To determine the comparative efficacy of non-biologic treatments for remission maintenance in ANCA-associated vasculitis.

We identified all randomized trials comparing leflunomide, azathioprine, methotrexate or mycophenolate mofetil in adult patients with granulomatosis with polyangiitis or microscopic polyangiitis. Relapse-free survival was compared through hazard ratios (HR) using a Bayesian fixed-effects network meta-analysis. Multiple sensitivity analyses were performed to explore biases identified in one trial using original trial data.

Three trials were available (leflunomide-methotrexate, methotrexate- azathioprine, azathioprine-mycophenolate). Mycophenolate was inferior to all treatments, although the 95% credible interval (CrI) of the HR relative to methotrexate crossed 1. Leflunomide was superior to azathioprine (HR 0.43 [95% CrI: 0.14–1.3]) and methotrexate (HR 0.47 [95% CrI: 0.18–1.2]), although the 95% CrI also crossed 1. There was a 90% probability that leflunomide was the best treatment. After down weighting the effect of leflunomide vs. methotrexate for early trial termination and slow MTX dose escalation, there remained a 55% probability leflunomide was best.

Based on indirect evidence, leflunomide is effective in maintaining remission in granulomatosis with polyangiitis or microscopic polyangiitis relative to other non-biologic treatments. Further randomized trials of leflunomide are needed for confirmation.