OBJECTIVE. The purpose of this study was to analyze the utility of the Prostate Imaging Reporting and Data System version 2 (PI-RADSv2) in the detection of a clinically significant cancers in patients with prostate cancers with a biopsy Gleason score of 6. MATERIALS AND METHODS. A group of 182 consecutively registered patients with biopsy-proven prostate cancer with a Gleason score of 6 underwent MRI and radical prostatectomy. Clinically significant cancer was surgically defined as Gleason score of 7 or greater, tumor volume of 0.5 cm3 or greater, or tumor category T3 or greater. Clinical parameters (prostate-specific antigen level, greatest percentage of biopsy core, and percentage of positive cores) and the PI-RADSv2 ratings by three independent readers (experienced readers 1 and 2, inexperienced reader 3) were investigated. Cutoffs and the diagnostic performance of PI-RADSv2 for clinically significant cancer were analyzed. RESULTS. Clinically significant cancer was found in 87.4% (159/182) of patients. The cutoff PI-RADSv2 score for clinically significant cancer was 4 for readers 1 and 2 and 5 for reader 3. The AUCs were 0.829 and 0.853 for readers 1 and 2 (p < 0.001) and 0.602 for reader 3 (p = 0.067). For reader 1, sensitivity was 89.9% (143/159); specificity, 69.6% (16/23); positive predictive value, 95.3% (143/150); negative predictive value, 50.0% (16/32); and accuracy, 87.4% (159/182). The corresponding values for reader 2 were 81.1% (129/159), 82.6% (19/23), 97.0% (129/133), 38.8% (19/49), and 81.3% (148/182). For the experienced readers, 66.7-81.3% of patients with false-negative results had clinically significant cancers with tumor volume less than 1 cm3. CONCLUSION. PI-RADSv2 may help experienced readers identify clinically significant prostate cancers in patients with a biopsy Gleason score of 6. However, some small (< 1 cm3) clinically significant cancers can be missed when PI-RADSv2 is used.
All Science Journal Classification (ASJC) codes
- Radiology Nuclear Medicine and imaging