Epigenetic mechanisms of PARP inhibitor resistance in ovarian cancer: A systematic review with bioinformatic analysis of clinically actionable genes

PARP inhibitors (PARPi) improve ovarian cancer (OC) outcomes, but resistance remains a major challenge without reliable prognostic biomarkers. This study identified epigenetic hallmarks of PARPi resistance by integrating 27 studies (22 preclinical, 5 clinical) from the past 15 years, and validating candidate genes using web-based bioinformatics tools and public microarray/RNA-seq datasets from non-relapsed, primary OC tissues. We hypothesised that early aberrant expression of these epigenetically altered, PARPi resistance-related genes in tumours may be linked to disease progression (PFS) and could serve as early biomarkers to be associated with PARPi resistance during first-line treatment. We confirmed epigenetic involvement in PARPi resistance across 36 genes linked to epigenetic modifications. Of these, 10 genes (n = 614-1435)-including RNASEH2B (HR=1.41), VHL (HR=1.26), ATM (HR=1.22), XRCC1 (HR=1.20), NRP1 (HR=1.16), KAT2B (HR=1.16), EZH2 (HR=1.15), CREBBP (HR=1.14), FZD10 (HR=0.87), and CARM1 (HR=0.86)-showed significant prognostic value for PFS (all: p < 0.05). This 10-gene signature remained collectively significant (HR 1.27, p = 0.014). RNA-seq validation showed differential expression of these genes, with highest fold-change overexpression in tumours for FZD10 (4.20), EZH2 (3.56), and CARM1 (1.61), and lowest in ATM (0.22), KAT2B (0.33), and NRP1 (0.44). GO and KEGG analyses revealed these genes are enriched in key resistance pathways, including impaired DNA repair, reduced replication stress, immune evasion, and stemness maintenance. This review with bioinformatic validation identified a 10-gene epigenetic signature associated with PARPi resistance and disease progression. These clinically actionable genes, aberrantly expressed before treatment, may serve as early biomarkers for risk stratification. Further validation in PARPi-sensitive and -resistant ovarian cancer cohorts is needed.