Racial variation in the reliability of prostate cancer indicators in men undergoing subsequent prostate biopsy.

115 Background: Many men with an initial negative prostate biopsy have a persistently elevated prostate specific antigen (PSA) prompting physicians to perform repeat biopsies. African American men (AA) are at particular risk as they have a greater prostate cancer (PCa) incidence and mortality, yet traditional PSA testing may be less reliable in this cohort. We sought to determine the predictors of PCa and PCa severity in a racially diverse population on subsequent biopsy following an initial benign biopsy. Methods: Upon receiving Institutional Review Board approval, a retrospective analysis was performed on men with repeat prostate biopsies at Tulane Medical Center and Southeast Louisiana Veterans Health Care Services in New Orleans, Louisiana from 2003-2015. Inclusion criteria included patients with a benign initial prostate biopsy and underwent subsequent repeat prostate biopsy within 5 years. Race, age, serum PSA, PSA density (PSAD), and prostate volume by transrectal ultrasound (TRUS) were evaluated to determine if they correlate with the presence and severity of PCa. Aggressive PCa was defined as Gleason score >6. Results: A total of 209 men were included; 127 (61%) were AA, and 82 (39%) were Caucasian American men (CA). The two groups were similar with respect to PSA, PSAD, and TRUS. More AA (25.2% vs. 17.1%) had a repeat biopsy showing any PCa. Of those with PCa, 28.1% of AA and 28.6% of CA had aggressive PCa. PSAD positively correlated with any PCa (p=.015). TRUS negatively correlated with any PCa (p=.008). PSA levels (p<.001) and PSAD (p<.001) positively correlated with aggressive PCa in CA but not in AA. Conclusions: The goal of prostate biopsy particularly for those with a prior negative biopsy is to detect aggressive PCa. PSA and PSAD positively correlated with finding any PCa in both AA and CA but not with aggressive PCa in AA. PSA and PSAD are less accurate predictors of aggressive PCa in AA, and novel biomarkers are needed. [Table: see text]