Abstract 6219: Endogenous human retroviruses in epithelial ovarian cancers: An exploratory analysis

Abstract Objectives: Endogenous human retroviruses (ERVs) are remnants of exogenous retroviruses that gained access to germline cells allowing integration into the human genome. Emerging data suggest that some ERVs may become activated allowing epigenetic alterations through DNA methylation or histone modification, which can further alter gene regulation. This may serve as a targeted therapeutic opportunity in modulating oncogenesis, through aberrant oncogene activation or tumor suppressor gene inactivation. This is an emerging area of exploration in ovarian cancer. Methods: We applied our ERV mapping tools (1) to RNA-seq data from 24 female patients with ovarian, fallopian tube, or primary peritoneal cancers to investigate expression of ~550,000 ERV elements from the Human Endogenous Retrovirus database (HERVd) (2, 3). Univariate Cox regression models were constructed with normalized ERV expression data and the most significantly differentially expressed ERVs were then filtered using a penalized Lasso-Cox proportional-hazards model. ERV expression, alongside the available clinical data (including age, histotypes, and tumor stage) were provided as inputs and linear predictors generated by the model separated samples into either high or low risk categories. We identified a panel of 15 predictive ERVs and these risk evaluations allowed for Kaplan-Meier analyses of survival by clinical parameters alone versus combined with ERV signature. We also performed a secondary analysis of platinum resistant ovarian cancers using RNA-seq data (GSE102118) to investigate the effect of demethylating agent guadecitabine on ERV expression (n=9) (4). Results: Exploratory analyses of ERVs demonstrated significantly different expression between histologies (clear cell, mucinous, endometrioid, high grade serous) and FIGO disease stage. In our prognostic model, Kaplan-Meier survival analysis using only clinical parameters resulted in a significance level of 0.0013, however the supplemented model combining the 15-ERV panel and the clinical data discriminated the two risk groups for time to recurrence at a much higher significance level of p = 7.076 × 10−8. Included in the 15 ERV panel are HERVK, HERV3, and several putative ERV promoters including LTR12, LTR7, LTR3 and LTR4 (5). In our secondary analysis of platinum-resistant patients, the ERV transcripts were compared for each patient pre- and post-guadecitabine treatment and there were several ERVs with risk-predictive values induced in patients receiving guadecitabine including LTR7, LTR12 and HERVK. Conclusions: In summary, ERV RNA expression in ovarian cancer is significantly different between histologies and disease stages. The ability to successfully classify patients as either high-risk or low-risk for disease progression is of considerable value for patient management and ERVs may improve prognostication and guide future therapeutic targeting. Citation Format: Jill Alldredge, Vinay Kumar, Brook Sanders, Farahnaz Rahmatpanah. Endogenous human retroviruses in epithelial ovarian cancers: An exploratory analysis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 6219.