Asthma is a complex and heterogeneous disorder.1, 2 The presence of eosinophils in asthmatic inflammation has been recognised for many years, and eosinophilic asthma is a common phenotype that is usually responsive to corticosteroid therapy.3 Eosinophilic airway inflammation, as shown by raised sputum eosinophil concentrations, appears to be closely related to the risk of severe asthma exacerbations and loss of asthma control with inhaled corticosteroid withdrawal, although the pathogenetic mechanisms remain undefined.4, 5, 6 The tailoring of asthma therapy based on maintaining sputum eosinophils at 2–3% or less is effective in decreasing asthma exacerbations in patients with severe disease,4, 7 and asthma therapies targeting eosinophils are effective in reducing the incidence of asthma exacerbations and improving markers of asthma control for patients with severe eosinophilic asthma,8, 9, 10 and for patients with moderate-to-severe asthma and eosinophilia.11
Sputum eosinophil percentages of 2% or more to 3% or more of the total cells, depending on the study, have been used to define eosinophilic asthma.2, 4, 11 However, sputum induction is impractical in non-specialised clinical settings. Instead, peripheral blood eosinophil counts are easily obtained, and their use as a biomarker for increased disease burden or exacerbation risk is a topic of ongoing study. An inverse correlation between blood eosinophil counts and forced expiratory volume in 1 s (FEV1) was recorded in an earlier small study.12 In a randomised controlled trial of patients with severe eosinophilic asthma, a progressive increase in risk of exacerbation was found with increasing baseline blood eosinophils,9 and in another study of severe eosinophilic asthma, blood eosinophil counts were independently associated with both risk of exacerbation and treatment response to anti-interleukin-5 therapy, whereas sputum eosinophils did not predict response.13
Possible associations between blood eosinophil counts and overall disease burden in asthma need further study in the general population of patients with asthma, outside of clinical trials for severe asthma. Whereas a recent validation study reported that blood eosinophil counts were an accurate biomarker for identifying sputum eosinophilia,14 other studies report a lack of concordance between presence of sputum eosinophilia and blood eosinophilia.15, 16 Therefore, rather than identifying sputum eosinophils, it is probably more important to determine whether blood eosinophil counts can be used to monitor asthma control or exacerbation risk in clinical practice. In recent observational studies in the USA, raised blood eosinophil counts have been associated with increased prospective risk of asthma exacerbations and excessive short-acting reliever use,17 as well as increased historical risk of exacerbations.18, 19 A need exists to replicate these findings in other settings and databases, to study larger numbers of patients, and to examine patient-reported outcomes.
Research in context
Evidence before this study
We searched PubMed for papers published from Jan 1, 2000, to June 15, 2015, investigating the relation between blood eosinophil count and asthma outcomes, including asthma control and exacerbations, for adult patients with asthma in the general population. We used various combinations of the search terms “asthma”, “eosinophils/eosinophilia”, “exacerbation rate/risk”, and “asthma control”, and restricted our search to publications in English. We reviewed the PubMed search results and reference lists of relevant papers to identify observational studies not restricted to patients with severe or uncontrolled asthma. We identified four observational studies reporting the association between raised blood eosinophil count and increased historical or prospective risk of asthma exacerbations in general populations of patients with asthma.
Added value of this study
Our results support and extend the findings of three of these previous studies to a 40 times or greater general population of over 130 000 patients with asthma in the UK. The large cohort size enabled assessment of the prevalence of raised blood eosinophils among patients with asthma, and availability of questionnaire data for 10% of patients enabled us to look at risk in relation to Global Initiative for Asthma (GINA) current clinical control. We found that patients with raised eosinophil counts had more severe exacerbations and had poorer asthma control (more disease burden) over a subsequent year than those with a blood eosinophil count of 400 cells per μL or less; we detected a clear and consistent count–response relation between blood eosinophil count and our database-derived measures during the outcome year. Our subanalysis finding that eosinophilia was associated with an increased risk of exacerbations within GINA partly controlled and uncontrolled categories provides independent information on risk and agrees with earlier clinical trial findings of a dissociation between symptoms and risk of exacerbations for patients with severe asthma.
Implications of all the available evidence
Our findings, together with those of previous studies, suggest that patients seen in primary care with asthma and blood eosinophilia are potentially at increased risk of future exacerbations regardless of current GINA control status and should be counselled and monitored accordingly. The question remains whether the raised blood eosinophil phenotype is stable, and further research is needed to examine blood eosinophil counts in relation to timing of oral corticosteroid bursts, therapy with oral corticosteroids, and therapy or adherence with inhaled corticosteroids, again in the wider general population of patients with asthma.
Anonymised data from high-quality electronic primary care records of several million patients are available in the UK, permitting the study of very large, heterogeneous populations of patients with asthma. The primary objective of this historical primary care cohort study was to investigate the relation between blood eosinophil count and severe asthma exacerbations and asthma control during the subsequent year. A subanalysis was done to examine the relation between severe exacerbations and Global Initiative for Asthma (GINA)-defined current clinical control. Secondary objectives were to identify a potential relation between demographic and clinical characteristics and the prospective risk of raised eosinophil counts.