Do the Benefits of Finasteride Outweigh the Risks in the Prostate Cancer Prevention Trial? Steven Grover,* Ilka Lowensteyn, David Hajek, John Trachtenberg, Louis Coupal and Sylvie Marchand From the Centre for the Analysis of Cost-Effective Care and the Divisions of General Internal Medicine and Clinical Epidemiology, The Montreal General Hospital, Departments of Medicine and Epidemiology & Biostatistics, McGill University, Montreal, Quebec (SG, IL, LC, SM), and the Department of Surgical Oncology/Urology, The Princess Margaret Hospital, University Health Network (DH, JT), and the Department of Surgery/Urology, University of Toronto, Toronto, Ontario, Canada Purpose: The Prostate Cancer Prevention Trial demonstrated that finasteride could reduce the incidence of prostate cancer by 25%. However, its use was also associated with an increased risk of high grade cancer resulting in uncertainty surrounding the net benefits of therapy. Materials and Methods: We used the Montreal Prostate Cancer Model, a validated Markov model of prostate cancer progression, to compare the forecasted survival in treated and untreated men. The conditions of the model were varied to reflect different assumptions about whether the cancer grade difference observed in the PCPT was real or a treatment associated artifact, and whether cancers detected on end of study biopsies were clinically significant. Results: For a hypothetical cohort of 1,000, 62-year-old men treated with finasteride, an increased survival of 140 life-years (0.14 years per individual) is forecasted if all diagnosed cancers are considered. If tumor grade differences are held to be artifactual, the forecasted benefits increase to 200 life-years. However, if the tumor grade difference is real and only clinically detected cancers are considered, estimated increased survival is only 20 life-years (0.02 years per individual). Conclusions: The primary prevention of prostate cancer with finasteride looks promising. However, at the present time it should only be considered with caution until we have answered critical questions surrounding the difference in cancer grade observed in the PCPT and the clinical significance of cancers detected on protocol directed end of study biopsies. Key Words: finasteride, prostatic neoplasms, prevention and control, risk assessment P rostate cancer is the most commonly diagnosed vis- ceral malignancy in the United States and ranks only behind lung cancer as a cause of cancer death. 1 In the United States the lifetime risk of carcinoma of the prostate developing is approximately 16% and the lifetime risk of death from the disease is 3% to 4%. 2 The development of prostate cancer is characterized by the appearance of latent, low grade cancers in the third and fourth decades of life and, in a minority of men, slow pro- gression over decades to clinically significant invasive dis- ease. These characteristics make prostate cancer a viable target for primary chemoprevention, and strategies using dietary constituents, vitamins, trace minerals and hormonal agents have been explored. 3 Finasteride is a potent inhibitor of 5-reductase type 2, which catalyzes the conversion of testosterone to dihydrotes- tosterone, the most biologically active androgen, within the prostate. Previous studies demonstrated the efficacy of fin- asteride in relieving symptoms and preventing progression of benign prostatic hyperplasia. 4,5 While there is no direct evidence tying increased testosterone levels to prostate can- cer, 3 several observations suggested the hypothesis that fin- asteride might delay or prevent the development of the disease. Notably, Japanese men, who have a low risk of prostate cancer, have significantly lower levels of 5-reduc- tase activity than men in higher risk populations. 6 The PCPT was a multicenter randomized placebo con- trolled trial that tested the hypothesis that finasteride would reduce the prevalence of prostate cancer. After 7 years of treatment, cancer prevalence was almost 25% lower among treated men than among controls. However, the risk of a high grade tumor developing (Gleason score 7 or greater) was significantly higher among men treated with finasteride (RR 1.67). 7 High grade tumors are much more likely than low grade tumors to progress to metastatic dis- ease and prostate cancer death. 8 Accordingly the overall impact of primary prevention with finasteride on prostate cancer morbidity or mortality remains unclear. Formal risk-benefit analysis can be used to help answer complex clinical questions when outcomes of clinical trials do not themselves permit clear conclusions. For example, mathematical modeling has been used to predict the impact Submitted for publication April 21, 2005. Nothing to disclose. * Correspondence: Division of Clinical Epidemiology, The Mon- treal General Hospital, 1650 Cedar Ave., Montreal, Quebec H3G 1A4 Canada (telephone: 514-934-1934; FAX: 514-934-8293; e-mail: nbouchard@epimgh.mcgill.ca). Editor’s Note: This article is the fourth of 5 published in this issue for which category 1 CME credits can be earned. Instructions for obtaining credits are given with the questions on pages 1178 and 1179. 0022-5347/06/1753-0934/0 Vol. 175, 934-938, March 2006 THE JOURNAL OF UROLOGY ® Printed in U.S.A. Copyright © 2006 by AMERICAN UROLOGICAL ASSOCIATION DOI:10.1016/S0022-5347(05)00424-6 934