INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: difficult asthma; exhaled nitric oxide; children; endobronchial biopsy

Exhaled nitric oxide (FE_NO) has been proposed as a noninvasive marker of airway inflammation in asthma (1). Nitric oxide (NO) is produced from the conversion of L-arginine to L-citrulline, a reaction catalyzed by nitric oxide synthase (NOS). The inducible form of this enzyme, iNOS, has been identified in a range of cells, including epithelial and inflammatory cells, and is induced, among other stimuli, by proinflammatory mediators in asthma (2, 3). FE_NO has been shown to be elevated in patients with mild asthma not treated with inhaled steroids (4). Treatment of patients with mild to moderate asthma with inhaled steroids leads to a fall in FE_NO, and reducing the steroid dose results in a rise in FE_NO (5). Patients with difficult asthma have also been studied, and this has led to the identification of a subgroup of patients, with elevated FE_NO levels despite treatment with oral corticosteroids (6, 7).

A number of studies have examined the relationship between FE_NO and direct measures of airway inflammation, such as induced sputum, bronchoalveolar lavage (BAL), and endobronchial biopsy (8). The latter is considered the gold standard for assessment of airway inflammation in asthma (13). These studies have demonstrated an association between FE_NO and sputum eosinophils in asthma, which is most marked in corticosteroid-naive patients (8). This relationship is less clear in corticosteroid-treated patients. Lim and colleagues demonstrated an association between FE_NO and BAL eosinophils, but did not find any relationship between FE_NO and eosinophils in endobronchial biopsy specimens from both corticosteroid-naive and corticosteroid-treated adults with asthma (11, 12). No such studies have been performed in children.

Current guidelines for the treatment of asthma recommend the use of inhaled or oral corticosteroids to control airway inflammation (14, 15). Patients with difficult asthma have persistent symptoms despite maximal conventional therapy (16). Currently there is little evidence to guide the choice of further treatment in these patients. It has been recognized that persistent eosinophilic airway inflammation is present in only a proportion of patients with difficult asthma (17, 18). This has implications for further treatment. Demonstration of ongoing eosinophilic inflammation despite oral corticosteroids strengthens the argument for the use of alternative antiinflammatory treatments. Conversely, the absence of persistent eosinophilia should lead to consideration of treatments, such as long-acting bronchodilators or subcutaneous terbutaline, aimed at targets other than inflammation. Recently, we and others have begun to make this type of individual management decision based on the presence or absence of persistent eosinophilic inflammation (19, 20).

Further investigation of the relationship between FE_NO and airway eosinophilia in patients with difficult asthma is important, as the demonstration of an association between the two would allow clinicians to use FE_NO to assess the degree of ongoing eosinophilic inflammation, and to monitor the response to further antiinflammatory therapy, without the need for more invasive measures, such as fiberoptic bronchoscopy.

The aims of this study were to investigate the relationship between FE_NO and mucosal eosinophilic inflammation in a group of children with difficult asthma, following treatment with oral corticosteroids, and to establish whether FE_NO can be used as a surrogate for airway eosinophilia in this group. Detailed descriptions of the pathology of difficult asthma in children, and its relation to the clinical phenotypes, will be the subject of further reports.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

We studied children with difficult asthma (Table 1). Asthma was diagnosed according to American Thoracic Society (ATS) criteria (21). Difficult asthma was defined as "symptoms requiring rescue bronchodilator on  3 d/wk, despite  1600 µg/d of inhaled budesonide (or equivalent), and regular long-acting ₂-agonist (or a previous unsuccessful trial of long-acting ₂-agonist), and/or regular prednisolone" (16).

Patients underwent a full conventional assessment (22). If symptoms persisted, they received a corticosteroid trial, followed by bronchoscopy and endobronchial biopsy, to examine the airway pathology and guide individual management (19, 23).

Corticosteroid Trial

Patients recorded symptoms and bronchodilator use in a diary for 2 wk, followed by spirometry and measurement of FE_NO. Prednisolone, 40 mg/d, was given for 2 wk, and the diary was continued. Spirometry and FE_NO were repeated, serum prednisolone and cortisol levels were measured (6, 25), and bronchoscopy was performed the next day.

Control Subjects without Asthma

Endobronchial biopsy was performed in seven children without asthma undergoing bronchoscopy for other clinical indications.

Ethics Approval

The study was approved by the RBH Ethics Committee. Written informed consent was obtained from the subjects' parents.

Spirometry

Spirometry was performed using a portable spirometer (Compact Vitalograph), according to ATS guidelines (26).

FE_NO Measurement

FE_NO was measured according to ERS guidelines (27), as previously described (6). Subjects breathed room air. Ambient NO levels did not exceed 120 ppb (28). Subjects kept the flow during expiration within the required range of 200-280 ml/s by the use of auditory and visual guides.

Prednisolone and Cortisol Levels

Serum prednisolone and cortisol levels were measured using high-performance liquid chromatography. Adherence was considered satisfactory if there was detectable prednisolone, with cortisol < 100 nM (25, 29).

Clinical Response to Prednisolone

Diaries were examined for the last 7 d of the corticosteroid trial. Patients with satisfactory adherence were grouped as follows. Persistently symptomatic: symptoms requiring rescue bronchodilator on 7 of 7 d. Asymptomatic: symptoms on  1 of 7 d. Intermediate: patients fitting neither of the above criteria (i.e., symptoms on 2-6 of 7 d). Patients in whom satisfactory adherence was not demonstrated were classed as "adherence unknown."

Fiberoptic Bronchoscopy

Bronchoscopy with endobronchial biopsy was performed under general anesthesia using a 4.9-mm bronchoscope (Olympus, Keymed, UK), as previously described (19).

Processing of Biopsies

Processing of biopsies and detection of eosinophils using a mouse monoclonal antihuman major basic protein antibody (MBP, Monosan, TCS Biologicals Ltd, Botolph Claydon, Buckingham, UK) were performed as previously described (12).

Quantification

Coded sections were counted blind by a single observer (D.P.). Three sections were counted for each patient. The largest possible area of submucosa, in a zone 70 µm deep below the reticular basement membrane, was examined at ×1000 magnification (oil immersion). MBP immunostaining was assessed by point counting, using a 121-point grid (30). Each point overlying submucosal tissue was counted as positive or negative for immunostaining. Blood vessels, glands, muscle, and artifactual spaces caused by processing were excluded (31). At least 1210 points were counted (equivalent to 49,000 µm²). Immunostaining was expressed as the number of positive points, divided by the total number of points counted, given as a percentage (the "eosinophil score").

Statistical Analysis

Nonparametric tests were used. Comparison between groups was made using Kruskal-Wallis testing. Correlations were made using Spearman's rank correlation.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Thirty-one children with difficult asthma (Table 1) and seven control subjects without asthma (median age 10, range 8-13, yr) were studied. Evidence of satisfactory adherence to prednisolone was obtained in 26. Of these 26 children, 14 were persistently symptomatic, 7 were asymptomatic, and 5 were in the intermediate group. Technically reliable FE_NO readings were obtained from 29 of 31 children. Bronchoscopy was generally well tolerated. In the majority of patients anesthesia was induced with intravenous propofol, and maintained with an inhalational agent (isofluorane, sevofluorane, or halothane). Desaturation (a drop in oxygen saturation >10% from baseline) occurred in one patient with asthma, and was managed by removing the bronchoscope and increasing the inspired oxygen concentration. There were no other perioperative complications, in particular no instances of pneumothorax or significant bleeding. One patient with asthma developed a fever following the bronchoscopy and required readmission to hospital the day after bronchoscopy. BAL grew Streptococcus pneumoniae and the patient was treated as an inpatient for several days with intravenous antibiotics. Two other patients with asthma experienced an increase in symptoms for several days following bronchoscopy. Endobronchial biopsies suitable for analysis were obtained from 23 of 31 children with asthma, and 6 of 7 control subjects without asthma. For repeated measurements of eosinophil score on a single section, the coefficient of variation was 3.4%. Spearman's rank correlation coefficient (r²) for two sets of measurements on 15 sections was 0.96.

Group with Asthma versus Group without Asthma

The highest eosinophil score in the group without asthma was 5.4%. There was no significant difference in eosinophil score between the whole group of patients with asthma and the group without asthma (median [range] 3.8 [0-51.5] versus 0.8 [0.1- 5.4]%, p = 0.3). There was also no difference between the eosinophil score for any of the subgroups with asthma compared with the group without asthma.

Relationship between FE_NO and Eosinophil Score and Symptom-free Days

There was no correlation between symptom-free days and FE_NO (Figure 1), or between symptom-free days and eosinophil score (Figure 2). For all patients with asthma with both a satisfactory FE_NO reading and a biopsy suitable for analysis (n = 21), there was a significant correlation between FE_NO and eosinophil score (r = 0.67, p = 0.001) (Figure 3). When the patients in whom evidence of satisfactory adherence could not be obtained were excluded, the correlation remained significant (r = 0.54, p = 0.03, n = 17).

View larger version (7K): [in this window] [in a new window]   Figure 1.   FENo in children with difficult asthma, related to symptom-free days.

View larger version (8K): [in this window] [in a new window]   Figure 2.   Eosinophil score in children with difficult asthma, related to symptom-free days.

View larger version (7K): [in this window] [in a new window]   Figure 3.   Correlation between FENO and eosinophil score.

Use of FE_NO as a Marker for Airway Eosinophilia

For all patients with suitable FE_NO and biopsy measurements, and evidence of adherence to prednisolone (n = 17), FE_NO < 7 ppb was associated with an eosinophil score within the nonasthmatic range (Figure 4). FE_NO > 7 ppb was associated with an eosinophil score outside the nonasthmatic range in the persistently symptomatic group, but not in the intermediate and asymptomatic groups.

View larger version (10K): [in this window] [in a new window]   Figure 4.   Eosinophil score in subgroups with asthma, based on FENO level.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This is the first study to investigate the relationship between FE_NO and airway eosinophilia in endobronchial biopsies in children with difficult asthma. In pediatric practice the use of bronchoscopy and biopsy in asthma can currently be justified only in children with poorly controlled disease, in order to clarify the diagnosis and guide management (19, 32). It is therefore important that opportunities for obtaining tissue for research during clinically indicated procedures are maximized (35, 36). The patients in this study were all children with severe asthma, with frequent symptoms despite maximal conventional therapy, who underwent a clinical protocol for the assessment of their disease, which included bronchoscopy and endobronchial biopsy. The purpose of the protocol was to investigate the underlying basis for the poor control, in order to guide treatment (17, 19, 23). The aim of this study was to examine the relationship between FE_NO and mucosal eosinophilic inflammation, in order to see whether FE_NO could be used as a surrogate marker for airway eosinophilia, thus obviating the need for endobronchial biopsy.

In this study, we have demonstrated a significant correlation between FE_NO and mucosal eosinophilic inflammation in a group of children with difficult asthma, following treatment with oral prednisolone. This relationship was strongest in those patients who continued to have daily symptoms requiring rescue bronchodilator during the prednisolone course, despite evidence of adherence to treatment. For the whole group of patients, FE_NO < 7.0 ppb was associated with an eosinophil score within the nonasthmatic range, regardless of symptoms.

A potential weakness of the study is the difficulty in defining a normal range for airway eosinophilia in children. Endobronchial biopsy was performed in seven children undergoing fiberoptic bronchoscopy for other clinical indications. We cannot be certain that these patients had a normal lower airway. However, it is unethical to perform bronchoscopy and endobronchial biopsy on normal children solely for the purpose of research.

The technique of measurement of FE_NO differs in some respects from the method recommended in the ATS guidelines (37). Exhalation flow was maintained between 200 and 280 ml/s. This is similar to the range recommended by the ERS guidelines (27), which were the only guidelines published at the time that we begun this study. The current ATS guidelines recommend an exhalation flow rate of 50 ml/s. Patients breathed room air before measurements were taken, rather than NO-free air. A recent review of FE_NO measurement in childhood asthma concluded however that plateau values appear to be unaffected by ambient air (38). Using a method similar to our own, but with a flow rate of 5-6 L/min, Piacentini and colleagues demonstrated that ambient NO values up to 150 ppb do not affect plateau FE_NO values (28). In our study ambient NO levels were recorded and did not exceed 120 ppb.

A number of previous studies have demonstrated a significant correlation between FE_NO and airway eosinophilia in both induced sputum and bronchoalveolar lavage (8). The relationship is strongest in patients with corticosteroid-naive asthma. Treatment with corticosteroids appears to alter the relationship, suggesting that FE_NO and airway eosinophils may not be equally sensitive to the effects of corticosteroids. Given that the patients in this current study had all received 2 wk oral prednisolone, along with long-term treatment with high-dose inhaled corticosteroids, it is somewhat surprising that a significant correlation between FE_NO and airway eosinophils persisted. Looking at the response to prednisolone, it emerges that the relationship between the two parameters was strongest in those patients with persistent symptoms despite prednisolone. The relationship is less clear in those patients who showed some clinical response to prednisolone. We speculate that for patients who are clinically corticosteroid insensitive, corticosteroids may have little effect on FE_NO and eosinophils as well as on symptoms. This might explain why the relationship between FE_NO and airway eosinophils in this group is similar to that seen in corticosteroid-naive patients. This hypothesis requires further testing. In patients who respond to corticosteroids, the relationship between FE_NO and eosinophils is less clear, in keeping with previously published studies.

One of the aims of this study was to investigate whether FE_NO could be used as a surrogate for airway eosinophils, and therefore as an aid to the clinical management of children with difficult asthma. Our findings suggest that FE_NO < 7.0 ppb after prednisolone is associated with an eosinophil score within the nonasthmatic range, regardless of symptoms. For patients with persistent symptoms after prednisolone, FE_NO > 7.0 ppb appears to reflect an eosinophil score above the nonasthmatic range. This is not the case for patients in the intermediate and asymptomatic groups. Although the number of patients studied is very small, we suggest that measurement of FE_NO may have a role in the clinical setting, particularly in those patients with persistent symptoms in whom a decision regarding the use of unconventional therapies is being considered.

The upper limit of the normal range for FE_NO in our laboratory is 12.5 ppb (6). However, our findings suggest that NO > 7 ppb reflects significant eosinophilic airway inflammation in children with asthma who remain symptomatic despite oral prednisolone. One interpretation of these results might be that a proportion of healthy "control" subjects has significant eosinophilic airway inflammation. It is more likely that if oral prednisolone were given to normal children, which would of course be unethical, the upper limit of normal under these circumstances would be considerably lower than 12.5 ppb. It should also be noted that the relationship that we have demonstrated in children with difficult asthma after prednisolone therapy should not be extrapolated uncritically to other groups, for example, children treated with low-dose inhaled corticosteroids.

Difficult asthma in children is clearly a heterogeneous condition (16). The objective identification of different subgroups, based on the underlying pathology, may have important implications for the choice of further treatment. In those patients with daily symptoms and evidence of persistent eosinophilic inflammation despite prednisolone, it would seem logical to try another form of antiinflammatory therapy. Along with other investigators (17, 18), we have also identified a small number of patients with persistent symptoms, but evidence of little or no eosinophilic inflammation. Other mechanisms, such as neutrophilic airway inflammation, noninflammatory bronchial hyperreactivity, or a reduction in baseline airway caliber, may be responsible for ongoing symptoms in these patients (39, 40). The use of longacting bronchodilators or a continuous subcutaneous infusion of terbutaline may be more appropriate therapy for symptomatic patients with persistent bronchodilator reversibility in whom a noninflammatory basis for symptoms is suspected (20, 41). We have begun to evaluate the role of endobronchial biopsy in management decisions in individual children with difficult asthma. Further investigation of the underlying pathology should help to develop rational guidelines for the management of difficult asthma.

In summary, this study provides evidence of a relationship between FE_NO and mucosal eosinophilic inflammation in children with difficult asthma, following treatment with prednisolone. This relationship is most obvious in children with persistent symptoms despite prednisolone. Measurement of FE_NO in this group of patients may be helpful in identifying those with ongoing airway eosinophilia, therefore obviating the need for bronchoscopy and endobronchial biopsy.