The Simplified PADUA REnal (SPARE) nephrometry system: a novel classification of parenchymal renal tumours suitable for partial nephrectomy

Objective To simplify the original Preoperative Aspects and Dimensions Used for an Anatomical (PADUA) classification of renal tumours, generating a new system able to predict equally or better the risk of overall complications in patients undergoing partial nephrectomy (PN); and to test if the addition of the contact surface area (CSA) parameter improves the accuracy of the original PADUA and new Simplified PADUA REnal (SPARE) nephrometry classification systems. Patients and Methods We analysed the clinical records of 531 patients who underwent PN (open, laparoscopic and robot-assisted) for renal tumours at five tertiary academic referral centres from January 2014 to December 2016. The ability of each variable included in the PADUA classification to predict overall complications was tested using binary logistic regression analysis. The variables that were not statistically significant were excluded from the SPARE classification. In addition to the original PADUA and SPARE systems, another two models were generated adding tumour CSA. Receiver operating characteristic curve analysis was used to compare the ability of the four different models to predict overall complications. Binary logistic regression was used to perform both univariable and multivariable analyses looking for predictors of postoperative complications. Linear regression analysis was used to identify independent predictors of absolute change in estimated glomerular filtration rate (eGFR; ACE). Results The SPARE nephrometry score system including: (i) rim location, (ii) renal sinus involvement, (iii) exophytic rate, and (iv) tumour dimension; showed equal performance in comparison with the original PADUA score (area under the curve [AUC] 0.657 vs 0.664). Adding tumour CSA to the original PADUA (AUC 0.661) or to the SPARE (AUC 0.658) scores did not increase the accuracy of either system to predict overall complications. The SPARE system (odds ratio 1.2, 95% confidence interval 1.1-1.3) was an independent predictor of postoperative overall complications. Age (P < 0.001), body mass index (P < 0.001), Charlson Comorbidity Index (P = 0.02), preoperative eGFR (P < 0.001), and tumour CSA (P = 0.005) were independent predictors of ACE. Limitations include the retrospective design and the lack of central imaging review. Conclusions The new SPARE score is comprised of only four variables instead of the original six and its accuracy to predict overall complications is similar to that of the original PADUA score. Addition of tumour CSA was not associated with an increase in prognostic accuracy. The SPARE system could replace the original PADUA score to evaluate the complexity of tumours suitable for PN.