The Challenge of Colorectal Cancer: A Review Book Editor
Esther U帽a Cid贸n Clinical Oncology Department, Digestive Oncology Unit Clinical University Hospital of Valladolid, Spain
Research Signpost, T.C. 37/661 (2), Fort P.O., Trivandrum-695 023 Kerala, India
Published by Research Signpost 2011; Rights Reserved Research Signpost T.C. 37/661(2), Fort P.O., Trivandrum-695 023, Kerala, India Editor Esther U帽a Cid贸n Managing Editor S.G. Pandalai Publication Manager A. Gayathri Research Signpost and the Editor assume no responsibility for the opinions and statements advanced by contributors ISBN: 978-81-308-0460-6
Contents
Contributors Foreword Introduction Chapter 1 Colorectal cancer epidemiology. Assessing and managing risk. Prevention María Luque Cabal Chapter 2 Molecular biology of colorectal cancer Enrique Grande-Pulido, Alejandro Riquelme-Oliveira Javier Ballesteros-Bargues, Carmen Guillén-Ponce and Alfredo Carrato
1
35
Chapter 3 Imaging of colorectal cancer: Established techniques and emerging modalities Rachel E. Hyland and Fahmid U. Chowdhury
53
Chapter 4 Local management of primary colorectal cancer: Focus on surgery Henar Núñez, Esther Uña, José Herreros, Carlos Abril and Alejandro Romero
75
Chapter 5 Adjuvant systemic treatment for colon cancer José María Vieitez de Prado, Paula J Fonseca and Madalina Frunza
103
Chapter 6 Metastatic colorectal cancer: Focus on systemic treatment Paula García Teijido, Teresa Sampedro Gimeno and Pilar Blay Albors
139
Chapter 7 Site-specific therapy of metastatic colorectal cancer: Focus on surgical treatment of liver metastases Miguel Ángel Suárez Muñoz, Julio Santoyo Santoyo José Luis Fernández Aguilar, Belinda Sánchez Pérez José Antonio Pérez Daga and Mª Carmen Leiva Vera
157
Chapter 8 Lung metastases of colorectal cancer: Focus on surgery José Luis Duque-Medina, Manuel Castanedo and Henar Borrego
187
Chapter 9 Peritoneal carcinomatosis of colorectal cancer Bernardino Rampone
213
Chapter 10 The role of radiotherapy in colorectal cancer Francisco López-Lara and Patricia Diezhandino García
227
Chapter 11 Locoregional techniques for liver-limited metastatic colorectal cancer Virginia Arrazubi, José Luis del Cura and María Teresa Pérez-Hoyos
263
Chapter 12 Prognostic and predictive factors in colorectal cancer: The importance of reliable markers for effective selection of therapy Miriam López-Gómez, Enrique Casado, Paloma Cejas and Jaime Feliu
285
Chapter 13 Inherited syndromes of colorectal cancer and genetic counselling Ignacio Blanco, Conxi Lázaro and Gabriel Capella
309
Chapter 14 Issues in colorectal cancer survivors Esther Uña Cidón, Jesús Crespo and Rosa Bustamante
331
Contributors María Luque-Cabal Clinical Oncology Service Central University Hospital of Oviedo, Spain Enrique Grande-Pulido Medical Oncology Department “Ramón y Cajal” University Hospital, Madrid, Spain Alejandro Riquelme-Oliveira Medical Oncology Department “Ramón y Cajal” University Hospital, Madrid, Spain Javier Ballesteros-Bargues Medical Oncology Department “Ramón y Cajal” University Hospital, Madrid, Spain Carmen Guillén-Ponce Medical Oncology Department. “Ramón y Cajal” University Hospital, Madrid, Spain Alfredo Carrato Head of Medical Oncology Department. “Ramón y Cajal” University Hospital, Madrid, Spain Rachel E. Hyland Consultant Radiologist The General Infirmary at Leeds, Leeds, UK Fahmid U. Chowdhury Consultant Radiologist and Nuclear Medicine Physician St James’s University Hospital, Leeds, UK Henar Núñez Colorectal Cancer Surgery Unit Clinical University Hospital of Valladolid, Spain Esther Uña Clinical Oncology Department, Digestive Oncology Unit Clinical University Hospital of Valladolid Spain
José Herreros Colorectal Cancer Surgery Unit Clinical University Hospital of Valladolid, Spain Carlos Abril Colorectal Cancer Surgery Unit Clinical University Hospital of Valladolid, Spain Alejandro Romero Colorectal Cancer Surgery Unit Clinical University Hospital of Valladolid, Spain José María Vieitez de Prado Medical Oncology Department Central University Hospital of Asturias, Oviedo, Spain Paula J Fonseca Medical Oncology Department Central University Hospital of Asturias, Oviedo, Spain Madalina Frunza Surgery Department, Central University Hospital of Asturias Oviedo, Spain Paula García Teijido Medical Oncology Service, “San Agustín” Hospital, Spain Teresa Sampedro Gimeno Medical Oncology Service, Hospital of Cabueñes, Spain Pilar Blay Albors Medical Oncology Service Central University Hospital of Asturias, Oviedo, Spain Miguel Ángel Suárez Muñoz General, Digestive and Transplantation Surgery Service “Carlos Haya” University Hospital, Málaga, Spain Julio Santoyo Santoyo General, Digestive and Transplantation Surgery Service “Carlos Haya” University Hospital, Málaga, Spain José Luis Fernández Aguilar General, Digestive and Transplantation Surgery Service “Carlos Haya” University Hospital, Málaga, Spain
Belinda Sánchez Pérez General, Digestive and Transplantation Surgery Service “Carlos Haya” University Hospital, Málaga, Spain José Antonio Pérez Daga General, Digestive and Transplantation Surgery Service “Carlos Haya” University Hospital, Málaga, Spain Mª Carmen Leiva Vera General, Digestive and Transplantation Surgery Service “Carlos Haya” University Hospital, Málaga, Spain José Luis Duque-Medina Thoracic Surgery Department Clinical University Hospital of Valladolid, Spain Manuel Castanedo Thoracic Surgery Department Clinical University Hospital of Valladolid, Spain Henar Borrego Department of Pathology Clinical University Hospital of Valladolid, Spain Bernardino Rampone Department of Surgery Pineta Grande Hospital, Castel Volturno, Italy Francisco López-Lara Head of Radiation Oncology Department Clinical University Hospital of Valladolid, Spain Patricia Diezhandino. García Radiation Oncology Department Clinical University Hospital of Valladolid, Spain Virginia Arrazubi Department of Medical Oncology, Hospital de Basurto, Bilbao, Spain José Luis del Cura Department of Radiology, Hospital de Basurto, Bilbao, Spain María Teresa Pérez-Hoyos Department of Medical Oncology, Hospital de Basurto, Bilbao, Spain
Ignacio Blanco Head of the Genetic Counselling Unit, Hereditary Cancer Programme Catalan Institute of Oncology, Spain Conxi Lázaro Head of the Molecular Diagnostic Counselling Unit, Hereditary Cancer Programme, Catalan Institute of Oncology, Spain Gabriel Capella Head of the Molecular Diagnostic Counselling Unit, Hereditary Cancer Programme, Catalan Institute of Oncology, Spain Jesús Crespo-Sanjuan Biochemical and Molecular Biology Department, Clinical University Hospital of Valladolid, Spain Rosa Bustamante Head of Biochemical and Molecular Biology Department Clinical University Hospital of Valladolid, Spain Miriam López-Gómez Department of Medical Oncology, “Infanta Sofía” Hospital of San Sebastián de los Reyes, Madrid, Spain Enrique Casado Department of Medical Oncology “Infanta Sofía” Hospital of San Sebastián de los Reyes, Madrid, Spain Paloma Cejas Department of Medical Oncology, La Paz University Hospital Madrid, Spain Jaime Feliú Department of Medical Oncology, La Paz University Hospital Madrid, Spain
Foreword Dr. Uña has written a key manual on colorectal cancer. I am pleased to see such a resource available to physicians. “THE CHALLENGE OF COLORECTAL CANCER: REVIEW BOOK” should be on every instructor’s bookshelf, and an integral part of patient care. As a professor of medicine and department head at one of the leading oncology institutions in the world, I have seen a need for a greater analytical tool in this area. Dr. Uña’s insights, so concisely communicated, offer the bridge between the study of this complex disease and the practitioners who address these challenges with their patients face to face everyday. Dr. Uña has used her years of experience to describe the public health problem of colorectal cancer in this well-organized book. Her personal experiences in dealing with colorectal cancer patients add a tremendous amount of real-world relevance. One of the most impressive aspects of this book is its concise coverage of all the relevant issues colorectal cancer patients and their doctors face when dealing with this disease. I wish an almanac like THE CHALLENGE OF COLORECTAL CANCER: A REVIEW BOOK is available to every physician treating colorectal cancer patients. It should help many emerging physicians in this field avoid unnecessary mistakes and hours of research, since so much territory is covered in such an accessible way. The greatest benefit in the end is to the patient, who no doubt will receive more comprehensive care from the book’s reader. Fairooz Kabbinavar, M.D. FACP Professor of Medicine and Urologic Oncology Chair in Kidney Cancer Research Medical Director Thoracic Oncology and Kidney Cancer Programs Co-Director- Hematology-Oncology Fellowship Program David Geffen School of Medicine at UCLA 924 Westwood Blvd, Suite 1050 Los Angeles, CA 90024 fkabbina@mednet.ucla.edu
Introduction Colorectal cancer (CRC) is a major public health problem worldwide. Each year, this disease affects thousands of people, and it is still the second leading cause of mortality. In fact it has been reported that CRC affects nearly 800,000 individuals around the world and is associated with more than 500,000 deaths. Over the past years, significant advances have been made in the screening and early detection of CRC as well as in the different treatment modalities and approaches, including surgery, radiation therapy, and chemotherapy, which are now available for patients regardless the stages of the disease. Beyond doubt, CRC is now a highly preventable and curable disease through regular screenings and early detection, and highly treatable even in more advanced stages. In this book, we have condensed a wealth of information on the epidemiology, molecular biology, diagnosis, different treatment approaches for patients with this disease, hereditary syndromes and issues affecting CRC survivors, and we present essential information in a practical, easy to understand and readable format. It has been crucial that a team of physicians from various specialties, including surgery, radiation oncology, medical oncology, pathology, radiology, molecular biology, provide their special expertise in developing a deep review about all the aspects related to CRC, namely, the epidemiology, screening and prevention, diagnosis, treatment plans and problems affecting survivors which are increasing. The first chapter focuses on the epidemiology and prevention. The second one gives a condensed explanation about the molecular biology of CRC and the key pathways involved in cancer formation. The third which is devoted to update the diagnosis techniques, well established ones and emerging modalities. Then there are eight chapters which develop a nice overview of the different treatment approaches for this disease (adjuvant and palliative systemic treatments, the integration of biologic agents in treatment regimens, local and locoregional treatments including surgery and special techniques such as radiofrequency ablation and there is also a chapter about radiotherapy). All these chapters provide a timely review of the most up-to-date information presently available. The chapter 12th provides us with an interesting review about the role of prognostic and predictive biomarkers which have been developed or which are in developing for this disease.
The last two chapters focus on the hereditary syndromes of CRC and the issues affecting survivors which are a group whose members are increasing in number in a continuous way thanks to the improvements in screening, prevention and treatment. I hope that this book can serve as a source of practical information that can be used by physicians and other health care professionals actively involved in the daily care of patients with CRC. It represents the efforts of many dedicated people whose dedication to this disease is reflected in the high quality of their contributions. Finally, this book should be considered as a work in progress, and our expectation would be to update it in the future and to include new treatments and strategies that reflect the rapid advances in CRC. Esther U帽a Cid贸n, MD, PhD, Professor Digestive Oncology Unit Clinical Oncology Department Clinical University Hospital and Faculty of Medicine Valladolid, Spain
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 1-33 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
1. Colorectal cancer epidemiology. Assessing and managing risk. Prevention María Luque Cabal Clinical Oncology Service, Central University Hospital of Oviedo, Spain
Abstract. Colorectal cancer (CRC) is an important public health problem: there are nearly one million new cases of CRC diagnosed world-wide each year and half a million deaths. Recent reports show that it is one of the most frequent form of cancer among people aged 75 years and older. Given that the majority of cancers occur in elderly people and with the ageing of the population in mind, this observation gives further impetus to investigating prevention and treatment strategies among this subgroup of the population. Screening recommendations and implementation is nowadays a priority and achieving CRC control is the immediate challenge.
Introduction Most CRC, at least two-thirds and perhaps as much as 90%, arise from benign, adenomatous polyps lining the wall of the bowel. Those which grow to a large size and have a villous appearance or contain dysplastic cells are the most likely to progress to cancer. Correspondence/Reprint request: Dr. María Luque Cabal, Clinical Oncology Service, Central University Hospital of Oviedo, Spain. E-mail: malucab@hotmail.com
2
MarĂa Luque Cabal
The natural history and the role of several risk factors in the aetiology of CRC are becoming more clearly understood and the genetic events involved in this disease susceptibility are more clear for us. While there are many questions to be solved, it is apparent that many facets of CRC are becoming increasingly understood and prospects for prevention are becoming apparent.
1. Epidemiology of colorectal cancer Colorectal carcinoma (CRC) is the fourth most common cancer in men and the third most common cancer in women worldwide. In 2002, cancers of the colon and rectum accounted for around 1 million of all new cancer cases (9.4% of the world total), and in contrast to cancers in most other sites, only a minimal difference in incidence was observed between men and women (ratio, 1.2:1) (1). More than 142,000 cases of CRC will have been diagnosed in the United States in 2010. These will have included around 40,000 cases of rectal carcinoma. In 2008, the age-standardised incidence rates were 34.1 per 100,000 men and 25 per 100,000 women (2). In the European Union, 332,000 new cases were reported in 2008. Agestandardised incidence rates ranged between 27.6 per 100,000 men in Finland and 60.7 per 100,000 men in the Czech Republic (3). Five year survival rates as high as 90% have been reported when CRC is diagnosed at an early stage. In general, the mortality rate is around half of the incidence rate. In 2002, for example, approximately 529,000 CRC-related deaths occurred in the world. Worldwide, the prevalence of CRC is second only to that of breast cancer. Research has estimated that 2.8 million CRC patients are still alive within 5 years of diagnosis (1). It is estimated that a total of 51,370 deaths will have occurred secondary to CRC in the United States in 2010, and these will have accounted for 9% of all cancer deaths. Mortality rates for CRC have decreased over the past two decades, particularly in recent years, as a result of improvements in early detection and treatment (2). The most recent data from the European Union are those from 2008, during which 148,000 deaths occurred secondary to CRC. A recent study compared population-based data from the 109,953 colon cancer patients of the Surveillance, Epidemiology, and End Results (SEER) programme with those from the 134,206 colon cancer patients of the National Cancer Data Base (NCDB) (4). This demonstrated that 5-year survival ranged between 76% +/- 0.3% for patients with stage I (T1-2N0) disease, and was
Colorectal cancer epidemiology
3
26.5% for patients with stage IIIC (T4N2) disease. The relative survival (i.e., cancer survival in the absence of other causes of death) was longer in cases with early stage disease, and reached a level of 97.1% +/- 0.4% for stage I. These data indicate that most patients with stage I disease in this cohort died from non-CRC related causes. An analysis of data from the population-based registries of the National Cancer Institute (NCI) SEER programme yielded an adjusted 5-year conditional probability of survival (i.e., the probability of survival following a 5-year survival period) in which the greatest improvement was seen in cases with an advanced stage of CRC disease, and in which the rates of survival reached ≼ 80%. One exception was stage IV disease, for which the adjusted 5-year conditional probability of survival was 48% (5).
• Subsite distribution Approximately 70% of all CRCs are located in the distal or left large bowel. However, several studies have demonstrated a shift towards an increase in proximal lesions, particularly in the incidence of right-sided cancers (6-8). It is unclear whether this is a genuine phenomenon or simply an artefact of factors such as lack of consensus across studies concerning the most appropriate division of the colorectum into anatomical subsites, or an increase in the excision of adenomatous polyps in the descending colon following screening with flexible sigmoidoscopy. An alternative hypothesis is that screening colonoscopy is more effective in protecting the left colon from cancer compared with the right colon (9). However, this effect may also be related to particular aspects of the colonoscopy procedure, such as poor right-sided bowel preparation, incomplete colonoscopy, and the presence of anatomical configurations which compromise visibility. Research suggests that there are biological differences between right- and left-sided colon cancers. In a recent study, Meza et al. reported that adenoma initiation rates were highest for right-sided lesions, whereas adenoma growth rates were highest for left-sided lesions. This would explain why the incidence of right-sided colon cancer shows the greatest increase after the age of 70 years (10). Histological and molecular differences have also been described, including a higher incidence of the mucinous histotype, microsatellite instability, and defects in mismatch repair genes (11-13). Although the biological characteristics of proximal tumours are associated with a more favourable prognosis than those of distal lesions, proximal lesions are usually diagnosed at more advanced stages (14-15). This explains the discrepancy in survival rates across studies (14-20).
4
María Luque Cabal
• Age The risk of developing CRC increases with age. Diagnosis is rare before the age of 40 years, and incidence begins to increase significantly between the ages of 40 and 50. Around 91% of cases are diagnosed in individuals aged 50 years and above. In the United States, the probability of developing invasive CRC in individuals aged between 40 and 59 years is 1 in 110 for males and 1 in 139 for females. In individuals above the age of 70 years, the incidence increases to 1 in 22 for males and 1 in 24 for females (2).
• Gender Although the incidence of CRC is similar in men and women, some gender-related differences have been reported. Studies have shown that compared to men, female colon cancer patients tend to be significantly older at presentation (21-23), have longer survival in the case of advanced disease (23-24), present more frequently in an emergency setting (23), and are more likely to develop a right-side tumour (14). These findings have implications for screening, diagnosis, and prognosis.
• Geographical distribution The incidence and mortality of CRC vary markedly between countries, as shown in Figures 1 and 2. Center et al. analyzed variations in the rate of CRC across countries from 1953–1957 through to1998–2002 using data from the Cancer Incidence in Five Continents databases (25). The registries with the highest incidence of CRC were located predominantly in Europe, North America, and Oceania, which probably reflects the influence of dietary and lifestyle factors. In these regions, CRC represents 12.6% of all incident cancer cases in men and 14.1% of all incident cancer cases in women (26). From 1983–1987 to 1998–2002, the incidence rates from registries in Western Europe remained stable or showed only a slight increase with the exception of Spain, where the incidence of CRC showed a marked increase. Age-standardised incidence rates of fewer than 25 per 100,000 men 20 years ago had increased to 39.7 per 100,000 men by 2008 (3, 25). This increase may be partly attributable to a change in dietary habits from a “Mediterranean” diet rich in fruits and vegetables to a western diet containing a higher proportion of red meat and saturated fat. In men and women in the United States, the incidence of CRC increased during the period 1975–1985, and then showed a marked decline during
Colorectal cancer epidemiology
5
Figure 1. Age-standardised incidence and mortality rates in men (modified from) Ferlay J, Shin HR, Bray F, Forman D, Mathers C and Parkin DM. GLOBOCAN 2008, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 10 [Internet]. Lyon, France: International Agency for Research on Cancer; 2010. Available from: http://globocan.iarc.fr.
6
MarĂa Luque Cabal
Figure 2. Age-standardised incidence and mortality rates in women (modified from). Ferlay J, Shin HR, Bray F, Forman D, Mathers C and Parkin DM. GLOBOCAN 2008, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 10 [Internet]. Lyon, France: International Agency for Research on Cancer; 2010. Available from: http://globocan.iarc.fr.
Colorectal cancer epidemiology
7
1985–1995. This was followed by a short and non-significant increase during the period 1995–1998. A marked decline was observed during 1998–2006, which was probably due to the initiation of large-scale screening programmes. The most rapid annual rate of decline occurred among men and women aged ≥65 years. By contrast, short-term incidence trends showed an annual increase in individuals aged <50 years (27). The lowest rates were observed in registries from Asia, Africa, and South America. However, from 1983–1987 through to 1998–2001, the incidence of CRC increased in regions undergoing economic transition, such as countries in Eastern Europe, most parts of Asia, and certain countries in South America. These increases were mainly observed in men. Asian registries also recorded high rates in Japan, Singapore, and Israel. The recent increase in the incidence of CRC in these three countries is probably attributable to changes in environment and/or lifestyle (25). Variation in 5-year survival has also been observed between continents. This has been estimated to be 65% in North America, 54% in Western Europe, 34% in Eastern Europe, and 30% in India (1).
• Race The influence of race on the incidence of CRC and survival remains unclear, although studies of migrants have suggested that environmental factors play a major role in disease aetiology (28). From 1999 to 2004, a total of 814,697 cases of invasive CRC were reported in the United States. A recent analysis of these cases revealed that the incidence was higher among black migrants (57.2 per 100,000) compared with white- (50.8 per 100,000) or Asian migrants (38.9 per 100,000). This effect was most pronounced in individuals aged <65 years. In all age groups, localized and regional cancers were diagnosed most frequently in whites and non-Hispanics, and least frequently in blacks. Conversely, distant or unstaged cancers were diagnosed more frequently in blacks (29). It is probable that these discrepancies arise from a complex interplay between genetic and environmental risk factors as well as variability in population-level screening rates and access to care.
2. Non genetic risk factors a. Behavioural risk factors Numerous epidemiological studies have suggested that dietary and lifestyle factors influence the risk of colon cancer. However, their findings should be interpreted with caution, since many of these studies involved
8
María Luque Cabal
small sample sizes and sources of bias. Most of the studies performed to investigate dietary risk factors for colon cancer have used an ecological or case-control design. The cohort method is considered to be the most valid design for an observational epidemiological study. However, its use is less common in this particular field of research due to the relatively low incidence of colon cancer and the requirement for prolonged monitoring. Furthermore, many risk factors, such as obesity, a high fat diet, lack of physical activity, alcohol consumption and smoking, are intercorrelated, and thus caution should be exercised when interpreting the results for possible confounding factors.
• Behavioural factors that may increase the risk of CRC ○
Body mass index (BMI). Ecological studies have reported higher rates of obesity and CRC in industrialised countries. CRC ranks second in terms of both incidence and mortality in industrialised nations, compared with fifth in less developed countries (30). Two recent meta-analyses have reported an association between high BMI and CRC. The first included data from 23 cohort studies and eight case-control studies, which included a total of 70,906 CRC patients (31). Most of the study populations were from western countries. The pooled estimate indicated that individuals with a BMI of ≥ 30 kg/m2 had a 40% greater risk of CRC compared with individuals with a BMI of <25 kg/m2 (Relative risk (RR): 1.40; 95% CI: 0.31–1.51). Evidence was found for both significant heterogeneity across cohort studies and publication bias, and the true estimate of effect for the association between obesity and CRC in this analysis was around 20% (RR: 1.19; 95% CI: 1.11–1.29). A proportion of the observed heterogeneity in estimates across studies was explained by differences in cancer location (higher association with colon cancer than with rectal cancer) and gender (higher association in men than in women). No substantial change in the results was observed following adjustment for potential confounders such as diet and physical activity. The second meta-analysis involved cohort studies only. This analyzed data from 15 studies and a total of 1,058,883 participants, including 6,458 CRC patients (32). The pooled RR for CRC was 1.37 (95% CI: 1.21–1.56) for men who were overweight (BMI = 25–29.9) or obese (BMI ≥ 30), and 1.07 (95% CI: 0.97–1.18) for women. Available studies also suggest that obesity and obesity-related inflammation and metabolic disorders may affect the prognosis of CRC (33-34), although further research is necessary to confirm this hypothesis.
Colorectal cancer epidemiology
9
○
Cigarette smoking. Three recent meta-analyses have investigated the association between cigarette smoking and CRC (35-37). The first included observational studies. The other meta-analyses included prospective studies only, the majority of which were cohort studies. Compared to never-smokers, current smokers had a 7–17% higher risk of developing CRC. However, this association only reached significance in the study by Huxley et al. The RR for CRC mortality was significantly higher in ever-smokers compared to never-smokers in the studies of Botteri (RR: 1.28; 95% CI: 1.15–1.42) and Liang (RR: 1.40; 95% CI: 1.06–1.84). In both studies, comparisons between former-smokers and never-smokers revealed a significantly higher incidence of CRC (17–25%) and mortality (15–23%) in former smokers. In general, smokers showed a high tendency to an increased risk of rectal rather than colon cancer. The risk of CRC was also found to increase with the number of cigarettes smoked daily and the duration of smoking.
○
Diabetes mellitus. Most epidemiological studies investigating the relationship between diabetes and the risk of CRC have reported a positive association. Furthermore, insulin resistance and type 2 diabetes are known to be correlated with calorie-high diet, physical inactivity, and obesity, all of which are known risk factors for CRC. A meta-analysis published in 2005 included six case-control studies and nine cohort studies (38). This found that individuals with diabetes had a significant 30% increase in risk (RR: 1.30; 95% CI: 1.20–1.40). No heterogeneity across studies was found. The subgroup meta-analysis revealed similar results across studies for women and in men, and that the estimated risk was similar for colon cancer and rectal cancer. An analysis restricted to studies that had controlled for obesity and physical activity revealed a positive association between diabetes and CRC (summary RR: 1.34; 95% CI: 1.20–1.49), with no statistically significant heterogeneity across studies. A second meta-analysis of 15 cohort studies also suggested that the risk of CRC was 23% higher in individuals with diabetes compared with unaffected individuals (RR: 1.23; 95% CI: 1.17–1.30) (35).
○
Western diet. In 1971, Burkitt observed that the diet and stools of native South Africans differed from those of westerners, and that these two groups showed a marked difference in the incidence of colon cancer. He proposed that the high fibre content of the native diet had a protective effect (39). A formal analysis of firstepidemiological data suggested a positive association between the risk of colon cancer and dietary fat and meat, and a negative association with dietary fibre (40-43).
10
María Luque Cabal
However, more recent studies and one meta-analysis have indicated that dietary fat per se is not associated with an increased risk of CRC (44-45). In a large cohort study of nearly 149,000 men and women, participants were asked to complete a questionnaire concerning their eating habits at two time-points, i.e., 1982 and 1992/1993 (46). A high intake of red and processed meat in 1992/1993 was associated with a higher risk of colon cancer after adjusting for age and energy intake. However, this association was not found after adjustment for BMI, cigarette smoking, and other covariates. In terms of long-term consumption, individuals who had reported the highest intake of processed meat (1 oz of processed meat 5 or 6 days per week for men, and 2 or 3 days per week for women) in both 1982 and 1992/1993 had a higher risk of distal colon cancer (RR: 1.50; 95% CI: 1.04–2.17) compared to individuals who were in the lowest tertile of consumption at both time-points. This was also the case for the ratio of red meat to poultry and fish (RR: 1.53; 95% CI: 1.08–2.18). The most recently published meta-analysis included 26 cohort studies and data from more than 15,000 CRC cases (35). The pooled estimate for the highest vs. the lowest level of consumption was RR 1.21 (95% CI: 1.13–1.29) for red meat intake, and RR 1.19 (95% CI: 1.12–1.27) for processed meat intake. No evidence of heterogeneity across studies was found (P=0.72), and there was no significant difference in the estimates for colon cancer and rectal cancer. These results are similar to those of a meta-analysis published in 2006. Interestingly, the authors of the 2006 meta-analysis found no change in their results after adjusting for potentially confounding factors such as physical activity, BMI, smoking, alcohol intake, total energy, and calcium (47). ○
Alcohol consumption. Alcohol consumption is one of the most important known causes of human cancer, and several studies have suggested that increased alcohol intake is a risk factor for CRC. The most recently published meta-analysis included data from 21 cohort studies (35). The authors estimated that the risk of developing a malignant tumour in the colorectum was 56% greater in individuals who were categorized as “heavy drinkers” (RR: 1.56; 95% CI: 1.42–1.70), although the criterion for inclusion in this category (i.e., the level of alcohol intake) was not specified. No evidence of heterogeneity across studies was found. The effect size estimates for colon and rectal cancer were similar, and both were statistically significant. A previous meta-analysis of data from 16 cohort studies reported a 15% increase in the risk of CRC for every 100 g increase in alcohol intake per week. This association was found for both colon and rectal cancer (48).
Colorectal cancer epidemiology
11
• Behavioural factors that may reduce the risk of CRC ○
Intake of dietary fibre, fruit, and vegetables. Although early observations suggested that a diet rich in fibre exerted a protective effect, subsequent studies have yielded conflicting results. A large cohort study of 76,947 women and 47,279 men (49) and a meta-analysis of 13 prospective cohort studies (50) found only a minimal association between high fibre intake and a lower risk of CRC. However, this result was not statistically significant after adjustment for potential confounding variables. The meta-analysis found no association between CRC risk and the consumption of fruit or vegetables. In 2000, the results of a randomised trial performed to determine whether dietary supplementation with wheat-bran fibre reduces the rate of recurrence of colorectal adenomas were published (51). This had involved 1,429 adenoma patients who had undergone the removal of one or more histologically confirmed colorectal adenomas during the 3-month period before recruitment. The patients had been randomised to receive either a high- or low-fibre cereal supplement for a period of 6 weeks. Although compliance with the study protocol was high, no significant differences in recurrence rates were observed between the experimental and placebo subjects, even after adjustment for known risk factors such as randomisation period, sex, smoking status, alcohol consumption, and energy intake. Possible causes of this negative finding include an imbalance in other risk factors, the short intervention period, or the possibility that fibre may only protect against malignant change in the case of large adenomas.
○
Physical activity. A sedentary life style is associated with a higher risk of obesity, which is a known risk factor for CRC. It may also promote the development of cancer through its effects on immune function and on hormone and growth factor levels. Two meta-analyses of epidemiological studies have been published to date. The first included case-control studies and cohort studies (52), whereas the most recent meta-analysis included 27 cohort studies (35). Both meta-analyses found that the estimated reduction in colon cancer risk secondary to physical activity was around 20%. The protective effect conferred by physical activity appeared to be stronger for men than women, and for colon cancer rather than rectal cancer, and recreational physical activity rather than occupational physical activity.
○
Calcium and vitamin D. Most epidemiological studies performed to date have suggested that increased calcium consumption (either dietary
12
María Luque Cabal
or as a supplement) is associated with a slight decrease in CRC risk. Cho et al. performed a meta-analysis of 10 cohort studies. The pooled multivariate RR for CRC was 0.85 (95% CI: 0.78–0.94) for individuals who consumed 250 g/day of milk or more compared with participants who consumed less than 70 g/day. Every 500-g/day increase in milk consumption was associated with a 12% reduction in the risk of CRC (53). This association was found in both sexes, and was restricted to cancers of the distal colon and rectum. Grau et al. performed a large multicentre trial to investigate the effect of calcium supplementation on the recurrence of large bowel adenomas in patients with a recently diagnosed colorectal adenoma. A total of 930 subjects were randomised to either placebo or calcium (3 g of calcium carbonate, or 1200 mg of elemental calcium) (54-55). Among subjects assigned to calcium supplementation, a statistically significant reduction in the risk of one or more recurrent adenomas was only observed in those with a baseline 25-(OH) vitamin D level that was above the median (RR: 0.71; 95% CI: 0.57–0.89; P for interaction =0.012). Calcium supplementation had no effect in subjects whose baseline 25-(OH) vitamin D levels were at or below the overall median (29.1 ng/mL). The calcium risk ratio was higher for histologically advanced neoplasms (0.65; 95% CI: 0.46– 0.93) than for hyperplastic polyps (0.82; 95% CI: 0.67–1.00) or tubular adenomas (0.89; 95% CI: 0.77–1.03). The preventive effect was most pronounced in individuals with a high dietary intake of calcium and fibre and a low intake of fat. However, these interactions were not statistically significant. A meta-analysis of 17 epidemiological studies showed an inverse association between the level of circulating 25-(OH) vitamin D and the risk of developing colorectal adenomas (Odds Ratio (OR): 0.7; 95% CI: 0.56– 0.87) (56).
b. Inflammatory bowel disease In 1925, Crohn and Rosenberg were the first to describe a case of CRC in association with inflammatory bowel disease (IBD) (57). Many subsequent studies have produced evidence in support of an association between IBD and CRC risk. Early studies may have overestimated the risk associated with IBD, since their results were frequently based upon data from hospitalised patients. Although estimates of cancer risk in patients with ulcerative colitis (UC) differ between studies, almost all studies have demonstrated a significant association between UC and CRC. The standardised incidence ratio (SIR) ranges from 2 to 30. The cumulative incidence 25–35 years following the
Colorectal cancer epidemiology
13
assignment of a diagnosis of UC ranges from 8% to 43% (58-71). These wide differences are probably attributable to a range of factors, such as variation in dietary patterns, genetic factors, use of colonoscopy surveillance, the frequency of prophylactic colectomy, and treatment with aminosalicylates, which appear to have a protective effect. In 2001, Eaden et al. published a meta-analysis of 116 studies which had investigated the risk of CRC in patients with UC. These studies included a total of 54,478 patients, of whom 1,698 had a diagnosis of CRC (72). The overall estimate of the prevalence of CRC was 3.7% (95% CI: 3.2–4.2) for any UC patient, and 5.4% (95% CI: 4.4–6.5) for patients with pancolitis. The incidence was higher than that in the general population. The CRC incidence rate was 3 per 1000 person years duration (pyd) (95% CI: 2/1000–4/1000) for any UC patient, and 4 per 1000 pyd (95% CI: 3/1000–6/1000) for patients with pancolitis. The CRC incidence rate in the general population was estimated to be 0.6 per 1000 pyd. No publication bias was found. For the first decade of UC, the overall incidence rate was 2 per1000 pyd (95% CI: 1/1000–2/1000). For the second and third decades, the estimated overall incidence rates were 7 per 1000 pyd (95% CI: 4/1000– 12/1000) and 12 per 1000 pyd (95% CI: 7/1000–19/1000), respectively. Thus, the risk of CRC in UC was estimated to be 2% at 10 years, 8% at 20 years, and 18% at 30 years. These rates were very similar for patients with pancolitis. However, the CIs were wide, which is probably attributable to the small number of studies in the pancolitis group (n=6). In adult UC patients, the age-at-onset had no statistically significant bearing on cancer risk. A study of 723 UC patients identified longer disease duration, extensive colitis, primary sclerosing cholangitis (PSC), and the presence of dysplasia in the biopsy specimen as risk factors for CRC (73). Several studies have reported that PSC is a risk factor for CRC in patients with UC. However, these studies generally involved small sample sizes (74-75). The cumulative risk for CRC has been reported to be 14–16% after 10 years of IBD (74,76) and 31% after 20 years (76). The underlying biological mechanism remains unknown, although several hypotheses have been proposed. One such hypothesis is that the abnormal bile composition in PSC exerts a carcinogenic effect (77-78). An alternative hypothesis is that since UC and PSC are often characterised by low disease activity (79); these patients tend to undergo colectomy and sulphasalazine therapy less frequently than other IBD patients. The results of several epidemiological studies suggest that patients with Crohn’s disease (CD) have a 1.5–20-fold increase in CRC risk (80-83). In a meta-analysis of 12 studies, the overall pooled estimate for the colon cancer RR was 2.5 (95% CI: 1.3–4.7) (84). After selecting studies performed
14
María Luque Cabal
in patients with colonic or ileal disease, the only statistically significant association was between colonic disease and CRC (RR: 4.5; 95% CI: 1.3– 14.9), with no association being found with ileal disease (RR: 1.1; 95% CI: 0.8–1.5). The cumulative risk of CRC in CD patients was 2.9% (1.5–5.3%) at 10 years; 5.6% (3.1–10.4%) at 20 years; and 8.3% (4.5–15.1%) at 30 years. Patients with IBD tend to develop CRC 15–20 years earlier than patients with sporadic cancers. In a study of 80 IBD cases, the median age at CRC diagnosis was 54.5 years (range, 32–76 years) in patients with CD, and 43 years (range, 17–75 years) in patients with UC. During the same period, the median age at diagnosis of sporadic CRC in 5,266 patients with no history of IBD (P<0.001) was 65 years (range, 7–99) (P<0.001) (85). Patients with UC and CD tend to develop CRC in differing anatomical locations. Patients with UC develop tumours predominantly in the rectosigmoid area, whereas patients with CD frequently develop tumours in the right rectosigmoid colon. Patients with CD also have a higher risk of small bowel carcinoma compared with UC patients (84-85). In a cohort of 920 IBD patients who were followed up for a median period of 14.8 years (86), the occurrence of CRC was not associated with an increase in mortality.
c. Colonic polyps Adenomatous polyps have malignant potential and therefore require clinical intervention (87-88). They are classified histologically as: 1) tubular, 2) tubulovillous, or 3) villous. The reported frequency of each histological subtype varies across studies. In two large case series, tubular adenomas were the most frequent subtype (65–75%), followed by the tubulovillous (20–25%), and villous (5–9%) subtypes (89-90). The risk of invasive malignancy was also found to differ according to histological subtype. Malignant potential was reported to be rare in tubular adenomas (2–3%), but higher in tubulovillous (6–8%) and villous adenomas (10–18%). Cancer risk also increased with polyp size, reaching 6.5–17% for adenomas that were larger than 2 cm. The most frequent anatomical location was the sigmoid colon, followed by the descending colon (89-90). In 1992, Atkin et al. investigated a cohort of 1,618 rectosigmoid adenoma patients to assess the long-term risk of CRC following endoscopic polypectomy (91). Each patient was followed up for a mean period of 13.8 years. The incidence of CRC was found to be significantly increased in patients with a history of a rectosigmoid tubulovillous or villous adenoma, or an adenoma that was larger than 1 cm (SIR: 3.6; 95% CI: 2.4–5.0). In patients with multiple adenomas, the SIR reached 6.6 (95% CI: 3.3–11.8).
Colorectal cancer epidemiology
15
3. Inherited risk factors The aetiology of CRC involves both genetic and environmental factors. Although some specific genetic disorders are associated with a particularly high risk of CRC, they account for less than 5% of all cases. The majority of cases have a multifactorial aetiology. Approximately 20% to 25% of cases occur in patients with a positive family history of CRC, and the remaining cases are sporadic.
â&#x20AC;˘ Familial adenomatous polyposis Familial adenomatous polyposis (FAP) is an autosomal dominant disease that is characterized by the presence of hundreds of adenomatous polyps in the colon and rectum. If left untreated, this disorder can progress to CRC. FAP accounts for 1% of all CRC cases, and affects around 1 per 10,000 individuals. The penetrance of the disease approaches 100% by the age of 40 years (92). Attenuated FAP (AFAP) is a less aggressive variant of FAP. It is characterised by the presence of fewer colorectal adenomatous polyps (usually 10 to 100), a later age-at-onset, and a lower cancer risk. Certain types of lesions (skull and mandible osteomas, dental abnormalities, and fibromas on the scalp, shoulders, arms, and back) are indicative of the Gardner variant of FAP. Patients with Turcot syndrome present with FAP and a medulloblastoma of the brain (93). In 1991, the gene responsible for the vast majority of FAP cases, the adenomatous polyposis coli (APC) gene, was identified. In 5â&#x20AC;&#x201C;30% of FAP patients, no APC mutation is identifiable through currently available genetic tests. In 2003, Varesco et al. showed that 'APC-negative' FAP patients may carry biallelic mutations in the MYH gene (94). Until the 1950s, almost all untreated FAP patients died from CRC between the ages of 40 and 50 years. FAP also involves a variety of extracolonic manifestations. These include osteomas, epidermoid cysts, desmoid tumours, upper gastrointestinal polyps, congenital hypertrophy of the retinal pigment epithelium, and other malignancies (duodenal carcinoma, thyroid carcinoma, and hepatoblastoma). The introduction of screening programmes has reduced the prevalence of CRC and improved survival in the majority of FAP patients. It has also led to a substantial change in the pattern of mortality. At the time of writing, the most frequent causes of death in patients detected through screening programmes are duodenal cancer and desmoid tumours (95-98).
16
María Luque Cabal
• Hereditary nonpolyposis colorectal cancer Hereditary nonpolyposis colorectal cancer (HNPCC) is also known as Lynch syndrome. This disorder has an autosomal dominant mode of inheritance, and is associated with an increased risk of CRC and other neoplasias, including cancers of the endometrium, stomach, ovary, urinary tract, hepatobiliary tract, pancreas, and small bowel (99). HNPCC accounts for around 2% to 3% of all CRC cases (100) and is caused by germline mutations in DNA mismatch repair (MMR) genes, in particular MLH1, MSH2, and MSH6 (101). A study in Finland compared cancer incidence in the general population with that of a cohort of 1,763 members of 50 HNPCC families. In all cases, the diagnosis had been established through genetic testing. In mutation carriers, the authors reported an SIR for CRC of 68 (95% CI: 56–81) (102). The SIR for CRC was higher in men (SIR=83) than in women (SIR=48). The male-to-female ratio was 1.7 (95% CI: 1.2–2.7). An increased SIR was also observed for cancers of the endometrium (62; 95% CI: 44–86), ovaries (13; 95% CI: 5.3–25), biliary-tract (9.1; 95% CI: 1.1–33), uro-epithelium (7.6; 95% CI: 2.5–18), stomach (6.9; 95% CI: 3.6–12), and kidney (renal-cell adenocarcinoma) (4.7; 95% CI: 1–14), as well as for tumours of the central nervous system (4.5; 95% CI: 1.2–12). In mutation carriers, the cumulative incidence rates for CRC at 70 years were 100% in men and 54% in women. Lynch syndrome is characterised by an early age-at-onset and a predominance of right-sided colonic tumours (103). In a cohort of mutation carriers who were participating in a surveillance programme, the mean age at initial cancer diagnosis was 40 years (104). Most first lesions arise proximal to the splenic flexure, and are usually synchronous (two or more distinct tumours separated by normal bowel) or metachronous (new nonanastomotic tumours which develop at least 6 months after the initial diagnosis) (105).
• Family history of CRC In a survey of around 36,000 households in the United States, approximately 5% of respondents reported having one or more first-degree relatives with CRC (106). The number of affected family members, their degree of kinship, and the age-at-onset in relatives are all factors which influence the risk of CRC. Two meta-analyses estimated familial CRC risk in first-degree relatives of CRC- and colorectal adenoma patients (107-108). Both of these studies generated a pooled estimate of RR in the presence of a first degree relative with CRC of around 2.25. This increased to 3.87 in cases in which the relative had been diagnosed with CRC before the age of 45
Colorectal cancer epidemiology
17
years. The RR increased to 3.97â&#x20AC;&#x201C;4.25 in cases with at least two affected relatives. The CRC risk was found to be greater for the relatives of patients with colon cancer compared to the relatives of patients with rectal cancer.
4. Prevention In 1990, Fearon and Vogelstein described the molecular basis of colorectal neoplasia as a multistep, multipathway and multifocal process that ultimately requires cumulative damage to several genes within and across cellular generations (109). Research into the pathogenesis of CRC has shown that the progression from normal epithelium to adenoma and carcinoma may occur over a period of decades. It is generally accepted that most CRC evolve from adenomatous polyps, and that the removal of such polyps can prevent the transition to CRC. In the National Polyp Study, a cohort of 1,418 patients who had undergone complete colonoscopy involving the removal of a minimum of one adenomatous polyp were followed up with periodic colonoscopy for an average of 5.9 years (110). The incidence of colon cancer was 88% to 90% lower than in patients from other studies who had not undergone polyp excision, and 76% lower than in the general population. Two main strategies to prevent CRC have been described. The first is the introduction of screening programmes for the detection and removal of precancerous adenomas. The second is chemoprevention, which involves the use of dietary substances or medications to avoid either the development of adenomas or the transition to adenocarcinoma.
a. Chemoprevention A simple first step towards the prevention of CRC is the adoption of a series of healthy dietary and lifestyle habits. Epidemiological studies have identified several dietary and lifestyle factors that confer risk or protective effects in CRC, as described in the section â&#x20AC;&#x153;non-inherited risk factorsâ&#x20AC;?. The World Health Organisation recommends the avoidance of alcohol and tobacco, fatty foods and the excessive consumption of red meat, and advocates foods rich in calcium or fibre, fresh vegetables and fruit, and daily physical exercise as being beneficial (111). A number of medications have been shown to confer a potential preventive effect. However, since studies have shown that simple biannual faecal occult blood testing achieved a 21% reduction in mortality at 18-year follow-up, and that one-off screening with sigmoidoscopy, led to a decrease in mortality of 31% after 10 years compared to no screening, the risk-benefit ratio of the use of a chemopreventive drug needs to be carefully evaluated
18
María Luque Cabal
(112-113). Furthermore, as with screening programmes, the long-term prescription of a chemopreventive agent to otherwise healthy individuals may be associated with poor compliance. Finally, the cost-effectiveness of this strategy has not yet been demonstrated.
• Non-steroidal anti-inflammatory drugs Non-steroidal anti-inflammatory drugs (NSAIDs) have been shown to suppress malignant transformation and tumour growth experimentally, principally by inhibiting cyclooxygenase-2 (COX-2)-mediated prostaglandin E (PGE)-2 synthesis in neoplastic tissues. Several hypotheses concerning the contribution of COX-2 to tumourigenesis have been proposed. These include the inhibition of apoptosis, an increase in angiogenesis or invasiveness, modulation of inflammation/immune suppression, upregulation of PGE, and conversion of procarcinogens to carcinogens (114-116). Dubé et al. carried out a systematic review of studies assessing the effect of chemoprophylaxis on the incidence of adenoma and CRC and on CRC mortality in populations without genetic risk factors (117). The authors identified significant heterogeneity across studies in terms of dosage as well as the duration and frequency of administration which necessitated careful consideration in the formation of subgroups for the analyses. An analysis of randomised clinical trials showed that regular ingestion of aspirin reduced the incidence of new colonic adenomas in patients with a previous history of colonic adenomas (RR: 0.82; 95% CI: 0.7–0.95), but not in the average-risk population. The analysis of cohort studies showed that regular ingestion of aspirin was associated with a reduction of 22% in the RR for CCR incidence. However, two randomised clinical trials of low-dose aspirin failed to show any protective effect. Only one study assessed mortality rates, and this identified no difference between patients assigned to low-dose aspirin and those assigned to placebo. The benefits of chemoprevention were most pronounced when high dose aspirin was used for a period of more than 10 years. However, aspirin use was associated with a dose-related increase in the incidence of gastrointestinal complications. The same group reported similar findings from a systematic review of case-control and cohort studies of non-aspirin-NSAIDs chemoprophylaxis (118). One cohort study showed that non-aspirin-NSAIDs significantly reduced the incidence of colorectal adenomas in patients with a previous history of adenomas (RR: 0.64; CI: 0.48–0.85). Several case control studies showed that NSAIDs reduced the incidence of colorectal adenomas in average-risk populations (RR: 0.54; CI: 0.4–0.74). This analysis also showed that non-aspirin NSAIDs had a statistically significant dose-dependent and
Colorectal cancer epidemiology
19
duration-dependent effect on CRC risk. A risk reduction of up to 30% was found for colon cancer, whereas no benefit was observed for rectal cancer alone. A single study investigated the effect of ibuprofen on CRC mortality. This demonstrated no protective effect for ibuprofen, but rather a statistically significant increase in all-cause mortality (119). Although no statistically significant increase in all-cause mortality was observed for NSAID use, these drugs were associated with a greatest risk of peptic ulceration and gastrointestinal haemorrhage, as well as cardiovascular events in patients at high risk of cardiovascular disease. In patients with a history of colorectal adenomas, COX-2 inhibitors (mainly celecoxib) were found to reduce the incidence of all adenomas and advanced adenomas over a 3-year period of follow-up (pooled RR: 0.72; 95% CI: 0.68–0.77 vs. RR: 0.56; CI: 0.42–0.75). Although COX- 2 inhibitors were associated with fewer gastrointestinal adverse effects than non-aspirin NSAIDs, data from the APPROVe study demonstrated that they were associated with an increased risk of peptic ulceration compared with placebo (120). This trial also demonstrated an association between the use of rofecoxib vs. placebo and an increase in the risk of cardiovascular events (16 per 1000 events). This finding led to the withdrawal of rofecoxib from the market. A subsequent polyp prevention study (Adenoma Prevention with Celecoxib [APC]) also found that celecoxib was associated with an increased risk of cardiovascular events (risk relative to placebo: 1.6; 95% CI: 1.0–2.5), in particular for those patients with pre-existing atherosclerotic heart disease (121). It may therefore be concluded that the risk-benefit ratio for chemoprevention with non-aspirin NSAIDs or COX-2 inhibitors is non-favourable in average-risk individuals and patients with a history of colorectal adenomas. In 2000, Steinbach et al. conducted a randomised clinical trial in patients with FAP to investigate the effect of two doses of celecoxib on colorectal polyps (122). After 6 months, the authors observed a 28% reduction in the mean number of colorectal polyps (P=0.003 for the comparison with placebo), and a 30.7% reduction in the polyp burden (the sum of polyp diameters) (P=0.001) in patients receiving the higher dose of celecoxib (400 mg twice daily). In the placebo group, respective reductions of 4.5% and 4.9% were observed. The lower dose of 100 mg twice daily was associated with a less pronounced effect. The authors reported an 11.9% reduction in the mean number of colorectal polyps (P=0.33 for the comparison with placebo) and a 14.6% reduction in the polyp burden (P=0.09). In 2002, a randomised controlled trial of patients with FAP reported a significant reduction in the number of areas affected by duodenal polyps following 6 months treatment with 400 mg of celecoxib administered twice daily (123).
20
María Luque Cabal
• 5-aminosalicylic acid CRC is one of the most serious complications of IBD. 5-aminosalicylic acid (5-ASA) is an anti-inflammatory agent which is used in the treatment of mild to moderate active UC. It is usually well tolerated and is associated with minimal systemic adverse effects and gastrointestinal toxicity (124). Epidemiological studies performed in the early 1990s suggested that the chronic consumption of aminosalicylates, such as sulphasalazine, may also confer a degree of protection against CRC in patients with UC, and that this effect may be mediated, at least in part, via a similar mechanism to that of NSAIDs (125). A meta-analysis of nine studies (three cohort studies and six case-control studies) involving a total of 1,932 patients reported an association between 5-ASA use and a significant reduction in CRC risk (OR: 0.51; 95% CI: 0.37– 0.69), as well as a protective effect against a combined endpoint of CRC and dysplasia (OR: 0.51; 95% CI: 0.38–0.69) (126). Two studies analyzing the duration and frequency of 5-ASA use concluded that there was a doseresponse effect, and that regular use was associated with a lower risk of CRC (OR: 0.28; 95% CI: 0.10–0.81).
• Ursodeoxycholic acid Ursodeoxycholic acid (UDCA) is a well-tolerated, low-risk agent that is used in the clinical management of primary biliary cirrhosis (PBC) and PSC. Tung et al. performed a cross-sectional survey of 59 patients with co-morbid UC and PBC who were undergoing colonoscopic surveillance (127). A significant association was found between UDCA use and a decrease in the risk of colonic dysplasia (OR: 0.18; 95% CI: 0.05–0.61; P=0.005). This association was also found after adjustment for the use of sulphasalazine or 5-ASA. Pardi et al. followed up a cohort of patients with comorbid UC and PBC who had participated in a randomised, placebo-controlled trial of UDCA therapy assessing the effect of UDCA on the development of colorectal dysplasia and CRC compared to that of a placebo (128). UDCA therapy was found to be associated with a reduction in the risk of both colorectal dysplasia and CRC (RR: 0.26; 95% CI: 0.06–0.92; P=0.034). Although many patients assigned to the placebo group eventually received open-label UDCA, this did not significantly alter the results.
•Folic acid The results of early epidemiological studies suggested that there is an inverse association between folate levels and the risk of CRC. Although more
Colorectal cancer epidemiology
21
than 20 case control studies have been performed, the findings have been inconsistent. A meta-analysis of cohort studies reported an association between reduced CRC risk and dietary folate (i.e., folate from food alone; RR for high vs. low intake: 0.75; 95% CI: 0.64–0.89) but not with total folate (i.e., folate from both food and supplements; RR for high vs. low intake: 0.95; 95% CI: 0.81–1.11) (130). The authors found no significant heterogeneity between studies. However, recent research in patients with a history of colorectal polyps has suggested that folate supplements may induce a transition to malignancy in undetected precursor lesions. This hypothesis is consistent with the role of folate in nucleotide synthesis and cell proliferation (130). It is therefore possible that folate acts as a ‘dual-modulator’ in colorectal carcinogenesis. According to this hypothesis, moderate dietary increments initiated before the establishment of neoplastic foci may exert a protective effect, whereas excessive folate intake in individuals with early undetected lesions may increase the risk of tumourigenesis (131).
b. Screening for colorectal cancer The risk of CRC is higher in individuals with a history of adenomatous polyps, IBD, or inherited genetic disorders. Given that most CRCs arise from adenomatous polyps, and that the progression to carcinoma may take several years, the early detection and removal of adenomas is probably the most effective method of reducing the incidence and mortality of CRC, as suggested by the results of several studies (132-134). Available screening tests:
• Stool-based tests Screening for the presence of occult blood in the stool exploits the fact that most cancers, and some adenomatous polyps, tend to bleed. Faecal occult blood testing (FOBT) detects faecal hemoglobin, and employs a variety of techniques including immunochemical or DNA analysis and guaiac-testing. Stool DNA testing is expensive, and its suitability as a screening strategy has not yet been assessed within the context of large-scale prospective trials. Several factors may affect the accuracy of FOBT. These include stool rehydration (increases sensitivity; decreases specificity and positive predictive value), heme degradation, medications, and dietary substances such as peroxidases and heme from meat. Current recommendations state that testing should be conducted on two or three samples taken from stool specimens on consecutive days, since multiple consecutive sampling
22
María Luque Cabal
increases the likelihood of detecting blood. Following a positive test, diagnostic endoscopy must be performed to determine the source of the occult blood (135). Three large prospective randomised trials have been performed in average-risk populations to assess the effect on mortality of FOBT-based screening programmes. The programmes examined in these studies were conducted on an annual or biennial basis, and all subjects with positive results were examined by colonoscopy (136-139). A systematic review of these three studies and data from a Swedish trial estimated an overall significant reduction in CRC mortality of 16% (RR: 0.84; 95% CI: 0.77– 0.93) (140). This increased to 23% when the estimated RR was adjusted for attendance for screening in individual studies. A shift to the detection of earlier stage disease, which is associated with superior clinical outcome, was observed in all four studies.
• Sigmoidoscopy Sigmoidoscopy is a procedure which involves direct endoscopic examination of the distal part of the colon. Diagnostic biopsies can be obtained during its performance. If a neoplastic lesion is found during sigmoidoscopy, the entire large bowel should be evaluated by colonoscopy. Nonprospective, case-control studies have indicated that screening sigmoidoscopy can reduce the incidence of distal CRC by 50% to 66% (141-143), and reduce overall CRC mortality by up to 79% (142). A prospective controlled trial in Norway offered sigmoidoscopic screening to 400 randomly selected individuals from the general population. A control group of 399 individuals from the same population did not undergo screening (144). In cases in which polyps were discovered, a colonoscopy was performed and during this procedure, all polyps were removed. Colonoscopy was repeated 2 and then 6 years later. After 13 years of follow up, the authors found a significant reduction in the incidence of CRC (risk ratio: 0.2; 95% CI: 0.03–0.96; P=0.02). However, the study found a higher overall mortality among patients who had undergone screening (risk ratio: 1.57; 95% CI: 1.03–2.4; P=0.03). Further studies are warranted to investigate this finding.
• Colonoscopy Colonoscopy is regarded as the gold-standard for the diagnosis of CRC since it allows direct examination of the full length of the large bowel as well
Colorectal cancer epidemiology
23
as the excision of polyps and biopsy sampling. A recent systematic review found that the ability to detect adenomas was directly correlated with adenoma size. The sensitivity of a single colonoscopy was 97.8% for adenomas ≥ 10 mm. This was reduced to 87% and 74% for adenomas sized 5–10 mm and 1–5 mm, respectively (145). No published prospective study has demonstrated a direct reduction in CRC mortality secondary to primary screening colonoscopy. However, much of the demonstrated benefit of other screening techniques such as FOBT or sigmoidoscopy is attributable to the use of colonoscopy in cases with abnormal results. This is also illustrated by the reduction in mortality following colonoscopy and polipectomy that was observed in the study by the National Polyp Study Workgroup (132). A retrospective study of 1,177 average-risk and asymptomatic individuals who had undergone colonoscopy in a medical centre in Israel reported prevalence rates of 20.9%, 6.3%, and 1.1%, for colorectal neoplasia, advanced neoplasia, and cancer, respectively (146). Detection of proximal neoplasia was particularly frequent among patients aged 65 to 75 years. In this age-group, the prevalence of proximal neoplasia in the absence of distal lesions reached 60%. This finding suggests that colonoscopy could be of particular benefit in this population. By contrast, a recent Canadian case-control study reported an association between colonoscopy and fewer deaths from left-sided CRC (adjusted conditional OR: 0.33; CI: 0.28–0.39). This association was not found for right-sided CRC (adjusted conditional OR: 0.99; CI: 0.86–1.14) (147).
• “Virtual Colonoscopy” (computed tomographic colonography) Computed tomographic colonography (CTC) is a minimally invasive procedure in which two-dimensional (2D) or three-dimensional (3D) images of the entire colon are generated from a spiral computed tomographic scan. If a suspected colorectal lesion is detected, a conventional colonoscopy may be performed. Kim et al. compared patients who had undergone screening with CTC or conventional colonoscopy (148). Each group was comprised of more than 3,000 patients. Similar detection rates for advanced neoplasia were found (3.2% vs. 3.4%, respectively). However, polypectomy and complications were less common in the CTC group. The ACRIN (American College of Radiology Imaging Network) National CT Colonography Trial involved a total of 2,600 participants, all of whom underwent CTC followed by conventional colonoscopy (149). The physician who performed the colonoscopy was blind to the CTC results. For
24
MarĂa Luque Cabal
Table 1. Recommendations for screening in average-risk populations.
Colorectal cancer epidemiology
25
Table 2. Recommendations for screening in high-risk populations.
large adenomas (>10 mm) and cancers, the mean (±SE) per patient estimates of the sensitivity, specificity, and positive and negative predictive values were 0.90±0.03, 0.86±0.02, 0.23±0.02, and 0.99±<0.01, respectively. This technique may be limited to the identification of flat and depressed adenomas and polyps that are smaller than 10 mm in diameter, and two
26
MarĂa Luque Cabal
studies have estimated that CTC is less cost-effective than standard colonoscopy (150-151). These methods must be organised into a screening programme that targets appropriately aged individuals and identifies those with an increased likelihood of disease. The interval between repeat screening should be clearly defined for individuals with negative initial findings. When an abnormal result is obtained, additional diagnostic testing including biopsy or excision should be performed as appropriate. Furthermore, a surveillance programme should be in place for patients who have undergone the removal of a polyp or cancer. Screening guidelines have been established for average-risk populations (asymptomatic and without a personal or family history of adenomatous polyps or other illness, e.g., IBD, FAP, HNPCC, that predispose to CRC) and individuals who are at high risk. Recommendations for CRC screening need to take multiple factors into account. These include effectiveness, sensitivity, false-positive rate, safety, cost, cost-effectiveness, and patient preference. The guidelines of the World Gastroenterology Organisation (WGO) propose different approaches depending upon local resources, cultural preferences, and national health policies (152). The guidelines of the American Cancer Society and the United States Multi-Society Task Force on Colorectal Cancer (ACS-MSTF) distinguish between tests that can detect cancers at an early and treatable stage (e.g., stool tests), and tests that can also detect adenomas and thus lead to cancer prevention (e.g., sigmoidoscopy, colonoscopy, CTC) (153). The American College of Gastroenterology advocates colonoscopy as the preferred screening/prevention test and faecal immunochemical testing as the preferred method of screening/detection in patients who decline cancer prevention tests (154). A summary of the recommendations of each set of guidelines for average- and high-risk populations is provided in Tables 1 and 2 respectively.
Conclusion CRC is a common and lethal disease. The risk of developing CRC is influenced by both environmental and genetic factors and the most relevant risk factor is older age. The transition from identification of theoretically avoidable causes of this disease to implementation of preventive strategies depends on the delineation of exposures considered to be causally associated with development of CRC.
Colorectal cancer epidemiology
27
There are several factors considered to be associated with the development of CRC. In this way, the risk is clearly increased by a Western diet but also there are genes responsible for the most common forms of inherited CRC which have also been identified. With this background it seems appropriate to changer dietary habits, to practice regular physical activity and maintain a healthy weight, together with targeted screening programs and early therapeutic intervention. All these measures could, in time, substantially reduce the morbidity and mortality associated with CRC.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
Parkin, DM, Bray, F, Ferlay J and Pisani, P. 2005, CA Cancer J Clin., 55, 74. American Cancer Society. 2010, Atlanta, GA, American Cancer Society. Ferlay, J, Shin, HR, Bray, F, Forman, D, Mathers, C and Parkin, DM. 2010, Available from: http://globocan.iarc.fr Gunderson, LL, Jessup, JM, Sargent, DJ, Greene, FL, and. Stewart, AK. 2009, J Clin Oncol., 28, 264. Chang, GJ, Hu, C, Eng, C, Skibber, JM, and Rodriguez-Bigas, MA. 2009, J Clin Oncol., 27, 5938. Rabeneck, L, Davila, JA, and El-Serag, HB. 2003, Am J Gastroenterol., 98, 1400. Wu, X, Chen, VW, Martin, J, Roffers S, Groves, FD, Correa, CN, et al. 2004, Cancer Epidemiol Biomarkers Prev, 13, 1215. Cress, RD, Morris, C, Ellison, GL, and Goodman, MT. 2006, Cancer, 107, 1142. Baxter, NN, Goldwasser, MA, Paszat, LF, Saskin, R, Urbach, DR and Rabeneck, L. 2009, Ann Intern Med., 150, 1. Meza, R, Jeon, J, Renehan, AG, and Luebeck, EG. 2010, Cancer Res, 70(13), 5419. Di Pietro, M, Sabates Bellver, J, Menigatti, M, Bannwart, F, Schnider, A, Russell, A, et al. 2005, Gastroenterology. 129(3), 1047. Azzoni, C, Bottarelli, L, Campanini, N, Di Cola, G, Bader, G, Mazzeo, A, et al. 2007, Int J Colorectal Dis. , 22(2),115. Samowitz, W, Curtin, K, Ni Ma, Khe, Schaffer, D, Coleman, LW, Leppert, M, and Slattery, M. 2001, Cancer Epidemiol Biomarkers Prev., 10(9), 917. Cheng, X, Chen, VW, Steele, B, Ruiz, B, Fulton, J, Liu, L, et al. 2001, Cancer, 92(10), 2547. Ikeda, Y, Koyanagi, N, Mori, M, Minagawa, S, Toyomasu, T, Ezaki, T, et al. 1998, Hepatogastroenterology, 45(23), 1535. Coco, C, Magistrelli, P, Vecchio, FM, Roncolini, G, Granone, P, D'Ugo, D, Picciocchi, A., 1991, Ann Ital Chir., 62(4), 355. De Leon, ML, Schoetz, DJ Jr, Coller, JA, and Veidenheimer, MC. 1987, Dis Colon Rectum, 30(4), 237. Halvorsen, TB and Seim, E. 1987, Scand J Gastroenterol., 22(1), 124.
28
María Luque Cabal
19. Wolmark, N, Wieand, HS, Rockette, HE, Fisher, B, Glass, A, Lawrence, W, et al. 1983, Ann Surg., 198(6), 743. 20. Böhm, B, Nouchirvani, K, Hucke, HP, and Stock, W. 1990, Med Klin (Munich), 85(10), 586. 21. Brozek, W, Kriwanek, S, Bonner, E, Peterlik, M and Cross, HS. 2009, Anticancer Research, 29(9), 3721. 22. Rim, SH, Seeff, L, Ahmed, F, King, JB and. Coughlin, SS. 2009, Cancer, 115, 1967. 23. Paulson, EC, Wirtalla, C, Armstrong, K and Mahmoud, NN. 2009, Dis Colon Rectum, 52(12), 1982. 24. Hendifar, A, Yang, D, Lenz, F, Lurje G, Pohl, A, Lenz, C. et al. 2009, Clin Cancer Res., 15(20), 6391. 25. Center, MM, Jemal, A and Ward, E. 2009, Cancer Epidemiol Biomarkers Prev., 18, 1688. 26. Boyle, P and Langman, JS. 2000, Br Med J, 321, 805. 27. Edwards, BK, Ward, E, Kohler, BA, Eheman, C, Zauber, A, Anderson, R, et al..2010, Cancer, 116, 544. 28. Flood, DM, Weiss, NS, Cook, LS, Emerson, JC, Schwartz, SM and Potter, JD. 2000, Cancer Causes Control, 11(5), 403. 29. Anonymous. 2009, Cancer, 115, 1967. 30. Stewart BW, Kleihues P, editors. World Cancer Report. Lyon: IARC Press; 2003. 31. Moghaddam, AA, Woodward, M and Huxley, R. 2007, Cancer Epidemiol Biomarkers Prev., 16(12), 2533. 32. Dai, Z, Xu, YC, and Niu, L. 2007, World J Gastroenterol., 13(31), 4199. 33. Siegel, EM, Ulrich, CM, Poole, EM, Holmes, RS, Jacobsen, PB and Shibata, D. 2010, Cancer Control, 17(1), 52. 34. Moon, HG, Ju, YT, Jeong, CY, Jung, EJ, Lee, YJ, Hong, SC et al. 2008, Ann Surg Oncol., 15(7), 1918. 35. Huxley, R, Moghaddam, AA, Clifton, P, Czernichow, S, Parr, CL and Woodward, M. 2009, Int. J. Cancer, 125, 171. 36. Botteri E, Iodice S, Bagnardi V, Raimondi S, Lowenfels AB, Maisonneuve P. 2008, JAMA, 300(23), 2765. 37. Liang, PS, Chen, TY and Giovannucci, E. 2009, Int. J. Cancer, 124, 2406. 38. Larsson, SC, Orsini, N and Wolk, A. 2005, J Natl Cancer Inst., 97, 1679. 39. Burkitt, DP. 1971, Cancer, 28, 3. 40. Rozen, P, Hellerstein, SM and Horwitz, C. 1981, Cancer, 48(12), 2692. 41. Graham, S, Dayal, H, Swanson, M, Mittelman, A and Wilkinson, G. 1978, J Natl Cancer Inst., 61(3), 709. 42. Yeung, KS, McKeown-Eyssen, GE, Li, GF, Glazer, E, Hay, K, Child, P, et al. 1991, J Natl Cancer Inst., 83(1), 46. 43. Parnaud, G and Corpet, DE. 1997, Bull Cancer, 84(9), 899. 44. Alexander, DD, Cushing, CA, Lowe, KA, Sceurman, B and Roberts, MA.2009, Am J Clin Nutr., 89(5), 1402. 45. Liu, L, Zhuang, W, Wang, RQ, Mukherjee, R, Xiao, SM, Chen, Z, et al. 2010, Eur J Nutr., August 10. [Epub ahead of print].
Colorectal cancer epidemiology
29
46. Chao, A, Thun, MJ, Connell, CJ, McCullough, ML, Jacobs, EJ, Flanders, WD, et al. 2005, JAMA, 293, 172. 47. Larsson, SC and Wolk, A. 2006, Int. J. Cancer, 119, 2657. 48. Moskal, A, Norat, T, Ferrari, P and Riboli, E. 2007, Int J Cancer, 120(3), 664. 49. Michels, KB, Fuchs, CS, Giovannucci, E, Colditz, GA, Hunter, MJ, and Willett, WC. 2005, Cancer Epidemiol Biomarkers Prev., 14(4), 842. 50. Park, Y, Hunter, DJ, Spiegelman, D, Bergkvist, L, Berrino, F, Van den Brandt, PA., et al. 2005, JAMA, 294(22), 2849. 51. Alberts, DS, Martínez, ME, Roe, DJ, Guillén-Rodríguez, JM, Marshall, JR, Van Leeuwen, JB, et al. 2000, N Engl J Med., 342(16), 1156. 52. Samad, AK, Taylor, RS, Marshall, T and Chapman, MA. 2005, Colorectal Dis., 7(3), 204. 53. Cho, E, Smith-Warner, SA, Spiegelman, D, Beeson, WL, Van den Brandt, PA, Colditz, GA, et al. 2004, J Natl Cancer Inst., 96(13), 1015. 54. Grau, MV, Baron, JA, Sandler, RS, Haile, RW, Beach, ML, Church, TR, and Heber, D. 2003, J Natl Cancer Inst., 95(23), 1765. 55. Wallace, K, Baron, JA, Cole, BF, Sandler, RS, Karagas, MR, Beach, MA, et al. 2004, J Natl Cancer Inst., 96(12), 921. 56. Wei, MY, Garland, CF, Gorham, ED, Mohr, SB and Giovannucci, E. 2008, Cancer Epidemiol Biomarkers Prev., 17(11), 2958. 57. Crohn, B, Rosenberg, H. 1925, Am J Med Sci., 170, 220. 58. Broström, O, Löfberg, R, Nordenvall, B, Ost, A and Hellers, G. 1987, Scand J Gastroenterol., 22(10), 1193. 59. Devroede, GJ, Taylor, WF, Sauer, WG, Jackman, RJ and Stickler, GB. 1971, N Engl J Med., 285(1), 17. 60. No author listed. 1964, Gut, 5(1), 15. 61. Gilat, T, Fireman, Z, Grossman, A, Hacohen, D, Kadish, U, Ron, E, et al. 1988, Gastroenterology, 94(4), 870. 62. Greenstein, AJ, Sachar, DB, Smith, H, Pucillo, A, Papatestas, AE, Kreel, I, et al. 1979, Gastroenterology, 77(2), 290. 63. Gyde, SN, Prior, P, Allan, RN, Stevens, A, Jewell, DP, Truelove, SC, et al. 1988, Gut, 29(2), 206. 64. Katzka, I, Brody, RS, Morris, E and Katz, S. 1983, Gastroenterology, 85(1), 22. 65. Kewenter, J, Ahlman, H and Hultén, L. 1978, Ann Surg., 188(6), 824. 66. Lennard-Jones, JE, Morson, BC, Ritchie, JK and Williams, CB. 1983, Lancet, 2(8342), 149. 67. Maratka, Z, Nedbal, J, Kociánová, J, Havelka, J, Kudrmann, J and Hendl, J. 1985, Gut, 26(1), 43. 68. Mir-Madjlessi, SH, Farmer, RG, Easley, KA and Beck, GJ. 1986, Cancer, 58(7), 1569. 69. Prior, P, Gyde, SN, Macartney, JC, Thompson, H, Waterhouse, JA and Allan, RN. 1982, Gut, 23(6), 490. 70. Wandall, EP, Damkier, P, Møller, Pedersen, F, Wilson, B, and Schaffalitzky de Muckadell, OB. 2000, Scand J Gastroenterol., 35(3), 312.
30
María Luque Cabal
71. Ekbom, A, Helmick, C, Zack, M and Adami, HO. 1990, N Engl J Med., 323(18), 1228. 72. Eaden, JA, Abrams, KR and Mayberry, JF. 2001, Gut, 48(4), 526. 73. Lakatos, L, Mester, G, Erdelyi, Z, David, G, Pandur, T, Balogh, M, et al. 2006, Inflamm Bowel Dis., 12(3), 205. 74. Kornfeld, D, Ekbom, A and Ihre, T. 1997, Gut, 41(4), 522. 75. Lindberg, BU, Broomé, U and Persson, B. 2001, Dis Colon Rectum., 44(1), 77. 76. Claessen, MM, Vleggaar, FP, Tytgat, KM, Siersema, PD and Van Buuren, HR. 2009, J Hepatol., 50(1), 158. 77. Hill, MJ, Melville, DM, Lennard-Jones, JE, Neale, K and Ritchie, JK. 1987, Lancet, 2(8552), 185. 78. Shetty, K, Rybicki, L, Brzezinski, A, Carey, WD and Lashner, BA. 1999, Am J Gastroenterol., 94(6), 1643. 79. Gurbuz, KB, Giardiello, FM and Bayless, TM. 1995, Dis Colon Rectum., 38, 37. 80. Yano, Y, Matsui, T, Uno, H, Hirai, F, Futami, K and Iwashita, A. 2008, J Gastroenterol Hepatol., 23(11), 1683. 81. Larsen, M, Mose, H, Gislum, M, Skriver, MV, Jepsen, P, Nørgård, B, and Sørensen, HT. 2007, Am J Gastroenterol., 102(1), 163. 82. Gillen, CD, Andrews, HA, Prior, P and Allan, RN. 1994, Gut, 35(5), 651. 83. Weedon, DD, Shorter, RG, Ilstrup, DM, Huizenga, KA and Taylor, WF. 19873, N Engl J Med., 289(21), 1099. 84. Canavan, C, Abrams, KR and Mayberry, J. 2006, Aliment Pharmacol Ther., 23(8), 1097. 85. Choi, PM and Zelig, MP. 1994, Gut, 35(7), 950. 86. Masala, G, Bagnoli, S, Ceroti, M, Saieva, C, Trallori, G, Zanna, I, et al. 2004, Gut, 53(9), 1309. 87. Enterline, HT, Evans, GW, Mercado-Lugo, R, Miller, L, and Fitts, WT. 1962, JAMA, 179, 322. 88. Fenoglio, C. M. and Lane, N. 1974, Cancer, 34, 819. 89. Shinya, H and Wolff, WI. 1979, Ann Surg., 190(6), 679. 90. Gillespie, PE, Chambers, TJ, Chan, KW, Doronzo, F, Morson, BC and Williams, CB. 1979, Gut, 20(3), 240. 91. Atkin, WS, Morson, BC and Cuzick, J. 1992, N Engl J Med., 326(10), 658. 92. Bisgaard, ML, Fenger, K, Bülow, S, Niebuhr, E and Mohr, J. 1994, Hum Mutat., 3(2), 121. 93. Half, E, Bercovich, D and Rozen, P. 2009, Orphanet J Rare Dis., 12 (4), 22. 94. Varesco L. 2004, Tech Coloproctol., 8 Suppl 2:s305. 95. Arvanitis, ML, Jagelman, DG, Fazio, VW, Lavery, IC, and McGagnon, E. 1990, Dis Colon Rectum., 33, 639. 96. Nugent, K, Spigelman, AD and Phillips, RKS. 1993, Dis Colon Rectum., 36, 1059. 97. Galle, TS, Juel, K and Bülow, S. 1999, Scand J Gastroenterol., 34, 808. 98. Bertario, L, Presciuttini, S, Sala, P, Rossetti, C, and Pietroiusti, M. 1994, Semin Surg Oncol., 10, 225. 99. Mecklin, JP and Jarvinen, HJ. 1991, Cancer, 68, 1109.
Colorectal cancer epidemiology
31
100. Umar, A, Boland, CR, Terdiman, JP, Syngal, S, de la Chapelle, A, Rüschoff, J et al. 2004, J Natl Cancer Inst., 96, 261 101. Peltomäki, P and Vasen, H. 2004, Dis Markers, 20, 269. 102. Aarnio, M , Sankila, R , Pukkala, E, Salovaara, R, Aaltonen, LA, de la Chapelle, A, et al. 1999, Int J Cancer, 81, 214. 103. Chou, CL, Lin, JK, Wang, HS, Yang, SH, Li, AF and Chang, SC. 2010, Int J Colorectal Dis., 25(1), 47. 104. Järvinen, HJ, Renkonen-Sinisalo, L, Aktán-Collán, K, Peltomäki, P, Aaltonen, LA and Mecklin, JP. 2009, J Clin Oncol., 27(28), 4793. 105. Kouraklis, G and Misiakos, EP. 2005, Dig Dis Sci., 50(2), 336. 106. Ramsey, SD, Yoon, P, Moonesinghe, R and Khoury, MJ. 2006, Genet Med., 8, 571. 107. Butterworth, AS, Higgins, JP and Pharoah, P. 2006, Eur J Cancer, 42, 216. 108. Johns, LE and Houlston, RS. 2001, Am J Gastroenterol., 96, 2992. 109. Fearon ER and Vogelstein B. 1990, Cell, 61(5), 759. 110. Winawer, SJ, Zauber, AG, Ho, MN, O´Brian, MJ, Gottlieb, LS, Sternberg, SS et al. 1993, N Engl J Med., 329(27), 1977. 111. World Health Organization 2007, The World Health Organization’s Fight Against Cancer: Strategies That Prevent, Cure and Care, Who Press, Inc., Geneva. 112. Mandel, JS, Church, TR, Ederer, F and Bond, JH. 1999, J Natl Cancer Inst., 91, 434. 113. Atkin, WS, Edwards, R, Kralj-Hans, I, Wooldrage, K, Hart, AR, Northover, J et al. 2010, Lancet, 375, 1624. 114. Fosslien E. 2000, Crit Rev Clin Lab Sci., 37, 431. 115. Tsujii, M, Kawano, S, Tsuji, S, Sawaoka, H, Hori, M, DuBois, RN. 1998, Cell, 93, 705. 116. DuBois, RN, Giardiello, FM and Smalley, WE. 1996, Gastroenterol Clin North Am., 25, 773. 117. Dubé C, Rostom A, Lewin G, Tsertsvadze, A, Barrowman, N, Code, C, et al. 2007, Ann Intern Med., 146(5), 365. 118. Rostom A, Dubé C, Lewin G, Tsertsvadze, A, Barrowman, N, Code, C, et al. 2007, Ann Intern Med., 146(5), 376. 119. Lipworth, L, Friis, S, Blot, WJ, McLaughlin, JK, Mellemkjaer, L, Johnsen, SP, et al. 2004, Am J Ther.,11, 156. 120. Baron, JA, Sandler, RS, Bresalier, RS, Quan, H, Riddell, R, Lanas, A, et al. 2006, Gastroenterology, 131, 1674. 121. Bertagnolli, MM, Eagle, CJ, Zauber, AG, Redston, M, Breazna, A, Kim, K,et al. 2009, Cancer Prev Res (Phila)., 2(4), 310. 122. Steinbach, G, Lynch, PM, Phillips, RK, Wallace, MH, Hawk, E, Gordon, GB, et al. 2000, N Engl J Med., 342(26), 1946. 123. Phillips, RK, Wallace, MH, Lynch, PM, Hawk, E, Gordon, GB, Saunders, BP, et al. 2002, Gut, 50(6), 857. 124. Bernstein, CN, Blanchard, JF and Metge, C. 2003, Am J Gastroenterol., 98, 2784. 125. Pinczowski, D, Ekbom, A and Baron, J. 1994, Gastroenterology, 107, 117.
32
María Luque Cabal
126. Velayos, FS, Terdiman, JP and Walsh, JM. 2005, Am J Gastroenterol., 100(6):1345. 127. Tung, BY, Emond, MJ, Haggitt, RC, Bronner, MP, Kimmey, MB, Kodwley, K, et al. 2001, Ann Intern Med., 134(2), 89. 128. Pardi, DS, Loftus, EV Jr, Kremers, WK, Keach, J and Lindor, KD. 2003, Gastroenterology., 124(4), 889. 129. Sanjoaquin MA, Allen N, Couto E, Roddam, AW, and Key, TJ. 2005, Int J Cancer, 113, 825. 130. Luebeck, EG, Moolgavkar, SH, Liu, AY, Boynton, A and Ulrich, CM. 2008, Cancer Epidemiol Biomarkers Prev., 17(6), 1360. 131. Hubner, RA and Houlston, RS. 2009, Br J Cancer, 100(2), 233. 132. Winawer, SJ, Zauber, AG, Ho, MN, O'Brien, MJ, Gottlieb, LS, Sternberg, SS, et al. 1993, N Engl J Med., 329(27), 1977. 133. Nozaki, R, Takagi, K, Takano, M and Miyata, M. 1997, Dis Colon Rectum, 40(10 Suppl), S16. 134. Citarda, F, Tomaselli, G, Capocaccia, R, Barcherini, S and Crespi, M. 2001, Gut, 48(6), 812. 135. Hawk, ET and Levin, B. 2005, J Clin Oncol., 23(2), 378. 136. Mandel JS, Bond JH, Church TR, Snover, DC, Bradley, GM, Schuman, LM, and Ederer F. 1993, N Engl J Med., 328(19), 1365. 137. Mandel, JS, Church, TR, Ederer ,F, Bond, J. 1999, J Natl Cancer Inst., 91, 434. 138. Hardcastle, JD, Chamberlain, JO, Robinson, MH, Moss, SM, Amar, SS, Balfour, TW, et al.1996, Lancet, 348, 1472. 139. Kronborg, O, Fenger, C, Olsen, J, Jørgensen, OD, and Søndergaard, O. 1996, Lancet, 348, 1467. 140. Towler, B, Irwig, L, Glasziou, P, Kewenter, J, Weller, D, and Silagy, C. 1998, BMJ., 317, 559. 141. Selby JV, Friedman GD, Quesenberry CP Jr, and Weis, NS. 1992, N Engl J Med., 326, 653. 142. Newcomb PA, Norfleet RG, Storer BE, Surawic, TS, and Marcus, PM. 1992, J Natl Cancer Inst., 84, 1572. 143. Muller, AD and Sonnenberg, A. 1995, Ann Intern Med., 123, 904. 144. Thiis-Evensen, E, Hoff, GS, Sauar, J, Langmark, F, Majak, BM, and Vatn, MH.1999, Scand J Gastroenterol., 34, 414. 145. van Rijn, JC, Reitsma, JB, Stoker, J, Bossuyt, PM, van Deventer, SJ and Dekker, E. 2006, Am J Gastroenterol., 101(2), 343. 146. Strul, H, Kariv, R, Leshno, M, Halak, A, Jakubowicz, M, Santo, M, et al. 2006, Am J Gastroenterol., 101(2), 255. 147. Baxter, NN, Goldwasser, MA, Paszat, LF, Saskin, R, Urbach, DR and Rabeneck, L. 2009, Ann Intern Med., 150(1), 1. 148. Kim DH, Pickhardt PJ, Taylor AJ, Leung, WK, Winter, TC, Hinshaw, JL et al. 2007, N Engl J Med., 357(14), 1403. 149. Johnson CD, Chen MH, Toledano AY, Heiken JP, Dachman A, Kuo MD et al. 2008, N Engl J Med., 359(12), 1207. 150. Sonnenberg, A, Delco, F and Bauerfeind, P. 1999, Am J Gastroenterol., 94, 2268.
Colorectal cancer epidemiology
33
151. Ladabaum, U, Song, K and Fendrick, AM. 2004, Clin Gastroenterol Hepatol., 2, 554. 152. World Gastroenterology Organisation (WGO). Practice guidelines: colorectal cancer screening. Paris (France): World Gastroenterology Organisation (WGO); 2007. 18 p. 153. Levin B, Lieberman DA, McFarland B, Andrews KS, Brooks D, Bond Jet al.. 2008, CA Cancer J Clin., 58(3), 130. 154. Davila, RE, Rajan, E, Baron, TH, Adler, DG, Egan, JV, Faigel, DO, et al.2006, Gastrointest Endosc., 63(4), 546.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 35-51 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
2. Molecular biology of colorectal cancer Enrique Grande-Pulido, Alejandro Riquelme-Oliveira Javier Ballesteros-Bargues, Carmen Guillén-Ponce and Alfredo Carrato Medical Oncology Department, “Ramón y Cajal” University Hospital, Madrid, Spain
Abstract. An estimated one million people in the world are diagnosed each year with CRC cancer. Over 30% of these people will die of disseminated disease. In recent years, we have furthered our knowledge of the molecular mechanisms responsible for the onset, development, and spread of the disease to other organs. The Vogelstein model of CRC tumorigenesis, published in the early 1990s, pioneered our understanding of how solid tumours developed. Major advances in the field of tumour neoangiogenesis and greater knowledge of the function of epidermal growth-factor receptors have since added new specifically targeted drugs, such as bevacizumab, cetuximab, and panitumumab, to the armamentarium. Moreover, we know that certain changes in the intracellular signalling cascade triggered by these receptors can influence drug response. In this chapter we summarise our knowledge of tumour-cell physiology itself, the microenvironment surrounding colon cancer, and the possible involvement of epigenetic phenomena and stem cells in the genesis and development of these tumours. Correspondence/Reprint request: Dr. Enrique Grande-Pulido, Medical Oncology Department, “Ramón y Cajal” University Hospital, Madrid, Spain. E-mail: egrande@oncologiahrc.com
36
Enrique Grande-Pulido et al.
Introduction The relevance of CRC in neoplastic diseases and general gastrointestinal disorders has been amply demonstrated. With a worldwide incidence of more than one million cases a year, CRC is the second leading cause of cancer deaths in adults, with a mortality rate of 33% in developed countries [1]. CRC is a paramount health problem for the World Health Organization (WHO). Despite advances in screening, diagnostics, and treatment in this field in recent decades, which have led to increased cure and response rates and improved quality of life, much of the natural history of the disease, especially its origins, remains unknown. Research in recent years has thus focused on the molecular biology of the disease. The underlying goal is not only to improve prevention and early diagnosis, but also to discover new weapons to fight the disease [2]. Advances in molecular biology have led to new therapeutic targets and new molecular prognostic markers and predictors of response to current treatments [3]. The aim is to try to individualise treatments, in accordance with the tumour's characteristics, and to obtain therapies that act directly on the tumour [4]. Our knowledge of how tumours develop, however, is still insufficient; we lack a complete understanding of carcinogenesis at the molecular level. We continue to investigate the different pathways by which a normal mucosa or a simple adenomatous polyp turns into a neoplasia, with the ability to infiltrate and trigger distant dissemination. Understanding these molecular pathways will enable us to generate hypotheses for more-specific and less-toxic future therapies and molecular treatments that will also be much more effective. The great hope for the future appears to lie in increasing our knowledge of the molecular biology of CRC [5].
Sporadic versus hereditary molecular bases of colorectal cancer About 20-30% of all colorectal cancers are due to a hereditary, or familial predisposition. Patients with a history of first or second degree relatives with CRC have a 20% higher risk of suffering the disease. This risk increases considerably, to 80-100% in some cases, if the disease is associated with specific hereditary syndromes [6]. Cases associated with well-defined hereditary syndromes like Lynch Syndrome or Familial Adenomatous Polyposis account for less than 5% of all colorectal cancers. The remaining 20-25%, associated with some type of hereditary form, tend to occur as a result of polymorphisms or low-
Basic science behind colorectal cancer
37
penetrance mutations with no clear association to a clear hereditary syndrome [6]. These cases are less genetically determined than with hereditary syndromes, which we will only briefly discuss, outlining just a few of their characteristics, as they are discussed in much greater depth elsewhere in this book.
Hereditary nonpolyposis colorectal cancer (HNPCC), or Lynch syndrome HNPCC accounts for 2-4% of all colorectal cancers. Patients have an 80% risk of developing cancer during their lives. HNPCC is an autosomal dominant hereditary condition. It is associated with endometrial tumours (40-60%) and less commonly with gastric, ovarian, urinary, bowel, pancreatic, and brain tumours [7]. It is genetically characterised by high levels of microsatellite instability due to mutations in genes of the mismatchrepair (MMR) system. In 90% of cases, mutations are found in MSH2 or MLH1, and occasionally in MSH6 or PMS2 [8]. Apart from the diagnosis of HNPCC through genetic testing, clinical criteria can also be applied, as described in the Amsterdam II Criteria and Bethesda guidelines.
Familial adenomatous polyposis (FAP) FAP accounts for about 1% of all colorectal cancers. It is characterised by the appearance of hundreds to thousands of adenomatous polyps in the colon during adolescence. It is an autosomal dominant hereditary disorder, with 100% penetrance (if left untreated, all patients would develop colorectal cancer) [7]. FAP is genetically characterised by germline mutation in the Adenomatous Polyposis Coli (APC) tumour-suppressor gene and is associated with extracolonic manifestations (GI polyposis, congenital hypertrophy of the retinal pigment epithelium, fibromas, desmoid tumours, dental abnormalities). An attenuated variant (with >10 and <100 colonic polyps) exists. This attenuated variant has a specific hereditary variant called Gardner syndrome (FAP, osteomas, fibromas of the skin and epidermoid cysts). All patients must undergo colectomy at an early age [9].
MUTYH associated polyposis (MAP) MAP is a rare, autosomal recessive hereditary syndrome. It has a similar clinical presentation to attenuated FAP, with polyposis. However, the polyps are not only adenomatous, but also hyperplastic and sessile serrated. MAP is
38
Enrique Grande-Pulido et al.
caused by a biallelic germline mutation in MUTYH (MYH), a DNA baseexcision repair gene that acts when DNA damage is caused by oxidative stress [10].
Hamartomatous polyposis syndromes There are many hamartomatous polyposis syndromes, which entail an increased risk of colorectal cancer, but their incidences are very low. The two most common syndromes in this family are: Peutz-Jeghers syndrome.Æ This autosomal dominant syndrome is characterised by the presence of multiple hamartomatous polyps in the small intestine, stomach, and colon. Presenting symptoms are usually obstruction and intestinal bleeding. Common extraintestinal manifestations are pigmentation of the lips and oral and periorbital mucosa. The only known genetic alteration is the STK11 (LKB1) mutation, which is involved in the TGFβ pathogenic pathway. These patients have an 81-93% risk of CRC[11]. Juvenile polyposis syndrome.Æ This autosomal dominant syndrome presents with juvenile polyps in the GI tract, most often in the colon, with a clinical presentation similar to Peutz-Jeghers syndrome. The associated genetic alterations are mutations in SMAD4 and BMPR1A, which are also involved in the TGFβ pathway. The risk of developing CRC is 39% [12].
Carcinogenesis in colorectal cancer: From Vogelstein’s tumour model to the present day When a neoplasm develops from the epithelium of normal colonic mucosa, with transition into adenomatous epithelium, it presents a wellestablished adenoma-carcinoma sequence. In 1990, Vogelstein and Fearon proposed a multistep model in tumour carcinogenesis [13]. This model was the starting point for a course of research that has led to today's knowledge of tumour carcinogenesis triggered by cellular genomic instability. The loss of genomic stability opens the door to the development of colorectal cancer. Three different types of pathological pathways lead to genomic instability: chromosomal instability (CIN), microsatellite instability (MSI), and the CpG-island methylator phenotype (CIMP). These pathways are not mutually exclusive and therefore more than one may be involved in both sporadic and hereditary development of cancer. The alterations necessary for cancer development are not precisely known, nor are when in
Basic science behind colorectal cancer
39
the sequence such alterations must occur, but what is indeed known is the moment when certain components of these pathways play a major role [14,15].
Chromosomal instability (CIN) CIN occurs in 65-70% of sporadic colorectal cancers [16]. CIN is due to defects in different phases of the cell cycle at a genomic level, which often lead to the activation of oncogenes and inactivation of tumour-suppressor genes. The most common mechanisms that cause CIN are: abnormalities in chromosome segregation at mitosis (with reference to aberrations at checkpoints as well as centrosomal aberrations and alterations in Aurora and Polo kinases), telomeric dysfunction (excessive activation of telomerase, which is very common in CRC), alterations in the machinery responding to DNA damage (the most characteristic of which are alterations in the p53 gene), and loss of heterozygosity of alleles (through aberrations in chromosomal recombination, or deletions, for example). Over a hundred genes are involved in the pathways of chromosomal instability [17]. We will discuss the most important genes. APCÆ APC is a tumour-suppressor gene found on chromosome 5q. This gene is the most frequently mutated (in 80% of adenomas and CRC) and is considered to be the site of the earliest genetic alterations in the sequence of carcinogenesis. Its mutation in the germline is responsible for Familial Adenomatous Polyposis. The APC gene encodes the APC protein. Among the latter's many cellular regulatory functions is its responsiblity for the degradation of the β-catenin protein involved in Wnt stem signaling pathway [18,19]. Degradation of the β-catenin protein is prevented through the aberrant Wnt pathway, by accumulation in the cytosol and entrance to the nucleus, where it binds to the lymphoid enhancer factor/T-cell factor (LEF/TCF), producing an increased transcription of genes that stimulate cell growth and inhibit apoptosis (such as c-myc, c-jun, Fra-1). If an APC mutation prevents formation of the APC protein, or causes aberrant protein formation, then degradation of β-catenin is prevented, thus permitting uncontrolled activation of the Wnt pathway. Although other alterations occur at other points in this pathway, the APC mutation is the most common [20]. KRASÆ KRAS is an oncogene that mutates in 30-50% of CRCs. These mutations are thought to be the second alteration in the sequence of carcinogenesis and occur in precursor lesions of adenomas and carcinomas. The KRAS mutations block the action of the enzyme GTPase, permitting the
40
Enrique Grande-Pulido et al.
uncontrolled activation of the ras cascade, which results in inhibition of apoptosis and in cell proliferation [21]. TP53Æ TP53 is a tumour-suppressor gene, known as “the guardian of the genome”. Mutations here are believed to be the third alteration that occurs in the sequence of carcinogenesis, since they appear in the transition from adenoma to carcinoma. The p53 protein is responsible for detecting changes in the DNA (direct damage, aberrant proliferative signals, oxidative stress), which result in putting the cell cycle on hold (cell arrest) in order for these aberrations to be repaired, or to proceed to apoptosis if repair is not achieved. Mutations in TP53 are universal alterations in human tumours and appear in 50-75% of CRCs [22]. OthersÆ In addition to mutations in the above genes, many other alterations take place along the sequence. Other oncogenes are activated, such as BRAF, PIK3CA (related to the Akt/mTOR pathway), and CTNNB1 (the β-catenin gene); also, inactivation of many tumour-suppressor genes, such as those caused by loss of heterozygosity of 18q, SMAD2 and SMAD4 (TGF-β pathway mediators), and DCC. Overexpression of COX-2 has also been observed (and the consequent overproduction of PGE2) in the production of proangiogenic factors [22-25].
Microsatellite instability (MSI) MSI occurs in 12-17% of colorectal cancers (3% of which are hereditary) [26]. This pathogenic pathway develops from mutations in the DNA mismatch-repair systems (MMR). Resulting tumours present common clinical features, such as onset in adulthood, occurrence in the proximal colon, and a better prognosis. The presence of this pathway is called the “mutator phenotype” [27]. DNA polymerases sometimes incorporate an incorrect number of bases during replication in long repetitive DNA sequences such as microsatellites. These errors are usually detected by the MMR system, which is responsible for halting and repairing the replication. However, mutation in any component of this system leads to aberrations, resulting in the appearance of abnormal microsatellites. These, in turn, tend to occur in gene-promoter regions that regulate cell growth, in genes that are involved in carcinogenesis (TGF-β, BAX, Caspase-5, IGF-I, etc), thus leading to abnormal and uncontrolled cell growth [28]. DNA mismatch-repair systems contain different homologous genes that transcribe the proteins responsible for repairing transcription mismatch
Basic science behind colorectal cancer
41
errors. The genes involved are: MLH1, MLH3, MSH2, MSH3, MSH6, PMS1, and PMS2. In sporadic forms of MSI, MLH1 is the most commonly mutated gene, which is silenced by promoter methylation [29]. Lynch syndrome, or HNPCC (hereditary nonpolyposis colorectal cancer), occurs through the MSI pathway in germlines, mainly by mutations in MSH2 and MLH1, which leave a single functioning copy of the protein in each cell. The characteristics of Lynch syndrome will be discussed in more detail later.
CPG island methylator phenotype (CIMP) DNA methylation is an epigenetic form of regulating gene transcription. In normal physiological function, CpG islands (methylation of repetitive sequences between cytosine and guanine) can be found in the promoters of certain genes, leading to their silencing. In the same way, in a pathological process, they provide a pathway for carcinogenesis. Methylation through promoter CpG islands leads to the silencing of gene transcription, which happens in tumour-suppressor genes during tumour carcinogenesis, resulting in inactivation (e.g. p16, MINT1, NEUROG1, MLH1) [30]. This pathway has been observed in 24-51% of colorectal cancers, depending on the detection techniques used [31]. The most significant and frequent methylation occurs on MLH1 in sporadic colorectal cancers, producing high levels of MSI. However, the presence of CIMP+ together with MSI- has been demonstrated to have a worse prognosis. This finding shows that these disease pathways are not mutually exclusive; they coexist within the same tumour, and their different combinations define the individual characteristics of each tumour [32].
Role of the Epidermal Growth-Factor Receptor (EGFR) pathway in colorectal cancer development As we have already seen, many molecular pathogenic pathways influence the origin, development, growth, and local and distant invasion of tumours of colorectal origin. This discovery is very important in many practical and clinical aspects and means that we can further our knowledge of the disease and develop specific therapeutic targets. Within this group of pathways, we would like to highlight the Epidermal Growth Factor (EGF), and in particular, the family of receptors to which EGF binds, and the pathways that the receptors activate [33].
42
Enrique Grande-Pulido et al.
EGFR The family of ErbB (HER) receptors - transmembrane receptors associated with tyrosine kinases - is associated with many pathogenic pathways of carcinogenesis in different tumours. These receptors have four subtypes: ErbB1 (HER1/EGFR), ErbB2 (HER2/neu), ErbB3 (HER3), and ErbB4 (HER4) [34]. EGFR activation occurs through ligand-dependent (e.g. EGF, TGFÎą), ligand-independent (e.g. urokinase, plasminogen), and overexpression mechanisms (in cancer cells, the result is an activation by ligand-independent receptor dimerisation). In all cases, EGFR activation leads to a conformational change in the receptor, with subsequent activation of the associated tyrosine kinase. The activation of EGFR generates the activation of a cascade of multiple associated pathogenic signalling pathways that are involved in cell development, growth, and apoptosis. Alteration of these pathways leads to a state of cell growth and resistance to apoptosis [35]. Uncontrolled activation of EGFR activates the different associated intracellular pathogenic pathways. The RAS-RAF-MAPK pathway is activated, which also occurs through errors in the control of the RAS and RAF oncogenes, leading to cell growth, differentiation, and resistance to apoptosis [36]. Another pathway that is activated is the phosphatidylinositol 3-kinase (PI3K) pathway, which culminates by regulating the activation of AKT/PKB. This pathway is also related to cell growth and survival, by means of proapoptotic protein inactivation. In turn, a suppressor gene, PTEN, negatively regulates this pathway. Mutation of PTEN generates an accumulation of PI3K, making this pathway hyperactive [37]. In addition to these pathways, stress mechanisms activate EGFR and its protein kinase, which in turn activates transcription-factor pathways, such as Protein Kinase C, Jak, and STAT. We know of many different activated pathways, but others are likely yet to be discovered, for the end result is a highly elaborate cascade of activated phenomena, with many intertwined pathways [38,39].
Targeting the EGFR system Several molecules have been tested against the EGFR pathway, and many more are under development. Some are already available in clinical practice [40]. CetuximabĂ&#x2020; This molecule is a monoclonal antibody (IgG1) that binds to the extracellular domain of EGFR causing internalisation and degradation
Basic science behind colorectal cancer
43
of the receptor without activating it. The result is the inhibition of EGFRdependent cell growth [41]. PanitumumabĂ&#x2020; This molecule is another monoclonal antibody (IgG2) that also binds to the extracellular domain of EGFR to competitively inhibit other ligands. It prevents EGFR dimerisation, and therefore the EGFR cascade is not activated [42].
Angiogenesis in colorectal cancer New vessel formation is mediated by the balance between proangiogenic factors (vascular endothelial growth factor (VEGF), fibroblast growth factors (FGF), platelet-derived growth factor (PDGF), insulin-like growth factor (IGF), and transforming growth factor (TGF)) and antiangiogenic factors (thrombospondin-1, angiostatin, and endostatin) [43]. Neovascularisation that accompanies tumour growth is aberrant and therefore has different characteristics from normal tissue vascularisation. Tumour-induced vessels are fenestrated, i.e., they have pores through which plasma leaks into the extravascular space, with increased interstitial fluid pressure, leading to vessel collapse, greater ischaemia, and acidosis in the tumour tissue. Paradoxically, despite the increased number of vessels, tumour tissues are poorly oxygenated. This tissue hypoxia leads to increased production of pro-angiogenic factors that contribute in turn to generating more vessels, which sets up a cycle favouring further malignant formation. The acidotic, hypoxic environment leads to greater genetic and chromosomal instability, therefore enhancing the malignant potential of these tumours. In turn, the fenestration in the newly formed vessels encourages tumour dissemination, because it is easier for the tumour cells to penetrate the new vessels and metastasise at a distance, therefore worsening the prognosis [44,45]. VEGF is the main growth factor involved in angiogenesis in colorectal cancer. It activates different intracellular signalling cascades, leading to endothelial cell growth, migration, differentiation, and enhanced vascular permeability. The VEGF family of growth factors is composed of several proteins encoded by different genes: VEGF-A, VEGF-B, VEGF-C, VEGF-D, VEGFR-E, and placental growth factor (PIGF). These growth factors are activated when they bind to different tyrosine-kinase-type receptors present in the cell membranes of endothelial cells (VEGFR-1, -2, and -3) [46]. As the tumour grows, the demand for oxygen and nutrients increases, which leads to the intracellular stabilisation of the Îą subunit of the hypoxia-
44
Enrique Grande-Pulido et al.
inducible factor (HIF-α), thus preventing proteasomal degradation. If HIF-α remains in the cell for a longer time, it binds to the β subunit, forming a HIF-αβ complex that acts as a transcription factor at a nuclear level for many pro-angiogenic factors, including VEGF, FGF, and EGF [47]. We know of three VEGF receptors, which are characterised by having tyrosine-kinase domains in their intracellular portions. Each of these receptors binds with different affinity to the different isoforms of VEGF (-A, -B, -C, -D, -E, and PIGF), playing an important role in tumour angiogenesis, especially in the case of VEGFR-2, which binds to VEGF-A. Also, some coreceptors called neuropilins (-1 and -2) function to increase the affinity of the receptors and their ligands. Another mechanisms in the regulation of angiogenesis functions independently of VEGF and its receptors. These mechanisms operate through alternative signalling pathways that involve numerous ligands and theirreceptors, such as angiopoietins and their receptors (tie-2) [48], Ephrin B2 and EphB4 [49], and Delta and Notch [50,51]. Thus, angiopoetin-1 (Ang1) binds to tie-2, activating the PI3K/AKT/mTOR signalling cascade. Ang1 plays a pro-angiogenic role, whereas angiopoietin-2 (Ang2) acts as an antiangiogenic agent. The binding of EphrinB2 and EphB4 stimulates the formation of new capillaries. In the same way, Delta activates the Notch pathway leading to the formation of new vessels. IL-8 is an important factor involved in angiogenesis that appears to regulate angiogenic activation under hypoxic conditions independently of VEGF activation [52]. In fact, patients with stage III CRC who are carriers of a specific IL8 polymorphism (T-251A) overexpress this factor, which implies a worse prognosis [53]. Because angiogenesis is so important in tumour growth and survival, numerous target therapies for inhibiting angiogenesis have been proposed, ranging from anti-VEGF or anti-VEGFR-2 antibodies to receptor tyrosinekinase inhibitors [54].
Microenvironment in colorectal cancer: Beyond angiogenesis Precise regulation of intestinal tissue stroma is needed to ensure a perfect balance and permit constant cell renewal. At the base of intestinal crypts there is a specific compartment of stem cells that migrate upwards from the proliferation phase to form enteroendocrine cells in the luminal surface of the digestive tract. The stroma play an important role in establishing a positional gradient to effect signalling and correct binding between the different ligands. Myofibroblasts are the specific stromal cells that perform this process [55].
Basic science behind colorectal cancer
45
Stromal regulation is necessary to preserve normal tissue architecture. Myofibroblasts produce a variety of growth factors, prostaglandins, cytokines, and components of the extracellular matrix that facilitate tissue repair and survival [56]. Myofibroblasts are one of the most abundant cell types in the stroma associated with tumour growth. In fact, as the tumour grows, the microenvironment plays an increasingly important part in malignant formation. Cancer-associated or peritumoural myofibroblasts acquire different phenotypic characteristics, increasing production of growth factors and remodelling of proteases in the extracellular matrix, which facilitate tumour migration and invasion [57]. The growth factors in tumour stroma modulates the accumulation of nuclear β-catenin through PDGF. Stromal cells may also provide a suitable microenvironment for the maintenance of tumour stem cells both at the original tumour site and in invasive behaviour and distant migration. The cellular microenvironment therefore acts as a co-effector of a tumour's metastatic ability to increase its malignant potential by providing a suitable substrate for growth. As stromal cells modulate both tumour growth and accumulation of nuclear β-catenin, specific characteristics of these cells can be selected during tumorigenesis so that they continue to provide a suitable microenvironment [58]. In addition to the stromal fibroblasts, other cell types in the peritumoural microenvironment are involved in tumour growth and development. These include T cells. Increased tumour infiltration by these cells should be associated with a better prognosis; however, a subtype of T cell called Tregs that stimulates tumour growth through immunosuppression prevents autoimmunity and permits the growth of commensal bacterial flora. Patients with CRC have an increased number of this cell type in their peripheral blood [59,60]. Other cell types in the microenvironment that may play an important role in tumour growth are cancer-associated macrophages and immature myeloid dendritic cells. All these factors suggest that the inflammatory response is key to tumour survival, extravasation, and metastatic formation. In fact, tumour growth and development depend on the ability of the different stromal factors involved to modulate activity of β-catenin. Other non-cellular factors in the microenvironment are involved in tumour development and growth, and even in tumour initiation. The hypoxic conditions that characterise tumour progression are associated with the stabilisation of HIF1alpha (hypoxia-inducible factor-1 alpha), which is correlated with a worse prognosis in colon cancer. HIF1alpha binds directly to β-catenin in the nucleus, which is why activation of β-catenin changes under hypoxic conditions [61].
46
Enrique Grande-Pulido et al.
Stem-cell routes involved in colorectal cancer maintenance The onset of tumorigenesis could occur through the action of a single undifferentiated cell known as a tumour stem cell. These cells are not always destroyed by antitumour treatments and appear to play an important role in resistance to cytotoxic and cytostatic drugs, radiotherapy, and to the new antitarget drugs [62]. Stem cells are undifferentiated cells that undergo asymmetric division to produce two different daughter cells. One daughter cell is identical to the original parent cell, and the other is a more specialised cell. Stem cells maintain their undifferentiated state as a result of this asymmetric division, although they are tied to the different tissues in which they are located and are responsible for maintaining balance and making repairs after any tissue stress [63]. Some studies have shown that errors in DNA replication may occur when stem cells undergo active division, which would mediate in tumour pathogenesis. These tumour stem cells derive from the division of normal stem cells as a result of abnormal differentiation, from stem cells that differentiate directly into tumour cells, or by reprogramming themselves and thus acquiring tumour behaviour. The colonic crypts of Lieberkühn house the functional unit that produces continuous cell turnover. This complex process is regulated by stem cells located in these crypts. This physical environment is known as the stem-cell niche, where the subepithelial myofibroblasts are located, which are not only responsible for providing the right environment for tumour growth as described in the previous paragraph but are also actively involved in stemcell division and differentiation through the activation of numerous growth factors. These niches ensure the correct balance between stem-cell division and differentiation [64]. The base of the intestinal crypt is characterised by the activation of the Wnt signalling pathway and is the location of nuclear β-catenin. The majority of sporadic colorectal cancers are caused by the constitutive activation of Wnt due to mutations in the APC suppressor gene or in the β-catenin oncogene, thereby leading to accumulation of nuclear β-catenin. This accumulation leads to the activation of the Wnt pathway and other signalling pathways. We know that nuclear β-catenin binds to different factors, activating several signalling pathways and promoting tumour growth and malignancy [65]. Although the tumour stem-cell theory is based on animal models, which may underestimate the tumourigenic potential of these cells, malignant
Basic science behind colorectal cancer
47
transformation is hypothesised to occur in the normal stem cell, resulting in a genetically identical population of tumour cells in which only a few cells retain the characteristics of the original parental cells, thus contributing to tumour progression [66]. On the other hand, the clonal-evolution model hypothesises that any normal cell can be transformed, and that all its offspring can acquire additional mutations, forming a mass of tumour cells with different genetic variations that will promote tumour progression. From a therapeutic point of view, differences depend on the model that is embraced. With the clonal model, the tumour would be heterogeneous, and all its cells would be treated in a targeted manner. However, if the tumour stem-cell theory is embraced, the stem cells would be the therapeutic target since they would be responsible for maintenance and recurrence following the different therapies [67,68]. The main surface markers of intestinal stem cells are Ms1, CD29, Lgr5, and DCAMKL1. Of all the different surface markers expressed on colorectal tumour stem cells, no specific marker has yet been identified, although CD133 may be the most important and most widely expressed in these cells [69]. Data on the expression of this marker, however, are contradictory. ESA, CD44, CD166, Msi1, CD29, CD24, Lgr5, and ALDH1 are other surface markers expressed on colorectal tumour stem cells, and, in fact, during the progression from adenoma to carcinoma, the number of positive cells for CD133, CD44, and ALDH is clearly increased following the gradient in the crypts [70].
Epigenetics in colorectal cancer: Does it really matter? Epigenetic instability appears to play an important role in colorectal carcinogenesis. DNA methylation is an epigenetic mechanism of regulating gene transcription. Aberrant methylation of genes is a mechanism of gene inactivation in patients with CRC[71,72]. DNA methylation is present in much of the human genome in a relatively stable form. A significant number of CpG dinucleotides are carriers of epigenetic modifications, constituting the so-called CpG islands that are normally found in unmethylated form. An aberrant methylation of one of these regions is quite often accompanied by transcriptional silencing. If this aberrant methylation occurs in certain genes, such as MLH1, MGMT, and HIC1, tumour pathogenesis may be affected. In fact, aberrant methylation of MLH1 is found in 80% of sporadic colorectal cancers with microsatellite instability. Aberrant promoter methylation of certain genes (HLTF, SLC5A8, MGMT, MINT1, and MINT31) occurs in the early stages of the adenomacarcinoma sequence [73,74].
48
Enrique Grande-Pulido et al.
A subset of CRC called CIMP, defined by its high proportion aberrant gene promoter methylation, differs from the CIMP-negative phenotype. However, it is not clear whether these tumours are really a molecular subset of tumours or whether they are a tumour group found in the tail of the normal distribution of all tumours that really present aberrant gene promoter methylation [75]. Most CIMPs present a BRAF mutation, and those that do not have a K-ras mutation [75]. These two mutations therefore appear to be mutually exclusive in these tumours, indicating that this EGFR/BRAF/RAS signalling pathway is critical to the development of these tumours. DNA methylation can affect transcription. Methylation can affect transcription by directly inhibiting transcription factors and methylation of promoters, but methylation of CpG islands can also activate transcription silencing, by altering binding proteins and histone deacetylases. These altered histones appear to cooperate with DNA methylation in the transcription of tumour-suppressor genes in carcinogenesis [76]. Epigenetic alterations that are commonly observed in CRC are likely to be involved in the development of tumours, activation of oncogenes, and deactivation of suppressor genes. However, the specific mechanisms that contribute to these alterations are currently under investigation, and we do not know the exact role they play in tumour initiation, growth, and maintenance [77].
Conclusion Despite of great efforts performed in last decades, the prognosis of patients with metastatic CRC is poor. Life expectancy is less than 2 years in most of randomized clinical trials. Both the deeper knowledge of the molecular mechanisms involved in pathogensis of CRC and the improvements of molecular engineering techniques to design new drugs against different targets make this field one of the most attractive in terms of research in solid tumors. Nowadays, there are several new targeted designed compounds under clinical evaluation in colorectal cancer. Systemic treatment of advanced solid tumors in general, and CRC properly, require an adequate individualization. We would need to select the best drug for the right patient according to molecular features that define each tumor. Therefore, we would not only need more active and safe drugs but also accurate biologic predictors of response to these drugs that will be determined by well established pathology knowledge from the molecular point of view.
Basic science behind colorectal cancer
49
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
Jemal, A., Siegel, R., Xu, J., and Ward, E. 2010, CA Cancer J. Clin., 60, 277. Pritchard, C.C. and Grady, W.M. 2011, Gut, 60, 116. Sartore-Bianchi, A., Bencardino, K., Cassingena, A., Venturini, F., Funaioli, C., Cipani, T., et al. 2010, Cancer Treat. Rev., 36 Suppl 3, S1. Hawkes, E. and Cunningham, D. 2010, Lancet Oncol., 11, 1020. Ferté, C., André, F., and Soria, J.C. 2010, Nat. Rev. Clin. Oncol., 7, 367. Cunningham, D., Atkin, W., Lenz, H.J., Lynch, H.T., Minsky, B., Nordlinger, B., and Starling, N. 2010, Lancet, 375, 1030. Lynch, H.T., Lynch, J.F., Lynch, P.M., and Attard, T. 2008, Fam. Cancer, 7, 27. Jensen, L.H., Dysager, L., Lindebjerg, J., Kolvra, S., Byriel, L., and Cruger, D.G. 2010, Eur. J. Cancer, 46, 1823. Rowley, P.T. 2005, Annu. Rev. Med., 56, 539. López-Villar, I., Ayala, R., Wesselink, J., Morillas, J.D., López, E., Marín, J.C., et al. J. 2010, BMC Cancer, 10, 408. Thakur, N., Reddy, D.N., Rao, G.V., Mohankrishna, P., Singh, L., and Chandak, G.R. 2006, BMC Med. Genet., 7, 73. Doxey, B.W., Kuwada, S.K., and Burt, R.W. 2005, Clin. Gastroenterol. Hepatol., 3, 633. Fearon, E.R. and Vogelstein, B. 1990, Cell, 61, 759. Jass, J.R. 2006, Crit. Rev. Oncog., 12, 273. Rajagopalan, H., Nowak, M.A., Vogelstein, B., and Lengauer, C. 2003, Nat. Rev. Cancer, 3, 695. Pino, M.S. and Chung, D.C. 2010, Gastroenterology, 138, 2059. Rao, C.V., Yamada, H.Y., Yao, Y., and Dai, W. 2009, Carcinogenesis, 30, 1469. Alberici, P. and Fodde, R. 2006, Genome Dyn., 1, 149. Caldwell, C.M. and Kaplan, K.B. 2009, Adv. Exp. Med. Biol., 656, 51. Aoki, K., Aoki, M., Sugai, M., Harada, N., Miyoshi, H., Tsukamoto, T., et al. 2007, Oncogene, 26, 3511. Judson, H., Stewart, A., Leslie, A., Pratt, N.R., Baty, D.U., Steele, R.J., and Carey, F.A. 2006, J. Pathol., 210, 344. Calistri, D., Rengucci, C., Seymour, I., Leonardi, E., Truini, M., Malacarne, D., et al. 2006, Cell. Oncol., 28, 161. Hadjihannas, M.V., Brückner, M., Jerchow, B., Birchmeier, W., Dietmaier, W., and Behrens, J. 2006, Proc. Natl. Acad. Sci. U. S. A., 103, 10747. Royce, S.G., Alsop, K., Haydon, A., Mead, L., Smith, L.D., Tesoriero, A.A., et al. 2010, Colorectal Dis., 12, 213. Wang, D., Mann, J.R., and DuBois, R.N. 2004, Cell Cycle, 3, 1512. Boland, C.R. and Goel, A. 2010, Gastroenterology, 138, 2073. Ogino, S. and Goel, A. 2008, J. Mol. Diagn., 10, 13. Blanes, A. and Diaz-Cano, S.J. 2006, World J. Gastroenterol., 12, 5932. Vilar, E. and Gruber, S.B. 2010, Nat. Rev. Clin. Oncol., 7, 153. Iacopetta, B., Kawakami, K., and Watanabe, T. 2008, Int. J. Clin. Oncol., 13, 498.
50
Enrique Grande-Pulido et al.
31. Dahlin, A.M., Palmqvist, R., Henriksson, M.L., Jacobsson, M., Eklof, V., Rutegard, J., Oberg, A., and Van Guelpen, B.R. 2010, Clin. Cancer Res., 16, 1845. 32. Kawasaki, T., Ohnishi, M., Nosho, K., Suemoto, Y., Kirkner, G.J., Meyerhardt, J.A., et al. 2008, Mod. Pathol., 21, 245. 33. Van Cutsem, E. 2006, Oncologist, 11, 1010. 34. Markman, B., Javier Ramos, F., Capdevila, J., and Tabernero, J. 2010, Adv. Clin. Chem., 51, 71. 35. Ciardiello, F. and Tortora, G. 2001, Clin. Cancer Res., 7, 2958. 36. Siddiqui, A.D. and Piperdi, B. 2010, Ann. Surg. Oncol., 17, 1168. 37. Silvestris, N., Tommasi, S., Petriella, D., Santini, D., Fistola, E., Russo, A., et al. 2009, Oncology, 77 Suppl. 1, 69. 38. Lee, H.Y., Crawley, S., Hokari, R., Kwon, S., and Kim, Y.S. 2010, Int. J. Oncol., 36, 941. 39. Xiong, H., Zhang, Z.G., Tian, X.Q., Sun, D.F., Liang, Q.C., Zhang, Y.J., et al. 2008, Neoplasia, 10, 287. 40. Markman, B., Capdevila, J., Elez, E., and Tabernero, J. 2009, Immunotherapy, 1, 965. 41. Vincenzi, B., Zoccoli, A., Pantano, F., Venditti, O., and Galluzzo, S. 2010, Curr. Cancer Drug Targets, 10, 80. 42. Keating, G.M. 2010, Drugs, 70, 1059. 43. Rasheed, S., McDonald, P.J., Northover, J.M., and Guenther, T. 2008, Pathol. Res. Pract., 204, 501. 44. Kerbel, R.S. 2008, N. Engl. J. Med., 358, 2039. 45. Kerbel, R.S. 2006, Science, 312, 1171. 46. Rmali, K.A., Puntis, M.C., and Jiang, W.G. 2007, Colorectal Dis., 9, 3. 47. Mizukami, Y., Kohgo, Y., and Chung, D.C. 2007, Clin. Cancer Res., 13, 5670. 48. Cao, Y., Sonveaux, P., Liu, S., Zhao, Y., Mi, J., Clary, B.M., Li, C.Y., Kontos, C.D., and Dewhirst, M.W. 2007, Cancer Res., 67, 3835. 49. Liu, W., Jung, Y.D., Ahmad, S.A., McCarty, M.F., Stoeltzing, O., Reinmuth, N., et al. 2004, Br. J. Cancer, 90, 1620. 50. Jubb, A.M., Turley, H., Moeller, H.C., Steers, G., Han, C., Li, J.L., et al. 2009, Br. J. Cancer, 101, 1749. 51. Yan, M. and Plowman, G.D. 2007, Clin. Cancer Res., 13, 7243. 52. Fens, M.H., Storm, G., and Schiffelers, R.M. 2010, Expert Opin. Investig. Drugs, 19, 1321. 53. Landi, S., Moreno, V., Gioia-Patricola, L., Guino, E., Navarro, M., de Oca, J., Capella, G., and Canzian, F. 2003, Cancer Res., 63, 3560. 54. Iwasaki, J. and Nihira, S. 2009, Jpn. J. Clin. Oncol., 39, 543. 55. Shaker, A. and Rubin, D.C. 2010, Transl. Res., 156, 180. 56. Simon-Assmann, P., Turck, N., Sidhoum-Jenny, M., Gradwohl, G., and Kedinger, M. 2007, Cell Biol. Toxicol., 23, 241. 57. Vermeulen, L., De Sousa, E.M.F., van der Heijden, M., Cameron, K., de Jong, J.H., Borovski, T., et al. 2010, Nat. Cell Biol., 12, 468. 58. Le, N.H., Franken, P., and Fodde, R. 2008, Br. J. Cancer, 98, 1886.
Basic science behind colorectal cancer
51
59. Su, X., Ye, J., Hsueh, E.C., Zhang, Y., Hoft, D.F., and Peng, G. 2010, J. Immunol., 184, 1630. 60. Nosho, K., Baba, Y., Tanaka, N., Shima, K., Hayashi, M., Meyerhardt, J.A., et al. 2010, J. Pathol., 222, 350. 61. Kaidi, A., Williams, A.C., and Paraskeva, C. 2007, Nat. Cell Biol., 9, 210. 62. Todaro, M., Francipane, M.G., Medema, J.P., and Stassi, G. 2010, Gastroenterology, 138, 2151. 63. Yeung, T.M. and Mortensen, N.J. 2009, Dis. Colon Rectum, 52, 1788. 64. Hynes, M.J., Huang, K.M., and Huang, E.H. 2009, Vet. Pathol., 46, 819. 65. Fodde, R. and Brabletz, T. 2007, Curr. Opin. Cell Biol., 19, 150. 66. Pohl, A., Lurje, G., Kahn, M., and Lenz, H.J. 2008, Clin. Colorectal Cancer, 7, 92. 67. Fabrizi, E., di Martino, S., Pelacchi, F., and Ricci-Vitiani, L. 2010, World J. Gastroenterol., 16, 3871. 68. Todaro, M., Francipane, M.G., Medema, J.P., and Stassi, G. 2010, Gastroenterology, 138, 2151. 69. Cui, L., Ohuchida, K., Mizumoto, K., Moriyama, T., Onimaru, M., Nakata, K., et al. 2010, PLoS ONE, 5, e12121. 70. Lugli, A., Iezzi, G., Hostettler, I., Muraro, M.G., Mele, V., Tornillo, L., et al. 2010, Br. J. Cancer, 103, 382. 71. Liu, M. and Chen, H. 2010, J. Genet. Genomics, 37, 347. 72. Lichter, P. 2008, Int. J. Cancer, 123, ix. 73. Nystrom, M. and Mutanen, M. 2009, World J. Gastroenterol., 15, 257. 74. Timp, W., Levchenko, A., and Feinberg, A.P. 2009, Cell Cycle, 8, 383. 75. Hinoue, T., Weisenberger, D.J., Pan, F., Campan, M., Kim, M., Young, J., et al. 2009, PLoS ONE, 4, e8357. 76. Kaneda, A. and Yagi, K. 2011, Cancer Sci., 102, 18. 77. Toyota, M. and Suzuki, H. 2010, Adv. Genet., 70, 309.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 53-73 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
3. Imaging of colorectal cancer: Established techniques and emerging modalities 1
Rachel E. Hyland1 and Fahmid U. Chowdhury2 Consultant Radiologist, The General Infirmary at Leeds, Leeds, UK 2 Consultant Radiologist and Nuclear Medicine Physician St James’s University Hospital, Leeds, UK
Abstract. Colorectal cancer (CRC) is an important health problem worldwide with an overall 5-year survival of approximately 50%, this being heavily dependent on the stage of disease at presentation, with survival varying from 83% for early-stage disease to 3% for patients presenting with metastatic tumour. Therefore, accurate staging and restaging of disease using different imaging techniques is of paramount importance not only for directing the most appropriate therapeutic options but also for indicating prognosis and outcome of this relevant disease.
Introduction Colorectal cancer (CRC) is the third most common cancer in the UK, with over 38, 000 new cases diagnosed in 2008 [1]. The overall 5-year survival is approximately 50%, but this is heavily dependent on the stage of Correspondence/Reprint request: Dr. Fahmid U. Chowdhury, MRCP(UK) FRCR, Consultant Radiologist and Nuclear Medicine Physician, St James’s University Hospital, Leeds, UK E-mail: fahmid.chowdhury@leedsth.nhs.uk
54
Rachel E. Hyland & Fahmid U. Chowdhury
disease at presentation, with survival varying from 83% for early-stage disease to 3% for patients presenting with metastatic tumour [2]. In the European Union, there were more than 333,000 new cases of bowel cancer diagnosed in 2008, and the American Cancer Society estimated that in the same year, about 148,810 people would be diagnosed with colorectal cancer in the United States and that about 49,960 people would die of the disease [3-6]. Patients with CRC in some European countries and in the US seem to fare better than patients in the UK, with 5-year survival of 57-66%, which may partly be explained by earlier diagnosis due to more opportunistic use of screening facilities [3-6]. Curative treatment of CRC involves surgical resection of the primary tumour, with a role for adjuvant chemotherapy in patients with nodal involvement (stage III disease) and neoadjuvant chemoradiotherapy in patients with rectal tumours. Furthermore, ongoing therapeutic refinement may result in more patients with earlier (stage II) disease undergoing adjuvant treatment, and there is an ever-increasing role for surgical intervention in limited metastatic disease, with favourable 5-year survival of 26-64% in patients undergoing hepatic or pulmonary metastectomy [7-11]. Accurate staging and restaging of disease is of paramount importance not only for directing the most appropriate therapeutic options but also for indicating prognosis and outcome. A multimodality approach is required in the investigation of CRC involving faecal occult blood testing, colonoscopy, cross-sectional imaging with computed tomography (CT) and magnetic resonance imaging (MRI), contrast studies, ultrasound, and hybrid imaging with positron emission tomography/ computed tomography (PET/CT). This chapter will examine the roles of the major imaging techniques in the evaluation of CRC, with particular emphasis on the emerging uses of combined functional and morphological imaging, especially integrated PET/CT.
Established imaging modalities in the evaluation of CRC Contrast studies Barium enema was the traditional investigation of choice to rule out a synchronous colorectal tumour, said to occur in 4-5% of patients [12]. There is growing evidence that the sensitivity of barium studies for detection of early cancers and adenomas is inferior to alternative tests, such as colonoscopy and CT colonography (virtual colonography) (Figure 1). In a series of 580 patients undergoing both barium enema and colonoscopy
Imaging of colorectal cancer
55
Figure 1. Annular carcinoma of the sigmoid colon. (a) Double contrast barium enema showing annular stricture in the sigmoid colon (black arrow), causing significant narrowing to the lumen. (b) Axial CT from the same patient, demonstrating circumferential thickening (white arrow) of the proximal sigmoid colon, with stranding of the pericolonic fat in keeping with T3 tumour. Note the prominent peritumoural vessels in keeping with vascular invasion (arrowhead).
studies, the American National Polyp study found the sensitivity for detection of adenomas > 10 mm to be just 48% for barium enema [13]. The use of barium enemas in the diagnosis of CRC is declining, with single contrast studies serving a more useful role in the assessment of post-operative complications, such as the evaluation of anastomotic integrity. Ultrasound Endoscopic rectal ultrasound (ERUS) provides high-resolution images of the rectal wall, which can be visualised as 5 sonographic layers. The main disadvantage of ERUS is the inability to image the mesorectal fascia, which is extremely reliably demonstrated by MRI. Given the importance of the relationship of the tumour to the mesorectal fascia in planning treatment, the role of ERUS is best limited to T-staging early rectal tumours, as the sensitivity of T-stage has been reported to be as high as 96% [14]. The technique is less reliable in assessment of lymph nodes, because of difficulty accessing nodes along the draining vascular pedicle and the pelvic sidewall, with sensitivity ranging from 64-83%[14]. Ultrasonic features suggestive of nodal malignancy include size > 10 mm, rounded shape, hypoechoic centre, and a smooth border [15].
56
Rachel E. Hyland & Fahmid U. Chowdhury
Computed tomography (CT) CT remains the mainstay technique for staging and re-staging patients with CRC. The typical CT examination includes a volume acquisition through the chest, abdomen and pelvis, with intravenous iodinated contrast. Abdominal images are obtained in the portal venous phase (60-90 seconds after contrast injection) to optimise detection of liver metastases. Bowel distension is achieved by oral ingestion of 1000-1500 ml of positive or negative (i.e. water) contrast agent. CT colonography requires air insufflation per rectum, dual patient positioning and anti-spasmodic to optimise bowel distension. CT has traditionally been used for identification of the location and size of the tumour, demonstration of local extension, detection of distant metastases or complications such as perforation, obstruction or pericolic abscess formation. It is possible with the advent of multidetector CT (MDCT) to demonstrate poor prognostic factors such as depth of extramural spread, lymph node involvement, extramural vascular invasion and peritoneal infiltration [16-17]. Ongoing studies (e.g. FOXTROT) are evaluating the benefit of identifying these patients pre-operatively and down-staging them with neoadjuvant chemoradiotherapy [18]. Tumour is typically seen as circumferential thickening of the colonic wall. The presence of tumour extension beyond the muscularis propria is demonstrated by stranding of fat, spiculation and spread of tumour beyond the serosa and pericolic fat (Figure 2). Identification of T4 disease relies on loss of fat planes between the tumour and adjacent structures indicative of invasion, penetration of peritonealised surface or perforation. A recent metaanalysis found the sensitivity and specificity of differentiating between early (T1/T2) and locally advanced (T3/T4) disease at staging CT was 86% and 78% respectively [19]. The best results were demonstrated by the studies that utilized slice thicknesses of < 5 mm using spiral CT or MDCT, and rectal insufflation techniques. Pooled results for detecting tumour invasion in studies utilizing MDCT is much higher (sensitivity 93%, specificity 86%) [19]. The use of thin-slice reconstruction and endoluminal views has led to CT colonography being used for primary diagnosis as well as staging, with sensitivity > 90% for the detection of polyps > 10 mm and colonic tumours in asymptomatic adults [20]. Detection of involved lymph nodes by imaging is always limited by the inability to detect micro-metastases. Current CT criteria used to diagnose involved nodes include size > 10 mm, or a group of 3 or more nodes. Any enhancing nodule along the draining vascular pedicle is suspicious for involvement, but should be discriminated from peritoneal or omental nodules,
Imaging of colorectal cancer
57
Figure 2. CT colonography of locally advanced recto-sigmoid colon cancer. 2(a) CT colonography. Axial CT showing eccentric thickening of the rectosigmoid colon, with tumour clearly visible beyond the bowel wall (arrow). 2(b) Endoluminal view from CTC examination showing eccentric filling defect arising from the colonic wall.
which indicates T4 or metastatic disease. The same meta-analysis found the overall CT sensitivity and specificity for the presence of nodal disease was 70% and 78% respectively [19]. Magnetic resonance imaging (MRI) Advances in preoperative therapies for rectal cancer require an accurate preoperative staging technique to select patients who are most likely to benefit from neoadjuvant treatment. Several studies have demonstrated the ability of MRI to accurately stage rectal cancer [21-25]. The Magnetic Resonance Imaging and Rectal Cancer European Equivalence Study (MERCURY) study showed that highresolution MRI can accurately predict involvement of the surgical resection margin and extramural tumour invasion [25]. Preoperative staging by MRI is currently considered the reference standard in the management of rectal cancer. T2-weighted sagittal and axial images are obtained, using a highresolution axial sequence through the tumour. Imaging in a coronal plane is used in the assessment of low rectal tumours to show the relationship to the levator muscles. MR imaging can assess the relationship of the tumour to surrounding structures and the bowel wall. On T2-weighted imaging, perirectal fat is seen as high signal intensity surrounding the low signal of the muscularis propria (Figure 3). The mesorectal fascia is seen as a thin, low-signal
58
Rachel E. Hyland & Fahmid U. Chowdhury
Figure 3. High resolution axial MRI showing abnormal thickening of the lower rectum, but preservation of the muscularis propria (arrow) of the bowel wall, in keeping with T1/2 tumour.
layer encasing the perirectal fat and rectum. The mesorectal fascia is a critical structure that defines the surgical excision plane in anterior total mesorectal resection. MRI diagnosis of a T3 lesion is based on the presence of tumour extending into the perirectal fat. Stage T4 tumours are those that have perforated, or invaded the peritoneum, an adjacent organ or structure (eg levator ani). Extramural vascular invasion is an adverse prognostic factor, seen in up to 50% of surgical specimens, and can be identified as tumour expanding perirectal and pelvic vessels. Careful evaluation should be undertaken as to whether these threaten the circumferential resection margin (Figure 4). There is growing interest in the use of diffusion-weighted imaging (DWI) in the assessment of CRC. Recent research has shown that DWI may aid detection of small tumours and define tumour extent, particularly where there is peri-tumoural inflammation [26,27]. Research is also evaluating tumour response to chemoradiotherapy, both to try to predict early nonresponse, and complete response at the end of treatment [28-31]. Similarly, in the assessment of CRC liver metastases, high pre-treatment diffusion coefficient values have been shown to be a predictor of a poor response (Figure 5) [32]. Similar to CT staging, nodal staging by MRI traditionally relies on the size of the nodes, but criteria based on morphology, such as the outline of the node and features of signal intensity, have been shown to be more reliable than size alone [33,34]. The use of ultrasmall superparamagnetic iron oxide
Imaging of colorectal cancer
59
Figure 4. Same patient as figure 2. 4(a) Sagittal MRI showing T3 rectosigmoid tumour, with tumour nodularity extending posteriorly, with vascular invasion (arrow). 4(b) Axial high resolution MRI showing eccentric thickening of the rectum, with the advancing border inferolaterally, at the 5 oâ&#x20AC;&#x2122;clock position, with extra-mural tumour (white arrow) coming within 5 mm of the mesorectal fascia (black arrow).
Figure 5. Patient with colorectal liver metastases. Two hepatic metastases are clearly shown as regions of restricted diffusion on the diffusion weighted imaging (DWI). These are more difficult to appreciate on other sequences. (a) DWI, (b) corresponding apparent diffusion coefficient (ADC) map, (c) gadolinium-enhanced T1, and (d) T2 weighted images. (Images courtesy of Dr Andrew Scarsbrook, St Jamesâ&#x20AC;&#x2122;s University Hospital, Leeds, UK.)
60
Rachel E. Hyland & Fahmid U. Chowdhury
particles (USPIO) as a contrast medium may improve the detection of metastatic lymph nodes by MRI. Following intravenous administration, the particles are taken up by nodal macrophages. In normal or reactive lymph nodes this leads to decreased signal intensity on T2- and T2*-weighted images due to susceptibility effects [35]. A recent study looking at the diagnostic accuracy of the pattern of nodal uptake of USPIO found a higher diagnostic specificity (96% USPIO vs. 75% MRI morphology alone), but the same sensitivity (65%) as morphologic findings in pathologically matched mesorectal lymph nodes [36]. USPIO agents are not yet available for routine clinical use.
Combined metabolic and anatomic imaging of colorectal cancer: PET/CT PET/CT: Principles and technique For most oncological indications, PET/CT is currently performed using 2-[ F] labelled- fluoro-2-deoxy-D- glucose (FDG), a non-physiological glucose analogue that provides functional information about tumour metabolism by demonstrating augmented glycolysis in cancer cells compared with normal tissues. However, FDG uptake is not tumour-specific, as metabolically active benign processes such as infection and inflammation show increased glucose utilization [37-38]. FDG PET is also limited by its lack of anatomical landmarks and relatively poor spatial resolution, and, although CT can accurately show anatomical changes in malignancy, it does not provide information on the functional significance of morphological abnormalities. Combining the two techniques in a single examination using an integrated PET/CT scanner provides co-registered PET and CT images that allow precise localisation of functional abnormalities. This ability of PET/CT to combine the benefits of anatomical and metabolic imaging confers a distinct diagnostic advantage by facilitating more accurate staging and restaging of malignancy with significant impact on patient management [39-41]. Patient preparation includes a 4-6 hour fast, with an optimal blood glucose level of less than 10 mmol/l. Typically, a patient is injected with 370400 MBq (10-11 mCi) FDG, and, after a 60-minute uptake period, a scout CT is obtained from the skull base to upper thigh, with subsequent non-contrastenhanced multidetector CT. The PET acquisition follows immediately, with no change in patient position, and covers the same range, with 2-4 minutes per bed position. The patient maintains normal tidal respiration throughout 18
Imaging of colorectal cancer
61
the study, which lasts for approximately 30 minutes. The CT information is used to provide attenuation correction for the PET data and for anatomical co-registration. Standard multimodality workstations enable viewing of fused PET/CT images within minutes of completion of the study. The PET data can be analysed qualitatively and semi-quantitatively, by measuring the standardised uptake value, which is usually expressed as its maximum (SUVmax). The SUV is defined as the ratio of activity within the tissue (Bq/ml) and decay-corrected total activity injected divided by body weight (Bq/g).
Initial staging Although it has been shown that 95-100% of known intraluminal primary colon cancers are visible on FDG PET, false-negatives may be encountered with mucinous tumours and small tumour foci in tubulovillous adenomas [42-43]. Local T-staging can be conventionally performed with multidetector CT and MRI, with accuracy rates of > 80% [42-45]. Moreover, defining the anatomical relationship of the primary lesion to the mesorectal fascia is the single most important factor in the local staging of rectal tumours, and this is best demonstrated on MRI [46]. Lymph node involvement is an important prognostic indicator in CRC, with CT and MRI demonstrating a relatively low sensitivity of 55-66% for detection of nodal disease [45]. Metabolic changes often precede alteration in morphology, allowing FDG PET to demonstrate disease involvement in normal-sized nodes. However, practical limitations in spatial resolution, and inability of PET to detect disease in nodes with a small tumour burden, leads to an overall sensitivity of FDG PET of only 29-37% for local nodal staging, although the specificity is significantly higher at 87% [42, 47]. Use of FDG PET/CT in initial staging of CRC is reserved for high-risk patients (e.g. those with raised carcinoembryonic antigen (CEA) levels >10 ng/ml, locally advanced disease or equivocal findings on conventional imaging), where it has been shown that incorporating FDG PET/CT at the outset can lead to a management change in 18-24% of patients, by demonstrating unsuspected metastatic disease or clarifying the nature of indeterminate lesions [48-49]. It should be emphasised, however, that FDG PET has a low sensitivity for detection of hepatic metastases <10 mm in size compared with contrast-enhanced MRI (67% vs. 81%) [50-51]. Given these limitations, routine use of FDG PET/CT in the initial staging of CRC is not justified, and its use is reserved for problem-solving and evaluation of high-risk patients (Figure 6) (Table 1).
62
Rachel E. Hyland & Fahmid U. Chowdhury
Figure 6. Staging PET/CT in a patient with advanced rectal cancer. PET maximum intensity projection (MIP) and selected axial fused PET/CT images show (a) an extremely hypermetabolic locally advanced rectal cancer with local nodal involvement (arrowhead), (b) FDG avid retroperitoneal nodes, (c) multiple necrotic liver metastases and (d) a left adrenal metastasis (arrowhead). Table 1. Established indications for PET/CT in colorectal cancer.
Imaging of colorectal cancer
63
Incidental detection of colorectal tumours and PET/CT in screening Physiological uptake of FDG in the bowel is variable and unpredictable, but usually takes the form of a diffuse or segmental pattern with no corresponding morphological abnormality on the CT component of the PET/CT study. Focal FDG uptake in the colon, which is seen in up to 2-3% of all patients undergoing PET/CT scans, is more worrisome, as at least 40-60% of focal bowel uptake represents either a premalignant adenomatous polyp or early bowel cancer, especially if there is corresponding morphological abnormality, and hence requires further evaluation with colonoscopy (Figure 7) [52-53]. The evaluation of PET in screening for early colonic tumours has proved disappointing, as demonstrated by Yasuda et al in a study of 110 patients who underwent colonoscopy and FDG PET [54]. Fifty-nine adenomatous polyps, ranging in size from 5-30 mm, were found in 30 subjects by total colonoscopy. PET findings were positive for only 14 of the 59 adenomas (24%), although the sensitivity was considerably higher for
Figure 7. Incidental detection of metastatic colorectal cancer on staging PET/CT in a patient with squamous cell cancer of the oesophagus. PET maximum intensity projection (MIP) shows (a) an FDG avid primary oesophageal tumour, (b) focal uptake in the right iliac fossa and (c) FDG avid lesions in the liver. Axial CT, PET and fused PET/CT images show a colonic tumour, which had failed to undergo detection on prior CT, with local nodal uptake (arrowhead).
64
Rachel E. Hyland & Fahmid U. Chowdhury
adenomas > 13mm in size (90%). It is reasonable to conclude that FDG PET/CT does not have sufficient accuracy for the detection of early precursors of colorectal cancer, and this would also have to be considered in the context of the significant radiation burden from PET/CT performed in asymptomatic patients with no known malignancy, ranging from 14-32 mSv in adult patients depending on the CT protocol [55]. Detection of recurrence Colorectal tumour recurrence occurs in approximately 30% of patients within the first 2 years of surgery, and a rise in serum carcinoembryonic antigen (CEA) is the earliest indicator of recurrence in 60% [56]. A rise in CEA can precede clinical symptoms by a median lead-time of 4.5-8 months [57]. In patients with raised CEA and negative or equivocal conventional imaging, FDG PET/CT sensitivity for identifying the site of recurrence has been shown to be as high as 67-75% [58]. Importantly, FDG PET/CT also has a very high negative predictive value of 95-100% in this situation [58-59].
Figure 8. Detection of recurrence in a patient with rising CEA and negative conventional imaging. PET maximum intensity projection (MIP) demonstrates a focus of markedly avid FDG uptake in the right iliac fossa (arrow). Axial CT, PET and fused PET/CT images show that this localizes to a site of previously resected recurrence in the right iliopsoas muscle (arrow). Further metastasis surgery was performed.
Imaging of colorectal cancer
65
FDG PET/CT should be considered in patients with suspected recurrence presenting with a rising CEA and inconclusive conventional tests, to help confirm and localise the site of recurrence and facilitate appropriate further management (Figure 8). Most patients develop a fibrotic mass in the pre-sacral space following surgery and/or radiotherapy for treatment of rectal cancer, and it can be difficult to distinguish local tumour recurrence from benign post-therapy change using conventional imaging. Local recurrence occurs in 20-30% of these patients, with the majority recurring within the first 3 years of surgery [60]. Accurate imaging of recurrent disease is vital, as it is now clear that curative revision surgery is possible for a proportion of these patients [61]. In suspected local recurrence, FDG PET/CT can distinguish between benign and malignant pre-sacral abnormalities with high sensitivity and specificity
a
b
Figure 9. Detection of extra-hepatic disease prior to planned liver resection. PET maximum intensity projection (MIP) shows (a) hepatic metastasis and (b) focal uptake at the right pulmonary hilum. Panel of axial CT, fused PET/CT and PET images confirm liver recurrence at the site of prior metastatectomy (a) and a moderately FDG avid right hilar node (b), which was subsequently confirmed as a further site of metastatic disease.
66
Rachel E. Hyland & Fahmid U. Chowdhury
(100% and 96%, respectively) with an impact on clinical management in 47% of patients [60]. However, false-positive FDG uptake may persist in inflammatory or infective post-surgical complications, and findings should always be interpreted cautiously in the context of a protracted or complicated post-surgical recovery [62]. Recurrence of mucinous tumour may give a false-negative result with FDG PET. Direct comparison between PET and MR for detection of local recurrence in rectal cancer has been performed in small groups of patients, showing that the techniques are complementary and have similar accuracy [63]. Routine imaging follow-up or surveillance of treated patients with CRC is performed with conventional CT. A recent study randomised 130 patients to either being followed-up with a conventional work-up or with FDG PET [64]. Recurrences were detected after a shorter interval (12.1 vs. 15.4 months; P=0.01) in the PET group, in whom recurrences were also more frequently cured by surgery (R0). This raises an interesting conundrum as to whether PET/CT may in future replace conventional CT as the technique of choice in surveillance of treated CRC patients as this may detect recurrence and allow curative treatment at an earlier stage. However, it is difficult to justify such an approach at present without more robust and reproducible evidence that this would be a cost-effective and clinically worthwhile change in imaging strategy. Re-staging of patients with potentially limited, operable metastatic disease The liver is the commonest site of colorectal metastases. Between 25-30% of patients have hepatic metastases at presentation, and recurrent disease is seen in up to 30% of patients within 2 years of initial resection, which in the majority will manifest as liver metastases [56, 65]. Surgical resection of limited metastatic disease with a curative intent is a well-established treatment strategy, and recent studies have demonstrated favourable long-term outcomes [7-11]. At least 17% of patients have a >10-year disease free survival after liver metastectomy, with 5-year survival of 26-50% [7-10]. Five-year survival of up to 64% has also been demonstrated following resection of pulmonary metastases [11]. Appropriate patient selection is crucial, however, as 10-50% of patients with disease apparently limited to the liver on conventional imaging have historically been shown to have inoperable findings at laparotomy [66]. Performing PET in patients with colorectal liver metastases can reduce futile laparotomy rates from 45% to 28%, largely by identifying sites of unsuspected metastatic disease (Figures 9-10) [67]. In a recent randomized controlled trial, FDG PET was added to conventional diagnostic work-up in patients with colorectal liver metastases,
Imaging of colorectal cancer
67
Figure 10. Improving the accuracy of re-staging with PET/CT. Three panels of axial CT, fused PET/CT and PET images in a patient under consideration for pulmonary metastatectomy. (a) FDG avid right lower lobe pulmonary metastases. (b) Small liver metastasis (arrow). (c) Local recurrence in the presacral space and within the primary surgical field (arrowheads). None of the extra-pulmonary disease had been detected on prior conventional imaging.
with 75 patients included in each arm. The increased diagnostic performance of PET led to a 38% reduction in the rate of futile laparotomy [68]. This is in alignment with other data that has shown that FDG PET/CT is more sensitive than CT at identifying extra-hepatic disease with a sensitivity of 63% versus 25% for CT alone, thereby avoiding unnecessary surgery in 20-40% of patients [69-71]. There have also been several analyses of the cost-effectiveness of incorporating FDG PET/CT into the pre-operative staging of patients with potentially operable metastatic recurrence. These studies have shown a wide variation in cost savings of between $429-$5, 269 per patient [72-75]. They have all concluded that the cost savings from averted futile invasive surgical procedures readily offset the added cost of PET/CT, with direct benefits to the individual patient and optimal use of resources within the health care system.
68
Rachel E. Hyland & Fahmid U. Chowdhury
Assessment of therapy response The limitations of conventional morphological imaging in assessing therapy response to neoadjuvant, adjuvant or palliative treatment are well recognised [76]. There are preliminary data to suggest that assessment of metabolic response by FDG PET can allow prediction of long-term outcome by demonstrating a reduction in tumour SUV [77-78]. Detection of early metabolic response may become increasingly important in patients treated with novel drugs and therapeutic interventions that are costly and/or may lead to unnecessary toxicity in non-responders [79]. Conventional imaging also faces considerable challenges in revealing residual or recurrent tumour at sites treated by radiofrequency ablation. In theory, ablated liver cells are no longer able to take up FDG, and PET/CT can aid in the detection of recurrence after hepatic ablation or metastectomy, when altered anatomy or scarring can cause interpretative difficulties [80-81]. Radio-embolisation with yttrium- 90 microspheres injected directly into the hepatic artery after percutaneous catheterisation is an emerging technique in the management of inoperable CRC liver metastases, and PET/CT has also shown promise in improving the accuracy of therapy response assessment in these patients [82].
Future developments Modifications in scanning protocols that are already technically feasible such as intravenous contrast-enhanced PET/CT, motion-corrected ‘4D PET/CT’ acquisition with respiratory gating (for pulmonary nodules and lesions at the superior liver surface), and PET/CT colonography may gain increasing clinical acceptance if they can be shown to provide greater accuracy in diagnosis and staging and allow convenient ‘one-stop’ imaging [83-85]. Newer targeted oncological therapies aimed at oncogenic molecular pathways may benefit from novel non-FDG PET radiotracers that image intracellular molecular processes, other than glycolysis. For instance, [18F]fluorothymidine (FLT) is a marker of malignant cell proliferation that relies on thymidine kinase activity and can be used to detect cellular amino acid utilisation and DNA synthesis [86]. [18F]-Fluoromisonidasole (F-MISO) assesses the degree of tumour hypoxia and may allow response evaluation to anti-angiogenesis agents [87]. It is possible that early functional assessment of tumour response to novel therapies will lead to increasing use of PET/CT in these patients, although these novel non-FDG tracers have yet to enter into routine clinical practice. There may be an emerging role for PET/CT in intensity-modulated radiotherapy planning by more accurately depicting gross tumour volume. FDG PET has been shown to influence the
Imaging of colorectal cancer
69
determination of target volumes for non-small cell lung cancer and oesophageal tumours, but there is less evidence for its role in CRC [88-89]. In rectal cancer, incorporation of PET/CT in the planning process may help to improve standardisation of target volume delineation and minimise the radiation dose to surrounding normal structures and consequently reduce radiotherapy-induced morbidity [90].
Conclusion Accurate imaging of colorectal cancer requires a coordinated multimodality approach. While conventional US, CT and MRI continue to provide the mainstay of imaging in these patients, there is a rapidly emerging role for more novel methods that combine functional evaluation of tumour with anatomic imaging, such as integrated PET/CT and functional MRI.
References 1.
2.
3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
Cancer Research UK. Cancer statistics key fact- bowel cancer. http://info.cancerresearchuk.org/prod_consump/groups/cr_common/@nre/@sta/d ocuments/generalcontent/crukmig_1000ast-2806.pdf (accessed 30 October 2010). National Institute for Health and Clinical Excellence. Improving outcomes in colorectal cancers. Manual update. London: National Institute for Health and Clinical Excellence, 2004. Berrino, T, Angelis, R, Sant, M, Rosso, S.; Lasota, M.B, Coebergh, J.W.; et al. 2007, Lancet Oncol, 8, 773. Verdecchia, A, Francisci, S, Brenner, H, Gatta, G, Micheli, A, Mangone, L, et al. 2007, Lancet Oncol, 8, 784. Oâ&#x20AC;&#x2122;Connell, JB, Maggard, MA, and Ko, CY. 2004, J Natl Cancer Inst, 96,1420. American Cancer Society. Colorectal Cancer Facts & Figures 2008-2010. Atlanta:American Cancer Society, 2008. Carpizo, DR, and D'Angelica, M. 2009, Lancet Oncol, 10, 801. Tomlinson JS, Jamagin WR, De Matteo RP, Fong, Y, Kornprat, P, Gonen, M, et al. 2007, J Clin Oncol, 25, 4575. McLoughlin, JM, Jensen, EH, and Malafa, M. 2006, Cancer Control, 13, 32. Fong, Y, Fortner, J, Sun, RL, Brennan, MF, and Blumgart, LH. 1999, Ann Surg., 230, 309. Ashley, AC, Deschamps, C, and Alberts, SR. 2006, Clin Colorectal Cancer, 6, 32. Finan, PJ, Ritchie, JK, and Hawley, PR. 1987, Br J Surg., 74, 945. Winawer, SJ, Stewart, ET, Zauber, Bond, JH, Ansel, H, Waye, JD AG, et al. 2000, N Engl J Med, 342, 1766. Savides, TJ, and Master, SS. 2002, Gastrointest Endosc., 56, S12.
70
Rachel E. Hyland & Fahmid U. Chowdhury
15. Catalano, MF, Sivak, MV Jr, Rice, T, Gragg, LA, and Van Dam, J. 1994, Gastrointest Endosc., 40, 442. 16. Burton, S, Brown, G, Bees, N, Norman, A, Biedrzycki, O, Arnaout, K, et al. 2008, Br J Radiol., 81, 10. 17. Smith, NJ, Bees, N, Barbachano, Y, Norman, AR, Swift, RY, and Brown, G. 2007, Br J Cancer, 96, 1030. 18. Gray, PR, Morton, MD, and Seymour, PM. FOxTROT fluoropyrimidine, oxaliplatin & targeted receptor pre-operative therapy for colon cancer. A randomised trial assessing whether preoperative chemotherapy and/or an antiEGFR monoclonal antibody improve outcome in high-risk operable colon cancer. UK Clinical Research Network; 2007. 19. Dighe, S, Purkayastha, S, Swift, I, Tekkis, PP, Darzi, A, A´Hern, R, and Brown, G. 2010, Clin Radiol., 65, 708. 20. Halligan, S, Altman, DG, Taylor, SA, Mallett, S, Deeks, JJ, Bartram, CI, and Atkin, W. 2005, Radiology, 237, 893. 21. Blomqvist, L, Rubio, C, Holm, T, Machado, M, and Hindmarsh, T. 1999, Br J Radiol., 72, 18. 22. Brown, G, Davies, S, Williams, GT, Bourne, MW, Newcombe, RG, Radcliffe, AG et al. 2004, Br J Cancer, 91, 23. 23. Brown, G, Kirkham, A, Williams, GT, Bourne, M, Radcliffe, AG, Sayman, J, et al. 2004, Am J Roentgenol., 182, 431. 24. Brown, G, Radcliffe, AG, Newcombe, RG, Dallimore, NS, Bourne, MW, and Williams, GT. 2003, Br J Surg., 90, 355. 25. MERCURY Study Group. 2006, BMJ, 333, 779. 26. Figueiras, RG, Goh, V, Padhani, AR, Naveira, AB, Caamaño, AG, and Martin, CV. 2010, Am J Roentgenol., 195, 54. 27. Ichikawa, T, Erturk, SM, Motosugi, U, Sou, H, Lino, H, Araki, T, et al. 2006, Am J Roentgenol., 187, 181. 28. Hein, PA, Kremser, C, Judmaier, W, Griebel, J, Pfeiffer, KP, Kreczy, A, et al. 2003, Eur J Radiol., 45, 214. 29. Cui, Y, Zhang, XP, Sun, YS, Tang, L, and Shen, L. 2008, Radiology, 248, 894. 30. Dzik-Jurasz, A, Domenig, C, George, M, Wolber, J, Padhani, A, Brown, G, and Doran, S. 2002, Lancet, 360, 307. 31. DeVries, AF, Kremser, C, Hein, PA, Griebel, J, Kreczy, A, Ofner, D, et al. 2003, Int J Radiat Oncol Biol Phys., 56, 958. 32. Koh, DM, Scurr, E, Collins, D, Kanber, B, Norman, A, Leach, MO, et al. 2007, Am J Roentgenol., 188, 1001. 33. Brown, G, Richards, CJ, Bourne, MW, Newcombe, RG, Radcliffe, AG, Dallimore, NS, et al. 2003, Radiology, 227, 371. 34. Kim, JH, Beets, GL, Kim, MJ, Kessels, AG, and Beets-Tan, RG. 2004, Eur J Radiol., 52, 78. 35. Taupitz, M, Schmitz, S, and Hamm, B. 2003, Rofo, 175, 752. 36. Koh, DM, George, C, Temple, L, Collins, DJ, Toomey, P, Raja, A, et al. 2010, Am J Roentgenol., 194, 505.
Imaging of colorectal cancer
71
37. Metser, U, Miller, E, Lerman, H, and Evan-Sepir, E. 2007, Am J Roentgenol., 189, 1203. 38. Liu, Y, Ghesani, NV, and Zuckier, LS. 2010, Semin Nucl Med., 40, 294 39. von Schulthess, GK, Steinert, HC, and Hany, TF. 2006, Radiology, 238, 405. 40. Blodgett, TM, Meltzer, CC, and Townsend, DW. 2007, Radiology, 242, 360. 41. Hilner, BE, Siegel, BA, Liu, D, Shields, AF, Gareen, IF, Hanna, L, et al. 2008, J Clin Oncol., 26, 2155. 42. Abdel-Nabi H, Doerr, RJ, Lamonica, DM, Cronin, VR, Galantowicz, PJ, Carbone, CM, and Spaulding, MB. 1998, Radiology, 206, 755. 43. Kantorová, I, Lipská, L, Bêlohlávek, O, Visokai, V, Trubaĉ, M, and Schneiderová M. 2003, J Nucl Med., 44, 1784. 44. Dighe, S, Swift, I, and Brown, G. 2008, Clin Radiol., 63, 1372. 45. Bipat, S, Glas, AS, Slovs, FJM, Zwinderman, AH, Bossuyt, PM, and Stoker, J. 2004, Radiology, 232, 773. 46. Klessen, C, Rogalla, P, and Taupitz, M. 2007, Eur Radiol., 17, 379. 47. Furukawa, H, Ikuma, H, Seki, A, Yokoe, K, Yuen, S, Aramaki, T, and Yamaguchi, S. 2006, Gut, 55, 1007. 48. Park, IJ, Kim, HC, Yu, CS, Riu, MH, Chang, HM, Kim, JH, et al. 2006, Eur J Surg Oncol., 32, 941. 49. Llamas-Elvira, JM, Rodríguez-Fernández, A, Gutiérrez-Sainz, J, Gómez-Río, M, Bellón-Guardia, M, Ramos-Font, C, et al. 2007, Eur J Nucl Med Mol Imaging, 34, 859. 50. Sahani, DV, Kalra, SP, Fischman, AJ, Kadavigere, R, Blake, M, Hahn, PF, and Saini, S. 2004, Am J Roentgenol., 185, 239. 51. Bipat, S, Leeuwen, MS, Comans, EFI, Milan, EJ, Bossuyt, PM, Zwinderman, AH, et al. 2005, Radiology, 237, 123. 52. Gutman, F, Alberini, JL, Wartshi, M, Vilain, D, Le Stanc, E, Sarandi, F, et al. 2005, Am J Roentgenol., 185, 495. 53. Kamel, EM, Thumshirn, M, Truninger, K, Schiesser, M, Michael, F, Barbara, P, et al. 2004, J Nucl Med., 45, 1804. 54. Yasuda, S, Fujii, H, Nakahara, T, Nishiumi, N, Takahashi, W, Ide, M, and Shohtsu, A. 2001, J Nucl Med., 42, 989. 55. Huang, B, Law, MW, and Khong, P. 2009, Radiology, 251, 166. 56. Valk, PE, Abella-Columna, E, Haseman, MK, Pounds, TR, Tesar, RD, Myers, RW, et al. 1999, Arch Surg., 134, 503. 57. McCall, JL, Black, RB, Rich, CA, Harvey, JR, Baker, RA, Watts, JM et al. 1994, Dis Colon Rectum, 37, 875. 58. Flamen, P, Stroobants, S, Van Cutsem, E, Dupont, P, Bormans, G, De Vadder, N, et al.. 1999, J Clin Oncol., 17, 894. 59. Flanagan, FL, Dehdashti, F, Ogunbiyi, OA, Kogner, IJ, Siegel, BA. 1998, Ann Surg., 227, 319. 60. Even-Sapir, E, Parag, Y, Lerman, H, Gutman, M, Levine, C, Rabau, M, et al.. 2004, Radiology, 232, 815. 61. Bakx, R, van Tinteren, H, van Lanschot, JJ, and Zoetmulder, FA. 2004, Eur J Surg Oncol., 30, 857.
72
Rachel E. Hyland & Fahmid U. Chowdhury
62. Chowdhury, FU, Shah, N, Scarsbrook, AF, and Bradley, KM. 2010, Postgrad Med J., 86, 174. 63. Ito, K, Kato, T, Tadokoro, M, Ishiguchi, T, Oshima, M, Ishigaki, T, and Sakuma, S. 1992, Radiology, 182, 549. 64. Sobhani, I, Tiret, E, Lebtahi, R, Aparicio, T, Itti, E, Montravers, F, et al. 2008, Br J Cancer., 98, 875. 65. Scheele, J, Stang, R, Altendorf-Hofmann, A, and Paul, M. 1995, World J Surg., 19, 59. 66. Hughes, KS, Simon, R, Songhorabodi, Adson, MA, Ilstrup, DM, Fortner, JG, et al.1986, Surgery, 100, 278. 67. Ruers, TJ, Wiering, B, van der Sijp, JR, Roumen, RM, de Jong, KP, Comans, EFI, et al. 2009, J Nucl Med., 50, 1036. 68. Wiering, B, Adang, EM, van der Sijp, JR, Roumen, RM, de Jong, KP, Comans, EF, et al.2010, Nucl Med Commun, 31, 938. 69. Fong, Y, Saldinger, PF, Akhurst, T, Macapinlac, H, Yeung, H, Finn, RD, et al. 1999, Am J Surg., 178, 282. 70. Desai, DC, Zarvos, EE, Arnold, MW, Burak, WE Jr, Mantil, J, and Martin, EW Jr. Ann Surg Oncol 2003; 10:59-64. 71. Truant, S, Huglo, D, Hebbar, M, Ernst, O, Steinling, M, Pruvot, FR. 2005, Br J Surg., 92, 362. 72. Rohren, EM, Turkington, TG, and Coleman, RE. 2004, Radiology, 231, 305. 73. Lejeune, C, Bismuth, MJ, Conroy, T, Zanni, C, Bey, P, Bedenne, L, Faivre, J, et al. 2005, J Nucl Med., 46, 2020. 74. Park, KC, Schwimmer, J, Shepherd, JE, Phelps, ME, Czernin, JR, Schiepers, C, and Gambhir, S. 2001, Ann Surg., 233, 310. 75. Zubeldia, JM, Bednarczyk, EM, Baker, JG, and Nabi, HA. 2005, Cancer Biother Radiopharm., 20, 450. 76. Eisenhauer, EA, Therasse, P, Bogaerts, J, Schwartz, LH, Sargent, D, Ford, R, et al. 2009, Eur J Cancer., 45, 228. 77. Kalff, V, Duong, C, Drummond, EG, Matthews, JP, Hicks, RJ. 2006, J Nucl Med., 47, 14. 78. Capirci, C, Rubello, D, Chierichetti, F, Crepaldi, G, Fanti, S, Mandoliti, G, et al. 2006, Am J Roentgenol., 187, W202. 79. Wahl, RL, Jacene, H, Kasamon, Y, and Lodge, MA. 2009, J Nucl Med., 50(suppl 2), 122S. 80. Kuehl H, Antoch G, Stergar H, Veit-Haibach, P, Rosembaum-Krumme, S, Vogt, F, et al. 2008, Eur J Radiol., 67, 362. 81. Travaini, LL, Trifiro, G, Ravasi, L, Monfardini, L, Della Vigna, P, Bonomo, G, et al. 2008, Eur J Nucl Med Mol Imaging., 35, 1316. 82. Miller, FH, Keppke, AL, Reddy, D, Huang, J, Jin, J, Mulcahy, MF; and Salem, R. 2007, Am J Roentgenol., 188, 776. 83. Soyka, JD, Veit-Haibach, P, Strobel, K, Breitenstei, S, Tschopp, A, Mende, KA, et al. 2008, J Nucl Med., 49, 354. 84. Badiee, S, Franc, BL, Webb, EM, Chu, B, Hawkins, RA, and Coakley, F. Am J Roentgenol 2008; 191:1436-1439.
Imaging of colorectal cancer
73
85. Veit-Haibach, P, Kuehle, CA, Beyer, T, Stergar, H, Kuehl, H, Schmidt, J, et al. 2006, JAMA, 296, 2590. 86. Francis, DL, Freeman, A, Visvikis, D, Costa, DC, Luthra, SK, Novelli, M, et al. 2003, Gut, 52, 1602. 87. Cher, LM, Murone, C, Lawrentschuk, N, Ramdave, S, Papenfuss, A, Hannah, A, et al. 2006, J Nucl Med., 47, 410. 88. Chowdhury, FU, Bradley, KM, and Gleeson, FV. 2008, Clin Radiol., 63, 1297. 89. Gregoire, V, Haustermans, K, Geets, X, Roels, S, and Lonneux, M. 2007, J Nucl Med., 48, 68S. 90. Ciernik, IF, Dizendorf, E, Baumert, BG, Reiner, B, Burger, C, Davis, JB, et al. 2003, Int J Radiat Oncol Biol Phys., 57, 853.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 75-101 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
4. Local management of primary colorectal cancer: Focus on surgery
1
Henar Núñez1, Esther Uña2, José Herreros1, Carlos Abril1 and Alejandro Romero1
Colorectal Cancer Surgery Unit, Clinical University Hospital of Valladolid, Spain 2 Digestive Oncology Unit, Clinical University Hospital of Valladolid, Spain
Abstract. CRC survival is primarily related to the stage of disease at diagnosis. Surgery is still the only curative treatment and sometimes a surgical procedure is also used as a palliative measure for patients with metastases. The operative procedure must be tailored to the location of the cancer inside the bowel, but should include complete, en bloc resection of the cancer and its lymphatic drainage, including locally invaded structures. The bowel margins of resection should be at least 5 cm from the tumor to minimize anastomotic recurrences. The resection of lymph nodes has prognostic and therapeutic implications and the use of sentinel lymph node biopsy is feasible but has not yet been proved clinically useful. Laparoscopic colectomy has been shown to be as safe and effective as open colectomy for the treatment of this disease although early experiences with laparoscopic colectomy were unfavorable with higher than expected rates of wound tumor Correspondence/Reprint request: Dr. Esther Uña, Digestive Oncology Unit, Clinical University Hospital of Valladolid, Spain. E-mail: aunacid@hotmail.com
76
Henar Núùez et al.
implants and concerns about short and long-term compromised oncologic outcomes. Nowadays, the level of evidence supports the general feasibility and recovery advantage as well as cancer equivalence between both techniques.
Introduction In about 80% of cases of colon cancer this disease is localized to the bowel and regional lymph nodes at diagnosis (1). Surgery is the only curative modality for localized CRC and also provides a potentially curative option for selected patients with limited metastatic disease in liver and/or lung. In many cases, even in those who cannot be curated by surgery, this procedure could benefit them through palliation of symptoms such as obstruction and bleeding from the primary tumor (2). The overall 1 and 5 year survival rates are 83% and 62% respectively, and this rate continues to decline to 55% at 10 years (2). Nowadays, adjuvant treatment have shown to increase chances of cure in surgically resected patients who are at high risk of recurrence, and at the same time that efforts are made to improve adjuvant therapies, there is renewed interest in improving the surgical techniques (3). At this time the optimal treatment of these tumours should maximize outcomes with low morbidity and mortality (4). There is wide evidence that surgical treatment could affect the outcome significantly. In this way this chapter tries to review surgical treatment for this disease (5).
1. Preoperative evaluation as the first step to choose the tailored approach Once the diagnosis of CRC has been made, which is usually performed through a colonoscopy, a complete history and physical examination should be performed in all patients. It is very important to get a family history of CRC and other extracolonic cancers as the patient may be a member of a kindred with a hereditary predisposition. All these findings could alter the surgical approach, leading to a consideration of subtotal or total colectomy in high-risk individuals (2). The preoperative evaluation should also include a complete blood count, serum electrolytes, liver enzymes, and carcinoembryonic antigen (CEA), urinalysis, coagulation profile, electrocardiogram, chest x-ray, and computed tomography (CT) scan of the abdomen and pelvis (6-8). This evaluation allows the surgeon to clinically stage the patient, evaluate for the presence of comorbidities, and plan an elective surgical procedure.
Colorectal cancer surgery
77
A brief review of preoperative diagnosis Ideally, each patient should undergo to a complete colonoscopy prior to surgery to know about the tumor location, to get a histopathologic diagnosis, to characterize the lesion (mucosal or submucosal) and to evaluate for the presence of extrinsic colonic compression and/or synchronous carcinomas and adenomas which are the precursor lesions for CRC (6-8). Synchronous invasive CRC are found in 3-5% of cases, while the prevalence of synchronous adenomas is as high as 30% (6-8). Colonoscopy is not the perfect diagnostic test as it was shown in the study by Rex et al. who reported a rate of missed adenomas of 24% (9). If full colonoscopy cannot be performed prior to elective colon resection because of obstruction or poor preparation, a barium enema or computed axial tomography or magnetic resonance (MR) colonography can be used to evaluate the entire large bowel. Barium enema may show the classic â&#x20AC;&#x153;apple coreâ&#x20AC;? sign (see Figure 1), which is characteristic of a CRC, as well as its location. However, other potentially significant lesions such as small cancers or adenomas may be missed (10).
Figure 1. This is a barium enema. The area with the arrow shows an advanced cancer of colon that has produced an apple core lesion.
78
Henar Núùez et al.
Computed Tomography and MR colonography are useful to evaluate the proximal colon in cases of obstructive cancer, to identify synchronous neoplasms and tumour location, although data are limited (11). The endoscopic rectal ultrasound (ERUS) is not routinely used in the preoperative workup of colon cancer (in contrast to rectal cancer, where it is routinely used). This is the most accurate modality for assessing local depth of invasion of rectal cancer into the rectal wall layers (T stage). Unfortunately ERUS is not as good for predicting nodal metastases as it is for tumour depth, which could be related to the unclear definition of nodal metastases. Overstaging is more frequent than understaging mostly due to inflammatory changes and the limitations this technique has are operator and experience dependency, limited tolerance of patients and limited range of depth of the transducer. Its role is limited to the evaluation of submucosal or extraluminal masses and it has the advantage of allowing a tissue biopsy through EUS-guided needle, especially of submucosal or extracolonic lesions (12, 13). MR is a useful modality for the evaluation of rectal cancer, providing superior anatomic and pathologic visualization when compared with ERUS and computed tomography. Endorectal MR has an accuracy comparable to ERUS for T staging of superficial tumours. The problem would be also an overstaging risk due to fibrotic desmoplastic reactions, which mimic true mesorectal tumour invasion. MR can also identify infiltration of extramural veins and the peritoneal fold. Identification of metastatic lymph nodes is the most challenging issue in rectal cancer imaging and also a key in planning the treatment sequence. Using a size cutoff identifies enlarged nodes but it cannot distinguish between metastatic or benign etiologies. To date, the most accurate criteria for suspicious lymph nodes are irregular contour, spiculated or indistinct borders and heterogeneous signal but there are no known signal intensity properties that can reliably distinguish metastatic from reactive nodes (14-16).
2. General surgical principles for colorectal cancer The goal of colon cancer surgery is to perform a complete removal of the tumor through resection of the affected colonic segment along with the major vascular pedicle feeding, and the lymphatic drainage basin (17). Although segmental resection alone may be sufficient for primary tumor removal, wider resection is generally needed to achieve a sufficient lymphadenectomy (18). The blood vessels should be divided at their origin in order to obtain a wide resection and maximize the number of lymph nodes in the specimen.
Colorectal cancer surgery
79
Figure 2. En bloc resection of colorectal cancer.
A proper lymphadenectomy should extent to the level of the feeding artery and all lymph nodes should be removed en bloc with the tumourbearing segment of colon to get the best results. The nodes at the origin of the feeding vessel should be tagged for pathologic analysis. Lymph nodes that feel suspicious but lie outside the field of resection should be sampled and if these nodes are positive for malignancy, the entirety of that lymph node basin should also be resected. To fulfill the criteria for entry into adjuvant trials, at least one lymph node must be examined, and for entry into adjuvant trials but for negative lymph nodes, at least 12 lymph nodes must be examined (18-21). En bloc resection of contiguous structures is indicated if there is attachment or adhesion of the tumor to those structures. In those cases the tumor bed should be marked with radiopaque clips at the time or surgery to facilitate adjuvant locoregional treatment if necessary (17). In such cases, omentum could also be mobilized and placed in the postoperative bed to increase distance between the small bowel and the site of future irradiation. Following resection, bowel continuity should be reestablished if possible, preferably at the time of resection, which means to perform a primary anastomosis. When performing an anastomosis, there should be no tension at that because the risk of failure.
80
Henar Núùez et al.
Resection margins The ideal extent of a bowel resection is defined by removing the blood supply and the lymphatics at the level of the origin of the primary feeding arterial vessel. When the primary tumour is equidistant from two feeding vessels, both of them should be resected at their origin. Although there is controversy, proximal and distal margins should be at least 5 cm from the tumor. These margins should allow for an adequate resection of the appropriate segment of the bowel with its vascular supply and associated lymphatics. For patients undergoing right hemicolectomy, the length of ileum does not influence local recurrence rates. So, consequently, the shortest length of the ileum should be excised to prevent malabsorption syndromes (22, 23).
Regional lymphadenectomy Lymph node resection has prognostic and therapeutic implications (20). An adequate lymph node resection should extend to the level of the origin of the primary feeding vessel and in all cases the resection must be radical. To obtain this the excision should remove nodes en bloc. Apical nodes, which are located at the origin of feeding vessels should be removed when it is possible (17). Whenever suspected lymph nodes are detected outside the field of resection they need also to be sampled. Depending on their pathological results there are four possibilities: *
*
* *
If these results are positive for disease and the nodes are resected along with the apical ones, this resection is considered to be complete (called R0). If these results are positive for disease but they are not clinically involved, and they are not resected, the resection is microscopically incomplete (R1). If biopsy results are negative and the apical nodes positive for disease are resected, the resection is considered to be complete (R0). If biopsy results for lymph nodes left behind are positive for disease, the resection is incomplete (R2).
As a conclusion, patients with R1 or R2 resections are not considered to be eligible for adjuvant trials. The randomized clinical trial carried out by Rouffet et al (24) has assessed the value of a radical lymphadenectomy for left-sided colon cancer. There was no evidence of a benefit for a wider lymphadenectomy.
Colorectal cancer surgery
81
On the other hand, several retrospective studies (25-29) have yielded conflicting results on the value of extended lymphadenectomy and the conclusion has been that if any benefit could appear it would be in Dukes' stage C. The apical nodes may also have prognostic significance in addition to the number of lymph nodes positive for disease in the specimen (30). This provides important prognostic information that guides adjuvant treatment and is of therapeutic value as well. There is a direct correlation between the number of lymph nodes evaluated per patient after surgical resection and survival (31, 32). It has long been recognized that the frequency of detection of nodal disease is influenced by surgical technique and the methodology of lymph node handling and assessment. Surgical technique may contribute to variation in the number of nodes contained in a resection specimen. As an example, complete mesocolic excision and central venous ligation has been reported to remove more tissue around the tumor which could potentially yield more lymph nodes (33). On the other hand diligence in the search for nodes the use of fat clearing or other techniques to increase macroscopic visualization of nodes, and the pathologist´s threshold for an acceptable number of evaluated nodes are variables in pathologic technique. Consensus guidelines recommend that at least 12 lymph nodes be assessed for adequate staging (34-36). By definition, regional lymph nodes include those nodes along the course of major vessels, along the vascular arcades of the marginal artery, and those adjacent to the colon along its mesocolic border. The regional nodes differ for each segment of the large bowel. Involved lymph outside the primary nodedraining basin and extraregional lymph nodes are staged as metastatic disease (37).
Sentinel node mapping According to the sentinel lymph node hypothesis, tumor cells migrating from a primary tumor colonize one or a few lymph nodes before involving other lymph nodes. When this concept is taken into account, location, removal and pathologic analysis of this SLN provides an important information about the status of regional lymph nodes (38). The accuracy and advantages of SLN mapping for malignancies of for example breast cancer and melanoma, have been well documented in multiple trials (39).
82
Henar Núùez et al.
For CRC the principal advantage of this sampling is improvement in the accuracy and ease of determination of lymph node status after resection. The sentinel lymph node (SLN) procedure is a selective sampling technique that can be used to stage regional nodes. These SLN are the first nodes to receive lymphatic drainage from a primary tumour and the place where it is more likely to find metastases. The status of this SLN will not change the extent of lymph nodes resection because en bloc resection of the primary colon cancer includes regional lymph nodes (40). One of the advantages this technique has is the opportunity to focus on a more detailed exam using immunohistochemistry (IHC) and molecular techniques of one or a few lymph nodes which is more likely to have malignant cells. This could improve the accuracy of staging in this disease but it has also led to an increase in the finding of occult micrometastases that are not detected by routine hematoxylin and eosin staining. Upstaging has been reported in 11 to 29% of patients with node-negative colon cancer when the nodes are examined by IHC and / or molecular techniques (41-43). Injection of vital blue dye with or without radiolabeled colloid around the area of the tumor permits identification of an sentinel lymph node in the majority of patients, and its status accurately predicts the status of the remaining regional lymph nodes. The study by Bilchik et al. has concluded that the detection of occult micrometastases could improve the selection of patients for adjuvant therapy (44). However, no study has addressed whether adjuvant chemotherapy provides a similar magnitude of benefit for patients with occult micrometastatic disease as it does in the setting of macroscopically positive SLN. As a conclusion not all investigators have reported success with SLN mapping in colon cancer where it has not proved its clinical utility (45). There are several studies single and multiple institutions in which multilevel sectioning of SLN failed to predict the nodal metastases in a significant number of patients, mainly in T3 / T4. All these data suggest that SLN mapping may not improve the accuracy of standard nodal staging in colon cancer (46-49).
3. Surgical procedures in oncological colorectal cancer surgery Right hemicolectomy Right colon tumors have a frequency of 15 to 20% of the total of CRC. It has been estimated that in latest 25 years there has been a migration from
Colorectal cancer surgery
83
distal to proximal tumors. In some series tumors in right colon are even more frequent than those placed in left colon. This surgical technique is usually performed for tumours of the cecum and ascending colon, and for some located in the hepatic flexure. Once the complete preoperative evaluation is performed, a classic right hemicolectomy begins with mobilization of the right colon and ileocecal area by dividing the peritoneum lateral to the ascending colon and cecum. The ileocolic, right colic and right branch of the middle colic vessels are divided and removed with the contiguous mesentery with cautious to avoid damage of the right ureter, ovarian or testicular vessels and the duodenum. During mobilization of the hepatic flexure it is important to avoid injuring the duodenum, which lies beneath this part of the colon. If the omentum is attached to the tumor, it should be removed en bloc with the specimen. Proximal margin has to include about 10 cm of distal ileum and reconstruction of the intestinal transit should be done by laterolateral anastomosis (ileotransverse anastomosis) through manual or authomatic stapler because either of these two techniques has shown to be superior to the other (50).
Extended right hemicolectomy This technique could be used for proximal, mid or even distal transverse colon cancers, although the last ones are more often resected by left hemicolectomy. In this surgery the ileocolic, right colic and middle colic vessels with their contiguous mesentery are divided and removed. The inferior mesenteric vein may be divided and included in the specimen. Care must be taken to protect the duodenum and the pancreas. The inferior mesenteric vein is also divided and included in the resected specimen. Both the hepatic and splenic flexures may need to be mobilized in order to achieve a tension-free anastomosis. When mobilizing the splenic flexure, care must be taken not to apply much traction to the omentum or colon as this will invariably result in splenic capsule tears. Whenever the tumor is located in the transverse colon, the entire omentum overlying the neoplasia must be taken. At this point, the entire right colon and transverse colon can be lifted with their vascular supply from the retroperitoneum, if the tumour is not adherent to adjacent structures. If the tumour is adherent to adjacent structures, these must be taken en bloc with the tumour. The transverse colon is divided at the bifurcation of the middle colic artery. The transverse colon is resected along with the middle colic vessels
84
Henar Núùez et al.
and its mesentery. The right branch of the middle colic artery must be identified and ligated at its origin. At this point the blood supply to the distal colon is dependent on the patency of the inferior mesenteric artery and collateral pathways (marginal artery of Drummond and the arcade of Riolan).
Left hemicolectomy This is appropriate for tumours in the distal transverse or descending colon and for selected patients with proximal sigmoid colon cancer. The left colon is mobilized by incising the peritoneum lateral to the descending colon through an incision performed proximally toward the splenic flexure and lowered with caution not to damage the marginal artery. The left colon can now be lifted on its vascular pedicle and the inferior mesenteric artery is divided at its origin from the aorta. The inferior mesenteric vein is identified as it separates from the inferior mesenteric artery. This vein joins the splenic vein behind the pancreas. Ligation of the inferior mesenteric vein high up under the inferior border of the pancreas greatly lengthens the reach of the transverse colon to the rectum. The segment of the colon with the tumour could now be isolated with linear staplers. If the tumor is adherent to adjacent organs these must be removed en bloc to prevent spillage of malignant cells into the peritoneum. The mesentery and marginal artery are now divided, and the specimen is removed. The colorectal anastomosis can then be constructed in a stapled or handsewn fashion, depending on the surgeon´s preference. In some cases, a segmental colectomy may be performed as long as adequate resection margins and lymphadenectomy are achieved (51).
Sigmoid colectomy For sigmoid colon cancers, segmental or sigmoid colectomy is appropriate (see Figure 3 and 4). The procedure is similar to a left colectomy except that less colon is resected. The inferior mesenteric artery is divided at its origin and dissection proceeds just under the superior rectal vessels toward the pelvis until adequate margins are obtained. It needs caution while mobilizing the sigmoid and descending colon to identify the left ureter and the left ovarian or testicular vessels. Whenever the tumour is located in sigmoid colon the surgeon needs to ligate inferior mesenteric artery and if the tumour is located in rectum-sigmoid junction the surgeon needs to ligate superior hemorrhoidal artery.
Colorectal cancer surgery
85
Figures 3 and 4. Two surgical specimens of partial colectomy (removal of the colon) showing a cancer.
It is nessessary to mobilize the splenic flexure of the colon to avoid a tension anastomosis. Reconstruction is made through laterolateral anastomosis with mechanical sutures (20,52).
Subtotal and total colectomy These two procedures are indicated whenever there are synchronous carcinomas on the right and left sides of the colon.
86
Henar Núùez et al.
Occasionally in patients with obstructing-left sided tumours or with hereditary nonpolyposis CRC (HPNCC) who present a colon cancer, total abdominal colectomy would be the choice procedure. They are also indicated in cases with a tumour recurrence in the colon after an operation for the primary tumour to be removed. In cases with a proximal perforation and distal obstruction in which the reconstruction would be ileorectal anastomosis, it is also used one of these two techniques (53).
Locally advanced primary lesions About 10% of colon cancer have invasion of contiguous structures or inflammatory adhesions involving neighboring organs. There are different guidelines such as the National Comprehensive Cancer Network (NCCN), National Cancer Institute (NCI) and American Society of Colon and Rectal Surgeons, which have highlighted the relevance of surgical management for locally advanced colon cancer. They have concluded that a multivisceral resection with a negative margin of the adjacent structure must be done (19, 54, 55). This kind of resection is associated with improved local control and overall survival rates (56-59). There are several series (60, 61) reporting results similar to those who do not have organ adhesions although perioperative morbidity rates may be higher (60). The plane of adherence tumour-organ must not be broken because 40% of these adhesions are malignant and this could increase the recurrences rates or persistence of disease (62). Despite the benefits of multivisceral resections this kind of surgical procedure is not often used. The study by Govindarajan et al. has shown that only one third of locally advanced tumours are managed with multivisceral resection en bloc (63).
4. Specific issues in rectal cancer surgery Although anatomic definition of the rectum could be highly variable, rigid proctoscopy is a highly reproducible method of determining the level of the rectum and is less dependent on the operator or on the technique. The anal verge is the preferred anal landmark, since the edge of the tumour and the verge can be visualized simultaneously during rigid proctoscopy. There have been many discrepancies for defining the union between colon and rectum, but in the end it has been agreement that for colon it is considered
Colorectal cancer surgery
87
greater than 12 cm from the anal verge by rigid proctoscopy and for rectum 12 cm or less from anal verge. This definition is based on the differences in natural history between tumors located above or below this limit of 12 cm. In this way, tumours above have a local recurrence rate of 9.6% (64) compared to more than 30% for mid and low rectal cancers. There are some retrospective studies supporting this observation (65-68). At present, the optimal treatment of rectal cancer should maximize sphincter preservation with low morbidity and mortality. Moreover, favorable oncological outcomes in terms of a low local recurrence rate and a high survival rate are also important considerations. In this way, the surgical technique plays an important role to achieve these goals. Nowadays the preferred treatment would be a low anterior resection with total mesorectal excision and sphincter preservation. This procedure consists of performing a complete removal of the tumorâ&#x20AC;&#x2122;s lymphatic and vascular set with free margins of disease. Aiming those objectives it would be also relevant to keep in mind these two parameters: distal mesorectal spread and the circumferential mesorectal margin.
Distal resection margin The removal of lower rectal tumors (< 5 cm from the anal verge) with sphincter preservation has been possible by the introduction of surgical staplers, and revision of the traditional rule of 5 cm resection margin. Several studies have shown that distal intramural spread is rare and is found beyond 1 cm in only 4% to 10% of rectal cancers (69-72). In this way, survival and local recurrence have been shown to be acceptable with a 2 cm or greater distal bowel margin. Whenever distal spread occurs beyond 1.5 cm, it is usually from a poorly differentiated cancer and the prognosis is poor, regardless of the length of the distal margin (73). There are authors claiming for margins even smaller than 2 cm because these don´t increase the local recurrences rates or compromise 5-year survival (74). Therefore, it is currently accepted that due to distal intramural spread beyond 1 cm occurs rarely and it is associated with tumors of advanced stage at diagnosis and histologically aggressive disease, namely, poorly differentiated cancer and lymphovascular and perineural invasion, the associated poor prognosis is not improved by a longer distal margin (75). A National Cancer Institute (NCI) Expert Panel Guidelines series published in 2000, recommended a distal margin length of 2 cm as ideal, with margins of 1 cm being acceptable in low rectal tumors (17).
88
Henar Núùez et al.
A 1 cm distal margin may be adequate according to others (76, 77). With preoperative radiation therapy and chemotherapy, a simple, clear margin may be adequate, but this has not been accepted as a general guideline (78). In patients with ultra-low rectal cancer, the technique of transanal resection of a part or the entire internal anal sphincter was introduced aiming to preserve the sphincter and to keep bowel continuity with proper distal margins. This technique was first described by Schiessel et al who undertook the procedure to enable restorative resection and avoidance of a permanent stoma (79). Intersphincteric resection is a technical development of extended low anterior resection for low rectal cancer which involves the mucosa of the upper anal canal. The key steps of the operation include a laparotomy with mobilization of the left colon including the splenic flexure, autonomic nerve sparing rectal mobilization with total mesorectal excision down to the pelvic floor including division of Waldeyerâ&#x20AC;&#x2122;s fascia and anorectal mobilization within the puborectalis sling. The perineal phase involves circumferential anal canal transaction just deep to internal anal sphincter, which must be preserved in its distal part, and dissection upwards to meet in the plane of the abdominal dissection. This enables complete mobilization of anorectum with its tumour and permits delivery of the specimen, usually trans-anal, to minimize contamination. Big quantity of lavage of the anal canal and rectum with a tumouricidal agent, 10% Povidone iodine solution and purse string closure of the lower end of anorectum early after trans-anal mobilization are useful additions to the operation that would help minimise tumour seeding and recurrence. Reconstruction through trans-anal hand sewn colo-anal anastomosis. Using intersphincteric resection in 92 patients with a tumor at 1.5-4.5 cm (mean 3 cm) from the anal verge, Rullier et al achieved negative margins in 98% of cases; local recurrence was found in 2% (80).
Mesorectal excision Total mesorectal excision (TME) was the term introduced by Heald et al. (81) in 1982 and this concept revolutionized the surgical management of rectal cancer. TME has gained worldwide acceptance as a standard surgical technique. Actually, its use can result in a dramatic reduction of the local recurrence rate (82). Its principles have been advocated for sharp pelvic dissection based on pelvic anatomy, subsequently resulting in not only en-bloc removal of rectal
Colorectal cancer surgery
89
cancer and surrounding mesorectum containing a lymph node, but also pelvic autonomic nerve preservation (82). The quality of the macroscopic specimen for rectal cancer has been regarded as an important parameter for prediction of prognosis (83, 84). Oncologically, there must be no damage to the mesorectum during dissection, and a clear circumferential resection margin and distal resection margin should be obtained after resection. a. Mesorectum The mesorectum is defined as the lymphovascular, fatty and neural tissue that is circumferentially adherent to the rectum, starting at the level of the sacral promontory, where the superior hemorrhoidal vessel divides into right and left branches. This structure tapers and then diminishes below Waldeyer's fascia (which is the fascia of the levators) around the levator ani muscles at the level of the distal third of the rectum. The clinical significance of this structure and subsequently its surgical clearance is supported by the demonstration of the presence of tumor deposits within the mesorectum far away from the primary tumor (81) and by the demonstration of a strong correlation between the extent of the mesorectal tumor spread and cancer outcomes (85). Mesorectal spread can occur by direct tumor extension or as lymph nodes, perineural invasion, or isolated mesenteric deposits (84). Mesorectal spread has shown to be an important indicator of disease severity (85). Several studies (86-88) have shown higher rates of recurrence in patients with lateral resection margins that are positive for disease (89). The fact that lateral clearance rather than lateral spread is associated with the risk of local failure in a large study supports the relevance of radial clearance surgery. Further support derives from the fact that the practice of mesorectal clearance is associated with small rates of local recurrence (90-92). Data from pathologic assessments of rectal cancer specimens with attention to mesorectal deposits suggest that mesorectal clearance of at least 4 cm distal to the tumor should be sufficient (93). As a conclusion, the goal of rectal cancer surgery should be to get radial clearance of mesorectal tissue, including the primary tumor and lymphatic, vascular, and perineural set around the tumor. Wide anatomic resection includes presacral dissection under direct visualization, preservation of mesorectal fascia propria integrity, at least 4-cm clearance of attached mesorectum distal to the tumor, and pathologic confirmation of mesorectum
90
Henar Núñez et al.
attached to the bowel, distal to the tumor. This is accepted as the standard technique for rectal cancer procedures.
Distal mesorectal margin Heald et al. (90) pioneered the use of total mesorectal excision (TME), and reported distal mesorectal spread of 4 cm from the distal tumor edge. Hida et al (93) reported that in patients with pT3 and pT4 rectal cancer, the extent of distal mesorectal spread was related to tumor location. The longest distance to a metastatic node was 2 cm. in rectosigma, 4 cm. in upper rectum, and 3 cm. in lower rectum. They concluded that a mesorectal margin of at least 5 cm. was required in the surgical treatment of locally advanced rectal cancer and hypothesized that blockage of the upward lymphatic flow by the locally advanced cancer produced a downward spread in the mesorectum. They also suggested a 4-cm. mesorectal margin as an adequate margin to ensure oncologic resection.
Circumferential resection margin (CRM) The CRM also called “radial resection margin”, corresponds to the nonperitonealized surface of the resection specimen created by dissection of the subperitoneal aspect at surgery. The posterior CRM is triangular, and runs up towards the sigmoid mesocolon; the anterior CRM is located in the most distal aspect of the specimen. To be more precise it is defined as the minimum distance from tumour to mesorectal fascia, and determines the surgical resection plane necessary to achieve tumour-free margins. The CRM is the most powerful predictor of local recurrence rate. It can be used to identify tumours with close or involved margins. Thus, requiring neoadjuvant treatment and more aggressive surgery. Magnetic resonance (MR) imaging is a promising tool in preoperative local staging and may also provide measurements of the distance to the mesorectal fascia as the potential CRM in TME. The preoperative identification of patients at high risk of a positive CRM prior to surgery has improved with advances in this technique. It appears that histopathologic tumor involvement of the CRM is an independent predictor of local recurrence and distant metastases and may therefore influence overall survival after primary resection of rectal cancer (94,95). Marijnen et al (96) have showed that, even after preoperative radiation therapy, patients in whom the CRM was histopathologically involved had a substantially worse prognosis than patients with tumor-free CRM.
Colorectal cancer surgery
91
Recent data have confirmed the accurate prediction of both T stage and CRM clearance of 1 mm. of the resection margin using MRI. The accurate determination of the CRM status is essential, because it is the single most important factor for predicting the risk of local recurrence in patients with rectal cancer. A positive CRM is defined as continuous or discontinuous tumor extension, or the presence of a positive lymph node < 1 mm. from the radial, nonperitonealized soft tissue edge or when tumour is seen within 1 mm of the mesorectal fascia or has visibly invaded it. A tumour-free CRM is assumed when the closest visible tumour extension, mesorectal tumour deposits or suspicious lymph nodes are over 6 mm from the mesorectal fascia, but there is no consensus on the status of tumours located 2 to 5 mm from the mesorectal fascia. All these situations are associated with higher disease stage, higher histology grade, and tumor infiltration (97). A radial margin of less than 1 mm. was predictive of an increased risk of distant metastases (37% vs 15%) and shorter survival (70% vs 90%). There are other factors directly related to a positive CRM, such as the surgical technique used or the tumor location (lower and anterior rectal tumours are at greater risk which could be explained by the thinner mesorectum here). In this way, CRM was positive in 7.3% of 1113 patients after TME or PME (partial mesorectal excision) compared to 17% of 2450 patients after conventional blunt rectal dissection. Bernstein et al (98) studied 3194 patients with known CRM status, and made the conclusion that a CRM of 2 mm. or less had an impact on the prognosis of T2 and T3 tumors located 6-15 cm above the anal verge, but not on lower tumors. A CRM of 2 mm. or less confers a poorer prognosis, and patients should be considered for neoadjuvant treatment. The use of preoperative neoadjuvant chemoradiation has also been associated with CRM involvement (99,100).
Local surgical procedures for rectal cancer In selecting a treatment approach for a patient with rectal cancer, there are three points to consider. Firstly the patient´s conditions, secondly the tumour location and the stage of this tumour. Major resection for rectal cancer involves high morbidity rates of 20% to 30% and even mortality rates of 2% to 3%. Therefore, there are several patients too sick to undergo an abdominal operation or even sometimes continence issues will push patient and surgeon toward local excisions if these are possible.
92
Henar Núùez et al.
The ideal tumour for local excision is the tumour with a predictably low potential for lymph node involvement. And this risk is related to the depth of tumour penetration into the bowel wall. In this way, the rate of metastatic mesorectal nodal disease for T1 / T2 / T3 and T4 is 5-11%, 20-30%, 50-65%, and 75-80% respectively (101,102). Subsequently, an accurate determination of the depth of tumour penetration is critical when considering a local approach for a rectal cancer. Then, the absolute size of the tumour must be taken into consideration. This tumour should be less than 3 cm in size and involve no more than 30% to 40% of the circumference of the rectal wall. Lesion larger than this would be vary difficult to manage through the anus and also increasing tumour size seems to be related to increasing nodes positivity rate and higher rates of local recurrence. However, more recent data suggest that tumour size is not an independent predictor of lymph node positivity in multivariate analyses. Also tumour histology has significant value in choosing the surgical approach. Poorly differentiated tumours or lymphovascular or perineural invasion have a high rate of lymph node positivity and this finding lead us to consider a more radical approach than simple local excision alone. Mucinous tumours also behave aggressively and several studies have shown this as a negative feature for local excision. As a conclusion, the ideal tumour for local excision would be a nonulcerated tumour less than 3 cm in size and located well beneath the middle rectal valve with a histology showing good or moderate differentiation with no lymphovascular or perineural invasion and no mucinous features. There are different techniques for excision of rectal cancer but a deeply review is out of the scope of this chapter. One of these techniques is endoscopic submucosal dissection (ESD) which allows en-bloc resection of a lesion, regardless of its size (103). ESD has been established as a standard method for the endoscopic ablation of malignant tumors in the upper gastrointestinal (GI) tract in Japan (104), but although its use for CRC has been studied, it is not yet established as a standard therapeutic method. The current proposed indications for colorectal ESD are for lesions difficult to remove en bloc with a snare Endoscopic Mucosal Resection (EMR), such as nongranular laterally spreading tumors, lesions showing a type VI pit pattern, and large lesions of the protruded type suspected to be carcinogenic; or lesions with fibrosis due to biopsy or peristalsis; sporadic localized lesions in chronic inflammation ie ulcerative colitis; and local residual carcinoma after EMR.
Colorectal cancer surgery
93
Saito et al treated a total of 400 patients for 405 lesions with ESD. En-bloc resection rate was 87% and curative resection rate of 86% with a perforation rate of 3.5% (105). It has been concluded that ESD is a feasible technique for treating large superficial CRC because it provides a higher en-bloc resection rate with less invasive procedure than surgical resection and provides also a histologic information (106).
5. Special surgical scenarios Obstructed tumours Optimal management for an obstructing cancer depends on the patient condition and tumor location. In cases in which the patient is fit for surgery several surgical alternatives would include resection of the tumour with primary anastomosis with or without a temporary proximal diversion, which is the preferred procedure. In other cases a resection without an anastomosis and with an end colostomy and proximal diversion with a mucous fistula or a loop colostomy, to temporize the situation with an elective definitive resection at a second operation. Primary anastomosis is not recommended if there is diffuse peritonitis or free perforation and/or the patient is medically unstable. Because the right and transverse colon have lower bacterial counts and contain liquid stool, right colectomy or extended right colectomy and primary ileocolonic anastomosis without proximal diversion can be safely performed for obstructing right-sided cancers even with an unprepped bowel. On the other hand, in cases of obstructing left-sided tumours many surgeons routinely perform a temporary proximal diverting colostomy following the resection. However, some studies suggest that obstructing left-sided lesions can be safely managed with a one-stage procedure, with acceptable morbidity (107-109). There is a different temporizing alternative for an obstructing cancer which is endoscopic placement of an expandable metal stent. Following the placement of the stent, standard bowel cleansing can be achieved and an elective procedure performed avoiding an emergent one. Experience with this approach is limited and potentially fatal complications (perforation, stent migration) have been described. We reserve this approach for the rare patient with a single obstruction and carcinomatosis who may get a response to chemotherapy and is a poor candidate for colostomy.
94
Henar Núùez et al.
Perforated tumours The management of perforated tumours has to be individualized. Treatment alternatives will depend on the patient´s overall condition and whether peritonitis is localized or generalized. In stable patients with localized peritonitis a tumor resection with a primary anastomosis can be performed. For an unstable patient, percutaneous drainage or a proximal diverting colostomy can be performed. However, if drainage is performed, there will be the potential for seeding of the drain tracts; therefore, at the time of definitive resection, an en bloc resection including the abdominal wall with drain tract will need to be performed. Similarly, perforation may cause the tumour to adhere to other organs and an en bloc resection will need to be performed.
Timing of ostomy closure If an ostomy was performed at the time of primary colon surgery, the timing of stoma closure depends on factors such as the time of recovery from the events that led to the decision to create the stoma (perforation, peritonitis, obstruction, cardiovascular instability). If adjuvant chemotherapy is being administered, closure should be delayed to avoid unnecessary treatment delays. Finally, if a primary anastomosis was performed, it must be studied for integrity and patency prior to closure of the stoma, and globally colonoscopy is indicated prior to closure if the entire colon was not evaluated prior to the stoma formation due to the emergent situation or due to an obstructing tumour.
6. Laparoscopic colectomy This technique was first described in 1991 (110). However, because of concerns over port site recurrences and the adequacy of the oncologic procedure, laparoscopic colectomy was not accepted by the surgical community until the completion of multiple studies comparing safety and oncologic equivalency to open colectomy. At least six large prospective randomized controlled trials have been published. None suggest a significant detrimental impact on recurrence or survival for laparoscopic as compared to open colectomy (111-117). Three separate meta-analyses have reached the same conclusion (118-120). In the largest trial, the United States Intergroup Clinical Outcomes
Colorectal cancer surgery
95
of Surgical Therapy (COST) trial, 872 patients with colonic adenocarcinoma were randomly assigned to open colectomy or laparoscopic colectomy by well prepared surgeons, who each performed at least 20 laparoscopic resections and underwent videotape analysis of their technique before enrolling patients on the study. Operative time was significantly longer in the laparoscopic group (150 vs 95 minutes) and 21% of laparoscopic cases required an open procedure. Besides this, the laparoscopic group had modestly but significantly shorter durations of hospital stay (five versus six days) and parenteral analgesic use (three versus four days). There were no significant differences with regard to intraoperative or postoperative complications, perioperative mortality rates, readmission or reoperation rates, or rate of surgical wound recurrence. At a median follow up of seven years there were also no significant differences in the five year disease-free survival (69 versus 68 % in the laparoscopic group and open colectomy groups, respectively) or overall survival (76 versus 75%) (121, 122). Long-term outcomes from laparoscopic versus open surgery for CRC have been addressed in two meta-analyses of randomized trials, most of which enrolled exclusively or predominantly patients with colon cancer (123, 124). One analysed results from seven published randomized trials, while the other focused on four trials with the end point of survival that enrolled more than 150 patients (111- 114 , 116). Both agreed that laparoscopic colectomy provides oncologic outcomes comparable to those achieved with an open approach. Of the seven randomised trials reporting this outcome, only three of 826 patients randomised to laparoscopic surgery developed an incisional recurrence, compared to one of 801 patients undergoing open surgery (0.36 versus 0.12%) (119). Patients with tumours invading adjacent organs or with perforated tumours should not be considered candidates for laparoscopic procedures. Otherwise, the same oncologic principles that apply to open surgery for resectable colon cancer apply to laparoscopic. As a conclusion, in the hands of experienced laparoscopic surgeons, results from laparoscopic-assisted colectomy are oncologically similar to open colectomy in appropriately selected patients with resectable colon cancer, and rates of operative morbidity and mortality are comparable. Patients with tumours invading adjacent organs or with perforated or obstructed tumours are not appropriate candidates for laparoscopic colectomy. Although operative time is longer, laparoscopic colectomy offers the advantages of a minimally invasive procedure such as less pain or faster return of bowel function and shorter hospital stay. About a 11 to 20% of patients
96
Henar Núñez et al.
who undergo laparoscopic colectomy will require conversion to an open approach (111-113, 116).
Conclusion Improved results in the treatment of rectal cancer have been due to the introduction of TME and of the neoadjuvant therapies. All of these advances have been possible thanks to the improvement in diagnosis techniques which are more accurate in determining the depth of tumour penetration and also the nodal involvement. Both are key points to keep in mind to tailor the best treatment for the patient. On the other hand, a renewed interest has appear in improving surgical techniques, trying to optimize the outcomes with low morbidity and mortality and looking for the best patient´s quality of life after these potentially curable treatments. In fact, laparoscopic colectomy has been shown to be as safe and effective as open colectomy for the treatment of this disease and nowadays, the level of evidence supports its feasibility and recovery advantage as well as cancer equivalence outcomes between both techniques. Much efforts remain to be done to go on improving all these surgical procedures.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.
Rossi, H, and Rotthemberg, DA. Surgical treatment of colon cancer. Surg Onc Clin N Am, 2006, 15(1), 109. Lynch, ML, and Brand, M. 2005, Clinics in colon and rectal cancer surgery, 18(3), 163. NIH Consensus Conference. 1990, JAMA, 264, 1444. Sjövall, A, Granath, F, Cedermark, B, et al. Ann Surg Oncol 2007, 14, 432. Yu, YY. 2010, World J Gastrointest Oncol., 2(5), 229. Kim, MS, and Park, YJ.2007, World J Gastroenterol, 13, 4108. Langevin, JM, and Nivatvongs, S. 1984, Am J Surg, 147:330. Passman, MA, Pommier, RF, Vetto, JT. 1996, 39, 329. Rex, DK, Cutler, CS, Lemmel, GT et al. 1997, Gastroenterology, 112, 24. Winawer, SJ, Stewart, ET, Zauber, AG, et al. 2000, N Engl J Med, 342, 1766. Kim, JH, Kim, WH, Kim, TI et al. 2007, Yonsei Med J., 48:934. Bipat, S, Glas AS, Slors, FJM et al. 2004, Radiology, 232, 773. Chowdhury, P, Davies, R, and Roberts, A. Brown G ed. Cambridge University Press, 2007, p 136-156. Goh, V, Halligan, S, and Bartram, CI. 2004, Clin Radiol, 59(3), 215. Maier, A, and Fuchsjager, M. 2003, Eur J Radiol, 47(2), 89.
Colorectal cancer surgery
97
16. Moran, B, Brown, G, Cunningham, D, Daniels, I, Heald, R, Quirke, P, et al. 2008, Colorectal Dis., 10(3), 242. 17. Nelson, H, Petrelli, N, Carlin, A, et al. 2001, J Natl Cancer Inst., 93, 583. 18. Sonoda, T, and Milsom, JW. Chapter 34, Segmental Colon Resection 2006. 19. National Comprehensive Cancer Network (NCCN). Available in http://www.nccn.org/professionals/quality_measures/default.asp. 20. American Society of Clinical Oncology (ASCO). Available in http://www.asco.org/portal/site/ASCO/menuitem.5d1b4bae73a9104ce277e89a320 041a0/?vgnextoid1b08fcd4eb46c010 VgnVCM100000ed730ad1RCRD. 21. Otchy, D, Hyman, NH, Simmang, C, et al. 2004, Dis Colon Rectum, 47, 1269. 22. Curti, G, Maurer, CA, and Buchler, MW. 1998, Dig Surg., 15, 193. 23. Bruch, HP, Schwandner, O, Schiedeck, TH, and Roblick, UJ. 1999, Arch Surg., 384, 167. 24. Rouffet F, Hay JM, Vacher B, Fingerhut A, Elhadad A, Flamant Y, et al. Curative resection for left colonic carcinoma: hemicolectomy vs. segmental colectomy. A prospective, controlled, multicenter trial. French Association for Surgical Research. Dis Colon Rectum 1994;37:651–9. 25. Slanetz CA Jr. Effect of no touch isolation on survival and recurrence in curative resections for colorectal cancer. Ann Surg Oncol 1998;5:390–8. 26. Slanetz CA Jr, Grimson R. Effect of high and intermediate ligation on survival and recurrence rates following curative resection of colorectal cancer. Dis Colon Rectum 1997;40:1205–18; discussion 1218–9. 27. Tagliacozzo S, Tocchi A. Extended mesenteric excision in right hemicolectomy for carcinoma of the colon. Int J Colorectal Dis 1997;12:272–5. 28. Enker WE, Laffer UT, Block GE. Enhanced survival of patients with colon and rectal cancer is based upon wide anatomic resection. Ann Surg 1979;190:350–60. 29. Toyota S, Ohta H, Anazawa S. Rationale for extent of lymph node dissection for right colon cancer. Dis Colon Rectum 1995;38:705–11. 30. Malassagne B, Valleur P, Serra J, Sarnacki S, Galian A, Hoang C, et al. Relationship of apical lymph node involvement to survival in resected colon carcinoma. Dis Colon Rectum 1993;36:645–53. 31. Chang, GJ, Rodriguez-Bigas, MA, Skibber, JM, Moyer, VA. 2007, J Natl Cancer Inst, 99, 433. 32. Wong, SL, Ji, H, Hollenbeck, BK et al. 2007, JAMA, 298, 2149. 33. West, NP, Hohenberger, W, Weber, K et al. 2010, J Clin Oncol, 28, 272. 34. Compton, CC, Fielding, LP, Burgart, LJ, et al. 2000, 124, 979. 35. McGory, ML, Shekelle, PG, Ko, CY. 2006, J Natl Cancer Inst, 98, 1623. 36. Otchy, D, Hyman, NH, Simmang, C et al., 2004,47,1269. 37. AJCC Cancer Staging Manual, 7th ed, Edge, SB, Byrd, DR, Compton, CC, et al. (Eds), Springer, New York 2010, 143. 38. Scabini, S. 2010, World J Gastrointest Surg, 2(1), 6. 39. Morton, DL, Wen, DR, Wong, JH, Economou, JS, Cagle, LA, Storm, FK, et al. 1992, Arch Surg, 127, 392. 40. Bilchik, AJ, and Compton, C. 2007, Ann Surg., 245, 864. 41. Bembenek, AE, Rosenberg, R, Wagler, E, et al. 2007, Ann Surg., 245, 858.
98
Henar Núñez et al.
42. 43. 44. 45. 46. 47. 48. 49. 50.
Bilchik, AJ, Hoon, DS, Saha, S, et al. 2007, Ann Surg., 246, 568. Stojadinovic, A, Nissan, A, Protic, M, et al. 2007, Ann Surg, 245, 846. Bilchik, AJ, Hoon, DS, Saha, et al. 2007, Ann Surg 246, 568. De Haas, RJ, Wicherts, DA, Hobbelink, MG, et al. 2007, Ann Surg Oncol 14, 1070. Redston, M, Compton, CC, Miedema, BW, et al. 2006, J Clin Oncol., 24, 878. Lim, SJ, Feig, BW, Wang, H, et al. 2008, Ann Surg Oncol., 15, 46. Faerden, AE, Sjo, OH, Sjo, O, et al. 2008, Dis Colon Rectum, 51, 891. Cahill, RA, Bembenek, A, Sirop, S, et al. 2009, Ann Surg Oncol., 16, 2170. Baker, RJ, and Fisher, JE. El Dominio de la Cirugía. 4ª ed, vol 2, Cirugía del Colon y del Recto. Ed Panamericana 2004. Rouffet, F, Hay, JM, Vacher, B, et al. 1994, Dis Colon Rectum, 37, 651. Goligher, JC, Lee PWR, Macfie, J, Simpkins, KC, and Lintott, DJ. 1979, Surg Gynecol Obstet, 148, 516. Faerden, AE, Sjo, OH, Andersen, SN, Hauglann, B, Nazir, N, Gravdehaug, B, et al. 2008, Dis Colon Rectum, 51(6), 891. McGory, ML, Shekelle, PG, Ko, CY. 2006, J Natl Cancer Inst, 98, 1623. Otchy, D, Hyman, NH, Simmang, C, et al. 2004, Dis Colon Rectum, 47, 1269. Lopez, MJ. 2001, J Surg Oncol., 76,1. Poeze, M, Houbiers, JG, Van de Velde, CJ, et al. 1995, Br J Surg, 82, 1386. Hunter, JA, Ryan, JA, Jr, and Schultz, P. 1987, Am J Surg, 154, 67. Lehnert, T, Methner, M, Pollok, A, et al. 2002, Ann Surg., 235, 217. Nakafusa, Y, Tanaka, T, Tanaka, M, et al. 2004, Dis Colon Rectum, 47, 2055. Lehnert, T, Methner, M, Pollok, A, et al. 2002, Ann Surg., 235, 217. Darakhshan, A, Lin, BP, Chan, C, et al. 2008, Ann Surg., 247, 650. Govindarajan, A, Coburn, NG, Kiss, A, et al. 2006, J Natl Cancer Inst., 98, 1474. Pilipshen, SJ, Heilweil, M, Quan, SH, Sternberg, SS, and Enker, WE. 1984, Cancer, 53, 1354. Pilipshen, SJ, Heilweil, M, Quan, SH, Sternberg, SS, and Enker, WE. 1984, Cancer, 53, 1354. Adloff, M, Arnaud, JP, Schloegel, M, and Thibaud, D. 1985, Dis Colon Rectum, 28, 413. McDermott, FT, Hughes, ES, Pihl, E, Johnson, WR, Price, AB. 1985, Br J Surg., 72, 34. Moossa, AR, Ree, PC, Marks, JE, Levin, B, Platz, CE, and Skinner, DB, 1975, Br J Surg, 62, 727. Williams NS, Dixon MF, Johnston D. Reappraisal of the 5 centimetre rule of distal excision for carcinoma of the rectum: a study of distal intramural spread and of patients' survival. Br J Surg 1983;70:150–4. Williams NS. The rationale for preservation of the anal sphincter in patients with low rectal cancer. Br J Surg 1984;71:575–81. Pollett WG, Nicholls RJ. The relationship between the extent of distal clearance and survival and local recurrence rates after curative anterior resection for carcinoma of the rectum. Ann Surg 1983;198:159–63. Wolmark N, Fisher B. An analysis of survival and treatment failure following abdominoperineal and sphincter-saving resection in Dukes' B and C rectal
51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69.
70. 71.
72.
Colorectal cancer surgery
73.
74.
75.
76.
77.
78. 79. 80.
81. 82. 83.
84.
85.
86.
87.
99
carcinoma. A report of the NSABP clinical trials. Nation-al Surgical Adjuvant Breast and Bowel Project. Ann Surg 1986;204:480–9. Pollett WG, Nicholls RJ. The relationship between the extent of distal clearance and survival and local recurrence rates after curative anterior resection for carcinoma of the rectum. Ann Surg 1983;198:159–63. Mezhir JJ, Smith KD, Fichera A, Hart J, Posner MC, Hurst RD. Presence of distal intramural spread after preoperative combined-modality therapy for adenocarcinoma of the rectum: what is now the appropriate distal resection margin? Surgery. 2005;138:658–663; discussion 663-664. Chandler PJ, Orkin BA. Rectal carcinoma: operative treatment. In: Beck DE, Wexner SD, editors. Fundamentals of anorectal surgery. 2nd ed. London (U.K.): Saunders Co.; 1998. p. 301–67 Vernava AM III, Moran M, Rothenberger DA, Wong WD. A prospective evaluation of distal margins in carcinoma of the rectum. Surg Gynecol Obstet 1992;175:333–6. Andreola S, Leo E, Belli F, Lavarino C, Bufalino R, Tomasic G, et al. Distal intramural spread in adenocarcinoma of the lower third of the rectum treated with total rectal resection and coloanal anastomosis. Dis Colon Rectum 1997;40:25–9. Shirouzu K, Isomoto H, Kakegawa T. Distal spread of rectal cancer and optimal distal margin of resection for sphincter-preserving surgery. Cancer 1995;76:388–92 Schiessel R, Karner-Hanusch J, Herbst F, Teleky B, Wunderlich M. Intersphincteric resection for low rectal tumours. Br J Surg. 1994;81:1376–1378. Rullier E, Laurent C, Bretagnol F, Rullier A, Vendrely V, Zerbib F. Sphinctersaving resection for all rectal carcinomas: the end of the 2-cm distal rule. Ann Surg. 2005;241:465–469. Heald RJ, Husband EM, Ryall RD. The mesorectum in rectal cancer surgery: the clue to pelvic recurrence? Br J Surg. 1982;69:613–616. MacFarlane JK, Ryall RD, Heald RJ. Mesorectal excision for rectal cancer. Lancet. 1993;341:457–460. Maslekar S, Sharma A, Macdonald A, Gunn J, Monson JR, Hartley JE. Mesorectal grades predict recurrences after curative resection for rectal cancer. Dis Colon Rectum. 2007;50:168–175. [PubMed] 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. Cawthorn SJ, Parums DV, Gibbs NM, A'Hern RP, Caffarey SM, Broughton CI, et al. Extent of mesorectal spread and involvement of lateral resection margin as prognostic factors after surgery for rectal cancer. Lancet 1990;335:1055–9. Ng IO, Luk IS, Yuen ST, Lau PW, Pritchett CJ, Ng M, et al. Surgical lateral clearance in resected rectal carcinomas. A multivariate analysis of clinicopathologic features. Cancer 1993;71:1972–6. Goldberg PA, Nicholls RJ. Prediction of local recurrence and survival of carcinoma of the rectum by surgical and histopathological assessment of local clearance. Br J Surg 1995;82:1054–6.
100
Henar Núñez et al.
88. Adam IJ, Mohamdee MO, Martin IG, Scott N, Finan PJ, Johnston D, et al. Role of circumferential margin involvement in the local recurrence of rectal cancer. Lancet 1994;344:707–11. 89. Stocchi L, Nelson H, Sargent D, Allmer C, Pepper J, Wiesenfeld M. Impact of individual surgeon on rectal cancer outcome within 3 North Central Cancer Treatment Group (NCCTG) protocols [abstract]. Proc ASCO 1999;18:23a. 90. Heald RJ, Moran BJ, Ryall RD, Sexton R, MacFarlane JK. Rectal cancer: the Basingstoke experience of total mesorectal excision, 1978–1997. Arch Surg 1998;133:894–9. 91. Lopez-Kostner F, Lavery IC, Hool GR, Rybicki LA, Fazio VW. Total mesorectal excision is not necessary for cancers of the upper rectum. Surgery 1998;124:612–7; discussion 617–8. 92. Zaheer S, Pemberton JH, Farouk R, Dozois RR, Wolff BG, Ilstrup D. Surgical treatment of adenocarcinoma of the rectum. Ann Surg 1998;227:800–11. 93. Hida J, Yasutomi M, Maruyama T, Fujimoto K, Uchida T, Okuno K. Lymph node metastases detected in the mesorectum distal to carcinoma of the rectum by the clearing method: justification of total mesorectal excision. J Am Coll Surg 1997;184:584–8. 94. Cawthorn SJ, Parums DV, Gibbs NM, et al. Extent of mesorectal spread and involvement of lateral resection margin as prognostic factors after surgery for rectal cancer. Lancet 1990;335:1055–1059. 95. 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. 96. Marijnen CA, Nagtegaal ID, Kapiteijn E, et al. Radiotherapy does not compensate for positive resection margins in rectal cancer patients: report of a multicenter randomized trial. Int J Radiat Oncol Biol Phys 2003;55:1311–1320. 97. Parfitt JR, and Driman DK. The total mesorectal excision specimen for rectal cancer: a review of its pathological assessment. J Clin Pathol 2007;60:849-855. 98. Bernstein, TE, Endreseth, BH, Romundstad, P, and Wibe, A. 2009, Br J Surg., 96, 1348. 99. Wibe, A, Rendedal, PR, Svensson, E, Norstein, J, Eide, TJ, Myrvold, HE, et al. 2002, Br J Surg., 89, 327. 100. Nagtegaal, ID, Marijnen, CA, and Kranenbarg, EK. 2002, Am J Surg Pathol., 26, 350. 101. Saclarides, TJ, Bhattacharyya, AK, Britton-Kuzel, C, Szeluga, D, and Economou, SG. 1994, Dis Colon Rectum, 37, 52. 102. Sitzler, PJ, Seow-Choen, F, Ho, YH, and Leong, AP. 1997, Dis Colon Rectum, 40, 1472. 103. Ono, H, Kondo, H, Gotoda, T, Shirao, K, Yamaguchi, H, Saito, D, et al. 2001, Gut, 48, 225. 104. Uraoka, T, Kato, J, Ishikawa, S, Harada, K, Kuriyama, M, Takemoto, K, et al. 2007, Gastrointest Endosc, 66, 836. 105. Saito, Y, Sakamoto, T, Fukunaga, S, Nakajima, T, Kiriyama, S, and Matsuda, T. 2009, Dig Endosc., 21 Suppl 1: S7.
Colorectal cancer surgery
101
106. Zhou, PH, Yao, LQ, and Qin, XY. 2009, Surg Endosc., 23, 1546. 107. Lee, YM, Law, WL, Chu, KW, and Poon, RT. 2001, J Am Coll Surg., 192, 719. 108. Chiappa, A, Zbar, A, Biella, F, and Staudacher, C. 2000, Am Surg., 2000, 66, 619. 109. Lau, PW, Lo, CY, and Law, WL. 1995, Am J Surg., 169, 406. 110. Jacobs, M, Verdeja, JC, and Goldstein, HS. 1991, Surg Laparosc Endosc., 1, 144. 111. Lacy, AM, Delgado, S, Castells, A, et al. 2008, Ann Surg., 248, 1. 112. Anonymous. 2004, N Engl J Med, 350, 2050. 113. Fleshman, J, Sargent, DJ, Green, E, et al. 2007, Ann Surg., 246, 655. 114. Buunen, M, Veldkamp, R, et al. 2009, Lancet Oncol, 10, 44. 115. Liang, JT, Huand, KC, Lai, HS, et al. 2007, Ann Surg Oncol, 14, 109. 116. Jayne, DG, Guillou, PJ, Thorpe, H, et al. 2007, J Clin Oncol, 25, 3061. 117. Leung, KL, Kwok, SP, Lam, SC, et al. 2004, Lancet, 363, 1187. 118. Bonjer, HJ, Hop, WC, Nelson, H, et al. 2007, Ann Surg., 142, 298. 119. Jackson, TD, Kaplan, GG, Arena, G, et al. 2007, J Am Coll Surg., 204, 439. 120. Kuhry, E, Schwenk, WF, Gaupset, R, et al. 2008, Cochrane Database Syst Rev, CD003432. 121. Anonymous. 2004, N Engl J Med, 350, 2050. 122. Fleshman, J, Sargent, DJ, Green, E, et al. 2007, Ann Surg., 246, 655. 123. Bonjer, HJ, Hop, WC, Nelson, H, et al. 2007, Ann Surg., 142, 298. 124. Jackson, TD, Kaplan, GG, Arena, G, et al. 2007, J Am Coll Surg., 204, 439.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 103-138 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
5. Adjuvant systemic treatment for colon cancer José María Vieitez de Prado1, Paula J Fonseca1 and Madalina Frunza2 1
Medical Oncology Department, 2Surgery Department, Central University Hospital of Asturias (Oviedo), Spain
Abstract. The prognosis associated with colon cancer remains poor, as less than 60% of patients are potentially curable with radical resection of the primary tumour [1]. For this reason, adjuvant therapy has been combined with surgery, and this has made a significant impact on cure rates for this disease. The benefits of adjuvant therapy in colon cancer will be outlined in the following chapter.
Introduction Bolus of fluorouracil and leucovorin has been accepted as the standard adjuvant therapy in stage III colon cancer for many years. New drugs such as irinotecan, oxaliplatin and oral fluoropyrimidines have all completed phase III randomised evaluation in colon cancer and several of these studies have been reported recently. As a result of these studies, oxaliplatin-based chemotherapy has emerged as the new standard of care in adjuvant treatment of stage III colon cancer. Correspondence/Reprint request: Dr. José María Vieitez de Prado, Medical Oncology Department, Central University Hospital of Asturias (Oviedo), Spain. E-mail: josemariavieitez@yahoo.es
104
JosĂŠ MarĂa Vieitez de Prado et al.
The advent of monoclonal antibodies such as cetuximab and bevacizumab has further broadened the treatment horizon for CRC and they were the focus of the recent randomised studies in adjuvant therapy of this disease. In stage II colon cancer, adjuvant treatment remains controversial, but it is recommended in several subsets including poorly differentiated histology, T4 lesions, bowel perforation presentation and inadequately sampled lymph nodes (<13). This chapter will review deeply all these treatments, thereby assisting clinicians in deciding the optimal adjuvant treatment for patients in routine clinical practice.
1. The history of adjuvant therapy in colon cancer After many years of belief that adjuvant chemotherapy was of no benefit to patients with resectable adenocarcinoma of the colon, studies published in the 1980s evoked a great change in the surgical and medical approaches to this disease. Chemotherapy in colon cancer began in the 1950s with use of the drugs, thiotepa, floxuridine and the fluoropyrimidines in patients with or without curative resections. A major therapeutic impact was not observed with these drugs, probably because knowledge concerning their pharmacological effects and mechanism of action was limited, the dose intensity offered was insufficient, and the patient groups studied were too small. Although the benefits noted were minimal, they were enough to stimulate research into finding better agents and combinations, leading to an improvement in adjuvant chemotherapy. The 1970s saw the introduction of the concept of biomodulation of 5-fluorouracil (5FU) by several different agents (levamisole (LEVA), leucovorin (LV), methotrexate, interferon, and N-(phosphonacetyl)L-aspartate (PALA)), as well as immunotherapy, represented at first by the well-known Bacillus Calmette-GuĂŠrin (BCG) vaccine. The studies using methotrexate, interferon, or PALA with 5FU concluded that those modulators conferred no statistically significant benefit, and at the same time, increased toxicity. In addition, at around the same time, in the American National Surgical Adjuvant Breast and Bowel Project (NSABP) CO-1 study [2], conducted between 1977 and 1983, 1116 patients were randomly assigned to observation, chemotherapy with MOF (MeCCNU or semustine, plus vincristine and 5FU), or immunotherapy with BCG. This was the first largescale trial that found a small increase in disease-free survival (DFS) and overall survival (OS) in patients receiving chemotherapy, especially in those with right-sided colonic tumours, while BCG did not influence any of the parameters studied. Despite this benefit, it was noted that meCCNU and other
Adjuvant systemic treatment for colon cancer
105
alkylating agents could be leukemogenic and have renal toxicity, and so in subsequent studies these kinds of drugs were abandoned.
1.1. 5FU plus levamisole Two American trials published in 1989 and 1990 showed significant benefits in using adjuvant chemotherapy, and subsequently changed the standard of care for patients with resectable colon cancer. The first, a study designed by the North Central Cancer Treatment Group (NCCTG) [3], assigned 408 patients (35% stage II and 65% stage III) to observation or to receive 1 year of the immunomodulatory agent LEVA with or without 5FU (table 1). When given with 5FU, patients treated with LEVA (5FU/LEVA), showed a small improvement in DFS and OS and with only mild toxicity. Results from this study stimulated the design of the next, the Intergroup (INT)-0035 trial, with 1,247 patients (25% stage II and 75% stage III) who were randomised to observation or to LEVA with or without 5FU (using the same regimens as in the NCCTG study). 5FU/LEVA reduced the recurrence rate by 33% compared with surgery alone (95% confidence interval (CI): 16â&#x20AC;&#x201C;47%; hazard ratio (HR): 0.67, p= 0.0007) and reduced the risk of cancer recurrence by 40% (p<0.0001) [4,5]. The most common toxicities reported with this combination were mild and included myelosuppression, mild elevation of hepatic transaminases, dysgeusia, arthralgia, neurotoxicity and depression. In 1990, the magnitude of the benefit associated with the combination of 5FU/LEVA demonstrated in both of these studies resulted in FDA (Food and Drugs Administration) approval of the regimen for stage III CRC, and in recommendation by the USA National Cancer Institute consensus conference for its use as standard adjuvant therapy for stage III colon cancer [6].
1.2. 5FU-based regimens Once 5FU/LEVA had become the acknowledged standard adjuvant therapy in CRC, it was used in studies conducted in the 1990s as the reference for new protocols. Between 1990 and 1999, results of large clinical trials defined the relative efficacy of 5FU-based regimens in a variety of doses and schedules in advanced disease. The next step was testing them as adjuvant therapies; this is presented in the six relevant studies below. The first, the International Multicentre Pooled Analysis of Colon Cancer Trials (IMPACT) study [7], involving 1,526 patients, enrolled three independent trials in Italy, Canada and France, measuring the efficacy of 5FU and high-dose LV (5FU/LV) after surgery for stage II (56%) and stage III
106
José María Vieitez de Prado et al.
Table 1. Chemotherapy regimens. REGIMEN FLUOROPYRIMIDINES 5FU + LEVA NCCTG Mayo Clinic: FULV NCCT 89-4651 High-dose LV regimen: mFULV IMPACT Roswell Park: FL NSABP C-O4 De Gramont: LV5FU2 GERCOR C96.1 Raltitrexed PETTAC1 Capecitabine X-ACT UFT + LV NSABP C-O6 OXALIPLATIN FOLFOX4 MOSAIC FOLFOX6
mFOLFOX6 NSABP C-O6
FOLFOX7
mFOLFOX7
FLOX NSABP C-O7 XELOX NO16968 IRINOTECAN IFL CALGB-89803
FOLFIRI PETACC3-EORTC BEVACIZUMAB NSABP CO-8 CETUXIMAB NCCTG-N0147
CHEMOTHERAPY LEVA 150 mg/d 3 d + 5FU 450 mg/m2/d bolus 5 d and next Q1w starting on d28 LV 20 mg/m2/d bolus d1-5 + 5FU 425 mg/m2/d iv bolus d1-5 LV 200 mg/m2/d iv d1-5 + 5FU 400 mg/m2/d iv 15´d1-5 LV 500 mg/m2 iv over 2 h + 5FU 500 mg/m2 bolus 1 h after the start of LV LV 200 mg/m2 over 2 h + 5FU 400 mg/m2 bolus followed by a 600 mg/m2 IVCI 22 h 3 mg/m2 d1 1250 mg/m2 po bid UFT 100 mg/m2 po every 8 h + LV 30 mg po every 8 h LV 200 mg/m2 iv over 2 h + 5FU 400 mg/m2 bolus and then 600 mg/m2 IVCI 22 h + Oxaliplatin 85 mg/m2 iv over 2 h only d1 LV 400 mg/m2 iv over 2 h d1 + 5FU 400 mg/m2 bolus d1 followed by 2400 mg/m2 IVCI 46 h + Oxaliplatin 100 mg/m2 iv over 2 h d1 LV 400 mg/m2 iv over 2 h d1 + 5FU 400 mg/m2 bolus d1 followed by 2400 mg/m2 IVCI 46 h + Oxaliplatin 85 mg/m2 iv over 2 h d1
SCHEDULE LEVA: Q2w 5FU: Q1w, 1 year Q4-5w, 6 cycles Q4w, 6 cycles Q1w for 6 of 8 weeks, 3-4 cycles D1 and 2, Q2w, 12 cycles Q3w, 8 cycles 14 d Q3w, 8 cycles 4 weeks Q5w, 5 cycles
D1 and 2, Q2w, 12 cycles Q2w, 12 cycles Q2w, 12 cycles
LV 400 mg/m2 iv over 2 h d1 + followed by 5FU 2400 mg/m2 IVCI 46 h + Oxaliplatin 130 mg/m2 iv over 2 h d1
Q2w, 12 cycles
LV 400 mg/m2 iv over 2 h d1 + followed by 5FU 2400-3000 mg/m2 IVCI 46 h + Oxaliplatin 100 mg/m2 iv over 2 h d1
Q2w, 12 cycles
LV 500 mg/m2 iv over 2 h + 5-FU 500 mg/m2 bolus + Oxaliplatin 85 mg/m2 iv 2 h before 5FU Capecitabine 1000 mg/m2 po bid x 14 d + Oxaliplatin 130 mg/m2 iv d1 LV 20 mg/m2 bolus + 5FU 500 mg/m2 bolus + Irinotecan 125 mg/m2 iv LV 400 mg/m2 iv over 2 h d1 + 5FU 400 mg/m2 bolus d1 followed by 2400 mg/m2 IVCI 46 h + Irinotecan 180 mg/m2 iv over 90´ d1 5-10 mg/kg iv over 30-90´ d1 400 mg/m2 iv over 2 h d1, and then 250 mg/m2 iv over 1 h
5FU and LV, Qw Oxaliplatin week 1, 3, 5 Q8w, 3 cycles Q3w, 8 cycles
Qw for 4 of 6 weeks, 3-4 cycles Q2w, 12 cycles Q2w, 24 cycles Q1w, 6 months
Qw= weekly, Q2w= every 2 weeks, Q3w= every 3 weeks, ´= minute, h=hour, d=day, bid= twice a day, po=oral, iv=intravenous, IVCI=intravenous continuous infusion
Adjuvant systemic treatment for colon cancer
107
(44%) colon cancer. The three trials used the same regimen (table 1) and compared it with surgery alone. 5FU/LV was associated with a significant increase in 3-year event-free survival (EFS) (from 62% to 71%, HR: 0.67, p<0.0001) and OS (from 78% to 83%; HR 0.77, p=0.03), as well as with a significant reduction in both mortality by 22% (95% CI: 3â&#x20AC;&#x201C;38%; p=0.029) and in events by 35% (95% CI 22â&#x20AC;&#x201C;46%; p<0.0001) (table 2). Grade 4 toxicities, the most common of which being of the gastrointestinal (GI) tract, were reported in less than 3% of the patients, and more than 80% of patients completed the planned treatment. These analyses concluded that 5FU plus high-dose LV was an effective, well-tolerated 6-month adjuvant therapy for colon cancer. The second, the American NSABP CO-4 study [8], compared the Roswell Park regimen (FL) (table 1), with or without LEVA, with the standard regimen of 5FU/LEVA for a period of 12 months in 2,151 patients (41% stage II and 58% stage III). Results showed that 5FU/LV was superior to 5FU/LEVA in DFS (64% versus 60%; p=0.04) and 5-year OS (74% versus 69%; p=0.07). The three-drug regimen, 5FU/LV/LEVA, was intermediate in efficacy (table 2). This study suggested that 5FU plus high-dose LV for six cycles was superior to the standard 12 months of 5FU/LEVA. Grade 3 toxicity was reported in 36% of patients who received 5FU/LV, 38% of those who received 5FU/LV/LEVA and 28% of those who received 5FU/LEVA. The third, the German adjCCA-01 study [9], randomised 680 patients with stage III colon cancer to 12 months of 5FU/LV (5FU 450 mg/m2 plus LV 100 mg/m2, 5 days every 4 weeks) or LEVA, and demonstrated that 5FU/LV improved both DFS and OS (p=0.04 and 0.01, respectively). The fourth study, the American NCCTG 894651 trial [10], involved 915 patients with stage II and III CRC and compared the standard regimen of 5FU/LEVA, with the Mayo Clinic regimen of 5FU plus low-dose LV (FULV) (table 1), with or without LEVA, both regimens administered for 12 or 6 months. The study demonstrated that 12 months of 5FU/LEVA was equally as effective as 6 months of the three-drug regimen, 5FU/LEVA/LV, and more effective than 6 months of 5FU/LEVA (table 2). The toxicity of the tripledrug regimen was greater in terms of diarrhea and stomatitis. In the fifth, the American INT-0089 study [11], which involved 3,794 patients with stage II (20%) and stage III (80%) colon cancer, the standard regimen of 5FU/LEVA for 12 months was compared with both the 24-week (6 months) regimen of 5FU plus low-dose LV (Mayo Clinic regimen, table 1), with or without LEVA, and the 32-week (8-months) regimen of 5FU plus high-dose LV (table 1). The results demonstrated that the 5-year DFS (60% to 59%) and OS (65% to 63%) rates for patients with high-risk stage II (obstructing and/or perforated, node-negative lesions) and stage III colon
108
José María Vieitez de Prado et al.
cancers were no different when using 5FU plus high- or low-dose LV, and no benefit was registered with the addition of LEVA (table 2). These results contrasted somewhat with those of the NSABP CO-4 study [8], which showed that 5FU plus high-dose LV was superior to 5FU/LEVA. It is important to mention that, in the NSABP CO-4 trial [8], 5FU/LV therapy was administered over six cycles (12 months), whereas in the INT-0089 trial [11] the same regimen was administered only for four cycles (8 months). This trial also revealed clinically important differences in toxicity profiles among the various treatments according to age and gender. 5FU plus low-dose LV regimens, with or without LEVA, were associated with a significantly higher incidence of mucositis than 5FU/LEVA or 5FU plus high-dose LV regimens. On the other hand, diarrhea was almost three times more common with 5FU plus high-dose LV than with 5FU/LEVA. Patients older than 70 years (unpublished data) and females had significantly higher rates of mucositis and neutropenia; therefore, clinicians should pay more attention when using 5FU/LV regimens in these groups of patients, and should only consider 5FU/LEVA as an alternative once severe toxicity occurs. The sixth, the INK Colon Trial Group study [12], randomised 500 patients with stage III colon cancer to 5FU/LEVA or to LV (20 mg/m2 i.v.), 5FU and LEVA for 1 year. The addition of low-dose LV increased toxicity (especially mucositis and conjunctivitis) without a significant increase in 5-year DFS (without LV: 49%, LV-group: 46%; log-rank test, p=0.86) or in OS (55% versus 59%; log-rank test, p=0.96) (table 2). Therefore, the study concluded that LV, 5FU and LEVA were not recommended in a 12-month adjuvant regimen of stage III colon cancer. In summary, available data in 1999 indicated that 5FU/LV given for 6 months was at least as effective as 5FU/LEVA for 12 months. Only the NSABP CO-4 study [8] showed that 5FU/LV was better than 5FU/LEVA. Thus, use of the 24-week (6 months) NCCTG schedule [10], the Mayo Clinic regimen (“FULV”: table 1), the 32-week (8 months) NSABP schedule [8] or the Roswell Park regimen (“FL”: table 1) were then considered the preferred adjuvant treatments in patients with resected stage III colon cancer, leading to a 5% to 10% improvement in OS. Subsequent studies performed between 1996 and 2004 were designed to clarify the optimal duration of therapy, the advantage of high- or low-dose LV and the effectiveness of bolus compared with infusional treatment. The five most relevant are described below. The first was a German trial [13], which analysed bolus or infusional 5FU, with or without LV, in stage III colon cancer using three different regimens: 1) 5FU 450 mg/m2 and LV 100 mg/m2, 5 days every 4 weeks; six cycles; 2) 24 h infusion of high-dose 5FU 2600 mg/m2 and LV 500 mg/m2,
Adjuvant systemic treatment for colon cancer
109
weekly for 6 of 8 weeks, two cycles (16 weeks, 4 months); and 3) 24 h infusion of high-dose 5FU 2600 mg/m2, weekly for 6 of 8 weeks, two cycles (16 weeks, 4 months). The study enrolled only 145 patients (although 325 patients per treatment arm were programmed) between 1997 and 2001, and it was stopped in 2001 due to the slow recruitment. No statistically significant difference in efficacy and the percentage of patients with severe toxicity was found between the three treatment arms (table 2). There was, however, a significant difference in the type of toxicity as we will describe below. The second was the French Groupe d’Etude et de Recherche Clinique en Oncologie Radiotherapies (GERCOR) C96.1 study [14,15], which enrolled 905 patients with stage II (43%) and III (57%) colon cancer over 3 years (1996–1999). The 2 × 2 factorial study compared a semimonthly De Gramont’s infusion regimen (LV5FU2: table 1) with a monthly regimen (mFULV with high-dose LV European regimen: table 1), while also comparing the 24 week (6 months) versus 36 week (9 months) duration of each regimen. This mFULV reference regimen differs from the monthly Mayo Clinic regimen by a higher dose of LV (200 versus 20 mg/m2), a different period of administration (15 minutes versus bolus), and a slightly lower dose per day of 5FU (400 versus 425 mg/m2). At 3 years, the DFS and OS had not been influenced by the duration of the therapy (6 or 9 months), by the type of administration (bolus or infusional) or by the frequency of administration (semi-monthly or monthly) (table 2). Although patients in the LV5FU2 group received twice the dose of 5FU compared with those in the mFULV group (930 versus 463 mg/m2/week), the incidence of grade 3–4 neutropenia, diarrhea and mucositis was higher with mFULV (p=0.001), whereas grade 3–4 nausea and vomiting were lower in the LVFU2 group, although not significantly (p=0.093). In later studies using De Gramont’s schedule, an increase in hand-foot syndrome and in toxicity related to the implantable port was noticed; however, the study detailed here did not report any of these adverse events. The overall rate of toxicity is lower with continuous infusion of 5FU than with bolus, and the toxicity patterns of these two means of administration are completely different. The present trial does not offer any evidence of a difference in efficacy between the semimonthly LVFU2 and monthly FULV regimens. However, the small size of the study precludes any definitive conclusion about relative survival benefit. We cannot conclude that these two treatment regimens are equivalent, firstly because the upper boundary of the CI falls outside 0.8 to 1.25, often regarded as an acceptable region of equivalence, and also because the sample size is insufficient to confirm small benefits in DFS and OS (within this interval, the relative difference in the HR does not exceed 20%).
110
José María Vieitez de Prado et al.
The third, a British study [16] conducted between 1993 and 2003, randomised 801 patients with stage II (44%) and III (54%) colon cancer to the 6-month Mayo Clinic regimen or to a 12-week protracted venous infusion (PVI) of 5FU alone (300 mg/m2 daily). The trial showed that reducing treatment duration to 12 weeks did not influence the efficacy of infused 5FU (table 2), although it did reduce the probability of neutropenia, diarrhea and mucositis (p<0.0001), as well as nausea and vomiting (p=0.46), but with a higher incidence of hand-foot syndrome, albeit not significantly (p=0.09). The major side effects due to the implantable port in the PVI FU regimen were pneumothorax (0.6%), septicemia (1.2%) and thromboses (7%). In the fourth, the American Southwest Oncology Group (SWOG) 9415/INT-0153 study [17], 1,135 patients with stage II (16%) and III (84%) colon cancer were enrolled between 1994 and 1999, and received the 6-month Mayo Clinic regimen plus LEVA or the PVI low-dose 5FU (250 mg/m2 day for 56 days every 9 weeks, three cycles) plus LEVA (50 mg three times a day (tid) 3 days every other week). There were no differences in the 5-year DFS and OS between the two regimens. This trial showed that neither LEVA nor continuous infusion of 5FU added any benefit to the treatment (table 2), which, as mentioned by the authors themselves, could also have been compromised by hand-foot syndrome and thromboses related to catheters. The estimated 5-year DFS was 78%, 67% and 47%, while the 5-year OS was 83%, 74% and 55% for stages N0, N1 and N2–3, respectively, based on the 4th TNM edition [18]. The authors of this trial suggest that the poor survival of patients with four or more positive nodes (N2-3) could have been the consequence of including some with undetected metastases in the study. At the same time, they recommend a more rigorous imaging staging to identify the presence of distant metastases, in which case the treatment would be different, and more intensive chemotherapy might be considered. The fifth, the Pan European Trials in Adjuvant Colon Cancer 2 (PETACC-2) study [19] between 1997 and 2004, randomised 1,624 patients with stage III colon cancer to receive the Mayo Clinic FULV regimen, PVI 5FU, the weekly high-dose German Arbeitsgemeinschaft Internische Onkologie (AIO) regimen, the De Gramont LV5FU2 regimen, or the Spanish weekly high-dose TTD regimen. There was no difference among them, either in the 5-year DFS (p=0.9) or in the 5-year OS (p=0.44) (table 2). Grade 3–4 neutropenia, mucositis and diarrhea were more frequent in the bolus arm, while hand-foot syndrome was more frequent in the infusion arm. This trial showed no increased benefit for continuous infusion, although this delivery was associated with less toxicity.
Adjuvant systemic treatment for colon cancer
Table 2. Randomised trials with 5FU.
111
112
José María Vieitez de Prado et al.
1.3. Portal vein and peritoneal chemotherapy It is well known that colon carcinoma dissemination frequently involves the liver via the portal vein system; a fact that led to the hypothesis that cytotoxic therapy delivered into the portal vein could destroy microscopic metastases and, therefore, increase the OS. The first study of interest regarding portal vein infusion was published by Taylor et al in 1977 [20] and updated in 1985. It randomised 243 patients to 5FU and heparin via the portal vein or to surgery alone. Patients receiving intraportal therapy experienced a significantly increased 5-year survival rate (p=0.002), although this was limited to patients with stage II colon cancer. Between 1984 and 1988, NSABP CO-2, an interesting study [21] involving 1,158 patients, demonstrated not only an increase in DFS but also in OS, even though it did not have a significant impact on liver metastases. It was suggested that the benefit of intraportal therapy was probably due to the timing of administration (intraportally when given immediately after surgery, whereas 30 to 40 days postoperatively it was given as intravenous chemotherapy) rather than to the way it was administered. The concept of “immediate postoperative delivery of chemotherapy” was being tested in INT-0136, a phase III study in which 7 days of 5FU administered by intravenous infusion was initiated within 24 h of surgery and compared with a standard 5FU/LEVA program started within 35 days of the colon resection. The Swiss Group for Clinical Cancer Research (SAKK) study [22], involved 533 patients who were randomised to a single course of intraportal mitomycin (10 mg/m2, one dose) and 5FU (500 mg/m2 24-h continuous infusion for 7 days) starting immediately after surgery, or to surgery alone. The differences in OS and DFS were attributed to a consistent reduction of recurrences (local and distant metastases) in the treated group, rather than to a reduction of the liver metastases only, concluding that part of the benefit was due to the systemic effects of the intraportal chemotherapy. A meta-analysis published in 1997 [23] of 4,000 patients belonging to 10 studies suggested that despite showing a statistically significant 4.7% improvement in 5-year OS (p=0.006), intraportal infusion was too toxic, too complicated to administer and clinically irrelevant. Randomised studies involving more patients are needed to confirm the results of this meta-analysis before considering intraportal infusion as a standard treatment. A phase III Gastrointestinal Tract Cancer Cooperative Group of the European Organization for Research and Treatment of Cancer (GITCCG-EORTC 1983-1987) trial [24] included 199 patients from different European countries who were randomised to intraportal heparin alone (5000 IU daily × 7 days), to intraportal heparin (5000 IU daily × 7 days) and intraportal 5FU (500 mg/m2
Adjuvant systemic treatment for colon cancer
113
daily Ă&#x2014; 7 days), or to surgery (the control group). There were no differences in the postoperative complications and toxicity between those that received heparin or those receiving heparin and 5FU. Although intraportal 5FU infusion was safe and had a tolerable toxicity, there was no statistically significant improvement in the 5-year OS and DFS. A more recent multicentre European study [25] randomised about 1,500 patients to the same heparin and intraportal 5FU schedule or to surgery alone, obtaining no significant differences between survival rates and liver dissemination. An Italian ACOI/GIVIO/GISCAD group study [26] demonstrated that while 5FU-based adjuvant chemotherapy after surgical resection of colon cancer was the standard treatment, the best route of administration, either systemic or regional, remained controversial. In this trial, 1,084 eligible patients were randomised to the intraportal regimen (i.e., intraportal heparin and 5FU mentioned previously), to systemic (SY) (bolus 5FU 370 mg/m2 and LV 100 mg/m2, both daily for 5 days every month for six cycles, with treatment initiated 15â&#x20AC;&#x201C;35 days after surgery), or to intraportal and SY. The OS and EFS rates were similar in all regimens and the combined regimen was no better than the single regimen alone. In conclusion, the difficulties related to finding eligible patients, accomplishing the protocol and achieving homogeneity among different centres, makes it difficult to attribute small benefits to intraportal adjuvant therapy, at this point, it should, therefore, only be considered as an investigational approach rather than a standard one [27]. The treatment of peritoneal carcinomatosis is based on cytoreduction followed by hyperthermic intraperitoneal chemotherapy (HIPEC) and is combined with adjuvant chemotherapy. In 2003, a randomised trial [28], comparing systemic chemotherapy alone with cytoreduction followed by HIPEC and systemic chemotherapy, showed that the latter significantly increased the survival of patients affected by peritoneal carcinomatosis from CRC. However, it is still not known if intraperitoneal chemotherapy could prevent peritoneal dissemination in patients with stage II or III colon cancer. The administration of chemotherapy through the hepatic artery has been examined using arterial injection of Tc99 [29]. This study demonstrated that the concentration of drugs reaching liver metastases was higher when cytotoxic agents were delivered by this route than through the portal vein. This is because the main blood supply of the normal liver is provided by the portal vein, while metastases are mainly irrigated by the hepatic artery. However, despite this benefit, only a small number of studies of drug delivery via the hepatic artery have been completed because of the difficulty of the procedure.
114
JosĂŠ MarĂa Vieitez de Prado et al.
1.4. Monoclonal antibodies: 17-1A [30] Despite enthusiasm for evaluating monoclonal antibodies as adjuvant therapy for colon cancer, as demonstrated, for example, by the German phase III trial of 17-1A monoclonal antibody (edrecolomab), which reported a survival benefit resulting from the prevention of distant metastatic disease, such findings have not been confirmed by two large, international studies. These studies, which recruited 4,600 patients, investigated edrecolomab in combination with 5FU [31,32]. Some authors have suggested that the lack of antibody activity could be explained by the fact that chemotherapy has the potential to alter the immune response which could mask any benefit from an immunotherapeutic agent in CRC.
1.5. Vaccines A randomised Eastern Cooperative Oncology Group Study E5283 phase III trial of adjuvant immunotherapy with an autologous tumour cell- BCG vaccine (412 patients) showed no significant clinical benefit [33]. Targeting CEA epitopes as a vaccine in colon cancer also failed to show benefit in phase I studies and its use therefore remains limited to an investigational setting [34].
1.6. Other antimetabolites At the beginning of the 21st century, the inconvenience and cost of a central venous line and ambulatory infusion pump prompted research into other antimetabolites, such as raltitrexed and oral fluoropyrimidines (capecitabine and UFT). A) Raltitrexed The European PETACC-1 trial [35] was closed prematurely when 17 (1.9%) raltitrexed-related deaths, while using the standard approved drug dose in metastatic disease, were reported. Moreover, this study failed to demonstrate non-inferiority of raltitrexed compared to 5FU and LV for relapse-free survival and OS in stage III colon cancer (table 3). Despite this, another study showed that raltitrexed adjuvant therapy could be administered safely and effectively in patients for whom further 5FU was contraindicated, such as those with heart or vascular disease [36].
Adjuvant systemic treatment for colon cancer
115
B) Oral fluoropyrimidines Two strategies were developed to avoid the erratic intestinal absorption of 5FU after oral administration: the coadministration of an inhibitor of dihydropyrimidine dehydrogenase (DPD)-uracil with oral 5FU (tegafur), together known as UFT (uracil/tegafur, ratio 1:4), and the administration of a 5FU prodrug that is not catabolised by DPD (capecitabine). UFT Between 1997 and 1999, the NSABP C-06 trial [37] compared UFT and LV (table 1) with the Roswell Park regimen in 1,608 patients with stage II (47%) and III (53%) colon cancer and demonstrated similar rates of DFS and OS between the two treatment arms, with a comparable toxicity profile (table 3). Although UFT was used in Europe and Asia because it was convenient to administer, it was withdrawn by the manufacturer in the United States. Nowadays, it has been replaced by capecitabine in most European centres. Capecitabine Between 1998 and 2001, the International X-ACT phase III trial [38], involving 1,987 patients with stage III colon cancer, compared capecitabine with the Mayo Clinic regimen (table 1). Despite their equivalence in DFS, which led to its approval by the EMEA (European Medicines Agency), it is still uncertain as to whether the results would have been equivalent if capecitabine had been compared with a more tolerable schedule of intravenous continuous 5FU infusion (table 3). Table 3. Randomised trials with other antimetabolites, raltitrexed and oral fluoropyrimidines. Study
Patients Arms
PETACC1 Raltitrexed
1921
NSABP C-06 UFT-LV
X-ACT Capecitabine
5-year p DFS
FULV 6 m Raltitrexed 6 m
-
1608 II (47%), III
FL 6 m UFT+LV 6 m
68.2% 67%
1987 II (0%), III
FULV 6 m Capecitabine 6 m
60.6% 64.2%
5-year OS
Ns
-
0.96
78.7% 78.5%
0.12
77.6% 81.3%
p
Conclusions
Ns
DFS and OS: Ns at 4 years
0.9
Equivalent efficacy and toxicity
0.051 3-year DFS, OS: equivalent. Lower toxicity: Capecitabine
116
José María Vieitez de Prado et al.
Nowadays, for patients who are candidates only for monotherapy without oxaliplatin, oral fluoropyrimidines might be used instead of the more toxic, less cost-effective intravenous 5FU/LV, which is also less favoured by patients.
1.7. New drugs associated with fluoropirimidines (Oxaliplatin and Irinotecan) A) Oxaliplatin Oxaliplatin (L-OHP) is active as a single agent and is synergistic when combined with 5FU, possibly due to oxaliplatin-induced downregulation of thymidylate synthetase (TS). Three phase III clinical trials evaluated oxaliplatin and a fluoropyrimidine in the adjuvant treatment of colon cancer. In Europe, the Multicenter International Study of Oxaliplatin and 5-FU and LV (FOLFOX4) in the Adjuvant Treatment of Colon Cancer (MOSAIC) study [39], randomised 2,246 patients, with stage II (40%) and stage III (60%) colon cancer to receive 6 months of De Gramont´s LV5FU2 regimen with or without oxaliplatin 85 mg/m2 on day 1 (table 1) between 1998 and 2001. The 5-year follow-up confirmed an increase in the DFS for patients who received oxaliplatin (73.3% versus 67.4%, HR 0.80, p=0.003). After a longer followup, there was a trend for improvement in the 6-year OS for the whole population, which was significant for stage III (73.3% versus 67.4%; HR 0.80; 95% CI: 0.65–0.97; p=0.023) (table 4). In the USA, Canada, Australia and New Zealand, the NSABP C-07 trial [40] randomised 2,492 patients with resected stage II (29%) or stage III (71%) colon cancer to receive 6 months of the Roswell Park FL regimen with or without oxaliplatin (85 mg/m2 weekly over 6 weeks, followed by a 2-week rest period) (table 1) between 2000 and 2002. There was a benefit in the 3-year DFS rate (76.1% versus 71.8%; HR 0.80; 95% CI: 0.69–0.93; p=0.0034) for patients who received oxaliplatin (FLOX regimen), except for those older than 65 or those without pathological adenopathies; however, there was still no statistically significant difference in the 6-year OS (table 4). Comparing the toxicity profile of all patients treated in both studies with 5FU and oxaliplatin, the incidence of neuropathy was significantly lower in the NSABP C-07 trial than in the MOSAIC trial (6.9% versus 12%, p<0.0001), as a result of the lower accumulative dose of oxaliplatin given (676 mg/m2 in C07 versus 894 mg/m2 in Mosaic) than that planned (1020 mg/m2 versus 900 mg/m2 respectively). Although grade 1 neurotoxicity persisted for more than 2 years in at least 10% patients in the oxaliplatin arm
Adjuvant systemic treatment for colon cancer
117
in both studies, grade 2â&#x20AC;&#x201C;3 neuropathy decreased to 1.3% at 6 months after FOLFOX4 and to 0.5% at 6 months after FLOX [41]. Despite this, the benefit in the DFS was similar between both oxaliplatin regimens; therefore, FLOX may offer an advantage to patients who are at risk of neuropathy. However, grade 3â&#x20AC;&#x201C;4 diarrhea was more frequent with FLOX than with FOLFOX (38% versus 10.8%). Moreover, of all the patients with severe GI toxicity who required hospitalisation, 64.6% belonged to the FLOX and 35.4% to the FOLFOLX4 regimen (p<0.01). From a practical point of view, FOLFOX therapy is more difficult to manage because it requires the placement of a central venous catheter, a procedure which may involve complications (infections, port malfunction, or clot formation in the subclavian vein), as well as the necessity of carrying a pump for the 5FU infusion. On the other hand, FLOX therapy requires more time in the clinic and more frequent visits to the hospital (18 visits rather than 12 visits for FOLFOX4 during the course of the therapy). Despite the advantages and inconveniences of each of them, FOLFOX was the most used therapy for resectable stage III colon cancer, but FLOX was also an effective alternative. A third international study conducted between 2003 and 2004, known as the XELOXA trial (NO16968) [42], randomised 1,886 patients with resected stage III colon cancer to receive either XELOX (capecitabine and oxaliplatin, 8 cycles) (table 1) or bolus 5FU and LV (Mayo Clinic or Roswell Park regimen) at the discretion of each centre. Five-year DFS was significantly superior with XELOX than with 5FU/LV (66.1% versus 59.8%, respectively; p=0.0045) but there were no statistically significant differences in the 5-year OS (77.6% versus 74.2%, respectively; p=0.14) (table 4). XELOX was associated with fewer cases of diarrhea, hematological toxicity, mucositis and alopecia, but more cases of neurotoxicity, vomiting, and hand-foot syndrome when compared to the Mayo Clinic regimen. In addition, it was associated with fewer GI but more hematological events when compared to the Roswell Park regimen [43]. To prevent neurotoxicity, different neuromodulators have been studied such as calcium and magnesium (CaMg), xaliproden, glutathione, gabapentin, carbamazepine and venlafaxine, giving controversial results. Although the NCCTG N04C7 trial demonstrated the reduction of neurotoxicity by intravenous CaMg, another study, as yet unpublished (presented at ASCO 2009 by Grothey et al), found a decrease in the efficacy of FOLFOX in the group that received CaMg [44]. Another study from the same group, confirmed that CaMg reduced accumulative but not acute neurotoxicity [45]. In a phase III trial, xaliproden was shown to be efficient in reducing oxaliplatin-induced acute or accumulative neurotoxicity without influencing FOLFOX4 antitumour activity in metastatic disease [46].
118
José María Vieitez de Prado et al.
B) Irinotecan At least four randomised trials of adjuvant irinotecan with either bolus or infusional 5FU and LV have been reported. In the USA between 1999 and 2001, the Cancer and Leukemia Group B (CALGB)-89803 trial [47] randomised 1,264 patients with resected stage III colon cancer to receive either the Roswell Park regimen for 8 months, or the Saltz regimen of irinotecan for 7.5 months (IFL) (table 1). Surprisingly, although IFL had proven superior to 5FU and LV in patients with metastatic disease, it did not improve either the DFS or OS when administered as an adjuvant therapy (table 4). A higher early all-cause mortality rate was found with IFL (2.8%; p=0.008) and was attributable to GI syndrome or thromboembolic events, but how irinotecan was implicated was not clear. The incidence of neutropenia was greater in the IFL than in the FULV arm (43% versus 5%, respectively; p< 0.00001), while other common adverse events such as diarrhea and vomiting were similar in both arms. In Europe, the phase III multicentre PETACC-3-EORTC trial [48] randomised 3,333 patients with resected stage II (29%) or III (71%) colon cancer to receive infusional 5FU and LV (De Gramont LV5FU2 or the German AIO regimen) with or without irinotecan (180 mg/m2 every 2 weeks in LV5FU2 (FOLFIRI) and 80 mg/m2 weekly in AIO, respectively) (table 1) for 6 months. The addition of irinotecan did not significantly improve the 5-year DFS (56.7% with irinotecan/LV5FU2 and 54.3% with LV5FU2 alone; p=0.106), nor the 5-year OS (73.6% versus 71.3%, respectively; p=0.094) (table 4), but it did increase the incidence of grade 3–4 GI events and neutropenia without changing the early mortality rate (1%), suggesting that irinotecan might be better tolerated with infusional 5FU and LV. A smaller European trial published in 2006, the ACCORD 02 trial [49], randomised 400 patients with resected high-risk stage III colon cancer (N2, or N1 with colonic obstruction or perforation) to receive adjuvant therapy with LV5FU2 alone or with irinotecan (FOLFIRI) (table 1). After a median followup of 36 months, the rate of 3-year DFS was actually poorer in patients who had received irinotecan (table 4). This finding was also demonstrated in another European trial, the Aventis V307 trial. In summary, the addition of irinotecan to 5FU and LV in each of these four clinical trials resulted in increased toxicity without a meaningful improvement in outcome. The overall negative results of PETACC-3 [48] might be explained by the increased incidence of death in the early follow-up of the experimental arm and the inclusion of more T4 patients, without the benefit from the addition of irinotecan that was reported in the ACCORD2 trial [49].
Table 4. Randomised trials with Oxaliplatin- or Irinotecan-based regimens.
Adjuvant systemic treatment for colon cancer 119
120
JosĂŠ MarĂa Vieitez de Prado et al.
When combined with infusional 5FU in patients with metastatic disease, unlike the adjuvant therapy setting, the benefit of irinotecan is similar to that observed with oxaliplatin. There are several cell biology-related theories that could explain this discrepancy, one of them being that irinotecan acts on a phase of cell cycle that is rarer in micrometastases, while oxaliplatin activity is independent of the cell cycle. At the same time, irinotecan target protein, topoisomerase I is found at lower levels in micrometastases, while oxaliplatin has less specific targeting. Moreover, irinotecan sensitivity may change during the various stages of tumour progression while oxaliplatin is more stable [50].
1.8. Targeted agents A) Bevacizumab Bevacizumab is a humanised monoclonal antibody directed against vascular endothelial growth factor (VEGF), the addition of which to a variety of cytotoxic agents results in enhanced therapeutic outcomes in patients with advanced CRC. The next step was to test the potential benefit and safety of bevacizumab in the adjuvant setting, and to do this, four large studies were developed as we will describe below. The NSABP C-08 phase III study [51] randomised 2,672 patients with resected stage II (25%) or III (75%) colon cancer to modified FOLFOX6 (mFOLFOX6) (table 1) or to mFOLFOX6 for 6 months plus bevacizumab 5 mg/kg every 2 weeks for 12 months. Addition of bevacizumab did not result in an overall statistically significant prolongation in the 3-year DFS (HR: 0.87; p=0.089); however, there was a transient benefit in DFS during the first 1-year interval that bevacizumab was administered (HR: 0.6; p=0.0004). Futures studies should, therefore, consider a longer duration of bevacizumab administration. At the same time, more clinical trials in the adjuvant setting that address the possibility of angiogenesis inhibition as well as its influence on metastatic behaviour and subsequent course of the disease could be necessary. The studies also showed a lower dose-intensity effect and more grade 3 toxicity in patients â&#x2030;Ľ70 years (81% versus 73% in younger people; p<0.001), leading to caution in the use of complex regimens in elderly people [52]. Between December, 2004 and June, 2007, the international AVANT trial [53] compared 6 months of XELOX or FOLFOX4, either alone or with bevacizumab, for 1 year in 3,451 patients with stage III (83%) or high-risk stage II (17%) colon cancer; however, the DFS and OS results are still not available. Toxicity in this trial was comparable to the safety profile in metastatic CRC and in the NSABP C-08 trial. The arms with bevacizumab
Adjuvant systemic treatment for colon cancer
121
reported more cases of grade 3â&#x20AC;&#x201C;4 hypertension, but similar venous or arterial thromboses, bleeding, wound healing complications, abscess/fistula and GI perforations. There were no differences in adverse events due to bevacizumab between the capecitabine and 5FU-based regimens. All-cause mortality within 60 days of treatment start was less than 1% in all arms. Other studies currently using bevacizumab in an adjuvant setting include the ECOG5202 trial, which has randomised patients to no adjuvant treatment if chromosome 18q loss of heterozygosity (18q-LOH) was absent and microsatellite instability (MSI) was present, or to FOLFOX6 with or without bevacizumab if 18q-LOH and microsatellite stability (MSS) were present. Also being conducted is the English Quick And Simple And Reliable (QUASAR)2 trial, which has randomized patients to capecitabine alone or with bevacizumab. The results of these studies, to be communicated in the near future, will likely help to determine the convenience of bevacizumab in the adjuvant setting. B) Cetuximab Cetuximab is a human-murine chimeric monoclonal antibody that targets the epidermal growth factor receptor (EGFR). Autophosphorylation of the intracellular tyrosine kinase domain of the EGFR activates downstream signalling pathways, including the Ras/Raf/mitogen-activated protein kinase pathway. KRAS mutation status is an independent prognosis factor in advance CRC and KRAS wild-type is predictive for cetuximab responsiveness in this setting. On the other hand, KRAS mutation predicts resistance to cetuximab, and thus its status can be included in the treatment decision algorithm in advanced CRC [54]. The benefit from cetuximab in the adjuvant setting was addressed in the NCCTG-N0147 trial [55], in which 1,760 patients with resected, KRAS wildtype stage III colon cancer were randomly assigned to mFOLFOX6 with or without cetuximab (250 mg/m2/weekly with 400 mg/m2 as first dose). Results did not show an increase in survival, instead showing an impaired DFS and a trend towards impaired OS. Another trial with FOLFOX4 alone or with cetuximab is PETTAC-8, the interim analysis of which is expected in 2011.
1.9. Summary As previously emphasised, only a handful out of 100 stage II patients will benefit from adjuvant chemotherapy. Most of them will receive undue treatment and will have to face toxicity. Similarly, in stage III, over approximately 50% are cured by surgery and might not need adjuvant
122
JosĂŠ MarĂa Vieitez de Prado et al.
chemotherapy. An additional 15.3% of patients would be cured at 3 years by adjuvant 5FU/LV and 6.9% more by FOLFOX; however, metastasis will develop in 27.8% of patients despite adjuvant treatment. Among those potentially cured by FOLFOX, a majority would have also been cured by oral capecitabine, and would not have experienced neurotoxicity or the inconvenience of the i.v. route. In addition, among stage III, it is sub-stage IIIC patients that benefit most from adjuvant FOLFOX (11% benefit on 4-year DFS as compared to 7.2% in N1 stage and 8.6% for the overall stage III population). There is an urgent need for prognostic and predictive markers to improve decisions on who should receive adjuvant chemotherapy and which type. Some have already been identified, including 18q-LOH and MSI, which are considered criteria of inclusion in NSABP C-O8, and others are being prospectively studied in ongoing clinical trials. More are known to be valuable for improving the efficacy of 5FU (e.g., TS, thymidine phosphorylase) [56], but they are not yet established as decision-making markers and are not routinely available.
2. Controversies As mentioned above, there are no clear recommendations for adjuvant treatment in certain groups (in stage II colon cancer, in elderly patients, or in some clinical practice contexts).
2.1. Adjuvant in stage II colon cancer The vast majority of patients with stage II colon cancer has a good prognosis after surgery alone and never develops recurrence. With a 5-year survival rate of approximately 80%, physicians question whether the small potential gain with adjuvant treatment justifies subjecting the entire population to the risks, toxicities and costs of the therapy. However, considering that adjuvant chemotherapy is indicated in patients with stage III colon cancer, and that stage IIB (T4No) has a 5-year survival rate inferior to stage IIIA (T3N1), we might speculate that a subset of patients with stage II could benefit, even if the biological similarity of a tumour confined to the bowel wall and one that presents as nodal metastasis is still a debated issue [57]. Moreover, it is difficult to gather 4,700 patients per group, the sample size that is required to detect a 4% survival difference between the treatment and control arm (90% power with a significance level of 0.05); therefore, physicians need to identify prognostic markers capable of reducing the number needed to treat (NNT) [58,59].
Adjuvant systemic treatment for colon cancer
123
Nowadays, high-risk stage II disease is defined as the presence of T4 tumour, tumour perforation, poorly differentiated histology, extramural venous or lymphatic invasion, an elevated pre-surgical CEA value, residual tumour after surgery or fewer than 10 sampled lymph nodes. There are other factors: size of tumour, left-sided tumour, mucinous histology, perineural invasion, high proliferation index, peritoneal disease, absence of MSI, 18qLOH, bax expression, loss of p27 expression, increased mitotic index, low bcl2 or p53 expression, as well as other more recently identified molecular markers (such as indicators of angiogenesis (vascular count, VEGF)) and markers of invasion/metastasis (plasminogen-related molecules, matrix metalloproteinases) which should be confirmed on a prospective basis [60]. Although the first high-risk factors mentioned above are unquestionably associated with a poor prognosis, whether adjuvant therapy could improve outcome in this group is still not clear [60]. To discover whether an adjuvant treatment is necessary, there are three groups of clinical trials to which we can refer: those with the control arm of surgery or of fluoropyrimidines, and meta-analyses. Meta-analysis is used to increase statistical power to reveal small benefits from studies with small sample sizes such as those of stage II CRC. A) Clinical trials There is a lack of well-designed trials specifically conducted to investigate the efficacy of adjuvant therapy in stage II colon cancer. Despite the fact that the INT-0035 trial (5FU/LEVA) [4,5] was one of the first studies to demonstrate the benefit of chemotherapy in resected colon cancer, it could not confirm its utility in patients with stage II disease (table 5). The INT-0089 trial [11] demonstrated the same 5-year DFS and OS rates for patients with high-risk stage II (obstructing and/or perforating nodenegative lesions) as those with stage III using 5FU plus high or low-dose LV and at the same time, showed no benefit and more toxicity with LEVA (table 5). A study from the Austrian Breast and CRC Study Group, published by Schippinger et al [61] investigated the use of adjuvant 5FU/LV only in patients with stage II colon cancer (500 patients). The results demonstrated a trend towards a lower risk of relapse in patients treated with chemotherapy but it failed to detect small improvements in the DFS or OS because of the limited number of patients enrolled (table 5). The Netherlands Adjuvant CRC Project (NACCP) conducted a prospective trial [62], in which 1,029 patients with stage II (45%) and III colon cancer were randomised to receive 1 year of 5-FU/LEVA or surgery
124
JosĂŠ MarĂa Vieitez de Prado et al.
alone; the results support the hypothesis that adjuvant therapy might be equally effective in stage II colon cancer (p=0.007). The translational studies of the PETACC3/EORTC40993/SAKK 60-00 trial [63] (5FU/LV or 5FU/LV/irinotecan adjuvant therapy) showed that there was a higher incidence of MSI in stage II (22%) versus stage III (12%) and that MSI was a significant prognostic factor for DFS (HR: 0.265; p=0.0044) and OS (HR: 0.159; p=0.011) in stage II colon cancer. Although a statistically significant benefit was not observed in those patients with stage II disease in the MOSAIC study, a 5.4% absolute improvement in DFS was noted in patients with high-risk stage II disease, defined as the presence of T4 tumour, bowel obstruction, tumour perforation, poorly differentiated histology, venous invasion, or fewer than 10 examined lymph nodes (table 5). The QUASAR study [64] enrolled 3,239 patients (91% with stage II disease) between 1994 and 2003 who were randomised to 5FU/LV with or without LEVA (63% received 5FU 370 mg/m2 with low-dose LV 25 mg 5 days per month for 6 months) or to surgery alone. The study concluded that patients with stage II CRC treated with chemotherapy had a small but statistically significant absolute improvement in survival of 3.6% (95% CI: 1.0-6.0; p=0.001). In comparison with other studies, such as those carried out by Moertel et al [4,5], QUASAR did not find an increased mortality rate due to non-tumoural causes in the arm treated with chemotherapy (table 5). Another study, NSABP C-06 [37] involving 1,608 patients (47% stage II), compared 5FU with an oral fluoropyrimidine, UFT, and found that UFT was comparable to 5FU in patients with stage II and stage III colon cancer. B) Meta-analyses The IMPACT B2 meta-analysis [65], which included 5 separate trials (1,016 patients with B2 colon cancer who were randomised to 5FU/LV or observation), did not detect a significant DFS or an OS difference among treated versus untreated patients (table 4). Although, none of the four individual NSABP (C-01, C-02, C-03 (MOF vs 5FU) and C-04) trials [66] were designed to evaluate a treatment benefit in the subpopulation of patients with stage II colon cancer, a meta-analysis of all of them concluded that adjuvant chemotherapy benefited this subset of patients. Moreover, the reduction in mortality was superior in stage II compared to that in stage III (30% versus 18%) independent of the presence of risk factors (T4, obstruction or perforation) (table 4). However, it is important to mention that this analysis could have been influenced by both the heterogeneity of the statistical method used and the heterogeneity in
Adjuvant systemic treatment for colon cancer
125
chemotherapy regimens (intraportal administration, the use of alkylating agents). The Mayo Clinic group conducted a meta-analysis [67] of 3,302 patients with stage II (44%) and III colon cancer from seven randomised trials comparing 5FU/LV or 5FU/ LEVA to surgery alone and tested the value of treatment, age, sex, tumour location, T stage, nodal status and grade as both prognostic and predictive factors. The results showed that the only factors which influenced the DFS and OS were nodal status, T stage, and grade. An increase in DFS (p=0.049) with adjuvant chemotherapy in stage II colon cancer but not in OS (table 5) was also observed. A meta-analysis performed by The American Society of Clinical Oncology in collaboration with the Cancer Care Ontario Practice Guideline Initiative (CCOPGI) [68] included 37 trials and 11 meta-analyses (20,317 patients) comparing adjuvant therapy to surgery alone. They found that adjuvant therapy was associated with a benefit in DFS, but not in OS, for patients with stage II colon cancer (table 5). However, there is concern that the inclusion of rectal cancer patients and immunotherapy trials could have caused confusion. Table 5. Chemotherapy in stage II CRC . Stage II patients (n) Study
Arms
OS (p)
5FU+Leva 5FU+LV 5FU+LV
0.10 0.49 0.001
Studies with observation as control arm INT 0035 Schippinger et al QUASAR
318 500 3239
Studies with chemotherapy in control and experimental arm
INT 0089
752
NSABP C-06 MOSAIC NSABP C-07
746 899 718
5FU+Leva 5FU+LV 5FU+LV+Leva FL vs. UFT+LV FOLFOX vs. FU5LV2 FLOX vs. FL
IMPACT NSABP
1016 1565
5FU+LV 5FU-based regimens
CCOPGI
20317
Mayo Clinic
1440
5FU-based regimens inmunotherapy 5FU+LV or 5FU+Leva
>0.05 >0.05 0.986 NA
Meta-analyses 0.057 0.01 (with risk factors) 0.26 (without risk factors) 0.07 0.112
126
JosĂŠ MarĂa Vieitez de Prado et al.
C) Microsatellite instability (MSI) as a predictive or prognostic factor One of the risk factors that plays an important part in locally advanced colon cancer is MSI. It reflects a defective DNA mismatch repair mechanism (MMR), which results in somatic alterations in the size of simple repeat nucleotide sequences (microsatellites); MLH1 and MSH2 are the most frequently altered genes [69]. Tumours may be characterised based on: high-frequency MSI (MSI-H) if two or more of the five markers show instability, low-frequency MSI (MSI-L) if only one of the five markers shows instability and MSS if none of the markers shows instability. The existence of MSI is related to certain clinical characteristics, such as: female gender, mucinous tumours, proximal localisation of the neoplasm, poorly differentiated histology and small number of lymph nodes invaded [70]. Moreover, Teipja et al. [71] found that its presence differed from stage to stage: 22% in stage II, 12 % in stage III and 3.5% in stage IV. After corroboration with the results of other studies [72], it was suggested that MSI-H could be a protective factor against metastases (including node metastases) since functionally active lymphocytes infiltrate the MSI-positive colon cancer, which explains why this alteration was seen more frequently in non-advanced tumours. Moreover, the MSI-H colorectal tumours seem to share a less aggressive clinical course than stage-matched MSI-L or MSS, which is also associated with a better prognosis independent of tumour stage [73,74]. However, the studies of Hemminki et al and Elsaleh et al found that patients with MSI-H have a poor response to 5FU therapy [75,76]. This finding was confirmed in the study by Ribic et al, which included 570 patients from five clinical trials with previous colon surgery, whose aim was to detect the benefit of adjuvant chemotherapy (three of these trials were using 5FU/LV and the other two 5FU/LEVA) [77]. When all 570 patients were considered, it was found that chemotherapy did not make a significant difference (p=0.11), but when the group of patients with MSS were analysed it was found to benefit from the chemotherapy (p=0.02). For the reciprocal group that presented with MSI-H, adjuvant treatment had negative effects, although these were not statistically significant (p=0.1). Furthermore, there was a trend towards better survival in MSH-L patients from the control arm (p=0.004). Kim et al [78] reported contradictory results. From 542 cases in the NSABP studies (C-01, C-02, C-03, C-04) they found a favourable DFS for the 103 patients with MSI, but this was not reflected in the OS or in multivariate analyses with both parameters. They did, however, find concordance with the relationship between MSI-H and patient characteristics, such as gender, age, stage and tumour localisation. The difference in results among these trials
Adjuvant systemic treatment for colon cancer
127
could be explained by differences in the techniques used, or by the fact that only two of the studies had surgery as the control arm. Labianca et al at the European Society of Medical Oncology (ESMO) congress 2009 [79], extending the work of Ribic et al [77], presented data from 1,027 patients with MSI- positive stage II colon cancer and showed that adjuvant treatment was associated with a decrease in survival (p=0.04). A meta-analysis of 32 studies (14 with patients with localised and metastatic disease, 15 with stages Iâ&#x20AC;&#x201C;III and three with stage IV) [80] carried out in 16 countries and with a total of 7,742 patients looked for a correlation between survival and MSI which was detected in 16% of the patients. In 13 of the 32 trials statistically significant differences were found and the authors concluded that survival was increased in the group with MSI. Moreover, even though the results concerning the effect of the chemotherapy were limited, they did report a lack of benefit from adjuvant therapy for the MSI group (HR 1.24; 95% CI: 0.72â&#x20AC;&#x201C;2.14). We should add that the meta-analyses included patients whose MSI status was defined only by genetic study and not by immunohistochemical analysis (IHC). It is possible that the relationship between TS overexpression and MSI could explain both the lack of benefit of 5FU-based adjuvant treatment and the efficacy of other agents that use molecular mechanisms not involving TS [81]. Oxaliplatin acts by binding to DNA in a way that cannot be detected by the MMR system, and thus its activity might not be modified in MSI-positive tumours. Zaanan et al [82] presented a retrospective study of 233 stage III patients, 39 of who were part of the MOSAIC study. It compared the patients treated before October, 2003 (5FU/LV regimen) and those treated after this date (FOLFOX regimen). They found, not only in patients with p53 mutations but also in those with MSI, that the DFS in the FOLFOX group was superior to that in the 5FU group. We should analyse these data carefully, however, because not only was the study retrospective, but the number of patients included in it was also very small: 123 patients with a p53mutation and 32 with MSI. There are three retrospective studies of patients treated with FOLFOX in which MSI does not correlate with treatment efficacy but they were nonrandomised, two of them included patients in stage IV and the third enrolled a heterogeneous group (9.6%, 80%, and 10.4% in stages II, III, and IV, respectively) [83,84,85]. Surprisingly, a similar result was found in a retrospective analysis of Cancer and Leukemia Group B 89803 trial [86], which randomised patients with stage III CRC to 5FU/LV or IFL. Although the results were negative, a slight increase in DFS in MSI patients treated with IFL (p=0.07) was observed.
128
José María Vieitez de Prado et al.
When completed, the ECOG 5202 study may clarify the indication of chemotherapy for stage II. The patients at high risk (MSS, 18q-LOH) were randomised to FOLFOX with or without bevacizumab, while those with good prognosis were treated exclusively with surgery. Starting a study of patients with MSI randomised to FOLFOX or to a regimen without fluoropyrimidines being used for stage IV, such as an irinotecan and oxaliplatin combination (IROX), has nevertheless been suggested [87]. Finally, we should find out how this knowledge is going to help us in everyday medical practice [88]. Knowing that FOLFOX in patients younger than 65 years and with stage III colon cancer may be a standard treatment, the determination of MSI seems unlikely to bring about any changes in the treatment of those patients. In fact, caution should be exercised; validation in prospective trials is essential before the introduction of routine MSI testing as an aid to prognosis and choice of treatment regimen. By contrast, patients with high-risk stage II CRC tend to be offered adjuvant treatment, but in this case if MSI was determined it could be beneficial. Its presence could indicate a better prognosis and could inform the avoidance not only of adjuvant treatment, but also of certain other agents. Also, in patients older than 65 years with a poor general state of health and in those who are planned to be treated exclusively with fluoropyrimidines, the finding of MSI could lead to the implementation of FOLFOX treatment. On the other hand, for patients older than 70 years, FOLFOX seems to have no more efficacy than monotherapy with fluoropyrimidines and so MSI detection could lead to abstinence from adjuvant treatment [89]. However, it is important to take into account that MSI can be produced by two different pathways: 1) germline mutations, and 2) epigenetic silencing leading to different behaviour of tumours. The first refers to MMR gene mutation (especially MLH1, MSH2, and MSH6), which is associated with the most common genetic syndrome with a predisposition for colon cancer, hereditary non-polyposis CRC (Lynch syndrome). The second results from hypermethylation of CpG islands in MLH1 in sporadic CRC. Although MSI is associated with a better prognosis in sporadic tumours than in hereditary ones, it frequently coexists with the BRAF mutation, which is a poor prognostic factor [90]. Moreover, it is not clear whether these MSI groups (sporadic or hereditary) have a different response to chemotherapy [91]. In summary, it is not clear if over-treatment of 95% of patients with stage II is justified for the 2–4% added benefit from chemotherapy after surgery that has been the marginal, but consistent, finding in the studies and meta-analyses completed to date. The decision to offer adjuvant therapy for stage II disease therefore needs to be individualised to the circumstances of each patient, explaining the
Adjuvant systemic treatment for colon cancer
129
benefits and the possible risks the choice involves. In clinical practice, it is likely that stage II patients without T4 tumours, neither occluded nor perforated, with more than 12 examined lymph nodes and treated in hospitals that can guarantee a multidisciplinary approach, have a very high surgical cure rate. In spite of this, many studies have shown that patients are willing to accept adjuvant therapy for little or even no clinical improvement; therefore, the explanation about adjuvant treatment advantages and disadvantages should be provided by a medical oncologist with a lot of experience in this field. Furthermore, with approximately 40â&#x20AC;&#x201C;60% of patients with stage II currently receiving adjuvant chemotherapy, it would be more appropriate to encourage them to participate in randomised trials [92]. In the meantime, treatment should be individualised and oral fluoropyrimidines should be considered a valuable alternative in this good prognosis population.
2.2. Clinical practice attitude Several reviews found significant variations in the use of chemotherapy in stage III colon cancer (48â&#x20AC;&#x201C;62%) among different centres, which could put patients at risk if they failed to receive appropriate adjuvant therapy. The variation depended on: A) Patient characteristics It was noticed that older patients (>75 years) received adjuvant chemotherapy significantly less often than younger patients, older age being considered a predictor of treatment failure. Increased comorbidity (congestive heart failure, chronic obstructive pulmonary disease and diabetes) as well as duration of hospitalisation after surgical resection or rehospitalisation were found to be negative predictors for the use of adjuvant treatment. Subjective judgments about sex and race, probably in relation to social and environmental factors, have also influenced the use of chemotherapy, and explain why women and non-white patients have received less adjuvant treatment [93]. B) Tumour characteristics Patients with more positive lymph nodes had a greater overall use of chemotherapy; on the other hand, tumour size was not considered a significant predictor.
130
José María Vieitez de Prado et al.
C) Oncologist versus surgeon approach There are three factors that could influence good results of chemotherapy in stage III colon cancer: a) Referral to an oncologist: better results were seen when the treatment was conducted by an oncologist than by a surgeon as the former is more capable of completing administration of all the therapy cycles. However, there were no differences found in the number of patients who started treatment with either an oncologist or a surgeon. b) Acceptance of the treatment by the patients: patient refusal accounted for 29.6% of non-use and this was due to failure to communicate information or to the recommendation of no treatment being given by the physician rather than to the non-acceptance of the patient. c) The ability to maintain a regimen of chemotherapy: 22–31% of patients were unable to complete treatment, which led to a higher mortality rate (20–47%) than in the group of patients that finished therapy [94]. Although the data from these reviews are provided to improve day to day clinical practice, we should not forget that every patient is a different entity and the opportunity of prolonging life should surpass any social or economic barrier.
2.3. Should 3-year disease free survival be a surrogate of overall survival? The traditional end point for adjuvant chemotherapy clinical trials is OS. However, Sargent et al demonstrated in meta-analyses published in 2005 [95] and then in 2007 [96], that 3-year DFS is an appropriate surrogate end-point and an excellent predictor of 5-year OS in clinical trials of adjuvant therapy in colon cancer. This requires a smaller sample size with a shorter follow-up period, which reduces the cost and time of reporting such trials and accelerates the improved therapeutic strategies. For this reason, in 2004, the FDA accepted that DFS was an adequate basis for regular drug approval. The ACCENT meta-analysis [89] included 20,898 patients enrolled onto 18 phase III trials and it concluded that DFS outcomes after a 2- or 3-year median follow-up were excellent predictors of 5-year OS for trials in which the majority of patients were stage III. The correlation of HR within trials was 0.92 (95% CI: 0.85–0.95) for stage III patients and 0.70 (95% CI: 0.44–0.80) for stage II patients. Moreover, DFS with 1-year minimum follow-up demonstrated a perfect negative predictive value because all trials negative at
Adjuvant systemic treatment for colon cancer
131
1-year for DFS were negative for 5-year OS. This will help when discontinuation of a trial is necessary and when drugs have a negative oneyear interim evaluation. Nowadays, the existence of more accurate imaging techniques for the diagnosis of metastases and the availability of new, more effective options to treat them means that instead of attributing OS effects solely to adjuvant chemotherapy, other parameters can now be identified. This may result in DFS becoming a more appropriate outcome than OS in CRC. Moreover, while the DFS in the first 2 years was transiently favourable in the bevacizumab arm, the 3-year DFS predicted negative results in OS in the NSABP C-08 study [51]. Nevertheless, one aspect we must take into account, as we have shown in this chapter, is that all adjuvant clinical trials do not measure an identical end point, OS or DFS, and this can make it difficult to compare them.
3. Recommendations In this section, we will design an algorithm for an alternative adjuvant approach in the context of different patients.
3.1. Stage III resected disease (node-positive) (Grade 1A recommendation) A 6-month course of an oxaliplatin-based regimen is recommended: XELOX or mFOLFOX6. While FOLFOX4 was the regimen used in the adjuvant registration trials, mFOLFOX6 could be chosen because of its convenient administration and FOLFOX7 or modified FOLFOX7 (mFOLFOX7) (table 1), which eliminates the use of bolus 5FU completely, is a less aggressive alternative, with significantly less hematological toxicity. In particular situations, these regimens could be modified as detailed below: •
•
• • •
•
When patients cannot tolerate oral drugs or have a dihydropyrimidine dehydrogenase (DPD) deficiency: mFOLFOX6. When patients have a high risk of pneumothorax, septicemia or thromboses: the use of an implantable port should be avoided, and it is therefore better to use XELOX or FLOX. When patients have low risk of neuropathy: FLOX. When patients have high risk of neuropathy: fluoropirimidine alone. When patients have severe cardiovascular disease: raltitrexed plus oxaliplatin. When patients have renal failure: UFT/LV.
132
JosĂŠ MarĂa Vieitez de Prado et al.
3.2. Stage II resected disease Based on the available data [61-68], adjuvant chemotherapy should not be considered as a standard of care for all patients. It is also necessary to discuss the risks and benefits of adjuvant chemotherapy with patients who have highrisk disease as defined by any of the following: -
Inadequate node retrieval: fewer than 13 nodes in the surgical specimen T4 lesion Poorly differentiated histology, signet ring and mucinous histology Aneuploidy, high S-phase fraction or deletion of 18q Perforated lesion Vascular, lymphatic or perineural invasion
We suggest capecitabine or 5FU/LV alone, rather than an oxaliplatin-based regimen for most patients in stage II, but it is important to take into consideration that stage IIB (T4N0) has a lower OS than stage IIIA (T1-2N1) because of the increase of recurrences (Grade 2B recommendation). An online tool, Adjuvant! Online, can help the clinician estimate the risk of death within 5-years based on clinicopathologic features (age, comorbidities, T stage, N stage, number of nodes retrieved and histological grade) and the relative benefits of chemotherapy. Available at: http://www.adjuvantonline.com/index.jsp.
3.3. Radiation therapy The evidence is inconclusive regarding the benefit of adding radiotherapy to chemotherapy for patients at high risk of a local recurrence in colon cancer. In accordance with NCCN guidelines, adjuvant radiotherapy should be given to patients with T4 tumours that infiltrate a fixed structure or have positive resection margins (Grade 2C) [97].
4. Follow-up There are two aims of surveillance after curative resection of CRC : -
To identify recurrence that could be completely resectable. To identify second CRC s.
Although there is no standard schedule for follow-up, different groups have published guidelines concerning this issue: ASCO, ESMI, Cochrane, American Cancer Society, the US Multi-Society Task Force on CRC and others.
Adjuvant systemic treatment for colon cancer
133
Based on these guidelines and on the available literature, we recommend the following [98]: -
-
-
-
-
-
-
Evaluation of the general condition, concomitant pathology and organ function that determine therapeutic strategy. Thus, in cases where salvage surgery or systemic treatment would not be feasible, the need for a strictly controlled strategy will not be so important. Clinical suspicion of metastatic disease should always be confirmed by adequate radio-imaging (usually a computed tomography scan). Histopathologic or cytologic confirmation should be obtained whenever there is atypical presentation or very late presentation after the primary cancer. Resectable metastases do not need histologic confirmation before resection. An FDG-TEP can identify further lesions when performing a planned resection of metastases. History evaluation, physical examination and CEA should be determined every 4 months for the first 3 years, every 6 months during 4 and 5 years, and subsequently at the discretion of the physician. Ultrasonography of the liver and chest X-ray should be carried out every 6 months for 5 years. CT scan of the chest and abdomen instead of ultrasonography and X-ray could be considered in patients with stage III-C cancer or other tumoural conditions with a higher risk for recurrence. For rectal cancer, a pelvic CT scan should be also considered, especially for patients with several poor prognostic factors, including those who have not been treated with radiation. Other imaging tests should also be conducted if there are some symptoms or specific signs. Another colonoscopy should be performed after a year since diagnosis and thereafter every 3â&#x20AC;&#x201C;5 years to look for metachronous adenomas or cancer. In high-risk rectal cancer, a flexible proctosigmoidoscopy should be performed every 6 months for 2â&#x20AC;&#x201C;5 years. Blood count, routine blood chemistry (liver function tests) and other laboratory examination are not recommended unless patients have suspicious symptoms. Molecular or cellular markers should not influence the surveillance strategy based on available evidence.
Conclusion During the past 10 years, substantial progress has been made in the treatment of CRC. In patients with potentially resectable tumors, advances in
134
José María Vieitez de Prado et al.
surgery, radiation, and chemotherapy have all contributed to increased rates of cure. Higher volume medical centers have become models for improving surgical quality. Oxaliplatin has been incorporated into adjuvant treatment programs for colon cancer, and new targeted agents are currently in preclinical development and ongoing clinical trials. As in the past, further progress depends on the completion of well-designed RCTs.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
Wilkinson, NW, Yothers, G, Lopa, S, Costantino, JP, Petrelli, NJ, and Wolmark, N. 2010, Ann Surg Oncol, 17(4), 959. Wolmark, N, Fisher, B, Rockette, H, Redmond, C, Wickerham, DL, Fisher, ER, et al. 1988, J Natl Cancer Inst., 80(1), 30. Laurie, JA, Moertel, CG, Fleming, TR, Wieand, HS, Leigh, JE, Rubin, J, et al. 1989, J Clin Oncol,7(10),1447. Moertel, CG, Fleming, TR, Macdonald, JS, Haller, DG, Laurie, JA, Goodman, PJ, et al. 1990, N Engl J Med, 322(6), 352. 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. Anonymous. 1990, JAMA, 264, 1444. Anonymous. 1995, Lancet, 345(8955), 939. Wolmark, N, Rockette, H, Mamounas, E, Jones, J, Wieand, S, Wickerham, DL, et al. 1999, J Clin Oncol, 17(11), 3553. Porschen, R, Bermann, A, Löffler, T, Haack, G, Rettig, K, Anger, Y, et al. 2001, J Clin Oncol., 19(6), 1787. O'Connell, MJ, Laurie, JA, Kahn, M, Fitzgibbons, RJ, Erlichman, C, et al. 1998, J Clin Oncol., 16(1), 295. Haller, DG, Catalano, PJ, Macdonald, JS, O'Rourke, MA, Frontiera, MS, Jackson, DV, Mayer RJ. 2005, J Clin Oncol., 23, 8671. Bleeker, WA, Mulder, NH, Hermans, J, Otter, R, and Plukker, JT. 2000, Ann Oncol., 11(5), 547. Arkenau, HT, Rettig, K, and Porschen, R. 2005, Int J Colorectal Dis, 20(3), 258. André, T, Colin, P, Louvet, C, Gamelin, E, Bouche, O, Achille, E, et al. 2003, J Clin Oncol., 21, 2896. André, T, Quinaux, E, Louvet, C, Colin, P, Gamelin, E, Bouche, O, et al. 2007, J Clin Oncol., 25(24), 3732. Chau, I, Norman, AR, Cunningham, D, Tait, D, Ross, PJ, Iveson, T, et al. 2005, Ann Oncol., 16(4), 549. Poplin, EA, Benedetti, JK, Estes, NC, Haller, DG, Mayer, RJ, Goldberg, RM, et al. 2005, J Clin Oncol., 23(9), 1819. Hermanek, P and Sobin, LH. 1992, TNM Classification of Malignant Tumours. Fourth Edition 2nd Revision. Springer-Verlag (Ed.), Berlin, 52-55.
Adjuvant systemic treatment for colon cancer
135
19. Carrato, A, Köhne, C, Bedenne, L, Popov, I, Bouche, O, Gaspar, E, et al. 2006, J Clin Oncol., 24(18S), abst 3563. 20. Taylor, I, Brooman, P, and Rowling, JT. 1977, Br Med J, 2(6098), 1320. 21. Wolmark, N, Rockette, H, Wickerham, DL, Fisher, B, Redmond, C, Fisher, ER, et al. 1990, J Clin Oncol., 8(9), 1466. 22. Anonymous. 1995, Lancet, 345(8946), 349. 23. Anonymous. 1997, J Natl Cancer Inst., 89(7), 497. 24. Nitti, D, Wils, J, Sahmoud, T, Curran, D, Couvreur, ML, Lise, M, et al. 1997, Eur J Cancer, 33(8), 1209. 25. Rougier, P, Sahmoud, T, Nitti, D, Curran, D, Doci, R, De Waele, B, et al. 1998, Lancet, 351(9117), 1677. 26. Labianca, R, Fossati, R, Zaniboni, A, Torri, V, Marsoni, S, Nitti, D, et al. 2005, J Natl Cancer Inst., 96(10), 750. 27. Taylor, I. 2005, Eur J Surg Oncol., 31(1), 1. 28. Vervaal, VJ, Bruin, S, Boot, H, Gooike, VS, and Harm VT. 2008, Ann Surg Oncol., 15(9), 2426. 29. Ridge, JA, Bading, JR, Gelbard, AS, Benua, RS, Daly, JM, et al. 2006, Cancer, 59(9), 1547. 30. Riethmüller, G, Holz, E, Schlimok, G, Schmiegel, W, Raab, R, Höffken, K, et al. 1998, J Clin Oncol., 16(5), 1788. 31. Punt, CJ, Nagy, A, Douillard, JY, Figer, A, Skovsgaard, T, Monson, J, et al. 2002, Lancet, 360(9334), 671. 32. Fields, AL, Keller, A, Schwartzberg, L, Bernard, S, Kardinal, C, Cohen, A, et al. 2009, J Clin Oncol., 27(12), 1941. 33. Harris, JE, Ryan, L, Hoover, HC Jr, Stuart, RK, Oken, MM, Benson, AB 3rd, et al. 2000, J Clin Oncol.,18(1), 148. 34. Durrant, LG, Maxwell-Armstrong, C, Buckley, D, Amin, S, Robins, RA, Carmichael, J, and Scholefield, JH. 2000, Clin Cancer Res., 6(2), 422. 35. Popov, I, Carrato, A, Sobrero, A, Vincent, M, Kerr, D, Labianca, R, et al. 2008, Eur J Cancer, 44(15), 2204. 36. Wilson, KS, Fitzgerald, CA, Barnett, JB, Gill, S, and Khoo, KE. 2007, Cancer Invest., 25(8), 711. 37. Lembersky, BC, Wieand, HS, Petrelli, NJ, O'Connell, MJ, Colangelo, LH, Smith, RE, et al. 2006, J Clin Oncol., 24(13), 2059. 38. Twelves, C, Wong, A, Nowacki, MP, Abt, M, Burris, H 3rd, Carrato, A, et al. 2005, N Engl J Med., 352(26), 2696. 39. André, T, Boni, C, Navarro, M, Tabernero, J, Hickish, T, Topham, C, et al. 2009, J Clin Oncol., 27(19), 3109. 40. Kuebler, JP, Wieand, HS, O'Connell, MJ, Smith, RE, Colangelo, LH, Yothers, G, et al. 2007, J Clin Oncol., 25(16), 2198. 41. Land, SR, Kopec, JA, Cecchini, RS, Ganz, PA, Wieand, HS, Colangelo, LH, et al. 2007, J Clin Oncol., 25(16), 2205. 42. Haller, DG, Cassidy, J, Tabernero, J, Maroun, JA, De Braud, FG, Price, TJ, et al. ASCO GI 2010, abstr 284.
136
José María Vieitez de Prado et al.
43. Schmoll, HJ, Cartwright, T, Tabernero, J, Nowacki, MP, Figer, A, Maroun, J, et al. 2007, J Clin Oncol., 25(1), 102. 44. Nikcevich, DA, Grothey, A, Sloan, JA, Kugler, JW, Silberstein, PT, Dentchev, T, et al. 2008, J Clin Oncol., 26, abstr 4009. 45. Grothey, A, Nikcevich, DA, Sloan, JA, Kugler, JW, Silberstein, PT, Dentchev, T, et al. 2009, J Clin Oncol., 27(15s), abstr 4025. 46. Cassidy, J, Bjarnason, GA, Hickish, T, Topham, C, Provencio, M, Bodoky, G, et al. 2006, J Clin Oncol., 24(18s), abstr 3507. 47. Saltz, LB, Niedzwiecki, D, Hollis, D, Goldberg, RM, Hantel, A, Thomas, JP, et al. 2007, J Clin Oncol., 25(23), 3456. 48. Van Cutsem, E, Labianca, R, Bodoky, G, Barone, C, Aranda, E, Nordlinger, B, et al. 2009, J Clin Oncol., 27(19), 3117. 49. Ychou, M, Raoul, JL, Douillard, JY, Gourgou-Bourgade, S, Bugat., R, Mineur., L, et al. 2009, Ann Oncol., 20(4), 674. 50. Nallapareddy, S, Eckhardt, G, and Messersmith, W. 2008, Current Colorectal Cancer Reports, 4, 167. 51. Wolmark, N, Yothers, G, O'Connell, MJ, Sharif, S, Atkins, JN, Seay, TE, et al. 2009, J Clin Oncol., 27(18s), abstr LBA4. 52. Allegra, CJ, Yothers, G, O'Connell, MJ, Sharif, S, Petrelli, NJ, Colangelo, LH, et al. 2011, j. Clin Oncol., 29(1), 11. 53. Rousseau, B, Chibaudel, B, Bachet, JB, Larsen, AK, Tournigand, C, Louvet, C, et al. 2010, Cancer J., 16(3), 202. 54. Lièvre, A, Bachet, JB, Boige, V, Cayre. A, Le Corre. D, Buc. E, et al. 2008, J Clin Oncol., 26(3), 374. 55. Goldberg, RM, Sargent, DJ, Thibodeau, SN, Mahoney, MR, Shields, AF, Chan, E, et al. 2010, J Clin Oncol., 28(15s), abstr 3508. 56. Jensen, SA, Vainer, B, Witton, CJ, Jørgensen, JT, and Sørensen, JB. 2008, Acta Oncol., 47(6), 1054. 57. Gangadhar, T, and Schilsky, RL. 2010, Nat Rev Clin Oncol., 7(6), 318. 58. Benson, AB 3rd, Schrag, D, Somerfield, MR, Cohen, AM, Figueredo, AT, Flynn, PJ, et al. 2004, J Clin Oncol., 22(16), 3408. 59. André, T, Colin, P, Louvet, C, Gamelin, E, Bouche, O, Achille, E, et al. 2001, Semin Oncol., 28(Suppl 1), 35. 60. Hammond, ME, Fitzgibbons, PL, Compton, CC, Grignon, DJ, Page, DL, Fielding, LP, et al. 2000, Arch Pathol Lab Med., 124(7), 958. 61. Schippinger, W, Samonigg, H, Schaberl-Moser, R, Greil, R, Thödtmann, R, Tschmelitsch, J, et al. 2007, Br J Cancer, 97(8), 1021. 62. Taal, BG, Van Tinteren, H, and Zoetmulder, FA. 2001, Br J Cancer, 85(10), 1437. 63. Javle, M, and Hsueh, CT. 2010, J Hematol Oncol., 3:11. 64. Gray, R, Barnwell, J, McConkey, C, Hills, RK, Williams, NS, and Kerr, DJ. 2007, Lancet, 370(9604), 2020. 65. Anonymous. 1999, J Clin Oncol., 17(5), 1356. 66. Mamounas, E, Wieand, S, Wolmark, N, Bear, HD, Atkins, JN, Song, K, et al. 1999, J Clin Oncol., 17(5), 1349.
Adjuvant systemic treatment for colon cancer
137
67. Gill, S, Loprinzi, CL, Sargent, DJ, Thomé, SD, Alberts, SR, Haller, DG, et al. 2004, J Clin Oncol., 22(10), 1797. 68. Figueredo, A, Charette, ML, Maroun, J, Brouwers, MC, and Zuraw, L. 2004, J Clin Oncol., 22(16), 3395. 69. Peltomäki, P. 2003, J Clin Oncol., 21(6), 1174. 70. Cottrell, S, Bicknell, D, Kaklamanis, L, and Bodmer, WF. 1992, Lancet, 340(8820), 626. 71. Tejpar, S, Bertagnolli, M, Bosman, F, Lenz, HJ, Garraway, L, Waldman, F, et al. 2010, Oncologist, 15(4), 390. 72. Lynch, HT, Smyrk, TC, Watson, P, Lanspa, SJ, Lynch, JF, Lynch, PM, et al. 1993, Gastroenterology, 104(5), 1535. 73. Koopman, M, Kortman, GA, Mekenkamp, L, Ligtenberg, MJ, Hoogerbrugge, N, Antonini, NF, et al. Br J Cancer, 100(2), 266. 74. Gryfe, R, Kim, H, Hsieh, ET, Aronson, MD, Holowaty, EJ, Bull, SB, et al. 2000, N Engl J Med., 342(2), 69. 75. Hemminki, A, Mecklin, JP, Järvinen, H, Aaltonen, LA, and Joensuu, H. 2000, Gastroenterology, 119(4), 921. 76. Elsaleh, H, Powell, B, Soontrapornchai, P, Joseph, D, Goria, F, Spry, N, et al. 2000, Oncology, 58(1), 52. 77. Ribic, CM, Sargent, DJ, Moore, MJ, Thibodeau, SN, French, AJ, Goldberg, RM, et al. 2003, N Engl J Med., 349(3), 247. 78. Kim, GP, Colangelo, LH, Wieand, HS, Paik, S, Kirsch, IR, Wolmark, N, et al. 2007, J Clin Oncol., 25(7), 767. 79. Labianca, R, and Sargent, D. 2009, Eur J Cancer, (Suppl), 7(2), 46. 80. Popat, S, Hubner, R, and Houlston, RS. 2005, J Clin Oncol., 23(3), 609. 81. Capdevila, J, Elez, E, Peralta, S, Macarulla, T, Ramos, FJ, and Tabernero, J. 2008, Expert Rev Anticancer Ther., 8(8), 1223. 82. Zaanan, A, Cuilliere-Dartigues, P, Guilloux, A, Parc, Y, Louvet, C, de Gramont, A, et al. 2010, Ann Oncol., 21(4), 772. 83. Kim, ST, Lee, J, Park, SH, Park, JO, Lim, HY, Kang, WK, et al. 2009, Med Oncol., [Epub ahead of print]. 84. Des Guetz, G, Mariani, P, Cucherousset, J, Benamoun, M, Lagorce, C, Sastre, X, et al. 2007, Anticancer Res., 27(4C), 2715. 85. Kim, ST, Lee, J, Park, SH, Park, JO, Lim, HY, Kang, WK, et al. 2010, Cancer Chemother Pharmacol., 66(4), 659. 86. Bertagnolli, MM, Niedzwiecki, D, Compton, CC, Hahn, HP, Hall, M, Damas, B et al. 2009, J Clin Oncol., 27(11), 1814. 87. Ng, K, and Schrag, D. 2010, J Clin Oncol., 28(20), 3207. 88. André, T, Boni, C, Navarro, M, Tabernero, J, Hickish, T, Topham, C et al. 2009, J Clin Oncol., 27(19), 3109. 89. McCleary, NAJ, Meyerhardt, J, Green, E, Yothers, G, de Gramont, A, Van Cutsem, E, et al. 2009, J Clin Oncol., (suppl), 27(15s), abstr 4010. 90. Deng, G, Bell, I, Crawley, S, Gum, J, Terdiman, JP, Allen, BA, et al. 2004, Clin Cancer Res., 10(1 Pt 1), 191.
138
José María Vieitez de Prado et al.
91. Roth, AD, Tejpar, S, Delorenzi, M, Yan, P, Fiocca, R, Klingbiel, D, et al. 2010, J Clin Oncol., 28(3), 466. 92. Di Costanzo, F and Doni, L. 2005, Ann Oncol., 16(Suppl 4), 69. 93. McGory, ML, Zingmond, DS, Sekeris, E, Bastani, R, and Ko, CY. 2006, Dis Colon Rectum., 2006, 49(3), 319. 94. Rogers, SO Jr, Ayanian, JZ, Ko, CY, Kahn, KL, Zaslavsky, AM, Sandler, RS, and Keating NL. 2009, Ann Surg., 250(6), 895. 95. Sargent, DJ, Wieand, HS, Haller, DG, Gray, R, Benedetti, JK, Buyse, M, et al. 2005, J Clin Oncol., 23(34), 8664. 96. Sargent, DJ, Patiyil, S, Yothers, G, Haller, DG, Gray, R, Benedetti, J, et al. 2007, J Clin Oncol., 25(29), 4569. 97. Hung, AY, Canning, CA, Patel, KM, Holland, JM, and Kachnic, LA. 2006, Hematol Oncol Clin North Am., 20(2), 287. 98. Gravalos, C, Maurel, J, Rivera, F, Salazar, R, Sevilla, I, Sastre, J, et al. 2010, Clin Transl Oncol., 12, 724.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 139-155 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
6. Metastatic colorectal cancer: Focus on systemic treatment Paula García Teijido1, Teresa Sampedro Gimeno2 and Pilar Blay Albors3 1
Medical Oncology Service of Hospital “San Agustín”, Spain 2 Medical Oncology Service of Hospital Cabueñes 3 Medical Oncology Service of Central University Hospital of Asturias
Abstract. The treatment of advanced colorectal cancer (CRC) has significantly evolved in the last decade with the approval of new therapeutic agents. The median survival time of patients with stage IV disease without treatment is only 5-6 months. At present, with the use of systemic therapy, the median survival time has increased, and many patients live more than 20 months. Five different classes of drugs now have significant antitumour activity: fluoropyrimidines (5-FU, capecitabine and UFT), oxaliplatin, irinotecan, an anti-vascular endothelial growth factor (VEGF) monoclonal antibody (bevacizumab), and two antiepidermal growth factor receptor (EGFR) monoclonal antibodies (cetuximab and panitumumab). These therapeutic agents are mostly used in combination. The fluoropyrimidines, either 5-FU or capecitabine, are the backbone of most combination treatments. Molecular targeted therapies have emerged as a new generation of treatments that increase anticancer activity with minimal side effects. Correspondence/Reprint request: Dr. Paula García Teijido, Medical Oncology Service , “San Agustín” Hospital Spain. E-mail: paulagteijido@hotmail.com
140
Paula GarcĂa Teijido et al.
Despite efforts in clinical research, the best way to combine and sequence all these drugs to optimise treatment is not yet established.
Introduction Metastatic disease is present at diagnosis or ultimately develops in as many as 50-60% of patients with CRC, often in the liver. The majority of patients with metastatic colon or rectal cancer (mCRC) cannot be cured, although a subset with hepatic and/or lung metastases are potentially curable with surgery. For the rest, treatment is palliative and generally consists of systemic therapy. In the last decade, the integration of new drugs such as oxaliplatin and irinotecan has allowed us to improve the overall survival (OS) of mCRC patients. This chapter will review the data from clinical trials evaluating systemic chemotherapy with and without targeted therapies for nonresectable mCRC, focusing on the integration of anti-vascular endothelial and anti-epidermal growth-factor therapies.
1. Chemotherapy versus best supportive care The median OS time for patients with unresectable mCRC who receive the best supportive care is approximately six months. Systemic chemotherapy improves not only progression-free survival (PFS) but also median OS time [1]. Although no randomised trial has compared modern regimens containing oxaliplatin or irinotecan to best supportive care alone, median OS time is now nearly two years. Chemotherapy is effective in prolonging time to disease progression and survival in patients with advanced CRC.
2. Chemotherapy for metastatic disease 2.1. Fluoropyrimidines Different presentations of 5-Fluorouracil (5-FU) have been used for many years in mCRC. The main mechanism of toxicity is based on the impairment of DNA synthesis by the inhibition of thymidylate synthase. Administration of a 5-FU bolus administration may contribute to the inhibition of RNA synthesis. Monotherapy with 5-FU can be administered by bolus or by continuous intravenous infusion. Interpatient variation in the activity of the critical metabolising enzyme dihydropyrimidine dehydrogenase may account for differences in toxicity. Response rates of tumours are significantly higher for continuous infusion of 5-FU than for the 5-FU bolus, with a slight increase of OS [2]. Leucovorin enhances the cytotoxicity of 5-FU
Systemic treatment for metastatic colorectal cancer
141
associated with a twofold higher response rate and a small but statistically significant improvement in one-year survival [3]. Chronomodulation is a method where drug administration varies over a 24-hour period. These schedules of administration generally improve response rates and lessen toxicity, but the true value of this approach has not been fully established. Infusional 5-FU is associated with less diarrhoea, mucositis, nausea, vomiting, and neutropenia than 5-FU administered by bolus. Capecitabine is a fluoropyrimidine carbamate absorbed and converted to 5-FU by sequential enzymatic reactions. One of the enzymes, thymidine phosphorylase, is present at higher levels in tumours, which provides a higher efficacy with minor toxicity. Capecitabine has demonstrated a similar efficacy to 5-FU/LV in two randomised trials for first-line treatment of mCRC [4,5], and only hyperbilirubinaemia and hand-foot syndrome are more frequent with the oral schedule. Dose-limiting toxicity is manifested with diarrhoea. Capecitabine is frequently associated with hand-foot syndrome (up to 60%), generally in a dose-related manner. Although the standard dose in monotherapy is 1250 mg/m2 twice a day for 14 of every 21 days, a reduced dose is usually needed. UFT is a 1:4 molar combination of ftorafur and uracil, which competitively inhibits the degradation of 5FU. The association of UFT plus oral LV has a comparable efficacy and better tolerability compared to intravenous bolus 5FU [6,7].
2.2. Irinotecan Irinotecan is a topoisomerase I inhibitor. In patients with mCRC, it has demonstrated clinical benefit after failure of 5-FU as a single agent, with improvement in survival and quality of life [8]. Administration weekly, every two weeks, and every three weeks has yielded similar efficacies, but infusion every three weeks seems to have a better profile of toxicity with less grade-3 diarrhoea, which is the dose-limiting side effect [9]. In three randomised trials, the combination of irinotecan and 5FU has also been demonstrated to benefit survival when compared to 5FU/LV [10,11,12]. Drug sequencing and method of administration influence toxicity. Bolus 5-FU-based regimens such as IFL are associated with more gastrointestinal toxicity than infusional 5-FU-based regimens such as FOLFIRI [13]. Pharmacokinetic variability of irinotecan has been related to biliary excretion and inherited alterations in the hepatic metabolic pathways that control drug disposition, such as polymorphisms in uridine diphosphoglucuronosyl transferase 1A1 (UGT1A1) [14]. The combination of irinotecan and capecitabine has also been explored and showed efficacy, but in some cases gastrointestinal and haematologic toxicity
142
Paula GarcĂa Teijido et al.
limited the schedules for dose and administration [15,16]. Dose-limiting toxicity is manifested as diarrhoea and neutropenia but is also associated with asthenia, alopecia, nausea, and hepatotoxicity.
2.3. Oxaliplatin Oxaliplatin is a second-generation platinum derivative with significant antitumour activity when combined with 5FU, especially in mCRC, but also in other digestive tumours such as gastric neoplasia. As a single agent, oxaliplatin has shown a low level of activity with a 20-25% response rate and is therefore considered an inappropriate choice as a first-line treatment [17,18]. However, 5FU seems to synergise with oxaliplatin. Two randomised trials combine oxaliplatin with 5FU/LV, one with the de Gramont regimen [19] and the other with a low-dose Mayo regimen [20]. In both, adding oxaliplatin improves response rates to 48-51% and progression-free survival time to 7.9-9 months without a significant benefit in overall survival. We must consider, though, that patients receiving oxaliplatin develop dose-related neurotoxicity, which substantially limits the administration of the drug. The combination of oxaliplatin and capecitabine (XELOX or CAPOX) has been widely explored, and results suggest comparable efficacy with a different profile of toxicity [21]. The XELOX regimen produces higher rates of hand-foot syndrome, thrombocytopenia, grade 3-4 nausea-vomiting, and neuropathy. The standard regimen should include capecitabine at 1000 mg/m2/12h for 14 days and oxaliplatin at 130 mg/m2 on day 1, all repeated every 3 weeks, but lower doses of oxaliplatin can be considered for elderly patients. Neutropenia, nausea, fatigue, and diarrhoea are common symptoms of toxicity, but the most prevalent is neuropathy. Two different syndromes are involved: an acute neurosensory complex, and a cumulative sensory neuropathy, with distal loss of sensation and dysesthesias. The dose-limiting, delayed neuropathy seen with oxaliplatin is comparable to that observed with cisplatin.
2.4. Irinotecan versus oxaliplatin Available data suggest that the efficacies of first-line treatments with oxaliplatin/5FU/LV and irinotecan/5FU/LV are similar. Many different regimens of both drugs have been compared, with equivalent results but different profiles of toxicity [22,23,24]. Grothey et al. [25] also compared CAPOX and CAPIRI in first-line treatments with a cross-over for second-line therapy. Median overall survival (OS) time was again similar for the two groups (17.7 and 17.8 months, respectively). The choice of which regimen to
143
Systemic treatment for metastatic colorectal cancer
use should be based on toxicity profiles of each regimen and patient preferences.
3. Biological agents 3.1. Agents targeting VEGF: Bevacizumab Bevacizumab is a recombinant, humanised, monoclonal antibody that blocks the activity of the vascular endothelial growth factor (VEGF), a central mediator of tumour-related angiogenesis, expression of which is prognostic for a number of solid tumours. In CRC, increased expression of VEGF correlates with invasiveness, vascular density, metastasis, recurrence, and poor prognosis. Although the single-agent activity of bevacizumab is dismal, several phase III trials have demonstrated that the addition of bevacizumab to chemotherapy prolongs disease-free and overall survival. These advances, however, have come at the cost of treatment-related side effects that include bleeding, hypertension, bowel perforation, and thromboembolic events. The most prevalent symptom of toxicity is hypertension, at a rate between 23-67%. Irinotecan and 5-FU/LV clearly benefit from the addition of bevacizumab (Table1), but these differences are less evident with oxaliplatin-based regimens. Table 1. Selected trials with bevacizumab as part of initial treatment in mCRC. Trial Kabbinavar 2003 [26]
Kabbinavar 2005 [27] Hurwitz 2004 [28]
Fuchs 2007 [16], 2008 [29]
Saltz 2008 [30]
Grothey 2008 [31] Giantonio 2007 [32]
Treatment Arms 5-FU/LV 5-FU/LV + BEV5 5-FU/LV + BEV10 5-FU/LV 5-FU/LV + BEV IFL IFL + BEV 5-FU/LV + BEV FOLFIRI + BEV mIFL + BEV CapeIri (closed) FOLFOX/CAPOX FOLFOX/CAPOX BEV CT + BEV FOLFOX FOLFOX + BEV10 BEV 10
PFS (months) 5.2 9.0 (p=0.005) 7.2 (p=0.217) 5.5 9.2 (p=0.0002) 6.2 10.6 (p<0.001) 8.8 (p=0.4192) 11.2 NS 8.3
OS (months) 13.8 21.5 16.1 12.9 16.6 (p=0.16) 15.6 20.3 (p<0.001) 18.3 (p=0.2521) 28.0 (p=0.037) 19.2
8.0 9.4(p<0.0023)
19.9 21.3 (p<0.077)
9.9 4.7 7.3 (p<0.0001) 2.7
25.1 10.8 12.9 (p=0.0011) 10.9
BEV=bevacizumab, 5=5 mg/kg, 10=10 mg/kg, NS=non significant, CT=chemotherapy, OS=overall survival
144
Paula García Teijido et al.
3.2. Agents targeting EGFR: Cetuximab and panitumumab EGFR is overexpressed in about 70% of patients with mCRC and has been considered a poor prognostic factor [33]. Two monoclonal antibodies targeting the EGFR are active for mCRC: cetuximab and panitumumab. Molecular events such as overexpression or mutation of genes involved in the signaling pathway of EGFR identify patients that are more likely to respond to anti-EGFR drugs. For example, mutations in KRAS are present in 40% of patients and predict poorer outcome and resistance to cetuximab or panitumumab [34]. A wild-type KRAS, though, does not ensure response to these treatments. On the other hand, activating mutations of BRAF, mutually exclusive with mutations in K-RAS, occur in about 10% of patients and are also a poor prognostic factor. The NCCN guideline recommends routine K-RAS testing at the time of diagnosis of metastatic disease to facilitate the planning of treatment across a continuum of care [35]. 3.2.1. Cetuximab Cetuximab is a recombinant, human/mouse chimeric immunoglobulin (Ig) G1 monoclonal antibody that binds specifically to the extracellular domain of the human EGFR. Cetuximab blocks the activation of the EGFR by preventing ligand-mediated tyrosine-kinase phosphorylation and downstream signal transduction. Cetuximab in monotherapy has been compared with best supportive care in a study conducted by Jonker et al. [36]. The median OS time was 6.1 months in the cetuximab group and 4.6 months in the group assigned to best supportive care. This monoclonal antibody is also useful in combination with irinotecan in patients with irinotecan-refractory mCRC (23% response rate versus “vs” 11% with cetuximab in monotherapy) [37]. Two studies evaluated the addition of cetuximab to first-line chemotherapy in mCRC: the phase II OPUS study (FOLFOX4 with or without cetuximab) [38] and the phase III CRYSTAL study (FOLFIRI with or without cetuximab) [39]. In both trials, cetuximab improved response rate and PFS, and the CRYSTAL study also improved OS (in wild-type K-RAS). A meta-analysis of both trials presented at ASCO 2010 [40] confirmed the benefit in wild-type K-RAS of adding cetuximab: median OS time was extended by 4 months (23.5 vs 19.5, HR, 0.81, p=0.0062). The first results of the COIN study [41], however, showed no improvement in OS among patients with wild-type K-RAS. Further studies are needed to evaluate the impact of cetuximab in the survival of patients with wild-type K-RAS.
Systemic treatment for metastatic colorectal cancer
145
The most common adverse effects of cetuximab are weakness, an acneiform rash, nausea, infusion-related reactions, and hypomagnesaemia. A relation among development, severity of treatment-induced skin rash, and response to anti-EGFR monoclonal antibodies has been suggested [37]. 3.2.2. Panitumumab Panitumumab is a fully human, Ig G2, anti-EGFR, monoclonal antibody that binds with high affinity to the EGFR. Panitumumab has activity in monotherapy, with a response rate of 10% [42]. A phase III PRIME study [43] evaluated the efficacy and safety of FOLFOX 4 with or without panitumumab in first-line treatment. Panitumumab significantly improved PFS, with a median increase of 9.6 vs 8.0 months (HR=0.80, p=0.02). A trend towards an OS benefit was also observed. In patients with mutated-KRAS tumours, outcomes were inferior with panitumumab. No data are available comparing panitumumab and cetuximab (or either vs a bevacizumab-containing regimen), and the place of panitumumab, particularly for first-line therapy of mCRC, remains uncertain [35].
3.3. Dual antibody therapy combination of VEFG and EGFR inhibitors Preclinical models have demonstrated that EGFR stimulation leads to increased VEGF expression, and VEGF expression is induced by increased circulating levels of EGF, indicating an interaction between these pathways and providing a rationale for the hypothesis that dual inhibition would offer greater inhibition of tumour growth. The benefit of associating a VEGF and an EGFR inhibitor to standard chemotherapy has been addressed in three trials to date: BOND-2 (cetuximab + bevacizumab +/- irinotecan) [44], PACCE (bevacizumab + oxaliplatin or irinotecan-based chemotherapy +/- panitumumab) [45], and CAIRO-2 (XELOX + bevacizumab +/- cetuximab) [46]. After surprisingly negative results, combining VEGF and EGFR inhibitors is not being further studied as a treatment for CRC.
3.4. New approaches 3.4.1. Anti-VEGF drugs Multitargeted VEGF inhibitors Several inhibitors of the VEGF pathway, mostly tyrosine-kinase inhibitors such as sunitinib, are in different phases of clinical development.
146
Paula GarcĂa Teijido et al.
However, no phase II or III trials have demonstrated a benefit either as monotherapy or in combination with chemotherapy. Aflibercept (AVE-0005) is a fully human, recombinant, fusion protein composed of the second Ig domain of VEGFR-1 and the third Ig domain of VEGFR-2, fused to the Fc region of human IgG1. It binds to all isoforms of VEGF-A and to PIGF, thereby preventing VEGFs from binding to their cell receptors and stimulating angiogenesis. The combination of aflibercept and irinotecan is being tested in the metastatic setting. Cediranib is a highly potent inhibitor of VEGFR-2 (KDR) that inhibits VGF-dependent signaling, angiogenesis, and neovascular survival. It is also a potent inhibitor of VEGFR-1 and VEGFR-3 tyrosine kinases. The development of cediranib as a treatment for CRC is still proceeding. 3.4.2. Extra-cellular targeted therapies TRAIL-receptor inhibitors TRAIL/Apo2L is a cytokine member of the TNF superfamily that activates apoptosis through cell-surface death receptors. In CRC studies, results from a phase II study of mapatumumab (TRAIL-R1 Ab) as a single agent failed to demonstrate any objective responses [47]. Despite these results as a single agent, the strategy of targeting this pathway in CRC is currently being evaluated in several clinical trials, predominantly in combination with chemotherapy where TRAIL-receptor inhibitors are thought to be most effective. IGF-1R inhibitors The IGF pathway is a key regulator of normal cell proliferation, differentiation, and apoptosis and has been shown to be important in the development, maintenance, and progression of cancer. In fact, increased levels of IGFs are associated with an increased risk of the development of CRC, because IGF-1R is widely expressed in CRC [48]. Several studies have examined the potential role of IGF-1R inhibitors in CRC both as a single agent and in combination with chemotherapy or EGFR antibodies. C-MET/HGF inhibitors The tyrosine-kinase receptor, c-MET, and its ligand, HGF, are expressed in both normal and malignant cells. In cancer, strong evidence supports the role of the c-MET pathway in tumorigenesis, and the presence of receptor and/or ligand overexpression has been associated with a poor prognosis [49].
Systemic treatment for metastatic colorectal cancer
147
A trial evaluating the HGF antibody, AMG102, in combination with panitumumab vs panitumumab alone in patients with wild-type KRAS metastatic CRC has not yet reported results. Hedgehog inhibitors In adults, the Hedgehog pathway appears to be involved in the regulation of stem cells that govern tissue maintenance and regeneration [50]. A randomised study of GDC-0449, an SMO antagonist, with chemotherapy in CRC is ongoing.
3.4.3. Intracellular targeted therapies RAS inhibitors: Investigation previously focused on the inhibition of farnesylation of the RAS protein, but failure of this inhibition is frequent, perhaps due to the ability to undergo alternative pathways. Focus has now shifted to targeting the down-stream effectors such as B-RAF or MEK. B-RAF inhibitors: Preclinical studies exhibit the efficacy of a B-RAF inhibitor as monotherapy or in combination with capecitabine +/- bevacizumab in a CRC-xenograft model [51]. PI3K/AKT/MTOR inhibitors: The combination of these agents with chemotherapy and inhibitors of angiogenesis is of interest, because the VEGF pathway in endothelial cells signals through the PI3K pathway. MEK inhibitors: A randomised phase II study of AZD6244 vs capecitabine in patients with refractory disease failed to demonstrate any benefit [52]. Predictive biomarkers are being pursued because mutation of upstream genes is not sufficient for selecting patients that will benefit from these drugs in CRC. Poly(ADP-Ribose) polymerase inhibitors: Few phase II studies have been performed of the PARP inhibitors, olaparib and ABT-888, as single agents or in combination with chemotherapy. ABT-888 has been tested in CRC cells with microsatellite instability (MSI). MSI cell lines deficient in double-strand break repair have shown a high sensitivity to PARP inhibitors [53], but these results need to be confirmed in further clinical trials. SRC: The Src pathway is activated in 80% of all human colorectal tumours and is associated with a poor clinical prognosis [54]. Dasatinib, a multitargeted tyrosine-kinase inhibitor that inhibits src, BCR/ABL, c-Kit, and other targets, is being tested in clinical trials with chemotherapy.
148
Paula GarcĂa Teijido et al.
4. Treatment goals When treating a patient with mCRC, the first determination is whether the patient can be surgically cured. If the case does not appear curable, the main goals are to prolong OS and to maintain quality of life for as long as possible. Two different scenarios and associated-treatment goals can be identified:
4.1. Potentially resectable disease A small percentage of patients with stage IV disease are potentially curable by a surgical resection of metastases either at the time of diagnosis or after downsizing initially unresectable metastases by neoadjuvant chemotherapy. The term â&#x20AC;&#x153;conversion therapyâ&#x20AC;? has been proposed to designate the use of induction chemotherapy in patients with initially unresectable CRC liver metastases [55]. The appropriate chemotherapy in this setting is uncertain, but some reasonable options are listed in Table 2. All these regimens can be given either alone or in combination with an appropriate monoclonal antibody (bevacizumab or cetuximab if K-RAS wild-type). Table 2. Conversion therapies.
Systemic treatment for metastatic colorectal cancer
149
4.2. Nonresectable disease Approximately 15-25% of patients with mCRC present with synchronous liver metastases, although most are initially considered unresectable [58]. For asymptomatic nonresectable patients, inducing a tumour response is less important than delaying tumour progression for as long as possible. In this setting, the access of patients to all active agents is crucial to increase the median overall survival.
5. Timing of chemotherapy Many patients are asymptomatic at diagnosis of metastatic disease. The value of early chemotherapy vs deferral of treatment until symptoms develop is controversial. Only two randomised trials have been performed, both with 5-FU and LV, one negative but another positive for earlier treatment [59]. Resection is the standard approach for the local treatment of metastatic disease that is initially resectable or converted to a potentially curable status following chemotherapy [60].
6. Duration In advanced CRC, continued chemotherapy until progression or symptoms of toxicity appear is one option, but intermittent therapy followed by a chemotherapy â&#x20AC;&#x153;holidaysâ&#x20AC;? is commonly used (especially with oxaliplatin regimens). The rationale for using treatment-free intervals includes an increase in the quality of life, reduction in cumulative toxicities, prevention of premature discontinuation of therapy, preservation of the ability to administer further therapy, and reduction in cost [61]. With oxaliplatin-based therapies, more than 50% of patients discontinue treatment for reasons other than disease progression, i.e. mainly neurologic or haematologic toxicity [62]. For nearly all patients, oxaliplatin should be discontinued prior to the development of neuropathy [61]. Several trials in the past have tested the influence of chemo-holidays in patient outcome, defining different types of treatment breaks: a.
Complete chemotherapy-free intervals, stopping all therapeutic agents, as explored in the OPTIMOX-2 trial [63]. b. Stopping only those agents with significant toxicity while continuing others as maintenance therapy, e.g. OPTIMOX-1 [64] and CONcePT [65] trials. c. Discontinuation of all conventional cytotoxic agents but continuation of biological agents.
150
Paula García Teijido et al.
Data from the OPTIMOX-2 trial showed that continued treatment with maintenance chemotherapy (non-oxaliplatin) as compared with a completely drug-free interval resulted in a significantly longer duration of median PFS (8.6 vs 6.6 months) and a trend towards a longer survival time (24 vs 20 months) [63]. Additional evidence comes from the phase III NO16966 trial in which early discontinuation of bevacizumab and chemotherapy (without maintenance) was associated with a shorter PFS than expected [30]. According to these results, a full break in therapy would not be routinely recommended. Some clinical trials are currently under way to elucidate the role of chemo-holidays with chemotherapy plus targeted agents. Patients in the MACRO study were randomised to receive maintenance therapy with bevacizumab + XELOX or bevacizumab alone after 6 cycles of induction therapy with bevacizumab + XELOX [66]. Odds ratio, PFS, and OS were not statistically different between the two arms. Agents can be discontinued after a predefined number of treatment cycles, when a “best response” is achieved or when a long-term, stable disease is observed. Re-initiation of complete schema can be planned after a predefined interval (as in the OPTIMOX-1 and CONcePT trials) when relevant tumour progression has been observed (OPTIMOX-2) or after specific toxicities have been resolved. Whether survival is adversely impacted by a “stop and go” strategy of chemotherapy when a complete response is not achieved is unclear, but such a strategy seems necessary in the case of combinations with oxaliplatin. The latest guidelines recommend a maintenance therapy, except after two cycles of chemotherapy following a complete response [35].
7. Initial doublet or triplet combination versus sequential single agents 7.1. Initial doublet vs sequential single agents The proportion of patients exposed to the three conventional agents of chemotherapy (fluoropyrimidines, irinotecan, and oxaliplatin) during the course of therapy correlates strongly with median survival in all large phase III trials in the last ten years [67]. The caveats of using sequential single agents include the lack of all drugs having documented single-agent activity (oxaliplatin and bevacizumab), the potential synergistic effect of combinations, and partial dropout of patients along the course of the treatment. OS was not found to be associated with the order in which these drugs were received. The issue of initial combination vs single-agent therapy was directly addressed in the FOCUS [62] and CAIRO [68] trials. The first study compares first-line 5-FU/LV followed by irinotecan, 5-FU/LV followed by irinotecan
Systemic treatment for metastatic colorectal cancer
151
plus 5-FU, and either FOLFIRI/FOLFOX. In the CAIRO trial, previously untreated patients were randomised to capecitabine followed by irinotecan and then XELOX, or initial XELIRI followed by XELOX. In both trials the median OS was similar for sequential and initial combination therapy, although PFS was superior with combination therapy (either FOLFOX/FOLFIRI or XELOX/XELIRI). Only 23-62% of all patients received second-line therapy, and the median OS times for all groups in both trials were lower than expected (17 months on combination arm). The available data support initial combination therapy for most patients with mCRC, not only those who have limited liver metastases that might become potentially resectable but also for inoperable mCRC. This therapy is achieved by using well-established doublets such as FOLFOX, XELOX, or FOLFIRI. In addition, a complete response to initial therapy (which was more likely with a combination of drugs) was an independent predictor of survival [69].
7.2. Initial triplet Whether a combination of all three active chemotherapies (FOLFOXIRI: 5-FU-LV, irinotecan, and oxaliplatin) as first-line treatment is better than 5-FU-LV with irinotecan was explored in two small clinical trials, giving contradictory data. A Greek trial showed no benefit for FOLFOXIRI [70]. In contrast, an Italian trial with 244 patients found that FOLFOXIRI was significantly superior for improving response rate (66 vs 41%), median PFS (9.8 vs 6.9 months), and OS time (22.6 vs 16.7 months). Only the response rate, however, was the primary endpoint in the latter study [57]. A regimen such as FOLFOXIRI might be considered for conversion therapy in patients with initially unresectable liver metastases and good performance status (PS).
8. Planning therapy 8.1. Initial treatment The choice of therapy is based on a consideration of the goals of therapy, prior chemotherapy administered, and profiles of toxicity. Although data are limited in support, chemotherapy should start at diagnosis of metastases. The treatment should not be restricted to patients with PS < 2. A study from nine trials (more than 6000 patients) that evaluated the benefits and risks associated with intensive first-line treatment in the setting of mCRC according to patient PS showed similar therapeutic efficacy for patients with PS 2 or less compared with control groups, but increased gastrointestinal toxicities for patients with PS 2 [71].
152
Paula GarcĂa Teijido et al.
A chemotherapeutic doublet such as FOLFOX, FOLFIRI, or CapeOX is the backbone for patients able to tolerate it. FOLFOX and FOLFIRI have similar firstline efficacies but different profiles of toxicity. Many options of initial treatment are available (35): FOLFOX, FOLFIRI, or CapeOX, all with or without bevacizumab. With wild-type K-RAS, other choices are the same options for chemotherapy as above but with or without cetuximab or panitumumab.
8.2. Treatment after tumour progression Results are poor in this context, with responses between 4-16%. If oxaliplatin was used in first-line treatment, switching to irinotecan is indicated. If the patient was treated with a combination with irinotecan, two options are available: either add cetuximab or panitumumab at the time of progression (if KRAS is wild-type), or switch to FOLFOX or XELOX. If bevacizumab is part of the initial treatment, the lack of data cannot justify continuing bevacizumab in second-line treatment after progressing on first-line treatment.
Conclusions Some patients with mCRC can be cured with an interdisciplinary approach that includes surgery. Most patients, however, are considered nonresectable, and chemotherapy is offered to prolong overall survival. Although many patients tolerate a chemotherapeutic doublet, not all need it. Molecular targeted therapy increases anticancer activity and has improved outcomes, but not as much as expected. Clinicians should first analyse the goal of therapy when treating a patient with mCRC. Identifying initially resectable tumours is the first step, and conversion chemotherapy can select another small group of patients that could benefit from surgery. For most patients, however, prolonged survival and improved quality of life are the primary goals. Recent advances in treatment have extended median patient survival times by more than two years. Maximising efficacy must be balanced with minimising toxicities, and therapy should be individualised based on known clinical factors.
References 1. 2. 3.
Scheithauer, W., Rosen, H., Kornek, G.V., Sebesta, C., and Depisch, D. 1993, BMJ, 306, 752. Meta-analysis Group in Cancer. 1998, J. Clin. Oncol., 16, 301. Thirion, P., Michiels, S., Pignon, J.P., Buyse, M., Braud, A.C., Carlson, R.W., et al. . 2004, J. Clin. Oncol., 22, 3766.
Systemic treatment for metastatic colorectal cancer
4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.
153
Hoff, P.M., Ansari, R., Batist, G., Cox, J., Kocha, W., Kuperminc, M., et al. 2001, J. Clin. Oncol., 19, 2282. Van Cutsem E, Twelves C, Cassidy J, Allman D, Bajetta E, Boyer M et al 2001, J Clin Oncol, 19, 4097. Carmichael, J., Popiela, T., Radstone, D., Falk, S., Borner, M., Oza, A., et al. 2002, J. Clin. Oncol., 20, 3617. Douillard, J.Y., Hoff, P.M., Skillings, J.R., Eisenberg, P., Davidson, N., Harper, P., et al. 2002, J. Clin. Oncol., 20, 3605. Cunningham, D., Pyrhonen, S., James, R.D., Punt, C.J., Hickish, T.F., Heikkila, R., et al. 1998, Lancet, 352, 1413. Fuchs, C.S., Moore, M.R., Harker, G., Villa, L., Rinaldi, D., and Hecht, J.R. 2003, J. Clin. Oncol., 21, 807. Douillard, J.Y., Cunningham, D., Roth, A.D., Navarro, M., James, R.D., Karasek, P., et al. 2000, Lancet, 355, 1041. Saltz, L.B., Cox, J.V., Blanke, C., Rosen, L.S., Fehrenbacher, L., Moore, M.J., et al. 2000, N. Engl. J. Med., 343, 905. Kohne, C.H., Van Cutsem, E., Wils, J., Bokemeyer, C., El-Serafi, M., Lutz, M.P., et al. 2005, J. Clin. Oncol., 23, 4856. Van Cutsem, E., Douillard, J.Y., and Kohne, C.H. 2001, N. Engl. J. Med., 345, 1351. Innocenti, F., Kroetz, D.L., Schuetz, E., Dolan, M.E., Ramirez, J., Relling, M., Chen, P., Das, S., Rosner, G.L., and Ratain, M.J. 2009, J. Clin. Oncol., 27, 2604. Patt, Y.Z., Lee, F.C., Liebmann, J.E., Diamandidis, D., Eckhardt, S.G., Javle, M., et al. 2007, Am. J. Clin. Oncol., 30, 350. Fuchs, C.S., Marshall, J., Mitchell, E., Wierzbicki, R., Ganju, V., Jeffery, M., et al. 2007, J. Clin. Oncol., 25, 4779. Becouarn, Y., Ychou, M., Ducreux, M., Borel, C., Bertheault-Cvitkovic, F., Seitz, J.F., et al.1998, J. Clin. Oncol., 16, 2739. Rothenberg, M.L., Oza, A.M., Bigelow, R.H., Berlin, J.D., Marshall, J.L., Ramanathan, R.K., et al. 2003, J. Clin. Oncol., 21, 2059. de Gramont, A., Figer, A., Seymour, M., Homerin, M., Hmissi, A., Cassidy, J., et al. 2000, J. Clin. Oncol., 18, 2938. Grothey, A., Deschler, B., Kroening, H., Ridwelski, K., Reichardt, P., Kretzschmar, A., et al. 2002, Proc. Am. Soc. Clin. Oncol., 21, Abstr. 512. Arkenau, H.T., Arnold, D., Cassidy, J., Diaz-Rubio, E., Douillard, J.Y., Hochster, H., et al. 2008, J. Clin. Oncol., 26, 5910. Delaunoit T, Goldberg RM, Sargent DJ. 2004, Cancer, 101, 2170 Tournigand, C., Andre, T., Achille, E., Lledo, G., Flesh, M., Mery-Mignard, D., et al. 2004, J. Clin. Oncol., 22, 229. Colucci G, Gebbia V, Paoletti G, Giuliani F, Caruso M, Gebbia N et al. J Clin Oncol 2005;23:4866 Grothey, A., Jordan, K., Kellner, O., Constantin, C., Dietrich, G., Kroening, H., et al. 2002, Ann. Oncol., 13 (Suppl. 5), Abstr. 258. Kabbinavar, F., Hurwitz, H.I., Fehrenbacher, L., Meropol, N.J., Novotny, W.F., Lieberman, G., et al. . 2003, J. Clin. Oncol., 21, 60.
154
Paula GarcĂa Teijido et al.
27. Kabbinavar, F.F., Schulz, J., McCleod, M., Patel, T., Hamm, J.T., Hecht, J.R., et al. 2005, J. Clin. Oncol., 23, 3697. 28. Hurwitz, H., Fehrenbacher, L., Novotny, W., Cartwright, T., Hainsworth, J., Heim, W., et al. 2004, N. Engl. J. Med., 350, 2335. 29. Fuchs, C.S., Marshall, J., and Barrueco, J. 2008, J. Clin. Oncol., 26, 689. 30. Saltz, L.B., Clarke, S., DĂaz-Rubio, E., Scheithauer, W., Figer, A., Wong, R., et al. 2008, J. Clin. Oncol., 26, 2013. 31. Grothey, A., Sugrue, M.M., Purdie, D.M., Dong, W., Sargent, D., Hedrick, E., and Kozloff, M. 2008, J. Clin. Oncol., 26, 5326. 32. Giantonio, B.J., Catalano, P.J., Meropol, N.J., O'Dwyer, P.J., Mitchell, E.P., Alberts, S.R., et al. 2007, J. Clin. Oncol., 25, 1539. 33. Messa, C., Russo, F., Caruso, M.G., and Di Leo, A. 1998, Acta Oncol., 37, 285. 34. Etienne-Grimaldi, M.C., Formento, J.L., Francoual, M., Francois, E., Formento, P., Renee, N., et al. 2008, Clin. Cancer Res., 14, 4830. 35. National Comprehensive Cancer Network. 2010, v.3.2010, www.nccn.org/professionals/physician_gls/f_guidelines.asp. 36. Jonker, D.J., O'Callaghan, C.J., Karapetis, C.S., Zalcberg, J.R., Tu, D., Au, H.J., et al. 2007, N. Engl. J. Med., 357, 2040. 37. Cunningham, D., Humblet, Y., Siena, S., Khayat, D., Bleiberg, H., Santoro, A., et al. 2004, N. Engl. J. Med., 351, 337. 38. Bokemeyer, C., Bondarenko, I., Makhson, A., Hartmann, J.T., Aparicio, J., de Braud, F., et al. 2009, J. Clin. Oncol., 27, 663. 39. Van Cutsem, E., Kohne, C.H., Hitre, E., Zaluski, J., Chang Chien, C.R., Makhson, A., et al. 2009, N. Engl. J. Med., 360, 1408. 40. Bokemeyer, C., Kohne, C., Rougier, P., Stroh, C., Schlichting, M., and Van Cutsem, E. 2010, J. Clin. Oncol., 28 (Suppl. 15S), Abstr. 3506. 41. Maughan, T.S., Adams, R., Smith, C.G., Seymour, M.T., Wilson, R.H., Meade, A.M., et al. 2010, J. Clin. Oncol., 28 (Suppl. 15S), Abstr. 3502. 42. Van Cutsem, E., Peeters, M., Siena, S., Humblet, Y., Hendlisz, A., Neyns, B., et al. . 2007, J. Clin. Oncol., 25, 1658. 43. Douillard, J.Y., Siena, S., Cassidy, J., Tabernero, J., Burkes, R., Barugel, M., et al. 2010, J. Clin. Oncol., 28, 4697. 44. Saltz, L.B., Lenz, H.J., Kindler, H.L., Hochster, H.S., Wadler, S., Hoff, P.M., et al. 2007, J. Clin. Oncol., 25, 4557. 45. Hecht, J.R., Mitchell, E., Chidiac, T., Scroggin, C., Hagenstad, C., Spigel, D., et al. 2009, J. Clin. Oncol., 27, 672. 46. Tol, J., Koopman, M., Cats, A., Rodenburg, C.J., Creemers, G.J., Schrama, J.G., et al. 2009, N. Engl. J. Med., 360, 563. 47. Trarbach, T., Moehler, M., Heinemann, V., Kohne, C.H., Przyborek, M., Schulz, C., et al. . 2010, Br. J. Cancer, 102, 506. 48. Weber MM, Fottner C, Liu SB, Jung MC, Engelhardt D, Baretton GB et al. 2002, Cancer, 95, 2086. 49. Boccaccio, C. and Comoglio, P.M. 2006, Nat. Rev. Cancer, 6, 637. 50. Beachy, P.A., Karhadkar, S.S., and Berman, D.M. 2004, Nature, 432, 324.
Systemic treatment for metastatic colorectal cancer
155
51. Kolinsky, K.D., Su, F., Bollag, G., Lee, R., Packman, K., Schostack, K., et al. 2009, Gastrointestinal Cancers Symposium, Abstr. 362. 52. Bennouna, J., Lang, I., Valladares-Ayerbes, M., Boer, K., Adenis, A., Escudero, P., et al. 2010, Invest. New Drugs. 53. Vilar Sanchez, E., Chow, A., Raskin, L., Iniesta, M.D., Mukherjee, B., and Gruber, S.B. 2009, J. Clin. Oncol., 27 (Suppl. 15S), Abstr. 11028. 54. Aligayer, H., Boyd, D.D., Heiss, M.M., Abdalla, E.K., Curley, S.A., and Gallick, G.E. 2002, Cancer, 94, 344. 55. Petrelli, N.J. 2007, Proc. Am. Soc. Clin. Oncol., Plenary Program discussion. 56. Hochster, H.S., Hart, L.L., Ramanathan, R.K., Childs, B.H., Hainsworth, J.D., Cohn, A.L., et al. 2008, J. Clin. Oncol., 26, 3523. 57. Falcone, A., Ricci, S., Brunetti, I., Pfanner, E., Allegrini, G., Barbara, C., et al. 2007, J. Clin. Oncol., 25, 1670. 58. Adam, R., Delvart, V., Pascal, G., Valeanu, A., Castaing, D., Azoulay, D., et al. 2004, Ann. Surg., 240, 644. 59. Ackland, S.P., Jones, M., Tu, D., Simes, J., Yuen, J., Sargeant, A.M., et al. 2005, Br. J. Cancer, 93, 1236. 60. Abdalia, E.K. 2009, Am. J. Surg., 197, 737. 61. Grothey, A. 2009, 2009 Educational Book, Am. Soc. Clin. Oncol., Alexandria, USA, 209. 62. Seymour, M.T., Maughan, T.S., Ledermann, J.A., Topham, C., James, R., Gwyther, S.J., et al. 2007, Lancet, 370, 143. 63. Maindrault-Goebel, F., Lledo, G., Chibaudel, B., Mineur, L., Andre, T., Bennamoun, M., et al. 2007, J. Clin. Oncol., 25 (Suppl. 18S), Abstr. 4013. 64. Tournigand, C., Cervantes, A., Figer, A., Lledo, G., Flesch, M., Buyse, M., et al. 2006, J. Clin. Oncol., 24, 394. 65. Grothey, A., Hart, L.L., Rowland, K.M., Ansari, R.H., Alberts, S.R., Chowhan, N.M., et al. . 2008, J. Clin. Oncol., 26 (May 20 Suppl.), Abstr. 4010. 66. Tabernero, J., Aranda, E., Gomez, A., Massuti, B., Sastre, J., Abad, A., et al. 2010, J. Clin. Oncol., 28 (Suppl. 15S), Abstr. 3501. 67. Goldberg, R.M., Rothenberg, M.L., Van Cutsem, E., Benson, A.B., 3rd, Blanke, C.D., Diasio, R.B., et al. 2007, Oncologist, 12, 38. 68. Koopman, M., Antonini, N.F., Douma, J., Wals, J., Honkoop, A.H., Erdkamp, F.L., et al. 2007, Lancet, 370, 135. 69. Dy, G.K., Krook, J.E., Green, E.M., Sargent, D.J., Delaunoit, T., Morton, R.F., et al. 2007, J. Clin. Oncol., 25, 3469. 70. Souglakos, J., Androulakis, N., Syrigos, K., Polyzos, A., Ziras, N., Athanasiadis, A., et al. 2006, Br. J. Cancer, 94, 798. 71. Sargent, D.J., Kohne, C.H., Sanoff, H.K., Bot, B.M., Seymour, M.T., de Gramont, A., et al. 2009, J. Clin. Oncol., 27, 1948.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 157-186 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
7. Site-specific therapy of metastatic colorectal cancer: Focus on surgical treatment of liver metastases Miguel Ángel Suárez Muñoz, Julio Santoyo Santoyo José Luis Fernández Aguilar, Belinda Sánchez Pérez José Antonio Pérez Daga and Mª Carmen Leiva Vera General, Digestive and Transplantation Surgery Service “Carlos Haya” University Hospital, Málaga, Spain
Abstract. Over the past 10 years, the criteria for defining resectability for patients with CRC metastases have expanded. In the past, such features as the number of metastases, the size of the tumour lesion, and a mandatory 1-cm margin of resection dictated who was “resectable”. The criteria for resectability have recently been expanded to include any patient in whom all disease can be removed with a negative margin and who has adequate hepatic volume/reserve. Specifically, instead of resectability being defined by what is removed, decisions concerning resectability now centre around what will remain after resection. Under this new paradigm, the number of patients with resectable disease can be expanded by increasing/preserving hepatic reserve (e.g., portal vein embolisation, two-stage hepatectomy), combining resection with ablation, and Correspondence/Reprint request: Dr. Miguel Ángel Suárez Muñoz, General, Digestive and Transplantation Surgery Service, “Carlos Haya” University Hospital, Málaga, Spain. E-mail: masuarez59@gmail.com
158
Miguel テ]gel Suテ。rez Muテアoz et al.
decreasing tumour size (preoperative chemotherapy). The criteria for resectability have also expanded to include patients with extrahepatic disease. The presence of both intraand extrahepatic disease, rather than being an absolute contraindication to surgery, should potentially allow patients to be considered for resection based on strict selection criteria. The expansion of criteria for resectability of colorectal liver metastases requires a much more nuanced and sophisticated approach to the patient with advanced disease. A therapeutic approach that includes all aspects of multidisciplinary and multimodal care is required to select and treat this complex group of patients. We can only hope that future progress in the multidisciplinary care of patients with colorectal liver metastasis will lead to decreasing rates of recurrence that parallel the successes achieved in improving OS.
Introduction CRC is the third most common noncutaneous malignancy in both men and women, with approximately 150 000 cases diagnosed annually in the United States, and 25 000 in Spain, 25% of which present with liver metastases. In total, half of these patients either present with hepatic metastases or develop them during the course of their disease. These hepatic metastases are discovered synchronously in 15-25% of patients, while 20-25% of patients develop hepatic tumours metachronously. In 30-50% of patients with either synchronous or metachronous liver metastases, the liver is the only site of metastatic disease. Other organs involved in metastatic disease, in order of prevalence, are lungs, bone, and brain. Venous drainage (via the portal system) is the primary mode of dispersion. This mode of dispersion explains, in part, why distal rectal cancers more often metastasise to the lungs: the inferior rectal vein drains directly into the inferior vena cava instead of the portal system [1]. The management of colorectal liver metastases (CLM) has changed dramatically in the last decade. Modern multimodal therapies have improved OS. Liver resection remains the most important modality in the treatment of CLM. The evolution of the criteria for resectability has resulted in more patients being offered a hepatectomy. This increase is further augmented with the use of adjuncts to liver resection, including portal vein embolisation and local ablative techniques. Two-stage hepatectomy is also being used to increase resectability. OS is improved by the deployment of new chemotherapeutic agents and the use of combination chemotherapy [2]. Surgical resection is the treatment of choice in patients with CLM, with five-year survival rates reported in the range of 40-58%. Resection of liver metastases, when possible, remains the preferred therapy for potential cure. Data from several series support the expanding application of this approach,
Surgical treatment for liver metastases from colorectal cancer
159
including the LiverMetSurvey, an international registry of patients who have undergone hepatectomy for CLM. With 13 172 patients included up to December 2010, 40% remained alive at five years and 24% at 10 years [3]. Patients with untreated metastatic CRC have a short median survival time of approximately 12 months, and survival beyond five years is uncommon. Given that surgical resection remains the best chance for cure, expanding the criteria for resectability in patients with colorectal metastasis using other adjunctive and adjuvant approaches has attracted considerable interest.
1. Prognostic factors and clinical risk scores To date, numerous studies have correlated a number of demographic, clinical, operative, and pathological factors with disease-free and OS of patients treated with hepatic resection for CLM using statistical analysis comprising univariate and multivariate analyses. In comparison, prognostic scoring systems have received limited attention. In a recent review article, Gomez and Cameron [4] identified 12 prognostic scoring systems from 1996 to 2009. Nordlinger et al. [5] were the first group to propose a prognostic scoring system. They found age, size of the largest primary tumour, carcinoembryonic antigen (CEA) level, stage of the primary tumour, diseasefree interval, number of liver nodules, and resection margins to impact fiveyear survival, and proposed a clinical risk score that divided patients into three risk groups. The most recent prognostic scoring system was devised by Konopke et al. [6] based on their analysis of 201 patients over 13 years. The median followup was 31 months (range 6-143 months), and the authors included both disease-free and OS in their analyses. The disease-free and overall five-year survival rates were 28% and 43%, respectively. Only patients with a clear resection margin were included in the analysis. Multivariate analysis identified synchronous primary colorectal tumours and liver metastases, the presence of four or more liver metastases, and a CEA level over 200 ng/ml as adverse prognostic factors. Patients were divided into high-, intermediate-, and low-risk groups; the high-risk group (patients with two or more of these variables) had a median survival of 38 months, whereas the low-risk group (patients who did not exhibit any of these variables) had a median survival of 67 months. Fongâ&#x20AC;&#x2122;s clinical risk score from the Memorial Sloan-Kettering Cancer Center is the most well-known prognostic scoring system. Fong et al. [7] included 1001 patients in their study (1985-1998), which had a median follow-up of 22 months (range 0-89 months). The study identified seven independent prognostic variables derived from their multivariate analysis:
160
Miguel Ángel Suárez Muñoz et al.
lymph-node-positive in the primary tumour, a disease-free interval from primary tumour to liver metastasis of at least 12 months, more than one hepatic tumour, largest hepatic metastasis of five cm or more, CEA level over 200 ng/ml, the presence of extrahepatic disease, and an involved resection margin. Of these, margin involvement and the presence of extrahepatic disease were the most influential. In formulating their preoperative scoring system, Fong et al. excluded patients with extrahepatic disease and tumour involvement at the resection margin. Using the first five preoperative variables described above, Memorial’s clinical scoring system showed a declining five-year disease-free survival in patients scoring 0 (60%) compared with patients scoring 5 (14%). Currently, the clinical management of patients with CLM for hepatic resection lacks an ideal prognostic scoring system. Because of its potential for cure and long-term disease-free survival, hepatic resection should be offered to all patients with CLM in whom macroscopic disease can be addressed. It is important to note that clinical risk scores are prognostic tools and should not be used to deny patients a surgical resection [8]. Hence, prognostic scoring systems to select patients for resection would be considered obsolete. Although these prognostic scoring systems are clinically relevant with respect to survival, their broad application currently has limited value for patient stratification for clinical management. Any future prognostic scoring system should address patient selection for neoadjuvant or adjuvant chemotherapy and determine the surveillance protocol required by patients at higher risk of recurrence and poorer survival following hepatic resection [4,9].
2. Principles of indications for surgical treatment The philosophy (and definition) of hepatic resectability in colorectal liver metastatic disease has changed in the last few years where more extended resections are offered in selected cases somewhat regardless of the number of metastases, their locoregional invasiveness, or even the presence of demonstrable extrahepatic disease. This approach has been predicated on several important factors. Firstly, considerable improvements have been made in chemotherapeutic agents and scheduling with a greater use of neoadjuvant strategies to downstage previously defined inoperable cases to those of potential resectability [10]. This neoadjuvant strategy has recently been extended to include selected cases of first presentations of advanced metastatic disease with the primary colorectal tumour in situ [11]. Secondly, this aggressive approach has been coupled with improvements in preoperative staging that better delineates potentially resectable cases and that have shown enhanced survival when used preoperatively. This has developed in
Surgical treatment for liver metastases from colorectal cancer
161
association with technical advances in hepatic parenchymal transection and methodology of vascular isolation as well as improvements in perioperative care, each of which has resulted in very low rates of perioperative mortality even after extended resection. This level of safety has been maintained in formal hepatectomy independent of patient age, so that advanced age cannot be regarded as a medical contraindication to hepatic resection for CLM [12,13]. This extended approach towards curative hepatectomy has fostered a different attitude to the use of preoperative investigation. Efforts have become more directed towards the discovery of significant extrahepatic disease where magnetic resonance imaging (MRI) may have limitations in the detection of peritoneal and pulmonary disease. The emphasis is likely to shift in favour of initial FDG-PET and PET/CT scanning in the assessment algorithm despite limited knowledge of the economic impact, poor tissue definition, and limited availability [10]. This modality will likely be initially used selectively to define extrahepatic resectability, in situ liver recurrence, the delineation of widespread extra-abdominal disease, and in some cases of locally advanced disease at presentation with metastatic disease and an unresected primary tumour. In the experience of Fernandez et al. [14], screening by FDG-PET is associated with excellent postresection five-year OS for patients undergoing resection of hepatic metastases from CRC, allowing the definition of a new cohort of patients in whom tumour grade is a very important prognostic variable. By contrast, for Ramos et al. [15], PET/CT provided useful information in only 6.4% of cases in the selection of patients with CLM being considered for surgical therapy. In the authorsâ&#x20AC;&#x2122; opinion, PET-CT should thus be used mainly in patients with high risk of local recurrence. Contrast-enhanced helical computed tomography (HCT) is the most commonly used imaging modality for the preoperative evaluation of CLM. HCT is widely available at most institutions and allows the evaluation of both intra- and extrahepatic disease. In a recent study, Yang et al. [16] identified that the time interval between imaging and surgery is a significant factor that contributes to the discrepancy in values of HCT sensitivity reported in the current literature. Any future studies of imaging sensitivity in CLM evaluation in which intraoperative findings are used as a reference should thus control for and report the time interval to surgery. Regardless of the initial quality of imaging studies, their utility will decline unless surgery is performed in a timely manner. No standard currently defines how soon imaging should occur prior to proceeding to hepatic resection surgery, but Tanaka et al. [17] identified the mean tumour doubling time for CLM as 60.1 days. HCT investigations are proposed to be performed no more than 30 days prior to the date of surgery in order to minimise any unnecessary surgery. According to
162
Miguel テ]gel Suテ。rez Muテアoz et al.
current models, a time interval of 30 days can provide 89% sensitivity for the detection of lesions and 83% accuracy for the prediction of resectability. Intraoperative ultrasonography (IOUS) remains an important component of hepatic surgery. In addition to providing detailed visualisation of intrahepatic anatomy and guidance for needle biopsy or ablation, IOUS has long been considered important for improved intrahepatic staging and detection of lesions in patients undergoing liver surgery for colorectal metastases (Fig. 1). HCT with multidetectors and multidisciplinary meetings are the most important factors for electing surgery of liver metastases with a high rate of resectability; intraoperative echography is useful for the detection of 10% more liver metastases [18,19]. Until recently, evaluation of the liver focused on the location and number of lesions and the length of disease-free intervals. Classical contraindications for resection included: bilobar disease, more than four lesions, large tumours (over 5 or 10 cm), impossibility of obtaining a resection margin of at least one cm, and extrahepatic disease. Even though the prognostic impact of these factors is acknowledged, none are considered absolute contraindications for surgery in current practice. Evaluation of the surgical candidate must now determine if a complete and safe resection can be performed. Complete (R0) resection is defined as elimination of all metastatic lesions with free margins of resection. If liver function is normal, and at least 25-30% of the liver will
Figure 1. Intraoperative ultrasound in three different cases showing close relation of metastases with vascular structures.
Surgical treatment for liver metastases from colorectal cancer
163
remain, the resection can be done safely. The San Francisco Consensus Conference [20] represented an important shift in the focus of the evaluation before liver resection. Instead of being defined by what is removed, resectability should now be determined by what will remain. This shift will result in an increasing number of patients being eligible for surgery. In this paradigm, resectability is defined by four main criteria: 1. Complete resection. An R0 resection of both the intra- and extrahepatic disease sites must be feasible. 2. Preservation of at least two adjacent liver segments. 3. Preservation of vascular inflow and outflow as well as biliary drainage to the remaining segments. 4. Retention of an adequate volume of liver (i.e., the future liver remnant) after resection (which usually means at least 20% of the total estimated volume of the liver for normal parenchyma, 30-60% if the liver is injured by chemotherapy, steatosis, or hepatitis, or 40-70% in the presence of cirrhosis, depending on the degree of underlying hepatic dysfunction). These new criteria of resectability (Fig. 2) depend less on dogmatic parameters such as tumour number, size, or location than on clinical judgment. Resectability should thus be assessed by an experienced hepatobiliary surgeon
Figure 2. Defining resectability in surgical treatment of colorectal liver metastases.
164
Miguel テ]gel Suテ。rez Muテアoz et al.
in difficult or borderline cases [21]. Operative mortality within two months should be under 5%, and postoperative morbidity around 30% [22]. Most patients with CLM will present with unresectable disease at the time of diagnosis. The current definition of unresectability includes: 70% of the liver or more than six segments affected (Fig. 3). Other causes of unresectability are invasion of both portal branches or the three hepatic veins. Those patients who present with unresectable disease can be rescued by neoadjuvant chemotherapy. The prognosis of unresectable CLM might be improved if a radical surgical resection (rescue surgery) of CLM could be performed after a response to chemotherapy.
Figure 3. Bilobar unresectable colorectal metastatic disease.
3. Increasing resectability a. Portal vein embolisation Portal vein embolisation (PVE) is used to increase the volume and function of the liver that will remain after extensive liver resection. Operative outcomes are improved in properly selected patients who undergo PVE and experience adequate future liver remnant (FLR) hypertrophy. Absolute volume and volume change of the FLR after PVE (interpreted in context of liver disease) predict adequate liver function postresection [23]. The concept emerged from the recognition that tumour invasion of the portal vein causes contralateral lobar hypertrophy and ipsilateral atrophy. A variety of conditions have been shown to inhibit regeneration. These include biliary obstruction,
Surgical treatment for liver metastases from colorectal cancer
165
diabetes, chronic alcohol consumption, malnutrition, gender, aging, and infection [2]. In general, 20% of the total volume of the liver appears to be the minimum safe volume that can be left following extended resection in patients with a normal underlying liver. The cutoff threshold for the FLR, however, may need to be greater in those patients with underlying liver injury due to steatosis or steatohepatitis (30% FLR) or cirrhosis (40% FLR). Computed tomography (CT) or magnetic resonance imaging can now provide an accurate, reproducible, preoperative measure of FLR volume. The right liver (right lobe) accounts for about two thirds of the total volume of the liver, and the left liver (left lobe) accounts for about one third. Patients with multiple hepatic colorectal metastases are often treated by resection of the right hemiliver and segment IV (extended right hepatectomy or right trisectionectomy). On average, this procedure removes 84% of the total volume of the liver in the absence of compensatory hypertrophy resulting from tumour growth. Lobar and segmental intrahepatic volumetric distribution, however, varies considerably. To avoid operating on patients with a low FLR volume, any patient who fails to show compensatory hypertrophy as a result of tumour growth and who has an FLR below 20% (or 30% in patients with a very fatty liver) should be considered for PVE to induce hypertrophy of the contralateral liver lobe. PVE involves cannulating the left or right portal vein (Fig. 4) under fluoroscopic guidance and embolising the appropriate vessel(s) using embolic material (e.g., coils, thrombin, polyvinyl alcohol, cyanoacrylate, or tris-acryl gelatin microspheres) (Fig. 5).
Figure 4. First step in PVE: cannulation of right portal vein.
166
Miguel テ]gel Suテ。rez Muテアoz et al.
Figure 5. Portal occlusion with coils and cyanoacrylate.
PVE is safe, with a complication rate of 5-8%. PVE has been shown to increase the size of the FLR from 8% to 16%, depending on the extent of the underlying liver disease. Patients with normal liver function who underwent a right hepatectomy did not benefit from PVE. In general, PVE should be performed only in patients with normal liver function who are being considered for an extended right hepatic resection; PVE is rarely necessary prior to extended left hepatectomy because the right posterior sector typically constitutes about 30% of the total volume of the liver. The selective use of PVE may enable safe and potentially curative hepatic resection in a subset of patients with advanced colorectal metastases who would otherwise have been marginal candidates for resection because of an inadequate FLR or significant underlying liver disease. PVE is most useful as part of a multimodal approach that includes preoperative chemotherapy and surgery. Likewise, preoperative PVE may reduce the rate of intrahepatic recurrence after liver resection for unilobar CLM [24]. Some authors advise against performing major resection in patients in whom no hypertrophy of the FLR is observed after PVE, because the lack of hypertrophy is predictive of poor outcome. Chemotherapy in conjunction with PVE does not appear to be detrimental to liver hypertrophy with the exception of the use of bevacizumab, as recently reported by Aussilhou et al. [25].
b. Two-stage hepatectomy In patients with multiple CLM in both sides of the liver, a two-stage hepatectomy may be the best therapeutic approach. Specifically, a two-stage
Surgical treatment for liver metastases from colorectal cancer
167
approach may be the only potentially curative therapy for patients with extensive bilateral CLM that cannot be resected with or without ablation in a single procedure while sparing an adequate FLR [26]. Initial experience with the two-stage procedure without PVE was associated with a relatively high incidence of liver failure, resulting from insufficient functional volume of the FLR, and high mortality (9-15%) [27]. More recently, two-stage hepatectomy combined with PVE was reported with no operative mortality and acceptable morbidity. The procedure is performed with an overall curative intent. The initial stage of the hepatic resection is intended to remove the highest possible number of metastases. The remnant liver hypertrophies, and systemic chemotherapy limits the growth and spread of the remaining tumour deposits. The second hepatectomy is performed if it is potentially curative, in the absence of significant tumour progression, and when adequate parenchymal hypertrophy has reduced the risk of postoperative liver failure. Metastases in the FLR (usually the left lateral sector) are usually resected in the first stage due to the concern that nodules in the FLR after PVE may progress more rapidly than those in the nontumoural remnant hepatic parenchyma. PVE is then performed, if indicated, and the liver is allowed to hypertrophy for 3-4 weeks. Another advantage of performing a limited resection in the first stage is the preservation of a maximal amount of liver parenchyma that will hypertrophy after PVE to become the FLR [28]. Similarly, as an alternative to PVE, portal-vein ligation can sometimes be performed [29]. Transection instead of ligation of the branches of the right portal vein during the first-stage results in the technically easier complete dissection of the hepatoduodenal ligament during second-stage hepatectomy [30].
c. Combined local therapy: Resection plus radiofrequency ablation Different techniques for local ablation have been developed in recent years. Combining local ablation to hepatic resection has increased resectability. Thermal techniques using radiofrequency ablation (RFA) are the most commonly used methods [31]. Other thermal ablative techniques include cryotherapy [32], laser interstitial thermotherapy, microwave coagulation therapy, hot saline injection, and high-intensity focused ultrasound [33]. In RFA, an alternating current (460 kHz) is applied via electrode(s) implanted into the centre of the metastases under ultrasonic guidance. The current raises the temperature within the cells, killing the cells and forming a zone of coagulative necrosis around the electrode. RFA can be applied percutaneously, laparoscopically (Fig. 6), or during open surgery. RFA is especially useful for lesions smaller than five cm (preferably smaller than
168
Miguel テ]gel Suテ。rez Muテアoz et al.
Figure 6. Laparoscopic radiofrequency ablation.
three cm) and located at a certain distance from important vessels. The dissipation of heat produced by the blood flow in an adjacent vessel is a serious limitation of RFA that greatly diminishes its efficacy. Intraoperative use of RFA is an important tool in achieving an R0 resection in cases otherwise considered unresectable. RFA thus increases the number of patients who can be surgically treated. In cases of recurrences, some doctors prefer percutaneous RFA to repeat hepatectomy when feasible and safe because it is less invasive, using repeat hepatectomy only when RFA is contraindicated or fails.
d. Extreme liver surgery Advanced liver metastases occasionally invade major blood vessels such as the inferior vena cava (IVC), major hepatic veins, or the hepatic venous confluence. Complete removal of such tumours requires patients to undergo vascular resection and reconstruction. In the past, involvement of the IVC has been considered a contraindication to resection of advanced hepatic tumours, because surgical risks were high and long-term prognosis was poor. Liver resection combined with IVC resection and reconstruction has been reported to be a feasible procedure with improved long-term outcome that can be performed with acceptable operative risk. In some patients, reimplantation of a main hepatic vein may be necessary in addition to IVC replacement (with a ringed Gore-Tex tube graft) or reconstruction (primary or caval plasty).
Surgical treatment for liver metastases from colorectal cancer
169
At the Paul-Brousse hospital, nine of 22 patients in whom IVC was replaced presented metastatic colorectal disease. Five survived to follow-up (range 11-60 months), one died on the seventh postoperative day due to septic shock, and the remaining three died of recurrence of the disease (range 9-11 months) [34]. In the study by Tanaka et al. [35], twenty-five patients who underwent resection and reconstruction of the retrohepatic vena cava, major hepatic veins, or the hepatic venous confluence during hepatectomies showed 16% morbidity without operative mortality. Their one-, three-, and five-year rates of OS were 87%, 58.6%, and 24.9%, respectively, and disease-free rates at these time points were 21.9%, 8.8%, and 8.8%, respectively. In the opinion of these authors, hepatic resection, including major-vessel resection/reconstruction for CLM, can be performed with acceptable operative risk. Such aggressive approaches, however, are beneficial mainly in patients responding to effective prehepatectomy chemotherapy.
e. Downstaging unresectable patients In the past, 5-fluorouracil boluses and leucovorin had achieved response rates of only about 20%, but over the last decade, additional chemotherapeutic and biological agents have been developed that have significantly greater activity against CRC. These newer agents have led not only to higher response rates, but also to a notably longer survival time for patients with traditionally nonresectable disease. With combination therapies that include oxaliplatin and irinotecan, response rates over 50% can be achieved. This superior efficacy of chemotherapeutic agents has allowed clinicians to treat a subset of previously unresectable patients so that they can undergo liver surgery following tumour downsizing. Specifically, about 15-20% of patients with initially unresectable disease (defined as either the inability to resect all tumourous tissue while leaving 20% of the total volume of the liver, or the concomitant presence of extrahepatic metastases) have significant tumour downsizing to the point that the metastatic disease can ultimately be considered resectable [21]. By reconsidering the initial unresectability of patients, hepatic resection and longterm survival may be achieved in a subgroup of patients who otherwise would have a poor outcome. Adam et al. [36] reported that rescue surgery for unresectable CLM downsized by chemotherapy resulted in a five-year survival rate of 33%. Other investigators, as well as expert consensus panels, have substantiated these findings, and resection of initially nonresectable liver metastases following systemic chemotherapy has become increasingly more common [37-39]. The rapid expansion in the use of improved combination regimens of chemotherapy with or without biologics to render initially unresectable
170
Miguel テ]gel Suテ。rez Muテアoz et al.
metastases resectable has increased the percentage of patients eligible for potentially curative surgery [40]. The current staging criteria for CRC patients with metastatic disease, however, do not reflect these recent changes nor the fact that the survival of patients with stage IV CRC varies largely. For example, the chance of survival for a patient with a solitary, resectable liver metastasis is better than that for a patient with stage III disease. Poston et al. [41,42] therefore suggest the need of a new staging system that acknowledges both the improvements in surgical techniques for resectable metastases and the impact of modern chemotherapy on rendering initially unresectable CRC liver metastases resectable, while at the same time distinguishing between patients with a chance of cure at presentation and those for whom only palliative treatment is possible. In the setting of unresectable disease, the goal of preoperative chemotherapy is to convert the extent of intrahepatic disease from unresectable to resectable based on the criteria outlined above. Preoperative chemotherapy should generally thus be stopped once the intrahepatic disease has been downsized to the point where hepatic resection is feasible. Not only does treating to maximal effect prior to surgery fail to provide a measurable oncologic benefit, but prolonged courses of preoperative chemotherapy may also have a detrimental effect on the hepatic parenchyma. Chemotherapy that has effectively downsized previously unresectable lesions should not be continued indefinitely. Rather, the patient should be seen by the hepatic surgeon or surgical oncologist regularly during chemotherapy. The surgeon can thereby help to continually assess disease response/progression and reconsider resection if appropriate. An important question arises when considering a patient for surgery following response to preoperative chemotherapy: do all the original sites of disease require resection or ablation in order to derive a survival benefit from local therapy? The answer is generally yes. Benoist et al. [43] examined metastatic sites in which complete radiologic response was achieved. Of the 66 lesions evaluated, the authors reported persistent tumour on pathology in 83% of cases. Based on these data, resection should be performed on all the original sites of disease that were noted on the initial pretherapy scans, not just the residual disease seen on posttherapy imaging. At the time of planning the surgical resection, therefore, the first HCT image rather than images acquired after chemotherapy must be taken into account. Hepatic surgery should be performed as soon as the disease becomes resectable and during the therapeutic window provided by the response to chemotherapy. Otherwise we will be operating a patient during tumour progression, and the results will be poor [8].
Surgical treatment for liver metastases from colorectal cancer
171
4. Challenges for surgical treatment of metastatic hepatic disease a. Chemotherapy: When? The concept is well established and accepted that the only curative option for treating patients with CLM is surgery with chemotherapy. The dilemma we face today is whether chemotherapeutic agents should be given before or after surgery. Definitions must therefore be clarified. For example, neoadjuvant therapy is the administration of preoperative systemic therapy for resectable hepatic metastases. In general, neoadjuvant therapy does not include any treatment after hepatic resection. The administration of chemotherapy before and after hepatic resection is referred to as perioperative. Adjuvant therapy is the administration of systemic therapy after hepatic resection. Lastly, conversion therapy refers to systemic chemotherapy used for patients with unresectable hepatic metastases in an attempt to make the metastases resectable [44]. The last two decades have seen great improvements in the treatment of hepatic colorectal metastases. The introduction of systemic treatments such as oxaliplatin, irinotecan, bevacizumab, and cetuximab have increased favourable outcomes by increasing the number of patients eligible for resection, as well as by decreasing recurrence. In addition, one of the most important recent discoveries in the field was that K-ras mutations in tumours can be used as reliable predictors of response to cetuximab and oxaliplatin. Fifteen to 20% of patients with a previously unresectable disease now have a disease that is routinely downstaged by effective chemotherapy to a resectable state. On the basis of these data, many oncologists have extended the preoperative use of chemotherapy to include not only those patients with unresectable disease (downstaging strategy), but also those patients with initially resectable disease (neoadjuvant strategy). Many theoretical justifications have been provided for the routine use of neoadjuvant chemotherapy (Table 1). Moreover, increasing evidence suggests that preoperative chemotherapy may have substantial detrimental effects. The syndrome of chemotherapyassociated steatohepatitis is now well recognised. This syndrome is characterised by steatosis, splenomegaly, and thrombocytopenia - results of liver damage and portal hypertension. The clinical implications of such tissue damage on complications and postoperative recovery after liver resection are increasingly being reported. Given that no trial has yet clearly proven a role for preoperative chemotherapy, such neoadjuvant use of chemotherapy should not be regarded as the standard of care [45]. What should surgeons do now? If the patient has synchronous primary tumour and metastatic disease that can be safely removed in the same
172
Miguel テ]gel Suテ。rez Muテアoz et al.
Table 1. Justifications for the routine use of neoadjuvant chemotherapy (modified from [45]).
Surgical treatment for liver metastases from colorectal cancer
173
operation, a combined resection is justified. If the primary CRC has been removed, a delay in resection of synchronous secondaries may be justified by the need to recover from the primary resection. Such use of chemotherapy before subsequent liver resection is particularly attractive for patients with extensive regional nodal metastases or positive margins on the primary tumour resection. For patients with metachronous metastases, preoperative chemotherapy is only currently justified if disease is borderline or unresectable. All others should undergo resection followed by adjuvant chemotherapy, except in the setting of trials.
b. Timing of surgery The optimal timing of the surgical treatment of synchronous CLM is a matter of controversy. Studies that compared combined versus staged procedures demonstrated that the OS is similar, but the complication rate for the combined procedure was significantly lower than for a staged procedure. Combined procedures tended to include more right colectomies and more limited hepatic resections. Staged procedures tended to include more difficult low anterior or abdominal perineal resections or major liver resections. In highly specialised centres, the trend is towards combined procedures whenever possible, because simultaneous liver resection is safe, avoids two operations, avoids the possible hepatotoxicity of neoadjuvant chemotherapy, and allows resection of CLM that may ultimately metastasise further. The general policy is to avoid combined resection in those patients who need a major liver resection [22]. The classic staged surgical approach consists of treatment of the primary tumour followed by treatment of the CLM. In many patients with advanced synchronous CLM, however, the metastases progress during treatment of the primary tumour, precluding curative treatment. Based on this observation, Mentha et al. [11] designed a management strategy that involves initial highimpact chemotherapy, then resection of CLM, and finally removal of the primary tumour in those patients with adverse prognostic factors. In the experience of these authors, this new reverse approach produced resectability and survival rates better than those expected from the published data on patients with disease of similar severity. The obvious candidate for this novel approach would be a patient with a nonobstructive primary colonic tumour. The rationale for the reverse approach is that metastasis is the lesion that kills the patient. Other clinical presentations, though, may dictate the approach to treating synchronous disease: patients who present with obstruction, perforation, and bleeding should receive immediate surgical attention to the
174
Miguel Ángel Suárez Muñoz et al.
primary tumour. Both simultaneous and sequential liver surgery can now be safely performed by laparoscopy [46,47]. Surgery for CLM should not be performed on patients with tumours that progress during neoadjuvant chemotherapy. Liver resection can offer longterm survival to patients with multiple CLMs provided that the metastatic disease can be controlled by chemotherapy before surgery. Tumour progression before surgery is associated with a poor outcome, even after potentially curative hepatectomies. Tumour control before surgery is crucial to offer a chance of prolonged remission in patients with multiple metastases.
c. Surgical morbidity related to chemotherapeutic toxicity The use of the combination of 5-flourouracil and leucovorin is linked to the development of hepatic steatosis, and translates into increased rates of postoperative infection. Therapy with intraarterial floxuridine damages the extrahepatic biliary tree in addition to causing parenchymal liver damage and is associated with increased morbidity after hepatic resection [48]. A form of non-alcoholic steatohepatitis related to chemotherapy and otherwise known as chemotherapy-associated steatohepatitis (CASH) is closely linked to irinotecan-based therapy and is associated with inferior outcomes, including an increase in mortality, following hepatic surgery, mainly due to hepatic insufficiency and poor regeneration [49]. Rubbia-Brandt et al. [50] were the first to report oxaliplatin-associated sinusoidal obstruction syndrome (SOS) in the nontumourous liver specimens of patients undergoing hepatic resection following treatment with oxaliplatin. The histological features of SOS, also known as “blue liver syndrome”, consist of sinusoidal congestion and dilatation, disruption of the sinusoidal membrane, and deposition of collagen within the perisinusoidal space. These patients appear to have an increased risk for intraoperative bleeding and a decreased hepatic reserve. A recent study from Soubrane et al. [51] identified low preoperative platelet count and high APRI (aspartate aminotransferase to platelet ratio index) scores as the most reliable indicators of SOS severity related to oxaliplatin-based chemotherapy. In a study from Hôpital Ambroise Paré, Karoui et al. [52] identified a correlation between postoperative morbidity and the number of cycles of chemotherapy administered before surgery but not to the type of chemotherapy (Fig. 7). The authors concluded that prolonged neoadjuvant systemic chemotherapy alters liver parenchyma and increases morbidity after major resection but does not increase operative mortality. The optimal timing of hepatic resection after completion of chemotherapy varies among institutions, although a consensus is evolving for a minimum
Surgical treatment for liver metastases from colorectal cancer
175
Figure 7. Influence of the number of cycles of chemotherapy on the percentage of postoperative morbidity (modified from Karoui et al. [52]).
interval of four weeks to allow the liver to recover in the hope of reducing morbidity and mortality. In an observational study, Welsh et al. [53] analysed 252 patients who had received chemotherapy and compared them to 245 patients not requiring chemotherapy. Operative time, blood loss, and respiratory and septic complications were greater in the group receiving chemotherapy. Complications were significantly lower in patients undergoing surgery after an interval of 9-12 weeks (2.6%) when compared to surgery performed at 5-8 weeks (5.5%) or four weeks or less (11%). With the advent of other new, more-powerful agents, steatohepatits and sinusoidal injury must be considered in the evaluation process and active efforts made to consider the effects these histological changes may have on patient outcome [54]. Furthermore, there are other concerns regarding the use of antibody therapies, such as bevacizumab, a monoclonal antibody to vascular endothelial growth factor A (VEGF-A). Substantial evidence shows that inhibition of VEGF impairs the healing of wounds, and VEGF is important for liver regeneration. Clinical experience with bevacizumab and cetuximab (an antibody to the epidermal growth-factor receptor) has reflected these concerns as well as â&#x20AC;&#x153;blue liversâ&#x20AC;?, proving an increased prevalence of delayed hepatic regeneration and impaired wound healing. Cetuximab or bevacizumab may be widely used because they seem to improve the efficacy of chemotherapeutic agents but may increase the risk of bleeding. To prevent these related side effects, the surgical procedure must be scheduled 6-8 weeks after the discontinuation of antibody treatment [55,56].
176
Miguel Ángel Suárez Muñoz et al.
d. “Disappearing” liver metastases Despite the evidence favouring hepatectomy when feasible, many medical oncologists prefer to continue systemic chemotherapy to maximal benefit, resulting in a population of CRC patients with no residual nodules on followup imaging modalities, termed “disappearing” liver metastases. The debate rages over whether to resect lesions based upon pretreatment localisation or to observe closely and only resect areas of defined recurrence. Benoist et al. [43] examined whether a complete radiographic response indicated a cure of hepatic CRC metastases. In their series, they identified 38 of 586 patients with CRC liver metastases who had a “disappearance” of at least one metastasis following chemotherapy. These 38 patients had 183 metastases, 66 of which “disappeared” after chemotherapy. At the time of surgery, 20 of 66 (30%) tumours that had “disappeared” on preoperative imaging had macroscopic disease. An additional 12 of 15 “disappearing” tumours that were resected had microscopic disease identified on pathologic examination. Twenty-three of 31 tumours that were labelled as “disappearing” and left in situ reappeared during follow-up. Overall, a true pathologic response was achieved in only 17% of patients. Based on these results, the authors suggested that the site of the lesion that “disappeared” should be resected when feasible. Adam et al. [57] reported a pathologic complete response in 4% of patients following neoadjuvant chemotherapy for CLM. In their series, tumour size under three cm, CEA levels over 30 ng/ml, and age over 60 were associated with a pathologic response, and the five-year survival following hepatectomy for this sub-group of patients was 76%. This pathologic response may be considered a new outcome endpoint after resection of CLM [58]. Elias et al. [59] followed the course of 16 patients with “missing” liver metastases (defined as those tumours that vanish on imaging studies) that remain undetected during hepatectomy and are then left in place. The authors noted that during a mean follow-up of 51 months, metastases recurred at the previous locations in only 38% of patients. In addition, they suggested that preoperative hepatic-artery-infusion chemotherapy correlated with the definitive eradication of missing metastases. In these highly selected patients, the three-year rates of disease-free and OS were 64% and 94%, respectively. Auer et al. [60] reported their experience treating 435 patients with CRC liver metastases. Thirty-nine patients (9%) had a total of 118 disappearing lesions following chemotherapy. Sixty-eight disappearing liver metastases were resected, and 50 were followed clinically. Seventy-five (66%) disappearing liver metastases were true complete responses; 44 pathologic complete responses were found after resection, and 31 durable clinical responses determined after a one-year follow-up. In this
Surgical treatment for liver metastases from colorectal cancer
177
series, a complete response was independently associated with hepatic-arteryinfusion chemotherapy, inability to detect the disappearing metastases with MRI, and normalisation of CEA levels. In summary, complete pathologic disappearance of all liver metastases after chemotherapy is associated with considerably better OS and is a strong predictor of both prolonged survival and disease cure [61]. Four to 13% of patients with liver metastases from CRC will have a complete response following neoadjuvant chemotherapy. Of this sub-group, rates of true complete responses are between 17% and 65%. When patients with a true pathologic complete response undergo resection, rates of five-year survival are as high as 75%. These data would suggest that all lesions should be resected even if they seem to “disappear”.
e. Resection margins Resections of CLM must be approached with “intent to cure”. A marginnegative resection is known to positively impact the long term survival and decrease the rate of local recurrence. Improved outcomes in terms of local recurrence and survival have been associated with a minimum margin of one cm. This led to the adoption of the traditional one-cm margin policy at many centres. However, curative hepatectomies have been reported with margins larger than one cm [62] or more than one mm [63]. More recently, a microscopically positive resection margin (R1 resection) by necessity was not found to impact five-year OS. de Haas et al. [64] believe that positive surgical margins expose a higher risk of recurrence but should not be the only criterion considered. By stating R1 resection as a contraindication to surgery, what should be the alternative? Even with the increasingly efficient regimens of chemotherapy, median survivals currently reach 20-24 months. By combining liver resection (even R1), a quadrupling of the survival expectancy is possible, since median survival in their study was 88 months after R1 resection. Welsh et al. [65] analysed 1005 consecutive liver resections performed in 929 patients and described a simple and accurate model for quantifying the risk of a positive margin following liver resection for colorectal metastases. These data support the published literature, demonstrating a significant disadvantage in survival for patients with an R1 resection. On multivariate analysis, five risk factors independently predicted a positive resection margin: a non-anatomical or extended resection, more than three hepatic metastases involving more than 50% of the liver, repeat hepatic resection, bilobar disease, and an abnormal preoperative liver-function test. These factors not only reflect tumour load, but also the necessity for a non-anatomical surgery and the propensity for underestimating the extent of disease. With all these possible scenarios, achieving an R0 resection is challenging. This model
178
Miguel テ]gel Suテ。rez Muテアoz et al.
offers a unique opportunity to identify patients at high risk of an R1 resection preoperatively who may benefit from neoadjuvant therapy prior to hepatectomy, thus minimising the risk of a positive margin. In an attempt to offer a curative option to some patients with unresectable bilateral CLM, Elias et al. [66] proposed a new approach, called postradiofrequency trans-metastasis hepatectomy. This technique uses radiofrequency energy to first ablate an ill-located liver metastasis along the planned hepatectomy resection line and then to perform the hepatectomy through path of the initially ablated metastases. At the Gustave Roussy Cancer Center, 20 of 21 patients treated by this technique were alive (one died postoperatively) at a median follow-up of 19.4 months.
f. Extrahepatic disease Extrahepatic disease has traditionally been considered a contraindication to resection of hepatic CRC metastasis. However, improvements in rates of morbidity and mortality following hepatectomy, as well as the advent of more-effective systemic chemotherapeutic agents, have prompted several investigators to attempt surgery for some patients. Local extension to adjacent structures, intraluminal biliary tumour thrombus, and locoregional recurrence should not be considered true extrahepatic disease, and none of these should be considered a contraindication to surgery when an R0 resection is feasible. In contrast, there has been significantly more controversy regarding hepatic resection for colorectal metastases in the presence of pulmonary metastases, hilar lymph node metastases, carcinomatosis, and metastasis of other extraabdominal sites [67]. The lungs are the second most frequent site of metastatic disease, accounting for roughly 20-25% of the metastases in patients who have undergone curative resection of their colorectal primary tumour. Approximately 5-10% of patients who present with metastatic disease have a combination of liver and lung metastases, and therefore there has been interest in how this subset of patients should be treated. Several centres have reported five-year rates of survival of more than 30% for patients undergoing combined lung and liver resection. In fact, Shah et al. [68] reported a five-year rate of survival that exceeded 74%. Thus, patients should not be denied a chance at curative hepatectomy simply because of a history of pulmonary metastases, assuming all sites can be resected. A number of factors, however, do appear to be associated with a particularly poor prognosis. Patients with bilateral disease or who have more than six pulmonary metastases were 5070% more at risk for disease-specific death than other patients. Particular attention must therefore be paid when considering these high-risk patients for
Surgical treatment for liver metastases from colorectal cancer
179
combined liver and lung resection [21]. In addition, a recent study from Marudanayagam et al. [69] showed that sequential liver and lung resection for CLM is associated with good long-term survival in selected patients, but not in those presenting with synchronous lung and liver metastases. Unlike pericolic nodal disease, hilar lymph nodes are felt to be â&#x20AC;&#x153;metastases from metastasesâ&#x20AC;? and are associated with a poor outcome. Some investigators have therefore traditionally considered hilar nodal metastasis as a contraindication to hepatic resection of colorectal liver disease. Although early studies reported few, if any, long-term survivors with hilar lymph node metastases, later studies have reported long-term survival in some patients with hilar nodal metastases and have concluded that this patient population may still benefit from hepatic resection. When considering patients with hilar lymph node involvement, the location of the nodal metastases may be important. Two areas of hilar lymph node involvement should be distinguished: the hepatoduodenal-retropancreatic area and the common hepatic artery/celiac-axis region. Whereas patients with positive hilar lymph nodes in the region of the hepatic artery/celiac axis had a very poor prognosis, some patients with hilar lymph node metastases involving the hepatoduodenal or retropancreatic area achieved long-term survival. Specifically, the threeyear rate of survival following hepatectomy in patients with involvement of the hepatic pedicle lymph nodes limited to the hepatoduodenal-retropancreatic area was 38%, compared with 0% for patients with involvement of the common hepatic artery/celiac axis area. Another critical factor may relate to whether the hilar lymph nodes are macroscopically positive. Although patients with microscopic involvement may benefit from hepatic resection, gross involvement of the hilar nodes should be considered a relative contraindication to resection. Currently, the presence of extrahepatic disease is no longer considered an absolute contraindication to hepatic resection. A careful selection of patients must be conducted. The minimum requirements are: disease responsive to chemotherapy and the expectation of an R0 resection.
g. Recurrence of hepatic disease Although five-year survival approaches 55% following resection of colorectal liver metastasis, most patients develop recurrent disease that is often isolated to the liver. Advances in surgical and medical oncology have resulted in prolongation of survival for patients with colorectal liver metastasis, but many patients still develop recurrent disease. Recurrence will appear in 50-60% of patients following hepatic resection of colorectal liver metastasis [70,71]. The disease recurs solely as isolated intra-hepatic disease
180
Miguel Ángel Suárez Muñoz et al.
in a subset of patients. The first site of recurrence following curative intent surgery for colorectal liver metastasis was intra-hepatic only in over 40% of patients [72]. Most hepatobiliary centres have reported that about 10-15% of patients who underwent liver resection for colorectal metastasis eventually underwent a second operation, with rates of mortality of less than 1%. Regardless of the extent of disease, hepatectomy should only be repeated using the same criteria as the initial resection. In particular, surgery should only be undertaken when all disease can be resected with a microscopically negative (R0) margin [73]. A multi-institutional study of 1669 patients by de Jong et al. [74] found an overall five-year survival of 47.1% for the first hepatectomy and 32.6% for the second hepatic resection. Ablation techniques may be helpful tools for the surgeon in these patients with recurrent disease. In conclusion, repeat hepatectomies - even third hepatectomies [75] - can provide rates of long-term survival similar to those of first hepatectomies.
h. Liver metastases in patients with peritoneal carcinomatosis Until recently, the presence of extrahepatic disease was considered an absolute contraindication to hepatectomy for CLM; only a few patients received a combined treatment, and they had a reported five-year rate of survival near 0%. This dogma was also based on the principle stipulating that liver metastases are “confined” and carry a better prognosis than extrahepatic disease, which reflects systemic involvement. Several concepts that invalidate this belief have evolved [76]. Micrometastases and circulating cancer cells in blood, bone marrow, and lymph nodes are now known to almost always present in advanced-stage cancers; the majority of the cells derived from the primary tumour are not stopped by the liver and enter the systemic circulation. Moreover, cancer cells are usually found in the blood spilled into the surgical field. A radical operation is therefore, in some respects, only a cytoreductive operation (a treatment that does not completely eradicate all tumour cells, even if it entirely resects all visible and detectable disease). However, cure can be achieved with the assistance of the patient’s natural immunological defenses and with the latest generation of combination chemotherapy. Surgical reduction of tumour burden may provide an immunological benefit, because tumour cells produce substances that interfere with normal defenses. According to the log-kill hypothesis, each dose of a chemotherapeutic agent kills a constant fraction of cells rather than a specific number of cells. Therefore, by reducing the initial volume of the tumour, one increases the likelihood that chemotherapy will reduce the number of viable tumour cells towards the desired endpoint of zero. A frequently used treatment scheme consisted of
Surgical treatment for liver metastases from colorectal cancer
181
induction chemotherapy, followed by optimal cytoreductive resection of residual tumour and then additional systemic chemotherapy. Patients undergoing complete cytoreductive surgery (CS) (R0) have five-year rates of survival of up to 29%, and CCR is one of the main prognostic factors. Another significant prognostic factor is the number of liver metastases; data suggest that patients should not have more than two or three liver metastases. The incidental intraoperative discovery of extrahepatic disease during abdominal exploration should also restrain the surgeon from proceeding with resection. In conclusion, treating with curative intent selected patients with synchronous presence of CRC, liver metastases, and peritoneal carcinomatosis is possible using an aggressive surgical modality of treatment. For the combined treatment, data suggest the selection of patients with no more than two or three liver lesions and a moderate degree of peritoneal carcinomatosis that can be completely resected (R0) [77]. The surgical treatment of the entire macroscopically visible tumour is important, leaving the treatment of residual microscopic disease to the associated chemotherapy.
i. Liver transplantation for colorectal liver metastases Liver transplantation for CLM is no longer considered due to the poor outcome observed up to the 1990s. According to the European Liver Transplant Registry, one- and five-year rates of patient survival following transplantation for CLM performed prior to 1995 were 62% and 18%, respectively. However, 44% of graft loss or patient deaths were not related to tumour recurrence. Dramatic progress has been made in patient survival after liver transplantation over the last 20 years, thus survival after transplantation for CLM today might be expected to far exceed past outcomes. By using new imaging techniques for proper patient selection, modern chemotherapy, and aggressive multimodal treatment against metastases, long term survivors and even cure could be expected. Preliminary data from a pilot study show a rate of OS of 94% after a median follow up of 25 months [78]. No consensus currently exists concerning the extent and frequency of follow-up after hepatic resection for CLM. Most patients undergo serial serum CEA level tests and CT scans for five years or more following liver resection in an attempt to identify early recurrence that may be amenable to further resection for cure. The results of a recent study by Pulitano et al. [79] confirm that late recurrence (more than 5 years) may occur, and re-resection may still be possible for some of these patients. In the authorsâ&#x20AC;&#x2122; opinion, cure after liver resection of CLM should be defined as 10-year survival after initial hepatectomy.
182
Miguel テ]gel Suテ。rez Muテアoz et al.
Conclusion Liver resection has clearly been established as the standard treatment for resectable colorectal liver metastases. In patients with synchronous CLM, resection of the primary tumor and liver metastases is reccomended to be performed simultaneously whenever this approach is safe and allows a complete resection of the tumoral tissue. Conversion to resectability is possible for about 30% of the patients with initially unresectable CLM, using several therapeutical strategies, giving the chance of a long-term survival to these patients. Effective chemotherapy, as well as increasing data demonstrating effectiveness and safety of combined hepatectomy and ablative therapies, have further expanded the pool of patients now selected for resection. Presence of the extrahepatic colorectal metastases do not represent a contraindication to liver resection, as long as complete removal of the metastases is possible. The prolonged survival of most patients treated by hepatectomy has allowed a long-term analysis of the patterns of recurrence, which emphasize the importance of controlling liver disease for prolongation of life. Liver re-resection is the treatment of choice for patients with recurrent liver metastases, survival rates being similar to those achieved after the first liver resection. In present, the treatment of CLM is multimodal, involving the surgeon, oncologist, and radiologist. Taking into account the great variability of the patients and the multiple therapeutical possibilities, the treatment should be taylored to each patient.
References 1.
Abdel-Misih, S.R., Schmidt, C.R., and Bloomston, P.M. 2009, World J. Surg. Oncol., 7, 72. 2. Mohammad, W.M. and Balaa, F.K. 2009, Clin Colon Rectal Surg, 22, 225. 3. Anonymous. 2009, https://www.livermetsurvey.org. 4. Gomez, D. and Cameron, I.C. 2010, HPB, 12, 227. 5. Nordlinger, B., Guiguet, M., Vaillant, J.C., Balladur, P., Boudjema, K., Bachellier, P., and Jaeck, D. 1996, Cancer, 77, 1254. 6. Konopke, R., Kersting, S., Distler, M., Dietrich, J., Gastmeier, J., Heller, A., Kulisch, E., and Saeger, H.D. 2009, Liver Int., 29, 89. 7. Fong, Y., Fortner, J., Sun, R.L., Brennan, M.F., and Blumgart, L.H. 1999, Ann. Surg., 230, 309. 8. Gonzalez, H.D. and Figueras, J. 2007, HPB, 9, 251. 9. Gregoire, E., Hoti, E., Gorden, D.L., de la Serna, S., Pascal, G., and Azoulay, D. 2010, Eur. J. Surg. Oncol., 36, 568. 10. Chiappa, A., Makuuchi, M., Lygidakis, N.J., Zbar, A.P., Chong, G., Bertani, E., Sitzler, P.J., Biffi, R., Pace, U., Bianchi, P.P., Contino, G., Misitano, P., Orsi, F.,
Surgical treatment for liver metastases from colorectal cancer
11. 12. 13. 14. 15. 16. 17. 18. 19.
20. 21. 22. 23. 24. 25. 26. 27. 28.
29. 30.
31. 32.
183
Travaini, L., Trifiro, G., Zampino, M.G., Fazio, N., Goldhirsch, A., and Andreoni, B. 2009, Crit. Rev. Oncol. Hematol., 72, 65. Mentha, G., Majno, P.E., Andres, A., Rubbia-Brandt, L., Morel, P., and Roth, A.D. 2006, Br. J. Surg., 93, 872. Nagano, Y., Nojiri, K., Matsuo, K., Tanaka, K., Togo, S., Ike, H., and Shimada, H. 2005, J. Am. Coll. Surg., 201, 511. Figueras, J., Ramos, E., Lopez-Ben, S., Torras, J., Albiol, M., Llado, L., Gonzalez, H.D., and Rafecas, A. 2007, Clin. Transl. Oncol., 9, 392. Fernandez, F.G., Drebin, J.A., Linehan, D.C., Dehdashti, F., Siegel, B.A., and Strasberg, S.M. 2004, Ann. Surg., 240, 438. Ramos, E., Martinez, L., Gamez, C., Torras, J., Valls, C., Rafecas, A., Llado, L., Jorba, R., Ruiz, S., Serrano, T., and Fabregat, J. 2008, Cir Esp, 84, 71. Yang, S., Hongjinda, S., Hanna, S.S., Gallinger, S., Wei, A.C., Kiss, A., and Law, C. 2010, HPB, 12, 605. Tanaka, K., Shimada, H., Miura, M., Fujii, Y., Yamaguchi, S., Endo, I., Sekido, H., Togo, S., and Ike, H. 2004, World J. Surg., 28, 263. van Vledder, M.G., Pawlik, T.M., Munireddy, S., Hamper, U., de Jong, M.C., and Choti, M.A. 2010, Ann. Surg. Oncol., 17, 2756. Figueras, J., Planellas, P., Albiol, M., Lopez-Ben, S., Soriano, J., Codina-Barreras, A., Pardina, B., Rodriguez-Hermosa, J.I., Falgueras, L., Ortiz, R., Maroto, A., and Codina-Cazador, A. 2008, Cir Esp, 83, 134. Vauthey, J.N. 2006, Improving resectability in patients with hepatic colorectal metastases, AHPBA 2006 Consensus Conference, San Francisco. Pawlik, T.M., Schulick, R.D., and Choti, M.A. 2008, Oncologist, 13, 51. Figueras, J., Torras, J., Valls, C., Llado, L., Ramos, E., Marti-Rague, J., Serrano, T., and Fabregat, J. 2007, Dis. Colon Rectum, 50, 478. Abdalla, E.K. 2010, J. Surg. Oncol., 102, 960. Oussoultzoglou, E., Bachellier, P., Rosso, E., Scurtu, R., Lucescu, I., Greget, M., and Jaeck, D. 2006, Ann. Surg., 244, 71. Aussilhou, B., Dokmak, S., Faivre, S., Paradis, V., Vilgrain, V., and Belghiti, J. 2009, Ann. Surg. Oncol., 16, 1553. Tsai, S., Marques, H.P., de Jong, M.C., Mira, P., Ribeiro, V., Choti, M.A., Schulick, R.D., Barroso, E., and Pawlik, T.M. 2010, HPB, 12, 262. Jaeck, D., Oussoultzoglou, E., Rosso, E., Greget, M., Weber, J.C., and Bachellier, P. 2004, Ann. Surg., 240, 1037. Wicherts, D.A., Miller, R., de Haas, R.J., Bitsakou, G., Vibert, E., Veilhan, L.A., Azoulay, D., Bismuth, H., Castaing, D., and Adam, R. 2008, Ann. Surg., 248, 994. Capussotti, L., Muratore, A., Baracchi, F., Lelong, B., Ferrero, A., Regge, D., and Delpero, J.R. 2008, Arch. Surg., 143, 978. Homayounfar, K., Liersch, T., Schuetze, G., Niessner, M., Goralczyk, A., Meller, J., Langer, C., Ghadimi, B.M., Becker, H., and Lorf, T. 2009, Int. J. Colorectal Dis., 24, 409. Joosten, J. and Ruers, T. 2007, Crit. Rev. Oncol. Hematol., 62, 153. Yan, T.D., Padang, R., and Morris, D.L. 2006, J. Am. Coll. Surg., 202, 100.
184
Miguel テ]gel Suテ。rez Muテアoz et al.
33. Flanders, V.L. and Gervais, D.A. 2010, J. Vasc. Interv. Radiol., 21 (Suppl. 8), S214. 34. Azoulay, D., Andreani, P., Maggi, U., Salloum, C., Perdigao, F., Sebagh, M., Lemoine, A., Adam, R., and Castaing, D. 2006, Ann. Surg., 244, 80. 35. Tanaka, K., Matsuyama, R., Takeda, K., Matsuo, K., Nagano, Y., and Endo, I. 2009, World J Hepatol, 1, 79. 36. Adam, R., Delvart, V., Pascal, G., Valeanu, A., Castaing, D., Azoulay, D., Giacchetti, S., Paule, B., Kunstlinger, F., Ghemard, O., Levi, F., and Bismuth, H. 2004, Ann. Surg., 240, 644. 37. Baize, N., Gerard, B., Bleiberg, H., Caroli-Bosc, F., Berthier, F., Legendre, H., Pector, J.C., and Hendlisz, A. 2006, Gastroenterol. Clin. Biol., 30, 1349. 38. Reddy, S.K., Kattan, M.W., Yu, C., Ceppa, E.P., de la Fuente, S.G., Fong, Y., Clary, B.M., and White, R.R. 2009, HPB, 11, 592. 39. Abdalla, E.K., Adam, R., Bilchik, A.J., Jaeck, D., Vauthey, J.N., and Mahvi, D. 2006, Ann. Surg. Oncol., 13, 1271. 40. Adam, R., Wicherts, D.A., de Haas, R.J., Ciacio, O., Levi, F., Paule, B., Ducreux, M., Azoulay, D., Bismuth, H., and Castaing, D. 2009, J. Clin. Oncol., 27, 1829. 41. Poston, G., Adam, R., and Vauthey, J.N. 2006, J. Clin. Oncol., 24, 2702. 42. Poston, G.J., Figueras, J., Giuliante, F., Nuzzo, G., Sobrero, A.F., Gigot, J.F., Nordlinger, B., Adam, R., Gruenberger, T., Choti, M.A., Bilchik, A.J., Van Cutsem, E.J., Chiang, J.M., and D'Angelica, M.I. 2008, J. Clin. Oncol., 26, 4828. 43. Benoist, S., Brouquet, A., Penna, C., Julie, C., El Hajjam, M., Chagnon, S., Mitry, E., Rougier, P., and Nordlinger, B. 2006, J. Clin. Oncol., 24, 3939. 44. Petrelli, N.J. 2008, J. Clin. Oncol., 26, 4862. 45. Karakousis, G. and Fong, Y. 2009, Ann. Surg. Oncol., 16, 2086. 46. Tranchart, H., Diop, P.S., Lainas, P., Pourcher, G., Catherine, L., Franco, D., and Dagher, I. 2011, HPB, 13, 46. 47. Sanchez Perez, B., Palomo Torrero, D., Suarez Munoz, M.A., Antunez Martos, S., and Santoyo Santoyo, J. 2009, Cir Esp, 85, 262. 48. Khan, A.Z., Morris-Stiff, G., and Makuuchi, M. 2009, J. Hepatobiliary. Pancreat. Surg., 16, 137. 49. Vauthey, J.N., Pawlik, T.M., Ribero, D., Wu, T.T., Zorzi, D., Hoff, P.M., Xiong, H.Q., Eng, C., Lauwers, G.Y., Mino-Kenudson, M., Risio, M., Muratore, A., Capussotti, L., Curley, S.A., and Abdalla, E.K. 2006, J. Clin. Oncol., 24, 2065. 50. Rubbia-Brandt, L., Audard, V., Sartoretti, P., Roth, A.D., Brezault, C., Le Charpentier, M., Dousset, B., Morel, P., Soubrane, O., Chaussade, S., Mentha, G., and Terris, B. 2004, Ann. Oncol., 15, 460. 51. Soubrane, O., Brouquet, A., Zalinski, S., Terris, B., Brezault, C., Mallet, V., Goldwasser, F., and Scatton, O. 2010, Ann. Surg., 251, 454. 52. Karoui, M., Penna, C., Amin-Hashem, M., Mitry, E., Benoist, S., Franc, B., Rougier, P., and Nordlinger, B. 2006, Ann. Surg., 243, 1. 53. Welsh, F.K., Tilney, H.S., Tekkis, P.P., John, T.G., and Rees, M. 2007, Br. J. Cancer, 96, 1037. 54. Gomez-Ramirez, J., Martin-Perez, E., Amat, C.G., Sanz, I.G., Bermejo, E., Rodriguez, A., and Larranaga, E. 2010, Cir Esp, 88, 404.
Surgical treatment for liver metastases from colorectal cancer
185
55. Morris-Stiff, G., Tan, Y.M., and Vauthey, J.N. 2008, Eur. J. Surg. Oncol., 34, 609. 56. Thomay, A.A. and Charpentier, K.P. 2010, J. Surg. Oncol., 102, 1002. 57. Adam, R., Wicherts, D.A., de Haas, R.J., Aloia, T., Levi, F., Paule, B., Guettier, C., Kunstlinger, F., Delvart, V., Azoulay, D., and Castaing, D. 2008, J. Clin. Oncol., 26, 1635. 58. Blazer, D.G., 3rd, Kishi, Y., Maru, D.M., Kopetz, S., Chun, Y.S., Overman, M.J., Fogelman, D., Eng, C., Chang, D.Z., Wang, H., Zorzi, D., Ribero, D., Ellis, L.M., Glover, K.Y., Wolff, R.A., Curley, S.A., Abdalla, E.K., and Vauthey, J.N. 2008, J. Clin. Oncol., 26, 5344. 59. Elias, D., Goere, D., Boige, V., Kohneh-Sharhi, N., Malka, D., Tomasic, G., Dromain, C., and Ducreux, M. 2007, Ann. Surg. Oncol., 14, 3188. 60. Auer, R.C., White, R.R., Kemeny, N.E., Schwartz, L.H., Shia, J., Blumgart, L.H., Dematteo, R.P., Fong, Y., Jarnagin, W.R., and D'Angelica, M.I. 2010, Cancer, 116, 1502. 61. Tanaka, K., Takakura, H., Takeda, K., Matsuo, K., Nagano, Y., and Endo, I. 2009, Ann. Surg., 250, 935. 62. Figueras, J., Burdio, F., Ramos, E., Torras, J., Llado, L., Lopez-Ben, S., CodinaBarreras, A., and Mojal, S. 2007, Ann. Oncol., 18, 1190. 63. Vandeweyer, D., Neo, E.L., Chen, J.W., Maddern, G.J., Wilson, T.G., and Padbury, R.T. 2009, HPB, 11, 499. 64. de Haas, R.J., Wicherts, D.A., Flores, E., Azoulay, D., Castaing, D., and Adam, R. 2008, Ann. Surg., 248, 626. 65. Welsh, F.K., Tekkis, P.P., O'Rourke, T., John, T.G., and Rees, M. 2008, Surg. Oncol., 17, 3. 66. Elias, D., Manganas, D., Benizri, E., Dufour, F., Menegon, P., El Harroudi, T., and de Baere, T. 2006, Eur. J. Surg. Oncol., 32, 213. 67. Dimitroulis, D., Nikiteas, N., Troupis, T., Patsouras, D., Skandalakis, P., and Kouraklis, G. 2010, World J. Gastroenterol., 16, 3484. 68. Shah, S.A., Haddad, R., Al-Sukhni, W., Kim, R.D., Greig, P.D., Grant, D.R., Taylor, B.R., Langer, B., Gallinger, S., and Wei, A.C. 2006, J. Am. Coll. Surg., 202, 468. 69. Marudanayagam, R., Ramkumar, K., Shanmugam, V., Langman, G., Rajesh, P., Coldham, C., Bramhall, S.R., Mayer, D., Buckels, J., and Mirza, D.F. 2009, HPB, 11, 671. 70. de Jong, M.C., Mayo, S.C., Pulitano, C., Lanella, S., Ribero, D., Strub, J., Hubert, C., Gigot, J.F., Schulick, R.D., Choti, M.A., Aldrighetti, L., Mentha, G., Capussotti, L., and Pawlik, T.M. 2009, J. Gastrointest. Surg., 13, 2141. 71. Petrowsky, H., Gonen, M., Jarnagin, W., Lorenz, M., DeMatteo, R., Heinrich, S., Encke, A., Blumgart, L., and Fong, Y. 2002, Ann. Surg., 235, 863. 72. Antoniou, A., Lovegrove, R.E., Tilney, H.S., Heriot, A.G., John, T.G., Rees, M., Tekkis, P.P., and Welsh, F.K. 2007, Surgery, 141, 9. 73. Pessaux, P., Lermite, E., Brehant, O., Tuech, J.J., Lorimier, G., and Arnaud, J.P. 2006, J. Surg. Oncol., 93, 1. 74. de Jong, M.C., Pulitano, C., Ribero, D., Strub, J., Mentha, G., Schulick, R.D., Choti, M.A., Aldrighetti, L., Capussotti, L., and Pawlik, T.M. 2009, Ann. Surg., 250, 440.
186
Miguel テ]gel Suテ。rez Muテアoz et al.
75. Adam, R., Pascal, G., Azoulay, D., Tanaka, K., Castaing, D., and Bismuth, H. 2003, Ann. Surg., 238, 871. 76. Izzo, F., Piccirillo, M., Palaia, R., Albino, V., Di Giacomo, R., and Mastro, A.A. 2009, J. Surg. Oncol., 100, 345. 77. Elias, D., Benizri, E., Pocard, M., Ducreux, M., Boige, V., and Lasser, P. 2006, Eur. J. Surg. Oncol., 32, 632. 78. Foss, A., Adam, R., and Dueland, S. 2010, Transpl. Int., 23, 679. 79. Pulitano, C., Castillo, F., Aldrighetti, L., Bodingbauer, M., Parks, R.W., Ferla, G., Wigmore, S.J., and Garden, O.J. 2010, HPB, 12, 244.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 187-212 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
8. Lung metastases of colorectal cancer: Focus on surgery José Luis Duque-Medina1, Manuel Castanedo1 and Henar Borrego2 1
Thoracic Surgery Department of Clinical University Hospital of Valladolid, Spain 2 Department of Pathology of Clinical University Hospital of Valladolid, Spain
Abstract. The natural history of colorectal cancer (CRC) progresses from localised to metastatic stages and lung metastases are one of the most important locations. Lung metastasis appear mainly in the parenchyma and to a lower extent and generally later, grow endobronchially. The International Registry of Lung Metastases study (IRLM) sets two reference parameters for metastasis surgery: risk factors and prognosis groups. Nowadays, these factors are clinical, histopathological and immunohistochemical. They may specify the prognosis and surgical indication. Surgical treatment of lung metastases from CRC is not new. Extensive experience has confirmed that it can prolong survival in some patients. The fundamental goal of this surgery is to achieve a complete resection, universally acknowledged as a consistent prognostic factor because it is the only one in which we can truly have an influence on improving survival in these patients. Correspondence/Reprint request: Dr. José Luis Duque-Medina, Thoracic Surgery Department of Clinical University Hospital of Valladolid, Spain. E-mail: jduquemedina@gmail.com
188
JosĂŠ Luis Duque-Medina et al.
However, collaboration between thoracic surgeons and oncologists is needed to offer this surgery only to patients for whom it may improve survival and quality of life. A review of these topic will be performed in this chapter.
Introduction The natural history of CRC progresses from localised stage to metastatic one. In the majority of cases, stage IV disease is beyond the reach of curative surgery; however, in some situations surgery is considered the best treatment option. This is the case for isolated or a small number of lung metastases. CRC metastases to the respiratory system are fairly infrequent. In most cases, metastases enter the lungs through haematogenous dissemination (via the portal and inferior vena cava) and less frequently through lymphatic dissemination. Metastases to the bronchial system can also occur. According to different published series, lung metastases appear in 10% to 30% of CRC patients (1-3), and most are diagnosed 5â&#x20AC;&#x201C;60 months after primary tumour treatment (4, 5). These metastases can appear as the only metastatic lesions (2â&#x20AC;&#x201C;4% of cases) (6), but they can also sometimes appear subsequently to liver metastases. Intrabronchial lesions are rare and can present with dyspnea, cough, fever or repeated infections. In these cases, bronchoscopy is the best method to diagnose the metastasis. By contrast, lung metastases are asymptomatic and incidental radiographic findings. Mediastinal or hilar adenopathies arise from lymphatic spread or a nearby lesion. Their real frequency is unclear because surgical analyses of nodal status are not usually performed (7). Despite this, studies have recorded incidences of mediastinal node metastasis ranging from 3.5% to 16.7% (8, 9). All occurrences of mediastinal node metastases have a negative impact on survival. Surgical treatment of lung metastases is not new. The first metastasectomy was performed in 1885, and Blalock (1) performed the first CRC lung metastasectomy in 1944. After this procedure was first carried out, several studies were conducted on the topic, but patient heterogeneity led to controversial results. In general, surgical treatment for metastatic disease was thought to hold the potential to extend overall survival. In 1991 the International Registry of Lung Metastases (IRLM) was created, and an analysis of long-term metastasectomy results in 5206 patients with lung metastases from different primary tumours treated at European and American Hospitals was published in 1997 (10). This study focused on risk factors of lung cancer metastases and prognostic groups, although the results could not be validated due to sample heterogeneity. Other prognostic factors were subsequently introduced, such as type of lung surgical resection, completeness
Lung metastases of colorectal cancer: Focus on surgery
189
of surgery, tumour histology and nodal involvement. On the other hand, several immunohistochemical factors are also currently being considered to individualise effective treatments. The IRLM study changed the paradigm of surgical treatment in patients with lung metastases. Its results confirmed surgery as an effective treatment in terms of survival. In fact, although chemotherapy has improved the median survival to 2 years, the 5-year survival rate of patients who do not undergo a surgical procedure is less than 5% (11). However, when surgery is not able to completely remove the whole tumour, median survival is 9â&#x20AC;&#x201C;15 months (4, 10). Surgery carries a risk of complications, which can be very significant, particularly if procedures are bilateral, repeated, or combined with systemic or radiotherapeutic treatments. Knowing which patients are likely to benefit most from this decision is critical.
1. Preoperative evaluation Just 1â&#x20AC;&#x201C;8% of patients with CRC lung metastases can undergo metastasectomy. Surgical resection of lung metastases is determined by two essential factors: the surgical risk and the extent of metastatic disease. Evaluation of general health and respiratory function to assess surgical risk The surgical risk evaluation is a complex process. It should be individualised and must include the patientâ&#x20AC;&#x2122;s performance status, comorbidity, cardiopulmonary function and the extent of lung resection (20,21). Although most planned resections are minor, in many cases the resection may be more extensive than anticipated, or the patient may require multiple resections. These two factors should be kept in mind during evaluation to predict the respiratory functional status after the surgical procedure. A number of published series have shown that concomitant diseases such as arterial hypertension (diastolic pressure > 110 mmHg), diabetes mellitus, hypoalbuminemia (serum albumin < 2.5 g/dL), renal failure, weight loss > 10%, body mass index < 18.5 kg/m2 or > 30 kg/m2, ECOG > 2 or Karnofsky index < 50%, EPOC and neoadjuvant treatment have a negative impact on postsurgical mortality. On the other hand, patient age should not be considered a negative prognostic factor (22) and thus, by itself, should not preclude surgical resection. The mortality rate among elderly people treated with surgical resection of lung metastases is more closely related to their comorbidities than to age itself, so a select group of these patients can undergo the procedure after proper support treatment.
190
JosĂŠ Luis Duque-Medina et al.
The most determinant factors to predict surgical risk are cardiovascular and/or respiratory status. The mortality rate secondary to lung metastases resection is directly related to heart or respiratory complications, indicating the need for a very thorough evaluation of these two functions prior to surgery. A known history of cardiopathy significantly increases mortality risk (23), so its evaluation is strongly recommended (24). With this aim, all patients should undergo a thorough medical history and physical examination, electrocardiogram (ECG) and careful evaluation of Revised Cardiac Risk Index (RCRI) (25, 26). RCRI values > 2, or the presence of either cardiopathy requiring medication, suspicion of undiagnosed cardiopathy or trouble breathing when climbing stairs should be more carefully evaluated following the American Heart Association (AHA) and American College of Cardiology (ACC) criteria (25). If detected, unstable coronary syndrome, untreated cardiac congestive failure, significant arrhythmia or severe valve disease should be treated and corrected before planning a surgical resection. Patients with a low cardiac risk or with these problems successfully solved cardiopathy should undergo a routine respiratory function evaluation with spirometric testing and lung diffusion capacity for carbon monoxide (DLCO). Several series have shown that reduced forced respiratory volume at the first second (FEV1) and DLCO have prognostic significance related to postsurgical mortality and other complications. Either FEV1 or DLCO > 80% indicates that a pneumonectomy is possible, in contrast to cases with FEV1 or DLCO < 80%. In the former cases, a cardiopulmonary exercise test (CPET) to measure the maximum oxygen consumption rate (VO2 max) is necessary to understand the cardiopulmonary system reaction to stress and estimate the physiological capacity after surgery. In institutions that lack the capacity to perform a CPET, an exercise test such as climbing stairs or the shuttle walk test, although less preferred, can be used as an alternative. Those unable to climb 22 metres of stairs should be candidates to perform a regular CPET (24). VO2 max < 35% or 10 mL kg-1 min-1 indicates high risk; thus, lobectomy or pneumonectomy are not recommended. On the other hand, VO2 max > 75% or 20 mL kg-1 min-1 would allow a pneumonectomy to be performed. When VO2 max values are between these two limits, postoperative FEV1 (ppoFEV1), DLCO (ppoDLCO) and VO2 max (ppoVO2 max) values must be estimated. Various techniques may be used to evaluate the number of functional segments to be resected, such as radionuclide pulmonary perfusion scanning, anatomic quantification, computed tomography (CT) or bronchoscopy. All result in a similar estimation, although radionuclide scanning is preferred for the estimation of
Lung metastases of colorectal cancer: Focus on surgery
191
postpneumonectomy function because the anatomic quantification methods underestimate these values. The postoperative values can then be calculated by the following equation: ppo values = (preoperative value / T) Ă&#x2014; R, where T is the number of obstructed lung segments and R = T â&#x20AC;&#x201C; the number of segments that will be resected. When ppoFEV1 and ppoDLCO are less than 30% and ppoVO2 max is less than 35% (<10 mL/kg/min), resection should be precluded because the estimated risk is too high. Caution is necessary in borderline cases, because these figures predict the functional status at 3â&#x20AC;&#x201C;6 months after surgery and could be approximately 30% lower in the days immediately following surgery (27). Those patients selected for surgery who cannot undergo cardiopulmonary tests due to comorbidity should be considered high risk, and their treatment and support should be provided in advanced care units. Oncological evaluation: New diagnostic tests It is widely accepted that one of the most important criteria when considering lung metastases surgery is the global extent of the disease. Good control of the primary tumour, no detected extrapulmonary metastases, and the possibility of complete resection are absolutely necessary. Among patients who underwent primary colorectal tumour surgery, 14.4% presented with local recurrence, 3.2% with intraluminal recurrence in the anastomotic point and 1.3% with a metachronous cancer (28). Colonoscopy is considered the best method to diagnose these recurrences (29). However, this technique is unable to detect extraluminal recurrence, leading some authors to prefer the use of virtual colonography (30). Virtual colonography allows the examination of colonic, pericolonic and abdominal organs, with a negative predictive value of 97% to 100% depending on the lesion (31). The use of FDG-PET provides a sensitivity of 93% in detecting local recurrence, compared to 53% using conventional CT (32). However, PET findings may have different interpretations related to intestinal physiological absorption, lesion size, poor anatomical definition and previous treatment, which can induce inflammation leading to a false-positive result (33). Combining these two techniques could improve results by reducing the rate of false-positive and -negative findings. A recent study of FDG-PET/CT reported a significantly higher sensitivity and specificity than multidetector CT (MDCT) in the diagnosis of locoregional recurrence (98.1% and 75% versus 66.7% and 62.5%, respectively). Therefore, PET/CT should be considered as an elective diagnostic method before using invasive techniques.
192
JosĂŠ Luis Duque-Medina et al.
In the search for occult metastases, FDG-PET (Fig. 1) has proven to be a more effective tool than conventional imaging (thoracic and abdominal CT). This finding may change the therapeutic strategy in about one-third of patients (35), decreasing the number of futile surgeries. Due to its discriminatory ability in the context of distorted anatomy, combination PET/CT is particularly useful in the diagnosis of recurrence after completion of hepatectomy, showing a sensitivity of 100% compared to 50% obtained with CT alone (36). However, its effectiveness may decrease in patients who have undergone neoadjuvant chemotherapy (37) and those with mucinous adenocarcinomas (38). The detection of metastases in abdominal lymph nodes can be difficult when they are small. The sensitivity of FDG-PET in these cases hardly reaches 30%, and its interpretation is affected by the activity of neighbouring structures. In these cases the combination of PET and CT improves reliability. A basic requirement in the resection of pulmonary metastases is the possibility of complete resection. Mapping the number and anatomic arrangement of metastases as accurately as possible is therefore essential. In this regard, the identification of millimetric nodules may be underestimated by conventional imaging, leading a significant number of surgeons to prefer final confirmation by palpation in open surgery. The use of helical CT underestimates the presence of metastatic nodules in approximately 20% of pulmonary metastasectomies, with a false positive rate of 29% (39). However, the recent introduction of MDCT technology offers a sensitivity of 97% for non-sarcomatous nodules (40) (Fig. 2), similar to that of combined FDG-PET for nodules smaller than 1 cm (41).
Figure 1. TC-PET imaging: lung metastases from CRC.
Lung metastases of colorectal cancer: Focus on surgery
193
Figure 2. Lung metastases from CRC: micronodules.
2. Prognostic factors Since Thomford and colleagues (42) established the basis for the surgical treatment of pulmonary metastases, various clinical factors have been described whose presence appears to alter patient prognosis or further define the indication for surgical treatment. Multiple determinants of survival have been recognised with varying degrees of evidence, and always with the uncertainty resulting from the absence of comprehensive, prospective and homogenous studies. In a large study, Kanemitsu and colleagues (43) identified five predictors of poor prognosis: prethoracotomy CEA level, number of lung metastases, neoplastic involvement of hilar and mediastinal lymph nodes, primary tumour histology and presence of extrathoracic disease Factors such as disease-free interval (DFI), the multilateral and/or synchronised nature of injuries, lung tumour burden, the previous stage of primary tumour, liver metastases or recurrent metastases may also reduce survival. Lymph node involvement Mediastinal or hilar lymph node involvement is generally considered a poor prognostic factor in the overall context of pulmonary metastases. However, despite being present in between 12% and 19.2% of patients with lung metastases of colorectal carcinoma (44), few studies have addressed their impact. This may be the result of these patients not being considered
194
José Luis Duque-Medina et al.
amenable to surgery, since the disease has overtaken the lung station, and the fact that mediastinal node dissection is not performed routinely in many thoracic surgery centres. Various publications stress that if preoperative evidence indicates mediastinal lymph node metastases, diagnosis should be confirmed prior to considering treatment with metastatic resection. Most series that analysed the impact of lymph node metastases were based on intraoperative findings of lymph node involvement, either through "sampling" or by formal lymph node dissection. These studies (8, 13, 14, 45) have established a relationship between the presence of hilar or mediastinal lymph node involvement and poor prognosis, with 5-year survival ranging from 0–19%. Carcinoembryonic antigen Elevated CEA level is a consistent factor of poor prognosis in most series (44). Higashiyama and colleagues (46) found a significant association between elevated prethoracotomy CEA levels and extrathoracic metastases, especially in the brain. In one study, 5-year survival was 18.9% (range 0–36%) in patients with CEA values higher than 5–10 ng/mL and 59.3% (range 42.7–86.9%) in patients with normal CEA levels. This prognostic difference was more pronounced in patients with CEA levels greater than twice the reference value (14). Disease-free interval In the study of 5026 cases of lung metastases conducted by IRLM, Pastorino and colleagues (10) found significant prognostic differences according to the disease-free interval (DFI) between primary tumour resection and metastasis. These findings were confirmed by Rena and colleagues (12), who observed 5-year survival rates of 22.6%, 38.6% and 55% in patients with DFI of 0–11 months, 12–35 months and >36 months, respectively. In the same vein, one study (47) found that the synchronised nature of the metastasis was a prognostic factor, while other authors reported the predictive value of DFI in partnership with intrapulmonary tumour load and number of metastases (17). Number, size and distribution of metastases A significant number of published case series involved predominantly single metastases. Although some authors (8, 43, 48) found better survival in cases of solitary metastasis compared to multiple lesions, the real impact of the number of metastases on survival has not yet been established. In a recent
Lung metastases of colorectal cancer: Focus on surgery
195
study, Onaitis and colleagues (49) reported a significant relationship between the existence of three or more metastases and the possibility of recurrence, especially when associated with a DFI less than 1 year. Surgery was not curative in any of these patients, indicating non-surgical treatment. Few studies have investigated the size of metastases as a prognostic factor, probably due to the diversity of samples. In this respect, assessment of total tumour burden may be more appropriate (17). Only two series to date (48, 50) have reported a significant relationship between tumour size and 5-year survival, which may fall below 11% in patients with tumours larger than 3.75 cm. With respect to the distribution of metastases, non-conclusive evidence indicates that bilateral distribution is, by itself, an adverse prognostic factor. Primary cancer stage The initial state of the primary tumour is a factor rarely considered in most studies of pulmonary metastasis resection, and few have reported it to influence survival. However, Melloni and colleagues (51) observed a better prognosis in stage T1-T2 primary colorectal carcinoma compared to T3-T4 tumours, with 5-year survival rates of 63% and 34% respectively. These results are similar to those of Inoue and colleagues (14) using Dukes’ staging, who reported 5-year survival rates of 68.7% for stage A primary tumours compared to 32.8% for stages B–D. The anatomical location of the primary tumour has also been speculated to influence outcome. Several studies have demonstrated a greater tendency for lung metastases to originate from rectal cancers, which are associated with an inferior survival compared to tumours of colonic origin (52, 53). Liver metastases The liver and lung are the most common sites of CRC metastasis. Approximately 5–10% of CRC patients will develop liver and lung metastases (44), of which 6.9–30.8% will be synchronous respect to the primary tumor. In patients with both lung and liver metastases, the overall 5-year survival after the first metastasectomy (liver or lung) was between 74% and 30% (52), with no significant differences between patients with lung only as the first site of metastases compared to those with both lung and liver involvement. The prognosis associated with the synchronous occurrence of both lesions still remains unclear. While some large studies have reported no prognostic significance (54, 55), Kobayashi and colleagues (56) observed a 5-year survival rate of 22% among patients with synchronous metastases compared with 50% in patients with metachronous metastases. The results of other
196
JosĂŠ Luis Duque-Medina et al.
studies (52, 57) appear to be in line with these results, in that they associate the concomitance of lesions or a DFI < 1 year with poorer survival rates. Recurrence Insufficient experience is available to determine the prognostic role of metastatic recurrence in the lung. Some series (8, 58) have shown that in select patient groups, repeated resection of lung metastases can lead to survival rates ranging from 29% to 53.8%. However, further studies are needed to determine what types of patients may benefit from surgical treatment, since early recurrence (DFI < 6 months) (59) and number of metastases may be poor prognostic factors (58).
3. Pathological prognostic factors The formation of lung metastases is a multifactorial process. Factors related to the original tumour and other genetic factors, such as the activation of genes that promote metastasis or inactivation of genes that inhibit it, are not yet well understood. Metastases reach the lungs through the blood and also via the lymphatic system, and once confirmed, the metastatic nature of a lung tumour nodule can be defined by a number of these factors. The clinical behaviour of CRC pulmonary metastases is unpredictable. Although adverse prognostic factors have been identified, they have not been conclusive (51). Exhaustive studies are necessary, with larger groups and selected patient series. The molecular profile of aggressive CRC with high metastatic potential includes the expression and overexpression of proteins due to alteration of the p53, K-ras, DCC and nm23 genes, among others (60,61,62). The elevation of proteolytic enzymes (metalloproteinases 2 and 9 and cathepsin B), adhesion molecules, and a high rate of angiogenic factor expression can lead to increased vascular endothelial growth and a high density of microvessels. All these factors and others are promising sources for new antiangiogenic and targeted therapies. New prognostic factors determined using immunohistochemistry or molecular biology, which may also help in the selection of patients more or less likely to be susceptible to these new therapies, are currently in the early stages of investigation and remain still controversial. Histopathology Histopathological factors that may be related to tumour behaviour, and therefore to prognosis, include aerogenous spread with floating cancer cell
Lung metastases of colorectal cancer: Focus on surgery
197
clusters (ASFC) to alveoli in areas near the site of infiltration, vascular invasion in the tumour (63), lymphatic invasion and pleural invasion. These are all poor prognostic factors regarding tumour behaviour, particularly the association of the first two, which some studies have identified as independent prognostic factors (19, 60). Immunohistochemical No definitive prognostic immunohistochemical marker has been identified to date, but many can be informative. E-cadherin is a glycoprotein with a significant role in modulating the invasiveness of tumour cells (Fig. 3).
Figure 3. Expression of the E-cadherina.
Its low expression in tumour cells is associated with increased tumour invasion and worse behaviour, also expressed by primary tumours of the colon (60). Although not supported by all authors, E-cadherin expression may help to select patients for chemotherapy. -
The transcription factor CDX-2 is expressed in normal colonic epithelium and in the majority of adenocarcinomas of the colon. CDX-2 can potentially be used for diagnosis but has no prognostic value (64,65). This antibody is also specific for the identification of neuroendocrine tumours of intestinal origin and/or diagnosis of their metastases.
198
JosĂŠ Luis Duque-Medina et al.
-
-
-
-
Cathepsin B belongs to a family of lysosomal enzymes and is involved in the regulation of certain growth factors in malignancies including breast cancer and CRC. Cathepsin B could be an acceptable tumour marker, as it can also be quantified in the blood and may be valuable, especially in cases of remote metastases (66). The insulin-like growth factor 1 receptor (IGF1_R), B catenin, viral oncogene homolog B FBJ murine osteosarcoma (FOS-B), melanoma antigen (MAGE) and other proteins have not demonstrated statistically significant prognostic value (60). Vascular endothelial growth factor (VEGF) is a glycoprotein related to lymphangiogenesis (and thus metastasis via the lymphatic system). Additionally, higher VEGF levels in tumour tissue have been associated with poorer prognosis. This factor identified by immunohistochemistry may be an independent prognostic marker of tumour behaviour and used to indicate poor prognosis (67,68), although other studies have not confirmed these results (69). Epidermal growth factor receptor (EGFR) overexpression and/or amplification, as determined by immunohistochemistry and molecular biology, respectively, correlates with advanced tumour stages and metastasis development.
Genetic factors Many genes are involved in the molecular profile of CRC, either via the activation of oncogenes involved in the process of neoplastic proliferation (MYC, RAS) or inactivation of oncogene suppressors (APC, P53, DCC, nm23) (60, 61). Among tumour suppressors, p53 (located on chromosome 17p) plays an essential role in tumour angiogenesis. The p53 gene (Fig. 4) encodes a protein that acts as a tumour suppressor. Its mutation occurs in 70â&#x20AC;&#x201C;80% of CRC. Overexpression of p53 protein, detected immunohistochemically, is used as a marker for p53 mutations and associated with a poorer prognosis in terms of lower DFI, but has not been shown to affect survival (60). The nm23 gene is regarded as a suppressor of the metastasis process. The expression of two proteins, nm23-H1 and nm23-H2, and its significance in CRC is controversial (61). While some authors found that reduced expression of nm23 was associated with advanced stages of disease and metastasis, others found that nm23 overexpression was associated with recurrence, liver metastasis and decreased survival. This apparent contradiction was explained by Berney and colleagues (61), who suggested that nm23 protein overexpression was due to a mutant nm23 protein resulting from deletion of
Lung metastases of colorectal cancer: Focus on surgery
199
Figure 4. Overexpression of the p53.
the nm23 gene. However, other recent studies have failed to relate the expression of nm23 with staging and/or prognosis. Mutation or loss of the DCC (deleted in colorectal cancer) gene, which is related to adhesion molecules, is also involved in the progression of adenoma to carcinoma in primary tumour development and associated with more advanced tumour stage and poorer prognosis. Mechanisms related to the C4.4A gene may also be associated with tumour invasion (70). K-ras is currently the most important CRC biomarker. K-ras belongs to a family of oncogenes that encode a protein with a key role in EGFR signalling. K-ras mutations are found in 30% to 40% of CRC patients. They are highly correlated with both primary tumours and metastatic disease, and observed more frequently in metastasis. Patients with K-ras-mutated tumours have a lower DFI after lumpectomy, suggesting a worse prognosis. Patients with the mutated gene also do not respond to targeted anti-EGFR drug treatment (62).
4. Surgical treatment Surgery for lung metastases from colorectal carcinoma is universally accepted as one of the main therapeutic strategies in these patients, although, to date, no studies have assessed the precise benefit of this treatment. However, the surgical approach to these lesions carries a low morbidity and virtually nonexistent mortality and also occasionally leads to the discovery of a second
200
JosĂŠ Luis Duque-Medina et al.
primary tumour, with the possibility of providing correct oncological treatment thereof. The fundamental goal of this surgery is to achieve a complete resection. Complete resection is universally acknowledged as a consistent prognostic factor, and it is the only one in which we can truly have an influence on improving survival in these patients. When resection is complete, surgical treatment of patients with metastases has reported survival rates ranging from 30% to 63% (2, 12, 61) at 5 years and 18% at 10 years. By contrast, median survival after incomplete surgery is only 9 to 15 months (3, 4, 7), emphasising the need for careful collaboration between thoracic surgeons and oncologists to offer this surgery only to patients for whom it may improve survival and quality of life. Resection criteria indications Excluding the occasional diagnostic surgery, surgery for metastases should be considered a therapy for local control of extensive disease. Indications and surgical techniques have been modified with respect to the improved diagnoses provided by imaging techniques and the precise use of cancer treatments in the context of multimodal therapy, indirectly leading to a greater number of candidates for resection (48, 71) as well as a gradual improvement in survival. In general, surgical treatment to remove a solitary metastasis, especially if it is peripheral, is usually considered favourable. Indications for the surgical removal of bilateral multiple metastases and repeat metastasectomy of lung recurrences are more controversial. In this regard, the National Comprehensive Cancer Network (NCCN) has released NCCN Colon Cancer Guidelines Version 1.2011, which states the criteria for resectability of metastases and locoregional therapies within surgery (Table 1). Patients who do not meet these criteria may have to try alternative treatments for local disease control such as cryotherapy, stereotactic radiotherapy or radiofrequency ablation. Radiofrequency ablation, indicated firstly for metastases of less than 5 cm and not hilar, is an effective treatment even at the expense of non-negligible morbidity (mainly pneumothorax) and high recurrence requiring repeated ablation (72). The existence of affected lymph nodes is another point of controversy. Most authors (8, 44, 73, 74) consider lymph node involvement to negatively impact survival, regarding it as one of the most important prognostic factors. One study reported a 5-year survival of 0% for patients with mediastinal lymph node involvement (8), leading these authors to consider it a contraindication for extensive excision.
Table 1. NCCN Guidelines TM Version 1.2011 Colon Cancer. Principles of surgery: Criteria for respectability metastases and locoregional therapies within surgery.
Lung metastases of colorectal cancer: Focus on surgery 201
202
JosĂŠ Luis Duque-Medina et al.
The optimal timing of resection is also an important consideration, and the best time to resect pulmonary metastases has been a source of some controversy. Most authors state that metastasectomy should be performed when metastasis is diagnosed, if patients meet requirements. However, in a retrospective study, Tanaka and colleagues (75) observed that after potentially curative metastasectomies, new lesions appeared in 50% of patients due to further growth of previously undetected lesions, and found increased survival in patients in whom surgery was delayed more than 3 months. It is important to remember that smaller lesions have a shorter doubling time and therefore a shorter growing time. However, other tumour characteristics, mainly the number, size and aggressiveness of metastases and the type of surgical approach, must also be considered before deciding upon surgery. A metastasis that increases rapidly in size can produce both parenchymal spread and adenopathy, but can reduce the chances of a lesser resection. Waiting 3 months to conduct the metastasectomy may be indicated for no peripheral lesions smaller than 1 cm because this allows us to assess the presence of metastases initially hidden or absent, as well as their best display in radiological images and subsequent palpation. On the other hand, from a purely technical standpoint, the resection of a solitary peripheral metastasis using video-assisted thoracoscopic surgery (VATS) has minimal risk and generates few adhesions that interfere with subsequent resection in case of recurrence. By contrast, the need for open surgery due to central and/or multiple nodes poses a greater risk of adhesions, morbidity and increased recovery time. In these cases, the decision to wait may be particularly favourable. Another issue surrounding surgical indications is with the presence of liver metastases. Although some reports (77) consider these a negative prognostic factor of survival and recurrence, their synchronous or metachronous appearance should not be a contraindication. Several studies indicate that 5-year survival is around 30% and can reach 74% (13, 16, 52, 54, 56, 58). These results are comparable to surgery for lung metastases only, confirming the benefits of resection treatment in these patients (76). Recent studies (56, 78) have indicated better survival in metachronous metastases compared to synchronous metastases. The strategy for surgical intervention is controversial: sequential, simultaneous, or dependent on the severity of their location. Most groups favour a sequential process dealing first with liver metastases (1). Approaches The surgical approach must meet two requirements: it should be the less aggressive as possible and allow all metastases to be resected. The information provided by imaging is becoming more detailed and accurate, but
203
Lung metastases of colorectal cancer: Focus on surgery
recent studies have emphasised that these methods still demonstrate low sensitivity to millimetric injury (39). Thus, another requirement in this type of surgery is manual palpation. Several factors determine the type of approach in this type of surgery, including the number, size and anatomical location of metastatic lesions and positive lymph nodes, as well as the functional status of the patient and the preferences and personal experience of the different surgical teams. Regardless of preference per se, each approach has its advantages and disadvantages (Table 2). Table 2. Characteristic surgical approaches. Surgical approaches
Visual metastases
Lung palpation
Standard thoracotomy Lateral thoracotomy Sternotomy Clamshell Thoracoscopy
Excellent Excellent Excellent Excellent No
VATS
Excellent Excellent Optimal (except posterior lesions) Excellent Medium-Optimal (IF peripherals) Optimal
Surgical approaches Standard thoracotomy Lateral thoracotomy Sternotomy Clamshell Thoracoscopy VATS
Pain Severe Moderate Moderate Intense Mild Mild-moderate
Morbidity CRM + + CRM + CRM + CRM + + CRM CRM - +
Medium
SM SM SM SM SM SM
+ -+ -+ + + +
CRM: Cardio-Respiratory Morbidity SM: Surgical Morbidity VATS: Video-assisted Thoracoscopic Surgery
Classical posterolateral thoracotomy, previously used widely, is gradually falling into disuse because it is very aggressive and painful. Lateral thoracotomy should be the approach of choice for unilateral lesions. It allows a correct visualisation of parenchymal lesions and proper mediastinal lymph node evaluation. Axillary thoracotomy also has limited applications in specific metastatic sites. If metastatic disease is present in both hemithorax (Fig. 5), a choice must be made between sequential treatment with bilateral thoracotomy and simultaneous treatment, often with a sternotomy.
204
JosĂŠ Luis Duque-Medina et al.
Figure 5. Bilateral lung metastases and mediastinal nodal involvement.
Sternotomy, which had its peak in the 1980s, allows the simultaneous exposition of both hemithoraces and is of course indicated for bilateral lesions. Its postoperative management is less complex overall than thoracotomy, and it is thought to be a less painful approach and functionally less aggressive. Its limitations are determined by the difficulty in addressing posterior lesions, especially those in the lower left lobe. A variant of sternotomy is the transverse "clamshell" sternotomy used in lung transplants and also occasionally considered for addressing bilateral lesions. This technique allows proper access to both hemithoraxes but is more painful than sternotomy and requires sectioning of the mammary arteries. Thoracoscopy is also a reasonable surgical approach and its outcome, in terms of survival, is similar to open surgery (44, 57, 80). It provides all of the a priori advantages of minimally invasive surgery: less pain (79), less impairment of lung function (80) and less surgical trauma, with decreased morbidity, reduced hospital stay and improved aesthetics. Thoracoscopy also reduces the possibility of pleuropulmonary adhesions, which is of importance in the event that iterative surgery is indicated for multiple metastases or diagnostic purposes. At the therapeutic level, imaging with PET and high resolution CT was previously required to determine the exact number and location of metastases. Most authors consider thoracoscopy mainly for metastases that are single or double, non-recurring, peripheral and less than 3 cm. For lesions smaller than 1 cm, peripheral, but more than 5 mm of the visceral pleura, the use of CTguided microcoils as â&#x20AC;&#x153;pointersâ&#x20AC;? is indicated; the added complication rate is very low. Intraoperative ultrasound can also help localise initially undetectable lung nodules. However, it is clear that manual palpation of lung nodules provides
Lung metastases of colorectal cancer: Focus on surgery
205
greater security. In this respect, VATS with an additional small minithoracotomy allows the advantages of VATS and open surgery to be combined, and is in principle the approach of choice in most peripheral metastases. However, there are supporters of the VATS approach for the treatment of virtually all pulmonary metastases (81). Additionally, there is always the option, in case of doubt, for conversion to open surgery. Type of removal The purpose of this type of surgery is to achieve a complete (R0) resection while preserving the lung parenchyma as much as possible. Thus, the excision should be tailored to the lung injury, although in terms of survival, minor and major resections do not significantly differ. In this respect, as in all forms of lung surgery, surgery can be divided into three basic categories: minor resections (wedge resection and atypical), anatomical resections (segmentectomy, lobectomy and pneumonectomy), and extended resections. In general, the greatest number of resections, approximately 70% (7, 71), correspond to minor resections, due to the peripheral location of most injuries. On the other hand, the minimum margin of resection has not been defined yet, although most surgeons are in favour of limiting it to 1 cm (Fig. 6). If this limit cannot be assured using an autosuture machine and common endocutter, one must resort to manual suture resection with electrocautery and, more recently, laser cautery (Fig. 7).
Figure 6. Lung metastases: wedge resection.
206
JosĂŠ Luis Duque-Medina et al.
Figure 7. Laser resection of lung metastases.
These latter methods are also indicated for the resection of multiple small peripheral metastases. In general, as they tend to achieve hemostasis and correct aerostasis. In the case of larger or more central metastases, multiple metastases in the same lobe, and whenever there is a doubt about the origin of the lesion (second primary tumour vs metastasis), anatomical resection is required (segmentectomy, lobectomy or pneumonectomy). Anatomical resections may also be used for patients with lymph node tumour (N1) between the vessels and segmental bronchi. Pneumonectomy should be considered exceptionally, only if patients meet all indicated requirements for the surgery. Similarly, extended surgery (such as wall and mediastinum) is indicated only if patients meet these requirements and in the absence of response to other cancer treatments. The role of lymphadenectomy in its various forms is controversial. Most published studies (3, 7) do not involve systematic lymph node dissection, unless there is evidence of enlarged lymph nodes at the time of surgery. However, recent evidence that patients with lymph node involvement have a poorer prognosis (4, 8, 14, 73, 74) may indicate the potential need for systematic selective mediastinal lymph node examination (mediastinoscopy) (82, 84), ideally selectively, for screened patients whose imaging tests (CT, PET) suggest mediastinal lymph node involvement (cN2) associated with metastatic disease. In these situations, surgical resection is not contraindicated, but, particularly in the case of metastatic relapses, other methods such as stereotactic radio frequency radiation may be indicated instead.
Lung metastases of colorectal cancer: Focus on surgery
207
Lymphadenectomy associated with metastasectomy ensures consistent complete resection (7). Some authors noted an improvement in survival, (8,84) and others a decrease in recurrences (7. 83). Although at present no randomised studies have confirmed the benefits of lymph node dissection, these current findings suggest that lymphadenectomy associated with surgery for metastases should be considered. Reoperations Metastatic disease recurrence is common in colorectal carcinomas, perhaps due to the dubious effectiveness of adjuvant chemotherapy (7,83). Reported rates of both resectable and unresectable recurrences range from 46% to 70% in different series (4, 58). Although specific prognostic factors influencing recurrence have not been assessed in detail, according to some studies, recurrences are more frequent in patients younger than 65, with shorter DFI, and with more than three previous metastases (49). The indication for reoperation has been disputed, particularly in patients with confirmed mediastinal lymph node involvement, but the majority of reports favour reoperation after reevaluation and in compliance with requirements (17, 58, 83). However, only a minority of patients meet the criteria indicating a second resection. In this selected group of patients, survival rates ranging from 30% to 53% have been reported (8, 12, 16, 17, 58, 82, 83), with acceptable mortality. These results are similar to those after the first metastasectomy, particularly in cases of isolated relapse in the absence of lymph node involvement (83). Similarly, highly select patients may undergo a third metastasectomy, but with a progressively lower benefit in terms of both disease control and survival. Endobronchial metastases Endobronchial colorectal metastases are uncommon, but may also be underdiagnosed (85, 86). They are more frequent in advanced stages of disease, often coexist with lung metastases (86), and can be generated via the impact of occasional lymphatic tumour emboli on the bronchial submucosa or, more often, due to progression of parenchymal injury. A specific feature of this location is its tendency to cause airway obstruction (Fig. 9). The most common presenting symptoms are dyspnea, cough and hemoptysis, although it may be a radiological finding in asymptomatic patients. Only 55% of endobronchial metastases are detected by CT (86). Bronchoscopy is the ideal method for diagnosis, with sensitivity close to 100% in central lesions. Therefore, bronchoscopy or echobroncoscopy should
208
JosĂŠ Luis Duque-Medina et al.
Figure 8. Endobronchial metastasis, left lower lobe atelectasis.
perhaps be indicated systematically in patients with lung metastases, with a standard procedure for evaluation. Laser photoresection, photodynamic therapy, cryotherapy and brachytherapy are palliative methods that can substantially improve pulmonary symptoms. Cryotherapy, photodynamic therapy and even brachytherapy should be considered in central and localised lesions, photocoagulation and laser vaporisation in bleeding injury, and stenting in the case of coexisting extrinsic compression. As in lung metastases, excision surgery is an option, but only in very selected cases. In general, patientsâ&#x20AC;&#x2122; very advanced disease precludes excision.
Conclusion With the increasing incidence of CRC, the number of patients with distant metastases is also increasing, and for this reason, various treatment methods have been introduced, and treatment protocols have been changed. The lung is the most frequent extraperitoneal metastatic site and these metastases develop in about 10% of the patients who undergo a curative resection for this cancer. Most lung metastases are disseminated at the time of diagnosis and curative resection is not indicated in these cases, but a conservative treatment or chemotherapy, are good alternatives we can use. Although this surgery could be performed for selected cases, getting a 5-year survival rate from 9% to 60%, it would be necessary a close collaboration between thoracic surgeons and oncologists to offer this
Lung metastases of colorectal cancer: Focus on surgery
209
treatment only to patients for whom it may improve survival and quality of life.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
Pop, D., Venissac, N., Leo, F., Karimdjee, B-S., Lopez, S., and Mouroux J. 2004. Ann. Chir., 129, 589. Demmy, T.L., and Dunn, K.B. 2007. Surg. Oncol. Clin. Am., 16, 579. Mongil, R., Pagés, C., Ruiz, J.A., Roca, J., Arrabal, R., Benitez, A., et al. 2009. Arch. Bronconeumol., 45, 235. Milleron, B., Capron, F., Calais, G., Mangiapan, G., Rosencher, L., Wdowick, A., et al. 1996. Encycl. Méd. Chir., Pneumologie, 6-002-G-70. Goldberg, R.M., Tangen, T.M., Moertel, C.G., Macdonald, J.S., Haller, D.G., and Laurie, J.A. 1998. Ann. Intern. Med., 129, 27. Brister, S.J., De Varennes, B., Gordon, P.H., Sheiner, N.M., and Pym, J. 1988. Dis. Colon Rectum, 31, 786. Loeche, F., Kobinger, S., Hatz, R.A., Helmberger, T., Loehrs, U., and Fuerst, H. 2001. Ann. Thorac. Surg., 72, 225. Welter, S., Jacobs, J., Krbek, T., Poetgen, C., and Stamatis, G. 2007. Eur. J. Cardiothorac. Surg., 31, 167. Veronesi, G., Petrella, F., Leo, F., Solli, P, Maissoneuve, P., Galetta, D., et al. 2007. J. Thorac. Cardiovasc. Surg., 133, 967. The International Registry of Lung Metastases, Writing Committee: Pastorino, U., Buyse, M., Friedel G., Ginsberg, R.J., Girard, P., Goldtraw, P., Johnston, M., McCormack, P., Pass, H, and Putnam Jr., J.B. 1997. J. Thorac. Cardiovasc. Surg., 113, 37. Meyerharrdt, J.E., and Mayer, R.J. 2005. N. Engl. J. Med., 352, 476. Rena, O., Casadio, C., Viano, F., Cristofori, R., Ruffini, E., Filosso, P.L., and Maggi, G. 2002. Eur. J. Cardiothorac. Surg., 21, 906. Pfannschmidt, J., Muley, T., Hoffmann, H., Murley, T., Dienemann, H., and Herpel, E. 2003. J. Thorac. Cardiovasc. Surg., 126, 732. Inoue, M., Ohta, M., Iuchi, K., Matsumura, A., Ideguchi, K., Yasutmisu, T., et al. 2004. Ann. Thorac. Surg., 78, 238. Sakamoto, T., Tsubota, N., Iwanaga, K., Yuki, T., Matsuoka, H., Yoshimura, M. 2001. Chest, 119,1069. Saito, Y., Omiya, H., Kohno, K., Kobayashi, T., Itoi, K., Teramachi, M., et al. 2002. J. Thorac. Cardiovasc. Surg., 124, 1007. Zink, S., Kayser, G., Gabius, H.J., and Kaiser, K. 2001. Eur. J. Cardiothorac. Surg., 19, 908. Kandioler, D., Krömer, E., Tüchler, H., End, A., Müller, M.R., Wolner, E., and Eckersberger, F. 1998. Ann. Thorac. Surg., 65, 909. Shiono, S., Ishii, G., Nagai, K., Yoshida,J., Nishimura, M., Murata, Y., et al. 2005. Ann. Thorac. Surg., 79, 278. Duque, J L., Ramos G, Castrodeza, J, Cerezal J, Castanedo M, Yuste G., and Heras, F. 1997. Ann. Thorac. Surg., 63, 944.
210
José Luis Duque-Medina et al.
21. Kozower, B.D., Sheng, S., O'Brien, S.M., Liptay, M.J., Lau, C.L., Jones, D.R., et al. 2010. Ann. Thorac. Surg., 90, 875. 22. Colic, G.L., Shafazand, S., Griffin, J.P., Keenan, R., Bolliger, C.T. 2007. Chest, 132, 161S. 23. Duque J.L., and Bronchogenic Carcinoma Cooperative Group of the Spanish Society of Pneumology and Thoracic Surgery (GCCB-S). 2000. Lung Cancer, 29, 135. 24. Brunelli, A., Charloux, A., Bolliger, C.T., Rocco, G., Sculier, J.P., Varela, G., et al. 2009. Eur. Respir. J., 34, 17. 25. Lee, T.H., Marcantonio, E.R., Mangione, C.M., Thomas, E.J., Polanczyk, C.A., Cook, E.F., et al. 1999. Circulation., 100, 1043. 26. Auerbach, A., and Goldman, L. 2006. Circulation., 113, 1361. 27. Varela, G., Brunelli, A., Rocco, G., Novoa, N., Refai, M., Jiménez, M.F., et al. 2007. Eur. J. Cardiothorac. Surg., 31, 518. 28. Renehan, A.G., Egger, M., Saunders, M.P., and O’Dwyer, S.T. 2002. BMJ, 324, 1. 29. Rex, D.K., Kahi, C.J., Levin, B., Smith, R.A., Bond, J.H., Brooks, D., et al. 2006. CA Cancer J. Clin., 56, 160. 30. Choi, Y.J., Park, S.H., Lee, S.S., Choi, E.K., Yu, C.S., Kim, H.C., and Kim, J.C. 2007. AJR Am. J. Roentgenol., 189, 283. 31. Kim, H.J., Park, S.H., Pickhardt, P.J., Yoon, S.N., Lee, S.S., Yee, J., et al. 2010, Radiology. [Epub ahead of print] 32. Selzner, M., Hany, T.F., Wildbrett, P., McCormack, L., Kadry, Z., and Clavien, P.A. 2004. Ann. Surg., 240, 1027. 33. Smeets, P., Ham, H., Ceelen, W., Boterberg, T., Verstraete, K., and Goethals, I. 2010. Q. J. Nucl. Med. Mol. Imaging, 54, 327. 34. Metser, U., You, J., McSweeney, S., Freeman, M., and Hendler, A. 2010. AJR Am. J. Roentgenol., 194, 766. 35. Wiering, B., Krabbe, P.F., Jager, G.J., Oyen, W.J., and Ruers, T.J. 2005. Cancer, 104, 2658. 36. Vogel, W.V., Wiering, B., Corstens, F.H,M., Ruers, T.J.M, and Oyen, W.J.G. 2005. Cancer Imaging, 5, S143. 37. Lubezky, N., Metser, U., Geva, R., Nakache, R., Shmueli, E., Klausner, J.M., et al. 2007. J. Gastrointest. Surg., 11, 472. 38. Berger, K.L., Nicholson, S.A., Dehdashti, F., and Siegel, B.A. 2000. AJR Am. J. Roentgenol., 174, 1005. 39. Parsons, A.M., Detterbeck, F.C., and Parker, L.A. 2004. Ann. Thorac. Surg., 78, 1910. 40. Kang, M.C., Kang, C.H., Lee, H.J., Goo, J.M., Kim, Y.T., and Kim, J.H. 2008. Eur. J. Cardiothorac. Surg., 33, 473. 41. Divisi, D., Di Tommaso, S., Di Leonardo, G., Brianzoni, E., De Vico, A., and Crisci, R. 2010. Thorac. Cardiovasc. Surg., 58, 422. 42. Thomford, N.R., Woolner, L.B., and Clagett, T. 1965. J. Thorac. Cardiovasc. Surg., 49, 357. 43. Kanemitsu, Y., Hirai, T., and Yasui, K. 2004. Br. J. Surg., 91, 112.
Lung metastases of colorectal cancer: Focus on surgery
211
44. Pfannschmidt, J., Dienemann, H., and Hoffmann, H. 2007. Ann. Thorac. Surg., 84, 324. 45. Okumura, S., Tsuboi, M., Nakayama, H., Asamura, H., Tsuchiya, R., and Naruke, T. 1996. J. Thorac. Cardiovasc. Surg., 112, 867. 46. Higashiyama, M., Higaki, N., Takami, K., Murata, K., Kameyama, M., and Yokouchi, H. 2003. Jpn. J. Thorac. Cardiovasc. Surg., 51, 289. 47. Regnard, J.F., Nicolosi, M., Coggia, M., Spaggiari, L., Fourquier, P., Levi, J.F., and Levasseur, P. 1995. Gastroenterol. Clin. Biol., 19, 378. 48. Iizasa, T., Suzuki, M., Yoshida, S., Motohashi, S., Yasufuku, K., Iyoda, A., et al. 2006. Ann. Thorac. Surg., 82, 254. 49. Onaitis, M.W., Petersen, R.P., Haney, J.C., Saltz, L., Park, B., Flores, R., et al. 2009. Ann. Thorac. Surg., 87, 1684. 50. Vogelsang, H., Hierholzer, C., Berger, U., Siewert, J.R., and Präuer, H. 2004. Br. J. Surg., 91, 1066. 51. Melloni, G., Bandiera, A., Carretta, A., Ciriaco, P., Arrigoni, G., and Zannini, P. 2006. Ann. Thorac. Surg., 81, 2008. 52. Neeff, H., Hörth, W., Makowiec, F., Fischer, E., Imdahl, A., Hopt, U.T., and Passlick, B. 2009. J. Gastrointest. Surg., 13, 1813. 53. Mitry, E., Guiu, B., Cosconea, S., Jooste, V., Faivre, J., Bouvier, A.M. 2010, Gut, 59(10), 1383. 54. Headrick, J.R., Miller, D.L., Nagorney, D.M., Allen, M.S., Deschamps, C., Trastek, V.F., and Pairolero, P.C. 2001. Ann. Thorac. Surg., 71, 975. 55. Shah, S.A., Haddad, R., Al-Sukhni, W., Kim, R.D., Greig, P.D., Grant, D.R., et al. 2006. J. Am. Coll. Surg., 202, 468. 56. Kobayashi, K., Kawamura, M., and Ishihara, T. 1999. J. Thorac. Cardiovasc. Surg., 118, 1090. 57. Miller, G., Biernacki, P., Kemeny, N.E., Gonen, M., Downey, R., Jarnagin, W.R., et al. 2007. J Am Coll Surg., 205, 231. 58. Kim, A.W., Faber, L.P., Warren, W.H., Saclarides, T.J., Carhill, A.A., Basu, S., et al. 2008. Surgery, 144, 712. 59. Hwang, M.R., Park, J.W., Kim, D.Y., Chang, H.J., Kim, S.Y., Choi, H.S., et al. 2010. Ann. Thorac. Surg., 90, 398. 60. Shiono, S., Ishii, G., Nagai, K., Murata,Y., Tsuta, K., Nitadori, J., et al. 2006. Eur. J. Surg. Oncol., 32, 309. 61. Berney, CR., Yang, JL., Fisher, RJ., Russell, PJ., and Crowe, PJ. 1998, World J Surg., 22, 484. 62. Deschoolmeeter, V., Baay, M., Specenier, P., Lardon, F., Jan, B., and Vermorken, J. 2010, Oncologist, 1:699. 63. Watanabe, I., Arai, T., Ono, M., Sugito, M., Kawashima, K., Ito, M., et al. 2003, Brits J Surg., 90, 1436. 64. Werling, RW., Yaziji, H., Bacchi, CE., and Gown, AM. 2003, Am J Surg Pathol., 27, 303. 65. Barbareschi, M., Murer, B., Colby, TV., Chilosi. M., Macri, E., Loda, M., Doglioni, C. 2003, Am J Surg Pathol., 27(2), 141. 66. Padilla-Valverde, D. TesisDoctoral David Padilla Valverde. Departamento de especialidades Médico- quirúrgicas. Universidad de Córdoba 2001
212
José Luis Duque-Medina et al.
67. Tamura, M., Oda, M., Tsunezuka, Y., Matsumoto, I., Kawakami, K., and Wanatabe, G. 2004. J. Thorac. Cadiovasc. Surg., 128, 517. 68. Zafirellis, K., Agrogiannis, G., Zachaki, A., Gravani, K., Karameris, A., and Kombouras, C. 2008. J. Surg. Res., 147, 99. 69. Pfannschmidt, J., Bade, S., Hoheisel, J., Muley, T., Dienemann, H., and Herpel, E. 2009. Thorac. Cardiovasc. Surg., 57, 403. 70. Hansen, L., Gárdsvoll, H., Nielsen, B., Lund, L., Dano, K., Jensen, O, and Ploug, M. 2004, Biochem J., 380 (3), 845. 71. Rama, N., Monteiro, A., Bernardo, J.E., Eugenio, L., and Natunes, M. 2009. Eur. J. Cardiothorac. Surg., 35, 444. 72. Yamakodo, K., Hase, S., Matsuoka, T., Tanigawa, N., Najatsuka, A., Takaki, H., et al. 2007. J. Vasc. Interv. Radiol., 18, 393. 73. Ayarra-Jarne, J., Jimenez-Merchan, R., Congregado-Loscetales, M., Girón, J.K., Gallardo, G., Treviño, A.I., Arenas, C., and Loscertales, J.. 2008. Arch. Bronconeumol., 44, 525. 74. Poncelet, A.J., Lurquin, A., Weynard, B., Humbler, Y., and Nouhomme, P. 2007. Eur. J. Cardiothorac. Surg., 31, 173. 75. Tanaka, Y., Maniwa, Y., Nishio, W., Yoshimura, M., and Okita, Y. 2008. Eur. J. Cardiothorac. Surg., 33, 1135. 76. Ambiru, S., Miyazaki, M., Nakahawa, K., Shimizu, H., Omoto, H., and Nakajima, N. 1998. Cancer, 83, 1049. 77. Landes, L., Robert, J., Perneger, T., Mentha, G., Ott, V., Morel, P., and Gervaz, P. 2010. BMC Surg., 10, 17. 78. Limmer, S., Oevermann, E., Killaitis, C., Kujath, P., Hoffmann, M., and Bruch, H-P. 2010. Langenbecks Ann. Surg., 395, 1129. 79. Landreneau, R.J., Mack, M.J., Keenan, R.J., Hazdrigg, S.R., Dowling, R.D. and Ferson, P.F. 1993. Ann. Thorac. Surg., 56, 615. 80. Ninomiya, N., Nakayima, J., Tanaka, M., Takenchi E, Murakawa, T., and Fukami, T. 2001. Jpn. J. Thorac. Cardiovasc. Surg., 49, 17. 81. Nakajima, J., Murakawa, T., Fukami,T., and Takamoto, S. 2008, J Interactive Cardiovasc Thorac Surg., 7, 212. 82. Menon, A., Milton, R., Torpe, J.A.C., and Papagiannopoulos, K. 2008. Eur. J. Cardiothorac. Surg., 33, 1135. 83. Ogata, Y., Matono, K., Hayashi, A., Takamor, S., Miva, K., Sasatomi, T., et al. 2005. World J. Surg., 29, 363. 84. Baron, O., Hamy, A., Roussel, J-C., Galletta, D., Al Habash, O., Duveau, D., et al. 1999. Rev. Mal. Respir., 16, 809. 85. Coriat, R., Diaz, O., de la Fouchardière, C., Desseigne, F., and Négrier, S. 2007. Oncology, 73, 395. 86. Fournel, C., Bertoletti, L., Nguyen, B., and Vergnon, J.M. 2009. Respiration, 77, 63.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 213-225 ISBN: 978-81-308-0460-6 Editor: Esther U帽a Cid贸n
9. Peritoneal carcinomatosis of colorectal cancer Bernardino Rampone Department of Surgery, Pineta Grande Hospital, Castel Volturno, Italy
Abstract. Peritoneal Carcinomatosis (PC) from CRC accounts for approximately 13% of metastatic spread among patients who die from the disease (1). Despite significant advances in chemotherapy with regimens such as FOLFOX and FOLFIRI, approximately 13% of patients with PC achieve the 5-year survival mark with chemotherapy alone. Cytoreductive surgery (CS) combined with hyperthermic intraperitoneal chemotherapy (HIPEC) has shown impressive results with 5-year survival rates of approximately 50% in patients with disease limited to the peritoneum (2-4).
Introduction PC is, after liver metastasis, one of the most common routes of the dissemination of colorectal cancer (CRC) and represents the second most frequent cause of death in patients with this disease. Peritoneal involvement is encountered in 7 % of patients at primary surgery, while it develops in about 4% to 19% of patients after curative surgery and in up to 44% of patients with recurrent CRC (1,5). Correspondence/Reprint request: Dr. Bernardino Rampone, Department of Surgery, Pineta Grande Hospital Castel Volturno, Italy. E-mail: ramponebernardino@virgilio.it
214
Bernardino Rampone
PC is a sequence of events, as follows: access of cancer cells to the peritoneal cavity, adhesion to the mesothelial surface, invasion into subperitoneal space, proliferation and vascular neogenesis (6). The high incidence of tumour implantation on the peritoneal surface in CRC can occur by intraperitoneal tumour emboli as a result of serosal penetration. It can also be the consequence of surgical management through the leakage of malignant cells from lymphatic vessels or through dissemination due to tumour trauma as a result of dissection, with subsequent fibrin entrapment and tumour promotion of the entrapped cells (7). Peritoneal involvement from colorectal malignancies has traditionally been considered as a manifestation of terminal disease due to the limited response to conventional surgical and chemotherapeutic treatments (8,9). The current standard of care for colorectal PC is evolving from chemotherapy to CS combined with HIPEC for patients with disease limited to the peritoneum. PC from CRC treated with chemotherapy alone results in a median survival of 5â&#x20AC;&#x201C;13 months. In contrast, CS with HIPEC for early PC from CRC results in a median survival of 48â&#x20AC;&#x201C;63 months and the 5-year survival mark is achieved in 51% of patients (2-4). Dr Sugarbaker, one of the treatment pioneers for this disease, has suggested that PC is a locoregional cancer spread as a result of a molecular crosstalk between cancer cells and host elements (10). For this reason, the only chance to achieve long-term survival in these patients is the eradication of microscopic residual disease.
1. Cytoreductive surgery and hyperthermic intraperitoneal chemotherapy In the 1930s, Meigs was the first to strongly recommend the use of CS followed by adjuvant radiotherapy in patients with ovarian cancer (11). Thirty years later, Munnell and Griffiths demonstrated improved survival rates through extensive surgery, with the size of the residual disease as the most important prognostic factor (12,13). In 1980, after an experimental study of hyperthermic peritoneal perfusion in dogs, Spratt was the first to clinically test CS followed by HIPEC with thiotepa in a case of pseudomyxoma peritonei (14). Later, Sugarbaker introduced intraperitoneal chemotherapy as a new innovative therapeutic option for selected patients with PC (15,16). This treatment is based on surgical cytoreduction and HIPEC. The underlying rationale for this combined approach is to address macroscopic disease through an aggressive surgical approach combining visceral resection and peritonectomy, and to address residual microscopic disease with HIPEC (17).
Management of peritoneal carcinomatosis from colorectal cancer
215
Drugs selected for intraperitoneal administration are usually hydrophilic and have a large molecular size so that they pass slowly through the peritoneal-plasma barrier and become more effectively sequestered in the peritoneal cavity. Adding hyperthermia to intraperitoneal chemotherapy may increase tumour response to chemotherapeutic drugs by several mechanisms. Firstly, heat alone has a direct anti-tumour effect. Hyperthermia induces selective cytotoxicity of malignant cells. Secondly, the cytotoxic effects of some chemotherapeutic agents (doxorubicin, platinum complexes, mitomycin C, melphalan, docetaxel, irinotecan and gemcitabine) are augmented by applying mild hyperthermia (18). This locoregional therapy provides an elevated and persistent drug concentration to the tumour, while limiting the systemic concentration of the drug. HIPEC with mitomycin C, cisplatin and doxorubicin or oxaliplatin has been used for gastrointestinal PC. Oxaliplatin has been used intraperitoneally in Europe for colorectal PC, as pioneered by Elias (19). Oxaliplatin has a very low area under the curve ratio, which means that it is rapidly absorbed and does not require a long dwell time in the peritoneal cavity for maximal locoregional effect (19). However, the extent of intraperitoneal chemotherapy penetration into tumour nodules by passive diffusion is limited to a few cell layers. For this reason, Elias first suggested that intraperitoneal oxaliplatin should be combined with intravenous 5-FU administered just before HIPEC (bidirectional intraoperative chemotherapy) (19). By combining intraoperative intravenous and intraoperative intraperitoneal cancer chemotherapy, a bidirectional diffusion gradient is created through the intermediate tissue layer that contains the cancer nodules. This offers opportunities for optimising cancer chemotherapy delivery to the target peritoneal tumour nodules. Nevertheless, further pharmacologic studies are needed to determine the most efficient administration regimen (continuous vs. bolus vs. repeated bolus), doses and choice of chemotherapeutic drugs for this bidirectional approach.
2. Staging system of peritoneal carcinomatosis 2.1. Gilly staging Gilly introduced this system in 1994. This staging takes into account the size and distribution of malignant granulations (Table I). Its two main advantages are simplicity and reproducibility. The usefulness of this technique was shown in many studies and in particular in a prospective multicentre study of 370 patients with PC from non-gynecological malignant disease. Patients with PC stage 1 and 2 (malignant granulations < 5 mm) survived significantly longer than those with stage 3 and 4 (malignant granulations â&#x2030;Ľ 5 mm).
216
Bernardino Rampone
Table 1. Gilly Staging of PC (20). Stage Description Stage 0 Stage 1 Stage 2 Stage 3 Stage 4
No macroscopic disease. Malignant granulations less than 5 mm in diameter. Localised to one part of the abdomen. Malignant granulations less than 5 mm in diameter. Diffuse to the whole abdomen. Localised or diffuse malignant granulations 5â&#x20AC;&#x201C;20 mm in diameter. Localised or diffuse large.
The shortcoming of this staging system is that it does not clearly indicate the potential resectability of carcinomatosis. Stage 2 disease could consist of diffuse PC with nodules of less than 5 mm that are non-resectable. Conversely, stage 3 and 4 disease could include diffuse and localised PC with nodules of 5 mm or greater that are resectable (20). 2.2. Peritoneal cancer index Jacquet and Sugarbaker described this staging system in 1996 (21). The Peritoneal Cancer Index (PCI) is a more precise assessment of carcinomatosis quantification and distribution. It evaluates quantitatively the distribution and implant size of the cancer throughout the abdomen and pelvis (Figure 1). The abdomen and the pelvis are divided by lines into nine regions (AR0-8). The small bowel is then divided into four regions. Region 9 and 10 define the upper and lower portions of jejunum, and regions 11 and 12 define the upper and lower portions of the ileum. The lesion sizes of the largest implants are scored in each abdominopelvic region. Implants are scored as lesion size 0 through to 3 (LS-0 to LS-3). LS-0 means that no implants are seen throughout the region; this measurement is made after a complete lysis of all adhesions and the complete inspection of all parietal and visceral peritoneal surfaces. LS-1 refers to visible implants up to 0.5 cm in diameter. LS-2 identifies nodules greater than 0.5 cm and up to 5 cm. LS-3 refers to implants 5 cm or greater in diameter. This method quantifies the extent of disease within each region of the abdomen and pelvis, which can then be summed as a numerical score (from 1 to 39) for the peritoneal cavity as a whole (21). The combination of Gilly staging system and PCI contributes to the precise intraoperative description of carcinomatosis implants within the abdominopelvic cavity. Using both classifications provides an accurate map of the lesion.
Management of peritoneal carcinomatosis from colorectal cancer
217
Figure 1. Peritoneal Cancer Index (21).
In addition, the PCI can be used to estimate the probability of complete CR and has value as a prognostic factor. Colorectal carcinomatosis patients with a PCI greater than 20 should be treated only with palliative intent (22,23). 2.3. Completeness of cytoreduction score The size of tumour nodules remaining after cytoreduction (CR) has been shown to predict prognosis by estimating the possibility of cancer eradication by HIPEC. Several studies (24-26) have shown a direct relation between the completeness of CS and survival for carcinomatosis from all primary cancer locations. The Lyon group (25) has successfully used complete (R0â&#x20AC;&#x201C;R1) or incomplete (R2) CR to assess the completeness of the surgical clearance of cancer. Confirming an R0 resection is difficult in patients with carcinomatosis; thus, R0 and R1 can be grouped together because the outcome of these two groups is very similar (27-29). Jacquet and Sugarbaker (21) used the completeness of CR score (CC score) to assess the surgical clearance of carcinomatosis. A CC-0 score indicates that no peritoneal seeding was exposed during the complete exploration; a CC-1 score indicates that the tumour nodules persisting after
218
Bernardino Rampone
CR are less than 2.5 mm in diameter; a CC-2 score indicates that the nodules are between 2.5 mm and 25 mm in diameter; and a CC-3 score indicates that the nodules are greater than 25 mm in diameter or may include unresectable tumour nodules at any site within the abdomen and the pelvis (30,31).
3. HIPEC technique There are two types of HIPEC: one in which the abdomen is closed during chemotherapy (closed-abdomen HIPEC) and one in which the abdomen is left open (open-abdomen HIPEC). HIPEC is delivered once tumour CR has been concluded and before any digestive reconstruction or diversion is made. The rationale for this timing in relation to GI tract reconstruction concerns the opportunity of exposing bowel section lines to the chemotherapy solution in an effort to minimise the chance for anastomotic or staple line recurrence. Although this is the classical way to do it, some groups perform anastomoses before the administration of HIPEC with no apparent increase in anastomotic recurrences. The open method is usually performed by the “Coliseum technique”, as described by Sugarbaker (32). Once the cytoreductive phase has been finalised, a Tenckhoff catheter and four closed suction drains are placed through the abdominal wall and made watertight with a purse string suture at the skin. A different number of temperature probes secured to the skin edge may be used for intraperitoneal temperature monitoring; at least one in the inflow line and another one at the pelvis are employed. The skin edges of the abdominal incision are suspended to a Thompson self-retaining retractor by a running monofilament number 1 suture to create an open space in the abdominal cavity. A plastic sheet is incorporated into this suture to prevent the chemotherapy solution from splashing. A slit in the plastic cover is made to allow the surgeon’s double gloved hand access to the abdomen and pelvis. Impervious gown and protection goggles are mandatory. The smoke evacuator is placed under the plastic sheet to clear chemotherapy particles that may be liberated during the procedure. During the 30 min to 90 min perfusion, all the anatomic structures within the peritoneal cavity are uniformly exposed to heat and chemotherapy by continuous manipulation of the perfusate. A roller pump forces chemotherapy perfusion into the abdomen through the Tenckhoff catheter and pulls it out through the drains, with a flow rate around 1 L/min. A heat exchanger keeps the fluid being infused at 43–45°C so that the intraperitoneal fluid is maintained at 41–43°C. The single-use circuit tubing is commercially available from the HIPEC manufacturer or from the cardioplegia industry, and most of them incorporate a reservoir, useful when the chemotherapy solution needs to be quickly extracted from the abdomen
Management of peritoneal carcinomatosis from colorectal cancer
219
for any complication or in cases where the perfusate volume calculated cannot be fully accommodated by the peritoneal cavity capacity. The perfusate is first recirculated between the reservoir and the heat exchanger so that it can be heated to an adequate temperature. At this point, full circulation of the perfusate in and out of the peritoneal cavity is established until a minimum intraperitoneal temperature of 41.5°C is achieved and maintained. The drug is then added to the circuit and the timer for the perfusion is started. The main benefit of the Coliseum technique is that heated chemotherapy is adequately distributed throughout the abdominal cavity and there is no pooling of temperature or chemotherapy. One disadvantage of the open technique is heat dissipation, which makes it more difficult to initially achieve a hyperthermic state. Another possible disadvantage is the increased exposure of operating room personnel to chemotherapy. As the surgeon is manipulating chemotherapy throughout the perfusion, an increased potential for contact exposure exists. Furthermore, because the abdomen is open during the perfusion, the heated chemotherapy could form an aerosol, creating a risk of inhalation exposure. The safety of operating room personnel during the Coliseum technique was evaluated in a study published in 2002 (33). Urine from members of the operating team was assayed for chemotherapy levels. Air sampled below and above the plastic sheet was also analysed. Finally, sterile gloves commonly used in the operating room were examined for permeability to chemotherapy. All assessments of potential exposures were found to be negative and in compliance with established safety standards. The open technique has theoretical advantages over the closed technique due to improved distribution of heated chemotherapy; however, superiority has not been definitively proven in a randomised controlled trial. A variation of the open technique mainly used in Japan uses a device called â&#x20AC;&#x153;peritoneal cavity expanderâ&#x20AC;? (PCE). The PCE is an acrylic cylinder containing in-flow and out-flow catheters secured over the wound. When filled with heated perfusate, the PCE can accommodate the small bowel, allowing it to float freely and be manually manipulated in the perfusate. After HIPEC is completed, the perfusate is drained and the PCE is removed. By using the expander, a more uniform distribution is theoretically achieved compared to the closed technique. The main disadvantage of the PCE technique is the risk of exposure to chemotherapy of the operating room personnel, as in the Coliseum technique (34). In the closed technique, catheters and temperature probes are placed in the same fashion but the laparotomy skin edges are sutured watertight so that perfusion is done in a closed circuit. The abdominal wall is manually agitated during perfusion in an attempt to promote uniform heat distribution. A larger
220
Bernardino Rampone
volume of perfusate is generally needed to establish the circuit compared with the open technique and a higher abdominal pressure is achieved during the perfusion, which may facilitate drug tissue penetration. After perfusion, the abdomen is reopened and the perfusate is evacuated. Appropriate anastomoses are performed and the abdomen is closed in the standard fashion. A major advantage of the closed technique is the ability to rapidly achieve and maintain hyperthermia as there is minimal heat loss. In addition, there is minimal contact or aerosol exposure of the operating room staff to the chemotherapy. The only way for exposure to occur is from leakage through the surgical wound or catheter wounds. The main disadvantage is the lack of uniform distribution of the chemotherapy. When methylene blue was instilled using the closed technique, uneven distribution was observed. Uneven distribution of HIPEC is problematic, because hyperthermia has a narrow therapeutic index. Tumouricidal activity is manifested at 41–43°C; therefore, in-flow temperature usually exceeds 45°C (35). Rats exposed to intraperitoneal temperatures of 45°C suffered significant morbidity and mortality (36). Therefore, inadequate circulation of heated perfusate leads to pooling and accumulation of heat and chemotherapy in dependent parts of the abdomen. This may result in increased systemic absorption and foci of hyperthermic injury that could contribute to postoperative ileus, bowel perforation and fistula. Consequently, other intra-abdominal areas may be undertreated. CR and the HIPEC closed technique can be performed safely in combination, as different centres have reported (37,38). The morbidity associated with the closed procedure includes myelosuppression, ileus and fistula, as with the open technique. Heterogeneous distribution inside the closed abdomen may increase the rate of intra-abdominal complications. In the last few years, increased interest in HIPEC has led to the commercial development of hyperthermic intraperitoneal perfusion systems. These are compact devices that contain roller pumps, a heating device, a heat exchanger and temperature monitors in a single apparatus. A computer integrates and displays information from the temperature probes, inflow and outflow rates. Several options are commercially available at this time. The role of the anesthesiologist is crucial during HIPEC, as it is during the whole complex cytoreductive procedure. Specific training is desirable. During the whole lengthy surgical procedure, knowledgeable fluid management needs to be carried out, keeping a balance between the use of crystalloids and colloids to achieve adequate central venous pressures and urine output without incurring fluid overload. The latter is a common undesirable side effect observed after this surgery with consequences that range from acute pulmonary edema to cerebral edema when anesthesiologists
Management of peritoneal carcinomatosis from colorectal cancer
221
not familiarised with this procedure are assigned to these cases. A minimal urine output of 100 cc (desirable 150 cc) every 15 min during the administration of HIPEC is mandatory to avoid renal toxicity caused by the cytotoxic drug employed. The utilisation of a low-dose dopamine perfusion is commonly used to that effect. Central temperature is monitored by an esophageal probe and may be expected to rise up to 39°C or more; different cooling measures need to be implemented at this time to avoid sustained central hyperthermia starting by turning off the air heating blankets and moving to the intravenous administration of cold crystalloids or the placement of ice packs around the head and neck of the patient.
4. Result to date in treating colorectal PC Since the introduction of HIPEC, promising results have been reported by several groups in the treatment of PC (7,39). In all published series, the most important prognostic factors in predicting survival are the involvement of disease encountered at laparotomy (measured by peritoneal cancer index, PCI) and the completeness of resection (5). A retrospective analysis of prognostic factors in 71 patients treated by CS plus HIPEC for PC from CRC reported a median survival of 41 months in patients with a PCI < 20, while in cases with PCI > 20, the survival was 16 months (23). When there is extension of the disease to six or seven regions of the abdomen, patients have a poor prognosis, with a median survival of 5.4 months vs. 29 months in patients with a lower number of regions affected (23,24). Glehen et al. (40), in a study of 523 patients with PC from CRC, have reported a survival of 32.4 months in patients who had complete cytoreduction, compared with 8.4 months in patients in whom radical CS was not possible. The morbidity rates reported after HIPEC combined with CS range from 20% to 50% and mortality rates from 1% to 10% (41,42). The main morbidities associated with this advanced treatment are caused by complications of surgery and hematological toxic effects. The most frequent surgical complications include anastomotic leakage, intestinal perforation, pancreatitis, prolonged ileus, bile leak, intra-abdominal bleeding/sepsis, wound dehiscence, pulmonary embolism and renal failure. Intra-abdominal sepsis and enteric fistulas often necessitate re-operation. Most of these complications can be attributed to the extensive surgery performed, especially when the patient has had multiple previous operations. This procedure on average may take up to 10â&#x20AC;&#x201C;12 h with a median blood loss of 3â&#x20AC;&#x201C;5 L and necessitates blood transfusion, complex anesthetic management, intensive care and re-operation. To add to the complexity, Sugarbaker stresses that the dosing of intraperitoneal chemotherapy often needs to be modified to
222
Bernardino Rampone
keep the morbidity and mortality for these patients at 30% and 2%, respectively. He suggests using a one-third dose reduction in patients who have had prior extensive surgeries, extensive CS, prior chemotherapy or radiation therapy, if multiple anastomoses need to be performed, and in patients who are older than 65 years (43). Results of multivariate analyses have shown that the independent factors affecting morbidity are the duration of surgery, extent of carcinomatosis, number of anastomoses performed, and gender (44,45). The advanced stage of neoplastic disease and the immunodeficient status of patients previously subjected to chemotherapy were found to be important factors that probably contributed to the occurrence of septic complications after an extended surgical procedure. Furthermore, there are several technical and procedural nuances that mandate that only those who have considerable experience should perform these procedures. Indeed, both the CS and administration of HIPEC are technically demanding procedures with clear learning curves. In a study of 70 patients treated with HIPEC combined with CS, Yan et al. (46) reported a mortality of 4%, moderate morbidity in 44% of patients and severe morbidity in 20%; the investigators suggested that there is a learning curve associated with this advanced treatment in order to achieve an acceptable morbidity rate. Smeenk et al. (47) observed that the peak of the learning curve, graded by the percentage of complete cytoreductions, was reached after approximately 130 procedures. Therefore, such a complex, aggressive modality can be a considerable drain on the resources of institutions in order to provide a benefit for a small group of patients. Recently, Elias et al. (48) reported a multi-institutional retrospective analysis of CS and HIPEC for patients with PC from CRC. The study involved the treatment of 523 patients over a 17-year period. Almost 30% of the patients (n = 152) were from a single institution, whereas, importantly, 18 centres recorded fewer than 20 cases each. The purpose of the study was to evaluate the short-term and long-term efficacy of this combined approach and identify prognostic factors affecting outcome. The overall mortality was 3.3%, with 31% grade 3 to 4 morbidity, and the median survival was 30.1 months. The risk of postoperative morbidity and mortality was significantly influenced by two factors: the peritoneal index (reflecting the peritoneal metastatic burden) and the centre in which the treatment was performed (inexperienced centres were those with < 7 years of practice). The centre experience was an independent predictor for achieving the important goal of CS success (49). Currently, patients with CRC who develop peritoneal recurrence might benefit from CS and HIPEC even in the presence of resectable liver metastasis (3,42,49-52).
Management of peritoneal carcinomatosis from colorectal cancer
223
Because this treatment has been shown to be most effective in patients with absent or minimal residual disease, there might be a role for preventive HIPEC in patients at high risk of peritoneal tumour spread. In a study of 29 patients with CRC at high risk of peritoneal recurrence, PC that was not detected by preoperative imaging was present in 16 patients. In these cases, HIPEC resulted in a disease-free status in half of the patients after a median follow-up of 27 months (53,54).
Conclusion The clinical outcome after a diagnosis of PC has radically changed during the last two decades due to combined locoregional treatment. Currently, HIPEC provides a promising therapeutic option for patients with PC from CRC. Patients who have minimal residual disease as a result of cytoreductive surgery are candidates for HIPEC. This approach has become an important part of CRC treatment and should become a standard modality for the prevention and treatment of cancer that involves the peritoneal surface (55). The national health care systems of the Netherlands and France have approved this approach for colon carcinomatosis. In other countries in Europe, approval is given on a case-by-case basis. To achieve maximum benefit, correct patient selection, radical CR and a proper training are necessary. However, further controlled studies will help standardise indications and techniques to improve morbidity and mortality rates.
References 1. 2. 3. 4. 5. 6. 7. 8.
Jayne, DG, Fook, S, Loi, C, and Seow-Choen. 2002, Br J Surg., 89(12), 1545. Elias, D, Lefevre, JH, Chevalier, J, Brouquet, A, Marchal, F, Classe, JM, et al. 2009, J Clin Oncol., 27(5), 681. Verwaal, VJ, van Ruth, S, Witkamp, A, Boot, H, van Slooten, G, and Zoetmulder, FA. 2005, Ann Surg Oncol., 12, 65. Yan, TD, Black, D, Savady, R, and Sugarbaker, PH. 2006, J Clin Oncol., 24(24), 4011. Koppe, MJ, Boerman, OC, Oyen, WJ, and Bleichrodt, RP. 2006, Ann Surg., 243(2), 212. Confuorto, G, Giuliano, ME, Grimaldi, A, and Viviano, C. 2007, Surg Oncol., 16(Suppl 1), S149. Witkamp, AJ, de Bree, E, Kaag, MM, Boot, H, Beijnen, JH, van Slooten, GW, van Coevorden, F, and Zoetmulder, FA. 2001, Eur J Cancer., 37, 979. Chu, DZ, Lang, NP, Thompson, C, Osteen, PK, and Westbrook, KC. 1989, Cancer, 63(2), 364.
224
9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.
33. 34. 35.
Bernardino Rampone
Sadeghi, B, Arvieux, C, Glehen, O, Beaujard, AC, Rivoire, M, Baulieux, J, et al. 2000, Cancer, 88(2), 358. Sugarbaker, PH. 2007, J Surg Oncol., 95(2), 93. Meigs, JV. 1935, Tumours of the female pelvic organs, New York, Macmillan (ed). Munnell, EW. 1968, Am J Obstet Gynecol., 100, 790. Griffiths, CT. 1975, Natl Cancer Inst Monogr., 42, 101. Spratt, JS, Adcock, RA, Muskovin, M, Sherrill, W, and McKeown. 1980, Can Res., 40(2), 256. Sugarbaker, PH, Cunliffe, WJ, Belliveau, J, de Bruijn, EA, Graves, T, Mullins, RE, and Schlag, P. 1989, Semin Oncol., 16(4 Suppl 6), 83. Sugarbaker, PH. 1996, Cancer Treat Res., 81, 149. Sugarbaker, PH. 2003, Surg Oncol Clin N Am., 12(3), 703. Van der Speeten, K, Stuart, OA, and Sugarbaker, PH. 2009, Curr Drug Discov Technol., 6, 72. Elias, DM, and Sideris, L. 2003, Surg Oncol Clin N Am., 12, 755. Glehen, O, Mohamed, F, and Gilly, FN. 2004, Lancet Oncol., 5(4), 219. Jacquet, P, and Sugarbaker, PH. 1996, Peritoneal carcinomatosis: principles of management, In Sugarbaker PH (Ed)., Kluwer Academic Publishers, Boston, USA, 359. Sugarbaker, PH. 1998, Semin Surg Oncol., 14, 254. da Silva, RG, and Sugarbaker, PH. 2006, J Am Coll Surg, 203, 878. Elias, D, Blot, F, El Otmany, A, Antoun, S, Lasser, P, Boige, V, et al. 2001, Cancer, 92, 71. Glehen, O, Mithieux, F, Osinsky, D, Beaujard, AC, Freyer, G, Guertsch, P, et al. 2003, J Clin Oncol., 21, 799. Shen, P, Levine, EA, Hall, J, Case, D, Russell, G, Fleming, R, et al. 2003, Arch Surg., 138, 26. Beaujard, AC, Glehen, O, Caillot, JL, Francois, Y, Bienvenu,, J, Panteix G, et al. 2000, Cancer, 88, 2512. Sayag-Beaujard, AC, Francois, Y, Glehen, O, Sadeghi-Looyeh, B, Bienvenu, J, Panteix, G, et al. 1999, Anticancer Res., 19, 1375. Gilly, FN, Beaujard, A, Glehen, O, Grandclement, E, Caillot, JL, Francois, Y, et al. 1999, Anticancer Res., 19, 2317. Sugarbaker, PH, and Chang, D. 1999, Ann Surg Oncol., 6, 727. Pestieau, SR, and Sugarbaker, PH. 2000, Dis Colon Rectum, 43, 1341. Sugarbaker, PH. 2005, Handbook for the Integration of Cytoreductive Surgery and Perioperative Intraperitoneal Chemotherapy into the Surgical Management of Gastrointestinal and Gynecologic Malignancy, (4th Ed), Grand Rapids, Ludann Company, Michigan. Stuart, OA, Stephens, AD, Welch, L, and Sugarbaker, PH. 2002, Ann Surg Oncol., 9, 186. Fujimura, T, Yonemura, Y, Fujita, H, Michiwa, Y, Kawamura, T, Nojima, N, et al. 1999, Int Surg., 84, 60. Elias, D, Detroz, B, Debaene, B, Damia, E, Leclercq, B, Rougier, P, Lasser, P. 1994, Hepatogastroenterology, 41, 207.
Management of peritoneal carcinomatosis from colorectal cancer
225
36. Fumagalli, U, Trabucchi, E, Soligo, M, Rosati, R, Rebuffat, C, Tonelli, C, and Montorsi, M. 1991, J Surg Res., 50, 82. 37. Loggie, BW, Fleming, RA, McQuellon, RP, Russell, GB, and Geisinger, KR. 2000, Am Surg., 66, 561. 38. Fujimoto, S, Takahashi, M, Kobayashi, K, Kasanuki, J, and Ohkubo, H. 1996, Cancer Treat Res., 81, 239. 39. Roviello, F, Marrelli, D, Neri, A, Cerretani, D, de Manzoni, G, Pedrazzani, C, et al. 2006, World J Surg., 30, 2033. 40. Glehen, O, Cotte, E, Schreiber, V, Sayag-Beaujard, AC, Vignal, J, and Gilly, FN. 2004, Br J Surg., 91, 747. 41. Shen, P, Hawksworth, J, Lovato, J, Loggie, BW, Geisinger, KR, Fleming, RA, and Levine, EA. 2004, Ann Surg Oncol., 11, 178. 42. Glehen, O, Kwiatkowski, F, Sugarbaker, PH, Elias, D, Levine, EA, De Simone, M, et al. 2004, J Clin Oncol., 22, 3284. 43. Sugarbaker, PH. Re: Verwaal, VJ, van Tinteren, H, Ruth, SV, and Zoetmulder, FA. 2004, J Surg Oncol., 85, 61. J Surg Oncol., 88, 276. 44. Glehen, O, Osinsky, D, Cotte, E, Kwiatkowski, F, Freyer, G, Isaac, S, et al. 2003, Ann Surg Oncol., 10, 863. 45. Stephens, AD, Alderman, R, Chang, D, Edwards, GD, Esquivel, J, Sebbag, G, et al. 1999, Ann Surg Oncol., 6, 790. 46. Yan, TD, Links, M, Fransi, S, Jacques, T, Black, D, Saunders, V, and Morris, DL. 2007, Ann Surg Oncol., 14, 2270. 47. Smeenk, RM, Verwaal,, VJ, and Zoetmulder FA. 2007, Br J Surg., 94, 1408. 48. Elias, D, Gilly, F, Boutitie, F, Quenet F, Bereder, JM, Mansvelt, B, et al. 2010, J Clin Oncol., 28, 63. 49. Verwaal, VJ, van Ruth, S, de Bree, E, van Sloothen, GW, van Tinteren, H, Boot, H, Zoetmulder, FA. 2003, J Clin Oncol., 21, 3737. 50. Elias, D, Benizri, E, Pocard, M, Ducreux, M, Boige, V, and Lasser, P. 2006, Eur J Surg Oncol., 32, 632. 51. Kianmanesh, R, Scaringi, S, Sabate, JM, Castel, B, Pons- Kerjean, N, Coffin, B, et al. 2007, Ann Surg., 245, 597. 52. Esquivel, J, Sticca, R, Sugarbaker, P, Levine, E, Yan, TD, Alexander, R, et al. 2007, Ann Surg Oncol., 14, 128. 53. Elias, D, Goéré, D, Di Pietrantonio, D, Boige, V, Malka, D, Kohneh-Shahri, N, et al. 2008, Ann Surg., 247, 445. 54. Brouquet, A, Goéré, D, Lefèvre, JH, Bonnet, S, Dumont, F, Raynard, B, and Elias, D. 2009, Ann Surg Oncol., 16, 2744. 55. Rampone, B, Schiavone, B, Martino, A, and Confuorto, G. 2010, World J Gastroenterol., 16(11), 1299.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 227-261 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
10. The role of radiotherapy in colorectal cancer Francisco López-Lara and Patricia Diezhandino García Radiation Oncology Department, Clinical University Hospital of Valladolid, Spain
Abstract. The optimal management of CRC today requires a multidisciplinary approach that integrates gastroenterologists, radiologists, pathologists, surgeons, medical oncologists, and radiation specialists. Although surgery remains the mainstay of treatment, the rate of local recurrence in rectal and rectosigmoid junction cancer is more than negligible. Both EBRT and chemotherapy, therefore, play fundamental roles in increasing the rates of both local control and OS of this disease. This chapter attempts to clarify the current status of EBRT with or without chemotherapy in rectal and rectosigmoid junction cancer, based on scientific evidence.
Introduction CRC is one of the most common cancers in Spain and also in the world, but this term includes two diseases that share aspects such as etiology, diagnosis, pathology, and treatment of metastatic disease, but that differs significantly in the treatment of localised disease, locally advanced disease, and locoregional recurrences rates. Correspondence/Reprint request: Dr. Francisco López-Lara, Radiation Oncology Department, Clinical University Hospital of Valladolid, Spain. E-mail: flopezlara@hotmail.com
228
Francisco L贸pez-Lara & Patricia Diezhandino Garc铆a
The rectum is considered to range from the last 12-15 cm of the most distal intestine to the anus, although no anatomic definition is uniformly accepted because of constitutional differences by age and sex. It is considered that the rectum ends at the dentate line, 1-2 cm below the anorectal ring. The mainstay of treatment remains surgery, achieving cure rates of approximately 50% (1). The anatomical location, however, determines the natural history of the disease, with a greater propensity for local recurrence, which entails a different therapeutic approach than its counterpart in colon cancer. Rectal cancer thus tends to present two patterns of recurrence after surgery with curative intent. One pattern, which is shared with colon cancer, is the appearance of distant metastases in 25% of cases, and the other pattern is pelvic-presacral recurrence (25-30%) (2) that is mostly exclusive to rectal cancer. Rates of recurrence vary also according to the series evaluated; some studies report numbers of local recurrences of 10-20% in T3-T4 (3) and up to 50-60% in patients with positive nodes (N) (4). This pattern of recurrence is a serious problem because it has a negative impact on OS and also causes a significant deterioration in the quality of life.(QOL). Approximately 70% of patients have symptoms of pelvic relapse resulting from the infiltration of vascular structures, nerve plexus, soft tissues, or bone structures. Most cases manifest as bleeding, neuropathic pain, infection, etc. The usefulness of cancer treatments in these patients is limited by the high morbidity associated with the treatments. Numerous efforts have been made to prevent local recurrences. The root cause is postulated to be the lateral growth of malignant cells incompletely removed at surgery. The introduction of total mesorectal excision (TME) as a surgical technique has reduced rates of local recurrence up to 5% at 5 years (5). During the past decades different treatment modalities have been examined such as postoperative chemoEBRT with different 5FU based schedules, preoperative EBRT short course and long course alone or in combination with 5FU based regimens or even with new drugs and intraoperative radiation therapy (IORT). This chapter aims to update the treatment of rectal and rectosigmoid junction cancer by focusing primarily on treatment with EBRT.
1. Techniques and dose fractionation External-beam radiation therapy (EBRT) in rectal cancer should include the primary tumour and regions of draining lymph nodes of the pelvis. Correct and optimal planning of treatment is required to minimise the dose to organs at risk, thus limiting toxicity to the patient.
Radiotherapy in colorectal cancer
229
The superior port edge is placed at the L4/L5 interspace usually in the mid-L5 vertebral body. The distal port edge should be 5 cm below palpable tumour for patients receiving preoperative treatment. For postoperative cases the distal port edge is about 5 cm below the best estimate of the preoperative tumour bed and (if an abdomino-perineal has been performed) below the perineum. Anterior and posterior portals must have at least a 1.5-cm margin on the pelvic brim. Lateral treatment portals should encompass the entire sacrum posteriorly, 1.5-2 cm behind the sacrum anterior margin of the sacrum (6). A radiopaque marker should be placed at the posterior aspect of the anus to ensure that blocks in the posterior-inferior aspect of the portal do not impinge on targeted portions of the anorectum. The anterior margin should be at least 4 cm anterior to the rectum, as determined by the rectal-contrast simulation. The use of intrarectal contrast simulation during portals ensures that the rectum will be at maximum distension. If the tumour has considerable extrarectal extension, then these guidelines should be modified to make certain that all macroscopic disease (determined by CT scan) is encompassed with about a 4-cm margin (7). The definition of volumes of current planning with 3D techniques (8) are based on patterns of recurrence from the period before the introduction of TME surgery, and most provide references to RT in 2D. Following the guidelines of the International Commission on Radiation Units and Measurements (ICRU 50 and 62), the volumes planned are: â&#x20AC;˘ â&#x20AC;˘
â&#x20AC;˘
GTV (Gross Tumour Volume): includes the gross rectal tumour CTV (Clinical Target Volume): includes the rectal mesentery and regions of lymph nodes at risk. Must include the presacral space, internal and distal common iliac nodes. The risk of lymph paraortical is very low so irradiation is not needed. The area of the external iliac nodes should not be included either, only be performed for lesions that extend to the dentate line or in cases of affected structures with lymph drainage through them (vagina, uterus, prostate, bladder) and lymph inguinal nodes if there is involvement of the anal or distal third of the vagina. PTV (Planning Target Volume): includes the GTV and CTV with a margin of safety. This margin is not symmetric, because the movement of the rectum is anisotropic, i.e. movement is not the same on all axes and globally has been estimated at 1 cm for all margins (9).
In the event of amputation abdomino-pelvic surgery, postoperative RT irradiating the perineal region with a margin of 1.5-2 cm, and sometimes depending on the energy used in dosimetry, placing a bolus in the perineal incision will be necessary to optimise the dose. A study conducted at the Mayo Clinic from 1976 to 1984 showed a decrease in local recurrence and scarring,
230
Francisco L贸pez-Lara & Patricia Diezhandino Garc铆a
when the scar is radiated following an intervention of abdomino-perineal type, but not in the anterior resections where this scar does not exist (10). The technique of administration of radiation therapy should be precisely performed to avoid side effects as much as posible, because side effects can have an impact on survival. One example is the study of the Swedish group (11) in which preoperative EBRT with a dose of 25 Gy in 5 fractions of 5 Gy in two parallel opposed fields, anteroposterior and posteroanterior (AP and PA) to 849 patients, showing a mortality 8% at 30 days compared to 2% in nonirradiated patients (p <0.01). Mortality was secondary to infectious complications, cardiovascular complications, and progression of the disease. A Scandinavian study (12) using the same dose administered with a 3-field technique (AP, PA, and later) saw decreased mortality, at 4% in the RT group compared to 5% in the non-irradiated group (p = 0.4).
Dose and fractionation Rectal cancer exhibits a dose-response relationship. A review of phase II studies at Princess Margaret Hospital observed higher rates of pathological responses in patients who had received a dose of 50 Gy compared to those who received 40 or 46 Gy (13). This is the reason why, recommended doses range from 46 to 50 Gy in 23-25 fractions of 2 Gy. When administered concomitantly, CT is performed at a fractionation of 1.8 Gy per session to a total of 50.4 Gy, with an overlay to 54 Gy in the primary tumour in postoperative RT (14). If the volume of small bowel included in the scope of treatment could be minimised, a boost of 9 Gy overlay is administered, which will be of 5.4 Gy in the case can not be excluded bowel. A comparison of the fractions used in several studies found that Rt in 5 sessions of 5 Gy was less effective in stage III. A standard fractionation scheme of 46-50 Gy at 2 Gy per session is recommended to minimise toxicity. Intensitymodulated radiation therapy (IMRT) with image-guided systems did not confirm toxicity (15) and even if different doses and fractionation schemes have been compared, doses of 46-50 Gy (scheme long) is the standard recommended by all groups (16). Indeed, shorter fractionation RT combined to CT treatment is impossible. The two schemes have not been compared using current techniques. A study in 2010 concluded that the biological effective dose (BED), the actual dose at the cellular level taking into account cell division, must be greater than 30 Gy. Data from this study demonstrate that preoperative RT with a BED > 30 Gy (17) is more efficient in reducing local recurrence and mortality than preoperative RT with a BED < 30 Gy and is independent of the scheme of fractionation used (short or long course). A phase III trial is currently under way in Stockholm (18) in which patients are randomised to 3 arms: EBRT in 5
Radiotherapy in colorectal cancer
231
fractions of 5 Gy with immediate surgery, 5 fractions of 5 Gy with delayed surgery, and 50 Gy with standard fractionation and delayed surgery. Determining the best regimen of treatment must await the results of this study.
Immobilisation Historically, EBRT patients with rectal cancer were placed in the prone position to reduce the volume of irradiated small bowel (19). Combined with proper planning using non-coplanar beams, a combination of different energies, and the Belly-Board system, consisting of a table with a central opening for the patient's abdomen, this positioning allows better tolerance of the RT on the pelvis with lower acute and late toxicities (20). Currently, however, the use of EBRT multileaf collimators, 3D and IMRT (Intensity Modulated Radiation Therapy), and CT imaging systems (computerised tomography), and MRI (magnetic resonance imaging) to plan treatments (21) and several studies have noted that the prone position is so difficult to maintain in many patients, and irreproducible (22). There is a small bowel volume included in the planned volume similar in the two positions in both prone and supine so some authors recommend the use of supine tolerance for the best when planning the treatments (23). See Figures 1, 2, 3, and 4.
Figure 1. Definition of structures in preoperative rectal cancer in supine position. GTV: gross tumour in red, yellow ganglion chains, CTV and PTV in blue. Small intestine: orange. MasterPlan Oncentra Planner version 3.2.
232
Francisco L贸pez-Lara & Patricia Diezhandino Garc铆a
Figure 2. Definition of structures in preoperative rectal cancer in the prone position. CTV: node chains in red and PTV in blue. Small intestine in yellow.
Figure 3. Planning for preoperative rectal cancer with 3 shaped fields MLC 80, side impact and rear wedge.
Radiotherapy in colorectal cancer
233
Figure 4. Planning for preoperative rectal cancer 3 shaped fields MLC 80 with incidents and anterior lateral wedges.
The advent of IMRT systems and image-guided RT (IGRT) has allowed a decrease in intestinal toxicity with good coverage in the PTV (24). In planning a course of radiotherapy in rectal cancer, we must consider other factors such as margins and volumes. In 2009, a Dutch study observed 28 patients for variation of the mesorectum prior to treatment during daily imaging by image-guided systems, CT, and Cone Beam CT (CBCT). The mesorectum can vary in a heterogeneous and anisotropic inter-fraction, which differs between women and men (25). Changes in organs at risk in prone and supine positions are scarce.
Special techniques of radiation therapy a. Intraoperative Electron Radiotherapy (IORT) Rectal cancer after curative surgery has a high local recurrence rate of 10-15% depending on the stage, surgical technique, and risk factors. Unresectable recurrence or palliative resection has a poor prognosis with a median survival of 7-12 months (26). Many patients will have breakthrough bleeding, ulcers, perineal or pelvic pain, ureteral obstruction, and secondary sepsis. Some patients develop distant disease and die of complications from local recurrence. If we study the topography of local recurrence observed in
234
Francisco López-Lara & Patricia Diezhandino García
the presacral space, the perineum is the predominant anatomical site of relapse: 67% presacral, perineal 13%, 4% perianastomotic or pelvic side wall, 0.5% in the posterior wall of the vagina, 2% in the prostate or bladder, and 14% in the anastomotic line (27). No equivalent detailed information on patterns of relapse in the pelvic area after preoperative RT-CT and radical surgery (28) have been published. Local recurrences usually occur with fixing organs or structures and if there was an option for surgery, it must be very radical or resection margins would be positive, so that surgery would become palliative. This situation coupled with the majority of patients having been previously submitted to EBRT, limits the possibility of resectability. EBRT as a unique tool get very good results in pain control from 70% to 100% in recurrences (29) but with a cure rate of less than 5% so that the patient will die within 2 years (30). EBRT dose has to exceed 60 Gy for microscopic residual disease and further to reach a tumouricidal dose to the gross tumour (31). However, the small intestine as the target organ of the administered dose limits risk to EBRT. Since both surgery and EBRT as rescue treatments for pelvic local recurrences have a major constraint, it appears the possibility of increasing the radiation dose and the capacity of drawing better the RT field with IORT. Thus IORT is indicated for locally advanced tumours to deliver a boost and in cases of tumour recurrences, where surgery cannot eradicate the tumour (32). IORT mainly benefits patients with very low margins of resection or with minimal residual disease. When in doubt, we can use the surgical margins or the IORT boost on small volumes by postoperative EBRT (33). Small institutions with IORT programmes reported high rates of local control in both adjuvant RT (34) and in preoperative RT-CT (35-37). IORT can be administered by two techniques: • •
IORT with electron accelerator Intraoperative High Dose Rate (IOHDR) brachytherapy.
Although the potential advantage of brachytherapy is the flexibility of the applicators that allow a more composed treatment, it is a lengthy procedure, and IORT is most often used. During IORT, after the surgeon removes the tumour, the radiation oncologist defines the area to be irradiated, and then the radiation physicist selects the prescribed dose and sets the depth ranging from few millimeters to 4 cm. The area of the tumour is irradiated, excluding risk organs and tissues. The main difficulties with the use of IORT are the problems associated with transporting the patient from the operating room to the radiation-therapy bunker. The two rooms are generally in different locations due to the physical risks to the patient and the time required for
Radiotherapy in colorectal cancer
235
anesthesia. Recent decades have seen the emergence of accelerators, designed exclusively to run IORT, the MEVATRON, MOBETRON, and NOVAC, that may be located in the operating room, thus avoiding transport of the patient. Patients are treated during surgery in a prone or supine position. The dose is calculated as the 90% isodose line depending on the characteristics of the disease (38): • • • •
close but negative margins: 7.5-12.5 Gy microscopic margin: 10-12.5 Gy macroscopic residual disease, 2 cm or less: 15 Gy macroscopic residual disease, 2 cm or more: 17.5-20 Gy
If the dose of the EBRT is limited by problems of the disease, the IORT dose is increased to achieve an equivalent dose of 50 Gy for resection with negative margins and of 60 Gy for patients with residual disease. EBRT vs conventional pre- or postoperative IORT could present some potential advantages such as: 1) Direct action in the tumour bed. 2) Exclusion in the irradiated field structures with less toxicity to healthy organs at risk 3) Combination with EBRT 4) The biological equivalent dose delivered to the IORT is 2-3 times higher than the same dose administered with conventional fractionation (39). Many studies have retrospectively compared the results of the EBRT and EBRT with IORT as boost with promising results (Table 1), although there are no randomised trials to provide standard instructions. Also, very few centers currently have the facilities for IORT. The University Hospital of Navarra has researched the long-term effects of IORT over 5 years of followup (40). Table 1. IORT studies compared with other techniques. Author/year Wong 1998 (41) Guiney 1997 (42) Mannaerts 2001 (43)
Treatment
n
EBRT 519 surgery + EBRT 39 146 EBRT only vs EBRT + surgery vs EBRT + surgery + IORT
Survival Survival 3 years 5 years 5% 9% 14% 11% 60%
local relapse 93% 82% 3 years 90% 3 years 86% 3 years 27%
DFS 2% 8% 0 43%
236
Francisco López-Lara & Patricia Diezhandino García
b. Brachytherapy Although endocavitary brachytherapy data from 1914 when using the endocavitary Radium was not until the time of Papillon (44) when he showed the utility as contact RT boost after EBRT or as exclusive treatment. High dose rate (HDR) brachytherapy is used as treatment in three situations (45): 1. T1N0 polypoid tumours: small, up to 3 cm, exophytic, with minimal penetration of rectal wall without lymph node and less than 10-12 cm from the anal margin 2. smaller T2 and T3 tumours: used as an overlay or after surgery or boost EBRT, with doses of 20-30 Gy 3. presacral recurrences: HDR brachytherapy has not been evaluated in phase III trials, but its use is justified in phase II trials or treatment protocols. Brachytherapy can be performed with permanent seed called Seed MiniMick and can be performed using intraoperative or perioperative high dose rates. The techniques and applicators may be of several types: -
Technical Papillon applicator and needles Implant technique with "fork" Technical Créteil Technical cylindrical applicator.
According to the studies shown in Table 2, brachytherapy obtained good results in OS, with excellent preservation of the anal sphincter and without major genitourinary or gastrointestinal toxicity. Maignon et al. (46) observed late rectal effects grade 3 in 3.8% and sphincter preservation in 82% of the included patients. Gerard observed no grade 3-4 toxicity in any of the patients, only acute rectal proctitis, which was not the cause of the discontinuation of treatment, and anal sphincter preservation was 92%. Centres with more experience are the French, (Léon Bérard in Lyon) with major retrospectives publications (49) and several placed in Spain such as ICO (Catalan Oncology Institute), University of Navarra, or Hospital Gregorio Maranon, but with few publications. Table 2. Studies of survival with brachytherapy. Author Papillón (47) Maignon (46) Gérard (48)
N 90 (T1-2) 151 (T1-3) 63 (T1-2)
OS at 5 years 77,8% 59,6% 65,4%
Sphincter preservation 95,7% 98% 92%
Radiotherapy in colorectal cancer
237
c. Photodynamic therapy Photodynamic therapy (PDT) involves systemic or topical administration of a photosensitising compound that preferentially accumulates in tumour tissues. The subsequent irradiation of these tissues with visible light causes the formation of highly reactive oxygen species, which are responsible for the selective destruction of tumour cells. PDT has a number of advantages over other local or systemic treatments of cancer such as: • • •
low systemic toxicity, because the photosensitiser is activated only in the presence of light ability to selectively destroy tumours, with a consequent reduction of side effects on other tissues, and may be given alone or in combination with other therapies such as chemotherapy, EBRT, immunotherapy, or surgery. The results of its application are many, from the delay in tumour growth in advanced cancers to a complete destruction of the tumour.
The study of the photodynamic reaction has advanced in the last decade, extending its spectrum of activity to other diseases such as inflammatory and infectious diseases. Approved PDT indications are actinic keratoses, superficial and nodular basal cell carcinomas, and Bowen's disease, but PDT has also been applied to cases of recurrence and colorectal, gastric, pancreatic, and other gastrointestinal tumours in the last few decades. It is used in CRC recurrences and anal tumours after failure of radical treatment, with good long-term results and in obstructions caused by recurrences (50). Allison et al. (51) used PDT in recurrence of anal cancer after radical treatment with EBRT and chemotherapy. Follow-ups 18-48 months later indicated good control of local disease and retention of sphincter function (52). All published studies are out of testing and protocols are performed within each center.
2. Adjuvant and neoadjuvant radiotherapy with or without chemotherapy EBRT has become an essential therapeutic tool in the treatment of cancer of the rectum both as a single modality and combined with chemotherapy. Concurrent use is supported by the evidence that chemotherapy acts as a radiosensitiser at the cellular level (2) so that both treatments synergise their activities for the eradication of residual, subclinical micrometastases after surgery with radical intent. Although surgery alone remains the mainstay of treatment, the rate of local recurrence is 25% with OS of 40-50% for stage
238
Francisco L贸pez-Lara & Patricia Diezhandino Garc铆a
T3-4 N1-2 (53), while surgery combined with EBRT-chemotherapy reduces local relapse to 10-15% and increases survival to 50-60%.
Early localized tumours Early tumours are neoplasms limited to the rectal wall (c/p T1 and T2). These tumours represent 3-5% of rectal cancers. Cases underwent standard surgery for an early tumour do not need further therapy. However, patients after a local surgical procedure are at risk for disease recurrence in the rectal wall or the local nodes. At stage T1, tumours without adverse pathologic factors have a low rate of local failure of approximately 5% and usually do not need adjuvant therapy but there is not an evidence to show that the outcome would be equivalent to radical surgery (54). Favourable histological (55) features include well or moderately differentiated tumours, absence of perineural and lymphovascular invasion, location of injury from 8 to 10 cm from the anal margin, with less than 4 cm and less than one third of the circumference of the rectum affected. On the contrary, pT1 with unfavourable pathological factors have to undergo a radical resection and if patients refuse surgery or if general conditions are compromised, adjuvant radiation therapy with or without chemotherapy could be considered. On the other hand, whenever adverse pathologic factors are present or the tumour invades into or through the muscularis propria (T2), the local failure rate increases to 17% and the incidence of positive nodes is about 10% (56). In all these cases local excision seems to be insufficient and a radical surgery is recommended, but in cases in which this procedure cannot be performed or it is refused, adjuvant therapy should be considered. Preoperative short-course EBRT in T2 operable tumours results in an even lower risk of local recurrence but it is not indicated since the absolute risk is very low providing a very high quality surgery has been performed (57). Cases with medically inoperable or who refuse surgery could receive preoperative radiation followed by local excision. It is usually administered with 5FU based concurrent chemotherapy. In cases of patients not fit for prolonged EBRT with or without chemotherapy can receive short course EBRT alone and delayed surgery. This management must be limited to only this subset of patients.
Intermediate stage tumours These stages include T3, T4, and/or N (+) resectable tumours. There are two approaches for these patients. One of them is initial radical surgery
Radiotherapy in colorectal cancer
239
followed by adjuvant combined treatment. The other is preoperative EBRT with or without chemotherapy followed by surgery and then adjuvant chemotherapy could be considered, although there is insufficient evidence on the benefit of adjuvant chemotherapy after preoperative chemoradiation (58). Exploratory analyses of this topic suggest that only patients who are downstaged from T3-4 to ypT0-2 benefit from 5FU based adjuvant chemotherapy. These data supports a significant survival benefit of 3-4% with 5FU based chemotherapy. Bolus and continuous infusion 5FU and capecitabine have been combined in several phase II studies with oxaliplatin or irinotecan plus preoperative EBRT and have also been delivered alone as adjuvant treatment after surgery. However, the compliance of 5FU/LV in the adjuvant setting in two large randomized phase II trials was suboptimal. We expect that ongoing phase III studies will help clarify the role of the combination of these drugs (59, 60). As a conclusion, the role of adjuvant treatment strategy after preoperative chemoradiation is still being investigated (61). a. Adjuvant and neoadjuvant treatments In general, radical surgery has a lower rate of local recurrence of 7%. However, adjuvant or neoadjuvant treatments are needed for these stages to better control the disease (62). The characteristics of rectal cancer with a high rate of local recurrence and evidence of the palliative effect of EBRT in unresectable tumours led to studies with postoperative EBRT. Most of these studies were noncomparative, or compared modern with historical groups, and showed reductions in the number of relapses (<10%) and improvements in survival in stage II, reaching 65-75% at 5 years (63,64). In 1979, Ghossein et al. (65) demonstrated that postoperative EBRT decreased the rate of local recurrence in patients with intermediate stages rectal cancer (T3-4 N0 or Tx N 1-2). Since then, many studies in this field have addressed the benefits of this treatment (66, 67). A study at the MD Anderson Cancer Center published in 1977 included 62 patients treated with a EBRT dose of 50 Gy 3-6 weeks after surgery without total mesorectal excision. A local control of 92% and a disease-free survival of 79% at 48 months were achieved, but at the cost of significant gastrointestinal toxicity (68). A randomised study conducted at the Mayo Clinic on patients with these intermediate stages of rectal cancer or unresectable local recurrence showed that the combination of EBRT and 5FU improved symptoms control rate and duration and increased survival and the number of responses.
240
Francisco López-Lara & Patricia Diezhandino García
All the studies from this period included few patients and short followups. Therefore, subsequent randomised studies were needed to assess the effectiveness of adjuvant treatment. A Dutch multicenter prospective trial published by Treurniet-Donker (69) et al. studied 172 patients randomised to two arms. The experimental arm consisted of an EBRT dose of 50 Gy in five weeks (standard fractionation of two Gy per fraction), while the control group received no adjuvant treatment after surgery. The adjuvant EBRT showed a decrease in the rate of local recurrence but without statistical significance (24% vs 33%, p < 0.1) or an increase in disease-free or OS. These results were similar to those obtained in other studies in Europe and USA; then postoperative EBRT alone was accepted as routine treatment after surgery in resectable rectal cancer. a.1. Preoperative radiation therapy alone A study by the Medical Research Council Rectal Cancer Working Party compared a control arm with surgery alone to the experimental arm with surgery preceded by EBRT dose of 40 Gy with standard fractionation. The study included 479 patients with recruitment between 1981 and 1989 from over 20 regional centres in the UK (70). The patients were followed up for a minimum of 5 years or to death. 217 patients died, 114 of 140 allocated surgery alone and 103 of 139 allocated preoperative EBRT with median survival times of 24 months and 31 months, respectively. The hazard ratio (HR) for OS was 0·79 (95% CI 0·60—1·04, p=0·10). At 5 years' follow-up 65 patients allocated surgery alone and 50 who received preoperative EBRT had local relapse (HR 0·68 [0·47—0·98], p=0·04); the corresponding numbers of patients with distant recurrence were 67 and 49 (HR 0·66 [0·46-0·95], p=0·02). There was a significant benefit of EBRT on DFS (HR 0·76 [0·58-1·0], p=0·05). There was no increase in postoperative or late complications in the EBRT group. The authors interpreted these results as providing further evidence that preoperative EBRT reduces the rate of local recurrence of rectal cancer in patients with locally advanced disease. The results concerning survival, however, are still equivocal, and we must await the results of a meta-analysis of all trials that may provide definitive results, particularly for rates of survival. a.2. Preoperative versus adjuvant EBRT A systematic overview (71) of more than 8000 patients from 28 randomised, controlled trials compared the outcomes of surgery for rectal
Radiotherapy in colorectal cancer
241
cancer combined with preoperative or postoperative EBRT with those of surgery alone. Data from individual patients were analysed from 22 randomised comparisons between preoperative (6350 patients in 14 trials) or postoperative (2157 in eight trials) EBRT and no EBRT. OS in patients receiving EBRT was only marginally better than in those allocated to surgery alone (62% vs 63% mortality, p = 0.06), and preoperative EBRT failed to improve rates of curative resection (85% EBRT vs 86% control). The yearly risk of local recurrence was 46% lower in patients receiving preoperative EBRT, and 37% lower in patients receiving postoperative EBRT, compared to those who had surgery alone (p = 0.00001 and p = 0.002, respectively. Fewer patients who had preoperative EBRT died from rectal cancer than did those who had surgery alone (45% vs 50%, respectively, p=0.0003), but early (</=1 year after treatment) deaths from other causes increased (8% vs 4% died, p<0.0001). The authors concluded that preoperative EBRT (at biologically effective doses â&#x2030;Ľ 30 Gy) reduced the risk of local recurrence and death from rectal cancer. Postoperative EBRT also reduces local recurrence, but short schedules appeared to be at least as effective as longer ones. If the safety of preoperative EBRT could be improved without compromising effectiveness, then OS would moderately improve, especially for young, high-risk patients. a.3. Combined treatments Concomitant treatments involve the administration of systemic drugs producing cellular radiosensitisation that increases the destructive effect of radiation therapy. The most commonly used radiosensitisers in rectal cancer include intravenous fluoropyrimidines such as 5-fluorouracil (5-FU), oral fluoropyrimidines such as capecitabine and UFT, and antifolates such as raltitrexed. Oral fluoropyrimidines are an alternative to 5-FU. Pharmacokinetic studies have found no differences between oral and intravenous fluoropyrimidine 5-FU (72). Oral treatment during the course of irradiation is more comfortable for the patient than the continuous intravenous administration of 5-FU combined with EBRT. The Fisher Group trial (73), National Surgical Breast and Bowel Project R-01 (NSABP R-01), evaluated 555 patients treated by curative resection between 1977 and 1986. Patients were randomised to three arms, one with surgery alone compared to a second arm with surgery followed by postoperative EBRT, and a third arm with surgery followed by postoperative chemotherapy with 5-FU, vincristine, and semustine (MOF). The EBRT dose was 46-47 Gy in 26-27 fractions, with a later overlay in the perineal area of 51-53 Gy.
242
Francisco López-Lara & Patricia Diezhandino García
Local relapse decreased in the adjuvant EBRT group compared to the surgery-alone control group (25% vs 16%, p = 0.06). DFS (p = 0.4) or OS (p ═ 0.7) did not differ significantly with the use of EBRT. The chemotherapy group compared to the control group, though, showed a statistically significant improvement in OS (p = 0.05) and DFS (p = 0.006). When evaluated according to sex, the benefit for chemotherapy at 5 years, both in DFS (29% vs. 47%, p <.001; relative odds, 2.00) and in OS (37% vs. 60%, p 0.001, relative odds, 1.93), was restricted to males. When males were tested for age trend with the use of a logistic regression analysis, chemotherapy was to be more advantageous found in young patients. The global test for interaction to identify heterogeneity of response to radiation within subsets of patients was not significant. This investigation has demonstrated the benefit of adjuvant chemotherapy (MOF) for the management of rectal cancer. The observed advantage was restricted to males but postoperative EBRT reduced the incidence of local and regional recurrence but failed to affect OS and DFS. The North Central Cancer Treatment Group (NCCTG) and the Mayo Clinic (74) conducted a trial comparing adjuvant EBRT at a dose of 45- 50.4 Gy after surgery with a combination of adjuvant EBRT and chemotherapy with 5-FU preceded and followed by a cycle of systemic treatment with 5-FU and semustine (methyl-CCNU). The 204 patients in the combined-therapy arm demonstrated superior results. The relative reduction in recurrence was 34% (the 95% confidence interval was 12-50%). Combined therapy significantly reduced the length of time to recurrence (p = 0.0025) and reduced the death rate by 29% (p = 0.043, the 95% confidence interval was 7-45%). The combination of postoperative local radiation and systemic therapy with a fluorouracil-based regimen substantively improved treatment of rectal carcinomas having a poor prognosis, as compared to postoperative EBRT alone. In a trial by the Gastrointestinal Tumour Study Group (GITSG) (75) in1975, 202 patients at stages B2 and C were randomised to 4 arms, (1) surgery alone, (2) Adjuvant chemotherapy with 5-FU, (3) EBRT (40-48 Gy split course, or two courses), and (4) Chemotherapy and adjuvant EBRT (40-44 Gy). This study was criticised for its small number of patients and the use of interrupted EBRT. After nine years, the combination of EBRT and chemotherapy provided an increased OS of 54% compared to 27% in the surgery-alone arm. The combined arm had a longer time to recurrence, and the local recurrence decreased compared with surgery alone, 11% vs 24% respectively, and distance recurrence was also improved 26% vs 34%.
Radiotherapy in colorectal cancer
243
This study thus concluded that the combined-arm increased OS almost double than in surgery alone arm. Following the publication of these two trials, the American National Institutes of Health (NIH) recommended, in 1990, the use of adjuvant treatments after surgery for rectal cancer in stages II and III (76). Optimisation of chemotherapy The following studies were conducted to optimise the different patterns of chemotherapy. One study showed that 5-FU was better than methyl-CCNU being this more toxic and without any kind of benefit in local recurrence and OS. The administration of continuously infused 5-FU during EBRT was better than a bolus, with a risk of distant metastases of 31% vs. 41% and DFS at 4 years of 70% vs. 60%, although the rates of local recurrence did not differ. One of the most significant studies was that of Intergroup INT 0144 (77) that evaluated the benefits of continuous infusion during EBRT and then after completion. After resection of T3-4, N0, M0 or T1-4, N1, 2 M0 rectal adenocarcinomas, 1,917 patients were randomly assigned to arm 1, with bolus 5-FU in two 5-day cycles every 28 days before and after radiotherapy plus 5-FU via protracted venous infusion (PVI) (dose of 225 mg/m2/d during EBRT; arm 2 (PVI-only arm), with PVI 42 days before and 56 days after EBRT + PVI; or arm 3 (bolus-only arm), with bolus5-FU + LV in two 5-day cycles before and after EBRT, plus bolus 5-FU + LV (levamisole was administered each cycle before and after EBRT). Differences in DFS, local recurrence, or OS were not observed after a median follow-up of 5 years, and hematologic toxicity was lower in the group treated with 5-FU in continuous infusion. Despite these results, continuous infusion is considered effective and is the most widely used scheme in many institutions. The optimal time for administering EBRT, however, is not well defined and remains an important question. Most studies start after one or two cycles of chemotherapy. One study randomised 308 patients who received EBRT after the first or after the third cycle of adjuvant chemotherapy (5-FU/LV). DFS after four years was better in the patients who received early EBRT (81% vs 70%, p = 0.043), but OS in the two groups was essentially the same (84% vs 82%, p = 0.387). The group receiving EBRT after the first cycle had 23 recurrences (78), compared to 38 recurrences in the group receiving EBRT after the third cycle. Although no consensus exists, postoperative EBRT should start 3-6 weeks after surgery according to institutional protocols.
244
Francisco L贸pez-Lara & Patricia Diezhandino Garc铆a
T3N0 favourable cases Several studies show that a subgroup of patients do not need adjuvant treatment for the low risk of recurrence after surgery. In 1999, researchers at the Memorial Sloan-Kettering Cancer Center (MSKCC) (79) evaluated 95 patients with T3N0 rectal adenocarcinoma treated with surgery alone. The aim was to investigate whether favourable T3N0 cases could forgo additional treatment. Seventy-nine patients underwent low anterior resection, ten to coloanal anastomosis, and 6 to abdomino-perineal resections with total mesorectum excision (TME). At a follow-up after 53.5 months, 6% had local recurrence, distant metastasis 13% and 3% local and distant recurrence. Of the risk factors analysed, the presence of lymphovascular invasion was the most important histological factor for local recurrence. No other factors such as surgical technique, type of resection, tumour location, or extent of the margin were important in determining recurrence. This study showed that the TME anterior resection of abdominoperineal T3N0 tumours result in low rates of local tumour recurrence (<10%) without adjuvant therapies. Other groups have supported this result, but even if a limited subset of T3N0 patients get excellent results (80) with surgery alone, no randomised study has yet supported the omission of adjuvant therapies in these patients. a.4. Neoadjuvant radiation therapy Preoperative, or neoadjuvant, treatments have some important advantages over postoperative, or adjuvant, treatments. They have the potential to reduce the tumour to facilitate surgery, reduce the risk of tumour spread eliminating micrometastases, and reduce the problems of hypoxia that increase after surgery. The patient is in good physical condition, and treatments are usually administered at full dose with a better tolerance reducing the toxicity of treatment and tumour seeding that occurs during surgery. The administration of concurrent preoperative EBRT and chemotherapy increases efficiency providing a better resectability and therefore allowing the performance of sphincter-sparing surgery (81). Despite these advantages, neoadjuvant treatment also has a number of drawbacks. One of the main problems is the delay in surgical treatment, because patients must wait one month after completion of EBRT, and the side effects of EBRT increase the difficulties of surgery (81). The occurrence of postoperative perineal infections, delayed healing, and the overtreatment of patients with early-stage (T1-T2 N0) or disseminated tumours at diagnosis, despite the use of high-definition imaging techniques, are the main
Radiotherapy in colorectal cancer
245
disadvantages of preoperative EBRT (82,83). The selection of the best treatment strategy is based on a good initial staging. In this context diagnostic tests such as MRI, helical Computed Tomography, endorectal ultrasound, and PET could help to decide which cases could benefit more with neoadjuvant therapy than with adjuvant. Preoperative EBRT has a number of physical and biological advantages compared to postoperative EBRT. Preoperative EBRT is less toxic, and the absence of adhesions or other complications secondary to pelvic surgery is beneficial for the small intestine, which is very radiosensitive (84). Tumour cells are well oxygenated and so are more radiosensitive. A decrease in tumour mass and therefore a decrease in the resected cancer stage (downstaging) allows an increased curative resection at a lower cost compared to postoperative EBRT. Preoperative EBRT and first treatment are about six times less costly, are easier to complete, and are less toxic than equivalent doses after surgery (85). The first randomised studies with preoperative EBRT were performed with the aim of making unresectable lesions resectable. In 1914, Symonds (86) reported the first use of preoperative EBRT as a treatment for rectal cancer. Symonds used radium in a patient with a tumour in a low rectal location, with good results three months after surgery. Several studies have used moderate-dose preoperative EBRT, observing a consistent local control with or without a minimal increase in OS (87, 88). In 1960, a study was conducted at MD Anderson Cancer Center (92) with 126 patients, 52 of whom received only EBRT with a fractionation of two Gy per session and a total of 50 Gy prior to surgery for rectosigmoid tumours, resectable or unresectable. The objective was to evaluate the tolerance of patients to preoperative EBRT. 50% (8 of 16) unresectable lesions were operated. No tumour was found in pathology samples in 10 of 52 patients, and preoperative EBRT was well tolerated with few complications after treatment. Later clinical trials of the Medical Research Council among others, showed a significant decrease in local recurrence after preoperative EBRT (89). Rouanet et al. (90) administered 40 Gy preoperatively to 27 T2-T3 patients who were evaluated three weeks later. When the response was good (30%), patients received an additional 20 Gy, achieving anterior resection in 63% of patients and transanal local excision in four patients (15%). In all, 78% of the patients underwent sphincter-preserving surgery. In a later study by Mohiuddin et al (91)., 70 patients with tumours 2 cm from the anal margin received 40-45 Gy at 1.8-2 Gy/session, and the worst cases continued with a boost of 10-15 Gy. Surgery was performed at 5-10 weeks, and the results were very promising: 48 patients received radical resection, and 22 received full thickness local excision. At a mean follow-up of four years, 60 patients (86%)
246
Francisco L贸pez-Lara & Patricia Diezhandino Garc铆a
had succeeded in maintaining sphincter function, with local failure in nine patients (13%) and distant metastases in 12 (17%) without presenting gastrointestinal toxicity grade III- IV. These data led to many studies designed with preoperative EBRT fractionation schemes and standards, but in 1960, the MSKCC (93) designed a study that was administered in a hypofractionated EBRT, i.e. the dose per fraction increased, with the intention of reducing treatment time and preventing a delay to surgery. The trial was randomised to evaluate the usefulness of preoperative, low-dose EBRT (to 20 Gy in 8 fractions of 2.5 Gy) for operable tumours and was aimed at increasing OS. The study included 800 patients and found that the EBRT did not increase OS in patients with tumours or those who were node-positive or negative. This study was criticised because a large number of the non randomised patients were included and they received a very heterogeneous radiobiological doses and fractionations. Thus, with these methodological flaws, this study showed no advantages of preoperative EBRT for local control or survival compared to surgery alone. Randomised trials have subsequently shown benefits of preoperative EBRT. Cedermark et al., in a series of 847 patients (94), obtained an increase in local control by decreasing the rate of recurrence using a preoperative dose of 25 Gy in 5-7 days (hypofractionation). The EBRT arm, however, experienced more toxicity than the group that only underwent surgery. Other randomised trials have reported a significant reduction in local recurrence. Such a reduction, compared to surgery alone (p < 0.05), was found by a group of the Imperial Cancer Fund (1) in a randomised set of 468 patients receiving preoperative EBRT doses of 5 Gy in 3 sessions. The European Organisation for Research and Treatment of Cancer (EORTC) (2), and others, later showed a trend towards increased OS after preoperative EBRT. The most consistent data, however, were obtained in the Swedish Rectal Cancer Trial (1) that randomised 1168 patients to receive preoperative EBRT hypofractionated into doses of 25 Gy in 5 sessions or to receive surgery alone. Both local control (89% vs 73%) and OS (58% vs 48%, p < 0.05) increased with preoperative EBRT. A Dutch trial (2) with 1861 patients failed to demonstrate improvement in local control with the best surgery (TME). Patients received high-dose EBRT followed by surgery or surgery alone with TME in the two arms. Preoperative EBRT reduced the rate of local recurrence at 2 years (2.4% vs 8.2%) but failed to increase OS (82% in both arms). Patients in the EBRT arm also experienced problems such as poor healing of wounds after abdominal-perineal surgery, sexual dysfunction, and fecal incontinence (p = 0.008). The rates of local relapse after 6 years were 10.9% vs 5.6% with statistical significance (p < 0.001) favoring the preoperative EBRT group, with equivalent survival rates in both groups.
Radiotherapy in colorectal cancer
247
Two meta-analyses of approximately 6000 patients each explored the benefit of preoperative EBRT. One analysis included 14 randomised, controlled trials and found that neoadjuvant EBRT was associated with significantly fewer local recurrences, improved version-specific survival, and improved OS. The second meta-analysis, provided by the CRC Collaborative Group, included 22 randomised, controlled trials and concluded that preoperative EBRT significantly reduced the risk of local recurrence and death from rectal cancer (3). a.5. Neoadjuvant combined treatments Two fundamental strategies were developed to improve the results of preoperative EBRT: increasing intraoperative doses and adding chemotherapy to EBRT to increase the radiosensitivity of the tissues. The second of these strategies was tested in the EORTC 22921 (1) trial that randomised 1011 patients receiving preoperative EBRT doses of 45 Gy with standard fractionation or receiving hypofractionated EBRT with preoperative chemotherapy (5-FU/LV). Total excision of the mesorectum was performed in 40% of the patients. The rates of pathological complete response (pCR) were 13.7 % vs 5.3 % for the combined arm. The sizes of tumours decreased, but survival at 5 years and the rate of preservation of the anal sphincter in the two arms remained unchanged. The FFCD 9203 trial by the Fédération Francophone de la Canérologie Digestive obtained similar results. This study included 773 patients with T3-4 Nx M0 in the distal rectal region randomised to receive EBRT or EBRT-chemotherapy preoperatively. The EBRT group received 45 Gy with standard fractionation, and the EBRT-chemotherapy group received concomitant chemotherapy (5-FU/LV). The patients subsequently underwent surgery. The primary endpoint was OS. No differences between the groups were seen in sphincter preservation, but the group with preoperative EBRT-chemotherapy had a higher rate of pCR (11.4% vs 3.6%, p < 0.05) and a lower rate of local recurrence (8.1% vs 16.5%, p < 0.05) compared to the group with preoperative EBRT. No differences in OS between groups were observed (2). a.6. Neoadjuvant versus adjuvant combined treatments To determine which of the two therapies was better, three phase III trials were designed that compared EBRT and chemotherapy prior to or after surgery. The first trial (RTOG 94-01/intergroup (INT) 0417) compared 5-FU/LV + preoperative EBRT (50.4 Gy) versus the same pattern after
248
Francisco L贸pez-Lara & Patricia Diezhandino Garc铆a
surgery. This study included 53 patients and was closed for poor recruitment of patients. The second trial (NSABP-RO3) (95) had a design similar to the INT 0147 trial and compared the administration of EBRT (45 Gy) + CT (5-FU/LV) before or after surgery. The trial began in 1993 and recruited 267 of the 900 patients originally planned (due to low enrollment). The objectives of the study were to monitor sphincter preservation rate and both DFS and OS. Few data were published, but the study showed a similar toxicity in both groups, and no patients progressed during preoperative treatment. Sphincter preservation was higher in the preoperative treatment group (50% vs 33%), which had an 8% pCR. No significant differences in disease-free survival at 3 years (70% vs 65%) or OS (85% vs 78%) were observed. A subsequent analysis reported that patients who had achieved a pCR showed an OS at 3 years of 100%, while those with partial response had 95% OS and in cases with stable disease OS was of 83% (96). The third and final study (CAO/ARO/AIO-94 by the German Rectal Cancer Group) (97) in favour of preoperative EBRT randomised 823 patients with stage T3-4 or a nodepositive diagnosis to receive EBRT (50.4 Gy at 1.8 Gy per fraction) + 5-FU in continuous infusion (weeks 1 and 5). In all cases of resection of the mesorectum, surgery was followed by adjuvant chemotherapy. The postoperative group received a boost of 5.4 Gy. In a follow-up after 5 years, neoadjuvant therapy was associated with a lower rate of local recurrence (6% vs 13%). A significant downstaging after pretreatment was seen with a pCR of 8%. Disease-free or OS did not differ between groups. Both acute and chronic toxicities were lower in patients receiving the neoadjuvant. The results of these randomised studies show that the scheme of preoperative EBRT-chemotherapy for locally advanced rectal cancer is well tolerated, with no contribution to surgical mortality and morbidity, disregarding technical difficulties after surgery. Preoperative EBRT-chemotherapy (or even EBRT alone are able to decrease the rate of recurrence (98). Together with meticulous surgical techniques, preoperative EBRT-chemotherapy today is improving the riskbenefit of combination therapy for locally advanced carcinoma. Multidisciplinary approach can achieve good oncologic results at an equally good QOL. a.7. Neoadjuvant and adjuvant radiation therapy This technique, called the "Sandwich technique", has the advantage of a preoperative low dose (5-15 Gy) followed immediately by surgery, preventing the cells from spreading. A full-dose EBRT is given after surgery, as high as
Radiotherapy in colorectal cancer
249
45-50 Gy in the worst cases. The RTOG (Radiation Therapy Oncology Group) (1) trial administered 5 Gy preoperatively to 350 patients and then an adjuvant 45 Gy after surgery for T3 or N-positive patients. No therapeutic benefit was demonstrated as there were no differences in local or distant failure.
3. Palliative radiation therapy Locally advanced and/or unresectable stage tumours represent very heterogeneous diseases, and the definition of resectability differs among schools and authors. From the surgical point of view, resectability is when a tumour can be resected without leaving residual tumour material, either microscopic or macroscopic. In this context, we find that different clinical and multidisciplinary approaches are essential to the patient. Locally advanced cases are rare and have a poor prognosis for survival and QOL of the patients. After completing a perfect staging, the indication is to perform preoperative chemotherapy + EBRT at a dose of 45-50 Gy with concurrent chemotherapy schemes, provided that the patient can tolerate them. Responses can be achieved in 50-70% of patients with pCR rates of 10-30%, depending on the series, with increased gastrointestinal toxicity. Four to six weeks after surgery, the type of resection needed will depend on the location and extent of the disease (1). Once excised, the tissue is examined, and the surgical field must be examined to identify areas where residual disease and the presence of microscopic or macroscopic positive margins may occur. In cases in which residual disease is possible and the technology is available, IORT at doses of 10-20 Gy could be administered over the area of risk of relapse, to improve local control. Subsequent chemotherapy (4-6 cycles) gives significant rates of control of local disease. For patients with only local recurrence after a radical treatment with preoperative EBRT + chemotherapy, total or posterior pelvic exenteration can produce prolonged disease-free survival. IORT for patients with locally recurrent disease who previously received RT may improve local control with acceptable levels of morbidity. The symptoms that accompany both locally advanced colorectal and presacral recurrences are varied: bleeding, pain, infiltration of the sacral plexus, bone destruction, intestinal obstruction, etc. Metastases of colorectal origin occur, in order of frequency, in the liver, lung, and bone and also display the symptoms described above. Surgical resection of lung and liver metastases can be performed in selected patients with survival to five years. Palliative treatments improved symptoms in 70% (5) of patients with bone metastases. RT has an analgesic effect widely known since the 1960s. Pain is relieved in 70% of cases and disappears in 20-60%, beginning 3-10 days after treatment, depending on the fractionation used (99).
250
Francisco L贸pez-Lara & Patricia Diezhandino Garc铆a
The treatment of bone metastases has no standard time frames or dose rates. Many schemes have been analysed with different doses and fractionations, number of locations, onset of symptoms, histology of the primary tumour, and volumes of irradiation. None have shown significant differences in strength and response to pain (1). A meta-analysis in 2003 and the Cochrane review (Sze et al.) (100) reported no differences between single and standard subdivisions in response to pain (2,3). For liver metastases many therapeutic options have been studied and described, and all are palliative. The liver palliative RT is a great unknown in the palliation of symptoms. But some studies such as RTOG, in his study 76/05, communicate schemes with total doses of 20 to 30 Gy in 7 to 16 fractions over the entire liver, with or without superimposition of single lesions (20 Gy). The rates of improvement of signs and symptoms evaluated after four weeks of EBRT were: 55% for abdominal pain (1), 49% for nausea and vomiting, 45% for sweating and fever, 49% for ascites, and 39%for alkaline phosphatase levels (2). We must not forget, however, that the average survival of patients undergoing EBRT for liver metastases is only 4 months (3).
4. Reirradiation Irradiation in patients treated for rectal cancer is a major problem because of the risk of damage to organs involved in the treatment and the toxicity that occurs with EBRT and chemotherapy. This was long the thinking of many radiation oncologists, but recent years have seen a change in the management of irradiated patients. Oncologists should consider the previous radiotherapy, dose, fractionation, technique, and length of interval between treatments, the first of which is sometimes enough to repair cell damage. Even though we have different techniques, we should use for reirradiation cautiously and be conscious of possible complications to the patient. Different techniques include: 3D radiotherapy, IMRT, stereotactic RT, IORT, particle and proton therapy, and brachytherapy with both high and low rates. All these techniques are in clinical trials or treatment protocols. Studies, reviewed by Mohiuddin et al. (91), have shown good long-term outcomes in patients with local pelvic recurrence. In one study, 103 patients with presacral recurrence received a combined treatment with chemotherapy and EBRT (50.4 Gy) and subsequent surgery, then EBRT (reirradiation) with chemotherapy (5-FU continuous infusion). Patients were irradiated with doses of 30 Gy (1.2 Gy Twice a day) or 30.6 Gy (1.8 Gy Every Day) Followed by a boost of 6 to -20 Gy to gross tumour volume with a 2 cm margin) with Opposed laterals or a three-field technique with a posterior field and two laterals to the presacral area and gross tumour volume with 2 - to 4-cm
Radiotherapy in colorectal cancer
251
margin. Forty-one patients were subsequently explored surgically, 34 of whom received surgery that preserved the anal sphincter. OS at 5 years was 19%. Patients receiving surgery had better survival with acceptable acute and late toxicities. The palliation of symptoms such as bleeding was achieved in 100% of patients. Another study, by Haddock et al. from the Mayo Clinic and Mayo Medical School (101), highlighted a series of 51 previously irradiated patients who underwent IORT with a mean reirradiation dose of 20 Gy. The results revealed a tendency to local control but a short OS time due to the development of distant metastases. Peripheral neuropathy and ureteral obstruction were among the toxicities seen with IORT. Glimelius et al (102)., at the Uppsala University Hospital, reported that patients who suffered recurrence after receiving pelvic EBRT at a dose of 50 Gy could be treated with doses of at least 30 Gy with the new technologies described above. All strategies that succeed in increasing the dose in recurrences can bring shortterm benefits but do not change survival. All treatments with reirradiation should be considered when treatment options are minimal.
5. Toxicity of radiation therapy The toxicity of EBRT is of two types: acute and chronic. According to the criteria of the RTOG and the National Cancer Institute published in the Common Toxicity Criteria (CTC) is considered acute toxic to all the changes that occur during treatment and 90 days and chronic or delayed toxicity complications that occur after the 3 months to years after EBRT (103). The acute toxicity of EBRT in rectal cancer appears as intestinal (2) and genitourinary symptoms as are the parts that are included in the treatment volume. Symptoms of acute gastrointestinal toxicity (80-90% of patients) include: nausea and vomiting, fullness, anorexia, and fatigue, all of which may appear from the first session of EBRT; proctocolitis: increased bowel movements, diarrhea, rectal urgency, tenesmus, and rectal bleeding; and radiation enteritis, which appears in the second to third week of EBRT and consists of watery diarrhea, frequent bulky stools, and abdominal cramping. Radiation enteritis usually subsides within 2-3 weeks after the completion of EBRT (1). Symptoms of chronic gastrointestinal toxicity (60-90%) (104) include: complications such as intestinal obstruction that is often preceded by acute colitis; drilling with acute abdominal pain, rectal bleeding, altered bowel ulceration, thickening of folds, narrowing of bowel segments, and mesenteric adhesions that cause malabsorption; chronic proctitis with rectal urgency, abdominal pain, mucous discharge, and bleeding; and ulceration, fistulation, and rectal stenosis, which is more common in patients who
252
Francisco López-Lara & Patricia Diezhandino García
received vaginal brachytherapy. Symptoms of acute genitourinary toxicity (80-90%) include acute cystitis from bladder toxicity, clinically characterised by dysuria, and increased urinary frequency and urgency. Symptoms of chronic genitourinary toxicity (10-20%) usually appear within 12-20 months of EBRT and include reduced bladder capacity, hematuria by telangiectasias, hemorrhagic cystitis, chronic irritation, or the formation of vesicovaginal fistulas. All these symptoms depend on the total dose, fractionation used, and above all, the treatment volume and planning. Hypertension, diabetes, previous abdominal surgery, and pelvic inflammatory disease have been described as predisposing factors for the development of radiation-induced intestinal toxicity. Diarrhea (1-3) is the most common symptom of acute toxicity in these patients, which is increased by the concomitant administration of chemotherapy (5-FU and capecitabine) (105-107).
Fisiopathology of radiation toxicity The pathophysiology of toxicity begins early in EBRT, with an inhibition of mitosis, depletion of intestinal mucosal cells with high proliferative capacity, degenerative changes, and necrosis of the deep cells of the intestinal glands (crypts of Lieberkühn). Capillary conjunctival hyperemia and swelling develop later. Congestion of the vessels appears and hyalinisation begins. Mitosis resumes but starts late and programmed cell death (apoptosis) occurs. All these effects result in malabsorption of fats, carbohydrates, proteins, and bile salts. The intestinal mucosae usually recover within 1-3 months after completion of EBRT. In the connective tissues of the mucous membranes, submucosal and subserosal thickening of vascular adventitia appears and mucohialino tissue with the formation of collagen. Symptoms during the late phase of EBRT are more serious and irreversible. EBRT can cause atrophy of the mucosa and thickening and proliferation of endothelial lipid deposition, and endoarteritis obliterans with anoxia and subsequent necrosis may occur later (108). Many of these symptoms can be controlled with a low-fat diet, iron supplements, and vitamins B12, A, and D, but sometimes surgery is necessary, which is complicated by extensive multiple injuries. The distinction between tumour recurrence and radiation-induced toxicity is difficult to discern, and in some cases both can coexist. Prevention is the most important measure and avoid treating of the small intestine is a priority. Proper treatment, especially the choice of technique, is essential to minimise side effects.
Radiotherapy in colorectal cancer
253
Emami (109) tolerance doses to the small intestine and bladder are specified in Table 3, with the volumes of organ at risk receiving more than the written dose are the following: Volume rectal V72 <30%, V60 <50%, V40 <60%. In the case of the bladder, V70 <10%, V60 <40% (see Table 3 to better understanding). Table 3. Normal tissue tolerance to therapeutic irradiation. TD5/5 volume
TD5/50 volume
Organ
1/3
2/3
3/3
1/3
2/3
3/3
complications
Small intestine
50Gy ---
40Gy
60Gy ---
45Gy
Obstruction/perforation fistula
Colon
55Gy ---
45Gy
65Gy ---
55Gy
Obstruction/perforation
Bladder
----
80Gy 65Gy
---
85Gy 80Gy
Symptomatic bladder contractions and volume loss
Kidney
50Gy 30Gy 23Gy
---
40Gy 28Gy
Nephritis
TD 5/5: tolerance dose of 5 years, 5% complications
For the assessment of possible acute and late complications, various scales have been developed to standardise terms for the various symptoms/signs in EBRT: •
•
•
CTC (Common Toxicity Criteria) Scale: In 1982, the National Cancer Institute developed criteria for assessing the toxicity of treatment with chemotherapy, for use in clinical trials. The last scale was published in December, 2003 (CTCAE version 3.0 (commom Terminology Criteria for Adverse Events)) and is used for both chemotherapy and EBRT (110). RTOG/EORTC (Radiation Therapy Oncology Group/European Organisation for Research and Treatment of Cancer): In 1983, these two organisations differentiated between acute and chronic toxicity (103). This scale assesses organ toxicity and is commonly used in used in clinical practice in Radiation Oncology (111,112). SOMA/LENT: In 1995, the consensus conference of late effects in normal tissues introduced a new scale for late toxicity, also developed by the RTOG and EORTC, called SOMA/LENT (subjectivism, objectivism, Management, and Analytic/The Late Effects on Normal tissue) (113). The main objective of this scale was to create a classification system of late toxicity to vital organs as a result of different oncological treatments. This
254
Francisco L贸pez-Lara & Patricia Diezhandino Garc铆a
scale is used for each organ included in the radiation field, and consists of four aspects: 1. 2. 3. 4.
Subjective: description of symptoms Objective: signs, such as edema or weight loss, can be detected on physical examination Management: possible treatment and reversibility of toxicity Analysis: special laboratory tests or imaging techniques (Computed tomography, MRI), and measurable procedures.
In general, toxicity is lower in groups receiving preoperative EBRT + chemotherapy, compared to those receiving postoperative EBRT + chemotherapy. Minsky et al. (56) compared the toxicity of EBRT + chemotherapy applied preoperatively and postoperatively and found less grade 3-4 toxicity in preoperative treatment (13% vs 48%, p = 0.045). Two patients had grade 3 gastrointestinal toxicity in the preoperative group and postoperative group had toxicities grade 3-4 gastrointestinal and 2 genitourinary in 7 and 2 patients respectively. In the NSABP study (95), toxicity was similar in both treatments. Diarrhea was worse in the preoperative group, and complications of surgery were related to the type of surgery. Frequent complications of anastomosis arose in the preoperative group, and the postoperative group experienced perineal complications. In the CAO/ARO/AIO-94 study (114), rates of grade 3-4 gastrointestinal toxicity were significantly lower in the preoperative treatment group (acute: 12% vs 18%, p = 0.04; late: 9% vs 15%, p = 0.07). With chemotherapy, the incidence of diarrhea is higher for continuous infusion of 5-FU than for bolus 5-FU. Hand/foot syndrome is also more common in infusional therapy (1). Late complications are less frequent but more serious. Late symptoms typically appear 6-18 months after the end of EBRT and frequently involve the intestine: persistent diarrhea, proctitis, and obstructive symptoms and/or intestinal perforation. Urinary symptoms such as incontinence and cystitis can appear but are less common. The incidence of intestinal obstruction in patients who received postoperative pelvic EBRT and surgery is 4-12%, and some authors found no differences in the frequency of occurrence of this complication when compared with patients who did not receive EBRT (1). A very debilitating late side effect of postoperative EBRT is described in patients undergoing low anterior resection: rectal urgency with very frequent bowel movements. These patients usually have 10 or more bowel movements a day, increasing at night, and are occasionally incontinent. These symptoms are involved in the development of anastomotic stricture
Radiotherapy in colorectal cancer
255
and fibrosis neorectum (115). Several studies have attempted to minimise testing the effects of pelvic EBRT using Octreotide (a somatostatin analogue), but results were poor: some symptoms worsened in the group treated with Octreotide (116). In general, data from randomised EBRT + chemotherapy trials indicate that preoperative therapy is less toxic than postoperative therapy (7% vs 14%) (1). In studies of postoperative combination treatment, severe acute toxicity vary between 24-40% (2,3), however, the numbers of acute grade 3 to 4 in phase II studies of preoperative EBRT + chemotherapy are between 15-20% (2,5) so it is generally accepted that combined preoperative treatment does not appear to increase postoperative morbidity. Special techniques such as brachytherapy or IORT are not without morbidity. Some well-described symptoms are sensory neuropathy (16%), ureteral stenosis (23%) (117), and particularly anovaginal fistulae. As reirradiation there is no long-term studies showing low toxicity and effectiveness of these treatments, so they are performed in centers with good protocols and advanced technology experience that may cause little toxicity, ie IMRT. The treatment of CRC must consider both acute and chronic toxicity, the patient's general condition, status of disease, and the chances of a cure. Medical care should be provided using a multidisciplinary approach, offering the patient all the best treatment options for CRC.
6. Future of radiation therapy and combined treatments in colorectal cancers Several trials are studying ways to increase doses in rectal cancer for obtaining better local control. An Italian trial is comparing a pattern of accelerated hyperfractionation with concomitant boost and Capecitabine to conventional EBRT with Oxaliplatin and Capecitabine. This is the INTERACT-LEADER trial. Additionally, this test hopes that cT3N0-1 patients can skip the subsequent surgery if it is achieved pCR. The optimal schedule for RT, whether short or long course, must be, and is being, addressed. as we said earlier the Stockholm phase III trial will have short term results early. On the other hand, how could we administer chemotherapy in cases of short RT (5 fractions 5 Gy). To resolve this question we must await the results of the SCRIPT (Simply capecitabine in rectal cancer after irradiation plus TME) trial. All these studies will influence the way we work in the future. They will contribute to the prevention of colorectal tumours, better diagnosis, and better control at both the local and systemic scale.
256
Francisco L贸pez-Lara & Patricia Diezhandino Garc铆a
Conclusion Currently, neoadjuvant therapy, consisting of 5-FU based chemotherapy and long-course radiation therapy, is the standard of treatment for locally advanced rectal cancer. Local recurrence rates of 6-8% are consistent in randomised clinical trials. There does not appear to be a benefit in OS with preoperative chemoradiation. However, these regimens are nearing the point of tumour eradication, although the surgical resection of rectal cancer continues being necessary. Future randomised controlled trials will need to determine markers of complete response and the most effective combination of chemotherapy and radiation therapy to optimize outcomes in rectal cancer. The promising new therapies may have a role in selected patients although the research has to continue.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.
Stenbeck, M., Rosen, M., and Holm, L.E. 1995, Acta Oncol., 34, 881. Steel, G.G. and Peckham, M.J. 1979, Int. J. Radiat. Oncol. Biol. Phys., 5, 85. Edge, S.B., Byrd, D.R., Compton, C.C., Fritz, A.G., Greene, F.L., and Trotti, A. 2010, Cancer Staging Handbook, 6th Edition, Springer, New York. McCall, J.L., Cox, M.R., and Wattchow, D.A. 1995, Int. J. Colorectal Dis., 10, 126. MacFarlane, J.K., Ryall, R.D., and Heald, R.J. 1993, Lancet, 341, 457. Mohiuddin, M, Czito, BG, and Willett, CG. Colon and rectum. In: Edward C. Halperin,Carlos A. Perez,Luther W. Brady, ed. Principles and practice of radiation oncology. 5 Ed. Philadelphia: Lippincot-Raven; 2008.p. 1371. Camma, C., Giunta, M., Fiorica, F., Pagliaro, L., Craxi, A., and Cottone, M. 2000, JAMA, 284, 1008. Myerson, R.J., Valentini, V., Birnbaum, E.H., Cellini, N., Coco, C., Fleshman, J.W., Gambacorta, M.A., Genovesi, D., Kodner, I.J., Picus, J., Ratkin, G.A., and Read, T.E. 2001, Int. J. Radiat. Oncol. Biol. Phys., 50, 1299. Langen, K.M. and Jones, D.T. 2001, Int. J. Radiat. Oncol. Biol. Phys., 50, 265. Schild, S.E., Martenson, J.A., Jr., Gunderson, L.L., Ilstrup, D.M., Berg, K.K., O'Connell, M.J., and Weiland, L.H. 1989, Int. J. Radiat. Oncol. Biol. Phys., 17, 55. Stockholm Rectal Cancer Study Group. 1990, Cancer, 66, 49. Pahlman, L., Glimelius, B., and Graffman, S. 1985, Br. J. Surg., 72, 961. Wiltshire, K.L., Ward, I.G., Swallow, C., Oza, A.M., Cummings, B., Pond, G.R., et al. 2006, Int. J. Radiat. Oncol. Biol. Phys., 64, 709. Vigliotti, A., Rich, T.A., Romsdahl, M.M., Withers, H.R., and Oswald, M.J. 1987, Int. J. Radiat. Oncol. Biol. Phys., 13, 999. Klenova, A., Georgiev, R., Kurtev, P., and Kurteva, G. 2007, J BUON, 12, 227. Kubecova, M., Korinek, P., Sejdova, M., and Ambrus, M. 2010, Hepatogastroenterology., 57, 246.
Radiotherapy in colorectal cancer
257
17. Viani, G.A., Stefano, E.J., Soares, F.V., and Afonso, S.L. 2010, Int. J. Radiat. Oncol. Biol. Phys., 15;80(4):985. 18. Pettersson, D., Cedermark, B., Holm, T., Radu, C., Pahlman, L., Glimelius, B., and Martling, A. 2010, Br. J. Surg., 97, 580. 19. Gallagher, M.J., Brereton, H.D., Rostock, R.A., Zero, J.M., Zekoski, D.A., Poyss, L.F., Richter, M.P., and Kligerman, M.M. 1986, Int. J. Radiat. Oncol. Biol. Phys., 12, 1565. 20. Shanahan, T.G., Mehta, M.P., Bertelrud, K.L., Buchler, D.A., Frank, L.E., Gehring, M.A., Kubsad, S.S., Utrie, P.C., and Kinsella, T.J. 1990, Int. J. Radiat. Oncol. Biol. Phys., 19, 469. 21. O'Neill, B.D., Salerno, G., Thomas, K., Tait, D.M., and Brown, G. 2009, Br. J. Radiol., 82, 509. 22. Allal, A.S., Bischof, S., and Nouet, P. 2002, Strahlenther. Onkol., 178, 259. 23. Drzymala, M., Hawkins, M.A., Henrys, A.J., Bedford, J., Norman, A., and Tait, D.M. 2009, Br. J. Radiol., 82, 321. 24. Guerrero Urbano, M.T., Henrys, A.J., Adams, E.J., Norman, A.R., Bedford, J.L., Harrington, K.J., Nutting, C.M., Dearnaley, D.P., and Tait, D.M. 2006, Int. J. Radiat. Oncol. Biol. Phys., 65, 907. 25. Nijkamp, J., de Jong, R., Sonke, J.J., van Vliet, C., and Marijnen, C. 2009, Radiother. Oncol., 93, 285. 26. Wanebo, H.J., Koness, R.J., Vezeridis, M.P., Cohen, S.I., and Wrobleski, D.E. 1994, Ann. Surg., 220, 586. 27. Phillips, R.K., Hittinger, R., Blesovsky, L., Fry, J.S., and Fielding, L.P. 1984, Br. J. Surg., 71, 17. 28. Pucciarelli, S., Toppan, P., Friso, M.L., Fornasiero, A., Vieceli, G., Marchiori, E., and Lise, M. 1999, Dis. Colon Rectum, 42, 1276. 29. Knol, H.P., Hanssens, P.E., Rutten, H.J., and Wiggers, T. 1997, Strahlenther. Onkol., 173, 43. 30. Overgaard, M., Overgaard, J., and Sell, A. 1984, Radiother. Oncol., 1, 217. 31. Lybeert, M.L., Martijn, H., de Neve, W., Crommelin, M.A., and Ribot, J.G. 1992, Int. J. Radiat. Oncol. Biol. Phys., 24, 241. 32. Anonymous. 2010, National Cancer Institute, http://www.cancer.gov/cancertopics/factsheet/Therapy/radiation. 33. Rodel, C., Grabenbauer, G.G., Matzel, K.E., Schick, C., Fietkau, R., Papadopoulos, T., et al. 2000, Dis. Colon Rectum, 43, 312. 34. Gunderson, L.L., Nelson, H., Martenson, J.A., Cha, S., Haddock, M., Devine, R., et al. 1997, Int. J. Radiat. Oncol. Biol. Phys., 37, 601. 35. Eble, M.J., Lehnert, T., Herfarth, C., and Wannenmacher, M. 1997, Front. Radiat. Ther. Oncol., 31, 200. 36. Ca帽贸n, R., Azinovic, I., Aristu, J.J., et al. 2000, Long-term results after preoperative chemoradiation with or without intraoperative radiotherapy (IORT) in locally advanced rectal carcinoma, Proc. 2nd Congress Int. Soc. Intraoperative Radiation Therapy, Boston, 62 (Abstr.). 37. Calvo, F.A., Gomez-Espi, M., Diaz-Gonzalez, J.A., Alvarado, A., Cantalapiedra, R., Marcos, P., et al. 2002, Radiother. Oncol., 62, 201.
258
Francisco López-Lara & Patricia Diezhandino García
38. Gunderson, L.L., Nelson, H., Martenson, J.A., Cha, S., Haddock, M., Devine, R., et al. M.J. 1996, Dis. Colon Rectum, 39, 1379. 39. Abuchaibe, C.O. 1992, Radioterapia Intraoperatoria en Cáncer Colorrectal, Universidad de Navarra, Navarra. 40. Azinovic, I., Calvo, F.A., Puebla, F., Aristu, J., and Martinez-Monge, R. 2001, Int. J. Radiat. Oncol. Biol. Phys., 49, 597. 41. Wong, C.S., Cummings, B.J., Brierley, J.D., Catton, C.N., McLean, M., Catton, P., and Hao, Y. 1998, Int. J. Radiat. Oncol. Biol. Phys., 40, 427. 42. Guiney, M.J., Smith, J.G., Worotniuk, V., Ngan, S., and Blakey, D. 1997, Int. J. Radiat. Oncol. Biol. Phys., 38, 1019. 43. Mannaerts, G.H., Rutten, H.J., Martijn, H., Hanssens, P.E., and Wiggers, T. 2001, Dis. Colon Rectum, 44, 1749. 44. Papillon, J. 1975, Cancer, 36, 696. 45. Gerard, J.P., and Sentenac, I. 1997, Anus and rectum, A practical manual of brachytherapy, B. Pierquin and G. Marinello (Eds.), Medical Physics Publishing, Madison. 46. Maignon, P., Guerif, S., Darsouni, R., Salas, S., Barillot, I., d'Hombres, A., BoneLepinoy, M.C., Fraisse, J., and Horiot, J.C. 1998, Int. J. Radiat. Oncol. Biol. Phys., 40, 1077. 47. Papillon, J., Montbarbon, J.F., Gerard, J.P., Chassard, J.L., and Ardiet, J.M. 1989, Int. J. Radiat. Oncol. Biol. Phys., 17, 1161. 48. Gerard, J.P., Chapet, O., Ramaioli, A., and Romestaing, P. 2002, Int. J. Radiat. Oncol. Biol. Phys., 54, 142. 49. Papillon, J. 1990, Radiother. Oncol., 17, 275. 50. Spinelli, P., Mancini, A., and Dal Fante, M. 1995, Semin. Surg. Oncol., 11, 307. 51. Allison, R.R., Sheng, C., Cuenca, R., Bagnato, V.S., Austerlitz, C., and Sibata, C.H. 2010, Photodiagnosis Photodyn Ther, 7, 115. 52. Nakfoor, B.M., Willett, C.G., Shellito, P.C., Kaufman, D.S., and Daly, W.J. 1998, Ann. Surg., 228, 194. 53. Cancer Stanging Handbook. Six edition (2010). Springer New York, Inc. 54. Stamos, MJ, and Murrell, Z. 2007, Clin Cancer Res., 13, 6885s. 55. Sengupta, S. and Tjandra, J.J. 2001, Dis. Colon Rectum, 44, 1345. 56. Minsky, BD, Mies, C, Rich, TA, and Recht, TA. 1989, Int J Radiat Oncol Biol Phys, 17, 311. 57. Peeters, KC, Marijnen, CA, Nagtegaal, ID, Kranenbarg, EK, Putter, H, Wiggers, T, et al. 2007, Ann Surg., 246, 693. 58. Valentini, V, Beets-Tan, R, Borras, JM, et al. Radiother Oncol 2008, 87, 449. 59. Bosset, JF, Collette, L, Calais, G, et al. N Engl J Med 2006, 355, 1114. 60. Gérard, JP, Conroy, T, Bonnetain, F, et al. J Clin Oncol 2006, 24, 4620. 61. QUASAR collaborative group, Gray R, Barnwell, J, McConkey, C et al. Lancet 2007, 370, 2020. 62. Tepper, J.E., Cohen, A.M., Wood, W.C., Orlow, E.L., and Hedberg, S.E. 1987, Int. J. Radiat. Oncol. Biol. Phys., 13, 5. 63. Romsdahl, M.M. and Withers, H.R. 1978, Arch. Surg., 113, 446. 64. Vigliotti, A., Rich, T.A., Romsdahl, M.M., Withers, H.R., and Oswald, M.J. 1987, Int. J. Radiat. Oncol. Biol. Phys., 13, 999.
Radiotherapy in colorectal cancer
259
65. Ghossein, N.A., Ager, P.J., Ragins, H., Turner, S.S., DeLuca, F., Alpert, S., and Lowy, S.J. 1979, Surg. Gynecol. Obstet., 148, 917. 66. Adjuvant radiotherapy for rectal cancer: A systematic overview of 8507 patients from 22 randomised trials. Colorectal Cancer Collaborative Group. Lancet 2001;358:1291. 67. Prolongation of the disease-free interval in surgically treated rectal carcinoma. Gastrointestinal Tumor Study Group. N Engl J Med 1985;312:1465. 68. Withers, H.R. and Romsdahl, M.M. 1977, Int. J. Radiat. Oncol. Biol. Phys., 2, 1069. 69. Treurniet-Donker, A.D., van Putten, W.L., Wereldsma, J.C., Bruggink, E.D., Hoogenraad, W.J., Roukema, J.A., et al. 1991, Cancer, 67, 2042. 70. Medical Research Council Rectal Cacner Working Party. 1996, Lancet, 348, 1610. 71. Camma, C, Fiorica, F, Craxi, A, Cottone, M. 2002, The Lancet, 359(9311), 1068. 72. Ho, D.H., Pazdur, R., Covington, W., Brown, N., Huo, Y.Y., Lassere, Y., and Kuritani, J. 1998, Clin. Cancer Res., 4, 2085. 73. Fisher, B., Wolmark, N., Rockette, H., Redmond, C., Deutsch, M., Wickerham, D.L., et al. 1988, J. Natl. Cancer Inst., 80, 21. 74. Krook, J.E., Moertel, C.G., Gunderson, L.L., Wieand, H.S., Collins, R.T., Beart, R.W., et al. 1991, N. Engl. J. Med., 324, 709. 75. Douglass, H.O., Jr., Moertel, C.G., Mayer, R.J., Thomas, P.R., Lindblad, A.S., Mittleman, A., Stablein, D.M., and Bruckner, H.W. 1986, N. Engl. J. Med., 315, 1294. 76. Anonymous. 1990, JAMA, 264, 1444. 77. Gastrointestinal Tumour Study Group, 1992, J. Clin. Oncol., 10, 549. 78. Smalley, S.R., Benedetti, J.K., Williamson, S.K., Robertson, J.M., Estes, N.C., Maher, T., et al. 2006, J. Clin. Oncol., 24, 3542. 79. Merchant, N.B., Guillem, J.G., Paty, P.B., Enker, W.E., Minsky, B.D., Quan, S.H., et al. 1999, J. Gastrointest. Surg., 3, 642. 80. Willett, C.G., Badizadegan, K., Ancukiewicz, M., and Shellito, P.C. 1999, Dis. Colon Rectum, 42, 167. 81. Valentini, V. 1998, Tumouri, 84, 238. 82. Beets-Tan, R.G. and Beets, G.L. 2003, Int. J. Colorectal Dis., 18, 385. 83. Kim, J.H., Beets, G.L., Kim, M.J., Kessels, A.G., and Beets-Tan, R.G. 2004, Eur. J. Radiol., 52, 78. 84. Deschner, E.E. and Gray, L.H. 1959, Radiat. Res., 11, 115. 85. Van Den Brink, M., Van Den Hout, W.B., Stiggelbout, A.M., Klein Kranenbarg, E., Marijnen, C.A., Van De Velde, C.J., and Kievit, J. 2004, J. Clin. Oncol., 22, 244. 86. Symonds, C.J. 1914, Proc. R. Soc. Med., 7, 152. 87. Kligerman, M.M., Urdaneta, N., Knowlton, A., Vidone, R., Hartman, P.V., and Vera, R. 1972, Am. J. Roentgenol. Radium Ther. Nucl. Med., 114, 498. 88. Higgins, G.A., Humphrey, E.W., Dwight, R.W., Roswit, B., Lee, L.E., Jr., and Keehn, R.J. 1986, Cancer, 58, 352. 89. McLean, C.M. and Duncan, W. 1995, Clin. Oncol. (R. Coll. Radiol.), 7, 349.
260
Francisco López-Lara & Patricia Diezhandino García
90. Rouanet, P., Fabre, J.M., Dubois, J.B., Dravet, F., Saint Aubert, B., Pradel, J., Ychou, M., Solassol, C., and Pujol, H. 1995, Ann. Surg., 221, 67. 91. Mohiuddin, M., Regine, W.F., Marks, G.J., and Marks, J.W. 1998, Int. J. Radiat. Oncol. Biol. Phys., 40, 569. 92. Allen CV., Fletcher WS. Observations on preoperative irradiation of rectosignoid carcinoma. Am J Roentgen 1970;108:136. 93. McCall JL., Cox MR., Wattchow DA. 1995;Int J Colorectal Dis., 10,126-132. 94. Cedermark, B., Johansson, H., Rutqvist, L.E., and Wilking, N. 1995, Cancer, 75, 2269. 95. Hyams, DM, Mamounas, EP, Petrelli, N, Rockette, H, Jones, J, Wieand, HS, et al. Dis Colon Rectum 1997;40:131. 96. Roh, MS, Colangelo, L, Wieand, S, O'Connell, M, Petrelli, N, Smith, R, et al. Proc ASCO 2004;22:246 [abstract 3505]. 97. Sauer, R, Becker, H, Hohenberger, W, Rödel, C, Wittekind, C, Fietkau, R, et al. N Engl J Med 2004;351:1731. 98. Read, TE, Ogunbiyi, OA, Fleshman, JW, Birnbaum, EH, Fry, RD, Myerson, RJ, and Kodner, IJ. 2001, Diseases of the Colon & Rectum, 44(12):1778. 99. Murillo EC, Duque AA, Virizuela JAE. Tratamiento etiológico o causal del dolor oncológico. En: Sanz Ortiz J, editores. El control del sufrimiento evitable. Terapia analgésica. Madrid: You&Us;2000.p.33-49. 100. Sze, W.M., Shelley, M., Held, I., and Mason, M. 2008, Cochrane Database of Systematic Reviews 2002, Issue 1, http://onlinelibrary.wiley.com/o/cochrane/clsysrev/articles/CD004721/frame.html; jsessionid=78C3249326E336ED5F6FFE0F5433F9C5.d01t04. 101. Haddock, M.G., Gunderson, L.L., Nelson, H., Cha, S.S., Devine, R.M., Dozois, R.R., and Wolff, B.G. 2001, Int. J. Radiat. Oncol. Biol. Phys., 49, 1267. 102. Glimelius, B. 2003, Colorectal Dis., 5, 501. 103. Roh, M., Colangelo, L., Wieand, S., O'Connell, M., Petrelli, N., Smith, R., Mamounas, E., Hyams, D., and Wolmark, N. 2004, J. Clin. Oncol., 22 (Suppl. 14S), Abstr. 3505. 104. Yeoh, E, Horowitz, M, Russo, A, Muecke, T, Ahmad, A, Robb, T, and Chatterton, B. Int J Radiat Oncol Biol Phys 1993; 26(2):229. 105. Kozelsky, TF., Meyers, GE., Sloan, JA., Shanahan, TG., Dick, SJ., Moore, RL., et al. 2003, J Clin Oncol., 21, 1669. 106. Martenson, JA., Bollinger, JW., Sloan, JA., Novotny, PJ., Urias, RE., Michalak, JC., et al. 2000, J Clin Oncol, 18, 1239. 107. Martenson, JA Jr., Hyland, G., Moertel, CG., Mailliard, JA., O`Fallon, JR., Collins, RT., et al. 1996, Int J Radiat Oncol Biol Phys., 35, 299. 108. Rubin, P., and Casarett, GW. Clinical Radiation Phatology. Philadelphia, WB Saunders, 1968. 109. Emami, B., Lyman, J., Brown, A., Coia, L., Goitein, M., Munzenrider, J.E., Shank, B., Solin, L.J., and Wesson, M. 1991, Int. J. Radiat. Oncol. Biol. Phys., 21:109-122. 110. Common Terminology Criteria for Adverse Events v3.0 (CTCAE) ctep.cancer.gov/forms/CTCAEv3.pdf 111. http://www.rtog.org/members/toxicity/acute.html
Radiotherapy in colorectal cancer
261
112. http://www.rtog.org/members/toxicity/late.html 113. LENT SOMA scales for all anatomic sites. 1995, Int J Radiat. Oncol. Biol. Phys., 31(5), 1049. 114. Sauer, R, Fietkau, R, Wittekind, C, Rรถdel, C, Martus, P, Hohenberger, W, et al. 2003, Colorectal Dis.,5(5), 406. 115. Kollmorgen, CF., Meagher, AP., Wolff, BG., Pemberton, JH., Martenson, JA., and Illstrup, DM. 1994, Ann Surg., 220, 676. 116. Martenson, JA., Halyar, MY., Sloan, JA., Proulx, GM., Miller, RC., Deming, RL., et al. 2008, J of clin oncol., 26, 5248. 117. Alektiar, KM., Zelefsky, MJ., Paty, PB., Guillem, J., Saltz, LB., Cohen, AM., et al. 2000, Int J Radiat Oncol Biol Phys., 48, 219.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 263-283 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
11. Locoregional techniques for liver-limited metastatic colorectal cancer Virginia Arrazubi1, José Luis del Cura2 and María Teresa Pérez-Hoyos1 Departments of 1Medical Oncology and 2Radiology, Hospital de Basurto, Bilbao, Spain
Abstract. Local destruction techniques for colorectal liver metastases, particularly radiofrequency, have been incorporated into clinical practice in the past 5 years. Only 10% to 25% of patients with colorectal hepatic metastases have tumours that are potentially resectable. Local techniques represent a chance of treatment in conjunction with resection or as a standalone procedure. It is difficult to compare ablation modalities. Results of studies examining these techniques are mostly restricted to reports of patient cases, which are often not presented in a standardised manner. Additionally, randomised controlled trials are rare. Radiofrequency is the most extensively studied treatment modality, and response rates and survival after radiofrequency therapy have been established. Cryoablation and microwave ablation are employed in the same clinical setting as radiofrequency, but have less favorable safety profiles. Other local ablation techniques such as electrolysis, chemoembolization or high-intensity focused ultrasound are under development for colorectal liver metastases. Radioembolization and hepatic arterial Correspondence/Reprint request: Dr. Virginia Arrazubi, Department of Medical Oncology, Hospital de Basurto Bilbao, Spain. E-mail: varrazubi@gmail.com
264
Virginia Arrazubi et al.
chemotherapy should be considered as therapeutic options for patients who are not candidates for resection or radiofrequency.
Introduction Liver resection is the treatment of choice with the best chance for long-term cure in colorectal cancer patients with liver metastases[1-3]. However, most patients with liver-only metastases are not candidates for resection. The metastases may be inoperable due to a number of factors relating to the tumour distribution within the liver, the size of the tumour(s), and the presence of co-morbidities that preclude major surgery[4]. Novel approaches such as neoadjuvant chemotherapy, preoperative portal vein embolization, and two-stage hepatectomy have been used, but still not all metastases are safely resectable. Over the past decade, a number of novel, less invasive methods of ablation have been developed. In patients requiring extensive liver resection, ablative techniques allow parenchyma-sparing treatment of hepatic metastases. In addition, some patients are not candidates for surgery owing to the presence of co-morbidities. In these patients, percutaneous ablation allows local treatment with lower morbidity and mortality. Therefore, this approach could increase the number of patients who are candidates for treatment. Ablative techniques may be considered alone or in conjunction with resection. Several of these techniques have been developed specifically to treat unresectable liver metastases by inducing in situ coagulative necrosis. These techniques can be usefully divided into categories depending on their method of energy delivery or their mode of action. The purpose of this chapter is to review the effectiveness of radiofrequency, cryoablation, microwave ablation, radioembolization, hepatic arterial chemotherapy, and other ablation techniques in the treatment of colorectal liver metastases.
1. Radiofrequency Mechanism of action Radiofrequency (RF) ablation induces a thermal injury to the tumour through electromagnetic energy deposition. The term radiofrequency does not refer to the emitted wave but to an alternating electric current that oscillates in the range of high frequency (200â&#x20AC;&#x201C;1,200 kHz). The patient becomes a part of an electric circuit that includes a generator, grounding pads attached to the skin of the patient (usually on the thighs), and an electrode needle inserted into the tumour. When the generator is switched on, an alternating electric
Treatment of colorectal metastases to the liver
265
field is created within the tissue of the patient. The ions in the tissue that surrounds the electrode experience an agitation following the changes in direction of alternating electric current. This ionic agitation creates friction within the surrounding tissue and releases heat around the electrode[5]. The thermal damage caused by radiofrequency is dependent on both the tissue temperature achieved and the duration of the heating. Heating of tissue at 50–55ºC for 4–6 min produces irreversible cellular damage. At temperatures between 60ºC and 100ºC, tissue immediately coagulates, causing irreversible damage to mitochondrial and cytosolic enzymes of the cells. At more than 100–110°C, tissue vaporises and carbonises[5]. To ensure destruction of the tumour it is necessary to subject the entire tumour to cytotoxic temperatures for a certain period of time. A minimum may be to maintain a 50–100ºC temperature throughout the entire tumour for at least 6 min. In fact, given the relatively slow thermal conduction from the electrode through the tissues, usually the duration of application is maintained during a minimum of 12–30 min, depending of the size of the tumour being treated[5]. The goal of the procedure is to create a necrotic region that includes the tumour and a safety margin surrounding the target. Most authors recommend a safety margin of 1 cm beyond the boundaries of the tumour[5]. However, this is a controversial issue. Radiofrequency ablation is usually performed under intravenous sedation or, alternatively, general anaesthesia. The procedure may be performed
Figure 1a. Liver sonogram with a focal hyperechoic lesion (arrow) in a patient with a history of a surgically removed colon cancer. A liver metastasis was demonstrated in the biopsy.
266
Virginia Arrazubi et al.
Figure 1b. Radiofrequency ablation of the lesion. Using sonographic guidance, an electrode (arrowheads) has been percutaneously inserted through the liver into the lesion. Some minutes after starting the ablation, echogenic gas bubbles progressively appear around the tip of the electrode (arrows) as the ablation of the tumour progresses.
percutaneously, laparoscopically, or at open surgery. When a percutaneous approach is used, guidance of the placement of the electrode is more frequently performed with ultrasound (Figures 1a and 1b), and CT, but it can be performed using MRI as well[5].
Limits The main problems with radiofrequency ablation are the relatively slow thermal conduction from the electrode surface through the tissues and the need to avoid carbonisation and vaporisation around the tip of the electrode due to excessive heating. Both carbonised tissue and vapour act as insulators, blocking the transmission of the electric current. Two strategies have been used to increase the volume of ablation: to prevent overheating of the surrounding tissue, and to increase the number of active heads for treatment. The results are different kinds of electrodes: -
Internally cooled electrodes: the tip of the electrode is cooled by an inner circuit of circulating fluid to minimise carbonisation around the needle tip. In this way, the transmission of the electric current is not blocked, and a larger volume of tissue can be ablated.
Treatment of colorectal metastases to the liver
-
-
267
Multiple electrodes: several electrodes (usually up to three) work simultaneously to increase the total volume of ablation. Multitined expandable electrodes: these electrodes have several prongs that are deployed once the tip of the needle is in the tumour. The ablation volumes of each of these prongs combine to produce a larger total ablation volume. Some of these modified electrodes can ablate an area over 7 cm in diameter. Multitined perfused electrode: in this kind of electrode, a small volume of a saline solution is continuously injected through the tip into the surrounding tissues during the ablation. This fluid increases the conductivity of the treated tissue, allowing the radiofrequency current to penetrate further into the tissue, increasing tissue heating and necrosis.
Inadequate coagulation can also be due to the cooling effect of blood flow that can reduce the extent of thermal damage (the “heat sink effect”). Several strategies for reducing blood flow during ablation therapy have been proposed: -
Pringle manoeuvre (at open laparotomy and at laparoscopy). Embolization of the tumour-feeding artery. Combining thermal ablation with chemoembolization or transarterial administration of drug-eluting beads[5].
Experimental studies suggest that adjuvant chemotherapy may also increase the ablation volume. In an experiment in two animal models, Ahmed et al. found that RF ablation followed by IV liposomal doxorubicin resulted in a larger diameter of tumour necrosis than RF or doxorubicin alone. Combined therapy, as compared with liposomal doxorubicin therapy alone, was also associated with increased doxorubicin accumulation in the target organs[6]. When performing radiofrequency ablation in the liver, achieving a safety margin is more critical when treating liver metastases than when treating a primary tumour because metastases are usually surrounded by healthy liver parenchyma, throughout which the heat disperses easily. In the treatment of the hepatocellular carcinoma, the surrounding tissue is usually cirrhotic liver, which transmits heat poorly. So, the surrounding tissue concentrates the heat (the “oven effect”), potentiating the effect of the radiofrequency. Thus, the possibility of an incomplete treatment is higher in metastases than in hepatocellular carcinoma.
Results Several studies have been published on the effectiveness of radiofrequency ablation in colorectal metastases to the liver (Figures 2a and 2b).
268
Virginia Arrazubi et al.
Figure 2a. One month after radiofrequency, a CT shows an area of non-enhancement (arrows) in the treated lesion. No signs of residual tumour are observed.
Figure 2b. On 2-year follow-up MRI after radiofrequency, the residual lesion has decreased in size and no signs of recurrence are observed.
269
Treatment of colorectal metastases to the liver
It is important to specify that all of these studies have been carried out in the subset of patients not eligible for surgical resection and thus, the results should not be compared to results achieved with surgical resection. There is a great deal of variability in the reported 5-year patient survival rate (14% to 55%) and local tumour recurrence rate (3.6% to 60%) in colorectal metastases treated with radiofrequency ablation. The variability of these rates can be attributed to increased experience with the technique as it becomes more widely used, and to differences in tumour selection criteria. The highest ablation success rates are achieved in patients with solitary metastases or with a small number of tumours, each tumour measuring less than 3 cm in diameter[5, 7-15]. Several studies with a large number of patients are listed in the Table 1. In a recent study published by Otto et al. in two groups of patients with colorectal hepatic metastases amenable to surgery, patients treated with radiofrequency ablation as a primary option showed a higher frequency Table 1. Results of radiofrequency studies. Study
Number of patients
Access
3-year survival
5-year survival
Lencioni[10]
423
Percutaneous
47%
24%
Gillams[11]
309
Percutaneous
34%
24%
Siperstein[12]
234
Laparoscopic
20%
18.4%
Solbiati[13]
117
Percutaneous
46%
36%
Veltri[14]
122
Percutaneous, Open
Machi[15]
100
Percutaneous, Open, Laparoscopic
42%
31%
Sorensen[16]
100
Percutaneous, Open
64%
44%
Chen[17]
96
Percutaneous
25.1%
Amersi[18]
74
Percutaneous, Laparoscopic
Jackobs[19]
68
Percutaneous
68%
Hildebrand[20]
56
Percutaneous, Open, Laparoscopic
42%
22%
30%
270
Virginia Arrazubi et al.
of local recurrence (32% vs. 4%) and a shorter time to progression (203 vs. 416 days), as compared to patients treated initially by surgery. However, after primary treatment, RF ablation patients and surgery patients were amenable to repeated RF ablation or repeated surgery, resulting in identical rates of disease-free and 3-year overall survival in both treatment groups (67% and 60%, respectively)[21].
Complications Complication rate is proportional to the volume of tissue treated and increases in open radiofrequency ablation. Complication rates ranged between 13–27% after open treatment and 1.8–13% in percutaneous radiofrequency ablation. The incidence of major complications is 3.5–13% after open technique and 1.8–13% for percutaneous treatment. The published mortality rates in radiofrequency ablation for colorectal metastases have been 0–3.7% for open ablation and 0–0.5% after percutaneous treatment[5, 7-20]. Complications of radiofrequency ablation are similar for all ablative techniques and can be divided into two categories: a. Related to the placement of the electrode - Bleeding The risk for bleeding is relatively low and depends on tumour location, vascularity of the tumour, and the presence of underlying liver disease. Given the relatively large bore of the electrodes used and the multiple insertions sometimes required, it is absolutely necessary to check for an appropriate coagulation status before performing the procedure[5, 9, 22]. - Infection Hepatic abscess and sepsis are known complications of radiofrequency ablations. Radiofrequency ablation is a surgical procedure, so a strict adherence to sterile technique is compulsory to minimise the risk of infection. Diabetes and the presence of a bilioenteric anastomosis also increase the risk of sepsis after ablation[9, 22]. Non-specific fever after the ablation (“post-ablation syndrome”) is not uncommon but it is usually low-grade. When fever persists after 2–3 weeks, infection should be suspected[7, 22].
Treatment of colorectal metastases to the liver
271
- Tumour seeding Track seeding has been reported as a late and infrequent complication (0.5%). Superficial location is one factor that increases the risk of seeding. To minimise this complication, it is important to meticulously ensure optimal positioning on the first pass to avoid repeated entries into the tumour with the electrode. Also, heating the electrode as it is withdrawn after the ablation (â&#x20AC;&#x153;hot withdrawalâ&#x20AC;?) helps to destroy any tumour cells remaining in the track[5, 7, 9, 22]. - Pneumothorax Pneumothorax can occur when the electrode crosses the pleural recess. This occurs more frequently when CT is used as the method of guidance given that, ideally, an axial plane should be used to monitor the path of the electrode[5, 7-9, 22].
b. Related to thermal therapy - Thermal damage to adjacent structures The heat delivered can affect not only the intended target but it can also damage adjacent structures. The bile ducts are the most frequently damaged structures and bile duct strictures can occasionally be observed after the procedure. Also, the gallbladder and bowel are very sensitive to thermal insults, and cholecystitis and bowel perforation may appear when the tumour is located near these structures. Perforation occurs more frequently in the colon[5, 7, 9, 22]. Perforation of the gastrointestinal wall is probably the most feared complication and can be fatal. A history of previous abdominal surgery in the region adjacent to the tumour and chronic cholecystitis are both associated with an increased risk of perforation. It is possible that heat is more easily transmitted through the scars and adhesions caused by previous interventions. The colon, which is fixed to retroperitoneum, is the part of the gastrointestinal system most frequently damaged during ablations, due to its lack of mobility[22]. Careful planning before ablation is essential to avoid damaging these structures. One option to avoid thermal damage and to protect sensitive structures is to percutaneously inject air or fluid between the treatment area and the potentially damaged structure in order to create a protective gap. Another possibility is to use a laparoscopic or an open approach. A thermocouple, placed next to the structure that should be protected, can also be used to monitor the temperature and warn the surgeon when a dangerous level is reached.
272
Virginia Arrazubi et al.
- Hepatic failure Ablation of a large volume of tissue as a result of treatment of multiple lesions or aggressive creation of a wide ablative margin may be harmful. It may damage hepatic structures and may preserve too little normal tissue to permit adequate hepatic function, leading to a hepatic failure, especially in patients with previous hepatic surgery[22]. - Grounding pad burns Large grounding pads are used to disperse the electric current and close the electric circuit in the patient. These pads receive the same energy loads and produce the same amount of heat as the electrode, but over a larger surface. An inadequate placement of these pads may concentrate the heat in a small area causing second- and third-degree skin burns. Care should be taken to ensure that the grounding pads are placed equidistant from the electrode[22]. This is currently a rare complication which usually appears only in the initial phases of the learning curve. - Interferences with pacemaker and cardioverters Despite some concerns regarding the possible risk associated with performing radiofrequency ablation in patients with a pacemaker, no problems have been reported in these cases. However, some concern still exists regarding patients with an implantable cardioverter-defibrillator. In these patients, it may be advisable to inactivate the ventricular arrhythmia sensor if radiofrequency ablation is to be performed[5, 7-9, 22]. - Pleural effusion This occurs most frequently when treating lesions located near the diaphragm. It is the consequence of the thermal damage to the diaphragm and the adjacent pleura[7].
2. Cryoablation Mechanism of action The physiological basis of cryotherapy is the rapid formation of intracellular ice crystals, resulting in direct cell damage. In addition, hypoxia secondary to the disruption of the surrounding microvascular structures induces cell destruction. Cryoablation uses repetitive freezing and thawing of
273
Treatment of colorectal metastases to the liver
tissue to produce necrosis and irreversible tissue destruction. Both liquid nitrogen and argon gas can be used as coolants and are capable of producing temperatures as low as -40ยบC. Cell destruction is directly proportional to the rapidity and duration of freezing and the rate of thawing. Cryoablation of 1 cm of normal liver parenchyma around the treatment area is recommended to obtain complete ablation. Cryoablation is easy to monitor with ultrasound, which allows for accurate, real-time assessment during treatment.
Limits The local recurrence rate is higher for metastases adjacent to a major hepatic vessel. Even though technically this area is not a heat sink, it is thought that inflowing warm blood has a protective effect and prevents adequate freezing[4]. Larger lesions may be difficult to treat because large volumes of normal liver parenchyma may have to be subjected to lethal temperatures and conduction of cold is substantially reduced as the distance from the probe increases.
Results The rate of local recurrence varies from 9% to 44% in different reports[23-25]. These studies are difficult to compare due to the heterogeneity of the patients (number and size of lesions, local and systemic treatments received). One of the few trials comparing RF and cryotherapy found a 12% Table 2. Results of cryoablation studies. Study
Weaver[27] Seifert and Morris[28] Sheen[29] Yan[30] Kerkar[31] Seifert[32] Joosten[33] Brooks[34] Yan[35] Niu[36]
Recurrence
38% 33%
15% 24% 20% 39% 78%
Survival 3-year 40% 8% 14% 41% 48% 42% 40% 43% 43% 43%
5-year 20%
19% 28% 26% 19% 23% 24%
274
Virginia Arrazubi et al.
local recurrence rate with cryoablation and 14% with RF (no significant difference)[26]. Survival rates are difficult to establish because reports of 3- and 5-year survival rates are uncommon in the literature, and cryotherapy is frequently combined with other treatments. A summary of the results of cryoablation studies is presented in Table 2.
Complications The most serious complication of cryotherapy is the post-treatment syndrome referred to as cryoshock. This is a systemic response that consists of coagulopathy, thrombocytopenia, pleural effusions, an acute respiratory distress syndrome-like illness, and myoglobinuria. Reports in the literature suggest a mortality rate of 0% to 8%, although it may be responsible for 18% of perioperative deaths[37]. The cryoshock phenomenon correlates with the duration and volume of freezing and with hepatocellular injury[38]. Other major complications such as hemorrhage, subphrenic abscesses, biliomas and biliary fistulae frequently occurred in each of the studies, and there is a direct relationship between the volume of liver ablated and the incidence of complications. A few retrospective studies comparing radiofrequency with cryoablation found higher rates of complications, blood loss, and length of stay in patients treated with cryoablation[4, 26].
3. Microwave ablation Mechanism of action Microwaves with frequencies in the range of 900â&#x20AC;&#x201C;2,450 MHz induce tumour destruction. Within this electromagnetic field, polar molecules align themselves in the direction of the current. As the direction changes constantly, this continuous realignment causes a heating effect, and electromagnetic energy induces cell death by coagulative necrosis. This procedure may be performed percutaneously, laparoscopically, or during open surgery. Limits One of the main limitations of this technique is the difficulty of creating consistent ablation zones larger than 3 cm. Microwave ablation has principally been used in hepatocellular carcinoma but is also used in some colorectal liver metastases. In colorectal liver
275
Treatment of colorectal metastases to the liver
metastases, its use is still in development, and few studies have been published. Results Published studies so far are limited, with a small number of patients and a lack of long-term comparative survival data. Primary and secondary tumours treated with microwave ablation have a local recurrence rate of 6% in reference centres, and select patients have a 3-year overall survival of 40%. The main studies are shown in Table 3. Table 3. Results of microwave studies. Study Shibata[39] Tanaka[40] Bhardwaj[41]
Number of patients 14 16 31
Recurrence
3-year survival
22% 6%
14% 39% 40%
Complications A high rate of complications (as high as 20.6%) has been reported in patients treated for colorectal metastases[42]. These complications are bleeding, bilioma formation, tumour dissemination, portal vein thrombosis, or post-ablation pneumothorax. The incidence of complication increases when the tumour size exceeds 4 cm [42].
4. Radioembolization Mechanism of action Radioembolization or selective internal radiation therapy with microspheres loaded with the radionuclide yttrium-90 (90Y) enables multiple hepatic metastases to be targeted in a single procedure. The 90Y-resin microspheres lodge within the malignant microvasculature, where they deliver high, localised therapeutic doses of β-radiation to the tumour over approximately 14 days, while maintaining the radiation exposure of normal liver within tolerable levels[43]. In a study that measured tumour dosimetry in the human liver following hepatic yttrium-90 microsphere therapy, the authors found that the average level of radiation within the tumour periphery ranged from 200 Gy to 600 Gy, while in the normal liver parenchyma the average level was 8.9 Gy[44].
276
Virginia Arrazubi et al.
Limits A simulation of the treatment is first performed by injecting tracer doses of technetium-99m-labelled macroaggregated albumin (99Tc-MAA) intraarterially. The distribution of the 99Tc-MAA is accepted as an accurate representation of the distribution of the microspheres. The tracer uptake in the body is then quantified. If the lungs absorb >20% of the injected dose, treatment is contraindicated. During the interventional procedure, all collateral vessels that can transport microspheres from the targeted liver vasculature to extrahepatic organs are occluded by embolization to prevent systemic deposition of 90Y-microspheres. If there is abdominal 99Tc-MAA deposition outside of the liver, it is recommended to repeat the angiogram and the prophylactic embolization procedure, and to re-inject the 99Tc-MAA. Additionally, the presence of main portal vein thrombosis or pulmonary arterial thrombosis is a contraindication for the treatment[45, 46].
Results This procedure is being used increasingly to manage liver tumours of various histological origins. In some studies of colorectal liver metastases, 90 Y-microspheres have been combined with systemic chemotherapy or hepatic arterial infusion, with an encouraging median survival and response rate[47, 48]. Very few randomised controlled trials of radioembolization in colorectal liver metastases have been published. In phase I and II studies, radioembolization significantly decreased the average tumour size and increased the time to progression. The main studies related to radioembolization are summarised in Table 4. In a phase III randomised study, patients received hepatic artery chemotherapy with or without a single injection of 90Y-resin microspheres. Patients receiving radioembolization had a significantly higher tumour response rate and a lower rate of hepatic progression compared with those receiving artery chemotherapy alone[49]. Another phase III trial assessed the efficacy of 90 Y-microspheres in chemotherapy-refractory liver-limited colorectal cancer metastases[45]. The authors compared protracted fluorouracil intravenous infusion with or without radioembolization. The combined treatment improved the time to progression compared with fluorouracil alone. Therefore, radioembolization with 90Y-resin microspheres should be considered a valid therapeutic option for patients with liver-limited metastatic colon cancer, but questions remain regarding its timing and optimal combination schedule. Potentially, more effective chemotherapeutics combined with targeted therapies have been introduced into the clinic, and the clinical value of 90Y-resin microspheres when compared to these novel treatments warrants confirmation and validation.
Table 4. Results of radioembolization studies.
Treatment of colorectal metastases to the liver
277
278
Virginia Arrazubi et al.
Complications Radioembolization can lead to fatigue, vomiting, anorexia, fever, abdominal discomfort, and cachexia. Hepatic toxicity can be assessed by changes in the levels of liver enzymes and metabolites; these changes can be severe and can lead to significant morbidity and mortality. Hepatic toxicity rates following radioembolization are between 15% and 20%. Various factors such as the presence of altered hepatic function at baseline, age, and radiation distribution may predispose patients to the hepatotoxic effects of radioembolization. Radiation-induced liver disease (RILD) has been observed in 4% of patients. Less than 2% of patients develop severe biliary toxicity, and cholangitis has also been reported following radioembolization. Portal hypertension due to postembolization fibrosis is possible, but the clinically significant occurrence of portal hypertension is low[45, 46]. The inadvertent spread of microspheres to the gastrointestinal tract is responsible for potentially severe complications such as ulceration. Radiation pneumonitis has been observed when the lung shunt fraction is greater than 13%[45, 46]. Adverse events with radioembolization are mainly low and manageable. Only a limited number of patients develop gastrointestinal ulceration, bleeding or microsphere embolization in other organs.
5. Hepatic arterial infusion (HAI) of chemotherapy Mechanism of action The rationale for hepatic arterial infusion (HAI) is that hepatic metastases are fed by the hepatic artery, while the normal liver parenchyma is fed mostly by the portal vein. Infusing cytotoxic agents directly into the hepatic artery leads to prolonged elevated levels of the drug within tumour cells, with relative sparing of normal liver parenchyma. Agents with high hepatic extraction rates are particularly attractive for HAI. Floxuridine (FUDR: 5-fluoro-2´-deoxyuridine) has a 95% hepatic extraction rate when given via HAI. Phase I/II trials are ongoing to assess the efficacy of other cytotoxic agents such as oxaliplatin, cisplatin, mitomycin-C and irinotecan using HAI[54].
Limits All patients should have a nuclear medicine macro-aggregated albumin scan to assess whether the perfusion is contained to the liver or if there is extrahepatic perfusion. If there is misperfusion to the stomach or duodenum,
Treatment of colorectal metastases to the liver
279
ulceration or diarrhea can result. A subcutaneous pump connected to an arterial catheter must be surgically implanted.
Results Early clinical series using FUDR delivered as continuous HAI produced encouraging results. In subsequent years, prospective randomised phase III trials were published. In the 1990s, seven studies compared HAI FUDR with systemic FUDR or 5-FU Âą LV. The response rates were high, but benefit in overall survival was found only in two studies. Two meta-analyses included these trials and showed an increase in response and a survival advantage[55, 56]. However, two more recent European trials did not show any increase in survival [57, 58]. Some drawbacks to these studies have been discussed: 1) the number of patients was small; 2) a crossover design was allowed; 3) patients with extra-hepatic disease were included; and 4) many who were assigned to the HAI arms did not receive treatment. To avoid these biases, the CALGB-9481 study was initiated[59]. This study compared HAI FUDR against systemic 5-fluorouracil/lecovorin. The HAI group showed a significant increase in overall survival (24.4 months vs. 20 months; p = 0.0034), and the time to progression was superior in the HAI arm (9.8 months vs. 7.3 months; p = 0.12). Finally, a meta-analysis pooling individual patient data from 10 studies was published[60]. The response rate was 42.9% vs. 18.4% for HAI vs. systemic treatment, and the median overall survival was 15.9 months vs. 12.4 months for the HAI and systemic groups, respectively (HR = 0.9; p = 0.24). Despite a number of randomised clinical trials, the therapeutic impact of fluoropyrimidine-based HAI is still a matter of debate. HAI provides a tumour response advantage when compared with fluoropyrimidine-based systemic infusion; however, modern chemotherapy schedules can obtain tumour response rates similar to or even higher than those observed with HAI. Furthermore, an advantage in overall survival needs to be clarified. HAI added to systemic chemotherapy might control hepatic disease and prevent extra-hepatic metastases. New HAI protocols that use concomitant or alternating modern systemic chemotherapy have increased response rates, leading to increased resectability of liver metastases. However, published series are short, retrospective, or are early phase trials. Further trials comparing HAI plus systemic chemotherapy vs. systemic chemotherapy alone are needed to determine the relative merits of each.
Complications Biliary sclerosis determines the dose-limiting toxicity because the hepatic artery supplies blood to the bile ducts. The addition of dexamethasone to FUDR reduces biliary toxicity.
280
Virginia Arrazubi et al.
Early complications are more likely to involve misperfusion of the liver. Other early complications are pocket infections or hematomas, incomplete perfusions, and thrombosis. Late complications include occlusions, dislodgement, catheter occlusions, or arterial thromboses. Very rarely, the catheter can come out of the artery and abdominal blood loss can occur. Hepatotoxicity depends on the drugs used and the duration of treatment[54].
6. Other locoregional ablation techniques Other techniques, such as electrolysis, chemoembolization, and highintensity focused ultrasound, have been developed to treat local liver tumours. Electrolysis uses direct current passed between two electrodes. The negatively and positively charged ions in the tissue are attracted to the anode and cathode, respectively. This results in tissue destruction around the anode and cathode. The principal disadvantage of this technique compared with microwave and radiofrequency ablation is the relatively long period of time required to create similarly sized lesions; however, its safely profile is acceptable. Long-term survival data are awaited, and if these data are encouraging, electrolysis must be further refined to induce lesions more quickly[61]. Hepatic arterial embolization with the injection of vasoocclusive particles into the hepatic artery can occlude the blood supply to the tumour metastases. Chemoembolization involves local entrapment of the drug in an appropriate vehicle (the embolic agent), thus providing prolonged exposure of tumour to the drug with minimal systemic drug circulation. This strategy has been employed using doxorubicin, mitomycin-C, and cisplatin in the embolic agent lipiodol, and, more recently, with embolic doxorubicin as a drug-eluting bead. This later strategy has demonstrated response rates in hepatocellular carcinoma but is rarely used to treat liver metastases from colorectal cancer as a single entity[61]. High-intensity focused ultrasound (HIFU) ablation for liver tumours is an extracorporeal non-invasive treatment method using focused ultrasound beams that can cause complete coagulative necrosis of target lesions. For treating liver metastases from colon cancer, HIFU seems to be safe, but its efficacy is questionable. Therefore, further research is warranted[62].
Conclusion Although surgery is currently the only potentially curative treatment for liver metastases from CRC, however, fewer than 25% of cases are
Treatment of colorectal metastases to the liver
281
candidates for curative resection. A number of other locoregional therapies, such as radiofrequency or microwave ablation, cryotherapy, and chemotherapy, may be offered to patients with unresectable but isolated liver metastases. RF is the ablation procedure most used and it may be performed intraoperatively, laparoscopically, or percutaneously. The percutaneous approach is associated with the least procedural risk and may be performed under local anesthesia. This approach should be considered a safe, effective and potentially curative option as a primary treatment for patients with unresectable liver tumours or conditions that prohibit general anesthesia or abdominal surgery. Its long-term results are comparable with those of investigations using surgical resection. The factors such as lesion size, the number of lesions and location determine its success. Cryoablation and microwave ablation are employed in the same clinical setting but with less favorable safety profiles. Other local ablation techniques such as electrolysis, chemoembolization or high-intensity focused ultrasound are under development for CRC liver metastases. Finally, radioembolization and hepatic arterial chemotherapy should be considered as therapeutic options for patients who are not candidates for resection or radiofrequency.
References 1.
Abdalla, E.K., Vauthey, JN, Ellis, LM, Ellis, V, Pollock, R, Broglio, KR et al. 2004, Ann Surg, 239(6), 818. 2. Aloia, T.A., Vauthey, JN, Loyer, EM, Ribero, D, Pawlik, TM, Wei, SH, et al. 2006, Arch Surg, 141(5), 460. 3. Belghiti J, Hiramatsu K, Benoist S, Massault P, Sauvanet A, and Farges O. 2000, J Am Coll Surg, 191(1), 38. 4. Pearson AS, Izzo F, Fleming RY, Ellis LM, Delrio P, Roh MS, et al. 1999, Am J Surg, 178(6), 592. 5. Lencioni, R., Della Pina, C, and Bartolozzi C. 2009, Abdom Imaging, 34(5), 547. 6. Ahmed M, Liu Z, Lukyanov AN, Signoretti S, Horkan C, Monsky WL, et al. 2005, Radiology, 235(2), 469. 7. Wong, SL, Mangu, PB, Choti, MA, Crocenzi, TS, Dodd III, GD, Dorfman, GS, et al. 2009, J Clin Oncol., 28(3), 493. 8. Stang, A., Fischbach, R, Teichmann, W, Bokemeyer, C, and Braumann, D. 2009, Eur J Cancer, 45(10), 1748. 9. Flanders, V.L. and D.A. Gervais, 2010, J Vasc Interv Radiol., 21(8 Suppl), S214. 10. Lencioni R, Crocetti L, Cioni R, Mussi A, Fontanini G, Ambrogi M, et al.2004, Cardiovasc Intervent Radiol, 27(6), 581. 11. Gillams, A.R. and Lees, WR. 2009, Eur Radiol, 19(5), 1206. 12. Siperstein AE, Berber E, Ballem N, and Parikh RT. 2007, Ann Surg, 246(4), 565.
282
Virginia Arrazubi et al.
13. Solbiati, L, Livraghi, T, Goldberg, SN, Ierace, T, Meloni, F, Dellanoce, M, et al. 2001, Radiology, 221(1), 159. 14. Veltri, A., Sacchetto, P, Tosetti, I, Pagano, E, Fava, C, and Gandini, G. 2008, Cardiovasc Intervent Radiol, 31(5), 948. 15. Machi J, Oishi AJ, Sumida K, Sakamoto K, Furumoto NL, Oishi RH, and Kylstra JW. 2006, Cancer J, 12(4), 318. 16. Sorensen, S.M., Mortensen, FV, and Nielsen, DT. 2007, Acta Radiol, 48(3), . 253. 17. Chen, MH, Kun Yan, WY, Ying Dai, WG, Wang, YB, Zhang, XP, Yin, SS. 2005, World J Gastroenterol, 11(40), 6395. 18. Amersi FF, McElrath-Garza A, Ahmad A, Zogakis T, Allegra DP, Krasne R, and Bilchik AJ. 2006, Arch Surg, 141(6), 581. 19. Jakobs TF, Hoffmann RT, Trumm C, Reiser MF, and Helmberger TK. 2006, Anticancer Res, 26(1B), 671. 20. Hildebrand P, Leibecke T, Kleemann M, Mirow L, Birth M, Bruch HP, and Bürk C. 2006, Eur J Surg Oncol, 32(4), 430. 21. Otto, G, Düber, C, Hoppe-Lotichius, M, König, J, Heise, M, and Bernhard Pitton, M. 2010, Ann Surg., 251(5), 796. 22. Rhim H, Dodd GD 3rd, Chintapalli KN, Wood BJ, Dupuy DE, Hvizda JL, et al. 2004, Radiographics, 24(1), 41. 23. Adam, R, Akpinar, E, Johann, M, Kunstlinger, F, Majno, P, and Bismuth, H. 1997, Ann Surg, 225(1), 39. 24. Onik, G, Porterfield, B, Rubinsky, B, Zemel, R, Weaver, L, Diamond, D, and Cobb, C. 1991, Cancer, 67(4), 901. 25. Mala T, Edwin B, Mathisen Ø, Tillung T, Fosse E, Bergan A, et al.2004, Scand J Gastroenterol, 39(6), 571. 26. Kornprat P, Jarnagin WR, DeMatteo RP, Fong Y, Blumgart LH, and D'Angelica M. 2007, Arch Surg, 142(11), 1087. 27. Weaver, M.L., J.G. Ashton, and R. Zemel, 1998, Semin Surg Oncol, 14(2), 163. 28. Seifert, J.K. and Morris, DL. 1999, Br J Surg, 86(2), 234. 29. Sheen, A.J., Poston, GJ and Sherlock, DJ. 2002, Br J Surg, 89(11), 1396. 30. Yan, D.B., Clingan, P, and Morris, DL. 2003, Cancer, 98(2), 320. 31. Kerkar S, Carlin AM, Sohn RL, Steffes C, Tyburski J, Littrup P, and Weaver D. 2004, Surgery, 136(4), 770. 32. Seifert JK, Springer A, Baier P, and Junginger T. 2005, Int J Colorectal Dis, 20(6), 507. 33. Joosten J, Jager G, Oyen W, Wobbes T, and Ruers T. 2005, Eur J Surg Oncol, 31(10), 1152. 34. Brooks, AJ, Wang, F, Alfredson, M, Yan, TD, and Morris, DL. 2005, Surgeon, 3(4), 265. 35. Yan, T.D., Padang, R and Morris, DL. 2006, J Am Coll Surg, 202(1), 100. 36. Niu R, Yan TD, Zhu JC, Black D, Chu F, and Morris DL. 2007, Ann Surg Oncol, 14(7), 2078. 37. Seifert, J.K, and Morris, DL. 1999, World J Surg, 23(2), 109. 38. Seifert JK, Stewart GJ, Hewitt PM, Bolton EJ, Junginger T, and Morris DL. 1999, World J Surg, 23(10), 1019. 39. Shibata T, Niinobu T, Ogata N, and Takami M. 2000, Cancer, 89(2), 276.
Treatment of colorectal metastases to the liver
283
40. Tanaka, K., Shimada, H, Nagano, Y, Endo, H, Sekido, H, and Togo, S. ,2006, Surgery, 139(2), 263. 41. Bhardwaj N, Strickland AD, Ahmad F, El-Abassy M, Morgan B, Robertson GS, and Lloyd DM. 2010, Eur J Surg Oncol., 36(3), 264. 42. Shimada S, Hirota M, Beppu T, Matsuda T, Hayashi N, Tashima S, et al. 1998, Surg Today, 28(11), 1130. 43. Lau WY, Leung TW, Ho S, Chan M, Leung NW, Lin J, et al. 1994, Br J Radiol, 67(794), 136. 44. Campbell, A.M., Bailey, IH., and Burton, MA. 2001, Phys Med Biol, 46(2), 487. 45. Hendlisz, A, Eynde, MV den, Peeters, M, Maleux, G, Lambert, B, Vannoote, J. et al. 2010, J Clin Oncol., 28(23), 3687. 46. Riaz, A, Lewandowski, RJ, Kulik, L, and Salem, R. 2009, Q J Nucl Med Mol Imaging, 53(3), 311. 47. Kennedy, AS, Coldwell, D, Nutting, C, Murthy, R, Wertman, DE Jr, Loehr, SP, et al. 2006, Int J Radiat Oncol Biol Phys, 65(2), 412. 48. Jakobs, TF, Hoffmann, RT, Dehm, K, Trumm, C, Stemmler, HJ, Tatsch, K, et al. 2008, J Vasc Interv Radiol, 19(8), 1187. 49. Gray, B, Van Hazel, G, Hope, M, Burton, M, Moroz, P, Anderson, J, and Gebski, V. 2001, Ann Oncol, 12(12), 1711. 50. Sharma, RA, Van Hazel, GA, Morgan, B, Berry, DP, Blanshard, K, Price, D, et al. 2007, J Clin Oncol, 25(9), 1099. 51. Van Hazel, GA, Pavlakis, N, Goldstein, D, Olver, IN, Tapner, MJ, Price, D, et al. 2009, J Clin Oncol, 27(25), 4089. 52. Cosimelli, M, Golfieri, R, Cagol, PP, Carpanese, L, Sciuto, R, Maini, CL, et al. 2010, Br J Cancer, 103(3), 324. 53. Van Hazel G, Blackwell A, Anderson J, Price D, Moroz P, Bower G, et al. 2004, J Surg Oncol, 88(2), 78. 54. Power, D.G., and Healey-Bird, BR, and Kemeny, NE. 2008, Clin Colorectal Cancer, 7(4), 247. 55. Anonymous. 1996, J Natl Cancer Inst., 88(5), 252. 56. Harmantas, A., Rotstein, LE and Langer, B. 1996, Cancer, 78(8), 1639. 57. Kerr, DJ, McArdle, CS, Ledermann, J, Taylor, I, Sherlock, DJ, Schlag, PM et al. 2003, Lancet, 361(9355), 368. 58. Lorenz, M. and Muller, HH. 2000, J Clin Oncol, 18(2), 243. 59. Kemeny, NE, Niedzwiecki, D, Hollis, DR, Lenz, HJ, Warren, RS, Naughton, MJ, et al. 2006, J Clin Oncol, 24(9), 1395. 60. Mocellin, S, Pilati, P, Lise, M, and Nitti, D. 2007, J Clin Oncol, 25(35), 5649. 61. Bhardwaj, N, Strickland, AD, Ahmad, F, Dennison, AR, and Lloyd, DM. 2010, Surg Endosc., 24(2), 254. 62. Park, MY; Jung, SE; Cho, SH; Piao, XH; Hahn, ST; Han, JY; and Woo, IS. 2009, Int J Hyperthermia, 25(3), 180.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 285-308 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
12. Prognostic and predictive factors in colorectal cancer: The importance of reliable markers for effective selection of therapy
1
Miriam López-Gómez1, Enrique Casado1, Paloma Cejas2 and Jaime Feliu2
Department of Medical Oncology, Infanta Sofía Hospital, San Sebastián de los Reyes, Madrid, Spain; 2Department of Medical Oncology La Paz University Hospital, Madrid, Spain
Abstract. In recent years, significant advances have been made in the study of colorectal cancer, (CRC) prognosis and outcome. A better understanding of the molecular basis of this disease, as well as the development of new therapeutic approaches to treat it, has dramatically altered its management. It is essential for physicians to have an effective methodology by which to plan treatment, project prognosis and measure outcome. In this chapter, we discuss predictive and prognostic markers identified in CRC in terms of their utility in assisting the clinician to select the most efficacious and least toxic therapeutic options for each patient. Correspondence/Reprint request: Dr. Miriam López-Gómez, Department of Medical Oncology, Infanta Sofía Hospital, San Sebastián de los Reyes, Madrid, Spain. E-mail: miriam.lopez@telefonica.net
286
Miriam López-Gómez et al.
Introduction CRC constitutes one of the leading causes of cancer-related deaths in the Western world. Selection of the most beneficial treatment regimes in CRC remains a challenge and is hindered by a lack of well established prognostic markers (markers that correlate with survival or DFS (disease-free survival) and predictive markers (markers that predict response to a particular therapy). The American Joint Committee on Cancer (AJCC) established a Colorectal Working Group to develop a system for the evaluation of prognostic marker values. The goal of this group was to categorise prognostic markers according to their strength and reliability based on data found in the literature, which would allow physicians to predict the aggressiveness of the disease and the likelihood of recurrence after surgery based on analysis of these markers. These factors will be reviewed in the first part of this chapter. Unfortunately, selection of the most appropriate therapeutic approach has been plagued by the development of drug resistance. In recent years, several studies have attempted to identify critical molecular and/or biochemical markers that can be used to predict response to chemotherapy. The primary aim of such predictive biomarker testing is to allow treatment to be designed according to the molecular phenotype of the tumour. In the second part of this chapter, we will examine some of the potentially clinically important predictive markers for chemotherapy in CRC.
1. Prognostic factors The AJCC convened a Prognostic Factors Consensus Conference to evaluate the role of biologic, genetic, molecular and other nonanatomic factors in staging cancer (2). The first edition of this staging manual was published in 1977 (3), and established the use of T (tumour extent), N (lymph node status) and M (the presence or absence of metastases) in an organised staging outline that allowed clinicians to uniformly describe the extent of the disease. Recently, an increasing number of nonanatomic factors have been identified and studied. Some of these factors have been shown to influence outcome predictions and treatment decisions, and have been classified as “prognostic factors”. Under the auspices of the College of American Pathologists, a multidisciplinary group of clinical (medical oncology, surgical oncology and radiation oncology), pathologic and statistical experts reviewed relevant medical literature and stratified prognostic factors into categories that reflect the strength of published evidence demonstrating their prognostic value. The
Prognostic and predictive factors in colorectal cancer
287
factors fall into five categories (from the College of American pathologists consensus statement; 2). Category I is comprised of factors definitively proven to be of prognostic importance based on evidence from multiple statistically robust published trials, and are widely used in patient management. Category IIA includes factors that have been extensively studied biologically and/or clinically, and have repeatedly shown to have prognostic value for outcome and/or predictive value for therapy that is of sufficient import to be included in the pathology report, but remains to be validated in statistically robust studies. Factors in Category IIB include those shown to be promising in multiple studies, but lack sufficient data for inclusion in Category I or IIA. Category III includes potential factors not yet sufficiently studied to determine their prognostic value. Category IV includes factors that have been well studied and shown to have no prognostic significance (Table 1). Table 1. Classification of prognostic markers in colorectal adenocarcinoma (2).
Classification of prognostic markers in colorectal adenocarcinoma Category I Well supported by literature, generally used in patient management, and of sufficient importance to modify TNM stage groups. Category IIA Extensively studied biologically and/or clinically. Prognostic value for therapy, sufficient to be noted in pathology report. Category IIB Well studied but not sufficiently established. Category III Less well studied and not yet sufficiently established. Category IV
Studied and shown to have no consistent prognostic significance.
Category I: Factors well supported by the literature and generally used in patient management a) Pathological assessment of tumour extent (pT) Tumour in situ (Tis). The designation “Tis” (i.e., carcinoma in situ, a malignancy that has not yet penetrated the basement membrane of the epithelium to invade the underlying lamina propria) is used to refer both to intraepithelial malignancies (“Tie”) and intramucosal carcinomas (“Tim”), tumours that have invaded the mucosal estroma (the lamina propria, up to and including the muscularis mucosae). These designations are of great prognostic
288
Miriam LĂłpez-GĂłmez et al.
significance in that patients classified with these types of tumours have an extremely good prognosis. Invasion of parietal peritoneum. The highest category of local extent of colorectal tumour (T4) includes both extension into an adjacent structure or organ and involvement of the parietal peritoneum (serosal involvement), which has been demonstrated to have independent adverse prognostic significance (4). Patients with pT4 tumours that have penetrated the visceral peritoneum have a shorter median survival time after surgical resection compared to patients with pT4 tumours that lack serosal involvement, in either the presence or absence of distal metastases (5). Based on this adverse effect on outcome, it has been suggested that tumours in the T4 category be further classified as T4a (tumours that invade adjacent structures or organs) and T4b (tumours that involve the visceral peritoneum).
b) Regional lymph node metastases (pN) Metastasis to regional lymph nodes as determined by pathologic assessment is among the factors that most strongly predict outcome following surgical resection, second only to distant metastatic disease in importance. It is recommended that all identified lymph nodes be sectioned, and a minimum of 12â&#x20AC;&#x201C;15 negative lymph nodes are required to confirm regional node negativity (6). When less than 12 negative lymph nodes are harvested, the dissection is considered insufficient. These patients have a higher incidence of postoperative cancer death than patients with a sufficient dissection (p<0.001) in stage II CRC (7). Therefore, insufficient lymph node dissection constitutes an independent risk factor for postoperative cancer death in patients who undergo CRC surgery.
c) Presence or absence of blood or lymphatic vessel invasion T1 colorectal tumours may invade submucosal vessels, either venous or lymphovascular (i.e., small nonmuscularised vessels that represent either postcapillary venules or lymphatics). The invasion of lymphovascular vessels has been associated with a significantly increased risk of regional lymph node metastasis (8), and invasion of the submucosal venous system has been associated with the development of liver metastasis (9). The presence of one of these factors in a T1 tumour may influence the decision to perform more extensive surgical excision. For these tumours, the T1 category should be further divided into T1a (no evidence of lymphatic or venous invasion) and T1b (the presence of lymphatic or venous invasion).
Prognostic and predictive factors in colorectal cancer
289
d) Residual tumours The finding of tumour tissue at a surgical resection margin indicates that the tumour has not been completely removed from the patient at the surgical interface. The surgical margin status should always be examined and reported, as residual tumour tissue is related to a worse prognosis and a higher rate of local or distant recurrence. e) Elevated serum carcinoembryonic antigen (CEA) CEA is the most widely accepted and frequently used tumour marker for colon cancer. Its detection is relatively inexpensive and easy. The threshold value of CEA varies, but the standard is 2.0â&#x20AC;&#x201C;2.5 ng/ml, dependent on the measurement test. Due to frequent false-positive outcomes caused by benign gastro-intestinal disorders and smoking, the generally accepted threshold value in follow-up testing for colorectal carcinoma is 5.0 ng/ml. Preoperative CEA levels >5.0 ngr/ml have been shown to have an adverse impact on prognosis (i.e., survival) that is independent of tumour stage (10, 11). Tumours with elevated CEA levels at presentation should be differentiated from those without CEA elevation by TNM staging (Cx, serum CEA cannot be assessed; C0, serum CEA not elevated; C1, CEA levels elevated â&#x2030;Ľ5 ng/ml).
II. Category IIA: Factors shown to have prognostic value for outcome and/or predictive value for therapy but still need to be validated in statistically robust studies a) Presence of residual tumour in the resection specimen following neoadjuvant therapy (ypTNM) Any remaining viable tumour found in a resection specimen following neoadjuvant therapy is associated with a worse prognosis. Therefore, the region of resection should always be evaluated.
b) Radial margins The radial margin represents the adventitial soft tissue margin of a nonperitonealized surface. This acquires special relevance in rectal disease (12, 13). By contrast, the colon is encased in segments by a peritonealised (serosal) surface (e.g., the cecum, transverse colon and sigmoid colon) and the only radial margin remaining is the mesenteric resection margin. Due to this, studies examining the relationship between the radial margin and adverse
290
Miriam López-Gómez et al.
outcome in colon carcinoma are lacking. However, multivariate analyses of rectal disease have suggested that involvement of this margin in tumour formation may be the most critical factor in predicting local recurrence (14). The radial resection margin is considered to be involved when a tumour is present ≤ 1 mm from the surface of the specimen (12, 13). In rectal cancer, this is associated with local recurrence and indicates a need for additional therapy. Regardless of whether a colorectal tumour is classified as T3 or T4b, the excision is considered complete only if all surgical margins are negative, including the radial margin. If a radial margin is involved in a tumour, additional therapy, such as local radiation, should be considered.
c) Histologic grade Traditionally, tumours have been stratified into three or four grades as follows: Grade 1 tumours are well differentiated, Grade 2 tumours are moderately differentiated, Grade 3 tumours are poorly differentiated, and Grade 4 tumours are mostly undifferentiated. Multivariate analyses have shown histologic grade to be of independent prognostic significance, with undifferentiated tumours associated with a worse prognosis (15, 16, and 17). However, in a number of studies, the number of grades has been reduced and tumours are classified as either low or high grade tumours. It is hoped that these new grading systems will reduce interobserver variability. Therefore, tumours in the low grade include well differentiated or moderately differentiated tumours and poorly differentiated or undifferentiated tumours are considered high grade tumours (18).
d) Tumour border configuration The growth pattern of the tumour at the advancing edge (tumour border) has been shown to have prognostic significance independent of stage. This pattern may also predict metastasis to the liver. In both univariate (19) and multivariate (20) analyses, an “irregular, infiltrating pattern of growth”, as opposed to a “pushing border”, has been demonstrated to be an independent adverse prognostic factor.
III. Category IIB: Factors that are well studied but lack sufficient evidence for inclusion in Categories I or IIA a) Lymphocytic infiltration of tumour or peritumoural tissue Lymphocytic infiltration of a tumour is indicative of an immunologic response to the invasive malignancy and has been shown by some studies to
Prognostic and predictive factors in colorectal cancer
291
be a favorable prognostic factor (21). By contrast, other studies have failed to confirm this prognostic significance (22) or have demonstrated it only by univariate analyses (23). One type of lymphoid reaction is characterised by direct lymphocytic infiltration of the tumour, also known as “tumour infiltrating lymphocytes”, and is associated with tumours that carry DNA mismatch repair gene mutations and numerous DNA replication errors (RER+) together with microsatellite instability (MSI). Because MSI is associated with improved prognosis, tumour-infiltrating lymphocytes may prove to be a favourable prognostic marker (24). MSI is classified as high (MSI-H) or low (MSI-L) based on the number of markers that exhibit instability. MSI-H is defined as instability in two or more loci, while MSI-L describes a tumour with instability at a single locus (25). Microsatellite stability is defined as 0% unstable loci. Recent studies confirm the association between MSI and a significantly better prognosis for both OS and disease-free survival in CRC patients compared to those with intact mismatch repair (26, 27). Nevertheless, its predictive value for chemosensitivity remains controversial, and further studies are needed (28).
b) Histologic types Signet cell-type adenocarcinoma and neuroendocrine (small cell) carcinoma: The signet ring cell type of adenocarcinoma and neuroendocrine carcinoma are subtypes of colonic carcinoma that have been demonstrated by multivariate analysis to have an independent adverse impact on prognosis (29). These tumours are respectively assigned grade 3/4 (poorly differentiated) and grade 4/4 (undifferentiated). Therefore, both of these types of tumour are considered high grade and associated with an unfavourable prognosis.
c) Tumour tissue molecular markers Several molecular markers in colorectal carcinoma have been identified, but their clinical relevance remains unconfirmed. To establish their role as prognostic factors, they must be further evaluated by prospective randomised trials. 18q/DCC. The allelic loss of a region on the long arm of chromosome 18 is commonly observed in CRC; it is also known as “Loss of heterozygosity” (LOH). This region contains several tumour suppressor gene(s), the best known of which is the DCC gene. LOH at this site inhibits expression of the encoded protein, and some previous studies have identified it as an adverse
292
Miriam López-Gómez et al.
prognosis factor for Stage II colorectal adenocarcinoma (30). Stage II patients with 18q LOH behave clinically as stage III patients and, by contrast, stage II patients without 18q LOH behave as stage I patients. A recent meta-analysis revealed that the results from studies investigating the relationship between CRC survival and chromosome 18q allelic imbalance (AI)/loss of DCC expression (LOE) have been inconsistent. Considerable variation exists in the techniques and the assessment methods used in these reports. Nevertheless, it appears that cancers with chromosome 18q loss have a poorer prognosis (31). Further studies are required to determine the clinical utility of this marker, which may allow the molecular staging of stage II patients into two separate categories, good and poor prognosis. However, current data do not support classification of chromosome 18q AI as a marker of survival and it should not be considered outside clinical trials (32). K-ras. Ras mutations occur early in the development of colorectal carcinoma, and are present among 12–75% of colorectal carcinomas. The majority of K-ras mutations in colon adenocarcinoma affect codons 12 and 13. Results regarding the prognostic role of K-ras mutation are controversial. Ras mutations often occur in cancers with other poor prognostic factors. For example, there is a significantly higher mutation rate (65%) observed in tumours with lymphatic or hematogenous metastases compared with tumours lacking these features (33). Likewise, only 28% of carcinomas limited to the muscularis propria (Tis, T1 and T2) contain mutated ras compared with 41% of deeply invasive tumours (T3 and T4) (34). Therefore, there is a trend toward the acceptance of ras mutation as indicative of a worse prognosis. Moreover, the prevalence of kras mutations in codons 12 and 13 occurs in 25% of patients with non-recurrent disease versus 71% in patients with recurrent tumours. (33). Nevertheless, no consensus has been reached regarding its prognostic role. Although the large multicentre RASCAL study failed to show an association between the presence or absence of K-ras mutations and Dukes’ stage (35), multivariate analysis suggested that the presence of mutation increased the risk of recurrence and death (35). In addition, the FOCUS study reported that mutation in either KRAS or BRAF was a poor prognostic factor for OS (hazard ratio (HR): 1.40; 95% CI: 1.00 to 1.36; p=0.05) (36). Meanwhile, other studies do not support the prognostic value of KRAS; results from the PETACC-3 showed that KRAS mutations had no major prognostic value regarding DFS or OS (37). Further studies will be required to establish the prognostic role of K-ras mutations in colorectal adenocarcinoma. Microsatellite instability (MSI). Mutations in one of several mismatch repair (MMR) genes leads to the development of hereditary nonpolyposis CRC (HNPCC) and are responsible for 15–20% sporadic colon carcinomas
Prognostic and predictive factors in colorectal cancer
293
(38). Individuals with HNPCC inherit a mutation in one allele of one of the MMR genes and acquire a second somatic mutation in the same gene in tumours with MSI. MSI corresponds to alterations in the length of simple, repetitive microsatellite sequences that occur throughout the genome and appears to predict improved patient survival (39). Results from a recent metaanalysis confirmed an association between MSI and a favourable prognosis based on both the overall and disease-free survival of CRC patients (26). Thymidylate synthase. Thymidylate synthase (TS) converts deoxyuridine monophosphate (dUMP) to deoxythymidine monophosphate (dTMP), a step essential for DNA synthesis. TS is also an important target of fluoropyrimidine drugs that are widely used in the treatment of colon carcinoma, a feature that will be discussed later in the chapter. Regarding its prognostic role, a number of studies have investigated the relationship between thymidylate synthase (TS) expression and survival in CRC patients. Most of these studies have reported poorer overall and progression-free survival with high TS expression, but the methodology used in these studies was not consistent (40). Therefore, additional studies are needed to define the precise prognostic value of TS. p27. p27 is a cyclin dependent kinase inhibitor that acts as a cell cycle inhibitor and a potential tumour suppressor. Previous data showed that decreased expression levels of p27 in colorectal carcinoma correlated with poor prognosis (risk ratio for death = 2.9; 41). Recent studies corroborated this theory: Bertagnolli et al showed that patients with p27 negative tumours had a reduced OS [66% 5-year OS (95% CI: 0.59–0.72) versus 75% 5-year OS (95% CI: 0.70–0.79); 42]. Therefore, loss of p27 appears to correlate with reduced survival in stage III colon cancer. Bcl-2. Bcl-2 belongs to a protein family involved in regulating cell death and survival. Bcl-2 suppresses programmed cell death (apoptosis), conferring a survival advantage for these cells. The loss of Bcl-2 expression appears to be correlated with an increased number of relapses in stage II CRC, and could be therefore a potentially useful marker (43). Moreover, the expression of Bcl-2 has been shown to be a valuable indicator of good prognosis in CRC in the distal colorectum (44). Nevertheless, further studies are needed. p53. The gene that encodes p53 is located on the short arm of chromosome 17. p53 is also known as the “guardian of the genome” because it plays a role in controlling cell proliferation, cell differentiation, DNA repair and synthesis and programmed cell death (45). When the normal growth regulatory activity of the protein is lost, genetic instability is promoted. This instability diminishes the probability of apoptosis and contributes to unregulated cell growth (46).
294
Miriam LĂłpez-GĂłmez et al.
Regarding the prognostic role of p53, a distinction exists between overexpression of the protein and the presence/absence of a mutation. Reports regarding the prognostic significance of p53 overexpression are controversial. Although some studies have shown a significant correlation between high p53 levels and prolonged disease-free survival, especially among stage III cancer patients (47) other studies failed to show a relationship between p53 expression and prognosis (48). The disparity of these results is due to methodological problems with respect to the measurement of p53, which is usually assessed by immunoassay, a technique that has several limitations in the interpretation of the results (49). Regardless, it is agreed that p53 mutation and allelic loss represent useful markers for prognosis, and it is thought that tumours harbouring p53 mutations are at higher risk of metastasis (50).
Category III: Factors not yet studied sufficiently to meet criteria for inclusion in Categories I, IIA or IIB Factors grouped in this category include DNA content, other molecular markers (except loss of heterozygosity 18q/DCC and MSI-H), perineural invasion, microvessel density, tumour cell-associated proteins or carbohydrates, peritumoural fibrosis, peritumoural inflammatory response, focal neuroendocrine differentiation, nuclear organizing regions, and proliferation indices. All of these factors lack sufficient data for specific recommendations.
Category IV: Factors that are well studied and have been shown to have no prognostic significance To our knowledge, no evidence of an association between tumour size and outcome has been reported (51). Regarding histologic type, only a single multivariate analysis has demonstrated that mucinous carcinoma is an independent predictor of adverse outcome (52). Therefore, neither tumour size nor histologic tumour type is of prognostic significance in patients with colorectal carcinoma.
2. Predictive factors In CRC, clear evidence exists that adjuvant chemotherapy improves overall and disease-free survival in patients with resected Dukesâ&#x20AC;&#x2122; stage C. 5-Fluorouracil (5-FU)-based chemotherapeutic regimens are the standard
Prognostic and predictive factors in colorectal cancer
295
treatment for these patients. However, the response rates for patients given 5-FU as a single-first line treatment in advanced CRC are only 10–15% (53). Combining 5-FU with newer chemotherapies, such as irinotecan (CPT-11) and oxaliplatin, has improved the response rates of patients with advanced CRC to 40–50% (54, 55). Despite these improvements, new therapeutic strategies are needed. The use of novel biological agents, such as the monoclonal antibodies, cetuximab (an anti epidermal growth factor receptor (EGFR) inhibitor) and bevacizumab (a Vascular Endotelial Grow Factor Receptor (VEGFR) inhibitor), have recently been shown to provide additional clinical benefit to patients with metastatic CRC (56, 57). These agents are now under intense investigation as adjuvant therapies. Resistance to chemotherapy limits the effectiveness of current cancer strategies, including those used to treat CRC. Drug resistance is either intrinsic or acquired during treatment, and is believed to be the cause of treatment failure in over 90% of patients with metastatic cancer. Furthermore, drug resistant micrometastatatic tumor cells are likely to reduce the effectiveness of adjuvant chemotherapy following surgery. Therefore, overcoming drug resistance is one of the main challenges of current cancer research. There are several factors that can affect drug sensitivity: the pharmacokinetic profile of the drug, drug activation and inactivation, alterations in the drug target, the processing of drug-induced damage and the evasion of apoptosis. Currently, several studies are underway that are designed to identify critical molecular and/or biochemical markers that can be used to predict response to chemotherapy. The primary aim of such predictive biomarker testing is to allow treatment to be formulated according to the molecular phenotype of the patient tumour.
3. 5-Fluorouracil 5-FU is converted into several active metabolites, the most important of which include 2’-deoxy-5-fluorouridine-5’monophosphate (FdUMP), 2’-deoxy-5-fluorouridine-5’triphosphate (FdUTP), and 5’-fluorouridine 5’triphosphate (FUTP). 5-FU is converted to 5’fluouridine-5’monophosphate (FUMP) either directly by orotate phosphoribosyl transferase (OPRT) or indirectly via 5-fluorouridine by the sequential action of uridine phosphorylase (UP) and uridine kinase (Figure 1).
296
Miriam López-Gómez et al.
Capecitabine DPD
5-FU
LV 5’DFUR
Degradation
Cell membrane
5-FU
TP
FUDR
CH2THF dUMP
DPD
FUMP
FdUMP OPRT
Degradation FUDP
FdUDP
FUTP
TS dUTPase
Nucleotide pools imbalance
dUTP misincorporation
FdUTP RNA damage
dUTP
DNA damage
Figure 1. Mechanism of action of 5-FU. Abbreviations: 5-FU=5 Fluorouracil; CH2THF= 5,10-methylenetetrahydrofolate; DPD=dihydropyrimidine dehydrogenase; dUTP= deoxyuridine triphosphate; dUTPase=dUTP pyrophosphatase; 5’DFUR= 5’deoxy-5’fluorouridine; FdUDP=5’-fluoro-2’deoxyuridine-5’monophosphate; FDUR= 2’deoxy-5-fluorouridine; FUMP=5’-fluorouridine-5’triphosphate; LV=leucovorin; TP=thymidine phosphorylase; TS=thymidylate synthase.
Thymidine phosphorylase Thymidine phosphorylase (TP) catalyses the conversion of 5-FU to 2’-deoxy-5-fluorouridine (FUDR). TP overexpression has been correlated with increased sensitivity to 5-FU, possibly due to enhanced formation of FdUMP (58). By contrast, results from analysis of TP messenger RNA (mRNA) expression in colorectal tumours showed that tumours with high TP expression were less likely to respond to 5-FU (59). The authors hypothesised that TP acts as an angiogenic factor, based on the finding that it contains a sequence identical to that of platelet-derived endothelial cell growth factor, a well-documented angiogenic factor. Therefore, elevated TP expression might be a marker of a more invasive and malignant tumour phenotype with increased resistance to chemotherapy.
Prognostic and predictive factors in colorectal cancer
297
Thymidylate synthase FdUMP is a metabolite of 5-FU and forms a stable ternary complex with the folate-dependent enzyme TS and the reduced folate cofactor 5,10methylenetetrahydrofolate. The formation of this complex results in inhibition of TS enzyme activity, leading to an imbalance in deoxynucleotide pools and subsequent inhibition of DNA synthesis and repair. Co-administration of Leucovorin increases the intracellular levels of this complex, thereby enhancing and stabilising the ternary complex formation and optimising TS enzyme inhibition. 5-FU can also cause DNA damage through misincorporation of FdUTP into DNA. Furthermore, FUTP can be misincorporated into nuclear and cytoplasmic RNA (Figure 1). Numerous studies have demonstrated that TS is a key determinant of sensitivity to 5-FU. High TS expression correlates with increased resistance to 5-FU (60, 61). A meta-analysis performed by Popat et al analysed 20 studies and included over 3,000 patients, and concluded that tumours with elevated TS expression had poorer OS compared with tumours that expressed low levels of TS (62). Similar results have been shown in metastatic disease: detection of TS activity in liver metastasis was linked to 5-FU activity and is related to clinical responsiveness to 5-FU (63). Because TS expression is critical for the efficacy of fluoropyrimidines, identification of the genetic alterations that regulate TS gene expression should be crucial for developing predictive markers. However, the results are contradictory. In some studies, low TS levels were associated with improved survival, which might indicate that these tumours are more sensitive to 5-FU-based therapy (64). By contrast, other larger studies concluded that tumours with elevated TS levels are more likely to benefit from 5-FU-based chemotherapy (65, 66). The heterogeneity observed between studies of TS expression might reflect differences in the method of assessing TS. The optimal method by which to measure TS expression remains unclear. Although real-time PCR is more specific than immunohistochemistry, as it allows a quantitative determination of TS expression, its widespread use has limited applicability due to the requirement for fresh tissue. Immunohistochemical analysis of TS expression is more efficient and clinically applicable, although it only allows for a semiquantitative determination of TS expression. Therefore, results from studies that use different TS measurement techniques cannot be compared. Dihydropyrimidine dehydrogenase (DPD) DPD is the rate-limiting enzyme in 5-FU catabolism. More than 80% of an administered dose of 5-FU is normally catabolised by the liver.
298
Miriam López-Gómez et al.
Therefore, variation in the expression levels of DPD has a direct effect on the bioavailability of 5-FU, as higher levels result in increased 5-FU metabolism and decreased levels of the drug (67). Conversely, patients who possess inactivating mutations of the DPD gene are deficient in DPD enzyme activity and cannot degrade 5-FU. As a result, these patients experience severe gastrointestinal and hematological toxicity when exposed to 5-FU (68). Regarding its predictive value, an inverse correlation has been reported between tumour DPD expression and response to 5-FU and capecitabine. Kobayashi et al reported that high DPD activity correlated with low chemosensitivity to 5-FU, while Tsuji et al found that low tumor DPD expression in patients with stage II and III CRC correlated with better response to oral 5-FU chemotherapy; 69, 70). Salonga et al reported that measuring tumour DPD levels in conjunction with TS levels significantly increased the ability to predict responsiveness to 5-FU-based chemotherapy (71). The results of these studies demonstrate that while the utility of DPD as a marker of toxicity has been firmly established, its role in predicting patient response to 5-FU needs to be further defined (72). Capecitabine Capecitabine is an oral fluoropyrimidine that is not degraded by DPD in the gut mucosa (73). Instead, capecitabine is absorbed intact through the gastrointestinal wall and is then converted to 5’deoxy-5 fluoridine (DFUR) in the liver by the sequential activity of carboxylesterase and cytidine deaminase (Figure 1). Initially, the conversion of DFUR to 5-FU was attributed solely to TP. However, recent studies suggest that another enzyme, uridine phosphorylase (UP), might play an important role in capecitabine activation (74). Therefore, it would be expected that capecitabine would be most effective in tumours expressing high levels of TP and/or UP. Indeed, among patients with stage II and III CRC treated with oral capecitabine, those with high TP and low DPD expression had the best OS (75).
4. Irinotecan Irinotecan (CPT-11) is a camptothecin analog that improves median survival in advanced CRC compared to 5-FU alone (76). It targets DNA topoisomerase I (Topo I) causing inhibition of DNA replication and subsequent cell death (Figure 2). To exert its toxic effect, irinotecan must be converted to SN-38 (7 ethyl10-hydroxy-camptothecan) by carboxylesterase (CES, an enzyme found mainly in serum, the liver, and intestine). SN-38 is detoxified in the liver, primarily
Prognostic and predictive factors in colorectal cancer
299
CPT-11 EGF
CES
Cell membrane EGFR SN-38 UGT1A
Topo I
B-Gluc
SN38/Topo 1/ss break
Decreased Topo I expression/mutation
SN-38G
DNA repair
SN38/Topo 1/ds break
SN-38G inactivation
SN-38G Carboxylate
Degradation
Lactone
DNA damage
Figure 2. Activation and metabolism of Irinotecan. Abbreviations: β Gluc = betaglucuronidase; CES = carboxylesterase; SN-38= 7-ethyl-10hydroxy-camptothecan; SN-38G: SN-38 glucuronide; ds = double strand; Topo I= topoisomerase I; ss = single strand; UGT = uridine diphosphate glucuronosyltransferase.
by the 1A1 isoform of uridine diphosphate glucuronosyltransferase (UGT). UGT catalyses the conversion of SN-38 to SN-38G (SN-38 glucuronide), which is excreted in bile and urine (77). Clinical observation revealed that the use of irinotecan is associated with significant toxic effects (irinotecan-related mortality has been reported in up to 7% of patients), the most frequent of which is grade 4 diarrhea, which can be life-threatening (78). The high mortality rate associated with irinotecan use has spurred several efforts to identify the responsible molecular mechanisms. To this end, a polymorphic variant in UGT1A1 has been identified. UGT1A1*28 is associated with a significant decrease in SN-38 glucuronidation, resulting in reduced SN-38 detoxification (79). Patients that exhibit this UGT1A1 polymorphic variant are associated with increased gastrointestinal and bone marrow toxicity (80). Similarly, as CPT-11 requires CES to become active, high levels of CES are associated with increased grade 3/4 neutropenia (81). Just as is the case with 5-FU, the identities of reliable markers for predicting the efficacy of irinotecan in patients with CRC need to be clearly
300
Miriam L贸pez-G贸mez et al.
defined. The expression of Topo-I, the target enzyme of 5-FU, is currently being investigated as a potential marker for response to irinotecan. Studies in human CRC xenografts have shown that the levels of Topo-I complexes are predictive of response to irinotecan (82). However, the clinical value of Topo-I as a predictive marker for response to irinotecan has yet to be definitively demonstrated, as some studies have shown that Topo-I expression levels in patients receiving 5-FU/irinotecan had no influence on response, time to progression or overall patient survival (83).
5. Oxaliplatin Oxaliplatin is a third generation platinum analog in which the amine groups of cisplatin are substituted by a 1,2-diaminocyclohexane (DACH) ligand. Cytotoxic platinum compounds form positively charged compounds that block DNA replication and transcription and may initiate apoptosis (84). Once oxaliplatin induces DNA damage, the cell attempts to repair the damage through the Nucleotide Excision Repair (NER) pathway (85) (Figure 3). Oxaliplatin ABC superfamily
Cell membrane Drug Efflux
Oxaliplatin GST-P1
Detoxification by GSH
Pt-DNA adducts
XRCC1 Intra/extra strand crosslinks
DNA repair
ERCC1 XPD
Enhanced nucleotide excision repair
Survival
DNA damage
Figure 3. Activation and metabolism of Oxaliplatin. Abbreviations: ERCC1 = excision repair cross complementing protein 1; XRCC1 = x-ray repair cross complementing 1; XPD = xeroderma pigmentosum group D; GSH = glutathione; GST = glutathione-S-transferase; pt = platinum.
Prognostic and predictive factors in colorectal cancer
301
The protein ERCC1 is a member of the NER pathway. ERCC1 forms a complex with xeroderma pigmentosum group F (XPF), which recognises and splits the damaged DNA strand, repairing damage to the DNA. DNA repair is an important mechanism for resistance to platinum-based chemotherapy. Therefore, several studies have been performed to establish a correlation between the expression of ERCC1 and clinical outcome after oxaliplatin therapy. It has been shown that patients with elevated ERCC1 levels do not respond well to platinum therapy (86). Another study demonstrated that the level of ERCC1 mRNA expression was an independent predictive marker of survival in 5-FU/oxaliplatin chemotherapy (p<0.001; 87). Interestingly, polymorphic variants within the ERCC1 gene are also associated with clinical outcome in patients receiving 5-FU/oxaliplatin (88). Xeroderma pigmentosum group D (XPD) is also important in the NER pathway. A polymorphic variant of this protein has been correlated with lower response rates in patients receiving 5-FU/oxaliplatin (p=0.02; 89). Finally, another family of enzymes important for oxaliplatin metabolism are glutathione-S-transferases (GST), which play a key role in the detoxification of oxaliplatin. Some studies have identified polymorphisms in GST enzymes that show a correlation with response to platinum agents (90).
6. Cetuximab and panitumumab The EGFR is a member of the human epidermal growth factor receptor (HER)-erbB family of receptor tyrosine kinases. Its activation stimulates a cascade that enhances tumour growth and progression, including proliferation, angiogenesis, invasion and development of metastasis. Whereas tyrosine kinase inhibitors have little effect in CRC, EGFRtargeted monoclonal antibodies play an important role. Cetuximab, a chimeric mouse-human monoclonal antibody, was the first anti-EGFR therapy approved for CRC treatment (91). Recently, panitumumab, a whole human antibody, has been incorporated into the therapeutic arsenal (92). Both appear to have similar efficacy, though panitumumab is less immunogenic. Positive EGFR protein expression determined by immunohistochemistry was long considered to be a predictive factor for anti-EGFR treatment. However, subsequent studies obtained objective responses even in patients with low or negative EGFR expression (93). As a result of these findings, intense research was done to identify alternative predictive molecular biomarkers able to help physicians identify which patients are most likely to benefit from EGFR-targeted therapy. Currently, we know that the growth of many tumours is secondary to the activation of signalling pathways downstream of the EGFR, regardless of
302
Miriam López-Gómez et al.
whether the EGFR is activated or pharmacologically inhibited. The interlinked RAS-MAPK and PI3K signalling pathways play an important role in tumourigenesis via phosphorylation of various proteins and transcription factors that directly control cell growth, differentiation and apoptosis (94). Mutations in KRAS, BRAF and PI3KCA result in constitutive activation of downstream pathways. Although KRAS mutations have been explored as prognostic biomarkers, they do not seem to have a stage-specific prognostic value. Nevertheless, the role of KRAS as a predictive marker has been strongly established. Lievre et al first were the first to report a lack of responsiveness to anti-EGFR therapy in tumours with KRAS mutations (95). A pivotal randomised phase III study of panitumumab in monotherapy was the first to confirm the negative predictive value of KRAS mutations (96). Results from the CRYSTAL and the OPUS studies, in which cetuximab was combined with irinotecan or oxaliplatin, respectively, confirmed that patients carrying KRAS mutations did not benefit from anti-EGFR therapy (97, 98). BRAF is another downstream molecule in the EGFR signalling pathway. BRAF and KRAS mutations are mutually exclusive. A recently published retrospective analysis of patients who received panitumumab or cetuximab showed that those with tumours that carried BRAF V600E mutations did not respond to EGFR inhibition and had a statistically significant shorter progression-free interval and OS than patients with tumours with wild-type BRAF (99). PIK3CA mutations and loss of PTEN expression also appear to confer resistance to cetuximab, as revealed by in vitro studies (100). In a clinical analysis, Sartore-Bianchi et al (101) found that PIK3CA mutations and PTEN loss in colorectal cancers were significantly associated with a lack of response to panitumumab or cetuximab. PIK3CA mutations and loss of PTEN expression was also associated with progression-free survival and poorer OS. Nevertheless, further investigation and prospective data are required to confirm these findings before they can be integrated into clinical practice. The characteristic “acneiform” skin rash observed in patients treated with EGFR inhibitors has also been studied as a potential predictive marker (103). Skin toxicity has been linked with higher response rates and longer survival among patients treated with panitumumab (92) or cetuximab (93). However, there are several limitations on the use of rash as an early marker of efficacy, as there are no criteria for evaluating toxic effects involving skin rash in EGFR-targeted treatment. Some authors highlight the fact that because EGFR is expressed in the skin, the rash may indicate local receptor saturation, although other factors, such as immune status, might alter an individual’s susceptibility to rash.
Prognostic and predictive factors in colorectal cancer
303
7. Bevacizumab The VEGF pathway plays an important role in tumour growth and angiogenesis. The formation of new blood vessels carrying oxygen, nutrients, growth factors and hormones is indispensable for the proliferation of tumour cells. VEGF has become a common target for therapeutic intervention. Bevacizumab is a recombinant humanised monoclonal antibody that targets VEGF and significantly increases both overall and progression-free survival in metastatic CRC (57). Therefore, the identification of biomarkers that can help the clinician to identify patients who will benefit from antiangiogenic therapy is of great interest. Preclinical data suggest that dysregulation of the Ras/Raf/Mek and p53 pathways may have some role in this setting; it has been suggested that tumour cells deficient in p53 experience decreased apoptosis in hypoxic tissue, which might explain their increased responsiveness to antiangiogenic therapy (103). Other studies have focused on the role of potential predictors of response to bevacizumab, such as microvessel density (MVD), thrombospondin (THBS), and VEGF, but no statistical difference in OS was found (104). Elevated baseline levels of Ca 19.9 have shown to have a predictive value, as patients with increased expression benefited significantly from bevacizumab treatment (105). In addition, other studies have evaluated the role of bevacizumab induced hypertension as a marker of bevacizumab efficacy. A study by Ryanne et al showed that patients who developed any grade of hypertension while on bevacizumab treatment had an adjusted HR for death of 0.32 (p=0.03) compared to those without hypertension (106).
Conclusion Despite recent efforts to identify individual genes for predicting response to therapy and disease prognosis in colorectal cancer, a definitive list of predictive and prognostic markers still does not exist. As the response to treatment and disease progression is the result of complex pathways and not single molecules, the analysis of individual factors should be supplemented by the identification of gene expression profiles that can predict the response to chemotherapy and help to classify patients according to their predictive outcome. To date, there are only retrospective analyses, and the design of prospective trials is crucial. Identification of these factors would provide
304
Miriam López-Gómez et al.
clinicians the ability to tailor therapeutic intervention to the pharmacogenetic profile of the patient, which would significantly improve the treatment of CRC.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
Ahmedin Jemal, Rebecca Siegel, Jiaquan Xu and Elizabeth Ward. 2010, . Cancer J Clin 2010 doi: 10.3322/caac.20073. Compton CC, Fielding P, Burgart LJ, Conley B, Cooper HS, Hamilton SR, et al. 1999, Arch Pathol Lab Med,124, 979-994. American Joint Committee on Cancer. Cancer staging manual. 1st edition. 1977. Sheperd N, Baxter K, and Love S. 1997, Gastroeneterology, 112, 1096-1102. Zeng Z, Cohen AM, Hajdu S, Sternberg SS , Sigurdson ER, and Enker W. 1992, Cancer, 70, 737-40. Ratto C, Soío L, Ippoliti M, Merico M, Bossola M, Vecchio FM, et al. 1999, Dis Colon Rectum., 42, 143-158. Ishizuka M, Nagata H, Takagi K, and Kubota K. 2010, Eur Surg Res., 7, 46(2), 57-64. Muller S, Chesner IM, Egan MJ, Rowlands DC, Collard MJ, Swarbrick ET, Newman J. 1989, Gut, 30, 1385-91. Cooper HS, Deppisch LM, Gourley WK, Kahn EI, Lev R, Manley PN, et al. 1995, Gastroenterology, 108, 657-65. Onetto M, Paganuzzi M, Secco GB, Radelli R, Santi F, Rovida S et al.1985, Biomed Pharmacother, 39, 392-5. Harrison LE, Guillem JG, Paty P, Cohen AM. 1997, J. Am Coll Surg, 185, 55-9. Quirke P, Durdey P, Dixon M, and Williams NS. 1986, Lancet, 2, 996-8. Quirke P, and Scott N. 1992, Surg Oncol Clin N Am., 3, 1-17. Adam IJ, Mohamdee MO, Martin IG, Scott N, Finan PJ, Jonhston D et al. 1994, Lancet, 344, 707-11. Chapuis PH, Dent OF, Fisher R, Newland RC, Pheils MT, Smyth E et al. 1985, Br J Surg, 72, 698-702. Scott NA, Wieand HS, Moertel CG, Cha SS, Beart RW, and Lieber MM. 1987, Arch Surg., 122, 1375-9. Freedman L, Macaskill P, and Smith A. 1984, Lancet, 2, 733-6. Blenkinsopp WK, Stewart-Brown S, Blesovsky L, Kearney G, and Fielding LP. 1981, J Clin Pathol., 34, 509-13. Thynne GS, Weiland LH, Moertel CG, and Silvers A. 1980, Mayo Clin Proc., 55, 243-5. Jass JR. 1986, J Clin Pathol., 39, 585-9. Halvorsen TB, and Seim E. 1989, J Clin Pathol., 42, 162-6. Roncucci L, Fante R, Losi L, DiGregorio C, Micheli A, Benatti P et al. 1996, Eur. J Cancer, 32ª, 295-302. Carlon CA, Fabris G, Arslan –Pagnini C, Pluchinotta AM, Chinelli E, and Carniato S. 1985, Dis Colon rectum., 28, 47-50.
Prognostic and predictive factors in colorectal cancer
305
24. Jass JR, Do K-A, Simms LA, Lino H, Wynter C, Pillary SP et al. 1998, Gut, 42, 673-9. 25. Boland Cr, Thibodeau SN, Hamilton Sr, Sidransky D, Eshleman JR, Burt RW, et al. 1998, Cancer Res, 52, 48-57. 26. Guastadisegni C, Colafranceschi M, Ottini L, Dogliotti E.2010, Eur J Cancer, 46(15), 2788-98. 27. Popat S, Hubner R, and Houlston RS. 2005, J Clin Oncol., 23(3), 609-618. 28. Des Guetz G, Uzzan B, Nicolas P, Schischmanoff O, and Morere JF. 2009, Target Oncol., 4(1), 57-62. 29. Ofner D, Riedmann B, Maier H, Hittmair A, Rumer a, Totsch M et al. 1995, Patho, 175,441-8. 30. Jen J, Kim H, Piantadosi S, Liu A-F, Levitt RC, Skistonen PK et al. 1994, N Engl J Med, 331, 213-21. 31. Popat S, and Houlston RS. 2005, Eur J Cancer., 41(14), 2060-70. 32. Popat S, Zhao D, Chen Z, Pan H, Shao Y, Chandler I, and Houlston RS. 2007, Anticancer Res., 27(1B), 627-33. 33. Finkelstein SD, Sayegh R, Bakker A, and Swalsky P. 1993, Arch Surg., 128, 526-32. 34. Cerottini JP, Caplin S, Sarage E, Givel JC, and Benhattar J. 1998, Am J Surg., 175, 198-202. 35. Andreyev HJN, Norman AR, Cunningham D, Oates J, Dix BR, Iacopetta BJ et al. 2001, Br J Cancer, 85(5), 692-696. 36. Richman SD, Seymour MT, Chambers P, Elliott F, Daly CL, Meade AM et al. 2009, J Clin Oncol, 27 (25), 5931-5937. 37. Roth AD, Tejpar S, Delorenzi M, Yan P, Fiocca R, Klingbiel D et al. 2010, J Clin Oncol, 28(3), 466-474. 38. Shibata D, Peinado MA, Ionov Y, Malkhosyan S, and Perucho M. 1994, Nat Genet., 6, 263-81. 39. Boland CR and Goel A. 2010, Gastroenterology, 138(6), 2073-2087.e3. 40. Popat S, Matakidou A, and Houlston RS. 2004, J Clin Oncol., 22(3), 529-36. 41. Loda M, Cukor B, Tam SW, Lavin P, Florentino M, Draetta GF et al. 1997, Nat Med, 3,331-4. 42. Bertagnolli MM, Warren RS, Niedzwiecki D, Mueller E, Compton CC, and Redston M. 2009, Clin Cancer Res., 15(6), 2116-22. 43. Poincloux L, Durando X, Seitz JF, Thivat E, Bardou VJ, and Giovannini MH. 2009, Surg Oncol., 18(4), 357-65. 44. Manne U, Weiss HL, and Grizzle WE. 2000, Int J Cancer., 89(5), 423-30. 45. Levine AJ, Momand J, and Finlay CA. 1991, Nature, 315, 453-5. 46. Hartwell L. 1992, Cell, 71, 543-6. 47. Adrover E, Maestro ML, Sanz-Casla MT, del Barco V, Cerdรกn J, and Fernรกndez C. 1999, Br J Cancer, 81(1), 122-6. 48. Starzynska T, Bromley M, Marlicz K, Roberts SA, Ucinski M, and Stern PL. 1997, Eur J Gastroenterol Hepatol., 9, 183-6. 49. Lotter O, Amr A, Safi FG. 2010, Med Sci., 8, Doc24. 50. Hsieh JS, Lin SR, Chang MY, Chen FM, Lu CY, and Huang TJ. 2005, Am Surg., 71(4), 336-43.
306
Miriam L贸pez-G贸mez et al.
51. Griffin MR, Bergstralh EJ, Coffey RJ, Beart RW, and Melton LJ. 1987, Cancer, 60, 2318-24. 52. Carlon Ca, Fabris G, Arslan-Pagnini C, Pluchinotta AM, Chinelli E, and Carniato S. 1985, Dis colon Rectum, 28, 47-50. 53. Jonhston PG, Kaye S. 2001, Anti-cancer drugs, 12(8), 639-46. 54. Douillard JY, Cunningham D, Roth AD, Navarro M, James RD, and Karasek P. 2000, Lancet, 355, 1041-1047. 55. Giacchetti S, Perpoint B, Zidani R, Le Bail N, Faggiuolo R, and Focan C. 2000, J. Clin. Oncol., 18, 136-147. 56. Cunningham D, Humblet Y, Siena S, Khayat D, Bleiberg H, and Santoro A. 2004, N. Engl. J. Med., 351, 337-345. 57. Hurwitz H, Fehrenbacher L, Novotny W, Cartwright T, Hainsworth J, and Heim W. 2004, N. Engl J. Med., 351, 337-345. 58. Evrard A, Cuq P, Ciccolini J and Cano JP. 1999, Br J Cancer, 80(11), 1726-1733. 59. Metzger R, Danenberg K, Leichman CG, Salonga D, Schwartz EL, and Wadler S. 1998, Clin Cancer Res., 4, 2371-2376. 60. Johnston PG, Lenz HJ, Leichman CG, Danenberg KD, Allegra CJ, and Danenberg PV. 1995, Cancer Res., 55, 1407-1412. 61. Cascinu S, Graziano F, Valentini M, Catalano V, Giordani P, and Staccioli MP. 2001, Ann Oncol., 12, 239-244. 62. Popat S, Matakidou A, and Houlston RS. 2004, J Clin Oncol., 22, 529-536. 63. Etienne MC, Chazal M, Laurent-Puig P, Magn茅 N, Rosty C, and Formento JL. 2002, J Clin Oncol., 20, 2832-2843. 64. Edler D, Kressner U, Ragnhammar P, Johnston PG, Magnusson I, and Glimelius B. 1997, Eur J Cancer, 33, 2278-2281. 65. Edler D, Glimelius B, Hallstr枚m M, Jakobsen A, Johnston PG, and Magnusson I. 2000, J. Clin Oncol., 20, 1721-1728. 66. Johnston PG, Fisher ER, Rockette HE, Fisher B, Wolmark N, and Drake JC. 1994, J Clin Oncol., 12, 2640-2647. 67. Diasio RB, and Harris BE. 1989, Clin Pharmacokinet., 16, 215-237. 68. Lyss AP, Lilenbaum RC, Harris BE, and Diasio RB. 1993, Cancer Invest., 11, 239-240. 69. Kobayashi N, Kubota T, Watanabe M, Otani Y, Teramoto T, and Kitajima M. 1999, J Infect Chemother., 5, 144-148. 70. Tsuji T, Sawai T, Takeshita H, Nakagoe T, Hidaka S, and Yamaguchi H. 2004, Cancer Chemother. Pharmacol., 54, 531-536. 71. Salonga D, Danenberg KD, Johnson M, Metzger R, Groshen S, and Tsao-Wei DD. 2000, Clin Cancer Res., 6, 1322-1327. 72. Van Kuilenburg AB. 2004, Eur J Cancer, 40, 939-950. 73. Johnston PG, and Kaye S. 2001, Anticancer Drugs, 12, 639-646. 74. Cao D, Russell RL, Zhang D, Leffert JJ, and Pizzorno G. 2002, Cancer Res., 62, 2313-2317. 75. Nishimura G, Terada I, Kobayashi T, Ninomiya I, Kitagawa H, and Fushida S. 2002, Oncol Rep., 9, 479-482. 76. Douillard JY, Cunningham D, Roth AD, Navarro M, James RD, and Karasek P. 2000, Lancet, 355, 1041-1047.
Prognostic and predictive factors in colorectal cancer
307
77. Humerickhouse R, Lohrbach K, Li L, Bosron WF, and Dolan ME. 2000, Cancer Res., 60, 1189-1192. 78. Fuchs CS. 2003, Semin Oncol., 30, 9-17. 79. Iyer L, Das S, Janish L, M Wen, J Ramírez, T Karrison, et al:. 2002, Pharmacogenomics J, 2, 43-47. 80. Iyer L, Das S, Janisch L, Wen M, Ramírez J, and Karrison T. 2002, Pharmacogenomics J., 2(1), 43-7. 81. Cecchin E, Corona G, Masier S, Biason P, Cattarossi G, and Frustaci S. 2005, Clin Cancer Res., 11, 6901-6907. 82. Lansiaux A, Bras-Goncalves RA, Rosty C, Laurent-Puig P, Poupon MF, and Bailly C. 2001, Anticancer Res., 21, 471-476. 83. Paradiso A, Xu J, Mangia A, Chiriatti A, Simone G, and Zito A. 2004, Int J Cancer, 111, 252-258. 84. Adjei AA. 1999, Br J Clin Pharmacol 48: 265-277. 85. Reardon JT, Vaisman A, Chaney SG, and Sancar A. 1999, Cancer Res., 59, 3968-3971. 86. Metzger R, Leichman CG, Danenberg KD, Danenberg PV, Lenz HJ, and Hayashi K. 1998, J Clin Oncol., 16(1), 309-16. 87. Shirota Y, Stoehlmacher J, Brabender J, Xiong YP, Uetake H, and Danenberg KD. 2001, J Clin Oncol., 19, 4298-4304. 88. Park DJ, Zhang W, Stoehlmacher J, Tsao-Wei D, Groshen S, and Gil J. 2003, Clin Adv Hematol Oncol., 1, 162-166. 89. Park DJ, Stoehlmacher J, Zhang W, Tsao-Wei DD, Groshen S, and Lenz HJ. 2001, Cancer Res., 61, 8654-8658. 90. Stoehlmacher J, Park DJ, Zhang W, Groshen S, Tsao-Wei DD, and Yu MC. 2002, J Natl Cancer Inst., 94, 936-942. 91. Cunningham D, Humblet Y, Siena S, Khayat D, Bleiberg H, and Santoro A. 2004, N Engl J Med., 351(4), 337-345. 92. Van Cutsem E, Peeters M, Siena S, Humblet Y, Hendlisz A, and Neyns B. 2007, J Clin Oncol., 25(13), 1658-1664. 93. Chung KY, Shia J, Kemeny NE, Shah M, Schwartz GK, Tse A, Hamilton A, Pan D, Schrag D, Schwartz L, Klinstra DS, Frigman D, Kelsen DP, Saltz LB. 2005, J Clin Oncol., 23(9), 1803-10. 94. Scaltriti M, and Baselga J. 2006, Clin Cancer Res., 12(18), 5268-5272. 95. Lièvre A, Bachet JB, Le Corre D, Boige V, Landi B, and Emile JF. 2006, Cancer Res., 15, 66(8), 3992-5. 96. Amado RG, Wolf M, Peeters M, Van Cutsem E, Siena S, Freeman DJ, Juan T, Sikorski R, Suggs S, Radinsky R, Patterson SD, and Chang DD. 2008, Clin Oncol., 26(10), 1626-34. 97. Bokemeyer C, Bondarenko I, Makhson A, Hartmann JT, Aparicio J, and De Braud F. 2009, J Clin Oncol., 27(5), 663-71. 98. Van Cutsem E, Köhne CH, Hitre E, Zaluski J, Chang Chien CR, Makhson A, D´Haens G, Pintér T, Lim R, Bodoky G, Roh JK, Folprecht G, Ruff P, Stroh CH, Tejpar S, Schlichting M 2009, N Engl J Med., 360(14), 1408-17.
308
Miriam L贸pez-G贸mez et al.
99. Di Nicolantonio F, Martini M, Molinari F, Sartore-Bianchi A, Arena S, Saletti P. 2008, J Clin Oncol, 26(35),5705-12. 100. Jhawer M, Goel S, Wilson AJ, Montagna C, Ling YH, Byun DS. 2008, Cancer Res., 68(6), 1953-61. 101. Sartore-Bianchi A, Martini M, Molinari F, Veronese S, Nichelatti M, Artale S. 2009, Cancer Res., 69(5),1851-7. 102. Jonker DJ, O'Callaghan CJ, Karapetis CS, Zalcberg JR, Tu D, Au HJ. 2007, N Engl J Med., 357(20), 2040-8. 103. Yu JL, Rak JW, Coomber BL, Hicklin DJ, Kerbel RS. 2002, Science, 295(5559), 1526-8. 104. Jubb AM, Hurwitz HI, Bai W, Holmgren EB, Tobin P, and Guerrero AS. 2006, J Clin Oncol., 24(2), 217-27. 105. Formica V, Massara MC, Portarena I, Fiaschetti V, Grenga I, Del Vecchio Blanco G. 2009, Cancer Biomark., 5(4), 167-75. 106. Ryanne Wu R, Lindenberg PA, Slack R, Noone AM, Marshall JL, and He AR. 2009, J Gastrointest Cancer, 40(3-4), 101-8.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 309-330 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
13. Inherited syndromes of colorectal cancer and genetic counselling 1
Ignacio Blanco1, Conxi Lázaro2 and Gabriel Capella3
Head of the Genetic Counselling Unit, Hereditary Cancer Programme, Catalan Institute of Oncology, Spain; 2Head of the Molecular Diagnostic Counselling Unit, Hereditary Cancer Programme, Catalan Institute of Oncology, Spain; 3Director of the Hereditary Cancer Programme Catalan Institute of Oncology, Spain
Abstract. CRC is the third most common cancer and the fourth most frequent cause of cancer deaths worldwide. Approximately 6% of colorectal cancers can be attributed to recognizable heritable germline mutations. The public-health implications of hereditary CRC are substantial because close relatives of an index patient could benefit from genetic counselling and, potentially, from mutation testing. Referral should be made to a cancer geneticist so that sensitive and appropriate counselling, and appropriate management and surveillance, can be offered. The two most common hereditary colorectal-cancer syndromes are Lynch Syndrome (LS) and Familial Adenomatous Polyposis (FAP). Lynch syndrome is an autosomal dominant disorder characterised by early onset of CRC with microsatellite instability. Mutations in mismatch-repair genes lead to a lifetime risk of colon cancer of 85% in these patients; carcinomas of the endometrium, ovary, and Correspondence/Reprint request: Dr. Ignacio Blanco, Head of the Genetic Counselling Unit, Hereditary Cancer Programme, Catalan Institute of Oncology, Spain. E-mail: iblanco@iconcologia.net
310
Ignacio Blanco et al.
other organs also occur with increased frequency. FAP is a well-described autosomal dominant syndrome classically presenting hundreds to thousands of adenomatous colorectal polyps that are caused by mutations in the APC gene. Adenomas typically develop in the mid-teens in these patients, and CRC is a virtual certainty if this condition is untreated. Morbidity and mortality from hereditary forms of CRC should be reduced once a patientâ&#x20AC;&#x2122;s familial or hereditary risk is established and a highly targeted programme of cancer surveillance and management is undertaken. Prevention will be aided by the identification of the causative germ-line mutation in a patientâ&#x20AC;&#x2122;s family, thus confirming the risk.
Introduction CRC is the third most common cancer and the fourth most frequent cause of cancer deaths worldwide [1]. This cancer is more common in developed than developing countries. In developed countries, it is the second most common tumour, with a lifetime incidence of 5%, but its incidence and rate of mortality are now decreasing. The worldwide variability of outcome is proportional to access to specialists and availability of modern drug therapy. Every year worldwide, more than 1 million individuals will develop CRC, and the diseasespecific mortality rate is nearly 33% in the developed world [1]. The classic description of colorectal carcinogenesis is the adenomacarcinoma sequence and multistep tumorigenesis that is determined by gatekeeper and caretaker molecular pathways, which take years to decades to develop [2]. Tumour-suppressor gatekeepers act directly to regulate cell proliferation and/or apoptosis and are rate limiting for tumorigenesis, an example being the APC gene in human colorectal tumorigenesis [3]. Caretakers, also affected by inactivating mutations in cancer, do not directly regulate proliferation. When mutated, they lead to accelerated conversion of a normal cell to a neoplastic cell via an increased frequency of mutations in other cellular genes, particularly genes that are rate-determining in tumour development. These traits imply that CRC is a genetic disease but is not necessarily inherited. All colorectal neoplasms result in part from genes that mutate somatically [4]. Only a minority of colorectal neoplasms, however, arise from an inherited gene that confers susceptibility in development [3,5]. As we have mentioned, most cases of CRC arise sporadically [6]. Risk factors include increasing age, male sex, previous colonic polyps or CRC, and environmental factors (eg, red meat, high-fat diet, inadequate intake of fibre, obesity, sedentary lifestyle, diabetes mellitus, smoking, and high consumption of alcohol). Inflammatory bowel disease (ulcerative colitis and Crohnâ&#x20AC;&#x2122;s disease) accounts for roughly two-thirds of the incidence, and the risk increases with duration of illness (2% at 10 yrs, 18% by 30 yrs) and severity and extent of inflammation.
Inherited syndromes of colorectal cancer
311
About 20% of all patients with this cancer are estimated to have some component of familial risk without fulfilling the strict criteria for hereditary CRC. Roughly 5-10% of all CRCs develop in the setting of defined hereditary cancer syndromes [3]. The most common of these is Lynch syndrome (LS) [7]. It is characterised by the development of CRC, endometrial cancer, and various other cancers and is caused by a mutation in one of the mismatchrepair (MMR) genes: MLH1, MSH2, MSH6, or PMS2 [3]. Familial adenomatous polyposis (FAP) is another well-described inherited syndrome and is responsible for <1% of all CRC cases [8]. It is characterised by the development of hundreds to thousands of adenomas in the colorectum. FAP is transmitted as an autosomal dominant trait and is caused by truncating mutations in APC. MUTYH has been identified as another gene able to cause polyposis [9]. The public-health implications of hereditary CRC are substantial, because close relatives of an index patient could benefit from genetic counselling and, potentially, from mutation testing. Referral should be made to a cancer geneticist so that sensitive and appropriate counselling, and appropriate management and surveillance, can be offered [3]. Identification of inherited susceptibility to colon cancer is now readily possible [10]. The process of identifying genetic susceptibility, though, is complicated, due to the presence of clinical phenotypes whose appearances overlap with normal variation and other genetic syndromes. Multiple genes may need to be evaluated, and the sensitivity of available mutational tests is limited (60-80%). Suboptimal sensitivity is compounded by the potential for detecting variants in gene sequences that are simply polymorphisms or variants of uncertain significance. Interpretation of genetic tests for colorectal carcinoma is not as straightforward as interpreting a blood-sugar test to ascertain whether it is elevated, normal, or lowered [10]. The wide variety of tests available, the different clinical contexts in which they may be obtained, and the difficulty in interpretation in some results highlight the complexity of the issue and the necessity of undertaking a good process of genetic counselling. In this review we will first describe the principles of genetic counselling and their applicability to CRC and then discuss the most common hereditary CRC syndromes.
1. Genetic counselling in hereditary colorectal cancer Genetic counselling is a dynamic process of communication between the patient and the counsellor who provides education and support within a multidisciplinary team. The National Society of Genetic Counselors states that
312
Ignacio Blanco et al.
â&#x20AC;&#x153;the purpose of cancer genetic counseling is to educate clients about their chance of developing cancer, help them derive personal meaning from cancer genetic information, and empower them to make educated, informed decisions about genetic testing, cancer screening, and cancer preventionâ&#x20AC;? [11]. In addition to educating and empowering patients to make decisions, genetic counsellors also assess psychosocial issues that may arise for patients or their families while experiencing this process. Counselling is targeted at individuals who are interested in having an estimate of personal or familial risk [12]. Those at greatest risk for developing malignancy may benefit from accessing current screening tests, whereas those at lower risk will be reassured that their risk for CRC is lower than they thought. Genetic counselling is appropriate when the personal or family history is suggestive of a heritable predisposition for malignancy [13]. In general, the process of counselling is composed of discrete elements: risk assessment, informative counselling, supportive counselling, and follow-up [11]. Tailoring sessions to individual needs and presenting information in clear and sensitive language is essential for effective counselling.
Risk assessment The presence or absence of a family history of benign and malignant neoplasms helps to determine whether a person has an increased probability of having an inherited susceptibility for malignancy. Other familial traits can be characteristic signs of a genetic syndrome and suggestive that an increased risk for cancer may be present in the family (i.e. desmoid tumours). Pedigree analysis is multigenerational and includes details from both sides of the proband's family [14]. Attempts should be made to document all diagnoses of malignancies. For each case reported, obtaining information on relatives at least one generation up and down is significant. Current age or age at death is ascertained for everyone, because early death may explain the absence of a positive family history. Most family histories are not suggestive of an autosomal dominant or recessive mode of inheritance for CRC, and genetic testing is not indicated. Nevertheless, such individuals may benefit from the other aspects of genetic counselling, such as the reassurance that their risk is lower than they thought. Although pedigree analysis is vital in providing an assessment of individual risk, past medical history or the role of environmental influences (such as diet) should not be ignored, even if an inherited susceptibility exists. In addition to estimating the patient's actual risk, counselling must also assess the person's perceived risk for developing a malignancy. Aspects of family
Inherited syndromes of colorectal cancer
313
structure and interpersonal relations are often disclosed that assist the counsellor in identifying individual patient needs, thereby allowing better case management. Informative and supportive counselling Providing risk assessment is only one part of the overall educational process in genetic counselling [15]. Explaining what genes are, how they are passed to subsequent generations, and the natural history of a neoplastic disease is an integral part of genetic counselling. Other educational aims include the recommendation or provision of screening modalities for early detection and preventive options. Because most genes predisposing to CRC are organ specific, not all organs are equally susceptible to malignancy. A focus on organs that are more vulnerable to malignancy optimises surveillance. Medical interventions, such as prophylactic surgery or chemoprevention, and the possible limitations of these treatments are issues for members of families with hereditary cancer. Obtaining knowledge may empower the individual and family by dispelling myths about malignancy and by providing guidelines that help in making decisions [15]. Knowledge also establishes the foundation of informed consent for genetic testing. Genetic testing is discussed when the patient's family history of CRC follows an autosomal or recessive dominant mode of inheritance. The decision to undergo genetic testing is a personal one based on informed consent. The elements of informed consent include information on the gene being tested and the implications, limitations, and impact of results for the person being tested and for family members [16]. We emphasise that genetic testing is voluntary, and patients need to be aware of alternatives. Education and discussion of the benefits and risks of genetic testing through each patient's viewpoint promotes autonomy and lays the foundation for informed consent. Genetic testing may be highly specific but quite insensitive, so the interpretation of results may not be straightforward when a mutation has not been identified in the family. In this instance, patients need to understand the difference between a true negative result and a non-informative negative result. An ambiguous test result may be more distressing than a positive test result. Moreover, genetic testing often yields results that are probabilistic; that is, a mutation carrier does not know for sure if and when a malignancy will develop. The psychological burden of knowing that one is at high risk for developing a neoplasm may outweigh the possible benefits from intervention [17]. Individuals need to know the implications and possible impacts of the results of genetic tests before testing. Surveillance or surgery can be targeted
314
Ignacio Blanco et al.
to specific sites that are more prone to neoplasia, depending on the identity of the gene involved. Unlike other medical tests, genetic testing can reveal information about ourselves and about relatives with whom we share a common genetic legacy. Genetic testing should always be voluntary, and a person should be well informed. Patient autonomy is vital in testing for hereditary susceptibility: the patient has the right not to know [16]. Despite the possible benefits of testing for susceptibility to malignancy, the uncertainty and limitations of current efforts to lower mortality or morbidity contribute to the psychosocial and ethical complexities. As with other medical information, genetic information must remain confidential. Follow-up The management of individuals who may be at increased risk for an inherited CRC is complex. Knowledge of genetic information can contribute to clinical management at several stages during patient care. Recommendations for surveillance and treatment in the most common hereditary CRC syndromes are described in the next part of this review.
2. Lynch syndrome Definition Hereditary Non-Polyposis CRC (HNPCC) was described as an inherited form of CRC that was initially defined to separate adult inherited CRCs from FAP for research purposes in an attempt to uncover the underlying genetic causes [7]. A workshop in Amsterdam in 1989 agreed upon the name HNPCC, because at that time the syndrome was unknown to most doctors. Clinical criteria, known as the Amsterdam criteria (AMSI) [18], were developed to identify the kindreds in order to provide uniform familial material required for international collaborative studies. These criteria enriched the familial clustering of hereditary CRCs by selecting early-onset cases, vertical transmission, and high number of cases per family. Following the identification of the MMR genes MLH1 and MSH2, mutations in these genes were clearly associated with endometrial and other cancers [19]. The AMSI criteria were revised to include these cancers, (AMSII criteria) (Table 1), aiming at identifying HNPCC [19]. At an international meeting in Bethesda in 2004, most participants considered the term HNPCC to be inappropriate, since the syndrome is also associated with many other tumours, and agreed to use the term Lynch syndrome (LS) to cover defects in MMR genes [7]. Families that meet AMSI criteria â&#x20AC;&#x201C; considering
Inherited syndromes of colorectal cancer
315
Table 1. Amsterdam criteria II and revised Bethesda guidelines.
CRC cases only in the family - but show no evidence of MMR deficiency (up to 60% of all families identified) are referred to as having familial colorectal cancer. Lindor et al. [20] reported that these families had a lower annual incidence rate of CRC and a later age of onset. The underlying defect: MMR deficiency The mismatch-repair system scans DNA in search of mismatches and insertion/deletion loops ranging from one to ten or more bases [21]. MSH2 forms heterodimers with MSH3 and MSH6 that are able to recognise these errors. hMutSβ (hMSH2-hMSH3) and hMutSι (hMSH2-hMSH6) complexes may show distinct specificity in error recognition. Once errors are identified, a second complex formed by hPMS2 and hMLH1 completes DNA excision
316
Ignacio Blanco et al.
after strand discrimination. Finally, polymerases, endonucleases, and other proteins contribute to complete the process of repair [3]. Microsatellite instability (MSI), characterised by small deletions or insertions within short tandem repeats in tumour DNA, is the molecular symptom of a defective mismatch-repair machinery and characterises a distinctive pathway of tumour progression known as the Microsatelite Mutator Phenotype (MMP) [22]. MSI is found in >90% of colon cancers associated with Lynch syndrome, whereas MSI is found in only 8-15% of the sporadic cases [3]. A similar proportion of other types of human tumours (endometrial and gastric cancers) exhibit ubiquitous MSI. In the majority of sporadic cases, the phenotype is due to epigenetic inactivation of MLH1 that may occur via epigenetic changes such as DNA methylation of the MLH1 transcriptional regulatory sequences. The loss of protein expression of the causative gene can be shown with immunohistochemical (IHC) analysis using antibodies against the four MMR proteins [23]. Germline mutations - mainly truncating point mutations but also gross deletions - in MSH2 and MLH1 are the hallmark of Lynch syndrome [3]. Inactivating mutations in other MMR genes (PMS1, PMS2, and MSH6/GTBP) have been seen in a small fraction of those with LS [7]. Altogether, germline mutations in the known mismatch-repair genes have been detected in only 1-2% of CRC patients. Characteristics of Lynch syndrome Carriers of an MMR mutation have a high lifetime risk of developing CRC (28-75% in men, 24-52% in women), endometrial cancer (27-71%), and other associated cancers, mainly ovarian, gastric, and urinary tract cancer (1.13%) [7,24]. The cancers observed in families with Lynch syndrome are diagnosed at an unusually early age, may be multiple, are predominantly located in the right colon, and show specific pathological features including a Crohnâ&#x20AC;&#x2122;s-like reaction, a mucinous phenotype, and significant infiltration of lymphocytes [25]. Identification Identification of family members carrying a defect in an MMR gene is important so that colonoscopic surveillance can be restricted to these individuals. Those without a gene defect can be reassured and spared intensified surveillance [5]. Mutation analysis is rather expensive because four genes may need to be analysed. Moreover, comprehensive screening of these genes is required because their mutational spectra are wide [10]. Currently, the Amsterdam II/Revised Bethesda criteria [26] (Table 2) are appropriate tools to
Inherited syndromes of colorectal cancer
317
select patients with CRC for molecular genetic and/or immunohistochemical analysis of the tumour, and those with evidence of MSI or loss of MMR expression, as assessed by IHC, are offered mutation analysis. The sensitivity of MSI analysis is slightly higher than that of IHC analysis. In families with a high probability of having a mutation (Amsterdam II criteria), IHC is the best first step because it may direct mutation analysis. Table 2. Extracolonic cancer risks in FAP. Malignancy Desmoid Duodenum Thyroid Brain Ampullary Pancreas Hepatoblastoma Gastric
Relative Risk
Absolute Lifetime Risk (%)
852.0 330.8 7.6 7.0 123.7 4.5 847.0 â&#x20AC;&#x201C;
15.0 3.0-5.0 2.0 2.0 1.7 1.7 1.6 0.6
Figure 1. Comprehensive molecular analysis of defects in mismatch-repair genes in suspected cases of Lynch syndrome.
318
Ignacio Blanco et al.
In other families, either MSI or IHC analysis can be the first step. The results of pedigree and MSI/IHC analyses should be discussed in a multidisciplinary setting (pathologist, clinical/molecular geneticist, gastroenterologist, surgeon, etc.) [23] (Fig. 1). Even with the use of these guidelines, however, a significant proportion of mutation carriers may be missed. Since most missed mutation carriers are diagnosed with CRC between the ages of 50 and 60, increasing the age at diagnosis below which MSI-analysis is recommended may be appropriate. Due to increasing evidence that MSI/IHC is an important prognostic factor and may predict the response to chemotherapy, these tests might in future be performed on a much larger scale, if not in all CRC cases, which could lead to the population-based molecular screening of Lynch syndrome [27]. Surveillance and management Periodic examination by colonoscopy leads to detection of CRC at an earlier stage, a 63% reduction of the risk of CRC, and a significant reduction of the mortality associated with CRC [21]. A 3-yr interval between examinations has proven to be (at least partly) effective. In view of the observation of advanced CRC detected 2-3 yr after colonoscopy, the optimal interval probably lies between 1 and 2 yrs. Surveillance should start between age 20-25 yrs [25]. Decisions on the upper age limit of surveillance should be made on an individual basis. In families with clustering of CRC but without evidence of MMR-deficiency (non-Lynch syndrome families), a less intensive surveillance protocol is recommended, i.e., colonoscopy at 3-5yr intervals, starting 5-10 yrs before the first diagnosis of CRC, or >45 yrs. A substantial proportion (estimated at 5-10% per 10 yrs of follow-up) of patients, however, develop interval cancers under surveillance [20]. The value of surveillance for endometrial cancer is unknown, and scarce data are available on the effectiveness of surveillance for endometrial cancer. Surveillance by gynaecological examination, transvaginal ultrasound, and aspiration biopsy starting from age 30-35 yrs may lead to detection of premalignant lesions and early cancers [28]. Prophylactic hysterectomy and salpingo-oophorectomy may be an option for women with Lynch syndrome since it substantially reduces site-specific cancers [7]. For patients with Lynch syndrome who present with CRC, the surgical choice lies between partial resection and more extensive surgery such as subtotal colectomy and ileorectal anastomosis [25]. In view of the increased risk of developing a second tumour and the evidence for improved life expectancy after extensive surgery, the best option appears to be a subtotal colectomy, especially for young patients. However, because such an extensive
Inherited syndromes of colorectal cancer
319
surgical procedure might have a significant impact on the quality of life, a randomised controlled trial should be performed that includes assessment of the quality of life and functional outcome after the two procedures. We must emphasise that evidence supporting the recommendations summarised so far is scarce, and many of these recommendations are based on expert opinion [25].
3. Adenomatous polyposis syndromes Numerous polyposis syndromes, both hereditary and non-hereditary, involve the gastrointestinal (GI) tract. The key points to correctly classify a patient diagnosed with gastrointestinal polyposis are: 他 他 他 他
number of polyps type of polyps familial history of colorectal polyps or cancer extra-intestinal manifestations
Correct classification will help us to proceed with both a correct surveillance programme and a useful genetic counsel to the entire family. In this chapter, we will review the different syndromes characterised by the presence of multiple (more than 5) adenomas affecting the colon and rectum.
a. Familial adenomatous polyposis Definition FAP is an autosomal dominantly inherited syndrome that arises from germline mutation of APC. FAP is characterised clinically by the occurrence of hundreds to thousands of adenomas throughout the colorectum at an early age [8]. This disorder, estimated to affect one in 10 000 individuals and is nearly 100% penetrant, occurs worldwide and affects men and women equally. Familial adenomatous polyposis is an autosomal dominantly inherited syndrome. However, 15-20% of cases are "de novo" without clinical or genetic evidence of FAP in the parents. Recent studies indicated the presence of mosaicism in approximately 15% of such cases [29,30]. Characteristics of APC-associated FAP The hallmark of FAP is the development of hundreds of adenomatous polyps in the colon and rectum, usually in adolescence, with an almost
320
Ignacio Blanco et al.
inevitable progression to CRC by the age of 35-40 yrs, significantly younger than those with sporadic cancers [5]. Other features have been observed in FAP (Table 2). Upper gastrointestinal tract polyps and cancer Upper gastrointestinal polyps (gastric and duodenal adenomas) are present in nearly 90% of FAP patients by the age of 70 yrs. Roughly twothirds of duodenal adenomas occur in the papilla or periampullary region. Advanced duodenal adenomas confer an increased risk of small-bowel cancer, which is the third leading cause of death in FAP patients (8.2%), after metastatic CRC (58.2%) and desmoid tumours (10.9%). The severity of the duodenal disease is determined by the Spigelman classification (Table 3) [8]. FAP patients are also at an increased risk of fundic gland polyps (FGPs) in the stomach. Table 3. Spigelman classification for duodenal polyposis in FAP patients. Grade (points)* Criteria
1
2
3
Polyp number
1-4
5-20
>20
Polyp size (mm) Histology Dysplasia
1-4 Tubular Mild
5-10 Tubulovillous Moderate
>10 Villous Severe
*Stage 0 (0 points); Stage I (1-4 points); Stage II (5-6 points); Stage III (7-8 points); Stage IV (9-12 points).
Congenital hypertrophy of the retinal pigment epithelium Congenital hypertrophy of the retinal pigment epithelium (CHRPE) refers to the presence of characteristic pigmented fundus lesions that are thought to occur in roughly 70-80% of patients with FAP. Ophthalmological examination for CHRPE could be an early diagnostic test for family members at risk. Desmoid tumours Desmoid tumours are rare, locally invasive fibromatoses that are a major cause of morbidity and the second leading cause of death in FAP patients. The
Inherited syndromes of colorectal cancer
321
overall prevalence of desmoid disease in FAP is 15%. The majority of tumours occur within the abdomen (50%), usually involving mesentery of the small bowel, or in the abdominal wall (48%). Traditionally, desmoids have been linked to trauma, particularly abdominal surgery such as prophylactic colectomy. Sex (female), family history, presence of osteomas, and germline mutations after codon 1399 have also been identified as independent risk factors for desmoid occurrence. Thyroid cancer Another malignancy associated with FAP is thyroid cancer, with an estimated incidence of 1-2% and an overwhelming predominance in females (17:1). The most common histological type of thyroid cancer in these patients is papillary (>75%), with an unusual cribriform pattern. Most tumours are multicentric and unilateral. Hepatoblastomas Hepatoblastomas are rapidly progressive embryonal liver tumours, usually affecting children under the age of 2.5 yrs, with a male:female ratio of 2.3:1. The incidence of hepatoblastoma among children of FAP patients is 1 in 235, compared to 1 in 100 000 in the general population. Attenuated FAP (AFAP) Attenuated familial adenomatous polyposis (AFAP) is a phenotypically distinct variant of FAP characterised by the presence of fewer than 100 adenomas, a more proximal colonic location of polyps, and delayed age of CRC onset (15 yrs later than patients with classic FAP). The clinical diagnosis of AFAP is more difficult. Recently, diagnostic criteria for AFAP have been proposed [31]. According to these criteria, a diagnosis of AFAP requires (a) at least two family members over 30 yrs of age with 10-99 adenomas, or (b) one family member over 30 yrs of age with 10-99 adenomas and a first-degree relative with CRC with few adenomas, and applying to both criteria, no family members with more than 100 adenomas before the age of 30 yrs. The cumulative risk of CRC by the age of 80 yrs is estimated to be 69%, and 75% of tumours occur in the proximal colon. Extra-intestinal manifestations have been observed but are less frequent than in the classical form of FAP.
322
Ignacio Blanco et al.
The underlying defect of FAP and AFAP: The APC gene In more than 70% of patients with typical FAP, a mutation can be identified in APC [32]. The occurrence of mutations in APC is much lower in patients with AFAP (approximately 25%). APC acts as a tumour suppressor, both in FAP and in about 80% of sporadic colorectal tumours. APC encodes a large protein, the most common isoform of which comprises 2843 amino acids. The protein functions primarily as part of a complex with axin and glycogen synthase kinase 3-beta that phosphorylates the Wnt effector beta-catenin, thereby targeting it for proteasomal degradation. The activity of the complex appears to be largely controlled by GSK3-beta phosphorylation; the complex is consequently inactive when Wnt signalling is active [33]. Almost all disease-associated germline mutations truncate the protein, although exonic and whole-gene deletions can occur. About 15-20% of mutations arise de novo. Most germline APC mutations occur between codons 168 and 1580 with hotspots at codons 1061 and 1309, perhaps because these two codons contain short repeat sequences that are prone to spontaneous slippage. Mutations close to codon 1309 are associated with florid polyposis (typically several thousand adenomas) and early-onset CRC, whereas most other mutations typically produce between 100 and 1,000 adenomas. Other genotype-phenotype associations have been described, including more-severe upper-gastrointestinal polyposis and desmoid disease in carriers of germline APC mutations after codon 1400. A particularly interesting association is found in patients with AFAP. These patients tend to have germline APC mutations in one of three regions of the gene: before codon 168, in exon 9, or after codon 1580 [31]. Genetic counselling and mutation analysis should be offered to all patients with FAP and AFAP. If a pathogenic mutation has been identified in the index patient, mutation analysis should be offered to the first-degree unaffected relatives [26,34].
b. MUTYH-associated polyposis (MAP) More recently, an autosomal recessive type of oligopolyposis has also been recognised involving the human MutY homologue (MYH, or more accurately MUTYH) gene, referred to as MYH-associated polyposis (MAP) [9]. In 2002, Al-Tassan et al. [35] demonstrated a role for base-excisionrepair (BER) in hereditary CRC. They identified bi-allelic germline mutations in the base-excision-repair gene MUTYH in a British family with
Inherited syndromes of colorectal cancer
323
three affected members and recessive inheritance of multiple colorectal adenomas and carcinoma. Further studies found bi-allelic MUTYH mutations in 26-29% of patients with 10-100 polyps, and 7-29% of patients with 100-1,000 polyps [9]. No bi-allelic mutations were found in patients with fewer than 10 adenomas, but bi-allelic mutations have been reported in some CRC-only patients. Based on these findings, patients with more than 10 adenomas should be referred for genetic counselling, and mutation analysis of the MUTYH gene should be considered. MAP patients have a colonic phenotype similar to attenuated or mild FAP [36]. Duodenal polyposis may also occur in MAP. To date, intestinal malignancies and FAP-associated extra-intestinal lesions such as desmoid tumours, osteomas, and CHRPE have been reported only sporadically in MAP patients. Family members with mono-allelic mutations in MUTYH are probably not at increased risk of CRC and therefore do not need colonoscopic surveillance. The underlying defect in MAP: The MUTYH gene MUTYH, located on the short arm of chromosome 1, is a base-excisionrepair gene preventing mutations from products of oxidative damage, particularly the oxidised guanine lesion 8-oxodG [35]. Oxidative damage can cause the mutant base 8-oxoguanine to be erroneously incorporated into DNA in place of guanine. MUTYH removes adenine residues that have been incorporated opposite 8-oxogunaine due to a tendency for these bases to mispair. If MUTYH is deficient thymidine can be incorporated opposite the adenine following replication. Consequently, germline mutations in MUTYH, which lead to absent or severely deficient glycosylase activity, cause an excess of G:C>T:A mutations in genes such as APC and K-ras [9]. The frequency of carriers of MUTYH mutations is about 1% in western European populations. Although mutations can occur throughout the gene, the most common alleles in the UK and most of northwestern Europe are Y165C and G382D, both of which severely diminish protein function. Other ethnic groups tend to have different ‘common’ alleles, such as Y90X in Pakistanis, E466X in Indians, and c.1395-7delGGA in Caucasians from southern Europe [31]. The frequency of these variant alleles in these populations has not been estimated. Good evidence supporting an association between different MUTYH mutations and different clinical phenotypes is currently lacking. MAP tumours develop along a specific genetic pathway characterised by G:C>T:A somatic hypermutation. Bi-allelic mutations in APC are the norm, with co-selection of the ‘two hits’ and a low frequency of loss of heterozygosity. Unlike FAP adenomas, K-ras mutations are frequent in MAP adenomas. Unusually, all K-ras mutations in MAP are identical (GGT>TGT,
324
Ignacio Blanco et al.
G12C), presumably because this particular guanine is susceptible to replacement by 8-oxoguanine and/or to subsequent misincorporation of adenine. MAP CRCs tend to be near-diploid and MSI-negative. Identification Testing patients with adenomatous polyposis, including those with fewer than 100 adenomas, is more problematic. One unresolved problem is where to draw the lower limit on numbers of adenomas. This problem has no simple answer, given that some AFAP patients, for example, have no detectable adenomas when in their forties. Moreover, patients who have had screening by colonoscopy - for example, those with a family history of bowel tumours will tend to have small lesions detected that might never otherwise have come to clinical attention. Of individuals who present de novo, about 50% with more than 10 adenomas have detectable germline APC or MUTYH mutations. For those being screened as a result of their family history, the presence of any family member with similar numbers of adenomas suggests that a search for a mutation in MUTYH or APC is more likely to be successful than a search in some other gene, such as one of the mismatch-repair genes
Figure 2. Algorithm for clinical and molecular evaluation of patients with polyposis.
Inherited syndromes of colorectal cancer
325
(although HNPCC patients occasionally have several adenomas). The genetic basis of disease, if any, of the individuals with more than 10 adenomas and no detectable mutation in MUTYH or APC (or mismatch-repair genes) currently remains unknown [31,32]. We suggest that the algorithm below (Figure 2) could be used to direct mutation screening and diagnosis for those with more than 10 adenomas. We emphasise that, wherever possible, diagnoses of all syndromes of polyposis should be molecular rather than clinical. Diagnosis The diagnosis of FAP can be made clinically by the identification of hundreds to thousands of colorectal adenomatous polyps during colonoscopic examination. Individuals with attenuated FAP or MAP exhibit oligopolyposis (fewer than 100 colorectal adenomas). Histology is the key to differentiating FAP from the other known syndromes of polyposis, such as lymphoid hyperplasia or hyperplastic polyposis, which may mimic FAP endoscopically. Family histories and the presence/absence of vertical transmission can help to differentiate between FAP/AFAP and MAP [8]. Management and screening First-degree relatives of patients with FAP should undergo screening for FAP between 10 and 12 yrs of age. The screening test of choice is genetic testing for mutation in APC, and genetic counselling is an essential part of genetic testing. Counselling should include patient education, screening and management recommendations, discussion of the possible consequences of genetic testing, and written informed consent for APC testing obtained from the patient and/or parents. Once the disease-causing mutation is identified in an individual affected with FAP, other family members can be tested, and endoscopic surveillance can be directed only at those who test positive for the mutation. If the pedigree mutation is not found, or if informative genetic testing cannot be done, all firstdegree family members should undergo endoscopic screening. Current screening recommendations include yearly sigmoidoscopy starting at 12 yrs of age, with a reduction in screening frequency with each subsequent decade up to age 50 yrs (every 2 yrs after age 25 yrs; every 3 yrs after age 35 yrs), after which screening should conform to the guidelines for people of average risk [8]. For upper-GI screening in patients affected with FAP, upper endoscopy (with procurement of biopsy specimens and cytological specimens by brushing) of the stomach, duodenum, and peripapillary region with forward-
326
Ignacio Blanco et al.
and/or side-viewing endoscopes is recommended, starting at 25 yrs of age in asymptomatic patients with FAP. The frequency of upper-GI endoscopies will be determined by the degree of duodenal disease measured by the Spigelman Classification [37]. For individuals at risk for attenuated FAP or MAP, colonoscopic screening should start no later than the age of 20, and then every 2-3 yrs. Upper GI-endoscopy is advised starting at an age between 25 and 30 yrs. The recommended intervals between screenings depend on the severity of disease [8]. Treatment Patients with FAP require surgical therapy. To eliminate the risk of CRC, colectomy is the recommended treatment. Prophylactic surgery should be performed shortly after the diagnosis of FAP is established clinically by endoscopy. Several surgical options are available, including colectomy with ileorectal anastomosis, total proctocolectomy with ileostomy, and total proctocolectomy with mucosal proctectomy and ileoanal pull-through (with pouch formation). Because of the risk of cancer in the retained rectal segment, surgical approaches that eliminate the rectum are advocated [8]. Patients with subtotal colectomy require routine endoscopic surveillance of the remaining rectum about every 6 months for recurrent adenomas and/or carcinomas. Factors associated with increased risk of subsequent rectal cancer include high number of rectal polyps, long rectal segment (>10-15 cm), inadequate endoscopic surveillance, and colon cancer at the time of colectomy. Use of nonselective and selective cyclooxygenase (COX-2) inhibitors (sulindac and celecoxib, respectively) to prevent or induce regression of polyps in the retained rectum of patients with FAP has been shown to be effective in short-term trials and in a long-term study of sulindac. Currently, celecoxib is approved both by the Food and Drug Administration in the United States and the European Medicines Agency for this indication [26]. Administration of sulindac for primary chemoprevention of FAP failed to prevent the development of adenomatous polyposis in unaffected carriers with genetic mutations. Consequently, treatment with sulindac or celecoxib is not recommended for presymptomatic carriers, nor are these drugs advocated for the primary treatment of polyposis in patients with an intact colon. Several endoscopic techniques are available for managing duodenal polyposis, including polypectomy, thermal ablation, and photodynamic therapy. Because of the high incidence of peripapillary carcinoma among patients with FAP, prophylactic duodenal resection by an appropriately skilled surgeon should be seriously considered in those with stage IV duodenal polyposis.
Inherited syndromes of colorectal cancer
327
The management of desmoid tumours is complex. NSAIDs, specifically sulindac, have been associated with unexpected resolution of desmoid tumours in patients with FAP who were treated with those agents for other reasons. Anti-estrogenic therapy has also been used for the management of desmoid tumours, with response rates of approximately 50%. The role of chemotherapy remains controversial, although several chemotherapeutic agents have been used in the management of desmoid tumours. The role of radiotherapy in the treatment of intra-abdominal desmoid tumours remains controversial. Currently, the use of radiation therapy for treatment of desmoid tumours is not advocated as a primary treatment. Radiotherapy, however, may have a role as an adjuvant therapy in patients with desmoid tumours amenable to surgical resection and with positive margins. Complete radical resection of intra-abdominal mesenteric desmoids is often impossible. The frequency of local recurrence after surgical resection may be as high as 70%, and the removal of a considerable length of small bowel is often required. Because operative treatment of intra-abdominal desmoid tumours is associated with serious morbidity and mortality, and may stimulate tumour growth, surgery is often limited for specific complications of desmoid tumours such as obstruction of the small bowel or ischemia. Most patients with AFAP or bi-allelic mutations in MUTYH have an attenuated phenotype. The removal of these polyps endoscopically is possible in many patients because of the small number of adenomas. If surgery is required, an ileorectal anastomosis will be sufficient in most cases to eliminate the risk of cancer. However, if rectal polyposis is severe, an ileal pouch-anal anastomosis could be advised. Morbidity and mortality from hereditary forms of CRC should be reduced once a patientâ&#x20AC;&#x2122;s familial or hereditary risk is established and a highly targeted programme of cancer surveillance and management is undertaken. Prevention will be aided by the identification of the causative germ-line mutation in a patientâ&#x20AC;&#x2122;s family, thus confirming the risk. In addition to diagnostic methods, physicians must also be familiar with the available screening methods and with the options for surgical prophylaxis, particularly prophylactic colectomy in patients with FAP, and prophylactic colectomy and prophylactic bilateral salpingo-oophorectomy (the latter when childbearing is complete) in patients with HNPCC.
Conclusion About a 6% of individuals with CRC have an identifiable inherited genetic predisposition to this malignancy. Genetic testing and rational clinical management recommendations currently exist for the management of
328
Ignacio Blanco et al.
individuals with a variety of CRC syndromes, including hereditary nonpolyposis colorectal cancer (HNPCC, also known as Lynch syndrome), familial adenomatous polyposis (FAP), MUTYH-associated polyposis (MAP), and the hamartomatous polyposis syndromes (Peutz-Jeghers, juvenile polyposis, and Cowden disease). These syndromes also predispose carriers to a variety of extracolonic cancers. The elucidation of the genetic basis of several CRC predisposition syndromes over the past decades has allowed for better management of individuals who are either affected with, or at-risk for inherited CRC syndromes. Appropriate multidisciplinary management of these individuals includes genetic counseling, genetic testing, clinical screening, and treatment recommendations.
Acknowledgements The authors would like to particularly acknowledge the Asociación Española contra el Cáncer (AECC), which recognizes our group as a one of the Grupos Estables de Investigación Oncológica 2010. Contract grant sponsor: Asociación Española contra el Cáncer - junta de Barcelona, Spanish Health Research Fund; Carlos III Health Institute; Catalan Health Institute and Autonomous Government of Catalonia. Contract grant numbers: ISCIIIRETIC: RD06/0020/1051, RD06/0020/1050, RD06/0020/0028, 2009SGR290, PI10/0748.
References 1. 2. 3. 4. 5. 6. 7. 8.
Center, M.M., Jemal, A., Smith, R.A., and Ward, E. 2009, CA. Cancer J. Clin., 59, 366. Cho, K.R. and Vogelstein, B. 1992, Cancer, 70, 1727. de la Chapelle, A. 2004, Nat. Rev. Cancer, 4, 769. Markowitz, S.D. and Bertagnolli, M.M. 2009, N. Engl. J. Med., 361, 2449. Foulkes, W.D. 2008, N. Engl. J. Med., 359, 2143. Cunningham, D., Atkin, W., Lenz, H.J., Lynch, H.T., Minsky, B., Nordlinger, B., and Starling, N. 2010, Lancet, 375, 1030. Lynch, H.T., Lynch, P.M., Lanspa, S.J., Snyder, C.L., Lynch, J.F., and Boland, C.R. 2009, Clin. Genet., 76, 1. Vasen, H.F., Moslein, G., Alonso, A., Aretz, S., Bernstein, I., Bertario, L., Blanco, I., Bulow, S., Burn, J., Capella, G., Colas, C., Engel, C., Frayling, I., Friedl, W., Hes, F.J., Hodgson, S., Jarvinen, H., Mecklin, J.P., Moller, P., Myrhoi, T., Nagengast, F.M., Parc, Y., Phillips, R., Clark, S.K., de Leon, M.P., RenkonenSinisalo, L., Sampson, J.R., Stormorken, A., Tejpar, S., Thomas, H.J., and Wijnen, J. 2008, Gut, 57, 704.
Inherited syndromes of colorectal cancer
9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
21. 22. 23.
24. 25.
26. 27.
329
Sampson, J.R., Dolwani, S., Jones, S., Eccles, D., Ellis, A., Evans, D.G., Frayling, I., Jordan, S., Maher, E.R., Mak, T., Maynard, J., Pigatto, F., Shaw, J., and Cheadle, J.P. 2003, Lancet, 362, 39. Pineda, M., Gonzalez, S., Lazaro, C., Blanco, I., and Capella, G. 2010, Mutat. Res., 693, 19. Cabrera, E., Blanco, I., Yague, C., and Zabalegui, A. 2010, Patient Educ. Couns., 78, 382. Bancroft, E.K. 2010, J. Adv. Nurs., 66, 710. Hampel, H., Sweet, K., Westman, J.A., Offit, K., and Eng, C. 2004, J. Med. Genet., 41, 81. Hemminki, K. and Eng, C. 2004, J. Med. Genet., 41, 801. Etchegary, H. and Perrier, C. 2007, J. Genet. Couns., 16, 419. Offit, K. and Thom, P. 2010, Cancer Treat. Res., 155, 1. Vos, J., Gomez-Garcia, E., Oosterwijk, J.C., Menko, F.H., Stoel, R.D., van Asperen, C.J., Jansen, A.M., Stiggelbout, A.M., and Tibben, A. 2010, Psychooncology. Vasen, H.F., Mecklin, J.P., Khan, P.M., and Lynch, H.T. 1991, Dis. Colon Rectum, 34, 424. Park, J.G., Vasen, H.F., Park, K.J., Peltomaki, P., Ponz de Leon, M., RodriguezBigas, M.A., Lubinski, J., Beck, N.E., Bisgaard, M.L., Miyaki, M., Wijnen, J.T., Baba, S., and Lynch, H.T. 1999, Dis. Colon Rectum, 42, 710. Lindor, N.M., Rabe, K., Petersen, G.M., Haile, R., Casey, G., Baron, J., Gallinger, S., Bapat, B., Aronson, M., Hopper, J., Jass, J., LeMarchand, L., Grove, J., Potter, J., Newcomb, P., Terdiman, J.P., Conrad, P., Moslein, G., Goldberg, R., Ziogas, A., Anton-Culver, H., de Andrade, M., Siegmund, K., Thibodeau, S.N., Boardman, L.A., and Seminara, D. 2005, JAMA, 293, 1979. Gruber, S.B. 2006, Gastroenterology, 130, 577. Yamamoto, H., Imai, K., and Perucho, M. 2002, J. Gastroenterol., 37, 153. Vasen, H.F., Moslein, G., Alonso, A., Aretz, S., Bernstein, I., Bertario, L., Blanco, I., Bulow, S., Burn, J., Capella, G., Colas, C., Engel, C., Frayling, I., Rahner, N., Hes, F.J., Hodgson, S., Mecklin, J.P., Moller, P., Myrhoj, T., Nagengast, F.M., Parc, Y., Ponz de Leon, M., Renkonen-Sinisalo, L., Sampson, J.R., Stormorken, A., Tejpar, S., Thomas, H.J., Wijnen, J., Lubinski, J., Jarvinen, H., Claes, E., Heinimann, K., Karagiannis, J.A., Lindblom, A., Dove-Edwin, I., and Muller, H. 2010, Fam. Cancer, 9, 109. Pineda, M., Castellsague, E., Musulen, E., Llort, G., Frebourg, T., BaertDesurmont, S., Gonzalez, S., Capella, G., and Blanco, I. 2008, Genes Chromosomes Cancer, 47, 326. Vasen, H.F., Moslein, G., Alonso, A., Bernstein, I., Bertario, L., Blanco, I., Burn, J., Capella, G., Engel, C., Frayling, I., Friedl, W., Hes, F.J., Hodgson, S., Mecklin, J.P., Moller, P., Nagengast, F., Parc, Y., Renkonen-Sinisalo, L., Sampson, J.R., Stormorken, A., and Wijnen, J. 2007, J. Med. Genet., 44, 353. Lynch, P.M. 2008, Fam. Cancer, 7, 65. Dinh, T.A., Rosner, B.I., Atwood, J.C., Boland, C.R., Syngal, S., Vasen, H.F., Gruber, S.B., and Burt, R.W. 2011, Cancer Prev. Res. (Phila.), 4, 9.
330
Ignacio Blanco et al.
28. Grindedal, E.M., Blanco, I., Stormorken, A., Maehle, L., Clark, N., Gonzalez, S., Capella, G., Vasen, H., Burn, J., and Moller, P. 2009, Fam. Cancer, 8, 145. 29. Hes, F.J., Nielsen, M., Bik, E.C., Konvalinka, D., Wijnen, J.T., Bakker, E., Vasen, H.F., Breuning, M.H., and Tops, C.M. 2008, Gut, 57, 71. 30. Castellsague, E., Gonzalez, S., Guino, E., Stevens, K.N., Borras, E., Raymond, V.M., Lazaro, C., Blanco, I., Gruber, S.B., and Capella, G. 2010, Gastroenterology, 139, 439. 31. Nielsen, M., Hes, F.J., Nagengast, F.M., Weiss, M.M., Mathus-Vliegen, E.M., Morreau, H., Breuning, M.H., Wijnen, J.T., Tops, C.M., and Vasen, H.F. 2007, Clin. Genet., 71, 427. 32. Rivera, B., Gonzalez, S., Sanchez-Tome, E., Blanco, I., Mercadillo, F., Leton, R., Benitez, J., Robledo, M., Capella, G., and Urioste, M. 2010, Ann. Oncol. 33. Fearnhead, N.S., Britton, M.P., and Bodmer, W.F. 2001, Hum. Mol. Genet., 10, 721. 34. Nieuwenhuis, M.H., Mathus-Vliegen, L.M., Slors, F.J., Griffioen, G., Nagengast, F.M., Schouten, W.R., Kleibeuker, J.H., and Vasen, H.F. 2007, Clin. Gastroenterol. Hepatol., 5, 374. 35. Al-Tassan, N., Chmiel, N.H., Maynard, J., Fleming, N., Livingston, A.L., Williams, G.T., Hodges, A.K., Davies, D.R., David, S.S., Sampson, J.R., and Cheadle, J.P. 2002, Nat. Genet., 30, 227. 36. Nielsen, M., Morreau, H., Vasen, H.F., and Hes, F.J. 2010, Crit. Rev. Oncol. Hematol. 37. Spigelman, A.D., Williams, C.B., Talbot, I.C., Domizio, P., and Phillips, R.K. 1989, Lancet, 2, 783.
Research Signpost 37/661 (2), Fort P.O. Trivandrum-695 023 Kerala, India
The Challenge of Colorectal Cancer: A Review Book, 2011: 331-351 ISBN: 978-81-308-0460-6 Editor: Esther Uña Cidón
14. Issues in colorectal cancer survivors 1
Esther Uña Cidón1, Jesús Crespo2 and Rosa Bustamante2
Clinical Oncology Department, Clinical University Hospital of Valladolid, Spain 2 Biochemical and Molecular Biology Department, Clinical University Hospital of Valladolid, Spain
Abstract. Colon and rectal cancers (CRC) are very frequent and their incidence doubles with each successive decade of life beyond 50 years. Consequently, the absolute number of CRC patients is expected to grow substantially in the coming years as the population progressively ages. On the other hand, improvements in treatment are expected to increase the number of survivors in the near future. A better understanding of the late effects and health-care needs of long-term CRC survivors will thus be imperative and is the aim of this chapter.
Introduction The incidence of CRC increases with age, mainly after 50 years, and the population is now progressively aging. An increasing number of people are therefore expected to develop this disease in the near future [1,2]. On the other hand, advances in detection (prevention and screening programmes), treatment (adjuvant and neoadjuvant regimens) [3-7], and even the improvement of Correspondence/Reprint request: Dr. Esther Uña Cidón, Clinical Oncology Department, Clinical University Hospital of Valladolid, Spain. E-mail: aunacid@hotmail.com
332
Esther U帽a Cid贸n et al.
surgical techniques such as mesorectal excision are contributing to improved rates of survival. For all these reasons, we can predict that the number of survivors of CRC is very likely to increase in the coming years [4]. Many relevant issues related to the treatments for this disease, or related to or favoured by possible comorbidities in the target population (elderly people) [8], merit our attention due to their potential impact on the quality of life (QOL) of these patients. This chapter will discuss these issues and several strategies for helping to improve the QOL of survivors.
1. Issues related to treatment The treatment of CRC is often multidisciplinary and depends mainly on the location of the primary tumour (rectum or colon) and also on the stage of the disease at diagnosis. In general terms, the treatment modalities for early stages could include surgical resection and adjuvant chemotherapy for colon cancer and additional radiation therapy for rectal cancers. All three modalities might have lasting associated morbidities that could affect long-term function and QOL for survivors of disease after the cancer is treated. We will describe the effects related to each of these treatments.
Issues related to surgical procedure The main curative treatment for CRC is still complete surgical resection [9]. This standard oncologic surgery consists of en bloc bowel resection (colon and/or rectum) with appropriate proximal and distal resection margins, the resection of the associated blood vessels, and more than 12 harvested lymph nodes [10,11]. Most of these surgical procedures are usually performed through an open vertical incision in the abdominal wall, a procedure called laparotomy. This kind of surgery has several risks that could be relevant for people surviving the disease.
a. Laparotomy Several problems associated with laparotomy can occur in the short and long term. Although long-term complications are more relevant for survivors, problems secondary to surgery in the acute period can be very important and may require additional invasive procedures. The most frequent complications during this period include wound infections (3-26%), anastomotic leaks (2-10%), and intra-abdominal infections (2-5%). All of these can be severe and may require invasive procedures [12-14]. Long-term complications,
Colorectal cancer survivors
333
though, are the most relevant. Bowel obstructions, hernias of the abdominal wall, and functional problems are the three most frequent complications. a.1. Bowel obstructions secondary to surgery Early postoperative small-bowel obstruction is a common and serious complication after colectomy. Shin and Hong [15] analysed 504 patients who underwent colectomy for CRC. These patients were monitored, and those assessed to have an obstruction within the first 30 days after surgery and lasting for at least two days presented nausea, vomiting, and abdominal distention. Early postoperative obstruction, the most frequent complication during the early perioperative period, occurred in 8.1% of the patients. The authors showed that pelvic surgeries did not lead to a higher rate of early obstruction compared with colonic surgeries. Local remnant tumours and poor systemic condition were independent risk factors for this complication after colectomies for CRC. Obstruction is particularly relevant for survivors of CRC. This complication has the potential for tumour recurrence, which can reduce the patientsâ&#x20AC;&#x2122; QOL [16]. Jeong et al. [17] studied 2586 patients operated for a primary CRC. During the follow-up, 5% of the patients presented adhesive small-bowel obstructions. The observed incidence rate was 0.0013 per patient-month. Most patients (80%) were successfully treated in less than one week (range 1-22 days) by conservative procedures such as intestinal decompression with gastrointestinal tubes. The remaining patients needed a new surgical procedure due to strangulation or the lack of improvement. These results indicate that initial conservative management is recommended for these patients. a.2. Abdominal-wall hernia secondary to surgery Incisional hernia is a long-term complication of laparotomy. Its exact frequency varies according to different authors but is always near 10% [18,19]. In cases with infected surgical wounds, though, the rate can be as high as 20% [20]. This complication can lead to pain, limitation of patient activity, and possibly reparative surgery or emergency surgery for bowel strangulation. Only 50% of these hernias become evident within six months of the operation. The remainder appear well after recovery from surgery [21,22]. Millikan [21] reported that 100 000 repairs are performed annually in the USA, with a rate of recurrence of 5-10%. Incisional ventral hernia has many predisposing factors that might reduce the overall incidence of these hernias,
334
Esther U帽a Cid贸n et al.
such as factors related to patients conditions, type of closure, and materials used. Some patients present associated systemic diseases (chronic obstructive pulmonary disease, obesity, severe cardiopathies, immunodeficiencies, etc.) that favour or increase the risk of an incisional hernia. Hidalgo et al. [20] evaluated 72 high-risk patients (42 with CRC) selected for surgical intervention through median infra-umbilical laparotomy. During laparotomy, the preperitoneal space was dissected at a point where a lowmolecular-weight polypropylene mesh was to be placed when closing the peritoneum. Meshes were about 7-8 cm wide with a variable length depending on the length of the surgical incision. Follow-up was 3-5 years. No noteworthy complications or operative mortality occurred, and no mesh required removal. In two patients who developed liver metastases and needed a second surgical intervention, the health of the abdominal wall and the absence of hernia were confirmed. None of the 72 patients developed an incisional hernia. These authors concluded that the prophylactic use of a low-molecular-weight polypropylene mesh in abdominal surgery may be useful for the prevention of incisional hernia. Current controversies remain to be answered concerning the methods for repairing hernias (open versus laparoscopic procedure) and the types of fixation (partial- versus full-thickness abdominal muscular/fascial wall) necessary to stabilise the position of the mesh while tissue growth occurs [21].
b. Minilaparotomy Ishida et al. [23] evaluated minilaparotomy, a minimally invasive alternative to laparoscopy, for performing curative resections of colon cancer. They studied the feasibility, safety, and early oncological outcome among 73 patients (first group), in whom a curative resection of colon cancer was performed via minilaparotomy (skin incision < or = 7 cm) using specific instruments (North-bridge retractor system), and 94 patients (second group), in whom a similar procedure was performed without using specific instruments. These two groups did not differ in the incidence of postoperative complications, length of postoperative hospital days, or overall survival, although the second group had a higher frequency of prior abdominal surgery (38.3 vs. 21.9%; P = 0.03) and a shorter median operating time required for a standard lymph-node dissection (120 vs. 135 min; P = 0.03). The authors thus concluded that improved techniques and experience precluded the need of specific instruments for the performance of a curative colectomy and that minilaparotomy is a good alternative to laparoscopy.
Colorectal cancer survivors
335
c. Laparoscopy Several studies, however, have demonstrated the advantages of laparoscopic surgery, another less aggressive surgical procedure [24-27]. Its smaller incisions lead to less pain, shorter length of stay, lower rate of incisional hernia, and probably less adhesion formation. Several randomised trials comparing this procedure with open incisions have demonstrated that patients who underwent laparoscopic resections recovered earlier with less need of blood support and lower morbidity. Some studies have even suggested a lower risk of tumour recurrence and death from any cause or from cancer-related causes [27]. More laparoscopically assisted surgical resections will likely be performed in the future. Minimally invasive surgery has multiple benefits, including a shorter hospital stay and less postoperative opioid use, without compromising long-term [27,28] outcomes or increasing the costs of health care [29]. Laparoscopic resection of the colon was first described in nineties [30]. Although techniques and equipment were at first cumbersome, laparoscopic colectomy for benign and malignant conditions of the colon soon became a reality. Early reports of laparoscopic-assisted colectomy revealed a faster recovery from surgery and fewer surgical complications [31,32]. However, wound-site recurrence, which reached 21% in some studies, raised significant concerns about this technique [33]. A randomised controlled trial comparing the efficacy of laparoscopicassisted colectomy with open colectomy found that patients having the former approach recovered faster, had less blood loss, and had lower morbidity (P < 0.001). Finally, the authors report that the probability of cancer-related survival was higher in the laparoscopic-assisted group (P = 0.02); the Cox model showed that the laparoscopic-assisted approach was independently associated with reduced risk of tumour relapse, death from any cause, and death from a cancer-related cause compared to open colectomy [24]. Another trial of the laparoscopic versus open approach for treatment of colon cancer has also been performed in North America. Regarding QOL issues, Weeks et al. demonstrated that the global QOL was significantly higher at two weeks following the laparoscopic approach compared to open surgery [25]. Additionally, the laparoscopic patients required significantly fewer days of parenteral and oral narcotics. Of note, however, is that no differences in QOL were demonstrated at two months following surgery. Importantly, survival and recurrence rates from this trial were not statistically different, which thus suggests that laparoscopic resections of colon cancer can be performed safely in appropriately performed operations [27].
336
Esther U帽a Cid贸n et al.
Two recent meta-analyses reviewed the current literature on laparoscopic resection in rectal cancer. One meta-analysis showed that this approach was associated with lower morbidity but a longer operative time; wound infection, anastomotic leakage, and mortality were similar in both open and laparoscopic groups [34]. The other analysis revealed a reduction in length of hospital stay and time to first bowel movement and stomal function in patients who underwent laparoscopic surgery. Specifically, in the set of patients requiring abdominoperineal resection, laparoscopic patients required fewer parenteral analgesics and had a reduced rate of postoperative wound infection [35]. Increasingly more laparoscopically assisted surgical resections of colon and rectal cancers will likely be performed in the future; however, with the laparoscopic approach, the performance of appropriately indicated and safe resections of cancers remains a priority.
d. Other issues related to surgical procedures 1. Bowel changes and dysfunction The goal in the treatment of rectal cancer should be recovery from the disease with the best faecal continence and QOL. Certainly, one of the functions of the rectum is the storage of faecal material. Whenever the rectum is resected for cancer, the storage capacity of the replacement, which is the colon (i.e., usually the descending or sigmoid colon), is much reduced. Consequently, more-frequent, clustered, and/or incomplete bowel movements can appear. On the other hand, because of possible nerve damage from surgery or radiation therapy, the functioning of the anal sphincter may be further affected, and the degree of incontinence may worsen. Radiation therapy also has a role in these kinds of symptoms. In fact, Kollmorgen et al. [36] found that adjuvant chemoradiotherapy for rectal cancer had a major long-term detrimental effect on bowel function, as explained later in this chapter. This study did not find a significant correlation between the level of the anastomosis and postoperative stool frequency or incontinence. Previous trials had demonstrated such a correlation, with lower anastomoses being associated with greater stool frequency and more incontinence [37-39]. These contradictory results may be partly due to the selection criteria used in the study, which were not used in the previous investigations. Short-term follow-up, during which the greatest deterioration in bowel function occurs, was not performed, and patients who had sutured colo-anal anastomoses or who had extensive resections, dysfunctioning stomas, or clinically significant anastomotic leaks were excluded. In these groups, very low anastomoses may be more common, and postoperative
Colorectal cancer survivors
337
function may be even worse [37-39]. Various studies have also detected that bowel function is often compromised in the early postoperative period after anterior resection, with frequent bowel movements and faecal incontinence. Bowel function improves over the ensuing one to two years, and this improvement correlates well with increasing capacity of the "neorectum" [40-42]. Franco et al. [43] tried to compare QOL and manometric results in patients treated with neo-adjuvant chemotherapy and rectal low-anterior resection (LAR). They recruited 50 patients with advanced (T3-T4) rectal cancer who underwent neo-adjuvant chemotherapy. Forty-one patients subsequently underwent LAR with or without a colonic pouch. QOL was evaluated by questionnaire after a few months, and the patients later underwent manometric evaluations measuring resting, squeeze, and rectal compliance. The manometric results and the scores from the questionnaire agreed in 75% of the patients. Patients with a hypotonic sphincter had a good QOL if a LAR with pouch had been performed compared to the patients without a pouch. The authors concluded that performing LAR with a colonic pouch after neoadjuvant chemotherapy in patients with a hypotonic sphincter improves QOL. Preoperative anorectal manometry could thus select patients who would benefit from pouch construction to avoid a deterioration in QOL. Chatwin et al. [44] tried to evaluate clinical outcomes and QOL in terms of anal function, among others, after LAR for rectal cancer. They evaluated 43 patients with low rectal cancers. Twenty-seven were not given adjuvant radiotherapy, and 16 received preoperative adjuvant radiotherapy of 1.6 Gy twice daily for 13 days. Twenty-three patients reported normal defecation (53%), nine had incontinence of flatus (21%), five had occasional minor soiling (12%), two had frequent major soiling (5%), and four had total faecal incontinence (9%). The authors concluded that despite the reported faecal dysfunction most patients were satisfied with their QOL. The authors strongly recommended counselling at the time of operation as a means of contributing to personal satisfaction. Most patients prefer a sphincter-sparing procedure, although they need to understand that bowel dysfunction may also occur. Although the location of the anastomosis is very relevant for maintaining function, defecatory problems can occur as a result of surgical trauma or the effects of radiation therapy on the anal sphincter and associated nerves, even in cases in which a sphincter-preserving procedure has been performed. A low anastomosis is more associated with a higher frequency of defecation and faecal leakage and incontinence than a higher anastomosis. Although these are the most frequent symptoms after surgery for rectal cancer, some degree of diarrhea or constipation and excessive flatus have also been described. All these issues could impact the patients´ QOL.
338
Esther Uña Cidón et al.
2. Ostomy issues in survivors Intestinal ostomy is a surgical procedure for treating several benign or malignant diseases, but rectal cancer is the most common reason for a person to undergo this kind of surgery [45]. Many patients undergo ostomy each year, and every effort is made to preserve intestinal integrity [46,47]. This therapeutic approach can be either temporary or permanent and creates many challenges for QOL and maintenance of function. In colostomic and ileostomic surgeries, normal bowel function is interrupted, and waste is passed through the abdominal wall via an opening called a stoma into an appliance that must be emptied periodically [46,47]. Ostomy is an unpleasant surgery. In many cases this procedure leads to intensified distress for patients who will inevitably suffer from physical and emotional problems, isolation, and fear of cancer recurrence or death. The stoma is usually red and swollen and may cause stress as a consequence of skin irritation and rash around the ostomic site (76%), pouch leakage (62%), bad odour (59%), reduction in pleasurable activities (54%), and depression or anxiety (53%) [46]. The procedure impacts negatively on the patients´ QOL. Most patients report significant limitation in physical activity after their ostomy. Social and family relationships are also affected, although these improved over time. A patient’s family must learn how to cope with the new situation, and the patients may complain about their relationships with their intimate partners [45]. A study by Brown and Randle [48] showed that patients with stomas tend to worry about sexual issues, especially in the early postoperative period, leading to a further deterioration of their QOL. Symms et al. [49] reported that almost half of patients who were sexually active before ostomy became inactive after this procedure. The passage of time is the most important factor in adapting successfully to life after ostomy. As the study by Ohman [50] has shown, many problems such as anxiety from faecal leakage, offensive odour, bowel noise, and loss of libido can decrease over time. Counselling and evaluation of sexual health after recovery from surgery are important. After an abdominoperineal resection, the presence of a permanent colostomy has a strong influence on survivors. Stoma-related problems are common. In a series of 203 patients with end sigmoid colostomies, the 13-year actuarial risk of paracolostomic complications was 58%. Paracolostomic hernia was the most common complication (36% at 10 years). Other stomarelated complications may occur in survivors, including stomal prolapse (12%), skin-related problems (e.g., excoriation) (12%), and stenosis of the stomal opening (7%) [51]. Several studies have demonstrated a decreased health-related quality of life (HRQOL) in patients with stomas. US war
Colorectal cancer survivors
339
veterans with stomas were surveyed using the City of Hope Quality of LifeOstomy questionnaire. Qualitative analysis was performed on the basis of the City of Hope QOL for Ostomates format of HRQOL (physical, psychological, social, and spiritual). For colostomic patients, family and spousal relationships were frequently considered important in the psychological category, while sexual relationships were important in the social category. The authors concluded that awareness of patients' social, psychological, and medical status allows surgeons to identify those likely to have problems and to use resources to help these patients [52]. Ostomic patients depend on the integrity of their peristomal skin to maintain a normal lifestyle. Peristomal-skin problems are thought to be common and may interfere with the use of ostomic pouching systems. These problems constitute a special category not commonly seen by dermatologists. Loss of skin integrity may be related to chemical injury, mechanical destruction, infectious conditions, immunological reactions, and diseaserelated conditions. Peristomal irritant dermatitis caused by skin contact with ostomic effluent is by far the most ordinary condition seen [53]. Mechanical trauma, infection, and aggravation of pre-existing skin diseases are also seen. Allergic contact dermatitis, which is often cited as the cause of peristomal skin problems, appears to be a rare condition with an estimated prevalence of only 0.6%. Despite the importance of the integrity of peristomal skin, the topic is poorly described in the literature. Existing publications suggest that although peristomal skin disease can be diagnosed and treated, additional information on both patients and physicians is necessary to optimise patient care [54,55]. In a survey of almost 400 ostomates, 51% had skin problems (e.g., rashes) and 36% had leakage; 80% reported some change in lifestyle. In addition to reports of complications, several studies have compared postoperative psychosocial adjustment and QOL in ostomic and nonostomic patients, including concerns about sexuality, limitations of activity, and bowel function. It is important that clinicians make use of referral of patients to enterostomal therapists who are able to address both the physical and psychosocial sequelae of having a stoma. Their consultation is valuable throughout the course of survivorship from preoperative to short- and longterm periods. Stomal support groups exist in many communities and are another resource [45]. Therapeutic procedures may not only treat disease but also affect patient QOL. Therefore, QOL should be measured in order to assess the impact of disease and therapeutic procedures. To identify clients' problems, the assessment of several dimensions of QOL, including physical, spiritual, economic, and social aspects, is necessary. To address these issues, we
340
Esther U帽a Cid贸n et al.
conducted a qualitative study to explore QOL and its dimensions in ostomic patients referred to the Iranian Ostomy Association. Fourteen patients were interviewed about their QOL dimensions by purposeful sampling. Data were gathered by semistructured interviews and analysed using the content-analysis method. Nine main themes emerged, including physical problems related to colostomy, impact of colostomy on psychological functioning, social and family relationships, travel, nutrition, physical activity, exual function, and religious and economic issues. The findings of the study identified a number of challenges in QOL for patients with ostomy. The results can be used by providers of health care to create a supportive environment that promotes better QOL for their ostomic patients [45]. 3. Sexual dysfunctions A growing body of evidence suggests that the prevalence of sexual dysfunction among men and women is high following treatment for CRC [56-59]. Several descriptive cross-sectional and longitudinal studies of CRC survivors have concluded that overall health-related HRQOL after treatment is good, although many survivors, even those who reported good QOL, also report significant difficulties in sexual functions. Sexual function is one element of QOL that may be significantly damaged following treatment for rectal cancer, but its incidence and contributing risk factors are generally poorly understood. An increasing number of studies have recently reported sexual dysfunction following treatment of rectal cancer. The lack of a standard definition of sexual dysfunction, though, complicates matters. The absence of sexual activity can be used as a surrogate marker for sexual dysfunction, but this is confounded by an individual's desire and opportunity for sexual activity and may not be an accurate reflection of physiological functionality [56-59]. Sexual problems are associated with surgical and radiation therapies that affect the pelvic tissues and/or organs and their innervation. In males, the main sexual problem is erectile dysfunction related to disruption of the parasympathetic nerve and ejaculatory difficulties (inability to ejaculate or retrograde ejaculation) secondary to injury to sympathetic nerves [60]. A conventional resection of rectal cancer in men is associated with postoperative impotence and retrograde ejaculation, or both, in 25-100% of cases [61-63]. The incidence of these problems is higher in older men and in men who underwent an abdominoperineal resection (removal of the rectum and placement of a permanent colostomy) compared to those who underwent an anterior resection [62,63].
Colorectal cancer survivors
341
In females, the most common postoperative sexual problem is dyspareunia, which may include loss of vaginal lubrication and inability to achieve orgasm. Because both symptoms are very difficult to measure, sexual activity is used as a surrogate marker. Various studies have found that among those women who were sexually active before surgery, 47-86% remained sexually active after surgery. Other studies have demonstrated that this number is related to the kind of procedure patients have undergone. After sphincter-preservation surgery, 55-58% of women remained sexually active, but only 10-39% were active after an abdominoperineal resection [60-64]. Havenga et al. [62] surveyed 54 women after total mesorectal excision with autonomic-nerve preservation for rectal cancer (44 LAR, 10 abdominoperineal resection). Ninety-five percent of women remained interested in sex, 86% remained sexually active, 85% continued to experience vaginal lubrication during arousal, and 91% maintained their ability to achieve orgasm postoperatively. In all these cases of men and women, the best possible outcome is achieved by careful sharp dissection with preservation of the pelvic autonomic nerves. Marijnen et al. [65] evaluated the QOL of 990 patients with rectal cancer who did or did not undergo preoperative short-term, high-dose radiation therapy prior to rectal excision. At three months after surgery, patients who underwent preoperative radiotherapy reported significantly reduced daily activity. Both men and women receiving preoperative radiation therapy reported a significant decrease in sexual activity, although HRQOL was not otherwise affected. Relatively nonspecific problems such as changes in level of sexual activity, a lack of sexual enjoyment, and alterations in body image have also been identified in both men and women following treatment for CRC. Lee et al. [66] evaluated the effects of surgery for rectal cancer on postoperative voiding and sexual function over time. Data from 28 male patients who underwent autonomic nerve-preserving surgery for rectal cancer were retrospectively analysed. Operations were performed between October 2005 and July 2007, and all patients were followed-up for more than three years. Preoperatively, all patients underwent urodynamic studies, including uroflowmetry, and were assessed for their International Prostate Symptom Score (IPSS). The evaluation of sexual function consisted of an ErectileFunction domain score from the International Index of Erectile Function (IIEF-EFD) and an Ejaculation domain score from the Male Sexual Health Questionnaire (MSHQ-EjD). Data from uroflowmetry and the questionnaires were examined. At three years postoperation, the prostate volume was similar to the preoperative value (P = 0.727). No statistically significant postoperative changes were found in the average maximum flow rate (15.9 ml/s vs.
342
Esther U帽a Cid贸n et al.
16.2 ml/s, P = 0.637) and the post-void residual urine volume (34.7 ml vs. 36.8 ml, P = 0.809). No statistically significant differences were observed in the IPSS (13.2 vs. 12.2, P = 0.374). However, even though pelvic autonomicnerve preservation was performed, a significant proportion of rectal-cancer patients suffered from sexual dysfunction, and the average IIEF-EFD and MSHQ-EjD scores were lower postoperatively for three years (25.1 vs. 16.1 and 28.3 vs. 14.2 respectively, P < 0.001). The authors concluded that voiding function was not affected after autonomic-nerve-preserving surgery for rectal cancer, but sexual function was significantly aggravated. We recommend that the baseline genitourinary function should be evaluated before treatment for male patients with rectal-cancer, and penile rehabilitation is necessary for their QOL after treatment. In light of the improving prognosis for patients with rectal cancer, the quality of functional outcomes has become increasingly important [67,68]. Large multicentre studies show that urogenital dysfunction remains a common problem after treatment for rectal cancer, despite the good functional results achieved by expert surgeons. More than half of patients experience a deterioration in sexual function, consisting of ejaculatory problems and impotence in men and vaginal dryness and dyspareunia in women. Urinary dysfunction occurs in one-third of patients treated for rectal cancer, with surgical nerve damage being the main cause. Radiotherapy appears to have a role in the development of sexual dysfunction, without affecting urinary function. Pelvic autonomic nerves are especially at risk in cases of low rectal cancer and during abdominoperineal resection. Data concerning nerve damage during laparoscopic surgery for resection of rectal cancer are needed. Structured education of surgeons on pelvic neuroanatomy, and systematic registration of identified nerves, could well be the key to improving functional outcome for these patients. Meanwhile, patients should be informed of all associated risks before their operation, and their functional status should be evaluated before and after surgery [67,68]. 4. Urologic dysfunctions Bladder dysfunction has been reported to occur in 7-68% of patients after resection of low rectal cancer, although its incidence is generally around 30% in most reports [69]. This kind of dysfunction includes various symptoms such as incomplete bladder voiding, mictional urgency, overflow or stress incontinence, loss of bladder sensation, dysuria, and chronic infections of the urinary tract. As with sexual dysfunction, most of these difficulties have a neurogenic origin related to parasympathetic denervation. After an abdominoperineal resection in men, 50% will have a neurogenic bladder, but
Colorectal cancer survivors
343
with recovery within 3 to 6 months after surgery in most cases [69,70]. Also as with sexual problems, the incidence is higher in elderly people, due to the presence of benign prostatic hypertrophy (particularly in this age group), which contributes to urologic dysfunction. Another associated cause of these problems is mechanical: surgery can cause an angulation that contributes to these symptoms. These dysfunctions are more often associated with abdominoperineal resection than with anterior resection [71]. Once again, favorable outcomes may be achieved by careful sharp dissection with preservation of pelvic anastomotic nerves [71].
Issues related to radiation therapy Several randomised studies have evaluated the role of radiation therapy and chemotherapy as adjuvant treatments in rectal cancer. After these trials, two consensus conferences, one American (NIH Consensus Conference 1990) and the other German (German Cancer Society Consensus Conference 1998), concluded that the combination of radiation therapy and chemotherapy was the preferred adjuvant treatment for patients with rectal cancer stages II and III [72-78]. Radiation therapy decreases local recurrence, and chemotherapy based on fluoropyrimidines increases survival at five years by 10-15%. Kollmorgen et al. [36] have assessed the long-term effect of postoperative chemoradiotherapy on bowel function in a retrospective way. They studied patients undergoing anterior resection for Astler-Coller stage B2 or C rectal carcinoma who were given postoperative radiation therapy with chemotherapy. One hundred patients were suitable for inclusion and participated in a telephone questionnaire; 41 patients had postoperative chemoradiotherapy, and 59 did not. Both groups were well matched for basal characteristics. The authors found that the group receiving chemoradiotherapy had more bowel movements per day than the other group (median of 7 vs. median of 2, P < 0.001); the former group had higher frequencies of "clustering" of bowel movements (42% vs. 3%, P < 0.001), nocturnal movements (46% vs. 14%, P < 0.001), occasional or frequent incontinence (39% and 17% vs. 7% and 0%, P < 0.001), pad use (41% vs. 10%, P < 0.001), and an inability to defer defecation for more than 15 minutes (78% vs. 19%, P < 0.001). The group that received chemoradiotherapy also had liquid stools, used antidiarrheal medications, had perianal skin irritation, was unable to differentiate stool from gas, and needed to defecate again within 30 minutes of a movement significantly more often than the group that did not receive chemoradiotherapy. From these findings, the authors concluded that adjuvant postoperative chemoradiotherapy for rectal carcinoma has a major long-term detrimental effect on bowel function.
344
Esther U帽a Cid贸n et al.
Bowel function after radiation therapy, then, is an important functional issue. Several studies have found consistent results showing that bowel function in patients, as measured by frequency, urgency, evacuation, sensation, and/or continence, is impaired after radiation therapy when compared with patients not treated with radiation. The Swedish Rectal Cancer randomised controlled trial [79] has shown that preoperative high-dose radiotherapy improves survival and decreases local recurrence in patients who underwent anterior resection for rectal cancer. The trial also found that the median frequency of bowel movements was higher in the group receiving radiotherapy and surgery compared to the surgery-only group (20 vs. 10 bowel movements per week, P < 0.0001). Also, urgency, voiding difficulties, and incontinence for loose stools were more common in the radiation group (all P < 0.0001). In terms of QOL, 30% of the radiation group stated that their social life was impaired because of bowel dysfunction compared to 10% of the surgery-only group (P < 0.01). Dehni et al. [80] found similar results that the irradiated group had more diarrhea (39% vs. 13%, P = 0.005) and more nocturnal defecation (36% vs. 15%, P = 0.03) compared with the nonirradiated group. Pucciarelli et al. [81] evaluated the long-term complications after preoperative chemoradiotherapy for rectal cancer. They recruited 123 consecutive patients with locally advanced mid-low rectal cancer who underwent preoperative chemoradiotherapy. Complications were defined as late if they occurred more than 6 months after surgery. At a median follow-up of 95 (range, 56-160) months, 50 late complications occurred in 41 patients, 21 of whom required surgery. Several of these complications were clearly related to chemoradiotherapy and were significantly associated with the total dose of radiation delivered (P < 0.05). These authors concluded that late morbidity after this treatment is relevant and related to the radiotherapeutic dose used. In a study of men with symptoms of radiation damage 2-6 years after radiation therapy for prostatic carcinoma, resting anal pressure and anal sphincter length were significantly decreased compared with age-matched controls [82]. Significant reductions in rectal capacity and compliance have also been found [83]. Histologic examination of specimens from patients who underwent a proctectomy for radiation injury commonly revealed hypertrophy of the muscularis mucosae and the muscularis propria, with degeneration of both Meissner's and Auerbach's plexi, and significant damage to the tissues surrounding the anus and rectum [84]. All these problems could contribute to long-term changes in bowel function. Magnetic resonance imaging of the pelvis after radiation therapy has demonstrated alterations in the signal from striated muscle, with thickening of the perirectal fascia and presacral space [85]. All these physiologic and pathologic changes evolve over a prolonged period.
Colorectal cancer survivors
345
In a study of 20 patients who underwent preoperative radiation for rectal carcinoma, the anal-sphincter pressures had not decreased when measured four weeks after the radiation was completed [86]. Furthermore, histologic examination of the excised specimens revealed only minimal changes. A different study reported nine patients who underwent postoperative radiation therapy after anterior resection [87]. Although the selection criteria for this study were not stated, the neorectal capacity and compliance were lower in these patients than in patients who had not undergone radiation therapy. With all these aspects in mind, we must conclude that radiation therapy, although with benefits, can have potentially serious late effects for the survivors of rectal cancer. Acute complications are discussed in detail in another chapter of this book. For all these survivors who underwent radiation therapy, the evaluation of morbidity, pelvic-floor function, and QOL is important. Globally, delayed radiation toxicities have been reported several times and include radiation enteritis (4%), small-bowel obstruction (5%), and rectal stricture (5%), in addition to the bowel, sexual, and urinary dysfunction discussed earlier that may be aggravated by radiation-induced injuries to pelvic nerves. Nathanson et al. [88] evaluated patients with rectal cancer 2-8 years following surgical resection with preoperative radiation, postoperative radiation, or no radiation. The postoperative radiation group had more episodes of clustered bowel movements (P < 0.02) than either the preoperative radiation group or the no-radiation group. The authors attributed the adverse effects of postoperative radiation therapy to irradiation of the neorectum, which is spared when radiation is given preoperatively.
Issues related to chemotherapy Many patients with CRC have been treated with oxaliplatin in combination with fluoropyrimidines in the adjuvant or neoadjuvant context. This combination is generally well tolerated. The secondary effects most often reported have been haematologic, rarely exceeding grade 3, digestive (such as nausea, vomiting, diarrhea, etc.), mucositis, early-onset cold-induced dysesthesias, and a cumulative peripheral sensory neuropathy. These symptoms are acute, leading to a delay of cycles or even a cessation of chemotherapy, except for neuropathy.
a. Neurosensory syndrome The occurrence of peripheral sensory symptoms was noted during the first phase I clinical trial [89]. The symptoms reported can be separated into two
346
Esther U帽a Cid贸n et al.
distinct categories: a cold-induced dysesthesia with a rapid onset of hours to days following treatment, and a late-onset cumulative sensory neuropathy observed after multiple cycles of therapy. Acute onset is common (80-85% of patients); in fact, it is the most common form of acute neuropathic toxicity. It appears within hours of infusion and is usually short-lived without any sequalae. It is rarely debilitating since patients can adapt by avoiding cold stimuli. The temporary loss of awareness of breathing is relatively uncommon but is probably the most alarming symptom. Patients must be warned about these symptoms and their transient nature prior to the administration of chemotherapy. Acute oxaliplatin-associated neurotoxicity can result in other disturbing symptoms such as visual-field cuts, blurred vision, and ptosis [90]. A dose-limiting peripheral sensory neuropathy occurs in 10-15% of patients after a total cumulative dose of 780-850 mg/m2 [91-93]. This neuropathy begins as a persistent paresthesia after multiple cycles of oxaliplatin. A physical examination detects decreases in proprioception, vibration perception, and finepoint discrimination. The patients complain about difficulty in writing and the impairment of such fine manipulations as buttoning a shirt. The duration of these symptoms increases with an increase in the number of cycles. A 25% reduction in dose is thus recommended when these symptoms become persistent between cycles. Cessation of treatment is recommended in cases where worsening symptoms lead to functional impairment. This syndrome is reversible in most cases after discontinuation of the treatment, with a median time of recovery of 15 weeks. These symptoms, however, sometimes persist and reduce the QOL of the survivors, with a perception of increased difficulty with current or daily activities. These patients generally have received a cumulative dose higher than 780 mg/m2 [91-93]. The combination of oxaliplatin with 5-FU is associated with persistent neuropathy in 48% of these patients. Due to stringent guidelines for dose modification, only a small proportion of patients experience serious functional impairment. However, as the cumulative dose of oxaliplatin increases, patients are more likely to experience paresthesia due to a cumulative sensory neuropathy. In patients who receive more than six cycles of chemotherapy (projected maximum cumulative dose given every three weeks of 780 mg/m2), the neuropathy can become persistent and affect the subject's ability to perform routine activities of daily living. Oxaliplatin-associated cumulative sensory neuropathy is slowly reversible in most patients [91-93]. Oxaliplatin-related neurotoxicity has no standard treatment. A variety of strategies have been employed to prevent or treat neurotoxicity to oxaliplatin [92,94,95], including treatment with carbamazepine, gabapentin, alpha lipoic acid, amifostine, glutathione, or celecoxib. The largest study of oxaliplatin-
Colorectal cancer survivors
347
related neurotoxicity is the retrospective analysis by Gamelin et al. [96]. It evaluated the benefit of infusing magnesium and calcium prior to administration of oxaliplatin in 96 patients compared to 65 patients who did not receive magnesium/calcium. The percentage of patients with grade 3 distal paresthesia was lower in the Ca/Mg group (7% vs. 26%, P = 0.001), and acute symptoms such as distal and lingual paresthesia were much less frequent and severe in the Ca/Mg group. The authors concluded that Ca/Mg infusions seemed to reduce the incidence and intensity of acute oxaliplatin-induced symptoms and might delay cumulative neuropathy. Despite these data, a large, randomised trial will be required to clarify the link between acute, transient symptoms and the likelihood of development of chronic sensory neuropathy, and to confirm whether strategies such as Ca/Mg infusions reduce the neurotoxicity without impacting on anti-tumour efficacy. Optimising the QOL of cancer survivors is of paramount importance, and continued research on oxaliplatin will help achieve this goal while also providing further progress in clinical benefit outcomes.
Conclusion The number of long-term survivors from CRC (> 5 years after diagnosis) continues to rise and increasing attention to the health problems and needs of this population has appeared. Many CRC survivors could return to a normal functioning after the completion of treatment and could live relatively without symptoms. However, cancer and its treatments can also provoque a wide spectrum of physical and psychological problems without improvement with time, which can affect the patients´ QOL. Although it is necessary more research to better understand these problems, the identification, surveillance and management of survivorship issues is a very relevant part of a survivorship care plan which would allow cancer survivors to be aware about the problems they are expected to present and the adequate strategies they could begin to alleviate and mitigate the impact of these long-term issues can have in their lives.
References 1. 2.
Jemal, A., Murray, T., Samuels, A., Ghafoor, A., Ward, E., and Thun, M.J. 2003, CA. Cancer J. Clin., 53, 5. American Cancer Society. 2008, Cancer Facts and Figures 2008, http://www.cancer.org/acs/groups/content/@nho/documents/document/2008cafffi nalsecuredpdf.pdf.
348
3. 4. 5. 6.
7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.
Esther U帽a Cid贸n et al.
Balch, G.C., De Meo, A., and Guillem, J.G. 2006, World J. Gastroenterol., 12, 3186. Meyerhardt, J.A. and Mayer, R.J. 2005, N. Engl. J. Med., 352, 476. Rao, S. and Cunningham, D. 2003, Scand. J. Surg., 92, 57. O'Connell, M.J., Laurie, J.A., Kahn, M., Fitzgibbons, R.J., Jr., Erlichman, C., Shepherd, L., Moertel, C.G., Kocha, W.I., Pazdur, R., Wieand, H.S., Rubin, J., Vukov, A.M., Donohue, J.H., Krook, J.E., and Figueredo, A. 1998, J. Clin. Oncol., 16, 295. Arkenau, H.T., Bermann, A., Rettig, K., Strohmeyer, G., and Porschen, R. 2003, Ann. Oncol., 14, 395. Hewitt, M., Greenfield, S., and Stovall, E. (Eds.). 2005, From Cancer Patient to Cancer Survivor: Lost in Transition, National Academies Press, Washington. Schwartz, S.I., Shires, G.T., Spencer, F.C., Daly, J.M., Fischer, J.E., and Galloway, A.C. (Eds.). 1998, Principles of Surgery, McGraw Hill, New York. Heald, R.J., Moran, B.J., Ryall, R.D., Sexton, R., and MacFarlane, J.K. 1998, Arch. Surg., 133, 894. Joseph, N.E., Sigurdson, E.R., Hanlon, A.L., Wang, H., Mayer, R.J., MacDonald, J.S., Catalano, P.J., and Haller, D.G. 2003, Ann. Surg. Oncol., 10, 213. Gordon, P.H. and Nivatvongs, S. 2001, Surgery of the Colon, Rectum, and Anus, Saunders, London. Corman, M.C. 2001, Surgery of the Colon and Rectum, Lippincott Williams & Wilkins, Philadelphia. Smith, R.L., Bohl, J.K., McElearney, S.T., Friel, C.M., Barclay, M.M., Sawyer, R.G., and Foley, E.F. 2004, Ann. Surg., 239, 599. Shin, J.Y. and Hong, K.H. 2008, World J. Surg., 32, 2287. Edna, T.H. and Bjerkeset, T. 1998, Eur. J. Surg., 164, 587. Jeong, W.K., Lim, S.B., Choi, H.S., and Jeong, S.Y. 2008, J. Gastrointest. Surg., 12, 926. Cameron, J. (Ed.). 2001, Current Surgical Therapy, Mosby, St. Louis. Cahalane, M.J., Shapiro, M.E., and Silen, W. 1989, Lancet, 1, 146. Hidalgo, M.P., Ferrero, E.H., Ortiz, M.A., Castillo, J.M., and Hidalgo, A.G. 2011, Hernia, Feb. 12. [Epub ahead of print] Millikan, K.W. 2003, Surg. Clin. North Am., 83, 1223. Flum, D.R., Horvath, K., and Koepsell, T. 2003, Ann. Surg., 237, 129. Ishida, H., Ishiguro, T., Ohsawa, T., Okada, N., Yokoyama, M., Kumamoto, K., Ishibashi, K., and Haga, N. 2010, Tech. Coloproctol., 14, 153. Lacy, A.M., Garcia-Valdecasas, J.C., Delgado, S., Castells, A., Taura, P., Pique, J.M., and Visa, J. 2002, Lancet, 359, 2224. Weeks, J.C., Nelson, H., Gelber, S., Sargent, D., and Schroeder, G. 2002, JAMA, 287, 321. Lian, L., Kalady, M., Geisler, D., and Kiran, R.P. 2010, Surg. Endosc., 24, 2039. Clinical Outcomes of Surgical Therapy Study Group, C.O.o.S.T.S. 2004, N. Engl. J. Med., 350, 2050. Bonjer, H.J., Hop, W.C., Nelson, H., Sargent, D.J., Lacy, A.M., Castells, A., Guillou, P.J., Thorpe, H., Brown, J., Delgado, S., Kuhrij, E., Haglind, E., and Pahlman, L. 2007, Arch. Surg., 142, 298.
Colorectal cancer survivors
349
29. Delaney, C.P., Kiran, R.P., Senagore, A.J., Brady, K., and Fazio, V.W. 2003, Ann. Surg., 238, 67. 30. Phillips, E.H., Franklin, M., Carroll, B.J., Fallas, M.J., Ramos, R., and Rosenthal, D. 1992, Ann. Surg., 216, 703. 31. Schwenk, W., Haase, O., Neudecker, J., and Muller, J.M. 2005, Cochrane Database Syst. Rev., 3, CD003145. 32. Schwenk, W., Bohm, B., and Muller, J.M. 1998, Surg. Endosc., 12, 1131. 33. Berends, F.J., Kazemier, G., Bonjer, H.J., and Lange, J.F. 1994, Lancet, 344, 58. 34. Gao, F., Cao, Y.F., and Chen, L.S. 2006, Int. J. Colorectal Dis., 21, 652. 35. Aziz, O., Constantinides, V., Tekkis, P.P., Athanasiou, T., Purkayastha, S., Paraskeva, P., Darzi, A.W., and Heriot, A.G. 2006, Ann. Surg. Oncol., 13, 413. 36. Kollmorgen, C.F., Meagher, A.P., Wolff, B.G., Pemberton, J.H., Martenson, J.A., and Illstrup, D.M. 1994, Ann. Surg., 220, 676. 37. Karanjia, N.D., Schache, D.J., and Heald, R.J. 1992, Br. J. Surg., 79, 114. 38. Lewis, W.G., Holdsworth, P.J., Stephenson, B.M., Finan, P.J., and Johnston, D. 1992, Br. J. Surg., 79, 1082. 39. Carmona, J.A., Ortiz, H., and Perez-Cabanas, I. 1991, Int. J. Colorectal Dis., 6, 108. 40. Williams, N.S., Price, R., and Johnston, D. 1980, Br. J. Surg., 67, 203. 41. Suzuki, H., Matsumoto, K., Amano, S., Fujioka, M., and Honzumi, M. 1980, Br. J. Surg., 67, 655. 42. Pedersen, I.K., Christiansen, J., Hint, K., Jensen, P., Olsen, J., and Mortensen, P.E. 1986, Ann. Surg., 204, 133. 43. Franco, S., Foresta, G., Cordio, S., Marletta, F., Lodin, M., Marino, G., and Giannone, G. 2004, G. Chir., 25, 117. 44. Chatwin, N.A., Ribordy, M., and Givel, J.C. 2002, Eur. J. Surg., 168, 297. 45. Dabirian, A., Yaghmaei, F., Rassouli, M., and Tafreshi, M.Z. 2010, Patient Prefer Adherence, 5, 1. 46. Krouse, R.S., Grant, M., Rawl, S.M., Mohler, M.J., Baldwin, C.M., Coons, S.J., McCorkle, R., Schmidt, C.M., and Ko, C.Y. 2009, J. Psychosom. Res., 66, 227. 47. Smith, D.M., Loewenstein, G., Rozin, P., Sherriff, R.L., and Ubel, P.A. 2007, J. Res. Pers., 41, 787. 48. Brown, H. and Randle, J. 2005, J. Clin. Nurs., 14, 74. 49. Symms, M.R., Rawl, S.M., Grant, M., Wendel, C.S., Coons, S.J., Hickey, S., Baldwin, C.M., and Krouse, R.S. 2008, Clin. Nurse Spec., 22, 30. 50. Ohman, U. 1982, Am. J. Surg., 143, 742. 51. Popek, S., Grant, M., Gemmill, R., Wendel, C.S., Mohler, M.J., Rawl, S.M., Baldwin, C.M., Ko, C.Y., Schmidt, C.M., and Krouse, R.S. 2010, Am. J. Surg., 200, 640. 52. Nybaek, H. and Jemec, G.B. 2010, J. Eur. Acad. Dermatol. Venereol., 24, 249. 53. Richbourg, L., Thorpe, J.M., and Rapp, C.G. 2007, J. Wound. Ostomy Continence Nurs., 34, 70. 54. Mitchell, K.A., Rawl, S.M., Schmidt, C.M., Grant, M., Ko, C.Y., Baldwin, C.M., Wendel, C., and Krouse, R.S. 2007, J. Wound. Ostomy Continence Nurs., 34, 524.
350
Esther U帽a Cid贸n et al.
55. Lynch, B.M., Hawkes, A.L., Steginga, S.K., Leggett, B., and Aitken, J.F. 2008, J. Wound. Ostomy Continence Nurs., 35, 424. 56. Havenga, K., Maas, C.P., DeRuiter, M.C., Welvaart, K., and Trimbos, J.B. 2000, Semin. Surg. Oncol., 18, 235. 57. Hendren, S.K., O'Connor, B.I., Liu, M., Asano, T., Cohen, Z., Swallow, C.J., Macrae, H.M., Gryfe, R., and McLeod, R.S. 2005, Ann. Surg., 242, 212. 58. Sideris, L., Zenasni, F., Vernerey, D., Dauchy, S., Lasser, P., Pignon, J.P., Elias, D., Di Palma, M., and Pocard, M. 2005, Dis. Colon Rectum, 48, 2180. 59. Guren, M.G., Eriksen, M.T., Wiig, J.N., Carlsen, E., Nesbakken, A., Sigurdsson, H.K., Wibe, A., and Tveit, K.M. 2005, Eur. J. Surg. Oncol., 31, 735. 60. Beck, D.E. and Wexner, S.D. (Eds.). 1998, Fundamentals of Anorectal Surgery, Saunders, London. 61. Havenga, K., DeRuiter, M.C., Enker, W.E., and Welvaart, K. 1996, Br. J. Surg., 83, 384. 62. Havenga, K., Enker, W.E., McDermott, K., Cohen, A.M., Minsky, B.D., and Guillem, J. 1996, J. Am. Coll. Surg., 182, 495. 63. Enker, W.E., Havenga, K., Polyak, T., Thaler, H., and Cranor, M. 1997, World J. Surg., 21, 715. 64. van Driel, M.F., Weymar Schultz, W.C., van de Wiel, H.B., Hahn, D.E., and Mensink, H.J. 1993, Eur. J. Surg. Oncol., 19, 183. 65. Marijnen, C.A., van de Velde, C.J., Putter, H., van den Brink, M., Maas, C.P., Martijn, H., Rutten, H.J., Wiggers, T., Kranenbarg, E.K., Leer, J.W., and Stiggelbout, A.M. 2005, J. Clin. Oncol., 23, 1847. 66. Lee, D.K., Jo, M.K., Song, K., Park, J.W., and Moon, S.M. 2010, Korean J. Urol., 51, 858. 67. Ho, V.P., Lee, Y., Stein, S.L., and Temple, L.K. 2011, Dis. Colon Rectum, 54, 113. 68. Lange, M.M. and van de Velde, C.J. 2011, Nat. Rev. Urol., 8, 51. 69. Rothenberger, D.A. and Wong, W.D. 1992, World J. Surg., 16, 478. 70. Havenga, K., Enker, W.E., Norstein, J., Moriya, Y., Heald, R.J., van Houwelingen, H.C., and van de Velde, C.J. 1999, Eur. J. Surg. Oncol., 25, 368. 71. Balslev, I. and Harling, H. 1983, Dis. Colon Rectum, 26, 785. 72. Gastrointestinal Tumor Study Group, G.T.S. 1985, N. Engl. J. Med., 312, 1465. 73. Krook, J.E., Moertel, C.G., Gunderson, L.L., Wieand, H.S., Collins, R.T., Beart, R.W., Kubista, T.P., Poon, M.A., Meyers, W.C., Mailliard, J.A., and et al. 1991, N. Engl. J. Med., 324, 709. 74. O'Connell, M.J., Martenson, J.A., Wieand, H.S., Krook, J.E., Macdonald, J.S., Haller, D.G., Mayer, R.J., Gunderson, L.L., and Rich, T.A. 1994, N. Engl. J. Med., 331, 502. 75. Wolmark, N., Wieand, H.S., Hyams, D.M., Colangelo, L., Dimitrov, N.V., Romond, E.H., Wexler, M., Prager, D., Cruz, A.B., Jr., Gordon, P.H., Petrelli, N.J., Deutsch, M., Mamounas, E., Wickerham, D.L., Fisher, E.R., Rockette, H., and Fisher, B. 2000, J. Natl. Cancer Inst., 92, 388.
Colorectal cancer survivors
351
76. Tepper, J.E., O'Connell, M., Niedzwiecki, D., Hollis, D.R., Benson, A.B., 3rd, Cummings, B., Gunderson, L.L., Macdonald, J.S., Martenson, J.A., and Mayer, R.J. 2002, J. Clin. Oncol., 20, 1744. 77. Smalley, S.R., Benedetti, J.K., Williamson, S.K., Robertson, J.M., Estes, N.C., Maher, T., Fisher, B., Rich, T.A., Martenson, J.A., Kugler, J.W., Benson, A.B., 3rd, Haller, D.G., Mayer, R.J., Atkins, J.N., Cripps, C., Pedersen, J., Periman, P.O., Tanaka, M.S., Jr., Leichman, C.G., and Macdonald, J.S. 2006, J. Clin. Oncol., 24, 3542. 78. Anonymous 1990, JAMA, 264, 1444. 79. Dahlberg, M., Glimelius, B., Graf, W., and Pahlman, L. 1998, Dis. Colon Rectum, 41, 543. 80. Dehni, N., McNamara, D.A., Schlegel, R.D., Guiguet, M., Tiret, E., and Parc, R. 2002, Dis. Colon Rectum, 45, 1635. 81. Pucciarelli, S., Gagliardi, G., Maretto, I., Lonardi, S., Friso, M.L., Urso, E., Toppan, P., and Nitti, D. 2009, Ann. Surg. Oncol., 16, 893. 82. Varma, J.S., Smith, A.N., and Busuttil, A. 1986, Gut, 27, 528. 83. Varma, J.S., Smith, A.N., and Busuttil, A. 1985, Br. J. Surg., 72, 875. 84. Varma, J.S. and Smith, A.N. 1986, Br. J. Surg., 73, 285. 85. Hricak, H. 1990, Magn. Reson. Q., 6, 276. 86. Birnbaum, E.H., Dreznik, Z., Myerson, R.J., Lacey, D.L., Fry, R.D., Kodner, I.J., and Fleshman, J.W. 1992, Dis. Colon Rectum, 35, 757. 87. Lewis, W.G., Williamson, M.E.R., Kuzu, A., Stephenson, B.M., Holdsworth, P.J., Finan, P.J., and Johnson, D. 1993, Gut, 34 (Suppl. 1), S1. 88. Nathanson, D.R., Espat, N.J., Nash, G.M., D'Alessio, M., Thaler, H., Minsky, B.D., Enker, W., Wong, D., Guillem, J., Cohen, A., and Paty, P.B. 2003, Dis. Colon Rectum, 46, 888. 89. Extra, J.M., Espie, M., Calvo, F., Ferme, C., Mignot, L., and Marty, M. 1990, Cancer Chemother. Pharmacol., 25, 299. 90. Leonard, G.D., Wright, M.A., Quinn, M.G., Fioravanti, S., Harold, N., Schuler, B., Thomas, R.R., and Grem, J.L. 2005, BMC Cancer, 5, 116. 91. Gamelin, E., Gamelin, L., Bossi, L., and Quasthoff, S. 2002, Semin. Oncol., 29, 21. 92. Grothey, A. 2003, Semin. Oncol., 30, 5. 93. Andre, T., Boni, C., Mounedji-Boudiaf, L., Navarro, M., Tabernero, J., Hickish, T., Topham, C., Zaninelli, M., Clingan, P., Bridgewater, J., Tabah-Fisch, I., and de Gramont, A. 2004, N. Engl. J. Med., 350, 2343. 94. Lersch, C., Schmelz, R., Eckel, F., Erdmann, J., Mayr, M., Schulte-Frohlinde, E., Quasthoff, S., Grosskreutz, J., and Adelsberger, H. 2002, Clin. Colorectal Cancer, 2, 54. 95. Agafitei, R.D., Schneider, S., Iqbal, S., Yang, D., Groshen, S., and Lenz, H.J. 2004, J. Clin. Oncol., 22 (July 15 Suppl.), Abstr. 3600. 96. Gamelin, L., Boisdron-Celle, M., Delva, R., Guerin-Meyer, V., Ifrah, N., Morel, A., and Gamelin, E. 2004, Clin. Cancer Res., 10, 4055.