Background Epithelial ovarian cancer (EOC) is the most lethal gynecologic malignancy. Greater than 75% of cases are diagnosed with advanced stage, which is associated with 5-year survival rates of less than 20% compared with 92% if caught early. Previous attempts at discovering early detection biomarkers for EOC have largely failed because of reliance on sub-optimal discovery approaches and the lack of comprehensive patient cohorts. DNA methylation changes are among the earliest alterations in cancer and methylation represents a stable and robust modification on DNA. In addition, several groups including ours have reported the detection of tumor-associated methylation changes in cfDNA extracted from plasma. Accordingly, in this study, we performed comprehensive DNA methylation analysis by reduced representation bisulfite sequencing (RRBS) to identify DNA methylation biomarkers that can be detected in circulation and used for the early detection of ovarian cancer. Methods For the biomarker discovery phase, a cohort of 23 stage 1 serous EOC tissues, and 10 unmatched, adjacent normal tissues from fallopian tubes were analyzed by RRBS. We used Metilene, a binary segmentation algorithm combined with a two-dimensional statistical test, to detect differentially methylated regions (DMRs). For validation, we performed targeted bisulfite amplicon sequencing. Briefly, bisulfite primers were designed to regions of interest, optimized, multiplexed and sequenced to depths ranging from 500-25,000X on the Illumina MiSeq platform. The validation cohort consisted of cfDNA isolated from the plasma of 51 independent EOCs and 10 normal individuals as well as the genomic DNA from tissue used in the discovery cohort. Results We identified 10972 statistically significant (q-value <0.05) DMRs between stage 1 EOC and normal samples. Of these, 227 were hypermethylated and had an average differential methylation value of greater than 30% in tumors compared to normal. Unsupervised clustering analysis of these DMRs demonstrated clear separation of tumor from normal, with the possibility of three distinct molecular subtypes of serous EOC. Six DMRs were prioritized and further validated in both tissue and plasma. Targeted bisulfite amplicon sequencing performed revealed hypermethylation in greater than 50% of tumor samples across the DMRs. Additionally, the area under the curve (AUC) for the DMRs interrogated ranged from 0.91 and 0.94. Conclusion In conclusion, our RRBS based comprehensive biomarker discovery identified a six-DMR panel that demonstrated preliminary robust diagnostic accuracy in EOC samples. Further evaluation of these DMRs in multicenter retrospective and prospective cohorts might lead to the development of a non-invasive diagnostic, as well as a population-screening assay for OC patients.

Citation Format: Gerald C. Gooden, Christophe Legendre, Monique Spillman, Ajay Goel, Bodour Salhia. Circulating epigenetic biomarkers for the early detection of ovarian cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3309.