INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A significant proportion of patients with chronic obstructive pulmonary disease (COPD) often exhibit little reversibility of airflow obstruction after the acute inhalation of bronchodilators (1), with up to one-third showing no or minimal response. However, bronchodilators have been shown to confer symptomatic relief and improvement of the functional capacity even in patients with "fixed" airflow obstruction (2, 4, 5). Thus, their use in the symptomatic treatment of the disease is generally recommended in current guidelines for management of COPD (1).

Inhaled short-acting ₂-adrenoceptor agonists, like salbutamol, and anticholinergic agents, like ipratropium bromide, are the preferred bronchodilator agents for treatment of COPD, with theophylline being added as necessary (1, 2). Ipratropium bromide has been reported to produce greater bronchodilation than conventional doses of a short-acting ₂-adrenoceptor agonist in one study (6), whereas equivalence of the two treatments has been shown in another (7). Because of its longer duration of action and better tolerability compared with short-acting ₂-adrenoceptor agonists, ipratropium bromide has been considered more suitable for use on a regular basis.

The development of inhaled ₂-adrenoceptor agonists with a prolonged duration of action, such as formoterol and salmeterol (8), has represented a useful therapeutic advance for the management of COPD (9). These agents may become an effective alternative to ipratropium bromide for regular treatment of the disease, especially in patients with nighttime or early morning symptoms, but more information is required on their efficacy and safety during long-term use in patients with COPD and in comparison with ipratropium (17).

The purpose of this study was to compare the effects of a 12-wk treatment with formoterol, ipratropium bromide, and placebo (with on demand salbutamol) on airflow obstruction, symptoms, and quality of life, and to assess the safety profile of the two active medications in patients with COPD.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Patients were male or female outpatients aged  40 yr, who were either current or ex-smokers of > 10 pack-years, had a diagnosis of COPD according to the American Thoracic Society guidelines (2, 18), and gave written informed consent.

Inclusion criteria required that their forced expiratory volume in 1 s (FEV₁) was < 70% of predicted and > 0.75 L, with the ratio FEV₁/vital capacity (VC) < 88% of that predicted in men and < 89% of that predicted in women (1, 18). Daytime and/or nighttime symptoms were to be present on at least 4 of the last 7 d of the run-in period.

Specific exclusion criteria were current or past diagnosis of asthma, a respiratory tract infection in the previous month, need for long-term oxygen therapy, QT_c > 0.46 s, initiation or discontinuation of inhaled corticosteroids, or change in daily dose of these drugs in the month preceding the screening visit, treatment with parenteral or oral corticosteroids in the previous month, current treatment with theophylline (any formulation), oral or inhaled anticholinergics, and oral or inhaled long-acting ₂-adrenoceptor agonists.

Patients were recruited after approval of the local ethics committees.

Study Design

This was a multicenter, double-blind, randomized, parallel-group, placebo-controlled study. Ipratropium bromide was chosen as the active comparator because it is considered the first-line long-term therapy in patients with COPD (1). Blinding was obtained by double-dummy dosing.

After screening at Visit 1, patients entered the run-in period of 10- 21 d during which they became accustomed to the trial procedures and baseline measurements were performed. During this period, patients received placebos matched to both formoterol and ipratropium bromide, and inhaled salbutamol (100 µg/puff) as rescue medication.

Eligible patients were then randomized to receive one of the following four regimens for 12 wk: (1) formoterol 12 µg (F12): inhaled formoterol dry powder inhalation capsules 12 µg twice daily via the single dose, breath-activated Aerolizer inhaler (Foradil Aerolizer, Novartis) and placebo matching ipratropium bromide four times daily; (2) formoterol 24 µg (F24): same as above but with formoterol 24 µg; (3) ipratropium (IPR): ipratropium bromide 40 µg four times a day via a pressurized metered dose inhaler (Atrovent, Boehringer Ingelheim) and placebo matching formoterol twice daily; and (4) placebo (PL): placebo matching formoterol twice daily and placebo matching ipratropium bromide four times daily.

Patients on stable inhaled corticosteroid treatment were allowed to remain on that treatment throughout the trial. Rescue medication with inhaled salbutamol was allowed throughout the study up to a maximum of 8 puffs/d. Short courses (< 15 d, not more than twice, and separated by at least 2 wk) of antibiotics, oral corticosteroids, and/or oxygen were permitted in case of exacerbation or respiratory infection. Patients who needed additional medications for COPD were to be withdrawn from the study.

Our primary efficacy variable was the area under the curve (AUC) for FEV₁ measurements performed over 12 h following the morning dose of study medications after 12 wk of treatment. FEV₁ was recorded by spirometry before the morning dose of study medications and then at 5, 15, 30 min, 1 h, and hourly up to 12 h after the morning dose. The AUC (in liters × minutes) obtained by computing absolute FEV₁ values versus time (12 h) was divided by the time (in minutes) the patient was actually observed to obtain the normalized AUC (in liters). The patients had to abstain from taking rescue medication within 6 h before the start of serial spirometric measurements.

Secondary efficacy variables were the normalized AUC for FEV₁ measurements performed over 12 h after the first dose of study medications, FEV₁ at all the individual time points during the 12 hour spirometry, predose FEV₁ at all visits, daily morning premedication peak expiratory flow (PEF), daily number of puffs of rescue salbutamol and daily total patient symptom scores from the patient diary averaged over 12 wk, frequency of COPD exacerbations, and quality of life (QOL) scores at the end of treatment.

The patients were given a patient diary that contained spaces for recording the use of rescue medication and for scoring the following symptoms on a four-point scale (0 = best to 3 = worst) up to a maximum of 18/d: ability to perform usual daily activity, breathlessness over the past 24 hours, waking at night due to respiratory symptoms, breathlessness on rising, cough, and sputum production.

Three levels of COPD exacerbation were identified as follows: first level: "bad days," defined as days with at least two individual symptom scores of 2 or more and/or a reduction in PEF from baseline (averaged value over the last 7 d of the run-in period) of more than 20%; second level: course of additional therapy (corticosteroids, antibiotics, or oxygen); third level: COPD-related hospitalizations.

QOL was measured before the first dose of study medications and at the end of the treatment period by the validated St. George's Respiratory Questionnaire (SGRQ) (19, 20).

Vital signs were obtained during the screening visit. At randomization and at the end of the 12-wk study period, they were obtained before the morning dose and at 1, 2, 4, and 12 hr postdose. Electrocardiograms and clinical laboratory evaluations were performed at randomization and at the end of treatment. Adverse events were recorded throughout the study.

Statistical Analysis

On the basis of previous studies of the sponsor, the between-patient standard deviation for the primary efficacy variable was assumed to be 400 ml and a difference of 120 ml between treatment groups was considered clinically relevant. This is in agreement with the data on FEV₁ variability reported in a recent meta-analysis of studies carried out in patients with reversible airflow obstruction (21). The sample size was calculated as 175 patients per treatment group (two-sided  = 0.05, 1   = 80%). To allow for an expected dropout rate of 15%, the recruitment of 824 patients was initially planned (206 patients per group). This number was reduced to 770 when it was found that the actual dropout rate was around 10%.

Statistical analysis was carried out according to the intent-to-treat principle. The analysis of covariance (ANCOVA) was used to estimate all treatment differences for the normalized AUC-FEV₁. Baseline FEV₁ value (last FEV₁ measured before randomization) was used as a covariate. All the other variables were analyzed by ANCOVA or the van Elteren test, as appropriate. For QOL, a difference of four points or more was considered clinically relevant (19).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

A total of 935 patients were screened and 780 were randomized into this study. Of these 82 patients discontinued from the study prematurely (13 on F12, 23 on F24, 29 on PL, and 17 on IPR). The main reason for discontinuation was adverse events: 6 on F12, 14 on F24, 11 on PL, and 9 on IPR. Treatment with F12 was associated with less than half the number of discontinuations compared with PL. Otherwise, there were no meaningful differences in numbers of patients discontinuing or in the reasons for discontinuation among the treatment groups.

The demographic and baseline characteristics of patients are summarized in Table 1. Concomitant medications were inhaled corticosteroids, which were used by 47%, 53%, 54%, and 52% of the patients on F12, F24, PL, and IPR, respectively. The mean daily dose of the most frequently used inhaled corticosteroids, budesonide and beclomethasone dipropionate, was 996 µg and 854 µg in the F12 group, 1,046 µg and 848 µg in the F24 group, 963 µg and 775 µg in the PL group, and 774 µg and 902 µg on IPR. ₂-Adrenoceptor agonists were used by 22%, 21%, 21%, and 23% of the patients on F12, F24, PL, and IPR, respectively. The percentage of patients who required oral corticosteroids over the treatment period was 7% on F12, 8% on F24, 8% on PL, and 10% on IPR. Corresponding values for patients who required antibiotics were 13%, 14%, 12%, and 14%.

Following the inhalation of 200 µg salbutamol at the screening visit, and on the basis of their FEV₁ response within 30 min, 56.2%, 55.7%, 59.5%, and 60.3% of the total patients randomized to F12, F24, PL, and IPR, respectively, presented with a change in FEV₁ of < 15% from baseline value and < 200 ml in absolute value. This is a finding suggestive of a relatively "fixed" airflow obstruction in these patients (3). With the exception of only one patient taking ipratropium, no patient in the study attained a postsalbutamol FEV₁ compatible with asthma (i.e., an FEV₁ > 80% of predicted normal) (3).

Primary Efficacy Variable

ANCOVA of the normalized AUC-FEV₁ at the end of the treatment period showed that both F12 and F24 were superior to PL (Table 2). The differences were highly statistically significant and both exceeded the 120-ml improvement deemed to be clinically relevant.

The comparison between IPR and PL showed a statistically significant and clinically relevant difference in favor of IPR (Table 2), confirming the sensitivity of the trial.

Secondary Efficacy Variables

The ANCOVA of the normalized AUC-FEV₁ after the first dose of trial medications showed that both F12 and F24 were superior to PL. The differences were significant (p < 0.001) and clinically relevant (217 ml for F12 and 223 ml for F24). Both F12 and F24 also showed statistically significant improvements when compared with IPR (p  0.024). IPR gave an estimated improvement over PL that was significant both statistically (p < 0.001) and clinically (137 ml).

The 12-h profile plots of mean FEV₁ values following the first dose of trial medications and after 12 wk of treatment are displayed in Figure 1. After the first dose of trial medication, both F12 and F24 produced a significant improvement of FEV₁ over PL that exceeded 120 ml at every time point from 5 min after dosing up to 12 h (all p < 0.001). Similar statistically and clinically significant results in favor of F12 and F24 were obtained after the last dose of trial medication (all p < 0.001 between 5 min and 12 h). F12 or F24 was significantly better than IPR at most time points (Figure 1). The estimated improvement by F12 over IPR was in excess of 120 ml at 5, 15, and 30 min, 1, 2, 4, and 5 h following the last dose of trial medication.

View larger version (32K): [in this window] [in a new window]   Figure 1.   Twelve-hour profile of mean FEV1 measurements performed following the morning dose of study medications on the first day (A) and after 12 wk (B) of treatment. The crosses indicate time points when both formoterol doses were statistically significantly better than placebo. The asterisks indicate statistically significant differences between formoterol and ipratropium. At each time point, a difference in mean FEV1 of 120 ml between treatment groups was considered clinically relevant. Intent-to-treat population: F24 = formoterol 24 µg; F12 = formoterol 12 µg; IPR = ipratropium bromide; PL = placebo.

The mean values of morning premedication PEF were averaged over the entire treatment period and analyzed by ANCOVA, with the mean values over the last 7 d of the run-in period (baseline) used as a covariate. Treatment contrasts and the associated 95% confidence interval (CI) are reported in Table 3. Both F12 and F24 were significantly more effective than PL and IPR. No significant difference between IPR and PL was detected.

Improvement in mean total diary symptom score, averaged over the whole treatment period, was statistically significant for both F12 (p < 0.001) and F24 (p = 0.007) when compared with PL. F12 also produced a significant improvement when compared with IPR (p = 0.009), whereas F24 approached significance (p = 0.060) (Figure 2). There was no significant difference between IPR and PL (p = 0.439).

View larger version (23K): [in this window] [in a new window]   Figure 2.   Mean total diary symptom scores averaged over the last 7 d of the run-in period (baseline value) and over all the treatment period. Intent-to-treat population. F24 = formoterol 24 µg; F12 = formoterol 12 µg; IPR = ipratropium bromide; PL = placebo. *p < 0.001 versus placebo and p = 0.009 versus ipratropium; **p = 0.007 versus placebo.

The mean daily number of puffs of inhaled salbutamol averaged over the treatment period was 1.2 in the F12 group, 1.7 in the F24 group, 2.5 in the PL group, and 2.0 in the IPR group. Both F12 and F24 produced a significant reduction in the use of rescue medication versus PL (all p < 0.001) and versus IPR (all p  0.014). There was no statistically significant difference seen between IPR and PL (p = 0.147).

The percentage of "bad days" was significantly smaller in the F12 and F24 groups than in the PL group (p < 0.001 and p = 0.001, respectively) and compared with the IPR group (p < 0.001 and p = 0.01, respectively). No significant difference was seen between IPR and PL (p = 0.414).

There was no significant difference among treatment groups in terms of number of days of additional therapy required for COPD exacerbations (corticosteroids, antibiotics, or oxygen). A total of 14 patients had a COPD-related hospitalization during the study, with two hospitalizations in each of the two F12 and F24 groups, four in the PL group, and six in the IPR group.

QOL was assessed before and at the end of the treatment period and the mean absolute SGRQ Total score and subscores are reported in Table 4. To test for treatment effects, the data were analyzed by ANCOVA, fitting the scores at the randomization visit (baseline) as a covariate. After 12 wk of treatment, both F12 and F24 showed statistically significant improvement in SGRQ Total score when compared to PL (Figure 3). F12 demonstrated a difference that exceeded the four points considered to be clinically relevant (5.06) (Figure 3). When compared with IPR, F12 showed statistically significant improvement that approached clinical relevance (3.79). The effect of IPR on total SGRQ score was not different from PL.

View larger version (15K): [in this window] [in a new window]   Figure 3.   Treatment group contrasts of SGRQ Total scores after 3 mo of treatment. Results of the ANCOVA are expressed as the estimated differences between least-square treatment means and the associated 95% confidence interval. Baseline SGRQ Total score measured at the randomization visit was fitted as a covariate. Negative differences indicate a reduction in score and improved quality of life. The horizontal dashed line shows the threshold of clinically significant differences. F24 = formoterol 24 µg; F12 = formoterol 12 µg; IPR = ipratropium bromide; PL = placebo; SGRQ = St. George's Respiratory Questionnaire.

Considering the three domains of the total SGRQ scores, F12 produced statistically and clinically significant improvements in Symptoms (p = 0.003, -5.12), Activity (p = 0.01, 5.42), and Impacts (p = 0.001, 5.00) subscores in comparison with PL. F12 also induced a statistically significant improvement in all three subscores when compared with IPR (all p  0.036), with treatment differences that were clinically relevant for the Activity (4.25) and Impacts (4.04) subscores. F24 produced statistically significant improvements in Symptoms and Activity subscores when compared with PL (all p  0.037), and the difference was clinically relevant for Symptoms (4.05). As seen with the Total scores, the effect of IPR was not different from PL for any of the three subscores.

Adverse Events and Other Safety Variables

Similar adverse event incidence was observed among all treatment groups. The most frequent adverse events were viral infections, exacerbations of COPD, headache, upper respiratory tract infection, pharyngitis, chest pain, coughing and dyspnea, without significant differences between treatment groups. Thirteen patients reported cardiovascular adverse events (one on F12, two on F24, five on PL, and five on IPR). Heart rate and rhythm disorders were uncommon across the treatment groups (six on F12, four on F24, eight on PL, and eight on IPR).

In total, 167 adverse events in 106 patients were considered drug related by the investigator: 22 patients (11%) on F12, 37 patients (19%) on F24, 24 patients (12%) on PL, and 23 patients (12%) on IPR. The most frequently occurring drug-related events were headache, tremor, dry mouth, muscle cramps, coughing, COPD exacerbation, dyspnea, and pruritus.

A total of 28 serious adverse events occurred during the trial, 10 of which led to premature discontinuation. There was no death. Of the 28 serious adverse events, eight were on F12, three on F24, nine on PL, and eight on IPR. Eighteen were respiratory events (four on F12, three on F24, five on PL, and six on IPR). Of the 10 serious adverse events leading to premature discontinuation, one was on F12, two on F24, four on PL, and three on IPR.

The numbers of patients reporting COPD-related adverse events were 25 (13%) on F12, 37 (19%) on F24, 37 (19%) on PL, and 37 (19%) on IPR. Very few of these adverse events were considered to be serious, with only seven leading to premature discontinuation from the trial (two on F24, three on PL, and two on IPR).

There was a very low incidence of clinically relevant abnormal serum potassium or fasting glucose level and no difference among treatment groups. Only one patient had a clinically significant electrocardiogram alteration after 12 wk that was not present at baseline. This patient suffered from atrial fibrillation and he was receiving treatment with F12. The percentage of patients whose QT_c became longer than 0.46 s after treatment was not significantly different among treatment groups.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

An effective treatment of COPD with bronchodilators should safely reduce airflow obstruction, decrease the frequency and severity of symptoms by reducing the amount of dynamic hyperinflation, and improve the QOL (1, 22, 23). In this respect, the results of this study indicate that 12-wk treatment with formoterol dry powder inhalation capsules 12 and 24 µg twice daily is safe and effective in patients with COPD, and that it is more effective than treatment with ipratropium bromide 40 µg four times daily for most outcomes, as discussed below.

Analysis of the primary variable, the normalized AUC-FEV₁ following the morning dose of study medications at the end of the treatment period, revealed that the bronchodilator effect of both doses of formoterol was statistically and clinically significant compared with placebo and statistically significant compared with ipratropium. Similar results were obtained with the analysis of the normalized AUC-FEV₁ after the morning dose of study medications on the first day of treatment.

The analysis of FEV₁ at individual time points after the morning dose of study medications showed that a significant bronchodilator effect over placebo was detectable with both doses of formoterol as early as 5 min after dosing and that it persisted for at least 12 h. The bronchodilating action of formoterol was significantly better than that of ipratropium at most time points. Estimation of the treatment effect on morning premedication PEF confirmed that bronchodilation by both doses of formoterol was long lasting and still evident 12 h after dosing, whereas the effect of ipratropium was not sustained from the evening to the morning measurement.

We compared the mean bronchodilating effects of formoterol and ipratropium and did not evaluate the existence of individual differences in patient response to these agents. Although some studies have indicated that certain patients with COPD may respond better to anticholinergic than to ₂-adrenoceptor agonists (24, 25), other reports have failed to document these findings (26, 27). Individual patients in this study may be more responsive to ipratropium than formoterol, but our evaluation of mean responses will not demonstrate this.

In this study, the improvement in pulmonary function by formoterol was associated with a reduction in daily symptoms, use of rescue medication, and frequency of "bad days" (see STUDY DESIGN for definition), whereas ipratropium had no significant effect on these parameters. Although the study was probably too short to capture a significant numbers of disease deteriorations, the ability of formoterol to reduce the number of "bad days" suggests that it may alter the course of acute exacerbations of COPD. This may be related to the inhibitory effect of ₂-adrenoceptor agonists on plasma exudation and neutrophil migration (28, 29). Because about 50% of the patients in each treatment group used inhaled corticosteroids, it is also possible that the addition of a long-acting ₂-adrenoceptor agonist may have produced independent or additional reductions in the expression of adhesion molecules and activation of inflammatory cells in the airways as observed in patients with asthma (30). However, in our investigation the "bad days" were predominantly defined on the basis of PEF values and therefore the beneficial effect of formoterol in comparison with ipratropium may simply be due to its more potent bronchodilator activity, with no effect on the actual course of the exacerbations.

The patients on formoterol, particularly those on treatment with the dose of 12 µg twice daily, perceived the beneficial impact on the quality of their daily living. Compared with placebo, this treatment induced a statistically and clinically significant improvement in all the dimensions of the QOL questionnaire, including distress due to respiratory symptoms (Symptoms domain), disturbance to mobility and physical activity (Activity domain), psychosocial impact of the disease (Impacts domain), and global estimation of respiratory health (Total score). Compared with ipratropium, this treatment induced a statistically significant improvement in all the dimensions of the SGRQ, which was also clinically relevant for the Activity and Impacts scores and approached clinical significance for the other domains. Although ipratropium also significantly improved pulmonary function, this did not translate into clinical benefit as its effect on indicators of the daily level of disease control, such as symptom scores and rescue use, was not different from placebo. This was reflected in QOL scores where the effect of ipratropium again did not differ from the placebo effect.

In a previous study (16) comparing the efficacy and safety of salmeterol (42 µg twice daily) and ipratropium bromide (36 µg four times daily) for 12 wk in a similar population of patients with COPD, both salmeterol and ipratropium significantly improved QOL. The disparity between the results of this study and our results may be due to the use of different questionnaires to assess QOL and different statistical methods to test for the treatment effect.

The outcome of the present studythat the lower formoterol dose seemed to perform better than the higher dose for some QOL variableswas unexpected, given the other efficacy and safety results. However, a similar outcome was also reported in a study (14) investigating the effects on QOL of 16-wk treatment with salmeterol 50 and 100 µg twice daily in patients with COPD.

One limitation of our study is that ipratropium bromide was tested at the dose of 40 µg (measured ex valve and corresponding to 36 µg in the United States) four times a day. This dose may be suboptimal in some patients (6, 7, 31), and the use of doses two or three times higher is reported in some guidelines for COPD treatment (2, 22). However, the dose of ipratropium tested here is the usual maintenance dose approved for clinical use in patients with COPD by health authorities worldwide. What should also be noted is that though ipratropium was given four times a day and formoterol twice a day, it was formoterol that gave a better 24-h coverage to the patient.

We did not compare the efficacy of formoterol with that of the other available long-acting ₂-adrenoceptor agonist, salmeterol. Most of the published comparison studies in COPD (9, 10, 32) investigated the acute effects of formoterol and salmeterol. In these studies, the onset of bronchodilation was more rapid with formoterol than salmeterol (9, 32), salmeterol 50 µg and formoterol 12 µg provided comparable bronchodilation over 12 h in patients with a range of COPD severity (32), and salmeterol 50 µg was more active than formoterol 12 µg and 24 µg in a group of patients with severe disease (10).

In a study comparing the effect of regular treatment with formoterol and salmeterol in patients with reversible obstructive airway disease (33), formoterol 12 µg twice daily and salmeterol 50 µg twice daily, both formulated as dry powders, demonstrated similar efficacy and safety profile over a 6-mo period of treatment. To our knowledge, a similar direct comparison of efficacy and safety during regular treatment in patients with a specific diagnosis of COPD is still lacking.

This study was not specifically addressed to investigate the occurrence of tachyphylaxis upon repeated administrations of formoterol in patients with COPD, but the magnitude and duration of the bronchodilator response following the first and last morning dose were similar, indicating that the bronchodilating activity of formoterol did not diminish during the treatment period. This supports the view that the bronchodilator effect of long-acting ₂-adrenoceptor agonists is fairly stable in patients with COPD on regular treatment with these agents (17).

In conclusion, in this study formoterol was statistically and clinically better than placebo and ipratropium bromide in improving lung function over the 12 wk of treatment. Compared with ipratropium bromide, considered first-line therapy for the long-term treatment of stable, symptomatic COPD (1), formoterol significantly reduced symptoms, the frequency of "bad days," and the need for short-acting bronchodilators on demand for symptom relief, while showing a similar safety and tolerability profile. Importantly, it was effective in improving the patient QOL whereas ipratropium had an effect on QOL similar to placebo.

The twice-daily dosing schedule of formoterol, its fast onset of action, and the perception of benefit by the patients should enhance compliance and minimize the possibility of overdosage. The clinically significant improvement of QOL by treatment with formoterol could reduce disability, lost productivity, and the medical care costs.

Together, these observations indicate that formoterol 12 or 24 µg twice daily could be used as the first-line therapy for stable, symptomatic COPD.