Repeated measures analysis of patient-reported outcomes in prostate cancer after abiraterone acetate

Background: Metastatic castration-resistant prostate cancer (mCRPC) is typically associated with declining health-related quality of life (HR-QoL). Objective: To assess patient experience with abiraterone acetate (hereafter abiraterone) plus prednisone longitudinally. Methods: COU-AA-302 was a phase 3, multinational, randomized, double-blind study that enrolled asymptomatic or mildly symptomatic, chemotherapy-naïve patients with mCRPC. Patients were randomized to 1 g abiraterone daily plus 5 mg prednisone BID (n = 546) or placebo plus prednisone (n = 542) in continuous 28-day cycles. Patient-reported outcomes (PROs) were collected using the Functional Assessment of Cancer Therapy-Prostate (FACT-P) questionnaire, consisting of 4 well-being subscales (physical, social/family, emotional, functional) and a prostate cancer-specific subscale (PCS). The trial outcome index (TOI) is a composite of the physical well-being, functional well-being, and PCS scores. Least squares mean change from baseline at each cycle up to 1 year (cycle13) was compared between treatment arms using a mixed-effects model for repeated measures, which assumed that data were "missing at random." A pattern-mixture model (PMM) with multiple imputation was performed to address the assumption that data were "missing not at random." Results: Significant differences favoring abiraterone-prednisone were observed for FACT-P total, TOI, and PCS scores, and for all well-being subscales except social/family well-being over the first year of treatment. These results were supported by the PMM with multiple imputation. Limitations: Attrition after 1 year limited our ability to analyze the PRO data beyond that time point. Conclusions: Abiraterone-prednisone confers sustained HR-QoL benefits over the course of treatment