Background and objective: The PEACE-3 trial demonstrated overall survival benefit for enzalutamide plus radium-223 versus enzalutamide alone in metastatic castration-resistant prostate cancer (mCRPC), but survival curves showed initial crossing followed by progressive separation, indicating non-proportional hazards. We aimed to characterize the time-dependent treatment effects through comprehensive analysis.
Methods: Individual patient datas (IPD) were reconstructed from published Kaplan-Meier curves of the PEACE-3 trial. Time-dependent treatment effects were evaluated by restricted mean survival time (RMST) analysis at predefined time points (18-72 months), time-dependent Cox regression with treatment-by-time interaction, piecewise Cox regression across 6 time intervals, landmark analyses, and Fleming-Harrington weighted log-rank tests.
Results: A total of 446 IPDs were reconstructed. Validation confirmed consistency with the original trial. During the initial 18 months, combination therapy showed no survival advantage (RMST difference: -0.36 months, 95% CI, -0.90 to 0.18, p = 0.20) and was associated with increased hazard (HR = 2.14, 95% CI, 1.48-3.10, p < 0.001). Significant survival benefits were observed after 60 months, with RMST differences of 4.34 months (95% CI, 0.49-8.19, p = 0.03) at 60 months and 6.25 months (95% CI, 1.56-10.95, p = 0.01) at 72 months. Time-dependent Cox regression confirmed a significant treatment-by-time interaction (p < 0.01). Piecewise analysis revealed the most substantial benefit for 60-72 months (HR = 0.20, 95% CI, 0.05-0.77, p = 0.02). Landmark analyses consistently demonstrated increasing treatment benefit with longer follow-up.
Conclusions: Enzalutamide plus radium-223 demonstrates delayed but substantial survival benefit in mCRPC, becoming statistically significant after 60 months with over 6 months survival advantage at 72 months. REGISTRY AND THE REGISTRATION NO.
Of the study/trial: Not applicable.
Keywords: enzalutamide; metastatic castration‐resistant prostate cancer; non‐proportional hazards; radium‐223; shiny method.
© 2026 The Japanese Urological Association.