In a recent Lung Alert reviewing our study of mepolizumab in severe eosinophilic asthma,1 Barratt states that the study population represented a minority of patients with asthma and that they had corticosteroid-resistant disease.2 These comments require clarification. While we accept that the population studied by us represents about 3% of the total asthma population, it was 30% of patients with refractory asthma, a population with the greatest unmet need for new treatment. In contrast, only 13% of this population meet criteria set out by the manufacturer for treatment with omalizumab, and only a minority of these meet NICE criteria for use of the agent.

Our patients were corticosteroid resistant in the sense that they continued to have morbidity despite high-dose inhaled corticosteroids and, in many cases, oral prednisolone. However, many responded well to higher-dose oral prednisone, and a post hoc analysis showed that those who did tended to do better with mepolizumab treatment (table 1). In contrast, there was a poorer response to mepolizumab in patients with marked bronchodilator reversibility. The clear implication is that mepolizimab works best in patients who have airflow limitation and symptoms as a result of corticosteroid-responsive airway inflammation rather than airway smooth muscle contraction. Interestingly, the exact opposite relationship between acute bronchodilator response and clinical efficacy has recently been reported for treatment with anti-tumour necrosis factor α,3 a treatment which targets airway smooth muscle function but not eosinophilic airway inflammation.4 It therefore appears that the population that does well with these different anti-cytokine strategies can be recognised by identifying the mechanism of airflow limitation.

We believe that clinicians who see large numbers of patients with inflammatory airway diseases will readily recognise patients with prednisolone-responsive bronchodilator-resistant airflow limitation and/or symptoms in their asthma, cough and COPD clinics. The challenge is that many of these patients have disease that is difficult to classify using current physiology-based classification systems. The optimum development of mepolizumab and other promising agents does require new and better ways of assessing and classifying inflammatory airways disease.