Abstract A007: DNA methylation analysis comparing high-grade serous ovarian cancer to the normal ovary with adjustment for cell type variation identifies tumor-specific methylation alterations

Abstract High-grade serous ovarian carcinoma (HGSOC) is the most common form of ovarian cancer and the most fatal gynecologic cancer, accounting for approximately 12, 800 deaths in 2022. DNA methylation is an epigenetic modification involved in lineage specification and cellular differentiation that has distinct patterns by cell type. Once defined, cell-specific DNA methylation patterns can be leveraged as biomarkers of cell types in deconvolution approaches where cell proportions are inferred from measured composite DNAm signals in common biospecimens like blood and tumor. While epigenetic modifications, specifically methylation of DNA, are known to contribute to carcinogenesis and disease progression, to date there has been limited application of approaches accounting for cellular heterogeneity in the ovarian tumor microenvironment. Adjusting for cellular heterogeneity addresses potential confounding by cell type in bulk DNA methylation analyses. Using publicly available HGSOC tumor data (GSE133556) and GTEx normal ovarian data, we applied Hierarchical Tumor Immune Microenvironment Epigenetic Devonvolution (HiTIMED) to profile cell types in the tumor microenvironment. We also performed repetitive element methylation prediction using the REMP package and copy number variation analysis using the conumee package in R. DNA methylation data from all tumor (n = 99) and nontumor normal samples (n=111) was deconvoluted into six major cell types: tumor, lymphocyte, myeloid, endothelial, epithelial, and stromal. Data from HGSOC tumors were divided into tumor high (25% tumor burden) and tumor low cell proportions (25% tumor burden). Assessing non-tumor cell proportions, we identified significant differences in lymphocyte, endothelial, epithelial, and stromal cell type proportions in HGSOC samples with a high tumor burden compared to normal, and we identified significant differences in lymphocyte, endothelial, epithelial, stromal, and myeloid cell type in HGSOC samples with low tumor burden compared to normal. Assessing repetitive element methylation, we found that Alu repetitive elements are significantly hypomethylated in HGSOC samples in comparison to normal ovarian tissue. Performing copy number variation analysis, we identified regions of copy number gain and loss associated with HGSOC. Our study identifies tumor-specific DNA methylation alterations in high-grade serous ovarian carcinoma samples compared to normal ovarian tissue and demonstrates differences in specific cell proportions in the tumor ovarian microenvironment, which may play a greater role in ovarian cancer etiology. Citation Format: Irma M. Vlasac, Brock C. Christensen, Lucas A. Salas. DNA methylation analysis comparing high-grade serous ovarian cancer to the normal ovary with adjustment for cell type variation identifies tumor-specific methylation alterations abstract. In: Proceedings of the AACR Special Conference on Ovarian Cancer; 2023 Oct 5-7; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84 (5 Suppl₂): Abstract nr A007.