The first two authors contributed equally to this work.
Clinical characteristics of eosinophilic and noneosinophilic asthma in children
Article first published online: 1 NOV 2012
©2012 The Author(s)/Acta Pædiatrica ©2012 Foundation Acta Pædiatrica
Volume 102, Issue 1, pages 53–57, January 2013
How to Cite
Lee, Y. J., Kim, K. W., Choi, B. S., Sohn, M. H. and Kim, K.-E. (2013), Clinical characteristics of eosinophilic and noneosinophilic asthma in children. Acta Paediatrica, 102: 53–57. doi: 10.1111/apa.12046
- Issue published online: 11 DEC 2012
- Article first published online: 1 NOV 2012
- Manuscript Accepted: 28 SEP 2012
- Manuscript Revised: 24 SEP 2012
- Manuscript Received: 3 JUN 2012
- Korea Healthcare Technology R&D Project, Ministry for Health, Welfare and Family Affairs
- Airway inflammation;
- Asthma in children;
Asthma is a heterogeneous and complex chronic inflammatory disease of the airways. Asthma can be classified as eosinophilic asthma (EA) or noneosinophilic asthma (NEA). We investigated whether children with EA manifest different clinical characteristics than those with NEA.
We enrolled 288 steroid-naive asthmatic children and classified them, based on the cell counts in induced sputum, into EA (158 children) and NEA (89 children) groups.
No significant differences were observed between the groups with regard to age, sex, family history of atopy, secondary smoking or asthma exacerbations. Moderate-to-severe asthma was more frequent in the EA group than in the NEA group. Blood eosinophil counts and serum eosinophil cationic protein were higher in EA patients than in NEA patients. The forced expiratory volume in 1 sec was lower in children with EA than in those with NEA (% of predicted value, 88.6 ± 18.5 vs. 93.6 ± 15.6, p < 0.05). The sputum eosinophil (in EA) and neutrophil (in NEA) counts increased with increasing asthma severity.
Airway inflammation, especially eosinophilic inflammation, was associated with asthma severity and reduced pulmonary function in children.
- Moderate-to-severe symptoms are more frequent in eosinophilic asthma (EA) than in noneosinophilic asthma (NEA) in children.
- Eosinophilic inflammation is associated with severity of EA.
- Neutrophilic inflammation is associated with severity of NEA.
Asthma is a heterogeneous disease characterized by various degrees of airway obstruction, airway hyper-responsiveness (AHR) and chronic airway inflammation. There are several biologic markers for the diagnosis and assessment of asthma, including blood eosinophil levels, serum eosinophil cationic protein (ECP), exhaled nitric oxide and induced sputum. Among these biomarkers, induced sputum is useful parameter for assessing airway inflammation because it reflects both upper and lower airway inflammation . Asthma can be classified as eosinophilic asthma (EA) or noneosinophilic asthma (NEA) based on the induced sputum .
In adults, EA, as opposed to NEA, has been associated with more severe asthma , more severe airway obstruction, increased airway reactivity to methacholine  and a better response to treatment with inhaled corticosteroids (ICS) . NEA has been reported to be associated with refractory asthma , pollutants, smoking, occupational allergens and obesity . However, a recent study has shown that the patterns of the inflammatory phenotypes differ between adults and children , and there are only a few studies concerning the relationship between clinical characteristics and airway inflammation. In children, cough frequency  and nonatopic asthma  were associated with sputum neutrophilia, while frequent asthma exacerbations were associated with sputum eosinophilia .
In this study, we investigated whether children with EA and those with NEA (as defined by the inflammatory patterns of induced sputum) manifested with different clinical characteristics. Additionally, we studied the extent to which airway inflammation in each phenotype was associated with asthma severity, pulmonary function and AHR.
Patients and Methods
This study enrolled 288 children who visited the Asthma and Allergy Clinic at the Severance Children's Hospital from July 2006 to July 2009 to confirm asthma because of recurrent symptoms of asthma. Their ages ranged from 6 to 15 years. The inclusion criteria were recurrent symptoms of wheezing, dyspnoea or chest tightness during the previous year, a provocative concentration of methacholine causing a 20% decrease in FEV1 (PC20) of <16 mg/mL or a change in forced expiratory volume in 1 s (FEV1) of ≥12% between pre- and postbronchodilator (BD) inhalation. The symptoms were reported by the parents. The exclusion criteria were the use of inhaled or systemic corticosteroids or leukotriene receptor antagonists, infection of the respiratory tract within 1 month of enrolment and refusal to undergo sputum test or inability to produce adequate sputum. This study was approved by the Institutional Review Board of the Severance Hospital. Written consent for participation was obtained from the parents with verbal assent from the children.
Forty-one children failed to produce adequate sputum and were excluded from the analysis. Their average age was 7.6 (mean) ± 1.8 (standard deviation) years, compared with 8.7 ± 2.8 years in attending children (p < 0.05). Their clinical characteristics, blood eosinophil levels, serum ECP, pulmonary function results and PC20 values were not significantly different from those of NEA patients.
Sputum eosinophilia was defined as eosinophils comprising >2.5% of the cells counted in induced sputum . Patients with sputum eosinophilia were classified as having EA and patients without sputum eosinophilia as having NEA.
Asthma severity was classified as intermittent, mild persistent or moderate-to-severe persistent based on the extent of symptoms and the pulmonary function results. Patients in the intermittent group had daytime symptoms less than or once a week, nocturnal symptoms less than or twice a month and pre-BD FEV1 (% of predicted value) ≥ 80%. The mild persistent group had day symptoms more than once a week but less than once a day, nocturnal symptoms more than twice a month and pre-BD FEV1 (%) ≥ 80%. The moderate-to-severe persistent asthma patients had symptoms daily, nocturnal symptoms more than once a week and pre-BD FEV1 (%) < 80%. Atopy was defined by a positive skin prick test upon exposure to one or more extracts of 12 common aeroallergens or a serum total immunoglobulin E (IgE) concentration of ≥150 IU/mL. Nonatopy was defined by a negative skin prick test with a serum total IgE concentration of <150 IU/mL . The allergens tested included Dermatophagoides pteronyssinus, Dermatophagoides farinae, cat and dog epithelia, sagebrush, ragweed, oak, birch, Aspergillus spp., Alternaria spp. and American and German cockroach antigens (Torii & Co., Tokyo, Japan). Saline was used as a negative control, and 0.5% histamine HCl was used as a positive control. The wheal diameter was measured after 15 min, and a positive reaction was defined as a wheal diameter ≥3 mm. A parental history of atopy was present when at least one parent had atopic dermatitis, asthma or allergic rhinitis.
The pulmonary function and methacholine challenge tests were performed on different days and when the patients were stable and without acute exacerbation. A pulmonary function test (VIASYS Healthcare Inc., Conshohocken, PA, USA) was performed according to ATS guidelines before and after BD inhalation . A methacholine challenge test was performed according to a standardized method . Each subject inhaled increasing concentrations of methacholine (0.075, 0.15, 0.31, 0.62, 1.25, 2.5, 5, 10, and 25 mg/mL) nebulized by a Mefar MB3 dosimeter (Brescia, Italy) until FEV1 was reduced by 20% from a postnebulized saline value. Bronchial response to methacholine was expressed as a provocative concentration of methacholine causing a 20% decrease in FEV1 (PC20 in mg/mL) and was calculated by linear interpolation of the log dose–response curve . Patients with an estimated pre-BD FEV1 (%) of <70% did not undergo the methacholine challenge test.
Sputum induction and processing were performed as previously described . The induced sputum was considered adequate if the patient inhaled 3% saline for longer than 4 min, the induced sputum volume was ≥1 mL, the sample contained >400 inflammatory cells and the proportion of squamous epithelial cells was ≤80% .
The NE-8000 system (Sysmex, Kobe, Japan) was used to count eosinophils automatically in peripheral blood, while serum total IgE, specific IgE and ECP were measured with the CAP system (Pharmacia-Upjohn, Uppsala, Sweden).
Statistical analyses were performed using the Statistical Package for the Social Sciences (version 15.0; SPSS, Inc., Chicago, IL, USA). Numerical variables are expressed as means and standard deviations. Numeric data were compared using the two-sample t-test. Frequencies were compared using the chi-square test. Analysis of variance (ANOVA) was used to compare sputum inflammatory cells or pulmonary function among the asthma severity groups. A p-value of <0.05 was considered statistically significant.
There were 158 children who had EA and 89 who had NEA. The frequency of moderate-to-severe persistent asthma in the EA group was greater than that in the NEA group (Table 1). Blood eosinophils and serum ECP were higher in the EA than in the NEA group (Table 2). Both pre-BD and post-BD FEV1 were lower in the EA than in the NEA (Table 2). However, pre-BD FEV1/FVC values and changes in FEV1 did not differ between the groups. When the lung function findings were compared separately in the subgroups constructed according to the severity of asthma, the differences between the EA and NEA groups were seen in children with mild asthma but not in those with moderate-to-severe asthma (data not shown).
|Characteristic||EA group (n = 158)||NEA group (n = 89)|
|Age (year)||8.6 ± 2.5||8.9 ± 3.4|
|Sex [M/F (%)]||94 (59)/64 (41)||60 (67)/29 (23)|
|Atopy, n (%)||137 (87)||69 (77)|
|Parental history of atopy, n (%)||84 (53)||45 (51)|
|Secondary smoking, n (%)||37 (23)||25 (28)|
|Household pets, n (%)||9 (6)||4 (5)|
|Number of asthma exacerbations in the last 12 months, n (%)|
|0||101 (64)||62 (70)|
|1||41 (26)||20 (23)|
|≥2||16 (10)||7 (7)|
|Asthma severity (%)a|
|Intermittent||26 (16)||27 (30)|
|Mild persistent||83 (52)||44 (50)|
|Moderate-to-severe persistent||49 (31)||18 (20)|
|Parameter||EA group (n = 158)||NEA group (n = 89)|
|Blood eosinophils (per μL)||542 ± 343a||371 ± 338|
|Total IgE (IU/mL)||600 ± 681||529 ± 609|
|Serum ECP (μg/L)||46.6 ± 43.6a||30.4 ± 36.7|
|Number of positive skin tests||2.3 ± 1.7||2.1 ± 2.2|
|Abnormal FEV1, n (%)||41 (26)||15 (17)|
|Pre-BD FEV1 (% predicted)||88.6 ± 18.5b||93.6 ± 15.6|
|Pre-BD FEV1/FVC (%)||86.7 ± 11.2||88.3 ± 10.0|
|Pre-BD FEF25–75% (% predicted)||77.9 ± 30.6b||87.1 ± 31.1|
|Post-BD FEV1 (% predicted)||97.1 ± 18.3b||102.0 ± 15.8|
|Change in FEV1 (%)||10.7 ± 10.9||8.9 ± 20.2|
|PC20 (mg/mL)||8.5 ± 12.9||10.8 ± 14.3|
The mean percentages of macrophages, neutrophils, eosinophils and lymphocytes in the induced sputum samples were 46.7 ± 25.4% (mean ± standard deviation), 30.6 ± 22.6%, 22.1 ± 20% and 0.2 ± 1%, respectively, in the EA group, and 61.8 ± 6.4% (p < 0.001), 37.6 ± 26.2% (p < 0.05), 0.4 ± 0.6% and 0% (p < 0.001), respectively, in the NEA group.
Sputum eosinophils increased with asthma severity in the EA patients (Fig. 1). However, there was no association between asthma severity and sputum neutrophils in the EA group (data not shown). Similar to sputum eosinophils in the EA patients, in the NEA patients, the percentage of sputum neutrophils in the moderate-to-severe persistent asthma group was higher than in the intermittent asthma and the mild persistent asthma groups (Fig. 2).
In this study, we found that moderate-to-severe persistent asthma was more frequent in EA than in NEA and that children with EA had worse pulmonary function results than those with NEA. There were significant differences in levels of eosinophilic biomarkers between EA and NEA patients.
In the intermittent and mild persistent asthma groups, the pre-BD FEV1, pre-BD FEF25–75%, and post-BD FEV1 were lower in the EA patients than in the NEA patients. However, we did not observe the same results in the moderate-to-severe persistent groups. This could indicate that airway inflammation decreases lung function in patients with severe asthma regardless of its pattern. It is also possible that factors other than asthma severity and airway inflammation, such as age of asthma onset, genetics, infection and lung structure, affect lung function in patients with severe asthma. The changes in FEV1 and PC20 did not differ between the groups. These findings probably indicate that eosinophilic inflammation affects pulmonary function rather than AHR or airway reversibility. On the other hand, pre-BD FEV1/FVC did not differ significantly between the groups. Most of the patients in the study had intermittent or mild persistent asthma and normal pulmonary function, which may be why the pre-BD FEV1/FVC results did not associate significantly with asthma severity in asthmatic children with normal pre-BD FEV1 .
Eosinophilic airway inflammation is known to be associated with decreased pulmonary function, frequent exacerbations and poor prognosis [9, 17]. We used induced sputum to evaluate airway inflammation. According to previous studies, induced sputum has been a good tool for the evaluation of airway inflammation because it has been found to be highly correlated with bronchial wash results, but is less invasive [18, 19]. However, it is limited in that it is an indirect method. The exhaled nitric oxide (eNO) concentration has also been reported to be closely correlated with sputum eosinophil counts in adults with asthma  and can be useful to identify EA in a less invasive manner than induced sputum. This parameter was not investigated in the current study of paediatric asthma patients, and further studies are needed to evaluate its utility of eNO compared with induced sputum for the evaluation of airway inflammation.
In the present study, total blood eosinophil counts and serum ECP levels were higher in the EA patients than in the NEA patients. Blood eosinophil counts and serum ECP were weakly associated with sputum eosinophil counts in the whole study population, but not in the EA groups (data not shown). Blood eosinophil and serum ECP levels may also be increased in other allergic conditions, such as atopic dermatitis, and allergic rhinoconjunctivitis. Therefore, blood eosinophilic markers may have limited use as predictors of eosinophilic airway inflammation.
The ratios of patients with atopy, number of positive skin tests and total IgE levels were high and similar in both groups. These findings were different from those previously reported, and NEA has been reported to be associated with nonatopic factors including pollutants, smoking, occupational allergens and obesity . In children, nonatopic asthma has been associated with sputum neutrophilia . This difference might have been because some patients were already sensitized to allergens but did not develop eosinophilic airway inflammation. The frequency of asthma exacerbation in the last year was low in both of our study groups because asthma exacerbation was defined as a condition that a patient needed systemic steroids to avoid hospitalization or intubation.
This study had a few limitations. First, we did not evaluate the reproducibility of the induced sputum. Although the induced sputum test is a simple method of evaluating airway inflammation, it is uncomfortable for children to perform it repeatedly. Moreover, many of our patients had mild persistent asthma but had not undergone treatment with ICS before this study. ICS were administered immediately after the tests, which were therefore not performed repeatedly. Second, we used the relative percentage of inflammatory cells in sputum to represent airway inflammation, without indicating the absolute cell counts. However, it has been suggested that absolute cell counts are relatively poorly reproducible and able to yield only limited information [21, 22].
In conclusion, we found that pulmonary function in children with EA was more reduced than in those with NEA and that asthma severity was affected by the degree of eosinophilic inflammation in EA or neutrophilic inflammation in NEA.
This research was supported by grant A090399 from the Korea Healthcare Technology R&D Project, Ministry for Health, Welfare and Family Affairs, Republic of Korea.
- 12Standardization of spirometry, 1994 update. American Thoracic Society. Am J Respir Crit Care Med 1995; 152: 1107–36., , , , , , et al.