To my wife, Anna,

and to my son, Giuseppe

G.A.S.

To my/amily
G.DiF.
Springer-Verlag Italia SrI.
Giulio Aniello Santoro • Giuseppe Oi Falco

Atlas of Endoanal
and Endorectal
Ultrasonography
Staging and Treatment Options
for Anorectal Cancer

With contributions by

D.C.C. Bartolo • C.H. Crane • A. D' Annibale • c.o. Finne

1.1. Gunderson • R. Labianca • U. Hildebrandt • T. Hull • M. Hiinerbein
c.G.S. Hiischer • A. Infantino • J.B. Kruskal· J. Marescaux· G. Melotti
K. Miller • T. Muto • C. Ratto • M. Trompetto • V. Valentini • W.D. Wong

Foreword by

B. Cola

Springer
Giulio Aniello Santoro, M.D., Ph.D.
Head, Section of Anal Physiology and Ultra sound
Coloproctology Service
Department of Surgery
Regional Hospital-Treviso
Italy

Giuseppe Di Falco, M.D.

Chairman, Department of Surgery
Regional Hospital-Treviso
Italy

Contributors
D.C.C. Bartolo· C.H. Crane· A. D'Annibale· C.O. Finne
L.L. Gunderson· R. Labianca .' U. Hildebrandt • T. Hull • M. Hünerbein
C.G.S. Hüscher • A. Infantino • J.B. Kruskal· J. Marescaux • G. Melotti
K. Miller • T. Muto • C. Ratto • M. Trompetto • V. Valentini· W.D. Wong

springeronline.com

© Springer-Verlag Italia 2004

Originally published by Springer-Verlag Italia Milan in 2004
Softcover reprint ofthe hardcover 1st edition 2004
ISBN 978-88-470-2176-1 ISBN 978-88-470-2129-7 (eBook)
DOI 10.1007/978-88-470-2129-7
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcast-
ing, reproduction on microfilm or in any other way, and storage in data banks. Duplieation of this pub-
lication or parts thereof is permitted only under the provisions of the Italian Copyright Law, in its cur-
rent version, and permission for use must always be obtained from Springer-Verlag. Violations are liable
for prosecution under the Italian Copyright Law.
The use of general descriptive names, registered names, trademarks, ete. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protec-
tive laws and regulations and therefore free for general use.
Product Iiability: The publishers cannot guarantee the accuracy of any information about dosage and
application contained in this book. In every individual case the user must check such information by
consulting the relevant literature.

Cover design: Simona Colombo, Milan

Typesetting: ITG sas, Turin

SPIN 10971680
 Foreword

Endoanal and endorectal ultrasound represents considerable progress in the field of

anorectal imaging.
Its importance in the staging of rectal and anal cancer, in identifying perianasto-
motic neoplastic recurrences, in defining acute and chronic suppurative disease and
in assessing sphincteral anatomical integrity, is now a confirmed reality.
The evidence provided by endoanal and endorectal ultrasonography guides the
colorectal surgeon in his choice among the various therapeutic approaches and
helps in assessing the effects of the treatments applied.
The great flexibility and high degree of reliability make this exploratory method
an indispensable tool at all stages of the clinical course in the majority of patients
with anorectal problems, including the fundamental follow-up stage.
The publication of this excellent volume on intraanal and intrarectal ultrasonog-
raphy is therefore to be appreciated, especially since it is in the form of an atlas. As
we are dealing with imaging, it is particularly appropriate that it is the images that
tell the story.
Like all important things, this volume does not come out of nothing.
It is the result of many years of skilled work by Drs. G.A. Santoro and G. Di Falco
and is based on the experience of internationally renowned professional experts.
The ten sections of the atlas begin with general information on ultrasonography,
gradually progressing to the technical problems of the method and the aspects rela-
tive to anorectal tumors.This last section leads to a review of traditional and mod-
ern therapeutic possibilities.
The inclusion at the end of each chapter of an invited commentary, signed by per-
sonalities with experience on a par with the authors, provides an interesting com-
parison between different points of view on extremely topical matters.
A very interesting book, therefore, useful both to the newly qualified and to the
more expert. I hope it will become a volume that is found on everyone's shelf, be they
a surgeon, a gastroenterologist, an oncologist, a radiologist or a pathologist, who is
involved or interested in coloproctology.

Prof. Bruno Cola, M.D., F.A.C.S.

Chairman of the i d Surgical Clinic
University of Bologna
President of the European Council of Coloproctology

Bologna, November 2003

 Preface

The great transformation in rectal surgery in the early 1990'S, from surgery oriented
towards extended resections, with disappointing results for the patients' social life, to
the current sphincter-saving and minimally invasive procedures, has increased the
importance of accurate preoperative staging in decision-making. Endorectal ultra-
sonography for the staging of rectal cancer was first reported at the beginning of 1985
by Hildebrandt and Feifel, from the Department of Surgery at the University of
Saarland, in Homburg, Germany. They proposed an ultrasonic classification based on
the tumor node metastases (TNM) system. Currently this technique remains the best
modality for accurately staging depth of penetration and presumptive nodal status in
rectal neoplasms, guiding therapeutic decisions by identification of lesions suitable
for local excision, transanal endoscopic microsurgical excision or laparoscopic resec-
tion. Recently, three-dimensional ultrasound has resulted in a better understanding
of normal and abnormal anorectal patterns and has improved overall accuracy in
preoperative staging and follow-up of anorectal tumors. In parallel, the progressive
advances in both operative technique and instrumentation has greatly enabled the
possibility of minimally invasive surgery. Combined modality treatments with pre-
operative neoadjuvant radiochemotherapy and intraoperative radiotherapy have
encouraged this development, reducing the volume of tumors which would otherwise
not be resectable, thus enabling previously impossible minimally invasive surgery to
be carried out successfully. The complexity of problems associated with rectal sur-
gery has led to the need of a multidisciplinary approach to anorectal tumors with
close collaboration between surgeons, radiologists, gastroenterologists, oncologists
and radiotherapists.
With these considerations in mind we planned this volume, fruit of over ten years
of experience, as both a textbook and surgical atlas to describe the endoanal and
endorectal ultrasonographic techniques and their clinical relevance along with the
main operative techniques in the field of rectal surgery. The three initial sections of
the book present a broad base of information on physical principles of ultrasound,
currently available ecographic equipment and techniques, future technological
trends and basic surgical and endosonographic anatomy of the rectum and anal
canal. The following sections (IV-VII) provide a considerable detailed description of
the role of endoanal and endorectal ultrasonography in preoperative staging and
postoperative follow-up of anorectal cancers and summarize our personal knowl-
edge in this field with a detailed review of recent ultrasound literature. The more
practical aspects of surgical management, including traditional surgery, total
mesorectal excision with autonomic nerve preservation, sphincter-preserving proce-
dures, trans anal endoscopic microsurgical techniques, advanced laparoscopic proce-
dures, intraoperative laparoscopic ultrasound and robotic techniques, are discussed
in the eighth section taking into account indications, contraindications, risks, bene-
VIII Preface

fits and limitations. The final two sections review combined modality therapy for
rectal cancer and treatment options for anal carcinoma. A number of highly regard-
ed specialists in the field, internationally recognized for their direct personal experi-
ence and scientific expertise, have been selected to provide a critical discussion of
each of the different sections of this atlas. This book is designed to be didactically
sonography and practice-oriented. For this reason considerable space has been ded-
icated to drawings illustrating anatomy and techniques and to two-dimensional and
three-dimensional ecographic figures derived from our experience in clinical prac-
tice compared with radiological, endoscopic, pathologic and histologic images in
order to show the many different facets of a given disease and enable the examiner to
make the correct diagnosis with greater facility. Finally, a comprehensive bibliogra-
phy accompanies each section, which will help the reader further explore the topic.
We are convinced that endoanal and endorectal ultrasonography represent the
first diagnostic approach in staging of anorectal cancer. In many cases ultrasonogra-
phy provides an immediate diagnosis or suggests subsequent diagnostic tools. As
editors we hope this book will give greater insight into the value of this procedure
and will be useful to surgeons training in colorectal surgery to learn how to see and
interpret ecographic images, providing more experienced proctologists with an
opportunity to review and reassess their techniques. Many of the sections, however,
should interest radiologists, gastroenterologists, pathologists, oncologists and radio-
therapists.
The authors wish to express their deep appreciation to all colleagues, among the
foremost investigators in this clinical and research area, who have contributed
towards the realization of this book. Without their expertise and cooperation, this
volume would not have been possible. Thanks must go to our Hospital, whose
advanced technological support gave us the possibility to accomplish to this project,
to the medical illustrator Mr. Massimiliano Crespi who has realized the numerous
artistic drawings, to Mr. Giancarlo Pengo for his help in collecting the radiological
and photographic illustrations, to Mr. Fabrizio Giavenni, managing director of B-K
Medical Italia for gathering much of the data and photographic material of the tech-
nological equipment, and to Mrs. Antonella Cerri, executive editor at Springer-Verlag
Italia, for her constant advice throughout the development of the project, following
every stage of the editorial work.

Treviso, November 2003 G.A. Santoro, G. Di Falco

 Contents

SECTION I
Basic principles of ultrasonography ........................................................................ 1
G.A. Santoro, G. Di Falco

SECTION II
Endoanal and endorectal ultrasonographic techniques ....................................... 9
G.A. Santoro, G. Di Falco

Invited commentary: J.B. Kruskal ............................................................................. 22

SECTION III
Basic anatomy ............................................................................................................. 25
G.A. Santoro, G. Di Falco

III.I. Surgical and endosonographic anatomy of the anal canal ..................... 27

III.2. Surgical and endosonographic anatomy of the rectum ........................... 37

Invited commentary: C. Ratto ................................................................................... 42

SECTION IV
Endoluminal ultrasound in the preoperative staging of rectal carcinoma ........ 49
G.A. Santoro, G. Di Falco

IV.I. Introduction .................................................................................................. 51

IV.2. Stage uTO: Villous adenoma ........................................................................ 61
IV.3. Stage uTI: Submucosal invasion ................................................................. 67
IV4 Stage uT2: Invasion of the muscolar layer ................................................. 74
IV.5. Stage uT3: Perirectal fat invasion ............................................................... 83
IV.6. Stage uT4: Extensive local invasion ............................................................ 91

Invited commentary: D.M. Schaffzin, W.D. Wong .................................................... 95

IV.7. Stage UNI: Lymph node metastases ............................................................ 98

Invited commentary: T. Muto .................................................................................... 109

X Contents

SECTION V
Staging following preoperative chemoradiotherapy
for advanced rectal cancer .............. ......... ......... ................... ........... ..... ..... ....... ...... ... 111
G.A. Santoro, G. Di Falco

Invited commentary: M. Hiinerbein ......................................................................... 119

SECTION VI
Postoperative follow-up of rectal cancer ................................................................. 123
G.A. Santoro, G. Di Falco

VI.I. Postsurgical evaluation ................................................................................ 125

VI.2. Local recurrences .......................................................................................... 130

Invited commentary: U. Hildebrandt ........................................................................ 137

SECTION VII
Endoanal ultrasonography in the staging of anal carcinoma .............................. 139
G.A. Santoro, G. Di Falco

Invited commentary: A. Infantino ............................................................................ 152

SECTION VIII
Surgical treatment options for rectal cancer .......................................................... 155

VIII.I. Introduction .................................................................................................. 157

G.A. Santoro, G. Di Falco
VIII.2. Traditional local excision for rectal cancer ............................................... 158
M. Trompetto

Invited commentary: C.O.Finne ................................................................................ 163

VIII.3. Transanal endoscopic microsurgery .......................................................... 165

G.A. Santoro, C. Pastore, G. Di Falco

Invited commentary: K.Miller ................................................................................... 173

VIII.4. Conventional techniques ............................................................................. 178

G.A. Santoro, G. Di Falco

Invited commentary: 1. Zorcolo, D.C.C. Bartolo ..................................................... 189

Invited commentary: T. Hull ...................................................................................... 193

VIII.5. Laparoscopic techniques ............................................................................. 199

G.A. Santoro, G. Di Falco

Invited commentary: T.H.A. Arulampalam, C.G.S. Hiischer .................................. 213

Invited commentary: G. Melotti ................................................................................ 215
 Contents XI

VIII.6. Intraoperative laparoscopic liver ultrasonography ................................. 217

G.A. Santoro, G. Di Falco

Invited commentary: J.B. Kruskal ............................................................................. 227

VIII.7. Robotic rectal resections ............................................................................. 228

A. D'Annibale, E. Morpurgo

Invited commentary: J. Marescaux, F. Rubino ......................................................... 234

SECTION IX
Combined modality therapy for rectal cancer ....................................................... 239

IX.I. Preoperative combined chemoradiation and adjuvant therapy

for rectal cancer ............................................................................................ 241
G.D. Beretta, S. Mosconi, 1. Milesi, M.A. Pessi, R. Labianca

Invited commentary: V. Valentini .............................................................................. 253

IX.2. Intraoperative radiotherapy in rectal cancer ............................................ 256

V. Valentini, M. Campitelli, G. Mantini

Invited commentary: 1.1. Gunderson, M.G. Haddock ............................................ 263

SECTION X
Treatment options for anal carcinoma .................................................................... 275
A. Infantino, 1. Pisegna Cerone

Invited commentary: C.H. Crane .............................................................................. 285

SUBJECT INDEX ........................................................................................................ 287

 Contributors

THANJAKUMAR H.A. ARULAMPALAM, M.D. ANNIBALE D' ANNIBALE, M.D.

I st Department of General Surgery Chief, Department of General Surgery
San Giovanni - Addolorata Hospital Teaching Center of Minimally Invasive
Roma, Italy and Robotic Surgery
Ospedale di Camposampiero
DAVID c.c. BARTOLO, M.S., ER.C.S. Padova, Italy
Consultant Colorectal Surgeon
Colorectal Unit CHARLES O. FINNE, M.D.
Western General Hospital Division of Colon and Rectal Surgery
Edinburgh, Scotland, UK University of Minnesota
Minneapolis, Minnesota, USA
GIORDANO D. BERETTA, M.D.
Unita Semplice di Oncologia LEONARD L. GUNDERSON M.D., M.S.,
Gastroenterologica Getz Family Professor and Chair
Ospedali Riuniti di Bergamo of Radiation Oncology
Bergamo, Italy Mayo Clinic Scottsdale
Mayo Medical School
MAURA CAMPITELLI, M.D. Deputy Director for Clinical Affairs
Resident, Radiotherapy Department Mayo Clinic Cancer Center
Institute of Radiology Scottsdale, Arizona, USA
Universita Cattolica "S. Cuore"
Roma, Italy MICHAEL G. HADDOCK, M.D.
Associate Professor and Consultant
LIA PISEGNA CERONE, M.D. in Radiation Oncology
Department of Surgery Mayo Clinic Rochester
Ospedale "S. Maria dei Battuti" Mayo Medical School
S.Vito al Tagliamento, Pordenone, Italy Rochester, Minnesota, USA

CHRISTOPHER H. CRANE, M.D. ULRICH HILDEBRANDT, M.D.

Associate Professor Professor of Surgery
Program Director, Gastrointestinal Klinikum Salzgitter GmbH
Section Department of Radiation Chirurgische Klinik I
Oncology Salzgitter, Germany
Associate Director, Gastrointestinal Center
University of Texas
M.D. Anderson Cancer Center
Houston, Texas, USA
XIV Contributors

TRACY HULL, M.D. LAURA MILESI, M.D.

Head, Section of Anal Physiology Department of Oncology
and Ultrasound Ospedali Riuniti di Bergamo
Department of Colorectal Surgery Bergamo, Italy
The Cleveland Clinic Foundation
Cleveland, Ohio, USA KARL MILLER, M.D.
Associate Professor of Surgery
MICHAEL HONERBEIN, M.D., PH.D. Surgical Department,
Department of Surgery and Surgical Krankenhaus Hallein and Ludwig
Oncology Boltzmann Institute for
Charite University Hospital Gastroenterology
Campus Berlin Buch and Helios Hospital and Experimental Surgery
Berlin, Germany Hallein/Salzburg, Austria

CRISTIANO G.S. HOSCHER, M.D., EA.C.S. EMILIO MORPURGO, M.D.

Head, 1st Department of General Surgery Department of General Surgery
Ospedale San Giovanni - Addolorata Teaching Center of Minimally Invasive
Roma, Italy and Robotic Surgery
Ospedale di Camposampiero
ALDO INFANTINO, M.D. Padova, Italy
Chief, Department of Surgery
Ospedale "S. Maria dei Battuti" STEFANIA MOSCONI, M.D.
S. Vito al Tagliamento, Pordenone, Italy Department of Oncology
Ospedali Riuniti di Bergamo
JONATHAN B. KRUSKAL, M.D., PH.D. Bergamo, Italy
Chief, Abdominal Imaging
Beth Israel Deaconess Medical Center TETSUICHIRO MUTO, M.D., PH.D.
Associate Professor of Radiology Director, Cancer Institute Hospital
Harvard Medical School Department of Surgery
Boston, MA, USA Toshima-ku, Tokyo, Japan

ROBERTO LABIANCA, M.D. CLAUDIO PASTORE, M.D.

Head, Department of Oncology Head, Coloproctology Service
Ospedali Riuniti di Bergamo Department of Surgery
Bergamo, Italy Regional Hospital
Treviso, Italy
GIOVANNA MANTINI, M.D.
Assistant Professor MARIA A. PESSI, M.D.
Radiotherapy Department Department of Oncology
Institute of Radiology Ospedali Riuniti di Bergamo
Universita Cattolica "S. Cuore" Bergamo, Italy
Roma,Italy
CARLO RATTO, M.D.
JACQUES MARESCAUX, M.D., ER.C.S. Assistant Professor
IRCAD - European Institute Department of Surgical Sciences
of Telesurgery Universita Cattolica "S. Cuore"
Univerite Louis Pasteur Roma, Italy
Strasbourg, France
FRANCESCO RUBINO, M.D.
GIANLUIGI MELOTTI, M.D.,EA.C.S. IRCAD - European Institute
Chief, Department of General Surgery of Telesurgery
Sant'Agostino Hospital Univerite Louis Pasteur
Modena, Italy Strasbourg, France
 Contributors XV

DAVID M. SCHAFFZIN, M.D. W. DOUGLAS WONG, M.D., EA.C.S.,

Colorectal Service ER.C.S.(C)
Memorial Sloan-Kettering Chief, Colorectal Service
Cancer Center Memorial Sloan-Kettering
New York, NY, USA Cancer Center
New York, NY, USA
MARIO TROMPETTO, M.D.
Head, Coloproctology Unit LUIGI ZORCOLO, M.D.
Ospedale di Ivrea Visiting Surgeon, Colorectal Unit
Torino, Italy Western General Hospital
Edinburgh, Scotland, UK
VINCENZO VALENTINI, M.D. Research Fellow,
Associate Professor University of Cagliari
Radiotherapy Department Cagliari, Italy
Institute of Radiology
Universita Cattolica "S. Cuore"
Roma, Italy
 SECTION I
Basic principles of ultrasonography
 Basic principles of ultrasonography
G.A. Santoro, G. Di Falco

In examinations using external energy sources The term ultrasound describes the general
such as X-rays, ultrasound and injected radioac- form of mechanical energy emitted at a frequency
tive substances, the information contained in the above the limit of human audibility. The maximum
images is produced by the interaction between the frequency audible to man is in the range of 16-
energy and the organ under examination. This 20000 cycles per second, called hertz (Hz), where-
interaction can be visualized in two different as clinical echography now uses frequencies from
ways: an image may be formed by the energy 2.5 to 15 million Hz (MHz), corresponding to wave-
which manages to pass through the organ or its lengths of 0.6-0.01 mm. Elastic waves at ultrasonic
parts, or the energy reflected or scattered from frequency are generated by a perturbation which
portions of the structure under examination may causes the particles of a given medium to vibrate.
be transferred into images. The two techniques The vibration of the particles of the medium is a
are respectively termed visualization by transmis- basic characteristic of the propagation of elastic
sion and visualization by reflection. In the visual- waves. It is therefore impossible for such waves to
ization by transmission technique, the energy is advance in a vacuum. Different modes of propaga-
propagated through the object, i.e. the biological tion are possible and in diagnostic applications
tissues; the energy, which is not absorbed, scat- longitudinal waves are used. In this type of wave
tered or reflected by the tissues, is visualized. the particles, which make up the propagation
Traditional radiology makes use of this technique. medium, vibrate forwards and backwards around
The visualization by reflection method uses the their intermediate position, so that the energy is
energy, which has been scattered back from or transmitted through the medium as a perturbation
reflected by the tissues. In most cases, diagnostic without transfer of matter, and the vibrations of
ultrasound currently uses the technique of forma- the particles are in the direction of propagation of
tion of images by reflection. the wave (Fig. 1.1). The velocity at which the energy

Fig. 1.1. Elastic waves are generated by a

perturbation which causes the particles
of a medium to vibrate. The vibrations of
Impulse Particles of Ultrasonic wave the particles are in the direction of prop-
agation of the wave (redrawn and modi-
the medium
fied from [1])
4 Atlas of Endoanal and Endorectal Ultrasonography

Table 1.1. The values of the absorption coefficient for

various media

Material Absorption coefficient (dB/em)

Water 0.002 a
Blood 0.2
Fat 0.6
Skeletal muscle 1-5- 2 .5
Bone 10
Air 35

b
is transmitted through the medium, which coin-
cides with the velocity of propagation v, depends Fig. 1.2. Specular reflection. Normal incidence (a), oblique
on the elastic properties of the medium and on the incidence (b)
average value of the density of the medium. The
velocity of propagation can be roughly considered
constant in the frequency interval used in diagnos- acoustically uniform medium, reaches an inter-
tic applications. The wavelength A and the fre- face with a medium of differing characteristic
quency f are linked to the velocity of propagation v mechanical impedance (Table 1.2), reflection and
by the equation: v = f x A. One of the characteris- refraction of the wave occurs. Image formation
tics of the propagation of ultrasound within bio- depends entirely on the returning echoes. For inci-
logical tissues is acoustic absorption, mainly due dence of a wave over a theoretically plane surface,
to the transformation of the ultrasonic energy into along the direction of the perpendicular to the
thermal energy. In soft tissues, absorption is surface (normal incidence) (Fig. 1.2a), part of the
roughly proportional to the frequency. Absorption incident energy is transmitted whilst the rest is
in biological tissues is mainly linked to the protein reflected (specular reflection). When the incidence
content. Absorption is very low in fluid media, of the ultrasound wave between two media is not
intermediate in soft tissue and very high in bone normal (oblique incidence) (Fig. 1.2b), part of the
and gas at the frequencies used in diagnostics incident wave is reflected, with the angle of reflec-
(Table 1.1). tion equaling the angle of incidence of the beam to
The intensity at each point of a wave is defined the interface, whilst the rest is refracted, with an
as the energy flow per unit of time through the angle of refraction different from the angle of inci-
unit area perpendicular to the direction of propa- dence (Fig. 1.3). Using a rotating endoprobe with-
gation at the point considered. The amplitude of in mainly circular structures such as sphincters,
the ultrasound beam is expressed in decibels (dB). most specular reflections will be at right angles
Attenuation of ultrasonic intensity of the ultra- and so maximal (Fig. 1.4). In cases of interfaces
sound beam is not only due to absorption, but also much smaller than the wavelength, the ultrasound
to divergence of the beam, scattering and reflec- beam is scattered uniformly in all directions
tion. When an ultrasonic wave, propagating in an (diffraction) (Fig. 1.5). Scattering within biological
tissues reduces the propagating energy so that it
contributes to attenuation together with reflec-
tion, refraction and absorption. Although the
Table 1.2. The values of the resonance coefficient for properties of the biological tissues contribute to
various interfaces the formation of discontinuity of characteristic
impedance and their variations, the physics of
Resonance coefficient scattering or reflection of widespread structures is
not yet possible to establish. The amount of colla-
Interface between water and fat 3·5% gen fibers is the dominant element in determining
Interface between water and bone 68% the value of characteristic impedance. Fat, gas and
Interface between water and air 100% bone have very different densities, but various soft
 Section /. Basic principles of ultrasonography 5

ultrasound frequency depends on the thickness

of the crystal. When the mechanical vibration
2
encounters an obstacle (the interfaces between
one tissue and another, discontinuity in the bio-
logical tissue, cavities full of fluid, air bubbles or
foreign bodies) as the wave front advances part of
the energy is reflected and the rest continues. The
3------------------~----------------- same transducer used to emit the ultrasonic
impulse responds to mechanical vibrations by
generating a corresponding electrical signal. The
signal is transmitted to a series of electronic cir-
cuits and frequently presented as a luminous sig-
nal on the screen of an oscilloscope (Fig. 1.6). By
making a correlation between the luminous point
on the screen and the distance travelled by the
Fig. 1.3. Refraction of the ultrasound wave along the path-
way with oblique incidence between two media with differ- ultrasonic waves advancing within the tissues, an
ent propagation velocities: 1) incident wave, 2) reflected image is formed of the organization of the layers
wave, 3) interface, 4) normal line, 5) refracted or transmitted encountered by the ultrasound beam in the direc-
wave tion in which the transducer is pointing. The
images used in clinical echography may appear as
images of the amplitude of the echo signals in
tissues have very similar densities and changes in relation to the distance from the probe along a
impedance are due to alteration in the elasticity of single observation direction (A-mode or ampli-
the tissues.
The ultrasonic energy required to form an
image must be supplied from an external source,
which can be continuous or impulsive. The for-
mation of images by reflection makes use of
T I I
impulsive ultrasound energy. The basic transduc-
er, also called probe, which transmits and receives
the ultrasonic impulses is made up of a piezoelec-
tric disk (which is a ceramic crystal, usually lead-
zirconate-titanate). After the application of an
electrical current, it vibrates and transmits these
vibrations into the object in contact with it. The

Fig. 1.4. Using a rotating endoprobe within mainly circular Fig. 1.5. Diffraction develops when the interface is much
structures most specular reflections will be at right angles smaller than the wavelength of the incident wave that is scat-
and so maximal tered uniformly in all directions (T = transducer)
6 Atlas of Endoanal and Endorectal Ultrasonography

Fig. 1.6. Diagram of echographic equip-

ment with A-mode display: a) transmit-
ter, b) receiver, c) oscilloscope, d) trigger
signal, e) pulse electrical signal, f) piezo-
electric transducer, g) pulse mechanical
signal, h) medium 1, i) medium 2,1) medi-
um3

tude mode) or the return signals may be dis- optimum range of resolution for the transducer to
played as luminous dots on the screen (B-mode work at. For the B-K Medical 10 MHz transducer,
or brightness mode), whose brightness is ampli- the axial resolution is <0.05 mm and the lateral
tude dependent. Gray scale means that the bright- resolution 0.5-1 mm, with a focal range of 5-45
ness is assigned to one of a fixed series of shades mm. As the beam goes through tissue it is attenu-
of gray, usually 256 in number. The luminous dots ated at approximately 1 dB/cm per MHz.
are displayed on the screen in the same sequence Reflections from deeper structures are weaker, due
of time and geometry according to which the to greater signal attenuation. Accurate adjustment
ultrasonic signal encounters a differing acoustic of amplification is important in order to enable
impedance along its path, whilst the ultrasonic adequate screen display. Gain adjustments influ-
beam moves in the scanning plane. This system ence both axial and lateral resolution obtained in
permits two-dimensional imaging of a section mode B display (Fig. 1.7). Different images are
(tomography) in the scanning plane. If a obtained for the same objects, at the same dis-
sequence of scans is performed at a sufficiently tance, with the same transducer and hence with
high frequency, a continuous real time display is the same potential lateral resolution (Fig. 1.8). As
supplied of sections of the organ, even though it the gain increases the rendering of detail in the
is in motion. Information on the absorption of images deteriorates.
different tissues and the depth of the organ to be Reverberation is an artifact due to a gross mis-
studied allows a choice of probes for various match of acoustic impedance at an interface, usu-
applications. The highest frequencies give better ally an air/tissue interface, causing a very strong
resolution and better display of detail, but the reflection which is only partially absorbed by the
trade-off for this is a short focal zone and limited transducer and which is reflected back to the inter-
penetration. face where it is again reflected back to the trans-
Axial resolution is the capacity to resolve two ducer. The beam therefore reverberates back and
point reflectors in the direction of the axis of the forth from the interface and transducer, creating a
ultrasound beam. Lateral resolution signifies the series of reflections separated by a constant dis-
capacity for resolving two point reflectors at an tance, which indicates the distance of the reflec-
equal distance from the transducer but situated in tion from the transducer (Fig. 1.9). Reverberation
two different directions from it. With a mechani- echoes are common in anal endosonography when
cally rotated single crystal, this is fixed by the there is loss of acoustic contact within the canal
design of the transducer. Focusing describes the (Fig. 1.10).
 Section /. Basic principles of ultrasonography 7

MODE

I I
Fig. 1.7. Diagram of the influence of gain
I I adjustment of echographic equipment on
I : I
the resolution obtained in mode B dis-

~ ~I
MO D I! •••
play. As the gain increases (right-hand
I figure) the rendering of details in the
images deteriorates

a b
Fig. I.S. The uniform time gain compensation makes the image in (b) more brighter

The echogenicity of any structure may be reflectivity, but is usually moderately reflective.
characterized by the level of echoes within it Collagen is more reflective and whiter, whereas
(hyper- or hypoechoic). The echo pattern may be muscle fibers are poorly reflective and blacker.
homogeneous or inhomogeneous. Water has the Solid tumors are usually poorly reflective and
lowest reflectivity and appears black. Fat varies in inhomogeneous.
 8 Atlas of Endoanal and Endorectal Ultrasonography

T
jI'II ....

1 2 ~
J 1
3 .) ls~
Fig. 1.9. Reverberation echoes develop
1+5 when there is a very strong reflection
returning to the transducer. Part will be
1+2 reflected from the transducer back into
the tissue. This in turn will be reflected
'·2·3 back to the transducer and so on. This
creates a series of equally spaced reflec-
1+2·3·6 _ __ tions. The gap indicates the distance
between the transducer and the initial
reflection

• -3 ;;
',~-

,
I,. " ,'. .j fi
.~
... .. - f '
.-.. ~!...
.~t'!~
.• ~
i.- , '
-.
'

""
,,'
t
• 0 ~,
\t~~ __ . ;'.'
.t€ll~\ ' "---- ~.: . ·
':~r
'.
) .... .
~-~'4
. . . . ."; . ~,·l.

a 3 b
Fig. 1.10. In (a) the image has been taken with the probe standing free in air. The gas impedance mismatch at the cone/air inter-
face creates major reflection echo patterns with an equally spaced series of concentric rings. In (b) the probe has been inserted
beyond the anal canal into the rectal ampulla. Acoustic contact has been lost from 9-11 o'clock producing a reverberation echo

References

1. Bartram CI, Frudinger A. Basic principles of ultra-

sonography. In: Handbook of anal endosonography.
Wrightson Biomedical Publishing Ltd; Petersfield, UK:
1977
 SECTION II
Endoanal and endorectal
ultrasonographic techniques
 Endoanal and endorectal
ultrasonographic techniques
G.A. Santoro, G. Di Falco

In order to obtain meaningful ultrasonic images, pressing of a button. All of these probes have a
the operator must have an overall understanding frequency range from 5 to 10 MHz. The focal
and therefore correct use of the controls available range of a 7.0 MHz transducer is 2-5 cm, whereas
on the ultrasound device. In fact, unsuitable regu- a 5.5 MHz transducer has a focal range of 1-4 cm.
lation of the equipment produces poor images and The longer focal length provides an advantage in
can lead to false positive or negative diagnosis. examining perirectal structures. The 10.0 MHz
Many types of ultrasound probes have been transducer is preferred in the anal canal as it
developed to evaluate the rectal wall and anal gives improved resolution of the sphincter mech-
sphincter. Most of these have been made to exam- anism over the 7.0 MHz transducer.
ine the prostate gland and are not suitable for Ultrasound transducers at the tip of a colono-
evaluating the wall of the anorectum and the scope (echoendoscope) (Fig. 11.1) can be used to
immediately adjacent tissues. The types of evaluate the wall of the colon all the way around to
endorectal probes include mechanical sector the cecum [1]. These ultrasonic colonofiberscopes
probes with a single transducer, which may have provide both endoscopic and sonographic imag-
a limited field of view of 120 to 210 degrees or ing and can offer simultaneous observation on the
may incorporate radial probes with a full 360- same monitor (picture in picture presentation).
degree field of view, and linear and curved array The angulation capability of 180 0 UP/DOWN and
probes, which have a limited 120-degree to 210- 1600 LEFT/RIGHT enables a thorough exploration
degree field of vision. Some of these probes are of affected areas (Fig. 11.2). Fiberscopes include a
biplanar and can be changed from the axial plane working biopsy channel allowing for mucosal
to the longitudinal (sagittal, coronal) plane by biopsy to be carried out during ultrasound proce-

1800 UP 1600 LEFT

1800 DOWN 1600 RIGHT

Fig. 11.1. Ultrasonic colonofiberscope (Olympus CF type Fig. 11.2. Ultrasonic colonofiberscope: angulation range
UM20-12) (Olympus CF type UM20-12)
12 Atlas of Endoanal and Endorectal Ultrasonography

Fig. II.3. Ultrasonic colonofiberscope: biopsy forceps Fig. II.S. Ultrasonic colonofiberscope: balloon contact
(Olympus CF type UM20-12) method (Olympus CF type UM20-12)

dures (Fig. 11.3). Two models with different fre- We currently use an Hawk 2102 EXL B-K
quencyoptions (7.5 MHz, 12 MHz) are now avail- Medical scanner (Fig. 11.6) with 1850 transducer
able. The 7.5 MHz frequency provides clear and radial array probe which gives a real-time 360-
accurate sonographic imaging of excellent quali- degree axial view of the rectal wall (B-K Medical
ty. The highly detailed imaging provided by the 12 A/S, Mileparken 34, DK-2730 Herlev, Denmark)
MHz frequency is utilized for intensive examina- (Fig. 11.7) [2,31. The radial probe has a 24-cm metal
tion of the five layers of the colon wall. The trans- shaft with a rotating transducer at its tip. The
ducer and resultant ultrasonic beam allow for transducer's crystal has a frequency range from 5
360 0 sector scanning (mechanical, radial scan- to 10 MHz with a focal length of 2-5 cm and a 90-
ning) (Fig. 11-4). Two sonographic contact meth- degree scanning plane and is rotated at 4-6 cycles
ods can be used: the water immersion method per second to get a radial scan of the rectum and
(deaerated water is used to completely fill the surrounding structures (Fig. 11.8) [4,51. The probe
colon) and the balloon contact method (Fig. 11.5). unit must be prepared carefully. The rectal tube
The echoendoscopes are not further discussed in (UA 0799) fits over the central axle that holds the
this chapter. transducer. The rotating transducer is pushed into
the end of the axle and is covered by a latex balloon
that is held in place by two metal rings, one of
which screws onto the metal shaft (Figs. 11.9, 10). A
syringe filled with 50 ml of tap water is manipulat-
Distal End Ultrasonic ed to remove air bubbles, then attached to a spigot

,,-...
Cap' Transducer at the base of the probe. Water is gently injected
I I into the balloon; the water is then withdrawn again
~
I with the transducer tip in the dependent position.
Nozzle In this way all of the air is removed from the water-
i
filled balloon. This procedure may have to be
J~. i:J -Ugh t Guide
Lens repeated to remove all of the air (Fig. lI.n). The
Inst~ument latex balloon filled with degassed water allows
Cha nnel ~ acoustic coupling between the transducer and the
Outlet rectal wall (Fig. 11.12). The rectum can be of vary-
Objective Lens ing diameters and therefore the volume of water in
the balloon may have to be adjusted intermittently.
Fig. 11.4. Ultrasonic colonofiberscope: ultrasonic transducer
Endoluminal ultrasound is usually performed
(Olympus CF type UM20- 12) with the patient in the left lateral decubitus posi-
 Section /I. Endoanal and endorectal ultrasonographic techniques 13

Fig. II.S. B-K Medical rotating endoprobe type 1850

Fig. 11.9. B-K Medical anorectal probe type 1850 assembled

with latex balloon

Fig. 11.10. Endoluminal ultrasound probe with a transducer

Fig. 11.6. B-K Medical Scanner: Hawk 2102 EXL surrounded by a water-filled balloon at its tip

Fig. 11.7. B-K Medical anorectal probe type 1850 (dismantled), with rectal tube, 10 MHz transducer, plastic cone and latex bal-
loon
14 Atlas of Endoanal and Endorectal Ultrasonography

stenotic annular lesion, the finger can check to

determine whether it will allow easy passage of
the probe [6, 71. The entire shaft of the probe is
coated with a thin layer of warm gel using a paper
towel. The probe tip is gently inserted through the
anal canal and then angled posteriorly and
advanced cephalad to as high a level as possible,
with the bony sacrum used as a curved landmark.
The patient should be instructed before the exam-
ination that no pain should be experienced. If
pain should occur, the study should be halted
until the cause of the pain is elucidated. Under no
circumstances should force be used to advance
the probe. The examiner should never try to push
the tip through a narrow stenotic lesion. However,
in most instances passage can be achieved,
although the volume of the fluid in the balloon
will have to be substantially reduced in order to
Fig. 11.11. The balloon can be inflated and deflated using the withdraw the probe through the stenotic portion.
syringe at the base of the 24-cm metal shaft In some instances, a plastic cup, which is then
filled with water, will facilitate the imaging of a
stenotic lesion. Once the tip is advanced to as high
tion. Before the probe is inserted into the rectum, a level as possible, usually 10-14 em from the anal
a digital rectal examination may be performed to verge, the balloon can be inflated with 50 ml of
identify the size, fixation, morphology and loca- water. Now the rotating transducer is activated
tion of the tumor, if it is low enough. If there is a and the rectal wall visualized. When the spigot for

Fig. 11.12. Diagrammatic representation

of a transrectal endosonography
 Section //- Endoanal and endorectal ultrasonographic techniques 15

air or stool gets between the balloon and rectal

wall, it will prevent visualization of the wall. To
avoid this we administer an enema 2 hours before
the examination, but despite this, problems can
arise and it may be necessary to remove the probe
and suction out the rectum with reintroduction in
order to optimize the image.
P. C'I1M
With the probe at the highest level possible
Fig. 11.13. Anorectal endosonography. Lesion localization on and with good visualization of the rectal wall,
the screen when the patient is in the left lateral position and images are obtained at l-cm intervals as the
the spigot for introducing water into the balloon is pointed probe is withdrawn. The exact level of the trans-
towards the ceiling ducer tip can be read off the metal shaft of the
ultrasound probe. More closely spaced images
(0.5 cm) are obtained in the area of any abnor-
introducing water into the balloon is pointing mality. The balloon may have to be deflated and
towards the ceiling, by convention the anterior reinflated to maintain good acoustic contact with
aspect of the rectum will be at the top (12 o'clock) the rectal wall as the probe is withdrawn down
on the screen, right lateral will be left (9 o'clock) the rectum. Once the entire rectum down to the
on the screen, left lateral will be right (3 o'clock) anal sphincter has been evaluated the balloon is
on the screen and posterior will be inferior (6 fully deflated and the probe is removed from the
o'clock) on the screen (just like the image on axial rectum. The entire length of the rectal tumor has
CT scans) (Fig. II.13). The tip of the ultrasound to be carefully examined and it is not uncommon
probe should be maintained in the center of the to require several passes along the full length of
rectal lumen to gain optimal imaging of the rectal the tumor in order to gain all the information
wall and perirectal structures. Some adjustments that is necessary. In some instances, two to six
may have to be made in the gain of the ultrasound passes may be required to properly stage a rectal
scanner to provide optimal imaging. Occasionally, cancer. In most instances the use of a large bore
it is possible to perfectly depict all five layers of proctoscope (Sapimed, Alessandria, Italy) serves
the rectum circumferentially, but usually only a several purposes (Fig. II.14). It allows visual
portion of the rectal wall can be optimally imaged examination of the rectal tumor with exact deter-
at one time and minor adjustments will have to be mination of its location both with respect to cir-
made in the location of the probe relative to the cumferential involvement of the rectal wall and
rectal wall at various locations to optimally image the distance from the anal verge. Secondly it
all five layers clearly. The amount of water in the allows suctioning of any residual stool or enema
balloon may have to be increased to provide com- fluid that might interfere with the acoustic path-
plete acoustic coupling with the rectal wall. The ways of the ultrasound waves, which may distort
examiner should never distend the balloon with the image. Most importantly, however, it allows
more than 80 ml of degassed water, as it may rup- easy passage of the probe above the tumor to
ture. If this occurs, the probe must be removed insure that the transducer is advanced above the
from the rectum and cleaned, a new balloon rectal lesion to allow complete imaging of the
installed and the whole procedure started over. If rectal tumor. This is of extreme importance as the

Fig. 11.14. The use of a large proctoscope

allows easy passage of the probe above
the rectal lesion
16 Atlas of Endoanal and Endorectal Ultrasonography

Fig. II.15. B-K Medical multiplane tran-

srectal endosonic probe type 8551

lower border of a rectal cancer can differ signifi- ducer is limited to lesions with a maximum dis-
cantly in the depth of invasion than the center or tance of 10 em from the anocutaneous line. The
upper portions of the cancer and lymph nodes in scanning plane rotates about a common axis
the perirectal region are often just above the level through all oblique planes from transverse to lon-
of the tumor and will be missed if complete imag- gitudinal (multiplane scanner). The transrectal
ing is not obtained. Small distal lesions can be biopsy is performed through a built-in needle
adequately imaged with the ultrasound inserted guide under real-time endosonographic control
blindly and advanced above the lesion, but for (Fig. 11.16). If the lesion is a localized perineal one,
most mid rectal bulky tumors, the use of a proc- a trans-perineal needle biopsy under local anes-
toscope will facilitate the passage of the trans- thesia can be performed using the B-K Medical
ducer. Once the 20 em score mark on the shaft of 1850 endoprobe (Figs. 11.17, 18). A disadvantage of
the probe is at the proximal end of the procto- the trans-perineal route is that it does not allow
scope, the proctoscope is then pulled back on the the examiner to follow the biopsy needle along the
probe as far as possible thus exposing the trans- entire pathway leading to the ultrasound visual-
ducer for 7 em beyond the end of the proctoscope ized node, since a transverse ultrasound probe
and thus positioned above the rectal cancer. The must be used for a trans-perineal procedure.
balloon is then instilled with 30-60 cc of water, If the anus is being evaluated, the same probe
the volume of fluid usually needed to gain opti- is used with a transducer that has a shorter focal
mal imaging. zone of 1-4 em. A water-filled hard plastic cone
In every patient with suspicious lesions or sus- (WA 0543), made of a sonolucent polymethyl pen-
pected tumoral recurrence ultrasound-guided tene (TPX) plastic and 1.7 em in outer diameter, is
needle biopsies can be performed to obtain a used in place of the latex balloon (Fig. 11.19) [8]. A
specimen for histological examination. If the hole at the top allows air to escape as the assembly
lesion is a localized intramural one or is near the is filled with degassed water (Fig. II.20).A gel con-
rectal wall «1.5 em) a transrectal needle biopsy taining condom is then placed over the probe and
without anesthesia can be performed using a B-K a thin layer of water-soluble lubricant is placed on
Medical 8551 multiplane-endoprobe. The 8551 type the exterior of the condom. Any air interface will
is a rigid sector scanner with a sector angle of 112°, cause a major interference pattern. The probe is
using a frequency of 7.5 MHz with a focal length of now ready for insertion. The outer walls of this
1 to 6 em (Fig. 11.15). This probe cannot be insert- cone are parallel, so that the probe may be moved
ed via a rectoscope, so the use of the 8551 trans- within the anal canal without causing any

Fig. 11.16. Diagrammatic representation

of a trans rectal biopsy performed
through a built-in needle guide using B-K
Medical 8551 endoprobe
 Section /I. Endoanal and endorectal ultrasonographic techniques 17

Fig. 11.17. Equipment used to perform a trans-perineal needle

biopsy utilizing the B-K Medical 1850 endoprobe: (a) disman-
tled, (b) assembled b

anatomical distortion (Fig. 11.21). This is very floor, to assess sphincter contraction. At the origin
important when assessing the canal at different of the canal the "u" shaped sling of the puborec-
levels. There does not appear to be any benefit talis is the main landmark and should be used for
from asking the patient to "squeeze" their pelvic final adjustment [9].

a L-______________________________________ ~~ __ ~

Fig. 11.18. Diagrammatic representation

of a trans-perineal needle biopsy per-
formed using the B-K Medical 1850 endo-
b L-__ ~~ ________ ~~ __________________ ~
probe (a, b)
 18 Atlas of Endoanal and Endorectal Ultrasonography

. \~
. - ,:,~ -- "
a L.~ _~~

b
Fig. 11.19. Endoluminal ultrasound probe with a transducer
surrounded by a hard plastic cone at its tip (a, b)

Fig. 11.20. The hard plastic cone can be filled with water using
the syringe at the base of the 24-cm metal shaft

Fig. 11.21. Diagrammatic representation

of a trans anal endosonography
 Section /I. Endoanal and endorectal ultrasonographic techniques 19

b
Fig. 11.24. B-K Medical ultrasound probe type 1850 and col-
orectal pullback mover (a, b) used to acquire 3D anorectal
endo images

It is now possible to perform the three-dimen-

sional (3D) reconstruction of two-dimensional
(2D) images [10]. It is not necessary to use new
Fig. 11.22. B-K Medical L3Di-2000!2100 for acquisition, ultrasound probes, but to connect the ultrasound
reconstruction and visualization of 3D ultrasound images apparatus to a computer equipped with software
(L3Di-2000/2100) allowing for 3D reconstruction
(Fig. 11.22). A 3D reconstruction is based on a high
number of parallel transaxial images (Fig. 11.23)
acquired using a special colorectal pullback
mover (UA0552) with the B-K Medical ultrasound
probe type 1850 (Fig. 11.24). The colorectal pull-
back mover is a computer-controlled, motor-driv-
en device that can be operated at different levels of
resolution. For the endoanal application, the usual
setting is 0.2-0.3 mm between adjacent transaxial
images. Scanning the anal canal with these set-
tings over a pullback distance of 35 mm will typi-
cally yield 175 parallel images. The data from a
series of closely spaced two-dimensional images
is combined to create a 3D volume displayed as a
cube (Fig. 11.25). The 3D image does not remain
fixed, it can be freely rotated, rendered, tilted and
sliced to allow the operator to infinitely vary the
different section parameters and visualize the
lesion at different angles and to get the most
information out of the data (Fig. 11.26). After the
data have been acquired it is immediately possible
to select coronal anterior-posterior or posterior-
anterior as well as sagittal right-left views. The
Fig. 11.23. Schematic model for acquisition of 3D anorectal
multiview function allows seeing up to six differ-
endo images as parallel transverse 2D images ent and specialized views at once with real-time
20 Atlas of Endoanal and Endorectal Ultrasonography

reconstruction (Fig. 11.27). In addition the images

may be easily saved, reviewed and studied. The
advantage of 3D reconstruction for rectal pathol-
ogy consists in better visualization of the perirec-
tal region, notably of the mesorectum, the possi-
bility to visualize the perirectal infiltration of a
stenosing tumor and a better yield for trans-rectal
biopsies [11].
Extensive anorectal examinations require
moving the transducer head. Probe movement
may cause artifacts and change anatomical pre-
Fig. 11.25. Schematic model for reconstruction of the sentation. The B-K Medical 2050 anorectal trans-
acquired set of 2D images into a 3D volume displayed as a ducer is designed so that no moving parts come
cube into contact with human tissue. The transducer's

Fig. 11.26. 3D image of a T2 rectal tumor

extending from the 3 to 9 o'clock posi-
tion. The sagittal projection suggests that
the lateral margin of the tumor is con-
fined to the muscolaris propria

Fig. 11.27. The scan image in multi-view

with 6 types of cube image presentation
 Section /I. Endoanal and endorectal ultrasonographic techniques 21

3600 rotating head, the proximal-distal actuation

mechanism and the electronic mover are fully
enclos~d within the housing of the slim probe
(Fig. I1.28). Both 3D data set acquisition and high
precision positioning of the scan head over a lon-
gitudinal distance of 60 mm are accomplished at
the touch of a button, allowing getting the infor-
mation without having to move the probe's posi-
tion. The 2050'S double crystal covers a frequency
range from 6 to 16 MHz. With a shaft length of
270 mm, the probe is long enough to thoroughly
cover the entire rectum plus the sigmoideum. It
also can pass through a 200 mm X 20 mm recto-
scope.

Fig. 11.28. B-K Medical anorectal transducer type 2050

References 6. Hildebrandt U, Feifel G. Preoperative staging of rectal

cancer by intrarectal ultrasound. Dis Colon Rectum
1. Hunerbein M, Totkas S, Ghadimi B, Schlag PM. 1985;28:42-6
Preoperative evaluation of colorectal neoplasms by 7. Lohnert MSS, Doniec JM, Henne- Bruns D.
colonoscopic miniprobe ultrasonography. Ann Surg Effectiveness of endoluminal sonography in the iden-
2000;232:46-50 tification of occult local rectal cancer recurrences. Dis
2. Bartram CI, Frudinger A. Handbook of anal endosonog- Colon Rectum 2000;43:483-91
raphy. Petersfield, UK: Wrightson Biomedical Publishing 8. Saclarides TJ. Endorectal ultrasound. Surg Clin North
Ltd,1997 Am 1998;78:237-49
3. Hildebrandt U, Feifel G, Schwarz HP, Scherr O. 9. Hussain SM, Stoker J, Schutte HE, Lameris JS. Imaging
Endorectal ultrasound: instrumentation and clinical of the anorectal region. Europ J RadioI1996;22:n6-22
aspects. Int J Colorectal Dis 1986;1:203-7 10. Giovannini M. Three-dimensional endorectal ultra-
4. Law PJ, Bartram CI. Anal endosonography: technique sound: gadget or technology of the future? Acta
and normal anatomy. Gastrointest RadioI1989;14:349-53 Endoscopica 2000;30:19-25
5. Kumar A, Scholefield JH. Endosonography of the anal n. Hunerbein M, Schlag PM. 3D-endosonography for
canal and rectum. World J Surg 2000;24:208-15 staging of rectal cancer. Ann Surg 1997;225:432-8
 Invited commentary
J.B. Kruskal

The authors have provided a thorough overview anus since the hard anal cap provides sufficient
of current techniques employed for imaging the coupling with the anal wall, even in patients with
anus and rectum using an endoluminal ultra- lax sphincters. I cannot emphasize enough the
sound approach. The techniques discussed are importance of assembling the transducers
quite vendor-specific and the reader should be appropriately. It is important to eliminate all
aware that other manufacturers also produce bubbles within the hard anal cap or the balloon,
transducers capable of imaging the rectum and given that these may produce artifact and limit
anus. Many of these transducers however are the overall utility of the study. It is also impor-
designed for endovaginal imaging and are thus tant to obtain appropriate acoustic coupling in
not optimized for the unique requirements of the rectum, and this may require additional
imaging the rectum or anus. The equipment cleansing prior to commencing the study. For
described in this chapter is however the most bowel cleansing, we routinely administer an
technically optimal, given the ability of these enema immediately prior to the rectal study, but
transducers to image the hollow lumen of the do not require any cleansing for anal imaging.
rectum and anus for the entire 360 degrees. This When inserting the rectal transducer, we prefer
circumferential imaging ability is especially to insert the transducer as proximal as possible
advantageous both in the rectum and anus; in prior to inflating the balloon, in other words as
the rectum, the 360-degree field of view is impor- high up the rectum as is comfortable. This limits
tant when staging tumors and looking for extra pain in patients with large polypoid intraluminal
rectal spread of disease. In the anus, the ability to masses and allows us to get more accurate stag-
image circumferentially is especially important ing of lesions in their more proximal extent in
when evaluating the complex anatomy of the the colon. The operator should also be aware that
internal and external anal sphincters in a real- the balloon can be de- or re-inflated as needed
time manner. The operator should also be aware when evaluating rectal wall tumors. When insert-
of the limited depth of field of this equipment, ing the transducer in patients with enlarged
which limits the overall utility of ultrasound for prostate glands, the transducer should be angled
staging rectal and anal tumors. MRI is clearly more posteriorly to limit the extent of discom-
better for evaluating extra-rectal spread, but fort. If patients experience discomfort when
ultrasound provides accurate tumor staging inserting the transducer, this may be because of
information that is clinically useful for the sur- the presence of a rectal tumor, constricting
geon planning a resection. As far as the scanning lesions or changes resulting from radiation or
technique is concerned, we do not use the water inflammatory conditions such as Crohn's dis-
immersion method for providing acoustic cou- ease. In these situations, the balloon should
pling in the rectum. This results in unnecessary always be deflated since it may catch on the
leakage and prolongs the procedure time. Also, tumor and move this in conjunction with the
the balloon contact method is not required in the transducer being inserted.
 Section /I. Endoanal and endorectal ultrasonographic techniques 23

It is important to keep the transducer in the advocated in this chapter, is not widely used in
center of the lumen. By inserting the transducer our experience and the needle cannot be visual-
into an eccentric location, compression of layers ized along its entire path. Given that important
of the rectal wall may occur, falsely leading to the structures may exist in the path of the needle, we
impression of a hypo echoic mass or falsely result- would recommend transrectal rather than
ing in overstaging of tumors. We never distend the transperineal biopsies.
rectal balloon with more than 50 ml of degassed It is important, when performing anal ultra-
water in all cases. The operator should also be sound, that the operator understands the indica-
aware of the spectrum of pitfalls that may occur tions for the study. Not only is ultrasound used for
when staging rectal tumors. These pitfalls may be staging tumors, but it is also widely used for eval-
due to the presence of air or stool between the uating patients with incontinence or prolapse. In
transducer balloons and the rectal wall. Similarly, this clinical setting, the operator should be well
air trapped within a balloon can also produce aware of the complex anatomy, not only of the
acoustic shadowing, limiting full evaluation of a internal and external anal sphincters, but of the
rectal mass. Other pitfalls may arise from adjacent longitudinal muscle, the puborectalis and the leva-
loops of large or small bowel, diverticuli in the tor ani muscles. Their relationship and location are
colon wall and lesions that do not permit ade- important to facilitate surgical planning for these
quate distension of the rectal lumen. It is also patients. The authors also include a discussion of
important for the operator to obtain a good histo- three-dimensional imaging. While 3D imaging is
ry from the patient, especially if lesions have being used with increasing frequency with ultra-
undergone biopsy or radiation treatment. Both of sound, the overall utility has not been proven at
these may produce artifacts, which not only this stage and awaits prospective clinical studies.
change the imaging appearance, but may result in The authors do not discuss the use of color
overstaging of the tumor. flow and ultrasound contrast agents. While none
As far as biopsies are concerned, we choose to of these have been shown to be useful at this stage,
do all our biopsies through the transducer via the their increasing use in evaluation may result in
endorectal route. It is easy to identify the lesion their utility being proven, not only for the identi-
and use the built-in guide while inserting the nee- fication of tumors, but for characterization and
dle into a lesion. The transperineal approach, as staging.
SECTION III
Basic anatomy
 111.1.
Surgical and endosonographic
anatomy of the anal canal
G.A. Santoro, G. Di Falco

The anal canal is 2-4 cm long. The dentate line of cylinder that encompasses the internal sphincter
the mucosa denotes the squamocolumnar junc- (Figs. lIb, 2). The levator ani is a thin muscle aris-
tion. The circular smooth muscle of the rectal wall ing from the side wall of the pelvis, supporting the
continues downward as the internal anal sphinc- pelvic contents, and separating the ischio-anal
ter (lAS). Anteriorly the muscle fibers from its fossa below from the supra elevator space above
anterior superior part run into the perineal body. (Fig. I11.3) [1]. The iliococcygeus lies postero-Iat-
The outer longitudinal component of the musco- erally, arising from the ischial spine to insert into
laris propria extends as the longitudinal layer of the coccyx and anococcygeal raphe. The ischio-
the anal canal. It acts as an anchor, with its fibroe- coccygeus is a small, often rudimentary, subdivi-
lastic component permeating through the subcu- sion of this. The pubococcygeus originates from
taneous part of the external sphincter, terminat- the pubic bone, along with the puborectalis, and
ing in the perianal skin. The external anal sphinc- from a tendinous arch formed by obturator fascia
ter (EAS) is a voluntary muscle arising from the running posteriorly towards the ischial spine.
levator ani and puborectalis muscle to form a Pubococcygeal fibers pass posteriorly towards the

Fig. IILl. Coronal anatomy of the anal

canal. The muscolaris propria layer of the
rectal wall consists of both circular and
longitudinal smooth muscle fibers. The
circular layer is in continuity with the cir-
cular internal anal sphincter muscle. The
longitudinal layer extends into the anal
canal where the muscle fibers mix with
external
.;>UIU ".""IIU'OUS
the external sphincter. The external
sphincter muscle sphincter extends further down than the
internal sphincter
28 Atlas of Endoanal and Endorectal Ultrasonography

coccyx in a plane just cranial to the iliococcygeus. rior aspect of the vagina. In males the perineal
Fibers also cross the midline to form rectal and body is smaller [1].
vaginal hiati. The puborectalis arises from the The area around the anorectum is divided into
pubis, making a distinct sling around the anorec- spaces. The perianal space surrounds the lower
tal junction (Fig. I1L4) [1]. anal canal. Lateral to the sphincter is the
The anatomy of the EAS remains controversial. ischiorectal fossa, which is bounded laterally by
It is described as having three parts (Fig. IlLS): the obturator internus muscle and superiorly by
the levator ani muscle. The intersphincteric space
1) the deep part joins with fiber from the pub- is not a recognized anatomical term but is used to
orectalis muscle and should be considered as describe the area between the internal and the
one anatomic muscle group; external sphincter. The supraelevator space lies
2) the superficial part is attached to the superfi- above the levator ani muscle and is demarcated
cial transverse perineii, anteriorly and is superiorly by the pelvic peritoneum.
inserted to the coccyx via the anococcygeal Endoluminal ultrasound defines anatomy of
ligament, posteriorly; the anal canal and pelvic floor. Five and possibly
3) the subcutaneous part lies below the internal six hypo echoic and hyperechoic layers can be
anal sphincter. seen (Fig. I1L6) [1]. From inner to outer, these are:

The anterior part of the external sphincter dif-

fers between the sexes. In women, the muscle 1) HYPERECHOIC: interface with the plas-
fibers unite in the inferior part anteriorly and so tic cone
imaging just above this level may show an appar-
ent defect. In men the external sphincter is sym- 2) HYPOECHOIC: mucosa
metrical at all levels.
The perineal body is a junctional zone where 3) HYPERECHOIC: subepithelial tissues
fibers from the internal sphincter, the external
sphincter, and the longitudinal muscle converge 4) HYPOECHOIC: internal anal sphincter
and fuse into muscles of the anterior urogenital
triangle to attach the anal canal to the ischial and 5) HYPERECHOIC: longitudinal muscle
pubic bones. The anatomy of the perineal body
differs between the sexes. In females a large block 6) MIXED ECHOGENICITY: external anal
of fibers from the internal sphincter and the lon- sphincter
gitudinal muscle run horizontally into the poste-

Fig. 111.2. The voluntary and involuntary

muscle of the anal canal (cadaveric speci-
men)
 Section 11/- Basic anatomy 29

Puborectalis -lilococcygeus muscles

Puborectalis muscle

a muscle
L-______________________________________________ ~

Anococcygeal Fig. I1I.3. Normal anorectal and pelvic

c ~ ________________________________________________________ ~

anatomy (a-c)
30 Atlas of Endoanal and Endorectal Ultrasonography

Fig. IlIA. The puborectalis muscle

swings around the anal canal at its most
proximal limit to encircle the sphincter
posteriorly

A thin hypo echoic layer representing the Therefore although the bulk of the sphincter may
mucosa may sometimes be seen. The hypoechoic increase with age, the number of functioning
layer that represents the internal anal sphincter muscle fibers decreases. Any internal anal sphinc-
can be traced superiorly into the circular muscle ter >4 mm thick should be considered abnormal
of the rectum. Its thickness varies from 1.5 to 4 whatever the patient's age; conversely a sphincter
mm (mean 3.5±0.5 mm) and increases with age of 2 mm is normal in a young patient, but abnor-
owing to the presence of more fibrous tissue as mal in an elderly one. In older age subjects the
the absolute amount of muscle decreases [2]. sphincter becomes thicker and loses its uniform

Fig. 111.5. Coronal diagram of the three

different anal canal levels
 Section 11/. Basic anatomy 31

low level echogenicity, which is characteristic of ultrasound imaging of the anus can be divided
smooth muscle throughout the gut, to become into three levels: deep, mid and superficial por-
more echogenic and inhomogeneous in texture tions [7-8]. With Bartram's system the level refers
[3]. The longitudinal muscle is 2.5±0.6 mm in to the following anatomical structures:
males and 2.9±0.6 mm in females. This muscle is
moderately echogenic, which is surprising as it is
mainly smooth muscle, however an increased
fibrous stroma may account for this. The average • DEEP: the sling of the puborectalis and
thickness of the EAS is 8.6±1.1 mm in males and the deep part of the external
7.7±1.1 mm in females, respectively. The thickness sphincter
of the lAS and the EAS should be measured at the
• MID: the anococcygealligament, superfi-
3 and 6 o'clock positions in the midlevel of the
cial part of the external sphincter,
anal canal. A postulated association between
internal sphincter, perineal body
manometric function of the sphincters and their and vagina in females
sonographic appearance remains controversial in
literature; however some authors have found no SUPERFICIAL: the subcutaneous part of
correlation between muscle thickness and muscle the external sphincter
performance, neither resting nor squeeze pres-
sure [4]. Scanning anal sphincter muscle may
allow for determination of their integrity, but not
for their morphometric properties. Vaginal deliv- The muscles of the lower and the upper part of
eries among nulliparous patients carry a substan- the anal canal are different. The lower part is sur-
tial risk of development of sphincter injuries and rounded by the internal anal sphincter, the longi-
a significant number of these injuries are unrec- tudinal muscle layer and the external anal sphinc-
ognized at the time of delivery [5-6]. Little is ter (Fig. 111.6), whereas the upper part is com-
known about the consequences of occult sphinc- prised of the internal anal sphincter, the longitu-
ter injuries. It was proposed that the young age of dinal muscle layer, and the puborectalis muscle
patients could compensate a defect in the EAS by (Fig. III.7). The latter muscle slings the anal canal
the action of the puborectalis and the pelvic floor instead of completely surrounding it. At its upper
muscle. Symptoms will develop later in life. The end the puborectalis muscle is attached to the
evaluation of anterior injuries in EAS on endolu- funnel-shaped levator ani muscle and the levator
minal ultrasound is sometimes difficult. ani anchors the sphincter complex to the inner
Bartram and Frudinger [I] established that side of the pelvis [9]. In some cases the external

Fig. 111.6. Normal ultrasound anatomy of

the anal sphincter
 32 Atlas of Endoanal and Endorectal Ultrasonography

I
'~ I.!
\~\\;., ~ ~)
)

'.1 / ..•.• i

\ \
." .;.
.,.' •
.' 6
.
. ...;

a b

Fig. III.7. Normal ultrasound anatomy of

the deep level of the anal canal in male (a)
and female (b) and schematic representa-
c tion (c)

sphincter is similar in echogenicity to the longitu- may be seen extending anteriorly. The perineal
dinal muscle of the rectum and so is indistin- body is seen as a complex structure of concentric
guishable from it. However, in most cases the rings with a hypo echoic or hyperechoic center
external sphincter is mainly hypo echoic and seen [10] and the transverse perineii are imaged at 11
as a concentric ring around the longitudinal mus- and 1 o'clock. The anoccoccygealligament is seen
cle. Anteriorly the circular fibers of the deep and as a posterior hypo echoic triangle.
superficial parts of the external sphincter are not The first ultrasound image recorded is nor-
recognizable. This is particularly true in females mally at puborectalis level, where the perineal
where the anterior bundle of the external sphinc- body is also seen in females. This image is nor-
ter is seldom detected, whereas in males thin arcs mally documented and labeled HIGH (Fig. III.?).
of muscle from the deeper parts of the sphincter In a normal patient, moving the probe a few mm
 Section 11/. Basic anatomy 33

TI . n,wt'l'Se pt'rilleii Coml>lt'te ring

Allococcygealligament

Fig. 111.8. Normal ultrasound anatomy

of the formation of the anterior ring (a)
b and schematic representation (b)

in the distal direction will show an intact anterior 3D endoanal ultrasound offers a significant
EAS forming just below the superficial transverse advantage over conventional2D ultrasound [11]. It
perineal muscles (Fig. 111.8). This image is a mid- makes it possible to project the anal canal in the
canal projection where the lAS, conjoining longi- entire proximal-distal anterior-posterior exten-
tudinal muscle, and the superficial EAS all are sion (Fig. 111.11). The data acquired can be dis-
identified. This image will be labeled MID (Fig. played in ways that makes it easier to abstract the
111.9). When the probe is pulled further out, the information. It is easy to manipulate the data cube
image of the lAS will disappear and only the to see the 3D image from different angles (Fig.
sub epithelium and the subcutaneous segment of I1I.12). The cube can be sliced at any point and at
the LM + EAS will be seen. This last image will be any angle, making it possible to see what is going
labeled LOW (Fig. 111.10). on in the exact location of interest.
 34 Atlas of Endoanal and Endorectal Ultrasonography

a b

Fig. III.9. Normal ultrasound anatomy of

the mid level of the anal canal in male (a)
and female (b) and schematic representa-
tion (C)
 Section 11/. Basic anatomy 35

a b

Fig. 111.10. Normal ultrasound anatomy

of the superficial level of the anal canal in
male (a) and female (b) and schematic
representation (c)
c

Fig. 111.11. Normal ultrasound anatomy

of the sphincter complex in 3D
 36 Atlas of Endoanal and Endorectal Ultrasonography

a b
Fig. m.12. Normal ultrasound anatomy of the sphincter complex in 3D in a 33-year old nulliparous female. Mid anal canal (a).
Low anal canal (b)

References 6. Sultan AH, Kamm MA, Hudson CN, Thomas JM,

Bartram CI. Anal-sphincter disruption during vaginal
1. Bartram CI, Frudinger A. Handbook of anal endosonog- delivery. N Engl J Med 1993;329:1905-11
raphy. Petersfield, UK: Wrightson Biomedical Publishing 7. Kumar A, Scholefield JH. Endosonography of the anal
Ltd,1997 canal and rectum. World J Surg 2000;24:208-15
2. Nielsen MB, Hauge C, Rasmussen 00, Sorensen M, 8. Hussain SM, Stoker J, Schutte HE, Lameris JS.
Pederson JF, Christiansen J. Anal sphincter size mea- Imaging of the anorectal region. Europ J Radio11996;
sured by endosonography in healthy volunteers. Effect 22:116-22
of age, sex and parity. Acta RadioI1992;33:453-6 9. Tjandra JJ, Milsom JW, Stolfi VM, Lavery I, Oakley J,
3. Burnett SJD, Bartram CI. Endosonographic variations Church J, Fazio V. Endoluminal ultrasound defines
in the normal internal anal sphincter. Int J Colorectal anatomy of the anal canal and pelvic floor. Dis Colon
Dis 1991;6:2-4 Rectum 1992,35:465-70
4. Enck P, Heyer T, Gantke B, Schmidt WU, Schafer R, 10. Zetterstrom JP, Mellgren A, Madoff RD, Kim DG, Wong
Frieling T, Haussinger D. How reproducible are mea- WD. Perineal body measurement improves evaluation
sures of the anal sphincter muscle diameter by of anterior sphincter lesions during en do anal ultra-
endoanal ultrasound? AJG 1997;92:293-6 sonography. Dis Colon Rectum 1998;41:705-13
5. Zetterstrom JP, Mellgren A, Jensen LL, Wong WD, Kim 11. Williams AB, Bartram CI, Halligan S, Marshall MM,
DG, Lowry AC, Madoff RD, Congilosi SM. Effect of Nicholls RJ, Kmiot WA. Multiplanar anal endosonog-
delivery on anal sphincter morphology and function. raphy- normal anal canal anatomy. Colorectal Dis
Dis Colon Rectum 1999;42:1253-60 2001;):169-74
 111.2.
Surgical and endosonographic
anatomy of the rectum
G.A. Santoro, G. Di Falco

The normal rectum is 11-15 cm long and has a na in the female (Fig. III.I4b). The rectal wall con-
maximum diameter of 4 cm. It is continuous with sists of 5 layers surrounded by perirectal fat (Fig.
the sigmoid colon superiorly at the level of the 111.15).
third sacral segment and courses inferiorly along On ultrasound the normal rectal wall is 2-3
the curve of the sacrum to pass through the mm thick and is composed of a five-layer struc-
pelvic diaphragm and become the anal canal ture [1]. There is some debate as to what the actu-
(Fig. III.I3). It is surrounded by fibrofatty tissue allayers represent anatomically. Hildebrandt and
that contains blood vessels, nerves, lymphatics Feifel [2] believe that three layers are anatomical,
and small lymph nodes. The superior one third is while the other layers represent interfaces
covered anteriorly and laterally by the pelvic between the anatomical layers. Beynon et aI, how-
peritoneum. The middle one third is only covered ever, have produced both experimental and clini-
with peritoneum anteriorly, where it curves ante- cal evidence that the five anatomic layers are rec-
riorly onto the bladder in the male and onto the ognizable. These five layers cannot be seen in all
uterus in the female. The lower one third of the patients and at all levels. Good visualization
rectum is below the peritoneal reflection and is depends on maintaining the probe in the center of
related anteriorly to the bladder base, ureters, the lumen of the rectum and having adequate dis-
seminal vesicles and prostate in the male (Fig. tension of the water-filled balloon with good
III.I4a) and to the lower uterus, cervix and vagi- acoustic contact with the rectal wall [3].

Ext 1ft I ""I

splln I

Fig. III.l3. Coronal anatomy of the

anorectum
38 Atlas of Endoanal and Endorectal Ultrasonography

The layers represent (Fig. III.16):

1) the hyperechoic interface between the water-filled balloon and the mucosa;

2) the hypoechoic deep mucosa (lamina propria + muscolaris mucosae);

3) the hyperechoic submucosa;

4) the hypo echoic muscolaris propria (in rare cases seen as 2 layers: inner circular and outer lon-
gitudinallayer);

5) the hyperechoic interface between the rectal wall and the perirectal fat tissue.

levatOI ilni I

Fig. III.14. Sagittal anatomy of the

L -__________~~~____~~__~________________~b
anorectum in men (a) and women (b)
 Section 11/. Basic anatomy 39

a b
Fig. III.15. Diagrammatic representation of five-layer structure of the normal rectal wall (a). Comparable histological section (b)

Fig. 1II.16. Normal ultrasound anatomy

of the rectal wall (a) and schematic repre-
b sentation (b)
40 Atlas of Endoanal and Endorectal Ultrasonography

Fig. 111.17. Sonographic view of blood vessels in two cross sections at different levels (arrows)

Fig. III.l8. Sonographic view of the rectum at the level of the seminal vesicles in a male patient

The ultrasonographer must have a clear The fibrofatty tissue surrounding the rectum
understanding of what each of these five lines contains blood vessels, nerves and lymphatics and
represent anatomically. When staging a rectal has an inhomogeneous echo pattern. Very small,
cancer, various levels of the tumor must be opti- 2- to 3-mm, round to oval hypo echoic lymph
mally imaged and the integrity of the lines care- nodes may be seen and must be distinguished
fully assessed [41. Attention must be focused on from blood vessels which are also circular hypoe-
the third hyperechoic layer. Once it has been choic areas but when followed longitudinally, they
ascertained that the middle hyperechoic line is seem to extend further than the corresponding
broken, then an invasive lesion is recognized and diameter and can often be seen to branch and to
attention is then turned to the thickness of the elongate in a longitudinal fashion, confirming that
muscolaris propria and the integrity of the outer this is a blood vessel and not a node (Fig. III.ll).
hyperechoic line to see if the perirectal fat is Anteriorly the bladder, seminal vesicles (Fig.
invaded. III.I8) and prostate can be identified in the male
 Section 11/. Basic anatomy 41

Fig. III. I 9. Sonographic view of the rectum at the level of the Fig. III.20. Sonographic section through the distal rectum in
prostate in a male patient (arrow) a female patient showing the vagina anteriorly (arrows)

Interface. hyperechoic
2 layers. hypoechoic
(mucosae)
~t .ft.'.
Itllt,
----a.....
~ . ,'.~,,~p,
'IJ !,
'. ;' ;,
. ';lp..,~
'~ .

. '/'-'. ' ....

.
(
}
.)
'!

1 layer. hyperechoic ~h~

(submucosa) 7J/f:0l"
2 layers. hypoecholc .............~ , : t
(muscolarrs propr ia) ~~. / l l. '. ,~~'
.: ~ c.. \'
1 layer. hypoechoic ~ \'i? '. . .
(perrrectal fat t .. ~ )~ .,". 11 ,/
-./ .., t. i ( •.\-. Fig. m.21. Normal ultrasound anatomy
of the rectal wall in 3D

(Fig. 111.19) and the uterus, cervix and vagina in of the rectal wall are clearly illustrated in the coro-
the female (Fig. 111.20) [5]. nal plane as well as in the transaxial and the lon-
3D rectal ultrasound offers a valuable supple- gitudinal image planes (Fig. m.21).
ment to conventional ultrasound [6]. The 5 layers

References canal and rectum. World J Surg 2000; 24:208-15

4. Hildebrandt U, Feifel G. Preoperative staging of rectal
1. Bartram CI, Frudinger A. Handbook of anal endosonog- cancer by intrarectal ultrasound. Dis Colon Rectum
raphy. Petersfield, UK: Wrightson Biomedical Publishing 1985;28:42-6
Ltd,1997 5. Hussain SM, Stoker J, Schutte HE, Lameris JS. Imaging
2. Hildebrandt U, Feifel G, Schwarz HP, Scherr O. of the anorectal region. Europ J RadioI1996;22:116-22
Endorectal ultrasound: instrumentation and clinical 6. Hunerbein M, Schlag PM. 3D-endosonography for
aspects. Int J Colorectal Dis 1986;1:203-7 staging of rectal cancer. Ann Surg 1997;225:432-8
3. Kumar A, Scholefield JH. Endosonography of the anal
 Invited commentary
C.Ratto

The imaging of pelvic floor structures is presently and rectum, being (together with the simplicity of
of great interest. In the last two decades increasing use, comfort for patients, and low costs) one of the
attention has been dedicated in order to both most interesting characteristics of this kind of
increase knowledge on the pelvic floor anatomy investigation. Particularly in the last decade, due
(possibly related to physiology and pathophysiolo- to their diagnostic possibilities, an increasing
gy) and improve technologies for diagnosis. In par- interest toward the EAUS and ERUS has been reg-
ticular, ultrasonography (US) and magnetic reso- istered, and the wider use of these procedures has
nance (MR) have acquired the right to be the most allowed precision in clinical indications, enlarg-
useful procedures in pelvic floor imaging. Their ing the fields of application.
contributions should be effectively integrated with The ambitious aim of this diagnostic tool is to
other diagnostic steps (Le., endoscopy, anorectal correctly identify very small and thin structures,
manometry and electrophysiology, defecography) which have no precise interfaces and limits
to have a complete assessment of neoplastic and toward the adjacent other structures, being often
functional anorectal diseases. This integration will very difficult to measure. Moreover, there is a wide
permit planning the best treatment for patients. anatomic variability between different subjects,
This approach has been demonstrated to be very depending on sex, age, ultrastructural character-
useful for colorectal surgeons, particularly if the istics, and a number of other not well-known fac-
colorectal surgeon himself is directly implicated in tors. Consequently, significant confusion and con-
the diagnostic process. In this regard, it has been trast have been registered for a long time [3]. Both
well demonstrated that the anatomic knowledge of US and MR have contributed to the modifications
the anorectal region and the ability in performing of previous, approximated knowledge about the
anorectal operations are superior for specialized anorectal anatomy and adequately correlate the
surgeons than for general surgeons. US diagnostic imaging with the pelvic floor diseases [4-6].
procedures are well suited to both daily clinical Significant improvement in minimizing investiga-
work and research in surgical proctology. tional problems has been more recently obtained
Of course, the first problem encountered by by using more sophisticated devices (Le., 3D
users of endoanal ultrasound (EAUS) and acquisition systems, probe pull through systems,
endorectal ultrasound (ERUS) has been to believe and newer probe with integrated 3D and pull
in the reliability of US imaging. The intraobserver through devices) which allow to evaluate the anal
and interobserver agreement are statistically stat- canal and rectum in a variety of projections,
ed as "good" or "very good" [1], even if measure- including the axial, longitudinal and all the possi-
ments of each anal structure do not provide ble diagonal views. Methods for measuring not
homogeneous morphometric results [2]. The ulti- only linear distance or thickness, but also volume
mate technology, using a 36o-degree endoprobe, are now easily available.
has shown fascinating features in delineating However, considering both diagnostic poten-
accurately the anatomical structures of the anus tials and problems of EAUS and ERUS, it is actu-
 Section 11/. Basic anatomy 43

ally mandatory to standardize the technique of dure is able to give a quantity of useful clinical
examination, as well as interpretation criteria, information, but can offer only indirect, not very
modalities in taking images (showing visible and reliable computerized reconstruction of the canal,
clear documentation of the diagnosed disease based on the registration of the resting pressure
also by other doctors), and criteria in reporting (produced by the resting tone of the whole anal
the examination results. canal, not only the lAS). On the other hand, the
endosonographic 360-degree imaging of the lAS
is realistic, its modifications being well correlated
Endoanal ultrasound (EAUS) to functional anorectal disorders. In a large num-
ber of them, an integrated evaluation with EAUS
Accuracy of EAUS in correct definition of and anorectal manometry (and, if necessary, with
anatomic layers of the anal canal has been con- electrophysiologic examination), is mandatory,
firmed in autoptic studies [7] as well as during which may explain both structural and physiolog-
surgery [4]. Concerning the single structures of ic characteristics of the anal canal at different lev-
the anal canal, the importance of an accurate def- els (Fig. III.22).Although it seems normal, without
inition of the internal anal sphincter (lAS) by the any lesion, sometimes lAS has differences in
endoanal ultrasound has been very well illustrat- echogenicity and thickness. Recently, an increas-
ed by Drs. Santoro and Di Falco. It is a very well ing interest in lAS degeneration has been regis-
delineated circular muscle along the major part of tered. In such conditions (firstly described with
the anal canal, and represents a useful landmark EAUS), the lAS appears as intact but thinner than
for the correct orientation between the different normal and hyperechoic; it has been regarded as a
anatomical entities, internally and externally to cause of passive fecal incontinence [8]. Vaizey et
the lAS. This structure, always rightly regarded as al. [8] reviewed the EAUS examinations of 38
fundamental in the physiology of defecation and patients with passive fecal incontinence and intact
continence, has been traditionally studied with anal sphincters. All patients had ultrasonographic
anorectal manometry. Actually, the latter proce- appearance of lAS degeneration as described

superior third I

inferior third \
\
\

Fig. 111.22. Integrated anal canal evaluation in a normal subject with EAUS and anorectal manometry
44 Atlas of Endoanal and Endorectal Ultrasonography

above, combined with reduced resting pressure only the tumor dimension. A number of non-con-
and normal squeeze pressure, rectal sensitivity cordant EAUS-based staging systems have been
and pudendal latency. Incontinent patients with proposed. Boman et al. [13] and Goldman et al.
lAS degeneration were older than those with [14] classified as T2, tumors infiltrating lAS or
obstetric trauma incontinence [9]. An apparently EAS (they staged as Tl tumors confined to the
opposite US condition is the abnormal increased subepithelium). Recently, a more useful classifica-
thickness of the lAS. It seems typical for older tion has been proposed by Tarantino and Berstein
ages [9-11] without differences of anal canal levels [12], in which tumors limited to the subepithelium
[9]. Interestingly, decreased thickness of lAS can are staged as Tl, cancers infiltrating only the lAS
be frequently observed in patients with chronic as T2a, and those extended to the EAS as T2b.
anal fissure, sign of an increased sphincter tone. One of the most important contributions of
Imaging of an lAS break following internal EAUS has been in the correct imaging of the EAS,
sphincterotomy could help to monitor the clinical in particular differentiating males and females.
results of the operation or the unexpected seque- Indeed, while in males there are no significant dif-
lae (persistence of anal fissure and pain, or, on the ferences in the length of the sphincter between the
other hand, fecal incontinence). Moreover, lAS has anterior and posterior side, in females it has been
been accurately investigated. with EAUS in cases very well documented that a shorter EAS exists on
of anal fistula or abscess (particularly in order to the anterior side than on the posterior side. This
define the relationship between the sphincter and latter feature is particularly frequent in middle-
the fistula track, and, then, plan the treatment). old age female subjects, especially if multiparous.
Some surgical operations could produce detri- Three-dimensional longitudinal imaging can
mental effects on the lAS; obstetric injuries can offer suggestive features of these anatomic char-
involve not only the external anal sphincter acteristics of EAS. The most difficult characteris-
(EAS), but also the lAS; during fistula surgery a tic of EAS to be assessed is the outer border, and
too much wider fistulotomy could be performed, then, there could be disagreement in measuring
damaging the lAS (and EAS also, eventually); the the sphincter thickness. However, in Frudinger et
incorrect use of the stapler during anterior resec- al.'s study [9] the EAS thickness was difficult to
tion of the rectum for cancer, or muco-prolassec- define in only 2% of patients at all the three levels
tomy (Longo's operation) for hemorrhoids or examined, and in 3% at the subcutaneous level
obstructed defecation syndrome could also cause only. A significant negative correlation with the
an lAS tear. In other cases fecal incontinence fol- patients'age was also demonstrated in this study,
lows an anterior resection, but EAUS is able to at all anal canal levels. In particular, the anterior
show the absence of any lAS defect suggesting EAS part was found significantly thinner in older
redirectioning the investigations toward pelvic subjects. EAS atrophy was first described by
neurogenic disorders. Concerning the main argu- endoanal MR as thinner sphincter (with replace-
ment of this publication, EAUS shows better diag- ment of muscle by fat) and correlated with idio-
nostic sensibility when compared to computer- pathic fecal incontinence [15-18]. Features shown
ized tomography, magnetic resonance, and digital by EAUS can help to differentiate incontinent
examination. Tarantino and Bernstein [12] found patients with intact EAS and those with sphincter
a 100% agreement between EAUS and surgery in lesions, identifying number, site and extension of
patients undergoing anal cancer excision. In our the defects. The EAS tears, very often iatrogenic,
experience, EAUS is the first line diagnostic are the most frequent causes of fecal inconti-
modality to stage the tumor. Being most patients nence. Today, an increasing number of therapeu-
addressed to neoadjuvant chemoradiation thera- tic options are available for patients suffering
py plus interstitial brachytherapy (in the case of from this disabling condition. However, the best
neoplastic residuals), we use EAUS also for restag- results can be obtained only if a correct decision
ing, and in patients with complete response, for making process is followed, and, in this regard,
follow-up evaluation. The EAUS relationship EAUS has a key role, so that, it will contribute in
between lAS and anal canal tumor should be pre- choosing sphincteroplasty or graciloplasty or
cised in order to formulate a correct staging and artificial anal sphincter or other less invasive
establish the treatment schedule. However, classi- therapeutic procedures. Moreover, patients oper-
fication criteria are conflicting. TNM classifica- ated on could be re-evaluated during the follow-
tion does not consider such relationships, but up for monitoring the results of surgery. However,
 Section 11/- Basic anatomy 45

these diagnostic possibilities displayed by the functionally the anterior from posterior per-
EAUS could overestimate the occurrence of EAS ineum; perineal body lesions could predispose to
lesions because of the peculiar conformation of anterior or posterior perineal prolapse of the
the EAS in females confirmed in EAUS [4, 7], and pelvic organs. Although EAUS is able to precise
MR studies [5].Also anatomic studies detailed the differences of perineal body between males and
lower length of ventral EAS in females, showing females, it remains difficult to reliably measure
that the differences with males were already pre- this structure because of the lack of clear limits.
sent in fetal age [191. In examining a female sub- Also the proposed use of a finger introduced into
ject, the ultrasonographic differences between the the vagina as a landmark seems to be of poor ben-
natural gaps (hypo echoic areas with smooth, reg- efit, altering its normal configuration and the
ular edges) and the sphincter ruptures (mixed measurement due to the digital compression on
echogenicity, due to scarring, with irregular the central perineum.
edges) occurring at the upper anterior part of the The perianal spaces, known as ischioanal
anal canal must be kept in mind [20]. Bolland et spaces, are constituted by the soft surrounding tis-
al. [20] found manometritic confirmation of the sue, which result as an ultrasonographic inhomo-
EAS gaps identified by EAUS. Differences between geneous area around the anal canal. Inside this
genders have also clinical implications in diagno- space septic fluids could be identified as hypoe-
sis of anal fistulas involving the EAS on the ante- choic areas; primary and secondary fistula exten-
rior anal canal and, consequently, planning a sions are visualized as branched hypo echoic
sphincter-saving surgical procedure. tracks; infiltration of an anal cancer beyond the
Neoplastic infiltration of the EAS can be accu- EAS can be detected, staging the tumor as uT3,
rately assessed by EAUS for tumor staging. As for eventually with isoechoic enlarged lymph nodes
the lAS involvement, the criteria to define infiltra- suspected to be metastasized.
tion of the EAS has been also differently proposed Supralevator space can be visualized by EAUS,
by various Authors; however, in all staging sys- allowing the identification of normal or patholog-
tems the assessment of tumor spread within or ic processes surrounding the recto-anal junction.
beyond EAS is of greatest importance. Both This is a crucial site for extension of neoplastic
Boman et al. [13] and Goldman et al. [141 classified infiltration from anal or low-rectal cancers; this
tumors within EAS as T2, while those spreading condition can be investigated and detected by
beyond it as T3, while Tarantino and Bernstein EAUS, or ERUS, or both in a combined examina-
[121 distinguished these two conditions as T2b tion. Also diffusion of acute sepsis and extension
and T3, respectively. of fistulas in complex configurations (Le., extras-
The intersphincteric space, in which the longi- phincteric and suprasphincteric) can be observed:
tudinal muscle is located, presents a wide variabil- EAUS could be useful to accurately delineate such
ity in thickness, and is not always distinctly visible conditions.
along the entire anal canal. However, the ability to EAUS has been criticized because believed
identify this structure in normal subjects, differ- highly operator dependent and for its limited field
entiating it from the lAS and EAS, could be useful of view and contrast resolution, which would
in assessing diseases (anal cancer, anal fistula/ make it less useful for pelvic floor muscle imaging;
abscess) involving the intersphincteric space. so, MR has been suggested as a better diagnostic
The necessity to identify and assess the per- procedure. However, differences in definition of
ineal body with the EAUS has been well under- anal canal anatomy have been described in rela-
lined by Drs. Santoro and Di Falco. This is a cen- tion to the technique used. The endoanal coil has
tral structure of the perineum where a variety of been used for a long time [5,15,21,221, but recent-
muscle fibers are confluent. Such anatomic config- ly it has been believed to distort the anatomy and
uration gives a fundamental function of support a phased-array technique has been preferred [231.
of all muscolo-ligamentous components of the However, with this procedure all the main features
pelvis to the perineal body anchoring the anal of the anal canal morphology showed with EAUS
canal to the pelvic bones. The lack of these con- are similarly confirmed: good resolution of the
nections could be one of the most important rea- lAS; shorter EAS at the anterior anal canal in
sons of a perineal descent, determining also dam- females; not precised subdivision of the EAS into
ages to the pelvic nerves (Le., pudendal nerves) two or three parts; difficulty to measure perineal
and muscles (Le., levator ani), and disconnecting body. The only significant advantage of phase-
46 Atlas of Endoanal and Endorectal Ultrasonography

array MR over endoanal MR and EAUS is the wall by the development of edema, fibrosis, or
imaging of a wider field-of-view. Considering micro abscess. Primary poor inflammation of the
technical characteristics, time-consuming, costs, rectal mucosa can alter ultrasonographic interpre-
and availability of the instruments in Hospitals, in tation.
our opinion MR should be used in cases of clinical The main diagnostic advantage in using
complexity, when EAUS is unable to give reliable ERUS is the realibility in distinguishing the first
information. hypoechoic layer (interface between the balloon
and the mucosa) from the outer hypoechoic lam-
ina propria / muscolaris mucosae, and the latter
Endorectal ultrasound (ERUS) one from the hyperechoic submucosa. Indeed, a
number of cancerous adenomas or early rectal
Over the last two decades an increasing interest for cancers need to be correctly staged and
ERUS in diagnosis and treatment of rectal cancers addressed to an adequate treatment. Therapy has
has been registered. The realibility of ERUS find- to be radical but, possibly, as less destructive as
ings in identifying the layers of the normal rectum possible. So that, a correct staging of an adenoma
has played a determinant role in the correct defi- or a very early carcinoma will be treated with
nition of local tumor spread. Consequently, endoscopic resection, while patients with carci-
endoscopy and computerized tomography (CT) noma non-infiltrating the bowel beyond the sub-
have progressively lost some roles in the diagnos- mucosa will undergo a transanal excision (if the
tic process. Only MR is actually able to compete in tumor is located at the middle-lower rectum) or
accuracy of tumor staging [24]. Technology an anterior resection (in case of upper rectal
progress has permitted to obtain an increased res- tumor). The hypoechoic muscolaris propria is an
olution of the five distinct layers of the rectal wall, ultrasonographically well visible layer: its neo-
and more and more detailed US staging systems plastic involvement could be assessed accurately.
are being proposed, following the TNM system. However, a crucial problem is to distinguish
The 360-degree imaging is actually the gold stan- tumors limited to the muscolaris propria from
dard because the panoramic view of the rectum those spreading beyond it, invading the perirec-
and the perirectal tissues and organs significantly tal fat (and, eventually, adjacent organs). In the
helps to identify relationships between structures. former condition surgery is indicated, but in the
Moreover, the 3D image acquisition system allows latter situation patients should benefit if they
the evaluation and measurement of the normal receive neoadjuvant radiation or chemoradiation
rectal wall layers and tumor spread in a variety of therapy. The effects of these treatments usually
projections, increasing the accuracy of examina- significantly alter endosonographic morphology
tion. Using ERUS, the accuracy for the staging of of the rectal wall, which becomes thickened,
the tumor infiltration ranges between 64 and 95%, more hypoechoic, with poorer definition of each
while that of the lymph node metastases from 73 to layer. Consequently, tumor restaging by ERUS
83%· after neoadjuvant therapies could be more diffi-
Despite these very good results, the loss in cult [27]. Finally, ERUS can be very useful in the
accuracy could be due to a few very important patients' follow-up, showing a very good accura-
technical problems, with consequent distortion of cy rate in early detection of local recurrences, in
rectal wall morphology [25, 26]. Filling the con- asymptomatic patients [28]. Lohnert et al. [28]
dom covering the probe with too much water will reported better results using ERUS compared
decrease the possibility of separately distinguish- with tumor markers, digital examination and
ing each layer. Also the choice of the US frequency rectos copy. The CT scan has proven to have a
is significantly related to the quality of imaging. sensitivity ranging between 40 to 100%, but the
An incorrect angulation of the probe will result in local recurrence must be at least 2 em in diame-
a possible misinterpretation of the images. The ter, and can be confused with fibrosis or inflam-
presence of air inside the condom or between the mation; metal clips can generate artifacts. MR
condom and the rectal wall, as well as residual could be conditioned by these metal clips. It has
feces in the rectal ampulla could create sono- an accuracy of 75 to 93% and a sensitivity of 91 to
graphic artifacts. Sometimes previous procedures 100%. For both CT and MR costs and possible
performed during endoscopic assessment (Le., repeated examinations are the main problems for
biopsy, polipectomy) can alter the normal rectal routine use.
 Section 11/. Basic anatomy 47

References 15. DeSouza NM, Puni R, Zbar A, Gilderdale DJ, Coutts

GA, Krausz T. MR imaging of the anal sphincter in
1. Gold DM, Halligan S, Kmiot WA, Bartram CI. multiparous women using an enoanal coil: correlation
Intraobserver and interobserver agreement in anal with in-vitro anatomy and appearances in fecal incon-
endosonography. Br J Surg 1999;86: 371-5 tinence. AJR Am J RoentgenoI1996;167:1465-71
2. Enck P, Heyer T, Gantke B, Schmidt WU, Schafer R, 16. Briel JW, Stocker J, Rociu E, Lameris JS, Hop WC,
Frieling T, Haussinger D. How reproducible are mea- Schouten WR. External anal sphincter atrophy on
sures of the anal sphincter muscle diameter by endoanal magnetic resonance imaging adversely
en doanal ultrasound? Am J Gastroenterol 1997;92: affects continence after sphincteroplasty. Br J Surg
293-6 1999;86:1322-7
3. Dalley AF. The riddle of the sphincters. The morpho- 17. Williams AB, Bartram CI, Modhwadia D, et. AI. Endocoil
physiology of the anorectal mechanism reviewed. Am magnetic resonance imaging quantification of external
Surg 1987;53:298-306 sphincter atrophy. Br J Surg 2001;88: 853-9
4. Sultan AH, Kamm MA, Hudson CN, Thomas JM, 18. Williams AB, Malouf AJ, Bartram CI, Halligan S, Kamm
Bartram CI. Anal-sphincter disruption during vaginal MA, Kmiot WA. Assessment of external anal sphincter
delivery. N Engl J Med. 1993;329:1905-11 morphology in idiopathic fecal incontinence with
5. Hussain SM, Stoker J, Lameris JS. Anal sphincter com- endocoil magnetic resonance imaging. Dig Dis Sci
plex: endoanal MR imaging of normal anatomy. 2001;46:1466-71
Radiology 1995;197:671-7 19. Fritsch H, Brenner E, Lienemann A, Ludwikowski B.
6. Schafer A, Enck P, Furst G, Kahn T, Frieling T, Lubke HJ. Anal sphincter complex: reinterpreted morphology
Anatomy of the anal sphincters. Comparison of anal and its clinical relevance. Dis Colon Rectum 2002;
endosonography to magnetic resonance imaging. Dis 45:188-94
Colon Rectum 1994;37:777-81 20. Bollard RC, Gardiner A, Lindow S, Phillips K, Duthie
7. Sultan AH, Kamm MA, Talbot IC, Nicholls RJ, Bartram GS. Normale female anal sphincter: difficulties in inter-
CI. Anal endosonography for identifying external pretation explained. Dis Colon Rectum 2002;45:171-75
sphincter defects confirmed histologically. Br J Surg 21. Hussain SM, Stoker J, Zwamborn AW, et al. Endoanal
1994;81:463-5 MR imaging of the anal sphincter complex: correlation
8. Vaizey CJ, Kamm MA, Bartram CI. Primary degenera- with cross-sectional anatomy and histology. J Anat
tion of the internal anal sphincter as a cause of passive 1996;189:677-82
faecal incontinence. Lancet 1997;349:612-5 22. Rociu E, Stoker J, Eijkemans MJC, Lameris JS. Normal
9. Frudinger A, Halligan S, Bartram CI, Price AB, Kamm anal sphincter anatomy and age- and sex-related vari-
MA, Winter R. Female anal sphincter: age-related dif- ations at high-spatial-resolution endoanal MR imag-
ferences in asymptomatic volunteers with high-fre- ing. Radiology 2000;217:395-401
quency endoanal US. Radiology 2002;224:417-23 23. Morren GL, Beets-Tan GH, van Engelshoven MA.
10. Burnett SJ, Spence-Jones C, Speakman CT, Kamm MA, Anatomy of the anal canal and perianal structures as
Hudson CN, Bartram CI. Unsuspected sphincter dam- defined by phase-array magnetic resonance imaging.
age following childbirth revealed by anal endosonog- Br J Surg 2001;88:1506-12
raphy. Br J RadioI1991;64:225-7 24. Maier A, Fuchsjager M. Preoperative staging of rectal
11. Nielsen MB, Rasmussen 00, Pedersen JF, Christiansen cancer. Eur J RadioI2003;47:89-97
J. Anal endosonographic findings in patients with 25. Akasu T, Sugihara K, Moriya Y, Fujita S. Limitations
obstructed defecation. Acta RadioI1993i34:35-8 and pitfalls of transrectal ultrasonography for staging
12. Tarantino D, Bernstein MA. Endoanal ultrasound in of rectal cancer. Dis Colon Rectum 1997;40 (Suppl.)
the staging and management of squamous-cell carci- SlO-S15
noma of the anal canal: potential implications of a new 26. Kumar A, Scholefield JH. Endosonography of the anal
ultrasound staging system. Dis Colon Rectum 2002 canal and rectum. World J Surg 2000;24:208-15
Jan;45:16-22 27. Napolean B, Pujol B, Berger F, Valette PJ, Gerard JP,
13. Boman BM, Moertel CG, O'Connell MJ, Scott M, Souquet Jc. Accuracy of endosonography in the stag-
Weiland LH, Beart RW, Gunderson 11, Spencer RJ. ing of rectal cancer treated by radiotherapy. Br J Surg
Carcinoma of the anal canal. A clinical and pathologic 1991;78:785
study of 188 cases. Cancer 1984;54:114-25 28. Lohnert MSS, Doniec JM, Henne-Bruns D.
14. Goldman S, Norming U, Svensson C, Glimelius B. Effectiveness of endoluminal sonography in the iden-
Transanorectal ultrasonography in the staging of anal tification of occult local rectal cancer recurrences. Dis
epidermoid carcinoma. Int J Colorect Dis 1991;6:152-7 Colon Rectum 2000;43:483-91
 SECTION IV
Endoluminal ultrasound
in the preoperative staging
of rectal carcinoma
 IV.I.
Introduction
G.A. Santoro, G. Di Falco

Staging of rectal cancer has assumed new mean- The American Joint Committee on Cancer
ing and importance in recent times. Dukes' classi- (AJCC) and the Union International Contre Ie
fication (1932) remains the most commonly used Cancer (VICC) have developed a universal staging
staging system today [1]. It was intended to pro- system for all anatomic sites. The TNM system
vide a pathologic basis for prognosis in patients emphasizes the depth of invasion (Fig. IV.I), the
with rectal cancer. Rectal lesions were divided into number and site of metastatic lymph nodes and
three categories: the presence or absence of distant metastases:

A: cancer confined to rectal wall. No lymph Primary tumor (T)

node metastases
Tx: primary tumor cannot be assessed
B: cancer extends into perirectal tissues. No TO: no evidence of primary tumor
lymph node metastases TI: tumorinvadessubmucosa
T2: tumor invades muscolaris propria
C: lymph node metastases present T3: tumor penetrating through the musco-
laris propria to involve the perirectal fat
T4: tumor perforates the visceral peri-
In 1954, Astler and Coller [2] modified Dukes' toneum or directly invades other organs
classification by subdividing B lesions into two cat- or structures
egories depending on whether or not the tumor
penetrated through the muscolaris propria, and C Regional lymph nodes (N)
cases into those without full-thickness wall pene-
tration and those with complete penetration: Nx: regional lymph nodes cannot be
assessed
NO: no regional lymph nodes
A: limited to mucosa NI: metastasis in 1 to 3 pericolic or perirec-
tal lymph nodes
BI: into muscolaris propria but not penetrating N2: metastasis in 4 or more pericolic or
through perirectal lymph nodes

B2: penetrating through muscolaris propria Distant metastases (M)

CI: confined to wall with lymph node metas- Mx: presence of distant metastases cannot be
tases assessed
MO: no distant metastases
C2: penetrating wall with lymph node metas- MI: distant metastases
tases
52 Atlas of Endoanal and Endorectal Ultrasonography

Endorectal ultrasonography is an effective tool

for preoperative staging of rectal cancer. This is
important as it may provide useful prognostic
information by accurately assessing the local
depth of invasion and involvement of mesorectal
lymph nodes, the two most important prognostic
factors. It also may influence the choice of opera-
tion and identify cases suitable for techniques of
local excision such as trans anal endoscopic
microsurgery (TEM) and patients with advanced
local disease who should be considered for preop-
erative radiotherapy or chemoradiation to down
stage the disease. Endorectal ultrasonography can
be used to follow-up patients after rectal cancer
resection, for early detection of local recurrences. Fig. IV.I. The TNM staging emphasizes the depth of invasion
The ultrasonic staging of rectal cancer, pro- (T)
posed by Hildebrandt and Feifel, corresponds to
the TNM classification as all anatomic layers of
the rectum can be imaged (Figs. IV.2-6) [3]. The
prefix "u" should be used to indicate ultrasonic phy in assessing the level of rectal wall invasion
staging: was 69%, with 18% of the tumors overstaged and
13% understaged. The positive predictive value
was 72% and the negative predictive value was
uTO: villous adenoma 93%. There are several drawbacks to the use of
this technique in preoperative staging of rectal
uT!: tumor confined to the submucosa
with an intact bright middle hypere-
choic layer

uT2: tumor invading the muscolaris pro-

pria with no disruption of the third
hyperechoic layer

uT3: tumor penetrating through the mus-

colaris propria to involve the perirec-
tal fat. The tumor edge is usually irreg-
ular with sawtooth projection disrupt-
ing the third hyperechoic layer

uT 4: tumor invading an adjacent structure

uNO: lymph node metastases absent

uNl: lymph node metastases present

Most of the studies have reported the accuracy

rates of ERUS in the evaluation of rectal tumor
invasion to be 81-94% (Table IV.l) [4-20].
Overestimation occurred in 10% of cases and
underestimation occurred in 5% of cases. In a Fig. IV.2. uTO rectal tumor (villous adenoma). The first
hypoechoic layer representing the mucosa is thickened. The
recent study from the University of Minnesota second hyperechoic layer, representing the submucosa is seen
[15], overall accuracy of endorectal ultrasonogra- to be continuous beneath the lesion (arrows)
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 53

Fig. Iv'3. UTI rectal tumor. Note the preservation of the mid-
dle hyperechoic line of the submucosa

Fig. IV.S. uT3 rectal tumor. The arrows show areas of invasion
of the perirectal fat

Fig. IV,4. uT2 rectal tumor. The surrounding hyperechoic Fig. IV.6. uT 4 rectal tumor. The arrows show disruption of the
layer corresponding to the perirectal fat interface remains hyperechoic layer corresponding to the perirectal fat interface
intact (arrows) and spread beyond the rectal wall

tumors. Principally, it has been accepted that 1) It may be very difficult to differentiate between
interpretation of ultrasonograms, owing to its an adenoma and a very early cancer involving
dynamic nature, is highly operator-dependent only mucosa. Adenoma and Tl lesions are
and consequently fraught with a learning curve. affected by biopsy-taking, resulting in some-
Apart subjective factors, there are certain clinical what lower staging accuracy for this group of
situations that limit the reliability of ERUS in the rectal neoplasia.
assessment of rectal lesions [11,21,22]. There are 2) Overstaging of tumor invasion depth is caused
many pathologic clinical causes of error in rectal mostly by tumor invasion close to the deeper
cancer staging using ERUS: uninvolved layer, inflammatory cell aggrega-
54 Atlas of Endoanal and Endorectal Ultrasonography

Table IV.I. Accuracy of endorectal ultrasonography (ERUS) in determining the depth of tumor invasion in the rec-
tal wall

Author Sensitivity Specificity Positive Negative Overstaged Understaged

predictive predictive
value value

Herzog [4] 98.3% 75% 89·2% 95·4% 10.2% 0.8%

Sentovitch [5] 79% UT2 - 81% UT2 - 88% 17% 4%
#. ,
I}'
uT3/T4 -90% uT3/T 4 -100%
Anderson [6] pTI-70% pTI-100% UTI-100% UTl- 85% 15% 30%
pT2 -100% pT2-44% UT2 - 55% UT2 -100%
Harnsberger [7] 55% 40% 5%
Glaser [8] pT3 - 97% pT3 - 90% 90% pT3 - 98%
Holdsworth [9] 96% 50% 87% 80%
Orrom [10] 75% 22% 3%
Hulsmans [11] 97% 24%
Yamashita [12] 78% 17%
Dershaw [13] 50% 90% 75% 25%
Beynon [14] 97% 92%
Garcia-Aguilar [15] 72% 93% 18% 13%
Akasu [16] early pTI - 99% early pTI - 74% early pTI - 97% early pTI - 87%
deep pTI - 98% deep pTI - 88% deep pTI - 98% deep pTI - 88%
pT2 - 97% pT2 - 93% pT2 - 97% pT2 - 90%
pT3 - 96% pT3 - 83% pT3 - 85% pT3 - 96%
Jochem [17] pT3 - 92% pT3 -76%
Sailer [18] pTO/TI - 81% pTo/Tl- 98%
pT2 - 41% pT2 - 92%
Kim [19] pTO - 83.1% pTO - 96.5% pTO - 27% pTO -11.2%
pTl- 93.3% pTl- 97.9% pTl-lp% pTl- 5.6%
Santoro [20] 82% 93% 89% 94% 8% 3%

tion, desmoplastic change and hypervasculari- are most prone to overestimation (uT3) that
ty around the tumor mimicking tumor inva- results in a rather poor sensitivity for these
sion on ERUS. This is because the echogenici- tumors, with far fewer than one-half of all T2
ty of the tumor is similar to that of the musco- cancers being staged correctly. Taking into
laris propria and inflammatory infiltrate. consideration the extremely subtle ultrasonic
Peritumoral inflammatory changes, especially reflex of 1 mm or less delineating the layers of
following biopsies, simulate infiltration of the mucosal submucosa (Tl) or the muscolaris
deeper structures that can lead to overstaging. propria (T2), it is not surprising that any dis- .
The accuracy of ERUS in assessing the depth turbance in the bowel wall architecture caused
of rectal wall invasion varies with tumor stage. by the physiologic inflammatory reaction fol-
According to Sailer et al. [18], T2 carcinomas lowing biopsies, can impede the correct inter-
 Section IV· Endoluminal ultrasound in the preoperative staging of rectal carcinoma 55

pretation of neoplastic infiltration. On the especially useful for the visualization of small
other hand, once the tumor has penetrated and early-stage cancers or peduncolated
into the perirectal tissue (T3), inflammatory tumors.
changes do not alter endosonographic staging
because all relevant ultrasonic structures have Understaging of a tumor is mostly the result of
already been breached. Whereas the ultra- microscopic invasion beyond the estimated layers.
sonographic involvement of adjacent organs Understaging occurs even if the depth of penetra-
such as the prostate gland, seminal vesicles, the tion is correctly estimated but the presence of
vagina or bladder could again be the result of nodal disease is undetected.
either malignant infiltration or peritumoral Overstaging of rectal tumors clearly has major
inflammation. The inability of ERUS to distin- clinical consequences. When a tumor is under-
guish between the two results again in a some- staged and the patient undergoes local excision,
what lower staging accuracy with regard to T4 additional surgery can be performed for complete
cancers. rectal excision if necessary. However overstaging
3) Difficulties in endosonographic examination of Tl tumors as T2 may result in patients under-
occur where the tumor is close to the anal going complete rectal excision when local exci-
canal or the tumor is adjacent to one of the sion may have been adequate management. For
valves of Houston. this reason, a useful modification of the endorec-
4) Polypoidal tumors and bulky tumors that do tal ultrasonographic staging system was proposed
not lie within the focal length of the transduc- to reduce errors in the preoperative staging of rec-
er can also lead to misinterpretation. tal tumors:
5) Frank stenosis also precludes precise
endosonographic evaluation.
6) Angulation of the probe to the tumor axis also uTw: uTO/early Tl
can cause misinterpretation. Only those struc-
tures seen at a 90° angle to the probe can be uTx: deep Tl/early T2
assessed correctly.
7) Incomplete field of vision can cause shadow- uTy: deep T2/early T3
ing, reverberation or a mirror image.
8) Difficulties in distinguishing the layers of the uTz: deep T3/early T4
rectal wall can be due to pressure from the
balloon compressing the layers of the wall UNI significance: Probable; Definite;
together (Fig. IV.7). For this purpose Katsura Equivocal
et al. [23] and Boyce at al. [24] have introduced
a free filling of the rectum with de-aerated
water. Although slightly complicated com- Whether tumor site (in terms of height) and
pared with the balloon method, it makes it position (with respect to rectal circumference)
possible to describe the rectal wall as a five or have an influence on the reliability of endolumi-
seven-layer structure more clearly and it is nal ultrasound staging is not settled, as yet.

Fig. IY.7. Difficulties in distinguishing the

layers of the rectal wall can be due to pres-
sure from the balloon compressing the
layers of the wall together
56 Atlas of Endoanal and Endorectal Ultrasonography

Table IV.2. Comparative studies for determining depth of tumor invasion by digital examination, endorectal ultra-
sonography (ERUS), computed tomography (CT) and magnetic resonance imaging (MRI)

Author Year Patients Digital examination ERUS CT MRI

Romano [26] 1985 23 87% 83%

Rifkin [27] 1989 102 (ERUS) 72% 53%
81 (CT)
Holdsworth [9] 1988 36 (ERUS) 86% 94%
17 (CT)
Waizer [28] 1989 58 83% 77% 66%
Beynon [14] 1989 42 62% 95% 71%
Goldman [29] 1991 32 81% 52%
Akasu [16] 1990 41 80% 46%
Waizer [30] 1991 13 85% 77%
Herzog [4] 1993 87 91% 75%
Thaler [31] 1994 34 (ERUS) 88% 82%
35 (MRI)
Starck [32] 1995 34 (ERUS) 68% 88% 66%
35 (MRI)
Joosten [33] 1995 15 66% 66%
Zerhouni [34] 1996 III (CT) 74% 58%
79 (MRI)
Zagoria [35] 1997 10 70% 80%
Kim [36,37] 1999 89 (ERUS) 81.1% 65·2% 81%
69 (CT) (T2 = 50%) (T2 =15%) (T2 = 57.1%)
73 (MRI) (T3 = 87%) (T3 = 88%) (T3 = 89%)

Sentovich et al. [5] and Senesse et al. [25] reported scanner. For examinations considered adequate
a significantly better result for tumors within 6 by the surgeon, the distance of the tumor from the
cm from the anal verge. This is in contradiction to anal verge did not influence accuracy.
the study conducted by Herzog et al. [4] who A number of comparative studies have been
found a significantly poorer accuracy rate for performed to assess the efficacy of endorectal
tumors of the distal third. The reason for the less ultrasonography, computed tomography, magnet-
accurate staging in the lower rectum is a technical ic resonance imaging and digital examination in
one, because of the difficulty in reaching all sites the preoperative staging of rectal cancer (Table
of the ampulla recti with a rigid probe. However, IV.2) [26-37]. Some studies have shown a clear
Sailer et al. [18] reported that tumor site and superiority for endorectal ultrasonography,
tumor position do not influence the predictive whereas other studies have shown little difference.
value of ERUS. Analysis showed that there was no CT and MR imaging are accurate in assessing
difference between the various locations, which spread beyond the rectal wall, invasion of con-
means that all tumors are equally amenable to tiguous structures, spread to regional nodes or
ultrasound staging if they are within reach of the distant metastases (Figs. IV.8, 9). The lateral pelvic
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 57

Fig. IV.S. Computed tomography scan

showing a rectal tumor confined to the
rectal wall (a) with preservation of the
rectovaginal septum (arrow) and a rectal
tumor invading the seminal vesicles (b)
(arrow)

lymph nodes, such as the obturator nodes, are identification of neoplastic infiltration of perirec-
located too far from the rectum to be imaged tal fat (T3) endorectal balloon computed tomog-
effectively with rectal probes. Therefore, possible raphy had 100% sensitivity, 78.7% specificity and
advantages of MRI and CT can be considered in 86.8% accuracy. The CT sensitivity for detecting
assessing the lateral pelvic lymph nodes, pelvic lymph node metastases was 52.6%, specificity
wall invasion and involvement of levator ani mus- 85.3% and accuracy 73.6%. ERUS is minimally
cle. However, CT and MR imaging currently lack invasive and less expensive than CT.
the accuracy in determining the depth of wall MRI with endorectal coils has been studied in a
invasion required by the surgeon. Overall accura- number of small studies for the evaluation and
cy of CT for the staging of rectal tumors is staging of rectal tumors [39,40]. With the addition
approximately 50% to 75%. Goldman et al. [29] of endorectal surface coils to conventional MR
compared CT with ERUS and found accuracy imaging, spatial resolution is increased and
rates of 52% and 81%, respectively, for perirectal anatomic definition is improved. T2-weighted
fat invasion and 64% and 68%, respectively, for turbo spin-echo sequences allow to distinguish the
lymph node involvement. Similarly, Beynon et al. five layers of the rectal wall. These layers consist of
[14] showed that ERUS was significantly more an inner layer of high signal intensity (fluid
accurate than CT for both depth of tissue invasion between the coil and the rectal wall), a second layer
and lymph node involvement. Accuracy rates were of low signal intensity (mucosa and muscolaris
68%,82% and 91% for 44 patients evaluated with mucosae), a third layer of high signal intensity
digital examination, CT and ERUS respectively. (submucosa), a fourth layer of low signal intensity
Other investigators have also demonstrated a (muscolaris propria) and an outer layer of high
superior accuracy and reliability of ERUS com- signal intensity (perirectal fat). Rectal carcinoma
pared with CT. Civelli et al. [38] reported in the in T2-weighted turbo spin-echo sequences gives

Fig. IV.9. Computed tomography scan

showing a rectal tumor confined to the
rectal wall (a) (arrow) and a rectal tumor
invading the prostate (b) (arrow)
58 Atlas of Endoanal and Endorectal Ultrasonography

ERUS (Fig. IY.lO). It is unlikely, however, that MRI

will gain wid@!;pread UYllge because of lack of
widespread availability and significantly increased
financial costs. One advantage of endorectal MRI
is the ability to visualize the normal and patholog-
ical anatomy in multiple scan planes.
Three-dimensional (3D) imaging offers
improved knowledge of various anatomic struc-
tures and tumors. Additional scan planes (longi-
tudinal and coronal) can be used to determine the
size, location and local extent of the lesion pre-
cisely (Figs. IV.n, 12) [41]. Kim et al. performed a
prospective study to verify whether 3D ERUS
enhances the accuracy of rectal staging, as com-
pared with 2D ERUS [42]. The accuracy of 3D
ERUS was 90.9% for pT2 and 84.8% for pT3,
whereas that of 2D ERUS was 84.8% and 75.8%
respectively, thereby showing no difference
between these two methods. The lymph node
Fig. IV.IO. Magnetic resonance images showing a rectal
tumor with pararectal lymph nodes (arrow) metastasis was accurately predicted by 3D ERUS
in 84.8% of patients, whereas 2D ERUS predicted
the disorder in 66.7%. The difference was not sta-
tistically significant. Although the findings did
medium-to-Iow signal intensity higher than the not show 3D ERUS to have a significant advantage
muscular layer. MRI and ERUS demonstrate equiv- over 2D ERUS for the accurate evaluation of rectal
alent efficacy in the preoperative staging of rectal cancer, stereoscopic visualization provided easier
tumors. Overall accuracy rates of 70% to 90% have and complete understanding of both focal lesions
been reported for staging of rectal tumors using and lymph nodes. Additional studies are neces-
MRI with endorectal coils. In the evaluation of sary to further evaluate the efficacy of 3D ERUS
lymph nodes, MRI does not offer significant and to develop imaging protocols for special clin-
improvement in accuracy rates compared with ical situations.

Fig. IV.H. 3D image of an uT3 rectal

tumor. The sagittal projection suggests
that the the tumor spread to the perirectal
fat
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 59

Fig. IV.l2. 3D image of an uT3 rectal

tumor extending from the 3 to 9 o'clock
position. The projection suggests that the
tumor spread to the perirectal fat

References 11. Hulsmans FJ, Castelijns JA, Reeders JW, Tytgat GN.
Review of artifacts associated with transrectal ultra-
1. Dukes CEo The classification of cancer of the rectum. J sound: understanding, recognition and prevention of
PathoI1932;35:323-32 misinterpretation. J Clin Ultrasound 1995;23:483-94
2. Astler VB, Coller F. The prognostic significance of 12. Yamashita Y, Machi J, Shirouzu K, Morotomi T,Isomoto
direct extension of carcinoma of the colon and rec- H, Kakegawa T. Evaluation of endorectal ultrasound for
tum. Ann Surg 1954;139:846-51 the assessment of wall invasion of rectal cancer: report
3. Hildebrandt U, Feifel G. Preoperative staging of rectal of a case. Dis Colon Rectum 1988;31:362-8
cancer by intrarectal ultrasound. Dis Colon Rectum 13. Dershaw DD, Enker WE, Cohen AM, Sigurdson ER.
1985;28:42-6 Transrectal ultrasonography of rectal carcinoma.
4. Herzog U, von Flue M, Tondelli P, Schuppisser JP. How ac- Cancer 1990;66:2336-40
curate is endorectal ultrasound in the preoperative stag- 14. Beynon J, Mortensen NJ, Foy DM, Channer JL, Virjee J,
ing of rectal cancer? Dis Colon Rectum 1993;36:127-34 Goddard P. Preoperative assessment of local invasion in
5. Sentovich S, Blatchford G, falk P, Thorson A, Christensen rectal cancer: digital examination, endoluminal sonog-
M. Transrectal ultrasound of rectal tumors. Am J Surg raphy or computed tomography Br J Surg 1986;73:1015-7
1993;166:638-41 15. Garcia-Aguilar J, Pollack J, Lee S-H, Hernandez de
6. Anderson BO, Hann LE, Enker WE, Dershaw DD, Anda E, Mellgren A, Wong WD, Finne CO,
Guillem JG, Cohen AM. Transrectal ultrasonography Rothenberger DA, Madoff RD. Accuracy of endorectal
and operative selection for early carcinoma of the rec- ultrasonography in preoperative staging of rectal
tum. J Am Coli Surg 1994;179:513-7 tumors. Dis Colon Rectum 2002;45=10-5
7. Harnsberger JR, Charvat P, Longo WE, Vernava AM, 16. Akasu T, Sunouchi K, Sawada T, Tsioulias GJ, Muto T,
Salimi Z, Arends T, Daniel G. The role of intrarectal Morioka Y. Preoperative staging of rectal carcinoma:
ultrasound in staging of rectal cancer and detection of prospective comparison of trans rectal ultrasonogra-
extrarectal pathology. Am Surg 1994;60:571-6 phy and computed tomography. Gastroenterology
8. Glaser F, Schlag P, Herfarth C. Endorectal ultrasonog- 1990;98:268
raphy for the assessment of invasion of rectal tumors 17. Jochem RJ, Reading CC, Dozois RR, Carpenter HA,
and lymph node involvement. Br J Surg 1990;77:883-7 Wolff BG, Charboneau JW. Endorectal ultrasonograpic
9. Holdsworth PJ, Johnston D, Chalmers AG, Chennells P, staging of rectal carcinoma. Mayo Clin Proc
Dixon MF, Finan PJ, Primrose IN, Quirke P. 1990;65:1571-7
Endoluminal ultrasound and computed tomography 18. Sailer M, Leppert R, Bussen D, Fuchs KH, Thiede A.
in the staging of rectal cancer. Br J Surg 1988;75=1019-22 Influence of tumor position on accuracy of endorectal
10. Orrom WJ, Wong WD, Rothenberger DA, Jensen LL, ultrasound staging. Dis Colon Rectum 1997;40:1180-6
Goldberg SM. Endorectal ultrasound in the preopera- 19. Kim HJ, Wong WD. Role of endorectal ultrasound in
tive staging of rectal tumors. A learning experience. the conservative management of rectal cancers. Semin
Dis Colon Rectum 1990;33:654-9 Surg Oncol 2000;19:358-66
60 Atlas of Endoanal and Endorectal Ultrasonography

20. Santoro GA, Pastore C, Barban M, Di Falco G. Role of nance imaging are superior to clinical examination in
endorectal ultrasound in the management of rectal the preoperative staging of rectal cancer. Eur J Surg
cancers. Hepato-Gastroenterology 2001;48 (SI):CXIX 1995;161:841-5
21. Kruskal JB, Sentovich SM, Kane RA. Staging of rectal 33. Joosten FB, Jansen JB, Joosten HJ, Rosenbusch G.
cancer after polypectomy: usefulness of endorectal Staging of rectal carcinoma using MR double surface
US. Radiology 1999;2ll:31-5 coil, MR endorectal coil and intrarectal ultrasound:
22. Mier AG, Barton PP, Heuhold NR, Herbst F, Teleky BK, correlation with histopathologic findings. J Comput
Lechner GL. Peritumor tissue reaction at transrectal Assist Tomogr 1995;19:752-8
US as a possible cause of overstaging in rectal cancer: 34. Zerhouni EA, Rutter C, Hamilton SR, Balfe DM,
histopathologic correlation. Radiology 1997;203:785-9 Megibow AJ, Francis IR, Moss AA, Heiken JP, Tempany
23. Katsura Y, Yamada K, Ishizawa T, Yoshinaka H, CM, Aisen AM, Weinreb JC, Gatsonis C, McNeil BJ. CT
Shimazu H. Endorectal ultrasonography for the and MR imaging in the staging of colorectal carcino-
assessment of wall invasion and lymph node metasta- ma: report of the Radiology Diagnostic Oncology
sis in rectal cancer. Dis Colon Rectum 1992;35:362-8 Group II. Radiology 1996;200:443-51
24. Boyce GA, Sivak MV, Lavery IC, Fazio VW, Church JM, 35. Zagoria RJ, Scharlb CA, Ott DJ, Bechtoldt RI, Wolfman
Milsom J, Petras R. Endoscopic ultrasound in the pre- NT, Scharling ES, Chen MY, Loggie BW. Assessment of
operative staging of rectal carcinoma. Gastrointest rectal tumor infiltration utilizing endorectal MR
Endosc 1992;38:468-71 imaging and comparison with endoscopic rectal
25. Sene sse P, Khemissa F, Lemanski C, Masson B, Quenet sonography. J Surg OncoI1997;64:312-7
F, Saint-Aubert B, Simony J, Ychou M, Dubois JB, 36. Kim NK, Kim MJ, Yun SH, Sohn SK, Min JS.
Rouanet P. Contribution of endorectal ultrasonogra- Comparative study of transrectal ultrasonography,
phy in preoperative evaluation for very low rectal can- pelvic computerized tomography and magnetic reso-
cer. Gastroenterol Clin BioI 2001;25:24-8 nance imaging in preoperative staging of rectal cancer.
26. Romano G, de Rosa P, Vallone G, Rotondo A, Grassi R, Dis Colon Rectum 1999;42:770-5
Santangelo ML. Intrarectal ultrasound and computed 37. Kim NK, Kim MJ, Park JK, Park SI, Min JS.
tomography in the pre- and postoperative assessment Preoperative staging of rectal cancer with MRI: accu-
of patients with rectal cancer. Br J Surg 1985; racy and clinical usefulness. Ann Surg Oncol 2000;7:
72(suppl): Sl17-9 732-7
27. Rifkin MD, Ehrlich SM, Marks G. Staging of rectal car- 38. Civelli EM, Gallino G, Mariani L, Cozzi G, Biganzoli E,
cinoma: prospective comparison of endorectal US and Salvetti M, Gallo R, Belli F, Bonfanti G, Bertario L,
CT. Radiology 1989;170:319-22 Andreola S, Leo E. Double-contrast barium enema and
28. Wazier A, Zitron S, Ben-Baruch D, Baniel J, Wolloch Y, computerised tomography in the pre-operative evalu-
Dintsman M. Comparative study for preoperative stag- ation of rectal carcinoma: are they still useful diagnos-
ing of rectal cancer. Dis Colon Rectum 1989;32:53-6 tic procedures? Tumori 2000;86:389-92
29. Goldman S, Arvidson H, Normig U, Lagerstedt U, 39. Meyenberger C, Huch Boni RA, Bertschinger P, Zala
Magnusson I, Frisell J. Transrectal ultrasound and GF, Klotz HP, Krestin GP. Endoscopic ultrasound and
computed tomography in preoperative staging of endorectal magnetic resonance imaging: a prospec-
lower rectal adenocarcinoma. Gastrointestinal Radiol tive, comparative study for preoperative staging and
1991;16:259-63 follow-up of rectal cancer. Endoscopy 1995;27:469-79
30. Waizer A, Powsner E, Russo I, Hadar S, Cytron S, 40. Gualdi GF, Cas ciani E, Gadalaxara A, d'Orta C,
Lombrozo R, Wolloch Y, Antebi E. Prospective compar- Polettini E, Pappalardo G. Local staging of rectal can-
ative study of magnetic resonance imaging versus cer with transrectal ultrasound and endorectal mag-
transrectal ultrasound for preoperative staging and netic resonance imaging: comparison with histologi-
follow-up of rectal cancer: preliminary report. Dis cal findings. Dis Colon Rectum 2000;43:338-45
Colon Rectum 1991;34:1068-72 41. Hunerbein M, Pegios W, Bau R, Vogl TJ, Felix R, Schlag
31. Thaler W, Watzka S, Martin F, Guardia GL, Psenner K, PM. Prospective comparison of endorectal ultrasound,
Bonatti G, Fichtel G, Egarter-Vigl E, Marzoli GP. three-dimensional ultrasound and endorectal MRI in
Preoperative staging of rectal cancer by endoluminal the preoperative evaluation of rectal tumors.
ultrasound vs. magnetic resonance imaging, prelimi- Preliminary results. Surg Endosc 2000;14:1005-9
nary results of a prospective comparative study. Dis 42. Kim JC, Cho YK, Kim SY, Park SK, Lee MG.
Colon Rectum 1994;37:1189-93 Comparative study of three-dimensional and conven-
32. Starck M, Bohe M, Fork FT, Lindstrom C, Sjoberg S. tional endorectal ultrasonography used in rectal can-
Endoluminal ultrasound and low-field magnetic reso- cer staging. Surg Endosc 2002;16:1280-5
 IY.2.
Stage uTO: Villous adenoma
G.A. Santoro, G. Di Falco

Sonographic evaluation of a villous rectal lesion is that expand the inner hypoechoic layer and are
helpful in determining the presence of tumor surrounded by a uniform middle hyperechoic
invasion. Although the presence of ulceration or layer are considered villous adenoma (Fig. IV.I3),
induration in these lesions correlates with the and lesions that expand the inner hypo echoic
presence of cancer, approximately 10% of villous layer and have distinct echo defects of the middle
polyps harbor a malignancy that is not clinically hyperechoic layer are considered UTI tumors.
apparent and is frequently missed on random They reported that the positive predictive value of
biopsy. Accurate preoperative staging of such ERUS for detecting a malignant focus was 66.7%,
lesions may determine the choice between a sub- the negative predictive value was 88.7%;sensitivi-
mucosal or a full-thickness excision or even a rad- ty was 50% and specificity was 94%. When an
ical resection [1]. Given the limitations of digital optimal image was obtained, all cancers that pen-
examination and CT scan for detecting malignan- etrated the submucosa were detected. Sensitivity
cy within rectal villous lesions, one only could of the technique was compromised in some large
advise a patient on the higher incidence of malig- exophytic lesions and in those at the level of the
nancy in a larger lesion. The potential for both anal sphincter because of artifacts produced in
undertreating and overtreating lesions is great. the ultrasonographic image. Garcia-Aguilar et al.
ERUS is evolving into a powerful tool for assess- [5], using these criteria, have reported a high
ment of rectal cancers [2]. There is however less degree of accuracy discriminating between vil-
published data available to clarify the role of lous adenomas and early invasive cancer. Glaser et
ERUS in preoperative assessment of villous ade- al. [6] correctly identified all six cancers in a
noma. The presence of an intact hyperechoic sub- group of 31 villous adenoma, identifying them by
mucosal interface indicates lack of tumor inva- the presence of echo-poor foci within the homo-
sion into the submucosa. Heintz et al. [3] believe geneously echo-rich adenoma, rather than by sub-
that ERUS cannot differentiate between villous mucosal layer echo defects (Fig. IY.I4).
adenoma and invasive cancer because neither the Technical difficulties associated with scan-
muscolaris mucosae nor the submucosa is sono- ning villous adenoma can be due to very large
graphically definable and the first hypoechoic lesions that tend to attenuate rectal layers, mak-
layer corresponds anatomically with the mucosa ing them difficult to differentiate from complex
and the submucosa. They suggest that uTO and internal echoes of the polyp. Lesions with a very
UTI tumors that appear as a broadening of the large exophytic component place the rectal wall
first hypoechoic layer should be classified togeth- at the limits of the probe's focal length, reducing
er. Instead, Adams and Wong [4] disagree with resolution (Fig. IV.Is). According to Adams and
this interpretation; they consider the first hypoe- Wong [4] some improvements may be gained by
choic layer as the mucosa and muscolaris directly filling the rectum with water for acoustic
mucosae and the middle hyperechoic layer as the coupling to avoid the compression and distor-
submucosa. Consequently for such authors lesions tion that a balloon produces in the lesion. In
 62 Atlas of Endoanal and Endorectal Ultrasonography

a b

Fig. IY.l3. uTO rectal tumor (villous adenoma). Lesion expands the inner hypoechoic layer and is surrounded by the uniform
middle hyperechoic layer of the submucosa (arrows) (a). Endoscopic appearance (b)

Fig. IV.I4. Early invasive cancer identified by the presence of Fig. IV.IS. The complex structure of this large exophytic
echo-poor foci (arrows) within the homogeneously echo-rich lesion produce artifacts in the ultrasonographic image due to
villous adenoma poor acoustic contact of the probe (arrows). Note the rever-
beration echo from 10 to 11 o'clock position
 Section IV· Endoluminal ultrasound in the preoperative staging of rectal carcinoma 63

large carpeting lesions, careful evaluation of the treatment for rectal villous lesions. In this study
entire tumor is necessary to determine that a the overall accuracy of ERUS in the evaluation of
small area of invasion has not been overlooked. tumor invasion depth was 60% in villous lesions
In some polyps the complex structure produces and 91% in nonvillous lesions. In villous lesions
fixed artifacts over one part of the rectal wall, 37% were over staged and 3% understaged and in
obscuring the image. The rectal wall thickness nonvillous lesions 6% were overstaged and 3%
near the anal sphincter and the layers become understaged.
less distinct. Scanning is also complicated at this ERUS can effectively detect the presence of
level by the balloon separating from the rectal associated lymphadenopathy. In benign villous
wall at the angle produced by puborectalis. Snare lesions, there should not be any enlarged perirec-
biopsy of lesions before referral to ERUS pro- tal nodes. Presence of lymphadenopathy should
duces a burn artifact, which can lead to tumor alert the examiner that malignant degeneration
overstaging [7]. with lymphatic metastases is possible and exci-
Konishi et al. [8] reported that ERUS-based sion of the rectal segment with perirectal nodes
evaluation· cannot determine the appropriate may be indicated as opposed to local excision.

References 5· Garcia-Aguilar J, Pollack J, Lee S-H, Hernandez de

Anda E, Mellgren A, Wong WD, Finne CO,
1. Bernick PE, Wong WD. Staging. What makes sense? Rothenberger DA, Madoff RD. Accuracy of endorectal
Can the pathologist help? Surg Oncol Clin North Am ultrasonography in preoperative staging of rectal
2000;9:703-23 tumors. Dis Colon Rectum 2002;45:10-5
2. Hildebrandt U, Feifel G. Preoperative staging of rectal 6. Glaser F, Schlag P, Herfarth C. Endorectal ultra-
cancer by intrarectal ultrasound. Dis Colon Rectum sonography for the assessment of invasion of rectal
1985;28:42-6 tumors and lymph node involvement. Br J Surg 1990;
3. Heintz A, Buess G, Frank K, Kuntz C, Strunck H, 77:883-7
Junginger T. Endoluminal ultrasonic examination of 7. Orrom WJ, Wong WD, Rothenberger DA, Jensen LL,
sessile polyps and early carcinomas of the rectum. Goldberg SM. Endorectal ultrasound in the preopera-
Surg Endosc 1989;3:92-5 tive staging of rectal tumors. A learning experience.
4. Adams WJ, Wong WD. Endorectal ultrasonic detection Dis Colon Rectum 1990;33:654-9
of malignancy within rectal villous lesions. Dis Colon 8. Konishi K,Akita X, Kaneko K et al. Evaluation of endo-
Rectum 1995;38:1093-6 scopic ultrasonography in colorectal villous lesions.
Int J Colorectal Dis 2003;18:19-24
 Clinical cases

Stage uTO. Case 1. Large villous adeno-

ma in a 54Y male occupying 40% of the
rectal circumference in left posterolateral
aspect and located at the level of the sem-
inal vesicles (a). The tumor is 4cm in
greatest diameter and about 3cm in thick-
ness. The 3D projections (b, c) show that
the hyperechoic submucosal interface is
intact surrounding the lesion. Endoscopic
appearance (d). Macroscopic view (e)

b c

~________________________~ e
d
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 65

Stage uTO. Case 2. Villous adenoma in a 57Y male in the pos-

terior rectal wall and located at the level of the seminal vesi-
cles (a). The submucosal layer is intact consistent with benign
lesion. Endoscopic appearance (b)
b

a b
Stage uTO. Case 3. Small polyp in a 67Y female in the left lateral rectal wall just above the dentate line (a). The submucosal layer
is intact consistent with benign lesion. Endoscopic appearance (b)
 66 Atlas of Endoanal and Endorectal Ultrasonography

IViliOUS tumorl

Stage uTO. Case 4. Villous adenoma in a 80 Y male in the left

lateral rectal wall and located at the level of the seminal vesi-
cles (a). The lesion occupies 40% of the rectal circumference
and has the appearance of a benign lesion. Endoscopic
b appearance (b)

a b
Stage uTO. Case 5 Small polyp in a 43Y female in the posterior rectal wall just above the dentate line (a). The submucosal layer
is intact consistent with benign lesion. Endoscopic appearance (b)
 IV.3.
Stage uTI: Submucosal invasion
G.A. Santoro, G. Di Falco

Accurate preoperative staging of early cancers is because of their obliterated deeper margin. Such
important if local excision is to be considered a errors are frequently encountered in tumors with
treatment alternative (Fig. IV.16). ERUS appears to irregular surfaces and in pedunculated rather than
be more accurate than CT or MRI for determining sessile lesions. Free filling of the rectum with
the depth of wall invasion in rectal cancer. The deaerated water may be indicated for ERUS of early
ERUS accuracy rate for evaluating early rectal can- stage tumors to avoid compression of the five-layer
cer has been reported to be between 81% and 92% echo structure of the rectal wall by the water-filled
[1-6]. Overstaging and understaging has been balloon. This is accomplished by instillation of
reported to be between 8% and 23% and between water by a 60-ml syringe via the accessory channel
0% and 18%, respectively [1-6]. Strunk et al. [7] of the instrument. Volumes greater than 250 ml are
found that 92% of all lesions with an inhomoge- sometimes necessary to eliminate air artifact. This
neously echo-poor pattern represented malignant also prevents refraction artifacts from being seen
transformations. Submucosal invasion is diag- around tumor edges. Reverberation artifacts
nosed by disruption of the hyperechoic submucos- caused by feces from incomplete bowel prepara-
al interface without alteration of the muscolaris tion obscure the deeper tumor margin to create
propria. The tumor is hypo echoic, and invasion of overstaging. Another source of error can come
the submucosa results in destruction of the hyper- from previous biopsy or from polypectomy that
echoic line that represents the submucosal layer. has caused an inflammatory reaction with result-
This layer is usually well seen and its disruption is
a reliable sign of tumor invasion (Fig. IV.17). Care
must be taken not to overinterpret slight irregular-
ities of the outer margin of the muscolaris propria
as tumor extension [8]. Irregularities of the sub-
mucosal layer can occur as a stippling or thicken-
ing but should not constitute a distinct break in the
submucosal layer. If a break is seen in the submu-
cosallayer, the muscolaris propria has been invad-
ed (uT2Iesion). Source of errors in ERUS was very
well summarized by Kim et aI. [8]. Overstaging of
Tl tumors may be due to compression of the rectal
wall by the water-filled balloon causing oblitera-
tion of the interface between the tumor and the
muscolaris propria (Fig. IV.18). Because air or
water in a gap between the tumor surface and the
balloon produce either reverberation or posterior Fig. Iy'16. Diagrammatic representation of Tl rectal tumor
enhancement, tumors appear to be more advanced invading submucosa
68 Atlas of Endoanal and Endorectal Ultrasonography

nodes. Source of errors in the evaluation of early

rectal cancer by ERUS can also frequently be
caused by examiner confusion or a tendency to
overestimate a malignant lesion because of con-
cern for undertreatment despite clean ERUS
imaging [8]. Image recording appears to allow
correction of interpretative errors by review, espe-
cially for an inexperienced examiner.
Akasu et al. [6] evaluated the accuracy of pre-
operative staging by ERUS and its contribution to
treatment of early stage rectal cancer. TI-Iesions
were grouped as TI-slight and TI-massive.
Sensitivity, specificity, overall accuracy rate, posi-
tive predictive value and negative predictive value
for detection of slight submucosal invasion were
99%, 74%, 96%, 97% and 87%, respectively.
Sensitivity, specificity, overall accuracy rate, posi-
tive predictive value and negative predictive value
for detection of massive submucosal invasion
were 98%, 88%, 9?'Yo, 98% and 88%, respectively.
Fig. IY.17. UTI rectal tumor. Lesion expands the inner middle
hyperechoic layer of the submucosa (arrows) and is sur- The incidence of lymph node metastases was 26%
rounded by the uniform hypoechoic layer corresponding to for uTI-massive tumors versus 0% for uTI-slight
the muscolaris propria tumors (p=0.0003%). If the cut-off value for pos-
itive nodes was 5 mm or larger, the specificity and
overall accuracy rate were high (94% and 89%,
ing edema, abscess or fibrosis thus obliterating the respectively), but the sensitivity was low (38%). If
anatomic plane. Evaluation of associated perirec- the cut-off value for positive nodes was 4 mm,
tal nodes is also important since up to 14% of sensitivity, specificity, overall accuracy rate for
tumors with submucosal invasion may have nodal detecting lymph node metastases were 63%, 76%
metastases and therefore require proctectomy and 73%, respectively. With the cut-off value for
instead of local excision. The small vessels, ure- positive nodes set at 3 mm, sensitivity, specificity,
thra, seminal vesicles and small bowel are known overall accuracy rate for detecting lymph node
to be mistaken frequently for metastatic lymph metastases were 75%, 49% and 53%, respectively.

Fig. IV.IS. Difficulties in staging of TI tumor may be due to compression of the rectal wall by the water-filled balloon (a: vol-
ume of 60 ml; b: volume of 20 ml) causing obliteration of the interface between tumor and muscolaris propria
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 69

Fig. IY.19. Early invasive cancer identified by the presence of

echo-poor area (arrow) within the homogeneously echo-rich
villous adenoma (a). Endoscopic view of the polypoid lesion
(b). Pathologic finding confirms a tubolovillous adenomas
c ~ ______________ ~ ____ ~ ____________ ~
showing invasive adenocarcinoma within the core of the
polyp (c). (H&E stain)

Thus a cut-off value of 3 mm seems to be best for node metastases. Its detection is clinically impor-
pTI tumors. On the other hand, it is worthy of tant because the choice of treatment, whether to
note that, the larger the detected node, the higher perform radical surgery or local excision/endo-
is the probability of metastases. This study con- scopic polypectomy, should be based on whether
firmed that ERUS is accurate for preoperative lymph node metastases are present. Patients with
evaluation of transmural invasion depth for early uTI-slight tumors staged by ERUS are at low risk
stage rectal cancer (Fig. IV.I9). The classification of positive nodes and good candidates for local
of UTI into TI-slight and TI-massive can be treatment. Those with uTI-massive lesions should
deemed rational. The incidence of lymph node be treated with a radical operation because of the
metastases was 0% in pTI-slight tumors and 22% high incidence of positive nodes, as confirmed in
in pTI-massive tumors. Thus massive submucosal a recent study by Nascimbeni et al. [9] from the
invasion can be considered a risk factor for lymph Mayo Clinic.
70 Atlas of Endoanal and Endorectal Ultrasonography

References 5. Sailer M, Leppert R, Kraemer M, Fuchs KH, Thiede A.

The value of endorectal ultrasound in the assessment
1. Hildebrandt U, Feifel G. Preoperative staging of rectal of adenomas, Tl- and T2-carcinomas. lnt J Colorectal
cancer by intrarectal ultrasound. Dis Colon Rectum Dis 1997;12:214-9
1985;28:42-6 6. Adams WJ, Wong WD. Endorectal ultrasonic detection
2. Boyce GA, Sivak MV, Lavery IC, Fazio VW, Church JM, of malignancy within rectal villous lesions. Dis Colon
Milsom J, Petras R. Endoscopic ultrasound in the pre- Rectum 1995;38:1093-6
operative staging of rectal carcinoma. Gastrointest 7. Strunk H, Heintz A, Frank K, Buess G, Braunstein S.
Endosc 1992;38:468-71 Endosonographic staging of rectal tumors. Fortschr
3. Bernick PE, Wong WD. Staging. What makes sense? Rontgenstr 1990;153:373-8
Can the pathologist help? Surg Oncol Clin North Am 8. Kim JC, Yu CS, Jung HY, Kim HC, Park SK, Kang GH,
2000;9:703-23 Lee MG. Source of errors in the evaluation of early rec-
4. Akasu T, Kondo H, Morioka Y, Sugihara K, Gotoda T, tal cancer by endoluminal ultrasonography. Dis Colon
Fujita S, Muto T, Kakizoe T. Endorectal ultrasonogra- Rectum 2001;44:1302-9
phy and treatment of early stage rectal cancer. World J 9. Nascimbeni R, Burgart LJ, Nivatvongs S, Larson DR.
Surg 2000;24:1061-8 Risk of lymph node metastasis in Tl carcinoma of the
colon and rectum. Dis Colon Rectum 2002;45:200-6
 Clinical cases

Stage uTI. Case 1. Tumor in a 68Y male

located in the right posterolateral rectal
wall15cm from the anal verge. The lesion
is approximately 3.5 cm in diameter. The
hyperechoic layer corresponding to the
submucosa is interrupted at several
points consistent with tumor invasion
(a). The hypo echoic layer representing
the muscolaris propria is intact as is the
I Submucosa I hyperechoic layer corresponding to the
perirectal fat interface. There are no visi-
b
e perirectal lymph nodes. These find-
ings correspond to an UTINO tumor. (b)
Double contrast barium enema showing
a large polypoid lesion. (c) Endoscopic
view of a friable tumor with central
ulceration. (d) Small depressed foci of
invasive carcinoma within a protruding
adenoma. (e) Pathology confirms a mod-
a erately differentiated adenocarcinoma
invading the submucosa (pTINO). (H&E
stain)

b c

d L-__________________________________ ~

e
 72 Atlas of Endoanal and Endorectal Ultrasonography

Stage uTI. Case 2. Tumor in a 55Y female located in the left

anterolateral rectal wall 8 cm from the anal verge. The hyper-
echoic layer corresponding to the submucosa is interrupted at
several points consistent with tumor invasion (a). The hypoe-
choic layer representing the muscolaris propria is intact as is
the hyperechoic layer corresponding to the perirectal fat
interface. There are no visible perirectal lymph nodes. These
findings correspond to an UTINO tumor. (b) Double contrast
barium enema showing the lesion (arrows). (c) Endoscopic
view. (d) Macroscopic view. (e) Pathology confirms a moder-
ately differentiated adenocarcinoma invading the submucosa.
d (H&E stain) Lymph nodes are free of tumor (pTINO).
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 73

a b

C L -________________________ ________
d
~ ~~ ~

Stage uTI. Case 3. Tumor in a 56Y female located in the right anterolateral rectal wall 7 cm from the anal verge. (a)
Ultrasonography shows a hyperechoic submucosal layer interrupted at several points consistent with tumor invasion. The
hypoechoic layer representing the muscolaris propria is intact as is the hyperechoic layer corresponding to the perirectal fat
interface. There are no visible perirectal lymph nodes. These findings correspond to an UTINO tumor. (b) Endoscopic view of a
friable tumor with central ulceration. (c) Macroscopic view. (d) These findings accord histologically with a moderately differ-
entiated adenocarcinoma invading the submucosa (H&E stain). Lymph nodes are free of tumor (pTINO)
 IY.4.
Stage UT2: Invasion of the muscular layer
G.A. Santoro, G. Di Falco

Sonographic diagnosis of tumor invasion of the compared postoperatively with the histological
muscolaris propria is based on the thickening of result at the University Hospital of Wurzburg.
this layer (Figs. IV.20-22) [1]. The muscolaris pro- The sensitivity for adenomas and TI tumors was
pria is represented by a thin hypo echoic layer 81%, the specificity 98%. For T2 tumors, the sen-
adjacent to the hyperechoic submucosal interface. sitivity was only 41% and the specificity 92% as
Since the tumor is also hypoechoic, early muscu- the majority of pT2 neoplasias were overstaged
lar invasion is difficult to detect [2]. The sur- (uT3). Authors concluded that adenomas and TI
rounding hyperechoic layer corresponding to the tumors could be assessed with a high grade of
perirectal fat interface remains intact (Fig. IV.23). accuracy using ERUS. In these tumors ERUS can
Lymph node metastases occur in approximately be used to assist clinical decision-making. Owing
15% to 20% of patients with T2 tumors. ERUS is to the lack of sensitivity ERUS is of no help in the
important to distinguish UT2 and UTI lesions, assessment of T2 carcinomas. Katsura et al. [5]
because local therapy is not routinely recom- reported that the predictive value of positive rate
mended for UT2 rectal lesions [3]. in the assessment of rectal wall invasion by ERUS
In a prospective study Sailer et al. [4] exam- was 96.2% in UTI and 87.5% in UT2. Kim et al. [6]
ined the value of ERUS in the preoperative stag- performed a comparative study of ERUS, CT and
ing of potentially locally excisable tumors. MRI in preoperative staging of rectal cancer (Fig.
During the study period a total of 160 rectal IV.24). In pT2 tumors, accuracy rates were 50% by
tumors were staged endosonographically and ERUS, 15% by CT and 57.1% by MRI. Overstaging

Fig. IV.20. Diagrammatic representation of T2 rectal tumor Fig. IV.2!. Small adenocarcinoma invading muscolaris pro-
invading muscolaris propria pria (H&E stain)
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 75

Fig. IY.23. UT2 rectal tumor. Sonographic diagnosis of tumor

invasion of the muscolaris propria is based on the thickening
of this hypoechoic layer. The surrounding hyperechoic layer
corresponding to the perirectal fat interface remain intact

is a particular problem with T2 tumors [7-10 J.

Fig. IY.22. Sonographic diagnosis of tumor invasion of the
muscolaris propria (UT2) is based on thickening of this layer
Among the interpretative errors, severe inflam-
matory infiltrate underlying a tumor, which is
sonographically indistinguishable from malig-
nant tissue, can inhibit accurate evaluation of
tumor invasion and appears to cause inevitable
error. Understaging, on the other hand, may be
caused by a failure to detect microscopic cancer
infiltration owing to the limits of resolution of

a
b
Fig. IY.24. Endoluminal ultrasound showing a rectal tumor invading the muscolaris propria (UT2NO) (a). Computed tomogra-
phy scan showing the same lesion confined to the rectal wall (b)
 76 Atlas of Endoanal and Endorectal Ultrasonography

a
Fig. IV.25. (a) T2 tumor overstaged as T3. Endorectal ultrasound scan shows marked thickening of the muscolaris propria layer
with focal nodularity that appears to extend into the perirectal fat. Histopathological analysis demonstrated no extension into
the perirectal fat. (b) T3 tumor understaged as T2. Endorectal ultrasound scan shows a large tumor in the left side of the ante-
rior rectal wall. The tumor clearly extends into the hypo echoic muscolaris propria layer with a focal area of nodularity (arrows)
abutting the perirectal fat, thought to be due to tumor-induced desmoplasia of the muscolaris propria. Histopathologic analysis
demonstrated tumor extension into the perirectal fat

a b

Fig. IV.26. Three-dimensional section enables assessment of a

T2 rectal tumor in coronal (a), sagittal (b) and transverse (c)
scan planes. The sagittal and transverse planes reveal no
c extension into the perirectal fat (arrows)
 Section IV· Endoluminal ultrasound in the preoperative staging of rectal carcinoma 77

the equipment (Fig. IV.2S). Distinguishing (2D) ERUS for the accurate evaluation of rectal
between stage T2 and T3 tumors may be a prob- cancer (Fig. IV.26). In a preliminary study, Kim et
lem, particularly if the patient has received pre- al. [11] showed that the accuracy of 3D ERUS was
operative radiotherapy. 90.9% for pT2 whereas that of 2D-ERUS was
Three-dimensional (3D) ERUS offers a signifi- 84.8%.
cant advantage over conventional bi-dimensional

References 6. Kim NK, Kim MJ, Yun SH, Sohn SK, Min JS.
Comparative study of transrectal ultrasonography,
1. Hildebrandt U, Feifel G. Preoperative staging of rectal pelvic computerized tomography and magnetic reso-
cancer by intrarectal ultrasound. Dis Colon Rectum nance imaging in preoperative staging of rectal cancer.
1985;28:42-6 Dis Colon Rectum 1999;42:770-5
2. Boyce GA, Sivak MV, Lavery IC, Fazio VW, Church JM, 7. Akasu T, Kondo H, Morioka Y, Sugihara K, Gotoda T,
Milsom J, Petras R. Endoscopic ultrasound in the pre- Fujita S, Muto T, Kakizoe T. Endorectal ultrasonogra-
operative staging of rectal carcinoma. Gastrointest phy and treatment of early stage rectal cancer. World J
Endosc 1992;38:468-71 Surg 2000;24:1061-8
3. Bernick PE, Wong WD. Staging. What makes sense? 8. Adams WJ, Wong WD. Endorectal ultrasonic detection
Can the pathologist help? Surg Oncol Clin North Am of malignancy within rectal villous lesions. Dis Colon
2000;9:703-23 Rectum 1995;38:1093-6
4. Sailer M, Leppert R, Kraemer M, Fuchs KH, Thiede A. 9. Kim JC, Yu CS, Jung HY, Kim HC, Park SK, Kang GH,
The value of endorectal ultrasound in the assessment Lee MG. Source of errors in the evaluation of early rec-
of adenomas, Tl- and T2-carcinomas. Int J Colorectal tal cancer by endoluminal ultrasonography. Dis Colon
Dis 1997;12:214-9 Rectum 2001;44:1302-9
5. Katsura Y, Yamada K, Ishizawa T, Yoshinaka H, 10. Heriot AG, Grundy A, Kumar D. Preoperative staging
Shimazu H. Endorectal ultrasonography for the of rectal carcinoma. Br J Surg 1999;86:17-28
assessment of wall invasion and lymph node metas- 11. Kim JC, Cho YK, Kim SY, Park SK, Lee MG.
tasis in rectal cancer. Dis Colon Rectum 1992; Comparative study of three-dimensional and conven-
35:362-8 tional endorectal ultrasonography used in rectal can-
cer staging. Surg Endosc 2002;16:1280-5
 Clinical cases

l
a d

Stage uT2. Case 1. Tumor in a 54Y female occupying the

entire posterior hemicircumference of the rectal wall, located
7cm from the anal verge. The lesion is approximately 3.5 cm in
diameter and 1.2 cm in depth. The hyperechoic layer corre-
sponding to the submucosa is destroyed and the hypoechoic
layer representing the muscolaris propria is thickened consis-
tent with tumor invasion (a). The hyperechoic layer which
corresponds to the perirectal fat is intact. (b) Computed
tomography scan showing the rectal tumor confined to the
rectal wall. (c) Magnetic resonance images. The neoplasm of
the posterior wall of the rectum involves the low signal inten-
sity layer representing the muscolaris propria with no exten-
sion into the outer layer of high signal intensity representing
the perirectal fat. (d) Double contrast barium enema showing
an irregular surface (arrows) of the posterior rectal wall due
to tumor infiltration. (e) Macroscopic view of the tumor with
c central ulceration
 Section IV· Endoluminal ultrasound in the preoperative staging of rectal carcinoma 79

IBladderl

L -_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
c
~ d

Stage uT2. Case 2. Tumor in a 62 Y male encompassing two-thirds of the rectal circumference, located at the level of the semi-
nal vesicles. (a) The hyperechoic layer of the submucosa is destroyed and the hypoechoic layer representing the muscolaris pro-
pria is thickened consistent with tumor invasion. The hyperechoic layer which corresponds to the perirectal fat is intact. The
sonographic findings are representative of an uT2NO tumor. (b) Computed tomography scan showing the rectal tumor confined
to the rectal wall. (c) Double contrast barium enema showing an irregular surface of the rectal wall due to tumor infiltration
(arrows). (d) Macroscopic view. The nearly circumferential lesion has some villous features but the base is hard
 80 Atlas of Endoanal and Endorectal Ultrasonography

b c

Stage uT2. Case 3. Tumor in a 7SY male located in the left lateral rectal wall. (a) Sonography shows that the lateral borders of
the tumor (T) appear to be a villous adenoma. The central portion shows destruction of the hyperechoic layer corresponding to
the submucosa and thickening of the hypo echoic layer representing the muscolaris propria. These findings represent tumor
invasion. In the villous portion of the tumor these layers remain intact. The hyperechoic layer which corresponds to the perirec-
tal fat is intact. A large lymph node about 0.8 cm in diameter is seen proximal to the tumor near the left lateral rectal wall. This
lymph node is hypoechoic and has an irregular outline consistent with tumor metastasis. (b) Endoscopic view. (c) Macroscopic
view
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 81

Stage uT2. Case 4. Tumor in a 65Y male located in the right

lateral rectal wall. (a) Sonographic findings correspond to
those of a T2 tumor (T). One lymph node with a high proba-
bility of metastatic disease is visible adjacent to the tumor
(arrow). (b) Double contrast barium enema showing an irreg-
ular outline of the rectal wall due to tumor infiltration
(arrows). (c) Macroscopic view of the tumor with central
C L-______________________________________ ~

ulceration
 82 Atlas of Endoanal and Endorectal Ultrasonography

I Submucosa I

a
b

Stage uT2. Case 5. Tumor in a 42Y male located in the right lateral rectal wall at the level of the prostate. (a) Sonographic find-
ings correspond to those of a T2 tumor. (b) Double contrast barium enema image

Stage uT2. Case 6. Tumor in a 64Y female located in the ante-

rior rectal wall. (a) Sonographic findings correspond to those
of a T2 tumor. (b) Endoscopic view b
 IY.S.
Stage uT3: Perirectal fat invasion
G.A. Santoro, G. Di Falco

Perirectal fat invasion is diagnosed sonographi- with accuracy since nodes are often located prox-
cally by the presence of irregularities of the outer imal to the tumor. The incidence of regional
hyperechoic layer, which correspond to the lymph node metastases in uT3 tumors is approxi-
perirectal fat interface (Figs. IV.27-29). These mately 30% to 50% [4].
findings should be associated with disruption of
the hyperechoic layer corresponding to the sub-
mucosa and thickening of the hypo echoic layer
representing the muscolaris propria [1-3].
Contiguous organs are not involved. About 10% of
such tumors are unfortunately accompanied by a
narrowing of the lumen and angulation and it
may be difficult or impossible to advance the
probe proximal to the tumor (Fig. IY.30). To per-
form a complete staging by ERUS a residual
lumen of 2 cm is necessary because only those
structures seen at a 90° angle to the probe can be
assessed correctly. Under these circumstances the
study may be incomplete and the presence of
enlarged lymph nodes may not be ascertained

Fig. IY.28. Sonographic diagnosis of tumor invasion of the

Fig. IY.27. Diagrammatic representation of T3 rectal tumor perirectal fat (uT3) is based on irregularity of the outer hyper-
infiltrating into the perirectal fat echoic interface
 84 Atlas of Endoanal and Endorectal Ultrasonography

Fig. IY.30. It may be impossible to advance the probe proxi-

mal to large, locally advanced tumor, accompanied by a nar-
rowing of the lumen

The recognition of perirectal fat invasion is an

important determination to select appropriate
patients for preoperative combined chemothera-
py and radiation therapy followed by surgery. One
of the most important drawbacks in endosono-
graphic staging is the distinction between the T2
tumor invading most of the muscolaris propria
and the T3 tumor, which slightly invades the
perirectal fat. Indeed most errors are understag-

c
Fig. IY.31. T2 tumor overstaged as T3. Endorectal ultrasound
Fig. IV.29. Sonographic diagnosis of tumor invasion of the scan shows irregularity of the hyperechoic perirectal fat inter-
perirectal fat (uT3) is based on irregularity of the outer hyper- face (arrows). Histopathologic analysis demonstrated no
echoic interface (a-c) extension into the perirectal fat
 Section IV· Endoluminal ultrasound in the preoperative staging of rectal carcinoma 85

Fig. IV.32. Endoluminal ultrasound showing a rectal tumor

invading the hyperechoic layer representing the perirectal fat
(arrows). Adjacent to the tumor is a large lymph node appear-
ing as a possible nodal metastases (uT3Nl) (a). Computed
tomography scan showing the same lesion confined to the
a rectal wall (b)

ing of small pT3 tumors or overstaging of pT2 operative staging of rectal cancer. In pT3 tumors,
tumors (Fig. Iv'31). Glaser et al. [4) reported that accuracy rates were 87% by ERUS, 88% by CT and
the sensitivity of ERUS for detection of perirectal 89% by MRI (Fig. IV.32).
fat infiltration was 97%, specificity was 90%, neg- Three-dimensional (3D) ERUS offers a signifi-
ative predictive value was 98% and positive pre- cant advantage over conventional bi-dimensional
dictive value was 90%. Garcia-Aguilar et al. [5) (2D) ERUS for the accurate evaluation of rectal
reported that the overall accuracy rate in the cancer (Fig. IV.33). In a preliminary study, Kim et
assessment of rectal wall invasion for the T3 stage al. [7) showed that the accuracy of 3D ERUS was
by ERUS was 70%. Kim et al. [6) performed a 84.8% for pT3 whereas that of 2D-ERUS was
comparative study of ERUS, CT and MRI in pre- 75·8%.

a
 86 Atlas of Endoanal and Endorectal Ultrasonography

Fig. IV.33. Three-dimensional section

enables assessment of a T3 rectal tumor
in sagittal (a), coronal and transverse (b,
c) scan planes. The sagittal and transverse
planes reveal extension into the perirectal
c fat (arrows)

References 5. Garcia-Aguilar J, Pollack J, Lee S-H, Hernandez de

Anda E, Mellgren A, Wong WD, Finne CO,
1. Hildebrandt U, Feifel G. Preoperative staging of rectal Rothenberger DA, Madoff RD. Accuracy of endorectal
cancer by intrarectal ultrasound. Dis Colon Rectum ultrasonography in preoperative staging of rectal
1985;28:42-6 tumors. Dis Colon Rectum 2002;45:10-5
2. Boyce GA, Sivak MV, Lavery IC, Fazio VW, Church JM, 6. Kim NK, Kim MI, Yun SH, Sohn SK, Min JS.
Milsom J, Petras R. Endoscopic ultrasound in the pre- Comparative study of transrectal ultrasonography,
operative staging of rectal carcinoma. Gastrointest pelvic computerized tomography and magnetic reso-
Endosc 1992;38:468-71 nance imaging in preoperative staging of rectal cancer.
3. Bernick PE, Wong WD. Staging. What makes sense? Dis Colon Rectum 1999;42:770-5
Can the pathologist help? Surg Oncol Clin North Am 7. Kim IC, Cho YK, Kim SY, Park SK, Lee MG.
2000;9:703-23 Comparative study of three-dimensional and conven-
4. Glaser F, Schlag P, Herfarth C. Endorectal ultrasonog- tional endorectal ultrasonography used in rectal can-
raphy for the assessment of invasion of rectal tumors cer staging. Surg Endosc 2002;16;1280-5
and lymph node involvement. Br J Surg 1990;77:883-7
 Clinical cases

_ ____ b
a
Stage uT3. Case 1. Th.mor in a 45Y female located 7cm from the anal verge occupying about 40% of the rectal circumference
along the right lateral rectal wall. (a) The hyperechoic layer corresponding to the submucosa is interrupted and the hypoechoic
layer representing the muscolaris propria is thickened consistent with tumor invasion. The hyperechoic layer corresponding to
the perirectal fat interface is mostly intact although some irregularity (arrows) is present which represent focal tumor invasion.
(b) Sigmoidoscopy showing a polypoid ulcerated lesion

~ __________________________________ ~ b

Stage uT3. Case 2. Tumor in a 87 Y male located on the left lat-

eral rectal wall. (a) The hyperechoic line corresponding to the
submucosa is destroyed and the hypo echoic line representing
the muscolaris propria is thickened consistent with tumor
invasion. The hyperechoic layer corresponding to the perirec-
tal fat interface shows irregularity (arrows) characteristic of
a tumor invasion. (b) Macroscopic view
 88 Atlas of Endoanal and Endorectal Ultrasonography

~ ________________________________________ ~ e

Stage uT3. Case 3. Bulky tumor in a 52Y male present 6cm

from the anal verge occupying nearly 90% of the rectal cir-
cumference. (a) Deep penetration is represented by destruc-
tion of the hyperechoic layer corresponding to the submu-
cosa, marked thickening of the hypoechoic muscolaris pro-
pria and invasion of the hyperechoic perirectal fat interface.
One huge lymph node with high probability of nodal metas-
tases is visible adjacent to the tumor. (b) Computed tomogra-
phy scan showing the rectal tumor infiltrating the perirectal
fat (arrows). (c) Double contrast barium enema. The lumen is
narrowed with an irregular surface due to tumor infiltration.
(d) Sigmoidoscopy showing a locally advanced tumor, accom-
panied by a narrowing of the lumen. (e) Macroscopic view
showing a nearly circumferential lesion invading the perirec-
tal fat (arrows)
 Section IV, Endoluminal ultrasound in the preoperative staging of rectal carcinoma 89

b c

Stage uT3. Case 4. Tumor in a S6Y male located on the left lateral rectal wall. (a) The hyperechoic line corresponding to the sub-
mucosa is destroyed and the hypo echoic line representing the muscolaris propria is thickened consistent with tumor invasion.
The hyperechoic layer corresponding to the perirectal fat interface shows irregularity characteristic of tumor invasion. (b)
Endoscopic image. (c) Macroscopic view showing a lesion invading the perirectal fat (arrows)
 90 Atlas of Endoanal and Endorectal Ultrasonography

________________________________________ c
b
~ ~

Stage uT3. Case 5. Tumor in a 43Y female occupying half the rectal circumference. (a) The hyperechoic line corresponding to
the submucosa is destroyed and the hypoechoic line representing the muscolaris propria is thickened consistent with tumor
invasion. The hyperechoic layer corresponding to the perirectal fat interface shows irregularity characteristic of tumor invasion.
(b) Endoscopic image. (c) Macroscopic view
 IV.6.
Stage uT 4: Extensive local invasion
G.A. Santoro, G. Di Falco

uT 4 lesions are locally invasive into contiguous invasion of contiguous structures, spread to
organs such as bladder, uterus, cervix, vagina, regional nodes or distant metastases (Fig. IV.36).
prostate and seminal vesicles (Fig. IV.34). These The lateral pelvic lymph nodes, such as the obtura-
advanced lesions are clinically fixed or tethered. tor nodes, are located too far from the rectum to be
Sonographically there is a loss of the normal imaged effectively with rectal probes [41.
hyperechoic interface between tumor and the adja- Therefore, possible advantages of MRI and CT can
cent organ (Fig. IV.3S) [1,21. The inability of ERUS be considered in assessing the lateral pelvic lymph
to distinguish between malignant infiltration or nodes, pelvic wall invasion and involvement of lev-
peritumoral inflammation results in a somewhat ator ani muscle. The use of ERUS to assess extra-
lower staging accuracy with regard to T4 cancers. rectal pelvic invasion has been mentioned briefly
Frank stenosis also precludes precise endosono- in literature [41. It is still debatable if the results of
graphic evaluation, and angulation of the probe to a preoperative ERUS examination in patients with
the tumor axis also can cause misinterpretation. large rectal tumors can predict whether preopera-
Rectal cancer staging using a CT scan or MRI tive radiation would be beneficial. Future efforts
showed high accuracy rates for locally advanced focusing on common protocols and carefully
tumors (80% to 90%) [31. CT and MR imaging are defined data points are needed to precisely define
accurate in assessing spread beyond the rectal wall, the role of ERUS in pelvic disease management.

Fig. IV.34. Diagrammatic representation of locally advanced

T4 rectal tumor with deep invasion beyond the rectal wall
 92 Atlas of Endoanal and Endorectal Ultrasonography

a b

Fig. IV.35. Endoluminal ultrasound showing a rectal tumor invading beyond the rectal wall (arrows) (a). Three-dimensional
transverse section reveals extension beyond the perirectal fat (arrows) (b)

Fig. IV.36. Computed tomography scan showing a locally

advanced rectal tumor (T) invading into uterus (U). (B: blad-
der)

References

1. Hildebrandt U, Feifel G. Preoperative staging of rectal

The treatment of uT4 tumors requires com- cancer by intrarectal ultrasound. Dis Colon Rectum
bined modality chemotherapy, radiation therapy 1985; 28:42-6
and potential continuity organ resection. These 2. Bernick PE, Wong WD. Staging. What makes sense?
advanced rectal cancers have a poor prognosis Can the pathologist help? Surg Oncol Clin North Am
and are resected for cure in less than half of the 2000; 9:703-23
cases. The use of preoperative radiation therapy 3. Boyce GA, Sivak MV, Lavery IC, Fazio VW, Church JM,
can shrink the tumor and allow for increased Milsom J, Petras R. Endoscopic ultrasound in the pre-
resectability and decreased local recurrence rates. operative staging of rectal carcinoma. Gastrointest
Intraoperative radiation therapy when used in Endosc 1992; 38:468-71
combination with high-dose preoperative radio- 4. Milsom JW, Graffner H. Intrarectal ultrasonography in
therapy appears to improve local control of local- rectal cancer staging and in the evaluation of pelvic
ly advanced rectal cancers. disease. Ann Surg 1990; 2l2:602-6
 Clinical cases

b c
Stage uT4. Case 1. Large rectal cancer in a 57Y male occupying nearly 70% of the rectal circumference. (a) The tumor is most
extensive in the right rectal wall where it is up to 5cm in thickness and extends to the pelvic side wall. (b) Sigmoidoscopy show-
ing a locally advanced tumor, accompanied by a narrowing and distortion of the rectal lumen. (c) Double contrast barium
enema. The lumen is narrowed with an irregular surface due to tumor infiltration (arrows)
 94 Atlas of Endoanal and Endorectal Ultrasonography

a b
Stage uT4. Case 2. Extensive anterior rectal cancer in a 59Y male. (a) The sonographic layers of the rectal wall are completely
disrupted by the tumor. These findings correspond to those of an advanced deeply invasive rectal cancer extending through the
bowel wall. (b) Sigmoidoscopy showing a locally advanced tumor, accompanied by a narrowing and distortion of the rectal
lumen

Stage uT4. Case 3. (a) Large anterior rectal cancer located at

the level of the prostate. The normal sonographic layers are
no longer visible due to extensive tumor invasion. The tumor
appears to invade the prostatic capsule anteriorly (arrow).
Two lymph nodes with a moderate probability of nodal
metastases are visible adjacent to the tumor. (b) Computed
tomography scan showing tumor invading the prostate
b
 Invited commentary
D.M. Schaffzin, W.D. Wong

The authors of this section should be congratulat- Accuracy of ascertaining the depth of penetra-
ed on a well-presented summary of endoluminal tion by stage is variable. While reports for UTI
ultrasound staging of rectal carcinoma. Although lesions have been as low as 50%, a meta-analysis
preoperative staging of rectal cancer is not perfect, of multiple studies evaluating ELUS found UTI
we agree with the authors that, given current tech- lesions to be accurately staged 97% of the time [1,
nological options, endoluminal ultrasound (ELUS) 5]. This underscores the variability of this staging
provides an accurate and reproducible method of modality.
staging these lesions in order to stratify them A point to be emphasized is that ELUS is
according to the need for neoadjuvant chemora- operator dependant [6, 7]. This has been stressed
diotherapy, the option to perform local excision, or in many studies, and deserves special attention.
the decision to take the patient directly for radical Orrom et al. reviewed their experience with
surgery. Of the available technologies, namely com- ELUS [7], dividing their series into three time
puted tomography (CT), magnetic resonance periods. The first time period had instituted no
imaging (MRI), endoluminal coil MRI (EL-MRI), standardization in technique or training for
and ELUS, the latter is an office-based procedure of ELUS; accuracy was found to be only 58%, with
short duration, and is well tolerated by patients. EL- significant numbers of tumors being either over-
MRI requires patients to be in a prone position staged or understaged. The second time period
with the probe for a lengthy period of time, which instituted standardization in technique, includ-
would seem to be associated with a fair degree of ing the use of a rigid proctoscopy, and perfor-
discomfort. Given that the EL-MRI and ELUS are mance or overseeing of performance of ELUS by
essentially equivalent with respect to accuracy, sen- a single physician. Accuracy in this group
sitivity, and specificity, the choice between ELUS improved to 77')10. The third time period stan-
and EL-MRI seems easy [1]. An advantage of ELUS dardized interpretation of the ELUS based on the
is that it provides an opportunity to take a biopsy at now agreed-upon 5-1ayer model. Accuracy here
the time of the procedure. In addition, if the equip- improved to 95%.
ment is available, ELUS can be performed during Garcia-Aguilar et al. also evaluated perfor-
any given office visit whether it has been scheduled mance based on which surgeon performed the
or not. An advantage of EL-MRI, however, is that it exam (each surgeon conducting more than 50
allows visualization of adjacent structures and the examinations) [6]. For accuracy in evaluation of
relationship of the tumor to these structures in depth of penetration, there was a statistically sig-
multiple planes, utilizing multiple techniques [2]. nificant difference even among these experienced
The advent of three-dimensional ELUS (3D-ELUS) observers.
may overcome this potential deficit, which current- The deficits in ELUS are based in part on diffi-
lyexists in two-dimensional ELUS [2]. CT has con- culties distinguishing between certain uT2 and
sistently been found to be less accurate than ELUS uT3 tumors. The scalloping of the outer border
or EL-MRI [1,3,4]. seen in deeper uT2 lesions means that these
96 Atlas of Endoanal and Endorectal Ultrasonography

lesions are often overstaged as uT3, decreasing residual air and fluid prior to introducing the
accuracy of staging for these levels [8]. Because T3 probe.
lesions are associated with a higher incidence of At MSKCC, we have attempted to improve the
lymph node metastases, this distinction is impor- treatment selection criteria of rectal cancer by
tant in the decision to utilize neoadjuvant modifying the ELUS staging system. We have
chemoradiation [9, 10]. The distinction between stratified both T2 and T3 lesions respectively
uT3 and uT4 is also potentially prone to over- or into deep and superficial categories [14]. The
understaging. This, however, is of lesser impor- superficial ELUS correlates well with pathologic
tance with regard to therapy, as the treatment superficial T2 tumors, defined as occupying less
options are much the same [11]. than 30% of the muscle thickness. Superficial T3
The well know study by Garcia-Aguilar et al. is tumors are defined by ELUS as having radial
one of the largest to evaluate the accuracy of ELUS extension into the perirectal fat of less than 2
in the preoperative staging of rectal cancers. The mm. The pathologic correlate is radial extension
overall accuracy was significantly lower than in of less than 4 mm, which also correlates well with
many other published studies for both T-stage the ELUS findings. Node positivity for deep T3
(69%) and N-stage (64%) [6]. However, the exclu- lesions is 70%, versus 30% for superficial T3
sion of patients who underwent neoadjuvant ther- lesion. This system may allow for improved
apy, and thus the exclusion of most stage II and III patient selection for local excision and for
lesions, appeared to have had a significant detri- neoadjuvanttherap~
mental effect on these numbers. Accuracy for dif- We have proposed a modification of the
ferentiating benign (TO) and malignant (Tl) uTNM system (Table IV.3) in the hopes that this
lesions, however, was found to be 87%. might help clarify clinical considerations related
ELUS is also accurate in staging the nodal sta- to tumor stage. The stratification of uT3 tumors
tus of rectal cancers. The overall accuracy ranges into superficial and deep categories allows for
from 65 to 88%, with overstaging ranging from 3 grouping of lesions into a treatment -oriented
to 25% and understaging ranging from 0 to 26% staging system. For example, uTO and UTI lesions
[2, 6, 12, 13]. Some of the reasons for these dis- can be distinguished by endorectal ultrasound as
crepancies in staging may be related to the use of one group which may be amenable to local exci-
different size criteria to determine nodal status, sion, whereas most UT2 and select superficial T3
the differentiation between benign or inflamma- lesions constitute a second group that is
tory nodes versus malignant ones, and the amenable to radical resection without neoadju-
inability of ELUS to detect micrometastases. Size vant therapy. The third group that can be accu-
criteria vary from as low as 3 mm to 10 mm or rately categorized by endorectal ultrasound
larger, with accompanying variable sensitivities comprises deep T3 and all T4 lesions, which are
and specificities. At Memorial Sloan-Kettering best treated with neoadjuvant therapy followed
Cancer Center (MSKCC) we tend to utilize 5 mm by radical resection. Hence, the stratification of
as our lower limit. Other factors that indicate T3 lesions into superficial and deep constitutes
probable malignancy include irregular borders the major advantage of this new sub classifica-
and length/width discrepancies. Additionally, an tion. However, the ability of endorectal ultra-
inflammatory lymph node often has a hypere- sound to stratify T2 lesions into superficial and
choic center. deep, allows for consideration of local excision of
Technical considerations also improve the some superficial T2 lesions identified in moder-
accuracy of ELUS. The authors note that filling the ate-to-high-risk patients, even though this might
rectal lumen with fluid may decrease mechanical not be the optimal treatment in an otherwise
distortion by the balloon. We utilize this method, good-risk patient.
switching to the plastic cap in the fluid medium to We are currently exploring new modalities of
best visualize the tumor. In addition, the use of the staging utilizing 3D-ELUS. There are multiple
proctoscope is invaluable. A subtle point to be methods of rendering the image to accentuate the
made, though, is to not insufflate much air during layers of the rectal wall. Our hope is that, in lesions
proctoscopy, and to decompress the lumen before that are borderline UT2 or uT3, these methods
introducing the ELUS probe, as this adds signifi- might improve the accuracy of ELUS in the areas
cant artifact. We often introduce a red rubber where it is now most prone to overstaging and
catheter into the lumen to further decompress understaging.
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 97

Table Iv'3. Proposed modification of ELUS staging system

uTO Confined to mucosa

UTI To but not through submucosa

UT2 Into but not through muscularis propria

uT3 Through bowel wall into perirectal fat

uT4 Involving adjacent structures

uNO No definable lymph nodes by ultrasound

UN1 Ultrasonographically apparent lymph nodes

uTw: uTOI UTI Amenable to local excision

uTy: uT2/superficial uT3 Recommend radical surgery

uTz: deep uT3/any uT 4 Recommend neoadjuvant therapy followed by radical surgery

UN1: probable or definite Recommend neoadjuvant therapy followed by radical surgery

UN1: equivocal Base treatment on T stage and pathologic features

In conclusion, ELUS is currently the best modal- tion with other modalities such as MRI or PET.
ity available in the preoperative staging of rectal However, until further improvements are made, the
cancer. Future improvements may include integra- speediest and best tolerated modality is ELUS.

References 8. Akbari RP, Wong WD. Endorectal ultrasound and the

preoperative staging of rectal cancer. Scand J Surg
1. Kwok H, Bissett IP, Hill GL. Preoperative staging of rec- 2003;92:25-33
tal cancer. Int J Colorectal Dis 2000;lS:9-20 9. Sitzler PJ, Seow-Choen F, Ho YH, Leong AP. Lymph
2. Hunerbein M, Pegios W, Rau B, YogI TJ, Felix R, Schlag node involvement and tumor depth in rectal cancers:
PM. Prospective comparison of endorectal ultrasound, an analysis of 80S patients. Dis Colon Rectum 1997;
three-dimensional endorectal ultrasound, and endorec- 40:1472-6
tal MRI in the preoperative evaluation of rectal tumors. 10. Janjan NA, Khoo YS, Abbruzzese J et al. Tumor down-
Preliminary results. Surg Endosc 2000;14:100S-9 staging and sphincter preservation with preoperative
3. Herzog U, von Flue M, Tondelli P, Schuppisser JP. How chemoradiation in locally advanced rectal cancer: the
accurate is endorectal ultrasound in the preoperative M. D. Anderson Cancer Center experience. Int J Radiat
staging of rectal cancer? Dis Colon Rectum 1993;36: Oncol BioI Phys 1999;44:1027-38
127-34 11. Rivadeneira DE, Wong WD. Preoperative staging of

4. Rifkin MD, Ehrlich SM, Marks G. Staging of rectal car- rectal cancer. Clin Colon Rectal Surg 2001;15:17-25
cinoma: prospective comparison of endorectal US and 12. Kim NK, Kim MJ, Yun SH, Sohn SK, Min JS.
CT. Radiology 1989;170:319-22 Comparative study of transrectal ultrasonography,
5. Marusch F, Koch A, Schmidt U et al. Routine use of tran- pelvic computerized tomography, and magnetic reso-
srectal ultrasound in rectal carcinoma: results of a nance imaging in preoperative staging of rectal cancer.
prospective multicenter study. Endoscopy 2002;34:38S-90 Dis Colon Rectum 1999;42:770-5
6. Garcia-Aguilar J, Pollack J, Lee SH et al. Accuracy of 13. Adams DR, Blatchford GJ, Lin KM, Ternent CA,
endorectal ultrasonography in preoperative staging of Thorson AG, Christensen MA. Use of preoperative
rectal tumors. Dis Colon Rectum 2002;4S:10-5 ultrasound staging for treatment of rectal cancer. Dis
7. Orrom WJ, Wong WD, Rothenberger DA, Jensen LL, Colon Rectum 1999;42:1S9-66
Goldberg SM. Endorectal ultrasound in the preopera- 14. Nissan A, Shia J, Klimstra D et a1. Can endorectal ultra-
tive staging of rectal tumors. A learning experience. sound stratify T-staging of primary rectal cancer?
Dis Colon Rectum 1990;33:654-9 Booth P23. Dis Colon Rectum 2001;44:A27-AS9
 IV.7.
Stage uN1: Lymph node metastases
G.A. Santoro, G. Di Falco

Metastatic involvement of the mesorectal lymph The lack of a correlation between nodal isolated
nodes is a major independent prognostic factor. It tumor cells and either recognized prognostic indi-
has been observed that the presence of >3 nodes is cator or survival confirms that they do not repre-
associated with a poor prognosis [1]. Moreover, sent true metastases.
identification of a metastatic perirectal lymph The currently published data support the
node is important as these patients may benefit notion that ERUS is superior to other modalities in
from preoperative adjuvant radiotherapy, and assessing perirectal lymph node involvement,
some of the early Tl or T21esions with mesorectal although such an assessment requires much greater
node involvement are not suitable for local exci- operator experience than for assessment of tumor
sion. The term micrometastasis has been in wide infIltration in the bowel wall. Sonographic evalua-
use to represent a small focus of tumor cells often tion of lymph node metastases is somewhat less
not identified because of the way in which the accurate than depth of invasion. The accuracy of
lymph node was sampled. It is essential to differen- ERUS in assessing lymph node involvement varies
tiate between isolated tumor cells and micrometas- from 58% to 86% (Table IY.4) [3-19]. Undetectable
tases, real metastases 0.2 cm or less in greatest or benign appearing lymph nodes are classified as
dimension. As very well stated by Hermanek et al. uNo. Malignant appearing lymph nodes are classi-
[1] "this terminology may be quite a misnomer in a fied as uN!. Normal, non-enlarged perirectal nodes
sense because only in metastases has there been are not usually seen on ERUS. The criteria used to
arrest and implantation of tumor cells in the organ identify metastatic lymph nodes in most of the
involved, with extravasation, proliferation and studies are echogenicity, border demarcation and
often a stromal reaction. In contrast isolated tumor node diameter [20-21]. Inflamed, enlarged lymph
cells have no contact with vessels or lymph sinus nodes appear hyperechoic, with ill-defined bor-
walls or extravasation or extravascular prolifera- ders. Much of the sound energy is reflected because
tion". Therefore the term isolated tumor cells, pre- the lymphatic tissue has not changed (Fig. IY.37). In
ferred by the IUAC is appropriate to represent contrast, metastatic lymph nodes that have been
occult tumor cells. Choi et al. [2] performed a study replaced with tumor do not provide the normal tis-
to investigate whether presence of isolated tumor sue architecture and appear hypo echoic with an
cells in lymph nodes has a prognostic significance. echogenicity similar to the primary tumor (Figs.
There was no difference in five-year survival IY.38, 39). Malignant lymph nodes tend to be circu-
between positive and negative groups for isolated lar rather than oval, have discrete borders and are
tumor cells (82.8% and 85.9%, respectively). From most commonly found adjacent to the primary
these results it is speculated that nodal isolated tumor or in the mesorectum proximal to a tumor
tumor cells may represent tumor cells, which have (Figs. IY.40-43). Rifkin and Wechsler [24] suggest-
gained access to the lymphatics, but which are des- ed regarding all endosonographic discernible
tined to be destroyed by the host's immune mech- lymph nodes as potentially malignant. Doing so
anism before implantation and growth can occur. they attained a surprisingly high specificity of 91%
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 99

Table IV.4. Accuracy of endorectal ultrasonography (ERUS) in determining lypmh node metastases from rectal
tumors

Author Sensitivity Specificity Accuracy Positive Negative

predictive value predictive value

Hildebrandt [3] 72% 83% 78%

Herzog [4] 89.2% 73-4% 80.2% 71.2% 9°·4%
Sentovich [5] 100% 100% 73%
Fleshman [6] 37·5% 100%
Orrom [7] 88% 90% 88%
Glaser [8) 78% 80% 79% 76% 82%
Holdsworth [9) 57% 64% 61% 5°% 70%
Milsom [10) 81% 80% 7°% 74% 86%
Saitoh [11) 82% 73%
Rifkin [12) 6?,Yo 91%
Garcia-Aguilar [13) 33% 64% 64% 52% 68%
Akasu [14] 100% 60% 65% 26% 100%
Jochem [15) 9°% 66% 72% 47% 95%
Frascio [16) 74% 73% 75%
Kim [17) 53·3% 75% 63·5% 70.6% 58.8%
Napoleon [18] 75% 92% 85% 86% 85%
Santoro [19) 70% 79% 74% 72% 84%

a b

Fig. IV.37. Sonograms of benign appearing lymph nodes

(arrows) (a-c). Lymph nodes appear hyperechoic, with ill-
c defined borders
 100 Atlas of Endoanal and Endorectal Ultrasonography

a a

b
Fig. IV.39. (a) Malignant lymph node (In) tend to appear with
an echogenicity similar to the primary tumor (T). The corre-
sponding micro section of the lymph node confirms the
metastatic involvement (b), (H&E stain)

Fig. IY.38. Sonogram of an enlarged, hypoechoic lymph node

(In) appearing as a possible nodal metastasis (a). The corre-
sponding micro section of the lymph node confirms the
destruction of the tissue architecture by tumor cells (H&E
stain)

and a sensitivity of 67%. Holdsworth et al. [9] car-

ried out endorectal sonography by means of a 5.5
MHz transducer. They identified lymph node
metastases with a sensitivity of 57% and a specifici-
ty of 64% and concluded that the technique was not
reliable to identify metastases. Feifel et al. [25] have
demonstrated that lymph nodes cannot be evaluat- Fig. IV.40. Extensive deeply infiltrating rectal cancer. The
ed for metastases with the low resolution of 5.5 hyperechoic layer corresponding to the perirectal fat interface
is irregular consistent with tumor invasion (UT3). Adjacent to
MHz. With a resolution of 7.5 MHz however the
the tumor is a large lymph node which is 11 mm in diameter.
accuracy rate is significantly higher. According to Node is hypo echoic with irregular borders with a high prob-
Hildebrand et al. [3], the sonographic features of ability of metastatic disease
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 101

Fig. IV.4I. Rectal cancer 7 em from the anal verge. The hyper-
echoic layer corresponding to the perirectal fat interface is
irregular consistent with tumor invasion (uT3). Two lymph
nodes ranging from 0.5 to 1 em in diameter are visible. They
appear with an echogenicity similar to the primary tumor and
have a high probability of containing metastatic disease.

b
Fig. IV.43. Sonographic view of lymph nodes (In) in two cross
sections at different levels (a, b). Malignant lymph nodes are
commonly found in the mesorectum proximal to a tumor (T)

lymph nodes generally can be distributed into four

groups. If lymph nodes are not visible by ultra-
sound, the probability of lymph node metastases is
low. Hyperechoic lymph nodes are often benign
and result from non-specific inflammatory
changes. Hypoechoic lymph nodes larger than 5
mm are highly suggestive of lymph node metas-
tases. Lymph nodes larger than 5 mm that are visi-
Fig. IV.42. Tumour located in the posterior rectal wall. The ble with mixed echogenic patterns cannot be clas-
hypoechoic layer representing the muscolaris propria is sified accurately but should be considered
markedly thickened. The hyperechoic layer corresponding to metastatic. On size characteristics alone, sono-
the perirectal fat interface is intact (uT2). One lymph node
5mm in diameter, circular in outline and hypoechoic is visible
graphically detected nodes in the mesorectum
adjacent to the tumor. These findings correspond to a moder- greater than 5 mm in diameter have a 50-70%
ate probability of nodal metastasis chance of being involved, whereas those smaller
102 Atlas of Endoanal and Endorectal Ultrasonography

a ratio of long axis to short axis diameter, referred

to as the roundness index, lobulations (multiple
notches), echogenicity, inhomogeneity (not uni-
form), border delineation, presence of an echo-
poor rim (the outer rim being more hypoechoic
than the rest of the node), presence of a peripheral
halo and presence of a hilar reflection. The authors
showed that 3 ultrasonographic features of a node
significantly related to its being benign or malig-
nant at histopathologic examination are short axis
diameter, degree of inhomogeneity and presence or
absence of hilar reflection. Overstaging and under-
staging can occur during assessment of lymph
node involvement. Edematous lymph nodes trans-
mit more sound energy and appear in echo pat-
terns which are similar to metastases. The cross-
sectional appearance of blood vessels in the
Fig. IY.44. Sonogram shows two lymph nodes (In) 3 mm in perirectal fat may be confused commonly with pos-
diameter adjacent to the primary tumor (T). Pathologic find- itive lymph nodes (Fig. IV.45). The sonographic
ing confirmed a micrometastases present in one node
continuity of hypo echoic vessels over a distance
greater than the cross-sectional diameter is the cri-
terion used to distinguish vessels from hypoechoic
than 4 mm have a less than 20% chance. However lymph nodes. With careful scanning, blood vessels
up to 20% of patients have involved nodes of less appear to branch or extend longitudinally. In addi-
than 3 mm, limiting the accuracy of the technique tion, it may be difficult to differentiate islands of
(Fig. IV.44). Size alone is not an adequate criterion tumor outside the bowel wall from involved nodes.
for defining lymph node involvement. Akasu et al. With careful scanning, one can demonstrate conti-
[14] found that approximately 50% of lymph nodes nuity with the main tumor that may not have been
with diameters between 3 and 5mm demonstrated recognized initially. Even with an improved under-
metastases, which makes prediction based on size standing of the characteristics of malignant lymph
difficult. Herrera-Ornelas et al. [26] found that two nodes and utilizing criteria of shape, echogenicity
thirds of metastatic lymph nodes in colorectal can- and border character, micrometastases and granu-
cer pathologic specimens were smaller than 5 mm lomatous inflammation will probably be difficult if
in diameter. Katsura et al. [27] found that 18% of not impossible to differentiate by ERUS. If a whole
lymph nodes less than 4 mm in diameter contained node is replaced by a tumor or the node is enlarged
metastases. Hildebrandt et al. [3] reported that by secondary to it, detection is more likely. However, if
interpreting hypo echoic and hyperechoic nodes as only a small deposit or a micrometastases is pre-
involved and uninvolved, respectively, they could sent in a node, the characteristics of the node are
obtain a specificity of 83% and a sensitivity of 72%. unlikely to be sufficiently altered to allow detection.
In another study Hildebrandt et al. [22] showed an This explains in part the lower accuracy rates for
accuracy of 74% in detecting mesorectal lymph lymph nodal detection with current, conventional
nodes in 27 patients. Sunouchi et al. [28] studied ultrasonography. Grossly malignant lymph nodes
hypo echoic lesions larger than 5 mm and found located at a distance from the primary tumor also
that 20% were tumor deposits and 68% were remain undetected if they exceed the depth of pen-
metastatic lymph nodes. Beynon et al. [23], based etration of the transducer. This is particularly true
on identification of circular or oval echo-poor of nodes in the proximal mesorectum above the
lesions in the mesorectum, obtained a sensitivity of reach of the rigid probe. To obtain high sensitivity
86% and a specificity of 79% for detection of and high specificity, the combination of a small cut-
involved nodes. Hulsmans et al. [29] studied sever- off value and ERUS-guided needle biopsy or ERUS-
al features by correlating pathologic and sono- guided fine-needle aspiration biopsy may be help-
graphic findings in the lymph nodes of specimens ful (Fig. IV.46). Milsom et al. [10] reported the fea-
obtained from a series of 21 consecutive patients sibility and safety of needle biopsy of parectal
with resected rectal cancer. These features included lymph nodes under ERUS guidance. A 7.0 MHz lon-
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 103

Fig.IY.45. Sonographic view of blood vessels in two cross sections at different levels. The continuity of hypo echoic vessels over
a distance greater than the cross- sectional diameter is the criterion used to distinguish vessels from hypoechoic lymph nodes.
With careful scanning, blood vessels appear to branch or extend longitudinally

a L-________________________________________ ~~L_ __ ~

Fig. IV.46. Diagrammatic representation

of a trans-perineal needle biopsy per-
formed using the B-K Medical 1850 endo-
b ~ ____ ~~ ____________ ~ ____________________________ ~ probe (a, b).
 104 Atlas of Endoanal and Endorectal Ultrasonography

Fig. IY.47. Diagrammatic representa-

tion of a trans rectal biopsy performed
through a built-in needle guide using
B-K Medical 8551 endoprobe

gitudinally oriented endoluminal probe (model demonstrates enlarged nodes but does not pro-
8551, Bruel & Kjaer sector scanner encompassing a vide criteria that allow tissue differentiation with-
110° field of imaging) was used to locate the lymph in nodes. Civelli et al. [301 reported that in the
nodes. This probe is equipped with a needle guide, identification oflymph node metastases, endorec-
allowing for a passage of an 18-gauge, 21 em long tal balloon computed tomography had 52.6% sen-
spring-loaded core biopsy needle through the sitivity, 85.3% specificity and 73-6% accuracy.
probe and into the pararectal tissues under direct Goldman et al. [31] compared CT with ERUS and
ultrasound guidance (Fig. IV.47). The overall accu- found accuracy rates of 64% and 68%, respective-
racy rate was 77%, with a sensitivity of 71%, a speci- ly, for lymph node involvement. Similarly, Beynon
ficity of 89%, a positive predictive value of 92% and et al. [231 showed that ERUS was significantly
a negative predictive value of 62% for prediction of more accurate than CT for lymph node involve-
lymph node metastases from rectal cancer. ment. Other investigators also have demonstrated
The accuracy of ERUS in the detection of a superior accuracy and reliability of ERUS com-
lymph node metastases compares favorably with pared with CT. ERUS is minimally invasive and
the accuracy of CT. Computed tomography less expensive than CT. In the evaluation of lymph

a b

Fig. IY.48. Endoluminal ultrasound showing a large lymph node appearing as a possible nodal metastases adjacent to a rectal
tumor invading the hyperechoic layer representing the perirectal fat (UT3Nl) (a). Magnetic resonance image showing the same
tumor with two pararectal lymph nodes (arrow) (b)
 Section IV· Endoluminal ultrasound in the preoperative staging of rectal carcinoma 105

Fig. IV.49. Three-dimensional section

enables assessment of a T3 rectal tumor
in coronal, sagittal and transverse scan
planes. The transverse planes reveal pres-
ence of two additional malignant lymph
b
nodes (arrows) (a, b)

nodes, MRI does not offer significant improve- Three-dimensional (3D) imaging offers
ment in accuracy rates compared with ERUS (Fig. improved knowledge of various anatomic struc-
IV.48). It is unlikely, however, that MRI will gain tures and tumors. Additional scan planes (longi-
widespread usage because of lack of widespread tudinal and coronal) can be used to determine the
availability and significantly increased financial size, location and local extent of the lesion pre-
costs. One advantage of endorectal MRI is the cisely. Kim et al. [32] performed a prospective
ability to visualize the normal and pathological study to verify whether 3D ERUS enhances the
anatomy in multiple scan planes. accuracy of rectal staging, as compared with 2D
 106 Atlas of Endoanal and Endorectal Ultrasonography

Fig. IY.SO. Three-dimensional endorectal

ultrasound showing malignant lymph
nodes (arrows) (a, b)
b

ERUS. The lymph node metastases were accurate- scopic visualization provided easier and complete
ly predicted by 3D ERUS in 84.8% of patients, understanding of both focal lesions and lymph
whereas 2D ERUS predicted the disorder in nodes (Figs. IV.49, 50). Additional studies are nec-
66.7%. The difference was not statistically signifi- essary to further evaluate the efficacy of 3D ERUS
cant. Although the findings did not show 3D ERUS and to develop imaging protocols for special clin-
to have a significant advantage over 2D ERUS for ical situations.
the accurate evaluation of rectal cancer, stereo-
 Section IV' Endoluminal ultrasound in the preoperative staging of rectal carcinoma 107

References prospective comparison of transrectal ultrasonogra-

phy and computed tomography. Gastroenterology
1. Hermanek P, Hutter RV, Sobin LH, Wittekind Ch. 1990;98:268
International Union Against Cancer. Classification of 15. Jochem RJ, Reading CC, Dozois RR, Carpenter HA, Wolff
isolated tumor cells and micrometastasis. Cancer BG, Charboneau Jw. Endorectal ultrasonograpic stag-
1999;86:2668-73 ing of rectal carcinoma. Mayo Clin Proc 1990;65:1571-7
2. Choi HJ, Choi YY, Hong SH. Incidence and prognostic 16. Frascio F, Giacosa A. Role of endoscopy in staging col-
implications of isolated tumor cells in lymph nodes orectal cancer. Semin Surg Oncol 2001;20:82-5
from patients with Dukes B colorectal carcinoma. Dis 17. Kim HJ, Wong WD. Role of endorectal ultrasound in
Colon Rectum 2002;45: 750-6 the conservative management of rectal cancers. Semin
3. Hildebrandt U, Klein G, Schwarz HP. Endosonography Surg Onco12000;19:358-66
of parectallymph nodes. In vitro and in vivo evalua- 18. Napoleon B, Pujol B, Berger F, Valette PJ, Gerard JP,
tion. Dis Colon Rectum 1990;33:863 Souquet JC. Accuracy of endosonography in the stag-
4. Herzog U, von Flue M, Tondelli P, Schuppisser JP. How ing of rectal cancer treated by radiotherapy. Br J Surg
accurate is endorectal ultrasound in the preoperative 1991;78:785-8
staging of rectal cancer? Dis Colon Rectum 19. Santoro GA, Pastore C, Barban M, Di Falco G.
1993;36:127-34 Endorectal ultrasonography for the assessment of
5. Sentovich S, Blatchford G, Falk P, Thorson A, pararectal lymph node involvement in rectal cancer.
Christensen M. Transrectal ultrasound of rectal Hepato-Gastroenterology 2001;48 (SI):CXIX
tumors. Am J Surg 1993;166:638-41 20. Heriot AG, Grundy A, Kumar D. Preoperative staging
6. Fleshman JW, Myerson RJ, Fry RD, Kodner lJ. Accuracy of rectal carcinoma. Br J Surg 1999;86:17-28
of transrectal ultrasound in predicting pathologic 21. Kumar A, Scholefield JH. Endosonography of the anal
stages of rectal cancer before and after preoperative canal and rectum. World J Surg 2000;24:208-15
radiation therapy. Dis Colon Rectum 1992;35:823-9 22. Hildebrandt U, Feifel G, Schwarz HP, Scherr o.
7. Orrom WJ, Wong WD, Rothenberger DA, Jensen LL, Endorectal ultrasound: instrumentation and clinical
Goldberg SM. Endorectal ultrasound in the preopera- aspects. Int J Colorectal Dis 1986;1:203-7
tive staging of rectal tumors. A learning experience. 23. Beynon J, Mortensen NJ, Foy DM, Channer JL, Rigby
Dis Colon Rectum 1990;33:654-9 H, Virjee J. Preoperative assessment of mesorectal
8. Glaser F, Schlag P, Herfarth C. Endorectal ultrasonog- lymph node involvement in rectal cancer. Br J Surg
raphy for the assessment of invasion of rectal tumors 1989;76:276
and lymph node involvement. Br J Surg 1990; 77:883-7 24. Rifkin MD, Wechsler RI. A comparison of computed
9. Holdsworth PJ, Johnston D, Chalmers AG, Chennells P, tomography and endorectal ultrasound in staging rec-
Dixon MF, Finan PJ, Primrose IN, Quirke P. tal cancer. Int J Colorect Dis 1986;1:219-23
Endoluminal ultrasound and computed tomography 25. Feifel, G, Hildebrandt U, Dhom G. Die endorektale
in the staging of rectal cancer. Br J Surg 1988;75:1019-22 sonographie beim rektumkarzinom. Chirurg 1985;56:
10. Milsom JW, Czyrko C, Hull TL, Strong SA, Fazio VW. 398-402
Preoperative biopsy of pararectal lymph nodes in rec- 26. Herrera-Ornelas L, Justiniano J, Castillo N, Petrelli NJ,
tal cancer using endoluminal ultrasonography. Dis Stulc JP, Mittelman A. Metastases in small lymph
Colon Rectum 1994;47:364-8 nodes from colon cancer. Arch Surg 1987;122:1253-6
11. Saitoh N, Okui K, Sarashina H, Suzuki M, Arai T, 27. Katsura Y, Yamada K, Ishizawa T, Yoshinaka H,
Nunomura M. Evaluation of ecographic diagnosis of Shimazu H. Endorectal ultrasonography for the
rectal cancer using intrarectal ultrasonic examination. assessment of wall invasion and lymph node metas-
Dis Colon Rectum 1986;29:234-42 tases in rectal cancer. Dis Colon Rectum 1992;35:362-8
12. Rifkin MD, Ehrlich SM, Marks G. Staging of rectal car- 28. Sunouchi K, Sakaguchi M, Higuchi Y, Namiki K, Muto
cinoma: prospective comparison of endorectal US and T. Small spot sign of rectal carcinoma by endorectal
CT. Radiology 1989;170: 319-22 ultrasonography: histologic relation and clinical
13. Garcia-Aguilar J, Pollack J, Lee S-H, Hernandez de impact on postoperative recurrence. Dis Colon
Anda E, Mellgren A, Wong WD, Finne CO, Rectum 1998;41:649-53
Rothenberger DA, Madoff RD. Accuracy of endorectal 29. Hulsmans FJ, Bosma A, Mulder PJ, Reeders JW, Tytgat
ultrasonography in preoperative staging of rectal GN. Perirectal lymph nodes in rectal cancer: in vitro
tumors. Dis Colon Rectum 2002;45:10-5 correlation of sonographic parameters and histologic
14. Akasu T, Sunouchi K, Sawada T, Tsioulias GJ, Muto T, findings. Radiology 1992;184:553
Morioka Y. Preoperative staging of rectal carcinoma: 30. Civelli EM, Gallino G, Mariani L, Cozzi G, Biganzoli E,
108 Atlas of Endoanal and Endorectal Ultrasonography

Salvetti M, Gallo R, Belli F, Bonfanti G, Bertario L, computed tomography in preoperative staging of

Andreola S, Leo E. Double-contrast barium enema and lower rectal adenocarcinoma. Gastrointestinal Radiol
computerised tomography in the preoperative evalua- 1991;16:259-63
tion of rectal carcinoma: are they still useful diagnos- 32. Kim JC, Cho YK, Kim SY, Park SK, Lee MG. Comparative
tic procedures? Tumori 2000;86:389-92 study of three-dimensional and conventional endorec-
31. Goldman S, Arvidson H, Normig U, Lagerstedt U, tal ultrasonography used in rectal cancer staging. Surg
Magnusson I, Frisell J. Transrectal ultrasound and Endosc 2002;16:1280-5
 Invited commentary
T.Muto

Total mesorectal excision (TME) has been widely risk factors are not independent of each other,
recognized as a standard operation for rectal car- multivariate analysis and scoring systems may
cinoma. The presence or absence of nodal well be useful for this purpose. The risk of con-
involvement in the mesorectum, Duke's C or not, taining metastases in large (more than 10 mm
is relevant to the occurrence of local recurrence across) and small (less than 5 mm across) nodes
and distant metastases. It is also an essential piece with the same echoic characteristics should be
of information for performing local excision of different but are not always clearly documented
rectal tumors. Therefore, preoperative accurate in literatures.
assessment of positive lymph nodes is of vital One of the problems of US diagnosis is techni-
importance in order to design surgical procedures cal difficulty. Although only simple procedures are
more effectively. It is well known that the presence involved and it seems to be easy to handle, metic-
of positive nodes in the mesorectum indicates a ulous techniques are required to effectively visu-
higher risk for the presence of mesorectal alize the mesorectallymph nodes. The visualized
micrometastases elsewhere and for distant metas- nodes can be more precisely assessed in situ dur-
tases as well; therefore a meticulous TME proce- ing the US procedures and we have to entirely rely
dure is essential in this situation. on the examiner's ability, which is not readily
The figures shown in this chapter are beauti- assessed. In other words, accuracy of US diagnosis
ful, convincing and instructive. There seems to is rather person-dependent and obviously a learn-
be several risk factors suggesting positive nodes, ing curve also exists. One receives only one spot of
e.g. size, echogenicity, margin irregularity, loca- the whole procedure of US in a representative fig-
tion and number of enlarged nodes etc. These ure. Video-US may well be a solution to better
factors tend to be independently described and information exchange.
assessed in the individual cases, which are con- On the other hand CT and MRI have a better
vincing in the text. However, one must remem- standardized quality for visualizing positive-
ber that these examples are the most representa- nodes. Recent reports on the nodes in the
tive and carefully selected ones, and most cases mesorectum by MRI seem promising (Brown Z et
we meet in everyday practice, may not always aI., Radiology 2003;227:371) and there is no such
have characteristics of positive nodes as shown learning curve in this procedure, as required in
in the presented figures. In reality, there are def- US. The problem of CT and MRI is a high cost and
initely positive/negative cases and doubtful they may not be available in every institution. In
cases in-between, which need to be carefully putting all this information and these require-
analyzed in order to improve accuracy in the ments together, one needs to compare the three
preoperative assessment of positive nodes. As procedures in terms of accuracy, reliability and
mentioned in the text, it is vitally important to cost, that has not been well documented yet.
distinguish positive nodes from enlarged nodes The presence or absence of meso rectal node
due to inflammation. As the above-mentioned metastases is also relevant to the risk of lateral
110 Atlas of Endoanal and Endorectal Ultrasonography

Table IY.S. Clinical characteristics associated with lateral node involvement in 697 patients with RO rectal cancer

Depth of tumor invasion Tumor location (from Anal Verge)

~5cm 5 cm-8 cm 2:8 cm

Tis (N=58) 0% (0123) 0% (0/7) 0% (0128)

Tl (N=108) 3.4% (1129) 0% (0/28) 0% (0/51)
T2 (N=148) 15·3% (9/59) 2.4% (1/42) 0% (0/47)
T3 (N=364) 26.8% (26/97) 11.0% (13/118) 0.8% (11149)
T4 (N=19) 33·3% (3/9) 0% (0/3) 14·2% (1/7)

CIH 2001

node involvement, which influences the outcome lateral node dissection is not well accepted and
after curative resection of rectal carcinoma. In not popular in the Western world, it has to be
our experience there was a 17% risk of lateral remembered that there is a definite risk of lateral
node involvement in cases with positive mesorec- node involvement as shown in Table IV.s and
tal nodes. The lower rectal carcinomas with deep- nearly half of the lateral node positive patients
er invasion have a much higher risk of lateral can be saved by this procedure.
nodes involvement (Table IV.S). A low rectal T3 As described here, the accurate diagnosis of
tumor with mesorectal nodal involvement has positive nodes in the mesorectum by US is useful
the highest risk of lateral node involvement, and essential for the adequate treatment of rectal
which can be safely cured by lateral node dissec- carcinomas, and further improvements are
tion with 46% of s-year survival (Sugihara K et required to treat patients with rectal carcinoma
aI., Cancer 1996;78:1871). Although the concept of more effectively.
 SECTION V
Staging following
preoperative chemoradiotherapy
for advanced rectal cancer

ISUbmucosa l

IPerirectal fat I
 Staging following preoperative chemoradiotherapy
for advanced rectal cancer
G.A. Santoro, G. Di Falco

Preoperative high-dose radiation for treatment of ment of wall invasion by the ERUS [1,21. This is
patients with advanced rectal cancer has been probably because after irradiation the rectal wall
used to enhance resectability, facilitate sphincter- is thickened, more hypoechoic and the different
saving procedures, reduce the local recurrence layers are less clearly visualized (Fig. V.I). The
rate and potentially improve survival. More outer limit of the rectal wall is especially hard to
recently concomitant chemotherapy as a sensitiz- delineate. Post-radiotherapy inflammation and/or
ing agent has been added in attempt to improve fibrosis explain the echographic changes. Tumor
these results. One explanation of the greater effi- eradicated from an area may also leave behind
cacy and tolerance to preoperative radiotherapy is edema. A transrectal ultrasound image may
that radiation is much more effective in a surgi- reflect this as the presence of tumor. Small dis-
cally undisturbed bed. Well-oxygenated and vas- ruptions of the interfaces between bowel wall lay-
cularized tissue is vital for radiotherapy cell- ers on the ERUS may be due to areas of tumor
killing and a postoperative surgical bed is natu- invasion, artifact or edema. In these patients over-
rally hypoxic. Evaluation of patients who have staging is more of a problem than understaging
received preoperative radio chemotherapy often [3,41 (Figs. V.2-4). The tendency for overestimat-
shows significant tumor regression. Accurate ing tumor invasion, however, is due to a fear of
staging following radiation is more difficult since missing an area of tumor invasion and not treat-
radiotherapy can significantly hamper the assess- ing the tumor aggressively enough. The rectopro-

Isubmucosa l

Fig.v.l. The effect of previous radiation

is to blur the echogenic distinctions of the
I
Perirectal fat I layers of the bowel wall. After irradiation
the rectal wall is thickened, more hypoe-
choic, hypervascularized and the differ-
ent layers are less clearly visualized
 114 Atlas of Endoanal and Endorectal Ultrasonography

Fig. V.2. Sonogram before chemoradia-

tion (a) shows a large rectal tumor occu-
pying about 40% of the rectal circumfer-
ence along the anterior rectal wall. There
a is a loss of the hyperechoic layer corre-
sponding to the submucosa and marked
thickening of the hypoechoic layer corre-
sponding to the muscolaris propria con-
sistent with tumor invasion. The hypere-
choic layer representing the perirectal fat
interface is irregular. Three hypoechoic
IReSldual tumor? I lymph nodes 0.4 to 0.5 cm in diameter
are present adjacent to the tumor
(arrows) (Endosonographic staging:
uT3NO). (b) Following chemoradiation
the layers of the bowel wall are somewhat
difficult to distinguish due to the previ-
0us radiotherapy. There is only a small
area in the left anterolateral rectal wall
with a moderate probability of residual
tumor. The hyperechoic layer represent-
Iperlrectal tat I ing the perirectal fat interface is intact
(Endosonographic restaging: UT2NO). In
this case on pathological examination the
tumor has been eradicated (Pathologic
staging: pTONO)
b

static septum and anterior perirectal fat between with 86% in those patients not given preoperative
the rectum and the prostatic capsule compose an radiotherapy (P<0.05), while the endosonograph-
important interface. This area may become ede- ic accuracy for lymph node involvement was sim-
matous after irradiation of an anterior rectal ilar in the two groups (85% compared with 84%).
tumor, making it difficult to delineate an area of These results correlate with data from Williamson
normal tissue between the rectum and the et al. [6] in which ERUS accuracy in evaluating T3
prostate. On the other hand, the interpretation of and T4 lesions drops significantly after radiation
lymph node involvement seems to be unaffected from 88% to 58% and 78% of uNI lesions were
by radiotherapy. This discrepancy may be due to pNO on final evaluation. Fleshman et al. [I] found
the fibrotic involution of metastatic lymph nodes a poor correlation between post-radiochemother-
after radiotherapy. Thus the hypoechoic nodes apy ERUS and pathologic staging and reported
could be more easily visualized with ERUS (Fig. that ERUS correctly determined the depth of inva-
V.5). Napoleon et al. [5] showed that the accuracy sion in only 58% of patients after radiation and
of ERUS in assessing local invasion of tumor after overstaged the tumor in the remaining 42%. The
preoperative radiotherapy was 47% compared involvement of lymph nodes with metastatic can-
 Section V· Staging following preoperative chemoradiotherapy for advanced rectal cancer 115

IPerirectal fat I

Fig. V.3. (a) Sonogram before chemoradiation shows a large

rectal tumor (T) along the right anterolateral rectal wall. The
hyperechoic layer corresponding to the submucosa is
destroyed and the hypoechoic layer which represents the
muscolaris propria is thickened consistent with tumor inva-
sion. The hyperechoic layer that corresponds to the perirectal
fat interface is irregular in some areas and these findings rep-
resent focal tumor invasion (Endosonographic staging:
uT3NO). (b) Two months later, after chemoradiotherapy, the
layers of the rectal wall are difficult to distinguish. Residual
tumor is present in the anterior rectal wall. The hyperechoic
layer representing the perirectal fat interface is irregular
(Endosonographic restaging: UT3NO). (c) No residual tumor
was found on the computed tomography scan which shows
only a marked thickening of the layers of the rectal wall fol-
lowing chemoradiation. In this case no invasion of the
perirectal fat was found on microscopic examination
c (Pathologic staging: pT2NO)
 116 Atlas of Endoanal and Endorectal Ultrasonography

Fig. V.4. Prior to preoperative chemoradi-

ation (a) the sonographic layers of the
anterior rectal wall are completely dis-
rupted by the tumor (T). These findings
correspond to those of an advanced
deeply invasive rectal cancer extending
through the bowel wall (arrows). Multiple
large (0.5-1 em) hypoechoic lymph nodes
(In) are present consistent with tumor
invasion (Endosonographic staging:
uT 4N1). (b) Following chemoradiation
the layers of the bowel wall are somewhat
difficult to distinguish due to previous
radiotherapy. There is no regression of
the tumor (T) (arrows). The previously
seen lymph nodes are no longer present
b (Endosonographic restaging: uT4NO)

cer was correctly determined in 68% patients. Rau neoadjuvant radio chemotherapy. Wall invasion
et al. [2] evaluated ERUS accuracy in staging was correctly ascertained in 50% of patients, with
locally advanced primary rectal cancer (uT3/uT4) under estimation of 13% and overestimation of
at 4 to 6 weeks after completion of preoperative 37%. The accuracy to predict the depth of tumor
 Section V' Staging following preoperative chemoradiotherapy for advanced rectal cancer 117

72% of patients with lymph node involvement.

Positive predictive values of ERUS after irradia-
tion were 72% and 56% for wall penetration and
lymph node status, respectively. Negative predic-
tive values of ERUS after irradiation were 100%
and 82%, respectively. Gavioli et al. [8] assessed
the advantages of ERUS after preoperative radio-
therapy in rectal cancer, its reliability in tumoral
staging and its capacity to identify completely
sterilized lesions. Morphologically and quantita-
tively, postradiation ERUS showed the reappear-
ance of anatomic cleavage planes, a considerable
shrinkage of the tumor and in low rectal tumors
an increase in the distance from the anorectal ring
in more than 50% of cases. Histological examina-
tion showed that fibrosis was the most dominant
component of the irradiated lesions, varying by
more than 50 to 100% of the lesion. A comparison
Fig. V.S. Metastatic lymph nodes (In) visualized at ERUS after
of postradiation ERUS with histopathology
radiotherapy. A 2-cm area of residual tumor (T) is present in
the left posterolateral rectal wall revealed that fibrosis became the morphologic
basis of ultrasound images; therefore, after radio-
therapy, what ERUS staged was no longer the
tumor but the extent of fibrosis in the rectal wall.
A histopathological examination showed that the
infiltration was found to correlate with downs tag- residual tumor, when present, was always within
ing. Neither tumor distance from the anal verge the fibrosis, never outside or separate from it.
nor tumor location correlated with the staging Postradiation ERUS showed echo-pattern
accuracy. Lymph node involvement was correctly changes. Some of the changes (more echo genic
assessed in 57% of patients. Bernini et al. [7] and non-homogeneous lesions) were histological-
reported an overall downstaging of 70% after pre- ly related to the persistence of the tumor to a con-
operative radio chemotherapy. Downstaging was siderable degree; other changes (reappearance of
seen in 53% of patients with wall invasion and in parietal layers) were related to complete steriliza-
tion of the lesions. This report presents a com-
pletely new concept: that six to eight weeks after
radiotherapy, ERUS no longer stages the tumor,
but rather the fibrosis that takes its place. The
extent of fibrosis in the rectal wall is a direct indi-
cation of the depth of residual cancer. Kahn et al.
[3] assessed the ability of digital rectal examina-
tion, computed tomography, endorectal ultra-
sound and magnetic resonance imaging to predict
the absence of disease after radio chemotherapy
(Fig. V.6). Most irradiated lesions were overstaged
by radiological assessment and physical examina-
tion. No technique could reliably distinguish
between postradiation fibrosis and residual can-
cer. The negative predictive value for digital rectal
examination was 24%. CT accurately staged 23%
of lesions and ERUS predicted 17% of lesions cor-
rectly.

Fig. V.6. Computed tomography scan showing a marked

thickening of the layers of the rectal wall following chemora-
diation
118 Atlas of Endoanal and Endorectal Ultrasonography

References radiotherapie pre-operatoire? Gastroenterol Clin BioI

1993;17:287-91
1. Fleshman IW, Myerson RI, Fry RD, Kodner IJ. Accuracy 5. Napoleon B, Pujol B, Berger F, Valette PI, Gerard IP,
of transrectal ultrasound in predicting pathologic Souquet Ie. Accuracy of endosonography in the stag-
stage of rectal cancer before and after preoperative ing of rectal cancer treated by radiotherapy. Br I Surg
radiation therapy. Dis Colon Rectum 1992;35:823-9 1991;78:785-8
2. Rau B, Hunerbein M, Barth C, Wust P, Haensch W, 6. Williamson PR, Hellinger MD, Larach SW, Ferrara A.
Riess H, Felix R, Schlag PM. Accuracy of endorectal Endorectal ultrasound of T3 and T4 rectal cancers
ultrasonography after preoperative radiochemothera- after preoperative chemoradiation. Dis Colon Rectum
py in locally advanced rectal cancer. Surg Endosc 1999; 1996;39:45-9
13:980-4 7. Bernini A, Deen, KI, Madoff RD, Wong WD.
3. Kahn H, Alexander A, Rakinic I, Nagle D, Fry R. Preoperative adjuvant radiation with chemotherapy
Preoperative staging of irradiated rectal cancers using for rectal cancer: its impact on stage of disease and the
digital rectal examination, computed tomography, role of endorectal ultrasound. Ann Surg Oncol 1996;
endorectal ultrasound and magnetic resonance imag- ):131-5
ing does not accurately predict ToNo pathology. Dis 8. Gavioli M, Bagni A, Piccagli I, Fundaro S, Natalini G.
Colon Rectum 1997;20:140-4 Usefulness of endorectal ultrasound after preoperative
4. Burtin P, Cellier P, Croue A, Arnaud IP, Carpentier S, radiotherapy in rectal cancer: comparison between
Boyer I. Place de l'echographie endorectale dans Ie sonographic and histopathologic changes. Dis Colon
bilan d'extension du cancer du rectum: avant ou apres Rectum 2000;43:1075-83
 Invited commentary
M. Hiinerbein

Surgery is the primary modality for cure in radio chemotherapy results in down-staging and
patients with rectal cancer. However, despite tumor shrinkage of rectal cancer [Boulis-Wassif S
potential curative surgery, there is a high risk of et al., 1984; Rau B et al., 1998]. However, broad
recurrence. The lower risk of treatment related acceptance of preoperative therapy is limited in
toxicity such as small bowel damage and higher the absence of reliable techniques for accurate
efficacy are arguments for the use of preoperative preoperative staging of the residual tumor. Our
radiotherapy instead of postoperative radiothera- experience and the data from other investigators
py. Recently, the beneficial effect of short-term demonstrate that endorectal ultrasound (ERUS)
preoperative radiotherapy was proven in a ran- does not allow reliable restaging of rectal cancer
domized trial comparing preoperative radiother- following neoadjuvant therapy. The accuracies for
apy (5 x 5 Gy) followed by total mesorectal exci- the assessment of tumor infiltration depth and
sion (TME) with TME alone in patients with lymph node involvement are approximately from
resectable rectal cancer. Among patients who 50%-60% and 60%-70%, respectively [Bozetti F et
underwent a macroscopically complete local al., 1996; Glaser F et al., 1993; Napoleon B et al.,
resection, the rate of local recurrence at two years 1991]. Several factors have been identified that
was significantly reduced from 8.2 percent in the contribute to the unsatisfactory accuracy of
surgery group to 2.4% in the surgery plus radio- endorectal ultrasonography with respect to TNM
therapy group [Kapiteijn E et al., 2001]. classification. First, ERUS is limited by the inabil-
Histological alterations i.e. less inflammatory ity to make a tissue specific diagnosis and to dis-
reaction and more pronounced fibroblastic reac- tinguish fibrotic transformation from the residual
tion occur even after 5 days of irradiation of rec- tumor. Overestimation of tumor penetration is
tal carcinomas [Nagtegaal ID et al., 2002]. the main problem. After eradication the tumor is
However, it is unlikely that these changes have sig- replaced by fibrotic tissue that shows a low
nificant influence on preoperative staging and the echogenicity similar to rectal cancer (Fig. V.7).
choice of the surgical procedure, because the Radiation induced edema of the normal rectum
tumor size is usually not reduced by short-term makes it difficult to delineate the individual layers
radiation. of the rectal wall. Although the size of the tumor
A completely different situation must be con- may be dramatically reduced by pre-treatment,
sidered in patients with advanced rectal cancer there is no complete restoration of the wall layers.
who undergo intensive preoperative radio- Consequently, no imaging method, including
chemotherapy. Despite a lack of randomized data ERUS, is capable of documenting complete remis-
demonstrating clinical benefit, preoperative sion of rectal cancer according to the TNM-classi-
chemoradiation has been increasingly used to fication. On the other hand ERUS cannot rule out
improve the prognosis of the patients and to microscopic foci of residual disease in the tissue
increase the rate of sphincter preserving opera- (Fig. V.8). Nonetheless, ERUS remains the most
tions. It has been shown that preoperative important technique for the planning of surgery
 120 Atlas of Endoanal and Endorectal Ultrasonography

a b
Fig. Y.7. a) uT3 rectal cancer [9] b) complete destruction of the wall layers (yuT3) after radiochemotherapy, but complete remis-
sion (ypTO)

a b
Fig. Y.S. a) uT3 rectal cancer [9] b) partial restoration of the wall layers without visible tumor (uTO), but microscopic residual
disease (ypT2)

because it provides essential information on the based on functional imaging techniques such as
size of the tumor and its relation to adjacent power Doppler sonography, a variety of harmonic
organs. The reduction of tumor size may be the imaging techniques and dynamic contrast
best predictor of the treatment results that can be enhanced ERUS. These techniques allow assessing
obtained with conventional ERUS. Other criteria changes in flow and vascularity that occur in
and/or examination techniques are required to response to neoadjuvant therapy and may better
identify patients with complete response after reflect tumor eradication than ultrasound mor-
radio chemotherapy, in whom aggressive surgery phology alone. Elastography is a new ultrasound-
may be avoided. Promising new methods are technique that provides information on mechani-
 Section V' Staging following preoperative chemoradiotherapy for advanced rectal cancer 121

Fig. V.9. Fronfire EUS of a stenotic tumor:

Multiplanar display of 3D-ERUS showing
an uT3 tumor

cal tissue properties. During the examination a tional ERUS [Hiinerbein M et al., 2000]. However,
sequence of echographic images is acquired while the field of view of these probes is limited to 3-4
the tissue is slightly compressed by the ultrasound cm and may not allow the depiction of extensive
probe. Using numerical analysis of image pairs for tumors entirely. Front fire probes with an anterior
the acquired sequence, the tissue strain is calcu- scan plane may also be used for the imaging of
lated to represent the spatial elasticity distribu- stenotic tumor, but the interpretation of the
tion of a specific cross-section. Preliminary images is difficult (Fig. V.9). In the meantime 3D-
results suggest that ultrasound elastography has frontfire probes have been developed that acquire
the potential to detect malignant tissue areas, three-dimensional endorectal ultrasound data.
which are not shown in the B-mode image Reconstructed transverse scan planes can be
[Lorenz A et al., 2000]. obtained by multiplanar reformatting.
Another limitation of ERUS is that the exami- Clearly there are some limitations of conven-
nation may not be performed in patients with tional ERUS in the reevaluation of rectal cancer
stenotic tumors because the instrument cannot be after neoadjuvant therapy. However, the results of
passed across the tumor. This problem can be ERUS are at least comparable to those of other
solved by using small caliber echo graphic probes imaging modalities i.e. CT, MRI [Kwok H et al.,
that can be introduced through the working chan- 1999]. It seems likely that the performance of
nel of the colonoscope. It has been reported that endorectal ultrasound in the restaging of rectal
the accuracy of miniprobes in the staging of col- cancer after neoadjuvant therapy can be further
orectal cancer is comparable to that of conven- improved by technical advances.

References with the determinant of pathological parameters in

rectal cancer. J PathoI2002;197:20-7
1.Kapiteijn E, Marijenen CA, Nagteegaal ID, Putter H, 3. Rau B, Wust P, Hohenberger P, Loffel J, Hiinerbein M,
Steup WH, Wiggers T, Rutten HJ, Pahlman L, Glimelius Below C, Gellermann J, Speidel A, Vogi T, Riess H, Felix
B, van Krieken JH, Leer, JW, and van de Velde, CT. R, and Schlag PM. Preoperative hyperthermia com-
Preoperative radiotherapy combined with total bined with radiochemotherapy in locally advanced
mesorectal excision for resectable rectal cancer. N Engl rectal cancer: a phase II clinical trial. Ann Surg
J Med 2001;345:690-2 1998;22n80-9
2. Nagtegaal ID, Marijnen CA, Kranenberg EK, Mulder- 4. Boulis-Wassif S, Gerard A, and Loygue J. Final results
Stapel A, Hermans J, van der Velde CJ, and van Krieken of a randomized trial on the treatment of rectal cancer
JH. Short-term preoperative radiotherapy interferes with preoperative radiotherapy alone or in combina-
122 Atlas of Endoanal and Endorectal Ultrasonography

tion with 5-fluouracil, followed by radical surgery. 7. Glaser F, Kuntz C, Schlag P, and Herfarth C. Endorectal
Cancer 1984;53:1811-8 ultrasound for control of preoperative radiotherapy of
5. Napoleon B, Pujol B, Berger F, Valette pJ, Gerard JP, and rectal cancer. Ann Surg 1993;217:64-71
Souquet Je. Accuracy of endosonography in the stag- 8. Lorenz A, Ermert H, Sommerfeld HJ, Garcia-
ing of rectal cancer treated by radiotherapy. Br J Surg Schurmann M, Senge T, and Philippou S. Ultrasound
1991;78:785-8 elastography of the prostateA new technique for
6. Bozetti F, Andreola S, Rossetti C, Zucali R, Meroni E, tumor detection. Ultraschall Med 2000;21:8-15
Baratti D, Bertario L, Doci R, and Gennari 1. 9. Hiinerbein M, Totkas S, Ghadimi BM, and Schlag PM.
Preoperative radiotherapy for resectable cancer of Preoperative evaluation of colorectal neoplasms by
the middle-distal rectumits effect on the primary colonoscopic miniprobe ultrasonography. Ann Surg
lesion as determined by endorectal ultrasound using 2000;232:46-50
flexible echo colonoscope. Int J Colorectal Dis 1996; 10. Kwok H, Bisset IP, and Hill GL. Preoperative staging of
11:283-6 rectal cancer. Int J Colorectal Dis 1999;15:9-20
 SECTION VI
Postoperative follow-up
of rectal cancer
 VI.I.
Postsurgical evaluation
G.A. Santoro, G. Di Falco

The intensity of postoperative surveillance for lems caused by artifacts from metallic clips may
rectal cancer varies greatly between institutions, occur. Magnetic resonance imaging (MRI) has
and recommendations in literature are inconsis- been accepted as superior to CT in tissue charac-
tent [1]. It is still unclear which method of detec- terization with an accuracy of 75% to 93% and a
tion will reveal local recurrence before clinical sensitivity of 91% to 100% for detecting local
symptoms or markers of advanced growth appear. recurrence. For the early detection of local recur-
Lohnert et al. [2] very well summarized the effec- rence, frequent assessments are required, but the
tiveness of common follow-up schedules. repeated use of both CT and MRI in a follow-up
Anamnesis, tumor markers, digital examination program is undesirable because of high costs and
and endoscopy have to detect local recurrence in the case of CT, radiation exposure. Endorectal
and ultrasound of the liver and chest x-ray have to ultrasonography (ERUS) is being used frequently
detect distant metastases. Anamnesis can reveal as a diagnostic modality in order to obtain follow-
advanced recurrences that produce pain by sec- up information in patients who had previous
ondary invasion of the sacral bone or other adja- endoscopic excision of rectal lesions, as well as in
cent organs, stenosis by extended intraluminal patients who had surgical local excision or wide
growth, compression by extramural masses or resection of a rectal tumor with anastomosis [2-
bleeding. Because therapy will be palliative in 5]. ERUS combines much of the diagnostic poten-
such cases, anamnesis is an insufficient tool for tial of endoscopy (endoluminal tumor detection)
early detection of local recurrence. Digital rectal and CT or MRI (extramural tumor detection). It
examination is limited to the distal 8 cm of the can be performed on an outpatient basis, is well
anorectum and exclude one-third to two-thirds of tolerated, is simple to execute, does not involve x-
all rectal anastomotic regions from follow-up. ray exposure and is inexpensive. In summary
Furthermore, small pararectal lesions go unno- ERUS is an ideal technique in the monitoring of
ticed. Endoscopy plays a major part in rectal can- postoperative patients. In practice, every patient
cer follow-up, but the frequent extramural growth can be examined if the probe can reach the level of
of local recurrence will not permit an early endo- anastomosis. Previous surgical manipulation pro-
scopic diagnosis. Computed tomography (CT) has duces scarring at the site of excision and at the
proven to be effective in the detection of local anastomosis, which is visible on ERUS and must
recurrences, with sensitivities of 40% to 100%. often be distinguished from recurrent or residual
Although long considered the best current tech- tumor. According to Lohnert et al. [2] the
nique, CT has important limitations. Lesions must endosonographic appearance of the normal sta-
be at least 2 cm in diameter for accurate diagnosis pled anastomosis is characterized by small, local-
and it may be difficult to distinguish local recur- ized, bright spots with a dorsal shadow (suturing
rence from postoperative alterations of the nor- clips) (Fig. VLI). The typical five-layer structure of
mal anatomy by fibrosis or inflammation, or prob- the rectal wall is interrupted by the anastomosis,
126 Atlas of Endoanal and Endorectal Ultrasonography

presenting as a ring with a hyperechoic or

echomixed thicker inner layer, followed by a
smaller echo-poor layer and surrounded by a
small, white, interface layer as a border signal to
the perirectal tissue.
The same authors reported the interpretation
of findings in digital examination, rectoscopy,
endorectal ultrasonography and tumor markers
in the follow-up of rectal cancer (Table VLI). The
overall accuracy rates for detection of locally
recurrent rectal cancer was 57.8% for tumor
markers, 23.3% for digital examination, 34.5% for
rectoscopy and 79.3% for ERUS. Accuracy rose to
100% when transrectal or transperineal needle
biopsies were used to differentiate between malig-
nant tumors and benign lesions.
Benign-appearing rectal polyps may be iden-
tified and removed during endoscopy. If adeno-
carcinoma is identified in these polyps, then stag-
Fig. VI.1. Rectal endosonography of a stapled anastomosis. ing these tumors is important for clinical deci-
Suturing clips are visible as bright spots with a dorsal shadow. sion-making. Endorectal ultrasonography is an
The typical five-layer structure of the rectal wall is interrupt- accurate technique for localizing tumors in or
ed by the anastomosis, presenting as a ring with hyperechoic
or echomixed thicker inner layer, followed by a smaller echo- beyond the rectal wall in patients who have
poor layer and surrounded by a small, white, interface layer as undergone diagnostic polypectomy (Fig. VI.2).
border signal to the perirectal tissue

Table VI.1. Interpretation of findings in digital examination, rectoscopy, endosonography and tumor markers in
follow-up of rectal cancer (From Lohnert MSS, Doniec JM, Henne-Bruns D. Effectiveness of endoluminal sonog-
raphy in the identification of occult rectal cancer recurrences. Dis Colon Rectum 2000;43:483-91)

Interpretation Digital Rectoscopy Endorectal CEA CA 19-9

examination endosonography

Normal No tumor palpable No visible tumor No appearance <2.5 ng/ml <40 ng/ml
or impression of circum script
of the rectum areas in the anastomotic
region or perirectal tissue
Suspicious Smooth pararectal tumor Macroscopically Echomixed, hyperechoic >2.5 ng/ml >40 ng/ml
inflammation or or both circumscript areas <6.0 ng/ml <60 ng/ml
granulation, impression at the anastomotic region,
of the rectum or both perirectal tissue or both
Suspected Hard pararectal tumor Macroscopically malignant Hypoechoic lesions at >6.0 ng/ml >60 ng/ml
recurrence palpable or irregular tumor or ulcer visible anastomotic region
anastomotic surface or perirectal tissue with
infiltration or destruction
of rectal wall or adjacent
organs
 Section VI' Postoperative follow-up of rectal cancer 127

Site of excision

Isubmucosal __ ...."',...

Fig. VI.2. (a) Endoscopic excision of a 2

cm polypoid lesion located on the right
side of the rectal wall. (b) Endorectal
sonography performed 2 months after
excision shows focal widening of the
hypoechoic layer corresponding to the
muscolaris mucosa and disruption of the
hyperechoic layer which corresponds to
the submucosa. These changes are con-
b sistent with postfulguration scar tissue

However, after polypectomy, tissue changes may Kruskal et al. [6] compared the results of ERUS
occur that limit accurate demonstration of the staging of rectal cancer after polypectomy with
depth of tumor invasion. For these reasons ERUS the histopathologic findings after surgery. ERUS
should be performed 3-4 weeks after polypecto- had a sensitivity of 100%, specificity of 44%, pos-
my. Focal hypoechoic abnormalities identified at itive predictive value of 64% and negative predic-
the biopsy site, cannot often be distinguished tive value of 100% for detection of residual tumor
from residual tumor. Histopathologic analysis of after polypectomy. Although the precise T stage
the abnormal areas demonstrates fibrous tissue, was correctly predicted with ERUS in only 44% of
desmoplasia, and hemorrhage, none of which can patients, ERUS was able to demonstrate whether
be reliably distinguished from residual tumor the tumor was limited to the bowel in 89% of
with ERUS. This may result in overstaging of cases. ERUS should therefore remain the local
some To lesions (tumor completely removed with staging technique of choice in this specific
polypectomy) as subtle Tl lesions (Fig. VI.3). patient population.
 128 Atlas of Endoanal and Endorectal Ultrasonography

a b

Sit. of 8x(:lal,on I

Fig. VI.3. Rectal wall scarring misinter-

preted as Tl tumor. A polypoid rectal
d
lesion was excised with a colonoscope 3
months ago (a-c). Sigmoidoscopy shows
IProstate I no residual lesion and the site of excision
is healed (d). Endorectal sonography
shows an area of thickening adjacent to
Islte otexclslon l the prostate, approximately 1.5 cm in
diameter (e). Lesion is hypoechoic and
there is disruption of the hyperechoic
layer which corresponds to the submu-
cosa immediately beneath the hypoe-
choic layer representing the muscolaris
Isubmucosa l propria. The hyperechoic layer corre-
sponding to the perirectal fat is intact. No
residual tumor was identified at
e histopathologic analysis
 Section VI- Postoperative follow-up of rectal cancer 129

References staging and follow-up of rectal cancer. Endoscopy

1995;27:469-79
1. Brethauer SA, Magrino TJ, Riffenburgh RH, Johnstone 4. Beynon J, Mortensen NJ, Foy DM, Channer JL, Rigby
PAS. Management of recurrent colorectal carcinoma. H, Virjee J. The detection and evaluation of locally
Colorectal Dis 2002;4:246-53 recurrent rectal cancer with rectal ultrasonography.
2. Lohnert MSS, Doniec JM, Henne-Bruns D. Dis Colon Rectum 1989;32:509-17
Effectiveness of endoluminal sonography in the iden- 5. Charnley RM, Heywood MF, Hardcastle JD. Rectal
tification of occult rectal cancer recurrences. Dis endosonography for the visualization of the anasto-
Colon Rectum 2000;43:483-91 mosis and its relevance to local recurrence. Int J
3. Meyenberger C, Huch Boni RA, Bertschinger P, Zala Colorect Dis 1990;5:127-9
GF, Klotz HP, Krestin GP. Endoscopic ultrasound 6. Kruskal JB, Sentovich SM, Kane RA. Staging of rectal
and endorectal magnetic resonance imaging: a cancer after polypectomy: usefulness of endorectal
prospective, comparative study for preoperative US. Radiology 1999;211:31-5
 VI.2.
Local recurrences
G.A. Santoro, G. Di Falco

Local recurrences in rectal carcinoma appear Whereas recurrence within the bowel lumen is
mostly within the first two years after surgery. easily accessible to endoscopy, it is often extreme-
Different groups have reported huge variations in ly difficult to identify extramural recurrence.
the recurrence rates of rectal cancer, ranging from Radiological methods including CT, MRI,
40/0 to 430/0. The causes oflocal recurrence are pos- immunoscintigraphy (Fig. VI.4) and positron
itive histological margins, implantation of viable emission tomography (PET) have been used with
cancer cells at the time of surgery and incomplete limited success for the detection of pelvic recur-
excision of the mesorectum. At diagnosis most rence of rectal cancer. A sensitivity ranging from
recurrences are extraluminal within the pericolic 41% to 88% has been reported for CT. Computed
fat, mesentery and lymph nodes and by the time tomography scanning usually·. fails to detect
they become symptomatic or intraluminal they lesions smaller than 1.5 cm in diameter. In addi-
are fixed and therefore unresectable. Pollard et al. tion, postoperative fibrosis cannot\be differentiat-
[1] achieved a five-year survival rate of 290/0 of ed from tumor recurrence by this technique.
patients if radical resection of the tumor was pos- Magnetic resonance imaging seems to be able to
sible. Such repeated curative re-operation distinguish late fibrosis from tumor recurrence by
depends mainly on the early detection of recur- low signal intensity on T2-weighted images. Early
rence and several follow-up programs have been diagnosis of recurrence in asymptomatic patients
developed to achieve this aim. requires frequent examinations for early detection

Fig. VI.4. Recurrent rectal cancer detect-

ed by immunoscintigraphy (CEA-scan)
 Section VI' Postoperative follow-up of rectal cancer 131

Fig. VI.6. Rectal endosonogram showing at 10 o'clock an

extraluminal recurrence involving part of the rectal wall
(arrow). At 5 o'clock an hypoechoic areas appearing like
enlarged lymph node (arrow)

anastomotic region that may infiltrate the perirec-

tal fat of the rectal wall next to the anastomosis
(Figs. VI.5-8). At times it may be difficult to differ-
b
entiate between a recurrent tumor and fibrosis.
Fig. VI.S. Large recurrence occupying nearly 80% of the rec-
tal circumference at the site of the previous colorectal carci- The typical problem of misinterpretation of post-
noma. (a) The tumor penetrates deeply beyond the bowel operative changes and scars can be solved by a
wall. The various layers of the rectal wall are not well seen due baseline ERUS three months after surgery. Usually
to scarring from previous surgery and subsequent radiother- if no definitive diagnosis can be made by ERUS, a
apy. (b) Computed tomography scan showing recurrent rectal
repeat examination is performed after four to six
cancer at the site of the colorectal anastomosis (arrows).
weeks. If the lesion has increased in size, it is most
likely to be a tumor; if it has not increased in size,
further examinations are mandatory at short
intervals [6]. This procedure results in consider-
of any change of the perirectal tissue, but the cost- able discomfort for the patient and can also delay
benefit problem should be kept in mind. ERUS treatment of disease. Consequently ERUS-guided
with 7.5 or 10 MHz transducers has been shown to transrectal or transperineal needle biopsies could
be an effective tool in detecting extrarectal recur- be extremely valuable for therapeutic decisions in
rence [2-5]. The sonographic anatomy of the neo- doubtful cases [7-8].
rectum is no different from the original rectum One must be wary about postoperative pelvic
apart from the fact that the anastomotic area may anatomy on ERUS because the uterus or seg-
look more thickened than the rectal wall. The ments of the small bowel may prolapse alongside
presence of metallic clips does not affect the inter- the neorectum and can be mistaken for a recur-
pretation of the images as the clips are seen as rent tumor. Many studies have reported the effec-
small bright echoes (Fig. Vb). The sonographic tiveness of ERUS in detecting local recurrence
diagnosis of an anastomotic recurrence is based after resection of rectal cancer. Mascagni et al. [9]
on hypoechoic, irregularly shaped areas in the detected 17 (14%) recurrences in 120 patients reg-
 132 Atlas of Endoanal and Endorectal Ultrasonography

c
Fig. VI.7. Hypoechoic lesion in the left posterolateral rectal
wall at the level of the seminal vesicles. The sonographic lay-
ers are difficult to distinguish clearly due to scarring from
previous surgery (a, b). There is a thickening of the hypoe-
choic layer corresponding to the muscolaris propria. The
perirectal fat interface is intact. (c) Computed tomography
scan showing recurrent rectal cancer at the site of the col-
orectal anastomosis (arrow). (d) Sigmoidoscopy showing an
ulcerated lesion located posteriorly at the site of the previous
d anastomosis
 Section VI' Postoperative follow-up of rectal cancer 133

ital rectal examination in 5, CT in 9, colonoscopy

in 8 and the CEA was increased in 4. In the study
by Meyenberger et al. [I4] four of six recurrences
(6iYo) were detected by ERUS alone, without
endoscopic signs of recurrence. The accuracy and
positive predictive value of ERUS was 93% and
86%, respectively. In this series, ERUS was com-
parable to endorectal coil MRI in detecting and
excluding recurrent rectal cancer. The advantages
of ERUS are its low cost, availability and the small
diameter of the instrument, allowing an exami-
nation without any discomfort for the patient. In
contrast endorectal coil MRI is not operator-
dependent, compared to ERUS. The methods
should therefore be recommended in all patients
with curatively resected rectal cancer in a routine
follow-up program, or at least in all patients with
elevated CEA or symptoms suggesting tumor
Fig. VI.S. Recurrent posterior extraluminal rectal carcinoma recurrence without any evidence of distant
at site of resection appearing like 4X4 cm mixed echogenicity metastases. Further comparative investigations
rounded mass with regular controls in a larger number of
patients should conclusively define the value of
the two methods in detecting local cancer recur-
rence and improving the overall outcome for
ularly investigated with ERUS at three-month these patients.
intervals. The overall accuracy in predicting 3D image analysis facilitates assessment of sus-
tumor recurrence was 97%, with a sensitivity of picious pararectal lesions by the contemporary
94%, a specificity of 98%, a positive predictive display of three perpendicular scan planes (sec-
value of 85% and a negative predictive value of tion display) or volume reconstructions of the
99%. Tarroni et al. [10] detected 58 recurrences scanned area (volume display) (Fig. VI.9). The for-
with ERUS by following 335 patients. Of the 58, 18 . mer facilitates delineation of small pararectal
patients were asymptomatic. Hildebrandt et al. lesions such as lymph nodes, because these can be
[4] detected 22 recurrences in his series, but only easily distinguished from vessels or artifacts. The
six (27%) were diagnosed with ERUS alone, with- latter is very useful in enhancing the understand-
out digital or endoscopic signs of recurrence. ing of the anatomy and defining the spatial rela-
Charnley et al. [11] identified 15 recurrences by tionship of the tumor with normal structures,
performing follow-up ERUS in 92 patients who bladder, seminal vesicles, uterus. However, despite
had undergone surgery for rectal cancer. Two of improved imaging facilities, differentiation of
these patients had extraluminal recurrences that benign and malignant lesions remains difficult.
were not demonstrated by digital or sigmoido- ERUS-guided biopsy has a considerable impact on
scopic examination. Beynon et al. [12] imaged 22 the diagnostic accuracy of ERUS. Only a few
recurrences in 85 patients, of which only 3 (14%) reports, however, have described guided biopsies
were detected solely by ERUS. Milsom et al. [13] of recurrent rectal cancers or recurrent pelvic
showed that ERUS was much more effective in masses. Transrectal biopsies are usually confined
detecting local recurrence than CT. Novell et al. to the lower rectum (Fig. VI. 10 ). Transperineal
[6] compared the results obtained by ERUS with biopsies will provide specimens taken from the
other means of assessment in the follow-up of middle and upper part of the rectum (Fig. VI.11).
rectal cancer. In 21 patients local recurrence was A major disadvantage of a transperineal approach
diagnosed. All 21 showed evidence of local recur- is that the examiner cannot observe the pathway
rence by ERUS examination, 14 by digital rectal of the needle, because only a transverse section
examination, 16 by colonoscopy, 18 by CT and the through the lesion is available. Hunerbein et al.
CEA level was high in 13 cases. In 12 patients who [15] reported a strong agreement between trans-
were asymptomatic ERUS was positive in 12, dig- rectal biopsy results and the final diagnosis, the
134 Atlas of Endoanal and Endorectal Ultrasonography

Fig. VI.9. Three-dimensional sections

enable assessment of a recurrent rectal
tumor in coronal, sagittal and transverse
scan planes (arrows)

Fig. VI.lO. Diagrammatic representation

of a transrectal biopsy performed
through a built-in needle guide using B-K
Medical 8551 endoprobe

sensitivity and specificity being 91% and 93%, rectal cancer. 3D image display allowed precise
respectively. In contrast, clinical examination, CT control of the position of the biopsy needle with-
or ERUS showed only a moderate level of agree- in the target. Representative material was collect-
ment with the histopathologic diagnosis, mainly ed in 93% of lesions and overall endosonographic
because of the limited specificity of all 3 methods diagnosis was changed in 28% of patients, thus
(65% vs. 46% vs. 57%). ERUS-guided biopsy was avoiding overtreatment in patients with benign
significantly more accurate than CT and ERUS lesions, whereas appropriate treatment could be
(P<O.OI). The same authors investigated the role initiated promptly after histological confirmation
of 3D ERUS for evaluation and biopsy of recurrent of malignancy.
 Section VI- Postoperative follow-up of rectal cancer 135

a L -________________________________________ ~L_ __ ~

Fig. VI.n. Diagrammatic representation

of a transperineal needle biopsy per-
formed using the B-K Medical 1850 endo-
b L-____ ~~ ____________ ~ ______________________________ probe (a, b)

References 4. Hildebrandt U, Feifel G, Schwarz HP, Scherr O.

Endorectal ultrasound: instrumentation and clinical
1. Pollard SG, Macfarlane R, Everett WG. Surgery for aspects. Int J Colorectal Dis 1986;1:203-7
recurrent colorectal carcinoma - is it worthwhile? Ann 5. Lohnert MSS, Doniec JM, Henne-Bruns D.
R Coli Surg EngI1989;71:293-8 Effectiveness of endoluminal sonography in the identi-
2. Heriot AG, Grundy A, Kumar D. Preoperative staging fication of occult rectal cancer recurrences. Dis Colon
of rectal carcinoma. Br J Surg 1999;86:17-28 Rectum 2000;43:483-91
3. Kumar A, Scholefield JH. Endosonography of the anal 6. Novell F, Pascaul S, Viella P, Trias M. Endorectal ultra-
canal and rectum. World J Surg 2000;24:208-15 sonography in the follow-up of rectal cancer. Is it a
136 Atlas of Endoanal and Endorectal Ultrasonography

better way to detect early local recurrence? Int J endosonography. Br J Surg 1988;75=1232
Colorectal Dis 1997;12:78-81 12. Beynon J, Mortensen NJ, Foy DM, Channer JL, Rigby
7. Nielsen MB, Pederson JF, Hald J, Christiansen J. H, Virjee J. The detection and evaluation of locally
Recurrent extraluminal rectal carcinoma: trans rectal recurrent rectal cancer with rectal ultrasonography.
biopsy under sonographic guidance. AJR 1992; Dis Colon Rectum 1989;32:509-17
158:1025-7 13. Milsom JW, Lavery IC, Stolfi VM, Czyrko C, Church
8. Hunerbein M, Totkas S, Moesta KT, Ulmer C, Handke JM, Oakley JR, Fazio vw. The expanding utility of
T, Schlag PM. The role of transrectal ultrasound-guid- endoluminal ultrasonography in the management of
ed biopsy in the postoperative follow-up of patients rectal cancer. Surgery 1992;112:832-40
with rectal cancer. Surgery 2001; 129:164-9 14. Meyenberger C, Huch Boni RA, Bertschinger P, Zala
9. Mascagni D, Corbellini L, Urciuoli P, Di Matteo G. GF, Klotz HP, Krestin GP. Endoscopic ultrasound and
Endoluminal ultrasound for early detection of local endorectal magnetic resonance imaging: a prospec-
recurrence of rectal cancer. Br J Surg 1989;776:1176-80 tive, comparative study for preoperative staging and
10. Tarroni D, Mascagni D, Urciuoli P, Di Pietrantonio M, follow-up of rectal cancer. Endoscopy 1995;27:469-79
Di Matteo G. Endoluminal ultrasonographic accuracy 15. Hunerbein M, Dohmoto M, Haensch W, Schlag PM.
in rectal cancer. Br J Surg 1992;79:S30 Evaluation and biopsy of recurrent rectal cancer using
11. Charnley RM, Pyf G, Amar SS, Hardcastle JD. The early three-dimensional endosonography. Dis Colon
detection of recurrent rectal carcinoma by rectal Rectum 1996;39:1373-8.
 Invited commentary
U. Hildebrandt

The aim of intensive follow-up after initial cura- ERUS the high percentage of recurrences is in
tive treatment for rectal cancer is to increase long itself somewhat astonishing.
term survival. The purpose of postoperative ERUS Local pelvic recurrences occur not by addi-
is the detection of isolated perirectal recurrent tional tumor spread, but rather by inadequate ini-
disease with the chance of a salvage operation. tial treatment of the primary site with surgery.
Follow-up guidelines are inconclusive and Bulow et al. [5] compared conventional surgery
clinical practice varies widely. Only rarely is ERUS with mesorectal excision for rectal cancer. The
included in postoperative follow-up protocols. cumulative 3-year local recurrence rate was 11 per
Protocols not including ERUS come to different cent after mesorectal excision compared with 30
results. In one study [1] adherence to a strict fol- per cent after conventional surgery. When preop-
low-up program was ineffective for improving erative radiotherapy was combined with total
long-term survival for patients who underwent mesorectal excision the rate of local recurrence at
re-operation with curative intent. In another two years was 2.4% in the radiotherapy-plus-
study the outcome of salvage therapy among surgery group and 8.2% in the surgery-only group
patients with single-site pelvic recurrence was [6]. These results of recent studies, which were
investigated. The chance of a long-term cure with confirmed by many others, demonstrate the supe-
such intervention was approximately 27% [2]. In a rior results of changing surgical technique.
third study [3] mortality related to cancer was However, if occult rectal cancer recurrence is
reduced by 9-13% if computed tomography and detected, primarily by ERUS, surgery or radio-
frequent measurements of carcinoembryonic therapy are treatment options. Even in advanced
antigen were used by follow-up. ERUS was most cases an overall three-year survival rate of 60%
effective in a follow-up study conducted by with a local control rate of 73% could be
Lohnert et al. [4]. A total of 116 (34.3%) out of 338 achieved [7]
patients were shown to have local recurrence, Today and in the future the prevention of local
which was suggested by endoluminal ultrasound recurrence by a proper selection of primary cases,
and proven by endoluminal ultrasound-guided including the staging tool ERUS, the use of mod-
needle biopsy in all cases of unclear pararectal ern surgical techniques, mesorectal excision, and
structures. 25 patients in whom rectal cancer the optimal use of radiotherapy is the way for-
recurrence was detected solely by ERUS had a ward to improve results. Hopefully this will make
good outcome after salvage operation. Despite the postoperative follow-up, including ERUS, unnec-
successful identification of rectal recurrences by essary.
138 Atlas of Endoanal and Endorectal Ultrasonography

References 4. Lohnert MSS, Doniec JM, Henne-Bruns, D.

Effectiveness of endoluminal sonography in the iden-
1. Secco GB, Fardelli R, Rovida S, Gianquinto D, Baldi E, tification of occult local rectal cancer recurrences. Dis
Bonfante P, Derchi L, Ferraris R. Is intensive follow-up Colon Rectum 2000;43:483-91
really able to improve prognosis of patients with local 5. Bulow S, ChristensenIJ, Harling H, Kronborg 0, Fenger
recurrence after curative surgery for rectal cancer? C, Nielsen HJ. Recurrence and survival after mesorec-
Ann Surg OncoI2000;7:32-7 tal excision for rectal cancer. Br J Surg 2003;90:974-80
2. Tepper JE, O'Connell M, Hollis D, Niedzwiecki D, Cooke 6. Kapiteijn E, Marijnen CA, Nagtegaal ID, Putter H,
E, Mayer RJ. Analysis of surgical salvage after failure of Steup WH, Wiggers T, Rutten HJ, Pahlman L, Glimelius
primary therapy in rectal cancer: results from inter- B, van Krieken JH, van de Velde CJ. Preoperative radio-
group study 0114. J Clin Oncol 200P1:3623-8 therapy combined with total mesorectal excision for
3. Renehan AG, Egger M, Saunders MP, O'Dwyer ST. resectable rectal cancer. N Engl J Med 2001; 345:638-46
Impact on survival of intensive follow-up after curative 7. Wiggers T, Mannaerts, GH, Marinelli AW, Martijn H,
resection for colorectal cancer: systematic review and Rutten HJ. Surgery for locally recurrent rectal cancer.
meta-analysis of randomised trials. BMJ 2002; 324:813 Colorectal Dis 2003;5:504-7
 SECTION VII
Endoanal ultrasonography in the
staging of anal carcinoma
 Endoanal ultrasonography in the staging
of anal carcinoma
G.A. Santoro, G. Di Falco

Histologically the mucosa of the anal canal can be and mostly affect women. Anal tumors are classi-
divided into three zones (Fig. VII.I). The upper fied according to the WHO histological classifica-
part is covered with colorectal type mucosa. The tion (Table VI!.I).
middle part is the anal transitional zone (ATZ), Squamous cell carcinoma is the most common
which is covered by a specialized epithelium with malignancy of the anal canal (85%) and repre-
varying appearances; it extends from the dentate sents 3% of gastrointestinal carcinomas and
line and on average 0.5-1.0 cm upwards. The lower approximately 1.1% of large bowel malignancies
part extends from the dentate line and down- (Figs. VII.2, 3). It is a malignant epithelial neo-
wards to the anal verge and has formerly been plasm that is frequently associated with chronic
called the pecten. It is covered by squamous HPV infection. It typically occurs among patients
epithelium, which may be partly keratinized, par- in their 6th or i h decade of life. However it may
ticularly in the case of mucosal prolapse. The occur in young adults, particularly in patients
perianal skin (the anal margin) is defined by the with cellular immune incompetence. Still a rela-
appearance of skin appendages. tively rare disease (incidence rates between 0.5
Anal tumors account for about 2.5-5% of all and 1.0 per 100000 in women and between 0.3 and
malignant tumors of the colon. Most frequently 0.8 per 100000 in men), anal squamous cell carci-
they occur in patients between 55-60 years of age noma has shown a remarkable increase in inci-

Fig. VII.I. Histologically the anal canal

can be divided into three zones: the upper
part is covered with colorectal type
mucosa, the middle part is the anal tran-
sitional zone which is covered by a spe-
Dentate line cialized epithelium and the lower part is
covered by squamous epithelium
142 Atlas of Endoanal and Endorectal Ultrasonography

Table VII. I. WHO histological classification of tumors of the anal canal

EPITHELIAL TUMORS Intraephitelial neoplasia (dysplasia)

- Squamous or transitional epithelium
- Glandular
- Paget disease

Carcinoma
- Squamous cell carcinoma
- Adenocarcinoma
- Mucinous adenocarcinoma
- Small cell carcinoma
- Undifferentiated carcinoma
- Others

Carcinoid tumor

MALIGNANT MELANOMA

NON-EPITHELIAL TUMORS

SECONDARY TUMORS

dence during the past half century, because of a ually transmitted diseases. Sexually transmittable
rise in the prevalence of cases reported in younger human papilloma viruses, particularly serotypes
males. For both men and women, urban popula- 16 and 18, are detected in the majority of these
tions are at higher risk than rural populations and tumors. States of cellular immunosuppression are
there are considerable racial differences in inci- associated with increased risk of anal squamous
dence. In the United States, blacks tend to have cell carcinoma. This has been observed for
higher incidence rates than whites, while Asians patients with HIV infection and AIDS and for
and Pacific Islanders appear to be at very low risk. renal transplant recipients. Anal intraepithelial
Homosexual men appear to constitute a group at neoplasia is often an unexpected finding in minor
particular risk (incidence 11 to 34 times higher surgical specimens. Clinical manifestations of
than in the general male population). Other sexu- anal cancer are often late and non-specific and are
al factors strongly associated with anal squamous mainly related to the tumor size and extent of
cell carcinoma include number of sexual partners, infiltration. They include anal pruritus, discom-
receptive anal intercourse and co-existence of sex- fort in sitting position, sensation of a pelvic mass,

Fig. VII.2. Ulcerating squamous cell carcinoma of anus Fig. VII.3. Squamous cell carcinoma of anus. (H&E stain)
 Section VII' Endoanal ultrasonography in the staging of anal carcinoma 143

Fig. VII.4. Locally advanced rectal adeno-

carcinoma extending downward to the
anal canal and deeply invading the
sphincteric muscles

pain, change in bowel habit, incontinence due to be related only to the stage at diagnosis and is
sphincter infiltration, discharge, bleeding, fissure poorer than that for squamous cell carcinoma.
or fistula. The initial non-specificity of clinical Basal cell carcinoma of the anal margin is pri-
features, attributed to more common anorectal marily found on sun-exposed areas. The etiology
disorders such as hemorrhoids, explains why is unknown and there is no evidence ofHPV infec-
diagnosis can be delayed. The diagnosis of anal tion. The tumor commonly presents as an indurat-
cancer is made on physical examination and biop- ed area with raised edges and central ulceration,
sy-proven histology. Digital rectal examination located in the perianal skin. Histologically it can
reveals size, location and relative fixation of the show the same variability in morphology as basal
tumor. Biopsy for tissue diagnosis is easy because cell carcinoma elsewhere. It is important to distin-
lesions are very distal and within simple to reach. guish it from squamous carcinoma, and this may
The distinction between anal canal and anal be particularly difficult.
margin squamous cell carcinoma may be diffi- Anal melanoma is rare (Fig. VII.6). It is an adult
cult, as tumors often involve both areas at the disease; most patients are white. Presentation is
time of diagnosis. This may account for the vary- usually with mass and rectal bleeding, but tenes-
ing data on prognosis, but this is generally better mus, pain and change in bowel habit also occur.
for anal margin carcinoma than for anal canal Lesions may be sessile or polypoid. Pigmentation
carcinoma. of the lesion is often appreciated. Satellite nodules
Anal canal adenocarcinoma is an adenocarci- may occur. Anal melanomas spread by lymphatics
noma arising in the anal canal epithelium, includ- to regional nodes and hematogenously to the liver
ing the mucosal surface, the anal glands and the and hence to other organs. Metastases are frequent
lining of fistulous tracts. The clinical features of at the time of presentation, and the prognosis is
anal adenocarcinoma of colorectal type do not poor; the s-year survival is less than 10%.
differ from those of anal squamous cell carcino- In 1997 the International Union Against Cancer
ma. Most adenocarcinomas found in the anal and the American Joint Committee on Cancer
canal represent downward spread from an adeno- developed a common staging system for anal car-
carcinoma in the rectum or arise in colorectal cinoma based on TNM classification (Table VIl.2)
type mucosa above the dentate line (Fig. VIlA). [1]. Although this system is widely accepted as the
Macroscopically and histologically they are indis- standard staging system, it has little impact on
tinguishable from ordinary colorectal type ade- treatment decisions because most patients receive
nocarcinomas and do not seem to represent a spe- chemoradiation therapy as a first-line treatment
cial entity except for their low location (Fig. VII.S). regardless of their stage. The existing TNM classi-
The prognosis for anal adenocarcinoma seems to fication for anal cancer, based on the result of rec-
 144 Atlas of Endoanal and Endorectal Ultrasonography

IAdenocarcinoma I

Fig. VII.S. (a) Anal canal adenocarcino-

ma deeply invading the internal sphinc-
, teric muscle. (b) Larger view of a. (H&E
b stain)

tal examination, where only the margins of the VII.7-1O). Depth of penetration may be more
tumor around the anal circumference and its prox- important than size alone because some studies
imal and distal ends are assessed, emphasizes have demonstrated that the size of the primary
tumor size and not depth of penetration (Figs. tumor makes little difference in prognosis in

a ~ ____________________________________ ~

Fig. VII.6. (a) Histologic view of malignant melanoma of the anus. (b) High-power view showing some malignant cells rich in
melanin and others amelanic. (H&E stain)
 Section VII' Endoanal ultrasonography in the staging of anal carcinoma 145

Table VII.2. AJCC anal cancer staging

Primary tumor (T)

Tx Primary tumor cannot be assessed
TO No evidence of primary tumor
Tis Carcinoma in situ
T1 Tumor 2 em or less in greatest dimension
T2 Tumor more than 2 em but not more than 5
em in greatest dimension
T3 Tumor more than 5 em in greatest dimension
T4 Tumor of any size invades adjacent organs
(vagina, urethra, bladder)

Regional lymph nodes (N)

Nx Regional lymph nodes cannot be assessed Fig. VII.S. Diagrammatic representation of T2 anal canal
NO No regional lymph node metastases tumor (see Table Vn.2)
N1 Metastases in perirectal lymph node(s)
N2 Metastases in unilateral internal iliac and/or
inguinal lymph node(s)
N3 Metastases in perirectal and inguinal lymph
nodes and/or bilateral internal iliac and/or
inguinal lymph nodes

Metastases (M)
MO No distant metastases
M1 Distant metastases

Fig. VII.9. Diagrammatic representation of T3 anal canal

tumor (see Table Vn.2)

Fig. VII.7. Diagrammatic representation of T1 anal canal Fig. VII.lO. Diagrammatic representation of T4 anal canal
tumor (see Table VIL2) tumor (see Table VII.2)
146 Atlas of Endoanal and Endorectal Ultrasonography

Table VII.3. Ultrasound anal cancer staging system Table VII.4. Ultrasound anal cancer staging system
from Boman et al. [2] from Bartram and Burnett [3]

Tl Tumor confined to the sub epithelium Tl Tumor limited to internal anal sphincter
T2 Tumor limited to the sphincter muscles T2 Tumor involves external anal sphincter
T3 Tumor penetrating through the external anal T3 Tumor extends outside external anal
sphincter sphincter
T4 Tumor involves adjacent organ T4 Tumor involves adjacent organ
(vagina, bladder) (vagina, bladder)

patients recelvmg chemoradiation therapy. that distinguishes clearly early tumors limited to
Clinical staging has the disadvantage of being a the submucosa and also differentiates lesions
method that depends on the experience of the based on the depth of invasion into the sphincter
investigator; it allows no objective documentation. complex. This system may have a more practical
Endoanal ultrasonography (EAUS) enables relevance when determining treatment options.
assessment of spread in terms of proximal and An anal carcinoma in an EAUS image appears
circumferential extension and infiltration of deep as a hypo echoic mass, most commonly not homo-
layers. Furthermore EAUS allows detection of geneous, with areas of degeneration and with
involved lymph nodes and enables the follow-up irregular outlines (Figs. VII.1l-15). In literature
of irradiated carcinomas. Anal cancer is hypoe- EAUS has been reported to be accurate in staging
choic and the extent of the tumor is quite easy to the anal canal tumors (overall accuracy 86%). In a
define and relate to the sphincters. To have a high study by Novell et al. [5] accuracy of EAUS in local
resolution of the tissues near the transducer the staging of anal cancers was 85.7%. The sensitivity
examination should be performed with a 10 MHz of endosonography in visualizing enlarged lymph
probe with a focal length of 1-4 cm covered with nodes was 83%.
the hard plastic cone. In the anal canal a water- Lymph nodes larger than 3 mm are already vis-
filled balloon is compressed and an exact deter- ible on EAUS. The studies comparing EAUS and
mination of the depth of infiltration of a small magnetic resonance imaging (MRI) show that
tumor may be very difficult and inaccurate. MRI is inferior to EAUS in lymph node staging
Various ultrasound-staging systems have evolved (39-95% vs. 62-83%, respectively) [6]. The accuracy
for anal cancers. Boman et al. [2] (Table VII.3) of CT is between 22-73%. The size oflymph nodes
proposed a system in which Tl lesions are limited appeared to be an unreliable criterion, whereas
to the submucosa and invasion into the "sphincter their echogenicity appeared to be more reliable.
complex" is considered a T2lesion; no distinction Hypoechoic lymph nodes representing metastases
is made between invasion into the internal and were predicted with a sensitivity of 72%.
external sphincter muscles. Large uT 4 carcinomas Indinnimeo et al. [7] suggested a follow-up
cannot be seen in their outermost circumference protocol for anal canal carcinoma which consists
whereas computed tomography (CT) allows an
exact staging and shows the correlation between
the tumor and adjacent structures.
Bartram and Burnett [3] (Table VII.4) refined
staging of anal carcinoma with anal endosonog- Table VII.S. Ultrasound anal cancer staging system
raphy, by identifying tumor spread in relation to from Tarantino and Bernstein [4]
sphincter involvement, differentiating tumors that
invade only the internal sphincter muscle from Tl Tumor confined to the submucosa
those that penetrate into the external sphincter. T2a Tumor limited to internal anal sphincter
However they did not make any allowance for T2b Tumor involves external anal sphincter
tumors limited to the submucosa. T3 Tumor invades through the sphincter into the
Tarantino and Bernstein [4] (Table VII.5) pro- perianal tissues
posed a modified ultrasound staging system T4 Tumor involves adjacent organ
based on the TNM classification of anal tumors (vagina, bladder)
 Section VII- Endoanal ultrasonography in the staging of anal carcinoma 147

Fig. VII.lI. UTI squamous-cell carcinoma of the anal canal

(arrows) (a, b). The lesion appears as a hypoechoic mass
invading the internal sphincter. The hyperechoic layer corre-
sponding to the external sphincter is intact. (c) Sigmoidoscopy
shows a soft polypoid lesion in the right lateral wall located
c 2cm from the anal verge
148 Atlas of Endoanal and Endorectal Ultrasonography

Fig. VII.12. UT2 anal carcinoma.

Hypoechoic fingers of tumor (arrows) are
pushing their way into the surrounding
external sphincter

in performing rectal examination, anoscopy, Tru- resolution of edema and inflammatory changes in
cut needle biopsies and EAUS for local control, the tissue (Fig. VII.I6). Using this time frame,
and inguinal ultrasound and CT to evidence dis- ultrasound could be used to monitor response to
tant metastases. EAUS can be used to evaluate treatment, with biopsies indicated only in the
tumor response to chemoradiation. Optimal tim- select group of patients in which there is a ques-
ing for initial posttreatment studies is as yet unde- tion of residual tumor on EAUS. It is also possible
termined. Six or eight weeks after completion of to do ultrasound-guided biopsies with the aid of
treatment is too soon to evaluate the patient with an ultrasound probe with a biopsy channel. EAUS
EAUS and in most cases there are still residual may also have a role in evaluating other treatment
effects of therapy. Sixteen to twenty weeks postra- modalities [8]. EAUS could allow early compar-
diation is a sufficient time frame to allow for the isons to be made in the efficacy of various treat-
full effect of radiation to occur and to allow for ment regimens by giving examiners an objective
means to document tumor response. EAUS may
have a role in surveillance of patients after suc-
cessful treatment of the initial disease. Ultrasound
may be useful in distinguishing between thicken-
ing caused by tumor recurrence and that caused
by scarring. Using ultrasound to follow abnormal
areas before there is clinical evidence of recur-
rence can still allow for earlier detection. Scarred
areas will remain unchanged or improve with suc-
cessive studies, whereas tumor recurrence will
progress. In the case of progression of lesions,
biopsies with ultrasound guidance may identify
the recurrence before it is clinically evident [8].
3D image analysis facilitates assessment of sus-
picious anal lesions by contemporary display of
three perpendicular scan planes (section display)
or volume reconstructions of the scanned area
(volume display) (Fig. VII.I7). The latter is very
useful in enhancing the understanding of the
Fig. VII. 13. UT2 anal carcinoma invading subcutaneous part
anatomy and defining the spatial relationship of
of the external anal sphincter (arrows) the tumor with normal structures.
 Section VII, Endoanal ultrasonography in the staging of anal carcinoma 149

Fig. VII.I4. Nearly circumferential anal

carcinoma located at the level of the pub-
orectalis muscle (a, b). The lesion appears
as a markedly thickened area completely
invading the internal sphincter and
extending through the puborectalis mus-
b cle (ultrasound staging: uT3)

Fig. VII.IS. uT4 anal carcinoma extending outside external

sphincter (arrows) and infiltrating os coccyx
 150 Atlas of Endoanal and Endorectal Ultrasonography

Fig. VII.16. uT3 anal carcinoma before

combined radio chemotherapy, 55-year-
old male (a, b). Four months after com-
bined radio chemotherapy note residual
disease limited to internal anal sphincter
c (ultrasound staging: UTI) (c)
 Section VII- Endoanal ultrasonography in the staging of anal carcinoma 151

Fig. VII.I7. 3D projection of a small uTI

anal tumor limited to internal anal
sphincter (arrows). The layers of the anal
canal can clearly be seen with the sagittal
view

References of ilie anal canal: potential implications of a new ultra-

sound staging system. Dis Colon Rectum 2002;45:16-22
1. American Joint Committee on Cancer. Anal canal. In: 5. Novell F, Trias M. Intraluminal anorectal ultrasonogra-
Fleming ID, Cooper JS, Henson DE, eds. AJCC cancer phy in the staging of anal canal cancer. Rev Esp
staging manual. 5th ed. Philadelphia: Lippincott- Enferm Dig 1993;84:153-5
Raven, 1997:91-6 6. Sudol-Szopinska I, Szczepkowski M, Jakubowski W.
2. Boman BM, Moertel CG, O'Connell MJ, Scott M, Anal ultrasound in the diagnosis of anal carcinoma.
Weiland LH, Beart RW, Gunderson LL, Spencer RJ. Radiol Oncol 2001;35:273-6
Carcinoma of the anal canal. A clinical and pathologic 7. Indinnimeo M, Cicchini C, Stazi A, Ghini C, Mingazzini
study of 188 cases. Cancer 1984;54:114-25 P, Laghi A. Analysis of a follow-up program for anal
3. Bartram CI, Burnett SJ. Atlas of anal endosonography. canal carcinoma. J Exp Clin Cancer Res 2001;20:199-203
Oxford: Butterworth-Heinemann, 1991 8. Herzog U, Boss M, Spichtin HP. Endoanal ultrasonog-
4. Tarantino D, Bernstein MA. Endoanal ultrasound in ilie raphy in the follow-up of anal carcinoma. Surg Endosc
staging and management of squamous-cell carcinoma 1994;8:1186-9
 Invited commentary
A. Infantino

The incidence of anal cancer, mainly represented T2NO by EAUS had a complete response to treat-
by anal canal epidermoid carcinoma, has ment with reduced local recurrences, than tumors
increased in the last few decades although it still classified as Tl-T2NO according to UICC classifi-
represents a rare condition. The clinical suspicion cation (94.5% vs. 80%, respectively; P=0.008), and
should always exist when evaluating more com- a multivariate statistical analysis demonstrated
mon benign proctologic lesions. Drs. Santoro and that the depth of tumor invasion (T stage) evalu-
Di Falco have focused on the importance of a cor- ated by EAUS was the only predictive factor of
rect staging even though studies designed in patient survival [2].
order to demonstrate the opportunity of different EAUS is a valuable procedure in the follow-up
treatment strategies for the different stages of anal of anal cancer distinguishing a scar from residual
canal epidermoid cancer (NCI, Le lutte contre Ie cancer or a recurrence. At successive examina-
cancer, UKCCCR) are still in progress. In particu- tions a scar shows no variation or a reduction in
lar, low chemoradiation doses or radiotherapy size while the volume of a local recurrence
alone could eradicate early stage anal cancers with increases with time. Moreover, it is possible to
lower local and systemic toxicities. Following this undertake an EAUS-guided trucut needle biopsy
statement we need a method for a correct staging of any suspicious lesion [3].
of this neoplasia. Nevertheless, considering results of studies on
Endoanal ultrasonography (EAUS) has been EAUS in the staging of rectal cancer, we can
demonstrated to be accurate in evaluating the cor- assume that skilled operators can achieve a high
rect depth of tumor invasion and, to some extent, level of accuracy; in fact the best correlation with
the presence of meso rectal lymph node metas- pathological specimens is reached after 50 exami-
tases. As a consequence of the high efficacy of this nations [4].
diagnostic procedure, new staging systems have Even in the absence of large trials comparing
been proposed [1]. In a multicentric prospective EAUS to magnetic resonance imaging or to com-
trial the sonographic staging system resulted in puted tomography, we can conclude that EAUS is
having a higher prognostic value compared with highly efficient in the staging of anal cancer, it is
the common clinical UICC staging system. In this useful in the follow-up, it is easily repeatable and
study a higher rate of anal tumors, classified as Tl- has a low cost.
 Section VII' Endoanal ultrasonography in the staging of anal carcinoma 153

References (ERUS). Results of a prospective multicenter study.

Endoscopy 2001;33:231-6
1. Tarantino D, Bernstein MA. Endoanal ultrasound in the 3. Indinnimeo M, Cicchini C, Stazi A, Ghini C, Mingazzini
staging and management of squamous-cell carcinoma P, Laghi A. Analysis of a follow-up program for anal
of the anal canal: potential implications of a new ultra- canal carcinoma. J Exp Clin Cancer Res 2001;20:199-203
sound staging system. Dis Colon Rectum 2002;45:16-22 4. Mackay SG, Pager CK, Joseph D, Stewart PJ, Solomon
2. Giovannini M, Bardou VJ, Barclay R, Palazzo L, Roseau MJ. Assessment of the accuracy of transrectal ultra-
G, Helbert T, Burtin P, Bouche 0, Pujol B, Favre O. Anal sonography in anorectal neoplasia. Br J Surg 2003;
carcinoma: prognostic value of endorectal ultrasound 90:346-50
 SECTION VIII
Surgical treatment options
for rectal cancer

n
!l C
"q;r:
-iio
"·i
.,3 .....
~~o
~

.,"
 VIII.2.
Traditional local excision for rectal cancer
M. Trompetto

In patients with a rectal cancer confined to the Clinical evaluation of the tumor permits to
bowel wall, local excision is commonly accepted define its size and mobility, two important criteria
as the procedure of choice, permitting cure rates for choosing the optimal surgical approach. Today
equivalent to those obtained with abdominal a very accurate preoperative anatomical staging
approaches. Furthermore this technique has a of rectal cancer can be achieved by endorectal
negligible mortality, a very low morbidity and ultrasonography (ERUS) that provides a precise
provides excellent functional results. evaluation of the invasion of the bowel wall and
Unfortunately no more than 8% of patients mesorectum [8], permitting to distinguish
with rectal cancer can be treated using only a local between non-invasive lesions (pTo tumors) and
excision [1,2]. invasive rectal cancers (pT1 tumors) [9].Although
Of course local excision must also be consid- more costly and not so accurate as endoscopic
ered in the case of patients with extensive dissem- ultrasound, magnetic resonance imaging (MRI)
inated disease treated to exclusively achieve local with phase-arrayed coils has excellent sensitivity
control, for patients who cannot have major in detecting transmural invasion of rectal cancer
surgery because of the high risk of peri/postoper- [10 ].
ative complications related to their general condi- Of course to make a careful local excision pos-
tions or concomitant disorders and for patients sible the tumor must be accessible, with its upper
who do not accept colostomy as a part of the sur- edge at no more than 6-8 cm from the dentate
gical treatment. line. In 1983 Hager et al. [11] in one of the first
The main difficulty dealing with rectal cancer papers published on local excision for rectal can-
is in identifying patients with "early" disease, par- cer used a very strict criterion of selection:
ticularly those patients without lymph node tumors less than 3 cm in diameter, confined to the
metastases. An inaccurate selection of cases can mucosa or submucosa; malignancy grade 1 or 2
lead to dangerous undertreatment due to an unre- and no signet-cell carcinoma. They obtained a
sected mesorectum with a high possibility of five-year survival rate of 89±21.7% for patients
regional and/or diffuse recurrences of the disease. with submucosal invasion and 78±49.9% for cases
Pathological studies on rectal cancers treated with involvement of muscularis propria. Grigg et
by radical surgery demonstrated that the inci- al. [12] using the same selection criteria plus the
dences of lymph node metastases was 3-17% in presence of a pedicle or pseudopedicle obtained a
pT1 tumors [3,4], increasing to 38% in pT2 tumors 100% five-year cancer specific survival, with an
[5] and reaching 61% in pT3 tumors [6,7]. In addi- incidence of 6.2% of lymph node metastases in
tion mucinous, undifferentiated or poorly differ- pT1 tumors.
entiated rectal cancers or tumors with lymphatic Specific selection criteria for the local treat-
invasion have a greater risk of lymph node ment of rectal tumors with curative intent are
involvement and always need a radical resection summarized in Table VII I. I.
with complete removal of the mesorectum.
 VIII.I.
Introduction

G.A. Santoro, G. Di Falco

Precise staging of rectal carcinomas is mandatory scopic spread of tumors is rare, occurring in less
in order to improve the long-term results and to than 7% of cases, so reducing the distal margin of
individualize therapeutic approach. Endorectal resection from 5 cm to 2 cm. In recent years total
ultrasound is important in accurately staging the mesorectal excision (TME) has been accepted as
tumor by identifying both depth of invasion and the optimal treatment for rectal cancer. This tech-
presumptive nodal status. Conventional local nique is associated with low local recurrence and
excision (LE) of a properly selected group of increased disease free survival.
patients with low risk rectal cancers can provide The role of minimally invasive surgery for rec-
long-term survival, with minimal morbidity, neg- tal cancer continues to be evaluated critically. The
ligible mortality and excellent functional results. laparoscopic approach has been generally shown
Under a curative intention, local excision of distal to be technically and oncologically feasible, with
rectal tumors (located 4 to 18 cm from the anal overall patient survival and local and distant
verge) can be performed by transanal endoscopic recurrence rate comparable to those with conven-
microsurgery (TEM). In the future (neo-) adju- tional open resection. Laparoscopic liver sonogra-
vant therapies can further improve results of phy assists surgeons in critical decision-making
these procedures. by providing a more complete neoplastic staging
Recent improvements in surgical techniques at time of surgery.
have allowed more patients with rectal cancer to Advances in technology continue at a rapid
undergo sphincter-sparing operations. The pace. Robotics and telemedicine are taking
biggest change in surgery for rectal cancers came surgery well into the 21st century.
with the anatomical concept that the distal micro-
 Section VIII· Surgical treatment options for rectal cancer 159

Table VIII.I. Specific selection criteria for the local Trans-sacral resection
treatment of rectal tumors with curative intent

Criteria Indications The technique was presented by Kraske in 1885 [17]

even if it was first described by Kocher in 1874. It is
Fixation Mobile performed with the patient placed in the classical
Size 3-5 cm prone jack-knife position. With this approach the
Pelvic nodes Not involved (ERUS) rectum is reached by removing the coccyx, with or
Histology Well (moderate) differentiated without portioning of the sacrum through a trans-
Lymphatic invasion Absent verse incision over the anococcygeal ligament.
ERUS staging UTl-uT2 After entering the presacral space the fibers of the
Technically accessible Yes levator ani muscle are longitudinally divided in the
midline plane and the rectum can be easily mobi-
lized, and delivered through the incision, permit-
Transanal excision ting the positioning of a right angle bowel clamp
distal to the tumor. A second bowel clamp is placed
Transanal excision is mostly performed with the proximally to the tumor, the rectum is resected and
patient placed in a lithotomy position for tumors of an end-to-end anastomosis is performed.
the posterior wall of the rectum, and in the prone This approach is rarely considered and few
jack-knife position for anterior lesions. Lateral series in literature report the results of trans-
position can also be used, but only few colorectal sacral resections [18].
surgeons consider it mandatory for lateral tumors.
A complete mechanical bowel preparation is
always recommended as well as antibiotic preop- Trans-sphincteric procedure
erative administration.
The use of submucosal injection of saline solu-
tion to reduce intraoperative bleeding can be use- This old technique was reintroduced by York-
ful and moreover there is no evidence that this Mason in 1970 [19]. It consists in a local resection
can cause the spreading of tumor cells. through a posterior approach without a concomi-
In the case of a preoperative diagnosis of can- tant abdominal approach. The transection of the
cer a full-thickness excision of the tumor with a entire sphincteric compartment is associated with
free margin of 1-2 cm is mandatory, whereas in the a high percentage of incontinence and this tech-
case of a villous adenoma a submucosal removal nique has been very rarely considered after the
of the lesion can be accepted [13] (Fig. VIlLI). introduction of staplers.
Some lateral and proximal sutures can be very Both trans-sacral and trans-sphincteric tech-
useful for traction in cases of lesions in the mid- niques present a great number of complications,
dle/upper rectum, where special care must be as anastomotic leaks [20], wound breakdown,
taken to avoid peritoneal perforation, particularly fecal incontinence, enterocutaneous fistula for-
in cases of anterior-sided tumors in patients with mation [21,22], and the majority of surgeons con-
a deep pouch of Douglas [14]. siders them only of historic interest.
The closure of the rectal wound is performed
using interrupted absorbable 2-0 or 3-0 sutures.
The suture must be transverse to avoid any rectal Specific considerations
stenosis and must be placed through the full-
thickness of the rectal wall.
Recently some new instruments initially If the removed tumor is found to have a poor dif-
designed for laparoscopic surgery have been test- ferentiation, positive margins, lymphatic and/or
ed for local excision of rectal cancers [15, 16]. muscularis propria invasion, the performed local
These approaches could be very attractive, but up excision must be considered only as a macrobiop-
to date no prospective study concerning their sy and a major surgical resection is recommend-
oncological results has been published. ed [23].
The excised specimen must be oriented, Patients, who undergo immediate resection for
labeled and fixed on a cork for the pathologist. the presence of adverse pathologic findings after
 160 Atlas of Endoanal and Endorectal Ultrasonography

a d

b ' -_ _ _ _ _ _ ___ .....

e
~_~ ~.;..... _""""'"~d.

c
Fig. VIII. I. Large villous adenoma occupying 40% of the rectal circumference in right anterolateral aspect and located at 5 cm
from the anal margin (a). Saline solution is injected in the submucosa to reduce intraoperative bleeding (b). The mucosa is
incised near the base of the tumor (c). Complete submucosal dissection of the lesion with a 5 mm tumor free resection margin
(d). Defect is closed with a transverse suture (e). Surgical specimen of the large polypoid lesion that is 3 cm at its greatest diam-
eter (f)
 Section VIII' Surgical treatment options for rectal cancer 161

local excision of rectal tumors, have a better prog- treatment are represented by the presence of
nosis when compared to patients resected after positive margins, endothelial-lined space inva-
local recurrence [24], so in these cases early sion, poor differentiation, perineural invasion or
abdominal surgery is advisable. invasion of the muscularis propria [32]. Some
If the selection criteria are complied and the reports concerning this combined therapy are
pathologic findings are favorable, transanal exci- encouraging, with good results also after long-
sion of rectal carcinomas can be performed with term follow-up [33] but others authors [34] clear-
good overall survival and local control [25-27]. ly demonstrated an elevated local recurrence
Gao et al. [28] reported a 91% overall s-year sur- rate in cases of local excision of pT2-T3 tumors
vival rate after local excision of rectal cancers even after adjunction of postoperative radio-
(94.4% in pTI tumors and 83.3% in pT2 tumors, therapy.
respectively). However other authors [29,30], who Preoperative radiation followed by full-thick-
demonstrated a high risk of local recurrence, par- ness trans anal excision has never been tested in a
ticularly in patients with pT2 cancers, are more prospective study but could be a possible future
cautious. Chorost et al. [31] reported that local neoadjuvant treatment for T2-T3 cancers unsuit-
excision has an unacceptable high rate of local able for resective surgery.
recurrence and that long-term results after sal- All patients undergoing local treatment for
vage surgery remain unknown. rectal cancer require careful long-term follow-up
Many reports on postoperative radiotherapy because they are at significant risk for local recur-
after local excision of rectal cancers have been rence and distant failure [3S]. Only a close follow-
published. Main indications for this adjuvant up may allow a radical salvage surgery [36].

References 10. Gagliardi G, Bayar S, Smith R, Salem RR. Preoperative

staging of rectal cancer using magnetic resonance
1. Rouanet P, Aubert BS, Fabre JM et al. Conservative imaging with external phase-arrayed coils. Arch Surg
treatment for low rectal carcinoma by local excision 2002; 137:447-51
with or without radiotherapy. Br J Surg 1993; 80:1452-6 11. Hager T, Gall FP, Hermanek P. Local excision of cancer
2. Fortunato L,Ahmad NR, Yeung Rs et al. Long-term fol- of the rectum. Dis Colon Rectum 1983; 26:149-51
low-up oflocal excision and radiation therapy for inva- 12. Grigg M, Mc Dermott FT, Pihl EA, Hughes ES. Curative
sive rectal cancer. Dis Colon Rectum 1995; 38:1193-5 local excision in the treatment of carcinoma of the rec-
3. Hojo F, Koyama Y, Moriya Y. Lymphatic spread and its tum. Dis Colon Rectum 1984; 27: 81-3
prognostic value in patients with rectal cancer. Am J 13. Bailey HR, Huval Wv, Max E et al. Local excision of car-
Surg 1982;144:350-4 cinoma of the rectum for cure. Surgery 1992; 1l1:555-61
4. Banerjee AK, Jehle EC, Shorthouse AJ, Buess G. Local 14. Nivatvongs S, Wolff BG. Technique of per anal excision
excision of rectal tumors. Br J Surg 1995; 82:1165-73 for carcinoma of the low rectum. World J Surg 1992;
5. Killingback MJ. Local excision of carcinoma of the rec- 16:447-50
tum: indications. World J Surg 1992; 16: 437-46 15. De Gennaro VA, Lescher TC. Transanal excision of rec-
6. Liu SK, Church JM, Lavery IC, Fazio VW. Operation in tal tumors using a laparoscopic stapler. Dis Colon
patients with incurable colon cancer- is it worthwhile? Rectum 1995; 38:327-8
Dis Colon Rectum 1997; 40:11-4 16. Qureshi MA, Monson JR, Lee PW. Transanal Multifire
7. Billingham RP. Conservative treatment of rectal can- Endo GIA technique for rectal polipectomy. Dis Colon
cer. Extending the indications. Cancer 1992; 70:1355-63 Rectum 1997; 40:116
8. Mackay SG, Pager CK, Joseph D, Stewart PJ, Solomon MJ. 17. Kraske P. Extirpation of high carcinomas of the large
Assessment of the accuracy of transrectal ultrasonogra- bowel. Dis Colon Rectum 1984; 27: 499-503
phy in anorectal neoplasia. Br J Surg 2003; 90:346-50 18. Huber PJ, Reiss G. Rectal tumors: treatment with a
9. Starck M, Bohe M, Simanaitis M, Valentin 1. Rectal posterior approach. Am J Surg 166:760-3
endosonography can distinguish benign rectal lesions 19. Mason AY. The place oflocal resection in the treatment
from invasive early rectal cancers. Colorectal Dis 2003; of rectal carcinoma. Proc Roy Soc Med 1970; 63:1259-62
5:246-50 20. Localio SA, Eng K, Coppa GF. Abdomino-sacral resec-
162 Atlas of Endoanal and Endorectal Ultrasonography

tion for mid-rectal cancer: a fifteen-year experience. sion carcinoma in early stage. World J Gastroenterol
Ann Surg 1983; 198:320-4 2003; 9:871-3
21. Lazorthes F, Fages P, Chiotasso P et al. Resection of the 29. Mellgren A, Sirivongs P, Rothenberger DA, Madoff RD,
rectum with construction of a colonic reservoir and Garcia-Aguilar J. Is local excision adequate therapy for
colo-anal anastomosis for carcinoma of the rectum. Br early rectal cancer? Dis Colon Rectum 2000; 43:1064-71
J Surg 1986; 76:136-8 30. Sengupta S, Tjandra JJ. Local excision of rectal cancer:
22. Lazorthes F, Fages P, Chiotasso P et al. Synchronous what is the evidence ? Dis Colon Rectum 2001; 44:
abdomino-trans-sphincteric resection of low rectal 1345-61
cancer: new technique for direct colo-anal anastomo- 31. Chorost MI, Petrelli NJ, McKenna M et al. Local exci-
sis. Br J Surg 1986; 73:573-5 sion of rectal carcinoma. Am Surg 2001; 67:774-9
23. Masaki T, Sugiyama M, Atomi Y, MatsuokaH, Abe N, 32. Mendenhall WM, Morris CG, Rout WR, Zlotecki RA,
Watanabe T, Nagawa H, Muto T. The indication of local Lind DS, Hoch SN, Schell SR, Copeland EM 3rd . Local
excision for T2 rectal carcinomas. Am J Surg 2001; excision and postoperative radiation therapy for rectal
181:133-7 carcinoma. Int J Cancer 2001;96: S89-96
24. Baron PL, Enker WE, Zakowski MF, Urmacher C. 33. Minsky BD, Enker WE, Cohen AM, Lauwers G. Local
Immediate vs. salvage resection after local treatment excision and postoperative radiation therapy for rectal
for early rectal cancer. Dis Colon Rectum 1995;38:177-81 cancer. Am J Clin Onco11994; 17:411-6
25. Hoth JJ, Waters GS, Pennell TC. Results of local exci- 34. Pigot F, Dernaoui M, Castinel A, Juguet F, Chaume JC,
sion of benign and malignant rectal lesions. Am Surg Faivre J. Local excision with postoperative radiothera-
2000; 66:1099-103 py for T2 or T3 distal rectal cancer. Long-term results.
26. Koscinski T, Malinger S, Drews M. Local excision of Ann Chir 2001; 126:639-43
rectal carcinoma not-exceeding the muscular layer. 35. Moore HG, Guillem JG. Local therapy for rectal cancer.
Colorectal Dis 2003; 5:159-63 Surg Clin North Am 2002; 82:967-81
27. Visser BC, Varma MG, Welton ML. Local therapy for 36. Rothenberger DA, Garcia-Aguilar J. Role of local exci-
rectal cancer. Surg Onco12001; 10:61-9 sion in the treatment of rectal cancer. Semin Surg
28. Gao JD, Shao YF, Bi JJ, Shi SS, Liang J, Hu Yh. Local exci- Onco12000; 19:367-75
 Invited commentary
C.O. Finne

Local excision for rectal cancer is a popular but have enthusiastically championed local therapy
unproven treatment for localized rectal cancer. for selected rectal cancers, has prompted caution
Only one clinical trial comparing local excision in our enthusiasm for this modality. We found an
with standard resection has ever been done [1]. 18% local recurrence rate for pathologically staged
This trial compared the outcomes of local recur- Tl cancers (pTl) with 2% dying from cancer and
rence and five-year survival rates between 24 37% local recurrence rate for pT2 cancers with 7%
patients having transanal endoscopic microsurgi- dying and 4% alive with disease following salvage
cal excision (TEM) versus 26 patients undergoing [6]. Most of these patients had been followed care-
standard resection for pathologically staged Tl fully and regularly and could be salvaged by fur-
node negative carcinomas of the rectum. The out- ther surgery, but the high local recurrence rate
comes were not statistically different with a short prompted a reexamination of our rectal cancer
mean follow-up of slightly less than four years. experience. A comparison of our local excision
Consistent with expectations, complications (21% patients with similar patients treated by radical
vs. 34%), blood loss (143 cc vs. 745 cc), operative excision during the same time frame revealed a
times (103 minutes vs.147 minutes), analgesic use, clear advantage to those treated by radical means.
and length of hospitalization (6 days vs. 15 days) There were no local recurrences in 30 Tl cancers
all favored the local procedure. (5% cancer deaths) treated by radical means, but
Local therapies for rectal cancer became accept- 18% local recurrence (5% cancer deaths) in 69 Tl
able over the past 40 years precisely because mor- cancers treated with local excision. T2 cancers
bidity and mortality are low when compared to tra- treated by radical means experienced 6% local
ditional surgery. Medicare claims data in the recurrence (9% cancer deaths) versus 47% local
United States shows that perioperative mortality for recurrence after local excision with 19% cancer
colectomy rises from 3.3% in patients 66-69 years of deaths, indicating that salvage was not as reliable
age to 9.3% in patients 85 or older [2]. These data after recurrence of T2lesion treated locally [7].
are corroborated by data from Australia and France Long-term follow-up of three other series has
[3, 4]. Operative mortality associated with local shown unexpectedly high recurrence rates, but
excision is an order of magnitude less: even when also questioned the efficacy of adjuvant radiation
combined with radiation averages 0.5% or less [5]. therapy for T2 cancers. A combined series from
This is especially significant when one realizes that Emery University and Harvard demonstrated a
most series of local excisions are composed of median time to recurrence in the irradiated group
patients considered unfit for conventional surgery. of 55 months versus 13.5 months in the non-irradi-
Local therapy is also associated with avoidance of ated group [8]. Local recurrence and actuarial
colostomy, often a bigger factor in choice of treat- cancer mortality was unacceptably high for T2
ment by patients than oncologic outcome. cancers in their series, as well: 67% (67% cancer
Reexamination of local excision data from my deaths) in the non-irradiated group and 15% (24%
institution, the University of Minnesota, where we cancer deaths) in the irradiated group. Memorial
164 Atlas of Endoanal and Endorectal Ultrasonography

Sloan-Kettering reexamined their series in 2002 The concept of local excision as a total biopsy
and found data similar to ours with actuarial is a realistic way to improve assessment of small
recurrence rates of 17% at ten years (26% cancer cancers. Our experience and literature emphasize
deaths) for Tl cancers and for T2 cancers 28% local that strict selection criteria should be used to
recurrence (25% cancer deaths) in patients treated accept this modality as definitive treatment.
without radiation. They found that 28% of their Evidence exists that definitive resection as initial
cancer deaths occurred beyond 5 years of follow- treatment is to be preferred over salvage resection
up. Irradiation did not significantly improve local at the time of recurrence [10 J.
recurrence or cancer specific survival with either The proof that lumpectomy and radiation is
Tl or T2 cancers [9J. The CALGB study published adequate treatment for breast cancer compared to
favorable results at 48 months with 3% local recur- radical and modified radical mastectomy (proven
rence rate and 3% cancer deaths for Tl cancers (all by clinical trial) has prompted the assumption
non-irradiated), 14% local recurrence and 12% that rectal lumpectomy is equivalent to radical
deaths for T2 cancers (all had chemoradiation), resection for certain rectal cancers (not proven
but long term follow-up has shown the same except by one small clinical trial with short fol-
delayed recurrence and deaths beyond five years low-up). Such acceptance oflocal therapy remains
for both groups, with six-year actuarial cancer free an assumption and deserves careful clinical
survival of 83% for Tl and 78% for T2 disease. scrutiny and further clinical trial.

References 6. Garcia-Aguilar J, Mellgren A, Sirivongs P, Buie D,

Madoff RD, Rothenberger DA. Local excision of rectal
1. Winde G, Nottberg H, Keller R, Schmid KW, Bunte H. cancer without adjuvant therapy: a word of caution.
Surgical cure for early rectal carcinomas (Tl). Ann Surg 2000;231:345-51
Transanal endoscopic microsurgery vs. anterior resec- 7. Mellgren A, Sirivongs P, Rothenberger DA, Madoff RD,
tion. Dis Colon Rectum 1996;39:969-76 Garcia-Aguilar J. Is local excision adequate therapy for
2. Whittle J, Steinberg EP, Anderson GF, Herbert R. Results early rectal cancer? Dis Colon Rectum 2000;43=1064-
of colectomy in elderly patients with colon cancer, based 71;discussion 1071-4
on Medicare claims data. Am J Surg 1992;163:572-6 8. Chakravarti A, Compton CC, Shellito PC, et al. Long-
3. Payne JE, Chapuis PH, Pheils MT. Surgery for large term follow-up of patients with rectal cancer managed
bowel cancer in people aged 75 years and older. Dis by local excision with and without adjuvant irradia-
Colon Rectum 1986;29:733-7 tion. Ann Surg 1999;230:49-54
4. Arnaud JP, Schloegel M, Ollier JC, Adloff M. Colorectal 9. Paty PB, Nash GM, Baron P, et al. Long-term results of
cancer in patients over 80 years of age. Dis Colon local excision for rectal cancer. Ann Surg 2002;236:522-
Rectum 1991;34:896-8 9;discussion 529-30
5. Finne CO. Local Treatment of Rectal Cancer. In: 10. Baron PL, Enker WE, Zakowski MF, Urmacher C.
Edelstein PS, ed. Colon and Rectal Cancer. New York: Immediate vs. salvage resection after local treatment
Wiley-Liss, 2000 for early rectal cancer. Dis Colon Rectum 1995;38:177-81
 VIII.3.
Transanal endoscopic microsurgery
G.A. Santoro, C. Pastore, G. Di Falco

Traditional transanal procedures are, mostly for laparoscopic techniques, specially designed
technical reasons, limited to tumors in the distal instruments (Fig. VIII.S) including forceps,
10 cm of the rectum. Large, flat villous tumors and curved scissors, curved tissue graspers (Fig.
those of the upper rectum and distal sigmoid are VIII.6), suction and irrigation instruments, elec-
virtually impossible to resect with the conven- trocautery, needle-tip high-frequency diathermy
tional trans anal approach. The introduction of knife (Fig. VIII.7), needle holders (Fig. VIII.8) and
transanal endoscopic microsurgery (TEM) by ultrasonically activated scalpel (Fig. VIII. 9) [2]
Buess et al. [1] in 1983 realized an interesting tech- can be introduced through individual gas-tight
nical advance in localized surgical treatment of ports in the faceplate of the rectoscope (Fig.
rectal tumors. It is an endoluminal, minimally VIII.1O, 11). A binocular stereoscopic optic gives a
invasive technique, which allows the peranal exci- clear three-dimensional view (Fig. VIII.12) or in
sion of rectal lesions located 4 to 18 cm from the alternative a 30 degrees monocular optical system
anal verge (Fig. VIII.2). Surgery is performed can be hooked up to a monitor for visualization
through a special operative rectoscope of 4 cm in (Fig. VIII.13). During the procedure a continuous
diameter and 12 to 20 cm long (Fig. VIII.3), which pressure-controlled (10-12 mmHg) insufflation of
is connected to the operating table via a support- carbon dioxide keeps the rectum open for expo-
ing arms system, which can be adjusted to any sure. The tip of the rectoscope is beveled down-
desired position (Fig. VIII.4). In the manner of ward. The patient's position for surgery therefore

Fig. VIII.2. Diagrammatic representation

of the operative system for trans anal
endoscopic microsurgery
166 Atlas of Endoanal and Endorectal Ultrasonography

depends on the anterior-posterior and lateral ori-

entation of the tumor (Fig. VIII.14). To optimize
access to the entire lesion, the rectoscope orienta-
tion must be changed frequently to compensate
for the limited operating field and length of the
surgical instruments. Distal rectal lesions just
above the dentate line are not appropriate for
TEM, because the gas seal may be easily lost,
impairing vision. Instrumentation is also more
difficult for distal lesions. Anteriorly placed
lesions above the peritoneal reflection should be
Fig. VIII.3. Operative rectoscope of 4 cm in diameter and 15 dissected carefully in the mucosectomy plane to
cm long for transanal endoscopic microsurgery (Karl Storz - prevent perforation of the rectal wall and avoid
Endoskope)
pneumoperitoneum.
This minimally invasive procedure has some
essential advantages: improved exposure of the
entire gas-fllled rectum, the superior optics and the
opportunity to address lesions in the upper rectum
[31. The disadvantages of TEM include costly equip-
ment, a steep learning curve and long operating
times combined with the relatively low incidence of
the pathologies for which this procedure is indicat-
ed [31. Complications are infrequent and include
bleeding, intraperitoneal perforation, fistula forma-
tion, incontinence, emboli and wound breakdown.
On the other hand postoperative recovery is rapid,
oral intake can be established early and most
patients are discharged within three to five days [41.
Prolonged anal stretching by 4 em diameter opera-
tive rectoscope induces few sphincter function
problems. Kennedy et al. [51 reported no significant
change in mean continence score, squeeze or cough
pressure, pudendal nerve terminal motor latency,
Fig. VIII.4. Supporting arms system to connect the recto- anal mucosal electrosensitivity or rectal balloon
scope to the operative table (Karl Storz - Endoskope)
study volumes after surgery. There was a reduction
in mean anal resting pressure with complete recov-
ery within a short-term period.

Fig. VIII.S. Instruments for transanal

endoscopic microsurgery (Karl Storz -
Endoskope)
 Section VIII' Surgical treatment options for rectal cancer 167

Fig. VIII.6. Forceps, scissors and tissue graspers for transanal Fig. VIII.7. Multifunctional instruments unit coagulation
endoscopic microsurgery (Karl Storz - Endoskope) with suction and irrigation (Karl Storz - Endoskope)

Fig. VIII.8. Needle holders for trans anal endoscopic micro-

surgery (Karl Storz - Endoskope)

Fig. VIII.9. Ultrasonically activated scalpel (Ultracision -

Ethicon Endo-Surgery)

The ideal indication for TEM are all types of

adenoma located 4 to 18 em from the anal verge
that cannot be treated by colonoscopy [6]. Various
excision methods are possible: mucosectomy, par-
tial thickness excision or full-thickness excision
(Figs. VIII.15-17). Demartines et al. [7] recom-
mended a full-thickness resection for large sessile
adenoma to decrease the risk of missing a small
rectal cancer, which is found in up to 20% of such
lesions. The size of adenoma proved to be the
most important influencing factor for the pres-
Fig. VIII. 10. Gas-tight ports in the faceplate of the rectoscope
(Karl Storz - Endoskope) ence of invasive carcinoma.
 168 Atlas of Endoanal and Endorectal Ultrasonography

a b
Fig. VIII.n. Instruments for transanal endoscopic microsurgery introduced through individual gas-tight ports of the recto-
scope (a-b) (Karl Storz - Endoskope)

Fig. VIII.12. Binocular stereoscopic device (Wolf, Germany) Fig. VIII. 13. Monocular 30 degrees optical system (Karl Storz
- Endoskope)

Local excision of early carcinoma of the rectum Incidences oflymph node metastases in early stage
is controversial. The primary factor limiting the rectal cancers staged by ERUS as uTI-slight, UTI-
effectiveness of local treatment is lymph node massive and uT2 were 0%, 26% and 36%, respec-
invasion. Careful patient selection is crucial to tively. Incidences oflymph node metastases in pTI-
TEM outcome. The extent of rectal cancer invasion slight, pTI-massive and pT2 cases were 0%, 22%
can be evaluated by digital examination, endorec- and 30%, respectively. This study confirmed that
tal ultrasonography (ERUS) (Fig. VIIL18), comput- ERUS is accurate for preoperative evaluation of
ed tomography (CT) and magnetic resonance transmural invasion depth for early stage rectal
(MR). The accuracy rates with ERUS for preopera- cancer. Patients with uTI-slight tumors are at low
tive staging of rectal cancer have been reported to risk for positive nodes and are good candidates for
be superior to those with other modalities. Akasu local excision. Those with uTI-massive or T2
et al. [8] evaluated the accuracy of ERUS for early lesions should be treated with radical operations
stage rectal cancer. Sensitivity, specificity and over- because of the high incidence of positive nodes. To
all accuracy rates for detection of slight submucos- date, in most published series, treatment of rectal
al invasion (uTI-slight tumor), massive submucos- cancer by TEM is generally accepted for Tllow risk
al invasion (uTI-massive tumor) and muscolaris cancer. The recurrence rate following this applica-
propria invasion (uT2) were 99%/74%/96%, tion lies between 4% and 8% compared with a local
98%/88%/97% and 97%/93%/96%, respectively. recurrence rate of up to 30% for TI high-risk can-
 Section VIII' Surgical treatment options for rectal cancer 169

cer (Table VIII.2) [9-16]. Mahmoud et al. from the

Division of Colorectal Surgery at the University of
Minnesota reported a study comprising 46
patients with rectal cancer treated by TEM and 43
patients with rectal cancer treated by trans anal
excision (TA). TEM reduced the incidence of posi-
tive margins. Recurrence rates were significantly
different in the two groups: 9% in the TEM group
vs. 33% in the TA group. In a series of 113 patients,
Buess et al. [12] found 5 residual tumors among a
subset of 39 patients (12.8%) who underwent an
a ~ __________ ~~~ ________________ ~
anterior resection immediately after TEM. In a
series of 12 cancers resected by TEM reported by
Demartines et al. [7] 4 required re-operation by
total mesorectal excision. Over the follow-up peri-
od of 30 months the recurrence rate of Tl tumors
was 8.3%. Winde et al. [15] published a prospective,
randomized study comprising 52 patients with Tl
tumors treated by TEM or anterior resection (AR).
There were no significant differences in group out-
come: the 5-year survival was 96%, the local recur-
rence rate was 4.1% for TEM and 0% for AR and
the metastases rate was 0% for TEM and 4.1% for
AR. Heintz et al. [9] reported similar results in a
study comprising 46 patients with low-risk Tl
tumors treated by TEM and 34 patients with low-
risk Tl tumors treated by anterior resection. There
were no significant differences in group outcome:
the 5-year survival was 79% in the TEM group vs.
81% in the AR group. Ziprin et al. [18] performed 19
TEM for rectal tumors. Recurrence rates for Tl and
T2 tumors were 0% and 43% respectively. They
concluded that the role of TEM in the surgical
management of rectal cancer should be limited to
those with early disease (Tl).
Lezoche et al. [19] achieved results similar to
those observed after conventional surgery in 35
patients with T2 rectal cancer who underwent
trans anal endoscopic microsurgical excision
combined with preoperative high-dose radiother-
apy (5.040 cGy divided over 5 weeks). Forty days
after completion of radiotherapy, the complete
full-thickness local excision of the rectal lesion
including adjacent perirectal fat was performed
by TEM. The median follow-up of the patients was
38 months. One local recurrence (2.85%) was
noted. The probability of surviving 96 months
after completion of treatment was 83%.
TEM is ideal for the elderly who are reluctant
c to undergo a major operation and for palliative
Fig. VIII. 14. The patient is placed in a lithotomy position for treatment at any tumor stage. In such cases,
posterior lesions (a), in a prone jack-knife position for lesions radio chemotherapy should precede palliative sur-
located anteriorly (b) or in a lateral position (c) gical treatment.
 170 Atlas of Endoanal and Endorectal Ultrasonography

a ~ ____________________________________ ~
~ __________________________________ ~ b

Fig. VIII.IS. Extent of excision is marked using needle-tip high-frequency diathermy knife (a, b), aiming for a margin of 5 mm
for benign lesions and for a margin of at least 1 em for tumors

Fig. VIII. 16. Submucosal dissection of the polyp using elec- Fig. VIII.I7. Defect is closed with continuous sutures
trocautery and forceps. A suction tube is placed close at hand

Table VIII.2. TEM in the management of malignant rectal tumors (modified from [4])

Authors N° of Stage Follow-up Complications Local Survival

patients Recurrences
Tl T2 T3

Heintz [9] 58 58 0 0 42.8-52 months 3% 10 70%

Lezoche [10] 34 10 19 5 35 months 16% 5·9 71%

Saclarides [11] 36 28 8 0 > 1 year Bleeding, 25

incontinence

Buess [12] 69 52 13 4 29 months 8% 5·8

Smith [13] 51 30 15 6 15% 25

Steele [14] 21 7 14 0 10

Winde [15] 24 5 years 8% 4. 2 96%

Stipa [16] 21 10 9 2 10.3 months 5·6% 4·7

 Section VIII' Surgical treatment options for rectal cancer 171

Fig. VIII.IS. Large villous adenoma in a

54Y male occupying 40% of the rectal cir-
cumference in left posterolateral aspect
and located at the level of the seminal vesi-
cles (a). The tumor is 4 cm at its greatest
diameter and about 3 cm in thickness.
Endoscopic view of a large polypoid lesion
a
I Submucosa I (b). Double contrast barium enema (c)

b c

References 4. Sengupta S, Tjandra J. Local excision of rectal can-

cer: what is the evidence? Dis Colon Rectum 2001;
1. Buess G, Theiss J, Hutterer F. Die transanale 44:1345-61
endoskopische Rektum-operation: Erprobung einer 5. Kennedy ML, Lubowski DZ, King DW. Transanal endo-
neuen Methode im Tierversuch. Leber Magen Darm scopic microsurgery excision: is anorectal function
1983il3:73-7 compromised? Dis Colon Rectum 2002;45:601-4
2. Langer C, Markus P, Liersch T, Fuzesi L, Becker H. 6. Saclarides TJ. Transanal endoscopic microsurgery.
Ultracision or high-frequency knife in trans anal endo- Surg Clin North Am 1997;77:229-39
scopic microsurgery (TEM)? Advantages of a new pro- 7. Demartines N, von Flue MO, Harder FH. Transanal
cedure. Surg Endosc 2001;15:513-7 endoscopic microsurgical excision of rectal tumors:
3. Mortensen N, Mayer J. Transanal endoscopic micro- indications and results. World J Surg 2001;25:870-5
surgery: a forgotten minimally invasive operation. Br J 8. Akasu T, Kondo H, Morioka Y, Sugihara K, Gotoda T,
Surg 1995;82:435-7 Fujita S, Muto T, Kakizoe T. Endorectal ultrasonogra-
172 Atlas of Endoanal and Endorectal Ultrasonography

phy and treatment of early stage rectal cancer. World J Initial experience from three centres in the United
Surg 2000;24:1061-8 Kingdom. Br J Surg 1996;83:207-10
9. Heintz A, Morschel M, Junginger T. Comparison of 15. Winde G, Nottberg H, Keller R, Schmid KW, Bunte H.
results after trans anal endoscopic microsurgery and Surgical cure for early rectal carcinomas (T1).
radical resection for T1 carcinoma of the rectum. Surg Transanal endoscopic microsurgery vs. anterior resec-
Endosc 1998;12:1145-8 tion. Dis Colon Rectum 1996;39:969-76
10. Lezoche E, Guerrieri M, Paganini AM, Feliciotti F. 16. Stipa S, Chiavellati L, Nicolanti V, Stipa F. Microscopic
Transanal endoscopic microsurgical excision of irradi- endoluminal tumorectomy. Dis Colon Rectum
ated and non irradiated rectal cancer. A 5-year experi- 1994;37:S81-5
ence. Surg Laparosc Endosc 1998;8:249-56 17. Mahmoud N, Madoff R, Rothenberger D, Finne C.
11. Saclarides TJ. Transanal endoscopic microsurgery: a Transanal endoscopic microsurgery (TEM) reduces
single surgeon's experience. Arch Surg 1998;133:595-8 the incidence of positive margins compared with
12. Buess G, Mentges B, Manncke K, Starlinger M, Becker trans anal excision for rectal tumors. Dis Colon
HD. Technique and results of trans anal endoscopic Rectum 2001;44:A27
microsurgery in early rectal cancer. Am J Surg 18. Ziprin P, Bethune R, Benoist S, Ridgway PF, Gould S,
1992;163:63-70 Darzi AW. The role of trans anal endoscopic micro-
13. Smith LE, Ko ST, Saclarides T, Caushaj P, Orkin BA, surgery in the management of malignant rectal
Khanduja KS. Transanal endoscopic microsurgery. tumors. Br J Surg 2002;89:11-2
Initial registry results. Dis Colon Rectum 1996;39:S79- 19. Lezoche E, Guerrieri M, Paganini AM, Feliciotti F.
84 Long-term results of patients with pT2 rectal cancer
14. Steele RJ, Hershman MJ, Mortensen NJ, Armitage NC, treated with radiotherapy and transanal endoscopic
Scholefield JH. Transanal endoscopic microsurgery. microsurgical excision. World J Surg 2002;26:1170-4
 Invited commentary
K.Miller

Minimally invasive surgical techniques have senting with Tl tumors well to moderately differ-
emerged as a prominent part of patient care in entiated. Recurrence rates after TEM ranges from
many diseases. Transanal Endoscopic Microsurgery 4% to 8% in low-risk cancer [4,5,141 and 30% in TI
(TEM) is a minimally invasive technique to remove high-risk cancer, respectively [91. However it is the
certain tumors from the rectum. This means that method of choice when local resection of rectal
patients may benefit by having smaller incisions cancer is indicated. TEM may be an alternative for
and a faster recovery. The operation is performed the treatment of stenoses after fistulae or anasto-
through a scope placed into the anal canal and was motic stenoses. Patients should be able to tolerate
pioneered by Gerhard Buess of Tubingen, Germany anesthesia and must be in reasonable health. The
in the early 1980'S [2,31. He designed, developed,
and first used TEM for selected mid-rectal and
upper rectal tumors. TEM was quickly taught to
other German surgeons, and is widely used in Table VIII.3. Indications for Transanal Endoscopic
Europe though rarely used in the u.s. Like other Microsurgery
minimally invasive surgical techniques, there are
certain criteria and indications. The technique is Indications in benign lesions:
not generally established yet, because of the neces- Lesions located 4-20 em from the anal verge
sary special instrumentation and tools, the unusual Any size (even circumferential)
technical aspects of the approach, and the stringent Adenomas, carcinoids, endometriosis, lipomas
patient selection criteria.
Indications in cancer:
Cancers located 4-20 em from the anal verge
Well to moderate differentiation (GI, GIl)
Indication and patient selection No vascular or lymphatic invasion
Mobile by digital examination
Although the selection criteria are very specific, the Less than 3 em
indications for TEM have been reported in several uTI (uT2) rectal cancers
large series. These indications are summarized in Palliation of bleeding or obstruction
Table VIII.3. The main indication consists of ade-
nomas located within 4 to 18 em from the anal Other indications and uses:
verge that cannot be treated by colonoscopy. A full Rectal prolapse
thickness resection is recommended to ensure an Re-biopsy of suspicious sites
appropriate margin of safety. Encapsulated cancers Strictures
in adenomas have been reported in up to 31% of Flap closure of high fistulas
cases (Winburn, 1998). TEM is also used for the Full thickness biopsies of the rectum
removal of malignant neoplasms in patients pre- (e.g. Hirschsprung's disease)
174 Atlas of Endoanal and Endorectal Ultrasonography

surgery is usually performed under general anes- The tumor should be exposed and the bound-
thesia but epidural or spinal anesthesia is possible. aries of resection with a distance of S mm to 20
Some patients who need open surgery but cannot mm must be reached with the instruments. The
tolerate a general anesthesia may be candidates for positioning of the patient depends upon the local-
TEM. Bowel preparation is similar as formal col- ization of the tumor. Right-handed surgeons start
orectal surgery and perioperative single short with grasping the tissue near the base of the tumor
antibiotic prophylaxis is recommended. Pre- on the right side (Fig. VIII.20). The mucosa is dis-
operative assessment to stage and grade the type sected and incised with electrocautery or with an
and severity of the tumor is crucial to success. ultrasonically activated scalpel (UltraCision -
Clinical examination with rectoscopy and a biopsy, ETHICON Endosurgery, Norderstedt, Germany) at
endorectal sonography using a 360-degree endo- a right angle into the perirectal fat tissue (Fig.
probe (7 MHz) and anal manometry is recom- VIII.21). Isolated vessels are visualized and selec-
mended. tively coagulated. The potential advantages of
ultrasound dissection using UltraCision, rather
than conventional electrosurgery are described by
Operating technique Langer et al. [13]. The advantages of ultrasound
dissection are magnified under the particular con-
and instrumentation ditions of minimally invasive endoscopic rectal
surgery by means of TEM. In principle, all the
The operative rectoscope (Wolf GmbH known risks associated with the application of
Knittlingen, Germany, or Storz Endoskope, electric current can be avoided by using ultra-
Tuttlingen, Germany) is inserted after gentle sound technology.
dilatation. Carbon dioxide is insufflated with a The region is closed by a running suture (Fig.
pressure of 4 to 10 mmHg, to enlarge the intrarec- VIII.22). The rectal wall is sutured with a 3-0
tal space and facilitate precise resection. absorbable monofilament suture that is fixed with
Laparoscopic surgical instruments (S mm) are silver clips (Wolf GmbH Knittlingen, Germany).
used although specially designed graspers, for- We recommend a running suture with 3-0 PDS*II
ceps, a suction probe, and a needle holder to work Lahodny Polydioxanon (7 em MICS4E or 10 em
through the scope are provided with the operative MICSSE) Suture (ETHICON Johnson & Johnson
rectoscope. The carbon dioxide flow is regulated Intl.). The suture is fixed with a PDS clip (Fig.
by the preadjusted pressure and is similar to the VIII.23). The suture is followed transversely until
high flow used during conventional laparoscopy complete closure of the rectal wall is achieved.
(Fig. VIII.19). Finally the specimen is affixed to a preparation

Fig. VIII.20. The dissection starts with grasping the tissue

near the base of the tumor on the right side. The mucosa is
dissected and incised with electrocautery or with UltraCision
(ETHICON Endosurgery, Norderstedt, Germany) at a right
Fig. VIII.19. The operative rectoscope and laparoscopic sur-
angle into the perirectal fat tissue
gical instruments
 Section VIII· Surgical treatment options for rectal cancer 175

plate to allow the pathologist a precise description

of the tumor free resection margin.
Cicatricial stenosis in anorectal surgery is dif-
ficult to manage. We employed a Y-V rectoplasty
pedicle flap and obtained satisfactory to excellent
results with rEM.
Patients are allowed to walk and sit up after
anesthesia. A liquid diet is maintained for 24
hours and a fiber free diet is recommended for 14
days. Patients are discharged on postoperative day
2 to 4. Follow-up every 3 months during the first
year, every 6 months during the second year and
yearly thereafter is recommended to detect cancer
Fig. VIII.21. Complete resection with tumor free resection recurrences.
margin

Complications

Overall, the rate of local complications associated

with transanal resection range between 4% and
8% of cases while complications with traditional
surgery range in literature between 14% and 21%
[14, 21]. Perioperative and late complications in
our series ranged from 7% and S%, respectively
(Table VIII.4). Fatal cases are rarely reported and
one out of 3000 cases due to retroperitoneal
phlegmon leading to septic shock and death is
described in literature [s]. Prolonged anal dilata-
tion of 4 cm (concentric) induces few sphincter
function defects. Grade II incontinence ranges
from 4% in our series to IS% in literature [S, 10],
Fig. VIII.22. Closure of the defect by means of a running with full postoperative recovery after 3 months
suture demonstrated with manometric analysis [1].

Discussion and conclusion

The technique of trans anal endoscopic rectal

operation is designed for the removal of broad
based, sessile adenomas. More than two-thirds of
colonoscopically found polyps are located in the
rectum and lower sigmoid colon [6]. Sessile
polyps generally show a greater tendency of
becoming malignant than pedunculated polyps
[IS]. Both the incidence of adenomas located in
the lower part of the rectum and the high proba-
bility that they become malignant make complete
resection of these tumors necessary. Broad-based
adenomas should be removed to provide a total
Fig. VIII.23. Transverse running suture with 3-0 PDS*II
biopsy, since forcep and snare biopsies obtained
Lahodny Polydioxanon (ETHICON Johnson & Johnson lnt!.).
The suture is fIxed with a PDS clip do not provide representative information as to
176 Atlas of Endoanal and Endorectal Ultrasonography

Table VIII.4. Complications and follow-up after TEM their biological significance [11]. Indication for
(Landesklinik fuer Chirurgie Salzburg and Hallein, local excision of carcinomas is based on the rec-
Ludwig Boltzmann Institut fuer Gastroenterologie und ommendations of Hermanek and Gall [11] who
experimentelle Chirurgie, Jan. 1990 - Jan 2002) consider local excision of pTI carcinomas to be
sufficient when patients present synchronous col-
TEM n= 85 orectal cancer, including Tl rectal cancer. Local
excision of the rectal lesion via TEM can help to
Total Perioperative Complications 6 (7%) improve the patient's quality oflife without affect-
ing the curability of the disease. Complication
Bleeding 2 (2%) rates must be compared to those of Mason and
Anastomotic leak 3 (3.5%) Kraske (cited in Schildberg's paper [16]), because
Phlegmon 1 (1%) when a transanal procedure is used, the majority
Reoperation 4 (5%) of polyps will be located in a too high position to
Conversion to laparotomy 2 (2%) be reached. According to these procedures the
complication rate is up to 31% [7,8]. Complication
Total Late Complications 4 (5%) and local recurrence rates are obviously higher
with conventional local transanal procedures [7-
Fecal incontinence Grade III 3 (3.5%) 9, 16]. Winde et al. demonstrated similar local
Rectal stenosis 1 (1%) recurrence and survival rates in a prospective
randomized study, TEM vs. conventional anterior
Local recurrence of Carcinoma* resection of the rectum, with lower morbidity
(Follow-up mean 72 months, range 12-144 months) favoring the TEM technique [21]. Comparable
results in survival rate to the gold standard ante-
Total rior resection are objective arguments for choos-
ing the adequate surgical procedure. For early rec-
pT2 GIl 2 (6%) tal cancer, the minimally invasive TEM technique
pT1 GIl 1 (3%) should be preferred because of superior overview
during operation with safer suturing after metic-
5 Year Survival 100% ulous full-thickness excision.
The possibility of performing wide sphincter
* Patients with Carcinoma total n=31 dilatation which, postoperatively, leads to com-
plete incontinence, makes the rectal cavity, other-
wise difficult to access, an ideal region for endo-
scopic operations. Surgical procedures can be car-
ried out without the need to divide healthy struc-
tures. The advantage is rapid discharge of the
patients from the hospital, and less complications
in the postoperative course. These major advan-
tages of the TEM technique for the patient are
coupled with the effect of the saving of costs by
reducing the period of hospitalization and time
off work.
Training, in common with other endoscopic
operative procedures, must be given within the
framework of a training program. We therefore
Fig. VIII.24. Video-supported training program of the TEM offer video-supported training courses (Fig.
technique VIII.24) (www.miller.co.at/TEM).
 Section VIII· Surgical treatment options for rectal cancer 177

References carcinoma. Pathology, diagnosis, surgical treatment.

Int J Colorect Dis 1986;1: 79-84
1. Banerjee AK, Jehle EC, Kreis ME, Schott UG, Claussen 12. Ikeda Y, Koyanagi N, Mori M, Akahoshi K, Ueyama T,
CD, Becker HD et al. Prospective study of the procto- Sugimachi K. Transanal endoscopic microsurgery for
graphic and functional consequences of transanal T1 rectal cancer in patients with synchronous colorec-
endoscopic microsurgery. Br J Surg 1996;83:211-3 tal cancer. Surg Endosc 1999;7:710-2
2. Buess G, Kipfmueller K, Hack D, Gruessner R, Heintz 13. Langer C, Markus P, Liersch T, Fuzesi L, Becker H.
A, Junginger T. Technique of transanal endoscopic UltraCision or high-frequency knife in transanal
microsurgery. Surg Endosc 1988;2:71-5 endoscopic microsurgery (TEM)? Advantages of a new
3. Buess G. Review: transanal endoscopic microsurgery procedure. Surg Endosc 2001;5: 513-7
(TEM). J R Coll Surg Edinb 1993;4: 239-45 14. Mentges B, Buess G, Effinger G, Manncke K, Becker
4. de Graaf EJ, Doornebosch PG, Stassen LP, Debets JM, HD. Indications and results of local treatment of rectal
Tetteroo GW. Transanal endoscopic microsurgery for cancer. Br J Surg 1997;84: 348-51
rectal cancer. Eur J Cancer. 2002;7: 904-10 15. Morson BC, Sobin LH Histological typing of intestinal
5. Demartines N, von Flue MO, Harder FH. Transanal tumors. International histological classification of
endoscopic microsurgical excision of rectal tumors: tumors, WHO, Geneva 1976 vol 15: 56
indications and results. World J Surg 2001;25:870-5 16. Schildberg FW, Wenk H. Der posteriore Zugang zum
6. Frimberger E, Kuhner W, Seib HJ, Ottenjann R. Rektum. Chirurg 1986;57: 779-91
Kolorektale Adenome. Dtsch Med Wochenschr 17. Sitzler PJ, Seow-Choen F, Ho YH, Leong AP. Lymph node
1978;103: 649-52 involvement and tumor depth in rectal cancers: an anal-
7. Grigg M, McDermott FT, Phil EA, Hughes ES. Curative ysis of 805 patients. Dis Colon Rectum 1997;40: 1472-6
local excision in the treatment of carcinoma of the rec- 18. Warneke J, Petrelli NJ, Herrera L. Local recurrence
tum. Dis Colon Rectum 1984;27:81 after sphincter-saving resection for rectal adenocarci-
8. Haring R, Karavias T, Konradt J. Die posteriore noma. Am J Surg 1989;158: 3-5
Proctorectotomie. Chirurg 1978;49: 265-71 19. Weber TK, Petrelli NJ. Local excision for rectal cancer:
9. Heintz A, Morschel M, Junginger T. Comparison of an uncertain future. Oncology (Huntingt) 1998;12:933-
results after trans anal endoscopic microsurgery and 43;discussion 944-7
radical resection for T1 carcinoma of the rectum. Surg 20. Winburn GB. Surgical resection of villous adenomas
Endosc 1998;12: 1145-8 of the rectum. Am Surg 1998;64: 1170-3
10. Hemingway D, Flett M, McKee RF, Finlay IG. Sphincter 21. Winde G, Nottberg H, Keller R, Schmid KW, Bunte H.
function after transanal endoscopic microsurgical Surgical cure for early rectal carcinomas (T1).
excision of rectal tumors. Br J Surg 1996;83: 51-2 Transanal endoscopic microsurgery vs. anterior resec-
11. Hermanek P, Gall FP. Early (microinvasive) colorectal tion. Dis Colon Rectum 1996;39: 969-76
 VIII.4.
Conventional techniques

G.A. Santoro, G. Di Falco

A number of surgical options are available for the from the anal margin. Tumors of the high and
management of rectal cancer. The location of the midrectum (10 em or higher from the anal verge)
primary tumor is probably the most important are treated by radical distal sigmoid and rectal
variable in the choice of operation, usually deter- resection with an adequate distal margin and
mining the extent of rectal resection. Measurement anastomosis of the descending colon to the residu-
of the distance between the anal verge and the allow rectum (Fig. VIII.25). In patients who have
lower border of the tumor are generally made by tumors between 7 to 10 em above the anus sphinc-
preoperative rigid proctosigmoidoscopy. However ter preservation is often technically possible by
rectal mobilization during pelvic dissection will performing a very low anterior resection or a
allow the rectum to be risen up out of the pelvis colo anal anastomosis (Fig. VIII.26). The use of gas-
significantly, increasing the distance of the tumor trointestinal stapling devices has facilitated the

Fig. VIII.2S. Carcinoma of the upper rectum is appropriately Fig. VIII.26. Carcinoma between 7-10 cm from the anal mar-
treated by radical sigmoid colectomy and resection of adequate gin are best treated by low anterior resection or colo anal
upper rectum to gain a wide clear margin below the tumor anastomosis
 Section VIII' Surgical treatment options for rectal cancer 179

100
lower third of the rectum (at or below 6 em from
90 the anus) are best treated by abdominoperineal
80 resection that includes the distal sigmoid colon,
70
the entire rectum, the anus, the sphincteric muscu-
",,80 lature and associated lymph nodes (Fig. VIII.28). A
50 AR
40 permanent end colostomy is always necessary.
30 -----~~-----_/ In addition to surgical considerations, the sur-
20 eM geon must consider quality of life issues to
10 LE improve the overall well being of the patient. This
a~nnn~~~nnn80~~~ME~ chapter will focus on the recent developments in
r.ars treating patients with rectal cancer, along with
controversial aspects.
Fig. VIII.27. Surgical options for the management of rectal
cancer. Until 1960 sphincter preserving procedures were con-
sidered appropriate for only 30-50% of patients. These tech-
niques are now the accepted operation for 80-90% of all
patients with rectal tumor (AR= anterior resection; APE=
Inferior mesenteric artery ligation
abdominoperineal excision; CAA= colo anal anastomosis;
LE= local excision)
There is some controversy, in treating carcinoma
of the rectum, over the optimal point for dividing
the lymphovascular bundle (Fig. VIII.29). The
performance of such anastomosis. Until 1960 highest point to which it is practical to extend the
sphincter-preserving procedures were considered lymph node dissection is the junction of the infe-
appropriate for only 30-50% of patients. These rior mesenteric artery and the aorta. In literature it
techniques are now the accepted norm for 80-90% was reported that a significant number of patients
of all patients with rectal cancer (Fig. VIII.27). (7-11%) has lymph node metastases between the
Provided that a clear distal margin of rectum is point of ligation of the artery at the aorta and the
resected, five-year survival and recurrence rates level of its left colic branch [1]. However Pezim and
are similar between abdominoperineal resection Nicholls [2] and Surtees et al. [3] have failed to
and sphincter-saving operations. Tumors of the show a significant improvement in terms of local

LtllcoHc ""try

Mlddlt
htmonholdol orttry

InI..to.
htmormoklol t"try

Fig. VIII.2B. Abdominoperineal resection is appropriate for Fig. VIII.29. Division of the inferior mesenteric artery at the
low rectal cancers that do not allow adequate length of resid- junction with the aorta (a) or just distal to the take-off of the
ual distal rectum for anastomosis after removal of the tumor left colic artery (b)
180 Atlas of Endoanal and Endorectal Ultrasonography

recurrence or survival with high ligation instead of Mesorectal excision

ligation of the inferior mesenteric artery just distal
to the left colic artery. Slanetz and Grimson [4]
reviewed the results of a series of 2409 patients Mesorectal involvement in rectal tumors by both
who underwent curative resections for rectal can- direct infiltration and diffusion to lymph nodes is
cer. High ligation was able to provide lower rates of related to a higher incidence of local recurrences.
local recurrence and improved long term survival In 1982, Heald et al. [10] popularized the total
as compared to intermediate levels of ligation. mesorectal excision (TME) to prevent local recur-
Conventional low ligation may be sufficient in the rences in all rectal cancers. The principle of this
treatment of Dukes' A and B lesions. High ligation approach is to dissect the rectum and what Heald
is of significant value to the patients with a calls the "oncologically dangerous" tissue of the
Dukes'C tumor with nodal metastases at a level mesorectum, made up of the terminal branches of
proximal to the left colic artery but not beyond the the inferior mesenteric artery, the associated veins
origin of the inferior mesenteric artery. High liga- and lymphatics and a variable pad of supporting
tion has not value if proximal lymphatic spread perirectal fat, intact surrounded by its fascia (Fig.
beyond the origin of the inferior mesenteric artery VIII.30). The perirectal fascia was originally rec-
has occurred. From a technical standpoint high ognized by Thoma Jonnesco as a serofibrous
ligation allows complete mobilization of the left envelope to provide fixity to the rectum [11].
colon for a tension-free anastomosis, while con- Gerota also demonstrated that the lymphatics of
ventional low ligation may preserve those auto- the rectum were contained within this envelope
nomic nerves, which split around the origin of the [11]. Later the perirectal fascia was described by
inferior mesenteric artery. Waldeyer under the name "fascia propria recti"
[11]. Chapuis et al. [11] have recently provided a
detailed description on the anatomic concepts for
Distal margins correct mobilization of the rectum without
breaching the perirectal fascia. This can be
achieved by identifying the correct plane between
What constitutes a reliably safe distal margin has the rectum covered by its fascia and the sacrum
been under much debate. The S cm rule of distal covered by the parietal pelvic fascia. The retrorec-
margin was standard practice for many years. tal space between these two fascial layers is an
Based on a better understanding of distal intra- avascular space of loose areolar tissue. Heald [12]
mural spread, several reports have suggested that coined the term "holy plane" to emphasize how
the traditionals cm distal margin is not necessary. proceeding along this plane using either sharp or
Pollett and Nicholls [5] demonstrated that a distal diathermy dissection will avoid hypogastric
margin of less than 2 cm for rectal cancers does nerves injury, ensuring the complete excision of
not adversely affect local recurrence or survival. all fat enclosed within the perirectal fascia (Fig.
Wilson and Beahrs [6] reported that a 2 cm distal VIII.31). Despite Heald's report [10] concentrated
margin appears adequate to encompass any distal on the importance of TME in all rectal cancers, it
intramural lymphatic spread. Heimann et al. [7] has subsequently become clear that the complete
found a high recurrence rate (36%) when distal removal of the distal mesorectum is not always
margins of </= 1 cm were compared to distal mar- necessary for upper rectal cancers [13]. For
gins >2 cm. Shirouzu et al. [8] found a distal patients with tumors above the peritoneal reflec-
spread in only 3.8% of patients who underwent tion the appropriate procedure is high anterior
curative resection for rectal cancer. Distal spread resection and mesorectal division S cm below the
was seen in patients with advanced disease and lower edge of the cancer, preserving the laterallig-
was confined to within a 1 cm length. The authors aments, to ensure that any tumorous lymphatic
suggested that a 1 cm distal margin may be ade- invasion close to the primary lesion is removed.
quate for most rectal tumors. However Tonak et al. The mesorectum should be removed as part of the
[9] reported a local recurrence rate of 33% for dis- proper clearance of all mid and low rectal tumors.
tal margins <3 cm and only 13% for margins >3 TME-based operations in comparison with
cm. Despite this report, a 2 cm distal margin seems conventionally practiced blunt surgery for rectal
adequate for most rectal carcinomas, especially cancer are associated with significantly lower
when it may allow a sphincter-saving operation. rates oflocal recurrences (4-8% vs. 20-40%), a sig-
 Section VIII- Surgical treatment options for rectal cancer 181

Fig. VIII.30. Diagrammatic representa-

tion of a total meso rectal excision for rec-
tal cancer by removing the distal-most
portion of the mesorectum

nificantly higher rate of survival (45-50% vs. 75%) disease-free survival was 78% in 135 patients who
and significantly lower long-term morbidity underwent curative resection of rectal carcinoma
(higher rates of both sphincter and autonomic according to the principles of TME. Enker et al.
nerve preservation, thus preserving sexual and [21] observed a 5-year survival rate of 74.2% and a
urinary function in most patients) (Table VIII.5) recurrence rate at 5 years of 7.3% on 246 patients
[7,9,10,14-23]. Arbman et al. [17] compared the 5- who underwent TME-based operations. Recently,
year survival rate and the incidence of local recur- Ratto et al. [24] evaluated if mesorectal neoplastic
rences in two groups of over 200 patients before microfoci not connected to the tumor mass may
and after the introduction of TME. Local recur- affect the prognosis of patients with rectal cancer.
rence developed in 19% and 6%, respectively and Microscopic foci were found in 44.2% of 77
the 5-year survival improved from 50% to 70% patients with extraperitoneal rectal cancer resect-
with TME. MacFarlane et al. [22] reported that the ed with TME. The incidence of neoplastic foci was
recurrence rate at 5 years was 4% and the 5-year high even in early staged tumors (18.8% of Stage I
tumors, 46.9% of Stage II tumors and 59.3% of
Stage III tumors) and despite aggressive preoper-
ative treatment. Microscopic foci significantly
affected overall survival rate (43.4% in patients
with mesorectal microfoci compared to 63.3% in
patients without deposits, P= 0.016). In the stud-
ies by Reynolds et al. [25], Ueno and Mochizuki
[26] and by Ono et al. [27] mesorectal foci were
found in 34%,35.2% and 20% of patients respec-
tively, further confirming the importance of a cor-
rect total mesorectal excision.
Maeda et al. [28] evaluated if the total mesorec-
tal excision combined with the pelvic autonomic
Fig. VIII.31. The typical smooth, bilobed, encapsulated
nerve preservation may interfere with the opera-
appearance of the specimen produced following low anterior
resection with total meso rectal excision tive radicality. Although nodes along autonomic
182 Atlas of Endoanal and Endorectal Ultrasonography

Table VIII.S. Comparison between conventional surgery and TME-based operations in terms of local recurrence
rate and 5-year survival

Author N° Total Local 5-year

of patients mesorectal recurrence survival
excision rate

Tonak [9] 248 21%

Phillips [14] 1988 18% 41%
Heimann [7] 329 16%
Holm [15] 347 13%
Fegiz [16] 131 28·9% 45%
Arbman [17] 553 19% 5°%
250 + 6% 7°%
Heald [10] 135 + 3·5% 74·2%
McCall [18] 1033 + 7·1%
Moriya [19] 306 + 6.2%
Aitken [20] 103 + 0%
Enker [21] 118 + 7·3% 74.2%
MacFarlane [22] 135 + 5% 78%
Tocchi [23] 53 + 9% 75%

nerves were identified in 94% of cases, the inci- Enker et al. [29] have examined the functional
dence of node metastases was minimal (4%). outcomes related to TME and autonomic nerve
Nodal involvement occurred in patients with local- preservation for low rectal cancer (Fig. VIII.32).
ly advanced lower rectal tumors who had poor Sexual function was maintained in 86% of
prognosis in spite of combined resection of auto- patients after low anterior resection and in 57% of
nomic nerves. Therefore these authors considered patients after abdominoperineal resection of the
that "doing TME with autonomic nerve preserva- rectum. No significant urinary morbidity was
tion does not imply a less radical surgery for rectal encountered.
carcinoma from the point oflymphatic spread".

Fig. VIII.32. Pelvic autonomic nerve

anatomy. Sympathetic (hypogastric) and
parasympathetic (sacral) branches and
plexus (From: Guillem JG. Ultra-low ante-
rior resection and coloanal pouch recon-
struction for carcinoma of the distal rec-
tum. World J Surg 1997;21:721-7, with per-
mission)
 Section VIII' Surgical treatment options for rectal cancer 183

Extended pelvic lymphadenectomy increased to 8.8% for T3 and 31% for T4 rectal
tumors. The level of the tumor greatly influenced
the incidence of lateral node metastases. For
The lymphatic spread of rectal cancer consists in tumors with a lower margin above 6 em from the
both an upward spread (pararectal, superior rec- dentate line, lateral node involvement was negligi-
tal and inferior mesenteric lymph nodes) and a ble (0.6%). Once the lower margin of the tumor
lateral spread (middle rectal, obturator and inter- was below 5 em above the dentate line, lateral
nal iliac lymph nodes) (Fig. VIII.33). For tumors node involvement was of clinical importance
located above the peritoneal reflection, lymphatic (7.5% for tumors between 4.1-5 em, 16.?,Yo for
spread can occur only in the direction of upward tumors between 2.1-3 em and 29.6% for tumors
mesenteric spread. When the tumor is located at between 0.1-1 em).
or below the peritoneal reflection, lymphatic The lateral ligaments playa critical role for the
spread can occur both in the direction of upward lateral lymphatic flow [32]. Lymphatic vessels pass
mesenteric spread and lateral extramesenteric from the lower rectum toward the iliac lymph
spread. The incidence of lateral lymph node nodes through the lateral ligament, around the
metastases is reported at 9-18%. Moriya et al. [30] middle rectal artery.
found that the incidence of upward lymphatic These studies suggest that low rectal cancer
spread was 94% and the incidence of laterallym- (less than 5 em from the dentate line) spreads
phatic spread was 27% in 218 patients with widely beyond the scope of TME. The rationale for
Dukes'C disease. The incidence of lateral spread radical pelvic lymphadenectomy in rectal cancer
increased in proportion to deepening wall inva- is to achieve a complete clearance of laterallym-
sion, amounting to 14% in patients with T2 phatic spread in order to improve survival and
tumors, to 30% in patients with T3 tumors and to reduce local recurrence. In a retrospective study
40% in those with T4. Takahashi et al. [31] retro- on 437 patients with carcinoma of the middle and
spectively analyzed a series of 764 patients who lower rectum who received radical lateral node
underwent curative resection for rectal cancer. dissection, Hojo et al. [33] reported an overall
Very low incidences of lateral node involvement pelvic recurrence rate of 12.4%. By contrast, Glass
were found in Tl (2.8%) and T2 (3.4%) rectal et al. [34] found no improvement in 5-year sur-
tumors. Once the tumor penetrated the rectal vival rates in 75 patients who underwent extended
wall, the incidence of lateral node involvement pelvic lymphadenectomy, compared to the results
of conventional operations. Takahashi et al. [31]
reported that for all cases proved to have lateral
node involvement the 5-year survival rate was
42-4% in contrast with 90.1% for negative node
involvement. Local recurrence rate for patients
with negative nodes was 1.9% and 21.6% for posi-
tive lateral nodes.
Due to the unacceptable incidence of urinary
(difficulty in voiding urine) and sexual (loss of
erection and ejaculation) dysfunction, an auto-
nomic nerve-sparing surgery has been developed.
Moriya et al. [30] reported promising results using
nerve-sparing operations with lateral node dis-
section in 448 patients with advanced lower rectal
cancer. The overall 5-year survival was 70% and
the overall incidence of local recurrence was 9.3%.
In 84% of patients an acceptable level of urinary
Middle rectal function was preserved.
lymph node. To date the appropriate use of lateral lymph
node dissections remains to be defined.
inguinal lymph nod..

Fig. VIII.33. Lymphatic drainage from the rectum

184 Atlas of Endoanal and Endorectal Ultrasonography

Sentinel lymph node mapping function may be compromised (Table VIII.6) [37-
41]. Incontinence rates of 5-20% have been
reported in different series following straight
Accurate staging of rectal cancer directly impacts coloanal anastomosis [42]. Parks and Percy [40]
treatment and prognosis. To adequately stage rec- reported that 48% of patients experienced urgen-
tal cancer approximately 13 lymph nodes should cy of defecation or increased frequency (>3 per
be examined. When fewer than 13 lymph nodes are day) after resection and straight colo anal anasto-
examined, there is a significant chance of under- mosis for rectal carcinoma. A frequency of at least
staging a rectal cancer [35]. Esser et al. [36] per- three motionsi24h persisted for more than a year
formed a prospective evaluation of sentinel lymph postoperatively in 21% of 48 patients from St.
node mapping in colorectal cancer patients to Mark's Hospital [39] and in 66% of 36 patients
identify lymph nodes most likely to contain tumor from the Hospital Purpan in Toulouse [37].
metastases. During surgery Lymphazurin blue dye Keighley and Mathenson [41] noted in 12 patients
was injected into the rectal wall around the tumor treated by rectal excision and straight colo anal
site. Sensitivity of 67%, specificity and positive anastomosis no significant differences between
predictive value of 100% and negative predictive preoperative and postoperative mean resting
value of 94% were found when sentinellypmh pressure (76±24 cmH20 and 62±24 cmH20,
nodes were identified. A controversial aspect of respectively) and mean maximum squeeze pres-
sentinel lymph node technique is the lack of sure (122±39 cmH20 and 123±41 cmH20, respec-
impact on surgical technique, in contrast to breast tively). However the mean rectal capacity before
and melanoma surgery, where sentinel lymph and after operation was significantly reduced
node biopsy changes the extent of surgery that is (268±24 ml vs. 123±41 ml; P<O.OOI) and urgency
performed. Two additional controversial aspects of defecation was reported in 80% of patients.
are the biologic significance of micrometastic dis- Suzuki et al. [43] reported comparable results in a
ease and presence of skip metastases. A larger series of 16 sphincter-saving resections. Function
clinical trial will be needed to confirm that lym- of the internal and external anal sphincter
phatic mapping techniques helps prevent under- appeared uninfluenced by the operation. However
staging cancer and allows more precise stage of significantly lower rectal compliance was found
disease. in resected patients than in normal subjects (3.33
mllcmH20 vs. 15.65 mllcmH20; P<O.OOI).
Frequent bowel movements and occasional soil-
Formation of a neorectum ing were experienced by most patients in the early
postoperative period. Williams et al. [44] report-
ed a mean maximal tolerated volume of 206ml
Most patients with tumors in the lower two-thirds after 20 straight colo anal anastomosis compared
of the rectum can now be treated by sphincter- to 350 ml in normal subjects (P<O.OI). Most
saving surgery. Although a permanent stoma is patients had more frequent bowel actions than
avoided and the operation is oncologically sound, the controls.

Table VIII.6. Functional results after straight colo anal anastomosis. Reported experience

Author N° Full Leakage Frequency

of patients continence Soiling >3124h
Incontinence Urgency

Lazorthes [37] 45 78% 22% 66%

Nicholls [38] 15 70% 30% 60%
Sweeney [39]
48 72% 28% 21%
Parks [40] 70 90% 10% 48%
Keighley [41] 10 99% 20% 80%
 Section VIII' Surgical treatment options for rectal cancer 185

a b
Fig. VIII.34. J-colonic pouch with colo anal anastomosis (a). The diagram of sagittal section shows the reservoir in the pelvis (b)

These studies suggest that there is a close cor- tal resection and colonic pouch reconstruction.
relation between control of bowel movements and Lazorthes et al. [37] performed 20 colonic pouch
rectal compliance after straight colo anal anasto- and 45 straight colo anal anastomoses. After one
mosis. Surgical removal of the rectum leaves the year 86% of the patients with a reservoir had one
patient with a high pressure, highly motile, rela- or two bowel movements per day compared with
tively non-compliant sigmoid or descending 33% of the patients without a reservoir (p<O.Ol).
colon anastomosed to the anal canal. A situation The frequency of bowel movements was inversely
where neorectal motility will overcome the anal related to the maximal tolerable volume
sphincter and will cause leakage, soiling or incon- (P<O.OOl). In a study from the Royal Infirmary of
tinence. However, within 6 months to a year fol- Edinburgh, Santoro et al. [46] achieved a satisfac-
lowing surgery most patients develop quite satis- tory fecal continence with a stool frequency of 1-2
factory function in evacuating the rectum and in per day in 85.7% of patients who underwent
controls of flatus or liquid stool. colo anal anastomosis with J-colonic pouch for
Over the last 15 years, surgical developments rectal cancer. Functional results did not appear to
have aimed at improving function after restora- depend on the presence of a diverting stoma and
tion of bowel continuity essentially by creating a were not influenced by the technique used (sta-
neorectum to increase the capacity and compli- pled or hand-sewn anastomosis).
ance in the pulled-through segment. The best- Although a J-pouch improves some aspects of
known and most widely practiced operation function, the results are not perfect [47, 48]. The
involves formation of a colonic J-pouch (Fig. use of a reservoir may lead to difficulty in evacu-
VIII.34). Clinical trials have shown a significant ation. Emptying frequency is related to pouch vol-
advantage of colonic pouch anal-anastomosis. ume, which increases with time. Parc et al. [45]
Nicholls et al. [38] compared 13 patients who used 12 cm limbs of colon and 25% of 24 patients
underwent rectal excision with colonic pouch treated had difficulty in securing spontaneous
reconstruction and 15 patients who underwent defecation and had to resort to enemata or sup-
rectal excision and straight colo anal anastomosis. positories to empty their reservoir. The same
All patients with a reservoir had stool frequency complications have been reported by Lazorthes et
<2 per day at only a mean of 7 months post-oper- al. [37] and Williams et al. [44] who used 12 cm
atively, while 40% of those without a reservoir had and 10 cm limbs of colon, respectively. Dynamic
a stool frequency >2 per day. Patients with pouch proctography showed a large volume pouch with
significantly increased rectal sensitivity threshold an anal-pouch angle so acute that evacuation was
volume and maximal tolerable volume (83 ml and impossible. Santoro et al. [46] used 7 cm limbs of
317 ml, respectively) compared with patient with- colon. Postoperatively the anal-pouch angle
out reservoir (52 ml and 174 ml, respectively). Parc appeared acute during squeezing and widened
et al. [45] achieved full continence during day and during straining (Fig. VIII.35). Obstructed defeca-
night, with a mean number of 1.1 bowel move- tion was not a problem and 92.8% of patients
ments per day in 24 patients who underwent rec- referred to empty completely their pouch.
 186 Atlas of Endoanal and Endorectal Ultrasonography

a b

Fig. VIII.3S. Dynamic proctography in a patient who under-

went coloanal anastomosis with J-colonic pouch for rectal
cancer. At rest (a) the anal-pouch angle is 100°. Pouch capaci-
ty is 150ml. During squeezing (b) the anal-pouch angle
becomes acute (85°) and during straining (c) the anal-pouch
c angle widens to 140°

4. Slanetz CA Jr, Grimson R. Effect of high and interme-

References diate ligation on survival and recurrence rates follow-
ing curative resection of colorectal cancer. Dis Colon
1. McElwain JW, Bacon HE, Trimpi HD. Lymph node Rectum 1997;40:1205-18
metastases: experience with aortic ligation of inferior 5. Pollett WG, Nicholls RJ. The relationship between the
mesenteric artery in cancer of the rectum. Surgery extent of distal clearance and survival and local recur-
1954;35:513-31 rence rates after curative anterior resection for carci-
2. Pezim ME, Nicholls RJ. Survival after high or low liga- noma of the rectum. Ann Surg 1983;198:159-63
tion of the inferior mesenteric artery during curative 6. Wilson SM, Beahrs OH. The curative treatment of car-
surgery for rectal cancer. Ann Surg 1984;200:729-33 cinoma of the sigmoid, rectosigmoid, and rectum. Ann
3. Surtees P, Ritchie JK, Phillips RKS. High versus low lig- Surg 1976;183:556-65
ation of the inferior mesenteric artery in rectal cancer. 7. Heimann TM, Oh C, Steinhagen RM, Greenstein AJ,
Br J Surg 1990;77:618-21 Perez C, Aufses AH. Surgical treatment of tumors of
 Section VIII' Surgical treatment options for rectal cancer 187

the distal rectum with sphincter preservation. Ann the prognosis of patients with rectal cancer. Dis Col
Surg 1992;216:432-6 Rectum 2002;45:733-43
8. Shirouzu K, Isomotu H, Kakegawa T. Distal spread of 25. Reynolds JV, Joyce WP, Dolan J, Sheahan K, Hyland JM.
rectal cancer and optimal distal margin of resection Pathological evidence in support of total meso rectal
for sphincter-preserving surgery. Cancer 1995;76:388- excision in the management of rectal cancer. Br J Surg
92 1996;83:1112-5
9. Tonak J, Gall FP, Hermanek P, Hager TH. Incidence of 26. Ueno H, Mochizuki H. Clinical significance of extra-
local recurrence after curative operations for cancer of bowel skipped cancer infiltration in rectal cancer. Surg
the rectum. Aust N Z J Surg 1982;52:23-7 Today 1997;27:617-22
10. Heald RJ, Husband EM, Ryall RD. The mesorectum in 27. Ono C, Yoshinaga K, Enomoto M, Sugihara K.
rectal cancer surgery - the clue to pelvic recurrence? Discontinuous rectal cancer spread in the mesorectum
Br J Surg 1982;69:613-6 and the optimal distal clearance margin in situ. Dis
11. Chapuis P, Bokey L, Fahrer M, Sinclair G, Bogduk N. Col Rectum 2002;45:744-9
Mobilization of the rectum. Anatomic concepts and 28. Maeda K, Maruta M, Utsumi T, Hosoda Y, Horibe Y.
the bookshelf revisited. Dis Col Rectum 2002;45:1-9 Does perifascial rectal excision (i.e. TME) when com-
12. Heald RJ. The holy plane of rectal surgery. J R Soc Med bined with the autonomic nerve-sparing technique
1988;81:503-8 interfere with operative radicality? Colorectal Disease
13. Hainsworth PJ, Egan MJ, Cunliffe WJ. Evaluation of the 2002;4:233-9
policy of total mesorectal excision for rectal and rec- 29. Enker WE, Havenga K, Polyak T, Thaler H, Cranor M.
tosigmoid cancers. Br J Surg 1997;84:652-6 Abdominoperineal resection via total mesorectal exci-
14. Phillips RK, Hittinger R, Blesovsky L, Fry JS, Fielding sion and autonomic nerve preservation for low rectal
LP. Local recurrence following 'curative' surgery for cancer. World J Surg 1997;21:715-20
large bowel cancer: II. The rectum and rectosigmoid. 30. Moriya Y, Sugihara K, Akasu T, Fujita S. Importance of
Br J Surg 1984;71: 17-20 extended lymphadenectomy with lateral node dissec-
15. Holm T, Cedermark B, Haggmark T, Wilking N. Local tion for advanced lower rectal cancer. World J Surg
recurrence after rectal cancer surgery. A surgical and 1997;21:728-32
oncological challenge. Lakartidningen 1994;91:232-5 31. Takahashi T, Ueno M,Azekura K, Ohta H. Lateral node
16. Fegiz G, Indinnimeo M, Gozzo P, Del Grande E, Cataldi dissection and total mesorectal excision for rectal can-
S, Brozzetti S. Low rectal cancer-what is the choice? cer. Dis Colon Rectum 2000;43:S59-68
Dis Colon Rectum 1994;37:S35-41 32. Takahashi T, Ueno M, Azekura K, Ohta H. Lateralliga-
17. Arbman G, Nilsson E, Hallbook 0, Sjodahl R. Local ment: its anatomy and clinical importance. Semin Surg
recurrence following total mesorectal excision for rec- Oncol 2000;19:386-95
tal cancer. Br J Surg 1996;83:375-9 33. Hojo K, Sawada T, Moriya Y. An analysis of survival
18. McCall JL. Total mesorectal excision: evaluating the and voiding, sexual function after wide iliopelvic lym-
evidence. Aust N Z J Surg 1997;67:599-602 phadenectomy in patients with carcinoma of the rec-
19. Moriya Y, Sugihara K, Akasu T, Fujita S. Patterns of tum, compared with conventional lymphadenectomy.
recurrence after nerve-sparing surgery for rectal ade- Dis Colon Rectum 1989;32:128-33
nocarcinoma with special reference to loco-regional 34. Glass RE, Ritchie JK, Thompson HR, Mann cv. The
recurrence. Dis Colon Rectum 1995;38:1162-8 results of surgical treatment of cancer of the rectum by
20. Aitken RJ. Mesorectal excision for rectal cancer. Br J radical resection and extended abdomino-iliac lym-
Surg 1996;83:214-6 phadenectomy. Br J Surg 1985;72:599-601
21. Enker WE, Thaler HT, Cranor ML, Polyak T. Total 35. Ratto C, Sofo L, Ippoliti M, Merico M, Bossola M,
mesorectal excision in the operative treatment of car- Vecchio FM, Foglietto GB, Crucitti F. Accurate lymph-
cinoma of the rectum. J Am Coll Surg 1995;181:335-46 node detection in colorectal specimens resected for
22. MacFarlane JK, Ryall RD, Heald RJ. Mesorectal exci- cancer is of prognostic significance. Dis Col Rectum
sion for rectal cancer. Lancet 1993;341:457-60 1999;42:143-58
23. Tocchi A, Mazzoni G, Lepre L, Liotta G, Costa G, 36. Esser S, Reilly WT, Riley LB, Eyvazzadeh C, Arcona S.
Agostini N, Miccini M, Scucchi L, Frati G, Tagliacozzo The role of sentinel lymph node mapping in staging of
S. Total meso rectal excision and low rectal anastomo- colon and rectal cancer. Dis Col Rectum 2001;44:850-6
sis for the treatment of rectal cancer and prevention of 37. Lazorthes F, Fages P, Chiotasso P, Lemozy J, Bloom E.
pelvic recurrences. Arch Surg 2001;136:216-20 Resection of the rectum with construction of a colonic
24. Ratto C, Ricci R, Rossi C, Morelli U, Vecchio FM, reservoir and colo anal anastomosis for carcinoma of
Doglietto GB. Mesorectal microfoci adversely affect the rectum. Br J Surg 1986;73:136-8
188 Atlas of Endoanal and Endorectal Ultrasonography

38. Nicholls RJ, Lubowski DZ, Donaldson DR. Comparison 44. Williams NS, Price R, Johnston D. The long-term effect
of colonic reservoir and straight coloanal reconstruc- of sphincter preserving operations for rectal carcino-
tion after rectal excision. Br J Surg 1988; 75:318-20 ma on function of the anal sphincter in man. Br J Surg
39. Sweeney JL, Ritchie JK, Hawley PRo Resection and 1980;67:203-8
sutured peranal anastomosis for carcinoma of the rec- 45. Parc R, Tiret E, Frileux P, Moszkowoski E, Loygue J.
tum. Dis Col Rectum 1989;32:103-6 Resection and coloanal anastomosis with colonic
40. Parks AG, Percy JP. Resection and sutured colo anal reservoir for rectal carcinoma. Br J Surg 1986;73:138-41
anastomosis for rectal carcinoma. Br J Surg 1982; 46. Santoro GA, Makhdoomi KR, Eitan BZ, Bartolo DCC.
69:301-4 Functional outcome after coloanal anastomosis with J-
41. Keighley MRB, Matheson D. Functional results of rec- colonic pouch for rectal cancer. Ann Ital Chir
tal excision and endoanal anastomosis. Br J Surg 1980; 1998;LXIX,4:485-9
67:757-61 47. Brown SR, Seow-Choen F. Preservation of rectal func-
42. Cavaliere F, Pemberton JH, Cosimelli M, Fazio VW, tion after low anterior resection with formation of a
Beart RW. Coloanal anastomosis for rectal cancer. neorectum. Semin Surg Oncol 2000;19:376-85
Long-term results at the Mayo and Cleveland Clinics. 48. Sagamata Y, Takase Y, Oya M. Scintigraphic compari-
Dis Col Rectum 1995;38:807-12 son of neorectal emptying between clonic J-pouch
43. Suzuki H, Matsumoto K, Amano S, Fujioka M, anastomosis and straight anastomosis after stapled
Honzumi M. Anorectal pressure and rectal compliance low anterior resection. J Colorectal Dis 2003;18:355-60
after low anterior resection. Br J Surg 1980;67:655-7
 Invited commentary
1. Zorcolo, D.C.C. Bartolo

The surgical treatment of rectal cancer is depen- margins. Thus the main challenge for rectal sur-
dent on the level of the lesion. Tumors of the geons is to have an accurate means of planning
upper rectum have a risk of local recurrence, treatment according to strict scientific protocols.
which is comparable to sigmoid tumors, and for One of the problems highlighted is selecting
this reason are usually treated with a standard the level of inferior mesenteric artery division.
anterior resection [I]. The Dutch trial [2] has Literature suggests that high legation does not
shown that there was no benefit from adding modify the poor prognosis of those patients
radiotherapy to tumors of 10 cm and above. Mid found to have positive apical nodes [3, 4].
(5-10 cm from anal verge) and low rectal «5 cm) Nevertheless, from a technical standpoint high
cancers have a different biological behavior and legation, together with mobilization of the splenic
their treatment presents a considerable manageri- flexure, are usually necessary in order to allow the
al challenge. The controversy has been ably pre- descending colon to reach the pelvis for a tension-
sented in this chapter. free anastomosis. Moreover, obtaining histology
The major advance in recent years has been of the apical nodes allows more accurate staging
with preoperative staging and imaging of rectal and prognostication.
tumors. Unfortunately, endorectal ultrasonogra- One of the milestones in the management of
phy tends to over stage and the sensitivity is only rectal cancer has been the fact that low recurrence
about 70% in predicting lymph node involve- was dependent on lateral clearance rather than
ment. Magnetic resonance may be marginally bet- distal resection [5-9]. This allowed the "5 cm rule"
ter since it is less operator dependent. Once again, to be abandoned. Heald [8] has popularized Total
the message from the Dutch trial is that radio- Mesorectal Excision and he and others have
therapy is mainly of benefit in node positive mid- shown that meticulous surgery that removes the
rectal tumors. The challenge therefore is to select rectum with a clean mesorectal facial envelope is
accurately those patients who will benefit because associated with a substantial reduction in the
their nodes are positive and to protect the remain- incidence oflocal recurrence rates. It is now stan-
der of patients from the potential hazards of dard practice to limit distal resection margins to 2
radiotherapy that will serve them no actuarial cm or less [10]. The key is to assess the lateral
benefit in terms of survival or reduction of local spread at the distal margin. Clearly in some
recurrence rates. It has been demonstrated that instances it is advisable to resect more when a
radiotherapy is associated with poor healing of bulky tumor has extended laterally and this may
the perineal wound and in elderly patients any in some patients necessitate doing an abdomino-
delay in following short course radiotherapy with perineal resection. Current practice along these
early surgery beyond 3 days is associated with an lines has reduced the permanent stoma and local
increase in cardiovascular morbidity and mortal- recurrence rates.
ity. The benefits of radiotherapy are also most The aim of rectal surgery has to be the removal
marked in bulky T3 tumors with short lateral of the rectum with all its surrounding mesorectum.
190 Atlas of Endoanal and Endorectal Ultrasonography

Local recurrence is highly dependent on involve- changing with the concept of AME (Adequate
ment of the circumferential resection margin Mesorectal Excision) and several authors [19, 21,
(CRM) [11]. Moreover, even when the lateral mar- 22,23] consider it safe to transect the mesorectum
gins are clear, the macroscopic evidence of an 2-5 cm beyond the tumor in order to obtain a
incomplete meso rectal excision is associated with longer rectal cuff for the anastomosis.
higher risk of LR [12]. A correct sharp anatomical Mesorectal excision obviously cannot treat the
dissection must remove all the "oncologically dan- potential metastases in the iliac lymph nodes,
gerous" perirectal tissue, whilst at the same time which according to Takahashi T et al. [24] are pre-
preserving the important neuro-vascular elements. sent in 8.6 percent of all rectal cancers and 16.4
The results of TME in terms of lowering local percent of low-lying rectal cancer cases. However,
recurrence rates have been already presented in survival after extended pelvic lymphadenectomy
this chapter. What we would like to emphasize is is reported to approximate that obtained with
that the introduction of TME concept has acceler- TME alone [25], while the incidence of urinary
ated the creation of colorectal specialists. In the and sexual complications are unacceptably high.
pre-TME era, percentages of LR varied from 10 up We suggest that such advanced cases should be
to 50% [13], meaning that rectal surgery was clear- identified by accurate preoperative staging and
ly performed in different ways and by surgeons treated by neo-adjuvant chemoradiation.
with different skills all around the world. After the The other area discussed is the neorectum and
results presented by Heald, the technique has been the use of a pouch to optimize bowel function
standardized and it has become evident that rec- after ultra-low anastomosis. The creation of a J
tal tumors should always be treated by skilled sur- colonic pouch results in better early functional
geons after specific training. TME has been rec- outcome compared to straight colo-anal anasto-
ommended in guidelines for the management of mosis, but the benefits seem to be minimal in the
rectal cancer published by The Royal College of long-term [26]. A J-pouch is not without compli-
Surgeons of England and the Association of cations and to avoid defecation disorders its size
Coloproctology of Great Britain and Ireland [14] has to be restricted to approximately scm. A
and in Scandinavia the term "TME Surgeon" has recent randomized controlled trials (RCT) com-
been introduced [15]. paring short J-pouch and straight anastomosis
One of the aspects which is still debated is found that the functional outcome was better in
whether mid-rectal tumors (5-10 cm from anal the pouch group but the neorectal capacity was
verge) should always be treated with complete decreased to a similar degree in both groups com-
removal of the distal "tongue" of mesorectum, as pared to the preoperative rectal volume. Thus, it
described by Heald, or if the mesorectum can be was postulated that the advantage of the colonic J-
transected at some point beyond the tumor, in pouch was not the creation of a larger reservoir
order to preserve a longer rectal cuff for anasto- but construction of a segment with decreased
mosis and improved post-operative function. motility [27].
Complete removal of the distal mesorectum Transverse coloplasty has been proposed as an
means extending the level of resection into the attractive alternative to a J-pouch [28]. This pro-
lower third of the rectum and requires an ultra- cedure consists of making a longitudinal incision
low anastomosis. These are then associated with along the colonic wall and closing the defect
higher anastomotic leak rates [16], a higher transversely. Preliminary results from RCT have
requirement for temporary stomas [17] and an shown a functional outcome comparable to that
increased incidence of rectal urgency and inconti- of the J-pouch [29,30], while others [31] found
nence [18]. Moreover, the oncological value of that coloplasty pouches resulted in higher per-
complete removal of the distal "tongue" of centage of anastomotic leakage. An alternative to
mesorectum has never been proven [19], whilst it the creation of a pouch is the side-to-end anasto-
has been shown that distal diffusion in the mosis, whose results seem to be acceptable and
meso rectal fat is rare and often associated with comparable with those of more complicated pro-
aggressive tumors, whose prognosis is mainly cedures [32].
related to systemic rather than local recurrence In conclusion, the surgery of rectal cancer con-
[20]. For these reasons the concept of TME is tinues to pose major challenges in terms of
improving oncological and functional outcomes.
 Section VIII- Surgical treatment options for rectal cancer 191

References ation of rectal cancer resection specimen: clinical sig-

nificance of the pathologist in quality control. J Clin
1. Lopez-Kostner F, Lavery IC, Hool GR, Rybicki LA, One 2002; 20:1729-34
Fazio VW. Total mesorectal excision is not necessary 13. McCall JL, Cox MR, Wattchow DA. Analysis of local
for cancers of the upper rectum. Surgery 1998; 124:612- recurrence rates after surgery alone for rectal cancer.
7; discussion 617-8 Int J Colorectal Dis 1995; 10:126-32
2. Peeters KC, Kapiteijn E, van de Velde CJ; Dutch 14. Royal College of Surgeons of England, Association of
ColoRectal Cancer Group. Managing rectal cancer: the Coloproctology of Great Britain, and Ireland.
Dutch experience. Colorectal Dis 2003; 5:423-6 Guidelines for the management of colorectal cancer.
3. Hida 1, Yasutomi M, Maruyama T, Fujimoto K, 199 6; P25
Nakajima A, Uchida T, Wakano T, Tokoro T, Kubo R, 15. Heald RJ. Total mesorectal excision is optimal surgery
Shindo K. Indication for using high ligation of the for rectal cancer: a Scandinavian consensus. Br J Surg
inferior mesenteric artery in rectal cancer surgery. 1995; 82:1297-9
Examination of nodal metastases by the clearing 16. Carlsen E, Schlichting E, Guldvog I, Johnson E, Heald
method. Dis Colon Rectum 1998; 41:984-7; discussion RJ. Effect of the introduction of TME for the treatment
987-91 of rectal cancer. Br J Surg 1998; 85:526-9
4. Adachi Y, Inomata M, Miyazaki N, Sato K, Shiraishi N, 17. Heald RJ, Karanjia ND. Results of radical surgery for
Kitano S. Distribution of lymph node metastasis and rectal cancer. World J Surg 1992; 16:848-57
level of inferior mesenteric artery ligation in colorec- 18. Nesbakken A, Nygaard K, Lunde ~C. Mesorectal exci-
tal cancer. J Clin Gastroenterol1998; 26:179-82 sion for rectal cancer: functional outcome after low
5. Penfold JC. A comparison of restorative resection of anterior resection and colorectal anastomosis without
carcinoma of the middle third of the rectum with reservoir. Colorectal Disease 2002; 4:172-6
abdominoperineal excision. Aust NZJ Surg 1974; 19. Beart RW Jr. Mesorectal excision for rectal carcinoma:
44:354-6 the new standard? Adv Surg 1999; 32:193-203
6. Williams NS, Dixon MF, Johnston D. Reappraisal of the 20. Scott N, Jackson P, al-Jaberi T, Dixon MF, Quirke P,
5 centimetre rule of distal excision for carcinoma of Finan PJ. TME and local recurrence: a study of tumor
the rectum: a study of distal intramural spread and of spread in the mesorectum distal to rectal cancer. Br J
patient's survival. Br J Surg 1983; 70:150-4 Surg 1995; 82:1031-3
7. Pollett WG, Nicholls RJ. The relationship between the 21. Sjodahl R. The role of total mesorectal excision in rec-
extent of distal clearance and survival and local recur- tal cancer surgery. EJSO 2001; 27:440-1
rence rates after curative anterior resection for carci- 22. Beck DE, Wexner SD. Fundamentals of anorectal
noma of the rectum. Ann Surg 1983; 198:159-63 surgery. WB Saunders Company Limited 1998
8. Heald RJ, Husband EM, Ryall RDH. The mesorectum 23. Killingback M, Barron P, Dent 0 F. Local recurrence after
in rectal cancer surgery - the clue to pelvic recur- curative resection of cancer of the rectum without total
rence? Br J Surg 1982; 69:613-6 mesorectal excision. Dis colon Rectum 2001; 44:473-86
9. Quirke P, Durdey P, Dixon MF, Williams NS. Local 24. Takahashi T, Ueno M, Azekura K, Ohta H. Lateral node
recurrence of rectal adenocarcinoma due to inade- dissection and total mesorectal excision for rectal can-
quate surgical resection: histopathological study of cer. Dis Colon Rectum 2000; 43:S59-68
lateral tumor spread and surgical excision. Lancet 25. Moreira LF, Hizuta A, Iwagaki H, et al. Laterallymphn-
1986; 2:996-8 ode dissection for rectal cancer below the peritoneal
10. Andreola S, Leo E, Belli F, Bonfanti G, Sirizzotti G, reflection. Br J Surg 1994; 81:293- 6
Greco P, Valvo F, Tomasic G, Gallino GF. 26. Ho YH, Seow-Choen F, Tan M. Colonic J-pouch func-
Adenocarcinoma of the lower third of the rectum sur- tion at six months versus straight coloanal anastomo-
gically treated with a <lO-MM distal clearance: pre- sis at two years: randomized controlled trial. World J
liminary results in 35 No patients. Ann Surg Oncol Surg 2001;25:876-81
2001; 8:611-5 27. Furst A, Burghofer K, Hutzel L, Jauch KW. Neorectal
11. Birbeck KF, Macklin CP, Tiffin NJ, Parson W, Dixon MF, reservoir is not the functional principle of the colonic
Mapstone NP et al. Rates of circumferential resection J-pouch: the volume of a short colonic J-pouch does
margin involvement vary between surgeons and pre- not differ from a straight coloanal anastomosis. Dis
dict outcome in rectal cancer surgery. Ann Surg 2002; Colon Rectum 2002;45:660-7
235:449-57 28. Z'graggen K, Maurer CA, Buchler MW. Transverse
12. Nagtegaal ID, van de Velde CJH, van der Worp E, coloplasty pouch. A novel neorectal reservoir. Dig Surg
Kapiteijn E, Quirke P, van Krieken. Macroscopic evalu- 1999;16:363-6
192 Atlas of Endoanal and Endorectal Ultrasonography

29. Pimentel JM, Duarte A, Gregorio C, Souto P, Patricio J. KW, Tang CL. Comparison of J-pouch and coloplasty
Transverse coloplasty pouch and colonic J-pouch for pouch for low rectal cancers: a randomized, controlled
rectal cancer: a comparative study. Colorectal Dis trial investigating functional results· and comparative
2003;5:465-70 anastomotic leak rates. Ann Surg 2002;236:49-55
30. Furst A, Suttner S, Agha A, Beham A, Jauch KW. 32. Machado M, Nygren J, Goldman S, Ljungqvist 0
Colonic J-pouch vs. coloplasty following resection of Similar outcome after colonic pouch and side-to-end
distal rectal cancer: early results of a prospective, ran- anastomosis in low anterior resection for rectal can-
domized' pilot study. Dis Colon Rectum 2003;46:1161-6 cer: a prospective randomized trial. Ann Surg
31. Ho YH, Brown S, Heah SM, Tsang C, Seow-Choen F, Eu 2003;238:214-20
 Invited commentary
T. Hull

The management of rectal cancer has significant- the IMA is ligated close to the aorta and the infe-
1y evolved over the past several decades. The rior mesenteric vein ligated parallel with the
excellent comprehensive overview by Drs. Santoro artery (and again just beneath the pancreas if
and Di Falco outlines many of these. With every needed for mobilization), the colon can be divid-
surgical advancement toward avoidance of a per- ed at the descending colon-sigmoid junction.
manent stoma, the first priority has been to cure Combined with splenic flexure mobilization, the
the patients of their rectal cancer. Since the mid distal colonic end has sufficient reach and blood
1980'S function and quality of life after low anas- supply so that a tension free, well-vascularized
tomosis has become an important secondary pri- anastomosis can be performed. This will decrease
ority. Even with new thoughts and approaches to the risk of leak or ischemic stricture.
rectal cancer treatment, some aspects remain
unsettled and controversial. I will focus on these
while addressing the excellent chapter by the Distal margins
Authors. Additionally, I will reinforce some
important points previously made.
The exact distal margin needed for cancer clear-
ance in low rectal cancers remains controversial.
Inferior mesenteric artery ligation Agreeing that survival rates and local recurrence
rates should be similar between those undergoing
an abdominal perineal resection versus a low
At the Cleveland Clinic high ligation of the inferi- colo anal anastomosis, there are some patients
or mesenteric artery (IMA) and generally the "no with cancers in the distal third of the rectum that
touch" technique is practiced. This stems from can avoid permanent colostomy. When reviewing
work by Rupbert Turnbull at this institution. He literature reporting local recurrence rates for all
found that tumor cells were disseminated in the studies involving rectal cancer, it is important to
bloodstream with manipulation of the tumor. He be sure the authors are obeying oncologic princi-
believed that initial high ligation of the vessel ples. During the conduct of the operation, prior to
would avoid unnecessary blood borne tumor cells consideration of an anastomosis vs. an abdomino-
[1]. Studies can be found (as evidenced in this perineal resection (APR), total mesorectal exci-
review) supporting both principles of high liga- sion (TME) is needed for low rectal cancer. Older
tion of the IMA flush with the aorta versus more studies were performed prior to the recognition
distal ligation. The most important point of this of the importance of avoiding entering the invest-
discussion and also stressed by Drs. Santoro and ing envelope of the rectal mesentery. Any study
Di Falco, is that appropriate resection based on regarding rectal cancer must be critically assessed
anatomy will provide well-vascularized bowel with this in mind, especially older studies. Quirke
ends and sufficient length for anastomosis. When [2] reported 38% of specimens had a positive lat-
194 Atlas of Endoanal and Endorectal Ultrasonography

eral margin (tangential) after surgery for rectal Extended pelvic lymphadenectomy
cancer. In more than half, the surgeon thought he
had a curative resection. Today most surgeons
who perform rectal cancer surgery must be acute- Lateral lymphadenectomy is a concept popu-
ly aware of this dissection plane for adequate larised and practiced in Japan. When reviewing
oncological technique in the low pelvis. recent continuing educational program syllabuses
With the thought that a S cm distal margin for across North America designed to provide updates
rectal cancer is unnecessary, in 1997 we reviewed and reviews in rectal cancer treatment, the topic of
our results for cure in patients with rectal tumors extended pelvic lymphadenectomy is currently
S-7 cm from the anal verge as measured by rigid almost never included. It is unclear why the
proctoscopy. Based on previous teachings from Japanese find the increased incidence of lateral
Dr. Turnbull we have always removed the entire node involvement. Looking at the referenced arti-
mesorectum for distal rectal cancers. Of those in cles in Drs. Santoro's and Di Falco's review, Moriya
our database (reflecting patients starting in 1980) found lateral spread in 14% of T2 patients and 30%
that met the criteria, 162 had low anastomosis and of T3 [S]. This compared to Takahashi who found
99 had APR. With a mean follow-up of over 60 lateral node involvement in 2.8% of T1, 3.4% of T2,
months, there was no statistical difference in local and 8.8% of T3 [6]. This large dichotomy may be
recurrence (8 VS.ll% P=.41), distant metastasis (23 related to level of tumor as it was also reported by
vs. 28% P=.3S) or five year disease free survival Takahashi that tumors below S cm have clinically
(70.S vs. 62.3% p=.2) for anastomosis vs. APR significant lateral node involvement.
respectively. This was also true comparing low If the lateral nodes were not removed at the
anastomosis and APR stage for stage [3]. time of the rectal cancer surgery, involvement
Therefore, we feel for low rectal cancers that should be reflected in a positive local recurrence
total excision of the rectal mesentery is needed. as the lateral nodes are pelvic structures. Hence,
We will perform an anastomosis if a resection when omitting lateral node resection as part of the
margin of one cm can be safely obtained. initial operation, there should be high local recur-
rence rates. When reviewing the Cleveland Clinic
Rectal Cancer database from 1980 until 1998 there
Mesorectal excision were 1349 curative resections for rectal cancer
(unpublished data). SS% were cancers in the lower
rectum, 0-7 cm from the verge. The local recur-
As mentioned above, the recognition and promo- rence rate was 8% for all stages. If the incidence of
tion of sharp dissection TME has allowed sur- lateral node involvement was significant, as stated
geons to perform similar oncologically sound above, I would anticipate that the local recurrence
resection of rectal cancers in the distal and mid rate would be higher as we do not do lateral node
rectum. This should unequivocally lead to dissection. Therefore, I am not sure why rectal can-
reduced local recurrence rates. However at the cer surgery differs in the US, but the risk of uri-
Cleveland Clinic resection of upper rectal cancers nary and sexual dysfunction entailed with lateral
has been identical to sigmoid cancer. Namely a S node dissection outweighs the risk of recurrence
cm margin along with the accompanying mesen- for lateral node involvement in our population.
tery is resected. It is important to take the fullS cm
of mesentery and avoid the phenomena of "coning
in". When reviewing our results comparing upper Formation of a neorectum
rectal cancer (n=229) resection to sigmoid
(n=22S) and lower rectal (n=437) cancer resec-
tion, the risk of recurrence and cancer death were Historically it appears as attention focused on the
similar for upper rectal and sigmoid cancers. quality of life patients had following ileal pelvic
However, lower rectal cancers were not similar to pouch surgery, surgeons became aware of bowel
upper rectal cancers. The risk of recurrence and dysfunction which patients experienced after rectal
cancer death were 2.7 and 0.8 times more likely in resection with low colorectal and coloanal anasto-
those with lower rectal cancer vs. upper rectal mosis. These problems believed to be related to
cancer. Nonetheless, this supports the notion that resection of the rectal reservoir include frequent
upper rectal cancers can be resected in a similar stooling, urgency, nocturnal bowel movement,
fashion as sigmoid cancers. and fragmentation of stooling. This constellation
 Section VIII' Surgical treatment options for rectal cancer 195

is frequently referred to as the anterior resection colonic reservoir [17]. They further reported more
syndrome [7]. It may also include incontinence. work on this novel pouch [18,19]. Based on the
Since 1986, the colonic J-pouch has gained pop- simplicity and results obtained, we felt this pouch
ularity for reconstruction after rectal resection could be used for the 25% of patients where a
with a low colorectal or colo anal anastomosis [8, colonic J-pouch could not be performed. Initially
9] Through studies it has been determined that the we used this "coloplasty" pouch on a trial basis.
reservoir should be 5-8 cm [10, ll] and used when [20, 21] Comparing three groups of patients
the remaining rectal remnant is equal to or less colonic J-pouch (n=16), coloplasty (n=20), and
than 4 cm [12]. It has been shown that the number straight anastomosis (n=17), postoperative
of stools daily and the problems with urgency are sphincter pressure were similar in all three
less when the colonic J-pouch is incorporated into groups. However, the maximum tolerated volume
the low colorectal or coloanal anastomosis [13]. and compliance were statistically better in the
The benefit of a colonic J-pouch has clearly coloplasty and colonic J-pouch groups compared
been shown to effect the patients bowel pattern to the straight group (Chi-square P<0.05).
for at least one to two years after surgery [13-15]. Examining functional results, there was no sta-
When we compared 119 consecutive patients look- tistical difference comparing all three groups when
ing to see if improvement in functional outcome looking at use of antidiarrheal medication and
was sustained from the period of 5-9 years from continence. However, the straight group had statis-
their original surgery, there were 62 with a colonic tically more bowel movements daily (4.5 range 1-8)
J-pouch and 57 with a straight anastomosis. [16] when compared to the coloplasty group (2.6 range
Patients with the colonic J-pouch had significant- 1-5) or the colonic I-pouch group (3.1 range 2-6) (t-
ly better continence scores for the duration of the test P<0.05).
study including 5-9 years after surgery when com- Figures VII1.36-39 outline construction of the
pared with those having a straight anastomosis. 5 coloplasty. We recently compared 69 coloplasty
to 9 years after surgery the median number of
nighttime bowel movements was also significant-
1y lower in those with a colonic J-pouch (P=0.02).
Patients with colonic J-pouch continued to expe-
rience significantly less evacuation difficulties
and less urgency of defecation than those with the
straight anastomosis. Therefore, the teaching that
in the "long run" a colonic J-pouch offers no ben-
efit to patients over a straight anastomosis may be
inaccurate. Even after accommodation to the
change in anatomy with the straight anastomosis,
some people never reach the superior function
offered by the colonic J-pouch. 8-10cm
Many surgeons feel the colonic I-pouch should
be used routinely whenever the rectal remnant is 4
cm or less. Unfortunately, this is not always techni-
cally feasible. When looking at our results, about
one quarter of patients were unable to have a
colonic I-pouch for technical reasons (unpub-
lished data from the Cleveland Clinic Foundation).
These reasons included a bulky colonic pouch
4-6cm
from excessive fat in the mesentery, which would
not fit into through the pelvis (especially in a nar-
row male pelvis), limitations in the ability to reach
the pelvic floor (usually due to a fatty mesocolon
which hindered the pouch mobility), and difficul-
ty in pulling a bulky colonic J-pouch through long Fig. VIII.36. A conventional resection is done to remove the
rectum. 4-6 em from the distal cut end of the colon an 8-10 em
tight anal muscles.
longitudinal colostomy is made between the tenia [20].
In 1997, Z'graggen and colleagues from (Reprinted with permission from Diseases of the Colon and
Switzerland described (in abstract form) a "novel" Rectum)
196 Atlas of Endoanal and Endorectal Ultrasonography

Fig. VIII.37. Stay sutures are placed midway between the ends
of the colostomy so that the longitudinal opening can be closed
transversely. After the stay sutures, the next suture is a sero-
muscular stitch in the middle. It is not tied initially but held as
a traction suture. Then starting at each end (marked by stay Fig. VIII.38. If a handsewn anastomosis is required prior
sutures) and working toward the middle suture, seromuscular planning is needed so that sufficient distance remains from
sutures are placed to close the colostomy. We use 2-0 polygly- the distal colostomy to the cut edge of the colon. The goal is
colic acid suture. Once the colostomy is closed, it is tested by that the coloplasty suture line will be positioned above the
instilling saline in the open end and occluding the bowel just sphincters after the anastomotic suture line is completed and
above the suture line to rule out a leak [20j . (Reprinted with not in the anal canal [20j. (Reprinted with permission from
permission from Diseases of the Colon and Rectum) Diseases of the Colon and Rectum)

patients, 43 colonic J-pouch patients, and 50 randomized fashion to see if there are differ-
patients with a straight anastomosis who had ences or benefits of the coloplasty over the
surgery from 1998 to 2001 [22]. The choice of colonic J-pouch.
reconstruction was left up to the surgeon's pref-
erence. The coloplasty and colonic J-pouch
patients had significantly fewer nighttime bowel Conclusion
movements than the straight group. The colo-
plasty group also had fewer bowel movements
during the day than the straight group and used Surgical treatment of rectal cancer has evolved in
less antidiarrheal medications. We concluded many aspects in recent decades. Drs. Santoro's
that the coloplasty offered superior function to and Di Falco's excellent review highlights this
the straight anastomosis and was similar to the fact. The necessity of total rectal mesenteric
colonic J-pouch. Therefore it is an excellent alter- resection and avoiding the "coning in" phe-
native when the colonic J-pouch cannot be per- nomenon has become standard surgical conduct
formed. We currently have a large multi-institu- for low rectal cancers. The minimal distal margin
tional study underway looking prospectively in a required for safe resection of rectal cancers is still
 Section VIII' Surgical treatment options for rectal cancer 197

References
1. Turnbull RB, Kyle K, Watson F, et al. Cancer of the
colon: the influence of the no-touch technique on sur-
vival rates. Ann Surg 1967;166:420-7
2. Quirke P, Durdey P, Dixon M, et al. Local recurrence of
rectal adenocarcinoma due to inadequate surgical
resection. Histopathology study of lateral tumor
spread and surgical excision. Lancet 1986;2:996-9
3. Lavery IC, Lopex-Kostner F, Fazio VW, et al. Chances of
cure are not compromised with sphincter-saving pro-
cedures for cancer of the lower third of the rectum.
Surgery 1997;122:779-85
4. Lopez-Kostner F, Lavery IC, Hool G, et al. Total
mesorectal excision is not necessary for cancers of the
upper rectum. Surgery 1998;124:612-8
5. Moriya Y, Sugihara K, Akasu T, et al. Importance of
extended lymphadenectomy with lateral node dissec-
tion for advanced lower rectal cancer. World J Surg
1997;21:728-32
6. Takahashi T, Ueno M, Azekura K, et al. Lateral node
dissection and total mesorectal excision for rectal can-
cer. Dis Colon Rectum 2000;43:S59-68
7. Ortiz H, Armendariz P. Anterior resection: do the
patients perceive any clinical benefit? Int J Colorectal
Dis 1996;11:191-5
Fig. VIII.39. A stapled anastomosis can also be performed. 8. Parc R, Tiret E, Frileux P, et al. Resection and colo-anal
This figure illustrates a double pursestring used to tie around anastomosis with colonic reservoir for rectal carcino-
each part of the gun for the anastomosis. The distal rectal
ma. Br J Surg 1986;73:139-41
stump can also be stapled and a double stapled anastomotic
technique performed with the circular stapler [20]. 9. Lazorthes F, Fages P, Chiotasso P, et al. Resection of the
(Reprinted with permission from Diseases of the Colon and rectum with construction of a colonic reservoir and
Rectum) colo-anal anastomosis for carcinoma of the rectum. Br
J Surg 1986;73=136-8
10. Dennett ER, Parry BR. Misconceptions about the
colonic J-pouch: what the accumulating data show. Dis
Colon Rectum 1999;42:804-11
11. Lazorthes F, Gamagami R, Chiotasso P, et al.
not definite, but 5 em is not needed for low rectal Prospective, randomized study comparing clinical
cancer. The need to divide the IMA at the aortic results between small and large colonic J-pouch- fol-
level is debatable, but it does facilitate and lowing colo anal anastomosis. Dis Colon Rectum
improve mobility, oncologic resection and anas- 1997;40:1409-13
tomosis so should be highly considered. The 12. Hida J, Yasutomi M, Maruyama T, et al. Indications
occurrence of positive lymph nodes along the lat- for colonic J-pouch reconstruction after anterior
eral pelvic sidewall is not well understood, but resection for rectal cancer: determining the opti-
would not be considered standard for oncologic mum level of anastomosis. Dis Colon Rectum 1998;
resection of rectal cancers in the US. When there 41:558-63
is a low colorectal or colo anal anastomosis, 13. Seow-Choen F, Goh HS. Prospective randomized trial
reconstruction with a neorectum should receive comparing colonic J-pouch anastomosis and straight
strong consideration. The coloplasty is an excel- coloanal reconstruction. Br J Surg 1995;82:608-10
lent alternative when the colonic J-pouch cannot 14. Lazorthes F, Chiotasso P, Gamagami RA, et al. Late
be used. It remains to be proven if the coloplasty clinical outcome in a randomized prospective compar-
will be superior in any respect to the colonic J- ison of colonic J-pouch and straight colo anal anasto-
pouch. mosis. Br J Surg 1997;84:1449-51
198 Atlas of Endoanal and Endorectal Ultrasonography

15. Joo JS, Latulippe JF, Alabaz 0, et al. Long-term func- 19. Z'graggen K, Maurer CA, Buchler MW. Transverse
tional evaluation of straight colo anal anastomosis and coloplasty pouch: a novel neorectal reservoir. Dig Surg
colonic J-pouch: is functional superiority of the 1999;16:363-6
colonic J-pouch sustained? Dis Colon Rectum 1998; 20. Fazio VW, Mantyh CR, Hull TL. Colonic "coloplasty":
41:740-6 novel technique to enhance low colorectal or coloanal
16. Harris GJC, Lavery IC, Fazio Vw. Function of a colonic anastomosis. Dis Colon Rectum 2000;43=1448-50
J-pouch continues to improve with time. Br J Surg 21. Mantyh CR, Hull TL, Fazio VW. Coloplasty in low col-
2001;88:1623-7 orectal anastomosis: manometric and functional com-
17. Z'graggen K, Maurer CA, Mettler D, et al. A novel parison to straight and colonic J-pouch anastomosis.
colonic reservoir for rectal construction: description Dis Colon Rectum 2001;44:37-42
of the technique. Gastroenterology 1997;112:A1487 22. Zutshi M, Remzi F, Lavery IC, et al. Functional outcome
18. Z'graggen K, Maurer CA, Mettler K, et al. A novel colon and complications of coloplasty pouch for low
pouch and its comparison with a straight colo anal and colo anal and colorectal anastomosis. Podium presen-
colon J-pouch-anal anastomosis: preliminary results tation at the American Society of Colon and Rectal
in pigs. Surgery 1999;125:105-12 Surgeons Chicago Illinois June 6, 2002
 VIII.S.
Laparoscopic techniques
G.A. Santoro, G. Di Falco

Controversy remains regarding the appropriate- Surgical techniques

ness of laparoscopic methods for the curative
resection of colorectal neoplasm. Laparoscopic The patient is initially placed in a supine position
surgery for rectal cancer is technically feasible [1]. with legs in a 20° to 25° abducted position sup-
However a laparoscopic colorectal procedure is ported by padded, adjustable stirrups (Fig.
never indicated solely because it can be done VIII.40). The legs are kept as straight as possible
laparoscopically. According to Fazio {2] "one good because movement of the laparoscopic instru-
reason for considering changing the approach to ments may be limited by the legs if the hips are
conventional surgery should be the proven supe- flexed. The patient must be positioned so that the
riority of the new technique". Oncological short- pelvis is just above the break at the lower end of
term outcome does not differ from the results the table to allow the surgeon to have free access
achieved by open techniques [1-3]. Howeverlong- to the perineum for pelvic manipulation or
term oncological results (local recurrence rates, transanal anastomosis. Almost all of the retrac-
disease-free survival and 5-year survival rates) tion of the bowel for exposure during the opera-
after minimally invasive resection must be shown tion is achieved by varying the inclination and
to be equivalent or better than those after open rotation of the operating room table. Thus it is
resection to make a definitive assessment of the essential that the patient remain securely attached
place of laparoscopic surgery in oncological inter- to the table. A moldable bean bag form is placed
ventions done in curative intent. under the patient's head and torso to stabilize the
A consensus of opinion has been reached of body on the table. Hands and arms are tucked at
what constitutes "good laparoscopic cancer the patient's side. Surgery is usually performed in
surgery" [1-7]. The same oncologic principles, the Trendelemburg position (30° head-down tilt)
which apply to laparotomy, should be adhered to with the patient tilted 19° right side down.
for laparoscopic surgery. Intraoperative ultra- For the first phase of the operation the surgeon
sonographic staging, complete lymphadenectomy and the second assistant (camera person) usually
with high ligation of the inferior mesenteric stand on the patient's right side, looking at a mon-
artery, total mesorectal excision with autonomic itor placed on the patient's left side. The first assis-
nerve preservation for carcinomas of the middle tant and the nurse stand opposite the surgeon
and lower third, adequate proximal, lateral and looking at a monitor placed near the patient's
distal resection margins, minimal manipulation right knee (Fig. VIII.41). During mobilization of
of the tumor and wide en bloc resection of the the left colonic flexure the surgeon changes posi-
malignant lesion should be accomplished by tion standing between the patient's legs while the
laparoscopic techniques. monitor is moved near the patient's left shoulder.
The surgical technique of laparoscopic low The first step is to establish a C02 pneu-
anterior resection for rectal cancer is described in moperitoneum of 15 mmHg or less with either a
detail below. Veress needle or an open Hasson technique
 200 Atlas of Endoanal and Endorectal Ultrasonography

~ ________________ ~ b

Fig. VIII.40. (a) Position of the patient on the operating table

for laparoscopic surgery. The legs are supported by padded
a L.-_ _....._ _ ~...........; stirrups and abducted. A bean bag is placed under the patient
to stabilize the body on the table. (b) Diagrammatic represen-
tation

inserted through a small supraumbelical incision. liver. The dissection commences with an incision
The first 10 mm trocar is inserted into the in the retroperitoneum just to the left of the duo-
abdomen in the same location and the peritoneal denojejunal flexure to identify the origin of the
cavity is immediately inspected with a 10 mm,300 inferior mesenteric vein (Fig. VIII.43). Using care-
telescope, to verify the diagnosis, exclude other ful blunt dissecting techniques a peritoneal win-
diseases and in case of a malignancy, determine dow is made just lateral to the inferior mesenteric
any metastatic dissemination of the tumor. We vein (Fig. VIII.44). Through the peritoneal win-
utilize a laparoscopic ultrasound to assess the dow the left mesocolon is bluntly dissected from
liver for metastases. If the operation is deemed the underlying retroperitoneal structures includ-
possible additional ports are placed. The number ing the pancreas and Gerota's fascia (submesen-
and placement of ports is tailored to each specific teric dissection). The line of dissection (white line
procedure. The port site must be close enough to of Toldt) is parallel to the inferior mesenteric vein.
the operating field for the instruments to reach. Dissection is continued laterally to the peritoneal
The ports should be placed far enough apart from attachments of the left colon and towards the
each other to avoid the sword fighting of the spleen, visible at the superior extent of dissection.
instruments. For laparoscopic resection of the Under continuous traction the peritoneum is
rectum we utilize five ports: one for the laparo- incised caudad toward the direction of the origin
scope, two for the operating surgeon and two for of the inferior mesenteric artery (Fig. VIII.45).
the assistant. We have adopted the trocar place- The vessel is isolated, clipped and divided 1.5 cm
ment indicated in Fig. VII1.42. Two 5 mm ports are to its origin from the aorta (Fig. VIII.46). Next the
positioned in the left lateral abdominal wall. The inferior mesenteric vein is divided near its inser-
lower left port will be incorporated in the incision tion into the splenic vein in an identical manner
necessary to remove the specimen. Two addition- (Fig. VIII.47). Dissection is continued laterally
al ports are placed in the right lateral abdominal beneath the inferior mesenteric artery and vein as
wall. A 5 mm port is placed at the level of the the left ureter and the gonadal vessels are identi-
umbilicus lateral to the rectus sheath. The stapling fied and swept posteriorly (Fig. VIII.48). In the
device will be inserted in the caudad 12 mm port next phase, the lateral attachments of the sigmoid
on the right side. The surgeon and assistant sur- colon are dissected free with the ultrasonic har-
geon, on opposite sides of the table, use two monic scalpel (Ultracision, Ethicon Endo-
instruments each through these ports. surgery) (Fig. VIII.49). The grasper in the first
The procedure is started with the bed in a deep assistant's right hand is used to better expose
Trendelemburg position. The omentum is grasped these attachments. Once the peritoneal reflection
with autraumatic grasping forceps and displaced is incised along the white line of Toldt, the
cephalad above the transverse colon and onto the mesosigmoid is easily bluntly pushed toward the
 Section VIII- Surgical treatment options for rectal cancer 201

first

Fig. VIII.41. (a) Position for the operating

team and instruments for laparoscopic
resection of the rectum. (b) Diagrammatic
representation (redrawn and modified

b ~ ,'=-1
______________________________ ~ ______________ ~
from Decanini C et al. Laparoscopic onco-
logic abdominoperineal resection. Dis
Colon Rectum 1994;37:552-8)

midline. In the course of this dissection, the sur- splenic flexure is mobilized, the descending colon
geon must remain close to the bowel edge lateral- and left transverse colon are then pushed down-
ly and between Gerota's fascia and the colonic ward and to the right to complete the dissection
mesentery. Dissection progresses in a cephalad on the posterior surface of the mesocolon, leaving
manner in the direction of the splenic flexure the psoas muscle and the left ureter exposed (Fig.
(Fig. VIIl.so). As the splenic flexure is approached, VIII.S2). The dissection should proceed as far as
attention is directed to the left transverse colon. the surgeon deems necessary to allow the
The gastrocolic omentum is then divided in a rel- descending colon to reach the pelvis. At this point
atively avascular plane, progressing toward the the mobilization of the rectum is initiated along
greater curvature of the stomach in the direction the right side of the pelvis (Fig. VIII.S3). The sur-
of the lienocolic ligament (Fig. VIII.Sl). After the geon grasps the rectum as distal as possible and
 202 Atlas of Endoanal and Endorectal Ultrasonography

a b

Fig. VIII.42. (a, b) Position of the ports

for laparoscopic resection of the rectum.
c ~ ______________________________________________ ~

(c) Diagrammatic representation

retracts it cephalad. The small bowel is pulled out cation. This area is referred to as the lateral rectal
of the pelvis and pushed cephalad. The right stalks. Nerve branches to the mesorectum and
pelvic peritoneum is divided with a harmonic rectum are divided, while those to the deeper
scalpel until the midline is reached. The dissec- pelvic structures are preserved. Dissection on the
tion is continued posterior to the rectum at the left side follows the same plane as selected on the
level of the sacral promontory following the avas- right side (Fig. VIII.SS). Anterior pelvic dissection
cular plane between parietal pelvic fascia and the is usually performed last. The parietal peri-
fascia propria of the rectum (Fig. VIII.S4). The toneum immediately anterior to the peritoneal
mesorectum must always be removed intact with- reflection is incised (Fig. VIII.S6). The planes of
out injuring the pelvic autonomic nerves (Total dissection on the left and right sides are connect-
Mesorectal Excision with Autonomic Nerve ed across the midline. Dissection is in the areolar
Preservation). The pelvic autonomic plexus con- plane anterior to Denonvillier's fascia, exposing
sists of the anterior parasympathetic sacral nerve the seminal vesicles bilaterally. With the rectum
roots from S2-S4 and the hypogastric nerves pre- fully mobilized, the distal margin of resection is
viously identified anterolaterally the aortic bifur- precisely determined, at least 2 cm below the
 Section VIII' Surgical treatment options for rectal cancer 203

or three applications of the roticulating, 45 or 60

mm linear stapler (TYCO Healthcare), that has
been passed through the right-lower quadrant 12
mm port (Fig. VIII.57).
The next step is the proximal segment resec-
tion. The left-lower quadrant 5 mm trocar site is
extended as a transverse incision to approximately
3-5 cm. A wound protector (LAP DISC) is inserted
and the specimen is withdrawn (Fig. VIII.58). The
proximal margin of the transection is selected
based upon vascularity and condition of the
bowel. The remaining mesentery is divided at this
point and the marginal vessels are ligated. A purse-
Fig. VIII.43. Incision in the retroperitoneum just to the left of
string device is applied and the colon is transect-
the duodenojejunal angle to identify the origin of the inferior ed. A proximal purse-string suture is created to
mesenteric vein secure the anvil of the circular stapler (either 29
mm or 31 mm) into the descending colon lumen.
The colon is returned into the abdominal cavity
lesion. For lesions in the upper third of the rectum (Fig. VIII.59), the wound protector is twisted (Fig.
a 5 cm distal margin including the upper mesorec- VIII.60) and the pneumoperitoneum is reestab-
tum is achieved. The rectum is divided with two lished.

a b

c d
Fig. VIII.44. Under continuous traction of the inferior mesenteric vein (a), using careful blunt dissecting techniques a peri-
toneal window is made just lateral to the inferior mesenteric vein (b, c), on the plane of the white line of Toldt (d)
 204 Atlas of Endoanal and Endorectal Ultrasonography

Fig. VIII.4S. Dissection continues toward the origin of the

inferior mesenteric artery from the aorta

____
a
~-.....I b

Fig. VIII.46. The inferior mesenteric artery is isolated (a),

carefully clipped (b) and divided 1.5 cm to its origin from the
aorta (c)
c
 Section VIII' Surgical treatment options for rectal cancer 205

a b

Fig. VIII.47. The inferior mesenteric vein is isolated (a), care-

fully clipped (b) and divided (c)
c

The final stage of the operation is the creation

of an end-to-end colorectal or colo anal anasto-
mosis. The head of the 29 mm or 31 mm circular
stapler is inserted through the anus and advanced
up the rectum to the stapled closure. The plastic
white trocar is screwed through the center of the
staple line (Fig. VIII.61a). A sponge mounted on a
grasping instrument pushes against the rectal
wall as the trocar is advanced. This helps to pre-
vent a tear in the rectal wall. The spike is dis-
lodged from the center rod and removed through
a port (Fig. VIII.61b). Verified the orientation of
the bowel and its mesentery and the lack of ten-
sion, under laparoscopic guidance the anvil of the
Fig. VIII.48. Dissection is continued laterally beneath the
stapler is inserted into the orange collar of the cir-
inferior mesenteric artery and vein with identification of the
left ureter and the gonadal vessels cular stapler (Fig. VIII.61C). The stapling device is
screwed closed and fired (Fig. VIII.61d). The sta-
pler is opened and withdrawn from the rectum.
The integrity of the resulting anastomosis is veri-
fied by inspection of the doughnuts. A drain is
inserted through a trocar and positioned near the
anastomosis (Fig. VIII.62).
 206 Atlas of Endoanal and Endorectal Ultrasonography

Fig. VIlI.49. Initial dissection of the lateral attachments of Fig. VIII.SO. Dissection progresses in a cephalad manner in
the sigmoid and left colon the direction of the splenic flexure

Specific consideration

While it is clear that laparoscopic colorectal

surgery can be performed safely, the question
remains whether laparoscopic colorectal surgery
can be performed safely for patients with malig-
nant disease [8]. Prerequisites for the laparoscopic
treatment of colorectal cancer include appropriate
patient selection, adequate physician training,
skilled assistance and extensive technical support.
The potential advantages of laparoscopic surgery
include decreased postoperative pain, ileus and
hospitalization, faster return to normal activity,
reduced wound infection, a favorable effect on
cell-mediated immunity and better cosmesis.
These results have been carefully analyzed by
a Agachan and Wexner [9]. This report found no dif-
ferences in outcome between laparoscopic and
conventional surgery. For instance the length of
hospitalization ranges in most series from 1 to 40
days. The criterion of return to normal activity
cannot be quantified because many patients with
carcinoma are elderly and retired. Moreover an
objective scientifically valid quantification and
qualification of pain is even more difficult to com-
pare. For these authors currently the only univer-
sally accepted benefit of laparoscopic colorectal
procedures is improved cosmesis. Furthermore
there have been several reports of port site recur-
rences [1, 4, 10-17]. The incidence of this phe-
nomenon has been cited at between 1.5 and 21%.
b
However most evidence derives from advanced
Fig. VIII.S1. Dissection is continued between the greater colorectal cancer and experienced centers report
omentum and the colon (a), progressing toward the greater rates of 0-1.1% (Table VIII.7) [1,4,10-17]. Several
curvature of the stomach in the direction of the lienocolic lig-
ament (b) possible mechanisms contribute to tumor cell dis-
 Section VIII' Surgical treatment options for rectal cancer 207

semination. The wide variation in port-site recur-

rences reported in literature may relate to surgical
technique. The most likely mechanism is shedding
of tumor cells from operative manipulation.
Feasibility of following the same oncologic
principles applied to an open procedure for malig-
nant disease has been already demonstrated [1-7].
If at any time the surgeon is uncertain that the
accustomed standard of technique can be achieved,
the laparoscopic procedure should be converted to
an open operation. What constitutes a reliably safe
distal margin has been under much debate. The 5
cm rule of distal margin was the standard practice
Fig. VIII.52. The left ureter is identified along the left sides of
for many years. Pollett and Nicholls [18] demon-
the pelvis as it crosses the iliac artery strated that a distal margin of less than 2 cm for
rectal cancers does not adversely affect local recur-
rence or survival. Sphincter-preserving operations
(ultra-low anterior resection or coloanal anasto-

a b
Fig. VIII.53. Right pelvic dissection (a). The peritoneum is divided until the midline is reached (b)

Fig. VIII. 54. Total mesorectal excision with autonomic nerve Fig. VIII.55. Left pelvic dissection. The left ureter is identified
preservation at the level of the sacral promontory along the left sides of the pelvis
 208 Atlas of Endoanal and Endorectal Ultrasonography

a b
Fig. VIII.56. Mobilization of the rectum anteriorly (a). Dissection is in the areolar plane anterior to Denonvillier's fascia, expos-
ing the seminal vesicles bilaterally (b)

a b

c d
Fig. VIII.57. The roticulating laparoscopic linear stapler (TYCO Healthcare) has been passed through the right-lower quadrant
12 mm port (a) and has been placed obliquely across the rectum at least 2 em below the lesion (b). The rectum is transected with
two or three applications of the stapler (c). Rectal staple line (d)
 Section VIII' Surgical treatment options for rectal cancer 209

a b
Fig. VIII.58. A wound protector is inserted through the 5 cm transverse incision in the left-lower quadrant (a) and the speci-
men is withdrawn (b)

Fig. VIII.59. The end of the colon containing the anvil-shaft Fig. VIII.60. The wound protector device (a=opened,
assembly is returned to the abdomen b=twisted) (LAP DISC)

mosis) are technically feasible laparoscopically. 4, 21-24]· The need for dissection of the lateral
Most authors reported similar margins of resection pelvic wall for extended lymph node resections
when patients who underwent laparoscopic or remains controversial. Long-term follow-up of
open colorectal resection were compared [1-4]. patients does not seem to support the use of radi-
Advantages for patients treated by high liga- cal pelvic dissections [25]. Furthermore, increased
tion of the inferior mesenteric artery (between morbidity in terms of urological and sexual com-
the origin and the left colic artery) remain con- plications, the length of hospital stay and the
troversial. Some recent investigations suggest that transfusion rate per patient have been reported.
there is no advantage in terms of 5-year survival Recent experience indicates that total
in performing a high ligation of the inferior meso rectal excision with wide lateral margins and
mesenteric artery [19-20]. However division of autonomic nerve preservation is easily done
this vessel close to its origin will more readily through the laparoscope. Weiser and Milsom [5]
allow the descending colon to reach into the pelvis performed approximately 21 rectal resections
for a colorectal anastomosis. with TME for low rectal adenocarcinoma and
Many recent studies have failed to demon- observed acceptable morbidity. Hartley et al. [7]
strate any significant differences between the completed 21 TME totally laparoscopically.
numbers of lymph nodes removed when compar- Histologic parameters (total specimen length,
ing laparoscopy and laparotomy (Table VIII.B) [1, number of lymph nodes harvested, longitudinal
 210 Atlas of Endoanal and Endorectal Ultrasonography

a b

c d
Fig. VIII.6I. The plastic white trocar is screwed through the center of the staple line (a) and removed through a port (b). The
anvil of the stapler is inserted into the orange collar of the circular stapler (c). The stapling device is screwed closed and fired (d)

and radial margin of excision), early survival and

recurrence appeared to be comparable to open
surgery. These authors failed to demonstrate a
clear advantage for the laparoscopic approach in
terms of major morbidity or hospital stay.
Abdominoperineal resection for cancers
requires separate consideration because of con-
cern over sacrificing the anal sphincter unneces-
sarily and high urogenital complication rates. As a
policy if there is any possibility of preserving the
anal sphincter with conventional open tech-
niques, no attempts should be done for laparo-
scopic resection. Leung et al. [26] reported that
compared to open surgery, laparoscopic-assisted
abdominoperineal resection allowed earlier post-
operative recovery, with equal oncological clear-
ance, morbidity, mortality, disease-free interval
and 4-year survival.
Currently, laparoscopic surgery should be
indicated for early rectal carcinoma (ultrasono-
graphic staging UTl-uT2) or advanced cancer
Fig. VIII.62. Pelvic drain placement near the anastomosis (uT3-uT4) after preoperative radio chemotherapy.
 Section VIII' Surgical treatment options for rectal cancer 211

Table VIII.7. Series review of port-site recurrence after laparoscopic colorectal cancer surgery

Authors N° Port-site Mean follow-up Stage Time to recurrence

of operations recurrence (Dukes or TNM)

Prasad [10] 50 4% 30 months A,B A: 26 months

B: 6 months

Berends [u] 14 21% 30 months B,C,D,

Boulez [12] 86 3·5%

Ramos [13] 208 1.4% 12 months C 6 months, 8 months, 21 months

Wexner [4] 22 4·5% B

Fingerhut [14] 92 3·3% 24 months A,B

Jacquet [I5] 445 1.6% A,B,C,D A: 9 months; B: 6 months,

6 months; C: 1, 5, 9 months;
D: 2 months

Fleshman [16] 372 1.1% 22.6 months Stage I, II, III

Vukasin [17] 480 1.1% 12 months Stage III 2,3,7,15,21 months

Buchmann [15] 35 0% 7 months

Huscher [15] 146 0% 16 months

Milsom [I] 42 0% 18 months

Table VIII.8. Comparison of lymph node harvest in open and laparoscopic rectal resection for cancer

Authors N° of patients Procedures No. of nodes (average)

Lord [21] 49 19 Laparoscopic anterior resection 7·8

30 Laparotomic anterior resection 8.9

Tate [22] 25 u Laparoscopic anterior resection 10

14 Laparotomic anterior resection 13

Darzi [23] 28 12 Laparoscopic abdominoperineal resection 9·5

16 Laparotomic abdominoperineal resection 6

Zucker [24] 8 4 Laparoscopic abdominoperineal resection 8

4 Laparotomic abdominoperineal resection 7·3

Wexner [4] 26 12 Laparoscopic anterior resection 19

14 Laparotomic anterior resection 14

Milsom [I] 28 13 Laparoscopic anterior resection 6·5

15 Laparotomic anterior resection 7·3
212 Atlas of Endoanal and Endorectal Ultrasonography

References 14. Fingerhut A. Laparoscopic-assisted colonic resection:

the French experience. In: Jager R, Wexner SD (eds).
1. Milsom JW, Bohm B, Hammerhofer KA, Fazio VW, Laparoscopic colorectal surgery. Churchill
Steiger E, Elson P. A prospective, randomized trail Livingstone, New York, 1996:253-7
comparing laparoscopic versus conventional tech- 15. Tomita H, Marcello PW, Milsan JW. Laparoscopic
niques in colorectal cancer surgery: a preliminary surgery of the colon and rectum. World J Surg
report. J Am Coll Surg 1998;187:46-57 1999;23:397-405
2. Fazio VW, Lopez-Kostner F. Role of laparoscopic 16. Fleshman JW, Fry RD, Birnbaum EH, Kodner IJ.
surgery for treatment of early colorectal carcinoma. Laparoscopic assisted and minilaparotomy approach-
World J Surgery 2000;24:1056-60 es to colorectal diseases are similar in early outcome.
3. Wexner SD, Cohen SM, Johansen 0, Nogueras J, Dis Colon Rectum 1996;39:15-22
Jagelman D. Laparoscopic colorectal surgery: a 17. Vukasin P, Ortega AE, Greene FL, Steele GD, Simons AJ,
prospective assessment and current prospective. Br J Anilione GJ, Weston LA, Beart RW. Wound recurrence
Surg 1993;80:1602-5 following laparoscopic colon resection: results of the
4. Wexner SD, Cohen SM. Port site metastases after American Society of Colon and Rectal Surgeons
laparoscopic surgery for cure of malignancy: a plea for Laparoscopic Registry. Dis Colon Rectum 1996;39:
caution. Br J Surg 1995;82:295-8 S20-3
5. Weiser M, Milsom JW. Laparoscopic total meso rectal 18. Pollett WG, Nicholls RJ. The relationship between ilie
excision with autonomic nerve preservation. Semin extent of distal clearance and survival and local recur-
Surg Oncol 2000;19:396-403 rence rates after curative anterior resection for carci-
6. Lezoche E, Feliciotti F, Paganini AM, Guerrieri M, noma of the rectum. Ann Surg 1983;198:159-63
Campagnacci R, De Sanctis A. Laparoscopic colonic 19. McElwain JW, Bacon HE, Trimpi HD. Lymph node
resections versus open surgery: a prospective non- metastases: experience with aortic ligation of inferior
randomized study on 310 unselected cases. Hepato- mesenteric artery in cancer of the rectum. Surgery
gastroenterology 2000;47:697-708 1954;35:513-31
7. Hartley JE, Mehigan BJ, Qureshi AE, Duthie GS, Lee 20. Pezim ME, Nicholls RJ. Survival after high or low liga-
PWR, Monson JRT. Total mesorectal excision: assess- tion of the inferior mesenteric artery during curative
ment of the laparoscopic approach. Dis Colon Rectum surgery for rectal cancer. Ann Surg 1984;200:729-33
2001;44:315-21 21. Lord SA, Larach SW, Ferrara A, Williamson PR, Lago
8. Scheidbach H, Schmeider C, Hugel 0, Scheurlein H, CP, Lube MW. Laparoscopic resections for colorectal
Barlehner E, Konradt J, Wittekind C, Kockerling F. carcinoma. A iliree-year experience. Dis Colon Rectum
Oncological quality and preliminary long-term results 1996;39:148-54
in laparoscopic colorectel surgery. Surg Endosc 22. Tate JJ, Kwok S, Dawson JW, Lau WY, Li AK.
2003;17:903-10 Prospective comparison of laparoscopic and conven-
8. Agachan F, Wexner SD. Controversies in laparoscopic tional anterior resection. Br J Surg 1993;80:1396-8
colorectal surgery. G Chir 1996;17:77-83 23. Darzi A, Lewis C, Menzies-Gow N, Guillou PJ, Monson
9. Prasad A, Avery C, Foley RJG. Abdominal wall metas- JR. Laparoscopic abdominoperineal excision of the
tases following laparoscopy. Br J Surg 1994;81:1697 rectum. Surg Endosc 1995;9:414-7
11. Berends FJ, Kazemier G, Bonjer HJ, Lange JF. 24. Zucker KA, Pitcher DE, Martin DT, Ford RS.
Subcutaneous metastases after laparoscopic colecto- Laparoscopic-assisted colon resection. Surg Endosc
my. Lancet 1994;344:58 1994;8:12-8
12. Boulez J, Herriot E. Multicentric analysis of laparo- 25. Glass RE, Ritchie JK, Thompson HR. The results of sur-
scopic colorectal surgery in FDCL group: 274 cases. Br gical treatment of cancer of the rectum by radical
J Surg 1994;81:527 resection and extended abdominosacral lym-
13. Ramos JM, Beart RW, Goes SM, Ortega AE, Schlinkert phadenectomy. Br J Surg 1985;72;599-601
RT. Role of laparoscopy in colorectal surgery. A 26. Leung KL, Kwok SPY, Lau WY, Meng WCS, Chung CC,
prospective evaluation of 200 cases. Dis Colon Rectum Lai PBS, Kwong KH. Laparoscopic-assisted
1995;38:494-501 abdominoperineal resection for low rectal adenocarci-
noma. Surg Endosc 2000;14:67-70
 Invited commentary
T.H.A. Arulampalam, C.G.S. Hiischer

The management of rectal cancer is constantly Research Council (MRC) CLASICC trial has pro-
evolving and improving with recent advances in duced robust preliminary randomized data sug-
imaging the disease [1,2] and impressive results for gesting that there is no difference in circumferen-
neoadjuvant radiotherapy [3]. The role of surgery tial resection margin (CRM), lymph node harvest,
has also been pivotal to these advances. It is well morbidity and mortality with acceptable rates of
documented that recurrence following surgery for conversion (approximately 30%) [7]. The final
rectal cancer poses difficult management problems CLASICC data analysis should definitively answer
and the morbidity and mortality are both signifi- the questions regarding long-term safety of the
cant. The contribution of total mesorectal excision laparoscopic approach. It must be noted that only
(TME) has been crucial in reducing local recur- 381 rectal cancers were recruited, a number that is
rence rates to an acceptable level. The natural pro- not sufficient to reveal statistically significant dif-
gression in the surgical management of the disease ferences. It is, therefore, necessary to either accu-
has been to move towards a laparoscopic approach mulate further data using meta-analysis in collab-
with its potential benefits of better vision during oration with other similar trials, such as the NIH
surgery, decreased tissue trauma and improved trial and Singapore NMRC trial or await the results
cosmesis. There is also a perception that postoper- of larger trials (COLOR II) [8].
ative pain should be less than with conventional The questions regarding port-site recurrence
open surgery with faster return of bowel function are addressed succinctly by the paper and the evi-
and consequent reduced hospital stay. These points, dence now suggests that this is not a major prob-
although contentious to some, appear to hold true. lem (refer to Table VIII.7), but randomized trial
data is awaited.

Laparoscopic resection:
a safe approach? Surgical technique

The most important consideration is that oncolog- The technique described is detailed and one to
ical principles are not compromised. The preced- which little can be added. The only important com-
ing chapter addresses these issues in a systematic ments to make are that fewer than 5 ports can be
way and re-iterates our experience of over 1600 used (usually 3 or 4) and dissection is best carried
colorectal resections performed using a laparo- out using a harmonic scalpel. This instrument
scopic approach. The technique is safe in the hands allows accurate dissection with extremely good
of appropriately trained surgeons and this is borne haemostatic capability. It is not this author's prac-
out by the fact that short-term outcomes are tice, but steps to gain control of the inferior mesen-
equivalent to open surgery [4,5]. Lacy's seminal teric artery as the first operative maneuver have
paper on laparoscopic colonic resection interest- been employed. An alternative method of mobiliz-
ingly suggests that there may even be some sur- ing the splenic flexure is to divide the gastrocolic
vival advantages [6]. The United Kingdom Medical ligament, expose the pancreas and continue dissec-
214 Atlas of Endoanal and Endorectal Ultrasonography

tion around the flexure from a cephalad position. A The future

flexible approach must be intrinsic to the surgeon
in order to successfully progress in this field. One
would also stress the importance of well-trained The momentum for laparoscopic resection contin-
support staff in the operating room. The range of ues, but this must be tempered by proper evidence
instruments and the different technologies used in based data. The results of COLOR II and CLASICC
addition to the length of the procedure require are awaited. Development of laparoscopic instru-
concentration from skilled nursing staff. ments has probably reached a plateau, but
advances are being made in video technology,
optics and telerobotics. Telerobotic systems are
Training seen as viable for the future, but cost restraints will
limit their introduction [12]. In addition, the extra
benefit gained is still debatable.
The availability of adequate training is a difficult
problem to address and does vary throughout the
world. A coordinated approach to training junior Conclusion
surgeons is essential if the benefits that have already
been proven are to be widely available. Issues of
safety must be taken into account and novel meth- The challenges that face the surgical community
ods may be adopted. A combination of expert are to apply the technology appropriately and this
courses, live demonstrations and mentoring can be requires adequate training and research. The pre-
used along with training on inanimate simulators ceding chapter highlights many important points
[9,10], but introduction into clinical practice may and our experience suggests that this approach
be limited [11]. A system of fellowships may be of will become part of the standard skills required of
more benefit in the immediate future until the num- the colorectal surgeon of the future.
ber of institutions carrying out these advanced pro-
cedures increases. Training requires a substantial
investment of resources both in terms of operating
time and reduced service commitments.

References ed colectomy versus open colectomy for treatment of

non-metastatic colon cancer: a randomized trial.
1. Vining DJ Virtual colonoscopy. Gastrointest Endosc Lancet 2002;359:2224-9
Clin N Amer 1997;7: 285-91 7. MRC CLASICC Collaborators. Proceedings of the
2. Arulampalam TH, Costa DC, Loizidou M, Visvikis D, Tripartite Colorectal meeting, Melbourne October
Ell PJ, Taylor I. Positron emission tomography and col- 2002
orectal cancer. Br J Surg 2001;88:176-89. 8. Jayne D. Laparoscopic Colorectal Surgery: where to
3. Kapiteijn E, Marijnen CA, Nagtegaal ID, Putter H, after CLASICC? Ass Coloproctology of Great Britain
Steup WH, Wiggers T, Rutten HJT, Pahlman L, and Ireland - Update 2003;4-7
Glimelius B, van Krieken JHJM, Leer JWH, Van de 9. Fried GM, Derossis AM, Bothwell J, Sigman HH.
Velde C. Preoperative radiotherapy combined with Comparison oflaparoscopic performance in vivo with
total mesorectal excision for resectable rectal cancer N performance measured in a laparoscopic simulator.
Engl J Med 2001;345: 638-46 Surg Endosc 1999;13:1077-81
4. Milsom JW, Bohm B, Hammerhofer KA, Fazio VW, 10. Rosser JC, Rosser LE, Savalgi RS. Objective evaluation
Steiger E, Elson P. A prospective, randomized trial of a laparoscopic skill program for residents and
comparing laparoscopic versus conventional tech- senior surgeons. Arch Surg 1998;133:657-61
niques in colorectal cancer surgery: a preliminary 11. Rogers DA, Elstein AS, Bordage G. Improving continu-
report. J Am ColI Surg 1998;187: 46-57 ing medical education for surgical techniques: apply-
5. Wexner SD, Cohen SM, Johansen 0, Nogueras J, Lago ing the lessons learned in the first decade of minimal
CP, Lube MW. Laparoscopic colorectal surgery: a access surgery. Ann Surg 2001:233:159-66
prospective assessment and current perspective. Br J 12. Ballantyne GH. Robotic surgery, telerobotic surgery,
Surg 1993;80: 1602-5 telepresence and telemonitoring. Review of early clin-
6. Lacy AM, Garcia V, Delgado S et al. Laparoscopy assist- ical results. Surg Endosc 2002;16:1389-402
 Invited commentary
G. Melotti

Whether the laparoscopic approach to the resec- Port-site recurrences of 0-1,4% in recently
tion of colorectal cancer has the same benefits of reported series, similar to wound recurrences in
laparoscopic cholecystectomy is controversial laparotomy, stress the importance of an accurate
because the presence of an anastomosis and an technique and a profound skill when performing
almost equal number of operative manipulations such difficult surgery.
in the abdominal cavity, such as regional lym- Excellence of surgical technique is of particu-
phadenectomy, as for an open procedure, may pre- lar relevance in the treatment of rectal cancer
vent early food intake and a rapid recovery after where a complete total mesorectal excision is the
the operation. Thus it has been difficult to demon- mainstay of good disease-free survival. Recent
strate the benefits of laparoscopic approaches to published works show that laparoscopic TME is
colorectal cancer using only the currently available feasible in the majority of patients undergoing an
clinical parameters, such as the day of first mobi- elective resection of middle and low rectal lesions
lization, the day of initial food intake, the length of [4]: cancer clearance, only in part reflected by the
post-operative hospital stay, and the time it takes absolute number of lymph-nodes retrieved in the
to attain rehabilitation to the preoperative level of specimen, can be achieved in almost all cases with
social life. These parameters are not only depen- lateral clearance of the side walls. The role of lym-
dent on the physician's directions, but also affected phatic mapping and biopsy in colorectal cancer,
both by the patient and by a variety of social con- with the yet unknown meaning of micrometas-
ditions. Traditional ways of assessing the success tases identified by a more thorough evaluation of
of a surgical intervention will fall short and suc- the index node, is undergoing intensive study
cess will take on new meaning with the implemen- even in open surgery. As a matter of fact,
tation of quality of life parameters [1]. laparoscopy allows the same identification rate
Nowadays, we can say that technical standard- and possibility oflymph-node retrieval than open
ization of the procedures, with a clearly defined surgery, showing once more, similar feasibility [5].
vascular supply and an established technique for A particular concern in rectal surgery with
anstomosis construction, laparoscopic resection TME is the risk of bladder and sexual dysfunc-
of colorectal cancers can be performed reliably tion, possible consequences of neuropraxia or
and safely, with long-term results comparable to permanent injury to the sympathetic and
open surgery [2]. parasympathetic nerves that course along the lat-
More than 2500 patients have been random- eral pelvic side-walls. Laparoscopy, in the opinion
ized in clinical trials (COLOR, CLASICC etc.) of many Authors, provides better visualization of
comparing conventional open resections with the the operative field in a deep and narrow pelvis. In
mini-invasive approach and results will be avail- addition, the magnified view (10-20 x) may offer
able in 2004-2006, but more and more observa- better precision in performing high ligation of the
tional studies show actuarial5-year survival super inferior mesenteric artery and preservation of
imposable with open surgery, even better for stage autonomic nerves [6]. As a matter of fact, an
III disease [3]. altered ejaculatory function is significantly more
216 Atlas of Endoanal and Endorectal Ultrasonography

frequent after laparoscopic surgery owing to the tive times. However it is increasingly considered a
technical difficulties of the resecting rectal safe oncologic approach and therefore appears as
tumors, particularly of advanced stage [71. a recommended option in colorectal treatment
Even conversion rates are much higher in guidelines, on the condition that the surgeon be
advanced tumors, stressing the need to exclude experienced both in colorectal and laparoscopic
from the laparoscopic approach T4 cancer cases surgery [81.
that do not respond to preoperative radiochemio- A team approach, with special skills in neoad-
therapy [41. This issue, once more, underscores juvant therapies, in accurate preoperative staging
the value of an accurate restaging by endoscopic coupled with a volume of cases sufficient to
ultrasonography after neoadjuvant treatments. acquire and maintain the appropriate skills, are
Laparoscopic resection of rectal cancer is diffi- the only guarantee for good short and long-term
cult, with a steep learning curve and long opera- results.

References 5. Saha S, Bilchik A, Wiese D. Sentinel lymph node biop-

sy in colorectal cancer. In H.S. Cody III, Martin Dunitz,
1. Inoue Y, Kimura T, Noro H et al. Is laparoscopic col- London 2002
orectal surgery less invasive than classical open 6. Sim R, Milsom JW. Laparoscopic resection of the rec-
surgery? Quantification of physical activity using an tosigmoid colon. In F.L. Greene, B. T. Heniford,
accelerometer to assess postoperative convalescence. Springer, New York 2001
Surg Endosc 2003;17:1269-73 7. Quah H, Jayne DG, Eu KW, Seow-Choen F. Bladder and
2. Scheidbach H, Scneider C, Hugel 0 et al. Oncological sexual dysfunction following laparoscopically assisted
quality and preliminary long-term results in laparo- and conventional open meso rectal resection for can-
scopic colorectal surgery. Surg Endosc 2003,17:903-10 cer. Br J Surg 2002;89:1551-6
3. Hazebroek EJ. COLOR: a randomized clinical trial 8. COR-CPO, Regione Piemonte - Linee guida clinico
comparing laparoscopic and open resection for colon organizzative sui tumori del colon-retto. Assessorato
cancer. Surg Endosc 2002;16:949-53 Sanita Regione Piemonte 2001
4. Morino M, Parini U, Giraudo G et al. Laparoscopic
total mesorectal excision. A consecutive series of 100
patients. Ann Surg 2003;237:335-42
 VIII.6.
Intraoperative laparoscopic liver ultrasonography
G.A. Santoro, G. Di Falco

Approximately 20 to 25% of patients with rectal below the surface of the liver. Furthermore, it is
cancer have hepatic metastases at the time of the difficult, sometimes even impossible, to obtain
presentation. Assessment for coincident metastatic biopsies from deep hepatic lesions and lymph
disease is a fundamental component of oncologic nodes with conventional laparoscopy. The intro-
surgery and may lead to major modifications in duction of laparoscopic ultrasound (LUS) has
the surgical tactics for primary disease and in been shown to give important clinical informa-
postoperative management. The search for liver tion, especially in the screening for hepatic metas-
metastases before surgery therefore forms an tases during laparoscopic rectal cancer surgery [3].
accepted part of colorectal surgical practice. The
combination of bimanual palpation of the liver at
laparotomy with intraoperative ultrasound (IOUS) Instrumentation
is particularly helpful in evaluating hepatic lesions
seen on preoperative imaging studies that may be
metastatic tumors. Most series reported that IOUS Several types of LUS probes are currently being
detected otherwise unrecognized liver metastases used. Most devices have small linear-array trans-
in 5-10% of patients with colorectal carcinoma [1- ducers mounted on rigid or flexible probes. These
2]. The recent development of laparoscopic col- laparoscopic probes are designed to be inserted
orectal resection for malignancy poses some prob- through a standard laparoscopic port of 10-11 mm
lems, because during this type of surgery, such an diameter. B-K Medical's type 8666 transducer (B-
intraoperative assessment of the liver is not possi- K Medical AlS, Mileparken 34, DK-2730 Herler,
ble. For example, it is not possible for the surgeon Denmark) (Fig. VII1.63) is an example of the
to gain tactile information, nor it is possible to see instrumentation available for laparoscopic ultra-

a ~==o:;;:;;:;;=~===
Fig. VIII.63. (a) B-K Medical's type 8666 transducer, designed for laparoscopic ultrasound examinations. (b) Convex multifre-
quency electronic transducer
 218 Atlas of Endoanal and Endorectal Ultrasonography

a b

c d

Fig. VIII.64. Using controls on the transducer's handle (a),

the tip of the transducer can be moved up to ninety degrees in
e each of four directions: up (b), down (c), right (d) and left (e)

sound scanning of the liver. This device has a where it is needed and to explore any part of the
small convex-array electronic transducer mount- liver at varying depths and angles.
ed on a 10 mm diameter flexible probe with a shaft The transducer also supports Doppler and
length of 30 cm. Its tip can be moved up to ninety color flow imaging, making it easy to distinguish
degrees in each of four directions - up, down, left between vascular and biliary structures. The color
and right (Fig. VIII.64) - and can be fixed at any Doppler signals are shown as an overlay, coded in
desired angle using controls on the transducer's shades of red for flow towards and blue for flow
handle. This, together with the transducer's three away from the transducer, though a variety of
scanning frequencies (s MHz, 6.S MHz and 7.S other color maps may be selected (Fig. VIII.6S).
MHz), makes it possible to place the tip exactly The scanner can provide simultaneous imaging
 Section VIII' Surgical treatment options for rectal cancer 219

and Doppler (known as "duplex scanning" from

their operations in two modes) so that the
Doppler-sensitive gate can be positioned precisely
over the vessel of interest (Fig. VIII.66). Normal
flow in the portal vein gives a continuous signal
above the baseline, indicating flow into the liver.
There may be slight phasic variations correspond-
ing to transmitted cardiac and respiratory pressure
changes (Fig. VIII.66). Normal flow in the hepatic
artery gives a typical biphasic signal (Fig. VIII.67).
Dedicated equipment for LUS-guided biopsy
makes it possible to obtain histological or cyto-
logical confirmation of metastases and to per-
form LUS-guided interventional procedures such
Fig. VIII.6S. Laparoscopic intraoperative image of a trans-
verse section of normal liver with the color-flow Doppler in as radio frequency ablation (Fig. VIII.6S) [4-5]. A
use. The right portal venous system is shown in tints of red in biopsy attachment to the probe with a needle
scan. (IVC = inferior vena cava) channel constrains the flexible needle along a
path that runs diagonally across the scan area; its
line is indicated on the screen and once this has
been positioned over the lesion, the needle should
pass in the correct direction (Fig. VIII.69).
New technology involves reconstruction of
two-dimensional (2D) images into a 3D image
generated by a computer (Fig. VIII.70).

Technique

LUS is performed during the initial phase of the

laparoscopic operation. The ultrasound probe,
Fig. VIII.66. Color Doppler of normal portal vein. In this controls and electronic cord are covered with ster-
duplex spectral image the position of the Doppler-sensitive
gate is indicated by the white bars on the scan (arrow), while
ile sheaths (Fig. VIII.71). Probe cover must be
the spectral trace itself gives a continuous signal above the filled With 3 to 5 ml of sterile saline solution to
baseline, indicating normal flow in the portal vein establish acoustic coupling at the transducer sur-
face. To perform diagnostic laparoscopy, two
ports are placed through the abdominal wall: the
laparoscope is inserted via a 10 mm periumbilical
port while the ultrasonographic probe is placed in
a 12 mm right subcostal port (Fig. VIII.72). The
patient is tilted in the reverse Trendelenburg posi-
tion. The operator generally stands to the right
side of the patient (Fig. VIII.73a). This will allow
the examiner to have full access to controls on the
scanner (Fig. VIII.73b). In addition some surgeons
prefer to stand between the legs of the patient
during laparoscopic liver ultrasound (Fig.
VIII.74). The ultrasound image can be observed
simultaneously with the laparoscopic camera
Fig. VIII.67. Color Doppler of normal hepatic artery. In this image using picture-in-picture video equipment.
duplex spectral image the position of the Doppler-sensitive Each survey of the liver should completely
gate is indicated by the white bars on the scan (arrow), while
assess the entire liver parenchyma. In the 1950'S,
the spectral trace itself gives a biphasic signal, indicating nor-
mal pulsatile flow in the hepatic artery Couinaud refined the functional anatomy of the
 220 Atlas of Endoanal and Endorectal Ultrasonography

J b
Fig. VIII.68. B-K Medical's type 8666 transducer with biopsy system (a). A flexible needle for histological biopsy can be insert-
ed through the dedicated puncture guide (b)

liver and demonstrated that it was divided into

four sectors, each consisting of two segments.
Functionally, then, the liver is now understood to
be divided along vertical and oblique planes
defined by the three main hepatic veins, and along
a transverse plane that follows a line drawn
through the right and left portal branches (Fig.
VIII.7S). The four traditional segments (right
anterior, right posterior, left medial and left poste-
rior) have been replaced by four sectors (right
anterior, right posterior, left anterior and left pos-
terior), each of which is further divided into two
segments by the transverse line mentioned (Fig.
VIII.76). The eight segments of the liver are num-
bered clockwise in the frontal plane. Each seg-
Fig. VIII.69. LUS-guided biopsy. The needle is monitored
ment is an independent functional unit supplied
continuously as it is advanced into the lesion by a single portal triad.
The standard full liver scan should commence
from the right lobe as high up and as close to the
diaphragm as possible and extend caudally onto

b
Fig. VIII.70. 3D Power Doppler image (a) and 3D Power Doppler image with tissue subtraction (b)
 Section VIII- Surgical treatment options for rectal cancer 221

the inferior surface of the liver and toward the

left side (Fig. VIII.n). Marchesa et al. [3]
described a methodical scanning technique to
completely assess the liver parenchyma (Fig.
VII1.78). In position 1, starting just to the right of
the falciform ligament, as dorsally as possible,
the ultrasonographer detects the inferior vena
cava and the three main hepatic veins and scans
segments I, IVa and IVb (Fig. VII1.79). In position
2 the right portal venous and arterial system, the
right hepatic ducts and the parenchymal seg-
ments VIII and V can be identified. In position 3,
the operator moves the probe tip laterally to
Fig. VIII.71. LUS probe covering. The sterile sheath covers
achieve a clear evaluation of the distal branches
both the probe and shaft as well as the controls and electrical of the right hepatic vein, the right portal system
cord connection

M
Fig. VIII.n. Position of ports for ultra-
sonography during laparoscopic colorec-
tal resection (a). Diagram showing the
laparoscopic probe inserted via a right
subcostal port while the laparoscope is
b L-____ ~L- __________________________________ ~ placed in a periumbilical port (b)
222 Atlas of Endoanal and Endorectal Ultrasonography

b
Fig. VIII.73. Position of operating team and equipment for intraoperative laparoscopic ultrasound (a). Examiner stands on the
patient's right side to have full access to controls on the scanner (b)

1st Assist nt
Surgeon

Fig. VIII.74. Diagram showing position

of operating team and equipment for
intraoperative laparoscopic ultrasound.
Surgeon stands between the legs of the
patient
 Section VIII' Surgical treatment options for rectal cancer 223

Fig. VIII.7S. The eight segments of the

liver

a ~ ________________________________- J ~ ________________________________- J b

Fig. VIII.76. (a) Anterior and posterior; (b) Inferior and superior views of the liver, showing the eight functional segments

and segments VI and VII. In position 4 the exam- nective tissue framework. Small round or tubular
iner places the probe tip just to the left of the fal- transonic structures are usually visible within this
ciform ligament to detect the left hepatic and framework and should not be confused with
portal venous system and to scan segments II pathological changes such as metastases. They
and III (Fig. VIII.8o). Finally the probe can be represent small veins, either portal branches or
placed beneath the right and left lobes, respec- tributaries of the hepatic veins. Portal branches
tively, to scan the parenchyma from the inferior are easily located at the hilum and are surround-
surface. ed by a band of stronger echoes (Fig. VIII.79). The
Normal liver exhibits a uniform pattern of lin- hepatic veins are easily identified in the sub-
ear echoes arising from small vessels and the con- diaphragmatic region (Fig. VIII.8!).
224 Atlas of Endoanal and Endorectal Ultrasonography

Fig. VIII.77. Methodical scanning tech-

nique to completely assess the liver
parenchyma. The probe is gently glided to
the right and left over the liver surface
using clockwise rotations

Fig. VIII.78. Probe positions on liver surface. (a) Position I. (b) Position II. (c) Position III. (d) Position IV. The shadowed areas
indicate segments scanned in each position (Redrawn and modified from [3])
 Section VIII' Surgical treatment options for rectal cancer 225

Fig. VIII.SO. Probe is in position IV just to the left of the fal-

Fig. VIII.79. Probe is in position II to scan parenchymal seg- ciform ligament to scan parenchymal segments II and III.
ments VIII and V. The ultrasound image and the laparoscopic Distal branch of the left portal vein (PV) is seen as echo-free
image are both projected in the video monitor (picture-in- tubular structure with prominent wall
picture). The normal liver's uniform stippled texture of mid
grey level is shown. Portal vein (PV) is seen as an echo-free
tubular structure with a prominent wall while the thin walls of
the hepatic vein (HV) are not apparent. (lve = inferior vena
cava, HA= hepatic artery)

Clinical applications detected 21 additional tumors (9.5%) in 11 patients

(20.0%). These tumors missed by CT ranged in
The role of laparoscopic ultrasound in the intra- size from 0.3 to 2.7 cm. Small tumors tended to be
operative staging of rectal carcinoma is still being missed by CT scan (28.6% of the lesions <1 cm,
assessed, but it may be of great value for detecting 15.8% of those 1-2 cm,4% of those 2-3 cm and 0%
and assessing liver metastases. of those >3 cm), as did those in the segment III
The appearance of a metastases at ultrasound and IV. Milsom et al. [7] reported a prospective,
is extremely variable (Fig. VIII.81). Most common- blind comparison of LUS versus contrast-
ly, single or multiple hyperechoic areas are seen. enhanced computerized tomography for liver
Alternatively a well-defined hypo echoic area sur- assessment in patients undergoing laparoscopic
rounded by normal parenchyma is found often colorectal cancer surgery. They found a higher
with fine echoes within. A third pattern is that of yield for detecting liver lesions using LUS com-
a "target" lesion in which an echogenic central pared with the preoperative CT (94.6% vs. 78.4%).
area is surrounded by an echo-poor halo. A fourth Two patients with negative preoperative CT scans
group, with ultrasonic features of mixed patterns, underwent treatment of their metastatic lesions
diffuse irregularity or poorly defined masses is based on the information provided by the LUS.
more difficult to characterize. Another recent work by Rurtley et al. [8] has
Most studies emphasize that the additional shown LUS to be superior to extracorporeal ultra-
lesions disclosed by LUS are generally quite small, sound (94.2% vs. 88.1%) and as effective as mag-
frequently less than 1 cm in diameter and there- netic resonance imaging (94.2% vs. 90.7%) in the
fore below the limit of resolution of most preop- detection of colorectal hepatic metastases. The
erative imaging devices. Foroutani et al. [6] use of LUS, however, did not alter the manage-
reported that the LUS detected all 201 tumors seen ment of patients within this study.
on preoperative computed tomography (CT) and
 226 Atlas of Endoanal and Endorectal Ultrasonography

a b

Fig. VIII.81. Liver metastases from a rectal carcinoma(a). The echogenicity of the lesions (arrows) is not markedly different
from the normal liver. The thin hypoechoic rim makes it possible to distinguish the lesion from the adjacent liver. The hepatic
veins (hv) are well visualized. (b) Macroscopic view of the specimen

References Percutaneous radio-frequency ablation of hepatic

metastases from colorectal cancer: long-term results
1. Kruskal JB, Kane RA. Intraoperative ultrasonography in 117 patients. Radiology 2001;221:159-66
of the liver. Crit Rev Diagn Imaging 1995;36:175-226 6. Foroutani A, Garland AM, Berber E, String A, Engle K,
2. Takeuchi N, Ramirez JM, Mortensen NJ, Cobb R, Ryan TL, Pearl JM, Siperstein AE. Laparoscopic ultra-
Whittlestone T. Intraoperative ultrasonography in the sound vs. triphasic computed tomography for detect-
diagnosis of hepatic metastases during surgery for col- ing liver tumors. Arch Surg 2000;135:933-8
orectal cancer. Int J Colorect Dis 1996;11:92-5 7. Milsom JW, Jerby BL, Kessler H, Hale JC, Herts BR,
3. Marchesa P, Milsom JW, Hale JC, O'Malley CM, Fazio O'Malley CM. Prospective, blinded comparison of
VW. Intraoperative laparoscopic liver ultrasonography laparoscopic ultrasonography vs. contrast-enhanced
for staging of colorectal cancer: initial experience. Dis computerized tomography for liver assessment in
Colon Rectum 1996;39:S73-8 patients undergoing colorectal carcinoma surgery. Dis
4. Durup Scheel-Hincke J, Mortensen MB, Pless T, Colon Rectum 2000;43:44-9
Hovendal C. Laparoscopic four-way ultrasound probe 8. Hartley JE, Kumar H, Drew PJ, Heer K, Avery GR,
with histologic biopsy facility using a flexible tru-cut Duthie GS, Monson JRT. Laparoscopic ultrasound for
needle. Surg Endosc 2000;14: 867-9 the detection of hepatic metastases during laparo-
5. Solbiati L, Livraghi T, Goldberg SN, lerace T, Meloni F, scopic colorectal cancer surgery. Dis Colon Rectum
Dellanoce M, Cova L, Halpern EF, Gazelle GS. 2000;43:320-5
 VIII.7.
Robotic rectal resections
A. D'Annibale, E. Morpurgo

Laparoscopy has many differences compared to Description of the Da Vinci® system

traditional open surgery, and it is considered
more difficult and technically more demanding.
For this reason surgeons may be reluctant to use The Da Vinci robotic surgical system (Intuitive
laparoscopy for complex procedures like colorec- Surgical Inc. Sunnyvale, CA USA) has three com-
tal resections [1-5]. ponents. The first is the vision cart, which holds
The major pitfalls of laparoscopy are: the dual light source and image processor for the
two three-chip cameras that are installed on the
1) loss of three-dimensional view; single endoscope, which provides a 3D image for
2) loss of dexterity, due to several factors: straight the surgeon. The second component is the master
instruments have a fulcrum at the port entry console where the operating surgeon sits (Fig.
and consequently motions are reversed; loss of VIII.82). The console contains the computers that
the wrist-like movements due to the impossibil- process the combined images into a true 3D
ity of performing intra-abdominal articulation image. The master control grips drive the robotic
of instruments tips; the length of the instru- arms and the instruments. Foot pedals at the con-
ments may amplify the physiologic hand sole provide control for the electrocautery, as well
tremor; as a clutch mechanism to drive the camera arm.
3) camera is held by an assistant or by a nurse The last component of the system is the surgical
and can therefore be unstable and not always cart consisting of the three robotic arms mounted
be focused where the surgeon wants [6]. on a movable chassis (Fig. VIII.83). Three manip-
ulators (arms), which are covered with sterile
In order to overcome these disadvantages the drapes for the procedures, are mounted on a cen-
possibility of using robotic systems has been pro- tral column that is placed on the wheeled surgical
posed and with the advancement of technology cart. Two arms are designed to hold the surgical
telerobotic systems have started to playa role in tools and respond to the movements of the sur-
surgery. The new generation of robotic instru- geon's hands, and one is designed to hold the
ments has been designed in order to overcome all three-dimensional stereo endoscope and camera.
the pitfalls of traditional laparoscopic instru- The arms have three degrees of freedom (pitch,
ments: a) they offer a tridimensional view; b) they yaw and insertion); they allow the responsible
hold the camera with a stable platform and the instrument to move two additional degrees of
camera can be moved directly by the surgeon; c) freedom in the wrist and provide two additional
they translate exactly the movements of the sur- motions for tool actuation. The surgeon looks
geon's hand to the tip of the instruments, that down into the viewer as if looking into the surgi-
have a wrist-like articulation. In addition, the sur- cal field. Holding the control handles he carefully
geon operates while sitting in a comfortable posi- guides the tools mounted at the robotic arms
tion. For these reasons more complex surgical inside the patient's body. As the surgeon moves
procedures can potentially be accomplished with the manipulators on the console, manipulators
a minimally invasive approach. closely follow the input motions. A vision cart has
 Invited commentary
J.B. Kruskal

Laparoscopic ultrasound of the liver is being used capsular locations, and adjacent to the portal
with increasing frequency for staging of tumors in veins. The operator should also be aware that dif-
the liver, for characterization of liver lesions, and ferent types of lesions can co-exist in the liver.
for facilitating surgical resection when these are For instance, identification of lesions with differ-
identified. The study also extends beyond the liver ent appearances should induce a careful evalua-
and is important for identifying portal and tion of each lesion since benign entities, such as
retroperitoneal lymphadenopathy. The learning hemangiomas, frequently coexist with metas-
curve for use of laparoscopic ultrasound is long tases. Certain specific characteristics of rectal
and operators should be aware that the probes tumors may be evident. Some mucinous tumors
need to be separately inserted on both sides of the of the rectum or colon may produce stippled cal-
falciform ligament in order to obtain adequate cifications in the liver, facilitating characteriza-
visualization of both the left and right lobes of the tion of metastases.
liver. Certain areas of the liver are traditionally The laparoscopic ultrasound exam is not just
more difficult to image with laparoscopic ultra- diagnostic, but can also be used for guiding pro-
sound. These include the bare area on the posteri- cedures, such as biopsy, and cryo- or radio fre-
or right as well as the high dome of the right lobe quency ablation of these tumors. In addition, the
of the liver. The transducer needs to be placed on laparoscopic study can be used to guide resection
the naturally moistened surface of the liver and of tumors, in which setting the operator is expect-
carefully maneuvered in a manner to allow ade- ed to provide accurate localization of lesions
quate visualization of all segments of the liver. within different segments of the liver and to
Many patients with rectal or anal cancers are in define important vascular anatomy. This would
chemotherapy, which is associated with hepatic include identification of replaced or accessory left
steatosis or fatty replacement. In the setting of or right hepatic arteries, which need to be identi-
focal fatty replacement or sparing, the intraoper- fied to prevent complications such as bleeding or
ative laparoscopic study can be especially useful ischemia to the remaining liver or if the surgeons
for distinguishing these areas from intrahepatic are planning to insert intraarterial chemothera-
metastases. Typically, areas of focal fatty sparing peutic pumps. Laparoscopic ultrasound is being
or replacement are not rounded, but have a geo- used with increasing frequency by both radiolo-
graphic appearance, are soft and typically com- gists and surgeons, and it is expected to gain an
pressible with the ultrasound transducer, and even more prominent role both in the staging of
may have vessels traversing their center. In addi- rectal tumors and in facilitating their intraab-
tion, fatty replacement or sparing is also more dominal and intrahepatic identification. This
likely to occur in specific locations such as adja- chapter has provided a useful and practical intro-
cent to the gallbladder fossa, in peripheral sub- duction to this technique.
 Section VIII' Surgical treatment options for rectal cancer 229

Fig. VIII.82. Master console. The sur-

geon works using the handles to move the
robotic arms and instruments and the
pedals to move the camera

Fig. VIII.83. The Da Vinci robotic arms

during the dissection of the lower rectum

a conventional two-dimensional monitor for the components of the system in the room is crucial.
assistant, two camera light sources and a trans- As a general rule, the robotic arms are placed on
mitter for the three-dimensional camera. the same side of the patient as the lesion.
Therefore for surgery of the left colon and rectum
the arms are placed on patient's left side. The
Room preparation and instrument vision cart is positioned at the patient's feet so that
the assistant and the scrub nurse have an optimal
set-up view. The master console is on the patient's right
side, about 10 feet away. This allows enough space
The Da Vinci robotic device is a heavy and bulky for the robotic arms, the scrub nurse and the assis-
instrument and a large operating room is needed. tant to move. Furthermore, from this position the
In order to minimize the need to move the device surgeon has a complete visual control of the surgi-
in the room, the proper positioning of the three cal field and of the robotic arms (Fig.VIII.84). All
230 Atlas of Endoanal and Endorectal Ultrasonography

VISIoN
-.
It • -t
1
CART

Fig. VIII.84. Position of the console, the robotic arms, the

vision cart and of the surgeons in the operating room

Fig. VIII.8S. Port placement for the dissection of the left

the instruments are prepared in the room before colon. Robotic arms work through ports 2 and 3; the assistant
works through port 4. The camera is in port 1
the patient's arrival. After anesthesia is induced
and the patient is monitored, the robotic arms are
wrapped with sterile plastic sheets and are even-
tually approached to the operating table.

Port placement

Optimal port and patient position is also crucial

because of the difficulty of moving the device after
it has been positioned. Rectal resections are usual-
ly performed in two steps (the take down of the
splenic flexure with the dissection of the left colon
and the dissection of the rectum). For the first step
of the operation the positions of the ports are
shown in Fig. VIII.8S. The camera is positioned in
port access 1, and the surgeon works with robotic •.1
instruments in port 2 and 3. Port 3 is at the right
limit of the pfannestiel incision that will be even- . ~
tually needed to exteriorize the specimen and to ,,- .4-t
insert the anvil of the stapler. Port 4 is used by the 3
assistant to insert the clip applier or to help with
any additional retractor. For the dissection of the
rectum the robotic device has to be moved to the
left coxal of the patient and one additional access
is needed (Fig. VIII.86). The robot works now with
port access 4 and S to dissect the rectum and the
Fig. VIII.86. Port placement for the dissection of the lower
assistant works through port 3. From this access rectum. A fifth trocar is needed. Robotic arms work through
the assistant will introduce the stapler to staple the port 4 and 5. The assistant works though port 3. The camera is
rectum. The site of port S will be used to exterior- in port 1
 Section VIII' Surgical treatment options for rectal cancer 231

ize the terminal stoma if a Miles procedure is per-

formed and therefore its location has to be chosen
very carefully preoperatively.

Surgical technique

The patient is supine with the legs open, in the

Trendelemburg position, rotated to the right. This
position allows the small bowel to fall in the right
upper abdominal quadrant, so that the left colon
with its mesentery can be visualized. The robotic
arms are first positioned at the patient's left side Fig. VIII.88. Initial dissection of the inferior mesenteric vein
(Fig. VIII.87). The inferior mesenteric vein is iden-
tified at the ligament of Treitz (Fig. VIII.88). After
the vein has been clipped, it is lifted together with
the mesentery of the left colon and it is sharply dis- The vessels can be cut with scissors or with a stapler
sected from the fascia of Gerota and from the after the mesentery of the left colon is completely
retroperitoneum. Starting from the right iliac dissected from the retroperitoneum and from the
artery, the pre aortic area is freed and the inferior ureter and the gonadal vessels (Fig. VIII.89).
mesenteric artery is identified and isolated. The Starting from the arell: of dissection of the infe-
nerves of the aortic plexus are identified and pulled rior mesenteric vein, the mesentery of the distal
downwards. The identification and isolation of part of the transverse colon is divided from the tail
both the vessels is greatly facilitated by the wrist- of the pancreas. This maneuver allows the gas to
like movement of the tips of the instruments that enter the lesser sac and will greatly help the dis-
easily allows to go behind the origin of the vessels. section of the omentum from the transverse colon.
Once this step is accomplished, the take down of
the splenic flexure is completed. The take down of
the splenic flexure is a crucial part of the left hemi-
colectomy and of the anterior resection of the rec-
tum. This procedure is again facilitated by the high
degree of movement of the tips of the instruments
that facilitate the dissection of the splenic flexure
from behind, which is usually difficult with regular
laparoscopic instruments. When the splenic flex-

Fig. VIII.87. Position of the robotic arms during dissection of

the left colon and during take down of the splenic flexure
Fig. VIII.89. The inferior mesenteric vein and artery have
been clipped and cut
232 Atlas of Endoanal and Endorectal Ultrasonography

ure is taken down, the rectum with its mesorectum

is divided from the presacral fascia, and the
hypogastric nerves are recognized and spared.
Due to the fact that the rectum is a medial organ,
if the tumor is located in the lower third of the rec-
tum, in order to allow the lower dissection, the
robotic arms have to be moved downwards, to the
patient's left thigh (Fig. VII1.90) and a fifth trocar
is needed. This facilitates the dissection of the rec-
tum up to the levator of the anus (Fig. VIII.91).
Once the rectum is freed, it is stapled by the assis-
tant through port 3. A small pfannestiel incision
including the port site 3 is performed and after
wound protection the left colon is exteriorized and Fig. VIII.91. The rectum (held by the grasper) is dissected
is divided. The anvil of the stapler is inserted into from the seminal vesicles
the descending colon, the abdominal incision is
sutured and the pneumoperitoneum is re-estab-
lished. The trans-anal stapled anastomosis is then
performed laparoscopically. If the suture line
needs to be reinforced with sutures, the Da Vinci is
used again. The high dexterity of the instrument
tips greatly helps any kind of intra-abdominal
suture that will otherwise be impossible with stan-
dard laparoscopic instruments. If a Miles opera-
tion is required, the rectum is exteriorized from
the perineal incision and therefore the mini-
laparotomy is not needed. The terminal stoma is
located at the site of port 5 (Fig. VIII.92).

Fig. VIII.92. End of operation after robotic Miles procedure.

The stoma is located at the site of port 5

Conclusions

Our experience is based on 194 cases performed

with the Da Vinci robotic system from May 2001
to July 2003. Of these, 58 are resections of the rec-
tum or recto-sigmoid colon. We can conclude that
surgical steps of robotic resection of the rectum
do not differ from conventional laparoscopic
techniques. The robotic device is designed in
order to enhance the surgeon's skills, which are
limited by the traditional laparoscopic instru-
ments [7]. With the Da Vinci system some maneu-
vers are greatly facilitated during anterior resec-
tions of the rectum, the take down of the splenic
flexure, the dissection and isolation of the inferior
Fig. VIII.90. Position of the robotic arms during dissection of
mesenteric vessels, the dissection of a narrow
the lower rectum pelvis and the over suture of the anastomosis.
 Section VIII· Surgical treatment options for rectal cancer 233

Acknowledgements

We thank Nicoletta Vettore RN, for her help with

drawing preparation.

References robot-assisted laparoscopic surgery. Surg Laparosc

Endosc Percutan Tech 2002;12:41-5
1. Ballantyne GH. Robotic surgery, telerobotic surgery, 5. Weber PA, Stephen M, Wasielewski A, Ballantyne GH.
telepresence, and telementoring. Surg Endosc 2002; Telerobotic-assisted laparoscopic right and sigmoid
16:1389-402 colectomies for benign disease. Dis Col Rectum
2. Cadiere GB, Himpens J, Germay 0, Izizaw R, Degueldre 2002;45:1689-96
M, Vadromme J, Capelluto E, Bruyns J. Feasibility of 6. Kim VB, Chapman WHH, Albrecht RJ, Marcus Bailey
robotic laparoscopic surgery: 146 cases. World J Surg M, Young JA, Wiley Nifong L, Chitwood WR. Early
2001;25:1467-77 experience with telemanipulative robot-assisted
3. Chapman WHH, Albrecht RJ, Kim VB, Young JA, laparoscopic cholecystectomy using da Vinci. Surg
Chitwood WR. Computer-assisted laparoscopic Laparosc Endosc Percutan Tech 2002;12:33-40
splenectomy with the Da Vinci surgical robot. J 7. Marescaux J, Smith MK, Folscher D, Jamali F. Telerobotic
Laparoendosc Adv Surg Tech 2002;12:155-9 laparoscopic cholecystectomy: initial clinical experi-
4. Ruurda JP, Broeders IAMJ, Simmermacher RPM, Borel ence with 25 patients. Ann Surg 2001;234:1-7
Rinkes IHM, Van Vroonhoven TJMY. Feasibility of
 Invited commentary
J. Marescaux, F. Rubino

The technical challenges that laparoscopy poses not robotic assistance does deliver its potential
to surgeons mostly depend on the physical sepa- advantages of increasing dexterity and enabling
ration between the surgeon and the tissues and stabilization of surgical maneuverss. Or, perhaps,
organs. This requires some degree of adaptation we should formulate the question in another way,
to having instruments whose fulcrum lies far which is: is the enhanced dexterity and stabiliza-
from their tips as well as the need to reconstruct a tion afforded by surgical robots really improving
three-dimensional anatomic situation from a two- our performance and outcomes in general
dimensional image on a video-screen. surgery?
This limitation, however, is at the same time the We are afraid we have no answer yet to this
truly revolutionary aspect of laparoscopy as this question.
physical gap gives access to robotic and computer Indeed, whereas clinical and experimental
interfaces. Robotic systems include computers that studies have demonstrated feasibility and safety
digitize the surgical movements and images. Once of robotic assistance without evidence of specific
digitized, this information can be modified to fil- complications, it is a common finding that the use
ter and exclude non-finalized movements (i.e. the of robot assistance usually increases the operative
surgeon's physiologic tremor) [11 potentially time of procedures. Furthermore, all clinical stud-
resulting in greater dexterity and precision for the ies performed so far failed to provide any evi-
performance of difficult tasks [21. dence of specific patient benefits in general surgi-
Since the mid-nineties robotic assistance has cal procedures.
been used for virtually any type of abdominal For instance, at the 2003 SAGES Meeting, the
operation. In this issue of the Atlas of Endoanal data of a multicentric prospective randomized
and Endorectal Ultrasonography, D' Annibale and study performed in the USA were presented [3, 41.
Morpurgo describe their technique of robotic The study did actually compare robotic assisted
assistance for laparoscopic rectal surgery using cholecystectomy and Nissen fundoplication using
the "Da Vinci" surgical system. They suggest that the Zeus system to the respective procedures per-
the use of the robot may overcome the difficulty formed by conventional laparoscopy. In both
of performing complex tasks laparoscopically, operations, robotic assistance was performed
such as the mobilization of the splenic flexure and with significantly longer operative time than in
dissection deep in the pelvis, as well as help in conventional laparoscopy, however, complica-
over sewing the anastomosis in a narrow space. tions, blood loss and functional outcomes were
Their experience of almost 60 cases is a reliable similar between robotic and conventionallaparo-
validation for the technique they have described. scopic surgery.
This is consistent with the knowledge that robotic There are therefore some limitations that cur-
assistance is feasible and safe for a large number rently hamper the ability of surgeons to demon-
of abdominal operations. strate clear benefit of robotic assistance. However,
However, the question remains on whether or on the other hand, the potential benefits of using
 Section VIII • Surgical treatment options for rectal cancer 235

robotic technology in surgery are such that they Indeed, it is well known that adequate technical
warrant efforts to overcome the current limita- preparation and skills are key in rectal cancer
tions. Rectal surgery is probably one of the best- surgery. In fact, in spite of the improvements
suited fields of application for robotic technology achieved through the use of adjuvant and neoad-
among general surgery operations. This is for a juvant therapies, the surgeon remains an impor-
number of reasons. tant factor in the accomplishment of tumor con-
The laparoscopic approach for colorectal dis- trol and reduction of morbidity [7, 8]. This stress-
eases includes increased patient comfort and ear- es the need for an efficient educational system to
lier return to daily activities [5]. In contrast with achieve a highly standardized standard of care and
early concerns about its oncologic adequacy, robotic technology may represent a great means to
recent literature suggests that laparoscopy can improve surgical education (Fig. VIII.93).
also achieve oncologic outcomes that compare The use of a surgical robot indeed is associat-
favorably with those ofits open counterpart [6]. ed to the concept of distantiation, which,
There are some features of laparoscopy that stretched to its extreme, translates into the field of
make it potentially better than open surgery even remote surgery. Our group recently demonstrated
in cancer patients. One of its major advantages that current telecommunication technology
lies in the magnification that is offered by the allows performance of a surgical operation
endoscopic camera allowing greater precision and through transoceanic distances [9]. This has great
better identification of critical structures. For educational potential as expert assistance can be
instance, during rectal surgery and total mesorec- ensured much more easily and reach virtually any
tal excision, this may help in the preservation of hospital. The possibility of having active inter-
the nervous plexus. Furthermore, the ability to vention from an expert at a distance could boost
perform deep pelvic dissection in full view of the the application of minimally invasive techniques
whole operating team along with the magnifica- for those advanced procedures, such as laparo-
tion offered by the scope will undoubtedly accel- scopic rectal cancer surgery, that young inexperi-
erate the teaching of rectal cancer surgery and enced surgeons would be reluctant to undertake.
possibly lead to a better standardization of the Furthermore, robotic systems are truly informa-
surgical procedure. Moreover, laparoscopic col- tion systems, where computer interfaces are
orectal surgery allows the monitoring and the placed between the surgeon and the patient.
recording of the differences in operative tech- Under this perspective, the future developments
niques, so that standardized surgery can be docu- of computer technologies and their applications
mented photographically and videos used to might result in significant enhancement of per-
prove adherence to strict oncologic criteria. formance of surgical robots. Current technology,
Laparoscopic rectal surgery, however, is a indeed, already allows 3D reconstruction of
technical challenge to the surgeon who is novice anatomical and pathological structures; transla-
to this approach and specific training and skills tion of medical information contained in images
are needed to achieve all the possible benefits of into a set of 3D models allows the development of
the laparoscopic approach. Dissecting deep in a a very real interaction of virtual instruments with
narrow pelvis becomes a difficult task with the the virtual organs. By simulating a surgical proce-
classic limitation of laparoscopic instruments, dure in a virtual environment one can perform it
which have only four degrees of freedom and the several times and rehearse it till the best possible
amplification of the physiologic tremor of the performance is delivered. In theory, the digital
surgeon's hand. The need for a human assistant to data of the selected best-performed procedure
hold and move the camera in a small space adds can be stored and transferred to a robotic system
further disadvantages; fatigue, tremors, orienta- to be automatically reproduced on the actual
tion errors and an unstable camera may compro- patient, with obvious advantages.
mise the smoothness of the operation. Robotic Although this may sound as too far away from
technology may diminish the degree of technical now, we have learned a lesson from seeing the
difficulty linked to the reasons mentioned above, pace by which technology advances and how this
especially when all the new and useful energy affects surgery. For instance, a few months before
directed instruments will be available at the tip of the first transatlantic operation was performed,
the robotic arms. But this is not the only advan- authoritative reviews on the topic of distant
tage. surgery clearly stated that remote surgery would
 236 Atlas of Endoanal and Endorectal Ultrasonography

b
Fig. VIII.93. Laparoscopic (a) and robotic (b) surgical training at IRCAD - European Institute of Telesurgery, Strasbourg,
France

never be possible for distances greater than a few logical devices nowadays, then we should not feel
hundred miles [10 J. If we just think of how rapid- surprised if during our lifetime we see automated
ly our daily life is revolutionized by new techno- robotic surgery become a reality.
 Section VIII- Surgical treatment options for rectal cancer 237

References 5. Braga M, Vignali A, Gianotti L, Zuliani W, Radaelli G,

Gruarin P, Dellabona P, Di Carlo V. Laparoscopic ver-
1. Garcia-Ruiz A, Gagner M, Miller JH, Steiner CP, Hahn sus open colorectal surgery: a randomized trial on
JF Manual vs. robotically assisted laparoscopic surgery short-term outcome. Ann Surg 2002;236:759-66
in the performance of basic manipulation and sutur- 6. Lacy AM, Garcia-Valdecasas JC, Delgado S, Castells A,
ing lasks. Arch Surg 1999;133,957-61 Taura P, Pique JM, Visa J. Laparoscopy-asslsted colec-
2. Reichenspurner H, Boehm D, Reichart B. Minimally tomy versus open colectomy for treatment of non-
invasive mitral valve surgery using three-dimensional metastatic colon cancer: a randomized trial. Lancet.
video and robotic assistance. Semin Thorac Cardiovasc 2002;359:2224-9
Surg 1999;11:235-40 7. McArdle CS, Hole D. Impact of variability among sur-
3. White P, Carbajal-Ramos A, Gracia C, Nunez-Gonzales geons on postoperative morbidity and mortality and
E, Bailey R, Broderick T, DeMaria E, Hollands C, Soper ultimate survival. BMJ 1991;302:1501-5
N. A prospective randomized study of the Zeus robot- 8. Birbeck KF, Macklin CP, Tiffin NJ, Parsons W, Dixon
ic surgical system for laparoscopic anti-reflux surgery. MF, Mapstone NP, Abbott CR, Scott N, Finan PJ,
Proceedings of the SAGES 2003 Meeting, Los Angeles, Johnston D, Quirke P. Rates of circumferential resec-
CA, March 12-15. S027, 81 tion margin involvement vary between surgeons and
4. White P, Carbajal-Ramos A, Gracia C, Nunez-Gonzales predict outcomes in rectal cancer surgery. Ann Surg
E, Bailey R, Broderick T, DeMaria E, Hollands C, Soper 2002;235:449-57
N. A prospective randomized study of the Zeus robot- 9. Marescaux J, Leroy J, Gagner M, Rubino F, Mutter D, Vix
ic surgical system for laparoscopic cholecystectomy. M, Butner SE, Smith MK. Nature 2001;413:379-80
Proceedings of the SAGES 2003 Meeting, Los Angeles, 10. Mack MJ. Minimally invasive and robotic surgery.
CA, March 12-15. S061, 89 JAMA 2001;285:568-72
 SECTION IX
Combined modality therapy
for rectal cancer
 IX.I.
Preoperative combined chemoradiation
and adjuvant therapy for rectal cancer
G.D. Beretta, S. Mosconi, L. Milesi, M.A. Pessi, R. Labianca

Introduction that the risk of residual local disease and local

failure is considerable.
It is estimated that the rate of recovery (surgi-
Rectal cancer is often a curable disease when cal resection with curative intent) is 92% for the
localized. Although adenocarcinoma of the rec- colon and 84% for the rectum. The prognosis is
tum and adenocarcinoma of the more proximal clearly related to the depth of tumor invasion in
colon have a similar histologic appearance, the the bowel wall and to the presence of lymph node
natural history of rectal and colon cancer differs, metastases. Other prognostic factors are the
due to epidemiological, anatomical and patho- degree of differentiation and the histological type
physiological reasons. In high-risk populations (signet-ring cell, oat-cell, mucinous adenocarci-
the incidence of colon and rectal tumors is 2:1, noma in patients <45 years of age) [1,2]. The best
with rectal cancer that tends to be more frequent predictor of local recurrence is the lateral margin
in males (20-50%). (radial, circumferential, tangential) [3]. With tra-
Despite the upper margin of the rectum being ditional rectal resection, positive margins
anatomically located at 15 cm from the anal verge, occurred in 23% of patients and local recurrences
only cancers which are located at 12 cm or less developed in 80% of cases. With total mesorectal
from the anal verge are considered true rectal excision (TME), the rate of local failure decreased
cancers. This anatomical definition is based on to iYo at 5-year follow-up [4].
the analysis of biology and the pattern of failure: Adjuvant and neoadjuvant combined treat-
tumors at or below the peritoneal reflection have ments have been developed for tumor downstag-
a more direct route for metastases in the pelvis, ing and downsizing, in order to increase the rate
which has a more complex lymphatic system, so of cure, recovery and quality of life (Table IX.l).

Table IX.I. Impact of surgical and pathologic variables in rectal cancer: a United States Community and
Cooperative Group Report [5]

Local failure and survival depend on:

No adjuvant therapy can ever Surgeon's operative experience
make good what one has failed Surgeon's operative ability to achieve a negative margin
to do surgically[ 6] Adhesion of tumor to adjacent structures
Regional lymph nodes involvement
The number of involved nodes
242 Atlas of Endoanal and Endorectal Ultrasonography

Table IX.2. Incidence of recurrence after curative rence (T3-T 4, N +) on the bases of pathological
resection [81 specimens, better knowledge of tumor extent at
the time of surgery and the identification of the
S-1O% anatomical areas at major risk of local failure
IS-30% according to the operative findings [10]. On the
3S-4S% other hand, the disadvantages include an
increased volume of small bowel in the radiation
field, a potentially hypoxic post-surgical bed and,
the inclusion of perineal scar in the case of
abdominoperineal resection [11].
Adjuvant therapy: radiotherapy In stage I rectal tumors (TI-2, No) the treat-
alone ment of choice is conservative surgery but in
elderly or high operative risk patients, adjuvant
radiotherapy can control subclinical disease in
As already mentioned, the natural history of rec- the mesorectum, pelvic nodes and tumor bed (for
tal cancer differs from that of colon cancer for T2 or more advanced cancer the risk of nodal
the major risk of the development of local recur- involvement is 20% or higher and additional
rence. radiotherapy ± chemotherapy or standard surgi-
In 1974 Gunderson [7] reported that following cal resection is required) [12].
curative resection for adenocarcinoma of the rec- Historical studies of radiotherapy as comple-
tum many patients developed posterior extension mentary to surgery are listed in Table IX.3 [9, 13-
of pelvic recurrence, with distal sacral and coccyx 18]. In all trials the local control was quite uniform
involvement; local recurrences frequently occurred ranging from 71% to 84%. Statistically it was sig-
in perianastomotic tissue and involved adjacent nificantly superior to surgery alone in the MRCIII
organs. Often there was a late diagnosis. Local fail- trial [18]. Severe toxicity rate approached 18% and
ure was responsible for deaths and preceded dis- late bowel complications, which contributed to
tant metastases. treatment-related death [18] and required second
The incidence of local recurrence depends on surgery, occurred in up to 10% of cases [15,16]. No
the stage of the disease at the time of surgery impact on survival was detected in all trials.
(Table IX.2) [8]. Radiotherapy alone is not considered the stan-
Radiotherapy alone was firstly used in adju- dard adjuvant treatment in rectal cancer, although
vant settings in patients with rectal cancer [9]. To these results belong to a period in which there was
achieve evident advantages with postoperative no improvement in radiation planning and the
radiotherapy it is fundamental to have the best techniques that are now employed minimize
selection of patients with a higher risk of recur- treatment-related complications [19].

Table IX.3. Randomized trials of adjuvant radiotherapy alone following radical surgery for rectal cancer versus
follow-up (pathological stage II-III)

Author Treatment % p % P
(N" ofpts.) LF Survival

Baslev [13] So Gy (244) 19 NS 6S B137 C NS

NSABP R-Ol [141 Follow-uP/46.S Gy (1841184) 24/16 NS 43/41 NS
Bentzen [ls1 Follow-up (2S0) 23 NS 68 B/27 C NS
GITSG [9] Follow-Up/40-48 Gy (So) 2S/16 NS 43/S0 NS
Treurniet [161 Follow-up/so Gy (84/88) 33/24 NS S7/4S NS
Arnaud (EORrC) [17] Follow-uP/46 Gy (88/84) 34/33 NS 61/63 NS
MRCIII [18] Follow-uP/40 Gy (23S/234) 34/21 0.001 46/S2 NS

Pts= patients; %LF= local failure; NS= not significant; Band C = Dukes'B and C patients
 Section IX' Combined modality therapy for rectal cancer 243

Adjuvant combined treatment: Despite the benefit on local recurrence and

survival, combined treatment increases severe
chemoradiotherapy toxicity, particularly related to small bowel
obstruction and causes treatment-related death in
Several randomized clinical trials have shown that 1-4% of patients during the follow-up period.
SFU is a radiosensitizer. This mechanism is due to Despite the achieved advances in preoperative
the inhibition of sub lethal damage rescue treatment modality, the statement published by
through the depletion of thymidylathe synthase NIH in 1991 should still be considered valid but
and the accumulation of cells in the S phase (syn- the choice of which regimen to recommend in the
chronization) [20,21]. clinical setting is controversial. The Intergroup
Radiochemotherapy is a combined modality trial revealed that prolonged infusion of SFU dur-
therapy that became standard in 1991 when a clini- ing irradiation is the best modality for treatment,
cal announcement published by the National it is more effective than bolus and has optimal tol-
Institute of Health (NIH) stated that postoperative erance. The role, the optimal sequence and the
radiation and concurrent SFU may be recommend- period of systemic chemotherapy are still to be
ed for patients with resected stage II and III rectal defined. Recently INTo144 directly compared
cancers [22]. This statement is the result that continuous infusion of SFU with bolus/LV /lev-
emerges from randomized clinical trials conducted amisole. This study showed that any of these three
in United States since the 1980'S. These trials showed arms are acceptable in clinical practice for similar
a significant reduction of local recurrence rate and relapse free survival and overall survival, with dif-
a significant benefit in survival for postoperative ferent toxicity profiles and differing needs for
combined modality therapy (Table IX.4) [9,23-29]. central line [30] (Table IX.S).
In some trials chemoradiotherapy was com- The most recent question is: are adjuvant ther-
pared with follow-up alone [9,26], or with radio- apies (radiation alone, even combined with
therapy alone [9,23] or with chemotherapy alone chemotherapy) always needed after total
[9, 2S]· In some of these studies sequential mesorectal excision (TME)? TME decreased the
chemotherapy was compared with concomitant local recurrence rate to 7% and increased the
chemoradiotherapy [9, 23, 2S]. In other trials awareness of the importance of surgical tech-
chemotherapy consisted in different schedules of niques, which is the cornerstone in rectal cancer.
SFU plus methyl-CCNU [24, 28], levamisole or Some physicians suggest that adjuvant radiother-
folinic acid [29]. The Intergroup trial (INT- apy ± chemotherapy is not needed by all patients.
NCCTG, 24) compared continuous infusion of To answer this question the Dutch Colorectal
SFU and concomitant irradiation with SFU bolus; Cancer Group [31] performed a randomized mul-
finally in another trial the optimal sequence of ticenter trial comparing surgery alone (TME tech-
chemotherapy and radiotherapy (advanced or nique) with surgery plus a short course of preop-
delayed) was investigated [27]. erative radiotherapy (500 cGy x S days). Data are
The postoperative combined modality showed preliminary and it is not possible to analyze sur-
an impact on local control and an advantage on vival rates because the follow-up is too short and
survival [9,23]. The rate of local control ranged any benefit should be diluted due to the inclusion
from 83% to 92% and resulted quite uniform in all of early stage disease patients who were also
studies. The benefit on survival was shown in the treated (TI-2No). The authors demonstrated that
GITSG 717S [9] and the NCCTG trial [23]: in the the advantage of preoperative radiotherapy is
former, chemoradiation reached a better survival most evident in stage III patients and in those
rate than surgery alone, in the latter, the combined whose tumor is located less than 10 em from the
modality showed a survival benefit as compared anal verge. The benefit in local recurrence was in
to radiotherapy alone. S.1% of patients treated with surgery alone and in
In Intergroup INT-NCCTG [24] prolonged 16.1% of those treated with RT plus surgery
infusion of SFU (22S mg/m2/d concomitant to (P<0.001), so no adjuvant treatment seems to be
irradiation for S weeks) during irradiation needed.
showed better survival than bolus of SFU. GITSG It is to be stressed that prognostic importance
7180 and Intergroup 0114 [27,28] did not demon- for local failure is high-quality TME and full eval-
strate better outcomes for patients treated with uation of lymph nodes by a pathologist.
methyl-CCNU nor levamisole. Incomplete TME or positive circumferential mar-
244 Atlas of Endoanal and Endorectal Ultrasonography

Table IX.4. Randomized trials of concomitant radio chemotherapy in adjuvant setting

Author Treatment (W of pts.) Results

GITSG 717S [91 Follow-up (S8) RCH: benefit on S (p=O.Ol),

40-44Gy (so) TTR (p=o.OOS) and rate
SFU + Me-CCNU x 18 months (48) of LR (p=0.06) vs. follow-up
4o-44Gy + sFU+Me-CCNU x 18 months (46)

NCCTG 7947S1 [231 4S-so.4Gy (100) RCH: benefit on S (p=O.Q2)

4S-so.4Gy + SFU/sFU+Me-CCNU x 3 months (104) and LR (P=O.OI)

NCCTG 8647S1 [241 SFU x 2 cycles/so.4Gy + SFU/SFU x 2 cycles (219) No benefit with
SFU + Me-CCNU x 2 cycles/so.4Gy + SFU/ Me-CCNU+SFU vs. SFU
SFU + Me-CCNU x 2 cycles (226)

NSABP R-02 [2S] MOF Xs or SFU+FAx 6 cycles (348) RCH: benefit on LR

4S GY+SFU/MOF x S cycles (P=0.02), not on S
or SFU + FAx 6 cycles (346) No benefit with MOF

Tveit [26] Follow-up (70) RCH benefit on S (p=O.Ol)

46Gy + SFU (76) and LR (P=O.OI)

GITSG 7180 [271 41.4Gy + SFU/SFU x 4 cycles (104) No benefit with

41.4Gy + SFU/SFU + Me-CCNU x 4 cycles (9S) Me-CCNU + SFU vs. SFU

INT 0114 [28] SFU x 2/4SGy + SFU/SFU x 2 No benefit with SFU+FA

SFU+FAx 2/4SGy + SFU +FA/SFU + FA x 2 or SFU + LEV or
SFU + LEV X2/4SGy + SFU/SFU + LEV x 2 SFU+ FA + LEV vs. SFU
SFU + FA + LEV x 2/4SGY+SFu/
SFU + FA + LEV x 2 (169S)

INT-NCCTG [24] sFU±Me-CCNU x 2/S0.4Gy Benefit on S (p=o.OOS),

+ SFU bolus/SFU ± Me-CCNU (332) TTR (p=O.Ol) and
SFU ± Me-CCNU x 2/S0.4Gy distant metastases
+ SFU ci/SFU ± Me-CCNU X2 (328) (P=0.03) with sFU ci

Lee [29] 4SGy + SFU + FA/SFU + FAx 6 cycles (ISS) Benefit on LR (P=0.047)
SFU + FAx 2/4SGy + SFU + FA and S (P=0.043) with
/SFU + FAx 2 cycles (IS3) sFU + FA before RT

RCH = concomitant radio chemotherapy; S = survival; LR = local recurrence; TTR = time to recurrence; FA = folinic acid;
LEV = levamisole; SFU = fluorouracil; Me-CCNU = methyl-lomustine; MOF = fluorouracil + vincristina + methyl-lomustine;
ci = continuous infusion

gins (CRM) (in half cases for inadequate surgical ed (particularly for tumors in the lower third of
technique) are cause of local failure [3] and this is the rectum) a more aggressive management in
not a negligible matter for several patients. Many term of chemoradiation may be indicated. The
multivariate analyses have demonstrated that CRM should be considered positive if the distance
tumor involvement of the CRM may be the single between the deepest extent of the tumor and the
most critical factor in predicting the likelihood of closest measures 0 to 1 mm on microscopic exam-
local recurrence: where a positive CRM is expect- ination (see Table IX.6) [32].
 Section IX' Combined modality therapy for rectal cancer 245

Table IX.S. Toxicity related to SFU schedules and con- Table IX.6. Risk of local recurrence with analysis of
comitant radiotherapy [301 CRM

1917 (4.6% ineligible) Toxicity (I1IIIV) CRM Risk of recurrence

ARM I 66%/42% >2mm 6%

ARM 2 68%/27% ::;2mm 16%
ARM 3 43%/50% ;:::lmm 38%

ARM I =bolus 5FU 500 mg/m2 5 days -28d before and 450
mg/m2 5 days -q28 after XRT (50.4-54 Gy) + 5FU via PVI
225 mg/m2/d during XRT; follow-up of 4.6 years, results in relapse free sur-
ARM 2 = PVI 300 mglm2/d 5FU 42 d before and 56 dafter vival and overall survival were similar (3 yrs. RFS
identical XRT + PVI as ARM 1; 68-69% P=0.4S and 3 yrs. OS 81-83% P=0.21). The
ARM 3 = bolus 5FU, LV (20 mg/m2) in 2 five days q 28 most important difference was in toxicity (see
before (425 mg/m2 5FU) and after (380 mg/m2 5FU) XRT + Table IX.S). Authors suggested that in clinical
bolus 5FU (400 mglm2)+LV d 1-4 of wk 1,5 of RT. ARM 3 practice the choice of these schedules depends on
received LEV 150 mgld dl-3, 14-16 each cycle before and different toxicity profiles.
after XRT In stage setting the sub-stratification of stage
III patients in three subsets [34] on the bases of T
and N (see Table IX.7) is recommended. This
In the same way, in order to predict the risk of stratification can predict the different risks of
pelvic relapse, it is necessary to detect a major relapse and death. In all cases patients can be dis-
number of nodes: prognosis in node-negative criminated into intermediate, moderately high
patients who have more total nodes identified (at and high-risk groups with different needs of
least 14) is better than for those who have fewer treatment strategies [3S]. From the analysis of
nodes [33]. NCCTG7947S1, NCCTG8647S1, INTu4, NSABP
Unfortunately most patients do not meet these ROI and NSABP R02 trials (3791 patients) progno-
criteria. TME is a component of adjuvant therapy sis (overall survival and disease free survival) is
first of all in patients who are found to have an related to TN stage and treatment (s-year follow-
extramural tumor or regional lymph-node up). For patients at intermediate-risk (TI-2Nl,
involvement (T3N1-N2, T4). Omission of adjuvant T3No) s-year OS was similar with chemotherapy
treatment in the case ofT3NO tumors can be con- and chemoradiotherapy (84-8S% with CT and 78-
sidered after histological review of the specimen, 83% with RT +CT, respectively). For patients at
which assures the integrity of the mesorectal sur- moderately high-risk (TI-2N2, T4No, T3Nl) OS
face and the margin negativity, exclusively. ranged from 43% to 70% with CT and from 22% to
80% with CT+RT. For high-risk patients (T3N2,
T4NI-2) OS ranged from 2S% to 4S% with CT and
What is new in adjuvant setting? from 29% to S7% with CT + RT. On the bases of this
pooled analysis the three-modality postoperative

During the last annual meeting in the USA known

as ASCO, many reports addressed neoadjuvant Table IX.7. Survival in subset of stage III patients
combined therapies and new combinations (see
neoadjuvant treatment). N° of Proposed Observed Relative
In adjuvant settings oral drugs (capecitabine patients stage survival survival
or UFT) seem suitable: they mimic SFU ci, offer- (%) (%)
ing the advantage of no devices for the adminis-
tration of chemotherapy (central venous line), 1043 IlIA (TI-2 NI) 55. 1** 67·2**
good tolerance and optimal patient-compliance. 2856 I1IB (T3-4 Nl) 35·3** 43·5**
Intergroup 0144 [30] compared SFU bolus versus 2089 I1IC (any T N2) 24·5** 29·8**
SFU ci before and after radiotherapy versus SFU
modulated by LV and levamisole. With a median ** P<O.OOOI
246 Atlas of Endoanal and Endorectal Ultrasonography

adjuvant treatment in patients at intermediate- surgical damage of micro vessels. Furthermore it

risk may be excessive. Further trials are needed to requires the irradiation of smaller treatment vol-
evaluate separate treatment regimens for interme- umes. Finally, pre-RT may induce a significant
diate versus moderately high or high-risk shrinkage of an initially fixed and marginally
patients. resectable lesion, resulting in a higher rate of sub-
Molecular markers may have the potential to sequent complete resection and/or sphincter
allow a better selection of patients for adjuvant preservation [10].
combined modality therapy. A large number of
molecular, protein and carbohydrate markers
have been investigated as possible prognostic fac- Combined modality:
tors but none have yet been validated for patient
care [I, 36, 37]. These included genotypic alter- chemoradiotherapy in preoperative
ations as c-myc amplification, Her-2/neu over settings
expression, or loss of heterozygosity at various
chromosomal sites as well as markers of cell pro- Combined chemoradiotherapy is the recom-
liferation or angiogenesis, protease or their recep- mended standard preoperative treatment for
tors and the expression of plasma membrane gly- patients with locally advanced rectal cancer in the
coproteins. Most of these studies are single-mark- United States. In preoperative settings this modal-
er, retrospective or small prospective investiga- ity of treatment preserves the advantages of pre-
tions. Large prospective cooperative group stud- RT plus those of chemotherapy (control of
ies are currently ongoing in order to assess the micrometastases). This strategy allows a spatial
real prognostic value of many of these factors. cooperation with a possible better local control of
disease through the intensification of the effects
achieving tumor shrinkage [40,41].
Neoadjuvant treatment: The first randomized clinical trial of combined
radio chemotherapy versus radiation alone
preoperative radiotherapy showed a decreased survival with the combined
modality, but chemotherapy with SFU was not
Many studies demonstrated that the toxicity of employed as a systemic therapy and radiotherapy
preoperative radiotherapy (pre-RT) is minimal did not use a conventional radiation modality
and does not increase surgical morbidity. A meta- [42]. Subsequently, three clinical trials showed an
analysis [38] and a systematic overview [39] have increase in the resectability rate (60-90% of the
both demonstrated an improvement in local con- preoperatively treated patients able to undergo
trol, disease-free survival and overall survival surgery with curative intent) and a tumor down-
with the use of pre-RT. One of these analyses also staging, with 10-20 % of pathologically-proven
showed a dose-effect relationship: a higher radia- complete remissions, in patients with locally
tion dose produced a larger therapeutic effect. advanced or unresectable disease treated with
The reason for the advantage of preoperative preoperative chemoradiotherapy (Table IX.8) [43-
radiotherapy is still unknown but may be related 46]. Some of the potential advantages of this
to an improved vascular supply of the tumor pre- approach include: reduced tumor seeding during
operatively, which would improve radiation sensi- surgery, decreased acute toxicity, increased
bility. There are many aspects that define preoper- radiosensitivity due to well oxygenated cells and
ative radiotherapy more suitable than postopera- enhanced sphincter preservation.
tive treatment: in the preoperative setting the Is very important to emphasize the necessity of
bowel is more likely to be mobile with a lower risk the diffusion of preoperative chemoradiation in
of radiation injury and morbidity. Preoperative clinical practice. As already stated, results in local-
irradiation can reduce the incidence of residual ly advanced rectal cancer (T3/T 4) treated with pre-
microscopic disease in the perirectal space operative RT showed the possibility of downstag-
beyond the resection margins, decreasing the ing, resulting in higher resection rates and reduc-
local recurrence rate and the dissemination of ing local recurrence [47]. Downstaging may be
viable cells at the time of surgery. Pre-RT may also achieved with the combination of chemotherapy
be more effective because of the oxygenation of and radiotherapy [48]. In fact as reported by the
tumor cells is not compromised by the possible CCCG, preoperative combined modality of treat-
 Section IX· Combined modality therapy for rectal cancer 247

Table IX.S. Rate of pCR (pathological complete response) obtained with preoperative therapy

Author Wofpts. EBRT (Gy) CT pCR (%)

Mendenhall [43] 132 30-50 NO 11

Rich [44] 77 45 5FU ci 29
Chan [45] 43 45 5FU-DDP 27
Minsky [46] 20 50 5FU-LV 20

Pts: patients; EBRT: external-beam radiotherapy; CT: chemotherapy; pCR: pathological complete response; ci: continuous
infusion; 5FU: fluorouracil; DDP: cisplatin; LV: leucovorin

ment can guarantee a higher pathologic downstag- trial the rate of disease free recurrence at 1 year
ing of primary tumor as compared to preoperative was 44% in patients treated with sphincter-saving
radiation alone and a lower incidence of acute tox- procedures and preoperative approach as com-
icity as compared to postoperative combined pared to 34% of patients who received postopera-
modality [49]. Downstaging (reduction in the tive therapy [50, 51]. The aim of preoperative
depth of tumor invasion in the bowel wall) and treatment is tumor downstaging and sphincter
downsizing (reduction in tumor size) allow preservation. The radiation treatment employed
sphincter-preservation in a rate of patients initial- in this setting is critical (intensive short-courses
ly considered for an abdominoperineal resection versus conventional fractionated radiation thera-
(APR). The possibility of conservative surgery pyas shown in Table IX.9). An analysis of two
increases, waiting at least 6 weeks from the end of previously published Scandinavian trials on
radio chemotherapy. intensive short -course radiation revealed a more
Initial preliminary results from two random- pronounced degree of downstaging when the
ized clinical studies seem to confirm that preop- interval between the completion of irradiation
erative chemoradiotherapy grants sphincter-sav- and surgery was at least 10 days [52].
ing surgery in approximately 20-30% of patients Radiobiologically this regimen is unfavorable
(initially judged to require an APR) without com- because it leads to a lower differentiation of the
promising the control of the disease. In an NSABP tumor response: the short interval between radio-

Table IX.9. Conventional and short course radiotherapy

Modality of preoperative treatment Conventional Short course

Doses (Gy) 25

Duration of treatment (weeks) 6

Chemotherapy YES NOT

Surgery (weeks) 4-8

Advantages Sphincter-preservation Short treatment

Less toxicity All pts treatable

Disadvantages Overtreating Mortality (?)

Undertreating Morbidity (?)
No down staging
Organization
248 Atlas of Endoanal and Endorectal Ultrasonography

therapy and surgery does not allow a tumor present unresectable rectal cancer, which is not
downstaging that can guarantee sphincter preser- sufficiently down staged by chemoradiation to
vation. Moreover the association with chemother- allow resection. There is therefore a need for more
apy it is not indicated. effective preoperative therapies in order to
Nowadays, the most common preoperative enhance sphincter-preserving surgical options if
recommended combined regimen is a 4S-S004 Gy the tumor shrinks significantly (downsizing) and
of pelvic radiation at 1.8 Gy per fraction, S frac- allow radical (Ro) surgery in a higher rate of
tions per week plus concurrent bolus sFU/leucov- patients (Table IX.8) [43-46].
orin or continuous infusion sFU with surgery per- Phase 1111 trials are examining the use of these
formed 4 to 7 weeks after the completion of irra- new drugs in combination with pelvic radiation
diation. This approach allows the recovery from therapy. Selected cytotoxic agents (i.e. uracil and
the acute side effects of radiation and adequate tegafur, raltitrexed, oxaliplatin, irinotecan and
time for tumor downstaging. Results are limited: capecitabine) as well as targeted agents such as
local recurrence rates <10%, significantly lower C22S, ZD1839 and OSI are being combined with
than expected in a primary advanced tumor; in pelvic radiation therapy. Phase III trials are need-
contrast metastases are the more frequent way of ed to determine whether these regimens offer an
recurrence and do not seem relevantly influenced advantage compared with sFU-based combined
by the reduction of local recurrence. modality therapy regimens.
Preoperative combined modality treatment Both irinotecan and oxaliplatin have shown
also allows the control of systemic disease and the radiation sensitization. There are ongoing stud-
difference of low and high-risk patient groups. ies on the use of these agents along with sFU and
Some studies suggest that patients with T3 or T4 radiation therapy in the preoperative setting. It is
rectal cancers, who achieved a pathologic tumor very important to pay attention to the timing of
downstaging after chemoradiation, have a chance radiation and drug delivery. Recent studies sug-
of s-year survival in the region of 90-100%. In gest that irinotecan should be given within 2
contrast, patients who still present tumor cells hours from radiation delivery to have a maximal
infiltrating through the bowel wall or in the sensitization [S3], whereas oxaliplatin should be
perirectal fat or positive involved lymph nodes administered a number of hours prior to radio-
after chemoradiation, have a very poor prognosis therapy [S4]. Oxaliplatin should also be given
with a s-year survival of 2S% [SI]. during the early part of the week to benefit from
Thepotential disadvantage of this procedure is its prolonged action in the tissues. Capecitabine
overtreatment: the preoperative approach may is also used to produce radiation sensitization
result in the inclusion of patients with an early due to its activity similar to continuous infusion
stage of disease due to overstaging or with of sFU: preliminary data suggests that it should
metastatic disease. The future hope is that further be given approximately 1 hour before the deliv-
improvements in accuracy of staging techniques ery of the radiation therapy to maximize the
(endorectal ultrasonography, pelvic computed effect.
tomography and magnetic resonance imaging) Mitchell et al. [ss] conducted a recent phase
may allow us to carefully select patients who 1111 study presented at ASCO 2003 to evaluate the
could benefit from preoperative treatment, spar- maximally tolerated dose and effectiveness of
ing those for whom radiotherapy may be not nec- CPT-H, sFU and RT preoperatively in clinical
essary. T3/T 4 adenocarcinoma of the rectum. They
reported a good rate of tolerance, with 39% of
complete or almost complete pathological
What is new in neoadjuvant setting? response to therapy. Encouraging results were
also recorded in another phase II study conducted
by Dunst et al. [s6] (Table IX.lO) [s6-62] regarding
Even though sFU has been the standard drug used continuous daily capecitabine combined with
in association with pelvic irradiation in the treat- radiotherapy as a neoadjuvant treatment of local-
ment of rectal cancer for many decades, many ly advanced rectal cancer. This combination
additional drugs have been developed for the proved to be well tolerated and resulted in a major
therapy of advanced disease. This is of funda- clinical response in 81% of patients (tumor down-
mental importance because several patients still staging reached in 72% of patients).
 Section IX· Combined modality therapy for rectal cancer 249

Table IX.10. What is new in neoadjuvant combined therapy for rectal cancer

Authors Treatment Results

Dunst [57), 45 pts Capecitabine + RT as neoadjuvant Downstaging: 72%

treatment for locally advanced Clinical CR or clinical PR: 81%
rectal cancer Only 8% remained inoperable

Lin [58), 54 pts Capecitabine + RT + concomitant boost Good tolerance

May replace 5FU ci

Glynne-Jones [59],17 pts. Phase I: preoperative RT Capecitabine recommended

+ Capecitabine + OXP dose: 650 mglm2/BD
in combination with
OXP 130 mg/m2 days 1 & 29
and synchronous preoperative RT
Tumor downstaging rates
are encouraging high
Ongoing phase II

Rosenthal [60],16 pts. Phase I: preoperative RT Safety with full-dose OXP+RT

+ concurrent 5FU ci + dose escalating Promising pathologic response
OXP followed by surgery.
Adjuvant 5FU and Leucovorin Recommended OXP dose
for phase II is 85 mt
Casado [61), 25 pts. Neoadjuvant OXP+ raltitrexed + RT Downstaging: 93%
CR: 14%
Toxicity: moderate

Balch [62],158 pts.: Prognostic significance of response Difference in 5 yrs. disease specific
38%: neoadjuvant CRT to neoadjuvant chemoradiation survival in pts. treated with
36%: adjuvant CMT therapy for rectal cancer Neoadjuvant CMT + RT when comparing
26%: surgery alone those with a pCR (100%), pPR (95%),
no response (65%) p=0.012
Response to neoadjuvant
CRT is a significant prognostic factor

Gemici [63), 14 pts. Preoperative RT High rate of pathologic response

+ mitomycin C and IFNa Tumor down-staging in all pts. (10110)

CRT = chemoradiotherapy; CMT = chemotherapy; RT = radiotherapy; CR = complete response; PR = partial response; OXP
= oxaliplatin

2. Adam IJMM, Martin IG. Role of circumferential

References involvement in the local recurrence of rectal cancer.
Lancet 1995;344:707-10
1. Compton CC, Fielding LP, Burgat LJ, et al. Prognostic 3. Nagtegaal ID, Van de Velde CJH, Van de Worp E, et al.
factors in colorectal cancer: College of American Macroscopic evaluation of rectal cancer resection
Pathologist Consensus Statement 1999. Arch Pathol specimen: clinical significance of the pathologist in
Lab Med 2000;124:979-94 quality control. J Clin Oncol 2002;20:1729-34
250 Atlas of Endoanal and Endorectal Ultrasonography

4. Dahlberg M, Pahlman L, Bergstrom R, et al. Improved 18. Medical Research Council Rectal Cancer Working
survival in patients with rectal cancer: a population- Party. Randomized trial of surgery alone versus
based register study. Br J Surg 1998i85:515-20 surgery followed by radiotherapy for mobile cancer of
5. Stocchi L, Nelson H, Sargent DJ, et al. North Central the rectum. Lancet 1996i348:1606-1O
Cancer Treatment Group. Impact of surgical and 19. Koelbl 0, Richter S, Flentje M. Influence of patient
pathologic variables in rectal cancer: a United States positioning on dose-volume histogram and normal
community and cooperative group report. J Clin Oncol tissue complication probability for small bowel and
2001i19:3895-902 bladder in patients receiving pelvic irradiation: a
6. Bleiberg H, Kemeny N, Rougier P, et al. Colorectal prospective study using a 3D planning system and a
Cancer: a clinical guide to therapy. Martin Dunitz, radiobiological model. Int J Radiat Oncol BioI Phys
London 2002, pp 173-84 1999i45:1193-8
7. Gunderson LL, So sin H. Areas of failure found at re- 20. Byfield JE. Theoretical basis and clinical applications
operation (second or symptomatic look) following of 5-fluorouracil as a radiosensitizer. An overview. In
curative surgery for adenocarcinoma of rectum. Rotman M, Rosenthal CT (Eds): Concomitant continu-
Clinicapathological correlations and implications for ous infusion chemotherapy and radiation. Springer
adjuvant therapy. Cancer 1974i34:1278-92 Verlag, Berlin 1991, PP115-27
8. Obrand DI, Gordon PH, et al. Incidence and pattern of 21. Lawrence TS, Maybaum J. Fluoropyrimidines as radia-
recurrence following curative resection for colorectal tion sensitizer. Sem Radiot OncoI1993i3:20-8
carcinoma. Dis Colon Rectum 1997i40:15-24 22. National Institute of Health Consensus Conference.
9. GITSG 7175: Prolongation of the disease-free interval Adjuvant therapy for patients with colon and rectal
in surgically treated rectal carcinoma. Gastrointestinal cancer. JAMA 1990i264:1444-50
Tumor Study Group. N Engl J Med 1985i312:1465-72 23. Krook JE, Moertel CG, Gunderson LL, et al. Effective
10. Valentini V, Micciche F. Quali pazienti radiotrattare e surgical adjuvant therapy for high risk rectal carcino-
quando. Tumori 2003i4:S87-S88 ma. N Engl J Med 1991i324:709-15
11. Minsky BD, Conti JA, Huang Y, et al. The relationship of 24. O'Connel MJ, Martenson JA, Wieand HS, et al.
acute gastrointestinal toxicity and volume of irradiated Improving radiation therapy for rectal cancer by com-
small bowel in patients receiving combined modality bining protracted infusion fluorouracil with radiation
therapy for rectal cancer. J Clin OncoI1995i13:1409-16 therapy after curative surgery. N Engl J Med 1994i331:
12. Sitzler PJ, Seow-Choen F, Ho YH, et al. Lymph node 502-7
involvement and tumor depth in rectal cancer: an anal- 25. Wolmark N, Wieand HS, Hyams DM, et al. Randomized
ysis of 805 patients. Dis Colon Rectum 1997i40:1472-6 trial of postoperative adjuvant chemotherapy with or
13. Balslev I, Pedersen M, Teglbjaerg PS, et al. Post- without radiotherapy for carcinoma of the rectum:
operative radiotherapy in Dukes'B and C carcinoma of National Surgical Adjuvant Breast and Bowel Project
the rectum and rectosigmoid. Randomized multicen- Protocol R-02. JNat Cancer Inst 2000i92:338-96
ter study. Cancer 1986i58:22-8 26. Tveit KM, Guldvog S, Hagen E, et al. Randomized con-
14. Fisher B, Wolmark N, Rockette H, et al. Postoperative trolled trial of postoperative radiotherapy and short-
adjuvant chemotherapy or radiation therapy for rectal term time-scheduled 5-fluorouracil against surgery
cancer. Results from NSABP protocol R-01. J Nat alone in the treatment of Dukes Band C rectal cancer.
Cancer Inst 1988i80:21-9 Br J Surg 1997i84:1130-5
15. Bentzen SM, Baslev I, Pedersen M et al. Time to locore- 27. GITSG 7180. Radiation therapy and fluorouracil with
gional recurrence after resection of Dukes'B and C col- or without semustine for treatment of patients with
orectal cancer with or without adjuvant postoperative surgical adjuvant adenocarcinoma of rectum.
radiotherapy. A multivariate regression analysis. Br J Gastrointestinal Tumor Study Group. J Clin Oncol
Cancer 1992i65:102-7 1992ilO:549-57
16. Treurniet Donker AD, Van Putten WL, Werwldsma JC. 28. Tepper JE, O'Connel M, Niedzwiecki D. Adjuvant ther-
Postoperative radiation therapy for rectal cancer. An apy in rectal cancer: analysis of stage, sex, and local
interim analysis of a prospective, randomized multicen- control. Final report of Intergroup 0114. J Clin Oncol
ter trial in the Netheralands. Cancer 1991i67:2028-42 2002i20:1774-50
17. Arnaud JP, Nordlinger B, Bosset JF, et al. Radical 29. Lee HL, Lee JH, Ahn JH. Randomized trial of postop-
surgery and postoperative radiotherapy as combined erative adjuvant therapy in stage II and III rectal can-
treatment in rectal cancer. Final results of phase III cer to define the optimal sequence of chemotherapy
study of the European Organization for Research and and radiotherapy: a preliminary report. J Clin Oncol
treatment of Cancer. Br J Surg 1997i84:352-7 2002i20:1751-8
 Section IX· Combined modality therapy for rectal cancer 251

30. Smaller SR, Benedetti J, Williamson J, et al. Intergroup probability of local control and survival in high-risk
0144-phase III trial of 5FU based chemotherapy regi- distal rectal cancer? Ann Surg 1992;215:696-705
mens plus radiotherapy (XRT) in postoperative adju- 44. Rich TA, Skibber JA et al: Preoperative infusional
vant rectal cancer. Bolus 5FU vs. prolong venous infu- chemoradiation therapy for stage T3 rectal cancer. Int
sion (PVI) before and after XRT + PVI vs. bolus 5FU + J Radiat One BioI Phys 1995;32:1025-9
leucovorin (LV)+Levamisole (LEV) before and after 45. Chan A, Wong A, Langevin J, Khoo R. Preoperative
XRT + bolus 5FU+LY. Proceedings of ASCO 2003;22: concurrent 5FU infusion, mitomycin and pelvic radia-
abstract n° 1006 tion therapy in tethered and fixed rectal carcinoma.
31. Kapiteijn E, Corrie M, Nagtegaal ID, et al. Preoperative Int J Radiat Oncol BioI Phys 1993;25:791-9
radiotherapy combined with total mesorectal excision 46. Minsky BD, Cohen AM, Enker WE, et al. Preoperative
for resectable rectal cancer. N Engl J Med 2001;34:638-46 5FU, low-dose leucovorin and radiation therapy for
32. Adam 1], Mohamdee MO, Martin IG, et al. Role of the locally advanced and unresectable rectal cancer. Int J
circumferential margin involvement in the local recur- Radiat Oncol BioI Phys 1997;40:515-22
rence of rectal cancer. Lancet 1994;344:707-11 47. Enami B, Pilepich M, Willett C, et al. Effect of preoper-
33. Tepper JE, O'Connell MJ, Niedwiecki D, et al. Impact of ative irradiation on resectability of colorectal carcino-
number of nodes retrieved on outcome in patients mas. Int J Radiat Oncol BioI Phys 1982;8:1335-9
with rectal cancer. J Clin OncoI2001;19:157-63 48. Bosset JF, Pavy JJ, Hamers HP, et al. Determination of
34. Greene FL, Steward AK, Norton HJ, et al. A new staging the optimal dose of 5-fluorouracil when combined
strategy for node-positive stageIII rectal cancer: An with low-dose D,L-Leucovorin and irradiation in rec-
analysis of 5988 patients. Proceedings of ASCO tal cancer: results of 3 consecutive Phase II studies. Eur
2003;22:abstract n° 1007 J Cancer 1993;33A:1406-10
35. Gunderson 11, Sargent D, Tepper J, et al. Impact of TN 49. Minsky BD, Cohen AM, Enker WE, et al. Combined
stage and treatment on survival and relapse in adju- modality therapy of rectal cancer. Decreased acute
vant rectal cancer pooled analysis. Proceedings of toxicity with the preoperative approach. J Clin One
ASCO 2003;22:abstract n° 1008 1992;10:1218-24
36. Compton CC. Update protocol for the examination of 50. Roh MS, Petrelli N, Wieand S, et al. Phase III random-
specimens removed from patients with colorectal car- ized trial of preoperative versus postoperative multi-
cinoma. Arch Pathol Lab Med 2000;124:1016-25 modality therapy in patients with carcinoma of the
37. Compton CC, Fenoglio-Preiser CM, Pettigrew N, et al. rectum (NSABP R-03). Proceedings of ASCO 2001,
American Joint Committee on Cancer Prognostic abstract n° 490
Factors Consensus Conference: Colorectal Working 51. Ahmed, et al. Pathologic complete response predicts
Group. Cancer SS 2000;1739-57 long-term survival following preoperative radiation
38. Camma C, Giunta M, Fiorica F, Pagliaro L, et al. therapy for rectal cancer. Int J Radiat Oncol BioI Phis
Preoperative radiotherapy for resectable rectal cancer. 1997;39:284, abstract n02087
JAMA 2000;23:30 52. GrafW, Dahlberg M, Osman MM, et al. Short-term pre-
39. Colorectal Cancer Collaborative Group: Adjuvant operative radiotherapy results in downstaging of rectal
radiotherapy for rectal cancer: A systematic overview cancer: a study of 1316 patients. Rad OncoI1997;43:133-7
of 8507 patients from 22 randomized trials. Lancet 53. Omura M, Torigoe S, Kubota N. SN-38, a metabolite of
2001;358:1291-304 the camptothecin derivative CPT-ll, potentiates the
40. Minsky BD. Multidisciplinary management of cytotoxic effect of radiation in human colon adeno-
resectable rectal cancer. Oncology 1996;10:1701-14 carcinoma cells grown as spheroids. Radiother Oncol
41. Rosenthal q, et al. Concomitant continuous infusion 1997;43=197-201
chemotherapy and radiation, Springer Verlag, Berlin, 54. Hess S. Oxaliplatin: In vitro and in vivo evidence of its
1991:1-9 radiation sensitizing activity-preclinical observations
42. Boulis-Wassif S, Gerard A, Loygue J, et al. Final results relevant to ongoing clinical trials. Proc AACR 2000;
of a randomized trial on the treatment of rectal cancer 41:A335
with preoperative radiotherapy alone or in combina- 55. Mitchell EP, Anne PR, Fry R, et al. Chemoradiation
tion with 5-fluorouracil, followed by radical surgery. with CPTll, 5FU in neoadjuvant treatment of locally
Trial of the European Organization on Research and advanced or recurrence adenocarcinoma of the rec-
Treatment of Cancer Gastrointestinal Tract Cancer tum: A phase lIII study update. Proceedings of ASCO
Cooperative Group. Cancer 1984;53=1811-8 2003;22:abstract n° 1052
43. Mendenhall WM, Bland KI, Copeland EM 3rd, et al. 56. Dunst J, Reese T, Hoelscher T, et al. Capecitabine com-
Does preoperative radiation therapy enhance the bined with radiotherapy as neoadjuvant treatment of
252 Atlas of Endoanal and Endorectal Ultrasonography

locally advanced rectal cancer. Proceedings of ASCO adjuvant 5FU, and leucovorin for locally advanced
2003;22: abstract n° 1113 (T3/4) rectal adenocarcinoma. Proceedings of ASCO
57. Lin EH, Skibber J, Delcos M, et al. A phase II study of 2003;22:abstract n° 1094
capecitabine and radiotherapy plus concomitant boost 60. Casado E, De Castro J, Feliu J, et al. Neoadjuvant ther-
in patients (pts) with locally advanced rectal cancer apy of rectal carcinoma with oxaliplatin and
(LARC): Preliminary safety analysis. Proceedings of raltitrexed plus radiotherapy. Proceedings of ASCO
ASCO 2003;22:abstract n° 1152 2003;22:abstract n° 1342
58. Glynne-Jones R, Sebag-Motefiore D, Me Donald A, et 61. Balch GC, Mithani SK, Shyr Y, et al. Prognostic signifi-
al. A phase I study of preoperative radiation and cance of response to neoadjuvant chemoradiation
capecitabine in combination with oxaliplatin in local- therapy for rectal cancer. Proceedings of ASCO
ly advanced rectal cancer. Proceedings of ASCO 2003;22:abstract n° 1047
2003;22:abstract n° 1174 62. Gemici C, Mayadagli A, Marti A, et al. Preoperative
59. Rosenthal Dr, Catalano P, Haller DG, et al. ECOG 1297: radiation in combination with 5-fluorouracil, mito-
A phase I study of preoperative radiation therapy (RT) mycin-C and interferon a in locally advanced rectal
with concurrent protracted continuous infusion 5FU adenocarcinoma. Proceedings of ASCO 2003;22:
and dose escalating oxaliplatin followed by surgery, abstract n° 1308
 Invited commentary
V. Valentini

The Authors reviewed the role of the preoperative indication for any adjuvant therapy after surgery.
chemoradiation and of the adjuvant therapies in Some clinical experiences identified better out-
rectal cancer analyzing the more relevant studies comes in T3-4 patients treated with preoperative
with a well-managed evidence based approach, radiotherapy + concomitant chemotherapy who
furthermore they addressed the ongoing ques- experienced downstaging (Table IX.n) [1-5].
tions in the scientific debate very correctly. Furthermore, a new pathological staging sys-
Between the many aspects treated in the tem was tested to verify the prognostic meaning of
manuscript for the preoperative setting, it seems the tumor regression after preoperative chemora-
relevant to stress the studies devoted to explore if diation. Tumor Regression Grade (TRG) has been
downsizing and downstaging could identify a quantified in 5 grades according to Mandard score:
population with better prognosis, so to reduce the TRG 1 (complete response) showed absence of

Table IX.II. Clinical outcomes in T3-4 patients treated preoperatively with radiotherapy ± concomitant
chemotherapy stratified according to pathological findings

Author pTO-z/Tot pts. Preoperative % 5-yearLC %5-year DFS %5-year SVV

treatment pTO-z pTO-z pTO-z

Janjan [1] 68/117 45 Gy + 5FU PVI n.a. 75-83* 93-100*

M.D.Anderson

Berger [2] 19/167** 44 Gy + 5FU PVI n.a. 87 9Z

Hospital Bretonneau

Kaminsky-Forrett [3] 21/88 40 Gy ± 5FU bolus + FA 940 94 0 1000

Alexis Vautrin Centre

Mohiuddin [4] 22/77 45 Gy + 5-15 Gy 1000 1000 1000

University of Kentucky + 5FU PVI

Catholic University [5] 77/165 45 Gy 88-100 80-82 80-90

+5FU +MMC
+5FU + DDP

LC = Local control; DFS = Disease Free Survival; SVV = Survival, n.a. = not available; MMC = Mitomycin C; DDP = D-
Cisplatin; * 3-year; ** pTO-l; ° only pNo patients
254 Atlas of Endoanal and Endorectal Ultrasonography

residual cancer and fibrosis extending through the regression may help to predict the outcome for
bowel wall [6]. TRG 2 was characterized by the rectal cancer in patients treated with preoperative
presence of rare residual cancer cells; TRG 3 was chemoradiation [5].
characterized by an increase in the number of The pending outcomes of randomized trials
residual cancer cells with predominant fibrosis; (EORTC, Italian study) on the contribution of
TRG 4 showed residual cancer outgrowing fibrosis adjuvant chemotherapy after preoperative radio-
and TRG 5 was characterized by the absence of therapy + concomitant chemotherapy will clarify
regressive changes. In our Institution 144 patients the need for adjuvant therapy and the more rele-
were evaluated after preoperative chemoradiation. vant prognostic factors for identifying the sub-
The analysis used was a statistical competing risk groups which could exploit its contribution [7,8].
analysis (metastases vs. death) with respect to Another aspect addressed in the manuscript
TRGs. Median follow-up was 46 months (range: 3- explores the contribution of new drugs in enhanc-
143). The cumulative incidence rates of distant ing sphincter saving surgery procedures, not only
metastases across TRG levels were significantly the tumor response. In Table IX.12 [9-12] the per-
different. (p<O.01). The 5-year cumulative inci- centage of sphincter saving surgery recorded in
dence rates for metastases were 7.6% (95% CI: 0- 247 T3 patients treated with 4 different chemoradi-
18%) for TRGl, 5.6% (95% CI: 0-16.7%) for TRG2, ation schedules between 1990-2002 shows how the
10% (95% CI: 0-21.2%) for TRG3 and 44.1% (95% use of new intensification modalities improved the
CI: 31-57.2%) for TRG4, respectively. We conducted outcomes. However, the use of new drugs in com-
a univariate analysis to test whether pathologic bination with preoperative radiotherapy requires
stage is associated with metastases-free survival. careful methodology in defining the dose of the
On univariate analysis, pathologic stage was a sig- drug to be used. Phase III studies are addressed to
nificant predictor of metastasis-free survival define the new drugs' contribution on sphincter
(P<O.OI). In a multivariate analysis including both saving surgery and on the control of the disease;
these factors, TRG still remained an independent Phase I and Phase II studies are promoted this year
predictor (p<O.01). These data suggest that tumor in 2 Italian cooperative groups.

Table IX. 12. Impact on sphincter-saving surgery in a sequence of 4 chemoradiation studies undertaken at Catholic
University between 1990-2002

Study Trial Design Sphincter saving procedures

According to the distance
of the lower pole of the tumor
from the anorectal marging
0-100 mm 0-30 mm

FUMIR [9] 38 GyERT +

MMC 10 mg/m2, day 1
5FU 1000 mg/m2, days 1-4

PLAFUR [10] 50.4 GyERT + 69/80 (86%)

DDP 60 10 mg/m\ days 1,29
5FU 1000 mg/m2, days 1-4,29-32

TOM-RT[l1] 50.4 GyERT +

Raltritexed 3 mg/m2, days 1,19,38

TOMOX-RT [12] 50.4 GyERT + 3115 (80%)

Raltritexed 3 mg/m2, days 1,19,38
Oxaliplatinum 130 mg/m2, days 1,19,38

ERT = External Radiotherapy, MMC = Mitomycin C, DDP = D-Cisplatin

 Section IX' Combined modality therapy for rectal cancer 255

It seems useful to focus on the fact that radio- could receive less adjuvant therapy has to be
therapy promotes better local control in any set- addressed very carefully, as these are strategies
ting: preoperatively even in combination with a which do not guarantee the ongoing low rate of
certified TME, postoperatively in combination local recurrence.
with concomitant chemotherapy. Local control I compliment the Authors of the manuscript
has a tremendous impact on the patients' quality for their very clear exposition of the evidence-
of life. Among the many aspects analyzed in the based strategies in the combined treatment of rec-
postoperative setting it seems relevant to stress tal cancer.
that the research of subgroups of patients, who

References 7. Skibber JM, Hoff PM, Minsky BD. Cancer of the rec-
tum in De Vita V, Hellman S, Rosemberg SA: Cancer,
1. Janjan NA,Abbruzzese J, Pazdur R, Khoo VS, Cleary K, principle & practice of oncology' 6th edition,
Dubrow R, Ajani J, Rich TA, Goswitz MS, Evetts PA, Lippincott Williams & Wilkins Eds, Philadelphia, 2001,
Allen PK, Lynch PM, Skibber JM Prognostic implica- 1271-319
tions of response to preoperative infusional chemora- 8. Cionini L, Cartei F, Manfredi B, Laliscia C, Sainato A,
diation in locally advanced rectal cancer. Radiother Valentini V, Lupattelli M, Pizzi GB, Osti M, Santoni R.
OncoI1999;51:153-60 Randomized study of preoperative chemoradiation
2. Berger C, de Muret A, Garaud P, Chapet S, Bourlier P, (CTRT) in locally advanced rectal cancer. Preliminary
Reynaud -Bougnoux A, Dorval E, de Calan L, Huten N, Ie results Int J Radiat Oncol BioI Phys 1999;45: supple-
Folch 0, Calais G. Preoperative radiotherapy (RT) for ment 178,
rectal cancer: predictive factors of tumor downstaging 9. Valentini V, Coco C, Cellini N, Picciocchi A, Genovesi
and residual tumor cell density (RTCD): prognostic D, Mantini G, Barbaro B, Cogliandolo S, Mattana C,
implications. Int J Radiat Oncol BioI Phys 1997; 37:619-27 Ambesi-Impiombato F, Tedesco M, Cosimelli M.
3. Kaminsky-Forrett MC, Conroy T, Luporsi E, Peiffert D, Preoperative chemoradiation for extraperitoneal T3
Lapeyre M, Boissel P, Guillemin F, Bey P Prognostic rectal cancer: acute toxicity, tumor response, and
implications of downstaging following preoperative sphincter preservation. Int J Radiat Oncol BioI Phys
radiation therapy for operable T3-T4 rectal cancer. Int 1998;40:1067-75
J Radiat Oncol BioI Phys 1998;42:935-41 10. Valentini V, Coco C, Cellini N, Picciocchi A, Fares MC,
4. Mohiuddin M, Hayne M, Regine WF, Hanna N, Rosetto ME, Mantini G, Morganti AG, Barbaro B,
Hagihara PF, McGrath P, Marks GM Prognostic signif- Cogliandolo S, Nuzzo G, Tedesco M, Ambesi-
icance of postchemoradiation stage following preop- Impimbato F, Cosimelli M, Rotman M. Ten years of
erative chemotherapy and radiation for advanced! preoperative chemoradiation for extraperitoneal T3
recurrent rectal cancers. Int J Radiat Oncol BioI Phys rectal cancer: acute toxicity, tumor response and
2000;48:1075-80 sphincter preservation in three years consecutive
5. Vecchio FM, Valentini V, Padula G, Venkatraman E, studies. Int J Radiat Oncol BioI Phis 2001;51,2:371-83
Fares MC, Micciche F, Ricci R, Moranti AG, 11. Gambacorta MA, Valentini V, Morganti AG, Mantini G,
Gambacorta MA, Maurizi F, Minsky BD, Coco C. A Micciche F, Ratto C, Di Miceli D, Rotondi F, Alfieri S,
pathologic score system after preoperative combined Doglietto GB, De Paoli A, Rossi C, Cellini N.
modality therapy as a predictor of metastases free sur- Chemoradiation with Raltitrexed ('Tomudex') in pre-
vival in the management of rectal carcinoma. Int J operative treatment of stage IIIIII resectable rectal
Radiat Oncol BioI Phys 2002;54:207 cancer: A Phase II Study (In press).
6. Mandard AM, Dalibard F, Mandard JC, Marnay J, 12. Gambacorta MA, Valentini V, Morganti AG, Smaniotto
Henry-Amar M, Petiot JF, Roussel A, Jacob JH, Segol P, D, Micciche F, Mantini G, Barbaro B, Garcia-Vargas JE,
Samama G, Ollivier JM, Bonvalot S, Gignoux M. Coco C, Cogliandolo S, Picciocchi A, Cellini N.
Pathologic assessment of Tumor Regression Grade Chemoradiation with Raltitrexed and Oxaliplatin in
after preoperative chemoradiotherapy of esophageal preoperative treatment of stage 1I/I11 resectable rectal
carcinoma; Cancer 1994;73:2680-6 cancer: A Phase I and A Phase II Studies ( In press).
 IX.2.
Intraoperative radiotherapy in rectal cancer
-----
V. Valentini, M. Campitelli, G. Mantini

Intraoperative radiation therapy (IORT) refers to therapy department during surgical procedure or,
the delivery of irradiation at surgery. A large sin- alternatively, the use of a dedicated operating
gle fraction of irradiation (in the range of 10-15 room with a linear accelerator.
Gy) is delivered to a limited body region, with dis- Therefore, more recently, some alternative
placement of uninvolved and dose-limiting tis- techniques have been proposed, such as small
sues. IORT dedicated linear accelerators which can be
IORT has been proposed in the treatment of transferred to different operating rooms during
patients in whom the risk oflocal failure is promi- the same day, or IORT without external beams
nent for the cure or for preserving a good quality produced by linear accelerators, based on the
of life. positioning of high activity radioactive sources,
The first known experience similar to IORT using specially designed surface applicators.
was performed by Comas and Prio in 1905 [1].
Other initial experiences were described by Beck
in 1909, for patients with advanced gastric and Multidisciplinary approach
colon carcinoma [2] whose lesions were exposed
to an abdominal injury kept open for more than
three weeks. Between 1930 and 1950, a number of IORT is a treatment modality designed to com-
institutions used higher energy orthovoltage bine surgery and radiation therapy to increase
units for IORT [3, 4]. The modern approach to local control rates. The knowledge and expertise
IORT began with Abe's studies at the University of of both surgeon and radiation oncologist are
Kyoto in the early 1960s [5]. His approach was to essential in treatment decision-making through-
perform surgical excision immediately followed out the operation procedure: target approach, tar-
by a single massive dose (25-30 Gy) of radiation get delimitation, organ displacement, dose pre-
during surgery. This work stimulated significant scription and treatment report.
laboratory and clinical investigation into the use IORT offers the surgeon a back-up to the effec-
of IORT throughout Japan, Europe and the United tiveness of operative procedures. Even in
States. In the USA, the first IORT facility was con- instances where the tumor cannot be completely
structed at the Howard University Hospital, where resected, the possibility of precisely defining
a suite of the radiotherapy department was intraoperatively the limits of the gross tumor
equipped as an operating room [6]. and/or the areas of potential spread is important
Today, most clinical experiences of IORT are in the improvement of clinical results.
based on electron beams produced by linear The anesthesiology team is critically and
accelerators (Fig. IX.l); however, this kind of closely involved in all stages of surgery and
treatment is associated with organizational disad- radiotherapy. In fact, an anesthetized patient
vantages, such as patients transfer to the radio- undergoing a surgical procedure is generally
 Section IX· Combined modality therapy for rectal cancer 257

Fig. IX.I. Linear accelerator for IORT

administration

transferred to a radiation therapy unit where selected that encompasses the target, usually with
anesthesia must be safely continued while the a 1 cm margin. It can often function adequately as
attending personnel are not near the patient dur- a normal tissue retractor to hold sensitive normal
ing tissue irradiation. structures out of the IORT field, otherwise surgi-
Furthermore, the IORT procedure also chal-
lenges the nursing and technical staff with many
new responsibilities.

10RT delivery

The IORT dose is delivered in dedicated suites or

after transfer of the patient from the operating
room to the radiation therapy suite. The dedicat-
ed suite represents the ideal solution as these
facilities simplify the combination of IORT with
surgery. There is no requirement for a transport
process and the personnel remain in the same
familiar environment. The main limitation of this
situation is related to the high costs. In many
institutions, the patient must be transferred from
the operating room to the radiation oncology
department for treatment and back to the operat-
ing room if necessary.
Before proceeding to IORT administration, it
is very important to have a good hemostasis so as
to create an optimal operatory field for IORT.
For the delivery of IORT, a set of treatment
applicators must be made available with full
physic calibration (Fig. IX.2). An applicator is Fig. IX.2. Applicator tube for the delivery of IORT
258 Atlas of Endoanal and Endorectal Ultrasonography

cal packing or lead covered with saline-soaked Advantages of IORT

gauze are used to prevent normal tissue irradia-
tion. The applicators have to be docked to LINAC
before irradiation to guarantee optimal beam The high dose of additional radiation treatment
alignment. The selection of the dose and electron provided by IORT yields:
energy depends on the amount of residual tumor
after maximal resection [7]. The delivered dose - improved local control of selected tumors,
should be accurately delivered according to IORT thanks to the direct tumor visualization and
treatment guidelines [8]. administration of the dose on the target;
- higher effective doses of radiation, while dose
limiting structures are displaced, leading to a
New technologies reduced toxicity due to irradiation;
- improved therapeutic ratio of local control
versus complications, because of the possibili-
Recently, radioactive devices and mobile LINACs ty of excluding all or a part of the dose-limit-
are used to facilitate IORT. The first method, ing structures through intraoperative mobi-
based on the use of radioactive sources, is called lization and shielding (or by using appropri-
high-dose-rate intraoperative irradiation (HDR- ates radiation energies).
IORT). It is based on the use of superficial appli-
cators composed of flexible material cut to specif-
ic size and thickness. These applicator tubes are Tolerance and toxicity
incorporated to guide the positioning of radioac-
tive sources. After the surgeon completes the
resection, the target to be treated with HDR-IORT IORT dosage is based on normal tissue tolerance,
is delineated jointly by the surgeon and the radia- estimated by laboratory tests and clinical studies
tion oncologist. The applicator is then placed onto [9]. A single large intraoperative dose has a nega-
the appropriate surface and fixed with packing or tive impact on normal tissues if they are included
sutures. The applicator is then connected to the in the IORT field.
treatment machine (a remote after-loader that IORT toxicity is related to the dose and the
projects the sources into the tubes). After the kind of structures involved in the treatment vol-
delivery of the prescribed dose, the guide tubes ume. It is mainly a late toxicity: peripheral nerves
are disconnected from the remote after-loader have been definitely assessed as dose limiting and
and the applicator itself is removed. At this point, ureters as dose-sensitive in IORT experience. Only
the surgical team can complete the surgical proce- anecdotal reports have described severe toxicity
dure. in other organs as bone (vertebral collapse), ves-
Mobile operating rooms LINACs (Mobetron sels (fatal bleeding), and brain (demyelinization).
[8] and NOVAC-7 [9]) were introduced in the late However, tissue toxicity and second neoplasms
1990'S. The electron beam energy control is have frequently been reported for IORT doses
achieved without resorting to a bending magnet, exceeding 20-25 Gy [10].
thereby avoiding the high levels of radiation leak-
age produced by that system. These machines are
brought to the patient's bed after surgical expo- Selection criteria for IORT
sure of the target. The applicator is readily con-
nected to the mobile LINAC; the dose is delivered
in a few minutes. The time of IORT procedure is IORT application fields are mainly: cancer of the
shortened and the machines can be used in adja- stomach, pancreas, colo rectum, soft tissue sarco-
cent operating rooms, thus permitting a wider uti- ma and breast cancer. Their risk of local recur-
lization of IORT. rence is in fact high and a main cause of failure.
Finally there is the possibility of employing IORT represents an interesting model of thera-
orthovoltaic radiation produced by miniaturized peutic integration and in particular it can increase
sources. the efficacy of the traditional association between
surgery and radiotherapy especially when there is
a low local control rate achievable with surgery
 Section IX' Combined modality therapy for rectal cancer 259

alone and no medical contraindication exists for ventional fractionated EBRT ± chemotherapy and
gross tumor resection. On the other hand IORT is maximal surgical resection [11].
proper when external beam radiation therapy In primary tumors unresectable for cure, in
(EBRT) doses required for adequate local control locally advanced colorectal cancer and in local
exceed normal tissue tolerance. Finally metastatic recurrence it has been shown that local control
disease is a criterion of exclusion. and long-term survival improve especially when
IORT is combined with multidisciplinary treat-
ment programs of EBRT, surgery and chemother-
Sequencing of treatment in IORT apy, in comparison with results achieved with
conventional treatment. The results are most
beneficial in patients undergoing complete
Decision about the optimal sequencing of surgery resection as compared to patients undergoing
and the external beam component of irradiation partial resection only. These findings are consis-
for locally advanced cancers should involve sur- tent with those 12, 13], Mayo [14, 15], Pamplona
geon, radiation oncologist and medical oncologist [16] and Rome [17]). In an initial MGH report
as a multidisciplinary approach. For many [12] of 32 patients treated with EBRT, maximal
patients with locally advanced lesions, preopera- resection and IORT, 16 patients with initially
tive EBRT of 45-50 Gy in 1.8-2.0 fractions (± unresectable primary lesions received EBRT
chemotherapy as indicated by disease site) fol- before resection and IORT. When results where
lowed by exploration and resection in 3-5 weeks compared with historical controls treated with
offers advantages over a sequence of resection EBRT and resection, disease relapse within irra-
and IORT followed by EBRT: diation fields was 0% vs. 43% for IORT vs. non-
IORT patients, and 1- and 2-year survival rates
- exclusion of patients with metastases detected were statistically better with IORT. In another
at restaging, thus sparing the potential risk of MGH IORT analysis [18], 5-year survival rate was
aggressive surgical resection ± IORT; 43% in 42 IORT patients with locally advanced
- potential tumor down staging with a better primary lesions. Both 5-year local control and
chance to deliver IORT without gross residual disease-free survival appeared better if the sur-
disease; geon was able to accomplish a gross total resec-
- possible alteration of implantability of cells tion prior to IORT.
that may be disseminated during surgery; In 1996 Mayo Clinic published a study [19]
- reduction of treatment interval between EBRT which evaluated disease control and survival in a
and IORT (when resection and IORT are done multimodality approach containing IORT for
initially, postoperative complications could locally recurrent, previously unirradiated colorec-
delay postoperative EBRT). tal cancer. From April 1981 and August 1995, 123
patients received IORT as a "boost" to EBRT dur-
ing maximal resection. All received EBRT ± con-
Clinical experiences in colorectal comitant S fluorouracil (SFU). From the analysis
of the results they obtained a central failure (with-
cancer in the IORT field) in 13 of 123 patients (11%) with a
5-year rate of 26% and a local relapse (in EBRT
The interest in the use of IORT in colorectal can- field) in 24 patients (20%) with a s-year rate of
cer is explained by the natural history of the dis- 37%. Distant metastases occurred in 66 patients
ease, burdened with a high rate of local recur- (54%) with a s-year rate of 72%. Median survival
rence and by the possibility of preventing it with was 28 months, with overall survival at 2-, 3-, and
high doses of radiation. s-years of 62%, 39% and 20% respectively.
In Japanese IORT trials instituted in the 1960s, Moreover, disease control within the IORT and the
as well as early U.S. trials, IORT was the sole radi- external field was decreased when the surgeon
ation modality using single doses of 20 to 40 Gy. was unable to accomplish a gross total resection.
In pilot trials started at the Massachusetts General The multimodality treatment including IORT
Hospital (MGH) in 1978 and the Mayo Clinic in resulted in improved local control and long-term
Rochester in 1981, investigators preferred to use survival rates of 20% vs. an expected 5% with con-
IORT as a "boost" dose in combination with con- ventional techniques.
260 Atlas of Endoanal and Endorectal Ultrasonography

In a following experience from Mayo Clinic, a reduced to 20-25% adding 5FU during EBRT;
comparison was made between two sequential besides distant metastases still occurred in 50% of
series: one of 17 patients having no IORT (group patients in view of the high metastatic potential in
1), the other of 55 patients receiving IORT (group patients with locally advanced colorectal cancer.
2). Patients were treated for locally advanced This high risk of systemic failure provides the
non-metastatic primary rectal cancer. There was rational for the addition of maintenance
local disease relapse in 76% of group 1 patients chemotherapy (sFU ± leucovorin) to concomitant
within 18 months of surgery, whereas in group 2 radio chemotherapy and resection.
the local relapse rate was 18%. The improvement In the experience of the Universita Cattolica of
in local control appeared to result in improved 5- Rome [17],47 patients with locally recurrent non-
year survival rates associated with IORT (42% vs. metastatic rectal carcinoma received preoperative
24%; P<O.OOI) [20]. Prognostic factors that had a external beam radiation plus concomitant 5 fluo-
statistically positive impact on disease control rouracil and mitomycin C (FUMIR) ± surgical
and survival in IORT patients included EBRT + resection. Eleven resected patients received an
SFU vs. EBRT alone (local recurrence 11% vs. 24%, IORT boost on the tumor bed. Thereafter, adju-
distant recurrence 36% vs. 77%, s-year survival vant chemotherapy (5 fluorouracil and leucov-
53% vs. 25%, respectively), treatment sequence of orin) was prescribed for a total of 6 to 9 courses.
preoperative EBRT + SFU vs. postoperative EBRT Patients treated with IORT had a significantly bet-
+ 5FU (local recurrence 14% vs. 0%, distant ter local control rate as compared with those who
recurrence 35% vs. 44%, 5-year survival 62% vs. received external beam or brachytherapy boost or
17%), microscopic residual after maximal resec- those who did not receive boost (5-year local con-
tion vs. gross residual (local recurrence 5% vs. trol rate 79% vs. 42% and 23%, respectively;
25%, distant recurrence 39% vs. 75%, 5-year sur- P<0.05). The survival rate was twice as high in the
vival 59% vs. 21%) and colon vs. rectal primary IORT group (5-year overall survival rate 41% vs.
(local recurrence 6% vs. 16%, distant recurrence 19% and 16%, respectively; P value not significant)
28% vs. 58%, s-year survival 63% vs. 38%) [20]. In [17]. Radically resected patients had better local
another Mayo Clinic analysis of 106 patients with control and survival that were respectively at 5
recurrent rectal cancer without extrapelvic dis- years of 68% and 58%; postoperative adjuvant
ease treated with palliative resection ± external chemotherapy improved all clinical outcomes
beam radiotherapy, 42 received IORT as treat- even if not at significant levels.
ment component. Patients treated with IORT In 2001 MGH published a report to update and
showed a lower incidence of local failure (at 3 summarize its experience of a treatment program
years, 40% vs. 93%). The survival rate was signif- of high-dose preoperative irradiation, surgical re-
icantly influenced by the presence of pain as a resection and IORT for recurrent rectal or recto
symptom, the degree of tumor fixation and IORT sigmoid carcinoma [22]. From June 1978 to
delivery and by the amount of residual disease (5- February 1997, in 49 patients who received IORT
year survival: IORT 19% vs. no IORT 7%; the 5-year overall survival, local control and dis-
P<O.OOI) [21]. ease-free survival rates were 27, 35 and 20%
That the amount of residual disease influenced respectively, with a significant difference between
local control and survival was also demonstrated the s-year local control of the completely resected
in a Mayo IORT analyses [15] in which 56 patients group with negative margins (56%) and the par-
received IORT combined with EBRT ± 5FU and tially resected group, including patients with
surgical resection. The 3- and 5- year survival in microscopic residual disease (17%), thus confirm-
patients with micro-residual was 69% vs. 28% for ing the literature data.
those with gross residual with median survival of Despite the good results of these important
67 months vs. 22 months. The study confirmed studies, it is necessary to underline that a ran-
that both overall survival and disease control are domized trial of the effect of IORT in recurrent or
improved with the addition of IORT to standard locally advanced colorectal cancer still does not
treatment. In the same study another important exist.
evaluation regarded disease persistence or relapse In 2002 a Norwegian study [23] criticized the
within the IORT and EBRT fields and distant con- heterogeneity of IORT studies and the confusion
trol. In the IORT series the incidence of disease with which results were presented in a multitude
progression within the irradiation field was of ways, and difficult to compare. They proposed
 Section IX' Combined modality therapy for rectal cancer 261

to analyse survival and local recurrence in a metastatic colorectal cancer, 34% s-year survival
group of 107 patients with recurrent rectal cancer rate reported here suggests that aggressive locore-
with regard to surgical resection stages and IORT. gional approaches which include IORT should be
They found that survival was nearly identical for investigated further in randomized study and that
the total IORT and non-IORT groups with as-year patients with isolated locally advanced nodal
survival of approximately 30%. For patients with metastases from colorectal cancer should not be
Ro, R1 and R2 resections the s-year survival was excluded from curative-intent treatment pro-
around 60, 20 and 0%, respectively for both grams.
groups. There was no statistical significant differ-
ence between IORT and non-IORT groups for
local recurrence too. This study did not reveal any Conclusion
effect of IORT, but it is not a randomized study,
thus still indicating the need for randomization.
Finally, the recent study of Mayo Clinic on Encouraging trends exist in different IORT stud-
IORT in patients with advanced unresectable or ies with regard to improvement in local control
recurrent colorectal cancer involving lymphnodes and survival of locally advanced primary or
focused its role in promoting long-term survival recurrent colorectal cancer. A combined multidis-
after curative-intent, combined treatment [241. ciplinary approach is mandatory for this aim.
The Authors identified 48 patients with advanced Given the poorer results in patients with gross
paraaortic and mesenteric nodal metastases who residual disease, efforts should focus on strategies
had been treated, between April 1981 and January to increase the likelihood of gross total resection
2000, with EBRT ± chemotherapy and IORT after such as adding chemotherapy during EBRT. Also
maximal tumor resection. At 3 years, the local con- dose modifiers in conjunction with IORT (sensi-
trol within the external beam and the central con- tizers, etanidazole) have to be explored to reduce
trol within the intraoperative radiotherapy fields the risk of relapse within the irradiation field.
were 81% and 93%, respectively. Macroscopically Distant relapse remains a problem, so adjuvant
complete resection and colonic primary sites were chemotherapy should be evaluated as a means of
predictors of survival and disease control. The improving tumor response rate and survival.
median survival time and s-year survival rate Finally randomized studies are claimed to proper-
were 3S months and 34%, respectively. Despite the ly confirm the role of IORT in locally advanced or
generally poor prognosis associated with recurrent colorectal cancer.

References 6. Goldson A. Preliminary clinical experience with intra-

operative radiotherapy. J Nat Med Assoc 1978;
1. Comas C, Prio A. Irradiation roentgen preventive 70:408-16
intra-abdominale, apres l'intervention chirurgicale 7. Willet CG, Gunderson LL, Busse PM, Nagorney D,
dans un cas de cancer de l'uterus. Communication an Tepper JE, Calvo FA. IOERT treatment factor. In:
III an. Congres International d'Electrologie 1906. Current Clinical Oncology: Intraoperative Irradiation:
Barcelona: Imprenta Francesco Badia 1907; 5-14 Techniques and Results. Ed. Gunderson LL, et al.
2. Beck C. On external roentgen treatment of internal Totowa, NJ: Humana Press 1999.
structures (eventration treatments). NY Med J 1909; 8. Tepper JE, Gunderson LL, Goldson AL, et al. Quality
89:621 control parameters of intraoperative radiation thera-
3. Eloesser L. The treatment of some abdominal cancers py. Int J Radiat Oncol BioI Phys 1986; 12:1687-95
by irradiation through the open abdomen combined 9. Emami B, Lyman J, Brown A, et al. Tolerance of normal
with cautery excision. Ann Surg 1937; 106:645-52 tissue to therapeutic irradiation. Int Radiat Oncol BioI
4. Barth G. Erfahrungen und Ergebnisse mit der Phys 1991; 21:109-22
Nahbestrahlung operative freigeleger Tumor. 10. Sindelar WF, Johnstone PAS, Hoekstra HJ, Kinsella TJ.
Strahlentherapie 1959; 91:481-527 Normal tissue tolerance to intraoperative irradiation.
5. Abe M, Takahashi M. Intraoperative radiotherapy: the In: Current Clinical Oncology: Intraoperative
Japanese experience. Int J Radiat Oncol BioI Phys 1981; Irradiation: Techniques and Results. Ed. Gunderson
7:863-8 LL, et al. Totowa, NJ: Humana Press 1999.
262 Atlas of Endoanal and Endorectal Ultrasonography

11. Gunderson LL. Past, present and future of intraopera- 18. Wallace HJ., Willet CG., Shellito PC., Co en JJ., Hoover
tive irradiation for colorectal cancer. Int J Radiation HC. Intraoperative radiation therapy for locally
Oncol BioI Phys 1996;34:741-4 advanced recurrent rectal or rectosigmoid cancer. J
12. Gunderson LL, Cohen AM, Dosoretz DE, et al. Surg OncoI1995;60:122-7
Residual, unresectable, or recurrent colorectal cancer: 19. Gunderson LL, Nelson H, Martenson JA, Cha S,
external beam irradiation and intraoperative electron Haddock M et al. Intraoperative electron and external
beam boost ± resection. Int J Radiat Oncol BioI Phys beam irradiation with or without 5-fluorouracil and
1983;9:1597-606 maximum surgical resection for previously unirradi-
13. Willett CG, Shellito PC, Tepper JE, Eliseo R, Convery K, ated, locally recurrent colorectal cancer. Dis Colon
Wood we. Intraoperative electron beam radiation Rectum 1996;39:1379-95
therapy for primarily locally advanced rectal and rec- 20. Farouk R, Nelson H, Gunderson LL. Aggressive multi-
tosigmoid carcinoma. J Clin OncoI1991;9:843-9 modality treatment for locally advanced irresectable
14. Gunderson LL, Martin JK, Beart RW, et al. External rectal cancer. Br J Surg 1997;84:741-9
beam intraoperative electron irradiation for locally 21. Suzuki K, Gunderson LL, Devine RM, et al.
advanced colorectal cancer. Ann Surg 1988;207:52-60 Reoperation ± irradiation for locally recurrent rectal
15. Gunderson LL, Nelson H, Martenson JA, et al. Locally cancer. Cancer 1995;8:118-27
advanced primary colorectal cancer: intraoperative 22. Lindel K, Willett CG, Shellito PC et al: Intraoperative
electron and external beam irradiation ± 5Fu. Int J radiation therapy for locally advanced recurrent rectal
Radiat Oncol BioI Phys 1997;37:601-14 or rectosigmoid cancer. Radiother OncoI2001;58:83-7
16. Azinovic I, Calvo FA, Aristu JJ, et al. Intraoperative 23. Wiig IN, Tveit KM et al. Preoperative irradiation and
radiation therapy as a treatment component in prima- surgery for recurrent rectal cancer. Will intraoperative
ry rectal cancer: ten-year experience. (Personal radiotherapy (IORT) be of additional benefit? A
Communication) prospective study. Radiother OncoI2002;62:207-13
17. Valentini V, Morganti AG, De Franco A, et al. 24. Haddock MG, Nelson H, Donohue JH et al.
Chemoradiation with or without intraoperative radia- Intraoperative electron radiotherapy as a component
tion therapy in patients with locally recurrent rectal of salvage therapy for patients with colorectal cancer
carcinoma: prognostic factors and long term outcome. and advanced nodal metastases. Int J Radiat Oncol
Cancer 1999;86:2612 BioI Phys 2003;56:966-73
 Invited commentary
1.1. Gunderson, M.G. Haddock

The preceding chapter by Valentini et al. presents ysis from the Radium Hospital in Norway in
the rationale for and results of IORT as a compo- which apparent advantages were not seen in IORT
nent of treatment for patients with locally vs. non-IORT patients with recurrent rectal can-
advanced primary and locally recurrent rectal cer. The authors concluded by encouraging the
cancers. Their philosophies with regard to the evaluation of:
potential advantages of IORT plus external beam
irradiation (EBRT) over EBRT alone and the need 1) concurrent chemotherapy during preoperative
for a multidisciplinary team approach closely EBRT to increase the rate of Ro and R1 resec-
mirror the philosophies in the two U.S. institu- tions;
tions in which I was involved with IORT efforts 2) dose modifiers in conjunction with IORT (sen-
(Massachusetts General Hospital [MGH] and sitizers) to reduce the rate of in-field relapse;
Mayo Clinic). In addition, we are in general agree- 3) adjuvant chemotherapy to decrease the rate of
ment with regard to selection criteria for IORT systemic disease relapse.
and sequencing of treatment when both IORT and
EBRT are utilized. Information was presented on In the remainder of the commentary, we will
the delivery of IORT and new technologies that present the rationale for and results of WERT as a
include mobile electron accelerators and high- component of treatment in our institution,
dose-rate brachytherapy machines (HDR-IORT) including a discussion of the marked change in
that can be brought into an operating room set- technologies for WERT from 1978 to the present.
ting. The tolerance issues pertaining to IORT were WERT results at MGH and HDR-IORT results at
briefly discussed. Memorial Sloan Kettering Cancer Center
After laying an excellent foundation with the (MSKCC) will be summarized. Results in three
above sections, Valentini et al. then summarized additional European series will also be briefly
nicely the results of a number of institutions that presented (Pamplona, French IORT Group,
use an electron beam (WERT) as the method of Eindhoven). Future directions in clinical research
IORT. Data was presented from WERT series in will also be discussed.
the U.S. (MGH, Mayo Clinic) and the Catholic
University of Rome. The authors note that
improvements in local control with WERT and Rationale for 10RT
maximal resection, when combined optimally
with EBRT plus concurrent and maintenance (IOERT or HDR-IORT)
chemotherapy, appear to translate into survival
advantages for patients with both locally Most of the major irradiation advances in the
advanced primary and locally recurrent rectal treatment of cancer have been due to differences
cancers in series from these three institutions. in dose distribution between tumor and dose-lim-
The authors also briefly discussed a separate anal- iting normal tissues. For most tumors, the likeli-
264 Atlas of Endoanal and Endorectal Ultrasonography

hood of obtaining local control improves as Gy in 1.8 to 2.0 Gy for microscopic residual, 70
increasing irradiation doses can be delivered [1,21. to 90 Gy for gross residual or un-resected dis-
In many clinical situations, however, the dose of ease).
irradiation that can be safely delivered is limited 3) IORT will be performed at the time of a
by virtue of dose-limiting normal tissues in close planned surgical procedure.
proximity to the tumor volume. IORT with either 4) The IORT plus EBRT technique would theoret-
electrons (IOERT) or high-dose-rate brachythera- ically result in a more suitable therapeutic
py (HDR-IORT) evolved as an attempt to achieve ratio between cure and complications by per-
higher effective doses of irradiation while dose- mitting direct irradiation of marginally resect-
limiting structures are surgically displaced [3, 41. ed or un-resected tumor, with single or abut-
For patients with locally advanced abdominal ting fields, while surgically displacing or
or pelvic malignancies in whom all disease cannot shielding dose-limiting structures or organs.
be removed surgically with negative margins, 5) There is no evidence of distant metastases or
EBRT (± chemotherapy) is often only palliative, peritoneal seeding (rare exceptions: resectable
since doses greater than 45 to 50 Gy in 25 to 28 single organ metastasis, excellent chemothera-
fractions cannot be delivered safely. If microscop- py options, slow progression of systemic dis-
ic residual exists after gross total resection, EBRT ease).
doses necessary to accomplish local control are
~60 Gy in 1.8 to 2 Gy fractions. Dose requirements In addition to the disease sites mentioned by
would be even higher if gross residual remains Valentini et al., IOERT is also utilized as a compo-
after maximal resection. With doses ~60 Gy, the nent of treatment at Mayo Clinic for selected
radiation tolerance of numerous organs and patients with locally recurrent gynecologic and
structures in the abdomen and pelvis would be renal malignancies as well as locally advanced
exceeded [51. esophagus/E-G junction and biliary cancers.
A preferred treatment alternative for patients
with locally advanced malignancies is to give tol-
erable EBRT doses of 45 to 50 Gy preoperatively Sequencing of treatment
(1.8 Gy fractions) and deliver IORT as a supple-
ment at the time of surgical exploration and max-
imal resection [3,41. Dose-limiting organs can be We are in agreement with Valentini et al. that opti-
excluded as a result of surgical mobilization or mal sequencing of surgery and EBRT for locally
shielding. This unique approach thereby allows an advanced cancers should be discussed and deter-
increase in local control (local control curve shift mined at the time of a joint multi-specialty con-
to the left) with a lower risk of complications than sultation involving a surgeon, radiation oncolo-
with an EBRT-only approach (complication curve gist and medical oncologist. This allows input
shift to the right). from all specialists with regard to studies that
would be helpful for IORT and EBRT treatment
planning as well as whether IORT may be appro-
Patient selection and evaluation priate. For many patients with locally advanced
colorectal cancer, preoperative EBRT of 45-50 Gy
in 1.8 to 2.0 Gy fractions (± chemo as indicated by
The following general criteria have guided the disease site) followed by exploration and resec-
selection of appropriate patients for IORT at Mayo tion in 3-5 weeks offers theoretical advantages
Clinic [3, 41: over a sequence of resection and IORT followed
by EBRT. Potential advantages are as follows:
1) Surgery alone will not achieve acceptable local
control (Le., ~microscopic residual disease 1) deletion of patients with metastases detected
after maximal resection). By definition, there at the restaging work-up or laparotomy, thus
must be no medical contraindications for a sparing the potential risks of aggressive surgi-
surgical attempt at gross total resection. cal resection ± IORT;
2) EBRT doses needed for adequate local control 2) possible tumor shrinkage with an increased
following subtotal resection or no resection possibility of achieving a gross total resection;
would exceed normal tissue tolerance (60 to 70 3) potential alteration of implantability of cells
 Section IX' Combined modality therapy for rectal cancer 265

that may be disseminated at the time of a ment is placed into clearer perspective by a com-
marginal or partial surgical resection; parison with tumor-related morbidity. For
4) reduction of treatment interval between EBRT instance, when EBRT is used as the main treat-
and 10RT (if resection/IORT precede EBRT, a ment modality for locally advanced rectal can-
postoperative problem could delay EBRT/ cers, more than 90% of patients have local persis-
chemo excessively). tence or relapse of disease and most are dead in 2
to 3 years (end result is nearly 100% tumor related
morbidity/mortality).
IORT and EBRT technique Peripheral nerve is the principal dose limiting
normal tissue for IOERT in the pelvis ± retroperi-
and tolerance toneum [7-10]. These observations have been made
in animal studies from National Cancer Institute
Irradiation doses and technique (NCI) and Colorado State University (CSU) [8,9]
and clinical analyses from Mayo Clinic Rochester
The method of EBRT has been fairly consistent in [10] and the NCI [8]. When treating the pelvic side-
most u.s. and European single institution and wall for marginally resected rectal cancer, periph-
group studies. In the Mayo Clinic and MGH trials, eral nerves are often adjacent to or involved by
doses of 45 to 54 Gy are delivered in 1.8 Gy frac- tumor. Even when uninvolved, it is usually impossi-
tions,5 days per week over 5 to 6 weeks in patients ble to move or shield nerves from the IOERT field.
who have had no prior irradiation. With GI malig- In Mayo Clinic Rochester analyses of IOERT
nancies, chemotherapy is often given during regimens in 178 patients with locally advanced pri-
external irradiation with 5FU based regimens mary or locally recurrent colorectal cancer [11,12],
(bolus vs. infusion 5FU; alone vs. combined with tolerance data suggested a relationship between
other drugs - Leucovorin, Cisplatin, Mito C). For IOERT dose and the incidence of grades 2 or 3 neu-
previously irradiated patients, an attempt is made ropathy (Table IX.13; EBRT factors appeared con-
to re-irradiate with low-dose preoperative EBRT stant). The incidence of grades 2-3 neuropathy was
doses of 20 to 30 Gy in 1.8 to 2.0 Gy fractions (plus -5% in primary plus locally recurrent patients with
concurrent infusion 5FU). an IOERT dose of 10-12.5 Gy versus - 20% with a
The dose of IOERT and HDR-IORT and the dose of 15-20 Gy (p=om). This trend is consistent
energy of IOERT are dependent on the amount of with animal data that suggest a correlation between
residual disease remaining after maximal resec- IOERT dose and the incidence of clinical and elec-
tion, and on the EBRT component that is feasible. tro-physiologic neuropathy in dogs [8-10]. Grade 2
For patients who can or have received preopera- neuropathy was usually manifested as pain requir-
tive doses of 45 to 54 Gy (1.8 Gy fractions, five days ing narcotics, and not infrequently occurred with a
per week), the 10RT dose usually varies from 10 to delayed onset of 6-9 months after IOERT. Grade 3
20 Gy: :::; microscopic residual- 10 to 12.5 Gy, gross neuropathy (uncontrolled pain, severe weakness)
residual 15 to 20 Gy. In previously irradiated was uncommon except with an IOERT dose of
patients, the IORT dose is usually 15 to 20 Gy if 20Gy, when the incidence was 22% (2 of 9 patients).
EBRT doses of 20 to 30 Gy can be safely given pre-
or postoperatively. Summary - Many recurrent rectal cancer
The biologic effectiveness of single-dose IORT patients who are candidates for IOERT present
is considered equivalent to 1.5 to 2.5 times the with pain from recurrent tumors due to neurolog-
same total dose of fractionated EBRT [6]. The ical tumor compression or invasion. While all
effective dose in the IOERT boost field, when potential IOERT patients are given an informed
added to 45 to 50 Gy given with EBRT, is 60 to 70 consent about possible nerve related side effects,
Gy for an IOERT dose of 10 GY,75 to 87.5 Gy with they are aware that with uncontrolled tumor they
a 15 Gy boost, and 85 to 100 Gy with 20 Gy IOERT. will often have similar side effects. On the basis of
both human and animal data, when a full compo-
Dose-limiting structures nent of EBRT options exist, (Le. 45 to 54 Gy frac-
(IORT Tolerance) tionated EBRT can be delivered), an 10RT boost
dose of 10 to 20 Gy should be used dependent on
In patients with locally advanced malignancies, the amount of tumor remaining after maximal
the issue of morbidity following aggressive treat- resection. If a marginal gross total resection can
266 Atlas of Endoanal and Endorectal Ultrasonography

Table IX.13. Colorectal IOERT, IOERT dose vs. neuropathy (grades 2 or 3) (Mayo Clinic results)

IOERTDose
$12.5 Gy ~15Gy
Disease Presentation N° (%) N° (%) Pvalue

Primary* 1/29 (3) 6/28 (21) 0.03

Recurrent, No Prior EBRTA 2/29 (7) 191101 (19) 0.12
Primary + Recurrent 3/58 (5) 251129 (19) 0.01

*57 IOERT fields in 55 evaluable patients. A130 IOERT fields in 123 patients
Incidence of grade 3 neuropathy by dose: :::;12.5 Gy - 0 of 9,15 or IJ.5 Gy - 1 of 19 or 5%, 2:20 Gy - 2 of 9 or 22%
Modified from references 10-12

be accomplished after full dose preop EBRT, the The HDR-IORT dose at MSKCC has been 12 or IS
IORT dose can be limited to 12.5 Gy or less, with a Gy as calculated at a O.S cm distance from the
decreased risk of neuropathy. For patients with HAM applicator surface [23].
gross residual disease, when a full component of
preoperative EBRT has been feasible, the IOERT
dose is usually limited to 15-17.5 Gy, in view of an Primary unresectable cancers
apparent increase in motor neuropathy at the 20
Gy IOERT dose.
MGH results - EBRT + resection, IOERT

Colorectal IORT (IOERT, Sixty-four patients with locally advanced primary

rectal cancer had full-dose preoperative EBRT
HDR-IORT) (alone or plus SFU) followed by resection and
IOERT at MGH [13]. The 5-year actuarial local
EBRT has been combined with resection and control (LC) and disease-specific survival (DSS)
chemo for locally advanced colorectal cancers. In for 40 patients with complete resection plus
separate series using EBRT alone or combined IOERT was 91% and 63%, respectively (Table
with systemic therapy, the local relapse rate was IX.14). For 24 patients undergoing partial resec-
>90% in evaluable patients. Although a combina- tion, LC and DSS correlated with the extent of
tion of surgical resection and EBRT (alone or residual cancer: 6S% and 47%, respectively, for
combined with 5 fluourouracil [5FU]) for patients microscopic residual disease, and S7% and 14%,
with residual disease after subtotal resection or respectively, for gross residual.
initially unresectable disease produces better
local control than no resection, local relapse
remains too high at 30-50%. For locally recurrent Mayo Clinic results -EBRT ± 5FU,
rectal cancers, EBRT alone or plus chemotherapy Resection, IOERT
results in excellent short-term palliation (6-12
mo) but rare long-term survival (0-5%,5 yr). In a Mayo Clinic comparison of 17 non-IOERT
In an attempt to decrease local recurrence and 56 IOERT + EBRT patients with locally
and improve survival, institutions in the U.S., advanced primary rectal or colon cancers [11],
Europe, Japan, and Scandinavia have combined local control was 24% vs. 87%, median survival
an IOERT or HDR-IORT boost with fractionated was 18 vs. 40 months, and the 3-year survival was
EBRT (45 to 50 cGy in 1.8 Gy fractions), with or 24% vs. 5S% (Table IX.lS) [11, 12, 16, 17, 19]. All
without resection [11-23]. In U.S. studies, the relapses in non-IORT patients occurred before 18
IOERT dose varies from 10 to 20 Gy dependent months. Prognostic factors that had a statistically
on the amount of residual after maximal resec- positive impact on disease control and survival in
tion: microscopic, 10 to 12.S Gy; gross <2 cm, IS IOERT patients included EBRT + SFU vs. EBRT
Gy; gross ~2 cm or unresectable, 17.S to 20 Gy. alone, treatment sequence of preoperative EBRT +
 Section IX' Combined modality therapy for rectal cancer 267

Table IX.I4. Primary colorectal IOERT - disease control and survival by degree of resection

MGH - Rectal Mayo - Colorectal

(5-yr Actuarial) (5-yr Actuarial)
Degree of Resection, #Pts LC (%) DSS (%) # Pts LC (%) DF (%) OS (%)
Amount Residual

No Tumor 2 100 0 100

Complete Resection (Ro) 40 91 63 18 93 54* 69

Partial Resection: 24 63 35
- Microscopic residual (R1) 17 65 47 19 86 50* 55
- Macroscopic residual (R2) 7 57 14 16 73 83* 21

No Resection 0

Total 64 56 84 59 46

Pts = patients; LC = Local Control; DSS = Disease Specific Survival; DF = Distant Failure; OS = Overall Survival
*3-year actuarial DF of 43%, 38%, and 66% for complete resection, microscopic residual or gross residual
(Modified from Willett CG, Shellito PC and Gunderson LL, reference 13)

5FU vs. postoperative EBRT + 5FU, colon vs. rectal ed, as seen in Table IX.14 [11]. The 5-year survival
primary and ::;; microscopic residual disease after for the entire group of 56 colorectal IOERT + EBRT
maximal resection [11]. patients was 46%. Patients with microscopic or
The impact of degree of resection and amount less residual fared better than those with gross
of residual disease on disease control and survival residual with a 5-year actuarial overall survival of
in the Mayo Clinic IOERT series was also evaluat- 59')10 vs. 21% (P=O.0005). Relapse within an irradi-

Table IX.IS. Locally advanced primary and recurrent colorectal cancer - EBRT ± IOERT (Mayo Clinic results)

Treatment Ref. Pts. Median Survival % Relapse (3 yr) %

at risk (mo) 2-yr 3-yr 5-yr Local Distant

Primary Disease
External (EBRT)* * 17 18 35 24 24 76 59

EBRT + IOERT 11 56 40 70 55 46 16 49

Localized Recurrence
a) Suzuki et al. 17
No IOERT 64 17 26 18 7 93 54
IOERT ± EBRT 42 30 62 43 19 t 40 60
b) IOERT + EBRT
No prior EBRT 12 123 28 62 39 20 II 25 64
Prior EBRT 16 51 23 48 28 12 ** 49 71

*All deaths within 30 mo; LF range 3-15 mo; DF range 3-17 mo; t p =.0006
II5-yr survival of 27% in 159 recurrent rectal cancer patients in later analysis [191
** 5-yr survival of 20% in 145 recurrent rectal, prior EBRT patients in later analysis [191
268 Atlas of Endoanal and Endorectal Ultrasonography

ation field occurred in 4 of 16 patients (2S%) who was 24% and 66%, respectively. No correlation
presented with gross residual after partial resec- between local control and post-treatment extent
tion vs. 2 of 39 (S%) with microscopic or less resid- of tumor extension into or beyond the rectal wall
ual after gross total resection (P=O.OI). and/or lymph node involvement was observed. In
A separate Mayo Clinic analysis was done the most recent MGH analysis, 47 patients with
regarding the use of EBRT (with or without locally advanced rectal cancer received 4S to SO.4
chemotherapy) alone or plus IOERT as a supple- Gy preoperative EBRT (usually with concurrent
ment to maximal resection for 103 patients with sFU based chemotherapy) and had complete
locally advanced colon cancers [14]. The s-year resection with clear resection margins by patho-
actuarial local relapse rate was 10% for patients logical exam. None of the patients received IOERT
with no residual disease, S4% for patients with since the response to preoperative EBRT was
microscopic residual, and 79% for patients with favorable or IOERT was not technically feasible.
gross residual (p <0.0001). For 24 patients with no residual tumor or tumor
For patients with residual disease, local relapse confined to the rectal wall after preoperative
occurred in only 11% of patients receiving IOERT EBRT, the s-year actuarial local relapse rate was
plus EBRT compared with 82% of patients receiv- only 13%. In contrast, the s-year actuarial local
ing only EBRT (P=O.02). The s-year actuarial sur- relapse rate was 68% for 27 patients with persis-
vival rate was 66% for patients with no residual tent transmural tumor and/or lymph node metas-
disease, 47% for patients with microscopic residu- tases despite a favorable response to preoperative
al and 23% for patients with gross residual disease EBRT and no clearly defined indication for IOERT
(P=0.0009). The s-year survival rate for patients at surgery (i.e., no remaining tumor adherence or
with residual disease was 76% for patients receiv- compromised soft tissue margins). Therefore, the
ing IOERT plus EBRT and 26% for patients receiv- extent of tumor regression after preoperative
ing EBRT alone (P=O.04). EBRT is no longer used at MGH as an absolute
A more recent Mayo Clinic analysis evaluated indication or contra-indication for IOERT.
IOERT as a component of treatment for 48 col-
orectal cancer patients with nodal relapse or
bulky mesenteric or para-aortic nodal disease at Locally recurrent colorectal cancer
time of presentation [IS]. Patients received preop-
erative EBRT plus concurrent sFU based
chemotherapy followed by an attempt at resection MGH results - IOERT ± EBRT
and IOERT. Salvage at s-yrs was achieved in -3S%
of all patients at risk and -40% of those with In an MGH IOERT analyses, S yr actuarial DFS
colon cancer. This strongly supports the use of was 16% and OS of 30% in 41 patients with locally
abdomen CT as a component of follow-up III recurrent lesions [16] (exceeds expected long-
resected high-risk colon cancer patients. term survival of :::;s% when treated with standard
techniques). Patients with gross residual had five-
year LC and DFS of 21 and 7% vs. 47 and 21% with
Local relapse vs. disease extent after preop clear or micro positive resection margins.
EBRT (no IOERT)

The incidence of local relapse as a function of dis- Mayo Clinic results - IOERT ± EBRT
ease extent after preoperative EBRT alone or plus
concurrent SFU (no IOERT) has been evaluated in Additional data supporting the use of IOERT for
three different MGH analyses [13]. For patients locally recurrent disease is found in a Mayo Clinic
with locally advanced (T4) rectal cancer treated analysis of 106 patients with palliative resection of
with preoperative EBRT and curative resection in locally recurrent rectal cancers from 1981-1988, as
the initial series, S of 8 patients (62.S%) with per- seen in Table IX.lS [17]. None had evidence of
sistent gross tumor extension beyond the rectal extrapelvic disease and 42 received IOERT as a
wall relapsed in the pelvis. In 28 patients with component of treatment. EBRT was given in 41
tethered (T3 vs. early T4) rectal cancers treated patients C~4S Gy in 38). Significant factors with
with preoperative EBRT and resection, the s-year regard to 3-yr and s- yr survival included amount
actuarial local relapse and disease-free survival of residual (micro versus gross - 3 yr, 44 vs. 26%; S
 Section IX· Combined modality therapy for rectal cancer 269

yr, 33 vs. 9%; P=0.032) and IOERT versus none (3 SR was 12% vs. 3S% (P=0.03). The actuarial local
yr,43 vs. 18; 5 yr, 19 vs. 7%; p=0.0006). For IOERT control rate at 36 months was increased for
patients,3 yr SR with gross residual or presenta- patients treated with EBRT + IOERT vs. IOERT
tion with pain was 44% and 43%, respectively vs. alone at 30% vs. 0% (P=0.03). In patients treated
- 15 % in non-IOERT patients. with adjuvant EBRT, the preoperative sequence
Mayo Clinic results in 123 locally recurrent col- appeared to cause better survival rates than post-
orectal cancer patients without prior EBRT [12,16] operative EBRT with median survival of 23 months
are seen in Table IX.lS. Survival results paralleled vs. 10 months, respectively (P=O.Ol).
those seen in the earlier Suzuki analysis with
2-year survival of 62% and s-year survival of 20%. French lORT Group
In an IOERT series of 51 patients with recurrent
colorectal cancer who had received prior EBRT, Similar findings have been observed by investiga-
median survival was 23 months and s-yr survival tors from the French IORT group as seen in Table
was 12% [IS]. IX.16 [20-22]. In 30 patients treated with IORT
A series of 304 patients with locally recurrent alone, no long-term survivors were found after 42
rectal cancer receiving IOERT as a component of months vs. 70% for the 16 patients treated with
treatment was presented by Haddock et al. at IOERT plus EBRT.Actuariallocal control was 60%
ISIORT 2002, in Aachen [19]. In the 159 without for EBRT plus IORT vs. 0% with IORT alone.
prior EBRT, s-yr survival was 27%, and in the 145
with prior EBRT, s-yr survival was 20%. Eindhoven (The Netherlands) WERT Results
Improvements in long-term survival appeared to
be related to more intensive treatment in recent From 1994 to 1995, 37 patients with locally recur-
years (patients usually receive PVI SFU during rent rectal cancer (without distant metastasis)
EBRT in both groups; prior EBRT patients often received IOERT as a component of treatment in
receive re-irradiation EBRT doses of 25.2 to 30.6 Eindhoven [22]. Patients with no prior irradiation
Gy plus PVI SFU), including the use of mainte- (n=17) received preoperative EBRT of 16 to 60 Gy
nance systemic chemotherapy. Local control (16 Gy - 1; 50.4 Gy - 15; 60 Gy - 1) alone or with
appeared to improve in prior EBRT patients concomitant sFU based chemotherapy (n=2).
retreated with >25.2 Gy (p=.06). Those who had received prior EBRT (n=20) either
had no further EBRT (n=IS) or in the latter part of
European results - IORT ± EBRT the study were given 30 Gy in 2 Gy fractions (n=s).
Disease control and survival at 3 years were local
The current philosophy in Italy and other parts of control (LC) of 60%, disease-free survival (DFS) of
Europe is closely related to the U.S. concept, which 32%, and overall survival (OS) of SS% (Table IX.16)
utilizes IORT as a segment of a multidisciplinary [20-22]. The ability to achieve a gross total resec-
approach in cancer management. A component of tion with negative (Ro) or microscopically positive
EBRT ± sFU-based chemotherapy is always resection margins (R1) impacted all three parame-
attempted, either before or after surgery, if no pre- ters when compared with R2 resection. This was
vious EBRT has been delivered. shown as an improvement in OS (Ro/Rl - 74% vs.
R2 - 35%, P <0.05), DFS (Ro/R1- 54% vs. R2 - 11%,
Pamplona Experience P = o.oooS), and LC (p=O.Ol). The ability to
achieve a gross total resection was higher for
In a Pamplona series [20], 37 patients were treated patients who received preoperative EBRT (Ro/RI
with IOERT for locally recurrent colorectal carci- vs. R2, P=O.OOl). Patients who received preopera-
noma; 12 were treated with IOERT alone, since tive EBRT had improved survival and disease con-
they had received previous EBRT for their prima- trol (OS, p=o.oos, DFS, p=O.OOI; metastases,
ry disease (Table IX.16) [20-22]. The remaining 25 p=0.00007; LC, p=o.oS).
were treated with EBRT + IOERT (11 with postop-
erative EBRT and 14 with preoperative chemoradi- Colorectal HDR-IORT, MSKCC Results
ation; 40-50 Gy in I.S-2 Gy fractions). Median sur-
vival from initiation of treatment was 15 months The predominant initial experience in HDR-IORT
vs. 22 months for patients treated with IORT alone at MSKCC has been the management of patients
vs. those treated with EBRT plus IOERT, and 3 year with locally advanced primary or locally recurrent
270 Atlas of Endoanal and Endorectal Ultrasonography

Table IX.16. European results with IORT ± EBRT for locally recurrent colorectal cancer

Institution Ref. W of patients LC* (P) Survivalt (P)

A) Pamplona 20
IOERT+EBRT 25 30% 38%
IOERT alone 12 0% 12%
B) France 21
IORT + EBRT 16 61% 68%
IORT alone 30 0% 24%:1:
C) Eindhoven 22 37 60% 58%
EBRT + IORT 22 NS 0.08 NS 0.005
IORT alone 15 NS NS
RO/R1 23 79% 0.01 74% <0.05
R2 14 21% 35%**

*LC: actuarial local control rates; NS = not stated

tSurvival: 3-year actuarial survival rates; :l:No long-term survivors beyond 42 months; ** No DFS beyond 22 months

colorectal cancer [23]. From 1992 through 1996, 112 most ideal situation is to place a facility within the
patients with rectal cancer were explored, and 68 OR suite, as done at Mayo Clinic in Rochester,
were treated with HDR-IORT (primary unre- MGH, MDACC and some European facilities for
sectable - 22; locally recurrent - 46). If there was IOERT and at Beth Israel (NYC) for HDR-IORT.
no prior EBRT, preoperative chemoradiation was Either approach simplifies the treatment of
given followed by maximal resection and HDR- patients, necessitates fewer re-operations (refused
IORT (range 10-20 Gy, median 12 Gy). Early results by some patients and physicians), and avoids
are similar to results with IOERT series. Local con- transportation and sterility problems. It also pre-
trol at 2-years was dependent on margin status vents the need to shut down the outpatient treat-
after maximal resection in both groups of patients ment machine for a "potential" case. However, the
(primary: negative margin - 92%, positive margin dedicated IOERT option in an OR setting is quite
- 38%; locally recurrent: negative margin 82%, expensive if an existing OR has to be retrofitted
positive margin - 19%). Disease-free survival at 2 for proper shielding and a new linear accelerator
years was 69% and 47% for primary and recurrent is purchased as the electron source.
cases, respectively and was again related to margin At Mayo Clinic in Rochester, the transition
status (primary: negative margin vs. positive 77% from a non-dedicated to a dedicated facility
vs. 38%; recurrent: negative margin vs. positive resulted in a marked yearly increase in the num-
71% vs. 0%). Results were best in previously unir- ber of patients who received IOERT as a compo-
radiated patients who could receive all compo- nent of treatment. For the 8 years in which a non-
nents of treatment. dedicated facility was used for IOERT from April
1981 to April 1989, a total of 241 patients received
IOERT (avg of - 30fyr). After several years in the
Future possibilities dedicated facility, the yearly IOERT volumes
ranged from 100-140 patients per year.
New technologies should improve the avail-
Technology and selection of patients ability of IORT from the perspective of cost effec-
tive alternatives. These technologies include
Some of the technical problems and nuisance mobile HDR-IORT units and the mobile IOERT
aspects of IORT, encountered in the 1980s and machines - Mobetron [25] and Novac-7 [26] (Fig.
early 90S, can be overcome with a dedicated or IX.3). The initial Mobetron unit was evaluated at
semi-dedicated IOERT or HDR-IORT facility [24]. UCSF with subsequent units placed in other US
This can be built as an operating room (OR) in the institutions, including Mayo Clinic in Scottsdale
Radiation Oncology Department; however, the (Jan 2002), as well as European and Japanese
 Section IX' Combined modality therapy for rectal cancer 271

Fig. IX.3. The mobile IOERT machine

NOVAC-7

institutions. For the mobile HDR-IORT machine, a doses of 60 to 65 Gy, when a marginal resection
shielded facility is necessary in either the OR area has been performed. The EBRT plus IORT option
or in the radiation oncology department. Instead may improve both local tumor control and nor-
of shielding an entire OR room, however, technol- mal tissue tolerance.
ogy now exists to create a shielded box (room IORT could also potentially be used to replace
within a room) into which the patient can be a component of EBRT in node-negative patients,
placed for the HDR-IORT component of treat- as a method of shortening the course of EBRT.
ment after surgical resection and placement of the European pilot studies have been performed in
HAM applicator have been accomplished. The both breast (TI-2, No) and rectal cancer patients
Mobetron unit has built-in shielding in a C-arm (T3N 0) that demonstrate acceptable tolerance and
design and could theoretically be moved from one local tumor control. Mayo Clinic in Scottsdale ini-
operating room to another, if indicated. tiated a phase II breast cancer study Nov 2002 in
The existence of both dedicated facilities and which an IOERT supplement replaces 1-1.5 weeks
new technologies increases the likelihood of eval- of EBRT boost-dose irradiation.
uating IORT in combination with "curative resec-
tion" and reduced dose EBRT ± chemotherapy in
adjuvant disease settings where adjuvant EBRT Colorectal cancer treatment, primary and
doses necessary to achieve local control approach
or exceed an acceptable level of normal tissue tol- locally recurrent
erance. An excellent example of this philosophy
was the randomized NCI abdominal sarcoma trial Although encouraging trends exist with regard to
in which adjuvant level doses of EBRT resulted in improved LC and SR when IOERT is combined
excessive small bowel morbidity in addition to with standard treatment for locally advanced pri-
poor local control. For lesions of various histolo- mary and locally recurrent colorectal cancers, the
gies, the use of moderate dose EBRT (45 Gy in 25 incidence of systemic failure is ~500/0, and relaps-
fractions of 1.8 Gy over five weeks) plus IORT of 10 es within IOERT and EBRT fields are significant if
to 12.5 Gy may be preferable to high dose EBRT gross resection is not feasible. Early results with
272 Atlas of Endoanal and Endorectal Ultrasonography

HDR-IORT as a component of treatment parallel chemotherapy or other systemic therapy regi-

those with IOERT. In attempts to improve LC, mens need to be evaluated after, if not during
bolus 5FU plus leucovorin, infusion 5FU or other EBRT. This includes the evaluation of newer
enhancing or additive agents should be given dur- chemotherapy agents (i.e., irinotecan, oxaliplatin)
ing EBRT, and studies should be performed to and biologics (bevacizumab and Cetuximab).
evaluate the use of dose modifiers with IORT While we agree with Valentini et al. that it
(sensitizers, other). would be of scientific interest to randomly com-
With both locally advanced primary or locally pare standard treatment ± IORT, such trials did
recurrent colorectal cancers, distant metastases not accrue well in the U.S. or Europe and were
continue to be a major issue. Incorporation of sys- closed. Trials that are feasible will standardize the
temic therapy into locally aggressive treatment aggressive local treatment of EBRT, resection, and
regimens will be necessary in order to maximize IORT with IOERT or HDR-IORT and randomize
disease control and survival. Maintenance optimal chemotherapy, or other systemic therapy,
chemotherapy should become a standard compo- during, as well as after EBRT, and the presence or
nent of treatment regimens, and more aggressive absence of dose modifiers during IORT.

References 10. Gillette EL, Powers BE, Gillette SM, Gunderson LL,
Willett C. Peripheral nerve tolerance: experimental
1.Fletcher GH, Shukovsky LJ. The interplay of radio cur- and clinical. In: Intraoperative Irradiation: Techniques
ability and tolerance in the irradiation of human can- and Results. Ed. Gunderson LL et al. Totowa, NJ:
cers. J Radiol ElectroI1975;56:383-400 Humana Press 1999;165-74
2. Suit HD. Local control and patient survival. Int J 11. Gunderson LL, Nelson H, Martenson J, et al. Locally
Radiat Oncol Bioi Phys 1992;23:653-60 advanced primary colorectal cancer: intraoperative
3. Gunderson LL, Tepper JE, Biggs DJ, et al. Intraoperative electron and external beam irradiation ± 5FU. Int J
± external beam irradiation Curr Probl Cancer 198p:1- Radiat Oncol Bioi Phys 1997;37:601-14
69 12. Gunderson LL, Nelson H, Martenson J, et al.
4. Gunderson LL, Willett CG, Harrison LB, Calvo F. Intraoperative electron and external beam irradiation
Intraoperative irradiation: Techniques and results. with or without 5-fluouracil and maximal surgical
Humana Press 1999;1-550 resection for previously unirradiated, locally recurrent
5. Gunderson LL, Martenson JA. Gastrointestinal tract colorectal cancer. Dis Colon Rectum 1996;39:1379-95
radiation tolerance. In: Radiation Tolerance of Normal 13. Willett CG, Shellito PC, Gunderson L1. Primary col-
Tissues. Ed. Vaeth JM and Meyer JE. Karger, Basel. orectal EBRT and IOERT. In Intraoperative
Front Radiat Ther OncoI1989;23:277-98 Irradiation: Techniques and Results. Ed. Gunderson LL
6. Okunieff P, Sundararaman S, Chen Y. Biology of large et al. Totowa, NJ: Humana Press 1999;249-72
dose per fraction radiation therapy. In: Intraoperative 14. Schild SE, Gunderson LL, Haddock MG et al. The
Irradiation: Techniques and Results. Ed. Gunderson LL treatment of locally advanced colon cancer. Int J
et al. Totowa, NJ: Humana Press 1999;25-46 Radiat Oncol Bioi Phys 1997;37:51-8
7. Tepper JE, Gunderson LL, Goldson AL, et al. Quality 15. Haddock MG, Nelson H, Donohue JH et al.
control parameters in intraoperative radiation thera- Intraoperative electron radiotherapy as a component
py. Int J Radiat Oncol Bioi Phys 1986;12:1687-95 of salvage therapy for patients with colorectal cancer
8. Sindelar WE, Johnston PA, Hoekstra H, Kinsella TJ. and advanced nodal metastases. Int J Radiat Oncol
Normal tissue tolerance to intraoperative irradiation: Bioi Phys 2003;56:966-73
The National Cancer Institute experimental Studies. 16. Gunderson LL, Willett CG, Haddock MG et al.
In: Intraoperative Irradiation: Techniques and Results. Recurrent colorectal: EBRT with or without IOERT or
Ed. Gunderson LL et al. Totowa, NJ: Humana Press HDR-IORT. In: Intraoperative Irradiation: Techniques
1999;131-46 and Results. Ed. Gunderson LL et al. Totowa, NJ:
9. Gillette EL, Gillette SM, Powers BE. Studies at Colorado Humana Press 1999;273-306
State University of normal tissue tolerance of beagles 17. Suzuki K, Gunderson LL, Devine RM, et al.
to IOERT, EBRT, or a combination. In: Intraoperative Intraoperative irradiation after palliative surgery for
Irradiation: Techniques and Results. Ed. Gunderson LL locally recurrent rectal cancer. Cancer 1995;75:939-52
et al. Totowa, NJ: Humana Press 1999;147-64 18. Haddock M, Gunderson L, Nelson H, et al.
 Section IX' Combined modality therapy for rectal cancer 273

Intraoperative irradiation for locally recurrent colorec- 23. Harrison L, Minsky BD, White C, Cohen AM, Enker,
tal cancer in previously irradiated patients. Int J Radiat WE. HDR-IORT for colorectal cancer. In: Intra-
Oncol BioI Phys 2001;49:1267-74 operative Irradiation: Techniques and Results. Ed.
19. Haddock MG, Miller RC, Nelson H, Devine RM, Gunderson LL et al. Totowa, NJ: Humana Press 1999;
Gunderson LL. Intraoperative electron irradiation for 307-14
locally recurrent rectal cancer. ISIORT 2002, Aachen. 24. Gunderson LL, Willett CG, Calvo FA, Harrison LB.
Abstract 8.3 Conclusions and future possibilities. In: Intraoperative
20. Abuchaibe 0, Calvo FA, Azinovic I et al. Intraoperative Irradiation: Techniques and Results. Ed. Gunderson LL
radiotherapy in locally advanced recurrent colorectal et al. Totowa, NJ: Humana Press 1999;527-36
cancer. Int J Radiat Oncol BioI Phys 1993;26:859-67 25. Meurk ML, Goer DA, Spalek G, Cock T. The Mobetron:
21. Bussieres E, Gilly FN, Rouanet P, et al. Recurrences of a new concept for intraoperative radiotherapy. In: Role
rectal cancers: result of a multimodal approach with of Intraoperative Radiation Therapy in the Treatment
intraoperative radiation therapy. Int J Radiat Oncol of Cancer Ed. Vaeth J and Meyer J. S Karger: Basel.
BioI Phys 1996;34:49-56 Front Radiat Ther OncoI1997;31:65-70
22. Mannaerts GHH, Martijn H, Crommelin MA, et al. 26. Fantini M, Santori F, Soriani A, et al. 10RT - Novac 7, a
Intraoperative electron beam radiation therapy for linear accelerator for electron beam therapy. Abstract
locally recurrent rectal carcinoma. Int J Radiat Oncol in 6th International Symposium ofIORT, San Francisco,
BioI Phys 1999;45:297-308 1996
 SECTION X
Treatment options for anal carcinoma

Dentate line
 Treatment options for anal carcinoma
A. Infantino, 1. Pisegna Cerone

Although anal cancer is a rare condition its inci- strated that HPV-positive cancers are associated to
dence rate seems to have increased progressively. a higher cellular spreading rate and to a worse
As described in a recent study carried out in state of aneuploidy compared to HPV-negative
Connecticut [1], the incidence rate is 1.9 in males cancers [3]. Anti-HPVI6 antibodies (peptide E2
and 2.3 in females and is lower in white males and E7) were found in 55% of patients with anal
(0.41/100000) and higher in black females epidermoid cancer [4]. Furthermore, it has been
(0.74/100000). The increased incidence rate has noted that women with previous vulvar/vaginal or
also been confirmed by a study on three cancer cervical tumors are more prone to develop anal
registers in Great Britain where it is emphasized cancer (risk rate 15.4 and 4.3, respectively) and that
that situations at risks are unmarried status, patients with anal cancer can develop subsequent-
homosexual behavior, immunodeficiency syn- 1y a second neoplasia (risk rate 1.4), more fre-
dromes and immunosuppression [2]. quentlyvulvar or vaginal cancer (risk rate 12.3) [5].
Melanoma and adenocarcinoma represent the For many years demolitive surgery has been
most rare forms of anal cancers (15%), while the the treatment of choice for anal cancer. Five-year
majority is represented by squamous cell carcino- survival after recto-anal abdominoperineal resec-
mas (85%), including the cloacogenic tumor tion (APR) has been reported to be between 58
(35%). Human papillomavirus (HPV) is the recog- and 71%. However, we must bear in mind, that in
nized cause of the major part of epidermoid anal the past, an operative mortality rate ranging
carcinoma. HPV-6 and HPV-ll may cause condy- between 2.5% and 8% has been described with the
lomata and low-grade dysplastic lesions, while above-mentioned procedure. Thus, the greatest
HPV-16 and HPV-18 are more frequently involved hesitation toward this kind of surgery is due to the
in the development of high-grade dysplasias and high operative mortality rate, to the sacrifice of
epidemoid carcinoma, as compared with carcino- the anal sphincter necessitating a life long perma-
ma of the cervix. nent colostomy, to the considerable incidence of
A retrospective study carried out by SICP genitourinary disorders and to the poor control of
(Italian Society of Colonproctology) between lymph nodes spreading.
1984-1993, pointed out that 363 patients were Until the presentation of Nigro's protocol,
affected by anal epidermoid carcinoma, mostly radiotherapy (RT) was characterized by severe side
occurring in females. This study confirmed litera- effects. In 1974 Nigro referred to a combined treat-
ture data that women are at higher risk than men ment with chemotherapy (CT) and radiotherapy
(2:1). A possible reason could be that women are [6]. The 3-week period RT treatment with external
more prone to viral infections of the genital appa- fields was contemporaneously carried out for the
ratus. As mentioned previously HPV can be found first five days of therapy with a continuous infu-
in 46% of anal carcinomas, with HPV serotypes 16 sion of 5-fluorouracil (5FU). In addition, the
and 18 more frequently involved. Confirming the patient was given mitomycin C (MMC) as a single
"tumorigenicity" of this virus, it has been demon- bolus on the first day. Such a therapeutic combina-
278 Atlas of Endoanal and Endorectal Ultrasonography

tion, whose acronym is FUMIR, was found apt to Radiochemotherapy (RT-CT) (MMC+SFU), as
sterilize neoplasias in 74% of patients who subse- the sole method of treatment of anal carcinoma
quently underwent abdominoperineal resection [7,17] has raised the percentage of control of ini-
[7]. The excellent results of radiotherapy plus tial tumor >4 cm from 70% to 90% compared to
chemotherapy in epidermoid anal cancer were also the radiation therapy alone, with a s-year survival
confirmed in other studies, which reported that no rate ranging between 6S% and 80%. The above
residual cancer was histologically found in surgical schedule of treatment has been the most
specimens of patients who were operated on with widespread with very encouraging results: 97
an APR after neoadjuvant radio chemotherapy [8, patients with a survival rate of 71 months [18] and
9]. Even in the absence of studies comparing surgi- with anal sphincter function maintained in 8S% of
cal and RT-CT treatments, such results suggest that cases [18], also confirmed by more recent studies
demolitive surgery could have been avoided and even including anal margin cancer [19].
reserved solely for recurrent or persistent tumor The low incidence of this neoplasia makes it
not otherwise curable or for serious cases of local difficult to carry out randomized studies that
radio-toxicity. could help solve many controversies. In order to
Studies carried out on cultures of neoplastic settle the debate between supporters of the sole
cells of human origin as well as clinical surveys on RT and those using the chemoradiation therapy
patients with solid neoplasias have pointed out regimen, the European Organization for Research
the possibility of a more efficient radiotherapeu- and Treatment of Cancer (EORTC) presented the
tic effect in the presence of SFu' The efficacy result of a prospective comparative study between
resulted proportional to the dosage and to the RT (4S Gy + IS-20 Gy) and RT (same doses) + CT
hematologic concentration of sFU [10]. (SFU + MMC) in the treatment of anal tumors.
To reinforce the choice of conservative treat- The 3-year complete response was 53% vs. 77%
ment over demolitive surgery, in agreement with (p=0.0008) and local relapse was 60% vs. 40%,
the national SICP retrospective study and with the respectively. However, survival rate was not statis-
internationally accepted standard, come the tically significantly improved in the group of
results of another cooperative retrospective study patients who received RT +CT [20, 21]. A UKCCCR
on 142 patients from five university surgical cen- (United Kingdom Co-ordinating Committee on
ters in Germany that confirmed that a non surgi- Cancer Research) multicenter study reported sim-
cal approach is recommended for treatment of ilar results [22].
anal cancer [11]. Regardless of the results of the The RTOG-ECOG (Radiation Therapy
above mentioned reports, several authors contin- Oncology Group/Eastern Cooperative Oncology
ued to believe that radical external beam radia- Group) joint study considered the possibility of
tion alone was effective, adopting different excluding MMC from treatment, in order to reduce
dosages and techniques: fixed fields, addition of toxicity; nevertheless the results confirmed the
perineal field, box-technique, rotatory technique utility of MMC associated to 5FU: the two year
and brachytherapy. The doses that have been local relapse was 33% vs. 17% and complete
found to be effective for epidermoid carcinoma of response was 62% vs. 78%, respectively [23].
the anal canal are between so and 70 Gy. Doses Furthermore, an additional attempt of chemoradi-
and total treatment time on inguinal fields ation treatment seems justified before deciding for
depend on the presence of inguinal nodal metas- surgery in the case of residual local disease [24]. A
tases [12]. higher dosage of 5FU (1000 mglmq/24 h) and its
Moreover, the toxicity caused by curative continuous infusion resulted in an improved
radiotherapy does not seem to be correlated to the response between 80% and 95%, and reduced local
patient's age, but to the presence of concurrent relapse incidence to about 10% [25,26].
disease [13]. Although the best result is obtained through
Interstitial brachytherapy, suggested in order the combination of radio and chemotherapy, the
to avoid external radiotherapy-induced side problem of treatment cycles and splits remains
effects, is used to limit irradiation only to the open. A split between two cycles of treatment
tumor volume, thus allowing the use of greater reduces toxicity [27,28] even if it is believed that
doses and prolonged exposure with more efficacy local control of the disease is inversely related to
and reduction of toxicity to the adjacent organs the total treatment time [29]. It is necessary to
[14-16]. stress that pelvic irradiation reduces rectum com-
 Section X· Treatment options for anal carcinoma 279

Table X.I. Results of 22953 EORTC phase II study. The new protocol improved the results of previous trial 22861 [36]

Compliance: 93%

Complete response rate: 90.7%

Grade 3 toxicityies:
• Skin 28 %
• Diarrhoea 12 %
• Hematologic toxicityies 2%

3-year estimated rates for trials 22861 and 22953 are, respectively:
• 68% and 88% for local control
• 72% and 81% for colostomy-free interval
• 62% and 84% for severe late toxicity-free interval
• 70% and 81% for survival

pliance also in asymptomatic patients, and that split of two weeks after 36 Gy. Preliminary results
such conditions worsen with time [30]. showed a 3-5 toxicity degree in 63% of patients,
New strategies with cisplatin-based regimens with one death caused by septicemia; further-
are opening up. The published results using more, after two years, 30% of patients required a
CDDP+5FU instead of MMC+5FU showed that colostomy compared to 7% of the previous study
local response and survival rates were similar for (RTOG 8704) where a split was not foreseen and
both groups, but local relapses were lower with survival rate was lower than 75% [35]. However,
the new combined treatment: 6% vs. 24%, respec- Cummings [28] observed good results with a
tively [31-33]. Gerard et al. [32] reported that the treatment schedule including the split and advo-
introduction of CDDP, as a substitute for MMC, cated that during radio chemotherapy a split
reduced levels of toxicity with similar results in should not be abandoned. Efforts should be made
terms of local response, overall survival, quality of to identify the optimal radiochemotherapic asso-
life and anal sphincter function. These results ciation with good results but lower toxicity. A
confirmed what Gerard already suggested in 1998 recent EORTC study [36], on 43 patients with T2-
that CDDP+5FU regimen should be the new stan- T4, NO-3 tumors and varying cycles of radiother-
dard of treatment instead of FUMIR. More recent- apy (36 Gy over 4 weeks + 5FU 200 mg/m2/days 1-
1y Hung et al. [33] reached the same conclusion. 26 + MMC 10 mg/m2/day 1 + 16 days split, fol-
According to more recent reports, radio lowed by 23,4 Gy over 17 days + 5FU 200
chemotherapeutic treatment must be used irre- mg/m2/days 1-17 and MMC 10 mg/m2/day 1) has
spectively of anal cancer stage [18]. In a prospec- produced very good results as seen in Table X.!.
tive study, Schneider et al. [34] confirmed that a Following Svensson's study [37] which demon-
normal anal sphincter function can be preserved strated the possible advantage of neoadjuvant
in 80% of patients and reported an 80% 5-year chemotherapy prior to CT-RT, and Lee's study [38]
survival rate, a 71% disease free survival and an on RT alone in 23 patients, a phase II multicentric
83% local control rate. Moreover an inverse corre- study by Pfeiffert et al. [39], on patients with local-
lation has been demonstrated between disease ly advanced anal epidermoid cancer, is opening
free survival rate and the need to reduce or inter- new interesting prospects: complete as well as par-
rupt chemotherapy (5FU and MMC). tial response at the end of a cycle of neoadjuvant
Because various trials have reported contra- chemotherapy with 5FU+CDDP was 10% and 51%,
dictory results on the relation between total radi- respectively. Because of these results, two current
ation dose and local disease control, RTOG pro- studies of phase III, sponsored by National Cancer
moted a phase I-II study (RTOG 9208), increasing Institute (USA) and by Federation Nationale des
the total dose from 45-50 Gy of RTOG 8704 to 59 Centres de Lutte Contre Ie Cancer (France), have
Gy and reducing the treatment volume, with a started up and foresee the accrual of 650 patients
 280 Atlas of Endoanal and Endorectal Ultrasonography

great interest. The introduction of the technique to

identify the sentinel inguinal lymph node, in anal-
ogy to breast and melanoma surgery, might allow
more precise staging of disease (Fig. X.l) [41].
Most patients with local recurrence must be
selectively treated by salvage surgery with APR
and permanent colostomy. According to Smith et
al. [42], results of this operation, that often
requires large removal of adjacent organs, are
unsatisfactory. Other authors reported that over-
all 3- and 5-year survival was higher in patients
with residual tumor after RT +CT (72% and 60%,
respectively) compared to patients with local
a recurrence (29% and 0%, respectively). Other
reports seem to suggest that an early intervention
may give more optimistic results [43,44], even if
burdened by a greater number of complications
compared to non-irradiated patients [45].
Nevertheless delays in the diagnosis of a recur-
rent disease lead to more extensive salvage opera-
tions [46]. There are biological and technical
problems related to reconstruction that could
guarantee a primary closure of the wounds, and
b
when possible, restoring an organ function, for
Fig. X.I. In this patients with epidermoid cancer of the anal
example, sexual function with reconstruction of
canal the procedure of the sentinel lymph node mapping was the vagina. The biological problem is the conse-
applied. (a) shows the projection obtained in the course of quence of damages of post-irradiated tissues, with
scintigraphy, and (b) shows the two removed inguinal lymph a higher risk of dehiscence [47]. The technical
nodes, intraoperatively detected by gamma probe problem consists in filling or reconstructing
resected organs, mainly the vagina, if possible, by
in 5 years and 358 patients in 4 years, respectively. using whole undamaged neurovascular bundles.
Among phase III studies, we can even find one In 2 out of 4 patients we operated on with posteri-
sponsored by UKCCCR Anal Cancer Trial Working or pelvic exenteration, a sensate fascio-cutaneous
Party, with a foreseen accrual of 600 patients, flap vascularized by the terminal branches of the
which aims to confirm the good results now internal pudendal artery was used, but with a dif-
achieved with adjuvant chemotherapy. ferent anatomic model [48]. The postoperative
Particular attention goes to the problems due course was uneventful, without infections or
to synchronous or metachronous inguinal lymph dehiscence in any cases. The angle of inclination
node metastases. Among different methods of fac- of the vagina was physiological and natural, sexu-
ing the problem, Gerard et al. [40] proposed a sur- al intercourse was possible; the cosmetic result
gical dissection of synchronous inguinal node was excellent, with no donor-site morbidity (Figs.
metastases followed by CT (5FU+CDDP) + RT X.2, 3). The first patient, aged 42, died 23 months
(45-50 Gy in 5 weeks), 3 weeks after the first CT later of liver and lungs metastases, while the sec-
cycle. In the presence of metachronous inguinal ond one, aged 53, died of multiorgan metastases
node metastases, they used inguinal surgical dis- after I-year follow-up.
section followed by RT (45-50 Gy), but only uni- The highly effective antiretroviral therapy
laterally. The overall 5-year survival rates were (HAART) used in females HIV- with cervical
satisfactory: 54.4% in the synchronous and 41.4% intraepithelial neoplasia [49], as in patients HIV +
in metachronous presentations, respectively. with "anal squamous intraepitheliallesions" [50,
Because inguinal lymph nodes are the most 51] has given interesting results in modifying the
frequent site of relapse and represent one of the clinical course. Furthermore in patients HIV-
most important prognostic factors, every attempt with high-risk warts the use of a vaccine against
to adequately stage anal cancer is looked at with HpsE7 of HPV-16 favored the disappearance of
 Section X· Treatment options for anal carcinoma 281

c d

Fig. X.2. Perineal anal cancer recurrence infiltrating the vagi-

nal wall in a 42Y patient. Once resection was performed (a),
the pudendal flap of the thigh is drawn and incisions made
(b), it is rotated in order to fill the vaginal deficiency (c) and
it is sutured at the borders of the vagina (d). (e) shows the
e results two months later
 282 Atlas of Endoanal and Endorectal Ultrasonography

a b
Fig. X.3. 53 Y patient who underwent abdominoperineal resection for locally recurrent anal cancer after radiochemotherapy and
reconstruction with pudendal flap of the thigh. The results of the healing process were excellent, as seen at the donor site (a)
and for the almost complete disappearance of hairs due to a progressive mucosal metaplasia of the flap (b)

the anal-genital condylomata in 3 out of 14 through the combination of radiotherapy and

patients and their reduction in 10 out of 14 chemotherapy, while resection of the anal sphinc-
patients. No side effects were reported [52]. These ter complex should be reserved for recurrent
studies could represent a hope of reaching even tumors or in the case of unsuccessful local control.
better and more lasting results both in high- New multicenter studies are necessary for the
grade dysplasia and in HPV-correlated anal numerous open questions, as the utility of neoad-
intraepithelial neoplasia, as well as consolidating juvant or adjuvant therapy, the possibility to
good results of treatments for anal canal epider- include splits in the treatment schedule to reduce
moid carcinoma. toxicity, the need of a more accurate staging of
Finally, the best course of treatment for epider- inguinal lymph nodes, and the optimistic expecta-
moid carcinoma of the anal canal is achieved tions of using a specific vaccine in the near future.

References immunoglobulin G antibodies against human papillo-

mavirus type 16 capsids with anal epidermoid carci-
1. Melbye M, Rabkin C, Frisch M, Biggar RJ. Changing noma. J Nat! Cancer Inst 1995;87:437-40
patterns of anal cancer incidence in the United States, 5. Frisch M, Olsen JH, Melbye M. Malignancies that occur
1940-1989. Am J EpidemioI1994;139:772-80 before and after anal cancer: clues to their etiology.
2. Schoenfield JH, Jones HT, Cuzick J, Northover JMA. Am J EpidemioI1994;140: 9-12
Anal cancer and marital status. Br J Cancer 6. Nigro ND, Vaitnevicius VK, Considine B. Combined
1990;162:286-8 therapy for cancer of the anal cancer: a preliminary
3. Noffsinger AE, Hui YZ, Suzuk L, Yochman LK, Miller report. Dis Colon Rectum 1974;17:354-6
MA, Hurtubire P, Gal AA, Fenoglio-Preiser CM. The 7. Nigro ND, Seydel HE, Considine F, Vaitkevicius VK,
relationship of human papillomavirus to proliferation Leichman L, Kinzie J]. Combined preoperative radia-
and ploidy in carcinoma of the anus. Cancer tion and chemotherapy for squamous cell carcinoma
1995;75:958-67 of the anal canal. Cancer 1983;51:1826-9
4. Heino P, Eklund C, Fredriksson-Shanazarian V, 8. Flam MS, John M, Lovalvo LJ, Mills RJ, Ramalho LD,
Goldman S, Schiller JT, Dillner J. Association of serum Prather C, Mowry PA, Morgan DR, Lau BP. Definitive
 Section X' Treatment options for anal carcinoma 283

non surgical therapy of epithelial malignancies of radiotherapy and Gastrointestinal Cooperative Group
the anal canal: a report of 12 cases. Cancer 1983;51: Proc ASCO 14, 1995
1378-87 21. Bartelink H, Roelofsen F, Eschwege F, Rougier P, Bosset
9. Meeker WR Jr, Sickle-Santanello BJ, Philpott G, Kenady JF, Gonzalez DG, Peiffert D, van Glabbeke M, Pierart M.
D, Bland KI, Hill GH, Popp MB. Combined chemother- Concomitant radiotherapy and chemotherapy is supe-
apy, radiation and surgery for epithelial cancer of the rior to radiotherapy alone in the treatment of locally
anal canal. Cancer 1986;57:525-9 advanced anal cancer: results of a phase III randomized
10. Byfield JE. Theoretical basis and clinical applications trial of the European Organization for Research and
of 5FU as a radiosensiter - an overview. In: Rotman M Treatment of Cancer Radiotherapy and Gastrointestinal
and Rosenthal CJ. Concomitant continuous infusion Cooperative Groups. J Clin OncoI1997;15:2040-9
chemotherapy and radiation, Springer Verlag, Berlin 22. UKCCCR Anal Cancer Trial Working Party.
1991;115-27 Epidermoid anal cancer: results from the UKCCCR
11. Staib L, Gottwald T, Lehnert T, Ruf G, Sturm J, Becker randomised trial of radiotherapy alone versus radio-
HD, Farthmonn E, Herparth C, Post S, Trede M, Berger therapy, 5-fluorouracil, and mitomycin C. Lancet
HG. Sphincter-saving treatment in epidermoid anal can- 1996;348:1049-54
cer: cooperative analysis of 142 patients in five German 23. Flam MS, John M, Pajak TF, et al. Radiation (RT) and 5-
university surgical centers. Int J Colorect Dis Fluorouracil (5FU) vs. Radiation, 5FU, Mitomycin-C
2000;15:282-90 (MMC) in the treatment of anal carcinoma: results of
12. Martenson JA jr, Gunderson LL. External ratiation a Phase III randomized RTOG/ECOG intergroup trial.
therapy without chemotherapy in the management of Anticancer Research 1995;15:1740
anal cancer. Cancer 1993;71:1736-40 24. Flam MS, John M, Pajak TF, Petrelli N, Myerson R,
13. Pignon T, Horiot JC, Bolla M, van Poppel H, Bartelink Doggett S, Quivey J, Rotman M, Kerman H, Coia L,
H, Roelofsen F, Pene F, Gerard A, Einhorn N, Nguyen Murray K. Role of Mitomycin in combination with flu-
TD, Vanglabbeke M, Scalliet P. Age is not a limiting fac- orouracil and radiotherapy, and of salvage chemoradi-
tor for radical radiotherapy in pelvic malignancies. ation in the definitive non surgical treatment of epi-
Radiother OncoI1997;42:107-20 dermoidal carcinoma of the anal canal: results of
14. Wagner JP, Mahe MA, Romestaing P, Rocher FP, Berger phase III randomized intergroup study. J Clin Oncol
C, Trillet -Lencir V, Gerard JP . Radiation therapy in the 1996;14:2527-39
conservative treatment of carcinoma of the anal canal. 25. Cummings BJ, Keane TJ, O'Sullivan B, Wong CS, Catton
Int J Radiat Oncology BioI Phys 1994;29:17-23 CN. Mitomicyn in anal canal carcinoma. Oncology
15. Brunet R, Becouarn V, Pigneux J, et al. Cisplatine (P) et 1993;50:S63-9
Fluorouracile (FU) en chemiotherapie neoadjuvante 26. John MJ, Flam M, Lovalvo L, Mowry PA. Feasibility of
des carcinomes epidermoides du canal anal. Lyon Chir non surgical definitive management of anal canal car-
1991;87:77-78 cinoma. Int J Radiat Oncol BioI Phys 1987;13:299-303
16. Belliere A, Chapet 0, Coquard R, Romestaing P, Ardiet 27. Valentini V, Mantello G, Luzi S, Genovesi D, Smaniotto
JM, Gerard JP. Brachytherapy in carcinomas of anal D, Mantini G, Coco C, Ratto C, Sofo L, Cellini N. The
canal and rectum: techniques and results. Cancer anal canal cancer: 2 cycles of radiochemotherapy and
Radiother 2003;7:24-32 boost brachytherapy. Radiol Med 1997;93:451-6
17. Nigro ND. Mutidisciplinary management of cancer of 28. Cummings BJ. Anal canal cancer: To split or not to
the anus. World J Surg 1987;11:446-51 split. Cancer J Sci Am 1996;2:194-6
18. Cummings BJ. Concomitant radiotherapy and 29. Allal AS, Mermillod B, Roth AD, Marti MC, Kurtz JM.
chemotherapy for anal cancer. Semin Oncol The impact of treatments factors on local control in
1992;19:102-8 T2-T3 anal carcinomas treated by radiotherapy with or
19. Pfeiffert D, Bey P, Pernot M, Guillemin F, Luporsi E, without chemotherapy. Cancer 1997;79:2329-35
Hoffstetter S, Aletti P, Boissel P, Bigard MA, Dartois D, 30. Iwamoto T, Nakahara S, Mibu R, Hotokezaka M,
Baylac F. Conservative treatment by irradiation of epi- Nakano H, Takana M. Effect of radiotherapy on
dermoid cancers of the anal canal: prognostic factors anorectal function in patients with cervical cancer. Dis
of tumoral control and complications. Int J Radiat Colon Rectum 1997;40:693-7
Oncol BioI Phys 1997;37:313-24 31. Doci R, Zucali R, La Monica G, Meroni E, Kenda R,
20. Roelofsen F, Bosset JF, Eschwege F. Concomitant radia- Eboli M, Lozza L. Primary chemoradiation therapy
tion and chemotherapy superior to radiotherapy alone with fluorouracil and cisplatin for cancer of the anus:
in the treatment of locally advanced anal cancer. results in 35 consecutive patients. J Clin Oncol
Results of phase III randomized trial of the EORTC 1996;14:3121-5
284 Atlas of Endoanal and Endorectal Ultrasonography

32. Gerard JP, Ayzac L, Hun D, Romestaing P, Coquard R, 41. Peley G, Farkas E, Sinkovics I, Kovacs T, Keresztes S,
Ardiet JM, Mornex F. Treatment of anal canal carcino- Orosz Z, Koves 1. Inguinal sentinel lymph node biopsy
ma with high dose radiation therapy and concomitant for staging anal cancer. Scand J Surg 2002;91:336-8
fluorouracil-cisplatinum. Long-term results in 95 42. Smith AJ, Whelan P, Cummings BJ, Stern HS.
patients. Radiother OncoI1998;46:249-56 Management of persistent or locally recurrent epider-
33. Hung A, Crane C, Delclos M, Ballo M, Ajani J, Lin E, moid cancer of the anal canal with abdominoperineal
Feig B, Skibber J, Janjan N. Cisplatinum-based com- resection. Acta Oncol 2001;40:34-6
bined modality therapy for anal carcinoma: a wider 43. Po card M, Tiret E, Nugent K, Dehni N, Parc R. Results
therapeutic index. Cancer 2003;97:1195-202 of abdominoperineal resection for anal cancer after Rt
34. Schneider HF, Grabenbauer GG, Reck T, Kockerling F, (54 Gy - no Ct). Dis Colon Rectum 1998;41:1488-93
Sauer R, Gall FP. Combined radiation and chemother- 44. Allal AS, Laurencet FM, Reymond MA, Kurtz JM, Marti
apy for epidermoid carcinoma of the anal canal. Int J MC. Effectiveness of surgical salvage therapy for
Colorect Dis 1992;7:192-6 patients with locally uncontrolled anal carcinoma after
35. John M, Pajak T, Flam M, Hoffman J, Markoe A, Wolkov sphincter-conserving treatment. Cancer 1999; 86:405-9
H, Paris K. Dose escalation in chemoradiation for anal 45. Nilsson PJ, Svensson C, Goldman S, Glimelius B.
cancer: preliminary results of RTOG 9208. Cancer J Sci Salvage abdominoperineal resection in anal epider-
Am 1996;2:205-11 moid cancer. Br J Surg 2002;89:1425-9
36. Bosset JF, Roelofsen F, Morgan DA, Budach V, Coucke P, 46. van der Wal BC, Cleffken BI, Gulec B, Kaufman HS,
Jager JJ, Van der Steen-Banasik E, Triviere N, Stub en G, Choti MA. Results of salvage abdominoperineal resec-
Puyraveau M, Mercier M. European Organization for tion for recurrent anal carcinoma following combined
Research and Treatment of Cancer. Radiotherapy and chemoradiation therapy. J Gastrointest Surg 2001;
Gastrointestinal Cooperative Groups. Shortened irradi- 5:383-7
ation scheme, continuous infusion of 5-fluorouracil and 47. Joseph VT, Wee JTK. New technique of vaginal recon-
fractionation of mitomycin C in locally advanced anal struction using neurovascular pudendal-thigh flaps.
carcinomas. Results of a phase II study of the European In Strauch B, Vasconez 0, and Hall-Findlay EJ. Grabb's
Organization for Research and Treatment of Cancer. Encyclopedia of Flaps, 2nd Ed. Philadelphia:
Radiotherapy and Gastrointestinal Cooperative Lippincott-Raven 1998,1466-70
Groups. Eur J Cancer 2003;39:45-51 48. Pisegna Cerone L, Papadia A, Infantino A. Ricostruzio-
37. Svensson C, Goldman S, Friberg B, Glimelius B. ne "tailor-made" del perineo con lembi pudendi inter-
Induction chemotherapy and radiotherapy in loco- ni dopo resezione neoplastica. Atti del Congresso
regionally advanced epidermoid carcinoma of the anal SICP-UCP Bologna 2002
canal. Int J Radiat Oncol Bioi Phys 1998;41:863-7 49. Minkoff H, Ahdieh L, Massad LS,Anastos K, Watts DH,
38. Lee DH, Kim I, Song HH, Jung JY, Kim DY, Lee KW, Melnick S, Muderspach L, Burk R, Palefsky J. The effect
Kim TY, Heo DS, Bang Y], Ha SW, Park ]G, Kim NK. of highly active antiretroviral therapy on cervical
Induction chemotherapy followed by radiotherapy in cytologic changes associated with oncogenic HPV
the treatment of anal cancer. Oncol Rep 2003;10:101-4 among HIV-infected women. AIDS 2001;15:2157-64
39. Peiffert D, Giovannini M, Ducreux M, Michel P, 50. Fox P, Stebbing J, Portsmouth S, Winston A, Frances N,
Francois E, Lemanski C, Mirabel X, Cvitkovic F, Nelson M, Gazzard B, Bower M. Lack of response of
Luporsi E, Conroy T, Gerard JP. Digestive Tumors anal intra -epithelial neoplasia to highly active
Group of the French 'Federation Nationale des Centres antiretroviral therapy. AIDS 2003;17:279-80
de Lutte Contre Ie Cancer' High-dose radiation thera- 51. Palefsky JM, Holly EA, Ralston ML, Da Costa M,
py and neoadjuvant plus concomitant chemotherapy Bonner H, Jay N, Berry JM, Darragh TM. Effect of
with 5-fluorouracil and cisplatin in patients with local- highly active antiretroviral therapy on the natural his-
ly advanced squamous-cell anal canal cancer: final tory of anal squamous intraepitheliallesions and anal
results of a phase II study. Ann Oncol 2001;12:397-404 human papillomavirus infection. J Acquir Immune
40. Gerard JP, Chapet 0, Samiei F, Morignat E, Isaac S, Defic Syndr 2001;28:422-8
Paulin C, Romestaing P, Favrel V, Mornex F, Bobin JY. 52. Goldstone SE, Palefsky JM, Winnett MT, Neefe JR.
Management of inguinal lymph node metastases in Activity of HpsE7, a novel immunotherapy, in patients
patients with carcinoma of the anal canal: experience with anogenital warts. Dis Colon Rectum 2002;45:502-7
in a series of 270 patients treated in Lyon and review of
the literature. Cancer 2001;92:77-84
 Invited commentary
C.H. Crane

The treatment of anal cancer is a triumph of mul- University of Texas M. D. Anderson Cancer
tidisciplinary cancer care, and represents a model Center) without a reduction in efficacy. In the
organ preserving strategy that could be duplicated future, the use of therapies that specifically target
in other gastrointestinal tumor sites in the future. the tumor microenvironment (antiangiogenic
The current reference chemoradiotherapy regi- agents) and survival signals (epidermal growth
men (5-florouracil, mitomycin-C and radiothera- factor receptor, EGFR) are examples of ways to
py) was originally developed by a surgeon [1], and more specifically target the tumor and spare nor-
subsequently oncologists on both sides of the mal tissue effects. These strategies are already
Atlantic have tried to improve the efficacy as well being used with chemotherapy and radiotherapy
as tolerability of therapy in various ways. As Drs. and will hopefully enhance the therapeutic benefit
Pisegna Cerone and Infantino have pointed out in of chemoradiotherapy in anal cancer.
the current review, investigators have used radio- The uncommon histological subtypes of anal
therapy alone, split course radiotherapy tech- cancer generally have a different biologic behav-
niques and single agent 5FU with radiotherapy in ior, sensitivity to chemoradiation, and a worse
efforts to make treatment better tolerated. At least overall prognosis. For this reason, the manage-
for healthy patients with advanced disease (T3-T4 ment strategy is slightly different from that of epi-
or node positive), these approaches are not opti- dermoid carcinoma. In contrast to epidermoid
mal from an efficacy perspective. Combination lesions, adenocarcinomas of anal glandular origin
therapy with 5FU and cisplatin appears to com- have a much lower local control rate (approx. 50%
pare favorably to the standard sFU and mito- compared to 80% for epidermoid lesions) and a
mycin-C [2], but results from randomized trials much higher distant failure rate with definitive
comparing the two will not be available for sever- chemoradiation [4]. For this reason, abdomino-
al years (Radiation Therapy Oncology Group, perineal resection with pre- or postoperative
RTOG 98-11, 500/650 patients accrued). Additional chemoradiation followed by adjuvant chemother-
studies are focusing on the use of radio-protectors apy (similar to combined modality therapy for
(proposed study, RTOG), and the use of conformal low rectal cancer) is probably the best initial strat-
radiotherapy techniques [3] to overcome the toxi- egy for most of these patients. Primary melanoma
city limitations of therapy. Conformal techniques of the anal canal is effectively managed with lim-
including intensity modulated radiation therapy ited surgery (local excision) and postoperative
(IMRT) can spare the skin, small bowel, femoral hypo-fractionated radiotherapy [5]. Again, the
heads, and genitalia in ways that were not possible distant failure rate is high, but the local control
a decade ago. Additionally, better tolerated with that strategy is acceptable.
chemotherapy agents such oral capecitabine sub- In summary, we have come along way in the
stituted for i.v. 5FU and oxaliplatin substituted for treatment of anal carcinoma without surgery.
cisplatin, will probably lead to reduced gastroin- Organ preserving strategies would also be desir-
testinal and hematologic toxicity (current study, able in selected patients who have other diseases
286 Atlas of Endoanal and Endorectal Ultrasonography

where the surgical morbidity is high (esophageal and select the patients correctly, we may be able to
cancer, gastric cancer, low rectal cancer). If we can duplicate the success story of anal cancer in other
develop more effective chemoradiation regimens diseases.

References conformal therapy in the treatment of anal canal car-

cinoma with combined chemotherapy and radiothera-
1. Nigro ND, Seydel HE, Considine F, et al. Combined py: results of a phase II study. Int J Radiat Oncol BioI
preoperative radiation and chemotherapy for squa- Phys 2003;56:823-31
mous cell carcinoma of the anal canal. Cancer 1983; 4. Papagikos M, Crane CH, Skibber J, et al. Chemo-
51:1826-9 radiation for adenocarcinoma of the anus. Int J Radiat
2. Hung A, Crane C, Delclos M, et al. Cisplatin-based Oneol BioI Phys 2003;55:669-78
combined modality therapy for anal carcinoma: a 5. Ballo MT, Gershenwald JE, Zagars GK, et al. Sphincter-
wider therapeutic index. Cancer 2003;97:1195-202 sparing local excision and adjuvant radiation for anal-
3. Vuong T, Devic S, Belliveau P, et al. Contribution of rectal melanoma. J Clin Oncol 2002;20:4555-8
 Subject index

3D reconstruction 19,20,235 - colo anal anastomosis 178,179,184-188,191,193-195,197,198,

5-fiuorouracil243-254, 259, 260, 262, 265-280, 283, 284 205,207
5FU see s-fluorouracil - stapled anastomosis 126,197,232
anococcygealligament 27
A anorectal ring 117
abdominoperineal resection 179,182,187,210-212,242,247,277, anterior resection 44,46,125,164,169,172,176,177-180,181,182,
278,282,284 186,188,189,191,192,195, 197, 199, 207, 211, 212, 231
abscess 44-46, 68 antiretroviral therapy 280, 284
absorption 4, 6 HAART 280
- absorption coefficient 4 aorta 179,193, 200, 204
- acoustic absorption 4 applicator 257, 258, 266, 271
accelerator 256, 257, 270, 273 artifacts 20,23,46,59, 61-63, 67, 125, 133
adenoma 46,52,53,61,62,64-66,69,71,80,159,160,167,171 autonomic nerves 180-182,202,215
adjuvant therapy 164, 241, 242, 245, 250, 253-255, 282
AIDS 142, 284 B
amplification 6,235,246 balloon 12-16, 22, 23, 37, 38, 46, 55, 57, 61, 63, 67, 68, 96, 104, 146,
amplitude mode 6 166
A-mode 5,6 biopsy
anal cancer - transperineal needle biopsy 135
- anal canal adenocarcinoma 143, 144 - transrectal needle biopsy 16
- anal intra -epithelial neoplasia 284 bladder 37, 40, 55, 91, 92, 133, 145, 146, 215, 216, 250
- anal melanoma 143 bony sacrum 14
- anal squamous cell carcinoma 141-143 brachytherapy 44, 260, 263, 264, 278, 283
- basal cell carcinoma 143 brightness mode 6
- carcinoid tumor 142 B-mode 6,121
- epidermoid anal cancer 278,283
anal margin 141,143, 160, 178, 278 C
anal mucosal electro sensitivity 166 capecitabine 245,248,249,252, 285
anal sphincter cervix 37, 41, 91, 277
- external anal sphincter 27,28,31,44,47,146,148,184 chemotherapy
- internal anal sphincter 27,28,30,31,36,43,47,146,150,151 - adjuvant chemotherapy 250, 254, 260, 261, 263, 280, 285
anal transitional zone 141 - concomitant chemotherapy 113, 253-255, 284
ATZ 141 cisplatin 247,253, 254, 265, 279, 283-286
anal-pouch angle 185,186 coccyx 27, 28, 149, 159, 242
anastomosis colonofiberscope 11, 12
- anastomotic stenoses 173 coloplasty 190-192, 195-198
288 Atlas of Endoanal and Endorectal Ultrasonography

colostomy 158, 163, 179, 193, 195, 196, 277, 279, 280 G
condylomata 277, 282 gain 6,7,15,16,58,105,178,213,217,227
console 228-230 gamma probe 280
cough pressure 166 gastrocolic omentum 201
crystal 5, 6, 12, 21 gonadal vessels 200, 205, 231
iray scale 6
D
Da Vinci robotic device 229 H
density 4, 255 hepatic artery 219, 225
dentate line 27,65, 66, 141, 143, 158, 166, 183 hepatic metastases 217,225,226
diaphragm 37, 220 hepatic veins 220, 221, 223, 226
diffraction 4, 5 Hirschsprung's disease 173
doppler 120,218-220 HIV infection 142
downsizing 241, 247, 248, 253 HPV infection 141,143
downstaging 97, 117, 241, 246-249, 251, 253, 255, 259 hypogastric nerves 180, 202, 232
duodenojejunal flexure 200
dysplasia 142, 282
iliococcygeus 27, 28
E immunoscintigraphy 130
echoendoscope 11 impedance 4-6, 8
elastography 120-122 impulse 5
endometriosis 173 incontinence 23, 43, 44, 47, 143, 159, 166, 170, 175, 176, 184, 185,
epidermal growth factor receptor 285 190,195
EGFR 285 inferior mesenteric artery 179, 180, 186, 189, 191, 193, 199, 200,
excision 204,205,209,212,213,215,231
- full-thickness excision 61,159,167,176 inferior mesenteric vein 193, 200, 203, 205, 231
- local excision 52,55,63,67-69,95-98,109,125,157-159,161- inferior vena cava 219, 221, 225
164,168,169, 171, 176, 177, 179, 285, 286 intensity 4, 57, 58, 78, 125, 130, 285
- partial thickness excision 167 interface 4-6, 8, 16, 28, 38, 46, 53, 61, 64,67,68,71-75,83,84,87-
- trans anal excision 46,159,161,169,172 91,100,101,114,115,126,132
intersphincteric space 28,45
F intraoperative ultrasound 217
falciform ligament 221,223,225, 227 IOUS 217
fascia 27, 180, 200, 201, 208, 231, 232 irinotecan 248, 272
- Denonvillier's fascia 202, 208 ischial spine 27
- fascia propria of the rectum 202 ischio-anal fossa 27
- Gerota's fascia 200,201 ischiococcygeus 27
- obturator fascia 27 ischiorectal fossa 28
- parietal pelvic fascia 180
- perirectal fascia 180
- Waldeyer's fascia 202 J-pouch 185,190-192,195-198
fibrosis 46, 68,113,117,125, 130, 131, 254
fistula 44, 45, 143, 159, 166 L
flap 173,175,280-282 lamina propria 38, 46
focal length 11,12,16,55, 61, 146 laparoscopic surgery 159, 199, 200, 206, 210, 212, 216, 233, 234,
focal range 6, 11 237
folinic acid 243, 244 laparoscopic ultrasound 200,217,222, 225-227
follow-up 44, 46,52, 60, 123, 125-127, 129-131, 133, 135-138, 146, 151- lateral ligaments 180, 183
153,161,163,164,169,170,175,176,194,209,211, 241-245, 254, leakage 22,184,185,190,258
268,280 left colic artery 179, 180, 209
frequency 3-6, 11, 12, 16, 21, 23, 46, 47, 165, 170, 171, 177, 184, 185, left colonic flexure 199
219,226,227 leucovorin 247-249, 251, 252, 260, 265, 272
 Subject index 289

levamisole 243-245,251 pelvic peritoneum 28,37, 202

levator ani 23,27,28,31,45,57,91,159 pelvis 27,31,45,178,185,189,194,195,199,201,202,207,209,215,
lienocolic ligament 201, 206 232,234,235,241,264,265,268
line of Toldt 200, 203 perianal skin 27,141,143
longitudinal muscle 23,28,31-33,45 perianal space 28
Lymphazurin blue dye 184 perineal body 27, 28, 31, 32, 36, 45
lymph nodes perineum 45,199
- inferior mesenteric lymph nodes 183 perirectal fat 37, 38, 40, 46, 51-53, 57-59, 71-76, 78-80, 83-90, 92,
- inguinal lymph nodes 145,280,282 96,97,100-102,104,114,115,128,131,132,169,174,180,248
- internal iliac lymph nodes 183 perturbation 3
- lateral pelvic lymph nodes 57,91 piezoelectric transducer 6
- mesorectallymph nodes 52,98, 102, 109 plastic cone 13,16,18,28,146
- obturator lymph nodes 27, 28, 57, 91, 183 pneumoperitoneum 166, 199, 203, 232
- perirectal lymph nodes 51,71-73, 107 polypectomy 60, 67, 69, 126, 127, 129
- sentinel lymph node 184,187,216,280,284 portal branches 220,223
portal vein 219,225
M positron emission tomography 130, 214
manometry 42, 43, 174 pouch of Douglas 159
maximum tolerated volume 195 presacral space 159
mesocolon 195, 200, 201 probe
mesorectum 20,98,101,102,109,110,130,158,180,181,187,189, - curved array probes 11
190,191,194,202,203,232,242 - miniprobes 121
methyl-lomustine 244 - radial probes 11
micrometastases 96, 98, 102, 109, 215, 246 proctography 185, 186
mitomycin C 249, 253, 254, 260, 277, 283, 284 proctoscope 15,16,96
MMC 253,254,277-279, 283 prolapse 23,45,131,141, 173
mucosectomy 166, 167 prostate gland 11, 55
muscolaris mucosae 38, 46, 57, 61 psoas muscle 201
muscolaris propria 20, 27, 38, 40, 46, 51, 52, 54, 57, 67, 68,71-76, pubococcygeus 27
78-80,83,84,87-90,101,114,115,128,132,168 puborectalis muscle 27,30,31, 149
pudendal artery 280
N pudendal nerve terminal motor latency 166
neoadjuvant therapy 96, 97, 119, 120, 121, 252 pudendal nerves 45
neorectum 131,184,185,188,190,194,197 pullback mover 19

o R
obstructed defecation 44,47,185 radio chemotherapy
obturator internus muscle 28 - post-radiochemotherapy 114
oscilloscope 5, 6 - preoperative radio chemotherapy 113, 117-119, 210
overstaging 23, 53-55, 60, 63, 67, 74, 85, 96, 102, 113, 127, 248 radiotherapy
oxaliplatin 248,251,252, 255, 272, 285 - external radiotherapy 254, 278
- intraoperative radiotherapy 256, 261, 262, 273
P - postoperative radiotherapy 119, 161, 162, 242, 250
Paget disease 142 - preoperative radiotherapy 52, 77, 92, 113, 114, 117-119, 121, 122,
pancreas 193,200,213,231,258 13h138,214,243,246,251,253-255
papillomavirus 277, 282, 284 real time 6
parasympathetic sacral nerve 202 rectal artery 183
pelvic autonomic nerve preservation 181 - middle rectal artery 183
pelvic dissection 178,202,207, 235 rectal capacity 184
pelvic exenteration 280 rectal compliance 184,185,188
pelvic floor 17, 28,31,36,42,45,195 rectal sensitivity 44, 185
pelvic lymphadenectomy 183, 190, 194 rectoprostatic septum 113
290 Atlas of Endoanal and Endorectal Ultrasonography

rectoscope 16,21,165-168,174 T
recurrence telemedicine 157
- anastomotic recurrence 131 telementoring 233
- extraluminal recurrence 131 TEM see transanal endoscopic microsurgery
- local recurrence 46,92,109,113,119,125,129,130,131,133,136- thermal energy 4
138,152,157,161-164,168,169,176,177,179,180-183, 186, 187, TME see total mesorectal excision
189, 190, 191, 193, 194, 197, 199, 207,212, 213, 241-246, 248, total mesorectal excision (TME) 109,119,121,137,138,157,169,
249,251,255,258-261,266,280 180-183,187,189-191,193,194,197,199,202,207, 209, 212-216,
- pelvic recurrence 130,137,183,187,191,242 235,241,243,251, 255
reflection 3-6, 8, 37, 102, 166, 180, 183, 191, 200, 202, 241 transanal endoscopic microsurgery (TEM) 52,157,163,164-177
refraction 4,5, 67 transducer 5, 6, 8, 11-16, 18, 20-23, 55, 100, 102, 146, 217-220, 227
reservoir 162,185,187,188,190,191,194,195,197,198 transsacral resection 159
resolution 6,7,11,19,45-47,57,61,75,100,146,148,225 trans sphincteric procedure 159
- axial resolution 6 transverse perineii 28, 32
- lateral resolution 6 - deep transverse perineii 28
resonance coefficient 4 - superficial transverse perineii 28
resting pressure 43,44,166,184
retrorectal space 180 U
reverberation 6,8,55, 62, 67 ultrasonic energy 4, 5
robotics 157 ultrasonic intensity 4
- robotic arms 228-232, 235 ultrasonically activated scalpel 165, 167, 174
- robotic systems 228,234, 235 understaging 55, 67, 75, 84, 96, 102, 113, 184
- robotic technology 235 ureter 200,201, 205, 207, 231
urethra 68, 145
S urgency 184, 190, 194, 195
sacral promontory 202, 207 urogenital triangle 28
sacrum 14,37,159,180 uterus 37, 41, 91, 92, 131, 133, 261
scattering 4
seminal vesicles 37, 40, 55, 57, 64-66, 68, 79, 91, 132, 133, 171, 202, v
208,232 vaccine 280,282
soiling 184, 185 vagina 28,31,37, 41, 45, 55, 91, 145, 146, 280, 281
sphincter-saving operations 179 velocity of propagation 4
spleen 200 vibrations 3, 5
splenic flexure 189,193,201,206,213,230-232,234 viner is tina 244
splenic vein 200
squamocolumnar junction 27 w
squeeze pressure 31, 44, 184 wave 3,4,5
staging system 47,51,55,96,97,143,146,151-153,253 - elastic waves 3
staple line 205,208,210 - longitudinal waves 3
stapler 44, 161, 197, 203, 205, 208, 210, 230- 232 wavelength 4, 5
stomach 201,206,258 wound protector 203, 209
submucosa 38, 46, 51, 52-54, 57, 61, 62, 67, 68, 71-73, 78-80, 83, 87-
90,97,114,115, 127,128,146,158,160
supraelevator space 27,28