In current smokers with COPD, LABA+LAMA is superior to LABA+ICS, regardless of blood eosinophils. Data from the FLAME trial.

<bold>Background:</bold> Current smokers with COPD may be less or non-responsive to corticosteroids. This post-hoc analysis of the FLAME trial explored the impact of current smoking status on treatment response to LABA+ICS versus LABA+LAMA in relation to blood eosinophil count (BEC) <bold>Methods:</bold> We selected 481 current and 851 ex-smokers who were receiving ICS at baseline, and had their BEC measured both on and off ICS treatment, as we have previously showed these are not equivalent. We assessed the interaction between BEC off ICS, BEC change during ICS treatment and treatment effects on exacerbation outcomes among (a) current and (b) ex-smokers. Then, we explored the interaction between smoking status and treatment effects on exacerbations among participants anticipated to gain most benefit from ICS. We used negative binomial regression models and cox proportional hazard ratios for rate of and time-to-first exacerbation. <bold>Results:</bold> Higher BEC off ICS and BEC suppression during treatment were associated with favourable treatment response to LABA+ICS versus LABA+LAMA among ex-smokers (significant interaction), but not among current smokers, where LABA+LAMA was superior irrespective of BEC. Smoking status exerted significant interaction with the rate of and time-to-first severe and moderate or severe exacerbations (p≤0.02) among patients with BEC off ICS ≥200 cells/μL, or BEC suppression (BEC change <0 cells/μL). In these groups, ex-smokers appeared to gain more benefit from LABA+ICS vs LABA+LAMA. <bold>Conclusion:</bold> In current smokers with COPD, LABA+LAMA appeared superior to LABA+ICS in preventing exacerbations, irrespective of BEC.