Blood eosinophil levels and prognosis of hospitalized patients with acute exacerbation of chronic obstructive pulmonary disease

Background: Studies about the clinical significance of high eosinophil levels in chronic obstructive pulmonary disease (COPD) are conflicting, and it has been less studied in hospitalized patients with acute exacerbation of COPD (AECOPD).This study was to examine blood eosinophil levels in relation to the prognosis of hospitalized patients with AECOPD. Methods: This was a retrospective cohort study of patients with AECOPD as their primary diagnosis and admitted to Beijing Shijitan Hospital, Capital Medical University, from January 2010 to December 2016. The patients were assigned according to the proportion and count of eosinophil in peripheral blood at their first hospitalization. Patients were grouped as ≤100, 100-300, and ≥300 eosinophils/µL of peripheral blood. The use of glucocorticoids, duration of hospitalization, in-hospital mortality, and re-hospitalization were examined. Results: Compared with the 100-300 eosinophils/µL group, the ≤100 eosinophils/µL group showed higher frequencies of fever, respiratory failure, and the use of systemic glucocorticoids. Eosinophil counts were not associated with in-hospital mortality and duration of hospitalization. The multivariable analysis showed that GOLD3/4 (odds ratio (OR)=2.04, 95%CI: 1.20-3.44, P=0.008), neutrophil count (OR=1.21, 95%CI: 1.03-1.41, P=0.019), systemic glucocorticoids (OR=1.84, 95%CI: 1.41-2.98, P=0.012), mechanical ventilation (OR=2.66, 95%CI: 1.36-5.18, P=0.004), and acute exacerbation in the past year before hospitalization (OR=2.03, 95%CI: 1.27-3.23, P=0.003) were independently associated with acute exacerbation within 1 year after discharge. Eosinophil count was not associated with acute exacerbation within 1 year after discharge.


Background
Chronic obstructive pulmonary disease (COPD) is characterized by significant airflow limitation associated with a chronic inflammatory response in the airways and lungs, resulting in the destruction of lung tissue [1]. It commonly affects adults >40 years old who smoke, with an estimated worldwide prevalence of 4%-10% [1,2]. The disease course is usually progressive, with a long-term decline in lung function [3]. It is a preventable and treatable disease commonly associated with co-morbidities (such as cardiovascular disease) and significant systemic consequences (such as skeletal muscle dysfunction) [1].
COPD has several complications, including acute exacerbation, respiratory failure, and pulmonary hypertension. The 4-year mortality rates range from 28% for mild-to-moderate COPD to 62% for moderate-to-severe COPD [4,5].
Acute exacerbation of COPD (AECOPD) is characterized by an acute worsening in baseline symptoms such as cough, dyspnea, and/or sputum production beyond normal daily variations to the extent where it requires a change in therapy [1,6]. AECOPD is commonly caused by viral or bacterial infections, including pneumonia and air pollution [1,6,7]. Inhospital mortality for patients with AECOPD is around 2.5% in general and 10% with hypercarbia [1,8]. All-cause mortality within 3 years of hospitalization may be as high as 49% [1,6].
The pathogenesis of COPD involves inflammation-induced structural changes that result in small airway remodeling and narrowing and parenchymal destruction, decreased elastic recoil, and reduction in the ability of the airways to remain open [1]. COPD progression is associated with the accumulation of inflammatory mucous exudate and inflammatory exudate in the airway wall [9]. COPD is a heterogeneous disease among patients.
Eosinophil infiltration was previously thought to be limited to asthma, but it is now known that eosinophil infiltration constitutes a subset of COPD [10][11][12][13], with about 37% of 4 patients with persistently elevated eosinophil counts [14]. Among all patients with COPD, the patients with eosinophil infiltration show the greatest response to corticosteroid therapy [10][11][12][13]. Eosinophil numbers are increased in the sputum and peripheral blood during exacerbation episodes [15,16].
Nevertheless, the exact role of eosinophil counts in the management of patients with COPD remains controversial. The Copenhagen City Lung study showed that high peripheral eosinophil counts (>340 cells/µL) were associated with increased risk of AECOPD [17], as supported by other studies [15,16], but such association was not observed when using a threshold of 2% [14,17,18]. Further contributing to the controversy, elevated blood eosinophil counts are associated with better lung function, quality of life, and mortality rates [14,19,20]. Eosinopenia is associated with sepsis, pneumonia, and worse prognosis of AECOPD [21][22][23].
Studies are lacking about the clinical significance of high eosinophil levels in hospitalized patients with AECOPD. Therefore, the aim of the present study was to examine blood eosinophil levels in relation to the prognosis of hospitalized patients with acute exacerbation of chronic obstructive pulmonary disease.

Study design and patients
This was a retrospective cohort study of patients with AECOPD as their primary diagnosis and 2) complete clinical data, including one-year follow-up data. The exclusion criteria were: 1) bronchiectasis, pulmonary interstitial fibrosis, active tuberculosis, lung cancer, or other diseases; 2) diseases of the blood and endocrine system that affect blood eosinophil count; or 3) confirmed history of bronchial asthma.

Data collection
Basic data and clinical manifestations included sex, age, history of disease (hypertension, type II diabetes, ischemic heart disease, asthma, tuberculosis, and osteoporosis), history of smoking, course of COPD, COPD stage, use of inhaled corticosteroids in a stable period, and acute exacerbation in the past year before hospitalization were collected. The main clinical symptoms of acute exacerbation included wheezing, coughing, and expectoration with/without purulent sputum were also collected.
Results of the first examination after admission were collected, including white blood cell count, neutrophil percentage and count, eosinophil percentage and count, arterial blood gas analysis, sputum culture, chest imaging examination, and lung function (FEV1/FVC, FEV1%, bronchodilatation test).
Treatment and outcome included patients' use of glucocorticoids during hospitalization, including duration of use and total use, duration of hospitalization, and hospital mortality.

Follow-up
6 All patients were followed at 30 days and 1 year after discharge by telephone. Follow-up was censored on 2017.12.31. Whether the patient was hospitalized again due to acute exacerbation within 30 days and 1 year after discharge was recorded (acute exacerbation of re-hospitalization refers to coughing, sputum expectoration, or wheezing). SPSS 20.0 (IBM, Armonk, NY, USA) was used to analyze the data. Categorical data are expressed as n (%) and were analyzed using the chi-square test. Continuous data are presented as means ± standard deviation (SD) and were tested using ANOVA with the LSD post hoc test. Multivariable logistic regression analysis was used to analyze the association of eosinophil counts and percentage with patient outcomes. The regression method was Backward and the included variables had to be significant in the univariable analyses. The results are shown as odds ratio (OR) and 95% confidence interval (95%CI).

Statistical analysis
Two-sided P-values <0.05 were considered statistically significant.

Characteristics of the patients
A total of 1287 patients with AECOPD were first identified, and 132 were excluded because of concomitant asthma and 647 because they were readmissions. Among the 508 evaluable patients, 328, 142, and 38 were in the <100, 100-300, and >300 eosinophils/µL groups, respectively ( Figure 1). Table 1 presents the characteristics of the patients. There were no differences among the three groups, except that the ≤100 eosinophils/µL group, compared with the 100-300 eosinophils/µL group, showed higher frequencies of fever (50.6% vs. 35.2%, P=0.002) and respiratory failure (30.5% vs. 20.4%, P=0.025).  Table 3 presents the clinical outcomes. There were no differences among the three groups 14.8%, P=0.003). Eosinophil counts were not associated with in-hospital mortality and duration of hospitalization. Discussion 8 Studies about the clinical significance of high eosinophil levels in COPD are conflicting, and it has been even less studied in hospitalized patients with AECOPD. Therefore, the aim of the present study was to examine blood eosinophil levels in relation to the prognosis of hospitalized patients with acute exacerbation of chronic obstructive pulmonary disease.

Factors associated with acute exacerbation within 1 year after discharge
The results suggest that peripheral blood eosinophil counts are not associated with the 1year prognosis of AECOPD.
Eosinophils play important roles in inflammatory diseases, especially the initiation and modulation of inflammation [24]. Their role in diseases like asthma has been known for a long time, but they are now known to play roles in diseases like COPD, inflammatory bowel diseases, autoimmune myocarditis, and primary biliary cirrhosis, among others [10-13, 24, 25]. The mechanisms for their association with autoimmune diseases include degranulation of their granule content, autoantibody-dependent cell-mediated cytotoxicity, degradation of the extracellular matrix, cytokine secretion, antigen presentation, induction of fibrosis; and role in thrombosis [24,25]. Patients with underlying eosinophil-associated diseases might be at higher risk of autoimmune diseases [24][25][26][27].
COPD is a heterogeneous disease that differs in etiology, progression, and prognosis from one patient to the other [28,29]. About 37% of patients with COPD display a persistently elevated eosinophil count [14]. Despite the association of eosinophils with inflammatory and autoimmune diseases, the exact clinical implications of those elevated counts in COPD remain controversial [24,25]. On the one hand, the Copenhagen City Lung study showed that eosinophil count >340 cells/µL were associated with a higher risk of AECOPD, but that the threshold of 2% was not [17], as supported by Singh et al. [14] and Zysman et al. [18].
In addition, eosinophil counts are increased in the sputum and peripheral blood during exacerbation episodes [15,16]. Surprisingly when we consider their role in inflammation and autoimmune diseases, elevated peripheral blood eosinophils have been associated with better lung function, improved quality of life, and lower mortality rates in patients with COPD [14,19,20], but this could be due to the fact that patients with COPD and high eosinophil counts respond better to corticosteroids than patients with lower eosinophil counts [10-13, 19, 30, 31]. Low eosinophil counts have also been associated with a higher occurrence of complications of AECOPD, including sepsis, pneumonia, longer hospitalization, and mortality [21][22][23]32].
In the present study, the patients were grouped according to eosinophil counts of ≤100, 100-300, and ≥300 cells/µL, and no association was observed between eosinophil count and acute exacerbation within 1 year after discharge. A recent study suggested that eosinophil counts could predict exacerbation events in patients with COPD, but only in exsmokers [33]. Brusselle et al. [34] showed that eosinophil counts could predict the recurrence of AECOPD in patients who already had a history of AECOPD episodes. Therefore, because COPD is a complex disease, it is possible that eosinophil counts are associated with the prognosis of AECOPD only in subgroups of patients. In addition, future studies should account for the environmental factors associated with the development of COPD and AECOPD [35]. On the other hand, stratification of the patients of the SPIROMICS study according to eosinophil counts showed no association between blood eosinophil counts and AECOPD, while sputum eosinophil counts were associated with AECOPD [36], therefore supporting the present study.
This study has limitations. The sample size was relatively small and from a single center.
In addition, environmental factors like air pollution were not controlled for, and it is known that air pollution is elevated in Chinese cities [37], limiting the generalizability of the study. Finally, due to the retrospective nature of the study, only the data that were included in the patient charts could be analyzed.

Ethics approval and consent to participate
This study was approved by the Ethics Committee of Beijing Shijitan Hospital, Capital Medical University (#2018-10-66). The informed consent was waived because of the retrospective study.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests      Figure 1 Patient flowchart and selection process