The inflammatory profile of exacerbations in patients with severe refractory eosinophilic asthma receiving mepolizumab (the MEX study): a prospective observational study.

BACKGROUND

Clinical trials with mepolizumab, a humanised monoclonal antibody against interleukin-5, show a 50% reduction in severe asthma exacerbations in people with severe eosinophilic asthma. Exacerbations in patients treated with mepolizumab seem to be different to exacerbations in those given placebo, as patients treated with mepolizumab report fewer symptoms, have a lower sputum eosinophil count, and smaller fall in peak expiratory flow. We aimed to investigate the inflammatory phenotype and physiological characteristics of exacerbation events in patients with severe eosinophilic asthma who were treated with mepolizumab.

METHODS

This multicentre, prospective, observational cohort study was carried out at four UK specialist severe asthma centres. Participants were aged 18-80 years, with severe eosinophilic asthma (Global Initiative for Asthma steps 4 and 5), and were eligible for mepolizumab therapy. All participants received mepolizumab 100 mg subcutaneously every 4 weeks, had a scheduled study visit when stable on mepolizumab (≥3 months on treatment), and measured daily peak flow and completed symptoms diaries throughout the course of the study. Participants attended their study centre for unscheduled exacerbation assessment when symptoms worsened outside of their normal daily variation and before commencing rescue treatment. If a participant was unable to attend their study centre for exacerbation or had initiated rescue treatment before the study visit, clinical details of the missed exacerbation were collected by clinical staff. In this exploratory study, the endpoint was 100 clinical assessments at exacerbation completed across all sites for participants on mepolizumab before initiation of rescue treatment. Characteristics of those who had exacerbations on mepolizumab were compared with those who did not, peak flow and symptoms diaries were compared for assessed versus missed exacerbations, and exacerbation phenotypes defined by sputum eosinophil cell count were compared. The utility of fractional exhaled nitric oxide (FeNO) and C-reactive protein in determining exacerbation phenotype on mepolizumab treatment were also assessed. This study is registered with ClinicalTrials.gov, NCT03324230.

FINDINGS

Between Nov 30, 2017, and May 29, 2019, 145 participants were enrolled and treated with mepolizumab, five were excluded from the analysis. 172 exacerbations occurred, with 96 (56%) assessed before commencing rescue treatment. Compared with patients who did not exacerbate, patients who exacerbated had a higher exacerbation rate and more emergency department attendances in the year before commencing mepolizumab. The change in peak expiratory flow at nadir in the assessed exacerbation group was mean -40·5 L/min (SD 76·3) versus mean -37·0 L/min (93·0; p=0·84) in the missed exacerbation group, and there was no difference in reported symptom burden. When comparing exacerbations with a high sputum eosinophil count (≥2%; SE) with exacerbations with a low sputum eosinophil count (<2%; SE), the SE exacerbations were FeNO high (median difference 33 parts per billion [ppb; 95% CI 8 to 87]; p=0·0004), with lower FEV percent predicted (mean difference -15·9% [-27·0 to -4·8]; p=0·0075), lower FEV to forced vital capacity ratio (mean difference -10·3 [-17·0 to -3·6]; p=0·0043), and higher blood eosinophil counts (median difference 40 cells per μL [20 to 70]; p=0·0009). By contrast, SE exacerbations had higher C-reactive protein concentrations (median difference 12·7 mg/L [3·5 to 18·5]; p<0·0001), higher sputum neutrophil counts (median difference 52·7% [34·5 to 59·2]; p<0·0001), and were more likely to be treated with antibiotics (p=0·031). FeNO (≤20 or ≥50 ppb) was the most useful discriminator of inflammatory phenotype at exacerbation. The most common adverse event was hospital admission due to asthma exacerbation (17 [50%] of 34 events), none of the adverse events were study procedure related.

INTERPRETATION

Exacerbations on mepolizumab are two distinct entities, which can largely be differentiated using FeNO: non-eosinophilic events are driven by infection with a low FeNO and high C-reactive protein concentration, whereas eosinophilic exacerbations are FeNO high. The results of the MEX study challenge the routine use of oral corticosteroids for the treatment of all asthma exacerbation events on mepolizumab, as well as the switching of biological therapies for treatment failure without profiling the inflammatory phenotype of ongoing asthma exacerbations. The results highlight clinically available tools to enable profiling of these residual exacerbations in patients treated with mepolizumab.

FUNDING

UK Medical Research council.