The old discussion by the end of the last century about the use of inhaled corticosteroids (ICS) in chronic obstructive pulmonary disease (COPD) generated a lot of ink and dialectical confrontations – sometimes heated – between the defenders of the role of ICS in COPD, who based their arguments on the effect on the prevention of exacerbations (and the much discussed p value=0.052 of the TORCH1 study, – do you remember that?), and the ardent defenders of the FLAME2 study, which made ICS the enemy to beat, partly supported by clear signs of an increased risk of pneumonia associated with this treatment. The scientific evidence generated in recent years has been making both sides happy, while front-line clinicians have seen with astonishment the futility of the arguments being put forward on both fronts, far removed from clinical reality. This was well-reflected in the clinical audit carried out by the EPOCconsul study,3 which showed that 60% of patients with COPD in Spain received treatment with ICS. The incorporation of eosinophils as a biomarker of response to ICS treatment also helped to satisfy both sides, identifying a clear profile of responders that the GesEPOC4,5 guidelines previously classified as mixed phenotype or ACO (Asthma COPD overlap)6 and now see as an eosinophilic exacerbator. International guidelines such as GOLD7 have also opted to direct treatment with ICS to patients with higher peripheral blood eosinophil values in a first assessment, although they also leave its use open in patients with repeated exacerbations and peripheral blood eosinophil counts above 100cells/μL.

Eventually, all these discussions were settled after the publication of studies with fixed triple therapy in COPD (TRIBUTE,8 IMPACT9 and ETHOS10) that consolidated the previous evidence on the efficacy of ICS in preventing exacerbations in exacerbator patients with FEV1<50%.

Even after this point has been accepted, however, many clinicians who often follow severe COPD patients with frequent exacerbations caused by potentially pathogenic bacteria (PPB) (frequently Pseudomonas aeruginosa in the most severe cases)11 still wonder how treatment with ICS can influence the long-term outcome of these patients, considering this treatment's immunosuppressive properties, and even more pertinently, the ICS doses recommended by the guidelines, which are medium or high doses (1000μg of fluticasone propionate or equivalent). Not surprisingly, there is scientific evidence to show that ICS can modify the lung microbiome,12 selecting more aggressive strains such as Pseudomonas aeruginosa. However, what if we are using doses that are too high in these patients, who also receive frequent cycles of antibiotics and oral corticosteroids, due to their exacerbator profile? The use of ICS in patients with blood eosinophil counts<100 cells/μL undoubtedly increases the risk of infection and worsens the prognosis, as shown by the studies published by Martínez-Garcia et al.13 However, it is also possible that eosinophils are not such a good biomarker in the most severely ill patients, who are often underrepresented in clinical trials and only provide us with information about the risk of side effects in patients with repeated severely decreased eosinophil counts. Eosinopenia in COPD appears to be a mortality risk marker in exacerbations14 (also in SARS-CoV215 infection), leaving aside the use of ICs, but the efficacy of ICs in preventing exacerbations has been demonstrated, regardless of the number of eosinophils. Are the currently approved doses of ICS ideal for the most severe COPD patients who have frequent exacerbations with an infective profile? Should we not use lower doses16 if it is shown that these continue to prevent exacerbations? Anzueto et al.17 showed that half the approved dose of fluticasone propionate (250μg every 12h) was equally effective in reducing exacerbations. Meanwhile, research carried out by our group18 in a population with severe COPD has shown that high-dose ICS are associated with a higher risk of Pseudomonas aeruginosa infection and higher mortality – an effect not observed at low doses. In other words, the effect on the risk of infection and on mortality is probably not associated with the use of ICS by itself, but rather with the use of high doses of this treatment. Moreover, this effect is independent of the number of peripheral eosinophil counts. These findings have recently been reproduced in a Danish cohort of 21,000 COPD patients, showing a dose-dependent association between the use of ICS and the risk of Pseudomonas aeruginosa infection.19

To modify the recommendations for IC doses, especially in patients with more severe COPD, we need scientific evidence to show that in this profile the use of low-dose ICs is more effective than not using ICS at all. Recently, at the COPD-Integrated Research Program meeting in SEPAR, the “Study on the Doses of Inhaled Corticosteroids in ePOc” (EDIPO) was presented to test the hypothesis that low doses of ICS (corresponding to 100μg every 12h of fluticasone propionate or equivalent) are as effective in reducing exacerbations as the currently used high doses, but with a lower risk of BPP infections. If the proposed hypothesis is proven and the ICS dose recommendations modified, many questions would thus be cleared up with respect to the management of those more severe COPD patients who frequently overwhelm our pulmonology clinics.