manual excision total mesorecto heald

SAVE OFFLINE

Manual of

Total Mesorectal Excision

HEBK001-FM.indd 1

11/02/13 6:40 PM

HEBK001-FM.indd 2

11/02/13 6:40 PM

Manual of

Total Mesorectal Excision Edited by

Brendan Moran, MCh, FRCSI, FRCS, Consultant Surgeon Basingstoke and North Hampshire NHS Trust; Honorary Senior Lecturer, Southampton University, UK Richard John Heald, CBE, MCHIR, FRCS, Clinical Director, Pelican Cancer Foundation, Basingstoke, UK; Honorary Professor of Surgery, Southampton University, UK; President of the Colostomy Association

HEBK001-FM.indd 3

08/02/13 5:27 PM

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2013 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed on acid-free paper Version Date: 20130129 International Standard Book Number-13: 978-1-4441-1716-5 (Hardback) This book contains information obtained from authentic and highly regarded sources. While all reasonable efforts have been made to publish reliable data and information, neither the author[s] nor the publisher can accept any legal responsibility or liability for any errors or omissions that may be made. The publishers wish to make clear that any views or opinions expressed in this book by individual editors, authors or contributors are personal to them and do not necessarily reflect the views/opinions of the publishers. The information or guid ance contained in this book is intended for use by medical, scientific or health-care professionals and is provided strictly as a supplement to the medical or other professional’s own judgement, their knowledge of the patient’s medical history, relevant manufacturer’s instructions and the appropriate best practice guidelines. Because of the rapid advances in medical science, any information or advice on dosages, procedures or diagnoses should be independently verified. The reader is strongly urged to consult the drug companies’ printed instructions, and their websites, before administering any of the drugs recommended in this book. This book does not indicate whether a particular treatment is appropriate or suitable for a particular individual. Ultimately it is the sole responsibility of the medical professional to make his or her own professional judgements, so as to advise and treat patients appropriately. The authors and publishers have also attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright. com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-forprofit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

HEBK001-FM.indd 4

12/02/13 8:46 PM

Contents Contributors   1   2   3   4   5   6   7   8   9 10 11 12 13 14 15

16

The evolution of a concept: the total mesorectal excision story R.J. Heald Anatomy of the rectum, anal canal and pelvic floor Thilo Wedel Clinical ultrasound Oliver Shihab, Arcot K. Venkatasubramaniam Magnetic resonance imaging staging of rectal cancer Peter How, Gina Brown Radiological staging for systemic disease Gina Brown, Chris Hunter Preoperative radiotherapy and chemoradiotherapy for rectal cancer Rob Glynne-Jones, Mark Harrison Total mesorectal excision for rectal cancer Brendan Moran Abdominoperineal excision of the rectum Torbjörn Holm Laparoscopic surgery Katharine E. Bevan, Tom D. Cecil Robotic total mesorectal excision M. Chadwick, H.S. Tilney, A.M. Gudgeon Local excision and transanal endoscopic microsurgery Wolfgang B. Gaertner, David A. Rothenberger Pathology assessment Philip Quirke, Tim Palmer, Gordon G.A. Hutchins, Nick P. West Assessment and management of recurrence Peter J. Lee, Kirk K.S. Austin, Michael J. Solomon Lateral pelvic side-wall nodal involvement in rectal cancer Hideaki Yano, Brendan Moran Intestinal stoma and the role of defunctioning a low anastomosis after anterior resection David Mitchell, Kandiah Chandrakumaran, Steven Arnold Quality of life in patients undergoing abdominoperineal excision and anterior resection for rectal cancer Peter How, Kandiah Chandrakumaran

Index

HEBK001-FM.indd 5

vii 1 31 52 58 71 87 103 124 140 154 171 191 203 222

228

239

249

11/02/13 6:40 PM

HEBK001-FM.indd 6

11/02/13 6:40 PM

Contributors

Steven Arnold Consultant Surgeon, Department of Surgery, Basingstoke and North Hampshire Foundation Trust, Basingstoke, UK Kirk K.S. Austin Department of Colorectal Surgery and Surgical Outcome Research Centre (SOuRCe), Royal Prince Alfred Hospital, Sydney, Australia Katharine E. Bevan Basingstoke and North Hampshire Hospital Foundation Trust, Basingstoke, UK Gina Brown Consultant Radiologist and Honorary Senior Lecturer, Royal Marsden NHS Foundation Trust, Sutton, UK Tom D. Cecil Consultant Surgeon, Department of Surgery, Basingstoke and North Hampshire Hospital Foundation Trust, Basingstoke, UK Michael Chadwick Consultant Laparoscopic Colorectal Surgeon, St Helens & Knowsley Teaching Hospitals, NHS Trust, Merseyside, UK Kandiah Chandrakumaran Associate Specialist, Department of Colorectal/ Pseudomyxoma Surgery, Basingstoke and North Hampshire Hospital Foundation Trust, Basingstoke, UK

Mark Harrison Consultant Clinical Oncologist, Mount Vernon Cancer Centre, Northwood, UK R.J. (‘Bill’) Heald Director of Surgery, Pelican Cancer Foundation, Basingstoke, UK Torbjörn Holm Associate Professor of Surgery, Section of Coloproctology, Department of Surgical Gastroenterology, Karolinska University Hospital, Stockholm, Sweden Peter How Colorectal Research Registrar Pelican Centre, Basingstoke & Croydon University Hospital Chris Hunter Colorectal Surgeon, Royal Marsden NHS Foundation Trust and Croydon University Hospital, Surrey, UK Gordon G.A. Hutchins Department of Pathology and Tumour Biology, Leeds Institute of Molecular Medicine, University of Leeds; and Leeds Teaching Hospitals Trust, Leeds, UK Peter J. Lee Department of Colorectal Surgery and Surgical Outcome Research Centre (SOuRCe), Royal Prince Alfred Hospital, Sydney, Australia

Wolfgang B. Gaertner Division of Colon and Rectal Surgery, Department of Surgery, University of Minnesota, Minneapolis, MN, USA

David Mitchell Colorectal Fellow, Department of Surgery, Basingstoke and North Hampshire Foundation Trust, Basingstoke, UK

Rob Glynne-Jones Consultant Clinical Oncologist and Macmillan Lead Clinician in Gastro-Intestinal Cancer, Mount Vernon Cancer Centre, Northwood, UK

Brendan Moran Consultant Surgeon, Department of Surgery, Basingstoke and North Hampshire Foundation Trust, Basingstoke, UK

Mark Gudgeon Department of Colorectal Surgery, Frimley Park Hospital, NHS Foundation Trust, Frimley, UK

Tim Palmer Department of Pathology and Tumour Biology, Leeds Institute of Molecular Medicine, University of Leeds; and Leeds Teaching Hospitals Trust, Leeds, UK

HEBK001-FM.indd 7

11/02/13 6:40 PM

Philip Quirke Department of Pathology and Tumour Biology, Leeds Institute of Molecular Medicine, University of Leeds, Leeds, UK

Arcot K. Venkatasubramaniam Consultant Colorectal Surgery, Basingstoke and North Hampshire Foundation Trust, Basingstoke, UK

David A. Rothenberger Division of Colon and Rectal Surgery, Department of Surgery, University of Minnesota, Minneapolis, MN, USA

Thilo Wedel Head of the Center of Clinical Anatomy, Institute of Anatomy, Christian Albrechts University of Kiel, Kiel, Germany

Oliver Shihab Basingstoke and North Hampshire Trust, Basingstoke, UK

Nick P. West Department of Pathology and Tumour Biology, Leeds Institute of Molecular Medicine, University of Leeds; and Leeds Teaching Hospitals Trust, Leeds, UK

Michael J. Solomon Department of Colorectal Surgery and Surgical Outcome Research Centre (SOuRCe), Royal Prince Alfred Hospital and Discipline of Surgery, University of Sydney, Sydney, Australia H.S. Tilney Consultant Colorectal Surgeon, Frimley Park Hospital, NHS Foundation Trust, Surrey, UK

HEBK001-FM.indd 8

Hideaki Yano Consultant Surgeon, Department of Surgery, National Centre for Global Health and Medicine, Tokyo, Japan

11/02/13 6:40 PM

1 The evolution of a concept: the total mesorectal excision story (for surgeon or patient alike) R.J. (‘Bill’) Heald

Introduction One German professor, a great friend, suggested that I should write the story of how it became possible to conquer the whole world with an idea–a concept– a guiding principal of surgical technique–to tell in fact how the words Total Mesorectal Excision (TME) moved from humble beginnings in Basingstoke to become a part of any discussion about rectal cancer anywhere in the world (Figure 1.1). A personal reflection on the key components of the TME story over a period of 30 years may elucidate some of the issues and encourage others to follow an idea born out of clinical observation at the surgical coalface. In the late 1970s, when I was a young consultant, almost half of all patients with rectal cancer died from local recurrence – that is, regrowth of the cancer within the pelvis. The pelvis became gradually permeated by malignant tumour, which infiltrated all pelvic contents, particularly pelvic side-wall nerves and the sacral nerve roots to the legs, eventually causing intractable pain, paralysis and incontinence – all adding up to one of the most miserable forms of death ever dreamed up, and all terribly slow. Unsurprisingly, almost 100 years before this, H.W. Maunsell stated in the Lancet in 1892:

HEBK001-C01_p01-30.indd 1

All carcinomas of the lower sigmoid and upper rectum are tabooed by all practical surgeons on account of their anatomical inaccessibility. All are abandoned without hope to linger on for a few months until death relieves them of their loathsome condition.

On this background, all who tackled rectal cancer surgery had to contend with many failures and terrible outcomes in the patients who developed local recurrence. The TME story starts with its roots in embryology and its basis in anatomy.

Embryology Embryology was always a challenge for me as a student to grasp the mysteries of one cell turning itself into a fetus with a near certainty of becoming a human being. One special mystery was that ‘tube’ from mouth to anus that was somehow programmed to sprout the potential liver in one direction and the pancreas in another, and then, before the fetus is 1 cm long, to take a journey out of the primitive abdominal cavity and subsequently back into it, thus setting the components of the gastrointestinal tract in their appointed order

11/02/13 3:43 PM

2  The evolution of a concept: the TME story

Figure 1.1. Worldwide TME. Some of the TME venues, comprising more than 450 invited operations in more than 40 countries. At the time of going to press the Queen’s Birthday Honours List presented the CBE to Professor Richard Heald for services to UK surgery internationally.

and placement and establishing the marvel of the human alimentary system in situ. My personal story is of gradual realization that this midline gut tube can be redefined with its intrinsic ­lymphovascular ­surround as a midline envelope recognizable for surgeons by the spider’s web of areolar tissue around it. The distal part of the envelope is around the rectum and becomes the mesorectum, and the cobweb around it where the surgeon dissects is the ‘holy plane’. Each part of the envelope has an artery coming from the front of the aorta, rather than the side like everything else. The key hypothesis was that each part of the envelope, if very carefully removed en bloc, might have a very good chance of enveloping the whole primary field of spread of a cancer and thus curing all but the most advanced cases (Figure 1.2). The TME story is inextricably linked with embryology, and it is probable that similar concepts might be applicable to many other cancers, since the constraints to cancer growth, such as areolar planes that block angiogenesis, are relevant to malignant infiltration of many kinds and in many places ­(Figure 1.3).

HEBK001-C01_p01-30.indd 2

The growing surgeon: ‘on the shoulders of giants’ These seeds of wonder at embryology lay dormant in me as surgery became practice. Small flashes of wisdom from along the path of surgical training blinked as possible explanations for good or bad outcomes: ‘The best surgeon is the one with a capacity for taking infinite pains’ is one such pearl that continues to flash up with great regularity. Pains to do what exactly? Advocating TME eventually became a battle to justify turning a 1 h operation into an increasingly fastidious and precise operation taking 3 h or more. Initially the only obvious justification for such a change was the hope of fewer permanent colostomies. The potential for improved outcome, survival and cure would not become apparent for 5 years or more and was never dreamed of at the beginning. It is perhaps a measure of the fundamental ethics of the early British National Health Service (NHS) of the 1970s and 1980s that no financial or management pressure was ever exerted on me to desist from

11/02/13 3:43 PM

Basingstoke, Hampshire  3

Figure 1.2a. The mesorectum surrounded by the holy plane.

Figure 1.3. Pursuable cobweb areolar plane.

such an expensive experiment. Such benign acceptance of the time-consuming emerging concept did not, however, imply that I was not berated and condemned by the surgical establishment for the temerity to challenge their supremacy – as outlined later. One early public condemnation, from Prof Mike Perry, still rings in my ears today: ‘you are taking on a grave responsibility by deviating from the radicality of the established orthodoxy of abdominoperineal excision’.

Basingstoke, Hampshire

Figure 1.2b. A perfect TME (PO Nystrom).

HEBK001-C01_p01-30.indd 3

With an abiding interest in visceral surgery, the early 1970s finds me a consultant surgeon in the little-known town of Basingstoke in the pleasant countrified county of Hampshire, about an hour south of London, with a catchment population of approximately 250 000. I was the second full-time consultant general surgeon in the large new hospital under construction. My senior colleague was the remarkable Mr Frank Tovey, whose voluntary service in India and China, combined with his passion for research and teaching, were all truly serendipitous for me. I still remember his standing at my side in the operating room as I tediously tried to achieve an early TME. I was embarrassed at my ineptitude, but he said ‘I think you might have something there, Bill. It is very different from what everyone else is doing and is definitely worth persevering with. I will give up doing rectums and let you have them all.’ This was a key starting point, as becoming the only rectal cancer surgeon in a District Hospital was then a unique privilege. Specialization was commencing naturally by mutual consent.

11/02/13 3:43 PM

4  The evolution of a concept: the TME story

1975

Dividing the mesorectum (not called that then ....) No such word in GRAY’s “ANATOMY”

Figure 1.4. The node contains cancer!

Anterior resection, as I had been taught, involved manual extraction of the rectum with much rough tearing of the surrounding fat, sometimes ­uncontrollable blood loss, and desperate anxiety about leakage from the anastomosis and later of local recurrence. The latter was assumed to be largely in the hands of the almighty and pretty much inevitable. It was frequently stated by the great and the good that ‘surgery had gone as far as it could go’. More than 75 per cent of rectal cancers, and every cancer that could be felt digitally, was offered only abdominoperineal excision (APE) with permanent colostomy, which was considered much the most ‘radical’ operation and much the safest oncologically. Anterior resection was believed to add substantially to the risk of local recurrence. Cancer recurrence at the suture line dominated our ideas of follow-up after anterior resection, and this dictated a digital examination and sigmoidoscopy. Ultrasound, computed tomography (CT) and magnetic resonance imaging (MRI) did not exist.

The mesorectum in rectal cancer: the clue to local recurrence? Against this background, clinical experiences and histological findings with five particular patients suggested that mesorectal residues might be the source of those very common pelvic recurrences.1 This ‘tale of five patients’ has, I am told, become the most cited reference in the whole surgical literature on bowel cancer. The most memorable, and perhaps most significant, of the five thought provoking patients in this first paper was a publisher called Anthea. Like most

HEBK001-C01_p01-30.indd 4

of the TME story, the seminal moment occurred in the operating room. We were dealing with the all too common random bleeding in the fat behind the rectum when we noticed a small lymph node less than 1 cm in size. It looked harmless enough, but I lifted it out and asked the scrub nurse to send it for pathological analysis. A week later the node was reported as containing a focus of cancer (Figure 1.4). It thus became apparent that I had cut or torn through the field of spread of Anthea’s cancer. What to do? A week later we performed an APE to be properly ‘radical’, but no further cancer was found in the specimen. Five years later Anthea died of rectal cancer bony metastases with a quadriplegia – a memorable and terrible end to a historic individual in the story of rectal cancer. It still seems to me that cutting through cancer, or its field of spread, almost always leads to a cancer death, unless possibly on occasion when the tissues have been recently irradiated. The concept of total excision of the mesorectum was developing. The idea seemed strengthened by the referral of a suture line recurrence later that year when the emptying of the pelvis for an ultra-low anastomosis revealed a large mesorectal remnant with a cancer focus growing from it into the suture line (Figure 1.5). The other three cases pointed in the same direction. The paper was published in the British Journal of Surgery entitled ‘The mesorectum in rectal cancer: the clue to local recurrence?’1 This key publication in the rectal cancer literature was simply an idea based on clinical observation. One of my much loved former chiefs, Mr Peter Philip, wrote me only one letter; it said ‘That was the most sensible paper I have read for a long time.’

11/02/13 3:43 PM

MESORECTUM – THE WORD  5

1977

Figure 1.5. Suture line recurrence.

Re-Operation for Suture Line Recurrence 1977

The ‘Russian gun’ The ‘Russian gun’ saga preceded TME and landed me in trouble before the 1970s were out, and it certainly contributed to my preoccupation with TME. The legendary figure of John Goligher, then Professor of Surgery in Leeds, showed me the circular stapling gun from Russia, which he had bought from a bazaar in Turkey for a few stray coins. ‘Take an interest in this, young man: it may well have a place in low anterior resection.’ So I bought one on my travels, and, to my eternal shame, it sat unused in a cupboard for a year until my encounter with Anne, a young woman with a large ulcerated carcinoma 5 cm from the anal verge. For a young woman, recently divorced and in her early twenties, the thought of a permanent stoma was unbearable. At this time, the TME story in my own mind was at the evolution point where the challenge of meeting this patient’s hopes was just what was needed to push the TME idea to its logical conclusion. Perhaps even such a low cancer, at 5 cm, could be safely removed with all its relevant field of local spread without sacrificing the pelvic floor and the sphincter muscles. The mesorectum as a surgical craft entity was being born – a dream of relevant radicality without unnecessary mutilation.2 For Anne, I deemed it possible to achieve sphincter preservation, which, though extremely unorthodox, might just be achievable with that Russian gun. Tortured by the risk of local recurrence, I attempted my first TME in 1978. For me, the anastomosis with the

HEBK001-C01_p01-30.indd 5

Russian gun at about 3 cm was the lowest I had ever attempted, and the circumferential clearance was a long slow attempt to circumnavigate the cancer by dissecting in clean surgical planes. Anne thrived for 3 years and then sadly developed liver secondaries, for which I performed a hemi-hepatectomy. At this time she had found a new partner. Three years later she was watching television when she fell unconscious from a cerebral secondary and did not wake up – merciful but very, very sad. She never had pelvic recurrence, so the seeds of hope for lower anterior resections were sewn, and my first TME seemed to me at least to have been justified with local disease control.

Mesorectum – the word I had heard the word ‘mesorectum’, although it was not in Gray’s Anatomy or anywhere else that I ever found. My mentor and trainer at Guy’s, Mr Rex Lawrie, used the word ‘mesorectum’ to describe the fat behind the rectum that we divided during conventional anterior resection. This fat often bled in a tiresome way and so was the object of irritated abuse on occasions. We have since seen a report of the word ‘mesorectum’ in an ancient radiotherapy account of the use of radiotherapy in rectal cancer in 1914, but, for practical purposes, the word was undiscoverable at the time. It is attributed in my mind to one of my personal giants, Rex Lawrie, on whose shoulders much of my surgical thinking had certainly been based.

11/02/13 3:43 PM

6  The evolution of a concept: the TME story

Figure 1.6. Anatomist’s mesentery.

Words Matter In his book Talking Sense,2 Richard Asher points out that ‘physicians indulge in conjectural names’ and cause a great deal of confusion as a result. He quotes from Socrates: ‘He who first gave names and gave them according to his conception of the things they signified, if his conception was erroneous, shall we not be deceived by him?’ I believe that both Socrates and Richard Asher would be happy with the phrase ‘total mesorectal excision’ as it describes exactly what the surgeon should try to do. It is not easy to do well, but, very fortunately, the surgeon who believes in the TME concept and attempts to do TME is hugely rewarded by improved results. One enduring problem is that the word ‘mesorectum’ was, and remains, anathema to many anatomists. Even the current edition of Gray’s Anatomy describes the mesorectum as the mesentery of the sigmoid colon. Anatomists describe a ‘meso’ (e.g. mesocolon) as simply two layers of peritoneum with blood vessels and lymphatics between them and therefore question the use of the term ‘mesorectum’ (Figure 1.5). For me, however, and fortunately for the future of the concept of TME, this was the point where oncology and embryology met, and the choice of the word became descriptively critical and entirely acceptable to surgeons who were trying to define the primary field of spread of a rectal cancer. Thus, the mid-rectum is entirely encompassed circumferentially by its own fat, lymphatics and vascular supply, all enveloped in a fascial covering, with a recognizable practical surgical plane outside it. Anatomist Professor Morgado proposed that only a two-layered mesothelial structure surrounding and

HEBK001-C01_p01-30.indd 6

suspending a segment of gut may count as a mesentery and that the term ‘mesentery’ is not applicable to the rectum, since the normal rectum is devoid of an encircling, suspensory mesothelial sleeve (Figure 1.6).3 I replied: To the oncological surgeon the term ‘mesentery’ implies an integral visceral, lymphovascular and fatty entity that nourishes the organ concerned, and is developed embryologically in concert with it. In the case of the mid-rectum, this fatty lymphovascular structure surrounds the rectum completely, and no other word but ‘mesorectum’ seems readily available to describe it. Appreciation of the surgical anatomy of this entity, and exercising particular care when dissecting around it, are important components of good surgical technique in relation to rectal cancer. Graham Hill’s term ‘extrafascial excision of rectum’ may be more acceptable to morphologists – but is unlikely to displace the term TME from the surgical nomenclature …’

and ‘his’ fascia is indeed that which surrounds ‘our’ mesorectum.

The initial impact Early experiences of applying the circular stapling gun resulted in lower and lower anterior ­resections.4 This stimulated interest in the guns by other surgeons, which led to my first travelling phase, largely limited to England. As a young surgeon, I was

11/02/13 3:43 PM

Data collection and analysis and the pivotal role of Rosemary Sexton  7

Figure 1.7. My first travelling phase.

instructed to turn up with ‘the gun’ after the cancer had been removed by the host. The outcome of such a visit to St Mark’s Hospital in London to demonstrate the ‘new toy’ is nicely illustrated by my friend Mick Elliot (Figures 1.7–1.9). The Russian gun required that the staples be individually loaded in addition to the washer on to which the circular knife cut after firing of the staples. Undoubtedly the fastidious Professor Goligher measured each washer’s size and thickness, but I was content for others to do this, and I assumed the washers were uniform. What happened at St Mark’s was that the so-called ‘doughnuts’ were not cut through, although the staples had fired. The washer was too thin and the gun could not be removed. Perhaps the Basingstoke Gazette (Figure 1.10) was right! We all rejoiced when the US company Autosuture started to manufacture disposable, more reliable stapling instruments.

Data collection and analysis and the pivotal role of Rosemary Sexton Evidence was appearing of wide variations in the local recurrence and survival rates between surgeons, and Hermanek suggested that TME might be at the heart of this. The maverick practices of a surgeon in Basingstoke were unlikely ever to change the practice of surgery in a fundamental way. Only the most unchallengeable data with every possible discrepancy covered were likely to be of any value if what we were discovering was to turn out to be of any importance. Data collection and management were initiated by Rosemary Sexton, formerly my secretary, whose ­husband David worked for Xerox and who gave us an enormous computer for this purpose. We were ably backed by my cousin Pam who helped us to create our own dedicated software to correlate preoperative and operative detail with outcomes in a unique way. Thus, my initial data collections were

Figure 1.8. Staples fired, doughnut not released.

HEBK001-C01_p01-30.indd 7

11/02/13 3:43 PM

8  The evolution of a concept: the TME story

Figure 1.9. Goodbye, Bill!

comprehensive and consecutive so that there were truly no exclusions. The quality of the data was enormously enhanced by the fact that Rosemary made it her business to get to know every patient. Of course we made some mistakes, and the databases required continuing alterations; for example, initially we failed to record the quadrant of the rectum where the cancer was situated. Rosemary attended the outpatient clinics and was truly the forerunner of what is often now called the ‘key worker’ in the UK or ‘case manager’ in Europe as part of the modern multidisciplinary team management of cancer. She provided a communication lifeline in the sometimes complex and terrifying journey that follows a diagnosis of rectal cancer.

The embryonic multidisciplinary team and Roger Ryall The unsung hero of the TME story is Dr Roger Ryall, a clinical oncologist with an intellect ahead

of his time, with whom I did a joint clinic. Dr Ryall’s view of the role of radiotherapy in rectal cancer was indeed ahead of his contemporaries. Throughout the USA, and in much of the UK, radiotherapy was being given to many patients postoperatively, which he considered to be inferior to radical radiation given before surgery. We selected for the latter only patients considered to be locally inoperable as judged by palpable fixity to adjacent pelvic organs. The other members of our embryo multidisciplinary team were Rosemary Sexton and stoma therapist Sister Anne Leppington-Clarke, who pioneered irrigation in southern England as a method of managing a permanent end ­colostomy. She also worked out ­beautifully how to cope with the tiresome temporary right transverse colostomy that we generally performed to protect a low anastomosis for the first 6 weeks. Transverse colostomy remains my preference to reduce the formation of small bowel adhesion and abolish the high output sometimes associated with loop ileostomy, which enjoys greater popularity at this time.

Early publications

Figure 1.10. From the Basingstoke Gazette, 1979.

HEBK001-C01_p01-30.indd 8

One key early publication emanated from my presidential address to the section of surgery of the ­Royal Society of Medicine. Two years earlier Mr Geoff Oates’ presidential address had included the words ‘I don’t know why you don’t believe Bill Heald – he is a very truthful chap’ – life-saving for me sitting in the audience at the time. In my own address I advanced my preoccupation with the embryology

11/02/13 3:43 PM

Early publications  9

and ­anatomy to explain the astonishing local recurrence figures that were beginning to earn me so much derision. Unexpectedly good outcomes led me to choose as my title ‘The holy plane of rectal surgery’5 and to suggest that embryologically determined planes might be a sounder basis for optimal oncological specimens and optimal outcomes than random sacrifice of the anus. I suggested that there were widespread misunderstandings about the word ‘radical’. Abdominoperineal excision was much more radical and therefore much more ‘curative’ than an anterior resection, but this was in complete contrast to the surprising data that we were accumulating. Total mesorectal excision, I suggested, was ‘relevant precise radicality replacing unnecessary mutilation’.

Publication by a Surgeon and Radiotherapist About Patients Not Given Radiotherapy Our little multidisciplinary team published the 115 cases in the Lancet with just Heald and Ryall as the authors with the rather straightforward title ‘Recurrence and survival after total mesorectal excision for rectal cancer’.6 This paper, with outcomes at 2–6 years, was based on only 115 curative cases (i.e. patients without metastases). I recall quite primitive attempts to work out confidence intervals myself, eventually aided by

one of the medical students who understood log ranking and Kaplan–Meier curves. The local recurrence of 3.7 per cent was finally calculated based on these 115 patients. It is intriguing that, when the numbers had quadrupled, the original Kaplan–Meier prediction was within 0.2 per cent of this initial calculation, and indeed with gradual narrowing of the confidence intervals over the years very little has changed from those original predictions. When a local recurrence rate of 3.7 per cent in ‘curative’ 115 cases compared so favourably3 with the then accepted figures of 20–40 per cent in various publications, it is perhaps unsurprising that the original data were greeted with some scepticism. A pointer to the future comes from the following quote in the discussion section of this Lancet paper: ‘On this evidence it is often safe to limit mural clearance and thus preserve the sphincters provided the mesorectum is excised intact with the cancer.’6

More than 100 ‘Away Cases’ in Sweden and Norway By the late 1980s the literature suggested that a Basingstoke surgeon could achieve local recurrence rate of less than 5 per cent with anterior resection. Many felt that either he was lying or he had redefined a ‘curative’ resection in some way to ­optimize ­outcomes reported. The Scandinavians, particularly in ­Norway

Figure 1.11. Using the smaller camera.

HEBK001-C01_p01-30.indd 9

11/02/13 3:43 PM

10  The evolution of a concept: the TME story

Figure 1.12. Professor Pahlman's depiction of the effects of TME workshops across Sweden.

and Sweden, had identified the issues of local recurrence in rectal cancer and, after evaluation, had decided that the addition of radiotherapy could result in slight improvements but at much risk to the patients and cost to the health-care system. Professor Soreide of Oslo invited me as a visiting surgeon to his videobased ‘super radical cancer surgery’ annual workshops at the Rikshospital. Here in 1 week, assisted by a succession of Norwegian surgeons, I performed ten televised TME procedures while trying hard to dodge around an enormous broadcast-quality camera. This news-gathering television camera had been given to me by Sony Broadcast, serendipitously a Basingstokebased company (Figure 1.11). The Swedish approach differed from Norway’s in that three surgeons from the central Ostergotland district asked if they might come to Basingstoke for a week. After three cases, and intensive study of video clips from my Sony recording kit, Erik Nillsen and his colleagues returned to Motala and Norrkoping, clearly convinced of the potential of this tedious ‘TME surgery’. They grasped that only one surgeon in each Swedish district could possibly hope to hone the necessary skills and understanding of the complexities of the deep pelvic anatomy. The Swedish TME ‘wave’ (Figure 1.12) was carried forward by Professor Rune Sjodahl, who organized a series of workshops in Linkoping. This was followed by the first of several honorary professorships, many lasting friendships and ultimately over 80 visits across Sweden. Subsequently, Stockholm became especially important, with special friendships established through

HEBK001-C01_p01-30.indd 10

Professor Bjorn Cedemark and the then young Torbjörn Holm. They and their colleagues coordinated what we often refer to as ‘Stockholm 3’, a succession to the famous Stockholm 1 and 2 radiotherapy trials. Publication of the impact of a series of workshops in the mid-1990s at the Karolinska Hospital, Stockholm was the first report to demonstrate that workshop training in surgical technique can improve cancer outcomes in a whole population. Ongoing educational activities that were thus started have continued to make a real difference throughout Sweden. Looking back, it is clear that any world conquest by the TME story occurred very gradually as surgeons grasped the unexpected improvements in outcome that could be achieved by the expenditure of time and painstaking effort in pursuit of the TME principles. The first TME after our first workshop in Perth, Western Australia makes it very clear that a ‘better’ specimen with intact margins will be much less liable to spill cells or leave cancer-containing residues. I labelled the goal as ‘specimen-oriented surgery’, whereby the surgeon’s mind is forever prioritizing the quality of the specimen as he or she dissects around it. Nakedeye inspection and a recorded evaluation of the shape, contours and perfection of the specimen thus became the key tool in the audit of the quality of the surgery.

Histopathology workshops: Phil Quirke As the Scandinavian workshops spread to other countries, histopathological evaluation of the

11/02/13 3:43 PM

A new form of surgical teaching  11

margins of excised TME specimens became part of the TME ­crusade in a multidisciplinary fashion. What had been the ‘Travelling Bill Show’ became, with Phil Quirke from Leeds, the ‘Travelling Bill and Phil Show’, and the workshops increasingly embraced not only video relay of the live surgery but also workshops for pathologists with detailed audit of the margins and the quality of the specimens. Naked-eye examination and evaluation of the perfection of the specimen are potentially humiliating for many surgeons but a key component of introducing objectivity and honesty to improve outcomes.

Television, BBC2 and Phil Hammond Without doubt, the most exciting television appearance of the TME story occurred on BBC2, when a series entitled Trust Me I’m a Doctor was conducted by a dynamic doctor and investigatory reporter called Phil Hammond. His colourful coverage of the ‘Phil and Bill Show’ embraced the experiences of one young and engaging patient, a natural television personality who emphasized, straight into the camera, ‘When it’s your life, you’ve only got one chance.’ I had often made a similar observation in relation to rectal cancer that ‘what is omitted at the first operation is lost forever’. Phil Hammond interviewed Bjorn Cedermark, who stated on camera: [Abdominoperineal resections] and permanent colostomies were 50 per cent down to 20 per cent, the local recurrence from 22 per cent down to 5 per cent, and what happened during 1994–1995 was simply the workshops. The important thing about this is that what we didn’t really know, we read Bill Heald’s paper for the first time and Bill knows as well as everyone else that a lot of people did not believe him. They said this is not right, we have all been doing this for years, it cannot be true – well it is true! And I think we have proven that we can really do something by changing our surgical habits also for a population.

Stockholm was perhaps the single most significant episode in the establishment of an extraordinary lifelong series of practical video workshops.

HEBK001-C01_p01-30.indd 11

A new form of surgical teaching: the live surgery video workshop and the birth of nerve recognition and preservation The special problem that has beset the TME story from the beginning is that the operation is difficult to perform, although the concept of ‘embryological envelopes’ is easy enough to grasp in theory. Since good-quality data were accumulating through the late 1980s and the Scandinavian enthusiasm gathered momentum, the demand for people to see exactly what was involved grew steadily. Invitations flooded in from around the world so that the next 25 years have become for me a steady stream of ‘away cases’, armed with a video camera and backed up by the ever willing Phil Quirke to bring on the histopathology audit, which was the essence of convincing surgeons of the need to be ‘specimen-oriented’. Phil’s commitment has become legendary and he has criss-crossed the world with the salutary message that the histopathologist should not only report on a surgeon’s work but also audit the quality with ruthless tenacity (Figure 1.13), which is apt to strike fear into the surgical heart. Total mesorectal excision video workshops have introduced surgeons across the world to the true detail of deep pelvic anatomy. In the late 1980s and 1990s, the demonstration of the seminal vesicles during a workshop evoked amazement while video images of dissection between them and ­Denonvilliers’ fascia were hailed a first-time experience. Neither cadaver dissection nor CT scanning had ever conveyed an understanding of Denonvilliers’ fascia, a trapezoidal bib or sheet of collagen extending down from the apex of the peritoneal reflection between the rectal envelope behind and genitourinary organs in front. Denonvilliers described in 1837, in his doctoral thesis, ‘l’aponévrose pubio-rectale’. Understanding this medially tapering trapezoidal sheet is truly the key to a proper TME in a low anterior cancer in male patients and also the key to preserving potency in all TME dissections in male patients (Figure 1.14). The anterior mesorectal fat, whose very existence was previously ridiculed by the top surgical establishment, is very real but easy to tear, so we have always advocated dissection anterior to Denonvilliers’ septum (Figure 1.15). The emergence and widespread use of axial MRI finally established in everyone’s mind that the anterior

11/02/13 3:43 PM

12  The evolution of a concept: the TME story

Before After

Figure 1.13. Histopathology audit.

mesorectum did indeed exist and that the septum was often recognizable on a good-quality image.

Video surgery Many of the initial video workshops were in Basingstoke, and it soon became apparent that the quality of standard videotapes and video cameras was inadequate to convey the detailed images required. Serendipity intervened in that the advanced products division of Sony Broadcast was situated in Basingstoke, only a few hundred metres from the operating theatres. Here the first commercially aspiring high-definition television was being developed, and the education department in Sony Broadcast became friendly and supportive. As a result I, who had started life in my teens as a

HEBK001-C01_p01-30.indd 12

cinema operator, found myself with a massive amount of broadcast-quality video editing material capable of handling videotape non-linear editing. In the 1980s one whole room at my house was taken over by the equipment necessary, and the cost in today’s terms would have been well over £250 000, provided without charge by Sony. Thus, we were able to record on professional video format hundreds of operations performed in a variety of different countries, all in broadcast quality and capable of conveying to a watching surgeon the essence of holy plane dissection and the beginning of the identification of the layer of nerves that surround the mesorectal envelope. Throughout the development of the video technology necessary to identify the autonomic nerves of the pelvis, the improvement in picture quality and the recording and transmission of detail have

11/02/13 3:43 PM

The identification of the nerves of urogenital function  13

Figure 1.14. Denonvilliers’ septum.

been crucial. Further serendipity provided the TME story with Jason Flowers, a man of unique talent in the field of computers and computer visualization of images. It is with his help that we have led the world in visualizing the fine detail, including the autonomic nerves and plexuses.

The identification of the nerves of urogenital function ‘Man recognizes only what he knows.’ (Goethe)

In the early days some of the most distinguished surgeons totally rejected the idea that we were seeing the autonomic nerves responsible for male

Figure 1.15. Dissection anterior to Denonvilliers’ septum.

HEBK001-C01_p01-30.indd 13

erection, ejaculation, and control and normality of urinary and faecal voiding and continence. One notable exception was Sir William Slack, a distinguished London surgeon at the Middlesex Hospital. He had, with his histopathologists, looked carefully at the periphery of rectal cancer specimens and correlated patient impotence with the nerves inadvertently removed by the surgeon. I was not innocent of this myself, and I still show an earlier teaching video in which I cut through the superior hypogastric plexus by getting into the plane outside it at the aortic bifurcation and then back into the correct plane by cutting through the nerves on which ejaculation and to some extent the control of urinary function depend (Figure 1.16).

Figure 1.16. Right erigent pillar.

11/02/13 3:44 PM

14  The evolution of a concept: the TME story

Most surgeons had been unaware of these structures, and I was still failing to grasp fully the onionlike anatomy that must be respected if these nerves are to be identified, recognized and preserved. The idea of the onion came from a Japanese visiting surgeon, Professor Yoshi Moriya. The onion is a helpful anatomical concept, with the mesorectum as the core, the autonomic nerve layer as the first layer, the Wolffian ridge ureters and vesicles as the next layer, and so on. Holy plane dissection is thus the very careful entry into, and pursuit of, the innermost proper surgical plane around the core outside the mesorectal fascia, while recognizing and preserving the nerve layer outside it. The detail of the autonomic nerves was the next frontier and the next major controversy. Nerve preservation is one of the most difficult challenges of a good TME. Mesorectal fascial dissection liberates the inferior hypogastric plexus in an area that was formerly disastrously and improperly named the ‘lateral ligaments’. Video teaching dissections made it clear that no actual ligament existed, only areas of adherence. Simultaneously Japanese anatomists Sato and Sato had also noted that middle rectal arteries coming from the internal iliac vessels were uncommon, being present in only one in five patients on one side or the other and bilaterally in one in ten patients. The routine of clamping and dividing something called a ‘lateral ligament’ that also appeared to contain major arteries and veins (traditionally labelled the middle rectal vessels) had to be challenged. My own practical experience was, and is, that substantial vessels at this level are rare. If the surgeon dissects carefully between the mesorectal fat and the inferior hypogastric plexus, only a variable few nerves entering the mesorectum and some minor vessels are encountered. These observations suggest that the ‘lateral ligaments’ are no more than areas of adherence between inferior hypogastric plexus and mesorectum that tether the specimen inferolaterally. The vessels that we used to clamp were the lateral intramesorectal branches of the superior rectals and a part of the lymphovascular field that should have been encompassed and removed. This stage of the dissection is very challenging and is more achievable with precise monopolar diathermy. Diathermy here, however, does bring into question the matter of collateral nerve damage, and so low settings and optimal methodology are essential areas for ongoing technical research.

HEBK001-C01_p01-30.indd 14

The emergence of magnetic resonance imaging in a Swedish workshop Initial Swedish workshops at the Karolinska Hospital involved two adjacent operating theatres and transmitted images plus continuous review of MRI scans with Dr Lennart Blomqvist, their expert futuristic radiologist. Groups of up to 20 Swedish surgeons attended, taking it in turns to assist with the operation. The informal relationship between small groups of surgeons and personal interaction has been a unique part of what is undoubtedly a completely new form of surgical teaching. The sense of involvement mixed with informality adds considerably to the learning experience. Similar arrangements now work extremely well in my regular Heidelberg workshop series.

Who needs enemies? Professor Paul Hermanek, a great friend and worldrenowned pathologist, once told me over breakfast that the TME project would progress much faster if I didn’t have so many people, many from England and beyond, belittling it. But to every force there is a counterforce, and such antagonism often breeds silent supporters from the quiet majority, and opponents often do more benefit than harm. One curious, almost dramatic example of this was the assault by Professor Isbister, at the time a professor of surgery in Riyadh. His antagonism appeared to take the form of an attack on Basingstoke, the innocent town from which the TME idea had emanated. Various articles appeared under the titles of ‘Basingstoke visited’, ‘Basingstoke revisited’ and ‘Basingstoke visited again’. The essence of the articles was that the data were wrong and meaningless. Since he had never visited Basingstoke, my replies constantly reiterated a warm invitation to him to do so, to see what we were trying to teach, and to give us his comments on our work, favourable or unfavourable. He never came. Isbister’s criticisms demanded robust replies and undoubtedly the ‘noise’ helped enormously in spreading TME around the globe. Where there is no argument there is often little interest. One of the more colourful condemnations that I earned for my impudence was from the then president of the South African Surgical Association, who instructed

11/02/13 3:44 PM

The low colorectal and coloanal anastomosis  15

his members to take no notice of ‘that fellow Heald, he is a flamboyant charlatan’. I rather cherish this particular accolade. So it has indeed been down to friends and foes along the way to make the final judgements. The fact that TME is now a standard established operation in virtually every country in the world implies that the friends have tended to win.

John MacFarlane: a Canadian in Basingstoke Professor John MacFarlane, Professor of Surgery at the University of British Columbia, Vancouver, Canada, felt inclined to reject British and American scepticism concerning the excellent reported outcomes from Basingstoke. He offered to join the fray by spending a sabbatical examining the methods and validity of the data. In the event he recorded a slightly greater favourable impact of ‘TME surgery’ on outcomes than had been reported, in both local recurrence and survival data. His publication in the Lancet, ‘Mesorectal excision for rectal cancer’,7 was hailed by Professor Hans Troidl of Koln as the first example in the history of surgery of one professor visiting the unit of another to validate controversial data and thus advance the craft of surgery. Figure 1.17 illustrates his comparison between Basingstoke patients treated with TME, without radiotherapy or chemotherapy, compared with the results of the Krook–Moertel publication extolling

80 58.5% 60 40 37.3% 20 0

0

1

2

3

4

Years after Randomisation

The low colorectal and coloanal anastomosis The production and widespread use of disposable circular staplers coincided with the emergence of TME in the early 1980s. Completion of the anastomosis involved insertion of a manual purse-string in the anorectal stump. Various operative tricks were developed to facilitate this often challenging undertaking but have now been superseded by cross-stapling techniques and hand-sewn coloanal anastomosis for the very lowest or to salvage rupture of a cross-stapled anorectal stump. During the mid-1980s it became apparent that it was easier and quicker to perform a linear staple line across the anorectal stump after the routine washout of the lumen below a Satinsky or Lloyd-Davis right-angled clamp. One of the great advances in stapling was the ‘triple stapling system’ described by my registrar at the time, and subsequently my first colorectal colleague in Basingstoke, which I have named the ‘Moran triple stapling technique’. The details are outlined in Chapter 7. We have used this to facilitate TME for more than 20 years. Throughout and beyond the 1980s, a firm friendship and cooperation developed with the stapling Patients without Recurrence (%)

100

the extra benefits of radiotherapy and chemotherapy added to standard Mayo Clinic and other North Central Cancer Treatment Group surgery.

5

100 78%

80 60 40 20 0

0

1

2

3

4

5

Years after Surgery

Comparison of lifetables for recurrence-free interval between NCCGT series (left) and this series of TME (right).

Figure 1.17. Our comparable Dukes “B” and “C” stage patients compared with North Central Cancer Treatment Group patients.

HEBK001-C01_p01-30.indd 15

11/02/13 3:44 PM

16  The evolution of a concept: the TME story

gun manufacturer Autosuture, which later became Tyco and then Covidien. Mutual cooperation led to sponsorship and support for workshops at home and all over the world, and collaboration led to the development of the TA45 (45 mm across) linear stapler for the triple stapling technique (the original technique was described with the TA55 stapler). We advised that the average human rectum required something more than the TA30 stapler while the pelvis required something less than the TA55 stapler for safe ­occlusion in the depths of the pelvis. This was a fine example of clinical and industrial cooperation in the interests of better surgical technology. Similarly, sharp dissection, which was the essence of development of the holy plane, was increasingly performed by pencil monopolar diathermy dissection, and modern Valleylab Triverse units have evolved to maximize cut and coagulation and minimize collateral damage.

Why is the plane ‘holy’? The phrase ‘holy plane’ came from the practical reality that this was the most satisfying interface to develop and deliver ‘holy space’ – an idea that came to me when visiting Jerusalem. It is a useful concept because the line that is clearly visible around the mesorectum on MRI, with appropriate traction and counter-traction and sharp dissection, turns into a potential space. It has always seemed intrinsically probable to me that this potential space, which allows the juxtaposed surfaces to move slightly in life, implies an encompassing layer, which is likely to block angiogenesis, the important concept of Judah Folkman, who I once had the pleasure of meeting. Angiogenesis is the induction of new blood vessels to nourish a spreading tumour, and it does seem intrinsically likely that the areolar tissue between embryologically different organs such as the mesorectum and the surrounding nerve layer may constitute an angiogenic barrier to the spread of cancer. This is little more than theory based on the unexpectedly good results and simple clinical observation.

The low and the ultra-low Total mesorectal excision was a feasible procedure for upper and mid-rectal cancers and led to a smooth muscle tube that is ideal for stapling. There emerged, however, various differences of degree of difficulty according to the height of the cancer in

HEBK001-C01_p01-30.indd 16

relation to the build and sex of the patient. Thus, in a slim woman it often proved to be easy to perform a sphincter-preserving stapled anastomosis at or near the dentate line, whereas some cancers in large men as high as 7–8 cm could occasionally prove extremely difficult to circumnavigate. The lowest tumours continue to be problematic, especially anterior tumours, and the lower the anastomosis the greater the risk of poor anal function and continence. This relates particularly to the number of visits to the toilet each day, to the occurrence of episodes of actual incontinence, and also to the annoying symptom known as ‘clustering’, where multiple visits occur rather close together. Unfortunately, increased use of preoperative radiotherapy has greatly increased the incidence of these symptoms over the past two decades. Poor function has always been a worry with very low anastomoses, and the disappointment of a patient cured of rectal cancer but with worse bowel symptoms after surgery is very problematic. Karanjia, Schache and I published a comparison of the worst function expected from an anastomosis at 3 cm from the anal verge compared with one at 6 cm.8 This important message justifies a subtotal mesorectal excision if it is possible to clear the cancer in higher tumours with a clear 5 cm of mesorectum. Even after TME, careful preservation of a small rectal reservoir is desirable when oncologically safe.

Basil Morson During these exciting years of exploration of lower and lower anastomoses, one of my early friends and supporters was Dr Basil Morson, Head of Histopathology at St Mark’s Hospital and successor there to the famous Cuthbert Dukes of the Dukes Classification. From the very beginning, Basil helped the lift-off of the TME idea. ‘I’m a save-the-anus man’ he famously stated to the movie camera when the first TME 16 mm film was produced in the early 1980s. He also made a highly significant scientific observation: ‘The palpable lower edge of a rectal carcinoma is almost always the microscopic lower edge.’ Various papers had demonstrated that intramural downward spread from the lower margin of a rectal cancer occurred only rarely and only in cases of a particularly high malignant potential and a high risk of coexistent metastatic disease. This statement by a person so greatly respected by generations of young surgeons in a TME movie made at

11/02/13 3:44 PM

Washing out the rectal stump against malignant implantation  17

St Mark’s was a huge boost.9 It reinforced the seminal paper from the 1970s by Norman Williams and colleagues who examined, and discarded, what used to be called the ‘5 cm rule’ for distal rectal clearance below a cancer.10 Therefore during the 1990s, we began to focus on what we called the ‘close shave’ in rectal cancer and published widely on this subject. From the practical point of view, if one could get a finger and thumb distal to the lower margin of the cancer, and a stapler could be placed beyond this, there was a high probability that even margins of less than 1 cm would not compromise cure, provided proper TME is performed. Repeated distal margin frozen sections and histological analysis of the distal ­‘doughnut’ from the subsequent circular stapling firing confirmed cancer clearance, allowing salvage of the sphincters without sacrificing cure. One of my highlights in 1996 was presenting the Norman Nigro lecture at the American Society of Colon and Rectal Surgeons.11 Norman Nigro revolutionized the management of anal carcinoma from major surgery (APE) to chemoradiotherapy alone. Years later, with Norman sitting in the hall, I presented my personal experience of 136 consecutive rectal adenocarcinomas within 6 cm of the anal verge – that is, very low cancers that would routinely all have had an APE in that era. The rather provocative title ‘APE: an endangered operation’ was the same title he had used years earlier. It incorporated the concepts and advantages of TME, as 77 per cent of the patients with these ultra-low tumours had undergone sphincterpreserving surgery and yet the local recurrence was well below 5 per cent (confirmed on extended subsequent follow-up). My hypothesis was that the vast majority of abdominoperineal resections, by most surgeons, were being performed for tumours that could safely have been removed without sacrificing the levators or the anal sphincters. My own very small number of APEs had gone very badly, and this was widely quoted, particularly in Japan. The reasons for these poor APE outcomes may have been partly technical, although it also represented the worst small subgroup.

‘The sword of Damocles’: anastomotic leakage For all gastrointestinal surgeons, two major technical issues haunt us: patients where delay has occurred in the diagnosis of strangulating intestinal obstruction,

HEBK001-C01_p01-30.indd 17

and patients who have died from the consequences of anastomotic leaks. In theory, both of these causes of death could be eliminated, or markedly reduced, by constant awareness and timely appropriate intervention. In practice, especially with modern early discharge, and discontinuity of care aligned with suboptimal recording of patients’ temperature, pulse and so on, patients still die from these causes. Reoperation for leakage needs to be timely and effective. The words of Rex Lawrie are worth reiterating: ‘It is really important that the second operation should go well because the third and fourth never work!’ This may not be absolutely true, but it is important that the senior professional is always there at that second operation.

Why Does Leakage Still Occur, and Why Is It So Frequently Mismanaged? It is well established that the risks of leakage increase the nearer to the anal sphincters, with five to ten times the risk for anastomoses within 6 cm of the anal verge compared with higher anastomoses. Contributing factors include haematoma formation in the presacral space and blood supply at the bowel ends. Over the years we published three papers on this galling subject.13–15 The haunting memory that 3 of the first 100 TME patients died as a consequence of anastomotic leakage probably led to very few in the subsequent hundreds and largely a consequence of constant anxiety and early intervention combined with a little good fortune. Faecal peritonitis arising within the pelvic cavity, away from the pain-sensitive peritoneum of the anterior abdominal wall, commonly goes unrecognized or is treated by repeated doses of analgesia. Indeed, anastomotic leaks are commonly confused with myocardial infarction, pulmonary embolism or other causes of serious illness with little abdominal symptoms or signs. My Dutch friends refer to ‘weekend tragedies’ because the operating team is not working and the on-call team may have a lower sensitivity to the possibility of faecal peritonitis and less of a feeling of ‘ownership’ of the complication. My various phases of defunctioning all or selected patients finally settled to defunctioning all patients whose anastomosis is low enough to have followed a TME, somewhere between the dentate line and about 4 cm above it. Loop ileostomy has

11/02/13 3:44 PM

18  The evolution of a concept: the TME story

recently gained favour over my preference for right ­transverse colostomy, which I personally regret. I consider a loop colostomy less likely to give small bowel adhesions, to defunction more effectively with little or no faecal residue between the stoma and the join, and to obviate the risk of a highoutput stoma. This is particularly relevant with the trend towards omitting mechanical bowel preparation, although the pendulum appears to be swinging back in its favour for these low cancers.

Washing out the rectal stump against malignant implantation My preference had always been to wash the rectal stump below a clamp. Viable tumour cells have been shown in animal models to produce tumours, especially in crushed devitalized tissue. Thus, a staple line is excellent soil for a viable tumour seed to grow. A paper from Sweden has reported a reduction in the local recurrence from 10.2 per cent with no washout compared with 6.0 per cent with a ­washout.12 Although it may be that washout is a surrogate marker for attention to detail, this seems a simple procedure with no complications and major potential benefits.

The Gina and Lennart story: the emergence of magnetic resonance imaging The UK, Sweden and the Netherlands lead the world in the use of MRI in staging and treatment planning in visceral cancer management. All of this is a direct consequence of the ‘TME story’. The two radiologists who first grasped the key importance of the precise visualization of the holy plane on MRI were both intimately involved with TME at a practical level, Gina Brown in the UK and Lennart Blomqvist at the Karolinska Hospital, Sweden. Similarly, my visit to Maastricht with the ‘TME circus’ touched the work of Regina Beets Tan, who has also become one of the world’s great MRI leaders. All three had already recognized the potential of MRI to delineate cancer and bowel wall muscle, essential for T staging. The ability to delineate the embryologically determined holy plane was of even greater potential importance. Magnetic resonance imaging could predict a clear margin in the

HEBK001-C01_p01-30.indd 18

mesorectal fascial plane (i.e. TME holy plane) or direct more extended surgery, preoperative neoadjuvant therapy or combinations of both for locally advanced tumours where the potential circumferential margin was involved or threatened. Magnetic resonance imaging has revolutionized the comprehensive planning of treatment strategy, is part of UK and many other guidelines, and has been the ‘mother and father’ of the multidisciplinary team concept. With the ability of all to view the cancer and its surrounding structures, multidisciplinary team decision-making is a major advance, but the experienced surgeon’s finger remains a most sophisticated probe for tumour height, the type of distal margin, and the mobility of the tumour on the puborectal sling. Looking back, it is interesting to record how I ‘discovered’ Gina Brown in her formative years as a research registrar. Janet (now Dame Janet) Husband, a leading figure in diagnostic radiology, and a friend from my days at Guy’s Hospital, gave me Gina’s name as the person to talk on MRI at a meeting at the Royal Society of Medicine. This was when I was President of the Section of Coloproctology, where a paper was to be read that pointed out how valueless MRI was in the pelvis. Having foresight of this, I exercised ‘chairman’s prerogative’ and invited Gina to ‘trump’ this paper and tell us what she was achieving. This underpins an important reality, special sequences are required, fine (2–3 mm) slices, small fields, cuts perpendicular to suspect margins, and so on. ‘The devil, as with TME itself, is in the detail’. In tracking down Gina to put the record straight for this meeting, I rang her hospital in Cardiff, where the telephonist revealingly said: ‘Oh, I know where she is – even this late at night. She is always working in her room in radiology.’ That single-minded work ethic has driven her to the very front of one of the world’s most exciting imaging advances – rectal cancer treatment planning. It is thus through Gina and Lennart that the history of MRI planning is so intimately bound up with that of TME. So much of the future will be about imaging that this reality is a major part of our story.

The home front: Pelican My operating career and TME drive would have been shorter by at least 15 years if it had not been

11/02/13 3:44 PM

‘Metastatic’ total mesorectal excision to the liver and peritoneum  19

for the organizational skills of Sir Peter Michael, creator of the Quantel Electronics Empire, Classic FM and countless other enterprises that appealed to his fertile imagination. He sculpted the Pelican Centre, which gave an independent home, an office and a personal ‘carer’ for me. Of all professionals, surgeons seem to need these ‘carers’ most of all. Jenny, Juliet and Emma have exercised the Pelican hand on the rudder – their names come as worthy successors to Jill, Rosemary and Pat in the twentieth century. Throughout, the Medical Director Dr John Fowler has overseen the key decisions and direction of the whole foundation with unique wisdom. The conquest of the world would certainly have faltered without Emma and all at Pelican. Pelican also expanded the horizons. The name is short for ‘pelvic and liver cancer’. It sought to encompass the two poll positions that our little Basingstoke Hospital was aspiring to in the world of surgery – rectum and liver. At the same time it sought to extend at least our intellectual involvement to the management of prostate cancer. The disease was essentially different – that which is lethal obscured by red herrings, and the diagnosis by a fallacious blood test, prostate-specific antigen (PSA). So it was that the Pelican team branched out into focal therapy for prostate cancer, yet another story that spins off from TME. In particular we have focused on precision imaging, which has the potential to reduce so much suffering, but that is another story!

idea to encompass at the same time the concept of multidisciplinary cancer board meetings. Serendipity and the wisdom of Sir Peter Michael produced at the Pelican a series of firstclass chief executive officers – first, Diane Hayter, who put the show on the road, and then Chris Beagley CBE. Chris brought a determination of a naval ship’s captain and the negotiating skills of a former civil servant at the Ministry of Defence to bear at a critical moment. Sir Michael Richards was so determined that his colleagues said ‘He’ll get his way in the end’, and Chris delivered for Pelican the biggest contract in its history and the most outstanding government-backed training scheme ever, the ‘multidisciplinary team TME development project’. All but one colorectal cancer multidisciplinary teams in England attended between 2003 and 2007 (148 of 149). Out of all this came complete acceptance of the principle of multidisciplinary team discussion for every patient and mandatory MRI for each patient. Real planning had started. After 8 years Chris passed the helm to Sarah Crane, who has skilfully negotiated a mass of burgeoning bureaucratic obstructions to clinical research, and she now keeps Pelican securely on course.

Total mesorectal excision and British medical politics

Surgical expertise and endeavour propagate and ‘metastasize’, and thus Basingstoke has wandered into two other fields where attention to detail and surgical technique are key. The first development was the ‘hepatic element’ of TME history with the appointment of Merv Rees in the late 1980s to tackle the problem of colorectal secondaries in the liver. I had become aware of the unexpected benefit attached to what precision hepatic surgery might deliver, especially on encountering Johannes Scheele from Erlangen in Germany. It was a pleasure to be a catalyst in Merv’s friendship and cooperation with Johannes and ultimately establishing Basingstoke as a world-class centre for hepatic resection. In addition Merv has spearheaded the educational centre, which has been key to our home efforts in ongoing education and is properly called ‘Merv’s Ark’.

The greatest tragedy of the NHS is hugely powerful managers who do not understand the once powerful, but now impotent, consultants. Communication blocked is progress deferred. Only those who can build bridges are those who can still achieve real improvement in hospital practice. Sir Michael Richards has this in spades. Appointed as the first ‘cancer tsar’, he looked around him through the fog of UK-critical Eurocare data for projects to improve outcomes and increase real cures. He took the trouble to travel to Sweden to confirm our own wholly unlikely story of Englishmen and Irishmen training Swedes, and he returned with a burning determination to repeat and build on the TME workshop

HEBK001-C01_p01-30.indd 19

‘Metastatic’ total mesorectal excision to the liver and peritoneum

11/02/13 3:44 PM

20  The evolution of a concept: the TME story

Another key addition has been the treatment of peritoneal malignancy, in particular pseudomyxoma peritonei of the appendix. Visceral surgeons who innovate find themselves a part of an international club of eccentrics. My encounter commenced with a telephone call from a general practitioner friend, Steve Tristram, who had been contacted by a childhood friend of his wife whose young husband had been diagnosed at open-and-close laparotomy ‘with a rare kind of cancer called pseudomyxoma but nothing can be done’. The answer to the question ‘Bill, can you do anything?’ was ‘I can’t, but I know someone who can – I heard him last week in Hong Kong.’ This culminated in a visit to Basingstoke by Paul Sugarbaker from the Washington Cancer Centre; Paul, assisted by me and Brendan, performed a 12 h operation, which gave the man 12 years of life. It was the first of what is now approaching 1000 operative cases, and the name Basingstoke has emerged once again, spearheaded by Brendan and now Tom Cecil with three younger colleagues, Steve Arnold, Faheez Mohamed and Arcot Venkat, joining the colorectal team in April 2009. That first operation will be remembered for many things, including three emergency visits by the fire brigade as a consequence of fire alarm activation by the high-powered diathermy smoke.

Worldwide total mesorectal excision National Total Mesorectal Excision Projects Perhaps the most extraordinary feature of the TME story is the number of other countries that set up national training projects in TME technique before Britain. The Anglo-Saxon world was certainly the last to show any enthusiasm. Professor Norman Williams voiced dissatisfaction to the press at our failure in the UK to introduce the training aspects formally at an earlier stage. All the fundamental ideas were crystallizing across the world from UK origins, but in other cultures before ours – the ‘Bill, Phil and Gina show’ has extended to what the Canadians recently called the ‘fab four’, embracing our Pelican revolutionary oncologist Rob Glynne-Jones. Excluded was our new leader Brendan Moran, who stayed at home for the wedding of Prince William and Kate Middleton. Casting back to earlier days, it must be recorded that

HEBK001-C01_p01-30.indd 20

willingness to believe the unexpected and almost unbelievable results and their potential benefit also started in Norway and ­Sweden, although the importance of ‘Das Konzept’ was foreseen by Markus Buchler in Germany. Much of the Scandinavian story has already been told, along with some of the film scenarios of operating rooms invaded by coffee machines with Swedish and Norwegian surgeons eager to absorb the TME message. This early enthusiasm was greatly enhanced in both teachers and taught by the growing involvement of Brendan Moran, whose ready Irish charm and consummate surgical skill soon established him as the future captain of the ‘Good Ship TME’ and ultimately the clinical lead in both the second Danish national project and the somewhat belated British Low Rectal Cancer National Development Programme (LOREC).

Ireland Many a rejected Englishman, smarting at the indifference of his fellow countrymen, has found solace and welcome across the Irish Sea. Conversely, Brendan, like many Irish surgeons, found similar solace in leafy Hampshire. Many Dublin surgeons, Liam Kirwan in Cork and surgeons from across the Emerald Isle embraced and practised TME at an early stage. Liam Kirwan published figures identical to those from Basingstoke very soon afterwards, the first firm confirmation that our data were real. To me, Irish TME philosophy runs thus: ‘Violation of the mesorectum is the clue to surgical failure – treat it delicately like a virgin’.13

Germany For me, the worldwide TME story begins in and ends with Germany. An early great honour was that Markus Buchler travelled to Basingstoke to visit one of our first workshops back in the 1980s. This was followed by an invitation from Bern, Switzerland to present the Kocher Memorial Lecture. Researching the life of Theodor Kocher is inspirational: he was a brave honest surgeon who believed in slow meticulous technique and the primacy of precision, exactly like the TME dream. Markus Buchler

Professor Markus Buchler rapidly recognized the potential importance of TME, and our book

11/02/13 3:44 PM

Worldwide total mesorectal excision  21

Stunning variability between surgeons in Germany: Paul Hermanek

Figure 1.18. Das Konzept.

Das Konzept Der Totalen Mesorektalen Exzision14 ­(Figure 1.18) published the many implications of new thinking. Markus regards cancer surgery as a question of compartments and emphasized that TME was the careful emptying of the central pelvic compartment. This amplifies the embryological midline mesorectal entity and makes us aware that each lateral compartment contains the internal iliac vessels outside the nerve layer and a septum of significant parietal fascia, which separates paired parietal structures from the midline visceral block. The visceral envelopes have arteries emerging from the front of the aorta, and middle rectal arteries from the internal iliacs are uncommon. At the time the German book was published, Christophe Maurer translated many chapters for me, and he remains a leading German exponent of TME. This first visit by Markus Buchler and his two friends will never be forgotten – a totally unsuitable patient, no MRI, a ‘fixed cancer’ and no planning. Videos of the operation survive to illustrate my ‘skeletons in the cupboard’ talk. These show me cutting the left ureter while trying to clear the cancer by excising the seminal vesicles. Lessons were learned that day: we needed planning tools (now MRI), and the ureters in such advanced cases should always be stented. More amusing was the cocktail party in the back garden of our village home when Phil Quirke sat on a pillar and cut up the specimen and passed around the slices; the guests, champagne in their hands, were offered these alternating with cocktail fancies and sausages on sticks. My wife protested ‘What will people in the village think?’

HEBK001-C01_p01-30.indd 21

Paul Hermanek has been called the ‘father of ­German histopathology’. He is, along with Cuthbert Dukes, Basil Morson and more recently Phil Quirke, one of the all-time greats among ‘surgeons’ pathologists’ – those who involve themselves in the detail of the specimen, those whose room is always open to the surgeon bearing fruits fresh from the operating room of a recent battle against cancer, eager for consultation and discussion with a fellow warrior. Such pathologists are one of the greatest weapons we have in the war against cancer. Paul made two seminal contributions in his paper,15 curiously published in the Italian journal Tumori. His observations were crucial: first, there was a huge variability between operating surgeons in their local recurrence rates; second, these variations were reflected in long-term survival. The clear implication was that poorer surgeons were failing to remove tissue potentially containing cancer. This could then go on to metastasise. Recent times in Germany

It is no surprise to me that, despite speaking no German – except perhaps for ‘zug und gegen zug’ (‘traction and counter-traction’) – I now feel most welcome of all in Germany. Having grown up during the Second World War and witnessed a Messerschmitt 109 shot down in anger, my life has gone full circle – what a desperately sad and destructive century the twentieth was for our great continent of Europe. At least we have moved on and the TME story is one of friendship containing many key German contributions. Professor Paul Hermanek’s observation of the unique, totally unexpected variations in the local recurrence rates in seven excellent German hospitals was seminal to the concept. Equally important, and still defying the understanding of some clinical and medical oncologists, was Paul’s observation that surgeons who achieved lower local recurrence rates also had more true cures and fewer metastases. This of course is critical to the importance of TME, because the natural history of rectal cancer is such that it remains locoregional in one central pelvic compartment for a very long time. During this period a perfect TME will remove all the satellites of cancer and thus cure the patient. Residual mesorectum containing cancer or implanted cells on the raw

11/02/13 3:44 PM

22  The evolution of a concept: the TME story

pelvic side-wall are extremely likely to lead to local regrowth and ultimately death, from either direct extension or metastases from the cancer left behind. Friends in Heidelberg have been the principal basis of my ongoing practical teaching life in ­Germany in recent years, together with regular visits to Erlangen, the home of Paul Hermanek. Two-day craft workshops still seem the premier teaching mode, while the modern video connectivity of high-definition laparoscopy stacks and the Da Vinci Robot can potentially extend the experience to larger and larger audiences around the world. In April 2012 I was proud to be a part of the Heidelberg surgery team transmitting an open TME by satellite to the German Surgical Association in Berlin for Markus Buchler’s presidential meeting.

that one could stand a cup of coffee on the perineum. This option in a slightly modified form is now emerging again in the Netherlands under the influence of Harm Rutten of Eindhoven. There is no doubt that this difficult dissection in through the perineal body with its proximity to the urethra on the one hand and the front of the specimen on the other would be very much better effected without turning the patient if further studies show that equally precise dissection is possible that way. The question still remains open, and the most difficult challenge is certainly the perineal body and the neurovascular bundles subserving erection. The problem is that there are no natural planes to guide the surgeon and there is almost no room for error.

Total mesocolic excision: Werner Hohenberger

France

The influence of TME extends beyond the rectum. Professor Werner Hohenberger in Erlangen has put down robust TME type principles for mesocolic excision in colon cancer. Such principles do seem, to rigid oncological surgeons like him, me and Brendan, sometimes to have been compromised by some surgeons following the rise of minimally invasive surgery. Ligation of the ileocolic vessels flush with properly exposed superior mesenteric vessels and intact untorn colonic mesenteries are two examples. It is a personal opinion that we are beginning to see the rise of German surgery towards its formerly dominant position before the tragedies of the 1930s and 1940s. It is therefore a source of pride that the widespread teaching of TME along practical workshop lines has been embraced so enthusiastically in Germany and may become a part of this rise in other surgical fields. Joachim Strassburg

At an early stage in our discussions about how unsatisfactory our technique for APE indeed was, the ‘Berlin position’ entered the argument. This was put forward by Professor Joachim Strassburg from Berlin, who claimed that perfectly satisfactory access to the back of the prostate and the challenging dissection between it and the front of the specimen could be effected without turning the patient. Essentially the Berlin position meant a very steep Trendelenburg with shoulder support and maximal flexion and abduction of the hips; the patient was so steeply head-down that Joachim pointed out

HEBK001-C01_p01-30.indd 22

There are probably few stories of human endeavour where the French have not added their own unique dash of flair. So it is with our story. The word is laparoscopy, and the minimally invasive hall of fame is littered with the names of Frenchmen – Marescaux, Leroy, Rullier, Panis. Even before the laparoscopic revolution I overheard a most memorable compliment from Professor Rolland Parc addressing a group of French surgeons, describing TME as the ‘l’idée la plus importante du monde’. He, and later Professor Emmanuel Tiret, at the Hôpital St Antoine have established a centre of excellence with TME at the heart of its rectal cancer surgery. It was my privilege to undertake a three-screen review of their technique as long ago as 1995, and their APE results suggest that they are well ahead of us when it comes to avoidance of the ‘waist’ that bedevils APE. The rise of laparoscopy has been gradual from benign disease to malignancies, starting with smaller, easier and earlier, and progressing with caution to larger and deeper. Back in 1990, my videos of sharp dissection under direct vision seem to have inspired Professor Joel Leroy at a meeting in Tours: ‘I determined that day that we would reproduce precisely with the laparoscope what Heald was doing’. Over the years this process has progressed like two long-distance runners – me for example with threedirectional traction as the key tenet of teaching, and him adding a fourth dimension because of the pneumoperitoneum, and so on.

11/02/13 3:44 PM

Worldwide total mesorectal excision  23

Joel said: ‘I did probably the first laparoscopic TME with coloanal anastomosis in November 1991 in a young male patient [after chemoradiotherapy]. It was a successful procedure and satisfactory oncologically’. A postscript can be added in Joel’s own words that provides perhaps a fantasy glimpse of the future. I have news for you! I am writing for Archives of Surgery the first NOTES transanal TME without abdominal assistance and diverting stoma using a particular procedure to mobilize the sigmoid and descending colon and divide the mesenteric vessels. I called the approach PROGRESS – perirectal oncologic gateway for retroperitoneal endoscopic single site surgery: all the dissection is done by transanal retroperitoneal approach.

Eligio Floscoli of Covidien has also been persuading me of the merits of dissecting upwards in the holy plane from the perineum, along with John Marks from Philadelphia, the delightful son of the illustrious Gerald Marks. Who knows? The holy plane may perhaps be accessed entirely by the ‘unholy orifice’! As we go to press, the charismatic Spaniard Antonio Lacy and I are also focused on this possibility. Video-surgery

The primacy of surgical ideas being transmitted by video is reflected in the rise of great surgical teaching units with ever improving visual relay of ever more surgical detail. Nowhere is this more evident than in the magnificent Institut de Recherche contre les Cancers de l’Appareil Digestif (IRCAD) European Institute of Tele Surgery (EITS) in Strasbourg. Here I rejoice in the wonderful title of ‘Pope of the Rectum’, a manifestation of the ready wit of Jacques Marescaux, the great French surgical entrepreneur. The virtuosity of Leroy and of a rich line-up of invited surgeons provides unique training in the potential of colorectal laparoscopic surgery – always a wonderful weekend for everyone, faculty and visitors alike. In the background is the comprehensive Websurg providing accessible training on demand from cyberspace. Extraordinarily the Marescaux tentacles have replicated the IRCAD miracle in Taiwan and Brazil, where it is also my privilege to preen myself as ‘Pope’. The world owes Marescaux a lot. Elsewhere in France the saga has continued. An open TME operation by myself in Bordeaux at the

HEBK001-C01_p01-30.indd 23

invitation of the great laparoscopic surgeon Eric Rullier tested the idea that certain patients with very low cancers could only be performed open. In 2010 Eric believed that this was true, but he now believes that such patients do not really exist for the highly experienced laparoscopist, except occasionally for those with certain very large cancers. If we extend this story to liver secondaries, France can also boast one of the greatest of all liver surgeons – Professor Henri Bismuth of the Hôpital Paul Brosse. He has also exchanged visits with Basingstoke.

Sweden Sweden has already featured in the story, but, around the turn of the century, the Scandinavian data collection was to prove a new weapon of enormous power. Throughout my time on the council of the Royal College of Surgeons, the strenuous efforts of the maxillofacial surgeon John Williams to introduce compulsory registration of cancer and comprehensive clinical outcomes recording in the UK had been sidelined by the Department of Health. The ­manager-dominated NHS was far more interested in politically sensitive issues such as waiting lists. Not so the Swedes or Norwegians. National training projects in both countries gathered momentum, backed by compulsory pro forma registration of all the clinically relevant information – very similar to ‘Rosemary’s database’ back home in Basingstoke. What was being recorded, at last for a whole country, was the information that cancer doctors were interested in – that is, those things that might influence cure and patients’ suffering. Such enlightened backing has reaped many rewards. Most important was the ‘Martling papers’, which demonstrated without doubt the impact of our new workshops.16 More randomly, on a recent visit to Gothenburg, which had formerly resisted what Lars Pahlman called the ‘Heald wave’, real evidence emerged of the value of irrigation of the anorectal segment before stapled anastomosis. As a believer in the potential of the NHS for accumulating clinical data, I can only grieve that we have not explored its potential in the way that they have in Sweden. Fortunately the Association of Coloproctology of Great Britain and Ireland (ACPGBI) is now addressing this matter urgently.

11/02/13 3:44 PM

24  The evolution of a concept: the TME story

Norway The Norwegian data have a special interest of their own. As in other matters, the Norwegians have shown remarkable independence and non-compliance with the American National Institutes of Health (NIH) consensus advice that led to neoadjuvant therapy sweeping across the world. Their series reflects a very small usage of radiation and reports results very little affected thereby, and of course superior functional results because low anastomoses and radiotherapy are unhappy bedfellows. A substantial book was written early on about the Norwegian project, and Norway’s influence endures through the years.

The Netherlands The formidable Dutch entrepreneurial surgeon Cornelis (Cock) van der Velde first approached me in 1993. At that time he had no real funding and was indeed flying by the seat of his pants, with his dream of improving cancer surgery. From our early meetings came the Dutch TME trial, which has played in the colorectal concert halls for almost 20 years, generating more theses than any other project ever. The trial was a development of Cock’s earlier D2 Radical Gastrectomy Cancer project, and the train-

ing component was already under way with the Japanese colorectal surgeon Yoshihiro Moriya. He visited me at home in Basingstoke and we operated together several times. Memorable to me, and to the patient who I later told, was his comment about the male hypogastric nerves: ‘Englishman has very big sex nerves’ he announced loudly to the operating room. I declined to comment in the hope that we might, if this view was to be widely promulgated, thereby elevate the street credibility of the English. This may have succeeded as I heard it repeated some years later. Both Yoshi and I criss-crossed the traffic-packed roads of the Netherlands in a unique enterprise to change surgical habits for a whole nation of 20 million people. A posse of mentors followed behind us providing surgical discipline unique to the Netherlands. For me, a particular and special Dutchman was a surgeon from the Academic Medical Centre (AMC) in Amsterdam, Carlo Taat. It was he who took me to every major hospital in the Netherlands and attended every one of almost 50 demonstrations that I performed there. Together we realized that learning TME is a progressive process, the demonstration of the posterior dissection being much easier than the front. The demonstration of ‘Bill’s billen’ (Dutch, ‘buttocks’) by a surgeon is perhaps the first and easiest sign that he has grasped the importance of holy plane dissection (Figure 1.19).

Figure 1.19. Bill’s ‘buttocks’.

HEBK001-C01_p01-30.indd 24

11/02/13 3:44 PM

Worldwide total mesorectal excision  25

His high intellect contributed much unrecognized wisdom to the importance and impact of the Dutch TME trial in the world of coloproctology. He later drove me round Baden Wurtenburg in south-west Germany on a similar mission. Sadly, I later went to the Netherlands to say goodbye to him when he decided to exercise the right of Dutch citizens to end their own lives – his duodenum was obstructed by an irremovable intractably painful malignancy. As one of the two surgical trainers, I must confess to disappointment at the high level of reported margin involvement – 23 per cent – in the Dutch data. So, although we cannot interpret the surgery in the Dutch trial as being truly optimal TME surgery, we salute Cornelis and Carlo’s courage in tackling the enormous task of attempting to convert an entire country’s surgical community to long, slow tedious TME dissection and careful planning. The very holding of the workshops had a remarkable impact, as analysed by Peeters (Figure 1.19). Further major information is emerging after 12 years of fastidious follow-up of this near-perfect prospective randomized controlled trial of radiotherapy versus no radiotherapy with no chemotherapy. There are more people alive in the group that had not been irradiated than in those who had received short-course high-dose radiation. This was despite about 6 per cent of extra cancer cures, but these have been more than cancelled out by additional deaths from other causes – second malignancies in the irradiated pelvis, cardiovascular deaths, et cetera. No explanation has emerged for the several largely undocumented

deaths that clearly followed radiotherapy, apart from some second malignancies in the irradiated pelvis and a hint of additional cardiovascular deaths. This is surely a major challenge to the world of clinical oncology to demand an answer. If we get such an answer, then TME will make yet another chapter in the cancer story. Figure 1.20 shows that the greatest impact of the Dutch TME trial was that the surgical training improved national outcomes immediately – workshops are the thing.

Belgium An interesting case report from the Procare Project states: Patient had had a sigmoid to anorectum anastomosis . . . visit by me at 14 days: niggling fever, normal water-soluble radial opaque enema X-ray shows no leak. However, rectal examination reveals the unique odour of gangrene on the examining glove. Proctoscopy shows the colon to be black and necrotic despite the absence of a leak at that stage. Laparoscopy was performed to mobilize the splenic flexure, excise necrotic sigmoid and effect a new stapled anastomosis. Satisfactory outcome.

This case emphasizes the importance of the temperature chart, the need for constant awareness of the risk of anastomotic failure, and the importance of examining a suspect anastomosis with a proctoscope

Figure 1.20. The impact workshops, the Netherlands. RT, radiotherapy; TME, total mesorectal excision.

HEBK001-C01_p01-30.indd 25

11/02/13 3:44 PM

26  The evolution of a concept: the TME story

to establish viability: viable ends mean that all faeces must be washed through and the anastomosis can remain truly defunctioned and not need to be taken down.

Italy As in so many aspects of life, Italy TME has embraced TME with generosity. Many TMEs have been performed in Rome, Milan, Bologna, Naples, Palma and Bari. One Armani-suited elder on Palma station greeted me as ‘Signor Mesorectum!’

China and Japan I have been made particularly welcome in these two so different and important countries. Each deserves a whole chapter in its own right, for rectal cancer is common in both and extremely individual national attitudes to the disease are apparent. Controversy has raged for decades between Japan and the west over the oncological importance of the internal iliac and obturator nodes, the prophylactic removal of which has been quintessentially Japanese but completely ignored by myself. My TME story is entirely about what Markus Buchler calls the ‘central compartment’. Hideaki Yano and Brendan Moran has addressed this fascinating issue in Chapter 14. China is perhaps the deepest mystery in the world, and rectal cancer surgery is not excluded. Each time I return from China I am a little more confused. Several outstanding professors have befriended me there, and there is no doubt that TME has made an impact in Beijing, Shanghai, Chengdu and Wuhan, in all of which I have done open TME workshops. One most interesting difference from western practice is that low anastomoses are often simply drained for 2 weeks or more against the risk of faecal leakage but with no diverting stoma created. Professor Gu at the Beijing Cancer Hospital has undertaken extensive TME histopathology micro-analysis and established that small node positivity for cancer is far more common than is generally believed – a cogent further reason for the TME ‘oncological envelope’ concept (Figure 1.21). It is better simply to regard mesorectal fat as ‘dangerous stuff ’.

HEBK001-C01_p01-30.indd 26

Figure 1.21. A perfect abdominoperineal excision is a total mesorectal excision wrapped in levator and sphincters.

Hong Kong Hong Kong has played a major part in TME, with more than 20 visits, countless workshops and video-operations, and the huge talent of Professor Michael Li of the Chinese University. Some of the earliest laparoscopic versus open TME ‘contests’ took place here, and Hong Kong has developed the most modern endoscopic laparoscopic operating room, where all combinations of endoscopes are available at all times.

Russia, Poland and Eastern Europe As I near the end of my career as a teacher of surgery, the shift towards the eastern part of Europe has been most noticeable. In 2010–2011 I was invited to Russia, Poland, the Czech Republic, Hungary, the Ukraine, Serbia, Slovenia (Figure 1.22) and Romania. The incidence of colorectal cancer across the world is highest of all in these countries, the top (or bottom) of the league being Slovakia. The reasons are probably the usual mix of dietary and genetic factors, but the need for good affordable management plans with good outcomes is nowhere more urgent. The first visit to Moscow in 2009 was memorable, with many hundreds of surgeons attending from more than 80 cities in 9 countries. The patient

11/02/13 3:44 PM

Worldwide total mesorectal excision  27 % 100 90 80 70 60 50 40 30 20 10 0

Local recurrence

5 year survival (incl St IV)

Permanent stomas % Before After

Local recurrence

21

4

5 year survival (incl St IV)

34

77

Permanent stomas

67

18

All above p5,0.05; stage distribution and RT rate (54% and 52%) unchanged. Figure 1.22. Astonishing results in Slovenia: 195 patients before and after two TME video workshops in 2009.

proudly announced to Russian television that he was a wild boar hunter, and he has since sent me more than 50 pictures of his continued activities in this field after his successful operation. This visit was the brainchild of Professor Petr Tsarkov, who has become a firm friend and through whose mediation Russia bestowed upon us its highest medical honour, the Petrovsky Professorship and Medal. These were formally conferred 2 years later in 2011. They stand as a pivotal step forward in the international spread of the TME idea (Figure 1.23). The cost of the disease and the management of failure is enormous, with the ever present nightmare of massively expensive last-ditch treatments from big pharma being simply unaffordable. Generally these buy only weeks of rather unenjoyable life. The real major solutions are earlier diagnosis and better surgery. Wroclaw, Poland

Some visits become fact-finding experiences for me. In Wroclaw, Professor Marek Bebenek had lined up eight operations to be performed over 2 days on two tables side by side in one room. Four were to be sphincter-preserving and four were to be APEs. It is self-evident that every APE should encompass

HEBK001-C01_p01-30.indd 27

Figure 1.23. The Petrovsky Medal and Professorship.

a perfect TME (see Figure 1.21) but that the sphincters and perhaps the levators should be accurately surrounding the distal part of the TME. Torbjörn Holm had already suggested improving APE by turning the patient into the prone jackknife position, removing the coccyx, and doing the job from behind. In Poland this was already happening and the words of Kocher from 1897 were key: ‘The surgery of the sacrococcygeal region has attained greater interest since it has been recognized that . . . access to the pelvic organs from behind is for many reasons preferable to that obtained from the perineum.’ Marek Bebenek taught me how right Kocher’s words were. I returned to the role of resident and was taken through what I now like to call the ­Kocher–Holm APE. Torbjörn Holm’s own chapter in this book is well illustrated with the emerging excitement of one very special advantage of the Kocher–Holm exposure. Whether Kocher himself included in his ‘many reasons’ the ability to identify the neurovascular bundles as they converge to the bulb of the penis is not recorded. This major advance is indeed what we demonstrated and videoed for the very first time at one of our Pelican operative workshops. Tom Cecil performed the upper end and placed a clip on the hypogastric plexus; later, with me at the audience end, the whole

11/02/13 3:44 PM

28  The evolution of a concept: the TME story

unique cadaver dissections and animations – all flexibly coloured and labelled. In my opinion nothing comparable has ever been achieved in the field of teaching of surgical anatomy. I recommend it. Please contact Ayhan Kuzu for more information [email protected].

The USA and Canada Figure 1.24. Neurovascular bundles after abdominoperineal excision. Note the proximity to the divided right puborectal muscle.

neuroanatomy at the back of the prostate was laid bare, with visual continuity established by the clip (Figure 1.24). The ongoing APE challenge still remains and is covered in Chapter 8.

A ‘TME Blockbuster’ from Turkey Students of TME now have access to a real treat: the perfect companion to this book. Ayhan Kuzu, Professor of Surgery from Ankara, with help from many of our authors (Heald, Quirke, Brown, Rullier, Holm), has created a virtual university course on the subject in a pack of three DVDs. The contents include most of our best videos plus Ayhan’s

For this mighty continent and our impact upon it, a whole book is required. We will highlight only a few themes. First was the excellent surgical technique I learned as a resident at the Ochsner Clinic in New Orleans, where John Ray and his colleagues had established one of the world’s first colorectal specialist units. At that time, at the Mayo Clinic anything above 5 cm was still under the general surgeons. From Canada came the all important intervention from John MacFarlane from Vancouver, which was critical to the establishment of our credibility; from Warren Enker in New York came the same ideas presented robustly and backed by immaculate data – plus an enduring friendship for me and the still memorable operation in Basingstoke where the left-handed Enker did the patient’s left side and, I, right-handed, did the right side (Figure 1.25). Along the way have been many interactions with admired American friends – the Norman Nigro lecture, visits and lectures at Memorial, Beth Israel, Steve Wexner’s presidential invited lecture at the

Figure 1.25. Warren Enker from his Ernest Miles Centenary Lecture.

HEBK001-C01_p01-30.indd 28

11/02/13 3:44 PM

Conclusion  29

Society of American Gastrointestinal and Endoscopic Surgeons, and so many more. More recently, in 2011 a major project involved us again in Toronto, where all the concepts – MRI, surgical and pathological – were promoted for the 12 million people of Ontario. For years these concepts have been a key part of the standard thinking of so many good American friends, including Stan Goldberg, David Rothenburger and Phil Paty. Professor Steve Wexner has done much for international surgical training, with special emphasis on Central and South America, Russia and Israel. Rob Maddoff and others are promoting these training methods in their own constructive ways for the betterment of US cancer management. 2012 is a special year for TME as it will provide the content for the Harry Bacon lecture at the specialist meeting of the American Society of Colon and Rectal Surgeons in San Antonio, and later the centenary cancer lecture of the American College in Chicago. In summary, TME has established a firm place in both British and American practice.

Anecdotal dramas from the operating room These are mostly tales that could have had a very unhappy ending. On a visit to Istanbul, the electricity was cut off as my right hand was poised to cut and we entered 65 minutes of total darkness. A similar event occurred at a Basingstoke workshop, when the building works for the Ark cut through the power line and all electricity, including the emergency reserve, disappeared. The visiting audience returned to their hotels but were called back over an hour later when, mercifully, reconnection was achieved; the patient lives 10 years on, after a brief moment of fame on the front of the local paper. In this case the blood supply to the rectum had already been divided so that non-continuation was not an option. Transfer to another hospital would have involved descent of a stairwell with the patient anaesthetized and the abdomen open or packed shut, as the lift had also failed. In another Basingstoke workshop I experienced uncontrollable bleeding, subsequently discovered to have been a result of inferior vena cava (IVC) occlusion from a prophylactic IVC filter as the preoperative CT had shown common iliac thrombosis. Total mesorectal excision had been performed and

HEBK001-C01_p01-30.indd 29

preparations had been made for anastomosis. The pelvis was packed and Brendan completed the surgery 2 days later. The patient is cured 8 years on. Perhaps the most dramatic incident occurred in front of one of my largest audiences ever, from all around the Balkans. The city was Belgrade. Sanctions were at work, and so the reserve diathermy machine was of unspecified origin. It later transpired that the settings had been known to no-one and were five times the manufacturer’s recommendations. The video shows that the camera recorded something amiss as there was electricity in the air and each touch of the diathermy makes the picture jump. I point towards the right hypogastric nerve, an arc of electricity leaps to ground on the patient’s right external iliac artery and a fountain of blood hits the light. There is an attempt to control the flow, eventual common sense to ask for audience support, and the appearance of a friendly vascular surgeon and subsequently my carefully completed TME. Many lessons were learned that day, and the then 21-year-old woman still survives in good health. More quaintly, I have presided over the top end of a widely transmitted colovaginal stapled anastomosis. We did recognize and reverse it, so all was well. The sharing of both triumphs and disasters with visitors and audiences has perhaps been a bumpy road towards better surgery around the world, but we claim always to have been honest with our visitors.

Conclusion This chapter is not finished. Perhaps it never will be. There are many gaps in the detail and many individuals who have contributed so much along the way, and so many patients who have been central to the story. It is remarkable to see the same TME principles being maintained and strived for as new technologies are applied – laparoscopy, robotic, NOTES and all. The cost-effectiveness of new expensive technologies such as robots is debatable, but it is gratifying that all are striving to achieve the same aim and are adding another brick to the TME wall. Total mesorectal excision has hugely stimulated the development of MRI cancer planning, has added sophistication to scientific histopathology, and has provided the most reliable and reproducible reference standard for surgical technique, and

11/02/13 3:44 PM

30  The evolution of a concept: the TME story

each of these is constantly evolving over time. To list all the major players along the TME pathway is impossible, but together they have made the word ‘Basingstoke’ at the heart of any meaningful discussion on optimal surgical technique for cancer and a landmark in rectal cancer management.

To Our Registrars Brendan and I share a sense of profound gratitude to the many surgical registrars who have slaved for hours over the human pelvis in the unique quest for greater skill and understanding. We started to compile a list and settled on avoiding the offence that an omission might give. We mention one symbolic name – Margaret Farquharson – because she probably slaved longer than any of you. If you have read this far and feel a part of the TME tale, then this ‘thank you’ is to you personally and with great sincerity. You are the very stuff of which the whole concept was constructed – painstaking precise surgery combined with a passion for doing things better.

References   1. Heald RJ, Husband EM, Ryall RD. The mesorectum in rectal cancer surgery: the clue to pelvic recurrence? Br J Surg 1982; 69: 613–6.   2. Asher R. Richard Asher Talking Sense. London, Pitman, 1972.   3. Morgado P. Total mesorectal excision: a misnomer for a sound surgical approach. Dis Colon Rectum 1998; 41: 120–1.   4. Heald RJ. Towards fewer colostomies: the impact of circular stapling devices on the surgery of rectal cancer in a district hospital. Br J Surg 1980; 67: 198–200.   5. Heald RJ. The ‘holy plane’ of rectal surgery. J R Soc Med 1988; 81: 503–8.   6. Heald RJ, Ryall RD. Recurrence and survival after total mesorectal excision for rectal cancer. Lancet 1986; 1: 1479–82.

HEBK001-C01_p01-30.indd 30

  7. MacFarlane JK, Ryall RD, Heald RJ. Mesorectal excision for rectal cancer. Lancet 1993; 341: 457–60.   8. Karanjia ND, Schache DJ, Heald RJ. Function of the distal rectum after low anterior resection for carcinoma. Br J Surg 1992; 79: 114–6.   9. Karanjia ND, Schache DJ, North WR, Heald RJ. ‘Close shave’ in anterior resection. Br J Surg 1990; 77: 510–2. 10. Williams NS, Dixon MF, Johnston D. Reappraisal of the 5 centimetre rule of distal excision for carcinoma of the rectum: a study of distal intramural spread and of patients’ survival. Br J Surg 1983; 70: 150–4. 11. Heald RJ, Smedh RK, Kald A, Sexton R, Moran BJ. Abdominoperineal excision of the rectum: an endangered operation. Norman Nigro Lectureship. Dis Colon Rectum 1997; 40: 747–51. 12. Kodeda K, Holmberg E, Jörgren F, Nordgren S, Lindmark G. Rectal washout and local recurrence of cancer after anterior resection. Br J Surg 2010; 97: 1589–97. 13. Reynolds JV, Joyce WP, Dolan J, Sheahan K, Hyland JM. Pathological evidence in support of TME in the management of rectal cancer. Br J Surg 1996; 83(8): 1112–5. 14. Buchler MW, Heald RJ, Maurer CA, Ulrich B. Das Konzept Der Totalen Mesorektalen Exzision. Basel, Karger, 1998. 15. Hermanek P, Wiebelt H, Staimmer D, et al. Prognostic factors of rectal carcinoma: experience of the German Multicentre Study. Tumori 1995; 81: 60. 16. Martling AL, Holm T, Rutqvist LE, Moran BJ, Heald RJ, Cedemark B. Effect of a surgical training programme on outcome of rectal cancer in the county of Stockholm. Lancet 2000; 356: 93–6.

Further reading Heald RJ, Leicester RJ. The low stapled anastomosis. Dis Colon Rectum 1981; 24: 437–44. Karanjia ND, Corder AP, Bearn P, Heald RJ. Leakage from stapled low anastomosis after total mesorectal excision for carcinoma of the rectum. Br J Surg 1994; 81: 1224–6. Schache D, Stebbing A, Heald RJ. Management of the pelvic space following low anterior resection. Aus N Z J Surg 1989; 59: 339–42.

11/02/13 3:44 PM

2 Anatomy of the rectum, anal canal and pelvic floor Thilo Wedel

Introduction Undoubtedly, anatomy is the mother discipline of all surgical interventions. This general notion holds especially true for rectal cancer surgery, as many of the successful advances in surgical treatment of rectal cancer have been derived from anatomical considerations. In fact, revisited studies of the anatomy of the rectum, anal canal and pelvic floor have led to substantial procedural modifications for both anterior rectal resection and abdominoperineal excision. One of the seminal anatomical contributions with regard to rectal cancer surgery has been the clear description of the major routes of lymphatic cancer spread. Since the introduction of total mesorectal excision (TME), including the removal of the rectum together with its surrounding lymphatic vessels and lymph nodes, local recurrence rates have been considerably reduced. This conceptual change was advocated more than 100 years ago, exemplified by Berkeley Moynihan (1908), who stated that ‘we have not yet sufficiently realised that the surgery of malignant disease is not the surgery of organs – it is the applied anatomy of the lymphatic system.’ Another milestone set by anatomical considerations has been the preservation of functionally

HEBK001-C02_p31-51.indd 31

relevant structures located in close topographical proximity to the rectum. Comparable to the nerve-sparing procedures introduced in thyroid gland surgery and radical prostatectomy, preservation of autonomic nerve plexi passing adjacent to the rectum has become an important issue to improve postoperative outcomes with regard to urogenital functions. Thus, from the anatomical point of view, rectal cancer surgery comprises at least a twofold challenge: oncologically curative removal of the rectum along the mesorectal plane, and at the same time optimal preservation of the surrounding structures, in particular the nerves responsible for urinary continence and sexual function. This chapter focuses on the anatomy of the organs to be removed and on the anatomy of the structures to be preserved during rectal cancer surgery. With regard to the anatomical nomenclature, one has to be aware that over the past few decades many different names have been given to the same structures in the anatomical and surgical literature. These differing nomenclatures have led to ongoing confusion and, in some instances, controversies regarding the exact topography or even the existence of a given anatomical structure. Irrespective of these terminological issues, however, a general agreement has been reached concerning the most relevant

11/02/13 4:31 PM

32  Anatomy of the rectum, anal canal and pelvic floor

anatomical key landmarks required to meet the expectations of modern rectal surgery.

Embryology Development of the Rectum and Anal Canal The gastrointestinal tract is an endoderm-derived structure passing through the body cavity as a primitive gut. The primitive gut can be subdivided into the foregut, midgut and hindgut, which are all connected to the inner body wall by mesenteries serving as access routes for blood and lymphatic vessels. The hindgut includes the distal third of the transverse colon, the descending and sigmoid colon, and at its most distal end the rectum and upper anal canal. The anorectal tube opens into a dilated pouch termed the endodermal cloaca. Initially the cloaca corresponds to a common cavity in the development of both urogenital and gastrointestinal organs. The Wolffian and Müllerian ducts and the bladder with the allantois and ureters open into its ventral portion and the anorectal tube opens into its dorsal portion. The cloaca is closed by the cloacal membrane formed by an inner endodermal layer and an outer ectodermal layer. From outside, the cloacal membrane displays a depression lined by ectoderm and named proctodaeum (anal groove). At a later stage, the common endodermal cloaca is divided by the primitive urorectal septum into an anterior part (urogenital sinus) and posterior part (anorectal canal) (Figure 2.1). The urorectal septum is composed of mesenchymal tissue growing downwards to fuse with the cloacal membrane, which is then also separated into a urogenital and anal membrane. The fusion area between the urorectal septum and the cloacal membrane corresponds to the primitive perineum, which in turn will form the perineal tendinous centre or perineal body. After degeneration of the anal part of the cloacal membrane, the upper anal canal derived from the endoderm is continuous with the lower anal canal derived from the ectoderm. Studies have shown that epithelial differentiation from cylindrical singlelayered rectal epithelium to keratinized squamous

HEBK001-C02_p31-51.indd 32

cloacal membrane cloaca

allantois urorectal septum hindgut

proctodaeum

urogenital membrane perineum anal membrane

urinary bladder urorectal septum anorectal canal

Figure 2.1. Embryonal development of the cloaca. The cloacal cavity is divided by the primitive urorectal septum and separated during the seventh embryonal week into an anterior part (urogenital sinus) and posterior part (anorectal canal). (Reproduced from Schünke, M. et al., Prometheus Atlas of Anatomy, Vols. 1 and 2. Stuttgart, Germany, Thieme Publ. 2007/2009, with permission.)

epithelium along the anorectal canal occurs in a craniocaudal direction, starting inside the anal canal and proceeding externally.

Relevance for Rectal Cancer Surgery The significance of embryological considerations for rectal cancer surgery comes mainly from the observation that the vascular and lymphatic supply of almost the entire rectum is derived from visceral blood and lymphatic vessels travelling within the mesorectum.

11/02/13 4:31 PM

RECTUM AND ANAL CANAL  33

This notion has been particularly emphasized by F. Stelzner, who subdivided vertebrates into two separate organisms based on developmental studies, namely into an inner ‘visceral individuum’ enveloped by an outer ‘somatic individuum’.1 In the visceral individuum, all organs are singular (e.g. stomach, liver, spleen, intestines, rectum), with singular blood and lymphatic vessels; in the somatic individuum, the organ units are bilaterally symmetric (e.g. two kidneys and ureters, two gonadal systems, two levator ani muscles, two semicircles of external anal sphincter). Accordingly, the blood and lymphatic supply of the rectum including the upper anal canal is provided by the visceral individuum (superior rectal artery, lymphatic drainage towards the inferior mesenteric lymph nodes), whereas the pelvic floor muscles including the external anal sphincter are supplied by bilateral symmetric vessels (internal iliac and pudendal vessels) derived from the somatic individuum. The border of these two embryologically different ‘individuals’ resides along the transition zone between the upper and lower anal canal at the level of the dentate line. Moreover, embryological studies have highlighted the fact that not only the rectum but also its vascular and lymphatic supply are commonly ensheathed by a fascial system termed the mesorectum, which remains clearly discernible in postnatal life. The mesorectal fascia represents the embryological delineation of the visceral from the somatic individuum confining the main route of rectal cancer spread. Thus, precise surgical dissection along these embryologically defined planes will enable an oncologically safe removal of the specimen with minimal vascular and nerve damage.

The recto-anal segment is 15–20 cm long and extends from the rectosigmoid junction adjacent to the promontorium down to the anal orifice. It corresponds to the dorsal intrapelvic organ compartment descending in front of the concave sacrum (sacral flexure) towards the anal hiatus of the pelvic floor. Due to the contraction of the puborectal sling, the anal canal is angled in a dorsal direction at the anorectal junction (perineal flexure), forming an angle of 90–1008 with the rectum (anorectal angulation) (Figure 2.2).

RECTUM AND ANAL CANAL

Anal Canal

The rectum and the anal canal are the last segments of the gastrointestinal tract responsible for mediating stool continence and coordinating the defecation process. Under normal conditions, the rectum is empty and the anal canal is closed by means of the anal sphincter complex and the haemorrhoidal plexus. Filling of the rectum by mass movements from the descending and sigmoid colon initiates the defecatory reflex, which induces relaxation of the anal sphincters, contraction of the rectal wall and deflation of the haemorrhoidal cushions to allow the passage of stool.

The rectal ampulla narrows at the level of the anorectal junction and becomes the anal canal (pars analis recti). The anal canal extends down and backwards to the anal orifice and is 2.5–4 cm long. The anal canal is where the gastrointestinal tube of endodermal origin (‘visceral individuum’) fuses with the skin of ectodermal origin and the surrounding striated pelvic floor and anal sphincter muscles (‘somatic individuum’). Thus, the anal canal represents the topographical junction where the nerve, blood and lymphatic supplies of visceral and somatic structures merge and overlap. Accordingly, the anal canal is

HEBK001-C02_p31-51.indd 33

Rectal Ampulla The upper part of the rectum has the same diameter as the sigmoid colon (approximately 4–6 cm), but the lower part of the rectum widens to form the dilated rectal ampulla, with a diameter of up to 16 cm, depending on how full it is. The rectal wall differs morphologically from the colon by the confluence of taeniae coli to a continuous longitudinal smooth muscle layer, permanent semilunar transverse folds, and the absence of fatty appendices epiploica. Most of the rectum is not covered by peritoneum, and thus the rectum is predominantly an extraor subperitoneal organ. Only the upper two-thirds of the ventrolateral rectal wall is directly related to the peritoneum, which reflects on to the bladder and seminal vesicles in males (rectovesical pouch) and on to the uterus and posterior vaginal fornix in females (recto-uterine pouch, commonly known as the pouch of Douglas).

11/02/13 4:31 PM

34  Anatomy of the rectum, anal canal and pelvic floor

external anal sphincter (deep part) external anal sphincter (superficial part)

haemorrhoidal plexus dentate line (anal valves & crypts)

perianal skin

intersphincteric groove (anal verge)

divided into an upper and lower segment separated by a transitional zone at the dentate line. Segments of the anal canal

The upper segment of the anal canal extends from the anorectal junction, defined by the superior border of the external anal sphincter (anorectal ring) to the transitional zone. The inner lining consists of a pink-coloured mucosa with single-layered cylindrical epithelium (zona colorectalis). At the transitional zone (zona transitionalis), the wet mucosa changes into a dry stratified squamous non-keratinized epithelium displaying a ­histological mosaic of cylindrical, cubic and flat epithelial cells. The transitional zone exhibits 8–12 readily discernible vertical anal columns (Morgagni) containing terminal branches of the superior rectal artery. Between these anal columns extend anal sinuses that form pocket-like folds at their lower ends, the anal valves or crypts. The circular line of alternating anal columns and anal valves corresponds to the dentate line (pectinate line, crypt line) considered to be the junction between the

HEBK001-C02_p31-51.indd 34

conjoint longitudinal muscle proctodeal glands

anoderm

external anal sphincter (subcutaneous part)

internal anal sphincter

Figure 2.2. Rectum and anal canal. The frontal section shows the inner relief of the rectal ampulla and anal canal, and cross-sections of the haemorrhoidal plexus, the internal and external anal sphincters and levator ani muscle. (Reproduced from Schünke, M. et al., Prometheus Atlas of Anatomy, Vols. 1 and 2. Stuttgart, Germany, Thieme Publ. 2007/2009, with permission.)

endodermal (cloacal) and ectodermal (proctodeal) parts of the anal canal. The lower segment of the anal canal extends from the dentate line to the anocutaneous line and corresponds to the anoderm lined by a non-keratinized stratified squamous epithelium (zona squamosa). The anodermal skin is devoid of glands and hairs but richly equipped with sensory somatic nerve endings, making it very sensitive to touch, pain and temperature. At the lower end of the anal canal, an anal intersphincteric groove (anal verge) is palpable between the bulge of the internal and external anal sphincters. Whereas the ‘surgical’ anal canal is defined by the entire segments from the anal ring to the anal verge, the ‘anatomical’ anal canal is confined to the lower segment and involves only the anoderm. At the anal orifice, the anodermal skin changes into true skin. The hairless perianal skin is of a dull brown colour and displays radial folds caused by the contraction of the corrugator ani muscle. The skin contains sweat, sebaceous and apocrine glands and is supplied by perianal blood vessels originating from the inferior rectal artery.

11/02/13 4:31 PM

RECTUM AND ANAL CANAL  35

Haemorrhoidal plexus

The haemorrhoidal plexus corresponds to a vascular network extending within the submucosa of the upper anal canal best described as corpus cavernosum recti (anulus haemorrhoidalis, glomera venosa haemorrhoidalia). The submucosal vascular plexus is characterized by arteriovenous anastomoses and dilated veins due to sphincter-like constrictions that provide a taut elastic apposition of the anal columns and thus an air- and liquid-tight closure of the anal canal. The haemorrhoidal plexus is supplied by branches from the superior rectal artery, which reach the corpus cavernosum recti from the right (7 and 11 o’clock, lithotomy position) and left (3 o’clock) sides. The blood is drained by veins penetrating through the internal anal sphincter and collected in the external rectal venous plexus. Due to the transsphincteric blood drainage, the degree of filling of the haemorrhoidal plexus is determined primarily by the contraction of the internal anal sphincter. Normally the draining veins are compressed by the resting tone of the internal anal sphincter, resulting in a physiological cushion-like swelling of the haemorrhoidal plexus. Relaxation of the internal anal sphincter during defecation allows proper emptying and down-sizing of the haemorrhoidal plexus for the passage of stool.

Anal Sphincter Complex The anal sphincter complex is composed of two concentric muscles: the internal anal sphincter composed of smooth musculature derived from the rectal wall, and the external anal sphincter composed of striated musculature, which is intimately fused with the surrounding muscles of the pelvic floor. Internal anal sphincter

The internal anal sphincter consists of circular smooth muscle bundles and corresponds to a distal thickening of the circular muscle layer of the rectal wall. The ring-shaped muscle is 5–8 mm thick and 2–3 cm long. In relation to the anal canal, the internal anal sphincter extends from the anorectal junction down to the anocutaneous line, with the most prominent part projecting on to the anoderm. The subcutaneous part of the external anal sphincter overlies the lower part of the internal sphincter, cre-

HEBK001-C02_p31-51.indd 35

ating the intersphincteric groove (anal verge). Due to its permanent involuntary contraction, the internal anal sphincter is readily palpable as a rigid cylinder, in particular when the striated external anal sphincter is completely relaxed (e.g. under anaesthesia). The longitudinal muscle layer of the rectal wall also continues distally into the anal sphincter complex. Smooth muscle bundles descend between the internal and external anal sphincter, towards the perianal region, and are joined by striated muscle fibres from the puborectal muscle (conjoint longitudinal muscle). Distally, the muscle fibres diverge, become fibro-elastic and insert into the perianal skin producing radial wrinkles (corrugator ani muscle). The most peripheral muscular septa radiate outwards and pass between the subcutaneous and superficial parts of the external anal sphincter into the ischioanal space. These fibres insert in the superficial perineal fascia and contribute to the separation of the ischioanal space from the subcutaneous perianal space. External anal sphincter

The external anal sphincter surrounds the internal anal sphincter, but both muscles are separated by connective tissue (intersphincteric plane). The external anal sphincter forms an elliptical cylinder approximately 15 mm thick and is divided by septa into three parts. Although the external anal sphincter is a striated skeletal muscle, its fibres are composed mainly of slowtwitch type I fibres mediating prolonged contractions suitable for maintaining an adequate basal tone. The deep part of the external anal sphincter is the thickest and most cranially located segment delineating the anorectal ring. Its fibres blend with the puborectal muscle and are not attached posteriorly to the coccyx, and so this muscular portion is able to follow the contraction of the puborectal muscle anteriorly to create the anorectal angle. The superficial part is attached firmly to the perineal tendinous centre anteriorly and to the anococcygeal ligament posteriorly and is therefore elliptical in shape. The subcutaneous part circumscribes the anal orifice and extends deep to the skin below the lower border of the internal anal sphincter. Smooth muscle bundles from the conjoint longitudinal muscle intermingle with striated muscle fibres of the external anal sphincter. In males, the three parts of the external anal sphincter are of equal dimension along the circumference; in females, the external anal sphincter muscle is reduced anteriorly

11/02/13 4:31 PM

36  Anatomy of the rectum, anal canal and pelvic floor

to about one-third of its posterior thickness, in particular due to its less developed deep part.

Perirectal Fascia and Spaces Comprehensive understanding of the topographical anatomy of fascia and spaces that surround the rectum is of utmost importance and an essential prerequisite for rectal cancer surgery. Since the introduction of the TME procedure, it has become obvious that successful surgery depends first and foremost on mobilizing the rectum along the correct planes to achieve both oncologically safe margins and optimal preservation of urogenital functions.

Initial attempts to describe the perirectal fascial anatomy have led to differing nomenclatures and controversies regarding the exact topography of those fascial structures enveloping the rectum. Macroscopic dissection studies on fixed and fresh specimens, histological and immunohistochemical studies and subtle intraoperative observations have shown, however, that the system of intrapelvic fascial layers allows clear identification of perirectal spaces suitable for proper and safe mobilization of the rectum (Figure 2.3). Rectal fascia/mesorectum

The inner surface of the pelvic wall, including the pelvic muscles (internal obturator, levator ani, coccygeal

middle rectal artery

peritoneum

inferior hypogastric plexus

rectal fascia

presacral fascia pelvic splanchnic nerves hypogastric nerve

superior rectal artery

retrorectal space presacral space with presacral vessels

mesorectum with lymphatic vessels

rectoprostatic septum (Denonvillier´s fasica)

rectosacral fascia , (Waldeyer s fascia)

anococcygeal ligament ischioanal space perineal body

HEBK001-C02_p31-51.indd 36

Figure 2.3. Perirectal spaces, mediosagittal section of male pelvis, left view. The rectal fascia and presacral fascia are highlighted to illustrate the perirectal tissue (mesorectum) and both the retrorectal and presacral spaces. Note that the retrorectal space is free of blood vessels and nerves and corresponds to the correct surgical plane for mobilization of the rectum during total mesorectal excision. (Reproduced from Schünke, M. et al., Prometheus Atlas of Anatomy, Vols. 1 and 2. Stuttgart, Germany, Thieme Publ. 2007/2009, with permission.)

11/02/13 4:31 PM

RECTUM AND ANAL CANAL  37

and piriformis muscles) are covered by the parietal pelvic fascia. The visceral pelvic fascia (endopelvic fascia) ensheathes the pelvic organs, including the rectum. In contrast to urogenital pelvic organs, the rectum exhibits an almost completely closed annular envelope called the rectal fascia (‘proper’ perirectal fascia, fascia propria recti). The rectal fascia corresponds to a rather thin connective tissue sheath and constitutes a morphological barrier, thereby preventing early penetration of rectal cancer. For this reason, it has also been described by Stelzner as a ‘delimiting’ rectal lamellae (‘rektale Grenzlamellen’).1 An alternative, clinically widely used term for the rectal fascia is ‘mesorectal’, introduced by surgeons to describe the complete removal of the perirectal tissue together with the rectum during TME. This surgical concept is based on the observation that the mesorectum – like the true mesenteries of the small and large bowel – ensheathes the tissues harbouring the major routes of blood vessel supply and lymphatic drainage of the corresponding bowel segment. In fact, the mesorectum surrounds the fatty perirectal tissue that contains

the branches of the superior rectal artery and the perirectal lymph nodes and vessels draining in a cranial direction towards the inferior mesenteric lymph nodes. The perirectal tissue enclosed by the mesorectum is most developed at the dorsal aspect of the rectal wall, producing two bulges (‘mesorectal cheeks’). Anteriorly, the mesorectum border is the rectogenital septum (Denonvilliers’ fascia), where the perirectal tissue is much thinner. The dorsal and ventral parts of the mesorectum resemble a continuous plane, but laterally the rectal fascia is not completely closed. At these sites the annular pattern of the mesorectum is interrupted bilaterally due to the reflection of the rectal fascia towards the pelvic wall (lateral rectal ligaments, rectal pedicles) to give access for blood vessels and autonomic nerves passing from the inferior hypogastric plexus to the rectal wall. As a consequence, surgical mobilization of the rectum along the mesorectum requires sharp dissection laterally, while posteriorly and anteriorly it can be achieved by making use of ‘self-opening’ surgical planes (Figure 2.4).

ureter (duplex) superior hypogastric plexus ovarian vessels

presacral fascia

presacral space with presacral vessels hypogastric nerve

superior rectal artery

retrorectal space rectal fascia mesorectum

Figure 2.4. Perirectal spaces, dissected female pelvis, cranial view. The right forceps is inserted into the rectum and pushed forward to illustrate the rectal fascia that surrounds the perirectal tissue (mesorectum) and contains the superior rectal artery (red vessel loop). The left forceps grasps the presacral fascia that separates the retrorectal space from the presacral space. Note that the superior hypogastric plexus (yellow strip) and hypogastric nerves are embedded into and extend along the presacral fascia. (Reproduced from Schünke, M. et al., Prometheus Atlas of Anatomy, Vols. 1 and 2. Stuttgart, Germany, Thieme Publ. 2007/2009, with permission.)

HEBK001-C02_p31-51.indd 37

11/02/13 4:31 PM

38  Anatomy of the rectum, anal canal and pelvic floor superior rectal artery pelvic splanchnic nerves

presacral fascia retrorectal space

hypogastric nerve inferior hypogastric plexus

middle rectal artery

presacral space with presacral vessels

rectal fascia T-junction

mesorectum with lymphatic vessels

rectoprostatic septum (Denonvillier´s fasica)

neurovascular bundles

seminal vesicles prostate

urinary bladder

Figure 2.5. Perirectal spaces, transverse section of male pelvis, cranial view. The rectal fascia and presacral fascia are highlighted to illustrate the perirectal tissue (mesorectum) and both the retrorectal and presacral spaces. Anteriorly the rectal fascia is bordered by the rectoprostatic septum (Denonvillier’s fascia). Laterally the rectal fascia is pierced on both sides by rectal nerves diverging from the inferior hypogastric plexus and by small vessels of the middle rectal artery (rectal pedicles, T-junction). Nerves from the inferior hypogastric plexus continue in an anterior direction as neurovascular bundles to supply the urinary bladder, prostate and seminal vesicles.

Lateral rectal ligaments/rectal pedicles

On both sides, the rectum is loosely connected to the pelvic wall by condensed connective tissue, traditionally described as lateral rectal ligaments. Originally these were considered to function as true ligamentous structures to support the rectal ampulla above the pelvic floor. From anatomical and developmental studies, however, it has become obvious that these ligaments do not provide substantial mechanical fixation but serve primarily as access routes for the vascular and nerve supply of the rectum, leading to alternative and more appropriate terms, such as rectal pedicles, rectal stalks and ‘paraproctial’ compartment. The rectal pedicles contain small-sized branches from the middle rectal arteries (if present) and the autonomic nerve fibre branches (rectal nerves) that originate from the dorsolaterally located inferior hypogastric plexus and traverse medially to enter the rectal wall. This connection between the rectal wall and the parietal pelvic fascia harbouring the inferior hypogastric plexus is also called the T-junction. As

HEBK001-C02_p31-51.indd 38

mentioned above, for complete mobilization of the rectum, the T-junction has to be sharply divided on both sides at the level of the rectal pedicles. Care must be taken, however, to stay rather close to the rectum to avoid damage to the inferior hypogastric plexus, thereby preserving sexual function and urinary continence (Figure 2.5). Presacral fascia

Behind the rectal fascia (posterior mesorectum) extends another fascial envelope following the concavity of the sacrum, the presacral fascia. The presacral fascia is part of the parietal pelvic fascia that covers the periosteal surface of the sacrum dorsally and reflects laterally towards the rectogenital septum (Denonvilliers’ fascia). The rectal fascia and presacral fascia are clearly separated from each other by the retrorectal space. When descending down to the pelvic floor, however, approximately at the level of the fourth sacral vertebra, the rectal fascia fuses with the presacral fascia. This connective tissue bridge between the rectal and presacral fascia

11/02/13 4:31 PM

RECTUM AND ANAL CANAL  39

hypogastric nerve

pelvic splanchnic nerves inferior hypogastric plexus

neurovascular bundles

middle rectal artery

urinary bladder

presacral fascia

T-junction

seminal vesicles prostate

rectal fascia rectoprostatic septum , (Denonvillier s fascia)

mesorectum

superior rectal artery

Figure 2.6. Perirectal spaces, sagittal section of male pelvis, left view. The rectum and the surrounding mesorectum are pushed towards the contralateral side to illustrate the course of the autonomic pelvic nerves along the pelvic wall. The hypogastric nerve and the pelvic splanchnic nerves join the extensive network of the inferior hypogastric plexus. Rectal nerves leave the inferior hypogastric plexus to pierce the rectal fascia and enter the rectal wall (rectal pedicle, T-junction). The remaining nerves of the inferior hypogastric plexus extend in an anterior direction as neurovascular bundles to supply the urinary bladder, prostate and seminal vesicles. (Reproduced from Schünke, M. et al., Prometheus Atlas of Anatomy, Vols. 1 and 2. Stuttgart, Germany, Thieme Publ. 2007/2009, with permission.)

can be quite thick and is also termed the rectosacral fascia or Waldeyer’s fascia. The presacral fascia is an important surgical landmark, as it separates the retrorectal space (surgical plane for a TME procedure) from the presacral space and it is intimately associated with the hypogastric and pelvic splanchnic nerves that form the inferior hypogastric plexus. The fascial architecture displays an inner and outer lamella, giving the presacral fascia the aspect of a double-layered sheath. The inner lamella borders the posterior mesorectum, confining the retrorectal space, and contains the autonomic nerve plexus. For this reason, it has also been called the hypogastric sheath or the pre-hypogastric nerve fascia. Further anteriorly, the presacral fascia reflects at the level of the rectal pedicles and extends medially towards the rectogenital septum to fuse with the prostatic capsule in the male. The outer lamella extends

HEBK001-C02_p31-51.indd 39

between the iliac vessels on both sides and delineates the presacral space. Laterally it is pierced by pelvic splanchnic nerves that originate from sacral spinal nerves and follow along the presacral fascia to join the hypogastric nerves to form the inferior hypogastric plexus (Figure 2.6). Rectogenital septum

Between the genital organs (vagina in females, prostate and seminal vesicles in males) and the rectum extends a fascial structure of considerable density originally described by Denonvilliers as ‘aponévrose pubio-rectale’,2 which is now recognized as the rectogenital septum. In females the rectovaginal septum is related to the posterior vaginal wall; in males the rectoprostatic septum (Denonvilliers’ fascia) covers the prostate, seminal vesicles and the most distal segments of the vasa deferens and ureters, separating them from the anterior rectal wall.

11/02/13 4:31 PM

40  Anatomy of the rectum, anal canal and pelvic floor

Although intraoperatively the rectogenital septum seems to comprise a single layer, subtle macroscopic dissections and histological studies have described a posterior sheath being the anterolateral lining of the mesorectum and an anterior sheath adjacent to the prostate. The rectoprostatic septum is clearly identifiable as a fascial structure at the level of the seminal vesicles and upper half of the prostate, but it often joins the prostatic capsule at the lower half of the prostate and the apex, making an unequivocal identification more difficult. The rectogenital septum continues laterally to the presacral fascia, in particular to its inner lamella containing the inferior hypogastric nerve plexus. The rectogenital septum is composed of condensed connective tissue and smooth muscle cells and is generally well developed. It is subject, however, to morphological alterations (e.g. weakening, thinning) with increasing age or diseases of adjacent organs (e.g. rectocele or enterocele in females, benign prostatic hyperplasia in males). Its surgical significance derives from the observation that multiple nerve branches of the inferior hypogastric plexus pass lateral to and in front of the rectogenital septum. These autonomic nerve fibres approach the seminal vesicles, vasa deferens, distal ureters and prostate dorsolaterally and comprise the so-called neurovascular bundles responsible for urinary continence and sexual functions. Consequently, this bilateral nerve network is at risk during surgical mobilization of the anterior rectal wall, in particular if the surgical plane is opened not behind but in front of the rectoprostatic septum with direct exposure of the seminal vesicles and prostate.

narrows and ends at the level of the fourth sacral vertebra, where the rectal fascia fuses with the presacral fascia (rectosacral fascia). Presacral space

The presacral space extends between the presacral fascia and the concave periosteal surface of the sacrum. This space is by no means an avascular or nerve-free area, as it contains the medial and lateral sacral arteries, the presacral venous plexus and the parasympathetic pelvic splanchnic nerves originating from the third and fourth sacral nerves. Thus, the presacral space corresponds to a ‘forbidden plane’ during the TME procedure, as surgical dissection may lead to nerve injury and, in particular, troublesome venous bleeding. When starting dorsal mobilization of the rectum, it is essential to clearly identify the proper plane for TME and not to mistake the presacral space for the retrorectal space.

Blood Supply of the Rectum and Anal Canal As the rectum derives from the hindgut, the rectum, including the upper anal canal, is supplied mainly by the inferior mesenteric artery via the superior rectal artery (Figure 2.7). The middle and inferior rectal arteries originate from the internal iliac vessels and contribute to the blood supply of the lower anal canal and minor parts of the rectum via intramural anastomoses. Occasionally, the posterior part of the anal canal and internal anal sphincter are additionally supplied by the median sacral artery.

Retrorectal space

The retrorectal space can be considered as a fascial interface extending between the rectal fascia (posterior mesorectum) anteriorly and the presacral fascia posteriorly. This distensible space corresponds to the proper plane for dorsal mobilization of the rectum which can be achieved by means of gentle traction and counter-traction. Opening of the retrorectal space reveals loose areolar connective tissue, which has also been described quite illustratively as ‘angel’s hair’ (‘cheveux des anges’) due to its delicate web-like appearance. As the retrorectal space corresponds to an avascular and nerve-free area, careful dissection will not result in major bleeding or nerve injury. Occasionally blood vessels originating from the presacral venous plexus pass through the fascial interface. When ­approaching the pelvic floor, the retrorectal space

HEBK001-C02_p31-51.indd 40

Superior rectal artery

The superior rectal artery corresponds to the direct continuation of the inferior mesenteric artery after the sigmoid arteries have diverged to supply the sigmoid mesocolon. The superior rectal artery is of considerable size (diameter 3.0 6 1.1 mm) and contributes more than 80 per cent to the entire rectal blood supply. It approaches the dorsal aspect of the rectal wall by entering the perirectal tissue completely wrapped by the mesorectal fascia. On its course within the perirectal tissue, the main trunk generally divides into three branches that surround the posterolateral wall of the rectum. The branches ramify further, enter the rectal muscle layers and the submucosa, and descend towards the upper segment of the anal canal. Terminal branches supply the haemorrhoidal plexus

11/02/13 4:31 PM

RECTUM AND ANAL CANAL  41 inferior mesenteric artery

median sacral artery

Lymphatic Drainage of the Rectum and Anal Canal

superior rectal artery

internal iliac artery internal pudend artery middle rectal artery inferior rectal artery

Figure 2.7. Arteries of the rectum and anal canal, dorsal view. The rectal wall, including the upper anal canal, is supplied mainly by the superior rectal artery derived from the inferior mesenteric artery. The middle and inferior rectal arteries originate from the internal iliac vessels and contribute to the blood supply of the lower anal canal. (Reproduced from Schünke, M. et al., Prometheus Atlas of Anatomy, Vols. 1 and 2. Stuttgart, Germany, Thieme Publ. 2007/2009, with permission.)

(corpus cavernosum recti), creating haemorrhoidal cushions at predilected areas (3, 7 and 11 o’clock with the patient in the lithotomy position). Middle rectal arteries

The middle rectal arteries originate from the internal iliac arteries mostly as direct branches. They pass above the pelvic floor muscles and approach the inferior rectum via the rectal pedicles. Thus, the middle rectal arteries intermingle with rectal nerve fibres that originate from the inferior hypogastric plexus and also run within the rectal pedicles (Tjunction). In contrast to the superior rectal artery, however, the middle rectal arteries are much smaller in size and inconstantly developed and are present bilaterally in only about 10 per cent of individuals. Inferior rectal arteries

The lower anal canal and the anal sphincter complex are supplied by anal arteries from the inferior rectal arteries. They diverge from the internal pudendal arteries located within the Alcock canal and approach the anal region via the ischioanal space, dividing into ventral and dorsal branches. Functional intramural anastomoses are established between the inferior, middle and superior rectal arteries. These multiple

HEBK001-C02_p31-51.indd 41

sources of rectal blood supply are essential for effective wound healing of deep rectal anastomosis after removal of the rectum and its major blood supply, the superior rectal artery.

Analogous to the blood supply, the lymphatic drainage from the rectum is mainly in a cranial direction and to visceral lymph nodes. Intramural lymphatic vessels project to perirectal lymph nodes located within the fatty perirectal tissue, which is completely enveloped by the mesorectal fascia. Lymphatic collection and drainage occur unidirectionally towards the major lymph node stations along the superior rectal and the inferior mesenteric blood vessels. Depending on the degree of individual development of the lymphatic vessels, lymphatic drainage may also take place via the rectal pedicles towards the internal iliac lymph nodes. In rectal cancer, this route of lymphatic spread is generally confined to advanced tumour stages in which the rectal fascia has been penetrated towards the lateral pelvic wall. The lower anal canal and the perianal region are not hindgut derivates and therefore receive a somatic rather than a visceral lymphatic drainage. Lymphatic vessels project mainly to inguinal and external iliac lymph nodes.

Nerve Supply of the Rectum and Anal Canal Like the other segments of the gastrointestinal tract, the distal end comprising the rectal ampulla and upper anal canal, including the internal anal sphincter, is controlled by the autonomic nervous system. In contrast, the somatic nerve supply is confined to the striated musculature of the external anal sphincter and to the lower anal canal up to the dentate line (Figure 2.8). Inside the rectal wall resides the enteric nervous system, which extends throughout the muscular, submucosal and mucosal layers by forming intramural nerve plexi. The myenteric and submucosal plexi are composed of ganglia and interconnecting nerve fibre bundles, which regulate most of the intestinal functions (e.g. motility, perception, secretion, absorption, immune functions). The enteric nervous system is linked to the central nervous system by spinal visceral afferents and both sympathetic and parasympathetic nerve fibres and ganglia.

11/02/13 4:31 PM

42  Anatomy of the rectum, anal canal and pelvic floor

superior hypogastric plexus

hypogastric nerve

S1 S2

inferior hypogastric plexus

S3 S4 levator nerves

pelvic splanchnic nerves rectal plexus

pudendal nerve

rectal wall external anal sphincter internal anal sphincter

Figure 2.8. Somatic nerve supply of pelvic floor muscles (left side) and autonomic nerve supply of the anorectum (right side), schematic drawing. Somatic nerves derive from sacral spinal nerves (S2–4), release direct nerve branches (levator nerves) to the levator ani muscle, and form the pudendal nerve to innervate the external anal sphincter and the perineal region. Autonomic nerves are composed of lumbar sympathetic nerves (superior hypogastric plexus, hypogastric nerves) and parasympathetic nerves (pelvic splanchnic nerves), which fuse in the inferior hypogastric plexus responsible for the nerve supply of anorectum, including the internal anal sphincter. (Reproduced from Schünke, M. et al., Prometheus Atlas of Anatomy, Vols. 1 and 2. Stuttgart, Germany, Thieme Publ. 2007/2009, with permission.)

It is important to emphasize that sympathetic and parasympathetic nerves supplying the rectum originate from an extensive nervous network that is also responsible for the control of urinary continence and mediation of sexual functions. As the rectum resides in close proximity to the autonomic nerve fibre meshes extending towards the urogenital organs, these nerves are in danger of damage during surgical removal of the rectal specimen. Sympathetic nerves support the filling phase of the urinary bladder under resting conditions by relaxing the detrusor vesicae muscle and closing the internal urethral sphincter, thereby contributing to urinary continence. For controlled micturition, parasympathetic nerves suppress the sympathetic input and initiate continuous contraction of the detrusor vesicae muscle. Thus, injury of these nerves may lead to both urinary incontinence and urinary bladder dysfunction, including retrograde ejaculation. Both sympathetic and parasympathetic nerves are also involved in regulating sexual functions. Sympathetic nerves mediate ejaculation, and parasympathetic nerves are responsible for the filling of the

HEBK001-C02_p31-51.indd 42

cavernous bodies (erection in males, swelling of the clitoris in females) and for lubrification of the external genital organs in both genders. Sexual dysfunction due to nerve injury is clinically more evident in male patients, manifesting as sexual impotence characterized by erectile dysfunction and impaired ejaculation. The effects and impact of nerve injury in female patients, such as inability to experience orgasm and insufficient lubrication, should not be underestimated, however. The extent of autonomic nerve injury that results in definite postoperative impairment of urinary continence and sexual functions is unclear. In general, the autonomic innervation of organs is characterized by a certain functional redundancy, which is anatomically reflected by a bilateral organization of abundantly ramifying intrapelvic nerve plexi and collateral nerve fibres connecting both sides with each other. Thus, it has been suggested that unilateral damage may be compensated by the intact contralateral nervous input, although sexual dysfunction and urinary continence have also been described after partial damage of autonomic nerves (Figures 2.9 and 2.10).

11/02/13 4:31 PM

RECTUM AND ANAL CANAL  43

Figure 2.9. Autonomic and somatic nerves of the pelvis, sagittal section of male pelvis, left view. The hypogastric nerves diverge from the superior hypogastric plexus and extend laterally to the rectum into the inferior hypogastric plexus, which is joined by the pelvic splanchnic nerves. Rectal nerves leave the inferior hypogastric plexus to enter the rectal wall (rectal plexus), while the remaining nerve fibres extend further anteriorly as neurovascular bundles towards the urinary bladder, prostate and seminal vesicles. The neurovascular bundles include the cavernous nerves, which enter the cavernous bodies. The pudendal nerve extends below the levator ani muscle and provides the somatic nerve supply of the striated pelvic floor muscles and the skin of the perineal region and external genital organs. (Reproduced from Schünke, M. et al., Prometheus Atlas of Anatomy, Vols. 1 and 2. Stuttgart, Germany, Thieme Publ. 2007/2009, with permission.)

Figure 2.10. Perirectal spaces, dissected pelvis, cranial view. The rectum is retracted together with the rectal fascia to illustrate the retrorectal space and the presacral space (forceps inserted). Note that the presacral fascia is removed to expose the autonomic nerves (superior hypogastric plexus, hypogastric nerves) embedded within this fascia. From the inferior hypogastric plexus diverge nerve fibres via the rectal pedicles (T-junction) into the rectal wall.

HEBK001-C02_p31-51.indd 43

superior hypogastric plexus hypogastric nerve pelvic splanchnic nerves

neurovascular bundles cavernous nerves

inferior hypogastric plexus rectal plexus pudendal nerve levator ani muscle perineal nerves

dorsal penile nerve

inferior rectal nerves external anal sphincter

superior hypogastric plexus

superior rectal artery

presacral space with presacral vessels

hypogastric nerve

rectal fascia

inferior hypogastric plexus

T-junction

11/02/13 4:32 PM

44  Anatomy of the rectum, anal canal and pelvic floor

Sympathetic nerves

Lumbar sympathetic nerves pass along and in front of the descending aorta and form fairly well-defined periarterial nervous networks, the inferior mesenteric and superior hypogastric plexus. These networks enclose the inferior mesenteric artery up to 5 cm from its origin from the aorta. A high-tie arterial ligation is virtually impossible without damaging, at least in part, adjacent sympathetic nerve fibres. Thus, when disconnection of the rectal blood supply is carried out, a low-tie ligation of the inferior mesenteric artery is generally preferred to avoid impairment of both urinary continence and sexual functions at this initial surgical step. When the superior hypogastric plexus has entered the pelvic cavity in front of the promontorium, it divides into the left and right hypogastric nerve. The hypogastric nerve does not always present as a single clearly defined nerve; often it consists of different nerve fibre bundles arranged in parallel oriented intermingling strings. Both hypogastric nerves are attached to the inner lamella of the presacral fascia corresponding to the parietal pelvic fascia extending in front of the sacral concavity. Due to the intimate relationship between the hypogastric nerves and the presacral fascia, this fascial layer has also been termed the hypogastric sheath or the pre-hypogastric nerve fascia. Gentle traction on the hypogastric nerves will lift this fascial sheath, helping to visualize the course of the nerves down to the inferior hypogastric plexus. Parasympathetic nerves

Parasympathetic nerves derive from the sacral portion of the parasympathetic nervous system located within the sacral spinal cord. Together with thick ventral branches of the sacral spinal nerves (S2–4), they leave throughout the ventral sacral foramina as so-called pelvic splanchnic nerves. Due to one of their main functions, they are also termed erigent nerves or erigent pillars. The pelvic splanchnic nerves pierce the presacral fascia at the left and right sides and follow this fascial plane to join the inferior hypogastric plexus, where they intermingle with their sympathetic counterparts. Inferior hypogastric plexus/pelvic plexus

The inferior hypogastric plexus, also termed the pelvic plexus, is an extensive network of autonomic nerve fibres composed of sympathetic nerves origi-

HEBK001-C02_p31-51.indd 44

nating from the left and right hypogastric nerves and parasympathetic nerves originating from pelvic splanchnic nerves (Figure 2.11). The inferior hypogastric plexus is embedded in connective tissue of the parietal pelvic fascia. At the level of the rectal pedicles, small branches diverge from the main nerve plexus and enter the rectal wall (T-junction). The inferior hypogastric plexus extends further in an anterior direction as a fan-like network radiating cranially as high as the seminal vesicles, distal ureters and vasa deferens and caudally towards the apex of the prostate and the upper aspect of the perineal body (Figure 2.12). The lower part of the inferior hypogastric plexus has been described by Walsh and colleagues as neurovascular bundles extending dorsolaterally along the prostate to approach the apex of the prostate and penetrate the urogenital diaphragm.3 The neurovascular bundles contain the cavernous nerves responsible for erectile functions and terminal nerve fibres innervating the internal urethral sphincter. Thus, special attention should be given to this lower portion of the inferior hypogastric plexus, as these nerve fibres are closely related to the anterior rectal wall, before they continue along the transverse perineal muscles into the penile bulb. The same holds true in female patients in whom the lowermost portion of the inferior hypogastric plexus runs along the lower lateral wall of the vagina. The upper part of the inferior hypogastric plexus extending along the prostate and seminal vesicles is separated from the rectum by the rectoprostatic septum (Denonvilliers’ fascia) (Figure 2.13). As long as the anterior surgical dissection plane is behind the rectoprostatic septum, preservation of these nerves is feasible. Some communicating nerve fibres between the left and right inferior hypogastric plexus pass and cross in front of the rectoprostatic septum. This exchange of nerve fibres may explain why unilateral lesions of the inferior hypogastric plexus may be compensated by the intact contralateral plexus. Despite the presence of autonomic nerve fibres at the midline, this area is still the favoured approach to initiate anterior mobilization of the rectum, because nerve fibre density constantly increases towards the lateral sides of the rectoprostatic septum. Somatic nerves

The autonomic nerve supply of the rectum and anal canal ends at the level of the dentate line. Below this

11/02/13 4:32 PM

RECTUM AND ANAL CANAL  45

superior hypogastric plexus hypogastric nerve

Figure 2.11. Dissected left male hemipelvis with pelvic organs, right anterior view. The levator ani muscle is pulled back to expose the course of the autonomic nerves. The hypogastric nerve diverges from the superior hypogastric plexus and extends laterally to the rectum into the inferior hypogastric plexus, which is joined by the pelvic splanchnic nerves. Rectal nerves leave the inferior hypogastric plexus to enter the rectal wall, while the remaining nerve fibres extend further anteriorly as neurovascular bundles towards the urinary bladder, seminal vesicles, prostate and cavernous bodies.

pelvic splanchnic nerves

vas deferens

urinary bladder seminal vesicle

inferior hypogastric plexus rectal plexus

HEBK001-C02_p31-51.indd 45

cavernous nerves

levator nerves rectum

penile bulb

levator ani muscle

embryologically defined border between the gastrointestinal tract (‘visceral individuum’) and the pelvic floor, including the perineal region and the external anal sphincter (‘somatic individuum’), somatic nerves are in charge for both sensory and mo-

Figure 2.12. Dissected right male hemipelvis with pelvic organs partly reflected, left view. The entire rectum together with the intact mesorectum is mobilized and pulled out of the pelvic cavity, attached only to the perineal body anteriorly. This procedure allows optimal exposure of the right hypogastric nerve, pelvic splanchnic nerves and the extensive nervous network of the inferior hypogastric plexus (green strips). Nerve fibres of the inferior hypogastric plexus extend towards the ureter, seminal vesicle (yellow strip), urinary bladder and prostate, and down to the cavernous bodies (red vessel loop) as neurovascular bundles.

prostate

urethra

tor functions. The lower anal canal comprising the anoderm and perianal region is supplied by perianal branches of the pudendal nerves. In contrast to the autonomically innervated rectum, the anodermal segment is highly sensitive to touch, pressure, pain

ureter urinary bladder seminal vesicle prostate

hypogastric nerve

parietal pelvic fascia inferior hypogastric plexus neurovascular bundle perineal body

pelvic splanchnic nerves

anterior mesorectum

11/02/13 6:38 PM

46  Anatomy of the rectum, anal canal and pelvic floor

urinary bladder

peritoneum

seminal vesicle neurovascular bundle prostate

rectoprostatic septum (Denonvillier´s fasica) anterior rectal wall anal canal

and temperature, due to densely distributed somatosensory nerve endings. The external anal sphincter is innervated by inferior rectal nerve branches provided by the pudendal nerves. It is notable that a subgroup of healthy males and females reveals an asymmetry of external anal sphincter innervation, with a functional preponderance of either the left or right pudendal nerve.

PELVIC FLOOR The pelvic floor is composed of both striated and smooth muscles covered by fasciae and thus corresponds to a rhabdo- and lissomusculofibrous system. The pelvic floor has a twofold function: closure of the pelvic cavity for the support of intrapelvic organs, and controlled opening for micturition and defecation in both genders and parturition in females. The female pelvic floor is larger than the male, displaying wider urogenital openings, but its muscle strength is generally lower and the nerve supply is less developed. The female pelvic floor is therefore capable of allowing vaginal delivery on the one hand, but on the other hand it is more susceptible to functional insufficiency and pelvic organ prolapse. The pelvic floor can be subdivided into a pelvic diaphragm, which is composed mainly of the levator

HEBK001-C02_p31-51.indd 46

Figure 2.13. Dissected right male hemipelvis with pelvic organs, left view. The rectoprostatic septum (Denonvilliers’ fascia) is grasped by the right forceps and retracted dorsally together with the anterior rectal wall. The left forceps grasps the posterior wall of the urinary bladder. Between the rectoprostatic septum and the prostate/ seminal vesicles extend the neurovascular bundles.

ani muscle and a urogenital diaphragm comprising the transverse perineal muscles. In addition to these major striated muscles composed mainly of slowtwitch type I fibres, the pelvic floor is also equipped with smooth musculature. The smooth muscle components are located predominantly along the border of the urogenital and anal hiatus created by the levator ani muscle. Further smooth muscle fibres have also been described extending from the rectal wall to the vagina (rectovaginal muscle) to the urethra (recto-urethral muscle) and to the coccyx along the anococcygeal ligament (rectococcygeal muscle, retractor recti muscle or Treitz muscle).

Pelvic Diaphragm/Levator Ani Muscle The largest of the pelvic floor muscles is the levator ani muscle, and the term ‘levator ani’ is often used to mean the entire pelvic floor. In fact, the levator ani muscle covers most of the lower pelvic aperture and leaves only a midline gap for the urethra and vagina (urogenital hiatus) and the anal canal (anal hiatus) (Figure 2.14). This anterior slit-like opening is partly closed by the urogenital diaphragm at the level of the urogenital hiatus. The levator ani muscle corresponds to a broad, flattened, funnel-shaped muscle that originates from the pelvic wall. The muscular

11/02/13 4:32 PM

PELVIC FLOOR  47 urogenital hiatus internal obturator muscle tendinous arc anal hiatus

puborectal muscle pubococcygeal muscle ileococcygeal muscle coccygeal muscle piriformis muscle

bone, where it is attached by a tendinous plate. The pubococcygeal muscle adapts the shape of a hammock extending above the ileococcygeal muscles. Some muscle fibres decussate to the periurethral musculature and insert into the walls of the vagina (pubovaginal muscle) and rectum and anal canal (puboanal muscle). The puboanal fibres blend with fibres of the longitudinal rectal muscle to form the conjoint longitudinal muscle. Puborectal muscle

Figure 2.14. Pelvic floor, cranial view. The pelvic diaphragm is formed by the levator ani muscle composed of the puborectal muscle, pubococcygeal muscles and ileococcygeal muscles. Most of the levator ani muscle originates from the tendinous arc (‘white line’), which corresponds to a condensed connective tissue line of the obturator fascia. The puborectal sling forms a midline gap for the urethra and vagina (urogenital hiatus) and the anal canal (anal hiatus). The coccygeal muscles extend from the ischiadic spine to the lateral margins of the coccyx following the course of the sacrospinal ligaments. (Reproduced from Schünke, M. et al., Prometheus Atlas of Anatomy, Vols. 1 and 2. Stuttgart, Germany, Thieme Publ. 2007/2009, with permission.)

The puborectal muscle is inseparable from the pubococcygeal muscle at its origin at the pubic bone, but more posteriorly the muscle angles at the anorectal junction to form a sling behind the rectum. Contraction will compress the anal canal by pulling the anorectal junction towards its fixed point (the pubic bone), thereby reducing the anorectal angle. The puborectal sling is intimately fused with the deep part of the external anal sphincter. From the caudal part of the muscle, pre-rectal fibres decussate to insert into the perineal tendinous centre. The puborectal muscle delimits the levator ani gap bordering the urogenital and anal hiatus. Coccygeal muscles

funnel descends towards the anus and fuses with the external anal sphincter. The levator ani muscle is composed of the following parts: Ileococcygeal muscles

The ileococcygeal muscles make up the largest portion of the levator ani muscle and originate from the tendinous arc (arcus tendineus fasciae pelvis) formed by a condensed tissue line (‘white line’) of the obturator fascia. Thus, its muscular insertion sides are not rigid but correspond to a dynamic suspension, allowing a degree of compliance against traction and pressure forces. The flattened muscles attach to the coccyx and the last two sacral vertebrae and fuse posteriorly in a midline raphe. The funnelshaped muscle is generally very thin and displays intramuscular slit-like gaps, particularly in females. Pubococcygeal muscle

In contrast to the non-bony origin of the ileococcygeal muscles, the pubococcygeal muscle runs from the inner surface of the pubic bone to the coccygeal

HEBK001-C02_p31-51.indd 47

The coccygeal muscles are located dorsocranially to the levator ani muscle and extend from the ischial spine to the lateral margins of the coccyx. The muscle tissue is generally poorly developed and follows the course of the sacrospinal ligaments. Lying in the same plane as the levator ani muscle, the coccygeal muscles complete the pelvic diaphragm at its dorsocranial end.

Urogenital Diaphragm The midline gap left by the levator ani muscle for the urogenital hiatus is covered by the urogenital diaphragm. This diaphragm extends between both inferior branches of the pubic bone and corresponds to a musculofibrous plate. In elderly people, and in particular in females after multiple vaginal deliveries, the muscular tissue is almost completely replaced by connective tissue. In the midline, an anterior opening is left for the urethra surrounded by the omega-shaped external urethral sphincter. Posteriorly, the urogenital diaphragm has an opening in females for the vaginal wall.

11/02/13 4:32 PM

48  Anatomy of the rectum, anal canal and pelvic floor

Deep transverse perineal muscle

Most of the urogenital diaphragm is made up of the deep transverse perineal muscle. According to its name, the muscle fibres extend in a transverse direction between the left and right inferior pubic branch. The thin muscular plate is interwoven with connective tissue and widens in a posterior direction, adopting a triangular shape. The muscular portion adjacent to the passage of the vaginal tube is also termed the constrictor vaginae muscle. Superficial transverse perineal muscle

The dorsal aspect of the deep transverse perineal muscle is bordered by the superficial transverse perineal muscle. The slender muscle also extends between the inferior pubic branches near the ischial tubercles and is connected to the perineal tendinous centre (Figure 2.15).

All striated muscles of the pelvic floor are innervated by nerves diverging from the ventral branches of the sacral spinal nerves (S2–4). The somatomotor

external urethral sphincter deep transverse perineal muscle superficial transverse perineal muscle

piriformis muscle

puborectal muscle ileococcygeal muscle

anal hiatus

coccygeal muscle

Figure 2.15. Female pelvic floor, caudal view. The urogenital hiatus of the pelvic diaphragm is covered caudally by the urogenital diaphragm composed of the deep and superficial transverse perineal muscles. The urogenital diaphragm extends between both inferior branches of the pubic bone and corresponds to a thin musculofibrous plate. An anterior opening is left for the urethra surrounded by the external urethral sphincter. Posteriorly, the urogenital diaphragm gives way for the vaginal wall. (Reproduced from Schünke, M. et al., Prometheus Atlas of Anatomy, Vols. 1 and 2. Stuttgart, Germany, Thieme Publ. 2007/2009, with permission.)

HEBK001-C02_p31-51.indd 48

Direct sacral nerves/levator nerves

The direct sacral nerves for the innervation of the muscular pelvic diaphragm originate from the third and fourth sacral nerves. The small-sized nerves, also termed levator nerves, leave the anterior sacral foramina and run from the sacral concavity directly towards the levator ani muscle. They approach the levator ani muscle from above and penetrate into its muscle fibres. When reaching the pelvic floor from above during a TME procedure, care must be taken to preserve these direct nerve branches. Pudendal nerves

Nerve Supply of the Pelvic Floor and Perineal Region

internal obturator muscle

input is provided by direct sacral nerves and the pudendal nerves. The pudendal nerves also carry the somatosensory supply for the perineal region, including the highly innervated anodermal segment of the lower anal canal.

The pudendal nerves (Figure 2.16) arise from the second, third and fourth sacral nerves and initially leave the pelvic cavity together with the sciatic nerves via the greater sciatic foramen. Once they have passed through the infra-piriformis gap, they curve around the sacrospinal ligament at the level of the ischial spine to re-enter the pelvis via the lesser sciatic foramen. The re-entry takes place at the infralevatory level within the ischioanal space. The main pudendal nerve branch is ensheathed by a duplication of the obturator fascia (Alcock’s canal) together with accompanying internal pudendal blood vessels. Along its way towards the pubic bone, multiple branches leave Alcock’s canal, run throughout the fatty tissue of the ischioanal space and reach the entire perineal region with somatomotor and somatosensory terminal nerve fibres. Inferior rectal nerves innervate the external anal sphincter, the lower portion of the levator ani muscles, and the perianal skin and anoderm. Further ventrally, perineal nerves approach the urogenital diaphragm for innervation of the transverse perineal muscles and the external urethral sphincter and for the sensory nerve supply of the perineal region, including the scrotal and labial skin. The dorsal nerve of the penis and clitoris in males and females, respectively, travels above the urogenital diaphragm towards the cavernous bodies.

11/02/13 4:32 PM

PELVIC FLOOR  49 levator ani muscle

sacrotuberal ligament pudendal nerve branches

Figure 2.16. Female pelvic floor with dissection of the left perineal region, dorsocaudal view. The fascia (grasped by forceps) of the internal obturator muscle is incised and opened to release the pudendal nerve fibres from Alcock’s canal and to illustrate its terminal branches (green vessel loop). Inferior rectal nerves run towards the external anal sphincter and perianal region. Perineal nerves extend in an anterior direction to innervate the urogenital diaphragm and the vaginal introitus.

fascia of internal obturator muscle

The terminal branches of the pudendal nerves are located at the outer surface of the puborectal muscle and come into close proximity with the neurovascular bundles at the level of the prostatic apex. At this point, fine branches of the pudendal nerves intermingle with nerve fibres from the neurovascular bundles to innervate the urinary sphincter complex. This region is located near the anterior dissection plane, and caution should be taken to preserve these nerves, in particular if surgical resection is extended down to the pelvic floor in abdominoperineal excision of the rectum.

Ischioanal Space Below the pelvic diaphragm extends the ischioanal space, or fossa, which corresponds to the infradiaphragmatic or infralevatory pelvic compartment extending below the pelvic floor. In a frontal crosssectional plane, the ischioanal space appears as a triangle with its base oriented towards the perineal skin, the lateral side limited by the internal obturator muscle, and the medial side bordered by the funnel-shaped levator ani muscle (Figure 2.17). The apex represents the junction of both muscles along the tendinous arc. At the lateral side-wall extends Alcock’s canal, a duplication of the internal obturator fascia harbouring the main branches of the internal pudendal vessels and pudendal nerves.

HEBK001-C02_p31-51.indd 49

anococcygeal ligament

inferior rectal nerves external anal sphincter

ischiadic tubercle

anus perineal nerves vaginal introitus

The ventral part of the ischioanal fossa is caudally closed by the urogenital diaphragm. The dorsal part surrounds the lower anal canal and widens towards the sacrotuberal ligaments and the gluteus maximus

tendinous arc internal obturator muscle

Alcock´s canal with pudendal vessels & nerve external anal sphincter

ischioanal space superficial perineal fascia

internal anal sphincter

superficial perineal/ perianal space

Figure 2.17. Rectum, anal canal and pelvic floor, frontal section, anterior view. The levator ani muscle originates at the tendinous arc on both sides and forms the funnelshaped pelvic diaphragm extending down to the external anal sphincter. The triangular space delimited by the levator ani muscle, the internal obturator muscle and the superficial perineal fascia corresponds to the ischioanal space. The main pudendal nerve branch and internal pudendal blood vessels are ensheathed by a duplication of the obturator fascia (Alcock’s canal). (Reproduced from Schünke, M. et al., Prometheus Atlas of Anatomy, Vols. 1 and 2. Stuttgart, Germany, Thieme Publ. 2007/2009, with permission.)

11/02/13 6:38 PM

50  Anatomy of the rectum, anal canal and pelvic floor

muscles. The ischioanal fossa is filled with loosely arranged areolar fat (corpus adiposum perinei) and contains the terminal branches of the pudendal nerves and blood vessels. Posteriorly, the ischioanal space is crossed in the midline by the anococcygeal ligament extending from the external anal sphincter to the coccygeal bone. At the perianal region, the ischioanal space is caudally delimited by a thin fascia, the superficial perineal fascia, formed by diverging tendinous endings of the conjoint longitudinal muscle. Below this connective tissue plane extends the perianal space. This space corresponds to the subcutaneous layer underlying the perianal skin and contains small fat lobules separated by rigid connective tissue septa. The ischioanal space is particularly important in abdominoperineal excision for low rectal cancer. In conventional abdominoperineal excision, a classical TME is performed and the ischioanal space is dissected only up to the external anal sphincter by the perineal approach. This procedure often bears the risk of a ‘tissue waist’ at the tumour site and thus is associated with increased rates of positive circumferential resection margins and inadvertent bowel perforation. The modified surgical procedure, also described as ‘cylindrical abdominoperineal excision’, consists of an extended extralevator ­abdominoperineal ­excision (ELAPE). The ELAPE procedure avoids the tissue waist at the anorectal junction by stopping the TME above the pelvic floor and by extending the perineal dissection up to the origin of the levator ani muscle. For this reason, the ischioanal space has to be dissected along the outer surface of the external anal sphincter and of the funnel-shaped levator ani muscle until reaching its origin at the tendinous arc. Subsequently, the resulting specimen will include the entire anal sphincter complex together with major parts of the levator ani muscle completely enveloping the tumour-bearing region.

transverse perineal muscle and bulbospongious muscle in front, the superficial portion of the external anal sphincter at the back, and transversely running fibres of the puborectal muscle from the lateral sides. The perineal body serves as an anterior abutment for the anorectal junction. It is virtually impossible to define a clear surgical dissection plane in this area, because longitudinal smooth muscle bundles deriving from the anterior rectal wall (corrugator ani muscle) and the inferior continuation of the rectogenital septum are intimately connected to the perineal body. Moreover, both the urogenital diaphragm and the external anal sphincter are firmly attached to the perineal body. In males, fibres of the bulbospongious muscle regularly cross the perineal body and fuse with fibres of the superficial portion of the external anal sphincter. The condensation of smooth muscle fibres that connect the anorectal junction to the perineal body and thence to the urethra is also called the recto-urethralis muscle. In abdominoperineal excision, detachment of the anterior aspect of the anorectal specimen appears to be one of the most delicate steps, since this region lacks self-opening planes. Thus, dissection has to be carefully carried out behind the superficial transverse perineal muscle to avoid perforation of the bowel wall and damage of the urethral sphincter complex and its corresponding nerves.

Acknowledgements The author wishes to thank Dr Sigmar Stelzner (Department of General and Visceral Surgery, Dresden-Friedrichstadt General Hospital, Germany) for his valuable contributions regarding the anatomical dissections, and Günter-Rudolf Klaws and Stefanie Gundlach (Institute of Anatomy, Christian Albrechts University of Kiel, Germany) for their assistance in the anatomical dissections and photographing.

Perineal Body/Perineal Tendinous Centre

References

The perineal body corresponds to the centre of the perineal region that serves as a common insertion site for several muscles of the pelvic floor and therefore is also termed the perineal tendinous ­centre. Attached to the perineal body are the superficial

1. Stelzner F. Chirurgie an den viszeralen Abschlusssytemen. Stuttgart, Thieme, 1998. 2. Valleix MM et al. Bulletin Nr. 10, third série. Juin 1836. 3. Walsh PC, Lepor H, Eggleston JC. Radical prostatectomy with preservation of sexual function: anatomical and pathological considerations. Prostate 1983; 4: 473–85.

HEBK001-C02_p31-51.indd 50

11/02/13 4:32 PM

FURTHER READING  51

FURTHER READING Aigner F, Zbar AP, Ludwikowski B, Kreczy A, Kovacs P, Fritsch H. The rectogenital septum: morphology, function, and clinical relevance. Dis Colon Rectum 2004; 47: 131–40. Baader B, Herrmann M. Topography of the pelvic autonomic nervous system and its potential impact on surgical intervention in the pelvis. Clin Anat 2003; 16: 119–30. Blaivas JG, Barbalias GA. Characteristics of neural injury after abdomino-perineal resection. J Urol 1983; 129: 84–7. Clausen N, Wolloscheck T, Konerding MA. How to optimize autonomic nerve preservation in total mesorectal excision: clinical topography and morphology of pelvic nerves and fasciae. World J Surg 2008; 32: 1768–75. Fritsch H, Lienemann A, Brenner E, Ludwikowski B. Clinical anatomy of the pelvic floor. Adv Anat Embryol Cell Biol 2004; 175: 1–64. García-Armengol J, García-Botello S, Martinez-Soriano F, Roig JV, Lledó S. Review of the anatomic concepts in relation to the retrorectal space and endopelvic fascia: Waldeyer’s fascia and the rectosacral fascia. Colorectal Dis 2008; 10: 298–302. Havenga K, DeRuiter MC, Enker WE, Welvaart K. Anatomical basis of autonomic nerve-preserving total mesorectal excision for rectal cancer. Br J Surg 1996; 83: 384–8. Havenga K, Enker WE, McDermott K, Cohen AM, Minsky BD, Guillem J. Male and female sexual and urinary function after total mesorectal excision with autonomic nerve preservation for carcinoma of the rectum. J Am Coll Surg 1996; 182: 495–502. Heald BJ, Moran BJ. Embryology and anatomy of the rectum. Semin Surg Oncol 1998; 15: 66–71. Heald RJ, Smedh RK, Kald A, Sexton R, Moran BJ. Abdominoperineal excision of the rectum: an endangered operation. Dis Colon Rectum 1997; 40: 747–51. Hollabaugh RS Jr, Steiner MS, Sellers KD, Samm BJ, Dmochowski RR. Neuroanatomy of the pelvis: implications for colonic and rectal resection. Dis Colon Rectum 2000; 43: 1390–97. Holm T, Ljung A, Häggmark T, Jurell G, Lagergren J. Extended abdominoperineal resection with gluteus maximus flap reconstruction of the pelvic floor for rectal cancer. Br J Surg 2007; 94: 232–8. Kinugasa Y, Murakami G, Suzuki D, Sugihara K. Histological identification of fascial structures posterolateral to the rectum. Br J Surg 2007; 94: 620–6. Kinugasa Y, Murakami G, Uchimoto K, Takenaka A, Yajima T, Sugihara K. Operating behind Denonvilliers’ fascia for reliable preservation of urogenital autonomic nerves in total mesorectal excision: a histologic study using cadaveric specimens, including a surgical experiment using fresh cadaveric models. Dis Colon Rectum 2006; 49: 1024–32.

HEBK001-C02_p31-51.indd 51

Kinugasa Y, Niikura H, Murakami G, et al. Development of the human hypogastric nerve sheath with special reference to the topohistology between the nerve sheath and other prevertebral fascial structures. Clin Anat 2008; 21: 558–67. Kirkham AP, Mundy AR, Heald RJ, Scholefield JH. Cadaveric dissection for the rectal surgeon. Ann R Coll Surg Engl 2001; 83: 89–95. Kourambas J, Angus DG, Hosking P, Chou ST. A histological study of Denonvilliers’ fascia and its relationship to the neurovascular bundle. Br J Urol 1998; 82: 408–10. Lindsey I, Guy RJ, Warren BF, Mortensen NJ. Anatomy of Denonvilliers’ fascia and pelvic nerves, impotence, and implications for the colorectal surgeon. Br J Surg 2000; 87: 1288–99. Lindsey I, Warren BF, Mortensen NJ. Denonvilliers’ fascia lies anterior to the fascia propria and rectal dissection plane in total mesorectal excision. Dis Colon Rectum 2005; 48: 37–42. Marr R, Birbeck K, Garvican J, et al. The modern abdominoperineal excision: the next challenge after total mesorectal excision. Ann Surg 2005; 242: 74–82. Maurer CA. Urinary and sexual function after total mesorectal excision. Recent Results Cancer Res 2005; 165: 196–204. Nagtegaal ID, van de Velde CJ, Marijnen GC, van Krieken JHJM, Quirke P. Low rectal cancer: a call for a change of approach in abdominoperineal resection. J Clin Oncol 2005; 23: 9257–64. Schünke M, Schulte E, Schumacher U. Prometheus LernAtlas der Anatomie: Allgemeine Anatomie und Bewegungssystem. Stuttgart, Thieme, 2007. Schünke M, Schulte E, Schumacher U. Prometheus LernAtlas der Anatomie: Innere Organe. Stuttgart, Thieme, 2009. Standring S. Gray’s Anatomy: The Anatomical Basis of Clinical Practice, 39th edn. Edinburgh, Churchill Livingstone, 2004. Stelzner S, Holm T, Moran BJ, et al. Deep pelvic anatomy revisited for a description of crucial steps in extralevator abdominoperineal excision for rectal cancer. Dis Colon Rectum 2011; 54: 947–57. Takenaka A, Hara R, Soga H, Murakami G, Fujisawa M. A novel technique for approaching the endopelvic fascia in retropubic radical prostatectomy, based on an anatomical study of fixed and fresh cadavers. BJU Int 2005; 95: 766–71. Uchimoto K, Murakami G, Kinugasa Y, Arakawa T, Matsubara A, Nakajima Y. Rectourethralis muscle and pitfalls of anterior perineal dissection in abdominoperineal resection and intersphincteric resection for rectal cancer. Anat Sci Int 2007; 82: 8–15. Uhlenhuth E, Day EC, Smith RD, Middleton EB. The visceral endopelvic fascia and the hypogastric sheath. Surg Gynecol Obstet 1948; 86: 9–28.

11/02/13 4:32 PM

3 Clinical ultrasound Oliver Shihab and Arcot K. Venkatasubramaniam

Introduction Accurate preoperative staging of rectal cancer influences the choice of surgery and the use of neoadjuvant therapies, with the aim of achieving a clear resection margin. It is also possible that, in selected early tumours, resectional surgery may be avoided with the use of local excision. The endorectal ultrasound (EUS) staging system, its accuracy with respect to other imaging modalities, and the clinical role of EUS in the assessment of rectal cancer are discussed in this chapter.

Imaging technique

but give a poorer depth of penetration.2 In EUS, frequencies of 7.0 MHz or 10 MHz are most commonly used and give satisfactory images.

STAGING Primary Tumour Staging Using EUS, it is possible to clearly visualize five ultrasonic layers of the rectum and surround tissues, which correspond to the five anatomical layers. The EUS appearances alternate between echogenic and echo-poor:3 l l l l

Endorectal ultrasound is the oldest imaging modality for local staging of rectal cancer1 and can be conveniently performed in the office setting by the colorectal surgeon. Endorectal ultrasound typically uses a rigid probe with a rotating ultrasound transducer mounted at the tip. This ultrasound transducer contains crystals that emit sound waves when electrically stimulated, producing images orthogonal to the probe (Figure 3.1). A balloon filled with degassed water is at the tip of the probe, passing the emitted sound waves to the rectum via the water, so preventing image distortion by air. The depth of penetration of these waves is a function of the focal length of the crystals: those that produce higherfrequency waves produce a higher-resolution image

HEBK001-C03_p52-57.indd 52

l

mucosa (echogenic); muscularis mucosae (echo-poor); submucosa (echogenic); muscularis propria (echo-poor); serosa/perirectal fat (echogenic).

The traditional staging system using EUS follows the method of Hildebrandt and Feifel:4 They established the use of the TNM (tumour, node, metastasis)-based system for EUS, with the ‘u’ prefix to show that the staging had been predicted by ultrasound, using the relationship of the hypoechoic tumour to ultrasonic layers, as described above: l l

uT0: tumour confined to the mucosa only. uT1: tumour confined to the mucosa and submucosa.

08/02/13 6:41 PM

STAGING  53

Figure 3.1. Anorectal transducer (courtesy BK medical). Scanning plane for the anorectal transducer.

uT2: tumour invades muscularis propria but is confined to the rectal wall. l uT3: tumour invades perirectal tissues but does not involve adjacent organs. l uT4: tumour invades adjacent organs. l

Accuracy for depth of invasion is high, with studies showing this to be in the region of 74–96 per cent.5–13 In one study, however, after neoadjuvant radiotherapy the T-stage accuracy of EUS fell from 86 per cent in the non-irradiated patient cohort to 47 per cent in the irradiated cohort. It was suggested that irradiation and subsequent rectal wall thickening resulted in poor visualization of the rectal wall layers.14 Another study demonstrated that EUS correctly predicts cases of complete response following neoadjuvant therapy in 63 per cent of patients and correctly predicts T stage in 48 per cent of patients.15 In earlier cancers (SM1-T2), overall accuracy of EUS is particularly high, in the region of 92 per cent.9,16 This compares favourably with magnetic resonance imaging (MRI), which has an accuracy of 67–86 per cent.17–19 The lower accuracy seen with MRI is thought to be due to the poor differentiation of T1/2 cancer from very early T3 cancer, where it can be difficult to distinguish true mesorectal tumour invasion from desmoplastic reactions. Use of a 15 MHz probe (resulting in an increased resolution but decreased focal length) has a high accuracy (86 per cent) in distinguishing the earliest lesions

HEBK001-C03_p52-57.indd 53

of the submucosa (SM1) from those with greater depth of invasion.20 This is of clinical importance, because patients with SM1 or lower disease have an extremely low risk of lymph node metastasis.9,21 Accurate identification of this subgroup could therefore allow for safe local excision of the cancer.

Lymph Node Staging Prediction of lymph node involvement is problematic, because approximately half of all lymph nodes with metastatic spread are less than 5 mm in diameter22 and up to 18 per cent of lymph nodes of 4 mm or less may contain tumour metastases.23 Although there is an association with lymph node size and likelihood of tumour metastasis, there is considerable overlap in size between metastatic and non-metastatic nodes. As such, lymph node involvement lacks sufficient discrimination to be used as a staging criterion. It has been noted, however, that for grossly enlarged mesorectal nodes (.9 mm in short-axis diameter), there is almost invariably involvement by tumour.24 Other criteria used in staging include inhomogeneity, long-axis to short-axis diameter ratio (‘roundness index’), echogenicity, lymph node hilar reflectivity and border contour.23–26 A small number of studies have examined the role of fine-needle aspiration (FNA) cytology in

08/02/13 6:41 PM

54  Clinical ultrasound

improving the accuracy of nodal staging.27,28 Typically this uses a longitudinally oriented ultrasound probe, enabling the biopsy needle to accurately pass into the target area. Although this has proven technically feasible, the role in rectal cancer is yet to be defined. In the study by Harewood and colleagues, there was no significant improvement in accuracy of nodal staging using EUS and FNA, compared with EUS alone.28 Importantly, with FNA there is a theoretical risk of disseminating cancer cells into the surrounding mesorectum. Nodal disease continues to present a problem, regardless of the imaging modality used, and results are still disappointing when compared with the staging of direct tumour invasion. Accuracy for EUS has been quoted as varying from 61 per cent to 83 per cent,6–11 while MRI has an accuracy of 57–85 per cent.17,19 Hilar reflectivity and lymph node inhomogeneity appear to have the greatest discriminatory value for prediction of lymph node involvement.24

The Low Rectum and Anal Canal In the lower third of the rectum (below the origin of the levator muscles and usually within 6 cm of the anal verge), the mesorectum tapers sharply and no longer forms a protective barrier against tumour spread. The anal canal is generally considered to extend from the upper border of puborectalis to the anal verge and is formed by the circular muscles of the internal and external anal sphincters, which form a partially overlapping tube of muscle. At the level of the sphincters the distinction between T2 tumours (invading muscularis propria, which, at this level, is the internal anal sphincter), T3 tumours (through the internal sphincter into the intersphincteric space) and T4 tumours (invading the external sphincter) is very small and difficult to determine with any degree of certainty. The depth of tumour invasion at this level has important implications in optimal preoperative and operative management, particularly the aspects of sphincter salvage or removal. Many centres give neoadjuvant chemoradiotherapy based on the height of tumours, due to the greater rate of margin involvement at surgery. Careful anatomical staging allows refinement of this. The benefits of neoadjuvant therapy have to

HEBK001-C03_p52-57.indd 54

be balanced against the immediate and long-term complications, such as perineal wound breakdown in abdominoperineal excision (APE)29 and well-documented long-term effects, including pelvic fractures, increased risk of second primary tumours and alterations in urinary, bowel and sexual ­function.30 For these reasons, accurate imaging at this height is crucial in optimal management and poses particular challenges in the lowest rectal cancers at the level of the sphincters. Historically, several studies have demonstrated the ability of EUS to define the anal sphincters and outline sphincter defects (typically in patients investigated for faecal incontinence) and to define ­fistulae-in-ano.4,31–34 Furthermore, EUS can provide reliable measurements of the external anal ­sphincter.35 There have been very few studies, however, of the ability of EUS to define the extent of sphincter infiltration by rectal carcinoma. The one study found is limited by its small number of patients (n = 12).36 Endorectal ultrasound does, however, seem promising when compared with pathology of the excised specimen, with 100 per cent sensitivity and specificity for pT2 and pT4 tumours and 86 per cent sensitivity and specificity for pT3 tumours. Furthermore, 5 of the 12 tumours were post-chemoradiotherapy when assessed by EUS, and all of these were staged correctly. There are some reports on the staging accuracy of EUS in anal squamous cell carcinoma, which also has the potential to invade the sphincter complex. As with the rectal cancer studies, published reports on anal cancer and sphincter invasion are few in number and have small study sizes. One study involved 12 patients with anal cancer, all of whom were staged by EUS;37 of these, only 5 patients underwent surgery, allowing EUS staging to be compared with histopathology. Two tumours were classified as uT2a (invading the internal anal sphincter) and three were classified as uT2b (invading the external sphincter); histology was in exact concordance in these cases.

Three-dimensional endorectal ultrasound Three-dimensional ultrasound reconstructs radial and longitudinal images to produce a three-­dimensional image and may provide a more accurate anatomical

08/02/13 6:41 PM

References  55

view of the rectum and surrounding structures. Its use is said to improve both primary tumour and nodal staging.26,38,39 This improved accuracy has also been described in the selection of early tumours for local excision40 and is also said to produce accurate information with regard to the down-staging of tumours following neoadjuvant therapies.41

Limitations of endorectal ultrasound Technical aspects of EUS must also be considered. Problems are similar to those faced with the endorectal coils in MRI before the introduction of pelvic phased-array coils. The size of the probe used is significant, as it has been found that larger probes result in stretching of the anal canal. This distorts the anatomy of the region, so potentially affecting the image.42 This anatomical distortion is also seen with the use of a water-filled balloon, used for acoustic coupling, in the anal canal.26 It is possible to overcome this distortion by using a rigid plastic cone in place of the balloon. A tumour that is stenosing or that has a large polypoid intraluminal component may also provide a physical barrier to accurate staging, which occurs in around 5 per cent of rectal cancers.43 Unlike MRI, where planes can be planned so that they lie orthogonal to the tumour and anal canal, it is possible that EUS can produce images that are oblique to the canal. These oblique images have the effect of exaggerating the apparent stage of the tumour.44 Harewood and colleagues hypothesized that small numbers influenced reported accuracy of EUS.45 They reported an inverse relationship between study size and accuracy of EUS and speculated that this could partly result from publication bias. This may indicate that the accuracy of EUS has been historically overestimated. This is emphasized by the decline in the published accuracy of tumour staging over time.38 The biggest limitation for EUS in the staging of rectal cancer is that it permits only limited evaluation of the structures surrounding the rectal wall, including the mesorectal fascia. The mesorectal fascia represents a crucial landmark in the assessment of rectal cancer, as it is the circumferential resection margin for a total mesorectal excision of the rectum. It has been demonstrated that the involvement

HEBK001-C03_p52-57.indd 55

of the mesorectal fascia is more informative than T stage alone in determining the resectability of a rectal tumour,46,47 thereby influencing the decision for neoadjuvant chemoradiotherapy. The pelvic sidewall lymph nodes are also typically outside the field of view of the EUS probes, and so other imaging modalities will be required to fully stage this region to adequately plan treatment.

CONCLUSION The management of rectal cancer is complex, requiring decisions to be made as to the requirement for neoadjuvant therapies and the ability to perform sphincter-sparing surgery. Endorectal ultrasound is accurate for T staging of tumours and is particularly accurate in early disease, where it may be possible for selected patients to avoid resectional surgery. It also appears to have a promising role in the assessment of tumour relationship to the components of the anal sphincter complex and may develop a complementary role to MRI in the analysis of cancers of this region. The development of three-dimensional EUS may enhance the accuracy of EUS in these roles.

References   1. Klessen C, Rogalla P, Taupitz M. Local staging of rectal cancer: the current role of MRI. Eur Radiol 2007; 17: 379–89.   2. Saclarides TJ. Endorectal ultrasonography. J Surg Oncol 1996; 61: 239–41.   3. Beynon J, Foy DMA, Roe AM, Temple LN, Mortensen NJM. Endoluminal ultrasound in the assessment of local invasion in rectal cancer. Br J Surg 1986; 73: 474–7.   4. Hildebrandt U, Feifel G. Preoperative staging of rectal cancer by intrarectal ultrasound. Dis Colon Rectum 1985; 28: 42–6.   5. Skandarajah AR, Tjandra JJ. Preoperative locoregional imaging in rectal cancer. ANZ J Surg 2006; 76: 497–504.   6. Garcia-Aguilar J, Pollack J, Lee SH, et al. Accuracy of endorectal ultrasonography in preoperative staging of rectal tumors. Dis Colon Rectum 2002; 45: 10–15.   7. Mackay SG, Pager CK, Joseph D, Stewart PJ, Solomon MJ. Assessment of the accuracy of transrectal ultrasonography in anorectal neoplasia. Br J Surg 2003; 90: 346–50.

08/02/13 6:41 PM

56  Clinical ultrasound   8. Nesbakken A, Lovig T, Lunde OC, Nygaard K. Staging of rectal carcinoma with transrectal ultrasonography. Scand J Surg 2003; 92: 125–9.   9. Akasu T, Kondo H, Moriya Y, et al. Endorectal ultrasonography and treatment of early stage rectal cancer. World J Surg 2000; 24: 1061–8. 10. Glaser F, Schlag P, Herfarth C. Endorectal ultrasonography for the assessment of invasion of rectal tumours and lymph node involvement. Br J Surg 1990; 77: 883–7. 11. Orrom WJ, Wong WD, Rothenberger DA, Jensen LL, Goldberg SM. Endorectal ultrasound in the preoperative staging of rectal tumors. A learning experience. Dis Colon Rectum 1990; 33: 654–9. 12. Beynon J. An evaluation of the role of rectal endosonography in rectal cancer. Ann R Coll Surg Engl 1989; 71: 131–9. 13. Hildebrandt U, Feifel G, Schwarz HP, Scherr O. Endorectal ultrasound: instrumentation and clinical aspects. Int J Colorectal Dis 1986; 1: 203–7. 14. Napoleon B, Pujol B, Berger F, Valette PJ, Gerard JP, Souquet JC. Accuracy of endosonography in the staging of rectal cancer treated by radiotherapy. Br J Surg 1991; 78: 785–8. 15. Vanagunas A, Lin DE, Stryker SJ. Accuracy of endoscopic ultrasound for restaging rectal cancer following neoadjuvant chemoradiation therapy. Am J Gastroenterol 2004; 99: 109–12. 16. Glancy DG, Pullyblank AM, Thomas MG. The role of colonoscopic endoanal ultrasound scanning (EUS) in selecting patients suitable for resection by transanal endoscopic microsurgery (TEM). Colorectal Dis 2005; 7: 148–50. 17. Blomqvist L, Holm T, Rubio C, Hindmarsh T. Rectal tumours: MR imaging with endorectal and/or phasedarray coils, and histopathological staging on giant sections – A comparative study. Acta Radiol 1997; 38: 437–44. 18. Poon FW, McDonald A, Anderson JH, et al. Accuracy of thin section magnetic resonance using phasedarray pelvic coil in predicting the T-staging of rectal cancer. Eur J Radiol 2005; 53: 256–62. 19. Brown G, Radcliffe AG, Newcombe RG, Dallimore NS, Bourne MW, Williams GT. Preoperative assessment of prognostic factors in rectal cancer using highresolution magnetic resonance imaging. Br J Surg 2003; 90: 355–64. 20. Harada N, Hamada S, Kubo H, et al. Preoperative evaluation of submucosal invasive colorectal cancer using a 15-MHz ultrasound miniprobe. Endoscopy 2001; 33: 237–40. 21. Kikuchi R, Takano M, Takagi K, et al. Management of early invasive colorectal cancer: risk of recurrence and clinical guidelines. Dis Colon Rectum 1995; 38: 1286–95. 22. Kotanagi H, Fukuoka T, Shibata Y, et al. The size of regional lymph nodes does not correlate with the

HEBK001-C03_p52-57.indd 56

presence or absence of metastasis in lymph nodes in rectal cancer. J Surg Oncol 1993; 54: 252–4. 23. Katsura Y, Yamada K, Ishizawa T, Yoshinaka H, Shimazu H. Endorectal ultrasonography for the assessment of wall invasion and lymph node metastasis in rectal cancer. Dis Colon Rectum 1992; 35: 362–8. 24. Hulsmans FH, Bosma A, Mulder PJ, Reeders JW, Tytgat GN. Perirectal lymph nodes in rectal cancer: in vitro correlation of sonographic parameters and histopathologic findings. Radiology 1992; 184: 553–60. 25. Hildebrandt U, Klein T, Feifel G, Schwarz HP, Koch B, Schmitt RM. Endosonography of pararectal lymph nodes: in vitro and in vivo evaluation. Dis Colon Rectum 1990; 33: 863–8. 26. Giovannini M, Ardizzone S. Anorectal ultrasound for neoplastic and inflammatory lesions. Best Pract Res Clin Gastroenterol 2006; 20: 113–35. 27. Milsom JW, Czyrko C, Hull TL, Strong SA, Fazio VW. Preoperative biopsy of pararectal lymph nodes in rectal cancer using endoluminal ultrasonography. Dis Colon Rectum 1994; 37: 364–8. 28. Harewood GC, Wiersema MJ, Nelson H, et al. A prospective, blinded assessment of the impact of preoperative staging on the management of rectal cancer. Gastroenterology 2002; 123: 24–32. 29. Marijnen CAM, Kapiteijn E, van de Velde CJH, et al. Acute side effects and complications after short-term preoperative radiotherapy combined with total mesorectal excision in primary rectal cancer: report of a multicenter randomized trial. J Clin Oncol 2002; 20: 817–25. 30. Holm T, Singnomklao T, Rutqvist L, Cedermark B. Adjuvant preoperative radiotherapy in patients with rectal carcinoma: adverse effects during long term follow-up of two randomized trials. Cancer 1996; 78: 968–76. 31. Kumar A, Scholefield JH. Endosonography of the anal canal and rectum. World J Surg 2000; 24: 208–15. 32. Deen KI, Kumar D, Williams JG, Olliff J, Keighley MR. Anal sphincter defects: correlation between endoanal ultrasound and surgery. Ann Surg 1993; 218: 201–5. 33. Tjandra JJ, Milsom JW, Stolfi VM, et al. Endoluminal ultrasound defines anatomy of the anal canal and pelvic floor. Dis Colon Rectum 1992; 35: 465–70. 34. Schwartz DA, Wiersema MJ, Dudiak KM, et al. A comparison of endoscopic ultrasound, magnetic resonance imaging, and exam under anesthesia for evaluation of Crohn’s perianal fistulas. Gastroenterology 2001; 121: 1064–72. 35. Frudinger A, Halligan S, Bartram CI, Price AB, Kamm MA, Winter R. Female anal sphincter: age-related differences in asymptomatic volunteers with highfrequency endoanal US. Radiology 2002; 224: 417–23. 36. Maier AG, Kreuzer SH, Herbst F, et al. Transrectal sonography of anal sphincter infiltration in lower rectal carcinoma. Am J Roentgenol 2000; 175: 735–9. 37. Tarantino D, Bernstein M. Endoanal ultrasound in the staging and management of squamous-cell carcinoma of the anal canal. Dis Colon Rectum 2002; 45: 16–22.

08/02/13 6:41 PM

References  57 38. Kim JC, Kim HC, Yu CS, et al. Efficacy of 3-dimensional endorectal ultrasonography compared with conventional ultrasonography and computed tomography in preoperative rectal cancer staging. Am J Surg 2006; 192: 89–97. 39. Santoro GA, Fortling B. The advantages of volume rendering in three-dimensional endosonography of the anorectum. Dis Colon Rectum 2007; 50: 359–68. 40. Santoro GA, Gizzi G, Pellegrini L, Battistella G, Di Falco G. The value of high-resolution three-­ dimensional endorectal ultrasonography in the management of submucosal invasive rectal tumors. Dis Colon Rectum 2009; 52: 1837–43. 41. Murad-Regadas SM, Regadas FS, Rodrigues LV, et al. Role of three-dimensional anorectal ultrasonography in the assessment of rectal cancer after neoadjuvant radiochemotherapy: preliminary results. Surg Endosc 2009; 23: 1286–91. 42. Saranovic D, Barisic G, Krivokapic Z, Masulovic D, Djuric-Stefanovic A. Endoanal ultrasound evaluation of anorectal diseases and disorders: technique,

HEBK001-C03_p52-57.indd 57

indications, results and limitations. Eur J Radiol 2007; 61: 480–9. 43. Manger T, Stroh C. Accuracy of endorectal ultrasonography in the preoperative staging of rectal cancer. Tech Coloproctol 2004; 8: s14–15. 44. Hashimoto BE. Possible invasive rectal wall mass or anal sphincter break. In Sanders RC, Winter TC (eds). Clinical Sonography: A Practical Guide. Philadelphia, PA, Lippincott Williams & Wilkins, 2006: 203. 45. Harewood GC. Assessment of publication bias in the reporting of EUS performance in staging rectal cancer. Am J Gastroenterol 2005; 100: 808–16. 46. Wolberink SV, Beets-Tan RG, Nagtegaal ID, Wiggers T. Preoperative assessment of the circumferential margin in rectal cancer is more informative in treatment planning than the T stage. Techn Coloproctol 2006; 10: 171–6. 47. MERCURY Study Group. Diagnostic accuracy of preoperative magnetic resonance imaging in predicting curative resection of rectal cancer: prospective observational study. BMJ 2006; 333: 779.

08/02/13 6:41 PM

4 Magnetic resonance imaging staging of rectal cancer Peter How and Gina Brown

Introduction Rectal cancer is common, and accurate preoperative staging of tumours using high-resolution magnetic resonance imaging (MRI) is a crucial part of modern multidisciplinary team management. The importance of MRI in staging rectal cancer relates specifically to the ability to delineate tumour extension with respect to surgically relevant landmarks, primarily the mesorectal fascia and sphincter complex. This has significant therapeutic implications regarding the need or otherwise for neoadjuvant therapy, sphincter preservation and plane of surgery that can have a major impact upon the patient’s survival and quality of life. With the development of effective coil systems and high-resolution surface body coils, MRI looks likely to maintain and expand its current position as the imaging modality of choice for planning an effective therapeutic strategy in patients with advanced rectal cancer.

MRI and surgical planning for rectal cancer Undoubtedly, the most important advance within rectal cancer surgery over the past 30 years has been the development of total mesorectal excision (TME). Described by Heald and colleagues in

HEBK001-C04_p58-70.indd 58

1982,1 TME involves the meticulous dissection and complete removal of an intact surgical specimen containing the tumour-bearing rectum surrounded by the mesorectum, the embryologically derived lymphovascular envelope of the human rectum.1,2 The principles of TME are outlined in Chapter 7 and are based on the following principles: The rectum is surrounded by the mesorectum and an areolar tissue, the mesorectal fascia, that forms a surgically distinct avascular plane. l Perimesorectal plane dissection should result in complete removal of the mesorectum contained within an intact fascia. l Careful dissection of the mesorectal fascia allows identification and preservation of the autonomic nerves, important for genitourinary function. l

In addition, anal sphincter preservation by anterior resection is generally favoured where feasible, as it avoids the need for a permanent stoma and is generally considered oncologically superior to sphincter removal at abdominoperineal excision.3 TME facilitates restorative resection. Since the adoption of TME, the mesorectal fascia has come to be synonymous with the circumferential resection margin, and tumours located within 1 mm of this margin are considered to be margin-positive and at high risk of local recurrence. Brown and colleagues demonstrated

08/02/13 6:40 PM

ANATOMICAL LANDMARKS  59

ANATOMICAL LANDMARKS The important anatomical structures to consider in rectal cancer surgery, some or all of which may be visualized on MRI, are: l l l l l l l

Figure 4.1. Depiction of rectal tumour (red line) relative to the mesorectal fascia (blue line) on axial magnetic resonance imaging.

the feasibility and reproducibility of identifying such high-risk tumours on MRI by accurately depicting the proximity of the tumour to the mesorectal fascia (Figure 4.1).4 This is a key determinant for successful treatment of rectal cancer, as it is this relationship that often dictates the use of neoadjuvant therapy and helps determine the optimal surgical plane. In cases where tumour extends to or goes through the mesorectal fascia, the TME or ‘holy plane’ is deemed unsafe; neoadjuvant therapy (preoperative radio- or chemoradiotherapy) or an extended surgical procedure, or a combination of both, may then be required to achieve a clear circumferential resection margin. In a low rectal cancer, an enhanced extralevator approach may be needed. Such radiological guidance is of particular interest with respect to tumours of the lower rectum where the mesorectum tapers as it approaches the pelvic floor. It is here where breaches of the mesorectal fascia and infiltration of the levator are often observed in advanced tumours. This is outlined in detail later. In addition to the mesorectal fascia, the proximity of the tumour to other important pelvic viscera plays an important role in surgical planning. These viscera include the prostate, seminal vesicles, bladder, uterus and vagina. Curative resection may depend on removing parts or all of these adjacent structures in locally advanced tumours. Where tumour extends to or invades these structures, conventional surgery (anterior resection or abdominoperineal excision) will result in cutting through the tumour, and a more radical surgical approach such as pelvic exenteration may be required.

HEBK001-C04_p58-70.indd 59

normal rectal wall; mesorectum and mesorectal fascia; anal sphincter complex; Denonvilliers’ fascia/urogenital septum; presacral fascia; peritoneal reflection; pelvic nerve plexuses.

Normal Rectal Wall The rectal wall, in keeping with the rest of the large bowel, is composed of an inner mucosal layer, concentrically surrounded by submucosa and the muscularis propria, comprising an inner circular and outer longitudinal layer. Interposed between the two muscle layers lies the myenteric plexus contained within a thin layer of connective tissue. The mucosal layer on MRI is represented by a delicate low-signal intensity line, with the submucosal layer beneath this depicted as a thicker higher-signal intensity structure. With high-quality images, the muscularis propria can sometimes be visualized as two distinct layers of circular and longitudinal muscle. The outer muscle layer often has an irregular ridged appearance on account of blood vessels entering the rectal wall. The perirectal fat appears as highsignal-intensity tissue, contrasting well with the low signal of the muscularis propria (Figure 4.2).

Mesorectal Fascia and Mesorectum The mesorectal fascia is best seen on axial images and appears as a low-signal intensity linear structure surrounding the mesorectum (Figure 4.3). Inferiorly, this layer fuses with the endopelvic fascia that lies over the surface of the levator muscles; superiorly, the fascia fuses with the peritoneal reflection anteriorly and parietal fascia posteriorly. The mesorectum on MRI axial images appears as a high-signal intensity (similar to fat) package surrounding the rectum, containing blood vessels and lymphatic tissue. Lymph nodes within the mesorectum appear as high-signal intensity ovoid structures.

08/02/13 6:40 PM

60  MRI staging of rectal cancer

ICM, inner circular muscle; M, mucosa; OLM, outer longitudinal muscle; SM, submucosa. Figure 4.2. Layers of the bowel wall as they appear on magnetic resonance imaging and on histopathology.

Anal Sphincter Complex The anal canal is composed of inner circular smooth muscle forming the internal sphincter, surrounded by skeletal muscle forming the external sphincter. The anal canal is seen as a cylindrical structure that extends from the insertion of the puborectalis sling on to the rectum to the external anal orifice. The lower portion of the levator ani and external anal sphincter are anatomically indistinguishable (see Chapter 2). With high-quality T2 coronal and axial images, it is possible to accurately depict tumour proximity to the sphincter complex and delineate individual components of the sphincter complex

important for continence (i.e. internal and external sphincter, puborectalis, levator plate) (Figure 4.4).

Denonvilliers’ Fascia/Urogenital Septum Denonvilliers’ fascia or the urogenital septum (see Chapter 7) forms a distinctive anterior structure, more prominent in the male. The fascia represents

EAS, external anal sphincter; IAS, internal anal sphincter. Figure 4.3. Mesorectal fascia (blue line) surrounding mesorectum and benign lymph node (arrow).

HEBK001-C04_p58-70.indd 60

Figure 4.4. Coronal image showing relationship of tumour (red) to sphincter complex and its components.

08/02/13 6:41 PM

ANATOMICAL LANDMARKS  61

side-wall. The pelvic side-wall contains a conglomeration of lymph nodes that are related to branches of the internal and external iliac vessels and are therefore in a separate compartment from the mesorectum. The pelvic side-wall nodes are therefore not routinely visualized in rectal cancer surgery unless this compartment is opened up by dissecting through the presacral or parietal fascia (see Chapter 7, p. 103).

Peritoneal Reflection

Figure 4.5. Axial T2-weighted image showing Denonvilliers’ fascia as a low-signal layer.

a barrier to the spread of anterior tumours and separates the rectum and perirectal structures from the urogenital organs. It consists of a fibromuscular structure composed of a number of layers that are fused together and envelope the seminal vesicles. Denonvilliers’ fascia is visible on MRI as a low-signal layer that can be traced up to the peritoneum superiorly (Figure 4.5).

Presacral Fascia The presacral fascia appears on sagittal MRI as a lowsignal intensity linear structure overlying the presacral vessels (Figure 4.6). The presacral fascia lies posterior to the mesorectal fascia. It is a thickened parietal fascia covering the presacral veins and fat. It fuses with and directly covers the muscles and vessels of the pelvic

Figure 4.6. Sagittal magnetic resonance imaging showing the presacral fascia as a low-signal-intensity linear structure.

HEBK001-C04_p58-70.indd 61

On sagittal MRI, the peritoneal reflection appears as a low-signal intensity linear structure that extends over the surface of the bladder and can be traced posteriorly to its point of attachment on to the rectum. On axial section, this point of attachment gives a Vshaped configuration, with the anterior covering of the rectum by the peritoneal reflection widening in the cranial direction. It is important to note that the peritoneal covering of the rectum is not a surgical resection margin, as there is no adjacent structure – hence, tumour extending to within 1 mm of the peritoneum should not be considered as a potentially involved circumferential resection margin, although of course the margin of the specimen removed may well be involved with spread through the peritoneum and a high risk of subsequent peritoneal carcinomatosis (Figure 4.7).

Pelvic Nerve Plexuses Preservation of the autonomic nerve plexuses is crucial if genitourinary dysfunction is to be avoided.

Figure 4.7. Sagittal magnetic resonance imaging demonstrating the peritoneal reflection as a low-signal V-shaped linear structure extending over the bladder to the rectum.

08/02/13 6:41 PM

62  MRI staging of rectal cancer

The superior hypogastric plexus, situated below the bifurcation of the aorta, gives rise to the right and left hypogastric nerves that descend to join the inferior hypogastric plexus. The right and left inferior hypogastric plexus lie in a parasagittal plane on the pelvic side-wall. In males, the inferior hypogastric plexus lies posterolaterally to the seminal vesicles; in females, its anterior half lies against the upper third of the vagina. On MRI, the inferior hypogastric plexus lies inside and medial to the vessels on the pelvic side-wall but outside the mesorectal plane. The inferior hypogastric plexus is a distinctive rectangular fenestrated structure, typically 3–4 cm in anteroposterior length, lying in a parasagittal plane that can often be identified on MRI due to its high-signal intensity (Figure 4.8). By contrast, on axial and coronal oblique imaging, the plexus appears as a linear beaded structure.

STAGING RECTAL TUMOURS Tumour staging is critical for planning of therapy and determining likely prognosis with regard to local recurrence and survival. To develop an effective therapeutic strategy based on the patient’s individual circumstances, discussion of MRI staging in a multidisciplinary team context is key. It has been reported that multidisciplinary team discussion of preoperative MRI with implementation of a selective preoperative strategy significantly reduces R1 and R2 resections.5 For most cancers, tumour extension at diagnosis and treatment determines the likelihood of cure; this is particularly relevant to rectal cancer, where local and systemic recurrence are major issues. Cuthbert Dukes highlighted the importance of extramural spread in predicting local recurrence and survival in patients with rectal cancer.6 The system he described has stood the test of time; the so-called ‘Dukes classification’ focuses on two key aspects – the local extent of the tumour in relationship to the bowel wall, and the presence or absence of lymph nodal involvement. The Dukes system was based on the excised specimen subjected to pathological analysis and subsequent staging. Currently, however, rectal cancer staging is based mostly on the tumour–node–metastasis (TNM) and Union for International Cancer Control (UICC) staging systems, which have largely superseded the Dukes classification.

HEBK001-C04_p58-70.indd 62

Figure 4.8. Inferior hypogastric plexus appears as a high-signal beaded structure on paracoronal (A) and axial (B) imaging.

Like the Dukes system, the TNM system is a pathologically based system; the prefix ‘p’ indicates ‘pathological’ (e.g. pT1 indicates a T1 tumour at pathology). The main advantage of the TNM system is its universal applicability and the ability to predict the likely stage at preoperative imaging and alter the management appropriately. In rectal cancer, a T1 rectal carcinoma is limited to the submucosa and on MRI is depicted by intermediate signal intensity within the mucosa and submucosa, with preservation of a thin layer of submucosa lying deep to the tumour. The MRI appearances of such a tumour should be recorded as ‘mrT1’. A T2 tumour invades the muscularis propria and on MRI corresponds to tumour signal extending into

08/02/13 6:41 PM

STAGING RECTAL TUMOURS  63

the circular muscle, but without extending through the full thickness of muscle. Where the full thickness of the muscle coat is replaced by intermediate signal intensity, this would equate with a T2 or T3 tumour, which can be impossible to differentiate on MRI. A T3 rectal cancer extends beyond the muscularis propria and is depicted on MRI by a broad-based infiltrative margin of intermediate signal intensity, extending into perirectal fat. Clearly, however, T3 cancers are a very heterogeneous group, including tumours of variable prognostic significance related to the depth of extramural spread.7 More recent versions of the TNM staging system have sought to rectify this problem with the introduction of subgroups T3a (, 1 mm beyond muscularis propria), T3b (1–5 mm beyond muscularis propria), T3c (5–15 mm beyond muscularis propria) and T3d (. 15 mm beyond muscularis propria), which have prognostic significance.8 T4 rectal cancers are defined as tumours that involve the peritoneal surface or invades adjacent organs.

Imaging Technique MRI rectal cancer staging is relatively fast and straightforward, because no special patient preparation is required. Some centres advocate the use of an antispasmodic agent such as hyoscine butylbromide to prevent artefacts caused by small bowel peristalsis and to help distend the sigmoid and rectum.14 To effectively plan the initial localization sequences, it is important for the radiologist or radiographer to know the approximate tumour height on clinical assessment. The patient is then positioned on his or her back and a phased-array surface coil is placed on the pelvis, such that the edge of the coil is situated a good distance below the pubic bone. The coil is secured with belts and the patient is advanced head first into the magnetic field. The first series of localization images is the sagittal T2W-FSE, which identifies the primary tumour. The second series consists of large field-of-view axial sections of the whole pelvis, extending from the iliac crest to the pubic symphysis. The sagittal T2-weighted images are used to plan T2-weighted thin-section axial images through the rectal cancer and along the rectum down to the anus, while encompassing perirectal tissues including important pelvic landmarks. The images are achieved using 3 mm sections and a 16 cm field of view.

HEBK001-C04_p58-70.indd 63

The abrupt change in size of the rectal lumen at the anorectal junction limits the usefulness of oblique axial imaging performed proximally. Beyond the anorectal junction, axial imaging cannot demonstrate the rectal wall in its entirety and is unable to reliably determine separation of the rectal wall from the levator plate. This appearance could lead to potential overstaging of tumours. A high-­ resolution coronal imaging sequence, however, can accurately depict the levator, sphincter complex and intersphincteric plane, which has important surgical implications.

mrT Staging It is widely accepted that MRI is the imaging modality that offers the highest soft-tissue contrast and is therefore best suited to T staging rectal cancer. Initial results with MRI were disappointing due to poor spatial resolution achieved by whole-body coil systems. The development of endorectal coils improved resolution with T staging accuracies comparable to endorectal ultrasound, widely regarded as the gold standard for assessing local invasion of an early rectal cancer. As with endorectal ultrasound, however, endorectal MRI imaging is limited by a small field of view that permits adequate evaluation only of early-stage rectal cancer, and not advanced cancers, as visualization of surrounding pelvic anatomy is limited. The other major disadvantage of endoluminal techniques relates to the invasive nature; technical limitations such as stenosis, stricturing, pain, discomfort, bowel wall movement and coil migration may render the acquisition of good-quality images impossible, particularly in advanced tumours. The major breakthrough in MRI rectal cancer staging came with the development of high-­ resolution phased-array surface-coil systems. These surface coils can achieve a very high spatial contrast resolution while maintaining a large field of view, enabling both accurate depiction of the intestinal wall and important surrounding anatomy, including the mesorectal fascia. In recent studies, reported accuracy has varied from 86 per cent to 100 per cent;9–11 such variability largely reflects the difficulties in staging borderline tumours as T2 or T3. It is well recognized that overstaging T2 tumours is often caused by a desmoplastic reaction

08/02/13 6:41 PM

64  MRI staging of rectal cancer

of peritumoural tissue.12 The most clinically relevant feature of MRI, however, relates to its assessment of the mesorectal fascia and thus the potential circumferential resection margin in a patient undergoing a TME. The multicentre MERCURY study revealed MRI to be accurate within 0.5 mm in predicting local tumour spread and specificity for prediction of a clear margin by MRI to be 92 per cent.13

N Staging As with extent of extramural invasion, Dukes was one of the first to identify lymph node involvement as having significant prognostic value, with 5-year survival being related to the number of affected nodes.19 A number of studies have demonstrated that the involvement of four or more nodes significantly worsens survival and highlights the importance of adequate lymph node dissection during surgery and histopathological assessment. Current guidelines from the Royal College of Pathologists recommend that an average of more than 12 nodes should be sampled for all surgical specimens containing cancer. Predicting involvement of the lymph nodes on preoperative imaging has been considered by many to be a crucial but challenging task in rectal cancer staging. Accuracy rates for N staging reported in the literature show wide variation across all recognized imaging techniques. The reported accuracy for MRI ranges from 57 per cent to 85 per cent.20–23 Although it is impossible for any form of preoperative imaging to reliably exclude microscopic nodal involvement, the high-contrast resolution images offered by MRI have enabled the identification of consistent markers for lymph nodal involvement. Brown and colleagues demonstrated that an irregular contour and an inhomogeneous signal are the most reliable MRI criteria for lymph node metastasis when comparing preoperative MRI with histopathological assessment of the resected specimen (Figure 4.9).23 Conversely, normal or reactive nodes are characterized by uniform signal intensity and smooth, well-defined borders. Although it is natural to suspect that enlarged nodes are likely to be malignant, using size criteria alone would result in a high number of false positives. Considerable overlap in size between

HEBK001-C04_p58-70.indd 64

benign, reactive and malignant nodes has been observed; Brown and colleagues23 have reported that a previously purported cut-off size of 5 mm to differentiate involved from uninvolved nodes22 is unfounded. The importance of preoperative N staging is not limited to its prognostic value with regard to systemic recurrence, because lymph nodal involvement has also been linked to an increased local recurrence. The presence of involved lymph nodes close to the mesorectal fascia (circumferential resection margin in a patient with an upper or middle rectal cancer having a TME) has been reported to increase the risk of local recurrence,24 and many units consider suspicious nodes near to the mesorectal fascia to be an indication for neoadjuvant therapy. A publication from 2010 evaluated the importance of nodes near to the mesorectal fascia detected on MRI and correlated whether this resulted in a positive circumferential resection margin. In a retrospective review of 396 patients with rectal cancer, Shihab and colleagues identified 31 patients with suspicious nodes 1 mm or less from the circumferential resection margin. None of these patients had a positive circumferential resection margin owing to nodal involvement.25 This would suggest that with regard to nodes threatening the circumferential resection margin, only those with obvious extracapsular extension will result in a positive margin and would therefore require neoadjuvant therapy.

Pelvic Side-Wall Nodal Disease The use of 16 cm field-of-view thin-section MRI enables the pelvic side-wall compartment and the mesorectal compartment to be assessed. This is of particular value in identifying patients at risk of residual disease, despite a successful TME operation, due to pelvic side-wall nodal disease. It is generally thought that pelvic side-wall disease is a feature of aggressive disease and associated with poor survival.26 Undoubtedly involved nodes are more common in patients with low rectal cancer. In the UK and Europe, pelvic side-wall nodal dissection is seldom performed because of the unacceptable morbidity related to this operation. Currently in the West, a high risk of suspicious pelvic side-wall nodal involvement is generally treated by targeted radiotherapy or chemoradiotherapy.

08/02/13 6:41 PM

STAGING RECTAL TUMOURS  65

Figure 4.9. Irregular contour and inhomogeneous signal are the most reliable magnetic resonance imaging (MRI) criteria for lymph node metastasis on MRI/histopathological comparison.

Vascular Spread Extramural vascular invasion (EMVI) is defined as the presence of malignant cells within blood vessels beyond the muscularis propria in the vicinity of a colorectal tumour. Histological EMVI has long been recognized as an independent predictor of ­local and systemic recurrence and poorer overall survival27–29 and has been reported to occur in up to 52 per cent of all cases of colorectal cancer.30–34 Extramural vascular invasion is therefore considered an adverse prognostic feature and by some is considered an indication for neoadjuvant therapy, despite the circumferential resection margin remaining unthreatened by the primary tumour. Although MRI can detect EMVI, only recently have the radiological characteristics of EMVI been documented adequately. This characterization is important as it is not always possible to determine with absolute certainty whether a structure is vascular. Larger vessels on T2-weighted images may appear black due to signal void, while smaller vessels may be recognized because of tortuosity and branching, but there is a general lack of consistency of radiological features. In response to this, Smith and colleagues defined four criteria by which MRI-­ predicted EMVI should be assessed:35 Pattern of tumour margin: tumour invasion into small veins that radiate outward from the bowel wall produces a nodular border that can be distinguished from desmoplasia, which appears as fine stranding of low-signal intensity.

l

HEBK001-C04_p58-70.indd 65

Location of tumour relative to major vessels: presence of tumour signal within a vascular structure is highly suggestive of EMVI. l Calibre of vessel: as tumour invades along the lumen, the vessel expands. Tumour signal is intermediate (grey), and hence expansion of a low-signal vessel by tumour extension is readily identifiable. l Vessel border: tumour may eventually expand through the vessel wall to give variable results; the border may be smooth, irregular or nodular. l

Based on these criteria, a five-point grading system (0 to 4) for MRI-predicted EMVI has been proposed, with the lowest score (0) corresponding to an absence of suspicious features and the highest score (4) corresponding to the most overt features (Figure 4.10). A subsequent study comparing MRI and histopathological assessment of EMVI has demonstrated sensitivity and specificity of MRI to be 62 per cent and 88 per cent, respectively (Figure 4.11).36 Furthermore, relapse-free survival at 3 years was reported at 35 per cent for patients with an MRIEMVI score of 3–4 compared with 74 per cent for patients with an MRI-EMVI score of 0–2. Amalgamating five separate scores (0, 1, 2, 3, 4) into these two broader groups (0–2, 3–4) achieved a similar prediction of relapse-free survival as histopathological assessment of EMVI (34 per cent versus 74 per cent). MRI-predicted EMVI can therefore help predict disease recurrence and is an important part

08/02/13 6:41 PM

66  MRI staging of rectal cancer

MRIEMVI score

Imaging features

Illustration

0

1

Minimal extramural stranding/nodular extension seen, but not in the vicinity of any vascular structures.

2

Stranding demostrated in the vicinity of extramural vessels, but these vessels are of normal calibre, and there is no definite tumour signal seen within the vessel.

3

Intermediate signal intensity apparent within vessels, although the contour and calibre of these vessels is only slightly expanded

4

Obvious irregular vessel contour or nodular expansion of vessel by definite tumour signal.

Figure 4.10. Five-point grading system for magnetic resonance imaging-predicted extramural vascular invasion.

of preoperative staging. Although EMVI does predict for an increased risk of local recurrence, the real issue is the very high risk of systemic recurrence; in future, patients with MRI-detected EMVI may be treated with neoadjuvant chemotherapy rather than neoadjuvant chemoradiotherapy. It has also been demonstrated, however, that a greater proportion of patients with rectal cancer with MRI-predicted EMVI also have MRI-predicted pelvic side-wall disease. As alluded to earlier, affected PSW nodes are generally associated with disseminated systemic disease and in the West are an indication for neoadjuvant therapy. Lower rectal tumours in particular are noted to have a much

HEBK001-C04_p58-70.indd 66

higher incidence of PSW disease compared with mid- and upper tumours,37,38 thought to be related to differing patterns of lymphatic drainage.

Low Rectal Cancer and an Anatomically Based Staging System Low rectal cancer, as defined by a tumour with its lower edge at or below the origin of the levators at the pelvic side-wall, usually corresponds to a tumour within 6 cm from the anal verge. At this level, the mesorectum tapers out and the safety net of the mesorectum is lost. Thus, tumours at this

08/02/13 6:41 PM

STAGING RECTAL TUMOURS  67

Figure 4.11. Magnetic resonance imaging (MRI) and histopathological assessments of MRI.

level are more likely to involve the sphincters and neighbouring structures, and surgical treatment is more problematic. For example, extramural extension located anteriorly is likely to infiltrate the prostate and may require clearance of the anterior compartment; similar extramural extension located posteriorly may be amenable to surgical cure by conventional TME, even for low tumours. Indeed, although preoperative staging techniques are limited in differentiating between T2 and early T3 tumours, this has minimal therapeutic implications in the mid- and upper rectum, given that both have a good prognosis with optimal curative surgery and neoadjuvant therapy is not generally required. The low rectum is more complex because of anatomical constraints and the lack of mesorectum as a natural barrier and safety net. Unsurprisingly, interest is growing in anatomically based staging classifications that can accurately predict whether a tumour-free circumferential resection margin is likely to be achieved or not, regardless of the T stage. This would enable discrimination between patients with minimal mesorectal infiltration not requiring neoadjuvant therapy and those with threatened or infiltrated mesorectal fascia who would benefit from neoadjuvant therapy. Furthermore, such a system may help identify patients in whom a modified surgical approach is required such as the extralevator abdominoperineal excision (Figure 4.12) (see Chapter 8, p. 125).

HEBK001-C04_p58-70.indd 67

The complexity of low rectal cancer has stimulated increasing interest in assessment and management of cancers in the distal rectum. At the time of writing, the English National Cancer Action Team has funded the Low Rectal Cancer National Development Programme aiming to improve cancer outcomes and quality of life in patients with low rectal cancer. Some of the key objectives are to enhance patient-centred multidisciplinary team decision-making to improve patient outcome and quality of life by optimal preoperative therapy and surgical technique. Shihab and colleagues describe a two-plane approach to low rectal cancer – the ­mesorectal and extralevator planes.15,16 In response to the apparent oncological inferiority of the traditional abdominoperineal excision compared with low anterior resection in the management of low rectal cancer, the authors proposed a modified approach to abdominoperineal excision, as outlined in Chapter 8. Histopathological studies have demonstrated a reduction in circumferential resection margin positivity and specimen perforation rates with this approach,17 but possibly more perineal wound complications,18 highlighting the importance of appropriate patient selection. Shihab and colleagues advise that for low tumours located above the sphincters and more than 1 mm from the mesorectal fascia, the mesorectal plane may be appropriate and adequate.15,16 Similarly, for tumours at the anal canal, the mesorectal plane is

08/02/13 6:41 PM

68  MRI staging of rectal cancer

Resection lines Level at which resection lines meet

Figure 4.12. The extralevator abdominoperineal excision, whereby perineal dissection is performed outside the levators, which are divided close to their origin at the pelvic side-wall, with removal en bloc of the rectum and anus.

safe for those that show partial invasion or less of the internal sphincter – that is, the intersphincteric plane is not involved or threatened (Figure 4.13). Although the extralevator abdominoperineal excision remains a topic for debate, the authors seek to define the excision planes for this operation and, in so doing, offer appropriate selection criteria; the main indication is for low rectal tumours above the sphincters that are predicted to invade the levator

muscle or invade the full thickness of the internal anal sphincter, or beyond this, at the level of the sphincters. Given that MRI can often accurately depict component parts of the sphincter complex and lower rectum described earlier, a novel staging system for low rectal cancer has been proposed.15,16 In keeping with the aforementioned surgical planes, this system divides the lower rectum into suprasphincteric

Mesorectal fascia

Levator insertion

1

Figure 4.13. A novel magnetic resonance imaging-based staging system for low rectal cancer based on key anatomical landmarks of the sphincter complex.

HEBK001-C04_p58-70.indd 68

Intersphinteric plane 2

Puborectalis and external sphincter

Internal sphincter

08/02/13 6:41 PM

References  69

and sphincteric compartments, with clear selection criteria as to which plane of surgery may be appropriate, safe and effective. Retrospective evaluation of this staging system has yielded promising results, and prospective larger-scale studies are under way. It is hoped that this novel MRI-based staging system will help reduce circumferential resection margin positivity for low rectal cancers by the selective use of neoadjuvant therapy and planned surgery.

CONCLUSION The development of high-resolution MRI surface-coil systems has resulted in accurate preoperative staging, which has marked a new era in rectal cancer management. Since the widespread adoption of TME, more attention is being devoted to staging systems that can accurately predict circumferential resection margin status and reliably determine selective use of neoadjuvant therapy or suggest an alternative surgical approach. Despite its limitations in T staging, MRI is currently the only imaging modality capable of depicting the topographic relationship between tumour and the proposed surgical resection margin. Low rectal cancer in particular is more complex and with worse outcomes than middle and upper rectal cancer and is currently the focus of ongoing attempts to improve all aspects of management, including improved MRI staging based on key anatomical landmarks. Optimal staging, enhanced decision-making regarding preoperative therapy and optimal surgery will lead to improved oncological outcomes.

References   1. Heald RJ, Husband EM, Ryall RD. The mesorectum in rectal cancer surgery: the clue to pelvic recurrence? Br J Surg 1982; 69: 613–6.   2. Wibe A, Møller B, Norstein J, et al. A national strategic change in treatment policy for rectal cancer implementation of total mesorectal excision as routine treatment in Norway: a national audit. Dis Colon Rectum 2002; 45: 857–66.   3. Heald RJ, Smedh RK, Kald A, Sexton R, Moran BJ. Abdominoperineal excision of the rectum: an endangered operation. Norman Nigro Lectureship. Dis Colon Rectum 1997; 40: 747–51.   4. Brown G, Kirkham A, Williams GT, et al. High-resolution MRI of the anatomy important in total mesorectal

HEBK001-C04_p58-70.indd 69

excision of the rectum. AJR Am J Roentgenol 2004; 182: 431–9.   5. Burton S, Brown G, Daniels IR, et al. MRI directed multidisciplinary team preoperative treatment strategy: the way to eliminate positive circumferential margins? Br J Cancer 2006; 94: 351–7.   6. Dukes CE. The classification of cancer of the rectum. J Pathol Bacteriol 1932; 35: 323.   7. Cawthorn SJ, Parums DV, Gibbs NM, et al. Extent of mesorectal spread and involvement of lateral resection margin as prognostic factors after surgery for rectal cancer. Lancet 1990; 335: 1055–9.   8. Pollheimer MJ, Kornprat P, Pollheimer VS, et al. Clinical significance of pT sub-classification in surgical pathology of colorectal cancer. Int J Colorectal Dis 2010; 25: 187–96.   9. Beets-Tan RG, Beets GL, Vliegen RF, et al. Accuracy of magnetic resonance imaging in prediction of tumour-free resection margin in rectal cancer surgery. Lancet 2001; 357: 497–504. 10. Gagliardi G, Bayar S, Smith R, Salem RR. Preoperative staging of rectal cancer using magnetic resonance imaging with external phase-arrayed coils. Arch Surg 2002; 137: 447–51. 11. Blomqvist L, Holm T, Rubio C, Hindmarsh T. Rectal tumours: MR imaging with endorectal and/or phased-array coils, and histopathological staging on giant sections – a comparative study. Acta Radiol 1997; 38: 437–44. 12. Laghi A, Ferri M, Catalono C, et al. Local staging of rectal cancer with MRI using a phased array body coil. Abdom Imaging 2002; 27: 425–31. 13. MERCURY Study Group. Extramural depth of tumor invasion at thin–section MR in patients with rectal cancer: results of the MERCURY study. Radiology 2007; 243: 132–9. 14. Klessen C, Rogalla P, Taupitz M. Local staging of rectal cancer: the current role of MRI. Eur Radiol 2007; 17: 379–89. 15. Shihab OC, Heald RJ, Rullier E, et al. Defining the surgical planes on MRI improves surgery for cancer of the low rectum. Lancet Oncol 2009; 10: 1207–11. 16. Shihab OC, Moran BJ, Heald RJ, Quirke P, Brown G. MRI staging of low rectal cancer. Eur Radiol 2009; 19: 643–50. 17. West NP, Finan PJ, Anderin C, et al. Evidence of the oncologic superiority of cylindrical abdominoperineal excision for low rectal cancer. J Clin Oncol 2008; 26: 3517–22. 18. West NP, Anderin C, Smith KJ, et al. Multicentre experience with extralevator abdominoperineal excision for low rectal cancer. Br J Surg 2010; 97: 588–99. 19. Dukes CE, Bussey HJ. The spread of cancer and its effect on prognosis. Cancer 1958; 12: 309–20. 20. Kim NK, Kim MJ, Yun SH, Sohn SK, Min JS. Comparative study of transrectal ultrasonography, pelvic computerized tomography, and magnetic resonance

08/02/13 6:41 PM

70  MRI staging of rectal cancer imaging in preoperative staging of rectal cancer. Dis Colon Rectum 1999; 42: 770–5. 21. Chan TW, Kressel HY, Milestone B, et al. Rectal carcinoma: staging at MR imaging with endorectal surface coil – work in progress. Radiology 1991; 181: 461–7. 22. Schnall MD, Furth EE, Rosato EF, Kressel HY. Rectal tumor stage: correlation of endorectal MR imaging and pathologic findings. Radiology 1994; 190: 709–14. 23. Brown G, Richards CJ, Bourne MW, et al. Morphologic predictors of lymph node status in rectal cancer with use of high-spatial-resolution MR imaging with histopathologic comparison. Radiology 2003; 227: 371–7. 24. Adam IJ, Mohamdee MO, Martin IG, et al. Role of circumferential margin involvement in the local recurrence of rectal cancer. Lancet 1994; 344: 707–11. 25. Shihab OC, Quirke P, Heald RJ, Moran BJ, Brown G. Magnetic resonance imaging-detected lymph nodes close to the mesorectal fascia are rarely a cause of margin involvement after total mesorectal excision. Br J Surg 2010; 97: 1431–6. 26. Yano H, Moran BJ. The incidence of lateral pelvic sidewall nodal involvement in low rectal cancer may be similar in Japan and the West. Br J Surg 2008; 95: 33–49. 27. Bokey EL, Chapuis PH, Dent OF, et al. Factors affecting survival after excision of the rectum for cancer: a multivariate analysis. Dis Colon Rectum 1997; 40: 3–10. 28. Horn A, Dahl O, Morild I. Venous and neural invasion as predictors of recurrence in rectal adenocarcinoma. Dis Colon Rectum 1991; 34: 798–804. 29. Harrison JC, Dean PJ, el-Zeky F, Van der Zwaag R. From Dukes through Jass: pathological prognostic indicators in rectal cancer. Hum Pathol 1994; 25: 498–505.

HEBK001-C04_p58-70.indd 70

30. Talbot IC, Ritchie S, Leighton MH, et al. Spread of rectal cancer within veins: histologic features and clinical significance. Am J Surg 1981; 141: 15–17. 31. Dukes C, Bussey HJ. Venous spread in rectal cancer. Proc R Soc Med 1941; 34: 571–4. 32. Knudsen JB, Nilsson T, Sprechler M, Johansen A, Christensen N. Venous and nerve invasion as prognostic factors in postoperative survival of patients with resectable cancer of the rectum. Dis Colon Rectum 1983; 26: 613–7. 33. Krasna MJ, Flancbaum L, Cody RP, Schneibaum S, Ben Ari G. Vascular and neural invasion in colorectal carcinoma: incidence and prognostic significance. Cancer 1988; 61:1018–23. 34. Sunderland D. The significance of vein invasion by cancer of the rectum and sigmoid: a microscopic study of 210 cases. Cancer 1949; 2: 429–37. 35. Smith NJ, Shihab O, Arnaout A, Swift RI, Brown G. MRI for detection of extramural vascular invasion in rectal cancer. Am J Roent 2008; 191: 1–6. 36. Smith NJ, Barbachano Y, Norman AR, Swift RI, Abulafi M, Brown G. Prognostic significance of magnetic resonance imaging-detected extramural vascular invasion in rectal cancer. Br J Surg 2008; 95: 229–36. 37. Takahashi T, Ueno M, Azekura K, Ohta H. Lateral node dissection and total mesorectal excision for rectal cancer. Dis Colon Rectum 2000; 43(10 suppl.): S59–68. 38. Ueno M, Oya M, Azekura K, Yamaguchi T, Muto T. Incidence and prognostic significance of lateral lymph node metastasis in patients with advanced low rectal cancer. Br J Surg 2005; 92: 756–63.

08/02/13 6:41 PM

5 Radiological staging for systemic disease Gina Brown and Chris Hunter

Introduction The early accurate identification of metastatic disease is becoming increasingly important in rectal cancer, in both the preoperative and follow-up settings. A proportion of patients undergoing resection of metastatic disease now achieve long-term cure. 1 It is now well established that patients undergoing surgery for resectable liver metastases achieve a 5-year survival in the region or in excess of 40 per cent, compared with almost no survivors at 5 years for untreated patients.1–4 The proportion of patients in whom long-term cure is achieved following metastasectomy has increased over time, with more recent studies demonstrating 5-year survival approaching 50 per cent in some patients. 5 This is due in part to more accurate preoperative imaging allowing better patient selection.6,7 The early detection of metastatic disease increases the likelihood of identifying resectable disease. In the follow-up setting, it has been shown that the early detection of metastatic disease is of benefit in addition to informing patients of prognosis.8 Intensive follow-up regimes that include computed tomography (CT) identify local or distant recurrence 8.5 months earlier than those using clinical follow-up and carcinoembryonic

HEBK001-C05_p71-86.indd 71

antigen (CEA) alone, and 5-year survival is improved from 9 per cent to 13 per cent.9 The patient group in which metastasectomy may be indicated also continues to expand. Improvements in surgical technique have resulted in low operative morbidity and mortality associated with metastasectomy, and metastasectomy is increasingly offered to patients over 70 years of age. Even after metastasectomy, continued surveillance for further metastatic disease is of benefit, as repeat metastasectomy may improve survival.10,11 In a proportion of patients, even with optimal surveillance, metastatic disease will be inoperable at the time of diagnosis. Patients with unresectable metastatic disease treated with palliative chemotherapy have better survival at 6 months and 12 months, however, compared with patients treated with supportive care alone.12 Patients with unresectable metastatic disease may also be selected for palliative rather than aggressive surgical treatment. Surgical palliation is an effective treatment that alleviates symptoms, increases survival over non-­surgical supportive care, and may require only a short hospitalization.13 In this chapter we discuss the common mechanisms and sites of metastasis in rectal cancer, and the modalities available to identify metastases to these sites.

08/02/13 5:59 PM

72  Radiological staging for systemic disease

Mechanisms of spread of colorectal metastases Lymphatic Spread Lymph nodes are the most common site of metastasis in rectal cancer, with lymph node metastases occurring in 51.9 per cent of 1732 patients undergoing resection of their primary rectal cancer in the Dukes series.14 The incidence of lymph node metastases increases with increasing histological grade of the primary tumour, and with more locally advanced tumours. Lymph node metastases usually spread progressively along the lymphatic chain from lymph node to lymph node. In rectal cancer, lymphatic spread is initially to mesorectal lymph nodes adjacent to the primary tumour. As metastases move up the lymph node chain, they may follow the inferior or middle rectal arteries to pelvic side-wall lymph nodes, particularly in low rectal cancer (usually within 6 cm from the anal verge), and then the internal iliac lymph node chain. The main lymphatic drainage is to nodes along the superior rectal artery and thence to nodes around the inferior mesenteric artery.15 If metastases block the usual lymphatic drainage, then retrograde spread of metastases through the lymphatics may occur. This usually occurs only in advanced tumours with multiple lymph node metastases and is one of the mechanisms for spread to pelvic side-wall and inguinal nodes.16 Spread to inguinal lymph nodes in rectal cancer is rare (occurring in approximately 2 per cent of patients) and usually occurs only in low rectal cancers invading the anal canal.

Haematogenous Spread Extramural vascular invasion has been known to be a predisposing factor for the development of distant metastases for some time.17–20 It is probable that haematogenous spread is the main route for visceral metastases. It is therefore unsurprising that the most common site for colorectal metastases is the liver, as the venous drainage of the colon and rectum is primarily via the portal vein. In a population-based study of 1325 patients, 18.8 per cent of patients had liver metastases occurring within 6 months of diagnosis of colorectal

HEBK001-C05_p71-86.indd 72

cancer, and a further 10.5 per cent of patients developed metachronous metastases within 3 years.21 As well as being the most common site of visceral metastases in colorectal cancer, the liver is the only site of visceral metastases in 30–40 per cent of patients. Other, less common sites of distant metastases in colorectal cancer are the lung (11 per cent of patients have synchronous metastases and 5.8 per cent of patients develop metachronous metastases within 5 years),22 ovary (2–8 per cent of women within 5 years),23,24 bone (5 per cent) and brain (2 per cent).25 Other sites of colorectal metastases are rare but include the spleen, kidneys, pancreas, adrenals, breast, thyroid and skin. It appears, however, that unusual sites of metastasis are becoming more common with the increased use of systemic agents. There also appears to be a relationship to the number of systemic therapies received.26

Transcoelomic Spread The third mechanism of metastasis encountered in colorectal cancer is spread of tumour cells across the peritoneal cavity. This usually occurs when tumour has invaded through the peritoneum (T4 disease). This may be apparent on preoperative imaging, macroscopically at the time of operation, or on histopathological assessment postoperatively. Occasionally, this form of spread may result from peritoneal spillage of tumour cells at the time of surgical resection. Peritoneal deposits have a predilection for certain sites within the abdominal cavity; these include the superior and inferior paracolic recesses, the rectovesical pouch (pouch of Douglas), the under surface of the diaphragm and the transverse mesocolon.27–29

SITES OF METASTASES IN COLORECTAL CANCER Liver Metastases As mentioned previously, the liver is the most common site for visceral colorectal metastases. Figure 5.1 shows a contrast-enhanced CT demonstrating liver metastases. Over the period from 1984

08/02/13 5:59 PM

SITES OF METASTASES IN COLORECTAL CANCER   73

Figure 5.1. Contrast-enhanced computed tomography demonstrating liver metastases (arrowheads).

to 2005, 5-year overall survival for patients undergoing resection of hepatic metastases improved from 23 per cent to 46 per cent, despite patients undergoing operations with more severe disease.30 Improved survival following treatment for colorectal liver metastases has been attributed to a number of different factors, including improved techniques for anatomical resection,1,2,4 increased use of intraoperative ultrasound,31,32 reduction in perioperative morbidity and mortality,3 second resections for recurrent hepatic metastases,33 and the use of chemotherapy.34–37 Recent developments in interventional radiology have also had a significant impact on the treatment of colorectal liver metastases. Radiofrequency ablation (RFA) now has a significant role. Although there is an absence of evidence from randomized controlled trials, due in part to the fact that the two attempted randomized controlled trials of RFA closed early due to poor recruitment, there is a growing body of evidence from clinical series and non-randomized comparative studies that supports the use of RFA. Although resection of colorectal liver metastases remains the treatment of choice in suitable patients, RFA is a useful additional treatment modality for metastases of less than 3–4 cm in size, with local recurrence rates of 6.7–17 per cent.38–40 Mean and 5-year survival with combined chemotherapy and RFA are better than with either modality alone, and median survival

HEBK001-C05_p71-86.indd 73

of 24–59 months can be achieved in unresectable colorectal liver metastases.41 This compares favourably with median overall survival of 16–24 months with chemotherapy alone. Figure 5.2 shows a hepatic metastasis following RFA. Portal vein embolization may be used to induce liver hypertrophy and allow extended hepatectomy in patients who would otherwise have insufficient functional reserve.42 A number of different studies have identified risk factors that predict outcome after resection of liver metastases. One of the largest studies included more than 1000 patients and identified the following adverse prognostic factors: size of metastasis .5 cm, potential involvement of the resection margin, more than one liver metastasis, poor prognosis primary, synchronous primary tumour and liver metastases within 12 months, and extrahepatic metastatic disease. Overall 5-year survival was 60 per cent with none of these risk factors, falling to less than 20 per cent with four or more risk factors.43 It is therefore very important to have accurate preoperative imaging before treatment for liver metastases for careful patient selection. The aims of imaging for patients undergoing resection of liver metastases are: to accurately demonstrate anatomical distribution of liver metastases and segmental sparing; l to exclude widespread micro-metastatic disease within the liver; l to exclude extra-hepatic metastases; l to discriminate between benign and malignant liver lesions; l to stratify patients in terms of prognosis following metastasectomy. l

When assessing patients preoperatively for occult intra- or extra-hepatic malignancies, it is worth bearing in mind that the risk is strongly related to the number and distribution of hepatic metastases; both occult intra-hepatic metastases and extrahepatic metastases are rare with solitary unilobar metastases but much more common with multiple bilobar metastases.44

Lung Metastases The lungs are the second most common site of metastasis in rectal cancer. If resection can be

08/02/13 5:59 PM

74  Radiological staging for systemic disease

Figure 5.2. Liver magnetic resonance imaging demonstrating two metastases (in left lobe, black arrow; in right lobe, white arrow. (A) Before treatment; (B) following chemotherapy; (C) following resection of larger metastasis; (D) following radiofrequency ablation of smaller metastases.

undertaken safely, then excellent outcomes can be achieved, with 5-year survival of 55.4–71 per cent.45,46 Patients undergoing repeat thoracotomy for recurrent pulmonary metastases may also achieve good outcomes, but the presence of hilar or mediastinal lymph node involvement or extra-pulmonary metastases is associated with a poor survival rate.11 As with liver metastases, it is therefore important to identify lesions early, when curative resection may still be achieved, and to exclude poor prognosis features such as extra-pulmonary metastases and hilar or mediastinal lymph node involvement. Lung metastases usually present initially as small nodules, which are often less than 5 mm in diameter. Multidetector CT allows early detection and is superior to any other modality in the detection of these small lesions.47 Figure 5.3 illustrates pulmonary metastases evident on thoracic CT. Although

HEBK001-C05_p71-86.indd 74

this results in the identification of more benign lesions, a baseline CT and serial examinations allow indeterminate lesions to be interpreted more confidently during follow-up.

Ovarian Metastases Ovarian metastases occur in 6–8 per cent of women with colorectal cancer, making the ovaries the third most common site of visceral metastases in women. Radiologically, colorectal ovarian metastases may be easily mistaken for primary mucinous adenocarcinoma of the ovary.48 It is therefore important that the reporting radiologist is aware of any current or previous diagnosis of colorectal cancer when interpreting radiological imaging. Colorectal ovarian metastases may appear as large oval or lobulated,

08/02/13 5:59 PM

IMAGING MODALITIES AVAILABLE FOR STAGING SYSTEMIC DISEASE  75

Figure 5.3. Contrast-enhanced computed tomography demonstrating lung metastasis (arrowhead).

therefore usually occur in the presence of T4b disease, where the serosa has been breached by tumour. The pattern of peritoneal spread is related to tumour grade, with poorly differentiated tumours tending to produce diffuse seeding and well-differentiated tumours being more likely to produce solitary deposits. Peritoneal metastases are the most common cause of ureteric obstruction due to colorectal cancer and should be considered when unexplained ureteric obstruction is identified in patients with colorectal cancer. Although peritoneal metastases are usually managed palliatively, very carefully selected patients may have a good outcome if treated surgically,51,52 and so patients should be discussed in a multidisciplinary setting if peritoneal disease is identified. Figure 5.5 demonstrates peritoneal metastases on CT.

cystic or solid ovarian masses.49,50 Figure 5.4 shows a typical ovarian metastasis on CT.

IMAGING MODALITIES AVAILABLE FOR STAGING SYSTEMIC DISEASE

Peritoneal metastases occur through spread of tumour cells across the peritoneal cavity. They

The list of imaging modalities available for the systemic staging of rectal cancers continues to increase. No single imaging modality is currently optimal for identifying all distant metastases, however, so a multimodality approach is necessary.

Figure 5.4. Unenhanced computed tomography demonstrating bilateral ovarian metastases (arrowheads).

Figure 5.5. Contrast-enhanced computed tomography demonstrating peritoneal metastases (arrowheads).

Peritoneal Metastases

HEBK001-C05_p71-86.indd 75

08/02/13 5:59 PM

76  Radiological staging for systemic disease

Computed Tomography Computed tomography of the thorax, abdomen and pelvis is the most common imaging technique used for identifying distant metastases in rectal cancer, in both primary staging and follow-up. It is widely available, rapid and relatively cheap. Despite the fact that colorectal cancer metastasizes to the liver via the portal vein, colorectal liver metastases derive their blood supply from the hepatic artery.53 This fact can be exploited in the identification of liver metastases: liver CT performed during peak hepatic enhancement will identify most colorectal liver metastases due to their relative hypovascularity in this phase of enhancement. A contrast-enhanced CT demonstrating hepatic metastases is shown in Figure 5.1. This feature can be exploited further using CT portography. In this technique, the superior mesenteric artery is selectively catheterized, and CT scanning of the liver is performed approximately 60 s after the administration of contrast material via the catheter. This technique increases the sensitivity and specificity of CT in the identification of colorectal metastases54 and has been used in the preoperative assessment of patients being considered for hepatic metastasectomy to delineate the disease extent. The technique is invasive, however, and so has not been used in the routine primary staging or follow-up of patients with rectal cancer. The steady improvements in CT technology have also increased its accuracy in identifying liver metastases. Improved sensitivity and specificity in detecting colorectal liver metastases have been attributed to the use of helical CT with 5-mm ­collimation.55 Using this technique in the preoperative assessment of patients undergoing resection of liver metastases, 94 per cent of patients selected for resection by CT were resectable at operation, and 4-year survival was 58 per cent. Multidetector CT represents further technological improvement and is accurate in detecting lesions over 10 mm in size, with 1 mm isotropic voxels giving improved spatial resolution and the ability to reconstruct images in any plane.56 This has largely removed the need for CT portography, even in the preoperative assessment for metastasectomy. The advent of multidetector CT has also allowed imaging of the whole liver in around 10 s. Multidetector CT allows assessment of the liver during

HEBK001-C05_p71-86.indd 76

multiple phases of enhancement in the same sitting, by repeating images through the liver at different time delays after the injection of contrast. There is little evidence, however, that arterial phase imaging is of significant benefit in colorectal metastases.57,58 Colorectal liver metastases typically appear as hypointense lesions with a rim of enhancement. Occasionally they may contain central necrosis with a low-density ‘cystic’ centre. Liver cysts are the most common benign lesions identified. These are well-defined lesions with no rim of enhancement, no internal structure and very low attenuation.59 Although it is often relatively easy to characterize larger lesions, small liver lesions (, 10 mm) may be more challenging. In these circumstances, interval scanning or review of serial imaging is very important. A 3 month interval CT scan improves specificity from 90 per cent to 99 per cent.60 A meta-analysis from 2010 found that the perpatient sensitivity of CT for detecting liver metastases was 83.6 per cent.61 Computed tomography is therefore a good baseline screening investigation for distant metastases in patients with rectal cancer, in both the preoperative and follow-up settings. It may also identify patients with unresectable metastatic disease who do not require further imaging. In this study, however, the sensitivity of magnetic resonance imaging (MRI) and 18F-fluorodeoxyglucose positron-emission tomography (18F-FDGPET) were both higher than that of CT. Computed tomography also has limited ability to identify small-volume peritoneal and liver surface disease. Computed tomography is largely limited to size criteria for the identification of distant nodal metastases, which has limited sensitivity. Therefore, when planning metastasectomy, or if there is a high index of suspicion of metastatic disease and a normal CT, other imaging modalities should also be employed.

18F-Fluorodeoxyglucose Positron-emission Tomography 18F-Fluorodeoxyglucose positron-emission tomography is a form of functional imaging that uses an 18F labelled tracer, which is taken up by the glucose receptor but not metabolized. The tracer accumulates in cells with increased glucose metabolism and upon decay releases high-energy photons, which can be detected by dedicated receptors. As

08/02/13 5:59 PM

IMAGING MODALITIES AVAILABLE FOR STAGING SYSTEMIC DISEASE  77

increased glucose metabolism is exhibited by many tumour cells, this technique can be used to identify primary and metastatic disease in rectal cancer. There has been much interest in the use of 18F-FDGPET to identify metastatic disease in colorectal cancer. In 2005, a meta-analysis of 61 eligible articles suggested that on a per-patient basis, 18F-FDG-PET was more sensitive than CT or MRI in detecting liver metastases, with a sensitivity of 94.6 per cent, although on a per-lesion basis superparamagnetic iron oxide (SPIO)-enhanced MRI was more sensitive.62 Similar results were found in a further meta-analysis in 2010, where the sensitivity of 18F-FDG-PET for detecting colorectal liver metastases on a per-patient basis was 94.1 per cent and significantly higher than CT or MRI;61 again, on a per-lesion basis, there was no significant difference between MRI and 18F-FDG-PET, and MRI was significantly better than CT in detecting lesions under 1 cm in size. A number of studies have also demonstrated 18F-FDG-PET to be accurate in the identification of extra-hepatic disease. In a meta-analysis, PET/ CT had a sensitivity of 91.2 per cent and a specificity of 95.4 per cent in identifying extra-hepatic metastases.63 As a result of this high sensitivity for metastatic disease in rectal cancer, 18F-FDG-PET has been used in the preoperative assessment of patients being considered for resection of hepatic metastases. A number of studies have shown that 18F-FDG-PET significantly alters management in patients being considered for metastasectomy, predominantly by identifying additional extra-hepatic metastases in 10–21 per cent of patients,64,65 and so reduces the rate of negative laparotomies due to unexpected metastatic disease at the time of surgery by approximately 9 per cent.63 The identification of lung metastases may also allow successful resection of both lung and liver metastases. 18F-Fluorodeoxyglucose positron-emission tomography has also been used to identify pelvic recurrence in rectal cancer. Following surgery, scarring and fibrosis can be difficult to distinguish from recurrence on anatomical imaging, and this problem can be compounded by post-radiotherapy changes. A metaanalysis suggested that the sensitivity and specificity of 18F-FDG-PET in detecting pelvic recurrence in rectal cancer are both 94 per cent.66 18F-Fluorodeoxyglucose positron-emission tomography is also used widely to help identify metastatic

HEBK001-C05_p71-86.indd 77

disease in the presence of normal or equivocal anatomical imaging and endoscopy and a high index of suspicion of recurrent rectal cancer, for example in the presence of rising CEA. In these circumstances, Flamen and colleagues found that 18F-FDGPET had a per-lesion sensitivity of 75 per cent,67 although it should be noted that in this study 38 per cent of patients had equivocal findings on conventional imaging and 70 per cent had not undergone chest CT. Despite the apparent superiority of systemic staging incorporating 18F-FDG-PET, very few studies have looked at the impact of 18F-FDG-PET on the primary staging of rectal cancer. Heriot and colleagues assessed this issue in 46 patients with advanced primary rectal cancer and found that management was changed in 17 per cent of patients,68 although again it should be noted that patients did not undergo routine preoperative thoracic CT. There are a number of potential limitations in 18F-FDG-PET staging of rectal cancer. Establishing the precise anatomical location of increased FDG uptake may be difficult, and because of this it is not always possible to identify the anatomical location of liver metastases using 18F-FDG-PET. Physiological or benign increased FDG uptake may create false positives, such as granulomatous and inflammatory processes in the lung. Following radiotherapy, granulation tissue, fibroblast and macrophage activity may give false positives. This is a particular problem in the first 6 months following radiotherapy. The spatial resolution of 18F-FDG-PET is limited to around 7 mm, which means that lesions need to be about 1 cm or larger in size to be identified. This particularly limits the ability of 18F-FDG-PET to identify small-volume peritoneal and liver surface disease. Some authors have suggested that the incidence of extra-hepatic metastases in patients with a low clinical risk score being considered for hepatic metastasectomy is so low that 18F-FDG-PET is of limited value. Indeed, in this group of patients, the use of 18F-FDG-PET may incorrectly upstage patients and prevent them from undergoing potentially curative resection.69 To reduce the risk of false positives and false negatives in 18F-FDG-PET imaging, interpretation with correct anatomical information from crosssectional imaging is very important. For this reason, combined 18F-FDG-PET/CT has superseded 18F-FDG-PET in many situations.

08/02/13 5:59 PM

78  Radiological staging for systemic disease

Combined 18F-Fluorodeoxyglucose Positron-Emission Tomography and Computed Tomography Combining CT and 18F-FDG-PET in the same scanner allows attenuation correction for the PET portion of the imaging, but also allows fusion of the anatomical component of the CT scanning with the functional element of the 18F-FDG-PET. The risk of misregistration of PET and CT images is reduced (but not eliminated) by combining both types of imaging in the same machine at the same time. It is hoped that this combined functional and anatomical imaging will be more accurate in the identification of rectal cancer metastases than either imaging modality alone.70 Niekel and colleagues noted in a meta-analysis that there were not enough studies of 18F-FDGPET/CT in the detection of liver metastases to comment on its accuracy.61 In a single series of 467 patients, however, Orlacchio and colleagues have suggested that the combined 18F-FDG-PET/ CT modality is more accurate than either modality alone in identifying colorectal liver metastases, with reported per-patient sensitivity and specificity of 98 per cent.71 An 18F-FDG-PET/CT scan demonstrating liver metastases is illustrated in Figure 5.6, outlining lymph node metastases in Figure 5.7, and bilateral ovarian metastases in Figure 5.8.

Figure 5.6. Combined 18F-fluorodeoxyglucose positronemission tomography and computed tomography demonstrating liver metastases (arrowheads).

HEBK001-C05_p71-86.indd 78

Figure 5.7. Combined 18F-fluorodeoxyglucose positronemission tomography and computed tomography demonstrating internal iliac lymph node metastasis (arrowhead).

Davey and colleagues assessed the impact of 18F-FDG-PET/CT on the management of 83 patients with primary rectal cancer and found that management was altered in 12 per cent.72

Figure 5.8. Combined 18F-fluorodeoxyglucose positronemission tomography (FDG) and computed tomography demonstrating bilateral ovarian metastases (white arrowheads). Physiological FDG uptake is seen in the right ureter (black arrowhead).

08/02/13 5:59 PM

IMAGING MODALITIES AVAILABLE FOR STAGING SYSTEMIC DISEASE  79

Pelvic Magnetic Resonance Imaging As discussed in Chapter 4, pelvic MRI is used predominantly for the local staging of rectal cancer and is very accurate for this purpose. In some instances, however, pelvic MRI may also demonstrate systemic disease in rectal cancer. Ovarian metastases may be evident on pelvic MRI; they most commonly take the form of solid adnexal masses with intratumoural cysts, although they may also be predominantly solid or predominantly cystic lesions.50 Figure 5.9 shows a typical ovarian metastasis on MRI. Lymph node metastases may also be identified on pelvic MRI. Lymph node metastases occurring within the mesorectal envelope will usually be removed at the time of optimal total mesorectal excision surgery and so can be considered with the local staging. Distant lymph node metastases may also be identified along the internal iliac chain or less commonly in the inguinal lymph node groups, however. Although overlap between benign and malignant lymph nodes makes size criteria unreliable, morphological changes such as an irregular border or inhomogeneous signal intensity have a high positive predictive value, and lymph nodes demonstrating these characteristics should be treated as metastatic disease.73 Figure 5.10 illustrates a typical lymph node metastasis identified on MRI.

Figure 5.9. Magnetic resonance imaging demonstrating bilateral ovarian metastases (arrowheads).

HEBK001-C05_p71-86.indd 79

Figure 5.10. Magnetic resonance imaging demonstrating internal iliac lymph node metastasis (arrowhead).

Peritoneal disease may also be detected and on T2-weighted MRI has an intermediate signal intensity, which appears relatively hyperintense compared with the low-signal peritoneum. Figure 5.11 illustrates peritoneal disease on T2-weighted MRI. If peritoneal disease is suspected, fat-suppressed T1-weighted imaging with gadolinium enhancement may be of benefit in highlighting mesenteric deposits.74

Figure 5.11. Magnetic resonance imaging demonstrating peritoneal metastases (arrowhead).

08/02/13 5:59 PM

80  Radiological staging for systemic disease

Hepatic Magnetic Resonance Imaging Dedicated liver MRI is a highly accurate modality for identifying colorectal metastases to the liver. A metaanalysis of CT, MRI and PET/CT found that on a perlesion basis, for lesions over 1 cm in size, MRI was the most sensitive modality for identifying colorectal liver metastases.62 This study also found that gadoliniumenhanced and SPIO-enhanced MRI was significantly more accurate than unenhanced MRI. A more recent meta-analysis, which included only prospective studies, found that on a per-lesion basis, CT, MRI and 18F-FDG-PET had comparable sensitivity of 74.4 per cent, 80.3 per cent and 81.4 per cent. The sensitivity of MRI had increased since 2004, however, and in studies since 2004 the sensitivity of MRI on a per-lesion basis is 84.9 per cent; MRI is also the most accurate modality in identifying lesions less than 10 mm in diameter.61 A paper comparing liver specific contrast-enhanced MRI and 18F-FDG-PET/CT suggested that on a perpatient basis PET/CT and MRI had equally high sensitivity and specificity at 98 per cent and 100 per cent, respectively, but on a per-lesion basis for lesions less than 1 cm MRI was more sensitive.75 The mainstay of liver MRI is the T1-weighted gradient echo sequence. This allows imaging of the whole liver in a single breath-hold, reducing motion artefact. T1-weighted gradient echo sequences can also be acquired as a volume image and reconstructed in any plane, allowing accurate anatomical localization of lesions. Figure 5.12 illustrates hepatic metastases on a T1-weighted MRI.

Additional sequences are useful to help with lesion characterization and thus help to determine whether lesions represent colorectal liver metastases. In-phase and out-of-phase T1-weighted imaging may help assess areas of focal fatty infiltration, which may appear as apparent ‘perfusion defects’ on CT and mimic colorectal liver metastases. T2-weighted images, which may be acquired as either respiratorytriggered fast spin echo (FSE) sequences or breathhold single-shot FSE ­sequences, further help with the characterization of focal liver lesions. A number of different contrast agents have been used to increase the sensitivity and specificity of liver MRI. Dynamic gadolinium contrast enhancement can be obtained with images in arterial, portal venous, equilibrium and delayed phases. This is superior to dual-phase contrast-enhanced spiral CT, especially in lesion characterization.76 Superparamagnetic iron oxide may also increase sensitivity and specificity of liver MRI in detecting colorectal liver metastases. Superparamagnetic iron oxide is taken up by Kupffer cells, rendering the substance of liver dark on T1 gradient echo sequences. Liver metastases without functioning Kupffer cells therefore appear relatively hyperintense. Other lesions including cysts and cavernous haemangiomas may also be relatively hyperintense, and so images must be interpreted with T2-weighted images to avoid false positives.77 The American Food and Drug Administration has not approved SPIO contrast agents, however, and in 2011 all previously available SPIO preparations were withdrawn from the market.

Figure 5.12. Comparison of gadoxetic acid-enhanced magnetic resonance imaging (MRI) and unenhanced T1-weighted MRI. Gadoxetic acid MRI (left) shows liver metastases (black arrowhead) not seen on unenhanced T1-weighted MRI. Second metastasis (white arrowhead) is seen on both gadoxetic acid-enhanced and unenhanced MRI.

HEBK001-C05_p71-86.indd 80

08/02/13 5:59 PM

IMAGING MODALITIES AVAILABLE FOR STAGING SYSTEMIC DISEASE  81

Figure 5.13. Comparison of gadoxetic acid-enhanced magnetic resonance imaging (MRI) and contrast-enhanced computed tomography. Gadoxetic acid MRI (left) shows liver metastasis (arrowhead) not seen on contrast-enhanced CT.

Liver-specific gadolinium-based contrast agents such as gadoxetic acid and gadobenate dimeglumin are taken up by functioning hepatocytes. This causes increased signal in the liver parenchyma on T1-weighted images. Metastatic lesions remain hypointense. These liver-specific contrast agents increase the sensitivity of liver MRI for colorectal liver metastases, particularly for small lesions.78 Figure 5.13 illustrates colorectal liver metastases on gadoxetic acid-enhanced MRI. Diffusion-weighted imaging (DWI) is an MRI technique that identifies areas within tissues with restricted diffusion. As many tumours contain tightly packed cells with a poorly organized extracellular matrix, they frequently have restricted diffusion of water molecules compared with normal tissues. This

has been exploited in hepatic MRI to help identify and characterize metastatic lesions within the liver. Although evidence is still accumulating for this technique, a number of studies have suggested that the addition of DWI sequences to standard liver MRI protocols can increase the sensitivity.79,80 Figure 5.14 illustrates colorectal liver metastases on DWI.

Transabdominal Ultrasound The primary use of ultrasound in the systemic staging of rectal cancer has been to evaluate the liver for the presence of metastases. Liver metastases may lack contrast with the surrounding normal tissue, however, and therefore may be difficult to identify

Figure 5.14. Diffusion-weighted imaging (DWI) and corresponding gadolinium-enhanced magnetic resonance imaging. Metastasis (white arrowheads) demonstrates restricted diffusion of DWI.

HEBK001-C05_p71-86.indd 81

08/02/13 5:59 PM

82  Radiological staging for systemic disease

on ultrasound. A meta-analysis found that the sensitivity of ultrasound on a per-patient basis for identifying liver metastases was 55 per cent, compared with 72 per cent for CT, 75 per cent for MRI and 90 per cent for 18F-FDG-PET.81 Micro-bubble contrast agents have been used to improve the ability of ultrasound to identify liver metastases and detect around 17 per cent more lesions than conventional B-mode non-contrast-enhanced ultrasound.82 Even when using micro-bubble contrast agents, however, ultrasound is still less sensitive than CT in detecting liver metastases.83 It also does not offer the opportunity to identify metastases at other sites. Ultrasound is cheap and widely available. Therefore, ultrasound may still have a role as a screening or monitoring technique in some institutions and occasionally helps to clarify the nature of a hepatic abnormality. Computed tomography, PET and MRI are superior imaging modalities where these facilities are available, however.

Intraoperative Ultrasound Intraoperative ultrasound is used in many institutions for the intraoperative assessment of patients undergoing resection of hepatic metastases. In this setting, ultrasound is helpful; it alters surgery in 20–44 per cent of patients and demonstrates anatomy relative to the metastases to help plan resection.84,85 It is ­particularly useful in the intraoperative setting to detect small lesions for resection or radiofrequency ablation that may otherwise have been missed.86

Laparoscopy In certain circumstances, diagnostic laparoscopy with cytology or histology may be useful to assess the presence of metastatic disease if non-invasive investigations are equivocal. The steady improvement in non-invasive cross-sectional and functional imaging will probably continue to reduce the situations where laparoscopy is necessary, however.

Chest X-ray Chest radiography is less sensitive and specific in detecting colorectal liver metastases than thoracic

HEBK001-C05_p71-86.indd 82

CT. In a study of 1049 patients who underwent chest X-ray for the purposes of preoperative colorectal cancer staging, only 0.86 per cent of patients had lung metastases correctly identified.87 Chest radiography should therefore be reserved for situations where thoracic CT is not available.

Assessing response to chemotherapy Aside from identifying metastatic disease, imaging also has an important role to play in the assessment of response to treatment and in restaging disease following chemotherapy. The accuracy of 18F-FDG-PET in assessing metastatic disease appears to be reduced in the period immediately following chemotherapy. In a study of 138 patients investigating the accuracy of PET in identifying colorectal liver metastases, Glazer and colleagues found that PET had a negative predictive value of only 13.3 per cent when performed within 4 weeks of chemotherapy.88 This is probably due to the metabolic inhibition of chemotherapeutic agents preventing metastatic lesions from taking up FDG, despite harbouring viable tumour cells. Other studies have supported this finding, reporting a per-lesion sensitivity of PET of only 47–62 per cent,89,90 and a per-patient sensitivity of 63 per cent,91 after chemotherapy. The accuracy of anatomical imaging performed following chemotherapy appears to be comparable to the accuracy in the absence of chemotherapy. Superparamagnetic iron oxideenhanced MRI has a per-lesion sensitivity of 92 per cent following chemotherapy,89 and non-contrast MRI has a per-lesion sensitivity of 80 per cent.92 Contrast-enhanced CT has a per-lesion sensitivity of 65–76 per cent.90–92 The response of colorectal cancer metastases to treatment is therefore best assessed using anatomical imaging on either serial CT or MRI scans. An objective assessment of response can be obtained by repeated measurements of identified lesions; the Response Evaluation Criteria In Solid Tumors (RECIST) guidelines provide a structure for consistent assessment.93 Restaging of hepatic metastases following neoadjuvant chemotherapy before resection is best performed using MRI with a liver-specific contrast agent.

08/02/13 5:59 PM

References  83

Future developments Incremental improvements in both MRI and CT technology have steadily increased the speed and spatial resolution of these imaging modalities, and it is likely that similar advances will see the spatial and perhaps contrast resolution of anatomical imaging continue to increase. Functional imaging, and in particular fused functional and anatomical imaging, has undergone rapid development. Alternative radiotracers such as 39-deoxy-39-[18F]fluorothymidine (FLT) are being investigated, and it is possible that more specific tracers will improve the specificity of PET in the future. It is also likely that the spatial resolution of PET will continue to increase, and this may further increase the sensitivity of PET in the future. Technological hurdles have been overcome to allow PET and MRI scanning to be incorporated in the same machine as PET/MRI fusion. Ongoing research is under way to determine whether this offers an advantage over current PET/CT.

CONCLUSION The ideal modality for systemic staging of rectal cancer would be cheap, be widely available, have a high sensitivity for identifying rectal metastases in all sites, and have a high specificity. Unfortunately, no modality currently fulfils all these criteria. It is therefore necessary to use a multimodality approach and to adapt the staging strategy to the patient’s situation to try to maximize the benefit from available imaging resources. Computed tomography of the chest, abdomen and pelvis is a good baseline investigation for the primary systemic staging of rectal cancer. It identifies metastatic disease with good sensitivity in the most commonly affected sites and should be performed in all patients where this facility is available. Review of serial CT images increases the sensitivity and specificity of subsequent CT for identifying recurrent and metastatic disease in the follow-up period. Magnetic resonance imaging should be used to clarify the nature of indeterminate liver lesions identified on baseline or follow-up CT. Liver MRI should also be performed in patients being considered for resection of liver metastases, as this probably

HEBK001-C05_p71-86.indd 83

provides the most accurate per-lesion preoperative assessment of liver disease. 18F-Fluorodeoxyglucose positron-emission tomography (18F-FDG-PET), or 18F-FDG-PET/CT where available, should be used as a second-line investigation for patients with a high index of clinical suspicion for metastatic disease (such as rising CEA) and normal first-line investigations. It may also be of benefit in assessing pelvic recurrence and in excluding extra-pulmonary and extra-hepatic disease in patients being considered for metastasectomy. Further studies are required to define the role of modalities other than CT in the primary systemic staging of rectal cancer. It is possible that more intensive primary staging may be of benefit in rectal cancer, particularly in patients who are defined as high risk on MRI staging of the primary tumour. In the future, functional imaging may help in selection of patients most likely to benefit from neoadjuvant therapy before surgery and in determining the likelihood and presence of a complete pathological response and thus selecting patients in whom surgery can be deferred and occasionally avoided.

References   1. Fong Y, Cohen AM, Fortner JG, et al. Liver resection for colorectal metastases. J Clin Oncol 1997; 15: 938–46.   2. Jaeck D, Bachellier P, Guiguet M, et al. Long-term survival following resection of colorectal hepatic metastases. Br J Surg 1997; 84: 977–80.   3. Choti MA, Sitzmann JV, Tiburi MF, et al. Trends in long-term survival following liver resection for hepatic colorectal metastases. Ann Surg 2002; 235: 759–66.   4. Geoghegan JG, Scheele J. Treatment of colorectal liver metastases. Br J Surg 1999; 86: 158–69.   5. De Jong MC, Pulitano C, Ribero D, et al. Rates and patterns of recurrence following curative intent surgery for colorectal liver metastasis: an international multiinstitutional analysis of 1669 patients. Ann Surg 2009; 250: 440–8.   6. Wei AC, Greig PD, Grant D, Taylor B, Langer B, Gallinger S. Survival after hepatic resection for colorectal metastases: a 10-year experience. Ann Surg Oncol 2006; 13: 668–76.   7. Elias D, Blot F, el Otmany A, et al. Curative treatment of peritoneal carcinomatosis arising from colorectal cancer by complete resection and intraperitoneal chemotherapy. Cancer 2001; 92: 71–6.   8. Bleeker WA, Mulder NH, Hermans J, Otter R, Plukker JT. Value and cost of follow–up after adjuvant treatment

08/02/13 5:59 PM

84  Radiological staging for systemic disease

  9.

10.

11.

12.

13.

14. 15. 16.

17. 18.

19.

20.

21.

22.

23.

24.

25.

of patients with Dukes’ C colonic cancer. Br J Surg 2001; 88: 101–6. Renehan AG, Egger M, Saunders MP, O’Dwyer ST. Impact on survival of intensive follow up after curative resection for colorectal cancer: systematic review and meta-analysis of randomised trials. BMJ 2002; 324: 813. Shaw IM, Rees M, Welsh FK, Bygrave S, John TG. Repeat hepatic resection for recurrent colorectal liver metastases is associated with favourable long-term survival. Br J Surg 2006; 93: 457–64. Ogata Y, Matono K, Hayashi A, et al. Repeat pulmonary resection for isolated recurrent lung metastases yields results comparable to those after first pulmonary resection in colorectal cancer. World J Surg 2005; 29: 363–8. Simmonds PC. Palliative chemotherapy for advanced colorectal cancer: systematic review and meta-analysis. BMJ 2000; 321: 531–5. Cummins ER, Vick KD, Poole GV. Incurable colorectal carcinoma: the role of surgical palliation. Am Surg 2004; 70: 433–7. Dukes CE, Bussey HJ. The spread of rectal cancer and its effect on prognosis. Br J Cancer 1958; 12: 309–20. Grinnell RS. Lymphatic metastases of carcinoma of the colon and rectum. Ann Surg 1950; 131: 494–506. Grinnell RS. Lymphatic block with atypical and retrograde lymphatic metastasis and spread in carcinoma of the colon and rectum. Ann Surg 1966; 163: 272–80. Dukes C, Bussey HJ. Venous spread in rectal cancer. Proc R Soc Med 1941; 34: 571–3. Sunderland D. The significance of vein invasion by cancer of the rectum and sigmoid: a microscopic study of 210 cases. Cancer 1949; 2: 429–37. Talbot IC, Ritchie S, Leighton MH, Hughes AO, Bussey HJ, Morson BC. The clinical significance of invasion of veins in cancer of the rectum. Br J Surg 1980; 67: 439–42. Talbot IC, Ritchie S, Leighton MH, Hughes AO, Bussey HJ, Morson BC. Spread of rectal cancer within veins: histologic features and clinical significance. Am J Surg 1981; 141: 15–7. Leporrier J, Maurel J, Chiche L, Bara S, Segol P, Launoy G. A population-based study of the incidence, management and prognosis of hepatic metastases from colorectal cancer. Br J Surg 2006; 93: 465–74. Mitry E, Guiu B, Cosconea S, et al. Epidemiology, management and prognosis of colorectal cancer with lung metastases: a 30-year population-based study. Gut 2010; 59: 1383–8. Birnkrant A, Sampson J, Sugarbaker PH. Ovarian metastasis from colorectal cancer. Dis Colon Rectum 1986; 29: 767–71. Segelman J, Floter-Radestad A, Hellborg H, Sjovall A, Martling A. Epidemiology and prognosis of ovarian metastases in colorectal cancer. Br J Surg 2010; 97: 1704–9. Mongan JP, Fadul CE, Cole BF, et al. Brain metastases from colorectal cancer: risk factors, incidence, and the

HEBK001-C05_p71-86.indd 84

possible role of chemokines. Clin Colorectal Cancer 2009; 8: 100–5. 26. Sundermeyer ML, Meropol NJ, Rogatko A, Wang H, Cohen SJ. Changing patterns of bone and brain metastases in patients with colorectal cancer. Clin Colorectal Cancer 2005; 5: 108–13. 27. Demeo JH, Fulcher AS, Austin RF Jr. Anatomic CT demonstration of the peritoneal spaces, ligaments, and mesenteries: normal and pathologic processes. Radiographics 1995; 15: 755–70. 28. Coakley FV, Hricak H. Imaging of peritoneal and mesenteric disease: key concepts for the clinical radiologist. Clin Radiol 1999; 54: 563–74. 29. Meyers MA, Oliphant M, Berne AS, Feldberg MA. The peritoneal ligaments and mesenteries: pathways of intraabdominal spread of disease. Radiology 1987; 163: 593–604. 30. Andres A, Majno PE, Morel P, et al. Improved longterm outcome of surgery for advanced colorectal liver metastases: reasons and implications for management on the basis of a severity score. Ann Surg Oncol 2008; 15: 134–43. 31. Foroutani A, Garland AM, Berber E, et al. Laparoscopic ultrasound vs triphasic computed tomography for detecting liver tumors. Arch Surg 2000; 135: 933–8. 32. Cervone A, Sardi A, Conaway GL. Intraoperative ultrasound (IOUS) is essential in the management of metastatic colorectal liver lesions. Am Surg 2000; 66: 611–5. 33. Nordlinger B, Wind P. Repeat resections of primary hepatic malignancies. Cancer Treat Res 1994; 69: 53–6. 34. Clavien PA, Selzner N, Morse M, Selzner M, Paulson E. Downstaging of hepatocellular carcinoma and liver metastases from colorectal cancer by selective intraarterial chemotherapy. Surgery 2002; 131: 433–42. 35. Shankar A, Leonard P, Renaut AJ, et al. Neoadjuvant therapy improves resectability rates for colorectal liver metastases. Ann R Coll Surg Engl 2001; 83: 85–8. 36. Adam R, Avisar E, Ariche A, et al. Five-year survival following hepatic resection after neoadjuvant therapy for nonresectable colorectal. Ann Surg Oncol 2001; 8: 347–53. 37. Giacchetti S, Itzhaki M, Gruia G, et al. Long-term survival of patients with unresectable colorectal cancer liver metastases following infusional chemotherapy with 5-fluorouracil, leucovorin, oxaliplatin and surgery. Ann Oncol 1999; 10: 663–9. 38. Gillams AR, Lees WR. Radio-frequency ablation of colorectal liver metastases in 167 patients. Eur Radiol 2004; 14: 2261–7. 39. Hildebrand P, Kleemann M, Roblick UJ, et al. Radiofrequency-ablation of unresectable primary and secondary liver tumors: results in 88 patients. Langenbecks Arch Surg 2006; 391: 118–23. 40. Machi J, Oishi AJ, Sumida K, et al. Long-term outcome of radiofrequency ablation for unresectable liver metastases from colorectal cancer: evaluation of

08/02/13 5:59 PM

References  85

41.

42.

43.

44.

45.

46.

47.

48. 49.

50.

51.

52.

53.

54.

prognostic factors and effectiveness in first- and second-line management. Cancer J 2006; 12: 318–26. Stang A, Fischbach R, Teichmann W, Bokemeyer C, Braumann D. A systematic review on the clinical benefit and role of radiofrequency ablation as treatment of colorectal liver metastases. Eur J Cancer 2009; 45: 1748–56. Barbaro B, Caputo F, Tebala C, et al. Preoperative right portal vein embolisation: indications and results. Radiol Med 2009; 114: 553–70. Fong Y, Fortner J, Sun RL, Brennan MF, Blumgart LH. Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: analysis of 1001 consecutive cases. Ann Surg 1999; 230: 309–18, 318–21. Strasberg SM, Dehdashti F, Siegel BA, Drebin JA, Linehan D. Survival of patients evaluated by FDG-PET before hepatic resection for metastatic colorectal carcinoma: a prospective database study. Ann Surg 2001; 233: 293–9. Rama N, Monteiro A, Bernardo JE, Eugenio L, Antunes MJ. Lung metastases from colorectal cancer: surgical resection and prognostic factors. Eur J Cardiothorac Surg 2009; 35: 444–9. Borasio P, Gisabella M, Bille A, et al. Role of surgical resection in colorectal lung metastases: analysis of 137 patients. Int J Colorectal Dis 2011; 26: 183–90. Reinhardt MJ, Wiethoelter N, Matthies A, et al. PET recognition of pulmonary metastases on PET/CT imaging: impact of attenuation-corrected and nonattenuation-corrected PET images. Eur J Nucl Med Mol Imaging 2006; 33: 134–9. Cho KC, Gold BM. Computed tomography of Krukenberg tumors. AJR Am J Roentgenol 1985; 145: 285–8. Kuhlman JE, Hruban RH, Fishman EK. Krukenberg tumors: CT features and growth characteristics. South Med J 1989; 82: 1215–9. Kim SH, Kim WH, Park KJ, Lee JK, Kim JS. CT and MR findings of Krukenberg tumors: comparison with primary ovarian tumors. J Comput Assist Tomogr 1996; 20: 393–8. Glehen O, Kwiatkowski F, Sugarbaker PH, et al. Cytoreductive surgery combined with perioperative intraperitoneal chemotherapy for the management of peritoneal carcinomatosis from colorectal cancer: a multiinstitutional study. J Clin Oncol 2004; 22: 3284–92. Sugarbaker PH, Schellinx ME, Chang D, Koslowe P, von Meyerfeldt M. Peritoneal carcinomatosis from adenocarcinoma of the colon. World J Surg 1996; 20: 585–91, 592. Ridge JA, Bading JR, Gelbard AS, Benua RS, Daly JM. Perfusion of colorectal hepatic metastases. Relative distribution of flow from the hepatic artery and portal vein. Cancer 1987; 59: 1547–53. Schwartz L, Brody L, Brown K, et al. Prospective, blinded comparison of helical CT and CT arterial portography in the assessment of hepatic metastasis from colorectal carcinoma. World J Surg 2006; 30: 1892–9, 1900–1.

HEBK001-C05_p71-86.indd 85

55. Valls C, Andia E, Sanchez A, et al. Hepatic metastases from colorectal cancer: preoperative detection and assessment of resectability with helical CT. Radiology 2001; 218: 55–60. 56. Wong K, Paulson EK, Nelson RC. Breath-hold threedimensional CT of the liver with multi-detector row helical CT. Radiology 2001; 219: 75–9. 57. Scott DJ, Guthrie JA, Arnold P, et al. Dual phase helical CT versus portal venous phase CT for the detection of colorectal liver metastases: correlation with intra-operative sonography, surgical and pathological findings. Clin Radiol 2001; 56: 235–42. 58. Ch’en IY, Katz DS, Jeffrey RB Jr, et al. Do arterial phase helical CT images improve detection or characterization of colorectal liver metastases? J Comput Assist Tomogr 1997; 21: 391–7. 59. Jang HJ, Lim HK, Lee WJ, Lee SJ, Yun JY, Choi D. Small hypoattenuating lesions in the liver on singlephase helical CT in preoperative patients with gastric and colorectal cancer: prevalence, significance, and differentiating features. J Comput Assist Tomogr 2002; 26: 718–24. 60. Glover C, Douse P, Kane P, et al. Accuracy of investigations for asymptomatic colorectal liver metastases. Dis Colon Rectum 2002; 45: 476–84. 61. Niekel MC, Bipat S, Stoker J. Diagnostic imaging of colorectal liver metastases with CT, MR imaging, FDG PET, and/or FDG PET/CT: a meta-analysis of prospective studies including patients who have not previously undergone treatment. Radiology 2010; 257: 674–84. 62. Bipat S, van Leeuwen MS, Comans EF, et al. Colorectal liver metastases: CT, MR imaging, and PET for diagnosis: meta-analysis. Radiology 2005; 237: 123–31. 63. Wiering B, Krabbe PF, Dekker HM, Oyen WJ, Ruers TJ. The role of FDG–PET in the selection of patients with colorectal liver metastases. Ann Surg Oncol 2007; 14: 771–9. 64. Rosa F, Meimarakis G, Stahl A, et al. Colorectal cancer patients before resection of hepatic metastases: impact of (18)F-FDG PET on detecting extrahepatic disease. Nuklearmedizin 2004; 43: 135–40. 65. Truant S, Huglo D, Hebbar M, Ernst O, Steinling M, Pruvot FR. Prospective evaluation of the impact of [18F]fluoro-2-deoxy-d-glucose positron emission tomography of resectable colorectal liver metastases. Br J Surg 2005; 92: 362–9. 66. Zhang C, Chen Y, Xue H, et al. Diagnostic value of FDG-PET in recurrent colorectal carcinoma: a metaanalysis. Int J Cancer 2009; 124: 167–73. 67. Flamen P, Hoekstra OS, Homans F, et al. 2001. Unexplained rising carcinoembryonic antigen (CEA) in the postoperative surveillance of colorectal cancer: the utility of positron emission tomography (PET). Eur J Cancer 2001; 37: 862–9. 68. Heriot AG, Hicks RJ, Drummond EG, et al. Does positron emission tomography change management in

08/02/13 5:59 PM

86  Radiological staging for systemic disease

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

primary rectal cancer? A prospective assessment. Dis Colon Rectum 2004; 47: 451–8. Schussler-Fiorenza CM, Mahvi DM, Niederhuber J, Rikkers LF, Weber SM. Clinical risk score correlates with yield of PET scan in patients with colorectal hepatic metastases. J Gastrointest Surg 2004; 8: 150–57, 157–8. Wahl RL. Why nearly all PET of abdominal and pelvic cancers will be performed as PET/CT. J Nucl Med 2004; 45(suppl. 1): 82–95S. Orlacchio A, Schillaci O, Fusco N, et al. Role of PET/ CT in the detection of liver metastases from colorectal cancer. Radiol Med 2009; 114: 571–85. Davey K, Heriot AG, Mackay J, et al. The impact of 18-fluorodeoxyglucose positron emission tomography–computed tomography on the staging and management of primary rectal cancer. Dis Colon Rectum 2008; 51: 997–1003. Taylor FG, Swift RI, Blomqvist L, Brown G. A systematic approach to the interpretation of preoperative staging MRI for rectal cancer. AJR Am J Roentgenol 2008; 191: 1827–35. Cotton F, Pellet O, Gilly FN, Granier A, Sournac L, Glehen O. MRI evaluation of bulky tumor masses in the mesentery and bladder involvement in peritoneal carcinomatosis. Eur J Surg Oncol 2006; 32: 1212–6. Kong G, Jackson C, Koh DM, et al. The use of 18F-FDG PET/CT in colorectal liver metastases: comparison with CT and liver MRI. Eur J Nucl Med Mol Imaging 2008; 35: 1323–9. Semelka RC, Martin DR, Balci C, Lance T. Focal liver lesions: comparison of dual-phase CT and multisequence multiplanar MR imaging including dynamic gadolinium enhancement. J Magn Reson Imaging 2001; 13: 397–401. Reimer P, Jahnke N, Fiebich M, et al. Hepatic lesion detection and characterization: value of nonenhanced MR imaging, superparamagnetic iron oxide-enhanced MR imaging, and spiral CT-ROC analysis. Radiology 2000; 217: 152–8. Zech CJ, Herrmann KA, Reiser MF, Schoenberg SO. MR imaging in patients with suspected liver metastases: value of liver-specific contrast agent Gd-EOBDTPA. Magn Reson Med Sci 2007; 6: 43–52. Kenis C, Deckers F, de Foer B, van Mieghem F, van Laere S, Pouillon M. Diagnosis of liver metastases: can diffusion-weighted imaging (DWI) be used as a stand alone sequence? Eur J Radiol 2011 [Epub ahead of print]. Koh DM, Brown G, Riddell AM, et al. Detection of colorectal hepatic metastases using MnDPDP MR imaging and diffusion-weighted imaging (DWI) alone and in combination. Eur Radiol 2008; 18: 903–10. Kinkel K, Lu Y, Both M, Warren RS, Thoeni RF. Detection of hepatic metastases from cancers of the

HEBK001-C05_p71-86.indd 86

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

gastrointestinal tract by using noninvasive imaging methods (US, CT, MR imaging, PET): a meta-analysis. Radiology 2002; 224: 748–56. Yarmenitis SD, Karantanas A, Bakantaki A, Papantoniou Y, Gourtsoyiannis N. Detection of colorectal cancer hepatic metastases with contrast-enhanced ultrasound: comparison with conventional B-mode ultrasound. Dig Dis 2007; 25: 86–93. Larsen LP, Rosenkilde M, Christensen H, et al. Can contrast-enhanced ultrasonography replace multidetector-computed tomography in the detection of liver metastases from colorectal cancer? Eur J Radiol 2009; 69: 308–13. Torzilli G, del Fabbro D, Palmisano A, et al. Contrastenhanced intraoperative ultrasonography during hepatectomies for colorectal cancer liver metastases. J Gastrointest Surg 2005; 9: 1148–53, 1153–4. Mazzoni G, Napoli A, Mandetta S, et al. Intra-operative ultrasound for detection of liver metastases from colorectal cancer. Liver Int 2008; 28: 88–94. Siosteen AK, Elvin A. Intra-operative uses of contrastenhanced ultrasound. Eur Radiol 2004; 14(suppl. 8): P87–95. Gielen C, Sanli I, Stroeken L, Botterweck A, Hulsewe K, Hoofwijk A. Staging chest radiography is not useful in patients with colorectal cancer. Eur J Surg Oncol 2009; 35: 1174–8. Glazer ES, Beaty K, Abdalla EK, Vauthey JN, Curley SA. Effectiveness of positron emission tomography for predicting chemotherapy response in colorectal cancer liver metastases. Arch Surg 2010; 145: 340–5. Bacigalupo L, Aufort S, Eberle MC, Assenat E, Ychou M, Gallix B. Assessment of liver metastases from colorectal adenocarcinoma following chemotherapy: SPIO-MRI versus FDG-PET/CT. Radiol Med 2010; 115: 1087–100. Carnaghi C, Tronconi MC, Rimassa L, et al. Utility of 18F-FDG PET and contrast-enhanced CT scan in the assessment of residual liver metastasis from colorectal cancer following adjuvant chemotherapy. Nucl Med Rev Cent East Eur 2007; 10: 12–5. Spatz J, Holl G, Sciuk J, Anthuber M, Arnholdt HM, Markl B. Neoadjuvant chemotherapy affects staging of colorectal liver metastasis: a comparison of PET, CT and intraoperative ultrasound. Int J Colorectal Dis 2011; 26: 165–71. Van Kessel CS, van Leeuwen MS, van den Bosch MA, et al. Accuracy of multislice liver CT and MRI for preoperative assessment of colorectal liver metastases after neoadjuvant chemotherapy. Dig Surg 2011; 28: 36–43. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009; 45: 228–47.

08/02/13 5:59 PM

6 Preoperative radiotherapy and chemoradiotherapy for rectal cancer Rob Glynne-Jones and Mark Harrison

Introduction Colorectal cancer is one of the most common solid tumours, of which approximately 40 per cent of patients have rectal cancer. Historically, surgery alone for rectal cancer has been associated with a high incidence of local recurrence. Additionally, some 10–40 per cent of patients with rectal cancer require extirpative procedures leading to a permanent stoma. A landmark retrospective study highlighted the poor prognosis for patients with rectal cancer, with a 5-year survival of 6 per cent in patients with pathologically Dukes C cancer.1 In the light of these results, using post-surgical histology to define risk, several postoperative studies investigated chemotherapy and radiotherapy2 or a combination of both to prevent local recurrence and reduce metastases. Overall, these studies showed a significant benefit for the combination of chemoradiation.3,4 Interestingly, a retrospective analysis of pooled data from US trials showed similar 5-year overall survival for patients with pT3N0 rectal cancer treated with either surgery or chemotherapy alone (84 per cent) compared with those treated with postoperative chemoradiotherapy (74–80 per cent).5 Similar data have been reported from the National Cancer Data Base.6 With the introduction of better preoperative imaging such as computed tomography (CT),

HEBK001-C06_p87-102.indd 87

ultrasound and magnetic resonance imaging (MRI) to stage patients, the success of postoperative chemoradiation was extrapolated to the preoperative setting. The concept of spatial cooperation in preoperative chemoradiation for locally advanced rectal cancer is attractive.7 Chemotherapy may, as a component of chemoradiation, act as a radiosensitizing agent; it may also potentially eradicate distant micro-metastases. In general, concurrent chemotherapy has proved more effective than the sequential use of these modalities.8 A series of randomized controlled trials have clearly shown the superiority of preoperative radiation or chemoradiation therapy compared with postoperative treatment.9–13 Current recommendations suggest the use of preoperative chemoradiation followed by total mesorectal excision (TME) surgery as the standard treatment of choice for the majority of patients diagnosed with advanced rectal cancer. In the past two decades, greater understanding of the natural history of the disease and patterns of recurrence, and more precise histopathological reporting, have helped to define patients with a high risk of local recurrence and metastatic disease following ‘curative resection’.14 This particular focus on the circumferential resection margin (CRM) has driven technical advances in surgical technique with meticulous surgical dissection along ­embryological planes and highlighted the ­importance of

08/02/13 5:58 PM

88  Preoperative radiotherapy and chemoradiotherapy for rectal cancer

­ reoperative assessment of the tumour with MRI. p Even technically optimized surgery is unlikely to achieve a curative resection when the preoperative MRI shows a threatened or breached CRM. In addition, at the time of diagnosis, some 20–25 per cent of patients with rectal cancer are found to have overt metastatic disease, and a further 30–40 per cent subsequently develop metastases. The multimodality treatment of rectal cancer attempts to integrate the three modalities of surgery, radiotherapy and chemotherapy as strategies to eradicate cells at the margins or in discontinuous areas of tumour within the pelvis, in lymph nodes or in distant metastatic sites to improve both local control and overall survival. Systematic reviews and several meta-analyses have examined the role of radiotherapy (as opposed to chemoradiation) in resectable rectal cancer.15–17 These meta-analyses concluded that the evidence from relevant trials favoured preoperative rather than postoperative radiotherapy, and that a biologically equivalent dose greater than 30 Gy is more effective in reducing local relapse. A 2007 Cochrane review further supports these findings but failed to confirm a reduction in mortality or any advantage in terms of sphincter sparing from preoperative radiotherapy.18 The Cochrane review also highlighted the significant risks of increased pelvic and perineal infections when radiation and surgery are combined, and a detriment to the patient’s anorectal, urinary and sexual function. Since the early 1980s, the fluoropyrimidine 5-fluorouracil (5FU) alone, and more recently in combinations of cytotoxic chemotherapy using oxaliplatin or irinotecan, has represented the mainstay of chemotherapy treatment for patients with advanced and metastatic colorectal cancer. These combinations have been integrated into chemoradiation regimens and tested in a number of prospective randomized trials in rectal cancer, potentially to mirror the success of 5FU and oxaliplatin in dealing with distant micro-metastases in the adjuvant setting in colon cancer.19–21 Investigators have used these cytotoxic agents either as ­radiosensitizers22,23 or in an attempt to use systemically active chemotherapy preoperatively with an appropriate duration.24–26 Perhaps surprisingly when used as a radiosensitizer, these combinations of a cytotoxic agent and radiation have had only moderate success in improving outcome in rectal cancer.12,13,22,23,27

HEBK001-C06_p87-102.indd 88

More recently, a number of molecularly targeted agents have been integrated into standard palliative chemotherapy regimens for advanced metastatic colorectal cancer. These biologically targeted agents appear to improve response rates and extend progression-free and overall survival, albeit with varying success.28–31 Three monoclonal antibodies have now entered clinical practice in colorectal cancer – cetuximab, panitumumab and bevacizumab. A strategy to incorporate these newer biologically active targeted agents into chemoradiation schedules has also emerged. In this chapter, we examine the rationale for preoperative radiation and chemoradiation, the selection of patients who are appropriate for long-course preoperative radiotherapy and short-course preoperative radiation (SCPRT), the techniques available such as intensity-modulated radiotherapy (IMRT), optimal doses and field size, the integration of cytotoxic chemotherapy and biological agents, potential biological markers, and future potential strategies of treatment.

Defining the most appropriate treatment strategy The terms ‘favourable’, ‘early’ and ‘good’, ‘intermediate’ and ‘bad’, and ‘locally advanced’ and ‘ugly’ have historically been used for categorizing rectal cancer into clinical subgroups. The term ‘locally advanced’ has been commonly used in a very broad and undefined way and generally directed the patient to receive preoperative treatment. In the 1980s, preoperative assessment and clinical staging of rectal cancer was limited to digital rectal examination, possibly a rigid sigmoidoscopy and a barium enema. A clinical judgement based on these findings assessed tumour fixity, but there was no widely accepted and validated imaging method of defining either locally advanced rectal cancer or unresectable disease. Currently, we broadly consider patients to have easily resectable cancers (which can be cured by radical surgery alone), borderline resectable disease (i.e. a potentially ‘threatened’ or minimally breached circumferential margin as predicted by MRI), or unresectable cancers with disease outside the mesorectum (for whom surgery is not possible without leaving tumour within the pelvis).

08/02/13 5:58 PM

PREOPERATIVE RADIOTHERAPY  89

Our increased level of knowledge is based on the use of preoperative MRI, which allows division of rectal cancers into categories with different risks and different indications for preoperative radiotherapy. Preoperative treatment may be tailored to the radiological stage of the tumour to facilitate a reduction of risk of local recurrence, improved likelihood of curative resection, or identification of unresectable disease and prevention of an unnecessary surgical procedure. Use of MRI allows rectal cancer to be divided into the following groups: 1. very early (some cT1 sm1/2) 2. low risk (.cT1 sm2 to cT2, some cT3a (to T3c), CRM 2 /N0) 3. locally advanced (cT3, some T4 crm 2 /N1) 4. potentially unresectable advanced (cT3 CRM1 or cT4) 5. unresectable disease (extension to lateral pelvic side-wall/lateral pelvic lymph node involvement/disease outside the pelvis). Other well-recognized factors such as tumour height, proximity to the CRM, cN stage, lymphovascular invasion, extramural vascular invasion and perineural invasion are also relevant to the risk of local recurrence and the development of metastatic disease. Where possible, these factors should be mentioned in the MRI report with the proviso that they are categorized with the prefix ‘mr’, e.g. mrN1 or mrEMVI. The relevance of the historical clinical classification into fixed, tethered or mobile tumours is unclear if MRI is performed for decision-making, but it may be complementary, especially in low rectal tumours. In all patients, a history, clinical examination and rectal examination by the responsible surgeon is mandatory, and the surgical opinion on current and likely future anorectal function should be transmitted to the radiation oncologist, ideally at the multidisciplinary team colorectal conference.

PREOPERATIVE RADIOTHERAPY There are a number of theoretical advantages of preoperative radiation therapy compared with postoperative radiation therapy. To an extent, these advantages have been confirmed in clinical trials: Preoperative radiation can be delivered to an area where the blood supply has not yet been

l

HEBK001-C06_p87-102.indd 89



l



l



l



l



l



l

compromised by surgery and infection; theoretically, better oxygenation will improve the radiosensitivity of the tumour cells. An R0 resection is much more likely if the tumour undergoes a degree of shrinkage. Killing or sterilizing all the tumour cells with radiation before surgery may reduce the risk of tumour cells seeding in the wound during surgery, particularly in the posterior pelvis when the patient lies supine, thereby decreasing the local recurrence rate. With preoperative treatment, the tumour, draining lymph nodes and adjacent mesorectum can be targeted in the gross tumour volume and clinical target volume, rather than an uninvolved length of bowel and mesocolon, which may already have a compromised blood supply. The clinical target volume will not be required to cover the perineum, and there will be no radiation to the anastomotic site. The small bowel is much less likely to be included in the radiation field, unless the patient has had a hysterectomy or previous pelvic surgery. Although still unvalidated, the preoperative approach may define good and bad prognostic groups of patients by their response to ­treatment.32–36

For all these reasons, the expectation is that the acute and late toxicity of preoperative radiation will be less, and hence compliance will be better and more patients will receive the full dose of radiation. Three adjuvant radiation approaches have been in common use: SCPRT, long-course preoperative radiation and long-course preoperative chemoradiation. Long-course radiation alone has been shown to be less effective in terms of local control when compared with preoperative 5FU-based chemoradiation27,37 and hence is used only in patients who are unfit for chemotherapy; it will be discussed only briefly here. There are different rationales for SCPRT and chemoradiotherapy, and some differences in tumour target volume definition that will be discussed later. Short-course preoperative radiotherapy uses a short intensive course of radiation delivered with a dose of 25 Gy over 5 days. The single aim of treatment is to reduce the risk of pelvic recurrence. This approach has gained widespread acceptance in Europe following the publication of

08/02/13 5:58 PM

90  Preoperative radiotherapy and chemoradiotherapy for rectal cancer

the ­Swedish and Dutch rectal cancer studies.38,39 In contrast, the rationale for preoperative chemoradiotherapy is more attractive, as it combines early systemic chemotherapy with a locoregional treatment. Current criteria in our own units for the delivery of preoperative chemoradiation rely on MRI scanning in determining the extent of tumour either at (i.e. within 1 mm) or outside the mesorectal envelope or below the envelope extending into the levators and sphincter mechanism. Using these MRI criteria, it is possible to predict patients in whom there is a high risk that the surgeon will not be able to perform an R0 resection and will leave gross macroscopic or undetected microscopic tumour behind in the pelvis.

Rendering unresectable tumours resectable A minority of patients present with a rectal tumour that cannot be resected without leaving residual macroscopic or microscopic disease. The multimodality approach to rectal cancer management allows these patients to be identified before an (unsuccessful) attempt at surgical resection is made. This approach facilitates the use of preoperative radiotherapy or chemoradiotherapy. Before the development of pelvic MRI, most patients were identified by the presence of tumour fixity determined either in the clinic or during an examination under anaesthetic. Pragmatically this method of assessment is limited to those tumours where the examining finger can assess mobility of the tumour. Pelvic MRI can assist in demonstrating the relationship of the primary tumour to the surrounding mesorectal fascia and the surrounding organs, as well as enlarged lymph nodes with MRI appearances of tumour involvement. Thus, patients may be selected for preoperative radiation in this category if there is evidence of primary tumour involving or extending beyond the mesorectal fascia; primary tumour within 1–2 mm of the mesorectal fascia; involved lymph nodes outside the mesorectal fascia (usually iliac nodes); and primary tumour involving the levators. Since response and shrinkage of the tumour is required, these patients require preoperative chemoradiotherapy. Current approaches use a fluoropyrimidine combined with radiation, although many phase II studies of the use

HEBK001-C06_p87-102.indd 90

of combination chemotherapy schedules have been completed or are ongoing. Two randomized trials in unresectable disease have confirmed an advantage for 5FU-based chemoradiation over radiation alone.40 One very small phase III trial in unresectable rectal cancer compared radiation alone with chemoradiotherapy in Sweden, demonstrating improved resectability and local control with the use of chemoradiotherapy.41

Reducing local recurrence The majority of randomized trials in rectal cancer have been performed in patients with resectable rectal cancer, where the aim of treatment has been to lower the risk of local recurrence. Some of these trials were performed before 2000 prior to the acceptance of the benefits of meticulous surgical dissection in the form of mesorectal excision. The local recurrence rates in the surgery alone arm of these studies were in the range of 21–36.5 per cent.16 Three retrospective systematic reviews/meta-analyses have been published on the role of radiotherapy in rectal cancer.15–17 The Colorectal Cancer Collaborative Group meta-analysis identified 22 randomized controlled trials that have compared the use of both preoperative radiotherapy (14 trials, 6350 patients) and postoperative radiotherapy (8 trials, 2157 patients) in patients intended to undergo a curative resection for rectal cancer, versus surgery alone.16 In the entire group of patients, overall survival at 5 years was not significantly different, with a 5-year overall survival of 45 per cent in the patients receiving radiation versus 42.1 per cent in the patients treated with surgery alone. Despite the lack of effect on overall survival, there was an expected reduction in local recurrence. In the preoperative trials, the rate of local recurrence at 5 years was 12.5 per cent in the patients treated with radiation versus 22.2 per cent in the patients treated with surgery alone. Postoperative radiation also reduced the rate of local recurrence from 23.8 per cent in the patients treated with surgery alone to 16.9 per cent in the patients receiving radiation. Trials performed in individual countries have influenced their definition of the ‘standard of care.’ A German phase III trial reported improved outcome for preoperative chemoradiotherapy in this setting. A total of 823 patients were randomized between

08/02/13 5:58 PM

LONG-COURSE PREOPERATIVE 5-FLUOROURACIL-BASED CHEMORADIOTHERAPY  91

preoperative chemoradiotherapy and postoperative chemoradiotherapy (patients received postoperative adjuvant chemotherapy in both arms of the trial). Local recurrence and acute and late toxicity were statistically significantly reduced with the preoperative approach, but there was no difference in the distant metastases rate or overall survival.12 In contrast to the German approach, a short accelerated preoperative radiation schedule of 25 Gy in five fractions (SCPRT) with surgery performed within a few days of completion of radiation was developed in Sweden to reduce the risk of local recurrence. The Swedish Rectal Cancer Trial demonstrated a significant reduction in local recurrence and a 10 per cent absolute improvement in survival; this led to the widespread adoption of SCPRT in the Scandinavian and northern European countries.38 The Swedish Rectal Cancer Trial is the only major modern trial to show a benefit in survival from preoperative radiotherapy. This trial randomized 1168 patients with resectable rectal cancer to SCPRT followed by surgery, versus surgery alone. The radiation was delivered over 5 days with 5 Gy per fraction. The 5-year overall survival was significantly better in the radiation–surgery group (58 per cent) than in the surgery-only group (48 per cent) (P 5 0.004). The benefit of radiation persisted through a median of 13 years’ follow-up (38 per cent alive v. 30 per cent, P 5 0.008). Local recurrence as either a first or later event was less common in the group of patients treated with radiation (9 per cent v. 26 per cent, P , 0.001).42 The survival advantage seen in this trial has been suggested to derive from the large difference in local recurrence rates between the two groups, which may have been a reflection of the suboptimal surgery before the TME era. The Dutch Colorectal Cancer Group conducted a large, prospective, randomized trial using an identical radiation schedule to that used in the Swedish Rectal Cancer Trial. In the Dutch trial a series of surgical training programmes and mentoring attempted to standardize the surgical procedure, and all patients underwent surgery by TME principles. Of 1805 eligible patients, 897 were randomized to radiation and then immediate surgery (within 1 week of completion) and 908 were randomized to surgery alone. The overall survival at 5 years was similar in the two groups: 64.2 per cent in the preoperative radiation group compared

HEBK001-C06_p87-102.indd 91

with 63.5 per cent in the surgery-only group. Local recurrence at 5 years was 5.6 per cent and 10.9 per cent, respectively (P , 0.001). The low level of local recurrence rate in this trial has led to the acceptance of TME as the standard surgical technique. Mature outcome from the trial has confirmed a reduction in local recurrence but no survival benefit.43 In the UK, some colorectal multidisciplinary teams have adopted the non-selective use of routine SCPRT, while others use pelvic MRI to determine patients whose primary tumour is predicted to be clear of the CRM and in whom initial surgery is performed. The Dutch and Medical Research Council (MRC) CR07 trials have directly compared these two approaches. The Dutch trial examined the routine use of SCPRT followed by TME, with TME and selective postoperative radiotherapy in the event of histopathological evidence of involvement of the circumferential resection margin. The Dutch trial appears to show an absolute reduction in local recurrence of 6 per cent. Thus, if 100 patients are irradiated, 6 local recurrences are prevented, i.e. the number of patients needed to treat to prevent 1 local recurrence is 16.7. In the MRC CR07 trial, if a good mesorectal excision was performed,44 the number needed to treat appears to be over 20. As stated above, the present authors do not therefore support the widespread advocacy for routine adjuvant radiotherapy as used in the treatment arms of recent trials. There is increasing evidence that patients with very low tumours that require abdominoperineal excision are at higher risk of involvement of the CRM, local recurrence and inferior survival. Most agree that routine preoperative radiotherapy is indicated for this group of patients. It is a source of considerable debate whether SCPRT may be used in some patients, whether chemoradiotherapy is preferred in other patients, and whether a wider cylindrical surgical technique should be adopted to reduce the rate of an involved CRM.

LONG-COURSE PREOPERATIVE 5-FLUOROURACIL-BASED CHEMORADIOTHERAPY – 45–50 GY IN 5 WEEKS Randomized studies in resectable T3/T4 rectal cancers have compared preoperative chemoradiation

08/02/13 5:58 PM

92  Preoperative radiotherapy and chemoradiotherapy for rectal cancer

(45 Gy in 25 fractions over 5 weeks) and 5FU-based chemotherapy with radiation alone.27,37 Surgery is undertaken 4–12 weeks following chemoradiotherapy to allow the patient to recover and enable tumour shrinkage. The addition of 5FU to preoperative radiation increases the pathological complete response rate over radiotherapy alone and provides evidence for improvements in locoregional control but has not translated into an improvement in disease-free survival or overall survival. Trials in unresectable disease have also confirmed an advantage for chemoradiation over radiation alone.40 Local control was significantly better (82 per cent v. 67 per cent at 5 years, P 5 0.03), with a trend towards improved overall survival (66 per cent v. 53 per cent at 5 years, P 5 0.09).

Oral Alternatives to 5-Fluorouracil In the neoadjuvant setting in rectal cancer, the safety and efficacy of capecitabine45 and a combination of uracil and tegafur46 have been compared with that of a continuous intravenous infusion (CVI) of 5FU. For capecitabine, the authors used data from 345 patients treated with capecitabine and compared them with 197 patients treated with CVI 5FU.45 The pathological complete response was significantly higher for capecitabine than CVI 5FU (25 per cent v. 13 per cent).

When is it safe to proceed to surgery alone? The selection of patients in whom initial surgical resection is safe and who do not require preoperative radiotherapy remains a controversial issue. The present authors believe that patients in groups 1 and 2, and even some categories of group 3, can avoid preoperative radiotherapy in view of the significant associated morbidity in terms of increased pelvic and perineal post-surgical infections, and significant late anorectal, urinary and sexual dysfunction.18 About 5–10 per cent of patients experience grade 3 or 4 late morbidity.47,48 Small bowel tolerance is the main dose-limiting factor, and the volume of the small bowel in the radiation field is crucial,49 although only 1–2 per cent of patients require early surgical intervention for small ­bowel

HEBK001-C06_p87-102.indd 92

damage. Mature results of the Swedish Rectal Cancer Trial confirm problems such as late bowel obstruction and abdominal pain.50 Additionally, there are unexplained late cardiac effects and an increased risk of a second malignancy.51 As followup in the majority of studies is generally short, there is likely to be a major underestimate of the real risks of late effects. Our caveat would be that in some cases of low rectal cancer, it may be difficult to discriminate between a cT2 and cT3 or even cT4 at the level of the levators and below. High-quality MRI is not always available. Low rectal tumours behave more aggressively compared with cancers in the upper rectum. High-quality surgery cannot always achieve a curative resection for locally advanced cancers that extend below the levators. Currently, more aggressive surgical approaches such as the ‘cylindrical’ or extralevator approach are being explored, and these may eventually be shown to be superior to standard abdominoperineal excision. Preoperative chemoradiation still has a major role in many patients with low rectal cancer, however. It should be stressed that the choice of not using preoperative 5FU-based chemoradiotherapy represents a positive decision to forgo effective neoadjuvant treatment. If the surgical histopathology subsequently reveals a positive circumferential margin, the optimal window for treatment may have been lost.

SHORT-COURSE PREOPERATIVE RADIATION VERSUS CHEMORADIOTHERAPY In Europe and Australia, two trials have directly compared SCPRT with chemoradiation for patients with resectable rectal tumours.52,53 The Polish study found no difference in long-term outcomes of locoregional control, disease-free survival or overall survival (Table 6.1). Preliminary results from the Australian study also failed to show differences in late outcome. If the premise is accepted that in resectable cancers, where the CRM is not threatened, SCPRT and chemoradiotherapy are equivalent in terms of outcomes such as local recurrence, disease-free survival and overall survival, then these results raise questions about the importance of down-staging and the effect of pathological ­complete response

08/02/13 5:58 PM

HEBK001-C06_p87-102.indd 93

Table 6.1. Trial design: preoperative radiotherapy versus preoperative chemoradiation in resectable rectal cancer.@

Patients Trial EORTC

54

Patients

Primary

Local

Pathological

circumferential

Overall

free

complete

resection

Duration

(n)

Randomization

(n)

TME

end-point

recurrence

Metastases

survival

survival

response

margin

1972–1984,

247

34.5 Gy v. 5FU 1

121 v.

No

Overall

15% v.

30% overall

59% v. 46%

72% v.

2.5% v. 5%

Not Quirke

5% v. 14%

Not Quirke

3.6% v.

Not Quirke

12 years

34.5 Gy

survival

126

at 5 years

15% at

3 years

5 years EORTC 2292127

1992–2004,

1011

12 years

45 Gy in 25 fractions v. FUFA 1 45 Gy

506 v.

No

Overall survival

505

17.1% v. 8.7% at

34.4% overall

5 years FFCD 920337

1993–2005,

762

12 years

45 Gy in 25 fractions

367 v.

v. FUFA 1 45 Gy

375

25 Gy in 5 fractions/

155 v.

No

Overall survival

16.5% v.

No data

trial52

1999–2004,

316

SCPRT v. FUFA 1

5 years

Yes

157

50 Gy TROG 01-

2001–2006,

0453

5 years

326

25 Gy in 5 fractions/ SCPRT v. 50.4 Gy 1 PVI 5FU

No data

Yes

Sphincter-

11% v.

64.8% v.

54.4% v.

65.8% at

56.1% at

5 years

5 years

67.9% v.

No data

11.4%

67.4% at

8.1% at

5 years

5 years Polish

68% at

No data

67.2% v.

58.4% v.

preserving

16.5% at

66.2% at

55.6% at

surgery

5 years

4 years

4 years

Local recurrence

7.5% v. 4.1% at

72% v. 69%

74% v. 70% at 5 years

No data

1% v. 16%

13% v. 4% (CRM)

No data

Not Quirke v. no data

3 years

EORTC, European Organisation for Research and Treatment of Cancer; FFCD, Fédération Francophone de Cancérologie Digestive; 5FU, 5-fluorouracil; FUFA, 5-fluorouracil plus folinic acid; PVI, protracted venous infusion; TME, total mesorectal excision; TROG, Trans-Tasman Radiation Oncology Group.

SHORT-COURSE PREOPERATIVE RADIATION VERSUS CHEMORADIOTHERAPY  93

Pathology Disease-

08/02/13 5:58 PM

94  Preoperative radiotherapy and chemoradiotherapy for rectal cancer

on survival. Additionally, a positive CRM may not be as significant an adverse prognostic factor after SCPRT as after chemoradiotherapy.

Facilitating sphinctersparing procedures Many surgeons consider the low position of some rectal cancers (below 6 cm from the anal verge) to inevitably require an abdominoperineal excision resection, particularly if the sphincter is invaded. In addition, a bulky anterior tumour in an obese man with a narrow pelvis may prove technically demanding to achieve sphincter-sparing surgery. These situations have led to a commonly held belief that preoperative radiation will allow an increase in the rate of sphincter-preserving procedures performed. Possibly, but not consistently, if a bulky polypoidal tumour shrinks, this may allow the surgeon to safely navigate around the mesorectal fascia down on to the pelvic floor and achieve a low anterior resection. This would not be possible without prior radiation, in which case abdominoperineal excision may be required. More controversially is the choice of the distal resection margin. Should this be chosen as 1 cm below the distal edge of a tumour that has shown major regression, or is it only oncologically safe to choose the distal margin on the basis of the initial tumour location and size? One study compared immediate and delayed surgery after preoperative radiotherapy.55,56 Sphincter preservation was achieved in 79 per cent of patients with a long interval (4–6 weeks) following radiotherapy, compared with 69 per cent of patients with a shorter interval (2 weeks). A further randomized study from this group suggests that dose escalation of the radiotherapy with an endoluminal boost may offer a higher rate of complete clinical response and hence increase the chance of sphincter preservation from 44 per cent to 76 per cent.57 Several non-randomized surgical series have reported results of patients with clinically resectable rectal cancer, in whom clinical assessment by their surgeon categorized the height of the cancer from the anal verge as requiring an abdominoperineal excision. These patients then received long-course preoperative radiotherapy and were subsequently reassessed before definitive surgery. Some studies used radiotherapy alone,55,58,59 and

HEBK001-C06_p87-102.indd 94

others used synchronous chemoradiation.36,60,61 The putative hypothesis is supported by evidence from randomized studies. Preoperative chemoradiation appears to offer a 10 per cent13 or even a 20 per cent12,62 higher chance overall in achieving sphincter-saving surgery. These trials were not specifically designed to answer the question of the sphincter preservation rate, however, and these data represent subset analysis. The German data are particularly liable to unforeseen biases, as the Zelen method of randomization was used, and there are significant differences in the numbers in each group.

RADIOTHERAPY ISSUES Field Size The optimal clinical target volume in preoperative radiotherapy and chemoradiotherapy for resectable rectal cancer before TME remains poorly defined. Total mesorectal excision does not remove all nodal stations potentially harbouring subclinical disease. Contouring a clinical target volume is a balance between encompassing structures at risk of containing residual cancer cells after TME and the desire to minimize normal tissue toxicity and surgical m ­ orbidity. Many series in the pre-TME era mapped recurrence, including the seminal work from Gunderson at the Mayo Clinic, amassed from ‘second-look’ surgery.63 Previously, groups have made recommendations for clinical target volume contouring based on observed sites of local recurrence.64–72 Historically, wide-field radiotherapy using bony landmarks was delivered to almost all patients. Early studies defined the superior border as the junction of L5/ S1, but some studies treated up to the origin of the inferior mesenteric artery at L4 and were associated with significant morbidity. Field sizes delivered in the USA are generally larger than in Europe. The recommended field size in one randomized phase II study suggested ‘The superior border of the treatment volume was at the L5–S1 junction with the inferior border a minimum of 5 cm inferior to the distal-most extent of the tumour or the anal verge as identified by a marker on simulation’ [our italics].66 In the UK field sizes are likely to be at least 4–5 cm shorter in the craniocaudal aspect.

08/02/13 5:58 PM

RADIOTHERAPY ISSUES  95

The studies of local recurrence after TME relate to SCPRT, and there are very few relating to chemoradiotherapy. Lymph nodes and their precise locations are well described in Far Eastern series but not generally in Western series. Few studies recommend individualizing clinical target volume according to site, stage and risk. Data on quality of life and late morbidity are sparse. We recommend clinical target volume-customizing delineations based on site and stage, with adaptations for high-risk features according to clinical and MRI staging. Although these proposed clinical target volume delineations have not been evaluated in randomized trials, their use appears rational and based on current evidence.

Total Dose Conventionally, when 1.8 Gy per fraction is used, total doses in the range 45–50.4 Gy have been delivered in the preoperative setting, and 50.4 Gy with the option of a 5.4 Gy boost to the tumour bed in the postoperative setting. It is assumed that the treatment will be delivered 5 days per week, 1 ­fraction per day, 1.8 Gy per fraction. The most appropriate dose is not precisely known, but in association with chemotherapy the total radiation dose should be at least 45 Gy in fractions of 1.8–2 Gy. A boost of 4–6 Gy in two to four fractions to the primary tumour is often given, thus reducing the radiation dose to the entire volume when chemoradiotherapy is given. Ideal preoperative dose

Some SCPRT studies have used 20 Gy in five fractions,73 but the Swedish Rectal Cancer Study, the Dutch study and the CR07 study use 25 Gy in five fractions, which most clinicians accept is a dose fractionation that cannot safely be exceeded. With the use of 1.8–2.0 Gy per fraction, however, the total dose may range from 45 Gy to 54 Gy in most reported series. This upper limit is 20 per cent higher than 45 Gy, and the balance between improved local control and the increased risk of late toxicity is continually debated. The Colorectal Cancer Collaborative Group meta-analysis confirmed that when preoperative radiotherapy is used, a biologically equivalent dose greater than 30 Gy is more effective in reducing local relapse.16 A dose of 5 Gy daily for 5 days represents a biologically equivalent dose of 37.5 Gy accord-

HEBK001-C06_p87-102.indd 95

ing to the linear quadratic equation, but in practice may be even higher. Early retrospective studies suggest an increase in local control with a dose of 50 Gy compared with 40 Gy.74 More recent data from phase II studies suggest that higher doses may be associated with lower risks of local relapse.75–77 A sequential phase II study from Canada has been reported. Three sequential schedules combined radiation with infusional 5FU, escalated from 40 Gy in 20 fractions to 46 Gy in 23 fractions and finally to 50 Gy in 25 fractions. A statistically significant difference in terms of local control was observed for doses of 46 Gy and above, but there was no difference between doses of 46 Gy and 50 Gy.78 The same study also appeared to show a trend to higher pathological complete response rates with increasing radiation dose of 13 per cent, 21 per cent and 31 per cent for 40 Gy, 46 Gy and 50 Gy, respectively.

Hyperfractionation Short-course preoperative radiotherapy uses fewer fractions at a higher dose per fraction (5 3 5 Gy), which might be expected to cause more late morbidity64,66 and which cannot allow integration of chemotherapy. Hence, a number of studies have used hyperfractionated radiotherapy with a view to reducing late effects.79–81 Widder and colleagues reported SCPRT with twicedaily fractions of 2.5 Gy to a total dose of 25 Gy within 1 week.79 The authors reported a high local control rate (98 per cent after 4 years) along with low rates of toxicity. It should be kept in mind, however, that theoretical anti-tumour efficiency is significantly reduced in this fractionation protocol. A more recent study modified SCPRT in rectal cancer to deliver twicedaily fractions of 2.9 Gy to a total dose of 29 Gy in 1 week immediately before surgery.82 Adjuvant chemotherapy was intended to be offered subsequently to patients with any pathological tumour assessment of stage II or higher in the resected specimen.

Brachytherapy High-dose-rate intraluminal brachytherapy (HDRILBT) has the advantage of high conformality, i.e. a rapid fall-off of radiation dose, which allows the delivery of a high dose to the tumour while sparing normal surrounding structures such as the

08/02/13 5:58 PM

96  Preoperative radiotherapy and chemoradiotherapy for rectal cancer

contralateral normal rectal mucosa, bladder and small bowel.83,84 Limited data are available evaluating the advantages of HDR-ILBT with external beam radiotherapy (EBRT) compared with EBRT alone.85 High-dose-rate intraluminal brachytherapy for advanced or inoperable tumours of the rectum has been used both in the palliative setting and to dose escalate after chemoradiation for curative treatment.86,87 It has also been used as a single modality in the preoperative setting, with 30 per cent pathological complete response;83 again, however, there are limited data regarding HDR-ILBT as a boost to the gross tumour volume in combination with preoperative chemoradiation for rectal cancers. In principle HDR-ILBT as a component or following chemoradiation may avoid a colostomy by increasing down-staging and facilitating sphincter preservation57 or by potentiating a non-operative approach.84

Intensity-Modulated Radiotherapy Technical advances such as IMRT allow greater precision and sparing of normal surrounding structures such as small bowel compared with conventional two-dimensional or three-dimensional planning. Intensity-modulated radiotherapy may allow improved compliance or facilitate dose escalation without increasing late morbidity. In addition, there was no difference with regard to late toxicity for any of the four arms in the European Organisation for Research and Treatment of Cancer (EORTC) 22921 study trial: 522 patients retained their sphincters, and of these only 1.4 per cent required surgery for small bowel complications. This low level of late morbidity calls into question the need to deliver IMRT with the aim of reducing small bowel toxicity.

Other cytotoxic drugs The addition of oxaliplatin and irinotecan to 5FUbased chemotherapy or oral fluoropyrimidines has been explored within a chemoradiotherapy schedule in numerous phase II studies in an attempt to increase tumour shrinkage before surgery and ­potentially mirror the success of oxaliplatin in dealing with distant micro-metastases in the adjuvant setting in colon cancer.20,21

HEBK001-C06_p87-102.indd 96

Phase II trials provide additional toxicity data and potential regimens for future phase III randomized trials in patients with rectal cancer. The Radiation Therapy Oncology Group (RTOG) 0247 trial, a randomized phase II trial, compared capecitabine administered 5 days per week and weekly oxaliplatin plus radiation administered before surgery followed by nine cycles of postoperative oxaliplatin with 5FU and folinic acid chemotherapy (FOLFOX) with irinotecan on a weekly schedule substituted for oxaliplatin. Another phase II trial, E3204, combines capecitabine administered 5 days per week plus weekly oxaliplatin and bevacizumab with radiation every other week; postoperative patients receive 12 cycles of FOLFOX and bevacizumab. Two randomized phase III studies have examined one versus two cytotoxic drugs (Table 6.2).22,23 In the Italian Studio Terapia Adjuvante Retto (STAR) trial, oxaliplatin added little to early assessable end-points but increased toxicity when added to a standard preoperative chemoradiation regimen for locally advanced rectal cancer. There were no significant differences in pathological complete response, local tumour response or tumour down-staging between patients who received FU-based chemoradiation with weekly oxaliplatin infusions.22 The Fédération Francophone de Cancérologie Digestive (FFCD) Fédération Nationale des Centres de Lutte Contre le Cancer (FNCLCC) ACCORD 12 trial (NCT00227747) compares neoadjuvant preoperative capecitabine combined with 45 Gy in 25 fractions over 5 weeks against capecitabine and oxaliplatin partnered with 50 Gy in 25 fractions over 5 weeks. Only data on toxicity and early pathological and surgical end-points are currently available.23,37 The design of this study is poor, because capecitabine in combination with 45 Gy (arm A) is compared with capecitabine/oxaliplatin and 50 Gy (arm B). The pathological complete response appears higher (18.8 per cent) in the capecitabine/oxaliplatin arm (arm B) than in arm A (13.8 per cent), but it is not clear whether this reflects the addition of oxaliplatin or the higher radiotherapy dose.23 It remains to be seen whether disease-free survival and overall survival are improved by the strategy of adding oxaliplatin in this setting, but in the meantime the addition of irinotecan and ­oxaliplatin to 5FU-based chemoradiation in patients with resectable tumours remains investigational.

08/02/13 5:58 PM

THE FUTURE  97 Table 6.2. Ongoing phase III trials of neoadjuvant chemoradiation in resectable rectal cancer. Concurrent chemoradiotherapy

Radiotherapy

Study

(CRT) randomization

dose

Patients (n)

Preoperative treatment

Comments

NSABP R-04,

Capecitabine or 5FU/

50.4 Gy in 28

1606;

Recommended to consider

Primary end-

NCT00058474

CAO/ARO/AIO 04

CRT v. capecitabine

fractions and

closed to

or 5FU 1 oxaliplatin

55.8 Gy for

recruitment

(every 2 weeks)/CRT

fixed tumours

CI 5FU (days 1–15,

50 Gy in 25

ECOG 5204

point: 3-year locoregional disease control

1259;

5FU 34 or FOLFOX 34

Primary end-

22–35)/CRT v. CI

fractions over

closed to

point: disease-

5FU 1 oxaliplatin as

33 days

recruitment

free survival at 3 years

radiosensitizer (days 1, 8, 22, 29) PETACC 6/

Capecitabine/

45 Gy in 25

Open

Capecitabine 36 or XELOX

Primary end-

EORTC 40054,

radiotherapy v.

fractions over

36; adjuvant chemotherapy

point: disease-

NCT00766155

capecitabine 1

33 days 6

regimen continues

free survival

oxaliplatin (every

5.4 Gy boost

preoperative agents

week)/radiotherapy

CI, continuous infusion; ECOG, Eastern Cooperative Oncology Group; FOLFOX, oxaliplatin with 5-fluorouracil and folinic acid chemotherapy; 5FU, 5-fluorouracil; XELOX, capecitabine with oxaliplatin.

Integration of biologicals Integration of targeted drugs such as cetuximab into preoperative chemoradiation schedules in rectal adenocarcinoma is attractive in principle. The preliminary results of chemoradiation clinical trials with cetuximab, on the early clinical end-point of pathological complete response, are disappointing however. Cetuximab can lead to G1 or G2/M-cell cycle arrest; if only a small proportion of cells within the tumour are affected, this decrease in proliferation could impact on the chance of achieving a complete pathological response. A large multinational randomized phase II study EXPERT-C (NCT00383695) has compared neoadjuvant therapy comprising oxaliplatin, capecitabine and chemoradiotherapy with or without cetuximab in 164 patients. The study was completed in July 2008, and results may throw more light on combinations of cetuximab and chemoradiation in the clinical setting in locally advanced rectal cancer. The combination may yet reproduce the improvement in long-term results achieved when cetuximab has been combined with radiation alone in studies of head and neck cancer.88 Also, in a randomized phase II trial of squamous cell carcinoma of the oesophagus, the overall response rate according to Response Evaluation Criteria In Solid

HEBK001-C06_p87-102.indd 97

Tumors (RECIST) criteria was 19 per cent versus 13 per cent for the addition of cetuximab to cisplatin and 5FU.89 Median progression-free survival was 9.5 months for cetuximab with cisplatin and 5FU and 5.5 months for cisplatin and 5FU, demonstrating that the response may not be the best end-point. The integration of bevacizumab has considerable preclinical rationale. Experimental studies in human tumour xenograft models have shown that vascular endothelial growth factor (VEGF) blockade serves as a potent and non-toxic enhancer of radiation. Anti-VEGF decreases interstitial pressure, increases oxygenation, and reverses radiation resistance conferred by hypoxia. Few phase II trials have been reported, but the combination with chemoradiotherapy appears potentially deliverable with acceptable toxicity. There are five other ongoing or recently closed phase III trials registered on the http://clinicaltrials.gov website.

THE FUTURE The best opportunity to improve survival in patients with rectal cancer will require continued focus on adjuvant chemotherapy strategies, since the development of metastases is the predominant cause of recurrence and death. At the same time, we need to

08/02/13 5:58 PM

98  Preoperative radiotherapy and chemoradiotherapy for rectal cancer

achieve better documentation and understanding of pursuing detailed assessment of acute and chronic toxicity and the risk of second malignancies. Additionally, if treatment strategies are to be successful in patients with rectal cancer, then a degree of individualization in the cytotoxic chemotherapy and biological agents – and even further individualization of radiation – is required. To optimize treatment, we need a deeper understanding of tumour biology.

CONCLUSION A pelvic MRI scan is an essential part of the initial staging of all patients with rectal cancer 0–15 cm from the anal verge and can help to define the most appropriate preoperative strategy. There is a wellestablished role for both SCPRT and down-staging long-course chemoradiation. For chemoradiation, however, there is no consensus regarding the optimum dose, fractionation, field size or chemotherapy schedule. Current evidence does not support the degree of response to chemoradiation (e.g. pathological complete response; down-sizing the primary tumour; sterilizing the regional nodes; tumour regression grades; residual cell density) as a valid surrogate for long-term local control or survival. Current trials suggest that in resectable cancers where the CRM is not threatened, SCPRT and chemoradiotherapy are equivalent in terms of outcomes such as local recurrence, disease-free survival and overall survival. In view of the high risk of metastatic disease, research strategies should be explored to address the problem in the preoperative setting. More rationally designed preclinical and translational studies (with recognized negative predictive factors such as k-ras mutations, b-raf mutations and epidermal growth factor receptor (EGFR) gene copy numbers) may therefore help select out inappropriate patients and determine the optimal sequence of chemotherapy and biological triple combinations. Only then can we move on to perform large randomized phase III trials.

Recommendations A pelvic MRI scan is an essential part of the initial staging of all patients with rectal cancer 0–15 cm from the anal verge.

l

HEBK001-C06_p87-102.indd 98

Current evidence does not support the degree of response to chemoradiation (e.g. pathological complete response; down-sizing the primary tumour; sterilizing the regional nodes; tumour regression grades; residual cell density) as a valid surrogate for long-term local control or survival, although down-staging may predict a role for adjuvant chemotherapy. l There is no evidence that conventional preoperative chemoradiation can facilitate sphinctersparing surgery. It should not be a primary reason for administering preoperative chemoradiation, although in the case of excellent downstaging this may prove a bonus. l There is no evidence for increased efficacy of chemoradiotherapy over SCPRT in moderaterisk resectable cancers. l Preoperative chemoradiation using 45–54 Gy over 5–6 weeks with an interval to allow response and shrinkage is recommended where the MRI confirms a ‘threatened’ or breached circumferential margin, or where there is a radiologically unresectable cancer with disease outside the mesorectum, i.e. surgery is not possible without leaving tumour within the pelvis. l

References   1. Thomas WH, Larson RA, Wright HK, Cleveland JC. Analysis of 830 patients with rectal adenocarcinoma. Surg Gynecol Obstet 1969; 129: 10–14.   2. Fisher B, Wolmark N, Rockette H, et al. Postoperative adjuvant chemotherapy or radiation therapy for rectal cancer: results from NSABP protocol R-01. J Natl Cancer Inst 1988; 80: 21–9.   3. Gastrointestinal Tumor Study Group. Prolongation of the disease-free survival in surgically treated rectal carcinoma. N Engl J Med 1985; 312: 1465–72.   4. Krook JE, Moertel CG, Gunderson LL, et al. Effective surgical adjuvant therapy for high-risk rectal carcinoma. N Engl J Med 1991; 324: 709–15.   5. Gunderson LL, Sargent DJ, Tepper JE, et al. Impact of T and N stage and treatment on survival and relapse in adjuvant rectal cancer: a pooled analysis. J Clin Oncol 2004; 22: 1785–96.   6. Greene FL, Stewart AK, Norton HJ. New tumor–node– metastasis staging strategy for node positive (stage III) rectal cancer: an analysis. J Clin Oncol 2004; 22: 1778–84.   7. Bleehen NM. Combined drug–radiation treatments: indications and clinical problems. Bull Cancer 1981; 68: 127–31.

08/02/13 5:58 PM

References  99   8. Tannock IF. Treatment of cancer with radiation and drugs. J Clin Oncol 1996; 14: 3156–74.   9. Stockholm Rectal Cancer Study Group. Preoperative short-term radiation therapy in operable rectal carcinoma: a prospective randomized trial. Cancer 1990; 66: 49–55. 10. Dahl O, Horn A, Morild I, et al. Low-dose preoperative radiation postpones recurrences in operable rectal cancer: results of a randomized multicenter trial in western Norway. Cancer 1990; 66: 2286–94. 11. Pahlman L, Glimelius B. Pre- or postoperative radiotherapy in rectal and rectosigmoid carcinoma: report from a randomized multicenter trial. Ann Surg 1990; 211: 187–95. 12. Sauer R, Becker H, Hohenberger W, et al. Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med 2004; 351: 1731–40. 13. Roh MS, Colangelo LH, O’Connell MJ, et al. Preoperative multimodality therapy improves disease-free survival in patients with carcinoma of the rectum: NSABP-R03. J Clin Oncol 2009; 27: 5124–30. 14. Quirke P, Dixon MF, Durdey P, Williams NS. Local recurrence of rectal cancer due to inadequate surgical resection. Lancet 1986; 2: 996–9. 15. Camma C, Giunta M, Fiorica F, Pagliaro L, Craxi A, ­Cottone M. Preoperative radiotherapy for resectable rectal cancer: a meta-analysis. JAMA 2000; 284: 1008–15. 16. Colorectal Cancer Collaborative Group. Adjuvant radiotherapy for rectal cancer: a systematic overview of 8507 patients from 22 randomised trials. Lancet 2001; 358: 1291–304. 17. Munro AJ, Bentley AHM. Adjuvant radiotherapy in operable rectal cancer: a systematic review. Semin Colon Rectal Surg 2002; 13: 31–42. 18. Wong R, Tandan V, De Silva S, Figueredo A. Preoperative radiotherapy and curative surgery for the management of localized rectal cancer. Cochrane Database Syst Rev 2007; (2): CD002102. 19. Moertel CG, Fleming TR, Macdonald JS, et al. ­Fluorouracil plus levamisole as effective adjuvant therapy after resection of stage III colon carcinoma: a final report. Ann Intern Med 1995; 122: 321–6. 20. Kuebler JP, Wieand HS, O’Connell MJ, et al. Oxaliplatin combined with weekly bolus fluorouracil and leucovorin as surgical adjuvant chemotherapy for stage II and III colon cancer: results from NSABP C-07. J Clin Oncol 2007; 25: 2198–204. 21. André T, Boni C, Navarro M, et al. Improved overall survival with oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment in stage II or III colon cancer in the MOSAIC trial. J Clin Oncol 2009; 27: 3109–16. 22. Aschele C, Pinto C, Rosati G, et al. Preoperative (FU)-based chemoradiation with and without weekly oxaliplatin in locally advanced rectal cancer: pathologic response analysis of the Studio Terapia Adjuvante Retto (STAR)-01 randomized phase III trial. J Clin Oncol 2009; 27: 804s.

HEBK001-C06_p87-102.indd 99

23. Gérard JP, Azria D, Gourgou-Bourgade S, et al. Comparison of two neoadjuvant chemoradiotherapy regimens for locally advanced rectal cancer: results of the phase III trial ACCORD 12/0405-Prodige 2. J Clin Oncol 2010; 28: 1638–44. 24. Chau I, Brown G, Cunningham D, et al. Neoadjuvant capecitabine and oxaliplatin followed by synchronous chemoradiation and total mesorectal excision in magnetic resonance imaging-defined poor-risk rectal cancer. J Clin Oncol 2006; 24: 668–74. 25. Fernández-Martos C, Pericay C, Aparicio J, et al. Phase II, randomized study of concomitant chemoradiotherapy followed by surgery and adjuvant capecitabine plus oxaliplatin (CAPOX) compared with induction CAPOX followed by concomitant chemoradiotherapy and surgery in magnetic resonance imagingdefined, locally advanced rectal cancer: Grupo cancer de recto 3 study. J Clin Oncol 2010; 28: 859–65. 26. Chua YJ, Barbachano Y, Cunningham D, et al. Neoadjuvant capecitabine and oxaliplatin before chemoradiotherapy and total mesorectal excision in MRI-defined poor-risk rectal cancer: a phase 2 trial. Lancet Oncol 2010; 11: 241–8. 27. Bosset JF, Collette L, Calais G, et al. Chemotherapy with pre-operative radiotherapy in rectal cancer. N Engl J Med 2006; 355: 1114–23. 28. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus irinotecan, fluorouracil and leucovorin for metastatic colorectal cancer. N Engl J Med 2004; 350: 2335–42. 29. Van Cutsem E, Labianca R, Bodoky G, et al. Randomized phase III trial comparing biweekly infusional fluorouracil/leucovorin alone or with irinotecan in the adjuvant treatment of stage III colon cancer: PETACC-3. J Clin Oncol 2009; 27: 3117–25. 30. Tol J, Koopman M, Cats A, et al. Chemotherapy, bevacizumab, and cetuximab in metastatic colorectal cancer. N Engl J Med 2009; 360: 563–72. 31. Folprecht G, Gruenberger T, Bechstein WO, et al. Tumour response and secondary resectability of colorectal liver metastases following neoadjuvant chemotherapy with cetuximab: the CELIM randomised phase 2 trial. Lancet Oncol 2010; 11: 38–47. 32. Janjan NA, Crane C, Feig BW, et al. Improved overall survival among responders to preoperative chemoradiation for locally advanced rectal cancer. Am J Clin Oncol 2001; 24: 107–12. 33. Read TE, Ogunbiyi OA, Fleshman JW, et al. Neoadjuvant external beam radiation and protectomy for adenocarcinoma of the rectum. Dis Colon Rectum 2001; 44: 1778–90. 34. Hiotis SP, Weber SM, Cohen AM, et al. Assessing the predictive value of clinical complete response to neoadjuvant therapy for rectal cancer: an analysis of 488 patients. J Am Coll Surg 2002; 194: 131–5. 35. Ruo L, Tikoo S, Klimstra D, et al. Long term prognostic significance of extended rectal cancer response to

08/02/13 5:58 PM

100  Preoperative radiotherapy and chemoradiotherapy for rectal cancer preoperative radiation and chemotherapy. Am Surg 2002; 236: 75–81. 36. Valentini V, Coco C, Picciocchi A, et al. Does downstaging predict improved outcome after preoperative chemoradiation for extraperitoneal locally advanced rectal cancer? A long term analysis of 165 patients. Int J Radiat Oncol Biol Phys 2002; 53: 664–74. 37. Gerard JP, Conroy T, Bonnetain F, et al. Preoperative radiotherapy with or without concurrent fluorouracil and leucovorin in T3–T4 rectal cancers: results of FFCD 9203. J Clin Oncol 2006; 24: 4620–25. 38. The Swedish Rectal Cancer Trial. Improved survival with preoperative radiotherapy in resectable cancer. N Engl J Med 1997; 336: 980–7. 39. Kapiteijn E, Marijnen CA, Nagtegaal JD, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001; 345: 638–46. 40. Braendengen M, Tveit KM, Berglund A, et al. Randomized phase III study comparing preoperative radiotherapy with chemoradiotherapy in nonresectable rectal cancer. J Clin Oncol 2008; 26: 3687–94. 41. Frykholm GJ, Pahlman L, Glimelius B. Combined chemo and radiotherapy vs radiotherapy alone in the treatment of primary non-resectable cancer of the rectum. Int J Radiat Oncol Biol Phys 2001; 50: 433–40. 42. Folkesson J, Birgisson H, Pahlman L, et al. Swedish Rectal Cancer Trial: long lasting benefits from radiotherapy on survival and local recurrence rate. J Clin Oncol 2005; 23: 5644–50. 43. Peeters K, van de Velde C, Leer J, et al. Late side effects of short-course preoperative radiotherapy combined with total mesorectal excision for rectal cancer: increased bowel dysfunction in irradiated patients – Dutch Colorectal Cancer Group study. J Clin Oncol 2005; 23: 6199–206. 44. Quirke P, Steele R, Monson J, et al. MRC CR07/NCICCTG CO16 Trial Investigators, NCRI Colorectal Cancer Study Group. Effect of the plane of surgery achieved on local recurrence in patients with operable rectal cancer: a prospective study using data from the MRC CR07 and NCIC-CTG CO16 randomised clinical trial. Lancet 2009; 373: 821–8. 45. Saif MW, Hashmi S, Zelterman D, et al. Capecitabine vs continuous infusion 5-FU in neoadjuvant treatment of rectal cancer: a retrospective review. Int J Colorectal Dis 2008; 23: 139–45. 46. De la Torre A, Garcia-Berical MI, Arrias F, et al. Preoperative chemoradiotherapy for rectal cancer: randomized trial comparing oral uracil and tegafur and oral leucovorin vs. intravenous 5-fluorouracil and leucovorin. Int J Radiat Oncol Biol Phys 2008; 17: 102–10. 47. Dahlberg M, Glimerlius B, Pahlman L. Preoperative radiation effect functional results after surgery for rectal cancer: results from a randomised study. Dis Colon Rectum 1998; 41: 543–9.

HEBK001-C06_p87-102.indd 100

48. Tepper JE, O’Connell, Nieddzwiecki D, et al. Adjuvant therapy in rectal cancer: analysis of stage, sex and local control – final report of Intergroup 0114. J Clin Oncol 2002; 20: 1744–50. 49. Gallagher MJ, Brereton HD, Rostock RA. A prospective study of treatment techniques to minimise the volume of pelvic small bowel with reduction of acute and late effects associated with pelvic irradiation. Int J Radiat Oncol Biol Phys 1986; 12: 1565–73. 50. Birgisson H, Pahlman L, Glimelius B. Adverse effects of preoperative radiation therapy for rectal cancer: long-term follow-up of the Swedish Rectal Cancer Trial. J Clin Oncol 2006; 23: 8697–705. 51. Birgisson H, Pahlman L, Gunnarsson U, Glimelius B. Occurrence of second cancers in patients treated with radiotherapy for rectal cancer. J Clin Oncol 2005; 23: 6126–31. 52. Bujko K, Nowacki MP, Nasierowska-Guttmejer A, et al. Sphincter preservation following preoperative radiotherapy for rectal cancer: report of a randomized trial comparing short-term radiotherapy vs. conventionally fractionated radiochemotherapy. Radiother Oncol 2004; 72: 15–24. 53. Ngan S. A randomized trial comparing local recurrence rates between short-course and long-course preoperative radiotherapy for clinical T3 rectal cancer: an intergroup trial (TROG, AGITG, CSSANZ, RACS). J Clin Oncol 2010; 28: 15s. 54. Boulis-Wassif S, Gerrard A, Loygue J, et al. Final results of a randomised trial on the treatment of rectal cancer with pre-operative radiotherapy alone or in combination with 5-fluoro-uracil followed by radical surgery. Cancer 1984; 53: 1811–8. 55. Francois Y, Nemoz CJ, Baulieux J, et al. Influence of the interval between preoperative radiation therapy and surgery on downstaging and on the rate of sphincter-sparing surgery for rectal cancer: the Lyon R90-01 Randomised Trial. J Clin Oncol 1999; 17: 2396–402. 56. Glehen O, Chapet O, Adham M. Long-term results of the Lyon R90-01 randomized trial of preoperative radiotherapy with delayed surgery and its effect on sphincter-saving surgery in rectal cancer. Br J Surg 2003; 90: 996–8. 57. Gerard JP, Chapet O, Nemoz C, et al. Improved sphincter preservation in low rectal cancer with high dose preoperative radiotherapy: the Lyon R96-02 randomized trial. J Clin Oncol 2004; 22: 2404–9. 58. Wagman R, Minsky BD, Cohen AM, et al. Sphincter preservation in rectal cancer with preoperative radiation therapy and coloanal anastomosis: long term follow-up. Int J Radiat Oncol Biol Phys 1998; 42: 51–7. 59. Rouanet P, Fabre JM, Dubois JB, et al. Conservative surgery for low rectal carcinoma after high-dose radiation: functional and oncologic results. Ann Surg 1995; 221: 67–73.

08/02/13 5:58 PM

References  101 60. Grann A, Feng B, Wong M, et al. Preoperative combined modality therapy for clinically resectable uT3 rectal adenocarcinoma. Int J Radiat Oncol Biol Phys 2001; 49: 987–95. 61. Kuvshinoff B, Maghfoor I, Miedema B, et al. Distal margin requirements after preoperative chemoradiotherapy for distal rectal carcinomas: are , or = 1 cm distal margins sufficient? Ann Surg Oncol 2001; 8: 163–9. 62. Sauer R, Fietkau R, Wittekind C, et al. Adjuvant vs. neoadjuvant radiochemotherapy for locally advanced rectal cancer: the German trial CAO/ARO/AIO-94. Colorectal Dis 2003; 5: 406–15. 63. Gunderson LL, Sosin H. Areas of failure found at reoperation (second or symptomatic look) following ‘curative surgery’ for adenocarcinoma of the rectum: clinicopathologic correlation and implications for ­adjuvant therapy. Cancer 1974; 34: 1278–92. 64. Minsky B. Pelvic radiation therapy in rectal cancer: technical considerations. Semin Radiat Oncol 1993; 3: 42–7. 65. Bagatzounis A, Kolbl O, Muller G, et al. The locoregional recurrence of rectal carcinoma: a computed tomographic analysis and a target volume concept for adjuvant radiotherapy. Strahlenther Onkol 1997; 173: 68–75. 66. Minsky BD. Rectal cancer. In Leibl SA, Phillips TL (eds). Textbook of Radiation Oncology. Philadelphia, PA, WB Saunders, 1998: 686–71. 67. Lorchel F, Maingon P, Crehange G, et al. Cancer du rectum: volumes cibles de la radiotherapie preoperatoire. Cancer Radiother 2002; 6(suppl. 1): 93–9s. 68. Arcangeli S, Valentini V, Nori SL, et al. Underlying anatomy for CTV contouring and lymphatic drainage in rectal cancer radiation therapy. Rays 2003; 28: 331–6. 69. Myerson R, Drzymala R. Technical aspects of imagebased treatment planning of rectal carcinoma. Semin Radiat Oncol 2003; 13: 433–40. 70. Gunderson LL, Haddock MG, Gervaz PA. Clinical target volume in conformal and intensity modulated radiation therapy: a clinical guide to cancer treatment. In Grégoire V, Scalliet P, Ang KK (eds). Rectal and Anal Cancer in Conformal Radiotherapy Planning: Selection and Delineation of Lymph Node Areas. Heidelberg, Springer-Verlag, 2003: 187–97. 71. Portaluri M, Bambace S, Perez C, et al. Clinical and anatomical guidelines in pelvic cancer contouring for radiotherapy treatment planning. Cancer Radiother 2004; 8: 222–9. 72. Höcht S, Hammad R, Thiel HJ. Recurrent rectal cancer within the pelvis: a multicenter analysis of 123 patients and recommendations for adjuvant radiotherapy. Strahlenther Onkol 2004; 180: 15–20. 73. Marsh PH, James RD, Schofield PF. Adjuvant preoperative radiotherapy for locally advanced rectal carcinoma: results of a prospective randomized trial. Dis Colon Rectum 1994; 37: 1205–14.

HEBK001-C06_p87-102.indd 101

74. Fortier GA, Constable WC, Meyers H, et al. Preoperative radiation therapy for rectal cancer. Arch Surg 1986; 121: 1380–4. 75. Chan AK, Wong AO, Langevin J, et al. Preoperative chemoradiotherapy and pelvic radiation for tethered or fixed rectal cancer: a phase II dose escalation study. Int J Radiat Oncol Biol Phys 2000; 48: 843–56. 76. Ahmad NR, Marks G, Mohiuddin M. High-dose preoperative radiation for cancer of the rectum: impact of radiation dose on patterns of failure and survival. Int J Radiat Oncol Biol Phys 1993; 27: 773–8. 77. Mohiuddin M, Winter K, Mitchell E, et al. Randomized phase II study of neoadjuvant combined-modality chemoradiation for distal rectal cancer: Radiation Therapy Oncology Group Trial 0012. J Clin Oncol 2006; 24: 650–5. 78. Wiltshire KL, Ward IG, Swallow C, et al. Preoperative radiation with concurrent chemotherapy for resectable rectal cancer: effect of dose escalation on pathologic complete response, local recurrence-free survival, disease-free survival, and overall survival. Int J Radiat Oncol Biol Phys 2006; 64: 709–16. 79. Widder J, Herbst F, Dobrowsky W, et al. Preoperative short-term radiation therapy (25 Gy, 2.5 Gy twice daily) for primary resectable rectal cancer (phase II). Br J Cancer 2005; 92: 1209–14. 80. Brooks S, Glynne-Jones R, Novell R, et al. Short course continuous hyperfractionated accelerated radiotherapy (CHART) as pre-operative treatment for rectal cancer. Acta Oncol 2006; 45: 1079–85. 81. Coucke PA, Notter M, Matter M, et al. Effective timing of surgery on survival after preoperative hyperfractionated accelerated radiotherapy (HART) for locally advanced rectal cancer (LARC): is it a matter of days? Acta Oncol 2006; 45: 1086–93. 82. Guckenberger M, Wulf J, Thalheimer A, et al. Prospective phase II study of preoperative short-course radiotherapy for rectal cancer with twice daily fractions of 2.9 Gy to a total dose of 29 Gy: long-term results. Radiat Oncol 2009; 4: 67. 83. Kamikonya N, Hishikawa Y, Kurisu K, Taniguchi M, Miura T. Primary rectal cancer treated with high-doserate intraluminal brachytherapy following external radiotherapy. Radiat Med 1991; 9: 85–7. 84. Vuong T, Belliveau PJ, Michel RP, et al. Conformal preoperative endorectal brachytherapy treatment for locally advanced rectal cancer: early results of a phase I/II study. Dis Colon Rectum 2002; 45: 1486–93. 85. Kaufman N, Nori D, Shank B, et al. Remote afterloading intraluminal brachytherapy in the treatment of rectal, rectosigmoid, and anal cancer: a feasibility study. Int J Radiat Oncol Biol Phys 1989; 17: 663–8. 86. Jakobsen A, Mortensen JP, Bisgaard C, Lindebjerg J, Hansen JW, Rafaelsen SR. Preoperative chemoradiation of locally advanced T3 rectal cancer combined with an endorectal boost. Int J Radiat Oncol Biol Phys 2006; 64: 461–5.

08/02/13 5:58 PM

102  Preoperative radiotherapy and chemoradiotherapy for rectal cancer 87. Corner C, Bryant L, Chapman C, Glynne-Jones R, Hoskin PJ. High-dose-rate afterloading intraluminal brachytherapy for advanced inoperable rectal carcinoma. Brachytherapy 2010; 9: 66–70. 88. Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus cetuximab for squamous cell carcinoma of the head and neck. N Engl J Med 2006; 354: 567–78.

HEBK001-C06_p87-102.indd 102

89. Lorenzen S, Schuster T, Porschen R, et al. Cetuximab plus cisplatin–5-fluorouracil versus cisplatin–5-fluorouracil alone in first-line metastatic squamous cell carcinoma of the esophagus: a randomized phase II study of the Arbeitsgemeinschaft Internistische Onkologie. Ann Oncol 2009; 20: 1667–73.

08/02/13 5:58 PM

7 Total mesorectal excision for rectal cancer Brendan Moran

Introduction There have been several key advances in the optimal management of rectal cancer, but none so important as standardization and improvement in the appropriate surgical procedure. In addition to surgical advances, neoadjuvant and adjuvant radiotherapy and chemoradiotherapy, preoperative imaging (by computed tomography (CT) of the chest and abdomen looking for metastatic disease, and by magnetic resonance imaging (MRI) to assess local tumour extent), pathological assessment and audit have all contributed to better results for this complex but eminently curable cancer. These advances, and complex management decisions, are undoubtedly best coordinated by a multidisciplinary team approach, but surgery, and the surgeon, are the key to optimal decision-making and outcome. Multidisciplinary team management should be individualized to each patient with rectal cancer and should focus on helping the surgeon and the patient choose the best course of action. The preoperative MRI provides optimal local staging and is a key visual component in the multidisciplinary team discussion of the optimal management of a patient with rectal cancer. The combination of clinical examination and good-quality MRI facilitates selection for neoadjuvant (preoperative) radiotherapy or chemoradiotherapy. This decision is crucial

HEBK001-C07_p103-123.indd 103

and requires a balance between the desired effect on the tumour and the immediate and long-term side effects associated with radiotherapy for rectal cancer. The benefits of neoadjuvant therapy for locally advanced tumours are now generally accepted, with good evidence for both down-staging and downsizing after optimal neoadjuvant therapy. Debate persists as to what the term ‘locally advanced’ means, and until recently the definition lacked the objective fine detail of good-quality MRI. The evidence for neoadjuvant therapy is strongest for a reduction in local recurrence in operable rectal cancer and has been reported in a number of randomized controlled trials such as the Swedish, Stockholm, Dutch and UK CR07 trials. There is little evidence of a survival benefit from radiation in these trials, however, with well-documented reports of immediate and long-term side effects associated with radiotherapy. Unquestionably, tissue healing is impaired by radiotherapy, with an increase in perineal wound failure in patients who have an abdominoperineal excision and an increase in anastomotic leakage when reconstruction by anterior resection is performed. Additionally, in the medium to long term, bladder, bowel and sexual function are impaired by radiotherapy. There are also increasing reports of long-term complications such as an increase in second primary tumours, excess

11/02/13 3:51 PM

104  Total mesorectal excision for rectal cancer

cardiovascular deaths, and reoperation for bowel obstruction, all of which emphasize the need for a selective approach. There is no doubt that patients with advanced tumours benefit from radiotherapy and, if used in rectal cancer, radiotherapy should be given before surgery with randomized trials supporting this strategy. There is now general agreement that the optimal surgical treatment involves the concept of total mesorectal excision (TME), and that an R0 excision (all margins of the excised specimen free of tumour at detailed pathological assessment) is crucial. For this reason, the relationship of the tumour to the mesorectal fascia is key in that a tumour invading or breaching the mesorectal fascia will have an involved margin at TME surgery and will not be cured by TME alone. A key advance of staging MRI is the ability to visualize the mesorectal fascia and to accurately determine whether the mesorectal fascia is clear of tumour, definitely involved or possibly involved (the latter often categorized as ‘threatened’) by tumour. This crucial information on the relationship of the tumour to the mesorectal fascia allows an informed decision on whether TME alone is likely to be curative (clear margins), indicates patients who definitely need neoadjuvant therapy (involved margin), and facilitates discussion on the risks and benefits in patients with a threatened margin. These categories are more difficult to define in the low rectum (at or below the level of the levators), where the mesorectum tapers out and the margins are more likely to be threatened. For this reason, we have an ongoing interest, research programme and educational initiative in the staging and management of low rectal cancers (see www.lorec.nhs.uk). The broad principles of TME surgery remain the same, whether the technique is open surgery, laparoscopic TME or robotic TME. The principles of optimal visualization and more precise dissection, improved haemostasis and less collateral damage by diathermy are key. Modern advances in technology have helped in these areas. Knowledge of the anatomical structures, and their relationships in the pelvis, are essential for surgeons, regardless of the surgical technique used. It is noteworthy that surgical observations and attention to the fine detail of the anatomical structures at surgery have been a defining force in developing,

HEBK001-C07_p103-123.indd 104

refining and advancing the complex anatomy of the human pelvis (see Chapter 2).

Fundamental principles of total mesorectal excision surgery The fundamental components of TME surgery incorporate a number of principles and include the following: Peri-mesorectal ‘holy plane’ sharp dissection by diathermy and scissors under direct ­v ision: three-directional traction and counter-­ traction is a vital principle for diathermy dissection, as it is essential that the areolar tissue is ‘on stretch’ if holy plane dissection is to be accurate. l Specimen-oriented surgery and histopathology, with the objective of an intact mesorectum with no tearing of the surface and no circumferential resection margin (CRM) or distal margin involvement (naked eye or microscopic). l Personal naked-eye assessment as audit for obvious CRM involvement as the principal immediate outcome measure: this should be combined with objective assessment of the whole specimen by the pathologist, as this confirms the optimal planning and completion of the surgery. Surgeons and oncologists may also base postoperative therapy on this pathology report, although there must always be a sense of the ‘horse having bolted’ if cancer has perforated through the resected specimen and radiotherapy or chemoradiotherapy (CRT) was not given before surgery. An involved margin generally represents a failure of management planning or surgical technique, or both. l Recognition during surgery and preservation of the autonomic plexuses and nerves on which sexual and bladder function depend. l A major increase in anal preservation and reduction in the number of permanent colostomies by skilful extension into the depths of the pelvis. l Stapled low pelvic reconstruction, usually using the Moran triple stapling technique (described later), plus creation of a short colon pouch or a side-to-end anastomosis to the low rectum or anal canal. l

11/02/13 3:51 PM

BASIC PRINCIPLES OF SURGERY FOR RECTAL CANCER  105

BASIC PRINCIPLES OF SURGERY FOR RECTAL CANCER Rectal cancer, as defined by an adenocarcinoma with its lower edge at or within 15 cm of the anal verge, is a common cancer and accounts for approximately 30 per cent of all colorectal cancers. Surgical excision by anterior resection is curative in most cases. The term ‘anterior resection’ defines an operation whereby the inferior mesenteric artery has been ligated and reconstruction is by anastomosis of the proximal colon to the distal remnant. Anterior resection is neither required nor appropriate for all patients with rectal cancer. Early tumours (perhaps in the region of 5–10 per cent of all patients with rectal cancer, although this percentage may increase with screening) can be treated by local excision alone in selected patients (see Chapter 11), and tumours involving the anal sphincter complex or levator muscle (approximately 10–15 per cent) require excision of the sphincter complex with a permanent stoma (see Chapter 8). Thus, restorative anterior resection is required and feasible in approximately 70–80 per cent of all patients with rectal cancer. For each patient, however, the feasibility and benefits of restorative resection will ultimately depend on the tumour, the patient and, to a lesser extent, the surgeon. Tumour-dependent factors are predominantly the distance of the tumour from the anal verge, the fixity of the tumour, and the presence or absence of diffuse metastatic disease. Patient-dependent factors include the body habitus, the size and depth of the pelvis, and anal sphincter integrity and function. Surgeon-dependent factors include the availability of resources (e.g. stapling instruments, adequate retractors) and the availability of a suitably experienced surgical assistant and theatre team. The principles of anterior resection for rectal cancer currently revolve around what has been described as ‘circumferential awareness’ incorporating circumferential staging, circumferential down-staging, circumferential surgery and circumferential pathology. Circumferential staging incorporates clinical examination and, more latterly, the addition of cross-sectional imaging, in particular abdominal and pelvic CT and pelvic MRI. Circumferential down-staging or down-sizing is achieved by preoperative radiotherapy or chemoradiotherapy (neoadjuvant therapy; see Chapter 6). The concept of circumferential surgery has emanated from

HEBK001-C07_p103-123.indd 105

Figure 7.1. Inferior mesenteric angiogram outlining the blood supply to the rectum and demonstrating the absence of any significant arterial input from the lateral pelvic side-wall.

the description and popularization of TME with specimen-oriented surgery. Finally, circumferential pathology reports on the adequacy of excision margins on the resected specimen and allows continuous quality control of the staging, MDT decisions and surgical precision. The principles of optimal surgery for rectal cancer revolve around the embryology and anatomy of the rectum, whereby the lymphatic drainage (which is associated with the arterial blood supply; Figure 7.1) is almost exclusively proximal and is generally confined within the mesorectal fascia. The principles of TME focus on specimen-oriented surgery, whereby completeness and intactness of the specimen are crucial factors, such that ideally it should be one recognizable block of tissue whose orientation and former relations can be identified. In most cases, naked-eye inspection provides the initial necessary quality control. Visual inspection of the front of a well-performed TME specimen should show three clear landmarks: the cut edge of the peritoneal reflection; the smooth shiny anterior surface of the anterior mesorectum of the middle third (the rectogenital septum); l the almost bare anterior aspect of the anorectal muscle tube in the lowest anterior resections. l l

11/02/13 3:51 PM

106  Total mesorectal excision for rectal cancer

Figure 7.2. Total mesorectal excision specimen illustrating the shiny mesorectal fascia surrounding the fatty mesorectum. A linear stapling gun occludes the lumen, as used in the Moran triple stapling technique.

Laterally, the fatty mesorectum expands distally beyond an anteroposterior groove made by the nervi erigentes so that an embryologically perfect specimen has a lateral dilatation distally, corresponding with the part related to the inside of the levator muscles beyond their origins from the pelvic side-wall (Figure 7.2). Posteriorly, a perfect specimen exhibits perfectly curved ‘buttocks’ with a central midline groove corresponding to the anococcygeal raphe. Distal mesorectal spread of the tumour rarely extends more than 2–3 cm beyond the lower palpable luminal edge of the tumour, although for safety reasons a distal mesorectal clearance of 5 cm, where feasible and available, is recommended. This was the fundamental principle underpinning the concept of TME, which has now become accepted as the optimal surgical procedure for patients with resectable mid- and low rectal cancer. There has been some confusion that we have advocated TME for all rectal cancers; but TME is not necessary for tumours of the upper rectum provided the mesorectum and muscle tube can be divided 5 cm distal to the lower edge of the tumour and a mesorectal transection, rather than TME, can be performed, possibly diminishing the postoperative surgical complications such as a reduction in the anastomotic leak rate. Other major advances facilitating safe restoration of continuity, even in low tumours, were the development and widespread availability of sta-

HEBK001-C07_p103-123.indd 106

pling instruments,1 the recognition that adenocarcinoma rarely spreads distally in the muscle tube, and the recognition that a 2 cm clearance beyond the macroscopic tumour provides a safe distal margin and even less than 1 cm may be adequate for ultra-low tumours.2 It is important to recognize that, even at its most scientific, surgery is primarily a craft with very special challenges and rewards for excellence. Surgery is not amenable to study by the same methods as those applicable to drugs or radiotherapy treatments, and yet surgical technique has by far the greatest impact on rectal cancer outcomes. The randomized controlled clinical trial has so far contributed remarkably little to the development of surgical technique in colorectal surgery or most other surgical specialties. Even though variations in technique, such as open, laparoscopic or robotic surgery, lend themselves to comparative studies and indeed to randomization, in all such trials the cohort studied can be only a selected subset of the patient population, with the most difficult patients excluded and operated on by open surgical techniques. For this reason, the results of such trials are not applicable to all patients and merely represent comparative outcomes, generally in the more favourable cases. The technique of TME has never been critically evaluated by a randomized trial, and nor is likely to be. In this difficult and complex surgery for rectal cancer, the devil is in the detail, and attention to detail using TME concepts results in optimal ­outcome.

Preoperative Assessment, Planning and Preparation Preoperative preparation is essential, and to fail to plan is to plan to fail. The assessment of a rectal cancer, local staging and management of a rectal cancer are summarized in Figures 7.3 and 7.4, which, although specific for low rectal cancer, apply to any rectal adenocarcinoma. Critical decisions in lower-third cancers

One focal area of current interest centres on the anatomical and embryological fact that the mesorectal envelope tapers down in this (infralevator) lower third to appear very thin indeed, particularly on the crucial coronal oblique MRI cuts on

11/02/13 3:51 PM

BASIC PRINCIPLES OF SURGERY FOR RECTAL CANCER  107

Suspicion of Rectal Cancer

History, patients perspective and Rectal Examination (PR) by Experienced Surgeon + CNS ideally present

Too painful

PR

Rigid Sigmoidoscopy

Mobility

Distance in cms from anal verge to lower edge

EUA (Examination Under Anaesthetic)

Distance in cms from dentate line to lower edge

Colonoscopy Biopsy Surgeon and CNS Squamous Carcinoma

Treatment

Adenocarcinoma

Staging

Adenoma

Other: Prostate/Carcinoid

Repeat biopsy

Treatment

Systemic Staging CT - Chest and Abdomen

Disseminated disease

Resectable metastases

No metastases

Local Extent primary (Figure 7.5)

Pallative treatment ± Chemotherapy ± Radiotherapy ± Local Ablation CNS, clinical nurse specialist; CT, computed tomography. Figure 7.3. Suggested algorithm for assessment of rectal cancer.

HEBK001-C07_p103-123.indd 107

11/02/13 3:51 PM

108  Total mesorectal excision for rectal cancer Low Rectal Cancer

PR by experienced surgeon

MRI - Low rectal protocol

EOUS (Endo-anal ultrasound)

MDT ‘Good’ (early tumour uT1, mrT1)

‘Bad’ (mrT3N1-2, EMVI)

‘Ugly’ (Advanced uT3/T4, mr T3/T4)

SCRT/CRT?

CRT Restage Systemic (CT)

PR, MRI, EAUS Tumour present

Recurrence

Surgery Local excision perianal, TEM

Anterior Resection (AR) (Including Intersphincteric) Defunctioning Stoma

Intensive Follow-up

Abdomino Perineal Excision (APE) Intersphinteric

Intensive Follow-up

5–10 years

Extralevator ELAPE Perineal closure

Primary Suture

Primary Suture

Clinical and radiological complete response Consider Deferral of surgery ‘watch and wait’

Biological mesh

Omentum

Muscle flap

Back to MDT for post op discussion

CRT, chemoradiotherapy; CT, computed tomography; EAUS, endoanal ultrasound; MDT, multidisciplinary team; MRI, magnetic resonance imaging; PR, per rectal examination; SCRT, short-course radiotherapy; TEM, transanal endoscopic microsurgery. Figure 7.4. Suggested algorithm for local assessment of low rectal cancer.

which decisions in modern multidisciplinary teams are made. On such an MRI it is tempting to predict that this tapering and narrowing area of the mesorectum will constitute a hazardous margin; thus, a decision may be made to administer preoperative down-staging neoadjuvant therapy or even choose abdominoperineal excision for fear of margin involvement when in fact a carefully oriented axial oblique sequence with the axial cuts precisely at right-angles to the tumour segment may demonstrate a potentially safe clearance. In such cases, where the cancer is below the levator origins on MRI, it is essential that an experienced surgeon examines the patient to establish free mobility of

HEBK001-C07_p103-123.indd 108

the tumour in the conscious patient (with muscle tone). In the author’s opinion, this clinical observation of mobility on the sphincter complex and adjacent organs almost invariably means that a TME will be an achievable surgical objective (see Figure 7.5). It does not get away from the other issues of the higher incidence of internal iliac and particularly obturator node involvement in tumours less than 6 cm from the anal margin.3 A rectal neoplasm should be assessed by an experienced clinician performing a rectal examination and rigid sigmoidoscopy. The height of the lower edge of the tumour should be measured by rigid sigmoidoscopy with the patient awake and recum-

11/02/13 3:51 PM

BASIC PRINCIPLES OF SURGERY FOR RECTAL CANCER  109

pelvic anatomy, such as ovarian or uterine pathology in female patients and prostatic enlargement in male patients, all of which may limit access. In some patients, synchronous treatment of ovarian and rectal pathology may be performed. The clinical details and radiological investigations of all patients with rectal cancer should ideally be discussed at a colorectal multidisciplinary team meeting to optimize the management strategy, which in some cases includes neoadjuvant therapy. Patient consent and immediate preoperative care

Figure 7.5. Total mesorectal excision. Dissection follows the dotted line. Transection of the muscle tube should be 3–5 cm below the luminal distal edge of the tumour as distal mesorectal spread has been documented so that mesorectal transection may be adequate for upper rectal tumours.

bent in the left lateral position. The surgeon should assess the integrity of the anal canal and if possible assess the mobility, or otherwise, of a palpable tumour. The neoplasm should be biopsied to confirm the diagnosis. Biopsy may be deferred to a subsequent urgent colonoscopy, however. One useful procedure is to re-examine the lesion, particularly a low rectal cancer under sedation at colonoscopy. Occasionally an examination and biopsy under general anaesthetic is required, and even today this may be a useful addition in the work-up of a patient with a rectal cancer. The remainder of the colon should be assessed to look for synchronous neoplasia (present in 3–4 per cent of patients) by colonoscopy, CT colonography or barium enema. Colonoscopy has the advantage of allowing biopsy or removal of synchronous lesions such as polyps. The benefits of CT colonography are the ability to stage the abdominal cavity and to outline the proximal bowel in patients with stenotic lesions where colonoscopy may not be ­feasible. Current best practice also incorporates staging for systemic disease by chest and abdominal CT scan and local pelvic staging of the rectal tumour, ideally by MRI to assess the relationship of the tumour to the mesorectal fascia (see Chapter 4 and 5). Additionally, MRI can provide details of relevant

HEBK001-C07_p103-123.indd 109

Medical conditions such as diabetes, hypertension, and cardiac and pulmonary diseases should be ­optimized. Arrangements should be in place for postoperative critical care if required. The patient should be consented and sites marked in case a stoma is required. It is prudent to consent for a permanent stoma, in addition to a temporary defunctioning, should an unexpected event render this necessary. In addition to the complications after any intestinal resection, such as leakage and haemorrhage, the possibility of sexual and bladder dysfunction needs to be documented in the consent form for rectal cancer surgery. These latter complications usually result from injury to the pelvic autonomic nerves. In female patients, the possible need for oophorectomy should be discussed. Consent for oophorectomy should be documented in case the tumour involves the ovaries or the patient has a family history of ovarian cancer or requests oophorectomy. A personal preference is also to seek consent for appendicectomy to treat synchronous or prevent metachronous appendiceal pathology. Unlike current practice in surgery for colon cancer, an empty large intestine is desirable for restorative rectal cancer surgery, particularly if there is a need for temporary defunctioning. An ileostomy proximal to a full colon may not reduce the consequences of an anastomotic leak. Optimal bowel preparation can be achieved by combining clear fluids by mouth for 48 h before surgery with oral laxatives. There is emerging evidence that the non-prepared colon increases complications after ­restorative rectal cancer surgery, and many surgeons who abandoned bowel preparation on the back of experiences with colonic resection are reverting back to full mechanical bowel preparation for restorative rectal cancer surgery.

11/02/13 3:51 PM

110  Total mesorectal excision for rectal cancer

Systemic antibacterial agents, including anaerobic cover such as with metronidazole, are given at induction of anaesthesia and continued for 24 h post-surgery. Deep venous thrombosis (DVT) prophylaxis is commenced by a combination of heparin or its analogues (depending on the use of epidural anaesthesia). The risk of DVT must always be balanced against the risk of bleeding, which is catastrophic if into the dural space and potentially catastrophic during pelvic dissection. For these reasons, anticoagulation is best withheld in the immediate preoperative period. The use of mechanical calf compression devices, once the patient is positioned on the operating table, and in the postoperative period, has been shown to be effective and generally safe.

Patient Positioning The lithotomy–Trendelenburg position is optimal as it allows peranal palpation, inspection and washout, together with insertion of the circular staple gun to complete the anastomosis. Additionally, a second assistant can stand between the legs. The patient is kept horizontal during the abdominal phase of the operation and can then be tilted head down by 15–208 or more to facilitate the pelvic dissection. It is important not to maintain steep Trendelenburg positioning for extended periods, to reduce the risks of calf compartment syndrome. Good lighting is essential. Optimal lighting can be obtained by readjusting the movable focused operating lights during different phases, using a headlight (which many surgeons find irksome for a prolonged procedure), and using retractors with integrated lights.

On-Table Examination, Skin Incision and Abdominal Exploration A rectal examination before painting and draping is mandatory and must be supplemented by a vaginal examination in female patients. A long vertical midline incision provides optimal access to the abdomen and pelvis and may need to be extended from the pubic symphysis to the xiphisternum in patients who are overweight (Figure 7.6).

HEBK001-C07_p103-123.indd 110

Figure 7.6. Midline incision extending from the symphysis pubis, which may need to be extended to the xiphisternum. The optimal site for a defunctioning stoma is on the right side (defunctioning lower right ileostomy) and is marked preoperatively by an enterostomal therapist. A site is also marked in the left iliac fossa in case a permanent stoma is deemed necessary.

Although some advocate a long transverse incision or even a modified extended Pfannensteil incision, access to the splenic flexure and pelvis is inferior compared with a midline incision. The abdominal cavity is fully palpated, with particular attention directed to the liver and spleen, greater omentum, stomach and small bowel, and the entire colorectum, including the appendix. The surgical procedure is then planned, including the sequence. For example, if a low anterior resection is planned mobilization of the splenic flexure is almost always needed. Personal preference is to complete this manoeuvre at the beginning while still fresh and to avoid the temptation to omit this at the end of a long procedure and thus compromise on tension and blood supply to the neorectum.

Commencing the Dissection and Splenic Flexure Mobilization The operating surgeon stands on the left side of the patient. The assistant on the right lifts the sigmoid

11/02/13 3:51 PM

BASIC PRINCIPLES OF SURGERY FOR RECTAL CANCER  111

colon anteriorly and to the patient’s right. The peritoneal reflection on the left side of the colon (the white line of Toldt) is identified and divided by scissors, or more commonly diathermy, and followed cranially towards the splenic flexure. The plane in the left upper quadrant between the colon and the urogenital structures (Gerotas fascia surrounding the kidney and the gonadal vessels) is developed. At this juncture, if the spleen is mobile on the diaphragm, a large moist swab placed gently between the spleen and diaphragm helps to push the spleen into view and facilitates splenic flexure mobilization (Figure 7.7). The greater omentum is now retracted anteriorly and to the patient’s left, and the bloodless plane between the transverse colon and omentum is developed by sharp scissors dissection or diathermy incision. The apex of the splenic flexure attachments are visualized by downwards colonic traction from the right side of the patient with counter-traction by a retractor under the left ribcage. The assistant on the patient’s right, in addition to colonic traction, insinuates a finger behind the colon on the left. Division of the apical lateral colonic attachments is performed by an operator who stands either on the patient’s left or temporarily between the patient’s legs.

Figure 7.7. Mobilization of the splenic flexure, incising lateral attachments.

HEBK001-C07_p103-123.indd 111

Ligation and division of the inferior mesenteric vessels

The left-sided colonic mobilization is continued inferiorly by the left-sided operator identifying the ureter (usually positioned medial to the gonadal vessels and crossing the bifurcation of the common iliac artery) and the fascial covering of the uppermost part of the mesorectal package. This manoeuvre is facilitated by the right-sided assistant applying traction on the sigmoid, anteriorly and to the right, taking care not to damage the mesentery of the colon. Once the plane has been developed at the pelvic brim to just beyond the midline, a small swab is pushed behind the colon and rectal mesentery at the level of the pelvic brim. The sigmoid traction is reversed and the surgeon on the patient’s right can identify the exact place to incise the right-sided peritoneum by a combination of air in the tissues and displacement of the mesentery anteriorly. The swab helps to protect the autonomic nerves at the level of the pelvic brim by displacing the colonic mesentery anteriorly. The right-sided peritoneum is incised caudally to the pelvic brim and cranially towards the root of the inferior mesenteric artery. At this point, the surgeon on the patient’s left places the left index finger behind the pedicle, with left thumb anteriorly, palpating the vessel between index finger and thumb. The peritoneal attachments are divided and pre-aortic nerve structures mobilized away from the right side of the pedicle by sharp dissection. The index finger is then advanced cranially on the left side, parallel to the midline, where a window will be identified above the origin of the inferior mesenteric artery (IMA) between the aorta and the inferior mesenteric vein (IMV) and ascending left colic artery running side by side at this point. This window is opened, and the autonomic nerves are freed until the root of the IMA is clearly identified. It is important to check that the left ureter has not been elevated in this manoeuvre by visualizing the structures to the left of the pedicle. Once the IMA pedicle has been isolated, it is clamped, divided and ligated approximately 2 cm from the aorta to reduce injury to the pre-aortic nerves and to achieve a high but not flush tie of the IMA (Figure 7.8). For maximum length and mobility of the left colon, the IMV has to be divided above its last branch, at the inferior border of the pancreas, where it disappears upwards to join the splenic vein. In 5–10 per cent of patients, a substantial branch of

11/02/13 3:51 PM

112  Total mesorectal excision for rectal cancer

Figure 7.8. Ligation of the inferior mesenteric artery (large arrow) and high ligation of the inferior mesenteric vein (small arrow).

the superior mesenteric artery lies near the IMV at this point, supplying part of the arterial blood supply to the left colon. Judgement is required to determine whether this vessel should be divided to provide sufficient length or preserved if it is likely to be important for colonic viability.

MOBILIZATION OF THE MESORECTUM AND RECTUM The pelvic dissection is oncologically one of the most important stages of the operation. The surgeon must develop a mental picture of the exact position and extent of the tumour, based on the prior clinical and radiological assessment. The circumferential concepts of TME surgery are applied to ensure clear margins on the resected specimen. It is helpful to divide the descending colon well above the cancer at this stage, a so-called ‘division of convenience’ using a linear cutting stapler such as a GIA 60 (Figure 7.9). This particularly facilitates the posterior pelvic dissection. The ‘division of convenience’ allows optimal mobility of the top of the specimen and facilitates

HEBK001-C07_p103-123.indd 112

Figure 7.9. The colon has been divided using a linear cutting stapler and the posterior plane is being developed anterior to the superior hypogastric plexus and hypogastric (presacral) nerves.

gentle opening of the perimesorectal planes by traction and counter-traction in any direction throughout the pelvic dissection. Having divided the colon, the small bowel and right transverse colon can be packed away, upwards and to the right. Personal preference is to place a large pack under the root of the caecum wrapping the small bowel and place a gauze roll beneath the root of the mesentery. An extendable self-retaining retractor (we generally use a Finochetti chest retractor with the largest retractor blades) then opens up the space and allows excellent views into the top of the pelvis.

Posterior Dissection – Starting ‘Right’ The pelvic dissection for a rectal cancer is a dynamic procedure and involves circumferential disection commencing posteriorly, continuing laterally and anterolaterally and then anteriorly, and constantly moving the site of dissection rather than proceeding too far at one site. The easiest place is commonly posteriorly, and commencement in this area is crucial. For clarity the dissection sites are covered separately, although in reality it is a dynamic procedure

11/02/13 3:51 PM

MOBILIZATION OF THE MESORECTUM AND RECTUM  113

requiring alteration in the placement and direction of traction of the retractors. The ‘pedicle package’ – the clue to the top of the ‘holy plane’ – involves the key principle that dissection should proceed only in the areolar tissue plane (the ‘holy plane’) within (and thus sparing) the autonomic nerve plexuses, the non-visceral presacral fat pad (when present), the parietal side-wall fascia of the small pelvis, the hypogastric plexus, the vesicles, and the prostate in male patients and the vagina in female patients. All of the dissection should be performed sharp with diathermy or scissors under direct vision with good light. Throughout, the dedicated assistants should provide three-directional traction to open up the planes for the operator: diathermy can be used safely only when the areolar tissue is on stretch. The overall procedure is lengthy and time-consuming and requires prolonged attention to detail, as a careful TME plus pouch-to-anus reconstruction takes 3–5 h according to the detail of the patient’s build. Starting correctly involves three-directional traction on the colon and retroperitoneum to identify the plane between the back of the pedicle package and the gonadal vessels, ureter and pre-aortic sympathetic nerves, all of which must be carefully preserved. The key to this phase is the recognition of the shiny fascial-covered surface of the back of the pedicle – like a tapering longitudinal sausage with the inferior mesenteric vessels within. This must be gently lifted forwards. The avascular areolar tissue plane (mesorectal fascia) that surrounds the mesorectum is identified. It is worth remembering that the mesorectum resembles a bilobed lipoma. The rectum is lifted gently forwards from the bifurcation of the hypogastric nerves, and dissection commences in the midline using diathermy, aiming to minimize direct or collateral heat damage to the nerves. Dissection is extended downwards anterior to the curve of the sacrum on the surface of the mesorectal fascia. Once there is sufficient space, a St Marks rectal retractor (one with integral illumination, if available, is optimal) is introduced behind the specimen. This helps to spread and ‘tent’ the hypogastric nerves and aids identification. It is important to gently position the retractor and apply firm but gentle pressure to expose the mesorectal fascia and the layer of areolar tissue (sometimes called ‘angel hairs’) where dissection should proceed. In this manoeuvre, the

HEBK001-C07_p103-123.indd 113

operator standing on the patient’s left and the first assistant on the patient’s right have to position and control the angulation and retraction force, aided by the second assistant between the patient’s legs when more forceful retraction is needed. It is important to note that all four hands of the operator and first assistant are needed for retraction and dissection, and a suction device may be a useful retractor, in addition to its role in removing smoke and fluid. It is really only the operator and first assistant who can clearly determine the correct angulation and direction of the retractors. The second assistant’s role is predominantly supportive and to add force rather than direction to the traction and counter-traction. It is useful to wash out the pelvis on a regular basis; personal preference is to use water with dilute liquid proflavine (an antiseptic and cytocidal agent). In this context, water rather than normal saline is optimal, as it is hypotonic and therefore cytocidal. Additionally, the use of a clear solution helps to visualize the tissue planes, unlike equally effective agents such as povidone–iodine. Dissection proceeds in the ‘angel hair’ and should be predominantly from below upwards, allowing the hypogastric nerves to drop away posterolaterally. It is important to focus on circumferential mobilization rather than proceed too far posteriorly at this stage. Some dissection in the lateral and anterior plane is recommended at this juncture.

Lateral Dissection The lateral attachments are mobilized by extending the dissection plane forwards from the midline posteriorly around the side-walls of the pelvis. It is important to remember that the inferior hypogastric plexuses (formed by the hypogastric nerves and the pelvic parasympathetic nerves) curve forwards tangentially around the surface of the mesorectum in close proximity to it. The nervi erigentes (pelvic parasympathetic nerves on which male erection depends) lie more posteriorly in the same plane as the hypogastric nerves and should be visualized and preserved, although it is all too easy to ‘tent up’ the nerves and cut them at this point. The nervi erigentes curve forwards from the sacral foramina and converge like a fan to join the hypogastric nerves and form the neurovascular bundles of Walsh,4 as elegantly illustrated in Figure 7.10, adapted

11/02/13 3:51 PM

114  Total mesorectal excision for rectal cancer

Dissection Anteriorly prostate

vesicle

neurovascular bundle

Denonvilliers’ denonvilliers’ fascia

the ‘holy plane’

Figure 7.10. Schematic outline of the mesorectum and mesorectal fascia with the neurovascular bundles anterolaterally in the male pelvis.

from Walsh’s approach to radical prostatic resection for prostate cancer. Thus, the nerves lie at the outer edges of Denonvilliers’ fascia and are in danger at the 10 o’clock and 2 o’clock anterolateral positions, just behind the lateral edges of the seminal vesicles in the male. More distally, they curve forwards out of danger. As the lateral dissection moves deeper into the pelvis, one or two middle rectal vessels may be encountered and occasionally may have to be occluded by precise diathermy or ligation after application of a slender curved artery forceps. There are almost always some slender nerve branches at this point, and it is usually these branches that form the so-called ‘lateral ligament’. When medial traction is applied, these branches will ‘tent’ the plexus; it is important to divide them by sharp diathermy or scissors dissection on the mesorectal surface. The previously described clamping of the lateral ligaments is unnecessary and potentially injurious to the pelvic nerves. If bleeding is encountered, it is often wise to place a gentle pack (personal preference is to use an adrenaline-soaked swab) and move the dissection to another area, perhaps the other side or anteriorly.

HEBK001-C07_p103-123.indd 114

The traditional approach in low anterior resection in male patients was to incise the peritoneal reflection anteriorly, but a more satisfactory approach is to follow the plane forwards, from behind, anterolaterally on both sides until the vesicles are visualized. A small swab may be placed on the anterior surface of the specimen, and the plane immediately in front of Denonvilliers’ fascia is developed by sharp dissection in the midline anteriorly and then carefully extended laterally to meet the lateral dissection, remembering that autonomic nerves converge to form the neurovascular bundles at the outer edge of Denonvilliers’ fascia (see Figure 7.11). Denonvilliers’ fascia marks the anterior extent of the ‘tumour package’ and lies like an apron anterior to the anterior mesorectum, behind the vesicles, until it fuses posteriorly with the posterior fascia of the prostate. For this reason, Denonvilliers’ fascia must eventually be divided by scissors or diathermy to access the lowest few centimetres of the anterior rectum. This should

Denonvilliers’ fascia

Figure 7.11. Schematic view with rectum posteriorly. It is the author’s preference to dissect along the dotted line anterior to Denonvilliers’ fascia.

11/02/13 3:51 PM

MOBILIZATION OF THE MESORECTUM AND RECTUM  115

the ‘rectogenital septum’ (the female equivalent of the male Denonvilliers’ fascia). It is sometimes helpful to place a narrow retractor (e.g. narrow Kelly retractor) or a moist ‘swab-on-a-stick’ in the vagina to help identify the posterior fornix of the vagina and find this plane.

Denonvilliers’ Fascia and the Anterior Part of a Low Anterior Resection

Figure 7.12. Total mesorectal excision in female patients.

be well beyond the distal edge of the cancer, except in ultra-low resection of a distal rectal cancer. Anterior dissection in female patients

Anterior dissection in female patients is generally more straightforward, provided the uterus has been lifted well forward (Figure 7.12). One useful manoeuvre is to place a strong figure-of-eight suture through the apex of the uterus and suture this to the lower end of the wound, which helps considerably in uterine retraction. There is a condensation of fibrous tissue anteriorly, analogous to Denonvilliers’ fascia in the male but almost always a much more tenuous structure. It is often difficult to find a precise avascular plane behind the cervix and posterior fornix without encountering bleeding from the venous plexus. Often the peritoneal reflection may adhere to the posterior fornix. As in male patients, the plane may be best approached by continuation of the anterolateral dissection from the side-wall, thus finding the relatively avascular plane between the rectum and vagina. If bleeding occurs from the vagina, attempts to control it may be futile until the vagina is fully mobilized off the anterior rectum, allowing the stretched venous plexus to collapse down. Additionally, provided the cancer resection permits, the peritoneal incision may be made at the level of the posterior fornix to facilitate entering into the plane between the vagina and what is best termed

HEBK001-C07_p103-123.indd 115

There is ongoing controversy concerning the exact origin of Denonvilliers’ fascia and some debate as to whether it exists in female patients. Most now agree that the term ‘rectogenital septum’ is more comprehensive and encompasses Denonvilliers’ fascia, and that the most likely origin is two layers of peritoneum partly or fully fused together and obliterating what was once a peritoneal cavity extending to the pelvic floor. In anterior resection, one has to divide this layer from above to enter the plane between the rectum and the prostate in male patients, as the fascia fuses with the back of the prostate. The upper part of the septum is usually adherent to the anterior mesorectal fat of the middle third. In female patients the middle third has a rather thin and tenuous fatty layer between the rectum and vagina, with the fascia often being scant and difficult to identify. The ‘rectogenital septum’ in male patients is commonly a substantial rectangular or trapezoidal layer, like a bib between the hindgut behind and the vesicles and prostate in front. Just as peritoneum is the integral surface of the anterior aspect of the upper, intraperitoneal third of the rectum, so the rectogenital septum ­(Denonvilliers’ fascia) is integral to the fatty anterior mesorectum in the middle third in male patients.

Anterolateral Dissection Connecting the lateral to the anterior plane is a critical point in the dissection. It is usually best to continue the dissection from posterior to anterior, as the autonomic nerves will usually be visible and the correct plane is just medial to the autonomic bundles. There is a tendency to stray too far ­laterally with attendant risks of injury to the autonomic nerves or troublesome bleeding from the lateral pelvic side-wall vessels. Careful assessment

11/02/13 3:51 PM

116  Total mesorectal excision for rectal cancer

and reassessment is essential at this point to access the correct operative plane. The anterolateral peritoneal and subperitoneal incisions are curved medially towards the midline anteriorly to preserve the autonomic nerve structures.

Deep Posterior Dissection Posteriorly and posterolaterally, the areolar plane is well defined around the globular expanding bilobed mesorectum. A condensation of the fascia called the rectosacral ligament (or Waldeyer’s fascia) often presents a barrier to the surgeon posteriorly below the promontory. Just in front of the rectosacral fascia (also referred to as the rectosacral ligament), within the mesorectum, the superior rectal vessels can often be seen through the back of the mesorectal fascia, and around them cancerous nodes are likely to be present, often only millimetres away. An intact shiny visceral fascia over these must be jealously guarded and tearing avoided at all costs. This poses one of the greatest dangers of blunt manual extraction or of any haste or roughness, since the rectosacral ligament may be stronger than the surface fascia over the nodes. Thus, tearing into the lymphatic field by the inserted hand becomes a real risk and probably occurred often in the past. Sharp dissection under direct vision is crucial and good lighting is essential. Beyond this area of attachment, the plane is easy to recognize, except that the forward angulation demands strong anterodistal retraction to facilitate direct visualization. Understanding the importance of this forward angulation is critical to mastering the traction and counter-traction necessary in open, laparoscopic and robotic TME. A further reason to positively identify the ‘holy plane’ posteriorly, in front of the presacral fat pad (when present), is to avoid the risk of tearing thinwalled presacral veins, which often have no valves and can bleed prodigiously when cut or torn. Injury to these veins is much less likely if the correct plane is followed and tearing is avoided; if they are torn, a small pack and a considerable period of anterior dissection away from them will provide the safest way forward. The key to this dissection is always to remain on the yellow mesorectal fascia and to ­display the dissection plane by traction and ­counter-traction.

HEBK001-C07_p103-123.indd 116

Lateral Pelvic Dissection This involves forward extension of the plane from the back around to the sides, gently easing the adherent hypogastric nerves laterally off the mesorectal surface under direct vision. The freedom to lift the divided rectosigmoid forward often means that the tangentially running hypogastric nerves are first positively identified at this stage, the superior hypogastric plexus itself only becoming obvious proximal to the nerves after they have been dissected away from the mesorectal surface on each side. The superior hypogastric plexus may have been ensheathed by fatty tissue and not immediately recognized as a nerve bifurcation.

The ‘Lateral Ligament’ Area The so-called ‘lateral ligaments’ are now generally acknowledged not to be true ligaments and appear to be attachments between the rectum and pelvic side-wall by small nerves going directly from the pelvic plexus to the rectum, often accompanied by tiny arteries and veins. This particular point of adherence in the distal anterolateral sector is one of the most complex and difficult areas in a TME operation. The area is approached by following the ‘holy plane’ down towards the vesicles in male patients, with the expanding plexiform band of inferior hypogastric plexus outside it but increasingly adherent to it. In essence, there is no actual ligament, but there is an area of adherence between mesorectum medially and plexus laterally: small branches of nerves and vessels penetrate through at this point, but none generally reaches more than 1–2 mm in diameter. The key nerves entering this flattened band from above are largely sympathetic hypogastric nerves curving distally from the superior plexuses and more distally the erigent parasympathetic nerves coming forwards to it from behind. These arise from the front of the roots of the sacral plexus (especially S3, out of sight behind the parietal sidewall fascia). This fascia is quite robust laterally, and the surgeon will note that they usually cannot even see the internal iliac vessels, which are outside and positioned laterally. Posteriorly, these erigent pillars from the nerve roots around S3 curve forwards outside the parietal fascia, but medial to the branches of the internal iliac vessels.

11/02/13 3:51 PM

EXTENDED RESECTION IN SPECIAL CASES   117

A little way behind the vesicles in male patients, the erigent pillars pierce the fascia to join the inferior hypogastric plexus and often contribute nerve branches to the mesorectum and rectum. These neural T-junctions are the nearest structures to ‘lateral ligaments’ that the most careful surgeon will find with precise dissection. On occasions, there is a true middle rectal artery of substantial size at this point, but this is much less frequent than reported and may be no more than in the region of 20 per cent or so, and then only on one side. We believe that in the past, what the surgeon thought to be a middle rectal artery was most often a lateral intramesorectal artery, and the so-called ‘stalk’ being divided represented a ‘coning in’ to the mesorectum. Alternatively, dissection too laterally will result in damage to a pelvic side-wall vessel. Careful dissection between the mesorectum and the inferior hypogastric plexus may result in bleeding but usually nothing more than a tiny vessel that requires no more than a touch of diathermy. Thus, the clamp and cut routine generally implied a poor quality of dissection and damaged nerves, and left a substantial residue of dangerous tissue – probably all a part of the 30–40 per cent local recurrence rate that was once common. A more distal vessel from the prostatic branch in male patients or from the pelvic floor is often found later, and lower down, where it can cause troublesome bleeding. It is likely that the minimal nature of a true middle rectal blood supply also implies a minimal lymphatic component, accounting for the relatively few patients with spread to the pelvic side-wall nodes.3 With regard to the so-called ‘lateral ligament’, the surgeon should generally attempt to dissect at this point of adherence precisely between the outer aspect of the mesorectum and the triangulated neural band of nerve plexus, which should be left intact. This technique can be dubbed ‘mesorectal fat surface dissection’, because no actual loose areolar tissue exists in those areas where mesorectum and plexus are adherent. The final specimen will often lack the shiny fascial covering over this area.

male patients, as one works distally, there comes a point where the fascia must be divided transversely, as it becomes adherent to the posterior capsule of the prostate. Particular care is necessary during this step to avoid damage to the neurovascular bundles (of Walsh) that constitute the distal condensation of the inferior hypogastric plexuses, joined by numerous veins and small arteries, hence the title ‘neurovascular’. In a low anterior tumour in male patients, this can be critical: it is essential to avoid exposing malignant tissue on the front of the specimen at the very point where the nerves are curving acutely medially. Since they are often impossible to see in open surgery because of forward angulation behind the vesicles, bladder and prostate, they are in particular danger. In addition, surgeons who ‘take a slice off the back of the prostate’ are highly likely to cause impotence because of the close relationship of the neurovascular bundles to the back of the prostate. Hand in hand with this dissection anteriorly goes the development of the lateral side-wall dissection. The parasympathetic erigent nerves form posteroanterior lateral pillars on the pelvic side-wall. ­Cadaver dissections have led us all to be taught that the pelvic parasympathetic outflow is tripartite S2–3–4, but to the surgeon there is no doubt that a recognizable landmark is often a single or bifid pillar comprising a nerve root arising from the front of the S3 component of the main sacral plexus, which is out of sight posteriorly. The pillar-like appearance is due in part to the forcible forward traction on the prostate, vagina and bladder to see the structures during an open operation; this tends to bow the nerves medially and thus make them stand out. This retraction does not occur to the same extent in a laparoscopic operation, which may account for some of the reported higher incidence of nerve damage at laparoscopic rectal cancer surgery. These pillars and the hypogastric plexuses curve medially towards the back of the prostate in male patients, where they form the neurovascular bundles, which taper towards the urethra at the apex of the prostate. Here they become the erectile nerves of the corpora cavernosa or cavernous nerves.

Deep Anterior Dissection

EXTENDED RESECTION IN SPECIAL CASES

As outlined above, the key structure anteriorly is Denonvilliers’ fascia in male patients and the rectogenital septum equivalent in female patients. In

HEBK001-C07_p103-123.indd 117

With current optimal preoperative staging, it is uncommon to unexpectedly find a rectal cancer at

11/02/13 3:51 PM

118  Total mesorectal excision for rectal cancer

operation that extends into adjacent organs. It is pertinent to be aware that although attachment to or invasion of an adjacent organ may be an inflammatory adhesion, in approximately half of patients cancer invasion is present. Rupture of a malignant adhesion will almost certainly result in tumour spillage and dissemination or local recurrence. Consequently, if feasible, it is prudent to resect an adjacent, adherent structure rather than gamble on the adhesion being benign.

Uterus and Vagina Involvement of the vagina and uterus is usually detected at preoperative imaging and vaginal examination before surgery. A large fixed cancer, even with neoadjuvant chemoradiotherapy, is best removed by en-bloc resection of the uterus and rectum and as much as is needed of the posterior vaginal wall to clear the tumour safely. The vagina may be closed primarily in most such cases, but if the defect is large, particularly in a sexually active woman, reconstruction using a musculocutaneous flap may be needed. The need for this should have been anticipated, and appropriate assistance (perhaps by a plastic surgeon) and resources should be available.

Seminal Vesicles Involvement of the vesicles on one or both sides may be managed by dissection anterior to the vesicles, removing them en bloc with the rectum. The ureters are at risk and should be identified and preserved, and it may be prudent to consider preoperative ureteric stenting to facilitate identification. Additionally, the neurovascular bundles are at particular risk, and problems with urinary and sexual dysfunction may ensue.

Prostate It is possible, although technically difficult, to remove a part of the prostate involved by a rectal cancer. Major prostatic involvement may require pelvic exenteration, however, or, in selected patients, a nerve-preserving prostatectomy. Modern MRI imaging should predict this eventuality, and the

HEBK001-C07_p103-123.indd 118

advice and assistance of an experienced urologist should be sought.

Ureters The ureters may be involved in rectosigmoid or colonic cancers but are seldom involved in tumour of the mid- or low rectum, unless the tumour is very advanced. In all complex pelvic surgery, however, it is always prudent to visualize, and in most cases mobilize, the ureters. It is safe to divide the tissues anterior to the ureteric tunnels in male and female patients as the ureters are anterior to the pelvic plexuses and are crossed only by the vas deferens in males and the uterine vessels in females. A locally advanced tumour may invade the distal ureter and require en-bloc resection. Depending on the extent of the ureteric defect and the height of the ureteric resection, it may be possible to perform a scalloped end-to-end anastomosis over a ureteric stent or to reimplant the proximal end into the bladder or the opposite ureter. Again, urological assistance is strongly recommended.

Bladder Rectal cancer involving the bladder is usually predicted by preoperative pelvic MRI, and neoadjuvant chemoradiotherapy is generally indicated. Cystoscopy is recommended as part of the work-up to try to determine the site of involvement relative to the ureters. The operative strategy for rectal cancer involving the bladder requires consultation with a urologist and varies from partial cystectomy, to excising a disc of the involved bladder en bloc with the rectum (this will almost always be an upper rectal cancer in this scenario), to total cystectomy by pelvic exenteration, particularly if the trigone is involved. Bladder involvement is much more common in male patients. In female patients, the uterus and vagina intervene between the rectum and bladder, acting as a barrier to bladder invasion.

Inferior Hypogastric and Pelvic Plexuses Even though the focus in TME surgery is on identification and preservation of these nerves, a locally

11/02/13 3:51 PM

DISSECTION OF THE MOST DISTAL MESORECTUM   119

advanced tumour, adherent to the pelvic side-wall, may require nerve resection. The perivascular plane outside the nerves along the aorta and major vessels may be developed and followed. Depending on the site and extent of the tumour, it may be possible to limit dissection and nerve resection to one side. Bladder and sexual function is likely to be impaired, depending on the particular nerves excised.

Pelvic Side-Wall Nodal Involvement There is a wide variation in the reported incidence and clinical significance of pelvic side-wall involvement. Increasingly, there is general agreement that lateral nodal involvement is associated mainly with low rectal cancer, is indicative of tumours with a worse prognosis, and can often be predicted by pelvic imaging.3 The details are outlined in Chapter 14.

DISSECTION OF THE MOST DISTAL MESORECTUM The anatomy of the position of the mesorectal package in relationship to the pelvic floor becomes difficult for the surgeon to grasp from above because of its inaccessibility behind the vesicles and prostate in male patients and behind the vagina in female patients. The situation is complicated further by the fact that the levators are like a funnel in continuity with the tube of the external sphincter distally. Conceptually, and because of the distortion introduced by upward traction, surgeons tend to think of the pelvic floor as being much flatter than it is in vivo, especially if an assistant applies upward pressure on the perineum. If one doubts this, a careful look at the layers on a coronal MRI scan will make it evident. A clear three-dimensional perception of the now globular bilobed mesorectum in the depth of the pelvis and the surrounding neural lamella is the most elusive and challenging conceptual acquisition for the aspiring rectal cancer surgeon. Careful pursuit of the plane at this level eventually liberates the mesorectal package and takes the operator down to a clean muscle tube. Although crossed by a few small arteries and veins from the puborectal sling and some slips of sphincter muscle, the ‘holy plane’ here becomes the intersphincteric plane, which

HEBK001-C07_p103-123.indd 119

is familiar to proctologists from below – a tube of red skeletal muscle outside a tube of whiter smooth muscle within (the internal sphincter).

Partial Mesorectal Excision (High Anterior Resection and Mesorectal Transection) Tumours of the upper rectum (lower edge 11–15 cm from the anal verge) may not require TME and may be optimally managed by mesorectal transection 5 cm below the lower edge of the tumour. The characterization of rectal cancer into high, mid- and low has traditionally been measured from the anal verge in conscious patients using a rigid sigmoidoscope. It is worth documenting the height at MRI; indeed, after much debate, a new definition of low rectal cancer is emerging as a tumour with its distal margin at or below the level of origin of the levators on the pelvic side-wall (see www.lorec.nhs.uk). Rectal cancer heights measured under anaesthesia need to be from the dentate line, since the external sphincter dilates and retracts (add about 1.5 cm). Care is necessary: any instrument can push a mobile tumour upwards, and flexible instruments often give falsely high measurements. With these qualifications, it can be said that, in our view, most cancers with the lower edge at or below 12 cm should have TME as the key component of a radical cancer operation because the mesorectum is the primary field of lymphovascular spread. The decision as to whether the smaller partial mesorectal excision is adequate, however, is confirmed after the mobilization has been completed to the point where the mesorectum must be liberated from the adherent inferior hypogastric plexus – the region usually referred to as the ‘lateral ligament’. It has long been convention and a very sound rule, borrowed from ­German surgical practice, that a minimum of 5 cm of mesentery should always be excised both proximal and distal to any colorectal cancer. Although muscle tube margin may safely be reduced to 1 cm in the interest of anal conservation, we have always believed that if less than a TME is contemplated, a minimum of 5 cm of mesorectum distal to the lower edge of the cancer must be dissected in the perimesorectal (‘holy’) plane. If, therefore, after initial mobilization there is a clear 5 cm of mesorectum, then tapering into the mesentery, in the interest of

11/02/13 3:51 PM

120  Total mesorectal excision for rectal cancer

making a more minor operation and a higher anastomosis, becomes acceptable. An anastomosis in a patient who has mesorectal transection, provided it is airtight, is less likely to leak than after a TME. In this case, selected patients may avoid a temporary stoma.

Management of the Anorectum Distal to the Cancer: Stapling Techniques, Distal Washout and Anastomosis In more than 90 per cent of rectal cancers, it is technically feasible, although not necessarily optimal in terms of future function, to extend the dissection down to a clean muscle tube where a cross-clamp may be applied with a finger-and-thumb clearance beyond the lowest edge of the cancer. This is a difficult and challenging moment requiring both skill and experience. We have developed a preference for the use of the linear stapler in place of the rightangled clamp (the Moran triple stapling technique). The first TA-45, TA-30 (Covidien) staple line seals the muscle tube (Figure 7.13) so that the anorectal lumen beyond can be washed out with water or a tumoricidal solution (Figure 7.14). For very low tumours we have modified the technique and now leave the first TA-30 closed and fired and in place, washout below it and place a second TA-30 below the first, across the washed muscle tube (Figure 7.15). The bowel is sectioned with a scalpel between the two TA-30 guns. A proctoscope is introduced into the anal canal, and the lumen below the staple line is irrigated with repeated infusions using a 50 mL bladder syringe or through a catheter irrigation system. Water (not saline), povidone–iodine or dilute proflavine are recommended. The risk of incorporating viable exfoliated intraluminal cells in the second staple line is thus eliminated and a second TA-45 or TA-30 is fired through the washed bowel while the anatomy is distorted by upward traction on the first (pathologist’s) stapler. This process, in our view, justifies the cost of a second stapler because of the greater security against spillage of potentially malignant bowel contents. Only this washed staple line remains within the patient. The first of these two linear staple lines should be safely clear of the palpable distal edge of the cancer.

HEBK001-C07_p103-123.indd 120

Figure 7.13. Schematic drawing of a linear stapler applied across the anorectal tube below the cancer.

This is usually, although not invariably, the microscopic edge. Downward spread along the muscle tube is not a significant factor in recurrence: a 2 cm clearance is more than adequate, and 1 cm plus the ‘doughnut’ is acceptable when the tumour is very low and survival of the anal sphincters depends on such a narrow margin.

Figure 7.14. Schematic drawing of washout below an occlusive staple to remove any debris, which may contain viable tumour cells.

11/02/13 3:51 PM

DISSECTION OF THE MOST DISTAL MESORECTUM   121

For troublesome presacral, pelvic side-wall or other bothersome bleeding, a haemostatic agent such as Tachosil® may be helpful. Another strategy to keep in mind is that rather than repeated futile attempts at diathermy or suturing, packing the pelvis for 10–15 min will usually arrest bleeding.

Neorectal Reservoir

Figure 7.15. Modified ‘Triple stapling technique’. The washed muscle tube is sectioned between the two linear staplers. The muscle tube is divided on the proximal surface of the distal stapler (Moran triple stapling).

The rectal cancer specimen is now removed. Careful examination of the bowel end in the upper stapler will confirm clearance: frozen section is generally unnecessary. If there is doubt about clearance, the staples may be removed and the lumen inspected. Only after checking this should the distal (patient’s) stapler be removed. In the occasional case where the clearance is marginal, the distal stapler can be pulled up hard and a further linear stapler applied and fired beyond the distal one, although clearly this is less perfect than first-time clearance of the cancer. If there are no concerns with clearance after inspection (and removal of the staples if necessary), the distal linear stapler is removed, leaving a transverse staple line across the anorectal muscle tube. The pelvic cavity is washed copiously and inspected for bleeding. Haemostasis is secured with carefully applied diathermy or suturing if necessary.

HEBK001-C07_p103-123.indd 121

When TME has been performed and a coloanal anastomosis is required, a neorectal reservoir provides better functional outcome than a straight colonic anastomosis. Whatever technique is used, adequate length of colon from the splenic flexure mobilization is essential for the pouch to lie without tension in the hollow of the sacrum. For a reservoir, a side-to-end (side of colon to end of anorectum) anastomosis is acceptable and easiest to construct. Alternatives include construction of a colonic J-pouch or a coloplasty. For a side-to-end anastomosis, approximately half of the colonic staple line is excised and the lumen washed out. This allows assessment of the distal colonic blood supply and inspection of the mucosa. An appropriately sized circular stapler is selected (28–31 mm head is optimal), and the head is detached. The detached head of the gun is inserted into the lumen spike first, and the spike is brought out through the antimesenteric border, halfway between the taenia coli, approximately 4 cm from the distal colonic end. The defect in the staple line is closed with interrupted sutures. We tend to invert the remaining staple line as well. The side-to-end technique provides a flat surface for the proximal part of the anastomosis and leaves room for the thicker anorectal end. Several variations of pouch construction are available. Typically a GIA-60 is inserted via a colostomy 5 cm from the end of the fully mobilized colon to create a J-pouch. The CEEA-31 staple gun anvil is inserted into the same colostomy, which is purse-stringed around the shaft with 00 Surgipro.

Circular Stapled Anastomosis The anorectal remnant is palpated from between the legs. The anal canal may have to be dilated

11/02/13 3:52 PM

122  Total mesorectal excision for rectal cancer

gently to accommodate the lubricated circular staple gun. Relaxation of the anal sphincter by peranal application of glyceryl trinitrate (GTN) cream applied 30–60 min before or sublingual GTN spray applied 5 min before may facilitate introduction of the stapler by relaxing the internal sphincter. The body of the circular stapler, usually the CEEA-31, is inserted from below transanally. It is essential with ultra-low anastomoses to be certain that only the internal sphincter is purse-stringed into the instrument. To this end, it must be confirmed from above that only one thickness of muscle can be felt around the periphery of the cartridge. Care must be taken not to disrupt the transverse staple line, and the abdominal surgeon may have to bimanually assist in this step to ensure safe placement. A St Mark’s retractor helps to visualize the anorectal stump and anteriorly retract the vesicles and prostate in male patients and the vagina in female patients. Once the circular ring of the gun is visible clearly through the bowel wall, the gun is opened and the protruding spike guided through the bowel, ideally just behind the linear staple line. The anvil (head) of the gun in the proximal bowel is brought down, engaged with the shaft. The gun is closed slowly until the tissues are in apposition, as seen on the tissue indicator mechanism on the gun shaft (see Figure 7.16). At this point, it is mandatory to check the alignment of the proximal colon (including the transverse colon) to ensure there is not a 3608 twist of the colonic mesentery before firing the stapler. The circular stapler is fired according to the manufacturer’s instructions. A delay of a minute or so before opening the gun according to the manufacturer’s instructions is said to reduce the risk of staple line haemorrhage. The doughnuts are inspected for intactness of the tissue rings. If there is concern about tumour clearance in a low rectal cancer, the distal doughnut should be sent for histology. The anastomosis is gently palpated for integrity and can be air-tested by filling the pelvic cavity with water and insufflating air via the anal canal using a bladder syringe. If an air leak is identified, it may be possible to repair it with interrupted sutures, if necessary by a transanal approach. Two low-suction Abdovac drains are used for 48 h, unless there is copious drainage, when they may need to be left for longer. The objective is to avoid

HEBK001-C07_p103-123.indd 122

Figure 7.16. Side- (of the colon) to-end anastomosis using a circular stapling gun.

haematoma in the hollow of the sacrum, which may become infected, form an abscess, and point into the bowel at or near the anastomosis, thus creating a late leak 10–20 days later.

Defunctioning a Low Anastomosis Even if the anastomosis is airtight, consideration should be given to temporarily defunctioning all coloanal anastomoses after TME. A randomized trial in 2009 reported a 28 per cent leak rate in patients after TME without a loop stoma compared with 10 per cent in patients with a stoma.5 Factors that have been reported by many groups to increase the risks of anastomotic leakage are

11/02/13 3:52 PM

FURTHER READING  123

height of the anastomosis from the anal verge (particularly below 5 cm, which includes all patients who have had a TME), male patients, preoperative chemoradiotherapy, and intraoperative technical factors such as major bleeding and absence of a loop stoma. There is little to choose between a defunctioning loop transverse colostomy and loop ileostomy, although an ileostomy has the added risks of a high-output stoma and probably more adhesional obstruction risks in the longer term. Unquestionably, a loop stoma reduces the consequences of a leak and the need for emergency surgery. If the recovery is uneventful, the stoma may be closed 6–8 weeks later after a water-soluble enema and rectal palpation to check there is no leak and to dilate up any slight narrowing, which is quite common in the defunctioned anastomosis. If a temporary stoma has not been placed and there are concerns in the postoperative period, an anastomotic leak should be sought by a rectal contrast study (often performed in combination with an abdominal and pelvic CT scan). If a leak is detected, the patient should have immediate broad-spectrum antibiotics while being prepared for emergency reoperation surgery, which will need a stoma. On occasion, it is possible to salvage the anastomosis by a combination of a proximal loop colostomy, distal washout of the downstream colon and a pelvic drain. In some patients, however, anastomotic excision with an end stoma is required.

Conclusion Rectal adenocarcinoma is a common cancer and is curable by surgery alone in most cases. Recent advances in pelvic MRI, adoption of the surgical concepts of total mesorectal excision, and ­excellent mechanical stapling instruments have revolutionized the management and outcomes of this technically challenging but eminently curable cancer.

References 1. Moran BJ. Stapling instruments for intestinal anastomosis in colorectal surgery. Br J Surg 1996; 83: 902–9. 2. Karanjia ND, Schache DJ, North WR, et al. ‘Close shave’ in anterior resection. Br J Surg 1990; 77: 510.

HEBK001-C07_p103-123.indd 123

3. Yano H, Moran B. The incidence of lateral pelvic side-wall nodal involvement in low rectal cancer may be similar in Japan and the West. Br J Surg 2008; 95: 33–49. 4. Walsh PC, Schiegel PN. Radical pelvic surgery with preservation of sexual function. Ann Surg 1988; 208: 391–400. 5. Matthiessen P, Hallböök O, Rutegård J, Simert G, Sjödahl R. Defunctioning stoma reduces symptomatic anastomotic leakage after low anterior resection of the rectum for cancer: a randomized multicenter trial. Ann Surg 2007; 246: 207–14.

FURTHER READING Birbeck K, Macklin C, Tiffen N et al. Rates of circumferential resection margin involvement vary between surgeons and predict outcomes in rectal cancer surgery. Ann Surg 2002; 235: 449–57. Daniels IR, Fisher SE, Heald RJ, Moran BJ. Accurate staging, selective pre-operative therapy and optimal surgery improves outcome in rectal cancer surgery: a review of the recent evidence. Colorectal Dis 2007; 9: 290–301. Dukes CE. The classification of cancer of the rectum. J Pathol Bacteriol 1932; 35: 323. Edwards DP, Leppington-Clark A, Sexton R, Heald RJ, Moran BJ. Stoma related complications are more frequent after transverse colostomy than loop ileostomy: a randomized controlled trial. Br J Surg 2001; 88: 360–63. Heald RJ, Husband EM, Ryall RDH. The mesorectum in rectal cancer surgery: the clue to pelvic recurrence? Br J Surg 1982; 69: 613–6. Heald RJ. The ‘holy plane’ of rectal cancer. J R Soc Med 1988; 81: 503. Heald RJ, Moran BJ, Ryall RDH, et al. The Basingstoke experience of total mesorectal excision 1978–1997. Arch Surg 1998; 133: 894. Mercury Study Group. Diagnostic accuracy of pre-­ operative magnetic imaging in predicting curative resection of rectal cancer; prospective observational study. BMJ 2006; 333: 779–84. MacFarlane JK, Ryall RD, Heald RJ. Mesorectal excision for rectal cancer. Lancet 1993; 341: 457. Moran BJ, Docherty A, Finnis D. Novel stapling technique to facilitate low anterior resection for rectal cancer. Br J Surg 1994; 81: 1230. Quirke P, Durdey P, Dixon MF, et al. Local recurrence of rectal adenocarcinoma due to inadequate surgical resection: histopathological study of lateral tumour spread and surgical excision. Lancet 1986; 2: 996. Quinlan DM, Epstein JL, Careter BS, Walsh PC. Sexual function following radical prostatectomy: influence of preservation of neurovascular bundles. J Urol 1991; 145: 998–1002.

11/02/13 3:52 PM

8 Abdominoperineal excision of the rectum TorbjÖrn Holm

Introduction The earliest surgical approaches to rectal cancer were via the perineum, and the techniques used were exclusively extraperitoneal. The intraoperative and postoperative mortality was high, the postoperative functional results were extremely poor, and the local recurrence rate ranged up to 90 per cent. An important step in the development of the surgical treatment for rectal cancer was taken by W. Ernest Miles, a surgeon at St Mark’s Hospital in London, who on 19 December 1908 in the Lancet published a paper entitled ‘A method of performing abdomino-perineal excision for carcinoma of the rectum and of the terminal portion of the pelvic colon’.1 In Miles’ original description of the procedure, the rectum was bluntly mobilized down to the sacrococcygeal articulation posteriorly, to the prostate anteriorly and to the upper surface of the levator ani laterally. A colostomy was brought out and the abdominal wall was closed. The patient was then turned over and placed in the right lateral and semiprone position. The perineal part of the operation included a wide excision of skin and fat, and Miles emphasized that the levator muscles should be divided ‘as far outwards as their origin from the white line so as to include the lateral zone of spread’. The specimen was brought out through the perineum and the skin was closed over two drains.

HEBK001-C08_p124-139.indd 124

This paper had an enormous impact on the surgical community, and ‘Miles’ operation’ became the gold standard procedure for all rectal carcinomas for many decades. The concept of removing the entire rectum and the anus in all patients with rectal cancer gradually changed with time, however, and the increasing experience with bowel reconstruction, including the development of stapling instruments, led to the new concept of anterior resection and low anterior resection, which became the standard procedures for tumours in the upper and middle rectum.2–6 For tumours in the lower rectum, most surgeons continued to perform abdominoperineal excision (APE), although the extensive perineal approach described by Miles was more or less forgotten and the synchronous combined APE was introduced as a feasible procedure, which became popular and gained widespread use in the treatment of low rectal cancer.7 During the synchronous combined operation, the perineal part is carried out simultaneously with the pelvic part of the abdominal procedure, with the patient in the supine lithotomy or Lloyd Davies position. The rectum with its mesorectum is first mobilized down to the pelvic floor; the perineal surgeon then enters the pelvic cavity just in front of the coccyx, the levator muscles are divided on both sides, and the rectum is dissected off the prostate or the vagina and the specimen delivered through the perineum.

08/02/13 5:54 PM

PROBLEMS ASSOCIATED WITH CONVENTIONAL SYNCHRONOUS COMBINED  125

Despite the gradual improvements in rectal cancer treatment during the twentieth century, local control remained a major problem after surgery, with local recurrence rates of up to 50 per cent after potentially curative resections.8 Therefore, preoperative radiotherapy was evaluated in several randomized trials during the 1980s; it was shown to reduce the local recurrence rate by 50 per cent and also to improve cancer specific survival. It was with the development of total mesorectal excision (TME), however, as described by Bill Heald, that the picture changed dramatically.9,10 The TME technique for rectal cancer resection was introduced in many countries over the past 15–20 years, and subsequently the results with regard to local control and cancer survival have improved significantly. Local recurrence rates are now reported to be less than 10 per cent in population-based studies.11–13 The acknowledgement of TME as the standard surgical technique in the treatment of rectal cancer has resulted not only in improved local control but also in increasing rates of sphincter-saving procedures and improved survival. Therefore, since the mid-1990s, teaching in rectal cancer surgery has focused mainly on the operative technique of TME and anterior resection. Although the technique used for the abdominal part of APE was modified along the lines of TME, little attention was given to the perineal part of this procedure. Thus, most surgeons adopted the technique of sharp dissection under direct vision outside the mesorectal fascia down to the pelvic floor, with the aim to save autonomic nerves and to create a perfect specimen with an intact mesorectal fascia. The perineal part, however, has often been completed in the conventional way, with dissection outside the external sphincter and with the division of the levator muscles close to the rectal wall. With the patient in the supine lithotomy position, it is difficult to achieve an optimal view, especially anteriorly (in the front), and therefore parts of the perineal dissection are often done with blunt dissection when this approach is used.

PROBLEMS ASSOCIATED WITH CONVENTIONAL SYNCHRONOUS COMBINED ABDOMINOPERINEAL EXCISION In recent years, several authors have acknowledged the fact that local control and survival after APE

HEBK001-C08_p124-139.indd 125

have not improved to the same degree as that seen after anterior resection. In one study based on 561 patients in Leeds in the UK, it was reported that compared with patients who had an anterior resection during the same time period, patients undergoing APE had a higher local failure rate (22.3 per cent v. 13.5 per cent) and poorer survival (52.3 per cent v. 65.8 per cent).14 In another paper, based on data from five different European trials, it was reported that the APE procedure was associated with an increased risk of circumferential resection margin (CRM) involvement, an increased local recurrence rate and a decreased cancer specific survival.15 The difference in oncological results between the two procedures may be explained by several factors, including anatomical difficulties and the surgical technique associated with standard APE surgery. In the lower rectum, the surrounding mesorectum is reduced in size and disappears at the top of the sphincters. Below this level, the sphincter muscle forms the CRM. As mentioned above, the abdominal dissection during a conventional synchronous combined APE is often carried out along the mesorectum, all the way down to the pelvic floor and the top of the puborectalis muscle, with the mesorectum being mobilized off the levator muscles. The perineal dissection then follows the external sphincter to meet the pelvic dissection at the top of the anal canal (Figure 8.1a). With this technique, the retrieved specimen often has a typical ‘waist’ at 3–5 cm from the distal end, corresponding to the top of the external sphincter at the level of the puborectalis muscle and the lowest part of the mesorectum (Figure 8.1b). This inwards coning at the pelvic floor carries the dissection close to the rectal wall, and several studies have reported higher rates of bowel perforation and tumour involvement of CRM after APE compared with after anterior resection. Nagtegaal and colleagues assessed 846 anterior resection specimens and 373 APE specimens from the Dutch TME trial. They found that the plane of resection was within the sphincter muscle, the submucosa or lumen in more than a third of the APE cases and was on the sphincter muscles in the remaining cases. This resulted in a positive CRM rate of 30.4 per cent after APE versus 10.7 per cent after anterior resection, and a perforation rate of 13.7 per cent after APE versus 2.5 per cent after ­anterior ­resection.16 Similarly, population-based reports

08/02/13 5:54 PM

126  Abdominoperineal excision of the rectum A

Figure 8.1a. The pelvic dissection in a conventional synchronous combined APE is carried along outside the mesorectal fascia down to the top of the anal canal (blue line) and the perineal dissection is carried along the external sphincter (red line). The two dissection planes meet at the level of the puborectal muscle, which creates a waist on the specimen.

Levator ani Obturator internus Ischioanal fossa External sphincter Internal sphincter

B

Figure 8.1b. Photograph showing a fresh specimen after a conventional APE, with the typical waist at the level of the puborectal muscle.

from Sweden, Norway and the Netherlands have shown a threefold increase in perforation rates after APE compared with anterior resection (14–15 per cent v. 3–4 per cent).17 Thus, the differences in oncological outcome between the conventional type of APE and anterior resection may be explained to a substantial part by the increased risk of tumour-involved margins and inadvertent bowel perforations, as both of these factors are significantly related to local control and survival. With the development of TME, leading to substantially improved results after anterior resection, many surgeons have advocated low or ultra-low

HEBK001-C08_p124-139.indd 126

anterior resection, even for tumours of the lower rectum. It has also been shown that these procedures are feasible and oncologically safe, provided the tumour can be removed with a clear distal and circumferential margin. In dedicated and highly specialized centres adopting intersphincteric anterior resection for appropriate cases, the overall APE rate may be below ten per cent. On the other hand, the functional results after an ultra-low anterior resection may be poor, especially if the patient has received preoperative radiochemotherapy.18 In patients with a preoperative history of gas or faecal incontinence, careful counselling is therefore mandatory and information should be

08/02/13 5:54 PM

PREOPERATIVE PREPARATION  127 100 90 80 70 60 Figure 8.2. Graph from the Swedish Rectal Cancer Registry showing the proportion of patients, with rectal cancer below 6 cm, operated on with APE, AR or Hartmann’s procedure, annually since 1995.

LAR

50

Hartmann

40 30 20 10 0

1995

1997

given about the risk of a poor functional outcome after an anterior resection. In such patients, a permanent stoma may be preferable. If the tumour in the lower rectum is more advanced, growing close to or into the distal mesorectal fascia, the levator muscle or the external sphincter and thereby threatens the potential circumferential resection margin, it may not be possible to perform a safe anterior resection; in these cases, an APE is necessary. The decision on when to recommend an APE is therefore related to the patient and the tumour characteristics. Since such variables are interpreted differently between different surgeons, the rate of APE varies greatly between individual surgeons and between different institutions. Morris and colleagues reported that the rate of APE varied from 8.5 per cent to 52.6 per cent between different English hospitals.19 In Sweden, the rate of APE for low rectal cancer, as defined by tumours within 6 cm from the anal verge, has varied between 80 per cent and 92 per cent over the past 15 years (Figure 8.2). Thus, APE is still a common operation for low rectal cancer. Since the results have been suboptimal, it is important to change the concept of APE to reduce the rate of inadvertent bowel perforations and tumour-involved margins and thereby obtain improved oncological outcomes.

The new concept of abdominoperineal excision One obvious problem associated with the conventional type of synchronous combined APE is the lack of standardization. Although the abdominal part of

HEBK001-C08_p124-139.indd 127

APE

1999

2001

2003

2005

2007

2009

2011

the operation follows the standard TME principles, there has been no obvious agreement on the surgical details of the perineal part of the operation. This probably explains the significant variability in the observed rates of tumour-involved margins, bowel perforations, local recurrence and survival.20 Due to this variability, and the suboptimal results after APE, there has been a call for a different concept and a more standardized approach to APE.21 In recent years, a new concept of APE has therefore evolved, which takes into account the specific anatomical structures of the perineum and the pelvic floor and which aims to adopt and standardize the procedure according to the characteristics of the patient and the tumour. Three basic types of APE can be described in relation to the perineal approach and the extent of dissection: the intersphincteric APE, the extralevator APE and the ischioanal APE. The mobilization of the rectum and the mesorectum during the pelvic dissection in the abdominal part of the operation differs between the intersphincteric APE and the two other types. In addition, the indications are different for the three procedures, as shown in Table 8.1.

PREOPERATIVE PREPARATION All patients planned for an APE should be well informed about the extent of the procedure, the potential complications that may occur postoperatively, and the possible late sequel that they may have to live with. These adverse outcomes after APE surgery are discussed below. A crucial part of the preoperative preparation and information is to have the patient meet a stoma

08/02/13 5:54 PM

128  Abdominoperineal excision of the rectum Table 8.1. Indications for APE in rectal cancer Inter-sphincteric APE •  Patient unsuitable for bowel reconstruction •  Preoperative history of incontinence •  High risk of anastomotic leak •  Co-morbidity: crucial to prevent leakage + fatal outcome •  Patient’s preference Extra-levator APE •  Tumour extending less than 1 cm from dentate line (T2 – T4 cancer) •  Tumour threatening CRM Ischio-anal APE •  Locally advanced cancer infiltrating levator muscles, ischio-anal fat or peri-anal skin •  Perforated cancer with abscess or fistula in ischio-anal compartment

nurse, well ahead of the operation. The stoma nurse has an important role in informing the patient about the practicalities around stoma care and how to use bags and other aids. It is also very important that the placement of the stoma is assessed carefully to avoid a suboptimal placement, close to a skinfold or a scar. The patient needs to be able to see the stoma, which may be a problem in patients who are obese if the stoma is placed too low. Thus, the stoma site should always be marked in advance by the stoma nurse. Prophylaxis against deep venous thromboembolism should be administered the evening before surgery, and antibiotic prophylaxis against postoperative infection should be given, either orally in the morning or intravenously within 30 min of the abdominal incision. Per oral mechanical bowel preparation is not necessary for APE, but it is recommended to give an enema to clear the rectum in the morning before surgery. After administration of general anaesthesia, a bladder catheter is inserted. The main reason for this is to identify the urethra during the perineal phase of an extralevator APE or ischioanal APE. Our preference is to keep the catheter closed and to insert a suprapubic catheter once the abdomen is opened. The urethral catheter is removed directly after surgery and the suprapubic catheter kept postoperatively. In patients scheduled for an intersphincteric APE, there is no need for a urethral catheter, as the perineal dissection is carried out between the internal and external sphincter, at a safe distance from the urethra.

HEBK001-C08_p124-139.indd 128

The patient is placed in a modified lithotomy position, with the buttocks brought down to the edge of the table and the legs placed in soft stirrups, but the legs are not elevated during the abdominal phase. A preoperative briefing is important to allow the surgeon to share the plan with the entire operative team and to confirm the presence of appropriate instruments such as self-retaining retractors, St Mark’s pelvic retractors, a variety of staplers and other devices, including long instruments. Headlights or lighted retractors may facilitate dissection deep in the pelvis. The assistance of an experienced second surgeon is invaluable and strongly recommended. The surgeon should reassess the rectal cancer by digital rectal examination and confirm the degree of involvement of the anal sphincter or other organs and the level of the distal edge of the tumour. In female patients, the vagina must be examined to assess the relation of the tumour to the posterior vaginal wall. The abdomen and perineum, including the vagina in female patients, should be prepared and properly draped.

ABDOMINAL PROCEDURE IN ABDOMINOPERINEAL EXCISION With a few exceptions, the approach and the operative technique for the abdominal part of this procedure are identical to those used with TME and anterior resection, as described in Chapter 6. The abdomen is opened through a midline incision and the entire abdominal cavity is explored to detect any metastatic disease or other unexpected pathology. The small bowel is packed into the upper abdomen, the patient is placed in a slight Trendelenburg position, and a self-retaining retractor is inserted. The sigmoid colon may be mobilized from medial to lateral or vice versa, although it is our preference to start the mobilization laterally. The sigmoid colon is retracted to the right and the peritoneal attachments laterally are incised along the avascular plane (white line of Toldt), distally to the level of the promontory and as far proximally as needed. It is usually necessary to mobilize a portion of the descending colon to allow the later construction of a tensionfree end colostomy, elevated 1 cm above the skin level. A complete takedown of the splenic flexure, as is routinely done if a low anastomosis is planned, is usually unnecessary for an APE. The left ureter and

08/02/13 5:54 PM

PELVIC DISSECTION IN INTERSPHINCTERIC ABDOMINOPERINEAL EXCISION  129

gonadal vessels are identified and preserved by using sharp and gentle blunt dissection to separate the retroperitoneal tissues from the left colonic mesentery. The sympathetic nerve plexus in front of the aorta is identified, and the dissection continues in front of these nerves and from left to right, just posterior to the inferior mesenteric artery (IMA). The peritoneum on the right side of the colonic mesentery is then opened and the IMA is followed proximally to its origin at the aorta. There is no consensus on where to divide the IMA. Some surgeons prefer a high ligation at the origin from the aorta and suggest that this maximizes the lymph node yield and may improve oncological outcome. Other surgeons have a preference for a low ligation just distal to the left ascending colic artery and argue that this ensures a better blood supply to the remaining left colon and may prevent nerve damage at the base of the IMA, resulting in less functional impairment. There is not enough evidence to state that one approach is better than the other. After ligation of the IMA and the inferior mesenteric vein at the same level, the sigmoid mesentery is divided with diathermy coagulation of small vessels, division and ligation of the marginal artery, and finally division of the colon at the level of the proximal sigmoid colon, preferably with a linear stapler to prevent any faecal contamination.

PELVIC DISSECTION IN INTERSPHINCTERIC ABDOMINOPERINEAL EXCISION Pelvic dissection in intersphincteric APE is identical to that performed for anterior resection, which is described in detail in Chapter 6. In summary, the loose connective tissue plane (‘holy plane’) separating the mesorectal fascia from the parietal pelvic structures is identified and followed first posteriorly, then to the left and right, and finally anteriorly while the peritoneum is gradually divided. Gentle traction on the specimen and counter-traction with appropriate retractors is crucial to achieve a good view of this plane. The superior hypogastric plexus is identified at the sacral promontory and the hypogastric nerves should be identified, protected and preserved while the dissection gradually proceeds downwards in the pelvic cavity. Anteriorly the dissection is conducted just posterior to the vesicles and prostate

HEBK001-C08_p124-139.indd 129

in male patients and the vagina in female patients. The lower anterolateral dissection is the most difficult part of the abdominal phase of the operation because the correct plane is often difficult to find here, and the inferior hypogastric plexus must be carefully preserved to maintain postoperative sexual and urinary function. Reducing the angle of the Trendelenburg position or even shifting the patient to a reverse Trendelenburg position may facilitate the exposure for the anterior dissection. Mobilization of the rectum with an intact mesorectum is continued down to the pelvic floor and the puborectalis muscle. Because the intersphincteric APE is generally not performed when the tumour is close to the anus, a transverse stapler can be put across the rectum below the tumour to seal the bowel and to prevent leakage of mucus or faeces from the anus.

Perineal Part of Intersphincteric Abdominoperineal Excision The patient’s legs are elevated and the perineum is exposed. The surgeon and assistant now move from the abdomen to perform the perineal phase of the intersphincteric APE. The anal canal is washed out and an incision is made around the anus, just distal to the intersphincteric groove. A self-retaining retractor with hooks is recommended to optimize the view and to facilitate the intersphincteric dissection. Once the skin incision is made, the anus is closed with a running suture. The dissection then follows the intersphincteric plane between the internal and external sphincter, around the circumference of the anal canal, and all the way up to the puborectal sling and into the pelvic cavity (Figure 8.3A). The specimen is gently removed through the perineal incision or, if the mesorectum is large and bulky, lifted up from the pelvis and removed from the abdomen via the abdominal incision (Figure 8.3B). The perineal incision is closed subcutaneously with a running or interrupted suture in the puborectalis and external sphincter. It is our preference to use a running suture in three layers, where the most superficial suture line is placed subdermal to leave the skin unsealed in order to allow for discharge of fluid from the wound.

08/02/13 5:54 PM

130  Abdominoperineal excision of the rectum A

Figure 8.3a. The pelvic dissection in an inter-sphincteric APE is carried along outside the mesorectal fascia down to the top of the anal canal (blue line) and the perineal dissection is carried out between the internal and external anal sphincter (red line). The two dissection planes meet at the level of the puborectal muscle.

Levator ani Obturator internus Ischicanal fossa External sphincter Internal sphincter B

Figure 8.3b. Photograph showing a fresh specimen after an intersphincteric APE.

Finalizing the Abdominal Procedure When the perineum is closed, the patient’s legs are put down and the abdominal part of the operation continues with a washout of the pelvic cavity, preferably with sterile water or some other cytotoxic agent, and haemostatic control.

be mobilized from the transverse colon and from the greater curvature of the stomach to prepare an omentoplasty, which can help to fill out the empty pelvic cavity. We prefer to place a drain in the pelvic cavity, but there is no substantial evidence that this is of any value. Finally, a thorough control of haemostasis in the abdomen and pelvic cavity is mandatory before stoma formation and closure of the abdominal wall.

Omentoplasty

Stoma Formation

Bowel obstruction due to entrapment of the small bowel in the pelvic cavity is not infrequent after APE. Therefore, if the patient has a large omentum, this may

A circular incision is made in the skin at the marked stoma site, and the dissection is carried through the subcutaneous fat down to the aponeurosis covering

HEBK001-C08_p124-139.indd 130

08/02/13 5:54 PM

EXTRALEVATOR ABDOMINOPERINEAL EXCISION  131

the rectus abdominis muscle. There is no consensus on how to incise this aponeurosis; some surgeons make a longitudinal incision, others prefer a crossincision, while some make a round hole and remove a small disk of the aponeurosis. The fibres of the rectus abdominis muscle are gently separated and the posterior sheet and peritoneum is opened. The divided sigmoid or descending colon is now picked up and brought out through the stoma site. At this point it is vital to ensure that the bowel is of sufficient length and has a good blood supply. The bowel should not be opened at this point, and not until the abdominal wall is closed and the wound properly dressed. In recent years some authors have advocated placing a lightweight mesh in a sublay position, deep to the rectus abdominis muscle, to prevent parastomal hernia formation, a complication that is frequent in patients with a colostomy. The use of a preventive mesh in colostomy formation is described in more detail below. Once the bowel has been pulled through the stoma site, the abdominal wall is closed. This may be done in different ways, but we strongly advocate a monofilament absorbable or non-resorbable suture, a short distance between stitches, and a suture length to wound length ratio of at least 4 to reduce the risk of wound dehiscence and the later development of incisional hernia. The skin and wound are cleaned with saline and the wound is dressed with an appropriate bandage. The exteriorized bowel is now opened, everted and fixed to the skin with interrupted monofilament sutures. Finally, a stoma bag is placed to cover the stoma.

EXTRALEVATOR ABDOMINOPERINEAL EXCISION The main objective of an extralevator APE is to diminish the risk of inadvertent bowel perforation and tumour involvement of the circumferential resection margins. This is a consequence of excision of the levator muscles en bloc with the mesorectum to protect the most distal part of the bowel, thereby avoiding the ‘waist’ of the specimen, which has been so common after the conventional type of synchronous combined APE. Since the levator muscles should not be separated from the mesorectum, the

HEBK001-C08_p124-139.indd 131

pelvic dissection during the abdominal part of an extralevator APE differs notably from an anterior resection or an intersphincteric APE (Figure 8.4A).

Pelvic Dissection in Extralevator Abdominoperineal Excision The initial abdominal and pelvic dissection is identical to that described above, but with one very important difference. In both anterior resection and intersphincteric APE, the dissection continues all the way down to the pelvic floor and the puborectalis muscle and subsequently the mesorectum is lifted off the levator muscles. In extralevator APE it is crucial not to take the mobilization of the rectum and mesorectum as far down as the pelvic floor. Instead, the dissection should proceed only down to the sacrococcygeal junction dorsally, just beyond the inferior hypogastric plexus anterolaterally, and anteriorly dissection should stop just below the seminal vesicles in male patients or the cervix uteri in female patients. By terminating the mobilization of the rectum and mesorectum at this level, the mesorectum is still attached to the levator muscles of the pelvic floor, which is a crucial feature of extralevator APE. After completion of the dissection down to this level, the pelvic cavity is rinsed as described above. The final steps of the abdominal procedure are carried out just as for intersphincteric APE, the abdominal wall is closed and the stoma fashioned.

Perineal Part of Extralevator Abdominoperineal Excision The perineal part of extralevator APE differs considerably from the perineal part of intersphincteric APE. This part of the operation can be performed with the patient in the supine Lloyd Davies position, or in the prone jack-knife position. Our preference is the prone jack-knife position, due to the excellent exposure of the operative field. Other surgeons prefer the supine position, mainly to avoid the time-consuming process of turning the patient and subsequent preparation and dressing of the perineal area. The pros and cons of the two different positions during the perineal phase of the operation are discussed below. With the patient in the prone jack-knife position, the exposed, washed and sterile operative field should be well wide of the anus, laterally halfway

08/02/13 5:54 PM

132  Abdominoperineal excision of the rectum A

Figure 8.4a. The pelvic dissection in an extra-levator APE is carried along outside the mesorectal fascia but stops at the top of the levator muscle (blue line). The perineal dissection proceeds just outside the external sphincter and along the levator muscle fascia, up to its origin at the obturator internus muscle (red line).

Levator ani Obturator internus Ischieanal fossa External sphincter Internal sphincter

B

Figure 8.4b. Photograph showing a fresh specimen after an extralevator APE. The specimen is more cylindrical, without a waist, because the levator muscle is attached to the mesorectum.

along each buttock and dorsally up to the level of the mid-sacrum. In male patients the scrotum and penis should be covered and kept out of vision to expose the whole perineum. In female patients the vagina should be washed out and kept in the operative field. In male patients a urethral catheter must be in place even if a supra pubic catheter has been inserted during the abdominal phase of the operation. This is to ensure that any damage to the urethra during the dissection in the anterior plane, between the rectum and the prostate, can be detected and taken care of immediately if it occurs. The perineal phase starts with closure of the anus to avoid any spillage of faeces or mucus, which could contain tumour cells. The aim of anal closure is thus to reduce infection and reduce the risk of tumour contamination, which may result in local recurrence.

HEBK001-C08_p124-139.indd 132

Closing the anus can be done with a double pursestring suture or with an inverting running suture after the skin incision has been made around the anus. The latter technique is ­especially valuable in very low advanced tumours, which may protrude through the anus. After closure of the anus, it is recommended to wipe the operative field once more with an alcohol solution or another appropriate antibacterial solution. In extralevator APE, less skin and ischioanal fat are excised compared with Miles’ original description of the perineal part of the APE procedure. Instead, the skin is incised around the anus, with a margin of only about 3 cm anteriorly and laterally; posteriorly the incision is carried up to the level of the lower sacrum, i.e. 2–3 cm cranial to the sacrococcygeal junction.

08/02/13 5:54 PM

EXTRALEVATOR ABDOMINOPERINEAL EXCISION  133

With gentle traction and counter-traction on the skin edges, the dissection is continued in the ­subcutaneous fat. As the dissection proceeds deeper, it is important to identify the subcutaneous extension of the external sphincter. These fibres of striated muscle should be kept medially, and the dissection follows a plane between the external sphincter and the thin fascia covering the ischioanal fat in the ischioanal compartment (also called the ischiorectal fossa) on both sides. At the top of the external sphincter and puborectal muscle, the levator ani muscles are in direct continuity and the dissection is carried along the surface of the levator muscles all the way up to their insertion at the pelvic side-wall, i.e. the obturator internus muscle. The pudendal nerve, derived from S2–4, runs together with the pudendal vessels through Alcock’s canal and converges towards the midline on the outer anterior surface of the levator muscles. Small branches of blood vessels and nerves deriving from the pudendal vessels and nerve cross the space between the ischioanal compartment and the levator muscles, and these are divided by diathermy. Once the surface of the levator muscles is exposed all around the circumference, haemostasis must be controlled before entering the pelvic cavity. In the midline, the levator muscles are attached to the anterior surface of the coccyx and continue as the presacral fascia on the anterior lower aspect of the sacrum. The dissection follows the proximal portion of the levator muscles on both sides of the coccyx so that the coccyx is clearly exposed. An incision is made at the sacrococcygeal junction, which is easily identified by gentle moving of the coccyx. Once the cartilaginous connection between the sacrum and coccyx has been opened, the coccyx is pressed anteriorly to stretch the presacral fascia, which is then divided and an entrance into the pelvic cavity created. At this stage it is important to identify the mesorectum in order to not injure the mesorectal fascia. The pelvic floor, i.e. the levator muscle, is now divided from posterior to anterior, first on one side and then on the other side. As the division of the pelvic floor continues anteriorly, it is important to avoid division of the levator muscles too far laterally and too close to the ischial tuberosity as this may injure the main pudendal nerve and vessels in Alcock’s canal. The division of the pelvic floor continues until the dorsolateral part of the prostate or vagina can be

HEBK001-C08_p124-139.indd 133

palpated and visualized, or preferably 1–2 cm posterior to this point. The specimen is now still attached to the anterior aspect of the levator muscles and to the prostate or posterior wall of the vagina. The dissection in the anterior plane during the perineal phase of extralevator APE is the most difficult, and potentially most dangerous, part of the procedure because of the close relationship between the anterior rectal wall and the prostate or posterior vaginal wall. In addition, the neurovascular bundles derived from the inferior hypogastric plexus run anterolaterally on each side of the prostate or vagina and close to the rectum and are easily damaged if they are not recognized at this stage of the operation. The dissection along the anterior and lateral aspects of the lower rectum must therefore be performed meticulously and with great care. If the dissection is performed close to the rectal wall, there is a risk of inadvertent perforation or tumourinvolved margin; if the dissection is carried out too laterally or too anteriorly, there is a risk of damage to the neurovascular bundles or to the prostate or vagina. In anteriorly located tumours, it may be necessary to include the posterior vaginal wall or a slice of the posterior prostate with the specimen, and sometimes even to sacrifice the neurovascular bundle on one side, to be able to achieve a negative CRM. This extension of the procedure should ideally be planned in advance, however, so that the surgeon is prepared for it and the patient is well informed about the consequences, which may be impairment of bladder or sexual function. To facilitate the anterolateral dissection of the lower part of the rectum, it is recommended that the specimen is gently brought out of the pelvic cavity so that the anterior aspect of the bowel can be seen. It is now easy to look into the pelvic cavity and to recognize the seminal vesicles and upper part of the prostate in male patients and the posterior vaginal wall in female patients. The plane between Denonvillers’ fascia and the prostate or posterior vaginal wall is now followed carefully while the surgeon attempts to identify the neurovascular bundles on each side. Gradually, these planes of dissection are developed anteriorly and alternately on the right and left sides, and the remaining part of the levator muscles that are attached to the lowest part of the rectum are divided. Finally, the puborectal muscle on each side and the perineal body just posterior to the transverse perineal muscle is divided and the specimen can be

08/02/13 5:54 PM

134  Abdominoperineal excision of the rectum

delivered. The excised specimen is cylindrical, usually without a waist, due to the fact that the levator muscle is still attached to the mesorectum, forming a cuff around the rectal muscle tube (Figure 8.4B). As soon as the specimen is removed, it is crucial to control haemostasis at the back of the prostate or vagina, along the neurovascular bundle on each side and on the pelvic side-walls. When bleeding is completely controlled, the pelvic cavity is rinsed with sterile water or another appropriate cytotoxic solution. If an omentoplasty has been prepared, the omentum is now gently brought down into the pelvis by gentle traction. This manoeuvre may be difficult, especially if a large omentum is to be pulled down into a narrow pelvis. Therefore, it is recommended that the omentum is attached to the proximal end of the specimen with a couple of sutures during the abdominal phase of the operation. This significantly facilitates the placement of the omentum in the pelvic cavity, since the omentum slides down into the pelvis when the rectum is gently pulled out during the perineal phase of the operation. Care should be taken not to rotate the omentum by 360 as this will impair its blood supply and may result in omental necrosis. At this stage the pelvic drain, placed through the abdomen, is also brought down into the pelvic cavity. The next step of the operation is to reconstruct and close the pelvic floor. As discussed below, there are several alternative methods available to complete this reconstruction.

ISCHIOANAL ABDOMINOPERINEAL EXCISION In some patients, the rectal tumour is locally advanced and may infiltrate or even perforate the pelvic floor, i.e. the levator muscle. In other patients, a perianal abscess may be the presenting feature of a perforated low rectal cancer, and after drainage a fistula may persist between the low rectum and the perianal skin. In a few very low tumours, the growth may extend into the perianal skin (Figure 8.5). In these instances, extralevator APE may not be sufficient to achieve a safe tumourfree CRM, and ischioanal APE is usually required to obtain an oncologically secure margin. In this situation, the levator muscle must be removed covered with ischioanal fat, and the ischioanal fat must be removed to include the perianal fistula, which may

HEBK001-C08_p124-139.indd 134

Figure 8.5. In some very low and advanced tumours the cancer growth may infiltrate the perianal skin. In these patients a wide excision of the perianal skin and an ischio-anal APE is necessary to achieve a potentially curative resection.

contain tumour cells. Therefore, ischioanal APE is a valid procedure in these special situations. The abdominal part of ischioanal APE is exactly equivalent to the abdominal part of extralevator APE. Thus, the dissection stops just above the levator muscle and leaves the mesorectum attached to the pelvic floor (Figure 8.6A). When the abdominal part of the procedure is completed, with closure of the abdominal wall and formation of a ­colostomy, the patient is turned into the prone jack-knife ­position.

Perineal Part of Ischioanal Abdominoperineal Excision After proper preparation of the skin of the perineum, lower sacrum, the medial parts of the buttocks and the vagina in female patients, a double purse-string suture is placed to close the anus. The area of the skin incision in ischioanal APE depends on the extent of tumour involvement of the skin. Any tumour infiltration or fistula opening must be included in the excised skin area with a margin of at least 2–3 cm. As soon as the incision deepens into the subcutaneous space, the dissection should be directed laterally towards the ischial tuberosity and progress on to the fascia of the internal obturator muscle. Thus, ­contrary to extralevator APE, the dissection does not follow the external sphincter and levator muscle but instead is carried along the fascia of the internal obturator muscle. The dissection is performed along this plane up to where the levator muscle is inserted into the internal obtu-

08/02/13 5:54 PM

ISCHIOANAL ABDOMINOPERINEAL EXCISION  135 Figure 8.6a. The pelvic dissection in an ischio-anal APE is carried along outside the mesorectal fascia but stops at the top of the levator muscle (blue line). The perineal dissection is directed towards tuber os ischii and follows the obturator internus muscle fascia, in order to remove the fat in the ischio-anal compartment en bloc. The size of the skin incision depends on the extent of tumour involvement in the skin and may be extensive (left side) or similar to the skin incision in an extra-levator APE (right side).

A

Levator ani Obturator internus Ischio-anal fossa External sphincter Internal sphincter B

Figure 8.6b. Photograph showing a fresh specimen after an ischio-anal APE. In very advanced tumours, infiltrating the perineal skin, a wide skin excision and a complete clearance of both ischio-anal compartments may be necessary for a potentially curative operation.

rator muscle and hence includes the entire fat compartment of the ischioanal space. This dissection can be performed unilaterally or bilaterally, depending on the extent of tumour growth. When the dissection up to this level is completed, the sacrococcygeal junction is incised and the pelvic cavity is entered in the same fashion as with extralevator APE. The subsequent dissection is also similar to that of extralevator APE, as the levator muscles are divided along the fascia of the internal obturator muscle on to the

HEBK001-C08_p124-139.indd 135

prostate in male patients or the vagina in female patients. Once the specimen has been brought out of the pelvic cavity, the anterior and lateral dissection along the prostate or vagina is also carried out as in extralevator APE. As mentioned above, the difference between extralevator APE and ischioanal APE is that the fat in the ischioanal space is resected en bloc and attached to the levator muscle (Figure 8.6B). This procedure is very similar to what Miles described in 1908 in his original paper in the Lancet.

08/02/13 5:54 PM

136  Abdominoperineal excision of the rectum

LAPAROSCOPIC APPROACH TO ABDOMINOPERINEAL EXCISION The abdominal procedure in APE may be performed with minimally invasive techniques, either laparoscopic or robot-assisted. This approach is gaining increasing popularity, and it may well be that minimally invasive methods for the abdominal part of the operation will become predominant in the near future. Although some details differ between the open and minimally invasive approach, the main principles described above for the abdominal part of APE should be adopted, irrespective of the surgical method used. The principles and details of minimally invasive procedures in rectal cancer are described in Chapter 10.

Prone or supine position during perineal dissection In Miles’ original description of the APE procedure, the perineal part of the operation was performed in the right lateral and semiprone position after completion of the abdominal phase and the stoma creation. Subsequently, synchronous combined APE gained popularity and became standard for the vast majority of surgeons. With synchronous combined APE, the perineal part of the operation is performed with the patient in the supine Lloyd Davies position, and the final part of the pelvic dissection from the abdominal side is often performed simultaneously with the perineal dissection. In addition, historically this perineal dissection was often done blunt and often by the most junior surgeon in the operating team. As mentioned above, the consequence of this approach was repeatedly that the excised specimen had a waist just proximal to the puborectal sling and often also an inadvertent perforation or an involved circumferential resection margin. Due to the increasing awareness of the problems associated with synchronous combined APE, extralevator APE was introduced as a different and more radical procedure, performed via the posterior perineal approach in the prone jack-knife position, which closely mirrors the original Miles operation.22 The main purpose of this prone jack-knife position is to optimize the visibility of the operative field and to have full control of the perineal and pelvic floor anatomy. As a result, the correct anatomi-

HEBK001-C08_p124-139.indd 136

cal planes can be followed more easily and the risk of perforation or involved margins of the specimen by tumour can be reduced. In one study, it was shown that the risk of perforation in extralevator APE was significantly lower when perineal dissection was carried out in the prone jack-knife position rather than the lithotomy or Lloyd Davies position (6.4 per cent v. 20.6 per cent).23 As perforation of the specimen is a wellknown risk factor for adverse prognosis, it may be that the prone jack-knife position improves oncological outcomes. Thus, performing the perineal part of APE with the patient in the prone jack-knife position may potentially improve prognosis. In addition, this position gives an excellent view, which facilitates teaching of this part of the operation. With the patient in this position and with the legs spread apart, the surgeon stands between the legs and has one assistant on each side. The operative field is well exposed to all three surgeons, and trainees can perform a step-by-step approach to learn how to do the operation, with full control by the responsible surgeon. Many surgeons who have adopted extralevator APE for low rectal cancer also favour the jack-knife position for the perineal part of the procedure. Some surgeons still use the lithotomy or Lloyd Davies position, however, for the perineal part, even though they perform extralevator APE.24 There has been some debate about the pros and cons of the two approaches, and there have also been reports on excellent oncological outcomes after APE with the perineal dissection performed with patients in the supine position.25 It is likely that the position of the patient is not crucial, provided that skilled and properly trained surgeons perform a meticulous dissection to create a perfect specimen in the extralevator plane.26

REMOVAL OF THE COCCYX With the conventional synchronous combined APE, the pelvic dissection during the abdominal part of the operation is carried along the mesorectal fascia all the way down to the pelvic floor and beyond the coccyx. With this approach, there is no rationale for removing the coccyx while performing the perineal part of the operation, because the mesorectum has already been detached from the levator muscle of

08/02/13 5:54 PM

PERINEAL RECONSTRUCTION  137

the pelvic floor. Therefore, the division of the levator muscle during conventional APE is often instigated between the tip of the coccyx and puborectal sling and continued forwards bilaterally on to the prostate or the vagina. With this approach, a limited amount of the pelvic floor muscle is removed en bloc with the rectum and, as mentioned earlier, the specimen frequently has a waist just above the puborectal sling. During extralevator APE and ischioanal APE, the coccyx is often removed en bloc with the rectum. There are two main reasons for doing this: first, the levator muscle is attached to the anterior surface of the coccyx and continues cranially as the presacral fascia. If the levator muscle is to be completely removed and fixed to the mesorectum, as in extralevator APE, then the coccyx needs to be removed. Cutting the levator muscle below the tip of the coccyx will inevitably leave a part of the levator muscle behind. Second, it is sometimes difficult to bring out the specimen from the pelvic cavity through a narrow opening in the pelvic floor. Especially in male patients with a narrow pelvis and a large mesorectum, it can be hard to bring out the specimen before starting the challenging dissection of the anterior part of the rectum and mesorectum off the prostate or the vagina. By dividing between the sacrum and the coccyx and excising the coccyx en bloc, the surgeon creates a wider opening of the pelvic floor, which facilitates delivery of the specimen and the anterior dissection. Some authors have advocated an even more extensive resection including the last sacral vertebra to achieve better access and control of the perineal part of the operation.27 With the patient in the supine, lithotomy or Lloyd Davies position, removal of the coccyx is difficult due to limited access to the natal cleft. Therefore, surgeons who are proponents of removal of the coccyx usually also perform the perineal part of extralevator APE in the prone jack-knife position.

Using a mesh to prevent parastomal hernia One major problem after APE with formation of a permanent stoma is the subsequent development of a parastomal hernia in many patients, which by definition is an incisional hernia related to the presence of a stoma in the abdominal wall. This is

HEBK001-C08_p124-139.indd 137

a common clinical problem, reported to occur in 30–80 per cent of patients with a colostomy. The wide variation in incidence is probably related to the definition of a parastomal hernia, the method used to diagnose the condition and the length of follow-up. Although many parastomal hernias are only cosmetically disturbing or associated with mild symptoms, about 30–40 per cent of patients with parastomal hernias may require surgical repair of the hernia due to pain, obstruction, bleeding, increasing protrusion, poorly fitting stoma bag, faecal leakage or incarceration. Several different surgical methods are used to treat parastomal hernias, including relocation, primary closure of the fascia and mesh repair, but high complication and recurrence rates have been reported. The use of a prosthetic mesh to repair a parastomal hernia and to reconstruct the abdominal wall has gained increasing popularity and improved the rate of successful revisions, but these methods are not without problems. Erosion of the bowel (with subsequent infection and fistula formation) and recurrent hernia development have been reported. The high risk of parastomal hernia and the suboptimal results after repair have led to attempts to prevent its occurrence. Several studies have shown that the use of a polypropylene mesh at the time of stoma formation is safe and may prevent the occurrence of a parastomal hernia. A review of three randomized controlled trials and three prospective observational series reported parastomal hernia rates of 55 per cent in patients who did not have a mesh and 7 per cent in patients in whom a mesh was used. Postoperative morbidity rates were similar, irrespective of whether a mesh was used.28 Although the placement of a mesh, preferably by a sublay technique, seems promising, there are no data on long-term results in terms of hernia recurrence or possible late complications. Despite this, we recommend the use of a 10 3 10 cm lightweight synthetic mesh around the bowel behind the rectus muscle and anterior to the rectus sheet.

PERINEAL RECONSTRUCTION When the first major paper on APE was published by Miles in 1908, he described the closure of the perineal wound: ‘Finally, the skin margins are brought together with sutures and a large drainage-

08/02/13 5:54 PM

138  Abdominoperineal excision of the rectum

tube is inserted in the anterior and the posterior extremities of the median incision.’ Miles later abandoned this technique due to the high risk of perineal wound infections and the ensuing mortality, and instead the defect was left open and allowed to heal by secondary intention. The practice of leaving the perineal wound open was applied for many decades but is now used by very few surgeons. Although this method avoids some of the infectious complications associated with primary closure, the morbidity associated with a large, slowly healing wound is significant. Therefore, primary closure of the perineal wound has been the most common method of ­perineal reconstruction after synchronous combined APE. Although the clinical course after primary closure is often uneventful, complications due to the perineal wound are still one of the major problems associated with the conventional type of APE, especially in patients who have received preoperative radiotherapy. Wound problems have been reported in up to 50 per cent of patients receiving preoperative radiotherapy after APE with primary wound closure.29 Wound infection and delayed healing are the most common complications. These problems may become even more frequent in patients who have received a combination of preoperative radiotherapy and extralevator APE, with a more extensive excision of the pelvic floor. A variety of surgical alternatives to primary closure have been used to reconstruct the pelvic floor and to reduce the wound healing problems after APE. These procedures include omental pedicle flaps (omentoplasty) and different local rotational musculocutaneous flaps. Data from controlled studies support the use of musculocutaneous flaps for single-stage reconstruction after APE, especially in the presence of chemoradiotherapy. Several reports using the rectus abdominus, gluteus maximus or gracilis musculocutaneous flaps have been published.30,31 Rectus abdominus flaps have been most commonly used. The reported overall complication rates vary from 10 per cent to 50 per cent, while healing rates during follow-up vary from 95 per cent to 100 per cent.32,33 Most studies are small, with fewer than 50 reported patients; the small number of patients, the varying definition of wound complications and the varying follow-up times probably explain the difference in complication rates.

HEBK001-C08_p124-139.indd 138

Some experience with biological mesh reconstruction of the pelvic floor has also been reported. This option seems feasible with a reasonable complication rate.34 The number of reports is limited, however, and long-term results from biological mesh reconstruction of the pelvic floor are still lacking. In fact, there is no standard solution for pelvic floor reconstruction after APE, and the method used must be tailored according to the patient and the extent of excision. Thus, primary closure is generally appropriate after intersphincteric APE, while some kind of mesh or flap reconstruction is often used after extralevator APE. After ischioanal APE, a flap reconstruction is almost always necessary, especially if the excision of skin has been extensive. It is recommended to assess each patient carefully before surgery to determine the suitable type of pelvic floor reconstruction and to establish collaboration with a plastic surgeon for reconstruction after wider excisions.

References   1. Miles WE. A method of performing abdomino-perineal excision for carcinoma of the rectum and of the terminal portion of the pelvic colon. Lancet 1908; 2: 1812–13.   2. Collins DC. End-results of the Miles’ combined abdominoperineal resection versus the segmental anterior resection: a 25-year postoperative follow-up in 301 patients. Am J Proctol 1963; 14: 258–61.   3. Fick TE, Baeten CG, von Meyenfeldt MF, Obertop H. Recurrence and survival after abdominoperineal and low anterior resection for rectal cancer without adjunctive therapy. Eur J Surg Oncol 1960; 16: 105–8.   4. Groves RA, Harrison RC. Carcinoma of the rectum and lower sigmoid colon: abdominoperineal or anterior resection? Can J Surg 1962; 5: 393–403.   5. Slanetz CA, Herter FP, Grinnell RS. Anterior resection versus abdominoperineal resection for cancer of the rectum and rectosigmoid: an analysis of 524 cases. Am J Surg 1972; 123: 110–17.   6. Vandertoll DJ, Beahrs OH. Carcinoma of the rectum and low sigmoid: evaluation of anterior resection in 1766 favourable lesions. Arch Surg 1965; 90: 793–8.   7. Schmitz RL, Nelson PA, Martin GB, Boghossian HM. Synchronous (two-team) abdominoperineal resection of the rectum. AMA Arch Surg 1958; 77: 492–7.   8. Påhlman L, Glimelius B. Local recurrences after surgical treatment for rectal carcinoma. Acta Chir Scand 1984; 150: 331–5.   9. Heald RJ, Husband EM, Ryall RD. The mesorectum in rectal cancer surgery – the clue to pelvic recurrence. Br J Surg 1982; 69: 613–16.

08/02/13 5:54 PM

References   139 10. Macfarlane JK, Ryall RD, Heald RJ. Mesorectal excision for rectal cancer. Lancet 1993; 341: 457–60. 11. Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001; 345: 638–46. 12. Martling AL, Holm T, Rutqvist LE, Moran BJ, Heald RJ, Cedemark B. Effect of a surgical training programme on outcome of rectal cancer in the County of Stockholm: Stockholm Colorectal Cancer Study Group, Basingstoke Bowel Cancer Research Project. Lancet 2000; 356: 93–6. 13. Wibe A, Syse A, Andersen E, Tretli S, Myrvold HE, Soreide O. Oncological outcomes after total mesorectal excision for cure for cancer of the lower rectum: anterior vs. abdominoperineal resection. Dis Colon Rectum 2004; 47: 48–58. 14. Marr R, Birbeck K, Garvican J, et al. The modern abdominoperineal excision: the next challenge after total mesorectal excision. Ann Surg 2005; 242: 74–82. 15. Den Dulk M, Putter H, Collette L, et al. The abdominoperineal resection itself is associated with an adverse outcome: the European experience based on a pooled analysis of five European randomised clinical trials on rectal cancer. Eur J Cancer 2009; 45: 1175–83. 16. Nagtegaal ID, van de Velde CJ, Marijnen CA, van Krieken JH, Quirke P. Low rectal cancer: a call for a change of approach in abdominoperineal resection. J Clin Oncol 2005; 23: 9257–64. 17. Eriksen MT, Wibe A, Syse A, Haffner J, Wiig JN. Inadvertent perforation during rectal cancer resection in Norway. Br J Surg 2004; 91: 210–6. 18. Pollack J, Holm T, Cedermark B, Holmstrom B, Mellgren A. Long-term effect of preoperative radiation therapy on anorectal function. Dis Colon Rectum 2006; 49: 345–52. 19. Morris E, Quirke P, Thomas JD, Fairley L, Cottier B, Forman D. Unacceptable variation in ­abdominoperineal excision rates for rectal cancer: time to intervene? Gut 2008; 57: 1690–7. 20. Birbeck KF, Macklin CP, Tiffin NJ, et al. Rates of circumferential resection margin involvement vary between surgeons and predict outcomes in rectal cancer surgery. Ann Surg 2002; 235: 449–57. 21. Radcliffe A. Can the results of anorectal (abdominoperineal) resection be improved: are circumferential resection margins too often positive? Colorectal Dis 2006; 8: 160–7.

HEBK001-C08_p124-139.indd 139

22. Holm T, Ljung A, Haggmark T, Jurell G, Lagergren J. Extended abdominoperineal resection with gluteus maximus flap reconstruction of the pelvic floor for rectal cancer. Br J Surg 2007; 94: 232–8. 23. West NP, Anderin C, Smith KJ, Holm T, Quirke P. Multicentre experience with extralevator abdominoperineal excision for low rectal cancer. Br J Surg 2010; 97: 588–99. 24. Martijnse IS, Dudink RL, West NP, et al. Focus on extralevator perineal dissection in supine position for low rectal cancer has led to better quality of surgery and oncologic outcome. Ann Surg Oncol 2012; 19, 786–93. 25. De Campos-Lobato LF, Stocchi L, Dietz DW, Lavery IC, Fazio VW, Kalady MF. Prone or lithotomy positioning during an abdominoperineal resection for rectal cancer results in comparable oncologic outcomes. Dis Colon Rectum 2011; 54: 939–46. 26. Bebenek M. Abdominosacral resection is not related to the risk of neurological complications in patients with low-rectal cancer. Colorectal Dis 2009; 11: 373–6. 27. Janes A, Cengiz Y, Israelsson LA. Preventing parastomal hernia with a prosthetic mesh: a 5-year followup of a randomized study. World J Surg 2009; 33: 118–21, 122–3. 28. Bullard KM, Trudel JL, Baxter NN, Rothenberger DA. Primary perineal wound closure after preoperative radiotherapy and abdominoperineal resection has a high incidence of wound failure. Dis Colon Rectum 2005; 48: 438–43. 29. Khoo AK, Skibber JM, Nabawi AS, et al. Indications for immediate tissue transfer for soft tissue reconstruction in visceral pelvic surgery. Surgery 2001; 130: 463–9. 30. Nisar PJ, Scott HJ. Myocutaneous flap reconstruction of the pelvis after abdominoperineal excision. Colorectal Dis 2009; 11: 806–16. 31. Mcallister E, Wells K, Chaet M, Norman J, Cruse W. Perineal reconstruction after surgical extirpation of pelvic malignancies using the transpelvic transverse rectus abdominal myocutaneous flap. Ann Surg Oncol 1994; 1: 164–8. 32. Tobin GR, Day TG. Vaginal and pelvic reconstruction with distally based rectus abdominis myocutaneous flaps. Plast Reconstr Surg 1988; 81: 62–73. 33. Christensen HK, Nerstrom P, Tei T, Laurberg S. Perineal repair after extralevator abdominoperineal excision for low rectal cancer. Dis Colon Rectum 2011; 54: 711–7.

08/02/13 5:54 PM

9 Laparoscopic surgery Katharine E. Bevan and Tom D. Cecil

Introduction Since the 1940s surgeons have used a laparoscope to inspect the peritoneal cavity. Advances in equipment technology coupled with pioneering approaches have moved the role of the laparoscope from a diagnostic to an increasingly therapeutic tool in managing abdominal pathology. The first appendicectomy was performed laparoscopically in the 1980s, closely followed by laparoscopic cholecystectomy in 1985. Laparoscopic colonic resection, particularly for cancer, has been slower in development and uptake, however. This is unsurprising, considering the technical challenges involved, including the necessity to operate in all four quadrants of the abdomen and the complexity of the arterial and venous anatomy of the colon. The need to resect and remove a sometimes substantial specimen, and subsequently perform a difficult intestinal anastomosis, are further challenges. When considering mid- to lower rectal cancers, there are additional anatomical and technical difficulties. These include bulky tumours in a narrow pelvis, retraction and visualization (especially with the distal dissection), preservation of autonomic nerves, accurate circumferential resection (without jeopardizing the mesorectal envelope) and technological challenges (due to current limited angulation of laparoscopic staplers). Early anxiety regarding port site metastasis1 combined with a lack of randomized controlled

HEBK001-C09_p140-153.indd 140

trials and concerns regarding long-term oncological outcomes culminated in a logical reluctance to embrace laparoscopic colorectal cancer surgery. There are now a number of randomized controlled trials comparing laparoscopic and open colon cancer surgery. The short-term benefits for laparoscopic colonic resection have been summarized in a Cochrane review, with the main disadvantage being that the surgery is more time-consuming.2 A subsequent Cochrane review reported that laparoscopic surgery was feasible for upper rectal cancers, but the overall conclusion was that more randomized controlled trials were required to assess long-term outcomes.3 A further Cochrane review principally looking at the safety and efficacy of laparoscopic versus open total mesorectal excision (TME) for rectal cancer concluded that ‘the limited evidence suggests that laparoscopic TME has clinically relevant short-term advantages in selected patients with rectal cancer’.4 Controversy continues regarding the role of the laparoscope in rectal cancer as initial reports were from enthusiasts with small series, with little emphasis on the selection of appropriate patients or the challenges to be overcome before widespread adoption of safe laparoscopic rectal cancer surgery.5 The Conventional versus Laparoscopic-Assisted Surgery in Colorectal Cancer (CLASICC) trial was the first randomized controlled trial to include patients with rectal cancers. Outcomes were evaluated separately from colon cancer outcomes. The

08/02/13 5:50 PM

MODERN MANAGEMENT OF RECTAL CANCER AND THE ROLE OF THE LAPAROSCOPE  141

CLASICC trial highlighted concerns with regard to an increased circumferential resection margin (CRM) positivity and suggested that the routine use of the laparoscope for rectal cancer was not yet justified.6 In August 2006, the National Institute for Health and Clinical Excellence (NICE) issued revised guidelines recommending that laparoscopic surgery (including laparoscopically assisted surgery) should be offered as an alternative to open surgery if both techniques are suitable and the appropriate surgical skills are available.7 This chapter focuses on some issues unique to laparoscopic rectal cancer surgery. We review the evidence from randomized controlled trials and summarize our approach to laparoscopic rectal resection. The particular problems needed to be addressed in rectal cancer are explored to give a balanced and safe approach to the use of the laparoscope in rectal cancer and in particular TME.

of visceral metastases9 and useful for lymph node assessment.10 Peritoneal carcinomatosis, especially if localized and low volume,11 and to a lesser extent local tumour extension, continue to be a challenge to diagnose, however, despite major advances in imaging. Thus, there continues to be a role for a more sensitive modality such as laparoscopy, and in colorectal surgery this is particularly so for detection of early, potentially treatable localized carcinomatosis.12 A pivotal role for staging laparoscopy in colorectal cancer surgery is undoubtedly a useful initial investigation when a laparoscopic resection is being contemplated or planned. The ultimate decision as to whether laparoscopic resection is safe and feasible, or whether open surgery is required, is best made at this point, and experienced surgeons can rapidly decide on the appropriate modality of surgery in most cases.13

MODERN MANAGEMENT OF RECTAL CANCER AND THE ROLE OF THE LAPAROSCOPE

Patients with rectal cancer may, in some circumstances, require a defunctioning stoma, either as palliative treatment or before neoadjuvant therapy for very symptomatic locally advanced tumours. A stoma may be necessary for obstruction, uncontrollable diarrhoea, tenesmus or bleeding and often before chemoradiation, as either down-staging or definitive care. There is considerable variability in the frequency of defunctioning stoma formation before neoadjuvant therapy,14 but it undoubtedly has a role and may be needed before or during neoadjuvant therapy.15 The laparoscope can be used to perform a defunctioning ileostomy or colostomy alone, or as part of a laparoscopic colorectal resection, or to reduce the septic complications secondary to anastomotic leak. In addition, the laparoscopic approach should enable correct identification of the proximal and distal limbs and thus avoid ‘bringing out the wrong limb’, as surgeons of an older generation have experienced when fashioning a defunctioning trephine stoma. Some reports suggest shorter time to passage of flatus and faeces, earlier resumption of a liquid and solid diet, decreased morphine requirements, a reduction in 30-day morbidity, and earlier discharge in patients having a stoma formed ­laparoscopically.16 As a consequence, patients in whom defunctioning alone is required, without resection or anastomosis, are likely to achieve maximal benefit from a

The modern management of rectal cancer involves accurate preoperative staging, appropriate neoadjuvant therapy in selected cases (either preoperative radiotherapy or chemoradiotherapy), optimal surgical excision and detailed pathological assessment of the excised specimen.5 All these aspects interact with the role of the laparoscope in the management of rectal cancer. In the absence of longer-term outcome data from larger clinic trials, a surrogate marker is required to assess operative success. Macroscopic and microscopic assessment of the specimen, being a prognostic marker for local recurrence as reported in the Medical Research Council (MRC) CR07 trial, may be the best method for comparing early oncological results from both open and laparoscopic surgery.8

Staging and Diagnostic Laparoscopy Despite advances in non-invasive imaging, laparoscopy continues to have an important role in some aspects of staging rectal cancer. Cross-sectional imaging is highly sensitive and specific for detection

HEBK001-C09_p140-153.indd 141

Laparoscopic Stoma Formation

08/02/13 5:50 PM

142  Laparoscopic surgery

laparoscopic approach due to expedited recovery and therefore minimal interruption to treatment, in comparison with open surgery. There have been reports, however, of obstructive complications after any proximal defunctioning stoma. A laparoscopic colostomy is also a very useful training operation as part of the learning curve for more complex surgery involving mobilization of the colon.

Abdominoperineal Excision of the Rectum Abdominoperineal excision (APE) was the first resectional laparoscopic colorectal procedure to be performed for malignancy. As the oncological part of the operation is carried out by an open surgical technique from the perineum, it could be confidently attempted without the concern for a major compromise in the oncological outcome. Patients with low rectal cancer undergoing APE, particularly with the recent developments in extralevator APE, have perhaps most to gain by a laparoscopic approach.17 In extralevator APE, the main oncological part of the procedure is performed via the perineum, there is no bowel anastomosis, and the specimen can be removed through the perineum, such that no abdominal incision is required.17 The upper part of the TME plane is developed and completed laparoscopically, but without the need to continue the dissection very far distally. Subsequently, the perineal plane is developed, usually after completing the colostomy, closing the abdomen and turning the patient into the prone jack-knife position. The dissection proceeds until the perineal operation meets the extent of the abdominal operation. Additionally, the tumour requiring an APE is generally lower in the rectum, well below the peritoneal reflection, compared with cases where reconstruction by an anterior resection is being performed, and is therefore less likely to obscure the laparoscopic view or be subject to tumour rupture by traction or retraction. With the success of laparoscopic APE, enthusiasts deemed it as a natural progression to extend the abdominal surgery deeper into the pelvis in suitable patients to mobilize the rectum and subsequently proceed to reconstruction to achieve laparoscopic anterior resection.18,19

HEBK001-C09_p140-153.indd 142

Evidence for the laparoscopic approach to rectal cancer Short-Term Outcomes Although there are some conflicting findings, many of the short-term benefits of laparoscopic colonic resection have also been reported for laparoscopic rectal resection.4 These include the earlier return of bowel function and subsequently resumption of diet,6,20–23 shorter length of stay,6,20,23 quicker return to household activities,20 reduced pain and analgesia use,20,21 and decreased blood loss,21–23 sometimes equating to a reduction in the requirement for blood transfusion (Table 9.1).23 These findings are unsurprising and are consistent with the evidence of a lesser inflammatory response and decreased immunological disturbance in laparoscopic colorectal surgery; some of these beneficial effects may be due to less small bowel manipulation.24 On the negative side, the duration of the laparoscopic operation is longer6,20–23 and the cost of surgery greater.2,20,23

Lymph Node Harvest Lymph node harvest has often been considered a shortterm marker for oncological equivalence of open and laparoscopic resection specimens. A small randomized controlled trial included 73 patients (39 open surgery, 34 laparoscopic TME) and reported equivalent lymph node harvest in both groups.25 Subsequently other studies have reported similar findings.6,20–23

RESECTION QUALITY AND CIRCUMFERENTIAL RESECTION MARGIN It is now widely accepted that TME, as described by Heald in 1988, is the treatment of choice in ­resectable rectal cancer.26 Provided a clear CRM can be achieved, TME results in an improvement in cancer-specific survival. The effect of the quality of the resected specimen is also emerging as a prognostic indicator for local recurrence after rectal cancer. One group showed that both the prospective histopathological assessment of the plane of

08/02/13 5:50 PM

RESECTION QUALITY AND CIRCUMFERENTIAL RESECTION MARGIN   143 Table 9.1 Short- and longer-term outcome results reported in the five largest randomized controlled trials comparing laparoscopic and open surgery for rectal cancer. Study

Kang et al.21

Lujan et al.22

Braga et al.23

Jayne et al.29

Leung et al.20

Open surgery (n)

170

103

85

253

200

Laparoscopic

170

101

83

128

203

Greater in open

Greater in open

Greater in open

NS

NS

Increased in

.45.7 min in

surgery (n) Blood loss

group, significant Operative time

.48 min in

group, P 5 0.001 .21 min in

group, P 5 0.0002 .53 min in

laparoscopy

laparoscopy

laparoscopy

laparoscopy

group, P , 0.0001

group, P 5 0.020

group, P 5 0.0001

group

laparoscopy group, P , 0.001

Conversion rate (%)

1.2

7.9

7.2

34 (rectal)

23.2

Analgesic

Reduced in

NR

NR

NR

Reduced in

requirements

laparoscopy

laparoscopy

group, P , 0.001

group, P , 0.0001 Return to bowel

Reduced time in

function

laparoscopy

Return to diet

Reduced time in

NR

NR

NS

NS

,1 day in

Reduced time in

NS

Reduced time in

group, P , 0.0001

Length of stay

laparoscopy

laparoscopy

laparoscopy

laparoscopy

group, P , 0.0001

group, NS

group, P , 0.0001

group, P , 0.001

,1 day in

,1 day in

,3.6 days in

,2 days in

laparoscopy

laparoscopy

laparoscopy

laparoscopy rectal

group, NS

group, NS

group, P 5 0.004

group

Reduced in laparoscopy group, P , 0.001

Lymph node yield

NS

NS

NS

NS

NS

CRM positivity (%)

NS

NS

NS

12 (laparoscopy) v.

NR

6 (open) 30-day morbidity

NS

NS

NS

NS

NS

Mortality

NS

NS

NS

NS

NS

Local recurrence at

NR

5.3 (open) v. 4.8

NS at 3 years

10.1, NS

6.6 (laparoscopy)

NS

52.1 (open) v. 53.2

NS

5 years (%) Disease-free survival

NR

at 5 years (%) Overall survival at

(months)

81 (open) v. 84 (laparoscopy), NS

NR

5 years (%) Mean follow-up

v. 4.1 (open), NS

(laparoscopy), NS

75 (open) v. 72

(laparoscopy), NS NS

(laparoscopy), NS NR

34 (open) v. 32

52.9 (open) v. 60.3

NS

(laparoscopy), NS 53.6

56.3

52.7

(laparoscopy)

CRM, circumferential resection margin; NR, not reported; NS, not statistically significant.

surgery and the CRM margin were independent prognostic markers for local recurrence. Clear differences were noted with suboptimal surgery, with a statistically significant increase in the local recurrence rate at 3 years. Circumferential resection margin positivity is still one of the most important predictors of increased risk of local recurrence and decreased 3-year survival, regardless of the plane

HEBK001-C09_p140-153.indd 143

of dissection.27 It would therefore seem appropriate to use these two standards (plane of surgery achieved and CRM) as a means of quality control when comparing laparoscopic and open rectal cancer surgery. The evidence base for equivalent clinical outcomes in rectal cancer treated laparoscopically is limited but emerging.

08/02/13 5:50 PM

144  Laparoscopic surgery

The CLASICC study initially demonstrated concerns with regard to CRM. The study reported more positive CRMs in the laparoscopic group compared with the open anterior resection group (12 per cent v. 6 per cent). Although the results were not significantly different, they were concerning enough to recommend caution when treating rectal cancer ­laparoscopically.6 These concerns have been somewhat allayed by the 3and 5-year follow-up data, which have shown that the trend towards positive CRM did not translate into an increase in local recurrence rates.28,29 Five randomized controlled trials, all with at least 80 patients in each arm, have reported longer-term follow-up data. Three of these trials reported on rectal cancer outcomes alone.21–23 One trial included both colonic and rectal cancer and reported on the rectal subset separately.6 The final trial recruited both high rectal and sigmoid tumours, but the heterogeneity makes interpretation difficult, particularly as the tumours were not analysed independently.20 None of these trials demonstrated a statistically significant difference in CRM positivity (Table 9.2). Kang’s group looked at macroscopic quality of the TME dissection and showed no difference between open and laparoscopic surgery. Three of the randomized controlled trials have published long-term survival figures.6,22,23 These studies have shown no statistical difference in local recurrence, ­disease-free survival or overall survival between surgical approaches (see Table 9.1). Some groups have compared the quality of the mesorectal dissection specimen and reported no statistical difference between open and laparoscopic

resections.30,31 A comparison of the macroscopic quality of specimens after laparoscopic32 and open33 TME showed that there was a more complete TME with intact visceral pelvic fascia in the laparoscopic group. The investigators concluded that the improved visualization at laparoscopy may result in improved macroscopic specimens.34 Although the rates of local recurrence vary widely in the published literature, no significant differences between laparoscopic and open surgery were reported in the 2006 Cochrane review4 or in several randomized controlled trials.20,22,23 It remains to be seen whether long-term follow-up consistently gives results equivalent to those reported for open TME of less than 5 per cent in curative cases.35

Anastomotic leak and its influence on local recurrence An anastomotic leak after restorative resection of rectal cancer continues to be a major issue whether laparoscopic or open surgery is performed. There have been reports that anastomotic leak may significantly increase the local recurrence rates,36 with poorer overall and cancer-specific survival.37 Therefore, it is imperative that we have at least equal or preferably lower anastomotic leak rates when performing laparoscopic surgery for rectal cancer. None of the randomized controlled trials showed a statistically significant difference between the open and laparoscopic groups when comparing leak rates

Table 9.2 Comparison of leak rate within randomized controlled trials comparing laparoscopic and open surgery for rectal cancer. Trial

Kang et al.21

Lujan et al.22

Braga et al.23

Jayne et al.29

Leung et al.20

Open (n)

170

103

85

132

200

Laparoscopic (n)

170

101

83

160

202

Inclusions

,9 cm from anal

Mid- and low rectal

,15 cm from anal

.5 cm from anal

Rectosigmoid

verge

cancers: mean

verge

verge (67%

6.2 cm (open) v.

open v. 79%

5.5 cm (laparoscopy)

laparoscopy)

from anal verge Stoma formation (%)

88.4 (open) v. 91 (laparoscopy)

Leak rate (%)

0 (open) v. 1.2 (laparoscopy)

59 (open) v. 62 (laparoscopy) 12 (open) v. 6 (laparoscopy)

24.7 (open) v. 26.5

NR

NR

7 (open) v. 8

2 (open) v. 0.5

(laparoscopy) 7 (open) v. 6 (laparoscopy)

(laparoscopy)

(laparoscopy)

NR, not reported.

HEBK001-C09_p140-153.indd 144

08/02/13 5:50 PM

OUR APPROACH TO LAPAROSCOPIC ANTERIOR RESECTION  145

(see Table 9.2).6,20–23 Liberal use of defunctioning, either in all patients having TME or at a minimum all patients at increased risk of anastomotic leak (e.g. male patients, patients with a high body mass index, patients who have had neoadjuvant treatment, patients with a low or complex anastomosis), may significantly reduce the clinical consequences of a leak. There may be a case for routine defunctioning of all anastomoses after any mid- or low rectal cancer resection, as in the Comparison of Open versus Laparoscopic Surgery for Mid and Low Rectal Cancer after Neoadjuvant Chemoradiotherapy (COREAN) study.21 It has been noted, however, that the addition of a defunctioning stoma significantly increases length of stay after laparoscopic resection, thereby partly negating this potential benefit of the laparoscopic approach.21

Length of stay Reduction in length of stay is often quoted as being a particular benefit of laparoscopic surgery. Interestingly, two of the five randomized controlled trials did not demonstrate any statistically significant reduction in length of stay (see Table 9.1).21,22 Braga and colleagues did report a reduction in length of stay with laparoscopic surgery, but this study included high rectal cancers and had a subsequently low stoma rate.23 Leung and colleagues also demonstrated a reduction in length of stay, but their study included mainly rectosigmoid ­tumours.20 The studies of Kang and colleagues21 and Lujan and colleagues22 showed no significant reduction in length of stay. Of note, all patients in the COREAN trial and over 70 per cent of patients in the Spanish trial were operated on after neoadjuvant chemoradiotherapy, and consequently the rates of defunctioning ileostomy were very high (91 per cent and 62 per cent, respectively), which has previously been shown to increase length of stay.37

Morbidity and mortality The currently reported randomized controlled trials have not shown a significant difference with regard to perioperative mortality.20–23,28 Some studies have shown lower morbidity in laparoscopic patients,38,39 but this is a minority and many studies have shown

HEBK001-C09_p140-153.indd 145

equivalence.6,20–23 Taken in conjunction with studies on colonic resection,2,40 the available data suggest possible benefits in favour of laparoscopic surgery in frailer and older patients due to fewer complications in the laparoscopic group, especially in people aged over 70 years.40

OUR APPROACH TO LAPAROSCOPIC ANTERIOR RESECTION Preoperative Assessment The preoperative investigations that we carry out for laparoscopic anterior resection for rectal cancer are almost identical to those for open surgery. Staging computed tomography (CT) of the chest and abdomen and magnetic resonance imaging (MRI) of the pelvis are essential. Endoscopic assessment of the entire colon (unless an obstructing tumour limits progress) allows biopsy confirmation of carcinoma and allows assessment for synchronous pathology. Unlike open surgery, tattooing of the location of the tumour is essential. For colon cancers above the peritoneal reflection, we recommend marking 5 cm distal to the tumour, and usually in three separate locations around the circumference of the bowel at this level. For rectal cancer, however, this is unhelpful, as the tumour is below the peritoneal reflection and the tattoo can stain the mesorectal and fascial planes, making dissection more difficult. For rectal cancer, we rely on MRI assessment and digital examination or rigid sigmoidoscopy for localization of the tumour at surgery. We recommend full bowel preparation for any total mesorectal resection to prevent ongoing pelvic sepsis in a defunctioned patient from the column of faeces that will be left from an unprepared colon. Care needs to be taken to ensure that this is carefully documented and communicated to the ward and nursing staff, as bowel preparation is not routine, especially for patients in enhanced recovery programmes. All our patients having open or laparoscopic TME are counselled and sited preoperatively for a defunctioning right-sided ileostomy and left iliac fossa colostomy.

Operation Table Positioning The patient is placed supine on an electrically operated table, with the legs in a modified Lloyd Davies

08/02/13 5:50 PM

146  Laparoscopic surgery

Figure 9.1 Modified Lloyd Davies position of legs.

position. The legs are placed in leg supports, often requiring additional foam padding behind the calf. This may be placed laterally to reduce the risk of common perineal nerve compression. The legs are abducted with the knees at shoulder width apart and flexed to approximately 708. To prevent the right thigh getting in the way of the right hand of the operating surgeon, the hips are placed in neutral or slight extension (Figure 9.1). All patients have bladder catheterization and a nasogastric tube.

Port Insertion and Positioning The surgeon stands to the patient’s right, with the camera holder on the same side to the left of the operating surgeon. The assistant stands to the patient’s left. For mobilization of the splenic flexure, it is often useful for the operating surgeon to stand between the patient’s legs and the assistant to stand to the patient’s right. We commonly use four ports for an anterior resection. The first port is normally placed just below the umbilicus. Occasionally, if the distance from umbilicus to symphysis pubis is short, this port may be placed above the umbilicus. This port is placed using an open or Hasson technique to minimize the risk of injuring intra-abdominal viscera upon insertion. A 10-mm balloon port is our preference at

HEBK001-C09_p140-153.indd 146

the umbilicus. This minimizes both gas leak and the distance of extension of the port into the abdomen. A second 10–12-mm port is placed just medial and superior to the anterior superior iliac spine, with a 5-mm port at least a hand’s breadth cephalad. A further 5-mm port is placed laterally to the umbilicus on the left side of the patient. This is used as an assisting port to provide retraction to the sigmoid colon, and as an operating port (Figure 9.2). Extra ports are introduced as necessary. A 10-mm suprapubic port at the specimen extraction site can be useful if the small bowel is difficult to position, as the camera can be used through this port to give a good view of the omentum and small bowel in the upper abdomen and right upper quadrant. In addition, this port can be used to provide additional retraction in the pelvis. A further right upper quadrant port may be necessary to retract the small bowel or help with splenic flexure mobilization in large patients (see Figure 9.2).

Laparoscopy The first step is to perform a diagnostic laparoscopy and assess the abdomen. This involves evaluation of: l l

the tumour itself; the surface of the liver for liver metastasis;

08/02/13 5:50 PM

OUR APPROACH TO LAPAROSCOPIC ANTERIOR RESECTION  147

Transverse Colon

IMV

5 5

12

DJ flexure

5

12 Figure 9.3 Medial approach to the splenic flexure.

12 Figure 9.2 Standard ports are shown in blue and optional extra ports in green.

the peritoneal surfaces and abdominal cavity for small volume peritoneal disease; l the splenic flexure to assess the need for mobilization. Generally we have a low threshold for fully mobilizing the splenic flexure and would do this initially as in open surgery. We do not feel that mobilization is obligatory for all patients, e.g. in the case of a long sigmoid loop, it may be feasible to omit splenic flexure mobilization. l

Splenic Flexure Mobilization Positioning

(Figure 9.5). This allows the transverse mesocolon to be mobilized off the pancreas. The vein is then lifted and the two areas of dissection can be connected to allow a medial to lateral dissection out over Gerota’s fascia and the pancreas up to the splenic flexure. The omentum is then detached from the transverse colon, which can usually be achieved by sharp dissection with scissors and diathermy. In patients who are obese, it may be easier to divide the gastrocolic ligament with LigaSure™, starting medially and entering the lesser sac from above and moving out laterally. Finally, the lateral attachments of the descending colon and splenic flexure are divided to complete the mobilization. Splenic flexure mobilization can be challenging, and sometimes moving the point of dissection

The patient is positioned flat or in the reverse Trendelenburg position to allow the small bowel to fall into the pelvis. The omentum is retracted cephalad, above the liver and spleen. Some right tilt can be useful to encourage the small bowel mesentery to fall to the right to allow identification of the duodenal-jejunal flexure and inferior mesenteric vein (Figure 9.3). Approach

The inferior mesenteric vein is mobilized and divided above the ascending left colic vein as it dives below the pancreas. We use either clips or ­LigaSure™ bipolar diathermy to divide the vein (Figure 9.4). The lesser sac is opened above the pancreas and lateral to the middle colic vessels.

HEBK001-C09_p140-153.indd 147

Pancreas

Ascending left colic vein

IMV

Figure 9.4 High division of the IMV.

08/02/13 5:50 PM

148  Laparoscopic surgery

transverse mesocolon

lesser sac pancreas

Figure 9.5. Medial approach to the lesser sac.

­ etween a combination of medial, lateral and superior b approaches is required to continue to make progress. An alternative approach, when splenic flexure mobilization is considered imperative, is to start the operation with the patient in the right lateral position. If this is the preferred approach, several pointers that may be helpful are as follows: Mark the first port position in the left upper quadrant midclavicular line just below the costal margin while the patient is prone. Particularly in a patient with a pendulous abdomen, the port placement site can be difficult to assess when the patient is in the right lateral position. l Have the patient positioned at about 70–808 (not at 908). l Use two back supports at the level of the scapula and the anterior superior iliac spine. l Position the monitor above and lateral to the patient’s left shoulder. l The right knee and hip should be in slight flexion. l Cushion between the patient’s legs. l

Approach to the Inferior Mesenteric Artery Pedicle Positioning

The patient is placed in a steep Trendelenburg position with the right lateral tilt left in place. The small bowel needs to be moved into the right upper quadrant. Several tricks and tips can facilitate this: l

The omentum needs to be as cephalad as possible; the maximum degree of Trendelenburg is

HEBK001-C09_p140-153.indd 148

then required to allow gravity to assist the small bowel in falling out of the pelvis. l The small bowel mesentery needs to be ‘layered’, right and cephalad. We achieve this with the blunt graspers (Johan’s type) closed, using blunt manipulation of the leaves of the small bowel mesentery. l The terminal ileum and appendix may need to be mobilized if they are holding the small bowel in the pelvis. l If the small bowel persists in dropping into the field of view, a tonsil swab can be helpful in providing additional retraction. If this fails, a right upper quadrant port can allow a grasper to act as a retractor across the right iliac fossa. Finally, if small bowel positioning is still elusive, placing the camera through a 10- or 12-mm suprapubic port allows a better view of the right upper quadrant and more accurate positioning of the small bowel and omentum. Approach

The inferior mesenteric artery (IMA) is approached from a medial direction. To facilitate visualization of the IMA pedicle as it arises from the aorta, the sigmoid colon is retracted superiorly and cephalad. Avoid handling the bowel directly by grasping an appendix epiploicae for retraction. Dissection of the pedicle

The pelvic brim is identified and the dissection begun at the sacral promontory. The peritoneum is incised below and parallel to the ileocolic pedicle. Our preference is to mark the peritoneal dissection plane with closed diathermy scissors. The vascular space is then opened using small downward movements with the scissors. The peritoneal incision is extended in a cephalad direction over the origin of the IMA. The aim is to separate the mesenteric fat from the retroperitoneum. Continuing in this direction, Toldt’s plane will be reached (the plane between the posterior mesocolon and the anterior retroperitoneal fascia). Once in this plane, it is our preference to use a blunt dissector (e.g. with 5-mm LigaSure™) to extend the dissection laterally towards the abdominal wall. The correct plane should be avascular and allows the hypogastric nerves, gonadal vessels and ureter to be dropped back posteriorly (Figure 9.6). If the bare belly of the psoas muscle is visible, then the dissection plane is deep to the ureter. It is imperative that the ureter is positively

08/02/13 5:50 PM

OUR APPROACH TO LAPAROSCOPIC ANTERIOR RESECTION  149

IMA

left ureter

branches of hypogastic plexus

Figure 9.6 Mobilizing the inferior mesenteric artery pedicle.

identified before division of the pedicle. If this has not been possible, then a lateral approach may be helpful. It is our preference to carefully dissect out the IMA at its origin from the aorta and then apply two large Haemolock™ clips (gold). Dividing between the clips can be done with or without an energy device. If the pedicle is particularly bulky, it may be necessary to divide it with a laparoscopic stapler. A tonsil swab can be left at the most lateral point of dissection, above the ureter, to identify the plane when attention is turned to the lateral attachments. Following division of the IMA, the dissection can be continued in a cephalad direction (in Toldt’s plane) towards the splenic flexure. A second low division of the inferior mesenteric vein is then performed for complete mobilization. Again, if in the correct plane, Gerota’s fascia will be left intact and swept posteriorly. Dissection is continued in this direction until the physical extent of dissection is limited. Before taking down the lateral attachments, ensure that the plane has also been developed as far as is possible in an inferior direction. To release the lateral attachments, the sigmoid colon will need to be retracted in a medial direction. If a tonsil swab was left in situ, it will be visible as the attachments are taken down and the plane identified. If the medial to lateral dissection was adequate, this will be a thin layer.

Mobilization of the Rectum

mesorectal fascial plane. In laparoscopic surgery, traction and counter-traction on the mesorectum, or near the tumour, have to be performed with instruments and may be more traumatic (less soft and sensitive) than an expert surgeon’s hand. This may result in transgression of the ‘holy plane’, which has been shown to be associated with poorer outcomes.33,35 Using a tonsil swab underneath a Johan’s retractor can facilitate atraumatic retraction. Once the plane is opened up, air will enter the areolar plane and allow sharp dissection with either scissors or hook diathermy. At all time, retraction needs to be gentle and the peritoneal edges should be used for retraction. The uterus, or the peritoneum above the seminal vesicles, can be stitched up to the anterior abdominal wall to facilitate retraction if necessary. In addition, in female patients a ‘swab-on-a-stick’ in the vagina retracting it anteriorly may aid visualization of the planes. It is important at all times during mobilization to be aware of the position of the tumour. A combination of digital examination before and during the procedure, and a review of the preoperative MRI, are essential to compensate for the loss of tactile sensation at laparoscopic surgery compared with open surgery. Technique

The peritoneum can then be incised towards the pelvis, over the pelvic brim. Here it is important to identify the ‘holy plane’ or mesorectal fascial plane (Figure 9.7). The rectum is initially retracted anteriorly. This should be achieved with minimal trauma, particularly at the level of the tumour.

rectum

TME plane

hypogastric plexus

Principles

The concept of TME revolves around retraction and counter-traction26 to facilitate dissection in the

HEBK001-C09_p140-153.indd 149

Figure 9.7 Entering the total mesorectal excision plane at the top of the pelvis.

08/02/13 5:50 PM

150  Laparoscopic surgery

uterus rectum retracted to left

rectum swab retractor

holy plane right hypogastric

TME plane at back

nerve

Figure 9.8 Dropping back the right hypogastric nerve.

Figure 9.10 Use of a swab as retractor.

Care must be taken not to damage the hypogastric nerves as they descend over the sacral promontory into the pelvis. Dissection is continued in the TME plane posteriorly and laterally, focusing initially on the right side (Figure 9.8). When adequate mobilization on the right and posteriorly has been achieved, the peritoneum on the left can be divided carefully, dropping back the left hypogastric nerve (Figure 9.9). The rectum needs to be fully mobilized posteriorly with division of Waldeyer’s fascia and mobilized off the pelvic floor. The same principles apply as in open surgery, namely working down the back of the rectum and coming round to the sides with traction and counter-traction, often with a tonsil swab demonstrating the planes (Figure 9.10). As in open surgery, it is necessary to move the focus of the dissection from right

lateral to left lateral or anteriorly to facilitate distal dissection. Anteriorly, in both male and female patients, a cuff of anterior peritoneum should be preserved by continuing the dissection from the left and right sides and meeting in the middle. This acts as a safe area to grasp with a retractor and allows the surgeon to find the correct plane behind the seminal vesicles in male patients and the vagina in female patients (Figure 9.11). On the lateral pelvic side-wall the nervi erigentes and inferior hypogastric plexus are dropped off the rectum. The dissection should continue until 1–2 cm clear of the distal extent of the tumour, with at least 5 cm of mesorectum below the tumour or a total mesorectal excision. For TME, the rectum needs to be completely mobilized off the pelvic floor and on to the muscle tube as it enters the intersphincteric groove (Figure 9.12).

uterus

left hypogastric nerve

vagina

rectum cul de sac

Figure 9.9 Dividing peritoneum on the left.

HEBK001-C09_p140-153.indd 150

Figure 9.11 Dissecting the rectovaginal plane.

08/02/13 5:50 PM

CONCLUSION  151

inferior hypogastric plexus

vagina

pelvic floor rectum

Figure 9.12 Approaching the intersphincteric plane on the left.

Division of the Rectum and Anastomosis It is our practice to strive to perform the same operation laparoscopically as at open surgery, and therefore a rectal washout after cross-clamping and before rectal transection is imperative. Although washout is a simple procedure at open surgery, it is more complex laparoscopically due to difficulty in occluding the lumen distal to the tumour. A number of methods have been reported, including the use of a bulldog clip or Roeder knot or passing a piece of intravenous line tubing (or similar) via the left iliac fossa port, around the rectum and using this as an occlusive tourniquet.41 These techniques are technically challenging in many cases and potentially result in the procedure being abandoned and washout not performed. Following washout, the rectal transection can be performed. Rectal transection proximal to the pelvic floor can be challenging, as current laparoscopic staplers can angulate to a maximum of only 658. As in open surgery, the transection should be perpendicular to the rectal muscle tube. Ideally, this should be achieved with a single firing of the stapler, but practically it often requires two or three firings. Not only is this expensive and time-consuming, but it can also result in ‘dog ears’ or a zigzag transection staple line, raising the possibility of ischaemic areas that may subsequently increase the risk of leak. Furthermore, within the confines of the narrow male pelvis, especially for a mid- or low rectal cancer, it

HEBK001-C09_p140-153.indd 151

may be impossible to angulate the stapler across the rectum. The inability to accurately palpate and correctly identify the lower limit of the tumour makes mesorectal transection for lower tumours technically difficult. If there is still uncertainty, a rigid sigmoidoscope, or digital examination in low tumours, can be used to accurately determine the point of transection. If there is still uncertainty despite these manoeuvres, then it is our preference to insert the linear stapler via a Pfannenstiel incision; this incision is also used as the specimen extraction site. If the rectum has been fully mobilized, the stapler can be applied easily through an 8-cm incision and also facilitates formation of a defunctioning ileostomy.

Specimen Extraction To facilitate specimen location and retrieval, we leave a Johan’s retractor attached to the divided bowel end. If not performed already for cross-stapling, our preferred extraction site is via a Pfannenstiel incision. The patient should be left in the head-down position while the abdominal incision is made. A wound retractor is inserted; we currently favour the Alexis (Applied Medical)™ wound system. The bowel is removed and the proximal resection point identified. The anvil of the circular stapler can be inserted in the normal fashion, before dropping back into the abdomen, restoring the pneumoperitoneum and performing the anastomosis.

CONCLUSION In summary, with regard to the ‘abdominal’ part of a rectal cancer operation, it is likely that a laparoscopic minimal-access technique is feasible and beneficial, in the short term, although operating time may be prolonged. Equipment costs may be greater, but the shorter hospital stay and faster recovery may more than compensate for these additional initial expenses. Initial randomized controlled trials report equivalence in longer-term oncological outcomes, with some groups reporting comparable local recurrence rates. The advantage of magnification of the TME plane may give better visualization and

08/02/13 5:50 PM

152  Laparoscopic surgery

s­ ubsequently better macroscopic quality of the specimen, particularly when surgeons and operative teams have overcome the learning curve and gained experience. Ultimately, the individual surgeon’s approach to rectal cancer surgery will be determined by many factors. Each case has to be assessed on an individual basis, taking into account patient factors, anatomical considerations and the experience of the individual surgeon. The surgeon should also be willing to change the approach intraoperatively if there is any concern about the oncological outcome being compromised. Each surgeon should audit their laparoscopic rectal cancer outcomes and compare these with their outcomes from open surgery. Longer-term outcome data from clinical trials are awaited to answer some ongoing issues, such as long-term cure rates, incisional hernia rates and postoperative adhesions. The data suggest that optimal oncological colorectal cancer resection can be performed laparoscopically in selected patients. Nevertheless, there are still some ongoing concerns regarding laparoscopic low anterior resection. The group that described the first rectal cancer resection have reported higher than expected local recurrence rates in Union for International Cancer Control (UICC) stage III disease.42 Achieving good oncological outcomes for mid- and low rectal cancer is a continuing challenge, regardless of whether the approach is open or laparoscopic. Good-quality surgery, according to the experience and technical ability of the surgeon, should be the gold standard. Whether the marginal short-term benefits of the laparoscopic approach justify the potential oncological downside remains controversial.5 In rectal cancer surgery, safe retraction and countertraction, an optimal view, good-quality adherence to the TME ‘holy plane’ principles, optimal mesorectal excision or transection, and rectal division and anastomosis remain the aim in this common, often complex cancer.

References   1. Berends FJ, Kazemier G, Bonjer HJ, Lange JF. Subcutaneous metastases after laparoscopic colectomy. Lancet 1994; 344: 58.   2. Schwenk W, Haase O, Neudecker JJ, Müller JM. Short term benefits for laparoscopic colorectal resection. Cochrane Database Syst Rev 2005; (2): CD003145.

HEBK001-C09_p140-153.indd 152

  3. Kuhry E, Schwenk W, Gaupset R, Romild U, Bonjer HJ. Long-term results of laparoscopic colorectal cancer resection. Cochrane Database Syst Rev 2008; (2): CD003432.   4. Breukink S, Pierie J-P, Wiggers T. Laparoscopic versus open total mesorectal excision for rectal cancer. ­Cochrane Database Syst Rev 2006; (4): CD005200.   5. Marsden MR, Parvaiz A, Moran B. Resection of rectal cancer: laparoscopy or open surgery? Ann R Coll Surg Engl 2010; 92: 106–12.   6. Guilloi PJ, Quirke P, Thorpe H, et al. Short-term endpoints of conventional versus laparoscopically-assisted surgery in patients with colorectal cancer (MRC CLASICC trial): multicentre, randomised controlled trial. Lancet 2005; 365: 1718–26.   7. National Institute for Health and Clinical Excellence. Colorectal Cancer: Laparoscopic Surgery (Review). Technology appraisal TA105. London, National ­Institute for Health and Clinical Excellence, 2006 (http://www.nice.org.uk/Guidance/TA105).   8. Quirke P, Steele R, Monson R, et al. Effect of the plane of surgery achieved on local recurrence in patients with operable rectal cancer: a prospective study using data from the MRC CR07 and NCIC0CTG CO16 randomised clinical trial. Lancet 2009; 373: 821–8.   9. Robinson PJ. Imaging liver metastases: current limitations and future prospects. Br J Radiol 2000; 73: 234–41. 10. Klerkx WM, Bax L, Veldhuis WB, et al. Detection of lymph node metastases by gadolinium–enhanced magnetic resonance imaging: systematic review and meta-analysis. J Natl Cancer Inst 2010; 102: 244–53. 11. Dromain C, Leboulleux S, Auperin A, et al. Staging of peritoneal carcinomatosis: enhanced CT vs. PET/CT. Abdom Imag 2008; 33: 87–93. 12. Daniels IR, Fisher SE, Heald RJ, Moran BJ. Accurate staging, selective preoperative therapy and optimal surgery improves outcome in rectal cancer: a review of the recent evidence. Colorectal Dis 2007; 9: 290–301. 13. Neudecker J, Klein F, Bittner R, et al. Short-term outcomes from a prospective randomized trial comparing laparoscopic and open surgery for colorectal cancer. Br J Surg 2009; 96: 1458–67. 14. Morton DG, Sebag-Montefiore D. Defunctioning stomas in the treatment of rectal cancer. Br J Surg 2006; 93: 650–1. 15. Koea JB, Guillem JG, Conlon KC, Minsky B, Saltz L, Cohen A. Role of laparoscopy in the initial multimodality management of patients with near-obstructing rectal cancer. J Gastrointest Surg 2000; 4: 105–8. 16. Young CJ, Eyers AA, Solomon MJ. Defunctioning of the anorectum: historical controlled study of laparoscopic vs. open procedures. Dis Colon Rectum 1998; 41: 190–4. 17. West NP, Finan PJ, Anderin C, Lindholm J, Holm T, Quirke P. Evidence of the oncological superiority of cylindrical abdominoperineal excision for low rectal cancer. J Clin Oncol 2008; 26: 3517–22.

08/02/13 5:50 PM

References  153 18. Morino M, Parini U, Giraudo G, Salval M, Brachet Contul R, Garrone C. Laparoscopic total mesorectal excision: a consecutive series of 100 patients. Ann Surg 2003; 237: 335–42. 19. Leroy J, Jamali F, Forbes L, et al. Laparoscopic total mesorectal excision (TME) for rectal cancer surgery: long-term outcomes. Surg Endosc 2004; 18: 281–9. 20. Leung KL, Kwok SP, Lam SC, et al. Laparoscopic resection of rectosigmoid carcinoma: prospective randomised trial. Lancet 2004; 363: 1187–92. 21. Kang SB, Park JW, Jeong SY, et al. Open versus laparoscopic surgery for mid or low rectal cancer after neoadjuvant chemoradiotherapy (COREAN trial): short-term outcomes of an open-label randomised controlled trial. Lancet Oncol 2010; 11: 637–45. 22. Lujan J, Valero G, Hernandez Q, Sanchez A, Frutos MD, Parrilla P. Randomized clinical trial comparing laparoscopic and open surgery in patients with rectal cancer. Br J Surg 2009; 96: 982–9. 23. Braga M, Frasson M, Vignali A, Zuliani W, Capretti G, Di Carlo V. Laparoscopic resection in rectal cancer patients: outcome and cost–benefit analysis. Dis Colon Rectum 2007; 50: 464–71. 24. Hiki N, Shimizu N, Yamaguchi H, et al. Manipulation of the small intestine as a cause of the increased inflammatory response after open compared with laparoscopic surgery. Br J Surg 2006; 93: 195–204. 25. Pechlivanides G, Gouvas N, Tzortzinis A, et al. Lymph node clearance after total mesorectal clearance for rectal cancer: laparoscopic versus open approach. Dig Dis 2007; 25: 94–9. 26. Heald RJ. The ‘holy plane’ of rectal surgery. J R Soc Med 1988; 81: 503–8. 27. Birbeck KF, Macklin CP, Tiffin NJ, et al. Rates of circumferential resection margin involvement vary between surgeons and predict outcomes in rectal cancer surgery. Ann Surg 2002; 235: 449–57. 28. Jayne DG, Thorpe HC, Copeland J, Quirke P, Brown JM, Guillou PJ. Five-year follow-up of the Medical Research Council CLASICC trial of laparoscopically assisted versus open surgery for colorectal cancer. Br J Surg 2010; 97: 1638–45. 29. Jayne DG, Guillou PJ, Thorpe H, et al. Randomized trial of laparoscopic–assisted resection of colorectal carcinoma: 3-year results of the UK MRC CLASICC Trial Group. J Clin Oncol 2007; 25: 3061–8. 30. Biondo S, Ortiz H, Lujan J, et al. Quality of mesorectum after laparoscopic resection for rectal cancer: results of an audited teaching programme in Spain. Colorectal Dis 2010; 12: 24–31.

HEBK001-C09_p140-153.indd 153

31. Breukink SO, Grond AJ, Pierie JP, Hoff C, Wiggers T, Meijerink WJ Laparoscopic vs open total mesorectal excision for rectal cancer: an evaluation of the mesorectum’s macroscopic quality. Surg Endosc 2005; 19: 307–10. 32. Law WL, Lee YM, Choi HK, Seto CL, Ho JW. Impact of laparoscopic resection for colorectal cancer on operative outcomes and survival. Ann Surg 2007; 245: 1–7. 33. Quirke P. Training and quality assurance for rectal cancer: 20 years of data is enough. Lancet Oncol 2003; 4: 695–702. 34. Gouvas N, Tsiaoussis J, Pechlivanides G, et al. Quality of surgery for rectal carcinoma: comparison between open and laparoscopic approaches. Am J Surg 2009; 198: 702–8. 35. Heald RJ, Moran BJ, Ryall RD, Sexton R, MacFarlane JK. Rectal cancer: the Basingstoke experience of total mesorectal excision, 1978–1997. Arch Surg 1998; 133: 894–9. 36. Rahbari NN, Weitz J, Hohenberger W, et al. Definition and grading of anastomotic leakage following anterior resection of the rectum: a proposal by the International Study Group of Rectal Cancer. Surgery 2010; 147: 339–51. 37. Cartmell MT, Jones OM, Moran BJ, Cecil TD. A defunctioning stoma significantly prolongs the length of stay in laparoscopic colorectal resection. Surg Endosc 2008; 22: 2643–7. 38. Zhou ZG, Hu M, Li Y, et al. Laparoscopic versus open total mesorectal excision with anal sphincter preservation for low rectal cancer. Surg Endosc 2004; 18: 1211–5. 39. Vignali A, Braga M, Zuliani W, Frasson M, Radaelli G, Di Carlo V. Laparoscopic colorectal surgery modifies risk factors for postoperative morbidity. Dis Colon Rectum 2004; 47: 1686–93. 40. Allardyce RA, Bagshaw PF, Frampton CM, et al. Australasian Laparoscopic Colon Cancer Study shows that elderly patients may benefit from lower postoperative complication rates following laparoscopic versus open resection. Br J Surg 2010; 97: 86–91. 41. García-Granero E, Flor-Lorente B, García-Botello S, Muñoz E, Blanco F, Lledó S. The occlusive tourniquet: a simple method for rectal stump washout during open and laparoscopic surgery. Dis Colon Rectum 2008; 51: 1580–2. 42. Scheidbach H, Schneider C, Hügel O, et al. Oncological quality and preliminary long-term results in laparoscopic colorectal surgery. Surg Endosc 2003; 17: 903–10.

08/02/13 5:50 PM

10 Robotic total mesorectal excision M. Chadwick, H.S. Tilney and A.M. Gudgeon

Introduction The da Vinci robot is a sophisticated tool for performing minimal access surgery. The actions of the instruments are entirely under the control of a surgeon seated separately from the patient. The camera and instruments are controlled by hand and foot movements of the surgeon. The surgeon can operate the instruments from a comfortable position, with forearms supported and completely intuitive reproduction of hand movements to the operating instruments. The surgery is aided by high-definition three-dimensional optics. The surgery may be performed entirely using the robotic instruments or as a hybrid procedure combining standard laparoscopy with robotics.

History and development of robotic colorectal surgery Robots have been used in surgery since 1985, with the PUMA 560 first assisting a brain biopsy. In 1988 the first prostatic surgery was performed by the PROBOT at Imperial College, London. The first robot for clinical use in general surgery was the automated endoscopic system for optimal positioning (AESOP) (Computer Motion, Santa Barbara, CA, USA). In 1994 the United States Food and Drug Administration (FDA) approved AESOP for clinical use as a robotic camera-holder. In 1990 the Defence Advanced Research Projects Administration

HEBK001-C10_p154-170.indd 154

supported the development of a prototype robot by SRI International because of its potential to allow surgeons to operate remotely on soldiers wounded on the battlefield. As a result of this, in 1995 Intuitive Surgical Devices Inc. came into being; taking this prototype forward, the da Vinci system evolved. It was first tested in 1997 and then marketed in ­Europe in 1999. Sales in the USA were soon to follow in 2000, with FDA approval for general laparoscopic surgery. The ZEUS robotic surgical system developed by the rival company Computer Motion allowed the first telerobotic surgery in 2001, with the surgeon in New York and the patient in France. Fierce competition stalled further developments until, in 2003, Intuitive Surgical and Computer Motion merged and the ZEUS system was phased out in favour of the da Vinci system, which is now used across the world, particularly in South Korea, the USA and Europe. The latest refinement is the da Vinci Si, released in 2009. The most common procedure performed with the da Vinci system is radical prostatectomy, but many other urological, gynaecological, cardiothoracic, paediatric and general surgical procedures have been performed, including pyeloplasty, cystectomy, nephrectomy, ureteral reimplantation, hysterectomy, myomectomy, sacrocolpopexy, mitral valve repair, atrial septal defect closure, cardiac bypass, ablation, mediastinal tumour excision, cholecystectomy,1 Nissen fundoplication, Heller myotomy, gastric bypass, donor nephrectomy, adrenalectomy, splenectomy and hepatic resection.2

11/02/13 3:58 PM

ROBOTIC RECTAL SURGERY  155

The uptake of robotic colorectal surgery has been relatively recent. Most work has been done in South Korea, but increasing numbers of procedures are being performed in Italy and the USA, and its use is gradually expanding in a few centres in the UK. Procedures include right and left hemicolectomy, total mesorectal excision (TME), proctectomy and ileal pouch–anal anastomoses, and ventral mesh rectopexy. With increasing experience, confidence is building and increasingly advanced procedures are being performed. In 2007 Baik and colleagues reported simultaneous robotic TME, total abdominal hysterectomy for rectal cancer and uterine myoma.3 In 2009 Patriti and colleagues reported a series of robotic colectomies, including one rectal resection with synchronous hepatic metastatectomies.4 In 2002 Weber and colleagues reported three robotic right and sigmoid colectomies for benign disease using the da Vinci robotic system.5 In the following year, Delaney and colleagues compared robot-assisted laparoscopic colectomy with casematched results using standard laparoscopic approaches, focusing on clinical outcomes.6 Despite only six robotic colectomies being performed between December 2001 and April 2002, there were conclusions of feasibility and safety of the da Vinci system. D’Annibale and colleagues reported 53 robotic colorectal cases from May 2001 to May 2003, including 22 patients with malignant disease.7 They concluded that robotic techniques could achieve the same operative and postoperative results as conventional laparoscopic surgery. Rawlings and colleagues reported 30 consecutive robotic cases;8 the series included 17 right hemicolectomies and 13 anterior resections, with the conclusion that robotic surgery is technically feasible using the da Vinci system. Two years later the same authors reported the results comparing robotic with laparoscopic colectomy.9 The outcome in each group was similar.

ROBOTIC RECTAL SURGERY Robotic low anterior resection is performed in only a few centres worldwide. Surgery for rectal cancer is more difficult than colonic surgery because of the anatomical characteristics of the rectum and the pelvis. The principle underlying TME is precise dissection of an avascular plane between the presacral fascia and the fascia propria of the mesorectum.

HEBK001-C10_p154-170.indd 155

The view of the surgical field is limited within the pelvis during TME, particularly when operating in the narrow male pelvis, and this may have an impact on the quality of the resected mesorectum.10 The most important factor relating to rectal dissection is the grade of the mesorectal specimen, which in turn is closely related to oncological outcome.11,12 A good view is mandatory if precise mesorectal dissection is to be successful. The ability of the da Vinci surgical system to offer a three-dimensional view with magnification, filtering of hand tremor, fine dexterity and motion scaling suggests the potential for a technical surgical advantage over open or laparoscopic TME. A four-arm technique for robotic TME has been summarized by Baik and colleagues.13 The instruments used for dissection were a Cadiere grasper, a PreCise™ Bipolar grasper and a permanent cautery spatula. The Cadiere grasper provides the first traction and the PreCise Bipolar grasper provides the second proper traction. Moreover, these graspers can change to an L-shaped small retractor using the EndoWrist™ function that is the core technology of the da Vinci system. The robotic instrument can be used not only to create traction by grasping tissue but also to push the tissue in the narrow pelvic space, thereby achieving traction and countertraction to expose the ideal tissue planes. The combination of traction and counter-traction using the robotic instruments can provide an excellent surgical view during the rectal dissection in a confined space.

Evolution of Robotic Total Mesorectal Excision Until now, most studies relating to robotic low anterior resection have reported only initial experiences. The sample sizes have been small and they are generally non-randomized. Larger, randomized studies are necessary to assess the feasibility of the robotic system and whether there are real benefits of the robotic system compared with conventional laparoscopic surgery. Tables 10.1–10.3 summarize the results of existing studies comparing outcomes between robotic and laparoscopic rectal resections. In 2004 D’Annibale and colleagues compared 12 robotic rectal resections with laparoscopic rectal resection

11/02/13 3:58 PM

Period Study

Institution

Delaney et al.

6

Cleveland Clinic,

of study

2001–2002

Patients (n)

Type of study

Case-matched

Laparoscopic

surgery

surgery

Robotic surgery

Laparoscopic surgery

6

6

Right hemicolectomy (2), sigmoid

Right hemicolectomy (2), sigmoid

OH, USA D’Annibale et al.7 Padova, Italy

Pigazzi et al.

14†

City of Hope National

Procedures (n)

Robotic

resection (3), rectopexy (1) 2001–2003

2004–2005

Case-matched

Non-randomized,

53

53

Right hemicolectomy (10), ileocaecal

resection (3), rectopexy (1) Right hemicolectomy (13), ileocaecal

resection (0), transverse colectomy (0),

resection (1), transverse colectomy (1),

left hemicolectomy (17), sigmoid

left hemicolectomy (17),

resection (11),

sigmoid resection (4), anterior

anterior resection (10), abdominoperineal

resection (15), abdominoperineal

excision of rectum (1),

excision of rectum (0),

total colectomy (2), Hartmann’s (1),

total colectomy (1), Hartmann’s (1),

Hartmann reversal (0), rectopexy (1)

Hartmann reversal (2), rectopexy (0)

6

6

Low anterior resection

Low anterior resection

39

–

Rectal cancers: low anterior resection (22),

–

comparative

Medical Center, CA, USA City of Hope National

Hellan et al.15†

2004–2007

Case series

coloanal (11), abdominoperineal excision

Medical Center,

of rectum (6)

CA, USA City of Hope National

Baek et al.16†

Spinoglio et al.

19

2004–2008

Case series

64

–

Rectal cancers: colorectal anastomosis (34),

Medical Center,

coloanal anastomosis (18), abdominoperineal

CA, USA

excision of rectum (12)

Alessandria, Italy

2005–2007

Non-randomized, comparative

50

161

Right hemicolectomy (18), left hemicolectomy

–

Right hemicolectomy (50), left

(10), anterior resection (19), abdominoperineal

hemicolectomy (73), anterior resection

excision of rectum (1), transverse colectomy (1),

(26), abdominoperineal excision of

total colectomy (1)

rectum (7), transverse colectomy (2), total colectomy (3) (Continues)

156  Robotic total mesorectal excision

HEBK001-C10_p154-170.indd 156

Table 10.1. Overall characteristics of studies reporting outcomes of robotic rectal cancer surgery.

11/02/13 3:58 PM

HEBK001-C10_p154-170.indd 157

Table 10.1. Overall characteristics of studies reporting outcomes of robotic rectal cancer surgery. (Continued) Period Study

Institution

of study

Patients (n)

Type of study Robotic

Laparoscopic

Procedures (n)

surgery

surgery

Robotic surgery

Laparoscopic surgery

Baik et al.13*

Seoul, South Korea

2006–2007

Case series

9

–

Total mesorectal excision

–

Baik et al.20*

Seoul, South Korea

2006–2007

Randomized

18

18

‘Tumour-specific total mesorectal

‘Tumour-specific total mesorectal

controlled trial Baik et al. * 21

Seoul, South Korea

2006–2007

excision’

excision’

56

57

Low anterior resection

Low anterior resection

Case series

50

–

Rectal cancers

–

Case series

55

–

Anterior resection (17), abdominoperineal

–

Non-randomized, comparative

Choi et al.22

Korea University Anam 2007–2008 Hospital, Seoul, South Korea Milan, Italy

2007–2008

excision of rectum (7), coloanal (4), left hemicolectomy (27)

*Multiple publications by the same departments with likely duplication of reported outcomes.

†,

ROBOTIC RECTAL SURGERY  157

Luca et al.23

11/02/13 3:58 PM

Duration of surgery (min) Study

Delaney

Conversion

Length of stay (days)

Anastomotic

(range)

leak

Blood loss (mL) (range)a

Complications

Robotic

Laparoscopic

Robotic

Laparoscopic

Robotic

Laparoscopic

Robotic

Laparoscopic

Robotic

Laparoscopic

Robotic

surgery

surgery

surgery

surgery

surgery

surgery

surgery

surgery

surgery

surgery

surgery

surgery

–

–

216.5

150 (116–165)

100

87.5 (50–200)

3 (2–5)

2.5 (2–7)

–

–

1 atelectasis

1 incisional

(170–274)

et al.6 D’Annibale

(range)a

6 (2 to

et al.7

3

240 (661)

Laparoscopic

(50–350) 222 (677)

21 (680)

hernia 37 (6102)

–

–

0

1

2 bowel injuries, 1

laparoscopy,

CVA, 1 wound

4 to hand-

infection

assisted laparoscopy) Pigazzi

–

–

264 (192–318)

258 (198–312)

104

Hellan

150 (50–300)

4.5 (3–11)

3.6 (3–6)

–

–

1 prolonged ileus

1 pelvic abscess

–

4 (2–22)

–

4 (12.1%)

–

1 intraoperative

–

(50–318)

et al.14† –

–

285

–

200

(180–540)

et al.15†

(25–6000)

bleed, 2 wound infections

Baek

6 (9.4%)

–

270 (150–540)

–

200

2 (1 to

4

383.8b

266.3b

–

5 (2–33)

–

4 (7.7%)

–

4 pelvic

–

7.74

8.31

2

–

1 incisional hernia,

(20–6000)

et al.16† Spinoglio

–

et al.18

abscesses

laparoscopy,

1 atelectasis,

1 to

1 wound

laparotomy)

infection,

–

1 phlebitis, 1 CVA Baik

0

–

213 (153–315)

–

–

–

7 (5–10)

–

–

–

0

2

202.5

196.0 (114–297)

–

–

7 (5–10)b

9 (6–12)b

–

–

–

179 (100–360)

–

–

5 (5–10)b

6 (4–16)b

1

4

‘Serious

et al.13* Baik

Baik et al.21*

–

(149–315)

et al.20* 0b

6b

178 (120–315)

‘Serious

complication’

complication’

5.4%b

19.3%b

(Continues)

158  Robotic total mesorectal excision

HEBK001-C10_p154-170.indd 158

Table 10.2. Short-term outcomes from studies reporting outcomes of robotic rectal cancer surgery.

11/02/13 3:58 PM

–

a

HEBK001-C10_p154-170.indd 159

and demonstrated that robotic and laparoscopic techniques could achieve the same operative and postoperative results.7 Pigazzi and colleagues compared short-term outcomes between robotic TME and laparoscopic TME and concluded that robotic low anterior resection with TME and autonomic nerve preservation was possible.14 The same group subsequently reported a series of 39 consecutive unselected patients with primary rectal cancer and concluded that robotic-assisted surgery for rectal cancer could be carried out safely.15 Baik and colleagues reported the first Asian experience of robotic total mesorectal excision for rectal cancer patients in June 2006.17 The first published case of robotic abdominoperineal resection in Asia was performed in Hong Kong in August 2006.18 One study including 20 robotic rectal cancer excisions concluded that robotic colonic surgery was feasible and safe. Spinoglio and colleagues reported a significant increase in operating time (see Table 10.2),19 but in the authors’ experience this length reduces with experience as ‘docking time’ reduces. The use of a hybrid technique to perform vascular division and left colon/splenic flexure mobilization laparoscopically can also help to reduce operative times. Baik and colleagues launched the first prospective randomized trial of robotic versus laparoscopic low anterior resection in 2006.20 The short-term outcomes of this pilot study comparing 18 robotic low anterior resections with 18 laparoscopic cases highlighted the feasibility and safety of robotic low anterior resection. A trend towards a superior mesorectal grade in the robotic low anterior resection group did not achieve statistical significance, but macroscopic grading of the robotic group showed the mesorectum to be complete in 17 cases and nearly complete in 1 case. Similar results were shown in their larger, albeit non-randomized, comparative study from 2008.23 The quality of the resected specimen can be considered a surrogate marker for both surgical quality and outcome.24 If this can be shown to achieve significance in ongoing trials, then it would support a potential oncological benefit for robotic rectal cancer surgery, in oncological terms. In addition, the use of sharp dissection and simple diathermy in the avascular TME plane may aid identification and preservation of the pelvic autonomic nerves and therefore represent a secondary benefit in terms of long-term quality-oflife outcomes.

b

(0–600) (164–487) et al.22

Figures for continuous outcomes represent median values with range in parentheses unless 6 stated, which denotes standard deviation. Statistically significant results. †, *Multiple publications by the same departments with likely duplication of reported outcomes. CVA, cerebrovascular accident.

– – – – (4–17)

(5–24)

7.5 – 68 –

(190–485)

290 – 0 Luca

304.8 – – Choi

et al.2

surgery

– – – 4 (8.3%) – 9.2 – –

Laparoscopic

surgery surgery surgery

–

surgery surgery surgery surgery surgery surgery surgery

Laparoscopic Robotic Laparoscopic Robotic Laparoscopic Robotic Laparoscopic Robotic

surgery

Laparoscopic

leak (range)

Robotic

Anastomotic Length of stay (days)

Blood loss (mL) (range)a Duration of surgery (min)

(range)a Conversion Study

Table 10.2. Short-term outcomes from studies reporting outcomes of robotic rectal cancer surgery. (Continued)

Robotic

Complications

ROBOTIC RECTAL SURGERY  159

11/02/13 3:58 PM

Circumferential Lymph node yield (n)

resection margin

(range)a

Study

Distal resection margin (cm)

Robotic

Laparoscopic

surgery

surgery

Delaney et al.

–

–

–

–

–

D’Annibale et al.7

17 (610)

16 (69)

–

–

–

Pigazzi et al.14†

14 (9–28)

17 (9–39)

3.8 (1.8–9)

3.5 (2.2–5)

Hellan et al.15†

13 (7–28)

–

2.65 (0.4–7.5)

Baek et al.16†

14.5 (3–28)

–

3.4 (0.2–10)

6

Robotic surgery

Laparoscopic

involvementb

Total mesorectal excision assessment Robotic surgery

Laparoscopic surgery

Robotic

Laparoscopic

surgery

surgery

–

–

–

–

–

–

–

–

–

–

–

–

–

0

–

–

–

–

0

–

surgery

Spinoglio et al.

22.03

22.85

7.3

7.9

–

–

–

–

Baik et al.13*

22 (2–31)

–

2.5 (1–6.5)

–

Complete (8), nearly complete (1)

–

–

–

Baik et al.20*

18 (6–49)

22 (9–42)

4 (1–5.5)

3.5 (1.5–6.0)

Complete (17), nearly complete (1)

Complete (13), nearly complete (5)

–

–

Baik et al. *

17.5 (4–43)

17 (4–53)

4 (1–7)

3 (1–9)

Complete (52), nearly complete (4),

Complete (43), nearly complete

4

5

18

21

incomplete (0)

(12), incomplete (2)

Choi et al.2

20.6 (4–48)

–

1.9 (0.5–4.5)

–

–

–

1

–

Luca et al.22

–

–

–

–

–

–

–

–

Figures for continuous outcomes represent median values with range in parentheses unless 6 stated, which denotes standard deviation. tumour #1 mm from circumferential cut edge. †, *Multiple publications by the same departments with likely duplication of reported outcomes. a

b

160  Robotic total mesorectal excision

HEBK001-C10_p154-170.indd 160

Table 10.3. Oncological outcomes from studies reporting outcomes of robotic rectal cancer surgery.

11/02/13 3:58 PM

ROBOTIC RECTAL SURGERY  161

Overall, few studies have shown substantial differences in short-term outcomes between robotic and laparoscopic resections (Tables 10.1 and 10.2). Two studies from the same unit have revealed a significant reduction in length of stay and ­‘serious complications’ in patients undergoing robotic surgery, but most of these small, and almost exclusively non-randomized, studies have shown ­robotic surgery to be at least as safe as, and oncologically equivalent to, laparoscopic surgery (see Table 10.3).18,24 There is a need for further exploration in large-scale randomized trials to determine whether robot-assisted TME is superior in oncological and quality-of-life terms to laparoscopic and open surgery.

Patient Preparation The preoperative work-up for a patient with rectal cancer begins with a full history and physical examination. Blood tests should include a full blood count, electrolytes, liver function tests, carcinoembryonic antigen (CEA) levels and a 2 unit cross-match. In our experience, blood transfusion is rarely required during laparoscopic or robotic rectal cancer surgery, but it must be remembered that if serious bleeding occurs undocking the robot can lead to delays in controlling the bleeding point. An electrocardiogram is requested; where appropriate, cardiopulmonary exercise testing is increasingly used to help to stratify operative risk and optimize perioperative management. Full colonic imaging by colonoscopy, barium enema or computed tomographic (CT) pneumocolon evaluation is performed to exclude synchronous tumours, and histological diagnosis is confirmed on biopsy. The distance from the lower border of the tumour to the anal verge is measured using rigid sigmoidoscopy; accurate measurement is essential for surgical planning. Clinical evaluation of sphincter tone and careful assessment of premorbid anorectal function may help to identify patients in whom a low anastomosis would lead to unacceptable postoperative outcomes. Computed tomography of the chest, abdomen and pelvis is performed to assess for distant metastases and local disease infiltration. The rectal cancer is staged by pelvic magnetic resonance imaging (MRI) to measure the extent of local infiltration and identify patients with threat-

HEBK001-C10_p154-170.indd 161

ened resection margins(see Chapter 4). Endoluminal ultrasound (see Chapter 3) can help to differentiate between T1 and T2 tumours. In the authors’ unit, all patients undergoing TME with anastomosis have preoperative oral mechanical bowel preparation and a defunctioning loop ileostomy. Patients are counselled and sited by a stoma therapist preoperatively.

Position The patient is most commonly placed in a modified Lloyd Davies position, with the hips slightly extended and the knees flexed to 70–908 and placed just greater than shoulder width apart. In our experience, the use of a body-length gel mat in direct contact with the patient’s skin keeps the patient safely on the operating table without the need for shoulder supports. Thromboembolic deterrent stockings (unless contraindicated) and intermittent calf compression are used routinely. A warm air blanket is placed over the patient’s chest. The arms are wrapped at the patient’s sides and protected from the metalwork of the table. The patient is catheterized. Some surgeons prefer to use a lateral approach to the splenic flexure, in which case the patient is positioned accordingly – left side uppermost, with the hips straight and knees bent. Lateral supports are necessary against the patient’s back. A left arm support gutter is also used. (If the hips are flexed, the left thigh will hinder the smooth and efficient use of instruments via the left lateral port.) After mobilizing the splenic flexure, the patient is turned on to the back and positioned in the Lloyd Davies position, as described above.

Surgical Approach The authors favour routine mobilization of the splenic flexure when performing an anterior resection with TME. This is done at the start of the procedure as it is often technically difficult. Once completed, there is no temptation to compromise if the rectal dissection proves difficult and time-consuming. Although mobilization of the splenic flexure and left colon with high ligation of the inferior mesenteric artery and vein can be performed using the da Vinci system, the authors favour a hybrid approach, with standard laparoscopy

11/02/13 3:58 PM

162  Robotic total mesorectal excision

for the abdominal component of the operation followed by robotic rectal dissection. Port sites

The optimum port siting is the subject of much conjecture, and many arrays of ‘standardized’ port site are described. The authors’ preferred set-up has evolved from assessment of the published techniques and as a result of trial and error. A compromise in placement is required to facilitate safe and efficient laparoscopic mobilization, while still allowing effective robotic instrument use in the pelvis, if the overall number of port sites is to be minimized. A report by Luca and colleagues describes a port site set-up that allows for complete low anterior resection and does not require the robot cart to be moved.25 The authors’ favoured sites are as follows: A supraumbilical incision is made and a 12-mm internal diameter port is inserted using an open Hassan technique to accommodate the large-diameter three-dimensional robotic telescope and camera. The pneumoperitoneum is established with an insufflation pressure set to 12–15 cmH2O. A 10-mm laparoscope is used to assess the feasibility of a minimal access approach in terms of adhesions, adiposity of the abdominal wall and mesenteries, and anatomical anomalies, which may hinder or prevent progress. The procedure continues by insertion of the operating and assisting robotic 8-mm ports. These are inserted under direct vision via marked skin incisions positioned optimally to avoid clashing of the robotic arms. A minimum of 8 cm is recommended between ports. The robotic cannulae are marked with broad black bands that represent the fulcrums for angulation. These should lie just within the abdominal wall to allow effective movement and minimize postoperative pain. The authors favour the use of robotic ports in standardized positions during the laparoscopic part of the procedure. An additional 12-mm port in the right iliac fossa, 3–4 cm inferomedial to the anterior superior iliac spine, is used for laparoscopic instrumentation and for assistant access during the robotic phase of the procedure. This allows access for clip applicators, suction/irrigation devices, insertion and removal of tonsil and mastoid swabs, and stapling devices. Ports are sited to optimize triangulation of instruments, avoid clashing and provide maximum

HEBK001-C10_p154-170.indd 162

field of view. The camera port is therefore placed in the mid-position of the instrumental port site arc, ideally just above or to the right of the umbilicus, so that the field of vision is centred on the midline. The primary operative robotic 8-mm port is placed on the right, 8–10 cm laterally along the arc from the umbilicus to the right anterior superior iliac spine. The secondary operative robotic 8-mm port is placed 8–10 cm left laterally along the arc from the umbilicus to the left anterior superior iliac spine. The tertiary operative robotic 8-mm port is placed 8–10 cm laterally to this along the same arc at a site roughly corresponding to the 12-mm laparoscopic assistant port opposite (Figure 10.1). For a small patient, the second operating arm can be placed a few centimetres cranially to maintain optimum spacing between ports. Right lateral approach to the splenic flexure

If a lateral approach is chosen, the first port to be placed is in the left lateral port position using a 5-mm optical port, inside which a 5-mm laparoscope is used to enter the abdominal cavity under

A2

12 R1

R2 R3

A1

Figure 10.1. Robotic instrument ports R1, R2 and R3. R1 is the main operating port, with traction and countertraction provided by R2 and R3. The 12-mm port is for the camera just to the right of the umbilicus. A1 and A2 are assistant ports. A1 is also used for stapling the rectum.

11/02/13 3:58 PM

THE DA VINCI SURGICAL SYSTEM  163

vision. This port can be replaced later by the 8-mm robotic port. The laparoscopy is performed and the 12-mm umbilical port can then be inserted under direct vision. An additional 5-mm port is placed in the left upper midclavicular line for the purposes of laparoscopic splenic flexure mobilization. Mobilization of the splenic flexure

The right lateral position for splenic flexure mobilization is particularly useful in patients who are obese, as the entire enteric mass falls away under the influence of gravity; in some cases, the entire lateral mobilization of the left colon down to the pelvic brim can be achieved in this position. More commonly, however, splenic flexure mobilization is performed with the patient supine, tilted left side up towards the operating surgeon. The small bowel should fall away but may require gentle sweeping. Occasionally, peritoneal adhesions prevent this and require division to partially mobilize the duodenojejunal flexure. Once the lateral border of the duodenojejunal flexure is clearly identified, the inferior mesenteric vein can be identified running inferolaterally within the left colonic mesentery. The peritoneum is incised to allow identification and isolation of the inferior mesenteric vein. The vein is divided with an appropriate haemostatic energy source, or between clips, at the level of the lower border of the pancreas. Dissection continues medial to lateral in the plane between the mesentery and Gerota’s fascia towards the lateral abdominal wall. The lesser sac is entered just above the body of the pancreas, where the left branch of the middle colic artery marks the most medial aspect of the dissection. The posterior wall of the stomach is clearly seen superiorly, with the tail of the pancreas inferiorly and the transverse mesocolon anteriorly after its disconnection from the pancreas. The splenic flexure and descending colon are anterolateral. Lateral attachments of the descending colon and splenic flexure are divided, and the flexure is freed by re-entering the lesser sac from above by dissecting the greater omentum free from the transverse colon. Division of inferior mesenteric artery and sigmoid mobilization

The patient is placed in a steep head-down position, with the right lateral tilt maintained. The omentum is reflected over the liver and the mobilized flexure

HEBK001-C10_p154-170.indd 163

placed back in the left upper quadrant. The small bowel is gently swept out of the pelvis up into the right upper quadrant. The pelvic brim is identified at the sacral promontory, and the sigmoid colon is tented up to reveal the fossa beneath the inferior mesenteric artery (IMA) over the bifurcation of the aorta. The peritoneum is incised along a broad front to allow entry into, and then medial to lateral dissection within, Toldt’s plane. The retroperitoneal structures are carefully swept posteriorly, preserving the left ureter and gonadal vessels. Once these are identified and preserved, attention is focused on isolation of the IMA. Taken high, this allows for a single vessel division. It is important to identify and preserve the pre-aortic sympathetic nerves at this point and to avoid taking the vessel flush with the aorta, as this may cause nerve damage. There are many ways to secure the artery, including clips, stapling devices and energy instruments. When taken more distally, the ascending left colic artery and sigmoid branches may be encountered, necessitating further dissection and separate vessel division. The dissection is continued in Toldt’s plane as far out laterally as the abdominal wall and cranially until the previously dissected descending colonic mesentery is reached. Inferiorly, the dissection reaches the plane between the mesorectal fascia and the pelvic parietal fascia. The lateral attachments are then divided, taking particular care to avoid the left ureter across the pelvic brim, such that the entire left colon is then mobilized. It is then the authors’ preference to tie a nylon tape around the rectosigmoid junction to facilitate retraction of the rectum out of the pelvis during mesorectal dissection.

THE DA VINCI SURGICAL SYSTEM The robotic system consists of three main components: the operating console, the electronic tower and the patient cart. The console (Figure 10.2) is where the surgeon controls the robotic arms and instruments. It comprises scissor-handle-type master controls, which translate the surgeon’s hand movement directly to the instrument tips, including the ‘wristing’ angulation and rotation (Figure 10.3). There are foot pedals for camera control, disengaging instruments (clutch), focusing and the application of monopolar and bipolar diathermy (Figure 10.4).

11/02/13 3:58 PM

164  Robotic total mesorectal excision

cart with specially designed transparent sterile covers (Figure 10.6). The stereoscopic scope, camera and light lead are also prepared in a sterile fashion with a special disposable outer sleeve. There is a range of dedicated robotic instruments, including dissectors, graspers and retractors. Where appropriate, these instruments can be connected to monopolar or bipolar diathermy sources. Each instrument has a lifespan of ten procedures. To minimize cost, it is important to develop a technique using as few instruments as possible. The authors prefer the use of a diathermy hook or monopolar diathermy scissors (‘hot shears’) in the right-side operative port, Maryland bipolar forceps in the primary assistant port, and a large grasping retractor or Cadiere forceps in the secondary assistant port. The central robotic arm holds the camera. Arms 1, 2 and 3 hold the operating instruments introduced through the 8-mm reusable robotic ports.

Figure 10.2. Operating console.

The electronic tower holds the video-processing system, light source and insufflation equipment. All the electronics run through the control tower between the console and the patient cart. The cart also has a high-definition screen to transmit a two-dimensional image of the operating field and allows a trainer to draw lines on the screen to help a trainee sitting at the console locate the correct planes of dissection. The patient-side cart consists of a powered trolley holding three or four robotic arms (Figure 10.5). The scrub nurse is responsible for preparing the robot

Set-Up of Robotic Total Mesorectal Excision The patient remains in a steep head-down right lateral tilt to prevent the small bowel from falling into the pelvis. In female patients with an intact uterus, a straight-needled polypropylene suture is driven through the suprapubic skin through the fundus of the hitched-up uterus and back through the suprapubic skin and tied over a swab to prevent damage to the skin and to act as a reminder to remove the suture at the end of the operation. The authors find this gives better uterine elevation than lifting the uterus by the broad ligament

Figure 10.3. Hand controls for robotic surgical instruments.

HEBK001-C10_p154-170.indd 164

11/02/13 3:58 PM

THE DA VINCI SURGICAL SYSTEM  165

Figure 10.4. Foot pedals.

and fallopian tubes and also requires only a single suture.

Positioning the Robot The choice of robot docking position is entirely personal to the surgeon and will be influenced by both positive and negative experiences in terms of ease of dissection and minimization of arm clashes and instrument clashes. The unique disadvantage of colorectal resection, compared with prostatic surgery for example, is the necessity to operate in multiple, and comparatively larger, anatomical sectors (e.g. 30 cm for splenic flexure and descending colon and 20 cm for the rectum versus 4–5 cm for the retropubis). A low anterior resection with TME requires at least two fields of view, and double docking of the robot adds time to the procedure. There are few advantages to robotic splenic flexure mobilization compared with standard laparoscopic techniques, since the space is less confined and the surgery often requires relatively large retraction movements crossing the surgical field. As a result, a hybrid procedure is often preferred. This eliminates the compromised position of the robot positioned by the left leg and allows the more optimal between-the-legs docking position. The set-up involves manoeuvring the patientside cart into position between the patient’s legs. The central hub, trolley and camera arm of the robot should all be in line with the patient’s midline. This enables the ‘sweet spot’ of the camera arm to be found with ease and without the camera arm getting in the way of the other instrument arms.

HEBK001-C10_p154-170.indd 165

The role of the assistant is important and far from redundant. A scrubbed assistant knowledgeable in the set-up and function of the da Vinci system is absolutely essential to the smooth running and efficiency of the procedure. Set-up (docking) can be time-consuming and present risk to the patient if not performed carefully and meticulously. The robot is carefully docked between the patient’s legs, and the laparoscopic instruments, including the camera and scope, are stored in sterile fashion for later use. The pneumoperitoneum is maintained and the

Figure 10.5. Patient-side cart with operating arms.

11/02/13 3:58 PM

166  Robotic total mesorectal excision

Figure 10.6. Patient-side cart set up and draped.

ports remain in place. The robotic arms are positioned carefully under clutch control to the proximity of the robotic ports. Each is then carefully manoeuvred and fixed to its respective ports using their clasp attachments. The assistant is then responsible for insertion of the camera and instruments under the careful instruction of the surgeon now sitting at the console. The camera is introduced first to allow for safe introduction of the other robotic instruments. Although in most cases the 08 scope is suitable for visualization, when difficulty is encountered the 308 scope can be used with good effect. The angled view is fixed either ‘down’ or ‘up’ for the robotic surgeon, however, who does not have the advantage afforded the laparoscopic surgeon of being able to rotate the lens of the 308 scope to view the surgical field from varying angles while still maintaining the horizon. Each arm is fitted in turn with an appropriate robotic instrument. The primary arm usually holds the scissors or hook dissector with monopolar diathermy attachment. The secondary arm usually holds the Cadiere or Maryland bipolar diathermy graspers. The third arm holds a further grasper of the surgeon’s choice; a longer grasper that doubles as a retractor is a useful instrument for this arm. The assistant can lift the rectum out of the pelvis using another laparoscopic grasper holding the nylon tape as required. Thus, three-point traction can be achieved, facilitating precise dissection in the correct plane in a confined space. This dissection is under the continuous control of the operating surgeon at the console. The assistant must be continuously alert to the risk of robotic arm clashing and communicate regularly with the surgeon. The assistant must also be ready to improve vision by aspiration of the dia-

HEBK001-C10_p154-170.indd 166

thermy plume and provide tonsil and mastoid swabs as necessary. The assistant makes a valuable contribution throughout the procedure in helping to provide the essential traction and counter-traction to progress with the dissection; this can usually be performed with the suction and irrigation device. It is important to remember that each instrument attached to a robot arm has a fixed limited number of uses, and therefore careful instrument choice can limit wastage and unnecessary expense.

Pelvic Dissection Sharp dissection using a hook or scissors attached to monopolar diathermy is preferred by the authors. The harmonic scalpel can be used, but there is no hinge mechanism on the wrist of this instrument. The bloodless mesorectal plane between the mesorectum and parietal presacral pelvic fascia is demonstrated easily due to the enhanced view and by gentle anterior traction using a tonsil swab held by a grasping instrument. The swab provides a blunt broad compression of the posterior mesorectum, minimizing the risk of fascial breach. It also absorbs small amounts of blood and tissue fluid. The dissection continues as far as the pelvic floor behind and down the right side of the mesorectum. It is essential for the surgeon to remember that the sacral concavity arcs forwards as dissection progresses, but this is well demonstrated with the three-dimensional view offered by the robot. Extra care must be taken not to breach the parietal pelvic fascia; this is particularly important in the midline

11/02/13 3:58 PM

THE DA VINCI SURGICAL SYSTEM  167

posteriorly, where the presacral veins can be damaged, leading to haemorrhage that can be difficult to control. Lateral dissection should be performed predominantly from the right side; it is here that the hinged robotic instruments offer a significant advantage over standard laparoscopy. The peritoneum is opened laterally, taking care not to inadvertently damage the inferior hypogastric nerve or ureter. The posterolateral dissection is taken down to a comfortable level, where progress starts to become a little more difficult; at this stage, the anterior dissection is started. The peritoneum is divided anterior to and above the reflection. The anterior dissection continues behind the seminal vesicles in male patients, at which point Denonvilliers’ fascia is reached. Denonvilliers’ fascia may exist in one or two layers, the anterior layer behind the prostate and the posterior layer in front of the anterior mesorectum. When two layers are present, they fuse at the lower level of the prostate. At this point, the fascia is divided to enter the plane in front of the mesorectum. At the lateral corners of Denonvilliers’ fascia, the surgeon needs to be aware of the close proximity of the neurovascular bundles, as these are easily damaged at this point. In female patients, the dissection continues down the rectovaginal septum to the pelvic floor. The mesorectum is often very thin and easily breached at this level. If dealing with an advanced anterior T3 or T4 tumour in the lower rectum, it is often necessary to remove some of the posterior vaginal wall en bloc with the specimen. An advanced upper rectal tumour may require an en-bloc hysterectomy. The anterior dissection is perhaps the most challenging part of the operation and may require anterior retraction of the prostate to allow the dissection to proceed to the pelvic floor under direct vision. The mesorectal plane guides the surgeon along the pelvic floor to the rectal tube, which is encountered as it passes through the pelvic floor; at this point, the dissection can continue into the intersphincteric plane if required. The surgeon switches from anterior to posterior to lateral as dictated by view and ease of access. Laterally the retraction and three-dimensional view aids the identification and preservation of the inferior hypogastric plexus and nervi erigentes. A branch of the inferior hypogastric nerve passes medially to the rectum (previously known as the lateral ligament); this nerve is variable in thickness and will need to be divided. Occasionally the nerve is accompanied by the mid-

HEBK001-C10_p154-170.indd 167

dle rectal vessels, but these are by no means consistent and are generally small and insignificant. If progress is not being made in one area, it is important to move to another. It is usually possible to dissect the left side of the mesorectum from the right side of the patient. The mesorectal dissection is started on the right and posterolaterally to the left pelvic side-wall from the right side of the patient; when coming over to the other side of the rectum, all that remains is the division of the peritoneum and to complete the TME. Finally, any lower parts of the mesorectum are dissected off the pelvic floor to reveal a clean rectal tube. The muscle tube can then be divided at the pelvic floor.

Rectal Transection A further nylon tape is tied around the rectal tube, taking care that it is tightened well below the lower border of the tumour. The authors favour a distal rectal washout with a cytocidal solution of iodine and water. At this point, there are two choices. The rectum can be divided with the robotic cutting instrument, followed by the insertion of a purse-string suture using the robotic needle driver;26 this may facilitate transanal extraction of the specimen.27 Alternatively, if there is space for a linear laparoscopic stapler, this can be inserted to divide in the standard laparoscopic fashion. The instrument cart can then be undocked after careful removal of the robotic instruments and camera. An articulated 45-mm linear stapler cutter is then inserted via the 12or 15-mm (right iliac fossa) port. The distal nylon tape is used to position the rectal tube in the jaws of the stapler. Once precise positioning has been achieved, the stapler is closed and then fired after 20 s of tissue compression. The stapler is withdrawn and the nylon tape is grasped with a ratcheted toothed grasper so that the specimen can be easily found for extraction. There are often difficulties getting a laparoscopic stapler into a tight male pelvis, and it is vital not to compromise the distal resection margin because of inadequate instrumentation. Currently the technology is not available to provide the huge forces necessary to generate compression and closure of the device around an acute angle without a large joint between the stapler head and the driving mechanism. In these circumstances, it is better to enlarge the Pfannenstiel incision and apply an alternative narrow stapling device manually, as commonly used in open surgery. These staplers can be

11/02/13 3:58 PM

168  Robotic total mesorectal excision

manipulated laparoscopically providing a seal is established to prevent leakage of the CO2 pneumoperitoneum. Alternatively, the rectal muscle tube can be divided distal to the nylon tape using the da Vinci scissors; a purse-string suture can then be inserted using the da Vinci needle drivers, avoiding the need for a linear stapler. Other options include a sutured anastomosis performed either robotically or with a peranal approach.

Specimen Withdrawal A 5- to 7-cm suprapubic transverse incision is made and a self-retracting dual ring-reinforced wound protector is used to facilitate specimen extraction. When the mesorectum is not too bulky, the specimen may be removed through the anal canal. The left colon is divided at a point of convenience; there should be a good colonic blood supply based on the marginal artery, and sufficient length to reach the pelvic floor without tension. Some surgeons may also favour the fashioning of a colo-pouch. A circular stapler anvil is inserted in the proximal colon and secured with a purse-string suture. The colon is then returned to the abdomen and the wound protector loosened by unravelling the ring reinforcer to allow torsional closure. This is retightened using the ring reinforcer over a swab wrapped around the closed core. Pneumoperitoneum can then be re-established and an end-to-end, end-to-side or end-to-pouch anastomosis fashioned with the circular stapling device. The anastomosis can be tested for air leaks, but this is usually unnecessary as it is the authors’ routine practice to defunction all TME coloanal/low colorectal anastomoses with a loop ileostomy. It is our practice to carefully inspect the doughnuts and examine the anastomosis digitally to exclude obvious flaws.

Advantages and disadvantages of robotic colorectal surgery Technical Advantages Laparoscopic colorectal surgery is technically very demanding and the learning curve is often long. There is considerable restriction of movement with standard laparoscopic instruments, with a maxi-

HEBK001-C10_p154-170.indd 168

mum of four degrees of freedom. Movements are compromised somewhat by dependence upon the fulcrum effect of the instruments having to pass via ports through the abdominal wall. The intuitive and articulated instrument movements of the robot have seven degrees of freedom, which are particularly helpful in the confined space of the pelvis. The ports are simply a means of access rather than an integral part of instrument movement, as with standard laparoscopy. The robot eliminates tremor and scales down movements made at the surgeon console, enhancing the accuracy of dissection. The image of the operative field provided by highdefinition camera systems in laparoscopic surgery is now excellent. Depth of field has to be interpreted by the surgeon and may adversely affect performance. The high-definition three-dimensional view provided by the robotic optics is vastly superior to the high-definition two-dimensional views produced by standard laparoscopic cameras and display monitors. The camera is held perfectly still, avoiding the risk of assistant fatigue and reducing disorientation of the operating surgeon. These advantages allow for more precise, meticulous, sharp dissection of clearly identified tissue planes, resulting in less blood loss and reduced surgical trauma. The enhanced view, intuitive instrument movements and improved dissection quality are likely to steepen, and therefore shorten, the learning curve for minimally invasive TME and reduce the chance of the need to convert to a standard open procedure. Robotic TME may demonstrate improvements in oncological circumferential resection margins and pelvic nerve preservation. It is important to emphasize, however, that this is purely conjecture until the results of clinical trials such as the Robotic versus Laparoscopic Resection for Rectal Cancer (ROLARR) trial, designed to specifically investigate robotic TME compared with laparoscopic TME, become available.

Disadvantages One of the main disadvantages is the complete lack of tactile feedback from the instruments. This can potentially lead to tissue damage both in and out of the field of view. A second drawback is the time-consuming procedure of the docking and undocking of the robotic cart from the patient. If significant bleeding occurs, robotic instruments cannot be changed

11/02/13 3:58 PM

References  169

quickly to deal with the problem, and the undocking process may lead to greater blood loss before control is achieved either laparoscopically or by open surgery. Technological advances are in progress to create hinged harmonic/ultrasonic energy devices and robotically controlled suction devices, and the latter are now available on the Si model. Suction can easily be added via the assistant port, however, and also makes an excellent additional retractor to provide counter traction. Another disadvantage is the duration of surgery. This applies to the set-up (docking) and the procedure itself. Operating times will reduce with progression up the learning curve, but until mastery is achieved they will always be longer than with standard laparoscopic surgery or open surgery. This has implications for list planning, service provision and training of juniors in the future. A current main issue is cost, as the da Vinci system is expensive. This applies not only to the initial outlay for the robot but also to the cost of consumables and the annual servicing contract. These disadvantages must be weighed against the potential benefits of unrivalled views of the anatomy, reduced blood loss through accurate dissection of the tissue planes, and potentially improved nerve preservation and quality of the TME specimen.

THE FUTURE It is likely that robotic technology will improve with time and become more cost-effective, smaller and efficient. There is enormous potential to improve the quality of mesorectal dissection, and we believe this is another step of progress in the surgical treatment of rectal cancer.28 Cross-stapling the low rectum in a narrow pelvis remains a challenge with the minimal access staplers currently available, but robotic pursestring suturing may help overcome some of these difficulties in the short term. There are numerous case series and several comparative studies demonstrating equivalence or improvements in oncological and functional outcome measures. Robotic rectal cancer surgery is currently being evaluated through international collaboration in a worldwide multicentre randomized controlled trial of robotic versus laparoscopic TME (the ROLARR trial), whose main outcome measures are rates of conversion to open surgery. Secondary measures are quality of re-

HEBK001-C10_p154-170.indd 169

sected specimen, positivity of resection margins, local recurrence, quality of life, pelvic nerve function and health economics analysis. An excellent review by Mirnezami and colleagues confirms there is no evidence in the medical literature to demonstrate an advantage of using the da Vinci robot for rectal cancer surgery over standard laparoscopic surgery.28 There is no doubt that the view of the pelvic dissection provided by the da Vinci three-dimensional camera system is unrivalled, making it easier to identify important anatomical structures and dissect without damage to the pelvic nerves. The intuitive hinged instruments also provide an advantage when dissecting in the narrow confines of the pelvis. These are subjective rather than objective observations, and translating them into an evidencebased advantage may prove difficult. The literature does, however, conclude that robotic TME is at least as good and as safe as conventional laparoscopic surgery, and it is therefore important that we engage with and develop the technology with the hope of improving the quality of rectal TME surgery in the future.

References   1. Kang CM, Chi HS, Hyeung WJ, et al. The first Korean experience of telemanipulative robot-assisted laparoscopic cholecystectomy using the da Vinci system. Yonsei Med J 2007; 48: 540–5.   2. Choi SB, Park JS, Kim JK, et al. Early experiences of robotic-assisted laparoscopic liver resection. Yonsei Med J 2008; 49: 632–8.   3. Baik SH, Kim YT, Ko YT, et al. Simultaneous robotic total mesorectal excision and total abdominal hysterectomy for rectal cancer and uterine myoma. Int J Colorectal Dis 2008; 23: 207–8.   4. Patriti A, Ceccarelli G, Bartoli A, Spaziani A, Casciola L. Laparoscopic and robot-assisted onestage resection of colorectal cancer with synchronous liver metastases: a pilot study. J Hepatobiliary Pancreat Surg 2009; 16: 450–7.   5. Weber PA, Merola S, Wasielewski A, Ballantyne GH. Telerobotic-assisted laparoscopic right and sigmoid colectomies for benign disease. Dis Colon Rectum 2002; 45: 1689–94, 1695–6.   6. Delaney CP, Lynch AG, Senagore AJ, Fazio VW. Comparison of robotically performed and traditional laparoscopic colorectal surgery. Dis Colon Rectum 2003; 46: 1633–9.   7. D’Annibale A, Morpurgo E, Fiscon V, et al. Robotic and laparoscopic surgery for treatment of colorectal disease. Dis Colon Rectum 2004; 47: 2162–8.

11/02/13 3:58 PM

170  Robotic total mesorectal excision   8. Rawlings AL, Woodland JH, Crawford DL. Telerobotic surgery for right and sigmoid colectomies: 30 consecutive cases. Surg Endosc 2006; 20: 1713–8.   9. Rawlings AL, Woodland JH, Vegunta RK, Crawford DL. Robotic versus laparoscopic colectomy. Surg Endosc 2007; 21: 1701–8. 10. Baik SH, Kim NK, Lee KY, et al. Factors influencing pathologic results after total mesorectal excision for rectal cancer: analysis of consecutive 100 cases. Ann Surg Oncol 2008; 15: 721–8. 11. Heald RJ, Husband EM, Ryall RD. The mesorectum in rectal cancer surgery: the clue to pelvic recurrence? Br J Surg 1982; 69: 613–6. 12. Nagtegaal ID, van de Velde CJ, van der Worp E, et al. Macroscopic evaluation of rectal cancer resection specimen: clinical significance of the pathologist in quality control. J Clin Oncol 2002; 20: 1729–34. 13. Baik SH, Lee WJ, Rha KH, et al. Robotic total mesorectal excision for rectal cancer using four robotic arms. Surg Endosc 2008; 22: 792–7. 14. Pigazzi A, Ellenhorn JD, Ballantyne GH, Paz IB. Robotic-assisted laparoscopic low anterior resection with total mesorectal excision for rectal cancer. Surg Endosc 2006; 20: 1521–5. 15. Hellan M, Anderson C, Ellenhorn JD, Paz B, Pigazzi A. Short-term outcomes after robotic-assisted total mesorectal excision for rectal cancer. Ann Surg Oncol 2007; 14: 3168–73. 16. Baek JH, McKenzie S, Garcia-Aguilar J, Pigazzi A. Oncologic outcomes of robotic-assisted total mesorectal excision for the treatment of rectal cancer. Ann Surg 2010; 251: 882–6. 17. Baik SH, Kang CM, Lee WJ, et al. Robotic total mesorectal excision for the treatment of rectal cancer. J Robotic Surg 2007; 1: 99–102. 18. Ng SS, Lee JF, Yiu RY, Li JC, Hon SS. Teleroboticassisted laparoscopic abdominoperineal resection for low rectal cancer: report of the first case in Hong Kong

HEBK001-C10_p154-170.indd 170

and China with an updated literature review. World J Gastroenterol 2007; 13: 2514–8. 19. Spinoglio G, Summa M, Priora F, Quarati R, Testa S. Robotic colorectal surgery: first 50 cases experience. Dis Colon Rectum 2008; 51: 1627–32. 20. Baik SH, Ko YT, Kang CM, et al. Robotic tumor-specific mesorectal excision of rectal cancer: short-term outcome of a pilot randomized trial. Surg Endosc 2008; 22: 1601–8. 21. Baik SH, Kwon HY, Kim JS, et al. Robotic versus laparoscopic low anterior resection of rectal cancer: short-term outcome of a prospective comparative study. Ann Surg Oncol 2009; 16: 1480–7. 22. Choi DJ, Kim SH, Lee PJ, Kim J, Woo SU. Single-stage totally robotic dissection for rectal cancer surgery: technique and short-term outcome in 50 consecutive patients. Dis Colon Rectum 2009; 52: 1824–30. 23. Baik SH, Kim NK, Lee KY, et al. Factors influencing pathologic results after total mesorectal excision for rectal cancer: analysis of consecutive 100 cases. Ann Surg Oncol 2008; 15: 721–8. 24. Quirke P, Steel R, Monson J, et al. Effect of the plane of surgery achieved on local recurrence in patients with operable rectal cancer: a prospective study using data from the MRC CR07 and NCIC-CTG CO16 randomised clinical trial. Lancet 2009; 373: 821–8. 25. Luca F, Cenciarelli S, Valvo M, et al. Full robotic left colon and rectal cancer resection: technique and early outcome. Ann Surg Oncol 2009; 16: 1274–8. 26. Prasad LM, deSouza AL, Marecik SJ, Park JJ, Abcarian H. Robotic pursestring technique in low anterior resection. Dis Colon Rectum 2010; 53: 230–4. 27. Kang J, Min B, Hur H, Kim N, Lee K. Transanal specimen extraction in robotic rectal cancer surgery. Br J Surg 2012; 99: 133–6. 28. Mirnezami A, Mirnezami R, Venkatasubramaniam A, Chandrakumaran K, Cecil T, Moran B. Robotic colorectal surgery: hype or new hope? A systemic review of robotics in colorectal surgery. Colorectal Dis 2010; 12: 1084–93.

11/02/13 3:58 PM

11 Local excision and transanal endoscopic microsurgery Wolfgang B. Gaertner and David A. Rothenberger

Introduction The role of local excision (LE) in the treatment of rectal cancer is ever-evolving. Over the past 125 years, its use has varied widely, depending on the intent of therapy (curative, palliative or compromised); specific treatment goals; the safety, effectiveness and local availability of alternative forms of therapy for rectal cancer; and the decision-making process and treatment philosophy of those involved in choosing an optimal treatment plan, including the patient and the surgeon. At one time, LE through a posterior approach was the most common method used to treat patients with rectal cancer, for both curative and palliative intent. At other times in the twentieth century, LE was rarely used for curative-intent treatment of rectal cancer in normal-risk patients, who instead were treated by standard anterior resection or abdominoperineal excision (APE). Local excision was used primarily for palliation or in compromise situations, such as when a patient’s operative risks and comorbidities made it too dangerous to undergo standard radical resection or because the patient refused the recommended anterior resection or APE. Today, the goals for treating rectal cancer have broadened to include securing local and distant oncological control; minimizing treatment-related mortality and morbidity; performing restorative

HEBK001-C11_p171-190.indd 171

anastomosis to achieve near-normal continence and defecation; preserving genitourinary functions; and promoting rapid recovery from surgery, with a prompt return to normal activities. This broadened list of goals has led many surgeons to question the dogma that all normal-risk patients with a potentially curable invasive cancer are best served by standard radical resection (anterior resection or APE). Instead, they suggest a policy that uses LE in the curativeintent treatment of patients with selected stage I rectal cancer. Although this approach is highly controversial, recent data confirm that US surgeons are increasingly using LE for curative-intent treatment of rectal cancer, despite a lack of level I or II evidence that LE is oncologically equivalent to standard radical resection. Proponents of this practice justify their increased use of LE because it better achieves the other goals now considered important in treating patients with rectal cancer and because it is unclear that overall survival is uniformly better after radical resection for stage I cancer compared with LE. In addition to its use as the sole treatment modality in curative surgery for selected rectal cancer, LE is now being used with adjuvant or neoadjuvant therapies. These controversies and the indications, techniques and outcomes following conventional endoanal LE and transanal endoscopic microsurgery (TEM) for selected patients with rectal cancer are the subject of this chapter. The role of LE in the treatment of benign rectal neoplasms and

08/02/13 5:37 PM

172  Local excision and transanal endoscopic microsurgery

the potential negative consequences of LE of what is thought to be a benign rectal neoplasm, but is found by histopathology on the excised specimen to harbour a focus of invasive adenocarcinoma, are beyond the scope of this chapter.

Historical evolution of local excision techniques Surgeons have devised multiple techniques of LE for rectal cancer that can be grouped broadly by operative approach: posterior, perineal and endoanal.

Posterior Approach Local resection via the posterior approach was the preferred treatment for most rectal cancers in the late 1800s, when mortality following anaesthesia and abdominal surgery was common. The posterior transsacral approach was popularized in Europe by Kraske, who described making a posterior longitudinal incision over the distal sacrum and coccyx to the midline raphe of the perineum with the patient in the prone jack-knife position.1 The rectal tumour was exposed and resected with a 1 cm margin. The rectal wall defect was repaired and the wound closed. This became the most common approach for rectal cancer in Europe in the late 1800s and early 1900s. Mason subsequently modified this posterior approach by dividing the anal sphincters to achieve better exposure.2 The sphincter muscles were carefully preserved and tagged to facilitate suture repair in layers after excising the rectal lesion. Although still useful for several benign conditions, both the Kraske trans-sacral and the Mason trans-sphincteric posterior approaches to local resection for rectal cancer have now essentially been abandoned because of associated operative morbidity (posterior faecal fistulas, anal incontinence), the increased risk of incurable local recurrences, and the development of safer, more effective alternative techniques.

Perineal Approach Lockhart-Mummery advocated a two-stage approach to the local resection of rectal cancer.3 The first stage consisted of constructing a colos-

HEBK001-C11_p171-190.indd 172

tomy under regional anaesthesia. Two weeks later, through a transperineal incision, Lockhart-Mummery performed a wide local resection of the anus with its surrounding skin, the rectum, the contents of the ischiorectal fossas, part of the levators and as much of the sigmoid with its mesentery and lymph nodes as possible. In 1926, he reported a remarkably low (for the time) operative mortality of 8.5 per cent in 200 patients.3 The Lockhart-Mummery technique offered the potential advantages of a more aggressive clearance of the cancer compared with the limited resection done with the Kraske trans-sacral approach, but with significantly less mortality than the 42 per cent reported in 1908 by Miles after APE.4 It is interesting to note that Lockhart-Mummery had used the Miles technique but said he ‘had abandoned it since 1913, except for the high cancers at the rectosigmoidal junction’. For all other cancers of the rectum, Lockhart-Mummery preferred his transperineal approach because it ‘was attended with a lower mortality and was applicable to a larger proportion of cases’.

Endoanal Approach One of the earliest descriptions of use of the endoanal approach to LE of rectal cancer was by von Volkmann in 1878, a mentor to Kocher. He noted that it was possible to treat ‘a well-circumscribed tumour, with the removal of which requires excision of a small portion of the rectum’, providing it was possible to close the defect in the rectum primarily by suture, preferably in a transverse fashion to avoid the complication of a stricture. In 1977 Morson reported the St Marks Hospital, London, experience with LE of rectal cancers using the endoanal approach of Sir Alan Parks.5 His endoanal technique is the prototype of the conventional technique used for LE worldwide. The morbidity is markedly less than that observed after radical resection. Technical details and outcomes are described below. Transanal endoscopic microsurgery (TEM) was introduced by Gerhard Buess of Tubingen, Germany, in 1984.6 This is a modification of LE that combines the excellent visualization offered by endoscopy with advanced-instrument technology. The technique allows for improved endoanal access to the mid- and upper rectum, thus increasing the utility of LE. Visual imaging is achieved through a

08/02/13 5:37 PM

Current techniques of local excision  173

binocular stereoscope, which permits an optimum view during the procedure, thus enhancing the surgeon’s ability to accurately perform full-thickness excisions and to repair the rectal wall defect. The greater accuracy is reported to reduce the incidence of positive resection margin and local recurrence rates compared with conventional endoanal LE.7 As with conventional LE, TEM causes less operative morbidity and mortality than radical resection. Technical details and outcomes are described below.

Current techniques of local excision Surgeons continue to evolve techniques for LE. Today, the posterior approaches have essentially been abandoned for treatment of rectal cancer, while the endoanal LE and TEM approaches have increased in popularity. Both endoanal LE and TEM are increasingly used in conjunction with chemoradiation in an attempt to improve the outcomes achieved by these surgical techniques when used as sole therapy for select rectal cancer. Zerz and colleagues described a new combined approach, the endoscopic posterior mesorectal resection (EPMR), designed to allow sampling of the mesorectal nodes while maintaining the advantages of low morbidity and negligible mortality of LE.7 More recently, Atallah and colleagues described an LE technique using a single-incision laparoscopic port.8 These techniques are briefly described below.

Endoanal Local Excision Endoanal LE, as described by Parks in 1968,9 is the most popular technique of LE. The patient is positioned depending on surgeon’s preference and the location of the tumour. An anterior tumour is best approached with the patient prone, whereas a posterior tumour, if distally located, may be better accessed with the patient in the lithotomy position. Mid- and proximal posteriorly based lesions are best accessed with the patient in prone jack-knife position. A limited distal colonic bowel preparation and preoperative antibiotics are generally used. Preoperative deep venous thrombosis prophylaxis is recommended. The procedure may be performed under general or regional anaesthesia. The use of a

HEBK001-C11_p171-190.indd 173

pudendal nerve block assists with sphincter relaxation and postoperative analgesia. A Lone Star (Lone Star Medical Products, Inc., Stafford, TX, USA) retractor is used to efface the anus and facilitate exposure of the distal rectum. Retractors of the surgeon’s choice are used to dilate the anus and expose the cancer. A fibre-optic headlight or a retractor with a light source is needed to adequately visualize the lesion within the rectum. After exposure is achieved, electrocautery is used to define a 1-cm circumferential margin around the tumour. Traction sutures may be placed in the normal tissue around the tumour to facilitate better exposure and manipulation of the lesion. Care should be taken not to traumatize the lesion or place sutures into the cancer. A full-thickness excision is performed with the dissection extending into the surrounding mesorectal fat. One must avoid injury to the vagina in female patients and the prostate in male patients. Once the specimen is fully excised, it should be orientated and pinned before sending it to pathology. After ensuring haemostasis, the defect is closed transversely to minimize risk of stricture. For large defects, a sleeve anastomosis with interrupted absorbable sutures can be used to advance the proximal rectum over the defect. Although it is not our practice, if the rectal opening is below the peritoneal reflection, it can be left open with no added morbidity.10 A proctoscopy examination is performed at the end of the procedure to ensure the rectum is widely patent. Postoperatively, antibiotics should be stopped within 24 h. Patients may take a normal diet on the evening of the operation. Postoperative analgesia requirements are minimal and patients may be discharged on the day of the operation. Most complications after endoanal LE are minor, with rates ranging from zero to 22 per cent.11 Bleeding, local sepsis, urinary tract infection or retention, rectovaginal fistula and pulmonary emboli are the most common complications. Lesions located more than 10 cm from the dentate line can be difficult to expose and excise in one piece with an adequate margin. If radical resection is indicated after LE, it is important to wait at least a month for the wound to contract and heal, otherwise the wound could dehisce during rectal mobilization, thus potentially spilling tumour cells or faeces into the operative site.

08/02/13 5:37 PM

174  Local excision and transanal endoscopic microsurgery

Transanal Endoscopic Microsurgery Transanal endoscopic microsurgery has gained increasing popularity and was developed to allow excision of more proximal lesions than those amenable to endoanal LE. The operation is performed through a 40-mm operating rectoscope (12 cm and 20 cm in length), which is insufflated with carbon dioxide (CO2) to obtain vision of the proximal rectum. Flow can be increased to a rate of 6 L/min and intra-rectal pressure is continuously monitored and maintained at 12–15 cmH2O. Specially designed instruments are introduced through the rectoscope’s working ports, similar to the technique used for laparoscopic surgery. Transanal endoscopic microsurgery offers endoscopic magnification and illumination that provide excellent visualization as well as multiple instruments that allow precise resection and secure suture closure. This is a particular advantage in patients with a large body habitus, for patients with a proximally based rectal cancer, and when performing excision of locally recurrent lesions. A limited distal colonic bowel preparation and preoperative antibiotics are generally used. Bowel preparation is of particular importance if penetration into the peritoneal cavity should inadvertently occur. Informed consent should include possible laparotomy for lesions at high risk for entering the peritoneum, specifically anterior tumours in female patients where the location of the anterior peritoneal reflection is unpredictable and may be quite distal. The rectoscope is inserted up to the lesion under direct vision with manual insufflation and then secured to the operating table. The bevel of the scope must face down at the lesion. The binocular eyepiece and the accessory scope are inserted. Adequate pneumorectum should be obtained without signs of an air leak. The entire lesion should be visible, although if not, one can still proceed as long as the entire lesion can be pulled down into the operative field. Excision of the lesion follows the same guidelines as standard endoanal LE. Partial-thickness or full-thickness excisions can be performed with this approach, depending on the location and nature of the lesion. In general, cancers should be removed by a full-thickness excision with a 1 cm margin. Most of the bleeding during full-thickness excisions occurs during the mesorectal dissection. In addition to conventional haemostasis with elec-

HEBK001-C11_p171-190.indd 174

trocautery, one can use a variety of commercially available products such as a harmonic scalpel that can simplify and speed up this step of the operation. One must also remember that the rectoscope should be repositioned several times during the dissection to keep the lesion in the centre of the operative field. If the peritoneum is violated, it should be repaired promptly, but this does not mandate immediate conversion to laparotomy. If pneumorectum is adequately maintained, one can proceed with excision. With regard to the suturing technique, sutures are started and finished with silver shots applied to the thread with a specially designed applicator because traditional knot-tying is very difficult in the rectum. Endoscopic suturing devices (e.g. Endo Stitch™) are also being used for closure of the rectal wall defect left after the lesion has been excised; these add significant cost to the procedure, however. Short sutures are preferable, and bisecting the wound with an interrupted suture may also facilitate closure. Compared with endoanal LE, CO2 distention greatly facilitates closure of the wound. Most patients are discharged home the same day or the day following the procedure. Essentially, any lesion, regardless of size, location and degree of circumferential involvement, can be removed with TEM, as long as the upper and lower margins of the lesion can be reached and completely visualized. After excision of lesions that involve the entire rectal circumference, intestinal continuity can be re-established with a hand-sewn end-to-end anastomosis performed through the rectoscope. Distal rectal lesions are difficult to excise with TEM because of the difficulty obtaining and maintaining pneumorectum and difficulty using the instruments in the distal rectum. Whereas exposure in laparoscopy can be facilitated by inserting additional ports or to insert the scope from a different angle, this is not possible with TEM. The instruments are inserted and manipulated only in parallel, and the scope is fixed in one position. Transanal endoscopic microsurgery is thus associated with longer operative times for such distal lesions compared with traditional endoanal LE. Disadvantages of TEM include costly equipment and slightly longer operating times. Although TEM was thought to have a steep learning curve, more recent reports have suggested a short learning curve for the procedure in surgeons with prior experience in minimally invasive laparoscopic surgery. As with

08/02/13 5:37 PM

Historical role and outcomes of local excision  175

endoanal LE, TEM is also a safe procedure, with a low rate of complications. Reported complications include entry into the peritoneal cavity, conversion to laparotomy, bleeding, pneumoscrotum, perineal pain, rectal stenosis, wound dehiscence, urinary dysfunction and rectovaginal fistula. Kreissler-Haag and colleagues assessed complications in 288 patients undergoing TEM. They reported that tumour localization higher than 8 cm, tumour diameter of more than 2 cm, and localization of the tumour on the lateral wall of the rectum were risk factors for surgical complications, mainly bleeding.12 Although there has been concern about anal sphincter injury from prolonged stretch, early studies documented no lasting adverse effect on anorectal function. Most individual TEM case series report mild transient incontinence, which usually resolves within weeks. Dafnis and colleagues reported impaired continence in 18 of 48 patients (37 per cent) at a median follow-up of 22 months.13 These patients did not report improvement in continence over time (pre- v. post-1 year assessment), which conflicts with other studies reporting significant improvement within weeks.14,15 Radiotherapy may be another factor affecting functional results after TEM, since radiation has negatively impacted function when used in conjunction with radical surgery.16 Data regarding functional outcomes after radiotherapy plus TEM are scant and not well defined. Patients should be aware that significant changes in continence may occur in the early postoperative period. Potential risk factors include pre-existing decreased continence or sphincter defects and large lesions requiring a more extensive resection and prolonged operating time. The majority of patients experience a return of pre-procedure continence at 6 weeks to 3 months. The long-term effects of an irradiated rectum after TEM remain to be defined.

Endoscopic Posterior Mesorectal Resection One significant disadvantage of LE or TEM is the lack of information regarding the lymph node status of the mesorectum. Endoscopic posterior mesorectal resection includes TEM or endoanal LE of selected, favourable T1 rectal cancers followed by a minimally invasive transperineal resection of the posterior part of the mesorectum, including all relevant lymphatic tissue and clip-

HEBK001-C11_p171-190.indd 175

ping of the superior rectal artery (Figure 11.1).7 Its proponents claim this technique provides complete tumour staging with minimal morbidity after LE of T1 rectal cancers.

Laparoscopic Local Excision Atallah and colleagues reported their experience with using a single-incision laparoscopic surgery port for access to the rectum, replacing the conventional operative rectoscope and using ordinary laparoscopic instruments to perform LE.8 The authors found this technique to be a safe and feasible alternative to TEM, providing its benefits at a fraction of the cost (Figure 11.2).

Historical role and outcomes of local excision In the late eighteenth and early nineteenth century, LE via a posterior approach to the rectum was the most common method of treatment of rectal cancer. Surgeons adopted this practice because operative mortality following anaesthesia and abdominal surgery was common. Their goals were simple: to resect the tumour and keep the patient from dying from the operation. During the first half of the twentieth century, mortality after elective colorectal cancer surgery fell steadily from 41 per cent in 1916–1920 to 7 per cent in 1946–1950.17 As anaesthesia and abdominal surgery became safer, surgeons embraced the Miles APE because of its superior oncological control compared with that achieved by LE. Abdominoperineal excision became the operation of choice for rectal and rectosigmoid cancers in most centres around the world. In the 1950s, anterior resection and colorectal anastomosis was shown to achieve oncological outcomes identical to APE for distal sigmoid and proximal rectal cancers without the morbidity of a permanent colostomy. By the 1970s, mortality following radical resection for rectal cancer had dropped to 5 per cent or less in most series. A reliable circular stapler that facilitated distal colorectal anastomosis was introduced in 1978. Soon, anterior resection became the procedure of choice for all but the most distal rectal cancers, as surgeons pushed the distal limits of sphincter-sparing proctectomy. The goals of surgery

08/02/13 5:37 PM

176  Local excision and transanal endoscopic microsurgery

Figure 11.1. Technique of endoscopic posterior mesorectal resection. (A) Trocar positions. (B) Access to the retrorectal space using the index finger. (C) Establishment of a sufficiently large operating space using a dissecting balloon trocar. (D) Dissection of the mesorectum from the posterior wall of the rectum.

for rectal cancer had expanded beyond safety and local oncological control to also include sphincter preservation and avoidance of permanent colostomy. Local excision of rectal cancer seemed relegated to use only in high-risk patients, in patients refusing radical resection and for palliation. A dissenting opinion came in 1977, when Morson and colleagues reported the St Marks experience with LE of rectal cancers using the endoanal approach of Sir Alan Parks.18 They reported equivalent oncological results after LE compared with radical resection but with significantly less morbidity and mortality. In the 1980s, reports of high rates of local recurrence and major operative morbidity, including anastomotic leak, anterior resection syndrome and genitourinary dysfunction, after low anterior resection and distal anastomosis caused physicians to re-examine the preferred approach to rectal cancer. Radiation therapists and medical oncologists argued for postoperative chemoradiation to enhance oncological outcomes. Heald and others held workshops to teach surgeons how to optimize surgical technique by conducting a total mesorec-

HEBK001-C11_p171-190.indd 176

tal excision (TME) as part of anterior resection or APE.19 Various technical modifications to anterior resection and APE are now included under the term TME. Proponents argued that the surgeon could achieve better oncological outcomes, increase sphincter preservation with coloanal anastomosis, improve anorectal function by using a colo-pouch, and maintain genitourinary function by using a meticulous pelvic dissection technique. The role of LE in treatment of rectal cancer varied widely from centre to centre in the 1980s. Some surgeons rarely used LE for any indication, while others, noting the significant morbidity and less than ideal oncological results following standard anterior resection and APE, increasingly used LE as an alternative in patients marginally unfit for radical resection. At the same time, new and improved imaging techniques, including endorectal ultrasonography (ERUS), magnetic resonance imaging (MRI) and computed tomography (CT), held the promise of being able to reliably and accurately stage rectal cancer, thus minimizing the risk that the use of LE would undertreat a patient with rectal cancer. Our

08/02/13 5:37 PM

Current role and outcomes of local excision  177

from zero to 28 per cent for T1 lesions and from 11 per cent to 45 per cent for T2 lesions, whereas 5-year survival rates ranged from 74 per cent to 90 per cent for T1 lesions and from 55 per cent to 75 per cent for T2 lesions. The wide variation in reported outcomes probably reflects differences in patient selection, intent of surgery and surgical technique. We also found that radical salvage surgery in patients with local recurrence after LE was not always curative.24 It seemed that LE did not equal the oncological outcomes historically achieved with standard anterior resection or APE for similarly staged cancers. As a result, the use of LE for curative-intent treatment of rectal cancer was curtailed in many centres, including our own.

Current role and outcomes of local excision

Figure 11.2. Transanal minimally invasive surgery (TAMIS). Single-port access is used to facilitate minimally invasive surgery so that transanal excision of rectal lesions can be performed. This is a novel approach that is a hybrid between single-port laparoscopy and transanal endoscopic microsurgery. Shown here is the SILS™ Port in position within the anal canal (A). Insufflation of the rectum is established with a dedicated port. Through the remaining three cannulas, two are used for laparoscopic instrumentation and one is used for a 5-mm camera (B).

experience with increased use of curative-intent LE for marginally unfit patients was consistent with Morson’s favourable report.18 As a result, we slowly began using LE for curative-intent treatment in normal-risk patients. We initially selected what we thought were only the most favourable, small T1 cancers for LE; when results seemed reasonable, we gradually liberalized our selection criteria and began including selected T2 cancers. We were dismayed when our longer-term follow-up showed high rates of local recurrence following LE of both T1 and T2 lesions.20,21 Soon, results from other large single institutions that had similarly liberalized the use of LE also showed higher than expected local recurrence rates; some studies showed a decrease in overall survival following LE.22,23 Local recurrence rates ranged

HEBK001-C11_p171-190.indd 177

The current role of LE for treating patients with rectal cancer is controversial. Consensus opinion as noted in practice guidelines and surgical textbooks is that the standard curative-intent operation of choice for normal-risk individuals with rectal cancer including stage I lesions is radical resection (anterior resection or APE) performed with the meticulous dissection and other techniques now encompassed in TME. This is supported by data such as those offered by Blumberg and colleagues.25 They reported that patients with stage I rectal cancer who undergo radical anterior resection with sphincter-sparing anastomosis or radical abdominoperineal excision of the rectum with colostomy generally benefit from a high cure rate, with 5-year survival rates around 90 per cent. Given this consensus, the appropriate role of LE in curative treatment of stage I rectal cancer would seem to be very limited. It is thus surprising that You and collaborators reported that LE had been used increasingly for stage I rectal cancer.26 One may speculate as to why LE has increased in use in the USA. It is possible that the increased enrolment in national screening programmes for colorectal cancer has led to diagnosis of a higher proportion of early-stage rectal cancers, including rectal cancers arising in sessile polyps, some of which may be potentially amenable to LE. Additionally, new neoadjuvant treatment regimens may lead to more effective down-staging of stage II and

08/02/13 5:37 PM

178  Local excision and transanal endoscopic microsurgery

stage III rectal cancers, potentially making some patients whose cancers demonstrate a complete or near-complete clinical response amenable to curative-intent LE.27–31 Local excision in combination with adjuvant or neoadjuvant chemoradiotherapy for early-stage rectal cancer is being assessed in randomized trials. Its use in compromise situations and for palliation is more difficult to assess but is likely increasing. With an ageing population, there may be an increase in the number of patients for whom local therapy is selected as a compromise treatment rather than selecting a conventional radical treatment protocol for an elderly patient with significant operative risk factors. The extent to which any of these possible explanations underlies the apparent increased use of LE is unclear. Regardless, it is obvious that the current role of LE in the treatment of rectal cancer is still being defined.

Curative-Intent Local Excision You and colleagues used the National Cancer Database to examine choice of surgical therapy over time for 35 179 patients diagnosed with stage I rectal cancer between 1989 and 2003.26 They found that LE increased from 26.6 per cent to 43.7 per cent for T1 cancers and from 5.8 per cent to 16.8 per cent for T2 lesions. This is a remarkable trend given the lack of level I or II evidence that LE is oncologically appropriate for curative-intent surgery of stage I rectal cancer. They further analysed 2124 patients with stage I rectal cancer diagnosed between 1994 and 1996 to assess perioperative outcomes, local recurrence and survival. Local excision was used in 765 patients (601 with T1 cancer, 164 with T2 cancer) and standard radical resection was done in 1359 patients (493 with T1 cancer, 866 with T2 cancer). Patients treated by LE experienced significantly lower 30-day morbidity than patients undergoing standard radical resection (5.6 per cent v. 14.6 per cent). The 5-year local recurrence rate after LE versus radical resection, however, was significantly higher (12.5 per cent v. 6.9 per cent for T1 cancers; 22.1 per cent v. 15.1 per cent for T2 cancers). The 5-year overall survival, however, was influenced by age and comorbidities but not by choice of surgery in both T1 cancers (77.4 per cent after LE v. 81.7 per cent after radical resection) and T2 cancers (67.6 per cent after LE v. 76.5 per cent after radi-

HEBK001-C11_p171-190.indd 178

cal resection). The authors concluded that there is an increased use of LE for stage I rectal cancer, but there are associated risks of local failure and the benefits of LE must be balanced against such risks. Winde and colleagues randomized 50 patients with T1 rectal cancer to either TEM or anterior resection.32 At a mean follow-up of 46 months, there was no significant difference in the local recurrence rate (4.2 per cent v. 0 per cent) or survival rate (96 per cent v. 96 per cent) between the two groups. Lezoche and colleagues randomized patients with T2N0 rectal cancer to TEM or laparoscopic standard radical resection with TME after neoadjuvant chemoradiation therapy. Local recurrence (5.7 per cent v. 2.8 per cent) and survival (94 per cent for both groups) were not significantly different after 84 months of follow-up.33 For proponents of LE, the apparent lack of a deleterious effect of choice of therapy on 5-year overall survival was reassuring and added credence to their opinion that LE should play a significant role in curative-intent treatment of select rectal cancers. Additional support for this view derives from the observation that radical surgery does not cure all patients with stage I disease and is associated with a 30-day postoperative mortality of 1.6–4.8 per cent, an often prolonged recovery and major ­morbidity.34 Anastomotic leak, pelvic sepsis, perineal wound problems, colostomy complications, and bowel, sexual and urinary functional disturbances may adversely affect the patient’s long-term quality of life. By contrast, LE of rectal cancer will confer undisputed benefits of negligible mortality, minimal morbidity with rapid recovery, and reasonable preservation of genitourinary and anal sphincter function. In a recent update of the Cancer and Leukemia Group B trial on local excision of distal rectal cancer (CALGB 8984), Greenberg and collaborators presented data supporting the concept that patients with selected T1 cancers treated by LE have comparable rates of local recurrence, overall survival and disease-free survival as historic controls treated by radical resection.35 As noted in the discussion about Greenberg and colleagues’ paper, the use of historic controls and other design issues make this a less than totally convincing argument to routinely use LE for selected stage I rectal cancers. The critical and still unresolved controversial issues are: does LE for selected stage I rectal cancers provide oncological outcomes equivalent to

08/02/13 5:37 PM

Current role and outcomes of local excision  179

that achieved by radical resection, and if not, do the advantages of LE outweigh the oncological disadvantages?

Curative-Intent Transanal Endoscopic Microsurgery Local recurrence after TEM has been reported mainly in single-institution reviews with limited numbers, which makes comparisons difficult. Recurrence rates after TEM range from zero to 13 per cent for patients with T1 tumours and from zero to 80 per cent for patients with T2 tumours. Bach and colleagues reported the outcomes of 424 patients with rectal cancer treated with TEM and entered into a national database.36 A positive resection margin occurred in 11 per cent, 22 per cent and 42 per cent of patients with T1, T2 and T3 tumours, respectively. Patients who were observed after TEM (with or without postoperative radiation) were 15 times more likely to develop local recurrence than those who were converted to standard radical resection (anterior resection or APE) with TME based on unfavourable histological findings in the TEM specimen. We analysed our results with TEM at the University of Minnesota in 95 patients with rectal carcinoma (58 T1, 26 T2, 11 T3).37 Follow-up time was 49.5 months. Local recurrence rates for 83 malignant tumours were 9.8 per cent for T1 tumours, 23.5 per cent for T2 tumours and 100 per cent for T3 tumours. Ganai and colleagues studied the outcomes of 28 patients with malignant rectal tumours (8 T in situ (Tis), 12 T1, 4 T2, 4 T3) who underwent TEM.38 Margin positivity occurred in 14 per cent of patients (4/28). Mean follow-up time was 46 months. Local recurrence occurred in 15 per cent of patients. Depth of invasion was related to the likelihood of invasive local recurrence, and lesions larger than 4 cm had a higher rate of recurrence. Median time to recurrence was 15 months. The authors recommended close endoscopic follow-up after TEM. Stipa and colleagues reviewed the outcomes of 69 patients with early rectal cancer who underwent TEM.39 The 5-year local recurrence rate was 8 per cent for Tis tumours, 8.6 per cent for T1 tumours and 9.5 per cent for T2 tumours. The median time to recurrence was 10.5 months. Of 14 patients who received preoperative chemoradiation (2 T1, 12 T2

HEBK001-C11_p171-190.indd 179

tumours), no local recurrences occurred, and only 1 patient with T2 disease had distant recurrence. All local recurrences were managed surgically. The 5-year disease-specific survival rate was 100 per cent for Tis tumours, 100 per cent for T1 tumours and 70 per cent for T2 tumours.

Comparisons of Curative-Intent Endoanal Local Excision and Transanal Endoscopic Microsurgery Although You and colleagues did not distinguish traditional endoanal LE from TEM,26 there is increasing literature comparing the two techniques. Unfortunately, most studies are retrospective singleinstitution experiences (Table 11.1). Moore and colleagues compared endoanal LE (n 5 89) and TEM (n 5 82) and showed no differences in complications.5 Transanal endoscopic microsurgery was more likely to yield clear margins and a non-fragmented specimen. At a mean follow-up of 37 months, recurrence was less frequent after TEM than after traditional LE (5 per cent v. 27 per cent). At our institution, we retrospectively compared endoanal LE (n 5 129) and TEM (n 5 42). We evaluated quality of resection, local recurrence and survival rates in patients with stage I rectal cancer.40 Although surgical margins were less often positive with TEM, the estimated 5-year disease-free survival rate was similar between the groups (TEM 84 per cent v. endoanal LE 76 per cent). We found that tumour distance from the anal verge, resection margin status, T stage and use of adjuvant therapy, but not surgical technique (LE or TEM), were independent predictors of local recurrence and disease-free survival.

Curative-Intent Endoscopic Posterior Mesorectal Resection Two years after the description of EPMR by Zerz and colleagues, Tarantino and colleagues reported their experience with this promising technique.44 They performed EPMR 6 weeks after LE in 18 consecutive patients with T1 tumours (13 endoanal LE, 5 TEM) and compared their results with historical controls that underwent low anterior resection with TME. Both groups included patients with low- and high-risk histology with no significant differences.

08/02/13 5:37 PM

77

The median number of lymph nodes resected was 7 for EPMR and 11 for low anterior resection. Of importance, an anatomical study found a mean of 8.4 lymph nodes in the mesorectum.45 At a median follow-up of 23 months, one patient developed distant metastases but there was no evidence of local recurrence in either group. They concluded that although the lymph node harvesting of EPMR was not comparable to that of TME, EPMR allowed for adequate lymph node staging and showed equivalent outcomes for T1 tumours. This technique may prove useful to accurately identify otherwise occult metastatic mesorectal nodes and thus facilitate appropriate selection of curative-intent therapy for apparently-early stage rectal adenocarcinoma.

4 64

Curative-Intent Local Excision as Part of Multimodality Therapy Dissatisfaction with oncological control offered by LE as the sole therapy for rectal cancer led some centres to use postoperative radiation with or without sensitizing chemotherapy. Advocates argued that by doing LE first, the exact depth of tumour invasion and histological features of the primary tumour could be assessed properly. Some centres chose to use adjuvant chemoradiation therapy for T1 cancers with unfavourable features or for T2 cancers. The hypothesis was that adjuvant therapy following curative-intent LE would treat occult local spread or occult lymph node metastases and thus improve oncological outcomes without subjecting the patient to the complications of radical surgery. As neoadjuvant chemoradiation therapy has been more accepted for use in advanced-stage rectal cancers before radical surgery, its use was suggested for treatment of small early-stage rectal cancers in combination with LE. Its role in this setting is now the subject of an ongoing national trial in the USA. Advocates argue that one can use local tumour response to estimate the likelihood of persistent nodal disease. Whether this is true is debated. NR, not reported.

80 64 De Graaf et al.

43

1

5

0

24

75

75

0

94

NS 0

0 35

2.8

3.7 56

17

18

94

NR 10

5.7 0

0

84

7.6 20 21–43

2

Langer et al.

Lezoche et al.33

1

14

96

42

Lee et al.

41

HEBK001-C11_p171-190.indd 180

35

93

9.4 83 95 19.5 22 2

survival (%)

0 0 31–35 1

52

4.1

0

4.1

100

48

(%)

22

Overall recurrence

(%) (%)

n (months) T stage

Follow-up

n

(%)

(%)

(%)

(%)

Mortality Complications survival recurrence Complications

Mortality

Overall Local

TME TEM Study

Table 11.1. Studies comparing transanal endoscopic microsurgery (TEM) versus total mesorectal excision (TME) for early rectal carcinoma.

Local

180  Local excision and transanal endoscopic microsurgery

Adjuvant therapy

A multi-institutional phase II prospective trial evaluating adjuvant chemoradiation therapy after LE divided patients into three groups based on the final histopathological findings.46 Group 1 patients had T1

08/02/13 5:37 PM

Current role and outcomes of local excision  181 Table 11.2. Studies assessing chemoradiotherapy before local excision of early rectal carcinoma. Study

T stage

Pathological complete

Recurrence, local/

response (%)

distant (%)

Survival (%)

T1

T2

26

73

4

92

Bonnen et al.

–

26

54

11

86

Lezoche et al.28,29

35

–

32

9

94

Callender et al.

–

47

49

21

79

Yeo et al.50

–

11

73

18

89

Garcia-Aguilar et al.51

77

–

43

NR

NR

Kim et al.27 48

49

NR, not reported.

lesions with favourable histology and negative margins and were observed without adjuvant therapy. Group 2 patients had T1 tumours with unfavourable histology or lesions greater than or equal to T2 and received the standard long-course dose of chemoradiation therapy. Group 3 patients were the same as group 2, except the resection margins were positive for malignancy, and therefore they received a higher dose of chemoradiation therapy. After a median follow-up of 6.1 years, local-regional recurrences developed in 16 per cent of the patients (8/65). The risk of recurrence correlated with T stage (T1, 4 per cent; T2, 8 per cent; T3, 23 per cent) and the degree of involvement of the resection margin. For the two groups treated with chemoradiation therapy, the 5-year actuarial freedom from pelvic relapse was 86 per cent, comparing favourably with historical controls treated with TME and suggesting a beneficial effect of chemoradiation therapy. The Cancer and Leukemia Group B conducted a similar prospective multi-institutional trial comparing the outcomes of 59 patients with T1 tumours treated with LE alone and 51 patients with T2 lesions treated with LE and postoperative chemoradiation therapy.35 At a median follow-up of 7.1 years, the local recurrence rates were 8 per cent for T1 lesions and 18 per cent for T2 cancers. The 10-year actuarial overall survival and diseasefree survival rates were 84 per cent and 75 per cent for T1 cancers and 66 per cent and 64 per cent for T2 cancers, respectively. Neoadjuvant therapy

Although most studies evaluating neoadjuvant therapy have small numbers and are retrospective in nature, they do suggest that preoperative chemoradiation therapy can indeed down-stage tumours subjected to subsequent LE and also lower recur-

HEBK001-C11_p171-190.indd 181

rence rates compared with LE alone (Table 11.2). One of the first reports on preoperative radiotherapy was by Mohiuddin and colleagues.47 Thirty patients with T3 tumours were treated with preoperative radiotherapy followed by LE. They reported a pathological complete response in 11 patients (36 per cent), a 5-year recurrence rate of 10 per cent and a survival rate of 83 per cent. Bonnen and colleagues compared the outcomes of LE and radical resection with TME after preoperative chemoradiation therapy.48 Of 26 patients, 54 per cent (n 5 11) had a pathological complete response and 35 per cent had microscopic residual disease. The 5-year local recurrence rate in the patients undergoing LE was 6 per cent and overall survival was 86 per cent, compared with 8 per cent and 81 per cent, respectively, for the TME group. The results of this study were updated with a median follow-up of 63 months.49 Ten-year actuarial local recurrence was not significantly different between the LE and TME groups (10.6 per cent v. 7.6 per cent), and no significant difference in survival was found between the groups. Kim and collaborators at the Lee Moffitt Cancer Center retrospectively reviewed the outcomes of 26 patients who received neoadjuvant therapy followed by LE for T2 and T3 rectal cancers.27 Pretreatment ERUS staging included 5 T2N0, 13 T3N0, 7 T3N1 and 1 not done. Pathological complete and partial responses were achieved in 9 of 26 (35 per cent) and 17 of 26 (65 per cent) patients, respectively. At a mean follow-up of 24 months, only one tumour with partial response, in a patient who refused APE, had recurred. None of the tumours with complete pathological response recurred. Nair and colleagues reported a recurrence rate of 16 per cent after neoadjuvant chemoradiation therapy and endoanal LE for T2 and T3 rectal cancers.52 The overall 5-year survival was 84 per cent in

08/02/13 5:37 PM

182  Local excision and transanal endoscopic microsurgery

node-negative patients. The results of these studies contrast with other series that have shown higher rates of lymph node positivity after TME as well as lower responses to chemoradiation therapy. These differences probably reflect selection bias, with patients having more favourable tumours likely undergoing LE. In a prospective randomized trial, Lezoche and colleagues compared TEM and TME after neoadjuvant chemoradiation therapy in patients with T2N0 rectal cancer.28,29 The overall pathological complete response rate was approximately 35 per cent in both groups. In an update of this trial with a median follow-up of 84 months, 2 of 35 patients (5.7 per cent) had local recurrence in the TEM group and only 1 of 35 patients (2.8 per cent) had recurrence in the TME group. Diseasefree survival after 84 months of median follow-up was similar (94 per cent) in both groups. Overall, there seems to be a trend towards lower recurrence rates and higher disease-free survival with neoadjuvant or adjuvant therapy compared with LE alone. This may be more beneficial in highrisk histology tumours and T2 or T3 tumours. For patients with a pathological complete response documented on the surgical specimen, the reported rate of positive nodes in the mesorectum ranges from zero to 12 per cent. Therefore, LE may not guarantee a curative resection, even in patients with ypT0 tumours after chemoradiotherapy and LE. What needs to be defined is the risk of nodal metastasis in patients with clinically N0 tumours who become ypT0 after chemoradiation. Another question that remains unanswered is the efficacy of neoadjuvant therapy in T2N0 rectal adenocarcinoma followed by LE. This is currently being addressed by the American College of Surgeons Oncology Group trial Z6041, a multicentre phase II clinical trial using neoadjuvant chemoradiation therapy followed by LE in patients with ultrasound-staged T2N0 rectal cancer. Preliminary results presented as an abstract reported a 98 per cent resection rate with negative margins.51 Down-staging occurred in 49 of 77 patients (64 per cent), with a pathological complete response rate of 43 per cent.

Local Excision for Compromise and Palliation Local excision has long been used for patients judged to have comorbidities that put them at high

HEBK001-C11_p171-190.indd 182

operative risk for radical surgery. For the most part, even patients with major illnesses can tolerate LE, providing the rectal cancer does not exceed the technical limits of the operation. Similarly, when patients refuse a recommended radical surgery, usually because of the refusal to accept a permanent colostomy, the surgeon may consider LE as an alternative therapy. This can pose an ethical dilemma for the surgeon, since patients often cannot comprehend that use of LE, even for relatively early-stage rectal cancer as a compromise indication, may result in incurable painful local recurrence that often requires further palliative treatment. Local excision for palliation is used selectively, but the cancer cannot exceed the technical limits of the operation.

Patient and cancer selection for local excision Criteria for selecting patients with rectal cancer who would benefit from LE vary depending on the intent of therapy, the patient’s desires, the treatment philosophy and techniques offered by the treating physicians, and, most importantly, the stage and characteristics of the primary cancer. That LE can cure some rectal cancers is not in dispute, but the question is whether we have the ability to accurately distinguish patients whose rectal cancer is almost certain to be cured by LE from patients whose rectal cancer is unlikely to be cured by LE. There are risk factors that can be identified during pretreatment evaluation that are contraindications to curative intent LE, but there is no widely accepted evidencebased list of criteria that surgeons can use to select rectal cancers for curative intent LE. Clinical, pathological and imaging assessments of the primary rectal cancer are essential to determine whether LE is even to be considered as a viable option for curative intent treatment.

Assessing the Primary Rectal Cancer for Local Excision Not all rectal cancers can or should be resected by LE. The surgeon must assess the primary rectal cancer to determine whether LE is technically possible and whether doing so is likely to achieve the therapeutic goals. It is important that the surgeon

08/02/13 5:37 PM

Patient and cancer selection for local excision  183

recognizes when not to attempt LE even for palliation. Most surgeons agree that LE should be considered only if the cancer is in an accessible part of the rectum and small enough to be totally excised in one piece with a 1 cm margin of normal rectum. Additionally, most surgeons would agree that ideally the cancer is confined to the rectal wall (T1 or T2). Sometimes exceptions are made in cases of compromised intent therapy or palliative intent therapy. Clinical, pathological and imaging assessments are key elements to achieving pretreatment staging of the rectal primary tumour. Tumour size, location and morphology

Careful digital rectal examination with proctosigmoidoscopy to palpate and visualize the rectal cancer can determine the lesion’s size, precise level and location in relation to the anal sphincters and other structures, and the presence of gross features associated with a poor prognosis. Although tumour size is not a reliable predictor of depth of invasion or nodal metastasis, patients with tumours larger than 3–4 cm in diameter or involving more than 40 per cent of the circumference of the rectum are poor candidates for traditional endoanal LE, simply because adequate exposure is difficult to obtain with conventional retractors.53 Proponents of TEM report that this approach provides better exposure and visualization, such that the size of the lesion is less of an issue than with endoanal LE. Although tumour distance from the anal verge does not have independent prognostic value when stratified by tumour stage,54,55 it is an important consideration if LE is being considered. Traditional endoanal LE provides reliable access to distal and most mid-rectal lesions, while TEM provides access to mid- and proximal lesions but is difficult to use for the most distal cancers. An additional consideration for the surgeon is whether the location of the cancer allows full-thickness excision of the rectal wall without entering the peritoneal cavity. Distal and posteriorly based mid-rectal cancers are extraperitoneal and generally can be excised safely without much risk of peritonitis if the rectal wall defect fails to heal. Proximal lesions and anteriorly based mid-rectal cancers are likely to be intraperitoneal, and their full-thickness excision may increase the risk of peritonitis if a leak occurs and risk other organ injury to adjacent structures such as the vagina or prostate. Transanal endoscopic microsurgery reputedly

HEBK001-C11_p171-190.indd 183

offers the ability to more securely suture repair the rectal wall defect for proximal lesions than is possible with endoanal LE. Although tumour morphology does not have independent prognostic value when stratified by tumour stage,54,55 it is often used as a surrogate for depth of invasion, since most ulcerated lesions are T3 or T4. Thus, many surgeons avoid LE for ulcerated cancers. Histopathology

Biopsy of the primary lesion may reveal microscopic features associated with lymph node metastases, local recurrences, and a poor prognosis such as lymphovascular and perineural invasion or mucinous, signet ring or poorly differentiated histology. If these are present on pretreatment biopsy, LE is generally contraindicated. The rates of local recurrence and lymph node involvement in T1, grade 1, well-differentiated tumours are 0–3 per cent, whereas T1, grade 3 poorly differentiated tumours have 12 per cent nodal involvement and a 10–63 per cent chance of local recurrence.12,56 Ulcerated mucinous cancers and lesions with evidence of perineural and lymphovascular invasion are associated with local recurrence rates as high as 25 per cent.12,56 Chemoradiotherapy regimens are less effective in tumours with undifferentiated histology and lymphovascular invasion.57 Tumour budding, defined as isolated cancer cells or nests of cancer cells in normal tissue at the edge of the main tumour, has been associated with lymph node metastases in up to 25 per cent of patients with T1 cancers and has been identified as a predictor of worse survival independent of disease stage.58,59 The literature is unclear as to how often histopathology of the final specimen obtained by LE results in identification of one of these unfavourable features not noted on pretreatment biopsy. Depth of invasion, nodal metastases and local spread

Nodal metastases, direct intrapelvic spread and mesorectal spread are contraindications for curative intent LE in average-risk patients and relative contraindications for LE in other settings (palliation or compromise). All of the risk factors discussed above are surrogate markers for nodal metastases or local spread, but the most important correlate is depth of invasion of the primary cancer (Table 11.3). At a minimum, tumours that are considered for curative intent LE

08/02/13 5:37 PM

184  Local excision and transanal endoscopic microsurgery Table 11.3. Morphologic features of favourable and unfavourable T1 rectal cancers.

Differentiation Submucosal

Favourable/

Unfavourable/

low risk

high risk

Well differentiated

Poorly differentiated

(G1–G2)

(G3)

1

2–3

No

Yes

  invasion Lymphovascular   invasion Size

,3 cm

.3 cm

Wall circumference

,40%

.40%

should be localized to the rectal wall (T1 and T2). Nodal involvement occurs in 0–12 per cent of T1 tumours and 12–28 per cent of T2 tumours.56,60 Kudo first introduced the concept of dividing T1 cancers into three levels based on the extent of submucosal invasion, varying from the superficial third to the middle third to the deepest third (sm1, sm2, sm3).61 Several studies have demonstrated an association of deeper submucosal invasion with increasing risk of lymph node metastases in 0–3 per cent for sm1 lesions to 20–23 per cent for sm3 lesions,53,62 with increased rates of local recurrence for sm3 lesions.63 Additionally, gender may be a predictive marker for lymph node metastasis in early rectal cancer. Kobayashi and colleagues showed that 1 per cent of male patients with well-differentiated T1 tumours of the lower rectum had lymph node metastasis, compared with 30 per cent in female patients with histological types other than well-differentiated adenocarcinoma, even when the tumour did not invade the muscularis propria.64 The authors suggested that these patients should not undergo LE. Serum carcinoembryonic antigen may be useful as a baseline study. A CEA level that is elevated and drops to normal after resection is reassuring. Persistence of an elevated CEA level after treatment or a rising CEA level should alert the surgeon to the possible presence of metastatic disease or incomplete resection. Perez and colleagues showed that a CEA level below 5 ng/mL after chemoradiotherapy is a favourable prognostic factor for rectal cancer and is associated with increased rates of earlier disease staging and complete tumour regression.65 Conversely, an elevated pre-chemoradiotherapy serum CEA level (. 5 ng/mL) is associated with poor tumour response to chemoradiotherapy.66

HEBK001-C11_p171-190.indd 184

Imaging studies used to stage rectal cancer include CT, MRI and ERUS. New-generation CT scanning and surface-coil MRI are able to detect tumour invasion outside the rectal wall,67 but they cannot reliably distinguish between a T1 and a T2 lesion or between a benign lesion and a T1 lesion. Endorectal ultrasonography and endorectal coil or 3-tesla MRI can provide this level of detail, and either of these studies should be performed in the evaluation of any patient considered for LE. Although MRI is more expensive and less available than ERUS, it does not suffer from the interobserver variation that plagues endosonography, and it is the optimal way to assess mesorectal invasion and determine whether the circumferential radial margin is involved, threatened or free of tumour. In a meta-analysis including data from 90 publications, Bipat and colleagues found the sensitivity of ERUS and MRI for tumour invasion outside the rectal wall was as high as 90% and 82%, respectively.68 The sensitivity for lymph node involvement was significantly lower, however, at 67 per cent and 66 per cent, respectively. In a systematic review of 53 studies including 2915 patients, the accuracy of ERUS was 87 per cent for T stage and 74 per cent for lymph node involvement;69 for MRI with endorectal coil, the corresponding numbers were 84 per cent and 82 per cent. Data have shown that three-dimensional reconstruction increases the accuracy of ERUS in assessing the depth of rectal wall and submucosal invasion and may help in selecting patients for local excision.70 Because the gold standard for assessing the accuracy of any imaging study should be the final nonirradiated pathology specimen, many studies are biased by selective inclusion and exclusion criteria. Patients with advanced disease who undergo neoadjuvant therapy are generally excluded, and patients with clearly localized lesions who are offered LE are similarly often excluded. Thus, the patients left for evaluation are those with more equivocal findings on imaging studies. This selection bias has likely led to an underestimation of the accuracy of ERUS and MRI.

Assessing the patient with rectal cancer The goals of the pretreatment evaluation are to identify conditions and issues that will affect the

08/02/13 5:37 PM

Follow-up and salvage therapy  185

choice of therapy. In addition to carefully staging the primary rectal cancer as noted above, distant metastases, especially to the liver and lungs, and synchronous gastrointestinal or colonic abnormalities, diseases or prior surgery that may complicate the treatment plan must be assessed. Pre-existing comorbidities, baseline bowel function including continence and defecation and operative risks can largely be defined by a thorough history and physical examination, with additional laboratory or other testing as needed. The surgeon must understand the patient’s desired outcome and assess their psychosocial state, decision-making capability, and availability of family or others for support during the stressful periods likely to arise as the rectal cancer is treated. A complete review of the pretreatment assessment is beyond the scope of this chapter but is available elsewhere.71

Optimal therapy selection: role of a multidisciplinary team Despite new and improved imaging, pretreatment staging of rectal cancer has inherent significant inaccuracies, especially in detecting lymph node metastases and making subtle distinctions between T stages. Thus, what is thought to be a stage I rectal cancer may actually be a stage II or stage III rectal cancer. This means that the decision to use endoanal LE or TEM, techniques characterized by ‘total mesorectal neglect’, for an alleged stage I rectal cancer rather than performing a standard radical resection with TME may unwittingly compromise oncological outcomes. Since nodal metastases occur in 0–12 per cent of T1 cancers and 12–28 per cent of T2 lesions,56,60 there is a possibility of undetected nodal involvement, even in T1 tumours. Inaccuracies in staging can result in undertreatment of occult cancer spread to the mesorectum, mesenteric lymph nodes or other sites. Such undertreatment will inevitably result in increased local recurrences and decreased survival that likely would not have occurred had standard radical resection been used. Whether the technique of EPMR can overcome this shortcoming in staging lymph nodes accurately remains to be seen.9,44 These facts simply reflect the state of current medical knowledge and cannot be used as an

HEBK001-C11_p171-190.indd 185

excuse to avoid a rigorous evidence-based decision-­ making process for each patient presenting with rectal cancer. This admonition is especially relevant to the controversies regarding choice of LE versus radical resection for rectal cancer. Although the issues regarding appropriate choice of curative intent LE to avoid over- and undertreatment are not resolved, pretreatment review by a multidisciplinary team of rectal cancer experts from different disciplines (colorectal surgery, medical oncology, radiation oncology, pathology, diagnostic radiology) is recommended to minimize the risk of personal bias or incorrect interpretation of imaging, pathology and other relevant data (Figure 11.3). In addition, this approach couples the pretreatment assessment with the skill, experience and judgement of the treating physicians who have knowledge of locally available treatment modalities and outcomes achieved in the past as well as the patient’s specific situation. Intent of therapy (curative or palliative) and the occasional need to compromise a preferred treatment for a patient whose comorbidities create a prohibitive operative risk for radical surgery can be clarified. A final recommendation for the patient’s treatment can be made, informed consent can be obtained and a treatment protocol implemented.

Follow-up and salvage therapy The first decision to make following curative intent LE of a rectal cancer is whether the intended goal is likely to have been achieved. Borschitz and colleagues evaluated and compared oncological outcomes of patients undergoing immediate reoperation versus salvage surgery after LE in the setting of unsuspected T2 disease or unfavourable ­histology.30,31 At a median follow-up of 10 years, patients who underwent salvage surgery had significantly higher recurrence rates (37 per cent v. 8 per cent) and significantly lower tumour-free survival (54 per cent v. 86 per cent). The median time to local recurrence for patients undergoing salvage surgery was 12 months. Lee and colleagues also reviewed the outcomes of 36 patients with unfavourable histology and positive resection margins who underwent TEM.72 Of the 36 patients, 12 underwent salvage surgery and 24 either refused radical resection or had poor functional status and

08/02/13 5:37 PM

186  Local excision and transanal endoscopic microsurgery

CT, computed tomography; ERUS, endorectal ultrasound; LE, local excision; MRI, magnetic resonance imaging; SM1, superficial (third) submucosal invasion; TAE, transanal excision; TEM, transanal endoscopic microsurgery; TME, total mesorectal excision. Figure 11.3. Multimodality treatment algorithm for early rectal carcinoma.

would not tolerate a major operation. Of the 12 patients who underwent salvage surgery, 1 had systemic recurrence. Of the 24 patients who did not have radical surgery, 5 had recurrence (3 local, 2 distant). Thereafter, many have supported the role of immediate salvage surgery after TEM in patients with unfavourable histology, positive excision margins and a threatened resection margin (, 1 mm), reducing the recurrence rate to 6 per cent.73 Based on such data, most surgeons agree that if final pathology after LE or TEM does not confirm a rectal cancer with the anticipated favourable features and with clear margins (i.e. deeper invasion than anticipated, unfavourable histology features or a positive margin), then further treatment is indicated, by radical resection or chemoradiotherapy, or both. A potential concern when LE has to be converted to radical resection with TME is that of a violated mesorectal plane, which may result in a suboptimal TME. Although recurrences after TME

HEBK001-C11_p171-190.indd 186

reoperation have been reported, they are infrequent. We have found that radical reoperation a few weeks following LE has been safe. There is no agreed best practice protocol to follow patients long term after apparent successful LE of rectal cancer. Since local recurrence is a major concern following LE of rectal cancer, it would seem reasonable to regularly perform digital rectal, proctoscopic and ERUS examinations. Unfortunately, even with careful follow-up, it is difficult to detect residual disease as a curable local recurrence. Many tumours recur with unresectable local tumours or incurable distant metastases. Patients who are able to undergo curative intent therapy often require multimodality therapy with chemoradiotherapy and extensive resections. Five-year survival rates range from 50 per cent to 58 per cent after salvage surgery and are well below what is expected after initial radical surgery with TME. At the University of Minnesota, we followed patients after LE with

08/02/13 5:37 PM

CONCLUSION  187

digital rectal examination, proctoscopy and ERUS every 4 months. Despite this intensive surveillance, 27 of the 29 patients (93 per cent) we identified with recurrence after LE presented with advanced disease (stage II or higher).24 Curative resection (R0) was possible in only 79 per cent of patients. Of note, 81 per cent of the patients were asymptomatic at the time of their diagnosis.

Surgeon’s philosophy Given the controversies and evolving data regarding the use of curative intent LE or TEM, it is not surprising that colorectal surgeons have widely varying philosophies about their roles in normal-risk individuals with rectal cancer. Some surgeons, discouraged by our current inability to accurately stage the T and N stage of rectal cancer in the pretreatment evaluation and chastened by the data revealing compromised oncological outcomes, totally reject endoanal LE or TEM as a viable option for all but compromised or palliative situations because of the fear of undertreating a potentially curable rectal cancer. They argue that radical surgery with TME is the treatment of choice for all rectal cancers, including stage I lesions. Of the surgeons who do use curative intent LE or TEM, most attempt to restrict their use at least as sole therapy to highly selected, low-risk, small, T1 N0 M0 rectal cancers. They argue that despite the inaccuracies in our best current imaging studies, they can usually avoid the significant risk of lymph node metastases in T2 cancers and the significant morbidity of radical surgery. Other surgeons attempt to be even more restrictive by limiting curative intent LE or TEM to T1 cancers with sm1 invasion and otherwise favourable features. The problem is that this approach assumes, incorrectly, that we have reliable ways to identify precise levels of T stage and can accurately assess lymph node metastases before treatment. These surgeons often view the initial LE or TEM as a first step in the treatment of such lesions. If final pathology after LE or TEM does not confirm a rectal cancer with the anticipated favourable features and with clear margins, they then proceed with further treatment by radical resection or chemoradiotherapy, or both. Whether this approach compromises survivorship that may have been achieved had more aggressive therapy been used initially is debated.

HEBK001-C11_p171-190.indd 187

Strong advocates for the expanded use of LE or TEM for more advanced lesions emphasize the small but significant mortality, the high morbidity, the long-term functional disabilities and the slow recovery associated with traditional radical resection. By contrast, they point to the negligible mortality, low morbidity, generally excellent function and prompt return to normal life activities after LE or TEM. They note also that many studies have shown no significant difference in survival and that new techniques, including the minimally invasive sampling of mesorectal nodes and combination therapy with chemoradiotherapy, may offer expanded opportunities and improved oncological outcomes for these approaches. Endoanal LE or TEM may also be offered to patients who are medically unfit for a major operation or who refuse radical resection or a colostomy, with the option of pre- or postoperative chemoradiotherapy. In the setting of palliative excision, however, most primary tumours are bulky or stenotic and may not be amenable to LE.

Conclusion The ideal therapy for rectal cancer cures the patient of the primary lesion and any distant metastases without treatment-related mortality or major morbidity, while preserving pretreatment bowel, sexual and urinary function and allowing prompt return to a high quality of life. Over- and undertreatment are avoided. Unfortunately, the ideal is often not achievable, in part because there are numerous other factors still to be elucidated, such as tumour biology, tumour–host interactions and genetics, which if known may influence our decisions and, in part, because the information on which we base our decisions is less than complete and less than totally accurate. The appropriate role of endoanal LE or TEM, or newer modifications of these techniques, for treatment of rectal cancer is highly controversial. Although radical resection with TME continues to be the standard operation for most patients with rectal cancer, endoanal LE and TEM are acceptable alternatives, with significantly less morbidity. Most surgeons restrict their curative intent use to selected patients with T1 disease or to patients unfit for radical resection. Compared with endoanal LE, TEM offers

08/02/13 5:37 PM

188  Local excision and transanal endoscopic microsurgery

a higher likelihood of achieving clear resection margins, lower recurrence rates and the ability to successfully excise more proximal tumours. Although the number of studies comparing endoanal LE or TEM to radical resection with TME is small, the results are similar, in the sense that both show a higher recurrence rate, a trend towards decreased survival, and lower morbidity with LE techniques compared with radical resection. Significant disease progression can occur after LE despite intense surveillance, which may preclude curative salvage. The role of chemoradiation therapy and LE techniques in the treatment of rectal cancer is still under study.

References   1. Kraske P. Zur exstirpation hoch sitzender mastdarmkrebske. Verh Dtsch Ges Chir 1885; 14: 464–74.   2. Mason AY. Surgical access to the rectum: a transsphincteric exposure. Proc R Soc Med 1970; 63(suppl.): 91–4.   3. Lockhart-Mummery J. Two hundred cases of cancer of the rectum treated by perineal excision. Br J Surg 1926; 14: 110–24.   4. Miles WE. A method of performing abdomino-perineal excision for carcinoma of the rectum and of the terminal portion of the pelvic colon. Lancet 1908; 2: 1812–3.   5. Moore JS, Cataldo PA, Osler T, Hyman NH. Transanal endoscopic microsurgery is more effective than traditional transanal excision for resection of rectal masses. Dis Colon Rectum 2008; 51: 1026–30, 1030–1.   6. Buess G, Hutterer F, Theiss J, et al. A system for a transanal endoscopic rectum operation. Chirurg 1984; 55: 677–80.   7. Zerz A, Muller-Stich BP, Beck J, et al. Endoscopic posterior mesorectal resection after transanal local excision of T1 carcinomas of the lower third of the rectum. Dis Colon Rectum 2006; 49: 919–24.   8. Atallah S, Albert M, Larach S. Transanal minimally invasive surgery: a giant leap forward. Surg Endosc 2010; 24: 2200–05.   9. Parks AG. A technique for excising extensive villous papillomatous change in the lower rectum. Proc R Soc Med 1968; 61: 441–2. 10. Strong SA, Lavery IC, Oakley JR, Fazio VW, Church JM. Transanal excision of rectal tumors: is the case closed? Dis Colon Rectum 1992; 35: P25. 11. Sengupta S, Tjandra JJ. Local excision of rectal cancer: what is the evidence? Dis Colon Rectum 2001; 44: 1345–61. 12. Kreissler-Haag D, Schuld J, Lindemann W, et al. Complications after transanal endoscopic microsurgical resection correlate with location of rectal neoplasms. Surg Endosc 2008; 22: 612–6.

HEBK001-C11_p171-190.indd 188

13. Dafnis G, Pahlman L, Raab Y, et al. Transanal endoscopic microsurgery: clinical and functional results. Colorectal Dis 2004; 6: 336–42. 14. Cataldo PA, O’Brien S, Osler T. Transanal endoscopic microsurgery: a prospective evaluation of functional results. Dis Colon Rectum 2005; 48: 1366–71. 15. Herman RM, Richter P, Walega P, Popiela T. Anorectal sphincter function and rectal barostat study in patients following transanal endoscopic microsurgery. Int J Colorectal Dis 2001; 16: 370–6. 16. Peeters KC, van de Velde CJ, Leer JW, et al. Late side effects of short-course preoperative radiotherapy combined with total mesorectal excision for rectal cancer: increased bowel dysfunction in irradiated patients: a Dutch colorectal cancer group study. J Clin Oncol 2005; 23: 6199–206. 17. Grinnell RS. Results in the treatment of carcinoma of the colon and rectum: an analysis of 2,341 cases over a 35 year period with 5 year survival results in 1,667 patients. Surg Gynecol Obstet 1953; 96: 31–42. 18. Morson BC, Bussey HJ, Samoorian S. Policy of local excision for early cancer of the colorectum. Gut 1977; 18: 1045–50. 19. Heald RJ, Ryall RD. Recurrence and survival after total mesorectal excision for rectal cancer. Lancet 1986; 1: 1479–82. 20. Garcia-Aguilar J, Mellgren A, Sirivongs P, et al. Local excision of rectal cancer without adjuvant therapy: a word of caution. Ann Surg 2000; 231: 345–51. 21. Mellgren A, Sirivongs P, Rothenberger DA, et al. Is local excision adequate therapy for early rectal cancer? Dis Colon Rectum 2000; 43: 1064–71. 22. Paty PB, Nash GM, Baron P, et al. Long-term results of local excision for rectal cancer. Ann Surg 2002; 236: 522–30. 23. Madbouly KM, Remzi FH, Erkek BA, et al. Recurrence after transanal excision of T1 rectal cancer: should we be concerned? Dis Colon Rectum 2005; 48: 711–9. 24. Friel CM, Cromwell JW, Marra C, et al. Salvage radical surgery after failed local excision for early rectal cancer. Dis Colon Rectum 2002; 45: 875–9. 25. Blumberg D, Paty PB, Picon AI, et al. Stage I rectal cancer: identification of high risk patients. J Am Coll Surg 1998; 186: 574–9. 26. You YN, Baxter NN, Stewart A, et al. Is the increasing rate of local excision for stage I rectal cancer in the United States justified? A nationwide cohort study from the National Cancer Database. Ann Surg 2007; 245: 726–33. 27. Kim CJ, Yeatman TJ, Coppola D, et al. Local excision of T2 and T3 rectal cancers after downstaging chemoradiation. Ann Surg 2001; 234: 352–8. 28. Lezoche E, Guerrieri M, Paganini AM, et al. Long-term results in patients with T2–3 N0 distal rectal cancer undergoing radiotherapy before transanal endoscopic microsurgery. Br J Surg 2005; 92: 1546–52.

08/02/13 5:37 PM

REFERENCES  189 29. Lezoche E, Guerrieri M, Paganini AM, et al. Transanal endoscopic versus total mesorectal laparoscopic resections of T2-N0 low rectal cancers after neoadjuvant treatment: a prospective randomized trial with a 3-years minimum follow-up period. Surg Endosc 2005; 19: 751–6. 30. Borschitz T, Gockel I, Kiesslich R, Junginger T. Oncological outcome after local excision of rectal carcinomas. Ann Surg Oncol 2008; 15: 3101–8. 31. Borschitz T, Wachtlin D, Mohler M, et al. Neoadjuvant chemoradiation and local excision for T2–3 rectal cancer. Ann Surg Oncol 2008; 15: 712–20. 32. Winde G, Nottberg H, Keller R, et al. Surgical cure for early rectal carcinomas (T1): transanal endoscopic microsurgery vs. anterior resection. Dis Colon Rectum 1996; 39: 969–76. 33. Lezoche G, Baldarelli M, Guerrieri M, et al. A prospective randomized study with a 5-year minimum followup evaluation of transanal endoscopic microsurgery versus laparoscopic total mesorectal excision after neoadjuvant therapy. Surg Endosc 2008; 22: 352–8. 34. Hodgson DC, Zhang W, Zaslavsky AM, et al. Relation of hospital volume to colostomy rates and survival for patients with rectal cancer. J Natl Cancer Inst 2003; 95: 708–16. 35. Greenberg JA, Shibata D, Herndon JE 2nd, et al. Local excision of distal rectal cancer: an update of cancer and leukemia group B 8984. Dis Colon Rectum 2008; 51: 1185–91, 1191–4. 36. Bach SP, Hill J, Monson JR, et al. A predictive model for local recurrence after transanal endoscopic microsurgery for rectal cancer. Br J Surg 2009; 96: 280–90. 37. Tsai BM, Finne CO, Nordenstam JF, et al. Transanal endoscopic microsurgery resection of rectal tumors: outcomes and recommendations. Dis Colon Rectum 2010; 53: 16–23. 38. Ganai S, Kanumuri P, Rao RS, Alexander AI. Local recurrence after transanal endoscopic microsurgery for rectal polyps and early cancers. Ann Surg Oncol 2006; 13: 547–56. 39. Stipa F, Burza A, Lucandri G, et al. Outcomes for early rectal cancer managed with transanal endoscopic microsurgery: a 5-year follow-up study. Surg Endosc 2006; 20: 541–5. 40. Christoforidis D, Cho HM, Dixon MR, et al. Transanal endoscopic microsurgery versus conventional transanal excision for patients with early rectal cancer. Ann Surg 2009; 249: 776–82. 41. Lee W, Lee D, Choi S, et al. Transanal endoscopic microsurgery and radical surgery for T1 and T2 rectal cancer. Surg Endosc 2003; 17: 1283–7. 42. Langer C, Liersch T, Suss M, et al. Surgical cure for early rectal carcinoma and large adenoma: transanal endoscopic microsurgery (using ultrasound or electrosurgery) compared to conventional local and radical resection. Int J Colorectal Dis 2003; 18: 222–9.

HEBK001-C11_p171-190.indd 189

43. De Graaf EJ, Doornebosch PG, Tetteroo GW, et al. Transanal endoscopic microsurgery is feasible for adenomas throughout the entire rectum: a prospective study. Dis Colon Rectum 2009; 52: 1107–13. 44. Tarantino I, Hetzer FH, Warschkow R, et al. Local excision and endoscopic posterior mesorectal resection versus low anterior resection in T1 rectal cancer. Br J Surg 2008; 95: 375–80. 45. Canessa CE, Badia F, Fierro S, et al. Anatomic study of the lymph nodes of the mesorectum. Dis Colon Rectum 2001; 44: 1333–6. 46. Russell AH, Harris J, Rosenberg PJ, et al. Anal sphincter conservation for patients with adenocarcinoma of the distal rectum: long-term results of radiation therapy oncology group protocol 89-02. Int J Radiat Oncol Biol Phys 2000; 46: 313–22. 47. Mohiuddin M, Marks G, Bannon J. High-dose preoperative radiation and full thickness local excision: a new option for selected T3 distal rectal cancers. Int J Radiat Oncol Biol Phys 1994; 30: 845–9. 48. Bonnen M, Crane C, Vauthey JN, et al. Long-term results using local excision after preoperative chemoradiation among selected T3 rectal cancer patients. Int J Radiat Oncol Biol Phys 2004; 60: 1098–105. 49. Callender GG, Das P, Rodriguez-Bigas MA, et al. Local excision after preoperative chemoradiation results in an equivalent outcome to total mesorectal excision in selected patients with T3 rectal cancer. Ann Surg Oncol 2010; 17: 441–7. 50. Yeo SG, Kim DY, Kim TH, et al. Local excision following pre-operative chemoradiotherapy-induced downstaging for selected cT3 distal rectal cancer. Jpn J Clin Oncol 2010; 40: 754–60. 51. Garcia-Aguilar J, Qian S, Thomas CR Jr., et al. Pathologic complete response (pCR) to neoadjuvant chemoradiation (CRT) of uT2uN0 rectal cancer (RC) treated by local excision (LE): results of the ACOSOG Z6041 trial. J Clin Oncol 2010; 28(suppl.): 15s. 52. Nair RM, Siegel EM, Chen DT, et al. Long-term results of transanal excision after neoadjuvant chemoradiation for T2 and T3 adenocarcinomas of the rectum. J Gastrointest Surg 2008; 12: 1797–806. 53. Kitajima K, Fujimori T, Fujii S, et al. Correlations between lymph node metastasis and depth of submucosal invasion in submucosal invasive colorectal carcinoma: a Japanese collaborative study. J Gastroenterol 2004; 39: 534–43. 54. Leong AF, Seow-Choen F, Tang CL. Diminutive cancers of the colon and rectum: comparison between flat and polypoid cancers. Int J Colorectal Dis 1998; 13: 151–3. 55. Chambers WM, Khan U, Gagliano A, et al. Tumour morphology as a predictor of outcome after local excision of rectal cancer. Br J Surg 2004; 91: 457–9. 56. Hermanek P, Gall FP. Significance of local control of colorectal cancer. Fortschr Med 1985; 103: 1041–6.

08/02/13 5:37 PM

190  Local excision and transanal endoscopic microsurgery 57. Chakravarti A, Compton CC, Shellito PC, et al. Longterm follow-up of patients with rectal cancer managed by local excision with and without adjuvant irradiation. Ann Surg 1999; 230: 49–54. 58. Hase K, Shatney C, Johnson D, et al. Prognostic value of tumor ‘budding’ in patients with colorectal cancer. Dis Colon Rectum 1993; 36: 627–35. 59. Wang HS, Liang WY, Lin TC, et al. Curative resection of T1 colorectal carcinoma: risk of lymph node metastasis and long-term prognosis. Dis Colon Rectum 2005; 48: 1182–92. 60. Gall FP, Hermanek P. Cancer of the rectum: local excision. Surg Clin North Am 1988; 68: 1353–65. 61. Kudo S. Endoscopic mucosal resection of flat and depressed types of early colorectal cancer. Endoscopy 1993; 25: 455–61. 62. Ishizaki Y, Takeda Y, Miyahara T, et al. Evaluation of local excision for sessile–type lower rectal tumors. Hepatogastroenterology 1999; 46: 2329–32. 63. Kikuchi R, Takano M, Takagi K, et al. Management of early invasive colorectal cancer: risk of recurrence and clinical guidelines. Dis Colon Rectum 1995; 38: 1286–95. 64. Kobayashi H, Mochizuki H, Kato T, et al. Is total mesorectal excision always necessary for T1–T2 lower rectal cancer? Ann Surg Oncol 2010; 17: 973–80. 65. Perez RO, São Julião GP, Habr-Gama A, et al. The role of carcinoembryogenic antigen in predicting response and survival to neoadjuvant chemoradiotherapy for distal rectal cancer. Dis Colon Rectum 2009; 52: 1137–43.

HEBK001-C11_p171-190.indd 190

66. Park YA, Sohn SK, Seong J, et al. Serum CEA as a predictor for the response to preoperative chemoradiation in rectal cancer. J Surg Oncol 2006; 93: 145–50. 67. Muthusamy VR, Chang KJ. Optimal methods for staging rectal cancer. Clin Cancer Res 2007; 13: 6877–84s. 68. Bipat S, Glas AS, Slors FJ, et al. Rectal cancer: local staging and assessment of lymph node involvement with endoluminal US, CT, and MR imaging – a metaanalysis. Radiology 2004; 232: 773–83. 69. Kwok H, Bissett IP, Hill GL. Preoperative staging of rectal cancer. Int J Colorectal Dis 2000; 15: 9–20. 70. Santoro GA, D’Elia A, Battistella G, et al. The use of a dedicated rectosigmoidoscope for ultrasound staging of tumours of the upper and middle third of the rectum. Colorectal Dis 2007; 9: 61–6. 71. Rothenberger D, Ricciardi R, Madoff RM. Procedures for rectal cancer. In Ashley SW (ed.). ACS Surgery: Principles and Practice. Philadelphia, PA, BC Decker, 2010: 933–48. 72. Lee WY, Lee WS, Yun SH, et al. Decision for salvage treatment after transanal endoscopic microsurgery. Surg Endosc 2007; 21: 975–9. 73. Borschitz T, Heintz A, Junginger T. The influence of histopathologic criteria on the long-term prognosis of locally excised pT1 rectal carcinomas: results of local excision (transanal endoscopic microsurgery) and immediate reoperation. Dis Colon Rectum 2006; 49: 1492–506, 1500–05.

08/02/13 5:37 PM

12 Pathology assessment Philip Quirke, Tim Palmer, Gordon G.A. Hutchins and Nick P. West

Introduction Pathologists play a key role in the multidisciplinary team management of rectal cancer. The pathological analysis of the excised specimen provides important prognostic information on the stage of the tumour, the accuracy of radiology and the quality assurance of the surgery. Quality assurance emanates from reporting the completeness of tumour excision and the precise planes of excision, as represented by the macroscopic appearance of the specimen. The pathologist can assess the response to preoperative therapy and help to determine the need for postoperative or adjuvant therapy, either radiotherapy or chemotherapy. The pathologist can also identify patients at high risk of metachronous disease and screen for possible hereditary nonpolyposis colorectal cancer using immunohistochemistry for mismatch repair gene expression. Pathology may also help to educate radiologists and assure the quality of their reporting. More recently, pathologists have undertaken further molecular testing to help predict the type of therapy that may be effective; for example, Ki-ras mutations may be used to predict response to anti-epidermal growth factor receptor (anti-EGFr) antibodies. The current era is moving into whole-genome sequencing, which will have far-reaching implications for the diagnosis, prognosis and treatment of colorectal cancer. A further crucial responsibility is submission of high-quality pathology reports, preferably

HEBK001-C12_p191-202.indd 191

via electronic proformas, to cancer registries so that the work of individuals, teams and institutions can be compared nationally and internationally and so that excellent and potentially suboptimal practice can be identified.

Applied anatomy The rectum is situated with the confines of the bony pelvis and adjacent to important structures. The pelvic autonomic nerves run very close to the perimesorectal resection planes, the ureters are located laterally, and the prostate in male patients and the posterior vaginal wall in female patients are in close proximity to the anterior surface of the rectum. The sigmoid and upper and middle rectum are encased in peritoneum, which decreases in extent distally, but anteriorly at the peritoneal reflection the surgeon must incise the peritoneum and create the anterior surgical plane that is so critical to a successful surgical operation. Posteriorly the surgeon enters the perimesorectal fascial plane, following either the visceral or the parietal plane. This can be distinguished by the thickness of the fascia applied to the mesorectum. The fascia blends with the peritoneum laterally above the peritoneal reflection, enclosing the surgical ‘package’, which needs to be removed intact. Posteriorly, deep within the mesorectum, close to the mesorectal fascial surgical circumferential resection margin

08/02/13 5:36 PM

192  Pathology assessment

(CRM), is the ­superior rectal artery and its division into its left and right branches. This artery is important because lymphatic drainage is associated with the arterial blood supply and lymph nodes can be found clustered near to the artery and adjacent to the mesorectal fascia. The mesorectum narrows superiorly in its cranial portion, bulges at the level of the mid-rectum and then narrows to taper out at the puborectal sling. The levators are applied to the lowest part of the mesorectum and can be taken with the mesorectum when excising low rectal cancers or large cancers involving the anal canal. The anatomy of the anal sphincters varies between individuals, with variable bulk of the muscle layers and differing anal canal lengths. It is a highly complex evacuation control mechanism and deserves further anatomical study. The intersphincteric plane, an important potential surgical plane between the internal and external sphincters, can be used to achieve a very low reconstruction with partial or even complete removal of the internal sphincter, thus avoiding the need for abdominoperineal excision (APE). The anatomical structures of the anal sphincter are not well known by pathologists but are worthy of detailed study to analyse and improve low rectal cancer surgery and to accurately stage these tumours. Radiologists, surgeons and pathologists are increasingly skilled at understanding the anatomical relationships of rectal cancer; optimal preoperative images help in providing the anatomical road map for the surgeon to plan the appropriate planes for surgical intervention. The pathologist can then determine how closely the surgeon followed the intended plane.

The circumferential margin In 1982, while discussing the mismatch between the infrequently observed involvement of the distal margin of resection and the high rate of local recurrence in rectal cancer, we decided to look at the ‘lateral’ surgical margins on the resected specimens, i.e. the margin created by the surgeon around the rectum. It rapidly became apparent that this margin, initially referred to as the ‘lateral resection margin’ and subsequently termed the ‘circumferential resection margin’, was frequently involved in rectal cancer. Fifty-two cases were ana-

HEBK001-C12_p191-202.indd 192

lysed, all patients were followed up for 2 years and the outcomes were published in 1986.1 Just before this, Chan and colleagues from Hong Kong published 50 cases showing tumour involvement of the same margin but without clinical follow-up.2 We were able to demonstrate the importance of this margin as a cause of local recurrence (although ‘progressive persistent disease’ might be a more accurate term). We reported the high frequency of local recurrence and the importance of a margin greater than 1 mm rather than 0 mm, as had previously been the gold standard. From this emanated the concept of an involved margin where tumour was found within 1 mm of the surgical resection margin. This initial 1986 publication was followed by reports on two larger series3,4 and by others summarized in a review article.5 The principles of CRM involvement have been documented in numerous studies, including many prospective randomized trials (Dutch Total Mesorectal Excision (TME) Trial, Medical Research Council (MRC) CR07 Trial, MRC Conventional versus Laparoscopic-Assisted Surgery in Colorectal Cancer (CLASICC) Trial) and have confirmed the importance of a positive or negative CRM. More recently, a meta-analysis was reported,5 and even with a group of mesorectal trained surgeons it retains its importance.6,7 Involvement of the CRM occurs most frequently anteriorly, where the mesorectum is at its thinnest. Additionally, the CRM is at high risk in the low mesorectum from surgical coning at the puborectalis and sphincter muscles when the surgeon undertakes standard APE. Anteriorly there is the least tissue surrounding the muscle tube,8,9 and following the intended plane under excellent direct vision is essential. For advanced posterior and lateral tumours, removing the levators, or if appropriate the ischiorectal fat, can increase the potential clearance. Tumour involvement either at the margin or 1 mm or closer to the margin carries the highest risks of local recurrence,5,10 and this risk decreases as the distance increases between the tumour and the resection margin. The mode of margin involvement does not seem to matter and can be by direct, discontinuous, venous, lymphatic or neural spread.4 The presence of tumour within a lymph node within 1 mm of the margin does not appear to confer as high a risk in the two small series identified, but the database on this is very small. These areas are

08/02/13 5:36 PM

PLANES OF SURGERY  193

reviewed elsewhere,5 but the hypothesis is that if the tumour is confined to lymphatic pathways, then spread will be within this system and will thus be upwards towards higher nodes rather than towards the CRM. This does, however, depend on the surgeon removing the TME plane perfectly. The importance of an involved CRM increases after preoperative treatment, with an increase in the risks of local failure,5,11–14 and there is a strong association with an increased local recurrence but an equivalently poor survival. Achieving a complete excision locally is essential in the curative management of rectal cancer. Although preoperative treatment helps to achieve a clear margin in some advanced cases, if the margin is involved after neoadjuvant therapy the outcome is poor. It has been proposed that consideration should be given to the strategy of trying alternative chemotherapy approaches to achieve down-staging before intervening with definitive radiotherapy and subsequent surgery. The delivery of induction chemotherapy before chemoradiotherapy in the Expert C trial appears to provide additional down-staging,15 and in patients resistant to the first therapy an alternative combination may be tried before adding radiation. An alternative approach is being studied by the Swedish trialists, where 6 weeks of chemotherapy is given following short-course radiotherapy. These approaches are still experimental and being evaluated.

Planes of surgery Heald and colleagues described TME in 198216 and excellent outcomes in 1986.17 Subsequent audit of these results confirmed the survival benefit of this operation in highly skilled hands.18 The differences in specimen appearances and surgical planes of standard anterior resection and Heald’s TME specimens became apparent when Quirke, Heald and others participated in a pathological dissection workshop in 1993 in Oslo as part of the Norwegian rectal cancer education courses. Total mesorectal excision achieves a smooth surgical margin by following the anatomical plane provided by the outer surface of the mesorectal fascia. The standard surgical planes achieved by the Norwegian surgeons at that time were either on the muscularis propria, where the surgical excision extended down on to the muscularis propria; intramesorectal, where the

HEBK001-C12_p191-202.indd 193

plane was within the mesorectum; or mesorectal, where the planes were followed accurately. Coning of the specimen near to the distal excision margin was also frequent, bringing the surgical CRM closer or into the area of the cancer. This classification into mesorectal, intramesorectal and muscularis propria specimen descriptions was developed from these observations.19,20 The classification was assessed prospectively by local pathologists in the MRC CR07 trial21 and retrospectively by central review of photographs in the Dutch TME Trial.22 Both studies demonstrated that the plane of surgery was related to local recurrence; additionally, the Dutch study showed the plane of surgery was related to survival. Three further studies have confirmed the relationship between specimen quality and local recurrence.23–25 Unsurprisingly, the plane achieved was related to CRM positivity rates, with the lowest positive CRM rates in mesorectal plane surgery. The specimen plane was not related directly to the stage of the tumour, with many suboptimal specimens in early-stage tumours. The frequency of CRM involvement and the plane of surgery have been documented in the Magnetic Resonance Imaging in Rectal Cancer European Equivalence (MERCURY) study,7 with the lowest CRM positivity in mesorectal plane excisions and a higher rate of mesorectal excisions than seen in the MRC CR07 study. Surgeons showed improvement in the frequency of mesorectal excisions over the trial. There was an additive effect of radiotherapy on all grades of surgery; with this combination, good surgery led to a 1 per cent local recurrence rate compared with 6 per cent in patients who received no irradiation.21 In many studies, the quality of surgery was much poorer in APE specimens. There was no improvement in quality over the period of the trial in CR07 in APE, whereas there was an improvement in quality in anterior resection. Thus, achieving optimal planes improves outcomes in patients with rectal cancer undergoing anterior resection; focusing on this during surgery, and its subsequent pathological assessment, has major benefits to patients. This can be seen in the reduction in local recurrence rates and improving survival rates seen in rectal cancer where such techniques have been adopted in ­Norway,26 Stockholm,27 the Netherlands,28 Vancouver29 and British Columbia;30 many other results are expected over the next few years.

08/02/13 5:36 PM

194  Pathology assessment Table 12.1. CRM involvement in anterior resections compared with abdominoperineal excisions. Abdominoperineal

Anterior

n

excision (%)

resection (%)

686

36.5

22.3

282

33

13

Nagtegaal et al.

1586

29

13

Guillou et al.31

400

21

10

21

10

Study Marr et al.

10

Taylor et al.6,7 22

M. Peters,  personal communication, 2008 Quirke et al.21

1350

17

8

Wibe et al.32

2136

12

5

Abdominoperineal excision The outcome after APE for rectal cancer is generally inferior to that after anterior resection, with an 8–10 per cent worse overall survival in many studies. The main cause appears to be related to the high CRM positivity (Table 12.1) and specimen perforation (Table 12.2) rates in APE. This may be explained on the basis of the surgical planes used when resecting these tumours. This aspect of APE specimens was first identified in 200236 and subsequently verified in a joint study of the APE specimens in the Dutch TME Trial.33 In the latter study, a third of cases showed the surgical plane to extend into the lumen, submucosa or internal sphincter. In the other twothirds of cases, the typical ‘waist’ seen in most standard APEs was noted. A solution to reduce margin involvement and specimen perforation was identified at the Karolinska Institute, Stockholm, when the concept of what is now termed an ‘extralevator APE’ was performed and proTable 12.2. Perforation rates in anterior resections compared with abdominoperineal excisions. Abdominoperineal

Anterior

Study

n

excision (%)

resection (%)

Nagtegaal et al.33

1586

13.7

2.5

Wibe et al.

2136

16

4

Eriksen et al.34

2873

15.4

3.6

Stockholm audit35

613

12

4

32

HEBK001-C12_p191-202.indd 194

moted by Holm.37 The concept of this operation involves removing the levators attached to the lower mesorectum and the entire anal canal with internal and external sphincters and a greater or lesser degree of ischiorectal and ischioanal fat. This occurs under direct vision, with the patient in the prone position for the perineal part of the procedure. The supine position has also been promoted with an emphasis on excellent direct vision.38 This approach can provide the critical few extra millimetres of protection around a ­locally advanced low rectal tumour and prevent perforation of the anal canal. Holm has categorized excision of the sphincters and anal canal into three types of operation: the standard APE, with its perilous waist, high margin positivity and high perforation rates; the extralevator APE, where the levators are removed en bloc; and the extended extralevator APE, with a unilateral or bilateral extension into the ischiorectal fossa, dependent on the local extension of the cancer. The latter is more commonly needed for advanced recurrent squamous cancer, although on occasions it may be required for very advanced adenocarcinoma. Any of these operations can include removal of other local organs such as the posterior wall of the vagina in female patients or variable amounts of the prostate in male patients. A study compared the specimens from Holm’s extralevator operation with the standard operation APE specimens from our own institution in Leeds and demonstrated a markedly reduced CRM positivity and perforation rate.8 An increase in the amount of tissue removed by extralevator surgery compared with standard surgery was also documented. These studies were extended to collect extralevator specimens from colleagues in the UK and Europe and found identical results, with a drop in margin involvement and specimen perforation.9 On reviewing the literature, it also became apparent that there were other units in Paris and Wroclaw that used this approach. The Hôpital St Antoine in Paris reported excellent outcomes, with 10 per cent local recurrence in all patients and 5 per cent in curative excisions and a 5-year survival of 76 per cent after APE for rectal cancer.39 Discussions with the authors and photographic proof from Emmanuel Tiret and Roland Parc confirmed that they adopted an extralevator approach. Similarly, Bebenek has reported what he terms ‘abdomino-sacral APE’

08/02/13 5:36 PM

REPORTING RECTAL CANCER  195

with excellent results of 8 per cent CRM positivity, 4 per cent perforation, 4.4 per cent local recurrence and 5-year survival of 68 per cent in more advanced stage low rectal cancers.40 Thus, there are now substantial data on the problems of the ‘standard’ APE and accumulating evidence of the superiority of the extralevator approach to APE.9 Ongoing issues pertain to patient selection, the optimum approach, whether prone or supine,38,41 whether open, laparoscopic42,43 or robotic,44 and the best methods of closing the pelvic floor to optimize wound healing and reduce perineal hernias.45,46 While awaiting the outcomes of these issues, however, the potentially avoidable local recurrences and impaired survival from standard APE should not be discounted. The National Health Service in England has commissioned a pilot programme on training and education in this area for English multidisciplinary teams (www. lorec.nhs.uk), and there are similar initiatives in Denmark and other health-care systems.

Reporting rectal cancer Accurate reporting of rectal cancer specimens is crucial because the details may have a major impact on the multidisciplinary management of what is a complex and common cancer. It is therefore important to ensure that adequate time and resources are available to undertake this task. Pathologists need to feel like an important component of the team, and to have seen good-quality TME and extralevator APE specimens so that they understand the anatomy and potential surgical planes that they need to evaluate and categorize. Ideally, pathologists should understand the magnetic resonance imaging (MRI) images so that preoperative scans can guide specimen dissections and provide accurate feedback to radiologists, surgeons and others in the multidisciplinary team. Increasingly, radiologists can accurately describe the site of the tumour, and report the distance between the margin of the tumour or tumour deposit and the CRM, whether this is the mesorectal fascia, levators, anal sphincters, ischiorectal fat or other structures. The distance from the muscularis propria to the edge of the tumour, and large vessel extramural vascular invasion, may also be recordable; on occasion, peritoneal involvement may be documented.

HEBK001-C12_p191-202.indd 195

The accuracy of MRI with regard to nodal deposits is debatable, as larger nodal size does not always equate with tumour involvement and many nodal deposits are in small lymph nodes. Indeed, large nodal size can equate with a better immune response and thus a better outcome. The texture of nodes and their irregularity may give a better indication of involvement, but it is still suboptimal. The MRI images provide a surgical road map for the operation, but it is important to note that inaccurate calling of MRI images can have a profound impact on patient treatment, with either over- or under treatment as a consequence.

Preparation of the Specimen The following preparation and dissection notes were prepared for the National Cancer Research Institute (NCRI) Aristotle trial (see www.controlled-trials.com/ISRCTN09351447) and are very similar to those used within the Low Rectal Cancer and MERCURY studies. They should be used in conjunction with the Royal College of Pathologists’ dissection guidelines and the reporting proformas (see www.rcpath.org/Resources/RCPath/ Migrated%20Resources/Documents/G/G049ColorectalDataset-Sep07.pdf and www.rcpath.org/ Resources/RCPath/Migrated%20Resources/ Documents/G/G049ColorectalDatasetAppendixCSep07.doc). Whether the specimen is received fresh or fixed, the handling should be equivalent. Before opening the specimen, the anterior and posterior surfaces should be photographed to document the quality of removal by assessing the anatomical planes that have been achieved. Close-up images of the front and back of the levator/anal sphincter facilitate assessment of extralevator APE specimens. Each photograph should include a centimetre scale placed beside the specimen. The pathologist should grade the quality of the surgery for both the mesorectum and the levator/anal sphincter area. It is advantageous to paint the surgically created margin as early as possible to allow the dye to adhere to the specimen as securely as possible so it does not spread widely on dissection. The adherence of the dye increases with the length of fixation. The specimen can then be opened from the proximal margin distally to 2–5 cm above the tumour. The distal end should be kept intact. If the specimen

08/02/13 5:36 PM

196  Pathology assessment

is received unfixed, and fresh material is required for storage or research, it should be taken at this stage. The tumour can be everted through the upper anterior specimen incision and luminal tumour removed, or a jumbo biopsy forceps can be used to obtain tumour material. A piece of foam or tissue paper soaked in formalin can be inserted through the tumour if appropriate to improve fixation; if received fixed, the specimen can be inflated with formalin and sealed at both ends. The specimen can then be placed in formalin. Before dissection it is acceptable to inflate the specimen with formalin, fix and then take photographs. The anterior surface of the tumour should never be opened as it destroys the assessment of the anterior CRM, the area most frequently involved by cancer by direct invasion, peritoneal involvement or perforation. The luminal size or area of the tumour does not convey prognostic information. The anterior high rectum is covered by peritoneum; this area, especially the paracolic gutters, should be inspected carefully for peritoneal involvement. Although peritoneal involvement is an adverse prognostic indicator, peritoneal involvement does not constitute an involved CRM, as this is not a surgically created margin.

Dissection Consideration should be given to enhancing the lymph node yield by the use of methylene or patent blue injection47 into the artery at the time the specimen is removed by the surgeon, or by use of glacial acetic acid, ethanol, distilled water and formaldehyde (GEWF) fixation48 or post-dissection using xylene clearance.49,50 This can be especially useful for post-treatment resections where nodal size is reduced.51,52 Anterior and posterior non-peritonealized surfaces are painted with ink. It should be remembered that the CRM applies only to the surgically incised mesorectal planes and not the peritonealized surfaces. The mesorectal surface is larger posteriorly and extends up to a higher level than it does anteriorly. After the resection surfaces have been inked, the specimen is fixed in formalin for a minimum of 48 h, but longer if possible. The firmer the mesorectum after fixation, the thinner the slices can be and the more thoroughly the tumour can be examined.

HEBK001-C12_p191-202.indd 196

The macroscopic description should be completed, specifically noting the presence of a perforation of the tumour or mucosa and the site of the perforation. It should be stated whether the tumour perforation is present in an area covered by peritoneum or a surgical margin, and whether it is above or at the level of the sphincters. The presence or absence and completeness of the attachment of levator ani to the mesorectum on the specimen should be described. The descriptions of grading are given below. The specimen should be sliced as thinly as possible, starting from the distal margin to 2–5 cm above the tumour. These slices should be laid out in good light, starting with the most distal slice at the top left-hand corner, and the most proximal slice ending up as the last slice. The face presented to the camera should be consistent in all the slices. The slices should be photographed including a centimetre scale. The minimum distance of the tumour to the CRM should be described, as should the maximum depth of invasion through the muscularis propria and the structures invaded by the tumour. If the CRM is free of tumour, it should be noted whether there is normal tissue at the margin or whether it is fibrotic tissue following tumour regression potentially indicating a previously involved margin. If the CRM is involved (confirmed on histology), then the mode of involvement and the distance of involvement at that site should be stated (e.g. small area measuring 2 mm or more extensive involvement measuring 10 mm). It is preferable to sample the main tumour by embedding each tumour-bearing slice and cutting one to four large mount sections, although this is not possible in many laboratories; where it is not possible, it is important to ensure that a minimum of five blocks of tumour are taken, including the CRM where the tumour is close and any other important macroscopic pathology (see below). As many lymph nodes as possible should be dissected from the specimen. Lymph nodes after preoperative therapy are generally smaller and frequently around 1 mm in size; this should be borne in mind, as they will be more difficult to locate. The use of methylene blue injection or GEWF fixation will increase the number of nodes found. Involvement of the peritoneum is defined as per Shepherd and colleagues’ recommendations.53,54

08/02/13 5:36 PM

REPORTING RECTAL CANCER  197

Extramural vascular invasion (EMVI) is defined by involvement of a thick vascular structure with a smooth muscle wall that will contain elastin on elastic staining. This should be looked for closely. If an isolated tumour deposit is present close to an arterial structure, without an accompanying vein, consider vascular invasion. Increased reporting frequencies of EMVI have been found following the addition of elastic staining to haematoxylin and eosin.55,56

T Staging of Low Rectal Cancers The T staging of cancers above the sphincters is straightforward and is the same as that for colon cancers. As some mid-rectal and all low rectal adenocarcinomas have a proportion of the lesion within the region of the sphincters, however, T staging is problematic. Currently, T staging of adenocarcinoma in the area of the sphincters is unsound. TNM 7 states that such tumours should be staged as anal cancers by tumour size. In the absence of a robust staging system, the only solution is to describe the maximal anatomical extent of spread, both above the sphincter and at the level of the sphincter separately, to allow subsequent analysis. We propose that the maximum level of invasion above and at the level of the sphincter should be recorded separately by extent of maximal spread. This should be captured by describing the extent of spread (e.g. submucosal, internal sphincter, intersphincteric space, external sphincter) and also measuring the extent of spread from the muscularis propria and, where applicable, from the edge of the internal sphincter when a tumour involves the sphincters. We need to know whether one or both sphincters are involved by tumour. This is relatively easy to do, as the internal sphincter is an easily identified reference point, being composed of pearly-white smooth muscle. Involvement of striated muscle indicates involvement of the levators, puborectalis or external sphincter.

Assessment of Quality of Surgery: Grading The mesorectum and the anal canal/levator parts of the specimen should be graded separately. For an anterior resection specimen there will only be one grade –

HEBK001-C12_p191-202.indd 197

the grade for the mesorectum. For APE, there will be a grade for the mesorectum and another grade for the levator and anal canal area below the mesorectum.

Quality of Resection of the Mesorectum The quality of a mesorectal resection can be easily assessed macroscopically and graded as shown in Table 12.3. This classification has been used in several series and two trials. The plane of surgery correlates with frequency of CRM positivity, with the lowest CRM positivity rates in mesorectal plane excision specimens and the highest rates in muscularis propria excision specimens.

Quality of Resection in Abdominoperineal Surgery The quality of surgery of the levator/anal canal area below the mesorectum can be assessed as shown in Table 12.4. Table 12.3. Summary of mesorectal grading. Fascial plane

Description

Mesorectal

The mesorectum is smooth, with no violation  of the fat and good bulk to the mesorectum anteriorly and posteriorly. The distal margin appears adequate, with no coning near the tumour. No defect is more than superficial or 5 mm deep.

Intramesorectal

There is moderate bulk to mesorectum but  irregularity of the mesorectal surface. There is moderate coning of the specimen towards the distal margin. At no site is the muscularis propria visible, with the exception of the area of insertion of levator muscles. There is moderate irregularity of the CRM.

Muscularis

There are areas of substantial loss of

  propria

 mesorectal tissue. Deep cuts and tears down on to the muscularis propria are present. On cross-section there is a very irregular CRM, with little bulk to the mesorectal fat; the muscularis propria forms the CRM in places.

08/02/13 5:36 PM

198  Pathology assessment Table 12.4. Summary of abdominoperineal excision grading. Plane

Description

Levator

The surgical plane lies external to the levators,  with the levators being removed en bloc with the specimen. This creates a cylindrical specimen with the levators forming an extra protective layer on the sphincters.

Sphincteric

Either there are no levator muscles attached  to the specimen or there is only a very small cuff and the resection margin on the surface of the sphincters. The specimen has a waisted/apple-core appearance.

Intrasphincteric/

The surgeon has inadvertently entered the

  submucosal

 sphincters or even deeper into the submucosa or perforated the specimen at any point.

For an anterior resection specimen, there will be a single mesorectal grade. For an APE specimen there will be two grades.

Definitions Used in Pathology Position of tumour

The position of the tumour should be noted accurately. Initially this involves documentation of the surface involvement (i.e. anterior quadrant, posterior quadrant, lateral quadrant, combinations of these). To correlate the position with the MRI report, however, the tumour position should be ­reported from the distal resection margin with the mesorectum posteriorly and the peritoneal reflection anteriorly. This can be documented as a relationship to a clock face on the reporting proforma, with anterior at the 12 o’clock position (e.g. ‘maximal extent of tumour present between 10 o’clock and 3 o’clock’). Relationship to peritoneal reflection

A crucial landmark for recording the site of rectal cancers is the peritoneal reflection. This is identified from the exterior surface of the anterior aspect of the specimen. Rectal cancers are classified according to whether they are: entirely above the level of the peritoneal reflection anteriorly; l astride or at the level of the peritoneal reflection anteriorly; l

HEBK001-C12_p191-202.indd 198

entirely below the level of the peritoneal reflection anteriorly.

l

Relationship to circumferential resection margin

Anteriorly the upper rectum is covered by peritoneum. Only the area below the peritoneal reflection is at risk of surgical CRM involvement. Posteriorly this area and the area above it, a triangular bare area running up to the start of the sigmoid mesocolon, is at risk not only from direct tumour spread but also from metastatic deposits in lymph nodes that lie against the circumferential margin. It is recommended that the whole of this margin (i.e. the mesorectum) be painted with a marker such as India ink or Alcian blue followed by acetic acid to fix the dye on the surface before dissecting the specimen. The tumour is then best sliced serially at 3- to 4-mm intervals to select blocks from the area above and below the tumour to look for metastatic deposits. If lymph nodes lie against the circumferential margin, then these should be included in the block. Relationship to extent of extramural invasion of tumour

When assessing the relationship to the CRM, on the whole-mount section the corresponding relationship between the outer surface of the muscle coat and the maximum depth of extramural invasion needs to be measured. This is performed using the Vernier scale on the microscope or by overlaying a scale on the microscope slide. Lymph nodes

All lymph nodes found in the specimen should be sampled and counted, regardless of their site and size. The number of positive lymph nodes must be equal to or less than the number of lymph nodes sampled. Extramural tumour deposits measuring 3 mm or larger are counted as involved lymph nodes, even if no residual lymph node structure can be identified, as defined in TNM 5. If lymph nodes are present but the structure cannot be discerned, the presence of satellite nodules should be recorded as well. Smaller satellite deposits are regarded as apparent discontinuous extensions of the main tumour, and their presence should be recorded. In TNM 5, pN1 (where ‘p’ corresponds to ‘pathological’) denotes involvement of one to three nodes

08/02/13 5:36 PM

REPORTING RECTAL CANCER  199

and pN2 denotes involvement of four or more nodes. We do not recommend using TNM 6 or 7 because of the many issues outlined in several papers.57–59 Distance to distal resection margin

This is measured from the nearest cut end of the specimen to the tumour, not the circumferential margin. It is necessary to examine the margins histologically only if the tumour extends macroscopically to within 30 mm of the distal or circumferential margin. For tumours further away from the margin, it can be assumed that the cut ends are not involved. Exceptions to this recommendation are adenocarcinomas that are found on subsequent histology to have an exceptionally infiltrative growth pattern, show extensive vascular or lymphatic permeation, or are undifferentiated carcinomas. Relationship to dentate line

This can be measured only for low rectal tumours in APE specimens. Tumour perforation

If the tumour has perforated into the peritoneal cavity or through the mesorectal fascia or sphincter complex, then these cases should be recorded as a perforation. The TNM staging system states that only perforation of the peritoneum should be called pT4, but perforation/surgical rupture of tumour in such areas confers a poor prognosis34 and many of them would be reported as pT4 in routine practice in the UK. Tumour differentiation

The differentiation of the tumour should be defined on the dominant area of tumour as well differentiated or moderately differentiated, equivalent to low grade and poorly differentiated or undifferentiated in the World Health Organization ‘blue book’,57 which is equivalent to high-grade tumour. Other types of differentiation (e.g. mucinous adenocarcinomas, signet ring, undifferentiated) and the presence of high-grade dysplasia should be documented. Assessment after preoperative therapy

The prefix ‘y’ indicates either a staging scan or pathological assessment after preoperative (neoadjuvant) therapy. Such tumours should be staged with the prefix ‘y’ before pTNM staging. There is no doubt

HEBK001-C12_p191-202.indd 199

that lowering the ypT stage is of some value, but the relative risk/benefit ratio is still uncertain for many patients. Some studies suggest that it improves outcome.14 The occurrence of a complete pathological response (ypT0 ypN0) is very important, as patients with this finding at pathology have excellent cancer-related outcomes.58,59 If ypT0 and ypNx are included as complete pathological response, then the frequency of complete response increases by 7 per cent. One major issue with complete response is the pathological dissection and sampling protocol followed. Lax pathological dissection and sampling lead to a higher frequency of complete response. For the Sanofi Capecitabine, Oxaliplatin, Radiotherapy and Excision (CORE) Trial, an international panel recommended taking five blocks of tumour and, if no tumour was identified, to cut three further levels. If tumour is still absent, then the whole area of abnormality should be embedded and, if necessary, a further three levels sectioned. If at this stage no tumour is identified, then this should be considered a complete response. It is critical for the continued use of complete pathological response that the sampling is standardized. The reports of tumour regression grading studies have been disappointing. Rödel and colleagues tested regression grading in the German trial. They failed to show a significant relationship to survival with five categories and then amalgamated the groups and obtained significance.60 Gosens and colleagues13 and Rullier and colleagues14 did not find any significant benefits correlated to tumour regression grading. It is clear that a complete response ypT0 ypN0 has an excellent prognosis and that good responders, usually regressing to pT1 or early pT2 N0, also have excellent cancer related outcomes.14 In our experience, it has become apparent that it is impossible to distinguish between no response and a poor response if looking at randomized trial material blinded to trial arm.

Optimal Reporting As outlined above, optimal reporting of a rectal cancer specimen requires extensive knowledge of the anatomy, radiology, pathology and treatment modalities. Excellent reporting is key for the best modern management of this disease. Pathology reporting goes beyond the crucial benefits to

08/02/13 5:36 PM

200  Pathology assessment

the individual patient. The pathological findings allow us to give feedback and audit to radiologists and surgeons and help oncologists to understand the effects of current and new therapies. This mechanism creates a virtuous loop of continuous improvement. Optimal practice and continuous feedback offers our best hope of improving the outcomes in colorectal cancer, one of the most common curable gastrointestinal tract tumours. The combination of screening, optimal surgery and appropriate preoperative and adjuvant therapy should allow us to make inroads into the mortality associated with this condition.

References   1. Quirke P, Durdey P, Dixon MF, et al. Local recurrence of rectal adenocarcinoma due to inadequate surgical resection: histopathological study of lateral tumour spread and surgical excision. Lancet 1986; 328: 996–9.   2. Chan KW, Boey J, Wong SK. A method of reporting radial invasion and surgical clearance of rectal carcinoma. Histopathology 1985; 9: 1319–27.   3. Adam IJ, Mohamdee MO, Martin IG, et al. Role of circumferential margin involvement in the local recurrence of rectal cancer. Lancet 1994; 344: 707–11.   4. Birbeck KF, Macklin CP, Tiffin NJ, et al. Rates of circumferential resection margin involvement vary between surgeons and predict outcomes in rectal cancer surgery. Ann Surg 2002; 235: 449–57.   5. Nagtegaal ID, Quirke P. What is the role for the circumferential margin in the modern treatment of rectal cancer? J Clin Oncol 2008; 26: 303–12.   6. Taylor FGM, Quirke P, Heald RJ, et al. One millimetre is the safe cut-off for magnetic resonance imaging prediction of surgical margin status in rectal cancer. Br J Surg 2011; 98: 872–9.   7. Taylor FGM, Quirke P, Blomqvist L, et al. Preoperative assessment of the circumferential resection margin by MRI predicts survival outcomes and recurrence in rectal cancer: 5 year follow up results of the MERCURY study [paper submitted]. Lancet Oncol.   8. West NP, Finan PJ, Anderin C, Lindholm J, Holm T, Quirke P. Evidence of the oncologic superiority of ­cylindrical abdominoperineal excision for low rectal cancer. J Clin Oncol 2008; 26: 3517–22.   9. West NP, Anderin C, Smith KJE, Holm T, Quirke P. Multicentre experience with extralevator abdominoperineal excision for low rectal cancer. Br J Surg 2010; 97: 588–99. 10. Marr R, Birbeck K, Garvican J, et al. The modern abdominoperineal excision: the next challenge after total mesorectal excision. Ann Surg 2005; 242: 74–82.

HEBK001-C12_p191-202.indd 200

11. Mawdsley S, Glynne-Jones R, Grainger J, et al. Can histopathologic assessment of circumferential margin after preoperative pelvic chemoradiotherapy for T3–T4 rectal cancer predict for 3-year disease-free survival? Int J Radiat Oncol Biol Phys 2005; 63: 745–52. 12. Luna-Pérez P, Bustos-Cholico E, Alvarado I, et al. Prognostic significance of circumferential margin involvement in rectal adenocarcinoma treated with preoperative chemoradiotherapy and low anterior resection. J Surg Oncol 2005; 90: 20–25. 13. Gosens MJ, Klaassen RA, Tan-Go I, et al. Circumferential margin involvement is the crucial prognostic factor after multimodality treatment in patients with locally advanced rectal carcinoma. Clin Cancer Res 2007; 13: 6617–23. 14. Rullier A, Laurent C, Capdepont M, Vendrely V, BioulacSage P, Rullier E. Impact of tumor response on survival after radiochemotherapy in locally advanced rectal carcinoma. Am J Surg Pathol 2010; 34: 562–8. 15. Chua YJ, Barbachano Y, Cunningham D, et al. Neoadjuvant capecitabine and oxaliplatin before chemoradiotherapy and total mesorectal excision in MRI-defined poor-risk rectal cancer: a phase 2 trial. Lancet Oncol 2010; 11: 241–8. 16. Heald RJ, Husband EM, Ryall RD. The mesorectum in rectal cancer surgery: the clue to pelvic recurrence? Br J Surg 1982; 69: 613–6. 17. Heald RJ, Ryall RD. Recurrence and survival after total mesorectal excision for rectal cancer. Lancet 1986; 1: 1479–82. 18. MacFarlane JK, Ryall RD, Heald RJ. Mesorectal excision for rectal cancer. Lancet 1993; 341: 457–60. 19. Quirke P, Dixon MF. The prediction of local recurrence in rectal adenocarcinoma by histopathological examination. Int J Colorectal Dis 1988; 3: 127–31. 20. Quirke P. Limitations of existing systems of staging for rectal cancer: the forgotten margin. In Soreide O Norstein J (eds). Rectal Cancer Surgery Optimization, Standardization, Documentation. Berlin, SpringerVerlag, 1997: 63–81. 21. Quirke P, Steele R, Monson J, et al. Effect of the plane of surgery achieved on local recurrence in patients with operable rectal cancer: a prospective study using data from the MRC CR07 and NCIC–CTG CO16 randomised clinical trial. Lancet 2009; 373: 821–8. 22. Nagtegaal ID, van de Velde CJH, van der Worp E, et al. Macroscopic evaluation of rectal cancer resection specimen: clinical significance of the pathologist in quality control. J Clin Oncol 2002; 20: 1729–34. 23. Maslekar S, Sharma A, Macdonald A, Gunn J, Monson JR, Hartley JE. Mesorectal grades predict recurrence after curative excision for rectal cancer. Dis Colon Rectum 2006; 50: 168–75. 24. García-Granero E, Faiz O, Muñoz E, et al. Macroscopic assessment of mesorectal excision in rectal cancer: a useful tool for improving quality control in a multidisciplinary team. Cancer 2009; 115: 3400–11.

08/02/13 5:36 PM

REFERENCES  201 25. Leite JS, Martins SC, Oliveira J, Cunhna MF, CastroSousa F. Clinical significance for macroscopic completeness of excision in rectal cancer. Colorectal Dis 2011; 13: 381–6. 26. Wibe A, Rendedal PR, Svensson E, et al. Prognostic significance of the circumferential resection margin following total mesorectal excision for rectal cancer. Br J Surg 2002; 89: 327–34. 27. Martling AL, Holm T, Rutqvist L-E, et al. Effect of a surgical training programme on outcome of rectal cancer in the County of Stockholm. Lancet 2000; 356: 93–6. 28. Kapiteijn E, Putter H, van de Velde CJ, et al. Impact of the introduction and training of total mesorectal excision on recurrence and survival in rectal cancer in the Netherlands. Br J Surg 2002; 89: 1142–9. 29. Phang PT, McGahan CE, McGregor G, et al. Effects of change in rectal cancer management on outcomes in British Columbia. Can J Surg 2010; 53: 225–31. 30. Phang PT, Woods R, Brown CJ, Raval M, Cheifetz R, Kennecke H. Effect of systematic education courses on rectal cancer treatments in a population. Am J Surg 2011; 201: 640–4. 31. Guillou PJ, Quirke P, Thorpe H, et al. Short-term endpoints of conventional versus laparoscopic-assisted surgery in patients with colorectal cancer (MRC CLASICC trial): multicentre, randomised controlled trial. Lancet 2005; 365: 1718–26. 32. Wibe A, Syse A, Andersen E, et al. Oncological outcomes after total mesorectal excision for cure for cancer of the lower rectum: anterior vs. abdominoperineal resection. Dis Colon Rectum 2004; 47:48–58. 33. Nagtegaal ID, van de Velde CJ, Marijnen CA, et al. Low rectal cancer: a call for a change of approach in abdominoperineal resection. J Clin Oncol 2005; 23: 9257–64. 34. Eriksen MT, Wibe A, Syse A, et al. Inadvertent perforation during rectal cancer resection in Norway. Br J Surg 2004; 91: 210–6. 35. Anderin C, Martling A, Hellborg H, Holm T. A population-based study on outcome in relation to the type of resection in low rectal cancer. Dis Colon Rectum 2010; 53: 753–60. 36. Marr R, Birbeck K, Garvican J, et al. The modern abdominoperineal excision: the next challenge after total mesorectal excision. Ann Surg 2005; 242: 74–82. 37. Holm T, Ljung A, Häggmark T, Jurell G, Lagergren J. Extended abdominoperineal resection with gluteus maximus flap reconstruction of the pelvic floor for rectal cancer. Br J Surg 2007; 94: 232–8. 38. Martijnse IS, Dudink RL, West NP, et al. Focus on extralevator perineal dissection in supine position for low rectal cancer has led to better quality of surgery and oncologic outcome. Ann Surg Oncol 2012; 19: 786–93. 39. Dehni N, McFadden N, McNamara DA, Guiguet M, Tiret E, Parc R. Oncologic results following

HEBK001-C12_p191-202.indd 201

abdominoperineal resection for adenocarcinoma of the low rectum. Dis Colon Rectum 2003; 46: 867–4. 40. Bebenek M. Abdominosacral amputation of the rectum for low rectal cancers: ten years of experience. Ann Surg Oncol 2009; 16: 2211–7. 41. Holm T. Abdominoperineal resection revisited: is positioning an important issue? Dis Colon Rectum 2011; 54: 921–2. 42. Singh B, Lloyd G, Nilsson PJ, Chaudri S. Laparoscopic extralevator abdominoperineal excision of the rectum: the best of both worlds. Tech Coloproctol 2012; 16: 73–5. 43. Vaughan-Shaw PG, King AT, Cheung T, et al. Early experience with laparoscopic extralevator abdominoperineal excision within an enhanced recovery setting: analysis of short-term outcomes and quality of life. Ann R Coll Surg Engl 2011; 93: 451–9. 44. Marecik SJ, Zawadzki M, Desouza AL, Park JJ, Abcarian H, Prasad LM. Robotic cylindrical abdominoperineal resection with transabdominal levator transection. Dis Colon Rectum 2011; 54: 1320–5. 45. Anderin C, Martling A, Lagergren J, Ljung A, Holm T. Short-term outcome after gluteus maximus myocutaneous flap reconstruction of the pelvic floor following extra-levator abdominoperineal excision of the rectum [published online ahead of print 8 Oct 2011]. ­Colorectal Dis 2011. 46. Horch RE, D’Hoore A, Holm T, Kneser U, Hohenberger W, Arkudas A. Laparoscopic abdominoperineal resection with open posterior cylindrical excision and primary transpelvic VRAM flap. Ann Surg Oncol 2012; 19: 502–3. 47. Martijnse IS, Dudink RL, Kusters M, Rutten HJ, ­Nieuwenhuijzen GA, Wasowicz-Kemps DK. Patent blue staining as a method to improve lymph node detection in rectal cancer following neoadjuvant treatment. Eur J Surg Oncol 2012; 38: 252–8. 48. Iversen LH, Laurberg S, Hageman-Madsen R, Dybdahl H. Increased lymph node harvest from colorectal cancer resections using GEWF solution: a randomised study. J Clin Pathol 2008; 61: 1203–8. 49. Jass JR, Miller K, Northover JM. Far clearance method versus manual dissection of lymph nodes in specimens of rectal cancer. Int J Colorectal Dis 1986; 1: 155–6. 50. Cawthorn SJ, Gibbs NM, Marks CG. Clearance technique for the detection of lymph nodes in colorectal cancer. Br J Surg 1986; 73: 58–60. 51. Sanchez W, Luna-Perez P, Alvarado I, Labastida S, Herrera L. Modified clearing technique to identify lymph node metastases in post-irradiated surgical specimens from rectal adenocarcinoma. Arch Med Res. 1996; 27: 31–6. 52. Wang H, Safar B, Wexner SD, Denoya P, Berho M. The clinical significance of fat clearance lymph node harvest for invasive rectal adenocarcinoma following neoadjuvant therapy. Dis Colon Rectum 2009; 52: 1767–73.

08/02/13 5:36 PM

202  Pathology assessment 53. Shepherd NA, Baxter KJ, Love SB. The ­prognostic importance of peritoneal involvement in colonic ­cancer: a prospective evaluation. Gastroenterology 1997; 112: 1096–102. 54. Ludeman L, Shepherd NA. Serosal involvement in gastrointestinal cancer: its assessment and significance. Histopathology 2005; 47: 123–31. 55. Vass DG, Ainsworth R, Anderson JH, et al. The value of an elastic tissue stain in detecting venous invasion in colorectal cancer. J Clin Pathol 2004; 57: 769–72. 56. Howlett CJ, Tweedie EJ, Driman DK. Use of an elastic stain to show venous invasion in colorectal carcinoma: a simple technique for detection of an important prognostic factor. J Clin Pathol 2009; 62: 1021–5. 57. Hamilton SR, Bosman FT, Bofetta P, et al. Carcinoma of the colon and rectum. In Bosman FT, Carneiro F,

HEBK001-C12_p191-202.indd 202

Hruban RH (eds). WHO Classification of Tumours of the Digestive System. Lyon, World Health Organization, 2010: 131–93. 58. Capirci C, Valentini V, Cionini L, et al. Prognostic value of pathologic complete response after neoadjuvant therapy in locally advanced rectal cancer: long-term analysis of 566 ypCR patients. Int J Radiat Oncol Biol Phys 2008; 72: 99–107. 59. Maas M, Nelemans PJ, Valentini V, et al. Long-term outcome in patients with a pathological complete response after chemoradiation for rectal cancer: a pooled analysis of individual patient data. Lancet Oncol 2010; 11: 835–44. 60. Rödel C, Martus P, Papadoupolos T, et al. Prognostic significance of tumor regression after preoperative chemoradiotherapy for rectal cancer. J Clin Oncol 2005; 23: 8688–96.

08/02/13 5:36 PM

13 Assessment and management of recurrence Peter J. Lee, Kirk K.S. Austin and Michael J. Solomon

Introduction Historically, surgery for locally recurrent rectal cancer (LRRC) has not been a widely accepted modality of treatment. It was believed to incur high morbidity and mortality. Cancer-related deaths with LRRC are unpleasant, with intractable pain, offensive perineal discharge and rectal bleeding, all culminating in a poor quality of life.1–3 The prognosis of such patients with incurable or inoperable recurrent rectal cancer without treatment is invariably poor, with a median survival of 9 months and less than 5 per cent 5-year survival.4–6 Surgical resection, defined as pelvic exenteration, was first described over 50 years ago, but it was not until the 1990s that substantial case series were published. As a consequence, rather than surgical referral and aggressive treatment, non-invasive palliative treatments such as chemotherapy or radiotherapy were implemented. This would relieve symptoms for only a short term and had a median survival of 10–17 months.2,7,8 Improvements in surgical techniques and postoperative care and the implementation of multimodality therapy have contributed to a significant reduction in operative mortality to somewhere in the region of 1 per cent in specialized units, but morbidity remains high, at about 30 per cent.9,10 It is pertinent to note that after pelvic exenteration, patients report an improvement in their

HEBK001-C13_p203-221.indd 203

quality of life, with a reduction in pain. In addition, their quality of life is better than in patients who receive non-surgical palliative treatment.11–13 Improved surgical techniques and adjuvant therapies for primary rectal cancer have reduced the rates of local recurrence to less than 10 per cent.14 Surgery for LRRC, or pelvic exenteration surgery, has become more complex, requiring more radical extended surgery and the involvement of a multidisciplinary team approach. The most important factor in exenteration surgery is to achieve a microscopically clear (R0) resection margin.4,9,10,15–18 Exenteration surgery is challenging and can be very difficult due to the anatomical confines of the bony pelvis and undefined surgical planes from previous surgery or radiotherapy. Therefore, such surgery is best done in specialized units. The major series on surgery for LRRC in the past 20 years have reported overall 5-year survival rates of 20–35 per cent.9,10,15,16,19 Data have shown that exenteration surgery for LRRC can be performed successfully with minimal mortality rates of 0–1 per cent.9,10,15,17 Furthermore, the extent of resection does not influence long-term survival.9,10,16,18 Surgery is the only modality that offers a chance for long-term survival.9,10,15,19 Surgery does entail a significant morbidity rate, but unrelated to the extent of resection.10 Morbidity rates from the recent larger series are in the range 24–27 per cent.9,10,17

08/02/13 5:34 PM

204  Assessment and management of recurrence

Bedrosian and colleagues suggest there is a biological variance between truly locally recurrent rectal cancers and those that disseminate.15 This is supported by Wong and colleagues,20 but their analysis was of colonic cancer. They concluded there are two types of colonic cancers – locally active cancers and cancers that have the tendency to metastasize. Approximately 50 per cent of patients with local pelvic recurrence do not have distant disease.21–23 Approximately two-thirds of this group can have a curative surgical resection.10,14 These patients with R0 pathology have 5-year survival rates of up to 49 per cent.10 The encouraging results of the more recent larger series have fuelled the growing interest in surgical resection for LRRC. This chapter defines and outlines the principles of assessment and surgical management of LRRC.

Presentation Patients usually present with symptoms that reflect the location of the recurrence.4,5,9 Symptoms include pelvic pain or referred pain, per rectal bleeding and discharge, altered bowel habit and tenesmus; the most common is pain. Pain is caused by local invasion of bone, the sacrum or pelvic sidewall, or invasion into nerves, sacral nerve roots or the sciatic nerve. Referred pain is in the distribution of the involved sciatic nerve. Asymptomatic patients are identified during routine colorectal cancer follow-up. Pelvic recurrences may be detected during a digital rectal or vaginal examination, colonoscopy or routine follow-up computed tomography (CT) scan. A rising trend in the carcinoembryonic antigen (CEA) should raise the suspicion for recurrence in an asymptomatic patient. This will initiate a cascade of investigations. In a Cochrane review published in 2007, intensive follow-up was found to be associated with improved overall survival and earlier detection of recurrence, suggesting that earlier detection may permit earlier intervention leading to improved survival.24 Locally recurrent rectal cancer most commonly recurs within the first 2–3 years after seemingly curative resection.5,24 In a study by Wanebo and colleagues, 28 per cent of patients presented within 1 year after initial resection, while 36 per cent, 17 per cent and 15 per cent presented within 2 years,

HEBK001-C13_p203-221.indd 204

3 years and more than 3 years after initial surgery, respectively.25 In a study by Heriot and colleagues, time to recurrence was within a year in 28 per cent of patients, in the second year in 29.5 per cent, in the third year in 18 per cent and more than 3 years after surgery in 24.5 per cent.10 In a large series from the Mayo Clinic, 19 per cent, 26 per cent and 55 per cent of patients presented within 1 year, 2 years and after 2 years, respectively.9

Preoperative assessment and preparation for surgery All patients should have localized pelvic disease confirmed by CT, magnetic resonance imaging (MRI) and positron-emission tomography (PET) before consideration of proceeding to surgery. Although CT and PET imaging give vital information regarding metastatic disease, MRI is most useful in determining resectability and meticulous planning of the surgical approach. The current consensus amongst the institutions performing pelvic exenteration is to perform all three scans. With the emergence and popularity of PET/CT, routine standard CT scans can be avoided. Less than 5 per cent of pelvic exenterations are said to be irresectable at the time of surgery after preoperative clinical and radiological assessment including CT, MRI and PET.4

Clinical Assessment Clinical assessment with an examination under general anaesthetic is frequently performed before pelvic exenteration. Examination without anaesthesia is invariably painful and unpleasant for the patient. Tumour fixation and involvement of other pelvic organs can be assessed. Additionally, biopsy for tissue confirmation of recurrence, cystoscopy and transanal ultrasound can all be performed without causing pain to the patient.

Carcinogenic Embryonic Antigen A rise in CEA prompts investigation for recurrence, although it does not always rise with recurrence and does not help to determine tumour resectability.26

08/02/13 5:34 PM

Preoperative assessment and preparation for surgery  205

The specificity is up to 84 per cent, but sensitivity is poor, at 59 per cent. The CEA level may provide prognostic information, as some studies have demonstrated that an elevated CEA is associated with worse survival in recurrent pelvic disease.27,28

Colonoscopy Colonoscopy should be performed to biopsy intraluminal recurrent tumour and exclude synchronous colonic malignancies. Anastomotic recurrences are generally discovered during postoperative surveillance colonoscopies.

Computed Tomography A CT scan of the chest, abdomen and pelvis is performed routinely and is the first step for staging recurrence before pelvic exenteration. In recent years, PET scans have become more precise in locating the recurrence, and the incorporation of CT in the form of a PET/CT scan is optimal.

Pelvic Magnetic Resonance Imaging Magnetic resonance imaging is an essential and optimal mechanism for locoregional staging of the recurrence and for planning pelvic exenteration. An MRI scan is the best imaging modality available for soft tissue, lymph nodes and tumour definition. The images are superior to those of a CT scan. The MRI scan provides accurate assessment of the tumour, and extent of involvement of adjacent pelvic viscera, including invasion into bone, nerve, vessels and ligaments. A significant disadvantage of MRI is the difficulty in distinguishing between recurrent disease and post-radiotherapy or post-surgical fibrosis and inflammation;29 the addition of a PET scan, or even better a PET/CT scan, complements MRI to help distinguish between the two.30–32 In an early study by Popovic and colleagues, MRI was found to have an accuracy of 83 per cent in determining patients’ eligibility for pelvic exenteration.33 The accuracy of MRI will continue to improve with future developments in technology and the implementation of thinner MRI sections.34

HEBK001-C13_p203-221.indd 205

Positron-Emission Tomography Positron-emission tomography scans have been shown to be highly accurate in the detection of disseminated disease and have been shown to alter the management of 20–40 per cent of patients by demonstrating distant disease otherwise undetected on conventional imaging.26,35 One of the major limitations of CT and PET scans, however, is that peritoneal disease may go undetected until the time of surgery. Furthermore, PET scans are less accurate for the mucinous variant of adenocarcinoma and can be false positive with any inflammatory condition.

Chemotherapy and Radiotherapy Before surgery, patients not previously exposed to radiotherapy would usually receive long-course radiotherapy with a chemosensitizer. In an era where radiotherapy is used increasingly, most patients with recurrent rectal cancer are likely to have already been exposed to radiotherapy; moreover, most patients will have received the maximum permissible dose. To circumvent the issue of tissue toxicity, intraoperative radiotherapy (IORT) has been developed with the ability to deliver a boost of up to 10 Gy directly to the tumour bed, which has a biological activity of twoto threefold that of external-beam radiotherapy.36 Intraoperative radiotherapy can be performed as either intraoperative electron-beam radiotherapy or intraoperative high-dose-rate brachytherapy.37,38 Combined with tissue shielding, toxicity to the surrounding tissue can be effectively minimized; therefore, IORT is feasible in previously irradiated patients.10,37–40 This has been advocated for R2 resections at the time of surgery, but unfortunately it is not readily available in all units. Although some authors have demonstrated significant survival benefit with IORT, it does not replace inadequate surgery.40,41 Intraoperative radiotherapy becomes less significant with the advancements in pelvic exenteration, in particular with the encouraging results of increasing proportion of complete tumour clearance rates (R0) for laterally recurrent cancers using the technique of en-bloc resection of the internal iliac vasculature and bone.42 Intraoperative radiotherapy will therefore be beneficial for the scenario where resection was proposed as potentially complete but intraoperatively inadvertent dissection through the tumour occurs.10

08/02/13 5:34 PM

206  Assessment and management of recurrence

Relative and absolute contraindications for pelvic exenteration Many authors have reported absolute and relative contraindications for pelvic exenteration (Table 13.1). Pelvic exenteration surgery has advanced considerably over the past decade, pushing the boundaries and redefining the extent of surgical techniques. As a consequence, the lists of contraindications are obsolete and continuously being redefined, and the number of absolute contraindications is diminishing. Nonetheless, some of the absolute contraindications have remained; these include patient performance, multiple distant metastases, including para-aortic lymph node disease, and sacral recurrences necessitating resection of the whole S1 vertebral body. Indications for curative pelvic exenteration surgery depend on patient factors and the extent of disease. Due to the high morbidity rates and extent of physiological insult induced during and after pelvic exenterations, the patient must be relatively fit (with an American Society of Anesthesiologists (ASA) score of 3 or less) and have tumour recurrence localized to the pelvis without distant metastases, unless distant metastases are resectable with curative intent. Pelvic exenteration with curative intent has been performed in patients with limited visceral metastases with accept-

able results.44 They include patients with unilateral liver metastases or solitary pulmonary recurrence. Nonetheless, these represent only a small group of highly selected patients. Lateral pelvic recurrences are no longer an absolute contraindication, as reported by some encouraging results.42 Extended resection laterally is possible, and extended lateral resections with lateral pelvic bone and ligament resections have been performed. This provides access to, and the ability to resect, some recurrences that extend to and through the greater sciatic foramen. Lateral nerve involvement is not an absolute contraindication, because as long as the lumbosacral nerve and S1 nerve roots are preserved adequate lower limb motor function can be preserved. External iliac vessel encasement is not an absolute contraindication if the vessel can be reconstructed after obtaining an R0 resection. Lower limb oedema is also not an absolute contraindication. Similarly, ureteric involvement is not a contraindication unless curative resection was deemed impossible preoperatively. Preoperative decisions and preparations for unilateral nephrectomy, ureterectomy and reimplantation or, if necessary, radical cystectomy and conduit need to be planned. Essentially, if there is inoperable metastatic disease or an oncological resection is not possible with

Table 13.1. Absolute and relative contraindications for pelvic exenteration. Study Boyle et al.

4

Absolute Contraindications

Relative Contraindications

Encasement of external iliac vessels

Distant metastases

Extension of tumour through the sciatic notch

Primary stage IV disease

Presence of lower limb oedema from lymphatic or venous

Extensive pelvic side-wall involvement

obstruction Poor performance status

Inability to achieve R0 resection Sacral invasion above S2/3

Pawlik et al.

43

Distant metastases

Ureteral obstruction

Involvement of common or external iliac vessels

Significant medical comorbidity

Para-aortic lymph node metastases

Poor performance status/inability to care for stomas

Involvement of sacrum above S1 Tumour extension through sciatic foramen Pelvic side-wall involvement Ogunbiyi et al.6

Tumour invading above S2

Distant metastases

Involvement of pelvic side-wall or pelvic nerves

Diffuse intra-abdominal nodal metastases

Involvement of ureters or presence of hydronephrosis on imaging

HEBK001-C13_p203-221.indd 206

08/02/13 5:34 PM

Anatomy of exenteration surgery  207

a clear pelvic margin, then this would be the major contraindication for pelvic exenteration. Metastatic disease no longer automatically precludes patients from pelvic exenteration. There is some evidence to suggest that patients with limited and resectable metastasis can survive, and some even beyond 5 years. 41,44–49 Nevertheless, these patients are a highly select group, and pelvic exenteration was performed after considerable deliberation and debate. Similar to the situation in primary colorectal cancers, there is no consensus on the optimal sequence of surgery. Some surgeons advocate staged procedures while others perform synchronous resections.41,44–50 Rarely, pelvic exenteration can be performed for palliation. These cases include uncontrollable malignant masses with small and large bowel to vesical, vaginal and or cutaneous fistulae, unmanageable malignant cutaneous and vaginal wounds, and intractable sciatic nerve or pelvic soft tissue pain. In addition to the indications and contraindications for surgery, quality-of-life implications and patient choice are significant factors to be considered before informed consent.

Multidisciplinary team planning and management All decisions regarding the management of patients considered for pelvic exenteration surgery should be made in a multidisciplinary setting with all involved relevant medical and surgical specialties. The multidisciplinary team meetings should optimally involve not only the various surgical, perioperative anaesthesiology and oncological specialties but also allied health specialists, including stoma therapy, rehabilitation, nutrition and psycho-oncology support and expertise.

Classification Several classification systems have been proposed. Anatomical classifications are most commonly used, although none is universally accepted. Suzuki and colleagues from the Mayo Clinic published a classification system based on symptoms and degree of fixation (Table 13.2).51 Although this study did not reveal any association between survival, symptomatic recurrence and points of fixation, a sub-

HEBK001-C13_p203-221.indd 207

sequent larger study by Hahnloser and colleagues from the Mayo Clinic found recurrence with pain and increased fixity of the tumour to be associated with greater operative challenges, difficulty with delivering intraoperative radiotherapy and significantly reduced survival.9 In 1999, Wanebo and colleagues proposed a classification system based on the Union for International Cancer Control (UICC) TNM staging system (Table 13.2), where the tumour recurrence stage increased depending on the depth of invasion through the bowel wall and beyond.25 The Memorial Sloan Kettering group, Leeds group and Yamada and colleagues described classification systems based on the anatomical location of local recurrence. In the classification system by the Memorial Sloan Kettering group, recurrence was classified as central including anastomotic, perirectal and mesorectal recurrences, anterior, posterior or lateral.52 The Leeds group classified recurrence as central, sacral, side-wall or composite, depending on the structures involved (Table 13.2).4 Yamada and colleagues classified recurrence into localized (central), lateral or sacral recurrences;28 in their study, lateral recurrences were associated with a significantly lower overall survival. Classification of LRRC is important for comparative and management purposes, but the sites or number of points of fixation is significantly important for determining clear margins and therefore prognosis.

Anatomy of exenteration surgery Patients who present with localized recurrent pelvic cancer are a heterogeneous group in terms of the involved pelvic structures; as such, the definition of extent of resection is debatable. As a result, there is no standard defined surgical procedure performed, but instead the type of operation is dependent on the site of the tumour, the size of the tumour and the number of organs involved. This means that a number of different pelvic organs, vasculature, muscles, ligaments, nerves and pelvic bone components are excised. To understand the operative approaches, the pelvis can be divided into four main compartments (Figure 13.1):

08/02/13 5:34 PM

208  Assessment and management of recurrence Table 13.2. Comparison between different classification systems. Study

Description

Suzuki et al.51

Symptoms   S0 Asymptomatic   S1 Symptomatic without pain   S2 Symptomatic with pain Degree of fixation  F0 Not fixed  F1 Single-point fixation  F2 Two-point fixation  F3 Three or more points of fixation

Wanebo et al.

25

TR1 Local recurrence at local excision site/anastomosis with invasion of submucosa TR2 Local recurrence at local excision site/anastomosis with full invasion of muscularis propria TR3 Local recurrence at anastomosis beyond muscularis with involvement of perirectal soft tissue TR4 Local/focally extensive invasion of rectum or other organs – anteriorly into vagina, uterus, prostate, bladder or   seminal vesicles, posteriorly into presacral tissue (tethered, not fixed) TR5 Extensive invasion of pelvis – bony/ligamentous pelvis

Guillem et al.52

Central – including mesorectal, anastomotic, perirectal recurrences or perineal recurrences after   abdominoperineal resection Anterior – uterine, vaginal, prostate, bladder, seminal vesicles Posterior – sacrum and presacral fascia Lateral – side-wall soft tissue or bony pelvis

Boyle et al.4

Central – tumour confined to pelvic organs without involvement or invasion into bone Sacral – tumour in presacral space and abuts or invades sacrum Side-wall – tumour involving structures on pelvic side-wall, including greater sciatic foramen, sciatic nerve,   through to piriformis and gluteal region Composite – simultaneous involvement of sacrum and pelvic side-wall

Yamada et al.28

Localized – localized to adjacent pelvic organs and connective tissue Sacral invasive – recurrent tumour that invades lower sacrum (S3–5), coccyx and periosteum Laterally invasive – tumour involving sciatic nerve, greater sciatic foramina, lateral pelvic wall or upper sacrum (S1–2)

Anterior compartment: consists of the bladder, prostate, seminal vesicles, vas deferens, urethra, urogenital diaphragm, dorsal vein complex, obturator internus and externus muscle, anterior half of the vagina, anterior pelvic floor muscle (pubococcygeus and puborectalis part of the levator ani), and pelvic bone (pubic symphysis, superior and inferior pubic rami). l Axial compartment: consists of the posterior half of the vagina, uterus, ovaries, fallopian tubes, broad ligament, round ligament of uterus, rectum, pelvic floor muscle (iliococcygeus part of levator ani), lower sacrum (S4 down) and coccyx. l Posterior compartment: consists of the rectum, pelvic floor (coccygeus muscle), internal iliac vessel branches and tributaries, piriformis muscle, sacral l

HEBK001-C13_p203-221.indd 208

nerves S1–S4, pelvic bone (sacrum and coccyx), anterior sacrococcygeal ligament, medial sacrotuberous and sacrospinous ligaments. l Lateral compartment: consists of the pelvic side-wall structures, ureter, internal iliac vessels, external iliac vessels, piriformis and obturator internus muscle around the ischial spine, coccygeus muscle, lateral sacrotuberous and sacrospinous ligaments attached to ischium, ischium including tuberosity and spine, lumbosacral trunk and sciatic nerve distal to ischial spine and obturator nerves and vessels. In general, the four compartments can be best understood by their central points as there is some degree of overlap of their peripheries. The central

08/02/13 5:34 PM

Surgical technique  209

Figure 13.1. Anatomical compartments of the pelvis. (Reproduced from O’Connell R, Madoff R, Solomon M (eds). Operative Surgery of the Colon, Rectum and Anus, 6th edn. London, Hodder Arnold, 2013, with permission.)

axis of the anterior compartment is the urethra, for the axial compartment it is the tip of the coccyx, for the posterior compartment it is the third sacral vertebra, and for the lateral compartment it is the ischial spine. In view of the heterogeneity in the types of resection, these are best defined as either partial or complete pelvic exenterations. A complete pelvic exenteration is defined as removal of the primary or recurrent tumour (with or without attached bone) with all remaining pelvic viscera – that is, all four anatomical components of the pelvis. Partial pelvic exenteration is defined as removal of the primary or recurrent tumour (with or without attached bone) with en-bloc resection of up to three anatomical components of the pelvis. Pelvic exenteration always involves an abdominal approach, usually with a perineal completion phase that can be done in the lithotomy or prone position. The anterior, axial and lateral compartments are best done through an abdominal combined with a perineal lithotomy approach. Posteriorly, resection of the sacrum from the fourth sacral vertebra (S4) down and the sacrospinous ligaments allows radical excision of posterior pelvic floor that is approached from the abdominal side and is often better visualized than in the prone position. Involvement of the third sacral vertebrae (S3), and above, because of the nature of the sacroiliac joint attachment, requires a prone approach unless resection involves only the anterior cortex of the midline bones up to the fifth lumbar vertebra

HEBK001-C13_p203-221.indd 209

(L5) and upper sacrum, as this can be done abdominally. Lateral higher sacrum and full vertebral excision of S2 and S3 require the posterior prone approach. Depending on the number and type of pelvic organs involved in the malignant process, the procedure requires a multidisciplinary team of highly skilled consultant surgeons from the surgical specialties of colorectal surgery, vascular surgery, urology, orthopaedics, and plastics and reconstructive surgery. At our institution, colorectal surgeons predominantly perform the surgery, with other surgical disciplines being involved at the appropriate time. Specialist anaesthetist and experienced theatre nursing staff are also required, as the procedure can take up to 8 to 20 hours.

Surgical technique The aim of surgery is to achieve an R0 resection. This is the principle predictor of long-term survival. The resection depends on the location of the recurrence in the pelvis and whether the surrounding organs and structures are involved. The pelvis is bounded by bone and, if necessary, the bone is resected as part of the en-bloc specimen. This should be performed only if stability of the pelvis can be preserved. The planes for exenteration are ill-defined due to the previous rectal dissection, quite commonly having been a total mesorectal excision (TME). There will inevitably be fibrotic tissue from previous surgery and radiotherapy, which can mimic malignancy. The surgeon must be aware of the necessity to obtain clear margins on the resected specimen and must dissect laterally to obtain wider circumferential margins. Involved lateral structures (ureter, nerve, vessels) are resected, often with bony margins if needed. The pathologist should be aware of the anatomical differences between primary resection and pelvic exenteration. Circumferential, proximal and distal margins in numerical values become less significant; instead, whether each margin is free of malignancy is of primary importance (i.e. R0, R1, or R2). The principles of surgical oncology are essential for curative intent. To achieve tumour-free resection margins (R0 resection), resection must be en bloc. Moore and colleagues observed that central or

08/02/13 5:34 PM

210  Assessment and management of recurrence

anterior pelvic recurrences were more likely to have R0 resections than other locations, especially lateral recurrences.53 Other institutions have shown that lateral pelvic side-wall recurrence is a significant negative prognostic indicator for survival.10,28 Heriot and colleagues classified pelvic exenteration into radical and extended radical excisions.10 Radical resection is confined to the recurrence involving the pelvis and neorectum. Radical resections can be appropriate for central recurrences. Predominantly, a radical resection requires an abdominoperineal excision but, a redo anterior resection or Hartmann’s procedure have been performed for selected cases.4,10,17,54 Extended radical resection involves at least one adjacent pelvic organ (bladder, prostate, seminal vesicles, uterus, ovaries, vagina, ureter, iliac vessels, small bowel, sacrum). The subgroups of anterior (bladder, prostate, seminal vesicles, uterus, vagina, ovaries), posterior (sacrum), anteroposterior, and lateral (internal iliac vessels, ureter) extended radical resection are defined by the organs and structures resected. Urinary tract involvement is common and can require complex urological interventions. Brunschwig described the first pelvic exenteration with ureteral diversion into the colon for locally advanced pelvic cancer in 1948.55 Wanebo, a pioneer and advocate for surgery for LRRC, has published extensively on abdominosacral resections.3,25,56,57 He describes a two-stage procedure. The first stage is performed in the modified Lloyd Davies position; the second stage, 2 days later, is done in the prone position. An extensive nodal dissection is performed from the distal aorta to the pelvic side-walls. The current contentious issues revolve around the traditional anatomical limits of resecting with a clear margin, such as involvement of any bone other than S2 down, and major vascular or lumbosacral, sciatic or femoral nerve involvement. This is in contrast to previous issues, when consideration of any surgical resection for recurrence was contentious. An understanding of outcomes both with and without exenteration, including long-term survival data, operative morbidity and mortality, length of hospital stay, length of rehabilitation and quality of life, must be considered and discussed in detail. Historically, the mean hospital stay was 5 weeks,16 but now the average is closer to 3 weeks, with recovery taking 3–6 months before a stable quality of life is achieved. Surgery for LRRC has advanced significantly in the past decade. New boundaries have been defined.

HEBK001-C13_p203-221.indd 210

Until recently, a lateral pelvic wall recurrence reflected a negative prognostic factor with consistently poor survival rates. Austin and Solomon reported R0 rates as high as 53 per cent for lateral recurrences. Pelvic exenteration included the excision of the lateral vasculature (internal iliac vessels) to achieve clear margins. The overall survival rate after local recurrence surgery was 69 per cent and disease-free survival was 72 per cent at a mean follow-up of 19 months.42

Classification of Operations The magnitude and type of exenteration can be broadly described in five approaches (the letters in the list below reflect the planes in Figures 13.2–13.5): A anterior plane for total pelvic exenteration or anterior pelvic exenteration; B anterior plane for axial pelvic exenteration with subtotal vaginectomy; in male patients, this includes B and the perineal part of A; C anterior plane for axial pelvic exenteration with posterior vaginectomy (in female patients only); D posterior plane for total pelvic exenteration or axial pelvic exenteration; E posterior plane for abdominosacral exenteration.

E

D

CB

A

Figure 13.2. Resection planes in pelvic exenteration (female). Reproduced from O’Connell R, Madoff R, Solomon M (eds). Operative Surgery of the Colon, Rectum and Anus, 6th edn. London, Hodder Arnold, 2013, with permission.

08/02/13 5:34 PM

Surgical technique  211 C

B

A

E

D A

Figure 13.3. Resection planes in pelvic exenteration (male). Reproduced from O’Connell R, Madoff R, Solomon M (eds). Operative Surgery of the Colon, Rectum and Anus, 6th edn. London, Hodder Arnold, 2013, with permission.

In broad principles the resection margin for the involved compartment of the pelvis is the soft tissue/ bone attachment or en-bloc excision of the involved bone. Attempting to obtain soft tissue margins of

A

R

Figure 13.5. Lateral pelvic resection. Reproduced from O’Connell R, Madoff R, Solomon M (eds). Operative Surgery of the Colon, Rectum and Anus, 6th edn. London, Hodder Arnold, 2013, with permission.

involved compartments will result in unacceptably high rates of margin positivity (R1 and R2).

Pelvic Exenteration: Abdominal Phase Perioperative preparations

Figure 13.4. Resection planes in pelvic exenteration (axial view, female). Reproduced from O’Connell R, Madoff R, Solomon M (eds). Operative Surgery of the Colon, Rectum and Anus, 6th edn. London, Hodder Arnold, 2013, with permission.

HEBK001-C13_p203-221.indd 211

Stoma sites are marked. If necessary, the elliptical cutaneous incision for the vertical rectus myocutaneous flap is marked with a surgical skin marker. If the patient already has a colostomy, this is covered with a swab and impervious plastic dressing so that it is not removed during the lengthy procedure. Bowel preparation is usually advisable, as the long duration of the exenteration and previous radiotherapy may contribute to a common requirement to repair damaged bowel. Bowel preparation can be omitted in patients with pre-existing stomas, but the colon should be prepared if a colostomy is to be converted into a colonic conduit. The insertion of ureteric stents before the laparotomy is especially useful for central exenterations where the bladder is to be preserved but is worth considering in all patients.

08/02/13 5:34 PM

212  Assessment and management of recurrence

Positioning

The patient is placed directly on a gel mattress to secure their position and prevent unexpected slippage during steep Trendelenburg positioning. The abdominal phase of pelvic exenteration is performed in the modified Lloyd Davies position, with both arms secured at the patient’s side. The perineal phase can also be performed in the modified Lloyd Davies or prone position. Sacrectomy above S3 is performed in the prone position. Partial anterior sacrectomy, or a distal sacrectomy below S3, can be performed from a transabdominal approach. The anus, if present, is sutured, preventing soiling during the operation. The vagina is also included in the skin preparation. When draping the patient, exposure should include the groin and thigh in preparation for vein harvest if vascular reconstruction is necessary with an interposition graft or patch. Laparotomy

Despite resectability having been predicted preoperatively by CT, MRI and PET scans in a multidisciplinary setting, the initial role of surgery is to exclude metastatic peritoneal disease that may not have been detected by preoperative testing. A thorough laparotomy and adhesiolysis is performed meticulously and cautiously to prevent enterotomies in radiotherapydamaged small bowel. If negative for small-volume missed peritoneal disease, then the planned exenteration surgery begins. Of note, lateral pelvic side-wall involvement and bone fixity are not contraindications for pelvic exenteration surgery, although these features are best assessed by MRI preoperatively rather than by palpation at laparotomy. Any nodule suspicious for peritoneal metastasis or hard tissue suspicious for carcinoma encountered during the pelvic dissection should be sent for frozen section. Confirmation of malignancy may change the course of the operation. Preparing the pelvis

Adhesiolysis is necessary for a re-laparotomy. The small bowel is mobilized out of the pelvis; the small bowel is often adherent to the tumour in recurrent disease, however, and requires en-bloc resection. The involved segments of small bowel are disconnected from the rest of the small bowel with a linear gastrointestinal stapling and cutting device. The appendix, if still in situ, is usually removed; this is to avoid potential future difficulty with abdominal access after exenteration with combinations of stomas,

HEBK001-C13_p203-221.indd 212

abdominal wall mesh reconstruction, and surgery often involving myocutaneous flaps. An open appendicectomy post exenteration has the difficulty of access as well as the potential risk of damaging the inferior epigastric vessels which will compromise. The ureters are identified, mobilized with adequate connective tissue to avoid damage to the ureteric blood supply, and protected with vessel loops. The ureters may have been displaced, usually laterally, due to previous pelvic dissection. The gonadal vessels are preserved if possible in male patients, but they are ligated in female patients during radical hysterectomy. Preservation of the abdominal portion of the gonadal vessels is important for the arterial blood supply of the ureters. The pelvic portions of the ureters are transected when an ileal conduit is planned and performed. Lateral pelvic wall lymph node dissection is routine with pelvic exenteration. The lymphadenectomy includes the nodes of the common iliac bifurcation, internal iliac chain and obturator nodes. In pelvic exenteration with sacrectomy, a metallic pin is inserted into the sacrum above the level of sacral involvement. An image intensifier can locate this pin in the prone position, thus ensuring the precision of the level of sacral transection.

Perineal Phase The perineal phase is performed via a wide lithotomy exposure, in the prone position or sequentially after prior lithotomy to widely resect the anterior compartment and then in the prone position for the posterior compartment. The prone phase is for the sacrectomy, if needed. The sacrectomy is the final step for the complete resection of the recurrence. The en-bloc specimen with the transected sacrum is delivered and the defect is closed, commonly with a myocutaneous vertical rectus abdominus muscle (VRAM) flap.

Different Surgical Approaches to Pelvic Exenteration Pelvic exenteration comprises a heterogeneous group of surgical operations commonly involving a combination of the different subcategories. This is to achieve an R0 resection, since resection margin is

08/02/13 5:34 PM

Surgical technique  213

the single most significant prognostic factor determining long-term survival. The different surgical approaches for pelvic exenteration have been classified into subgroups according to the anatomical classification of LRRC. Overlap between the groups is common, and inevitable in most cases, due to the nature of LRRC. Central (axial) pelvic exenteration

Central or axial recurrences represent anastomotic, perirectal and mesorectal recurrences without lateral pelvic wall involvement. This is surgery for central LRRC using the Memorial Sloan Kettering definition (see Table 13.2). Central recurrences can occur at the anastomosis without extraluminal spread. In such a case, redo low anterior resection (including the previous anastomosis) or abdominoperineal excision of the rectum can be performed without resection of any other structures (see Figure 13.3, planes D and C plus pelvic part of A). Heriot and colleagues reported that 63 of 160 patients (39 per cent) in their series had central recurrence.10 Unfortunately, redo anterior resection is uncommon and reported at 5–10 per cent.9,10 Abdominoperineal excision is more common and in the region of 26 per cent.10 The usual scenario is involvement of other organs and structures, and thus an extended resection is required. In female patients, axial pelvic exenteration involves en-bloc resection of the recurrence and hysterectomy with vaginectomy. The bladder and anterior wall of the vagina can usually be preserved (see Figure 13.2, planes D and C) as the uterus and posterior wall of the vagina provide protection. If these organs are involved, then total pelvic exenteration is performed (see Figure 13.2, planes D and A). In patients in whom the dome of the bladder is involved, wedge resection with primary closure is performed. If a unilateral distal ureter is involved, then resection of the ureter with reimplantation with a Boari flap or psoas hitch can be used. The TME plane no longer exists and therefore obtaining a clear circumferential margin becomes a great challenge. Additionally, the induration after radiotherapy usually adds considerably to the difficulty. Total pelvic exenteration

Total pelvic exenteration is an extension of central pelvic exenteration that includes the anterior compartment organs, namely the bladder and adjacent

HEBK001-C13_p203-221.indd 213

reproductive organs – that is, surgery for anterior and central LRRC (see Table 13.2). Additionally, lateral pelvic exenteration may also be required as part of total pelvic exenteration (see Lateral pelvic exenteration below). Total pelvic exenteration in male patients involves en-bloc resection of the anterior compartment organs (bladder, prostate and seminal vesicles), the rectum and previous anterior resection anastomosis with the recurrence (see Figure 13.3, planes D and A). This incorporates an abdominoperineal excision with radical cystoprostatectomy and formation of a urostomy. In female patients, the more extensive recurrence that invades beyond the protective barrier of the uterus and vagina, with involvement of the trigone of the bladder, warrants radical cystectomy and formation of a urostomy with total abdominal hysterectomy, bilateral salpingo-oopherectomy and abdominoperineal excision. The posterior plane (plane D) is the same for central pelvic exenteration and total pelvic exenteration. There is no anatomical plane but rather ill-defined fibrotic tissue as a result of the previous rectal resection. Hence, the exenteration dissection plane may include the presacral fascia to achieve clear margins. If there is any doubt that there is no posterior plane and the sacrum may be involved, then there should be no hesitation to proceed to abdominosacral exenteration. At this stage, a frozen section of ‘suspicious presacral tissue for carcinoma’ should be analysed for histological confirmation. This scenario occurs when the tumour has progressed despite no previous evidence of sacral invasion on MRI scan and post-radiotherapy or the presacral tissue has been interpreted at imaging as ‘postoperative scar tissue’. Abdominosacral exenteration

Abdominosacral exenteration is a total exenteration with the addition of sacrectomy. This is usually necessary in patients with extensive posterior LRRC (see Table 13.2). Sacrectomy can be performed at the S1/S2 junction without pelvic instability. After completion of the abdominal phase with reconstruction and abdominal wall closure, the patient is transferred to the prone jack-knife position for sacrectomy. During the abdominal phase, a metallic pin is inserted into the sacrum at the level of the proposed

08/02/13 5:34 PM

214  Assessment and management of recurrence

transection. With the aid of an image intensifier, the pin is located while the patient is in the prone jack-knife position to confirm the precise level of the sacrectomy. Sacrectomy does not have to be a complete resection. In selected patients an anterior cortical or partial resection of the sacrum can be performed. This is best performed in the abdominal phase. Additionally, distal sacrectomy, at the S4/S5 junction and below, can be performed during the abdominal phase, thus avoiding the necessity for the prone position. Before sacrectomy, especially proximal sacrectomy (S3/S4 junction and above), internal iliac vessel ligature, referred to as ‘pelvic devascularization’, is recommended to minimize pelvic haemorrhage. Ligation of the internal iliac vessels inherently involves the dissection of the lateral pelvic lymph nodes, starting from the common iliac lymph chain and down to the obturator lymph nodes. Bleeding from inadvertent injury of the internal iliac vessels, especially the veins, can be profuse and has led to uncontrollable and fatal haemorrhage.42 Arterial ligation distal to the first branch of the internal iliac artery is preferred to promote skin and muscle flap healing.2 The use of vessel loops of different colours helps to isolate the different structures and for proximal and distal vascular control. During sacrectomy, especially proximal sacral transection, sacral nerves are sacrificed. The ligation of the proximal sacral nerves (S1, S2) leads to significant morbidity. In contrast, the lower sacral nerves are relatively less important and adequate limb function is still achievable. The lumbosacral and S1 nerve root must be preserved to maintain motor function of the lower limb on that side. Gait without a foot-drop is preserved if the majority of the sciatic nerve is preserved. Ligation of S2 and lower sacral nerves will result in an atonic bladder. The bladder in a patient with recurrent rectal cancer has commonly been damaged previously by radiotherapy and prior surgery. Therefore, sacrectomy above S4 may require a radical cystectomy, even if there is no invasion into the bladder, prostate or seminal vesicles. The alternatives of self-catheterization and permanent suprapubic urinary catheterization should also be considered. Lateral pelvic exenteration

Previously, lateral side-wall recurrence with ureteric, neural or pelvic vascular involvement were contraindications for pelvic exenteration.4,9,10,15–17,22,28,51,57

HEBK001-C13_p203-221.indd 214

These limitations have been redefined. Over the past decade, the Royal Prince Alfred Group in Sydney, Australia has implemented a surgical approach for lateral pelvic wall recurrences with encouraging R0 resection rates and survival.42 Lateral pelvic exenteration is performed for lateral LRRC (see Table 13.2) and may be necessary in total pelvic exenteration. Internal iliac vessel ligation and resection is essential for lateral pelvic exenteration to ensure an R0 resection margin. The first step is to mobilize the pelvic ureters laterally and then ligate the internal iliac artery bilaterally. It is preferable, if possible, to ligate distal to the first branch of the internal iliac artery to maximize healing of the skin and muscles in the buttock region. The distal branches of the internal iliac vessels usually need to be ligated individually. Next, the venous system is ligated and resected en bloc with lateral tumours. This exposes the lumbosacral trunk and sacral nerve roots (S1, S2, S3) as they converge to form the sciatic nerve at the ischial spine. The piriformis muscle lies deep to the nerves. Lateral pelvic recurrences may involve sacral nerve roots and the sciatic nerve. If necessary, the sciatic nerve can be sacrificed to obtain an R0 resection, but at the cost of an altered gait and requiring an orthotic aid for the ankle joint. The ischial spine can be resected to gain maximal lateral exposure and reveal the sciatic nerve as it exits the pelvis via the greater sciatic foramen. Additionally, this helps free the lateral aspect of the sacrum by releasing the coccygeus muscle with the sacrospinous ligament. This is an important step and aids with the sacrectomy. Furthermore, the sacrotuberous ligament is exposed, allowing it to be transected if required. The Gigli saw is a useful tool to transect the spine. Tumour-encased external iliac vessels can be resected as part of the en-bloc specimen, with the assistance of a vascular surgeon. The vascular reconstruction results are most favourable with vein grafts using the contralateral femoral vein and are generally preferred over synthetic grafts. Where possible, arterial reconstruction is best performed immediately, rather than at completion of the exenteration, to avoid compartment syndrome. The encased external iliac vein may commonly be thrombosed and laden with clot. This should be identified during preoperative imaging and consideration given to an inferior vena caval filter

08/02/13 5:34 PM

Surgical technique  215

insertion before surgery to prevent propagation of the thrombus and pulmonary emboli. In chronic venous obstruction, reconstruction with an interposition venous graft is usually not necessary as collaterals are already established.

Reconstruction Urostomy: ileal and colonic urinary conduits

The ileum and colon have been used to create the urostomy.16,58 In one study, there was no difference in glomerular filtration rates or complications related to stenosis between the two types.58 Previously, ‘wet’ colostomies were performed to avoid double stomas (colostomy and urostomy). There are various methods for ureteric implantation. The most common techniques used include bilateral (Wallace) and single (Bricker) anastomoses. Ureteric stents are useful aids to promote healing and protect the anastomosis. Ureteric anastomotic leakage is a significant complication and may require temporary bilateral nephrostomies. The irradiated distal ileum and pelvic ureters increases the risk of anastomotic leaks. To minimize this risk, the abdominal rather than the pelvic ureter is used for the anastomosis. The blood supply of the ureters is protected by not ‘skeletalizing’ the ureters during ureterolysis and preserving the abdominal part of the gonadal vessels. The use of a colonic urostomy may be worth considering as this avoids the use of radiation-damaged distal ileum. Perineal flap reconstruction

The perineal defect may be closed primarily or with a pedicle flap. The most common types include the VRAM flap, gracilis myocutaneous flap and gluteal myocutaneous rotational flap.25 Closing the defect with flaps has decreased the rate of perineal wound complications, notably dehiscence. Shibata and colleagues showed a significant reduction in major complications of the perineal wound by closing the defect with a gracilis flap in patients undergoing proctectomy following preoperative radiation.59 Chessin and colleagues60 and Butler and colleagues61 reported that perineal closure with a rectus abdominis muscle flap post-irradiation significantly reduced major perineal complications, including dehiscence, compared with primary

HEBK001-C13_p203-221.indd 215

closure. Similarly, the Mayo Clinic recommended flap reconstruction over both primary closure and primary closure with pedicled omentoplasty.62 Perineal repair with a flap repair reduced wound-related complications and length of hospital stay. The pedicle flaps may also be used to reconstruct the vagina after posterior vaginectomy. Bell and colleagues reported a 20 per cent complication rate with vaginal flap reconstruction.63 A small proportion of patients will continue to have sexual intercourse after vaginal reconstruction.63

Unresectable Local Pelvic Recurrences Up to 50 per cent of local recurrences present without distant metastases and therefore are potentially amenable to resection.21–45 Patients eligible, and selected for pelvic exenteration may not proceed beyond a laparotomy due to the discovery of multifocal disease or peritoneal dissemination. Advancements in imaging modalities, especially PET/CT and MRI, have decreased the rate of unnecessary laparotomies where extensive disease is discovered and curative resection is not possible. Up to 5 per cent of patients do not proceed to pelvic exenteration due to previously undetected peritoneal disease or distant metastases detected intraoperatively.10

Factors Influencing Curative Resection Previously, LRRC has been deemed irresectable when there is extension into the sciatic notch, encasement of the external iliac vessels, extension to the sacral promontory, and the presence of lower limb oedema from lymphatic or venous obstruction.16 When iliac vessel involvement was determined intraoperatively, R0 resection was significantly less likely than when the vessels were not involved.53 Curative pelvic exenterations with transection of the involved sciatic nerve and resection of the external iliac vessels with reconstruction have been performed successfully, however. Radiological evidence of pelvic side-wall recurrence and hydroureter/hydronephrosis are significant factors that limit an R0 resection but are not contraindications for surgery. Involvement of the

08/02/13 5:34 PM

216  Assessment and management of recurrence

genitourinary organs, piriformis muscle, iliac vessels and sacrum do not influence margin status.53 If the recurrence is proximal to S1, however, most institutions would not aim for a curative resection.16,57 Sacrectomy above S2 has a high morbidity, including neuropathies and pelvic instability.

Table 13.3. Pelvic exenteration operative and admission detailsa. Median operating time

Complications The reported mortality rate has been low, ranging from 0.3 per cent to 5.4 per cent.4,9,10,15–17,25 In larger, more recent series, the mortality rate has been reported at less than 1 per cent.9,10,17 Occasional perioperative death may result from uncontrollable pelvic haemorrhage and cardiorespiratory events. The modern results are a significant improvement from the historically high mortality rates (10 per cent or more) of the pioneering era of pelvic exenteration.

HEBK001-C13_p203-221.indd 216

Median blood transfusion rate

6.6 units (range 0–17 units)

Mean number of specialties

3 (range 1–5)

  involved

Post-operation The patient should be transferred to the intensive care unit from the operating theatre for initial postoperative care. In general, patients who have had exenterations completed within 10 h can be extubated and transferred to the high dependency unit; patients whose surgery is prolonged over 10 h require continued intubation and ventilation. Early commencement of parenteral nutrition is essential. Prolonged ileus is a common complication following exenteration surgery, and therefore enteral nutrition is not recommended. Total parenteral nutrition is commenced on the first postoperative day. It is important to note that patients who undergo sacrectomy or VRAM flap reconstruction of the perineal or sacral defect should be nursed in a 308 lateral tilt position to prevent pressure on the surgical site. Upon stabilization, the patient can then be transferred to a ward with specialized nursing and allied health staff experienced in caring for patients with complex postoperative needs. Hospital stay can be prolonged and is usually dependent on postoperative complications. The median hospital stay for pelvic exenteration involving sacrectomy is 15 days (Table 13.3).64

9 h (range 3–16 h)

Median number of days in ICU

3 days (range 0–11 days)

Mean number of days in hospital

25 days (range 5–126 days)

Median number of days in

15 days

  hospital

Unpublished data from Royal Prince Alfred Hospital Pelvic Exenteration database, Sydney, Australia (n 5 240 cases).

a

ICU, intensive care unit.

The published morbidity rates range from 21 per cent to 72 per cent.1,49,51 Larger, more recent series from major institutions report morbidity rates of 23–27 per cent.9,10,17 The most common complications include perineal wound dehiscence, pelvic collections and cardiorespiratory problems. A more comprehensive list is shown in Table 13.4.

Prognosis Factors that affect survival have been extensively analysed and reported. Reported factors include margin status, site of pelvic recurrence, number of points of fixity, lymph node status at primary surgery, intraoperative radiotherapy, preoperative and postoperative adjuvant therapy, CEA levels, symptoms related to recurrence (especially pain), extent of resection, and operation type at primary resection.9,10,15–17,28,51,53,57 Statistically significant negative prognostic indicators are listed in Table 13.5. Resection margins have consistently been shown to be the most significant factor on survival.4,9,10,15–18 If microscopic clear margins are achieved, then survival is significantly improved compared with positive margins. R0 resection margins have been achieved in 38–84 per cent of patients.4,10,15–17 The 5-year overall survival rates for R0 resection range from 23 per cent to 49 per cent.9,10,16,17,57 If the margins are positive, however, then the 5-year survival decreases significantly to 0–23 per cent.9,10,15–17

08/02/13 5:34 PM

Prognosis  217 Table 13.4. Complications.

Table 13.5. Negative prognostic indicators for survival.

Perineal

Positive margin

  Wound dehiscence

Lateral pelvic side-wall recurrence

  Persistent perineal sinus

Number of fixed points in the pelvis

  Perineal flap necrosis

Positive lymph node at primary surgery

Pelvic collection

Elevated carcinoembryonic antigen level

  Wound infection

Presentation with pain

  Pelvic abscess/haematoma

Sciatic nerve involvement

Gastrointestinal   Prolonged ileus  Enterocutaneous fistula   Anastomotic leak   Bowel obstruction Cardiorespiratory/vascular   Atrial fibrillation   Myocardial infarction   Pneumonia  Deep venous thrombosis  Chronic lower limb oedema Urological  Urinary retention  Ileal conduit leak  Colovesical fistula Neurological   Sciatic nerve palsy  Obturator nerve palsy

An extensive pooled analysis of 1569 patients who underwent pelvic exenteration surgery was published in 2006.5 The overall 5-year survival rate was 30.7 per cent. Analysis of the specimens with clear margins (R0) showed an improvement in the 5-year survival rate up to 38.2 per cent. Heriot and colleagues reported similar survival rates from 160 patients across three different institutions.10 The overall 5-year survival was 36.6 per cent. The 5-year survival rate for negative resection margins (R0) was significantly higher when compared with those with involved margins. Overall, 61 per cent were R0 resections with a 5-year survival rate of 49 per cent, and 25 per cent were positive microscopically (R1) with a 5-year survival rate of 21 per cent. Macroscopic involvement in 9 per cent of patients resulted in a 5-year survival rate of 17 per cent. Similarly, the Mayo Clinic reported its series of 304 patients and demonstrated that positive resection margins

HEBK001-C13_p203-221.indd 217

(R1 or R2) had a lower 5-year survival rate at 16 per cent (R1, 22 per cent; R2, 15 per cent) when compared with a curative resection (R0, 37 per cent).9 These finding are consistent with the data from other exenteration units, including Leeds General Infirmary, the Memorial Sloan Kettering group and the Japanese reports (Table 13.6). The extent of resection did not significantly influence survival if R0 resection was achieved.9,10,16,18 This is consistent with the principles of surgical oncology. The ‘holy grail’ of exenteration surgery is thus an R0 resection. Patterns of pelvic recurrence have been studied by Yamada and colleagues.28 Five-year survival rates varied, depending on the original site of recurrence. Localized recurrence (i.e. recurrence involving adjacent organs) had a 5-year survival of 38 per cent, compared with 10 per cent in patients with sacral invasion and 0 per cent in patients with lateral recurrence. Lateral recurrence has been observed as a negative prognostic indicator by other authors.10,53 Lateral recurrences have inferior results because of the anatomical limitations determining resectability. Bone and major pelvic side-wall vessels limit the lateral extent of resection. In recent years, extended radical resections including the iliac vessels have been performed with encouraging results. Over the past decade, the Royal Prince Alfred Group in Sydney has implemented a surgical approach for lateral pelvic wall recurrences.42 The R0 resection rate for lateral pelvic recurrences was as high as 53 per cent, and the overall survival rate was 69 per cent at 19 months’ follow-up. R0 resection was associated with 46 per cent of patients remaining disease free, with an average disease-free interval of 30 months.3 An increasing number of fixed points of the recurrent cancer has been shown to have a poorer outcome.9,53

08/02/13 5:34 PM

218  Assessment and management of recurrence Table 13.6. Resection margins (R0–2) and 5-year survival rates. 5-year

5-year

Study

n

R0 (n [%])

survival (%)

Kusters et al.65

170

92 (54)

40

Heriot et al.10

160

98 (61)

Bell et al.

51

41 (80)

Bedrosian et al.15

134

65 (49)

30

Wiig et al.

150

66 (44)

27

Boyle et al.4

64

24 (38)b

Moore et al.53

119c

61 (51)

38 (31)

7 (7)

Moriya et al.

57

48 (84)

42

9 (16)d

Hahnloser et al.9

304

138 (48)

37

27 (6)

Yamada et al.

42

30 (71)

23

12 (29)

131

71 (54)

35

13 (10)

63

66

16

57

Salo et al.17

49

5-year

R1 (n [%])

survival (%)

R2 (n [%])

survival (%)

40 (25)

21

14 (9)

17

15

13 (10)

10 (20)

a

a

20 (15)

24 (38)b

9 (14)b

0 22 e

139 (46)

14

19 (15)

9

0 23

The difference after the sum of all the resections in those patients who had a laparotomy with curative intent but did not have a resection. a 4-year survival: R0 41%, R1 10%. b 3-year survival: R0 68%, R1 34%. c 18 recurrences from colonic cancers. d 9 positive margins. 3 had palliative intraoperative radiotherapy. 5-year survival was 0 for positive margins. All resections includes sacrectomies. e 12 palliative resections. All 42 patients had sacrectomies.

Some authors have concluded that an elevated CEA has a worse prognosis,15–17 but other reports are inconclusive.10,25

Survival Wanebo and colleagues reported their series in 1999.25 The overall 5-year survival rate was 31 per cent, with a 5-year disease-free survival rate of 23 per cent. More recent series report overall 5-year survival rates in the range 25–37 per cent, with a median overall survival of 31–43 months.9,10,15,16 The median cancer-specific survival has been reported at 48 months,10 with a cancer-specific 5-year survival rate in the range 36–41.5 per cent.10,15,16 The overall survival rate after a curative histological resection is in the range 23–49 per cent.2,4,9,10,15,17,18,53,57

Quality of Life The long-term quality of life in survivors of pelvic exenteration for recurrent and locally advanced primary rectal cancer is comparable to that in patients who have undergone primary rectal cancer resections. The

HEBK001-C13_p203-221.indd 218

quality of life survey performed by Austin and colleagues showed that patients suffered more in physical than mental terms following pelvic exenteration, but their overall quality of life was comparable to that in the general population.67 Miller and colleagues assessed quality of life and cost-effectiveness of therapy for LRRC; they concluded: … surgical resection was a cost-effective treatment for recurrent rectal cancer compared with non-operative and palliative strategies. Moreover, the procedure to elicit utility values showed that patients were more willing to gamble the risks of surgery and the possibility of developing pain or complications to have an opportunity to extend life than were health care providers. An incremental analysis was performed to combine the costs and benefits of each strategy. The incremental cost– utility ratio was $109 777 per QALY [qualityadjusted life-year] for the surgical resection compared with the non-operative therapy from the health care provider point of view, and $56  698 from the patient view. The diagnostic or palliative surgery strategy was always dominated by the non-operative option (less costly and more effective in terms of survival rate).13

08/02/13 5:34 PM

References  219

Following pelvic exenteration, patients report an improvement in their quality of life and a reduction in pain. In addition, their quality of life is better than in patients who receive medical palliative treatment.11–13

Re-recurrence A second local recurrence occurs in approximately a third of patients after initial surgery for local recurrence.68 Bedrosian and colleagues15 and GarciaAguilar and colleagues19 believe that LRRCs are biological variants with fewer tendencies to metastasize. Garcia-Aguilar and colleagues reported that 9 of 14 local recurrences did not metastasize.19 Bedrosian and colleagues showed that metastatic disease occurred earlier than the development of a second recurrence (9 months v. 16.5 months). Patients lived longer after the second local recurrence (41.5 months) compared with patients who developed metastatic disease (26.1 months).15 There was no significant difference in the rate of a second recurrence between an R0 or R1 resection from the initial local recurrence surgery.4 The authors stated that they were uncertain of the microscopic margin status of a second recurrence. Additionally, histological orientation and evaluation of the ‘true’ margins may be inaccurate, especially when the specimen becomes fragmented.4 Median survival is improved with R0 resection,54 but other series report no difference between R0 and R1 resections.17,66

Palliative surgery Consistent results have shown that a negative margin for malignancy is the key factor in survival – hence, the ethical debate about palliative surgery arises. Involved surgical margins after pelvic exenteration generally results in poor survival rates and poor quality of life,9,67 and therefore it is debatable as to whether it is justifiable to perform extensive surgery with considerable morbidity to prolong life for a short period. Patients with advanced recurrent pelvic disease commonly have intractable pelvic pain and malodorous perineal discharge associated with a significant decline in their quality of life. In this scenario, pelvic exenteration can be recommended only if surgery is for cure. Similarly, we recommend that

HEBK001-C13_p203-221.indd 219

patients with fistulating disease and small bowel obstruction have pelvic exenteration surgery only if cure is possible, otherwise a bowel bypass with anastomosis or a stoma should be considered in the palliative scenario. Patients with limited and resectable metastatic disease of the liver or lung may be considered for pelvic exenteration if there is potential for cure or at least disease control.

References   1. Bakx R, van Tinteren. Surgical treatment of locally recurrent rectal cancer. Eur J Surg Oncol 2004; 30: 857–63.   2. Gudderson LL, Sosin H. Areas of failure found at re-operation (second or symptomatic look) following ‘curative surgery’ for adenocarcinoma of the rectum: clinicopathologic correlation and implications for adjuvant therapy. Cancer 1974; 34: 1278–92.   3. Wanebo HJ, Koness RJ, Vezeridis MP, et al. Pelvic resection of recurrent rectal cancer. Ann Surg 1994; 220: 586–95.   4. Boyle K, Sagar P, Chalmers A, et al. Surgery for locally recurrent rectal cancer. Dis Colon Rectum 2005; 48: 929–37.   5. Heriot AG TP, Darzi A, Mackay J. Surgery for local recurrence of rectal cancer. Colorectal Dis 2006; 8: 733–47.   6. Ogunbiyi OA, McKenna K, Birnhaum EH, Fleshman JW, Kodner IJ. Aggressive surgical management of recurrent rectal cancer: is it worthwhile? Dis Colon Rectum 1997; 40: 150–5.   7. Ito K, Ohtsu A, Ishikura S, et al. Efficacy of chemoradiotherapy on pain relief on patients with intrapelvic recurrence of rectal cancer. Jpn J Clin Oncol 2003; 33: 180–5.   8. Wong CS, Cummings BJ, Brierley JD, et al. Treatment of locally recurrent rectal carcinoma: results and prognostic factors. Int J Radiat Oncol Biol Phys 1998; 40: 427–35.   9. Hahnloser D, Nelson H, Gunderson L, et al. Curative potential of multimodality therapy for locally recurrent rectal cancer. Ann Surg 2003; 237: 502–8. 10. Heriot AG, Byrne CM, Lee P, et al. Extended radical resection: the choice for locally recurrent rectal cancer. Dis Colon Rectum 2008; 51: 284–91. 11. Esnaola NF, Cantor SB, Johnson ML, et al. Pain and quality of life after treatment in patients with locally recurrent rectal cancer. J Clin Oncol 2002; 20: 4361–7. 12. Magrini S, Nelson H, Gunderson LL, et al. Sacropelvic resection and intraoperative electron irradiation in the management of recurrent anorectal cancer. Dis Colon Rectum 1996; 39: 1–9.

08/02/13 5:34 PM

220  Assessment and management of recurrence 13. Miller AR, Cantor SB, Peoples GE, et al. Quality of life and cost effectiveness analysis of therapy for locally recurrent rectal cancer. Dis Colon Rectum 2000; 43: 1695–701. 14. Kapiteijin E, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001; 345: 638–46. 15. Bedrosian I, Giacco G, Pederson L, et al. Outcome after curative resection for locally recurrent rectal cancer. Dis Colon Rectum 2006; 49: 175–82. 16. Moriya Y, Akasu T, Fujita S, Yamamoto S. Total pelvic exenteration with distal sacrectomy for fixed recurrent rectal cancer in the pelvis. Dis Colon Rectum 2004; 47: 2047–54. 17. Salo JC, Paty PB, Guillem J, et al. Surgical salvage of recurrent rectal carcinoma after curative resection: a 10 year experience. Ann Surg Oncol 1999; 6: 171–7. 18. Well BJ, Stotland P. Results of an aggressive approach to resection of locally recurrent rectal cancer. Ann Surg 2006; 14: 390–5. 19. Garcia-Aguilar J, Cormwell J, Marra C, et al. Treatment of the locally recurrent rectal cancer. Dis Colon Rectum 2001; 44: 1743–8. 20. Wong SKC, Jalaludin BB, Henderson CJA, et al. Direct tumour invasion in colon cancer: Correlation with tumour spread and survival. Dis Colon Rectum 2008; 51: 1331–8. 21. McDermott FT, Hughes ES, Pihl E, et al. Local recurrence after potentially curative resection for rectal cancer in a series of a 1008 patients. Br J Surg 1985; 72: 34–7. 22. Pilipshen SJ, Heilweil M, Quan SH, et al. Patterns of pelvic recurrence following definitive resections of rectal cancer. Cancer 1984; 53: 1354–62. 23. Rao AR, Kagan AR, Chan PM, et al. Patterns of recurrence following curative resection alone for adenocarcinoma of the rectum and sigmoid colon. Cancer 1981; 48: 1492–5. 24. Jeffery M, Hickey BE, Hider PN. Follow up strategies for patients treated for non-metastatic colorectal cancer. Cochrane Database Syst Rev 2007; (1): CD002200. 25. Wanebo HJ, Antoniuk P, Koness RJ, et al. Pelvic resection of recurrent rectal cancer: technical considerations and outcomes. Dis Colon Rectum 1999; 42: 1438–48. 26. Desai DC, Arnold MW, Burak WE, et al. Positron emission tomography affects surgical management in recurrent colorectal cancer patients. Ann Surg Oncol 2003; 10: 59–64. 27. Wanebo HJ, Marcove RC. Abdominosacral resection of locally recurrent rectal cancer. Ann Surg 1981; 194: 458–71. 28. Yamada K. Patterns of pelvic invasion are prognostic in the treatment of locally recurrent rectal cancer. Br J Surg 2001; 88: 988–93.

HEBK001-C13_p203-221.indd 220

29. Messiou C, Chalmers AG, Boyle K, Wilson D, Sagar P. Pre-operative MR assessment of recurrent rectal cancer. Br J Radiol 2008; 81: 468–73. 30. Arulampalam TH, Loizidou M, Visvikis D, et al. Positron emission tomography and colorectal cancer. Br J Surg 2001; 88: 176–89. 31. Schlag P, Strauss LG, Georgi P, Herfath C. Scar or recurrent rectal cancer. Positron Emission Tomography is more helpful for diagnosis than immunoscintigraphy. Arch Surg 124: 1989. 32. Takeuchi O, Koda K, Sarashina H, Nakajima N. Clinical assessment of positron emission tomography for the diagnosis of local recurrence in colorectal cancer. Br J Surg 1999; 86: 932–7. 33. Popovic MJ, Hricak H, Sugimura K, Stern JL. The role of MR imaging in determining surgical eligibility for pelvic exenteration. Am J Roentgenol 1993; 160: 525–31. 34. Brown G. Thin section MRI in multidisciplinary preoperative decision making for patients with rectal cancer. Br J Radiol 2005; 78: S117–27. 35. Delbeke DVJ, Sandler MP. Staging recurrent colorectal carcinoma with PET. J Nuclear Med 1997; 38: 1196–201. 36. Vermaas M, Ferenschild FT, Nuyttens JJ, et al. Preoperative radiotherapy improves outcome in recurrent rectal cancer. Dis Colon Rectum 2005; 48: 918–28. 37. Alektiar KM, Zelefsky MJ, Paty PB, et al. High-doserate intraoperative brachytherapy for recurrent colorectal cancer. Int J Radiat Oncol Biol Phys 2000; 48: 219–26. 38. Lowy AM, Rich TA, Skibber JM, et al. Preoperative infusional chemoradiation, selective intraoperative radiation, and resection for locally advanced pelvic recurrence of colorectal adenocarcinoma. Ann Surg 1996; 223: 177–85. 39. Lindel K, Willett CG, Shellito PC, et al. Intraoperative radiation therapy for locally advanced recurrent rectal or rectosigmoid cancer. Radiother Oncol 2001; 58: 83–7. 40. Mannaerts GH, Rutten HJ, Martijn H, et al. Comparison of intraoperative radiation therapy-containing multimodality treatment with historical treatment modalities for locally recurrent rectal cancer. Dis Colon Rectum 2001; 44: 1749–58. 41. Hashiguchi Y, Sekine T, Sakamoto H, et al. Intraoperative irradiation after surgery for locally recurrent rectal cancer. Dis Colon Rectum 1999; 42: 886–93. 42. Austin KK, Solomon MJ. Pelvic exenteration with en bloc iliac vessel resection for lateral pelvic wall involvement. Dis Colon Rectum 2009; 52: 1223–33. 43. Pawlik TMSJ, Rodriguez-Bigas MA. Educational review pelvic exenteration for advanced pelvic malignancies. Ann Surg Oncol 2006; 13: 612–23. 44. Hartley JE, Lopez RA, Paty PB, et al. Resection of locally recurrent colorectal cancer in the presence of distant metastases: can it be justified? Ann Surg Oncol 2003; 10: 227–33.

08/02/13 5:34 PM

References  221 45. Gagliardi G, Hawley PR, Hershman MJ, et al. Prognostic factors in surgery for local recurrence of rectal cancer. Br J Surg 1995; 82: 1401–5. 46. Hashiguchi Y, Sekine T, Kato S, et al. Indicators for surgical resection and intraoperative radiation therapy for pelvic recurrence of colorectal cancer. Dis Colon Rectum 2003; 46: 31–9. 47. Huguier M, Houry S. Treatment of local recurrence of rectal cancer. Am J Surg 1998; 175: 288–92. 48. Onodera H, Maetani S, Kawamoto K, et al. Pathologic significance of tumor progression in locally recurrent rectal cancer: different nature from primary cancer. Dis Colon Rectum 2000; 43: 775–81. 49. Wiggers T, de Vries MR. Surgery for local recurrence of rectal cancer. Dis Colon Rectum 1996; 39: 323–8. 50. Mirnezami AH, Sagar PM, Kavanagh D, et al. Clinical algorithms for the surgical management of locally recurrent rectal cancer. Dis Colon Rectum 2010; 53: 1248–57. 51. Suzuki K, Dozois RR, Devine RM. Curative reoperations for locally recurrent rectal cancer. Dis Colon Rectum 1996; 39: 323–8. 52. Guillem J RL. Strategies in operative therapy for locally recurrent rectal cancer. Semin Colon Rectal Surg 1998; 9: 259–68. 53. Moore H, Shoup M, Riedel E, et al. Colorectal cancer pelvic recurrences: determinants of respectability. Dis Colon Rectum 2004; 47: 1599–606. 54. Lopez-Kostner F, Fazio V, Vignali A, et al. Locally recurrent rectal cancer: predictors and success of salvage surgery. Dis Colon Rectum 2001; 44: 173–8. 55. Brunschwig A. Complete excision of pelvic viscera for advanced carcinoma: one-stage abdominoperineal operation with end colostomy and bilateral ureteral implantation into colon above colostomy. Cancer 1948; 1: 177–83. 56. Wanebo HJ, Gaker DL, Whitehill E, et al. Pelvic recurrence of rectal cancer: options for curative resection. Ann Surg 1987; 205: 482–95. 57. Yamada K, Ishizawa T, Niwa K, et al. Pelvic exenteration and sacral resection for locally advanced primary and recurrent rectal cancer. Dis Colon Rectum 2002; 45: 1078–84.

HEBK001-C13_p203-221.indd 221

58. Kristjánsson A, Wallin L, Månsson W. Renal function up to 16 years after conduit (refluxing or anti-reflux anastomosis) or continent urinary diversion. 1: Glomerular filtration rate and patency of uretero-intestinal anastomosis. Br J Urol 1995; 76: 539–45. 59. Shibata D, Hyland W, Busse P, et al. Immediate reconstruction of the perineal wound with gracilis muscle flap following abdominoperineal resection and intraoperative radiation therapy for recurrent carcinoma of the rectum. Ann Surg Oncol 1996; 6: 33–7. 60. Chessin DB, Hartley J, Cohen AM, et al. Rectus flap reconstruction decreases perineal wound complications after pelvic chemoradiation and surgery: a cohort study. Ann Surg Oncol 2005; 12: 104–10. 61. Butler CE, Gundeslioglu AO, Rodriguez-Bigas MA. Outcomes of immediate vertical rectus abdominis myocutaneous flap reconstruction for irradiated abdominoperineal resection defects. J Am Coll Surg 2008; 206: 694–703. 62. Radice E, Nelson H, Mercill S, et al. Primary myocutaneous flap closure following resection of locally advanced pelvic malignancies. Br J Surg 1999; 86: 349–54. 63. Bell SW, Dehni N, Chaouat M, et al. Primary rectus abdominis myocutaneous flap for repair of perineal and vaginal defects after extended abdominoperineal resection. Br J Surg 2005; 92: 482–6. 64. Melton GB, Paty PB, Boland PJ, et al. Sacral resection for recurrent rectal cancer: analysis of morbidity and treatment results. Dis Colon Rectum 2006; 49: 1099–107. 65. Kusters M, Dresen RC, Martijn H, et al. Radicality of resection and survival after multimodality treatment is influenced by subsite of locally recurrent rectal cancer. Int J Radiat Oncol Phys 2009; Dec; 75(5): 1444–9. 66. Wiig JN. Total pelvic exenteration with preoperative irradiation for advanced primary and recurrent rectal cancer. Eur J Surg 2002; 168: 42–8. 67. Austin KK, Young JM, Solomon MJ. Quality of life of survivors after pelvic exenteration for rectal cancer. Dis Colon Rectum 2010; 53: 1121–6. 68. Sagar PM. Surgical management of locally recurrent rectal cancer. Br J Surg 1996; 83: 293–304.

08/02/13 5:34 PM

14 Lateral pelvic side-wall nodal involvement in rectal cancer Hideaki Yano and Brendan Moran

Introduction The risk of metastasis to, and clinical significance of, lateral pelvic side-wall nodes in rectal cancer remains controversial, although there is now general agreement that involved pelvic side-wall nodes are more common with low rectal cancer.1 Removal of these nodes, i.e. lateral pelvic lymph node dissection (LPLD), in addition to total mesorectal excision (TME) is considered a major mechanism in reducing local recurrence and improving survival in Japan (Figure 14.1).2,3 The rationale behind this approach is that lateral pelvic side-wall nodal involvement is pathologically confirmed in 10–20 per cent of patients with low rectal cancer, and a considerable number of patients with lateral pelvic side-wall nodal involvement survived free of disease after LPLD even before the modern chemotherapy era. Over the past two decades, LPLD has been continuously developed and refined in Japan, with a more selective approach in high-risk patients and widespread use of autonomic nerve-preserving techniques.1 In Europe and North America, however, the approach has been either to ignore lateral pelvic sidewall nodes or to treat obviously involved nodes with radiotherapy or, more commonly, chemoradiotherapy. Lateral pelvic lymph node dissection

HEBK001-C14_p222-227.indd 222

has been practically abandoned and has not been performed due to the following issues: LPLD is associated with high morbidity; lateral pelvic side-wall nodal involvement was considered rare in the West compared with Japan and the East; l involved lateral pelvic side-wall nodes have generally been considered to indicate systemic disease not amenable to surgery. l l

Advances in pelvic imaging, however, particularly computed tomography (CT) and magnetic resonance imaging (MRI), with the ability to visualize and categorize lymph nodes, together with reviews of the published literature, suggest that lateral nodal involvement rates may be similar in the West and the East and are a particular problem in low rectal cancer.1

INCIDENCE OF LATERAL PELVIC SIDE-WALL NODE INVOLVEMENT In addition to the predominant lymphatic rectal drainage that runs upwards or proximally within the mesorectum, the lower rectum has lateral lymphatic drainage, which may result in involvement of lateral pelvic side-wall nodes (Figure 14.2). This anatomical

11/02/13 3:59 PM

INCIDENCE OF LATERAL PELVIC SIDE-WALL NODE INVOLVEMENT  223

Figure 14.1. Lateral pelvic lymph node dissection on the left pelvic side-wall.

feature pertaining to the low rectum has been supported not only by numerous experimental studies using dye injection and lymphoscintigraphy but also by clinical observations.1,4 A multicentre study in Japan that recruited 2916 patients with rectal cancer reported lateral pelvic side-wall node involvement in 20.1 per cent of patients with T3/4 low rectal cancer lying at or below the peritoneal reflection.2 It is noteworthy that the lower the tumour is, the higher the risk of lateral pelvic side-wall node involvement.5 A number of risk factors have been reported to predict lateral pelvic side-wall nodal metastasis. Modern pelvic imaging by CT or MRI allows visualization of suspicious pelvic side-wall nodes, and

studies are ongoing to define the sensitivity and specificity of imaging in determining the status of pelvic side-wall nodes. This correlation can only be truly documented in patients who have imaging followed by LPLD without neoadjuvant therapy. In addition to modern imaging of the nodes, reported predictive factors for lateral nodal involvement include positive mesorectal nodal status, height of the tumour, depth of tumour invasion, histological grade, lymphovascular invasion, size of the tumour and female sex.1,2,6 Some factors can be determined or reliably predicted before surgery, but others can be determined only after surgery. A prospective study reported a high diagnostic accuracy of CT scanning in predicting

Upward spread Lateral spread

Pelvic plexus

Figure 14.2. Lymphatic drainage from the low rectum exists in the space between the pelvic plexus and the lateral pelvic side-wall.

HEBK001-C14_p222-227.indd 223

11/02/13 3:59 PM

224  Lateral pelvic side-wall nodal involvement in rectal cancer

Figure 14.3. Imaging suggestive of right lateral pelvic side-wall node involvement.

lateral pelvic side-wall node involvement, allowing preoperative identification of patients most likely to benefit from extended dissection (LPLD) (Figure 14.3).7 In future, MRI or positron-emission tomography (PET) techniques may further enhance the selection of patients for appropriate treatment.

Lateral pelvic side-wall node involvement: local or systemic disease? Controversy exists as to whether involved pelvic side-wall nodes represent systemic disease or localized disease that is amenable to surgery. The previously mentioned large multicentre study reported that the 5-year survival rate of patients with lateral pelvic side-wall node involvement was 45.8 per cent, which surpasses the outcomes following liver resection for hepatic metastasis.2 Patients with lateral pelvic side-wall node involvement constitute a heterogeneous group. In some patients the disease is undoubtedly systemic and carries a poor prognosis, indicating a need for neoadjuvant and adjuvant treatment or even palliative care. Other patients have localized disease as demonstrated by favourable 5-year survival rates achieved with LPLD.

OPERATION In contrast to Western countries, LPLD has been continuously promoted, performed and refined

HEBK001-C14_p222-227.indd 224

in Japan. Key refinements have been a more patient-selective approach and nerve-preserving surgery.1 Lateral pelvic lymph node dissection for upper rectal cancer has been almost abandoned in Japan with a highly selective approach to LPLD. Based on the large multicentre study, the Japanese guidelines now suggest LPLD for T3/4 rectal cancer with its lower edge lying at or below the peritoneal reflection.8 There is a problem with these criteria because the precise depth of the tumour cannot be determined accurately preoperatively; lateral pelvic side-wall nodal involvement was confirmed in only 20 per cent of the patients who had LPLD; and lateral pelvic side-wall node involvement was found in 9.2 per cent of T2 low rectal cancers and would not have been removed based on these criteria. A further controversial issue is whether ‘prophylactic’ LPLD for advanced low rectal cancer offers any survival benefit. This depends on the accuracy of preoperative assessment of tumour depth, nodal involvement and other preoperative factors. Lateral pelvic lymph node dissection is known to be associated with significant morbidity, longer operating time, greater blood loss and genitourinary dysfunction. Nerve-sparing techniques have been developed and refined, with marked improvements in urinary and sexual function in recent years.9,10 There are concerns, however, as to whether preservation of autonomic nerves, particularly the inferior hypogastric plexus, may compromise the oncological outcomes when cancer cells are likely to have metastasized through the lateral lymphatic channels to the lateral pelvic side-wall nodes.

11/02/13 3:59 PM

OPERATION  225

Figure 14.4. Dissection between the external iliac artery and the psoas major muscle.

Figure 14.5. Dissection of the medial aspect of the external iliac vessels.

Operative Techniques

served. The ureter is then secured using a vessel loop or tape to reduce the risk of inadvertent damage. The external iliac artery is exposed and secured using a vessel loop or tape. By exposing the medial aspect of the psoas major muscle, the fat tissue between the external iliac artery and the psoas major is dissected and the obturator nerve is identified (Figure 14.4). Care should be taken not to damage the fifth lumbar vein that drains into either the inferior vena cava or the external iliac vein. Before dissection of the obturator region, the anterior aspect of the internal iliac artery is exposed and the fat tissue surrounding the external iliac vein is dissected from the cranial to the caudal direction. There is generally no need to follow the external iliac vessels to their distal ends, unless there is suspicious tissue there, as this may increase the risk of lymphoedema of the leg (Figure 14.5).

Lateral pelvic lymph node dissection is usually performed following completion of rectal resection using the principles of TME or sphincter resection in patients who require abdominoperineal excision. In restorative procedures, LPLD is performed before the anastomosis. If there are concerns about the margins of the suspicious nodes, the internal iliac artery or vein and the autonomic nerves, i.e. inferior hypogastric plexus (pelvic plexus), should be excised en bloc. The main aspects of LPLD on the right pelvic side-wall are as follows: The hypogastric nerve is secured using a vessel loop or tape. Caudally, the inferior hypogastric plexus and the pelvic splanchnic nerves (nervi erigentes) are dissected off the parietal endopelvic fascia and pre-

Figure 14.6. Dissection of the obturator region.

HEBK001-C14_p222-227.indd 225

11/02/13 3:59 PM

226  Lateral pelvic side-wall nodal involvement in rectal cancer

Figure 14.7. Dissection of the dorsal aspect of the obturator region.

Figure 14.9. Identification of the inferior vesical vessels.

By retracting the ureter and the external iliac vessels laterally, the optimal surgical field will be obtained. The dissection of the obturator region is commenced from where the internal and external iliac arteries branch off the common iliac artery. The superior vesical artery and the obturator vessels are then identified, both of which can be excised if necessary. Retracting the superior vesical artery medially, dissection is continued by exposing the obturator nerve and the internal obturator muscle as far as the point where the nerve enters the obturator canal (Figure 14.6). When using diathermy dissection, local injection of lidocaine is very helpful to prevent stimulation of the adductor muscles. Alternatively, ultrasonic coagulating shears may prove useful. On the craniomedial aspect, the lymphatic tissue is dissected off the internal iliac vessels that are

situated on the lumbosacral trunk and the sacral nerve plexus (Figure 14.7). Care should be taken, particularly if excising the internal iliac vessels, not to damage the small branches of the vein, which can result in considerable bleeding. The caudal part of the obturator area is dissected by pulling medially the superior vesical artery and the ureter. The levator ani muscle is exposed just caudal to the internal obturator muscle (Figure 14.8). More medially, identification of the inferior vesical vessels that are situated just lateral to the inferior hypogastric plexus is important. This is followed by division of the distal branches of the inferior vesical artery and vein at the point just proximal to the bladder (Figure 14.9). Lateral pelvic lymph node dissection is completed by dividing the inferior vesical vessels from the internal iliac vessels (Figures 14.10 and 14.11).

Figure 14.8. Dissection of the caudal aspect of the obturator region.

Figure 14.10. Division of the inferior vesical vessels.

HEBK001-C14_p222-227.indd 226

11/02/13 3:59 PM

References  227

  4.

  5.

  6.

Figure 14.11. View after completion of lateral pelvic lymph node dissection on right pelvic side-wall.

References   1. Yano H, Moran BJ. The incidence of lateral pelvic sidewall nodal involvement in low rectal cancer may be similar in Japan and the West. Br J Surg 2008; 95: 33–49.   2. Sugihara K, Kobayashi H, Kato T, et al. Indication and benefit of pelvic sidewall dissection for rectal cancer. Dis Colon Rectum 2006; 49: 1663–72.   3. Kusters M, Beets GL, van de Velde CJH, et al. A comparison between the treatment of low rectal cancer in

HEBK001-C14_p222-227.indd 227

  7.

  8.

  9.

10.

Japan and Netherlands, focusing on the patterns of local recurrence. Ann Surg 2009; 249: 229–35. Bell S, Sasaki J, Sinclair G, et al. Understanding the anatomy of lymphatic drainage and the use of bluedye mapping to determine the extent of lymphadenectomy in rectal cancer surgery: unresolved issues. Colorectal Dis 2009; 11: 443–9. Ueno M, Oya M, Azekura K, et al. Incidence and prognostic significance of lateral lymph node metastasis in patients with advanced low rectal cancer. Br J Surg 2005; 92: 756–63. Ueno H, Yamauchi C, Hase K, et al. Clinicopathological study of intrapelvic cancer spread to the iliac area in lower rectal adenocarcinoma by serial sectioning. Br J Surg 1999; 86: 1532–7. Yano H, Saito Y, Takeshita E, et al. Prediction of lateral pelvic node involvement in low rectal cancer by conventional computed tomography. Br J Surg 2007; 94: 1014–9. Japanese Society for Cancer of the Colon and Rectum (ed.). JSCCR Guidelines 2010 for the Treatment of Colorectal Cancer, 2nd edn. Tokyo, Kanehara & Co., 2010: 13–5. Moriya Y, Sugihara K, Akasu T, Fujita S. Nerve-sparing surgery with lateral node dissection for advanced lower rectal cancer. Eur J Cancer 1995; 31A: 1229–32. Matsuoka H, Masaki T, Sugiyama M, Atomi Y. Impact of lateral pelvic lymph node dissection on evacuatory and urinary functions following low anterior resection for advanced rectal carcinoma. Langenbecks Arch Surg 2005; 390: 517–22.

11/02/13 3:59 PM

15 Intestinal stoma and the role of defunctioning a low anastomosis after anterior resection David Mitchell, Kandiah Chandrakumaran and Steven Arnold

Introduction Surgeons have been constructing intestinal stomas for over 200 years, but the modern ileostomy and colostomy have been developed from our predecessors’ trial and error. Littre proposed the ‘inguinal colostomy’ in 1710 for infantile imperforate anus but never actually performed the operation. It took more than 65 years before the first stoma was performed in 1776, when French surgeon Pillore created a defunctioning stoma for obstructing rectal cancer on Monsieur Morel. This was a caecostomy and the patient died after 28 days from complications not directly related to the stoma; this was not reported in the medical press until 1840. In 1783, Dubois performed the first ‘Littre’s operation’ on a 3-day-old infant for imperforate anus, but the baby died on the tenth day. The first successful colostomy was performed in 1793 by Duret, on an infant with a maldeveloped perineum and absence of the rectum. With the advent of anaesthesia, diversion colostomy became popular for the management of large bowel obstruction. In the early days there was a high mortality rate. Amussant, a Parisian surgeon, attributed this high mortality rate to peritonitis as a result of opening the peritoneum. He suggested

HEBK001-C15_p228-238.indd 228

accessing the colon in the lumbar region without opening the peritoneum. A few surgeons attempted this procedure but failed, and the approach was rapidly abandoned. The first panproctocolectomy and ileostomy was performed by Miller in 1949. When total mesorectal excision (TME) is performed, with a very low division of the muscle tube to enable re-anastomosis, most surgeons would elect to use a proximal defunctioning stoma to reduce the consequences and possibly the incidence of anastomotic leakage. It has been well reported that when a low rectal anastomosis is performed (by definition, less than 7 cm from the anal verge on rigid sigmoidoscopy), then the risk of an anastomotic leak rises significantly; the lower the anastomosis, the higher the risk.1 The term ‘anterior resection’ by definition involves an operation whereby the superior rectal artery has been ligated and the anastomosis has been performed to the top of the rectum. The term is increasingly used for surgical management of pathology of the sigmoid colon, especially in the laparoscopic era, where the technique enables modern stapling guns to construct the anastomosis without the need for intracorporeal suturing. Even when resection of the upper rectum is added, the resulting anastomosis is relatively high, the risks of anastomotic leakage are much reduced, and a

08/02/13 5:29 PM

LOOP ILEOSTOMY  229

stoma is frequently unnecessary and can be avoided. An ‘anterior resection’ involving TME down to the pelvic floor, however, is a different matter, with much higher risks of leakage.

PREOPERATIVE PLANNING Almost all TME operations will be scheduled electively. This gives the patient and team the chance to prepare for the high likelihood of needing a stoma. Preoperatively, there may be the opportunity to meet another patient with a stoma, thereby allowing personal questions and expectations to be addressed. The patient should meet a specialist stoma nurse (stomatherapist), who has a key role to play in the preoperative preparation and peri- and postoperative management. Optimal abdominal wall stoma siting is a crucial step in successful outcome. The aim is to locate a flat smooth skin surface so that stoma appliances can be fitted with as near perfect seal as possible. This reduces the likelihood of leaking and skin excoriation caused by irritant gastrointestinal contents. The patient and stoma nurse carefully consider factors such as abdominal shape, previous scars, and where the patient can physically see when selecting and marking a suitable stoma site. Personal factors such as clothing styles and where belts are worn are also considered. Siting is carried out away from bony landmarks, and skinfolds are avoided to prevent the stoma from lying in a crease when the patient is sitting upright. Note that skinfolds may not be apparent when the patient is supine during the operation. Some surgeons have suggested that the risk of developing a parastomal hernia is reduced when the bowel is brought through the rectus abdominis muscle, and hence the stoma site is marked medial to the lateral edge of the rectus abdominis muscle. Once a suitable site is identified, the site is marked with an indelible marker pen. The site can then be covered with a clear adhesive plastic dressing, thereby preventing the mark from becoming faint or rubbing off. It is good practice to have both left- and right-sided stoma sites marked as surgery may not progress according to plan, thereby allowing both a defunctioning ileostomy or an end colostomy formation. Pre-existing marks allow whichever stoma is necessary to be formed in a good position. The authors favour the practice of

HEBK001-C15_p228-238.indd 229

marking stoma sites with either surgical skin clips or a silk stitch as soon as the patient is prepared and draped, as the ink marks may completely rub off during the operation, especially if the adhesive dressing is removed for skin preparation. In the emergency situation, it may not be possible to have optimal planning for siting a stoma, such as surgery for obstruction, bleeding, ischaemia or abdominal trauma. Nevertheless, the surgeon should take account of all of the above factors when choosing where to site a necessary stoma, as the ­patient may have the stoma long term and often indefinitely. A major role of the stomatherapist is to provide the patient with preoperative counselling. It has been reported that the presence of a stoma may delay the discharge of an otherwise well patient by as much as 4 days.2 With the drive towards enhanced recovery and early discharge, preoperative training in the care, emptying and changing of a stoma bag may hasten the patient’s discharge. It is essential to provide contact numbers of the stoma care team to the patient on discharge so that they can access specialist help quickly in the event of problems. It is important to note that a stoma may have profound impact on a patient’s life. Body dysmorphia is common and the impact is dramatically increased by leaking and poorly functioning stomas. Therefore, the surgeon should take the utmost care and exemplary surgical techniques to avoid any complications. This may be especially important to remember at the end of a long operation such as TME for rectal cancer or total colectomy.

LOOP ILEOSTOMY A loop ileostomy is most commonly used to defunction a low rectal anastomosis (Figure 15.1). A loop ileostomy appears to have become the procedure of choice in most units for defunctioning a low rectal anastomosis and has superseded the equally effective loop colostomy for defunctioning. A suitable point on the terminal ileum is selected. This is usually about 25–30 cm proximal to the ileocaecal valve. If it is nearer to the ileocaecal valve, then in theory back pressure on the segment at the time of stoma reversal may increase the likelihood of a leak from the ileostomy closure site. A disc of skin and underlying fat is then excised at

08/02/13 5:29 PM

230  Intestinal stoma and the role of defunctioning a low anastomosis after anterior resection

the marked site and the underlying sheath exposed. The assistant can attach a Littlewoods or similar instrument to the edge of the midline sheath, allowing gentle traction to straighten the abdominal wall so the stoma can be brought out straight. Traditionally a cruciate incision was made in the sheath, but a longitudinal incision is now recommended; the underlying rectus muscle fibres are gently separated with Langenbeck’s retractors to expose the posterior rectus sheath. This is then incised and then gently stretched to permit the formation of the stoma. Inserting two fingers through the site usually allows enough stretching for easy passage of the bowel, without leading to too big a defect with an increased risk of subsequent herniation. Babcock’s forceps are ideal for passing through the site, grasping the small bowel, and gently guiding the bowel through the abdominal wall. Orientation of the intra-abdominal small bowel is checked, preferably with the efferent limb positioned superiorly. At least 6 cm of bowel, measured on the antimesenteric border, should be delivered, to allow sufficient length for formation of a good spout. Before closing the abdomen, two or three dissolvable sutures are in-

serted inferiorly, attaching the small bowel to the skin. This holds the loop in place and marks where the flush afferent limb will be attached to the skin. Following closure and protection of the main wound, the ileostomy can then be formed. The ileum is opened transversely by an incision a couple of millimetres above the previously placed efferent small bowel-to-skin sutures. This incision is extended for about two-thirds of the circumference of the bowel, but preserving the mesenteric border, which is left in continuity. Three eversion sutures are then placed at the 12, 3 and 9 o’clock positions. Dissolvable sutures are preferred, such as 3.0 monocryl or vicryl rapide. The superior-most stitch is then used to take a seromuscular bite approximately 6 cm proximal to the cut end of the bowel, with the two side sutures taking a similar bite at approximately 5 cm from the cut end of the small bowel. Once all three are in situ, gentle traction on all three usually results in eversion of the bowel to form a spout approximately 2.5 cm long, facing slightly downwards. Sometimes gentle assistance is required to evert the spout, and the blunt end of Langenbeck’s retractors can be used to facilitate this. Once these stitches are secure, further sutures can be placed between them to secure the new mucocutaneous junction around the circumference of the stoma. Haemostasis is then checked, and a stoma bag can be cut to shape and fitted.

End ileostomy

Proximal spouted end Distal flush end

Figure 15.1. Loop ileostomy.

HEBK001-C15_p228-238.indd 230

In the event of a panproctocolectomy (e.g. a patient with rectal cancer on a background of ulcerative colitis), an end ileostomy must be fashioned. After removing the large bowel, a disc of skin is excised from the abdominal wall at the marked site, along with a disc of subcutaneous fat. The terminal ileum will usually have been cross-stapled and divided close to the caecum. After entering the abdominal cavity as for a loop ileostomy, the end of the ileum is grasped gently with Babcock’s forceps and withdrawn through the opening. The mesentery supplying the end of the bowel is also brought through; care must be taken not to traumatize the mesentery and risk a haematoma or disruption of the blood supply. Small bowel orientation is checked to ensure no rotation of the mesentery, and the laparotomy wound is closed.

08/02/13 5:29 PM

ALTERNATIVES  231

The staple line is cut off and a spout is created similar to the technique described above. The authors use four sutures placed at 6, 9, 12 and 3 o’clock on the circumference of the open small bowel, going from inside out and then through the serosa and muscle of the proximal small bowel 6 cm from the cut end for the uppermost two sutures and 5 cm from the end for the lower two sutures. Traction on all four sutures usually inverts the bowel to produce a spout; the slightly longer uppermost length produces a downward tilt at rest to direct liquid contents into a stoma bag. The sutures are tied and held on forceps while intervening sutures are placed.

Loop colostomy A transverse loop colostomy is a good alternative defunctioning stoma, and a premarked site will usually be in the right upper quadrant. Once the rectal anastomosis is completed, the portion of the transverse colon that can be exteriorized at the preoperatively marked site is selected. It is important that there is no undue tension on the mesentery or appearances of ischaemia; full mobilization of the hepatic flexure may be required. An ellipse of skin is removed and the subcutaneous fat is split and retracted with Langenbach’s retractors. The anterior rectus sheath is exposed and a cruciate or linear fascial incision made. The rectus abdominis muscle is split longitudinally. Great care must be taken to avoid injury to the inferior epigastric vessels. The peritoneum is opened and the selected portion of the colon is exteriorized with Babcock’s forceps. A useful technique is to pass a length of tubing around the transverse colon, through an avascular area in the mesentery, and use this to pull the colon through the abdominal wall opening. It is important at this stage to ensure that the orientation is correct, there is no undue tension and the opening is adequate. In patients who are obese, it may be necessary to widen the opening to ensure this. A couple of ‘stay’ sutures are applied and the laparotomy wound is closed in the usual manner. The wound is dressed and protected. A transverse enterotomy is made in the loop, to include approximately twothirds of the circumference, and the bowel ends are cleaned with povidone–iodine. Haemostasis is achieved, and interrupted absorbable stitches are applied to approximate the open edge of the colon

HEBK001-C15_p228-238.indd 231

to skin. These should be placed in an extramucosal fashion. It is common to use a ‘bridge’ under the colonic loop, either purpose-designed or an adapted piece of tubing (as described above to pull out the loop), to help support the stoma. This is removed at 5–7 days. The disadvantage of a bridge is that it interferes with appliance application and changing. A loop colostomy may also be required in some very symptomatic patients with rectal cancer before neoadjuvant chemoradiotherapy or occasionally to relieve obstruction. The traditional trephine colostomy method should now be almost exclusively replaced by a laparoscopic technique to obviate the reported risks of bringing the wrong end of a stoma or a rotated loop via a trephine. Difficulty often arose in correctly identifying the orientation, especially with redundant sigmoid or transverse colon. A laparoscopic technique is optimal to assess the mesenteric length, allow mobilization and confirm orientation. If a trephine sigmoid colostomy is being performed, a useful technique is to intubate the colon peranally using a flexible sigmoidoscope. If using a sigmoidoscope, it is important not to inflate too much air and to limit insufflation to a ­minimum.

End colostomy An end colostomy is performed after abdominoperineal excision of the rectum and in some emergencies in the form of a Hartmann’s procedure. In both cases, the distal end of the proximal colon is usually stapled closed during the laparotomy, until the specimen is dissected and removed. The closed end is then exteriorized through the ­preoperatively marked site, without tension, similar to a loop colostomy. Scissors can be used to excise the staple line, and the edges of the colon and skin are approximated with circumferential interrupted absorbable sutures.

ALTERNATIVES Although formal proximal defunctioning gives the best protection to a distal anastomosis, alternative measures have been described. Caecostomy (intubation of the caecum, usually with a urinary catheter) has been used, although

08/02/13 5:29 PM

232  Intestinal stoma and the role of defunctioning a low anastomosis after anterior resection

the procedure is largely historic. The appendix was amputated and the stump brought on to a small opening on the abdominal wall, in the right iliac fossa. The hypothesis was that a caecostomy allowed some venting of the accumulated gas, thereby reducing the luminal pressure on an anastomosis. It did not defunction the faecal stream, however, and this practice has now been superseded by loop ileostomy or colostomy if protection is deemed necessary. Other methods proposed to decrease the potential pressure across an anastomosis involve peranal intubation.3 Again, a urinary catheter has been used across a low anastomosis, but a custom-designed silastic stent that sits across a low anastomosis and has an aperture that allows flushing by a bladder syringe has been evaluated, with encouraging results.3 The tube or stent is removed at 5–7 days or when the patient has opened their bowel.

Closure of stomas Timing of stoma-reversal surgery is important and often is determined by multiple factors. Before stoma reversal, most surgeons advocate confirmation of anastomotic healing by a fluorographic water-soluble contrast enema. It is recommended that the surgeon performs a digital rectal examination before referral for a contrast study; often a soft stricture is encountered and should be digitally dilated. Care is needed in inserting the soft rubber enema tube to avoid undue trauma, especially to a very low anastomosis. Gastrograffin or a similar agent is then allowed to flow down the tube to fill the neorectum, and X-ray images are taken. These X-rays can be taken a few weeks after surgery, thereby allowing the reversal operation to be planned. If no adjuvant therapy is proposed, then the patient may be booked for reversal at 6–8  weeks after their resection, although generally the longer the time interval (even beyond this), the less difficult the procedure. The delay allows the adhesions between the stoma and abdominal wall to mature to a point where subsequent dissection is safer, with less peristomal and bowel oedema. Some surgeons rely on clinical examination alone to determine whether the anastomosis is intact. A careful digital examination is performed to feel for an intact join around the circumference of

HEBK001-C15_p228-238.indd 232

the bowel. This may be supplemented by luminal endoscopy. If the pathology of the resected specimen is reported as an advanced tumour, adjuvant therapy may be offered. The timing of adjuvant therapy is thought to be important, as undue delay from the original surgery may render adjuvant therapy less effective. Thus, reversal of a defunctioning stoma may delay treatment, so many oncologists prefer to treat before stoma reversal, and the reversal may have to be deferred for 6 months or longer. The principles of stoma reversal are similar for loop ileostomy or colostomy reversal. An incision is made around the stoma at the interface of the skin and mucosa. Some surgeons recommend a transverse elliptical incision a few millimetres away from the skin mucosa interface, with the ends tapering together at the 3 and 9 o’clock positions to improve the eventual scar. Once the edge of the bowel is identified, careful dissection proceeds circumferentially to free up the adhesions between the bowel and the abdominal wall. Sharp dissection is preferred to diathermy, thereby preventing thermal damage to the bowel. This should be a relatively bloodless plane; if difficulty is encountered with small bleeding points, it may be that there is a further layer of tissue still attached to the bowel wall. The dissection is deepened circumferentially, separating the bowel from the sheath and muscle layers, until the abdominal peritoneal cavity is ­entered. Within the peritoneal cavity, the remaining adhesions can be freed, often by sweeping a finger in the peritoneal cavity deep to the abdominal wall. This allows delivery of a good length of both bowel ends through the abdominal wall. Attention should then be focused on preparing the bowel for anastomosis. There may be a small rim of skin remaining on the edges of the bowel, and this should be excised. Bleeding during this manoeuvre is a good sign, as it signifies that there should be a healthy blood supply to the new anastomosis. The ends of a spouted ileostomy should then be unfolded. Often, as the stoma has been formed only a few weeks previously, the adhesions may be flimsy without any undue fibrosis. Sometimes gentle digital pressure alone is enough to evert the spout, but on other occasions careful sharp dissection may be required. This allows the ends of the bowel to be approximated, assisted by the existing continuity of the back wall. The bowel is then

08/02/13 5:29 PM

STOMA COMPLICATIONS  233

sutured. The authors favour using a 3/0 polydioxanone suture with interrupted stitches to form an end-to-end anastomosis. The first sutures should be placed at each end of the closure; it can be useful to hold these two tied sutures apart using artery forceps, thereby facilitating further stitch placement. Further extramucosal stitches are placed and tied, until the last suture is placed. Careful inspection of the anastomosis ensures no gaping defects, and the bowel can then be gently reduced into the abdominal cavity. If reduction is difficult, it is better practice to further open the sheath rather than risk haematoma formation or damaging a new anastomosis by forcing it through a tight gap. Haemostasis should be checked before fascial closure. An alternative way of fashioning the anastomosis is to use mechanical stapling instruments. Once the two limbs of the bowel have been freed, as described above, the two limbs of a linear stapler are inserted into the lumen of the loops of bowel and the device approximated but not closed. It is crucial at this point to check that the mesentery of the bowel is not caught in the stapler, as firing through this may lead to troublesome bleeding, with the risk of haematoma formation and bowel ischaemia. The stapler can then be closed fully and a final check made that the mesentery is clear. After firing, the linear stapler device is opened and removed, and a transverse stapler is applied perpendicular to the first staple line to seal the opened bowel ends. A linear stapler for this part of the technique is no longer recommended, as anastomotic leakage from the cross-stapling line has been reported. The excess bowel can then be removed using a scalpel. Many surgeons recommend re-enforcing the staple line with sutures and also placing a separate stitch approximately 1 cm beyond the extent of the first staple line to help relieve tension at this point, as this is an area reported as being prone to leakage. Once the bowel has been closed and reduced into the abdomen, the abdominal wall is closed. The sheath is sutured with a strong monofilament nylon using either continuous or interrupted sutures. Antiseptic can then be poured into the cavity, in an effort to reduce the risk of infection. Interrupted sutures or skin clips can be used to close the skin, allowing some of the clips or sutures to be removed easily in the event of a superficial infection.

HEBK001-C15_p228-238.indd 233

Due to the risk of postoperative infection, some surgeons prefer to leave the skin wound open. Some have advocated a circumferential skin incision at the skin/mucosa interface and then subsequently a nylon purse-string to draw the edges together and an alginate dressing as a plug in the centre. This requires regular dressing changes postoperatively, but the resulting scar is usually surprisingly neat and contracts down, with very low infection rates.

STOMA COMPLICATIONS There are a number of stoma-related complications. Complications related to temporary defunctioning stomas may be managed conservatively or tolerated until reversal, but patients who have a permanent stoma may require corrective surgery.

Stricture/Retraction Stricturing of the skin around a stomal orifice may be associated with a degree of retraction of the bowel. Minor separation of the mucocutaneous border is common and managed conservatively with careful assessment to ensure no collection or cellulitis develops. Subsequent healing by secondary intention may lead to a degree of scarring and subsequent stricturing. Dilation of a minor stricture may be successful, but if this fails surgery may be warranted if the stoma is permanent and the stricture problematic. Often, this is straightforward and requires only incision of the scarred area just away from the orifice, and a small sliver of the scar tissue is excised. The bowel is then re-approximated to the skin using absorbable sutures. A more significant and early retraction may be seen if there is too much tension on the bowel at the time of stoma creation. Abscess formation or cellulitis may occur in this instance, and a re-laparotomy may be required. This is often a difficult operation, with oedema and infection, resulting in friability of the tissues. For an ileostomy that gradually retracts and loses its spout, a local revision may be possible. This involves a circumstomal incision and gentle dissection of the layers between the bowel and the abdominal wall. Often, once freed circumferentially, the bowel can be advanced gently and the spout resutured satisfactorily.

08/02/13 5:29 PM

234  Intestinal stoma and the role of defunctioning a low anastomosis after anterior resection

Prolapse Prolapse is more common with a colostomy, particularly loop colostomy. Local amputation of the redundant prolapse of an end colostomy with resuture of the mucocutaneous junction is a straightforward procedure that can be done under local anaesthetic in an unfit patient. Recurrent prolapse is common, however. Loop stomas, particularly loop colostomies, are more prone to prolapse (almost always the distal limb) and are solved by stoma closure, provided reversal is safe and feasible. If reversal is not imminently possible, one option with a prolapsing loop is to mobilize the bowel circumferentially and then staple off the distal limb, which is reduced into the peritoneal cavity. The proximal limb is then refashioned into an end stoma, after amputating any redundant length.

Herniation Parastomal herniation is very common, but the risk of strangulation is low. Patients with herniation may have leaking bags caused by difficulty in getting appliances to adhere and seal properly. Various flanges and belts are available that may prevent this problem, but sometimes surgical intervention is r­ equired. Several options have been evaluated to repair these herniae, with variable success rates reported. All of them have a high incidence of hernia recurrence, and patients should be consented accordingly. Peristomal local mobilization, sac herniotomy and closure of the sheath with nonabsorbable sutures is probably easiest but also prone to the highest failure rate. A better repair may be achieved by making a medial incision away from the stoma site and approaching the sac anterior to the abdominal wall musculature. The mucocutaneous junction is kept intact and the stoma covered during this operation. Once the sac is reduced, the defect can be closed with nonabsorbable sutures and a synthetic mesh used to reinforce the area. Using synthetic mesh in this way still carries a risk of infection, but newer alternatives with collagen-based biological mesh may be helpful in this scenario. Biological meshes are expensive, however, and long-term follow-up is not currently available.

HEBK001-C15_p228-238.indd 234

Developments in laparoscopic surgery have extended to laparoscopic parastomal hernia repair. Laparoscopic ports are introduced and positioned well away from the stoma orifice, and intra-abdominal reduction of the parastomal contents allows the placement of synthetic mesh around the bowel. Procedure-specific meshes have antiadhesion backing layers to minimize adhesion formation. The mesh can be positioned and fixed into place, and the ends overlapped to ensure complete encirclement of the bowel where it passes through the abdominal wall. The other option to repair a parastomal hernia is to perform a re-laparotomy, take down the original stoma, and resite in a new premarked position. The defect at the original site can be closed with a strong non-absorbable monofilament suture. Some surgeons insert a prophylactic biological mesh to reduce the risk of longer-term hernia recurrence.

EVIDENCE FOR FAECAL DIVERSION AFTER TOTAL MESORECTAL EXCISION Anastomotic leakage following TME is impossible to prevent completely. Predicting risk and managing consequences therefore becomes the aim of the surgeon. Rates of leakage after anterior resection vary between institutions, but a consistent predictor of an increased risk is the height of the anastomosis from the anal verge.4 Colorectal anastomoses below the peritoneal reflection, or coloanal anastomoses, have clinical leak rates reported in the range 11–37.5 per cent.5 A meta-analysis of defunctioning stoma versus no stoma following rectal resection concluded that defunctioning resulted in a lower leak rate.6 The meta-analysis included data from four randomized controlled trials (RCTs)5,7–9 and 21 non-randomized studies, with 11 429 patients in total. Meta-analysis of the RCTs showed a lower clinical anastomotic leak rate and a lower reoperation rate in the stoma group. In the non-randomized studies, a lower clinical anastomotic leak rate, lower reoperation rate and lower mortality rate in the stoma group was reported. The strength of available evidence in favour of faecal diversion with proximal ostomy raises the questions of how best to divert the faecal stream, the likelihood that a defunctioning stoma will be permanent,

08/02/13 5:29 PM

EVIDENCE FOR FAECAL DIVERSION AFTER TOTAL MESORECTAL EXCISION  235

and the risks specifically associated with a defunctioning stoma and with stoma reversal.

Loop Colostomy or Ileostomy? A systematic review and meta-analysis of RCTs and observational studies comparing temporary loop ileostomy and loop colostomy for defunctioning of colorectal or coloanal anastomoses has reported findings in support of loop ileostomy.10 Clinically relevant events were grouped into four study outcomes: general outcome measures: dehydration, wound infection; l construction of the stoma outcome measures: parastomal hernia, stenosis, sepsis, prolapse, retraction, necrosis, haemorrhage; l closure of the stoma outcome measures: anastomotic leak or fistula, wound infection, occlusion, hernia; l functioning of the stoma outcome measures: occlusion, skin irritation. l

Five randomized controlled trials and seven observational studies were included in this analysis. Overall, the included studies reported on 1529 patients, just over half of whom (58 per cent) underwent defunctioning loop ileostomy. Loop ileostomy had a lower risk of prolapse and sepsis but was associated with a higher risk of dehydration due to a high output and intestinal obstruction after stoma closure. There were no other significant differences. Included in the analysis was a small randomized trial from the authors’ institution, which reported in favour of loop ileostomy versus loop colostomy after TME surgery, due to fewer abdominal wall problems (incisional hernia, wound infection) in the ileostomy group.11

Frequency of Non-Reversal of Defunctioning Stoma Although intended to be temporary, a substantial proportion of defunctioning stomas are never reversed and in effect the loop stoma becomes permanent. The Dutch Colorectal Cancer Group analysed data from a TME trial in patients with rectal cancer to identify factors that limit stoma reversal.12 In total, 924 patients

HEBK001-C15_p228-238.indd 235

with rectal cancer (459 in the radiotherapy group, 465 in the non-radiotherapy group) who underwent a low anterior resection had a defunctioning stoma in the prospective randomized TME trial involving a total of 1530 patients. Creation of stomas and time to stoma reversal were analysed retrospectively by use of multivariate analysis. In 523 of 924 patients (57 per cent), a primary stoma (defined as a stoma created at the time of TME) was constructed after a low anterior resection. Loop ileostomy was formed in 329 patients and colostomy in 194 patients. Secondary stoma, defined as any stoma created during a second or following procedure after TME, was necessary in 93 of 401 patients (23 per cent). The main reasons for secondary stoma formation were anastomotic leakage (n 5 61), abscess, sepsis or peritonitis (n 5 18) and fistula (n 5 6). The total number of patients who therefore received a temporary stoma either at initial surgery or at a secondary stage was 616 of 924 patients (67 per cent). For stomas that were closed, 97 per cent were done so within the first postoperative year. Nineteen per cent of these defunctioning stomas were never reversed during follow-up. Multivariate analysis demonstrated that preoperative radiotherapy was associated significantly with a decreased likelihood of stoma reversal for secondary stomas, but not for primary stomas. Older age, secondary stoma construction, an end ileostomy or colostomy, any postoperative complication and recurrence were identified as limiting factors for stoma reversal. The authors concluded that postoperative complications are an important limiting factor for stoma reversal because, unsurprisingly, after occurrence of these complications, patients and surgeons may be reluctant to reverse a defunctioning stoma. The most important trial addressing the issue of outcomes after TME for rectal cancer has been a Swedish RCT multicentre trial, which reported 6-year follow-up data.7,13 In total, 234 of 821 patients (28 per cent) undergoing TME were randomly assigned to have either a defunctioning stoma (n 5 116) or no defunctioning stoma (n 5 118). All patients randomized had a satisfactory anastomosis with a negative air-test. Strict definition of leakage, including pelvic abscess, rectovaginal fistula and late leaks, was adhered to. In the original analysis, symptomatic anastomotic leakage was present in 12 of 116 patients (10.3 per cent) with a defunctioning stoma versus 33 of 118 patients (28.2 per cent) without a defunctioning stoma (P , 0.001).7

08/02/13 5:29 PM

236  Intestinal stoma and the role of defunctioning a low anastomosis after anterior resection

This significant reduction in leak rates was observed irrespective of gender. The requirement for emergency reoperation was also significantly higher in the group with no stoma (30 of 118 patients, 25.4 per cent) compared with the group with a ­defunctioning stoma (10 of 116 patients, 8 per cent) (P , 0.001). Of great interest was that with meticulous follow-up, 40 per cent of leaks were diagnosed after discharge from hospital at a median of 24 days (range 13–172 range). At a median follow-up of 42 months postoperatively, more patients who started out with no stoma had a stoma (20 of 118 patients v. 16 of 116 patients; P 5 not significant). At 6-years’ follow-up, 1 patient had died and the total number of patients with stomas had risen from 36 to 45.13 The total group was analysed with regard to the presence of a permanent stoma, the type of stoma, the time point at which the stoma was constructed or considered as permanent, and the reasons for requiring a permanent stoma. During the study period, 45 of 233 patients (19 per cent) were considered to have a permanent stoma: 25 had an end sigmoid colostomy and 20 a loop ­ileostomy. End colostomies were constructed at a median of 22 months after TME, predominantly for anastomotic leakage. Loop ileostomies, performed either at initial surgery (n 5 12) or secondarily (n 5 8), were considered as permanent at a median of 12.5 months after the initial rectal resection. The reasons for a permanent loop ileostomy were metastatic disease, unsatisfactory anorectal function, deterioration in general medical condition, new non-colorectal cancer, patient’s refusal of further surgery, and chronic constipation. The risk of having a permanent stoma in patients with symptomatic anastomotic leakage was significantly higher compared with patients without symptomatic anastomotic leakage (56 per cent v. 11 per cent). Both the Dutch TME study and the Swedish RCT demonstrate the importance of consent at the time of original TME with regard to the possible permanence of a defunctioning stoma.

Risks and Complications of a Defunctioning Loop Ileostomy and its Subsequent Reversal Most reports on the morbidity of a defunctioning loop ileostomy, predictors of complications and

HEBK001-C15_p228-238.indd 236

non-closure of stoma, and complication rates following reversal have been retrospective reviews of a single-institution experience, with variable results. Chun and colleagues reported 123 patients who underwent a planned temporary defunctioning loop ileostomy performed at the time of a low rectal anastomosis.14 The primary outcome measures were the ileostomy complication rate for the entire spectrum of care, readmission and reoperation rates to treat ileostomy complications, and subsequent closure rate. Of these patients, 64.2 per cent experienced some form of morbidity during their care, the biggest group being patients readmitted for dehydration following ileostomy formation (11.4 per cent). The ileostomy was closed in 76.4 per cent of patients, with 8.6 per cent requiring a midline laparotomy (a non-closure rate of 23.6 per cent, which is higher than in the Dutch and Swedish groups). The overall ileostomy-related reoperation rate was 10.4 per cent (2.4 per cent during index hospitalization, 1.6 per cent at readmission, and 6.4 per cent following ileostomy closure). Obesity (body mass index $ 30 kg/m2) was associated with a higher overall ileostomy complication rate and outpatient complication rate. Age over 65 years and hypertension increased the risks of high ileostomy output and dehydration. Obesity and smoking decreased the likelihood of ileostomy closure. Akiyoshi and colleagues considered complications of stoma closure to evaluate the risks and benefits of a defunctioning stoma.15 Data were prospectively recorded in 125 consecutive patients who underwent an elective closure of loop ileostomy after primary rectal cancer resection. Postoperative complications developed in 21 patients (16.8 per cent), the majority being wound infections; ­ileus and anastomotic bleeding were also reported. There was no postoperative mortality. Risk factors for wound infection included male gender and surgical site infection after primary surgery. The mean length of postoperative hospital stay was significantly longer in patients with complications than in patients without complications. Luglio and colleagues reported the rate of complications after ileostomy reversal in 944 patients using standardized definitions and perioperative variables and 30-day outcomes.16 Anastomotic technique for reversal varied (sutured fold-over, 49.4 per cent; stapled, 33.4 per cent; hand-sewn end to end, 17.3 per cent). Patients who had sutured

08/02/13 5:29 PM

References  237

closure had longer operative times, increased times to bowel movement and initiation of soft diet, and longer time to discharge, but no difference in other variables. Overall, complications occurred in 203 patients (21.5 per cent), including 45 patients (4.8 per cent) who experienced a major complication. There was no post-reversal mortality. D’Haeninck and colleagues analysed 197 consecutive patients who underwent closure of a defunctioning loop ileostomy.17 Again, the method of closure varied (transverse closure of enterotomy, 75.6 per cent; segmental enterectomy with hand-sewn end-to-end anastomosis, 13.2 per cent; stapled sideto-side anastomosis, 11.2 per cent). Overall postoperative morbidity and mortality were 32.0 per cent and 0.5 per cent, respectively. The surgical complication rate was 30.5 per cent, including prolonged ileus (11.2 per cent), small bowel obstruction (4.1 per cent), anastomotic leak (3.0 per cent) and wound infection (4.6 per cent). Surgical complications were more frequent in male patients. Prolonged ileus was more frequent when the interval to stoma reversal exceeded 12 weeks. The incidence of complications was not influenced by the closure technique. Nineteen patients (9.6 per cent) required reoperation for anastomotic leak, wound infection, small bowel obstruction or repair of an incisional hernia. The complications following closure of a loop ileostomy are more frequent in male patients who are obese. There seems to be no significant difference between whether the closure is sutured or stapled. Although these risks should be borne in mind at the time of stoma construction, the risks are offset by the higher risks of anastomotic leakage following TME when not defunctioned, as documented in the Swedish study.7,13 Ideally, informed consent for TME should include discussions that specifically pertain to stoma-related complications.

CONCLUSION All colorectal anastomoses are at risk of leakage; the nearer the anastomosis to the anal verge, the higher the risk. A defunctioning stoma reduces the consequences (and perhaps the incidence) of leakage and reduces the requirement for emergency reoperation. Nevertheless, a stoma has complications at formation, in situ and at reversal, which should be

HEBK001-C15_p228-238.indd 237

borne in mind. Attention to detail and awareness of the risks and benefits favours consideration of routine defunctioning for all coloanal anastomoses after TME surgery.

References   1. Moran BJ, Heald RJ. Risk factors for and management of anastomotic leakage in rectal surgery. Colorectal Dis 2001; 3: 135–7.   2. Cartmell MT, Jones OM, Moran BJ, Cecil TD. A defunctioning stoma significantly prolongs the length of stay in laparoscopic colorectal resections. Surg Endosc 2008; 22: 2643–7.   3. Amin AI, Ramalingam T, Sexton R, Heald RJ, Leppington-Clarke A, Moran BJ. Comparison of transanal stent with defunctioning stoma in low anterior resection for rectal cancer. Br J Surg 2003; 90: 581–2.   4. Matthiessen P, Hallböök O, Andersson M, Rutegård J, Sjödahl R. Risk factors for anastomotic leakage after anterior resection of the rectum. Colorectal Dis 2004; 6: 462–9.   5. Ulrich AB, Seiler C, Rahbari N, Weitz J, Büchler MW. Diverting stoma after low anterior resection: more ­arguments in favor. Dis Colon Rectum 2009; 52: 412–8.   6. Tan WS, Tang CL, Shi L, Eu KW. Meta-analysis of defunctioning stomas in low anterior resection for rectal cancer. Br J Surg 2009; 96: 462–72.   7. Matthiessen P, Hallböök O, Rutegård J, Simert G, Sjödahl R. Defunctioning stoma reduces symptomatic anastomotic leakage after low anterior resection of the rectum for cancer: a randomized multicenter trial. Ann Surg 2007; 246: 207–14.   8. Pakkastie TE, Ovaska JT, Pekkala ES, Luukkonen PE, Järvinen HJ. A randomised study of colostomies in low colorectal anastomoses. Eur J Surg 1997; 163: 929–33.   9. Graffner H, Fredlund P, Olsson SA, Oscarson J, Petersson BG. Protective colostomy in low anterior resection of the rectum using the EEA stapling instrument: a randomized study. Dis Colon Rectum 1983; 26: 87–90. 10. Rondelli F, Reboldi P, Rulli A, et al. Loop ileostomy versus loop colostomy for fecal diversion after colorectal or coloanal anastomosis: a meta-analysis. Int J Colorectal Dis 2009; 24: 479–88. 11. Edwards DP, Leppington-Clarke A, Sexton R, Heald RJ, Moran BJ. Stoma-related complications are more frequent after transverse colostomy than loop ileostomy: a prospective randomized clinical trial. Br J Surg 2001; 88: 360–3. 12. Den Dulk, Smit M, Peeters KC, et al. A multivariate analysis of limiting factors for stoma reversal in patients with rectal cancer entered into the total

08/02/13 5:29 PM

238  Intestinal stoma and the role of defunctioning a low anastomosis after anterior resection mesorectal excision (TME) trial: a retrospective study. Lancet Oncol 2007; 8: 297–303. 13. Lindgren R, Hallböök O, Rutegård J, Sjödahl R, Matthiessen P. What is the risk for a permanent stoma after low anterior resection of the rectum for cancer? A six-year follow-up of a multicenter trial. Dis Colon Rectum 2011; 54: 41–7. 14. Chun LJ, Haigh PI, Tam MS, Abbas MA. Defunctioning loop ileostomy for pelvic anastomoses: predictors of morbidity and nonclosure. Dis Colon Rectum 2012; 55: 167–74.

HEBK001-C15_p228-238.indd 238

15. Akiyoshi T, Fujimoto Y, Konishi T, et al. Complications of loop ileostomy closure in patients with rectal tumor. World J Surg 2010; 34: 1937–42. 16. Luglio G, Pendlimari R, Holubar SD, Cima RR, Nelson H. Loop ileostomy reversal after colon and rectal surgery: a single institutional 5-year experience in 944 patients. Arch Surg 2011; 146: 1191–6. 17. D’Haeninck A, Wolthuis AM, Penninckx F, D’Hondt M, D’Hoore A. Morbidity after closure of a defunctioning loop ileostomy. Acta Chir Belg 2011; 111: 136–41.

08/02/13 5:29 PM

16 Quality of life in patients undergoing abdominoperineal excision and anterior resection for rectal cancer Peter How and Kandiah Chandrakumaran

Introduction Since the first description of the abdominoperineal excision (APE) by W. Ernest Miles in 1908,1 the surgical management of rectal cancer has evolved dramatically. Total mesorectal excision (TME), described by Heald and colleagues in 1982,2 has since been recognized as the gold standard treatment for most patients with rectal cancer. Other advances in surgical technique, including stapling devices, and multimodality treatment, have enabled anterior resection with sphincter preservation to become feasible for all but very distal rectal cancers. This has significantly reduced the number of patients burdened with a permanent stoma and the perceived limitations it imposes. Much of the impetus for this change was based on the belief that patients have a better quality of life (QOL) after sphincter preservation compared with patients with a permanent colostomy. This view is not supported universally, however, and remains a topic for debate. Over the past decade, interest has grown in patient-reported outcomes. The cultural shift within health care towards a more patient-centred approach has meant that assessment of treatment efficacy is no longer confined to classic biomedical indicators such as survival and mortality. Quality of life has emerged

HEBK001-C16_p239-248.indd 239

as an important outcome measure in its own right, and there is evidence for a positive relationship between QOL data or some QOL measures and survival in cancer patients.3

Quality-of-life assessment: general principles The term ‘quality of life’ intuitively evokes a sense of contentment that is influenced by economical, mental and physical status. Since the World Health Organization’s (WHO) definition of health in 1948 as ‘a state of complete physical, mental and social wellbeing and not merely the absence of disease or infirmity’, health-related quality of life (HRQOL) has come to represent a multidimensional construct pertaining to four distinct areas of wellbeing: physical, psychological, emotional and social. Attempts to measure HRQOL have focused largely upon these four domains with the use of specially designed questionnaires. These include, among others, patient-derived generic and cancer-specific modules such as the European Organization for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire – Core 30 (QLQ-C30) (Figure 16.1), EORTC QLQ-CR38,

08/02/13 5:28 PM

240  QOL in patients undergoing APE and anterior resection

EORTC QLQ-C30 (version 3) We are interested in some things about you and your health. Please answer all of the questions yourself by circling the number that best applies to you. There are no "right" or "wrong" answers. The information that you provide will remain strictly confidential. Please fill in your initials: Your birthdate (Day, Month, Year): Today's date (Day, Month, Year): 31 ___________________________________________________________________________ Not at A Quite Very All Little a Bit Much   1. Do you have any trouble doing strenuous activities, like carrying a heavy shopping bag or a suitcase?

1

2

3

4

  2.  Do you have any trouble taking a long walk?

1

2

3

4

  3. Do you have any trouble taking a short walk outside of the house?

1

2

3

4

  4.  Do you need to stay in bed or a chair during the day?

1

2

3

4

  5. Do you need help with eating, dressing, washing yourself or using the toilet?

1

2

3

4

During the past week:

Not at A Quite Very All Little a Bit Much

  6. Were you limited in doing either your work or other daily activities?

1

2

3

4

  7. Were you limited in pursuing your hobbies or other leisure time activities?

1

2

3

4

  8.  Were you short of breath?

1

2

3

4

  9.  Have you had pain?

1

2

3

4

10.  Did you need to rest?

1

2

3

4

11.  Have you had trouble sleeping?

1

2

3

4

12.  Have you felt weak?

1

2

3

4

13.  Have you lacked appetite?

1

2

3

4

14.  Have you felt nauseated?

1

2

3

4

15.  Have you vomited?

1

2

3

4

Figure 16.1. The European Organization for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire – Core 30 (QLQ-C30) questionnaire applicable to all cancers. Questions are grouped into functional scales, e.g. physical function (questions 1–5), symptom scales, e.g. pain (questions 9 and 19), and single items, e.g. insomnia (question 11). (Continues)

HEBK001-C16_p239-248.indd 240

08/02/13 5:28 PM

QUALITY-OF-LIFE ASSESSMENT: GENERAL PRINCIPLES  241

During the past week:

Not at A Quite Very All Little a Bit Much

16.  Have you been constipated?

1

2

3

4

17.  Have you had diarrhoea?

1

2

3

4

18.  Were you tired?

1

2

3

4

19.  Did pain interfere with your daily activities?

1

2

3

4

20. Have you had difficulty in concentrating on things, like reading a newspaper or watching television?

1

2

3

4

21.  Did you feel tense?

1

2

3

4

22.  Did you worry?

1

2

3

4

23.  Did you feel irritable?

1

2

3

4

24.  Did you feel depressed?

1

2

3

4

25.  Have you had difficulty remembering things?

1

2

3

4

26. Has your physical condition or medical treatment interfered with your family life?

1

2

3

4

27. Has your physical condition or medical treatment interfered with your social activities?

1

2

3

4

28. Has your physical condition or medical treatment caused you financial difficulties?

1

2

3

4

For the following questions please circle the number between 1 and 7 that best applies to you 29.  How would you rate your overall health during the past week?    1     2     3     4    5     6     7 Very poor           

  Excellent

30.  How would you rate your overall quality of life during the past week?    1      2      3      4     5     6      7 Very poor           

  Excellent

Figure 16.1. (Continued).

EORTC QLQ-CR29, short form 36 (SF-36), SF12, Functional Assessment of Cancer Therapy – Colorectal (FACT-C) and Functional Assessment of Cancer Therapy – General (FACT-G). Despite variation in content and format, certain principles and concepts remain central to all of these questionnaires, such as validity (measuring what it is intended to measure), reliability (measuring with

HEBK001-C16_p239-248.indd 241

sufficient precision), sensitivity (ability to detect changes), reproducibility (cross-cultural), simplicity (enabling self-administration), and easily interpretable (attaching clinical relevance). Furthermore, the questionnaires are intended to focus on dimensions that are influenced by health and can be quantified accurately and appropriately, according to its component indicators.4

08/02/13 5:28 PM

242  QOL in patients undergoing APE and anterior resection

Figure 16.2. Quality of life after surgery. APE, abdominoperineal excision; LAR, low anterior resection.

Quality of life and the preoperative pathway Although much of the literature regarding QOL in patients with rectal cancer focuses on the effects of surgery, most clinicians agree that appropriate support and counselling are vitally important from the time of diagnosis. The preoperative (pre-cancer resection) patient pathway is increasingly a complicated one, often encompassing multimodality treatment, restaging and, in some cases, surgery – all of which impact heavily upon QOL (Figure 16.2). Preoperative baseline QOL scores have been highlighted as prognostic indicators, and so it is important to provide support at an early stage. This should include stoma counselling, sex counselling, and counselling about potential faecal incontinence.

Stoma Counselling Often the most difficult and psychologically demanding hurdle to negotiate regarding rectal cancer surgery is the concept of a permanent stoma. It is common to encounter patients for whom such an idea is untenable. Retrospective studies show that patients who have access to stoma counselling before surgery enjoy a better quality of life postoperatively.5 In addition, accurate preoperative stoma siting and counselling by a stomatherapist helps to improve outcome in such patients.6 Currently, as befits a patient-tailored approach, there are no national guidelines pertaining to stoma

HEBK001-C16_p239-248.indd 242

counselling. It is common practice, however, to first establish the patient’s understanding of what a stoma is, followed by the introduction of a ‘starter pack’. This pack contains basic anatomical and functional information regarding what to expect both before and after surgery. It also includes stoma bags and related accessories to help prepare the patient, and their partner and family, for future daily maintenance. Education of the patient’s relatives and carers is beneficial and worthwhile. Up to half of patients (and their relatives) with an ileostomy and 30 per cent of patients with a colostomy have expressed dissatisfaction at the level of information they have received.7 Furthermore, up to 60 per cent of patients are poorly informed about stoma irrigation techniques,8 despite good evidence for its effectiveness in achieving faecal continence in patients with end colostomies.9 In addition, a greater proportion of patients are unhappy with their degree of participation within the decision-making process. It is clear that although the majority of patients speak highly of their stoma counselling experiences, there is a need for continually assessing and modifying this service, with the aim of improving general standards of quality of care.

Sex Counselling Sexual dysfunction remains one of the more longlasting side effects of pelvic cancer treatment. Sexual problems are associated with both poor physical health and emotional distress10 and are considered

08/02/13 5:28 PM

QUALITY-OF-LIFE OUTCOMES FOLLOWING SURGERY FOR RECTAL CANCER  243

to be a major determinant of global QOL. Despite the high risk of sexual dysfunction following rectal cancer surgery, sex counselling remains very much on the periphery of mainstream care. It has been suggested that an effective way to implement appropriate sex counselling is to provide routine screening and counselling for QOL issues related to cancer treatment. It is proposed that one person within the colorectal multidisciplinary team should be assigned the role of assessing and triaging patients for quality of life problems.11 This could be conducted via a brief 20- to 30-min interview centred upon validated questionnaires, which may be productive for clinical as well as research purposes. Sex counselling should include educating and warning patients about potential sexual problems arising from cancer treatment, helping patients overcome specific sexual problems (which may occur before surgery as a result of stress or neoadjuvant therapy), and preparing patients for physical changes that may adversely affect body image.12

Faecal Incontinence Counselling For patients looking to avoid a permanent stoma, many pay the price in terms of poor functional outcome, i.e. faecal incontinence. This is particularly true for patients with mid- and lower rectal tumours, where the closer the anastomosis to the anal canal, the poorer the functional outcome. Many patients prefer low anterior resection to APE, even if low anterior resection involves a risk of incontinence on a daily basis.13 Faecal incontinence rates have been reported to be as high as 61.5 per cent at 5 years following preoperative radiotherapy and TME.14 The potential need for sanitary towels and incontinence pads on a permanent basis combined with perianal excoriation is something that all patients undergoing restorative surgery should be made aware of. The aim is not to discourage patients before surgery but to provide realistic expectations of their likely postoperative quality of life. The possibility of long-term medication to help control symptoms and the option of further surgery as definitive treatment should also be made clear to give the patient an appropriate perspective on future events. Knowing that there are options if problems are encountered can help pre-empt fear and negativity in the lead-up to surgery and pro-

HEBK001-C16_p239-248.indd 243

mote a more positive outlook on the future. Unfortunately, precise risks of dysfunction have not been well documented, although research is ongoing.

Quality-of-life outcomes following surgery for rectal cancer The primary objective in managing rectal cancer is achieving satisfactory oncological clearance while maintaining QOL acceptable to the patient. Although much interest surrounds comparative QOL data between anterior resection and abdominoperineal excision, it is pertinent that there never has been, and now unlikely ever to be, a randomized controlled trial of APE compared with low anterior resection. Thus, comparative studies are difficult to interpret, and it is beneficial to look at aspects of QOL in the individual procedures to determine how each impacts upon QOL. It is also worth noting that patients treated with APE will have lower, more advanced tumours, and a higher percentage will receive neoadjuvant therapy,15 making direct comparisons with anterior resection more complex. Anterior resection has long been associated with superior oncological outcomes compared with APE,16 and a similar assumption is often made regarding QOL. In the following section, QOL outcomes following anterior resection and APE with regards to function and symptomatology are outlined. The results are interpreted to identify major determinants for good and poor outcomes in order to optimize treatment policy.

Functioning Scores Physical functioning has been shown to be better following anterior resection compared with APE in comparative studies.8,17 These findings are in keeping with a meta-analysis of several comparative studies, demonstrating statistically significantly higher physical function within the anterior resection group that also extends to patients with low rectal cancer.18 This is perhaps not surprising given the physical limitations imposed by a permanent stoma on being able to perform day-to-day tasks independently. Good agreement is generally observed between physical and role functions, and it is tempting to

08/02/13 5:28 PM

244  QOL in patients undergoing APE and anterior resection

speculate that if patients undergoing APE are physically limited by their stoma, then everyday work and leisure activities (role) may be affected adversely. Support for this theory comes from observations that QOL scores improve upon stoma reversal.19 This is not always a true reflection of functional status, however, as significant improvement in role functioning has been reported despite severe diarrhoea following stoma reversal.8 Indeed, despite significant improvements in body image and leisure activities, stoma reversal has not always resulted in improved QOL, while significant worsening of gastrointestinal problems may persist for up to a year following surgery.20 Interestingly, consistent data suggest equivalent8,21 or better emotional and cognitive function in patients undergoing APE.18 Reasons for this may include the definitive nature of the surgery, whereby the patient feels that by removing both the rectum and anus, there is less chance of recurrence and no further invasive surveillance is required per rectum. Patients undergoing anterior resection, on the other hand, remain troubled by the possibility of recurrence and further surgery. Although there is general agreement that fatigue, pain, nausea and dyspnoea are comparable between anterior resection and APE, there is conflicting evidence regarding gastrointestinal tract symptomatology. Much of the earlier literature within the TME era that reported inferior global QOL scores in patients undergoing APE demonstrated higher levels of gastrointestinal dysfunction (constipation, diarrhoea) compared with patients undergoing low anterior resection, including defecation problems.8 More recent studies, however, have demonstrated no differences in either global QOL or gastrointestinal dysfunction between the two groups,21 with some reporting higher scores for gastrointestinal dysfunction in the anterior resection group.17 It is difficult to explain these differences given that many of these studies employed similar methods using identical questionnaires. As patient demographics, tumour stage and follow-up time are well documented and broadly comparable, other factors should be considered. One reason may be due to the improved level and intensity of preoperative stoma counselling. Within the UK, it is now common for multidisciplinary teams to have both colorectal nurse specialists (offering general support and guidance) and stoma care nurses (offer-

HEBK001-C16_p239-248.indd 244

ing specific stoma counselling), reflecting increased specialization within nursing services. The importance of this is highlighted by the fact that stomarelated complications often do not impact upon how patients perceive their quality of care compared with the amount of information they receive, participation in decision-making and opportunity to discuss sexual matters.22 It would be reasonable to extrapolate this towards how they regard their QOL. Alternatively, equivalent results may be a result of adaptation or response shift, whereby patients who have survived life-threatening disease appear to have new internal standards and are more willing to tolerate the adverse effects of either surgery, i.e. having a stoma or incontinence. Such a phenomenon may also account for comparable QOL scores observed both pre- and postoperatively in previously reported and emerging studies.23 Despite improvements in APE global QOL scores, it is notable that body image is consistently reported to be higher in patients undergoing anterior resection, although this is not always statistically significant. Unlike functional scores, this aspect of rectal cancer surgery appears resistant to any form of counselling or psychotherapy and is probably a contributing factor to the inferior sexual function observed in patients undergoing APE. Interestingly, body image is significantly impaired in both sexes, although women demonstrate a relatively lower sex drive preoperatively, while in the months following surgery both men and women are affected negatively by a stoma.24 Such studies have called for greater attention to be paid to preoperative patient counselling with regard to body image and sexuality.

Sexual Dysfunction Following Rectal Cancer Surgery One critical aspect of TME, introduced by Heald and colleagues, is the meticulous preservation of the pelvic autonomic nerves, which has resulted in a reduction in sexual dysfunction following rectal cancer surgery. Sexual dysfunction continues to be a major complication following anterior resection, and APE however, with rates varying between 10 per cent and 60 per cent.25 This may be compounded further by neoadjuvant and adjuvant radiotherapy.26

08/02/13 5:28 PM

Quality-of-life outcomes following surgery for rectal cancer  245

There is widespread agreement within the literature that APE carries a higher risk of sexual dysfunction than anterior resection. One feature of this is altered body image associated with a permanent colostomy, described previously. Other reasons accounting for this, particularly regarding male sexual dysfunction, are a higher proportion of T4 tumours, lower rectal tumours, radiotherapy, comorbidity, and increased risk of nerve damage during surgery given the extent of the dissection and significant alteration in pelvic anatomy.27 With respect to physiological sexual dysfunction, it is perhaps best to focus on combined genitourinary function as a reflection of autonomic nerve preservation. Enker and colleagues demonstrated that 57 per cent of patients undergoing APE compared with 85 per cent of patients undergoing restorative surgery were able to maintain both urinary and sexual function.28 Similar emerging data on sexual function from our own multicentre prospective study echo these sentiments, despite global QOL scores remaining broadly equivalent.

Factors Affecting Functional Outcome Following Anterior Resection Level of anastomosis

Functional outcome is thought to be worse the lower the anastomosis because of smaller reservoir volume. It has been demonstrated that lower anastomotic level tends towards increased frequency of incontinence for solid stool and gas.21 These findings correlate well with the demonstration of more frequent or painful bowel movements in patients undergoing low anterior resection.8 Interestingly, a higher composite score for incontinence is not always reflected in QOL scores between patients with high and low anastomosis. This may be a reflection of the response shift phenomenon described earlier, whereby patients undergoing low anterior resection have lower expectations of their outcome compared with patients undergoing high anterior resection. It is not always correct, therefore, to assume that gastrointestinal symptomatology, depending on the level of the anastomosis, will affect QOL, although this does remain a possible reason for the observed variability between studies comparing patients with and without a stoma.

HEBK001-C16_p239-248.indd 245

Neorectal reservoir

If restorative surgery is possible for low rectal tumours, many surgeons perform a circular transanal double-stapled low colorectal or coloanal anastomosis, commonly without a pouch. Such patients often end up having poor function, particularly with regard to continence due to impaired neorectal function. Development of the colonic J-pouch anal anastomosis29,30 has been reported to reduce stool frequency and defecation urgency.31 Systematic reviews and meta-analyses of randomized controlled trials have reported improved function with colonic J-pouch anal anastomosis compared with a conventional straight coloanal anastomosis,32–35 suggesting better QOL scores for colonic J-pouch anal anastomosis.36 Despite good evidence for the J-pouch improving outcomes and QOL, the technique is not always possible. For example, a very narrow pelvis, a fatty mesentery or restrictions in length may mean this technique is impossible to perform in certain patients. Encouragingly, similar functional outcomes have been observed in side-to-end anastomosis,37 which is technically easier to perform and is a useful alternative when there is inadequate bowel length for a J-pouch. As an additional option, the technique of coloplasty has been developed with a similar aim of improving outcome and QOL. Coloplasty follows the principles of pyloroplasty in that a longitudinal incision is closed transversely and the reconstruction is completed by a stapled anastomosis or a hand-sewn end-to-end pouch anal anastomosis. Quality-of-life studies based on SF-36 have demonstrated better scores for coloplasty compared with straight anastomosis, including fewer bowel movements and less antidiarrhoeal medication,38 and equivalent functional outcome to colonic J-pouch anal anastomosis and side-to-end anastomosis.39,40 Radiotherapy

There is no doubt that the systemic effects of radiotherapy have a significant detrimental effect on QOL. Prospective studies using EORTC modules assessing the effects of neoadjuvant radiotherapy have shown significantly increased scores for diarrhoea, fatigue and appetite loss and worse outcomes for physical function, social function and global QOL.41 As QOL scores tend to return to pretreatment levels 4–6 weeks after radiotherapy, however,

08/02/13 5:28 PM

246  QOL in patients undergoing APE and anterior resection

it is perhaps more important to focus upon longterm local effects. Radiotherapy, by way of fibrosis, is purported to reduce compliance of the rectum, resulting in reduced reservoir function. It is also thought that radiotherapy-induced fibrosis affecting the myenteric plexus prevents adequate closure of the anal canal in its resting state. Preoperative radiotherapy, whether long or short course, has been associated with higher levels of frequency and urgency42 and faecal incontinence14,43,44 when assessed postoperatively in randomized controlled trial settings. A systematic review outlined similar late adverse effects related to radiotherapy in the treatment of rectal cancer.45 In addition, daily urinary incontinence has been reported more frequently after radiotherapy, with a similar adverse effect on social function.43 The benefits of radiotherapy in reducing local recurrence have to be balanced against the adverse effects of radiotherapy on bowel function and QOL.46 The most effective strategy to help improve this aspect of patient care would appear to lie in careful patient selection, so that substantial overtreatment is avoided.

Low rectal cancer surgery and quality of life: the heart of the matter Given that patients with high and mid-rectal tumours routinely undergo restorative surgery (except for patients with poor preoperative function), it is patients with distal rectal cancer (commonly defined as within 6 cm of the anal verge) who present the greatest challenge. This includes the technical challenges of low rectal cancer and the difficult decision as to whether to perform anterior resection or APE. Increasingly, it is oncologically feasible to perform either operation; provided the patient has been continent before the onset of the cancer and has no strong objections to a permanent stoma, the surgeon must decide which operation will give the best outcome. Techniques such as anal manometry have been suggested as reliable predictors of functional outcome, but evidence for this is inconclusive. Given that low rectal cancers are associated with more postoperative complications and poorer anorectal and sexual

HEBK001-C16_p239-248.indd 246

function47 and that patients with low rectal cancer have worse QOL than patients with upper or middle rectal cancers, further information on surgical options and outcomes in low rectal cancer would be helpful. Ideally, good-quality prospectively collected QOL data using established questionnaires would represent progress on this front, and this is currently in progress. There is an increasing interest in comparative QOL for APE and anterior resection for low rectal cancer, with several studies reporting interesting conclusions. Among these are better sexual functioning scores observed in patients undergoing APE (both genders) and worse scores for gastrointestinal symptoms in patients undergoing anterior resection.17 This has not impacted on the general trend of equivalent overall QOL outcomes, however (albeit with certain specific differences), even within larger studies.48 More recent studies reporting inferior QOL in patients with permanent stomas remain fewer and are generally hampered by smaller sample sizes.49 Although it is difficult to draw firm conclusions in the face of such contradictory evidence, there are variations between reported studies including, in some, a lack of preoperative QOL data, single institution involvement and response bias. In addition, length of follow-up is important given that patients undergoing restorative surgery commonly suffer from anterior resection syndrome (frequency, urgency, stool fragmentation, incontinence), which may persist beyond 1 year.50 Furthermore, there is often little or no information regarding the surgical technique (e.g. level and type of anastomosis, presence of pouch) or the level of preoperative counselling, factors known to influence outcome.

Conclusion Interest in QOL outcomes has grown significantly over recent years and is likely to continue, in keeping with growing importance placed on patientreported outcome measures. The remit of the surgeon is no longer only to remove cancer, but also to ensure the best possible outcome in keeping with the patient’s wishes and beliefs. This has important implications regarding the selection of patients for multimodality treatment and technical factors relating to surgical technique. Although every patient should be considered on an individual basis,

08/02/13 5:28 PM

REFERENCES  247

i­ nformation, advice and counselling must be centred on a strong collective evidence base. Long-term QOL prospective studies using internationally validated questionnaires currently represent the most assured way of achieving this.

References   1. Miles WE. A method of performing abdomino-perineal excision for carcinoma of the rectum and the terminal portion of the pelvic colon. Lancet 1908; 2: 1812–3.   2. Heald RJ, Husband EM, Ryall RD. The mesorectum in rectal cancer surgery: the clue to pelvic recurrence? Br J Surgery 1982; 69: 613–6.   3. Montazeri A. Quality of life data as prognostic indicators of survival in cancer patients: an overview of the literature from 1982 to 2008. Health Qual Life Outcomes 2009; 7: 102.   4. Lam CLK. Subjective quality of life measures: general principles and concepts. In Preedy VR (ed.). Handbook of Disease Burdens and Quality of Life Measures 1. Berlin, Springer-Verlag, 2009: 381–9.   5. Bass EM, Del Pino A, Tan A, Pearl RK, Orsay CP, Abcarian H. Does preoperative stoma marking and education by the enterostomal therapist affect outcome? Dis Colon Rectum 1997; 40: 440–2.   6. Crooks S. Foresight that leads to improved outcome: stoma care nurses’ role in siting stomas. Prof Nurse 1994; 10: 89–92.   7. Persson E, Gustavsson B, Hellström AL, Fridstedt G, Lappas G, Hultén L. Information to the relatives of people with ostomies: is it satisfactory and adequate? J Wound Ostomy Continence Nurs 2005; 32: 238–45.   8. Engel J, Kerr J, Schlesinger-Raab A, Eckel R, Sauer H, Hölzel D. Quality of life in rectal cancer patients: a fouryear prospective study. Ann Surg 2003; 238: 203–13.   9. Karadag A, Mentes BB, Ayaz S. Colostomy irrigation: results of 25 cases with particular reference to quality of life. J Clin Nurs 2005; 14: 479–85. 10. Laumann EO, Paik A, Rosen RC. Sexual dysfunction in the United States: prevalence and predictors. J Am Med Assoc 1999; 281: 537–44. 11. Schover LR. Sexuality and fertility after cancer. Hematology Am Soc Hematol Educ Program 2005; 523–7. 12. Schover LR. Sexual rehabilitation after treatment for prostate cancer. Cancer 1993; 71(3 suppl.): 1024–30. 13. Bossema E, Stiggelbout A, Baas-Thijssen M, van de Velde C, Marijnen C. Patients’ preferences for low rectal cancer surgery. Eur J Surg Oncol 2008; 34: 42–8. 14. Lange MM, den Dulk M, et al. Risk factors for faecal incontinence after rectal cancer treatment. Br J Surg 2007; 94: 1278–84. 15. Shihab OC, Brown G, Daniels IR, Heald RJ, Quirke P, Moran BJ. Patients with low rectal cancer treated by abdominoperineal excision have worse tumors and

HEBK001-C16_p239-248.indd 247

higher involved margin rates compared with patients treated by anterior resection. Dis Colon Rectum 2010; 53: 53–6. 16. Heald RJ, Smedh RK, et al. Abdominoperineal excision of the rectum: an endangered operation. Dis Colon Rectum 1997; 40: 747–51. 17. Kasparek MS, Hassan I, Cima RR, Larson DR, Gullerud RE, Wolff BG. Quality of life after coloanal anastomosis and abdominoperineal resection for distal rectal cancers: sphincter preservation vs quality of life. Colorectal Dis 2011; 13: 872–7. 18. Cornish JA, Tilney HS, Heriot AG, Lavery IC, Fazio VW, Tekkis PP. A meta-analysis of quality of life for abdominoperineal excision of rectum versus anterior resection for rectal cancer. Ann Surg Oncol 2007; 14: 2056–68. 19. Camilleri-Brennan J, Steele RJ. Prospective analysis of quality of life after reversal of a defunctioning loop ileostomy. Colorectal Dis 2002; 4: 167–71. 20. Siassi M, Hohenberger W, Lösel F, Weiss M. Quality of life and patient’s expectations after closure of a temporary stoma. Int J Colorectal Dis 2008; 23: 1207–12. 21. Guren MG, Eriksen MT, Wiig JN, et al. Quality of life and functional outcome following anterior or abdominoperineal resection for rectal cancer. Eur J Surg Oncol 2005; 31: 735–42. 22. Persson E, Gustavsson B, Hellström AL, Lappas G, Hultén L. Ostomy patients’ perceptions of quality of care. J Adv Nurs 2005; 49: 51–8. 23. How P, Stelzner S, Branagan G, et al. Comparative quality of life in patients following low anterior resection and abdominoperineal excision for low rectal cancer: preliminary results from a prospective multicentre study. Dis Colon Rectum 2010; 55: 400–6. 24. Burghofer K, Jauch KW. [Sexuality and body image in patients with rectal cancer]. MMW Fortschr Med 2006; 148: 36–7. 25. Banerjee AK. Sexual dysfunction after surgery for rectal cancer. Lancet 1999; 353: 1900–2. 26. Mannaerts GH, Schijven MP, Hendrikx A, Martijn H, Rutten HJ, Wiggers T. Urologic and sexual morbidity following multimodality treatment for locally advanced primary and locally recurrent rectal cancer. Eur J Surg Oncol 2001; 27: 265–72. 27. Hassan I, Cima RR. Quality of life after rectal resection and multimodality therapy. J Surg Oncol 2007; 96: 684–92. 28. Enker WE, Havenga K, Polyak T, Thaler H, Cranor M. Abdominoperineal resection via total mesorectal excision and autonomic nerve preservation for low rectal cancer. World J Surg 1997; 21: 715–20. 29. Parc R, Berger A, Tiret E, Frileux P, Nordlinger B, Hannoun L. [Colo-anal anastomosis with reservoir in the treatment of rectal cancer]. Ann Gastroenterol Hepatol (Paris) 1987; 23: 329–31. 30. Lazorthes F, Fages P, Chiotasso P, Lemozy J, Bloom E. Resection of the rectum with construction of a

08/02/13 5:28 PM

248  QOL in patients undergoing APE and anterior resection colonic reservoir and colo-anal anastomosis for carcinoma of the rectum. Br J Surg 1986; 73: 136–8. 31. Hallböök O, Hass U, Wänström A, Sjödahl R. Quality of life measurement after rectal excision for cancer: comparison between straight and colonic J-pouch anastomosis. Scand J Gastroenterol 1997; 32: 490–3. 32. Brown CJ, Fenech DS, McLeod RS. Reconstructive techniques after rectal resection for rectal cancer. Cochrane Database Syst Rev 2008; (2): CD006040. 33. Koh PK, Tang CL, Eu KW, Samuel M, Chan E. A systematic review of the function and complications of colonic pouches. Int J Colorectal Dis 2007; 22: 543–8. 34. Heriot AG, Tekkis PP, Constantinides V, et al. Metaanalysis of colonic reservoirs versus straight coloanal anastomosis after anterior resection. Br J Surg 2006; 93: 19–32. 35. Peng J, Zhan W, Zhao X, et al. [Comparison of colonic J-pouch and straight coloanal anastomosis after low anterior resection for rectal carcinoma: a meta-analysis of 8 randomized trials]. Zhonghua Wai Ke Za Zhi 2002; 40: 905–8. 36. Gosselink MP, Busschbach JJ, Dijkhuis CM, Stassen LP, Hop WC, Schouten WR. Quality of life after total mesorectal excision for rectal cancer. Colorectal Dis 2006; 8: 15–22. 37. Machado M, Nygren J, Goldman S, Ljungqvist O. Functional and physiologic assessment of the colonic reservoir or side-to-end anastomosis after low anterior resection for rectal cancer: a two-year follow-up. Dis Colon Rectum 2005; 48: 29–36. 38. Remzi FH, Fazio VW, Gorgun E, et al. Quality of life, functional outcome, and complications of coloplasty pouch after low anterior resection. Dis Colon Rectum 2005; 48: 735–43. 39. Matsuoka H, Masaki T, Kobayashi T, Sato K, Sugiyama M, Atomi Y. Comparison of functional and clinical outcomes: colonic J-pouch vs. coloplasty in patients with low rectal cancer. Hepatogastroenterology 2010; 57: 70–72. 40. Liao C, Gao F, Cao Y, Tan A, Li X, Wu D. Meta-analysis of the colon J-pouch vs transverse coloplasty pouch

HEBK001-C16_p239-248.indd 248

after anterior resection for rectal cancer. Colorectal Dis 12: 624–31. 41. Guren MG, Dueland S, Skovlund E, Fossa SD, Poulsen JP, Tveit KM. Quality of life during radiotherapy for rectal cancer. Eur J Cancer 2003; 39: 587–94. 42. Parc Y, Zutshi M, Zalinski S, Ruppert R, Fürst A, Fazio VW. Preoperative radiotherapy is associated with worse functional results after coloanal anastomosis for rectal cancer. Dis Colon Rectum 2009; 52: 2004–14. 43. Bruheim K, Guren MG, Skovlund E, et al. Late side effects and quality of life after radiotherapy for rectal cancer. Int J Radiat Oncol Biol Phys 2010; 76: 1005–11. 44. Murata A, Brown CJ, Raval M, Phang PT. Impact of short-course radiotherapy and low anterior resection on quality of life and bowel function in primary rectal cancer. Am J Surg 2008; 195: 611–5. 45. Birgisson H, Påhlman L, Gunnarsson U, Glimelius B. Late adverse effects of radiation therapy for rectal cancer: a systematic overview. Acta Oncol 2007; 46: 504–16. 46. Sebag-Montefiore D, Stephens RJ, Steele R, et al. Preoperative radiotherapy versus selective postoperative chemoradiotherapy in patients with rectal cancer (MRC CR07 and NCIC-CTG C016): a multicentre, randomized trial. Lancet 2009; 373: 811–20. 47. Hoerske C, Weber K, Goehl J, Hohenberger W, Merkel S. Long-term outcomes and quality of life after rectal carcinoma surgery. Br J Surg 2010; 97: 1295–303. 48. Tilney H, Fazio V, Lavery I, et al. Comparison of longterm survival and quality of life following restorative versus non-restorative surgery for lower third rectal cancer. Dis Colon Rectum 2007; 50: 703–4. 49. Fucini C, Gattai R, Urena C, Bandettini L, Elbetti C. Quality of life among five-year survivors after treatment for very low rectal cancer with or without a permanent abdominal stoma. Ann Surg Oncol 2008; 15: 1099–106. 50. Kakodkar R, Gupta S, Nundy S. Low anterior resection with total mesorectal excision for rectal cancer: functional assessment and factors affecting outcome. Colorectal Dis 2006; 8: 650–56.

08/02/13 5:28 PM

HEBK001-Index.indd 268

08/02/13 5:22 PM