br among patients with type I EC
among patients with type I EC. In patients with type II EC, there were three mutations in MSH6, one in MSH2, and one in PMS2. Two patients with clear cell histology, and one patient with serous histology had an MSH6 mutation. The PMS2 mutation occurred in a patient with serous histology, and the MSH2 mutation occurred with a grade 3 endometrioid tumor. There were no MLH1 mutations identified in this study.
The most commonly-mutated cancer predisposition 1642300-52-4 were CHEK2 (n = 8), MSH6 (n = 12), and ATM (n = 9). There were no differ-ences in incidence of these mutations when stratified for type I vs. type
Frequency of cancer predisposition gene mutations in specific EC subtypes.
No mutations were observed in MLH1, PALB2, or RAD51D. a Does not include CHEK2 p.Ile157Thr.
b Does not include variants in PMS2 pseudogene region or c.736_741del6ins11.
c All USC cases and mutations are included in the Type II frequencies.
Germline mutations in BRCA and LS genes were relatively rare in this cohort of unselected patients with EC. While the overall rate of BRCA1 mutations was similar to that in the general population, the rate of BRCA1 mutations eight times higher among patients with type II com-pared to type I EC. However, this difference was not statistically signifi-cant and was based on only four total mutations among the two groups. While our data suggest a potential association between type II EC and BRCA1 mutations, a larger cohort will be needed to confirm this finding. It is difficult to recommend changes in clinical management of patients with BRCA1 mutations based on these data. There was no difference in the incidence of BRCA2 mutations between the two groups, and no BRCA2 mutations were detected among patients with type II EC.
MSH6 mutations were significantly associated with EC overall and with both EC subtypes. However, associations between other LS gene mutations and EC were not informative due to limited sample size. We reliably ruled out CHEK2 and ATM as EC predisposition mutations, as these variants were not associated with EC despite a sufficient num-ber of mutations in our cohort. Additionally, we identified BLM as a po-tential EC predisposition gene, consistent with previous reports . Interestingly, about 4–5% of all patients with endometrial cancer were found to have a mutation when panel testing for all cancer
predisposition genes was performed. While this finding does not clarify the role of prophylactic hysterectomy in BRCA- or cancer predisposition gene mutation-positive patients, it may have implications for genetic screening for patients with EC.
Other studies have reported similar rates of BRCA mutations in pa-tients with USC. Pennington et al. performed germline panel testing on 151 unselected patients with USC and found a 2% rate of BRCA1/2 mutations. While this is higher than that in the general population , the authors concluded that the absolute risk of USC in BRCA-positive patients was probably low, given the rarity of this subtype . Another study of 56 patients with USC found no germline BRCA muta-tions. However, this study tested only a limited number of mutational variants, and a personal (11%) or family history (29%) of breast cancer was found in a significant proportion of patients, suggesting that some mutations may have been missed due to non-comprehensive testing .
In contrast, some studies have found a higher proportion of BRCA mutations among patients with USC. One study found a 27% rate of BRCA mutations in a cohort of 22 Jewish patients with USC. This rate is likely to be confounded by the increased frequency of BRCA mutations in Ashkenazi Jewish populations and has not been replicated in studies of unselected patients . Most genetic studies in USC are also con-founded by the prevalence of tamoxifen exposure in BRCA-positive co-horts. For instance, another investigation found an increased risk of EC in the overall cohort of BRCA-positive patients (standardized incidence ratio (SIR) 1.91, 95% CI 1.06–3.19, p = 0.03). However, the risk was greatest in the subset of BRCA-positive women with tamoxifen exposure (SIR = 4.14, 95% CI 1.92–7.87). Among BRCA-positive women who did not receive tamoxifen, the risk of USC was no higher than the standard-ized incidence found in age and region-matched controls (SIR 1.67, 95% CI 0.81–3.07) . While most authors have specifically examined the association between BRCA1/2 and USC, a recent prospective cohort study found an association when both serous and serous-like uterine cancers were included. These serous-like cancers included undifferenti-ated EC and carcinosarcoma, encompassing the majority of type II EC subtypes . For this reason, we examined the association of BRCA1/2 and other cancer predisposition gene mutations with both type II EC and USC separately.