Abstract PR-2: Investigation of mismatch repair deficiency in ovarian cancers

Abstract

Abstract PR-2 Background Defects in the mismatch repair (MMR) pathway are believed to be etiologically important in a reasonable proportion of epithelial ovarian cancers. Various laboratory methods can be used to identify MMR-deficient ovarian cancers, including microsatellite instability (MSI) analysis and immunohistochemistry (IHC). Specifically, in colorectal cancers (CRC), tumor characteristics previously shown to be useful in identifying MMR-deficient tumors include MSI-high (MSI-H) phenotype and loss of MMR protein expression on IHC analyses. Similar findings have been demonstrated in ovarian cancer, however the sample sizes of those studies have been limited. Objectives/Methods We sought to investigate tumor characteristics including: (1) MSI-high status, (2) Expression of 3 MMR proteins (MLH1, MSH2, and MSH6) through IHC studies, and (3) The concordance between MSI and IHC results. The tumors analyzed came from women between the ages of 18 and 80 years with newly-diagnosed epithelial ovarian cancer ascertained between December 13, 2000 and September 30, 2003in a population-based study covering a 2-county region of west central Florida. MSI-H was defined by instability in 2 or more of the 5 NCI-standardized microsatellite markers (ie: BAT25, BAT26, D2S123, D17D250, D5S346). Tissue microarrays were created to evaluate the loss of MMR protein expression on IHC based on a scoring system of 0-9 to evaluate both stain and intensity, and take the product of these 2 numbers; loss of expression of protein expression was defined as <3). These interim results are part of a larger multicenter study, currently underway, to investigate the clinical relevance of the MMR pathway in a population-based sample of ovarian cancers, including MSI analysis, IHC studies, MLH1 promoter hypermethylation, and germline mutation analysis of the MMR genes. Results Of the 232 who agreed to participation, tumor samples were collected on 219 (94%). The distribution of by race and ethnicity of study participants was similar to the distribution of cancer cases in the catchment area, with ~90% of participants being non-Hispanic Whites. The distribution of histologic subtypes, stage and median age at diagnosis was also similar to that seen in the general population. Of the 219 cases, all had IHC analysis and 201 had MSI analysis. The frequency of MSI-H was 29%. The frequency of loss of expression on IHC was 17%. The numbers of tumors which were concordant for MSI and IHC results was 141 (including 16 cases with MSI-H and loss of MMR protein expression and 125 cases with no MSI-H and presence of MMR protein expression). Conversely, there were 60 discordant cases (including 42 cases which were MSI-H, but MMR protein expression was present; and 18 cases which were not MSI-H, with loss of MMR protein expression). Conclusions In contrast to previous studies of CRC, our results indicate limited concordance between MSI-H status and loss of MMR protein expression. Thus, MSI analysis and IHC studies for MMR protein expression may be of limited utility in the identification of ovarian tumors with MMR deficiency. Citation Information: Cancer Prev Res 2008;1(7 Suppl):PR-2.