Objective To evaluate the association between the Prostate Health Index (PHI) and adverse pathology in a cohort of African American (AA) men undergoing radical prostatectomy. Materials and Methods Eighty AA men with prostate-specific antigen (PSA) of 2-10 ng/mL underwent measurement of PSA, free PSA (fPSA), and p2PSA prior to radical prostatectomy. PHI was calculated as [(p2PSA/fPSA) × (PSA)½]. Biomarker association with pT3 disease was assessed using logistic regression, and covariates were added to a baseline multivariable model including digital rectal examination. Biomarker ability to predict pT3 disease was measured using the area under the receiver operator characteristic curve. Results Sixteen men (20%) demonstrated pT3 disease on final pathology. Mean age, PSA, and %fPSA were similar in men with and without pT3 disease (all P > .05), whereas PHI was significantly greater in men with pT3 disease (mean 57.2 vs 46.6, P = .04). Addition of PHI to the baseline multivariable model improved discriminative ability by 12.9% (P =. .04) and yielded greater diagnostic accuracy than models, including other individual biomarkers. Conclusion In AA men with PSA of 2-10 ng/mL, PHI was predictive of pT3 prostate cancer and may help to identify men at increased risk of adverse pathology. Additional studies are needed to substantiate these findings and identify appropriate thresholds for clinical use.
ASJC Scopus subject areas