Correlation between clinical-pathological parameters and family history to detect mutations in MLH1, MSH2, and MSH6.

391 Background: The Amsterdam I/II and Bethesda criteria are used to select individuals for the study of mutations in hereditary nonpolyposis colon cancer (HNPCC). The aim of this study was to analyze whether specific clinical features and family history of individuals suspected of HNPCC were correlated with the detection of germline mutations in MLH1, MSH2 and MSH6. Methods: Between 2005-2008, the study of germline mutations in one or more of the genes MLH1, MSH2 or MSH6 was carried out on 124 individuals who fullfield the Amsterdam I/II criteria or Bethesda criteria with microsatellite instability (MSI) or loss of expression by immunohistochemistry (IHC) of any of the repair proteins of MMR genes. Subsequently, we applied univariate and multivariate analyses including clinicopathological characteristics and family history to see if they were related to the presence of germline mutations. The characteristics were: age, sex, age at diagnosis, whether they fulfilled the Amsterdam or Bethesda criteria, diagnosis of cancer, tumor type, presence of multiple tumors and age of first diagnosis of cancer in the family. Results: Out of 124 patients studied, 29 gene mutations were detected (detection rate 25%). Of all the parameters studied, only endometrial cancer increased the risk of mutation 7.3 times (confidence interval [CI] 95%, from 1.83 to 29.2) for colorectal cancer (p = 0.005). 20.7% of mutations in MMR genes were explained by the type of cancer (R square Nagelkerke = 0.207). In the multivariate analysis none of the variables predicted the presence of mutation. Conclusions: The most important clinical feature to predict the presence of a mutation in the genes MLH1, MSH2 and/or MSH6 in families with HNPCC is the diagnosis of endometrial cancer (univariate analysis). No significant financial relationships to disclose.