Abstract
BACKGROUND: In the management of anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV), accurate prognostication of renal outcome is critical. However, the clinical utility of highly precise multi-variable models, such as the ANCA Kidney Risk Score (AKRiS), is often hampered by their complexity. This study aimed to validate a simple alternative: the "Normal" Glomerular Score (NGS), defined as the percentage of normal glomeruli. METHODS: We retrospectively analyzed 188 consecutive patients with biopsy-proven AAV-glomerulonephritis (GN) from a single European center (2001-24). Patients were categorized by NGS: high (>50%), moderate (26%-50%), low (10%-25%) and very low (<10%). We assessed concordance between NGS, Berden classification and AKRiS to determine whether NGS serves as a valid proxy for complex systems. The primary endpoint was death-censored progression to end-stage kidney disease. RESULTS: The NGS strongly indicated presentation renal function; median estimated glomerular filtration rate (eGFR) was 31.4 mL/min in the high NGS group vs 10.8 mL/min in the low cohort (P < .001). It correlated moderately-to-strongly with AKRiS (Rho = -0.714, P < .001). Over a 46.6-month median follow-up, dialysis rates were 12.8%, 23.7%, 42.9% and 42.6% in the high, moderate, low and very low NGS groups (P = .003). There was no statistical difference in dialysis rates (P = .98) or long-term eGFR (28 vs 30 mL/min, P = .51) between the low and very low groups, indicating a prognostic plateau below 25% normal glomeruli. Cox regression showed a significant hazard ratio for dialysis of 1.432 with each worsening category (95% confidence interval 1.144-1.792). The NGS achieved a C-index of 0.812, statistically inferior to AKRiS (0.831). Subanalysis of post-2015 patients confirmed NGS's prognostic validity with modern rituximab-based induction. CONCLUSION: NGS predicts long-term renal survival in AAV-GN. While not replacing multi-variable models like AKRiS for precise risk assessment in clinical trials, it offers a rapid, accessible estimation of functional nephron mass for bedside stratification.