Research Project
8527662
DESCRIPTION (provided by applicant): Despite major investment by the National Institutes of Health into clinical trials on the subject of the chemoprevention of prostate cancer (prCA), unanswered questions remain regarding finasteride, a 5-alpha reductase inhibitor currently used for benign prostatic hyperplasia that may also protect against prCA.6 Specifically, concerns about the potential for finasteride to cause a more severe form of prCA, when it does occur, have prevented the uptake of the drug for chemoprevention since release of the PrCA Prevention Trial (PCPT) results in 2003.3, 6, 7 There are newly-emerging concerns from national drug-safety agencies regarding finasteride's potential to cause male breast cancer.5, 8 Another drug class, statins, is of chemopreventive interest and hypothesized to have pleiotropic effects affecting cancer initiation or progression. In four large prospective studies, statins were found to reduce the risk of advanced but not overall prCA.9 Recently, investigators have raised the possibility that the lipid-lowering action of statins (but not the statins themselves) may be the etiologic mechanism preventing severe prCA.12 If this hypothesis is true, all lipid-lowering drugs would affect severe prCA risk, not only statins. The research intent of this application is to cost-efficiently use a unique existing resource, the Saskatchewan Health Services Databases, in order to fill these knowledge gaps regarding the chemopreventive potential of these currently marketed drugs by conducting retrospective, pharmacoepidemiologic cohort studies of prostate and breast cancer outcomes nested in these databases. This study would inform clinical practice by providing longitudinal evidence complementary to clinical trials of drug benefit or harm in a 'real-world' population of users. The provincial government of Saskatchewan, Canada provides universal health coverage to all residents, and existing administrative data contain information on ~220,000 males from a defined geographic area with 15+ years of observation time and high-quality cancer information available from a provincial cancer registry.59, 60, 120 Drug exposure information is comprehensively captured in electronic prescribing records. We plan to use a matched cohort design and longitudinal analysis methods to estimate the risk of overall and advanced stage and high-grade prCA incidence, mortality and case-fatality among male users of 1) finasteride, 2) statins, and 3) non-statin lipid-lowering medications, compared to non-users. For finasteride, we will also study breast cancer incidence to inform the drug's safety. The diverse expertise of the research team includes pharmacoepidemiology, biostatistics, urology, and oncology. The proposed PI has a demonstrated history of manuscript productivity relevant to the publication plans of the project and is poised to gain experience with the databases through another funded project. The public health relevance of the proposal is substantial given the common occurrence of prCA and the urgent need for more information regarding high-grade prCAs and breast cancers potentially caused by 5-ARIs.
