A novel BRCA mutation classification system reveals differential responses to PARP inhibition and prognostic outcomes in epithelial ovarian cancer: a multicenter study.

Although poly (ADP-ribose) polymerase inhibitors (PARPis) have transformed the treatment landscape for the BReast CAncer susceptibility gene (BRCA)-mutated ovarian cancer, emerging evidence suggests significant heterogeneity in treatment efficacy. However, findings regarding the impact of mutation location and type remain debated, underscoring the necessity for more data and better subclassification. This study aimed to evaluate the association between novel BRCA mutation subgroups and PARPi response in epithelial ovarian cancer. This is a multicenter retrospective study conducted in China which enrolled BRCA1/2-mutated epithelial ovarian cancer patients between 2015 and 2024. Patients who achieved complete or partial response after surgery and platinum-based chemotherapy were included. Progression-free survival (PFS) was assessed with Kaplan-Meier curves (log-rank test) and Cox models to estimate hazard ratios with 95% confidence intervals. A total of 465 patients were included, of whom 319 (68.6%) received PARPi as maintenance therapy. We introduced a novel classification system that continuously segments the entire BRCA1 gene. Patients with mutations in the DNA-binding domain (DBD) and adjacent region [between DBD and C-terminal domain of BRCA1 (BRCT) (DB)] had poor prognosis without PARPi treatment but showed substantial benefit with PARPi (P = 0.0097 and P = 0.0016, respectively). Conversely, mutations in the Really Interesting New Gene (RING) domain and its downstream region (between RING and DBD) showed limited therapeutic benefit from PARPi and were associated with poorer overall outcomes. Patients with mutations located in the BRCT domain exhibited favorable prognosis regardless of PARPi treatment, diminishing the relative benefit of PARPi in this subgroup. According to the mutation type, patients with frameshift or nonsense mutations experienced significant improvement with PARPi (both P < 0.05). Our novel classification revealed that PARPi efficacy and prognosis varied significantly by mutation location, with the central region linked to superior outcomes compared with the N- and C-terminal regions. Comprehensive mutation profiling should guide treatment decisions to optimize outcomes in patients with BRCA-mutated ovarian cancer.