Article Text

Download PDFPDF

Evidence for eosinophil activation in bronchiectasis unrelated to cystic fibrosis and bronchopulmonary aspergillosis: discrepancy between blood eosinophil counts and serum eosinophil cationic protein levels
  1. Claus Kroegel,
  2. Marc Schüler,
  3. Martin Förster,
  4. Ruedi Braun,
  5. Paul Reinhard Grahmann
  1. Pneumology, Medical Clinic IV, Friedrich-Schiller-University, Erlanger Allee 101, D-7740 Jena, Germany
  1. Dr C Kroegel.


BACKGROUND Increased serum levels of eosinophil cationic protein (ECP) have been detected in adolescent patients with cystic fibrosis. However, ECP concentrations in adult patients with bronchiectasis unrelated to cystic fibrosis have not been studied.

METHODS Eosinophil numbers and serum concentrations of ECP were determined in 14 patients with known or newly diagnosed bronchiectasis and compared with age and sex matched patients with allergic bronchial asthma, chronic obstructive pulmonary disease (COPD), and controls in whom bronchiectasis or obstructive pulmonary disease could be excluded.

RESULTS Serum ECP levels were significantly raised both in patients with bronchiectasis (median (range) 22.5 μg/l (7–85)) and allergic asthma (35.0 μg/l (7–128)) compared with the sex and age matched subjects suffering from COPD (6.7 μg/l (1.5–28); p<0.006) and non-obstructive normal controls (7.5 μg/l (3.5–19); p<0.003). In contrast, significantly increased peripheral eosinophil numbers were observed in patients with bronchial asthma (305 × 106/l; p<0.01) but not in those with bronchiectasis (102 × 106/l), COPD (117 × 106/l), and healthy controls (101 × 106/l).

CONCLUSIONS The discrepancy between eosinophil counts and eosinophil numbers in patients with bronchiectasis suggests that serum ECP levels may be more relevant in assessing local eosinophil involvement than blood eosinophil numbers.

  • bronchiectasis
  • asthma
  • chronic obstructive pulmonary disease
  • eosinophil number
  • eosinophil cationic protein

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Bronchiectasis is associated with a vigorous bronchial inflammation due to frequent or persistent bacterial colonisation and increased numbers of neutrophils and neutrophil-derived products found in both bronchoalveolar lavage (BAL) fluid and histological specimens during the course of infection.1 Unlike neutrophils, eosinophils in BAL fluid and affected bronchoalveolar tissue have been detected in small numbers only.2 However, three recent reports3-5 have described increased levels of the eosinophil cationic protein (ECP) in the sputum and/or serum, indicating that eosinophils and their products may also contribute to the inflammatory process in cystic fibrosis. Since no data are available for subjects with bronchiectasis unrelated to cystic fibrosis, we prospectively assessed the serum ECP concentrations and blood leucocyte counts in adult patients admitted to our clinic in whom bronchiectasis had been previously established or subsequently diagnosed.



Between January 1992 and August 1994, 14 consecutive patients (eight men) with previously proven or newly diagnosed bronchiectasis seeking medical advice in the Department of Pneumology were included in the study. The patients were selected on the basis of (1) a history of persistent cough, (2) recurrent pneumonias and frequent haemoptysis, (3) large quantities and partially foul purulent sputum production, (4) persistently positive sputum cultures for more than three years, and (5) radiological evidence of bronchiectasis detected either by bronchography (n = 8) and/or high resolution computed tomographic (CT) scanning (n = 12). All patients had localised bronchiectases confined to one or two lobes.

The definite cause of bronchiectasis could not be found. None of the patients reported a family history of similar lung disease. In addition, hereditary diseases, including ciliary immotility syndrome, α1-antitrypsin deficiency, and cystic fibrosis as well as both IgG and IgA deficiency were ruled out, suggesting that the bronchiectasis was most probably acquired.

In 10 patients, infection with one or more of Pseudomonas aeruginosa, Streptococcus pneumoniae orHaemophilus influenzae was established by sputum culture.Candida spp were observed in the sputum of three patients. There was no clinical and serological evidence for viral infection or bronchopulmonary aspergillosis during the time of evaluation. In addition, sputum cultures did not reveal Aspergillusorganisms or other fungi, and dermal hypersensitivity toAspergillus antigen was absent. In two patients the total serum IgE levels were slightly above the upper normal limit (124 and 187 U/l). However, neither elevated specific IgE nor a positive skin test for common allergens, including grass pollen, birch pollen, and house dust mite, were found in these patients. In addition, no history of intermittent wheezing, chest tightness, and no family history of atopic diseases were present. None of the patients had received steroids or antibiotic treatment during the four weeks prior to admission.

The 14 adult patients (mean age 54.8 years; range 31–78) with bronchiectasis were compared with three sex and age matched control groups consisting of 14 patients each with either previously established allergic asthma, chronic obstructive pulmonary disease (COPD), or healthy controls in whom pulmonary disease was excluded. Asthma and COPD were diagnosed according to the criteria published previously.6 All patients were without clinical signs of a current infectious exacerbation during the previous four weeks. Informed consent for blood sampling was obtained in all cases. All subjects underwent a thorough clinical examination and lung function was assessed using whole body plethysmography.


Venous blood samples were collected at routine sampling for clinical evaluation and left at room temperature for 60 minutes as recommended by the manufacturers. After centrifugation at 1000g for 10 minutes at 4°C, plasma was frozen and stored at –20°C until further use.7 ECP was measured using a double antibody radioimmunoassay (Pharmacia, Uppsala, Sweden) (coefficient of variation 8.5%). In addition, total leucocyte counts were measured using standard cytometry (coefficient of variation 7.4%).


The two tailed Mann-Whitney U test and Kruskal-Wallis test were used to compare values in two or more groups, respectively. Correlations between continuous variables were calculated using Spearman’s rank test.


Total blood leucocyte counts were similar between all groups, with median values ranging from 6.5 × 109/l (controls) to 7.1 × 109/l (asthmatics). In addition, similar counts of neutrophils were observed ranging from 3.9 to 13.2 × 109/l with the highest median of 5.4 × 109/l (5.3–24 × 109/l) in patients with bronchiectasis followed by allergic asthmatic patients (4.3 × 109/l; 1.4–7.3 ×109/l), controls (4.1 × 109/l; 3.3–7.2 × 109/l), and patients with COPD (3.8 × 109/l; 3.5–12.8 × 109/l). Total leucocyte numbers and neutrophil counts were not statistically different between the groups studied (p > 0.05).


Blood eosinophil numbers in patients with bronchiectasis ranged from 1 to 32 × 107/l with a median of 10.2 × 107/l (fig 1A), and were not different from either controls (10.1 × 107/l; range 1.6–21.4) or patients with COPD (11.7 × 107/l; range 0.6–31.5) (fig 1A). However, significantly raised eosinophil counts were observed in the blood of subjects with allergic asthma (30.5 × 107/l; range 12.3–69.3) compared with that of other groups (all p<0.05).

Figure 1

(A) Absolute blood eosinophil counts and (B) serum ECP concentrations in 14 patients with bronchiectasis compared with age and sex matched control groups. Data given are individual cell counts. Bars indicate median values.


Serum ECP levels in the healthy control group ranged from 3.5 to 19 μg/l (median 7.5 μg/l), which is close to the ECP concentrations observed in patients suffering from COPD (1.5–28 μg/l; median 6.7 μg/l) (fig 1B). Patients with allergic asthma showed significantly raised serum ECP levels, ranging from 7 to 128 μg/l with a median of 35.0 μg/l (p<0.001 versus both controls and patients with COPD). In contrast to the eosinophil numbers, all except two subjects with bronchiectasis showed an increase in serum ECP concentration equal to or above the median concentration measured within the control and COPD groups. The median ECP value of 22.5 μg/l (range 7–85 μg/l) for patients with bronchiectasis was significantly higher than that of control subjects (p<0.003) or subjects with COPD (p<0.006). In contrast, no difference was detected between ECP serum levels in patients with bronchiectasis or bronchial asthma (p>0.05). In addition, no correlation was found between eosinophil counts and ECP levels among control subjects (r 2 = 0.09; p>0.05), asthmatics (r 2 = 0.002; p>0.05), and patients suffering from bronchiectasis (r 2= 0.02; p>0.05) or COPD (r 2 = 0.001; p>0.05).


The present study provides evidence that adult patients with bronchiectasis unrelated to either cystic fibrosis or bronchopulmonary aspergillosis show significantly increased serum ECP levels compared with sex and age matched controls and patients suffering from COPD. The magnitude of ECP values associated with bronchiectasis was similar to the concentrations measured in the serum of allergic asthmatic subjects. However, in contrast to patients with bronchial asthma, blood eosinophil counts in those with bronchiectasis were not increased above those obtained from normal controls or patients with COPD. Finally, the number of peripheral neutrophils was not different among the four groups studied.

To our knowledge, raised serum ECP in adult patients with bronchiectasis has not been previously described, and the reason for this observation in the absence of elevated blood eosinophil counts remains unclear. Possibly, secondary allergic sensitisation observed in bronchiectasis may have caused the increased ECP levels observed.1 However, none of the patients with bronchiectasis had a positive skin test or raised allergen specific IgE levels against common allergens. Thus, although we cannot totally exclude the possibility that a subclinical allergic component in our patients may have contributed to the increase in ECP levels, it appears unlikely to represent a dominant factor.

A central finding of our study is the discrepancy between the raised serum ECP levels and the number of blood eosinophils in patients with bronchiectasis, which were not different to those of the controls or patients with COPD. In subjects with allergic asthma, however, both serum ECP concentrations and eosinophil counts were significantly above the normal range, which is in good agreement with previous published studies.8 Though the reasons for the apparent discrepancy between raised ECP levels and normal eosinophil counts remain to be elucidated, data reported by Koller et al in adolescent subjects with cystic fibrosis suggest that peripheral eosinophils may be in a preactivated state, possibly due to the action of circulating mediators such as cytokines. However, future studies assessing serum cytokine concentrations from patients with bronchiectasis are required to substantiate this hypothesis.

In summary, together with previously published studies,3-5 the present data suggest that eosinophil derived granular proteins are increased in patients with lung disorders associated with bronchiectatic tissue destruction despite normal eosinophil counts. We therefore conclude that ECP does not merely reflect the absolute number of eosinophils but represents a more meaningful indicator for eosinophil participation in disease.


This work was supported by the County of Thüringia, Germany (01KC8906/1).