External Validation of the Arterial-Based Complexity Score and First Head-to-Head Comparison With the R.E.N.A.L. and PADUA Scores and C-index

The aim of the present study was to prove the arterial-based complexity (ABC) score validity by comparing it with the R.E.N.A.L. (radius, exophytic/endophytic tumor properties, nearness of tumor to deepest portion of collecting system or sinus, anterior/posterior descriptor, location relative to the polar line), PADUA (preoperative aspects and dimension for anatomic classification of renal tumors), and C-index scores. We performed a retrospective analysis of pre- and postoperative data from 234 patients who had undergone open and robot-assisted partial nephrectomy. An external urologist who was unaware of the outcomes reviewed all computed tomography scans to assign the nephrometry scores and determine tumor complexity. We found no statistically significant superiority for the ABC system. Introduction: We performed an external validation of the arterial-based complexity (ABC) score using a head-to-head comparison with the R.E.N.A.L. (radius, exophytic/endophytic tumor properties, nearness of tumor to deepest portion of collecting system or sinus, anterior/posterior descriptor, location relative to the polar line), PADUA (preoperative aspects and dimension for anatomic classification of renal tumors), and C-index scores for the prediction of surgical outcomes after partial nephrectomy. Materials and Methods: The data from a series of consecutive open or robot-assisted partial nephrectomies performed from January 2014 to July 2016 by 4 expert surgeons at a tertiary academic institution were reviewed. After dedicated training, 1 urologist not involved in the surgical procedures evaluated the cross-sectional imaging studies and assigned the nephrometry score using the 4 nephrometry scoring systems. The predictive performance of the ABC and other scoring systems was tested in univariate and multivariable fashion. Results: Overall, 234 patients were recruited (148 men and 86 women; age, 63 ± 10.9 years). The scores were all related to the estimated blood loss, use of hilar clamping, ischemia time, operative time, length of stay, and MIC (margin status, ischemia time, complications) score. They were not related to the occurrence of postoperative complications or, for the C-index and ABC score, the length of stay. In a head-to-head comparison, the ABC was not inferior only to the C-index relative to the occurrence of complications and MIC score, with borderline statistical significance. On multivariate analysis, the ABC score provided significant improvement only for the prediction of the operative and ischemia times. However, its performance was inferior to that of the other scoring systems. In addition, only the PADUA score improved the prediction of artery clamping and MIC score, and only the R.E.N.A.L. score showed an advantage for the prediction of the estimated blood loss. Conclusion: The predictive ability of ABC was inferior to that of well-established existing nephrometry scoring systems, such as the PADUA and R.E.N.A.L. scores.