We read with great interest the consensus document on the use of inhaled corticosteroids (ICS) in chronic obstructive pulmonary disease (COPD), published in Archivos de Bronconeumología,1 and we would like to consider some aspects.

The use of ICS in COPD remains a controversial subject, and a good number of studies argue either for it or against it. The main indication, which moreover achieves 100% consensus in the above-mentioned document, is in patients with COPD/asthma overlap syndrome – but the problem lies in defining this phenotype.2 A previous diagnosis of asthma may be the best criterion, but this approach assumes that the asthma diagnosis is correct, which may not always be the case.

The other most widely accepted indication is in frequent exacerbations. Several review articles have questioned whether the modest improvement achieved is clinically relevant for the patient. We were interested to see that the weight of consensus on whether to add ICS to the treatment of patient with frequent exacerbations, irrespective of their FEV1 value shifted significantly (54% vs 92%) between the first and second round of questions. This is probably a reflection of the participants’ uncertainty, possibly fueled by studies such as Withdrawal of Inhaled Steroids during Optimized Bronchodilator Management (WISDOM) and Indacaterol: Switching Nonexacerbating Patients with moderate COPD from Salmeterol/Fluticasone to Indacaterol (INSTEAD). Although the final result was taken as consensus, perhaps such a dramatic change should be studied in greater depth.

We would also like to examine the lack of agreement regarding the withdrawal of ICS in the case of eosinophilia. Evidence is growing on the role of elevated blood eosinophils in predicting COPD exacerbations. A recent study of fluticasone/vilanterol compared to vilanterol alone showed that exacerbations were only significantly reduced when blood eosinophilia was >2%, and that a much greater effect was found when eosinophil levels exceeded 6%.3 The consensus panel does not appear to be convinced of the benefit of ICS in exacerbations caused by infectious processes, since 37.5% agreed in the first round and 52% in the second. Bafadhel et al.4 point out that blood eosinophilia is a principal biomarker for non-infectious exacerbations, correlating with response to steroids during exacerbations.

We, personally, agree with Ernst et al.,5 who found that the best current marker for favorable response to ICS in COPD is the presence of eosinophilia. More biomarkers are needed for determining which patients within the wide spectrum of COPD will benefit most from long-term treatment with ICS.