Development and Validation of the Diameter and Sinus Score, a Novel Model to Predict Trifecta Failure for Robot-Assisted Partial Nephrectomy: A Multicenter Retrospective Study.

Objectives: To evaluate and restratify each component of the RENAL score and develop a novel model to predict the difficulty of robot-assisted partial nephrectomy. Subjects and Methods: This retrospective multicenter study included 847 patients with localized renal-cell carcinoma, including 401 in the development cohort and 446 in the validation cohort. Multivariate logistic regression analysis was performed on the subdivided variable of the RENAL score to develop a novel model. The predictive performance of the novel model was evaluated by external validation using the receiver operating characteristic curve and area under the curve (AUC). Results: The trifecta achievement rates in the development and validation cohorts were 71.8% and 74.9%, respectively. Multivariate analysis of the development cohort revealed that the R and N components of the RENAL score were significantly associated with trifecta failure. The R score was restratified into three categories of diameter score (<30 mm, 0 point, 30-49 mm, 1 point; ≥50 mm, 2 points), and the N score was restratified into four categories of sinus score (not in contact, 0 point; touching the sinus, 1 point; simple protrusion, 2 points; complex protrusion, 3 points). The sum of the diameter and sinus (DS) scores was developed. In the external validation cohort, the AUC of the DS score was 0.73 (95% confidence interval [CI]: 0.68-0.78), higher than that of the RENAL score (0.68, 95% CI: 0.62-0.73, p = 0.048) and SPARE model (0.69, 95% CI: 0.64-0.75, 0.06). The interobserver concordance for the DS score using Cohen's weighted kappa was considered excellent (κ = 0.94 [95% CI: 0.92-0.95]). Conclusions: We developed a simple and novel model comprising only two components. External validation revealed that its predictive performance was better than that of the RENAL score.