Adjustment for imbalances in baseline characteristics in the MAGNITUDE phase 3 study confirms the clinical benefit of niraparib in combination with abiraterone acetate plus prednisone in patients with metastatic prostate cancer.

BACKGROUND

MAGNITUDE (NCT03748641) demonstrated favourable outcomes with niraparib plus abiraterone acetate plus prednisone (+AAP) versus placebo+AAP in patients with BRCA1/2-altered metastatic castration-resistant prostate cancer (mCRPC). Imbalances in prognostic variables were reported between arms, which impacts estimation of both the clinical benefit and cost‑effectiveness of niraparib+AAP for healthcare systems. A pre-specified multivariable analysis (MVA) demonstrated improved overall survival (OS) with niraparib+AAP. Here, we used an inverse probability of treatment weighting (IPTW) model to adjust for covariate imbalances and assess time-to-event outcomes.

METHODS

IPTW analysis of time-to-event outcomes was conducted using data from patients with BRCA1/2-altered mCRPC (N = 225) in MAGNITUDE. Patients received niraparib+AAP or placebo+AAP. OS, radiographic progression-free survival, time to symptomatic progression, time to initiation of cytotoxic chemotherapy and time to prostate-specific antigen progression were assessed. Weighted Kaplan-Meier curves were generated for each endpoint, and adjusted hazard ratios (HR) were obtained from a weighted Cox model.

RESULTS

Improvements in survival outcomes were estimated for niraparib+AAP versus placebo+AAP: unadjusted median OS was 30.4 months versus 28.6 months, respectively (HR: 0.79; 95 % confidence interval [CI]: 0.55, 1.12; p = 0.183). Following IPTW, median OS increased to 34.1 months with niraparib+AAP versus a decrease to 27.4 with placebo (HR: 0.65; 95 % CI: 0.46, 0.93; p = 0.017). Similar improvements were observed for other time-to-event endpoints.

CONCLUSIONS

IPTW adjustment provided a more precise estimate of the clinical benefit of niraparib+AAP versus placebo+AAP in patients with BRCA1/2-altered mCRPC. Results were consistent with the pre-specified MVA, and further demonstrated the value of adjusting for baseline imbalances, particularly in smaller studies.

TRIAL REGISTRATION

NCT03748641 (MAGNITUDE).