Slowing the Deterioration of Asthma and Chronic Obstructive Pulmonary Disease Observed during Bronchodilator Therapy by Adding Inhaled Corticosteroids: A 4-Year Prospective Study

Methods

Patients

Patient selection has been previously described [6]. In short, 29 family physicians in the catchment area of the University of Nijmegen, Nijmegen, the Netherlands, selected all patients who were 30 years or older and had symptoms of asthma or COPD. Only patients who showed mild-to-moderate airway obstruction (FEV₁ > 50% of the predicted value [7]) or bronchial hyper-responsiveness to histamine (PC ₂₀ 8 mg/mL) were included in the study. Patients dependent on inhaled corticosteroids who had chronic heart failure, malignant disorders, or other severe life-threatening diseases were excluded from the study. Of these patients, 160 (59 with asthma and 101 with COPD) completed the bronchodilator trial. During the 2 years of bronchodilator treatment, a rapid decline in FEV₁ ( 80 mL/y) and a relatively high exacerbation rate ( 1/y) were observed in a subgroup of 56 patients (35%). Because of their unfavorable disease course, these patients were selected for additional treatment with inhaled beclomethasone for 2 years. The criteria for diagnosis of asthma or COPD were based on the standards of the American Thoracic Society [8]. Asthma was defined [6, 8] by a combination of factors: bronchial hyper-responsiveness to histamine (PC ₂₀ 8 mg/mL); reversible obstruction (an improvement in FEV₁ of more than 15% of the prebronchodilator value 60 minutes after the administration of both salbutamol, 400 g, and ipratropium bromide, 80 g); dyspnea; and allergy (defined as at least one positive result on seven radioallergosorbent tests that assessed sensitivity to pollen from weeds, grasses, and trees; cats and dogs; house dust mite; and Aspergillus fumigatus) or wheezing. Chronic obstructive pulmonary disease was defined [6, 8] by the combination of chronic cough or chronic sputum production for at least 3 months during at least 2 consecutive years; and continuous bronchus obstruction (FEV₁ 85% of the predicted value). The separate features of asthma and COPD overlap (for instance, some asthmatic patients had chronic cough, and some COPD patients had a PC ₂₀ 8 mg/mL), but the definitions based on feature combinations ensured that no patients with asthma also had COPD and vice versa [6]. The study was approved by the Medical Ethics Committee of the University of Nijmegen. All patients gave informed consent.

Study Design and Treatment

At the start of the 4-year intervention study, the patients were randomly assigned to one of two parallel treatment regimens: continuous bronchodilator therapy (four times daily) or treatment on demand (dry powder inhalations during symptomatic periods) [6]. The patients used salbutamol, 400 g, during 1 year and ipratropium bromide, 40 g, during the other year; both were administered as dry powder inhalations. The sequence of the drugs was determined by random allocation. During years 3 and 4, the 56 patients received 400 g of beclomethasone, two times daily, in combination with 400 g of salbutamol or 40 g of ipratropium bromide, four times daily (all dry powder inhalations). The bronchodilator inhaled during year 2 was also used in years 3 and 4.

During the first 2 years of the study, 27 of the 56 patients received bronchodilator therapy on demand (of the 27, 15 had asthma and 12 had COPD). For patients treated on demand, the mean (SE) daily number of dry powder inhalations of salbutamol or ipratropium bromide was 1.2 0.3 in those with asthma and 0.8 0.2 in those with COPD. During years 3 and 4, 28 patients received salbutamol (15 with asthma and 13 with COPD) and 28 received ipratropium bromide (13 with asthma and 15 with COPD).

Once every 3 months, inhalation technique and compliance with the prescribed medication were checked. Patients were instructed to rinse their mouths after the dry powder inhalations. During the second year of beclomethasone therapy, a single-blind prospective study was done to assess patient compliance with beclomethasone and the additional bronchodilator. Compliance was measured by counting capsules at the end of a 4-month period. Patients were unaware that their medication was counted after this period.

Lung Function, Nonspecific Bronchial Responsiveness, and Reversibility

All measurements were carried out by two qualified laboratory assistants during exacerbation-free periods. No bronchodilator was inhaled for at least 8 hours before the pulmonary function tests. At the start and after 24 and 48 months of the study, the inspiratory vital capacity (IVC), residual volume (RV), functional residual capacity (FRC), and total lung capacity (TLC) were assessed using the wet Gould spirometer (Sensormedics, Bilthoven, the Netherlands) according to the standards of the European Coal and Steel Community [7]. The FEV₁, bronchial responsiveness to histamine, and the reversibility of airway obstruction were assessed at 6-month intervals using the Microspiro HI-298 (Chest Corporation, Tokyo, Japan) [9]. Moreover, FEV₁ and reversibility were also assessed after 1 and 13 months of study [6]. The best of three forced expiratory maneuvers, with the highest sum of the forced vital capacity (FVC) and FEV₁, was used for data analysis. The bronchial responsiveness to histamine was measured according to the method described by Cockcroft and colleagues [10]. Results were expressed as the concentration of histamine that provoked a 20% decrease in FEV₁ (PC ₂₀). After the FEV₁ had returned to the baseline value, the bronchodilating response (reversibility) was assessed 60 minutes after the administration of both 80 g of ipratropium bromide and 400 g of salbutamol (metered dose aerosol) [6]. The bronchodilating response was expressed as the increase in FEV₁ relative to the predicted value of the FEV₁.

Peak Expiratory Flow Assessments

Once a week (on the same day and at the same time), peak expiratory flow rate (PEFR) was measured with the Assess peak flow meter (HealthScan Products, Cedar Grove, New Jersey) [11] in the morning and in the evening. The highest value of three measurements was included in the analysis. The diurnal PEFR index (absolute difference between the evening value and the morning value divided by the highest value) was calculated.

Exacerbations

Our definition of exacerbation was based on that of Fletcher as modified by Boman and colleagues [12]. When an exacerbation occurred, a 10-day course of oral prednisone was administered. Patients received 25 mg for 2 days, 20 mg for 2 days, 15 mg for 2 days, and so forth.

Symptoms and Adverse Effects

Using a scale of 0 to 4, all patients recorded, on a weekly basis, the presence and severity of symptoms (cough, phlegm, dyspnea, fatigue, disturbed sleep at night). The adverse effects of medication (dysphonia and oropharyngeal irritation) were recorded by the patients once every 3 months. Moreover, every 6 months, the presence and severity of oral candidiasis were assessed using a questionnaire (no, light, or severe symptoms).

Smoking

At the start of the study, smoking history was assessed in pack-years. During the study, the average number of cigarettes smoked per day was also recorded in weekly diary entries.

Power Calculations

Assuming that the clinically relevant, decreased annual decline in FEV₁ during beclomethasone treatment is 25 mL/y and that the residual standard deviation is 50 mL/y, the coefficient of variation is 25/50 or 0.5. Based on an of 0.05 and a of 0.20 (power:1 0.2, or 0.8), the required number of patients for the study would be 51. Based on an estimated dropout rate of 10%, the required initial number of study patients would be 56.

Statistical Analysis

Data on outcome variables obtained before and during beclomethasone therapy were compared. Differences were tested by repeated-measures analysis of variance, the paired Student t-test for normally distributed variables, and the Wilcoxon paired signed-rank test for non-normally distributed variables. Before the analysis, the PC ₂₀ values were ²log transformed. The annual changes in FEV₁ and PC ₂₀ were calculated by linear regression of individual FEV₁ and ²log PC ₂₀ values in the course of time. For group values, the individual regression coefficients were averaged. The weekly measured PEFR were averaged for 3-month periods and statistically compared with the value obtained just before the start of beclomethasone therapy. Patients with asthma and COPD were analyzed together and separately.

The gain in FEV₁ caused by an additional 2 years of treatment with beclomethasone was calculated. This calculation was done by comparing the end point of FEV₁ after the 2-year treatment with beclomethasone with the extrapolated end point if bronchodilator therapy alone had been continued during years 3 and 4 (point E in Figure 1). Because some degree of spontaneous improvement might have occurred during years 3 and 4 if beclomethasone had not been administered, we adjusted the extrapolated end point E for the disturbing influence of regression to the mean (point E in Figure 1). The effect of regression to the mean was calculated by the equations of Gardner and Heady [13] and Das and Mulder [14]. The gain in FEV₁ with and without correction for regression to the mean was calculated.

Results

Baseline Characteristics

Compared with patients who had COPD, asthmatic patients had lower values for smoking-related variables, were more often allergic, had a higher bronchodilating response in FEV₁, had more diurnal variation in peak flow, and had greater bronchial hyper-responsiveness at the start of the 4-year intervention study (Table 1). Of the 9 nonsmokers with COPD at the start of the 4-year study, 6 were ex-smokers. Only 3 patients with COPD had never smoked.

Withdrawal Rate

Of the 56 patients, 48 completed the 2-year treatment with beclomethasone. Reasons for withdrawing from the study included refusal to use corticosteroids (1 patient with asthma, 1 with COPD), bronchial carcinoma (1 patient with COPD), chronic heart failure (1 patient with COPD), persistent oral candidiasis and dysphonia (2 patients with COPD), and personal reasons (1 patient with asthma, 1 with COPD).

Compliance

The single-blind study of patient compliance with medication showed that mean compliance rates (SD) were 82% 30% for the prescribed amount of beclomethasone and 98% 29% for the additional bronchodilator. Compliance rates were not related to the changes in outcome measures (lung function, bronchial hyper-responsiveness, and exacerbations) during beclomethasone treatment (all P values > 0.2).

Forced Expiratory Volume in 1 Second

Because of patient selection, the annual decline in pre- and postbronchodilator FEV₁ before corticosteroid treatment was large (Table 2 and Figure 1). In the group as a whole, repeated-measures analysis of variance showed an overall treatment effect of beclomethasone during years 3 and 4 with respect to both the pre- and postbronchodilator FEV₁ (P = 0.0001 and P = 0.01, respectively). The increases in the pre- and postbronchodilator FEV₁ were 458 mL/y and 105 mL/y, respectively, during months 1 to 6 of beclomethasone therapy, which differed substantially from the annual decreases seen during bronchodilator therapy alone (difference, 618 mL/y [95% CI, 363 to 873 mL/y] and 203 mL/y [95% CI, 54 to 352], respectively) (Table 2). After the initial improvement, declines in the pre- and postbronchodilator FEV₁ were found during months 7 to 24 of corticosteroid treatment. However, the annual decline in prebronchodilator FEV₁ of 102 mL/y during months 7 to 24 of corticosteroid therapy was less than that observed before corticosteroid treatment (difference, 58 mL/y; 95% CI, 2 to 87) (Table 2). When changes in FEV₁ were assessed separately for patients with asthma and COPD, the effects of beclomethasone seemed to be more pronounced in patients with asthma than in those with COPD (Table 2).

The gain in prebronchodilator FEV₁ (adjusted for regression to the mean) was 356 mL in patients with asthma (95% CI, 197 to 515 mL) and 195 mL in patients with COPD (95% CI, 24 to 415 mL). No statistically significant gains in postbronchodilator FEV₁ values were observed. The effects of beclomethasone were the same with and without correction for regression to the mean.

Peak Expiratory Flow Rate

During the entire 2-year period of beclomethasone therapy, asthmatic patients showed significantly improved PEFR relative to the value at the start of beclomethasone treatment, with a mean maximal increase (SE) of 0.7 0.2 L/s observed during months 13 to 15 of beclomethasone treatment (P = 0.0006). The diurnal variation in the PEFR in patients with asthma decreased from 13.2% in the period of bronchodilator therapy alone to 9.7% during months 4 to 24 of beclomethasone treatment (P = 0.017). Patients with COPD showed a mean (SE) increase in the PEFR of 0.21 0.09 L/s during months 4 to 9 of beclomethasone therapy relative to the value at the start of steroid treatment (P = 0.04). Such an increase was not found for the rest of the beclomethasone treatment period. In patients with COPD, the diurnal variation in PEFR decreased from 10.7% in the period of bronchodilator therapy alone to 8.4% during months 4 to 24 of beclomethasone therapy (P = 0.0004).

Other Lung Function Indices

In patients with asthma, beclomethasone improved the course of several other lung function indices in addition to forced expiratory flow rates (Table 3). Relative to values observed with bronchodilator therapy alone, RV decreased by 0.49 L (95% CI, 0.18 to 0.80 L), the RV/TLC ratio diminished by 9% (95% CI, 3% to 15%), the IVC increased by 0.50 L (95% CI, 0.16 to 0.86 L), and the FIV₁ increased by 0.41 L (95% CI, 0.10 to 0.72 L) during treatment with beclomethasone. In patients with COPD, no statistically significant changes were observed (Table 3).

Histamine PC₂₀ Values

Histamine PC ₂₀ values over time are shown in Figure 2. In patients with asthma, the PC ₂₀ improved by 3.0 doubling doses/y (95% CI, 0.8 to 5.2 doses/y) during beclomethasone treatment relative to bronchodilator therapy alone. In patients with COPD, however, the PC ₂₀ showed no significant change. The annual change of 0.3 doubling doses/y before beclomethasone therapy was not different from the value of 1.6 doubling doses/y observed during beclomethasone treatment (P = 0.2) (see Figure 2).

Exacerbations

In patients with asthma, the number of exacerbations decreased from 1.3/y before beclomethasone therapy to 0.6/y during beclomethasone therapy (difference, 0.7/y; 95% CI, 0.4 to 1.0/y). In the first year of beclomethasone treatment, the duration of exacerbations decreased from 1.2 to 0.4 wk/y (difference, 0.8 wk/y; 95% CI, 0.5 to 1.1 wk/y). In patients with COPD, the number of exacerbations decreased from 1.7 to 1.3/y in year 2 of beclomethasone treatment (difference, 0.5/y; 95% CI, 0.1 to 0.9/y).

Symptoms

In patients with asthma, the severity of weekly recorded total symptoms showed a decrease of 17% during the first 3 months of beclomethasone therapy when compared with the period before beclomethasone therapy (P = 0.01). This also appeared to be the case for the dyspnea score (19%; P = 0.05) and the cough score (28%; P = 0.005). During the rest of the treatment period, no significant improvements were found for the total symptom score. In patients with COPD, the severity of cough, phlegm, and dyspnea was only reduced during months 7 to 12 of beclomethasone treatment (13%; P = 0.03).

Adverse Effects

On average, no increases in the severity of dysphonia and irritation of the oropharynx were observed during treatment with beclomethasone when compared with the severity observed just before treatment, neither in the group as a whole (all P values > 0.2) nor in asthma and COPD groups separately. However, the severity of oral candidiasis increased after 1 year of beclomethasone therapy when compared with the severity seen at the start of corticosteroid treatment (score, 0.21 compared with 0.02, P = 0.01). Because of oral candidiasis, six patients had to administer their medication with a spacer. Four patients were able to continue the study with a spacer, but two withdrew from the study because of persistent oral candidiasis and dysphonia.

Smoking

Among patients with asthma, 14 were smokers and 14 were nonsmokers during the first 2 years of treatment; 12 were smokers and 16 were nonsmokers during the third and fourth years (P > 0.05, two-sample paired proportion test). Two of the three asthmatics who stopped smoking had smoked less than 0.1 cigarettes/d during bronchodilator therapy alone. Therefore, these patients can be regarded as nonsmokers during the 4-year study period. One asthmatic patient started smoking during beclomethasone therapy. Among patients with COPD, 19 were smokers and 9 were non-smokers during the first 2 years of therapy; 17 were smokers and 11 were nonsmokers during beclomethasone treatment (P > 0.05, two-sample paired proportion test). One of the two patients with COPD who stopped smoking during beclomethasone therapy withdrew from the study for personal reasons.

Influence of Preceding Bronchodilator Treatment on Outcome Measures during Beclomethasone Therapy

Neither the treatment regimen in the first 2 years (continuous bronchodilator therapy or treatment on demand) or the sequence of the bronchodilators used during the study period (salbutamol-ipratropium bromide or ipratropium bromide-salbutamol) influenced the changes in outcome measures during beclomethasone therapy.

Delay in Reaching Severe Airflow Obstruction

Knowing the extent to which inhaled corticosteroids can induce a delay in reaching a certain low level of lung function (severe degree of chronic airflow limitation) is useful. Assuming that the decline in FEV₁ in the course of 10 to 20 years is stable, calculations based on data from our study indicate that the time in years before a certain (low) FEV₁ level of 0.5 to 1.0 L is reached in our patient population is almost doubled when inhaled corticosteroids are substituted (Figure 3).

Discussion

We found that in 48 patients with asthma or COPD, additional treatment with an inhaled corticosteroid (beclomethasone, 800 g daily) during a 2-year period improved the unfavorable disease course seen during bronchodilator treatment alone. The increase in FEV₁ reached a plateau during the first 6 months of beclomethasone treatment, after which a decline in FEV₁ was seen. However, the annual decline in FEV₁ during months 7 to 24 of beclomethasone treatment was significantly less than the decline seen during bronchodilator therapy alone.

Crossover and self-controlled study designs can produce results that are as statistically and clinically valid as results produced by a parallel-treatment study design, even with a lower required number of patients [15]. However, two conditions have to be fulfilled: No period or carry-over effects should be present. In our study, no carry-over effect of the preceding bronchodilator medication on either of the outcome measures during beclomethasone treatment was found. No period effect was assessed in the number of exacerbations and the FEV₁ decline. We can assume that the decline in FEV₁ would have continued if patients had not received additional therapy with beclomethasone during years 3 and 4 of the study. This assumption is based on three findings: First, a stable annual decline in FEV₁ during a period of several years (linear loss of FEV₁ in the course of time) was invariably found in studies in patients [16-19] and in random population samples [20-23]. Second, in the 56 patients included in our study, the annual decline in FEV₁ during bronchodilator therapy alone was stable. The mean declines (SE) during the first and second years were similar (149 42 mL/y and 156 60 mL/y, respectively). Third, regression to the mean did not explain the observed improvement in FEV₁ during beclomethasone therapy. No change in exacerbations occurred during the 4-year period in the patients who continued bronchodilator treatment during the third and fourth years of study. Our study meets the criteria for a valid, self-controlled trial. Therefore, from a methodologic point of view, a placebo group was not absolutely necessary and, more important, from an ethical point of view, a placebo group would have been unacceptable. It is unethical to assign patients who showed an average annual decline in FEV₁ of 160 mL/y and almost two exacerbations per year to a placebo group. The observed improvements in outcome measures during the third and fourth years of study must have been the result of additional beclomethasone therapy.

The large improvement seen in FEV₁ during the first 6 months of beclomethasone treatment probably was the consequence of a rapid initial decrease in the thickness of airway mucosa and submucosa by inhibition of edema, microvascular leakage, and mucus production from glands. The inflammatory processes underlying asthma or COPD in our study patients were probably not completely abolished by beclomethasone treatment because the annual decline in FEV₁ during months 7 to 24 of beclomethasone treatment was still larger than the physiologic decline of 20 to 40 mL/y observed in random population samples. Nevertheless, these findings indicate that inhaled corticosteroids not only have a short-term influence during several months but also have a long-term influence during 2 or more years. The clinically significant gain in FEV₁ after 2 years of beclomethasone treatment showed that patients were in a more favorable clinical condition than they would have been if bronchodilator treatment alone had been continued for another 2 years. The beneficial effect of beclomethasone was not statistically significant in the separate group of 22 patients with COPD. In our power analysis, we assumed that beclomethasone affected the annual decline in FEV₁ in both patients with asthma and COPD. For the entire group of 56 patients, however, we had sufficient statistical power to show that the annual decline in FEV₁ during beclomethasone treatment was significantly less than that seen before steroid therapy. The data suggest that this effect may be more evident in asthmatic patients (70 mL/y) than in patients with COPD (50 mL/y). Finally, our calculations showed that inhaled beclomethasone may double the time before patients reach low levels of lung function (FEV₁, 0.5 to 1.0 L) (see Figure 3). Of course, this is a simplification based on extrapolation from our FEV₁ data for a period of 10 to 20 years, but it gives an impression of the delay that inhaled corticosteroid therapy might induce in progression to severe degrees of chronic airflow limitation.

In addition to its effects on FEV₁, beclomethasone therapy had beneficial effects on static lung function indices and nonspecific bronchial responsiveness (in patients with asthma), peak-flow rate, exacerbations, and respiratory symptoms. The peak flow was statistically increased during the entire beclomethasone treatment period in patients with asthma, but such an increase was seen in patients with COPD only during months 4 to 9. Beclomethasone therapy had a beneficial effect on the indices of hyperinflation in patients with asthma (RV, RV/TLC), but no such effect was seen in patients with COPD. Improvements in these indices may also have a beneficial influence on the efficacy of gas exchange and may reduce the work of breathing [8]. The histamine PC ₂₀, which is an important indicator of the severity [10, 24] and perhaps even of the prognosis of asthma [25], improved in the study patients with asthma during beclomethasone therapy. The number of exacerbations, an important determinant of the subjective well-being of patients with asthma and COPD [8], decreased during beclomethasone therapy. The severity of weekly recorded symptoms decreased temporarily during beclomethasone treatment in both the asthma and COPD groups. Unfortunately, complaints of oral candidiasis increased during beclomethasone treatment, although all patients except two were able to continue beclomethasone therapy when the drug was inhaled with a spacer. With a spacer, the drug deposition in the oropharynx is diminished, thereby decreasing the severity of local side effects in this area [26, 27].

Morbidity and death due to asthma and COPD appear to have increased during the past two decades [1, 2]. It is possible that the administration of bronchodilator therapy without anti-inflammatory medication has contributed to this worldwide trend in morbidity and death. Bronchodilators are very effective in immediately diminishing bronchospasm and therefore in relieving symptoms directly, but they may have adverse effects on the control or progression of the disease when used continuously [5, 6]. Inflammation of the airway wall is a major factor in the pathophysiology of asthma [28] and perhaps of COPD [29], which makes anti-inflammatory treatment of major importance [30-32]. A problem of long-term treatment with inhaled corticosteroids is patient compliance. Patients may not notice the long-term improvements induced by corticosteroids. Indeed, among the patients with asthma in our study, only a temporary decrease in symptoms was found during the 2 years of beclomethasone therapy. In our study, compliance with beclomethasone, 400 g two times daily, was good but lower than compliance with the additional bronchodilator, four times daily. The good compliance rate in our study was probably the reason that we found no relation between individual compliance rates and the change in outcome measures during beclomethasone treatment. However, during uncontrolled conditions in clinical practice, compliance with medication can be much lower [26, 27]. Thus, physicians must disseminate proper information about prophylactic corticosteroid treatment and motivate patients to use their inhaled corticosteroids.

Some other long-term studies have shown the efficacy of inhaled corticosteroids in patients with asthma. Haahtela and colleagues [33] reported the results of a 2-year, randomized, double-blind study with budesonide, 600 g twice daily, and terbutaline, 375 g twice daily, in 103 newly identified patients with asthma. Budesonide was superior to terbutaline in its effect on the peak flow, the PC ₁₅, respiratory symptoms, and the use of rescue medication. Two recent 1-year studies by Juniper and colleagues [34, 35]one in 32 corticosteroid-dependent asthmatic patients and the other in 32 non-corticosteroid-dependent asthmatic patientshave shown that 1 year of budesonide therapy induced a gradual improvement in PC ₂₀, the severity of asthma symptoms, number of exacerbations, and bronchodilator use. The corticosteroid-dependent asthmatic patients also showed an improvement in FEV₁. Our study confirmed that inhaled corticosteroids are a first-line therapy for asthma, particularly when the need for bronchodilators is increasing [30-32].

The efficacy of inhaled corticosteroid therapy in COPD has long been a controversial subject [8]. No prospective long-term, controlled studies have been reported, although two retrospective studies have suggested that corticosteroid therapy has a positive effect on the annual decline in FEV₁ [36, 37]. Several short-term studies have found that corticosteroids have no effect on COPD, whereas others have observed the opposite [38]. In a recent editorial, Stoller [38] suggested that these drugs are potentially of benefit in COPD. In our 4-year prospective study, inhaled beclomethasone therapy had a beneficial influence on the course of the FEV₁, the diurnal variation in PEFR, the number of exacerbations, and the severity of symptoms in patients with COPD. Therefore, a rapid annual decline in FEV₁ during bronchodilator therapy alone is probably a good indication that additional treatment with inhaled corticosteroids in patients with COPD is necessary and effective.

Abbreviations

COPD: chronic obstructive pulmonary disease

FEV₁: forced expiratory volume in 1 second

FIV₁: forced inspiratory volume in 1 second

FRC: functional residual capacity

FVC: forced vital capacity histamine

PC₂₀: concentration of histamine that provokes a 20% decrease in FEV₁

IVC: inspiratory vital capacity

PEFR: peak expiratory flow rate

RV: residual volume

TLC: total lung capacity