The predictive performance of the ANCA renal risk score in patients over 65 years of age with renal ANCA-associated vasculitis

Background. The anti-neutrophil cytoplasmic antibody (ANCA) renal risk score (ARRS) for predicting renal survival in ANCA-associated vasculitis (AAV) had not previously been validated in adults over 65 years of age and presenting impairments associated with an aging kidney, a high cardiovascular comorbidity burden and prevalent microscopic polyangiitis. Methods. We retrospectively studied a cohort of 192 patients over 65 years of age [median (interquartile range) age: 73 (68-78) years], including 17.2% with renal-limited vasculitis, 49.5% with microscopic polyangiitis and 33.3% with granulomatosis with polyangiitis, at six centres in northern France. The primary study endpoint was the cumulative incidence of end-stage kidney disease (ESKD, maintenance of dialysis for at least 3 months) at 12 months, with death considered as a competing event. Results. The median serum creatinine concentration at diagnosis was 300 (202-502) mu mol/L, and 48 (25.0%) patients required dialysis at presentation. The ARRS was high in 43 (22.4%) patients, medium in 106 (55.2%) and low in 43 (22.4%). The cumulative incidence of ESKD at 12 months was 0% in the low-risk group, 13.0% (interquartile range 7.6-20.0) in the medium-risk group and 44.0% (29.0-58.0) in the high-risk group (P < .001). In the subgroup of 149 patients presenting a medium or high score, the ARRS had a C-index of 0.66 (0.58-0.74) for the prediction of ESKD at 12 months; this rose to 0.86 (0.80-0.90) when dialysis status at diagnosis was included. Conclusion. The ARRS was a poor predictor of kidney survival at 12 months among patients over 65 years of age with renal AAV involvement-especially in the high ARRS group. The addition of dialysis status at diagnosis as an additional clinical parameter might improve the predictive performance of the ARRS.