Abiraterone acetate plus prednisolone for metastatic patients starting hormone therapy: 5-year follow-up results from the STAMPEDE randomised trial (NCT00268476)

Abiraterone acetate plus prednisolone (AAP) previously demonstrated improved survival in STAMPEDE, a multi-arm, multi-stage platform trial in men starting long-term hormone therapy for prostate cancer. This long-term analysis in metastatic patients was planned for 3 yrs after the first results. Standard-of-care (SOC) was androgen deprivation therapy. The comparison randomized patients 1:1 to SOC-alone with or without daily abiraterone acetate 1000 mg + prednisolone 5 mg (SOC + AAP), continued until disease progression. The primary outcome measure was overall survival. Metastatic disease risk group was classified retrospectively using baseline CT and bone scans by central radiological review and pathology reports. Analyses used Cox proportional hazards & flexible parametric models, adjusted for baseline stratification factors. 1003 patients were contemporaneously randomized (Nov-2011--Jan-2014): median age 67 yr; 94% newly-diagnosed; metastatic disease risk group: 48% high, 44% low, 8% un-assessable; median PSA 97 ng/mL. At 6.1 yr median follow-up, 329 SOC-alone deaths (118 low-risk, 178 high-risk) and 244 SOC + AAP deaths (75 low-risk, 145 high-risk) were reported. Adjusted HR = 0·60 (95%CI:0·50—0·71; P = 0.31x10−9) favoured SOC + AAP, with 5-yr survival improved from 41% SOC-alone to 60% SOC + AAP. This was similar in low-risk (HR = 0·55; 95%CI:0·41—0·76) and high-risk (HR = 0·54; 95%CI:0·43—0·69) patients. Median and current maximum time on SOC + AAP was 2.4 yr and 8.1 yr. Toxicity at 4 yr post-randomisation was similar, with 16% patients in each group reporting grade 3 or higher toxicity. A sustained and substantial improvement in overall survival of all metastatic prostate cancer patients was achieved with SOC + abiraterone acetate + prednisolone, irrespective of metastatic disease risk group. This article is protected by copyright. All rights reserved.