Presentation Authors: Marco Roscigno*, Bergamo, Italy, Armando Stabile, Milan, Italy, Giovanni Lughezzani, Rozzano, Italy, Pietro Pepe, Catania, Italy, Lucio Dell'Atti, Ancona, Italy, Maria Nicolai, Giovanni La Croce, Michele Manica, Bergamo, Italy, Angelo Naselli, Milan, Italy, Giorgio Guazzoni, Rozzano, Italy, Alberto Briganti, Francesco Montorsi, Milan, Italy, Sandro Sironi, Luigi Filippo Da Pozzo, Bergamo, Italy
Introduction: We aim to evaluate whether multiparametric (mp) MRI alone could be used as a stand-alone test suggesting risk of reclassification in men on Active Surveillance (AS).
Methods: From 01/2016 to 09/2018, 340 pts underwent mpMRI before confirmatory/follow-up biopsy according to PRIAS protocol. Pts with negative (-) mpMRI received systematic random biopsy. Pts with positive (+) mpMRI (PI-RADS score =>3) underwent targeted biopsies (3 cores) + systematic random biopsies (12-18 cores). Multivariate logistic regression analyses (MVA) was used to create 3 model predicting the probability of disease reclassification (PCa GS =>3+4 at biopsy): a basic model including only clinical variables (age, PSAD and number (#) of positive cores at baseline); a MRI model including only PI-RADS score; a full model including both the previous ones. The predictive accuracy (PA) of each model was quantified using the AUC. The clinical net benefit deriving from the use of each model was assessed with decision curve analysis.
Results: Median patient age, PSA and PSAD were 67 yrs, 6.3 ng/ml, and 0.12 ng/ml/cm3. Median # of positive cores at initial biopsy was 1 (IQR:1,2).Â Eighty-four pts (24.7%) had mpMRI(-); 71 pts (20.9%), 146 (42.9%), and 39 (11.5%) had PI-RADS 3,4, and 5 lesions, respectively. At a median follow up of 12 months, 113 patients (33.2%) were reclassified. In pts with mpMRI(-) the rate of reclassification was 18%. while was 28%, 40% and 50% according to PI-RADS 3, 4 and 5, respectively. In the basic model, PSAD and # of positive cores at baseline were independent predictors of reclassification (p=0.001; OR 66.4 and p < 0.001; OR 2.2, respectively), with a PA of 69%. In the MRI model, PI-RADS 5 was predictor of reclassification (p=0.002; OR 4.7) and the PA was lower than in the basic model (AUC 62%). The full model had the best PA of 72%. PSAD (p=0.01; OR 28.6), # of positive cores at baseline (p < 0.001; OR 2.2) and PI-RADS 5 (p=0.02; OR 3.6) were predictors of reclassification. Fig.1 depicts clinical net benefit deriving from the use of the three evaluated models.
Conclusions: MRI alone should not be used in clinical practice as a stand-alone trigger for disease reclassification. The combination of MRI and other clinical variables still represents the most accurate approach to patients on AS.