Trends In Actionable Mutations For Recurrent Ovarian And Endometrial Cancer At A Single Institution

Objectives: To identify actionable gene mutations in recurrent ovarian and endometrial cancer and to evaluate the effect of these cancer specific mutations on survival outcomes at a single institution. Methods: This was an institutional review board (IRB) approved retrospective chart review of all patients at our institution with recurrent endometrial or ovarian cancer whose tumor underwent Next-Generation Sequencing using Foundation Medicine genomic profiling. Demographic and cancer related variables were extracted including age, cancer type, cancer stage at diagnosis, progression free survival and overall survival. Fisher's exact test was performed to identify specific cancer type associated with a particular mutation and a regression model was used to evaluate the cancer specific mutations’ correlation with age, cancer stage and survival. Results: A total of 117 patients who met criteria were identified. Twenty-six significant mutations were found among three cancer subtypes; endometrioid endometrial adenocarcinoma (n=30), serous endometrial carcinoma (n=21), serous ovarian carcinoma (n=32) (Table 1). The following mutations were significant associated with the following cancer type (p≤0.05): endometrioid endometrial adenocarcinoma- arid1a, blm, chd4, crebbp, ctcf, fgfr2, mll2, mll3, msh2, msh6, nfe2l2, pik3r, pten, rnf43, taf1, tp53; serous endometrial cancer- brip1, crkl, fbxw7, ppp2r1a, tp53; serous ovarian cancer-arid1a, nf1, pik3ca, terc, tp53. Overall survival and progression free survival did not correlate significantly with any of the mutations identified, aside from PTEN mutation (p<0.05) which correlated with improved progression free survival in all cancer types, although in a very small subset of patients (n=14). Conclusions: Twenty-six significant mutations were identified in three common gynecologic cancer subtypes. Of these 26 mutations, four mutations have never been clinically reported for the respective cancer subtype, to our knowledge (Table 1). Larger numbers are needed to correlate these mutations with survival outcomes and to determine their impact on clinical management. To identify actionable gene mutations in recurrent ovarian and endometrial cancer and to evaluate the effect of these cancer specific mutations on survival outcomes at a single institution. This was an institutional review board (IRB) approved retrospective chart review of all patients at our institution with recurrent endometrial or ovarian cancer whose tumor underwent Next-Generation Sequencing using Foundation Medicine genomic profiling. Demographic and cancer related variables were extracted including age, cancer type, cancer stage at diagnosis, progression free survival and overall survival. Fisher's exact test was performed to identify specific cancer type associated with a particular mutation and a regression model was used to evaluate the cancer specific mutations’ correlation with age, cancer stage and survival. A total of 117 patients who met criteria were identified. Twenty-six significant mutations were found among three cancer subtypes; endometrioid endometrial adenocarcinoma (n=30), serous endometrial carcinoma (n=21), serous ovarian carcinoma (n=32) (Table 1). The following mutations were significant associated with the following cancer type (p≤0.05): endometrioid endometrial adenocarcinoma- arid1a, blm, chd4, crebbp, ctcf, fgfr2, mll2, mll3, msh2, msh6, nfe2l2, pik3r, pten, rnf43, taf1, tp53; serous endometrial cancer- brip1, crkl, fbxw7, ppp2r1a, tp53; serous ovarian cancer-arid1a, nf1, pik3ca, terc, tp53. Overall survival and progression free survival did not correlate significantly with any of the mutations identified, aside from PTEN mutation (p<0.05) which correlated with improved progression free survival in all cancer types, although in a very small subset of patients (n=14). Twenty-six significant mutations were identified in three common gynecologic cancer subtypes. Of these 26 mutations, four mutations have never been clinically reported for the respective cancer subtype, to our knowledge (Table 1). Larger numbers are needed to correlate these mutations with survival outcomes and to determine their impact on clinical management.