Anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) involves autoimmune-mediated inflammation of predominantly small-sized blood vessels. Pediatric-onset AAV is often more severe than adult-onset AAV, with high rates of kidney dysfunction at diagnosis, and accrual of permanent kidney damage despite aggressive immunosuppressive therapy. Since children remain at risk of accumulating disease- and treatment-related damage, this highlights the need for pediatric-specific monitoring and predictive tools to inform tailored treatment administration. While adult-based studies have associated histopathological features on kidney biopsy and urinary metabolites with kidney outcomes and identified urine proteins capable of discriminating kidney disease activity, their utility in pediatric-onset AAV-associated glomerulonephritis (AAV-GN) has not been established. Using the largest repository of pediatric vasculitis clinical data and biological samples (stored at the University of British Columbia and BC Children’s Hospital Research Institute), my thesis evaluated the performances of adult-based histopathological classification schemas, the Berden Classification, the ANCA Renal Risk Score (ARRS) and the ANCA Kidney Risk Score (AKRiS) to predict 12-month kidney outcomes. Second, I explored urinary metabolites as non-invasive indicators of kidney dysfunction and chronic kidney disease (CKD), while attempting to validate candidate urinary protein biomarkers of kidney disease activity, soluble CD163 (usCD163) and CD25 (usCD25). The Berden Classification, ARRS and AKRiS predicted kidney outcomes in pediatric AAV-GN (N = 103) with comparable accuracy (AUROC range: 0.780–790). However, the ARRS performed better in myeloperoxidase (MPO)-AAV (n = 41; AUROC 95% CI: 0.892 0.784, 0.974), compared to proteinase-3 (PR3)-AAV (n = 34; AUROC 95% CI: 0.714 0.488, 0.917). Comparisons of urine proteins (n = 25) demonstrated that excretion of usCD163 (p =0.009), but not usCD25 (p = 0.167), significantly decreased from baseline (high kidney disease activity) to follow-up (inactive kidney disease). Baseline excretion of 17 and 23 urinary metabolites associated with at-diagnosis kidney dysfunction (n = 19) and 12-month CKD (n = 25) (p < 0.05), respectively. These findings highlight the limitations of applying adult-derived predictive tools to children and support further investigation of urinary metabolites as potential non-invasive biomarkers in pediatric AAV-GN. Integrating urinary biomarkers into clinical prediction models may optimize the monitoring of kidney disease trajectories in pediatric-onset AAV.
Sabrina Sefton (Fri,) studied this question.