Plasma cfDNA methylation markers for the detection and prognosis of ovarian cancer

Background Plasma cell-free DNA (cfDNA) methylation has shown the potential in the detection and prognostic testing in multiple cancers. Herein, we thoroughly investigate the performance of cfDNA methylation in the detection and prognosis of ovarian cancer (OC). Methods The OC-specific differentially methylated regions (DMRs) were identified by sequencing ovarian tissue samples from OC (n = 61), benign ovarian disease (BOD, n = 49) and healthy controls (HC, n = 37). Based on 1,272 DMRs, a cfDNA OC detection model (OC-D model) was trained and validated in plasma samples from patients of OC (n = 104), BOD (n = 56) and HC (n = 56) and a prognostic testing model (OC-P model) was developed in plasma samples in patients with high-grade serous OC (HG-SOC) in the training cohort and then tested the rationality of this model with International Cancer Genome Consortium (ICGC) tissue methylation data. Mechanisms were investigated in the TCGA-OC cohort. Findings In the validation cohort, the cfDNA OC-D model consisting of 18 DMRs achieved a sensitivity of 94.7% (95% CI: 85.4%-98.9%) at a specificity of 88.7% (95% CI: 78.7%-94.9%), which outperformed CA 125 (AUC: 0.967 vs 0.905, P = 0.03). Then the cfDNA OC-P model consisting of 15 DMRs was constructed and associated with a better prognosis of HG-SOC in multivariable Cox regression analysis (HR: 0.29, 95% CI, 0.11-0.78, P = 0.01) in the training cohort, which was also observed in the ICGC cohort using tissue methylation (HR: 0.56, 95% CI, 0.32- 0.98, P = 0.04). Investigation into mechanisms revealed that the low-risk group had higher homologous recombination deficiency and immune cell infiltration (P < 0.05). Interpretation Our study demonstrated the potential utility of cfDNA methylation in the detection and prognostic testing in OC. Future studies with a larger population are warranted. Copyright (c) 2022 The Author(s). Published by Elsevier B.V.