Early diagnosis of prostate cancer has always been a major concern. The Prostate Health Index (PHI), integrating total prostate-specific antigen (PSA), free PSA, and [-2]proPSA, is proposed to enhance specificity in detecting clinically significant prostate cancer (csPCa) compared to conventional PSA metrics. This systematic review and meta-analysis aims to evaluate the diagnostic accuracy of PHI for predicting csPCa.

A systematic search of PubMed, Embase, Web of Science and Cochrane Library (Up to March 1, 2025) identified studies reporting PHI’s diagnostic performance against histopathological confirmation of csPCa (Grade Group≥2 or ISUP ≥2). Pooled sensitivity, specificity, and hierarchical summary receiver operating characteristic curves were derived using bivariate random-effects models. Study quality was assessed via QUADAS-2. Subgroup analyses explored heterogeneity across patients‘ PSA range and continent.

Fifty-nine studies including 23019 patients were screened out in this study, PHI demonstrated a pooled sensitivity of 81% (95% CI: [78%-84%] ) and specificity of 62% (95% CI: [58%–67%]) for csPCa detection, with an AUC of 79% (95% CI [75%-82%]). In the subgroup analysis, Asia, Europe and North America showed different diagnostic accuracy with an AUC of 81% (95% CI: [77%–84%]),76% (95% CI: [72%–80%]) and 74% (95% CI: [70%-78%]).

Compared with previous studies, PHI exhibits superior specificity to traditional PSA parameters in identifying csPCa, potentially reducing unnecessary biopsies in patients with elevated PSA. However, threshold standardization and validation in diverse clinical settings are needed to optimize its integration into diagnostic pathways.