Progression-Free Survival with PARP Inhibitors According to Clinical Risk in Patients with Ovarian Cancer: An Indirect Comparison Using Reconstructed Data.

Background

Poly (ADP-ribose) polymerase (PARP) inhibitors (PARPi) are established maintenance treatments in ovarian cancer, but comparative efficacy across genetic profiles and relapse risk categories remains unclear. The aim of this study was to compare the efficacy of different PARPi as maintenance therapy in ovarian cancer across genetic profiles and relapse risk categories using reconstructed individual patient data (IPD) from randomized trials (RCTs).

Methods

IPD were reconstructed using the IPDfromKM method from published Kaplan-Meier curves of RCTs stratified by clinical risk subgroup. Progression-free survival (PFS) was the primary endpoint. Three comparisons were performed: Breast Cancer gene (BRCA)+ high-risk, Homologous Recombination Deficiency (HRD)+/BRCAwt high-risk, and BRCA+ low-risk populations. Restricted Mean Survival Time (RMST) was calculated as a supplementary measure, with curves truncated at 66 months.

Results

In the BRCA+ high-risk population, olaparib monotherapy (median PFS 41.2 months) and olaparib plus bevacizumab (median PFS 42.5 months) demonstrated the greatest PFS benefit, marginally outperforming niraparib (median PFS 31.2 months). RMST analysis showed a 14-month advantage for olaparib plus bevacizumab over bevacizumab alone. In the HRD+/BRCAwt high-risk population, olaparib plus bevacizumab and niraparib showed comparable efficacy, with no statistically significant inter-treatment difference. In the BRCA+ low-risk population, olaparib plus bevacizumab showed superior HR versus olaparib monotherapy, without reaching statistical significance. RMST analysis also indicated an advantage of 8.5 months for the combination, though this did not reach statistical significance.

Conclusions

PARPi treatment benefit in ovarian cancer is meaningfully influenced by genetic profile and relapse risk, supporting biomarker-driven treatment selection in clinical practice.