Francesco Barletta, Jean Baptiste Beauval, Pierre I. Karakiewicz, Alberto Briganti, Zhe Tian, Francesco Porpiglia, Guillaume Ploussard, Mathieu Roumiguié, Arnas Rakauskas, Giorgio Gandaglia, Stefania Zamboni, Armando Stabile, Bernard Malavaud, Massimo Valerio, Cristian Fiori, Francesco Montorsi, Agostino Mattei, Nicola Fossati, Simone Scuderi, Marco Moschini, Francesco De Cobelli, Gandaglia, G., Ploussard, G., Valerio, M., Mattei, A., Fiori, C., Roumiguie, M., Fossati, N., Stabile, A., Beauval, J. -B., Malavaud, B., Scuderi, S., Barletta, F., Moschini, M., Zamboni, S., Rakauskas, A., Tian, Z., Karakiewicz, P. I., De Cobelli, F., Porpiglia, F., Montorsi, F., and Briganti, A.
Background The combined role of multiparametric magnetic resonance imaging (mp-MRI), and magnetic resonance imaging (MRI)-targeted and concomitant systematic biopsies in the identification of prostate cancer (PCa) patients at a higher risk of adverse pathology at radical prostatectomy (RP) is still unclear. Objective To develop novel models to predict extracapsular extension (ECE), seminal vesicle invasion (SVI), or upgrading in patients diagnosed with MRI-targeted and concomitant systematic biopsies. Design, setting, and participants We included 614 men with clinical stage≤T2 at digital rectal examination who underwent MRI-targeted biopsy with concomitant systematic biopsy. Outcome measurements and statistical analyses Logistic regression analyses predicting ECE, SVI, and upgrading (ie, a shift from biopsy International Society of Urological Pathology grade group to any higher grade at RP) based on clinical variables with or without mp-MRI features and systematic biopsy information (the percentage of cores with grade group ≥2 PCa) were developed and internally validated. The area under the curve (AUC) was used to identify the models with the highest discrimination. Decision-curve analyses (DCAs) determined the net benefit associated with their use. Results and limitations Overall, 333 (54%), 88 (14%), and 169 (27%) patients had ECE, SVI, and upgrading at RP, respectively. The inclusion of mp-MRI data improved the discrimination of clinical models for ECE (67% vs 70%) and SVI (74% vs 76%). Models including mp-MRI, and MRI-targeted and concomitant systematic biopsy information achieved the highest AUC at internal validation for ECE (73%), SVI (81%), and upgrading (73%) and represented the basis for three risk calculators that yield the highest net benefit at DCA. Conclusions Not only mp-MRI and MRI-targeted sampling, but also concomitant systematic biopsies provide significant information to identify patients at a higher risk of adverse pathology. Although omitting systematic prostate sampling at the time of MRI-targeted biopsy might be associated with a reduced risk of detecting insignificant PCa and lower patient discomfort, it reduces the ability to accurately predict pathological features. Patient summary The combination of multiparametric magnetic resonance imaging (mp-MRI) with accurate biopsy information on MRI-targeted and systematic biopsies improves the accuracy of multivariable models based on clinical and mp-MRI data alone. Correct mp-MRI interpretation and proper extensive prostate sampling are both needed to predict adverse pathology accurately at radical prostatectomy.