Comparing Rituximab and Cyclophosphamide in Induction Therapy for Childhood-Onset Anti-Neutrophil Cytoplasmic Antibody-Associated Vasculitis: An ARChiVe Registry Cohort Study

Abstract

Objective: Granulomatosis with polyangiitis (GPA) and microscopic polyangiitis (MPA) are chronic life-threatening vasculitides requiring substantial immunotherapy. Adult trials identified rituximab (RTX) as an alternative to cyclophosphamide (CYC) for remission induction of GPA and MPA. Disease rarity has limited feasibility of similar trials with pediatric patients. We aim to evaluate the relative efficacy and toxicity of CYC and RTX for patients with childhood GPA and MPA through registry-based comparative evaluation. Methods: From A Registry of Childhood Vasculitis, we identified patients with GPA and MPA who received induction with RTX or CYC. Pediatric Vasculitis Activity Score (PVAS) and Pediatric Vasculitis Damage Index (pVDI) score evaluated disease activity and damage. Descriptive statistics summarized patient characteristics. RTX and CYC comparisons used logistic regression for primary outcomes of postinduction remission (PVAS = 0) or low disease activity (PVAS ≤ 2). Hospital admission for adverse events and pVDI scores were compared using logistic regression and ordinal regression, respectively. Results: Among 104 patients, 43% received RTX, 46% CYC, 11% both. Treatment groups did not significantly differ for diagnosis PVAS and onset age. There was no difference in remission among the groups (63% overall; odds ratio [OR] 1.07, 95% confidence interval [CI] 0.45–2.52). Hospitalizations occurred in 22% of patients receiving RTX versus 10% patients receiving CYC (OR 2.27, 95% CI 0.73–7.05). The median 12-month pVDI score was 1 in both groups (OR 0.98, 95% CI 0.43–2.22). Conclusion: This is the first study comparing CYC and RTX for induction in pediatric GPA and MPA. No significant differences were shown in rates of remission, severe adverse events, or organ damage. Limitations included lack of standardized treatment regimens, retrospectivity, and lack of longitudinal adverse drug-related event data.