Session: PD39: Prostate Cancer: Detection & Screening V
PD39-08: Inter-center variability in positive predictive values of PIRADS score in patients undergoing prostate biopsy: results from the PROMOD Study Group.
Introduction: Multiparametric magnetic resonance imaging (mp-MRI) of the prostate has a low specificity with up to 44% of PIRADS 4-5 lesions being falsely positive. Despite the effort of the PIRADS System Steering Committee to standardize parameters for image acquisition and radiology reports, mpMRI’s inter-reader and inter-center reproducibility remains moderate at best. The present study aims to evaluate inter-center variability of the positive predictive value (PPV) of MRI for the diagnosis of Prostate cancer (PCa). Methods: We retrospectively reviewed the PRostate Mri Outcome Database (PROMOD) including all man with clinical suspicion of PCa undergoing prostate biopsy with a positive MRI (PIRADS=3) in 24 European Institutions. The primary outcome of the study was csPCa defined as Biopsy ISUP Gleason Group (ISUP GG) =2. Overall and per institution PPVs were computed. Additionally, linear regression analysis was used to evaluate the association between PPV of PI-RADS scores with the csPCa prevalence as well as with proxy of csPCa prevalence such as median PSA density value and median ERSPC-Risk calculator predicted probability. Results: A total of 10185 patients from 24 institutions were identified from the PROMOD dataset. 8559 had a PI-RADS=3. Out of these, 5162 (60.3%) patients were diagnosed with PCa while 3585 (41.9%) were diagnosed with csPCa. PIRADS 3 median PPV was 18.1 (IQR: 13.6, 21.7), PIRADS 4 median PPV was 42.1 (IQR: 35.2, 49.8), PIRADS 5 median PPV was 73.6 (IQR: 63.7, 80.4). PPVs were significantly associated with center prevalence, median PSA density and median ERSPC-RC of each institution. Regression findings between PPV of each PIRADS score and characteristics of each cohort are graphically presented in Figure 1 Conclusions: Positive predictive value of MRI varies across institution. The variability is mainly due to differences in disease prevalence. Since disease prevalence is not available at the time of clinical decision, physician may use PSA density or Risk calculator to interpret mp-MRI results. The addition of these parameters to the PIRADS scoring system may help to define optimal strategies for selection of men who might benefit from prostate biopsy SOURCE OF Funding: None
