Clinical utility of the prostate health index density in patients with negative mpMRI or prior negative biopsy.

Background: Multiparametric MRI (mpMRI) has improved prostate cancer (PCa) detection; however, false-negative rates remain non-negligible, particularly among men with PI-RADS 1-2 lesions or those with persistently elevated PSA after a prior negative biopsy (PNB). The Prostate Health Index (PHI) and its volume-adjusted derivative, PHI density (PHID), have emerged as promising biomarkers for addressing these diagnostic gaps. This study evaluated the diagnostic performance of PHI and PHID, and examined whether incorporating age enhances PHID-based risk stratification in patients with negative mpMRI or PNB. Methods: This retrospective cohort included 261 men with PSA ≥4.0 ng/mL and either PI-RADS ≤2 mpMRI findings or PNB. Serum PSA, free PSA, and [-2]proPSA were measured, and PHI and PHID were calculated using standard formulas. All patients underwent transperineal biopsy within three months. Logistic regression and ROC analyses assessed predictors of clinically significant PCa. Age-stratified ROC analyses and bootstrap resampling (2,000 iterations) were performed to evaluate uncertainty in AUC and cutoff stability across the predefined age groups. Decision curve analysis (DCA) assessed clinical utility. Results: Among 117 biopsied men, 42 (35.9 %) had csPCa; in the mpMRI-negative subgroup, 28.8 % harbored csPCa. PHID demonstrated the strongest discriminatory ability (AUC 0.754; cutoff 1.568), and remained an independent predictor in multivariate analysis (OR 5.21). Incorporating age improved diagnostic performance (AUC 0.796; Conclusions: PHID is a robust marker for detecting csPCa in men with negative mpMRI or PNB, and age-adjusted PHID further improves diagnostic accuracy and clinical utility, although instability of cutoff values in younger men warrants caution.