Are African American men adequately represented in prostate cancer randomized controlled trials? A systematic review.

Abstract

6066 Background: African American (AA) men bear a disproportionate burden of incidence and mortality from prostate cancer compared with other racial groups (ACS Report, 2009). The clinical applicability (i.e., external validity) of results of randomized controlled trials (RCTs) depends on whether the sample of patients represents the disease population (NCI Report, 2004). Underrepresentation of AAs in clinical trials could impact the external validity of their findings leading to uncertainty about applicability of results to AA patients. A systematic assessment of participation of AA men in prostate cancer RCTs has not been done. Methods: We performed an electronic search of Medline and Cochrane databases to identify all prostate cancer RCTs for the years 2005-2009. Studies were included if they were phase III trials conducted in North America and evaluated prostate cancer prevention, screening, or treatment. Data were extracted on patient characteristics, interventions, controls, and outcomes. Dual extraction was undertaken for quality assurance. Meta-analysis using fixed effects model was performed to assess overall outcomes according to AA participation. Results: Our initial search identified 1206 citations of which 20 (29 comparisons) met the inclusion criteria. Overall, the methodological quality of included studies was high. Only 55% (11/20) of studies reported data on ethnic composition. Among those, 1 reported 30%, 4 reported 10-20%, and 6 reported <10% AA enrollment. Nine studies reported overall survival data. Pooling data according to the distribution of experimental versus standard treatments by participation of AA men showed no difference. The pooled hazard ratio in trials with >10% participation was 0.94 (95% CI; 0.87, 1.02), 0.97 (95% CI; 0.92, 1.03) in trials with <10% participation and 0.92 (95% CI; 0.80, 1.07) in trials with no reporting of ethnicity data. Conclusions: The results show AA men are underrepresented in prostate cancer RCTs in contrast to their disproportionate disease burden. Despite the conventional belief that AA men have poorer outcomes compared with other racial groups, our results show no difference in outcomes in prostate cancer RCTs.