INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Chronic obstructive pulmonary disease (COPD) is characterized by slowly progressive airflow limitation, and is often accompanied by inflammation of the airways (1). Smoking is the most important etiologic factor in COPD (2), and cessation of this habit is the best therapeutic measure to slow the deterioration of lung function (3). Oral corticosteroids are beneficial during exacerbations of COPD (4), but oral doses are given only for short periods because of side effects. Inhaled corticosteroids, however, are controversial in the treatment of COPD, and the role of these drugs is still a matter of intensive research.

Inhaled corticosteroids are effective in the treatment of asthma through reduction of airway-wall inflammation (5). This is reflected by rapid improvement of symptoms, lung function, and airway hyperresponsiveness (AHR) (6). A very sensitive marker of airway-wall inflammation is hyperresponsiveness to adenosine 5'-monophosphate (AMP), which provides an indirect stimulus causing bronchoconstriction in asthma via stimulation of mast cells and airway nerves. In nocturnal asthma, hyperresponsiveness to AMP increases more at night and improves more after treatment with inhaled corticosteroids than does hyperresponsiveness to methacholine (MCh) (7, 8). Thus, AMP is regarded as a better representative of airway-wall inflammation than MCh, which acts directly on airway smooth-muscle cells.

In COPD, inhaled corticosteroids have little or no beneficial effect in short-term studies (9). This may be due to a difference from asthma in number, type, and activation of inflammation cells in the airways. In COPD, the number of macrophages and lymphocytes is increased in the airway wall, and increased numbers of neutrophils are present in the airway lumen (12, 13). Mast cells may also play a role, since they have been found in increased numbers in mucous glands (14), and their products histamine and N-methylhistamine have been found in increased concentrations in the sputum (15) and urine, respectively, of bronchitic individuals (16). Alternatively, insensitive parameters may have been used to assess the effects of inhaled corticosteroids in COPD. Therefore, we investigated whether the indirect stimulus provided by AMP is more sensitive than the direct stimulus provided by MCh for assessing beneficial effects of budesonide in COPD. We also studied effects of budesonide on serum levels of interleukin-8 (IL-8) and histamine, which may play a role in the recruitment of airway neutrophils and activity of airway mast cells, respectively. Additionally, we studied the effects of budesonide on single doses of the anticholinergic drug ipratropium bromide and the antihistaminic drug terfenadine before the challenge tests. The administration of these drugs was also part of a protocol in which the mechanism of AMP-induced bronchoconstriction in COPD was studied. These two drugs were selected because they could provide insight into the two most important pathways by which AMP acts in asthma (17, 18).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

We studied 44 current smokers with COPD, according to American Thoracic Society (ATS) criteria (19). Subjects were recruited from our outpatient clinic and by advertisements in local newspapers. The study protocol was approved by the hospital ethical committee, and all subjects gave written informed consent prior to the study. Inclusion criteria at the beginning of the study were age between 45 and 75 yr, current smoking, FEV₁ and FEV₁/VC < %predicted  1.64 residual SD (20), an increase in FEV₁ < 10% predicted after inhalation of 1 mg of terbutaline via Turbuhaler (Pulmicort; Astra Pharmaceutica, Lund, Sweden), and hyperresponsiveness to both MCh (PC₂₀MCh < 8.0 mg/ ml) and AMP (PC₂₀AMP  80 mg/ml). All subjects were nonatopic, as defined by a negative history of atopy, a negative skin test for 18 common aeroallergens, negative specific serum IgE for 11 common aeroallergens, and a serum eosinophil count < 400 × 10⁹/ml. Subjects did not have acute upper-respiratory-tract infections, and received no oral or inhaled glucocorticoids, antibiotics, mucolytics, or theophylline in the month prior to the study.

Study Design

Patients were randomly treated with inhaled budesonide delivered via a Turbuhaler (1,600 µg/d) or placebo for 6 wk. Prior to and after this period they were challenged with MCh and AMP at the same time of day on three successive mornings, which were separated by 2 to 4 d (Figure 1). The challenge with AMP was performed 2 h after the start of the MCh challenge. Before beginning the study, we assessed the effects of MCh on subsequent AMP challenge in 16 patients with COPD. MCh did not significantly affect PC₂₀AMP. In a double-blind, double-dummy random fashion, subjects either ingested 180 mg of terfenadine (Triludan; Brocades Pharma, Leiderdorp, The Netherlands) 3 h before, or inhaled 120 µg of ipratropium bromide (Atrovent; Boehringer Ingelheim, Ingelheim, Germany) from an inhalet inhaler 45 min before, or ingested and inhaled placebo before the start of the MCh challenge. Subjects did not use bronchodilators in the 10 h prior to the measurements, and were asked not to smoke or drink tea or coffee on the morning of the measurement day.

View larger version (12K): [in this window] [in a new window]   Figure 1.   Study design. Before treatment with budesonide or administration of placebo, subjects were challenged with MCh and AMP on 3 d. On each study day they received pretreatment with terfenadine, ipratropium bromide, or placebo. These drugs were administered before the challenge test at the end of the treatment period, in the same order. Terf = terfenadine; IB = ipratropium bromide; MCh = challenge with methacholine bromide; AMP = challenge with adenosine 5'-monophosphate; Revers = reversibility in FEV1 with 1 mg terbutaline.

All subjects were allowed to use only terbutaline inhaled from a Turbuhaler 250 µg (Bricanyl; Astra Pharmaceutica) for symptom relief, but not within 10 h prior to MCh challenges or 4 h before peak expiratory flow (PEF) measurements. In case terbutaline did not give symptom relief, subjects were allowed to use ipratropium bromide 40 µg instead of terbutaline. All subjects kept daily diary cards throughout the study, on which they recorded symptom scores using a four-point severity scale (0 = no symptoms and 3 = severe symptoms), as well as the number of terbutaline or ipratropium bromide inhalations and the highest of three measured values of morning and evening PEF recorded with the Vitalograph PEF meter (Vitalograph Ltd, Buckingham, UK). The number of inhalations from the study Turbuhaler (budesonide or placebo) was also recorded in the second part of the study. Compliance was additionally checked at the end of the study, by counting the number of turns remaining in the study turbohalers before the red indicator became totally visible.

Pulmonary Function and Inhalation Challenge Tests

Spirometry was performed with a calibrated, water-sealed spirometer (Lobe BV, Groningen, The Netherlands) according to standardized guidelines (20). Baseline FEV₁ and inspiratory VC were measured until three reproducible recordings were obtained. Highest values were used for analysis. Reference values were those of the European Respiratory Society (20). FEV₁ was expressed as a percentage of the predicted FEV₁ (FEV₁ %predicted).

Inhalation challenge tests were performed according to a 2-min tidal breathing method adapted from Cockcroft and coworkers (21). Doubling concentration (DCs) of MCh (0.038 to 19.6 mg/ml) and AMP (0.04 to 320 mg/ml) were administered as aerosols generated from a starting volume of 3 ml in a DeVilbiss 646 nebulizer (DeVilbiss Co., Somerset, PA). Solution output was 0.13 ml/min. FEV₁ was measured at 30 s and 90 s after inhalation of each concentration of MCh and AMP. The test was terminated when a decrease in FEV₁ of 20% or more of the baseline value occurred. PC₂₀ values were calculated by linear interpolation between the last two data points on the logarithmic concentration-response curve. PC₂₀ values were determined by extrapolation if FEV₁ did not fall 20% or more from the baseline value. PC₂₀ values of 39.25 mg/ml and 640 mg/ml were given for MCh and AMP, respectively, if no value could be calculated by extrapolation.

Statistical Analysis

In order to show a difference in PC₂₀AMP between the budesonide and placebo groups, we calculated that each group needed at least 20 fully evaluable patients. This computation was based on an alpha level of significance of 5%, a power of 90%, a clinically significant difference of 1.0 in log₂ PC₂₀AMP, and an SD of 1.0 for that difference in log₂ PC₂₀AMP.

Data are presented as means and SDs unless stated otherwise. All analyses were performed with SPSS for Windows 6.0 (SPSS Inc., Chicago, IL). PC₂₀ values were analyzed after base-2 logarithmic transformation, a change of 1 unit in log₂ PC₂₀ representing 1 DC. Differences in FEV₁, PC₂₀MCh and PC₂₀AMP between terfenadine, ipratropium bromide, and placebo pretreatment days, and between budesonide and placebo treatment, were calculated through analysis of variance (ANOVA). Diary-card data were based on the mean of the data over the last 10 d of the run-in period and the double-blind treatment period, and were also analyzed with ANOVA. Relationships between variables were calculated with Pearson's correlation coefficient, r. Values of p < 0.05 were considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Forty-nine smoking subjects with mild to severe COPD entered the study. Five subjects dropped out before randomization because of erroneous inclusion, loss of contact, respiratory-tract infection, loss of motivation, and nausea after ingestion of the tablets on Visit 3. Forty-four subjects were randomized to treatment with 1,600 µg budesonide or placebo. The clinical characteristics of these patients are listed in Table 1. There were no differences between the budesonide and the placebo group, nor between the 44 subjects in these groups and the five subjects who dropped out. Of the 44 randomized subjects, four subjects in the placebo group dropped out, two because of exacerbation of their COPD, one because of a myocardial infarction, and one because of the development of angina pectoris. One subject in the budesonide group dropped out because of an exacerbation of his COPD. The high dose of budesonide was tolerated well. Only one subject complained of hoarseness, which disappeared after the study. Three subjects in the budesonide group and one in the placebo group had inhaled < 75% of the prescribed dose. Compliance could not be checked in two subjects in the budesonide group and one in the placebo group because of damage to or loss of the study Turbuhaler. Data for these seven subjects were included in the analyses. All subjects smoked the same number of cigarettes at the completion of the study as at the beginning of the study.

There were no significant differences between the budesonide and placebo groups in symptom scores, morning and evening PEF, or number of bronchodilators used per day (Table 2).

Table 3 shows the values of FEV₁, PC₂₀MCh, and PC₂₀AMP before and at the end of 6 wk of treatment with budesonide or placebo. It also contains the values after pretreatment with either ipratropium bromide or terfenadine. After placebo pretreatment, budesonide improved PC₂₀MCh and PC₂₀AMP by 0.51 and 0.40 doubling concentrations (DCs), respectively, which was not significantly different from placebo (0.07 and 0.62 DCs, respectively). Individual changes in PC₂₀MCh and PC₂₀AMP are shown in Figure 2. After ipratropium bromide pretreatment, budesonide improved PC₂₀MCh by 1.01 DC, which was significantly different from the 0.10 change in DC after placebo (Figure 3; p = 0.036). Budesonide improved PC₂₀AMP by 0.33 DC, which was not significantly different from the 0.33 DC change after placebo (p = 0.116). The change in PC₂₀MCh with budesonide treatment correlated positively with the number of cigarettes smoked per day (r = 0.53, p = 0.015; Figure 4). After terfenadine pretreatment, budesonide did not improve FEV₁ or increase PC₂₀MCh or PC₂₀AMP significantly as compared with placebo.

View larger version (32K): [in this window] [in a new window]   Figure 2.   Individual change in PC20AMP (upper panel ) and PC20MCh (lower panel ) before and during budesonide treatment or administration of placebo. Before the challenges, subjects received placebo pretreatment. Open circles represent mean values.

View larger version (15K): [in this window] [in a new window]   Figure 3.   Individual change in PC20MCh before and during budesonide treatment or administration of placebo. Before the challenges, subjects received ipratropium bromide pretreatment. Open circles represent mean values.

View larger version (9K): [in this window] [in a new window]   Figure 4.   Correlation between number of cigarettes smoked per day and the change in PC20MCh after budesonide treatment and ipratropium bromide pretreatment (see the study design). One unit in log2 PC20MCh represents one doubling concentration.

Table 4 shows the serum IL-8 and histamine concentrations before and at the end of 6 wk of treatment with budesonide. The budesonide group had significantly higher concentrations of IL-8 prior to treatment than the placebo group (p = 0.01). IL-8 decreased significantly during budesonide treatment (p < 0.001), but did not change significantly during placebo administration. Baseline IL-8 and decrease in IL-8 did not correlate with clinical characteristics or lung function. Differences between the budesonide and placebo groups in baseline serum histamine and change in serum histamine were not significant. Baseline histamine did not correlate with clinical characteristics or lung function.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We found that a 6-wk course of 1,600 µg budesonide daily did not improve symptoms, lung function, or AHR in smokers with COPD. The sensitivity to the indirect stimulus produced by AMP was not greater than that to the direct stimulus produced by MCh in the assessment of changes in AHR. Budesonide did decrease serum levels of IL-8, and improved the protective effect of ipratropium bromide on MCh-induced bronchoconstriction.

In contrast to observations in asthma (6), budesonide improved hyperresponsiveness neither to MCh nor to AMP in our study. Since inhaled corticosteroids reduce airway inflammation, as shown by a decrease in epithelial permeability and secretory activity in bronchoalveolar lavage fluid (BALF) in smokers with chronic bronchitis (10), we had expected an improvement in hyperresponsiveness to AMP in our smokers with COPD. The lack of improvement in hyperresponsiveness to AMP may have been due to a difference in the number and activation of inflammatory cells in the airway wall in COPD and asthma (22, 23). Eosinophils and mast cells predominate in the airway wall in asthma, in contrast to macrophages and T lymphocytes in COPD (22). These inflammatory cells may respond differently to corticosteroids, since eosinophils and mast cells have been shown to be more steroid sensitive than macrophages and T lymphocytes in vitro. Keatings and colleagues found a decrease in the percentage of eosinophils and their activation markers in induced sputum of asthmatic individuals after 2 wk of inhaled corticosteroids (24). However, they could not detect any changes in cell differentials or cell activation markers in induced sputum of subjects with COPD. Their data suggest that the eosinophil activation process in COPD is different from that in asthma, which may partly explain the relative resistance of COPD to treatment with corticosteroids.

It is also possible that our treatment period was too short to induce improvement in hyperresponsiveness to AMP. Corticosteroids may have a slower onset of action in COPD than in asthma (9, 25).

We studied smokers with COPD because current smokers are more hyperresponsive to AMP than ex-smokers with COPD (26). We therefore expected the changes produced by budesonide in PC₂₀AMP to be greatest in this patient group. Two studies have shown that smokers had a smaller improvement in lung function than ex-smokers after 2 yr of treatment with inhaled corticosteroids (25, 27). Continuing smoking and thereby promoting airway-wall inflammation may thus diminish positive effects of inhaled corticosteroids. This might also explain of lack of effect of budesonide on hyperresponsiveness to AMP.

We found a significant decrease in serum IL-8 during treatment with budesonide. IL-8 is a chemokine that predominantly attracts neutrophils to sites of inflammation (28), and has been found in increased concentrations in BALF and sputum of patients with COPD (29, 30). Although several cells produce IL-8 (31), it is likely that the decrease in serum IL-8 in our study was due to an effect of budesonide on epithelial cells, since 6 wk of treatment with beclomethasone decreased the epithelial-cell products lactoferrin and lysozyme in BALF of smokers with bronchitis (10). The decrease in serum IL-8 during budesonide treatment implies that chronic use of inhaled corticosteroids could be beneficial in subjects with COPD, since a decrease in IL-8 may reduce neutrophil influx into the lungs, thereby decreasing elastase burden. This does not have an immediate effect on lung function and hyperresponsiveness, but may slow the decline of lung function over time.

Budesonide improved the protective effect of ipratropium in MCh-induced bronchoconstriction, which was greatest in those subjects who smoked the most cigarettes per day. Statistically, there was no interaction between budesonide and ipratropium. We do not have a good explanation for these findings. A possible mechanism is that budesonide reduced airway-wall thickening and thereby facilitated the diffusion of ipratropium and MCh molecules to the cholinergic receptors on airway smooth muscle. Ipratropium would be rendered more effective than MCh by this decreased thickening, since it has a higher molar mass. The higher concentrations of ipratropium at the cholinergic receptor could then compete antagonistically with MCh for receptor binding. However, this explanation does not take into account the failure of budesonide to improve FEV₁, PC₂₀AMP, or PC₂₀MCh.

In vitro studies have shown that corticosteroids may increase the number of cholinergic receptors in airway tissue (32). Thus, the available binding sites for ipratropium bromide may be increased by budesonide treatment. More molecules of ipratropium could then bind to the receptor and prevent binding by MCh molecules via competitive antagonism. This would not result in an increase of FEV₁, because bronchodilation produced by the high-dose of ipratropium would already be maximal before budesonide treatment, and because of a ceiling effect, could not improve from the binding of more drug molecules. The greater effect of budesonide in those subjects who smoked more cigarettes may have been due to a decrease in number or sensitivity of cholinergic receptors caused by smoking. However, if budesonide did increase the number of cholinergic receptors, a decrease in PC₂₀MCh after placebo pretreatment would have been expected, and this was not observed. Nevertheless, the finding of an increased protective effect of ipratropium bromide may have clinical consequences. Our study shows that inhaled corticosteroids improve the action of ipratropium bromide, which is an important drug in the treatment of COPD.

In summary, a 6-wk course of budesonide decreased serum levels of IL-8, indicating a decrease in airway epithelial-cell activation, and improved the protective effect of ipratropium bromide in MCh-induced bronchoconstriction. Budesonide did not improve lung function, AHR, or symptoms in smokers with COPD. The indirect stimulus provided by AMP did not elicit greater sensitivity than the direct stimulus produced by MCh in the assessment of changes in hyperresponsiveness. Our findings suggest that long-term treatment with inhaled corticosteroids in COPD may be beneficial through a reduction of neutrophil load in the airways and improvement of the action of anticholinergic drugs. Data from long-term, multicenter trials will soon become available, and it is hopeful that they will help to clarify this issue.