Journal of Surgical Oncology Supplement 2:92-103 (1991)
Perspectives in Colorectal Cancer LEMUEL HERRERA, MD, MA, PEDRO LUNA, MD, JOSE R A M O N VILLARREAL, MD, MICHAEL BROWN, MD, JOSE SORRENTINO, MD, AND YlHAD KHALEK, MD From the Division of Surgical Oncology, The Medical Center of Delaware, Wilmington (L.H., M.B., I.S., Y.K.); Department of Surgery, lefferson Medical College, Philadelphia, Pennsylvania (L.H.); Hospital de Oncologia SlCLO XXI, Centro Medico Nacional IMSS, Mexico City (P.L.), and the Department of Medical Oncology, Hospital Universitario San Jose, Monterrey NL, Mexico (J.R.V.)
This paper presents an overview of recent developments pertaining to colorectal adenocarcinoma. It is aimed toward the practicing clinician. Topics discussed include epidemiologic observations; genetic predispositions; molecular biology findings; screening and early detection programs; endoscopy; principles of surgical resection; laser and radioimmunoguided surgery; staging; selection of patients for adjuvant chemotherapy; and considerations regarding biologic response modifiers and pain control in the advanced-disease setting. KEYWORDS: colorectal adenocarcinoma, screening programs, adjuvant chemotherapy
INTRODUCTION This paper is an attempt to bring to practicing clinicians some of the salient, clinically relevant points of recent advances and investigations in colorectal adenocarcinoma. Epidermoid carcinoma of the anal canal and other rarer large bowel malignancies are not discussed. EPIDEMIOLOGIC CONSIDERATIONS In a recent article based on data from the Connecticut Tumor Registry, Vukasin et al. noted the increasing incidence of cecal and sigmoid carcinoma in the last 50 years [l]. In men, the incidence of cecal carcinoma has increased from 3.6-16.7 cases per 100,000 population per year, and in women, it has increased from 4.9-14.2 cases per 100,000 population per year. The incidence of sigmoid carcinoma has increased from 8.7-1 8.7 cases per 100,000 population per year in men and from 7.7-12.8 cases per 100,000 population per year in women. The by-site distribution of colon cancer remained constant and the age-adjusted incidence increased at a constant rate, whereas the incidence of rectal cancer remained unchanged. The incidence of colon cancer continues to be higher in men than in women. These observations have led these authors to suggest that colon cancer is a twentieth-century epidemic and that its incidence will continue to increase. We strongly believe that this observation, although 0 1991 Wiley-Liss, Inc.
possibly correct, should be qualified, as proper reporting systems became available and the life expectancy at which this cancer would have a high prevalence was reached in the United States only after the 1930s [2]. On the other hand, these data reported from the Connecticut Tumor Registry compare well with data from an Italian study that demonstrated incidences of colorectal cancer at Trieste of 73.3 cases per 100,000 population per year in men and 56.2 cases per 100,000 population per year in women, and at Modena of 35.2 in men and 31.1 in women [3]. This supports the conclusion that as the population under medical surveillance continues to age and survive into decades at risk for colon cancer, the incidence of the diagnosis of colon cancer also will rise. Less clear is the comparative incidence in different races and possibly in individuals of different sexes or religions or with different sites of residence and diets. Certainly, the fact that there are specific countries and populations with a low incidence of colorectal cancer (e.g., Spaniards [Navarra], 8.1 cases; South Americans, 6.3 cases; and Filipinos, 4.1 cases, all age-standardized per 100,000 Accepted for publication May 30, 1991. Address reprint requests to Dr. Lemuel Herrerd, Division of Surgical Oncology, Department of Surgery, The Medical Center of Delaware, 501 West 14th Street, Wilmington, DE 19806. This study was partially supported by an educational grant from IVONYX.
Perspectives in Colorectal Cancer
population; and Mexicans*) makes studies in cancer control and comparative epidemiology relevant issues for the understanding of why and how colorectal cancers develop [4].
SCREENING AND EARLY DETECTION Screening for early detection of colorectal cancer are achievable tasks of proven benefit. Because colorectal cancer frequently is preceded by a precursor premalignant lesion, the adenomatous polyp, the goal of most screening programs is the detection of this lesion, as it clearly segregates and identifies a population with a high risk of developing large bowel cancer. Screening as a concept must be applied to ASYMPTOMATIC individuals from the general population. They must possess known high risk factors, such as a positive family history or a personal history of adenomatous polyps, inflammatory bowel disease, or cancer. Primary prevention of colorectal cancer still is not possible because a causal agent is not known; thus, the term “secondary prevention” has been developed as a pragmatic working concept. Secondary prevention consists of the following: 1. 2. 3. 4. 5.
Complete history and physical examination. A questionnaire for risk factors. Rectal digital examination. Test for occult blood in feces. Sigmoidoscopy.
We have reported previously that one fourth to one third of patients in whom colorectal polyps or cancer eventually was detected by either a retrospective or a prospective colonoscopy program had a positive fecal occult blood test [ 5 ] . Large-scale screening programs have shown that 3-6% of the general population will have a positive test for fecal occult blood [6,7]. Of these individuals, 30-50% will harbor a neoplasia; two thirds to three fourths of these patients will have a benign neoplasia (adenomatous polyp), and one fourth to one third will have a cancer. It is important to emphasize that in most of the reported trials of early detection, it has been shown that approximately three fourths of the cancers detected in the study populations were still “early” 181. An “early” cancer is defined as one that has not spread to the lymph nodes (Dukes’ A and B). In such cancers, the expected cure rate can be close to 80% or even higher in selected populations. More recently, in a prospective trial involving total colonoscopy in individ*In a Mexican study involving 9,412 consecutive autopsies in a general hospital caring for indigent populations, Albores-Saavedra found that 2,682 patients had died of cancer. Of these, 37 patients had rectal cancer and 25 had colonic cancer, demonstrating the Mexican population to be at low risk for colorectal cancer (personal communication, 1991).
93
uals 40 years of age or older, we have found that if there is a positive family history of a first-degree relative with large bowel polyps or cancer, the detection of a colonic neoplasia (polyp or cancer) in the screened individual will be approximately 40% 151. A significant percentage of these neoplasias would not have been detected if a total colonoscopy had not been performed. Thus, at present, a positive family history in a first-degree relative seems to be the most important and powerful predictive risk factor for an asymptomatic individual in the fourth decade or older to harbor a large bowel neoplasia. Therefore, a total colonoscopy is recommended for such individuals. Flexible sigmoidoscopy (66 cm) with an air-contrast barium enema is a second-best (because of the inability to biopsy and treat) but is an acceptable alternative for massive screening programs.
GENETICS There are recent reports that two vague categories of familial colorectal cancer exist. Although they do not have a defined mendelian pattern of inheritance, they clearly show a familial aggregation or clustering. These categories may be disclosed, at least in part, by the location of the primary tumor [9]. Of these two familial forms of colorectal cancer, the “nonpolyposis” form is associated with the development of tumors in a more proximal location (right colon) and a low incidence ( 15-25%) of associated adenomatous polyps (putative clinical markers), whereas the other form is linked to the inheritance of adenomatous polyps and it is often associated with tumors of the left colon. Of interest is that these clinical categories can be identified only after a meticulous pedigree and phenotypic data study. They are still far from clear-cut applications at the individual level, and yet, they are important when designing programs for cancer control and interpreting data from screening programs. The more defined genetic susceptibilities for colon cancer were recognized long ago. In the late nineteenth century, the syndrome of familial adenomatous polyposis (FAP) (previously known as “familial polyposis coli”) was described. More recently, through the efforts of Lynch et al., the hereditary nonpolyposis colorectal cancer (HNPCC) syndrome has been delineated [lo]. Because both premalignant susceptibilities are inherited in a mendelian autosomal dominant fashion and FAP can be correlated to clinical and molecular genetic markers, the development and implementation of surveillance programs for individuals with these inherited susceptibilities have demonstrated the potential to reduce the incidence and mortality of colorectal cancer secondary to these conditions. Specifically, in the case of FAP, the availability of genetic probes linked to the affected gene has made genetically directed screening of FAP relatives at risk a
94
Herrera et al.
real possibility. The emergence of these genetic markers with clinical utility in the FAP model has disclosed the potential for recognition of individuals who have inherited the diseased gene, who, until now, were discovered primarily by phenotypic manifestations (colorectal polyps) and analysis of the family pedigree. Relatives at risk who do not possess the defective gene after a careful study of all the data (family pedigree, endoscopy, genetic markers, age-onset table) then would avoid unnecessary endoscopic screening procedures and could be reassured that, being unaffected, their risk and that of their progeny of developing colorectal cancer is the same as that in the general population. These geno-phenotypic techniques also may be helpful in the investigation and elucidation of patients with unusual phenotypic features (e.g., familial aggregation of small numbers of adenomas) and in the calculation of familial risk of developing colorectal cancer. The FAP model has spearheaded the efforts to identify and correlate a variety of colonic and extracolonic phenotypic manifestations such as congenital pigmented ocular (retinal) lesions, osteomas of the jaw, and others to specific genetic markers [ 1 I]. Molecular genetic alterations associated with the morphologic abnormalities of premalignant polyps that supposedly mirror the progression from adenoma to carcinoma have become an important conceptual advance in the understanding of the pathogenesis of colorectal cancer. Studies of these alterations, including mutations in chromosome 5 , the ras gene (chromosome 12), mutation and deletion of the p53 gene (chromosome 17p), and deleted colon cancer gene (chromosome 1Sq), have culminated in the identification of the mutated gene for colon cancer (MCC) in the long arm of chromosome 5 [ 121. Evidence suggests that MCC can be an extremely rare type of mutation that can be inherited and that it may be the first gene mutation that sends normal cells into an abnormal growth pattern. Identification and quantification of some of these and other gene mutations and the overexpression of c-myc and K-ras oncogenes may play a relevant role in predicting colonic susceptibility to develop cancer, biologic commitment of adenomatous polyps for malignant transformation, and biologic aggressiveness of primary tumors. This understanding has the potential to enhance the predictive power of staging classifications and possibly could segregate individuals at higher risk of tumor recurrence, in whom more aggressive therapeutic interventions would be warranted [ 13,141. HNPCC syndrome is a disease inherited in an autosoma1 dominant mendelian fashion characterized phenotypically by the occurrence of cancer at a young age (mean age, polyp > tumor) and the poly-
amine pools [22]. ODC activity appears to be modulated by the environmental sites where metastases have occurred, so that the level of activity may change significantly if one compares samples from surrounding tissue with samples taken from primary or metastatic tumors [23]. Furthermore, the level of the enzyme activity varies significantly when compared with mRNA, indicating that this is a post-translational event [22]. ODC may be useful in predicting relatives at risk for FAP who have inherited the diseased gene. It may play a role in monitoring colorectal disease, as polyamine pools can be manipulated to modify cell growth. Others have shown relevant correlations between a high level of phorbin, a gene induced and controlled by PKC, and tumor invasiveness ~41. Flow cytometry has undergone significant refinements from permitting clinicians simply to calculate the DNA index, to becoming a possible additive, independent prognostic parameter [25]. It now permits the identification of specific cell subpopulations in phases of interest in the tumor when combined with the use of monoclonal antibodies. It also allows for a much more refined identification of markers such as oncogenes (c-myc, c-K-ras), hormone receptors, and others, as well as for their localization and allocation to specific tumor cell subpopulations. As the technique of flow cytometry and monoclonal antibodies is complemented with techniques such as the polymerase chain reaction (PCR), an enormous amount of information will be able to be derived from relatively small tumor samples containing less than 100 cells from which millions of copies of the gene of interest then can be obtained by PCR for subsequent analytical studies. Clearly, uncontrolled cell growth detected or controlled by oncogenes/antioncogenes does not confer on a tumor the property of being either benign or malignant. Other properties such as the ability to invade locally, to produce metastases regionally or distantly, or to be resistant to chemotherapeutic agents are all characteristics of prime oncologic importance. However, overexpression or underexpression of oncogenesiantioncogenes (i.e., differential expression of p53 and ras in colon, breast, and lung cancers) seems to be clearly implicated in the mechanism of many of these oncologic phenomena. At the nuclear level, DNA is known to be the site of hereditary changes responsible for some neoplasias. In mammals, all normal cells possess the same genes. Differentiation and development are believed to be dependent on activation of some genes but not others. Cell transformation seems to be influenced by the occurrence of genetic events that uncouple the cell from its normal regulatory mechanisms, and oncogenes and antioncogenes have been found to play a significant role under experimental conditions. An understanding of these control mechanisms and their signals and mediators
Perspectives in Colorectal Cancer could be of obvious importance for the disclosure of a potential biomarker. Great strides have been made in the area of genetic markers for diagnosis. For these markers to become clinically useful, it was first necessary that the patterns of inheritance be clearly elucidated and ascertained by clinicians, and then the associations were searched for to establish relationships between the markers and specific clinical phases of the disease. Thus, it was in order to understand large bowel cancer that FAP became the working model of choice because it was inherited in a mendelian autosomal dominant pattern, was 100% premalignant, and affected young individuals. The task of identifying potential markers was made much easier after a chromosomal site for the defective FAP gene was located [26]. The simultaneous wide availability of DNA probes and refinement of restriction fragment length polymorphism (RFLP) techniques, coupled with the analysis of pedigrees and DNA obtained from at least three generations of patients, made the genetically directed screening of individuals at risk from FAP kindreds and the search for associated genetic markers a sound investigative venture [27]. Many of the DNA markers are at sites physically very near to the precise location of the genetic defect and flanking it. Therefore, even if we do not know the exact location or the correct spelling of the abnormal gene, we still could make the diagnosis with accuracy by employing restriction endonucleases to a single strand of DNA to cut fragments of various sizes clearly traceable to the affected parent and then identifying these fragments by complementary pieces of radioactive DNA (RFLPs). The examination of DNA is performed by a technique known as “Southern blotting,” in which DNA fragments are separated by agarose gel electrophoresis. Singlestrand fragments then are obtained and “blotted” onto a sheet of nitrocellulose. These strands are exposed to prepared cDNA probes from a genomic library into which p32 has been incorporated (this labeling is known as “nick translation”). Hybridization then occurs between complementary strands of DNA, with the radioactive DNA corresponding to the gene of interest and making the DNA probe a marker associated with the genetic defect. In Northern blotting, p32 DNA fragments are hybridized to RNA, whereas in Western blotting, the technique calls for staining by means of an antibody directed against a protein product of a specific gene.
SURGICAL CONSIDERATIONS Despite advances in radiotherapy, chemotherapy, and immunotherapy in the last decade, surgery remains the mainstay of curative treatment for colorectal cancer. Surgery must include wide resection of the primary tumor, with documented healthy margins of at least 2 cm, but preferably 5 cm, and the removal of the draining
97
lymph nodes. Organs attached to the primary tumor should be resected en bloc, and the adhesions between them should not be freed by the surgical operator, as the peritoneal cavity may be contaminated with cancer cells, converting a potentially curative procedure (B3) into an incurable one. Surgery also plays a significant although less effective role in the management of metastasis. Nevertheless, substantial long-term disease-free survival can be obtained by the judicious application of extensive surgical techniques. On the other hand, palliation of bleeding, obstruction, or perforations can be obtained most quickly by surgical methods. We also have recommended that a bilateral salpingooophorectomy be performed at the time of the initial operation for the following reasons: 1. Colorectal cancer is mainly a disease of perimenopausal and postmenopausal individuals. 2. The presence of ovarian metastases abrogates the prognostic value of the Dukes’ staging system. 3. A small but definite number of patients (up to 5%) will develop primary ovarian cancer during subsequent follow-up. Because there is no known method for early diagnosis of ovarian cancer, oophorectomy should be considered as a prophylactic measure when the function of the ovaries is no longer considered necessary and especially when there is a positive family history in a firstdegree relative. 4. There is no increase in morbidity or mortality when oophorectomy is added to a planned primary resectional procedure [28].
For rectal cancer, we prefer an anterior resection and a colorectal anastomosis for those tumors located 10-15 cm from the anal verge and an abdominoperineal resection for those located 0-5 cm from the anal verge. We also have been performing an abdominoperineal resection for those tumors located between 5 and 10 cm from the anal verge because we have obtained good local control of the primary tumor (
Perspectives in Colorectal Cancer LEMUEL HERRERA, MD, MA, PEDRO LUNA, MD, JOSE R A M O N VILLARREAL, MD, MICHAEL BROWN, MD, JOSE SORRENTINO, MD, AND YlHAD KHALEK, MD From the Division of Surgical Oncology, The Medical Center of Delaware, Wilmington (L.H., M.B., I.S., Y.K.); Department of Surgery, lefferson Medical College, Philadelphia, Pennsylvania (L.H.); Hospital de Oncologia SlCLO XXI, Centro Medico Nacional IMSS, Mexico City (P.L.), and the Department of Medical Oncology, Hospital Universitario San Jose, Monterrey NL, Mexico (J.R.V.)
This paper presents an overview of recent developments pertaining to colorectal adenocarcinoma. It is aimed toward the practicing clinician. Topics discussed include epidemiologic observations; genetic predispositions; molecular biology findings; screening and early detection programs; endoscopy; principles of surgical resection; laser and radioimmunoguided surgery; staging; selection of patients for adjuvant chemotherapy; and considerations regarding biologic response modifiers and pain control in the advanced-disease setting. KEYWORDS: colorectal adenocarcinoma, screening programs, adjuvant chemotherapy
INTRODUCTION This paper is an attempt to bring to practicing clinicians some of the salient, clinically relevant points of recent advances and investigations in colorectal adenocarcinoma. Epidermoid carcinoma of the anal canal and other rarer large bowel malignancies are not discussed. EPIDEMIOLOGIC CONSIDERATIONS In a recent article based on data from the Connecticut Tumor Registry, Vukasin et al. noted the increasing incidence of cecal and sigmoid carcinoma in the last 50 years [l]. In men, the incidence of cecal carcinoma has increased from 3.6-16.7 cases per 100,000 population per year, and in women, it has increased from 4.9-14.2 cases per 100,000 population per year. The incidence of sigmoid carcinoma has increased from 8.7-1 8.7 cases per 100,000 population per year in men and from 7.7-12.8 cases per 100,000 population per year in women. The by-site distribution of colon cancer remained constant and the age-adjusted incidence increased at a constant rate, whereas the incidence of rectal cancer remained unchanged. The incidence of colon cancer continues to be higher in men than in women. These observations have led these authors to suggest that colon cancer is a twentieth-century epidemic and that its incidence will continue to increase. We strongly believe that this observation, although 0 1991 Wiley-Liss, Inc.
possibly correct, should be qualified, as proper reporting systems became available and the life expectancy at which this cancer would have a high prevalence was reached in the United States only after the 1930s [2]. On the other hand, these data reported from the Connecticut Tumor Registry compare well with data from an Italian study that demonstrated incidences of colorectal cancer at Trieste of 73.3 cases per 100,000 population per year in men and 56.2 cases per 100,000 population per year in women, and at Modena of 35.2 in men and 31.1 in women [3]. This supports the conclusion that as the population under medical surveillance continues to age and survive into decades at risk for colon cancer, the incidence of the diagnosis of colon cancer also will rise. Less clear is the comparative incidence in different races and possibly in individuals of different sexes or religions or with different sites of residence and diets. Certainly, the fact that there are specific countries and populations with a low incidence of colorectal cancer (e.g., Spaniards [Navarra], 8.1 cases; South Americans, 6.3 cases; and Filipinos, 4.1 cases, all age-standardized per 100,000 Accepted for publication May 30, 1991. Address reprint requests to Dr. Lemuel Herrerd, Division of Surgical Oncology, Department of Surgery, The Medical Center of Delaware, 501 West 14th Street, Wilmington, DE 19806. This study was partially supported by an educational grant from IVONYX.
Perspectives in Colorectal Cancer
population; and Mexicans*) makes studies in cancer control and comparative epidemiology relevant issues for the understanding of why and how colorectal cancers develop [4].
SCREENING AND EARLY DETECTION Screening for early detection of colorectal cancer are achievable tasks of proven benefit. Because colorectal cancer frequently is preceded by a precursor premalignant lesion, the adenomatous polyp, the goal of most screening programs is the detection of this lesion, as it clearly segregates and identifies a population with a high risk of developing large bowel cancer. Screening as a concept must be applied to ASYMPTOMATIC individuals from the general population. They must possess known high risk factors, such as a positive family history or a personal history of adenomatous polyps, inflammatory bowel disease, or cancer. Primary prevention of colorectal cancer still is not possible because a causal agent is not known; thus, the term “secondary prevention” has been developed as a pragmatic working concept. Secondary prevention consists of the following: 1. 2. 3. 4. 5.
Complete history and physical examination. A questionnaire for risk factors. Rectal digital examination. Test for occult blood in feces. Sigmoidoscopy.
We have reported previously that one fourth to one third of patients in whom colorectal polyps or cancer eventually was detected by either a retrospective or a prospective colonoscopy program had a positive fecal occult blood test [ 5 ] . Large-scale screening programs have shown that 3-6% of the general population will have a positive test for fecal occult blood [6,7]. Of these individuals, 30-50% will harbor a neoplasia; two thirds to three fourths of these patients will have a benign neoplasia (adenomatous polyp), and one fourth to one third will have a cancer. It is important to emphasize that in most of the reported trials of early detection, it has been shown that approximately three fourths of the cancers detected in the study populations were still “early” 181. An “early” cancer is defined as one that has not spread to the lymph nodes (Dukes’ A and B). In such cancers, the expected cure rate can be close to 80% or even higher in selected populations. More recently, in a prospective trial involving total colonoscopy in individ*In a Mexican study involving 9,412 consecutive autopsies in a general hospital caring for indigent populations, Albores-Saavedra found that 2,682 patients had died of cancer. Of these, 37 patients had rectal cancer and 25 had colonic cancer, demonstrating the Mexican population to be at low risk for colorectal cancer (personal communication, 1991).
93
uals 40 years of age or older, we have found that if there is a positive family history of a first-degree relative with large bowel polyps or cancer, the detection of a colonic neoplasia (polyp or cancer) in the screened individual will be approximately 40% 151. A significant percentage of these neoplasias would not have been detected if a total colonoscopy had not been performed. Thus, at present, a positive family history in a first-degree relative seems to be the most important and powerful predictive risk factor for an asymptomatic individual in the fourth decade or older to harbor a large bowel neoplasia. Therefore, a total colonoscopy is recommended for such individuals. Flexible sigmoidoscopy (66 cm) with an air-contrast barium enema is a second-best (because of the inability to biopsy and treat) but is an acceptable alternative for massive screening programs.
GENETICS There are recent reports that two vague categories of familial colorectal cancer exist. Although they do not have a defined mendelian pattern of inheritance, they clearly show a familial aggregation or clustering. These categories may be disclosed, at least in part, by the location of the primary tumor [9]. Of these two familial forms of colorectal cancer, the “nonpolyposis” form is associated with the development of tumors in a more proximal location (right colon) and a low incidence ( 15-25%) of associated adenomatous polyps (putative clinical markers), whereas the other form is linked to the inheritance of adenomatous polyps and it is often associated with tumors of the left colon. Of interest is that these clinical categories can be identified only after a meticulous pedigree and phenotypic data study. They are still far from clear-cut applications at the individual level, and yet, they are important when designing programs for cancer control and interpreting data from screening programs. The more defined genetic susceptibilities for colon cancer were recognized long ago. In the late nineteenth century, the syndrome of familial adenomatous polyposis (FAP) (previously known as “familial polyposis coli”) was described. More recently, through the efforts of Lynch et al., the hereditary nonpolyposis colorectal cancer (HNPCC) syndrome has been delineated [lo]. Because both premalignant susceptibilities are inherited in a mendelian autosomal dominant fashion and FAP can be correlated to clinical and molecular genetic markers, the development and implementation of surveillance programs for individuals with these inherited susceptibilities have demonstrated the potential to reduce the incidence and mortality of colorectal cancer secondary to these conditions. Specifically, in the case of FAP, the availability of genetic probes linked to the affected gene has made genetically directed screening of FAP relatives at risk a
94
Herrera et al.
real possibility. The emergence of these genetic markers with clinical utility in the FAP model has disclosed the potential for recognition of individuals who have inherited the diseased gene, who, until now, were discovered primarily by phenotypic manifestations (colorectal polyps) and analysis of the family pedigree. Relatives at risk who do not possess the defective gene after a careful study of all the data (family pedigree, endoscopy, genetic markers, age-onset table) then would avoid unnecessary endoscopic screening procedures and could be reassured that, being unaffected, their risk and that of their progeny of developing colorectal cancer is the same as that in the general population. These geno-phenotypic techniques also may be helpful in the investigation and elucidation of patients with unusual phenotypic features (e.g., familial aggregation of small numbers of adenomas) and in the calculation of familial risk of developing colorectal cancer. The FAP model has spearheaded the efforts to identify and correlate a variety of colonic and extracolonic phenotypic manifestations such as congenital pigmented ocular (retinal) lesions, osteomas of the jaw, and others to specific genetic markers [ 1 I]. Molecular genetic alterations associated with the morphologic abnormalities of premalignant polyps that supposedly mirror the progression from adenoma to carcinoma have become an important conceptual advance in the understanding of the pathogenesis of colorectal cancer. Studies of these alterations, including mutations in chromosome 5 , the ras gene (chromosome 12), mutation and deletion of the p53 gene (chromosome 17p), and deleted colon cancer gene (chromosome 1Sq), have culminated in the identification of the mutated gene for colon cancer (MCC) in the long arm of chromosome 5 [ 121. Evidence suggests that MCC can be an extremely rare type of mutation that can be inherited and that it may be the first gene mutation that sends normal cells into an abnormal growth pattern. Identification and quantification of some of these and other gene mutations and the overexpression of c-myc and K-ras oncogenes may play a relevant role in predicting colonic susceptibility to develop cancer, biologic commitment of adenomatous polyps for malignant transformation, and biologic aggressiveness of primary tumors. This understanding has the potential to enhance the predictive power of staging classifications and possibly could segregate individuals at higher risk of tumor recurrence, in whom more aggressive therapeutic interventions would be warranted [ 13,141. HNPCC syndrome is a disease inherited in an autosoma1 dominant mendelian fashion characterized phenotypically by the occurrence of cancer at a young age (mean age, polyp > tumor) and the poly-
amine pools [22]. ODC activity appears to be modulated by the environmental sites where metastases have occurred, so that the level of activity may change significantly if one compares samples from surrounding tissue with samples taken from primary or metastatic tumors [23]. Furthermore, the level of the enzyme activity varies significantly when compared with mRNA, indicating that this is a post-translational event [22]. ODC may be useful in predicting relatives at risk for FAP who have inherited the diseased gene. It may play a role in monitoring colorectal disease, as polyamine pools can be manipulated to modify cell growth. Others have shown relevant correlations between a high level of phorbin, a gene induced and controlled by PKC, and tumor invasiveness ~41. Flow cytometry has undergone significant refinements from permitting clinicians simply to calculate the DNA index, to becoming a possible additive, independent prognostic parameter [25]. It now permits the identification of specific cell subpopulations in phases of interest in the tumor when combined with the use of monoclonal antibodies. It also allows for a much more refined identification of markers such as oncogenes (c-myc, c-K-ras), hormone receptors, and others, as well as for their localization and allocation to specific tumor cell subpopulations. As the technique of flow cytometry and monoclonal antibodies is complemented with techniques such as the polymerase chain reaction (PCR), an enormous amount of information will be able to be derived from relatively small tumor samples containing less than 100 cells from which millions of copies of the gene of interest then can be obtained by PCR for subsequent analytical studies. Clearly, uncontrolled cell growth detected or controlled by oncogenes/antioncogenes does not confer on a tumor the property of being either benign or malignant. Other properties such as the ability to invade locally, to produce metastases regionally or distantly, or to be resistant to chemotherapeutic agents are all characteristics of prime oncologic importance. However, overexpression or underexpression of oncogenesiantioncogenes (i.e., differential expression of p53 and ras in colon, breast, and lung cancers) seems to be clearly implicated in the mechanism of many of these oncologic phenomena. At the nuclear level, DNA is known to be the site of hereditary changes responsible for some neoplasias. In mammals, all normal cells possess the same genes. Differentiation and development are believed to be dependent on activation of some genes but not others. Cell transformation seems to be influenced by the occurrence of genetic events that uncouple the cell from its normal regulatory mechanisms, and oncogenes and antioncogenes have been found to play a significant role under experimental conditions. An understanding of these control mechanisms and their signals and mediators
Perspectives in Colorectal Cancer could be of obvious importance for the disclosure of a potential biomarker. Great strides have been made in the area of genetic markers for diagnosis. For these markers to become clinically useful, it was first necessary that the patterns of inheritance be clearly elucidated and ascertained by clinicians, and then the associations were searched for to establish relationships between the markers and specific clinical phases of the disease. Thus, it was in order to understand large bowel cancer that FAP became the working model of choice because it was inherited in a mendelian autosomal dominant pattern, was 100% premalignant, and affected young individuals. The task of identifying potential markers was made much easier after a chromosomal site for the defective FAP gene was located [26]. The simultaneous wide availability of DNA probes and refinement of restriction fragment length polymorphism (RFLP) techniques, coupled with the analysis of pedigrees and DNA obtained from at least three generations of patients, made the genetically directed screening of individuals at risk from FAP kindreds and the search for associated genetic markers a sound investigative venture [27]. Many of the DNA markers are at sites physically very near to the precise location of the genetic defect and flanking it. Therefore, even if we do not know the exact location or the correct spelling of the abnormal gene, we still could make the diagnosis with accuracy by employing restriction endonucleases to a single strand of DNA to cut fragments of various sizes clearly traceable to the affected parent and then identifying these fragments by complementary pieces of radioactive DNA (RFLPs). The examination of DNA is performed by a technique known as “Southern blotting,” in which DNA fragments are separated by agarose gel electrophoresis. Singlestrand fragments then are obtained and “blotted” onto a sheet of nitrocellulose. These strands are exposed to prepared cDNA probes from a genomic library into which p32 has been incorporated (this labeling is known as “nick translation”). Hybridization then occurs between complementary strands of DNA, with the radioactive DNA corresponding to the gene of interest and making the DNA probe a marker associated with the genetic defect. In Northern blotting, p32 DNA fragments are hybridized to RNA, whereas in Western blotting, the technique calls for staining by means of an antibody directed against a protein product of a specific gene.
SURGICAL CONSIDERATIONS Despite advances in radiotherapy, chemotherapy, and immunotherapy in the last decade, surgery remains the mainstay of curative treatment for colorectal cancer. Surgery must include wide resection of the primary tumor, with documented healthy margins of at least 2 cm, but preferably 5 cm, and the removal of the draining
97
lymph nodes. Organs attached to the primary tumor should be resected en bloc, and the adhesions between them should not be freed by the surgical operator, as the peritoneal cavity may be contaminated with cancer cells, converting a potentially curative procedure (B3) into an incurable one. Surgery also plays a significant although less effective role in the management of metastasis. Nevertheless, substantial long-term disease-free survival can be obtained by the judicious application of extensive surgical techniques. On the other hand, palliation of bleeding, obstruction, or perforations can be obtained most quickly by surgical methods. We also have recommended that a bilateral salpingooophorectomy be performed at the time of the initial operation for the following reasons: 1. Colorectal cancer is mainly a disease of perimenopausal and postmenopausal individuals. 2. The presence of ovarian metastases abrogates the prognostic value of the Dukes’ staging system. 3. A small but definite number of patients (up to 5%) will develop primary ovarian cancer during subsequent follow-up. Because there is no known method for early diagnosis of ovarian cancer, oophorectomy should be considered as a prophylactic measure when the function of the ovaries is no longer considered necessary and especially when there is a positive family history in a firstdegree relative. 4. There is no increase in morbidity or mortality when oophorectomy is added to a planned primary resectional procedure [28].
For rectal cancer, we prefer an anterior resection and a colorectal anastomosis for those tumors located 10-15 cm from the anal verge and an abdominoperineal resection for those located 0-5 cm from the anal verge. We also have been performing an abdominoperineal resection for those tumors located between 5 and 10 cm from the anal verge because we have obtained good local control of the primary tumor (
