Colorectal Cancer (2006)

Colorectal Cancer �� Edited by

Jim Cassidy

University of Glasgow Scotland, U.K.

Patrick Johnston Queen's University Belfast Belfast, Northern Ireland, U.K.

Eric Van Cutsem

University Hospital Gasthuisberg Leuven, Belgium

New York London

DK3439_C000a.indd 3

08/02/2006 11:49:26 AM

Informa Healthcare USA, Inc. 270 Madison Avenue New York, NY 10016 © 2007 by Informa Healthcare USA, Inc. Informa Healthcare is an Informa business No claim to original U.S. Government works Printed in the United States of America on acid‑free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number‑10: 0‑8247‑2835‑1 (Hardcover) International Standard Book Number‑13: 978‑0‑8247‑2835‑9 (Hardcover) This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. 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‑for‑profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging‑in‑Publication Data Colorecal cancer / edited by Jim Cassidy, Patrick Johnston, Eric Van Cutsem. p. ; cm. Includes bibliographical references and index. ISBN‑13: 978‑0‑8247‑2835‑9 (alk. paper) ISBN‑10: 0‑8247‑2835‑1 alk. paper) 1. Colon (Anatomy)‑‑Cancer. 2. Rectum‑‑Cancer. I. Cassidy, Jim, 1958‑ II. Johnston, Patrick G., MD. III. Cutsem, Eric van. [DNLM: 1. Colorectal Neoplasms. WI 529 C71903 2006] RC280.C6C652 2006 616.99’4347‑‑dc22

2006048591

Visit the Informa Web site at www.informa.com and the Informa Healthcare Web site at www.informahealthcare.com

DK3439_C000a.indd 4

08/02/2006 11:49:26 AM

Dedication

,

I would like to dedicate this book to Sandi. Jim Cassidy

Preface

,

Colorectal cancer is a common disease in all developed nations of the World. Until about 10 to 15 years ago, it was largely a surgical disease, with chemotherapy or radiotherapy having little impact on survival. Few oncologists even treated the disease, and even fewer would admit to specialist interest in a disease that was considered ‘‘refractory.’’ We also had more knowledge of the underlying molecular events in this disease than in most other solid cancers of adults, but this had not been translated into any meaningful clinical interventions. However, the last two decades have seen advances in chemotherapy, radiotherapy, and use of molecular targeted agents that we could not have dreamed of 20 years ago. In the same timescale, we have made further and important advances in our knowledge and understanding of the genetics, molecular biology, and pathophysiology of this condition. This knowledge has been the driver for some of the clinical advances and promises to be important in the next generation of drugs that we hope will further improve outcomes. For example, it is not unreasonable to hypothesize that the prospect of chemopreventive agents for this disease can be realized within this generation. Our rapidly expanding knowledge base and the therapeutic advances based on it mean it is hard to keep up-to-date. This applies to both scientists trying to update therapy paradigms and clinicians trying to keep abreast of the molecular aspects. This book was designed to address these issues in a way that was accessible to the specialist as well as the generalist. This is important because some of the advances in colorectal cancer should be v

vi

Preface

applicable to other common malignancies that are currently regarded as ‘‘refractory.’’ The editorial team is currently based entirely in Europe, but all of us have the experience of practice in other nations and have been intimately involved in international research projects and clinical studies. We felt the time was ripe to produce a textbook that would update all aspects of our understanding in one publication. Thus, the goal was to produce a book that we would all like to own which would equip us enough to have informed discussion and conversation with experts within our own areas of expertise. This will be the key to making the molecular information that we are gaining at a frightening rate translate into clinical advances. We have collated contributions from world-class clinical and basic scientists in this book. Each was handpicked to represent their own disciplines and for their ability to describe complex issues in an understandable way. Each was tasked with producing text that was as comprehensive and up-to-date as possible in a rapidly moving field. There was a deliberate effort made to ‘‘globalize’’ the content of this book in order to maintain its relevance throughout a broad-based readership. The scope of the book encompasses risk factors, epidemiology, molecular pathogenesis, surgery, radiotherapy, chemotherapy, and even aspects of palliative care related to colorectal cancer. This text should serve as a useful and comprehensive reference for all who have an interest in colorectal cancer. In addition, it is designed so that the non-specialist can pick it up and access the topics that were of most interest to them without feeling overwhelmed. We thank our fellow editors and all the contributors for their diligence and help in the production of this book. Jim Cassidy Patrick Johnston Eric Van Cutsem

Contents

Preface . . . . v Contributors . . . . xiii 1. Genetic Susceptibility to Colorectal Cancer . . . . . . . . . . . . 1 Rebecca A. Barnetson and Malcolm G. Dunlop Introduction . . . . 1 Autosomal Dominant Disorders . . . . 2 Familial Adenomatous Polyposis . . . . 15 Other Defined Dominant Colorectal Cancer Susceptibility Syndromes . . . . 20 Autosomal Recessive Disorders . . . . 21 Low Penetrance Variants . . . . 24 Conclusions . . . . 30 References . . . . 31 2. Epidemiology of Colorectal Cancer . . . . . . . . . . . . . . . . . Julian Little and Linda Sharp Introduction . . . . 43 Descriptive Epidemiology of Colorectal Cancer . . . . 43 Groups at Increased Risk of Colorectal Cancer . . . . 46 Risk Factors for Colorectal Neoplasia . . . . 47 Conclusion . . . . 59 References . . . . 60 vii

43

viii

Contents

3. Colorectal Cancer Screening . . . . . . . . . . . . . . . . . . . . . . Robert J. C. Steele Introduction . . . . 77 Principles of Screening . . . . 78 Colorectal Cancer as a Suitable Target for Screening . . . . 80 Fecal Occult Blood Screening . . . . 82 Flexible Sigmoidoscopy . . . . 85 Colonoscopy . . . . 86 Radiology . . . . 88 Comparative Studies . . . . 88 Harm Caused by Screening . . . . 89 Economics of Screening . . . . 91 Novel Approaches to Screening . . . . 93 Conclusions . . . . 97 References . . . . 97 4. Pathology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Christian Wittekind Introduction . . . . 103 Definition . . . . 103 Site Distribution . . . . 104 Gross Morphology . . . . 104 Histomorphology . . . . 105 Special Clinical Types of Colorectal Cancer . . . . 107 Classification of Anatomical Extent Before Treatment . . . . 110 Classification of Anatomical Extent After Treatment . . . . 113 Histological Grading of Tumor Regression . . . . 114 Prognostic Factors . . . . 114 The Histopathological Report . . . . 114 Quality Management Within Pathology Departments . . . . 116 Pathology Findings in Resection Specimens Indicative of Oncological Quality of Surgery . . . . 116 Quality Assurance of Clinical Trials on Adjuvant and Neoadjuvant Therapy: The Surgical Pathologist’s Point of View . . . . 119 Malignant Tumors Other than Carcinomas . . . . 120 References . . . . 121

77

103

Contents

5. Familial Cancer Management . . . . . . . . . . . . . . . . . . . . Sabine Tejpar Introduction . . . . 125 Main Hereditary Colorectal Cancer Syndromes . . . . 127 Peutz–Jeghers Syndrome . . . . 139 Juvenile Polyposis . . . . 141 Cowden Syndrome or PTEN Hamartoma Syndrome . . . . 143 Familial Gastric Cancer . . . . 144 Familial Pancreatic Cancer . . . . 144 Hereditary Mixed Polyposis . . . . 144 Familial Colorectal Cancer . . . . 145 Aspects of Genetic Testing for GI Cancer Susceptibility . . . . 146 References . . . . 148 6. The Surgical Principles of Managing Colorectal Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ian R. Daniels and Richard J. Heald Introduction . . . . 151 The ‘‘Embryological Approach’’ to Rectal Cancer . . . . 152 The Multidisciplinary Approach . . . . 152 The Principles of Rectal Cancer Excision: Total Mesorectal Excision . . . . 152 Recurrence and Survival . . . . 156 Applying the Principles of Total Mesorectal Excision to Colonic Cancer . . . . 158 Conclusion . . . . 159 References . . . . 159 7. Adjuvant Therapy for Colorectal Cancer . . . . . . . . . . . . Geoff Chong and David Cunningham Introduction . . . . 163 Adjuvant Therapy for Stage III Colorectal Cancer . . . . 164 Adjuvant Therapy for Stage II Colorectal Cancer . . . . 173 Newer Agents for Adjuvant Therapy . . . . 176 Prognostic Factors and Adjuvant Therapy . . . . 183

ix

125

151

163

x

Contents

Recommendations for Adjuvant Therapy . . . . 185 Conclusions . . . . 186 References . . . . 187 8. The Role of Radiotherapy in the Treatment of Rectal Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rob Glynne-Jones and Rob Hughes Background . . . . 195 Indications for Adjuvant Radiotherapy . . . . 197 Risk Factors for Local Recurrence . . . . 198 The Evidence Base for Adjuvant Radiotherapy in Resectable Rectal Cancer . . . . 201 The Rationale for Short Course Preoperative Radiotherapy . . . . 203 Short Course Preoperative Radiotherapy vs. Surgery Alone or Postoperative Radiotherapy . . . . 204 More Recent Trials Performed with Chemoradiation . . . . 205 Fixed/Unresectable Rectal Cancer . . . . 208 Intraoperative Radiotherapy . . . . 209 Early Rectal Cancer . . . . 209 Pretreatment Clinical Assessment . . . . 211 Radiotherapy Planning Techniques . . . . 212 Acute Toxicity and Supportive Care During Radiotherapy . . . . 214 Timing of Surgery Following Preoperative Radiotherapy . . . . 215 Surgical Complications . . . . 216 Late Effects . . . . 216 Conclusion . . . . 217 References . . . . 219 9. The Treatment of Metastatic Colorectal Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eric Van Cutsem and Leonard Saltz Introduction . . . . 229 Cytotoxic Agents in Metastatic CRC . . . . 230

195

229

Contents

xi

Targeted Therapies for Metastatic CRC . . . . 236 References . . . . 244 10. Hepatic Directed Therapy . . . . . . . . . . . . . . . . . . . . . . Gregory D. Leonard and Nancy E. Kemeny Introduction . . . . 253 Hepatic Arterial Infusion Chemotherapy . . . . 254 Portal Vein Infusion Chemotherapy . . . . 271 Isolated Hepatic Perfusion Chemotherapy . . . . 272 Conclusion . . . . 274 References . . . . 275

253

11. Pharmacogenomics of Colorectal Cancer . . . . . . . . . . . . 287 Patrick Johnston, Wendy L. Allen, and Howard L. McLeod Introduction . . . . 287 Loss of Heterozygosity . . . . 288 Microsatellite Instability . . . . 289 TGFb II Mutation . . . . 290 5-Fluorouracil . . . . 291 Oxaliplatin . . . . 296 Irinotecan . . . . 298 Novel Therapies . . . . 299 Vascular Endothelial Growth Factor . . . . 301 New Technologies . . . . 302 Conclusions . . . . 306 References . . . . 306 12. Drug Development for Advanced Colorectal Cancer in the United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317 Igor Puzanov and Mace L. Rothenberg Introduction . . . . 317 Development of 5-FU and 5-FU/LV as First-Line Chemotherapy . . . . 318 Development of Capecitabine as an Oral Alternative to 5-FU . . . . 319 Development of Irinotecan as Second-Line Therapy . . . . 320 Integration of Irinotecan into Front-Line Chemotherapy . . . . 321 Oxaliplatin: Initial Data on Front-Line Therapy . . . . 322

xii

Contents

Oxaliplatin: Demonstration of Second-Line Efficacy . . . . 323 N9741: Comparison of First-Line Combination Regimens . . . . 324 Bevacizumab . . . . 325 Cetuximab . . . . 327 Regulatory Considerations of the U.S. FDA in the Approval of New Drugs for Treatment of Advanced Colorectal Cancer . . . . 328 Endpoints for New Drug Approval for Colorectal Cancer in the United States . . . . 329 How Will New Drugs be Developed for Colorectal Cancer in the Future? . . . . 331 References . . . . 333 Index . . . . 337

Contributors

Wendy L. Allen Drug Resistance Group, Centre for Cancer Research and Cell Biology, Queen’s University Belfast, Belfast, Northern Ireland, U.K. Rebecca A. Barnetson Colon Cancer Genetics Group, University of Edinburgh, School of Molecular and Clinical Medicine and MRC Human Genetics Unit, Western General Hospital, Edinburgh, U.K. Geoff Chong Department of Medicine, Royal Marsden Hospital, Sutton, Surrey, U.K. David Cunningham Department of Medicine, Royal Marsden Hospital, Sutton, Surrey, U.K. Ian R. Daniels Pelican Cancer Foundation, North Hampshire Hospital, Basingstoke, U.K. Malcolm G. Dunlop Colon Cancer Genetics Group, University of Edinburgh, School of Molecular and Clinical Medicine and MRC Human Genetics Unit, Western General Hospital, Edinburgh, U.K. Rob Glynne-Jones Mount Vernon Centre for Cancer Treatment, Northwood, Middlesex, U.K. Richard J. Heald Pelican Cancer Foundation, North Hampshire Hospital, Basingstoke, U.K.

xiii

xiv

Contributors

Rob Hughes Mount Vernon Centre for Cancer Treatment, Northwood, Middlesex, U.K. Patrick Johnston Drug Resistance Group, Centre for Cancer Research and Cell Biology, Queen’s University Belfast, Belfast, Northern Ireland, U.K. Nancy E. Kemeny Department of Gastrointestinal Oncology, Memorial Sloan-Kettering Cancer Center, New York, New York, U.S.A. Gregory D. Leonard Department of Gastrointestinal Oncology, Memorial Sloan-Kettering Cancer Center, New York, New York, U.S.A. Julian Little Department of Epidemiology and Community Medicine, University of Ottawa, Ottawa, Ontario, Canada Howard L. McLeod Department of Medicine, Washington University School of Medicine, St. Louis, Missouri, U.S.A. Igor Puzanov Division of Hematology/Oncology, Vanderbilt University Medical Center, Nashville, Tennessee, U.S.A. Mace L. Rothenberg Division of Hematology/Oncology, Vanderbilt University Medical Center, Nashville, Tennessee, U.S.A. Leonard Saltz Gastrointestinal Oncology Service, Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, New York, U.S.A. Linda Sharp National Cancer Registry, Cork, Ireland Robert J. C. Steele Department of Surgery and Molecular Oncology, Ninewells Hospital and Medical School, University of Dundee, Dundee, U.K. Sabine Tejpar Digestive Oncology Unit, University Hospital Gasthuisberg, Leuven, Belgium Eric Van Cutsem Digestive Oncology Unit, University Hospital Gasthuisberg, Leuven, Belgium Christian Wittekind Institut fu¨r Pathologie des Universita¨tsklinikums Leipzig, Leipzig, Germany

1 Genetic Susceptibility to Colorectal Cancer Rebecca A. Barnetson and Malcolm G. Dunlop Colon Cancer Genetics Group, University of Edinburgh, School of Molecular and Clinical Medicine and MRC Human Genetics Unit, Western General Hospital, Edinburgh, U.K.

INTRODUCTION Research into the genetic basis of cancer will undoubtedly lead to new understanding about disease etiology, not only shedding new light on disorders with a predominantly genetic contribution but also those where gene– environment interactions play a main role. Such understanding will also have important implications for instigating preventative measures, including environmental risk factor avoidance, chemoprevention and refining surveillance strategies, removal of premalignant lesions, and targeting prophylactic surgery. Finally, insight into the heritable genetic defects that cause cancer could lead to the design of novel anticancer agents, as well as inform understanding the response to chemotherapy and the mechanisms of drug resistance. Hence, there are substantial potential benefits that could result from the current substantial research endeavor in the genetics of colorectal cancer. Heritable genetic defects make a major contribution to the overall incidence of colorectal cancer. Many studies have shown aggregation of colorectal cancer in families (1) and there is even evidence to suggest that almost all colorectal neoplasia has a heritable component (2). However, familial aggregation does not, in itself, prove a genetic etiology because environmental risk 1

2

Barnetson and Dunlop

factors exposure is also frequently shared within families. However, twin studies comparing the incidence of colorectal cancer in monozygous and dizygous twins provide powerful evidence of a causal genetic involvement. From such twin data, it has been estimated that 35% of all colorectal cancer is attributable to genetic susceptibility (3). Despite this evidence for a substantial genetic contribution, only around 3% of the 32,000 incident cases arising annually in the United Kingdom are due to dominant syndromes for which the susceptibility gene responsible has been identified. The remainder are due to mutations in, as yet, undiscovered dominant genes with moderate to low penetrance, to recessive genetic traits, and to complex polygenic traits and multifactorial inheritance. Over recent years there have been considerable advances in understanding of the genetic basis of colorectal cancer in particular. The identification of a number of causative genes and the classification of the associated phenotype has meant that case-finding approaches can be set in place and surveillance instigated for at-risk relatives. It is now becoming possible to use genetic information clinically for prognosis as well as to tailor treatment and prophylaxis interventions. Indeed in some cases, the cancer risk is so high that prophylactic colectomy is merited. Here, we discuss hereditary colorectal cancer disorders that have been characterized to date and the implications for clinical practice. We also discuss low-penetrance alleles that, collectively, are likely to contribute more to overall colorectal cancer incidence than the known dominant disorders. Table 1 is a summary of the contribution of known genes to colorectal cancer incidence.

AUTOSOMAL DOMINANT DISORDERS HNPCC or Lynch Syndrome Hereditary nonpolyposis colon cancer (HNPCC) was one of the first disorders to be recognized as a hereditary cancer syndrome. First described in the 1890s by Aldred Warthin, who reported a high incidence of cancer of the colon and female organs in a seamstress’s family (4), the disease was further characterized by Henry Lynch in 1962, who was consulted by a patient who had a strong family history of colorectal and endometrial cancer, as well as several other cancer types (5). Segregation studies of other families with a predisposition to colon cancer strongly suggested an autosomal dominant pattern of inheritance. Further studies refined the definition of HNPCC and showed it was characterized by early age of cancer onset, a tendency for a greater proportion of tumors to be located in the proximal colon than in sporadic cancer, as well as a high frequency of synchronous and metachronous carcinomas (6). A typical HNPCC family is shown in Figure 1. Synchronous tumors are present in 5% to 20% of cases and metachronous tumors arise in 20% to 50% of cases (7). Despite the nomenclature, HNPCC

Genetic Susceptibility to Colorectal Cancer

3

Table 1 Relative Genetic Contribution of Genes Known to Predispose to Colorectal Cancer Gene Familial adenomatous polyposis HNPCC Rare dominant polyposis syndromes Peutz–Jegher’s syndrome Juvenile polyposis Multiple adenoma phenotype Familial Low-penetrance alleles

Gene–environment interactions

APC MMR STK11/LKB1 SMAD4, BMPR1A, PTEN MYH E-Cadherin, TGF-bRII, ?15q EpHx, GSTM1, GSTT1, NAT, CCND1 MTHFR, CYP1A1, CYP1A1 APCI1307K, APC-E1317Q, Hras APC-D1822V/fat (protective) MTHFR-A226V/folate

Contribution 0.07% 2.8% <0.01%


0.5% Unknown Unknown

RR 0.2 RR 0.8

Abbreviations: APC, adenomatous polyposis coli; MTHFR, methylene tetrahydrofolate reductase; MMR, mismatch repair; PTEN, phosphatase and tensin homolog; RR, relative risk; MYH, MutY homolog; EpHx, epoxide hydrolase.

is in fact associated with adenomatous colorectal polyps, although these are found at much lower frequency and are fewer in number compared with familial adenomatous polyposis (FAP). Extracolonic tumors are also well recognized as a feature of HNPCC. Endometrial carcinomas have

Figure 1 HNPCC family.

4

Barnetson and Dunlop

been reported in 20% to 60% of women with HNPCC (8), and the lifetime endometrial cancer risk of gene carriers is around 40% (9). Cancers of the stomach, small intestine, renal pelvis, ureter, and ovary are also frequently associated in HNPCC families (6). In some HNPCC families, sebaceous gland tumors and skin cancer are also a feature (Muir-Torre syndrome). Initially thought to be a distinct clinical entity, it is clear that Muir-Torre syndrome is an allelic variant of HNPCC (9a). HNPCC/Lynch Syndrome—Diagnostic Criteria and Clinical Features Because of the lack of a distinctive phenotype, a number of criteria have been developed to classify colorectal cancer families, mainly for unambiguous categorization of HNPCC for research purposes, but these are also used as a diagnostic tool. The best known of these are the Amsterdam criteria, which state that at least three relatives should have colorectal cancer and one of them should be a first-degree relative of the other two; at least two successive generations should be affected; in one of the relatives colorectal cancer should have been diagnosed before the age of 50 years; and families with FAP should be excluded. Although the majority of mutation carriers that have been identified fulfill the criteria, they are highly stringent and likely to exclude many mutation carriers in the population. Families that are small, adopted and illegitimate individuals, individuals with de novo mutations or individuals with a severe phenotype that do not reach childbearing age, or families where penetrance is incomplete are likely to be missed using these criteria. Additionally, as a result of a tendency for an attenuated phenotype, an appreciable proportion of patients with MSH6 germline mutations do not fulfill the Amsterdam criteria (10). Updated Amsterdam criteria (Amsterdam II) were developed to include carcinomas of the endometrium, upper gastrointestinal tract, and urinary tract (11). The Bethesda guidelines have also been developed (12) to include individuals with extracolonic carcinomas, synchronous and metachronous colorectal cancers, and right-sided carcinomas that are diagnosed under the age of 45 and have a histology that is characteristic of HNPCC tumors. Hence, the Amsterdam and Bethesda guidelines enrich for mutation carriers and increase efficiency of mutation searching. However, the problem is that both guidelines are stringent and make assumptions about penetrance such that the criteria are likely to exclude a considerable proportion of mutation carriers in the population. The true prevalence of HNPCC is yet to be determined in colorectal cancer cases, or indeed in the general population, but it is important to bear in mind that HNPCC alleles have been shown to be responsible for a substantial proportion of younger patients with colorectal cancer (13). Although there are no pathognomonic features of HNPCC-related tumors, certain patterns are more frequent in HNPCC than in ‘‘sporadic’’ colorectal cancer. There is a greater proportion of tumors located in the

Genetic Susceptibility to Colorectal Cancer

5

proximal colon and a high frequency of synchronous and metachronous tumors. Furthermore, colorectal cancer arising in HNPCC families is more frequently poorly differentiated and exhibits a mucinous histology (14). Tumor-infiltrating lymphocytes and a signet-ring cell component are also more common than in standard carcinomas (15). In addition to clinicopathology trends, unlike the majority of colorectal cancer, HNPCC tumors also tend to have a diploid DNA content (16). HNPCC Due to Mutations in DNA Mismatch Repair Genes It is now clear that the vast majority of HNPCC families are due to loss-offunction mutations in a class of DNA repair gene known as DNA mismatch repair genes (17). The DNA mismatch repair system has a number of functions (18), including identification and repair of G:T mispairs and of insertion deletion loops that are intermediaries of DNA replication slippage. DNA mismatch repair is also involved in cellular DNA damage response and links identification of DNA damage with apoptotic pathways to drive a cell death signal in the face of such lesions. Defective DNA mismatch repair has also been implicated in abnormalities of mitotic and meiotic recombination resistance to anticancer drugs and ionizing radiation, and defects in transcription-coupled repair. Tumors from HNPCC gene carriers almost universally show loss of expression of the corresponding protein (Fig. 2) due to somatic loss of heterozygosity (LOH), mutation, or epigenetic

Figure 2 (See color insert.) Expression of MLH1 in histological sections. (A) MLH1 expressed in the normal mucosa of a patient with colorectal cancer (B) loss of expression in tumor from a patient with a germline MLH1 mutation. Note that, even though there is complete loss of staining of malignant epithelial cells in B, the stromal cells stain normally.

6

Barnetson and Dunlop

silencing. Thus immunohistochemistry has utility identifying patients likely to carry mutations and targeting mutation analysis to specific genes, thereby improving efficiency and reducing costs and labor. From a historical perspective, around the time that tumor microsatellite instability (MSI) was recognized as an important feature in colorectal cancer (19), an HNPCC allele was mapped to the chromosomal locus 2p15–16 by genetic linkage analysis (20) and soon after this, a second locus at chromosome 3p21 was identified in three other HNPCC kindreds (21). In the same year, Strand et al. recognized that MSI in tumors was similar to the instability observed in bacteria with a mutation in the DNA mismatch repair genes mutS and mutL (22). These converging lines of investigation soon resulted in the identification of the human homologs for mutS and mutL at the 2p and 3p loci, respectively, with the identification of germline mutations in HNPCC patients following on shortly afterwards (23–25). Many reports of mutations in colorectal cancer families have followed. However, it has also become apparent that DNA mismatch repair gene defects are responsible for an appreciable proportion of apparently sporadic colorectal cancer arising in young patients (26) and in the generality of colorectal cancer (27,28). Of germline mutations reported in HNPCC kindreds to date, 50% have been identified in MLH1 and 40% in MSH2 (29,30). About 10% of germline mutations have been described in MSH6 (31,32) and far lower frequencies of mutation have been identified in other mutL homologs PMS2 (33) and MLH3 (34). Germline MSH6 mutations have a more attenuated HNPCC phenotype, with a high frequency of endometrial carcinomas, delayed age at onset, the MSI-L phenotype, and incomplete penetrance. In all, pathogenic mutations in DNA mismatch repair genes have an estimated population carrier frequency of approximately 1 in 3000 (9). MSI in Colorectal Cancer A particular molecular characteristic of tumors that arise in patients with germline DNA mismatch repair gene mutations is MSI, and an example is shown in Figure 3. Microsatellites are mononucleotide to pentanucleotide DNA repeat sequences that are present throughout the genome. MSI (also previously described as the replication error or RER phenotype) was identified by three independent studies in 1993 and is characterized by the presence of small deletions and insertions in tumor DNA compared with the germline DNA (19,35,36). While studying LOH in parallel with genetic linkage studies of chromosome 2 markers in HNPCC, rather than expected allelic loss, Aaltonen et al. found a high frequency of tumor phenotype characterized by variation in length of tandem repeats (35), which became known as MSI. Although tumor MSI is observed in the majority of tumors from HNPCC families, this is actually a minority of all MSI tumors because MSI is detected in 15% of sporadic colorectal carcinomas (36). MSI is actually a genome-wide phenomenon and because cancer-specific variations

Genetic Susceptibility to Colorectal Cancer

7

Figure 3 Microsatellite instability (MSI) in an HNPCC tumor. Chromatograph traces comparing normal DNA with tumor DNA for a poly-A BAT25 and polyCA repeat marker (D17S250). The overlaid traces clearly differ between the tumor and normal DNA in both markers (2 of the 5 standard set) indicating this is an MSI-H tumor. The additional tumor-specific peaks are arrowed.

at single tandem repeats is common, there is now a standard, validated, reference panel of five microsatellite markers (BAT25, BAT26, D2S123, D5S346, and D17S250) to rationalize categorization of tumors as MSI or otherwise. These are used widely to define tumor MSI phenotype (37), and tumors with instability at 2 markers are categorized as high-level microsatellite instability (MSI-H). This is the phenotype that is frequently identified in tumors from HNPCC patients, whereas the instability phenotype at only one marker is defined as low-level microsatellite instability (MSI-L). However, the significance of the MSI-L phenotype has yet to be elucidated with respect to the presence of germline mutations.

8

Barnetson and Dunlop

Although tumors arising in patients with germline mutations in DNA mismatch repair genes tend to exhibit the MSI phenotype, the corollary does not apply. In fact, most MSI tumors are not due to germline mutations but due to epigenetic silencing of MLH1 through promoter hypermethylation (38). Indeed, 95% of MSI-H sporadic colorectal carcinomas do not express MLH1 due to epigenetic silencing, despite having no germline abnormalities in this gene (39). Thus, germline mutations of DNA mismatch repair genes are not the only mechanism by which loss of function of DNA mismatch repair genes is brought about. This has important potential implications for clinical management of the 12% to 25% of cases of colorectal cancer that exhibit the MSI phenotype because only a minority of cases are due to germline defects, the rest being due to somatic events. Penetrance of HNPCC Systematic analysis of DNA mismatch repair mutation carriers in the population has shown that there is a 74% risk of developing colorectal cancer for males and 30% risk for females (40,41a). In fact female carriers have a higher risk of endometrial cancer than of colorectal cancer. Hence, preventive measures for females have to take into account the considerable gynecological cancer risk. Many women opt for prophylactic hysterectomy because the available screening modalities of ultrasound and pipelle biopsy are of no proven benefit and cause considerable inconvenience and distress in some women. Penetrance for colorectal and endometrial cancer in carriers of DNA mismatch repair genes is shown in Figure 4.

Figure 4 Penetrance of mutations in DNA mismatch repair genes. Age-dependent penetrance for all cancers and for colorectal cancer in males and females, and for uterine cancer in females. Note the substantial difference in cumulative colorectal cancer incidence between males and females.

Genetic Susceptibility to Colorectal Cancer

9

Strategies to Identify Gene Carriers and Potential Clinical Benefits It is clear that HNPCC families fulfilling stringent inclusion criteria such as the Amsterdam criteria are highly likely to carry mutations in DNA mismatch repair genes. However, as discussed earlier this approach is relatively insensitive. Another approach is prescreening of tumors by MSI analysis, followed by immunohistochemical assessment of DNA mismatch repair protein expression in tumor sections to guide gene testing (27,28). However, this again assumes that all mutations induce defective DNA mismatch repair and that this universally manifests as the tumor MSI phenotype. Another approach is to enrich for mutation carriers by analysis of all patients who develop colorectal cancer early in life. This has been particularly effective in identifying early onset cases due to germline DNA mismatch repair gene defects (13) and avoids a priori assumptions about familiality, penetrance, and tumor phenotype. However, at present it is not possible to mutation search all cases of colorectal cancer due to the considerable financial costs of such an undertaking. It would be ideal to identify mutation carriers prior to the development of cancer but population screening is even less feasible than testing patients with established cancer. Hence, for the foreseeable future, the best option is a synthesis of patient data comprising clinical, pathological, and family data linked to tumor MSI phenotype and immunohistochemistry findings. Looking for mutations in the MMR genes, our group has systematically screened a large populationbased, prospectively-collected series of patients who had developed colorectal cancer. We have identified a group of patients who are enriched for genetically determined disease, and constructed a clinically driven model that identifies patients where tumor MMR protein immunohistochemistry or MSI testing could further refine carrier prediction. The model that allows carrier prediction in surgical and oncology clinics can be accessed through a clinician-friendly URL (41b). Prognosis in HNPCC Colorectal Cancer There is some controversy in interpreting the prognostic implications of MSI in sporadic and HNPCC colorectal cancer. A considerable body of evidence indicates that patients with tumors exhibiting MSI have a better prognosis than those with microsatellite stable (MSS) tumors (19,36,42–49), although a number of studies also suggest there is no independent effect of MSI on overall outcome (50–55). A consensus of researchers in the field in 1997 considered that the evidence was not sufficiently robust for use of MSI as a prognostic variable in clinical practice (56) but population-based data have since been reported (44). It is now accepted that the generality of MSI tumors do indeed have a better overall prognosis than those which are MSS. The relative hazard of death is in the range 0.3–0.42 in patients

10

Barnetson and Dunlop

with MSI tumors compared with those with MSS tumors (42,44) indicating a clinically relevant effect. While there now seems little doubt that, stage-for-stage, colorectal cancers exhibiting MSI have a better prognosis than MSS lesions, there are some important factors to be taken into account. The age distribution of the cohorts analyzed in many studies reflects the fact that the majority of patients with colorectal cancer are elderly and that most MSI tumors occur in old people, due to MLH1 promoter hypermethylation and not due to germline MMR gene mutations (38). In contrast, MSI tumors observed in young colorectal cancer patients are predominantly due to germline DNA mismatch repair gene mutations (13,26). Hence, there is reason to question the validity of applying survival data from MSI analysis of older patient cohorts to estimation of prognosis for younger patients in whom the MSI phenotype is a manifestation of different molecular etiology. As discussed earlier, MSI is almost ubiquitous in tumors from patients who are members of HNPCC families but, since such gene carriers make up only a small proportion of all cases that exhibit MSI in unselected cases series, the majority of the observed difference in survival outcomes between MSI and MSS/ MSI-L tumors is attributable to sporadic MSI tumors. Nonetheless, germline carriage of DNA mismatch repair gene mutation is widely equated with the same effect on prognosis as that of the largely sporadic group. Hence, it is important to critically evaluate the available evidence for survival outcomes in HNPCC families and in MMR gene carriers specifically. Unfortunately there is a paucity of systematic data on which to gauge such prognostic implications of MSI and germline genotype status. Descriptive studies of HNPCC families report long-term survival to be a characteristic of cancers arising in members of such families (57–59). However, there are a few studies that have carried out robust analysis of cancer survival for patients with germline mutations. One Finnish study of MLH1 mutation carriers from HNPCC families suggested that survival was better than population controls (58), and another correlative study of MSH2 and MLH1 carriers in HNPCC families from the United States also suggested survival advantage for carriers compared to population controls (60). At face value, these findings might suggest that the observed beneficial effect of MSI in general was indeed representative of the situation in MMR gene carriers. However, there are several important potential biases incorporated in published studies of HNPCC families, including survival bias, length-time bias, ascertainment bias, and bias due to reproductive fitness. Furthermore, members of HNPCC families are much more likely to have undergone screening and for cancer to be detected at an early stage. In support of the notion that bias is a major factor on apparent prognosis in familial colorectal cancer, evidence from systematic, population-based studies have failed to show any better outcome for familial cases than nonfamilial. Three

Genetic Susceptibility to Colorectal Cancer

11

population-based studies of family history in substantial colorectal cancer patient cohorts were unable to detect any beneficial effect of cancer arising in either HNPCC (61,62) or in familial cases in general (63). Consistent with this, we found no statistically on clinically significant difference in survival between MMR mutation carriers and noncarriers in the only large prospective population-based patient series (64a). These findings emphasize the importance of early cancer detection in MMR mutation carriers. Chemotherapy and Defective DNA Mismatch Repair DNA mismatch repair is involved in cellular responses to certain types of DNA damage, including that caused by several chemotherapeutic agents. It is thus plausible that accumulation of mutations consequent upon defective mismatch repair might actually enhance the activity of cytotoxic or cytostatic agents. However, DNA mismatch repair is involved in effecting cell cycle arrest and apoptosis in response to DNA damage (64b,65), and so it is not surprising that resistance to a number of chemotherapy agents has been demonstrated including 5-FU (66), methylating agents (67), antimetabolites (68), and to a lesser degree cisplatin (69). In conflict with these studies showing resistance of DNA mismatch repair deficient tumor cells to specific chemotherapeutic agents, there is some evidence that patients with advanced colorectal cancer who exhibit MSI might be particularly sensitive to irinotecan through an indirect mechanism involving inactivation of genes such as BAX, due to the presence of coding repetitive DNA sequences that are disrupted as part of the MSI phenotype (70). Conventional ‘‘first line’’ chemotherapeutic regimens for colorectal cancer are primarily based on fluorinated thymidine analogs or thymidylate synthase inhibitors such as capecitabine. Hence, the effect of defective DNA mismatch repair on resistance to this class of chemotherapeutic agents is of particular interest and has been most extensively studied in the clinical setting. 5-FU causes DNA damage directly by the incorporation of FdUTP into DNA. The enzyme thymidylate synthase is also inhibited by formation of a complex that results in imbalance of dNTP pools as well as inhibition of DNA synthesis (71). In vitro cell line studies have yielded conflicting results as to the sensitivity of MMR deficient cells to 5-FU. One study examining cell lines mutant for both MLH1 and MSH2 showed that these cells were resistant to several chemotherapeutic agents compared to their MMR proficient counterparts, but were not resistant to 5-FU (72). The suggested explanation was the efficient removal of the low levels of 5-FU incorporated into DNA by uracil glycosylase. However, other studies suggest substantial resistance to 5-FU compared to proficient control lines (66,73). Furthermore, MMR deficient cells exhibit a defect in G2 cell cycle checkpoint arrest (73). Overall, the balance of published evidence from cell line assays suggests that MMR deficient cells are resistant to 5-FU-based chemotherapy. This has

12

Barnetson and Dunlop

considerable clinical implications because patients who have tumors that are defective for DNA mismatch repair may gain less benefit from those that have repair proficient tumors. However, this could only be definitively assessed by large randomized trials of placebo versus chemotherapy stratified by MSI status, but such trials are highly unlikely to be undertaken given that control arms require to receive current standard of care, namely, chemotherapy. Fortunately, there are many trials for which archival tumor material was kept such that some retrospective assessment of the effect of chemotherapy can be made. Chemotherapy in HNPCC and in Sporadic MSI Tumors The ability to predict benefit or otherwise for adjuvant chemotherapy would have potentially important clinical and health care resource implications. Several studies have examined the relative benefit of chemotherapy in patients with germline DNA mismatch repair gene mutations and/or with Lynch syndrome. However, there is an inherent difficulty in demonstrating differential responses because of a lack of sufficient numbers to provide statistical power. Studies demonstrating the benefit of adjuvant 5-FU for the generality of colorectal cancer require very large numbers to provide statistical power and which are impossible to undertake in HNPCC patients. There is also additional difficulty in identifying a suitable control group with which to compare outcomes. Hence, research has focused on sporadic MSI tumors rather than on MSI tumors arising in germline MMR gene mutation carriers. However, it is important that any conclusions reached from study of the generality of MSI colorectal cancers, which comprises MSI mainly as a consequence of somatic events, are not necessarily applied to the patients with germline mutations. With this caveat in mind, the impact of MSI status has been investigated in the two main clinical scenarios in which chemotherapy is employed in colorectal cancer, namely, palliative and adjuvant settings. Published data of 5-FU-based therapy in patients with metastatic disease provide conflicting conclusions. One study of patients receiving 5-FU showed no ability to discriminate response to therapy on the basis of MSI, p53, or k-ras status (74). Another larger study nonrandomly allocated 244 patients (93 MSI) to a chemotherapy regime, including leucovorin, or no chemotherapy, and demonstrated that MSI status and the administration of chemotherapy were independent favorable prognostic parameters (75). While the authors suggested this was due to increased chemosensitivity, it may also be a reflection of heterogeneity between patient groups with respect to biological tumor behavior. In view of such small numbers and the lack of rigorously controlled trials, it is difficult to draw any firm conclusions on the effect of 5-FU in the palliative setting. Post hoc analysis of MSI status of tumor material in the adjuvant setting has yielded some conflicting results. An Australian study of 656 patients (56 MSI) demonstrated a significant survival advantage for patients

Genetic Susceptibility to Colorectal Cancer

13

receiving chemotherapy who had MSI tumors (90% vs. 35%) whereas MSI status had no demonstrable effect on survival in patients who did not receive chemotherapy (76). However, the study groups were not randomized, and younger patients were significantly more likely to receive chemotherapy. Furthermore, right-sided tumors also had an enhanced response to chemotherapy, irrespective of MSI status and, since 20% of right-sided tumors exhibited MSI (the great majority of MSI tumors), it seems possible that there may have been some confounding effects responsible for the observed outcomes. A Finnish study also reached similar conclusions, although the numbers of patients receiving chemotherapy for stage C disease were very small, with only 11 of whom having MSI tumors (77). In another U.S. study, MSI was associated with improved disease-free five-year survival, enhancing clinical response to chemotherapy (78). Multivariate analysis demonstrated MSI status was an independent predictor of survival, although site of tumor was not included in regressional analysis. Although these studies (76–78) suggest there may be benefit for patients receiving 5-FU-based chemotherapy, lack of randomization means that the findings should be taken with caution. However, a number of studies have been undertaken to determine tumor MSI status in material collected on patients who were recruited to randomized controlled trials of 5-FU-based chemotherapy. The analysis of material from the U.K. AXIS trial of Dukes’ stage B or C tumors to assess the impact of MSI on survival of 368 patients undergoing chemotherapy (89 MSI tumors) found no difference in survival conditional on MSI status or site of tumor (79). Further support for a lack of a beneficial clinical effect reported by previous studies comes from a randomized trial of 570 patients (95 MSI cases) with stage II or III cancer (80). This study also demonstrated no survival benefit following chemotherapy in MSI tumors, whereas there was benefit of chemotherapy for MSS or MSI-L tumors. Furthermore, it appears that chemotherapy offered to patients with MSI tumors actually resulted in a worse outcome than if they had not received chemotherapy at all. Another recent study of the impact of tumor MSI on survival outcome in patients randomized to 5-FU-based chemotherapy in California also suggests that patients with MSI tumors do not benefit from chemotherapy (81). Hence, taking account of all published work, it seems there is no definitive evidence one way or the other to guide clinical practice because of the lack of robust trials, and that research involving assessment of tumor MSI status has some way to go before it can influence the decision of the patient and the clinician as whether or not to undergo chemotherapy in either the adjuvant or the palliative setting. Cancer Control in HNPCC There are several approaches to colorectal cancer prevention in HNPCC and the available evidence suggests that a substantial reduction in risk can

14

Barnetson and Dunlop

be achieved. Because the cancer risk is so high (40,41,82), surveillance is widely employed and guidelines are well established in clinical practice (83). The effectiveness of surveillance has never been formally assessed in randomized controlled trials and never will be, because of ethical issues about withholding surveillance from people with such a high risk. Comparative studies have been undertaken and these show a substantial beneficial effect of surveillance, with a 62% incidence reduction and 65% mortality reduction (effectively to zero) (84). The reduction in cancer incidence is probably due to the effect of removing premalignant adenomatous polyps that are small in number, but are highly likely to progress to cancer. However, it is important to note that the surveillance interval should be less than two years in view of a lesser benefit with longer time intervals (85). For clinicians managing patients with germline DNA mismatch repair gene mutations who have already developed colorectal cancer it is important that the considerable cancer risk to the retained colon and rectum is understood. Following a segmental colonic resection such as right hemicolectomy, the risk is around 16% (85), while for total colectomy and ileorectal anastomosis (IRA), the risk is 3.4% to 12% (85,86). Decision analysis in the surgical treatment of colorectal cancer due to a mismatch repair gene defect (87) suggests that subtotal colectomy, rather than a more limited resection, may be the preferred surgical treatment for established colorectal cancer arising in patients with known mutations or where there is a family history fulfilling Lynch syndrome criteria. This approach is effectively secondary prophylaxis, but what has not been studied is primary prophylaxis. Some patients opt for preemptive surgery because of concerns about the effectiveness of surveillance and also about the discomfort and potential complications of colonoscopy on at least a biennial basis until 75 years of age. However, at present there are no robust data on which to gauge advice to patients about the relative benefits of surveillance compared to the upfront risk of major intra-abdominal surgery with the aim of prophylaxis. It will be interesting to observe clinical practice with time as the cancer risk becomes more widely known and greater numbers of patients shown to carry pathogenic mutations who have relatives affected by cancer have the knowledge to allow them to make their own judgment about prophylactic surgery. It seems likely that primary prophylaxis will become more common than it has been to date. Considerable progress has been made in understanding the genetic basis of HNPCC, and this has allowed considerable progress in translational research because it is a relatively common condition. Similar progress in other genetic disorders that predispose to large bowel malignancy has been hampered by the rarity of those conditions. Nonetheless, there is much to learn indirectly from study of HNPCC. A summary of learning points are provided in Table 2.

Genetic Susceptibility to Colorectal Cancer

15

Table 2 Summary of Lynch Syndrome (HNPCC) Lynch syndrome is due to loss-of-function mutations in DNA mismatch repair genes DNA mismatch repair gene defects account for 3% of all colorectal cancer Cancer is very high for gene carriers, especially males Females are at an even greater risk of developing endometrial cancer Surveillance should be offered to gene carriers and people with a family history fulfilling established HNPCC criteria Tumor MSI phenotype has implications for prognosis Tumor MSI may impact on recommendations for chemotherapy in the future but currently understanding is insufficiently complete for definitive advice

FAMILIAL ADENOMATOUS POLYPOSIS FAP is an autosomal dominant disorder (incidence of 1:10,000 to 1:15,000) that is characterized by the development of hundreds to thousands of adenomatous polyps in the colon and rectum early in life. Figure 5 shows rectal polyposis in a girl aged 13 years with FAP. Almost all mutation carriers will manifest polyposis before age 40 and will develop colorectal cancer by age 50 unless prophlylactic colectomy is undertaken (88). With greater clinical awareness and predictive genetic testing, the majority of cases are detected early and undergo prophlylactic colectomy; less than 0.07% of colorectal carcinoma incident cases are attributed to FAP (89). As registries improve detection of at-risk family members, the proportion of colorectal cancer cases

Figure 5 Rectal polyps in familial adenomatous polyposis.

16

Barnetson and Dunlop

due to FAP should reduce, limited only by the proportion of new mutations. Around 25% of all cases arise as new mutations and so, although children are at 50% risk, preceding generations are at no greater risk than the general population (89a). In addition to colorectal cancer risk, FAP patients also have an increased risk of duodenal malignancy (90), and up to 32% have desmoid tumors (91). Although benign, these lesions cause distressing symptoms and are the most common cause of death in patients with FAP (92). Gardner’s syndrome was once thought to be a separate disorder, but it is, in fact, allelic with FAP. The syndrome is associated with polyposis phenotype combined with other manifestations including osteomas, dental abnormalities, epidermoid and sebaceous cysts, and neoplasms of the thyroid (93). The cancer risk is equivalent to that in FAP and so prophylactic colectomy is recommended. Molecular Pathology The gene responsible for FAP was identified by cytogenetics and genetic linkage studies, which located the gene to chromosome 5q21–22 (94,95). The adenomatous polyposis coli (APC) gene was subsequently cloned and germline chain-terminating mutations identified in a number of FAP patients (96,97). APC is a large gene encompassing 15 exons over 250 kb with an 8.5 kb transcript encoding a 2843 amino acid polypeptide. The vast majority of germline mutations reported in FAP patients/families result in premature truncation of APC (98,99). Around 80% of the APC mutations identified to date are in the large exon 15, with two specific mutations (codons 1061 and 1309) accounting for 15% to 20% of all APC mutations. However, the remainder are spread throughout the gene with no other ‘‘hotspots.’’ Short repeat sequences at the amino terminus of APC are predicted to form coiled-coil structures suggesting that normal APC functions as a homodimer. Thus mutations leading to a truncated APC protein may result in a heterodimer of mutant/wildtype APC protein that may abrogate the function of the normal protein in a dominant-negative manner. APC Gene Function The function of the APC gene product is the focus of much research interest and the complexities of the cellular role of APC have yet to be fully understood. The APC protein is expressed in epithelial cells in the upper portions of the colonic crypts, suggesting involvement in colonocyte maturation (100,101). Several functional domains are revealed in the protein sequence including the N-terminal homodimerization sequences, as well as numerous other cellular processes such as cellular adhesion, cell-cycle regulation, apoptosis, differentiation, and intracellular signal transduction. The central region of the protein contains b-catenin binding and regulatory domains as well

Genetic Susceptibility to Colorectal Cancer

17

as binding domains for the axin family of proteins. APC appears to influence cellular adhesion by affecting the interaction between catenins and E-cadherin, thus promoting the shedding and migration of epithelial cells. In conjunction with other proteins, axin, glycogen synthase kinase 3b (GSK), and other GSK binding proteins, APC plays a critical role in intracellular communication by modulating the levels of b-catenin-dependent transcription (102). b-catenin is an important transcription factor for oncogenic proteins such as cyclin D1 and c-myc (103). The pivotal role of abnormalities in the wnt signaling pathway in colorectal tumorigenesis is exemplified by the identification of somatic mutations in many of the components, such as APC, b-catenin, and axin, with somatic APC mutations being identified in
85% of all colorectal tumors (104). APC also plays a key role in the microtubule cytoskeleton, binding to microtubules as well as promoting their formation and presenting them to the kinetochore (105,106). The organization and structure of microtubules is vital to cell division and migration, and truncated forms of the APC protein appear to be unable to bind microtubules (107). Not only does APC bind microtubules but is also involved in mediating attachment of chromosomes to the spindle apparatus in order to accurately separate the sister chromatids during mitosis. Mouse cells homozygous for a truncating mutation, Apcmin, displayed abnormal chromosome patterns when compared with wildtype counterparts (105). This suggests that APC plays an important role in maintaining fidelity of chromosome segregation and thereby control of chromosome number. This is supported by the observation that aneuploidy occurs in the majority of colorectal cancers with APC mutations and underscores the complexity of the role that APC mutations play, suppressing tumorigenesis and progression. The carboxy terminus of the APC protein not only binds microtubules but also the microtubule binding protein EB1 (108) and the tumor suppressor protein Dlg (109), both of which are implicated in tumorigenesis. Genotype–Phenotype Correlations From the outline of APC function described earlier, it might be suspected that mutations in particular regions of APC might be expected to give particular phenotypes. However, this is not as straightforward as might be expected. Genotype–phenotype correlations could help inform the clinical management of at-risk individuals since it may help predict the likelihood of extracolonic manifestations, such as desmoid disease. Indeed the common 1309 mutation is associated with a dense polyp phenotype and a high cancer risk in the retained rectum after prophylactic colectomy and IRA (110), thus suggesting proctocolectomy and ileoanal pouch reconstruction may be the favored option. However, genotype–phenotype correlations are not well defined with various groups identifying families with identical APC

18

Barnetson and Dunlop

mutations that have diverse phenotype in terms of both colorectal polyposis and extracolonic disease (111). A number of variant syndromes are also caused by germline APC mutations. Attenuated FAP (AAPC or AFAP) describes patients with later onset disease and limited numbers of flat polyps. Mutations responsible for the AFAP phenotype tend to be in the first five exons of APC, exon 9 and at the 30 end of the gene (112). Desmoid disease has been correlated with 30 APC gene mutations (113). Another phenotypic variant of FAP is that of Turcot’s syndrome, which manifests as multiple colonic adenomas with early onset of colorectal cancer and also tumors of the central nervous system, particularly brain tumors. If the associated brain tumor is cerebellar medulloblastomas the genetic defect is likely to be in the APC gene, whereas if glioblastoma multiforme tumors are found, the defect is likely to be defective DNA mismatch repair (114). Although there are loose genotype–phenotype correlations, there are still many variables that impact on phenotype, even within a family, and it is likely that there are both genetic and environmental modifiers. Clinical Management The aims of clinical management are centered on carrier identification and prophylactic surgery involving removal of the majority of the large bowel epithelium, which is highly effective in reducing colorectal cancer risk. Genetic testing should be offered around the time of puberty although parents and child should be counseled that there is a small cancer risk even before this age (83). Unlike the situation in HNPCC there is currently no evidence to suggest differential response to chemotherapy in patients with germline APC gene mutations who have developed colorectal cancer in either the adjuvant or palliative setting. There is substantial benefit from APC gene mutation analysis in FAP cases because only one gene is involved and so it is relatively straightforward to survey the whole APC gene. Once a mutation has been identified, it is then straightforward to offer genetic testing to at-risk relatives, and so unaffected gene carriers can be offered prophylactic surgery. Operative Strategy in FAP FAP is a paradigm for prophylactic surgery in hereditary cancer and so merits some discussion about the nature and timing of such a major intervention in young healthy people. The aims of surgery are to reduce large bowel cancer to the smallest possible level but maintain the ability for a normal life, normal physiological function, and fertility. Total ablation of risk can be effected by a proctocolectomy and permanent ileostomy, and this was standard practice until relatively recently. However, patient preference and the considerable morbidity associated with such destructive surgery for young people has led to two other approaches that involve reconstruction and maintenance of bowel integrity. Colectomy and IRA involve reducing the

Genetic Susceptibility to Colorectal Cancer

19

surface area to only around 15 cm of rectum, and this requires at least annual surveillance for life. Restorative proctocolectomy and ileoanal pouch is another approach that reduces the cancer risk to very low levels but is technically more demanding and has some other potential disadvantages. However, the primary concern with these two operations centers around the residual cancer risk in the retained rectum after colectomy and IRA, which ranges from 10% at age 50, rising to 29% at age 60 (115). The cumulative risk after surgery has been reported as 7.7% at 10 years, 13.1% at 15 years, and 23.0% at 20 years (116). Further studies have demonstrated similar risks of developing rectal cancer (117,118) and have also shown that this is associated with mortality of 8% by age 55 and 12.5% by 65, with risk increasing if primary surgery was undertaken over the age of 25. The risk of cancer in residual large bowel epithelium is substantially reduced following proctocolectomy with ileoanal pouch, but this translated into an increased life expectancy of only 1.8 years over colectomy and IRA (118). There are limited data to guide practice with respect to identifying patients who are likely to require completion proctectomy following an ileorectal anastomosis, suggesting ileoanal pouch as a primary operation for these patients. Patients with dense rectal polyposis or large rectal polyps at primary operation will remain at risk with a retained rectum undergoing endoscopic surveillance. The risk of developing cancer in the rectal remnant has also been shown to vary according to genotype. Mutations between specific loci of the APC gene are associated with increased risk of requiring completion proctectomy, with mutations between codons 1200 and 1500 associated with increased risk (117). Mutation between codons 1250 and 1464 has been associated with a nine times increased risk of rectal excision (116). Mutations at codons 1309 or 1328 were associated with severe polyposis ( >1000 polyps), and almost all patients with these mutations who initially received colectomy and IRA required rectal excision due to rectal remnant polyposis and cancer concerns (119). Although there are descriptive and comparative studies that have described past practice, it is fair to say that there are no robust data with which to guide future surgical practice. Hence, all of the correlations discussed earlier should be taken with caution because of the lack of prospective randomized studies. Such studies are highly unlikely because of the rarity of FAP and the need for studies that last 25 years or more of follow-up. Hence patient involvement is important, with an honest appraisal of available information such that an informed choice can be made. This is particularly important because there are issues about undertaking major colorectal surgery that may have a detrimental effect on bowel function and on fertility, not to mention complications associated with any intra-abdominal operation, such as adhesions and bowel obstruction. Indeed, there is even conflicting evidence as to the relative benefits in terms of bowel function between colectomy and IRA and proctocolectomy with ileoanal

20

Barnetson and Dunlop

pouch in FAP. By one measure, significantly better outcomes have been observed following ileorectal anastomosis compared to ileoanal pouch, whereas the same study groups’ quality of life using the SF-36 form showed no difference between ileorectal and ileoanal pouch patients (120,121). Another concern about pelvic surgery centers around the possibility of decreased fertility. Proctectomy requires more extensive pelvic dissection, whereas colectomy and IRA may minimize damage to Fallopian tubes. A recent study of Scandinavian polyposis registers concluded that patients undergoing ileorectal anastomosis retained fecundity in line with the general population but ileoanal pouch reduced fertility, with a cumulative chance of pregnancy of 48% at one year and 61% at two years (122). In summary, surgical practice and prophylaxis in FAP is largely empirical and guided mainly by clinician preference to date. However, as the strength of observational data increases, some reasonably well-supported guidelines are now possible. However, it is essential that patients are fully informed about the possible detriments of surgery as well as the substantial benefits in cancer prevention that can be achieved. Turcot’s Syndrome The hallmarks of this disorder are the development of tumors of the central nervous system, particularly cerebellar medulloblastomas or glioblastomas, and multiple colorectal adenomas. This is a rarer variant of FAP because in the majority of cases, the underlying molecular defect is a germline mutation in the APC gene (114). However, mutations have also been identified in the DNA mismatch repair genes, MLH1 and PMS2. It is unclear as to whether the syndrome is dominantly or recessively inherited. Turcot’s syndrome is best not considered a distinct syndrome but part of both FAP and HNPCC syndromes, depending on the underlying genetic defect. OTHER DEFINED DOMINANT COLORECTAL CANCER SUSCEPTIBILITY SYNDROMES Peutz–Jeghers Syndrome Peutz–Jeghers syndrome is characterized by multiple gastrointestinal hamartomatous polyps, and mucocutaneous melanin deposits are found on the lips, perioral and buccal regions, hands, and feet in 95% of cases. This is a rare disorder with an incidence of 1 in 120,000 (123) and is associated with low penetrance. Affected individuals have about a 50% increased chance of developing gastrointestinal carcinomas or tumors of the pancreas, ovaries, testes, breast, and uterus (124–126). Germline mutations have been identified in the serine threonine kinase gene, STK11/LKB1 at 19p13.3 in 20% to 63% of patients with this disorder (127). Large bowel surveillance is recommended three times yearly for affected individuals from age 18 years (127A).

Genetic Susceptibility to Colorectal Cancer

21

Juvenile Polyposis Juvenile polyposis (JPS) is characterized by the development of multiple hamartomatous polyps throughout the gastrointestinal tract, usually when aged less than 10 years. Affected patients have a high risk of developing gastrointestinal cancer (128,129) that ranges from 9% to 68% and is probably around 50%. Because it is rare, there are no reliable estimates of the frequency of JPS in the general population, but around 1:50,000 is a reasonable estimate based on population registry data. It is associated with incomplete penetrance for both polyposis and also colorectal cancer, and less than 0.1% of all cases of colorectal cancer are attributable to JPS (129a). Unlike Peutz–Jeghers syndrome, JPS exhibits genetic heterogeneity, with mutations in at least three genes being responsible. The molecular basis of JPS in around 50% of cases is germline mutation of SMAD4 (130,131). SMAD4 is a tumor suppressor encoding a protein involved in the transforming growth factor-b (TGF-b) signaling pathway, which is involved in the regulation of cell proliferation and differentiation (132). Consistent with this, sporadic colorectal carcinomas are frequently found to have mutations in SMAD4 or LOH at this chromosome region (133). A second locus has been identified and germline nonsense mutations identified in the bone morphogenic protein receptor 1A gene (BMPR1A/ALK3) and, like SMAD4, this gene is also part of the TGF-b superfamily (134). There are a minority of JPS cases that are due to germline mutations in the protein phosphotase gene PTEN (135). PTEN is also mutated in patients with Cowden disease, which is also characterized by the presence of multiple gastrointestinal hamartomatous polyps as well as benign and malignant neoplasms of the thyroid, breast, uterus, and skin, but there is not an increased risk of colorectal cancer. Hence, there is a possibility of misclassification and that PTEN mutations are not associated with an excess colorectal cancer risk. Estimates of cancer risk have wide confidence intervals but are around 50% lifetime risk, and so surveillance is important. Consideration should also be given to prophylactic colectomy, much in the same way as in FAP (83). Large bowel surveillance for at-risk individuals is recommended one to two times yearly from the age of 15 to 18 years or even before if the patient has presented with symptoms. Screening intervals could be extended at age 35 years in at-risk individuals. However, documented gene carriers or affected cases should be kept under surveillance until the age of 70 years and prophylactic surgery discussed.

AUTOSOMAL RECESSIVE DISORDERS To date there is one recessive colorectal cancer susceptibility syndrome, but it seems likely that others remain to be discovered. Detecting such recessive alleles poses particular practical problems because of the lack of clear

22

Barnetson and Dunlop

Figure 6 Pedigree showing recessive inheritance of colorectal cancer.

family history in which to employ genetic approaches to gene identification, especially since colorectal cancer is so common in the general population. Hence it will be interesting to observe progress now that whole genome scanning approaches are being applied to detecting cancer susceptibility genes. These approaches allow the analysis to ignore presuppositions about the mode of inheritance. An example of a recessive colorectal cancer family is shown in Figure 6. MYH-Associated Polyposis In addition to the dominant inheritance of hundreds or thousands of polyps, another syndrome has recently become apparent in which there is a recessive mode of inheritance and a reduced number of adenomatous and of metaplastic/hyperplastic polyps, known as MYH-associated polyposis (MAP). Because many families have been included as FAP in genetic registries, the

Genetic Susceptibility to Colorectal Cancer

23

phenotype has not been fully described and is only now possible because of molecular testing of the underlying gene defects. Mutations were identified in the base excision repair (BER) gene, MYH, in around 25% of families that were originally categorized as FAP but with neither dominant transmission nor evidence of APC mutation (136–138). It is clear that the disorder is autosomal recessive because biallelic MYH mutations were required for the polyposis phenotype. Since transmission is as an autosomal recessive trait, this has substantial implications for genetic counseling, testing, and surveillance. MYH gene testing is now offered to patients with a phenotype resembling FAP when no clear evidence of vertical transmission is recorded and where no APC mutation is identified (138a). In addition to polyposis, recent studies have established that biallelic defects in BER genes predispose to colorectal cancer (139,140), with complete penetrance by age 60 years. However, it is not clear whether there is a heterozygous effect because very large numbers are required to show an effect. There is some supporting evidence that heterozygous mutations may be associated with a small excess risk of colorectal cancer (139,140), although the effect was only apparent for late onset disease and it is possible that some of the excess risk detected in heterozygotes is due to undetected variants on the other MYH allele. Indeed, data from mouse models suggest that on an ApcMin/þ background that monoallelic MYH inactivation does not increase tumor burden or the signature G:C to A:T transversions of the remaining Apc allele in the mouse tumors (141). Population frequency of heterozygous MYH mutations is around 0.6%, and the observed contribution to colorectal cancer is 2% for patients aged <40 years and 0.7% for patients aged <55 years. This has important clinical implications for siblings of carriers, who have at least a 1:4 risk of colorectal cancer before age 60 years. Molecular Pathology MYH is the homolog of the MutY protein in Escherichia coli, and functional analysis has revealed that human MutY protein has the same function as its bacterial counterpart, in that it is involved in base excision repair (142). The BER pathway plays a significant role in the repair of mutations caused by oxidative DNA damage. The role of MYH is to remove the adenine that mispairs with 8-oxoguanine, which arises due to replication of oxidized DNA. Thus, defective MYH results in an increased rate of transversion of G:C to A:T (143), and this has been observed in human colorectal cancer. Indeed, the role of MYH defects in colorectal cancer was elucidated through the observation of the high rate of G:C to A:T somatic inactivating transversions in the APC gene (136). Family members subsequently shown to carry two common missense variants, (Y165C and G382D), in MYH, which affect highly conserved residues over several species including E. coli. Codon 165 is located in a pseudo-helix-hairpin-helix motif and is predicted to function

24

Barnetson and Dunlop

in mismatch specificity, while codon 382 is located in the C-terminal MutTlike domain and may be involved in substrate binding. Functional assays of adenine glycosylase activity of both mutant proteins (Y165C and G382D) demonstrated that their activity was significantly reduced compared with wild type in E. coli using both the equivalent codons in the E. coli MutY gene (136) and the human MYH gene (143). Study of a mouse mutant equivalent to the G382D variant has shown loss of MYH function in vivo and an excess mutation rate, especially for G:C to A:T transversions (144). Thus, it is clear that inherited variants in MYH predispose to polyposis and to colorectal cancer. The two common variants have also been detected in three patients with extracolonic tumors and some of the features of FAP, including duodenal polyposis and congenital hypertrophy of the retinal pigment epithelium (138). Further clinical studies are required to assess whether mutations in this gene predispose to other extracolonic manifestations. Clinical Management of MAP Since the identification of MAP is so recent and there are only limited data available, it is unwise to make firm recommendations about clinical management. However, some broad conclusions can be made on the basis of findings of available studies. It is clear that the risk of colorectal cancer is very high for carriers of biallelic MYH mutations and so surveillance seems appropriate, based on experience with cancer risk reduction in FAP. Patients shown to have multiple polyps should be referred to a regional clinical genetics center for APC gene testing and if negative, MYH analysis should be undertaken. If a mutation is identified, then at-risk relatives should be offered counseling and genetic testing. The cancer risk is so high that prophylactic surgery should be discussed as part of the counseling process. Incidental prophylaxis is already possible, since patients who are undergoing colorectal cancer resection and who are known to carry an MYH mutation should have counseling about the potential benefits of subtotal colectomy and IRA or proctocolectomy and ileoanal pouch, where relevant. Although it is possible that defective BER has implications for adjuvant and palliative chemotherapy and radiotherapy, at present there are no data to guide practice. LOW PENETRANCE VARIANTS Common Genetic Variation and Susceptibility to Colorectal Cancer As shown in Table 1, genes responsible for colorectal cancer susceptibility syndromes account for only a small minority of all cases of colorectal cancer. However, the residual contribution of dominant high penetrance genes, exemplified by strong family history, is likely to be small and so the remaining high penetrance alleles are likely to be rare. A number of putative linkage

Genetic Susceptibility to Colorectal Cancer

25

signals have been identified, but one locus is particularly interesting on chromosome 15q in which a common risk haplotype has been identified (145). The problem with such a situation is that there is only one risk haplotype and so fine mapping of the region is problematical. However, once the causative mutation is identified, other smaller families and unselected cancer cases can be tested as a candidate colorectal cancer susceptibility gene, other than the family that has resulted in the linkage signal at the colorectal adenomata and carcinoma (CRAC) locus. Notwithstanding that some high-penetrance alleles have yet to be identified, it is likely that the majority of the unexplained genetic contribution to colorectal cancer is due to multiple low-risk alleles contributing to colorectal cancer etiology (146). However, only a very few alleles have been examined by statistically robust studies supported by replication studies. The evidence in most cases is inconclusive and much remains to be done to fully appraise the etiological role of such genetic variants. The problems center on lack of sufficient study size in case–control studies, subgroup analysis, publication bias, multiple testing (or ‘‘data-mining’’), lack of replication between different population groups, admixture, and population stratification. However, there are a number of putative colorectal cancer risk variants that are worthy of comment. Although alleles conferring RR of around 1.5 may not seem useful in the context of clinical genetics, such effects can provide novel insight into disease causation and lead to study of new pathways found to be aberrant in cancer. Furthermore, an RR ¼ 1.5 has considerable public health relevance because screening for colorectal cancer in the whole population is currently being recommended. Thus, targeting screening and more intensive approaches to those at higher risk could improve the efficiency of such public health measures. Nontruncating Variants in the APC Gene As discussed earlier, the vast majority of truncating variants in the APC gene cause a polyposis phenotype with a high-penetrance colorectal cancer phenotype. However, in view of its central gatekeeper role, APC is a strong candidate contributing to colorectal cancer susceptibility through common genetic variation, and a number of nontruncating alleles associated with excess colorectal cancer risk have been described. One interesting facet of APC-associated risk of colorectal neoplasia is that of the ‘‘premutation,’’ I1307K. I1307K is a germline variant that results in an unstable poly-A8 tract within the APC gene (147). The variant causes genomic instability in the APC gene and a marked increased frequency of somatic frameshift APC mutations in colonic epithelial cells. The resultant somatic inactivating APC mutations result in an attenuated polyposis phenotype and an excess colorectal cancer risk. However, although the variant is common in Ashkenazi Jews (6%), it is rare in other populations. Nonetheless, such premutations may be more common than previously considered in cancer genetics and will considerably

26

Barnetson and Dunlop

complicate the detection of risk alleles, since they in themselves do not appear to be pathogenic. Two other APC gene variants have sufficient weight of evidence to merit consideration as possible risk alleles, the E1317Q variant and the D1822V variant (through interaction with dietary fat). However, even with these alleles, the evidence suggesting a pathogenic role is conflicting. E1317Q has been reported to confer risk of colonic adenomatous polyps (148). The population prevalence of E1317Q is 2% to 3% and some evidence suggests it is associated with an 11-fold excess risk of multiple colorectal adenomas, accounting for around 4% of all patients with multiple colorectal adenomas. A recent study found no excess of E1317Q heterozygotes among patients with colorectal cancer or adenomatous polyps compared to the general population, but when allele frequencies were compared between disease groups and colonoscopically normal subjects, there was indeed a significant excess of the risk allele among colorectal cancer patients (149). However, the issue of subgroup analysis and lack of stringent matching by age/sex and environmental risk factor exposure remains a problem. Hence, the contribution of E1317Q to large bowel neoplasia requires investigation in very large case series with stringently matched controls to definitively address this issue. The D1822V APC variant may be associated with a protective effect from large bowel cancer. The valine variant is the most common and homozygotes (val/val) aged over 65 years may have a reduced risk of colon cancer [odds ratio (OR), 0.6; 95% confidence interval (CI), 0.4–1.0]. However, a substantial protective effect was observed when a low-fat diet was consumed (OR, 0.2; 95% CI, 0.1–0.5) relative to those homozygous for alleles encoding aspartate at codon 1822 and who consumed high-fat diets (150). The E1317Q and D1822V APC variants exemplify the problems with studies aimed at detecting genetic associations. Firstly, APC is an excellent candidate gene because it plays a pivotal role in tumor suppression in the colorectal epithelium. However, despite this the numbers of cases and matched controls required to definitively establish a causative role and to undertake confirmatory replication studies are very large indeed. Finally, there are problems in extrapolating these research findings to clinical benefit because of uncertainties about the causative role of the variants but also because the effect is insufficient to merit major interventions such as prophylactic surgery, while current understanding of the role of APC hinders the development of agents to counter the effect of such genetic variants at the protein level. Indeed, there are similar problems with variants in other genes that are less good candidates than APC as discussed below. Low-Penetrance Variants in DNA Mismatch Repair Genes Pathogenic variants, usually chain-terminating mutations, in MLH1 are generally highly penetrant and result in the phenotype of HNPCC or Lynch

Genetic Susceptibility to Colorectal Cancer

27

syndrome. However, recent evidence suggests that a G > C variant resulting in an aspartate to histidine amino acid substitution at codon 132 of MLH1 (D132H) is associated with excess cancer risk (151). The aspartate residue is highly conserved between species, and the variant appears to attenuate ATPase function and to reduce DNA mismatch repair efficiency but it is interesting to note that only a minority of tumors exhibit MSI. Carrier frequency is 1.3% in the Israeli population and so could potentially make an appreciable overall contribution to the burden of colorectal cancer. Penetrance is much lower than for truncating MLH1 mutations, with mean age at cancer onset for carriers of 70.1 years. The identification of such a low-penetrance allele in MLH1 invokes the possibility that there are other common alleles in DNA mismatch repair genes predisposing to colorectal cancer. Hence, it will be interesting to replicate studies of D132H MLH1 in large studies of other populations, as well as to define the contribution of other similar low-penetrance variants in genes, which are otherwise associated with high-penetrance alleles such as truncating mutations. Cyclin D1 Aberrant expression of the cell cycle gene, cyclin D1 (CCND1), is a common somatic event in colorectal cancer but there is conflicting evidence as to whether genetic variation contributes to colorectal cancer susceptibility. There is some data to suggest that CCND1 variants may modify the phenotype of HNPCC (152) but there have been no substantial replication studies to confirm or refute the findings of this small study. Similarly, there remains controversy as to whether CCND1 variants may be risk alleles of main effect in themselves. Reported putative associations should be taken with circumspection because the apparent association has only been detectable in certain subgroups (152,153). Furthermore, other reports suggest there is no effect (154) and so the jury remains out as far as the involvement of CCND1 in colorectal cancer susceptibility. Methylene Tetrahydrofolate Reductase Methylene tetrahydrofolate reductase (MTHFR) is centrally involved in regulation of folate metabolism, and thus indirectly in DNA synthesis and repair. Epidemiological evidence has suggested that dietary folate consumption is inversely related to colorectal cancer risk. Common variants in the MTHFR gene, C677T, and A1298C variants may be associated with an increased cancer risk in association with low dietary folate. The C677T variant influences enzyme activity, while the A1298C variant may be in LD with C677T, or both may be in LD with a separate pathogenic variant (146) Thus, there is a considerable rationale for studies investigating variation in MTHFR. However, despite the substantial background rationale, the evidence to date remains inconclusive that genetic variants in MTHFR are involved in

28

Barnetson and Dunlop

colorectal cancer susceptibility (155). At best, the evidence is only suggestive of a moderately reduced colorectal cancer risk associated with the 677TT genotype, especially when folate intake is high. However, there appears to be a complex relationship between alcohol consumption, folate, and genotype. Once again, several studies have undertaken subgroup analysis and appear to demonstrate an effect for some risk groups, but it is clear that there is a need for large population-based association studies in which risk exposure and dietary history are systematically recorded in order to determine whether or not genetic variation in MTHFR plays a role in colorectal cancer susceptibility. STK15 STK15 (Aurora-A) is involved in mitotic chromosomal segregation, and aberrant gene function has been implicated in neoplastic transformation and progression. A genetic variant, F31I, was identified through mapping a cancer susceptibility locus in outbred mice (156), and has been shown to have cell transformation capacity as well as being associated with aneuploidy in colorectal cancer cells. In a meta-analysis of case–control studies, there is a significantly increased risk of colorectal cancer in homozygotes, with an odds ratio of 1.5 (157). This is supported by biological rationale because aneuploidy and chromosomal instability are a hallmark of the majority of colorectal cancers. Interestingly the gene resides in a chromosomal area showing allelic imbalance in Finnish familial colorectal cancer cases (158). Thus, understanding the fundamental basis of cancer, such as aneuploidy in colon cancer cells, can allow convergence of data from different research strategies and provide evidence to support the causal role of particular genes in cancer susceptibility, as well as providing reassurance that data generated in animal models and in vitro systems have relevance to human cancer. STK15 is only one of many genes involved in mitotic chromosomal segregation, and so these findings hint at mechanisms that could induce tumor cell aneuploidy as a manifestation of colorectal cancer susceptibility. Cytochrome P-450 Cytochrome P-450 is involved in activation and detoxification of many carcinogenic substances and especially those in smoke. Cigarette smoking has been shown to be a risk factor for colorectal adenomas and colorectal cancer, especially in men. Polycyclic aromatic hydrocarbons are activated by cytochrome P4501A1 (CYP1A1) and detoxified by glutathione S-transferases. A number of genetic polymorphisms have been identified in genes that are known to be involved in the metabolism of polycyclic aromatic hydrocarbons and other tobacco-related carcinogens. Hence these are plausible candidate susceptibility loci, especially when considering interactions with environmental factors. However, the evidence remains inconclusive with several studies showing a

Genetic Susceptibility to Colorectal Cancer

29

marginal effect but others failing to replicate these findings. The largest and most recent study of almost 2000 cases of colorectal cancer and over 2000 controls (159) suggests that the CYP1A1 genotype alone is not associated with colon or rectal cancer. However, in combination with a risk GSTM1 allele and an allele associated with an NAT2 rapid-acetylator phenotype, there was a significantly increased risk of colon cancer (OR ¼ 1.7), especially among men who smoke (OR ¼ 2.5). A similar result was noted for male smokers with respect to rectal cancer. This large study provides substantial evidence for a contribution of GSTM1, CYP1A1, and NAT2 alleles, although replication studies are needed. Nonetheless, the data support the association between smoking and large bowel malignancy, as well as indicating a genetic modifying effect through CYP1A1 and GSTM1 genotypes. Cell Cycle Checkpoint Kinase 2 Cell cycle checkpoint kinase 2 (CHEK2) is involved in cellular responses to DNA damage that culminate in cell cycle arrest, apoptosis, and/or DNA repair. A variant in CHEK2, 1100delC, has been shown to be causally involved in a small proportion of breast cancer families and acts as a lowpenetrance susceptibility allele for breast cancer, with a twofold increased breast cancer risk for carriers. In some breast cancer families there is an excess of colorectal cancer and so by inference, CHEK2 has been proposed as a colorectal cancer susceptibility gene (160). However, although descriptive studies in breast and colorectal cancer families seem to indicate a common genetic etiology, namely, the CHEK2 1100delC allele, the largest association study to date has failed to show any convincing effect (161). A recent study has suggested an effect in certain risk subgroups (162), but again large studies are clearly required to determine definitively whether CHEK2 variants confer any excess risk of colorectal cancer, the level of any associated risk, and the overall contribution of such alleles to colorectal cancer disease burden. In short, the role of CHEK2 in colorectal cancer susceptibility remains unresolved. Other Putative Risk Alleles Apart from those discussed earlier, there are a substantial number of putative colorectal cancer risk alleles that have been proposed with varying degrees of support (146). We discuss a few of these in outline here in order to provide a flavor of those which appear to have at least some support for impacting on colorectal cancer risk. A common length polymorphism in gene encoding the type 1 receptor for TGFb, TGFbR16A, has been extensively studied in a number of cancers and a meta-analysis performed (163). The evidence for a role in colorectal cancer susceptibility is inconclusive since the effect appears to be related to population with an odds ratio of 1.38 in the data from studies of

30

Barnetson and Dunlop

populations living in the United States, while there was no effect for European populations. Nonetheless, components of the TGF signaling pathway constitute strong candidate genes in view of the antiproliferative and apoptosis-promoting role of TGFb. Furthermore, if confirmed in larger population-based studies, the TGFbR16A polymorphism could have considerable public health importance as the carrier frequency is
15%, implying an overall contribution of 1.2% to the total burden of colorectal cancer, greater than that of FAP and approaching that of HNPCC alleles. Loss of imprinting of the insulin-like growth factor II gene (IGF2) has been found in the normal colonic mucosa in around one-third of patients with colorectal cancer, a threefold excess in comparison to controls (164). Patients with colorectal cancer had a significantly greater loss of IGF2 imprinting (OR ¼ 21.7) in circulating lymphocytes, indicating it may be a marker for colorectal cancer. However, further evidence also indicates that not only is loss of IGF2 imprinting heritable but it also is associated with increased tumor burden in an animal model (165). Thus it is possible that epigenetic mechanisms are also involved in colorectal cancer susceptibility, further complicating attempts to identify the genes contributing to the disease burden. Apolipoprotein E (apoE) is involved in bile acid, cholesterol, and triglyceride metabolism. Variants in apoE have been associated with a number of diseases and recently their involvement in cancer has been proposed. Since bile acids have been causally implicated in colorectal cancer etiology, it is certainly plausible that colorectal cancer risk might be modified by apoE genotype. One sizable study suggests that the E3 apoE allele was associated with a protective effect, with around 30% excess risk for those not carrying the protective allele, especially those aged 65 years or older (166). There are many other putative risk alleles that have been identified in various populations, including insulin-like growth factor 1 (IGF1), its binding protein (IGFBP3), polymorphisms in COX-2 genes and in other genes encoding components of the arachidonic acid metabolism pathway, a length variant in the H-ras gene, variants in GSTT1, a C/T variant 13910 bp upstream of the lactase coding sequence in the Finnish population. However, each of these candidates, however plausible the gene or pathway, requires to be tested in large case–control series and then replicated in other populations, again with large case–control series. The size of the primary studies should be numbered in thousands given likely effects and the problems of multiple testing with similar numbers used to undertake the validation and replication studies. Only then can a reasonable view be taken on the likely involvement of common variation in large bowel malignancy. CONCLUSIONS Identifying genes responsible for the autosomal dominant colorectal cancer susceptibility syndromes such as HNPCC and FAP has resulted in major

Genetic Susceptibility to Colorectal Cancer

31

advances in patient management, translating directly into a reduction in death rate from malignancy for gene carriers. It is overinterpretation to assume that such dramatic benefits to individuals will result from identifying lower-penetrance alleles. However, major public health benefits could result from such discoveries, and it is likely that worthwhile benefits for the individual may also result. Perhaps the most important consequence of discovering low-penetrance alleles is that this will afford novel insight into disease causation. This will open up whole new disciplines of rational drug development aimed at risk reduction, much as has been the case for modern prevention of coronary heart disease. While genetic studies of highly penetrant alleles that lead to HNPCC and FAP are relatively straightforward using well-established strategies such as genetic linkage analysis, there are considerable practical and analytical obstacles to dissecting the role of common genetic variation in the etiology of colorectal cancer. Case–control, or genetic association, studies are the main strategy employed at present. However, the reported associations are not well supported by robust statistics and study design, combined with replication studies in similarly large collections of intensively phenotyped cases and controls. As new technologies come onstream that allow whole genome approaches to be brought to bear on the complex genetics of colorectal cancer, many more putative candidates will be identified. Thus, it is important that resource and effort is not diluted by undertaking underpowered association studies, many of which currently fill the literature with incomplete answers to very difficult problems. Hence, there is a pressing need for very large population-based studies in which the epidemiological design is both robust and comprehensive. The results of such studies are awaited with considerable anticipation and optimism.

ACKNOWLEDGMENTS The relevant work on genetic susceptibility to colorectal cancer ongoing in the laboratories of the Colon Cancer Genetics Group in which the work of RAB and MGD is funded by a Cancer Research UK Programme Grant (C348/ A3758) and by grants from the Chief Scientist Office (K/OPR/2/2/D333 and CZB/4/94) and the Medical Research Council (G0000657–53203).

REFERENCES 1. Dunlop MG. Screening for large bowel neoplasms in individuals with a family history of colorectal cancer. Br J Surg 1992; 79:488–494. 2. Cannon-Albright LA, Skolnick MH, Bishop DT, Lee RG, Burt RW. Common inheritance of susceptibility to colonic adenomatous polyps and associated colorectal cancers. N Engl J Med 1988; 319:533–537.

32

Barnetson and Dunlop

3. Lichtenstein P, Holm NV, Verkasalo PK, et al. Environmental and heritable factors in the causation of cancer-analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med 2000; 343:78–85. 4. Lynch HT, Smyrk TC, Watson P, et al. Genetics, natural history, tumor spectrum, and pathology of hereditary nonpolyposis colorectal cancer: an updated review. Gastroenterology 1993; 104:1535–1549. 5. Lynch HT, Smyrk T, Lynch JF. Molecular genetics and clinical-pathology features of hereditary nonpolyposis colorectal carcinoma (Lynch syndrome): historical journey from pedigree anecdote to molecular genetic confirmation. Oncology 1998; 55:103–108. 6. Lynch HT, de la Chapelle A. Genetic susceptibility to non-polyposis colorectal cancer. J Med Genet 1999; 36:801–818. 7. Terdiman JP, Conrad PG, Sleisenger MH. Genetic testing in hereditary colorectal cancer: indications and procedures. Am J Gastroenterol 1999; 94:2344–2356. 8. Vasen HF, Wijnen JT, Menko FH, et al. Cancer risk in families with hereditary nonpolyposis colorectal cancer diagnosed by mutation analysis. Gastroenterology 1996; 110:1020–1027. 9. Dunlop MG, Farrington SM, Nicholl I, et al. Population carrier frequency of hMSH2 and hMLH1 mutations. Br J Cancer 2000; 83:1643–1645. 9a. http://www.ncbi.nlm.nih.gov/entrez/dispomim.cgi?id¼120435. 10. Plaschke J, Engel C, Kruger S, et al. Lower incidence of colorectal cancer and later age of disease onset in 27 families with pathogenic MSH6 germline mutations compared with families with MLH1 or MSH2 mutations: the German Hereditary Nonpolyposis Colorectal Cancer Consortium. J Clin Oncol 2004; 22:4486–4494. 11. Vasen HF, Watson P, Mecklin JP, Lynch HT. New clinical criteria for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch syndrome) proposed by the International Collaborative Group on HNPCC. Gastroenterology 1999; 116:1453–1456. 12. Umar A, Boland CR, Terdiman JP, et al. Revised Bethesda Guidelines for hereditary nonpolyposis colorectal cancer (Lynch syndrome) and microsatellite instability. J Natl Cancer Inst 2004; 96:261–268. 13. Farrington SM, Lin-Goerke J, Ling J, et al. Systematic analysis of hMSH2 and hMLH1 in young colon cancer patients and controls. Am J Hum Genet 1998; 63:749–759. 14. Mecklin JP, Sipponen P, Jarvinen HJ. Histopathology of colorectal carcinomas and adenomas in cancer family syndrome. Dis Colon and Rectum 1986; 29:849–853. 15. Jass JR. HNPCC and sporadic MSI-H colorectal cancer: a review of the morphological similarities and differences. Fam Cancer 2004; 3:93–100. 16. Kouri M, Laasonen A, Mecklin JP, Jarvinen H, Franssila K, Pyrhonen S. Diploid predominance in hereditary nonpolyposis colorectal carcinoma evaluated by flow cytometry. Cancer 1990; 65:1825–1829. 17. de la Chapelle A. Genetic predisposition to colorectal cancer. Nat Rev Cancer 2004; 4:769–780. 18. Iaccarino I, Marra G, Palombo F, Jiricny J. hMSH2 and hMSH6 play distinct roles in mismatch binding contribute differently to the ATPase activity of hMutSalpha. EMBO J 1998; 17:2677–2686.

Genetic Susceptibility to Colorectal Cancer

33

19. Ionov Y, Peinado MA, Malkhosyan S, Shibata D, Perucho M. Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis. Nature 1993; 363:558–561. 20. Peltomaki P, Aaltonen LA, Sistonen P, et al. Genetic mapping of a locus predisposing to human colorectal cancer [see comments]. Science 1993; 260:810–812. 21. Lindblom A, Tannergard P, Werelius B, Nordenskjold M. Genetic mapping of a second locus predisposing to hereditary non-polyposis colon cancer. Nat Genet 1993; 5:279–282. 22. Strand M, Prolla TA, Liskay RM, Petes TD. Destabilization of tracts of simple repetitive DNA in yeast by mutations affecting DNA mismatch repair [see comments] [published erratum appears in Nature 1994 Apr 7;368(6471);569]. Nature 1993; 365:274–276. 23. Bronner CE, Baker SM, Morrison PT, et al. Mutation in the DNA mismatch repair gene homologue hMLH1 is associated with hereditary non-polyposis colon cancer. Nature 1994; 368:258–261. 24. Fishel R, Lescoe MK, Rao MR, et al. The human mutator gene homolog MSH2 and its association with hereditary nonpolyposis colon cancer [published erratum appears in Cell 1994 Apr 8;77(1):167]. Cell 1993; 75:1027–1038. 25. Leach FS, Nicolaides NC, Papadopoulos N, et al. Mutations of a muts homolog in hereditary nonpolyposis colorectal cancer. Cell 1993; 75:1215–1225. 26. Liu B, Farrington SM, Petersen GM, et al. Genetic instability occurs in the majority of young patients with colorectal cancer. Nature Medicine 1995; 1: 348–352. 27. Aaltonen LA, Salovaara R, Kristo P, et al. Incidence of hereditary nonpolyposis colorectal cancer and the feasibility of molecular screening for the disease. N Eng J Med 1998; 338:1481–1487. 28. Hampel H, Frankel WL, Martin E, et al. Screening for the Lynch syndrome (hereditary nonpolyposis colorectal cancer). N Engl J Med 2005; 352:1851–1860. 29. Mitchell RJ, Farrington SM, Dunlop MG, Campbell H. Mismatch repair genes hMLH1 and hMSH2 and colorectal cancer: a HuGE review. Am J Epidemiol 2002; 156:885–902. 30. Peltomaki P. Deficient DNA mismatch repair: a common etiologic factor for colon cancer. Hum Mol Genet 2001; 10:735–740. 31. Akiyama Y, Sato H, Yamada T, et al. Germ-line mutation of the hMSH6/ GTBP gene in an atypical hereditary nonpolyposis colorectal cancer kindred. Cancer Res 1997; 57:3920–3923. 32. Miyaki M, Konishi M, Tanaka K, et al. Germline mutation of MSH6 as the cause of hereditary nonpolyposis colorectal cancer. Nat Genetics 1997; 17: 271–272. 33. Nicolaides NC, Papadopoulos N, Liu B, et al. Mutations of two pms homologues in hereditary nonpolyposis colon cancer. Nature 1994; 371:75–80. 34. Wu Y, Berends MJ, Sijmons RH, et al. A role for MLH3 in hereditary nonpolyposis colorectal cancer. Nat Genet 2001; 29:137–138. 35. Aaltonen LA, Peltomaki P, Leach FS, et al. Clues to the pathogenesis of familial colorectal cancer. Science 1993; 260:812–816. 36. Thibodeau SN, Bren G, Schaid D. Microsatellite instability in cancer of the proximal colon. Science 1993; 260:816–819.

34

Barnetson and Dunlop

37. Dietmaier W, Wallinger S, Bocker T, Kullmann F, Fishel, Ruschoff J. Diagnostic microsatellite instability: definition and correlation with mismatch repair protein expression. Cancer Res 1997; 57:4749–4756. 38. Herman JG, Umar A, Polyak K, et al. Incidence and functional consequences of hMLH1 promoter hypermethylation in colorectal carcinoma. Proc Natl Acad Sci USA 1998; 95:6870–6875. 39. Thibodeau SN, French AJ, Cunningham JM, et al. Microsatellite instability in colorectal cancer: different mutator phenotypes and the principal involvement of hMLH1. Cancer Res 1998; 58:1713–1718. 40. Dunlop MG, Farrington SM, Carothers AD, et al. Cancer risk associated with germline DNA mismatch repair gene mutations. Hum Mol Genet 1997; 6:105–110. 41a.Quehenberger F, Vasen HF, van Houwelingen HC. Risk of colorectal and endometrial cancer for carriers of mutations of the hMLH1 and hMSH2 gene: correction for ascertainment. J Med Genet 2005; 42:491–496. 41b.(http://www1.hgu.mrc.ac.uk/softdata/MMRpredict.php). 42. Bubb VJ, Curtis LJ, Cunningham C, et al. Microsatellite instability and the role of hMSH2 in sporadic colorectal cancer. Oncogene 1996; 12:2641–2649. 43. Forster S, Sattler HP, Hack M, et al. Microsatellite instability in sporadic carcinomas of the proximal colon: association with diploid DNA content, negative protein expression of p53, and distinct histomorphologic features. Surgery 1998; 123:13–18. 44. Gryfe R, Kim H, Hsieh ETK, et al. Tumor microsatellite instability and clinical outcome in young patients with colorectal cancer. N Engl J Med 2000; 342:69–77. 45. Halling KC, French AJ, McDonnell SK, et al. Microsatellite instability and 8p allelic imbalance in stage B2 and C colorectal cancers. J Natl Cancer Inst 1999; 91:1295–1303. 46. Kim HG, Jen J, Vogelstein B, Hamilton SR. Clinical and pathological characteristics of sporadic colorectal carcinomas with DNA replication errors in microsatellite sequences. Am J Pathol 1994; 145:148–156. 47. Lothe RA, Peltomaki P, Meling GI, et al. Genomic instability in colorectal cancer: relationship to clinicopathological variables and family history. Cancer Res 1993; 53:5849–5852. 48. Lukish JR, Muro K, DeNobile J, et al. Prognostic significance of DNA replication errors in young patients with colorectal cancer. Ann Surg 1998; 227:51–56. 49. Wright CM, Dent OF, Barker M, et al. Prognostic significance of extensive microsatellite instability in sporadic clinicopathological stage C colorectal cancer. Br J Surg 2000; 87:1197–1202. 50. Curran B, Lenehan K, Mulcahy H, et al. Replication error phenotype, clinicopathological variables, and patient outcome in Dukes’ B stage II (T3,N0,M0) colorectal cancer. Gut 2000; 46:200–204. 51. Feeley KM, Fullard JF, Heneghan MA, et al. Microsatellite instability in sporadic colorectal carcinoma is not an indicator of prognosis. J Pathol 1999; 188:14–17. 52. Johannsdottir JT, Bergthorsson JT, Gretarsdottir S, et al. Replication error in colorectal carcinoma: association with loss of heterozygosity at mismatch repair loci and clinicopathological variables. Anticancer Res 1999; 19:1821–1826.

Genetic Susceptibility to Colorectal Cancer

35

53. Ko JM, Cheung MH, Kwan MW, et al. Genomic instability and alterations in Apc, Mcc and Dcc in Hong Kong patients with colorectal carcinoma. Int J Cancer 1999; 84:404–409. 54. Messerini L, Ciantelli M, Baglioni S, Palomba A, Zampi G, Papi L. Prognostic significance of microsatellite instability in sporadic mucinous colorectal cancers. Hum Pathol 1999; 30:629–634. 55. Salahshor S, Kressner U, Fischer H, et al. Microsatellite instability in sporadic colorectal cancer is not an independent prognostic factor. Br J Cancer 1999; 81:190–193. 56. Boland CR, Thibodeau SN, Hamilton SR, et al. A National Cancer Institute Workshop on Microsatellite Instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 1998; 58:5248– 5257. 57. Aarnio M, Mustonen H, Mecklin JP, Jarvinen HJ. Prognosis of colorectal cancer varies in different high-risk conditions. Ann Med 1998; 30:75–80. 58. Sankila R, Aaltonen LA, Jarvinen HJ, Mecklin JP. Better survival rates in patients with MLH1-associated hereditary colorectal cancer. Gastroenterology 1996; 110:682–687. 59. Watson P, Lin KM, Rodriguez-Bigas MA, et al. Colorectal carcinoma survival among hereditary nonpolyposis colorectal carcinoma family members. Cancer 1998; 83:259–266. 60. Lin KM, Shashidharan M, Ternent CA, et al. Colorectal and extracolonic cancer variations in MLH1/MSH2 hereditary nonpolyposis colorectal cancer kindreds and the general population. Dis Colon Rectum 1998; 41:428–433. 61. Bertario L, Russo A, Sala P, et al. Survival of patients with hereditary colorectal cancer: comparison of HNPCC and colorectal cancer in FAP patients with sporadic colorectal cancer. Int J Cancer 1999; 80:183–187. 62. Kee F, Collins BJ, Patterson CC. Prognosis in familial non-polyposis colorectal cancer. Gut 1991; 32:513–516. 63. Slattery ML, Kerber RA. The impact of family history of colon cancer on survival after diagnosis with colon cancer. Int J Epidemiol 1995; 24:888–896. 64a.Bignami M, Casorelli I, Karran P. Mismatch repair and response to DNAdamaging antitumor therapies. Eur J Cancer 2003; 39:2142–2149. 64b.Barnetson, Tenesa A, Farrington SM, et al. Identification of carriers of mutant DNA Mismatch Repair Genes in Patients with Colon Cancer. New Engl J Med 2006; 354: 2751–2763. 65. Karran P, Hampson R. Genomic instability and tolerance to alkylating agents. J Cancer Surv 1996; 28:69–85. 66. Carethers JM, Chauhan DP, Fink D, et al. Mismatch repair proficiency and in vitro response to 5-fluorouracil. Gastroenterology 1999; 117:123–131. 67. Kat A, Thilly WG, Fang WH, Longley MJ, Li GM, Modrich P. An alkylationtolerant, mutator human cell line is deficient in strand-specific mismatch repair. Proc Natl Acad Sci USA 1993; 90:6424–6428. 68. Griffin S, Branch P, Xu YZ, Karran P. DNA mismatch binding and incision at modified guanine bases by extracts of mammalian cells: implications for tolerance to DNA methylation damage. Biochemistry 1994; 33:4787–4793.

36

Barnetson and Dunlop

69. Aebi S, Kurdihaidar B, Gordon R, et al. Loss of DNA mismatch repair in acquired resistance to cisplatin. Cancer Res 1996; 56:3087–3090. 70. Fallik D, Borrini F, Boige V, et al. Microsatellite instability is a predictive factor of the tumor response to irinotecan in patients with advanced colorectal cancer. J Cancer Res 2003; 63:5738–5744. 71. Peters GJ, van Groeningen CJ, van der Wilt CL, et al. Time course of inhibition of thymidylate synthase in patients treated with fluorouracil and leucovorin. Semin Oncol 1992; 19:26–35. 72. Aebi S, Fink D, Gordon R, et al. Resistance to cytotoxic drugs in DNA mismatch repair-deficient cells. J Clin Cancer Res1997; 3:1763–1767. 73. Meyers M, Wagner MW, Hwang HS, Kinsella TJ, Boothman DA. Role of the hMLH1 DNA mismatch repair protein in fluoropyrimidine-mediated cell death and cell cycle responses. Cancer Res 2001; 61:5193–5201. 74. Rosty C, Chazal M, Etienne MC, et al. Determination of microsatellite instability, p53 and K-RAS mutations in hepatic metastases from patients with colorectal cancer: relationship with response to 5-fluorouracil and survival. 2001; 20;95:162–167. 75. Liang JT, Huang KC, Lai HS, et al. High-frequency microsatellite instability predicts better chemosensitivity to high-dose 5-fluorouracil plus leucovorin chemotherapy for stage IV sporadic colorectal cancer after palliative bowel resection. Int J Cancer 2002; 20;101:519–525. 76. Elsaleh H, Joseph D, Grieu F, Zeps N, Spry N, Iacopetta B. Association of tumor site and sex with survival benefit from adjuvant chemotherapy in colorectal cancer. Lancet 2000; 355:1745–1750. 77. Hemminki A, Mecklin JP, Jarvinen H, Aaltonen LA, Joensuu H. Microsatellite instability is a favorable prognostic indicator in patients with colorectal cancer receiving chemotherapy. Gastroenterology 2000; 119:921–928. 78. Watanabe T, Wu T, Catalano PJ, et al. Molecular predictors of survival after adjuvant chemotherapy for colon cancer. N Engl J Med 2001; 344:1196–1206. 79. Barratt PL, Seymour MT, Stenning SP, et al. DNA markers predicting benefit from adjuvant fluorouracil in patients with colon cancer: a molecular study. Lancet 2002; 360:1381–1391. 80. Ribic CM, Sargent DJ, Moore MJ, et al. Tumor microsatellite-instability status as a predictor of benefit from fluorouracil-based adjuvant chemotherapy for colon cancer. N Engl J Med 2003; 349:247–257. 81. Carethers JM, Smith EJ, Behling CA, et al. Use of 5-fluorouracil and survival in patients with microsatellite-unstable colorectal cancer. Gastroenterology 2004; 126:394–401. 82. Aarnio M, Sankila R, Pukkala E, et al. Cancer risk in mutation carriers of DNA mismatch-repair genes. Int J Cancer 1999; 81:214–218. 83. Dunlop MG. Guidance on gastrointestinal surveillance for hereditary nonpolyposis colorectal cancer, familial adenomatous polypolis, juvenile polyposis, and Peutz–Jeghers syndrome. Gut 2002; 51(suppl 5):V21–V27. 84. Jarvinen HJ, Aarnio M, Mustonen H, et al. Controlled 15-year trial on screening for colorectal cancer in families with hereditary nonpolyposis colorectal cancer. Gastroenterology 2000; 118:829–834.

Genetic Susceptibility to Colorectal Cancer

37

85. de Vos tot Nederveen Cappel WH, Nagengast FM, Griffioen G, et al. Surveillance for hereditary nonpolyposis colorectal cancer: a long-term study on 114 families. Dis Colon Rectum 2002; 45:1588–1594. 86. Rodriguez-Bigas MA, Vasen HF, Pekka-Mecklin J, et al. Rectal cancer risk in hereditary nonpolyposis colorectal cancer after abdominal colectomy. International Collaborative Group on HNPCC. Ann Surg 1997; 225:202–207. 87. de Vos tot Nederveen Cappel WH, Buskens E, van Duijvendijk P, et al. Decision analysis in the surgical treatment of colorectal cancer due to a mismatch repair gene defect. Gut 2003; 52:1752–1755. 88. Bisgaard ML, Fenger K, Bulow S, Niebuhr E, Mohr J. Familial adenomatous polyposis (FAP): frequency, penetrance, and mutation rate. Hum Mutat 1994; 3:121–125. 89. Bulow S, Bulow C, Nielsen TF, Karlsen L, Moesgaard F. Centralized registration, prophylactic examination, and treatment results in improved prognosis in familial adenomatous polyposis. Results from the Danish Polyposis Register. Scand J Gastroenterol 1995; 30:989–993. 89a.http://www.ncbi.nlm.nih.gov/entrez/dispomim.cgi?id¼175100. 90. Groves CJ, Saunders BP, Spigelman AD, Phillips RK. Duodenal cancer in patients with familial adenomatous polyposis (FAP): results of a 10 year prospective study. Gut 2002; 50:636–641. 91. Poritz LS, Blackstein M, Berk T, Gallinger S, McLeod RS, Cohen Z. Extended follow-up of patients treated with cytotoxic chemotherapy for intra-abdominal desmoid tumors. Dis Colon Rectum 2001; 44:1268–1273. 92. Arvanitis ML, Jagelman DG, Fazio VW, Lavery IC, McGannon E. Mortality in patients with familial adenomatous polyposis. Dis Colon Rectum 1990; 33:639–642. 93. Rustgi AK. Hereditary gastrointestinal polyposis and nonpolyposis syndromes. N Engl J Med 1994; 331:1694–1702. 94. Bodmer WF, Bailey CJ, Bodmer J, et al. Localization of the gene for familial adenomatous polyposis on chromosome 5. Nature 1987; 328:614–616. 95. Leppert M, Dobbs M, Scambler P, et al. The gene for familial polyposis coli maps to the long arm of chromosome 5. Science 1987; 238:1411–1413. 96. Kinzler KW, Nilbert MC, Su LK, et al. Identification of FAP locus genes from chromosome 5q21. Science 1991; 253:661–665. 97. Joslyn G, Carlson M, Thliveris A, et al. Identification of deletion mutations and three new genes at the familial polyposis locus. Cell 1991; 66:601–613. 98. Nagase H, Nakamura Y. Mutations of the APC (adenomatous polyposis coli) gene. Hum Mutat 1993; 2:425–434. 99. Mandl M, Paffenholz R, Friedl W, Caspari R, Sengteller M, Propping P. Frequency of common and novel inactivating APC mutations in 202 families with familial adenomatous polyposis. Hum Mol Genet 1994; 3:181–184. 100. Smith KJ, Johnson KA, Bryan TM, et al. The APC gene product in normal and tumor cells. Proc Natl Acad Sci USA 1993; 90:2846–2850. 101. Nathke IS, Adams CL, Polakis P, Sellin JH, Nelson WJ. The adenomatous polyposis coli tumor suppressor protein localizes to plasma membrane sites involved in active cell migration. J Cell Biol 1996; 134:165–179.

38

Barnetson and Dunlop

102. Sancho E, Batlle E, Clevers H. Signaling pathways in intestinal development and cancer. Annu Rev Cell Dev Biol 2004; 20:695–723. 103. He TC, Sparks AB, Rago C, et al. Identification of c-MYC as a target of the APC pathway. Science 1998; 281:1509–1512. 104. Polakis P, Hart M, Rubinfeld B. Defects in the regulation of beta-catenin in colorectal cancer. Adv Exp Med Biol 1999; 470:23–32. 105. Fodde R, Kuipers J, Rosenberg C, et al. Mutations in the APC tumor suppressor gene cause chromosomal instability. Nat Cell Biol 2001; 3:433–438. 106. Kaplan KB, Burds AA, Swedlow JR, Bekir SS, Sorger PK, Nathke IS. A role for the adenomatous polyposis coli protein in chromosome segregation. Nat Cell Biol 2001; 3:429–432. 107. Smith KJ, Levy DB, Maupin P, Pollard TD, Vogelstein B, Kinzler KW. Wildtype but not mutant APC associates with the microtubule cytoskeleton. Cancer Res 1994; 54:3672–3675. 108. Su LK, Burrell M, Hill DE, et al. APC binds to the novel protein EB1. Cancer Res 1995; 55:2972–2977. 109. Matsumine A, Ogai A, Senda T, et al. Binding of APC to the human homolog of the Drosophila discs large tumor suppressor protein. Science 1996; 272: 1020–1023. 110. Nagase H, Miyoshi Y, Horii A, et al. Correlation between the location of germline mutations in the APC gene and the number of colorectal polyps in familial adenomatous polyposis patients. Cancer Res 1992; 52:4055–4057. 111. Paul P, Letteboer T, Gelbert L, Groden J, White R, Coppes MJ. Identical APC exon 15 mutations result in a variable phenotype in familial adenomatous polyposis. Hum Mol Genet 1993; 2:925–931. 112. Knudsen AL, Bisgaard ML, Bulow S. Attenuated familial adenomatous polyposis (AFAP). A review of the literature. Fam Cancer 2003; 2:43–55. 113. Eccles DM, van der Luijt R, Breukel C, et al. Hereditary desmoid disease due to a frameshift mutation at codon 1924 of the APC gene. Am J Hum Genet 1996; 59:1193–1201. 114. Hamilton SR, Liu B, Parsons RE, et al. The molecular basis of Turcot’s syndrome. N Engl J Med 1995; 332:839–847. 115. Nugent KP, Phillips RK. Rectal cancer risk in older patients with familial adenomatous polyposis and an ileorectal anastomosis: a cause for concern. Br J Surg 1992; 79:1204–1206. 116. Bertario L, Russo A, Radice P, et al. Genotype and phenotype factors as determinants for rectal stump cancer in patients with familial adenomatous polyposis. Hereditary Colorectal Tumors Registry. Ann Surg 2000; 231:538–543. 117. Bulow C, Vasen H, Jarvinen H, Bjork J, Bisgaard ML, Bulow S. Ileorectal anastomosis is appropriate for a subset of patients with familial adenomatous polyposis. Gastroenterology 2000; 119:1454–1460. 118. Vasen HF, van Duijvendijk P, Buskens E, et al. Decision analysis in the surgical treatment of patients with familial adenomatous polyposis: a Dutch-Scandinavian collaborative study including 659 patients. Gut 2001; 49:231–235. 119. Wu JS, Paul P, McGannon EA, Church JM. APC genotype, polyp number, and surgical options in familial adenomatous polyposis. Ann Surg 1998; 227: 57–62.

Genetic Susceptibility to Colorectal Cancer

39

120. van Duijvendijk P, Slors JF, Taat CW, Oosterveld P, Vasen HF. Functional outcome after colectomy and ileorectal anastomosis compared with proctocolectomy and ileal pouch-anal anastomosis in familial adenomatous polyposis. Ann Surg 1999; 230:648–654. 121. van Duijvendijk P, Slors JF, Taat CW, et al. Quality of life after total colectomy with ileorectal anastomosis or proctocolectomy and ileal pouch-anal anastomosis for familial adenomatous polyposis. Br J Surg 2000; 87:590–596. 122. Olsen KO, Juul S, Bulow S, et al. Female fecundity before and after operation for familial adenomatous polyposis. Br J Surg 2003; 90:227–231. 123. Lindor NM, Greene MH. The concise handbook of family cancer syndromes. Mayo Familial Cancer Program. J Natl Cancer Inst 1998; 90:1039–1071. 124. Giardiello FM, Welsh SB, Hamilton SR, et al. Increased risk of cancer in the Peutz-Jeghers syndrome. N Engl J Med 1987; 316:1511–1514. 125. Spigelman AD, Murday V, Phillips RK. Cancer and the Peutz-Jeghers syndrome. Gut 1989; 30:1588–1590. 126. Lim W, Olschwang S, Keller JJ, et al. Relative frequency and morphology of cancers in STK11 mutation carriers. Gastroenterology 2004; 126:1788–1794. 127. Hemminki A, Markie D, Tomlinson I, et al. A serine/threonine kinase gene defective in Peutz-Jeghers syndrome. Nature 1998; 391:184–187. 127a.http://www.ncbi.nlm.nih.gov/entrez/dispomim.cgi?id¼175200. 128. Jarvinen H, Franssila KO. Familial juvenile polyposis coli; increased risk of colorectal cancer. Gut 1984; 25:792–800. 129. Desai DC, Neale KF, Talbot IC, Hodgson SV, Phillips RK. Juvenile polyposis. Br J Surg 1995; 82:14–17. 129a.http://www.ncbi.nlm.nih.gov/entrez/dispomim.cgi?id¼174900. 130. Howe JR, Roth S, Ringold JC, et al. Mutations in the SMAD4/DPC4 gene in juvenile polyposis. Science 1998; 280:1086–1088. 131. Woodford-Richens K, Williamson J, Bevan S, et al. Allelic loss at SMAD4 in polyps from juvenile polyposis patients and use of fluorescence in situ hybridization to demonstrate clonal origin of the epithelium. Cancer Res 2000; 60:2477–2482. 132. Heldin CH, Miyazono K, Tendijke P. Tgf-beta signalling from cell membrane to nucleus through SMAD proteins. Nature 1997; 390:465–471. 133. Thiagalingam S, Lengauer C, Leach FS, et al. Evaluation of candidate tumor suppressor genes on chromosome 18 in colorectal cancers. Nat Genet 1996; 13:343–346. 134. Howe JR, Bair JL, Sayed MG, et al. Germline mutations of the gene encoding bone morphogenetic protein receptor 1A in juvenile polyposis. Nat Genet 2001; 28:184–187. 135. Steck PA, Pershouse MA, Jasser SA, et al. Identification of a candidate tumor suppressor gene, MMAC1, at chromosome 10q23.3 that is mutated in multiple advanced cancers. Nat Genet 1997; 15:356–362. 136. Al-Tassan N, Chmiel NH, Maynard J, et al. Inherited variants of MYH associated with somatic G:C–>T:A mutations in colorectal tumors. Nat Genet 2002; 30:227–232. 137. Sampson JR, Dolwani S, Jones S, et al. Autosomal recessive colorectal adenomatous polyposis due to inherited mutations of MYH Biallelic germline mutations in MYH predispose to multiple colorectal adenoma and somatic

40

Barnetson and Dunlop

G:C–>T:A mutations Inherited variants of MYH associated with somatic G:C– >T:A mutations in colorectal tumors. Lancet 2003; 362:39–41. 138. Sieber OM, Lipton L, Crabtree M, et al. Multiple colorectal adenomas, classic adenomatous polyposis, and germ-line mutations in MYH. N Engl J Med 2003; 348:791–799. 138a.http://www.ncbi.nlm.nih.gov/entrez/dispomim.cgi?id¼608456. 139. Croitoru ME, Cleary SP, Di Nicola N, et al. Association between biallelic and monoallelic germline MYH gene mutations and colorectal cancer risk. J Natl Cancer Inst 2004; 96:1631–1634. 140. Farrington SM, Tenesa A, Barnetson R, et al. Germline susceptibility to colorectal cancer due to base-excision repair gene defects. Am J Hum Genet 2005; 77:112–119. 141. Sieber OM, Howarth KM, Thirlwell C, et al. Myh deficiency enhances intestinal tumorigenesis in multiple intestinal neoplasia (ApcMin/þ) mice. Cancer Res 2004; 64:8876–8881. 142. Gu Y, Parker A, Wilson TM, Bai H, Chang DY, Lu AL. Human MutY homolog, a DNA glycosylase involved in base excision repair, physically and functionally interacts with mismatch repair proteins human MutS homolog 2/ human MutS homolog 6. J Biol Chem 2002; 277:11135–11142. 143. Chmiel NH, Livingston AL, David SS. Insight into the functional consequences of inherited variants of the hMYH adenine glycosylase associated with colorectal cancer: complementation assays with hMYH variants and pre-steady-state kinetics of the corresponding mutated E.coli enzymes. J Mol Biol 2003; 327:431–443. 144. Hirano S, Tominaga Y, Ichinoe A, et al. Mutator phenotype of MUTYH-null mouse embryonic stem cells. J Biol Chem 2003; 278:38121–38124. 145. Jaeger EE, Woodford-Richens KL, Lockett M, et al. An ancestral Ashkenazi haplotype at the HMPS/CRAC1 locus on 15q13-q14 is associated with hereditary mixed polyposis syndrome. Am J Hum Genet 2003; 72:1261–1267. 146. Kemp Z, Thirlwell C, Sieber O, Silver A, Tomlinson I. An update on the genetics of colorectal cancer. Hum Mol Genet 2004; 13 Spec No 2:R177–R185. 147. Laken SJ, Petersen GM, Gruber SB, et al. Familial colorectal cancer in Ashkenazim due to a hypermutable tract in APC. Nat Genet 1997; 17:79–83. 148. Lamlum H, Al Tassan N, Jaeger E, et al. Germline APC variants in patients with multiple colorectal adenomas, with evidence for the particular importance of E1317Q. Hum Mol Genet 2000; 9:2215–2221. 149. Hahnloser D, Petersen GM, Rabe K, et al. The APC E1317Q variant in adenomatous polyps and colorectal cancers. Cancer Epidemiol Biomarkers Prev 2003; 12:1023–1028. 150. Slattery ML, Samowitz W, Ballard L, Schaffer D, Leppert M, Potter JD. A molecular variant of the APC gene at codon 1822: its association with diet, lifestyle, and risk of colon cancer. Cancer Res 2001; 61:1000–1004. 151. Lipkin SM, Rozek LS, Rennert G, et al. The MLH1 D132H variant is associated with susceptibility to sporadic colorectal cancer. Nat Genet 2004; 36:694–699. 152. Kong S, Wei Q, Amos CI, et al. Cyclin D1 polymorphism and increased risk of colorectal cancer at young age. J Natl Cancer Inst 2001; 93:1106–1108.

Genetic Susceptibility to Colorectal Cancer

41

153. Le Marchand L, Seifried A, Lum-Jones A, Donlon T, Wilkens LR. Association of the cyclin D1 A870G polymorphism with advanced colorectal cancer. JAMA 2003; 290:2843–2848. 154. Grieu F, Malaney S, Ward R, Joseph D, Iacopetta B. Lack of association between CCND1 G870A polymorphism and the risk of breast and colorectal cancers. Anticancer Res 2003; 23:4257–4259. 155. Sharp L, Little J. Polymorphisms in genes involved in folate metabolism and colorectal neoplasia: a HuGE review. Am J Epidemiol 2004; 159:423–443. 156. Ewart-Toland A, Briassouli P, de Koning JP, et al. Identification of Stk6/ STK15 as a candidate low-penetrance tumor-susceptibility gene in mouse and human. Nat Genet 2003; 34:403–412. 157. Ewart-Toland A, Dai Q, Gao YT, et al. Aurora-A/STK15 Tþ91A is a general low penetrance cancer susceptibility gene: a meta-analysis of multiple cancer types. Carcinogenesis 2005; 26:1368–1373. 158. Laiho P, Hienonen T, Karhu A, et al. Genome-wide allelotyping of 104 Finnish colorectal cancers reveals an excess of allelic imbalance in chromosome 20q in familial cases. Oncogene 2003; 22:2206–2214. 159. Slattery ML, Samowtiz W, Ma K, et al. CYP1A1, cigarette smoking, and colon and rectal cancer. Am J Epidemiol 2004; 160:842–852. 160. Meijers-Heijboer H, Wijnen J, Vasen H, et al. The CHEK2 1100delC mutation identifies families with a hereditary breast and colorectal cancer phenotype. Am J Hum Genet 2003; 72:1308–1314. 161. Kilpivaara O, Laiho P, Aaltonen LA, Nevanlinna H. CHEK2 1100delC and colorectal cancer. J Med Genet 2003; 40:e110. 162. de Jong MM, Nolte IM, Te Meerman GJ, et al. Colorectal cancer and the CHEK2 1100delC mutation. Genes Chromosomes Cancer 2005; 43:377–382. 163. Kaklamani VG, Hou N, Bian Y, et al. TGFBR16A and cancer risk: a metaanalysis of seven case-control studies. J Clin Oncol 2003; 21:3236–3243. 164. Cui H, Cruz-Correa M, Giardiello FM, et al. Loss of IGF2 imprinting: a potential marker of colorectal cancer risk. Science 2003; 299:1753–1755. 165. Sakatani T, Kaneda A, Iacobuzio-Donahue CA, et al. Loss of imprinting of Igf2 alters intestinal maturation and tumorigenesis in mice. Science 2005; 307:1976–1978. 166. Slattery ML, Sweeney C, Murtaugh M, et al. Associations between apoE genotype and colon and rectal cancer. Carcinogenesis 2005; 26:1422–1429.

2 Epidemiology of Colorectal Cancer Julian Little Department of Epidemiology and Community Medicine, University of Ottawa, Ottawa, Ontario, Canada

Linda Sharp National Cancer Registry, Cork, Ireland

INTRODUCTION Cancer of the large bowel is a major health problem. Worldwide each year, over 900,000 new cases are diagnosed, and almost 500,000 people die from the disease (1). About two-thirds of the incident cases occur in developed countries, where colorectal cancer is the third most common cancer in men and second most common in women (2). In developing countries, it is the fifth most common cancer in both sexes. Relatively few colorectal cancers occur in persons younger than 40. Rates increase rapidly with age thereafter, more markedly for colon than for rectal cancer (3). The burden of colorectal cancer is, therefore, expected to increase in the future as a result of population aging and increased life expectancy. This is particularly true for developing countries. DESCRIPTIVE EPIDEMIOLOGY OF COLORECTAL CANCER International Variations in Incidence After allowing for differences between the age structures of populations, there are substantial variations in incidence internationally (Fig. 1). In men, for the period 1993–1997, the highest rates, of 45 per 100,000 and 43

44

Little and Sharp

Figure 1 Age-standardized (world) incidence rates of colorectal cancer per 100,000 population: males and females, 1993–1997. Source: From Ref. 3.

above, occurred in Australia, New Zealand, parts of Japan (Miyagi), and parts of western Europe (Saarland, Germany; Bas-Rhin, France; northeast Italy) (3). Rates in the range of 35 to 45 per 100,000 were observed in the rest of western Europe, the United States, Canada, Hong Kong, and in Israeli Jews. In eastern Europe incidence was somewhat lower—around 25 to 35 per 100,000. Incidence rates of less than 15 per 100,000 occurred in Africa, India, Thailand, and Vietnam, and parts of the Middle East. For women, the geographical pattern was similar, but the age-standardized rates were about 60% to 80% of those in men. Time Trends in Incidence In 1971, Haenszel and Correa noted that colon cancer incidence was slowly increasing (4). Since then, moderate increases in colorectal cancer incidence is observed in many populations, although the timing and the magnitude of the increases have differed between populations (5). Rates have risen both in populations that, in earlier decades, had intermediate or high rates of colorectal cancer—such as Sweden, Denmark, Spain, Italy, Australia, New Zealand, Britain, and the United States—and in those that had low rates—such as

Epidemiology of Colorectal Cancer

45

Japan (6–15). Although the general pattern is similar, the magnitude of the increase differed between populations, as did the timing. In most, the increase was either more pronounced in men than women or observed only in men. While these trends are in part an artifact of improvements in the efficiency of cancer registration and increased detection rates resulting from the introduction of newer diagnostic tools, this seems unlikely to be the full explanation. The different patterns in males and females indicate strong sex-specific cohort effects, most likely associated with changes in exposures to environmental and lifestyle risk factors for the disease. Subsite of Tumor Between 60% and 70% of large bowel cancers occur in the colon (16). In western European and U.S. data, tumors of the right (proximal) colon are overrepresented among women (6). This is partly a function of age, because right tumors are more common among older persons and there are greater numbers of women in the older age groups than men. Tumors of the right colon have been reported to have become more common over the past 30 years (17–20). However, these observations are difficult to interpret for a number of reasons including different data categorizations and methods of statistical analysis, selection bias, population aging, and increasing use of colonoscopy and flexible sigmoidoscopy (5). It is, therefore, not at all clear whether the underlying incidence of right colon tumors is truly increasing. Variations in Incidence Within Countries Ethnic Origin In the United States in the period 1973–1997, the incidence and mortality due to large bowel cancer was higher in blacks than in other ethnic groups (15). The incidence in blacks in the United States is substantially higher than in Africa (3). In England and Wales in the period 1970–1985, the death rate due to large bowel cancer in people born in the Caribbean and East Africa was about half that of those born in England and Wales (21). Nevertheless, this would suggest a higher incidence than in the populations from which they originated. Death rates from the disease in those born in West Africa were similar to those born in England and Wales. These data from the United Kingdom contrasts with those from the United States, but there are differences in the pattern and circumstances of settlement and study period. As regards other ethnic groups in the United States, in men in the period 1973–1997, the incidence was lower in Asian/Pacific Islanders than in white males, and lower again in Hispanic and Native Americans and Alaskan Natives than in Asian/Pacific Islanders (15). Among women incidence rates were similar for Asian/Pacific Islanders, Hispanics and Native Americans, and Alaskans, and were lower than the levels observed in white and black women.

46

Little and Sharp

In addition to levels of incidence varying by ethnic group, there are some differences in time trends. In the United States while incidence has been declining in the white population since the mid-1980s, this trend has not been seen in the black population (15). Socioeconomic Status In most countries, the risk of colon cancer has been found to be higher in those with higher socioeconomic status (22). This has been observed in both men and women, both for incidence and mortality, and for diverse measures of socioeconomic status. The increase in risk with increasing socioeconomic status contrasts with most other types of cancer. No consistent association between cancer of the rectum and socioeconomic status has been observed. Survival and Mortality: International Variations and Time Trends In developed countries, colorectal cancer death rates have declined steadily over the past 20 to 30 years (23). This is due, at least in part, to declining proportions of patients presenting with more advanced disease over time (24,25), and most likely a consequence of increased availability and use of sigmoidoscopy, colonoscopy and, possibly, fecal occult blood (FOB) testing. The extent of disease at diagnosis is a strong predictor of survival for both colon and rectal tumors. USA Surveillance, Epidemiology and End Results (SEER) Program data for patients diagnosed with colon cancer in 1992–1997 show five-year relative survival of 91% for those whose disease was localized at diagnosis, 67% for those presenting with regional spread, and 9% for those with distant metastasis (15). For rectal cancer the corresponding figures were 87%, 57%, and 8%. Of all those for whom the extent of disease was known, 39% had localized disease, 40% regional spread, and 21% distant metastasis. In most of western and northern Europe, survival is lower than in the United States (26). This difference may be heavily influenced by a higher proportion of colorectal cancers that are adenocarcinomas in polyps diagnosed in the United States than in Europe (27). GROUPS AT INCREASED RISK OF COLORECTAL CANCER Several groups have an increased risk of developing colorectal cancer: those with inflammatory bowel disease or colorectal polyps, individuals in families affected by the autosomal dominant conditions hereditary nonpolyposis colorectal cancer (HNPCC) and familial adenomatous polyposis (FAP), and individuals who have a family history of colorectal neoplasia but are not part of families affected by HNPCC or FAP families. The risk of cancer in patients with longstanding ulcerative colitis or Crohn’s disease is hard to quantify, but is thought to be similar for patients

Epidemiology of Colorectal Cancer

47

with the two conditions (28). A meta-analysis of 116 studies estimated that the cumulative probability of cancer in a patient with ulcerative colitis was 2% by 10 years, 8% by 20 years, and 18% by 30 years (29). Studies have shown that adenomatous polyps left in situ progress from adenoma to cancer [reviewed in (30)]. These observations, coupled with indirect evidence, support the view that most colorectal carcinomas develop from adenomas. Although these lesions are usually removed when detected, the risk for recurrence three years after colonoscopic polypectomy is 30% to 40% (31,32). Investigation of factors related to occurrence and recurrence of polyps may provide information about the roles of exposures in the earlier stages of the adenoma–carcinoma sequence and, in turn, give clues as to likely routes for prevention of colorectal cancer. Adenoma recurrence is frequently used as a ‘‘model system’’ in intervention studies, the assumption being that an intervention that is effective in preventing recurrence in individuals with adenomas may also be effective in the prevention of colorectal cancer. Hyperplastic polyps also may exhibit malignant potential. These, and serrated adenomas, may be precursors of some right-sided colon cancers (33). Where pertinent, examples of studies of adenoma occurrence or recurrence are discussed later in this chapter. Fewer than 10% of incident colorectal cancers are due to HNPCC and FAP (34). Excluding these syndromes, carcinomas and adenomas aggregate in families. Individuals who have a first-degree relative with colorectal cancer have around a twofold increased risk of developing the disease themselves (35–37). This pattern is probably not entirely explained by familial clustering of environmental factors (38). This points to the potential importance of genetic susceptibility factors, and the interaction of these with each other and with environmental factors, in causing the disease. Genetic susceptibility is discussed further later in this chapter. RISK FACTORS FOR COLORECTAL NEOPLASIA The classic studies of Japanese migrants to the United States conducted in the 1960s revealed the overwhelming importance of environmental factors in colorectal cancer etiology (39) and the discussion below relates primarily to such factors. The evidence is summarized in Table 1. Physical Activity More than 40 case–control or cohort studies of physical activity and the risk of colorectal cancers have been carried out (40). These provide consistent evidence that physical activity is associated with a reduced risk of colon cancer, with relative risks for the highest category of activity compared with the lowest in the range 0.4 to 0.9 (41). The relationship has been observed in women as well as men, in various ethnic groups, and in diverse geographical

48

Little and Sharp

Table 1 Environmental Factors Associated with Colorectal Neoplasia Increasing risk Probable

Excess weight/BMI Alcohol

Possible

Tobacco smoking Insulin/hyperinsulinemia/ related factors

Reducing risk Physical activity Aspirin Hormone replacement therapy Vegetables Oral contraceptives Other NSAIDs Calcium Folate/folic acid

Abbreviations: BMI, body mass index; NSAIDs, nonsteroidal anti-inflammatory drugs.

areas. The association has been consistent in studies with widely different methods of assessing physical activity exposure, and persists after adjustment for other lifestyle factors. The data suggest that any activity is better than none (41) and that risk decreases in a dose–response fashion with increasing levels of activity (40). The volume of evidence specifically relating to cancer of the rectum is less substantial and suggests either a weak inverse association with higher levels of physical activity, or no association. The risk of adenomas is reduced among those reporting higher activity levels (42), and there is some suggestion that the relation may be stronger for adenomas with advanced features than for nonadvanced adenomas (43). Body Mass Index Excess weight raises the risk of developing colon cancer, with an increase of 15% in risk for an overweight person and 33% for an obese person (44). Similarly, the risk of adenomatous polyps is increased in individuals with a higher body mass index (45,46). There is little evidence of an association between weight and rectal cancer (42). Tobacco Smoking Tobacco smoking has consistently been found to be associated with an increased risk for adenomas and hyperplastic polyps (30,47–49). In the earlier studies of smoking and cancer, which mainly covered the 1950s and 1960s, there was no association between smoking and colorectal cancer, even among heavy smokers. In more recent studies, long-term smokers have been found to be at an elevated risk, with relative risks typically in the range of 1.5 to 3.0, following an induction period of 35 to 40 years (50). However, a recent review by the International Agency for Research on Cancer (IARC) concluded that it is possible that this association could be due to inadequate control of confounding (51).

Epidemiology of Colorectal Cancer

49

The aromatic amines, polycyclic aromatic hydrocarbons, and N-nitrosamines present in tobacco smoke are metabolized by a complex series of phase I and phase II activation and detoxification reactions. There is considerable interindividual variation in tobacco metabolism, and many of the genes controlling the production of the phase I and phase II enzymes are polymorphic. Several of these genes have been investigated in relation to colorectal neoplasia, with the most extensive evidence relating to the glutathionine-S-transferase genes GSTM1 and GSTT1, the cytochrome P450 1A1 gene CYP1A1, and the N-acteyltransferase genes NAT1 and NAT2. As regards GSTM1 and GSTT1, combined analysis of studies suggests that there is no association between the GSTM1 genotype alone and colorectal cancer (52–54), but that there may be a positive association with homozygosity for the GSTT1 deletion variant (52). However, for both polymorphisms, there is heterogeneity between studies, which is likely to be due in part to methodological differences (55), and in part to publication bias (56,57). On the basis of current evidence, it seems unlikely that either the GSTM1 or GSTT1 genotype strongly modifies the association between smoking and colorectal neoplasia (53,58–61), but most of the available studies have been hampered by a lack of statistical power to detect interactions. Studies of the CYP1A1 m1 and m2 polymorphisms and colorectal neoplasia have reported inconsistent results (62–69). In a large study, Slattery et al. found that presence of a CYP1A1 m1 or m2 variant allele modified the relationship between current smoking and colorectal cancer (69). No evidence of interaction between CYP1A1 genotype and smoking was found in two other studies, one of cancer and one of adenomas, but these were much smaller and so would have lacked statistical power to detect interactions (63,64). In a meta-analysis of 20 published case–control studies, the NAT2 genotype was not related to colon cancer risk (70). Similarly, pooled analysis of seven studies revealed no strong association between the NAT1 genotype and colorectal cancer (52). van der Hel et al. noted that cancer risk was raised in smokers who were imputed to be NAT2 rapid acetylators, compared to nonsmokers who were NAT2 slow acetylators (71). While this is compatible with the findings in a study of adenomas (72), other studies that have considered interactions between NAT1 or NAT2 and smoking have had inconsistent results (60,73–77). Exogenous Hormones The differences in the time trends in colorectal cancer in males and females (discussed earlier) could be explained by cohort effects in exposure to some sex-specific risk factor; one possibility that has been suggested is exposure to estrogens (19). There is, however, little evidence of an influence of endogenous hormones on the risk of colorectal cancer (78). In contrast, there is evidence that exogenous estrogens such as hormone replacement therapy (HRT), tamoxifen, or oral contraceptives might be associated with colorectal tumors.

50

Little and Sharp

In two large randomized controlled trials of the possible health benefits of HRT in postmenopausal women (79,80), the incidence of colorectal cancer was reduced by about one-third [relative risk (RR) in meta-analysis 0.64, 95% confidence interval (CI) 0.45–0.92] (81). These results were consistent with those of more than 20 case–control and cohort studies (82). When colon and rectal tumors have been considered separately, there was no evidence of an association between HRT and rectal cancer (83). Interpretation of the results of observational studies has not been straightforward. In meta-analyses, there was significant heterogeneity in the magnitude of the effect between studies (78,83–85). While the RR appear to be lower for current than for past HRT users and there is an attenuation of the risk several years after stopping hormone use (85), there is a lack of information on the effect by hormone type, dose, and duration of use. There has also been concern that unidentified confounding or the ‘‘healthy user effect’’ may have influenced the observed effect (78), but the observation of a similar effect in the two randomized controlled trials (RCTs) makes this a less likely explanation for the association. A potential issue of concern is that in the Women’s Health Initiative trial the colorectal tumors diagnosed in the group on estrogen plus progestin HRT were more advanced and had a greater number of positive lymph nodes than those that developed in women in the placebo arm (86). If confirmed, this would have important implications for the potential role for HRT in colon cancer prevention, in addition to the concerns about the increased risk of breast cancer, strokes, and pulmonary embolism (79,81). Regarding the use of the oral contraceptive pill, a meta-analysis of eight case–control studies and four cohort studies conducted up to 2000 was consistent with a moderate inverse association with risk of colorectal cancer (RR ¼ 0.82, 95% CI 0.70–0.97) (87). The relation was evident for both colon and rectal tumors. However, there was significant heterogeneity between the studies and risk did not decrease with increased duration of use. Although available data are sparse, the risk of colorectal cancer may also be increased among women taking tamoxifen therapy (85). Aspirin and Other Nonsteroidal Anti-inflammatory Drugs The relationship between colorectal cancer and aspirin use has been assessed in more than 20 observational studies. These consistently show that aspirin use is associated with a reduction in the risk for colorectal cancer of approximately 40% to 50% (88–95). Similar results have been reported for adenomas (88,96–101). There have also been four randomized controlled trials of aspirin in the prevention of colorectal neoplasia (97,102–105). Three of these found that aspirin, in doses between 81 and 325 mg/day, reduced risk of recurrence of adenomas (97,104,105). While the fourth trial failed to observe an effect, it had not been designed to evaluate colorectal neoplasia

Epidemiology of Colorectal Cancer

51

as endpoints and had limited statistical power (102,103). Despite this body of evidence, issues relating to the effective dose and duration of treatment that would be necessary for prevention are still unclear. These gaps in knowledge are particularly important because of the toxic effects of aspirin, particularly at high doses. As regards other types of nonsteroidal anti-inflammatory drugs (NSAIDs), three small randomized clinical trials have shown that sulindac reduces the number and size of colorectal polyps in patients with FAP, confirming the results of studies of nonrandomized case series (88,106). However, in one trial in patients who were genotypically affected with FAP but were phenotypically unaffected, sulindac did not prevent the development of colorectal adenomas (107). In a small randomized trial, no regression of small adenomatous polyps in patients without FAP was observed (108). Celecoxib and rofecoxib specifically inhibit cyclooxygenase-2 (COX-2). Celecoxib has been found to reduce the number of colorectal adenomas in patients with FAP (109). In an analysis of data from a prescription drug database in the elderly in Que´bec (Canada), there was an inverse association between colorectal adenomas and use of rofecoxib for a period of at least 90 days (110). In a secondary analysis relating to colorectal cancer, there were inverse associations with use of rofecoxib and celecoxib. It is likely that aspirin, or other NSAID, prophylaxis might be of benefit to particular subgroups of the population, but these groups have not yet been identified. However, there is intriguing evidence that genetic variation may modify the effect of NSAIDs on the development of colorectal neoplasia. Martinez et al. investigated the joint effects of aspirin use and a polymorphism in the ornithine decarboxylase gene (ODC) on the risk for recurrence of colorectal adenomas (111). Overall, both aspirin use and homozygosity for a G to A substitution in intron 1 of ODC were associated with a reduced risk of adenoma recurrence. The joint effect of aspirin use and homozygosity for the intron 1 variant was greater than would be expected on the basis of either an additive or multiplicative effect. Polymorphic genes encoding the two isoforms of prostaglandin H synthase [also known as cyclooxygenase (COX)], which are inhibited by NSAIDs, have also been investigated. Lin et al. reported that risks of both adenomas and colorectal cancers were associated with a rare COX-2 variant in AfricanAmericans (112). Cox et al., in an analysis of eight of the more frequent COX-2 polymorphisms in a study in Spain, observed that two variants in the untranslated region of exon 10 were associated with an increased risk of colorectal cancer (113). The protective effect of NSAIDs was not observed in those with the exon 10 variants, but this was based on small numbers. In a single study, in persons who carried either of two variants of COX-1, NSAID use was not associated with the decrease in adenoma risk observed in those without the variants (114). Other studies suggest interactions between aspirin and variants of the genes coding for interleukins IL6 (115)

52

Little and Sharp

and IL10 (116), the insulin receptor substrate 1 (IRS1) (117), the vitamin D receptor (VDR) (117), and the cyclin D1 gene (CCND1) (118). It will be important to determine whether these findings can be replicated (56,119), and whether they hold for different types of NSAIDs. Diet Diet has long been regarded as the most important environmental influence on colorectal cancer, and this is reflected in the volume of studies that have tested hypotheses about specific foods and nutrients. Virtually all of the studies have been observational and subject to three problems: (i) diet is related to other aspects of lifestyle, which may influence risk, (ii) people eat foods rather than nutrients, and (iii) misclassification of intake, both of the food group or nutrient being investigated, and of other food groups or nutrients that might confound the association could dilute or bias associations. In consequence, it has proved extremely difficult to identify the specific components of diet that influence risk. Vegetables and Fruit The comprehensive report of the World Cancer Research Fund (WCRF) and American Institute for Cancer Research (AICR) noted that of 21 case–control studies examining the association between vegetable and fruit consumption and colon cancer risk, 17 found some degree of reduced risk with higher consumption of at least one category of vegetable and fruit (120). Less consistent evidence was observed in the four cohort studies considered in the review. Of 10 case–control studies of rectal cancer in which statistical significance was reported, eight showed a significant inverse association with at least one category of vegetables and/or fruit, and the one cohort study in which rectal cancer risk was reported suggested an inverse relationship with consumption of green salad. On this basis, it was concluded that the evidence that diets rich in vegetables protect against cancers of the colon and rectum was convincing, however, no judgment was possible regarding the relationship with fruit. Recent evidence suggests the relation between vegetables and fruit and colorectal neoplasia is complex (121–123). For example, the association is much stronger in case–control than cohort studies (124), and case–control studies are potentially more susceptible to bias than cohort studies. Meat There have been two meta-analyses of meat consumption reported in the last few years, one based on cohort studies only (125), the other based on both case–control and cohort studies (126). The association between total meat consumption and colorectal cancer is inconsistent, and both meta-analyses show no statistically significant overall association. There is considerable

Epidemiology of Colorectal Cancer

53

heterogeneity between the results of case–control studies (126). In the cohort studies in which a positive association was found, the possibility that confounding factors might account for the results could not be excluded (125). Data on red meat and processed meat suggest positive associations with the risk for colorectal cancer. However, the volume of evidence on these is substantially less than for total meat consumption, and it is possible that publication bias has favored positive results. Heterocyclic amines are generated during the cooking of red meat at high temperatures, and increased consumption of well-done red meat has been associated with increased risk of colorectal neoplasia in some studies (127,128). For the heterocylic amines to be carcinogenic they must be metabolized by enzymes including glutathione-S-transferase (GST), N-acetyltransferase 1 (NAT1), and N-acetyltransferase 2 (NAT2). This has prompted investigation of interactions between variants in phase I and phase II metabolism genes and meat intake with regard to risk of colorectal neoplasia. Ishibe et al. observed a sixfold increased risk of adenomas among rapid NAT1 acetylators (defined as those carrying the NAT1
10 allele) who were estimated, on the basis of reported meat intake, cooking methods and doneness level, to consume more than 27 ng/day of the heterocyclic amine MeIQx, whereas among slow acetylators the increase in risk was twofold (129). While other investigators have also reported patterns in risk suggestive of interactions between particular genetic variants and meat intake [e.g., Welfare et al. for NAT2; Gertig et al. for GSTM1 and GSTT1; Turner et al. for GSTT1 and GSTP1; Cortessis et al. for microsomal epoxide hydrolase (mEH) (59,73,130,131)], the direction of the associations have not always been consistent with the underlying hypotheses (55). Other studies have failed to find any evidence that the relationship between red meat intake and colorectal neoplasia is modified by genotype (76,128, 132,133). In addition to differences between studies in the genes and polymorphisms that have been investigated, and different approaches to statistical analysis and low statistical power, the difficulty of adequately assessing exposure to carcinogens in cooked red meats further complicates this area of research. Fat and Fiber The contrast between low colorectal cancer rates in sub-Saharan Africa and high rates in industrialized countries was the basis for the suggestion that diet, in particular one with high levels of fat and low levels of fiber, might have a key role in causing the disease. The results of epidemiological studies on macronutrients (fat, proteins, and carbohydrates) have been less consistent in establishing an associated risk of cancer than those on food groups (134). Although the hypothesis that high fat intake is a major risk factor of the diet of industrialized countries has been investigated in many epidemiological and laboratory studies, no

54

Little and Sharp

clear relationship has been established with colorectal cancer (135). There is now increased emphasis on the effects of specific fatty acids (136). For example, in a single study, an intake of n-6 fatty acids above the median was associated with an increased risk of colon cancer in those who carried a variant of the promoter region of the COX-2 gene, but not in those who did not have the variant (137). As regards dietary fiber, in a large multicenter cohort study in Europe, a 40% reduction in risk of colorectal cancer among those with the highest dietary fiber intake was observed (138). In contrast, in large cohort studies in the United States and Finland, no association has been found (122, 139,140). Intervention studies examining the effect of bran and soluble fiber have not found any effect on adenoma recurrence, nor have trials of dietary modification to increase fiber and lower fat intake (141–143). The evaluation of the cancer–fiber relation is particularly challenging due to the varying composition of fiber from different sources and variations in assessment of intake (138,144). It has been suggested that higher levels of fat and lower levels of fiber might increase colorectal cancer risk by altering fecal characteristics. In particular, it has been postulated that development of colorectal neoplasia may be promoted either by a high fecal total bile acid concentration (145) or by an abnormal bile acid profile with a high ratio of lithocholic to deoxycholic acid (146–148). However, no consistent association between colorectal cancer and fecal bile acid concentrations has been observed (149–156). This inconsistency may be due in part to methodological factors such as selection bias and limited statistical power. It is also possible that fecal bile acid levels may have been affected by the presence of the tumor, either directly or indirectly; for example, as a result of changes in diet made because of symptoms or treatment. For this reason, colorectal adenomatous polyps have been investigated in some studies, however, the results have been inconclusive (150,152,157–159). Most of the studies have been based on small numbers of cases who have been ascertained as a result of symptoms. The factors causing the symptoms may have affected fecal constituents, including bile acids. In a study of asymptomatic subjects who had participated in FOB screening, no association between colorectal cancer and fecal bile acids was observed (160). Folate Vegetables, particularly green leafy vegetables, are a major source of folate. Folate is involved in the synthesis and methylation of DNA, and mechanisms have been postulated by which low folate status might increase the risk of malignancy [reviewed in (161)]. This has prompted considerable investigation of the role of folate, and its synthetic form folic acid in colorectal neoplasia. The majority of observational studies—either measuring blood folate or assessing intake—are compatible with an inverse association between

Epidemiology of Colorectal Cancer

55

folate level and risk of colon cancer and adenomas. Two of three prospective studies found an increased risk for colorectal cancer among people with reduced levels of serum or plasma folate (162–164), a short-term marker of folate intake. Two studies have reported an inverse association between red cell folate, a measure of folate status over a three- to four-month period, and adenoma risk (165,166). Almost all prospective studies of folate intake show an inverse association with risk of colon cancer (or colorectal cancer, where colon and rectal tumors have not been analyzed separately), with several reporting a dose–response relationship (162,164,167–173). While evidence from case–control studies is not as consistent, most have found at least a modest reduced risk of colon (or colorectal) cancer associated with higher intake, at least among subgroups (174–184). Moreover, use of dietary supplements containing folic acid has been related to lower risk of colon cancer in several studies and the association appears, albeit on limited evidence, to be stronger for longer periods of regular use, or for use of higher dose supplements (164,167,168,183,185,186). There appears to be no consistent association between folate and rectum cancer (162,170,172–175, 177,178,184). As regards adenomas, both cohort and case–control studies have reported risk of adenoma occurrence to be reduced among those with higher folate intake (165,178,187–192), but it is not currently clear whether dietary folate intake is associated with adenoma recurrence (193). Alcohol adversely affects the metabolism of folate (194), which has prompted interest in whether a composite dietary profile of lower folate and higher alcohol intake, together with low intakes of methionine and vitamins B6 and B12 (a ‘‘low methyl’’ diet) may be associated with colorectal neoplasia. Several studies suggest that persons with a low-methyl diet do indeed have higher risk for colon cancer than those with a high-methyl diet (162,167,169,172,180,184). Family history may impact on the folate–colon cancer relationship. Fuchs et al. found that a higher level of total folate intake had only a minimal protective effect on colon cancer risk among women without a family history of colorectal cancer in first-degree relatives but was associated with a substantial reduction in risk among women with a family history of the disease (195). Consistent with this, Slattery et al. observed a fivefold increased risk of colon cancer for a low-methyl diet among women who reported a first-degree family history of colorectal cancer, compared to a 1.5-fold risk among those without a family history (180). Many of the genes involved in the absorption, metabolism, and transport of folate contain common genetic variants (196). Several studies have investigated two common variants of the gene encoding 5,10-methylenetetrahydrofolate reductase (MTHFR), C677T and A1298C, in relation to colorectal neoplasia. In most studies, these variants are associated with moderately reduced colorectal cancer risk (197–201). Findings from six studies of C677T and adenomatous polyps are inconsistent (192,197,202).

56

Little and Sharp

In studies in which joint effects of MTHFR genotype and diet have been investigated, those homozygous for the C677T variant who had higher folate levels (or a high-methyl diet) had the lowest cancer risk (197). As yet, too few investigations on other polymorphisms in the folate pathway—such as variants of the gene coding for methionine synthase (MTR) (182,190,203,204) methionine synthase reductase (MTRR) (182), cystathione b-synthase (CBS) (182,205), or thymidylate synthase (TS) (206–210)—have been carried out to be conclusive. While the metabolism of any exposure is likely to depend on the balance between the relative activities of all the enzymes active within the metabolic pathway (211), to date joint effects of folate-pathway genes have only been little investigated (182,207). Carotenoids No consistent association between dietary carotenoids, or serum or plasma concentrations of beta-carotene, and colorectal cancer has been observed (212). None of the trials of beta-carotene supplementation suggests a decrease in the occurrence of colorectal cancer, and two randomized control trials provide evidence of a lack of efficacy of short-term supplementation of beta-carotene in preventing occurrence of colorectal adenomas (212). Calcium Several observational studies and three intervention trials have found a reduced risk of occurrence and recurrence of colorectal neoplasia associated with higher calcium intake, either from the diet or as supplements, but not all of the studies reached statistical significance (120,143,213–218). In a pooled analysis of 10 cohort studies of colorectal cancer, the relative risk for the highest versus lowest quintile of dietary intake was 0.86 (95% CI 0.78–0.95, p trend¼0.02); for total intake, combining dietary and supplemental sources, the relative risk was 0.78 (95% CI 0.69–0.88, p trend < 0.001) (219). It has been postulated that fecal calcium may protect against colorectal carcinogenesis (220), because calcium ions in the colon would also precipitate the bile acids as their calcium salts and so would modulate their toxicity (221). In a study of subjects participating in FOB screening, high levels of fecal calcium were associated with a reduced risk of both colorectal cancer and colorectal adenomas, but these associations were not statistically significant (160). Calcium homeostasis is maintained by vitamin D, in that the vitamin D metabolite 1–25(OH)2D3 mediates intestinal calcium absorption. Vitamin D mediates its effect through the vitamin D receptor (VDR). This has led to investigation of associations between polymorphisms in the VDR gene and colorectal neoplasia. The FokI polymorphism has been associated with risk of large adenomas and of colorectal cancer (222,223), but the direction of the relationship differed for the two types of neoplasm. Moreover, two

Epidemiology of Colorectal Cancer

57

other studies failed to find similar associations (224,225). However, the study of Slattery et al. did observe a relation between colon cancer and three other VDR polymorphisms (which are in linkage disequilibrium) (224). There is some evidence that the VDR genotype might modify the association between calcium intake and risk; three studies of adenomas and one of colorectal cancer found patterns consistent with an interaction (222,223,225,226). Alcohol In the WCRF/AICR report, an association was noted between colon cancer and alcohol intake in four out of five general population cohort studies, in three out of three cohort studies on rectal cancer, and two out of three cohort studies that did not distinguish between colon and rectal cancer (120). In 9 out of 18 case–control studies of colon cancer and 9 out of 17 case–control studies of rectal cancer, there was a positive association with alcohol intake. In a meta-analysis of studies published in the period 1966–1998 there was significant heterogeneity in the colon cancer–alcohol relationship between the cohort and case–control studies included (227). For the studies of rectal cancer, there was significant heterogeneity by study quality and gender. In a pooled analysis of eight cohort studies in five countries in North America and Europe, a small increase in risk (RR 1.23, 95% CI 1.07–1.42) of colorectal cancer associated with a reported intake of 30 g/day or more was observed (228). Alcohol is metabolized to the carcinogen acetaldehyde by oxidation by the enzyme alcohol dehydrogenase (ADH) and is subsequently detoxified into acetate by aldehyde dehydrogenase (ALDH). The ADH isoenzymes involved in these reactions include subunits encoded by the ADH3 gene, which is polymorphic. Two studies of adenomas have reported patterns consistent with an interaction between ADH3 genotype and alcohol intake (202,229). Among subjects in the male Health Professional Follow-up Study (HPFS), high consumers of alcohol with the slow catabolism genotype (
2/
2) had a substantially increased risk of disease [odds ratio (OR) > 30 g/day and
2/
2 vs. 5 g/day and
1/
1 ¼ 2.94, 95% CI 1.24–6.92] compared to those who consumed low levels of alcohol per day and carried the fast alcohol catabolism genotype (ADH3
1/
1). Those who consumed high quantities of alcohol but had the fast catabolism genotype had only minimally increased risk (OR > 30 g/day and
1/
1 vs.  5 g/day and
1/
1 ¼ 1.27, 95% CI 0.63–2.53) (202). The pattern of interaction described in the other study, from the Netherlands (229), was very similar to the HPFS result, and the relationship was apparent in both male and female subjects. Because of the link between alcohol and folate metabolism, Giovannucci et al. investigated, in the HPFS, whether ADH3 acted together with alcohol and folate intake to influence disease risk. Individuals with high alcohol and low folate and the slow catabolism genotype were at particularly high risk compared to fast catabolizers with low alcohol and high folate

58

Little and Sharp

intake (OR ¼ 17.1, 95% CI 2.13–137.0: p interaction ¼ 0.006), although the result was based on small numbers in the high alcohol/low folate/slow catabolism group (202). This study, along with three others of cancer and one of adenomas, has explored interactions between alcohol intake and the MTHFR genotype (192,199–202). Giovannucci et al. described a borderline significant interaction where the presence of the 677 TT genotype did not affect adenoma risk among persons consuming low amounts of alcohol (OR TT and 5 g/day vs. CC/CT and 5 g/day ¼ 0.79, 95% CI 0.42–1.49), but was associated with increased risk among those with a high alcohol intake (OR TT and >30 g/day vs. CC/CT and 5 g/day ¼ 3.52, 95% CI 1.41–8.78: p interaction ¼ 0.009) (202). Yin et al., in a study of 685 colorectal cancer cases and 778 controls in Japan, observed a similar pattern of risk for the A1298C polymorphism, but not for C677T (201). The other studies of MTHFR, alcohol, and colorectal neoplasia, all of which were smaller than the HPFS and the study of Yin et al., had inconsistent results (192,199,200). Insulin, Hyperinsulinemia, and Insulin-Like Growth Factors The similarity of risk factors for colon cancer and diabetes, and the observation that insulin promotes the growth of colon cells in vitro and colon tumors in vivo (230,231), prompted suggestions that hyperinsulinemia and insulin resistance may lead to colorectal cancer through growth-promoting effects of elevated levels of insulin, glucose, or triglycerides (232,233). While several strands of epidemiological evidence support the hypothesis, inconsistencies remain and a number of areas require clarification. Moderately increased risks of colorectal cancer and adenomas have been associated with type 2 diabetes (234–238), although the studies are not entirely consistent (239). Individuals with several risk factors consistent with insulin resistance syndrome (e.g., high systolic blood pressure, high BMI, etc.) were found to have an increased risk of death from colorectal cancer in two studies (240,241). Hyperglycemia has been associated with risk; higher fasting and nonfasting blood glucose levels are associated with an increased risk of colorectal cancer (incidence and mortality), carcinoma in situ, and adenomas (236,240–244). Two prospective studies observed a modest relationship between plasma insulin levels and colorectal cancer incidence, but a third study was negative (243,245,246). In two prospective studies from the United States, an increased concentration of plasma C-peptide, an indicator of insulin secretion, was associated with a significantly raised colorectal cancer risk (247,248). Two large studies have found an approximately two- to threefold increased risk of colorectal cancer associated with being in the highest, compared to the lowest, quintile of dietary glycemic load (249,250). A single study of adenomas, however, found no evidence that glycemic load or glycemic index of the diet were related to risk (251).

Epidemiology of Colorectal Cancer

59

One mechanism by which raised insulin levels could affect cancer risk is by increasing the bioactivity of insulin-like growth factor-1 (IGF-1) and inhibiting production of two main binding proteins, IGFBP-1 and IGFBP-2 (252). IGF-1 has mitogenic effects on normal and neoplastic cells, inhibiting apoptosis and stimulating cell proliferation (252). Three prospective studies of colorectal cancer have observed a greater than twofold increased risk among those in the highest quantile of IGF-1, compared with those in the lowest (247,253,254). A further prospective study reported a positive relationship with colon cancer (OR highest vs. lowest quantile ¼ 2.66; p trend ¼ 0.03) and a negative one for rectal cancer (OR ¼ 0.33; p trend ¼ 0.09), although the result for rectal cancer did not reach statistical significance (255). Risk of intermediate/late-stage adenomas, but not early stage adenomas, has also been found to be positively related to IGF-1 levels (254). One prospective study observed an inverse relationship between IGFBP-1 and IGFBP-2 and colorectal cancer (247), but two others have been null (245,246). A genetic variant at position 1663 in the human growth hormone-1 gene (GH1) is thought to be associated with lower IGF-1 levels. In a single study, the variant A allele was related, in a dose–response fashion, to a reduced risk of both colorectal cancer and adenomas (256). Also in a single study, polymorphisms in the genes encoding the insulin receptor substrates (IRS-1, IRS-2) were associated with risk of colon, but not rectal, cancer (257). In the same study, variants in the IGF-1 and IGFBP3 genes were not independently related to cancer but did appear to act together with IRS-1 to influence risk. Although requiring confirmation, the findings suggest that combinations of polymorphisms in the insulin-related signaling pathway may be important in colon cancer etiology. CONCLUSION Colorectal cancer continues to pose a major public health problem, with almost a million new cases being diagnosed each year worldwide, and over half a million deaths. The numbers are likely to increase as a result of population aging and increased life expectancy, especially in developing countries. Evidence that physical activity, a lower body mass index, use of aspirin and other NSAIDs, a higher intake of vegetables, and use of exogenous hormones in women are associated with decreased risk strongly suggest that there is considerable potential for primary prevention through lifestyle modification and, possibly, chemoprevention. While there remain challenges with implementation of lifestyle modification, and in developing methods of chemoprevention in either the general population or high-risk groups that maximize benefits and minimize harms, it is important that the potential power of primary prevention is not overlooked in developing strategies and guidelines for control of the disease, which tend to focus on treatment and screening.

60

Little and Sharp

An emerging theme is the investigation of associations with genetic polymorphisms, and interactions between these and established or putative risk factors. This is a challenging area of investigation. In many of the studies of gene–environment interaction and colorectal neoplasia that have been conducted to date, there has been limited statistical power to detect interaction. The methods used to test for the same putative interaction have differed between studies, making it difficult to integrate evidence across studies. It is important that evidence in these areas is synthesized and efforts made to minimize the likelihood of publication bias. Collaborative networks such as the Human Genome Epidemiology Network (258) should facilitate this.

REFERENCES 1. Ferlay J, Bray F, Pisani P, Parkin DM. Globocan 2000: Cancer Incidence, Mortality and Prevalence worldwide. Lyon: International Agency for Research on Cancer, 2001. 2. Parkin DM, Pisani P, Ferlay J. Estimates of the worldwide incidence of 25 major cancers in 1990. Int J Cancer 1999; 80:827–841. 3. Parkin DM, Whelan SL, Ferlay J, Teppo L, Thomas DB. (eds) Cancer Incidence in Five Continents. Vol. VIII. Lyon: International Agency for Research on Cancer, 2002. 4. Haenszel W, Correa P. Cancer of the colon and rectum and adenomatous polyps: a review of epidemiologic findings. Cancer 1971; 28:14–24. 5. Sharp L. Current trends in colorectal cancer: what they tell us and what we still do not know. Clin Oncol 2001; 13:444–447. 6. Thorn M, Bergstrom R, Kressner U, Sparen P, Zack M, Ekbom A. Trends in colorectal cancer incidence in Sweden 1959–93 by gender, localization, time period, and birth cohort. Cancer Causes Control 1998; 9:145–152. 7. Johansen C, Mellemgaard A, Skov T, Kjaergaard J, Lynge E. Colorectal cancer in Denmark 1943–1988. Int J Colorectal Dis 1993; 8:42–47. 8. Lopez-Abente G, Pollan M, Vergara A, et al. Age-period-cohort modeling of colorectal cancer incidence and mortality in Spain. Cancer Epidemiol Biomarkers Prev 2001; 6:999–1005. 9. De Angelis R, Valente F, Frova L, et al. Trends of colorectal cancer incidence and prevalence in Italian regions. Tumori 1998; 84:1–8. 10. Bell JC. Trends in colorectal cancer incidence and mortality in New South Wales 1973–1992. Med J Aust 1996; 166:175–176. 11. Cox B, Little J. Reduced risk of colorectal cancer among recent generations in New Zealand. Br J Cancer 1992; 66:386–390. 12. Hayne D, Brown RSD, McCormack M, Quinn MJ, Payne HA, Babb P. Current trends in colorectal cancer: site, incidence, mortality and survival in England and Wales. Clin Oncol 2001; 13:448–452. 13. Black RJ, Macfarlane GJ, Maisonneuve P, Boyle P. Cancer Incidence and Mortality in Scotland, 1960–1989. Edinburgh: Common Services Agency, 1995.

Epidemiology of Colorectal Cancer

61

14. Harris V, Sandridge AL, Black RJ, Brewster DH, Gould A. Cancer Registration Statistics Scotland, 1986–1995. Edinburgh: ISD Scotland Publications, 1998. 15. Reis LAG, Wingo PA, Miller DS, et al. The annual report to the nation on the status of cancer, 1973–1997, with a special section on colorectal cancer. Cancer 2000; 88:2398–2424. 16. Parkin DM, Whelan SL, Ferlay J, Raymond L, Young J. (eds) Cancer Incidence in Five Continents. Volume VII. Lyon: International Agency for Research on Cancer, 1997. 17. Jass JR. Subsite distribution and incidence of colorectal cancer in New Zealand 1974–1983. Dis Colon Rectum 1991; 34:56–59. 18. Devesa SS, Chow WH. Variation in colorectal cancer incidence in the United States by subsite of origin. Cancer 1993; 71:3819–3826. 19. dos Santos Silva I, Swerdlow AJ. Sex differences in time trends of colorectal cancer in England and Wales: the possible effect of female hormonal factors. Br J Cancer 1996; 73:692–697. 20. Rabeneck L, Davila JA, El-Serag HB. Is there a true ‘‘shift’’ to the right colon in the incidence of colorectal cancer?. Am J Gastroenterol 2003; 98:1400–1409. 21. Grulich AE, Swerdlow AJ, Head J, Marmot MG. Cancer mortality in African and Caribbean migrants to England and Wales. Br J Cancer 1992; 66:905–911. 22. Faggiano F, Partanen T, Kogevinas M, Boffetta P. Socioeconomic differences in cancer incidence and mortality. In: Kogevinas M, Pearce N, Susser M, Boffetta P, eds. Social Inequalities and Cancer. IARC Scientific Publications No. 138. Lyon: International Agency for Research on Cancer, 1997:65–176. 23. Coleman MP, Esteve J, Damiecki P, Arslan A, Renard H. Trends in Cancer Incidence and Mortality. Lyon: International Agency for Research on Cancer, 1993. 24. Robinson MHE, Thomas WM, Hardcastle JD, Chamberlain J, Mangham CM. Change towards earlier stage at presentation of colorectal cancer. Br J Surg 2002; 80:1610–1612. 25. Chu KC, Tarone RE, Chow WH, Hankey BF, Ries LAG. Temporal patterns in colorectal cancer incidence, survival, and mortality from 1950 through 1990. J Natl Cancer Inst 2002; 86:997–1006. 26. Sant M, Capocaccia R, Coleman MP, et al., EUROCARE Working Group. Cancer survival increases in Europe, but international differences remain wide. Eur J Cancer 2001; 37:1659–1667. 27. Gatta G, Ciccolallo L, Capocaccia R, et al. Differences in colorectal cancer survival between European and US populations: the importance of sub-site and morphology. Eur J Cancer 2003; 39:2214–2222. 28. Gillen C, Walmsley RS, Prior P, Andrews HA, Allan RN. Ulcerative colitis and Crohn’s disease: a comparison of the colorectal cancer risk in extensive colitis. Gut 1994; 35:1590–1592. 29. Eaden JA, Abrams KR, Mayberry JF. The risk of colorectal cancer in ulcerative colitis: a meta-analysis. Gut 2001; 48:526–535. 30. Cotton S, Sharp L, Little J. The adenoma-carcinoma sequence and prospects for the prevention of colorectal neoplasia. Crit Rev Oncogen 1996; 7(5&6):293–342. 31. Winawer SJ, O’Brien MJ, Waye JD, et al. Risk and surveillance of individuals with colorectal polyps. Bull World Health Organ 1990; 68:789–795.

62

Little and Sharp

32. Peipins LA, Sandler RS. Epidemiology of colorectal adenomas. Epidemiol Rev 1994; 16:273–297. 33. Hawkins NJ, Ward RL. Sporadic colorectal cancers with microsatellite instability and their possible origin in hyperplastic polpys and serrated adenomas. J Natl Cancer Inst 2001; 93:1307–1313. 34. Mecklin JP, Ponz de Leon M. Epidemiology of HNPCC. Anticancer Res 1994; 14:1625–1629. 35. Fuchs CS, Giovannucci EL, Colditz GA, Hunter DJ, Speizer FE, Willett WC. A prospective study of family history and the risk of colorectal cancer. N Engl J Med 1994; 331:1669–1674. 36. Winawer SJ, Zauber AG, Gerdes H, et al. Risk of colorectal cancer in the families of patients with adenomatous polyps. N Engl J Med 1996; 334:82–87. 37. Johns LE, Houlston RS. A systematic review and meta-analysis of familial colorectal cancer risk. Am J Gastroenterol 2001; 96:2992–3003. 38. Khoury MJ, Beaty TH, Liang KY. Can familial aggregation of disease be explained by familial aggregation of environmental risk factors? Am J Epidemiol 1988; 127:674–683. 39. Haenszel W, Kurihara M. Studies of Japanese migrants: mortality from cancer and other diseases among Japanese in the United States. J Natl Cancer Inst 1968; 40:43–68. 40. Friedenreich CM. Physical activity and cancer prevention: from observational to intervention research. Cancer Epidemiol Biomarkers Prev 2001; 10: 287–301. 41. McTiernan A, Ulrich C, Slate S, Potter J. Physical activity and cancer etiology: associations and mechanisms. Cancer Causes Control 1998; 9:487–509. 42. IARC Working Group. IARC Handbooks of Cancer Prevention Vol. 6. In: The Role of Weight Control and Physical Activity in Cancer Prevention. Lyon: International Agency for Research on Cancer, 2002. 43. Terry MB, Neugut AI, Bostick RM, et al. Risk factors for advanced colorectal adenomas: a pooled analysis. Cancer Epidemiol Biomarkers Prev 2002; 11:622–629. 44. Bergstrom A, Pisani P, Tenet V, Wolk A, Adami HO. Overweight as an avoidable cause of cancer in Europe. Int J Cancer 2001; 91:421–430. 45. Bird CL, Frankl HD, Lee ER, Haile RW. Obesity, weight gain, large weight changes, and adenomatous polyps of the left colon and rectum. Am J Epidemiol 1998; 147:670–680. 46. Boutron-Ruault MC, Senesse P, Belghiti C, Faivre J. Energy intake, body mass index, physical activity, and the colorectal adenoma-carcinoma sequence. Nutr Cancer 2001; 39:50–57. 47. Martinez ME, McPherson RS, Levin B, Glober GA. A case–control study of dietary intake and other lifestyle risk factors for hyperplastic polyps. Gastroenetrology 1997; 113:423–429. 48. Morimoto LM, Newcomb PA, Ulrich CM, Bostick RM, Lais CJ, Potter JD. Risk factors for hyperplastic and adenomatous polyps: evidence for malignant potential? Cancer Epidemiol Biomarkers Prev 2002; 11:1012–1018. 49. Lieberman DA, Prindiville S, Weiss DG, Willet W. Risk factors for advanced colonic neoplasia and hyperplastic polyps in asymptomatic individuals. JAMA 2003; 290:2959–2967.

Epidemiology of Colorectal Cancer

63

50. Giovannucci E. An updated review of the epidemiological evidence that cigarette smoking increases risk of colorectal cancer. Cancer Epidemiol Biomarkers Prev 2001; 10:725–731. 51. IARC Working Group. IARC Monograph on the Evaluation of Carcinogenic Risks to Humans. Volume 83: Tobacco Smoke and Involuntary Smoking. Lyon: International Agency for Research on Cancer, 2004. 52. de Jong MM, Nolte IM, te Meerman GJ, et al. Low-penetrance genes and their involvement in colorectal cancer susceptibility. Cancer Epidemiol Biomarkers Prev 2002; 11:1332–1352. 53. Smits KM, Gaspari L, Weijenberg MP, et al. Interaction between smoking, GSTM1 deletion and colorectal cancer: results from the GSEC study. Biomarkers 2003; 8:299–310. 54. Ye Z, Parry JM. A meta-analysis of 20 case–control studies of the glutathione transferase M1 (GSTM1) status and colorectal cancer risk. Med Sci Monit 2003; 9:SR83–SR91. 55. Cotton SC, Sharp L, Little J, Brockton N. Glutathione S-transferase polymorphisms and colorectal cancer. Am J Epidemiol 2000; 151:7–32. 56. Ioannidis JPA, Ntzani EE, Trikalinos TA, Contopoulos-Ioannidis DG. Replication validity of genetic association studies. Nat Genet 2001; 29:306–309. 57. Hirschhorn JN, Lohmueller K, Byrne E, Hirschhorn K. A comprehesive review of genetic association studies. Genet Med 2002; 4:45–61. 58. Katoh T, Nagata N, Kuroda Y, et al. Glutathione S-transferase M1 (GSTM1) and T1 (GSTT1) genetic polymorphism and susceptibility to gastric and colorectal adenocarcinoma. Carcinogenesis 1996; 17:1855–1859. 59. Gertig DM, Stampfer M, Haiman CH, Hennekens CH, Kelsey K, Hunter DJ. Glutathione S-Transferase GSTM1 and GSTT1 polymorphisms and colorectal cancer risk: a prospective study. Cancer Epidemiol Biomarkers Prev 1998; 7:1001–1005. 60. Slattery ML, Potter JD, Samowitz W, Bigler J, Caan B, Leppert M. NAT2, GSTM-1, cigarette smoking, and risk of colon cancer. Cancer Epidemiol Biomarkers Prev 1998; 7:1079–1084. 61. Slattery ML, Edwards S, Curtin K, Schaffer D, Neuhausen S. Associations between smoking, passive smoking, GSTM-1, NAT2, and rectal cancer. Cancer Epidemiol Biomarkers Prev 2003; 12:882–889. 62. Sivaraman L, Leatham MP, Yee J, Wilkens LR, Lau AF, Le Marchand L. CYP1A1 genetic polymorphisms and in situ colorectal cancer. Cancer Res 1994; 54:3692–3695. 63. Inoue H, Kiyohara C, Marugame T, et al. Cigarette smoking, CYP1A1 MspI and GSTM1 genotypes, and colorectal adenomas. Cancer Res 2000; 60:3749–3752. 64. Ishibe N, Stampfer M, Hunter DJ, Hennekens C, Kelsey KT. A prospective study of cytochrome P450 1A1 polymorphisms and colorectal cancer risk in men. Cancer Epidemiol Biomarkers Prev 2000; 9:855–856. 65. Kiss I, Sandor J, Pajkos G, Bogner B, Hegedus G, Ember I. Colorectal cancer risk in relation to genetic polymorphism of cytochrome P450 1A1, 2E1, and glutathione-S-tranferase M1 enzymes. Anticancer Res 2000; 20:519–522. 66. Butler WJ, Ryan P, Roberts-Thomson IC. Metabolic genotypes and risk for colorectal cancer. J Gastroenterol Hepatol 2001; 16:631–635.

64

Little and Sharp

67. Sachse C, Smith G, Wilkie MJV, et al. The Colorectal Study Group. A pharmacogenetic study to investigate the role of dietary carcinogens in the etiology of colorectal cancer. Carcinogenesis 2002; 23:1839–1849. 68. Ye Z, Parry JM. Genetic polymorphisms in the cytochrome P450 1A1, glutathione S-transferase M1 and T1, and susceptibility to colon cancer. Teratogenesis Carcinog Mutagen 2002; 22:385–392. 69. Slattery ML, Samowtiz W, Ma K, et al. CYP1A1, cigarette smoking, and colon and rectal cancer. Am J Epidemiol 2004; 160:842–852. 70. Ye Z, Parry JM. Meta-analysis of 20 case–control studies on the N-acetyltransferase 2 acetylation status and colorectal cancer risk. Med Sci Monit 2002; 8:CR558–CR565. 71. van der Hel OL, Bueno de Mesquita B, Sandkuijl L, et al. Rapid N-acetyltransferase 2 imputed phenotype and smoking may increase risk of colorectal cancer in women (Netherlands). Cancer Causes Control 2003; 14:293–298. 72. Probst-Hensch NM, Haile RW, Ingles SA, et al. Acetylation polymorphism and prevalence of colorectal adenomas. Cancer Res 1995; 55:2017–2020. 73. Welfare MR, Cooper J, Bassendine MF, Daly AK. Relationship between acetylator status, smoking, diet and colorectal cancer risk in the north-east of England. Carcinogenesis 1997; 18:1351–1354. 74. Potter JD, Bigler J, Fosdick L, et al. Colorectal adenomatous and hyperplastic polyps: smoking and N-acetyltransferase 2 polymorphisms. Cancer Epidemiol Biomarkers Prev 1999; 8:69–75. 75. Probst-Hensch NM, Haile RW, Li DS, et al. Lack of association between the polyadenylation polymorphism in the NAT1 (acetyltransferase 1) gene and colorectal adenomas. Carcinogenesis 1996; 17:2125–2129. 76. Barrett JH, Smith G, Waxman R, et al. Investigation of intercation between N-acetyltransferase 2 and heterocylic amines as potential risk factors for colorectal cancer. Carcinogenesis 2003; 24:275–282. 77. van der Hel OL, Bueno-de-Mesquita HB, Roest M, et al. No modifying effect of NAT1, GSTM1, and GSTT1 on the relation between smoking and colorectal cancer risk. Cancer Epidemiol Biomarkers Prev 2003; 12:681–682. 78. Beral V, Banks E, Reeves G, Appleby P. Use of HRT and the subsequent risk of cancer. J Epidemiol Biostat 1999; 4:191–215. 79. Writing group for the Women’s Health Initiative Investigators. Risks and benefits of estrogen plus progestin in healthy postmenopausal women. Principal results from the Women’s Health Initiative Randomized Controlled Trial. JAMA 2002; 288:321–333. 80. Hulley S, Furberg C, Barrett-Connor E, et al., HERS Research Group. Noncardiovascular disease outcomes during 6.8 years of hormone therapy: Heart and Estrogen/progestin Replacement Study follow-up (HERS II). JAMA 2002; 288:58–66. 81. Beral V, Banks E, Reeves G. Evidence from randomised trials on the long-term effects of hormone replacement therapy. Lancet 2002; 360:942–944. 82. Crandall CJ. Estrogen replacement therapy and colon cancer: a clinical review. J Womens Hlth Gender-based Med 1999; 8:1155–1166.

Epidemiology of Colorectal Cancer

65

83. Nanda K, Bastian LA, Hasselblad V, Simel DL. Hormone replacement therapy and the risk of colorectal cancer: a meta-analysis. Obstet Gynecol 1999; 93: 880–888. 84. Hebert-Croteau N. A meta-analysis of hormone replacement therapy and colon cancer in women. Cancer Epidemiol Biomarkers Prev 1998; 7: 653–659. 85. Grodstein F, Newcomb PA, Stampfer MJ. Postmenopausal hormone therapy and the risk of colorectal cancer: a review and meta-analysis. Am J Med 1999; 106:574–582. 86. Chlebowski RT, Wactawski-Wende J, Ritenbaugh C, et al. Estrogen plus progestin and colorectal cancer in postmenopausal women. N Engl J Med 2004; 350:991–1004. 87. Fernandez E, La Vecchia C, Balducci A, Chatenoud L, Franceschi S, Negri E. Oral contraceptives and colorectal cancer risk: a meta-analysis. Br J Cancer 2001; 84:722–727. 88. IARC Working Group. Non-Steroidal Anti-Inflammatory Drugs. Vol. 1. Lyon: International Agency on Research of Cancer, 1997. 89. Rosenberg L, Louik C, Shapiro S. Nonsteroidal antiinflammatory drug use and reduced risk of large bowel carcinoma. Cancer 1998; 82:2326–2333. 90. Neugut AI, Rosenberg DJ, Ahsan H, et al. Association between coronary heart disease and cancers of the breast, prostate, and colon. Cancer Epidemiol Biomarkers Prev 1998; 7:869–873. 91. Bucher C, Jordan P, Mickeleit V, Torhorst J, Mihatsch MJ. Relative risk of malignant tumors in analgesic abusers. Effects of long-term intake of aspirin. Clin Nephrol 1999; 51:67–72. 92. Smalley W, Ray WA, Daugherty J, Griffin MR. Use of nonsteroidal antiinflammatory drugs and incidence of colorectal cancer: a population-based study. Arch Intern Med 1999; 159:161–166. 93. Collet JP, Sharpe C, Belzile E, Boivin JF, Hanley J, Abenhaim L. Colorectal cancer prevention by non-steroidal anti-inflammatory drugs: effects of dosage and timing. Br J Cancer 1999; 81:62–68. 94. Langman MJS, Cheng KK, Gilman EA, Lancashire RJ. Effect of antiinflammatory drugs on overall risk of common cancer: case–control study in general practice research database. Br Med J 2000; 320:1642–1646. 95. Coogan PF, Rosenberg L, Louik C, et al. NSAIDs and risk of colorectal cancer according to presence or absence of family history of the disease. Cancer Causes Control 2000; 11:249–255. 96. Kahn HS, Tatham LM, Thun MJ, Heath Jr, CW. Risk factors for self-reported colon polyps. J Gen Intern Med 1998; 13:303–310. 97. Sandler RS, Galanko JC, Murray SC, Helm JF. Aspirin and nonsteroidal anti-inflammatory agents and risk for colorectal adenomas. Gastroenterology 1998; 114:441–447. 98. Garcia Rodriguez LA, Huerta-Alvarez C. Reduced incidence of colorectal adenoma among long-term users of nonsteroidal antiinflammatory drugs: a pooled analysis of published studies and a new population-based study. Epidemiology 2000; 11:376–381.

66

Little and Sharp

99. Breuer-Katschinski B, Nemes K, Rump B, et al. Long-term use of nonsteroidal antiinflammatory drugs and the risk of colorectal adenomas. Digestion 2000; 61:129–134. 100. Tangrea JA, Albert PS, Lanza E, et al. Non-steroidal anti-inflammatory drug use is associated with reduction in recurrence of advanced and nonadvanced colorectal adenomas (United States). Cancer Causes Control 2003; 14:403–411. 101. Chan AT, Giovannucci EL, Schernhammer ES, et al. A prospective study of aspirin use and the risk for colorectal adenoma. Ann Intern Med 2004; 140:157–166. 102. Gann PH, Manson JE, Glynn RJ, Buring JE, Hennekens CH. Low-dose aspirin and incidence of colorectal tumors in a randomized trial. J Natl Cancer Inst 1993; 85:1220–1224. 103. Sturmer T, Glynn RJ, Lee IM, Manson JE, Buring JE, Hennekens CH. Aspirin use and colorectal cancer: post-trial follow-up data from the Physicians’ Health Study. Ann Intern Med 1998; 128:713–720. 104. Baron JA, Cole BF, Sandler RS, et al. A randomized trial of aspirin to prevent colorectal adenomas. N Engl J Med 2003; 348:891–899. 105. Benamouzig R, Deyra J, Martin A, et al. Daily soluble aspirin and prevention of colorectal adenoma recurrence: one-year results of the APACC trial. Gastroenetrology 2003; 125:328–336. 106. Asano TK, McLeod RS. Nonsteroidal anti-inflammatory drugs and aspirin for the prevention of colorectal adenomas and cancer: a systematic review. Dis Colon Rectum 2004; 47:665–673. 107. Giardiello FM, Yang VW, Hylind LM, et al. Primary chemoprevention of familial adenomatous polyps with sulindac. N Engl J Med 2002; 346: 1054–1059. 108. Ladenheim J, Garcia G, Titzer D, et al. Effect of sulindac on sporadic colonic polyps. Gastroenetrology 1995; 108:1083–1087. 109. Steinbach G, Lynch PM, Phillips RK, et al. The effect of celecoxib, a cyclooxegenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000; 342:1946–1952. 110. Rahme E, Barkun AN, Toubouti Y, Bardou M. The cyclooxygenase-2selective inhibitors rofecoxib and celecoxib prevent colorectal neoplasia occurrence and recurrence. Gastroenterology 2003; 125:404–412. 111. Martinez ME, O’Brien TG, Fultz KE, et al. Pronounced reduction in adenoma recurrence associated with aspirin use and a polymorphism in the ornithine decarboxylase gene. Proc Natl Acad Sci USA 2003; 100:7859–7864. 112. Lin HJ, Lakkides KM, Keku TO, et al. Prostaglandin H synthase 2 variant (Val511Ala) in African Americans may reduce the risk for colorectal neoplasia. Cancer Epidemiol Biomarkers Prev 2002; 11:1305–1315. 113. Cox DG, Pontes C, Guino E, et al., for the Bellvitge Colorectal Cancer Study Group. Polymorphisms in prostaglandin synthase2/cyclooxygenase 2 (PTGS2/ COX2) and risk of colorectal cancer. Br J Cancer 2004; 91:339–343. 114. Ulrich CM, Bigler J, Sparks R, et al. Polymorphisms in PTGS1 (¼COX-1) and risk of colorectal polyps. Cancer Epidemiol Biomarkers Prev 2004; 13: 889–893.

Epidemiology of Colorectal Cancer

67

115. Landi S, Moreno V, Gioia-Patricola L, et al. Association of common polymorphisms in inflammatory genes interleukin (IL)6, IL8, tumor necrosis factor alpha, NFKB1, and peroxisome proliferator-activated receptor gamma with colorectal cancer. Cancer Res 2003; 63:3560–3566. 116. Macarthur M, Sharp L, Hold GL, Little J, El Omar EM. The role of pro- and anti-inflammatory cytokine gene polymorphisms in colorectal cancer in the north east of Scotland. Gut 2004; 53(Suppl 3):A42. 117. Slattery ML, Samowitz W, Hoffman M, Ma KN, Levin TR, Neuhausen S. Aspirin, NSAIDs, and colorectal cancer: possible involvement in an insulinrelated pathway. Cancer Epidemiol Biomarkers Prev 2004; 13:538–545. 118. Lewis RC, Bostick RM, Xie D, et al. Polymorphism of the cylcin D1 gene, CCND1, and risk for incident sporadic colorectal adenomas. Cancer Res 2003; 63:8549–8553. 119. Little J, Bradley L, Bray MS, et al. Reporting, appraising, and integrating data on genotype prevalence and gene-disease associations. Am J Epidemiol 2002; 156:300–310. 120. World Cancer Research Fund in Association with American Institute for Cancer Research. Food, Nutrition and the Prevention of Cancer: A Global Perspective. Menasha: American Institute for Cancer Research, 1997. 121. Michels KB, Giovannucci E, Joshipura KJ, et al. Prospective study of fruit and vegetable consumption and incidence of colon and rectal cancers. J Natl Cancer Inst 2000; 92:1740–1752. 122. Terry P, Giovannucci E, Michels KB, et al. Fruit, vegetables, dietary fiber and risk of colorectal cancer. J Natl Cancer Inst 2001; 93:525–533. 123. Flood A, Velie EM, Chaterjee N, et al. Fruit and vegetable intakes and the risk of colorectal cancer in the Breast Cancer Detection Demonstration Project follow-up cohort. Am J Clin Nutr 2002; 75:936–943. 124. IARC Working Group. IARC Handbook on Cancer Prevention. Volume 8: Fruit and Vegetables. Lyon: International Agency for Research on Cancer, 2003. 125. Sandhu MS, White IR, McPherson K. Systematic review of the prospective cohort studies on meat consumption and colorectal cancer risk: a metaanalytical approach. Cancer Epidemiol Biomarkers Prev 2001; 10:439–446. 126. Norat T, Lukanova A, Ferrari P, Riboli E. Meat consumption and colorectal cancer risk: dose-response meta-analysis of epidemiological studies. Int J Cancer 2002; 98:241–256. 127. Nowell S, Coles B, Sinha R, et al. Analysis of total meat intake and exposure to individual heterocyclic amines in a case–control study of colorectal cancer: contribution of metabolic variation to risk. Mutat Res Fund Mol Mechan Mutagen 2002; 506–507:175–185. 128. Murtagh MA, Ma K, Sweeney C, Caan BJ, Slattery ML. Meat consumption patterns and preparation, genetic variants of metabolic enzymes, and their association with rectal cancer in men and women. J Nutr 2004; 134:776–784. 129. Ishibe N, Sinha R, Hein DW, et al. Genetic polymorphisms in heterocyclic amine metabolism and risk of adenomas. Pharmacogenetics 2002; 12:145–150. 130. Turner F, Smith G, Sachse C, et al. Vegetable, fruit and meat consumption and potential risk modifying genes in relation to colorectal cancer. Int J Cancer 2004; 112:259–264.

68

Little and Sharp

131. Cortessis V, Siegmund K, Chen Q, et al. A case–control study of microsomal epoxide hydrolase, smoking, meat consumption, glutathione S-transferase M3, and risk of colorectal adenomas. Cancer Res 2001; 61:2381–2385. 132. Kampman E, Slattery ML, Bigler J, et al. Meat consumption, genetic susceptibility, and colon cancer risk: a United States multicenter case–control study. Cancer Epidemiol Biomarkers Prev 1999; 8:15–24. 133. Tiemersma EW, Kampman E, Bueno de Mesquita HB, et al. Meat consumption, cigarette smoking, and genetic susceptibility in the etiology of colorectal cancer: results from a Dutch prospective study. Cancer Causes Control 2002; 13:383–393. 134. Stewart BW, Kleiheus P (eds). World Cancer Report. Lyon: International Agency for Research on Cancer, 2003. 135. Kushi L, Giovannucci E. Dietary fat and cancer. Am J Med 2002; 113: S63–70S. 136. Nkondjock A, Shatenstein B, Maisonneuve P, Ghadirian P. Specific fatty acids and human colorectal cancer: an overview. Cancer Detect Prev 2003; 27:55–66. 137. Koh W, Yuan J, van der Berg D, Lee H, Yu MC. Interaction between cyclooxygenase-2 gene polymorphism and dietary n-6 polyunsaturated fatty acids on colon cancer risk: the Singapore Chinese Health Study. Br J Cancer 2004; 90:1760–1764. 138. Bingham SA, Day NE, Luben R, et al. Dietary fiber in food and protection against colorectal cancer in the European Prospective Investigation into Cancer and Nutrition (EPIC): an observational study. Lancet 2003; 361:1496–1501. 139. Fuchs CS, Giovannucci EL, Colditz GA, et al. Dietary fiber and the risk of colorectal cancer and adenoma in women. N Engl J Med 1999; 340:169–176. 140. Pietinen P, Malila N, Virtanen M, et al. Diet and risk of colorectal cancer in a cohort of Finnish men. Cancer Causes Control 1999; 10:387–396. 141. Schatzkin A, Lanza E, Corle D, et al., Polyp Prevention Trial Study Group. Lack of effect of a low-fat, high-fiber diet on the recurrence of colorectal adenomas. N Engl J Med 2000; 342:1149–1155. 142. Alberts DS, Martinez ME, Roe DJ, et al. Lack of effect of a high-fiber cereal supplement on the recurrence of colorectal adenomas. N Engl J Med 2000; 342:1156–1162. 143. Bonithon-Kopp C, Kronborg O, Giacosa A, Rath U, Faivre J. Calcium and fiber supplementation in prevention of colorectal adenoma recurrence: a randomised intervention trial. Lancet 2000; 356:1300–1306. 144. Greenwald P, Clifford CK, Milner JA. Diet and cancer prevention. Eur J Cancer 2001; 37:948–965. 145. Hill MJ. Bile acids and colorectal cancer: a hypothesis. Eur J Cancer Prev 1991; 1:69–74. 146. Owen RW, Dodo M, Thompson MH, Hill MJ. The fecal ratio of lithocholic to deozycholic acid may be an important aetiological factor in colorectal cancer. Biochem Soc Trans 1984; 12:861–862. 147. Owen RW, Day DW, Thompson MH. Fecal steroids and colorectal cancer: steroid profiles in subjects with adenomatous polyps of the large bowel. Eur J Cancer Prev 1992; 1:105–112.

Epidemiology of Colorectal Cancer

69

148. Owen RW. Fecal steriods and colorectal carcinogenesis. Scand J Gastroenterol 1997; 32:76–82. 149. Hill MJ, Drasar BS, Williams REO, et al. Fecal bile-acids and clostridia in patients with cancer of the large bowel. Lancet 1975; i:535–538. 150. Reddy BS, Wynder EL. Metabolic epidemiology of colon cancer. Fecal bile acids and neutral sterols in colon cancer patients and patients with adenomatous polyps. Cancer 1977; 39:2533–2539. 151. Moskovitz M, White C, Barnett RN, et al. Diet, fecal bile acids, and neutral sterols in carcinoma of the colon. Dig Dis Sci 1979; 24:746–751. 152. Mudd DG, McKelvey STD, Norwood W, Elmore DT, Roy AD. Fecal bile acid concentrations of patients with carcinoma or increased risk of carcinoma in the large bowel. Gut 1980; 21:587–590. 153. Murray WR, Blackwood A, Trotter JM, Calman KC, MacKay C. Fecal bile acids and clostridia in the etiology of colorectal cancer. Br J Cancer 1980; 41:923–928. 154. Kaibara N, Sasaki T, Ikeguchi M, Koga S, Ikawa S. Fecal bile acids and neutral sterols in Japanese with large bowel carcinoma. Oncology 1983; 40: 255–258. 155. Hikasa Y, Tanida N, Ohno T, Shimoyama T. Fecal bile acid profiles in patients with large bowel cancer in Japan. Gut 1984; 25:833–838. 156. Breuer NF, Dommes P, Jaekel S, Goebell H. Fecal bile acid excretion pattern in colonic cancer patients. Dig Dis Sci 1985; 30:852–859. 157. Tanida N, Hikasa Y, Shimoyama T, Setchell KDR. Comparison of fecal bile acid profiles between patients with adenomatous polyps of the large bowel and healthy subjects in Japan. Gut 1984; 25:824–832. 158. Breuer NF, Jaekel S, Dommes P, Goebell H. Fecal bile acids in patients with adenomatous polyps of the colon. A case–control study. Digestion 1986; 34:87–92. 159. Imray CHE, Radley S, Davis A, et al. Fecal unconjugated bile acids in patients with colorectal cancer or polyps. Gut 1992; 33:1239–1245. 160. Little J, Owen RW, Fernandez F, et al. Asymptomatic colorectal neoplasia and fecal characteristics: a case–control study of subjects participating in the Nottingham fecal occult blood screening trial. Dis Colon Rectum 2002; 45: 1233–1241. 161. Duthie SJ. Folic acid instability and cancer: mechanisms of DNA instability. Br Med Bull 1999; 55:578–592. 162. Glynn SA, Albanes D, Pietinen P, et al. Colorectal cancer and folate status: a nested case–control study among male smokers. Cancer Epidemiol Biomarkers Prev 1996; 5:487–494. 163. Ma J, Stampfer MJ, Giovannucci E, et al. Methylenetetrahydrofolate reductase polymorphism, dietary interactions and risk of colorectal cancer. Cancer Res 1997; 57:1098–1102. 164. Kato A, Dnistrian AM, Schwartz M, et al. Serum folate, homocysteine and colorectal cancer risk in women: a nested case–control study. Br J Cancer 1999; 79:1917–1921. 165. Bird CL, Swendseid ME, Witte JS, et al. Red cell and plasma folate, folate consumption, and the risk of colorectal adenomatous polyps. Cancer Epidemiol Biomarkers Prev 1995; 4:709–714.

70

Little and Sharp

166. Paspatis GA, Kalafatis E, Oros L, Xourgias V, Koutsioumpa P, Karamanolis DG. Folate status and adenomatous colonic polyps: a colonoscopically controlled study. Dis Colon Rectum 1995; 38:64–68. 167. Giovannucci E, Rimm EB, Ascherio A, Stampfer MJ, Colditz GA, Willett WC. Alcohol, low-methionine–low-folate diets, and risk of colon cancer in men. J Natl Cancer Inst 1995; 87:265–273. 168. Giovannucci E, Stampfer MJ, Colditz GA, et al. Multivitamin use, folate, and colon cancer in women in the Nurses’ Health Study. Ann Intern Med 1998; 129:517–524. 169. Su LJ, Arab L. Nutritional status of folate and colon cancer risk: evidence from NHANES I epidemiologic follow-up study. Ann Epidemiol 2001; 11:65–72. 170. Terry P, Jain M, Miller AB, Howe GR, Rohan TE. Dietary intake of folic acid and colorectal cancer risk in a cohort of women. Int J Cancer 2002; 97: 864–867. 171. Flood A, Caprario L, Chaterjee N, Lacey JVJ, Schairer C, Schatzkin A. Folate, methionine, alcohol, and colorectal cancer in a prospective study of women in the United States. Cancer Causes Control 2002; 13:551–561. 172. Konings EJM, Goldbohm RA, Brants HAM, Saris WHM, van den Brandt PA. Intake of dietary folate vitamers and risk of colorectal carcinoma. Cancer 2002; 95:1421–1433. 173. Harnack L, Jacobs DR Jr., Nicodemus K, Lazovich D, Anderson K, Folsom AR. Relationship of folate, vitamin B6, vitamin B12, and methionine intake to incidence of colorectal cancers. Nutr Cancer 2002; 43:152–158. 174. Benito E, Stiggelbout A, Bosch FX, et al. Nutritional factors in colorectal cancer risk: a case–control study in Majorca. Int J Cancer 1991; 49:161–167. 175. Freudenheim JL, Graham S, Marshall JR, Haughey BP, Cholewinski S, Wilkinson G. Folate intake and carcinogenesis of the colon and rectum. Int J Epidemiol 1991; 20:368–374. 176. Meyer F, White E. Alcohol and nutrients in relation to colon cancer in middleaged adults. Am J Epidemiol 1993; 138:225–236. 177. Ferraroni M, La Vecchia C, D’Avanzo B, Negri E, Franceschi S, Decarli A. Selected micronutrient intake and the risk of colorectal cancer. Br J Cancer 1994; 70:1150–1155. 178. Boutron-Ruault MC, Senesse P, Faivre J, Couillault C, Belghiti C. Folate and alcohol intakes: related or independent roles in the adenoma-carcinoma sequence? Nutr Cancer 1996; 26:337–346. 179. La Vecchia C, Braga C, Negri E, et al. Intake of selected micronutrients and risk of colorectal cancer. Int J Cancer 1997; 73:525–530. 180. Slattery ML, Schaffer D, Edwards SL, Ma KN, Potter JD. Are dietary factors involved in DNA methylation associated with colon cancer? Nutr Cancer 1997; 28:52–62. 181. Levi F, Pasche C, Lucchini F, La Vecchia C. Selected micronutrients and colorectal cancer: a case–control study from the Canton of Vaud, Switzerland. Eur J Cancer 2000; 36:2115–2119. 182. Le Marchand L, Donlon T, Hankin JH, Kolonel LN, Wilkens LR, Seifried A. B-vitamin intake, metabolic genes, and colorectal cancer risk (United States). Cancer Causes Control 2002; 13:239–248.

Epidemiology of Colorectal Cancer

71

183. Keku T, Millikan R, Worley K, et al. 5,10-methylenetetrahydrofolate reductase codon 677 and 1298 polymorphisms and colon cancer in African Americans and whites. Cancer Epidemiol Biomarkers Prev 2002; 11:1611–1621. 184. La Vecchia C, Negri E, Pelucchi C, Franceschi S. Dietary folate and colorectal cancer. Int J Cancer 2002; 102:545–547. 185. White E, Shannon JS, Patterson RE. Relationship between vitamin and calcium supplement use and colon cancer. Cancer Epidemiol Biomarkers Prev 1997; 6:769–774. 186. Jacobs EJ, Connell CJ, Patel AV, et al. Multivitamin use and colon cancer mortality in the Cancer Prevention Study II cohort (United States). Cancer Causes Control 2001; 12:927–934. 187. Giovannucci E, Stampfer MJ, Colditz GA, et al. Folate, methionine and alcohol intake and risk of colorectal adenoma. J Natl Cancer Inst 1993; 85:875–884. 188. Benito E, Cabeza E, Moreno V, Obrador A, Bosch FX. Diet and colorectal adenomas: a case–control study in Majorca. Int J Cancer 1993; 55:213–219. 189. Tseng M, Murray SC, Kupper LL, Sandler RS. Micronutrients and the risk of colorectal adenomas. Am J Epidemiol 1996; 144:1005–1014. 190. Chen J, Giovannucci E, Hankinson SE, et al. A prospective study of methylenetetrahydrofolate reductase and methionine synthase gene polymorphisms, and risk of colorectal adenoma. Carcinogenesis 1998; 19:2129–2132. 191. Breuer-Katschinski B, Nemes K, Marr A, et al. Colorectal adenomas and diet: a case–control study. Dig Dis Sci 2001; 46:86–95. 192. Boyapati SM, Bostick RM, McGlynn KA, et al. Folate intake, MTHFR C677T polymorphism, alcohol consumption, and risk for sporadic colorectal adenoma (United States). Cancer Causes Control 2004; 15:493–501. 193. Baron JA, Sandler RS, Haile RW, Mandel JS, Mott LA, Greenberg ER. Folate intake, alcohol consumption, cigarette smoking, and risk of colorectal adenomas. J Natl Cancer Inst 1998; 90:57–62. 194. Herbert V. Recommended dietary intakes (RDI) of folate in humans. Am J Clin Nutr 1987; 45:661–670. 195. Fuchs CS, Willett WC, Colditz GA, et al. The influence of folate and multivitamin use on the familial risk of colon cancer in women. Cancer Epidemiol Biomarkers Prev 2002; 11:227–234. 196. Johnson WG. DNA polymorphism-diet-cofactor-development hypothesis and the gene-teratogen model for schizophrenia and other developmental disorders. Am J Med Genet 1999; 88:311–323. 197. Sharp L, Little J. Polymorphisms in genes involved in folate metabolism and colorectal neoplasia: a HuGE review. Am J Epidemiol 2004; 159:423–443. 198. Curtin K, Bigler J, Slattery ML, Caan B, Potter JD, Ulrich CM. MTHFR C677T and A1298C polymorphisms: diet, estrogen and risk of colon cancer. Cancer Epidemiol Biomarkers Prev 2004; 13:285–292. 199. Jiang QT, Chen K, Ma XY, et al. A case-control study on the polymorphisms of methylenetetrahydrofolate reductases, drinking interaction and susceptibility in colorectal cancer. Zhonghua Liu Xing Bing Xue Za Zhi 2004; 25:612–616. 200. Kim D, Ahn Y, Lee B, Tsuji E, Kiyohara C, Kono S. Methylenetetrahydrofolate reductase polymorphism, alcohol intake, and risks of colon and rectal cancers in Korea. Cancer Lett 2004; 216:199–205.

72

Little and Sharp

201. Yin G, Kono S, Toyomura K, et al. Methylenetetrahydrofolate reductase C677T and A1298C polymorphisms and colorectal cancer: the Fukuoka Colorectal Cancer Study. Cancer Sci 2004; 95:908–913. 202. Giovannucci E, Chen J, Smith-Warner SA, et al. Methylenetetrahydrofolate reductase, alcohol dehydrogenase, diet and risk of colorectal adenomas. Cancer Epidemiol Biomarkers Prev 2003; 12:970–979. 203. Ma J, Stampfer MJ, Christensen B, et al. A polymorphism of the methionine synthase gene: assocation with plasma folate, vitamin B12, homocyst(e)ine, and colorectal cancer risk. Cancer Epidemiol Biomarkers Prev 1999; 8:825–829. 204. Goode EL, Ulrich CM, Bigler J, et al. Methionine synthase D919G polymorphism: possible interaction with dietary nutrients and association with colorectal adenoma risk. Proc Am Assoc Cancer Res 2002; 43:659–660. 205. Shannon B, Gnanasampanthan S, Beilby J, Iacopetta B. A polymorphism in the methylenetetrahydrofolate reductase gene predisposes to colorectal cancers with mircrosatellite instability. Gut 2002; 50:520–524. 206. Lenz HJ, Zhang W, Zahedy S, Gil J, Yu M, Stoehlmacher J. A 6 base-pair deletion in the 3 UTR of the thymidylate synthase (TS) gene predicts TS mRNA expression in colorectal tumors: a possible candidate gene for colorectal cancer risk. Proc Am Assoc Cancer Res 2002; 43:660. 207. Ulrich CM, Bigler J, Bostick R, Fosdick L, Potter JD. Thymidylate synthase promoter polymorphism, interaction with folate intake, and risk of colorectal adenomas. Cancer Res 2002; 62:3361–3364. 208. Chen J, Hunter DJ, Stampfer MJ, et al. Polymorphism in the thymidylate synthase promoter enhancer region modifies the risk and survival of colorectal cancer. Cancer Epidemiol Biomarkers Prev 2003; 12:958–962. 209. Chen J, Hunter DJ, Stampfer MJ, et al. Polymorphism in the thymidylate synthase promoter enhancer region modifies the risk and survival of colorectal cancer. Cancer Epidemiol Biomarkers Prev 2003; 12:958–962. 210. Adleff V, Hitre E, Koves I, Orosz Z, Hajnal A, Kralovanszky J. Heterozygote deficiency in thymidylate synthase enhancer region polymorphism genotype distribution in Hungarian colorectal cancer patients. Int J Cancer 2004; 108:852–856. 211. Wolf CR, Smith G. Chapter 18. Cytochrome P450 CYP2D6. In: Vineis P, Malats N, Lang M, et al. Metabolic polymorphisms and susceptibility to cancer. Lyon: International Agency for Research on Cancer, 1999:209–229. 212. IARC Working Group. IARC Handbooks of Cancer Prevention, Volume 2: Carotenoids. Lyon: International Agency for Research on Cancer, 1998. 213. Marcus PM, Newcomb PA. The association of calcium and vitamin D, and colon and rectal cancer in Wisconsin women. Int J Epidemiol 1998; 27: 788–793. 214. Baron JA, Beach M, Mandel JS, et al. Calcium supplements for the prevention of colorectal adenomas. N Engl J Med 1999; 340:101–107. 215. Martinez ME, Marshall JR, Sampliner R, Wilkinson J, Alberts DS. Calcium, vitamin D, and risk of adenoma recurrence (United States). Cancer Causes Control 2002; 13:213–220. 216. Terry P, Baron JA, Bergkvist L, Holmberg L, Wolk A. Dietary calcium and vitamin D intake and risk of colorectal cancer: a prospective cohort study in women. Nutr Cancer 2002; 43:39–46.

Epidemiology of Colorectal Cancer

73

217. McCullough ML, Robertson AS, Rodriguez C, et al. Calcium, vitamin D, diary products, and risk of colorectal cancer in the Cancer Prevention Study II Nutrition Cohort (United States). Cancer Causes Control 2003; 14:1–12. 218. Wallace K, Baron JA, Cole BF, et al. Effects of calcium supplementation on the risk of large bowel polyps. J Natl Cancer Inst 2004; 96:921–925. 219. Cho E, Smith-Warner SA, Spiegelman D, et al. Diary foods, calcium, and colorectal cancer: a pooled analysis of 10 cohort studies. J Natl Cancer Inst 2004; 96:1015–1022. 220. Newmark HL, Wargovich MJ, Bruce WR. Colon cancer and dietary fat, phosphate and calcium: a hypothesis. J Natl Cancer Inst 1984; 72:1323–1325. 221. Wargovich MJ, Eng VW, Newmark HL, Bruce WR. Calcium ameliorates the toxic effect of deoxycholic acid on colonic epithelium. Carcinogenesis 1983; 4:1205–1207. 222. Ingles SA, Wang J, Coetzee GA, Lee ER, Frankl HD, Haile RW. Vitamin D receptor polymorphisms and risk of colorectal adenomas (United States). Cancer Causes Control 2001; 12:607–614. 223. Wong HL, Seow A, Arakawa K, Lee HP, Yu MC, Ingles SA. Vitamin D receptor start codon polymorphism and colorectal cancer risk: effect modification by dietary calcium and fat in Singapore Chinese. Carcinogenesis 2003; 24:1091–1095. 224. Slattery ML, Yakumo K, Hoffman M, Neuhausen S. Variants of the VDR gene and risk of colon cancer (United States). Cancer Causes Control 2001; 12: 359–364. 225. Peters U, McGlynn KA, Chatterjee N, et al. Vitamin D, calcium, and vitamin D receptor polymorphism in colorectal adenomas. Cancer Epidemiol Biomarkers Prev 2001; 10:1267–1274. 226. Boyapati SM, Bostick RM, McGlynn KA, et al. Calcium, vitamin D, and risk of colorectal adenoma: dependency on vitamin D receptor BsmI polymorphism and non-steroidal anti-inflammatory drug use? Cancer Epidemiol Biomarkers Prev 2003; 12:631–637. 227. Corrao G, Bagnardi V, Zambon A, Arico S. Exploring the dose–response relationship between alcohol consumption and the risk of several alcohol-related conditions: a meta-analysis. Addiction 1999; 94:1551–1573. 228. Cho E, Smith-Warner SA, Ritz J, et al. Alcohol intake and colorectal cancer: a pooled analysis of 8 cohort studies. Ann Intern Med 2004; 140:603–613. 229. Tiemersma EW, Wark PA, Ock MC, et al. Alcohol consumption, alcohol dehydrogenase 3 polymorphism, and colorectal adenomas. Cancer Epidemiol Biomarkers Prev 2003; 12:419–425. 230. Koenuma M, Yamori T, Tsuruo T. Insulin and insulin-like growth factor 1 stimulate proliferation of metastatic varaints of colon carcinoma 26. Jap J Cancer Res 1989; 80:51–58. 231. Tran TT, Medline A, Bruce WR. Insulin promotion of colon tumors in rats. Cancer Epidemiol Biomarkers Prev 1996; 5:1013–1015. 232. McKeown-Eyssen G. Epidemiology of colorectal cancer revisited: are serum triglycerides and/or plasma glucose associated with risk? Cancer Epidemiol Biomarkers Prev 1994; 3:687–695. 233. Giovannucci E. Insulin and colon cancer. Cancer Causes Control 1995; 6: 164–179.

74

Little and Sharp

234. La Vecchia C, Negri E, Decarli A, Franceschi S. Diabetes mellitus and colorectal cancer risk. Cancer Epidemiol Biomarkers Prev 1997; 6:1007–1010. 235. Hu FB, Manson JE, Liu S, et al. Prospective study of adult onset diabetes mellitus (type 2) and risk of colorectal cancer in women. J Natl Cancer Inst 1999; 91:542–547. 236. Lund Nilsen TI, Vatten LJ. Prospective study of colorectal cancer risk and physical activity, diabetes, blood glucose and BMI: exploring the hyperinsulinaemia hypothesis. Br J Cancer 2001; 84:412–422. 237. Nishii T, Kono S, Abe H, et al. Glucose intolerance, plasma insulin levels, and colon adenomas in Japanese men. Jap J Cancer Res 2001; 92:836–840. 238. Marugame T, Lee K, Eguchi H, Oda T, Shinchi K, Kono S. Relation of impaired glucose tolerance and diabetes mellitus to colorectal adenomas in Japan. Cancer Causes Control 2002; 13:917–921. 239. Chang CK, Ulrich CM. Hyperinsulinaemia and hyperglycaemia: possible risk factors of colorectal cancer among diabetic patients. Diabetologia 2003; 46:595–607. 240. Trevisan M, Liu J, Muti P, Misciagna G, Menotti A, Fucci F. Markers of insulin resistance and colorectal cancer mortality. Cancer Epidemiol Biomarkers Prev 2001; 10:937–941. 241. Colangelo LA, Gapstur SM, Gann PH, Dyer AR, Liu K. Colorectal cancer mortality and factors related to the insulin resistance syndrome. Cancer Epidemiol Biomarkers Prev 2002; 11:385–391. 242. Yamada K, Araki S, Tamura M, et al. Relation of serum total cholesterol, serum triglycerides and fasting plasma glucose to colorectal carcinoma in situ. Int J Epidemiol 1998; 27:794–798. 243. Schoen RE, Tangen CM, Kuller LH, et al. Increased blood glucose and insulin, body size, and incident colorectal cancer. J Natl Cancer Inst 1999; 91: 1147–1154. 244. Teramukai S, Rohan T, Lee K, Eguchi H, Oda T, Kono S. Insulin-like growth factor (IGF)-1, IGF-binding protein-3 and colorectal adenomas in Japanese men. Jap J Cancer Res 2002; 93:1187–1194. 245. Palmqvist R, Stattin P, Rinaldi S, et al. Plasma insulin, IGF-binding proteins-1 and -2 and risk of colorectal cancer: a prospective study in northern Sweden. Int J Cancer 2003; 107:89–93. 246. Saydah SH, Platz EA, Rifai N, Pollak MN, Brancati FL, Helzlsouer KJ. Association of markers of insulin and glucose control with subsequent colorectal cancer risk. Cancer Epidemiol Biomarkers Prev 2003; 12:412–418. 247. Kaaks R, Toniolo P, Akhmedkhanov A, et al. Serum C-peptide, insulin-like growth factor (IGF)-1, IGF-binding proteins, and colorectal cancer risk in women. J Natl Cancer Inst 2000; 92:1592–1600. 248. Ma J, Giovannucci E, Pollak M, et al. A prospective study of plasma Cpeptide and colorectal cancer risk in men. J Natl Cancer Inst 2004; 96:546–553. 249. Franceschi S, Dal Maso L, Augustin L, et al. Dietary glycemic load and colorectal cancer risk. Ann Oncol 2001; 12:173–178. 250. Higginbotham S, Zhang Z, Lee I, et al. Dietary glycemic load and risk of colorectal cancer in the women’s health study. J Natl Cancer Inst 2004; 96: 229–233.

Epidemiology of Colorectal Cancer

75

251. Oh K, Willett WC, Fuchs CS, Giovannucci E. Glycemic index, glycemic load, and carbohydrate intake in relation to risk of distal colorectal adenoma in women. Cancer Epidemiol Biomarkers Prev 2004; 13:1192–1198. 252. Kaaks R. Nutrition, insulin, IGF-1 metabolism and cancer risk: a summary of epidemiological evidence. Novart Found Symp 2004; 262:247–260. 253. Ma J, Pollak MN, Giovannucci E, et al. Prospective study of colorectal cancer risk in men and plasma levels of insulin-like growth factor (IGF)-I and IGFbinding protein-3. J Natl Cancer Inst 1999; 91:620–625. 254. Giovannucci E, Pollak MN, Platz EA, et al. A prospective study of plasma insulin-like growth factor-1 and binding protein-3 and risk of colorectal neoplasia in women. Cancer Epidemiol Biomarkers Prev 2000; 9:345–349. 255. Palmqvist R, Hallmans G, Rinaldi S, et al. Plasma insulin-like growth factor 1, insulin-like growth factor binding protein 3, and risk of colorectal cancer: a prospective study in northern Sweden. Gut 2002; 50:642–646. 256. Le Marchand L, Donlon T, Seifried A, Kaaks R, Rinaldi S, Wilkens LR. Association of a common polymorphism in the human GH1 gene with colorectal neoplasia. J Natl Cancer Inst 2002; 94:454–460. 257. Slattery ML, Samowitz W, Curtin K, et al. Associations among IRS1, IRS2, IGF1, and IGFBP3 genetic polymorphisms and colorectal cancer. Cancer Epidemiol Biomarkers Prev 2004; 13:1206–1214. 258. HuGENet2: http://www.cdc.gov/genomics/hugenet/default.htm

3 Colorectal Cancer Screening Robert J. C. Steele Department of Surgery and Molecular Oncology, Ninewells Hospital and Medical School, University of Dundee, Dundee, U.K.

INTRODUCTION Colorectal cancer is a major problem worldwide, and the highest incidences are found in the most developed countries. In Europe, the incidence is currently very similar to that of lung and breast cancer (about 135,000 cases per year), and in the developed countries there are some 250,000 deaths attributable to the disease each year (1). The main symptoms of colorectal cancer consist of overt rectal bleeding, change of bowel habit, abdominal pain and anemia, and, unfortunately, a tumor giving rise to any of these symptoms is likely to be locally advanced. As a result symptomatic cancers are rarely early and, in the United Kingdom, only about 8% of colorectal cancers present at Dukes’ stage A with 25% having distant metastases at the time of diagnosis (2); in many instances ‘‘symptomatic’’ early cancer is probably discovered as a result of investigating symptoms arising from concurrent benign causes such as hemorrhoids or irritable bowel syndrome. It is well established that early-stage colorectal cancer carries a much better prognosis than does late-stage disease (3), but it is self-evident that relying purely on symptomatic presentation will never substantially increase the proportion of cancers treated early and thus with curative intent. The only reliable way to detect early disease consistently is to look for it actively in asymptomatic individuals; in other words to screen. Screening for colorectal cancer is now widespread throughout the developed world

77

78

Steele

although very few countries have a systematic screening policy. In this chapter we shall first briefly consider the principles of screening and then look at colorectal cancer as a suitable target for screening. The evidence for the major screening modalities, cost-effectiveness issues, and future approaches using novel methods will then be examined. PRINCIPLES OF SCREENING Screening can be defined as a process whereby a test is applied to individuals with a view to identify unrecognized disease at an early stage when treatment will be more effective. It is often stated that the individuals to whom screening is applied are asymptomatic but experience shows that this is not necessarily the case; an invitation to be screened may be more readily accepted in a patient with unreported symptoms. Indeed, recent work has indicated that about 50% of subjects accepting an invitation to be screened for colorectal cancer have significant symptoms, although these symptoms are unrelated to the findings on screening colonoscopy (4). It is also very important to be clear about the purpose of screening in a particular context. If the aim is to reduce the burden of disease on a community the correct approach is population screening; this requires the use of a test that is associated with a high uptake and low cost. If, on the other hand, the aim is to respond to an individual’s request for information regarding their disease status, uptake is not an issue and the emphasis must be on a test of high sensitivity and specificity. In this chapter, the emphasis will be largely on population screening, but it must be recognized that case finding on an individual basis forms the foundation of screening in many countries. The principles underlying an effective screening intervention were developed by Wilson and Jungner in 1968 (5), and these are summarized in Table 1. The essence of these principles is that the target disease process should be a common problem that has a better outcome when treated at an early stage, and that the test employed is acceptable and sufficiently sensitive, specific, and inexpensive to be cost-effective. That screening is a useful strategy may seem obvious, but the process of screening is associated with inherent biases that inevitably make screendetected disease appear to have a better prognosis than symptomatic disease whether or not the screening process has had a true effect on outcome. These biases are three in number: length-time bias, lead-time bias, and volunteer bias. Length-time bias arises from the fact that intermittent screening tests will tend to pick up slow-growing, indolent disease that is likely to have a better prognosis than the rapidly advancing disease, which is more likely to appear with symptoms between screening intervals (Fig. 1). Leadtime bias arises from early diagnosis itself; this will always lead to an apparent improved duration of survival merely by shifting the point of diagnosis forward and not necessarily by improving survival (Fig. 2). Volunteer bias is

Colorectal Cancer Screening

79

Table 1 Principles of Screening The condition should be an important health problem There should be an accepted treatment for patients with recognized disease Facilities for diagnosis and treatment should be available There should be a recognizable latent or early symptomatic stage There should be a suitable test or examination The test should be acceptable to the population The natural history of the condition, including development for latent to declared disease, should be adequately understood There should be an agreed policy on whom to treat as patients The cost of case-finding (including diagnosis and treatment of patients diagnosed) should be economically balanced in relation to possible expenditure on medical care as a whole Case finding should be a continuing process and not a ‘‘once and for all’’ project Source: From Ref. 5.

created by the fact that screening invitations tend to be accepted by healthconscious individuals who are likely to have a better outcome for reasons other than early detection of the tumor. The collective effect of these biases is to exaggerate the beneficial effect of screening, and to ensure that screening is producing a real benefit it is essential to carry out population-based randomized trials in which the group randomized to screening is analyzed as a whole and includes those who refuse the invitation to be screened and those who develop cancers that are not detected by screening. Only if the disease-specific mortality in the whole of this group is significantly lower than in the randomly selected group that is not offered screening can we be sure that the screening process is producing a truly beneficial effect.

Figure 1 Length bias. Screening tests tend to detect slow-growing disease, whereas rapidly progressive disease tends to arise and present between screening intervals.

80

Steele

Figure 2 Lead-time bias. Early diagnosis will always appear to lengthen survival whether or not it affects the rate of tumor growth.

Cost-effectiveness is a more arbitrary measure and is essentially dependent upon a society’s willingness to pay for prolonged high-quality survival. In screening cost-effectiveness can be calculated in purely monetary terms but it has to be remembered that screening produces morbidity both psychological and physical, and it is important that this is also factored into the equation when a cost-benefit analysis is carried out. The cost-effectiveness of colorectal cancer screening will be examined later in this chapter. COLORECTAL CANCER AS A SUITABLE TARGET FOR SCREENING Wilson and Jungner (5) stated that for screening to be successful: 1. the condition should be an important health problem 2. there should be an accepted treatment for patients with recognized disease 3. the natural history of the condition including development from latent to declared disease should be adequately understood 4. treatment of early-stage disease confers a benefit over treating the same disease at a later, symptomatic stage Using these criteria there is little doubt that colorectal cancer is a suitable candidate for screening. In the western world it is extremely common and there are well-established methods of diagnosis and treatment. Colonoscopy is the gold standard investigation for symptomatic patients and high-risk individuals, although, as will be discussed in detail, there is considerable debate as

Colorectal Cancer Screening

81

to whether it should be used as a primary screening tool in asymptomatic people. Barium enema is still widely employed, but the published evidence would indicate that this is an inferior method of investigation compared with colonoscopy (6), and most authorities would now recommend that it be supplemented by flexible sigmoidoscopy. The latest computed tomography (CT) technology and analytical software has resulted in the development of CT colography or ‘‘virtual colonoscopy,’’ and this is emerging as a highly sensitive and specific diagnostic modality for colorectal neoplasia, set to render barium enema obsolete and perhaps replace colonoscopy as a purely diagnostic procedure (7). Consensus regarding treatment protocols for colorectal cancer is also improving. Surgery is fairly well standardized, particularly for rectal cancer (8), although the role of laparoscopic surgery is still to be fully established (9). Adjuvant therapy, on the other hand, is a rapidly shifting area, and although chemotherapy for Dukes’ stage C disease is now widely accepted and backed up by high-quality randomized evidence, there is a great deal of debate around the ideal agents and the question of adjuvant radiotherapy for rectal cancer (10). This does not, however, detract from the basic principle that surgical excision is the only potentially curative approach in the majority of tumors, and that the earlier the tumor, the more likely it is to be successful. As far as natural history is concerned, the evidence for the adenoma– carcinoma sequence is strong although essentially circumstantial (Table 2), and it is now generally accepted that most, if not all, invasive cancers arise from preexisiting adenomas (11). This offers an opportunity to reduce the incidence of colorectal cancer if the screening process employed detects significant adenomas and allows for their removal. It is also well established that surgery for early-stage disease results in better outcomes than for latestage disease; the five-year survival for Dukes’ stage A in the United Kingdom

Table 2 Evidence Supporting the Adenoma–Carcinoma Sequence in Colorectal Cancer Adenomas and carcinomas are frequently contiguous The anatomical distribution of adenomas and carcinomas is similar Adenomas over 2 cm in diameter have a 50% risk of harboring invasive malignancy The prevalence of adenomas is similar to that of carcinomas, and the average age of adenoma patients is about five years younger In about one-third of all surgical specimens resected for carcinoma, synchronous adenomas will be found Familial adenomatous polyposis (FAP) is unequivocally premalignant Adenomas and carcinomas share similar patterns of chromosomal abnormality and genetic mutation There is indirect but strong evidence that colonoscopy and polypectomy are associated with a reduced incidence of carcinoma Source: From Ref. 11.

82

Steele

is currently around 90% and for stage C 40% (12). It must be appreciated, however, that part of this observation may be accounted for by lead-time bias, and it is important to evaluate the evidence provided by populationbased randomized trials of screening as outlined earlier. FECAL OCCULT BLOOD SCREENING There are various methods of testing for blood in the feces but the method that has been employed in all the published population-based screening trials involves the use of guaiac. Guaiac tests, by detecting peroxidases, react to heme in its free form or bound to protein (globin, myoglobin, and some cytochromes), and do not detect the degradation products of heme that are formed in the intestine as these lack peroxidase activity (13). Heme enters the proximal gastrointestinal tract as hemoglobin or myoglobin in food or as red cells from bleeding lesions, and relatively little is absorbed by the small intestine. However, in the colon, heme is modified by the microflora so that it loses its peroxidase activity, and, as a result, guaiac tests are more sensitive for distal (colonic) than for proximal (gastric) bleeding pathology. Using an unrehydrated guaiac test the clinical sensitivity (proportion of subjects with the disease who have a positive test) for colorectal cancer is only around 50% in a population screening context; this figure is derived from the interval cancer rate found in the randomized trials and will be discussed later. The reason for the low sensitivity of this test is presumably related to the fact that cancers bleed intermittently. The specificity of this test (proportion of subjects without the disease who have a negative test) is about 98% but, as the majority of the population do not have colorectal cancer, this still translates into a high false-positive rate. This is caused by a combination of factors including dietary hemoglobin, myoglobin, and peroxidase. The sensitivity of the test can be improved by rehydration before testing but at the expense of decreasing the specificity and thereby increasing the false-positive rate. Specificity is harder to deal with, and although appropriate dietary restriction for weakly positive tests has been employed, a recent meta-analysis suggests that this approach is ineffective and therefore unnecessary (14). The other issue is that traces of an individual’s own blood can be found in the stool for reasons other than colorectal neoplasia. More recently immunological fecal occult blood (FOB) tests have been introduced, and as these are specific for human hemoglobin or its early degradation forms, they are again more likely to detect distal rather than proximal disease. These tests, which are based on a variety of methods including reverse passive hemagglutination (using chicken erythrocytes that have been coated in antihuman hemogloblin, which agglutinate in the presence of human hemoglobin) and immunochromatography, can be set to a wide range of analytical sensitivities thus varying the clinical sensitivity (13).

Colorectal Cancer Screening

83

Studies have demonstrated that such tests can be highly clinically sensitive for colorectal cancer, but any increase in sensitivity is balanced by a decrease in specificity (15). All the population-based trials that have been reported used the guaiacbased Hemoccult II test, and were carried out in Minnesota (U.S.A.) (16), Nottingham (England) (17), Funen (Denmark) (18), Bordeaux (France) (19) and Goteborg (Sweden) (20). The first was the Minnesota study where volunteers were randomized to no screening, biennial screening, or annual screening using rehydrated Hemoccult II without dietary restriction. All subjects who had a positive test underwent colonoscopy and were statistically significant, 21% and 33% reductions in colorectal cancer mortality were observed in the biennial and annual groups, respectively, after 18 years (21). It has to be appreciated, however, that 10% of all tests were positive and 38% of the annually screened group underwent colonoscopy at least once. Thus, screening with unrehydrated hemoccult resulted in a high rate of colonoscopy and the implications of this study for an unselected and nonvolunteer population are not entirely clear. It is of great interest, however, that long-term follow-up of the Minnesota study has indicated that after 18 years the incidence of colorectal cancer in the groups offered screening was signficantly less than that in the control group (22). The underlying reasons for this observation are not clear, but as the colonoscopy rate in the screened groups was so high, it is likely to be related to polypectomy. In the Nottingham study (17) approximately 150,000 unselected subjects were randomized by household. The screened group was offered biennial nonrehydrated Hemoccult II testing. Dietary restriction was not specified but if the individual returned a weakly positive test they were offered a retest with dietary restriction. This led to a much lower rate of test positivity than the Minnesota study with a 2% investigation following the first (prevalence) round and 1.2% in subsequent (incidence) rounds. Over five screening rounds only 4% of the population offered screening underwent colonoscopy. In the Nottingham study, uptake varied from round to round, but overall 60% of the group offered screening completed at least one test. Screen-detected cancers tended to be highly favorable, with 57% being diagnosed at stage A (including polyp cancers). There were, however, a substantial number of interval cancers, and indeed about 50% of the cancers arising among those who had accepted at least one invitation to be screened were not detected by the screening process. This suggests that, for the purposes of population screening, the Hemoccult II test is only about 50% sensitive. Nevertheless, when the group offered screening was compared with the control group after a median of 7.8 years of follow-up, a statistically significant 15% reduction of colorectal cancer mortality was seen, and at a median of 11 years this was maintained at 13% (23). An interesting side effect of the Nottingham screening study has been highlighted by the observation that in the control group the percentage of

84

Steele

patients presenting with favorable stage rectal cancer (Dukes’ stage A) increased from 9% in the first half of the recruitment to 28% in the second half (24). This implies that the very existence of the screening program had an effect on the control group, presumably related to a heightened awareness of the significance of rectal bleeding among the general population and at primary-care level. Another encouraging finding was related to emergency presentation of colorectal cancer. Throughout the duration of the study, there were a total of 1962 cancers identified, and there were significantly fewer emergencies in the group offered screening (25), suggesting that a policy of screening should lead to a significant reduction in the emergency workload with favorable consequences for operative mortality. A Danish study carried out on the island of Funen that was almost identical in design to the Nottingham trial in all respects (other than for the use of dietary restriction from the outset) obtained very similar results (18). In this study 61,933 individuals were randomized either to be offered biennial screening or to form a control group. The uptake was higher than that in the Nottingham study with 67% completing the first screening round and with more than 90% accepting repeated screenings. The overall positivity rate was somewhat lower, however, being 1% following the first round and dropping to 0.8% in the second round. Interestingly, the Danish group found the positivity rate to increase with subsequent rounds and by round 5 it was 1.8%. As with the Nottingham study the stage at diagnosis of screen-detected cancers was extremely favorable, with 48% at stage A and only 8% with distant spread. Again, interval cancers were relatively common and making up approximately 30% of the cancers arising in the screening group. As might be expected the mortality reduction was also similar, with a statistically significant reduction of 18% after five rounds rising to 30% after seven rounds (26). In France the results of a population-based study using nonrehydrated hemoccult have recently been published (19). In this study small geographical areas were allocated either to screening or to no screening. This involved inviting 91,199 subjects between the ages of 50 and 74 years, and no dietary restriction was employed. Uptake in the first round was 52.8% and increased slightly in subsequent rounds. Positivity was 1.2% on the first round and 1.4% on average thereafter, and the overall colorectal cancer mortality reduction was 16%. In Sweden, all the 68,308 residents of Goteborg born between 1918 and 1931 were randomized into a control group or a group offered screening using the Hemoccult II FOB test (20). In the first round, uptake was 63% and dropped to 60% in later rounds. The positivity rate was 4.4% in the first round, and screen-detected cancers were found to be at a much more favorable stage than those arising in the control group. Unfortunately, mortality data are not available from this study. Thus there are five major studies investigating the role of guaiac-based FOB testing as a screening tool; four of these are randomized, four are truly

Colorectal Cancer Screening

85

population based, and four have reported mortality data. The message from these studies is remarkably uniform and it is clear that this approach can bring about a substantial reduction in deaths from colorectal cancer. A metaanalysis using data from all five studies has indicated a 16% reduction in colorectal cancer mortality in those offered screening, going up to a 23% reduction when adjusted for compliance (27). From an ideal perspective it can be argued that the test is insensitive, compliance is poor, and there has been educational ‘‘contamination’’ of the control groups that would tend to diminish the effect of screening. Despite this, however, the statistically significant effect is a very powerful indicator that the screening process is beneficial in colorectal cancer, even if the FOB test approach is not necessarily optimal. In the United Kingdom, when the National Screening Committee was considering a recommendation on colorectal cancer screening, it was decided that a demonstration pilot of FOB testing should be carried out to ensure that the results of the randomized trials could be reproduced in the U.K. National Health Service (28). This pilot was run in two areas, one in Scotland and one in England, and a total of 478,250 individuals were invited to take part over a two-year period to simulate the first round of a biennial screening program. Uptake was 56.8%, positivity was 1.9%, and 48% of all screen-detected cancers were at Dukes’ stage A with only 1% having metastasized by the time of diagnosis (29). The results of this pilot were independently evaluated and compared with the results in the Nottingham study (30). The similarities were striking and the clear implication is that a national screening program based on FOB test screening within the United Kingdom should bring about a useful reduction in colorectal cancer mortality. As a result of this pilot, the U.K. government has now given a firm commitment to develop a comprehensive colorectal cancer screening program (31,32). FLEXIBLE SIGMOIDOSCOPY The concept of using flexible sigmoidoscopy (examination of the distal colon and rectum with a 60-cm flexible instrument) is based on the assumptions that the majority of colorectal cancers are within reach of this instrument and that a finding of a distal significant adenoma is a marker for possible proximal cancer. It has been proposed that a single flexible sigmoidoscopy at around the age of 60 with removal of all small adenomas at the time of initial examination with colonoscopy reserved for those with high-risk polyps would be an effective intervention to reduce mortality from colorectal cancer and ultimately reduce the incidence of colorectal cancer by means of polypectomy (33). This hypothesis is being tested by two multicenter randomized controlled trials of identical design, one being carried out in the United Kingdom (34) and the other in Italy (35). In the U.K. trial, men and women

86

Steele

aged 60 to 64 in fourteen centers were sent a questionnaire in the mail to ask if they would attend for flexible sigmoidoscopy screening if invited. Of 354,262 people sent this questionnaire 194,726 (55%) responded in the affirmative, and of these, 170,432 eligible subjects were randomized using a 2:1 ratio of controls to those invited for screening. The screening protocol involved a flexible sigmoidoscopy with removal of all small polyps seen at the time of sigmoidoscopy with colonoscopy reserved for those with high-risk polyps (three or more adenomas, an adenoma greater than 1 cm in diameter, a villous or severely dysplastic adenoma) or invasive cancers. Of the 57,254 individuals invited for screening 40,674 (71%) attended. It must be appreciated therefore that this study is essentially a volunteer study and the extrapolated population compliance was no more than 30%. Of those undergoing flexible sigmoidoscopy distal adenomas were found in 12.1% and distal cancer in 0.3%. In those that went to colonoscopy proximal adenomas were found in 18.8% and a proximal cancer in 0.4%. Of particular importance was the stage of diagnosis, and it was found that 62% of the cancers were Dukes’ stage A. In the SCORE trial (the Italian arm of the once only flexible sigmoidoscopy study), similar results were found. In this case 236,568 people aged between 55 and 64 were sent letters of invitation but only 56,532 (23.9%) indicated that they would be prepared to be screened, and of the 17,148 assigned to screening 9999 (58%) attended. Fifty-four individuals were found to have colorectal cancer and 54% of these were diagnosed at Dukes’ stage A. A further randomized trial as part of a study looking at prostate, lung, colorectal, and ovarian cancer screening in the United States has looked at flexible sigmoidoscopy as a screening modality (36). To date there have been no data on uptake compliance or pathology yield published from this study although it is of some interest that repeat flexible sigmoidoscopy three years after an initial examination revealed advanced adenoma or cancer in the distal colon. The authors suggest that this highlights the need for repeated flexible sigmoidoscopy rather than the once only approach advocated by the U.K. and Italian studies. Thus the randomized evidence related to flexible sigmoidoscopy screening indicates that, although flexible sigmoidoscopy is an effective means of detecting early disease and adenomas, it does tend to miss proximal disease and currently compliance rates are modest. This calls into question the use of flexible sigmoidoscopy as a population screening tool, and although the randomized trials are likely to indicate mortality reductions further work requires to be done to estimate true population compliance. COLONOSCOPY In many countries there is considerable interest in using colonoscopy as a screening tool. The advantages are obvious. It is highly accurate with a

Colorectal Cancer Screening

87

specificity of virtually 100% and a very high sensitivity, although it has to be appreciated that sensitivity is not 100% as has been demonstrated by back-toback colonoscopy studies, which show that adenomas and occasionally carcinomas can be overlooked by even experienced colonoscopists (37). In addition, a recent study comparing state-of-the art CT colography with colonoscopy suggests that the sensitivity of colonoscopy for adenomatous polyps may be as low as 87.5% (7). Nevertheless colonoscopy is currently seen as the gold standard investigation for the colon and has the advantage of allowing immediate polypectomy with the potential for preventing colorectal cancer. Unfortunately there are no randomized trials of colonoscopy as a screening instrument and conclusions must necessarily be limited. One of the most influential studies in this area was the U.S. National Polyp Study that compared a cohort of subjects undergoing periodic colonoscopy with historical controls (38). In this study 1418 patients who had undergone total colonoscopy and removal of adenomas underwent subsequent colonoscopy during an average follow-up period of six years and the incidence of colorectal cancer in this group was compared with that in three reference groups including two cohorts in which polyps had not been removed. Ninety-seven percent of these subjects were followed up for a total of 8401 person years, and the majority (80%) had one or more follow-up colonoscopies. During this time five asymptomatic early-stage colorectal cancers were detected by colonoscopy and no symptomatic cancers were detected. When compared with the reference group this represented a much lower rate of diagnosis of colorectal cancer than would have been expected, and the conclusions were that colonoscopic surveillance in adenoma patients reduces the incidence of and subsequent mortality from colorectal cancer. Although a landmark study, the conclusions must be interpreted with caution as the comparison group was not derived from the same population as the cases and this is likely to have led to an overestimate of the efficacy of colonoscopy. In addition, it is difficult to extrapolate from polyp surveillance to screening asymptomatic populations. The most important study in the literature in terms of estimating the efficacy of screening colonoscopy is a case–control study conducted among U.S. military veterans (39). The study group consisted of 4411 veterans dying of colorectal cancer between 1998 and 1992. The control group was derived from living control patients and dead control patients without colorectal cancer matched by age, sex, and race to each case. Using this study design it was found that colonoscopy reduced death rates from colorectal cancer with an odds ratio of 0.41 (range 0.33–0.50). Further, comparison with the living control group revealed that the protective effects lasted for five years and that polypectomy was particularly protective. Similar results were found when the dead control group was employed. Again the study is far from perfect, particularly as the reasons for colonoscopy in the study group were varied and included investigation of symptomatic patients.

88

Steele

There are, of course, abundant uncontrolled data on screening colonoscopy and perhaps the most useful study was carried out in 13 veterans affairs medical centers to determine the utility of colonoscopy in detecting colorectal neoplasia in asymptomatic individuals aged 50 to 75 (40). Of 17,732 potential subjects 3196 were included and 3121 underwent total colonoscopy. The mean age was 62.9 years and 96.8% were males. An adenoma of at least 10 mm diameter was detected in 7.9% and invasive cancer in 1%. Of 1765 subjects with no adenomas distal to the splenic flexure 48% had proximal adenomas or cancers. It can be concluded from this study that if colonoscopy was used as a screening tool in men aged between 50 and 75 the uptake would only be 20% and only 1% of colonoscopies would detect colorectal cancer. Thus, although colonoscopy is widely used to screen asymptomatic individuals on demand, it would seem very unlikely that it could ever be used as an effective population screening modality. This question will be considered further in the section on cost-effectiveness. RADIOLOGY There are no reliable data to support barium enema as a screening tool but the new technology of CT colography (virtual colonoscopy) shows distinct promise. The most exciting results to date come from the National Naval Medical Center in Bethesda where 1233 asymptomatic individuals with a mean age of 57.8 years underwent CT colography and colonoscopy on the same day (7). The sensitivities and the specificities of the two investigations were calculated on the basis of a final unblinded colonoscopy as the reference standard. It was found that the sensitivity of CT colography was 93.8% for adenomas of 10 mm in diameter or more, whereas the sensitivity of standard colonoscopy for the same lesions was only 87.5%. The specificity of CT colography was 96% for adenomas of 10 mm or more. It would seem therefore that CT colography has the appropriate sensitivity and specificity characteristics for a colorectal neoplasia screening tool but there are as yet no data with which to assess its performance or costeffectiveness in population screening. COMPARATIVE STUDIES There are very few studies that directly compare different screening methods and of those that exist all address the relative merits of FOB testing and flexible sigmoidoscopy. The Nottingham group carried out a randomized study comparing FOB testing with a combination of flexible sigmoidoscopy and FOB testing (41). The neoplasia yield in those undergoing the combined approach was four times greater than in those doing the FOB test alone, but while compliance with FOB testing was 50% in those offered both tests only 20% went on to have flexible sigmoidoscopy. In Sweden a group of

Colorectal Cancer Screening

89

6367 individuals aged between 55 and 56 were randomized to be offered screening with Hemoccult II or flexible sigmoidoscopy (42). Compliance with the FOB test screening was 59% and with flexible sigmoidoscopy 49%. Of those who attended for FOBT screening 4% had a positive test and 13% had a neoplastic lesion greater than 1 cm in the rectum or sigmoid colon; the corresponding rate in the flexible sigmoidoscopy group was 2.3%. Overall, 10 individuals were diagnosed with a neoplastic lesion in the FOBT group compared with 31 in the flexible sigmoidoscopy group. In the Norwegian Colorectal Cancer Prevention (NORCCAP) Screening Study (43) 20,780 individuals aged between 50 and 64 were randomized to be invited for flexible sigmoidoscopy only or a combination of flexible sigmoidoscopy and FOB testing. Compliance was 65% and overall 41 (0.3%) cases of colorectal cancer and 2208 (17%) adenomas were found. The diagnostic yields in the two groups were identical in terms of colorectal cancer or high-risk adenomas indicating that there was very little benefit in adding a FOB test to a screening flexible sigmoidoscopy. These studies indicate that while compliance with flexible sigmoidoscopy tends to be less than that for FOB testing, the sensitivity of flexible sigmoidoscopy is much higher. On the other hand it has to be remembered that all the randomized studies of FOB test screening were based on repeated testing, and in a nonrandomized study from Denmark comparing once only flexible sigmoidoscopy plus FOB testing with FOB testing alone over 16 years found that the FOB test screening program had a diagnostic yield at least as high as a single flexible sigmoidoscopy (44). To date, the evidence relating to the relative merits of an FOB test program and once only flexible sigmoidoscopy is not of particularly high quality, and this question can only be fully resolved by a randomized trial directly comparing these two modalities. HARM CAUSED BY SCREENING Screening comes at a cost, and the cost is not only financial but can also be measured in terms of morbidity and mortality. The question of financial cost is dealt with in the section on economics, but the other two issues are no less important. While performing an FOB test is unlikely to occasion physical morbidity and flexible sigmoidoscopy is very safe, the possibility of complications of the subsequent colonoscopy for those with a positive test and of surgery for those who are diagnosed with cancer must not be overlooked. In addition, false-negative results caused by the low sensitivity of the FOB test and the propensity of sigmoidoscopy to miss proximal cancers might falsely reassure individuals and lead to delayed cancer diagnosis and poorer outcome (‘‘certificate of health effect’’). The Nottingham group has addressed these issues by examining the investigation and treatment-related mortality and the stage at presentation

90

Steele

of the interval cancers (45). There were no colonoscopy-related deaths and five deaths after surgery for screen-detected cancers; this represents a 2% operative mortality at a time when mortality after elective colorectal cancer surgery in the United Kingdom was estimated to be around 5% by a large national audit (46). Furthermore, the stage distribution of the interval cancers (cancers that were diagnosed after a negative FOB test or colonoscopy) was similar to that of the cancers in the control group, and the survival was significantly better than that for the control cancers—findings that are not consistent with an appreciable certificate of health effect. Nevertheless, these concerns have been highlighted by the finding that all-cause mortality is not affected by colorectal cancer screening and indeed, in the Nottingham study it was found to be increased in the group offered screening (47). However, colorectal cancer only accounts for around 2% of all deaths, and a 15% reduction in disease-specific death rate could only be expected to reduce all-cause mortality by 0.3%. To demonstrate a difference of this size with statistical power would require a trial too big to be feasible. Furthermore, unlike the difference in disease-specific mortality, the excess of all-cause deaths observed in the group offered screening was not statistically significant and therefore likely to represent a chance finding. Another important adverse effect of screening relates to psychological morbidity. In colorectal cancer screening there has been relatively little work done in this field, but there are two studies of note. In the Swedish randomized study of FOB test screening a questionnaire was administered to 2932 participants and it was found that 4.7% experienced worry from the invitation letter sufficient to influence daily life, and that this increased to 15% after a positive test (48). Worry decreased rapidly after the screening process was over, however, and at one year 96% declared that they had appreciated the opportunity to be screened. As part of the Nottingham trial a similar study was carried out using validated measures of psychiatric morbidity, and this was found to be highest in those with a positive test result, but in those with false-positive tests it fell the day after colonoscopy and remained low one month later (49). Thus it appears that the screening process does cause anxiety, but that is short lived. Finally, there is the issue of overinvestigation in the group with falsepositive tests. Despite the fact that the guaiac tests are very insensitive for upper gastrointestinal bleeding (q.v.), there is concern that ignoring a positive FOB test result in the face of a normal colonoscopy might be seen as negligent if significant upper gastrointestinal pathology is missed. To try to rationalize this fear the Nottingham group looked at a cohort of 283 FOB positive cases who had no neoplastic disease on colonoscopy (50). Fourteen (5%) of these underwent upper gastrointestinal endoscopy because of symptoms, and one was found to have gastric carcinoma. The rest, who were asymptomatic, were followed up for a median period of five years and only one, who had persistent symptoms after a previous partial gastrectomy, was subsequently

Colorectal Cancer Screening

91

diagnosed as having gastric cancer. Thus, the evidence supports a strategy of reassuring the majority of those who have a negative colonoscopy and reserving upper gastrointestinal endoscopy for those with relevant symptoms. ECONOMICS OF SCREENING Before deciding on a screening strategy, it is important to have information regarding cost-effectiveness. Unfortunately, however, there are no published data from the randomized controlled trials on the cost-effectiveness of colorectal cancer screening and the only approach is to use information provided by health economic models. In a recent study, 25 papers that were potentially relevant were identified although eight had to be excluded because the results were based on cost per cancer detected rather than cost per life year saved (51). Of the remaining papers the strategies assessed were yearly FOB testing, sigmoidoscopy every five years or colonoscopy every 10 years, or a combination of these strategies. It has to be stressed that there is no reliable information on biennial FOB testing or once only flexible sigmoidoscopy. This is unfortunate as these appear to be the most likely candidates for an effective population screening program. Overall, the models did not help to distinguish between the three approaches as the estimates of cost per life year saved were highly variable and overlapping (Fig. 3). The underlying reason for this huge variation is basically the lack of high-quality data on the efficacy of the different screening methods. As detailed earlier in this chapter, although we know a lot about the effect of FOB testing on colorectal cancer mortality from population-based randomized trials, we do not have the same level of information on flexible sigmoidoscopy and colonoscopy. Currently there are three randomized

Figure 3 Cost-effectiveness of fecal occult blood, sigmoidoscopy, and colonoscopy screening compared with no screening from 17 health economic papers. Abbreviation: FOB, fecal occult blood.

92

Steele

trials of flexible sigmoidoscopy but none have reported mortality data. As far as colonoscopy is concerned most of the models utilize the U.S. National Polyp Study (38), which is severely limited in this context owing to the selection criteria and the historical controls. The single case–control study that evaluated the efficacy of lower gastrointestinal endoscopy (39) was rarely used in the health economic models evaluated. To compound these shortcomings none of the models have followed international recommendations with regard to the uncertainty of data (52). This is usually evaluated using sensitivity analysis where one parameter is varied over a specific range, and the effect this has on the model is expressed as a range of incremental cost-effectiveness ratios. This does not allow for the distribution of the statistical uncertainty and assumes that any value within the 95% confidence intervals is equally possible. This is inaccurate as it does not take into account all the uncertainty that can exist. In order to overcome these problems specific techniques were used (Monte Carlo simulations of a Markov model). In this approach it was assumed that individuals were screened from 45 to 75 years of age and yearly FOB testing, sigmoidoscopy every five years, and colonoscopy every 10 years were compared with one another. The assumptions made are given in Table 3. The first simulation compared yearly FOB testing with no screening and suggested that FOB screening costs 48,900 per life year saved. This does not take into account the uncertainty but if this is incorporated it is still 95% certain that an annual FOB test is cost-effective provided society is willing to pay 430,000 per life year saved (Fig. 4). As this is below the threshold that most countries are prepared to pay it is possible to say with a high degree of certainty that FOB test screening is cost-effective. On the basis of the result of the first model, sigmoidoscopy and colonoscopy were compared with FOB test screening. Sigmoidoscopy was estimated to cost 48,000 per life year saved, which is in fact cheaper than FOB test screening, but when uncertainty was incorporated into the model it was not even possible to be 80% certain that sigmoidoscopy is costeffective compared with FOB test screening no matter how much is paid for each life year saved (Fig. 5). This uncertainty is caused by the lack of data on mortality reduction brought about by flexible sigmoidoscopy and will be resolved when the results of the randomized trials are available. As far as colonoscopy is concerned when 10 yearly examination was compared with annual FOB testing it was estimated that each life year saved would cost 428,500 and when uncertainty was taken into account it became clear that to be 95% certain of cost-effectiveness it would be necessary to pay 490,000 per life year saved (Fig. 6). This is well above the threshold that most countries are willing to pay and again reflects the lack of good data on efficacy. Cost-effectiveness studies are dependent on efficacy data and are currently of limited value in the area of colorectal cancer screening for the reasons outlined earlier. What can be said at present is that FOB test

Colorectal Cancer Screening

93

Table 3 Assumptions Made in the Models of Fecal Occult Blood, Sigmoidoscopy, and Colonoscopy Screening to Prevent Colorectal Cancer Mortality

Variable Efficacy of FOB (28) Efficacy of FS (24) Efficacy of colonoscopy (24) Compliance Referral for colonoscopy with FOB Referral for colonoscopy with FS No screen colonoscopy rate Colorectal cancer care costs Colonoscopy cost FS cost FOB cost Invitation/administration costs

Uncertainty assumed

Value

Distribution given to this parameter in the model

RR ¼ 0.77 RR ¼ 0.67 RR ¼ 0.45

95% CI 0.57–0.89 95% CI 0.54–0.82 95% CI 0.30–0.66

Log normal Log normal Log normal

50% 5%

Data for 100 cases Data for 100 cases

Beta Beta

5%

Data for 100 cases

Beta

0.2%

No

NA

49024

No

NA

4350 480 42.80 43.00

No No No No

NA NA NA NA

Abbreviations: RR, relative risk of colon cancer compared with no screen; FOB, unhydrated fecal occult blood test; FS, flexible sigmoidoscopy; CI, confidence interval; NA, not applicable. Source: From Ref. 51.

screening is cost-effective, flexible sigmoidoscopy screening might be costeffective but we have to await further data, and colonoscopy is unlikely to be cost-effective for population screening. The uncertainty related to sigmoidoscopy and colonoscopy is extremely high and while this will be resolved for flexible sigmoidoscopy in the future, this seems unlikely for colonoscopy as there are no population-based randomized trials planned. NOVEL APPROACHES TO SCREENING Research into improving screening methods has focused largely on stool tests and these can be subdivided according to whether they involve DNA. Turning to the non-DNA-based stool testing first, a number of proteins have been studied including transferrin, albumin, and a-1 antitrypsin. Immunological detection of transferrin has shown high sensitivity and specificity when used along with immunological estimation of hemoglobin levels (53) but has never been developed. Albumin, which is known to be increased in the stool of patients with colorectal neoplasia, can be estimated

94

Steele

Figure 4 Cost-effectiveness acceptability curve of annual fecal occult blood screening compared with no screening. Threshold of willingness to pay/life year saved that has a 50% chance of being cost-effective (dotted lines), and threshold of willingness to pay/life year saved that has a 95% chance of being cost-effective (dotted and dashed lines).

in stool but has a low sensitivity owing to its tendency to be degraded by colonic bacteria (54). a-1 antitrypsin, which inhibits proteolytic enzymes produced by both neoplastic and inflammatory lesions, is sensitive but not specific for neoplasia (55). There has been considerable interest in a calciumbinding protein known as calprotectin that is found in neutrophils, and a fecal calprotectin test has been developed that has been shown to be more sensitive in the detection of cancers and adenomas than FOB testing (56). Unfortunately, however, the specificity is highly variable from study to study, presumably related to the fact that it tends to be raised in inflammatory bowel disease. Another approach is stool cytology and this can be aided

Colorectal Cancer Screening

95

1.0 0.9 0.8

Probability cost-effective

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 1.0

5,000.9 10,000.8 15,000.7 20,000.6 25,000.5 30,000.4 35,000.3 40,000.2 45,000.1 50,000.0

Willingness to pay ( )

Figure 5 Cost-effectiveness of sigmoidoscopy every five years compared with annual fecal occult blood (FOB) screening. Even at an infinite willingness to pay per life year saved there is less than 80% certainty that sigmoidoscopy is more cost-effective than annual FOB screening.

by the immunohistochemical detection of MCM2, a protein expressed strongly by neoplastic epithelium (57). The rapidly expanding body of information on genetic mutations in colorectal neoplasia has sparked considerable interest in developing tests that can detect these genetic abnormalities in DNA extracted from the stool. This has met with varying degrees of success. The basic principle involves extraction of DNA from the stool and amplification of abnormal DNA using a polymerase chain reaction (PCR). The technical difficulties in achieving this are considerable and include sample collection, DNA extraction, removal of PCR inhibitors, and the choice of primers for the PCR (58). DNA extraction is now quite feasible but the main problem is related to the heterogeneity of genetic mutations. Not only do colorectal cancers and adenomas display mutations in a wide and nonuniform range of genes, but,

96

Steele

Figure 6 Cost-effectiveness of colonoscopy every 10 years compared with annual fecal occult blood screening. Threshold of willingness to pay/life year saved that has a 50% chance of being cost-effective (dotted lines), and threshold of willingness to pay/life year saved that has a 95% chance of being cost-effective (dotted and dashed lines).

in addition, the position of the mutation within the gene varies widely. Thus, in order to use this approach for screening it is essential to make a best guess as to the genes likely to be affected and the codons within these genes that are most frequently altered. The genes most commonly studied are K-ras, APC, and P53. In addition, markers of microsatellite instability (reflecting mutations in DNA mismatch repair genes) have been sought, and the most useful marker to date appears to be the mononucleotide BAT26 (58). In addition to looking for mutations DNA can also be utilized in a nonspecific way. Under normal circumstances cells shed from the colonic epithelium undergo apoptosis and nuclear endonucleases are activated giving rise to small fragments of DNA. Neoplasms on the other hand tend to shed long fragments of DNA

Colorectal Cancer Screening

97

so that an excess of long DNA in the stool can be used as a marker for neoplasia (59). Studies utilizing the detection in stool of single genetic mutations are insufficiently sensitive, and this has led to the development of panels of genetic abnormalities. Using this approach three groups have achieved high sensitivities for colorectal cancer and adenomas (59–61). In particular, Ahlquist’s group, using a combination of mutations in K-ras, APC and P53, BAT26 and long DNA achieved a sensitivity of 91% for cancer and 82% for adenoma with an initial specificity of 93% increasing to 100% if the K-ras marker was excluded from the panel. Molecular stool testing, although exciting, is still at a very early stage and is currently far too labor intensive to be considered seriously as a population screening tool. However, if the initial promise is sustained then automated analysis of inclusive panels of a wide range of abnormalities could be envisaged. CONCLUSIONS There is unequivocal, high-quality evidence that colorectal cancer screening reduces disease-specific mortality, and although this is strictly true only for guaiac-based FOB test screening, it is widely assumed that other, more sensitive modalities must be equally, if not more effective. This may be true as far as an individual is concerned, but given that the other methods have not been tested within the rigor of a randomized trial, it is impossible to be sure that the cost–benefit ratios associated with sigmoidoscopy, colonoscopy, or more sensitive fecal tests will be favorable. For flexible sigmoidoscopy, this is likely to change in the near future, but not for colonoscopy or for any other test, as the appropriate trials are not in place. However, the purpose of screening must be fully understood. The aim of population screening is to reduce the burden of disease on a community; this requires a test that is acceptable and affordable, and currently guaiacbased FOB testing is the only proven option. If, on the other hand, the aim of screening is to inform an individual, for whatever reason, regarding their disease status, then the sensitivity and specificity of the test are crucial. Here, colonoscopy, real or virtual, must be the test of choice. REFERENCES 1. Cancer Research Campaign. Cancer in the European Community. Factsheet 5.2 1992. 2. Slaney G, et al. Cancer of the large bowel. Clinical Cancer Monograph. Vol. 4. Macmillan Press, 1991. 3. Black RJ, Sharp L, Kendrick SW. Trends in Cancer Survival in Scotland 1968–1990. Edinburgh: ISD Publication, 1993.

98

Steele

4. Ahmed S, Leslie A, Thaha M, Carey FA, Steele RJC. Lower gastrointestinal symptoms do not discriminate for colorectal neoplasia in a faecal occult blood scree-positive population. Br J Surg 2005; 92:478–481. 5. Wilson JM, Jungner F. Principles and practice of screening for disease. In: Public Health Papers No. 34. Geneva: WHO, 1968. 6. Rex DK, Rahmani EY, Haseman JA, Lemmel GT, Kaster S, Buckley JS. Relative sensitivity of colonoscopy and barium enema for detection of colorectal cancer in clinical practice. Gastroenterology 1997; 112:17–23. 7. Pickhardt PJ, Choi JR, Hwang I, et al. Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med 2003; 349:2191–2200. 8. McFarlane JK, Ryall RDH, Heald RJ. Mesorectal excision for rectal cancer. Lancet 1993; 341:457–460. 9. Whelan RL, Young-Fadouk TM. Should carcinoma of the colon be treated laparoscopically? Surg Endosc 2004; 18:857–862. 10. SIGN (Scottish Intercollegiate Guidelines Network) Guidelines on the Management of Colorectal Cancer. SIGN, Edinburgh March, 2003. 11. Leslie A, Carey FA, Pratt NR, Steele RJC. The colorectal adenoma-carcinoma sequence. Br J Surg 2002; 89:845–860. 12. Cancer Research UK. Factsheets on Cancer. Large Bowel—UK, 2003. 13. Young GP, Macrae FA, St. John DJB. Clinical methods for early detection: basis, use, and evaluation. In: Young GP, Rozen P, Levin B, eds. Prevention and Early Detection of Colorectal Cancer. WB Saunders, 1996; 241–270. 14. Pignone M, Campbell MK, Carr C, Phillips C. Meta-analysis of dietary restriction during fecal occult blood testing. Eff Clin Pract 2001; 4:150–156. 15. Robinson MH, Marks CG, Farrands PA, Bostock K, Hardcastle JD. Screening for colorectal cancer with an immunological faecal occult blood test: 2-year follow-up. Br J Surg 1996; 83:500–501. 16. Mandel JS, Bond JH, Church JR, et al. Reducing mortality from colorectal cancer by screening for faecal occult blood. N Engl J Med 1993; 328:1365–1371. 17. Hardcastle JD, Chamberlain JO, Robinson MHE, et al. Randomised controlled trial of faecal occult blood screening for colorectal cancer. Lancet 1996; 348: 1472–1477. 18. Kronborg O, Fenger C, Olsen J, Jorgensen OD, Sondergaard O. Randomized study of screening for colorectal cancer with faecal occult blood test. Lancet 1996; 348:1467–1471. 19. Faivre J, Dancourt V, Lejeune C, et al. Reduction in colorectal cancer mortality by fecal occult blood screening in a French controlled study. Gastroenterology 2004; 126:1674–1680. 20. Kewenter J, Brevinge H, Engaras B, Haglin E, Ahren C. Results of screening, rescreening, and follow-up in a prospect randomized study for detection of colorectal cancer by fecal occult blood testing. Results for 68,308 subjects. Scand J Gastroenterol 1994; 29:468–473. 21. Mandel JS, Church TR, Ederer F, Bond JH. Colorectal cancer mortality: effectiveness of biennial screening for fecal occult blood. J Natl Cancer Inst 1999; 91: 434–437.

Colorectal Cancer Screening

99

22. Mandel JS, Church TR, Bond JH, et al. The effect of fecal occult-blood screening on the incidence of colorectal cancer. N Engl J Med 2000; 343:1603–1607. 23. Scholefield JH, Moss S, Sufi F, Mangham CM, Hardcastle JD. Effect of faecal occult blood screening on mortality from colorectal cancer: results from a randomised controlled trial. Gut 2002; 50:840–844. 24. Robinson MHE, Thomas WM, Hardcastle JD, Chamberlain J, Mangham CM. Change towards earlier stage at presentation of colorectal cancer. Br J Surg 1993; 80:1610–1612. 25. Scholefield JH, Robinson MH, Mangham CM, Hardcastle JD. Screening for colorectal cancer reduces emergency admissions. Eur J Surg Oncol 1998; 24: 47–50. 26. Jorgensen OD, Krongborg O, Fenger C. A randomised study of screening for colorectal cancer using faecal occult blood testing: results after 13 years and seven biennial screening rounds. Gut 2002; 50:29–32. 27. Towler B, Irwig L, Glasziou P, Kewenter J, Weller D, Silagy C. A systematic review of the effects of screening for colorectal cancer using the faecal occult blood test, Hemoccult. Br Med J 1998; 317:559–565. 28. Steele RJC, Parker R, Patnick J, et al. A demonstration pilot for colorectal cancer screening in the United Kingdom: a new concept in the introduction of health care strategies. J Med Screen 2001; 8:197–202. 29. Steele RJC for the UK Colorectal Cancer Screening Pilot Group. Results of the first round of a demonstration pilot of screening for colorectal cancer in the United Kingdom. Br Med J 2004; 329:133–135. 30. Evaluation of the UK colorectal screening pilot. A report for the UK Department of Health. http://www.cancerscreening.nhs.uk/colorectal/finalreport.pdf Department of Health, June 2003. 31. Department of Health Press Release 2003/0047. http://www.info.doh.gov.uk/ doh/intpress.nsf/page/2003–0047. 32. Cancer in Scotland. Action for Change. Bowel cancer framework for Scotland. NHS Scotland. Scottish Executive 2004. 33. Atkin WS, Edwards R, Wardle J, et al. Design of a multicentre randomized trial to evaluate flexible sigmoidoscopy in colorectal cancer screening. J Med Screen 2001; 8:137–144. 34. UK Flexible Sigmoidoscopy Screening Trial Investigators. Single flexible sigmoidoscopy screening to prevent colorectal cancer: baseline findings of a UK multicentre randomized trial. Lancet 2002; 359:1291–1300. 35. Segnan N, Senore C, Andreoni B, et al. SCORE working group. Baseline findings of the Italian multicenter randomized controlled trial of ‘‘once-only sigmoidoscopy’’–SCORE. J Natl Canc Inst 2002; 94:1763–1772. 36. Schoen RE, Pinsky PF, Weissfeld JL, et al. Results of repeat sigmoidoscopy 3 years after a negative examination. JAMA 2003; 290:41–48. 37. Rex DK, Cutler CS, Lemmel GT, et al. Colonoscopic miss rates of adenomas determined by back-to-back colonoscopies. Gastroenterology 1997; 112:24–28. 38. Winawer SJ, Zauber AG, Ho MN, et al. Prevention of colorectal cancer by colonoscopic polypectomy. The National Polyp Study Working Group. N Engl J Med 1993; 329:1977–1981.

100

Steele

39. Muller AD, Sonnenberg A. Protection by endoscopy against death from colorectal cancer. A case–control study among Veterans. Arch Intern Med 1995; 155:1741–1748. 40. Lieberman DA, Weiss DG, Bond JH, et al. Use of colonoscopy to screen asymptomatic adults for colorectal cancer. Veterans Affairs Cooperative Study Group 380. N Engl J Med 2000; 343:162–168. 41. Berry DP, Clarke P, Hardcastle JD, Vellacott KD. Randomized trial of the addition of flexible sigmoidscopy to faecal occult blood testing for colorectal neoplasia population screening. Br J Surg 1997; 84:1274–1276. 42. Brevinge H, Lindholm E, Buntzen S, Kewenter J. Screening for colorectal neoplasia with faecal occult blood testing compared with flexible sigmoidoscopy directly in a 55 years’ old population. Int J Colorectal Dis 1997; 12:291–295. 43. Gondal G, Grotmol T, Hofstad B, Bretthauer M, Eide TJ, Hoff G. The Norwegian Colorectal Cancer Prevention (NORCCAP) screening study: baseline findings and implementations for clinical work-up in age groups 50–64 years. Scand J Gastroenterol 2003; 38:635–642. 44. Rasmussen M, Fenger C, Kronborg O. Diagnostic yield in a biennial Haemoccult-II screening programme compared to a once-only screening with flexible sigmoidoscopy and Haemoccult-II. Scan J Gastroenterol 2003; 38: 114–118. 45. Robinson MHE, Hardcastle JD, Moss SM, et al. The risks of screening: data from the Nottingham randomised controlled trial of faecal occult blood screening for colorectal cancer. Gut 1999; 45:588–592. 46. Mella J, Biffin A, Radcliffe AG, Stamatakis JD, Steele RJC. Population-based audit of colorectal cancer management in two UK health regions. Br J Surg 1997; 84:1731–1736. 47. Black WC, Haggstrom DA, Welch HG. All-cause mortality in randomised trials of cancer screening. JNCI 2002; 94:167–173. 48. Lindholm E, Berglund B, Kewenter J, Halind E. Worry associated with screening for colorectal carcinomas. Scand J Gastroenterol 1997; 32:238–245. 49. Parker MA, Robinson MH, Scholefield JH, Hardcastle JD. Psychiatric morbidity and screening for colorectal cancer. J Med Screen 2002; 9:7–10. 50. Thomas WM, Hardcastle JD. Role of upper gastrointestinal investigations in a screening study for colorectal neoplasia. Gut 1990; 31:1294–1297. 51. Steele RJC, Gnauck R, Hrcka R, et al. ESGE/UEGF Colorectal Cancer– Public Awareness Campaign The Public/Professional Interface Workshop Oslo, Norway, June 20–22, 2003. Methods and Economic Considerations: Group 1 Report. Endoscopy 2004; 36:349–353. 52. Weinstein MC, Siegel JE, Gold MR, Kanlet MS, Russell LB. Recommendations of the Panel on cost-effectiveness in Health and Medicine. JAMA 1996; 276:1253–1258. 53. Miyoshi H, Ohshiba S, Asada S, Hirata I, Uchida K. Immunological determination of fecal haemoglobin and transferrin levels: a comparison with other fecal occult blood tests. Am J Gastroenterol 1992; 87:67–73. 54. Saitoh O, Matsumoto H, Sugimori K, et al. Intestinal protein loss and bleeding assessed by fecal hemoglobin, transferrin, albumin and alpha-1-antitrypsin levels in patients with colorectal diseases. Digestion 1995; 56:67–75.

Colorectal Cancer Screening

101

55. Moran A, Robinson M, Lawson N, Stanley J, Jones AF, Hardcastle JD. Fecal alpha 1-antitrypsin detection of colorectal neoplasia. An evaluation using HemoQuant. Dig Dis Sci 1995; 40:2522–2525. 56. Tibble J, Sigthorsson G, Foster R, Sherwood R, Fagerhol M, Bjarnason I. Faecal calprotectin and faecal occult blood tests in the diagnosis of colorectal carcinoma and adenoma. Gut 2001; 49:402–408. 57. Davies RJ, Freeman A, Morris LS, et al. Analysis of minichromosome maintenance proteins as a novel method for detection of colorectal cancer in stool. Lancet 2002; 359:1917–1919. 58. Mak T, Lalloo F, Evans DGR, Hill J. Molecular stool screening for colorectal cancer. Br J Surg 2004; 91:790–800. 59. Ahlquist DA, Skoletsky JE, Boynton KA, et al. Colorectal cancer screening by detection of altered DNA in stool: feasibility of a multitarget assay panel. Gastroenterology 2000; 119:1219–1227. 60. Rengucci C, Maiolo P, Saragoni L, Zoli W, Amadori D, Calistri D. Multiple detection of genetic alterations in tumors and stool. Clin Cancer Res 2001; 7:590–593. 61. Dong SM, Traverso G, Johnson C, et al. Detecting colorectal cancer in stool with the use of multiple genetic targets. J Natl Cancer Inst 2001; 93:858–865.

4 Pathology Christian Wittekind Institut fu¨r Pathologie des Universita¨tsklinikums Leipzig, Leipzig, Germany

INTRODUCTION Colorectal cancer pathology covers a wide spectrum of aspects beginning with etiology and ranging to metastasis and causes of death of colorectal cancer patients. In this chapter, only those areas of pathology that are relevant for treatment decisions, analysis of treatment results, and quality management are covered. Aspects of diagnosis and differential diagnosis, epidemiology, genetics, causal and formal pathogenesis, which are dealt with in other chapters of this book, are not included in this chapter. Because about 95% of all malignant tumors of the colon and rectum are carcinomas, predominantly the pathology of carcinomas is discussed. At the end a short overview of the other, very uncommon, malignant tumors (endocrine, mesenchymal, and lymphoid) is presented. DEFINITION Colorectal cancer comprises all malignant tumors of the colorectum, the most frequent of which is colorectal adenocarcinoma. In contrast to the stomach or small intestine, a neoplasm of the colon and rectum has metastatic potential only after invasion of at least the submucosa. For the colorectum in the biological and clinical sense, carcinoma is present only after the lamina submucosa has been invaded. This definition has been adopted by the World Health Organization (WHO) classification (1). Between intraepithelial neoplasia (formerly: dysplasia) and an invasive carcinoma as defined earlier, we find an intermediate step of malignant 103

104

Wittekind

Table 1 Differences in Classification Between Carcinoma In Situ and Invasive Carcinomas of the Colorectum T category Tis

T1

Tumor entities

ICD-O M code

Severe dysplasia High-grade dysplasia Intraepithelial neoplasia Intramucosal carcinoma Invasive carcinoma (¼invasion of submucosa)

8140/2

8140/3

Abbreviation: ICD-O, International Classification of Disease-Oncology. Source: From Refs. 1, 6.

progression, namely, a neoplastic lesion that shows invasive growth into the lamina propria mucosae or between the fibers of the muscularis mucosae, but does not invade the submucosa. Unfortunately, outside the United Kingdom and the Germanspeaking countries, the term ‘‘carcinoma’’ is not used uniformly. Thus, in any cancer statistics and in any report of treatment results, one has to make sure whether the data relate to invasive carcinoma only or include highgrade intraepithelial neoplasia (formerly: high-grade dysplasia). Differences in definitions and nomenclature are summarized in Table 1. Another classification of flat and depressed types of early colorectal cancer has recently been published in Japan (2). However, not all of the Japanese definitions fulfill the Western criteria of invasive colorectal carcinoma. SITE DISTRIBUTION The pathologies of carcinoma of the colon and rectum are essentially the same, although there are differences in epidemiology and etiology, which has led some to speculate about the existence of two completely different tumor entities. In the literature, the definitions of colon and rectum vary. This renders comparisons difficult and explains some differences among data. According to the updated International Documentation System for Colorectal Cancer (IDS for CRC) (3,4) carcinomas with a lower border of the tumor 16 cm or less from the anal verge (measured by a rigid rectosigmoidoscope) are classified as rectal carcinomas. The rectum is further subdivided in thirds: upper third 12 to 16 cm; middle third 6 to < 12 cm; and lower third: <6 cm. GROSS MORPHOLOGY Colorectal carcinomas are grossly differently dependent on their anatomic extent. In situ lesions can be polypoid or flat. Early carcinomas, i.e., tumors limited to the submucosa, are mostly polypoid, pedunculated, semipedunculated, or sessile. Sometimes flat lesions are observed with no or only slight elevation

Pathology

105

(not more than twice the height of the mucosa). In such flat lesions, a slight central depression may also be present. Advanced carcinomas (invading beyond the submucosa) are four types, similar to the Borrmann categories of gastric carcinoma:





Polypoid (protuberant) Ulcerated, with sharply demarcated margins Ulcerated without definite borders Diffusely infiltrating

In contrast to gastric carcinomas, the latter two types are uncommon. The most common type is the ulcerated type with sharply demarcated margins. HISTOMORPHOLOGY Histological Typing Tumors are classified according to the present WHO classification (1), which is shown in Table 2. Adenocarcinomas and mucinous adenocarcinomas Table 2 Histological Typing and Grading: World Health Organization Classification Histological type

ICD-O code

Adenocarcinoma Mucinous adenocarcinoma Signet-ring cell carcinoma Squamous cell carcinoma Adenosquamous carcinoma Medullary carcinoma

8140/3 8480/3

Undifferentiated carcinoma

8020/3

8490/3 8070/3 8560/3 8510/3

Definition Glandular epithelium More than 50% extracellular mucin More than 50% signet ring cells Exclusively squamous differentiation Adeno- plus squamous cell carcinoma Sheets of malignant cells (not graded), abundant cytoplasm, and prominent infiltration by intraepithelial lymphocytes No glandular structure or other features to indicate definite differentiation

Grading Low/high system (G1–G4) (L/H) 1–3 1–3

L/H L/H

3

H

1–3

L/H

1–3

L/H

4

H

Note: The small cell carcinoma mentioned in the World Health Organization classification is now classified as poorly differentiated endocrine carcinoma. Abbreviation: ICD-O, International Classification of Disease-Oncology. Source: From Ref. 1.

106

Wittekind

(sometimes still termed mucoid or colloid adenocarcinoma) account for 90% to 95% of carcinomas; all other types are rare. Mucinous adenocarcinomas are observed more frequently in the colon (15%) than in the rectum (10%). Some adenocarcinomas display areas of mucus production, scattered Paneth cells, endocrine cells, or small foci of squamous differentiation. Signet-ring cells may also be present in mucinous adenocarcinomas. For classification purposes, it has to be considered that more than 50% of a carcinoma should comprise signet-ring cells or mucin before it is classified as a signet-ring cell or mucinous carcinoma. The very uncommon undifferentiated carcinoma should be distinguished from poorly differentiated adenocarcinoma, poorly differentiated endocrine carcinoma (small cell carcinoma), malignant lymphoma, and leukemia infiltrates by use of mucin stains and especially immunohistochemistry. Extremely uncommon carcinomas not listed in the WHO classification, and reported in only a few cases, include microglandular goblet cell carcinoma, clear cell carcinoma, adenosquamous carcinoma, spindle cell and metaplastic carcinoma (carcinosarcoma), giant cell carcinoma, choriocarcinoma, carcinomas arising in endometriosis, melanotic adenocarcinoma, and Paneth cell rich papillary adenocarcinoma (5). Histological Grading Histopathological grading of tumors is performed to provide some indication of their aggressiveness, which relates to prognosis and/or choice of treatment. The traditional system of grading also used by the International Union Against Cancer (UICC) [tumor node metastasis (TNM) classification (6)] distinguishes four grades:





G1: G2: G3: G4:

well differentiated moderately differentiated poorly differentiated undifferentiated

The WHO (1) provides and recommends a grading system with two classes:
Low-grade, encompassing G1 and G2
High-grade, encompassing G3 and G4 This latter grading system fulfills all clinical requirements, and can be performed with higher reproducibility. We prefer this grading with only two categories. When a carcinoma shows different grades of differentiation, the higher grade should determine the final categorization. Thus a carcinoma that shows both low- and high-grade areas should be classified as high-grade. However, the disorganized glands seen commonly at the advancing edge of the carcinoma should not be considered as high-grade malignancy. Highgrade carcinomas account for 20% to 25% of resected carcinomas.

Pathology

107

Additional Histological Parameters For describing the histomorphology of an individual colorectal carcinoma, there are some further parameters which might be taken into consideration (7–9):











The character of the invasive margin (pushing or expanding, or well-circumscribed vs. irregular diffusely infiltrating) Peritumoral inflammation The presence of peritumorous lymphoid aggregates Invasion of lymphatic vessels (L classification) (6): L0, no lymphatic invasion, L1, lymphatic invasion; LX, lymphatic invasion cannot be assessed Venous invasion (V classification) (6): V0, no venous invasion; V1, microscopic venous invasion; V2, macroscopic venous invasion; VX, venous invasion cannot be assessed. In case of microscopic venous invasion, it is important to distinguish between involvement of intramural veins (submucosa, muscularis propria) and that of extramural veins (beyond muscularis propria) Invasion of perineural spaces [Pn (perineural) classification] (10): Pn0, no perineural invasion, Pn1, perineural invasion; PnX, perineural invasion cannot be assessed

In addition, the type of lymph node reactions may be recorded, because they reflect the host reactions. There may be follicular hyperplasia (in more than 50% or in 50% or less of regional lymph nodes) or paracortical hyperplasia or both (11). SPECIAL CLINICAL TYPES OF COLORECTAL CANCER There are some special types of colorectal cancer, which should be mentioned (12,13). Hereditary Nonpolyposis Colon Cancer These carcinomas have a tendency to occur in the right colon and at a much younger age (most cases between 35 and 45 years) than the usual sporadic carcinoma (preferred age 55–65 years). The otherwise uncommon histological feature of medullary carcinoma with numerous tumor-infiltrating lymphocytes (TILs) and marked peritumoral lymphoid reaction is frequently seen. Mucinous adenocarcinomas and high-grade tumors are observed relatively often. These listed special histologic types should raise some suspicion on a hereditary nonpolyposis colon cancer (HNPCC) carcinoma even in a patient who is older than 60 years. In HNPCC patients there is an increased incidence of metachronous multiple primary tumors. HNPCC accounts for at least 4% to 6% of all colorectal carcinomas.

108

Wittekind

Carcinoma Arising in Familial Adenomatous Polyposis Less than 1% of colorectal carcinomas arise in familial adenomatous polyposis (FAP) and in the ‘‘hereditary flat adenoma syndrome’’ (HFAS), a variant of FAP, usually with fewer than 100 adenomas, mostly of the flat type. The histological features are similar to those of sporadic cancers. However, there is a high proportion of multiple synchronous primary tumors in symptomatic cases (up to a third of cases). Carcinoma Developing in Inflammatory Bowel Disease Carcinomas in inflammatory bowel disease arise predominantly in extensive ulcerous colitis with a history of 10 years or longer, involving most of the large bowel (right-sided colitis) and with high activity of inflammation. There are often synchronous multiple carcinomas. The incidence of flat and diffusely infiltrating carcinomas, high-grade tumors, mucinous adenocarcinomas, and signet-ring cell carcinomas is higher than in ordinary colorectal carcinomas. Less than 1% of all colorectal carcinomas arise in inflammatory bowel disease. Tumor Spread Knowledge of tumor spread is of paramount importance for surgical procedure in treating colorectal carcinoma. The possible routes of tumor spread are shown in Figure 1. Local Spread The different types of local spread determine the extent of surgical resection, and lead to the demand for avoidance of local tumor spillage. Extent of Resection In colon carcinoma, the extent of resection is determined by the extent of lymph node dissection and the vascular supply. In rectal carcinoma, the distal margin of clearance may be critical, especially in tumors located in the lower rectum. In this context, it is important that histological examination of rectal carcinomas in curable stages show that the continuous distal local spread usually extends no more than some millimeters beyond the grossly recognizable margin of the tumor. In the perimuscular tissue discontinuous spread in the form of so-called satellites must be considered. Such microscopic tumor nodules without residua of lymph node tissue are found in the mesorectum, predominantly in the radial direction, but also distal, some centimeters from the lower tumor margin. Low anterior resection for tumors of the middle and lower rectal third therefore has to include total mesorectal excision down to the pelvic floor (14), while the margin of clearance within the proper wall of the rectum

Pathology

109

Figure 1 Tumor spread in colorectal carcinoma.

(muscularis propria, submucosa, mucosa) may be limited to 1 cm in situ, except in cases of high-grade tumors, which require a 2 cm distal margin (15). Avoidance of Local Spillage of Tumor Cells Local spillage of tumor cells results in a significantly increased incidence of locoregional recurrence (16). This is observed in rectal carcinoma after iatrogenic tumor perforation during mobilization or tumor perforation. Adherence of a tumor to adjacent organs may be caused by peritumorous inflammation or by tumorous invasion. Discrimination by gross inspection or palpation is sometimes not possible. In such cases, a primary multivisceral resection is indicated if a curative operation is intended. Any biopsy from the adherence should be avoided, because in cases of tumor invasion, biopsy results in local tumor spillage (17). Lymphatic Spread and Lymph-Node Dissection The extent of lymph-node dissection in radical surgery is determined by the lymph drainage. In the sixth edition of the UICC TNM classification these findings have been taken into consideration in the definition of regional lymph nodes for anatomical sites and subsites (6). While most parts of the colon drain into one direction, for both flexures and the right and left third of the transverse colon, a bidirectional lymph drainage exists (18). Tumors of

110

Wittekind

the hepatic flexure and the right third of the transverse colon drain into nodes along the right as well as the middle colic artery. Therefore, radical resection should be performed as extended right hemicolectomy (right and transverse colectomy), with removal of the nodes along the ileocolic, right, and middle colic arteries. The bidirectional lymph drainage of tumors of the splenic flexure and the adjacent left third of the transverse colon and upper third of the descending colon requires an extended left hemicolectomy (left and transverse colectomy) for radical resection. The main lymph drainage of rectal carcinoma occurs upwards to the nodes along the superior rectal and inferior mesenteric vessels. Skipping of nodes is uncommon, and occurs in about 3% of cases with nodal metastasis (17). In curable cases, retrograde lymph-node metastasis (along the inferior rectal arteries to inguinal nodes) is exceptional. The reports on the incidence of lateral lymphatic spread are controversial. If lateral metastasis is defined as metastasis along the iliac vessels outside the mesorectum, involvement of these nodes in curable cases is obviously rare. Thus an indication for additional lateral dissection of the iliac nodes is at least questionable, in particular when considering the significant morbidity of this procedure (19). CLASSIFICATION OF ANATOMICAL EXTENT BEFORE TREATMENT Although in the history of staging the introduction of a pathological stage classification for rectal carcinoma by Cuthbert Dukes (20) in 1930 was an important step, due to several modifications this system may lead to considerable confusion. Thus, today the anatomical TNM system is recommended for daily use and clinical trials by the NIH Consensus Conference on Adjuvant Treatment (21) and was introduced for general use in cancer hospitals. The present sixth edition system is shown in Table 3. Micrometastasis and Isolated Tumor Cells Micrometastasis is defined as metastasis 0.2 cm or less in greatest dimension. Cases with only micrometastasis may be identified by the addition of ‘‘(mi)’’ to the pN and/or pM category, e.g., pNl(mi) or pMl(mi). In micrometastasis, tumor cells penetrate the wall of lymph sinus or blood vessels and there is extrasinusoidal or extravascular proliferation (10,22). Micrometastasis has to be distinguished from isolated (disseminated or circulating) tumor cells in lymph nodes, blood or bone marrow, or at other distant sites. Isolated tumor cells (ITCs) are defined as single tumor cells or small clusters of cells not more than 0.2 mm in greatest dimension that are usually detected by immunohistochemistry or molecular methods [flow cytometry or polymerase chain reaction (PCR)], but which may be verified with H and E stains (6). ITCs do not

Pathology

111

Table 3 The UICC TNM/pTNM Classification of Tumors of the Colon and Rectum T/pT—Primary tumor TX/pTX T0/pT0 Tis/pTis T1/pT1 T2/pT2 T3/pT3

T4/pT4

N/pN—Regional lymph nodesd NX/pNX N0/pN0 N1/pN1 N2/pN2 M/pM—Distant metastasis MX/pMX M0/pM0 M1/pM1

Primary tumor cannot be assessed No evidence of primary tumor Carcinoma in situ: intraepithelial or invasion of lamina propriaa Tumor invades submucosa Tumor invades muscularis propria Tumor invades through muscularis propria into subserosa or into nonperitonealized pericolic or perirectal tissue Tumor directly invades other organs or structuresb,c and/or perforates visceral peritoneum Regional lymph nodes cannot be assessed No regional lymph node metastasis Metastasis in 1–3 pericolic or perirectal lymph nodes Metastasis in 4 or more pericolic or perirectal lymph nodes Distant metastasis cannot be assessed No distant metastasis Distant metastasis

Regional lymph nodes for each anatomical site or subsite the following are regional Appendix Ileocolic Cecum Ileocolic, right colic Ascending colon Ileocolic, right colic, middle colic Hepatic flexure Right colic, middle colic Transverse colon Right colic, middle colic, left colic, inferior mesenteric Splenic flexure Middle colic, left colic, inferior mesenteric Descending colon Left colic, inferior mesenteric Sigmoid colon Sigmoid, left colic, superior rectal (hemorrhoidal), inferior mesenteric, and rectosigmoid Rectum Superior, middle, and inferior rectal (hemorrhoidal), inferior mesenteric, internal iliac, mesorectal (paraproctal), lateral sacral, presacral, sacral promontory (Gerota) (Continued)

112

Wittekind

Table 3

The UICC TNM/pTNM Classification of Tumors of the Colon and Rectum (Continued )

Ramifications (i.e., optional subdivisions of existing TNM/pTNM categories) (10) pT3 pT3a Minimal: tumor invades through the muscularis propria into the subserosa or into nonperitonealized pericolic or perirectal tissues, not more than 1 mm beyond the outer border of the muscularis propria pT3b Slight: tumor invades through the muscularis propria into the subserosa or into nonperitonealized pericolic or perirectal tissues, more than 1 mm but not more than 5 mm beyond the outer border of the muscularis propria pT3c Moderate: tumor invades through the muscularis propria into the subserosa or into nonperitonealized pericolic or perirectal tissues, more than 5 mm but not more than 15 mm beyond the outer border of the muscularis propria pT3d Extensive: tumor invades through the muscularis propria into the subserosa or into nonperitonealized pericolic or perirectal tissues, more than 15 mm beyond outer border of the muscularis propria pT4 pT4a Invasion of adjacent organs or structures, without perforation of the visceral peritoneum pT4b Perforation of the visceral peritoneum Note: The definitions of the clinical classification (TNM) correspond to those of the pathological classification (pTNM). Stage grouping is shown in Figure 2. a This includes cancer cells confined within the glandular basement membrane (intraepithelial) or lamina propria (intramucosal) with no extension through the muscularis mucosae into the submucosa. b Direct invasion in T47pT4 includes invasion of other segments of the colorectum by way of the serosa, e.g., invasion of the sigmoid colon by a carcinoma of the cecum. c Tumor that is adherent to other organs or structures, macroscopically, is classified as T4. However, if no tumor is present in the adhesion, microscopically, the classification should be pT3. d A tumor nodule in the pericolic/perirectal adipose tissue without histological evidence of residual lymph node in the nodule is classified in the pN category as a regional lymph node metastasis if the nodule has the form and smooth contour of a lymph node. If the nodule has an irregular contour, it should be classifies in the T category and also coded as V1(microscopic venous invasion) or V2, if it was grossly evident, because there is a strong likelihood that it represents venous invasion. Source: From Ref. 6.

Pathology

113

Figure 2 International Union Against Cancer stage grouping. Source: From Ref. 6.

typically show evidence of metastatic activity (e.g., proliferation or stromal reaction) or penetration of vascular or lymphatic sinus walls. ITCs are not considered in the TNM classification, because their independent prognostic significance remains to be proven. It is recommended that positive morphological findings of ITC should be indicated by the addition of (iþ), and positive nonmorphological findings by the addition of (molþ), e.g., pN0(iþ) or M0(molþ).

CLASSIFICATION OF ANATOMICAL EXTENT AFTER TREATMENT While TNM and pTNM describe the anatomical extent of cancer before treatment, the residual tumor (R) classification deals with tumor status after treatment. It reflects the effects of therapy, influences further therapeutic procedures, and is the strongest predictor of outcome (6,10). Thus, following tumor resection, first of all, the pathologist has to examine the resection margins to obtain a reliable R classification, which is based on clinical as well as histopathological findings. In this regard, the main problem is the

114

Wittekind

examination of the circumferential (lateral, radial), i.e., mesorectal and mesocolon, resection margins. In general, at least two conventional blocks or a large-area (giant) block have to be submitted for histology (for detailed methodological recommendations, see references 23, 24). HISTOLOGICAL GRADING OF TUMOR REGRESSION In cases of neoadjuvant (preoperative) radio- and/or chemotherapy the pathologist may be requested to assess the tumor response. Unfortunately, there are not yet any internationally agreed systems for describing the regressive changes and staging methods in such cases. A proposal of Dworak et al. (25) is used in Germany. Complete regression is possible, but, from our experience is uncommon. In one randomized controlled trial it was observed in 8% of all primary carcinomas after neoadjuvant radiochemotherapy (26). The tumor region should be carefully worked up. Thus, for regression grading, detailed statements on workup should be given to enable estimation of the reliability of findings. A further problem, which is presently under discussion, is how to classify those neoadjuvant treated cases in the TNM classification. This has been addressed in the sixth edition (6). The prefix ‘‘y’’ should be used to identify these cases. PROGNOSTIC FACTORS Prognostic factors are variables (covariates) with independent influence on outcome. They may differ according to the various ways of measuring outcome (various endpoints) (e.g., overall survival, disease-free survival, relapse rate, and response to treatment) and also for different patient subgroups. The identification of prognostic factors, in particular the acceptance of new prognostic factors, should follow certain rules (27). Table 4 shows the prognostic factors for patients with complete resection of the tumor and those with residual tumor (R1, R2). The prognostic factors are divided into proven and probable. Another subdivision has been introduced recently distinguishing essential prognostic factors (necessary for treatment decisions), additional prognostic factors (may modify treatment and prognosis), and new and promising factors (9). Only recently sufficient data are available to demonstrate that surgical treatment and the individual surgeon are independent prognostic factors (28–31) in colorectal cancer. It should be emphasized that the independent prognostic significance of all biological and molecular factors (factors of the so-called new pathology or new and promising factors) remains to be proven (4,7–9). THE HISTOPATHOLOGICAL REPORT The histopathological report has to include all possible information that may be important in relation to tumor classification, description of surgical

CEA serum level ( >5 ng/mL) Comorbid disease (present, Higher ASA grade) Surgeon

Patient-related

For subgroups: multimodal therapy (not performed)

Peritumoral lymphoid cells/lymphoid aggregates (nonconspicuous/absent) Gender (male)

Tumor perforation/obstruction (present) Lymphatic and perineural invasion (present)

Anatomical site of primary (rectum)

Probable prognostic factors

Abbreviations: ECOG, European Cooperative Oncology Group; CEA, carcinoembryonic antigen; ASA, American Society of Anaesthesia; R1, microscopic residual tumor; R2, macroscopic residual tumor; pTNM, pathological tumor node metastasis. Source: From Refs. 7 and 9.

Treatment-related

Anatomical extent: pTNM and stage grouping (higher category) Histological grade Venous invasion (present, predominantly extramural) Histological pattern of tumor margin (infiltrative)

Proven prognostic factors

Patients with complete resection of tumor (no residual tumor, R0)

Tumor-related

B.

A. Patients with residual tumor (R1, R2) Proven prognostic factors Distant metastasis (present) Localization of residual tumor (distal) For patients with multiple distant metastases: performance status (increasing ECOG grade, decreasing Karnofsky score)

Table 4 Prognostic Factors in Colorectal Carcinoma (Unfavorable Level of Covariates Is Shown in Parentheses)

Pathology 115

116

Wittekind

procedure, and prognostic factors. This enables a reliable diagnosis, treatment decisions, estimation of prognosis, and analysis of treatment results and of the quality of diagnosis and treatment. Specific recommendations for the content of surgical pathology reports have been published in various countries since the 1980s. The minimal data to be included in a present-day surgical pathology report are listed in Table 5. It is based on recent publications (4,32–34). QUALITY MANAGEMENT WITHIN PATHOLOGY DEPARTMENTS Quality management is a requirement for diagnostic activities in pathology departments as well as for the treatment of cancer and for clinical studies. The methods of quality management within pathology departments have recently been described and summarized by Rosai (35). There are indicators of the quality of histopathological assessment and workup (Table 6). In all pathology departments, the respective data should currently be collected and analyzed. Any deviation from the usual values (ranges) and changes in frequencies should lead to careful analysis and response. Special attention should be paid to careful examination of the circumferential resection margins and lymph nodes because of the crucial prognostic significance of the respective findings (23,34). PATHOLOGY FINDINGS IN RESECTION SPECIMENS INDICATIVE OF ONCOLOGICAL QUALITY OF SURGERY The most important goal of surgical treatment is to achieve complete tumor resection (R0 resection). Thus the rate of R0 resections related to all patients is an important intermediate indicator of quality. The pathological findings below on resection specimens give further information on the oncological quality of surgery: 1. Evidence of local spillage of tumor cells: iatrogenic tumor perforation or tumor resection not en bloc with transsection of tumor tissue. 2. Length of resected bowel: limited (segmental) resection or radical resection with ligature of the trunk of the supplying vessels. 3. In cases of colon carcinomas with multidirectional lymphatic drainage: dissection of one or two lymph drainage areas. 4. Number of removed lymph nodes (provided that there is an adequate node examination technique). 5. In rectal carcinoma of the upper third: distal margin of clearance in muscular walls as well as in mesorectum (no coning) not less than 5 cm in situ corresponding to 3 cm measured on the fresh resection specimen without tension.

Pathology

117

Table 5 Minimal Data to Be Included in Surgical Pathology Reports on Colorectal Carcinoma Specimens Incision biopsies Gross description Histology

Number of pieces Extension (intraepithelial/intramucosal/invasion of submucosa) Histological type Histological grade

Polypectomies Gross description

Number of pieces Macroscopic type (flat/sessile/semi-pedunculated/ pedunculated) Greatest dimension (without stalk) Histology: tumor Histological type Histological grade Extension (intra-epithelial/intramucosal/invasion of submucosa) In semi-pedunculated and pedunculated polyps: extension of invasion of submucosa (none/ head/stalk) Lymphatic invasion (L classification) Venous invasion involvement Histology: margins Minimal distance of tumor from margin (mucosal margins, deep margin) Local excision (submucosal, full thickness) Gross description Number of pieces Tumor configuration (exophytic-fungating/ endophytic-ulcerative/diffusely infiltrating) Greatest dimension of tumor Margins: minimal distance of tumor from mucosal and deep margins Histology: tumor Histological type Histological grade Extension (pT classification) Lymphatic invasion (L classification) Venous invasion (intramural, extramural) Perineural invasion (Pn classification) Histology: margins Involvement. Minimal distance of tumor from mucosal and deep margins Histology Additional pathological findings (e.g., adenoma. intra-epithelial neoplasia) Resection specimens Gross description: Parts of colorectum removed resection specimen (Continued)

118

Wittekind

Table 5 Minimal Data to Be Included in Surgical Pathology Reports on Colorectal Carcinoma Specimens (Continued )

Gross description: tumor localization

Gross description: margins

Histology: tumor

Histology: margins

Histology: additional lesions

Adjacent organs removed Number of pieces received (resection en bloc/not en bloc) Number of malignant tumors In rectal carcinomas: Site in relation to peritoneal reflection (above/ at/below) Site of distal border of tumor (upper/middle/ lower third) In case of abdominoperineal excision: distance between distal border of tumor and anal verge/ method of measurement Greatest dimension Tumor perforation (spontaneous/iatrogenic) Minimal distance from proximal and distal margin/ method of measurement Minimal distance from circumferential (radial, lateral) margin/method of measurement Anatomical extent: pTNM classification Number of regional lymph nodes examined Number of regional lymph nodes involved Apical lymph node status Histological type Histological grade Lymphatic invasion (L classification) Venous invasion (intramural, extramural) Perineural invasion (Pn classification) Histological pattern of infiltrating margins (pushing-expanding/diffusely infiltrating) Involvement, minimal distance for Proximal margin Distal margin Doughnut Circumferential (radial, lateral) Resected adjacent organs Adenoma/intraepithelial neoplasia/familial adenomatous polyposis/ulcerative colitis/ other chronic inflammatory bowel diseases/other

Abbreviations: L, lymphatic invasion; pT, pathological T; Pn, perineural invasion; pTNM, pathological tumor node metastasis. Source: From Refs. 4, 32–34.

Pathology

119

Table 6 Quality Indicators for Pathological Diagnosis Range Parameter Tumor type Mucinous adenocarcinoma/ frequency Tumor grade High-grade/frequency R classification Frequency of R1 related to resections considered as complete by the surgeon Regional lymph nodes Frequency of node-positive cases related to radical resections for cure (R0 resections) Number of examined nodes in radical standard resections for cure (R0)b,c/mean Frequency of cases with fewer than 12 lymph nodesc

Colon

Rectum

15%

10%

Indicative of Adherence to WHO classification

20–25% 0–5%

5–10 (-20)%

Carefulness of histological examination of resection linesa

40–50%

20–30%

Carefulness of histological examination of lymph node drainage areaa

< 5%

a

Also influenced by the surgeon. Radical standard resection is defined as bowel resection with formal dissection of the lymph node drainage area. c Except cases with neoadjuvant therapy. Abbreviation: WHO, World Health Organization. Source: From Refs. 38 and 39. b

6. In rectal carcinoma of the middle and lower third:
Careful gross inspection of the surface of the specimen: appearance of the correctly mobilized mesorectum with intact smooth surface.
Distal margin of clearance in muscular wall not less than 1 cm measured on the fresh resection specimen without tension. QUALITY ASSURANCE OF CLINICAL TRIALS ON ADJUVANT AND NEOADJUVANT THERAPY: THE SURGICAL PATHOLOGIST’S POINT OF VIEW In adjuvant treatment, the quality of pathological examination of resection specimens influences the selection of patients and thus the results. Therefore, data indicating the quality of pathology (Table 6) should always be included

120

Wittekind

in reports on respective clinical trials. The frequency of all tumor resections, of resections without and with microscopic residual tumor (R0, R1), and the pN classification for all patients seen at the institution(s) during the study period should be stated. This information indicates the general surgical attitude as well as the quality of pathological examination. Tumor classifications performed according to international recommendations are important indicators of the quality of oncological studies. Any comparison of results will be made impossible by authors who do not classify their tumors according to the generally accepted international systems. MALIGNANT TUMORS OTHER THAN CARCINOMAS Traditionally, neuroendocrine tumors have been separated from epithelial tumors and classified in a special way. They are classified as:
Well-differentiated neuroendocrine tumor (formerly: carcinoid), ICD-O code 8240/1
Well-differentiated neuroendocrine carcinoma (formerly: malignant carcinoid), ICD-O code 8240/3
Poorly differentiated neuroendocrine carcinoma (small cell carcinoma), ICD-O code 8041/3 Most endocrine tumors of the appendix are localized at the distal tip, and cause local symptoms leading to appendectomy. Most tumors produce serotonin and show benign behavior. The uncommon malignant carcinoids require radical right hemicolectomy. In the colon neuroendocrine tumors are very rare, mostly of the poorly differentiated type, and affected patients have a poor prognosis. The rectum is the preferred site of endocrine tumors. They are mostly small (<1 cm), asymptomatic, well-differentiated neuroendocrine tumors. Metastasis may be seen in tumors larger than 2 cm or with invasion of the proper muscle. Exocrine carcinomas frequently show a minority of dispersed endocrine cells. Mixed exocrine–endocrine carcinomas should be diagnosed only in the case of quantitatively balanced amounts of endocrine and exocrine components. About 1% of all malignant colorectal tumors are gastrointestinal stromal tumors (GIST). Kaposi sarcoma in the colon and rectum is usually observed in patients with AIDS and Kaposi sarcoma of the skin and lymph nodes. In most cases, the involvement of the large intestine is clinically silent. Other malignant mesenchymal tumors are extremely rare. A few cases of primary malignant melanomas in the rectum (without involvement of the anal region) have been reported. Primary colorectal malignant lymphomas (no evidence of liver, spleen, nonmesenteric lymph node, or bone marrow involvement at the time of presentation) are very rare, most cases involving the ileocecal region and

Pathology

121

the rectum. Sometimes an association with ulcerative colitis or Crohn’s disease has been observed. Grossly, a nodular or polypoid mass or a diffuse infiltrate may be present. The classification is not yet standardized (for further details in noncarcinomatous malignant tumors see references 5,6,36,37). REFERENCES 1. Hamilton SR, Aaltonen LA, eds. World Health Organization (WHO) Classification of Tumors. Pathology and Genetics. Tumors of the Digestive System. IARC Press, Lyon, 2000. 2. Kudo S. Endoscopic mucosal resection of flat and depressed types of early colorectal cancer. Endoscopy 1993; 25:456–461. 3. Fielding LP, Arsenault PA, Chapuis PH, et al. Clinicopathological staging for colorectal cancer: An International Documentation System (IDS) and an International comprehensive Anatomical Terminology (ICAT). J Gastroenterol Hepatol 1991; 6:325–344. 4. Soreide O, Norstein I, Fielding LP, Silen W. International standardization and documentation of the treatment of rectal cancer. In: Soreide I, Norstein O, eds. Rectal Cancer Surgery. Optimisation—Standardisation—Documentation. Berlin: Springer, 1997:405–445. 5. Riddell RH, Petras RE, Williams GT, Sobin LH. Armed Forces Institute of Pathology. Atlas of Tumor Pathology, 3rd Series, Fascicle 32. Tumors of the Intestine. Washington: Am Registry Pathology, 2003:133–189. 6. Sobin LH, Wittekind Ch, eds. UICC Classification of Malignant Tumors. 6th ed. New York: Wiley, 2002. 7. Hermanek P, Sobin LH. Colorectal carcinoma. In: Hermanek P, Gospodarowicz MK, Henson DE, et al. Prognostic Factors in Cancer. Berlin: Springer, 1995:64–79. 8. Compton CC, Fenoglio-Preiser CM, Pettigrew N, Fielding LP. American Joint Committee on Cancer Prognostic Factors Consensus Conference: Colorectal Working Group. Cancer 2000; 88:1739–1757. 9. Hobday TJ, Erlichman CE. Colorectal Cancer. In: Gospodarowicz MK, Henson DE, Hutter RVP, et al., eds. Prognostic Factors in Cancer. 2nd ed. 2001:267–279. 10. Wittekind Ch, Henson DE, Hutter RVP, Sobin LH. UICC TNM Supplement. A Commentary on Uniform use. 3rd ed. New York: Wiley, 2003. 11. Dworak O. Morphology of lymph nodes in the resected rectum of patients with rectal carcinoma. Pathol Res Pract 1991; 187:1020–1024. 12. Green SE, Bradburn DM, Varma IS, Burn I. Hereditary non-polyposis colorectal cancer. Int J Colorect Dis 1998; 13:3–12. 13. Lynch HT, Lynch IF. Genetics of colorectal cancer. Digestion 1998; 59: 481–492. 14. Heald RG, Ryall KD, Husband E. The mesorectum in cancer in rectal cancer surgery: clue to pelvic recurrence. Br J Surg 1982; 69:613–616. 15. Hohenberger W, Hermanek P Jr., Hermanek P, Gall FP. Decision making in curative rectum carcinoma surgery. Onkologie 1992; 15:209–220. 16. Zirngibl H, Husemann B, Hermanek P. Intraoperative spillage of tumor cells in surgery for rectal cancer. Dis Colon Rectum 1990; 33:610–614.

122

Wittekind

17. Hermanek P. Pathology of colorectal cancer. In: Bleiberg H, Kemeny N, Rougier P, et al., eds. Colorectal Cancer. A Clinical Guide to Therapy. London, M. Dunitz, 2002:55–72. 18. Jatzko G, Lisberg P, Wette V. Improving survival rates for patients with colorectal cancer. Br J Surg 1992; 79:588–591. 19. Schoelefield JH, Northover IMA. Surgical management of rectal cancer. Br J Surg 1995; 82:745–748. 20. Dukes CE. The spread of cancer of the rectum. Br Surg 1930; 12:643–648. 21. NIH Consensus Conference. Adjuvant therapy for patients with colon and rectal cancer. JAMA 1990; 264:1444–1450. 22. Hermanek P, Hutter RVP, Sobin LH, Wittekind Ch. Classification of isolated tumor cells and micrometastasis. Communication International Union Against Cancer. Cancer 1999; 86:2668–2673. 23. Hermanek P, Wittekind Ch. The pathologist and the residual tumor (R) classification. Pathol Res Pract 1994; 190:115–123. 24. Wittekind Ch, Compton CC, Greene FL, Sobin LH. UICC Communication. TNM Residual tumor classification revisited. Cancer 2002; 94:2511–2519. 25. Dworak O, Keilholz L, Hoffmann A. Pathological features of rectal cancer after preoperative radiochemotherapy. Int J Colorect Dis 1997; 12:19–23. 26. Sauer R, Becker H, Hohenberger W, et al., for the German Rectal Cancer Study Group. Preoperative compared with postoperative chemoradiotherapy for locally advanced rectal cancer. New Engl J Med 2004; 351:1790–1792. 27. Hermanek P. Prognostic factor research in oncology. J Clin Epidemiol 1999; 52:371–374. 28. McArdle CS, Hole D. Impact of variability among surgeons on postoperative morbidity and mortality and ultimate survival. Br Med J 1991; 302: 1501–1505. 29. Kessler H, Mansmann U, Hermanek P Jr., et al. for the Study Group Colo-Rectal Carcinoma (SGCRC). Does the surgeon affect outcome in colon carcinoma? Semin Colon Rectal Surg 1998; 9:233–240. 30. Kessler H, Hermanek P, for the Study Group Colo-Rectal Carcinoma (SGCRC). Outcomes in rectal cancer surgery are directly related to technical factors. Semin Colon Rectal Surg 1998; 9:247–253. 31. Hermanek P. Impact of surgeon’s technique on outcome after treatment of rectal carcinoma. Dis Colon Rectum 1999; 42:559–562. 32. Association of Directors of Anatomical and Surgical Pathology. Recommendations for the reporting of resected large intestinal carcinomas. Hum Pathol 1996; 27:5–8. 33. Compton CC, Henson DE, Hutter RVP, et al., for Members of the Cancer Committee, College of American Pathologists. Updated protocol for the examination of specimens removed from patients with colorectal carcinoma. Arch Pathol Lab Med 1997; 121:1247–1254. 34. Quirke Ph. The pathologist, the surgeon and colorectal cancer—Get it right because it matters. Progr Pathol 1998; 4:201–213. 35. Rosai J. Ackerman’s Surgical Pathology. 9th ed. Edinburgh: Mosby, 2004. 36. Lewin KJ, Riddell RH, Weinstein WM. Gastrointestinal Pathology and its Clinical Implications. Tokyo: Igaku-Shoin, 1992.

Pathology

123

37. Fenoglio-Preiser CM, Noffsinger AE, Stemmermann GN, et al. Gastrointestinal Pathology. 2nd ed. Philadelphia: Lippincott-Raven, 1999. 38. Hermanek P. Qualita¨tsmanagement bei Diagnose und Therapie kolorektaler Karzinome. Leber Magen Darm 1996; 26:20–24. 39. Hermanek P. Qualita¨t der Chirurgie aus der Sicht des Pathologen. In: Bu¨chler MW, Heald RJ, Maurer CA, eds. Das Konzept der totalen mesorektalen Exzision. Basel: Karger, 1998.

5 Familial Cancer Management Sabine Tejpar Digestive Oncology Unit, University Hospital Gasthuisberg, Leuven, Belgium

INTRODUCTION Colorectal cancers (CRC), whether sporadic or hereditary, are caused by a defined set of molecular events. These genetic alterations have been described in detail elsewhere in this book. Briefly, genes functioning in diverse pathways such as maintenance of crypt homeostasis, cell cycle regulation, differentiation, genetic stability, and others are disrupted in colorectal epithelial cells. In sporadic cancers these alterations occur at a somatic level, meaning multiple alterations accumulate in a single, hitherto normal epithelial cell. This cell, which is then initiated, will clonally expand and progress to cancer through the acquisition of additional mutations. Hereditary cancers are driven by the same molecular events, but some of the first genetic alterations are already present in the germline and thus in all epithelial cells of the patient at birth. This leads to a greatly increased risk of tumor initiation and subsequent development of full-blown cancer. Some of the hereditary syndromes are caused by mutations in genes important in tumor initiation [familial adenomatous polyposis (FAP)], and others in genes important in tumor progression [hereditary nonpolyposis colorectal cancer (HNPCC)]. The net effect in both types of syndromes is a very high lifetime risk for CRC (60–100%) at a young age (median 42). Although these hereditary forms represent less than 5% of all CRCs, it is readily appreciable that correct recognition and preventive management

125

126

Tejpar

of these syndromes could dramatically impact on the cancer risk and survival of affected individuals. In addition to the hereditary occurrence of CRC due to defined genetic alterations, a large group of patients present with apparently familial clustering of the disease. This group is much larger than the hereditary forms; estimates range between 15% and 30% of all CRCs. Familial clustering cases represent a large, loosely defined group that can represent from two to over four cases in a family, with varying ages of onset and kinship. Probably different genes are responsible for these clusterings, but no clinical or molecular criteria are yet available, aiding the classification of these families. Due to the lack of knowledge and correct risk assessment in these families the approach to screening and prevention is more empiric. This large group of CRC patients and their relatives represent a major challenge in the coming years, both in terms of basic science research and validation of clinical management guidelines. This chapter will provide a ‘‘user friendly’’ overview of the hereditary and familial gastrointestinal (GI) cancer syndromes, aiming to help clinicians in the recognition and management of this disorder. Optimal management of familial cancer syndromes needs some out-of-the-ordinary resources. For instance, the rapid advances in molecular medicine can pose a challenge. Hence we emphasize issues such as when and how to initiate molecular and genetic testing, and to refer for genetic counseling and risk assessment. A multidisciplinary approach to these patients is crucial. Many organs are at risk, for example gynecological and urinary tract cancers in colorectal patients. In addition management options can vary from surveillance to extensive surgery. Moreover, dealing with whole families as opposed to individual patients involves new issues such as risk management over generations to come, collaboration with genetic services and registries, providing psychosocial support, addressing family planning issues, etc. Advancement in the management of these patients is expected from the growing number of multidisciplinary clinics being established worldwide and the heightened public awareness of these syndromes. Recognized hereditary syndromes account for only about 5% of CRCs. The commonest hereditary syndromes are familial adenomatous polyposis (FAP) and hereditary nonpolyposis colorectal cancer (HNPCC). Attenuated FAP, juvenile polyposis syndrome, Peutz–Jeghers and Cowden syndrome are rarer, mendelian causes of CRC (Table 1). The most important step leading to the diagnosis of a hereditary cancer syndrome is the compilation of a thorough family history of cancer (1). A detailed family history can be compiled by a patient and his relatives, aided by his physician, trained nurse or genetic counselor. The focus should be on identifying cancer of all types and sites, the family member’s age at the onset of cancer, any pattern of multiple primary cancers, any association with specific phenotypic features, and above all, documentation of pathological findings whenever possible. For many CRC syndromes, individuals are at

Familial Cancer Management

127

Table 1 Inherited Syndromes that Predispose to Gastrointestinal Cancer and Their Associated Genes Syndromes FAP, attenuated FAP HNPCC Peutz–Jeghers syndrome Juvenile polyposis Cowden’s syndrome Familial diffuse gastric cancer Familial pancreatic cancer

Associated genes APC, MYH (rare) Mismatch repair genes: MLH1, MSH2, MSH6, PMS2 (rare) STK11 SMAD4, BMPR1A PTEN E-cadherin BRCA2, mismatch repair genes, STK11, CDKN2A

Abbreviations: FAP, familial adenomatous polyposis; HNPCC, hereditary nonpolyposis colorectal cancer; APC, adenomatous polyposis coli; MYH, MutY associated polyposis; PTEN, phosphatase and tensin homolog.

risk for neoplasms outside of the colon and even the gastrointestinal tract. A constellation of clinical features should be sought out at this stage, leading to the potential recognition of a hereditary CRC syndrome in the family. Many primary care physicians and specialists may wish to refer the patient to a hereditary cancer specialist and genetic counselor for further evaluation if they are not immediately able to classify the family and for questions about the disorder’s clinical or molecular genetic diagnosis and the need for targeted surveillance and management. Once it is clear that a patient has a familial form of CRC, genetic counseling is mandatory and must provide the patient and the extended family with important details about their genetic risk of cancer at specific sites and options for surveillance and management, on the basis of the natural history of the disorder. MAIN HEREDITARY COLORECTAL CANCER SYNDROMES Familial Adenomatous Polyposis FAP until recently was known as an autosomal dominant disorder caused by germline mutations in the adenomatous polyposis coli (APC) gene. For patients with FAP, every cell in the body harbors a mutation at an APC allele, ensuring that a very large number of adenomatous polyps will occur once an inactivating event occurs in the second wild-type APC allele. Thus, it is nearly certain that polyps will develop several decades earlier in FAP than in the general population in which additional time is required to accumulate both mutational events. FAP is an autosomal dominant disorder that accounts for less than 1% of all CRCs. Approximately 30% of cases are spontaneous de novo mutations in an individual, thus with no apparent family history of the disorder.

128

Tejpar

In 2002 (2), mutations in the MYH gene, a base excision repair gene, were found to cause a recessive form of FAP, which is otherwise often clinically indistinguishable from classic APC-driven FAP (3,4). Mutations in MYH, if present on both alleles (inherited from both parents), will lead to loss of base excision repair in all cells. For unclear reasons, loss of base excision repair leads to frequent somatic APC mutations in colorectal and other cells. MAP patients (MYH-associated polyposis) develop a less florid polyposis than classic FAP (10–100 instead of 100–1000 polyps), but interestingly develop most of the extracolonic manifestations of FAP such as osteomas and congenital hypertrophy of the retinal pigment epithelium, despite the lack of germline APC mutation. Data on the incidence of biallelic MYH mutations in attenuated polyposis, sporadic, familial, or young-onset CRC kindreds are still being gathered. Possibly additional clinical syndromes will be attributed to loss of germline base excision repair in the future. Presently, MYH mutation testing is recommended in patients with polyposis where the disease segregation in the family appears recessive. Clinical management of MAP is the same as for FAP. Clinical Manifestations Classic FAP is characterized by the development of hundreds of intestinal adenomatous polyps, by an increased risk of extracolonic tumors (e.g., duodenal cancer, pancreatic cancer, and thyroid cancer), and sometimes by other extraintestinal features such as congenital hypertrophy of the pigmented retinal epithelium and desmoid tumors. In FAP, polyps present on average at age 16 years (range, 7–36 years), and colon cancer at a mean age of 39 years. Individuals with FAP often have duodenal adenomas and gastric fundic gland polyps and are at substantially increased risk of developing duodenal adenocarcinoma. In addition, Gardner’s syndrome is a variant form of FAP that presents with osteomas, odontomas, epidermoid cysts, fibromas, and desmoid tumors in addition to colonic adenomas. Attenuated FAP (AFAP), also caused by a subset of APC mutations, is characterized by an average of only 30 polyps and variable extraintestinal manifestations. Adenoma and cancer occurrence are delayed by at least 10 years compared with classic FAP. Presently, it is recommended that screening of relatives and genetic testing for AFAP be considered in any patient with 20 or more adenomas at one colonoscopy or over time. Extracolonic manifestations of FAP are summarized in Table 2. Clinical Management Individuals who inherit a mutant APC gene have a very high likelihood of developing colonic adenomas; the risk has been estimated to be over 90%. Patients who have an APC mutation or who have one or more first-degree relatives with FAP or an identified APC mutation (or both) are at high risk

Familial Cancer Management

129

Table 2 Extracolonic Features in Familial Adenomatous Polyposis and Hereditary Nonpolyposis Colorectal Cancer Syndrome FAP

Cancers

Other lesions

Brain (medulloblastoma)

Congenital hypertrophy of the retinal pigment epithelium Nasopharyngeal angiofibroma Osteomas Radiopaque jaw lesions Supernumerary teeth Lipomas, fibromas, epidermoid cysts Desmoid tumors Gastric adenomas, fundic gland polyps Duodenal, jejunal, ileal adenomas Cafe´ au lait spots Sebaceous gland adenomas, carcinomas Keratoacanthomas

Thyroid Duodenal Periampullary Pancreas Hepatoblastoma Biliary tree

HNPCC

Brain (glioblastoma) Stomach Small bowel Biliary tract Ureter and renal pelvis Uterus Ovary

Abbreviations: FAP, familial adenomatous polyposis; HNPCC, hereditary nonpolyposis colorectal cancer.

and should undergo screening by the age of 10 to 12 years. Endoscopic screening (Table 3) consists of yearly sigmoidoscopies or colonoscopies, starting at 12 years of age and reducing screening after the age of 50. By the age of 10 only 15% of FAP patients manifest adenomas; the probability rises to 75% by the age of 20. If polyposis develops, prophylactic subtotal colectomy is the recommended therapy to eliminate the development of CRC. Surgery is usually advised at the time of diagnosis to minimize the risk Table 3 Recommendations for Familial Adenomatous Polyposis Screening Procedure Colon

Colonoscopya

Stomach, duodenum

Upper GI endoscopy Genetic testing

a Preventive colectomy from 18 years if affected. Abbreviation: GI, gastrointestinal.

Age at start

Interval (years)

12–45 45–65 >65 30–35

1 2 5 1–3 depending on findings

12

130

Tejpar

Table 4 Modified Spigelman’s Score and Classification Factor No. of polyps Polyp size, mm Histology Dysplasia

1 point

2 points

3 points

1–4 1–4 Tubulous Low grade

5–20 5–10 Tubulovillous –

>20 >10 Villous High grade

Note: Classification—no polyps: stage 0; 1–4 points: stage I; 5–6 points: stage II; 7–8 points; stage III; 9–12 points: stage IV.

of malignancy, but because there is a relatively long time span between the onset of adenomas and the appearance of cancer, school and work schedules are taken into account in these teenage patients. Surgical options include subtotal colectomy with ileorectal anastomosis, total proctocolectomy with Brooke ileostomy (or continent ileostomy), and proctocolectomy with mucosal proctectomy and ileoanal pull-through (with pouch formation). Because CRC can occur in the rectal segment, the two latter procedures are favored. In patients with subtotal colectomy routine endoscopic surveillance of the remaining rectum every six months is mandatory. Even patients with proctocolectomies and ileoanal pouch should be followed because rare cases of adenomas in the pouch have been reported. Adenomas in the duodenum occur in the majority of FAP patients, with a lifetime risk of periampullary carcinoma of about 5%. There is limited knowledge about the causation, prevention, and management of duodenal polyposis in FAP. However, a recent detailed analysis of the natural history of duodenal polyposis in FAP patients shows a strong association between stage IV (according to the Spigelman classification, Table 4), duodenal adenomas, and duodenal cancer. Patients with Spigelman stage IV disease should be offered prophylactic surgery. Screening with upper GI endoscopy is recommended every one to three years, according to the Spigelman stage (Table 5) (5). As a Table 5 Proposed Program for Surveillance and Treatment of Duodenal Adenomatosis Spigelman Spigelman Spigelman Spigelman Spigelman

a

stage stage stage stage stage

0 I II III IV

Endoscopya at intervals 5 years Endoscopyb at intervals 5 years Endoscopy at intervals 3 years Endoscopy at intervals 1–2 years Endoscopic ultrasonography Consider pancreas-sparing or pylorussparing duodenectomy

Including random biopsies from mucosal folds in patients without polyps. Including multiple biopsies from polyps.

b

Familial Cancer Management

131

result of the increased risk of extracolonic tumors, including thyroid cancer and hepatoblastoma in young children, optimal management of FAP includes follow-up on thyroid function, liver function tests, and ultrasound of the liver in young children. Chemoprevention Patients with FAP who were treated with 400 mg of celecoxib, a selective inhibitor of cyclooxygenase-2, twice a day for six months had a 28.0% reduction in the mean number size of colorectal polyps when compared with patients in the placebo group. However, polyps may return while the patient is taking nonsteroidal anti-inflammatory drugs and certainly do at discontinuation of the drug. Moreover, selective COX-2 inhibitors and nonsteroidal anti-inflammatory drugs do not influence the progression of polyps toward malignancy. Currently, none of these chemoprevention strategies should replace screening or prophylactic colectomy. Potentially they can impact on the endoscopic management of the remaining rectum after subtotal coloctomy by reducing polyp burden. Genetic Testing for FAP As for all hereditary syndromes, genetic testing starts by a mutation analysis in an affected proband. If the mutation is found, at-risk family members can now be tested for this mutation. In classic FAP, the indications to start mutation screening in a family is classic polyposis (Table 9). At-risk family members will be tested for the specific mutation at age 10 to 12 years, the age at which colon cancer screening should start. For AFAP (6) it is recommended that mutation analysis be initiated in any patient with 20 or more adenomas at one colonoscopy or over time. MYH mutation analysis should be initiated in kindreds with a recessive inheritance pattern of CRC or moderate polyposis. Hereditary Nonpolyposis Colorectal Cancer HNPCC patients have a germline mutation in one copy of an MMR (mismatch repair) gene in every cell (most of the HNPCC kindreds are due to germline mutations in hMSH2 or hMLH1) (7). During the lifetime of this person, inactivation of the second allele will occur at random in certain cells. This event leads to absence of the gene product in these cells, causing MMR deficiency. MMR deficiency has been described in many malignancies, for example,endometrium,ovary,stomach,ureters,andothers. However,HNPCC patients seem to develop predominantly CRC. The lifetime risk for an HNPCC carrier to develop a colorectal carcinoma is 78%, and for endometrial carcinoma it is 48% (8). This is difficult to explain. Genes important in colon epithelial homeostasis might be particularly sensitive to MMR deficiency. For

132

Tejpar

example, the gene coding for the TGF-b type II receptor, which contains many microsatellite repeats, is frequently mutated in HNPCC CRCs. Clinical Manifestations HNPCC is characterized by early-onset colon cancer (median age 42) and extracolonic cancers (Table 2) (e.g., endometrial, ovarian, gastric, urinary tract, renal cell, biliary and gallbladder, central nervous system, and small bowel) in multiple individuals in a family (9). There is a predominance of right-sided tumors with frequent synchronous and metachronous tumors. Furthermore, cancers that arise in HNPCC appear to have an advanced rate of malignant transformation. The adenomas found in patients with HNPCC are more often histologically advanced and have areas of high-grade dysplasia. The true incidence of HNPCC is not well known, but it is thought to account for 5% of all CRCs. The disorder is inherited according to an autosomal dominant mode and is highly penetrant, so that about 80% of gene carriers will develop the disease. Clinically the disorder is often divided into two syndromes, called Lynch I and Lynch II, based on the presence or absence of extraintestinal malignancies (Table 2) (10). In Lynch I syndrome, neoplastic manifestations are confined to the large bowel, whereas in Lynch II syndrome affected individuals develop colorectal carcinoma as well as extracolonic tumors in the endometrium, stomach, small intestine, brain, hepatobiliary tract, urinary tract, and ovary. These extraintestinal manifestions should not be omitted when taking a family history of a CRC patient and might lead to the suspicion of HNPCC in certain atypical families. Criteria The most widely used criteria for the clinical diagnosis of HNPCC are the Amsterdam criteria (Table 6). These were developed to be able to focus genetic analysis in high-risk families and to allow uniform international studies (11). In clinical practice, however, the criteria are sometimes too strict and might lead to the exclusion of some true HNPCC families. For example, it is now widely recognized that endometrial cancer can be substituted for CRC in one of the family members. It is important for the clinician to bear in mind the possible variations in the presentation of an HNPCC kindred, due to predominant extraintestinal malignancies, for example, small family size, or later age of onset (12). Consequently, the Bethesda criteria (13) were developed to create a set of clinical criteria that would be more sensitive than the Amsterdam criteria and could be used to identify patients who should be considered for genetic testing for HNPCC. The Bethesda criteria turned out to be too unspecific, and were followed by the more stringent Amsterdam II (14) criteria. Most recently, in 2004, a revised version of the Bethesda criteria was published (15) that has been developed in another attempt to use clinical and pathological features to more accurately identify

Familial Cancer Management

133

Table 6 Clinical Criteria for Hereditary Nonpolyposis Colorectal Cancer Amsterdam criteria I (all criteria must be met) One member diagnosed with colorectal cancer before age 50 years Two affected generations Three affected relatives, one of them a first-degree relative of the other two FAP should be excluded Tumors should be verified by pathologic examination Amsterdam criteria II (all criteria must be met) There should be at least three relatives with an HNPCC-associated cancer (colorectal cancer or cancer of the endometrium, small bowel, ureter, or renal pelvis) One should be a first-degree relative of the other two At least two successive generations should be affected At least one should be diagnosed before age 50 years FAP should be excluded in the colorectal cancer cases Tumors should be verified by pathologic examination Revised Bethesda guidelines for testing colorectal tumors for microsatellite instability (meeting features listed under any of the criteria is sufficient) Colorectal cancer diagnosed in a patient who is less than 50 years of age Presence of synchronous, metachronous colorectal cancers or associated extracolonic cancers (Note: endometrial, ovarian, gastric, hepatobiliary, pancreas or small bowel cancer or transitional cell carcinoma of the renal pelvis or ureter, brain tumors usually glioblastoma, sebaceous gland adeomas and keratoacanthomas) regardless of age Colorectal cancer diagnosed in one or more first-degree relatives with an HNPCC-related tumor, with one of the cancers being diagnosed under age 50. Colorectal cancer diagnosed in two or more first- or second-degree relatives with an HNPCC-related tumor, regardless of age Colorectal cancers with MSI-H histology (presence of tumor infiltrating lymphocytes, Crohn’s-like lymphocytic reaction, mucinous/signet-ring differentiation or medullary growth pattern) diagnosed in a patient who is less than 60 years of age Abbreviations: FAP, familial adenomatous polyposis; HNPCC, hereditary nonpolyposis colorectal cancer; MSI-H, microsatellite instability high.

individuals who should have microsatellite instability (MSI) or immunohistochemistry as a pretest to determine who should go on to mutation analysis of the MMR genes. If any of these criteria are met one should then perfom the tumor tests (MSI and immunohistochemistry); if either is positive, then perform genetic testing. The issue of diagnostic criteria for HNPCC is complex mainly because there are two endpoints. On the one hand, the criteria should be broad enough to allow a clinician to suspect HNPCC in a kindred, thus encouraging the necessary endoscopic screening in this family. On the other hand, mutation analysis for this disorder is laborious and time-consuming. This restricts the amount of testing that can be performed

134

Tejpar

and implies the need for stricter selection criteria. At the moment, even in kindreds fulfilling the Amsterdam criteria, the rate of mutation detection is only about 50%. Nonetheless, a clinician or clinical geneticist will be faced with quite a few kindreds where molecular analysis is inconclusive or was not performed due to unfulfilled selection criteria. These families are still at risk of carrying HNPCC mutations and developing CRC and should be managed as such. An algorithm that incorporates recommendations based on more recent studies as well as the American Gastroenterological Association guidelines is shown in Figure 1. In general, it is recommended that if a patient is being considered for genetic testing, which includes either MSI testing of the tumor from an affected individual or germline mutation testing, the clinician should consult with a specialist who routinely manages HNPCC family members to determine an optimal diagnostic strategy that incorporates the clinical features and psychosocial issues of the family. Microsatellite Instability The key molecular feature of HNPCC tumors, MSI, can also be integrated in the selection of possible HNPCC mutation carriers (16,17). MSI is recognized by the frequent occurrence of insertion and deletion mutations in

Figure 1 Genetic testing for hereditary nonpolyposis colorectal cancer. Abbreviations: MSI, microsatellite instability; HNPCC, hereditary nonpolyposis colorectal cancer; IHC, immunohistochemistry.

Familial Cancer Management

135

microsatellite repeats, which are variable numbers of repeating sequences of mononucleotides, dinucleotides, and trinucleotides found throughout the human genome (18). Clinically, MSI is identified by a test that is done on DNA extracted from fresh or paraffin-embedded, formalin-fixed tumor tissue obtained from the tumors of individuals suspected of having HNPCC. MSI is present less often in adenomas than in cancers, but large adenomas with dysplasia in HNPCC are often MSI. It is important to bear in mind, however, that MSI is not restricted to HNPCC cancers. Somatic inactivation of MLH1 by aberrant methylation of the MLH1 promoter occurs in approximately 15% of sporadic colon cancers. The older the patient with an MSI tumor at diagnosis, the less likely this is due to a germline MMR defect. Conversely only 90% of HNPCC tumors show MSI. This could be due to limitations in detection of MSI or due to low levels of instability such as in hMSH6 mutants. Immunohistochemistry for MLH1, MSH2, MSH6, and PMS2 has been proposed as a tumor-based assay to identify families that should be considered for germline mutation testing. Loss of expression of these proteins is correlated with the presence of MSI and may suggest which specific mismatch repair gene is altered in a particular patient. As immunohistochemistry has a less than 100% sensitivity for detecting MSI, and because this rate can vary according to more or less experienced hands, immunohistochemistry is usually performed in parallel and not as a substitute for MSI (19). Clinical Management Endoscopic screening is still the cornerstone in the management of HNPCC patients. When genetic testing is positive, surveillance is mandatory, but even in families with negative test results and strong clinical suspicion for HNPCC, regular follow-up is necessary. This approach has been proved to be useful. Indeed, CRC rates in HNPCC kindreds can be reduced by screening. Jarvinen et al. (20) showed a decrease of CRC incidence by 62% in asymptomatic screened family members at 50% risk for HNPCC. To be effective, screening must be continued for a lifetime, and should be tailored to the natural history of the malignancies occuring in HNPCC (21). The current screening guidelines, as issued by the international collaborative group on HNPCC, are described in Table 7. These recommendations have been based on several studies. For example, the interval between colonoscopies was prompted by the observation of more rapid adenoma-to-carcinoma progression in HNPCC. International collaborative trials addressing these questions are still ongoing and might further improve the recommendations. In view of the high penetrance of the disease and of the high incidence of metachronous tumors, more aggressive measures such as prophylactic total colectomy in mutation carriers have been proposed. Prophylactic hysterectomy and oophorectomy have been proposed due to lack of validated

136

Tejpar

Table 7 Recommendations for Hereditary Nonpolyposis Colorectal Cancer Screening Site

Procedure

Colon Endometrium and ovaries

Stomacha Urinary tracta a

Colonoscopy Gyn. examination Transvaginal ultrasound CA 125 Gastroscopy Sonography Urine analysis

Lower age limit (years)

Interval (years)

20–25 30–35

1–2 1–2

30–35 30–35

1–2 1–2

Only if present in family.

screening procedures in the prevention of endometrial and ovarian cancer. Such measures, however, are still a controversial issue, and have to be debated in the individual setting. Genetic Testing A suggested approach (Fig. 1) for genetic testing in HNPCC is to go directly to mutation analysis of the MMR genes if the Amsterdam I or II criteria are met. If the revised Bethesda criteria are met, then proceed to MSI and immunohistochemistry, and if either one is positive, proceed to mutation analysis. Alternatively, if no tumor tissue is available in individuals meeting the Bethesda criteria, genetic testing should be performed. Genetic testing is usually recommended for family members at risk around age 20 to 25, as clinical screening would start around that age. Identifying which MMR gene is responsible can impact on the management, for example, the relative risk of gastric cancer, ovarian cancer, and cancer of the urinary tract has been shown to be higher in patients with mutations in MSH2 compared with MLH1. Similarly, women with MSH6 mutations seem to be more likely to develop endometrial cancer (22). Germline mutations in MSH6 are also associated with an atypical form of HNPCC that often does not meet Amsterdam criteria because of a later age at onset and tumors that display low-level MSI. In HNPCC, in general, there is improved survival for individuals with colon cancer even when corrected for stage. Germline mutations in MSH2 and MLH1 have been found in 45% to 70% of families that meet strict clinical criteria for HNPCC, and thus these two genes account for most HNPCC cases. HMSH6 and PMS2 mutations account for a small number of families. There is large spectrum of mutations in all these genes,

Mismatch repair genes

HNPCC

Peutz–Jeghers

Adenomatous

Adenomatous except stomach, fundic cystic glands

Polyp histology

Cowden’s

PTEN

Juvenile,

Juvenile Juvenile SMAD4, BMPR1A polyposis

Peutz– STK11 Jeghers syndrome

APC

FAP

Syndrome

Relevant gene

Benign extracolonic features

Malignant extracolonic features. cumm. life-time risk

(Continued )

Desmoid tumors: 8% males, Duodenal cancer: 3–5%, Stomach: 23–100%, 100% pancreatic: 2%, 15% females; fundic (39 years) duodenum: papillary thyroid: 2%, glands: 50%; epidermal AFAP: 50–90%, gastric: 0.5%, CNS: 1%, cysts, osteoma: 17%; 80% jejunum: 50%, hepatoblastoma: 1.6% CHRPE, dental (50 years) ileum: 20%, abnormality, colon: 100% nonfunctioning adrenal adenoma: 13% Uterine: 40%, ovarian: 80% Keratoacanthomas, Colon: 2–3-fold 12%, gastric: 13%, sporadic rate (44 years) sebaceous adenomas, urinary: 10%, biliary: epitheliomas 10–20%, renal, CNS, small bowel: rare 39% Pancreatic: 36%, gastric: Orocutaneous melanin Stomach: 24%, (46 years) 29%, small bowel: 19%, pigment spots small bowel: breast: 54%, ovarian: 90%, rectum: 21%, lung 15%, sex 24%, colon: 27% cord: 10–20% Stomach and small 9–68% Stomach and duodenum Macrocephaly, bowel: may (34 years) combined: 20% hypertelorism:20%, occur, colon: congenital anomalies usually Esophagus: 66%, About 10% Facial tricholemmomas, Follicular or papillary

CRC risk Polyp distribution (mean age of diagnosis) and frequency

Table 8 Features of Inherited Colorectal Cancers Syndromes

Familial Cancer Management 137

lipomas, inflammatory

Polyp histology stomach: 75%, duodenum: 37%, colon: 66%

CRC risk Polyp distribution (mean age of diagnosis) and frequency oral papillomas, multinodular goiter, fibrocystic breast disease, cerebellar gangliocytomatosis, hemangiomas

Benign extracolonic features

thyroid: 3–10%, breast: 25–50%, uterine 5–10%

Malignant extracolonic features. cumm. life-time risk

Abbreviations: CRC, colorectal cancer; FAP, familial adenomatous polyposis; APC, adenomatous polyposis coli; AFAP, attenuated FAP; HNPCC, hereditary nonpolypois colorectal cancer; CNS, central nervous system; CHRPE, congenital hypertrophy of the retinal pigment epithelium; PTEN, phosphatase and tensin homolog.

Relevant gene

Features of Inherited Colorectal Cancers Syndromes (Continued )

syndrome

Syndrome

Table 8

138 Tejpar

Familial Cancer Management

139

and very often the results obtained from mutation analysis may be inconclusive because the identified base pair change is a variant of unknown significance. Genetic testing in HNPCC is still complicated by factors such as patient selection, difficulty in detecting the heterogeneous mutations, and because not all the responsible genes nor their disease penetrance are known. A positive test result can bring about a lot of anxiety and psychological; on the other hand, it does allow a person to make informed decisions about the future. Genetic testing should always be performed in specialized genetic centers where correct pre- and post-test counseling and psychosocial follow-up can be offered. High-risk kindreds who cannot be offered molecular diagnosis will benefit from risk assessment and information on screening procedures by the genetic counselor.

PEUTZ–JEGHERS SYNDROME Peutz–Jeghers Syndrome (PJS) is a rare autosomal dominant disorder that is generally recognized by the association of melanocytic macules of the lips and buccal mucosa, and digits and GI hamartomatous polyps (23). There is still some debate on the matter, but a clinical diagnosis of PJS can probably be made if two Peutz–Jeghers polyps are found in the GI tract or if one Peutz–Jeghers polyp in the GI tract is found in association with classic Peutz– Jeghers pigmentation, or a family history of the syndrome. In addition to an increased risk of colon cancer, PJS is associated with an increased risk of various neoplasms, including sex cord tumors, pancreatic cancer, lung cancer, breast cancer, uterine cancer, melanoma, and gastric cancer (Table 8) (24,25). In light of this increased cancer risk in multiple tissue types, an extensive surveillance regimen has been recommended including upper and lower endoscopy, breast examination, and some form of surveillance for pancreatic and gynecologic malignancies. Guidelines, however, differ considerably over when to initiate screening, intervals between screening, and which techniques to use. The optimal surveillance strategy is unclear and to date the efficacy of intensive surveillance for cancer associated with PJS has not been established. Recommendations from the St. Mark’s Polyposis Registry consist of annual assessment including a full blood count, pelvic ultrasound in females, testicular ultrasound in males, and pancreatic ultrasound in all individuals. Every two years, patients should have an upper and lower endoscopy and some form of examination of the small bowel, preferably an enteroclysis. Patients should be screened for breast cancer according to the clinical recommendations for other high-risk individuals. Patients should be encouraged to perform monthly breast self-examination; clinical exams should be performed annually, beginning in the late teen years or as concerns arise. Mammography should begin at the age of 25 years. Pap smears should be done at least every three years (26).

Colon cancer risk (average age at diagnosis)

Colonoscopy every 3 years, starting late teens or at symptoms if earlier Colonoscopy every 3 years, starting early teens or at symptoms if earlier None established

Annual sigmoidoscopy beginning 12 years: colonoscopy late teens if AFAP Annual colonoscopy starting age 25

Colon cancer screening recommendations

Abbreviations: FAP, familial adenomatous polyposis; APC, adenomatous polyposis coli; AFAP, attenuated FAP; HNPCC, hereditary nonpolyposis colorectal cancer; MYH, MutY associated polyposis; PTEN, phosphatase and tensin homolog.

9%

Typical colonic polyps

PTEN (80–90%)

Cowden 25 years

5 or more juvenile polyps 9–68% (34 years)

Early teens or at SMAD4, BMPR1A (50%) symptoms if earlier

Any Peutz–Jeghers polyp 39% (46 years) or pigmentation

20 or more adenomatous FAP 100% (39 years), polyps AFAP 80% (50 years) Amsterdam II or revised 80% (44 years) Bethesda

Clinical signs that should initiate genetic testing

Juvenile polyposis

Late teens or at symptoms if earlier

20–25 years

MLH1, MSH2 (50–70%), MSH6, PMS2 STK11/LKB1 (50–60%)

HNPCC

Peutz–Jeghers

10–12 years

APC (80–90%) MYH

Age to consider predictive genetic testing if mut. known

FAP

Disorder

Causative gene and frequency mut. found

Table 9 Gene Frequency, Genetic Testing, and Management Guidelines

140 Tejpar

Familial Cancer Management

141

About 50% of patients with clinical Peutz–Jeghers disease are found to have a mutation in the STK11 gene (Table 9) (27). Additional genes probably exist, but have not yet been identified. Genetic testing for PJS can be initiated when a typical Peutz–Jeghers polyp is found (bands of arborizing smooth muscle in the lamina propria) or characteristic melanin pigmentation is present. In general, referring such patients to physicians with extensive experience with these disorders is recommended.

JUVENILE POLYPOSIS Juvenile polyposis (JPS) is an autosomal dominant syndrome that is associated with an increased risk of CRC. Classic juvenile polyps consist of stromal elements with a normal epithelial layer (dilated, mucus-filled cysts and abundant lamina propria that is lacking in smooth muscle) and are distinct from both adenomatous polyps and those of the Peutz–Jeghers type. Although solitary juvenile polyps are common in children, juvenile polyposis is marked by the presence of many polyps either in the colon or throughout the GI tract (28). Patients with juvenile polyposis typically present with benign complications, including GI bleeding, obstruction, and intussusception, in the first three decades of life. Malignant degeneration and complications start after the fourth decade. JPS is associated with a 10% to 38% lifetime risk of colon cancer, and the cancers appear to arise from adenomatous components present in some juvenile polyps (Table 8) (29). Colonoscopy every one to two years starting age 15 to 18 is advised, with removal of the larger polyps to be examined for adenomatous components. In addition, the lifetime risk of gastric and duodenal cancer appears to be approximately 15% to 21%. Thus the upper GI tract should be surveyed every one to two years and the larger polyps removed (30). In light of the recently found association between juvenile polyposis and hereditary hemorrhagic telangiectasia in patients with SMAD4 mutations, JPS patients with these mutations should be actively screened for the vascular lesions, especially occult arteriovenous malformations in visceral organs that may otherwise present acutely with serious consequences (31). Although juvenile polyps are also a feature of other genetic syndromes, juvenile polyposis is a distinct disorder that is caused by mutations in either SMAD4 or BMPR1A (32). SMAD4 and BMPR1A (33) genes account for about 50% of juvenile polyposis cases (Table 9). Mutations in other genes involved in TGF-b signaling are likely to be involved in the remaining families (34). Genetic testing is recommended in the first decade in families that have the disorder, since children often become symptomatic (rectal bleeding, abdominal pain, diarrhea) early on.

142

Tejpar

Table 10 International Cowden Consortium Operational Criteria for the Diagnosis of Cowden Syndrome Ver 2000 Pathognomonic criteria

Major criteria

Minor criteria

Mucocutaneous lesions Breast cancer

Other thyroid lesions (e.g., goiter) Trichilemmomas, facial Thyroid cancer, especially Mental retardation follicular thyroid cancer (IQ75) Acral keratoses, Macrocephaly (occipital frontal Hamartomatous papillomatous lesions circumference 97th percentile) intestinal polyps Fibrocystic disease of the breast Lipomas, fibromas Mucosal lesions Lhermitte–Duclos disease, defined as presence of a cerebellar dysplastic gangliocytoma Endometrial carcinoma Genitourinary tumors (e.g., uterine fibroids, renal cell carcinoma) or genitourinary malformation An operational diagnosis of Cowden syndrome is made if an individual meets any one of the following criteria: Pathognomic mucocutaneous lesions alone if there are: Six or more facial papules of which three or more must be trichilemmoma, or Cutaneous facial papules and oral mucosal papillomatosis, or Oral mucosal papillomatosis and acral keratoses, or Six or more palmo plantar keratoses Two major criteria but one must be either macrocephaly or Lhermitte–Duclos disease One major and three minor criteria (Continued )

Familial Cancer Management

143

Table 10 International Cowden Consortium Operational Criteria for the Diagnosis of Cowden Syndrome Ver 2000 (Continued ) Pathognomonic criteria

Major criteria

Minor criteria

Four minor criteria In a family in which one individual meets the diagnostic criteria for Cowden syndrome, other relatives are considered to have a diagnosis of Cowden syndrome if they meet any of the following criteria: A pathognomonic mucocutaneous lesion Any one major criterion with or without minor criteria Two minor criteria

COWDEN SYNDROME OR PTEN HAMARTOMA SYNDROME Germline mutation of PTEN leads to the development of the related hereditary cancer predisposition syndromes, Cowden’s disease and Bannayan–Zonana syndrome, wherein breast and thyroid cancer incidence is elevated (35). Hamartomas involve the skin, intestine, breast, and thyroid gland (Table 8). Eighty percent of patients present with dermatologic manifestations. The diagnostic criteria for Cowden’s disease are summarized in Table 10 (36). Only 35% of patients who meet the diagnostic criteria for Cowden’s disease have GI polyposis. Patients meeting the clinical criteria of the disease are found to have identifiable PTEN mutations in as many as 80% to 90% of cases (Table 9). The majority of patients with Cowden’s disease will have some form of benign thyroid or breast disease (37). In addition, the projected lifetime risk of thyroid malignancy is 10% and of breast malignancy is approximately 30% to 50% (Table 9). Clinical breast examinations are recommended annually for women beginning at age 25, and annual mammography starting at age 30 to 35. Breast cancer can also occur in men. A baseline thyroid ultrasound is also recommended at age 18, followed by an annual thyroid ultrasound thereafter. Women with Cowden syndrome should also undergo endometrial screening involving annual blind biopsies starting at age 35 to 40, and annual endometrial ultrasound after menopause, with biopsy of suspicious lesions. Annual urine analysis for the detection of renal cell carcinoma is also recommended along with annual urine cytology and renal ultrasound if there is a family history of renal cancer. Typical hamartamatous polyps can occur in the small and large bowel, with a low lifetime risk of CRC. There are no specific guidelines for endoscopic screening, but regular upper and lower GI endoscopy with biopsy of the polyps seems prudent.

144

Tejpar

FAMILIAL GASTRIC CANCER Hereditary diffuse gastric cancer arises from mutations of the E-cadherin gene that are inherited in an autosomal dominant fashion with a high penetrance over 70% (38). Only ‘‘diffuse type’’ histology arises, and this syndrome does not account for familial clustering of intestinal-type gastric cancer. Management of these families is very complex, but because of the very high penetrance of the disease, prophylactic surgery should be discussed. In general, referring such patients to physicians with extensive experience with these disorders is recommended. FAMILIAL PANCREATIC CANCER On the basis of detailed family history it is estimated that 15% to 38% of pancreatic ductal adenocarcinomas arise owing to inherited susceptibility. The genetic etiology relates to several syndromes in which pancreatic cancer is sometimes observed and also to additional families in which no specific syndrome can be identified. These syndromes include familial atypical multiple mole-melanoma (FAMM), breast–ovarian cancer syndrome, von Hippel-Lindau syndrome, HNPCC, PJS, ataxia-telangiectasia, and pure pancreatic cancer kindreds. In about 15% an inherited germline mutation can be identified. The genes that are known to give rise to familial pancreatic cancer include the BRCA2, p16, STK11, MLH1, FancC, and FancG genes and the cationic trypsinogen gene (39). Currently, the gene identified most commonly as mutated in familial pancreatic cancer families is the BRCA2 gene (40); 17% to 19% of patients with pancreatic cancer and two or more affected relatives carry a germline mutation in the BRCA2 gene. Germline BRCA2 mutations also have been identified in 5% to 10% of patients who present with pancreatic cancer without a family history of the disease. Presently, the optimal approach to screening this high-risk population for early pancreatic cancer is unknown and still under study. Minimally invasive tests and procedures such as Endoscopis Ultrasonography (EUS), magnetic resonance imaging/magnetic resonance cholangiopancreatography, computed tomography (CT), positron emission tomography, biochemistry, and CA19–9 are good candidates (41). In general, referring such patients to physicians with extensive experience with these disorders is recommended. HEREDITARY MIXED POLYPOSIS This apparently rare hereditary polyposis syndrome is characterized by the presence of a mixture of adenomatous polyps, hyperplastic polyps, juvenile polyps, and polyps with mixed histology in one individual. CRC has been reported in these kindreds at varying ages (42). The causative gene for this syndrome has not been identified, and there are no available tests to identify

Familial Cancer Management

145

asymptomatic carriers of hereditary mixed polyposis syndrome (HMPS) (43). Any individual suspected of having HMPS should be referred to a center with physicians who are specialized in polyposis syndromes and colon cancer family syndromes to obtain current management recommendations.

FAMILIAL COLORECTAL CANCER In addition to the well-recognized syndromes described earlier (FAP, HNPCC), clusters of CRCs occur in families much more often than would be expected by chance. Postulated reasons for this increased risk include ‘‘mild’’ and undetected mutations of APC and mismatch repair genes, as well as yet unknown polymorphisms in genes involved in nutrient or carcinogen metabolism (44). Candidate alleles that have been shown to be associated with modest increased frequencies of colon cancer include the APC, I1207K, and E1317Q polymorphisms and loss of imprinting of the IGF2 gene. However, none of these alleles have been characterized well enough to support its routine use in a clinical setting at this time. This familial clustering in about 10% to 20% of CRCs has implications for screening because the immediate family members of a patient with apparent sporadic CRC have a twofold to threefold increased risk of the disease (45). The magnitude of the risk depends on the age at diagnosis of the index case, the degree of kinship of the index case to the at-risk case, and the number of affected relatives. Thus, in addition to screening the easily identifiable high-risk groups such as FAP and HNPCC, care should be taken to recognize intermediate-risk patients and to provide them with appropriate screening recommendations (Fig. 2) (46). Because the molecular basis and the natural history of these intermediate-risk patients is largely unknown, screening recommendations are as yet more empirical. Future research into the molecular basis of these syndromes should allow more definite risk evaluation. Screening strategies have been developed to address the familial risk of commonly observed colon cancer. Screening recommendations are empiric and combine the known effectiveness of available screening tools with the observed risks associated with family history. If a person has a first-degree relative with colon cancer, average-risk colon cancer screening is recommended, but starting at age 40 years. The decreased age is given because the risk at age 40 years for those with an affected first-degree relative is similar to the risk at age 50 years for the general population. An individual with two first-degree relatives affected with colon cancer or one first-degree relative diagnosed under the age of 60 years should have colonoscopy beginning at age 40 years, or 10 years younger than the earliest case in the family. Colonoscopy should be repeated every five years if negative. An even stronger family history of colon cancer should suggest the consideration of one of the inherited syndromes of colon cancer.

146

Tejpar

Figure 2 Screening according to family history. Abbreviations: FAP, familial adenomatous polyposis; HNPCC, hereditary nonpolyposis colorectal cancer; CRC, colorectal cancer. Source: Ref. 46.

ASPECTS OF GENETIC TESTING FOR GI CANCER SUSCEPTIBILITY Genetic counseling should be a component of any genetic testing to ensure that optimal family histories are obtained and that appropriate risk assessment and test selection are performed (47). Because of the complexity of this rapidly evolving and vast field, a multidisciplinary approach is necessary, bringing together gastroenterologists, other organ specialists, geneticists, oncologists, and psychologists. Issues that become relevant during genetic testing are patient’s perception of risk, psychological stability and concerns, coping with anger, anxiety, responsibility, guilt, stress, self-image issues, survivor guilt, and optimal screening and prevention strategies (48). These questions need to be addressed by an experienced and multidisciplinary team. Genetic testing is indicated for each of the inherited syndromes when certain features of the syndrome are present (Table 9) (49). DNA is extracted from white blood cells and the relevant genes are analyzed to detect disease-causing mutations (50). The success of finding a mutation in a person in a family clinically identified as having one of the syndromes is also given in Table 9. Finding the mutation confirms the clinical diagnosis. Failure to find a mutation in a person suspected to have an inherited condition, however, does not rule out the syndrome. There are a number of technical reasons why mutations cannot always be found. In this case, further testing in relatives of the index person tested is not useful because

Familial Cancer Management

147

mutations would not be found in them either. If a mutation cannot be found in that first person, the genetic test is said to be ‘‘uninformative.’’ Cancer screening must be continued on all family members. But if a relevant mutation is found in the first person, or index case, then other family members can undergo ‘‘mutation specific’’ genetic testing. Only the exact mutation found in the first family member is tested for in other family members. Special screening can then be directed to those who have the disease-causing mutation, while those who do not have the mutation need no more than average-risk screening. In reality, genetic testing in a cancer syndrome very often leads to inconclusive results either because the mutation is not found, because a mutation of unknown significance is found, or because the syndrome was misclassified and the correct genes were not searched. Despite best efforts of everyone involved this scenario remains frequent. Current basic science knowledge on the pathogenesis of many of these syndromes is sorely lacking. Often mutations are found in an affected individual, but due to lack of knowledge on the exact function of the encoded protein, or lack of assays to test the effect of the mutation, the mutation will be classified as of unknown significance. No genetic testing or clinical guidelines can be based on a mutation of unknown significance. Another problem is posed by small families, lack of information on relatives or untraceable relatives, which can lead to misclassification of a syndrome. An optimal approach would include a registry-based, multidisciplinary approach, ensuring correct data gathering, verification, and updating over time. Without this many familial syndromes will go unrecognized and not be managed or tested correctly. A slightly provocative statement is that genetic testing should be considered the ‘‘icing on the cake’’ for many families. In these families in which a mutation is identified, genetic testing confirms the clinical suspicion, reinforces the surveillance recommendations, and allows predictive mutation-specific testing in at-risk individuals. However, the most important task is already performed by the clinicians recognizing the syndrome, and the most important remaining task is the implementation of tailored surveillance exams in these patients. The continuous need for motivating the patients and providing correct follow-up of the lifelong surveillance exams is a formidable task for the clinicians. For the majority of families, genetic testing will not be informative. For these families, there is still often the misconception that if no mutation is found, no hereditary cancer predisposition exists. Nothing is less true, and letting the guard down in the management of these families can lead to tragic situations. However, without molecular confirmation of the diagnosis, and with only often empiric and loosely defined risk assessments available, daily management of these families can be very difficult. Resources directed toward continuing education of physicians and providing administrative and other support are well directed. For these families the improvement in outcome and the potential lives saved will depend solely on the dedication of patients and physicians working together.

148

Tejpar

REFERENCES 1. Lynch HT, de la Chapelle A. Hereditary colorectal cancer. N Engl J Med 2003; 348(10):919–932. 2. Al-Tassan N, Chmiel NH, Maynard J, et al. Inherited variants of MYH associated with somatic G: C 3 T:A mutations in colorectal tumors. Nat Genet 2002; 30:227–232. 3. Sieber OM, Lipton L, Crabtree M, et al. Multiple colorectal adenomas, classic adenomatous polyposis, and germ-line mutations in MYH. N Engl J Med 2003; 348:791–799. 4. Jones S, Emmerson P, Maynard J, et al. Biallelic germline mutations in MYH predispose to multiple colorectal adenoma and somatic G: C 3 T:A mutations. Hum Mol Genet 2002; 11:2961–2967. 5. Bulow S, Bjork J, Christensen IJ, et al. DAF Study Group. Duodenal adenomatosis in familial adenomatous polyposis. Gut 2004; 53(3):381–386. 6. Brensinger JD, Laken SJ, Luce MC, et al. Variable phenotype of familial adenomatous polyposis in pedigrees with 3’ mutation in the APC gene. Gut 1998; 43:548–552. 7. Liu B, Parsons R, Papadopoulos N, et al. Analysis of mismatch repair genes in hereditary non-polyposis colorectal cancer patients. Nat Med 1996; 2:169–174. 8. Peltomaki P, Vasen HF. Mutations predisposing to hereditary nonpolyposis colorectal cancer: database and results of a collaborative study. The International Collaborative Group on Hereditary Nonpolyposis Colorectal Cancer. Gastroenterology 1997; 113:1146–1158. 9. Samowitz WS, Curtin K, Lin HH, et al. The colon cancer burden of genetically defined hereditary nonpolyposis colon cancer. Gastroenterology 2001; 121: 830–838. 10. Marra G, Boland C. Hereditary nonpolyposis colorectal cancer: the syndrome, the genes, and historical perspective. J Natl Cancer Inst 1995; 87:1114–1125. 11. Wijnen J, Khan PM, Vasen H, et al. Hereditary nonpolyposis colorectal cancer families not complying with the Amsterdam criteria show extremely low frequency of mismatch-repair-gene mutations. Am J Hum Genet 1997; 61:329–335. 12. Wijnen JT, Vasen HF, Khan PM, et al. Clinical findings with implications for genetic testing in families with clustering of colorectal cancer. N Engl J Med 1998; 339:511–518. 13. Rodriguez-Bigas MA, Boland CR, Hamilton SR, et al. A National Cancer Institute Workshop on Hereditary Nonpolyposis Colorectal Cancer Syndrome: meeting highlights and Bethesda guidelines. J Natl Cancer Inst 1997; 89: 1758–1762. 14. Park JG, Vasen HF, Park YJ, et al. Suspected HNPCC and Amsterdam criteria II: evaluation of mutation detection rate, an international collaborative study. Int J Colorect Dis 2002; 17:109–114. 15. Umar A, Boland CR, Terdiman JP, et al., Revised Bethesda Guidelines for hereditary nonpolyposis colorectal cancer (Lynch Syndrome) and microsatellite instability. J Natl Cancer Inst 2004; 96:261–268. 16. Terdiman JP, Gum JR Jr., Conrad PG, et al. Efficient detection of hereditary nonpolyposis colorectal cancer gene carriers by screening for tumor

Familial Cancer Management

17. 18.

19.

20.

21.

22.

23. 24. 25. 26.

27. 28. 29. 30.

31.

32.

33.

149

microsatellite instability before germline genetic testing. Gastroenterology 2001; 120:21–30. Jass JR, Cottier DS, Jeevaratnam P, et al. Diagnostic use of microsatellite instability in hereditary nonpolyposis colorectal cancer. Lancet 1995; 346:1200–1201. Boland C, Thibodeau S, Hamilton S, et al. National Cancer Institute workshop on microsatellite instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 1998; 58:5248–5257. Lindor NM, Burgart LJ, Leontovich O, et al. Immunohistochemistry versus microsatellite instability testing in phenotyping colorectal tumors. J Clin Oncol 2002; 20:1043–1048. Jarvinen HJ, Aarnio M, Mustonen H, et al. Controlled 15-year trial on screening for colorectal cancer in families with hereditary nonpolyposis colorectal cancer. Gastroenterology 2000; 118(5):829–834. Vasen HF, Wijnen JT, Menko FH, et al. Cancer risk in families with hereditary nonpolyposis colorectal cancer diagnosed by mutation analysis. Gastroenterology 1996; 110:1020–1027. Miyaki M, Konishi M, Tanaka K, et al. Germline mutation of MSH6 as the cause of hereditary nonpolyposis colorectal cancer (Letter). Nat Genet 1997; 17:271–272. Tomlinson IP, Houlston RS. Peutz–Jeghers syndrome. J Med Genet 1997; 34:1007–1011. Giardiello FM, Brensinger JD, Tersmette AC, et al. Very high risk of cancer in familial Peutz–Jeghers syndrome. Gastroenterology 2000; 119:1447–1453. Lim W, Olschwang S, Keller JJ, et al. Relative frequency and morphology of cancers in STK11 mutation carriers. Gastroenterology 2004; 126(7):1788–1794. Wirtzfeld DA, Petrelli NJ, Rodriguez-Bigas MA. Hamartomatous polyposis syndromes: molecular genetics, neoplastic risk, and surveillance recommendations. Ann Surg Oncol 2001; 8(4):319–327. Hemminki A, Markie D, Tomlinson I, et al. A serine/threonine kinase gene defective in Peutz–Jeghers syndrome. Nature 1998; 391:184–187. Woodford-Richens K, Bevan S, Churchman M, et al. Analysis of genetic and phenotypic heterogeneity in juvenile polyposis. Gut 2000; 46:656–660. Jass JR, Williams CB, Bussey HJ, Morson BC. Juvenile polyposis: a precancerous condition. Histopathology 1988; 13:619–630. Dunlop MG. Guidance on gastrointestinal surveillance for hereditary non-polyposis colorectal cancer, familial adenomatous polypolis, juvenile polyposis, and Peutz–Jeghers syndrome. Gut 2002; 51(suppl 5):V21–V27. Gallione CJ, Repetto GM, Legius E, et al. A combined syndrome of juvenile polyposis and hereditary haemorrhagic telangiectasia associated with mutations in MADH4 (SMAD4). Lancet 2004; 363(9412):852–859. Howe JR, Bair JL, Sayed MG, et al. Germline mutations of the gene encoding bone morphogenetic protein receptor 1A in juvenile polyposis. Nat Genet 2001; 28:184–187. Sayed MG, Ahmed AF, Ringold JR, et al. Germline SMAD4 or BMPR1A mutations and phenotype of juvenile polyposis. Ann Surg Oncol 2002; 9: 901–906.

150

Tejpar

34. Zhou XP, Woodford-Richens K, Lehtonen R, et al. Germline mutations in BMPR1A/ALK3 cause a subset of cases of juvenile polyposis syndrome and of Cowden and Bannayan–Riley–Ruvalcaba syndromes. Am J Hum Genet 2001; 69:704–711. 35. Grady WM. Genetic testing for high-risk colon cancer patients. Gastroenterology 2003; 124(6):1574–1594. 36. Pilarski R, Eng C. Will the real Cowden syndrome please stand up (again)? Expanding mutational and clinical spectra of the PTEN hamartoma tumour syndrome. J Med Genet 2004; 41(5):323–326. 37. Eng C. Constipation, polyps, or cancer? Let PTEN predict your future. Am J Med Genet 2003; 122A(4):315–322. 38. Brooks-Wilson AR, Kaurah P, Suriano G, et al. Germline E-cadherin mutations in hereditary diffuse gastric cancer: assessment of 42 new families and review of genetic screening criteria. J Med Genet. 2004; 41(7):508–517. 39. Vimalachandran D, Ghaneh P, Costello E, Neoptolemos JP. Genetics and prevention of pancreatic cancer. Cancer Control 2004; 11(1):6–14. 40. Hahn SA, Greenhalf B, Ellis I, et al. BRCA2 germline mutations in familial pancreatic carcinoma. J Natl Cancer Inst 2003; 95(3):214–221. 41. Canto MI, Goggins M, Yeo CJ, et al. Screening for pancreatic neoplasia in highrisk individuals: an EUS-based approach. Clin Gastroenterol Hepatol 2004; 2(7):606–621. 42. Whitelaw SC, Murday VA, Tomlinson IP, et al. Clinical and molecular features of the hereditary mixed polyposis syndrome. Gastroenterology 1997; 112: 327–334. 43. Thomas HJ, Whitelaw SC, Cottrell SE, et al. Genetic mapping of hereditary mixed polyposis syndrome to chromosome 6q. Am J Hum Genet 1996; 58:770–776. 44. Johns LE, Houlston RS. A systematic review and meta-analysis of familial colorectal cancer risk. Am J Gastroenterol 2001; 96:2992–3003. 45. Burt RW. Colon cancer screening. Gastroenterology 2000; 119:837–853. 46. Winawer SJ, Fletcher R, Rex D, et al. Colorectal cancer screening and surveillance: Clinical guidelines and rationale—update based on new evidence. Gastroenterology 2003; 124:544–560. 47. Giardiello FM, Brensinger JD, Petersen GM. AGA technical review on hereditary colorectal cancer and genetic testing. Gastroenterology 2001; 121:198–213. 48. Giardiello FM, Brensinger JD, Petersen GM. American Gastroenterological Association medical position statement: hereditary colorectal cancer and genetic testing. Gastroenterology 2001; 121:195–197. 49. Lindor NM. Recognition of genetic syndromes in families with suspected hereditary colon cancer syndromes. Clin Gastroenterol Hepatol 2004; 2(5): 366–375. 50. Burt RW, Winawer SJ, Bond JH, Levin B, Sandler RS. Preventing Colorectal Cancer: A Clinician’sGuide, AGA monograph http://www.gastro.org/edu/ CRCpreventionMonograph.pdf.

6 The Surgical Principles of Managing Colorectal Cancer Ian R. Daniels and Richard J. Heald Pelican Cancer Foundation, North Hampshire Hospital, Basingstoke, U.K.

INTRODUCTION Of infinite importance is the dissemination of cancer cells through the lymphatic channels, so that a knowledge of the lymphatic system is essential to the performance of any radical operation for cancer. —W.E. Miles, 1939 The key to the successful management of colorectal cancer is the understanding of the embryological origin of the colon and rectum. Surgery, performed along embryological planes, is the key to achieving a cure in colorectal cancer. With the development of accurate preoperative staging, recognition that the disease has spread into an area of tissue of differing origin influences the use of preoperative therapy or the performance of a more radical resection. Following resection, the pathologist influences further management by assessing the completeness of resection and identifying other factors that impact on the patient’s prognosis. In this chapter, we will address the surgical management of colorectal cancer, the lessons that have been learned during the advancement of treatment for rectal cancer, the surgical principles, and how they may be applied to colonic cancer. While it has been uniquely demonstrated, in the case of rectal cancer, that inadequate surgery leads to local failure and, therefore, distressing symptoms for the patient, the improvements we have now seen 151

152

Daniels and Heald

may be transferred to the treatment of colonic cancer. Indeed, in Sweden, a nation that has led the improvements in developing the use of radiotherapy for rectal cancer, for the first time in any nation, survival of patients with rectal cancer is better than with colon cancer. THE ‘‘EMBRYOLOGICAL APPROACH’’ TO RECTAL CANCER An understanding of the embryological development of the rectum underlies the principle to the successful management of rectal cancer. The artery of the rectum, the inferior mesenteric artery, is the integral vessel of the hindgut, the distal portion of the embryological gastrointestinal tract. Its relationship to the cloaca and the corresponding anatomy allow the surgeon to understand the basics of rectal cancer surgery. The mesorectal fascia and the contained mesorectum (the fatty pad surrounding the rectal muscle wall) represent the mesentery of the hindgut. This contains the vessels and lymphatics of the primitive hindgut. Containment of the tumor within this mesorectal package and the complete removal of the intact package during surgery are the key elements in the prevention of local recurrence (LR) and achieving a cure. Tumor spread beyond the mesorectal envelope or incomplete surgical excision leads to the development of LR in the patient. THE MULTIDISCIPLINARY APPROACH While the patient usually presents to the surgeon, the diagnosis and management of colorectal cancer is multidisciplinary. After recognition of symptoms, or clinical suspicion, the patient will be referred for radiological and/or endoscopic assessment. Prior to surgery, assessment of local disease spread and metastatic disease is performed. In rectal cancer, regimes of preoperative therapy have been advocated to downstage/downsize the tumor prior to surgery. In the colon, tumor invasion into other structures provides the surgeon with advance warning of the potential problems at operation. THE PRINCIPLES OF RECTAL CANCER EXCISION: TOTAL MESORECTAL EXCISION All carcinomas of the lower sigmoid and upper rectum are tabooed by all practical surgeons . . . on account of their inaccessibility. All are abandoned without hope to linger on for a few months until death relieves them of their loathsome condition. —H. Maunsell, 1892 In the century following Maunsell’s (1) observation, the management of rectal cancer was revolutionized. As surgeons, we have taken up the challenge, aiming to provide more sphincter-preserving surgery, better function, lower LR rates, minimal urogenital morbidity, and improved overall survival.

Surgical Principles of Managing Colorectal Cancer

153

However, at the beginning of the 20th century rectal cancer was treated, if at all, by attempted surgical resection. The standard operation for rectal cancer, in most parts of the western world, throughout the whole of the last century was the abdominoperineal excision, as advocated by the English surgeon Ernest Miles in 1908 (2). As recently as 1993, Murray and Veidenheimer (3) described it as the ‘‘gold standard’’ by which all other operations must be judged, not only for carcinomas of the distal third of the rectum but also for bulky tumors of the middle third. It is notable that the technique described by Miles had become standard practice soon after Miles’ original paper but its acceptance was not challenged. This must be contrasted with the work of Henri Hartmann (4,5), which was largely disregarded, leading to the additional morbidity involved in the extirpation of many anal canals. Miles’ theory behind the development of the abdominoperineal excision was based upon observations on the spread of rectal cancer. Miles recognized the upward spread of the lymphatics to the root of the inferior mesenteric artery and dismissed earlier work suggesting submucosal spread of carcinoma within the rectum concluding that: . . . Observations lead us to believe that detached cancer cells pass through the bowel wall somewhat rapidly by means of the intramural lymphatic system and, gaining the external lymphatics, give rise to extramural metastases scattered over a fairly wide area, long before the muscular coat has been penetrated by direct extension of the growth. While this may occur from time to time, we now realize that a relatively orderly progression from local to lymphovascular, and finally distant spread, is a much more common pattern. Miles felt that the extramural pathways existed in three divisions: 1. The ‘‘zone of downward spread’’—evidence for these channels came from Miles’ experience of locally recurrent disease within the ischioanal fossae from perineal resections. 2. The ‘‘zone of lateral spread’’—evidence for these came from the reported findings of plaques of tumor found within the levator ani. Miles also observed peritoneal deposits on the pelvic sidewalls and again implicated these channels. 3. The ‘‘zone of upward spread’’—the lymphatics of this zone were the retrorectal glands, those of the pelvic mesocolon, the glands situated at the bifurcation of the left common iliac artery, and the median lumbar (aortic) glands. Miles (6) recognized these as, ‘‘the most constant and, therefore, the most important of all the routes of spread.’’

154

Daniels and Heald

In 1939, while Miles (6) was publishing his book ‘‘Rectal Surgery. A Practical Guide to the Modern Treatment of Surgical Disease,’’ Dixon (7), in America, described the sphincter-preserving anterior resection (AR). Although previously alluded to by Moynihan (8) in the United Kingdom, it was Dixon who described the first large series. Dixon performed the operation in three stages: initially, by defunctioning the rectum with a colostomy, then an AR with the colostomy remaining, and lastly, closure of the colostomy. Dixon agreed with Miles that the ‘‘zone of upward spread’’ was the most important factor associated with the development of metastases. Dixon (7) believed that AR led to reduced perioperative morbidity and improved functional outcome and quality of life for the patients by restoring bowel continuity. The reports of AR raised questions about whether local control would be achieved. Such leading surgeons as Gabriel, of St. Mark’s, intoned against ‘‘the evils of such irresponsible challenges to established practice.’’ Those who were determined to persevere produced a further ‘‘standard’’ that had little scientific basis, namely, ‘‘the 5-cm rule’’ (9–11). However, later studies by Scott et al. (12) showed that microscopic tumor spread distal to the main tumor is rare, and occurs in cases with lymph node involvement and an almost invariably unfavorable prognosis. The discussions that limited the use of the AR in the 1940s resurfaced again in the 1970s and 1980s following the introduction of circular stapling devices. The early reports were followed by suggestions that the new technology would lead to increases in anastomotic failure and LR (13). Performing surgery within the confines of the pelvis is one of the most challenging aspects of colorectal surgery. The importance of accuracy in rectal cancer surgery is paramount in aiming to provide a cure for the patient. Conventional surgical techniques, using blunt dissection of the rectum and mesorectum (the embryological hindgut package), pose a high risk of damage to the hypogastric nerves and breaching of the mesorectum. A significant morbidity in bladder and sexual dysfunction is attributed to the damage of both the superior and inferior hypogastric nerve plexuses. This can vary between 20% and 60% of patients in different centers (14). When the mesorectum is excised intact, enveloped by the visceral pelvic fascia, all components of the pelvic autonomic nervous system can be preserved. This means that there is sparing of the superior hypogastric nerves, anterior roots of S2, S3, S4, and the nervi erigentes along the pelvic sidewalls. Similarly, breaching of the mesorectum influences LR of the tumor and therefore increases morbidity and mortality for the patient—symptoms include pelvic pain, ureteric obstruction, fistulation, and poor bowel function. LR rates, after potentially curative surgery, have varied between 5% and 40% in different reports. Five-year survival in patients with LR is <5%, with a median survival of only seven months (15). In these patients, palliation is difficult to achieve and, ultimately, our treatment has almost invariably failed. Therefore, total mesorectal excision (TME) has been introduced as

Surgical Principles of Managing Colorectal Cancer

155

a precise surgical technique with low morbidity, low LR rates, and increased survival in patients with rectal cancer (16). Maunsell (1) was the first to coin the term ‘‘mesorectum,’’ although it fails to appear in most editions of Gray’s Anatomy. In some recent editions, it is specifically defined as the mesentery of the sigmoid and rectosigmoid where peritoneum is present on both aspects—‘‘a sine qua non’’ for a mesentery for some anatomists. The term ‘‘mesorectum’’ has been popularized by Heald et al. (17) within the last 20 years. Most oncologists, and all surgeons, now accept this term as describing the integral visceral lymphovascular and fatty surround that provides the unique opportunity for the near-perfect oncological specimen or ‘‘monobloc.’’ Recently, controversy has been raised as to the strict anatomical definition of the technique of TME (18). Chapius states that: The description of the rectal mesentery is not correct as the structure itself is made up of the terminal branches of the inferior mesenteric artery, the associated veins and lymphatics, and a variable pad of fat within the envelope of the fascia propria. Hill and Rafique (19) offer the alternative term for TME as ‘‘extrafascial excision of the rectum.’’ In the 1980s in Basingstoke, enthusiasm for TME led to the development of a comprehensive rectal cancer database and the introduction of fastidious surgical efforts, directed at the complete excision of the mesorectum, by the careful pursuit of the innermost areolar plane around the integral visceral mesentery of the hindgut. The most independent pathological prognostic factors, following rectal cancer excision, are the relationship between tumor involvement of the circumferential resection margin (CRM) and peritoneal perforation by tumor (20). Therefore, for the surgeon, the operation of TME comprises five basic principles: 1. Perimesorectal ‘‘holy plane’’ sharp dissection with diathermy and scissors under direct vision. 2. Specimen-orientated surgery and histopathology, the object of which is an intact mesorectum with no tearing of the surface and no circumferential margin involvement [CRM negative (R0)]— naked eye or microscopic. High-quality pathological assessment, for circumferential completeness of resection, as the principal outcome measure of the success, or failure, of the surgery. 3. Recognition and preservation of the autonomic nerve plexuses on which sexual and bladder function depend. 4. A major increase in anal sphincter preservation and a reduction in the number of permanent colostomies.

156

Daniels and Heald

5. Stapled low pelvic reconstruction, usually using the Moran triple stapling technique, plus short colon-pouch anastomosis to the low rectum or anal canal (21). The rationale for TME is based upon the following principles: 1. That the regional spread of rectal cancer is confined, in the early stages, within the visceral pelvic fascia. Local spread beyond this fascial layer is usually associated with an advanced primary tumor (22). 2. Local failure is often caused by perforation of the visceral pelvic fascia by the surgeon, which may leave malignant tissue or spilled cells within the pelvic cavity. This may be apparent by CRM involvement (CRM þve) on the pathological specimen. 3. Distal spread within the mesorectum is encompassed by a complete mesorectal excision in mid- and low-rectal tumors. 4. Sharp dissection along the ‘‘holy plane,’’ between the visceral and parietal fascia, under direct vision, allows preservation of the pelvic autonomic nerves as well as delivery of an oncologically perfect specimen (23). The ‘‘holy plane’’ is the innermost (visceral) layer of fascia that surrounds the midline hindgut within its lymphovascular envelope (the core), while the two surrounding lamellae comprise a neural layer and a Wolffian ridge layer, which develop from the paired structures outside the hindgut. The outermost plane is the one around the adventitia of the great vessels. Therefore, the ‘‘holy plane’’ is around the integral visceral mesentery of the hindgut—the lymphovascular surround of the colon and rectum (mesorectum) (24). This is a completely fatty and lymphovascular surround, on all aspects of the middle third of the rectum. In the upper third, the anterior aspect is only covered by peritoneum and the posterior by mesorectum, enveloping the sides as the peritoneal reflection tapers forward. In the lower third of the rectum, there is virtually no fatty tissue between the rectum and front of the prostate. Denonvillier’s fascia is a septum between the anterior mesorectal fat of the mid-rectum and the loose areolar tissue behind the seminal vesicles (25). More distally, in front of the rectal muscle tube, the fascia is rather firmly adherent to the capsule of the prostate and there is little or no prerectal fat. Posteriorly and posterolaterally, the ‘‘holy plane’’ and fascia are well defined around the globular expanding bilobed mesorectum, eventually tapering to merge with the internal anal sphincter.

RECURRENCE AND SURVIVAL Rectal cancer surgery has often been associated with high LR rates, major morbidity, and a profound negative impact on survival. Rates may vary among surgeons and hospitals (26,27). Early papers from Basingstoke

Surgical Principles of Managing Colorectal Cancer

157

suggested such large improvements that the data were quite simply disbelieved by many or attributed to case selection. The concept of TME surgery for rectal cancer was introduced in 1982, with real promise that it would reduce LR, improve survival rates, and reduce surgeon variability. More recently, in the Norwegian Rectal Cancer Project, Wibe et al. reported an observational cohort. They looked at 3319 patients during the period between 1993 and 1997, who had undergone rectal cancer surgery, influenced by the Basingstoke/Norwegian Training Programme to introduce TME (28). Of these patients, 76% had a major resection, but only 54% of the total number had a resection that was likely to be ‘‘curative,’’ i.e., a specimen with a distance of >1 mm between the CRM and the tumor. The results showed that in 1994, 78% of patients had a TME and, by 1997, this figure had risen to 92%. The LR rate at 30 months was 6% in the TME group, compared with 12% in the conventional surgery group. The four-year survival was also greater in the TME group, 73% versus 60% in the conventional surgery group. The risk factors for LR identified as significant were male gender tumor at <6 cm from the anal verge, tumor stage, i.e., Dukes’ C, and operative technique (TME vs. conventional surgery). Critics of this type of paper would class this information as nonrandomized, circumstantial evidence. Ideally, the clinical and cost-effectiveness of the presumed better surgical performance of TME should be tested within a randomized controlled trial. However, once introduced to the technique, most surgeons regard TME, with its standardization of resection of the mesorectum as a package, as a better technique. Thus, a randomized controlled trial, with a suboptimal non-TME arm, would be unethical. What was needed was a measure of surgical proficiency, but the surgeon could not be controlled within the confines of a randomized clinical trial. The pathologist Quirke developed a grading system for surgery, based upon gross macroscopic examination of the mesorectum. This has been successfully used within the conventional surgery compared with laparoscopic-assisted surgery in treating patients with colorectal cancer (CLASICC) study, the Dutch TME and short-course radiotherapy (5 Gy
5 days) (SRT) trial, and the Magnetic Resonance Imaging and Rectal Cancer European Equivalance Study (MERCURY) (29). It seems that auditing the performance of individual institutions, using a population-based system with consecutive registration of all cases, is the way forward. Martling et al. (16) reported the surgeon as a prognostic factor after the introduction of TME for rectal cancer in Sweden. Patient outcome was significantly improved in units where the surgeons had colorectal subspecialty training and high case volume. This was reiterated by the Stockholm trials and by Read et al. (30,31), where LR and death rates were lower if the patient had been operated on by surgeons who had been specialists for more than 10 years. In the latter paper, pathology stage and the background of the surgeon were the only independent predictors of disease-free survival.

158

Daniels and Heald

In a historical comparison between the Dutch TME series and their previous series, the CRAB Study, the former showed a clear improvement in LR rates and survival as a result of ‘‘TME training.’’ Other evidence for the reduced LR rates following TME surgery can be found in the Danish TME Study. In this study, the reported three-year LR rate was 11% in patients who had undergone TME surgery, compared with 30% in patients who had conventional surgery. Three-year survival was also higher in the TME group (77% vs. 62%). Only advanced age and tumor in the lower third of the rectum were independent predictors of LR after TME. Heald and Ryall (32) have shown an LR rate as low as 5% and fiveyear survival of 75% for patients with all stages of rectal cancer. It would seem that with an optimal TME, the majority of rectal cancer cases have good outcome with surgery alone, but neoadjuvant therapy may prove useful in those cancers where the CRM is at risk of tumor infiltration. Tumors could be staged preoperatively using magnetic resonance imaging (MRI), as it correlates well with the histopathology of the resected specimen (33). Results of a large European prospective study of MRI staging are awaited—The MERCURY Study—Magnetic Resonance Imaging and Rectal Cancer European Equivalence Study—assessed the equivalence between high-resolution, fine-slice MRI and the corresponding histopathological whole mount section. In the original series described by Heald, there was a high leak rate (17.4% clinical and radiological) associated with TME. Reasons for this include a lower anastomosis at the pelvic floor and, possibly, the early practice of Heald to retain a skeletonized rectal tube after TME for function. Now, a colonic J pouch is used to improve function; a randomized controlled trial of colonic J pouch–anal anastomoses versus straight coloanal anastomoses gave a lower leak rate (2% vs. 15%) after pouch anastomoses (34). Studies that have shown leak rates of <5% after TME are quoted by Enker et al. (35) and Zaheer et al. (36). The obvious reason cited for higher leak rates in some TME series is that a greater proportion of patients are undergoing restorative surgery and anastomotic heights are therefore reducing (37). The Goldberg paper states that TME per se implies a greater need for a stoma, to provide proximal diversion for the anastomosis. However, Bissett reports a stoma rate of 40% for TME and 37% for conventional surgery. Bissett adds that although the treatment costs and hospital stay were the same for the two groups, LR was higher in the conventional group, therefore adding to overall management costs, mortality, and morbidity. APPLYING THE PRINCIPLES OF TOTAL MESORECTAL EXCISION TO COLONIC CANCER Recently, the focus has been on improving the outcome in rectal cancer. However, it must be remembered that two-thirds of colorectal cancer is

Surgical Principles of Managing Colorectal Cancer

159

colonic. In the early 20th century, colonic surgery was rare. The disease presented late and this led to poor survival. However, with the increasing use of imaging and endoscopy, early lesions are now being diagnosed and treated. While we have seen a trend of improved survival in rectal cancer, the trend for improvement in colonic cancer has not been as impressive. Indeed, in Sweden, survival following rectal cancer is now better than that following colonic cancer. This leads to the question: ‘‘How might the lessons of TME be applied to colonic cancer?’’ What we have learned with rectal cancer is the importance of detail of surgical excision along embryological planes. In rectal cancer, the importance of preoperative staging has led to a selective policy for preoperative therapy and staged surgical resection. The outcome has been improved with the specialization of surgeons and management with a multidisciplinary team approach. These areas need expanding within future clinical trials, directed towards improving outcome in colonic cancer. In this chapter, we have not addressed the issues of laparoscopic surgery and its role in the management of colorectal cancer. These are issues that will require further clinical trials and collection of accurate data, comparing the quality of resection, function, morbidity and mortality, and overall survival.

CONCLUSION In conclusion, the optimal operation for rectal cancer must take into account the spread of the tumor, the disabilities that would be inflicted by each component of the surgery, and consider the point at which cure is impossible by local surgery alone. Since the introduction of TME by Heald in 1982, the technique has been shown to reduce LR of rectal cancer, improve patient survival, and reduce the need for adjuvant therapy. The concept of TME has now become universally recognized as the standard surgical procedure for rectal cancer. However, LR still occurs and overall survival can probably be improved upon further. Quality of surgical resection is paramount and an accepted specimen grading may be used to compare surgeons performance. Accurate preoperative staging and targeted preoperative therapy, assessed by clinical trials, offer future hope to improving survival of patients with rectal cancer. However, the grim predictions of Maunsell have been beaten, not by the magic bullet of oncology but by careful attention to surgical details.

REFERENCES 1. Maunsell H. A new method of excising the two upper portions of the rectum and the lower segment of the sigmoid flexure of the colon. Lancet 1892; 2: 473–476.

160

Daniels and Heald

2. Miles WE. A method of performing abdomino-perineal excision for carcinoma of the rectum and terminal portion of the pelvic colon. Lancet 1908; 35: 320–321. 3. Murray JJ, Veidenheimer MC. Abdomino-perineal excision of the rectum. In: Fielding LP, Goldberg SM, eds. Rob and Smith’s operative surgery. London: Chapman and Hall, 1993:472. 4. Hartmann H. Ablation abdominale des cancers rectosigmoidienne. In: 30th Congres francais de chirurgie, Strasbourg, 1923. 5. Hartmann H. Procede nouveau d’extirpation des cancers de la partie terminale du colon. Bull et mem de la Societe de Chirurgie Paris 1923:1474. 6. Miles WE. Rectal surgery. In: A practical guide to the modern surgical treatment of rectal diseases. 1st ed. London: Cassell and Company Ltd., 1939: 198–286. 7. Dixon CF. Surgical removal of lesions occurring in the sigmoid and rectosigmoid. Am J Surg 1939; 46:12–17. 8. Moynihan BGA. The surgical treatment of cancer of the sigmoid flexure and rectum. Surg Gynaecol Obstet 1908; 6:463–466. 9. Cole PP. The intramural spread of rectal carcinoma. Br Med J 1913; 1: 431–433. 10. Connell JF Jr., Rottmo A. Retrograde spread of carcinoma in the rectum and rectosigmoid. Arch Surg 1949; 59:807–813. 11. Handley WS. The surgery of the lymphatic system. Br Med J 1910; 1:922–928. 12. Scott N, Jackson P, al-Jaberi T, et al. Total mesorectal excision and local recurrence: a study of tumour spread in the mesorectum distal to rectal cancer. Br J Surg 1995; 82:1031–1033. 13. Hurst PA, Prout WG, Kelly JM, et al. Local recurrence after low anterior resection using the staple gun. Brit J Surg 1982; 69(5):275–276. 14. Maurer C, Z’Graggen K, Renzuilli P, et al. Total mesorectal excision preserves male genital function compared with conventional rectal cancer surgery. Br J Surg 2001; 88:1501–1505. 15. Gunderson LI, Sosin H. Areas of failure at re-operation (second or symptomatic look) following ‘curative surgery’ for adenocarcinoma of the rectum. Cancer 1974; 34:1278–1292. 16. Martling A, Cedermark B, Johansson H, et al. The surgeon as a prognostic factor after the introduction of total mesorectal excision in the treatment of rectal cancer. Br J Surg 2002; 89(8):1008–1013. 17. Heald RJ, Moran BJ, Ryall R, et al. The Basingstoke experience of total mesorectal excision 1978–1997. Arch Surg 1998; 133:894–899. 18. Chapius P, Bokey L, Fahrer M, et al. Mobilization of the rectum. Anatomic concepts and the bookshelf revisited. Dis Colon Rectum 2002; 45(1):1–9. 19. Hill GL, Rafique M. Extrafascial excision of the rectum for rectal cancer. Br J Surg 1998; 85:809–812. 20. Hermanek P, Guggenmoos-Holzmann IP, Buttner PP. Prognostic factors in rectal cancer. Langenbecks Archiv fur Chirurgie—supplement II. Verhandlungen der par Deutschen Gesellschaft fur Chirurgie 1989:663–667. 21. Moran BJ. Stapling instruments for intestinal anastomosis in colorectal surgery. Br J Surg 1996; 83(7):902–909.

Surgical Principles of Managing Colorectal Cancer

161

22. 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(3):327–334. 23. Heald RJ. The ‘holy plain’ of rectal cancer. J Roy Soc Med 1988; 81:503–508. 24. Stelzner F. Technik und Ergebnisse der knappen Kontinenzresektion beim Rektumkarzinom. Zentralblatt fur Chirurgie 1992; 117(2):63–66. 25. Heald RJ, Moran BJ, Brown G, Daniels IR. Optimal total mesorectal excision for rectal cancer is by dissection in front of Denonvilliers’ fascia. Br J Surg 2004; 91(1):121–123. 26. Garcia-Granero E, Marti-Obiol R, Gomez-Barbadillo J, et al. Impact of surgeon organization and specialization in rectal cancer outcome. Colorect Dis 2001; 3(3):179–184. 27. Luna-Perez P, Reyna Huelga A, Labastida Almendaro S, et al. The surgeon as prognostic factor for local recurrence and survival in the anal sphincter preservation for mid-rectal cancer. Rev Invest Clin 1999; 51(4):205–213. 28. Wibe A, Eriksen MT, Syse A, et al. Total mesorectal excision for rectal cancer— what can be achieved by a national audit? Colorect Dis 2003; 5(5):471–477. 29. 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(9):638–646. 30. Martling AL, Holm T, Rutqvist LE, et al. Effect of a surgical training programme on outcome of rectal cancer in the County of Stockholm. Lancet 2000; 356:93–96. 31. Read TE, Myerson RJ, Fleshman JW, et al. Surgeon specialty is associated with outcome in rectal cancer treatment. Dis Colon Rectum 2002; 45(7):904–914. 32. Heald RJ, Ryall RD. Recurrence and survival after total mesorectal excision for rectal cancer. Lancet 1986; 1(8496):1479–1482. 33. Brown G, Richards CJ, Newcombe RG, et al. Rectal carcinoma: thin-section MR imaging for staging in 28 patients. Radiology 1999; 211(1):215–222. 34. Hallbook O, Hass U, Wanstrom A, et al. Quality of life measurement after rectal excision for cancer. Comparison between straight and colonic J-pouch anastomosis. Scand J Gastroenterol 1997; 32(5):490–493. 35. Enker W, Thaler HT, Cranor ML, et al. Total mesorectal excision in the operative treatment of carcinoma of the rectum. J Am Coll Surg 1995; 181(4):335–346. 36. Zabeer S, Pemberton JH, Farouk R, et al. Surgical treatment of adenocarcinoma of the rectum. Ann Surg 1998; 227(6):800–811. 37. Goldberg S, Klas JV. Total mesorectal excision in the treatment of rectal cancer: a view from the USA. Semin Surg Oncol 1998; 15(2):87–90.

7 Adjuvant Therapy for Colorectal Cancer Geoff Chong and David Cunningham Department of Medicine, Royal Marsden Hospital, Sutton, Surrey, U.K.

INTRODUCTION Despite progress in screening and treatment strategies, colorectal cancer remains a significant public health problem in the developed world. While a proportion of patients are cured by surgical resection alone, a significant percentage experience tumor relapse and ultimately die. Adjuvant chemotherapy has been shown in a series of randomized trials to prevent relapse in some patients, hence improving survival in defined risk groups. While to date, standard adjuvant therapy has been essentially restricted to 5-fluorouracil (5-FU)-based chemotherapy, this paradigm is rapidly changing. With the demonstration of survival advantages with oxaliplatin and irinotecan-based combinations in patients with advanced colorectal cancer, these agents are now being studied in the adjuvant setting. Capecitabine has been shown to be an equally efficacious and better-tolerated alternative to leucovorin-modulated 5-FU (5-FU/LV) in patients with metastatic disease, and can be substituted for 5-FU in oxaliplatin and irinotecan combinations. Also promising are the novel strategies of inhibiting epidermal growth factor receptor (EGFR) and vascular endothelial growth factor (VEGF) with monoclonal antibodies (mAbs). These molecules have shown additive benefits when combined with chemotherapy in patients with advanced colorectal cancer, naturally leading to the hypothesis that a similar approach might similarly improve adjuvant therapy.

163

164

Chong and Cunningham

This chapter outlines the evidence for adjuvant therapy for patients with colorectal cancer, the patient groups that may benefit from such therapy, and current and future research directions in this area. ADJUVANT THERAPY FOR STAGE III COLORECTAL CANCER According to the American Joint Committee on Cancer stage groupings, stage III colon cancer is defined as any T, N1/2, M0 under the tumor node metastasis (TNM) classification (1). Stage III tumors can be further subdivided into IIIA (T1/2, N1, M0), IIIB (T3/4, N1, M0), IIIC (any T, N2, M0). For untreated stage III colon cancer, the risk of relapse is 60% to 70%, and the five-year survival is approximately 50% (2). A sequence of large randomized trials was conducted by the U.S. National Surgical Adjuvant Breast and Bowel Project (NSABP) in the 1980s and 1990s, investigating the role of postoperative adjuvant therapy in reducing the risk of relapse. Concurrently, several North American and European cooperative group studies were also performed, further defining the role of adjuvant chemotherapy in patients with stage III colon cancer. For brevity in the following text, drug schedules of selected trials are specified in Table 1. 5-FU is a fluoropyrimidine that acts predominantly by inhibiting thymidylate synthase (TS) via one of its active metabolites, 5-fluorodeoxyuridine monophosphate (17). TS is critical for the production of thymidylate, which in turn is converted to thymidine triphosphate. This substrate is essential for DNA synthesis and repair. In addition, a second 5-FU metabolite, fluorouracil triphosphate, becomes incorporated into RNA, interfering with its normal functions. A number of preclinical studies have shown 5-FU to be active against human colorectal xenografts (18,19). Early clinical studies in patients with metastatic colorectal cancer confirmed modest antitumor activity (20,21). On this basis, 5-FU has formed the backbone of chemotherapy treatment for colorectal cancer for the past four decades. The C-01 trial was the first NSABP colon trial that was designed to evaluate outcomes for patients with resected Dukes’ B and C tumors (3). The randomization arms were: no additional treatment; chemotherapy with 5-FU, semustine and vincristine; and bacille Calmette-Gue´rin (BCG) vaccination. One thousand one hundred and sixty-six patients were randomized to one of the three arms, and the mean follow-up duration was 77.3 months. There was a statistically significant survival advantage seen in patients treated with chemotherapy, with a median five-year survival of 67% compared to 59% for no treatment ( p ¼ 0.05). Interestingly, the BCG arm experienced similar overall survival to the chemotherapy arm, although this appeared to be due to a reduction in non-cancer-related deaths. When this was corrected, the five-year survival was similar to the no-treatment group. This was the first randomized clinical trial to demonstrate disease-free and overall survival benefit using adjuvant chemotherapy in patients with resected Dukes’ B and C colon cancer. (Text continues on page 169.)

Stage Treatment arms

Wolmark II and III (A) Observation et al. (3) (B) 5-FU/semustine/ vincristine every 10 wk (C) BCG vaccination weekly II and III (A) Observation Moertel (B) 5-FU 450 mg/m2/LEV et al. days 1–5 then weekly (4,5) from day 29 (C) Levamisole (only for Dukes’ C) Wolmark (A) 5-FU/semustine/ et al. (6) vincristine every 10 wk (B) 5-FU 500 mg/m2 and LV 500 mg/m2 weekly for 6 of 8 wk O’Connell II and III (A) Observation et al. (7) (B) 5-FU 450 mg/m2 and LV 20 mg/m2 days 1–5 every 4–5 wk O’Connell II and III (A) 5-FU 450 mg/m2 et al. (8) and LEV (B) 5-FU 450 mg/m2 and LEV

Author(s)

12 mo

68%

60% (5 yr)

6 mo

891

74%

6 mo

63% (5 yr)

– 309

76%

48 wk

66% (5 yr)

50 wk 1045

71%

0.05 (4 arms)

0.01

0.003

55% (3 yr) NR 0.007 (A vs. C)

929

P

59% (5 yr) 0.05 (A vs. B), 67% 0.03 (B vs. C) 67%

3- or 5-year survival

1116

Eligible patients

12 mo

– 12 mo

40 wk

– 80 wk

Treatment duration

Table 1 Selected Trials Illustrating Progress in the Adjuvant Treatment of Colorectal Cancer

(Continued )

6 mo chemotherapy ¼ 12 mo chemotherapy

5-FU/LV superior to surgery alone

5-FU/LV superior to MOF

5-FU/LEV superior to surgery alone

Adjuvant chemotherapy (MOF) superior to surgery alone

Comments

Adjuvant Therapy for Colorectal Cancer 165

Stage Treatment arms

(B) 5-FU 450 mg/m2 and LEV (C) 5-FU 500 mg/m2 and LV 500 mg/m2 and LEV weekly 6 out of 8 wk

73%

48 wk

74% (5 yr)

48 wk

70%

67%

6 mo

48 wk

65%

6 mo

2078

66%

63% (5 yr)

12 mo

6 mo

63%

12 mo

3- or 5-year survival 70%

3759

Eligible patients

6 mo

Treatment duration

0.07 (A vs. B), 0.11 (B vs. C), 0.99 (A vs. C)

0.09 (A vs. B), 0.56 (B vs. C), 0.007 (A vs.C), 0.24 (B vs. D)

P

Selected Trials Illustrating Progress in the Adjuvant Treatment of Colorectal Cancer (Continued )

(C) 5-FU 370 mg/m2/LV 20 mg/m2 and LEV (D) 5-FU 370 mg/m2/LV 20 mg/m2 and LEV Haller II and III (A) 5-FU 450 mg/m2 and et al. (9) LEV days 1–5 then weekly from day 29 (B) 5-FU 425 mg/m2 and LV 20 mg/m2 days 1–5 every 4–5 wk (C) 5-FU 500 mg/m2 and LV weekly 6 out of 8 wk (D) 5-FU 425 mg/m2 and LV and LEV days 1–5 every 4–5 wk Wolmark II and III (A) 5-FU 500 mg/m2 and LV 500 mg/m2 weekly et al. (10) 6 out of 8 wk

Author(s)

Table 1

5-FU/LV equivalent survival to 5-FU/LEV; slightly better DFS

Addition of LEV to 5FU/LV not necessary. Weekly 5-FU/LV ¼ monthly 5-FU/LV

Comments

166 Chong and Cunningham

6 mo

6 mo

6 mo (C) 5-FU 370 mg/m2/LEV and LV 175 mg weekly or 4 weekly (D) 5-FU 370 mg/m2/LEV 6 mo and LV 25 mg weekly or 4 weekly Saini II and III (A) Continuous infusion 53 mo et al. (12) FU 300 mg/m2/day (B) 5-FU 425 mg/m2 and 6 mo LV 20 mg/m2 days 1–5 every 4 wk Andre II and III (A) LV5-FU2 every 2 wk 24/36 wk et al. (13) (B) 5-FU/LV monthly 24/36 wk Andre II and III (A) FOLFOX4 every 2 wk 6 mo et al. (14) (B) LV5-FU2 every 2 wk 6 mo Cassidy 6 mo III (A) Capecitabine 2500 mg/ et al. (15) m2/day days 1–14 every 3 wk

QUASAR II and III (A) 5-FU 370 mg/m2 Collaboand LV 175 mg weekly rative or 4 wk (B) 5-FU 370 mg/m2 and Group LV 25 mg weekly or 4 (11) weekly

88% 87.7% (3 yr) 86.6% 81.3% (3 yr)

1987

86% (3 yr)

83.2%

87.9% (3 yr)

0.07

NS

0.18

0.76

70.1% versus 71.0% 0.43 (A vs. B), 0.06 (C vs. D) (HDLV vs. LDLV), 69.4% versus 71.5% (LEV vs. no)

2246

905

692

4927

(Continued )

Capecitabine at least equivalent to bolus 5FU/LV

FOLFOX4 superior to LV5-FU2 in 3 yr DFS

LV5-FU2 ¼ monthly bolus 5-FU/LV

3 mo PVI 5-FU ¼ 6 mo bolus 5-FU/LV

Low dose LV ¼ high dose LV

Adjuvant Therapy for Colorectal Cancer 167

Stage Treatment arms

78.7%

0.88

P

UFT/LV ¼ bolus 5-FU/ LV

Comments

Abbreviations: 5-FU, 5-fluorouracil; LV, leucovorin; HDLV, high-dose leucovorin; LDLV, low-dose leucovorin; LEV, levamisolegiven 50 mg eight-hourly for three days every two weeks; QUASAR, quick and simple and reliable investigators; PVI, prolonged venous infusion; LV5-FU2, leucovorin 200 mg/m2 over two hours, 5-FU 400 mg/m2 bolus, 5-FU 600 mg/m2 over 22 hours on days 1 and 2; FOLFOX4, oxaliplatin 85 mg/m2 on day 1 in combination with LV5-FU2; NR, not reported; NS, not significant; DFS, disease free survival.

24 wk

78.7% (5 yr)

25 wk

3- or 5-year survival 77.6%

1608

Eligible patients

6 mo

Treatment duration

Selected Trials Illustrating Progress in the Adjuvant Treatment of Colorectal Cancer (Continued )

(B) 5-FU 425 mg/m2 and LV 20 mg/m2 days 1–5 every 4 wk Wolmark II and III (A) Tegafur 300 mg/m2/ day þ uracil (1:4 ratio) et al. (16) and LV 90 mg/day days 1–28 every 35 days (B) 5-FU 500 mg/m2 and LV 500 mg/m2 weekly for 6 of 8 wk

Author(s)

Table 1

168 Chong and Cunningham

Adjuvant Therapy for Colorectal Cancer

169

The next trial in the NSABP series was C-02, which studied the role of postoperative portal vein 5-FU infusion in patients with resected Dukes’ A, B, and C colon cancer (22). The main aims were to determine whether the subsequent incidence of liver metastases could be reduced using this strategy; and whether this could lead to improved survival. One thousand one hundred and fifty-eight patients were randomized to no treatment or to a seven-day hepatic vein infusion of 5-FU, commencing on the day of surgery. Significantly, there was no decrease in the incidence of liver metastases in patients receiving hepatic venous 5-FU. Despite this, there was improved disease-free survival (74% vs. 64% at four years; p ¼ 0.02), and a trend towards improved overall survival (81% vs. 73% at four years; p ¼ 0.07). The authors concluded that the survival differences might have been due to systemic effects of the hepatic vein–administered 5-FU. Leucovorin (LV), or folinic acid, is a reduced form of folic acid, which can enhance the activity of 5-FU by stabilizing a complex of the active metabolite (5-fluorodeoxyuridine monophosphate), TS, and the folate cofactor 5,10-methylenetetrahydrofolate (23). LV also increases the intracellular pool of 5,10-methylenetetrahydrofolate, and therefore enhances the TS inhibition mediated by 5-FU (17). A multiple-arm study found that 5-FU/LV produced superior survival compared to 5-FU alone in patients with advanced colorectal cancer (24,25). In order to test this strategy in the adjuvant setting, 5-FU/LV was evaluated in the NSABP C-03 trial (6). One thousand and forty-one eligible patients with resected Dukes’ B and C colon cancer were randomized to adjuvant treatment with 5-FU/LV or lomustine (MeCCNU), vincristine, and 5-FU (MOF). Treatment was for 12 months in each arm. The mean follow-up time was 47.6 months. There was an overall survival advantage in the 5-FU/LV arm with a three-year survival of 84% compared to 77% for those treated with MOF ( p ¼ 0.003). This study therefore provided the first prospective randomized evidence of improved survival after resection of Dukes’ B and C colon cancer using 5-FU modulated with LV. The NSABP C-04 study sought to build on the results of the previous study by comparing the new standard of 5-FU/LV with two levamisolecontaining combinations (10). The rationale behind the use of levamisole, an antihelmintic and immunomodulatory agent, was promising antitumor activity of 5-FU/levamisole demonstrated in xenograft models (26). Two thousand and seventy-eight patients were randomized to one of three arms: 5-FU/LV, 5-FU and levamisole, or 5-FU/LV and levamisole. Mean follow-up time was 86 months. While there was an advantage in five-year, disease-free survival in patients treated with 5-FU/LV compared to 5-FU/levamisole (65% vs. 60%; p ¼ 0.04); this did not translate into prolonged overall survival (74% vs. 70%; p ¼ 0.07). There was no difference in DFS or overall survival between the 5-FU/LV and 5-FU/LV/levamisole arms. Therefore, the addition of levamisole to 5-FU/LV did not appear to confer additional activity in the postoperative setting. On the basis of the slightly better

170

Chong and Cunningham

disease-free survival (DFS), 5-FU/LV was assessed to be the superior regimen to 5-FU/levamisole. On the basis of preclinical and early clinical data suggesting enhanced efficacy of 5-FU with interferon a-2a (IFN), the combination of 5-FU/LV/ IFN was compared to 5-FU/LV in the NSABP C-05 study (27). Two thousand one hundred and twenty-nine eligible patients with resected Dukes’ B or C colon cancer were randomized to receive 5-FU/LV with or without IFN. Mean follow-up was 54 months. The addition of IFN did not result in improved DFS (HR ¼ 0.93; 95% CI: 0.80–1.08; p ¼ 0.34), or overall survival (HR ¼ 0.92; 95% CI: 0.76–1.11; p ¼ 0.41). There was, however, increased toxicity demonstrated in the IFN-containing arm with fewer patients able to complete the protocol-specified duration of treatment compared to the 5-FU/LV arm. Interferon is therefore not used in the adjuvant treatment of colon cancer. While the NSABP studies were critical in demonstrating the efficacy of 5-FU/LV for resected Dukes’ B and C colon cancer, a parallel series of North American and European studies contributed important additional information on the optimal 5-FU-based adjuvant regimen. These trials attempted to answer questions relating to the duration and scheduling of therapy. INT-0089 was a large, four-arm randomized study, which addressed questions of the value of adding levamisole to 5-FU/LV as well as whether a weekly 5-FU/LV schedule was equivalent to a monthly schedule (9). Three thousand seven hundred and fifty-nine patients were randomized, and though this trial has only been published in abstract form, important conclusions can be drawn. The treatment arms were 5-FU/levamisole for 12 months, 5-FU/low-dose LV for seven to eight months, 5-FU/high-dose LV for seven to eight months, or 5-FU/low-dose LV/levamisole for seven to eight months. After a median follow-up of five years, the only statistically significant survival difference for stage III disease was between 5-FU/ levamisole and 5-FU/low-dose LV/levamisole (60% vs. 65%; p ¼ 0.0054). However, 5-FU/low-dose LV/levamisole was not superior to 5-FU/lowdose LV. This trial therefore suggested that levamisole was not an additive component of 5-FU-based adjuvant therapy. An intergroup study comprising the North Central Cancer Treatment Group (NCCTG), the Southwest Oncology Group (SWOG), and the Eastern Cooperative Oncology Group (ECOG) randomized 1247 eligible patients with resected Dukes’ B2 or C colon cancer to either observation or postoperative treatment with 5-FU/levamisole (4). Treatment was continued for 48 weeks in both arms. In addition, patients with stage C tumors could be randomized to levamisole alone. Results after a median follow-up of three years were published for the whole cohort in 1990 and showed a reduction in risk of cancer recurrence as well as improved overall survival in patients with Dukes’ C disease. The reduction in death rate in patients treated with 5-FU/levamisole was 33% (95% CI: 10–50%; p ¼ 0.006). In

Adjuvant Therapy for Colorectal Cancer

171

1995, mature results from the subset of patients with stage III tumors were reported (2). There were 929 eligible patients in this group with five-year follow-up (median 6.5 years). The earlier results were confirmed, with 5-FU/levamisole being associated with a 33% reduction in mortality rate (95% CI: 16–47%; p ¼ 0.0007) compared to patients being observed. There was no survival benefit detected for therapy with levamisole alone (6% reduction in death rate; p ¼ 0.57). Further evidence for the combination of 5-FU/LV as adjuvant therapy for resected Dukes’ B and C colon cancer was obtained by a pooled analysis of three randomized trials by the International Multicentre Pooled Analysis of Colon Cancer Trials (IMPACT) investigators (28). These independently conducted trials were performed by the Gruppo Interdisciplinare Valutazione Interventi Oncologia, the National Cancer Institute Canada Clinical Trials Group, and the Fondation Francaise de Cancerologie Digestive. The experimental arms were comparable across the three trials. A total of 1493 eligible patients were randomized to 5-FU/LV or observation; approximately half were Dukes’ C. For all patients, treatment with 5-FU/LV was associated with improved survival (HR 0.77; 95% CI: 0.62–0.96; p ¼ 0.018) as well as DFS (HR 0.65; 95% CI: 0.54–0.78; p < 0.0001). Treatment was well tolerated with fewer than 3% of patients experiencing grade 4 gastrointestinal toxicity. The Quick and Simple and Reliable (QUASAR) Collaborative Group conducted a large randomized trial to answer the question of whether the standard regimen of 5-FU/LV could be improved by the addition of levamisole or higher-dose LV (11). Patients with resected colorectal cancer were randomly assigned in a 2
2 factorial manner to 5-FU with either highor low-dose LV, and either levamisole or placebo. Patients were allowed to receive 5-FU/LV as five consecutive daily doses every four weeks for six cycles, or 30 doses given weekly. Four thousand nine hundred and twentyseven patients were randomized and the median follow-up for survivors was three years. Seventy-two percent of patients were of Dukes’ C stage and 68% had colonic tumors. No significant difference in survival was detected between high- and low-dose LV (70.1% vs. 71.0% at three years; p ¼ 0.43); although there was a trend to worse survival with levamisole compared to placebo (69.4% vs. 71.5% at three years; p ¼ 0.06). No statistically significant differences in recurrence rates were observed. The authors concluded that the results of this large randomized study represented strong evidence of the lack of benefit from higher-dose LV or addition of levamisole over standard 5-FU/LV in the adjuvant setting. This trial continued to accrue patients with no definite indication or contraindication for adjuvant therapy as the QUASAR-1 study. This will be described further in the section dealing with adjuvant therapy for stage II tumors. The German Arbeitsgemeinschaft Gastrointestinale Onkologie performed the adjCCA-01 randomized trial, which compared 5-FU/LV and

172

Chong and Cunningham

5-FU/levamisole in resected Dukes’ C patients. Six hundred and eighty eligible patients were randomized to treatment with one of these regimens for 12 months. In the initial study report with a median follow-up time of 46.5 months, survival was improved in the 5-FU/LV arm compared to the 5-FU/levamisole arm (29). Superior overall survival in the 5-FU/LV group was confirmed after a median follow-up time of 82 months: 88.9 months (95% CI: 85–93) versus 78.6 months (95% CI: 74–83; p ¼ 0.0035) (30). The number of toxicities experienced in both arms was similar, although grade 3/4 diarrhea, nausea, and vomiting occurred more frequently in the 5-FU/ LV arm ( p-values not given). The Dutch NACCP also demonstrated efficacy of 5-FU/levamisole, compared to no adjuvant treatment in a randomized trial of patients with colon or rectal cancer of stage II or III (31). One thousand and twenty-nine patients were randomized; patients in the treatment arm received 5-FU and levamisole. The combined treatment was continued for a total of 12 months. Of note, 49 patients randomized to chemotherapy never commenced treatment. Median follow-up time was 4.75 months, and the estimated five-year survival was significantly greater in the adjuvant treatment arm: 68% versus 58%; p ¼ 0.007. A survival difference was not detected on subanalysis of the rectal cancer group, although the number of patients in this group was small (n ¼ 229). A subsequent NCCTG trial established the efficacy of intensive six-month adjuvant therapy with 5-FU and low-dose LV compared with observation (7). Three hundred and seventeen patients with high-risk stage II or stage III colon cancer were randomized to six months of 5-FU 425 mg/m2 daily plus LV 20 mg/m2 daily for five days, every four to five weeks; or to observation alone. The proportion of patients alive at five years was 0.74 for the chemotherapy group compared to 0.63 for the observation group (p ¼ 0.02). The equivalence of six months’ treatment to 12 months of adjuvant therapy was demonstrated in another NCCTG trial (8). Eight hundred and ninety-one eligible patients were randomized to receive either six or 12 months of adjuvant therapy either with 5-FU/LV plus levamisole or 5-FU plus levamisole (total of four treatment groups). The 5-FU/levamisole schedule was the same as in the previous intergroup adjuvant trial (5). The incidence of grade 3/4 toxicities was similar across the four treatment arms, except that the frequency of diarrhea and stomatitis was higher in the 5-FU/LV/LEVtreated patients (p < 0.05). After a median follow-up time of 5.1 years, no improvement in survival was seen in patients treated for 12 months compared to 6 months of therapy. However, in patients treated for six months, there was inferior five-year survival in patients treated with 5-FU and levamisole, compared to 5-FU/LV/levamisole (60% vs. 70%; p < 0.01). Even though the studied regimens used levamisole, this trial was considered ‘‘proof of principle’’ of the equivalence of 6 and 12 months’ treatment.

Adjuvant Therapy for Colorectal Cancer

173

Another approach to 5-FU-based adjuvant therapy has been to utilize continuously infused 5-FU instead of bolus schedules. A potential advantage of continuous infusion is that greater dose intensity can be achieved, compared to bolus schedules. Given previous data suggesting a greater survival advantage in patients treated with higher planned doses of 5-FU, this could potentially be a strategy to further shorten the required duration of adjuvant therapy (32). Indeed, this question has been addressed in a randomized study of infused 5-FU given for three months versus bolus 5-FU/LV given for six months (12). Eight hundred and one eligible patients with Dukes’ B or C colorectal cancer were randomized to 5-FU 300 mg/m2 daily for 12 weeks; or to 5-FU 425 mg/m2 and LV 20 mg/m2 bolus days 1–5 every 28 days for six months. Updated results with a median follow-up of 5.3 years showed no difference in five-year, relapse-free survival (HR 0.8; 95% CI: 0.62–1.04; p ¼ 0.10) (33). There was a trend to improved five-year survival in the infused 5-FU arm: 75.7% versus 71.5% (HR 0.79; 95% CI: 0.61–1.03; p ¼ 0.083). The toxicity profile of infused 5-FU was more favorable compared to bolus 5-FU/LV, with significantly less grade 3/4 neutropenia, diarrhea, nausea/ vomiting, and stomatitis (p < 0.0001). The incidence of grade 3/4 palmarplantar syndrome was nonsignificantly increased in the infused 5-FU arm. Therefore, this study demonstrated the tolerability and activity of shortduration, infused 5-FU adjuvant treatment for patients with resected Dukes’ B and C colorectal cancer. Another schedule that has been widely studied in the advanced disease setting is bolus/infusion 5-FU/LV, otherwise known as the de Gramont schedule or LV5-FU2. This schedule consists of LV 200 mg/m2 given over two hours, followed by bolus 5-FU 400 mg/m2, followed by infused 5-FU 600 mg/m2 over 22 hours on days 1, 2 and repeated every two weeks. In a 2
2 factorial design, 905 patients with Dukes’ B2 and C colon cancer were randomized to LV5-FU2 given every two weeks or bolus 5-FU/LV (5-FU 400 mg/m2 and LV 200 mg/m2 days 1–5, repeated every 28 days). In addition, patients were randomized to a total of 24 or 36 weeks of therapy (13). After a median follow-up of 41 months, no statistically significant differences were observed between the studied treatment schedules or durations. Rates of grade 3/4 diarrhea, neutropenia, and mucositis were significantly lower in the LV5-FU2 arm (p < 0.001) for all toxicities. This trial thus represented important confirmation of the efficacy of different dosing schedules of 5-FU, as well as providing additional evidence supporting the use of six months’ therapy rather than longer durations of adjuvant treatment. ADJUVANT THERAPY FOR STAGE II COLORECTAL CANCER There is continuing controversy over the issue of adjuvant chemotherapy for patients with stage II colon cancer (T3/4, N0, M0). Part of the problem in resolving this issue has been the relative lack of randomized trials with

174

Chong and Cunningham

sufficient statistical power to detect a survival advantage in this subset of patients. Unfortunately, a comprehensive meta-analysis has not been conducted yet. Hence, the best available data are restricted to a single, large, randomized trial, and pooled analyses of selected randomized trials. In a similar manner to the original IMPACT study, the IMPACT B2 investigators conducted a pooled analysis of five randomized trials of adjuvant therapy in patients with Dukes’ B2 colon cancer (34). Their goal was to identify and quantitate any survival benefit in the subgroup of patients with Dukes’ B2 disease treated with adjuvant chemotherapy. The trials analyzed were performed by the Gruppo Interdisciplinare Valutazione Interventi Oncologia, the National Cancer Institute Canada Clinical Trials Group, the Fondation Francaise de Cancerologie Digestive, the NCCTG Intergroup, and the University of Siena group (35). One thousand and sixteen eligible B2 colon cancer patients were identified in these studies who had been randomized to adjuvant 5-FU/LV or observation. Patients in all studies were treated with comparable 5-FU/LV regimens. Duration of treatment was six months, except in the Siena study where treatment was for 12 months. With a median follow-up time of 5.75 years, no differences in event-free survival (HR 0.83; 90% CI: 0.68–1.01; p ¼ 0.061) or overall survival (HR 0.81; 90% CI: 0.64–1.01; p ¼ 0.057) were detected between treated and untreated patients. Thus, the authors concluded that the routine use of adjuvant 5-FU/LV in patients with Dukes’ B2 colon cancer was not supported by this data set. They did concede, however, that a larger data set might have revealed a statistically significant benefit, particularly in patients with poorer histologic grade. Another pooled analysis addressing the issue of adjuvant treatment for Dukes’ B colon cancer used the NSABP data set from trials C-01, C-02, C-03, and C-04 (36). Both Dukes’ B and C patients were included, but analyzed separately. Of a total of 3820 eligible patients, 1565 (41%) were Dukes’ B; 2255 (59%) were Dukes’ C. A weakness of this analysis is that the original trials differed substantially in the treatment interventions compared. Only the C-01 and C-02 trials had an observation arm. The treatment arm of trial C-02 was hepatic vein 5-FU infusion, whereas the other trials evaluated systemic chemotherapy treatment. The C-01 treatment arm was MOF, whereas the superior treatment arms for trials C-03 and C-04 were 5-FU/ LV. Individually, none of these trials showed a statistically significant survival advantage to adjuvant chemotherapy for patients with B2 disease. However, in this pooled analysis, a reduction in mortality for patients with B2 tumors without high-risk characteristics was seen when compared to observation/inferior therapy (HR 0.68; 95% CI: 0.50–0.92; p ¼ 0.01). Due to insufficient numbers, no significant improvement was seen for B2 patients with high-risk features (HR 0.80; 95% CI: 0.55–1.17; p ¼ 0.26). Contrary to the IMPACT B2 investigators, the NSABP authors concluded that there was sufficient evidence to warrant recommendation of adjuvant

Adjuvant Therapy for Colorectal Cancer

175

therapy to patients with resected Dukes’ B2 colon cancer, even in patients without adverse prognostic features. A recent pooled analysis utilized a data set including 3302 patients with stage II and III colon cancer from seven randomized trials (37). These trials comprised the five trials studied in the IMPACT B2 analysis, with the addition of an NCCTG trial and SWOG-INT 0035 (4,38). Each trial randomized patients to either 5-FU-based adjuvant therapy or no therapy following surgery. One thousand four hundred and forty patients had node-negative disease. On univariate analysis, five-year DFS in patients with stage II disease was improved by adjuvant therapy from 72% to 76% (p ¼ 0.049). However, five-year overall survival showed no difference: 80% versus 81% (p ¼ 0.11), in contrast to the clear survival advantage in patients with stage III disease. The most comprehensive systematic review published to date examined 37 clinical trials and 11 meta-analyses with respect to outcomes of adjuvant therapy versus observation in patients with stage II colon cancer (39). A meta-analysis was conducted using data on 4187 stage II patients and revealed a trend towards reduced mortality in patients receiving 5-FU-based chemotherapy compared to observation (HR 0.87; 95% CI: 0.75–1.01; p ¼ 0.07). As a result of this systematic review, the 2004 American Society of Clinical Oncology guidelines do not recommend routine treatment with adjuvant chemotherapy for patients with stage II colon cancer (40). However, the guidelines do state that patients with high-risk stage II disease may wish to consider the option of adjuvant chemotherapy. Unfortunately, the QUASAR-1 trial results were not available for inclusion in this analysis. As described later, the results of this trial may have significantly changed the result of the meta-analysis. Two stand-alone randomized trials have reported specifically on the role of adjuvant therapy in patients with Dukes’ B colon cancer. The Dutch NACCP study authors reported a survival benefit for adjuvant therapy in the subset patients with stage II disease. Of the total of 1029 patients, 468 had Dukes’ B colorectal cancer and appeared to experience a reduction in death rate of 19%, with an increase in five-year survival from 70% to 78% (95% CI and p-value not stated). This magnitude of benefit was similar to that seen for patients with Dukes’ C tumors. The QUASAR-1 study was a component of the original QUASAR trial, which randomized 3238 patients with ‘‘uncertain’’ indications for adjuvant therapy to either observation or 5-FU modulated by high- or low-dose LV, with or without levamisole (41). Ninety-one percent of patients had Dukes’ B stage and 71% had colon tumors. With a median follow-up duration of 4.2 years, there was reduced risk of recurrence with chemotherapy (HR 0.78; 95% CI: 0.67–0.91; p ¼ 0.001). In addition, five-year survival was improved in patients randomized to chemotherapy (HR 0.83; 95% CI: 0.71–0.97; p ¼ 0.02). The absolute improvement in five-year survival

176

Chong and Cunningham

was 2.9% (80.3% vs. 77.4%). Subgroup analysis of Dukes’ B patients alone still showed a five-year survival advantage to those treated with chemotherapy (p ¼ 0.04). The investigators cautioned that the quality-adjusted survival benefit obtained from adjuvant treatment in this population was inversely proportional to age. Nevertheless, this is the only large randomized trial to demonstrate a survival benefit, albeit small, in patients with stage II colon cancer. Unfortunately, confirmatory randomized trials of adjuvant therapy are unlikely to be performed in stage II colon cancer, due to the requirement for very large patient numbers. While a small benefit probably does exist, there is a need for a well-conducted meta-analysis of all comparable adjuvant trials to resolve this issue (42). In the meantime, it seems reasonable to recommend adjuvant treatment to selected patients with resected stage II colon cancer particularly in patients with adverse prognostic features, and with full consideration of the risk:benefit ratio.

NEWER AGENTS FOR ADJUVANT THERAPY Oral Fluoropyrimidines Oral fluoropyrimidines such as capecitabine, uracil/ftorafur (UFT), and S-1 offer an alternative to intravenously administered 5-FU, both in terms of patient convenience and altered toxicity profile. Capecitabine is a fluoropyrimidine prodrug, which is metabolized to 5-FU in a three-step process (43). The last step requires thymidine phosphorylase, which is significantly more active in tumor compared to normal tissues. Hence, the conversion of capecitabine to active metabolite occurs preferentially at tumor sites and may partially explain the different toxicity profile to intravenous (IV) 5-FU observed in clinical studies. The cytotoxic component of UFT is ftorafur (also known as tegafur), which is an orally bioavailable prodrug that is converted to 5-FU via the cytochrome P450 pathway. Uracil increases the bioavailability of 5-FU as it competes with 5-FU for dihydropyrimidine dehydrogenase (DPD). Thus, the orally administered combination of tegafur and uracil in a 1: 4 M ratio leads to therapeutic levels of 5-FU being delivered to tumor tissue (44). A recent meta-analysis of three randomized Japanese clinical trials, each comparing 12 months of oral adjuvant treatment with observation, has been reported (45). A total of 5233 patients with stage I, II, or III colorectal cancer were included in the meta-analysis; individual patient data were used, and only trials initiated prior to 1990 were included. Patients in the treatment arms received oral 5-FU, UFT, or carmofur; some patients additionally received mitomycin C. Disease-free survival was improved by the addition of oral fluoropyrimidines, with a hazard ratio of 0.85 (95% CI: 0.77–0.93; p ¼ 0.001). Overall survival was also superior in patients

Adjuvant Therapy for Colorectal Cancer

177

receiving oral adjuvant treatment. The overall hazard ratio was 0.89 (95% CI: 0.80–0.99; p ¼ 0.04), with no significant heterogeneity between the individual treatment arms. This meta-analysis suggests that oral fluoropyrimidines may be an effective adjuvant therapy when compared to observation. This is clearly consistent with the evidence supporting the use of IV 5-FU in the adjuvant colorectal cancer setting. Two studies have directly compared oral fluoropyrimidines to IV 5-FU/LV in an attempt to determine the equivalence or otherwise of these two adjuvant strategies. A randomized study of 1987 patients with resected Dukes’ C colon cancer compared 24 weeks of treatment with either oral capecitabine 2500 mg/m2 days 1–14 every 21 days; or IV 5-FU 425 mg/m2 and LV 20 mg/m2 days 1–5 every 28 days (15,46). The primary endpoint of the trial was equivalence of the two regimens with respect to DFS. Safety data showed significantly improved tolerability of capecitabine with respect to diarrhea, stomatitis, nausea/vomiting, and neutropenia. However, there was a significantly greater incidence of grade 3 palmar/plantar erythema with capecitabine (p < 0.001). Approximately 50% of patients required dose reduction in the capecitabine arm. Despite this, after a median follow-up of 3.8 years, equivalence of capecitabine to 5-FU/LV was demonstrated, with a strong trend to improved DFS in patients receiving capecitabine (HR 0.87; 95% CI: 0.75–1.00; p ¼ 0.0528). Likewise, there was a trend to improved overall survival, which did not reach statistical significance (HR 0.84; 95% CI: 0.69–1.01; p ¼ 0.0706) at the time of analysis. The NSABP C-06 trial was designed to compare outcomes in stage II or III colon cancer patients randomized to either adjuvant 5-FU/LV or oral UFT and LV (16). A total of 1608 patients with stage II or III rectal cancer were randomized to tegafur 300 mg/m2/day and uracil days 1–28, LV 90 mg/day, days 1–28, each 35 days for five cycles or 5-FU 500 g/m2, and LV 500 mg/m2 IV weekly for six of each eight weeks for three cycles. No differences in five-year DFS (66.9% vs. 68.3%; p ¼ 0.79) or five-year overall survival (78.7% vs. 78.7%; p ¼ 0.88) were demonstrated between the two arms after a mean 64 months follow-up. Both regimens had similar toxicity profiles and were generally well tolerated, with similar quality of life scores using multiple instruments. No subgroup analysis of stage II and III patients is yet available. When viewed together, these two large randomized trials comparing oral to IV fluoropyrimidine therapy represent a significant body of evidence demonstrating at least equivalence of oral therapy to the current standard of IV 5-FU/LV. In addition, capecitabine appears to be better tolerated than 5-FU/LV. Therefore, there are now sufficient data to recommend a change in standard practice from IV to oral fluoropyrimidine adjuvant chemotherapy for resected colon cancer. A recently reported study of adjuvant therapy with UFT with or without an oral protein bound agent, polysaccharide K (PSK), examined the

178

Chong and Cunningham

efficacy of an immunochemotherapy approach to postoperative therapy (47). The mechanism of action of PSK is not well defined, although it is thought to have immunomodulatory effects via the activation of lymphokineactivated killer cells and natural killer cells (48,49). Two hundred and seven patients were randomized to two years of treatment of either UFT alone, or UFT combined with PSK. At five-year follow-up, the mean DFS of the UFT/PSK group was 50.3 months (95% CI: 47.2–53.4 months) compared to 40.0 months (95% CI: 38.8–49.2 months; p ¼ 0.031). There was no difference in overall survival demonstrated; however, this could have been due to the small sample size. S-1 is an oral fluoropyrimidine that has three components: tegafur, 5-chloro-2,4-dihydroxypyridine, and potassium oxonate (50). Once tegafur is metabolized to 5-FU, its degradation is inhibited by the dihydropyrimidine dehydrogenase activity of 5-chloro-2,4-dihydroxypyridine, thus resulting in prolonged tissue concentrations of 5-FU. Potassium oxonate inhibits the phosphorylation of 5-FU within the intestine, thereby potentially reducing 5-FU-related gastrointestinal toxicity. A phase II study of 37 evaluable patients with metastatic colorectal cancer demonstrated a response rate of 24% (51). Significant diarrhea was noted, however, with over 35% of patients experiencing grade 3/4 diarrhea. No adjuvant colorectal studies of S-1 have been reported to date. Oxaliplatin Oxaliplatin is a third-generation platinum compound, which contains a 1,2-diaminocyclohexane carrier ligand (52). It is this bulky 1,2-diaminocyclohexane ligand that is thought to account for a broad spectrum of antitumor activity, which only partially overlaps with that of cisplatin. As with other platinum agents, oxaliplatin forms platinum-DNA adducts, which cause intra- and interstrand cross-linking of DNA molecules, thus inhibiting DNA replication. Unlike cisplatin, oxaliplatin displays substantial clinical activity against colorectal cancer, both as a single agent, and in combination with fluoropyrimidines (53,54). In patients with previously untreated advanced disease, oxaliplatin in combination with bolus/infusion 5-FU/LV (FOLFOX) has shown response rates in the region of 45% to 50% and overall survival durations of 16 to 19 months in randomized trials (55,56). The effect of adding oxaliplatin to 5-FU/LV as adjuvant colorectal therapy was investigated in the MOSAIC trial (14). Two thousand two hundred and forty-six patients with Dukes’ B2 or C colon cancer were randomized to six months’ treatment with FOLFOX4 or LV5-FU2 (de Gramont schedule). After a median follow-up time of 37 months, improved three-year DFS was evident in the FOLFOX4 arm (78.2%, 95% CI: 75.6– 80.7%) compared to LV5-FU2 (72.9%, 95% CI: 70.2–75.7%; p ¼ 0.002). A statistically significant reduction in relapse risk was seen in the Dukes’

Adjuvant Therapy for Colorectal Cancer

179

C subgroup (HR 0.76; 95% CI: 0.62–0.92), and a trend of similar magnitude was identified for patients with Dukes’ B2 tumors (HR 0.80; 95% CI: 0.56–1.15). At the cutoff date for this analysis, there was no overall survival advantage to FOLFOX4 (HR 0.90; 95% CI: 0.71–1.13). Grade 3/4 neutropenia was more frequent in the FOLFOX4 arm: 41.1% versus 4.7%. Peripheral neuropathy occurred in 92% of patients who received oxaliplatin. In total, 12.4% developed grade 3 neuropathy; however, this persisted for 12 months in only 1% of patients. All-cause treatment mortality was 0.5% in both arms. The major issue arising from the MOSAIC trial results is whether the demonstrated improvement in DFS is sufficiently compelling to warrant an immediate change in standard adjuvant treatment for stage III colon cancer from 5-FU/LV to FOLFOX4. The correlation between three-year DFS and five-year overall survival in adjuvant colon trials has now been addressed in a large, although single analysis (57). In this analysis of 15 trials, which comprised a total of 24 IV 5-FU-based treatment arms and nine observation arms, a robust correlation between three-year DFS and five-year survival was demonstrated. A small attenuation in benefit was observed between these parameters; however, in almost all trials, the presence of a statistically significant DFS advantage at three years persisted at five years. Importantly, this analysis did not include oxaliplatin-containing trials, so its applicability to studies such as the MOSAIC trial is currently unknown. However, it does provide strong support for the notion that the observed three-year DFS results in MOSAIC could predict for survival at five years. The NSABP C-07 trial randomized 2407 patients with stage II or III colon cancer to receive either adjuvant 5-FU/LV (weekly bolus, 6 out of 8 weeks) or 5-FU/LV (same schedule) plus oxaliplatin (FLOX) (58). The addition of oxaliplatin to the weekly bolus schedule of 5FU/LV improved 3-year DFS (76.5% vs 71.6%, p ¼ 0.004). The observed magnitude of improvement was strikingly similar to that documented in the MOSAIC trail, thus confirming the additional benefit from adding oxaliplatin of 5FU/LV in terms of DFS. Irinotecan Irinotecan is a semisynthetic derivative of camptothecin, which is a cytotoxic alkaloid extracted from Camptotheca acuminate. Together with its active metabolite, SN-38, irinotecan inhibits the action of topoisomerase I by preventing religation of single-strand DNA breaks, and causing doublestranded breaks, thereby leading to apoptosis (59). Activity in patients with advanced colorectal cancer has been demonstrated in a series of phase II and phase III studies, both as a single agent and in combination with fluoropyrimidines (60,61). At present, there are five planned, accruing or recently closed randomized trials evaluating irinotecan-based adjuvant therapy for resected colon

180

Chong and Cunningham

cancer (62). The CALGB 89803 randomized patients with stage III colon cancer to irinotecan 125 mg/m2, 5-FU 500 mg/m2, and LV 20 mg/m2 (IFL) given weekly for four weeks on and two weeks off for a total of 30 weeks; or 5-FU 500 mg/m2 and LV 500 mg/m2 weekly for six weeks on, two weeks off (total of 32 weeks) (63). This trial was prematurely closed to accrual in 2001 due to the findings of an unplanned interim analysis, which found a higher treatment-related death rate in patients treated with IFL (64). One thousand two hundred and sixty-four patients were enrolled and after a median follow-up time of 2.6 years, no advantage in failure-free survival (HR not stated; p ¼ 0.84) or overall survival (HR not stated; p ¼ 0.88) was observed. The investigators concluded that IFL was an inferior regimen compared to 5-FU/LV for the adjuvant treatment of colon cancer. The European Organisation for Research and Treatment of Cancer PETACC-3 trial was designed to compare 5-FU/LV to 5-FU/LV plus irinotecan in patients with resected stage III colon cancer. An important distinction between this trial and CALGB 89803 was the use of bolus/infusion schedules of 5-FU/LV rather than the bolus schedule of IFL. Individual centers had the choice of administering 5-FU/LV according to the de Gramont or AIO schedules, and irinotecan could be administered as a weekly dose of 80 or 180 mg/m2 every two weeks. A total of 3278 patients with stage II or III colon were randomized, of which 2333 had stage III disease. After a median follow-up of 32 months, the hazard ratio for 3-year replace-free survival for stage III patients was 0.87 (0.75–1.02) ( p ¼ 0.076). The hazard ratio for stage II and III patients combined was 0.87 (0.76–0.99) (p ¼ 0.038) leading to the conclusion that irinotecan may add to the efficacy of 5-FU/LV in the adjuvant setting, although these data are not as convincing as the MOSAIC and C-07 trials. Accrual of approximately 400 patients has been completed for the French ACCORD 2 trial. Patients with high-risk stage III colon cancer were randomized to adjuvant LV5-FU2, or LV5-FU2 plus irinotecan (FOLFIRI). High risk was defined as N2 disease or N1 disease with perforation or obstruction. The primary endpoint is event-free survival at three years, with overall survival and quality of life being secondary endpoints. QUASAR 2 is a U.K. and European trial that commenced accrual in late 2004. Patients with high-risk stage II or stage III colorectal cancer are randomized to one of three arms: 5-FU/LV (bolus or bolus/infusion), capecitabine plus irinotecan, or capecitabine/irinotecan plus bevacizumab. The accrual target is 3750 patients with a primary endpoint of disease-free survival; overall survival will be a secondary endpoint. In order to determine whether irinotecan-based adjuvant therapy is beneficial in stage II colon cancer, the PETACC 4 study is randomizing patients between surgery alone, and adjuvant therapy with 5-FU/LV plus irinotecan. The accrual target is 1960 patients.

Adjuvant Therapy for Colorectal Cancer

181

Monoclonal Antibodies Several mAbs have been examined in clinical trials for patients with resected colon cancer. Edrecolomab is a murine IgG2a mAb to the glycoprotein antigen 17-1A (or epithelial cell adhesion molecule—EpCAM), which is present on the cell surface of many epithelial cancers as well as normal tissues. An initial randomized study found edrecolomab to be superior to observation in patients with resected stage III colorectal cancer (65). On the basis of these promising data, a subsequent randomized study comprising 2761 patients compared 5-FU/LV, 5-FU/LV plus edrecolomab, and edrecolomab alone (66). Disappointingly, edrecolomab monotherapy was associated with significantly poorer disease-free and overall survival compared to 5-FU/LV. Furthermore, the addition of edrecolomab to 5-FU/LV did not improve disease-free or overall survival. A novel approach using an anti-idiotypic mAb strategy has been studied using the carcinoembryonic antigen (CEA) as the target. 3H1 (CeaVac1) is a murine mAb with an idiotype designed to mimic the three-dimensional structure of CEA (67). Via the anti-idiotypic network hypothesis, the immune system can be taught to recognize a self-antigen, if that antigen is presented in a different molecular environment, such as in the context of a murine antibody (68,69). The rationale of vaccination using 3H1 is to stimulate the immune system to generate specific antibodies targeting the 3H1 idiotype and therefore also CEA-bearing tumor cells. In a phase II study, 32 patients with resected Dukes’ B, C, and D colorectal cancer were treated postoperatively with weekly induction and monthly maintenance injections of 3H1. The treatment was generally well tolerated, and was found to induce potent anti-CEA cellular and humoral immune responses. A phase III study comparing 5-FU/LV and placebo to 5-FU/LV and 3H1 in patients with advanced colorectal cancer also demonstrated induction of anti-CEA immune responses; however, this did not translate into progression-free or overall survival benefit (70). The strategy of using mAb to specifically target tumor cell signaling pathways has been utilized with two novel mAbs and studied in patients with advanced colorectal cancer. Cetuximab is a chimeric mAb targeting EGFR, which is present in between 25% and 80% of colorectal cancers (71). EGFR is a member of a family of receptors, which have key roles in the maintenance of normal cellular function via their tyrosine kinasemediated effects on intracellular signaling. Inhibition of EGFR pathways in tumor cells results in cell cycle arrest and causes apoptosis (72,73). In the setting of irinotecan-refractory advanced disease, cetuximab treatment is associated with tumor response, whether given alone or in combination with irinotecan. In a randomized study, patients treated with irinotecan and cetuximab had higher responses than those treated with cetuximab alone, implying that cetuximab was able to resensitize tumor cells to the effect of irinotecan (74).

182

Chong and Cunningham

Additional toxicity from cetuximab was essentially limited to acneiform rash, which was variable in severity but generally well tolerated. Preliminary results from a phase II study of patients with metastatic colorectal cancer suggest enhanced response rates to FOLFOX when combined with cetuximab (75). In the adjuvant setting, cetuximab is being studied in a large U.S. intergroup trial (N0447), which is randomizing patients to FOLFOX (six months), FOLFIRI (six months), or FOLFOX (three months) followed by FOLFIRI (three months). In addition, within each chemotherapy arm, there is a further randomization to receive cetuximab or not. Once complete, this trial will provide valuable information in the adjuvant setting on the relative benefits of adding cetuximab to oxaliplatin versus irinotecan. Bevacizumab is a humanized mAb that targets VEGF, a growth factor essential for tumor angiogenesis and thus tumor growth. Preclinical studies have shown activity in a number of different tumor models (76). In a colorectal xenograft model, mice receiving bevacizumab showed a dose-dependent reduction in tumor number and volume compared to mice treated with a control mAb (77). A randomized phase II study of 104 untreated metastatic colorectal cancer patients evaluated 5-FU/LV, 5-FU/LV and bevacizumab (5 mg/kg), and 5-FU/LV and bevacizumab (10 mg/kg) (78). This trial demonstrated a higher response rate, progression-free and overall survival in patients treated with 5-FU/LV and low-dose bevacizumab compared to 5-FU/LV alone. A subsequent large randomized study of first-line treatment for metastatic colorectal cancer showed significant progression-free and overall survival benefit for patients treated with bevacizumab plus irinotecan/5-FU/LV (IFL) versus IFL alone (79). The main toxicities of bevacizumab were hypertension and possibly an increased but still small risk of bowel perforation compared to IFL alone. On the basis of these data in patients with metastatic colorectal cancer, adjuvant trials of bevacizumab-containing regimens have commenced. The NSABP C-08 trial has a two-arm randomized design with patients receiving modified FOLFOX 6 either alone or in combination with bevacizumab. The target accrual of 2623 patients is expected over 2.5 years. The currently accruing AVANT study is randomizing patients with stage III and high risk stage II colon cancer to FOLFOX4, FOLFOX4 plus bevacizumab or capecitabine/ oxaliplatin plus bevacizumab. The target accrual is 3450 patients. Other Novel Agents An alternative approach to EGFR and VEGF inhibition is the blockade of the respective receptor tyrosine kinase domains with small molecule inhibitors such as gefitinib or erlotinib. There is preliminary evidence of safety and efficacy of gefitinib in combination with FOLFOX in patients with advanced colorectal cancer (80). In a similar vein to cetuximab, future trials of gefitinib and other receptor tyrosine kinase domains with small molecule inhibitors may evaluate their role in the adjuvant setting.

Adjuvant Therapy for Colorectal Cancer

183

A number of other agents may in future be available for the adjuvant treatment of colorectal cancer. Proteasome inhibition with bortezomib is a strategy being evaluated in a number of malignancies, and has shown activity in a colon cancer xenograft model (81). The combination of bortezomib and irinotecan has demonstrated safety and tolerability in patients with relapsed or refractory colorectal cancer (82). While this and other new agents need to be further evaluated in patients with advanced disease, there is the potential for translation into colorectal cancer adjuvant therapy. PROGNOSTIC FACTORS AND ADJUVANT THERAPY Risk stratification on the basis of histopathologic TNM stage has been well documented and is the basis for adjuvant chemotherapy as described in the previous sections. In addition to presence or absence of lymph node metastases, the number of histologically examined nodes is, in itself, an important prognostic variable (83). While it has been suggested that a low number of examined nodes might be an indication for adjuvant therapy in stage II colon cancer, this has not been prospectively studied (84). With the advent of new molecular tools, several biomarkers have been described, which may help to more precisely define an individual patient’s risk for tumor recurrence. This progress has implications for the administration of adjuvant therapy, which in the future may be tailored according to a panel of several prognostic and/or predictive markers, rather than the relatively imprecise histopathology-based methods of today. While several histopathologically and clinically derived variables have been thought to confer additional risk of relapse after curative resection, these risks have been difficult to quantify. In an effort to incorporate some of these variables, a German group has developed a tree-based risk stratification method, which was developed using colorectal cancer registry data (85). Six hundred and forty-one patients with resected colon cancer were analyzed with respect to factors including colonic site, pT and pN stage, number of examined and involved nodes, histologic grade and tumor type, lymphovascular invasion, and emergency presentation. A prognostic tree was developed which contained pT, number of involved nodes, venous invasion, and emergency presentation as the critical variables. This model was then validated using a separate sample of 338 consecutive patients and compared with recognized stage grouping systems. Based on this validation sample, the prognostic tree was superior in predicting disease relapse compared to conventional stage grouping systems including International Union against Cancer TNM and Dukes’ classification. In another study, an analysis of 363 patients with resected colorectal cancer produced a simple four-point scale, which included the parameters of venous invasion, depth of primary tumor penetration, and regional lymph node status (86). In a validation sample of 231 patients, this scale was found

184

Chong and Cunningham

to be a better predictor of relapse compared to Dukes’, Astler-Coller, and TNM staging. Progress in defining molecular prognostic factors is being made. The most studied markers include EGFR, VEGF, p53, Ki-67, and TS. EGFR is strongly expressed in 20% to 35% of colon adenocarcinomas and is associated with reduced survival in several retrospective studies (87–89). Several other markers have also been investigated in association with EGFR. A study of 134 consecutive patients with stage II colon cancer treated with surgery alone (no adjuvant therapy) examined immunohistochemical expression of EGFR, p53, c-MET, and b-catenin (88). In a multivariate analysis, nuclear p53 expression (p ¼ 0.02) and strong EGFR (2þ/3þ) expression (p ¼ 0.01) were associated with poor survival. The prognostic value of VEGF expression is currently unclear. In one study, stage II colon tumors expressing VEGF were associated with poorer relapse-free survival compared to VEGF-negative tumors (90). However, in other studies that examined several factors via multivariate analyses, VEGF expression did not appear to be an independent risk factor for survival (89,91). A recent study examined the prognostic value of preoperative serum VEGF levels in 81 patients with colon carcinoma who were undergoing surgery (92). Patients with Dukes’ C or high-risk Dukes’ B tumors subsequently received adjuvant chemotherapy. Of a large panel of clinical and histopathological baseline variables, only Dukes’ stage and preoperative serum VEGF levels were predictive of DFS (p ¼ 0.001). In a similar way to tumor VEGF expression, high serum VEGF levels correlated with shorter DFS. Preoperative VEGF levels were also significantly lower in patients who were able to undergo curative surgery compared to noncurative surgery (p < 0.0001). Unfortunately, tumor VEGF expression was not measured in this study; however, the apparent association between serum VEGF level and DFS is provocative. High TS expression has been shown in several studies to be associated with poor survival; however, heterogeneous sample sizes and study methodologies have made interpretation of these studies difficult. A meta-analysis of 20 studies including both localized and advanced disease settings concluded that TS expression is indeed predictive for survival (93). Analysis of the seven adjuvant studies demonstrated an association between high TS expression and poorer overall survival but not progression-free survival (PFS). One of the studies included in the meta-analysis was a retrospective analysis of NSABP trials C-01–C-04 and merits discussion in its own right. TS, Ki-67, and p53 were assessed by immunohistochemistry in tumors of 706 patients and correlated with survival (94). Of the analyzed patients enrolled in these trials, 291 had Dukes’ B and 415 had Dukes’ C tumors. Multivariate analysis revealed that all three variables were significantly associated with disease-free or overall survival. Low expression of Ki-67 (p ¼ 0.03) and increased TS expression (p ¼ 0.02) were both associated with poor survival, while presence of p53 (p ¼ 0.01) and high TS (p ¼ 0.04) predicted for poor DFS.

Adjuvant Therapy for Colorectal Cancer

185

DNA microsatellite instability (MSI) is characteristic of hereditary nonpolyposis colorectal cancer, which is an autosomal dominant condition associated with germline mutations in several DNA mismatch repair genes including MLH1 and MSH2 (95). However, 10% to 15% of sporadic (nonhereditary) colorectal cancers also exhibit MSI (96,97). Several studies have demonstrated improved survival in patients with high-level MSI (MSIH) as compared to low-level MSI or microsatellite stable (MSS) patients (97–99). The mechanism for improved survival is currently unclear. A study of 100 patients with sporadic colorectal cancer examined the relationship between level of MSI, apoptosis and proliferation, and survival (100). High-level MSI patients had significantly higher proliferation based on Ki-67 immunohistochemistry, and this was associated with a trend (p ¼ 0.09) to improved survival in this group, but not in patients with lower levels of MSI. The interaction between DNA methylation and MSI has been examined in an analysis of 605 patients undergoing curative surgery for colorectal cancer (101). Patients with MSS had poorer survival compared with patients with MSI; however, presence of DNA methylation conferred even poorer prognosis on MSS patients. Interestingly, the level of DNA methylation did not appear to affect survival of patients with MSI. In order to definitively determine the value of incorporating molecular risk factors into algorithms for adjuvant therapy, well-designed prospective trials are required. In the future, an individual colorectal cancer patient may have a ‘‘risk profile’’ generated using a panel of multiple molecular and histopathological parameters. This approach might be particularly useful in patients with Dukes’ B tumors where the benefits appear to be relatively small when adjuvant treatment is applied across the board, but where defined high-risk subsets might derive significant benefit from such therapy. In addition, the presence of certain markers such as EGFR or VEGF clearly has potentially therapeutic implications for specific targeted agents.

RECOMMENDATIONS FOR ADJUVANT THERAPY As clear progress has been made in both risk stratification and adjuvant treatment strategies for colorectal cancer, treatment paradigms must necessarily change in order to accommodate the most recent data. Naturally, if further advances are to be made, both patients and clinicians must be encouraged to participate in well-designed clinical trials. However, for patients who do not have access to, or who decline to participate in such trials, the available evidence has changed in the practice of adjuvant therapy. On the basis of the MOSAIC and C-07 data, medically fit patients with resected stage III colorectal cancer should be offered adjuvant treatment with an oxaliplatin-based regimen. This recommendation is based on the clear benefit shown in both trials in three-year DFS, which is likely to translate

186

Chong and Cunningham

into five-year survival (14,57). Stage III patients who are deemed not suitable, or who decline such therapy, may be treated instead with an oral fluoropyrimidine, such as capecitabine, with the expectation of at least equivalent efficacy and reduced toxicity compared with IV 5-FU. This recommendation is based on statistically ro bust equivalence demonstrated in three-year DFS in the X-ACT trial, as well as five-year survival in NSABP C-06 (15). There is now little doubt that patients with stage II colorectal cancer obtain survival benefit from IV 5-FU-based adjuvant chemotherapy (41). Therefore, in the context of the X-ACT and NSABP C-06 data, patients with average-risk stage II tumors should be considered candidates for adjuvant treatment with an oral fluoropyrimidine. Medically fit stage II patients with poor-risk histopathological features should in addition be considered for oxaliplatin-based therapy, as the magnitude of benefit appears to be similar to that achieved in patients with stage III disease (14). Clearly, the risk:benefit ratio needs to be carefully assessed for each patient, especially in patients who have reduced life expectancy for noncancer reasons. There are Web-based resources available to assist oncologists and patients in making decisions about adjuvant therapy; one such on-line tool is found in the website mentioned in reference (102). The additional issue of whether capecitabine can be substituted for LV5-FU2 in oxaliplatin-containing regimens has not yet been answered in a randomized fashion; however, this seems likely to be the case based on comparable single agent efficacies of capecitabine and 5-FU.

CONCLUSIONS We live in an exciting era of progress in adjuvant colorectal cancer therapy. After over a decade where adjuvant 5-FU-based therapy was considered standard for patients with stage III colorectal cancer, the available data now demonstrate a measurable advantage to oral fluoropyrimidines and oxaliplatin-based combination therapy. These data have changed adjuvant practice, although overall survival data are still awaited. Because of the efficacy and favorable toxicity profile of molecularly targeted mAbs, these are now rapidly being incorporated into the latest generation of clinical trials. The optimal ways to combine and sequence the multiple agents now available are yet to be defined. Further novel agents may become available for adjuvant trials if efficacy is demonstrated in the setting of metastatic disease. There is now substantial evidence supporting a role for adjuvant treatment of stage II colorectal cancer. Progress is being made in identifying additional subgroups of patients using novel biomarkers, who might also benefit from adjuvant therapy. Hopefully, with the advances currently being made, the prognosis of patients with resected colorectal cancer will be significantly improved.

Adjuvant Therapy for Colorectal Cancer

187

REFERENCES 1. Colon and rectum. In: Greene FL, Page DL, Fleming ID, et al., eds. American Joint Committee on Cancer: AJCC cancer staging manual. 6th ed. New York, New York: Springer, 2002:113–124. 2. 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; 122:321–326. 3. Wolmark N, Fisher B, Rockette H, et al. Postoperative adjuvant chemotherapy or BCG for colon cancer: results from NSABP protocol C-01. J Natl Cancer Inst 1988; 80:30–36. 4. Moertel CG, Fleming TR, Macdonald JS, et al. Levamisole and fluorouracil for adjuvant therapy of resected colon carcinoma. N Engl J Med 1990; 322:352–358. 5. Moertel CG, Fleming TR, Macdonald JS, et al. Intergroup study of fluorouracil plus levamisole as adjuvant therapy for stage II/Dukes’ B2 colon cancer. J Clin Oncol 1995; 13:2936–2943. 6. Wolmark N, Rockette H, Fisher B, et al. The benefit of leucovorin-modulated fluorouracil as postoperative adjuvant therapy for primary colon cancer: results from National Surgical Adjuvant Breast and Bowel Project protocol C-03. J Clin Oncol 1993; 11:1879–1887. 7. O’Connell MJ, Mailliard JA, Kahn MJ, et al. Controlled trial of fluorouracil and low-dose leucovorin given for 6 months as postoperative adjuvant therapy for colon cancer. J Clin Oncol 1997; 15:246–250. 8. O’Connell MJ, Laurie JA, Kahn M, et al. Prospectively randomized trial of postoperative adjuvant chemotherapy in patients with high-risk colon cancer. J Clin Oncol 1998; 16:295–300. 9. Haller DG, Catalano PJ, Macdonald JS, Mayer RJ. Fluorouracil, leucovorin and levamisole adjuvant therapy for colon cancer: five year final report of INT-0089. Proc Am Soc Clin Oncol 1998; 17:abstract 982. 10. Wolmark N, Rockette H, Mamounas E, et al. Clinical trial to assess the relative efficacy of fluorouracil and leucovorin, fluorouracil and levamisole, and fluorouracil, leucovorin, and levamisole in patients with Dukes’ B and C carcinoma of the colon: results from National Surgical Adjuvant Breast and Bowel Project C-04. J Clin Oncol 1999; 17:3553–3559. 11. QUASAR Collaborative Group. Comparison of flourouracil with additional levamisole, higher-dose folinic acid, or both, as adjuvant chemotherapy for colorectal cancer: a randomised trial. Lancet 2000; 355:1588–1596. 12. Saini A, Norman AR, Cunningham D, et al. Twelve weeks of protracted venous infusion of fluorouracil (5-FU) is as effective as 6 months of bolus 5-FU and folinic acid as adjuvant treatment in colorectal cancer. Br J Cancer 2003; 88:1859–1865. 13. Andre T, Colin P, Louvet C, et al. Semimonthly versus monthly regimen of fluorouracil and leucovorin administered for 24 or 36 weeks as adjuvant therapy in stage II and III colon cancer: results of a randomized trial. J Clin Oncol 2003; 21:2896–2903. 14. Andre T, Boni C, Mounedji-Boudiaf L, et al. Oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment for colon cancer. N Engl J Med 2004; 350: 2343–2351.

188

Chong and Cunningham

15. Cassidy J, Scheithauer W, McKendrick J, Kroning H. Capecitabine vs bolus 5-FU/leucovorin as adjuvant therapy for colon cancer (the X-ACT study): efficacy results of a phase III trial. Proc Am Soc Clin Oncol 2004; 23:abstract 3509. 16. Wolmark N, Wieand S, Lembersky B, Colangelo L, Smith R. A phase III trial comparing oral UFT to FULV in stage II and III carcinoma of the colon: results of NSABP protocol C-06. Proc Am Soc Clin Oncol 2004; 23:abstract 3508. 17. Grem JL. Systemic treatment options in advanced colorectal cancer: perspectives on combination 5-fluorouracil plus leucovorin. Semin Oncol 1997; 24:S18. 18. Houghton JA, Weiss KD, Williams LG, Torrance PM, Houghton PJ. Relationship between 5-fluoro-20 -deoxyuridylate, 20 -deoxyuridylate, and thymidylate synthase activity subsequent to 5-fluorouracil administration, in xenografts of human colon adenocarcinomas. Biochem Pharmacol 1986; 35:1351–1358. 19. Houghton JA, Houghton PJ. On the mechanism of cytotoxicity of fluorinated pyrimidines in four human colon adenocarcinoma xenografts maintained in immune-deprived mice. Cancer 1980; 45:1159–1167. 20. Lavin P, Mittelman A, Douglass H Jr., Engstrom P, Klaassen D. Survival and response to chemotherapy for advanced colorectal adenocarcinoma: an Eastern Cooperative Oncology Group report. Cancer 1980; 46:1536–1543. 21. O’Connell MJ. A phase III trial of 5-fluorouracil and leucovorin in the treatment of advanced colorectal cancer. A Mayo Clinic/North Central Cancer Treatment Group study. Cancer 1989; 63:1026–1030. 22. Wolmark N, Rockette H, Wickerham DL, et al. Adjuvant therapy of Dukes’ A, B, and C adenocarcinoma of the colon with portal-vein fluorouracil hepatic infusion: preliminary results of National Surgical Adjuvant Breast and Bowel Project protocol C-02. J Clin Oncol 1990; 8:1466–1475. 23. Peters GJ, van Groeningen CJ, van der Wilt CL, et al. Time course of inhibition of thymidylate synthase in patients treated with fluorouracil and leucovorin. Semin Oncol 1992; 19:26–35. 24. Poon MA, O’Connell MJ, Wieand HS, et al. Biochemical modulation of fluorouracil with leucovorin: confirmatory evidence of improved therapeutic efficacy in advanced colorectal cancer. J Clin Oncol 1991; 9:1967–1972. 25. Poon MA, O’Connell MJ, Moertel CG, et al. Biochemical modulation of fluorouracil: evidence of significant improvement of survival and quality of life in patients with advanced colorectal carcinoma. J Clin Oncol 1989; 7: 1407–1418. 26. Van Ginckel R, Distelmans W, De Brabander M, et al. Levamisole plus 5-fluorouracil inhibits the growth of human colorectal xenografts in nude mice. Eur J Cancer 1992; 28A:1137–1139. 27. Wolmark N, Bryant J, Smith R, et al. Adjuvant 5-fluorouracil and leucovorin with or without interferon alfa-2a in colon carcinoma: National Surgical Adjuvant Breast and Bowel Project protocol C-05. J Natl Cancer Inst 1998; 90: 1810–1816. 28. Multicentre Pooled Analysis of Colon Cancer Trials (IMPACT) investigators. Efficacy of adjuvant fluorouracil and folinic acid in colon cancer. International Lancet 1995; 345:939–944.

Adjuvant Therapy for Colorectal Cancer

189

29. Porschen R, Bermann A, Loffler T, et al. Fluorouracil plus leucovorin as effective adjuvant chemotherapy in curatively resected stage III colon cancer: results of the trial adjCCA-01. J Clin Oncol 2001; 19:1787–1794. 30. Arkenau HT, Bermann A, Rettig K, Strohmeyer G, Porschen R. 5-Fluorouracil plus leucovorin is an effective adjuvant chemotherapy in curatively resected stage III colon cancer: long-term follow-up results of the adjCCA-01 trial. Ann Oncol 2003; 14:395–399. 31. Taal BG, Van Tinteren H, Zoetmulder FA. Adjuvant 5-FU plus levamisole in colonic or rectal cancer: improved survival in stage II and III. Br J Cancer 2001; 85:1437–1443. 32. Zalcberg JR, Siderov J, Simes J. The role of 5-fluorouracil dose in the adjuvant therapy of colorectal cancer. Ann Oncol 1996; 7:42–46. 33. Starling N, Chau I, Norman AR, Tait D, Iveson T. A randomised comparison between six months of bolus fluorouracil/leucovorin and twelve weeks of protracted venous infusion 5-FU as adjuvant treatment in colorectal cancer: an update with 5 years’ follow-up. Proc Am Soc Clin Oncol 2004; 23:abstract 3523. 34. International Multicentre Pooled Analysis of B2 Colon Cancer Trials (IMPACT B2) investigators. Efficacy of adjuvant fluorouracil and folinic acid in B2 colon cancer. J Clin Oncol 1999; 17:1356–1363. 35. Francini G, Petrioli R, Lorenzini L, et al. Folinic acid and 5-fluorouracil as adjuvant chemotherapy in colon cancer. Gastroenterology 1994; 106:899–906. 36. Mamounas E, Wieand S, Wolmark N, et al. Comparative efficacy of adjuvant chemotherapy in patients with Dukes’ B versus Dukes’ C colon cancer: results from four National Surgical Adjuvant Breast and Bowel Project adjuvant studies (C-01, C-02, C-03, and C-04). J Clin Oncol 1999; 17:1349–1355. 37. Gill S, Loprinzi CL, Sargent DJ, et al. Pooled analysis of fluorouracil-based adjuvant therapy for stage II and III colon cancer: who benefits and by how much? J Clin Oncol 2004; 22:1797–1806. 38. Laurie JA, Moertel CG, Fleming TR, et al. Surgical adjuvant therapy of large-bowel carcinoma: an evaluation of levamisole and the combination of levamisole and fluorouracil. The North Central Cancer Treatment Group and the Mayo Clinic. J Clin Oncol 1989; 7:1447–1456. 39. Figueredo A, Charette ML, Maroun J, Brouwers MC, Zuraw L. Adjuvant therapy for stage II colon cancer: a systematic review from the Cancer Care Ontario Program in Evidence-Based Care’s Gastrointestinal Cancer Disease Site Group. J Clin Oncol, published online, June 15, 2004. 40. Benson AB III, Schrag D, Somerfield MR, et al. American Society of Clinical Oncology recommendations on adjuvant chemotherapy for stage II colon cancer. J Clin Oncol, published online, June 15, 2004. 41. Gray RG, Barnwell J, Hills R, McConkey C, Williams N. QUASAR: a randomized study of adjuvant chemotherapy vs observation including 3238 colorectal cancer patients. Proc Am Soc Clin Oncol 2004; 23:abstract 3501. 42. Buyse M, Piedbois P. Should Dukes’ B patients receive adjuvant therapy? A statistical perspective. Semin Oncol 2001; 28:20–24. 43. Ishikawa T, Utoh M, Sawada N, et al. Tumor selective delivery of 5-fluorouracil by capecitabine, a new oral fluoropyrimidine carbamate, in human cancer xenografts. Biochem Pharmacol 1998; 55:1091–1097.

190

Chong and Cunningham

44. Maehara Y, Kakeji Y, Ohno S, Baba H, Sugimachi K. Scientific basis for the combination of tegafur with uracil. Oncology (Huntingt) 1997; 11:14–21. 45. Sakamoto J, Ohashi Y, Hamada C, Buyse M, Burzykowski T, Piedbois P. Efficacy of oral adjuvant therapy after resection of colorectal cancer: 5-year results from three randomized trials. J Clin Oncol 2004; 22:484–492. 46. Scheithauer W, McKendrick J, Begbie S, et al. Oral capecitabine as an alternative to i.v. 5-fluorouracil-based adjuvant therapy for colon cancer: safety results of a randomized, phase III trial. Ann Oncol 2003; 14:1735–1743. 47. Ohwada S, Ikeya T, Yokomori T, et al. Adjuvant immunochemotherapy with oral tegafur/uracil plus PSK in patients with stage II or III colorectal cancer: a randomised controlled study. Br J Cancer 2004; 90:1003–1010. 48. Harada M, Matsunaga K, Oguchi Y, et al. Oral administration of PSK can improve the impaired anti-tumor CD4þ T-cell response in gut-associated lymphoid tissue (GALT) of specific-pathogen-free mice. Int J Cancer 1997; 70: 362–372. 49. Hirose K, Zachariae CO, Oppenheim JJ, Matsushima K. Induction of gene expression and production of immunomodulating cytokines by PSK in human peripheral blood mononuclear cells. Lymphokine Res 1990; 9:475–483. 50. Shirasaka T, Shimamato Y, Ohshimo H, et al. Development of a novel form of an oral 5-fluorouracil derivative (S-1) directed to the potentiation of the tumor selective cytotoxicity of 5-fluorouracil by two biochemical modulators. Anticancer Drugs 1996; 7:548–557. 51. Van den BJ, Schoffski P, Schellens JH, et al. EORTC Early Clinical Studies Group early phase II trial of S-1 in patients with advanced or metastatic colorectal cancer. Br J Cancer 2003; 88:648–653. 52. O’Dwyer PJ, Johnson SW. Current status of oxaliplatin in colorectal cancer. Semin Oncol 2003; 30:78–87. 53. Becouarn Y, Ychou M, Ducreux M, et al. Phase II trial of oxaliplatin as firstline chemotherapy in metastatic colorectal cancer patients. Digestive Group of French Federation of Cancer Centers. J Clin Oncol 1998; 16:2739–2744. 54. de Gramont A, Vignoud J, Tournigand C, et al. Oxaliplatin with high-dose leucovorin and 5-fluorouracil 48-hour continuous infusion in pretreated metastatic colorectal cancer. Eur J Cancer 1997; 33:214–219. 55. Goldberg RM, Sargent DJ, Morton RF, et al. A randomized controlled trial of fluorouracil plus leucovorin, irinotecan, and oxaliplatin combinations in patients with previously untreated metastatic colorectal cancer. J Clin Oncol 2004; 22:23–30. 56. de Gramont A, Figer A, Seymour M, et al. Leucovorin and fluorouracil with or without oxaliplatin as first-line treatment in advanced colorectal cancer. J Clin Oncol 2000; 18:2938–2947. 57. Sargent DJ, Wieand S, Benedetti J, Labianca R, Haller DG. Disease-free survival vs overall survival as a primary endpoint for adjuvant colon cancer studies: individual patient data from 12915 patients on 15 randomised trials. Proc Am Soc Clin Oncol 2004; 23:abstract 3502. 58. Wolmark N, Wieand HS, Kuebler JP, Colangelo L, Smith RE. A phase III trial comparing FULV to FULV þ oxaliplatin in stage II or III carcinoma of the

Adjuvant Therapy for Colorectal Cancer

59.

60.

61.

62. 63.

64.

65.

66.

67.

68. 69. 70.

71.

72.

191

colon: Results of NSABP Protocol C-07. J Clin Oncol, 2005 ASCO Annual Meeting Proceedings. Vol 23, No. 160, Part I of II (June 1 Supplement), 2005: 3500. Voigt W, Matsui S, Yin MB, Burhans WC, Minderman H, Rustum YM. Topoisomerase-I inhibitor SN-38 can induce DNA damage and chromosomal aberrations independent from DNA synthesis. Anticancer Res 1998; 18: 3499–3505. Rougier P, Bugat R, Douillard JY, et al. Phase II study of irinotecan in the treatment of advanced colorectal cancer in chemotherapy-naive patients and patients pretreated with fluorouracil-based chemotherapy. J Clin Oncol 1997; 15:251–260. Douillard JY, Cunningham D, Roth AD, et al. Irinotecan combined with fluorouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal cancer: a multicentre randomised trial. Lancet 2000; 355:1041–1047. Douillard JY, Barbarot V, Bennouna J. Update on European adjuvant trials with irinotecan for colorectal cancer. Oncology (Huntingt) 2002; 16:13–15. Saltz LB, Niedzwiecki D, Hollis D, Goldberg RM, Hantel A. Irinotecan plus fluorouracil/leucovorin (IFL) versus fluorouracil/leucovorin alone (FL) in stage III colon cancer (intergroup trial CALGB C89803). Proc Am Soc Clin Oncol 2004; 23:abstract 3500. Rothenberg ML, Meropol NJ, Poplin EA, Van Cutsem E, Wadler S. Mortality associated with irinotecan plus bolus fluorouracil/leucovorin: summary findings of an independent panel. J Clin Oncol 2001; 19:3801–3807. Riethmuller G, Schneider-Gadicke E, Schlimok G, et al. Randomised trial of monoclonal antibody for adjuvant therapy of resected Dukes’ C colorectal carcinoma. German Cancer Aid 17-1A Study Group. Lancet 1994; 343: 1177–1183. Punt CJ, Nagy A, Douillard JY, et al. Edrecolomab alone or in combination with fluorouracil and folinic acid in the adjuvant treatment of stage III colon cancer: a randomised study. Lancet 2002; 360:671–677. Foon KA, John WJ, Chakraborty M, et al. Clinical and immune responses in resected colon cancer patients treated with anti-idiotype monoclonal antibody vaccine that mimics the carcinoembryonic antigen. J Clin Oncol 1999; 17: 2889–2895. Jerne NK. Towards a network theory of the immune system. Ann Immunol (Paris) 1974; 125C:373–389. Lindenmann J. Speculations on idiotypes and homobodies. Ann Immunol (Paris) 1973; 124:171–184. Chong G, Bhatnagar A, Cunningham D, et al. Phase III trial of 5-fluorouracil and leucovorin plus either 3H1 anti-idiotype monoclonal antibody or placebo in patients with advanced colorectal cancer. Ann Oncol 2006; 17:437–442. Herbst RS, Shin DM. Monoclonal antibodies to target epidermal growth factor receptor-positive tumors: a new paradigm for cancer therapy. Cancer 2002; 94:1593–1611. Moyer JD, Barbacci EG, Iwata KK, et al. Induction of apoptosis and cell cycle arrest by CP-358,774, an inhibitor of epidermal growth factor receptor tyrosine kinase. Cancer Res 1997; 57:4838–4848.

192

Chong and Cunningham

73. Karnes WE Jr., Weller SG, Adjei PN, et al. Inhibition of epidermal growth factor receptor kinase induces protease-dependent apoptosis in human colon cancer cells. Gastroenterology 1998; 114:930–939. 74. Cunningham D, Humblet Y, Siena S, et al. Cetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. N Engl J Med 2004; 351:337–345. 75. Tabernero J, Van Cutsem E, Sastre J, Cervantes A, Van Laethem J-L. An international phase II study of cetuximab in combination with oxaliplatin/5-fluorouracil/folinic acid in the first-line treatment of patients with metastatic colorectal cancer expressing epidermal growth factor receptor. Preliminary results. Proc Am Soc Clin Oncol 2004; 23:abstract 3512. 76. Kim KJ, Li B, Winer J, et al. Inhibition of vascular endothelial growth factorinduced angiogenesis suppresses tumour growth in vivo. Nature 1993; 362: 841–844. 77. Warren RS, Yuan H, Matli MR, Gillett NA, Ferrara N. Regulation by vascular endothelial growth factor of human colon cancer tumorigenesis in a mouse model of experimental liver metastasis. J Clin Invest 1995; 95:1789–1797. 78. Kabbinavar F, Hurwitz HI, Fehrenbacher L, et al. Phase II, randomized trial comparing bevacizumab plus fluorouracil (FU)/leucovorin (LV) with FU/LV alone in patients with metastatic colorectal cancer. J Clin Oncol 2003; 21: 60–65. 79. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004; 350:2335–2342. 80. Fisher GA, Kuo T, Cho CD, Halsey J, Jambalos CN. A phase II study of gefitinib in combination with FOLFOX-4 in patients with metastatic colorectal cancer. Proc Am Soc Clin Oncol 2004; 23:abstract 3514. 81. Yu J, Tiwari S, Steiner P, Zhang L. Differential apoptotic response to the proteasome inhibitor bortezomib [VELCADE, PS-341] in Bax-deficient and p21-deficient colon cancer cells. Cancer Biol Ther 2003; 2:694–699. 82. Dragovich T, Lenz HJ, Rocha Lima CMS, Kozuch P, Hochster H. Bortezomib  irinotecan in relapsed/refractory colorectal cancer: interim analysis results from phase 2b study. Proc Am Soc Clin Oncol 2004; 23:abstract 3591. 83. Le Voyer TE, Sigurdson ER, Hanlon AL, et al. Colon cancer survival is associated with increasing number of lymph nodes analyzed: a secondary survey of intergroup trial INT-0089. J Clin Oncol 2003; 21:2912–2919. 84. Prandi M, Lionetto R, Bini A, et al. Prognostic evaluation of stage B colon cancer patients is improved by an adequate lymphadenectomy: results of a secondary analysis of a large scale adjuvant trial. Ann Surg 2002; 235:458–463. 85. Radespiel-Troger M, Hohenberger W, Reingruber B. Improved prediction of recurrence after curative resection of colon carcinoma using tree-based risk stratification. Cancer 2004; 100:958–967. 86. Sternberg A, Sibirsky O, Cohen D, Blumenson LE, Petrelli NJ. Validation of a new classification system for curatively resected colorectal adenocarcinoma. Cancer 1999; 86:782–792. 87. Mayer A, Takimoto M, Fritz E, Schellander G, Kofler K, Ludwig H. The prognostic significance of proliferating cell nuclear antigen, epidermal growth

Adjuvant Therapy for Colorectal Cancer

88.

89.

90.

91.

92.

93.

94.

95. 96.

97.

98.

99.

100.

101.

102.

193

factor receptor, and mdr gene expression in colorectal cancer. Cancer 1993; 71: 2454–2460. Resnick MB, Routhier J, Konkin T, Sabo E, Pricolo VE. Epidermal growth factor receptor, c-MET, beta-catenin, and p53 expression as prognostic indicators in stage II colon cancer: a tissue microarray study. Clin Cancer Res 2004; 10:3069–3075. Khorana AA, Ryan CK, Cox C, Eberly S, Sahasrabudhe DM. Vascular endothelial growth factor, CD68, and epidermal growth factor receptor expression and survival in patients with Stage II and Stage III colon carcinoma: a role for the host response in prognosis. Cancer 2003; 97:960–968. Cascinu S, Staccioli MP, Gasparini G, et al. Expression of vascular endothelial growth factor can predict event-free survival in stage II colon cancer. Clin Cancer Res 2000; 6:2803–2807. Lee JC, Chow NH, Wang ST, Huang SM. Prognostic value of vascular endothelial growth factor expression in colorectal cancer patients. Eur J Cancer 2000; 36:748–753. De Vita F, Orditura M, Lieto E, et al. Elevated perioperative serum vascular endothelial growth factor levels in patients with colon carcinoma. Cancer 2004; 100:270–278. Popat S, Matakidou A, Houlston RS. Thymidylate synthase expression and prognosis in colorectal cancer: a systematic review and meta-analysis. J Clin Oncol 2004; 22:529–536. Allegra CJ, Paik S, Colangelo LH, et al. Prognostic value of thymidylate synthase, Ki-67, and p53 in patients with Dukes’ B and C colon cancer: a National Cancer Institute-National Surgical Adjuvant Breast and Bowel Project collaborative study. J Clin Oncol 2003; 21:241–250. Chung DC, Rustgi AK. The hereditary nonpolyposis colorectal cancer syndrome: genetics and clinical implications. Ann Intern Med 2003; 138:560–570. Jeong SY, Shin KH, Shin JH, et al. Microsatellite instability and mutations in DNA mismatch repair genes in sporadic colorectal cancers. Dis Colon Rectum 2003; 46:1069–1077. Colombino M, Cossu A, Manca A, et al. Prevalence and prognostic role of microsatellite instability in patients with rectal carcinoma. Ann Oncol 2002; 13:1447–1453. Wright CM, Dent OF, Barker M, et al. Prognostic significance of extensive microsatellite instability in sporadic clinicopathological stage C colorectal cancer. Br J Surg 2000; 87:1197–1202. Gryfe R, Kim H, Hsieh ET, et al. Tumor microsatellite instability and clinical outcome in young patients with colorectal cancer. N Engl J Med 2000; 342: 69–77. Michael-Robinson JM, Reid LE, Purdie DM, et al. Proliferation, apoptosis, and survival in high-level microsatellite instability sporadic colorectal cancer. Clin Cancer Res 2001; 7:2347–2356. Ward RL, Cheong K, Ku SL, Meagher A, O’Connor T, Hawkins NJ. Adverse prognostic effect of methylation in colorectal cancer is reversed by microsatellite instability. J Clin Oncol 2003; 21:3729–3736. www.mayoclinic.com/calcs/.

8 The Role of Radiotherapy in the Treatment of Rectal Cancer Rob Glynne-Jones and Rob Hughes Mount Vernon Centre for Cancer Treatment, Northwood, Middlesex, U.K.

BACKGROUND Approximately 15,000 people in the United Kingdom develop a cancer in the rectum each year. The rectum extends from the rectosigmoid junction to the anorectal ring. For practical purposes, the rectum is divided into a lower, middle, and upper portion. The proximal aspect of the rectum is usually defined in clinical trials as 12–16 cm from the anal verge and cancers above this level are categorized as rectosigmoid. The precise definition of the upper rectum influences the apparent local recurrence rate, since the risks of local failure are much lower for cancers in the upper rectum and rectosigmoid. In contrast, the rectum below the peritoneal reflection has no serosa. So, once a tumor has penetrated through the muscularis propria, tumor growth can extend deeply into perirectal fat. In this situation, there is a high risk of locoregional failure within the pelvis, even after potentially curative surgery (1,2). The risk of pelvic recurrence is particularly high when pathological review following surgery demonstrates very close (
1 mm) histological involvement of the circumferential resection margin as a result of inadequate lateral dissection (3). The last two decades have seen major advances in the understanding of the natural history of rectal cancer and its patterns of recurrence. This progress has led to significant improvements in treatment, especially for patients with clinically resectable rectal cancers, where surgery remains the 195

196

Glynne-Jones and Hughes

cornerstone of treatment, and has driven technical advances in surgical technique with the increasing use of meticulous sharp dissection and total mesorectal excision (TME). Some recent surgical series suggest that TME is associated with much lower rates of local recurrence, even when employed in stage III patients (4–6). Consequently, the role of adjuvant treatment with radiotherapy has also evolved with changing surgical practice and outcomes. Not only does local recurrence give rise to debilitating symptoms including intractable pain and intestinal obstruction, and a very poor quality of life (7), but is also often the main cause of death—frequently without any evidence of distant metastases. The prognosis following local recurrence is poor with a 90% chance of subsequent death from the disease (8). Once symptomatic the recurrence is almost invariably unresectable and for this reason an improvement in local pelvic control is a worthwhile objective ‘‘per se.’’ Hence, the prevention of local pelvic recurrence has been considered the ‘‘gold standard’’ endpoint for assessing rectal cancer treatments. Adjuvant radiotherapy (usually treating the posterior pelvis) both pre- and postoperatively has been used in an attempt to reduce the risk of local recurrence. Yet low local recurrence rates do not appear to impact on the risk of metastatic disease and overall survival. In the Erlangen series, the risk of metastatic disease remained at 25% between 1974 and 1991 (9)—despite major improvements in surgical technique during the period. A reduction in the rate of local recurrence may even lead to an apparent increase in the rate of distant metastases because these two events form competing risks that are not equally affected by local therapy. Therefore, the effectiveness of future treatment of rectal cancers may be better assessed on its effect on overall or disease-free survival rather than local recurrence. The use of modern staging investigations such as endorectal ultrasound, computerized tomography (CT) scanning, and positron emission tomography offer better visualization and superior information regarding tumor stage than in the past. Magnetic resonance imaging (MRI) is gaining a wider role in the United Kingdom in predicting the likelihood of involvement of the circumferential resection margin (CRM) and identifying which patients are at risk of the surgeon being unable to achieve an R0 resection (10–12). The availability of more accurate tumor imaging and multidisciplinary team working has led to the easier identification of individual patients likely to benefit from preoperative adjuvant radiotherapy. A large European multicenter study (MERCURY Study) of patients with rectal cancer has prospectively investigated the validity of this approach (13), and is likely to influence the design of future trials. Three adjuvant radiation techniques are in common use (postoperative chemoradiation, preoperative chemoradiation, and short course preoperative radiation). In the United States, postoperative concurrent chemotherapy

Radiotherapy in the Treatment of Rectal Cancer

197

and radiotherapy (CRT) has become the standard of care for stage II and III rectal cancer, following the publication of two randomized studies that demonstrated a survival advantage for postoperative CRT over chemotherapy or radiotherapy alone (14,15). This evidence led to the subsequent recommendation of the National Institute of Health consensus conference in 1990 (National Institute of Health Consensus) advocating postoperative chemoradiotherapy following the resection of locally advanced rectal tumors. In Europe, there has been more interest in the role of preoperative radiotherapy and also chemoradiation (despite the lack of any evidence base for an improved outcome over radiation alone). Centers on both sides of the Atlantic have extrapolated from the results of postoperative chemoradiation to the preoperative setting. In France in 1994, a national consensus statement recommended the use of preoperative chemoradiation in T3 and resectable T4 rectal cancer (16). The rationale for preoperative CRT is attractive, as it combines early systemic chemotherapy treatment simultaneously with a locoregional treatment. The third adjuvant radiotherapy technique commonly employed is short course preoperative radiotherapy (SCPRT). This strategy employs a short intensive course of radiotherapy over five days with the aim of reducing the risk of pelvic recurrence. This approach has gained widespread acceptance in many European and U.K. centers after the publication of the Swedish and Dutch rectal studies (17,18), which showed a reduction in local recurrence rates and in the former study an improvement in survival with this radiation schedule. In this present chapter, the current indications for pre- and postoperative radiation and SCPRT for rectal cancer, risk factors for local recurrence, the different treatment schedules, the techniques available, the potential complications and how to avoid them, and finally late effects are described. The role of synchronous chemoradiation is also discussed in the context of randomized trials. INDICATIONS FOR ADJUVANT RADIOTHERAPY The indications for delivering preoperative or postoperative radiotherapy continue to be debated on both sides of the Atlantic. The predominant aims of treatment are as follows. Reducing Local Recurrence The main indication for adjuvant radiation is to prevent local recurrence, and thereby improve survival. The rationale for this approach rests on the hypothesis that radiotherapy either in the preoperative or postoperative setting may sterilize malignant cells not removed by radical surgery.

198

Glynne-Jones and Hughes

Rendering Unresectable Tumors Resectable A further indication for using radiotherapy is to achieve tumor shrinkage or downstaging, or more precisely downsizing of the primary tumor, and thereby enabling a curative (R0) resection of an initially unresectable cancer. High preoperative doses of 45–50 Gy over four to six weeks are required to successfully downstage with a view to performing surgery approximately six weeks after the completion of radiotherapy. Short fractionation radiotherapy as used in the SCPRT is not suitable for this aim, as surgery is recommended within five days, and this interval allows insufficient time for downstaging to occur. Facilitating Sphincter-Sparing Procedures Many surgeons consider that low rectal cancers (3–6 cm from the anal verge) will inevitably require an abdominoperineal resection and creation of a permanent stoma, particularly if the sphincter is invaded. In addition, bulky anterior tumors especially in obese men with a narrow pelvis may prove technically demanding to achieve sphincter-sparing surgery. Preoperative radiotherapy may lead to tumor shrinkage at the distal margin, facilitating resection and sphincter preservation (19,20). However, the validity of this approach remains unproven, and data on late function of the sphincter mechanism following radiotherapy or chemoradiation remain elusive. RISK FACTORS FOR LOCAL RECURRENCE A high risk of local recurrence and poor survival (40–55% at five years) continues to be reported in the United States even for patients with mobile resectable rectal cancers, despite an apparently curative resection and postoperative chemoradiation (21). Many factors have been suggested to account for this high risk of local recurrence. Clinical Features Circumferential tumors or obstructing tumors have a worse prognosis (22). Also tumors that are exposed on the peritoneal surface above the peritoneal reflection have a high risk of intracoelomic and pelvic recurrence (23). Site In a series of patients with very low rectal tumors treated with TME and abdominoperineal excision of the rectum (APER), the local recurrence rate without preoperative radiation was over 30% (24). Low tumors 0–5 cm from the anal verge, or those undergoing an APER appear to have a higher risk of a positive circumferential margin and subsequent local pelvic recurrence (18). This observation is probably

Radiotherapy in the Treatment of Rectal Cancer

199

due to the narrowing of the pelvis—particularly in males—and funneling of the mesorectum preventing wide surgical clearance. T and N Stage The extent of disease as characterized by TNM staging or Dukes’ staging is an important prognostic factor for locoregional recurrence and survival. The presence of T4 disease—defined as evidence on CT imaging or clinical invasion of adjacent pelvic organs (bladder, prostate, sacrum, uterus, cervix, and vagina)—is associated with high risk of local recurrence. In the Dutch CKVO 95–04 trial, groups with a higher risk of local recurrence included stage III patients. Recent evidence also suggests that extracapsular nodal spread is a risk factor for local recurrence (25). Other factors include the depth of extension through the muscularis propria (26,27). Histological Grade Histological grade is not reliably reproducible on biopsy, and appears to have little significance in predicting the risk of local recurrence apart from defining patients where a local excision may not be appropriate. Lateral Pelvic Lymph-Node Involvement There are two pathways of lymphatic drainage for cancers in the mid and lower rectum (below the peritoneal reflection). The first pathway drains via the inferior mesenteric artery, and the second is the lateral lymphatic drainage along the internal iliac artery. Japanese surgeons have for this reason performed lateral pelvic lymph-node dissection for patients with mid and low rectal cancers. The incidence of lateral pelvic node metastases in lower rectal cancer is in the range of 15% to 20% (28,29). Histologically confirmed lymph node metastases in the lateral pelvic nodes confer a poorer prognosis both in terms of higher risk of metastatic disease and local recurrence. Standard TME does not involve removal of lateral lymph nodes; however, pelvic radiotherapy fields would usually encompass them. So some radiation oncologists believe that the benefit of radiation derives specifically from treatment of microscopic disease in these nodes. The Circumferential or Radial Margin of Excision Traditionally, histopathologists have reported on the proximal and distal surgical margins in rectal cancer. In contrast, Quirke et al. (3) aimed to define better the risks of local recurrence and carefully measured the minimum distance between the nearest extent of the tumor in the resected specimen and the CRM, i.e., the radial margin. A minimum distance of 1 mm appeared to discriminate between those patients with a high (85%) or low risk (3%) of local recurrence. Further evidence from this group with

200

Glynne-Jones and Hughes

a larger number of patients (30) lent further support to this hypothesis. A distance of
1 mm has therefore been defined as CRM involvement in several European studies (MRC CR07, the MRC Classic trial, and the Dutch CKVO 95–04 trial). Currently in the United Kingdom, now that TME is routinely practiced, reporting on the CRM is an essential component of surgical quality control. Of all the risk factors for local recurrence and metastatic disease, the circumferential margin is the only one that could potentially be influenced by treatment (surgery, chemotherapy, and radiation). Tumors that appear fixed or tethered on palpation have a higher risk that the surgeon will not be able to perform a curative resection. However, even in patients with an apparently mobile resectable rectal cancer according to digital rectal examination, a skilled surgeon will be unable to achieve a curative (histologically confirmed) R0 resection in 13% to 20% of cases. This is evident from the pathological findings in the following randomized trials: the Classic trial, the Dutch CKVO trial, the Polish trial, and the CR07 trial (Phil Quirke, personal communication) (31–33). Failure to achieve a negative resection margin (R0 histologically confirmed resection) not only confers a high risk of local recurrence but also doubles the risk of developing metastases (18,34–37). Thus a CRM of
1 mm heralds a very poor outcome even after TME resection (8). The data from the Dutch CKVO 95–04 trial suggest that preoperative radiotherapy using 25 Gy in five fractions does not compensate for a positive circumferential margin after TME surgery (32,38). In this study, the rates of local recurrence at two years are not significantly different whether the patients with a CRM
1 mm were randomized to preoperative radiotherapy or surgery alone. The addition of postoperative radiotherapy following the discovery of a close CRM also failed to demonstrate a significant change in local relapse rates (although this intervention was not randomized). If these results in terms of local failure are confirmed in the long term, this observation would support the use of more aggressive preoperative treatment to achieve more downstaging, i.e., preoperative chemoradiation. Quality of Surgery The surgeon or rather the quality of surgery is also a major prognostic factor for local recurrence (39,40) by virtue of inadequate clearance of the local tumor. Predominantly, this increase in risk reflects the circumferential margin, but tumor perforation and poor quality of the mesorectal excision even without a positive circumferential margin are important (41,42). Current studies in Europe use photographic evidence of the resected specimen to grade the quality of the surgical excision. All the above factors should be taken into account when assessing the risk of local recurrence, the need for pre- or postoperative radiotherapy or chemoradiation, and the field size required.

Radiotherapy in the Treatment of Rectal Cancer

201

THE EVIDENCE BASE FOR ADJUVANT RADIOTHERAPY IN RESECTABLE RECTAL CANCER Postoperative Radiotherapy and Chemoradiotherapy In the 1980s, preoperative assessment of tumor resectability was limited to digital rectal examination and assessment of tumor fixity. Consequently, histopathological examination of the surgical specimen offered the best prediction of outcome, and avoided overtreatment of early tumors. Therefore, a series of randomized trials tested chemotherapy, radiotherapy, and CRT in the postoperative adjuvant setting. The role of postoperative radiotherapy has been extensively tested in randomized controlled trials that use surgery alone as a control arm (14,43–50). Recent overviews/meta-analyses have demonstrated a reduction in local recurrence when postoperative radiation was administered. However, only one of the individual trials (44) has demonstrated a statistically significant reduction in local recurrence. Unfortunately, most of the trials studied in the meta-analyses were small and insufficiently powered to detect modest improvements. Postoperative adjuvant 5-FU-based chemotherapy alone following surgery was originally demonstrated to offer a survival advantage in stage II and III rectal cancer compared with surgery alone in the National Surgical Adjuvant Breast and Bowel Project (NSABP) RO1 study (48). But this study did not test the approach of combined chemoradiation. Two pivotal randomized American studies (GTSG 7175 and NCCTG 79/47/51) gave support to the use of synchronous chemoradiation for stage II and III rectal cancer (14,15). In the former study combined chemoradiation demonstrated a 24% survival advantage after a median follow-up of 94 months. The NCCTG study showed both a lower recurrence rate and improved survival over radiation alone. In the light of this evidence, the National Institute of Health consensus conference recommended 5-FU-based postoperative chemoradiation (45–50.4 Gy over five to six weeks) as the standard of care in the United States for patients with T3/T4 or node-positive disease. In contrast, with the widespread uptake of TME surgery, the approach in the United Kingdom is increasingly to select patients for postoperative chemoradiation only where there is evidence of involvement of the CRM (and where preoperative radiation has not been given). Increasingly, a positive CRM is considered to reflect a failure of preoperative staging. Seven trials have studied the role of concurrent CRT in the postoperative setting (14,15,48,49,51–53). The NSABP RO2 study (49) showed that the addition of radiotherapy produced significantly less locoregional failure as a first event (8% vs. 13% at five years) compared to adjuvant chemotherapy alone. However, the addition of radiotherapy did not influence overall relapse-free or diseasefree survival.

202

Glynne-Jones and Hughes

A European randomized trial of postoperative 5-FU-based chemoradiation to a dose of 46 Gy against surgery alone in Dukes’ B and C rectal cancer demonstrated a significant improvement in local control, disease-free survival, and overall survival for postoperative chemoradiation (52). Subsequently, studies in the postoperative setting have attempted to improve radiosensitization of the tumor by intensifying the chemotherapy component. This strategy has been associated with variable success. The American Intergroup 0014 study compared postoperative bolus 5FU alone versus 5-FU plus low-dose folinic acid, versus 5-FU and levamisole, or the combination of 5-FU levamisole and low-dose folinic acid with postoperative radiotherapy (21). None of these combination regimens gave a superior outcome over bolus 5-FU alone. The 0144 study (54,55) randomized 1917 patients with T3/T4 N0-3 rectal cancer between 1994 and 2000 to 50.4–54 Gy postoperative radiation in combination with either bolus or protracted venous infusion (PVI). With a median follow-up of 5.7 years, the disease-free and overall survival was similar for prolonged venous infusion (PVI) and the bolus regimens. However, toxicity appeared less in the arm using infusional 5-FU throughout. In contrast, the Intergroup 86–47–51 trial (56) showed a statistically significant improvement in disease-free survival and overall survival for the use of 5-FU given by continuous infusion (225 mg/m2/day) compared with bolus 5-FU in the first and fifth week of treatment when given synchronously with radiotherapy. Many of these studies have been criticized because the design delayed the use of fully adequate systemic doses of 5-FU until after the completion of the chemoradiation phase. There is no consensus regarding the optimum timing of radiotherapy in a postoperative chemoradiation protocol. However, a randomized study from Korea (57) suggested that the early timing of radiotherapy appeared to offer a significantly better disease-free survival compared with late timing of radiotherapy, although this had not produced an overall survival difference. Preoperative Radiotherapy In Europe, preoperative radiotherapy is widely considered more effective than postoperative. Almost all the preoperative randomized trials (18,44,58–73) have been performed on patients with clinically defined (usually digital rectal examination) resectable rectal cancer. In this setting, the randomized studies have compared preoperative radiotherapy to surgery alone, and in general have demonstrated improvements in local control. However, with the exception of the Dutch TME trial (18), these trials have almost certainly used conventional blunt dissection of the rectal fascia, and therefore should be interpreted with caution, as the lessons learned may not apply to present-day surgeons using TME techniques. Furthermore, only

Radiotherapy in the Treatment of Rectal Cancer

203

the Swedish Rectal Cancer trial has shown a significant improvement in overall survival following preoperative radiotherapy. Meta-analyses of Preoperative Radiotherapy Three recent systematic reviews/meta-analyses have been published on the role of preoperative radiotherapy (74–76). All three studies have pointed to a benefit from preoperative treatment. The overview published in 2000 identified 14 published controlled studies, which randomized patients between preoperative radiotherapy and surgery alone between 1970 and December 1999, but did not use individual patient data. The Colorectal Cancer Collaborative Group meta-analysis (75) identified 22 randomized controlled trials, which have compared the use of both preoperative radiotherapy (14 trials, 6350 patients) and postoperative radiotherapy (eight trials, 2157 patients) versus surgery alone. Individual patient data on surgery, site, and date of first recurrence were recorded. This meta-analysis confirmed that when preoperative radiotherapy is used a biologically equivalent dose of >30 Gy is more effective in reducing local relapse. A current Cochrane Review protocol is also examining preoperative radiotherapy in rectal cancer (77). The conclusions of these overviews/meta-analyses are limited due to disparity of the trials included. The long time period over which trials have taken place have seen considerable changes in radiotherapy technique. The use of inadequate dose schedules in the historical studies, older and cruder planning techniques, and unnecessarily large treatment fields may have contributed to a higher mortality in the treatment arms when compared to surgery alone. This observation may have obscured the benefits of preoperative radiotherapy. THE RATIONALE FOR SHORT COURSE PREOPERATIVE RADIOTHERAPY There are two approaches to preoperative radiotherapy—SCPRT, which is hypofractionated using large fractional doses (25 Gy in five fractions over five days), and conventionally fractionated long course radiotherapy (45–50.4 Gy in 25 to 28 fractions over five to six weeks). Long course radiotherapy often aims to shrink or downstage the tumor to allow an R0 resection to be performed, or to increase the chance of sphincter-sparing surgery. The primary aim of SCPRT is to kill cancer cells, which are not excised during surgery, in the draining lymph nodes or in the surrounding mesorectal tissue, and to reduce the risk of local recurrence. The Stockholm Colorectal Cancer studies, the Swedish Rectal Cancer trial, and the Dutch study (CKVO 95–04) have all confirmed that SCPRT reduces the risk of local recurrence compared to surgery alone (17,18,67,68). The short overall treatment time used in the SCPRT schedules does not allow a sufficient

204

Glynne-Jones and Hughes

interval for tumor shrinkage. Consequently, significant downstaging and sphincter sparing is not seen after SCPRT (78). SCPRT represents the ‘‘blanket’’ approach, without any selection, which offers a quick, effective, practical, and cheap schedule of preoperative radiotherapy. SHORT COURSE PREOPERATIVE RADIOTHERAPY VS. SURGERY ALONE OR POSTOPERATIVE RADIOTHERAPY In the Uppsala trial (79) SCPRT (25.5 Gy in five fractions) followed by immediate surgery was compared with postoperative radiotherapy to a dose of 60 Gy in 30 fractions. The short course schedule was more dose-effective in reducing local recurrence (13%) compared to postoperative radiotherapy (22%). Short course preoperative radiotherapy also led to less acute and late intestinal toxicity. The Swedish Rectal Cancer Study The Swedish Rectal Cancer trial recruited 1168 patients with resectable rectal cancer from 70 hospitals across Sweden (17). Patients were randomized to immediate surgery or SCPRT (25 Gy in five fractions over five days) followed by surgery. This preoperative schedule reduced the risk of local recurrence from 27% to 11% and improved overall five-year survival from 48% to 58%. The surgical techniques used during the Swedish rectal cancer trial were predominately conventional blunt dissection carried out by a large number of different surgeons across Sweden. It was therefore postulated that SCPRT may only compensate for inadequate or suboptimal surgery, and that the more widespread use of TME surgery would make SCPRT unnecessary. The Dutch Study (CKVO 95-04) For this reason, the Dutch study (18) aimed to clarify the role of SCPRT in the context of optimal TME surgery. The study recruited 1862 patients with resectable mobile tumors of the rectum. Patients were randomized either to TME surgery alone or to SCPRT (25 Gy in five fractions), followed by immediate TME surgery. The quality of the surgical technique in this study was ensured by a detailed program of training and instruction. The first five operations performed by any surgeon were directly supervised by an instructor. This study also achieved a successful program of quality assurance in the delivery/planning of the radiotherapy and also of the histopathological reporting of the rectal cancer specimen. The addition of SCPRT reduced the rate of local recurrence at two years from 8.4% to 2.4% (p < 0.0001). Despite the significant threefold reduction in local recurrence, no overall survival difference was detected at two years (82% after SCPRT to 81.8% after surgery alone). However, on this evidence many U.K. and European radiotherapy centers now offer SCPRT routinely to patients with mobile rectal tumors.

Radiotherapy in the Treatment of Rectal Cancer

205

A reduction in local recurrence was demonstrated for all tumor stages following irradiation. However, for TNM stage I tumors, this difference was not significant. The rate of recurrence after TME was also demonstrated to be very low following the resection of tumors within the upper rectum compared to those within the mid rectum or for those in the lower rectum requiring APER excision. Therefore, there seems little rationale from the evidence of this trial to offer radiotherapy to patients with clinically staged T1/T2, N0, or cancers more than 10 cm from the anal verge if a TME is to be performed, as the risks of local recurrence are small. The question whether SCPRT is sufficient adjuvant treatment for all stage II and III rectal tumors is controversial. The Dutch study showed that due to the short overall treatment time and early surgery, SCPRT did not appear to increase the rate of R0 or sphincter-sparing operations (78). SCPRT did not compensate for a positive CRM (80). Tumors where there is a high risk of incomplete surgical resection are therefore potentially more likely to benefit from long course preoperative chemoradiotherapy than from SCPRT. The recently completed MRC CR07 study aims to examine whether the standardized use of SCPRT or the selective use of postoperative chemo-radiotherapy in the presence of close <1 mm margins leads to better long-term local control (81). MORE RECENT TRIALS PERFORMED WITH CHEMORADIATION Recruitment into randomized clinical trials has often proved difficult in rectal cancer, and several studies have failed because of poor accrual. Phase II trials have usually examined patients with unresectable rectal cancer, and tested more active chemoradiation regimens (Table 1). In contrast, the phase III trials have defined a population of clinically or ultrasound staged resectable (T3/T4) patients. Modern randomized studies, which take the endpoint of survival in patients treated with concurrent CRT regimens, are identified in Table 2. The trials are remarkably similar in terms of radiation fractionation delivery schedule and specific chemotherapy agents used. Four trial strategies have been tested in this setting. Preoperative Chemoradiation vs. Postoperative Chemoradiation A popular randomized trial design has been to compare preoperative neoadjuvant CRT versus postoperative CRT. There are three trials in this setting—the NSABP RO3, the Intergroup trial INT-0147, and the German CAO/ARO/AIO-94. The NSABP R03 trial closed early because of poor accrual after randomizing only 267 patients between preoperative and postoperative CRT to a dose of 45 Gy (82,83). This study also included a neoadjuvant chemotherapy component delivered weekly for six weeks until the start of the CRT. Preliminary data have been presented, which show an increase in the proportion of patients with sphincter-sparing surgery and disease free at

206

Glynne-Jones and Hughes

Table 1 Nonrandomized Phase II Trials Employing Preoperative Chemoradiation in Rectal Cancer Showing the Percentage of Patients Achieving Pathological Complete Response RT dose (Gy)

No.

Rich et al., 1997 Bosset, 2000 Mehta, 2003 Gollins, 2004

45 45 50.4 45

390 66 32 31

Valentini, 2002 Glynne-Jones, 2003

50.4 45

16 16

Roedel, et al., 2003 Aschele, 2002

50.4 50.4

Trial

Chemotherapy

PCR rate (%)

5-FU by PVI 5-FU/leucovorin 5-FU/Campto 5-FU/Campto

26 14 37.5 20

Tomudex/oxali Xeloda/oxali

37.5 31

32

U/S T3 Clinically T3 U/S T3 Clinical locally advanced Stage II/III MRI T3/T4 (threatened CRM) U/S T3

Xeloda/oxali

19

38

U/S T2/T3/T4

5-FU (PVI)/oxali

29

Staged

Abbreviations: PCR, pathological complete response; 5-FU, 5-fluorouracil; MRI, magnetic resonance imaging; CRM, circumferential resection margin; PVI, protracted venous infusion; RT, radiotherapy; U/S, ultrasound.

one year in the preoperative arm. However, this advantage was at the expense of more frequent acute toxicity. The Intergroup trial INT-0147 also closed early because of poor accrual after randomizing only 53 patients between preoperative and postoperative CRT to a dose of 50.4 Gy. There are no results available. Table 2 Randomized Phase III Trials Employing Preoperative Chemoradiation in Rectal Cancer Showing the Percentage of Patients Achieving Pathological Complete Response Trial EORTC, 1984 INT 0147 NSABP R03 Frykholm, 2001 CAO/ARO/ AIO-94 Polish trial EORTC 22921 FFCD 9203

RT dose (Gy) 34.5 50.4 45 40 (split) 50.4 50 45 45

a Nonsignificant. Abbreviation: RT, radiotherapy.

No. 247 53 267 70 823

Chemoradiation 5% (6/126) Closed early 16% (21/130) 12% (3/29) 8% (29/363)

316 16% 1011 14% (66/479) 762 10%

Radiation 2.5% (3/121)a — — 4% (1/27)a — 1% 5% (24/476) 3%

Surgery alone — No data 0/137 — — — — —

Radiotherapy in the Treatment of Rectal Cancer

207

The German CAO/ARO/AIO-94 study protocol was initiated in 1995 to compare preoperative 5-FU-based chemoradiation with standard conventionally fractionated radiotherapy versus postoperative combined modality treatment for stage II/III resectable rectal cancer (84,85). The primary endpoints of the study were overall and relapse-free survival, and locoregional and distant control. The secondary endpoints included the rate of curative (R0) resections, sphincter-saving procedures, toxicity, and surgical complications. This study demonstrates a favorable therapeutic ratio for preoperative chemoradiation over postoperative chemoradiation in terms of local recurrence (6% vs. 13%) acute toxicity and late morbidity. Published data show no excess surgical morbidity for CRT. Due to premature closure of the Intergroup and the NSABP pre- versus postoperative CRT studies, the 800 patients recruited in the German study will offer crucial information on the role of preoperative 5-FU-based CRT. Preoperative Chemoradiation vs. Radiation Alone The second strategy has been to randomize between preoperative neoadjuvant chemoradiation and an identical schedule of preoperative radiotherapy alone. Two large studies, EORTC 22921 and FFCD 9203, have used this design. A major European study in T3/T4 resectable rectal cancer (EORTC 22921) represents a 2  2 factorial design. The study randomized between preoperative radiotherapy alone to a dose of 45 Gy versus CRT (with 5FU and leucovorin), and further randomized postoperatively to adjuvant chemotherapy with 5-FU and low-dose leucovorin or control. This trial closed in April 2003 and has accrued 1011 patients. Data have been presented (86,87) showing that chemoradiation is safe, and does not alter the feasibility of subsequent surgery. The EORTC trial should help to clarify the role of combined modality treatment compared to radiotherapy alone in the preoperative setting and the role of adjuvant chemotherapy in the postoperative setting. However, this trial was designed in 1991 and the endpoints are overall and disease-free survival. So this trial may not provide accurate information on the R0 resection rate after radiotherapy or CRT because of the lack of quality assurance of pathological reporting and the fact that TME surgery was not mandated. A very similar French trial (FFCD 9203) randomized a similar population of 762 patients over the same time period between preoperative radiotherapy to a dose of 45 Gy in 25 fractions over five weeks and an identical schedule of 5-FU modulated by leucovorin chemoradiotherapy. In contrast to the EORTC study, all patients were mandated to receive postoperative chemotherapy. The preliminary results showed that patients achieved a higher rate of pathological complete response (PCR) (88), but the addition of chemotherapy is at the expense of more acute toxicity. Late

208

Glynne-Jones and Hughes

outcome in terms of local recurrence, and disease-free and overall survival is awaited. Preoperative Chemoradiation vs. Preoperative Short Course Radiation A Polish study (33) randomized 316 patients between preoperative long fractionation chemoradiation and the short five-fraction schedule successfully employed in the Swedish Rectal Cancer trial (17) and the Dutch TME study (18). The study aimed to verify if conventional preoperative chemoradiation with surgery after an interval of four to six weeks to allow response offers an advantage in sphincter preservation in comparison to short fractionation radiation and immediate surgery. However, there was no significant difference. Although sphincter sparing is the main endpoint, this trial is important in that it is the first time that a long fractionation chemoradiation regimen has been directly compared against five fractions followed by immediate surgery. Crucially, chemoradiation reduced the positive CRM rate in this group of resectable patients, and increased the chance of an R0 resection from 87% after SCPRT to 96% (33). A trial in progress from the Trans Tasman Oncology Group randomizes patients with T3 clinically resectable cancers between preoperative short fractionation radiotherapy (25 Gy in five fractions over five days) and long fractionation chemoradiation (50.4 Gy in 28 fractions with continuous infusion 5-FU). Where it is considered uncertain whether preoperative radiotherapy is indicated, the patients are randomized to initial surgery, short course radiotherapy, or long course chemoradiation. The primary endpoint for these studies is local recurrence. Published clinical trials are mainly in the phase II setting (89,90), with little long-term outcome in terms of local recurrence, survival, or late function. There are no meta-analyses examining preoperative chemoradiation in rectal cancer or Cochrane reviews. There is no clearly defined standard neoadjuvant chemoradiation regimen in the treatment of locally advanced rectal cancer. So, many oncologists and surgeons remain very skeptical as to the benefit of fluoropyrimidine-based CRT, and feel that the enthusiasm for this approach is premature. FIXED/UNRESECTABLE RECTAL CANCER There have been no validated and widely accepted methods of clinically defining locally advanced rectal cancer or unresectability prior to surgery. Historically, patients with clinically staged T3 or T4 disease or where lymph nodes appear to be involved on clinical staging have been considered suitable for preoperative radiotherapy or chemoradiation (CRT). Selection has been based on the information obtained from digital rectal examination 3(tumors that are partially or totally fixed), transrectal ultrasound, and CT.

Radiotherapy in the Treatment of Rectal Cancer

209

There are no meta-analyses that have examined the role of radiotherapy or chemoradiation for initially unresectable cancers. Historical randomized studies of radiotherapy versus chemoradiotherapy in unresectable rectal cancer (91–93) were defined by digital rectal examination. The trials used poor methodology, are poorly controlled, and scantily reported with little detail and small numbers when compared to trials in the modern era. Since that time, only a single recent phase III study has been published looking at the role of CRT in unresectable disease (94). A total of only 70 patients with fixed inoperable rectal cancer were randomized. The trial compared pelvic radiotherapy alone (46 Gy in 23 fractions over five weeks) and a highly unusual regimen of chemoradiation with the primary endpoint of resectability (R0 resection). The chemoradiation delivered an unconventional alternating hyperfractionated split-course regimen of 10 Gy in five fractions over three consecutive days delivered alternate weeks to a total dose of 40 Gy over eight weeks in combination with synchronous methotrexate, 5-FU, and folinic acid. The trial demonstrated an advantage in terms of resectability and local control to the chemoradiation arm. Local recurrencefree survival at five years was 35% versus 66% (p ¼ 0.03), and five-year survival was 18% versus 29% (nonsignificant) for radiotherapy versus chemoradiation, respectively. The study is flawed, as with many chemoradiation studies in this era, by the use of a split-course schedule in the combined arm. However, the results do lend some support to the view that chemoradiation is more effective than radiotherapy alone. INTRAOPERATIVE RADIOTHERAPY Intraoperative radiotherapy has been enthusiastically endorsed by some selected centers for the treatment of locally inoperable tumors in addition to external beam radiotherapy and surgery. The ability to deliver this technique is limited, and currently there are no centers within the United Kingdom practicing this technique. The American series (95,96) used an intraoperative boost of 10–20 Gy following the external beam radiotherapy. More recently, intraoperative radiotherapy has been added to preoperative chemoradiation (97–99). The results of these studies have not been validated in randomized controlled trials, but show that due to the lack of resources they are not easily applied to U.K. practice. EARLY RECTAL CANCER Some groups have explored alternative more conservative surgical and nonsurgical approaches to avoid either an anterior resection in frail or elderly patients or an abdominoperineal excision of the rectum for distal rectal cancer. The rectum itself is sensitive to the effects of ionizing radiation, and is prone to early and late radiation effects. This sensitivity has limited

210

Glynne-Jones and Hughes

the curative use of radiotherapy as a sole modality of treatment to specialists with considerable expertise. However, techniques such as high- and lowdose rate endoluminal brachytherapy and wide local excision with adjuvant radiotherapy are becoming increasingly considered both in the palliative setting and as part of definitive management (100–103). The results of local treatment strategies need to be considered against the use of total mesorectal excision in the context of a randomized controlled trial. The local recurrence rate for TME in stage I (T1-2N0) disease in the Dutch TME study was 0.7% at two years with an associated 2.6% risk of postoperative mortality. However, it should be noted that the operative mortality in the 70 to 80-year-old age group increased to 6.4% (18). Wide Local Excision  Radiotherapy Wide local excision followed by selective postoperative radiotherapy or chemoradiation is used in some centers for the treatment of T1 and T2 tumors. There is a consensus, based on retrospective studies, that tumors <3 cm in size, which are limited to superficial layers of the muscularis propria and are well or moderately well differentiated, are appropriate for local excision alone. Tumors with these characteristics are associated with a low risk of microscopic lymph node involvement and a subsequent local recurrence rate of <10%, provided adequate surgical margins can be achieved. An approach taken by some investigators has been to follow local excision with postoperative adjuvant chemoradiation when the above selection criteria have not been met. One of the first to document this technique (104) reported a median follow-up of 27 months and disease-free survival of 88%. Unfortunately, there have been no trials that randomize a local treatment—e.g., local excision and adjuvant chemoradiotherapy versus standard AP excision of the rectum. The design of such a study is unlikely to be feasible or acceptable to patients. A number of single- and multicenter phase II studies have evaluated the role of postoperative radiation including the RTOG 8902 (105). The Cancer and Leukaemia Group B (CALGB) also applied these criteria for local excision in a prospective multicenter trial—CALGB 8984 (106). The standardized criteria for local excision in this trial were <4 cm in diameter occupying <40% of the bowel’s circumference and arising within 10 cm of the dentate line. The study registered 180 patients, and 110 patients with pathologically defined T1 or T2 lesions and negative margins were observed following local excision. Patients with T2 lesions were treated with postoperative adjuvant pelvic CRT with 5-FU to a dose of 54 Gy. For T1 lesions there were only three local failures (5%), one of which was successfully salvaged by AP excision of the rectum. However, in those patients with T2 lesions, there were seven local failures (14%) of which only four were successfully salvaged.

Radiotherapy in the Treatment of Rectal Cancer

211

Kim et al. (107) has suggested an alternative strategy. The authors raise the possibility that preoperative radiotherapy followed by local excision may provide an effective alternative to radical surgery for T2 and T3 tumors where there has been a PCR. Long-term survival and control after preoperative chemoradiation has also been demonstrated in locally advanced T3 rectal tumors in patients who either refused or had medical comorbidity, which prevented radical surgical resection (108). The introduction of these techniques may allow less surgically morbid approaches for the treatment of early rectal tumors. Endocavity Radiation Locally applied endocavity radiotherapy is a well-accepted method of radical treatment for clinically staged T1 and T2 tumors in the mid and distal rectum (101). Small histologically favorable lesions can be treated by high doses of low-energy radiotherapy (30 Gy), which are repeated over four weeks to a total dose of 120 Gy. The rectal mucosa will tolerate such very high doses to a small area. Advocates of this procedure treat until response and accept that lack of response is an indication for surgical excision. This technique is not usually considered appropriate if the lesions extend down into the anal canal, as there is a much higher risk of ulceration, which often fails to heal at this site. This approach requires rigorous patient selection, experience, and operator skill in the delivery of superficial radiation directly to the tumor on repeated occasions. The results of the Lyon experience (109) for 101 T1 tumors demonstrate a local control rate of 90%. The drawback to all radiotherapy series is that they are clinically defined and lack pathological assessment of the extent of invasion. The results are therefore difficult to stratify into T and N stage, since ultrasound may not be sufficiently sophisticated in the preoperative setting. Surgical series (110) suggest that the likelihood of having microscopic perirectal or mesorectal nodal involvement would be very rare in T1 tumors, but for T2 tumors this would be 10% to 30% and as high as 60% risk in T3 tumors. Local excision with or without adjuvant treatment should therefore be considered a controversial treatment for all except favorable T1 tumors. PRETREATMENT CLINICAL ASSESSMENT Standard clinical assessment has relied on digital rectal examination. Examination under anesthetic allows more detailed palpation of the anorectal and pelvic structures. Locoregional staging investigations should include whole body CT and MRI of the pelvis. The pelvic lymph nodes are frequently involved particularly with increasing tumor stage and in poorly differentiated tumors. Bulky, circumferential, tethered tumors at the level of the prostate or seminal vesicles, and very low rectal cancers are considered

212

Glynne-Jones and Hughes

appropriate on clinical grounds for preoperative treatment to shrink and downstage the tumor. The increasing use of preoperative pelvic MRI in predicting the likelihood of the surgeon failing to achieve an R0 resection (10,11) has resulted in improved selection, and MRI is now used widely in the United Kingdom as a staging preoperative tool upon which to base clinical decisions. The recent U.K.-based MERCURY Study (13) has evaluated the accuracy of MRI in defining the extent of tumor spread into the mesorectum and nodal involvement. Current criteria in our own units for the delivery of preoperative chemoradiation with a view to downstaging rely on the MRI in determining the extent of tumor either outside the mesorectal envelope or very close to the mesorectal envelope (i.e., within 1 mm) or below the mesorectal envelope extending into the lower rectum and the sphincter. Patients defined in such a way should have a high risk of not achieving an R0 resection. RADIOTHERAPY PLANNING TECHNIQUES The radiation fields employed by most radiation oncologists are still based to a great extent on patterns of relapse identified by Gunderson and Sosin (1) and have been derived from information obtained at laparotomy from second look operations. The risk of involvement of regional lymph nodes depends on whether the primary tumor lies in the upper middle or lower third of the rectum. Accurate target definition, and obtaining the best functional outcome from the combination of surgery and radiotherapy, requires a good collaboration between surgeons and radiation oncologists. Target Definition The radiotherapy plan aims to encompass the mesorectal envelope and all gross sites of disease with a margin. Therefore, treatment is usually delivered with a three or four field box technique using megavoltage radiotherapy encompassing the tumor plus a margin of 3 cm superiorly, inferiorly, and posteriorly with the anterior margin restricted to 2 cm. The conventional reference points in treatment of rectal cancer are bony landmarks within the pelvis (particularly the sacrum), and also the bladder and the external anal margin, which can be marked during simulation or CT planning with wire or ball bearings. Alignment tattoos are used as external reference points. Conventional Planning Patients are planned prone with a full bladder to push small bowel superiorly out of the high-dose volume. The prone position also allows placement of an anal marker, and assessment of the distance from the anal margin to the distal end of the rectal tumor by digital examination. Many units employ bowel displacement techniques such as the bellyboard (the use of a table cut

Radiotherapy in the Treatment of Rectal Cancer

213

out to allow the upper abdominal small bowel contents to fall forwards and upwards and further reduce the volume of small bowel irradiated) (104,111). Using this technique the target volume is defined clinically with reference to the pelvic bony anatomy. Computed Tomography Scanning Procedure The majority of patients undergoing radical treatment at our center are planned using a CT planning technique. Three-dimensional treatment planning (112) allows better visualization of the tumor on nonaxial planes with a beam configuration, and will allow a dose-volume histogram to derive maximal doses for both tumor and normal tissue. Similarly to conventional planning, patients are scanned in the prone position with instructions to maintain a full bladder to push the small bowel cranially out of the field as far as possible. A CT planning scan with 10 mm-thick slices is performed. A radio-opaque marker is useful to demonstrate the position of the anal verge. Patients are scanned from the superior aspect of L5, to 2 cm beyond the anal marker in order to cover the whole of the pelvis, rectosigmoid, and rectum. The anterior field borders can be defined in female patients with the use of a radio-opaque tampon placed in the vagina. The gross tumor volume can be defined with the help of CT scan, endoscopy, barium enema, positron emission tomography scan, and endoscopic ultrasound. The MRI scan is the most helpful and most accurate modality in defining both local tumor extent and macroscopically involved lymph nodes (12). Margins of 3 cm are added to allow for microscopic disease spread and patient movement. We accept a more modest margin superiorly if necessary in order to try and maintain the superior border and the extent of the field within the L5/S1 junction. The radiotherapy dose is prescribed to the intersection point ensuring that the 95% isodose covers the target volume. All fields are treated daily, five times a week. Simulation To preserve the function of the anal sphincter, we avoid including the anal canal if possible within the radiation field. Since it is difficult to image the sphincter mechanism with CT planning, a Foley catheter is introduced into the anus on the simulator couch, and the balloon filled with contrast. The catheter is gently withdrawn until resistance is encountered, indicating the balloon is pressing against the superior border of the upper anal canal (photo 1). For obvious reasons, this technique is not possible for low rectal tumors, as it risks shielding disease. Small Bowel Contrast Opacification of the small bowel is helpful; in this way customized shielding may be used to protect the small bowel. A number of approaches may be

214

Glynne-Jones and Hughes

used including 300 mL Baritop with 20 mL Gastrografin or dilute Gastrografin given orally 45 to 60 minutes prior to simulation of planned fields. Small bowel fixed in the pelvis is of considerable concern in patients who have undergone previous pelvic resectional surgery, or who are due to receive postoperative radiation. In these patients, small bowel imaging may allow a reduction in the volume of small bowel irradiated. Normal Tissue Tolerances The safe delivery of therapeutic doses of radiation is limited by the tolerance of the surrounding normal tissues. Decades of experience have defined what radiotherapy doses each particular structure will tolerate and the expression of acceptable late toxicity. With the advent of three-dimensional computerized planning, the concept of tolerance of an organ is in the process of being replaced by dose-volume histograms, which focus on the total dose delivered to the percentage of any specific organ. In the treatment of rectal cancer, the main organs of concern are the small bowel, bladder, ureters, and the femoral heads. All quoted normal tissue tolerances assume a daily dose of 1.8 Gy per fraction. However, it is wise to maintain the maximum dose per fraction to the small bowel, ureter, and bladder to 2 Gy or less, avoiding gross inhomogeneities of dose. ACUTE TOXICITY AND SUPPORTIVE CARE DURING RADIOTHERAPY Expected acute side effects of pelvic radiotherapy include diarrhea, proctitis, urinary frequency and dysuria, erythema, and moist desquamation of the perineum, particularly in the postoperative setting if the perineal scar is encompassed. With long fractionation radiotherapy and CRT, toxicities tend to develop after two to three weeks. Patients should be assessed at least once weekly with regard to toxicity and their overall tolerance to treatment. Acute toxicities during SCPRT are generally mild. In the Dutch study, side effects were predominantly mild to moderate (grade 1–2) gastrointestinal symptoms. However, the development of lumbosacral plexopathy, causing long-term lumbar and gluteal pain, in six patients after SCPRT in the Swedish study (113) had caused concern. In the Dutch study, transient neurological symptoms were documented in 53 patients, the majority of whom experienced only grade 1 toxicity (35,80). No patients have developed a permanent lumbosacral plexopathy—possibly due to use of better shielding of the sacrum posteriorly. In addition, the superior border of the radiotherapy field used in the Dutch study (L5-S1) was lower than that defined in the Swedish trials (mid L4). This toxicity is not reported in long fractionation schedules.

Radiotherapy in the Treatment of Rectal Cancer

215

TIMING OF SURGERY FOLLOWING PREOPERATIVE RADIOTHERAPY After Short Course Most studies with short fractionation preoperative radiotherapy (25 Gy in five fractions over five days) recommend a short interval (less than seven days) between the completion of radiotherapy and surgery. The aim of this strategy is to perform surgery before the side effects of radiotherapy have been expressed on the normal pelvic tissues. Early reports of SCPRT followed by surgery within a few days suggested significant downstaging at surgery (114). Downstaging was not observed in the Dutch TME trial (78). A decrease in tumor size and the number of recovered lymph nodes was observed, but there was no change in tumor or lymph node classification. This short gap therefore allows the original histology to be examined without added risks of morbidity or mortality. A current Swedish study is comparing SCPRT with a short or long (four to six weeks) interval to surgery and standard fractionation to 50 Gy with a similar long interval. In view of data on risks of mortality, the recommended maximum interval between SCPRT and surgery is five days. After Long Course The optimal timing of surgery following long course preoperative radiotherapy or chemoradiation remains undefined. Most clinicians accept that in bulky tumors, where significant regression is required to facilitate an R0 resection, long fractionation radiotherapy or chemoradiation is advisable and a minimum gap of four to six weeks is necessary. It would also appear obvious that sphincter preservation is likely to be achieved more frequently with preoperative radiotherapy after a delay to allow tumor shrinkage. Most surgeons prefer to operate when the initial inflammatory reaction has settled and perioperative bleeding is not a problem. This gap of four to six weeks has been extrapolated to rectal cancer from data on schedules of preoperative radiotherapy in bladder cancer prior to cystectomy. These historical studies defined the optimal interval in terms of minimizing toxicity. A European trial in resectable rectal cancer delivered preoperative radiotherapy to a total dose of 39 Gy in 13 fractions. The trial randomized patients between early surgical resection within two weeks of completing radiotherapy or delayed resection within six to eight weeks (115). Significantly more downstaging was observed in the long-interval group compared to those with an interval of only two weeks (6% vs. 10% pathological complete response rate, respectively). Sphincter preservation was achieved in 79% of patients with a long interval (four to six weeks) following radiotherapy, compared with 69% of patients where the interval was only short (two weeks). Surgical morbidity and other complications such as anastomotic

216

Glynne-Jones and Hughes

leaks were not significantly different between the two groups. Interestingly, local control was the same for both groups. Recent retrospective data from Memorial Sloan-Kettering suggest that increasing the interval between chemoradiation and resection will enhance the rate of pathological complete responses in rectal cancer (116). In contrast, a phase II study (117) failed to show that a longer time interval of 10 to 14 weeks versus 4 to 8 weeks resulted in any improvement in downstaging. SURGICAL COMPLICATIONS A concern of the Dutch TME study was that the rate of surgical complications would become unacceptable when combined with SCPRT. The rate of major complications requiring further intervention or surgery was similar whether preoperative radiotherapy was delivered or not (48% for the radiotherapy arm and 41% for the surgery alone arm). This difference was mainly attributable to slow perineal wound healing after radiotherapy (29% vs. 18%). Infections relating to wound healing in the sacral cavity and in the perineal areas following AP excision of the rectum remain one of the most frequent postoperative complications. Delayed perineal postoperative wound healing is reported to range between 7% and 63% (118). A recent study of preoperative five fractions of radiotherapy in rectal cancer showed that perineal wound healing was delayed in four out of 20 patients undergoing APER (20%) (119). Reports from early trials of preoperative radiotherapy do not suggest that radiotherapy enhances the risk of an anastomotic leak (MRC 2, Stockholm Study) (68,120), although a significantly higher leak rate was observed after preoperative radiotherapy in the Swedish Rectal Cancer trial. In the Dutch TME study, the clinical anastomotic leakage was 11% overall. There was no statistical difference whether the patient had received preoperative radiotherapy (11%) or surgery alone (12%). Similarly in the German AIO study (84), there was no difference between pre- and postoperative radiotherapy in terms of leak rate. No difference was observed in postoperative mortality, defined as any death within 30 days of surgery, in the Dutch study (4.0% vs. 3.3%). The postoperative mortality rate at 180 days was increased in the Dutch study (in a nonrandomized subanalysis), if there had been a delay longer than three days between the end of SCRT and surgery (4.1% for less than three days vs. 8.4% for greater than three days). This effect was most noticeable in patients over 75. LATE EFFECTS Common forms of late morbidity involve damage to the small bowel, bladder, rectum, soft tissues, and neuropathic problems. Small bowel tolerance has normally been considered the main dose-limiting factor when high doses of external beam radiotherapy are delivered to the pelvis. Late complications affecting the small bowel are estimated to occur with an

Radiotherapy in the Treatment of Rectal Cancer

217

approximate incidence of 5% following postoperative radiotherapy to doses in the region of 45–50 Gy in rectal cancer (121). The predicted late complications of pelvic radiotherapy depend on the volume, the radiation field, the overall treatment time, the fraction size, total dose, and technique, but the most crucial factor is the volume of the small bowel in the radiation field (122,123). As follow-up in the majority of studies is generally short, there is likely to be a major underestimate of the risks of late small bowel damage. There are also little data on long-term rectal and bladder morbidity following preoperative radiotherapy or chemoradiation for rectal cancer. Evidence suggests that 5% to 10% of patients will experience grade 3 or 4 late morbidity (112,124). Little is known about the late radiation effects on anal sphincter function, compliance of the rectal reservoir, fashioned colonic pouches, and other soft tissues within the pelvis. It is not clear how much rectal dysfunction is caused by fibrosis of the anal sphincter, loss of capacitance of the neorectum, and/or neuropathic damage from combination of surgery and radiation. Data on colonic J-pouches treated with preoperative chemoradiation for rectal cancer suggest that function is equally good as for patients who have undergone surgery alone (125). However, in the Swedish Rectal Cancer trial, irradiation of the anal canal was shown to be associated with much worse function, and patients were much more likely to wear an incontinence pad and to be incontinent to gas, liquid, and solid stool (124). A study from Poland (119) also confirmed late complications in 10% of patients following SCPRT. In the Polish study, 40% of the patients described a significant impairment of their social life because of anal rectal dysfunction and 18% described their quality of life as being severely affected. Tumors in the lower third of the rectum will inevitably have the anal sphincter irradiated, but almost all these patients will proceed to an AP excision of the rectum. So functional outcome is not an issue. The late effects of postoperative radiotherapy or chemoradiation appear to be worse than for preoperative chemoradiation (126–130). A 10% absolute increase of the risk of impotence following preoperative radiotherapy has also been demonstrated (78). Prospectively collected quality of life differences between the surgery and radiotherapy arms seem to evaporate within 12 months of completing treatment, and may simply reflect the increased use of a temporary ileostomy in the radiotherapy arm. CONCLUSION The Uppsala trial, The NSABP R-03 trial, and the German trial suggest that preoperative radiotherapy is more effective with less toxicity than postoperative (48,85). The development of biological tumor markers such as thymidilate synthetase to predict response to chemotherapy or radiotherapy treatment and future outcome have been followed with great enthusiasm, yet

218

Glynne-Jones and Hughes

have to date failed to offer a consistent message. Current technological advances mean that it is now possible to study the expression of literally thousands of genes using cDNA array and nucleotide polymorphisms. These techniques are enormously complex, but offer the opportunity to identify relevant molecular abnormalities and essential genetic changes. However, this approach will need to be applied systematically throughout all trials of rectal cancer in order to get the maximum information for the future. The most realistic future hope is of increasingly accurate methods of selecting patients—potentially by means of MRI staging—who would benefit from preoperative treatment, and sparing those for whom radiotherapy may be unnecessary. The R0 and CRM (1 mm) endpoints, with the emphasis on the circumferential margin, should therefore be mandated in all future studies in rectal cancer. Carefully controlled studies with large numbers of patients and much longer follow-up than is presently undertaken are required to assess accurately the long-term functional results of neoadjuvant preoperative chemoradiation. The role of the new agents, which act on targets predicted by our understanding of the molecular processes underlying the progression and development of malignancy, e.g., EGFR inhibitors, VEGF inhibitors, Cox-2 inhibitors, farnesyl transferase inhibitors, etc., is expanding. They provide a potential alternative strategy for inhibiting and containing the repopulation of tumor cells during radiotherapy/chemoradiotherapy (131). Future studies will need to focus on patients with cancers 0–10 cm from the anal margin, particularly those in the lower rectum, in whom an MRI predicts a high risk of a histopathological R1 resection. Points  Preoperative radiotherapy and/or chemoradiation are more effective and less morbid than postoperative radiotherapy.  It is vital that the radial or circumferential margin should be uninvolved if the patient is to have the best chance of cure.  MRI can define an unsafe circumferential margin in the mid and upper rectum, and hence clarify when preoperative treatment is indicated.  The majority of patients with a cancer in the low rectum where an abdominoperineal excision is envisaged should be considered for preoperative treatment.  Postoperative CRT is indicated when the above policy has failed!  Trials that do not use MRI to select the patients, TME to treat the patients, and ‘‘Quirke’’ style pathology of the surgical specimen will not tell us what we currently need to know.  The small minority of patients with early cancers suitable for local excision requires very careful selection.

Radiotherapy in the Treatment of Rectal Cancer

219

REFERENCES 1. Gunderson LL, Sosin H. Areas of failure found at reoperation (second or symptomatic look) following ‘‘curative surgery’’ for adenocarcinoma of the rectum: clinico pathologic correlation and implication for adjuvant therapy. Cancer 1974; 34:1278–1286. 2. McCall JL, Cox MR, Wattchow DA. Analysis of local recurrence rates after surgery alone for rectal cancer. Int J Colorect Dis 1995; 10(3):126–132. 3. Quirke P, Durdey P, Dixon MF, Williams NS. Local recurrence of rectal adenocarcinoma due to inadequate surgical resection. Histopathological study of lateral tumour spread and surgical excision. Lancet 1986(2):996–999. 4. Heald RJ, Moran BJ, Ryall RD, Sexton R, MacFarlane JK. The Basingstoke experience of total mesorectal excision 1978–1997. Arch Surg 1998; 133:894–899. 5. Enker WE, Thaler HT, Cranor ML, Polyak T. Total mesorectal excision in the operative treatment of carcinoma of the rectum. J Am Coll Surg 1995; 181: 335–346. 6. Cecil DT, Sexton R, Moran BJ, Heald RJ. Total mesorectal excision results in low local recurrence rates in lymph node-positive rectal cancer. Dis Colon Rectum 2004; 47:1145–1150. 7. Camilleri-Brennan J, Steele RJ. The impact of recurrent rectal cancer on quality of life. Eur J Surg Oncol 2001; 27:349. 8. Birbeck KF, Macklin CP, Tiffin NJ, et al. Rates of circumferential resection margin involvement vary between surgeons and predict outcome in rectal cancer surgery. Ann Surg 2002; 235(4):449–457. 9. Kockerling F, Reymond MA, Altendorf-Hoffman A, Dworak O, Hohenberger W. Influence of surgery on metachronous distant metastases and survival in rectal cancer. J Clin Oncol 1998; 16:324–329. 10. Brown G, Radcliffe A. Will MRI provide maps of lines of excision for rectal cancer. Lancet 2001; 357:495–496. 11. 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. 12. Brown G, Radcliffe AG, Newcombe RG, Dallimore NS, Bourne MW, Williams GT. Preoperative assessment of prognostic factors in rectal cancer using high-resolution magnetic resonance imaging. Br J Surg 2003; 90(3):355–364. 13. Daniels I. MRI predicts surgical resection margin status in patients with rectal cancer: results from the Mercury Study Group. Clin Oncol 2004; 16:S45–S46. 14. Gastrointestinal Tumor Study Group. Prolongation of the disease-free survival in surgically treated rectal carcinoma. N Engl J Med 1985; 312:1465–1472. 15. Krook JE, Moertel CG, Gunderson LL, et al. Effective surgical adjuvant therapy for high-risk rectal carcinoma. N Engl J Med 1991; 324:709–715. 16. Conference de consensus. Paris, France: Agence Nationale pour le Developpement de l’Evaluation Medicale, 1994. 17. Swedish Rectal Cancer trial. Improved survival with preoperative radiotherapy in resectable rectal cancer. N Engl J Med 1997; 336:980–987. 18. Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. Dutch Colorectal Cancer Group. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001; 345:690.

220

Glynne-Jones and Hughes

19. Mohiuddin M, Marks G. High dose preoperative radiation and sphincter preservation in the treatment of rectal cancer. Int J Radiat Oncol Phys 1987; 13:839–844. 20. Minsky BD. Pelvic radiation therapy in rectal cancer—technical considerations. Semin Radiat Oncol 1993; 3:42–47. 21. Tepper JE, O’Connell M, Niedzwiecki D, et al. Adjuvant therapy in rectal cancer: analysis of stage, sex, and local control—final report of Intergroup 0114. J Clin Oncol 2002; 20:1744–1750. 22. Myerson RJ, Singh A, Birnbaum EH, et al. Pretreatment clinical findings predict outcome for patients receiving preoperative radiation for rectal cancer. Int J Radiat Oncol Biol Phys 2001; 50(3):665–674. 23. Shepherd NA, Baxter KJ, Love SB. Influence of local peritoneal involvement on pelvic recurrence and prognosis in rectal cancer. J Clin Pathol 1995; 48: 849–855. 24. Heald RJ, Smedh RK, Kald A, Sexton R, Moran BJ. Abdominoperineal excision of the rectum—an endangered operation. Dis Colon rectum 1997; 40:747–751. 25. Heide J, Krull A, Berger J. Extracapsular spread of nodal metastasis as a prognostic factor in rectal cancer. Int J Radiat Oncol Biol Phys 2004; 58:773–778. 26. Merkel S, Mansmann U, Siassi M, Papadopoulos T, Hohenberger W, Hermanek P. The prognostic inhomogeneity in pT3 rectal carcinomas. Int J Colorect Dis 2001; 16:298–304. 27. Picon AI, Moore HG, Sternberg SS, et al. Prognostic significance of depth of gross or microscopic perirectal fat invasion in T3 N0 Mo rectal cancers following sharp mesorectal excision and no adjuvant therapy. Int J Colorect Dis 2003; 18:487–492. 28. Tagahashi T, Weno M, Azekura K. Lateral node dissection and total mesorectal excision for rectal cancer. Dis Colon Rectum 2000; 43(suppl):S59–S68. 29. Moriya Y, Sugihara K, Akasut A, Fujita S. Importance of extended lymph adenonectomy with lateral node dissection for advanced lower rectal cancer. World J Surg 1997; 21:728–732. 30. Adam IJ, Mohamdee MO, Martin IG, et al. Role of circumferential resection margin involvement in the local recurrence of rectal cancer. Lancet 1994; 344:707–711. 31. Guillou P, Quirke P, Bosanquet N, et al. The MRC Classic trial: results of short-term endpoints. Br J Cancer 2003; 88(suppl 1):S21 (abstract 6.5). 32. Marijnen CAM, Nagtegaal ID, Kapiteijn E, et al. Cooperative investigators of the Dutch Colorectal Cancer Group. Radiotherapy does not compensate for positive resection margins in rectal cancer patients: report of a multicentre randomised trial. Int J Radiat Oncol Biol Phys 2003; 55:1311–1320. 33. 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. 34. Ng IO, Luk IS, Yuen ST, et al. Surgical lateral clearance in resected rectal carcinomas. A multivariate analysis of clinicopathologic features. Cancer 1993; 71:1972–1976.

Radiotherapy in the Treatment of Rectal Cancer

221

35. De Haas-Kock DF, Baeten CG, Jager JJ, et al. Prognostic significance of radial margins of clearance in rectal cancer. Br J Surg 1996; 83:781–785. 36. Hall NR, Finan PJ, al-Jaberi T, et al. Circumferential margin involvement after mesorectal excision of rectal cancer with curative intent. Predictor of survival but not local recurrence? Dis Colon Rectum 1998; 41:979–983. 37. 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; 89:1067. 38. Nagtegaal ID, Marijnen CA, Kranenbarg EK. Circumferential margin involvement is still an important predictor of local recurrence in rectal cancer: not 1 millimetre but 2 millimetres is the limit. Am J Surg Pathol 2002; 26:350–357. 39. McArdle CS, Hole D. Impact of the variability among surgeons on postoperative morbidity and mortality and ultimate survival. Br Med J 1991; 302: 1501–1505. 40. Hermanek PJR, Mansmann U, Staimmer DS, Riedl S, Hermanek P. The German experience: the surgeon as a prognostic factor in colon and rectal cancer surgery. Surg Oncol Clin North Am 2000; 9:33–48. 41. Eriksen MT, Wibe A, Syse A, Haffner J, Wiig JN on behalf of the Norwegian Rectal Cancer Group. Inadvertent perforation during rectal cancer resection in Norway. Br J Surg 2004; 91:210–216. 42. Nagtegaal ID, Van der Velder CJH, Van der Worp E, Kapiteijn E, Quirke P, Van Krieken JH. Macroscopic evaluation of rectal cancer resection specimen: clinical significance of the pathologies in quality control. J Clin Oncol 2002; 20:1729–1734. 43. Balslev I, Pedersen M, Teglbjaerg PS, et al. Postoperative radiotherapy in Dukes’ B and C carcinoma of the rectum and rectosigmoid. A randomized multicenter study. Cancer 1986; 58:22–28. 44. Medical Research Council Rectal Cancer Working Party. Randomised trial of surgery alone versus radiotherapy followed by surgery for potentially operable locally advanced rectal cancer. Lancet 1996; 348:1605–1610. 45. Douglass HO Jr., Moertel CG, Mayer RJ. Survival after postoperative combination treatment of rectal cancer. N Engl J Med 1986; 315:1294–1299. 46. Arnaud JP, Nordlinger B, Bosset JF, et al. Radical surgery and postoperative radiotherapy as combined treatment in rectal cancer. Final results of a phase III study of the European Organization for Research and Treatment of Cancer. Br J Surg 1997; 84:352. 47. Treurniet-Donker AD, van Putten WL, Wereldsma JC, et al. Postoperative radiation therapy for rectal cancer, in interim analysis of a prospective randomised multi centre trial in Netherlands. Cancer 1991; 67:2042–2048. 48. Fisher B, Wolmark N, Rockette H, et al. Postoperative adjuvant chemotherapy or radiation therapy for rectal cancer: results from NSABP protocol R01. J Natl Cancer Inst 1988; 80:21–29. 49. Wolmark N, Wieand HS, Hyams DM, et al. Randomized trial of postoperative adjuvant chemotherapy with or without radiotherapy for carcinoma of the rectum: National Surgical Adjuvant Breast and Bowel Project Protocol R-02. J Natl Cancer Inst 2000; 92:361–362.

222

Glynne-Jones and Hughes

50. Marneghan H, Gray B, de Zwart J, et al. Adjuvant postoperative radiotherapy in rectal cancer: results from the ANZ bowel cancer trial (protocol 8202). Australas Radiol 1991; 35:50–61. 51. Mansour EG, Lefkopoulou M, Johnson R, Douglass H. A comparison of postoperative adjuvant chemotherapy, radiotherapy or combination therapy in potentially curable resectable rectal carcinoma. An ECOG study Est 4276. Proc Am Soc Clin Oncol 1991; 10:154 abstract. 52. Tveit KM, Guldvog I, Hagen S, et al. Randomized controlled trial of postoperative radiotherapy and short-term time-scheduled 5-fluorouracil against surgery alone in the treatment of Dukes B and C rectal cancer. Norwegian Adjuvant Rectal Cancer Project Group. Br J Surg 1997; 84:1130–1135. 53. Cafiero F, Gipponi M, Peressini A, Bertoglio S, Lionetto R. Preliminary analysis of a randomized clinical trial of adjuvant postoperative RT vs. postoperative RT plus 5-FU and levamisole in patients with TNM Stage II-III resectable rectal cancer. J Surg Oncol 2000; 75:80–88. 54. Smalley SR, Benedetti J, Williamson S, et al. Intergroup 0144-phase III trial of 5-FU based chemotherapy regimens plus radiotherapy (XRT) in postoperative adjuvant rectal cancer. Bolus 5-FU vs prolonged venous infusion (PVI) before and after XRT þ PVI vs bolus 5-FU þ leucovorin (LV) þ levamisole (LEV) before and after XRT þ bolus 5-FU þ LV. J Clin Oncol 2003; 22:251 (abstract 1006). 55. Smalley SR, Benedetti J, Williamson S, et al. Intergroup 0144 A phase III rectal surgical adjuvant study of pelvic radiation (XRT) plus 5-FU based chemotherapy (bolus 5-FU before and after PVI þ XRT vs PVI before during and after XRT vs biochemically modulated bolus 5-FU and XRT): mature outcome results and pelvic failure analysis. Int J Radiat Oncol Biol Phys 2004; 60(suppl):S137 (abstract 14). 56. O’Connell MJ, Martenson JA, Wieand HS, et al. Improving adjuvant therapy of rectal cancer by combining protracted infusional fluorouracil with radiation therapy after curative surgery. N Engl J Med 1994; 331:502–507. 57. Lee JH, Lee JH, Ahn JH, et al. Randomized trial of postoperative adjuvant therapy in stage II and III rectal cancer to define the optimal sequence of chemotherapy and radiotherapy: a preliminary report. J Clin Oncol 2002; 20:1751–1758. 58. Roswit B, Higgins G, Keehn R. Preoperative irradiation for carcinoma of the rectum and rectosigmoid colon: report of a National Veterans Administration randomised study. Cancer 1975; 35:1597–1602. 59. Kligerman MM. Preoperative radiation therapy in rectal cancer. Cancer 1975; 36:691–695. 60. Rider WD, Palmer JA, Mahoney LJ, Robertson CT. Preoperative irradiation in operable cancer of the rectum: report of the Toronto trial. Can J Surg 1977; 20:335–338. 61. Duncan W, Smith AN, Freedman LS, Alderson MR, Arnott SJ, Bleehan NM. The evaluation of low dose pre-operative x-ray therapy in the management of operative rectal cancer, results of a randomly controlled trial. Br J Surg 1984; 71:21–25. 62. Kutzner J, Bruckner R, Kempf P. Preoperative radiotherapy in rectal cancer. Strahlentherpie 1984; 160:236–238.

Radiotherapy in the Treatment of Rectal Cancer

223

63. Higgins G, Humphrey E, Dwight R, Roswit B, Lee L, Kreehn R. Preoperative radiation and surgery for cancer of the rectum. VASOG trial II. Caner 1986; 58:353–359. 64. Niebel W, Schultz U, Ried M. Five year results of a prospective randomised study: experience with combined radiotherapy and surgery for primary rectal carcinoma. Recent Results Cancer Res 1988; 110:111–113. 65. Gerard A, Buyse M, Nordlinger B, et al. Preoperative radiotherapy as adjuvant treatment in rectal cancer. Ann Surg 1988; 208:606–614. 66. Reis Neto JA, Quilici FA, Reis JA Jr. A comparison of nonoperative vs preoperative radiotherapy in rectal carcinoma. A 10 year randomised trial. Dis Colon Rectum 1989; 32:702–710. 67. Stockholm Rectal Cancer Study Group. Preoperative short-term radiation therapy in operable rectal carcinoma. A prospective randomised trial. Cancer 1990; 66:49–55. 68. Cedermark B, Johansson H, Rutqvist LE, Wilking N. The Stockholm I trial of preoperative short-term radiotherapy in operable rectal carcinoma. A prospective randomized trial. Stockholm Colorectal Cancer Study Group. Cancer 1995; 75:2269–2275. 69. Horn A, Halvorsen JF, Dahl O. Preoperative radiotherapy in operable rectal cancer. Dis Colon Rectum 1990; 33:823–838. 70. Marsh PJ, James RD, Schofield PF. Adjuvant preoperative radiotherapy for locally advanced rectal carcinoma. Results of a prospective randomised trial. Dis Colon Rectum 1994; 37:1205–1214. 71. Goldberg PA, Nicholls RJ, Porter NH, Love S, Grimsey JE. Long term results of a randomised trial of short course low dose adjuvant preoperative radiotherapy for rectal cancer: reduction in local treatment failure. Eur J Cancer 1994; 30A:1602–1606. 72. Sause WT, Pajak TF, Noyes RD, et al. Evaluation of preoperative radiation therapy in operable colorectal cancer. Ann Surg 1994; 220:668–675. 73. Swedish Rectal Cancer trial. Improved survival with preoperative radiotherapy in resectable rectal cancer. Br J Surg 1993; 80:1333–1336. 74. Camma C, Giunta M, Fiorica F, Pagliaro L, Craxi A, Cottone M. Preoperative radiotherapy for resectable rectal cancer: a meta-analysis. JAMA 2000; 284(8): 1008–1015. 75. Colorectal Cancer Collaborative Group. Adjuvant radiotherapy for rectal cancer: a systematic overview of 8507 patients from 22 randomised trials. Lancet 2001; 358(9290):1291–1304. 76. Munro AJ, Bentley AHM. Adjuvant radiotherapy in operable rectal cancer: a systematic review. Semin Colon Rectal Surg 2002; 13:31–42. 77. Wong R, Figueredo A, Tandan V. Preoperative radiotherapy and curative surgery for the management of localized rectal carcinoma (protocol for a Cochrane review). The Cochrane Library, Issue 3. Chichester, U.K.: John Wiley & sons Ltd, 2004. 78. Marijnen CAM, Nagtegaal ID, Kranenbarg EK, et al. for the Pathology Review Committee and the Cooperative Clinical Investigators. No downstaging after short-term preoperative radiotherapy in rectal cancer patients. J Clin Oncol 2001; 19:1976–1984.

224

Glynne-Jones and Hughes

79. Frykholm GJ, Glimelius B, Pahlman L. Preoperative or postoperative irradiation in adenocarcinoma of the rectum: final treatment results of a randomised trial and evaluation of late secondary effects. Dis Colon Rectum 1993; 36: 564–572. 80. Marijnen CAM, Kapiteijn E, van de Velde CJ, et al. Cooperative Investigators of the Dutch Colorectal Cancer Group. Acute side effects and complications after short-term preoperative radiotherapy combined with total mesorectal excision in primary rectal cancer: report of a multicentre randomised trial. J Clin Oncol 2002; 20:1976–1984. 81. Sebag-Montefiore D. An update on the MRC CR07 trial. Br J Cancer 2001; 85(suppl 1):85 (abstract 9.3). 82. Hyams DM, Mamounas EP, Petrelli N, et al. A clinical trial to evaluate the worth of preoperative multimodality therapy in patients with operable carcinoma of the rectum: a progress report of National Surgical Breast and Bowel Project Protocol R-03. Dis Colon Rectum 1997; 40:131–139. 83. Roh MS, Petrelli N, Wieand S, et al. Phase III randomised trial of preoperative versus postoperative multimodality therapy in patients with carcinoma of the rectum (NSABP R-03). J Clin Oncol Proc ASCO 2001; 20:abstract 450. 84. Sauer R, Becker H, Hohenberger W, Rodel C, et al. For the German Rectal Cancer Study Group. Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med 2004; 351:1731–1740. 85. Sauer R, Fietkau R, Wittekind C, et al. Adjuvant vs. neoadjuvant radiochemotherapy for locally advanced rectal cancer: the German trial CAO/ARO/ AIO-94. Colorect Dis 2003; 5(5):406–415. 86. Bosset JF, Calais G, Daban A, et al. Does the addition of chemotherapy to preoperative radiation increase acute toxicity in patients with rectal cancer: report of 22921 EORTC phase III trial. J Clin Oncol 2003; 22:Proc ASCO abstract 1179. 87. Bosset JF, Calais G, Mineur L, et al. Does the addition of chemotherapy to preoperative radiation increase the pathological response in patients with resected rectal cancer: report of 22921 EORTC phase III trial. J Clin Oncol 2004; 22:Proc ASCO abstract 3504. 88. Conroy T, Bonnetain F, Chapet O, et al. Preoperative (preop) radiotherapy (RT) þ 5-FU/folinic acid in T3, T4 rectal cancers: preliminary results of the FFCD 9203 randomized trial. J Clin Oncol 2004; 23:271. 89. Bosset JF, Magnin V, Maignon P, et al. Preoperative radiochemotherapy in rectal cancer: long-term results of a phase II trial. Int J Radiat Oncol Biol Phys 2000; 46:323–327. 90. 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; 24:107–112. 91. Moertel CG, Childs DS Jr., Reitemeier RJ, et al. Combined 5-fluorouracil and supervoltage radiation therapy of locally unresectable gastrointestinal cancer. Lancet 1969; ii:865–867. 92. Rominger CJ, Gunderson LL, Gelber RD, Conner N. Radiation therapy alone or in combination with chemotherapy in the treatment of residual or inoperable carcinoma of the rectum and rectosigmoid or pelvic recurrence following

Radiotherapy in the Treatment of Rectal Cancer

93.

94.

95.

96.

97.

98.

99.

100.

101. 102. 103.

104.

105.

106. 107.

225

colorectal surgery. Radiation Therapy Oncology Group study (76–16). Am J Clin Oncol 1985; 8:118–127. Overgaard M, Bertelsen K, Dalmark M, et al. A randomised feasibility study evaluating the effect of radiotherapy alone or combined with 5-fluorouracil in the treatment of locally recurrent or inoperable colorectal carcinoma. Acta Oncol 1993; 32:547–553. Frykholm GJ, Pahlman L, Glimelius B. Combined chemo- and radiotherapy vs. radiotherapy alone in the treatment of primary, nonresectable adenocarcinoma of the rectum. Int J Radiat Oncol Biol Phys 2001; 50:427–434. Gunderson LL, Martin JK, Beart RW, et al. Intraoperative and external beam irradiation for locally advanced colorectal cancer. Ann Surg 1988; 207: 52–60. Tepper JE, Cohen AM, Wood WC, Hedberg SE, Orlow E. Intraoperative electron beam radiotherapy in the treatment of unresectable rectal cancer. Arch Surg 1986; 121(4):421. Weinstein GD, Rich TA, Shumate CR, et al. Preoperative infusional chemoradiation and surgery with or without an electron beam intraoperative boost for advanced primary rectal cancer. Int J Radiat Oncol Biol Phys 1995 30; 32(1):197–204. Lowy AM, Rich TA, Skibber JM, Dubrow RA, Curley SA. Preoperative infusional chemoradiation, selective intraoperative radiation, and resection for locally advanced pelvic recurrence of colorectal adenocarcinoma. Ann Surg 1996; 23(2):177–185. Pacelli F, Di Giorgio A, Papa V, et al. Preoperative radiotherapy combined with intraoperative radiotherapy improves results of total mesorectal excision in patients with T3 rectal cancer. Dis Colon Rectum 2004; 47:170–179. Hoskin PJ, de Canha SM, Brownes P, Bryant L, Glynne-Jones R. High dose after loading intraluminal brachytherapy for advanced inoperable rectal carcinoma. Radiother Oncol 2004, In press. Papillon J. Rectal and anal cancers: conservative treatment by radiation—an alternative to surgery. New York: Springer Verlag, 1982. Gerard JP, Romestaing P, Baulieux J, Benchimol. Local curative treatment of rectal cancer by radiotherapy alone. Colorect Dis 2003; 5:442–444. 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; 22:2404–2409. Rich TA, Skibber JM, Ajani JA. Preoperative infusional chemoradiation therapy for stage T3 rectal cancer. Int J Radiat Oncol Biol Phys 1997; 32: 1025–1029. Russell AN, Haris J, Roseberg 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 Jan 15; 46(2):313–322. Steele GD Jr., Herndon JE, Bleday R, et al. Sphincter-sparing treatment for distal rectal adenocarcinoma. Ann Surg Oncol 1999; 6:433–441. Kim CJ, Yeatman TJ, Coppula D, et al. Local excision of T2 and T3 rectal cancers after downstaging chemoradiation. Ann Surg 2001; 234(3):352–358.

226

Glynne-Jones and Hughes

108. Bonnen M, Crane C, Feig B, et al. Long term results using local excision after preoperative chemoradiation among selected T3 rectal cancer patients. Program and abstracts of the American Society for Therapeutic Radiology and Oncology (ASTRO) 43rd annual meeting. Int J Radiat Oncol Biol Phys 2001; 12:51 (abstract 16). 109. Gerard JP, Baulieux J, Francois Y, et al. The role of radiotherapy in the conservative treatment of rectal canoma. Lyon experience. Acta Oncol 1998; 37(3):253–258. 110. Paty PB, Nash GM, Baron P, et al. Long term results for the excision of rectal cancer. Ann Surg 2002; 236:522–530. 111. Mak AC, Rich TA, Schultheiss TH, Kavanagh B, Ota DM, Romsdahl MM. Late complications of postoperative radiation therapy for cancer of the rectum and rectosigmoid. Int J Radiat Oncol Biol Phys 1994; 28:597–603. 112. Myerson RJ, Picas J, Read TE. Combining chemotherapy with radiation in the adjuvant care of rectal cancer: morbidity of treatment and its impact on management decisions. Perspect Colon Rectal Surg 1999; 12:113–129. 113. Frykholm JG, Sintorn K, Montelius A. Acute lumbo-sacral plexopathy during and after preoperative radiotherapy of rectal adenocarcinoma. Radiother Oncol 1996; 38:121–130. 114. Graf W, Dahlberg M, Osman, Holmberg L, Pahlman L. Short-term preoperative radiotherapy results in downstaging of rectal cancer: a study of 1316 patients. Radiother Oncol 1997; 43:133–137. 115. 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; 17:2396–2402. 116. Moore HG, Gittleman AE, Minsky BD, et al. Rate of pathological complete response with increased interval between preoperative combined modality therapy in rectal cancer resection. Dis Colon Rectum 2004; 47:279–286. 117. Stein DE, Mahmoud NN, Anne PR, et al. Longer time interval between completion of neoadjuvant chemoradiation in surgical resection and does not improve downstaging of rectal carcinoma. Dis Colon Rectum 2003; 46:448–453. 118. Irvin TT, Goligher JC. A controlled clinical trial of three different methods of perineal wound management following excision of the rectum. Br J Surg 1975; 62:287–291. 119. Bujko K, Nowaki MP, Oledzki J, Soipylo R, Skoczylas J, Chwalinski M. Sphincter preservation after short term preoperative radiotherapy for low rectal cancer. Presentation of own data and a literature review. Acta Oncol 2001; 40:593–601. 120. Pahlman L, Glimelius B. Pre- or postoperative radiotherapy in rectal and rectosigmoid carcinoma. Report from a randomized multicenter trial. Ann Surg 1990; 211(2):187–195. 121. Tepper JE, O’Connell MJ, Petroni GR, et al. Adjuvant postoperative fluorouracil-modulated chemotherapy combined with pelvic radiation therapy for rectal cancer: initial results of Intergroup 0114. J Clin Oncol 1997; 15: 2030–2039.

Radiotherapy in the Treatment of Rectal Cancer

227

122. 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; 12:1565–1573. 123. Letschert JGJ, Lebesque JV, Aleman VMP. The volume effect in radiation related late small bowel complications: results of a clinical study of the EORTC Radiotherapy Co-operative Group in patients treated for rectal carcinoma. Radiother Oncol 1994; 32:116–123. 124. Dahlberg M, Glimelius B, Pahlman L. Preoperative radiation effect functional results after surgery for rectal cancer: results from a randomised study. Dis Colon Rectum 1998; 41:543–549. 125. Gervaz P, Rotholtz N, Wexner SD, et al. Colonic J-pouch function in rectal cancer patients: impact of adjuvant chemoradiotherapy. Dis Colon Rectum 2001; 44:1667–1675. 126. Lundby L, Jensen VJ, Overgaard J, Laurberg S. Long-term colorectal function after postoperative radiotherapy for colorectal cancer. Lancet 1997; 350(9077): 564. 127. Lundby L, Krogh K, Qvist N. Severe late anorectal morbidity following postoperative radiotherapy for rectal cancer—long term results from a randomised trial. Radiother Oncol 2000; 57(suppl 1):abstract 25. 128. Kollmorgen CF, Meagher AP, Wolff BG, Pemberton JH, Martenson JA, Illstrup DM. The long-term effect of adjuvant postoperative chemoradiotherapy for rectal carcinoma on bowel function. Ann Surg 1994; 220:676–682. 129. Burmeister BH, Schache D, Burmeister EA, Bell A, Poulsen MG, Waplole ET, Mackintosh J. Synchronous postoperative adjuvant chemoradiation therapy for locally advanced carcinoma of the rectum. Int J Colorectal Dis 2004; 19: 55–59. 130. Nathanson DR, Espat NJ, Nash GM, et al. Evaluation of preoperative and postoperative radiotherapy on long-term functional results of straight coloanal anastomosis. Dis Colon Rectum 2003; 46:888–894. 131. Bonner JA, Harari PM, Giralt J, et al. Cetuximab prolongs survival in patients with local regionally advanced squamous cell carcinoma of the head and neck. A phase III study of high dose radiation therapy with or without Cetuximab. J Clin Oncol 2004; 22:489S (abstract 5507).

9 The Treatment of Metastatic Colorectal Cancer Eric Van Cutsem Digestive Oncology Unit, University Hospital Gasthuisberg, Leuven, Belgium

Leonard Saltz Gastrointestinal Oncology Service, Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, New York, U.S.A.

INTRODUCTION Colorectal cancer (CRC) is one of the most common causes of cancer worldwide (1). Around 25% of patients present with overt metastases and almost 50% of newly diagnosed patients will die of their disease. In patients with resectable liver or lung metastases, the five-year survival is 25% to 30% (2). Most patients, however, present with nonresectable metastases. For these patients, chemotherapy is the only therapeutic option. For almost 50 years, 5-fluorouracil (5-FU) has been the cornerstone of systemic cytotoxic chemotherapy for patients with metastatic CRC. In recent years, many important new treatment options have been developed, which changed the outcome for patients with metastatic CRC. Therapeutic milestones have included improved use of 5-FU, introduction of the new cytotoxic agents irinotecan and oxaliplatin, and the development of the oral fluoropyrimidines and novel targeted agents.

229

230

Van Cutsem and Saltz

CYTOTOXIC AGENTS IN METASTATIC CRC Benefit of Palliative Chemotherapy Untreated patients with metastastic CRC have a median survival of five to six months. It has been shown in randomized studies that chemotherapy for metastatic CRC prolongs the survival and maintains or improves the quality of life (3–5). In these trials, old 5-FU-based chemotherapy regimens were used. The median survival of the patients treated with chemotherapy was around 11 to 12 months compared to five to six months for the best supportive care (BSC) groups (3). A Nordic multicenter study randomized 183 asymptomatic patients with advanced CRC to initial chemotherapy with methotrexate, 5-FU, and folinic acid (FA) rescue for six months or to primary expectancy with chemotherapy only after the appearance of symptoms. Overall survival was significantly longer in the primary chemotherapy group (median 14 vs. nine months); however, observation intervals were relatively long, and more than a third of the observation group never received chemotherapy (4). It is also suggested that an objective response is beneficial in most patients since it is associated with survival prolongation and relief of tumor-related symptoms (5). A prolonged stationary disease is usually also of benefit for the patients, since this is usually also associated with prolonged survival and with a decrease in tumor-related symptoms. Regimens of 5-FU/-FA 5-FU has been the therapeutic mainstay for CRC for over 40 years. It is converted into its active metabolite, fluorodeoxyuridine monophosphate (FdUMP), within the cell. FdUMP inhibits the enzyme thymidylate synthase (TS), preventing pyrimidine and therefore DNA synthesis. This action confers a degree of S-phase specificity upon the drug. 5-FU is also falsely incorporated into RNA interfering with protein synthesis. The major dose-limiting toxicities of the common regimens are mucositis, diarrhea (which can usually be controlled by loperamide), plantar-palmar erythema, and mild myelosuppression. For many years, the standard regimen was daily bolus 5-FU given for five consecutive days every four weeks. Traditionally, a bolus injection over one to three minutes was given. Several clinical trials, using rigorous definitions of response, suggest that response rate (RR) to single-agent bolus administration of 5-FU is approximately 10%. Administration of a drug using a minibag system over 10 to 30 minutes has been used later more frequently. Due to the rapid catabolism of 5-FU, this prolongation of ‘‘bolus’’ 5-FU results in lower plasma peak levels and area-under-the-curve concentrations, and potentially fewer cytotoxic effects (5,6). More recently, two modifications have improved outcome either in terms of improved RRs or

The Treatment of Metastatic Colorectal Cancer

231

decreased toxicity. Firstly, the observations of S-phase specificity and short half-life of 5-FU led investigators to postulate that prolonged infusional 5-FU might increase the proportion of tumor cells susceptible to cytotoxicity of the drug. It has also been speculated that bolus 5-FU and infusional 5-FU act through different mechanisms (7). After a number of small, relatively underpowered trials, a meta-analysis in advanced CRC, comparing infusional 5-FU against the bolus regimen, found that infusional therapy delivered a higher RR (22% vs. 14%; p ¼ 0.0002) and also prolonged the survival ( p ¼ 0.04), although the median survival times were very close (12.1 months for the continuous infusion and 11.3 months for the bolus injections) (8). The pattern of toxicity was shifted with a decrease in mucositis and neutropenia and a higher frequency of palmar-plantar syndrome (9). Secondly, biomodulation of 5-FU with other drugs, including levamisole, methotrexate, and FA, was also investigated. FA administration, which increases the intracellular pool of reduced folate and stabilizes the FdUMP/TS complex, has been the most successful biomodulatory agent. Studies have shown that the addition of FA to bolus 5-FU improves RRs (23% vs. 11%) compared to single-agent 5-FU (10). The use of more optimal infusional 5-FU/FA regimens has changed the attitude towards the treatment of patients with advanced CRC. Several studies have shown a higher RR and a lower toxicity for infusional 5-FU/FA regimens compared to bolus 5-FU/FA regimens (11–15). The LV5FU2 regimen of de Gramont utilizes both infusion and biomodulation to improve outcome. It comprises a two-hour intravenous infusion of FA (200 mg/m2), followed by intravenous bolus of 5-FU (400 mg/m2) and a 22-hour intravenous infusion of 5-FU at a dose of 600 mg/m2. This is repeated on day 2 and the whole cycle repeated every 14 days (11). The ‘‘de Gramont’’ regimen has been tested recently against the bolus regimen of 5-FU/FA in a French trial and against a continuous infusion of 5-FU (single agent) and against a three-weekly short infusion of tomudex (a direct TS inhibitor) in a U.K. trial (11,12). In the French trial, the RR (33% vs. 14%; p < 0.002) and time to tumor progression (TTP) (27.6 vs. 22 weeks; p < 0.002) were superior for the de Gramont regimen compared to the bolus regimen. The overall survival was similar (14 vs. 13 months; p ¼ 0.07) (11). In the U.K. trial, RR and overall survival were equivalent for the de Gramont regimen compared to a continuous infusion, but the de Gramont regimen appeared to have some advantages with respect to patient acceptability and quality of life (12). Other regimens of infusional 5-FU/FA have been developed and used widely in some European countries (e.g., weekly 24 hours AIO regimen or weekly 48 hours 5-FU alone in the TTD regimen) (13–15). It is generally accepted that infusional regimens are a more optimal way of administering 5-FU/FA because the RR is higher and the TTP is longer, and because the toxicity is less pronounced (less mucositis and leucopenia/ neutropenia). The median survival for infusional regimens is, however, not

232

Van Cutsem and Saltz

longer compared to bolus regimens, and they require a central catheter in all patients (Portacath or Hickmann catheter). Second-Line Treatment with Cytotoxic Agents CPT-11 or irinotecan is a camptothecin analog that inhibits DNA topoisomerase I and induces single-strand DNA breaks and replication arrest. Oxaliplatin is a third-generation platinum analog that induces DNA crosslinkages and apoptotic cell death. The studies of irinotecan and oxaliplatin in 5-FU refractory patients have demonstrated the activity of these drugs and have also shown that patients with metastatic CRC can benefit from a second-line treatment. It has been shown in nonrandomized phase II studies that the RR of irinotecan and of the combination 5-FU/FA/oxaliplatin in patients progressive under 5-FU/FA is approximately 15%, and that 30% to 40% of patients can have tumor stabilization. This means that
50% of patients with progressive CRC will have tumor growth control for several months. It has been shown later in randomized phase III trials that this is translated into a survival benefit and a longer maintenance of the quality of life because of control of tumor-related symptoms (16–18). In the initial studies, irinotecan was administered as a single agent, but in later studies similar RRs have been shown with the combination of irinotecan þ 5-FU/FA (FOLFIRI) in patients resistant to 5-FU/FA. Because the combination irinotecan/5-FU/FA has a better tolerability profile than irinotecan monotherapy, this combination is often used also in 5-FU-resistant patients. It has been demonstrated in preclinical experiments that the combination of oxaliplatin plus 5-FU acts synergistically. Two pivotal randomized phase III studies of irinotecan versus best supportive care (BSC) and of irinotecan versus infusional 5-FU/FA demonstrated a survival benefit for irinotecan in the second-line treatment of patients with 5-FU-resistant metastatic CRC (16,17). The quality of life of patients treated with irinotecan was also superior compared to BSC and was at least identical compared to the quality of life of patients treated with infusional 5-FU/FA (19,20). The phase II studies with the combination 5-FU/FA/oxaliplatin (FOLFOX) suggest a higher activity for this combination than for oxaliplatin monotherapy (21–27). An excellent randomized phase III trial has confirmed this finding and has clearly shown that the combination 5-FU/ FA/oxaliplatin is more active than the LV5FU2 regimen and oxaliplatin monotherapy in the second-line treatment of CRC (28). Cytotoxic Agents in First-Line Treatment Raltitrexed, irinotecan, and oxaliplatin have been developed and extensively studied in the first-line treatment of metastatic CRC.

The Treatment of Metastatic Colorectal Cancer

233

Raltitrexed causes a direct and specific TS inhibition without secondary effects on RNA or protein synthesis. In three large trials, patients with advanced CRC were randomized between raltitrexed and a bolus regimen of 5-FU plus a low dose of FA (Mayo Clinic regimen) or a high dose of FA (in one trial) (29,30). Objective RRs were similar for raltitrexed and 5-FU/FA. The median survival was similar for raltitrexed and 5-FU/FA in two studies, but inferior survival was found for raltitrexed compared to 5-FU/FA in one trial (27). TTP was also inferior for raltitrexed in one trial (30). More recently, raltitrexed has been compared with two infusional 5-FU-based regimens (de Gramont and protracted continuous infusion) (12). Although similar RRs and survival were seen in the three arms, time to disease progression and quality of life were inferior in the raltitrexed arm. Several phase I and II trials have been performed with the combination of raltitrexed and 5-FU or irinotecan or oxaliplatin. Phase II studies have shown an interesting activity of some of these combinations, but since no phase III data are available these combinations are not frequently used in clinical practice (30). Raltitrexed has been abandoned for further evaluation in studies and is used only occasionally, also because of some unexpected toxicity issues. Studies with the combination of 5-FU/FA plus irinotecan and of 5-FU/FA plus oxaliplatin have shown that combination chemotherapy is more active than 5-FU/FA in patients with advanced CRC. In three randomized studies, a higher RR and a longer TTP or progression free survival (PFS) were demonstrated for patients treated with 5-FU/FA/irinotecan compared to patients treated with 5-FU/FA only. The median survival difference was significantly increased for the combination treatment compared to 5-FU/ FA alone in two of the trials (31–35). The RR and the progression-free survival were also higher for the combination 5-FU/FA/oxaliplatin compared to 5-FU/FA in the randomized trials. The median survival of the combination treatment was not significantly prolonged, probably because of the effect of crossover in the patients treated with 5-FU/FA alone and also because the trials were not powered to show a survival difference (34,35). The results of these important studies have demonstrated that combination treatment is more active than 5-FU/FA alone and it is therefore accepted that combination treatment is the standard option in the first-line treatment of advanced CRC. These studies do not answer the question whether all patients have to be treated in the first-line setting with a combination regimen of 5-FU/FA/irinotecan or 5-FU/FA/oxaliplatin or with 5-FU/FA only. It is also shown in these studies that, although the number of side effects was higher for combination treatment, the side effect pattern was acceptable and manageable and did not influence the quality of life of the patients. There was, however, some concern in the United States with the use of the bolus regimen of 5-FU/FA þ weekly irinotecan (IFL regimen) (36). A higher incidence of treatment-related deaths has been reported

234

Van Cutsem and Saltz

with this regimen compared to 5-FU/FA (36,37). In Europe, this has not been reported with the infusional regimens of 5-FU/FA plus irinotecan (37). In a French study, patients were randomized to a sequential combination study of 5-FU/FA/irinotecan followed by 5-FU/FA/oxaliplatin after progression or the inverse sequence (5-FU/FA/oxaliplatin followed by 5-FU/FA/irinotecan) (38). The TTP and survival were not statistically different for the two sequences. The median survival was more than 20 months with a sequential treatment. These data confirm that combination chemotherapy can have a clear impact on the outcome of patients with metastatic CRC. A U.S. intergroup study compared in a large randomized phase III trial the IFL regimen (bolus 5-FU/FA þ irinotecan) with infusional 5-FU/FA þ oxaliplatin (FOLFOX) and with oxaliplatin plus irinotecan in the first-line treatment of metastatic CRC. The RR, TTP, and survival were significantly higher and longer for the FOLFOX regimen compared to the IFL regimen (39). Moreover, the FOLFOX regimen led to less adverse events. This study led to the approval of oxaliplatin in the first-line treatment of metastatic CRC in the United States. In this study, however, a bolus regimen of 5-FU was used in the irinotecan combination, while in combination with oxaliplatin the more optimal infusional regimens of 5-FU were used. There was also a remarkable difference in the second-line treatment in this study, which can have contributed partly to the difference in survival. In other studies, in which infusional regimens of 5-FU were used as backbones for the oxaliplatin as well as the irinotecan combinations, the RR, TTP, and survival were similar (38,40). Therefore, oxaliplatin/5-FU/FA and irinotecan/5-FU/FA are considered as reference options in the first-line treatment of metastatic CRC with a similar activity, but with a different safety profile: oxaliplatin leads to cumulative neurotoxicity, and irinotecan can lead to diarrhea and alopecia. Two randomized studies have compared the triplet combination of 5-FU/FA plus oxaliplatin and irinotecan (FOLFOXIRI) with a doublet. A Greek study failed to show a significant advantage in survival, TTP, and RR of the triplet combination compared to the doublet of FOLFIRI in 283 patients (41). An Italian study, however, showed a higher RR, longer progression-free survival, and longer overall survival in 244 patients for the triplet FOLFOXIRI compared to the doublet FOLFIRI (42). However, in view of the limited data and the increased toxicity of the triplet combination, this combination cannot be recommended for routine clinical use today. Oral Fluoropyrimidines It has been shown that the oral fluoropyrimidines are at least as active as intravenous fluoropyrimidines (43–51). Three oral fluoropyrimidines are extensively investigated in CRC: uracil ftorafur (UFT), eniluracil, and capecitabine. UFT is a combination of uracil and tegafur (a prodrug of 5-FU) in a

The Treatment of Metastatic Colorectal Cancer

235

fixed molar ratio of 4:1. Uracil is a normal substrate for dihydropyrimidine dehydrogenase and competitively blocks the actions of this enzyme, allowing tegafur absorption and the availability of biologically active plasma concentrations of 5-FU. UFT has usually been administered with oral leucovorin (LV ¼ folinic acid). Eniluracil is a direct inhibitor of dihydropyrimidine dehydrogenase. The compound is given orally together with oral 5-FU. Capecitabine, a fluoropyrimidine carbamate, is a 5-FU prodrug that can be absorbed through the intestinal mucosa and is converted to 5-FU through three enzymatic steps. It has been shown in six large randomized trials in the first-line treatment of advanced CRC that these oral fluoropyrimidines have a similar activity to bolus IV 5-FU/FA and are less toxic (43–51). Eniluracil has been withdrawn from further development because some of the activity endpoints showed a trend for lower activity of the combination of oral eniluracil plus oral 5-FU compared to IV 5-FU/FA (50,51). In one randomized phase III study, a significantly higher RR was shown for capecitabine compared to IV 5-FU/FA (46). The combined analysis of the two large randomized phase III trials has also shown a higher RR for capecitabine compared to IV 5-FU/FA. The TTP and survival were identical (47). Although capecitabine induces hand–foot syndrome more frequently compared to the other oral fluoropyrimidines and IV 5-FU/FA, the oral fluoropyrimidines have clearly more toxicity, quality of life, and convenience advantages compared to intravenous 5-FU/FA (52). Phase I and II studies have investigated extensively the combination of capecitabine with irinotecan or oxaliplatin. Data on combination studies with UFT/LV are more limited. The data of several phase II trials and of a few underpowered randomized trials seem to indicate that the RR as well as the TTP and survival of the combinations of capecitabine plus oxaliplatin or capecitabine plus irinotecan are similar compared to the combination of 5-FU/FA with oxaliplatin or irinotecan. The available data of the phase II studies and of the small randomized studies seem to suggest that combination of capecitabine with oxaliplatin, and possibly also with irinotecan, will be able to replace IV 5-FU/FA in combination treatment with irinotecan or oxaliplatin (53–61). A large phase III trial randomizing an infusional regimen of 5-FU/FA/oxaliplatin (FOLFOX-4 regimen) and capecitabine plus oxaliplatin at a dose of 2  1000 mg/m2/day on days 1–14 and 130 mg/m2 IV every three weeks in approximately 1900 patients is expected to be reported in late 2006. A second randomization has been done in this study in a 2  2 design: bevacizumab or placebo. Although the efficacy data from the combination of capecitabine and irinotecan also seem to show a similar activity compared to IV 5-FU/FA/irinotecan, the safety profile and the dose of the combination of capecitabine and irinotecan seem to be less clear. The European Organisation for Research and Treatment of Cancer (EORTC) gastrointestinal cancer group addresses this question in a trial randomizing capecitabine/irinotecan with the FOLFIRI regimen. A second randomization was done in this trial: placebo or celecoxib. In this trial, as well as in

236

Van Cutsem and Saltz

some other trials, a dose of capecitabine of 2  1000 mg/m2/day on days 1–14 and of irinotecan of 250 mg/m2 IV every three weeks are used. The EORTC had to stop this trial because of increased toxicity with the regimen of capecitabine/irinotecan (62). The Dutch CAIRO trial is evaluating the same regimen in a sequential strategy, but has not yet reported important safety problems in a preliminary report of a part of the treated patients (63). The studies with the oral fluoropyrimidines in combination with the cytotoxic agents are important in view of the further developments and the combination with the biologicals, such as bevacizumab and cetuximab. Strategy with Cytotoxic Agents It can be concluded from the available data that:  The infused regimens of 5-FU/FA are a more optimal way of administering 5-FU/FA than the bolus regimen.  Combination of two cytotoxic agents is more active in the first line of metastatic CRC than 5-FU/FA alone. The combination of a triplet combination cannot be recommended today.  The combinations 5-FU/FA/irinotecan and 5-FU/FA/oxaliplatin have similar activities in the first-line treatment of metastatic CRC, but different safety profiles.  The oral fluoropyrimidines are at least as active as IV 5-FU/FA. Combination studies with capecitabine and oxaliplatin or irinotecan show that capecitabine may be an acceptable alternative to IV 5-FU/FA in these combinations.  Second-line treatment is indicated in patients with progressive disease in good condition. In oxaliplatin-failing patients, irinotecan or the combination 5-FU/FA/irinotecan are options and in irinotecan-failing patients 5-FU/FA/oxaliplatin is a reasonable option. Cetuximab-based regimens can also be considered (see later).  Exposure of patients with metastatic CRC to the three available cytotoxic agents (fluoropyrimidines, irinotecan, and oxaliplatin) increased the overall survival (64). TARGETED THERAPIES FOR METASTATIC CRC More recently, several newer agents have entered the clinical arena for the treatment of CRC. These so-called targeted agents exploit knowledge gained from preclinical studies on how both normal and malignant cells grow, divide, and survive. To some degree, the emphasis on targeting as a new concept is often a bit overstated. All of the cytotoxic therapies previously discussed in this chapter are, to some extent, targeted agents. We know the target of the fluoropyrimidines, the camptothecans, and the platinum-based compounds. More recently, two newer targets have been exploited: the

The Treatment of Metastatic Colorectal Cancer

237

epidermal growth factor receptor (EGFR) and the vascular endothelial growth factor (VEGF). The agents that have shown clinical activity in CRC through interference with these targets are reviewed. Angiogenesis Inhibitors Bevacizumab Bevacizumab is a humanized monoclonal antibody that targets and binds to VEGF, thereby inactivating this growth factor before it can bind to its intended receptor (65–67). VEGF is a central component in the process of angiogenesis or the development of new blood vessels. Up to a size of 1–2 mm, tumor cells are able to obtain required nutrients and oxygen from surrounding fluids via diffusion. Once the size of the tumor begins to increase further, however, new blood vessels must be developed to support the tumor, and VEGF is a key component in the signaling pathway employed to initiate the growth of these new vessels. Numerous strategies and compounds are being investigated with the intention of disrupting this angiogenic process. However, to date, bevacizumab is the only such agent to have demonstrated substantial activity in CRC. The first trial that suggested a potential role for bevacizumab in the management of CRC was a relatively modest-sized randomized phase II trial of two different doses of bevacizumab plus weekly 5FU and high-dose leucovorin (Roswell Park 5FU/LV schedule) (68). Patients were randomized to 5FU/LV alone, or with 5 or 10 mg/kg of bevacizumab given every other week. The RR, TTP, and overall survival were superior in the 5 mg/kg (low-dose bevacizumab) arm, with the 10 mg/kg bevacizumab arm appearing modestly superior to chemotherapy alone but inferior to the low-dose bevacizumab arm (68). Thrombosis was the most significant adverse event, and was fatal in one patient. Hypertension, proteinuria, and epistaxis were also seen. Since the role of this small randomized phase II study was to ‘‘pick the winner’’ for further development, the 5 mg/kg dose of bevacizumab was taken forward for phase III development. Hurwitz et al. (69) reported the pivotal phase III trial for bevacizumab in CRC. In this trial, the then standard regimen of weekly IFL (irinotecan, leucovorin, fluorouracil with fluorouracil given by bolus injection) was given to approximately 800 previously untreated CRC patients (32). Half of these patients were randomized to receive 5 mg/kg of bevacizumab in addition to IFL, while the other half received IFL plus a placebo. The group that received bevacizumab experienced better clinical results, with improved RR (45% vs. 35%; p < 0.003), improved progression-free survival (10.6 vs. 6.2 months; p < 0.00001), and improved overall survival (20.3 vs. 15.6 months; p ¼ 0.00003) compared to the control group receiving IFL alone (69). The survival benefit of 4.7 months is particularly noteworthy, in that no randomized study to date has shown a larger survival difference in CRC patients.

238

Van Cutsem and Saltz

While the subjective toxicity of bevacizumab is modest enough that a placebo control is credible, it would be wrong to characterize bevacizumab as without serious potential for toxicity. The toxicities more commonly associated with chemotherapy, such as nausea, vomiting, diarrhea, cytopenias, and asthenia, do not appear to be significant issues for bevacizumab. It should be noted, however, that bevacizumab has essentially no demonstrated activity as a single agent in CRC, and so must be given with standard cytotoxic regimens that do possess these toxicities. In addition, bevacizumab brings several toxicities of its own to the table: hypertension, proteinuria, arterial thrombosis, mucosal bleeding (mainly epistaxis), impaired wound healing, and a low chance of gastrointestinal perforations (1% to 1.5% of treated patients). The risk of venous thromboembolism was not higher in the bevacizumab arm compared to the placebo arm. The risk of arterial thrombosis is increased, especially in patients over 65 years and in patients with a history of arterial thrombosis. In addition to this one large, strongly positive study, another smaller trial also supports the role of bevacizumab in the first-line treatment of metastatic CRC. In a trial designed for those patients in whom irinotecan-based therapy was thought by the investigator to be inappropriate for the patient, the patients were randomized to 5-FU/FA using a weekly bolus schedule with high-dose leucovorin (Roswell Park schedule) or to the same schedule plus 5 mg/kg of bevacizumab every other week (70). While the study lacks adequate power to assess a survival advantage, improved RRs and TTP were seen with the bevacizumab-containing arm. In a combined analysis of three relatively small randomized trials of 5-FU/FA or IFL versus 5-FU/ FA plus bevacizumab a significant survival difference was shown: 17.9 versus 14.6 months. The progression-free survival was also significantly longer in the bevacizumab-treated patients: 8.8 versus 5.6 months (71). The ECOG 3200 phase III trial randomized 822 patients in the secondline treatment of metastatic CRC after progression on irinotecan/5-FU/FA, between the FOLFOX regimen with or without bevacizumab (10 mg/kg) and bevacizumab alone. Bevacizumab monotherapy was not active in this setting (72). Treatment with this combination regimen significantly improved median overall survival (OS) from 10.8 months with FOLFOX4 alone to 12.9 months (p ¼0.0018). Patients treated with bevacizumab plus FOLFOX had an increased incidence of grade 3/4 hypertension, bleeding, sensory neuropathy, and vomiting events compared with patients treated with FOLFOX4 alone. It should be noted that all of the patients on ECOG 3200 were bevacizumab-naive. Therefore, these data cannot be extrapolated to the question of second-line use of bevacizumab after failure of a first-line bevacizumab-containing regimen; that question remains unaddressed at this time. Furthermore, preliminary data from a phase II study of bevacizumab combined with capecitabine plus oxaliplatin (XELOX) in first-line treatment of 30 patients with metastatic CRC have shown this regimen to be active at a level similar to that of

The Treatment of Metastatic Colorectal Cancer

239

FOLFOX plus bevacizumab (73). The regimen was generally well tolerated. TREE-2 is a randomized phase II study of three oxaliplatin-based plus fluoropyrimidine (bolus 5-FU/FA, infused 5-FU/FA, and capecitabine) regimens in combination with bevacizumab (5 mg/kg every two weeks or 7.5 mg/kg every three weeks) in all three arms. The study shows a high RR and a long TTP, especially in the infused 5-FU/FA and capecitabine arms (74). Together, the safety data indicate that combining bevacizumab with oxaliplatincontaining and infusional 5-FU regimens is well tolerated and does not substantially alter the toxicity profiles of these regimens. To date, there are no large-scale randomized data involving bevacizumab with either an infusional 5FU-based regimen or an oxaliplatin-based regimen in the first-line treatment of metastatic CRC. However, the mechanism of action of bevacizumab, the data from the studies showing that bevacizumab increases the activity of 5-FU/LV, the data from phase II studies indicating a high activity of bevacizumab in combination with several oxaliplatin-based combinations and the activity of the 5-FU/FA/ oxaliplatin/bevacizumab in the second-line treatment support its use in the first-line treatment of metastatic CRC. The results of the randomized phase III study of FOLFOX compared to capecitabine plus oxaliplatin
bevacizumab in the first-line treatment are expected to be reported in late 2006 or early 2007. It has become a common current practice to use an appropriate firstline cytotoxic regimen, such as either FOLFIRI or FOLFOX, as a front-line regimen (35) and to add bevacizumab 5 mg/kg every other week to whichever regimen is selected. Whether this turns out to be the correct use of bevacizumab or not will have to await the maturation of further clinical trials. There are, however, several important points that should be considered in terms of where data do not support the use of bevacizumab in CRC. Firstly, there is no evidence to support the use of bevacizumab in the adjuvant setting. There are some reasons to be optimistic that bevacizumab may improve the effectiveness of adjuvant treatments, but it should also be remembered that an antiangiogenic strategy like bevacizumab might be ineffective against the micrometastases that are being treated by adjuvant therapy. Given that serious and even fatal toxicities from bevacizumab can occur, the use of bevacizumab in the adjuvant setting should be limited to appropriately designed clinical trials until and unless data supporting the risk/benefit ratio of bevacizumab in this setting are obtained. Secondly, it has been shown that the combination of bolus 5-FU/FA plus bevacizumab in CRC refractory to 5-FU, irinotecan, and oxaliplatin is not active (75). A question that remains unaddressed at this time is whether to continue bevacizumab with second-line therapy following failure of a bevacizumabcontaining first-line regimen. While some have advocated this, there are no data to support it. Given that bevacizumab can result in significant and even life-threatening toxicity, and that the economic cost of continuing

240

Van Cutsem and Saltz

bevacizumab is considerable, at this time the practice of continuing bevacizumab in second-line therapy after failure of bevacizumab in first-line therapy cannot be recommended. Studies looking at clinical, biochemical, and molecular markers for higher activity after treatment with bevacizumab failed also to show the usefulness of these markers in the discrimination of patients who will benefit more or less from bevacizumab-containing regimens. An analysis of predictive markers showed indeed that bevacizumab increased the activity of irinotecan plus 5-FU/LV regardless of the level of VEGF expression, thrombospondin expression, and microvessel density (76). Mutations of k-ras, b-raf, and p53 could not predict for a prolonged survival on bevacizumab plus IFL (77). Molecular markers for response prediction are under intensive investigation, but it is unlikely that they will influence in the short term our daily clinical practice. The data, therefore, so far mainly suggest that the selection of patients who will benefit most from a bevacizumabcontaining regimen should rely on the likelihood for an individual patient of developing toxicity and on efficacy as demonstrated in the larger studies. EGFR Inhibitors Cetuximab (C225) Cetuximab is a chimeric monoclonal antibody that binds selectively to the EGFR. The EGFR is a 170,000 kD transmembrane glycoprotein that is involved in signaling pathways affecting cellular growth, differentiation, and survival (78–81). By binding to the EGFR receptor binding site, cetuximab blocks ligands from binding to and activating EGFR, preventing phosphorylation of the tyrosine kinase on the intracellular domain of the receptor and thereby preventing receptor signaling. Preclinical models indicated that cetuximab had modest in vitro and in vivo single-agent activity, but had more significant activity when combined with cytotoxic chemotherapy. Based on this observation, and on a single anecdotal report of a major response to cetuximab plus irinotecan in a young irinotecan-refractory patient, a multicenter phase II trial was initiated (82). Patients were treated with the same dose and schedule of irinotecan that they had previously failed, with the addition of cetuximab at what is now the standard dose of 400 mg/m2 loading dose at week 1, followed by weekly 250 mg/m2 over one hour. Prior dose reductions of irinotecan that had been made before study entry were maintained upon entry to this trial. One hundred twenty patients who were determined by their treating physician to have failed prior CPT-11 were treated. The major objective RR for the patients who had previously failed irinotecan was 22.5%, as reported by an independent response assessment committee. As might have been expected, toxicity from irinotecan was modest in this population in whom dose modifications had already been made during prestudy treatment. Three percent of patients

The Treatment of Metastatic Colorectal Cancer

241

had to discontinue cetuximab therapy due to allergic, anaphylactoid reactions. Seventy-five percent of patients treated experienced a skin rash (12% grade 3), which has come to be recognized as characteristic of EGFR inhibitors. This rash appears superficially to resemble acne, and it has come to be referred to as an acneiform rash; however, it can be seen microscopically that this is not acne, and topical acne medications are not effective in management of this rash. The results seen in the above-described cetuximab plus irinotecan trial led to the initiation of a multicenter phase II trial to evaluate the activity of singleagent cetuximab in irinotecan-refractory CRC. Five of 57 patients (9%) achieved a partial response on the basis of an independent radiologic review (83). A subsequent larger trial (BOND trial) provided strong confirmatory evidence of the activity of cetuximab in CRC (84). This trial randomized 329 patients with EGFR-expressing irinotecan-refractory CRC between cetuximab/irinotecan and cetuximab alone in a 2:1 schema. The RRs were 22.9% for cetuximab/irinotecan and 10.8% for cetuximab alone, respectively, as reported by independent radiology review. These results closely corroborated the two previously reported trials. TTP in the BOND study was 4.0 months for the combination versus 1.6 months for single-agent cetuximab. Survival in the two arms was not significantly different, because crossover was allowed to cetuximab/irinotecan after progression on cetuximab alone and because the trial was not powered to show a survival difference. It should be noted that all patients received cetuximab in both arms, so this trial is in no way an assessment of whether or not cetuximab treatment confers a survival advantage. That question is being studied in a trial in chemorefractory CRC randomizing patients between cetuximab plus BSC versus BSC. Cetuximab thus has established activity in the salvage setting; however, its role in front-line therapy remains investigational at this time. Cetuximab has been investigated in a preliminary manner in combination with first-line therapy in phase II trials. A small phase II pilot trial of cetuximab plus weekly bolus IFL demonstrated a 44% RR (85). Small phase II experiences of cetuximab plus irinotecan and weekly infusional 5FU/LV, FOLFIRI, FOLFOX and weekly infusional 5-FU/FA plus oxaliplatin regimens have also been reported (86–88). All of these appear tolerable with encouraging activity. Thus far, no randomized trials of first-line cetuximab plus chemotherapy have been reported. As such, no data are available on what impact this drug will have on survival, and other efficacy endpoints in first-line combinations. Several such trials are in progress or in the planning stages, but, given the cutaneous toxicity as well as the significant economic cost of extended cetuximab treatment, front-line use of cetuximab must be regarded as investigational at this time. The question of addition of cetuximab to adjuvant therapy of stage III colon cancer is also being investigated, but again this remains a study question, and not part of standard care, at this time.

242

Van Cutsem and Saltz

Other EGFR-Targeting Agents Panitumumab, formally known as ABX-EGF, is a fully humanized monoclonal antibody that also targets the EGFR (89). A phase II study of this agent in patients demonstrated a 10% RR and 38% rate of stable disease. The median duration of response was 5.2 months (95% CI: 4.5–7.5 months). The median PFS was 2.0 months (95% CI: 1.9–3.8 months) and the median survival amounted to 7.9 months (95% CI: 5.7–9.9 months) (89). The vast majority of patients experienced some degree of acneiform rash. Only one of the 148 patients treated experienced a dose-limiting allergic reaction, suggesting that this agent might have a lower incidence of allergic reactions than cetuximab. Although the comparisons are nonrandomized, this agent appears to be quite similar to cetuximab, both in mode of action and in single-agent efficacy. A randomized phase III trial of panitumumab plus BSC versus BSC in four patients with EGFR-expressing oxaliplatin- and irinotecan-refractory patients has shown a significantly longer TTP for the panitumumabtreated patients (90). A small phase II study showed an interesting activity of the combination of panitumumab and irinotecan/5-FU/LV in the first-line treatment of metastatic CRC (91). Of 19 patients, 47% had an RR, and disease was stable in 32% (91). Matuzumab (EMD 72000), another fully humanized monoclonal antibody against the EGFR, has also shown some antitumor activity in CRC in early clinical trials; however, clinical data on this agent are more limited at this time (92). Compared with the anti-EGFR monoclonal antibodies, the role of tyrosine kinase inhibitors of EGFR has not been extensively investigated in metastatic CRC. However, it seems that gefitinib as a single agent is not active in chemorefractory CRC (93,94). In combination with FOLFOX, gefitinib does have a high RR (78% first line; 30% to 33% pretreated patients), but toxicity is also high, with grade 3/4 diarrhea occurring in 49% of patients (95,96). The combination of erlotinib with capecitabine and oxaliplatin is more feasible but less active, and FOLFOX plus the EGFR tyrosine kinase inhibitor EKB-569 also seems feasible (97,98). Combination of Anti-EGFR and Angiogenesis Inhibitors Animal models suggest additive efficacy and sometimes synergy can be achieved using EGFR inhibitors in combination with agents that inhibit the vascular endothelial growth factor receptor (VEGFR) (99). There is therefore good preclinical rationale for the BOND 2 study of cetuximab plus bevacizumab with or without irinotecan in irinotecan-refractory CRC patients not tested for the presence of the EGF receptor (100). In a randomized phase II study, irinotecan-refractory CRC patients received

The Treatment of Metastatic Colorectal Cancer

243

irinotecan (same dose and schedule as per their last treatment administration prior to study) plus cetuximab (400 mg/m2 loading dose, then weekly at 250 mg/m2) plus bevacizumab (5 mg/kg given every other week) versus cetuximab (400 mg/m2 loading dose, then weekly at 250 mg/m2) plus bevacizumab (5 mg/kg given every other week) (19). Use of the two antibodies plus irinotecan (N ¼ 41) is associated with a 37% RR and 7.9 months TTP, compared with an RR of 20% and 5.6 months TTP in patients treated with cetuximab plus bevacizumab alone (N ¼ 40). In the BOND study, the cetuximab/irinotecan arm achieved a RR of 23% and the TTP was 4.1 months, while the cetuximab-alone arm had a RR of 11% and a TTP of 1.5 months (84). In the BOND 2 study, no unexpected toxicities were encountered. Grade 3 was found in seven (17%) and grade 2 rash was observed in 25 patients (60%) on the cetuximab/bevacizumab/irinotecan arm, and in eight patients (20%) and 26 patients (65%), respectively, on the cetuximab/bevacizumab arm. Grade 3/4 diarrhea was observed in 24% of patients on the cetuximab/bevacizumab/irinotecan arm and 0% of patients on the cetuximab/bevacizumab arm (100). The combination of cetuximab/bevacizumab, alone or with irinotecan, appeared tolerable and active and warrants the investigation of this combination in front-line combination chemotherapy regimens. The Incidence and Significance of Rash Although more than 80% of patients suffer from an acneiform-like rash after the administration of the anti-EGFR monoclonal antibodies, rash grade 3/4 is present in less than 10% of the patients. The rash is only rarely a reason to stop the treatment (84,89). The association between rash and efficacy is proving intriguing. Retrospective analysis of the BOND data showed a clear association between higher grades of skin reaction and RR and median TTP (84). This was true also for OS, the median value rising from three months in patients with no rash to 14 months in those with rash of grade 3 severity. The same relationship was seen in the panitumumab study (89). The association between rash severity and survival seems to hold true across the range of clinical trials of cetuximab in CRC, and indeed in the treatment of tumors at other disease sites (82,84,101–104). However, the fact that all of these analyses are retrospective suggests that these data should be treated with caution. They should certainly not be made the basis of any decision by regulatory authorities to restrict continued dosing with EGFR inhibitors to patients showing a rash. A study in which the dose of cetuximab is escalated progressively until rash occurs or until a response is found (EVEREST study) to try to understand this relationship. Serial skin and tumor biopsies for expression of EGFR and markers of downstream signaling should help elucidate the link between rash and antitumor activity. The management of EGFR inhibitor rash and the associated postinflammatory changes is another challenge. Although randomized trials on

244

Van Cutsem and Saltz

the optimal patient management are lacking, the experience with topical treatments (thus far of minimal value) and with oral antibiotics is growing, and experience-based treatment advice is given (105). Patient Selection While EGFR expression is evident on immunohistochemistry in 70% to 80% of CRCs, data from the BOND study with cetuximab and from the panitumumab study show that intensity of staining does not correlate with RR (84). The panitumumab study involved an initial cohort of patients who were EGFR 2þ or 3þ in 10% or more of evaluated cells, and a later cohort of patients with less intense staining. The two cohorts seem not to differ appreciably in rates of response or stable disease (89). These findings may relate to the absence of a correlation between the EGFR status of the primary tumor and that of metastases (106). More recently, Chung et al. (107) have reported a 25% rate of response in 16 patients retrospectively identified as negative by immunohistochemistry for EGFR. No one has yet established a marker that is genuinely valuable in predicting response to EGFR inhibitors in patients with CRC. For example, there appears to be no equivalent of the EGFR mutation that seems to mediate response to gefitinib in patients with lung cancer. Many open questions and challenges remain in relation to the use of the anti-VEGF and anti-EGFR antibodies in metastatic CRC. Answers are needed to optimize the outcome for patients and to more optimally use the resources. A crucial challenge is to demonstrate to the anti-EGFR antibodies cetuximab and panitumumab. Until now, large studies validating molecular markers that are useful in the prediction of response to antiEGFR antibodies are not yet available in metastatic CRC (108). The clinical studies evaluating the activity of cetuximab and panitumumab have been carried out in EGFR-expressing tumors, as determined by immunohistochemistry. The intensity of EGFR immunostaining has not been related to antitumor activity, and clinical benefit has also been noted in patients whose tumors had no EGFR immunostaining. EGFR gene mutations have not been demonstrated to play a role in the response prediction in CRC. Although it has been reported in a small study that EGFR gene copy number, as assessed by fluorescence in situ hybridization, correlates with the propensity of CRC to respond to EGFR-directed antibodies, this finding is at the moment very controversial (109). REFERENCES 1. Boyle P, Ferlay J. Cancer incidence and mortality in Europe, 2004. Ann Oncol 2005; 16(3):481–488. 2. Fong Y, Cohen A, Fortner J, et al. Liver resection for colorectal metastases. J Clin Oncol 1997; 15:938–946.

The Treatment of Metastatic Colorectal Cancer

245

3. Scheithauer W, Rosen H, Kornek GV, et al. Randomised comparison of combination chemotherapy plus supportive care with supportive care alone in patients with metastatic colorectal cancer. Br Med J 1993; 306:752–755. 4. Nordic Gastrointestinal Tumor Adjuvant Therapy Group. Expectancy or primary chemotherapy in patients with advanced asymptomatic colorectal cancer: a randomized trial. J Clin Oncol 1992; 10:904–911. 5. Ragnhammer P, Hafstro¨m L, Nygren P, Glimelius B for the SBU-group. A systematic overview of chemotherapy effects in colorectal cancer. Acta Oncol 2001; 40:282–308. 6. Glimelius B, Jacobsen A, Graf W, et al. Bolus injection (2–4 minutes) versus shortterm (10–20 minutes) infusion of 5-fluorouracil in patients with advanced colorectal cancer: a prospective randomized trial. Eur J Cancer 1998; 34:674–678. 7. Sobrero A, Aschele C, Bertino J. Fluorouracil in colorectal cancer—a tale of two drugs: implications for biochemical modulation. J Clin Oncol 1997; 15: 368–381. 8. The Meta-analysis Group in Cancer. Efficacy of intravenous continuous infusion of fluorouracil compared with bolus administration in advanced colorectal cancer. J Clin Oncol 1998; 16:301–308. 9. The Meta-analysis Group in Cancer. Toxicity of fluorouracil in patients with advanced colorectal cancer: effect of administration schedule and prognostic factors. J Clin Oncol 1998; 16:3537–3541. 10. Advanced colorectal cancer meta-analysis project. Modulation of fluorouracil by leucovorin in patients with advanced colorectal cancer: evidence in terms of response rates. J Clin Oncol 1992; 10:896–903. 11. De Gramont A, Bosset J, Milan C, et al. Randomized trial comparing monthly low-dose leucovorin-5-fluorouracil bolus with bimonthly high dose leucovorin5-fluorouracil bolus plus continuous infusion for advanced colorectal cancer: a French Intergroup study. J Clin Oncol 1997; 15:808–815. 12. Maughan T, James R, Kerr D, et al. Comparison of survival, palliation, and quality of life with three chemotherapy regimens in metastatic colorectal cancer: a multicentre randomised trial. Lancet 2002; 359:1555–1563. 13. Weh H, Zschaber R, Braumann D, et al. A randomised Phase III study comparing weekly folinic acid (FA) and high-dose 5-fluorouracil (5-FU) with monthly 5-FU/FA (days 1–5) in untreated patients with metastatic colorectal carcinoma. Onkologie 1998; 21:403–407. 14. Kohne CH, Wils J, Lorenz M, et al. Randomized Phase III study of high-dose fluorouracil given as a weekly 24-hour infusion with or without leucovorin versus bolus fluorouracil plus leucovorin in advanced colorectal cancer: European Organization of Research and Treatment of Cancer Gastrointestinal Group study 40952. J Clin Oncol 2003; 21:3721–3728. 15. Aranda E, Diaz-Rubio E, Cervantes A, et al. Randomized trial comparing monthly low-dose leucovorin and fluorouracil bolus with weekly high-dose 48-hour continuous-infusion fluorouracil for advanced colorectal cancer: a Spanish Cooperative Group for Gastrointestinal Tumor Therapy (TTD) study. Ann Oncol 1998; 9:727–731. 16. Rougier P, Bugat R, Douillard J, et al. Phase II study of irinotecan in the treatment of advanced colorectal cancer in chemotherapy-naive patients and

246

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

Van Cutsem and Saltz patients pretreated with fluorouracil-based chemotherapy. J Clin Oncol 1997; 15:251–260. Rothenberg M, Eckardt J, Kuhn J, et al. Phase II trial of irinotecan in patients with progressive or rapidly recurrent colorectal cancer. J Clin Oncol 1996; 14:1128–1135. Van Cutsem E, Cunningham D, Ten Bokkel Huinink W, et al. Clinical activity and benefit of irinotecan (CPT-11) in patients with colorectal cancer truly resistant to 5-fluorouracil (5-FU). Eur J Cancer 1999; 35:54–59. Cunningham D, Pyrho¨nen S, James R, et al. Randomised trial of irinotecan plus supportive care versus supportive care alone after fluorouracil failure for patients with metastatic colorectal cancer. Lancet 1998; 352:1413–1418. Rougier P, Van Cutsem E, Bajetta E, et al. Randomised trial of irinotecan versus fluorouracil by continuous infusion after fluorouracil failure in patients with metastatic colorectal cancer. Lancet 1998; 35:1407–1412. Machover D, Diaz-Rubio E, de Gramont A, et al. Two consecutive Phase II studies of oxaliplatin (L-OHP) for treatment of patients with advanced colorectal carcinoma who were resistant to previous treatment with fluoropyrimidines. Ann Oncol 1996; 7:95–98. Diaz-Rubio E, Sastre J, Zaniboni A, et al. Oxaliplatin as single agent in previously untreated colorectal carcinoma patients: a phase II multicentric study. Ann Oncol 1998; 9:105–108. De Gramont A, Vignoud J, Tournigand C, et al. Oxaliplatin with high-dose leucovorin and 5-fluorouracil 48-hour continuous infusion in pretreated metastatic colorectal cancer. Eur J Cancer 1997; 33:214–219. Andre´ T, Louvet C, Raymond E. Bimonthly high-dose leucovorin, 5-fluorouracil infusion and oxaliplatin (FOLFOX3) for metastatic colorectal cancer resistant to the same leucovorin and 5-fluorouracil regimen. Ann Oncol 1998; 9:1251–1253. Maindrault-Goebel F, Louvet C, Andre´ T, et al. Oxaliplatin added to the simplified bimonthly leucovorin and 5-fluorouracil regimen as second-line therapy for metastatic colorectal cancer (FOLFOX6). Eur J Cancer 1999; 35: 1338–1342. Bertheault-Cvitkovic F, Jami A, Ithzaki M, et al. Biweekly intensified ambulatory chronomodulated chemotherapy with oxaliplatin, fluorouracil, and leucovorin in patients with metastatic colorectal cancer. J Clin Oncol 1996; 14:2950–2958. Van Cutsem E, Szanto J, Roth A, et al. Evaluation of the addition of oxaliplatin to the Mayo or German 5-FU regimen in advanced refractory colorectal cancer. Proc Am Soc Clin Oncol 1999; 18:243a. Rothenberg M, Oza A, Bigelow R, et al. Superiority of oxaliplatin and fluorouracil–leucovorin compared with either therapy alone in patients with progressive colorectal cancer after irinotecan and fluorouracil–leucovorin: interim results of a phase III trial. J Clin Oncol 2003; 21:2059–2069. Cunningham D, Zalcberg J, Rath U, et al. Final results of a randomised trial 1 comparing ‘Tomudex ’ (raltitrexed) with 5-fluorouracil plus leucovorin in advanced colorectal cancer. Ann Oncol 1996; 7:961–965. Van Cutsem E, Cunningham D, Maroun J, Cervantes A, Glimelius B. Raltitrexed: current clinical status and future directions. Ann Oncol 2002; 13:513–522.

The Treatment of Metastatic Colorectal Cancer

247

31. Douillard JY, Cunningham D, Roth A, et al. Irinotecan combined with fluorouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal cancer: a multicentre randomised trial. Lancet 2000; 355:1041–1047. 32. Saltz L, Cox J, Blanke C, et al. Irinotecan plus fluorouracil and leucovorin for metastatic colorectal cancer. N Engl J Med 2000; 343:905–914. 33. Kohne CH, Van Cutsem E, Wils J, et al. Phase III study of weekly high-dose infusional fluorouracil plus folinic acid with or without irinotecan in patients with metastatic colorectal cancer: European Organisation for Research and Treatment of Cancer Gastrointestinal Group study 40986. J Clin Oncol 2005; 23:4865. 34. De Gramont A, Figer A, Seymour M, et al. Leucovorin and fluorouracil with or without oxaliplatin as first-line treatment in advanced colorectal cancer. J Clin Oncol 2000; 18:2938–2947. 35. Giacchetti S, Perpoint B, Zidani R, et al. Phase III multicenter randomized trial of oxaliplatin added to chronomodulated fluorouracil–leucovorin as first line treatment of metastatic colorectal cancer. J Clin Oncol 2000; 18:136–147. 36. Rothenberg M, Meropol N, Poplin E, Van Cutsem E, Wadler S. Mortality associated with irinotecan plus bolus fluorouracil/leucovorin: summary findings of an independent panel. J Clin Oncol 2001; 19:3801–3807. 37. Van Cutsem E, Douillard JY, Ko¨hne C. Toxicity of irinotecan in patients with colorectal cancer. N Engl J Med 2001; 345:1351–1352. 38. Tournigand C, Andre T, Achille E, et al. FOLFIRI followed by FOLFOX6 or the reverse sequence in advanced colorectal cancer: a randomized GERCOR study. J Clin Oncol 2004; 22:229–237. 39. Goldberg R, Sargent D, Morton R, et al. A randomized controlled trial of fluorouracil plus leucovorin, irinotecan, and oxaliplatin combinations in patients with previously untreated metastatic colorectal cancer. J Clin Oncol 2004; 22:23–30. 40. Colucci G, Gebbia V, Paoletti G, et al. Phase III randomized trial of FOLFIRI versus FOLFOX4 in the treatment of advanced colorectal cancer: a multicenter study of the Gruppo Oncologico Dell’Italia Meridionale. J Clin Oncol 2005; 23:4811–4814. 41. Souglakos J, Androulakis N, Syrigos K, et al. FOLFOXIRI (folinic acid, 5-fluorouracil, oxaliplatin and irinotecan) vs FOLFIRI (folinic acid, 5-fluorouracil and irinotecan) as first-line treatment in metastatic colorectal cancer (MCC): a multicentre randomised phase III trial from the Hellenic Oncology Research Group (HORG). Br J Cancer 2006; 94:798–805. 42. Falcone A, Masi R, Murr G, et al. Biweekly, oxaliplatin and infusional 5FU/ LV (FOLFOXIRI) versus FOLFIRI as first-line treatment of metastatic colorectal cancer: results of a randomized, phase III trial by the Gruppo Oncologico Nord Ovest (GONO). Gastrointestinal cancers symposium, 2006: abstract 227. (http://www.asco.org) 43. Mayer R. Oral versus intravenous fluoropyrimidines for advanced colorectal cancer: by either route, it’s all the same. J Clin Oncol 2001; 19:4093–4096. 44. Van Cutsem E, Findlay M, Osterwalder B, et al. Capecitabine, an oral fluoropyrimidine carbamate with substantial activity in advanced colorectal cancer: results of a randomized phase II study. J Clin Oncol 2000; 18:1337–1345.

248

Van Cutsem and Saltz

45. Hoff P, Ansari R, Batist G, et al. Comparison of oral capecitabine v intravenous 5fluorouracil plus leucovorin (Mayo Clinic regimen) as first-line treatment in 605 patients with metastatic colorectal cancer: results of a randomized phase III study. J Clin Oncol 2001; 19:2282–2292. 46. Van Cutsem E, Twelves C, Cassidy J, et al. Oral capecitabine compared with intravenous 5-fluorouracil plus leucovorin (Mayo Clinic regimen) in patients with metastatic colorectal cancer: results of a large phase III study. J Clin Oncol 2001; 19:4097–4106. 47. Van Cutsem E, Hoff P, Harper P, et al. Oral capecitabine versus intravenous 5fluorouracil and leucovorin: integrated efficacy data and novel analyses from two large, randomised, phase III trials. Br J Cancer 2004; 90:1190–1197. 48. Douillard JY, Hoff P, Skillings J, et al. Multicenter phase III study of uracil/ tegafur and oral leucovorin versus fluorouracil and leucovorin in patients with previously untreated metastatic colorectal cancer. J Clin Oncol 2002; 20:3605–3616. 49. Carmichael J, Popiela T, Radstone D, et al. Randomized comparative study of tegafur/uracil and oral leucovorin versus parenteral fluorouracil and leucovorin in patients with previously untreated metastatic colorectal cancer. J Clin Oncol 2002; 20:3617–3627. 50. Van Cutsem E, Sorensen J, Cassidy J, et al. International phase III study of oral eniluracil (EU) plus 5-fluorouracil (5-FU) versus intravenous 5-FU plus leucovorin in the treatment of advanced colorectal cancer. ASCO annual meeting, 2001:abstract 522. (http://www.asco.org) 51. Levin J, Schilsky R, Burris H, et al. North American phase III study of oral eniluracil (EU) plus oral 5-fluorouracil (5-FU) versus intravenous 5-FU plus leucovorin in the treatment of advanced colorectal cancer. ASCO annual meeting, 2001:abstract 523. (http://www.asco.org) 52. Cassidy J, Twelves C, Van Cutsem E, et al. First-line oral capecitabine therapy in metastatic colorectal cancer: a favourable safety profile compared with intravenous 5-fluorouracil/leucovorin. Ann Oncol 2002; 13:566–575. 53. Cassidy J, Tabernero J, Twelves C, et al. XELOX (capecitabine plus oxaliplatin): active first-line therapy for patients with metastatic colorectal cancer. J Clin Oncol 2004; 22:2084–2091. 54. Scheithauer W, Komek G, Radere M, et al. Randomized multicenter phase II trial of two different schedules of capecitabine plus oxaliplatin as first-line treatment in advanced colorectal cancer. J Clin Oncol 2003; 21:1307–1312. 55. Patt Y, Leibmann J, Diamandidis D, et al. Capecitabine plus irinotecan (Xeliri) in first-line metastatic colorectal cancer: update on a phase II trial. Eur J Cancer 2003; S93:abstract 272. 56. Tewes M, Schleucher N, Achterrath W, et al. Capecitabine and irinotecan as first-line chemotherapy in patients with metastatic colorectal cancer: results of an extended phase I study. Ann Oncol 2003; 14:1442–1448. 57. Bajetta E, Di Bartolomeo M, Mariani L, et al. Randomized multicenter phase II trial of two different schedules of irinotecan as first-line treatment in metastatic colorectal carcinoma. Cancer 2004; 100:279–287. 58. Borner M, Bernhard J, Dietrich D, et al. A randomized phase II trial of capecitabine and two different schedules of irinotecan in first-line treatment of metastatic colorectal cancer: efficacy, quality-of-life and toxicity. Ann Oncol 2005; 16:282–288.

The Treatment of Metastatic Colorectal Cancer

249

59. Rea D, Nortier J, Ten Bokkel Huinink W, et al. A phase I/II and pharmacokinetic study of irinotecan in combination with capecitabine as first-line therapy for advanced colorectal cancer. Ann Oncol 2005; 16:1123–1132. 60. Arkenau H, Schmoll H, Kubicka S, et al. Infusional 5-fluorouracil/folinic acid plus oxaliplatin (FUFOX) versus capecitabine plus oxaliplatin (CAPOX) as first line treatment of metastatic colorectal cancer: results of the safety and efficacy analysis. ASCO annual meeting, 2005:abstract 3507. (http://www.asco.org) 61. Sastre J, Massuti B, Tabernero J, et al. Preliminary results of a randomized phase III trial of the TTD group comparing capecitabine and oxaliplatin (CapeOx) vs. oxaliplatin and 5-fluorouracil in continuous infusion as first line treatment in advanced or metastatic colorectal cancer. ASCO annual meeting, 2005:abstract 3524. (http://www.asco.org) 62. Ko¨hne C, De Greve J, Bokemeyer C, et al. Capecitabine plus irinotecan versus 5-FU/FA/irinotecan
celecoxib in first line treatment of metastatic colorectal cancer. Safety results of the prospective multicenter EORTC phase III study EORTC 40015. ASCO annual meeting, 2005:abstract 3525. (http://www.asco.org) 63. Punt C, Douma J, Hondoop A, et al. Randomised study of sequential versus combination chemotherapy with capecitabine, irinotecan, and oxaliplatin in advanced colorectal cancer, an interim safety analysis. A Dutch Colorectal Cancer Group (DCCG) phase III study. Eur J Cancer Suppl 2005; 3:abstract 600. 64. Grothey A, Sargent D, Goldberg R, Schmoll HJ. Survival of patients with advanced colorectal cancer improves with the availability of fluorouracil– leucovorin, irinotecan, and oxaliplatin in the course of treatment. J Clin Oncol 2004; 22:1209–1214. 65. Ferrara N, Kerbel R. Angiogenesis as a therapeutic target. Nature 2005; 438:967–974. 66. Ferrara N. VEGF as a therapeutic target in cancer. Oncology 2005; 69(suppl 3):11–16. 67. Carmeliet P. Angiogenesis in life, disease and medicine. Nature 2005; 438: 932–936. 68. Kabbinavar F, Hurwitz H, Fehrenbacher L, et al. Phase II, randomized trial comparing bevacizumab plus fluorouracil (FU)/leucovorin (LV) with FU/LV alone in patients with metastatic colorectal cancer. J Clin Oncol 2003; 21:60–65. 69. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004; 350:2335–2342. 70. Kabbinavar F, Schulz J, McCleod M, et al. Addition of bevacizumab to bolus fluorouracil and leucovorin in first-line metastatic colorectal cancer: results of a randomized phase II trial. J Clin Oncol 2005; 23:3697–3705. 71. Kabbinavar F, Hambleton J, Mass R, Hurwitz H, Bergsland E, Sarkar S. Combined analysis of efficacy: the addition of bevacizumab to fluorouracil/leucovorin improves survival for patients with metastatic colorectal cancer. J Clin Oncol 2005; 23:3706–3712. 72. Giantonio B, Catalano P, Meropol N, et al. High-dose bevacizumab improves survival when combined with FOLFOX4 in previously treated advanced colorectal cancer: results from the Eastern Cooperative Oncology Group (ECOG) study E3200. ASCO annual meeting, 2005:abstract 2. (http://www.asco)

250

Van Cutsem and Saltz

73. Fernando N, Yu D, Morse M, et al. A phase II study of oxaliplatin, capecitabine and bevacizumab in the treatment of metastatic colorectal cancer. ASCO annual meeting, 2005:abstract 3556. (http://www.asco.org) 74. Hochster H, Hart R, Ramanathan J, et al: Results of the tree-2 cohort: safety, tolerability, and efficacy of bevacizumab added to three oxaliplatin/fluoropyrimidine regimens as first-line treatment of metastatic colorectal cancer. Gastrointestinal cancer symposium, 2006:abstract 244. (http://www.asco.org) 75. Chen H, Mooney M, Boron M, et al. A multicenter phase II trial of bevacizumab plus fluorouracil and leucovorin in patients with advanced colorectal cancer that had progressed after standard chemotherapy: an NCI treatment referral center trial (TRC-0301). J Clin Oncol 2006. In press. 76. Jubb A, Hurwitz H, Bai W, et al. Impact of vascular endothelial growth factorA expression, thrombospondin-2 expression, and microvessel density on the treatment effect of bevacizumab in metastatic colorectal cancer. J Clin Oncol 2006; 24:217–227. 77. Ince W, Jubb A, Holden S, et al. Association of k-ras, b-raf, and p53 status with the treatment effect of bevacizumab. J Natl Cancer Inst 2005; 97: 981–989. 78. Harari P. Epidermal growth factor receptor inhibitor strategies in oncology. Endocrine Related Cancer 2004; 11:689–708. 79. Mendelsohn J, Baselga J. Status of epidermal growth factor antagonists in the biology and treatment of cancer. J Clin Oncol 2003; 21:2787–2799. 80. Wells A. Molecules in focus EGF receptor. Int J Biochem Cell Biol 1999; 31:637–643. 81. Ciardello F, Tortora G. A novel approach in the treatment of cancer: targeting the epidermal growth factor receptor. Clin Cancer Res 2001; 7:2958–2970. 82. Saltz L, Rubin M, Hochster H, et al. Cetuximab (IMC-C225) plus irinotecan (CPT-11) is active in CPT-11-refractory colorectal cancer (CRC) that expresses epidermal growth factor receptor (EGFR). ASCO annual meeting, 2001:abstract 7. (http://www.asco.org) 83. Saltz L, Meropol N, Loehrer P, et al. Phase II trial of cetuximab in patients with refractory colorectal cancer that expresses the epidermal growth factor receptor. J Clin Oncol 2004; 22:1201–1208. 84. Cunningham D, Humblet Y, Siena S, et al. Cetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. N Engl J Med 2004; 351:337–345. 85. Rosenberg A, Loehrer P, Needle M, et al.: Erbitux (IMC-C225) plus weekly irinotecan (CPT-11), fluorouracil (5FU) and leucovorin (LV) in colorectal cancer (CRC) that expresses the epidermal growth factor receptor (EGFr). ASCO annual meeting, 2002:abstract 536. (http://www.asco.org) 86. Folprecht G, Lutz M, Schoffski P, et al. Cetuximab and irinotecan/5-fluorouracil/folinic acid is a safe combination for the first-line treatment of patients with epidermal growth factor receptor expressing metastatic colorectal carcinoma. Ann Oncol 2006; 17:450–456. 87. Rougier P, Raoul J-L, Van Laethem J-L, et al. Cetuximab þ FOLFIRI as firstline treatment for metastatic colorectal CA. ASCO annual meeting, 2004:abstract 3513. (http://www.asco.org)

The Treatment of Metastatic Colorectal Cancer

251

88. Diaz-Rubio E, Tabernero J, Van Cutsem E, et al: Cetuximab in combination with oxaliplatin/5-fluorouracil (5-FU)/folinic acid (FA) (FOLFOX-4) in the first-line treatment of patients with epidermal growth factor receptor (EGFR)-expressing metastatic colorectal cancer: an international phase II study. ASCO annual meeting, 2005:abstract 3535. (http://www.asco.org) 89. Malik I, Hecht JR, Patnaik A et al. Safety and efficacy of panitumumab monotherapy in patients with metastatic colorectal cancer. ASCO annual meeting, 2005:abstract 2. Proc ASCO 2005;16S:251, A3520. (http://www.asco.org) 90. Peeters M, Van Cutsem E, Siena S, et al. A phase 3, multicenter, randomized controlled trial (RCT) of panitumumab plus best supportive care (BSC) vs BSC alone in patients (pts) with metastatic colorectal cancer (mCRC). 2006:abstract. (http://www.aacr.org) 91. Berlin J, Posey J, Tchekmedyian S, et al. First line therapy of panitumumab, a fully human antibody, in combination with FOLFIRI for the treatment of metastatic colorectal cancer. Eur J Cancer Suppl 2005; 3:abstract 653. 92. Vanhoefer U, Tewes M, Rojo F, et al. Phase I study of the humanized antiepidermal growth factor receptor monoclonal antibody EMD72000 in patients with advanced solid tumors that express the epidermal growth factor receptor. J Clin Oncol 2004; 22:175–184. 93. Mackenzie MJ, Hirte HW, Glenwood G, et al. A phase II trial of ZD1839 (Iressa) 750 mg per day, an oral epidermal growth factor receptor-tyrosine kinase inhibitor, in patients with metastatic colorectal cancer. Invest New Drugs 2005; 23:165–170. 94. Rothenberg ML, Lafleur B, Washington MK, et al. Changes in epidermal growth factor receptor signaling in serum and tumor biopsies obtained from patients with progressive metastatic colorectal cancer (MCRC) treated with gefitinib (ZD1839): an Eastern Cooperative Oncology Group study. ASCO annual meeting, 2004:abstract 3000. (http://www.asco.org) 95. Fisher G, Kuo T, Cho C, et al. A phase II study of gefitinib in combination with FOLOFOX-4 (IFOX) in patients with metastatatic colorectal cancer. ASCO annual meeting, 2004:abstract 3514. (http://www.asco.org) 96. Kuo T, Cho C, Halsey J, et al. Phase II study of gefitinib, fluorouracil, leucovorin, and oxaliplatin therapy in previously treated patients with metastatic colorectal cancer. J Clin Oncol 2005; 23:5613–5619. 97. Van Cutsem E, Beale P, Delord J, Verslype C, Brennscheidt U, Feyereislova A. Tarceva (erlotinib) in combination with Xeloda and oxaliplatin in patients with metastatic colorectal cancer (CRC)—a phase I dose escalation trial. Clin Cancer Res 2003; 9:A109. 98. Tejpar S, Van Cutsem E, Gamelin E, et al. Phase 1/2a study of EKB-569, an irreversible inhibitor of epidermal growth factor receptor, in combination with 5-fluorouracil, leucovorin, and oxaliplatin (FOLFOX-4) in patients with advanced colorectal cancer (CRC). ASCO annual meeting, 2004:abstract 3579. (http://www.asco.org) 99. Ciardiello F, Bianco R, Damiano V, et al. Antiangiogenic and antitumor activity of anti-epidermal growth factor receptor C225 monoclonal antibody in combination with vascular endothelial growth factor antisense oligonucleotide in human GEO colon cancer cells. Clin Cancer Res 2000; 6:3739–3747.

252

Van Cutsem and Saltz

100. Saltz L, Lenz H, Hochster H, et al. Randomized phase II trial of cetuximab/bevacizumab/irinotecan (CBI) versus cetuximab/bevacizumab (CB) in irinotecanrefractory colorectal cancer. ASCO annual meeting, 2005:abstract 3508. (http:// www.asco.org) 101. Lenz H, Mayer R, Gold P, et al. Activity of cetuximab in patients with colorectal cancer refractory to both irinotecan and oxaliplatin. ASCO annual meeting, 2004:abstract 3510. (http://www.asco.org) 102. Van Cutsem E, Mayer R, Gold P, et al. Correlation of acne rash and tumor response with cetuximab monotherapy in patients with colorectal cancer refractory to both irinotecan and oxaliplatin. Eur J Cancer Suppl 2004; 2:abstract 279. 103. Xiong H, Rosenberg A, LoBuglio A, et al. Cetuximab, a monoclonal antibody targeting the epidermal growth factor receptor, in combination with gemcitabine for advanced pancreatic cancer: a multicenter phase II Trial. J Clin Oncol 2004; 22:2610–2616. 104. Herbst R, Arquette M, Shin D, et al. Phase II multicenter study of the epidermal growth factor receptor antibody cetuximab and cisplatin for recurrent and refractory squamous cell carcinoma of the head and neck. J Clin Oncol 2005; 23:5578–5587. 105. Segaert S, Van Cutsem E. Clinical signs, pathophysiology and management of skin toxicity during therapy with epidermal growth factor receptor inhibitors. Ann Oncol 2005; 16:1425–1433. 106. Scartozzi M, Bearzi I, Berardi R, et al. Epidermal growth factor receptor (EGFR) status in primary colorectal tumors does not correlate with EGFR expression in related metastatic sites: implications for treatment with EGFR-targeted monoclonal antibodies. J Clin Oncol 2004; 22:4772–4778. 107. Chung K, Shia J, Kemeny N, et al. Cetuximab shows activity in colorectal cancer patients with tumors that do not express the epidermal growth factor receptor by immunohistochemistry. J Clin Oncol 2005; 23:1803–1810. 108. Vallbohmer D, Zhang W, Gordon M, et al. Molecular determinants of cetuximab efficacy. J Clin Oncol 2005; 23:3536–3544. 109. Moroni M, Veronese S, Benvenuti S, et al. Gene copy number for epidermal growth factor receptor (EGFR) and clinical response to antiEGFR treatment in colorectal cancer: a cohort study. Lancet Oncol 2005; 6:279–286.

10 Hepatic Directed Therapy Gregory D. Leonard and Nancy E. Kemeny Department of Gastrointestinal Oncology, Memorial Sloan-Kettering Cancer Center, New York, New York, U.S.A.

INTRODUCTION Colorectal cancer is the third most common cause of cancer in men and in women with a total of 146,940 new cases estimated in 2004 (1). It is also the third most common cause of cancer death in each sex, with 56,730 deaths estimated in 2004. Approximately 25% of patients have synchronous metastatic disease, and more than half of the patients with localized disease will develop metastatic disease (2). The liver is usually the first and the most common site of metastatic disease although low rectal primary tumors can skip the liver and metastasize to the lung. The optimal management of patients with metastatic disease located only in the liver is surgery. This select group of patients can have a prolonged survival with five-year survivals of up to 30% (2). However, less than 25% of patients are suitable for this liver resection. In unresectable patients, chemotherapy is the treatment of choice. Traditionally, chemotherapy has been considered a palliative approach, but in recent years advances in chemotherapy and biologic agents have significantly improved outcomes in patients with median survivals of 15–20 months (3–5). In some patients with unresectable metastatic disease, chemotherapy has been used as a neoadjuvant approach and has rendered patients resectable with favorable survival results (6). For 40 years, fluorouracil (5-FU) systemic chemotherapy was the treatment of choice for metastatic colorectal cancer. With the advent of

253

254

Leonard and Kemeny

irinotecan and oxaliplatin the standard of care has recently changed and continues to evolve with the development of biologic agents. In parallel with the development of new drugs, extensive investigation has also focused on the optimal schedule and route of administration of chemotherapy. As most colorectal metastases are located in the liver, it is intuitive that maximizing the cytotoxic effect of chemotherapy on the liver could result in a significant benefit for patients. In this chapter, we will analyze some of the methods by which chemotherapy can be directed towards the liver. Most emphasis will be on hepatic arterial infusion (HAI) of chemotherapy, which is the most extensively studied area in hepatic directed therapy. We will also briefly discuss other hepatic directed therapies such as portal vein infusion and isolated perfusion chemotherapy. HEPATIC ARTERIAL INFUSION CHEMOTHERAPY History and Rationale Colorectal metastases usually occur via hematogenous spread, initially via the portal circulation to the liver, then to the lungs and then to other organs (7). Early treatment of liver metastases therefore may prevent the sequential progression of metastases to extrahepatic sites. Liver metastases that grow beyond 2–3 mm are dependent on their blood supply from the hepatic artery, whereas smaller lesions and the normal liver parenchyma are supplied by the portal vein (8,9). Proof of this concept was obtained by injecting radiolabeled fluorodeoxyuridine (FUDR) into the hepatic artery or portal vein in patients with hepatic metastases (10). Liver and tumor biopsies were obtained directly after FUDR infusion. Liver parenchyma levels of FUDR were similar following HAI or portal vein infusion. Mean tumor levels of FUDR, however, were higher with HAI compared to portal vein infusion (p < 0.01). This study suggested that chemotherapy given via HAI might allow maximal concentration of the drug in the tumor. To confirm this, another study assessed tumor response to FUDR when given via HAI or portal vein infusion (11). HAI of FUDR resulted in a response in four of eight patients, but there was no response in 11 patients when FUDR was given via portal vein infusion. Nine patients who progressed while receiving portal vein infusion were then given the same drug by HAI, and three responded. Having established that the natural history of colorectal metastases could be disrupted using hepatic directed therapy and that the maximal effect on tumors could be achieved with HAI, it was then necessary to determine the optimal chemotherapy. Maximal drug exposure to tumor sites is achieved when the delivery route has low exchange rates and when the drug has a high total body clearance (12). As the hepatic artery has a high exchange rate, drugs with high extraction rates in the liver due to first pass metabolism are favored. In theory, drugs with high extraction rates in the

Hepatic Directed Therapy

255

liver should have maximal cytotoxic effect in the liver but with limited systemic toxicity. FUDR is an example of such a drug with 94% to 99% extraction in the liver during the first pass when given by HAI compared to 19% to 55% with fluorouracil (13,14). This estimated increased exposure to FUDR is 100–400-fold higher when given by HAI compared to the systemic route, whereas the increase is only 5–10-fold for 5-FU. Other drugs such as etoposide have not demonstrated an increased extraction rate when given by HAI (15). Evidence from trials comparing fluoropyrimidines given via an infusional or bolus approach found the infusional route to be more efficacious and further favored the use of HAI, which delivers chemotherapy over a protracted period of time (16). One of the first reports on the use of HAI chemotherapy was by Sullivan et al. in 1959 (17). This group subsequently reported their experience on 21 patients (nine patients with colon and rectum cancer) with primary and metastatic cancers of the liver and biliary system treated with HAI chemotherapy consisting of methotrexate, 5-fluorouracil, or 5-fluoro, 20 -deoxyuridine (18). The hepatic artery catheter was inserted at laparotomy and the correct positioning was determined by injection of a fluorescent dye and ultraviolet illumination to visualize the distribution of the dye. From 16 patients who received adequate treatment there were 13 objective responses and 10 patients had a clinical benefit. A similar study of 24 patients (13 patients with colon cancer) treated with HAI chemotherapy showed regression of metastases in nine of 16 evaluable patients (19). In this trial, the hepatic artery was cannulated via the brachial artery, a technique that was developed in the 1950s (20). These early studies confirmed the rationale that HAI chemotherapy could be an effective treatment for liver metastases. Subsequent studies have sought to address the optimal chemotherapeutics, the method of choice for cannulating the hepatic artery, and most importantly limiting toxicity so that single institution results can be replicated in the community. Hepatic Artery Pump Initial studies evaluating HAI therapy used external infusion pumps, often placed percutaneously, but these pumps had high rates of bleeding, thrombosis, catheter sepsis, and dislodgement (21). The use of a subcutaneously placed port was more convenient for the patient but was still associated with multiple complications including bleeding and clots of the port, hence not allowing treatment (22). Many of these complications were avoided with the development of an implantable pump. These pumps are inserted at laparotomy and can be accessed subcutaneously. The implantable pump was first tested in the early 1970s (23). It was first used in humans to deliver heparin via the hepatic artery for chronic anticoagulation and in 1980 was used to administer chemotherapy to five patients with primary or metastatic carcinoma of the liver (24,25). Comparison of the implantable pump (n ¼ 15)

256

Leonard and Kemeny

to a surgically placed catheter (n ¼ 12) or a percutaneously placed catheter (n ¼ 20) demonstrated mean chemotherapy durations of 115, 31, and 25 days, respectively (26). Another study compared the use of an implantable pump in 70 patients with a port system in 110 patients, and found technical complications in 41% and 62%, respectively, leading to an improvement in median complication-free survival in favor of the implantable pump ( p < 0.001) (27). Although implantable pumps appear to be superior to other forms of hepatic artery infusion devices, many studies used external infusion devices and ports; hence results of trials using these latter devices need to be interpreted with caution due to device-related complications and their limitations. The implantable pump is made from titanium, is slightly larger than a pacemaker, and consists of two chambers separated by metal bellows. One chamber contains a fluorocarbon liquid that is converted to gas by body heat and causes continuous pressure on the other chamber that contains the drug. The drug chamber releases the drug at a constant flow rate due to pressure from the gas-filled chamber and can be refilled percutaneously with about 30–50 mL of drug usually every two weeks. Some pumps also contain side ports, which bypass the reservoir and infuse directly into the hepatic artery. Pump Insertion Patients with portal vein thrombosis who therefore rely on the hepatic artery to supply the normal liver parenchyma are not candidates for HAI. Patients with extrahepatic disease are also unsuitable unless the extrahepatic disease is being concomitantly treated with systemic therapy. Relative contraindications are those with poor performance status and extensive liver metastases, as their perioperative morbidity and mortality are high (28). Prior to catheter insertion, patients require a hepatic arteriogram to rule out arterial anomalies. The common hepatic artery is a branch of the celiac axis, and the first branch of the hepatic artery is the gastroduodenal artery. Early autopsy studies revealed variable anatomy in many patients, and a recent study of arteriograms in 100 patients found normal anatomy in only 50% of patients (29,30). At laparotomy the surgeon should confirm the absence of extrahepatic disease, perform a cholecystectomy to prevent drug-related cholecystitis, and then place the catheter in the gastroduodenal artery after ligating any accessory or branching vessels. The catheter is placed in the gastroduodenal artery, where it emerges from the hepatic artery, so that the risk of occlusion of the hepatic artery is minimal (Fig. 1) (31). After insertion of the pump, fluorescein is injected via the side port and an ultraviolet light is used to determine adequate liver perfusion. Postoperatively, a similar procedure is performed using 99m-technetium macroaggregated albumin perfusion scan, which should be compared to a preoperative study. Extrahepatic perfusion on the technetium scan should be confirmed via angiography

Hepatic Directed Therapy

257

Figure 1 Diagram of hepatic arterial infusion pump.

with embolization of appropriate vessels (32). Recently, magnetic resonance imaging has been shown to be useful for evaluating HAI liver perfusion (33). The success of hepatic arterial infusion is dependent on the experience of the surgeon. In one study of 70 patients, inexperienced surgeons had a complication rate of 37% compared to 7% with experienced surgeons (p < 0.01) (34). Some institutions have demonstrated the ability to place hepatic artery catheters via laparoscopy (35). Complications of Pump Therapy Early complications from HAI occur due to technical difficulties with the insertion of the pump or the pump itself but are rare. Toxicity seen with the systemic administration of fluoropyrimidines such as diarrhea, mucositis, skin rash, and myelosuppression are rarely seen with HAI of fluoropyrimidines. Hepatobiliary toxicity is the most common toxicity with most patients exhibiting reversible liver enzyme elevations. A dose-limiting effect of HAI FUDR is biliary sclerosis, which is not surprising as the vascular supply to the bile ducts is from the hepatic artery (36). Biliary toxicity is manifest with elevated alkaline phosphatase and bilirubin and other serological markers of cholangitis. The etiology may be from an arteritis of vessels supplying

258

Leonard and Kemeny

the bile ducts or from chemotherapy-related biliary basement membrane damage, bile leakage, and consequent sclerosis (37). Acute and chronic inflammatory changes involving portal triads are evident in the livers of patients post HAI (38). In one study eight of 46 patients developed biliary strictures with HAI of FUDR, and in another study all 35 patients developed significant alkaline phosphatase elevation (39). The clinical picture mirrors that of sclerosing cholangitis, and endoscopic retrograde cholangiopancreaticogram can confirm the diagnosis radiologically. Dexamethasone in the pump with heparinized saline and ursodeoxycholic acid orally may help. Patients are managed in a similar fashion to those with sclerosing cholangitis, including the use of stents in patients with focal lesions. Discontinuation of HAI chemotherapy with increases in liver function tests is necessary to prevent this problem (40). In most patients, if hepatobiliary toxicity is noted early, treatment breaks and dose reductions should be used. If elevation of liver function occurs again, drug should be discontinued. The optimal treatment of hepatobiliary toxicity is prevention. Close observation of liver enzymes is essential but recent studies have examined the use of dexamethasone. Kemeny et al. (41) randomized 50 patients with metastatic colorectal cancer to receive FUDR with or without dexamethasone. Patients receiving dexamethasone had bilirubin elevation ( >3 ng/dL) in only 9% of patients compared to 30% of patients who did not receive dexamethasone ( p ¼ 0.07). Other investigators have used circadian-modified infusion schedules or have alternated FUDR with 5-FU and have also demonstrated improvements in toxicity profiles (42–44). Gastritis and peptic ulcer disease may also occur, predominantly due to inadvertent gastrointestinal perfusion of chemotherapy and can be avoided with close attention to perioperative perfusion scans and ligation or embolization of vessels as needed (45). Nonrandomized Trials of HAI Chemotherapy Phase II trials using HAI chemotherapy in patients with metastatic colorectal cancer produced high response rates and median survivals ranging from 12 to 26 months (Table 1). These results were even more impressive considering that many patients in these studies had been treated with prior chemotherapy. The results of these trials provided the basis for further investigation of HAI chemotherapy in randomized phase III trials. Randomized Trials of HAI Chemotherapy HAI for Metastatic Disease Ten prospective randomized phase III trials have been reported, which compare HAI chemotherapy to systemic chemotherapy in patients with unresectable liver metastases (Table 2). Although in most cases these trials

Hepatic Directed Therapy

259

Table 1 Nonrandomized Trials of Hepatic Arterial Infusion of Fluoropyrimidine Chemotherapy

Trial Oberfield et al. (46) Niederhuber et al. (47) Balch et al. (48) Kemeny et al. (49) Shepard et al. (50) Cohen et al. (51) Weiss et al. (52) Schwartz et al. (53) Johnson et al. (54) Howell et al. (55) Arai et al. (56)

Number of patients Chemotherapy 48 50 50 62 53 18 17 25 40 40 32

FUDR FUDR FUDR FUDR FUDR FUDR FUDR FUDR FUDR 5-FU 5-FU

Objective response (%)

Median survival (mos)

75 83 81 78 32 72 29 15 47 46 78

3.6–8.5 25 26 24.8 17 2.5–21 13 2–34 12 19 25.8

Abbreviations: FUDR, fluoro pyrimidine; 5-FU, 5-fluorouracil.

demonstrate HAI to be a favorable approach, the benefits are often underestimated due to study design or outdated treatment approaches. For example, small patient numbers, crossover design, extrahepatic disease, the use of subcutaneous ports as opposed to implantable pumps, and excessive technical difficulties have in some cases prevented a statistically significant survival advantage being evident for HAI. Others would argue that modern systemic chemotherapy is superior to the control arms of these studies although this could be obviated by the concomitant use of systemic chemotherapy with HAI. Also none of the early trials using HAI incorporated dexamethasone, which is now known to improve response rates and overall survival (40). The first randomized trial to be reported, by Kemeny et al. (57), was a single institution study from Memorial Sloan Kettering Cancer Center (MSKCC). From 99 patients, a statistically significant improvement in response rate (50% vs. 20%) was seen for HAI FUDR compared to intravenous chemotherapy ( p ¼ 0.001). The median survival (17 vs. 12 months) was not statistically significant although this might have been due to the high rate (60%) of crossover from intravenous to HAI chemotherapy. In patients who did not cross over, the median survival was eight months compared to 18 months for those who crossed over to HAI ( p ¼ 0.04). In the same year, the National Cancer Institute reported results from a similar trial, but once again the benefit of HAI may have been underestimated as 34% of the HAI group did not receive chemotherapy and 38% of this group had extrahepatic disease due to positive portal lymph nodes (58). Including all patients, twoyear survival was not statistically significant in favor of HAI (22% vs. 15%; p ¼ 0.27). When patients with hepatic-only disease were evaluated, the

54,57 145 68

FU/LV FU/LV FU/LV

FUDR, 5-FU FU/LV FUDR

39 81 51

FU/LV FU/BSC FU/BSC

FUDR FUDR FUDR

48 32 67 31

HAI

FUDR FUDR FUDR FU

IV

FUDR FUDR FUDR FUDR

HAI

145 67

57

35 82 49

51 32 76 10

IV

Number of patients

91 75 87

38 87

103 50 20

85 87 96 69,70

94 92 86 100

IV (%)

94 66 75 100

HAI (%)

Received assigned treatment

22 48

43,45

48 44 —

50 62 42 55

HAI (%)

19 25

20

21 9 —

20 17 10 20

IV (%)

Responses (CRþPR)

17.6 14.8 20

14.7 24.4a

10.5 11 7.5

12.6 15a 13.5a 12.7, 18.7

12 12 15.8 11.6

IV 17 17b 16.5 13.8

HAI

Median survival (mos)

b

Statistically significant median survival. Median survival calculated from Kaplan Meir survival curve published in original citation (excludes patients with extrahepatic disease). Abbreviations: BSC, best supportive care; CR, complete response; PR, partial response; HAI, hepatic arterial infusion; IV, intravenous; FUDR, floxuridine; FU, fluorouracil; LV, leucovorin.

a

Kemeny et al. (57) Chang et al. (58) Hohn et al. (59) Kemeny et al. (60) and Wagman et al. (61) Martin et al. (62) Rougier et al. (63) Allen-Marsh et al. (64) Lorenz and Muller (65) Kerr et al. (66) Kemeny et al. (67)

Trial

Chemotherapy

Table 2 Randomized Trials of Hepatic Arterial Infusion Chemotherapy for Unresectable Colorectal Liver Metastases

260 Leonard and Kemeny

Hepatic Directed Therapy

261

two-year survival was statistically significant (47% vs. 13%) ( p ¼ 0.03). The Northern California Oncology Group trial demonstrated an improvement in time to progression ( p
0.001), but possibly due to crossover of patients receiving intravenous treatment (43%) and extrahepatic disease there was no difference in survival (59). Similarly, the City of Hope study and the North Central Cancer Treatment Group (NCCTG) were negative studies but incorporated crossover design and the presence of extrahepatic disease, respectively (60–62). The French study was positive but was criticized as the control arm only used intravenous chemotherapy when patients were symptomatic (63). A similar trial design was used in the English study, which was also positive, but only 20% of patients in the systemic group in the English study received intravenous 5-FU (64). The English study also assessed quality of life and showed improvements in the HAI arm for physical symptoms ( p ¼ 0.04) and anxiety ( p ¼ 0.04). Two meta-analyses of the first seven trials were published. As many of the trials were underpowered, it was hoped that combining the data from these trials using a meta-analysis would provide more definite conclusions on the efficacy of HAI, although the presence of extrahepatic disease or inadequate chemotherapy administration could not be accounted for. One meta-analysis demonstrated a 27% relative survival advantage with HAI ( p < 0.001), although when the trials that incorporated best supportive care in the control arms were excluded the survival advantage of 19% was no longer statistically significant (68). The other meta-analyses evaluated six trials, but excluded the trials with best supportive care as the control arms, and found an improvement in one- and two-year survival (10% and 6%, respectively) although only the one-year survival was statistically significant ( p ¼ 0.041) (69). A subgroup analysis of trials that did not allow crossover demonstrated one- and two-year survivals of 19% and 9%, respectively, both of which were statistically significant. Three subsequent studies have been reported. The German Cooperative Group consisted of three arms with HAI FUDR, HAI 5-FU/LV, and intravenous 5-FU/LV. Response rates were doubled with the HAI arms, but there was no difference in survival between the arms although only 70% of patients in the HAI arms received therapy and 51% of patients crossed over (65). More patients experienced toxicity with 5-FU compared to FUDR. For example, stomatitis was seen in 8%, 75%, and 64%, and diarrhea in 0%, 11%, and 11% of the HAI FUDR, HAI 5-FU/LV, and 5-FU/ LV arms, respectively. The Medical Research Council/European Organization for the Research and Treatment of Cancer trial also used HAI 5-FU/ LV, and unlike other trials used the biweekly formulation of bolus and infusional 5-FU/LV in the intravenous arm (66). There was no survival advantage to HAI although 37% of patients did not start HAI therapy and another 29% had to stop before receiving the assigned treatment due to catheter failure. Patients in this trial had a subcutaneous port rather than

262

Leonard and Kemeny

an implantable pump, which may have contributed to the high complication rate (36%) in the HAI arm. The Cancer and Leukemia Group B 9481 trial is the most recent trial reported in abstract form (67). This trial differs from others in that it includes the use of dexamethasone in the HAI arm and there was no crossover. The response rate (48% vs. 25%; p ¼ 0.01) was higher in the HAI arm as was the median survival (24.4 vs. 20 months; p ¼ 0.0034). Interestingly, the time to hepatic progression was better in the HAI arm (9.8 vs. 7.3 months; p ¼ 0.034), but the time to extrahepatic progression was better in the systemic arm (7.7 vs. 14.8 months; p < 0.029). This result appears to support the use of HAI and systemic chemotherapy in order to optimize local and systemic control of the cancer. Due to conflicting results, there has been much controversy regarding the role of HAI chemotherapy for patients with colorectal liver metastases. Despite this, one cost-effective analysis comparing HAI, systemic chemotherapy, and best supportive care has shown HAI to be more favorable than systemic therapy (70). Close examination of the data recognizes multiple factors that may impair results in the HAI arms, and when taking this into account it appears that HAI chemotherapy is beneficial in many trials. Current investigation is focusing on the combination of HAI and systemic chemotherapy, which will be discussed later and may prove to be the optimal approach. Adjuvant HAI Post-resection of Liver Metastases The best possibility of achieving prolonged survival with metastatic colorectal cancer is rendering the patient free of disease, if the patient has localized disease amenable to resection. Nonrandomized studies have shown markedly superior outcomes for patients who have complete surgical resection of their liver metastases, and it is rare for patients to survive beyond three years without the use of surgery (71–74). A retrospective review of 1001 liver resections in patients with metastatic colorectal cancer at MSKCC demonstrated a five-year survival of 37% (75). A similar study from Europe of 1568 patients documented a five-year survival of 28% (76). In general five-year survival rates are around 30% (77). Moreover, repeat liver resection in suitable patients can result in a prolonged survival (78). The likelihood of survival is dependent on a number of features including the size and number of liver metastases, and scoring systems have been devised to help predict the outcome of patients (75,76,79). Despite careful selection of patients, most will recur. Three trials have evaluated the role of adjuvant therapy post–liver resection. O’Connell et al. treated 26 patients with 5-FU and semustine post–liver resection and compared them to 26 patients with closely matched prognostic factors treated with liver resection alone. There was no significant survival advantage in the adjuvant chemotherapy group (80). Portier et al. (81) randomized 173 patients to intravenous 5-FU or no chemotherapy post–liver resection and demonstrated a five-year survival of

Hepatic Directed Therapy

263

50% and 40%, respectively ( p ¼ 0.15). A similar trial by Langer et al. (82) reported a four-year survival of 57% and 47%, respectively, with a median survival of 53 months and 43 months, respectively ( p ¼ 0.39). A number of trials have evaluated the role of adjuvant HAI chemotherapy (Table 3). The German Cooperative Group was the first group to perform a large multi-institutional study of 226 patients comparing liver resection and HAI 5-FU/LV to resection alone (84). This study has been criticized, as chemotherapy was administered by surgeons and only 30% of the HAI group completed treatment. Not surprisingly with an intentionto-treat analysis there was no difference in median survival between the groups. However, in an analysis comparing only treated patients to control the survival was improved in the HAI arm (20 vs. 12.6 months). An MSKCC study randomized 156 patients post–liver resection to HAI FUDR and dexamethasone combined with systemic 5-FU/LV to systemic 5-FU/ LV alone (85). The trial was statistically powered only to demonstrate a difference between the two groups in their two-year survival. In this study, only 8% of the HAI group did not receive the assigned therapy. There was a statistically significant increase in the two-year survival (86% vs. 72%) in favor of the HAI group ( p ¼ 0.03). The hepatic-free survival was also improved ( p < 0.001). Although more patients were hospitalized due to treatment-related complications in the HAI group (39% vs. 22%; p ¼ 0.02), there was no significant difference in therapy-related deaths. Updated results with a median follow-up of 10 years demonstrate a 10 year survival of 41% in the HAI+SYS group and 27% in the systemic alone group. The progression free survival is 31 vs. 17.2 months for the HAI group vs. SYS alone group (p¼0.02). The trial conducted by the Eastern Cooperative Group and the Southwestern Oncology Group randomized 110 patients prior to liver resection to postoperative chemotherapy with HAI FUDR combined with systemic infusional 5-FU or no treatment (87). This trial was powered to detect a difference in disease-free survival but not overall survival. Of the 109 intent-to-treat population, only 75 received treatment and were assessable for analysis. For assessable patients, the four-year disease-free survival was improved in the HAI group (46% vs. 25%; p ¼ 0.03) and there was a trend towards improvement in four-year overall survival (62% vs. 53%; p ¼ 0.6). For the intent-to-treat population, the four-year survival was 37% and 49% for the HAI and no-treatment groups, respectively ( p ¼ 0.19). Preliminary results from a recent trial by O’Connell et al. (88) of 49 patients also demonstrate an improvement in five-year survival with adjuvant HAI (31% vs. 15%). As with the unresectable metastatic trials, the results of these adjuvant trials appear to be conflicting. However, closer scrutiny shows that in these trials the primary endpoints supporting the use of HAI that were appropriately powered were achieved with the exception of the German trial, where too many patients were inadequately treated to provide meaningful results.

FUDR Mit, Carb, Farm, FUb FU/LV FUDR þ IV FU/LV 5-FU þ oral 5-FU FUDR þ IV FU/LV FUDR alt with IV FU/LV

HAI

Chemotherapy IV 6 20 113 82 10 30 13

HAI 5 20 113 74 9 45 36

IV None None None FU/LV Oral 5-FU FU inf None — — 67c 90a 11 67a,d 29e

HAI (%) — — 63 60 60 43 26e

IV (%)

2-year HPFS

80 — 62% 86% 78% 62% 53%e

HAI

67 — 65% 72%a 50% 53% 23%e

IV

2-year overall survival

37.3 20a 44.8 72.2 62.6 63.7 31%f

HAI

28 11 39.7 59.3 39.9 49.7 15%f

IV

Median survival (mos)

b

Statistically significant. HAI therapy consisted of multidrug chemotherapy urographin, mitomycin, carboplastin, farmorubicin, leukovorin and immunotherapy interferon, interleukin, lipoiodol, and urographin. The survival quoted in this trial is the mean survival. c 1.5-year HPFS. d Four-year HPFS. e Three-year HPFS and survival. f Five-year survival. Abbreviations: HAI, hepatic arterial infusion; IV, intravenous; FUDR, floxuridine; FU, fluorouracil; LV, leucovorin; inf, infusional; HPFS, hepatic progression-free survival.

a

Wagman et al. (61) Lygidakis et al. (83) Lorenz et al. (84) Kemeny et al. (85) Tono et al. (86) Kemeny et al. (87) O’Connell et al. (88)

Trial

Number of patients

Table 3 Randomized Trials of Adjuvant Hepatic Arterial Infusion Chemotherapy Post-Resection of Colorectal Liver Metastases

264 Leonard and Kemeny

Hepatic Directed Therapy

265

Adjuvant HAI for Stage II or III Cancers Two trials have also evaluated the role of HAI in the adjuvant setting for patients with early-stage colon cancer. Sadahiro et al. (89) randomized 316 patients with stage II or stage III colon cancer to surgery with or without HAI 5-FU and dexamethasone. Adjuvant systemic chemotherapy given off protocol using an oral fluouropyrimidine was given in 81% of the HAI arm and 78% of the control arm ( p ¼ 0.57). Using an intent-to-treat analysis, 150 patients assigned to HAI (including 31 patients where the catheter was not inserted) had a significantly better five-year disease-free survival compared to the 155 patients in the control arm (86% vs. 68%; p < 0.001). Subset analyses revealed no benefit for stage II but a significant benefit for stage III cancers ( p < 0.001). The five-year overall survival was also improved with HAI (89% vs. 76%; p < 0.001) but the benefit was only evident in the stage III cancers ( p ¼ 0.001). In the 119 patients who received the assigned HAI, only one patient developed a liver metastases compared to 15% of patients in the control arm. Another trial by Ota et al. (90) evaluated 90 patients, postresection for Dukes’ C colon, who were treated with HAI 5-FU and compared them to a control group of 62 nontreated patients. The five-year overall survival for the HAI group was 84% and 65% for the non-HAI group ( p ¼ 0.0369). Further trials are required to validate these provocative results. HAI with New Chemotherapy Agents In prior decades, the median survival for patients with metastatic colorectal cancer was approximately one year, but in recent years a number of phase III clinical trials have reported median survivals approaching 20 months (3,5). Preliminary results using biologic agents after or in combination with chemotherapy have further added to the prolonged survival of patients (4,91–93). Only a few studies of HAI chemotherapy that incorporate the use of these new chemotherapy agents have been performed but appear encouraging (Table 4). Phase II trials using HAI irinotecan have reported conflicting results. Van Riel et al. (94) treated 25 patients with previously treated metastatic colorectal cancer with a five-day continuous infusion of HAI of irinotecan every three weeks at a dose of 20 mg/m2/day. From 22 assessable patients, 14% had a partial response and 40% had stable disease. The median survival was 8.1 months. These disappointing results may have been due to the lack of hepatic extraction of the drug despite the increased conversion of irinotecan to its active metabolite SN38 (112). However, Fiorentini et al. (95) treated 12 patients also with pretreated metastatic disease and demonstrated a partial response in 33% and stable disease in 25%. Similarly, Vogl et al. (96) noted a partial response of 33% in nine patients. A number of trials have evaluated HAI oxaliplatin combinations. Kern et al. (97) demonstrated an objective response rate of 59% using HAI oxaliplatin and 5-FU and folinic acid. Ducreux et al. (98), using a similar

Irinotecan Irinotecan Irinotecan Oxaliplatin/5-FU/LV Oxaliplatin þ systemic 5-FU/LV Oxaliplatin/5-FU/LV Oxaliplatin/Irinotecan/5-FU Oxaliplatin þ systemic 5-FU/LV Oxaliplatin þ systemic 5-FU/LV Oxaliplatin Oxaliplatin/5-FU/LV/mitomycin Cisplatin/mitomycin Cisplatin/5-FU Pirarubicin Mitomycin Mitomycin/5-FU Mitomycin/mitoxantrone/5-FU/LV Gemcitabine

Chemotherapy 14 33 33 35 76 37 39 17 69 46 80 35 59 33 15 39 62 —

PR (%) 0 0 0 24 9 0 0 0 — 0 0 0 0 — 4 7 4 —

CR (%) 41 25 33 22 14 37 33 75 — 21 0 33 27 — — 14 26 —

SD (%)

b

Two-year survival. Median progression-free survival. Abbreviations: PR, partial response; CR, complete response; SD, stable disease; FU, fluorouracil; LV, leucovorin; N, number.

a

22 12 9 18 21 11 18 12 16 15 5 44 22

Van Riel et al. (94) Fiorentini et al. (95) Vogl et al. (96) Kern et al. (97) Ducreux et al. (98) Pennucci et al. (99) Bouchahda et al. (100) Tomirotti et al. (101) Boige et al. (102) Mancuso et al. (103) Guthoff et al. (104) Fazio et al. (105) Kohnoe et al. (106) Rougier et al. (107) Makela et al. (108) Liu et al. (109) Link et al. (110) Sharma et al. (111) 46 56 50 10

N

Trial

Table 4 Selection of Trials Using Hepatic Arterial Infusion of New Chemotherapy Agents

8.1 13 — — 63%a — >12 — 9.4b 19 — 11.7 17 18þ 15 15 27.4 —

Median survival (mos)

266 Leonard and Kemeny

Hepatic Directed Therapy

267

regimen in a phase II study, found an objective response rate of 79% with 3 of 14 patients undergoing complete resection of their metastases after response to therapy and a six-month survival of 86%. Other trials have recorded response rates of 13% to 69% for HAI oxaliplatin combined with other chemotherapies, often given via the intravenous route (99–102). One trial assessed HAI of single-agent oxaliplatin in 15 patients and demonstrated a partial response of 46%, which is impressive considering that single-agent oxaliplatin given systemically has low response rates compared to the combination of oxaliplatin and 5-FU/LV (103,113). A study of HAI oxaliplatin, 5-FU/LV, and mitomycin in five patients documented a response rate of 80% (104). A number of other anticancer agents known to have minimal if any activity when given by the intravenous route as a single agent in colorectal cancer have been tested using HAI. For example, in preclinical and clinical studies cisplatin has not been shown to have significant activity in colorectal cancer (114). However, two trials of HAI cisplatin combinations demonstrated response rates of 35% and 68% and a median survival of 17 months in one of the trials (105,106). Similarly HAI pirarubicin (an Adriamycin derivative) has shown a response rate of 33%, while HAI mitomycin has produced a response rate of 20% as a single agent and 66% in combination with other chemotherapy given by HAI (107–110). Some studies have used HAI to administer nonchemotherapeutic agents such as an oncolytic adenovirus, interleukin-2, and activated leukocytes (111,115–117). HAI Chemotherapy Combined with Systemic Chemotherapy Although HAI chemotherapy has demonstrated response rates about double that of systemic chemotherapy, there is concern that despite good regional control of the liver with HAI, micrometastases located outside of the liver would continue to grow unless treated by systemic chemotherapy. This theory appears to be borne out in analysis of the data from the Cancer and Leukemia Group B study where hepatic metastases were seen less in the HAI arm compared to the systemic arm, whereas systemic metastases were seen more in the HAI arm compared to the systemic arm. Combining HAI with systemic chemotherapy should in theory be the optimal approach. Trials using HAI and systemic 5-FU/LV have demonstrated response rates of 17% to 86% and median survivals of 9 to 26 months (Table 5). The trials with lower response rates and median survivals used an intention-totreat analysis, which may have obscured some of the potential benefits for the combination of HAI and systemic therapy. Two trials were performed using HAI chemotherapy and irinotecan. Zelek et al. (124) performed a phase II trial using HAI pirarubicin and systemic irinotecan/5-FU/LV in 31 patients with unresectable liver metastases from colorectal cancer and demonstrated, using an intention-to-treat analysis, a response rate of 48%

Oxaliplatin Oxaliplatin Oxaliplatin FUDR 5-FU/epirubicin/mitomycin FUDR Cisplatin (CIþbolus) FUDR Pirarubicin Pirarubicin FUDRþdexamethasone FUDR

HAI 5-FU/LV 5-FU/LV 5-FU/LV Oxaliplatin/FU/LV 5-FU/LV 5-FU/LV CI 5-FU 5-FU/LV 5-FU/LV Irinotecan/5-FU/LV Irinotecan Oxaliplatin  CPT-11

Systemic 28 12 16 10 20 42 123 40 75 31 38 39

N 64 17 69 70 50 41 52 62 32 48 74 89

Objective response (%)

27 – – 9 18 13 18 18 19 20 20 37

Median survival (mos)

Abbreviations: HAI, hepatic arterial infusion; FUDR, fluoro pyrimidine; FU, fluorouracil; LV, leucovorin; CI, continuous infusion; N, number.

Ducreux et al. (98) Tomirotti et al. (101) Boige et al. (102) Pancera et al. (118) Cantore et al. (119) Copur et al. (120) Mancini et al. (121) O’Connell et al. (122) Fallik et al. (123) Zelek et al. (124) Kemeny et al. (125) Kemeny et al. (126)

Trial

Chemotherapy

Table 5 Selection of Trials Using Hepatic Arterial Infusion Combined with Systemic Therapy for Metastatic Disease

268 Leonard and Kemeny

Hepatic Directed Therapy

269

Table 6 Selection of Trials Using Hepatic Arterial Infusion Chemotherapy and Systemic Chemotherapy in the Adjuvant Setting Post-Resection of Liver Metastases Chemotherapy Trial Kemeny et al. (85) Tono et al. (86) Kemeny et al. (87) Kemeny et al. (127)

HAI

Systemic

Median survival (mos)

FUDR 5-FU FUDR FUDR

5-FU/LV Oral 5-FU 5-FU/LV Irinotecan

72.2 62.6 63.7 89%a

a Two-year survival. Abbreviations: HAI, hepatic arterial infusion, FUDR, floxuridine; FU, fluorouracil; LV, leucovorin.

and a median survival of 20.5 months, and one-third of patients were rendered resectable. The chemotherapy was well tolerated with no toxic deaths. The most common toxicities were neutropenia (48% of cycles had grade 3/4), diarrhea (22% of cycles), and mucositis (15% of cycles). Kemeny et al. (125) performed a phase I trial using HAI FUDR and systemic irinotecan also in patients with unresectable hepatic metastases from colorectal cancer. Unlike the prior trial where patients were untreated in the metastatic setting, in this trial 100% of the 38 patients who did not undergo cryosurgery had received one prior regimen and 42% had received two prior regimens. In these patients, the response rate was 74% and the updated median survival is 20 months. The dose-limiting toxicities were neutropenia and diarrhea. A potential reason for the higher response rate in the trial by Kemeny et al. may be due to the use of HAI FUDR rather than pirarubicin. Combining systemic oxaliplatin þCPT11 and HAI-FUDRþ Dex in previously treated patients a response late of 86% was obtained with an median survival of 37 months. 30% of patients were able to undergo hepatic resection. (126þupdate results) Most of the trials using HAI chemotherapy in the adjuvant setting also incorporated systemic chemotherapy and as previously discussed have shown a benefit over systemic chemotherapy alone (Table 6). Trials using HAI chemotherapy and systemic oxaliplatin combination chemotherapy are eagerly awaited, and eventually comparative trials of the combination of HAI and systemic therapy versus systemic chemotherapy alone will be performed and will provide essential information on the relative merits of these approaches. HAI as Neoadjuvant Therapy for Unresectable Liver Metastases A number of trials have reported using neoadjuvant systemic chemotherapy in patients with unresectable liver metastases in an attempt to render them resectable. Bismuth et al. (128) and Adam et al. (129) initially reported a resection rate of 16% and 13.5% in an updated report excluding patients

270

Leonard and Kemeny

with extrahepatic disease. In one trial of highly selected patients, surgery was attempted in 51% of patients, and 38% of all patients had a complete resection (6). However, a recent report on 795 patients with unresectable disease from the NCCTG N9741 trial found only 3.3% of patients able to undergo a curative operation after neoadjuvant systemic chemotherapy (130). As HAI chemotherapy has consistently demonstrated higher response rates than systemic chemotherapy, it is possible that HAI may be the optimal approach for improving resectability. Two early reports of HAI used as a neoadjuvant approach confirmed that resectability could be induced in formerly inoperable patients. Elias et al. (131) showed that 14 (5.8%) patients out of a total of 239 patients with initially unresectable hepatic tumors became resectable after HAI. Nine of the 14 patients had colorectal cancer. Link et al. reproduced these findings by evaluating four different HAI regimens in 168 patients with unresectable liver metastases. Although only 5% of patients were successfully downstaged to an extent that surgery was possible, the survival of one of these patients was 58 months (110). More recent studies found resection rates of 4.7% and 26% in previously unresectable patients with colorectal liver metastases despite the fact that most of these patients had previously failed systemic chemotherapy (132,133). Another trial of 36 patients with unresectable liver metastases used HAI 5-FU-based therapy (13 of the 36 were chemonaive, and also received systemic therapy), and from 31 assessable patients four became eligible for liver resection (unclear whether these four were chemonaive or previously treated) although one patient refused surgery (134). At MSKCC, 44 patients with unresectable liver metastases were treated with HAI FUDR þ dexamethasone and systemic oxaliplatin-based chemotherapy (FOLFOX or CPT-11/ oxaliplatin) as part of phase I protocols (126). Almost 90% of patients had previously been treated with chemotherapy and 70% had received CPT-11 prior to entering these protocols, yet the partial response rate was 77%. Nine patients had complete resection of their tumor and seven other patients are being considered for surgery. Further prospective trials are necessary to confirm these results. HAI as Second-Line Chemotherapy The current vogue for the optimal utilization of systemic therapy is the sequential use of modern chemotherapy agents. The median survival of patients correlates with the percentage of patients who receive all three effective chemotherapies for colorectal cancer (5-FU, CPT-11, oxaliplatin) at some point in their course (135). However, the choice of second-line therapy was not an important parameter in outcome. The best example of a study analyzing the sequential use of chemotherapy was performed by Tournigand et al. (3) and has resulted in the longest median survival seen in published trials to date. Patients were randomized to FOLFOX-6 followed by folinic acid, fluorouracil, and irinotecan (FOLFIRI) or the same drugs but in the opposite

Hepatic Directed Therapy

271

order. The second-line response rate for FOLFIRI was 4% and for FOLFOX-6 15%. Similarly, Rothenberg et al. (113) demonstrated response rates of 9.9%, 1.3%, and 0% for FOLFOX-4, oxaliplatin, and 5-FU, respectively, in patients who had previously progressed on CPT-11/5-FU/LV. A recent trial reported a second-line response rate of 15% for oxaliplatin/5-FU also in patients who had progressed on CPT-11-based chemotherapy (136). However, despite the low second-line response rates, the sequential use of all active agents has increased the median survival of patients from 12 to 14 months for trials performed at the beginning of the decade to 17 to 20 months for trials reported this year (3,4,137,138). The possibility that more effective second-line chemotherapy could further increase the median survival is an exciting prospect. As discussed with neoadjuvant chemotherapy, HAI combined with systemic chemotherapy has very high response rates as second-line therapy. The median survival for the 44 patients treated at MSKCC with FUDR and dexamethasone combined with systemic oxaliplatin-based chemotherapy was 25 months despite most patients having previously been treated with chemotherapy (126). In the subset of patients (n ¼ 23) who were treated with HAI FUDR and dexamethasone combined with systemic CPT-11 and oxaliplatin the median survival was 35 months despite the fact that 74% of patients had previously been treated with systemic CPT-11. The concept that tumors can be resensitized to previously inactive chemotherapy has been demonstrated with biologic agents. Single-agent cetuximab has a response rate of approximately 9%, whereas cetuximab and CPT-11 combined has demonstrated response rates of approximately 20% in patients who previously progressed on CPT-11 (91–93). Similarly, the use of HAI may render patients sensitive to drugs that they had previously progressed on (103,124,126). Future trials should incorporate HAI combined with systemic chemotherapy with or without biologic agents as a second-line approach and compare this to the current standard of the sequential use of systemic chemotherapy combined with biologic agents. PORTAL VEIN INFUSION CHEMOTHERAPY In 1957, Morales et al. (139) discussed the use of portal vein infusion of chemotherapy at the time of surgery in an attempt to prevent liver metastases. In 1978, Taylor (140) compared outcomes on 24 patients with nonresectable colorectal liver metastases who at the time of their primary colon surgery were treated with no subsequent therapy or hepatic artery ligation alone, portal vein infusion chemotherapy or hepatic artery ligation, and portal vein infusion. A mean survival of 3, 4, and 10 months, respectively, was seen, fueling further interest in portal vein infusion with or without hepatic artery ligation. Hafstrom et al. randomized 84 patients to portal vein infusion 5-FU and oral allopurinol or no treatment. The median survival for the portal vein

272

Leonard and Kemeny

infusion group was 17 months compared to 8 months for the control group ( p ¼ 0.0039) (141). Subsequent studies evaluated portal vein infusion versus HAI. As discussed previously, Daly et al. (11) compared portal vein infusion and HAI administration of FUDR and demonstrated higher response rates with HAI. Recently, Naredi et al. (142) compared portal vein infusion 5-FU and hepatic artery ligation with HAI 5-FU in patients also with colorectal liver metastases. The trial was stopped after randomization of 39 patients, as the mean survival was 19 months for HAI and only 13 months for portal vein infusion ( p ¼ 0.01). Due to the biological rationale for using HAI rather than portal vein infusion in patients with large colorectal liver metastases and because of the number of trials that have confirmed the efficacy of HAI in this setting, the use of portal vein infusion has diminished. Although portal vein infusion may not be as effective as other therapies for metastatic colorectal cancer, the role of portal vein infusion in the adjuvant setting for early-stage cancers is supported by the biological rationale that liver metastases usually develop due to hematogenous spread from the mesenteric veins to the portal system and that hepatic micrometastases can be successfully treated by portal vein infusion (8,9). In 1985, Taylor et al. (143) randomized 117 patients undergoing surgery for colorectal cancer to postoperative portal vein infusion and 127 patients to control. The portal vein infusion group had an improved survival although subset analysis revealed that the benefit was only seen in Dukes’ B patients and not in Dukes’ C patients. A large number of subsequent trials appeared to validate these findings, and so a meta-analysis of 10 trials was performed to clarify the role of portal vein infusion (144). Including the initial Taylor et al. trial, the absolute survival difference at five years was 4.7% ( p ¼ 0.006). Analyses of only the subsequent nine hypotheses testing trials demonstrated a survival difference of 3.6% ( p ¼ 0.04). Since this meta-analysis a number of trials have failed to demonstrate a significant benefit for adjuvant portal vein infusion (145–147). The largest trial to assess adjuvant portal vein infusion randomized 616 patients to portal vein infusion or no adjuvant therapy (146). There was no significant difference in the five-year survival between the two groups (72% vs. 73%, respectively). Due to the success of other treatments and the questionable benefit of portal vein infusion, current enthusiasm for this approach has waned.

ISOLATED HEPATIC PERFUSION CHEMOTHERAPY High-dose chemotherapy and bone marrow transplantation have demonstrated successful results in some solid tumors such as testicular cancer, but is predominantly used in hematopoietic malignancies. The limited use of high-dose chemotherapy in solid tumors is due to its uncertain efficacy over standard-dose chemotherapy and its high rates of morbidity and

49 (46) 50 (45) 29 (4) 9 (9) 9 12 (6) — 51 (51) 73 (73)

Aigner et al. (150) Schwemmle et al. (151) Hafstrom et al. (152) De Vries et al. (153) Marinelli et al. (154) Oldhafer et al. (155) Ku et al. (156,157) Bartlett et al. (158)

Rothbarth et al. (159)

5-FU/mitomycin, 5-FU 5-FU/mitomycin (þcisplatin n ¼ 4) Melphalan/cisplatin Melphalan/TNF Mitomycin Mitomycin — Melphalan/TNF (n ¼ 31), melphalan/TNF þ HAI (n ¼ 19) Melphalan

Chemotherapy

5.6%

6% (total) 8% 14% 33 0% 0% — 2% (total)

Operative mortality

7–18a 14 6.5 10.3 17 9 13 16,27 28.8

55 4

b

Median survival (mos) 26,79 68 20 83 13b 33 64b 77,74

Partial response (%)

67,16 22 0 0 13 0 0 0,0

Complete response (%)

Nonresponders and responders had median survivals of 7–8 months and 18 months, respectively. Tumor reduction was evident in 6/8 patients (75%). Abbreviations: CRC, colorectal cancer; FU, fluorouracil; HAI, hepatic arterial infusion; TNF, tumor necrosis factor; N, number.

a

N (CRC)

Trial

Table 7 Phase I and II Trials of Isolated Hepatic Perfusion

Hepatic Directed Therapy 273

274

Leonard and Kemeny

mortality. One potential way of facilitating high-dose chemotherapy exposure to a tumor while limiting systemic toxicity is to use regional directed chemotherapy. HAI relies on this technique to a certain extent by administering chemotherapy through the hepatic artery and selecting drugs that are predominantly extracted in the liver, thereby causing most of the cytotoxic effect to occur in the liver. Another method is to isolate the vascular supply to the liver by perfusing it with an extracorporeal bypass circuit and perfusing it with high doses of chemotherapy that do not enter the systemic circulation. This concept, known as IHP, was first devised by Klopp et al. in 1950 (148). Theoretical advantages of IHP over HAI are higher drug concentrations; drugs with low extraction rates can be used, as can hyperthermia, hypoxia, or other conditions facilitating cytotoxic activity. Its first application in humans was in 1961 when Ausman et al. (149) treated five patients with nitrogen mustard by IHP. Subsequent interest in this technique occurred due to the success reported for isolated limb perfusion in treating intransit metastases associated with melanoma. A number of trials have since evaluated IHP in patients with colorectal liver metastases (Table 7). Mortality rates have usually been below 10% and response rates have been seen in up to 83% of patients. The two most recent trials and the largest trials of IHP in colorectal cancer have shown particularly promising results. Bartlett et al. evaluated IHP melphalan and hyperthermia in 51 patients with 32 patients also receiving tumor necrosis factor. Patients were divided into two arms; one arm was treated with IHP melphalan alone (n ¼ 31) while the other arm was treated with IHP melphalan followed by HAI FUDR and leucovorin (n ¼ 19) (158). Only one patient died perioperatively and the response rates were 77% and 74%, respectively, with median survivals of 16 and 27 months, respectively. However, in the IHP-alone arm, the liver was the first site of failure in 42% of patients. Rothbarth et al. (159) used IHP melphalan in 73 patients with unresectable colorectal liver metastases and demonstrated an operative mortality of 5.6% (159). The response rate was 59% and the median survival 28.8 months with a three-year survival of 37%. Further studies are required to substantiate these findings, although the technical expertise and cost of IHP may limit its widespread applicability. CONCLUSION Hepatic directed therapy for colorectal cancer has demonstrated promising results. HAI chemotherapy appears to be the most effective of the hepatic directed strategies and is the approach in which there is the most experience. Despite a number of trials showing improvements in outcomes using HAI with or without systemic chemotherapy compared to the standard approach of systemic chemotherapy, there is considerable debate on the relative merits of HAI therapy. Proponents of HAI therapy highlight the superior outcomes in trials where HAI was appropriately delivered.

Hepatic Directed Therapy

275

The effectiveness of HAI therapy will be enhanced by more experience in the administration of HAI therapy, improvements in agents used for HAI, and further progress in the development of systemic therapies to complement HAI. Further progress in the field of HAI therapy may considerably reduce the morbidity and mortality associated with colorectal cancer, and so continued research in this area must be encouraged.

REFERENCES 1. Jemal A, Tiwari RC, Murray T, et al. Cancer statistics, 2004. CA Cancer J Clin 2004; 54:8–29. 2. Weber SM, Jarnagin WR, DeMatteo RP, et al. Survival after resection of multiple hepatic colorectal metastases. Ann Surg Oncol 2000; 7:643–650. 3. Tournigand C, Andre T, Achille E, et al. FOLFIRI followed by FOLFOX6 or the reverse sequence in advanced colorectal cancer: a randomized GERCOR study. J Clin Oncol 2004; 22:229–237. 4. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004; 350:2335–2342. 5. Goldberg RM, Sargent DJ, Morton RF, et al. A randomized controlled trial of fluorouracil plus leucovorin, irinotecan, and oxaliplatin combinations in patients with previously untreated metastatic colorectal cancer. J Clin Oncol 2004; 22:23–30. 6. 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; 10:663–669. 7. Weiss L, Grandmann E, Torhost J, et al. Hematogenous metastatic patterns in colonic carcinoma: an analysis of 1541 necropsies. J Pathol 1986; 150:195–203. 8. Breedis C, Young C. The blood supply of neoplasms in the liver. Am J Pathol 1954; 30:969–977. 9. Ackerman NB. The blood supply of experimental liver metastases. IV Changes in vascularity with increasing tumor growth. Surgery 1974; 75:589–596. 10. Sigurdson ER, Ridge JA, Kemeny N, et al. Tumor and liver drug uptake following hepatic artery and portal vein infusion. J Clin Oncol 1987; 5: 1836–1840. 11. Daly JM, Kemeny N, Sigurdson E, et al. Regional infusion for colorectal hepatic metastases. A randomized trial comparing the hepatic artery with the portal vein. Arch Surg 1987; 122:1273–1277. 12. Collins JM. Pharmacologic rationale for regional drug delivery. J Clin Oncol 1984; 2:498–504. 13. Ensminger WD, Rosowsky A, Raso V. A clinical pharmacological evaluation of hepatic arterial infusions of 5-fluoro-2-deoxyuridine and 5-fluorouracil. Cancer Res 1978; 38:3784–3792. 14. Ensminger WD, Gyves JW. Clinical pharmacology of hepatic arterial chemotherapy. Semin Oncol 1983; 10:176–182.

276

Leonard and Kemeny

15. van Tellingen O, Kuck MT, Vlasveld LT, et al. Unchanged pharmacokinetics of etoposide given by intra-arterial hepatic infusion as compared with i.v. infusion. Cancer Chemother Pharmacol 1996; 38:387–390. 16. Meta-analysis Group in Cancer. Efficacy of intravenous continuous infusion of fluorouracil compared with bolus administration in advanced colorectal cancer. J Clin Oncol 1998; 16:301–308. 17. Sullivan RD, Miller E, Sikes MP. Antimetabolite–metabolite combination cancer chemotherapy. Effects of intraarterial methotrexate-intramuscular Citrovorum factor therapy in human cancer. Cancer 1959; 12:1248–1262. 18. Sullivan RD, Norcross JW, Watkins E Jr. Chemotherapy of metastatic liver cancer by prolonged hepatic-artery infusion. N Engl J Med 1964; 270:321–327. 19. Clarkson B, Young C, Dierick W, et al. Effects of continuous hepatic artery infusion of antimetabolites on primary and metastatic cancer of the liver. Cancer 1962; 15:472–488. 20. Bierman HR, Byron RL Jr., Kelley KH, et al. Studies on the blood supply of tumors in man. III. Vascular patterns of the liver by hepatic arteriography in vivo. J Natl Cancer Inst. 1951; 12:107–131. 21. Barone RM, Byfield JE, Goldfarb PB, et al. Intra-arterial chemotherapy using an implantable infusion pump and liver irradiation for the treatment of hepatic metastases. Cancer 1982; 50:850–862. 22. Wickremesekera JK, Cannan RJ, Stubbs RS. Hepatic artery access ports: recognizing and avoiding the problems. Aust N Z J Surg 2000; 70:496–502. 23. Blackshear PJ, Dorman FD, Blackshear PL Jr., et al. The design and initial testing of an implantable infusion pump. Surg Gynecol Obstet 1972; 134:51–56. 24. Rohde TD, Blackshear PJ, Varco RL, et al. One year of heparin anticoagulation. An ambulatory subject using a totally implantable infusion pump. Minn Med 1977; 60:719–722. 25. Buchwald H, Grage TB, Vassilopoulos PP, et al. Intraarterial infusion chemotherapy for hepatic carcinoma using a totally implantable infusion pump. Cancer 1980; 45:866–869. 26. Yasuda S, Noto T, Ikeda M, et al. Hepatic arterial infusion chemotherapy using implantable reservoir in colorectal liver metastasis. Gan To Kagaku Ryoho 1990; 17:1815–1819. 27. Heinrich S, Petrowsky H, Schwinnen I, et al. Technical complications of continuous intra-arterial chemotherapy with 5-fluorodeoxyuridine and 5-fluorouracil for colorectal liver metastases. Surgery 2003; 133:40–48. 28. Mathur P, Allen-Mersh TG. Hepatic arterial chemotherapy for colorectal liver metastases. Hepatogastroenterology 2001; 48:317–319. 29. Michels NA. The anatomic variations of the arterial pancreaticoduodenal arcades: their import in regional resection involving the gall bladder, bile ducts, liver, pancreas and parts of the small and large intestines. J Int Coll Surg 1962; 37:13–40. 30. Kemeny MM, Hogan JM, Goldberg DA, et al. Continuous hepatic artery infusion with an implantable pump: problems with hepatic artery anomalies. Surgery 1986; 99:501–504. 31. Kemeny MM. The surgical aspects of the totally implantable hepatic artery infusion pump. Arch Surg 2001; 136:348–352.

Hepatic Directed Therapy

277

32. Bloom AI, Gordon RL, Ahl KH, et al. Transcatheter embolization for the treatment of misperfusion after hepatic artery chemoinfusion pump implantation. Ann Surg Oncol 1999; 6:350–358. 33. Uchino M, Takizawa K. Infusion MR arteriography during hepatic arterial infusion chemotherapy: evaluation of clinical usefulness. Nippon Igaku Hoshasen Gakkai Zasshi 2003; 63:558–564. 34. Campbell KA, Burns RC, Sitzmann JV, et al. Regional chemotherapy devices: effect of experience and anatomy on complications. J Clin Oncol 1993; 11: 822–826. 35. Urbach DR, Hansen PD. Laparoscopic placement of a continuous hepatic artery infusion pump. Semin Laparosc Surg 2000; 7:140–147. 36. Northover J, Terblance J. A new look at the arterial supply of the bile duct in man and its surgical implications. Br J Surg 1979; 66:379–384. 37. Kemeny M, Battifora H, Flayney D, et al. Sclerosing cholangitis after continuous hepatic artery infusion of FUDR. Ann Surg 1985; 202:176–181. 38. Doria MI Jr, Shepard KV, Levin B, et al. Liver pathology following hepatic arterial infusion chemotherapy. Hepatic toxicity with FUDR. Cancer 1986; 58:855–861. 39. Hohn D, Melnick J, Stagg R, et al. Biliary sclerosis in patients receiving hepatic arterial infusions of floxuridine. J Clin Oncol 1985; 3:98–102. 40. Brown K, Kemeny N, Berger M, et al. Obstructive jaundice in patients receiving hepatic artery infusional chemotherapy: etiology, treatment implications, and complications after transhepatic biliary drainage. J Vasc Interv Radiol 1997; 8:229–234. 41. Kemeny N, Steiter K, Diedzweiecki D, et al. A randomized trial of intrahepatic infusion of fluorouridine (FUDR) with dexamethasone versus FUDR alone in the treatment of metastatic colorectal cancer. Cancer 1992; 69:327–334. 42. Hrushesky W, Von Roemelling R, Lanning R, et al. Circadian-shaped infusions of floxuridine for progressive metastatic renal cell carcinoma. J Clin Oncol 1990; 8:1504–1513. 43. Stagg R, Venook A, Chase J, et al. Alternating hepatic intra-arterial floxuridine and fluorouracil: a less toxic regimen for treatment of liver metastases from colorectal cancer. J Natl Cancer Inst 1991; 83:423–428. 44. Davidson B, Izzo F, Chase J, et al. Alternating floxuridine and 5-fluorouracil hepatic arterial chemotherapy for colorectal liver metastases minimizes biliary toxicity. Am J Surg 1996; 172:244–247. 45. Hohn D, Stagg R, Price D, et al. Avoidance of gastroduodenal toxicity in patients receiving hepatic arterial 5-fluoro-2-deoxyuridine. J Clin Oncol 1985; 3:1257–1260. 46. Oberfield RA, McCaffrey JA, Polio J, et al. Prolonged and continuous percutaneous intra-arterial hepatic infusion chemotherapy in advanced metastatic liver adenocarcinoma from colorectal primary. Cancer 1979; 44:414–423. 47. Niederhuber JE, Ensminger W, Gyves J, et al. Regional chemotherapy of colorectal cancer metastatic to the liver. Cancer 1984; 53:1336–1343. 48. Balch CM, Urist MM, Soong SJ, et al. A prospective phase II clinical trial of continuous FUDR regional chemotherapy for colorectal metastases to the liver using a totally implantable drug infusion pump. Ann Surg 1983; 198:567–573.

278

Leonard and Kemeny

49. Kemeny N, Conti JA, Cohen A, et al. Phase II study of hepatic arterial floxuridine, leucovorin, and dexamethasone for unresectable liver metastases from colorectal carcinoma. J Clin Oncol 1994; 12:2288–2295. 50. Shepard KV, Levin B, Karl RC, et al. Therapy for metastatic colorectal cancer with hepatic artery infusion chemotherapy using a subcutaneous implanted pump. J Clin Oncol 1985; 3:161–169. 51. Cohen AM, Kaufman SD, Wood WC, et al. Regional hepatic chemotherapy using an implantable drug infusion pump. Am J Surg 1983; 145:529–533. 52. Weiss GR, Garnick MB, Osteen RT, et al. Long-term arterial infusion of 5-fluorouracil for liver metastases using an implantable infusion pump. J Clin Oncol 1983; 1:337–344. 53. Schwartz SI, Jones LS, McCune CS. Assessment of treatment of intrahepatic malignancies using chemotherapy via an implantable pump. Ann Surg 1985; 201:560–567. 54. Johnson LP, Wasserman PB, Rivkin SE. FUDR hepatic arterial infusion via an implantable pump for treatment of hepatic tumors. Proc Am Soc Clin Oncol 1983; 2:119 (abstr 465). 55. Howell JD, McArdle CS, Kerr DJ, et al. A phase II study of regional 2-weekly 5-fluorouracil infusion with intravenous folinic acid in the treatment of colorectal liver metastases. Br J Cancer 1997; 76:1390–1393. 56. Arai Y, Inaba Y, Takeuchi Y, et al. Intermittent hepatic arterial infusion of high-dose 5FU on a weekly schedule for liver metastases from colorectal cancer. Cancer Chemother Pharmacol 1997; 40:526–530. 57. Kemeny N, Daly J, Reichman B, et al. Intrahepatic or systemic infusion of fluorodeoxyuridine in patients with liver metastases from colorectal carcinoma. Ann Intern Med 1987; 122:459–465. 58. Chang AE, Schneider PD, Sugarbaker PH. A prospective randomized trial of regional versus systemic continuous 5-flouoroxyuridine chemotherapy in the treatment of colorectal liver metastases. Ann Surg 1987; 206:685–693. 59. Hohn DC, Stagg RJ, Friedman MA, et al. A randomized trial of continuous intravenous versus hepatic intra-arterial floxuridine in patients with colorectal cancer metastatic to the liver: the Northern California Oncology Group Trial. J Clin Oncol 1989; 7:1646–1654. 60. Kemeny MM, Goldberg D, Beatty JD, et al. Results of a prospective randomized trial of continuous regional chemotherapy and hepatic resection as treatment of hepatic metastases from colorectal primaries. Cancer 1986; 57:492–498. 61. Wagman LD, Kemeny MM, Leong L, et al. A prospective, randomized evaluation of the treatment of colorectal cancer metastatic to the liver. J Clin Oncol 1990; 8:1885–1893. 62. Martin JK Jr., O’Connell MJ, Wieand HS, et al. Intra-arterial floxuridine vs systemic fluorouracil for hepatic metastases from colorectal cancer. A randomized trial. Arch Surg 1990; 125:1022–1027. 63. Rougier P, Laplanche A, Huguier M, et al. Hepatic arterial infusion of floxuridine in patients with liver metastases from colorectal carcinoma: long-term results of a prospective randomized trial. J Clin Oncol 1992; 10:1112–1118.

Hepatic Directed Therapy

279

64. Allen-Mersh T, Earlam S, Fordy C, et al. Quality of life and survival with continuous hepatic artery floxuridine infusion for colorectal liver metastases. Lancet 1994; 344:1255–1260. 65. Lorenz M, Muller HH. Randomized, multicenter trial of fluorouracil plus leucovorin administered either via hepatic arterial or intravenous infusion versus fluorodeoxyuridine administered via hepatic arterial infusion in patients with nonresectable liver metastases from colorectal carcinoma. J Clin Oncol 2000; 18:243–254. 66. Kerr DJ, McArdle CS, Ledermann J, et al. Intrahepatic arterial versus intravenous fluouracil and folinic acid for colorectal cancer liver metastases: a multicentre randomized trial. Lancet 2003; 361:368–373. 67. Kemeny N, Neidwiecki D, Hollis D, et al. Hepatic arterial infusion (HAI) versus systemie therapy for efficacy, quality of life (QOL), cost effectiveness, and molecular markers. In Press.. 68. Meta-analysis Group in Cancer. Reappraisal of hepatic arterial infusion in the treatment of nonresectable liver metastases from colorectal cancer. J Natl Cancer Inst 1996; 88:252–258. 69. Harmantas A, Rotstein LE, Langer B. Regional versus systemic chemotherapy in the treatment of colorectal carcinoma metastatic to the liver. Is there a survival difference? Meta-analysis of the published literature. Cancer 1996; 78:1639–1645. 70. Durand-Zaleski I, Earlam S, Fordy C, et al. Cost-effectiveness of systemic and regional chemotherapy for the treatment of patients with unresectable colorectal liver metastases. Cancer 1998; 83:882–888. 71. Scheele J, Stang R, Altendorf-Hofmann A. Hepatic metastases from colorectal carcinoma: impact of surgical resection on the natural history. Br J Surg 1990; 77:1241–1246. 72. Wilson S, Adson M. Surgical treatment of hepatic metastases from colorectal cancers. Arch Surg 1976; 111:330–334. 73. Adson MA, van Heerden JA, Adson MH, et al. Resection of hepatic metastases from colorectal cancer. Arch Surg 1984; 119:647–651. 74. Kato T, Yasui K, Hirai T, et al. Therapeutic results for hepatic metastasis of colorectal cancer with special reference to effectiveness of hepatectomy: analysis of prognostic factors for 763 cases recorded at 18 institutions. Dis Colon Rectum 2003; 46(suppl 10):S22–S31. 75. Fong Y, Fortner J, Sun RL, et al. Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: analysis of 1001 consecutive cases. Ann Surg 1999; 230:309–318. 76. Nordlinger B, Guigiet M, Vaillant JC, et al. Surgical resection of colorectal carcinoma metastases to the liver. A prognostic scoring system to improve case selection, based on 1568 patients. Association Francaise de Chirurgie. Cancer 1996; 77:1254–1262. 77. Ruers T, Bleichrodt RP. Treatment of liver metastases, an update on the possibilities and results. Eur J Cancer 2002; 38:1023–1033. 78. Fong Y, Blumgart L, Cohen A, et al. Repeat hepatic resections for metastatic colorectal cancer. Ann Surg 1994; 220:657–662.

280

Leonard and Kemeny

79. Altendorf-Hofmann A, Scheele J. A critical review of the major indicators of prognosis after resection of hepatic metastases from colorectal carcinoma. Surg Oncol Clin N Am 2003; 12:165–192. 80. O’Connell MJ, Adson MA, Schutt AJ, et al. Clinical trial of adjuvant chemotherapy after surgical resection of colorectal cancer metastatic to the liver. Mayo Clin Proc 1985; 60:517–520. 81. Portier G, Rougier P, Milan C, et al. Adjuvant systemic chemotherapy (CT) using 5-fluorouracil (FU) and folinic acid (FA) after resection of liver metastases (LM) from colorectal cancer (CRC) origin. Results of an intergroup phase III study (trial FFCD-ACHBTH-AURC 9002). Proc Am Soc Clin Oncol 2002; 21:133a (abstr 528). 82. Langer B, Blieberg H, Labianca R, et al. Fluorouracil (FU) plus l-leucovorin (i-LV) versus observation after potentially curative resection of liver or lung metastases from colorectal cancer (CRC): results of the ENG (EORTC/NCIC CTG/ GIVO) randomized trial. Proc Am Soc Clin Oncol 2002; 21:149a (abstr 592). 83. Lygidakis NJ, Ziras N, Parissis J. Resection versus resection combined with adjuvant pre- and post-operative chemotherapy-immunotherapy for metastatic colorectal liver cancer. A new look at an old problem. Hepatogastroenterology 1995; 42:155–161. 84. Lorenz M, Muller H-H, Schramm H, et al. Randomized trial of surgery versus surgery followed by adjuvant hepatic arterial infusion with 5-fluorouracil and folinic acid for liver metastases of colorectal cancer. German Cooperative on Liver Metastases (Arbeitsgruppe Lebermetastasen). Ann Surg 1998; 228:756–762. 85. Kemeny N, Huang Y, Cohen AM, et al. Hepatic arterial infusion of chemotherapy after resection of hepatic metastases from colorectal cancer. N Engl J Med 1999; 341:2039–2048. 86. Tono T, Hasuike Y, Ohzato H, et al. Limited but definite efficacy of prophylactic hepatic arterial infusion chemotherapy after curative resection of colorectal liver metastases: A randomized study. Cancer 2000; 88:1549–1556. 87. Kemeny MM, Adak S, Gray B, et al. Combined-modality treatment for resectable metastatic colorectal carcinoma to the liver: surgical resection of hepatic metastases in combination with continuous infusion of chemotherapy—an intergroup study. J Clin Oncol 2002; 20:1499–1505. 88. O’Connell MJ, Bolton JS, Mahoney MR, et al. Final results of hepatic arterial infusion (HAI) plus systemic (SYS) chemotherapy after multiple metastasectomy in patients with colorectal carcinoma metastatic (M-CRC) to the liver: A North Central Cancer Treatment Group (NCCTG) phase II study. Proc Am Soc Clin Oncol 2004; 23:252 (abstr 3527). 89. Sadahiro S, Suzuki T, Ishikawa K, et al. Prophylactic hepatic arterial infusion chemotherapy for the prevention of liver metastases in patients with colon carcinoma. A randomized controlled trial. Cancer 2004; 100:590–597. 90. Ota M, Shimada H, Masui H, et al. Adjuvant hepatic arterial infusion chemotherapy after curative resection for Dukes C colorectal cancer: a pilot study. Hepatogastroenterology 2004; 51:124–127. 91. Saltz L, Rubin M, Hochster H, et al. Cetuximab (IMC-C225) plus irinotecan (CPT-11) is active in CPT-11-refractory colorectal cancer (CRC) that expresses

Hepatic Directed Therapy

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

281

epidermal growth factor receptor (EGFR). Proc Am Soc Clin Oncol 2001; 20:3a (abstr 7). Cunningham D, Humblet Y, Siena S, et al. Cetuximab (C225) alone or in combination with irinotecan (CPT-11) in patients with epidermal growth factor receptor (EGFR)-positive, irinotecan-refractory metastatic colorectal cancer (MCRC). Proc Am Soc Clin Oncol 2003; 22:252 (abstr 1012). Saltz LB, Meropol NJ, Loehrer PJ Sr., et al. Phase II trial of cetuximab in patients with refractory colorectal cancer that expresses the epidermal growth factor receptor. J Clin Oncol 2004; 22:1201–1208. van Riel JM, van Groeningen CJ, de Greve J, et al. Continuous infusion of hepatic arterial irinotecan in pretreated patients with colorectal cancer metastatic to the liver. Ann Oncol 2004; 15:59–63. Fiorentini G, Rossi S, Dentico P, et al. Irinotecan hepatic arterial infusion chemotherapy for hepatic metastases from colorectal cancer: a phase II clinical study. Tumori 2003; 89:382–384. Vogl TJ, Eichler K, Mack MG, et al. A phase I study of hepatic arterial infusion (HAI) of irinotecan in patients with inoperable liver metastases of colorectal cancer. Proc Am Soc Clin Oncol 2004; 23:298 (abstr 3713). Kern W, Beckert B, Lang N, et al. Phase I and pharmacokinetic study of hepatic arterial infusion with oxaliplatin in combination with folinic acid and 5-fluorouracil in patients with hepatic metastases from colorectal cancer. Ann Oncol 2001; 12:599–603. Ducreux M, Ychou M, Laplanche A, et al. Hepatic arterial oxaliplatin infusion plus intravenous hemotherapy in colorectal cancer with inoperable hepatic metastases: A Trial of the Gastrointestinal Group of the Federation Nationale des centres de lutte contre le cancer. J Clin Oncol 2005; 23(22):4881–4887. Pennucci MC, Del Freo A, Della Seta R, et al. A feasibility study of hepatic arterial infusion (HAI) with oxaliplatin (Oxa) 5-fluorouracil (FU) and leucovorin (LV) in patients with liver metastases (LM) from colorectal cancer (CRC). Ann Oncol 2002; 13(suppl 3):46 (abstr C50). Bouchahda M, Tanaka K, Adam R, et al. Hepatic artery chronomodulated infusion of irinotecan, 5-fluorouracil and oxaliplatin against liver metastases in heavily pretreated patients with colorectal cancer. Proc Am Soc Clin Oncol 2004; 23:300 (abstr 3722). Tomirotti M, Pallavicini EB, Tacconi F, et al. First-line activity of hepatic arterial infusion (HAI) of oxaliplatin and systemic chemotherapy with leucovorin plus 5-fluorouracil in metastatic colorectal cancer: preliminary data of a multicenter study. Proc Am Soc Clin Oncol 2002; 22:135b (abstr 2353). Boige V, Lacombe S, De Baere T, et al. Hepatic arterial infusion oxaliplatin combined with intravenous 5-FU and folinic acid in non resectable liver metastasis of colorectal cancer: a promising option for failures to systemic chemotherapy. Proc Am Soc Clin Oncol 2003; 22:291 (abstr 1170). Mancuso A, Giuliani R, Accettura C, et al. Hepatic arterial continuous infusion (HACI) of oxaliplatin in patients with unresectable liver metastases from colorectal cancer. Anticancer Res 2003; 23:1917–1922.

282

Leonard and Kemeny

104. Guthoff I, Lotspeich E, Fester C, et al. Hepatic artery infusion using oxaliplatin in combination with 5-fluorouracil, folinic acid and mitomycin C: oxaliplatin pharmacokinetics and feasibility. Anticancer Res 2003; 23:5203–5208. 105. Fazio N, Orsi F, Grasso RF, et al. Hepatic intra-arterial chemotherapy using a percutaneous catheter in pretreated patients with metastatic colorectal carcinoma. Anticancer Res 2003; 23:5023–5030. 106. Kohnoe S, Endo K, Yamamoto M, et al. Protracted hepatic arterial infusion with low-dose cisplatin plus 5-fluorouracil for unresectable liver metastases from colorectal cancer. Surgery 2002; 131:S128–S134. 107. Rougier P, Munck JN, Elias D, et al. Intra-arterial hepatic chemotherapy with pirarubicin. Preclinical and clinical studies. Am J Clin Oncol 1990; 13(suppl 1): S1–S14. 108. Makela J, Kantola R, Tikkakoski T, et al. Superselective intra-arterial chemotherapy with mitomycin C in hepatic metastases from colorectal cancer. J Surg Oncol 1997; 65:127–131. 109. Liu LX, Zhang WH, Jiang HC, et al. Arterial chemotherapy of 5-fluorouracil and mitomycin C in the treatment of liver metastases of colorectal cancer. World J Gastroenterol 2002; 8:663–667. 110. Link KH, Pillasch J, Formentini A, et al. Downstaging by regional chemotherapy of non-resectable isolated colorectal liver metastases. Eur J Surg Oncol 1999; 25:381–388. 111. Sharma S, Krug M, Tong W, et al. Phase I study of escalating doses of fixed rate intrahepatic infusion of gemcitabine via an intrahepatic pump for patients with hepatic metastases from colorectal carcinoma. Proc Am Soc Clin Oncol 2003; 22:316 (abstr 1268). 112. van Riel JM, van Groeningen CJ, Kedde MA, et al. Continuous administration of irinotecan by hepatic arterial infusion: a phase I and pharmacokinetic study. Clin Cancer Res 2002; 8:405–412. 113. Rothenberg ML, Oza AM, Bigelow RH, et al. Superiority of oxaliplatin and fluorouracil–leucovorin compared with either therapy alone in patients with progressive colorectal cancer after irinotecan and fluorouracil–leucovorin: interim results of a phase III trial. J Clin Oncol 2003; 21:2059–2069. 114. Kemeny N, Israel K, Niedzwiecki D, et al. Randomized study of continuous infusion fluorouracil versus fluorouracil plus cisplatin in patients with metastatic colorectal cancer. J Clin Oncol 1990; 8:313–318. 115. Reid T, Galanis E, Abbruzzese J, et al. Hepatic arterial infusion of a replicationselective oncolytic adenovirus (dl1520): phase II viral, immunologic, and clinical endpoints. Cancer Res 2002; 62:6070–6079. 116. Yoshimura K, Hazama S, Iizuka N, et al. Hepatic arterial infusion of IL-2 and chemotherapy for unresectable liver metastasis from colorectal cancer. Gan To Kagaku Ryoho 2002; 29:2117–2120. 117. Melichar B, Touskova M, Blaha M, et al. Hepatic arterial administration of activated leukocytes in patients with liver metastases. Cancer Biother Radiopharm 2002; 17:545–552. 118. Pancera G, Garassino MC, Domenico Beretta G, et al. A feasibility study of combined hepatic arterial infusion (HAI) with FUDR and systemic

Hepatic Directed Therapy

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

283

chemotherapy (SYS) with FOLFOX-4 in advanced colorectal cancer (CRC). Proc Am Soc Clin Oncol 2002; 22:105b (abstr 2233). Cantore M, Aitini E, Rabbi C, et al. Combined intra-arterial locoregional and systemic treatment of nonresectable hepatic metastases of colorectal carcinoma. G Chir 1997; 18:235–239. Copur MS, Capadano M, Lynch J, et al. Alternating hepatic arterial infusion and systemic chemotherapy for liver metastases from colorectal cancer: a phase II trial using intermittent percutaneous hepatic arterial access. J Clin Oncol 2001; 19:2404–2412. Mancini R, Tedesco M, Garufi C, et al. Hepatic arterial infusion (HAI) of cisplatin and systemic fluorouracil in the treatment of unresectable colorectal liver metastases. Anticancer Res 2003; 23:1837–1841. O’Connell MJ, Nagorney DM, Bernath AM, et al. Sequential intrahepatic fluorodeoxyuridine and systemic fluorouracil plus leucovorin for the treatment of metastatic colorectal cancer confined to the liver. J Clin Oncol 1998; 16:2528–2533. Fallik D, Ychou M, Jacob J, et al. Hepatic arterial infusion using pirarubicin combined with systemic chemotherapy: a phase II study in patients with nonresectable liver metastases from colorectal cancer. Ann Oncol 2003; 14:856–863. Zelek L, Bugat R, Cherqui D, et al. Multimodal therapy with intravenous biweekly leucovorin, 5-fluorouracil and irinotecan combined with hepatic arterial infusion pirarubicin in non-resectable hepatic metastases from colorectal cancer (a European Association for Research in Oncology trial). Ann Oncol 2003; 14:1537–1542. Kemeny N, Gonen M, Sullivan D, et al. Phase I study of hepatic arterial infusion of floxuridine and dexamethasone with systemic irinotecan for unresectable hepatic metastases from colorectal cancer. J Clin Oncol 2001; 19: 2687–2695. Kemeny N, Tong W, Gonen M, et al. Phase I study of weekly oxaliplatin plus irinotecan in previously treated patients with metastatic colorectal cancer. J Clin Oncol 2002; 13:1490–1496. Kemeny N, Jarnagin W, Gonen M, et al. Phase I/II study of hepatic arterial therapy with floxuridine and dexamethasone in combination with intravenous irinotecan as adjuvant treatment after resection of hepatic metastases from colorectal cancer. J Clin Oncol 2003; 21:3303–3309. Bismuth H, Adam R, Levi F, et al. Resection of nonresectable liver metastases from colorectal cancer after neoadjuvant chemotherapy. Ann Surg 1996; 224:509–522. Adam R, Avisar E, Ariche A, et al. Five-year survival following hepatic resection after neoadjuvant therapy for nonresectable colorectal (liver) metastases. Ann Surg Oncol 2001; 8:347–353. Delaunoit TP, Krook J, Sargent DJ, et al. Chemotherapy-allowed resection of metastatic colorectal cancer: a cooperative group experience. Gastrointest Cancers Symp Prog Proc 2004:138 (abstr 196). Elias D, Lasser P, Rougier P, et al. Frequency, technical aspects, results, and indications of major hepatectomy after prolonged intra-arterial hepatic

284

132.

133.

134.

135.

136.

137.

138.

139. 140. 141.

142.

143. 144.

145.

146.

Leonard and Kemeny chemotherapy for initially unresectable hepatic tumors. J Am Coll Surg 1995; 180:213–219. Meric F, Patt YZ, Curley SA, et al. Surgery after downstaging of unresectable hepatic tumors with intra-arterial chemotherapy. Ann Surg Oncol 2000; 7:490– 495. Clavien PA, Selzner N, Morse M, et al. Downstaging of hepatocellular carcinoma and liver metastases from colorectal cancer by selective intra-arterial chemotherapy. Surgery 2002; 131:433–442. Milandri C, Calazolari F, Giampalma E, et al. Combined treatment of inoperable liver metastases from colorectal cancer. Tumori 2003; 89(suppl 4): 112–114. Grothey A, Sargent D, Goldberg RM, et al. Survival of patients with advanced colorectal cancer improves with the availability of fluorouracil– leucovorin, irinotecan, and oxaliplatin in the course of treatment. J Clin Oncol 2004; 22:1209–1214. Aparicio A, Iqbal S, Yang S, et al. Pilot trial of 5-FU/oxaliplatin (CIFOX) for patients with metastatic colorectal cancer after progression with 5-FU/LV/ CPT-11 (IFL). Proc Am Soc Clin Oncol 2004; 23:313 (abstr 3773). Saltz LB, Cox JV, Blanke C, et al. Irinotecan plus fluorouracil and leucovorin for metastatic colorectal cancer. Irinotecan Study Group. N Engl J Med 2000; 343:905–914. de Gramont A, Figer A, Seymour M, et al. Leucovorin and fluorouracil with or without oxaliplatin as first-line treatment in advanced colorectal cancer. J Clin Oncol 2000; 18:2938–2947. Morales F, Bell M, Mcdonald GO, et al. The prophylactic treatment of cancer at the time of operation. Ann Surg 1957; 146:588–593. Taylor I. Cytotoxic perfusion for colorectal liver metastases. Br J Surg 1978; 65:109–114. Hafstrom L, Engaras B, Holmberg SB, et al. Treatment of liver metastases from colorectal cancer with hepatic artery occlusion, intraportal 5-fluorouracil infusion, and oral allopurinol. A randomized clinical trial. Cancer 1994; 74:2749– 2756. Naredi P, Oman M, Blind PJ, et al. A comparison between hepatic artery ligation and portal 5-Fu infusion versus 5-Fu intra arterial infusion for colorectal liver metastases. Eur J Surg Oncol 2003; 29:459–466. Taylor I, Machin D, Mullee M, et al. A randomized controlled trial of adjuvant portal vein cytotoxic perfusion in colorectal cancer. Br J Surg 1985; 72:359–363. Liver Infusion Meta-analysis Group. Portal vein chemotherapy for colorectal cancer: a meta-analysis of 4000 patients in 10 studies. J Natl Cancer Inst 1997; 89:497–505. Nitti D, Wils J, Sahmoud T, et al. Final results of a phase III clinical trial on adjuvant intraportal infusion with heparin and 5-fluorouracil (5-FU) in resectable colon cancer (EORTC GITCCG 1983–1987). European Organization for Research and Treatment of Cancer. Gastrointestinal Tract Cancer Cooperative Group. Eur J Cancer 1997; 33:1209–1215. Rougier P, Sahmoud T, Nitti D, et al. Adjuvant portal-vein infusion of fluorouracil and heparin in colorectal cancer: a randomised trial. European

Hepatic Directed Therapy

147.

148.

149. 150.

151. 152. 153.

154.

155. 156. 157. 158. 159.

285

Organisation for Research and Treatment of Cancer Gastrointestinal Tract Cancer Cooperative Group, the Gruppo Interdisciplinare Valutazione Interventi in Oncologia, and the Japanese Foundation for Cancer Research. Lancet 1998; 351:1677–1681. Focan C, Bury J, Beauduin M, et al. Adjuvant intraportal chemotherapy for Dukes B2 and C colorectal cancer also receiving systemic treatment: results of a multicenter randomized trial. Groupe Regional d’Etude du Cancer ColoRectal (Belgium). Anticancer Drugs 2000; 11:549–554. Klopp CT, Alford TC, Bateman J, et al. Fractionated intra-arterial cancer; chemotherapy with methyl bis amine hydrochloride; a preliminary report. Ann Surg 1950; 132:811–832. Ausman RK. Development of a technique for isolated perfusion of the liver. N Y State J Med 1961; 61:3993–3997. Aigner K, Walther H, Link KH. Isolated liver perfusion with MMC/5FU: surgical technique, pharmacokinetics, clinical results. Contr Oncol 1988; 29:229– 246. Schwemmle K, Link KH, Rieck B. Rationale and indications for perfusion in liver tumors: current data. World J Surg 1987; 11:534–540. Hafstrom LR, Holmberg SB, Naredi PL, et al. Isolated hyperthermic liver perfusion with chemotherapy for liver malignancy. Surg Oncol 1994; 3:103–108. de Vries MR, Rinkes IH, van de Velde CJ, et al. Isolated hepatic perfusion with tumor necrosis factor alpha and melphalan: experimental studies in pigs and phase I data from humans. Recent Results Cancer Res 1998; 147:107–119. Marinelli A, de Brauw LM, Beerman H, et al. Isolated liver perfusion with mitomycin C in the treatment of colorectal cancer metastases confined to the liver. Jpn J Clin Oncol 1996; 26:341–350. Oldhafer KJ, Lang H, Frerker M, et al. First experience and technical aspects of isolated liver perfusion for extensive liver metastasis. Surgery 1998; 123:622–631. Ku Y, Tominaga M, Iwasaki T, et al. Isolated hepatic perfusion chemotherapy for unresectable malignant hepatic tumors. Int J Clin Oncol 2002; 7:82–90. Ku Y, Fukumoto T, Tominaga M, et al. Percutaneous isolated hepatic perfusion for treatment of colorectal hepatic metastases. Geka 2001; 62:653–659. Bartlett DL, Libutti SK, Figg WD, et al. Isolated hepatic perfusion for unresectable hepatic metastases from colorectal cancer. Surgery 2001; 129:176–187. Rothbarth J, Pijl ME, Vahrmeijer AL, et al. Isolated hepatic perfusion with high-dose melphalan for the treatment of colorectal metastasis confined to the liver. Br J Surg 2003; 90:1391–7.

11 Pharmacogenomics of Colorectal Cancer Patrick Johnston and Wendy L. Allen Drug Resistance Group, Centre for Cancer Research and Cell Biology, Queen’s University Belfast, Belfast, Northern Ireland, U.K.

Howard L. McLeod Department of Medicine, Washington University School of Medicine, St. Louis, Missouri, U.S.A.

INTRODUCTION Over the last two decades there have been significant improvements in anticancer treatment; however, empiric dosing and an unpredictable drug response are the norm for anticancer therapy (1). There is inherited variability in drug metabolizing enzymes, drug transporters, DNA repair enzymes, and drug targets, which all play an important role in the unpredictability of anticancer treatment outcomes when drugs are administered uniformly to all patients (2–5). Some drugs are detoxified by polymorphic enzyme systems, which contributes to the great degree of variability in pharmacokinetics and toxicity; therefore, there is a need to evaluate genomic polymorphisms (6). In the last decade, the median survival for patients with metastatic colorectal cancer has nearly doubled from 12 to about 22 months, and new agents in late-phase clinical treatment may soon extend this survival benefit further. In the metastatic setting, single-agent 5-fluorouracil (5-FU) produces response rates of only 10% to 20% (7). More recently, 5-FU has been combined with new classes of drugs, oxaliplatin, and CPT-11, and this has significantly improved response rates to the 40% to 50% range in patients

287

288

Johnston et al.

with metastatic colorectal cancer (8,9). Despite these improvements, there is a need to identify novel panels of molecular and biochemical markers that can be used to predict response to traditional and novel therapies. In colorectal cancer, a limited number of predictive markers have been identified to date (10–14). The use of these predictive markers as single markers has led to somewhat contrasting results; however, when the predictive markers are combined there appears to be a greater ability to predict response to treatment (13). The aim of predictive marker testing is to tailor treatment according to an individual patient and tumor profile. Most active agents against colorectal cancer have only modest clinical activity. Therefore, research efforts must continue to identify novel targets, develop new therapeutic agents, and identify and characterize those patients most likely to respond to treatment. LOSS OF HETEROZYGOSITY Studies on allelic losses in familial adenomatous polyposis and nonfamilial adenomatous polyposis patients established that 5q loss of heterozygosity (LOH) occurred in a substantial proportion of adenomas, whereas 17p and 18q LOH were a later event (15,16). This served as the model for colorectal carcinogenesis proposed by Fearon and Vogelstein (17). The emphasis of the model was on the accumulation of mutations rather than the order in which they arise. It is interesting to note that no allelic losses have been observed in normal colonic epithelium surrounding colorectal neoplasms (18,19). Consistent allelic loss of a specific chromosomal region in a particular cancer type is generally taken as evidence for the location of a putative tumor suppressor gene in that region. Many studies have now demonstrated that loss of 5q [adenomatous polyposis coli (APC) gene] occurs at the transition phase of normal to benign adenoma, whereas loss of 17p (p53 gene) or 18q [deleted in colon cancer (DCC) gene] occurs at a later-stage transition of adenoma to carcinoma (18). However, additional genetic lesions may participate in the growth of adenomas, including point mutations in other genes such as the K-ras gene (20,21). This allelic picture has demonstrated an essential role for these genes in tumor progression. It has been reported that 70% of colorectal cancers have lost a portion of chromosome 17p or 18q or both (22). The 17p chromosome contains p53, which is an important tumor suppressor, and is reported to be mutated in 40% to 60% of colorectal cancers. Miyaki et al. (23) reported that LOH at the p53 allele mediated conversion to the highest grade of dyspasia. It has also been demonstrated that loss of 18q has been associated with a poorer prognosis in stage II and III tumors (24). Furthermore, retention of both 18q alleles in patients with stage II colorectal cancer results in a 93% five-year survival compared to 54% in patients who had lost one allele (24). Watanabe et al. (25) showed that patients with stage III colorectal

Pharmacogenomics of Colorectal Cancer

289

cancer who retained both 18q alleles have a more favorable outcome after adjuvant 5-FU—based chemotherapy. The 18q chromosome segment contains three candidate tumor suppressor genes: deleted in colon cancer (DCC), smad 2, and smad 4 (26). DCC is a cell adhesion molecule, and decreased DCC expression may lead to altered adhesion and contribute to enhanced tumor growth and metastatic spread of colorectal cancer (27). Smad proteins are transcription factors involved in the TGFb pathway (28). They are involved in signaling from the TGFb receptor and have been shown to regulate transcription of target genes including c-myc (29,30). In addition, 50% of colorectal cancers have lost a portion of chromosome 8p, and reintroduction of 8p into colon cancer cells decreased tumorigenicity and invasiveness. Moreover, loss of 8p is associated with a worse prognosis. The identity and function of the genes contained within this region have yet to be identified (31). MICROSATELLITE INSTABILITY Genetic instability has been recognized as a central element in the genesis of malignant lesions, resulting in clonal evolution of genetic events acquired in the course of tumor progression (32). Microsatellite instability (MSI) is common to many forms of cancer and is found in half to two-thirds of sporadic colon cancers. MSI is caused by mutations in the mismatch repair (MMR) genes such as hMSH2, hMLH1 and hMSH6, which results in failure of the DNA mismatch repair system to correct errors that occur during replication (33). Inactivation of mismatch repair genes causes replication errors, termed MSI, or the mutator phenotype. MSI is defined as variations in the numbers of repetitive di-, tri- and tetra-nucleotide repeats (called microsatellites) found in DNA. The DNA base–base mismatch repair system (MMR genes) has been shown to account for microsatellite instability in nearly all cases of hereditary nonpolyposis colorectal cancer (HNPCC) associated replication error (RER) cancer (34). The MMR genes (hMLH1, hMSH2, and hMSH6) are responsible for HNPCC, and patients with HNPCC inherit a single defective allele of a mismatch repair gene and require additional somatic mutation to inactivate the relevant gene. DNA mismatch repair is responsible for 10% to 15% of all colorectal cancers and accounts for greater than 90% of HNPCC (34). The majority of sporadic MSI tumors are due to transcriptional silencing of the hMLH1 gene, which is caused by promoter methylation (35). The cluster of MMR genes acts as a unit, and when mutations occur in any of these genes the mutation rate increases 100–1000-fold. An in vitro study carried out by Meyers et al. (36) demonstrated that restoration of hMLH1 activity in the MMR-deficient HCT116 cells increased their sensitivity to 5-FU. However, in in vivo studies the MSIpositive phenotype has been associated with increased survival in patients

290

Johnston et al.

with stage III colorectal cancer who receive adjuvant 5-FU-based therapy. Furthermore, several other studies have shown that there is a survival advantage and a less aggressive phenotype in patients who receive adjuvant 5-FU-based chemotherapy that have a high frequency of MSI (37). More recently, MSI-positive patients were shown to receive no benefit from adjuvant 5-FU, in contrast with MSI stable patients (38). The discrepancy in the in vitro and in vivo findings may be due to other intrinsic biological differences between MSI-positive and MSI-negative tumors. There also appears to be a link between MSI and TGFb RII mutation. A recent study demonstrated that 61% of stage III colorectal cancer patients with MSI-high tumors also had a TGFb RII mutation (25). This study also demonstrated that patients who had MSI-high tumors and TGFb RII mutations had a five-year survival of 74% following adjuvant 5-FU therapy compared to 46% in patients with MSI-high tumors without TGFb RII mutations (25). TGFb II MUTATION TGFb is a multifunctional polypeptide that regulates a number of cellular processes including growth, differentiation, deposition of the extracellular matrix (ECM), immunosuppression, embryogenesis, repair of soft and hard tissue, and regulation of hematopoiesis (39–42). TGFb exerts its effects through binding to specific cell surface proteins. These receptors have been termed type I (RI), type II (RII), and type III (RIII) (43–45). Type RI and RII are glycoproteins, while type RIII is a proteoglycan (46). It has been demonstrated that RI and RII are serine/threonine kinases, which are indispensable for TGFb signaling (47,48). Furthermore, cells that lose the ability to express or respond to TGFb are more likely to exhibit uncontrolled growth and over time become tumorigenic. RII, but not RI, can individually bind TGFb. Binding of RII to TGFb induces the assembly of the RII–RI heterodimer and transphosphorylation by RII of RI (48–50). A study of RI or RII mutants has shown that TGFb signaling can be abrogated by mutation of either the RII kinase domain, the RII binding site for TGFb, or sites on RII required for transphosphorylation of RI as well as the RI kinase domain (49,51,52). It has been shown that the TGFb receptors can be inactivated by somatic mutation or gene deletion in several types of human cancers, particularly in colon and gastric cancers (53,54). Certain colon cancers have somatic mutations that inactivate RII (via frameshift mutation); the mutation is found clustered in a 10-base pair polyadenine repeat (709–718), which is located within the RII coding region and known as the BAT-RII mutation (55,56). The BAT-RII mutations introduce early stop codons and encode RII proteins that are truncated between 129 and 161 amino acids compared to 565 amino acids of wild-type RII (57). The BAT-RII mutants are functionally inactive

Pharmacogenomics of Colorectal Cancer

291

and have been identified in seven colon cancer cell lines and associated with the absence of cell surface RII receptors (55). MSI (replication errors or RER) is common to many sporadic forms of cancer and is found in half to two-thirds of colon cancers (58). Furthermore, a study has demonstrated that RII is inactivated by BAT-RII frameshift mutation in 90% of RER colon cancers (56). TGFb RII is a downstream mutation target resulting in the disruption of growth regulation of HNPCC in both cell lines and tissues (55). It has also been observed that re-expression of RII in HCT116 colon cancer cells, which are resistant to TGFb, leads to a reversal of tumorigenicity both in vitro and in vivo (57,59). This suggests that TGFb plays a significant role in the suppression of malignancy, and mutations that inactivate RII are actively selected for in colon cancer and do not randomly accumulate (47,56). Taken together the data suggest that the TGFb pathway provides important targets for therapeutic intervention as well as for chemoprevention strategies. The biologic markers described above appear to offer a degree of precision for determining patient prognosis. This will aid the identification of patients who need more aggressive or experimental therapies. However, there is a need for predictive markers that are specific for individual chemotherapy agents (Table 1). 5-FLUOROURACIL A primary mechanism of action of 5-FU is inhibition of the nucleotide synthetic enzyme thymidylate synthase (TS) by its active metabolite fluorodeoxyuridine monophosphate (FdUMP) resulting in thymidylate depletion, which if prolonged causes apoptosis via the so-called thymineless death (60). TS is a cytosolic enzyme that catalyzes the reductive methylation of deoxyuridine monophosphate (dUMP) to yield deoxythymidine monophosphate, a precursor of deoxythymidine triphosphate, which is required for DNA synthesis and repair (61). FdUMP forms a stable ternary complex with TS and 5,10-methylene tetrahydrofolate (CH2THF), which blocks deoxythymidine monophosphate production, thereby inhibiting DNA synthesis and repair. A number of in vitro studies have demonstrated that TS is a prognostic and a predictive marker in fluoropyrimidine-based chemotherapy. The primary mechanism of resistance to fluoropyrimidines is an increase in TS expression (62). Furthermore, reverse transcription-polymerase chain reaction and immunohistochemical studies have demonstrated that patients with low tumoral TS expression have higher response rates to 5-FU than those with higher levels of TS expression (10–12). Several in vitro and in vivo studies have demonstrated that treatment of cancer cells with 5-FU acutely upregulates TS synthesis (12,63,64). The molecular basis for this acute upregulation appears to be the inhibition of an autoregulatory feedback loop, in

292

Johnston et al.

Table 1 The Potential Predictive Markers for 5-FU, Oxaliplatin, and CPT-11 in Colorectal Cancer Chemotherapy 5-FU

Marker

Comments

TS

Both TS overexpression and the 3R/ 3R TS gene polymorphism lead to resistance to 5-FU Low gene expression levels of TP are associated with response and survival There is an inverse correlation between DPD expression and response to 5-FU There are conflicting results as to the role of p53 as a predictive marker for 5-FU High mRNA expression of the ERCC1 and TS genes has been shown to be predictive of poor response in patients treated with oxaliplatin combined with 5-FU The XPD Lys751Gln polymorphism may be an important marker in the prediction of clinical outcome to oxaliplatin The GSTP1 Ile(105)Val polymorphism is associated with increased survival of patients with advanced colorectal cancer receiving 5-FU/oxaliplatin chemotherapy The polymorphism in the UGT1A1 gene, which is responsible for glucuronidation of the active metabolite of CPT-11, is associated with increased toxicity in patients treated with CPT-11

TP DPD p53 Oxaliplatin

ERCC1

XPD

GST-P1

CPT-11

UGT1A1

Abbreviations: TS, thymidylate synthase; TP, thymidine phosphosylase; DPD, dihydropyrimidine dehydrogenase; XPD, xeroderma pigmentosum group D; ERCC1, excision repair cross complementing 1; GST-P1, glutathione S-transferase P1, UGTIA1, UDP-glucuronosyltransferase 1A1; 5-FU, 5-fluorouracil; CPT-11, Irinotecan.

which ligand-free TS binds to, and inhibits, its own translation. However, when FdUMP stably binds to TS, it is no longer able to bind to its mRNA, resulting in increased TS protein expression (64). The role of acute TS inducion in 5-FU resistance was investigated in MDA435 breast cancer cells in which expression of a TS trans-gene was controlled by a tetracyclineregulated promoter. This study demonstrated that inducible expression of TS increased the IC50 dose of 5-FU by
3-fold, suggesting that acute induction of TS is a factor in determining sensitivity to 5-FU (65). There is, however, a

Pharmacogenomics of Colorectal Cancer

293

subset of patients who do not respond to 5-FU treatment and have low TS expression. These nonresponders may have other mechanisms of resistance at play such as high levels of thymidine phosphorylase (TP) and dihydropyrimidine dehydrogenase (DPD). This highlights the need and importance of measuring multiple markers of resistance to fluoropyrimidines. TS gene polymorphisms have the potential to predict clinical outcome and toxicity; therefore, it is an important factor to consider when deciding on patient treatment. The TS gene promoter is polymorphic and usually has two (TSER2) or three (TSER3) 28-base pair tandem repeat sequences (66). The tandem repeats may affect transcriptional and/or translational efficiency of the TS gene. It has been demonstrated that TSER3/TSER3 homozygous patients are less likely to respond to 5-FU than TSER2/ TSER2 homozygous, or TSER2/TSER3 heterozygous patients (67). This may be due to the fact that TS promoters with the TSER3 sequence have been reported to generate
3-fold higher mRNA than those with the TSER2 sequence (68), and therefore patients with this genotype may express higher levels of TS and be less responsive to 5-FU. This relationship is further enhanced by the presence of a single nucleotide protein (SNP) in the tandem repeat region (69). The identification of the polymorphism provides a means of selecting patients who are likely to respond to 5-FU-based chemotherapy and also identifies patients who will experience increased toxicity. Recently, a six base-pair polymorphic deletion in the 30 UTR of the TS gene has been identified; however, it is unclear at present whether this polymorphic change influences TS gene expression or mRNA stability. Dihydropyrimidine Dehydrogenase Dihydropyrimidine dehydrogenase (DPD) catalyzes the rate-limiting step in the catabolism of fluoropyrimidines. More than 80% of 5-FU administered is degraded in the liver by DPD. Thus, DPD limits the bioavailability of 5-FU (70). DPD is very active in the liver, but is also found in other tissues. Of note, DPD levels are higher in normal mucosa as compared to levels in tumor tissue. DPD has variable activity in human tumors, and tumoral DPD has been reported to be an important determinant of response to 5-FU both in vitro and in vivo (13,14). The differences in variation among patients who receive 5-FU must be due to genetic differences in the activity of the DPD gene. Patients deficient in DPD experience profound systemic toxicity when treated with 5-FU, which may prove fatal (71), as a result of its decreased catabolism of 5-FU resulting in higher systemic levels of 5-FU in patients. The DPD gene is also polymorphic and studies have identified a common G ! A point mutation in the invariant GT splice donor site flanking exon 14 (IVS14 þ 1G > A), which results in the loss of exon 14 and a truncated protein product. The mutated DPD activity is severely compromised,

294

Johnston et al.

especially in individuals homozygous for this mutation where the enzyme activity is virtually absent. Patients who are homozygous for this mutation are at higher risk from toxicity (71–73). There are some questions about the degree of penetrance of this polymorphism, as not all patients with this SNP will experience severe toxicity from 5-FU. More importantly, most patients with 5-FU-induced toxicity do not have the SNP, revealing a degree of sensitivity and specificity that is insufficient for clinical practice. Thymidine Phosphorylase Thymidine phosphorylase (TP) converts 5-FU to fluorodeoxyuridine (FUDR), which can then be converted to the active metabolite FdUMP. Initial preclinical studies demonstrated that increased TP expression correlated with increased sensitivity to 5-FU, probably due to increased synthesis of FUDR (74). However, analysis of TP mRNA expression in 38 colorectal tumors indicated that tumors with high TP were actually less likely to respond to 5-FU (74,75). TP is identical to platelet-derived endothelial cell growth factor, which is a well-established angiogenic factor. Therefore, high TP expression may be a marker for a more invasive and malignant tumor phenotype that is less responsive to chemotherapy. However, in vitro, the angiogenic effects of TP would not appear to be a factor and, the effects of TP on 5-FU activation may predominate with the result that high TP levels correlate with increased 5-FU sensitivity in cell line models. Salonga et al. examined the combined levels of TS, DPD, and TP in a series of colorectal tumors treated with 5-FU (Fig.1). Tumors that responded to 5-FUbased therapy had expression values of all three genes (TS, DPD, TP) that were below the nonresponsive cut-off values. This resulted in this group of patients having a 92% response rate. Those patients whose tumors did not respond had high levels of gene expression for at least one of the markers (13), once again highlighting the importance of multiple marker testing. The p53 Tumor Suppressor Gene The p53 tumor suppressor gene is the most frequently mutated gene in all human cancers and has been described as the universal sensor of genotoxic stress (76). p53 was first identified by the observation that the antibody against SV40 large T antigen coimmunoprecipitated with a cellular protein with a molecular mass of 53 kDa, subsequently called p53. p53 was found to be expressed at low levels in normal cells, but at high levels in transformed cells (77). As p53 is normally a short-lived protein, this difference in p53 expression levels was attributed to the stabilization of the protein in transformed cells (78). Evidence for the role of p53 in apoptotic processes was provided by the study of p53 knockout mice (p53/) (79). p53/ thymocytes and p53/ stem cells of the small and large intestine were found to be significantly more resistant to radiation-induced apoptosis than the

Pharmacogenomics of Colorectal Cancer

295

Figure 1 Salonga et al. analyzed thymidylate synthase (TS), thymidine phosphorylase (TP), and dihydropyrimidine dehydrogenase (DPD) mRNA levels in 33 pretreatment colorectal cancer biopsies from patients who went on to receive 5-FU and leucovorin. Those patients whose tumors did not respond had high levels of gene expression for at least one of the markers. This graph highlights the importance of multiple marker testing. Abbreviations: DPD, dihydropyrimidine dehydrogenase; TP, thymidine phosphosylase; TS, thymidylate synthase.

corresponding p53þ/þ cells. Furthermore, Lowe et al. (80) showed that p53deficient mouse embryonic fibroblasts were resistant to apoptosis when treated with various chemotherapies, including 5-FU. Longley et al. also established that p53 played a role in resistance to 5-FU. They developed two MCF-7 breast cancer cell lines with tetracyclineregulated expression of TS. The initial p53 wild-type line termed M7TS90 was transfected with HPV-E6 to generate the p53 null M7TS90-E6 cell line. Comparison of these isogenic cell lines demonstrated that inactivation of p53 significantly increased resistance to 5-FU. p53 also abrogated the cell cycle arrest and apoptosis induced by 5-FU (81). This study also agreed with that of Bunz et al. (82), who found that 5-FU-mediated apoptosis was attenuated in p53 null HCT116 colon cancer cells compared to the isogenic p53 wild-type parental lines. These in vitro findings have also been borne out in several clinical studies. For example, Ahnen et al. (83) found that patients with stage III colorectal cancer whose tumors overexpressed p53 did not derive significant survival benefit from adjuvant 5-FU-based treatment, whereas those without p53 overexpression did. p53 overexpression was used as a surrogate marker for p53 mutation in a number of studies. These studies found that p53 overexpression correlated with resistance (83–85). However, a number of other studies have found no correlation between p53 overexpression and response to 5-FU (25,86). Studies on xenografts have demonstrated that

296

Johnston et al.

only a small fraction of the nuclei of tumors with wild-type p53 alleles show p53 staining (87). These conflicting findings may be due to variations among the immunohistochemical protocols and also due to the use of different antibodies used to detect p53. Moreover, p53 overexpression does not necessarily result from p53 mutation in up to at least 30% to 40% of cases (88). Other technologies such as DNA sequencing or single-strand conformation polymorphism may provide a better assessment of p53 status. A recent sequencing study found that there was a significant correlation between p53 mutation and response to 5-FU (89). Due to these conflicting results the use of p53 as a predictive marker for 5-FU remains controversial. OXALIPLATIN Oxaliplatin is a third-generation platinum compound first synthesized in Japan in 1969. It was developed as one of several 1, 2-diaminocyclohexane (1, 2-DACH) platinums, in an attempt to generate compounds with a more favorable therapeutic index than cisplatin and carboplatin (90). Cytotoxic platinum compounds are activated by an aquation reaction in which the leaving group is replaced by water, forming a positively charged species that cross-links DNA and eventually leads to cytotoxicity (91). Oxaliplatin differs from both cisplatin and carboplatin, with respect to its carrier ligand (1, 2-DACH), causing it to form different Pt-DNA adducts. Oxaliplatin forms monoadducts that are eventually converted to diadducts over time; however, the 1, 2-DACH carrier ligands slow the conversion of the monoadduct to the diadduct (92,93). Due to its bulky 1, 2-DACH side chain the platinum adducts produced by oxaliplatin appear to be more resistant to DNA repair mechanisms than cisplatin. This may be due to the oxaliplatin– DNA adducts producing a greater deformation of DNA structure compared to cisplatin adducts (90,94). Interestingly, there appears to be no crossresistance between oxaliplatin and cisplatin in vitro (95,96). As a single agent, oxaliplatin has shown only modest response rates; however, a synergistic interaction has been documented between oxaliplatin and 5-FU. The combination of oxaliplatin and 5-FU/leucovorin in first-line treatment of advanced colorectal cancer has demonstrated response rates of 50% to 53% compared to 16% in patients treated with 5-FU/leucovorin alone (97). Oxaliplatin has proved effective as both a first- and second-line treatment for advanced colorectal cancer (98), and is now considered the standard in first-line treatment of patients with metastatic disease. Excision Repair Cross Complementing 1 Several important mechanisms have been identified that play a role in the development of resistance to oxaliplatin. These include decreased drug accumulation, drug inactivation, enhanced tolerance to platinum-DNA adducts,

Pharmacogenomics of Colorectal Cancer

297

and enhanced DNA repair (99). One of the major DNA repair systems in mammalian cells is the nucleotide excision repair (NER) pathway. NER removes bulky helix-distorting adducts produced by oxaliplatin. Excision repair cross complementing 1 (ERCC1) is a highly conserved protein and is an essential member of the NER pathway (100). The ERCC1XPF complex is involved in the cleavage of damaged DNA 50 to the DNA lesion. It has been shown that low ERCC1 gene expression levels have correlated with improved overall survival after 5-FU/oxaliplatin therapy in patients with advanced colorectal cancer refractory to first-line chemotherapy (101). Furthermore, an independent study has demonstrated that both low TS and low ERCC1 mRNA expressions are associated with significantly improved survival in patients treated with 5FU/oxaliplatin (101). The ERCC1 gene contains a common polymorphism at codon 118 (exon 4). This polymorphism is a single nucleotide change C ! T, which results in the same amino acid, asparginine. As it is a silent mutation, it is difficult to understand how it can influence mRNA levels. However, in a study of 32 patients, Park et al. (102) demonstrated that as the number of T alleles increased, a trend towards higher intratumoral ERCC1 levels were observed. A second polymorphism occurs in the 30 UTR at position 8092, which causes a nucleotide change of C ! A. This polymorphism has been shown to have some correlation with overall survival, whereby patients with the A allele have a survival benefit (102). Xeroderma Pigmentosum Group D The xeroderma pigmentosum group D (XPD) gene, also known as ERCC2, encodes a helicase that is a component of the transcription factor TFIIH and also an essential member of the NER pathway (103). The importance of XPD in platinum drug resistance has not been clearly established, and may be multifaceted, including an impact on NER pathways or cross talk with other factors involved in DNA repair, which are independent of NER. There are several common polymorphisms in the XPD gene with potential functional significance (104). There is a common polymorphism in the XPD gene at codon 751, which results in a nucleotide change of A ! C (lysine to glutamine). In one study patients with the A/A genotype had a higher response rate to chemotherapy and superior overall survival when they were treated with 5-FU/oxaliplatin compared to patients with the A/C or C/C genotypes (105). Glutathione S-Transferase P1 Glutathione S-transferase P1 (GST-P1) belongs to a superfamily of dimeric phase II metabolic enzymes and has been shown to be overexpressed in colorectal cancer tissues (106). GST-P1 catalyzes the conjugation of Glutathione (GSH) to a wide variety of toxic compounds, including platinum agents,

298

Johnston et al.

forming less toxic and more water-soluble conjugates that are exported out of cells. Therefore, GST-P1 plays an important role in detoxification (107). GST-P1 contains a single nucleotide polymorphism at codon 105, which results in a substitution of A ! C. The A ! C substitution results in an amino acid change of isoleucine to valine, which results in severely diminished enzymatic activity (108). In a retrospective analysis of 107 patients with refractory metastatic colorectal cancer treated with 5-FU/oxaliplatin, patients who possessed the valine allele had superior survival benefit (109).

IRINOTECAN Irinotecan (7-ethyl-10-[4(-1-piperidino)-1-piperidino] (CPT-11) carbonyloxy camptothecin) is a semisynthetic, water-soluble derivative of the plant alkaloid camptothecin. Irinotecan is a DNA topoisomerase I (DNA topo 1) inhibitor and was first introduced into the clinic in the late 1980s (110). DNA topo 1 is involved in the relaxation and recombination of torsionally strained supercoiled duplex DNA during replication and transcription (111). Inhibition of topo 1 induces DNA damage, resulting in cell death (112). Irinotecan is converted to SN-38 (7-ethyl-10-hydroxy-camptothecan) by carboxylesterases, which are abundant in the liver and also found in other tissues (113). SN-38 is 100–1000-fold more biologically active than irinotecan and exerts its cytotoxicity by trapping the complexes formed by topo 1 with DNA. The single-strand DNA breaks generated by SN-38 are not toxic, as they are highly reversible and rapidly repaired once the drug is removed. Lethal irreversible DNA damage occurs when DNA synthesis is ongoing and the replication fork encounters the DNA topo 1 complex, causing a double-strand break that can lead to cell death (114). A Japanese study demonstrated that irinotecan as a first-line therapy produced response rates of 22% to 25% in advanced colorectal cancer patients who had been pretreated with 5-FU/leucovorin (115). A multicenter randomized trial of 387 advanced colorectal cancer patients was set up to compare the combinations of 5-FU/leucovorin/irinotecan against 5-FU/ leucovorin in patients with previously untreated disease (8). The response rates and overall survival were more favorable for the 5-FU/leucovorin/irinotecan combination (35% vs. 22%, and 17.4 vs. 14.1 months, respectively), demonstrating the usefulness of irinotecan in combination with 5-FU and leucovorin as a first-line therapy in colorectal cancer. UDP-Glucuronosyl-Transferase 1A1 Hepatic UDP-glucuronosyl-transferase 1A1 (UGT1A1) glucuronidates SN38 to form the more polar and inactive glucuronide, which is eliminated in bile and urine (116).

Pharmacogenomics of Colorectal Cancer

299

The UGT1A1 promoter contains a number of TA repeats, with between five and eight repeats observed in the general population. A six repeat allele is most common, and an inverse relationship exists between the number of repeats observed and the expression of UGT1A1 (117,118). The presence of seven TA repeats instead of six TA repeats results in the variant allele UGT1A128. The UGT1A128 variant leads to decreased expression of UGT1A1 and therefore decreased SN-38 glucuronidation (119). Recently, it has been demonstrated that the UGT1A128 allele leads to significantly increased amounts of the irinotecan active metabolite SN-38, and also increased chance of developing severe toxicity from irinotecan therapy (119,120). The risk of leucopenia in patients homozygous for the UGT1A128 allele is as high as 50% with the 350 mg/m2 every-three-weeks regimen, but the genotype–toxicity relationship is also observed with the weekly schedule (69,121,122). Recently, a study has described an apparent influence of polymorphisms in UGT1A7 and UGT1A9 on toxicity and tumor response in patients receiving capecitabine/irinotecan (123). Prospective trials are now in progress to provide more definitive dosing guidelines for irinotecan, based on the UGT1A1 genotype. NOVEL THERAPIES Epidermal Growth Factor Receptor Epidermal growth factor receptor (EGFR) (c-erbB-1) is a type 1 receptor tyrosine kinase, which signals through PI3K/PKC, MAPK, and STAT3 to induce proliferation, cell cycle progression, and inhibition of apoptosis (124). Overexpression and/or mutation of EGFR has been detected in a number of human cancers and is often associated with aggressive disease and poor prognosis (125). More importantly, EGFR expression has been detected in 60% to 75% of colorectal cancer (126,127). EGFR activation can result from mutation of the receptor, overexpression, or by EGFR stimulation through autocrine loops involving excess production of its growth factors (127–129). A study by McKay et al. demonstrated that EGFR overexpression was a common event in colorectal carcinogenesis with high levels observed in 123/249 patients (49.4%). The results indicated that although they expressed high EGFR levels this did not influence patient prognosis and the equivalent expression was not displayed in metastatic tumors (130). These results question the effectiveness of EGFR in the advanced setting. A polymorphic (CA)n dinucleotide repeat is observed in intron 1 of the EGFR gene, which has been shown to be associated with gene expression (131). It has been shown that as the number of (CA)n repeats increases the level of transcription decreases (132). McKay et al. examined the (CA)n dinucleotide repeat length polymorphism in 114 colorectal tumors. They found that there was no association between the repeat length and EGFR protein expression as assessed by immunohistochemistry (130).

300

Johnston et al.

Several tyrosine kinase inhibitors and monoclonal antibodies have been developed to target EGFR. Iressa (ZD-1839), a tyrosine kinase inhibitor, is the most advanced in terms of clinical trials; IMC-C225, a monoclonal antibody targeted to EGFR, has shown antitumor activity in a wide variety of tumor types. Although EGFR inhibitors have shown encouraging activity, their response rates have been limited. Therefore, it is logical to combine these inhibitors with chemotherapy in an attempt to improve response rates. Iressa is a reversible anilinoquinazoline inhibitor that has been shown to inhibit EGFR phosphorylation and activity and thus the production of angiogenic growth factors (125). An in vitro study was established to provide preclinical evidence for optimizing the combination of Iressa and CPT-11. They found that the combination of Iressa and SN-38 was schedule and concentration dependent in the HT-29 and LoVo colon cancer cells. It was demonstrated that the most synergistic schedule was Iressa after SN-38; the combination resulted in the inhibition of topo 1, S phase arrest, and induction of apoptosis (133). A study carried out by Koizumi et al. combined Iressa and CPT-11 in a panel of human colon cancer cell lines. They demonstrated that the LoVo cells had the highest level of protein and autophosphorylation of EGFR and were the most sensitive to Iressa. CPT-11 and Iressa resulted in a supra-additive inhibitory effect in the COLO320DM, WiDR, and LoVo cells. CPT-11 increased the phosphorylation of EGFR in the LoVo and WiDR cells in a time- and dose-dependent manner. These results imply that Iressa and CPT-11 will be effective against colorectal tumor cells that express high levels of EGFR (134). The effect of Iressa on the radiation response of LoVo colon cancer cells has also been evaluated. Iressa significantly increased the antiproliferative effect of radiation treatment both in vitro and in vivo. The radiopotentiating effects of Iressa were more prominent when the radiation was delivered as a fractionated dose (135). A phase I/II study was carried out in metastatic colorectal cancer patients in which serial biopsies were taken pre- and post-treatment; 27 patients participated but paired data were available only from 17. Post-treatment samples showed a statistically significant reduction in the cancer cell proliferation index (31% pre- vs. 21% post-treatment). All pretreatment samples showed strong EGFR staining; moreover, phospho-EGFR, phospho-Akt, and phospho-ERK were lost in some patients after Iressa treatment. Several patients showed an induction in p27Kip1 staining and also had an increased apoptotic index. This study concluded that Iressa inhibited EGFR phosphorylation and proliferation in cancer cells of patients with metastatic colorectal cancer and that it may induce apoptosis through upregulation of p27Kip1 (136). C225 (Erbitux) is a chimeric IgG1 monoclonal antibody that binds competitively to the extracellular domain of EGFR, inhibiting EGF binding and receptor autophosphorylation, and inducing its internalization and

Pharmacogenomics of Colorectal Cancer

301

degradation (137). It has also been noted that C225 blocks the production of proangiogenic factors such as vascular endothelial growth factor (VEGF) and IL-8 (138). Ciardiello et al. carried out an in vivo study that focused on the antitumor and antiangiogenic activities of C225 and a human VEGF antisense oligonucleotide in human GEO colon cancer cells. They found that treatment with either agent alone in athymic mice resulted in a cytostatic reversible inhibition of tumor growth, whereas when both agents were combined there was a prolonged inhibition of tumor growth and significant improvement in mice survival compared to control or either agent alone (139). Furthermore, the combination of CPT-11 and C225 in HT-29 and DLD-1 xenografts resulted in a significant inhibition of tumor growth compared to either therapy alone. The combination of CPT-11 and C225 resulted in a regression rate of 100% in DLD-1 xenografts and 60% in HT-29 xenografts, respectively (140). Several phase II clinical trials have been carried out to investigate the effect of C225 alone and in combination with other chemotherapies. In 57 patients who had CPT-11 refractory and EGFR-positive metastatic colorectal cancer, treatment with C225 alone resulted in 10.5% showing a partial response and 36.8% with disease stabilization (141). Saltz et al. carried out a study in 121 patients who were CPT-11/5-FU refractory and EGFR positive. The patients received CPT-11 and C225 in combination, which resulted in 17% partial response and 31% with disease stabilization (141). The combination of C225 with combined 5-FU, leucovorin and irinotecan therapy (FOLFIRI) resulted in a 44% partial response and 22% disease stabilization in patients who had EGFR-positive metastatic colorectal cancer (142). Finally, Cunningham et al. carried out a randomized phase III trial that contained 329 patients with CPT-11 refractory, EGFR-positive colorectal cancer. They received either C225 alone or C225 in combination with CPT-11. The response rate was 9.9% versus 17.9%, respectively, and the time to progression was 45 versus 126 days, respectively (143). These trials provide evidence that C225 has activity both as a single agent and in combination with chemotherapy. An in vitro and in vivo study has recently demonstrated that combined treatment with suboptimal doses of Iressa and C225 resulted in a synergistic effect on cell proliferation and in superior inhibition of EGFR-dependent signaling and induction of apoptosis. The combination also results in complete and permanent regression of large tumor xenografts, whereas single agents result in transient tumor remission only at high doses (144). This study suggests that combined receptor targeting may be superior over single-agent receptor targeting. VASCULAR ENDOTHELIAL GROWTH FACTOR VEGF is the ligand for two tyrosine kinase receptors expressed on vascular endothelial cells, VEGF receptor 1 (VEGF R1) and VEGF R2. Binding

302

Johnston et al.

of VEGF to its receptors leads to endothelial cell proliferation, increased vascular permeability, mitogenesis, and chemotaxis (145). Many solid tumors secrete high levels of VEGF, which promotes their vascularization and initiates formation of metastases (146). It has been shown that increased VEGF expression correlates with tumor stage and poorer prognosis in colorectal cancer (147). VEGF is upregulated by hypoxia, growth factors such as EGF and TGFb, and oncogenes such as Ras (148,149). Ishigami et al. (150) have shown that the recurrence rate of resected stage III VEGF-positive colon cancer is 4.5 times higher than VEGFnegative colon cancer. Bevacizumab, which is a recombinant humanized monoclonal antibody against VEGF, is currently in clinical trials. The antibody inhibits the binding of VEGF to its endothelial cell receptors. A phase II study was established, which evaluated bevacizumab at two different doses in combination with 5-FU/LV. The trial contained 104 previously untreated patients with metastatic colorectal cancer, who received either 5-FU/LV alone, 5-FU/LV in combination with low-dose bevacizumab, or 5-FU/ LV in combination with high-dose bevacizumab. Higher response rates were seen in the bevacizumab-treated arms (17% vs. 40% vs. 24%, respectively). The median time to progression was also longer in the bevacizumab-treated arms (5.2 vs. 9.0 vs. 7.2 months, respectively) and the median survival was 13.8 versus 21.5 versus 16.1 months, respectively. It is interesting to note that only the low-dose bevacizumab reached statistical significance compared to 5-FU/LV alone (151). A phase III trial was carried out, which compared IFL (CPT-11/5-FU/LV) and placebo against IFL and bevacizumab in 815 patients with metastatic colorectal cancer. The combination of IFL with bevacizumab proved to be a superior combination with duration of response of 10.4 versus 7.1 months, progression-free survival of 10.6 versus 6.24 months, and overall survival of 20.3 versus 15.6 months (152). NEW TECHNOLOGIES The advent of high-throughput methodologies such as microarray-based gene expression profiling, proteomic profiling, comparative genomic hybridization (CGH) analysis, and the newly developed metabolomics enable tumor samples to be profiled on a global scale. This has major implications for the diagnostic capability and prognostic classification of tumors, where we can ultimately predict response of each individual tumor to chemotherapy. Whereas microarray expression profiling of colorectal cancer has been preformed, no comparable protein analysis has been reported. It would be of great interest to identify a group of consistently changing proteins whose functions may reveal insight into critical events in disease progression and which may hold value as potential therapeutic targets (153). It is important to also investigate the proteomic profile, as mRNA levels may not correlate with the amount of active protein within the cell. Furthermore, the gene

Pharmacogenomics of Colorectal Cancer

303

sequence does not describe the post-translational modifications that may be essential for protein function and activity, and finally the study of the genome does not provide information on dynamic cellular processes (154). A study by Friedman et al. made use of 2D difference gel electrophoresis coupled with mass spectrometry to investigate tumor-specific changes in the proteosome of human colorectal cancer and normal mucosa. The study was extremely small, as it contained only six samples, but the improved method has the potential to facilitate future proteomic profiling of large sets of colorectal cancer samples. In this method, the tumor samples and normal samples are labeled with cy3 and cy5 fluorescent dyes. The method allows each patient comparison to be preformed on proteins resolved in the same 2D gel separation, thereby removing the error caused by gel–gel variation and allows quantification of abundance of change for each protein pair (153). CGH allows the entire genome to be scanned, in a single step, for copy number aberrations in chromosomal material (155). CGH identifies specific chromosomal regions that are consistently gained or lost at a high frequency within colorectal cancer and has demonstrated an increase in the genetic grade of a tumor with disease progression (156,157). Rooney et al. carried out CGH analysis in 29 stage III colorectal cancer samples to assess any genomic aberrations and the overall level of chromosome instability. Also, eight colorectal cancer cell lines were used to evaluate their usefulness as model systems for colorectal cancer genomics. They found a high level of variation between both the number and type of genetic aberration detected in the 29 colorectal cancer samples and found that almost every chromosomal arm was detected as changed in at least one genome that was assessed. They were unable to demonstrate an association between any specific chromosomal aberration and patient survival; however, they did find a link between the number of aberrations and survival, with greater than two aberrations resulting in a better survival benefit (158). The most frequently used genome-wide approach in colorectal cancer is DNA microarray profiling. Maxwell et al. carried out a study to identify novel downstream mediators of tumor cell response to 5-FU. They found that 619 genes were upregulated greater than threefold by 5-FU. The genes that were consistently upregulated were spermine/speridine acetyl transferase, which was involved in polyamine metabolism; annexin II, which was involved in DNA synthesis, cell proliferation, and apoptosis; thymosin b-10, which was a G actin binding protein involved in apoptosis; chaperonin 10, which was a heat shock protein involved in the folding of mitochondrial proteins; and MAT-8, which was a member of the FXYD family of proteins that regulate chloride ion transport. They further demonstrated that these genes were upregulated in response to both tomudex and oxaliplatin and furthermore that the basal levels of these genes were upregulated in a 5-FU-resistant colorectal cancer cell line compared to the parental line (159). Mariadason et al. carried out gene expression profiling on 30 colorectal

304

Johnston et al.

cancer cell lines and correlated this with 5-FU sensitivity using three different assays of response. They were able to identify panels of genes that correlated with 5-FU sensitivity and further used leave-one-out cross-validation to demonstrate that these genes were predictive for 5-FU response. They noted that this gene set had a greater power to predict response than four ‘‘classical’’ determinants of 5-FU response: TS, TP, p53, and MMR status. Furthermore, they repeated the correlation analysis for sensitivity to CPT-11 and this generated a second gene set that showed great predictive power for sensitivity to CPT-11 (160). A second study then carried out gene expression profiling on 30 different colorectal cancer cell lines to select genes that could predict the apoptotic, response to oxaliplatin. Again the investigators used a leave-one-out cross-validation approach to determine this. They demonstrated that 80 genes best correlated with oxaliplatin-induced apoptosis, and that this gene set produced the most accurate prediction of oxaliplatin response (161). An interesting study was carried out by Clarke et al., which compared the gene expression profiles pre- and during treatment in patients who had rectal cancer. The patients were receiving 5-FU in combination with mitomycin C and fractionated radiation. The investigators wanted to assess if it was possible to use microarray profiling to detect altered gene expression in solid tumors in response to treatment. They found that altered gene expression profiles were apparent in all patients after treatment, with a major cluster of genes involved in synthesis and metabolism downregulated after treatment. They also found that there was downregulation in genes involved in RNA and protein synthesis and processing and in cellular metabolism. Importantly, one-third of the genes was found to be positively regulated by c-Myc; they also noted that c-Myc was downregulated following treatment (162). This is an interesting paper, as it highlights the possibilities of microarray profiling and has demonstrated that it is possible to detect global gene expression changes within solid tumor tissue in response to ongoing treatment. Lastly, Wang et al. also used gene expression profiling to identify markers for stage II colorectal cancer. The study contained 74 patients with stage II colorectal cancer. They used two supervised class prediction approaches to select markers from the 17,616 informative genes from the microarray. Firstly, the patients were divided equally into a training set and a test set. The training set was used to select markers and build a prognostic signature, while the test set was used to independently validate the training set. This approach yielded 60 genes from 38 patients. Secondly, the patients were divided into one of two groups based on unsupervised clustering results. From that, each subgroup was further divided into a training set and a test set and again were analyzed to select markers. This approach yielded 23 markers from the training set, which were then analyzed in the test set. The investigators then compared the predictive power

Pharmacogenomics of Colorectal Cancer

305

Figure 2 Flow charts that developed two supervised class prediction approaches to select markers from their microarray. (A) The first approach used the 74 patients as a single group. (B) The second approach identified patient subgroups and combined the markers to form a single signature for all patients as a whole. Abbreviations: S, disease free; R, relapse. Source: From Ref. 163.

of the 23 gene set and the 60 gene set and found that only the 23 gene set was predictive. The 23 gene set was then further validated in 36 independent patients and demonstrated an overall accuracy of 78% (Fig. 2) (163). This study highlights the power of predictive marker testing and also the need to carefully select the correct analysis for the purpose of the test.

306

Johnston et al.

The final and most recently developed high-throughput technology is metabolomics. Metabolomics is the global study of all the small molecules produced in cells, tissues, and organisms. While proteomic analysis allows the abundance and distribution of many proteins to be determined simultaneously, the functional consequences of changes to the proteome are only reported indirectly. However, a more direct approach is to measure the levels of these small molecules (e.g., enzyme substrates) or metabolites. These profiles are typically generated by high-throughput nuclear magnetic resonance spectroscopy and mass spectroscopy. These technologies have been used for decades for the study of metabolites, but what makes metabolomics different from these earlier efforts is the focus on the global metabolite profile rather than a limited sample. To date, metabolomic profiling has only been carried out on plants, and there are no known reports of metabolomics data on human tissue or cell lines derived from them. CONCLUSIONS This limited use of predictive markers highlights the importance and need for multiple marker testing in order to improve response rates and decrease toxicity. It has been previously demonstrated how a small number of genes can exert major effects on drug response, but it will be the case that the identification of key polymorphic genes and environmental factors will ultimately lead to the ability to predict enhanced response to chemotherapy while minimizing drug toxicity. The area of pharmacogenomics will lead to an era of predictive, individualized medicine and will undoubtedly make a huge contribution in the field of oncology. The ultimate goal of pharmacogenomics is to maximize the efficacy of therapeutic agents while minimizing their toxicity. Therefore, pharmacogenomics should help to eliminate the risk of ‘‘unpredictable’’ toxicities by prospectively identifying susceptible individuals who can receive individualized drug regimes based on their risk. The next step in the area of pharmacogenomics is to develop clinical trials that will assess prospectively the benefits of profiling a patient’s particular tumor, which should translate into improvements in both overall response and toxicity. If the goal of pharmacogenomics is realized, a new era of individualized or more rationale treatment will become a reality with an enhanced overall response rate and survival benefit for patients. REFERENCES 1. Desai AA, Innocenti F, et al. Pharmacogenomics: road to anticancer therapeutics nirvana? Oncogene 2003; 22(42):6621–6628. 2. Evans WE, Relling MV. Pharmacogenomics: translating functional genomics into rational therapeutics. Science 1999; 286(5439):487–491.

Pharmacogenomics of Colorectal Cancer

307

3. Ratain MJ, Relling MV. Gazing into a crystal ball-cancer therapy in the postgenomic era. Nat Med 2001; 7(3):283–285. 4. Innocenti F, Ratain MJ. Update on pharmacogenetics in cancer chemotherapy. Eur J Cancer 2002; 38(5):639–644. 5. Watters JW, McLeod HL. Cancer pharmacogenomics: current and future applications. Biochim Biophys Acta 2003; 1603(2):99–111. 6. Iqbal S, Lenz HJ. Targeted therapy and pharmacogenomic programs. Cancer 2003; 97(suppl 8):2076–2082. 7. Johnston PG, Kaye S. Capecitabine: a novel agent for the treatment of solid tumors. Anticancer Drugs 2001; 12(8):639–646. 8. Douillard JY, Cunningham D, et al. Irinotecan combined with fluorouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal cancer: a multicentre randomised trial. Lancet 2000; 355(9209):1041–1047. 9. Giacchetti SB, Perpoint, et al. Phase III multicenter randomized trial of oxaliplatin added to chronomodulated fluorouracil-leucovorin as first-line treatment of metastatic colorectal cancer. J Clin Oncol 2000; 18(1):136–147. 10. Johnston PG, Lenz HJ, et al. Thymidylate synthase gene and protein expression correlate and are associated with response to 5-fluorouracil in human colorectal and gastric tumors. Cancer Res 1995; 55(7):1407–1412. 11. Edler D, Blomgren H, et al. Immunohistochemical determination of thymidylate synthase in colorectal cancer—methodological studies. Eur J Cancer 1997; 33(13):2278–2281. 12. Lenz HJ, Danenberg KD, et al. p53 and thymidylate synthase expression in untreated stage II colon cancer: associations with recurrence, survival, and site. Clin Cancer Res 1998; 4(5):1227–1234. 13. Salonga D, Danenberg KD, et al. Colorectal tumors responding to 5-fluorouracil have low gene expression levels of dihydropyrimidine dehydrogenase, thymidylate synthase, and thymidine phosphorylase. Clin Cancer Res 2000; 6(4):1322–1327. 14. Takebe N, Zhao SC, et al. Retroviral transduction of human dihydropyrimidine dehydrogenase cDNA confers resistance to 5-fluorouracil in murine hematopoietic progenitor cells and human CD34þ-enriched peripheral blood progenitor cells. Cancer Gene Ther 2001; 8(12):966–973. 15. Vogelstein B, Fearon ER, et al. Genetic alterations during colorectal-tumor development. N Engl J Med 1988; 319(9):525–532. 16. Powell SM, Zilz N, et al. APC mutations occur early during colorectal tumorigenesis. Nature 1992; 359(6392):235–237. 17. Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990; 61(5):759–767. 18. Boland CR, Sato J, et al. Microallelotyping defines the sequence and tempo of allelic losses at tumour suppressor gene loci during colorectal cancer progression. Nat Med 1995; 1(9):902–909. 19. Herbergs J, Hopman AH, et al. In situ hybridization and flow cytometric analysis of colorectal tumours suggests two routes of tumourigenesis characterized by gain of chromosome 7 or loss of chromosomes 17 and 18. J Pathol 1996; 179(3):243–247. 20. Yamagata S, Muto T, et al. Polypoid growth and K-ras codon 12 mutation in colorectal cancer. Cancer 1995; 75(4):953–957.

308

Johnston et al.

21. Isogaki J, Shinmura K, et al. Microsatellite instability and K-ras mutations in gastric adenomas, with reference to associated gastric cancers. Cancer Detect Prev 1999; 23(3):204–214. 22. Kern SE, Fearon ER, et al. Clinical and pathological associations with allelic loss in colorectal carcinoma [corrected]. JAMA 1989; 261(21):3099–3103. 23. Miyaki M, Iijima T, et al. High incidence of protein-truncating mutations of the p53 gene in liver metastases of colorectal carcinomas. Oncogene 2002; 21(43):6689–6693. 24. Lanza G, Gafa R, et al. Prognostic significance of DNA ploidy in patients with stage II and stage III colon carcinoma: a prospective flow cytometric study. Cancer 1998; 82(1):49–59. 25. Watanabe T, Wu TT, et al. Molecular predictors of survival after adjuvant chemotherapy for colon cancer. N Engl J Med 2001; 344(16):1196–1206. 26. Papadimitrakopoulou VA, Oh Y, et al. Presence of multiple incontiguous deleted regions at the long arm of chromosome 18 in head and neck cancer. Clin Cancer Res 1998; 4(3):539–544. 27. Gotley DC, Reeder JA, et al. The deleted in colon cancer (DCC) gene is consistently expressed in colorectal cancers and metastases. Oncogene 1996; 13(4): 787–795. 28. Ijichi H, Ikenoue T, et al. Systematic analysis of the TGF-beta-Smad signaling pathway in gastrointestinal cancer cells. Biochem Biophys Res Commun 2001; 289(2):350–357. 29. Yagi K, Furuhashi M, et al. c-myc is a downstream target of the Smad pathway. J Biol Chem 2002; 277(1):854–861. 30. Frederick JP, Liberati NT, et al. Transforming growth factor beta-mediated transcriptional repression of c-myc is dependent on direct binding of Smad3 to a novel repressive Smad binding element. Mol Cell Biol 2004; 24(6): 2546–2559. 31. Paredes-Zaglul A, Kang JJ, et al. Analysis of colorectal cancer by comparative genomic hybridization: evidence for induction of the metastatic phenotype by loss of tumor suppressor genes. Clin Cancer Res 1998; 4(4):879–886. 32. Remvikos Y, Muleris M, et al. Colorectal carcinogenesis: from chromosomal evolution pathways to molecular pathogenesis. Cancer Genet Cytogenet 1997; 93(1):63–73. 33. Chai SM, Zeps N, et al. Screening for defective DNA mismatch repair in stage II and III colorectal cancer patients. Clin Gastroenterol Hepatol 2004; 2(11): 1017–1025. 34. Liu B, Parsons R, et al. Analysis of mismatch repair genes in hereditary nonpolyposis colorectal cancer patients. Nat Med 1996; 2(2):169–174. 35. Jacob S, Praz F. DNA mismatch repair defects: role in colorectal carcinogenesis. Biochimie 2002; 84(1):27–47. 36. Meyers M, Wagner MW, et al. Role of the hMLH1 DNA mismatch repair protein in fluoropyrimidine-mediated cell death and cell cycle responses. Cancer Res 2001; 61(13):5193–5201. 37. Elsaleh H, Shannon B, et al. Microsatellite instability as a molecular marker for very good survival in colorectal cancer patients receiving adjuvant chemotherapy. Gastroenterology 2001; 120(5):1309–1310.

Pharmacogenomics of Colorectal Cancer

309

38. Ribic CM, Sargent DJ, et al. Tumor microsatellite-instability status as a predictor of benefit from fluorouracil-based adjuvant chemotherapy for colon cancer. N Engl J Med 2003; 349(3):247–257. 39. Massague J, Cheifetz S, et al. TGF-beta receptors and TGF-beta binding proteoglycans: recent progress in identifying their functional properties. Ann N Y Acad Sci 1990; 593:59–72. 40. Roberts AB, Flanders KC, et al. Transforming growth factor-beta: multifunctional regulator of differentiation and development. Philos Trans R Soc Lond B Biol Sci 1990; 327(1239):145–154. 41. Shull MM, Ormsby I, et al. Targeted disruption of the mouse transforming growth factor-beta 1 gene results in multifocal inflammatory disease. Nature 1992; 359(6397):693–699. 42. Kulkarni AB, Huh CG, et al. Transforming growth factor beta 1 null mutation in mice causes excessive inflammatory response and early death. Proc Natl Acad Sci USA 1993; 90(2):770–774. 43. Boyd FT, Massague J. Transforming growth factor-beta inhibition of epithelial cell proliferation linked to the expression of a 53-kDa membrane receptor. J Biol Chem 1989; 264(4):2272–2278. 44. Lin HY, Lodish HF. Receptors for the TGF-beta superfamily: multiple polypeptides and serine/threonine kinases. Trends Cell Biol 1993; 3(1):14–19. 45. Moustakas A, Lin HY, et al. The transforming growth factor beta receptors types I, II, and III form hetero-oligomeric complexes in the presence of ligand. J Biol Chem 1993; 268(30):22215–22218. 46. Lopez-Casillas F, Wrana JL, et al. Betaglycan presents ligand to the TGF beta signaling receptor. Cell 1993; 73(7):1435–1444. 47. Lin JK, Chou CK. In vitro apoptosis in the human hepatoma cell line induced by transforming growth factor beta 1. Cancer Res 1992; 52(2):385–388. 48. Wrana JL, Tran H, et al. Two distinct transmembrane serine/threonine kinases from Drosophila melanogaster form an activin receptor complex. Mol Cell Biol 1994; 14(2):944–950. 49. Wrana JL, Attisano L, et al. TGF beta signals through a heteromeric protein kinase receptor complex. Cell 1992; 71(6):1003–1014. 50. Wrana JL, Attisano L, et al. Mechanism of activation of the TGF-beta receptor. Nature 1994; 370(6488):341–347. 51. Wieser R, Attisano L, et al. Signaling activity of transforming growth factor beta type II receptors lacking specific domains in the cytoplasmic region. Mol Cell Biol 1993; 13(12):7239–7247. 52. Carcamo J, Zentella A, et al. Disruption of transforming growth factor beta signaling by a mutation that prevents transphosphorylation within the receptor complex. Mol Cell Biol 1995; 15(3):1573–1581. 53. Park K, Kim SJ, et al. Genetic changes in the transforming growth factor beta (TGF-beta) type II receptor gene in human gastric cancer cells: correlation with sensitivity to growth inhibition by TGF-beta. Proc Natl Acad Sci USA 1994; 91(19):8772–8776. 54. Garrigue-Antar L, Munoz-Antonia T, et al. Missense mutations of the transforming growth factor beta type II receptor in human head and neck squamous carcinoma cells. Cancer Res 1995; 55(18):3982–3987.

310

Johnston et al.

55. Markowitz S, Wang J, et al. Inactivation of the type II TGF-beta receptor in colon cancer cells with microsatellite instability. Science 1995; 268(5215): 1336–1338. 56. Parsons R, Myeroff LL, et al. Microsatellite instability and mutations of the transforming growth factor beta type II receptor gene in colorectal cancer. Cancer Res 1995; 55(23):5548–5550. 57. Wang J, Sun L, et al. Demonstration that mutation of the type II transforming growth factor beta receptor inactivates its tumor suppressor activity in replication error-positive colon carcinoma cells. J Biol Chem 1995; 270(37):22,044–22,049. 58. Lim SB, Jeong SY, et al. Prognostic significance of microsatellite instability in sporadic colorectal cancer. Int J Colorect Dis 2004; 19(6):533–537. 59. Geiser AG, Burmester JK, et al. Inhibition of growth by transforming growth factor-beta following fusion of two nonresponsive human carcinoma cell lines. Implication of the type II receptor in growth inhibitory responses. J Biol Chem 1992; 267(4):2588–2593. 60. Houghton JA, Tillman DM, et al. Ratio of 20 -deoxyadenosine-50 -triphosphate/ thymidine-50 -triphosphate influences the commitment of human colon carcinoma cells to thymineless death. Clin Cancer Res 1995; 1(7):723–730. 61. Lehman NL. Future potential of thymidylate synthase inhibitors in cancer therapy. Expert Opin Invest Drugs 2002; 11(12):1775–1787. 62. Johnston PG, Drake JC, et al. Immunological quantitation of thymidylate synthase using the monoclonal antibody TS 106 in 5-fluorouracil-sensitive and -resistant human cancer cell lines. Cancer Res 1992; 52(16):4306–4312. 63. Swain SM, Lippman ME, et al. Fluorouracil and high-dose leucovorin in previously treated patients with metastatic breast cancer. J Clin Oncol 1989; 7(7):890–899. 64. Chu E, Koeller DM, et al. Regulation of thymidylate synthase in human colon cancer cells treated with 5-fluorouracil and interferon-gamma. Mol Pharmacol 1993; 43(4):527–533. 65. Longley DB, Ferguson PR, et al. Characterization of a thymidylate synthase (TS)-inducible cell line: a model system for studying sensitivity to TS- and non-TS-targeted chemotherapies. Clin Cancer Res 2001; 7(11):3533–3539. 66. Marsh S, McKay JA, et al. Polymorphism in the thymidylate synthase promoter enhancer region in colorectal cancer. Int J Oncol 2001; 19(2):383–386. 67. Marsh S, McLeod HL. Thymidylate synthase pharmacogenetics in colorectal cancer. Clin Colorectal Cancer 2001; 1(3):175–178; discussion 179–181. 68. Kawakami K, Omura K, et al. Polymorphic tandem repeats in the thymidylate synthase gene is associated with its protein expression in human gastrointestinal cancers. Anticancer Res 1999; 19(4B):3249–3252. 69. Marcuello E, Altes A, et al. Single nucleotide polymorphism in the 50 tandem repeat sequences of thymidylate synthase gene predicts for response to fluorouracil-based chemotherapy in advanced colorectal cancer patients. Int J Cancer 2004; 112(5):733–777. 70. Diasio RB, Harris BE. Clinical pharmacology of 5-fluorouracil. Clin Pharmacokinet 1989; 16(4):215–237. 71. Johnston SJ, Ridge SA, et al. Regulation of dihydropyrimidine dehydrogenase in colorectal cancer. Clin Cancer Res 1999; 5(9):2566–2570.

Pharmacogenomics of Colorectal Cancer

311

72. Wei X, McLeod HL, et al. Molecular basis of the human dihydropyrimidine dehydrogenase deficiency and 5-fluorouracil toxicity. J Clin Invest 1996; 98(3):610–615. 73. van Kuilenburg ABME, Haasjes J, Meinsma R, et al. Lethal outcome of a patient with a complete dihydropyrimidine dehydrogenase (DPD) deficiency after administration of 5-fluorouracil: frequency of the common IVS14þ1G > A mutation causing DPD deficiency. Clin Cancer Res 2001; 7(5):1149–1153. 74. Metzger R, Danenberg K, et al. High basal level gene expression of thymidine phosphorylase (platelet-derived endothelial cell growth factor) in colorectal tumors is associated with nonresponse to 5-fluorouracil. Clin Cancer Res 1998; 4(10):2371–2376. 75. Takebayashi Y, Miyadera K, et al. Expression of thymidine phosphorylase in human gastric carcinoma. Jpn J Cancer Res 1996; 87(3):288–295. 76. Liu ZG, Baskaran R, et al. Three distinct signalling responses by murine fibroblasts to genotoxic stress. Nature 1996; 384(6606):273–276. 77. Lane DP, Crawford LV. T antigen is bound to a host protein in SV40transformed cells. Nature 1979; 278(5701):261–263. 78. Oren M. The p53 cellular tumor antigen: gene structure, expression and protein properties. Biochim Biophys Acta 1985; 823(1):67–78. 79. Lowe SW, Schmitt EM, et al. p53 is required for radiation-induced apoptosis in mouse thymocytes. Nature 1993; 362(6423):847–849. 80. Lowe SW, Ruley HE, et al. p53-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell 1993; 74(6):957–967. 81. Longley DB, Boyer J, et al. The role of thymidylate synthase induction in modulating p53-regulated gene expression in response to 5-fluorouracil and antifolates. Cancer Res 2002; 62(9):2644–2649. 82. Bunz F, Hwang PM, et al. Disruption of p53 in human cancer cells alters the responses to therapeutic agents. J Clin Invest 1999; 104(3):263–269. 83. Ahnen DJ, Feigl P, et al. Ki-ras mutation and p53 overexpression predict the clinical behavior of colorectal cancer: a Southwest Oncology Group study. Cancer Res 1998; 58(6):1149–1158. 84. Elsaleh H, Powell B, et al. P53 alteration and microsatellite instability have predictive value for survival benefit from chemotherapy in stage III colorectal carcinoma. Clin Cancer Res 2001; 7(5):1343–1349. 85. Liang JT, Huang KC, et al. P53 overexpression predicts poor chemosensitivity to high-dose 5-fluorouracil plus leucovorin chemotherapy for stage IV colorectal cancers after palliative bowel resection. Int J Cancer 2002; 97(4):451–457. 86. Paradiso A, Simone G, et al. Thymidilate synthase and p53 primary tumour expression as predictive factors for advanced colorectal cancer patients. Br J Cancer 2000; 82(3):560–567. 87. Remvikos Y. Analysis of p53 protein in clinical research. Pathol Biol (Paris) 1993; 41(1):19–20. 88. Sjogren S, Inganas M, et al. The p53 gene in breast cancer: prognostic value of complementary DNA sequencing versus immunohistochemistry. J Natl Cancer Inst 1996; 88(3–4):173–182. 89. Etienne MC, Chazal M, et al. Prognostic value of tumoral thymidylate synthase and p53 in metastatic colorectal cancer patients receiving fluorouracil-based chemotherapy: phenotypic and genotypic analyses. J Clin Oncol 2002; 20(12):2832–2843.

312

Johnston et al.

90. Boudny V, Vrana O, et al. Biophysical analysis of DNA modified by 1,2-diaminocyclohexane platinum(II) complexes. Nucleic Acids Res 1992; 20(2):267–272. 91. Adjei AA. A review of the pharmacology and clinical activity of new chemotherapy agents for the treatment of colorectal cancer. Br J Clin Pharmacol 1999; 48(3):265–277. 92. Page JD, Husain I, et al. Effect of the diaminocyclohexane carrier ligand on platinum adduct formation, repair, and lethality. Biochemistry 1990; 29(4): 1016–1024. 93. Arnould S, Hennebelle I, et al. Cellular determinants of oxaliplatin sensitivity in colon cancer cell lines. Eur J Cancer 2003; 39(1):112–119. 94. Jennerwein MM, Eastman A, et al. Characterization of adducts produced in DNA by isomeric 1,2-diaminocyclohexaneplatinum(II) complexes. Chem Biol Interact 1989; 70(1–2):39–49. 95. Raymond E, Chaney SG, et al. Oxaliplatin: a review of preclinical and clinical studies. Ann Oncol 1998; 9(10):1053–1071. 96. Boyer JME, Aroori S, Wilson P, et al. Characterization of p53 wild-type and null isogenic colorectal cancer cell lines resistant to 5-fluorouracil, oxaliplatin, and irinotecan. Clin Cancer Res 2004; 10(6):2158–2167. 97. de Gramont AFA, Seymour M, Homerin M, et al. Leucovorin and fluorouracil with or without oxaliplatin as first-line treatment in advanced colorectal cancer. J Clin Oncol 2000; 18(16):2938–2947. 98. Machover D, Diaz-Rubio E, et al. Two consecutive phase II studies of oxaliplatin (L-OHP) for treatment of patients with advanced colorectal carcinoma who were resistant to previous treatment with fluoropyrimidines. Ann Oncol 1996; 7(1):95–98. 99. Reardon JT, Vaisman A, et al. Efficient nucleotide excision repair of cisplatin, oxaliplatin, and bis-aceto-ammine-dichloro-cyclohexylamine-platinum(IV) (JM216) platinum intrastrand DNA diadducts. Cancer Res 1999; 59(16): 3968–3971. 100. Wilson MD, Ruttan CC, et al. ERCC1: a comparative genomic perspective. Environ Mol Mutagen 2001; 38(2–3):209–215. 101. Shirota Y, Stoehlmacher J, et al. ERCC1 and thymidylate synthase mRNA levels predict survival for colorectal cancer patients receiving combination oxaliplatin and fluorouracil chemotherapy. J Clin Oncol 2001; 19(23): 4298–4304. 102. Park DJ, Stoehlmacher J, et al. Tailoring chemotherapy in advanced colorectal cancer. Curr Opin Pharmacol 2003; 3(4):378–385. 103. Sung P, Bailly V, et al. Human xeroderma pigmentosum group D gene encodes a DNA helicase. Nature 1993; 365(6449):852–855. 104. Lunn RM, Helzlsouer KJ, et al. XPD polymorphisms: effects on DNA repair proficiency. Carcinogenesis 2000; 21(4):551–555. 105. Park DJ, Stoehlmacher J, et al. A xeroderma pigmentosum group D gene polymorphism predicts clinical outcome to platinum-based chemotherapy in patients with advanced colorectal cancer. Cancer Res 2001; 61(24):8654–8658. 106. Terrier P, Townsend AJ, et al. An immunohistochemical study of pi class glutathione S-transferase expression in normal human tissue. Am J Pathol 1990; 137(4):845–853.

Pharmacogenomics of Colorectal Cancer

313

107. Zhang K, Mack P, et al. Glutathione-related mechanisms in cellular resistance to anticancer drugs. Int J Oncol 1998; 12(4):871–882. 108. Watson MA, Stewart RK, et al. Human glutathione S-transferase P1 polymorphisms: relationship to lung tissue enzyme activity and population frequency distribution. Carcinogenesis 1998; 19(2):275–280. 109. Stoehlmacher J, Park DJ, et al. Association between glutathione S-transferase P1, T1, and M1 genetic polymorphism and survival of patients with metastatic colorectal cancer. J Natl Cancer Inst 2002; 94(12):936–942. 110. Takimoto CH, Kieffer LV, et al. DNA topoisomerase I inhibitors. Cancer Chemother Biol Response Modif 1997; 17:80–113. 111. Pommier Y, Kohlhagen G, et al. DNA sequence- and structure-selective alkylation of guanine N2 in the DNA minor groove by ecteinascidin 743, a potent antitumor compound from the Caribbean tunicate Ecteinascidia turbinata. Biochemistry 1996; 35(41):13303–13309. 112. Misset JL, di Palma M, et al. Adjuvant treatment of node-positive breast cancer with cyclophosphamide, doxorubicin, fluorouracil, and vincristine versus cyclophosphamide, methotrexate, and fluorouracil: final report after a 16-year median follow-up duration. J Clin Oncol 1996; 14(4):1136–1145. 113. Rivory LP, Bowles MR, et al. Conversion of irinotecan (CPT-11) to its active metabolite, 7-ethyl-10-hydroxycamptothecin (SN-38), by human liver carboxylesterase. Biochem Pharmacol 1996; 52(7):1103–1111. 114. Hsiang YH, Liu LF. Identification of mammalian DNA topoisomerase I as an intracellular target of the anticancer drug camptothecin. Cancer Res 1988; 48(7):1722–1726. 115. Shimada Y, Yoshino M, et al. Phase II study of CPT-11, a new camptothecin derivative, in metastatic colorectal cancer. CPT-11 Gastrointestinal Cancer Study Group. J Clin Oncol 1993; 11(5):909–913. 116. Humerickhouse R, Lohrbach K, et al. Characterization of CPT-11 hydrolysis by human liver carboxylesterase isoforms hCE-1 and hCE-2. Cancer Res 2000; 60(5):1189–1192. 117. Beutler E, Gelbart T, et al. Racial variability in the UDP-glucuronosyltransferase 1 (UGT1A1) promoter: a balanced polymorphism for regulation of bilirubin metabolism? Proc Natl Acad Sci U S A 1998; 95(14):8170–8174. 118. Iyer L, King CD, et al. Genetic predisposition to the metabolism of irinotecan (CPT-11). Role of uridine diphosphate glucuronosyltransferase isoform 1A1 in the glucuronidation of its active metabolite (SN-38) in human liver microsomes. J Clin Invest 1998; 101(4):847–854. 119. Iyer L, Das S, et al. UGT1A128 polymorphism as a determinant of irinotecan disposition and toxicity. Pharmacogenomics J 2002; 2(1):43–47. 120. Ando Y, Saka H, et al. Polymorphisms of UDP-glucuronosyltransferase gene and irinotecan toxicity: a pharmacogenetic analysis. Cancer Res 2000; 60(24):6921–6926. 121. Rouits E, Boisdron-Celle M, et al. Relevance of different UGT1A1 polymorphisms in irinotecan-induced toxicity: a molecular and clinical study of 75 patients. Clin Cancer Res 2004; 10(15):5151–5159. 122. Innocenti F, Undevia SD, et al. Genetic variants in the UDP-glucuronosyltransferase 1A1 gene predict the risk of severe neutropenia of irinotecan. J Clin Oncol 2004; 22(8):1382–1388.

314

Johnston et al.

123. Carlini LE, Meropol NJ, et al. UGT1A7 and UGT1A9 polymorphisms predict response and toxicity in colorectal cancer patients treated with capecitabine/ irinotecan. Clin Cancer Res 2005; 11(3):1226–1236. 124. Zwick E, Hackel PO, et al. The EGF receptor as central transducer of heterologous signalling systems. Trends Pharmacol Sci 1999; 20(10):408–412. 125. Ciardiello F. Epidermal growth factor receptor tyrosine kinase inhibitors as anticancer agents. Drugs 2000; 60(suppl 1):25–32; discussion 41–42. 126. Mayer A, Takimoto M, et al. The prognostic significance of proliferating cell nuclear antigen, epidermal growth factor receptor, and mdr gene expression in colorectal cancer. Cancer 1993; 71(8):2454–2460. 127. Salomon DS, Normanno N, et al. The role of amphiregulin in breast cancer. Breast Cancer Res Treat 1995; 33(2):103–114. 128. Blume-Jensen P, Hunter T. Oncogenic kinase signalling. Nature 2001; 411(6835): 355–365. 129. Yarden Y. The EGFR family and its ligands in human cancer. signalling mechanisms and therapeutic opportunities. Eur J Cancer 2001; 37(suppl 4): S3–S8. 130. McKay JA, Murray LJ, et al. Evaluation of the epidermal growth factor receptor (EGFR) in colorectal tumours and lymph node metastases. Eur J Cancer 2002; 38(17):2258–2264. 131. Chi DD, Hing AV, et al. Two chromosome 7 dinucleotide repeat polymorphisms at gene loci epidermal growth factor receptor (EGFR) and pro alpha 2 (I) collagen (COL1A2). Hum Mol Genet 1992; 1(2):135. 132. Gebhardt F, Zanker KS, et al. Modulation of epidermal growth factor receptor gene transcription by a polymorphic dinucleotide repeat in intron 1. J Biol Chem 1999; 274(19):13176–13180. 133. Azzariti A, Xu JM, et al. The schedule-dependent enhanced cytotoxic activity of 7-ethyl-10-hydroxy-camptothecin (SN-38) in combination with Gefitinib (Iressa, ZD1839). Biochem Pharmacol 2004; 68(1):135–144. 134. Koizumi F, Kanzawa F, et al. Synergistic interaction between the EGFR tyrosine kinase inhibitor gefitinib (‘‘Iressa’’) and the DNA topoisomerase I inhibitor CPT-11 (irinotecan) in human colorectal cancer cells. Int J Cancer 2004; 108(3):464–472. 135. Williams KJ, Telfer BA, et al. ZD1839 (‘Iressa’), a specific oral epidermal growth factor receptor-tyrosine kinase inhibitor, potentiates radiotherapy in a human colorectal cancer xenograft model. Br J Cancer 2002; 86(7): 1157–1161. 136. Daneshmand M, Parolin DA, et al. A pharmacodynamic study of the epidermal growth factor receptor tyrosine kinase inhibitor ZD1839 in metastatic colorectal cancer patients. Clin Cancer Res 2003; 9(7):2457–2464. 137. Ciardiello F, Tortora G. A novel approach in the treatment of cancer: targeting the epidermal growth factor receptor. Clin Cancer Res 2001; 7(10):2958–2970. 138. Petit AM, Rak J, et al. Neutralizing antibodies against epidermal growth factor and ErbB-2/neu receptor tyrosine kinases down-regulate vascular endothelial growth factor production by tumor cells in vitro and in vivo: angiogenic implications for signal transduction therapy of solid tumors. Am J Pathol 1997; 151(6):1523–1530.

Pharmacogenomics of Colorectal Cancer

315

139. Ciardiello F, Bianco R, et al. Antiangiogenic and antitumor activity of anti-epidermal growth factor receptor C225 monoclonal antibody in combination with vascular endothelial growth factor antisense oligonucleotide in human GEO colon cancer cells. Clin Cancer Res 2000; 6(9):3739–3747. 140. Prewett MC, Hooper AT, et al. Enhanced antitumor activity of anti-epidermal growth factor receptor monoclonal antibody IMC-C225 in combination with irinotecan (CPT-11) against human colorectal tumor xenografts. Clin Cancer Res 2002; 8(5):994–1003. 141. Saltz LB, Meropol NJ, et al. Phase II trial of cetuximab in patients with refractory colorectal cancer that expresses the epidermal growth factor receptor. J Clin Oncol 2004; 22(7):1201–1208. 142. van Laethem R, Mitry. Cetuximab (C225) in combination with bi-weekly irinotecan, infusional 5-fluorouracil and folinic acid in patients with metastatic colorectal cancer expressing the epidermal growth factor receptor. Preliminary safety and efficacy results. Proc Am Soc Clin Oncol 2003; 22(A1058):264. 143. Cunningham D, Humblet Y, et al. Cetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. N Engl J Med 2004; 351(4):337–345. 144. Matar P, Rojo F, et al. Combined epidermal growth factor receptor targeting with the tyrosine kinase inhibitor gefitinib (ZD1839) and the monoclonal antibody cetuximab (IMC-C225): superiority over single-agent receptor targeting. Clin Cancer Res 2004; 10(19):6487–6501. 145. Kanno S, Oda N, et al. Roles of two VEGF receptors, Flt-1 and KDR, in the signal transduction of VEGF effects in human vascular endothelial cells. Oncogene 2000; 19(17):2138–2146. 146. Kim KJ, Li B, et al. Inhibition of vascular endothelial growth factor-induced angiogenesis suppresses tumour growth in vivo. Nature 1993; 362(6423): 841–844. 147. Cascinu S, Staccioli MP, et al. Expression of vascular endothelial growth factor can predict event-free survival in stage II colon cancer. Clin Cancer Res 2000; 6(7):2803–2807. 148. Ferrara N, Keyt B. Vascular endothelial growth factor: basic biology and clinical implications. Exs 1997; 79:209–232. 149. Neufeld G, Cohen T, et al. Vascular endothelial growth factor (VEGF) and its receptors. Faseb J 1999; 13(1):9–22. 150. Ishigami SI, Arii S, et al. Predictive value of vascular endothelial growth factor (VEGF) in metastasis and prognosis of human colorectal cancer. Br J Cancer 1998; 78(10):1379–1384. 151. Kabbinavar F, Hurwitz HI, et al. Phase II, randomized trial comparing bevacizumab plus fluorouracil (FU)/leucovorin (LV) with FU/LV alone in patients with metastatic colorectal cancer. J Clin Oncol 2003; 21(1):60–65. 152. Hurwitz H, Fehrenbacher L, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004; 350(23): 2335–2342. 153. Friedman DB, Hill S, et al. Proteome analysis of human colon cancer by twodimensional difference gel electrophoresis and mass spectrometry. Proteomics 2004; 4(3):793–811.

316

Johnston et al.

154. Anderson L, Seilhamer J. A comparison of selected mRNA and protein abundances in human liver. Electrophoresis 1997; 18(3–4):533–537. 155. Kallioniemi A, Kallioniemi OP, et al. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 1992; 258(5083):818–821. 156. Reid LH, Crider-Miller SJ, et al. Genomic organization of the human p57KIP2 gene and its analysis in the G401 Wilms’ tumor assay. Cancer Res 1996; 56(6):1214–1218. 157. Al-Mulla F, Keith WN, et al. Comparative genomic hybridization analysis of primary colorectal carcinomas and their synchronous metastases. Genes Chromosomes Cancer 1999; 24(4):306–314. 158. Rooney PH, Boonsong A, et al. Colorectal cancer genomics: evidence for multiple genotypes which influence survival. Br J Cancer 2001; 85(10):1492–1498. 159. Maxwell PJ, Longley DB, et al. Identification of 5-fluorouracil-inducible target genes using cDNA microarray profiling. Cancer Res 2003; 63(15):4602–4606. 160. Mariadason JM, Arango D, et al. Gene expression profiling-based prediction of response of colon carcinoma cells to 5-fluorouracil and camptothecin. Cancer Res 2003; 63(24):8791–8812. 161. Arango D, Wilson AJ, et al. Molecular mechanisms of action and prediction of response to oxaliplatin in colorectal cancer cells. Br J Cancer 2004; 91: 1931–1946. 162. Clarke PA, George ML, et al. Molecular pharmacology of cancer therapy in human colorectal cancer by gene expression profiling. Cancer Res 2003; 63(20): 6855–6863. 163. Wang Y, Jatkoe T, et al. Gene expression profiles and molecular markers to predict recurrence of Dukes’ B colon cancer. J Clin Oncol 2004; 22(9):1564–1571.

12 Drug Development for Advanced Colorectal Cancer in the United States Igor Puzanov and Mace L. Rothenberg Division of Hematology/Oncology, Vanderbilt University Medical Center, Nashville, Tennessee, U.S.A.

INTRODUCTION The beginning of the 21st century is an exciting time in the history of the treatment of colorectal cancer. After more than 30 years of relying on only one marginally effective drug—5-fluorouracil (5-FU)—whose activity was enhanced slightly by the addition of the biochemical modulator leucovorin (LV) (1)—the number of agents approved for use in the treatment of advanced colorectal cancer in the United States has increased to include irinotecan, capecitabine, oxaliplatin, bevacizumab, and cetuximab. The emergence of multiple new agents for the treatment of advanced colorectal cancer is associated with a substantial improvement in therapeutic outcome. Analyses of clinical trials that incorporate irinotecan, oxaliplatin, bevacizumab, and/or cetuximab in combination with 5-FU/LV show that patients diagnosed with advanced disease today can expect median survival times of 22 months or beyond (2–4) compared with only 11–13 months reported in the era when 5-FU was the only drug available to treat colorectal cancer (5). This improvement is not limited to patients treated in clinical trials. From 1974 to 1999, the five-year survival rates for Americans diagnosed with colorectal cancer rose from 50% to 63% (6), and it is reasonable to expect that data from subsequent years, when available, will show further improvements. Despite these additions to the treatment armamentarium, 317

318

Puzanov and Rothenberg

the statistics for colorectal cancer remain sobering. More than 146,000 new cases of colorectal cancer were diagnosed in the United States in 2004, and more than 56,000 men and women in the United States died from the disease, making it the third most common cause of cancer death (after lung cancer and prostate or breast cancer) (6). Importantly, slightly more than one-third of these people had their cancer diagnosed in an advanced stage and another 19% had distant metastases present at the time of diagnosis. DEVELOPMENT OF 5-FU AND 5-FU/LV AS FIRST-LINE CHEMOTHERAPY 5-FU (Adrucil1) was first described by Heidelberger in 1957 (7). The impetus for synthesis of fluorinated pyrimidines came from the observation that rat hepatomas use radiolabeled uracil more avidly than nonmalignant tissues (8). Clinical trials of single-agent 5-FU conducted prior to its approval in the United States in 1962 were limited by the tools available to assess therapeutic impact. Tumor assessment often relied on chest X ray, radionuclide scan, or physical examination. In many cases, responses did not have to be confirmed. In addition, limited insight into mechanisms of drug activation led to the clinical evaluation of a wide variety of drug dosages and schedules of administration. Therefore, it is not surprising that response rates reported for single-agent 5-FU in advanced colorectal cancer ranged from 3% to 45%. However, in phase III trials, response rates rarely exceeded 25%. Regardless of the exact response rate, it was clear that tumor shrinkage could be attained with single-agent 5-FU, and that these responding patients experienced a reduction in tumorrelated symptoms and disability. Given the criteria in effect at the U.S. Food and Drug Administration (FDA) at that time, this was sufficient evidence to garner new drug approval for 5-FU on April 25, 1962. With multiple new cytotoxic agents failing to demonstrate significant clinical activity in this disease, clinical trials performed over the ensuing two decades focused on optimization of 5-FU drug administration and biochemical modulation. The most extensively studied biochemical modulators were LV, methotrexate, and interferon. In 1992, the Advanced Colorectal Cancer Meta-Analysis Project (1) summarized the results of nine controlled clinical trials that compared single-agent 5-FU to LV-modulated 5-FU in patients with previously untreated advanced colorectal cancer. The 5-FU/LV combination was associated with a significantly higher response rate than single-agent 5-FU, 23% versus 11%, p < 10
7, but only a nonsignificant trend towards improved median survival: 11.5 versus 11.0 months, p ¼ 0.57. This analysis was updated in 2004 to include 19 trials and more than 3300 patients with a median follow-up of 45 months (9). In addition to a doubling of the objective response rate, this meta-analysis identified a significant improvement in median survival, 11.7 versus 10.5 months, p ¼ 0.004, for patients treated with 5-FU/LV.

Drug Development for Advanced Colorectal Cancer

319

Meanwhile, another question surrounding 5-FU was the optimal schedule of drug administration: bolus versus infusion. In the United States, the preferred methods for 5-FU drug administration were on a daily  5 basis every four to five weeks (the so-called Mayo Clinic regimen) or a weekly  6 basis every eight weeks (the so-called Roswell Park regimen). In Europe, 5-FU was typically administered on an infusional basis, ranging from a 24-hour infusion to infusions lasting several weeks or more. Although most phase III trials in which infusional 5-FU was compared to bolus 5-FU failed to identify a survival advantage for the infusional approach, a meta-analysis combining the results from seven phase III trials identified significantly higher response rates, 22% versus 14%, p ¼ 0.0002 and a longer median survival, 12.1 versus 11.3 months, p ¼ 0.04, for the infusional schedule (5). Despite these results, infusional 5-FU did not gain widespread popularity in the United States because of its increased expense—disposable pumps are not used— and perceived inconvenience and risk to the patient. In advanced colorectal cancer, bolus 5-FU regimens, such as the Mayo Clinic or Roswell Park regimens, were widely adopted in the United States, while infused schedules, such as the AIO or LV5FU2 regimens, were explored in Europe. A difference in toxicity profile also was confirmed; among patients treated with 5-FU bolus by the daily  5 schedule, 31% reported grade 3/4 toxicities compared with 4% of patients treated with 5-FU Cl ( p < 0.0001) (5). A greater percentage of patients receiving 5-FU bolus experienced hematologic toxicity and a greater percentage of those receiving infusional 5-FU experienced hand–foot syndrome, but there was no difference between groups in diarrhea, nausea/ vomiting, or mucositis (5). DEVELOPMENT OF CAPECITABINE AS AN ORAL ALTERNATIVE TO 5-FU 5-FU has unpredictable oral bioavailability and, therefore, must be administered intravenously. Capecitabine is an oral prodrug that is enzymatically cleaved to release 5-FU preferentially in cancer cells. The developmental strategy for capecitabine was to demonstrate therapeutic noninferiority compared to intravenous 5-FU and LV. The term ‘‘noninferiority’’ was used rather than ‘‘equivalence’’ due to the fact that it is virtually impossible to determine that two therapies are exactly equivalent. Two virtually identical phase III trials were performed that randomized patients to oral capecitabine 2500 mg/m2/day  2 weeks, every three weeks or to 5-FU/LV on the daily  5 (Mayo Clinic schedule), every four weeks (10,11). Each protocol was designed with 80% power to detect a difference of 10% in response rate with alpha ¼ 0.025. Because these trials were conducted simultaneously, using the same treatment regimes and the same eligibility criteria, it is useful to look at the results of the pooled data (12,13). Somewhat surprisingly, capecitabine was associated with a higher objective response rate than

320

Puzanov and Rothenberg

5-FU/LV: 22.4% versus 13.2% (by independent review criteria), ( p < 0.0001). Time to disease progression and overall survival were virtually identical between the groups. Time to tumor progression (TTP) was 4.6 months for capecitabine-treated patients and 4.7 months for 5-FU/LV-treated patients ( p ¼ 0.95, progression hazard ratio ¼ 1.0). Likewise, median survival was 12.9 months for both arms ( p ¼ 0.91, death hazard ration ¼ 1.0). Toxicity profiles differed between the two groups. Patients treated with daily bolus 5-FU/LV were more likely to experience diarrhea, neutropenia, stomatitis, nausea, and alopecia; patients treated with capecitabine were more likely to experience hand–foot syndrome and hyperbilirubinemia. Based on its similar therapeutic efficacy to 5-FU/LV, its convenience, and its acceptable tolerability profile, capecitabine was granted full approval by the FDA on September 7, 2001. DEVELOPMENT OF IRINOTECAN AS SECOND-LINE THERAPY Irinotecan (Campto1 and Camptosar1) binds the enzyme topoisomerase I, which prevents the uncoiling of DNA during DNA synthesis and repair, leading to apoptosis (14). Irinotecan first entered clinical trial in Japan in 1986. The results from phase I and early phase II studies began to appear in the literature in 1990. By the time the first clinical trials were initiated in France in 1990 and the United States in 1991, objective responses had been reported from Japan and there was emerging recognition of the clinical potential of irinotecan (15). There was clearly a need for another cytotoxic drug effective against advanced colorectal cancer. Patients with progressive CRC following 5-FU had limited therapeutic options beyond investigational agents. That is what made the objective responses observed in this group of patients in phase I such an important observation. The first phase II trial of irinotecan was performed in Japan by Shimada and colleagues who reported a 27% objective response rate in a group of previously treated and untreated patients. This prompted rapid movement of single-agent irinotecan into phase II trials in patients with advanced colorectal cancer that had progressed following 5-FU. Although objective response rates ranged from 11% to 27% in three phase II trials performed in the United States, the combined data demonstrated a 15% objective response rate, an additional 40% to 50% of patients with stable disease for at least four months, a median survival of nine months, and a reduction in tumor-related symptoms in a subset of patients with symptoms at baseline (14). Several potential designs for registration-directed phase III trials were considered. These included comparison of irinotecan to single-agent mitomycin C or carmustine (BCNU) or to best supportive care. No consensus could be achieved. Fortunately, another alternative had recently become available: accelerated approval.

Drug Development for Advanced Colorectal Cancer

321

Accelerated approval, or 21 CFR 314 Subpart H (Subpart H, for short) in the U.S. Code of Federal Regulations, was established in December 1992, as a response to the urgent need for effective drugs to fight lifethreatening diseases such as cancer and AIDS. This provision reflected acceptance of the fact that individuals with life-threatening illnesses such as cancer would be more willing to accept a greater degree of uncertainty, risk, or side effects from a new therapy than individuals with less severe illness or for whom effective therapies existed. This regulation allowed drugs that demonstrated beneficial effects on surrogate endpoints associated with clinical benefit to gain provisional approval based on the results of phase II trials. If granted, the approval was conditional on subsequent demonstration of benefit using more standard and stringent criteria, such as survival, in more definitive trials. Approval could be withdrawn if the confirmatory studies failed to demonstrate clear benefit. Using this mechanism in early 1996, the Oncologic Drugs Advisory Committee (ODAC) recommended unanimously to grant accelerated approval to irinotecan for the treatment of patients with progressive colorectal cancer following 5-FU. The FDA concurred with this recommendation, and in mid1996 irinotecan became the first new drug to be approved for colorectal cancer in the United States in more than 30 years. Two years later, the data from two phase III trials conducted in Europe demonstrated that irinotecan improved survival in this setting when compared to either best supportive care or infusional 5-FU. This enabled the FDA to convert irinotecan to full-approval status in 1999 (16,17). INTEGRATION OF IRINOTECAN INTO FRONT-LINE CHEMOTHERAPY Once a drug has been approved as a single agent in the refractory disease setting, the next step in its development is to combine the drug with existing front-line therapy to determine whether it improves therapeutic outcomes in that setting. Two phase III trials evaluated the combination of irinotecan and 5-FU/LV in comparison with 5-FU/LV alone as first-line chemotherapy for patients with advanced colorectal cancer. One, conducted primarily in the United States, Israel, and Australia, used irinotecan (125 mg/m2 infused over 90 minutes) added to bolus 5-FU (500 mg/m2) plus LV (20 mg/m2) given weekly for four weeks every six weeks. This became known as the IFL regimen (18). The other study, conducted primarily in Europe, compared irinotecan added to infusional 5-FU administered once every two weeks (LV5FU2 or de Gramont schedule) with the infusional 5-FU administered weekly (AIO schedule) (19). In both studies, overall survival was prolonged when irinotecan was added to 5-FU/LV. Both trials demonstrated significant improvements in response rate, progression-free survival (PFS), a two- to three-month improvement in median survival, and a 25%

322

Puzanov and Rothenberg

reduction in the death hazard ratio. Although these improvements were associated with an increase in the frequency and severity of adverse effects, including diarrhea, vomiting, and neutropenia, overall quality of life was not impaired. As a result, in March of 2000, ODAC recommended and the FDA granted expanded indication to irinotecan to include its use in the first-line treatment of advanced colorectal cancer. OXALIPLATIN: INITIAL DATA ON FRONT-LINE THERAPY Oxaliplatin (Eloxatin1) is a diaminocyclohexane platinum that is structurally and functionally distinct from cisplatin and carboplatin. The platinum-DNA adducts formed by oxaliplatin are bulkier and more hydrophobic—and, therefore, more difficult to repair—than those formed by other platinums. Oxaliplatin entered clinical testing in 1986, but it was not until 1990 that the single-agent maximum-tolerated dose was identified (20). The activity of oxaliplatin in advanced colorectal cancer was first reported by Le´vi and colleagues (21) who combined oxaliplatin with chronomodulated 5-FU and LV and observed a 58% response rate, a median PFS of 10 months, and a median overall survival of 15 months in a group of 93 patients, 49% of whom had received prior chemotherapy and/or radiotherapy. While these impressive results stimulated further evaluation of this combination in patients with advanced colorectal cancer, it also provoked numerous questions: How much of the antitumor effect was due to the three-drug combination and how much was due to oxaliplatin alone? How important was chronomodulation to the activity of these drugs? How active would this regimen have been in a more homogeneous group of patients with chemotherapy-na€
ve colorectal cancer? Over the ensuing years, other French investigators reported similar phase II efficacy data using nonchronomodulated drug administration schedules for this three-drug regimen. In order to better delineate the contribution of oxaliplatin to front-line chemotherapy, two phase III trials were undertaken in Europe in the mid-1990s using the oxaliplatin, 5-FU, and LV combination. One compared chronomodulated 5-FU and LV to the same chronomodulated regimen with oxaliplatin added (22). The other evaluated 5-FU given as a bolus and infusion combined with LV (the so-called LV5FU2 regimen) versus that same regimen with oxaliplatin added (23). While both phase III trials demonstrated that the addition of oxaliplatin to 5-FU and LV resulted in more than a doubling of objective response rates and a 40% improvement in PFS, overall survival was not significantly improved. Possible reasons offered for this therapeutic inconsistency included the small sample size of each study and the high proportion of poststudy cross-over of control patients to an oxaliplatin-containing second-line regimen. These questions were actively debated but never resolved. The improvement in PFS, the primary endpoint for both studies, was sufficient to garner drug approval

Drug Development for Advanced Colorectal Cancer

323

in France in 1996 and in the European Regulatory Agency in 1999. These data were brought before the U.S. Oncologic Drugs Advisory Committee in March of 2000. Ironically, this was the day on which ODAC reviewed the supplemental new drug application for irinotecan for the same indication (see previous section). While a survival advantage was achieved in both of the phase III trials of front-line irinotecan, 5-FU and LV over 5-FU and LV alone, no such survival advantage was seen for the trials involving oxaliplatin. Not surprisingly, ODAC recommended that the new drug application for oxaliplatin not be recommended for approval. At that time in the United States, overall survival was considered to be the most important measure of efficacy, especially when the new agent was a cytotoxic compound associated with clinically significant toxicity. Some tangible therapeutic advantages beyond increased response rate or prolonged PFS were needed in order for a drug to win regulatory approval in the United States. Lacking that, the data from these studies was considered insufficient to warrant new drug approval. OXALIPLATIN: DEMONSTRATION OF SECOND-LINE EFFICACY With data demonstrating a survival advantage when irinotecan was incorporated into front-line chemotherapy for metastatic disease, the combination of irinotecan, 5-FU, and LV became the new front-line treatment of choice for patients with advanced colorectal cancer. What this meant was that patients who developed progressive disease following front-line treatment with this combination regimen had no effective second-line therapeutic options. This opened the door for the evaluation of oxaliplatin as secondline chemotherapy and, in fact, created several important opportunities. The first was that the development strategy for oxaliplatin could shift to an accelerated approval focus, since the movement of irinotecan to the front-line setting created an unmet medical need for patients who progressed on combination front-line therapy. The second was that, unlike irinotecan in the mid-1990s, a phase III trial could be performed in the second-line setting. Several small phase II trials in the 1990s had reported clinical activity for infusional 5-FU in patients with disease progression on bolus 5-FU. Since 5-FU was given as a weekly bolus to patients on the IFL regimen, the most commonly used regimen in the United States at that time, the question of the impact of infusional 5-FU in this group of patients was a valid one. And lastly, data could be obtained regarding the activity of oxaliplatin as a single agent to help shed light on whether 5-FU was necessary at all for the efficacy of oxaliplatin in this setting. Four hundred and sixty-three patients were randomized to 5-FU and LV administered as the LV5FU2 regimen, single-agent oxaliplatin (85 mg/m2 every two weeks), or the FOLFOX4 regimen (LV5FU2 plus oxaliplatin) (24). The primary efficacy endpoint for a planned interim analysis was objective response rate (RR),

324

Puzanov and Rothenberg

with secondary endpoints of PFS and safety. The objective RR was 9.9% for the FOLFOX4 group, 0% for the LV5FU2 group, and 1.3% for the oxaliplatin monotherapy group. The difference between the FOLFOX4 and LV5FU2 groups was highly significant ( p < 0.0001). Patients treated with FOLFOX4 also experienced a longer TTP, 4.6 months, than those treated with LV5FU2, 2.7 months ( p < 0.0001). Although only an interim analysis, the FDA had agreed in advance that a significant improvement in agreedupon surrogate measures of efficacy—response rate and time to tumor progression—could provide the basis for accelerated approval of oxaliplatin in this recurrent disease setting. Therefore, on August 9, 2002, the FDA granted accelerated approval to oxaliplatin for use in combination with 5FU and LV for the treatment of patients with advanced colorectal cancer that had progressed despite prior treatment with irinotecan, 5-FU, and LV. Interestingly, although the superiority for FOLFOX4 over LV2FU2 in terms of response rate and TTP was maintained in the final analysis, this did not translate into an improvement in overall survival. Patients treated with FOLFOX4 had a median survival of 9.8 months compared to a median survival of 8.7 months for those treated with LV5FU2 ( p ¼ 0.07) [hazard ratio (HR ¼ 0.84)] (25). However, the impact of this observation was offset by surprising data regarding the efficacy of FOLFOX4 in the first-line metastatic setting that emerged around the same time. N9741: COMPARISON OF FIRST-LINE COMBINATION REGIMENS By the late 1990s, the challenges associated with multiple treatment options began to emerge in the field of colorectal cancer. For the first time, physicians had more than a single drug or treatment regimen to choose from. Various combinations involving irinotecan, oxaliplatin, and 5-FU þ LV had emerged, and there were little comparative data available for any of them. N9741 was originally designed to explore six combination chemotherapy regimens and compare them to a reference standard of 5-FU and LV on a Mayo Clinic daily  5 schedule (26). The experimental arms included irinotecan or oxaliplatin combined with bolus daily 5-FU and LV, IFL, FOLFOX4, and a combination of irinotecan and oxaliplatin, a regimen termed IROX. Based on emerging data on efficacy and safety, the trial was eventually truncated to just three arms: IFL, FOLFOX4, and IROX. This trial was particularly important because, at that time, IFL was the most popular first-line chemotherapy regimen for advanced colorectal cancer in the United States while FOLFOX4 was the most popular regimen in Europe. It was acknowledged that this trial was not designed to tease out the contributions of single variables but rather to compare popular treatment regimens in terms of safety and efficacy. For instance, IFL and FOLFOX are not only different in terms of the drugs contained in the regimen but also in terms of the way

Drug Development for Advanced Colorectal Cancer

325

in which 5-FU is administered: as a weekly bolus in IFL and as a bolus followed by infusion in FOLFOX. In addition, irinotecan was commercially available as second-line therapy to patients who received first-line FOLFOX, but oxaliplatin was not commercially available in the United States for patients who received first-line IFL. Despite these issues, the results of this trial proved quite impressive and quite surprising. Rather than demonstrating that FOLFOX and IFL were equivalent, the trial demonstrated significant advantages for FOLFOX over IFL in terms of response rate (45% vs. 31%; p ¼ 0.002), TTP (8.7 vs. 6.9 months; p ¼ 0.0014), and median survival (19.5 vs. 15.0 months; p ¼ 0.0001) (26). There were no significant differences among the groups in time to treatment discontinuation, although more patients treated with FOLFOX discontinued therapy due to reasons other than disease progression compared to patients treated with IFL. Based on the results of this trial, the approval of oxaliplatin was expanded to include first-line treatment of advanced colorectal cancer on January 12, 2004. BEVACIZUMAB For tumors to grow much beyond 1 mm3 in volume, a network of new blood vessels must be created to provide the growing tumor with an increasing supply of oxygen and nutrients. This process is called tumor angiogenesis and relies upon the stimulation and release of angiogenic growth factors by both the tumor and its surrounding stroma. The concept of controlling tumor growth by inhibiting tumor angiogenesis is an attractive one because these blood vessels are relatively small, leaky, and more reliant upon angiogenic growth factors than larger, normal blood vessels. In addition, endothelial cells were believed to lack the genetic instability of cancer cells and, as a result, considered less likely to mutate into a drug-resistant phenotype. Bevacizumab (Avastin1) is a recombinant humanized monoclonal antibody directed against the vascular endothelial growth factor, one of the most potent endothelial growth factors involved in tumor-associated angiogenesis. Bevacizumab binds to vascular endothelial growth factor before it can engage its receptor, acting like a molecular ‘‘sponge’’ and thereby depriving the tumor of new blood vessel formation required for further growth. The development of bevacizumab began with a small, firstline, randomized phase II trial in patients with metastatic colorectal cancer, comparing 5-FU and LV administered on a weekly basis for six out of eight weeks (Roswell Park schedule) with either bevacizumab (5 or 10 mg/kg every two weeks) or placebo (27). The results of this trial suggested that the 5 mg/kg dose of bevacizumab could be safely combined with bolus weekly 5-FU/LV and was associated with a very encouraging objective response rate of 40% [95% confidence interval (CI) ¼ 24–58%], median TTP of nine months (95% CI ¼ 5.8–10.9 months), and median survival of 21.5 months (range: 1.2–28.2þ months). More patients in the bevacizumab arms

326

Puzanov and Rothenberg

experienced a grade 3–4 adverse event including thrombotic events (5%), abdominal pain (3%), and hypertension (3%). Nevertheless, these results were considered sufficiently encouraging to prompt a large, registrationdirected phase III study. Because the standard of care for first-line treatment of metastatic colorectal cancer in the United States at the time was IFL, the trial was designed with three arms: IFL þ bevacizumab, IFL þ placebo, and 5-FU/LV (Roswell Park schedule) þ bevacizumab (4). This three-arm design was chosen because little clinical experience existed at that time regarding the safety of the IFL þ bevacizumab combination. The intention was that an independent data and safety monitoring committee would review the study after approximately 100 patients had been accrued to each arm and take appropriate action. If the IFL þ bevacizumab regimen was deemed too toxic, it would be closed and the study would proceed as a two-arm comparison of IFL þ placebo versus 5-FU/LV þ bevacizumab. As it turned out, the IFL þ bevacizumab combination was not found to be associated with an unacceptably high rate of excess toxicities, so the 5-FU/LV þ bevacizumab arm was closed and the trial proceeded to full accrual as a two-arm study comparing IFL þ bevacizumab to IFL þ placebo. The addition of bevacizumab to IFL significantly improved all measures of therapeutic efficacy, including objective response rate, 44.8% versus 34.8% (p ¼ 0.004); PFS, 10.6 versus 6.2 months median ( p < 0.001, HR for progression: 0.54); and overall survival, 20.3 versus 15.6 months ( p < 0.001, HR for death: 0.66). As observed in earlier trials, the addition of bevacizumab to chemotherapy increased the rates of grade 3 hypertension, proteinuria, arterial thrombotic events, and gastrointestinal perforations. For patients who were considered to be medically unfit for treatment with first-line IFL, a parallel phase III trial was performed in which patients 65 or older, Eastern Cooperative Oncology Group Performance Status (ECOG PS) 1 or 2, with a serum albumin of 3.5 g/dL, or who had received prior abdominopelvic radiation were randomized to receive 5-FU/LV þ bevacizumab or 5-FU/LV þ placebo (28). This was smaller than the IFL  bevacizumab trial, so some of the differences did not achieve statistical significance. The addition of bevacizumab to 5-FU/LV improved objective response rates from 15% to 26% ( p ¼ 0.055), PFS from 5.5 to 9.2 months ( p ¼ 0.0002, progression HR ¼ 0.50), and median survival from 12.9 to 16.6 ( p ¼ 0.16, death HR ¼ 0.79). Similar trends of increased toxicity, including hypertension, arterial thrombotic events, proteinuria, and GI perforation were observed in this study. A pooled analysis of three studies comparing 5-FU/LV to 5-FU/LV þ bevacizumab demonstrated significant improvements in all measures of efficacy (29). Based on the results of these studies, the FDA granted full approval to bevacizumab on February 26, 2004. The approval was for use with a 5-FUbased chemotherapy regimen in the first-line treatment of patients with metastatic colorectal cancer.

Drug Development for Advanced Colorectal Cancer

327

CETUXIMAB Cetuximab (Erbitux1) is also a mouse–human chimeric IgG1 monoclonal antibody with high affinity for the epidermal growth factor receptor (EGFR). Unlike bevacizumab, cetuximab binds directly to the EGFR, thereby blocking the growth signaling that occurs when one of the natural ligands bind to this receptor. The clinical development of cetuximab in colorectal cancer followed somewhat of a roundabout course. Based on preclinical data suggesting enhanced effect when given as part of a combination, the initial singlearm, phase II clinical trial evaluated cetuximab added to irinotecan in patients with EGFRþ tumors whose disease had failed to respond to irinotecan alone (30). In that trial, 121 patients were treated with cetuximab plus the same dose and schedule of irinotecan to which they had failed to respond or on which they had progressed. Twenty-seven patients (22.5%) achieved an objective response and an additional 37 (31%) achieved a best response of stable disease. Median PFS was approximately three months. When this trial was submitted to the FDA in late 2001 for accelerated approval, the agency issued a ‘‘refusal to file’’ letter, indicating that the application was so severely flawed that it did not merit formal review by the FDA. This decision was primarily based on the fact that this single trial was not designed with the rigor necessary for it to serve as a registration trial nor was the effect of single-agent cetuximab in this setting clearly established. A subsequent phase II trial of single-agent cetuximab performed in a similar group of patients yielded a 9% objective response rate and a 1.4month median PFS (31). Nevertheless, it took a larger, randomized Phase II trial to generate definitive data that would be required for drug registration. A subsequent, randomized Phase II trial was designed to test these therapies head to head in more rigorously defined group of patients with irinotecan-refractory disease (32). The trial was designed with a 2:1 randomization to the irinotecan þ cetuximab arm and required all patients to have EGFR-expressing tumors. The primary endpoint of the trial was objective response rate as assessed by an independent review committee. A 23% objective response rate was observed in the group of patients treated with the irinotecan þ cetuximab combination, whereas an 11% response rate was reported for those treated with cetuximab alone (p ¼ 0.007). A similar trend was observed for PFS, 4.1 months in the irinotecan þ cetuximab group and 1.5 months in the cetuximab-alone group, and for median survival, 8.6 months for the combination treatment group and 6.9 months for the single-agent treatment group. Although no correlation was observed between intensity of EGFR staining in tumor tissue and therapeutic outcome, a consistent correlation has been observed between intensity of skin rash and increased likelihood of response and prolonged survival. Toxicities of this therapy that appear to be related to cetuximab include acne-like skin

328

Puzanov and Rothenberg

rash and hypersensitivity reactions. The rash is generally of cosmetic concern, but for some patients can be clinically significant. When used in combination, cetuximab did not appear to lessen or exacerbate toxicities associated with irinotecan. The study performed by Cunningham and colleagues was a randomized phase II trial, not a phase III trial. There was no ‘‘control’’ arm since both arms of the study contained cetuximab and, thus, could be considered ‘‘experimental.’’ It is therefore justifiable to ask how a registration strategy could be built on this trial. The reason is that more than 60% of patients enrolled in this study had already received 5-FU, irinotecan, and oxaliplatin, all known effective therapies for advanced colorectal cancer at that time. In situations for which no standard treatment options exist, accelerated approval can be sought from the FDA on the basis of phase II data in which the therapy is associated with a beneficial outcome on a surrogate marker for survival, such as PFS or response rate in conjunction with tumor-related symptoms. It was on this basis that the FDA granted accelerated approval on February 12, 2004, to cetuximab—used in combination with irinotecan or used alone in irinotecanintolerant patients—for the treatment of refractory colorectal cancer.

REGULATORY CONSIDERATIONS OF THE U.S. FDA IN THE APPROVAL OF NEW DRUGS FOR TREATMENT OF ADVANCED COLORECTAL CANCER Full approval for an oncology drug in the United States requires substantial evidence of efficacy for the indication requested from adequate and wellcontrolled clinical trials (33). Recent interpretation by the U.S. FDA of the federal regulation that governs this area, known as 21 CFR 314, has concluded that the use of the plural ‘‘studies’’ requires that data from at least two studies be available to support the new drug application. The required characteristics for these studies are in listed in Table 1. Overall, the studies must be designed to allow a quantitative assessment of the drug’s beneficial as well as detrimental effects. In 1992, Subpart H was added to the regulations for new drug approval to allow for accelerated approval of drugs intended to treat serious or life-threatening diseases for which no effective therapy existed. The requirements for accelerated approval are not as stringent as those for full approval. While accelerated approval still requires that data be compiled from adequate, well-controlled clinical trials, it allows for the use of much broader endpoints than those usually required for full approval. To be considered useful, a surrogate endpoint must be clearly defined, measurable, and should have a reasonable probability of predicting clinical benefit to the patient. Cetuximab received accelerated approval from the FDA in 2004, based on objective response rate and PFS. Accelerated approval is

Drug Development for Advanced Colorectal Cancer

329

Table 1 21 CFR x 314.126: Qualities of Adequate and Well-Controlled Studies Clear statement of objectives and methods of data analysis planned and ultimately used Study design permits a valid comparison with a control to provide a quantitative assessment of the drug effect ‘‘Control’’ may consist of: Placebo concurrent control Dose-comparison concurrent control No treatment concurrent control Active treatment concurrent control Historical control, but only under exceptional circumstances Selection method must ensure that patients have the type and stage of cancer being studied Method of treatment assignment must minimize bias and ensure comparibility of groups with respect to pertinent variables such as: Age Sex Severity of disease Duration of disease Use of drugs or therapy other than the test drug (this is usually accomplished through the use of randomization, with or without stratification) Adequate measures must be taken to minimize the bias on the part of subjects, observers, and individuals involved in data analysis The methods to assess subjects’ response are well defined and reliable There is an analysis of the results adequate to assess the effects of the drug The test drug must be standardized as to its identity, strength, quality, purity, and dosage form

granted by the FDA with the understanding that larger, more definitive trials will be conducted to clarify the drug’s impact on the disease. In the case of cetuximab, a phase III trial has been completed in which cetuximab plus best supportive care was compared to best supportive care alone in patients with refractory colorectal cancer. Should this trial demonstrate an improvement in overall survival, with an acceptable toxicity profile, the accelerated approval for cetuximab will be converted to full approval. If, on the other hand, more definitive testing fails to show a survival improvement or should reveal unacceptable rates or severity of toxicity for a drug, the FDA has the ability to rescind accelerated approval and force withdrawal of a drug from the market in the United States. ENDPOINTS FOR NEW DRUG APPROVAL FOR COLORECTAL CANCER IN THE UNITED STATES In past years, overall survival was the most reliable indicator of patient benefit and the most widely used endpoint in large-scale phase III clinical trials

330

Puzanov and Rothenberg

in colorectal cancer. Clearly, a statistically significant improvement in survival provides the strongest evidence of benefit to the patient. Unfortunately, it is also the most remote endpoint of all. In past years, when there were few effective therapies available for advanced colorectal cancer, the effect of the new drug on survival could be isolated through a well-designed, phase III trial of first-line therapy. This is no longer the case. With five drugs now available for use in this group of patients, the impact of a new drug may be drowned out by the impact of second-, third-, and even fourth-line therapies. Therefore, while improvement in overall survival remains an important therapeutic goal, it is becoming less useful as a regulatory goal in colorectal cancer. Although often used interchangeably, TTP and PFS represent slightly different endpoints. TTP is the time from randomization until objective tumor progression, whereas PFS is the time from randomization until objective tumor progression or death. The advantages of using TTP or PFS as endpoints for drug development include the facts that they often correlate closely with survival, the endpoint is reached much earlier than overall survival, studies require fewer patients than those designed around overall survival, and, most importantly, that TTP or PFS reflect the effect of the investigational treatment on the disease and are not influenced by subsequent interventions. Disadvantages of using PFS or TTP as a regulatory endpoint include the fact that they may not always predict improvement in survival, as was the case for oxaliplatin when used as second-line therapy in advanced colorectal cancer (25). In addition, criteria for defining ‘‘progression’’ may not be as clear or straightforward as one might think. In addition to standard response evaluation criteria in solid tumors (RECIST) radiographic criteria, one also has to consider how one would interpret a patient with clear evidence of serologic progression of colorectal cancer, as reflected by a rising carcinoembryonic antigen (CEA) or recurrence of tumorrelated symptoms but without a significant increase in tumor dimensions on imaging studies. Would the TTP be the time at which tumor-associated serum or symptomatic changes occurred or the time of radiologic progression? Another problem with using PFS or TTP as a regulatory endpoint comes with the potential introduction of lead-time bias. Lead-time bias refers to an apparent difference between treatment that is due to the timing of reevaluation rather than due to a true difference in tumor progression. Registrationdirected clinical trials that use PFS or TTP as their primary endpoint must be designed in such a way as to minimize this variable and ensure that any difference observed is truly due to a difference in TTP. Radiologic objective response rate is one of the most commonly used measures of drug effect, especially in phase II clinical trials. It has the advantage of having objective criteria that can be used to independently confirm the observations made by the treating physician. Tumor shrinkage is a direct measure of antitumor activity of the new agent. It is also one of the earliest endpoints to emerge from a clinical trial, allowing decisions to be made long

Drug Development for Advanced Colorectal Cancer

331

before survival data become available. The disadvantage is that there may not be a tight correlation between tumor response rate and other clinically meaningful measures of patient benefit, such as prolongation of survival, delay of tumor progression, or even amelioration of tumor-related symptoms. For those reasons, response rates are never considered as the primary endpoint for full approval of a new oncology drug. However, response rate has been used, in conjunction with other patient-oriented endpoints, as the basis for accelerated approval for two drugs commonly used in the treatment of advanced colorectal cancer: irinotecan and cetuximab. Time to treatment failure is a composite endpoint that covers the time from randomization to the time of treatment discontinuation for any reason including tumor progression, death, or unacceptable toxicity. Although time to treatment failure might encompass clinically useful information, it is not an acceptable endpoint for regulatory purposes. The primary reason for this is its composite nature. Patients who discontinue treatment for any reason are all lumped into a single category, regardless of whether they discontinued treatment due to a failure to recover blood counts within a prespecified period of time, experienced tumor progression, or died from an unrelated event. Alleviation of tumor-related symptoms is a more recently identified endpoint for oncology drugs. Although improvement in this endpoint alone is unlikely to result in drug approval, this can be an important supporting endpoint in a registration strategy. Improvement in tumor-related symptoms is useful because it is a patient-centered endpoint that can be tailored to include the most common and troublesome symptoms associated with a particular tumor. In patients with recurrent colorectal cancer, for instance, approximately two-thirds have tumor-related symptoms, most commonly fatigue, pain, and weight loss (25). There are systematic ways in which these data can be collected and changes across treatment groups measured. The main limitations to this approach are that no one standard measure has emerged for measuring in tumor-related symptoms, it is difficult to determine the extent to which treatment-induced toxicities may offset relief of tumor-related symptoms, and there is no consensus regarding how to deal with missing data from patients who have gone off treatment due to progressive disease. Nevertheless, relief of tumor-related symptoms has contributed to the approval of several oncologic drugs, including oxaliplatin for the treatment of refractory colorectal cancer. HOW WILL NEW DRUGS BE DEVELOPED FOR COLORECTAL CANCER IN THE FUTURE? The emergence of five new drugs for the treatment of advanced colorectal cancer over the past 10 years has doubled the median survival in this disease from 11 to 22 months. Effective therapies exist not only for first-line

332

Puzanov and Rothenberg

treatment, but for second- and third-line treatment, as well. But our perspective must be tempered by the fact that very few patients with metastatic colorectal cancer are cured, and five-year survival rates are still <10%. While it is clear that further improvements are needed, it is not clear how these new drugs for advanced colorectal cancer will be developed. The most commonly used strategy in the past has been to establish activity of the new agent or regimen in a refractory disease setting, where no effective therapy exists. This approach was employed successfully in the initial development and approval of irinotecan, oxaliplatin, and cetuximab as second- or third-line treatment regimens. It will become more difficult to employ this strategy in the future, since patients who have exhausted all known effective drugs for advanced colorectal cancer will have already received three, four, or even five lines of therapy. The first problem is that this amount of prior treatment reduces the likelihood for response to a new agent. The second problem is that it also increases the likelihood of toxicity, especially in patients who might have residual toxicity from prior therapies. But the third and final problem may be the biggest of all: patients who still meet clinical trial eligibility criteria after all this therapy may not be representative of the entire population of patients with refractory colorectal cancer. Therefore, it becomes difficult, if not impossible, to extrapolate the results from fourth- and fifth-line clinical trials to the population at large and to make any claims as to the potential safety and efficacy in a broader, more heterogeneous, and potentially less fit group of patients. What other options exist? One is to demonstrate single-agent activity of the investigational drug by performing a ‘‘window of opportunity’’ study in which patients with advanced cancers are treated with the investigational drug as first-line therapy and monitored very closely. At the first sign of disease progression, the patient is switched to ‘‘standard’’ therapy. Should the drug demonstrate sufficient activity in this setting, it may then be easier to justify testing of the agent in an earlier line of therapy, either as a single agent or incorporated into a regimen with established activity in that particular setting. Contrary to the concerns of some about the potential adverse effect of delaying the initiation of known effective therapy, clinical trials have demonstrated that patients treated with an experimental drug in a window of opportunity setting enjoy the same overall survival as patients treated with standard first-line therapy. Another option is to develop the drug as part of a combination rather than as a single agent. In the past, combination phase I studies were initiated only after the drug demonstrated activity in single-agent phase II studies. Combination phase I studies were performed to determine whether the new agent could be safely combined with an established regimen, and these were followed by combination phase II studies to determine if the regimen had sufficient activity to warrant phase III, registration-directed studies. Although

Drug Development for Advanced Colorectal Cancer

333

methodical, this sequential approach may be unnecessarily slow. Combination phase I studies can be initiated even before the single-agent phase I trial has been completed. In this way, the maximum tolerated dose (MTD) for the new drug both as a single agent and as part of an established regimen for colorectal cancer should emerge within a relatively short period of time. This would allow initial Phase II trials to determine the activity of the combination regimen and determine whether the new regimen should be compared to the established regimen in a phase III, registration-directed trial. A final consideration in the design of future drug development efforts in advanced colorectal cancer is the incorporation of molecular or imaging techniques in the selection of patients who are most likely to respond to and/or least likely to suffer severe toxicities from the new agent. The identification of the increased risk of severe neutropenia in patients with UGT1A128 genotype who are treated with irinotecan is just the latest example of how pharmacogenomics—the association between genetic variations in the host and drug toxicity or tolerability—is likely to play an increasing role in drug development in the future. The selection of patients with favorable host or tumor genetic factors could potentially result in higher levels of clinical activity as well as reduced levels of clinical toxicity even at the earliest stages of clinical testing. This could result in a very efficient drug development strategy in which the outcome of patients who are selected based on favorable molecular profiles can be compared head to head against unselected patients treated either with that same agent or with standard therapy. Demonstration of superior activity and/or reduced toxicity could provide the basis not only for new drug registration but also for registration of the predictive assay. REFERENCES 1. Advanced Colorectal Cancer Meta-Analysis Project. Modulation of fluorouracil by leucovorin in patients with advanced colorectal cancer: evidence in terms of response rates. J Clin Oncol 1992; 10:896–903. 2. Tournigand C, Andre T, Achille E, et al. FOLFIRI followed by FOLFOX6 or the reverse sequence in advanced colorectal cancer: a randomized GERCOR study. J Clin Oncol 2004; 22:229–237. 3. Grothey A, Sargent D, Goldberg RM, Schmoll H-J. Survival of patients with advanced colorectal cancer improves with the availability of fluorouracil– leucovorin, irinotecan, and oxaliplatin in the course of treatment. J Clin Oncol 2004; 22:1209–1214. 4. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004; 350:2335–2342. 5. Meta-Analysis Group in Cancer. Efficacy of intravenous continuous infusion of fluorouracil compared with bolus administration in advanced colorectal cancer. J Clin Oncol 1998; 16:301–308.

334

Puzanov and Rothenberg

6. Jemal A, Tiwari RC, Murray T, et al. Cancer statistics, 2004. Cancer J Clin 2004; 54:8–29. 7. Heidelberger C, Chaudhuri NK, Danneberg P, et al. Fluorinated pyrimidines, a new class of tumour-inhibitory compounds. Nature 1957; 179:663–666. 8. Rutman RJ, Cantarow A, Paschkis KE. Studies on 2-acetylaminofluorene carcinogenesis: III. The utilization of uracil-2-C14 by pre-neoplastic rat liver. Cancer Res 1954; 14:119–126. 9. Meta-Analysis Group in Cancer. Modulation of fluorouracil by leucovorin in patients with advanced colorectal cancer: an updated meta-analysis. J Clin Oncol 2004; 22:2766–3775. 10. Van Cutsem E, Twelves C, Cassidy J, et al. Oral capecitabine compared with intravenous fluorouracil plus leucovorin in patients with metastatic colorectal cancer: results of a large Phase III study. J Clin Oncol 2001; 19:4097–4106. 11. Hoff PM, Ansari R, Batist G, et al. Comparison of oral capecitabine versus intravenous fluorouracil plus leucovorin as first-line treatment in 605 patients with metastatic colorectal cancer: results of a randomized Phase III study. J Clin Oncol 2001; 19:2282–2292. 12. Twelves C on behalf of the Xeloda Colorectal Cancer Group. Capecitabine as first-line treatment in colorectal cancer: pooled data from two large, phase III trials. Eur J Cancer 2002; 38:S15–S20. 13. Cassidy J, Twelves C, Van Cutsem, et al. First-line oral capecitabine therapy in metastatic colorectal cancer: a favourable safety profile compared with intravenous 5-fluorouracil/leucovorin. Ann Oncol 2002; 13:566–575. 14. Rothenberg ML. CPT-11 (irinotecan): an original spectrum of clinical activity. Semin Oncol 1996; 23(suppl 3):21–26. 15. Shimada Y, Yoshino M, Wakui A, et al. Phase II study of CPT-11, a new camptothecin derivative, in metastatic colorectal cancer. CPT-11 Gastrointestinal Cancer Study Group. J Clin Oncol 1993; 11:909–913. 16. Cunningham D, Pyrhonen S, James RD, et al. Randomised trial of irinotecan plus supportive care versus supportive care alone after fluorouracil failure for patients with metastatic colorectal cancer. Lancet 1998; 352:1413–1418. 17. Rougier P, Van Cutsem E, Bajetta E, et al. Randomised trial of irinotecan versus fluorouracil by continuous infusion after fluorouracil failure in patients with metastatic colorectal cancer. Lancet 1998; 352:1407–1412. 18. Saltz L, Cox JV, Blanke C, et al. Irinotecan plus fluorouracil and leucovorin for metastatic colorectal cancer. N Engl J Med 2000; 343:905–914. 19. Douillard JY, Cunningham D, Roth AD, et al. Irinotecan combined with fluorouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal cancer: a multi-centre randomised trial. Lancet 2000; 355:1041–1047. 20. Extra J-M, Marty M, Brienza, S, Misset J-L. Pharmacokinetics and safety profile of oxaliplatin. Semin Oncol 1998; 25(suppl 5):13–22. 21. Le´vi F, Misset J-L, Brienza S, et al. Chronopharmacologic phase II clinical trial with 5-fluorouracil, folinic acid, and oxaliplatin using an ambulatory multichannel programmable pump. Cancer 1992; 69:893–900. 22. Giacchetti S, Perpoint B, Zidani R, et al. Phase III multicenter randomized trial of oxaliplatin added to chronomodulated fluorouracil–leucovorin as first-line treatment of metastatic colorectal cancer. J Clin Oncol 2000; 18:136–147.

Drug Development for Advanced Colorectal Cancer

335

23. de Gramont A, Figer A, Seymour M, et al. Leucovorin and fluorouracil with or without oxaliplatin as first-line treatment in advanced colorectal cancer. J Clin Oncol 2000; 18:2938–2947. 24. Rothenberg ML, Oza AM, Bigelow RH, et al. Superiority of oxaliplatin and fluorouracil–leucovorin compared with either therapy alone in patients with progressive colorectal cancer after irinotecan and fluorouracil–leucovorin: interim results of a phase III trial. J Clin Oncol 2003; 21:2059–2069. 25. Rothenberg ML, Oza AM, Burger B, et al. Final results of a Phase III trial of 5-FU/leucovorin versus oxaliplatin versus the combination in patients with metastatic colorectal cancer following irinotecan, 5-FU, and leucovorin. Proc Am Soc Clin Oncol 2003; 22:252 (abstr 1011). 26. Goldberg RM, Sargent DJ, Morton RF, et al. A randomized controlled trial of fluorouracil plus leucovorin, irinotecan, and oxaliplatin combinations in patients with previously untreated metastatic colorectal cancer. J Clin Oncol 2004; 22:23–30. 27. Kabbinavar F, Hurwitz HI, Fehrenbacher L, et al. Phase II, randomized trial comparing bevacizumab plus fluorouracil/leucovorin with fluorouracil/leucovorin alone in patients with metastatic colorectal cancer. J Clin Oncol 2003; 21:60–65. 28. Kabbinavar FF, Schulz J, McCleod M, et al. Addition of bevacizumab to bolus fluorouracil and leucovorin in first-line metastatic colorectal cancer: results of a randomized phase II trial. J Clin Oncol 2005; 23:3697–3705. 29. Kabbinavar FF, Hambleton J, Mass RD, et al. Combined analysis of efficacy: the addition of bevacizumab to fluorouracil/leucovorin improves survival for patients with metastatic colorectal cancer. J Clin Oncol 2005; 223:3706–3712. 30. Saltz L, Rubin MS, Hochster HS, et al. Cetuximab plus irinotecan is active in irinotecan-refractory colorectal cancer that expresses epidermal growth factor receptor [abstr 7]. Proc Am Soc Clin Oncol 2001; 20:3a. 31. Saltz LB, Meropol NJ, Loehrer PJ Sr, et al. Phase II trial of cetuximab in patients with refractory colorectal cancer that expresses the epidermal growth factor receptor. J Clin Oncol 2004; 22:1201–1208. 32. Cunningham D, Humblet Y, Siena S, et al. Cetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. N Engl J Med 2004; 251:337–345. 33. Johnson JR, Williams G, Pazdur R. End points and United States Food and Drug Administration approval of oncology drugs. J Clin Oncol 2003; 21: 1404–1411.

Index

Abdominoperineal excision, Miles’ theory of, 153 Abdominoperineal excision of the rectum (APER), 198, 205 Aberration, chromosomal, 303 ACCORD 2 trial, 180 Adenine glycosylase, functional assays of, 23 Adenocarcinomas, 46, 105 Adenoma–carcinoma sequence, in colorectal cancer, 81 Adenomas, 130 distribution of, 81 prevalence of, 81 Adenomatous polyposis coli (APC) allele, 127 cellular role of, 16–17 gene, 127 gene function, 16–17 nontruncating variants in, 25–26 protein, 16 Adenomatous polyps, premalignant, 14, 48, 127, 288 Adjuvant chemotherapy, 163 Adjuvant radiotherapy facilitating sphincter sparing procedures by, 198 indications for, 197–198

[Adjuvant radiotherapy] reducing local recurrence by, 197 rendering unresectable tumors resectable by, 198 in resectable rectal cancer, 201 Adjuvant therapy, 119, 163 analysis of, 174 weakness of, 174 of colorectal cancer, 165–168 irinotecan for, 179–180 newer agents for, 176, 182 oral fluoropyrimidines for, 176–178 components of, 178 oxaliplatin for, 178–179 in patients, 174 prognostic factors and, 183–185 randomized trials of, 174 recommendations for, 185 for stage II colorectal cancer, 173–176 for stage III colorectal cancer, 164–173 Albumin, 93 Alcohol dehydrogenase (ADH), 57 Aldehyde dehydrogenase (ALDH), 57 Alignment tattoos, 212 Alpha-1 antitrypsin, 94 American Institute for Cancer Research (AICR), 52 Amsterdam criteria, 4

337

338 Anastomosis ileorectal, 19 tumor, 325 Anilinoquinazoline inhibitor, reversible, 299 Anorectal ring, 195 Anterior resection (AR), sphincterpreserving, 154 Antibodies, monoclonal, 180–182 Anticancer therapy, 287 Anti-EGFR and angiogenesis inhibitors, combination of, 242–243 Anti-idiotypicmAb strategy, 181 Apolipoprotein E (apoE), variants in, 30 Apoptosis, radiation-induced, 294 Aspirin, 50 Attenuated FAP (AFAP), 128 Autonomic nervous system, pelvic, 154 Autophosphorylation, receptor, 300 Autosomal dominant disorder, 2, 127 Autosomal dominant syndrome, 141 Autosomal recessive disorders, 21–22 Avastin1, 325

Bacille Calmette-Gue´rin (BCG) vaccination, 164 Bannayan–Zonana syndrome, 143 Barium enema, 81 Base excision repair (BER) gene, 22 bialleleic defects in, 23 Best supportive care (BSC) groups, 230 Bethesda guidelines, 4 Bevacizumab, 237–240 approval of, 325 in colorectal cancer, 327 development of, 325 in metastatic colorectal cancer, 237 use of, 239 Biases, 78 collective effect of, 79 lead/length time, 78 types of, 78–79 volunteer, 78 Biliary toxicity, 257 Bone morphogenic protein receptor 1A gene, 21

Index Bowel surveillance, 20 syndrome, 77 Brain tumors, 18 BRCA2 mutations, germline, 144 Breast cancer cell, 29, 295 Breast-ovarian cancer syndrome, 144

CAIRO trial, 236 Calcium homeostasis, 56 CALGB. See Cancer and Leukaemia Group B CALGB 89803 trial, 180 Calprotectin, 94 Camptotheca acuminate, 179 Cancer breast, 29 colonic, 151 colorectal, 303 defection of, 299 endometrial, 2 fixed/unresectable rectal, 208 gastric, 290 gastrointestinal, 21 hereditary, 125 rectal, 151, 195–196 risk of, 19 sporadic colon, 289 symptomatic, 77 Cancer and Leukaemia Group B (CALGB), 210 Cancer cell breast, 295 colon, 289, 291, 295 LoVo colon, 299 Cancer susceptibility, gastrointestinal aspects of genetic testing for, 146–147 Cancer syndrome, hereditary, 2 Cancer tissues, colorectal, 297 Capecitabine, 176 as an alternative to 5-FU, 319 approval of, 320 development of, 319 Carcinoembryonic antigen (CEA), 181

Index Carcinogenesis, colorectal, 288, 299 Carcinomas, 81 endometrial, 3 gastrointestinal, 20 submucosal spread of, 153 synchronous and metachronous, 2 CeaVac1, 181 Celecoxib and rofecoxib, 51 Cell adhesion molecule, 288 Cell cycle checkpoint kinase 2 (CHEK2), role of, 29 Cell growth factor, endothelial, 294 Cellular metabolism, 304 Cellular processes, dynamic, 303 Central nervous system, tumors of, 20 Cetuximab approval of, 328–329 in colorectal cancer, 327 development of, 327 for metastatic colorectal cancer, 240 therapy, 241 treatment, 181 21 CFR 314, 328 21 CFR 314 subpart H, 321 Chemoradiation postoperative, 205 preoperative, 205 trials performed with, 205 Chemotherapy, 11–12 adjuvant, 12, 163 benefit of, 12 cytotoxic, 229 5-FU-based, 12–13 fluorouracil (5FU) systemic, 253 hepatic arterial infusion, 254 isolated hepatic perfusion, 272 palliative, 230 portal vein infusion, 271 toxicities associated with, 238 Chemotherapy and radiotherapy (CRT) postoperative, 201–202 for rectal cancer, 196–197 regimens, 205 Chloride ion transport, regulation of, 303

339 Chromosome aberration, 303 segregation, 17 Chronic anticoagulation, 255 Circumferential resection margin (CRM), 155, 196, 200 Colectomy, 19–20 prophylactic subtotal, 129 Colon cancer cells, 289, (/) 291, 295 Colonic cancer, 151 sporadic, 289 total mesorectal excision to, principles of, 158–159 Colonic epithelium, 288 Colonic J-pouche(s), 217 anal anastomoses, versus straight coloanal anastomoses, 158 Colonoscopy, 78, 86–88 virtual, 81 Colorectal adenocarcinoma, 103 Colorectal adenomas multiple, risk of, 26 risk factor for, 28 Colorectal cancer (CRC) adenoma–carcinoma sequence, 81 adjuvant therapy of, 165–168 angiogenesis inhibitors for, 237 bevacizumab in, 325 cetuximab in, 240, 327 chemotherapeutic regimens for, 11 clinical types of, 107 in dizygous twins, 2 drug development for, 317 drugs in treatment of, 328 EGFR inhibitors for, 240–242 endpoints for new drug approval in United States for, 329–330 environmental influence on, 52 epidemiology of, 43 etiology, 30 familial, 145 FDA approval of drugs in, 320–325, 328–329 5-Fu/Fu regimens for, 230–232 future development of new drugs for, 331–332

340 [Colorectal cancer (CRC)] genetic variation and susceptibility to, 24 genetics of, 1 history of, 287, 317 immunohistochemistry in, 5 international variations in incidence for, 43 irinotecan in, 320 metastatic, 287, 302 bevacizumab in, 237 cytotoxic agents in, 230–232 first-line treatment of, 232 oxaliplatin-based regimen in, 239 second-line treatment for, 232 in monozygous twins, 2 mutations in, 6 MYH defects in, role of, 23 options in the development of drugs for, 332–333 other EGFR-targeting agents for, 242 oxaliplatin in, 322 palliative chemotherapy in, 230 pathology, 103 patients with, 244 penetrance for, 8 recessive inheritance of, 22 risk factor for, 23, 28, 58 socioeconomic status, 46 sporadic, 6 stage II, 173–176 stage III, 164 subsite of tumor, 45 as suitable target for screening, 80–82 surgical management of, 151 surgical treatment of, 14 symptoms of, 77 targeted therapies for, 236 time trends in incidence for, 44 tissues, 297 variations in incidence within countries, 45 ethnic origin, 45 Colorectal cancer cell, 303 aneuploidy in, 28

Index Colorectal cancer disorders, hereditary, 2 Colorectal cancer risk alleles, putative, 29 Colorectal cancer risk variants, putative, 25 Colorectal cancer screening, National Screening Committee recommendation on, 85 Colorectal cancer susceptibility syndromes, 20–21 recessive, 22 Colorectal cancer syndromes, 126 hereditary, 127–139 inherited, features of, 137–138 Colorectal carcinogenesis, 288, 299 Colorectal carcinoma pathologies of, 104 prognostic factors of, 114 sporadic, 6, 21 Colorectal malignant lymphomas, 120 Colorectal metastases, 254 Colorectal neoplasia, 1 APC associated risk of, 25 body mass index for, 48 calcium for, 56 carotenoids for, 56 diagnostic modality for, 81 exogenous hormones for, 49 fat and fiber for, 53 folate for, 54 meat consumption for, 52 physical activity for, 47 risk factors for, 47 tobacco smoking for, 48 vegetables and fruit for, 52 Colorectal tumorigenesis, 17 Comparative genomic hybridization (CGH) analysis, 302 Congenital hypertrophy, 128 Cowden syndrome, diagnosis of, 142–143 CRC. See Colorectal cancer Crohn’s disease, 46 CT colography, 81 Cyclin D1 (CCND1) variants, 27 Cyclooxygenase-2 (COX-2), 51

Index Cytochrome P-450, 28–29 Cytosolic enzyme, 291 Cytotoxic agents combination of, 236 in first-line treatment, 232–234 in second-line treatment, 232 strategy with, 236 Cytotoxic chemotherapy, 229 Cytotoxic effect, 255, 274 Cytotoxic platinum compounds, 296

D1822V APC variant, 26 de Gramont regimen, 231 de Gramont schedule, 173 Denonvillier’s fascia, 156 Deoxythymidine triphosphate, 291 Deoxyuridine monophosphate, 291 Desmoid disease, 17 Dexamethasone, 258 Diaminocyclohexane platinum, 322. See also Oxaliplatin Dietary folate consumption, variants in, 27 Dihydropyrimidine dehydrogenase (DPD), 293 DNA extraction, 95 DNA microarray profiling, 303 DNA mismatch repair genes, 5–6 germline mutations in, 8 low-penetrance variants in, 26–27 mutations in, penetrance of, 8 systematic analysis of, 8 DNA mismatch repair system, 289 defective, 11–12 effect of, 11 DNA replication slippage, 5 Drug metabolizing enzymes, 287 Drug resistance; platinum, 297 approval in United States for colorectal cancer, 329–330 bevacizumab, 325 cetuximab, 327 development for advanced colorectal cancer in the United States, 317 development in future, 331–332

341 [Drug resistance; platinum] endpoints in the approval of new, 329–330 irinotecan, 320–321 options involved in development of, 332–333 oxaliplatin, 322 PFS and TTP use as endpoints for development of, 330 in treatment of colorectal cancer, 328 Dukes B and C rectal cancer, 202 Dukes’ stage C disease, chemotherapy for, 81 Dukes’ staging, 199 Dutch CKVO trial, 199–200, 203 Dutch study (CKVO 95–04), 204–205 Dutch TME study, 208 Dutch TME trial, 202 Dynamic cellular processes, 303

E1317Q APC variant, 26 EGFR. See Epidermal growth factor receptor Embryonic fibroblasts, 295 Endocavity radiation, 211 Endometrial cancer, 2 penetrance for, 8 Endometrial carcinomas, 3 Endoscopic screening, 135 Endothelial cell growth factor, 294 receptors, 302 Eniluracil, 235 Enzyme(s), cytosolic, 291 drug metabolizing, 287 proteolytic, 94 Enzyme systems, polymorphic, 287 Epidermal growth factor receptor (EGFR), 163, 181, 237, 299 expressing tumors, 327 inhibitors for colorectal cancer, 240–242 staining, 327 targeting agents, 242 Epithelium, colonic, 288 Erlangen series, 196

342 Escherichia coli, 23 European multicenter study of rectal cancer, 196 European Organization for Research and Treatment of Cancer (EORTC), 235 European study in T3/T4 resectable rectal cancer (EORTC 22921), 207 Exocrine carcinomas, 120 Extracolonic tumors, 128 Familial adenomatous polyposis (FAP), 3, 14–16, 46, 108 chemoprevention, 131 clinical management of, 128–131 clinical manifestations of, 128 extracolonic features in, 129 genetic linkage study of, 16 genetic testing for, 131 operative strategy in, 18–19 patients germline mutations in, 16 rectal polyps in, 15 risk of, 15 screening, recommendations for, 129 Familial colorectal cancer, 145 Familial gastric cancer, 144 Familial pancreatic cancer, 144 FAP. See Familial adenomatous polyposis Fecal bile acid, 54 Fecal occult blood (FOB) screening, 82–85 tests, 82, 84 Flexible sigmoidoscopy, 81, 85–86 Florid polyposis, 128 Fluorinated thymidine analogs, 11 Fluorodeoxyuridine (FUDR), 254 Fluorodeoxyuridine monophosphate (FdUMP), 230 5-fluorouracil (5-FU), 163, 229, 317, 318 approach to, 173 based chemotherapy, 12–13, 201 bolus, 319 drug administration in United States, 319

Index [5-fluorouracil (5-FU)] efficacy of, 172 intravenous, use of, 176 levamisole, 202 new drug approval for, 318 regimens, 319 systemic chemotherapy, 253 toxicity profile of, 173 use of intravenous (IV), 176 versus 5FU and levamisole, 202 versus 5FU plus low-dose folinic acid, 202 Foki polymorphism, 56 Foley catheter, 213 FOLFIRI, 180 regimens, 232 FOLFOX4, 178–179 and LV5FU2 groups, difference between, 324 regimen, 323 FOLFOXIRI, 234 Folinic acid (FA), 165, 230 Follicular hyperplasia, 107 French trial (FFCD 9203), 207 Ftorafur, 176 Gardner’s syndrome, 16, 128 Gastric cancer(s), 290 familial, 144 Gastric carcinoma, Borrmann categories of, 105 Gastrointestinal (GI) cancer, 21 susceptibility, aspects of genetic testing for, 146–147 syndromes, 126 Gastrointestinal carcinomas, 20 Gastrointestinal stromal tumors (GIST), 120 Genetic aberration, types of, 303 Genetic polymorphisms, 28 Genetic susceptibility, 47 Genetic testing, 18 Genomic polymorphisms, 287 Genotype–phenotype correlations, 17–18 Germline BRCA2 mutations, 144

Index Germline mutations, 127 Glutathione S-transferase, 53, 297 Glycoprotein antigen, 180 Growth hormone-1 gene (GH1), 59 Guaiac tests, 82 HAI. See Hepatic arterial infusion Hamartamatous polyps, 143 Hamartomas, 143 Hand–foot syndrome, 320 Health professional follow-up study (HPFS), 57 Healthy user effect, 50 Heme, 82 Hemeagglutination, 82 Hemeoccult II test, 83 Hemoglobin, human, 82 Hemorrhoids, 77 Hepatic arterial infusion (HAI), 254, 257 chemotherapy, 254 nonrandomized trials of, 258 randomized trials of, 258 with systemic chemotherapy, 267 for metastatic disease, 258 as neoadjuvant therapy, 269 with new chemotherapy agents, 265 second-line chemotherapy, 270 Hepatic artery, 256 catheter, 255 pump, 255 Hepatobiliary toxicity, 257 Hereditary cancer syndrome, 2 Hereditary cancers, 125 Hereditary colorectal cancer disorders, 2 Hereditary colorectal cancer syndromes, 127–139 Hereditary flat adenoma syndrome (HFAS), 108 Hereditary mixed polyposis, 144–145 Hereditary nonpolyposis colon cancer (HNPCC), 2, 46, 107, 126, 131–139, 289 Amsterdam criteria for, 133 Bethesda criteria for, 132 cancer control in, 13–14 categorization of, 4 characteristics of, 4

343 [Hereditary nonpolyposis colon cancer (HNPCC)] chemotherapy in, 12–13 clinical criteria for, 132–134 clinical management of, 135–136 clinical manifestations of, 132 colorectal cancer, prognosis in, 9–11 diagnostic criteria and clinical features of, 4–5 family, 3 genetic testing for, 134, 136–139 microsatellite instability (MSI) in, 7 patients, germline mutations in, 6 penetrance of, 8–9 phenotype of, 26 study of, 10 tumors Hereditary polyposis syndrome, 144 Hereditary syndromes, 126 Heterocylic amines, 53 Histological parameters, 107 Histomorphology, 105 Histopathologic TNM stage, 183 Histopathological grading classes for, 106 classification of four grades for, 106 Histopathological report, 114 HNPCC. See Hereditary nonpolyposis colon cancer Hormone replacement therapy (HRT), 49 Hyperbilirubinemia, 320 Hyperglycemia, 58 Hyperinsulinemia, 58 Hyperplastic polyps, 47, 48 Hypertrophy, congenital, 128 Hysterectomy, prophylactic, 8 IFL regimen, 234, 321 IFL+bevacizumab regimen, 326 IFL+placebo, 326 Ileorectal anastomosis (IRA), 14, 19 Immunohistochemical protocols, 296 Immunohistochemistry, 133, 135

344 Inflammatory bowel disease, 108 Inherited syndromes, 127 Insulin-like growth factor-1 (IGF-1), 58 Insulin-like growth factor II gene (IGF2), 30 Interferon (IFN) a-2a, 170 Interleukin-6, genes coding for, 51 International Agency for Research on Cancer (IARC), 48 International Documentation System for Colorectal Cancer (IDS for CRC), 104 International Multicentre Pooled Analysis of Colon Cancer Trials (IMPACT), 171 B2 analysis, 175 Intestinal mucosa, 235 Irinotecan, 298 approval of, 321 comparison of, 320 development as second-line chemotherapy, 320 drug application for, 323 integration into front-line chemotherapy, 321–322 and oxaliplatin combination, 324 phase trials of, 320 regimen, 232 Irinotecan+cetuximab arm, 327 Irinotecan-refractory patient, 240 Irritable bowel syndrome, 77 Isolated hepatic perfusion chemotherapy, 272 Isolated tumor cells (ITC), definition of, 110 Juvenile polyposis (JPS), 20–21, 141 molecular basis of, 21 Kaposi sarcoma, 120 Lateral pelvic lymph node dissection, 199 Leucovorin (LV), 169, 235, 317

Index Levamisole, 170 Liver metastases, 254 adjuvant HAI post resection of, 262 neoadjuvant therapy for, 269 Local spread extent of resection for, 108 types of, 108 LoVo colon cancer cells, 299 LV5FU2. See de Gramont schedule LV5FU2 regimen, 231, 322–323 Lymphatic spread, and lymph node dissection, 109 Lymphatic system, intramural, 153 Lymph drainage, bidirectional, 110 Lymph node metastases, 183 Lynch syndrome, 132. See also Hereditary nonpolyposis colon cancer (HNPCC) criteria of, 14 summary of, 15 Lyon experience, 211

Magnetic resonance imaging, 257 Malignant tumors, 120 Mayo Clinic regimen, 319 Memorial Sloan Kettering Cancer Center (MSKCC), 259 MERCURY study, 196 U.K. based, 212 Mesorectum, 152, 155 Metabolism, cellular, 304 Metastases, extramural, 153 Metastatic colorectal cancer, 287, 298, 300, 302. See also Colorectal cancer (CRC), metastatic Metastatic tumors, 299 Methylene tetrahydrofolate, 169 Methylene tetrahydrofolate reductase (MTHFR), 27–28 5,10-Methylene tetrahydrofolate reductase (MTHFR) genotype, 55 genetic variation in, 28 Micrometastasis, definition of, 110 Microsatellite instability (MSI), 133–135, 184, 289–290

Index [Microsatellite instability (MSI)] characteristics of, 6 in colorectal cancer, 6–8 definition of, 289 in hereditary nonpolyposis colon cancer (HNPCC) tumor, 7 prognostic implications of, 9 in tumors, 6 Microsatellite instability (MSI) and MSS tumors, comparison of, 9 colorectal cancers, study of, 12 tumors in colorectal cancer patients, 10 multivariate analysis of, 13 sporadic, 10 chemotherapy in, 12–13 Microsatellite markers, 7 Microtubule cytoskeleton, structure of, 17 Mitogenic effects, 59 Mitotic chromosomal segregation, 28 MMR deficient cells, sensitivity of, 11 MMR gene, 136 Monoclonal antibodies (mAbs), 163, 180–182 Moran triple stapling technique, 156 Mortality rates, 274 MOSAIC trial, 179 Mouse–human chimeric IgG1 monoclonal antibody, 327 MSI. See Microsatellite instability Mucinous adenocarcinomas, 106 Muir-Torre syndrome, 4 Multifunctional polypeptide, 290 Muscularis propria, 210 Mutation accumulation of, 288 germline, 127 heterozygous, 23 MYH, 128 somatic, 289, 290 MutY protein, 23 MYH-associated polyposis (MAP), 22–24 clinical management of, 24

345 [MYH-associated polyposis (MAP)] molecular pathology of, 23 MYH gene, 127 MYH gene testing, 23 MYH mutations, 128 bialleleic, 24

N-acetyltransferase 1 (NAT1), 53 National Cancer Institute, 259 National Screening Committee recommendation on colorectal cancer screening, 85 National Surgical Adjuvant Breast and Bowel Project (NSABP), 164, 201 C-06 trial, 177 R-03 trial, 217 Neoadjuvant therapy, 119, 158 Neoplasia, 88 Neoplasms, 96 Nerve plexuses, hypogastric, superior and inferior, 154 Neuroendocrine tumors, classification of, 120 Nonsteroidal anti-inflammatory drugs (NSAIDs), 50 types of, 51 Normal tissue tolerances, 214 North central cancer treatment group (NCCTG), 170, 175, 261 Norwegian Colorectal Cancer Prevention (NORCCAP) Screening Study, 89 Norwegian rectal cancer project, 157 Nottingham screening, side-effect of, 83 Nuclear magnetic resonance spectroscopy, 306 Nucleotide excision repair (NER), 297

Oncologic Drugs Advisory Committee (ODAC), 321 Oncological quality, 116

346 Optimal chemotherapy, 254 Oral contraceptive pill, 50 Oral fluoropyrimidines, 234–236, 265 Ornithine decarboxylase gene (OC), 51 Oxaliplatin, 232, 296, 322 demonstration of second-line efficacy, 323 to front-line chemotherapy, 322 and irinotecan combination, 324 platinum-DNA adducts formed by, 322

Palliative chemotherapy, benefit of, 230 Pancreatic cancer, familial, 144 Pathogenic mutations, 14 Pathogenic variants, 26 Pathological complete response (PCR), 207 Perimuscular tissue discontinuous spread, 108 Perineural invasion, 107 Peritumorous inflammation, 109 PETACC-3 trial, 180 Peutz–Jeghers and Cowden syndrome, 126 Peutz–Jeghers pigmentation, 139 Peutz–Jeghers polyps, 139 Peutz–Jeghers syndrome (PJS), 20, 139–141 diagnosis of, 139 p53 Tumor suppressor gene, 294 Pharmacokinetics, 287 PJS. See Peutz–Jeghers syndrome Platinum compounds, cytotoxic, 296 Platinum drug resistance, 297 Polish trial, 200 Polycyclic aromatic hydrocarbons, metabolism of, 28 Polymerase chain reaction (PCR), 95, 110 Polymorphic enzyme systems, 287 Polymorphic genes, 51 Polymorphisms, genomic, 287

Index Polypeptide, multifunctional, 290 Polyposis adenomatous, 288 development, 129 florid, 128 hereditary, mixed, 144–145 juvenile, 141 phenotype, 25 rectal remnant, 19 Polyps, 131 adenomatous, 127 hamartamatous, 143 Polysaccharide K (PSK), 177 Portal vein infusion chemotherapy, 271 Portal vein thrombosis, 256 Potassium oxonate, 178 Potential therapeutic targets, 302 Preoperative chemoradiation vs. postoperative chemoradiation, 205–207 vs. preoperative short course radiation, 208 vs. radiation, 207 Proctectomy, 19 Proctocolectomy, 18 with ileoanal pouch, 19 Progression-free survival (PFS), 321, 327 improvement in, 322 use as endpoints for drug development, 330 Prolonged venous infusion (PVI), 202 Prophlylactic colectomy, 15 Prophylactic subtotal colectomy, 129 Prophylactic surgery, 18, 24 Protein, stabilization of, 294 Protein phosphotase gene, germline mutations in, 21 Proteolytic enzymes, 94 Proteomic analysis, 306 Protocol, screening, 86 Protracted venous infusion (PVI), 202 PTEN hamartoma syndrome, 143 Pump insertion, 256

Index Quality management, 116 Quick and simple and reliable (QUASAR) collaborative group, 171, 175

Radiation-induced apoptosis, 294 Radiation treatment, antiproliferative effect of, 300 Radio-opaque tampon, 213 Radiology, 88 Radiopotentiating effects, 300 Radiotherapy acute toxicity and supportive care during, 214 adjuvant, 197 after long/short course, 215–216 intraoperative, 209 preoperative, 202–203 meta-analyses of, 203 timing of surgery following, 215 Radiotherapy planning techniques, 212 computed tomography scanning procedure of, 213 conventional planning in, 212–213 definition of, 212 Raltitrexed, 232 Rash, incidence and significance of, 243–244 Receptor autophosphorylation, 300 Rectal cancer, 84, 151, 195–196 adjuvant radiation techniques for, 196–197 adjuvant therapy in, 201 chemotherapy and radiotherapy (CRT) for, 196–197 circumferential or radial margin of excision in, 199–200 early, 209–210 embryological approach to, 152 European multicenter study of, 196 excision, principles of, 152–153 fixed/unresectable, 208–209 late effects of, 216–217 local recurrence of lymphatic drainage pathways for, 199

347 [Rectal cancer] management of, 152 multidisciplinary approach, 152 operation for, 153 optimal operation for, 159 proximal and distal surgical margins in, 199 resectable, 201 short course preoperative radiotherapy (SCPRT) for, 197 spread of, 156 surgery basics of, 152 recurrence and survival of, 156–158 surgical complications of, 216 survival in, 159 TME surgery for concept of, 157 zone of downward spread, 153 zone of upward spread, 153 clinical features for, 198 histological grade for, 199 prognostic factors for, 198–200 quality of surgery for, 200 risk factors for, 198 sites for, 198–199 T and N stage in, 199 Rectal polyps, in familial adenomatous polyposis (FAP), 15 Rectosigmoid, 195 Rectum, extra-fascial excision of, 155 Refractory metastatic colorectal cancer, 298 Response rate (RR), 230 Reversible anilinoquinazoline inhibitor, 299 RNA and protein synthesis, 304 Roswell park regimen, 319 RTOG 8902, 210

Scandinavian polyposis registers, study of, 20 SCPRT. See Short course preoperative radiotherapy

348 Screening, 146 definition of, 78 economics of, 91–93 endoscopic, 135 fecal occult blood, 82–85 harm caused by, 89–91 Nottingham side-effect of, 83 principles of, 78–80 protocol, 86 Screening methods, comparative study, 88–89 Short course preoperative radiotherapy (SCPRT) aim of, 203 blanket approach in, 204 rationale for, 203–204 for rectal cancer, 197 versus surgery alone or postoperative radiotherapy, 204 Sigmoidoscopy, flexible, 81, 85–86 Signet-ring cells, 106 Small bowel contrast, 213–214 tolerance, 216 Somatic mutation, 289, 290 Southwest oncology group (SWOG), 170 Sphincter preserving surgery, 153, 198 anterior resection (AR), 154 Spigelman’s score and classification, 130 Spigelman stage, 130 Stage III colorectal cancer definition of, 164 division of, 164 Stapling devices, circular, 154 STK15 (Aurora-A), 28 Stockholm colorectal cancer studies, 203 Supra-additive inhibitory effect, 300 Surgery colorectal, 154 prophylactic, 24 rectal cancer, 152 sphincter preserving, 153, 198

Index Surveillance, Epidemiology and End Results (SEER) program, 46 Swedish Rectal Cancer trial, 203–204, 208 SWOG-INT trial, 175 Symptomatic cancers, 77 Syndrome autosomal dominant, 141 Bannayan–Zonana, 143 breast-ovarian cancer, 144 Gardner’s, 128 hand–foot, 320 hereditary, 126 colorectal cancer, 127–139 polyposis, 144 inherited, 127 Lynch I, 132 Lynch II, 132 Peutz–Jeghers and Cowden, 126 PTEN hamartoma, 143 von Hippel–Lindau, 144

T and N stage of rectal cancer, 199 T3T4 N0-3 rectal cancer, 202 T4 disease, definition, 199 Tegafur, 176, 178 Ternary complex, 291 TGF-b. See Transforming growth factor-b TGFb II mutation, 290–291 Therapeutic agents, development of, 288 Therapy adjuvant, 163 anticancer, 287 cetuximab, 241 Thymidine phosphorylase, 176, 294 Thymidylate synthase (TS), 164, 230 Thymidylate synthase inhibitors, 11 Thyroid malignancy, 143 Thyroid ultrasound, 143 Time to tumor progression (TTP), 231, 320 use as endpoints for drug development, 330 TNM staging, 199

Index Total mesorectal excision (TME), 154, 157, 196, 198 anatomical definition of, 155 Danish study, 158 Dutch series, 157 operation of, 155 rationale for, 156 resection, 200 surgery, 200 trial, 202 Transferrin, immunological detection of, 93 Transforming growth factor-b (TGF-b), 21 apoptosis-promoting role of, 30 Transphosphorylation, 290 Treatment, anticancer, 287 Tumor angiogenesis, 325 Tumor cells, local spillage of, 109 Tumor growth, 288 Tumor microsatellite instability (MSI) phenotype, 9 Tumor regression, histological grading of, 114 Tumor remission, transient, 301 Tumor resection, 116 Tumor response, 298 Tumor-infiltrating lymphocytes (TILS), 107 Tumors categorization of, 7 of central nervous system, 20 extracolonic, 3, 128 metachronous, 2 molecular characteristic of, 6 MRI staging of, study of, 158 synchronous, 2 Tumor shrinkage, 198 Tumor spread, 108

349 [Tumor spread] distal, microscopic, 154 Tumor suppressor gene, 288, 294 Turcot’s syndrome, 18, 20 Tyrosine kinase inhibitors, 299 receptors, 301

UICC TNM classification, 109 UK-based MERCURY study, 212 Ultrasound, thyroid, 143 Upsala trial, 204, 217 Uracil/ftorafur (UFT), 176 cytotoxic component of, 176

Valine variant, 26 Vascular endothelial growth factor (VEGF), 163, 237, 301 negative tumors, 184 Venous invasion, 107 Venous thromboembolism, 238 Virtual colonoscopy, 81 Visceral pelvic fascia, 156 Vitamin D receptor (VDR), 56 von Hippel–Lindau syndrome, 144

Wide local excision  radiotherapy, 210–211 Wolffian ridge layer, 156 World Cancer Research Fund (WCRF), 52

Xeroderma pigmentosum group D, 297

Figure 1.2

Expression of MLH1 in histological sections. (See p. 5)