Performance of tumor testing for Lynch syndrome identification in patients with colorectal cancer: A retrospective single-center study.

Abstract

To investigate the performance of tumor testing approaches in the identification of Lynch syndrome (LS) in a single-center cohort of people with colorectal cancer (CRC). A retrospective analysis of data stored in a dedicated database was carried out to identify patients with CRC suspected for LS who were referred to Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy, between 1999 and 2014. The sensitivity and specificity of immunohistochemistry (IHC) for mismatch repair (MMR) proteins and microsatellite instability (MSI) analysis (alone or combined) were calculated with respect to the presence of causative MMR germline variants. A total of 683 patients with CRC suspected for LS were identified. IHC results of MMR protein analysis and MSI were assessed in 593 and 525 CRCs, respectively, while germline analysis was performed in 418 patients based on the IHC or MSI test result and/or clinical features. Univariate and multivariate analysis revealed a significant correlation of pathogenic MMR germline variants with all clinicopathologic features including Amsterdam criteria, presence of endometrial cancer, CRC site, age at onset, stage, and grade. The highest odds ratio values were observed for IHC and MSI (17.1 and 8.8, respectively). The receiver operating characteristic curve and area under the curve values demonstrated that IHC alone or combined with other clinicopathologic parameters was an excellent test for LS identification. This study confirms the effectiveness of tumor testing to identify LS among patients with CRC. Although IHC and MSI analysis were similarly effective, IHC could be a better strategy for LS identification as it is less expensive and more feasible.