Oncology
To date, Lynch syndrome research has primarily been performed on cohorts that meet the Amsterdam criteria or Bethesda guidelines and population-based colorectal cancer (CRC) and endometrial cancer (EC) cohorts, possibly biasing results. To further evaluate the full phenotypic spectrum of mismatch repair (MMR) gene mutations, we performed a retrospective phenotype analysis of MMR and EPCAM mutation carriers ascertained through multi-gene panel testing (MGPT).
25,502 patients underwent MGPT including the MMR and EPCAM genes between March 2012 and March 2015. Clinical histories of those who tested positive for an MMR gene mutation (MMR+) were retrospectively reviewed and pairwise comparisons were performed for the various MMR genes using Fisher’s exact test and logistic regression multivariate analysis.
Overall, 471 MMR+ patients (1.8%) were identified. Patients with a second mutation in an MMR or other cancer gene (n=16) were excluded from subsequent analyses. Of the remaining 445, 61 (13.7%) had breast cancer (BC) only (no CRC or EC) and 42 (9.4%) had ovarian cancer (OC) only (no CRC or EC). Nine of these had both BC and OC, seven of which had MSH6 or PMS2 mutations. When comparing those with BC only to those with CRC but not BC, MSH6 and PMS2 mutations were more frequent than MLH1 and MSH2 mutations after controlling for age, ethnicity, and gender (p = 0.002). MSH6 mutations were more frequent among OC cases than MLH1, MSH2, and PMS2 mutations after controlling for age, ethnicity, and gender (p = 0.005, 0.025, and 0.002, respectively). In this cohort, 22% met only the National Comprehensive Cancer Network (NCCN) hereditary breast and ovarian cancer (HBOC) testing criteria and 7% met neither HBOC or Lynch syndrome testing criteria. MSH6 and PMS2 mutations were more frequent than MLH1 and MSH2 mutations among cases that met NCCN HBOC testing criteria, but did not meet NCCN Lynch syndrome testing criteria (p = 0.001).
Results from this study support the association of an HBOC phenotype with MSH6 and PMS2 mutations and highlight the need for further investigation of BC and OC risks for these mutation carriers. These data also support the inclusion of Lynch syndrome in the differential diagnosis for patients with an HBOC phenotype and highlight the limitations of current testing criteria in identifying these cases.