BACKGROUND:Chronic obstructive pulmonary disease (COPD) is a heterogeneous disorder encompassing different phenotypes with different responses to treatment. The present 1-year, two-center hospital-based study investigated whether the plasma immunoglobulin E (IgE) level and/or eosinophil cell count could be used as biomarkers to stratify patients with COPD according to predicted responses to inhaled corticosteroids (ICS)-based therapy. METHODS:A hospital-data based cohort study of COPD patients treated at two territory hospital centers was conducted for 1 year. Allergic biomarkers, including blood eosinophil counts and IgE levels, were assessed at baseline. Lung function parameters, including forced expiratory volume in 1 second (FEV1), forced vital capacity (FVC), and the COPD Assessment Test (CAT), were also evaluated. The frequencies of acute exacerbation (AE) and pneumonia were also measured. Eosinophilia and a high IgE level were defined as >3% and 173 IU/mL, respectively. RESULTS:A total of 304 patients were included. Among patients with eosinophilia and high IgE levels, ICS-based therapy was associated with significant improvements in FEV1, FVC, and CAT scores, compared with bronchodilator (BD) therapy (P≤0.042). ICS-based therapy was also associated with a significantly lower incidence of AE vs BD-based therapy (11.7% vs 24.1%; P<0.008). Among patients with only eosinophilia, ICS-based therapy yielded significantly better CAT score results vs BD-based treatment (7 vs 13; P=0.032). A receiver operating characteristic curve analysis found that the combination of a high plasma IgE level and eosinophilia most sensitively and specifically identified patients who would benefit from the addition of ICS to BD therapy. CONCLUSION:Our findings support the use of blood eosinophil cell counts plus IgE levels as predictive biomarkers of the ICS response in certain patients with COPD. Both biomarkers could potentially be used to stratify COPD patients regarding ICS-based therapy.