INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The pattern and characteristics of airway inflammation in children with acute asthma are not well described (1), and may differ from the typical infiltrate of eosinophils and mast cells that occurs in stable asthma in response to cytokines such as interleukin-5 (IL-5), secreted by T helper type 2 lymphocytes (2). Viral infections frequently cause exacerbations of asthma and may result in a neutrophil infiltrate in the airway mucosa of asthmatics during an exacerbation (3). Some studies in adults have suggested that the neutrophil may be the predominant inflammatory cell in the lower airway during acute exacerbations of asthma (4, 5), and that this may influence the clinical presentation, predisposing to a sudden severe attack (5). Because eosinophils, but not neutrophils, respond promptly to corticosteroid therapy, identification of the pattern of cellular infiltration in acute exacerbations of childhood asthma may have implications for therapy. At present, however, the extent to which airway inflammation occurs during acute exacerbations of asthma in children is unclear.

Studying airway inflammation in acute asthma is difficult. In patients with severe airflow obstruction, techniques such as bronchoscopy with bronchoalveolar lavage and biopsy are precluded because they may worsen airway obstruction. By contrast, analysis of induced sputum may be a promising technique because sputum can be induced from the majority of children over 7 yr and yield reproducible cell counts (6). Twaddell and coworkers examined the safety of sputum induction using normal saline (and not hypertonic saline), in a small group of children with acute asthma and reported no complications (7). These findings suggested that sputum induction could be used to examine airway inflammation in children with acute asthma.

The aim of this study was to describe the characteristics of airway inflammation in children with acute severe asthma. We sought to measure sputum cell counts and fluid-phase mediators and cytokines, and to compare inflammatory markers during the acute exacerbation and at resolution. We hypothesized that airway inflammation would be present in acute asthma, and would be characterized by a cellular infiltration and activation of eosinophils and neutrophils, which would be reduced upon resolution of the attack.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Children (n = 42) age 8 yr and older, attending the John Hunter Hospital Paediatric Emergency Department with an acute exacerbation of asthma were invited to participate in the study. The children had a prior diagnosis of asthma and we included only those who were seen by the investigator within 1 h of presentation at the Emergency Department and before administration of corticosteroid therapy in the Emergency Department. Asthma was diagnosed based upon a history of episodic respiratory symptoms, a prior doctor's diagnosis of asthma, and the use of inhaled asthma therapy (₂-agonists, corticosteroids, or cromolyn). Acute exacerbations were defined as increasing symptoms requiring emergency treatment and presentation to the hospital, and were confirmed by reference to records of the attending doctor.

A total of 42 children (21 male) were entered into the study, of whom four were excluded because of refusal to participate (n = 1) or inability to obtain satisfactory sputum samples (n = 3) (Table 1). The ages ranged from 8 to 17 yr, with a mean of 12 yr. The majority of the exacerbations (26, 68%) were attributed to viral infections, two to exercise, one to an inhalant (hairspray), and one to changes in the weather. Fifteen children (39%) had received oral steroids for the exacerbation at least 4 h before presentation to the Emergency Department. The mean percentage predicted FEV₁ during the acute exacerbation was 58.9% (SEM 3.5%). Twenty-four children (63%) were admitted to the ward.

Written informed consent was obtained from parents and children. The study was approved by the Hunter Area Research Ethics Committee and the University of Newcastle Ethics Committee.

Design

Upon arrival in the Emergency Department children underwent triage assessment, medical assessment, and initial therapy. The investigator was contacted, informed consent obtained, and a clinical asthma questionnaire was completed by interviewing the parent (Visit 1). The children were treated with nebulized albuterol, and sputum was induced using normal saline and lung function monitored. Upon completion of sputum induction, children were managed by staff in the Emergency Department with nebulized ₂-agonist and oral prednisolone. Those who produced a satisfactory sputum sample were contacted at least 2 wk later when they had recovered from their exacerbation and were invited to return to the hospital for a repeat sputum induction (Visit 2). The children were assessed for symptoms such as wheezing, cough, breathlessness, and increased use of bronchodilator. If they were symptomatic, the appointment was rescheduled until the symptoms had resolved and the children had returned to their baseline, preexacerbation state. At the follow-up visit spirometry was performed and bronchodilator response assessed, followed by sputum induction.

Sputum Induction: Acute Exacerbation

Sputum was induced during the acute exacerbation in the Emergency Department as described (7). Each child was pretreated with nebulized albuterol 5 mg. Lung function (FEV₁) was measured (Pneumocheck spirometer; Welch Allyn Inc., New York, NY); or if the FEV₁ could not be recorded because of inability to sustain expiration, then peak expiratory flow was recorded using a Wright Peak Flow Meter (Clement Clarke International, Essex, UK). Values were expressed as the percentage of the predicted value based upon age, gender, and height (8).

Sputum was induced using normal saline (0.9%) nebulized by an Omron NE-U06 ultrasonic nebulizer (Omron Corp., Tokyo, Japan; output 3 ml/min) with 9-cm corrugated tubing, and a plastic mouth piece and nose clips to ensure mouth breathing, Nebulized normal saline was given for intervals of 30 s, 1, 2, and 4 min, with lung function measured 1 min after each of these doses of normal saline. The child was encouraged to cough and expectorate sputum at each interval, and sputum induction was ceased once an adequate sample of sputum was obtained, once the maximum duration of nebulization was reached, or when lung function dropped to 40% predicted. After one of the three endpoints was reached, the child was managed by the staff in the Emergency Department.

The sputum sample was considered adequate when there was at least 0.5 ml of sputum containing three or more opaque, mucocellular clumps at least 1.5 × 3.0 mm in size (6). Supplemental ₂-agonist was administered if lung function dropped by more than 10% or if lung function dropped to 40% of predicted during the induction. If a child presented with lung function below 40% predicted, an attempt was made to obtain sputum spontaneously after inhalation of nebulized salbutamol. If sputum could not be obtained spontaneously, the child was excluded from the study.

Sputum Induction: Follow-up

Baseline lung function was recorded as the best of two forced expiratory maneuvers (Pulmonary Function System 1070, Series 2; Medgraphics, St. Paul, MN). Each child then inhaled 5 mg of nebulized albuterol with spirometry performed 10 min later. Sputum was induced using normal saline as described previously. If satisfactory sputum was not obtained, then 4.5% saline was inhaled using a DeVilbiss ultrasonic nebulizer (DeVilbiss Health Care Inc., Somerset) with 23-cm corrugated tubing and a Hans Rudolph 2700 two-way nonrebreathing valve box (Hans Rudolph Inc., Kansas City, MO) with rubber mouth piece and nose clip to ensure mouth breathing. Sputum cellularity is not altered by saline concentration or use of different nebulizers (9, 10).

Sputum Processing

Sputum was processed within 1 h of collection as described (7, 11). Briefly, opaque mucocellular clumps (MUCC) were selected from the saliva and 2 smears prepared for mast cell counts using the squash technique. From the remaining sputum, 300 µl of MUCC were drawn into a positive displacement pipette, and placed in 2.7 ml of dithiothreitol (DTT), 1:10 (Calbiochem, La Jolla, CA). This was placed in a shaking water bath for 30 min at 37° C for cell dispersion. The cell suspension was then filtered through 50-µm nylon gauze and a total cell count was performed. The remainder of the filtered cell suspension was made up to 5 ml using phosphate-buffered solution (PBS) and cells pelleted. The supernatant was aspirated and frozen (70° C) for later use. The remaining cell pellet was diluted with PBS to a concentration of 1 × 10⁶/ml and cytocentrifuge preparations were made (Cytospin 3; Shandon Scientific, Sewickey, PA) and air-dried. The quality of induced sputum samples was assessed based upon the presence of an adequate number of cells for enumeration, the presence of pulmonary macrophages on the slide, and the proportion of squamous epithelial cells. This gave a quality score ranging from 0 (poor quality) to 6 (good quality sample) (11).

Cytochemistry

A differential cell count was obtained by counting 400 nonsquamous cells on slides fixed with methanol and stained with May-Grünwald-Giemsa. Eosinophils were enumerated as the percentage of 400 cells on slides fixed with methanol and stained with chromotrope 2R. Metachromatic cells were counted as the percentage of 1,500 cells on smears fixed in Carnoy's solution and stained with acidic toluidine blue.

Immunochemistry

Cytospin slides were prepared for immunocytochemistry using the alkaline phosphatase-antialkaline phosphatase (APAAP) technique. Neutrophils were identified using a mouse monoclonal antibody directed at elastase which is present in the primary intracellular granules of neutrophils (DAKO Corp, Carpinteria, CA), on slides fixed in formalin/ethanol as described (7, 12). Freshly isolated peripheral blood neutrophils were used as positive controls. For immunostaining of eosinophils, a monoclonal mouse antibody (EG2) directed at the secreted form of human eosinophil cationic protein (ECP) (Kabi Pharmacia Diagnostic, Uppsala, Sweden) was applied (7, 13). Immunomagnetically isolated peripheral blood eosinophils were used as positive controls. Isotype antibody and substrate controls were included in all staining runs. ECP in sputum supernatant was measured in duplicate using radioimmunoassay (Pharmacia AB, Uppsala, Sweden). Myeloperoxidase (MPO), IL-8, and IL-5 were measured by ELISA (R&D Systems, Minneapollis, MN) with lower limit of detection being 1.6 ng/ml, 10 pg/ml, and 3 pg/ml, respectively.

Statistics

Cell counts were expressed as the median and interquartile range (IQR; Q1, Q3). Wilcoxon's rank sum test was used to compare paired data. Spearman's rank correlation was used to compare correlation for data that was not distributed normally. p Values < 0.05 were considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sputum Induction

Sputum was successfully obtained from 38 children during the exacerbation: 16 produced spontaneous samples and 22 required sputum induction. These children inhaled saline for a mean (SD) of 3.0 (3.6) min. At follow-up, 28 children attended and produced sputum. One child was able to produce sputum spontaneously, 14 children had sputum induction using normal saline, and 13 children had sputum induction using hypertonic saline. Sputum induction was well tolerated with a median decrease in FEV₁ of 5.8% at Visit 1 and 1.5% at Visit 2 (Table 1).

Sputum Markers in Acute Asthma

Sputum quality was good with a median score of 5 (IQR 5, 6) at Visit 1 and 5 (IQR 4, 6) at Visit 2. The median total cell count (TCC) at Visit 1 was 8.4 × 10⁶/ml (IQR 2.3, 29.0) (Table 2, Figure 1). There was a wide range of values, with the lowest total cell count being 0.5 × 10⁶/ml and the highest being 190.3 × 10⁶/ml. During the acute exacerbation there was a prominent neutrophilia as well as increased sputum eosinophils and mast cells (Table 2). At Visit 1 the median count of neutrophil-elastase positive cells was 3.4 × 10⁶/ml (n = 33), and there were 1.8 × 10⁶/ml (IQR 0.7, 11) EG2-positive cells (Figure 2). The percentage of activated eosinophils (EG2⁺) was negatively correlated with the severity of the exacerbation (r = 0.5, p = 0.02, Figure 3). Cell counts were similar in children admitted versus those discharged, suggesting that the intensity of the inflammatory cell infiltrate was not a major determinant of this clinical outcome.

View larger version (22K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   Figure 1.   Sputum total cell count (× 106/ml, panel A) and mast cell count (× 106/ml, panel B) during acute asthma (Visit 1) and at resolution of the attack (Visit 2). Individual data points with medians (horizontal bars); p < 0.05. In (B), the open squares represent data from seven subjects.

View larger version (16K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   Figure 2.   Activated sputum eosinophils (EG2+ cells, × 106/ml panel A), and neutrophils (elastase-positive cells, panel B) were high during the acute asthma exacerbation (Visit 1) and decreased significantly (p < 0.01) with resolution (Visit 2).

View larger version (9K): [in this window] [in a new window]   Figure 3.   The severity of airflow obstruction (percent predicted FEV1 or peak expiratory flow) during the asthma exacerbation was correlated with activated eosinophils (EG2+ cells) in induced sputum. r = 0.5, p = 0.02.

Sputum supernatant ECP was elevated during the exacerbation at 1,078 ng/ml, with values as high as 363,058 ng/ml recorded (Table 2, Figure 4). At resolution there was a significant decline in sputum ECP to 272 ng/ml (p < 0.01). MPO levels during the acute exacerbation were 222.0 (IQR 1.0, 526.6) ng/ml, and median IL-8 levels were 15.4 (IQR 3.04, 31.09) ng/ml (Figure 4). IL-5 was detected in 11 sputum samples during the acute exacerbation at a level of 110.5 (IQR 73.7, 279.1) pg/ml.

View larger version (17K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   Figure 4.   Sputum supernatant ECP (ng/ml, panel A) and IL-8 ( panel B, ng/ml) were high during the acute asthma exacerbation (Visit 1) and decreased significantly (p < 0.01) with resolution (Visit 2).

On recovery, there was a significant fall in all sputum inflammatory cells (p < 0.01, Table 2). At Visit 2 the median total cell count had fallen to 1.3 × 10⁶/ml (range 0.2 × 10.6/ml to 7.6 × 10⁶/ml; p < 0.001). Elastase-positive cells had fallen significantly to 0.4 × 10⁶/ml (p = 0.001, Figure 2) and EG2⁺ cells also fell significantly to 0.4 × 10⁶/ml (IQR 0.2, 0.6) at Visit 2 (p < 0.001, Figure 2). Sputum supernatant MPO and ECP both decreased significantly at resolution to 1.0 (IQR 1.0, 245.6) ng/ml (p = 0.02) and 272 (IQR 127, 1,083) ng/ml (p < 0.01, Figure 4), respectively. At follow-up, IL-8 levels were reduced to 2.76 (IQR 1.09, 5.4) ng/ml (p = 0.004, Figure 4). IL-5 could be detected in sputum supernatants from six children at Visit 2, at levels of 159.9 (IQR 119.2, 367.9) pg/ml (p > 0.05).

Spontaneous versus Induced Sputum

There was a nonsignificant trend for the total cell count to be higher in those who produced sputum spontaneously at Visit 1 than in those in whom sputum was induced by 0.9% saline [induced 4.9 × 10⁶/ml (Q1 2.1, Q3 22.8), spontaneous 18.0 × 10⁶/ ml (Q1 4.2, Q3 56.3), p = 0.1]. The absolute neutrophil count was significantly higher in spontaneous sputum (induced 1.8 × 10⁶/ml versus spontaneous 11.7 × 10⁶/ml, p = 0.005), as was the absolute lymphocyte count (0 versus 0.03 × 10⁶/ml, p = 0.05). Cell counts for other cell types were similar in spontaneous and induced sputum, as were fluid-phase measurements (IL-8, IL-5, ECP, MPO). The change in cell counts from acute exacerbation to resolution was also examined in the 15 children who had sputum induction performed at both time-points. With resolution of the exacerbation, there were significant decreases in total cell count (p = 0.005), eosinophil count (p = 0.003), neutrophil count (p = 0.05), and mast cell count (p = 0.02).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study has established that children who present to the Emergency Department with acute asthma have airway inflammation that is heterogeneous in nature and comprises eosinophils, neutrophils, and mast cells. There was evidence of both eosinophil and neutrophil activation, and secretion of cytokines known to be chemoattractant for neutrophils (IL-8) and eosinophils (IL-5). The airway inflammation had improved 2 wk later when the child's symptoms had also improved; however, sputum eosinophilia persisted. Although eosinophilic inflammation has been widely accepted as the hallmark of bronchial asthma, there was also evidence of neutrophil participation during the asthma exacerbation. This is an important observation which was initially reported in adults (4) and has implications for the cause of the exacerbations, their mechanisms, and treatment.

Sputum induction has been used as a noninvasive method of studying airway inflammation in asthma and in chronic obstructive airway disease in adults and children (6, 7). The technique is feasible and safe, and gives reproducible results. In particular, sputum can be safely obtained during an asthma exacerbation and in uncontrolled asthma (4, 7, 14) where bronchoalveolar lavage would be considered unsafe. Different techniques of sputum processing have been reviewed and found to be reproducible and valid (15). In this study, sputum portions were selected from the saliva as described by Pin (6). This reduces buccal squamous cell contamination and yields better quality slides and a significantly higher concentration of ECP (16). The quality of the slides in this study at both visits was good. In this study, the methods used to obtain sputum samples included spontaneous expectoration and sputum induction with normal or hypertonic saline. Prior work has established that the tonicity of saline used for induction has little effect on cellular differentials (9, 10). Although there was a predominance of spontaneous samples at visit 1, when only induced sputum samples were analyzed the same fall in cell counts was seen.

In stable asthma, sputum analysis demonstrates increased eosinophils, but neutrophils and total cell counts are not elevated (6, 11). There are few studies that have looked at inflammatory changes occurring during spontaneous exacerbations of asthma in vivo. The total cell counts in adults with spontaneous exacerbations of asthma were higher compared with stable asthmatics (4, 14, 17). This study showed that during an acute exacerbation of asthma, the inflammatory process was heterogeneous with infiltration of all cell types particularly neutrophils and eosinophils. The eosinophils were activated, and released ECP. In addition to the eosinophilic response, a prominent neutrophilic inflammation was observed during the acute exacerbation together with high levels of the neutrophil chemoattractant IL-8, and evidence of neutrophil activation. The dominance of the neutrophil could indicate that the attack was precipitated by a viral infection. Although we did not attempt to isolate virus, 68% of the exacerbations were attributed to an upper respiratory tract infection, and this is consistent with studies that demonstrate that most exacerbations of childhood asthma are associated with respiratory viruses (18). A postulated mechanism is the chemoattraction of neutrophils by IL-8 produced by macrophages or epithelial cells as a consequence of viral infection (3, 19). Coyle and coworkers (20) have recently described in mice how a viral infection occurring in the context of a local lung Th2 immune response can switch viral-specific CD8⁺ T cells to produce IL-5 and recruit eosinophils to the airways. This provides a potential mechanism for the sputum eosinophilia and IL-5 secretion in the exacerbations, and indicates how viral infection and allergy could interact to produce the heterogeneous inflammatory response seen in this study.

Sputum EG2⁺ eosinophils were negatively correlated with FEV₁, suggesting that eosinophils may be important determinants of the severity of airway obstruction in acute asthma. With resolution of the attack, there were impressive reductions in total cell counts, neutrophils, eosinophils, and sputum ECP and MPO levels. This provides further evidence that airway inflammation is important in acute attacks of asthma. Corticosteroids remain the most effective treatment for both chronic asthma and for acute exacerbations of asthma. However, their effect may be modest during an exacerbation (21, 22). Studies have shown that oral corticosteroids reduce sputum eosinophils and ECP in adult asthmatics with no effect on total cell counts, epithelial cells, neutrophils, or lymphocytes (14, 23). In this study, oral corticosteroids did not appear to switch off the inflammation during the exacerbation because even children taking oral steroid had elevated cell counts at presentation. It is possible that the duration of oral corticosteroids therapy was insufficient or inadequate to suppress airway inflammation. Alternatively, the main effect of steroids in acute asthma may be to reduce vascular leakage, and this was not measured.

The effects of corticosteroids in acute severe asthma do not seem as clear-cut as in mild exacerbations or stable asthma. A review of 15 studies of corticosteroids in acute asthma demonstrated a positive effect on lung function in six studies, whereas no improvement was seen in nine studies (22). Acute severe asthma differs from stable asthma in two ways. First, the trigger for the exacerbation is often a viral infection (18), and second, we have shown that neutrophils are a prominent cell contributing to airway inflammation in this setting. Because both these factors respond poorly to corticosteroid therapy (23, 24), this may explain the variable response to corticosteroids in acute asthma. These observations suggest a potential role for therapy directed at neutrophils or neutrophil chemokines such as IL-8 during severe exacerbations of childhood asthma.

In conclusion, this study has established that there is increased inflammation of the airways in acute exacerbations of asthma in children. The characteristics differ from stable asthma. In acute asthma both eosinophils and neutrophils are increased in number and are activated, probably as a consequence of the cytokines IL-5 and IL-8. This mixed inflammatory response is important in the pathogenesis of acute asthma and may have implications for response to therapy.