Mounting Evidence for Prediagnostic Use of Statins in Reducing Risk of Lethal Prostate Cancer

Abstract

during follow-up, reporting a relative risk of 0.39 (95% CI, 0.19 to 0.77) among men who were statin users compared with nonusers. The emphasis on lethal prostate cancer in that study was important, given the considerable biologic heterogeneity in metastatic potential of prostate cancer over a man’s life and also because the etiology of lethal prostate cancer clearly differs from that of indolent disease. 3,4 A large fraction of men harbor latent prostate cancer,andwiththehighprevalenceofprostate-specificantigen(PSA) screening in Western populations, most of the diagnosed prostate cancers are indolent in the sense that they never would have otherwise come to clinical attention. Much of the apparent inconsistency among studies on statins and prostate cancer disappears when one distinguishes associations with risk of advanced-stage—or preferably lethal—prostate cancer from overall incidence. A meta-analysis of 27 observational studies publishedin2012reportedapooledrelativeriskof0.93(95%CI,0.87 to 0.99) for total incident prostate cancer and a more pronounced inverse association for advanced disease, with a relative risk of 0.80 (95% CI, 0.70 to 0.90) on the basis of seven studies. 5 One challenge in the interpretation of thefindings is a lack of consistency in the definition of advanced prostate cancer, with investigators using various definitions and often classifying T3a or T2 as advanced. From the perspectiveofetiologyandprevention,weproposethatlethalprostate cancer is the optimal disease end point for epidemiologic investigations. Given the long natural history of prostate cancer, such studies require long-term and complete follow-up of large cohorts of men. Since the publication of the 2012 meta-analysis, five additional epidemiologic studies have reported on associations between statin use and lethal prostate cancer, all suggesting inverse associations. UsingSEERdatainWashingtonstate,Geybelsetal 6 reportedarelative risk of 0.19 (95% CI, 0.06 to 0.56) on the basis of 39 lethal cases in a case-only analysis; thesefindings are notably in contrast with the null resultsforstatinsandPSAprogressioninthesamestudy.Thisisakey issue in prostate cancer research, given that PSA recurrence is an imperfect classifier of the lethal potential of a cancer. In a Danish population registry-based study, statin users had significantly lower rates of prostate cancer mortality, with a relative risk of 0.81 (95% CI, 0.75 to 0.88). 7 However, these results were adjusted only for age and nototherriskfactorsforlethalprostatecancer.Inastudyshowingthat postdiagnostic use of metformin was associated with substantially reducedprostatecancermortalityindiabetics,Margeletal 8 alsonoted