Airway inflammation in chronic obstructive pulmonary disease (COPD) is characterised by increased concentrations of neutrophils,1 macrophages,2 proteases, interleukin 6 and 8, and T-helper-1 (Th1) cytokines,3 whereas airway inflammation in asthma is characterised by increased concentrations of eosinophils and Th2 cytokines.4 However, some research has challenged these presumed differences between the characteristic markers of asthma and COPD.
Research in context
Evidence before this study
We did a PubMed search for original research reports using the search terms “eosinophils”, “sputum”, “blood”, and “COPD” from April 15, 2014, to May 18, 2017, which yielded 154 articles, of which 32 were reviews. No publication date or language restrictions were used. Addition of “severity” as a search term reduced the publication number to 33 (seven reviews) and addition of “exacerbation” reduced the number to 35 (one review). However, many of these reports have further limitations. Some did not have sputum or blood eosinophil data for comparison, did not specifically focus on the severity of chronic obstructive pulmonary disease (COPD; eg, exacerbations), or were based on small numbers of patients (<100 per group), which limits the power to make conclusions for broader COPD populations. Generally, eosinophils in COPD have been linked to more frequent exacerbations and responsiveness to corticosteroid therapy, suggesting more severe disease. Often, studies are done primarily in populations that have met selection criteria for clinical trials, including the presence of COPD exacerbations. Thus, comparison of blood and sputum eosinophil concentrations for an association with severity of COPD phenotype has not been well studied in a general smoking population with a broad range of COPD severity, nor has possible substitution of blood eosinophils as a biomarker for sputum eosinophils in COPD populations been carefully examined.
Added value of this study
This study shows that in a large, comprehensively characterised smoking cohort with a broad range of COPD severity, increased sputum eosinophils, but not blood eosinophils alone, had significant associations with multiple measures of COPD severity, including exacerbations, increased emphysema and air trapping, St George Respiratory Questionnaire scores, and Global Initiative for Chronic Obstructive Lung Disease spirometric stage. Blood eosinophils showed weak association with sputum eosinophils and as a single biomarker had few significant associations with COPD severity and exacerbations. However, this study does show that increased blood eosinophils in combination with increased sputum eosinophils show associations with COPD exacerbations and severity.
Implications of all the available evidence
Increased sputum eosinophils in patients with a broad range of COPD severities identify patients who are more likely to have severe disease and exacerbations. Blood eosinophils as a single biomarker do not accurately predict sputum eosinophils, and do not show any association with disease severity or exacerbations unless observed in combination with increased sputum eosinophils. The findings from this study will be important in the design of therapeutic trials that target eosinophilic inflammation in COPD.
The ECLIPSE study1 reported that in COPD, sputum neutrophil concentrations were weakly associated with lung function and health status, but not associated with exacerbations, emphysema, or systemic inflammation. ECLIPSE reported a mean 1·3% (SD 2·6) sputum eosinophil concentration in 359 patients with COPD,1 but did not observe associations of blood eosinophils with radiological measures of emphysema or with COPD exacerbations and hospital admissions. ECLIPSE reported that concentrations of persistently 2% or more blood eosinophils (150 cells per μL) were associated with evidence of higher FEV1, lower St George Respiratory Questionnaire (SGRQ) score, and modified Medical Research Council score compared with intermittent concentrations or concentrations persistently less than 2%.5 Other COPD studies have reported that increased eosinophils in both blood and sputum are associated with respiratory exacerbations and greater hyperinflation when assessed by quantitative CT (QCT),6, 7 suggesting that Th2 inflammation might contribute to COPD progression. Furthermore, increased epithelial Th2 signature gene expression has been associated with more severe airflow obstruction in two COPD cohorts.8 Eosinophils might therefore be a potential biomarker in COPD because eosinophilia is related to corticosteroid responsiveness.1, 9, 10, 11 In a phase 2 clinical trial, anti-interleukin-5 receptor therapy reduced the occurrence of COPD exacerbations in a subgroup of patients with high concentrations of blood and sputum eosinophils.12
Establishing disease severity in patients with COPD is complex and involves more than lung function assessments; additional clinical characteristics have been incorporated in successive revisions of the Global Initiative for Chronic Obstructive Lung Disease (GOLD) severity stages.13 The current classification includes lung function, symptom scores, and exacerbation frequency. Thus, the severity of COPD is dependent on multiple characteristics, and eosinophilic inflammation might contribute.
Previous reports suggest that blood eosinophil counts might be a useful surrogate measure of airway eosinophils in COPD,11, 14 although blood eosinophils appear to correlate poorly with sputum eosinophils in asthma,15, 16 and do not distinguish between populations who are asthma dominant, COPD dominant, or those who have asthma–COPD overlap.17 However, larger studies of comprehensively phenotyped patients with COPD often do not have robust sputum eosinophil data either because sputum induction was not done or sputum induction cohorts were small.18, 19, 20, 21 Thus, whether or not peripheral eosinophils do accurately predict airway eosinophils is unknown.
We investigated the hypotheses that high concentrations of blood and sputum eosinophils in patients with a history of tobacco use are associated with a more severe COPD phenotype, identified by diminished lung function, QCT measurements of emphysema or air-trapping, clinical COPD characteristics, and exacerbations. We also investigated relationships between blood and sputum eosinophils to establish whether blood eosinophil concentrations reliably predicted sputum eosinophil concentrations. Measuring blood eosinophils is an easier and less expensive option than sputum induction in a clinical setting. Thus, an ability to predict sputum eosinophils accurately from blood eosinophil concentrations would be useful for clinical studies and patient care. These hypotheses were assessed in the comprehensively characterised SPIROMICS cohort.22 A portion of these studies were presented as an abstract at the 2016 American Thoracic Society meeting.23