Molecular basis for risk factors
BMJ 2000; 321 doi: https://doi.org/10.1136/bmj.321.7265.886 (Published 07 October 2000) Cite this as: BMJ 2000;321:886
- Robert G Hardy,
- Stephen J Meltzer,
- Janusz A Jankowski
Evidence for the molecular basis of colorectal cancer comes from genetic analysis of tissues either from patients with a family history of the disease or from patients with sporadic adenomatous colorectal polyps or extensive ulcerative colitis. The traditional view is that background rates of genetic mutation, combined with several rounds of clonal expansion, are necessary for a tumour to develop. It has recently been argued, however, that inherent genetic instability not only is necessary but may also be sufficient for cancer to develop.
Proposed adenoma to carcinoma sequence in colorectal cancer. Adenomatous polyposis coli (APC) gene mutations and hypermethylation occur early, followed by k ras mutations. Deleted in colon cancer (DCC) and p53 gene mutations occur later in the sequence, although the exact order may vary
Sporadic colorectal adenomas
More than 70% of colorectal cancers develop from sporadic adenomatous polyps, and postmortem studies have shown the incidence of adenomas to be 30-40% in Western populations. Polyps are asymptomatic in most cases and are often multiple. Flat adenomas, which are more difficult to detect at endoscopy, account for about 10% of all polyps and may have a higher rate of malignant change or may predispose to a more aggressive cancer phenotype.
Key molecular events in colorectal premalignancy: comparisons between the adenoma carcinoma sequence and ulcerative colitis associated neoplasia
Family history
Recognised familial syndromes account for about 5% of colorectal cancers. The commonest hereditary syndromes are familial adenomatous polyposis and heredity non-polyposis colon cancer. Patients with these syndromes usually have a family history of colorectal cancer presenting at an early age. Attenuated familial adenomatous polyposis, juvenile polyposis syndrome, and Peutz-Jeghers syndrome are rarer, mendelian causes of colorectal cancer. In familial adenomatous polyposis (a mendelian dominant disorder with almost complete penetrance) there is a germline mutation in the tumour suppressor gene for adenomatous polyposis coli (APC) on chromosome 5.
Factors determining risk of malignant transformation within colonic adenomatous polyps
High risk
Large size (especially >1.5 cm)
Sessile or flat
Severe dysplasia
Presence of squamous metaplasia
Villous architecture
Polyposis syndrome (multiple polyps)
Low risk
Small size (especially <1.0 cm)
Pedunculated
Mild dysplasia
No metaplastic areas
Tubular architecture
Single polyp
Heredity non-polyposis colon cancer also shows dominant inheritance, and cancers develop mainly in the proximal colon. Patients with heredity non-polyposis colon cancer show germline mutations in DNA mismatch repair enzymes (which normally remove misincorporated single or multiple nucleotide bases as a result of random errors during recombination or replications). Mutations are particularly demonstrable in DNA with multiple microsatellites (“microsatellite instability”).
The immediate family members of a patient with colorectal cancer will have a twofold to threefold increased risk of the disease
In addition to the well recognised syndromes described above, clusters of colorectal cancer occur in families much more often than would be expected by chance. Postulated reasons for this increased risk include “mild” APC and mismatch repair gene mutations, as well as polymorphisms of genes involved in nutrient or carcinogen metabolism.
Clinical and molecular correlates of familial adenomatous polyposis coli; attenuated familial adenomatous polyposis coli/hereditary flat adenoma syndrome; hereditary non-polyposis colon cancer/Lynch forms of hereditary colorectal cancer; and ulcerative colitis associated neoplasia
FAP=familial adenomatous polyposis coli; AFAP=attenuated familial adenomatous polyposis coli; HFAS=hereditary flat adenoma syndrome; HNPPC=hereditary non-polyposis colon cancer; UCAN=ulcerative colitis associated neoplasia.
Risk from ulcerative colitis
Several studies have indicated that patients with ulcerative colitis have a 2-8.2 relative risk of colorectal cancer compared with the normal population, accounting for about 2% of colorectal cancers. One of the factors influencing an individual's risk is duration of colitis—the cumulative incidence of colorectal cancer is 5% at 15 years and 8-13% at 25 years. The extent of disease is also important: patients with involvement of right and transverse colon are more likely to develop colorectal cancer (the relative risk in these patients is 15 compared with the normal population). Coexisting primary sclerosing cholangitis independently increases the relative risk of ulcerative colitis associated neoplasia (UCAN) by 3-15%. In addition, high grade dysplasia in random rectosigmoid biopsies is associated with an unsuspected cancer at colectomy in 33% of patients.
Factors affecting risk of colorectal cancer in patients with ulcerative colitis
High risk
Long duration of disease (especially >10 years)
Extensive disease
Dysplasia
Presence of primary sclerosing cholangitis
Family history of colorectal cancer
Coexisting adenomatous polyp
Low risk
Short duration of disease (especially <10 years)
Proctitis only
No dysplasia
No primary sclerosing cholangitis
No family history of colorectal cancer
No coexisting adenomatous polyp
Molecular basis of adenoma carcinoma sequence and UCAN
Cancers arising in colitis versus those in adenomas
Important clinical and biological differences exist between the adenoma carcinoma sequence and ulcerative colitis associated neoplasia. Firstly, cancer in ulcerative colitis probably evolves from microscopic dysplasia with or without a mass lesion rather than from adenomas. Secondly, the time interval from the presence of adenoma to progression to carcinoma probably exceeds the interval separating ulcerative colitis associated dysplasia from ulcerative colitis associated neoplasia. Thirdly, patients with a family history of colorectal cancer (but not ulcerative colitis associated neoplasia) and who also have ulcerative colitis are at further increased risk, suggesting additive factors.
Chromosomal and microsatellite instability
Molecular alterations in colorectal cancer can be grouped into two broad categories: chromosomal instability (subdivided into aneuploidy and chromosomal alterations) and microsatellite instability
As a consequence of these two phenomena, other specific genetic events occur at increased frequency
These include inactivation of tumour suppressor genes by deletion or mutation, activation of proto-oncogenes by mutation, and dysregulated expression of diverse molecules, such as the cell to cell adhesion molecule E cadherin and mucin related sialosyl-Tn antigen
Chromosomal instability
Aneuploidy indicates gross losses or gains in chromosomal DNA and is often seen in many human primary tumours and premalignant conditions. It has been shown that aneuploid “fields” tend to populate the epithelium of patients with ulcerative colitis even in histologically benign colitis. These changes may occur initially in some cases by loss of one allele at a chromosomal locus (loss of heterozygosity) and may imply the presence of a tumour suppressor gene at that site. Loss of both alleles at a given locus (homozygous deletion) is an even stronger indicator of the existence of a tumour suppressor gene. Loss of heterozygosity occurs clonally in both the adenoma carcinoma sequence and ulcerative colitis associated neoplasia. Many of these loci are already associated with one or more known candidate tumour suppressor genes. These include 3p21 (β catenin gene), 5q21 (APC gene), 9p (p16 and p15 genes), 13q (retinoblastoma gene), 17p (p53 gene), 17q (BRCA1 gene), 18q (DCC and SMAD4 genes), and less frequently 16q (E cadherin gene).
Downregulation of E cadherin (arrowed) within colonic adenoma. Normal membranous E cadherin staining (brown) is seen in non-dysplastic crypts in the right of the picture
The p53 gene locus is the commonest site demonstrating loss of heterozygosity. p53 is a DNA binding protein transcriptional activator and can arrest the cell cycle in response to DNA damage—hence its title “guardian of the genome.” The effect of normal (wild-type) p53 is antagonised by mutation or by action of the antiapoptotic gene Bcl-2, which is significantly less frequently overexpressed in ulcerative colitis associated neoplasia than in the adenoma carcinoma sequence. Most mutations in p53 cause the protein to become hyperstable and lead to its accumulation in the nucleus.
A second tumour suppressor gene necessary for development of sporadic colorectal cancer is APC, which is inactivated in >80% of early colorectal cancers. Consequently this gene has been termed the “gatekeeper” for adenoma development as adenoma formation requires perturbation of the APC gene's function or that of related proteins such as catenins. An important function of the APC gene is to prevent the accumulation of molecules associated with cancer, such as catenins. Accumulation of catenins can lead to the transcription of the oncogene c-myc, giving a proliferative advantage to the cell. APC mutations occur later and are somewhat less common in ulcerative colitis associated neoplasia (4-27%) than in sporadic colorectal adenomas and carcinomas. Catenins also bind E cadherin, which functions as a tumour suppressor gene in the gastrointestinal tract. It is currently thought that mutated catenins may not bind to APC and thus accumulate.
E cadherin mutations are not commonly associated with the adenoma carcinoma sequence, but loss of heterozygosity and, rarely, missense mutations have been reported in 5% of ulcerative colitis associated neoplasia
Staging and survival of colorectal cancers
Microsatellite instability
A further important category of alteration studied in the adenoma carcinoma sequence and ulcerative colitis associated neoplasia is microsatellite instability. This comprises length alterations of oligonucleotide repeat sequences that occur somatically in human tumours. This mechanism is also responsible for the germline defects found in heredity non-polyposis colon cancer. The incidence of microsatellite instability has been noted to be about 15% for adenomas and 25% for colorectal cancers overall. Microsatellite instability also occurs in patients with ulcerative colitis and is fairly common in premalignant (dysplastic) and malignant lesions (21% and 19% respectively). Indeed it has also been reported in “histologically normal” ulcerative colitis mucosa. It can therefore be considered to be an early event in the adenoma carcinoma sequence and in ulcerative colitis associated neoplasia.
Further reading
Prognosis
The prognosis of colorectal cancer is determined by both pathological and molecular characteristics of the tumour.
Pathology
Pathology has an essential role in the staging of colorectal cancer. There has been a gradual move from using Dukes's classification to using the TNM classification system as this is thought to lead to a more accurate, independent description of the primary tumour and its spread. More advanced disease naturally leads to reduced disease-free interval and survival. Independent factors affecting survival include incomplete resection margins, grade of tumour, and number of lymph nodes involved (particularly apical node metastasis—main node draining a lymphatic segment).
Molecular biology
Reports on correlations between tumour genotype and prognosis are currently incomplete. However, analysis of survival data from patients with sporadic colorectal cancer and from those with colorectal cancer associated with familial adenomatous polyposis and hereditary non-polyposis colon cancer has not shown any reproducible significant differences between these groups. In premalignancy, however, the onset of p53 mutations in histologically normal mucosa in ulcerative colitis suggests that detection of such mutations may be a useful strategy in determining mucosal areas with a high risk of dysplastic transformation.
Acknowledgments
This work was funded by the Cancer Research Campaign and the Medical Research Department of Veteran Affairs. Fiona K Bedford, Robert Allan, Michael Langman, William Doe, and Dion Morton provided useful comments.
Footnotes
-
The ABC of colorectal cancer is edited by D J Kerr, professor at the Institute for Cancer Studies, University of Birmingham; Annie Young, research fellow at the School of Health Sciences, University of Birmingham; and F D Richard Hobbs, professor of primary care and general practice, University of Birmingham. The series will be published as a book by the end of 2000.
-
Robert G Hardy is a Wellcome Trust clinical research fellow in the departments of medicine and surgery, University Hospital, Birmingham; Stephen J Meltzer is a professor of medicine and director, Functional Genomics Laboratory, University of Maryland Greenbarn Cancer Centre, Baltimore, USA; and Janusz A Jankowski is reader and consultant gastroenterologist in the department of medicine, University Hospital, Birmingham, and the Imperial Cancer Research Fund, London.