EP1128 Differential DNA methylation in High Grade Serous Ovarian Cancer (HGSOC) is associated with tumor behaviour

Introduction/Background The epigenome represents and adds another facet of complexity to cancer that needs to be understood to categorize patients into those at risk of disease, recurrence, or treatment failure. Our goal is to compare DNA methylation status between controls vs HGSOC, and relate it with clinical outcomes. Methodology A retrospective review of our institution9s advanced/ recurrent HGSOC patients yielded 97 cases with good quality DNA and RNA. 12 normal fallopian tubes were also collected and used as controls. Infinium Illumina methylationEPIC was used to characterize DNA methylation status and RNA seq to evaluate gene expression. T- tests were used to compare methylation patterns between controls vs HGSOC, primary vs recurrent, and optimal vs suboptimal surgical outcomes. Spearman9s rank correlation was used to evaluate association between degree of methylation and expression of identified genes. Validation of our findings used the cancer genome atlas (TCGA) database followed by C-statistics to assess degree of agreement. Results Out of 66,069 methylation probes that interrogate known genes, 5,852 probes were significantly different between normal and HGSOC. This includes genes that are enriched in several pathways such as the RAS signaling pathway. A similar analysis of the TCGA database showed 2,075 differentially methylated probes. 1,891 probes were significant in both datasets giving a 70.1% degree of agreement. When comparing differential DNA methylation between primary and recurrent disease, there were 57 probes which represent 17 genes that were significantly different. When comparing optimal vs suboptimal surgical outcomes, 99 probes were significantly different wherein 29 genes show expected inverse correlation between methylation status and gene expression. Conclusion There are significant differences in methylation patterns between HGSOC and fallopian tubes, primary vs recurrent disease, and optimal vs sub-optimal surgical outcomes. Cataloging these phenomena in a well characterized clinical population will aid in determining groupings for precision medicine treatment cohorts in the future. Disclosure Nothing to disclose