Phase II Randomized Study of MK‐2206 and Bicalutamide in Prostate Cancer Patients With Rising PSA After Primary Therapy (ECOG‐ACRIN E2809)

Background Combined inhibition of AKT and the androgen receptor (AR) might be more efficacious than AR inhibition alone in biochemical recurrence (BCR) of prostate cancer (PC) by additive effects on the two effector pathways. Patients and Methods E2809 assigned 108 high‐risk BCR patients (1:1, arm‐A: arm‐B) to two phases with 4‐week cycles: first (PreBic) phase (3 cycles), arm‐B received AKT inhibitor MK‐2206 200 mg 2x/week, arm‐A was observation; second phase, bicalutamide (Bic), 50 mg daily, was added to each arm. The PreBic phase assessed whether AKT inhibition changed PSA, reflecting AR activation. Exploratory primary endpoint (EP1) at 6–8 Bic cycles and secondary EP2 at cycle 14 compared the proportion of cases with ≥stable disease (SD) or progressive disease (PD). Results PreBic phase: 70% arm‐B vs 32% arm‐A patients had a ≥ 60% PSA rise ( p = 0.02). MK‐2206 skin toxicity (38% grade ≥ 3) drove high withdrawal prior to EP1. Bic phase: at EP1, inter‐arm PSA response (81%/82%) or PD (19%/18%) was equal. At EP2, PD was greater in arm‐A than arm B (36% vs 11%; p = 0.04). In subgroups defined by PreBic phase PD (PBPD; PSA rise ≥ 25%), 9/16 arm‐A vs 0/7 arm‐B cases experienced PD at EP2 ( p = 0.02); overall between PBPD‐subgroups, 73% arm‐A vs 27% arm‐B experienced PD. Conclusions The results suggest that latent improved outcome of high‐risk BCR patients (mean PSA doubling time 4.4 months) on combined MK‐2206+Bic versus Bic alone was attributable to a subgroup identified by crosstalk AR activation secondary to inhibition of AKT. Toxicity may affect tolerance of sustained AKT‐AR inhibition. Clinical Trial Registration ClinicalTrials. gov number, NCT01251861.