Prostate cancer, the second most prevalent malignancy in males, imposes a significant global health burden. Early and accurate diagnosis is pivotal for improving therapeutic outcomes and prognosis. Conventional screening methods such as prostate-specific antigen (PSA) testing suffer from limitations including limited specificity and suboptimal accuracy, failing to meet clinical demands. The Prostate Health Index (PHI), first clinically approved in 2012 as an adjunct for prostate cancer diagnosis, has demonstrated superior diagnostic performance compared to PSA. However, large-scale Asian population data validating the clinical utility of PHI in this demographic remain scarce, hindering its widespread clinical adoption.

We retrospectively analyzed clinical diagnostic data from patients with prostate cancer (PCa) or benign prostatic hyperplasia (BPH) treated at Tongji Hospital in Wuhan between January 2022 and December 2024. Participants who underwent PSA, PHI testing, and biopsy were included, while exclusion criteria encompassed concurrent infectious diseases or other malignancies. Demographic and clinical parameters—including age, total PSA (tPSA), free PSA percentage (fPSA%), PHI, PI-RADS scores, prostate volume, and histopathological results—were extracted. Receiver operating characteristic (ROC) curves, decision curve analysis (DCA), and regression analyses were performed to evaluate diagnostic performance metrics and clinical utility across methodologies.

A total of 2,091 patients were analyzed, including 724 PCa cases (83.8% were clinically significant PCa [csPCa]) and 720 MRI-evaluated individuals. ROC analysis revealed PHI’s superior diagnostic efficacy for PCa (AUC=0.854) over tPSA, fPSA%, and [-2]proPSA (p2PSA). PHI consistently outperformed other biomarkers in csPCa (AUC=0.879) and PSA gray zone (4–10 ng/mL) subgroups (AUC=0.777), with statistical significance (P<0.001). At a cutoff value of 47.15, PHI exhibited optimal diagnostic performance for csPCa (sensitivity=81.05%, specificity=82.61%). Regression analysis in the MRI cohort identified PHI and PI-RADS as independent diagnostic markers, with their combined model achieving an AUC of 0.938. DCA further demonstrated that PHI-PI-RADS integration provided a more reliable determinant for biopsy decision-making compared to individual parameters.

This study presents novel evidence supporting PHI’s diagnostic value in a Chinese population. PHI consistently outperformed conventional PSA across diverse clinical scenarios, reinforcing its utility for PCa detection. Notably, Asian patients exhibited higher PHI diagnostic thresholds compared to Western populations, highlighting the necessity of population-specific diagnostic criteria. Ultimately, the combination of PHI and PI-RADS scoring emerges as a promising strategy for precise prostate cancer diagnosis and biopsy necessity assessment.