2010:107–15. nary disease—United States, 2013. MMWR
294  Precision medicine in asthma and chronic obstructive pulmonary disease

64(11);2015, March 27:289–95. Available at Woolcock Institute of Medical Research. COPD:

http://www.cdc.gov/mmwr/. Chronic Obstructive Pulmonary Disease,
Wilkinson J, Grimley S, Collins A, Thomas NS, Version 1. 2015. Australia.
Holgate ST, and Morton N. Linkage of asthma Zedan MM, Osman AM, Laimon WN et al.
to markers on chromosome 12 in a sample Airway inflammatory biomarker: Could it tai-
of 240 families using quantitative phenotype lor the right medications for the right asthmatic
scores. Genomics 53(3);1998:251–9. patient? Iran J Immunol 13(2);2016:70–88.
 17
Customized DNA–directed precision
nutrition to balance the brain reward
circuitry and reduce addictive
behaviors

KENNETH BLUM, MARCELO FEBO, ERIC R. BRAVERMAN,

MONA LI, LYLE FRIED, ROGER WAITE, ZSOLT DEMOTROVICS,
WILLIAM B. DOWNS, DEBMALYA BARH, BRUCE STEINBERG,
THOMAS McLAUGHLIN, AND RAJENDRA D. BADGAIYAN

Introduction 295 and function, it should make a significant

DNA customization is a reality 298 contribution to overall health 301
Practical applications 298 Criterion 4: The beneficial effects should be
What is the importance and value of obvious, different, transformational, and
DNA-customized nutrition (why use it)? 298 sustainable for most patients 301
What exactly is a DNA-customized Conventional methods do not comply with
nutraceutical? 300 all of the four important criteria 302
What criteria need to be satisfied for DNA- Genetic testing: Fact or fiction 302
customized nutrition to be worthwhile? 300 Pharmacogenetic testing 303
Criterion 1: The biological effect should be Genetic Addiction Risk Score (GARS) 303
“body friendly” (i.e., biologically and Pharmacogenomics: Customized addiction
immunologically compatible) 301 medicine 303
Criterion 2: Gene polymorphisms identified Fiction 304
and targeted for nutrigenomic effects Facts 304
need to exert a significant influence on Benefits of GARS 305
the origin of a problem and the source of Have we created DNA-customized nutrition? 305
the targeted outcome 301 Genetic testing for customization 306
Criterion 3: The nutraceutical benefit should Conclusion 307
not only correct the target tissue structure References 308

INTRODUCTION opiate/opioid epidemic in America? The genomics

era has brought forth new hope for the future of
In 1969 we used precision to land on the moon, medicine and, in particular, a better understanding
so why can’t we use precision medicine to stop the of psychiatry. We now understand the important

295
296  Customized DNA–directed precision nutrition to balance the brain reward circuitry

roles of DNA, polymorphic associations, and brain happiness. However, in the twenty-first century
reward circuitry in a new perspective on addictive humans face daily reminders of horrific diseases,
behaviors. We present here a review of neurogenet- like cancer, that develop because of genetic and
ics and nutrigenomics as potential keys that pro- epigenetic differences and lead to fatalities and
vide more accurate genetic diagnoses to improve mental impairments that influence billions of
dopamine D2 agonist therapy and dopaminergic neurons and trillions of synapses. This molecular
activation. Through many experiments, we are
­ rearrangement of our genome makes each of us
ready to propose a novel approach that challenges unique. For example, how dopamine functions
the recovery field to incorporate genomic-based in our reward system may also be unique (Rena
tools in treatment programs at addiction clin- et  al., 2001). One example, among other gene
ics. We propose using the Genetic Addiction Risk variations involving brain reward, is that genetic
Score (GARS)™ for appropriate reward deficiency differences account for the presence of attention
syndrome (RDS) diagnosis, the Comprehensive deficit/hyperactivity disorder (ADHD), a subtype
Analysis of Reported Drugs CARD™ to determine of reward deficiency syndrome (RDS), in approxi-
both compliance and abstinence during treatment, mately 8%–12% of children in the United States
natural D2 agonistic therapy (KB220PAM™), and and 4% of adults worldwide (Blum et  al., 2008a).
eventually mRNA (patent pending) to determine You also may be surprised to know that, at birth,
pre- and postcandidate gene expressions in RDS. an estimated 100 million people in the United
We are, therefore, proposing a paradigm shift we States carry a gene form (allele) of just one genetic
have called “Reward Deficiency Solutions System variation that involves brain dopamine D2 recep-
(RDSS)™” (Blum et al., 2014a), most recently referred tors. We know that carriers of the allele DRD2Taq
to as “Precision Behavioral Management” (PBM). A1 have 30%–40% lower D2 receptors in the brain
Have you ever wondered why so many people in (Noble et al., 1991). So what does this mean regard-
America and across the globe are falling victim to ing our romance with getting high—“turning on”
the chain of addictive behaviors, part of the worst and “turning off” with potent psychoactive drugs
epidemic in the history of the word? The answer (e.g., alcohol, cocaine, and opiates)—and resultant
is in part both genetic and environmental (epigen- addiction and fatalities seen in our kids?
etic). Recently researchers found that epigenetic In 1990, the first association of a variant (A1)
effects on the chromatin structure of our DNA are on the dopamine D2 receptor gene (DRD2) and
a legacy that passes from generation to generation. severe alcoholism was discovered and published
Scientists like Stephen Hawking suggest that we by Blum and Noble et  al. in JAMA. Later experi-
are made up of self-assembled molecules generated ments showed that individuals who carry this vari-
over 14 billion years. More interesting is that we as ant have 30%–40% lower dopamine receptors than
Homo sapiens differ in our DNA by only 0.5%. The DRD2 A2 carriers (Noble et al., 1991). Being born
latest research demonstrates that each person has with this single gene variation (DRD2 A1 form),
on average about 60 new mutations in comparison which causes low dopamine receptors, sets an indi-
to their parents. Even more remarkable, the human vidual up to have a high addiction risk (vulnerabil-
brain contains billions of neurons working in con- ity) to any substance or behavior that stimulates
cert to provide us the gift of “well-being” free of the neuronal release of dopamine. In fact, in 1996,
mental disease and stress. The number of neurons in Blum’s laboratory used a mathematical model
the brain varies dramatically from species to species. (called the Bayesian theorem) to predict that an
One estimate (published in 2012) puts the human individual born with the A1 allele (variant) has
brain at about 85 billion neurons and approximately a 74.4% risk of developing an RDS behavior like
85 trillion synapses. It turns out that 20% of our an addiction (Blum et al., 1995). People with that
entire body’s energy is budgeted to keep our brain allele will have an initial acute response to using
working normally. The differences between individ- a psychoactive drug or experiencing pathological
ual humans are approximately the 4.25 billion neu- gambling, or whatever behavior stimulates enough
rons and 4.25 trillion synapses that make us unique. neuronal dopamine for them to feel normal well-
This difference affects the 7.4 billion humans being possibly for the first time. Unfortunately,
that roam our earth working and living together chronic consumption/experiences lead to epigene-
to achieve some degree of productivity and tic changes that further reduce dopamine receptor
 Introduction 297

numbers and a stronger need to abuse can lead to

unwanted uncontrollable behaviors and even nar- T A T A
cotic overdose followed by death. T A T A
How were the genes involved in reward found?
The chemical messengers (neurotransmitters) in GC SNP GC
the brain are like keys that turn on various func- A T G C
tions of genes. The neurotransmitters that partici- A T A T
pate in evoking pleasurable feelings in the reward T A T A
circuitry work in a cascading fashion throughout
the brain. These interactions (the brain reward
cascade [BRC]) may be viewed as activities of
T A T A
subsystems of a larger system with simultaneous
G C G C
cascades. The goal is the generation of feelings of
well-being by the eventual release of just the right
amount of dopamine at the reward site. In this Following 28 years of extensive research from
scenario, there are at least seven major neurotrans- many scientists worldwide, a panel of 11 reward
mitters and their pathways involved: Serotonin, gene risk variants called the Genetic Addiction
endocannabinoids, endorphin (enkephalin), GABA, Risk Score (GARS) has been developed. When
glutamine, acetylcholine, and dopamine. There are GARS was compared to the Addiction Severity
thousands of published studies about these reward Index (ASI) used in many clinical settings, it was
genes and pathways that influence the function found to predict the severity of both alcohol and
of these named neurotransmitters (Blum and drug dependency (Blum, 2015) significantly.
Kozlowski, 1990). This research involved the iden- Currently the GARS test is commercially avail-
tification of gene (DNA) variations or alleles that able and is the property of Geneus Health LLC.,
individuals are born with and epigenetic (environ- located in San Antonio, Texas. The GARS test is
mental RNA) changes that may alter the healthy, now available and for the first time, clinicians and
intended function of DNA. parents will be able to assess the vulnerability of
Dysfunctional DNA is due to what is referred their patients and, most important, of their chil-
to as single nucleotide polymorphisms, frequently dren to chemical dependency and RDS behaviors
called SNPs (pronounced “snips”). SNPs are like addiction, ADHD, and autism spectrum dis-
the most ubiquitous of genetic variations. Every orders. The common thread across all these risk
SNP represents a single nucleotide and they can gene variants is that they lead to a low dopamine
therefore replace cytosine with thymine within a (hypodopaminergia) function or deficit. There are
stretch of DNA. SNPs regularly occur throughout arguments against testing such as fear of labeling
individuals’ DNA; on average they can be found and arguments for testing like knowing the risk.
approximately every 300 nucleotides, translating The real issue or challenge, however, is what can be
to nearly 10 million SNPs in the human genome. done if risk alleles are found. It is understandable
They are commonly found at DNA points between that when there is one gene–one disease (OGOD)
genes and can act as biological markers of genes illnesses like in Huntington’s disease, and when
specific to various diseases. When SNPs occur treatment is unavailable, and prevention remains
within or near a gene (in a regulatory region), they a problem, why know the risk?
alter the gene’s function. If these SNPS show up in Have we found a safe gentle nonaddictive solu-
the brain reward cascade-set of genes (Blum et al., tion that will provide the brain a means to balance
2014c), the neurotransmission will be dysfunc- the neurotransmitters involved in the BRC culmi-
tional resulting in a loss of dopamine regulation nating in true dopamine homeostasis?
or balance (homeostasis). Too little dopamine will Despite variant genes, and epigenetic and envi-
at birth predispose people to “want” or “like” psy- ronmental insults, holistic approaches like mind-
choactive drugs or even behaviors like hypersexu- fulness, exercise, spirituality, and particularly
ality and gambling. Compromised DNA with risk amino acid therapy (KB220PAM formulations)
variations (alleles) can predispose them to become have been shown to reduce relapse and increase
victims to the chain of addictive behaviors. brain dopamine homeostasis (Blum et  al., 2012).
298  Customized DNA–directed precision nutrition to balance the brain reward circuitry

We suggest that clinicians will now be able to between nutrition, the response of genes, and their
genetically test our children for unwanted reward combined effects on overall health and behavior.
gene risk variants that predispose them to dopa- However, not all nutrigenomic technologies are
mine deficiency and lack of reward and risk for the same. To better evaluate the benefits of DNA
addiction, but possibly even use this genetic testing customization, a few questions need to be explored
to circumvent RDS behaviors (Blum et al., 2014b). and answered.
Genetic risk for substance abuse and other RDS
behaviors can be identified by the GARS test and ●● What is the importance and value of DNA-
explain why some individuals are vulnerable and customized nutrition (why use it)?
others not. With continued research, genetic and ●● What exactly is DNA-customized nutrition?
epigenetic dopamine deficiencies can be treated, ●● What criteria need to be satisfied for DNA-
relapse reduced, and we can free ourselves from customized nutrition to be worthwhile?
the clutches of powerful addictive behaviors and
bring balance and happiness to our lives. WHAT IS THE IMPORTANCE AND
VALUE OF DNA-CUSTOMIZED NUTRITION
(WHY USE IT)?
DNA CUSTOMIZATION
IS A REALITY As mentioned earlier, although we all have basic
similarities that may appear to make us pretty
The advent of DNA customization breaks through much the same, or at least make us need the same
the traditional paradigm of one-size-fits-all things in general, the fact is that there are many
dietary supplementation (and even drug therapy) distinct differences in humans that can alter basic
and can present a promising new, more accurate individual needs. Some of our most obvious differ-
and beneficial standard of nutrition. Individuals ences include gender, age, race, body types, body
can now receive supplementation of nutraceutical sizes, tastes and appetites, physical and mental
products that are designed for them and custom- capabilities and activities, talents and skills, eye
ized to meet their unique genetically determined and hair color, and blood types. Other less obvious
needs. No more trial-and-error guesswork is nec- differences can influence our bodies and behav-
essary. This new model provides nutritional sup- iors just as profoundly. Some of these differences
plements that offer significantly greater value to include, among others, variations in
the individual than conventional one-size-fits-all
formulations. ●● Metabolic processes (the ability to produce and
The introduction of DNA-customized technol- regulate energy)
ogy represents a fundamental change in dietary ●● Ability to tolerate stress
supplementation. As history demonstrates, in ●● Emotions
addition to availing hope and opportunity, change ●● Behaviors
can also be met with skepticism, doubt, and even ●● Lifestyle (i.e., food, drink, and activity choices)
resistance. For various reasons, many people just ●● Immune system
don’t want to change the status quo, especially
those with a heavily vested interest in conven- The wide range of variations in these differences
tional dogma. Yet there are others, who, seeing creates a variety of differences in our individual
the possibilities, want to partake in the exciting nutritional needs. These differences result from
new paradigm and market inferior versions of the genetic differences and are responsible for cultural
technology, albeit with anecdotal data. They offer diversity, as a root cause (that is, polymorphisms in
the promise of DNA customization with woefully genotypes and their gene expressions). An exam-
inadequate technology capabilities and tools. ple of this is the fact that alcohol affects Asians
and Native Americans quite differently than
Practical applications Caucasians. Genotypes represent specific gene
variants. An individual is not and does not have
The field of science out of which DNA custom- only one genotype but a vast number of genotypes.
ization has emerged is called nutrigenomics. In order to effectively utilize DNA-customized
Nutrigenomics is the study of the relationships formulations, it is necessary to know the many
 DNA customization is a reality  299

genotypes of an individual across multiple biologi- Consumers’ obsession with finding better ways
cal and metabolic pathways. to improve their health has resulted in a meteoric
In the earliest years of dietary supplementa- rise in the quantity and range of dietary choices
tion, vitamin and mineral preparations were and alternatives, especially in dietary supplements.
based on the lowest common denominator, that This is evidenced by the wide availability of a
is, what minimum amount of nutrients seemed to mind-numbing array of dietary supplements from
best address the most basic and common needs which to choose the “one that’s best for me”—clear
of the “masses.” Many supplement companies evidence of an intrinsic desire for customization
followed the Recommended Dietary Allowance and personalization. The search for the best sup-
(RDA) guidelines to meet the basement of bio- plement to achieve a health objective (e.g., weight
logical necessity. The inadequacies of that mini- loss or increased athletic performance) or address
malist model from those early days through to an ill-health condition (e.g., lower cholesterol,
the present day became evident. This realization resolve digestive disorders, address depression) is
has led to elementary types of “category custom- almost endless.
ization” to address the different nutritional needs The problem is that one often observes “what
caused by those factors cited earlier (gender, age, worked for somebody else” and then thinks that
athletics, blood type, etc.). Even with this sim- product will apply to oneself. We often also take
ple type of customization, a relative “sameness” the advice or recommendation of “informed” store
of the formulas is still maintained within those personnel to make supplement choices. In some
groups (e.g., all women get the same type of wom- cases, the product will work. But, in many cases,
en’s formula). they barely work or don’t work at all. Moreover,
While addressing some “general” differences sometimes the product effect backfires, promot-
within the categories, these conventional formu- ing negative results or reactions, thus confirm-
las were, and still are, one-size-fits-all, that is, for ing evidence that we have genetic differences.
example, young women need more iron and young Nevertheless, even dietary supplement companies
men need more zinc. One-size-fits-all formulas do try to get in on the popular trends and formulate
not meet an individual’s specific needs and can products very similar to popular-selling versions
inadvertently create some nutritional imbalances. on the market, perpetuating the one-size-fits-
Moreover, there are apparent flaws and inad- all and works-one-way formula fads. Innovation
equacies of common lifestyles which, when com- seems to be confined to the next best miracle fruit
bined with a one-size-fits-all supplement, are extract or the like.
ineffective in stopping the increasing levels of With the discovery of DNA and the advent of
chronic pain and mental disorders, such as obesity, the human genome project, an entirely new dimen-
cardiovascular disease, hypertension, Alzheimer’s, sion of and opportunity for nutritional custom-
ADHD, and the current opiate/opioid epidemic. ization was born in the field of nutrigenomics. An
Importantly, certain groups of people were, and important fact must now be punctuated: One size
are, more prone (i.e., predisposed) to suffer a doesn’t fit all for your clothes or your health needs.
higher incidence and severity of certain types Another important factor crucial to the success of a
of conditions. Examples include Caucasian and DNA-customized nutraceutical product, previously
Asian women being more prone to osteoporosis mentioned, is whether the genes being targeted for
and ­people of African descent being prone to more nutraceutical customization are the most appropri-
severe cardiovascular disease, diabetes, and obe- ate to provide the greatest benefit. In fact, DNA cus-
sity. This fact is underlined by the enormous quan- tomization eliminates the trial-and-error guesswork
tity and range of pharmaceutical drugs created to of finding the best product formulation for the indi-
address these disorders in different ways and the vidual, ensuring the greatest and most sustainable
plethora of adverse side effects that can occur due benefit. This is easier said than done, but remains a
to differences in how patients react to the drugs. laudable goal for the future of general medicine and,
Attempts to find natural remedies to address these in this context, addiction medicine. Blum’s group
ill-health conditions gave rise to an ever-increasing has already published on this possibility earlier
selection of one-size-fits-all “condition-specific” about KB220 variants in obesity (Blum et al., 2008d,
dietary supplement formulas as well. 2008e, 2012b, 2014a; Plomin et al., 1994).
300  Customized DNA–directed precision nutrition to balance the brain reward circuitry

WHAT EXACTLY IS A DNA- 14. Risk alleles of a number of important physi-

CUSTOMIZED NUTRACEUTICAL? ological pathways are selected on the basis
of a common thread (i.e., hypodopaminergic
DNA analysis and nutraceutical customization trait/state). Each risk allele must have been
is a real possibility now but will always require associated with a particular risk compared to
additional focused research for development of controls in many published studies.
advanced products. Following is an overview of
nutrigenomic DNA customization process:
WHAT CRITERIA NEED TO BE
1. Find the right science-based ingredients SATISFIED FOR DNA-CUSTOMIZED
that influence therapeutically important NUTRITION TO BE WORTHWHILE?
­pathways—a crucial factor.
2. Existing research must validate the dose- Any one part of your body cannot be
dependent mechanisms of action (MOA) and well, unless you care for the whole body.
beneficial effects of those ingredients.
3. Determine if, and by how much, those ingredi- —Attributed to Plato
ents influence gene expression.
4. Dose-dependent results are reviewed to It is clearly beneficial to address foundational
establish the baseline dose needed to achieve health factors that affect the whole body in con-
beneficial gene-expression effects in a reason- junction with other, more focused targets (syner-
able portion of the population. gistic symbiosis). For example, to address weight
5. Research is then either conducted or, if avail- loss without first considering genetic factors and
able, reviewed to determine a range of dose- the controlling involvement of the brain is a for-
dependent gene-expression results. mula for failure, repeatedly demonstrated with
6. Dose-dependent gene expression results are conventional methods. The shortcoming of most
then compared to results in the MOA and conventional therapeutic tactics, even nutraceuti-
efficacy research. cal ones, is that they rely solely on narrow, com-
7. Ingredients selected through this process partmentalized, and/or downstream symptomatic
become “DNA qualified” and are used in targets (Rana et al., 2016). Some examples of these
DNA-customized formulations. types of benefit targets include:
8. Research is then reviewed to determine which
genotypes exhibit the kind of gene expressions ●● Anti-inflammatory
noted in the gene expression studies. ●● Appetite suppression
9. Algorithms are then established for DNA- ●● Blocking carb absorption
customized dose-dependent applications for ●● Blocking fat absorption
each DNA-qualified ingredient based on the ●● Diuretics
genotypes for each pathway. ●● Stimulants
10. Then research is reviewed or conducted on ●● Calming and soothing
select populations of genotype-qualified ●● Libido enhancement
candidates to confirm and/or better define and ●● Cholesterol-lowering
catalog the efficacy parameters. ●● Blood sugar lowering
1 1. Once this process is completed, genuine DNA-
qualified ingredients can be customized for Conventional nutraceutical formulations that
the gene polymorphisms of an individual. address these types of “benefit” targets routinely
12. Specific genes of an individual are analyzed, ignore the gene-expression-based mechanisms of
via analysis of a buccal swab. underlying “upstream” causes promoting their
1 3. Genotypes for crucial pathways are deter- “downstream” conditions. They are generally
mined and ingredient potencies adjusted to targeted at mechanisms of action that achieve
accommodate gene-expression variations of superficial or temporary symptomatic relief at
the individual. best.
 What criteria need to be satisfied for DNA-customized nutrition to be worthwhile?  301

There could be many “causes” for a sup- formula’s ingredients have on the body. For exam-
posed malady and its symptoms. For instance, a ple, pharmaceutical drugs, stimulants, fat blockers,
depressed individual might try taking St. John’s and mega-dose nutrient overloading (pharmaco-
wort (Apaydin et  al., 2016) to feel better (trial- logic-like) effects, including all stimulants, sup-
and-error). Importantly, there are many causes pressants, and calorie deprivation tactics could
and types of depression. What works for one cause fight against what has been termed “body friendly.”
and type might not work for another or even make
other types worse. Knowing the genotypes of genes Criterion 2: Gene polymorphisms
in certain pathways in the brain would enable a identified and targeted for
DNA-customized formula to promote healthy nutrigenomic effects need to exert
brain function and reduced guesswork.
Another example of this premise is the use of
a significant influence on the origin
fat blockers to reduce fat calorie absorption and of a problem and the source of the
promote “weight” loss. The fact is that, in the end, targeted outcome
calorie blocking and deprivation work against the
body’s genetic survival instructions. The initial In simple terms, get to the genetic source. For
effect could be to reduce fat calorie absorption, example, in contrast to single-effect stimulants
resulting in some initial weight loss. But, such and calorie-deprivation tactics, weight loss needs
deprivation tactics create a survival crisis, trigger- to be the result of correcting biosystem inadequa-
ing a genetically mandated protection response. cies. At least five major pathways need to be simul-
The longer-term result of this effect is to lower taneously addressed:
the basal metabolic rate (rate of calorie burning), ●● Energy metabolism (production and
increase fat storage, and increase cravings and regulation)
food consumption to ensure survival, which pro- ●● Reward and craving management in the brain
motes greater weight regain (colloquially referred ●● Neuroendocrine management
to as yo-yo weight gain) (Casazza et al., 2015). ●● Stress management
Many conventional nutraceutical products, espe- ●● Immune system function
cially condition-specific formulas, target symptom-
atic relief. In contrast, by exerting a nutraceutical Criterion 3: The nutraceutical
influence on gene expressions, regarding precision benefit should not only correct
medicine, DNA-customized nutraceutical products
must address the whole body, and, importantly, the
the target tissue structure and
brain’s control of all bodily functions (neuro-nutrige- function, it should make a significant
nomics). The result corrects the effects of specific gene contribution to overall health
variants (polymorphisms), normalizing target tis-
To develop DNA-customized and nutrient-specific
sue function. Moreover, DNA-customized formulas
formulas for RDS (Downs et al., 2009) and all sub-
should be symbiotic, improve the overall health and
stance and nonsubstance-seeking addictive behav-
function of the body in a synergistic manner, and sat-
iors, all ingredients and their interactions must be
isfy the criteria required to provide efficient and sus-
considered as they work in concert to balance the
tainable outcomes. Those criteria are explained next.
BRC (Blum et al., 2016a) that will provide an effect
on overall health of an individual.
Criterion 1: The biological effect
should be “body friendly” (i.e., Criterion 4: The beneficial effects
biologically and immunologically should be obvious, different,
compatible) transformational, and sustainable
for most patients
DNA-customized nutrition is satisfying a need of
the body. Therefore, the body should not have to In vitro studies can simulate a powerful mecha-
protect, reject, or retaliate against the effects the nism of action in cell cultures. However, when
302  Customized DNA–directed precision nutrition to balance the brain reward circuitry

human studies are conducted on the same product Third, conventional methods do not make a sig-
to find a specific benefit (i.e., blocking fat absorp- nificant contribution to overall health. It is indisput-
tion to induce weight loss) the outcomes are usu- able that weight loss is associated with many health
ally skewed across a range of benefits. The results benefits. However, eventual reduced health and
could look like the following: amplified yo-yo weight fluctuation is highly prob-
able after significant weight decrease due to the
●● 30% of the people received strong benefits (also aforementioned first and second criteria, and the
a possible placebo effect) fact that absorption of healthy lipids like Omega-3s
●● 30% received minor benefits and fat-soluble vitamins are blocked in the process.
●● 15% received no benefits Fourth, conventional methods do not provide
●● 15% actually experienced reverse effects, that obvious, different, and positive transformational
is, increased weight gain effects that are sustainable for most patients. We
must keep in mind that most dietary supplement
According to a traditional perspective, this out- research, especially for weight loss products, evalu-
come would still be considered positive and have ates results over an 8- to 12-week period. This is
significant medical value. However, in this exam- a very short-term window in a long-term process.
ple, these results suggest that as much as 70% of the Considering the reality of the weight gain–rebound
patients will experience some level of disappoint- effect, this is an inadequate time to evaluate more
ment. This is the primary reason that many peo- telling longer-term effects. In fact, history reveals
ple stop taking a weight loss product within 3 to that conventional weight loss programs almost
6 months. This also disregards the yo-yo rebound without fail have indeed failed. If the products
weight gain effect (the failure effect), which work at all, conventional calorie-deprivation
appears to happen in at least 95% of the cases with weight loss tactics (including fat blockers), with-
time. The need for precision medicine will come of out exception, provide results for only a temporary
age especially when genetic testing could provide period. Ultimately, conventional methods lead to
meaningful results and medical benefits. the regaining of weight. Given this, and due to the
facts cited in the first three criteria, for the vast
Conventional methods do not comply majority of people, the results are not obvious, not
with all of the four important criteria different from other similar-type products, not
transformational, and not sustainable.
First, traditional methods might not be “body A major reason that human research results are
friendly.” In many cases products impose a reduced skewed is due to the wide variation in genetic fac-
fat and calorie intake (blocking) on the body. The tors. Identifying these factors enables the use of
initial effect could be weight loss (especially in the right therapeutic modalities in the appropri-
the 30% of people). However, the long-term effect ate amounts to achieve the greatest benefits in the
will be the result of the body’s attempt to pro- most people for the longest time. When done cor-
tect itself from the “famine” or “be thrifty” (Joffe rectly, the incredible value of DNA customization
and Zimmet, 1998) upshot. This effect might be will be indisputable.
significantly greater in certain individuals in the
70% group. The body will ultimately try to protect GENETIC TESTING: FACT
against reduced nutrient absorption with increased OR FICTION
fat storage, a natural survival mechanism.
Second, “ fat-blocking” methods do not address Although some physicians may think that by using
multiple metabolic or genetic causes of the problem. 23andMe genomic testing they could adequately
Just blocking fat absorption is not addressing the determine a person’s risk for RDS behaviors, they
many other issues related to obesity, whereby many are mistaken. In the first place, the FDA did not
genes and subsequent polymorphisms play impor- allow 23andMe to provide any interpretive data to
tant roles in the induction of aberrant cravings, an individual related to medical health issues and
enhancing brown fat synthesis (Polyák et al., 2016), in fact threatened removal of a such a misleading
and inducing metabolic anomalies (Higginson test. While 23andme obtained new clearance and
et al., 2016). approval on a number of gene-based predisposition
 Genetic testing: Fact or fiction  303

tests, the de novo approval does not cover RDS or adopting a Bayesian approach, determined a posi-
addiction. The following information will help tive predictive value (PPV) for the DRD2 A1 variant
inform the medical community about cautions (fewer D2 receptors) of 74% (Blum et al., 1995). This
concerning genetic testing in general. PPV demonstrated that children with this polymor-
Concerning molecular genetic testing, there phism have a very high risk of addiction disorders
are three types of current interest. They are phar- involving drugs, food, or other aberrant behaviors
macogenetics, primarily evaluating metabolizing in their future. Since these findings, global laborato-
enzymes for high and low metabolizers with for ries, including the National Institute on Drug Abuse
example opiates; Genetic Addiction Risk Score, to (NIDA) and the National Institute on Alcohol Abuse
determine through a panel of reward gene poly- and Alcoholism (NIAAA), have not only confirmed
morphisms stratification risk or vulnerability to all this early work (Xu et al., 2004) but also extended the
RDS behaviors including pain tolerance; and phar- scope of many other candidate genes, in particular
macogenomics, personalized addiction medicine genes and second in the reward circuitry of the brain
based on genotyping an individual and targeting (Goldman, 1995).
specific gene loci. For instance, Moeller et al. (2013) suggested that
drug cues contribute to relapse, and their neuroge-
Pharmacogenetic testing netic results have identified the DAT1R 9R-allele as
a vulnerability allele for relapse especially during
Various alleles in the P450 system are currently uti- early abstinence (i.e., detoxification). The DAT1R
lized in pain medicine clinics to evaluate metabolic 9R-allele influences the fast acting transport of
concerns to help identify high and low metaboliz- dopamine sequestered from the synapse leading to
ers. For the most part, this has not translated to sig- a hypodopaminergic trait.
nificant clinical utility, but may have some relevance It is practical to employ genetic testing to dis-
regarding buprenorphine/naloxone treatments. cover reward circuitry gene polymorphisms, espe-
When used in conjunction with GARS this could cially those related to dopaminergic pathways and
provide valuable information (De Fazio et al., 2008). opioid receptors. By determining genetic vulner-
abilities and risks, treatment outcomes can vastly
Genetic Addiction Risk Score (GARS) improve. Understanding reward circuitry involve-
ment in buprenorphine effects and genotypes
We now know that nature (genes), nurture (envi- provide more insight into personalizing patients’
ronment), and behavioral outcomes result from a clinical experience during opioid replacement
combination of 50% genes and 50% epigenetics. therapy (Blum et al., 2014b).
Thus, molecular genetic or DNA testing is criti- Blum’s genetic risk score represents a panel of
cal, especially in linking aberrant behaviors to any known reward genes and associated risk polymor-
individual. phisms providing genetic risk for addiction and
Blum’s laboratory proposed that disruptions from other behaviors including medical monitoring and
gene variations (polymorphisms) or the environment clinical outcome response.
(epigenetics) within this brain reward cascade result
in aberrant addictive behaviors or RDS. Despite the Pharmacogenomics: Customized
international research to discover specific or poten- addiction medicine
tial gene clusters from high-density SNP arrays,
many attempts have proven inconclusive. However, Furthering this work, Blum and Gerald Kozlowski
more recently, Palmer et  al. (2015) observed that identified the brain reward cascade (BRC;
25%–36% of the variance in vulnerability to sub- Blum et al., 2016b). BRC as a concept acts as a blue-
stance dependence is attributed to common single print explaining neurotransmitter interactions in
nucleotide polymorphisms. The effect is additive the reward circuit. In addition, respective reward
since such common single nucleotide polymor- genes that regulate such neurotransmitters ulti-
phisms are shared across the indicators of comorbid mately control the amount of dopamine released
drug problems. Recent research has also revealed into both the reward site and other brain regions.
that specific candidate gene variants account for The integrity of resting state functional connec-
risk prediction. Earlier work from Blum’s laboratory, tivity is crucial for normal homeostatic function.
304  Customized DNA–directed precision nutrition to balance the brain reward circuitry

Zhang et  al. (2015) showed that heroin addicts lacking panels of other candidate genes and out-
exhibit reduced connectivity between the dorsal come studies. Another example is that a well
anterior cingulate cortex (dACC) and rostral ante- known Pharmacogenetic company claims that hav-
rior cingulate cortex (rACC), as well as reduced con- ing one polymorphism in the mu opioid receptor
nectivity between the subcallosal anterior cingulate leads to opioid dependence. One gene alone cannot
cortex (sACC) and dACC. The heroin addicts’ varia- provide the necessary information to make such a
tions in functional connectivity of three subregions claim because it is more complex. Other misleading
of the ACC indicate that these three subregions, claims suggest that the patient results are compared
along with other key brain areas (such as the dor- to population controls. Our review of their “dis-
sal striatum, putamen, orbital frontal cortex, dor- ease-free” controls revealed substantial flaws, with-
sal striatum, cerebellum, amygdala) potentially out controlling for any RDS behaviors (Chen et al.,
modulate heroin addiction. Blum’s laboratory and 2005). For their claims to be true, they would have
Zhang’s group (Blum et al., 2015b) showed that in to utilize what has been termed “super controls.” In
abstinent heroin addicts, KB220Z™ (a putative dopa- short, population controls may carry many unseen
mine D2 agonist) increased BOLD activation in RDS behaviors that must be identified so that the
caudate-accumbens-dopaminergic pathways, com- control would be RDS free. Also the severity of the
pared to a placebo following one-hour acute admin- phenotype in question must be considered as well
istration. In addition, KB220Z™ also induced the otherwise spurious results will ensue.
reduction of resting state activity in the putamen. In To reiterate, otherwise, adopting genetic testing
the second phase of this pilot study, all 10 abstinent will lead to spurious and inaccurate results. Some
heroin-dependent subjects had three brain regions of these companies have selected genes for testing
of interest (ROIs), which were shown to be signifi- that may be involved in risky behavior, but they are
cantly activated from the resting state by KB220Z™ not using the proper variant in their tests, or they
compared to the placebo (P < 0.05). Increased func- use rare variants that do not truly prove addic-
tional connectivity was also observed in a system tion risk. Indeed, Mayer and Höllt (2006) cor-
involving the dorsal anterior cingulate, medial fron- rectly stated that the “vast number of non-coding,
tal gyrus, nucleus accumbens, posterior cingulate, intronic or promoter polymorphisms in the opioid
occipital cortical areas, and cerebellum. receptors may influence addictive behavior, but
Employing DNA-based testing, successful these polymorphisms are far less studied, and their
development of polymorphic gene testing enables physiological significance remains to be demon-
customized (personalized) antiobesity com- strated.” Such companies have never investigated
pounds, exemplifying the future of personalized the scientific integrity and predictability of their
addiction medicine utilizing Geneus Health’s genetic full-panel tests, let alone actual addiction
Genetic Addiction Risk Score (GARS). risk or any associated behaviors.

Fiction Facts
Although there is a slew of promising experiments Although we, the authors, may have a personal bias
for RDS behaviors involving candidate gene asso- due to the many years that the Blum laboratory has
ciations, there are also some negative outcomes dedicated to developing an accurate genetic test
(Doremus-Fitzwater and Spear, 2016). to predict true liability/risk for RDS and associ-
Currently, a few companies have entered into ated behaviors, we will attempt to explain why the
genetic testing in the addiction and pain medication successful development of the first GARS in con-
industry to offer what they claim to be “personal- junction with the Institute of Behavioral Genetics,
ized care.” However, many of these companies have University of Colorado, Boulder has the real poten-
not fully understood the science behind the appli- tial of impacting the entire addiction and pain
cation. They often make exaggerated commercial medicine field.
claims using Blum’s original work, stating that their To develop GARS™ we first selected (1) ten
genetic test is 74% predictive (Blum  et  al.,  1995). reward candidate genes (DRD1, 2, 3, 4; DAT1; sero-
This is obviously false as they used a single gene tonin transporter, COMT, MAO, GABA, Mu opi-
(DRD2) to support their predictability claim, ate receptor) and (2) a number of SNPs and point
 Have we created DNA-customized nutrition?  305

mutations that influence the net release of dopa- ●● Attenuation of guilt and denial
mine at the brain reward site. The SNPs and point ●● Corroboration of family geneograms
mutations were chosen to reflect a hypodopami- ●● Risk severity–based decisions about appropri-
nergic trait. To validate our methods we partnered ate therapies including pain medications and
with the developers of the Addiction Severity risk for addiction liability
Index-Media Version (ASI-MV), a test mandated ●● Appropriate level of care placement, for
in 18 states that determines both alcohol and drug example, in-patient, out-patient, intensive out-
severity risk scores (Butler et al., 2015). patient, residential
In the process, we collaborated with seven ●● Length of stay in treatment
diverse treatment centers across the United States, ●● Genetic severity–based relapse and recovery
obtaining a total of 393 subjects that we genotyped liability and vulnerability
using the GARS panel. All data were genotyped and ●● Pharmacogenetic medical monitoring for bet-
analyzed at the Institute for Behavioral Genetics. ter clinical outcomes, for example, the A1 allele
For 273 subjects, we discovered a marked associa- of the DRD2 gene reduces the binding to delta
tion between the summed score of GARS panel risk receptors in the brain, reducing naltrexone’s
alleles (variant forms) and both the drug (P < 0.05) clinical effectiveness
and ASI-MV alcohol (p < 0.004) severity indices.
In effect, the correlation indicated the higher HAVE WE CREATED DNA-
the number of risk alleles, the stronger the predic- CUSTOMIZED NUTRITION?
tion of alcohol or drug use severity. Notably, fam-
ily or domestic problems, psychological issues, and In light of these early studies and a scarcity of
medicalization significantly correlated as well. relevant research, we decided to design a study
Interestingly, though, a change in any specific SNP evaluating the process of DNA customization of a
resulted in no correlation. This signifies the impor- nutritional solution for both wellness and weight
tance of selected GARS panels, as any deviation management. We review the results of a number
will produce false results similar to scientifically of studies (Blum et  al., 2008c, 2008d, 2008e) in
vapid commercial tests. Weighting each allele by which Blum’s laboratory genotyped 1058 subjects
increasing its score power also resulted in lost sig- who were administered KB220z (formerly LG9939,
nificance suggesting the importance of a clustering Recomposize, Genotrim), a complex neuroadapta-
of genes implicated in the BRC rather than any one gen nutraceutical-dl-phenylalanine, along with
gene polymorphism by itself. chromium, l-tyrosine, and other select amino acids
and adaptogens, based on polymorphic outcomes.
BENEFITS OF GARS The customized formulas involved a minimum of
175 SNPs covering 16 genes important to the BRC
We found that a 10-gene panel driven by risk and, most importantly, involved in “dopamine
alleles for hypodopaminergic state, consisting of 11 homeostasis.” Within a small subset, simple t-tests
SNPs, was significantly associated with the level of comparing parameters before and after 80 days on
ASI-MV alcohol and drug risk severity score, sup- the nutraceutical were performed.
porting clinical relevance and utility. The clinical outcomes of the studies were weight
To our knowledge, this is the first and only corre- loss (p < 0.008), appetite suppression (p < 0.004),
lation of a panel of genes, with established polymor- sugar craving reduction (p < 0.008), snack reduc-
phisms that reflect the brain reward cascade with tion (p < 0.005), reduction of late night bing-
the ASI-MV alcohol and drug risk severity score ing (p < 0.007), increased energy (p < 0.004),
ever accomplished to date. While other studies are increased perception of overeating (p < 0.02),
required to further confirm and extend the GARS enhanced quality of sleep (p < 0.02), and increased
test to include other genes and polymorphisms that happiness (p < 0.02). Polymorphic correlates were
associate with hypodopaminergic trait, these results also obtained for various genes (PPAR gamma 2,
provide clinicians with a noninvasive genetic test. MTHFR, 5-HT2a, and DRD2) with positive clini-
Unlike 23andMe genomic testing, other than cal parameters tested in this study. Notably, in
finding ancestral heritage, GARS can be used to all the outcomes and gene polymorphisms, a sig-
improve clinical interactions and decision making: nificant Pearson correlation only occurred for the
306  Customized DNA–directed precision nutrition to balance the brain reward circuitry

DRD2 gene polymorphism (allele Al) with days food consumption, loss of inches around waist,
on treatment (r = 0.42, p = 0.045). This two fold weight loss, reduction in blood pressure, exercise
increase is a very important genotype for compli- performance improvement, reduction in drug-
ance in treatment. seeking behavior, reduction in hyperactivity, and
In addition, Blum’s group systematically evalu- reduction in cholesterol levels.
ated the impact of polymorphisms of these five
candidate genes as important targets for the devel- GENETIC TESTING
opment of a DNA-customized nutraceutical KB220z FOR CUSTOMIZATION
to combat obesity with particular emphasis on body
recomposition as measured by body mass index We therefore propose that utilizing this natu-
(BMI). A total of 21 individuals were evaluated in a ral dopaminergic-activating (potentially DNA-
preliminary investigational study of KB220z. customized) approach over time leads to neuronal
We based the experiment on the results of each DA release at the NAc, causing a proliferation of
subjects’ buccal swab genotyping and a customized D2 receptors (Downs et  al., 2013). We are enthu-
nutraceutical formula was provided specific to the siastic about previous results utilizing KB220Z in
function of measured gene polymorphisms of the neuroimaging and qEEG studies (Blum et al., 2010;
five gene candidates assessed. Subsequently, at the Miller et al., 2010) to demonstrate both enhanced
beginning and every 2 weeks each subject com- resting state functional connectivity in abstinent
pleted a modified Blum-Downs OPAQuE Scale™ heroin addicts (Blum et al., 2015b) and regulation
[Overweight Patient Assessment Questionnaire]. of widespread theta activity in the cingulate gyrus
The involved alleles included the DRD2 Al, MTHFR of abstinent psychostimulant abusers (Blum et al.,
C 677 T, 5HT2a 1438G/A, PPAR-γProl2Ala, and 2010), similar to effects observed in alcoholics and
Leptin Ob1875 <208 bp. Pre and post hoc analy- heroin addicts (Miller et al., 2010).
sis revealed a large difference between the starting Other support for anticraving effects of dopa-
BMI and the BMI after an average of 41 days (28–70 minergic activation in humans can be found in
d) of KB220z intake in 21 individuals. The pre- numerous peer-reviewed published clinical tri-
BMI was averaged at 31.2 (weight/Ht2) compared als including randomized double-blind placebo-
to the post-BMI of 30.4 (weight/Ht2), with signifi- controlled studies. In fact, decreased alcohol and
cance value P < 0.034 (one-tailed). Similarly the cocaine craving behavior was seen with animal
preweight in pounds was 183.52 compared to the gene therapy using cDNA vectors of the DRD2
postweight of 179 pounds, with a significance value gene implanted into the NAc (Myers and Robinson,
of P < (0.047). Trends for reduction of late night 1999; Thanos et al., 2001, 2005, 2008).
snacking, carbohydrate craving reduction, reduc- We the authors are cognizant that dopaminer-
tion of stress, and reduction of waist circumference gic activation in the long-term instead of blocking
were also observed. In the 41-day period, a weight dopamine (Blum et al., 2008b) should be utilized to
loss trend was seen as well: 71.4% of subjects lost treat not only alcohol, cocaine, and nicotine crav-
weight. Fifteen out of 21 subjects lost weight with a z ings, but glucose craving and other known behav-
score of 2.4 and significance value of P < (0.02) and ioral addictions (e.g., gambling, hypersexuality
this group (53%) overall lost on average over 2.5% of [Blum et al., 2015a]). Thus the coupling of genetic
their starting weight. antecedents such as GARS and nutrition may be a
Other preliminary conclusions requiring exten- very viable alternative approach for the treatment
sive investigation, using a path analysis (noncus- of all RDS behaviors (Blum et  al., 2008c, 2008d,
tomized KB220z), include notable associations 2008e).
regarding antiobesity-related behaviors. In a 1-year Additional research from Blum’s laboratory
cross-sectional open trial study of 24 unscreened developed a theoretical modeling study, which
subjects, administration of oral KB220z variant evaluated health and economic implications of
lead to substantial benefits: stress reduction, sleep a nutrigenomic product for weight loss. Blum’s
enhancement, increase in energy level, general- group (Meshkin and Blum, 2007) based a nutrig-
ized well-being, improvement in mental focus/ enomic economic model on linking (1) published
memory, reduction in cravings (sweets/carbs), study data on product efficacy of ingredients, (2)
improvement in blood sugar levels, reduction in validated clinical assessments already used in
 Conclusion 307

health economics data, and (3) data on condition addicts. In fact, the enhanced activation of rest-
prevalence and macroeconomic costs of illness. ing state functional connectivity was observed in
The model demonstrates that a DNA-customized a network involving the dorsal anterior cingulate,
nutraceutical positively decreases the cost of ill- medial frontal gyrus, nucleus accumbens, posterior
ness at the macroeconomic and microeconomic cingulate, occipital cortical areas, and cerebellum.
level, based on cost-effectiveness and cost-benefit In other published work at the University of
analysis. This model can thus forecast important Florida, Febo, Blum, and others found that KB220Z
prognostic health economic implications of a considerably stimulates (above placebo) seed
nutrigenomic intervention to demonstrate a the- regions of interest, such as the hippocampus, left
oretical model of nutrigenomic economics as it nucleus accumbens, anterior thalamic nuclei, cin-
relates to obesity. gulate gyrus, prelimbic and infralimbic loci. The
Reiterations of fifteen variant formulae allowed reaction caused by KB220Z establishes substan-
Blum’s group (Blum et  al., 2008c, 2008d, 2008e) tial functional connectivity and increased brain
to use polymorphic targets of various reward volume recruitment, as well as enhanced dopami-
genes (serotonergic, opioidergic, GABAergic, and nergic functionality across brain reward circuitry.
dopaminergic) to customize KB220 (neuroad- This robust and selective response is clearly clini-
aptogen-amino-acid therapy [NAAT]) with spe- cally significant (Febo et al., 2017).
cific algorithms. For 1000 obese subjects in the
Netherlands, a small subset was administered CONCLUSION
different KB220 formulae customized according
to individuals’ respective DNA polymorphisms, We, the authors, would like to now propose
which translated to significant decreases in both Precision Behavioral Management (PBM), a pro-
BMI and weight in pounds. tocol that promotes early diagnostic identification
Blum’s laboratory developed GARS along with and stratification of risk alleles through the use of
Brett Haberstick, Andrew Smolen, and others. GARS, thereby allowing nutrigenomic targeting
When 10 genes were selected with appropriate of these risk alleles through the alteration of neu-
variants, it appeared that there was a statistically ronutrient amino acid therapy ingredients as an
notable association between the ASI-MV alco- algorithmic function of polymorphic DNA SNPS.
hol and drug severity scores and GARS. This was A novel approach of this can produce the first ever
found true for 273 patients at seven treatment cen- nutrigenomic solution for  addiction, pain, and
ters. This correlation sets the framework for early other RDS addictive behaviors (see Figure 17.1).
clinical diagnosis and identification of addictive Welcome to the new era of genomic addiction
predispositions, showing personalized medicine as medicine (Blum et al., 2015c; Blum and Badgaiyan,
a nutrigenomic solution for RDS behaviors (Blum 2013). So maybe just like the famous moon walk, by
et al., 2014a). using precision medicine we will be able to differ-
Blum et al. (2015b) reported that a well-known entiate our cultural diversity here in America and
variant of KB220PAM actually increased resting globally and develop DNA-directed antiopiate/opi-
state functional connectivity in abstinent heroin oid precision medicine that promotes “dopamine

Genotyping Targeting Customized Dopamine Nutrigenomic

• Identify risk polymorphisms agent release solution
alleles using • Matching KB220z ingr. to • Alter NAAT-KB220Z • Increased • Treat addiction
GARS polymorphisms ingredients DA at NAc and pain

Figure 17.1  Schematic of DNA-customization application for a nutrigenomic solution to RDS.

308  Customized DNA–directed precision nutrition to balance the brain reward circuitry

homeostasis” instead of continued opioid locked- pilot study in The Netherlands. Gene Therapy
in addiction with associated zombie-like behaviors Mol Biol 12(1);2008d:129–40.
(Hill et al., 2013). Utilizing these simple principles Blum K, Chen TJ, Morse S et al. Overcoming
based on empirical genetic evidence concerning qEEG abnormalities and reward gene defi-
various RDS behaviors (Blum et  al., 1990, 1996, cits during protracted abstinence in male
2011; Levey et  al., 2014; Farris et  al., 2014; Ducci psychostimulant and polydrug abusers utiliz-
and Goldman, 2012; Gyollai et al., 2014; Yan et al., ing putative dopamine D2 agonist therapy:
2014; Guo et al., 2016), we are now poised to over- part 2. Postgrad Med 122(6);2010:214–26.
come the deviating unwelcomed opiate/opioid Blum K, Chen AL, Oscar-Berman M et al.
American second epidemic. Generational association studies of dopa-
minergic genes in reward deficiency
REFERENCES syndrome (RDS) subjects: Selecting appro-
priate phenotypes for reward dependence
Apaydin EA, Maher AR, Shanman R, Booth MS, behaviors. Int J Environ Res Public Health
Miles JN, Sorbero ME, Hempel S. A sys- 8(12);2011:4425–59.
tematic review of St. John’s wort for major Blum K, Chen TJH, Williams L et al. A short term
depressive disorder. Syst Rev 5(1);2016:148. pilot open label study to evaluate efficacy and
Blum K. “Dopamine Resistance” in Brain Reward safety of LG839, a customized DNA directed
Circuitry as a function of Genetic Addiction nutraceutical in obesity: Exploring nutrigenom-
Risk Score (GARS) polymorphisms in RDS: ics. Gene Ther Mol Biol 12(2);2008e:371–82.
Synaptamine Complex Variant (kb220z) Blum K, Febo M, Badgaiyan RD. Fifty years in the
induced “dopamine sensitivity” as a Pro- development of a glutaminergic-dopaminer-
Recovery Agent. Presented at the Reward gic optimization complex (KB220) to balance
Deficiency Syndrome (RDS) Inaugural brain reward circuitry in reward deficiency
Conference. San Francisco, November 16–18, syndrome: A pictorial. Austin Addict Sci
2015. 1(2);2016a: pii: 1006.
Blum K, Badgaiyan RD. Addiction research and Blum K, Febo M, Badgaiyan RD, Demetrovics Z,
therapy in the 21st century: Providing a forum Simpatico T, Fahlke C, Oscar-Berman M, Li
for evidence-based addiction medicine. J M, Dushaj K, Gold MS. Common neuroge-
Addict Res Ther 4;2013:1000e117. netic diagnosis and meso-limbic manipulation
Blum K, Badgaiyan RD, Gold MS. Hypersexuality of hypodopaminergic function in Reward
addiction and withdrawal: Phenomenology, Deficiency Syndrome (RDS): Changing the
neurogenetics and epigenetics. Cureus recovery landscape. Curr Neuropharmacol
7(10);2015a:e348. 15(1);2016b:184–94.
Blum K, Chen AL, Braverman ER. Attention- Blum K, Febo M, McLaughlin T, Cronjé FJ, Han
deficit-hyperactivity disorder and reward defi- D, Gold SM. Hatching the behavioral addic-
ciency syndrome. Neuropsychiatr Dis Treat tion egg: Reward Deficiency Solution System
4(5);2008a:893–918. (RDSS)™ as a function of dopaminergic neu-
Blum K, Chen AL, Chen TJ et al. Activation rogenetics and brain functional connectivity
instead of blocking mesolimbic dopaminergic linking all addictions under a common rubric.
reward circuitry is a preferred modality in the J Behav Addict 3(3);2014a:149–56.
long term treatment of reward deficiency syn- Blum K, Gardner E, Oscar-Berman M, Gold M.
drome (RDS): A commentary. Theor Biol Med “Liking” and “wanting” linked to Reward
Model 2008b;5:24. Deficiency Syndrome (RDS): Hypothesizing
Blum K, Chen AL, Chen TJ et al. LG839: anti-obe- differential responsivity in brain reward cir-
sity effects and polymorphic gene correlates cuitry. Curr Pharm Des 18(1);2012:113–8.
of reward deficiency syndrome. Adv Ther Blum K, Giordano J, Hauser M et al. Coupling
25(9);2008c:894–913. Genetic Addiction Risk Score (GARS™),
Blum K, Chen ALC, Chen TLC et al. Dopamine Comprehensive Analysis of Reported
D2 receptor Taq A1 allele predicts treatment Drugs (CARD™) dopamine agonist therapy
compliance of LG839 in a subset analysis of a (KB220z™): Reward Deficiency Solution
 References 309

System (RDSS). Abstract accepted to Casazza K, Brown A, Astrup A et al. Weighing

European Psychiatry Congress, Vienna, the evidence of common beliefs in obe-
Austria, 2014b March 18–20. sity research. Crit Rev Food Sci Nutr
Blum K, Kozlowski GP. Ethanol and neuromodula- 55(14);2015:2014–53.
tor influences. A cascade model of reward. Chen TJ, Blum K, Mathews D et al. Are dopa-
In: Alcohol and Behaviour: Basic and Clinical minergic genes involved in a predisposition
Aspects Progress in Alcohol Research. Ollat to pathological aggression? Hypothesizing
H, Parvez S, Parvez H (Eds), VSP International the importance of “super normal controls”
Science Publishers, Utrecht, Netherlands, in psychiatric genetic research of complex
1990, pp. 131–50. behavioral disorders. Med Hypotheses
Blum K, Liu Y, Wang W et al. rsfMRI effects of 65(4);2005:703–7.
KB220ZTM on neural pathways in reward cir- De Fazio S, Gallelli L, De Siena A, De Sarro G,
cuitry of abstinent genotyped heroin addicts. Scordo MG. Role of CYP3A5 in abnormal
Postgrad Med 127(2);2015b:232–41. clearance of methadone. Ann Pharmacother
Blum K, Noble EP, Sheridan PJ, Montgomery A, 42(6);2008:893–7.
Ritchie T, Jagadeeswaran P, Nogami H, Briggs Doremus-Fitzwater TL, Spear LP. Reward-
AH, Cohn JB. Allelic association of human centricity and attenuated aversions: An
dopamine D2 receptor gene in alcoholism. adolescent phenotype emerging from studies
JAMA 263(15);1990:2055–60. in laboratory animals. Neurosci Biobehav Rev
Blum K, Oscar-Berman M, Demetrovics Z et al. 70;2016:121–34.
Genetic Addiction Risk Score (GARS): Downs B, Oscar-Berman M, Waite R et al. Have we
Molecular neurogenetic evidence for predispo- hatched the addiction egg: Reward Deficiency
sition to Reward Deficiency Syndrome (RDS). Syndrome Solution System™ J Genet Syndr
Mol Neurobiol 50(3);2014c:765–96. Gene Ther 4(136);2013:14318.
Blum K, Oscar-Berman M, Stuller E et al. Downs BW, Chen AL, Chen TJ et al.
Neurogenetics and nutrigenomics of Nutrigenomic targeting of carbohydrate crav-
­neuro-nutrient therapy for Reward ing behavior: Can we manage obesity and
Deficiency Syndrome (RDS): Clinical rami- aberrant craving behaviors with neurochemi-
fications as a function of molecular neuro- cal pathway manipulation by Immunological
biological mechanisms. J Addict Res Ther Compatible Substances (nutrients) using a
3(5);2012b:139. Genetic Positioning System (GPS) Map? Med
Blum K, Sheridan PJ, Wood RC et al. The D2 Hypotheses 73(3);2009:427–34.
dopamine receptor gene as a determinant Ducci F, Goldman D. The genetic basis of
of reward deficiency syndrome. J R Soc Med addictive disorders. Psychiatr Clin North Am
89(7);1996:396–400. 35(2);2012:495–519.
Blum K, Simpatico T, Badgaiyan RD et al. Farris SP, Arasappan D, Hunicke-Smith S et al.
Coupling neurogenetics (GARS™) and a Transcriptome organization for chronic
nutrigenomic based dopaminergic agonist alcohol abuse in human brain. Mol Psychiatry
to treat Reward Deficiency Syndrome (RDS): 20(11);2014:1438–47.
Targeting polymorphic reward genes for Febo M, Blum K, Badgaiyan RD et al. Enhanced
carbohydrate addiction algorithms. J Reward functional connectivity and volume between
Defic Syndr 1(2);2015c:75–80. cognitive and reward centers of naïve rodent
Blum K, Wood RC, Braverman ER et al. The D2 brain produced by pro-dopaminergic agent
dopamine receptor gene as a predictor of KB220Z. PLoS One 12(4);2017:e0174774.
compulsive disease: Bayes’ theorem. Funct Goldman D. Candidate genes in alcoholism. Clin
Neurol 10(1);1995:37–44. Neurosci 3(2);1995:174–81.
Butler SF, McNaughton EC, Black RA. Tapentadol Guo P, Li Y, Eslamfam S et al. Discovery of novel
abuse potential: A post marketing evalua- genes mediating glucose and lipid metabo-
tion using a sample of individuals evaluated lisms. Curr Protein Pept Sci 18(6);2016:609–18.
for substance abuse treatment. Pain Med Gyollai A, Griffiths MD, Barta C et al. The genet-
16(1);2015:119–30. ics of problem and pathological gambling:
310  Customized DNA–directed precision nutrition to balance the brain reward circuitry

A systematic review. Curr Pharm Des dependence: Genetics of vulnerability to drug

20(25);2014:3993–9. dependence. Addiction 110(3);2015:530–7.
Higginson AD, McNamara JM, Houston AI. Plomin R, Owen MJ, McGuffin P. The genetic
Fatness and fitness: Exposing the logic of basis of complex human behaviors. Science
evolutionary explanations for obesity. Proc 264(5166);1994:1733–9.
Biol Sci 283(1822);2016: pii: 20152443. Polyák Á, Winkler Z, Kuti D et al. Brown adipose tis-
Hill E, Han D, Dumouchel P et al. Long term sue in obesity: Fractalkine-receptor dependent
Suboxone™ emotional reactivity as measured immune cell recruitment affects metabolic-
by automatic detection in speech. PLOS ONE related gene expression. Biochim Biophys Acta
8(7);2013:e69043. 1861(11);2016:1614–22.
Joffe B, Zimmet P. The thrifty genotype in type 2 Rana S, Kumar S, Rathore N et al. Nutrigenomics
diabetes: An unfinished symphony moving to and its impact on life style associated metabolic
its finale? Endocrine 9(2);1998:139–41. diseases. Curr Genomics 17(3);2016:261–78.
Levey DF, Le-Niculescu H, Frank J et al. Genetic Rena G, Begg F, Ross A et al. Molecular cloning,
risk prediction and neurobiological understand- genomic positioning, promoter identification,
ing of alcoholism. Transl Psychiatry 4;2014:e391. and characterization of the novel cyclic amp-
Mayer P, Höllt V. Pharmacogenetics of opioid specific phosphodiesterase PDE4A10. Mol
receptors and addiction. Pharmacogenet Pharmacol 59(5);2001:996–1011.
Genomics 16(1);2006:1–7. Thanos PK, Michaelides M, Umegaki H and
Meshkin B, Blum K. Folate nutrigenetics: A con- Volkow ND. D2R DNA transfer into the
vergence of dietary folate metabolism, folic nucleus accumbens attenuates cocaine
acid supplementation, and folate antago- self-administration in rats. Synapse
nist pharmacogenetics. Drug Metab Lett 62(7);2008:481–6.
1(1);2007:55–60. Thanos PK, Rivera SN, Weaver K, Grandy
Miller DK, Bowirrat A, Manka M et al. Acute intra- DK, Rubinstein M, Umegaki H, Wang GJ,
venous synaptamine complex variant KB220™ Hitzemann R, Volkow ND. Dopamine D2R
“normalizes” neurological dysregulation in DNA transfer in dopamine D2 receptor-defi-
patients during protracted abstinence from cient mice: Effects on ethanol drinking. Life
alcohol and opiates as observed using quan- Sci 77(2);2005:130–9.
titative electroencephalographic and genetic Thanos PK, Volkow ND, Freimuth P, Umegaki H,
analysis for reward polymorphisms: Part 1, Ikari H, Roth G, Ingram DK, Hitzemann R.
pilot study with 2 case reports. Postgrad Med Overexpression of dopamine D2 recep-
122(6);2010:188–213. tors reduces alcohol self-administration.
Moeller SJ, Parvaz MA, Shumay E et al. Gene x J Neurochem 78(5);2001:1094–103.
abstinence effects on drug cue reactivity in Xu K, Lichtermann D, Lipsky RH et al. Association
addiction: multimodal evidence. J Neurosci of specific haplotypes of D2 dopamine
33(24);2013:10027–36. receptor gene with vulnerability to heroin
Myers RD and Robinson DE. Mmu and D2 recep- dependence in 2 distinct populations. JAMA
tor antisense oligonucleotides injected in Psychiatry 61(6);2004:597–606.
nucleus accumbens suppress high alcohol Yan J, Aliev F, Webb BT et al. Using genetic
intake in genetic drinking HEP rats. Alcohol information from candidate gene and
118(2–3);1999:225–33. genome-wide association studies in risk pre-
Noble EP, Blum K, Ritchie T et al. Allelic associa- diction for alcohol dependence. Addict Biol
tion of the D2 dopamine receptor gene with 19(4);2014:708–21.
receptor-binding characteristics in alcoholism. Zhang Y, Gong J, Xie C et al. Alterations in brain
JAMA Psychiatry 48(7);1991:648–54. connectivity in three sub-regions of the
Palmer RH, Brick L, Nugent NR et al. Examining anterior cingulate cortex in heroin-dependent
the role of common genetic variants on individuals: Evidence from resting state fMRI.
alcohol, tobacco, cannabis and illicit drug Neuroscience 284;2015:998–1010.
Index

A Acute promyelocytic leukemia Altered liver enzyme activities, 5

(APL), 198 Alzheimer’s disease, 238
A1298C in exon 7 (GLU 429ALA,
Acyl-CoA thioesterase (ACOT4), Alzheimer’s Prevention Initiative,
RS1801131), 74
236 238
AAT, see Alpha-1-antitrypsin
ADA, see American Diabetes American Cancer Society (ACS),
(AAT)
Association (ADA) 50, 159
Abaloparatide, 246
Addiction Severity Index-Media American Diabetes Association
ABBV-075, 204
Version (ASI-MV) test, (ADA), 260
ABCC8 gene, 262
305 AML-MRC, 198
Aberrant DNA hypermethylation,
Adenomatous polyposis gene AML-NOS, 198
141
(APC), 52 Amyloid precursor protein (APP),
ABI SOLiD 3 platform, 152
AdoMet-dependent protein 238
Abnormal gene silencing, 53
methyltransferases, 204 Anaplastic lymphoma kinase
Acetaminophen, 68
Advanced medications (ALK), 5
Acetyl salicylic acid, see Aspirin
development, 226 Anaplastic thyroid cancer (ATC),
(acetyl salicylic acid)
a-ethinylestradiol, see Oral 10
Achilles’ heel, 52
contraceptive agents Aneuploidy, 53
ACS, see Acute coronary syndrome
Afatinib, 9 Angiogenesis, 159
(ACS)
Affymetrix, 149 ANKRD26, 206
ACS, see American Cancer Society
Aiolos (IKZF3), 183 Anthacycline, 200
(ACS)
Akt, 9 Antiangiogenic therapy, 159–161
Acute coronary syndrome (ACS),
Akt-PI3K pathways, 182 Antibiotic therapy, 287–288
227
Ala541Thr variant, 125 Anticancer drugs (colorectal
Acute myeloid leukemia (AML),
Aldehyde dehydrogenase activity cancers), 67
195–207
(ALDHhi), 141 fluoropyrimidine, 67
classification, 198–199
ALDHhi, see Aldehyde irinotecan, 67
current therapy and
dehydrogenase activity oxaliplatin, 67
experimental therapies,
(ALDHhi) Anti-CD20 monoclonal antibodies,
200–206
ALK, see Anaplastic lymphoma 274
diagnosis and treatment,
kinase (ALK) AntiEGFR monoclonal antibodies
196–198
Allograft, 207 (mAbs), 57
familial predisposition, 206
All-trans retinoic acid (ATRA), 199 Antiepileptic drugs (AEDs), 237
first models, 206–207
Alpha-1-antitrypsin (AAT), 227 AntiHIV drugs, 68
flow cytometry, 198
α 2 adrenergic-receptor blocker, Antiplatelet pharmacogenomics,
morphology, 198
264 223–224
prognostic stratification
α 2A adrenergic-receptor (α2AAR) Antiresorptive agents, 245
significance, 199
encoding gene, 264 APC membrane recruitment
therapeutic targeting of AML
α-ketoglutarate (α-KG), 202 protein 1 (AMER1), 253
stem cells, 206

311
312   Index

APL, see Acute promyelocytic Biomarker Development, 45 risk factors

leukemia (APL) Biomarker driven colon cancer age, 5
Apolipoprotein A-I, 227 therapy, implementation alcohol consumption, 5
Appropriate doses of drug, usage of, 52–63 bone density, 5
of, 226 Bisphosphonates, 245–246 BRCA1 and BRCA2
AQP2 (aquaporin2) rs296766 Blood eosinophils, 283 mutations, 5
T allele, 264 Blood-based DNA methylation breast tissue density, 5
A-RAF, 77 markers, 138 first menstrual cycle, age
Arg293X variants, 125 Blotting techniques, 148 of, 5
ARID1A mutation, 10, 152 Blum’s laboratory, 303 first pregnancy and parity,
ARID1B, 152 BMD, see Bone mineral density age of, 5
ARID2, 152 (BMD) menopause, age of, 5
Aromatase inhibitors, 165 BOADICEA, 7 oral contraceptives, usage
Array-comparative genomic Bone, 243 of, 5
hybridization (array- Bone mineral density (BMD), 243 positive family history, 5
CGH), 20 Bovine granulosa cells (GC), 153 previous atypical biopsies, 5
Arsenic trioxide (ATO), 199 BRAF, 57–59, 76–77 previous breast mass, 5
Aspirin (acetyl salicylic acid), 8 (V600e) inhibition, 76 chemotherapeutics, 5
Asthma phenotypes, 281 inhibitors, 76 diagnosis, 5
Astrocytoma, 39 PI3K/AKT pathway activation, drugs, 9
ATC, see Anaplastic thyroid cancer 76 surgical resection, 5
(ATC) RAS/RAF/MEK/ERK pathway Breast cancer 1 (BRCA1) gene
ATO, see Arsenic trioxide (ATO) inhibitors, 77 mutation, 4
ATP-binding cassette (ABC) targeted therapy, 76 Breast cancer epigenetic targets,
transporter proteins, 71 triplet therapy, 76 137–142
ABCC1/MRP1, 71 BRAF inhibitor, 45 epigenetic targets for
ABCC2/MRP2, 71 BRAF mutation, 10 prevention, 137–139
ABCG2/BCRP, 71 BRAF V600E mutation, 45 epigenetic targets for
ATRA, see All-trans retinoic acid B-RAF, 77 treatments, 139–142
(ATRA) BRAF-mutated and wild-type Bridge amplification, 150
Automated DNA sequencing tumors, differentiation, Bromodomain and extraterminal
method, 149 58 (BET) proteins, 204
Azacitidine and lenalidomide, Brain atrophy, 270 Bronchoconstriction, 280
combinations, 181 Brain gliomas, precision medicine Bronchodilators, 286
Azacitidine, 180 for Butyrate, 82
5-aza-2-deoxycytidine, 140–141 clinical scenario, 45–46
AZD9291, 9 current approaches for, 42
C
future directions, 46
incidence, 41 C-027, 282
B
potential targets, 43 C677T in exon 4 (Ala222Val,
BCL-2 inhibitor venetoclax, 204 prevalence, 39–41 rs1801133), 74
BCR-ABL fusion gene formation, 4 strategies facilitate the CA-125 glycoprotein, 158
BCRAT, 7 stratification of, 46 CAG, see Invasive coronary
Bead arrays, 149 survival and mortality, 41–42 angiography (CAG)
Beads, emulsions, amplification, treatment impact, 43–45 Calcitonin nasal spray, 246
and magnetics (BEAM), BRCA mutation analysis, 7 Calcitonin, 245
22 BRCA1 methylation role, 139 California, 149
BEAM, see Beads, emulsions, BRCA1 promoter methylation, 139 Camurati–Engelmann disease, 253
amplification, and BRCAPRO, 7 Cancer Genome Atlas Project, 152
magnetics (BEAM) Breast cancer Cancer mortality primary
Bevacizumab, 160 infiltrating ductal carcinoma, 5 cause, 17
 Index 313

Cancer risk prediction, 7 Chromosome conformation Cingulin (CGN), 153

Capecitabine, 72 capture (3C2)–based Circulating tumor cells (CTCS), 17
CapeOx, 75 methods, 84 biological relevance, 31–33
Carboplatin, 159 Chronic myeloid leukemia (CML), challenges and research
Carcinoembryonic antigen (CEA), 4 opportunities, 35–36
see TAAs expression Chronic myelomonocytic leukemia characteristics and utility of, 18
Caretaker genes, 52 (CMML), 176 clinical and research utility,
Carnitine palmitoyl transferase-1A Chronic obstructive pulmonary 20–21
(CPT1A), 141 disease (COPD) cytogenetic composition of, 20
Casein kinase1A1 (CSNK1A1), 183 approaches currently used, 281 general features of, 18
Cathepsin D, 19 biomarkers, 283–284 genotyping of, 20
Cathepsin K (CTSK) gene, 251 challenges and research health improvement of OSCC
CATSB mRNA, expression of, 62 opportunities, 284 patients, 34–35
CBFB-MYH11, 197 gene expression analysis, 288 molecular characterization and
CD4+ T cells, 280 classification, 281 cell analysis, 19–20
CD4+, 61; see also Peripheral epidemiology, 280 presence in peripheral blood, 20
T-effector cells genetic linkage, 287 RNA-based expression
CD45RO+ T cells, 61 genetics, 282–283 monitoring, 19
CD8+, 61; see also Peripheral molecular pathogenesis, selection approaches, 18–19
T-effector cells 280–281 study in precision medicine for
CDK12, 152 new approaches, 281 OSCC patients, 33–34
CDK8 amplification, 20 novel approaches, 287–288 technological development,
CDKN2A/B mutation, 10 biomarkers, 288 31–33
CEBPA, 206 epigenetic biomarkers, 288 Circulating tumor DNA (ctDNA),
Cediranib, 161 precision medicine, 284 17, 21–22
Cell surface marker-dependent classification, 285 characteristics and utility, 21–22
approach, 18 diagnosis, 285 clinical utility, 22
Cell-free DNA (cfDNA), 21 occurrence, 284–285 general features of, 21
CellSearch system (Veridex, pathogenesis, 285 molecular characterization and
New Jersey), 18 therapeutic options, 285–286 cell analysis, 21–22
Cereblon (CRBN), 183 exacerbation management, selection approaches, 21
Cetuximab, 54, 57, 59, 74 286 Cis-diamminedichloridoplatinum
CGN, see Cingulin (CGN) exacerbation (cisplatin), 10
Chemical sequencing method, 149 pharmacotherapy, Cisplatin and cyclophosphamide,
Chimeric antigen receptor (CAR)- 286–287 combination, 159
modified T cells, 204 monitoring and follow-up, Cisplatin metabolism, 10
Chloramphenicol, 68 287 c-Jun N-terminal kinase (JNK), 78
Chromatin, 64 pharmacotherapy, 286 CK19, 19
Chromosomal instability (CIN), 51, pulmonary rehabilitation, c-Kit and aflibercept/VEGF trap,
53–54 286 161
Chromosomal segregation process, treatment, 281 Classic adenomato-carcinoma
defects asthma medication, 281–282 progression, 53
chromosomal copy number, biological agents, 282 Clopidogrel, 224
changes in, 53 bronchial thermoplasty, 282 CML, see Chronic myeloid
chromosomal number, novel pharmacotherapeutic leukemia (CML)
imbalance in, 53 approaches, 282 CMML, see Chronic
loss of heterozygosity (LOH), Cilostazol, 224 myelomonocytic
high frequency of, 53 CIMP, see CpG island methylator leukemia (CMML)
subchromosomal genomic phenotype (CIMP) CNV, see Copy number variation
amplifications, 53 CIN, see Chromosomal instability (CNV)
Chromosome 18q, deletion, 55 (CIN) Cognitive-behavioral therapy, 288
314   Index

Colonoscopy, 50 CpG island methylator phenotype cell analysis, 21–22

Colorectal cancer (CRC), 9 (CIMP), 51, 86 chromosomal rearrangements, 22
approaches, 81–82 CpG islands, 154 clinical utility, 22
BRAF, 57–59 CpG sites (CpG dinucleotides), 64 molecular characterization,
cathepsin b (CATSB), focus on, CPMC, see Coriell Personalized 21–22
62–63 Medicine Collaborative use in KRAS or NRAS
epidemiology, 50 (CPMC) research study mutations, 22
epigenetics, 63 CPT1A, see Carnitine palmitoyl CTLA4 (cytotoxic T-lymphocyte
genetic and nongenetic transferase-1A (CPT1A) antigen-4) gene, 260
paradigms, 50–52 CPX-351, 201 Current clinical approaches
immune-related markers, 60–62 Craniodiaphyseal dysplasia, 253 chemotherapy, 4
interactomics, 84 CRBN, see Cereblon (CRBN) radiotherapy, 4
KRAS, 59–60 CRC, see Colorectal cancer (CRC) targeted therapy, 4
metabolomics reflects CRC epigenetics, 63 breast cancer, 4
“metabolic codes” of CRC, downregulated in hematological cancers, 4
patients, 83 hsa-mir-1, 65 lung cancer, 4
microarrays in diagnostics and hsa-mir-30a, 65 Customized addiction medicine,
treatment selection, 86 hsa-mir-126, 65 303–304
microbiome-driven hsa-mir-133b, 65 facts, 304
carcinogenesis in, 81–82 hsa-mir-143, 65 fiction, 304
next-generation analysis, 85–86 hsa-mir-145, 65 Customized DNA–directed
PIK3CA and PTEN, 56–57 hsa-mir-191, 65 precision nutrition
prognostic and predictive hsa-mir-192, 65 criteria, 300–302
biomarkers, 52–63 CRC tumorigenesis and LOH, DNA customization is a reality,
protease related interrelationship, 55 298
markers, 62–63 CRC, upregulated in importance and value,
telomere length, 63 hsa-mir-17-3p, 65 298–299
TP53, 55–56 hsa-mir-20a, 65 DNA-customized nutraceutical,
Common leukocyte antigen hsa mir-21, 65 300
CD45, 19 hsa-mir-25, 65 genetic testing, 302–303
COMPASS, see Coronary hsa-mir-31, 65 pharmacogenetic testing, 303
Obstruction Detection by hsa-mir-92, 65 genetic addiction risk score
Molecular Personalized hsa-mir-93, 65 (GARS), 303
Gene Expression hsa-mir-96, 65 genetic testing for
(COMPASS) study hsa-mir-106, 65 customization, 306–307
Complete EMT (epithelial/ hsa-mir-135b, 65 Cyclosporine,
mesenchymal+), hsa-mir-183, 65 see Immunosuppressants
see Phenotypes hsa mir-203, 65 CYP2C19*2 genotype, 224
Conventional methods, 302 Crenolanib, 201 CYP2C9, 225
COPD susceptibility genes in CRS, see Cytoreductive surgery Cytarabine, 200
smokers, 287 (CRS) Cytochrome P450 2C9 (CYP2C9),
Copy number variation (CNV), 21 CSMD3, 152 225
Core binding factor (CBF) CTC cells, monitored, 20 Cytochrome P450 enzymes (CYPs),
leukemia, 196 CTC chip in patients, 20 10
Coriell Personalized Medicine CTC protein fingerprinting, 19 Cytogenetic analysis, 174
Collaborative (CPMC) CTC-chips, 18 Cytokeratin (CK), see Epithelial cell
research study, 219 CTC-iChip, 19–20 adhesion markers
Coronary Obstruction Detection by CTCs, see Circulating tumor cells Cytokines, 61
Molecular Personalized (CTCs) Cytoplasmic TJ proteins, 153
Gene Expression CtDNA, 21; see also Circulating Cytoreductive surgery (CRS),
(COMPASS) study, 222 tumor DNA (ctDNA) 158–159
 Index 315

Cytosine–guanine (CpG) DMEs, see Drug metabolizing dUMP, see Deoxyuridine

dinucleotides, 154 enzymes (DMEs) monophosphate (dUMP)
Cytotoxic T cell-associated antigen DMRs, see Differentially Dynamin 1 (DNM1), 235
4 (CTLA-4), 61 methylated regions
Cytotoxic T lymphocyte antigen 4 (DMRs)
E
(CTLA4), 282 DNA hypermethylation, 64
Cytotoxic TILs, 54 DNA methylation (DNAm) EAD, see Entinostat/ATRA/
age, 138 doxorubicin (EAD)
DNA methylation role, 63–64, 140 EGAPP™, see Evaluation of
D
DNA methylome analysis, 138 Genomic Applications in
Dabrafenib, 76 DNA methyltransferases (DNMTs), Practice and Prevention
dACC, see Dorsal anterior cingulate 64, 154 (EGAPP™)
cortex (dACC) DNA spots, 154 EGFR mutation, 10
Danish study, 126 DNMT inhibitors, 140 EGFR, see Epidermal growth factor
DAT1R 9R-allele, 303 DNMTs, see DNA receptor (EGFR)
DDX41, 206 methyltransferases 18q loss of heterozygosity (LOH),
Decitabine, 180, 203 (DNMTs) 54–55
Defeat Glioblastoma Research Docetaxel (Taxotere®), 141, 159 Electronic microarrays, 149
Collaborative, 44 Dominantly Inherited Alzheimer Electrophoretic-based techniques,
Denaturing high-performance Network (DIAN), 238 147
liquid chromatography Dorsal anterior cingulate cortex Electrospray ionization-MS, 156
(DHPLC), 150 (dACC), 304 Elevated microsatellite alterations
Dendritic cell–based cancer Double-stranded DNA breaks at selected tetranucleotide
immunotherapy, 61 (DSBs), 139 repeats (EMAST), 54
Denosumab, 246 DPD, see Dihydropyrimidine ELN, see European Leukemia Net
Deoxythymidine monophosphate dehydrogenase (DPD) (ELN)
(dTMP), 72 DPD testing, clinical uptake, 72 EMAST, see Elevated microsatellite
Deoxyuridine monophosphate DPYD gene, 71 alterations at selected
(dUMP), 72 DPYD*2A (rs3918290) tetranucleotide repeats
Desmosine (DES), 285 polymorphism, 72 (EMAST)
DHPLC, see Denaturing high- DR15-DQ, 260 EMT, see Epithelial to mesenchymal
performance liquid DR15-DQ6 haplotype, 260 transition (EMT)
chromatography DR3-DQ2, 260 Endometrial adenocarcinoma, 154
(DHPLC) DR3-DQ2/ DR4-DQ8 haplotypes, Endometriosis, 154
Diabetes mellitus (DM), 259–265 260 ENO marker, 284
advantages and limitations, DR4-DQ8, 260 Entinostat/ATRA/doxorubicin
264–265 Droplet-based systems, see Digital (EAD), 141
application in type 1 dm, PCR EORTC 40993, 58
260–262 Drug Discovery, 45 Eosinophils marker, 281, 284
application in type 2 dm, Drug metabolizing enzymes EpCAM, see Epithelial cell
262–264 (DMEs), 69 adhesion markers
candidate genes, 263 Drugs targeting oncogenes Ependymoma, 39
classification, 261–262 ALK, 9 EPIC, see European Prospective
gestational diabetes mellitus, 264 BRAF, 9 Investigation into Cancer
Diamond–Forrester (DF) risk score, CKI, 9 and Nutrition (EPIC)
222 HER2, 9 Epidermal growth factor receptor
Differentially methylated regions KRAS, 9 (EGFR) blockers, 5, 9,
(DMRs), 138 DSBs, see Double-stranded DNA 44, 59
Digital PCR, 22 breaks (DSBs) Epigenetic (DNA methylation
Dihydropyrimidine dehydrogenase dTMP, see Deoxythymidine analysis), 154
(DPD), 71 monophosphate (dTMP) Epigenetic clock, 138
316   Index

Epigenomic analyses, 12 FAM20C gene, 253 Forkhead/winged helix gene

Epilepsy, 237–238 Familial adenomatous polyposis (FOXP3), 270
EPISPOT (Epithelial ImmunoSpot), (FAP), 8 Fracture Risk Assessment Tool
19 FAP, see Familial adenomatous (FRAX), 245
Epithelial (epithelial+/ polyposis (FAP) 5-FU and dihydropyrimidine
mesenchymal), FAT3, 152 dehydrogenase (DPD),
see Phenotypes FDA-approved pharmacogenomic 71–72
Epithelial cell adhesion markers, 18 biomarkers, 10 5-FU and MTHFR, 74
Epithelial to mesenchymal Fecal occult blood test (FOBT), 50 5-FU and thymidylate synthase
transition (EMT), 19, 31 Fetal liver ring finger (FLRF), 184 (TS), 72–74
ER, 5 Fibrinogen gamma chain, 227 5-FU metabolism, 71
ER+breast cancers, 140 Finnish family–based case-control 5-FU PGX, 71
ERBB2-4 mutation, 10 study, 124
ERK, see Proto-oncogenes Flow cytometry (FCM)
G
Erlotinib, 9 immunophenotyping, 177
Erythropoietic stimulating agents FLRF, see Fetal liver ring finger G protein subunit alpha O1
(ESAs), 180 (FLRF) (GNAO1), 235
ESAs, see Erythropoietic Fluidic flow cell design, 150 G84E mutation, 125
stimulating agents (ESAs) Fluorescence in situ hybridization GABRA6, 152
Estrogen, 246 (FISH), 20 Gadolinium contrast enhancement,
Estrogen deficiency, 247 Fluorescence-activated cell sorting 271
Estrogen receptor (ER) methylation, (FACS), 12 Gallbladder cancer, 10
role of, 139–140 Fluorescent dyes, 148 Gamma-aminobutyric acid type A
Estrogen receptor antagonists, Fluoropyrimidine, 67, 71 receptor alpha1 subunit
164–165 5-fluorouracil infusion leucovorin, (GABRA1), 235
ETV6, 206 and irinotecan Gamma-aminobutyric acid type
European Leukemia Net (ELN), 200 (FOLFIRI), 67 a receptor β3 gene
European Prospective Investigation Fluorouracil, leucovorin, and (GABRB3), 235
into Cancer and Nutrition oxaliplatin (FOLFOX), 67 GARS, see Genetic Addiction Risk
(EPIC), 138 FOBT, see Fecal occult blood test Score (GARS)
Evaluation of Genomic Applications (FOBT) GATA2, 206
in Practice and Prevention Folate antagonists, 162 Gatekeeper gene APC, 84; see also
(EGAPP™), 70 Folate receptor inhibitors, 162–163 Adenomatous polyposis
Everolimus, 79 Folate receptors (FR-α), 162 gene (apc)
Exome, 12 Folate-conjugated therapeutic GCS, see Glucosylceramide
Exposome, 52 agents, 164 synthase (GCS)
Exposomics, 7 FOLFIRI; see also 5-Fluorouracil GDM, see Gestational diabetes
Extracellular signal-regulated infusion leucovorin, and mellitus (GDM)
kinase (ERK1/2), 78 irinotecan (FOLFIRI) Gefitinib, 22
Extracellular β-amyloid (Aβ) fluorouracil, 71 Gemcitabine, 141, 159
protein, 238 irinotecan, 71 Gemtuzumab ozogamicin, 204
leucovorin, 71 Gene chips, 149
FOLFOX; see also Fluorouracil, Gene therapy, 43
F
leucovorin, and Genetic addiction risk score
FAB-M0, 198 oxaliplatin (FOLFOX) (GARS), 303
FAB-M3 (promyelocytic leukemia), infusional fluorouracil, 71 benefits, 305
198 leucovorin, 71 GeneChip p53 assay, 86
FAB-M7, 198 oxaliplatin, 71 General cancer information, 6
FACS, see Fluorescence-activated FOLFOX4, 58 Genetic Addiction Risk Score
cell sorting (FACS) Food and Drug Administration (GARS), 303
Faggot cells, 198 (FDA), 18 Genetic profile, 5
 Index 317

Genetic risk score (GRS), 80–81, 129 GRS, see Genetic risk score (GRS) Hereditary nonpolyposis colorectal
Genome and genetic GSDMB, 283 cancer (HNPCC), 52
characterization tools and Guadecitibine (SGI-110), Hereditary Prostate Cancer 1
techniques, 146 180, 203 (HCP1), 124
Genomic-based gene expression Gut microbial dysbiosis, 82 Herringbone, 19
score (ges), 221–222 GWAS, see Genome-wide Heteroduplex analysis (HDA), 147
studies, 222–223 association studies Heterozygous UGT1A1*28
454 Genome Sequencer FLX (GWAS) (TA7/6), 69
Titanium System, 150 GWAS to GRS, 80–81 HGP, see Human Genome Project
Genome-wide association studies Gynecologic oncology group study (HGP)
(GWAS), 66, 80–81, 218 (GOG-218), 160 High-grade serous ovarian
126, 219 cancer, 152
Genomic-based gene expression High-throughput genomic
H
score (GES), 221–222 technologies, 85
Genomics, 146 H3K18 histones, hypoacetylation, 65 High-throughput sequencing
Genotypes, 72 H3K27 trimethylation technologies, 150
2r/2r, 72 (H3K27me3), 65 Hip fractures, 244
2r/3r, 72 H4K12 histones, hypoacetylation, 65 HIPEC, see Hyperthermic
3r/3r, 72 HAD, see Heteroduplex analysis intraperitoneal
Genotyping or mutation analysis, (HDA) chemotherapy (HIPEC)
147–148 Hamartomatous polyposis Histamine and serotonin
GES, see Genomic-based gene syndromes, 52 antagonists, 282
expression score (GES) Haplotype B3, 72 Histone acetyltransferases
Gestational diabetes mellitus HATs, see Histone (HATs), 64
(GDM), 264 acetyltransferases Histone deacetylase (HDAC), 281
Gilteritinib, 201 (HATs) Histone deacetylase inhibitors
Glioblastoma, 39 HBCTC-chip, 19 (HDACIs), 204
subtypes HCP, see Hereditary Prostate Histone deacetylases (HDACS), 64
classical, 44 Cancer 1 (HCP1) class I
mesenchymal, 44 HCT116, 83 HDAC1, 64
neural, 44 HCT15, 83 HDAC2, 64
proneural, 44 HDAC inhibitors, 140 HDAC3, 64
Glioma therapy, 43 HDAC3 overexpression, 65 HDAC8, 64
Glu265X mutation, 124 HDAC3 silencing, 65 class II
Glucosylceramide synthase HDACIs, see Histone deacetylase HDAC4, 64
(GCS), 142 inhibitors (HDACIs) HDAC5, 64
Glucuronidation, 68 HDACs, see Histone deacetylases HDAC6, 64
Glutamate ionotropic receptor (HDACs) HDAC7, 64
NMDA type subunit 2A HDMs, see Histone demethylases HDAC9, 64
(GRIN2A), 235 (HDMs) HDAC10, 64
Glutamate, 202 Head and neck cancers, class III
Glutamic amino acid, 57 29; see also Oral SIRT1, 64
Glutamine, 202 squamous cell SIRT2, 64
Glycomics, 83–84 carcinoma (OSCC) SIRT3, 64
GOG-218, see Gynecologic Helicos HeliScope, 150–151 SIRT4, 64
oncology group study 218 Hematologic malignancies, 5 SIRT5, 64
(GOG-218) Hepatocyte nuclear factor 1-α SIRT6, 64
Grade II astrocytoma (low grade) (HNF1-α), 264 SIRT7, 64
patients, 40 HER2 inhibitors, 5 class IV
Grade IV astrocytoma (glioblastoma) Herceptin (trastuzumab), 9; see also HDAC11, 64
patients, 40 Breast cancer, drugs Histone demethylases (HDMs), 65
318   Index

Histone H3 lysine 4 [H3K4], 65 Hsa-miR-92a-1, 65 IL1RL1/IL18RL1 on 2q12, 283

Histone H3 lysine 9 [H3K9], 65 Hsa-miR-106a, 65 IL-2, 61
Histone H3 lysine 27 [H3K27], 65 Hsa-miR-221, 65 IL33 on chromosome 9p24, 283
Histone H3 lysine 36 [H3K36], 65 HSD17B4 methylation, 141 IL-6 induction, 183
Histone H3 lysine 79 [H3K79], 65 Human epidermal growth Illumina Human Methylation 27
Histone methylome (HMT), 204 factor receptor (HER) array, 86
Histone methyltransferases antagonists, 164 Illumina’s genome analyzer, 150
(HMTs), 65 Human Genome Project (2003), 66 Imatinib, 5, 9
Histone modifications, 64–65 Human Genome Project (HGP), 80 Immune-related markers, CRC,
posttranslational modifications Human mammary epithelial cells 60–62
exposure, 64 (HMECs), 138 innate immunity, 61
acetylation, 64 Hybrid fusion genes, 197 natural killer [NK] cells, 61
deamination, 64 Hyperacetylation, 64 tumor infiltrating
methylation, 64 Hypermethylation, 53, 55 macrophages [TIM], 61
phosphorylation, 64 Hyperpolarization activated unconventional T
ribosylation, 64 cyclic nucleotide gated lymphocytes, 61
acetylation/deacetylation, 64 potassium channel 1 adoptive immunity, 61
lysine amino acids, acetylation (HCN1), 235 intratumoural memory
reactions of, 64 Hyperthermic intraperitoneal CD8T cells, 61
methylation/demethylation, 64 chemotherapy (HIPEC), tumor in filtrating CD45RO+
HKN-RAS mutation, 10 159 memory T lymphocytes, 61
HLA-DQB1 on 6p21, 283 Hypoacetylation, 64 Immunogenic CRC, 61
HLA-DRA on 6p21, 283 Hypomethylating agents (HMAs), Immunoglobulin gamma heavy
HMAs, see Hypomethylating agents 202 chain, 227
(HMAs) Immunohistochemical (IHC), 56
HMECs, see Human mammary Immunohistochemical pattern of
I
epithelial cells (HMECs) p53, 180
HMT, see Histone methylome IBD, see Inflammatory bowel Immunosuppressants, 68
(HMT) disease (IBD) Immunotherapy, 43
HMTs, see Histone ICGC, see International Cancer Impaired bone resorption, 251–252
methyltransferases Genome Consortium Improved and safer drugs
(HMTs) (ICGC) development, 226
HNF1-α I27L (rs1169288) ICON-7, see International Increased bone formation, 253
polymorphism, 264 collaboration on ovarian Increased bone turnover, 253–254
HNPCC, see Hereditary neoplasms 7 (ICON-7) 7+3 induction regimen, 200
nonpolyposis colorectal ICUS, see Idiopathic cytopenias Inflammatory bowel disease
cancer (HNPCC) of undetermined (IBD), 51
Homozygous UGT1A1*28 genotype significance (ICUS) Inhibitor triple complex, 74
(TA7/7), 69 IDH2 inhibitor, 182 5-fluoro-2-deoxyuridine-
Hormone receptor (ER and HER2) Idiopathic cytopenias of 5-monophosphate
profile, 5 undetermined (5-FDUMP), 74
Housekeeping gene, 62; see also significance (ICUS), 180 5,10 methylenetetrahydrofolate, 74
Lysosomal cysteine IDS, see Interval debulking surgery TS, 74
proteases cathepsin B (IDS) INS/DEL polymorphism, 73
(CATSB) IFN-α, 61 DEL/DEL (homozygous for
HOXB13 gene, 125 IgE marker, 284 deletion of the 6 BP), 73
H-ras gene, 59; see also KRAS in Ikaros (IKZF1), 183 INS/DEL (heterozygous), 73
CRC IL-10 (AM0010), 61 INS/INS (homozygous for
Hsa-miR-17-3p, 65 IL13 on 5q31, 283 insertion of 6 BP), 73
Hsa-miR-21, 65 IL-13, 280 Integrative personal omics profiles
Hsa-miR-29a, 65 IL-16, 280 (iPOPs), 43
 Index 319

Integrin subunit alpha 6 JSCCR, see Japanese Society for Loci associated with forced vital
(ITGA6), 154 Cancer of the Colon and capacity, 283
Interferon regulatory factor 5 Rectum (JSCCR) BMP6, 283
(IRF5), 270, 272 Juvenile osteoporosis, 254 EFEMP1, 283
Intermediate EMT (epithelial+/ KCNJ2, 283
mesenchymal+), MIR129-2-HSD17B12, 283
K
see Phenotypes PRDM11, 283
International Cancer Genome Kadcyla (ado-trastuzumab WWOX, 283
Consortium (ICGC), 85 emtansine), 9; see also Logarithm of odds score (LOD),
International collaboration on Breast cancer, drugs 124
ovarian neoplasms 7 Karyotype analysis, 196 LOH, see Loss of heterozygosity
(ICON-7), 160 KB220z variant, 306 (LOH)
International Prognostic Scoring KB220Z™, 304 Loss of heterozygosity (LOH), 21,
System (IPSS), 176 KCNQ (Kv7), 282 54
International working group on Kirsten Ras In Colorectal Cancer clinical manifestations, 54
constitutional disorders Collaborative Group, 59 high-frequency, 54
of bone, 254 Knudson’s “two-hit” hypothesis, 55 Low malignancy potential (LMP)
Interval debulking surgery (IDS), KRAS, 74–77 tumors, 161
158–159 KRAS in CRC, 59–60 Low- and middle-income countries
Intracellular proteolysis, 62 404 sporadic and 94 hereditary (LMICs), 50
Invasive coronary angiography CRC patients, study, 60 Lung cancer
(CAG), 218 codon 13 mutations, 60 non-small cell, 5
Ion semiconductor sequencing, functional genes, 59 small cell, 5
151 in multivariate analysis, 60 treatment
Ion Torrent sequencing, 151 in rascal II study, 60 cryoablation, 5
IPSS, see International Prognostic PETTAC8 trial study, 60 radiofrequency ablation
Scoring System (IPSS) PG13D mutations, 60 (RFA), 5
Iressa (gefitinib), 9 primary tumors assessement, 60 radiotherapy, 5
Irinotecan and Ugt1a1, 68–71 KRAS exon 2, 75 surgical resection, 5
Irinotecan-based chemotherapy, K-ras gene, 59; see also KRAS in Lymph glands in the neck,
57 CRC metastasis, 30
Irinotecan Pgx (Irinogenetics), 54, Lymphoid-specific phosphatase
67–70 (LYP), 262
L
Irinotecan-related toxicities, 68 Lymphoma, 5; see also Hematologic
diarrhea, 68 Laser capture microdissection, 12 malignancies
myelosuppression, 68 Lenalidomide, 180 Lynch syndrome, 52
ISET system, 19 Leucovorin, 72 Lynx therapeutics, 150
Isodesmosine, 285 Leukemia, 5; see also Hematologic Lysosomal cysteine proteases
Isolation method, 18 malignancies cathepsin B (CATSB), 62
ITGA6, see Integrin subunit alpha 6 Leukemia stem cells (LSCs), 206
(ITGA6) Lifestyle modification strategy, 273
M
IVS7delTTA homozygous Li–Fraumeni syndrome (LFS), 55
genotype, 125 Light Cycler™, 148 Major histocompatibility complex
LINE-1/short interspersed (MHC), 61
nucleotide elements, 64 Major histocompatibility complex
J
Lipoprotein lipase (LPL) gene, 264 class I (MHC-I), 62
Jak-STAT pathways, 182 Liquid biopsies, 17, 29–36 Mammalian target of rapamycin
Japanese Society for Cancer of Liquid biopsy analysis, 86 (mTOR), 79–80
the Colon and Rectum LMICs, see Low- and middle- MAPK, see Mitogen-activated
(JSCCR), 70 income countries protein kinase (MAPK)
JARID1C, 152 (LMICs) Marker-independent approach, 18
320   Index

Markers for type 2 inflammation in 5,10-methylenetetrahydrofolate MLH1 gene, 53

asthma, 283 reductase (MTHFR), 74 MLL, 152
Massively parallel sequencing 5-methyltetrahydrofolate MLL2, 152
(NGS), 11 (CH3THF), 74 MLL3, 152
Matrix-assisted laser desorption/ MHC, see Major histocompatibility Modified mRNA (modRNA)
ionization (MALDI)-MS, complex (MHC) technologies, 13
156 MHC-I, see Major Molecular endotyping, 284
Maxam–Gilbert sequencing histocompatibility Monoclonal antibodies, 67, 274
method, 149 complex class I (MHC-I) bevacizumab, 67
mCRC, see Metastatic colorectal MI, see Microsatellite instability cetuximab, 67
cancer (mCRC) (MI) panitumumab, 67
MDCT, see Multidetector cardiac Microarrays, 148–149 Morphine, 68
computed tomography Microbiome (“second genome”), 81 MSH2 gene, 53
(MDCT) Microbiome-driven carcinogenesis MSH6 gene, 53
MDM–TP53 inhibitor, 162 in colorectal cancer MSI, see Microsatellite instability
MDR, see Multidrug resistance (metagenomics), 81–82 (MSI)
(MDR) Microdroplet technologies, 12 MSI-high (MSI-H), 54
MDS, see Myelodysplastic Microelectronic cartridges, 149 MSI-low (MSI-L) CRCs, 54
syndromes (MDS) Microfluidic platforms, see Digital MSI-low (MSI-L), 54
MDS genes, 178 PCR MSR1 gene, 125
MDS prognostic scoring system, Microfluidics-based cell MSS phenotype, 54
176 sorting, 12 MTHFR, see
Meayamycin B, 204 MicroRNA (miRNA) therapy, 43 5,10-Methylenetetrahy­
Megakaryopoiesis regulation, 183 MicroRNAs (miRNAs), 65–66 drofolate reductase
MEK, 78–79 Microsatellite instability (MI), (MTHFR)
Melanomaassociated antigen 53–54 mTOR, 9
gene (MAGE), see TAAs Microsatellite instability mTOR, see Mammalian target of
expression (MSI), 21 rapamycin (mTOR)
Meningeal inflammation, 270 Microsatellite stable (MSS) CRCs, MUC1, 19
Mepolizumab, 282 54 Mucin 1 (MUC1), see TAAs
Mesenchymal–epithelial transition Micrornas (miRNAS), 65–66 expression
(MET), 31 application of, 66 Mucinous adenocarcinoma, 57
MET, see Mesenchymal–epithelial circulating, 65 Multidetector cardiac computed
transition (MET) drug resistance or sensitivity, tomography (MDCT), 218
Metastasis genes, 141 role in, 66 Multidrug resistance (MDR), 141
PDGF-A, 141 dysregulation, 65 Multiferon, 61; see also IFN-α
SERPINB2, 141 let-7 family, 66 Multiple lysine sites, 65
TIMP-1, 141 miR-15a, 155 histone h3 lysine 4 [h3k4], 65
Met1Ile mutation, 124 miR-16, 155 histone h3 lysine 9 [h3k9], 65
Metabolic biomarkers, 83 miRNA profiles, 65 histone h3 lysine 27 [h3k27], 65
Metabolomics of colon cancer miRNA-143, 65 histone h3 lysine 36 [h3k36], 65
tissues, 83 miRNA-145, 65 histone h3 lysine 79 [h3k79], 65
Metagenomics, 81 miRNAs, see MicroRNAs Multiple sclerosis, 269–275
Metastasis (M), see TNM (miRNAs) age considerations, 270
Metastatic colorectal cancer Mitogen-activated protein kinase challenges and research
(mCRC), 57 (MAPK), 78 opportunities, 274–275
Metformin, 263 Mitomycin C, 54 diagnosis, 270–271
Methotrexate, 72 Mixed lineage leukemia gene, 198 biomarkers, 271
5,10-methylenetetrahydrofolate Mixture of barcodes tumors diagnostic techniques, 271
(CH2THF), 74 PRISM, 12 doubtful genes, 270
 Index 321

established biomarkers, 272 Nanoparticles loaded with NIDA, see National Institute on
ethnic communities, 270 low-dose DAC Drug Abuse (NIDA)
ethnic considerations, 270 (NPDAC), 141 Nine phase III trials, 59
gender considerations, 270 National Institute for Health and Nintedanib, 161
incidence, 270 Care Excellence (NICE) NMD, see Nonsense-mediated
genetics, 270 guidelines, 271 decay (NMD)
pathological basis, 270 National Institute on Alcohol Nonalcoholic fatty liver disease
general consideration, 273–274 Abuse and Alcoholism (NAFLD), 51
lifestyle modification strategy, (NIAAA), 303 Non-insulin-dependent diabetes
273 National Institute on Drug Abuse mellitus (NIDDM), 264
precision medicine approach, (NIDA), 303 Nonsense-mediated decay (NMD),
273 National Institutes of Health 153
therapeutic agents, 274 (NIH), 259 Non-small cell lung cancer
treatment options, 271–272 NCDS (NSCLC), 5
Mutanomes, 12–13 current diagnostic and large cell carcinoma, 5
Mutations predicted and verified by therapeutic approaches, lung adenocarcinoma, 5
various studies, 152 217–218 squamous cell carcinoma, 5
MYC, see Proto-oncogenes disadvantages Nonsynonymous variants, 74
Myelodysplastic syndromes (MDS) environmental and societal NPDAC, see Nanoparticles loaded
classification and criteria influences, 220 with low-dose DAC
(WHO), 177–178 genetic risk factors, 219–220 (NPDAC)
classification, 176–177 lifestyle risk factors, 220 NPDOX, see Nanoparticles loaded
concept of personalized, personalized medicine in CAD, with doxorubicin
184–185 220–221 (NPDOX)
cytological abnormalities, 176 cost concerns, 227 N-ras gene, 59; see also KRAS in CRC
diagnosis and treatment, 174 health care providers, Nrdp, see Neuregulin receptor
frequency and impact of chosen education of, 227 degradation protein-1
genetic mutations, 179 limited alternative therapy, (Nrdp1)
histopathology, 176 227 NSCLC, see Non-small cell lung
molecular and cellular challenges, 228–229 cancer (NSCLC)
mechanisms, 182–184 differential response to drug NTPs, see Nucleoside triphosphates
morphological evaluation, therapy, 221 (NTPs)
174–176 Neoadjuvant chemotherapy (NAC), Nucleoside triphosphates (NTPs),
somatic mutations, 178–179 159 150
therapies, 180–182 Neural subgroup, 44 Nucleosomes, 64
Myeloma, 5; see also Hematologic Neuregulin receptor degradation Number of cases and deaths
malignancies protein-1 (Nrdp1), 184 reported to WHO (2008
Neuroadaptogen-amino-acid and 2012), 122
therapy (NAAT), 307 Nutlin-3 alpha, 162
N
Next generation sequencing Nutrigenomic DNA-customization
NAAT, see Neuroadaptogen- (NGS), 7 process, 300
amino-acid therapy NF1, 152
(NAAT) NF1 tumor suppressor gene, 44
O
NAC, see Neoadjuvant NGS, see Massively parallel
chemotherapy (NAC) sequencing (NGS) Occludin (OCLN), 153
NAFLD, see Nonalcoholic fatty NGS, see Next generation OCLN, see Occludin (OCLN)
liver disease (NAFLD) sequencing (NGS) OCT-1, 263
Nanoparticles loaded with NIAAA, see National Institute Oligodendroglioma, 39
doxorubicin (NPDOX), on Alcohol Abuse and Oligodendroglioma grade II
141 Alcoholism (NIAAA) patients, 41
322   Index

Oligonucleotide ligation assay, 147 Osteoclasts, 246 Overall cost of health care,
Oligonucleotide primers, 147 Osteogenesis imperfect, 254 reduction, 226–227
Omalizumab, 281 Osteopathia striata, 253 Overall survival (OS), 20
Oncology Osteoporosis, 243–255 Oxaliplatin, 67
cancer, genetic mechanism of, 3 diagnosis, 245 Oxaliplatin-based chemotherapy,
current clinical approaches, drugs, 246 57
4–5 emerging treatments, 246 Oxygen therapy, 286
personalized oncology, 5–7 epidemiology, 244
cancer risk, predicting, 7–8 epigenetic mechanisms, 248
P
choosing personalized genetics, 247
treatment, 9 lifestyle, environmental, P16INK4a tumor suppressor gene,
personalized prophylactic and genetic factors 138
treatment, 8–9 importance, 247 P2Y12 receptor blockade, 224
epidemiology of, 3 morbidity and mortality rate, p53, 77
molecular features, 9 244 p53, see Tumor suppressor genes
personal metabolic features, prevention and treatment, (TSGs)
9–10 245–246 Paclitaxel (Taxol®), 159
predicting drug sensitivity Osteoporosis and bone diseases, Paget’s disease of bone (PDB), 253
and resistance, 10 243–255 Pan-cytokeratin, 19
precision medicine, future of, 10 Osteoporosis-pseudoglioma Panitumumab, 54, 57–58, 75
barcodes tumors prism, syndrome, 254 Parkinson’s disease, 238–239
mixture of, 12 Ovarian carcinoma, 145–166 PARP, see Poly(ADP-ribose)
next-generation sequencing, advancement of cutting edge polymerase (PARP)
11 technology, 146 PARP inhibitors, 139, 162–163
patient-derived tumor genomics, 146 Passenger mutations, 178
xenograft, 12 genotyping/mutation Patient-derived tumor xenograft, 12
mutanomes, 12–13 analysis, 147–148 Patient-derived xenograft (PDX), 12
single-cell genome microarrays, 148–149 Pazopanib, 161
sequencing, 11–12 semiquantitative approach, PBRM1 mutation, 10
One phase II trial, 59 148 PCM, see Personalized cancer
OPCML gene, 152 traditional methods medicine (PCM)
OPRT, see Orotate phosphoribosyl­ and next-generation PCR, see Polymerase chain reaction
transferase (OPRT) sequencing, 149–150 (PCR)
Optic coherent tomography (OCT), chemotherapy, 159 PCR amplification, 150
271 current scenario, 158 PD-1/PD-L1 interaction, 61
Oral cancer, 29 drugs targeting mutations, PDE4DIP phosphodiesterase
diagnosis, 30 162–165 4D interacting protein
etiology, 29 epigenetic alterations, 154–155 (PDE4DIP), 236
prognosis, 30 epigenetic changes, exploration PDGFRA, see Platelet derived
risk factors, 29 of, 156 growth factor receptor
treatment, 30 epigenetic role, 155 alpha (PDGFRA)
Oral contraceptive agents, 68 gene expression, exploration PD-L1, 61
Oral glucose tolerance test (OGTT), of, 155 PDX, see Patient-derived xenograft
263 micro-RNA expression, (PDX)
Oral squamous cell carcinoma 155–156 Pegylated liposomal doxorubicin
(OSCC), 29–36 miRNA expression and their (PLD), 159
ORMDL3, 283 role, exploration of, 157 PEGylated liposomal doxorubicin
Orotate phosphoribosyltransferase precision medicine, 159 (PLD), 164
(OPRT), 73 type I, 161 Pembrolizumab (Keytruda), 61
OSCC, see Oral squamous cell type II, 161 Pemetrexed, 162
carcinoma (OSCC) OVCAR3, 83 Peptidase, see Protease
 Index 323

Peptide hydrolase, see Protease PI3K/Akt/mTOR pathway, 79 PREDICT, 222–223

Periostin marker, 283–284 PI3Kgamma inhibitors, 281–282 Prednisone, 287
Peripheral T-effector cells, 61 PICK3CA gene, 56 Primary debulking surgery, 159
Perjeta (pertuzumab), 9; see also PIK3CA and PTEN in CRC, 56–57 Primary miRNAs, 65
Breast cancer, drugs PIK3CA mutation, 10 PRMT1, see Protein arginine
Peroxisome proliferator-activated Pilocytic astrocytoma, 40 N-methyltransferase 1
receptor gamma PIN, see Prostatic intraepithelial (PRMT1)
(PPARγ2), 260 neoplasia (PIN) Proapoptotic genes, 141
Personalized cancer medicine Plant-based therapy, 288 BAD, 141
(PCM), 85 Plasma DNA, 22 CASP9, 141
Personalized oncology, 5 SNVS detection, 21 COL18A1, 141
environmental factors, 5 whole-exome analysis of, 22 Probes, 148
single nucleotide Plasma miRNAs, 65 Proficient-MMR tumors, 54
polymorphisms (SNPs), 5 Plasma, 21 Programmed cell death 1 (PD-1)
Pertuzumab (Perjeta), 5; see also Platelet derived growth factor receptor, 61
HER2 inhibitors receptor alpha Programmed cell death receptor
Pertuzumab, 164 (PDGFRA), 44 1(PD-1), 61
PETTAC-3, 58 Platinum-based drugs, 139 Progression-free survival (PFS), 20
Peucedani Radix, 288 PLD, see Pegylated liposomal Prophylactic colectomy, 9
PGx for somatic mutations in doxorubicin (PLD) Prophylactic mastectomy, 9
anticancer agents, 74 PMS2 gene, 53 Pro-protein convertase subtilisin/
PGX, see Pharmacogenetics (PGX) Point mutation, 55 kexin type 9 (PCSK9), 219
Pharmacogenetics (PGX), 66–68, Polo-like kinase pathways, 182 Prostaglandin D(2), 281
223 Poly(ADP ribose) polymerase Prostate cancer, 121–131
Pharmacogenomics, 223 (PARP) inhibitors, 162 common and low-penetrance
Pharmacologic agents, 245 Poly(ADP-ribosylation) of histones, risk, 126–129
Pharmacogenetics (PGX), 66–68 162 genetic screening and
cancer susceptibility genes Polyacrylamide gel electrophoresis, personalized medicine,
identification, 67 149 124
genetic variation and adrs Polymerase chain reaction (PCR), prostate-specific antigen (PSA)
associations, 66 147 controversy, 123–124
modifiable factors, 67 Polymorphic markers, 124 rare high penetrance prostate
nonmodifiable genetic factors, Polymorphisms in IL2RA (CD25), cancer susceptibility
67 262 genes, 124–126
predictive biomarkers, 67 Polyposis syndromes, 52 SNPS clinical application,
toxicity biomarkers Polyps, 154 129–130
identification, 67 Positron emission tomography, 239 SNPS environmental effects, 130
Pharmacotherapy, 286 Postgenomic era, 66 Prostate cancer prevention trial,
Phenotypes, 19 Potassium channel activators, 282 129
Phosphatidylinositol-3-kinase Potassium sodium-activated Prostate-specific membrane antigen
(PI3K), 56 channel subfamily T (PSMA) markers, 19
Phosphatidylinositol-4,5- member 1 (KCNT1), 235 Prostatic intraepithelial neoplasia
bisphosphate 3-kinase, 56 PPAR-gamma, 262 (PIN), 125
Phosphodiesterase-4 inhibitors, 282 Prasugrel, 224 Protease, 62
Phosphoinositide 3-kinase, 182 Precision medicine approach, 273 Protease related markers, CRC,
Phosphoinositide 3-kinase (PI3K)/ Precision Medicine Initiative 62–63
AKT pathway, 76 (PMI), 259 Protein arginine
Phosphoinositide 3-kinase gamma Precision-guided immunotherapy, N-methyltransferase 1
(PI3Kgamma), 280 282 (PRMT1), 204
PI3K, see Phosphatidylinositol-3- Precocious screening for disease, 226 Protein phosphatase and tensin
kinase (PI3K) Precursor miRNAs, 65 homolog (PTEN), 57
324   Index

Protein tyrosine phosphatase, a Real-time polymerase chain RUNX1-RUNX1T1, 197

nonreceptor type 22 reaction (PCR), 148 RUNX3 hypermethylation, 140
(PTPN22) gene, 262 Receiver-operating characteristic
Proteinase, see Protease (ROC) curve, 222
S
Proteogenomic characterization, Receptor activator of nuclear factor
43, 46 kappa-B ligand (RANKL) S447S genotype, 264
Proteolytic activity, 62 antibody, 245 S447X variant, 264
Proteomics, 156–158, 227 Recessive genes, 55; see also Tumor sACC, see Subcallosal anterior
limitations of, 228 suppressor genes (TSGS) cingulate cortex (sACC)
proteomic studies, 227–228 Recombinant human parathyroid SAHA, see Suberoyl anilide
Proteomics-based biomarker, hormone analogues, 246 hydroxamic acid (SAHA)
Current progress, 158 Recommended Dietary Allowance SAKK, 58
Proto-oncogenes, 4 (RDA) guidelines, 299 SAM, see Sentrix Array Matrix
PTEN, see Protein phosphatase and REDUCE study, 126 (SAM)
tensin homolog (PTEN) REDUCE, see Reduction by Sanger sequencing method, 11, 149
PTEN, see Tumor suppressor genes Dutasteride of Prostate Santa Clara, 149
(TSGs) Cancer Events (REDUCE) SCFA, see Short-chain fatty acid
Pulmonary rehabilitation, 286 Reduction by Dutasteride of (SCFA)
Pycnodysostosis, 251 Prostate Cancer Events Sclerostin, 246
Pyrosequencing, 150 (REDUCE), 126 Sclerostin-related disorder, 253
Regional lymph nodes (N), see TNM ScreenCell approach, 19
Regorafenib, 78 Scriptaid, 140
Q
Rehabilitation with music therapy, scRNA-seq, 12
Quantitative computed tomography 288 Secondary AML, 201
(QCT), 245 Reporters, 148; see also Probes Selective estrogen receptor
Quantitative ultrasound Restriction fragment length modulators (SERM), 245
(QUS), 245 polymorphism (RFLP), Selumetinib, 79
Quizaritinib, 201 147 Sentrix Array Matrix (SAM), 149
Reversed-phase protein array Sentrix Bead Chip, 149
(RPPA), 156 Sequencing-by-ligation approach,
R
Reverse-transcription polymerase 151
Radiological neuroimaging, 237 chain reaction (RT-PCR), Sequestosome 1 [SQSTM1], 254
RAF kinases, 77–78 19 Ser217-Leu, 125
RAF-1 (C-RAF), 77 RFLP analysis, 147 Ser41Tyr variants, 125
Raine syndrome, 253 Ribonucleoprotein complex, 63 Serine arginine-rich splicing factor
Raloxifene, 9, 246 Rigosertib (ON01910.Na), 182 2 (SRSF2), 204
Raltegravir, see AntiHIV drugs RNA in situ hybridization (RNA- SETD2, 152
Raltitrexed, 162 ISH), 19 SETD4, 152
Rapamycin, 79 RNA sequencing (RNA-seq), 235 SETD6, 152
Rarecare European registry RNASEL gene, 124 SF3B1 mutations, 179
project, 41 Roche second-generation Short-chain fatty acid (SCFA), 82
RARβ2 reactivation, 141 sequencing, 150 Short-read sequencing technique,
RAS, see Proto-oncogenes Roflumilast, 282 150
RAS oncoprotein increased Romosozumab, 246 Signal recognition particle 72 gene
expression, 65 RPPA, see Reversed-phase protein (SRP72), 206
463 RAS-wild/BRAF-mutant CRC array (RPPA) Single nucleotide polymorphisms
patients, 59 RPS14haploinsufficiency, 183 (SNPs), 5, 67, 219
RAS-RAF-MAPK pathway, 59 RT-PCR, see Reverse-transcription Single nucleotide variation (SNV), 3
Rb, see Tumor suppressor genes polymerase chain Single-cell genome sequencing, 11–13
(TSGs) reaction (RT-PCR) Single-gene disorders of bone,
RB1, 152 RUNX1 GENE, 196, 206 248–250
 Index 325

Single-molecule sequencing, Statin pharmacogenetics, 224–225 Taxane, 159

150–151 Stem cells, 43 TC, see Theca cells (TC)
Single-strand conformation Stroke and other related T-cell receptor (TCR), 61
polymorphism (SSCP), neurological diseases, TCF7L2 gene, 264
147 235–239 TCGA, see The Cancer Genome
siRNAs, see Small interfering RNAs harnessing the power, 239 Atlas (TCGA)
(siRNAs) phenotype-based precision TCR, see T-cell receptor (TCR)
SKOV3, 83 medicine, 236–237 Telomere length, CRC, 63
SLCO1B1*5 variant, 224 precision genetics for precision cell senescence induction, 63
SMAD-4, 55 medicine, 236 shortening, 63
SMAD-4 protein inactivation, 55 precision medicine in other TERC, 206
Small cell lung cancer, 5 neurological diseases, 237 Teriparatide, 245–246
Small interfering RNAs (siRNAs), Alzheimer’s disease, 238 TERT, 206
141 epilepsy, 237–238 TGFB, see Transforming growth
Smart Cycler®, 148 Parkinson’s disease, 238–239 factor beta gene (TGFB)
SN-38 glucuronide (SN-38G), 68 Stupp protocol, 42 TGF-beta, see Transforming growth
SNPs, see Single nucleotide Subcallosal anterior cingulate factor-β1 (TGF-beta)
polymorphisms (SNPs) cortex (sACC), 304 Thalidomide group, 164
SNV, see Single nucleotide variation Suberoyl anilide hydroxamic acid The Cancer Genome Atlas (TCGA),
(SNV) (SAHA), 141 156
SNVs, see Somatic single nucleotide Supported Oligonucleotide Ligation Theca cells (TC), 153
variants (SNVs) and Detection (SOLiD Thr130Ile polymorphism, 264
Sodium voltage-gated channel alpha sequencing), 151 Thymidylate synthase (TYMS), 72
subunit 8 (SCN8A), 235 Surfactant proteins, 282 high expression group, 73
Solexa, 150 SYBR-Green I, 148 2r/3g genotype, 73
Solid phase-based techniques/ Synergistic research cores, 44 3c/3g genotype, 73
hybridization, 148 Synthesized mRNA, 13 3g/3g genotype, 73
SOLiD sequencing, see Supported Systemic corticosteroids, 286 low expression group, 73
Oligonucleotide Ligation 2r/2r genotype, 73
and Detection (SOLiD 2r/3c genotype, 73
T
sequencing) 3c/3c genotype, 73
Solute carrier family 35 member A2 T2-hyperintense lesion load, 271 TILs, see Tumor-infiltrating
(SLC35A2), 235 TAAs expression, 61 lymphocytes (TILs)
Somatic and germ-line mutation/ TAAs, see Tumor-associated Tivozanib, 79
variance, 151–152 antigens (TAAs) TJ protein 1 (TJP1), 153
Somatic mutations in epithelial Tacrolimus, see TJP, see TJ protein 1 (TJP1)
ovarian cancer, Immunosuppressants TNBCs, see Triple-negative breast
prevalence, 153 Tamoxifen, 9, 164 cancers (TNBCs)
Somatic single nucleotide variants TaqMan, 148 TNFalpha, 280
(SNVs), 21 Targeted deep sequencing TNFRSF11A gene, 251
Sonidebig, 204 capture-based approach, 22 TNM, 29
Sorafenib, 77, 201 PCR-based approach, 22 Toll-like receptor pathway, 183
Splicing factor 3 subunit b1 (SF3b1), Target discovery, 45 Topotecanin, 159
204 Targeted therapies, 74 Toxicity-related biomarkers, 68
Sporadic colorectal cancer (S-CRC), Targeting genes TP53 in crcs, 55–56
56 BRAF, 76–77 caretaker function, 56
SRP72, see Signal recognition KRAS, 74–76 mutation of, 56
particle 72 gene (SRP72) Targeting the pathways TP53 mutated gene, 44
SSCP, see Single-strand MEK, 78–79 TP53 mutation, 10
conformation MTOR, 79–80 TP53, 55–56
polymorphism (SSCP) RAF kinases, 77–78 Trametinib, 76
326   Index

Transcription factor 7-like 2 U W

(TCF7L2) gene, 262
U.S. Preventive Services Task Force Warfarin pharmacogenetics, 225
Transcriptome, 12
(USPSTF), 8 WDR36/TSLP on 5q22, 283
Transcriptomics, 153–154, 284
U2 small nuclear RNA auxiliary Wheezing phenotypes, 281
Transforming growth factor beta
factor 1 (U2AF1), 204 WHO, see World Health
gene (TGFB), 138, 253
U2AF1mutation, 181 Organization (WHO)
Transforming growth factor-β1
Ubiquitin-mediated decay, 153 Whole exome sequencing (WES),
(TGF-beta), 55
UGT1A1*28 genotyping test, 70 11
Transmembrane programmed
UGT1A7, 68 Whole genome sequencing (WGS),
cell death protein 1
UGT1A10, 68 7, 11
(PD1), 61
UGT1A8, 68 Whole metagenome shotgun
Transthyretin (TTR), 228
Upstream stimulatory factor 1 (WMS), 82
Trastuzumab (Herceptin), 5;
(USF-1), 73 Wide molecular diversity, 6
see also HER2 inhibitors
Uridine 5′-diphosphoglucurono­ Wild-type allele (TA6/6)
Trastuzumab (Herceptin®), 164
syltransferase (UDP- genotype, 69
Trebananib, 161
glucuronosyl transferase), Wild-type form, 72
Treg cells, 61; see also Peripheral
68 Wilms’ tumor gene 1 (WT1),
T-effector cells
USF-, see Upstream stimulatory see TAAs expression
Trichostatin A (TSA), 140, 281
factor 1 (USF-1) WMS, see whole metagenome
Triple-negative breast cancers
shotgun (WMS)
(TNBCs), 139
TRK, see Proto-oncogenes
V WNT, see proto-oncogenes
Wnt pathways, 182
T-scores, 245 V600E BRAF mutations, 58
Wnt signaling pathway inhibitors,
TSER*5, 73 VACCINATIONS, 286
204
TSGs, see Tumor suppressor genes Vadastuximab talirine, 204
Wnt/calcium pathway modulator
(TSGs) Valosin-containing protein (VCP),
cyclosporin A, 79
TTR, see Transthyretin (TTR) 254
WNT/β-catenin signaling, 253
Tumor mutational spectrum, 21 Van Buchem disease, 253
WNT-signaling pathway, 84
Tumor profiling, genetic screens, 7 Variant form, 72
World Health Organization
Tumor size (T), see TNM Variant HMECs (vHMECs), 138
(WHO), 3
Tumor suppressor genes (TSGs), 4 Vascular endothelial growth factor
inactivation of, 55 (VEGF), 159
Tumor-associated antigens (TAAs), Vascular endothelial growth factor
X
61 aflibercept, 67
Tumor-infiltrating lymphocytes VEGF, see Vascular endothelial Xelox
(TILs), 54 growth factor (VEGF) capecitabine, 71
Tumor-specific aberrations VEGF/VEGF receptor (VEGFR) oxaliplatin, 71
chromosomal rearrangements, signaling pathway, 159 Xenograft assay, 20
21 Vemurafenib, 45, 76; see also BRAF
epigenetic alterations, 21 inhibitor
somatic single nucleotide Vertebral fractures, 244 Y
variants (snvs), 21 vHMECs, see Variant HMECs
Yohimbine, 264
20-HETE, 280 (vHMECs)
TYMS expression, 72 Vintafolide, 164
TYMS—E2F1 correlation, 72 Vismodegib, 204
Z
Type 1 diabetes mellitus (T1D), 259 Vitamin D supplementation, 245
Type 2 diabetes mellitus (T2D), 259 Vitamin K epoxide reductase Zinc finger CCCH-type (ZRSR2),
Tyrer–Cuzick, 7 (VKORC1), 225 204
Tyrosine kinase formation, 4 VKORC1, 225 Z-scores, 245