Background: In the PAOLA-1/ENGOT-ov25 trial (NCT02477644), maintenance ola (a PARP inhibitor) + bevacizumab (bev) led to prolonged progression-free and overall survival vs placebo (pbo) + bev in AOC pts, specifically those with homologous recombination-deficient (HRD+) tumors (Ray-Coquard et al. ESMO 2022, LBA29). Here, we explore the efficacy of subsequent chemotherapy (CT) in pts after progression on first-line (1L) treatment in PAOLA-1. Methods: The efficacy of subsequent CT was explored by analyzing the median time from first subsequent therapy (FST) to second subsequent therapy (SST). Efficacy was compared by FST type, HRD status, and timing of progression (during/after ola). A multivariate Cox model was used in ola + bev pts to identify prognostic factors influencing time from FST to SST, including HRD status, platinum-free interval (PFI), clinical risk based on disease stage and surgical status, timing of progression, and response to 1L CT. All P values are exploratory. Results: In total, 544/806 (67.5%) pts progressed and received subsequent CT: 338/537 (62.9%) ola + bev pts and 206/269 (76.6%) pbo + bev pts. In the overall population and in the subgroup who received platinum-based combination therapy (PBC), time from FST to SST was shortest in pts who progressed during ola (Table). Pts who progressed after ola had similar outcomes to pbo + bev pts (Table). Efficacy was broadly consistent regardless of HRD status (Table). The multivariate analysis confirmed that progression after (vs during) ola prolonged time from FST to SST (hazard ratio [HR] 0.65, 95% confidence interval [CI] 0.50–0.84; P=0.0011) independently of other known prognostic factors, such as PFI ≥12 vs <12 months (HR 0.38, 95% CI 0.29–0.51; P<0.0001) or lower vs higher clinical risk (0.58, 0.42–0.81; P=0.0015). Conclusions: In this post hoc exploratory analysis of PAOLA-1, the efficacy of subsequent CT at relapse appeared dependent on whether progression occurred during or after the end of ola treatment. Efficacy was reduced when relapse occurred during ola but was comparable between pts who progressed after ola vs pbo + bev pts. Clinical trial information: NCT02477644.

*1/338 pts not treated. Kaplan–Meier estimates; median months with 95% CIs. HRD+ defined as tumor BRCA mutation and/or genomic instability score ≥42 (Myriad MyChoice HRD Plus assay)