Colorectal cancer is undoubtedly one of the most genetically studied malignancies. In the past 20 years, substantial progress has been made in understanding the genetic and molecular pathogenesis of this cancer. It is instructive that the genes involved in the inherited colon cancer syndromes are the same genes involved in the etiology of most colorectal cancers. Familial adenomatous polyposis (FAP) and hereditatry nonpolyposis colorectal cancer (HNPCC or Lynch syndrome) arise from mutations of the adenomatous polyposis coli (APC) gene and mismatch repair genes, respectively. Approximately 80% of sporadic colorectal cancers begin with APC mutations, while another 15% have a mismatch repair mutation occurring early in their pathogenesis. Additional mutations in different genes accumulate as polyps progress in both size and histopathology to colon cancer.
The genes mutated in tumors harboring APC mutations are said to be a part of the chromosomal instability pathway. This name is given because loss-of-heterozygosity changes are frequently found in the DNA of tumor tissue. Involved genes include K-ras and p53 among others. Tumors with mismatch repair mutations almost always exhibit microsatellite instability (MSI). Additional mutations in such tumors include TGF-beta, BAX, and others, and this mutation pathway is appropriately called the MSI pathway.
This review, which may not be apparent from the title, exclusively addresses the inherited syndromes of colorectal cancer, including their genetic etiologies, phenotypes, and relevant approaches to genetic testing and management. FAP and HNPCC are covered in most detail with appropriate attention given to the hamartomatous polyposis conditions, juvenile polyposis, and Peutz-Jeghers syndrome. The review is reasonably complete, very instructive, and of substantial interest to the practicing oncology specialist who must frequently deal with such syndromes and their diagnosis. The authors raise a number of issues in the review that are worthy of further comment.
‘Flat Adenoma’ Syndrome
The authors mention that attenuated FAP is also known as “the hereditary flat adenoma syndrome.” This connotation was derived from the observation of multiple small adenomas in the proximal colon of several families with an inherited predisposition to colorectal cancer. After it was determined that attenuated FAP arose from APC mutations, these “flat adenoma” families were found to actually be attenuated FAP families. Furthermore, the small adenomas in such families were not histologically consistent with the “flat adenoma” originally described by Morrison at the St. Marks Hospital in London, but instead were typical small adenomas. Thus, it would seem appropriate to discontinue reference to the hereditary flat adenoma syndrome in favor of attenuated FAP.
The authors note that FAP is a more commonly used term than Gardner syndrome. The designation Gardner syndrome came from descriptions by Eldon Gardner, PHD, of a family that exhibited colonic adenomatous polyposis typical of FAP, but also certain benign extraintestinal growths including osteoma and certain soft-tissue tumors, epidermoid cysts, and fibroma. It was not known at the time whether FAP and Gardner syndrome were genetically related.
Over time, additional extracolonic growths were added to the description of Gardner syndrome, including supernumerary teeth, desmoid tumors, adrenal adenomas, jaw bone opacities, odontomas, congenital hypertrophy of the retinal pigment epithelium (CHRPE), gastric fundic gland polyps, and duodenal adenoma. It became a semantic problem to decide whether only the syndrome seen in families with no extracolonic lesions should be called FAP, while all others should be called Gardner syndrome. Various combinations of lesions were evaluated in an attempt to determine if certain groupings could reliably separate the conditions.
Shortly after APC gene mutations were identified as the cause of FAP, it was determined that mutations in the same gene caused Gardner syndrome. Furthermore, identical lesions in that gene could give rise to both FAP and Gardner syndrome phenotypes. It seemed as though Gardner syndrome would quickly become relegated to the category of historical terms.
Then a number of “genotype-phenotype” correlations were made for lesions related to Gardner syndrome. For example, CHRPE lesions were found only when the APC mutations were between exon 9 and the middle portion of exon 15, and osteomas, fibromas, and desmoids were more common with mutations in the distal areas of the gene. Continued work has revealed no clear-cut phenotype or genotype that consistently separates FAP and Gardner syndrome. Instead, the extracolonic findings individually relate to some degree to the mutation location in the APC gene. Nonetheless, the term Gardner syndrome persists and is often applied to families when extraintestinal features, particularly osteoma and soft-tissue growths, are a predominant part of the phenotype.
Fundic Gland Polyps
It is now generally agreed that gastric cancer, although unusual in FAP (about 0.5% lifetime risk), can indeed arise from fundic gland polyps . These polyps likely develop dysplastic changes, sometimes leading to cancer. They can be a difficult management issue, as up to 30% of FAP patients diagnosed with the polyps exhibit some dysplastic change in them. Total gastrectomy seems to be the only sure treatment. In view of the morbidity of this procedure, however, it would seem inappropriate for most cases, considering the low risk of gastric cancer in FAP overall. On the other hand, the appearance of severe dysplasia should alert the clinician to consider intervention, possibly surgery.
Fundic gland polyps seem to occur in attenuated FAP with the same prevalence and phenotype as in typical FAP, although systematic study is needed. Interestingly, similar cases have been described in MYH-associated polyposis.
APC Mutation Detection
The authors refer to a frequently cited estimate of 90% to 95% as the deleterious mutation detection rate in the index case of families with clinically typical FAP. Recent work suggests that the rate is actually only 50% to 70%.[4,5] A few cases in which no APC mutation is found will exhibit pathogenic MYH gene mutations, but most remain unexplained. As the authors stress, once a deleterious mutation is found in the index case, other family members can be tested with mutation-specific testing with nearly 100% accuracy.
Lynch Syndrome and Genetic Testing
The authors carefully explain how the clinician determines who should undergo genetic testing for Lynch syndrome and how to go about that testing. Several points are worth emphasizing. First, MSI and immunohistochemistry (IHC) testing of tumor tissue are not only of use in determining who should have “mutation finding” genetic testing, but they are also helpful-particularly in combination-in ruling out Lynch syndrome. Almost all colorectal cancers associated with HNPCC will exhibit MSI, with the exception of some cases related to MSH6 mutations. Thus, if a colon cancer does not exhibit MSI and there is expression of all mismatch repair genes by IHC, Lynch syndrome is effectively ruled out. As this approach remains imperfect, however, genetic testing probably should still be done in families meeting Amsterdam criteria.
Colonic adenomas do not reliably exhibit MSI in Lynch syndrome, although “advance adenomas”-ie, those that are larger than 1 cm in diameter or exhibit villous or advanced histology-often do. Other HNPCC cancers are likewise inconsistent in their expression of MSI.
A second point concerns the clinical implications of a family meeting Amsterdam criteria, but with no mutations found by genetic testing, negative MSI results, and unrevealing IHC. Such families are often still thought to exhibit Lynch syndrome because they meet the Amsterdam criteria. Three groups (including ours, and one quoted by the authors) have now demonstrated that such families do not develop an excess of the extracolonic cancers found in HNPCC. Even the colon cancer phenotype, in terms of age of onset and risk to relatives, is less severe than that seen in Lynch syndrome families where a mutation has been found.
Third, as also mentioned in the review, the pathogenesis of up to 30% of colon cancers appears to include inheritance.[7,8] A number of groups are searching for the susceptibility genes in this setting-genes that appear to be less penetrant and result in a less severe phenotype than that observed in Lynch syndrome. Such families probably should undergo less aggressive screening than those who are genetically defined as having HNPCC, but more aggressive screening than the average-risk population.
The presently recommended screening strategy for persons in a family with multiple cases of colon cancer or a case diagnosed under the age of 50 years, but no apparent Lynch syndrome, involves colonoscopy starting at age 40 (or at 10 years younger than the earliest case in the family), which should be repeated every 5 years. These recommendations would also seem reasonable for Amsterdam criteria-positive families who do not appear to have Lynch syndrome, except that there should be the option to repeat colonoscopy every 3 years. The shortened interval is suggested because of the known propensity of precancerous lesions to progress much faster in Lynch syndrome and because there remains some possibility of that diagnosis despite the negative genetic testing. After individuals in such a family have been screened for a number of years, the screening might either be relaxed or made more aggressive, depending on follow-up findings.
Hamartomatous Polyposis Syndromes
These syndromes, once thought to be benign syndromes of intestinal hamartomatous polyps, are now known to carry a rather extreme malignancy risk, both in the gastrointestinal tract and in extraintestinal areas . It is questionable whether genetic testing would be helpful in Peutz-Jeghers syndrome, as the typical perioral pigmentation is almost universally present in those with this condition. Also, the polyps are histologically specific for the syndrome. Sporadic Peutz-Jeghers polyps are found but only rarely.
Juvenile polyposis syndrome is quite different. Although juvenile polyps are histologically characteristic, they are found in approximately 1% to 3% of children, sporadically in adults, and in several inherited syndromes. These syndromes include juvenile polyposis syndrome, Cowden syndrome, and Bannayan-Ruvalcaba-Riley syndrome. The first syndrome arises from mutations of SMAD4 and BMPR1A, as described in the review, while the second two arise from mutations of the PTEN gene. Thus, genetic testing is likely to be quite helpful in distinguishing these diseases.
The paper by Jeter and colleagues summarizes an important emerging area of medicine. Genetic testing is now available for each of the syndromes of inherited colon cancer, and screening strategies can effectively prevent cancer in affected patients and families. Thus, it is inherently important for physicians to recognize these syndromes, be familiar with the related cancer risks, appropriately apply genetic testing, and effectively institute established screening guidelines to prevent or detect the related cancers early in their course. To this end, the review represents an important source of information for physicians.