Long-term Consequences of Finasteride vs Placebo in the Prostate Cancer Prevention Trial

• Published in: JNCI : Journal of the National Cancer Institute 2016-12, Vol.108 (12), p.djw168
• Main Authors: Unger, Joseph M, Till, Cathee, Thompson, Jr, Ian M, Tangen, Catherine M, Goodman, Phyllis J, Wright, Jason D, Barlow, William E, Ramsey, Scott D, Minasian, Lori M, Hershman, Dawn L
• Format: Article
• Language: English
• Subjects: 5-alpha Reductase Inhibitors - therapeutic use; Administrative Claims, Healthcare; Aged; Clinical Trials as Topic; Depression - epidemiology; Finasteride - therapeutic use; Follow-Up Studies; Humans; Lower Urinary Tract Symptoms - epidemiology; Male; Medical Record Linkage; Medicare; Middle Aged; Prostatic Neoplasms - prevention & control; Protective Factors; Risk Assessment; Risk Factors; Time Factors; United States - epidemiology
• ISSN: 0027-8874; 1460-2105
• DOI: 10.1093/jnci/djw168

Finasteride has been found to reduce the risk of low-grade prostate cancer but to have no impact on overall survival. The long-term adverse and beneficial consequences of finasteride have not been examined. We used a linkage between data from the Prostate Cancer Prevention Trial (PCPT) and Medicare claims. Patients were examined by randomized study arm (finasteride vs placebo for 7 years) for long-term consequences of the intervention, including cardiac, endocrine, and sexual dysfunction, depression, diabetes, and benign prostatic hyperplasia (BPH)-related events. To examine time to events, we used cumulative incidence and Cox regression, adjusting for covariates. All statistical tests were two-sided. A total of 13 935 of 18 880 participants (73.8%) in the PCPT were linked to Medicare claims, with median Medicare follow-up assessment time of 16 years from trial registration. There were no differences between finasteride and placebo participants with respect to important baseline factors or amount of Medicare follow-up assessment time. Finasteride patients had a 10% higher risk of new claims for depression (hazard ratio [HR] = 1.10, 95% confidence interval [CI] = 1.01 to 1.19, P = .04) and a 6% lower risk of procedures for BPH-related events (primarily lower urinary tract symptoms; HR = 0.94, 95% CI = 0.89 to 1.00, P = .03). No other differences were found in rates of long-term consequences of intervention in the two study arms. Finasteride use is associated with reduced need for procedures for relief of BPH-related events and a modest increase in depression. Overall, there is little need to worry about long-term noncancer consequences of finasteride use in those who use it for treatment of symptomatic BPH, hair growth, or prevention of cancer.