INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In all types of asthma, irrespective of etiology, the airway wall is infiltrated with activated mast cells and eosinophils orchestrated by a subset of helper T-lymphocytes designated Th2 that secrete an array of cytokines encoded in the IL-4 gene cluster, IL-3, IL-4, IL-5, IL-9, and GM-CSF (1). It is these cytokines that are involved in the de novo maturation and/or recruitment of inflammatory cells as well as in their prolonged survival and priming for mediator secretion (4).

In all but the mildest forms of asthma, guidelines for management place inhaled corticosteroids as the mainstay of asthma treatment, with inhaled short-acting ₂-adrenoceptor agonists to be used for relief of symptoms only on an as-required basis (5, 6). Bronchial biopsy and lavage studies have clearly established the anti-inflammatory effect that inhaled corticosteroids exert on the airway inflammatory response. Their regular administration induces reduction in mast cell, eosinophil and T cell numbers together with their mediators in parallel with a reduction in asthma symptoms and restoration of pulmonary function, measured both as airway caliber and as bronchial responsiveness (7, 8). After studies demonstrating an increased benefit of adding the long-acting inhaled ₂-adrenoceptor agonist salmeterol to low dose inhaled corticosteroids (9, 10), this combination has now also been included in some asthma guidelines to bring deteriorating asthma under control as an alternative to further increasing the dose of inhaled corticosteroids (6, 11). The mechanism(s) underlying the added benefit of long acting inhaled ₂-adrenoceptor agonists is not known. It has been suggested that, although inhaled corticosteroids target the inflammatory response, this drug class acts functionally on the epithelium, microvasculature, and smooth muscle where the structure and function are altered as a consequence of the chronic inflammatory response (12).

However, there has been concern that regular high dose use of short-acting ₂-adrenoceptor agonists may result in deterioration of asthma control (13), although recent trials in mild asthma have been more reassuring (14). So far long-term clinical trials with both inhaled formoterol and salmeterol have failed to show any asthma deterioration, but there still remains some concern that regular treatment with these long-acting bronchodilators could mask an increase in bronchial inflammation. Conversely, it has not yet been established whether the increased degree of asthma control observed with the addition of inhaled long-acting ₂-adrenoceptor agonist to topical corticosteroids is associated with functional antagonism alone or through modulation of one or more of the mechanisms underlying the airway inflammatory response.

Because of uncertainties of the effects of inhaled long-acting ₂-adrenoceptor agonists on airway inflammation, the present study was designed to assess the effects of 8 wk of continuous treatment with inhaled formoterol on markers of airway inflammation as compared with treatment with the inhaled corticosteroid budesonide in patients with mild to moderate persistent asthma whose disease was being treated with as-required short-acting inhaled ₂-adrenoceptor agonists.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Design

A parallel group, double-blind, placebo-controlled study design was used to compare the effect of regular formoterol and budesonide on asthma symptoms, airway responsiveness, and indices of mucosal inflammation assessed in mucosal biopsies. The trial consisted of a 2 to 4 wk run-in, followed by a 9-wk treatment period, conducted outside the pollen season. Before and after 8 wk of treatment, fiberoptic bronchoscopy with bronchial biopsies and bronchoalveolar lavage (BAL) were performed. Asthmatic patients were recruited from three centers: Umeå, Boden, and Östersund, in the North of Sweden, but all bronchoscopies were performed at a single center in Umeå by the same investigator.

After the first bronchoscopy the patients received formoterol 24 µg, budesonide 400 µg, or placebo twice daily in a double-blind, double-dummy, randomized design. Formoterol (Foradil; Novartis, Basel, Switzerland), and placebo were given as inhalation capsules using a standard dry powder device (Aeroliser). Budesonide aerosol was delivered by metered-dose inhaler (MDI) via a large volume spacer. During the run-in and the treatment periods, patients used inhaled salbutamol by MDI (100 µg per puff) as required for symptom relief. The total number of salbutamol inhalations, symptoms, and peak expiratory flow (PEF) were recorded twice daily on diary cards. A methacholine bronchial provocation test was performed at the start of the run-in period, 2 to 4 wk before the baseline bronchoscopy, and at the end of the treatment period 12 to 16 h after the last formoterol dose, 1 wk after the second bronchoscopy.

Patients

Sixty-four subjects (28 female) 18 to 48 yr of age were enrolled into the study. They all had a typical asthma history and a methacholine PC₂₀ value between 0.1 and 6 mg/ml, positive skin prick tests to at least one common aeroallergen, and a baseline FEV₁ of more than 70% of that predicted for their age, sex, and height. The subjects either never smoked or had been ex-smokers for at least 5 yr. Their asthma was classified in the intermittent to mild persistent category as described in the GINA guidelines (10). Symptom control was maintained by the use of an inhaled short-acting ₂-adrenoceptor agonist as required during the preceding 2 mo. Subjects with other concomitant diseases, including respiratory tract infection during the preceding month, were excluded. The study was approved by the Ethics Committee of the University of Umeå, and written informed consent was given by each patient.

Protocol

Spirometry. A standard dry bellows spirometer was used throughout the study to measure FEV₁, and the highest of three consecutive measurements was recorded. FEV₁ was measured at visits before the morning dose of trial medication.

Diary card measurements (PEF, symptoms scores, as-required bronchodilator). PEF was taken as the best of three measurements with a Mini-Wright peak flow meter used every morning and evening before taking any medication. Asthma symptoms, PEF, use of inhaled salbutamol as required, and any adverse events were recorded on diary cards twice daily. Asthma symptoms were graded: 0 = none; 1 = symptoms not requiring rescue medication; 2 = symptoms requiring rescue medication; 3 = symptoms requiring an unscheduled physician contact. The use of salbutamol was recorded as the number of puffs taken since the previous diary entry.

Bronchial hyperresponsiveness. BHR was measured using the tidal breathing method described by Cockcroft (15). After inhaling isotonic saline, increasing concentrations of methacholine chloride were inhaled from a Wright's nebulizer, and the concentration that reduced the FEV₁ by 20% from postsaline baseline (PC₂₀) determined by linear interpolation.

Skin prick tests. Allergen testing was performed with 10 inhalant allergens (birch, mixed grass pollens, mugwort, cat fur, dog dander, horse, Dermatophagoides pteronyssinus, D. farinae, Alternaria, and Cladosporium Soluprick [ALK, Denmark]). The skin tests were considered positive if the diameter of a wheal exceeded by more than 3 mm that of the negative vehicle control.

Fiberoptic bronchoscopy. Fiberoptic bronchoscopy was undertaken under local anaesthesia in a dedicated endoscopy unit as previously described (16) and in accordance with the NHLBI guidelines (17). At each bronchoscopy four to six mucosal biopsies were taken from the subcarinae of one lung, and bronchoalveolar lavage (BAL) with two 60-ml aliquots of prewarmed physiological saline was performed in the opposite lung.

Processing and analysis of BAL fluid. The recovered BAL fluid was filtered and centrifuged at 400 × g for 15 min. The supernatant was separated and frozen at 20° C until analysis. After centrifugation the cells were resuspended and processed for total cell counts and cell differential counts as described (16). The mast-cell mediator tryptase was measured in the supernatant by radioimmunoassay (Pharmacia, Uppsala, Sweden) (detection limit, 0.5 µg/L; intra-assay and interassay variation < 3%). Eosinophil cationic protein (ECP) was analyzed by radioimmunoassay (Pharmacia) (detection limit, 2 µg/L; intra-assay and interassay variation < 8%). Albumin was assayed by a fluorometric method (Beckman Array Protein System; Beckman Instruments Inc., Brea, CA) (detection limit, 6 mg/L; interassay and intra-assay variability < 4%).

Bronchial mucosal biopsies. The mucosal biopsies measuring 1 to 2 mm were processed into glycol methacrylate resin (GMA) as described in detail previously (2, 18). After processing, the biopsies were transported to Southampton at 20° C for immunohistochemical staining using the streptavidin biotin peroxidase detection system, as described previously (2, 18). The monoclonal antibodies (Mabs) used are shown in Table 1, with nonbinding isotype-specific Mabs used as negative controls. The stained sections were coded and examined by the same investigator, who was blinded to the study design. Positive cells were counted in the submucosa for all antibodies, whereas only eosinophils and mast cells were counted in the epithelium.

Data Analyses

Wilcoxon's nonparametric Rank sum test was used for comparisons within treatments and the Mann-Whitney U test was used for comparisons between treatments for both bronchial biopsy, FEV₁ and BAL measurements. The change from baseline in density (number/mm²) of eosinophils (EG2⁺) and pan-T-cells (CD3⁺) in the airways submucosa were the primary end points in this study. PEF was subject to analysis of covariance. Bronchial responsiveness measured by interpolation as the PC₂₀ methacholine and were log-transformed and analyzed using Student's t test. Rescue medication use and asthma symptom scores were analyzed using van Elteren's nonparametric test. PEF, rescue medication, and asthma symptoms scores were analyzed for the whole period and for the 7 d preceding scheduled visits during the treatment period. Because in this group of patients with mild asthma there was a large variation in eosinophil count, to further assess treatment effects subanalyses were undertaken on biopsies containing  10 eosinophils/ mm² (Figure 1). Data are given either as medians with first and third interquartile ranges or as mean ± SD. A p value < 0.05 was considered significant.

View larger version (12K): [in this window] [in a new window]   Figure 1.   Baseline number of EG2+ eosinophils in the bronchial submucosa for formoterol, budesonide, and placebo treatment groups.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Clinical and Physiologic Measures

The 64 patients who were included in the study had a mean FEV₁ > 93% predicted and a PC₂₀ methacholine < 2 mg/ml. Four patients in the budesonide arm discontinued treatment prematurely: two because of poor compliance and two because of an intercurrent respiratory tract infection. The baseline values and effects of formoterol, budesonide, and placebo on symptom scores, rescue inhaled short-acting ₂-adrenoceptor agonists, PEF, FEV₁ and PC₂₀ values are presented in Table 2.

There were no significant differences between treatments in overall changes in symptom scores and changes in as-required salbutamol use. However, in the week preceding and the week after the final bronchoscopy, daytime asthma symptom scores were reduced with formoterol treatment when compared with placebo. This reduction in the week preceding the bronchoscopy was also significant compared with budesonide. A significant reduction in use of rescue medication in the formoterol group was observed in the week preceding the final bronchoscopy when compared with placebo.

Formoterol significantly improved morning PEF both overall and in the week after the final bronchoscopy compared with placebo (p < 0.05). This improvement in the week after the final bronchoscopy was also significant compared with budesonide (p < 0.05). No changes were observed in evening PEF. There was a decrease in FEV₁ during placebo (p < 0.05) but not over the active treatments (Table 2).

In all treatment groups there was a significant within-treatment increase in PC₂₀ methacholine (formoterol, p = 0.005; budesonide, p = 0.02; placebo, p = 0.04). There were no significant differences between treatment groups.

Bronchoalveolar Lavage and Bronchial Biopsy Analysis

The baseline and post-treatment values for tryptase, ECP, and albumin in BAL are displayed in Table 3. Comparison within and between groups showed no statistically significant differences in any of the measures. The information on the immunohistochemical staining for inflammatory cells in the submucosa and the epithelium before and after each of the three treatments for all the biopsies that were available for analysis are summarized in Tables 4 and 5.

When compared with the pretreatment biopsy, the number of mast cells identified by their tryptase content (AA1⁺) in the submucosa and in the epithelium decreased significantly after budesonide treatment, from a median value of 46 to 26/mm² (p < 0.01) and 1.1 to 0.3 mm (p < 0.05), respectively. The decrease in submucosal mast cell numbers after budesonide was also significantly different from that achieved with placebo (p < 0.01), but not formoterol (Table 4). Formoterol also reduced mast cell numbers when compared with pretreatment values (p < 0.05), but the decrease was not significantly different from placebo. In those patients with more severe airway inflammation (defined as an eosinophil count of  10/mm²) (Figure 1), both budesonide and formoterol resulted in a significant reduction in mast cell numbers when compared with pretreatment values (p < 0.05) and, in the case of budesonide, when compared with placebo (p < 0.05) (Figure 2).

View larger version (13K): [in this window] [in a new window]   View larger version (12K): [in this window] [in a new window]   Figure 2.   Submucosal mast cell (top panel ) and eosinophil (bottom panel ) counts in the submucosa of bronchial biopsies from asthmatic subjects with more severe airway inflammation (eosinophil count  10/mm2).

Eosinophil counts varied considerably in the submucosa of individual biopsies prior to any treatment (Figure 1 and Table 4). When all the available biopsies were analyzed neither budesonide nor formoterol had any significant effect on the submucosal eosinophil count within or between treatments (Table 4). However, these analyses were in part confounded by the low baseline eosinophil count in those subjects who received budesonide. Eosinophil count pretreatment was significantly lower (p < 0.05) when compared with either placebo or formoterol (Table 4). For the patients with more severe inflammation (an eosinophil count  10/mm² in the submucosa [Figure 1]), formoterol produced a highly significant reduction in the submucosal eosinophilia from a median baseline value of 32 to 6/mm² (p < 0.01) (Figure 2). The decrease in eosinophils after formoterol was also significant when the pretreatment to post-treatment change in cell counts were compared between formoterol and placebo (p < 0.01), but not between formoterol and budesonide. In this subgroup analysis, budesonide also resulted in a reduction in eosinophil numbers in four out of a total of five analyzable pairs of biopsies. This just failed to reach significance because of the small numbers of biopsies available for analysis (Figure 2). For the group as a whole, there was a significant relationship between the baseline levels of submucosal eosinophilia and the magnitude of the fall in eosinophils with regular formoterol (r_s = 0.47, p = 0.033), but not budesonide. Eosinophil counts in the epithelium were low in all of the treatment groups and, as a consequence, showed no overall change with any of the treatments when all the biopsies were analyzed. In the subgroup of more severely inflamed biopsies (number of eosinophils  10/mm² of submucosa), both formoterol and budesonide, but not placebo, showed a tendency to reduce the epithelial eosinophil content, but these changes were not significant (Table 5).

There were no significant differences either within or between treatments for CD3⁺, CD4⁺, and CD8⁺ cells (Table 4). When compared with baseline values budesonide, but not formoterol or placebo, reduced the expression of CD25 as a marker of T-cell activation (p < 0.01). A between-treatment comparison of CD25⁺ cells showed a significant reduction in favor of budesonide over formoterol (p < 0.05) and placebo (p < 0.01). In the subgroup with more severe inflammation, budesonide significantly (p < 0.05) reduced the number of CD4⁺ and CD8⁺ cells and the monocyte/macrophage (LCA⁺) population when compared with placebo (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In patients with mild atopic asthma, inhaled budesonide, when compared with placebo, reduced the overall number of airway mucosal mast cells, eosinophils, and activated (CD25⁺) T cells as shown by others. In addressing concern expressed over the regular use of inhaled formoterol the 8 wk of treatment with this drug alone did not result in any discernible increase in indices of airway inflammation believed to be relevant to the pathophysiology of asthma. On the contrary, there was evidence that regular use of formoterol resulted in a reduction in infiltrating mast cell and eosinophil numbers, particularly in a subgroup of patients whose airway inflammation, as judged by a high number of infiltrating eosinophils. Thus, rather than being proinflammatory, regular use of formoterol might be producing some of its long-term beneficial effect in asthma by reducing airway inflammation.

In the present study patients with mild persistent asthma, as defined by their sole use of inhaled short-acting ₂-adrenoceptor agonists, were deliberately selected (5, 6). On the basis of this patient selection, it is not surprising that the addition of either regular inhaled budesonide or formoterol resulted in little or no change in most of the clinical and physiologic indices followed during the 9-wk trial period. In the present study eosinophil counts in the submucosa covered a wide range, with almost one half of the subjects having eosinophil counts < 10/mm² (Figure 1). Although we conducted analyses of cell counts in the whole patient group, even if a treatment effect was present, the low number of eosinophils present in the airways of some of the patients would mitigate against its detection. It was for this reason, recognizing the possibility of regression towards the mean, that in addition to analyzing the complete data set, we also selected a cutoff of eosinophil count in the submucosa of 10 or more cells/mm² as a threshold against which to measure any treatment response. In future studies, where an index of airway inflammation is to be used as a primary end point for a therapeutic agent, we would recommend that the entry criteria for analyses of biopsies should include an eosinophil count of 10/mm² or greater.

Although there is still considerable debate as to whether regular short-acting inhaled ₂-adrenoceptor agonists lead to deterioration in clinical asthma, their abrupt cessation after a period of high dose regular use has been shown to cause rebound bronchial hyperresponsiveness (19). Regular short-acting inhaled ₂-agonists have also been shown to cause a differential reduction in bronchoconstriction induced by adenosine 5'-monophosphate (AMP) when compared with that produced by methacholine, which has been interpreted as selective ₂-adrenoceptor refractoriness operating at the level of mucosal mast cells (20). Regular salmeterol also results in a loss of protection against the bronchoconstrictor effects of methacholine (21) and the early response to allergen provocation (22). After abrupt cessation of treatment a more severe late-phase airway inflammatory response to allergen challenge is observed (23). In the present study regular use of inhaled formoterol did not result in a measurable loss of protection against the airway response to methacholine challenge as has been reported in some studies (21). Moreover, bronchial mucosal biopsies taken after 8 wk of regular treatment did not show any evidence of increased airway inflammation.

Instead, regular formoterol appeared to reduce the inflammatory cell numbers in the bronchial submucosa and epithelium of these mildly asthmatic patients, an effect that was especially apparent in patients with more intense eosinophil infiltration. On the basis of previous experience (24), an eosinophil count of more than 10 cells/mm² in the submucosa enables a reliable pharmacologic effect on these cell numbers to be evaluated. In addition, this cutoff level is based on data that separates eosinophil counts in the submucosa of asthmatic airways from the upper range of that found in atopic normal subjects (24). Importantly, formoterol did not reduce either the number of T cells or the proportion of activated (CD25⁺) T cells. This suggests that any effect on the inflammatory response is not due to general T-cell suppression, as has been suggested for corticosteroids (25). Whether and to what extent this is relevant in mild asthma, which is predominantly mast-cell- and eosinophil-driven, remains to be determined.

A possible anti-inflammatory effect of formoterol has been suggested by a number of in vitro and in vivo studies. Similar to salmeterol (26), inhaled formoterol inhibits allergen-induced late-phase bronchoconstrictor responses, although it has been suggested that this is due more to functional antagonism than to inhibition of the inflammation (27). While protecting against the late allergen response, inhalation of a single dose of salmeterol has no effect on the eosinophil content of induced sputum at 6 to 8 and 24 h after challenge (28), but with 7 d of continuous treatment, salmeterol reduces allergen-induced BAL eosinophilia (29). The regular administration of short-acting ₂-agonist terbutaline for a period of 3 mo has been reported to decrease the number of mast cells, T cells, and plasma cells in the bronchial epithelium but, interestingly, not eosinophils (7). In contrast, Jeffery and coworkers (30) have failed to find any effect of regular terbutaline inhalation on inflammatory cells when administered continuously for 4 wk. Two small studies of salmeterol given for 12 and 8 wk, respectively, have failed to show any effect on inflammatory cells, including eosinophils in BAL (31, 32), whereas in a single biopsy study, Li and coworkers (33) have reported an inhibitory effect of 6 wk of treatment with salmeterol on neutrophil but not on T-cell or eosinophil numbers in mild asthma.

The mechanism(s) of the effects of formoterol on eosinophil and mast cell numbers observed in our study are at present unclear. It is tempting to speculate that part of the response observed is due to stabilization of activated inflammatory cells such as mast cells, eosinophils, lymphocytes, monocytes, macrophages, and neutrophils, all of which express ₂-adrenoceptors. Both salmeterol and formoterol produce prolonged inhibition of preformed and newly generated mediator release from activated mast cells. Although it is recognized that human mast cells can also release cytokines, it is not known whether ₂-adrenoceptor agonists are able to similarly inhibit mast-cell cytokine production and release. A recent bronchial biopsy study has shown that regular theophylline, an inhibitor of cyclic AMP phosphodiesterase, is capable of reducing immunostaining for both IL-4 and IL-5 in submucosal mast cells of patients with asthma (34).

The recent FACET Study (35) in asthma has shown that the combination of formoterol with low- or high-dose budesonide reduces the incidence of severe asthma exacerbations and raises the possibility that this long-acting ₂-adrenoceptor agonist produces effects on asthmatic airways that extend beyond simple functional antagonism. FACET has shown that the beneficial effect of formoterol on asthma exacerbations is additive to that achieved by both low- and high-dose budesonide. Exacerbations of asthma, whether occurring spontaneously or triggered by allergen exposure or by virus respiratory tract infections are accompanied by an increased tissue eosinophil response. Our finding that regular formoterol is able to reduce the baseline tissue eosinophilia might provide a mechanism for the protective effect of formoterol in the long-term management of asthma, beyond its capacity to functionally antagonize bronchoconstrictor mediators. However, with our present understanding, we wish to emphasise that long-acting ₂-adrenoceptor agonists should still be used only in combination with a topical corticosteroid, as recommended in current guidelines, and should not be prescribed as a sole therapy.