Relationship between abiraterone exposure, prostate-specific antigen (PSA) kinetics, and overall survival (OS) in patients with metastatic castration-resistant prostate cancer (mCRPC) .
39 Background: Abiraterone, the active metabolite of abiraterone acetate (AA), is an effective androgen biosynthesis inhibitor for patients with metastatic castration-resistant prostate cancer (mCRPC). We conducted a sequential exposure-biomarker-survival modeling analysis to explore the relationship between prostate-specific antigen (PSA) kinetics and overall survival (OS) and to establish the exposure response for PSA kinetics and OS in chemotherapy-naïve and -pretreated patients with mCRPC following AA administration. Methods: The exposure-PSA-survival modeling framework was based on two phase III studies, COU-AA-301 (chemotherapy-pretreated, N = 1184) and COU-AA-302 (chemotherapy-naïve, N = 1081), and included a mixed-effects tumor growth inhibition (TGI) model to describe PSA dynamics in response to AA and a Cox proportional hazards survival model to evaluate the relationship between relative risk of death and PSA dynamic end points. Results: The TGI model best described the longitudinal PSA dynamics following AA treatment. Abiraterone exposure significantly increased PSA decay rate (maximum effect of 2.72, p < 0.0001). The estimated concentration for 50% of the maximum effect (EC50) was 4.75 ng/mL. The abiraterone effect on PSA kinetics was similar in chemotherapy-naïve and -pretreated subjects, and approximately 90% of subjects had a steady-state concentration greater than the EC50. All model-predicted PSA metrics were strongly associated with OS in both populations; model-based post-treatment PSA doubling time showed the strongest association (hazard ratios approximately 0.9 in both populations). Simulations showed that the modeling framework could accurately predict the survival outcome for both studies. Conclusions: The analysis revealed a similar effect of abiraterone on PSA kinetics and association between PSA kinetics and OS in chemotherapy-naïve and -pretreated subjects, providing additional evidence for surrogacy of PSA kinetics and the use of PSA end points to indicate clinical benefit of abiraterone in subjects with mCRPC regardless of prior chemotherapy. Furthermore, the study confirmed that the recommended 1,000 mg/d dose of AA leads to adequate clinical exposure above the effective level. Clinical trial information: NCT00638690, NCT00887198.