Sixteen rare sequence variants of the hMLH1 and hMSH2 genes found in a cohort of 254 suspected HNPCC (hereditary non-polyposis colorectal cancer) patients: mutations or polymorphisms?

5-8% of all colorectal cancer cases are assumed to be due to germline mutations in DNA mismatch repair genes. Mutation analysis of these genes in affected families enables one to identify subjects with an inborn susceptibility to colorectal tumorogenesis and to offer presymptomatic testing to family members at risk, provided that the mutation detected is a truncating one or a missense mutation that has either been judged as disease causing in other families or segregates with the disease and results in a microsatellite instability of the corresponding tumor. Segregation analysis within the family or microsatellite analysis of the tumor is, however, not always possible. In these cases, assessment of the relevance of the sequence variation identified is very difficult. On the other hand, discrimination between inactivating mutations and innocuous sequence polymorphisms is of extreme importance for clinical and genetic counseling of affected families. Here we report 16 rare sequence variants of the hMLH1 and hMSH2 genes including 11 different missense variations found in a cohort of 254 suspected HNPCC patients. We provide evidence, that missense variations in hMLH1 do not necessarily result in microsatellite instability of the corresponding tumor DNA. These patients would have been missed had one followed the recommendations of using only microsatellite analysis for the selection of patients at high risk of hereditary non-polyposis colorectal cancer for mutation analysis.