Mp49-13 Intercenter Variability in the Accuracy of Risk Calculators Based on Mri Parameters: Extensive External Validation of 7 Contemporary Risk Calculators

You have accessJournal of UrologyProstate Cancer: Detection however, MRI pathway is limited by high rate of false positive findings with high intercenter variability. The aim of this study was to assess the accuracy of seven risk calculators including MRI parameters in multiple heterogeneous cohorts. METHODS: We retrospectively reviewed the Prostate MRI Outcome Database (PROMOD) including all men with clinical suspicion of PCa undergoing prostate biopsy with a positive MRI (PIRADS≥3) at 24 institutions between 2013 and 2022. All patients underwent TBx+SBx. We included only patients who had positive MRI at target plus systematic biopsy. The probability of csPCa was computed using the logit function and coefficient of seven different Risk calculators: Radtke RC, vanLeeuwen RC, RAPID RC, PLUM RC, Mehralivand RC, Mount Sinai RC, IMPROD RC. Models were not recalibrated. The accuracy of each model in each institution was assessed using the area under the curve (AUC) and clinical benefit was assessed with decision curve analysis (DCA). RESULTS: We included 8391 patients from 24 institutions. There was a significant heterogeneity within each cohort in term of age, DRE, biopsy history, PSA values and PIRADS scores. CsPCa was detected in 3892 patients (45,6 %). The median prevalence of csPCa was 42% (IQR 29, 67). The best performing models were the PLUM RC (highest AUC in 8 institutions) and vanLeeuwen RC risk calculator (highest AUC in 5 institutions). All the RCs had a clinical benefit compared to the strategy of biopsy in all PIRADS≥3 above a probability of 15%. The vanLeeuwen RC had the highest net benefit in 10 institutions. CONCLUSIONS: We externally validated 7 risk calculators in 24 institutions and we found that all RCs had a clinical benefit compared to performing biopsy in every patients with a PIRADS≥3. However, given the high intercenter variability in their performance, retrospective validation of these model to assess their performance should be performed in each center before implementation in clinical practice. Source of Funding: None © 2024 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 211Issue 5SMay 2024Page: e788 Advertisement Copyright & Permissions© 2024 by American Urological Association Education and Research, Inc.Metrics Author Information Edoardo Tocci More articles by this author Ugo Falagario More articles by this author Anna Lantz More articles by this author Ivan Jambor More articles by this author Gian Maria Busetto More articles by this author Carlo Bettocchi More articles by this author Luigi Cormio More articles by this author Tobias Nordstrom More articles by this author Luigi Schips More articles by this author Paolo Gontero More articles by this author Luca Carmignani More articles by this author Pierluigi Bove More articles by this author Emanuele Montanari More articles by this author Carlo Trombetta More articles by this author Giuseppe Simone More articles by this author Vincenzo Ficarra More articles by this author Giuseppe Ludovico More articles by this author Vincenzo Mirone More articles by this author Alessandro Antonelli More articles by this author Francesco Porpiglia More articles by this author Pierfrancesco Bassi More articles by this author Peter J. Bostrom More articles by this author Ottavio De Cobelli More articles by this author Giuseppe Carrieri More articles by this author Expand All Advertisement PDF downloadLoading ...