Incremental Value of PI-RADS v2.1 for Adverse Pathology at Radical Prostatectomy: A Retrospective 1.5-T MRI Cohort

The incremental value of PI-RADS for predicting adverse pathology at radical prostatectomy in men diagnosed through systematic biopsy remains uncertain. We assessed whether PI-RADS improves discrimination beyond age, MRI-derived prostate-specificantigen density (PSAD), biopsy Grade Group, and quantitative biopsy tumour burden. This retrospective study included 284 men undergoing 1.5-T mpMRI, systematic biopsy, and radical prostatectomy between January 2020 and January 2026. Adverse pathology comprised prostatectomy Grade Group ≥ 3 and/or stage ≥ pT3a. Three logistic models were compared: age/PSAD/biopsy grade/tumour burden; age/PSAD/PI-RADS; and their combination. Internal validation used 2000 bootstrap samples. Adverse pathology occurred in 143 patients (50.4%). Apparent AUCs were 0.790 (95% CI 0.737-0.843), 0.713 (0.653-0.773), and 0.799 (0.747-0.851); optimism-corrected AUCs were 0.781, 0.703, and 0.787. Adding PI-RADS ≥ 4 yielded an adjusted OR of 1.996 (95% CI 1.150-3.465) and an apparent AUC difference of 0.009 (95% CI -0.015 to 0.033; descriptive DeLong p = 0.469). The combined model had a corrected calibration slope of 0.915 and Brier score of 0.187. PI-RADS from routine 1.5-T mpMRI was independently associated with adverse pathology, but its incremental discrimination beyond clinical and systematicbiopsy variables was small and uncertain. These findings concern a surgery-selected systematicbiopsy pathway, do not estimate the overall value of MRI, and require external validation.