599P Tumor mutation burden is a predictive biomarker for survival to patients with ovarian carcinoma

Abstract

With the development of immunotherapy, more and more patients with cancer, especially advanced cancers, clinically benefit from it. The panel of NCCN guidelines recommend that Pembrolizumab is acceptable recurrence therapies for epithelial ovarian cancer with high tumor mutation burden (TMB-H) at second-line therapy according to the finding of KEYNOTE158. However, KEYNOTE158 as a single-arm phase II clinical study, has some limitations. For example, a small number of patients with ovarian cancer were enrolled and it is difficult to distinguish whether TMB is a prognostic biomarker rather than a biomarker for immunotherapy. Therefore, we explored the prognostic role of TMB in ovarian cancer by public databases. Publicly available data from publicly available data from the Cancer Genome Atlas (n=617) was analyzed to calculate the TMB and investigate the association of TMB with the overall survival (OS) and disease-free survival (DFS) of patients. The correlation between survival and clinical characteristics including TMB was analyzed by Cox proportional hazards analysis. Furthermore, gene ontology enrichment analysis, and KEGG pathway analysis are conducted with RNA expression profiles of the TMB-H and non-TMB-H patients. We ranked TMB and divide patients into TMB-H and non-TMB-H by dichotomy (cutoff=1.4 mut/MB). The Kaplan-Meier analysis revealed that DFS (P＜0.001) and OS (P＜0.001) were significantly prolonged in TMB-H patients compared with non-TMB-H patients. In univariable analysis, Clinical factors statistically associated with DFS were age, race, clinical stages and TMB. Furthermore, multivariate Cox proportional hazards analysis revealed that TMB was a significant predictor of DFS and patients with non-TMB-H had worse DFS with Hazard Ratio of 1.850 (95% CI: 1.404-2.437, P＜0.0001). Of 280 genes differentially expressed between TMB-H and non-TMB-H patients. The most significant and meaningful terms of gene ontology were the presynapse, iron ion binding in tcellular component, and molecular function, respectively. In addition, the one pathway was predicted by the Kyoto Encyclopedia of Genes and Genomes database. TMB is a predictive biomarker for survival to patients with ovarian carcinoma.