Phenotypic Characteristics Associated With the APC Gene I1307K Mutation in Ashkenazi Jewish Patients With Colorectal Polyps FREE

Colorectal cancer (CRC) is the second leading cause of cancer death in North America. Approximately 15% to 20% of CRC occurs in familial aggregations.^{1 - 2} In 1997, a particular mutation in the adenomatous polyposis coli (APC) gene, an isoleucine-to-lysine variant at codon 1307 of the APC gene (I1307K), was reported to confer susceptibility to apparently sporadic CRC.³ The I1307K mutation was carried by 6.1% of unselected Ashkenazi Jewish individuals and 28% (7 of 25) of Jewish patients with CRC and a family history of CRC.³ In aggregate, subsequent studies^{4 - 14} have generally supported a role for I1307K in the pathogenesis of CRC in the Ashkenazi Jewish population but not in other ethnic groups or in Ashkenazi patients with other forms of cancer. Some reports have suggested that the mutation may be associated preferentially with early-onset CRC^{3 ,14} and certain distinguishing features, such as multiple colorectal adenomas.¹⁴ The prevalence of the mutation, specifically in individuals with sporadic colorectal polyps, has not yet been reported.

We sought to determine the I1307K carrier rate in Ashkenazi Jewish patients with a history of colorectal polyps but no CRC and to compare clinical characteristics and family history of carriers vs noncarriers.

The study involved 5 sites in Boston, Mass, including Brigham and Women's Hospital, Beth Israel Deaconess Medical Center, Dana-Farber Cancer Institute, Harvard Vanguard Medical Associates, and the Massachusetts General Hospital. Human subjects committees at each center approved the study protocol. Study staff reviewed colonoscopy, pathology, and clinic records at each site to identify subjects with at least 1 large-bowel polyp diagnosed from January 1, 1992, through January 31, 1999. Patient ethnicity was obtained from hospital or clinic registration data; patients with Jewish or undeclared religious affiliation were eligible for the study. Individuals with a history of familial polyposis or CRC were excluded.

With the permission of the patient's primary physician or gastroenterologist, eligible patients were mailed a study packet that included a letter describing the purpose of the study, a questionnaire, an informed consent form that included permission for medical record documentation and genetic analysis of the I1307K mutation, and a kit for collection of cheek cells. Interested patients returned a signed informed consent form, the questionnaire, and the cheek brushes via overnight mail in a prepaid envelope.

Ethnicity was confirmed by patient self-report. Specifically, individuals were asked, "Do you have any Ashkenazi (European-American) Jewish heritage?" Personal and family history of cancer and colorectal polyps were obtained via self-administered questionnaire. Medical records were obtained to confirm all colorectal polyp and cancer diagnoses in index patients; pathological confirmation was obtained for 91% of participants with colorectal adenomas. Pathology and endoscopy reports were reviewed for extraction of phenotypic characteristics of polyps, including size, number, and location.

DNA was isolated from blood or buccal swab samples using a QIAamp DNA Mini kit (Qiagen Inc, Valencia, Calif). Isolated DNA was amplified using primers specific for exon 15 of the APC gene, including the region that encompassed the published I1307K polymorphism. Amplified material was dot blotted onto Biotrans nylon membrane (ICN, Irvine, Calif) using a GIBCO/BRL 96-well dot blot apparatus (Life Technologies, Rockville, Md). Blotted membranes were hybridized with probes specific for either the normal APC gene sequence or the I1307K polymorphism.¹⁵ Results were determined by comparing the hybridization signal of positive controls to patient samples in relation to background.

Specimens and genetic analysis results were anonymized to preserve confidentiality (genetic test results were not linked to patient identifiers); thus, there was no disclosure of genetic test results to participants or their physicians.

Between May 1998 and May 1999, we mailed a total of 836 study packets to potentially eligible patients. A total of 272 individuals enrolled in the study, for a participation rate of 32%. Participation rates were similar across sites (ranging from 28% to 35%) and did not vary by type of practice or referring physician. Forty-one individuals were excluded from the study: 5 because of insufficient data, 19 because of a history of CRC; and 17 because of polymerase chain reaction amplification failure.

Of the remaining 231 participants, 27 were not Ashkenazi Jewish and 21 were of unknown ethnicity. There were 20 participants with hyperplastic polyps and 2 with inflammatory polyps. The 161 Ashkenazi Jewish patients with adenomatous polyps are the main focus of this report. A total of 153 (95%) were Ashkenazi Jewish on both maternal and paternal sides of the family; 8 (5%) were Ashkenazi only on 1 side of the family. There were 99 men (61%) and 62 women (39%). The mean age at first adenoma diagnosis was 63 years. Fifty-two individuals (32%) reported a first-degree relative with CRC, and 22 individuals (14%) had a first-degree relative with colorectal polyps.

To compare the characteristics of participants and those individuals who chose not to enroll, we randomly sampled 83 records (14%) of nonresponders. Comparison of demographic characteristics and findings on colonoscopy revealed no significant difference between the 2 groups except that the mean age of participants was 4 years younger than nonresponders.

Carrier rates were determined for several risk groups (Table 1). Among Ashkenazi Jewish patients with adenomas, 22 (14%) of 161 carried the I1307K mutation, whereas the alteration was found in only 1 (5%) of 20 Ashkenazi Jewish individuals with hyperplastic polyps. The 14% mutation rate in adenomatous polyps was significantly higher than reported mutation rates of 6.1% to 7.0% in 2 large series of Ashkenazi Jewish controls without CRC^{3 ,9} (2-sided P by Pearson χ² = .005). The results were not significantly different when the analysis was performed with the inclusion of data from the 19 patients who had been excluded because of a history of CRC.

Analysis of the characteristics of adenomatous polyps in Ashkenazi Jewish patients with colorectal polyps is presented in Table 2. Comparison of the mean age at first diagnosis of adenomas, the mean number of adenomas per colonoscopy, the mean size of adenomas, and the location of adenomas revealed no significant differences between carriers and noncarriers. Furthermore, there were no significant differences in the frequency of CRC, non-CRCs, or colorectal polyps in first-degree relatives between carriers and noncarriers (Table 3).

Cancer-predisposition genes have classically been associated with striking phenotypic features, such as early age at onset of neoplasia, multiple affected family members, rare tumors, and the presence of multiple tumors in the same individual. Therefore, we hypothesized that patients with adenomatous polyps who carried the I1307K mutation would exhibit unusual clinical characteristics and be more likely to have a positive family history of colorectal polyps or cancer than noncarriers. The I1307K carrier rate in Ashkenazi Jewish patients with adenomatous polyps (but not hyperplastic polyps) was increased compared with the population prevalence of the mutation and similar to the rate found in Ashkenazi Jewish patients with CRC,^{3 ,14} confirming its role early in the pathogenesis of colorectal neoplasia. However, in contrast to previous reports and our expectations, we found no differences in family history of colorectal neoplasia or any phenotypic features of adenomas in terms of age of onset and location, size, and number of polyps in I1307K mutation carriers vs noncarriers.

Multiple issues may affect patients' willingness to participate in genetics studies, including fear of confidentiality of individual results and stigmatization of the Jewish community,¹⁶ and these factors may have led to the suboptimal participation rate observed in our study. We attempted to address this issue with an analysis of nonresponders, which revealed similar demographic and polyp characteristics (except for a small difference in age) between eligible patients who chose and did not choose to participate. Ascertainment bias may still have affected enrollment, with a bias toward those who were undergoing endoscopic surveillance or had a family history of the disease. However, such a bias should not affect comparisons between carriers and noncarriers, since they would not be expected to differ between the mutation carrier and noncarrier groups.

When first identified, I1307K was hailed as the cause of a significant proportion of CRC in Jewish patients, and clinical recommendations were made advocating genetic testing for the mutation and intensified surveillance for carriers. Our detailed comparison of mutation carriers and noncarriers indicates that the I1307K mutation is distinctly different than classic high-penetrance cancer susceptibility genes, which are typically rare, segregate with disease, and are sufficient on their own to substantially increase risk with only minimal impact by environmental factors.¹⁷ With completion of the elucidation of the human genetic code, scientists will continue to identify other polymorphisms such as I1307K that are common and increase risk of cancer and other common diseases but can be strongly influenced by association with other genetic or environmental factors. Ultimately, the best estimation of risk is likely to be made by assessment of the combination and interaction of multiple low-penetrance mutations, behavioral factors, and personal medical history. In the meantime, it is important that patients who elect to undergo testing for I1307K appreciate that although the test is relatively cheap and easy to obtain, the results need to be interpreted in the context of a variety of other factors. A positive result does not mean the inevitability of cancer, and risk may be significantly affected by screening practices and behavioral factors, such as physical activity and diet. Equally important, however, is that a negative result does not imply protection from CRC, since there are surely other low-penetrance genes affecting CRC risk that have not yet been identified. The study of these complex interactions and clinical implementation of testing of low-penetrance genes clearly pose novel challenges for genetic epidemiologists, physicians, and patients.