Abstract
OBJECTIVES: To provide concise information to clinicians on how to better interpret multiparametric magnetic resonance imaging for prostate cancer risk stratification. MATERIALS AND METHODS: We analyzed 2 separate cohorts. For patients receiving a Prostate Imaging Reporting and Data System version 2 (PI-RADSv2) score of 1 or 2, we reviewed the charts of 226 patients who underwent multiparametric magnetic resonance imaging of the prostate ordered from 2015 to 2017 to determine who developed clinically significant prostate cancer (csPCa) by August 27, 2020. For patients receiving PI-RADSv2 a score of 3, 4, or 5, we reviewed the results of 733 fusion biopsies on solitary lesions. Statistical analysis was used to further determine risk factors for csPCa. RESULTS: Ten percent of men with PI-RADSv2 a score of 1 eventually developed csPCa. Seven percent with a score of 2 were eventually diagnosed with csPCa. Only 1 of 226 with a score of 1 or 2 developed metastasis. For PI-RADSv2 scores of 3, 4, and 5, csPCa was detected in 16%, 45%, and 67% of fusion biopsies. Peripheral zone (PZ) PI-RADSv2 score of 4 or 5 and prostate-specific antigen density (PSA-D) were significant predictors of csPCa on multivariable analysis. Using a PSA-D × PI-RADSv2 score of ≤0.39, we identified 38% of men with a PI-RADSv2 score of 3 in the PZ or 3, 4, or 5 in the transition zone who could have avoided a benign biopsy. CONCLUSIONS: The vast majority of patients with PI-RADSv2 scores 1 and 2 can be safely monitored with close surveillance. Lesions with PI-RADSv2 scores of 4 and 5 in the PZ should be biopsied. Peripheral zone lesions with a PI-RADSv2 score of 3 and transition zone lesions with 3, 4, or 5 can be risk-stratified using the PSA-D × PI-RADSv2 score to determine who may safely avoid a biopsy and who should proceed to fusion biopsy.