LAUSR

PURPOSE To examine interdisciplinary variability using two established preoperative nephrometry scores for prediction of conversion to nephrectomy (cNE) in patients with renal mass scheduled for partial nephrectomy (PN). MATERIALS AND METHODS 229 consecutive PN candidates were included at a single institution between January 2013 and May 2017. Patient, tumor and treatment characteristics were assessed. PADUA and RENAL scores were independently calculated by board-certified radiologists and urologic residents using computed tomography or magnetic resonance imaging. Statistical analyses (κ-analyses, ROC curve analyses, univariable/multivariable binary logistic regression analyses) were conducted. RESULTS PN was performed in 198/229 (86.5%) patients, whereas 31/229 (13.5%) patients were converted to nephrectomy. The prevalent tumor stage was pT1a (94/229,41.1%). The predominant histologic entity was clear cell carcinoma (128/229,55.9%). Interdisciplinary comparison (radiologist/urologist) of PADUA and RENAL score revealed a κappa of 0.40 and 0.56, respectively. ROC curve analyses demonstrated a higher AUC predicting cNE using the PADUA (uro: 0.79,rad: 0.782) compared to the RENAL score (uro: 0.731,rad: 0.766). Using a cutoff of ≥10 the PADUAUro had a sensitivity of 81%, a specifity of 71% and was independently associated with cNE (OR 10.98;p<0.001). CONCLUSIONS Our results indicate a higher prediction of cNE by PADUA when compared to RENAL score. Calculation of the PADUA and the RENAL score by physicians without specialized radiologic training is feasible and might achieve comparable results in predicting cNE as compared to the gold standard provided by board-certified radiologists. This information is helpful if nephrometry scores are not regularly included in the radiologic report.