Which men with non-malignant pathology at magnetic resonance imaging-targeted prostate biopsy and persistent PI-RADS 3-5 lesions should repeat biopsy?

Purpose: To assess predictors of clinically significant (cs) prostate cancer (PCa) in men who had a non-malignant Multiparametric mag- netic resonance imaging (mpMRI)-targeted biopsy and persistent Prostate Imaging-Reporting Data System (PI-RADS) 3 to 5 lesions in sub- sequent mpMRI. Materials and Methods: We retrospectively analyzed MRI-targeted biopsy database in three centers. Inclusion criteria: persistence of at least one PI-RADS ≥3 lesion found negative for cancer in a previous MRI-targeted plus systemic biopsy (baseline biopsy). Exclusion crite- ria: downgrading to PI-RADS 1-2. A logistic regression analysis was performed to estimate the predictors of csPCa. Results: Fifty-seven patients were included. Median interval between biopsies was 12.9(2.43) months. Median age was 68.0(12) years. Median PSA was 7.0(5.45) ng/ml. At follow-up, 24.6%, 54.4%, and 21% of patients had a PI-RADS score 3, 4, and 5 index lesion (IL), respectively. At re-biopsy, 28/57(49.1%) men were found to harbor PCa. Among these, 22(78.6%) had csPCa. csPCa was found outside the IL in only 2 patients. Eleven, 13, and 5 patients with PI-RADS 3, 4, and 5, respectively, had no cancer. Three patients with a PI-RADS 3 lesion had cancer (2 with Gleason score 3+3, 1 with Gleason score 3+4). 14/43 men with a PI-RADS 4/5 lesion harbored Gleason score ≥3 +4 PCa. Logistic regression analysis found that PSA (HR 1.281, 95% CI: 1.013−1.619, P = 0.039) and IL size (HR 1.146, 95% CI: 1.018 −1.268, P = 0.041) were the predictors of csPCa at re-biopsy. Conclusions: Patients with non-malignant pathology from PI-RADS ≥3 lesions targeted biopsy should be follow-up with mpMRI, and those with persistent PI-RADS 4 to 5 lesions should repeat MRI-targeted and systematic biopsy.