Introduction: African American (AA) race is a well-known risk factor for prostate cancer (PCa). However, the aggressive push to maximize detection of clinically significant cancer (CSC) is inevitably prone to high rates of over-biopsy. Prostate Health Index (PHI) is a Food and Drug Administration-approved tool to aid in the diagnosis of PCa in men with PSA 4-10 ng/ml. We sought to assess the effectiveness of PHI to avoid a prostate biopsy in AA men with an equivocal PSA.
Methods: Within the Duke Health System, AA men undergoing prostate biopsy with a PSA 4-10 ng/ml and PHI scores from 2016-2020 were identified. The cohort was analyzed for a PHI score cutoff for which 10% or less of men would be identified with CSC, defined as Gleason score =7. Representative 95% confidence intervals (CI) were calculated along with cancer detection rates. ROC curves for all PCa and CSC at biopsy were used to identify the area under the curve (AUC) and the diagnostic value of the proposed cutoff point.
Results: A total of 236 AA men with an average age of 61.2 years met inclusion criteria out of the 1,131 men who had PHI testing during the study period. Mean PHI and PSA for the entire cohort were 48.5 (SD 21.6) and 6.5 ng/ml (SD 1.7), respectively. Using the established lowest PHI cutoff of 27 led to six of 25 (24.0%) men diagnosed with CSC and 12 of 25 (48%) diagnosed with all PCa. Below a proposed PHI level of 23, one of 12 men were diagnosed with CSC (8.3% [CI: 0.002 – 0.385]) and two of 12 men were diagnosed with any PCa (16.7% [CI: 0.021 – 0.484]). Overall, PHI showed an AUC for all PCa and CSC of 0.705 (0.636 – 0.774) and 0.688 (0.621 – 0.755), respectively. PHI cut point of 23 was associated with a 98.7% sensitivity for all PCa and 99.0% for CSC, at a cost of specificities of 11.8% and 8.4%, respectively. Using this cut point 11 negative biopsies could be avoided, equating to 4.7% of the cohort.
Conclusions: The currently accepted lowest PHI cut point leads to an excessively high rate of PCa in this cohort of AA men. Alternatively, men with a PHI below a proposed cutoff of 23 had optimally low rates of cancer detection and excellent sensitivity to minimize false negative biopsies. While our cohort is the largest reported of AA men undergoing PHI, further validation is needed to confirm these findings. Though generally at higher risk, AA men with an equivocal PSA and a PHI below 23 should be considered candidates to avoid a prostate biopsy.