Association of Androgen Deprivation Therapy With Cardiovascular Death in Patients With Prostate Cancer: A Meta-analysis of Randomized Trials

CONTEXT Whether androgen deprivation therapy (ADT) causes excess cardiovascular deaths in men with prostate cancer is highly controversial and was the subject of a joint statement by multiple medical societies and a US Food and Drug Administration safety warning. OBJECTIVE To perform a systematic review and meta-analysis of randomized trials to determine whether ADT is associated with cardiovascular mortality, prostate cancer–specific mortality (PCSM), and all-cause mortality in men with unfavorable-risk, nonmetastatic prostate cancer. DATA SOURCES A search of MEDLINE, EMBASE, and the Cochrane Central Register of Controlled Trials databases for relevant randomized controlled trials in English between January 1, 1966, and April 11, 2011. STUDY SELECTION Inclusion required nonmetastatic disease, intervention group with gonadotropin-releasing hormone agonist–based ADT, control group with no immediate ADT, complete information on cardiovascular deaths, and median follow-up of more than 1 year. DATA EXTRACTION Extraction was by 2 independent reviewers. Summary incidence, relative risk (RR), and CIs were calculated using random-effects or fixed-effects models. RESULTS Among 4141 patients from 8 randomized trials, cardiovascular death in patients receiving ADT vs control was not significantly different (255/2200 vs 252/1941 events; incidence, 11.0%; 95% CI, 8.3%-14.5%; vs 11.2%; 95% CI, 8.3%-15.0%; RR, 0.93; 95% CI, 0.79-1.10; P = .41). ADT was not associated with excess cardiovascular death in trials of at least 3 years (long duration) of ADT (11.5%; 95% CI, 8.1%-16.0%; vs 11.5%; 95% CI, 7.5%-17.3%; RR, 0.91; 95% CI, 0.75-1.10; P = .34) or in trials of 6 months or less (short duration) of ADT (10.5%; 95% CI, 6.3%-17.0%; vs 10.3%; 95% CI, 8.2%-13.0%; RR, 1.00; 95% CI, 0.73-1.37; P = .99). Among 4805 patients from 11 trials with overall death data, ADT was associated with lower PCSM (443/2527 vs 552/2278 events; 13.5%; 95% CI, 8.8%-20.3%; vs 22.1%; 95% CI, 15.1%-31.1%; RR, 0.69; 95% CI, 0.56-0.84; P