Updated overall survival (OS) analysis for ProCAID: A randomized, double-blind, placebo-controlled phase II trial of capivasertib with docetaxel versus docetaxel alone in metastatic castration-sensitive prostate cancer (mCRPC)

108 Background: The AKT pathway is frequently deregulated in mCRPC. ProCAID tested addition of capivasertib, a potent selective inhibitor of all three AKT isoforms (AKT1/2/3) to docetaxel chemotherapy vs. placebo plus docetaxel for mCRPC. The primary analysis showed no difference between treatment arms for the primary endpoint of composite progression free survival (cPFS). However, OS, which was a secondary endpoint, was extended in the capivasertib plus docetaxel arm. cPFS and OS results were consistent irrespective of PI3K/AKT/PTEN pathway biomarker status (Crabb et al, J Clin Oncol 2021;39(3):190-201). Methods: An updated analysis of mature OS data was undertaken once events reached ≥65% by Cox proportional hazards model, adjusted for minimisation factors, within the intent to treat (ITT) population (n = 150). Patients (pts) and investigators remained blinded to treatment allocation. We also investigated OS outcomes within subsets based on prior androgen receptor targeted agent (ARTA) exposure to abiraterone and/or enzalutamide (abi/enza) and the balance of post-trial life extending treatment use by treatment arm. Funding: Cancer Research UK (C9317/A16029, CRUK/12/042) and AstraZeneca. Results: At this OS update, 99 pts (66.0%) had died, with 88 of these deaths (88.9%) due to prostate cancer. 5 pts (3.3%) remained on capivasertib or placebo. Median OS was 25.3 months for the capivasertib plus docetaxel arm vs. 20.3 months for placebo plus docetaxel (hazard ratio (HR) 0.70, 95% confidence interval (CI) 0.47 to 1.05; nominal p = 0.09). One, or more, subsequent life extending treatment options, including abiraterone, enzalutamide, radium-223 and cabazitaxel, were received by 99 pts (66.0%) and were balanced between treatment arms (68% capivasertib, 64% placebo). 101 pts (67.3%; 51 capivasertib, 50 placebo) had received abi/enza prior to entering ProCAID. Within this subgroup, OS benefit for capivasertib plus docetaxel (median OS 31.1 months) was maintained vs. placebo plus docetaxel (median OS 19.3 months; HR 0.57, 95% CI 0.36 to 0.91), but not in the remaining 49 pts who were naive to prior abi/enza (median OS 31.1 vs. not reached respectively; HR 1.43, 95% CI 0.63 to 3.23). These updated OS results remained consistent irrespective of biomarker status for PI3K/AKT/PTEN pathway activation. No clinically significant differences from the previously reported safety outcomes were seen with extended follow up of this trial. Conclusions: OS remains longer within the ProCAID ITT population with the addition of capivasertib to docetaxel for mCRPC. This does not appear to be explained by subsequent treatment choices. Exploratory analysis found prolonged OS with capivasertib within a subset of pts previously exposed to an ARTA which should be evaluated in prospective trials. Clinical trial information: NCT02121639.