Award Number : DAMD 17 - 03 - 1 - 0273 TITLE : Genes Involved in Oxidation and Prostate Cancer Progression

Background: Up to one third of early stage prostate cancer patients who underwent radical prostatectomy experienced disease progression. Different controls sampling methods applied in nested case-control studies, which searched for biomarkers that provide prognostic information and help understand disease progression mechanisms, have been adopted by researchers in different studies. Aims: Through a simulation, to compare the validity of the estimates of relative risk from three different controls sampling schemes. Methods: Hypothetical biomarker effects and disease progression probabilities were constructed by a stratified proportional hazard model. Three different controls sampling schemes, scheme 1 “incidence density sampling with replacement”, scheme 2 “incidence density sampling without replacement”, and scheme 3 “pure control”, were used and compared for their estimates of the true effect size. Results: Scheme 1 and 3 show unbiased and biased estimates of relative risk, respectively. Scheme 2 did not show evidence of bias because of few subjects have been selected as controls more than once. Conclusions: Nested case-control study, for searching biomarkers contributing to prostate cancer progression, within a defined cohort could efficiently provide valid estimate of relative risk with adequate sampling method like scheme 1.