Background/Purpose: Eosinophilic granulomatosis with polyangiitis (EGPA) (Churg–Strauss) is a small-vessel necrotizing vasculitis characterized by blood and tissue eosinophilia and asthma. Glucocorticoids (GCs) effectively control the disease, but relapses and/or GC-dependence are frequent. Recently, efforts were made to improve nosology of EGPA patients. Evolving concepts tend to distinguish vasculitis-related symptoms from asthma and/or ENT manifestations. That distinction has become even more important, since the development of new targeted biotherapies. This study aimed to describe and identify characteristics predicting long-term EGPA outcomes.

Methods: We created a multicenter European collaborative initiative that included 257 EGPA patients from tertiary referral centers. Based on recent consensus, we distinguished 4 EGPA-evolutionary profiles: GC-dependent asthma and/or ENT manifestations (requiring prednisone >7.5 mg/d), at least 1 vasculitis relapse (excluding asthma and/or ENT manifestations), both phenotypes, and prolonged remission (no GC-dependent asthma/ENT signs and no vasculitis relapse). Baseline and follow-up characteristics predicting those outcomes were analyzed.

Results: After median follow-up of 60 months, 24% had GC-dependent asthma and/or ENT manifestations, 18% had at least 1 vasculitis relapse, 8% had both phenotypes, and 50% were in prolonged remission (Table). Patients with GC-dependent asthma/ENT manifestations were younger at diagnosis, had more frequent asthma requiring GCs before overt EGPA and pulmonary infiltrates, less frequent general symptoms and ANCA-positivity, and tended to have lower eosinophil counts. Their daily GC dose and eosinophil counts were higher at every time point; at last follow-up, they had more active asthma and less frequent neurological sequelae. In contrast, patients with only vasculitis relapse(s) had more frequent general symptoms at diagnosis, ANCA-positivity and higher BVAS, and less frequent pulmonary infiltrates. Median diagnosis-to-1^st-vasculitis-relapse interval was 15 (9–42) months. During follow-up, their daily GC dose was lower than for those with GC-dependent asthma and/or ENT manifestations but similar to that of those in prolonged remission. At last follow-up, neurological sequelae were more frequent but active asthma less common. Finally, patients in prolonged remission were older, had less frequent asthma requiring GCs before EGPA, and lower daily GC dose and eosinophil counts during follow-up and less frequent sequelae.

Conclusion: Distinct baseline and follow-up characteristics defined 4 evolutionary EGPA profiles predicting patients’ long-term outcomes. Each evolutionary pattern was identifiable soon after diagnosis, which would allow early choices of the best therapeutic option in the future.

Table