Enhancing Prostate Cancer Diagnosis: The Combined Value of PHI and mpMRI.

Abstract

Background: Prostate cancer (PCa) diagnosis is often hindered by the need to detect clinically significant disease (csPCa) while minimizing unnecessary biopsies. The Prostate Health Index (PHI) and multiparametric magnetic resonance imaging (mpMRI) are promising tools to address these challenges.

Objective: To develop and internally validate a predictive model for PCa and csPCa by combining PHI and mpMRI in a high-risk population.

Methods: This retrospective study included 179 patients who underwent prostate biopsy between 2019 and 2023. Inclusion criteria comprised elevated PSA (> 3 ng/mL), suspicious digital rectal examination and/or family history, PHI values, and pre-biopsy mpMRI. Logistic regression models were developed, and model performance was assessed using C-statistics, calibration plots, and decision curve analysis (DCA).

Results: PCa was diagnosed in 40.2% of patients, and csPCa in 34.7% of them. A multivariate model including PHI, prostate volume, and mpMRI achieved an AUC of 0.81 for PCa. For csPCa, the best model combined PHI and prostate volume (AUC 0.76). In the PI-RADS 3 subgroup, PHI showed high discriminatory performance (AUC 0.81), surpassing PSA density (PSA-D). The DCA showed a superior net benefit of the multivariable models compared to single-parameter strategies.

Conclusion: Integrating PHI and mpMRI improves PCa diagnostic accuracy and clinical decision-making, especially in ambiguous cases such as PI-RADS 3 lesions, and reduces unnecessary biopsies in clinical practice.