Nationwide study of blood eosinophils in COPD: impact of exacerbations and corticosteroids
Abstract
Abstract Introduction Blood eosinophils are increasingly used as biomarkers to guide inhaled corticosteroid (ICS) and biologic therapy in chronic obstructive pulmonary disease (COPD). However, the long-term stability of eosinophil count and the impact of proximate clinical events remain uncertain. This study examines the longitudinal stability of blood eosinophil counts in a nationwide cohort of COPD patients and identifies acute exposures temporally linked to eosinophil measurement variability. Methods We used routinely collected Danish national registry data including 63 517 patients with COPD aged ≥40 years with ≥4 eosinophil measurements during 2–6 years of follow-up (median 3.4). Eosinophils were categorised as <0.15, 0.15–0.29 and ≥0.30×10⁹/L. Logistic regression assessed associations between stability and clinical exposures, adjusting for baseline eosinophils, age, sex, comorbidities, oral corticosteroid (OCS) use, antibiotics and exacerbations. Transitions were visualised with Sankey diagrams. Results Overall, 26% of patients remained within the same eosinophil category across all repeated measurements (stable patient phenotype), while 74% experienced ≥1 category shift. Baseline eosinophils ≥0.30×10⁹/L predicted phenotypic stability (OR 1.70, 95% CI 1.63 to 1.77). Severe exacerbations, OCS and antibiotics prior to testing were associated with individual measurement variability, yet >80% of unstable measurements occurred without these exposures. Among patients with an unstable phenotype, 39.7% shifted once, 43.2% two times and 17.1% three times. Use of triple inhaler therapy (long-acting beta-agonist, long-acting muscarinic antagonist, ICS) was slightly more common among in the unstable group. Conclusions Blood eosinophil counts in COPD are highly dynamic, with most patients shifting categories over time independent of clinical exposures. Single eosinophil measurements may be unreliable for guiding therapy; repeated assessments could improve identification of patients most likely to benefit from type 2 inflammation-targeted treatments.