Evaluation of a renal risk score for Japanese patients with ANCA-associated glomerulonephritis in a multi-center cohort study.

Abstract

In patients with anti-neutrophil cytoplasmic antibody (ANCA)-associated glomerulonephritis, prediction of renal survival should guide the choice of therapy, but a prediction of the histological classification has inconsistencies. To evaluate the usefulness of renal risk score (RRS) for Japanese patients with ANCA-associated glomerulonephritis (AAGN) and compare the prediction for end-stage renal disease (ESRD) between RRS and the histological classification. We retrospectively analyzed 96 patients with AAGN who underwent a renal biopsy. Renal survival was categorized by RRS, and the histological classification was assessed separately. We compared the predictive values for RRS and the histological classification. The median observational period was 37.5 (interquartile range [IQR] 21.5-77.0) months. The median RRS point at the time of renal biopsy was 2 (IQR 0-7.8), and the patients were categorized into low- (n = 29), medium- (n = 43), and high-risk groups (n = 24) using RRS. As expected, the renal prognosis was the worst in the "high-risk" group and the best in the "low-risk" group. In the histological classification, the survival deteriorated progressively from "focal" (best) to "mixed," "crescentic," and "sclerotic" (worst) classes, different from the order in the original proposal for this system. Multivariable Cox regression analysis revealed that RRS was independently associated with ESRD. The difference in prediction for renal survival between RRS and the histological classification was not significant using area under receiver-operating-characteristic curves. We evaluated the usefulness of RRS in Japanese patients with AAGN and found it a stable predictor of renal survival in such patients.