Purpose: Using radical retropubic prostatectomy (RRP) specimens as the reference standard, we assessed whether PI-RADS 4–5 lesions in the transitional zone (TZ) and peripheral zone (PZ) differ in their association with clinically significant prostate cancer (csPCa), whether the composition of non-csPCa lesions differs by zone, and whether multiparametric MRI (mpMRI) lesion size predicts pathological upstaging. Materials and methods: Preoperative mpMRI data from 55 patients (108 lesions) undergoing RRP were retrospectively analyzed. A single experienced radiologist scored lesions using PI-RADS v2.1; each was correlated with the specimen by anatomical compartment. PI-RADS 4–5 lesions were classified as (A) csPCa (ISUP Grade Group ⩾ 2), (B) ISUP GG1, or (C) no neoplastic correlate within the corresponding compartment. Upstaging was defined as clinical T2a/T2b to surgical ⩾T2c, or clinical T2c to surgical ⩾T3a. Results: csPCa was identified in 25.9% (7/27) of TZ PI-RADS 4–5 lesions versus 70.0% (35/50) in the PZ ( p < 0.001). Maximum lesion diameter was larger in patients with upstaging (18.7 ± 7.6 vs 11.8 ± 2.8 mm; p = 0.0003); the AUC was 0.808 (95% CI 0.694–0.922) and the Youden-selected cut-off was ⩾13 mm (sensitivity 87.9%, specificity 61.9%, OR = 11.78). After adjustment for age and PSA, lesion size remained the only significant predictor (aOR = 1.294 per mm, p = 0.003), and the association persisted when the endpoint was restricted to extraprostatic extension (per 10 mm: OR = 4.96, 95% CI 1.90–16.92). Among non-csPCa lesions, 11 of 20 in the TZ had no neoplastic correlate versus 1 of 15 in the PZ ( p = 0.004). Conclusion: Larger mpMRI lesion size was associated with pathological upstaging, an association that persisted after adjustment and under a stricter endpoint; the ⩾13 mm threshold is exploratory pending external validation. Most TZ PI-RADS 4–5 lesions were non-csPCa, and lesions lacking a neoplastic correlate were more frequent in the TZ, although this comparison rested on a single such lesion in the PZ and is hypothesis-generating
No takes yet. Share an insight, caveat, or question.
İşler et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: