PMS2 is the most frequently mutated Lynch syndrome (LS) gene in women with endometrial cancer (EC): what is the role that low penetrance LS genes play in EC?

Objectives: To determine the frequency and spectrum of LS pathogenic variants in EC patients using upfront germline testing for cancer susceptibility. Secondary objectives included determining the sensitivity and specificity of current LS screening. Methods: Women diagnosed with EC were prospectively enrolled at 9 institutions. Clinical germline testing (panel testing for common cancer susceptibility genes) was performed for 840 unselected EC patients between 10/1/2017 and 10/1/2020. Mismatch repair (MMR) IHC, MLH1 methylation status and relevant clinicopathologic data were abstracted from patients’ records. Detailed family history data were obtained for all subjects. Results: Germline testing revealed 26 of 840 EC patients (3.1%) had a pathogenic/likely pathogenic MMR gene mutation. Mutations in PMS2 were most frequent (10/26, 38%), followed by MSH6 (8/26, 31%), MSH2 (6/26, 23%), and MLH1 (2/26, 8%). Mean BMI for PMS2 carriers was greater than other mutation carriers (31.3 vs 27.8). Age at diagnosis was also higher for PMS2 carriers (59 vs 52). IHC combined with MLH1 methylation correctly predicted LS for all MLH1 and MSH2 mutation carriers. For MSH6 and PMS2, IHC and germline mutation status concordance was lower. Only 4/9 (44%) PMS2 carriers had isolated PMS2 loss. IHC was not completed for one PMS2 carrier (insufficient tumor). IHC correctly predicted MSH6 mutations in 7/8 (88%) carriers. Family history did not predict LS for PMS2 carriers: using PREMM5 (LS prediction model), all had low ( Conclusions: This is the largest cohort of prospectively enrolled EC patients to undergo germline genetic testing in an unselected manner, irrespective of patient age, family history or tumor results. It shows that a majority of LS mutations in EC are in the low-penetrance genes, MSH6 and PMS2. PMS2 mutations were the most common, 10/840 (1.2%). Universal screening with IHC would have missed three PMS2 carriers and one MSH6 carrier. Accurate identification of patients with MMR defects is critical not only for identification of LS, but potentially in consideration of treatment options. In the past, ascertainment of LS relied largely on a strong family cancer history, which is most often seen with MLH1 and MSH2 mutations. For all PMS2 carriers in our study, family history was not predictive of LS and the proband's EC diagnosis was the sentinel event leading to a LS diagnosis in the family. Our study may highlight an evolving LS phenotype in which gene/environment interactions are changing. The observation that EC patients with PMS2 mutations have higher BMIs leads us to hypothesize the existence of an obesity-mediated pathway important in PMS2-associated EC. It may be that in the general population, PMS2 mutation carriers face higher risks for EC than colon cancer.