Progression-free survival as a surrogate outcome for overall survival in ovarian cancer maintenance therapy randomized controlled trials.

Abstract

5545 Background: A traditional primary outcome in prospective trials is progression-free survival (PFS). PFS often functions as a surrogate outcome for overall survival (OS) to speed up the translation of research findings into practice. While PFS has been validated as a surrogate for OS in contemporary therapeutic trials for patients with advanced ovarian cancer, there is little evidence to support the validity of PFS as a surrogate for OS in contemporary trials of maintenance therapies. Our objective was to evaluate whether PFS is a reliable surrogate outcome for OS in patients with ovarian cancer receiving maintenance therapy after platinum-based chemotherapy. Methods: In May 2024, MEDLINE was queried for all phase 3 trials evaluating poly (ADP) ribose polymerase (PARP) inhibitors and bevacizumab in the maintenance setting for ovarian, fallopian tube and primary peritoneal cancers. Included trials studied PARP inhibitors, bevacizumab or both as an intervention compared to control. Enrollment numbers, median follow up, PFS, OS and hazard ratios were abstracted. Using a meta-analytic approach, correlation analysis was performed using weighted linear regression and Pearson’s correlation coefficient. Trials included contained complete survival data. Criteria for PFS surrogacy required R 2 >0.8. Results: Sixty trials were identified, 11 of which met inclusion criteria. Six trials investigated PARP inhibitors in the maintenance setting, while 4 trials investigated bevacizumab. One trial investigated both a PARP inhibitor and bevacizumab. The pooled sample size from all trials was n=6,243. Median follow up time was 60.35 months. Across all trials, the relationship between OS and PFS HRs was approximately linear. Corresponding R 2 values were low (R 2 =0.35, 95% CI 0-0.63). Pearson correlation when weighted by total study sample size, was of low strength (r=0.59, 95% CI -0.05-0.87). Four trials evaluating bevacizumab as a maintenance therapy demonstrated favorable PFS benefit with no statistically significant difference in OS, while only one PARP inhibitor trial demonstrated a statistically significant benefit in PFS and OS. Weighted Pearson correlation coefficient for bevacizumab trials demonstrated moderate correlation between PFS and OS HRs (r=0.81, 95% CI -0.75-0.99) while PARP inhibitor trials demonstrated a low strength of correlation (r=0.26, 95% CI -0.71-0.88). Conclusions: Phase 3 clinical trials assessing maintenance therapies for ovarian cancer demonstrate poor correlation between PFS and OS. This effect may be modulated by type of maintenance therapy and the inclusion of platinum-based chemotherapy in trial arms. PFS as a surrogate outcome in maintenance therapy ovarian cancer clinical trials must be supported by additional studies and caution should be taken prior to regulatory approval based on PFS data alone.