Clinical HNPCC with no evidence for a role of DNA mismatch repair in cancer susceptibility

1703

Objectives: Hereditary non-polyposis colorectal cancer (HNPCC) is an autosomal dominant cancer susceptibility syndrome caused by an inherited defect in DNA mismatch repair (DMMR). HNPCC-associated tumors exhibit microsatellite repeat instability (MSI) and loss of expression of one or more MMR proteins that is often characteristic of the underlying gene defect. We ascertained a large HNPCC kindred through two endometrial cancer probands (cousins). Tumor MSI and immunohistochemistry (IHC) studies for the probands’ tumors did not provide a clear indication as to which if any of the DMMR genes might be responsible for cancer susceptibility in the family. We therefore sought to determine whether a germline mutation in one of the DMMR genes (MLH1, MSH2, MSH6, and PMS2) was present.

Methods: A detailed family history and review of medical records was undertaken to construct a three generation pedigree. Blood samples were obtained for analysis of germline DNA from 39 family members. Highly informative polymorphic repeats from the MLH1, MSH2, MSH6, and PMS2 loci were genotyped using PCR methods and the co-segregation of markers and disease was assessed. Protein expression in the tumor was evaluated by IHC on all available specimens for MLH1, MSH2, MSH6, and PMS2.

Results: Four gynecologic carcinomas (3 endometrial, 1 ovarian), 3 colon carcinomas, 1 breast carcinoma, 1 prostate cancer, and 10 cases of adenomatous polyps were identified in family members across 3 successive generations. The family met the Amsterdam II criteria for HNPCC. However, it was unusual both with respect to late age-of-onset and location of colon carcinomas. The mean age at diagnosis in the kindred was 64 years (earliest age of diagnosis was ovarian cancer at 44 years), and all three colon carcinomas were left sided. Through marker genotype and haplotype analysis, we were able to exclude linkage among MLH1, MSH2, MSH6, and PMS2 as candidates for the cancer susceptibility gene in this HNPCC family. MSI analysis and IHC analysis for tumors did not suggest a role for DMMR.

Conclusions: We identified a large HNPCC family with no evidence of germline DMMR mutation. This family represents what is likely an important but poorly understood form of autosomal dominant inherited cancer susceptibility. Implications for cancer screening and surveillance, the need for genetic counseling and further molecular and clinical investigation of families like the one we described will be discussed.