The effect of different inhaled corticosteroid and long-acting bronchodilator combinations on the airway microbiome in patients with severe COPD: a randomised trial (MUSIC)

Background The microbiome is associated with exacerbation risk, quality of life and mortality in COPD. Inhaled corticosteroid (ICS) treatment has been reported to alter the microbiome through modulating host defence. How ICS alters the microbiome and whether effects are equal between different ICS preparations is debated. The aim of the MUSIC trial was to investigate whether commonly used ICS therapies have different effects on the airway microbiome in COPD. Methods This was a multicentre randomised controlled trial. After a 4-week washout period during which they withdrew from ICS, patients with COPD (forced expiratory volume in 1 s <50% predicted at baseline and/or a history of two or more exacerbations per year) were randomised to one of four treatments (budesonide/formoterol 400/12 µg (BF400), fluticasone/salmeterol 500/50 µg (FS500), fluticasone/salmeterol 250/50 µg (FS250) or aclidinium/formoterol 340/12 µg, twice daily). Patients were followed-up for 3 months with monthly induced sputum, oropharyngeal and nasopharyngeal swabs for bacterial load and 16S rRNA sequencing to characterise the microbiome. Inflammatory markers were measured in sputum and blood. The primary outcome was bacterial load in oropharyngeal swabs comparing BF400 versus FS500, with sputum bacterial load the key secondary end-point. Results 122 participants started the washout period. ICS withdrawal was poorly tolerated; 61 participants withdrew before randomisation with 45 experiencing an exacerbation. 61 patients were randomised. No statistically significant differences were observed for the primary comparison of BF400 versus FS500 in oropharyngeal bacterial load. There was, however, a significant increase in sputum bacterial load with FS500 compared to BF400 by month 3. This difference was not seen with FS250. No significant differences in microbiome α-diversity were observed over time. Adverse events were similar between the groups. Conclusion FS500 increased sputum but not upper airway bacterial loads. ICS withdrawal was poorly tolerated in severe COPD.