Objective: Eosinophilic granulomatosis with polyangiitis (EGPA) is one of the principal causes of vasculitic neuropathy. Recently, clinicopathological differences among patients with EGPA based on their anti-neutrophil cytoplasmic antibody (ANCA) status have been reported. Furthermore, the efficacy of mepolizumab (an anti-interleukin-5 (IL-5) antibody drug) for the treatment of patients with EGPA has been demonstrated. This study aimed to elucidate the significance of IL-5 on the clinicopathological features of EGPA.

Methods: The clinical features and pathological findings of sural nerve biopsy specimens from 45 patients with EGPA were examined. 

Results: Of the 45 patients, 26 (58%) exhibited elevated serum IL-5 levels, whereas 14 (31%) were positive for myeloperoxidase (MPO)-ANCA. MPO-ANCA positivity was observed more frequently in IL-5-negative patients than in IL-5-positive patients (27% vs. 37%, P = 0.5), although not statistically significant. Sural nerve biopsy revealed higher numbers of the following parameters in IL-5-positive patients compared with IL-5-negative patients: extravascular eosinophils in the epineurium (P < 0.01), eosinophils in the lumen of epineurial vessels (P < 0.001), eosinophils attached to epineurial vascular endothelial cells (P < 0.01), and epineurial vessels occluded by intraluminal eosinophils (P < 0.05). In contrast, from the viewpoint of MPO-ANCA status, the destruction of vascular structures, suggestive of vasculitis, was more frequently observed in MPO-ANCA-positive than in MPO-ANCA-negative patients (P < 0.05). Vessel walls that demonstrated fibrinoid necrosis were also more frequent in MPO-ANCA-positive patients (P < 0.01).

Conclusions: Our results suggest that in patients with EGPA, the pathogenesis of vasculitis was very closely related to their ANCA status, whereas eosinophil-associated tissue damage was provoked by IL-5.

Supplementary figure e-1:

A scatter diagram of serum IL-5 levels. 

In this study, serum IL-5 was considered positive if IL-5 levels were not lower than the detection limit for the assay (3.9 pg/ml, Quantikine®, R&D Systems, Minneapolis, USA). We also examined serum IL-5 levels in 10 healthy control subjects and confirmed that IL-5 levels in all subjects were lower than the detection limit. Even if the cut off value is increased to 10.0 pg/ml, the result of this study were similar as shown in Supplementary table e-2.

Supplementary figure e-2:

Pathologic findings in four groups in patients with EGPA.

Overall, classical findings suggestive of vasculitis were more likely to be found in MPO-ANCA-positive patients, irrespective of their IL-5 status, whereas eosinophil-associated vascular damage was more common in IL-5-positive patients, irrespective of their MPO-ANCA status.

Supplementary table e-1:

Clinical features and pathological findings of sural nerve biopsy specimens from patients with EGPA with and without ANCA

Supplementary table e-2:

Clinical features and pathological findings of sural nerve biopsy specimens from patients with EGPA (The cut off value of IL-5 is 10.0 pg/ml)