Treatment arms were evenly matched; median age was 71 years (range, 44-95) in the abiraterone acetate arm and 70 years (44-90) in the placebo arm. Median time from initial diagnosis to first dose was 5.5 and 5.1 years. Median PSA was 42.0 and 37.7. Bone metastases were present in 83% and 80%, with 48% and 47% having >10 bone lesions; 66% and 64% had pain 0-1 on the BPI Short Form, respectively.
In February 2012, the Independent Data Monitoring Committee (IDMC) unblinded the study at a second interim analysis (planned, 40% OS events; actual, 43% OS events); the investigators and sponsors remained blinded. The IDMC concluded that rPFS, OS, and secondary end points all favored the abiraterone acetate arm and unanimously recommended unblinding the study and that patients in the placebo arm be offered treatment with abiraterone acetate.
Median duration of follow-up for both arms was 22.3 months. At the time of the second interim analysis, median rPFS had not been reached for the abiraterone acetate arm; for the placebo plus prednisone arm, median rPFS was 8.3 months (HR, 0.43; 95% CI, 0.35-0.52; P<0.0001). The rPFS benefit of abiraterone acetate was demonstrated across the full spectrum of patient subgroups, Dr. Ryan said, including baseline ECOG, baseline BPI, bone metastasis only at entry, age, baseline PSA above median, baseline LDH above median, baseline ALK-P above median, and geographic region.
A strong trend was observed in the OS primary end point; median OS had not been reached in the abiraterone arm and, in the placebo arm, median OS was 27.2 months (HR, 0.75; 95% CI, 0.61-0.93; P=0.0097). Point estimates for OS favored abiraterone acetate in all patient subgroups.
Serologically, PSA decline ≥50% was 62% in the abiraterone acetate arm and 24% in the placebo arm (P<0.0001). RECIST defined objective response was 36% in the abiraterone acetate arm (11% complete response and 25% partial response; 61% of patients had stable disease and 2%, progressive disease) and 16% in the placebo arm (4% complete response, 12% partial response, 69% stable disease, and 15% progressive disease) (RR 2.273 [95% CI, 1.591-3.247]; P<0.0001).
A statistically significant improvement was also observed in all secondary end points, including time to opiate use for cancer-related pain, median not reached in the abiraterone acetate arm vs 23.7 months in the placebo arm (P=0.0001); time to chemotherapy initiation, 25.2 vs 16.8 months (P<0.0001); time to ECOG PS deterioration, 12.3 vs 10.9 months (P=0.0053); and time to PSA progression, 11.1 vs 5.6 months (P<0.0001), respectively. All secondary end points remained significant after adjusting for multiplicity testing. Patient reported outcomes favored abiraterone acetate; full data will be reported in the future.
Subsequent therapy was common in both treatment arms, with 44.3% of patients in the abiraterone acetate arm and 60.3% in the placebo arm receiving such treatment, primarily docetaxel, 37.9% and 53.0%, respectively. Other subsequent therapies included cabazitaxel, ketoconazole, and sipuleucel-T. Prior to unblinding (eg, not per protocol), 4.8% of patients in the abiraterone acetate arm received subsequent abiraterone acetate, as did 10.0% of patients in the placebo arm.