Background/Purpose: At month 24 (M24), CHUSPAN2-trial patients with nonsevere eosinophilic granulomatosis with polyangiitis (EGPA), who received 1 yr of azathioprine (AZA) and glucocorticoids (GC) as remission-induction therapy, did not have lower remission failure, vasculitis relapse-risk or asthma/rhinosinus-exacerbation (ARE) rates, or steroid-sparing (1). These EGPA patients’ long-term outcomes were analyzed to determine whether vasculitis-relapse or isolated ARE rates differed thereafter.

Methods: After M24, treating physicians chose treatments. Long-term outcomes were analyzed according to randomization group (AZA vs placebo). Flares and reasons for GC escalation or immunosuppressant use were reviewed to distinguish vasculitis relapses from isolated AREs according to EGPA Task Force recommendations. Trial entry parameters were evaluated as potential prognostic factors for vasculitis relapse or isolated ARE during follow-up.

Results: The 50 EGPA patients (25 men, mean age 53.4 yr, mean eosinophilia 5,900/µL) were followed for a median [IQR] of 6.3 [5.4‒7.6] yr. Two (4%) patients (1 after remission-induction failed) died 11 months postinclusion. At M24, 21/48 (43.8%) patients had vasculitis relapse and 14/50 (28%) had isolated AREs. Among the 48 follow-up patients analyzed, 1 died 4.8 yr postinclusion (probable infection); 19.6% (9/46) had subsequent vasculitis relapses of whom 3 had their 1^st relapse after M24; 25/45 (55.6%) patients had later isolated AREs, the 1^st for 17 only after M24. From M24 to last update, conventional immunosuppressants were prescribed for 22/48 (45.8%) patients to control flares (17 AZA, 5 MTX, 4 cyclophosphamide, 1 leflunomide) and, after immunosuppressants failed, rituximab for 7; 4 received mepolizumab and 3 omalizumab, mostly for AREs. At 5 yr, respective vasculitis relapse and isolated ARE rates were 48.3% (95% CI: 34.0–62.6) and 56.2% (95% CI: 41.7–70.8), with no long-term between-arm differences (P=0.16 and P=0.06). For the AZA and placebo arms, respective 5-yr overall survival (OS) rates were 100% and 91.3% (95% CI: 79.7–100; P=0.15). At last update, 78.7% of patients still required GC (median dose: 5 [3.0–8.0] mg/d) to control AREs and/or as vasculitis maintenance therapy. Damage resulted from asthma (n=42), followed by peripheral nerve involvement (n=28) and chronic sinusitis/nasal obstruction (n=27); Vasculitis Damage Indexes (mean±SD) were comparable for AZA (2.8±1.5) and placebo (2.2±0.9) groups (P=0.42). Among the 18 (36%) patients MPO-ANCA+ at entry, 13 had ANCA ELISA results at M24: 6 became ANCA– (none relapsed later) and 7 were ANCA+, after having been transiently ANCA– for 3, (2/7 subsequently relapsed). Analyses retained no entry clinical or biological parameter, including blood eosinophil count and ANCA status, as significantly associated with vasculitis relapse or ARE during follow-up.

Conclusion: These results confirmed that 1 yr of AZA and GC obtained good OS but had no long-term benefit for EGPA patients. Extrapulmonary vasculitis relapses, occurring mostly during the first 2 yr, and isolated AREs, predominating thereafter, remain worrisome and require further studies to prevent them.

1 Puéchal X. et al. Arthritis Rheumatol 2017;69:2175–86.

« Back to 2018 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/long-term-outcomes-of-patients-with-nonsevere-eosinophilic-granulomatosis-with-polyangiitis-given-azathioprine-and-glucocorticoids-for-remission-induction/