We previously reported that olaparib led to significantly longer imaging-based progression-free survival than the physician’s choice of enzalutamide or abiraterone among men with metastatic castration-resistant prostate cancer who had qualifying alterations in homologous recombination repair genes and whose disease had progressed during previous treatment with a next-generation hormonal agent. The results of the final analysis of overall survival have not yet been reported.

In an open-label, phase 3 trial, we randomly assigned patients in a 2:1 ratio to receive olaparib (256 patients) or the physician’s choice of enzalutamide or abiraterone plus prednisone as the control therapy (131 patients). Cohort A included 245 patients with at least one alteration in BRCA1, BRCA2, or ATM, and cohort B included 142 patients with at least one alteration in any of the other 12 prespecified genes. Crossover to olaparib was allowed after imaging-based disease progression for patients who met certain criteria. Overall survival in cohort A, a key secondary end point, was analyzed with the use of an alpha-controlled, stratified log-rank test at a data maturity of approximately 60%. The primary and other key secondary end points were reported previously.

The median duration of overall survival in cohort A was 19.1 months with olaparib and 14.7 months with control therapy (hazard ratio for death, 0.69; 95% confidence interval [CI], 0.50 to 0.97; P=0.02). In cohort B, the median duration of overall survival was 14.1 months with olaparib and 11.5 months with control therapy. In the overall population (cohorts A and B), the corresponding durations were 17.3 months and 14.0 months. Overall, 86 of 131 patients (66%) in the control group crossed over to receive olaparib (56 of 83 patients [67%] in cohort A). A sensitivity analysis that adjusted for crossover to olaparib showed hazard ratios for death of 0.42 (95% CI, 0.19 to 0.91) in cohort A, 0.83 (95% CI, 0.11 to 5.98) in cohort B, and 0.55 (95% CI, 0.29 to 1.06) in the overall population.

Among men with metastatic castration-resistant prostate cancer who had tumors with at least one alteration in BRCA1, BRCA2, or ATM and whose disease had progressed during previous treatment with a next-generation hormonal agent, those who were initially assigned to receive olaparib had a significantly longer duration of overall survival than those who were assigned to receive enzalutamide or abiraterone plus prednisone as the control therapy, despite substantial crossover from control therapy to olaparib. (Funded by AstraZeneca and Merck Sharp and Dohme; PROfound ClinicalTrials.gov number, NCT02987543.)