The Gastrointestinal Cancer Organ Site Committee with editorial assistance from Maryellen Daston

Welcome to the MMHCC Gastrointestinal Cancer Site. The introductory section provides background information on colorectal cancer and a general overview of current methods for diagnosis and treatment, molecular alterations occurring in colorectal cancer, and existing murine models of intestinal neoplasia. The introductory section is followed by several sections providing more in-depth discussion of the topics listed below. Links to other Internet sites that provide additional information are included throughout.

Introduction

Each year, more than 150,000 Americans are diagnosed with colorectal cancer (CRC) and more than 50,000 die from this disease. It is the fourth most commonly diagnosed cancer and the second leading cause of cancer-related deaths in the United States (1,2). A number of epidemiological factors have a clear relationship to CRC. First, it is a disease associated with increasing age. The incidence rate increases from less than 1 per 100,000 for people below the age of 20 to 450 per 100,000 for people over 85 years old. Second, there is strong evidence that the risk of CRC can be modified by lifestyle or environmental factors, particularly, diet and exercise. A low-fiber, high-fat diet combined with a sedentary lifestyle is associated with a high incidence of CRC. Epidemiological studies have demonstrated that smoking, meat consumption, and alcohol consumption are all risk factors, while there is an inverse association with vegetable consumption, prolonged use of non-steroidal anti-inflammatory drugs, estrogen replacement therapy and physical activity (3). Third, research on the molecular pathogenesis of this disease has revealed a number of germline mutations associated with inherited predisposition to CRC including APC and the mismatch repair (MMR) genes, hMSH2, hMSH6, hMLH1, hPMS1, and hPMS2. Fourth, adenomatous polyps are recognized as non-invasive precursors to CRC, providing the rationale for the use of screening tests such as the fecal occult blood test, sigmoidoscopy, colonoscopy, and barium enema to identify these lesions prior to malignant degeneration.

CRC incidence and mortality rates are higher among African Americans than among white Americans. The overall incidence for African American males is 57.8 per 100,000 as compared to 52.0 per 100,000 in white males. CRC occurs at a rate of 44.7 per 100,000 among African American females while white females are diagnosed at a rate of 36.8 per 100,000 individuals. Overall mortality rates are 27.5 and 19.7 per 100,000 among African American males and females, respectively, as compared to 20.6 and 13.9 per 100,000 for white men and white women (2). For more information on colorectal cancer statistics see: https://webarchive.library.unt.edu/eot2008/20090131090926/http://seer.cancer.gov
https://webarchive.library.unt.edu/eot2008/20090131090926/http://prg.nci.nih.gov

An increased understanding of the molecular pathways that control the growth and differentiation of intestinal epithelial cells (see GI Development and Biology) and the identification of those involved in tumor initiation and progression to malignancy are important achievements of the past ten years. Multiple genes with altered expression in colorectal tumors have now been clustered into three key intracellular signaling pathway: APC/Wnt, Ras, and TGF beta;. Another mechanism of carcinogenesis involves the cellular DNA MMR system. The identification of genes that are affected by the inactivation of these pathways has revealed potential therapeutic targets.

The development of mouse models that mimic human disease has played a vital role in our increased understanding of the molecular etiology of CRC. These models also facilitate the development and testing of novel targeted therapies. CRC is a disorder with multiple causes with differing molecular pathways to carcinogenesis, and the variety of available murine models reflects this heterogeneity. Murine models include those with chemically induced tumors; chemically mutagenized strains with germline mutations of relevant genes; and genetically engineered strains with deficiencies in tumor suppressor genes, overexpression of oncogenes, or impaired immune function. Strain-specific differences in lifespan, tumor location, histology, and age of onset may offer appropriate models for many aspects of human CRC. These differences are also instructive with respect to the precise roles of the affected genes in each strain.

Diagnosis and Treatment (See Tumor Classification and Staging)

A common clinical feature of CRC is blood loss in the GI tract, sometimes leading to anemia. Anemia may be detected on a routine blood test in someone without symptoms or may be detected due to symptoms emanating from the anemia, including weakness, fatigue, and/or shortness of breath on exertion. A stool test that reveals small amounts of blood (the fecal occult blood test or FOBT) should prompt further evaluation of the GI tract to locate sites of blood loss. This would include fiberoptic endoscopy of both the upper GI tract (esophagus, stomach, and duodenum) and the lower GI tract (colon and rectum). The advantage of fiberoptic endoscopy is that any suspicious lesions can be biopsied for histopathologic diagnosis. Other symptoms associated with colorectal cancer include: melena (dark, tarry-looking stools caused by partially digested blood usually emanating from upper-to-mid colon cancers), rectal bleeding (also known as hematochezia and usually associated with lower colon or rectal cancers), changes in bowel habits (persistent diarrhea or constipation, or alternating diarrhea and constipation), abdominal cramping, or pain.

Because of the prevalence of CRC, the enhanced survival of patients with early-stage lesions, and the relative simplicity and accuracy of screening tests, routine screening is recommended for all adults over the age of 50 years, and is especially important for those with higher-than-average risk. Groups identified with a high incidence of CRC include those with hereditary conditions, (e.g. familial polyposis or hereditary nonpolyposis colon cancer [HNPCC]), ulcerative colitis, personal or first-degree family history of CRC or adenomas, and personal history of ovarian or endometrial cancer. Screening tests for CRC include the FOBT, sigmoidoscopy, barium enema, and colonoscopy (4).

The choice of treatment for CRC is largely dependent on disease stage. If located within the colon, surgical removal of the entire cancer, surrounding area of normal colon, and regional lymph nodes is the standard therapy for patients with Stage I (or Dukes' Stage A) colon cancer. With surgery alone, long-term survival is excellent, with 5-year survival rates exceeding 90%. A similar approach is taken if the primary tumor is located in the rectum, with every attempt made to reconnect the bowel internally and preserve sphincter function to avoid the need for a colostomy.

Patients with adenocarcinoma of the colon that has spread to nearby lymph nodes (i.e., Stage III or Dukes' Stage C) have a 50% risk of death within 5 years if they are treated with surgery alone. This is due to the presence of subclinical (occult) tumor metastases that are not removed by surgery. Systemic therapy administered after surgical resection of all visible disease is termed "adjuvant therapy". Adjuvant therapy for patients with Stage III disease improves 5-year survival significantly to approximately 65%. Recent Phase III clinical trials have demonstrated that 6 months of therapy with 5-FU and leucovorin is as effective as 1 year of therapy with 5-FU and levamisole (5).

The role of adjuvant therapy for patients with colon cancer that has spread into or through the muscular layer of the colon, but not to lymph nodes or distant organs (Stage II or Dukes' Stage B disease) is controversial. Analysis of pooled data from multiple randomized trials has led to divergent conclusions. The relative reduction in mortality at 5 years for patients with Stage II colon cancer is approximately 30%, which is similar to the impact observed in Stage III patients (6). This has led some to conclude that adjuvant therapy should be considered the standard of care for these patients. However, this difference did not reach statistical significance in pooled analysis, leading others to conclude that adjuvant therapy should not be administered to all patients with Stage II disease and only those with the highest risk should receive post-operative treatment.

Unlike colon cancer, in rectal cancer the role of adjuvant therapy is well established for both for Stage II and Stage III disease due to its proven impact on reducing local as well as systemic relapses and improving overall survival. Treatment involves the combined use of chemotherapy and radiation. Typically, this has consisted of 2 cycles of systemic chemotherapy, usually 5-FU, followed by 5 weeks of external beam radiation and continuous infusion 5-FU, followed by 2 more months of 5-FU. This combined-modality approach to adjuvant therapy reduces 5-year mortality by approximately 30%, similar to the results obtained with chemotherapy alone in patients with colon cancer. Over the past few years, increasing attention has been given to administration of this therapy prior to surgery, an approach known as neoadjuvant therapy. Advantages include the potential to reduce the size of a locally advanced tumor so that less extensive surgery is required to remove all local disease, thus reducing the likelihood that a colostomy would be needed. The disadvantage of this approach is that it has never been compared head-to-head against standard post-operative adjuvant chemoradiotherapy and long-term results with this approach remain unknown.

For nearly 40 years, 5-FU administered in an intermittent intravenous bolus fashion has been the cornerstone of treatment for patients with CRC with distant metastases (Stage IV or Dukes' Stage D). Although the addition of leucovorin increases the objective response rate (i.e., the proportion of patients with at least 50% shrinkage of tumor size by classical bidimensional measurement on CT or MRI scan or ³ 30% shrinkage of tumor size on unidimensional measurement), it does not substantially improve survival. Administration of 5-FU by continuous intravenous infusion improves response rate and reduces some of the most troublesome toxicities, but does not result in a substantial improvement in survival and requires the insertion of an indwelling, semi-permanent intravenous catheter. An oral 5-FU prodrug, capecitabine (Xeloda), produces therapeutic results similar to intravenous, bolus 5-FU, with less mucositis and myelosuppression, but is associated with an increased risk of hand-foot syndromes. With any of these strategies, median survival is 10 to 14 months.

Irinotecan (CPT-11, Camptosar) converts the normal nuclear enzyme, topoisomerase-I, into a cellular poison. Irinotecan was first approved for use as a single-agent in patients who had progressive disease during or shortly after receiving 5-FU-based chemotherapy. Irinotecan improved survival when compared to either best supportive care or infusional 5-FU in patients with relapsed or refractory CRC. More recently, the combination of irinotecan, 5-FU, and leucovorin (IFL) has resulted in superior survival to that achieved with 5-FU and leucovorin. Median survival with the three-drug regimen is 15 to 17 months compared with 12 to 14 months for 5-FU and leucovorin.

For more information on CRC diagnosis, screening and treatment see: https://webarchive.library.unt.edu/eot2008/20090131090926/http://cancer.gov/cancer_information/cancer_type/colon_and_rectal

Overview of Molecular Alterations in Human Colorectal Cancer (For a more detailed discussion see Molecular Alterations in CRC)

The development of CRC is a multistep progression beginning with the transformation of normal colonic epithelium to an adenomatous polyp and, subsequently to an invasive cancer (7). This is a slow process requiring years, and possibly decades. Recently, several characteristic genetic alterations have been identified that contribute to this neoplastic progression. Most of the affected genes can be clustered into two key pathways, the APC/Wnt pathway and the TGFβ pathway. Mutations leading to inactivation of the tumor-suppressor gene adenomatous polyposis coli (APC) and dysregulation of the K-RAS protooncogene are early events in colorectal carcinogenesis. Loss of heterozygosity (LOH) on the long arm of chromosome 18 (18q) is a later event; candidate loci in this region include deleted in colorectal cancer (DCC), DPC4 and MADR2, with the latter two being components of the TGFβ pathway. The mutation of the tumor-suppressor gene, p53, is another late event in CRC that may allow cells of the growing tumor to evade cell cycle arrest and apoptosis (8).

Another class of gene alterations that contribute to the etiology of CRC is those that lead to deficiencies in DNA mismatch repair. This is thought to contribute to the growth of cancer cells through the mutation and inactivation of the genes encoding type II TGFβ receptor and insulin-like growth factor (IGF) II receptor. In HNPCC, deficiencies in MMR arise from germline mutation of the genes MSH2, MLH1, PMS1, or PMS2 while these genes are inactivated by an epigenetic mechanism, methylation of the promoter region of hMLH1in sporadic colorectal tumors (9,10).

Overview of Existing Mouse Models (For a more detailed discussion see Murine Intestinal Cancer)

CRC is a heterogeneous disorder that occurs in sporadic and inherited forms and includes a variety of morphological variants. It is characterized by multiple genetic alterations that are associated with different histopathologic features and stages in tumor progression. The complexity of CRC etiology is reflected in the variety of genetically manipulated mice that have been created to model different genetic, morphologic and clinical features of human CRC.

The first - and still the most widely used genetically engineered mouse (GEM)- ApcMin/+, was identified by screening the offspring of mice treated with the mutagen, ethylnitrosourea (ENU). Most new GEM models of gastrointestinal cancer are created using transgenic or knockout technologies to modify specific candidate genes. In addition to GEMs, intestinal neoplasia induced by exposure to the carcinogen azoxymethane (AOM) is another commonly used rodent model that predates GEM.

Mouse models have been generated with mutations in genes representing the major molecular pathways known to play a role in CRC initiation and progression (i.e. the Apc/Wnt pathway, the TGFβ pathway, and MMR genes), as well as other genes with altered expression in gastrointestinal tumors (e.g. Cdx2, n-cadherin and Muc2). Immune-deficient mouse lines that model inflammatory bowel disease are also available. In addition, intestinal neoplasias have been observed in some GEM that were developed in other contexts. The different murine models display considerable diversity with respect to disease severity, location of tumors within the gastrointestinal tract, gross tumor morphology and tumor histology. This phenotypic diversity provides models to study many characteristics of the human disease.