Mortality Prediction in ANCA-Associated Vasculitis with Kidney Involvement: Validation of the Danger Score

Background: The DANGER (Death in ANCA Glomerulonephritis-Estimating the Risk) score was developed in France to assess mortality risk in patients with ANCA-associated vasculitis (AAV). This study aimed to validate this score in a cohort of Latin American patients. Methods: Retrospective cohort study that included patients with AAV evaluated between 2000 and 2022. The DANGER score was calculated, and its performance evaluated by the c-statistic and time-dependent area under the ROC curve (AUROC). A multivariable Cox regression analysis was performed to identify variables that could improve the score’s predictive accuracy. Results: We included 154 patients, 104 (68%) female, with a median age of 52 years (IQR 38-61), and creatinine of 2.5mg/dL (IQR 1.7-2.5). Over a median follow-up of 74 months (IQR 32-126), 24 patients died, with mortality rates of 6.5%, 8.6%, and 11.9% at 1, 2, and 5 years, respectively. The leading cause of death was infection. Mortality rates at 1 and 3 years in the low, intermediate, and high-risk categories of the score were 1.0% and 3.1%, 14.0% and 16.8%, 40.0% and 70.0%, respectively (Figure 1A). The overall c-statistic of the DANGER model was 0.810 (95%CI 0.725-0.895), with AUROC of 0.809 (95%CI 0.703-0.914), 0.780 (95%CI 0.666-0.894), and 0.803 (95%CI 0.702-0.903) at 1, 3, and 5 years, respectively. Pulmonary involvement, especially interstitial lung disease and diffuse alveolar hemorrhage were associated with mortality in this cohort (Figure 1B). A complementary model incorporating age, creatinine, C-reactive protein, and pulmonary involvement had a c-statistic of 0.861 (95%CI 0.785-0.936). Conclusion: The DANGER score has good predictive accuracy for mortality in AAV patients with kidney involvement. In younger patients, the score may be modified to include variables such as C-reactive protein and pulmonary involvement to improve its performance.