Bronchodilator responsiveness as a phenotypic characteristic of established chronic obstructive pulmonary disease

Background Bronchodilator responsiveness is a potential phenotypic characteristic of chronic obstructive pulmonary disease (COPD). We studied whether change in lung function after a bronchodilator is abnormal in COPD, whether stable responder subgroups can be identified, and whether these subgroups experience different clinical outcomes. Methods 1831 patients with COPD, 285 smoking (SC) and 228 non-smoking (NSC) controls from the Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints (ECLIPSE) cohort. Spirometric reversibility to 400 mu g inhaled salbutamol was assessed on four occasions over 1 year. Results Forced expiratory volume in 1 s (FEV1) increase after salbutamol was similar in SC (mean 0.14 litres (SD 0.15)) and COPD (0.12 litres (0.15)) and was significantly greater than NSC (0.08 litres (0.14)). Reversibility status varied with repeated testing in parallel with the day-to-day variation in pre-bronchodilator FEV1, which was similar in control subjects and patients with COPD. Absolute FEV1 change decreased by Global initiative for chronic Obstructive Lung Disease (GOLD) stage in patients with COPD (GOLD II, mean 0.16 litres (SD 0.17); III, 0.10 litres (0.13); IV, 0.05 litres (0.08) as did chances of being classified as reversible. CT-defined emphysema was weakly related to the absolute change in FEV1 post salbutamol. Consistently reversible patients (n=227) did not differ in mortality, hospitalisation or exacerbation experience from irreversible patients when allowing for differences in baseline FEV1. Limitations Reversibility only assessed with salbutamol and defined by FEV1 criteria. The COPD population was older than the control populations. Conclusions Post-salbutamol FEV1 change is similar in patients with COPD and smoking controls but is influenced by baseline lung function and the presence of emphysema. Bronchodilator reversibility status varies temporally and does not distinguish clinically relevant outcomes, making it an unreliable phenotype.