Efficacy of umeclidinium/vilanterol according to the degree of reversibility of airflow limitation at screening: a post hoc analysis of the EMAX trial

BACKGROUND: In patients with chronic obstructive pulmonary disease (COPD), the relationship between short-term bronchodilator reversibility and longer-term response to bronchodilators is unclear. Here, we investigated whether the efficacy of long-acting bronchodilators is associated with reversibility of airflow limitation in patients with COPD with a low exacerbation risk not receiving inhaled corticosteroids. METHODS: The double-blind, double-dummy EMAX trial randomised patients to umeclidinium/vilanterol 62.5/25 µg once daily, umeclidinium 62.5 µg once daily, or salmeterol 50 µg twice daily. Bronchodilator reversibility to salbutamol was measured once at screening and defined as an increase in forced expiratory volume in 1 s (FEV(1)) of ≥ 12% and ≥ 200 mL 10−30 min post salbutamol. Post hoc, fractional polynomial (FP) modelling was conducted using the degree of reversibility (mL) at screening as a continuous variable to investigate its relationship to mean change from baseline in trough FEV(1) and self-administered computerised-Transition Dyspnoea Index (SAC-TDI) at Week 24, Evaluating Respiratory Symptoms-COPD (E-RS) at Weeks 21–24, and rescue medication use (puffs/day) over Weeks 1–24. Analyses were conducted across the full range of reversibility (−850–896 mL); however, results are presented for the range −100–400 mL because there were few participants with values outside this range. RESULTS: The mean (standard deviation) reversibility was 130 mL (156) and the median was 113 mL; 625/2425 (26%) patients were reversible. There was a trend towards greater improvements in trough FEV(1), SAC-TDI, E-RS and rescue medication use with umeclidinium/vilanterol with higher reversibility. Improvements in trough FEV(1) and reductions in rescue medication use were greater with umeclidinium/vilanterol compared with either monotherapy across the range of reversibility. Greater improvements in SAC-TDI and E-RS total scores were observed with umeclidinium/vilanterol versus monotherapy in the middle of the reversibility range. CONCLUSIONS: FP analyses suggest that patients with higher levels of reversibility have greater improvements in lung function and symptoms in response to bronchodilators. Improvements in lung function and rescue medication use were greater with umeclidinium/vilanterol versus monotherapy across the full range of reversibility, suggesting that the dual bronchodilator umeclidinium/vilanterol may be an appropriate treatment for patients with symptomatic COPD, regardless of their level of reversibility. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12931-021-01859-w.