Are sputum eosinophil cationic protein and eosinophils differently associated with clinical and functional findings of asthma?

Clin Exp Allergy. 2014;44(5):673-80. doi: 10.1111/cea.12236.



Silvana Cianchetti1, Elena Bacci1, Lucia Ruocco2, Tiziana Pavia2, Maria Laura Bartoli1, Cristina Cardini1, Francesco Costa1, Antonella Di Franco1, Laura Malagrinò1, Federica Novelli1, Barbara Vagaggini1, Alessandro Celi1, Federico Dente1 and Pierluigi Paggiaro1

  1. Cardio-Thoracic and Vascular Department, University of Pisa, Italy
  2. Laboratory Unit, University Hospital of Pisa, Italy



Silvana Cianchetti
Cardio-Thoracic and Vascular Department
Ospedale di Cisanello
Via Paradisa 2, 56124 Pisa, Italy



Background. Sputum eosinophil counts and ECP levels are usually increased in asthmatic patientsThe correlation between sputum eosinophils or ECP and clinical findings of asthma has been previously investigated but many of these studies have been performed on small samples of asthmatic patients, considering only few clinical indices and often including patients on oral or inhaled corticosteroids, which might be confounding when interpreting the relationship between disease activity and airway inflammation.

Objective. To assess whether sputum eosinophils and ECP were differently related to functional and clinical parameters of asthma in a large number of steroid-naïve asthmatic patients, taking into account several potential determinants of activity and chronicity of asthma.

Methods. 129 patients with mild-moderate asthma were studied. Sputum was induced by hypertonic saline inhalation and processed using the whole sample method.

Results. Sputum eosinophils and ECP significantly correlated to each other (r=0.41, p<0.001). In the overall sample, disease duration inversely correlated with sputum eosinophils, whereas FEV1 and PEF inversely correlated with sputum ECP. Rescue β2-agonist use and total symptom score positively correlated with both eosinophil counts and sputum ECP. Stepwise regression analysis showed that symptom score and disease duration accounted for 17.6% of sputum eosinophil variance, whereas symptom score and FEV1 accounted for 14.7% of sputum ECP variance.

Conclusions and Clinical Relevance. Both sputum eosinophils and ECP are weakly related to clinical markers of asthma severity. However, ECP was more closely related to lung function parameters than eosinophil counts.

Key words: asthma – eosinophils – eosinophilic cationic protein – induced sputum

PMID: 24245689



Asthma is a chronic inflammatory airway disease where eosinophils are potent effector cells. These cells release inflammatory mediators such as eosinophil cationic protein (ECP), upon activation by several environmental stimuli and intercellular adhesive interactions. ECP mediates several pathologic effects on bronchial epithelial cells, mucosal glands and airway smooth muscle [1].

Several studies have shown that ECP can be found in induced sputum at different concentrations, according to the severity of the disease. In general, data on its relationship with sputum eosinophil levels are inconsistent, probably as a result of different degrees of cell activation [2]. The correlation between sputum eosinophils or ECP and clinical and functional findings has shown conflicting results, with some studies showing substantial equivalence between these two sputum markers, others reporting better correlation of ECP to pulmonary function and/or bronchial hyperresponsiveness. However, many of these studies, performed on small samples of asthmatic patients, considered only few clinical indices and often included patients on oral or inhaled corticosteroids, which might be confounding when interpreting the relationship between disease activity and airway inflammation. Furthermore, the relationship between inflammatory markers and clinical or functional findings may be different according to whether indicators of disease “activity” (current symptoms, exacerbation rate, rescue medication use) or “chronicity” (disease duration, pulmonary function level) are considered.

We therefore studied a large number of corticosteroid-naïve, mild/moderate asthmatic patients, in order to assess whether sputum eosinophils and ECP are differently related to functional and clinical indices of the disease, thus providing further clues in asthma management.

We performed a retrospective analysis of the data obtained in 186 steroid-naive patients with mild/moderate asthma recruited from the asthma clinic of the Cardiothoracic and Vascular Department. The diagnosis of asthma was made according to internationally accepted criteria [3], in the presence of typical symptoms and risk factors, reversible airway obstruction and/or non-specific bronchial hyperresponsiveness to methacholine.

All patients were examined in a stable period of the disease, and patients with exacerbation of asthma in the last 4 weeks were excluded from analysis. None of the patients had received anti-inflammatory treatment in the last 3 months, and at the time of the study all patients had active symptoms and used bronchodilators as rescue medication. As part of our clinical routine, all patients performed: a) spirometry; b) peak expiratory flow (PEF) monitoring and symptom score recording on a diary card for 2 weeks; c) sputum induction with hypertonic saline (HS).


Table 1. Characteristics of asthmatic subjects

Sex, M/F 47/82
Age, yrs 37.8 ± 15
Smoking history, y/n 23/106
Atopy, y/n 87/42
Disease duration, yrs 8 (0.2-50)
Severity, mild/moderate 35/94
Daily β–agonist, puffs/day 0.6 (0-5.2)
Daily symptom score 1.1 (0-4.6)
FEV1, % pred 89.9 ± 15.2
Mean PEF, % pred 77.0 ± 18.7
PD20 FEV1, µg 123 (5-1294)

Age, FEV1 and PEF are expressed as mean±SD; Disease duration, Daily β–agonist , Daily symptom score are expressed as median (range); PD20FEV1 as Geometric mean. FEV1: forced expiratory volume in 1 s; PEF: peak expiratory flow; PD20FEV1, provocative dose of methacholine causing a 20% fall in FEV1.


A total of 27 out of 186 selected patients did not collect an adequate sputum sample; furthermore, 30 out of 159 patients showed a sputum sample with squamous cell percentages > 60%. All these 57 patients were not considered for analysis. Mean cell viability of sputum samples was 73.9+10.6% and mean percentage of squamous cells was 31.4+10.2%. Sputum eosinophils were > 2% in 65.1% of patients, whereas sputum ECP levels were > 75 mg/L in 55.8% of patients. The characteristics of subjects analyzed (n=129) are reported in Table 1.

 Figure 1. Eosinophil percentages and ECP concentrations (in log scale) in sputum of steroid-naive asthmatic patients. Horizontal and vertical lines show the cut-off values (normal upper limits) for ECP (75µg/L) and eosinophils (2%), respectively.

a: patients with eosinophils >2% and ECP >75µg/L; b: patients with eosinophils <2% and ECP <75µg/L; c: patients with eosinophils >2% and ECP <75µg/L; d: patients with eosinophils <2% and ECP >75µg/L.

Sputum eosinophils and ECP concentrations were significantly correlated to each other, both when eosinophils were expressed as percent of total inflammatory cells (r=0.41, p<0.001) and when expressed as absolute count/mL (r=0.46, p<0.001). There was, however, some data scattering, thus showing that the concordance between the two measurements was weak (Figure 1).

Because several clinical and functional variables were found to be associated with sputum eosinophils and/or ECP, multiple stepwise regression analysis was performed. The regression model included all the independent variables showing a significant correlation with sputum eosinophils or ECP in the single regression analysis, (disease duration, daily symptoms, FEV1), excluding independent variables related to each other (in particular daily β–agonist use, related to daily symptoms, and PEF, related to FEV1) (Table 2). Stepwise regression analysis showed that daily symptoms and disease duration predicted a proportion of the variance of sputum eosinophils with daily symptom score giving the greatest contribution. The results of stepwise regression for ECP showed that both daily symptoms and FEV1% entered the model, representing significant independent determinants for ECP variance, although daily symptoms provided the greatest contribution.

The aim of the present study was to evaluate whether sputum eosinophil counts and ECP concentrations are differently related to clinical and functional indices. We found that sputum eosinophil counts and ECP concentrations were both related to indices of asthma “activity”, such as current symptoms and rescue medication use, but differently related to indices of asthma “chronicity”: sputum eosinophil counts inversely correlated with the duration of the disease, whereas sputum ECP concentrations inversely correlated with pulmonary function, expressed as FEV1 and PEF. This observation may suggest that these two markers of eosinophilic inflammation, although concordant in their relationship with current asthma symptoms (markers of “activity”), may provide different clues about the chronic status of the disease, in terms of both disease duration and pulmonary function impairment.


Table 2. Stepwise regression analysis of the relationship between sputum eosinophil percentages or ECP levels and clinical/functional indices. The independent variables are listed according to the order of inclusion in the regression models.

Included variablesin regression model Global variance:adjusted R2 Partial regressionCoefficient ß (SE) p value
EOSINOPHILS: 0.176   <0.0001
Daily symptom score   0.34 (0.059) <0.001
Disease duration, yrs   -0.29 (0.005) <0.01
ECP: 0.147   <0.0001
Daily symptom score   0.30 (0.072) <0.001
FEV1, % pred   -0.19 (0.005) <0.05



Our study has some advantages over others previously published. First, the number of patients we studied is higher than that reported in previous studies [4]. Second, unlike in other studies, all our patients were corticosteroid-naïve in order to rule out the hypothesis that discrepancies between eosinophils and ECP might be due to their different sensitivity to corticosteroids.

In conclusion, our results demonstrate that sputum eosinophils and ECP are partly correlated to each other, but differ in their relationship with clinical and functional findings of asthma. Sputum eosinophils seem to be related more to the active inflammatory process that may fade with time, whereas sputum ECP seems to be related more to the reduced lung function.



1) Frigas E, Motojima S, Gleich GJ. The eosinophilic injury to the mucosa of the airways in the pathogenesis of bronchial asthma. Eur Respir J Suppl. 1991;13:123s-135s.

2) Tsoumakidou M, Papadopouli E, Tzanakis N, Siafakas NM. Airway inflammation and cellular stress in noneosinophilic atopic asthma. Chest. 2006; 129:1194-1202.

3) Global Initiative on Asthma. Global strategy for asthma management and preventionNHLBI workshop report. National Institutes of Health. Bethesda, MD: publication 95–3659, update December 2010.

4) Ronchi MC, Piragino C, Rosi E, Stendardi L, Tanini A, Galli G, Duranti R, Scano G. Do sputum eosinophils and ECP relate to the severity of asthma? Eur Respir J. 1997; 10:1809-13.