INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Eosinophilic bronchitis presents with chronic cough and is characterized by sputum eosinophilia similar to that seen in asthma, but unlike asthma the patients have no symptoms or objective evidence of variable airflow obstruction or airway hyperresponsiveness (1, 2). We (3) and others (4) have shown that eosinophilic bronchitis is a common cause of cough in patients presenting to a respiratory specialist. The etiology of eosinophilic bronchitis and the reason for the absence of lower airway hyperresponsiveness in this disease is unknown. One possible explanation is that the eosinophilic airway inflammation is less active than asthma, with less release of important effector mediators.

We and others have shown that various proinflammatory mediators, including cysteinyl-leukotrienes (cys-LT), prostanoids (PG) (5), and eosinophilic cationic protein (ECP) (6) can be measured in the induced sputum cell-free supernatant repeatably. We have used this technique to test our hypothesis by measuring eicosanoids, ECP, and histamine in the cell free supernatant from induced sputum in patients with eosinophilic bronchitis or asthma and in normal subjects.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Patients with eosinophilic bronchitis or asthma, and healthy volunteers were recruited from respiratory outpatient clinics and from staff at The Glenfield Hospital. The subjects with eosinophilic bronchitis (n = 8) had an isolated cough, no symptoms suggesting variable airflow obstruction, normal spirometric values, normal peak expiratory flow (PEF) variability (maximum within-day amplitude, % mean < 20% over 2 wk), a methacholine PC₂₀ > 16 mg/ml, a normal chest radiograph, and a sputum eosinophilia (> 3% nonsquamous cell). None had received oral or inhaled corticosteroids for at least 1 mo before entry into the study. Subjects with asthma (n = 17) gave a suggestive history and had objective evidence of variable airflow obstruction, as indicated by one or more of the following: (1) methacholine airway hyperresponsiveness (PC₂₀ < 8 mg/ml); (2) > 15% improvement in FEV₁ 10 min after 200 µg albuterol; or (3) PEF (> 20% maximum within-day amplitude from twice daily PEF measurements over 14 d). The subjects with asthma were clinically stable and treated, as required, with either -agonists only (n = 9) or with inhaled corticoste-roids and, as required, -agonists (n = 8). All the subjects with asthma had a sputum eosinophil count > 3%. The healthy subjects (n = 10) gave no history of respiratory diseases, had negative allergen skin prick tests, normal spirometry, and normal airway responsiveness. The sputum eicosanoid levels in the normal and asthmatic subjects have been previously reported (5). All subjects gave written informed consent to participate in the study. The protocol was approved by the Leicestershire Health Authority ethics committee.

Protocol and Clinical Measurements

Subjects attended on two occasions. At the first visit they had spirometry and allergen skin prick tests. At the second visit the subjects had a methacholine inhalation test followed on recovery by sputum induction. Spirometry was performed using a dry bellows spirometer (Vitalograph, Buckingham, UK) with the FEV₁ recorded as the best of successive readings within 100 ml. Allergen skin prick tests were performed to Dermatophagoides pteronyssinus, cat fur, grass pollen, and Aspergillus fumigatus solutions with normal saline and histamine controls (Bencard, Brentford, UK). A positive response to an allergen on the skin prick tests was recorded in the presence of a weal > 2 mm more than the negative control. The methacholine challenge was performed using the tidal breathing method (7) with doubling concentrations of methacholine (0.03 to 16 mg/ml) nebulized via a Wright nebulizer. Sputum was induced and processed as previously described (6, 8). Briefly, sputum was induced using 3, 4, and 5% saline inhaled in sequence for 5 min via an ultrasonic nebulizer (Medix, Harlow, UK; output 0.9 ml/min; mass median diameter, 5.5 µm). After each inhalation patients blew their noses and rinsed their mouths to minimize nasal contamination and expectorated sputum into a sterile pot. Subjects were pretreated with inhaled albuterol 200 µg 20 min before sputum induction to minimize bronchoconstriction. FEV₁ was measured after each nebulization, and if FEV₁ fell > 20%, or if troublesome symptoms appeared, the nebulization was stopped. If the FEV₁ fell by > 10%, < 20%, the following concentration of saline was not given. Once expectorated sputum was placed on ice and sputum processing was performed at 4° C.

Sputum free of salivary contamination was selected immediately after expectoration. The sputum was dispersed using four times the selected sputum volume of 0.1% dithiothreitol, which was mixed by gentle aspiration in and out of a Pasteur pipette and rocked on a bench rocker for 15 min. The sample was further diluted with four volumes of Dulbecco's phosphate-buffered saline (PBS). The suspension was filtered through a 48-µm gauze and centrifuged at 790 × g for 10 min. The cell-free supernatant was removed and stored at 80° C until analysis. The cell pellet was resuspended in PBS, and a total cell count and squamous cell contamination were performed using a Neubauer hemocytometer with cell viability assessed by the trypan blue exclusion method. The cell suspension was readjusted to 0.5 to 0.75 × 10⁶ cells/ml, and 75 µl was suspended in cytocentrifuge cups and centrifuged at 450 rpm for 6 min in a Shandon II cytocentrifuge (Shandon, Runcorn, Cheshire, UK). The suspension was air-dried, stained with Romanovski stain, and a differential cell count was obtained by counting > 400 nonsquamous cells. Cell counting was performed by an experienced observer blind to the subjects clinical characteristics.

Mediator Measurements

There was insufficient sputum supernatant to measure histamine in seven asthmatics and one normal subject, ECP in eight asthmatics and two normal subjects, and prostanoids in two normal subjects.

The concentration of the prostanoids PGD₂, PGE₂, PGF₂, and TXB₂ in the sputum supernatant were determined by modified stable isotope dilution assays that used gas chromatography-negative ion chemical ionization-mass spectroscopy (GC-NICI-MS) (5, 9). LTC₄/ D₄/E₄ were measured by enzyme immunoassay employing a cysteinyl-leukotriene polyclonal antiserum (Cayman Chemical, Ann Arbor, MI) (5, 9). Histamine was measured using a sensitive radioenzymic assay based on the conversion of histamine to [³H]methyl histamine by the enzyme, histamine-N-methyl transferase, in the presence of [³H]S-adenosyl methionine as the methyl donor (10). The concentration of eosinophilic cationic protein (ECP) was measured using a commercial fluroimmunoassay (Unicap; Pharmacia, Milton Keynes, UK), which produces results very similar to the previously described radioimmunoassay (6). The intra-assay coefficient of variability of the cysteinyl-leukotriene and histamine assay was 5 to 10% and the interassay coefficient of variability for cysteinyl-leukotriene was 10 to 15% across the range of concentrations measured.

Analysis

Sputum differential eosinophil counts and eicosanoid concentrations (corrected for the sputum dilution and expressed as nanograms per milliliter sputum) were log-transformed and described as geometric mean (log SEM). Other differential cell counts, total cell counts, viability, and squamous cell contamination were described as median and interquartile range. Sputum supernatant PGD₂ concentration was normally distributed and was described as mean (SEM). Sputum eosinophil count and supernatant mediator concentrations were compared between groups by one-way analysis of variance (ANOVA), and differences were expressed as fold or absolute change with 95% CI. Other sputum indices were compared by the Kruskal-Wallis test. Mediator concentrations were compared between subjects with asthma treated with inhaled corticosteroids and those treated with -agonist alone by Student's unpaired t test. Significance was accepted at the level of 95% and the Student-Newman-Keuls procedure was used to adjust for multiple comparisons.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The groups were well matched for age and sex (Table 1). Atopy was more common in the subjects with asthma than in those with eosinophilic bronchitis. There was no difference in the cysteinyl-leukotriene, ECP, PGD₂, or histamine concentrations between the asthma subjects treated with inhaled corticosteroids and those treated with -agonist alone (Figure 1).

View larger version (11K): [in this window] [in a new window]   Figure 1.   Geometric mean (log SEM) concentrations of ECP, cysteinyl-leukotriene, histamine and mean (SEM) PGD2 concentration in each group of subjects (closed triangles). Subjects with asthma treated with -agonist alone (open triangles). *p < 0.05 versus normal subjects by ANOVA.

The sputum cell counts and cell-free sputum supernatant mediator concentrations were as shown (Table 1 and Figure 1). The subjects with asthma and eosinophilic bronchitis were well matched for sputum eosinophilia. Sputum cysteinyl-leukotriene and ECP were a mean (95% CI) 1.6-fold (1.1 to 2.5) (p < 0.05) and 6.4-fold (1.4 to 28) (p < 0.01) higher in eosinophilic bronchitis and 1.9-fold (1.3 to 2.9) and 7.7-fold (1.2 to 46) (p < 0.01) higher in asthma than in control subjects. Sputum PGD₂ and histamine concentrations were significantly higher in eosinophilic bronchitis than in asthma (mean absolute difference in PGD₂ concentration, 0.47 ng/ml [95% CI, 0.19 to 0.74], p < 0.01 and mean-fold difference in histamine concentration 6.7 [95% CI, 1.7 to 26], p < 0.01), and normal subjects (0.64 ng/ml [0.36 to 0.90], p < 0.01 and 11-fold [3.3 to 36], p < 0.01), respectively. The concentration of the other prostanoids were greater in the subjects with eosinophilic bronchitis and asthma than in normal subjects, but these differences did not reach statistical significance (Table 1).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have demonstrated for the first time that induced sputum cysteinyl-leukotrienes and eosinophilic cationic protein concentrations are increased in subjects with eosinophilic bronchitis when compared with normal subjects. The increase was comparable to that seen in subjects with asthma with a similar degree of eosinophilic airway inflammation. Sputum histamine and PGD₂ concentrations were raised only in eosinophilic bronchitis. Thus there is evidence of active eosinophilic airway inflammation in eosinophilic bronchitis with on-going release of vasoactive, bronchoconstrictor and airway damaging mediators. This strongly suggests that the difference in airway function in eosinophilic bronchitis and asthma are not due to differences in mediator production.

We have measured a wide spectrum of mediators with different functions representing the major effector and airway-damaging mediators in asthma. Cysteinyl-leukotrienes, produced by eosinophils and mast cells, are potent airway smooth-muscle contractile agonists and increase mucus production and vascular permeability and may directly increase eosinophilic airway inflammation (11). PGD₂ and histamine produced by mast cells have similar effects on airway smooth muscle, although they are less potent. ECP is directly toxic to epithelial cells in vitro (12). PGE₂ may increase cough sensitivity by a direct effect on the cough receptors (13). The lack of an increase in PGE₂ concentration in eosinophilic bronchitis suggests this is not the mechanism of increased cough sensitivity previously observed in this condition (14).

We chose to estimate airway mediator production using induced sputum supernatant since this is noninvasive, simple (5), safe (15), and responsive (16), and previous studies have shown that sputum ECP and eicosanoid concentrations are significantly higher than in bronchoalveolar lavage (9, 10) and can be measured repeatably (5, 6). We have considered potential problems that may lead to an exaggeration of the between-category differences observed in this study. Cell viability was not different between groups, indicating that the differences in mediator levels measured were unlikely to be due to cytotoxic release. Furthermore, leukotrienes and prostanoids are not stored preformed, unlike histamine and ECP, and so their presence in sputum provides evidence of ongoing synthesis and release. We have previously shown that the concentrations of eicosanoids in sputum treated with agents blocking ex vivo production and breakdown are not different from untreated sputum (5), so significant ex vivo production of eicosanoids is unlikely to have affected our findings. It is possible that the mediator concentrations in the sputum from subjects with asthma or eosinophilic bronchitis may have been increased by the effect of hypertonic saline on mast cells and other mediator-producing cells. We consider this unlikely as subjects were pretreated with albuterol before sputum induction, which would be expected to attenuate acute mediator release (17).

This study illustrates that, like asthma, there is active airway inflammation in eosinophilic bronchitis with release of bronchoconstrictor mediators. Why then might an apparently similar and equally active airway inflammation in eosinophilic bronchitis be associated with different abnormalities of airway function? It is possible that the site of the airway inflammation is different in the two diseases. We (3) and others (2) have noted that upper airway symptoms are common in patients with eosinophilic bronchitis and have speculated that inflammation is confined to the upper airway. However, eosinophilic bronchitis is not typically associated with a nasal wash eosinophilia or upper airway hyperresponsiveness (14). Furthermore, Gibson and colleagues (18) have shown a similar degree of bronchoalveolar lavage (BAL) eosinophilia and GM-CSF and IL-5 gene expression in patients with asthma and eosinophilic bronchitis. These observations suggest that the site of inflammation in eosinophilic bronchitis is mainly in the lower airway.

One possible reason for the difference in airway responsiveness is that in eosinophilic bronchitis the epithelium is intact. With asthma the degree of airway responsiveness is correlated with loss of epithelial structure (19) and the appearance of epithelial cells in bronchoalveolar lavage fluid (20). In asthma the partial loss of an epithelial barrier may allow greater amounts of bronchoconstricting mediators to reach the smooth muscle, or there may be a reduction in bronchoprotective substances such as PGE₂ (21). We found no differences in epithelial cell count or PGE₂ concentration between groups, which would not support this view. Whether sputum epithelial counts reflect epithelial integrity is unclear and further bronchial biopsy studies are required to fully address this question.

An alternative explanation for the difference in airway responsiveness is that airway responsiveness may be increased by the airway inflammation in eosinophilic bronchitis but stays within the normal range because baseline airway responsiveness is far to the right of the normal range. We have recently observed such a phenomenon in a patient with eosinophilic bronchitis who developed worsening symptoms and airway hyperresponsiveness during an exacerbation of eosinophilic airway inflammation (22). Alternatively airway hyperresponsiveness and variable airflow obstruction may develop later in the course of the disease in eosinophilic bronchitis, perhaps as a result of airway remodeling. However, we have observed a patient with severe eosinophilic bronchitis who developed an accelerated decline in FEV₁, presumably secondary to airway remodeling, but did not develop airway hyperresponsiveness or other features of asthma (23).

Interestingly, although we found the cysteinyl-leukotriene and ECP sputum concentrations to be increased in asthma and eosinophilic bronchitis, the sputum histamine and PGD₂ concentrations were only significantly elevated in subjects with eosinophilic bronchitis. This difference was not due to some of the asthmatic subjects being treated with inhaled corticosteroids as there was no difference between the asthmatic subjects treated with inhaled corticosteroids and those treated with -agonist alone. The elevation of histamine in combination with PGD₂ is highly suggestive of mast cell activation since basophils, which also produce histamine, do not produce PGD₂ (24). Thus, mast cell activation appears to be a feature of eosinophilic bronchitis. In support of this finding, bronchial brushings have revealed a greater number of mast cells with eosinophilic bronchitis than with asthma (20). In asthma mast cells infiltrate the bronchial smooth muscle and may contribute to airway hyperresponsiveness and intermittent bronchoconstriction through the local release of tryptase and autocoid mediators (25). It is not known whether they infiltrate the bronchial smooth muscle in eosinophilic bronchitis. One possibility is that they are more localized to the epithelium, so that mediators reach airway smooth muscle in lower concentrations than in asthma, but are present in high concentrations in epithelial lining fluid and the sputum.

In conclusion, there is active eosinophilic airway inflammation with release of vasoactive, bronchoconstrictor mediators in patients with eosinophilic bronchitis. Further detailed and systematic comparison of the immunopathologic features of the eosinophilic airway inflammation in asthma and eosinophilic bronchitis will be important in identifying particular features of the inflammatory process that are functionally important.