MUPPITS-2 was a randomized, placebo-controlled clinical trial that demonstrated mepolizumab (anti-IL-5) reduced exacerbations and blood and airway eosinophils in urban children with severe eosinophilic asthma. Despite this reduction in eosinophilia, exacerbation risk persisted in certain patients treated with mepolizumab. This raises the possibility that subpopulations of airway eosinophils exist that contribute to breakthrough exacerbations.

We aimed to determine the effect of mepolizumab on airway eosinophils in childhood asthma.

Sputum samples were obtained from 53 MUPPITS-2 participants. Airway eosinophils were characterized using mass cytometry and grouped into subpopulations using unsupervised clustering analyses of 38 surface and intracellular markers. Differences in frequency and immunophenotype of sputum eosinophil subpopulations were assessed based on treatment arm and frequency of exacerbations.

Median sputum eosinophils were significantly lower among participants treated with mepolizumab compared with placebo (58% lower, 0.35% difference [95% CI 0.01, 0.74], P = .04). Clustering analysis identified 3 subpopulations of sputum eosinophils with varied expression of CD62L. CD62L^int and CD62L^hi eosinophils exhibited significantly elevated activation marker and eosinophil peroxidase expression, respectively. In mepolizumab-treated participants, CD62L^int and CD62L^hi eosinophils were more abundant in participants who experienced exacerbations than in those who did not (100% higher for CD62L^int, 0.04% difference [95% CI 0.0, 0.13], P = .04; 93% higher for CD62L^hi, 0.21% difference [95% CI 0.0, 0.77], P = .04).

Children with eosinophilic asthma treated with mepolizumab had significantly lower sputum eosinophils. However, CD62L^int and CD62L^hi eosinophils were significantly elevated in children on mepolizumab who had exacerbations, suggesting that eosinophil subpopulations exist that contribute to exacerbations despite anti-IL-5 treatment.