PD51-09: The Cancer of the Bladder Risk Assessment (COBRA) score for Estimating Cancer-Specific Survival after Radical Cystectomy: External Validation in a Large Bi-institutional European Cohort
Friday, May 15, 2020
7:00 AM – 9:00 AM
Tim Muilwijk, Murat Akand, Francesco Soria, Andrea Giordano, Uros Milenkovic, Lisa Moris, Liesbeth Demaegd, Gaetan Devos, Eduard Roussel, Thomas Gevaert, Hendrik Van Poppel, Maarten Albersen, Paolo Gontero, Steven Joniau
Introduction: Patients treated with radical cystectomy (RC) and lymph node dissection (LND) are at risk for cancer-related mortality. The COBRA score is an easy-to-use score that incorporates age, pT stage, and lymph node (LN) density to predict cancer-specific survival (CSS). We performed the first validation of the COBRA score in a European cohort of patients.
Methods: Patients treated with RC and LND between 05-1996 and 07-2017 were retrieved from the RC databases of Leuven and Turin. Primary outcome was CSS. Cox proportional hazard models were used to assess impact of variables on CSS. We performed a pairwise comparison using a log-rank test corrected by Bonferroni, and developed a simplified COBRA score with three risk categories. To compare models, we assessed C-indices, ROC curves with AUC, calibration plots and decision curve analysis (DCA).
Results: A total of 812 patients were included. Characteristics of the original SEER cohort and the Leuven/Turin cohort were comparable. All COBRA score variables had a
significant impact on CSS. Pairwise comparison of the COBRA score could not differentiate significantly between COBRA score levels (Figure 1A). Based on these findings, we developed a simplified COBRA score by introducing 3 categories within the following COBRA ranges: low [0-1] vs. intermediate [2-4] vs. high-risk [5-7] (Figure 1B). Pairwise comparison could discriminate significantly between all simplified COBRA risk categories. C-indices of original vs. simplified COBRA score were 0.69 and 0.71. AUCs, calibration plots (Figure 1C) and DCA (Figure 1D) were comparable for both models.
Conclusions: The COBRA score withholds its prognostic value for CSS in a European cohort of patients. We propose a simplified COBRA score with three risk categories with similar performance, which will allows for a more easy stratification. Usage of the COBRA score will allow for objective counseling of patients and usage of adjuvant treatment. Source of