INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Chronic obstructive pulmonary disease (COPD) is characterized by limitation of expiratory airflow (1, 2). Although airflow may improve significantly after treatment, by definition, airflow will never normalize (1). As a result, most patients with COPD are currently treated with regularly administered bronchodilators, with the goal of optimizing the limited airflow as continuously as possible. However, currently available inhaled agents have durations of action of 4 to 6 h and require administration four times daily or more.

Salmeterol is a long-acting -agonist bronchodilator (3). It stimulates ₂ receptors and increases cyclic AMP, similar to short-acting -agonists. However, salmeterol also interacts with an "exosite" in the receptor (4, 5). By this mechanism, salmeterol remains associated with the receptor, thus providing a prolonged duration of action. In patients with asthma and COPD, salmeterol is effective for up to 12 h (3, 6). This duration of action makes salmeterol suitable for twice daily administration and suggests that the drug may provide advantages over short-acting therapies when regular continuous bronchodilation is required.

The current study was designed to determine whether salmeterol is an effective bronchodilator in patients with COPD. Salmeterol administered twice daily was compared with the well-established bronchodilator ipratropium administered four times daily and with placebo. The bronchodilatory effects of each of these treatments were compared over a 12-h interval before, during, and after 12 wk of continuous therapy. In addition, the ability of the study treatments to improve subjective responses, such as dyspnea and quality of life measures, was assessed.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Enrollment Criteria

Patients were required to be at least 35 yr of age, have an FEV₁ of > 0.70 L and  65% of predicted, and an FEV₁/FVC ratio of  70% at initial screening. Patients with an FEV₁ < 0.70 L were eligible if the FEV₁ was  40% of predicted. Responsiveness to both inhaled albuterol (Ventolin Inhalation aerosol, 180-360 µg; Glaxo Wellcome, Research Triangle Park, NC) and ipratropium (Atrovent inhalation aerosol, 36 µg; Boehringer Ingelheim Pharmaceuticals, Ridgefield, CT) was assessed at entry. Patients were stratified in each treatment group by their response to albuterol. Patients with a  12% improvement and > 200-ml increase in FEV₁ after albuterol administration were considered responsive; those with less response to albuterol were considered nonresponsive. All patients had to score at least 1 (1 = troubled by shortness of breath when hurrying on the level or walking up a slight hill) on the Modified Medical Research Council (MMRC) five-point dyspnea scale (0 = not troubled with breathlessness except with strenuous exercise to 4 = too breathless to leave the house or breathless when dressing or undressing). Patients were excluded from the study if they experienced a pulmonary infection within the 4 wk before the study, had significant cardiovascular disease, a history of malignancy within the prior 2 yr, a significantly abnormal electrocardiogram (ECG), physical limitations that interfered with study compliance, abnormal laboratory results, or a history of hypersensitivity to any -agonist or anticholinergic compound. Patients were required to discontinue the use of theophylline, ipratropium (except as used in the ipratropium treatment group), and oral -agonists for the duration of the study. Patients who required daily continuous oxygen therapy or oral corticosteroid therapy greater than the equivalent of prednisone at 10 mg/d were excluded. Patients using inhaled corticosteroids at entry must have maintained a stable regimen for the duration of the study.

Study Design

This was a 12-wk, randomized, double-blind, placebo-controlled, parallel-group study comparing salmeterol xinafoate (Serevent inhalation aerosol; Glaxo Wellcome) (42 µg twice daily with placebo administered twice daily to maintain blinding), ipratropium (36 µg four times daily), and placebo (four times daily). All study medications were administered in identical canisters and actuators. The primary efficacy end points were area under the curve of forced expiratory volume in 1 s (FEV₁ AUC) and the Baseline Dyspnea Index/Transitional Dyspnea Index (BDI/TDI) assessment. Assessments of pulmonary function (spirometry and diffusing capacity), the 6-min walk and Borg dyspnea assessment, BDI (5, 10), symptom scores, and quality of life were conducted over a 2-wk baseline period. Safety assessments included adverse events, clinical laboratory tests, physical examinations, 12-lead ECG, and continuous 24-h ambulatory electrocardiograph (Holter monitor) in a subset of patients. Study visits were scheduled every 2 wk over the 12-wk treatment period. All sites had institutional review board (IRB) approval and all patients provided written informed consent.

Procedures

Pulmonary function. Pulmonary function tests were performed in triplicate with spirometric equipment that met or exceeded the minimal performance standards of the American Thoracic Society (11). Albuterol use was not allowed within 6 h before pulmonary function testing (PFT). At treatment Day 1 and Weeks 4, 8, and 12, PFTs were performed predose and at 0.5, 1, 2, 4, 6, 6.5, 8, 10, and 12 h after administration of study medication. If a patient required bronchodilator therapy during the serial PFT assessment, the data were not collected. The salmeterol group received salmeterol at time zero (t = 0) and placebo immediately after the Hour 6 PFT assessment. The ipratropium group received ipratropium at time zero and immediately after the Hour 6 PFT assessment. The placebo group received placebo at both t = 0 and immediately after the Hour 6 PFT assessment.

Each patient was instructed to perform triplicate peak expiratory flow (PEF) measurements with a Mini-Wright peak flow meter (Keller Medical Specialties, Waukegan, IL) each day in the morning before taking the first dose of study drug and in the evening immediately before taking the last dose of study drug. The highest of the triplicate values was recorded on diary cards. PEF measures taken within 4 h of supplemental albuterol use were not used in the analysis.

Dyspnea assessment. The Baseline Dyspnea Index (BDI) consists of three domains (functional impairment, magnitude of task, and magnitude of effort) that were summed to calculate the baseline focal score (10, 12). Each category was scored on a five-point scale from 0 to 4. Total baseline focal score could range from 0, very severe functional impairment and breathless at rest, to 12, no functional impairment and breathless only under extreme circumstances. The Transitional Dyspnea Index (TDI) was determined at each visit on treatment Weeks 2 through 12 and at the follow-up visit. Scores could range from 3 (major deterioration) to 0 (unchanged) to +3 (major improvement) for each of the three domains used to generate the baseline focal score. The three domains were summed to calculate the transition focal score, ranging from 9 to +9.

Six-minute walk. Baseline distance walked in 6 min was established during the run-in period. Patients performed two practice walks to minimize the effects of learning. At treatment Weeks 2, 6, and 10 patients performed a 6-min walk along a straight, flat corridor marked at 5- or 10-yd intervals. Patients were instructed to walk as far as possible during 6 min, stopping at any time if necessary. All investigating sites supervised the walks and gave a uniform set of instructions to study patients in order to minimize the effect of encouragement. Dyspnea was assessed immediately before and immediately after completing the 6-min walk, using the Borg Scale (0 = no dyspnea at all, 10 = maximal dyspnea). Dyspnea was defined as an uncomfortable need to breathe and did not include other sensations associated with exercise such as fatigue or increased breathing rate.

Patient self-assessments. Throughout the study, patients recorded their daytime and nighttime use of supplemental albuterol on their diary cards as number of puffs from the metered dose inhaler (MDI) or number of nebules used in nebulized treatments. Patients also rated severity of symptoms (before PEF measurements), recording a daytime and nighttime impression of chest tightness, shortness of breath, and cough. Daytime symptoms were recorded each evening on the basis of a five-point scale (0 = no symptoms at all to 4 = symptoms so severe as to preclude most daily activities). Nighttime symptom scores were recorded each morning (0 = no symptoms during the night to 4 = symptoms so severe that the patient did not sleep at all). Each morning, patients also recorded the total number of respiratory-related awakenings per night, including premature early morning awakenings.

COPD exacerbations. An exacerbation was defined as worsening symptoms of COPD requiring a change in drug therapy. Patients who experienced an exacerbation during the baseline period, required parenteral steroids, or were hospitalized for an exacerbation during the study were withdrawn. Treatment of exacerbations with oral corticosteroids for 14 d or less was permitted without withdrawing the patient from the study.

Quality of life. Quality of life (QOL) was assessed with the Chronic Respiratory Disease Questionnaire (CRDQ) before therapy, on treatment day 1, at treatment Weeks 2, 6, 8, and 12, and at the follow-up visit (13, 14). Day 1 scores were used as baseline values. The CRDQ measures four domains: dyspnea, fatigue, emotional function, and mastery. For the dyspnea domain, patients listed their five most frequent and important activities and their shortness of breath while performing these activities. Each domain question was rated on a seven-point scale (1 = the most severe impact and 7 = the least impact or most healthy situation). The values for all four domains were summed to give an overall score.

Statistical Analyses

It was estimated that 100 patients per treatment arm would provide > 80% power for detection of significant differences of 1 in the BDI/ TDI Transitional Focal Score and differences of 0.7 in the FEV₁ AUC at a significance level of 0.05 (10, 12). Statistical comparisons of treatment means were performed by analysis of variance except for Borg dyspnea scores (van Elteren test) and clinical laboratory data (Wilcoxon sign rank test). Comparisons between treatment proportions were performed using Cochran-Mantel-Haenszel tests except for adverse events (Fisher exact test) and time to exacerbation (Kaplan- Meier test). Analyses were performed on three groups of patients: (1) all patients, (2) albuterol-responsive patients, and (3) albuterol-nonresponsive patients. The Fisher exact test was used to assess the differences between groups for total numbers of dropouts and for each reason for dropout. Dropout rate was assessed by Kaplan-Meier analysis. Last observation carried forward (LOCF) analyses were performed on all efficacy measurements to investigate the impact of dropouts, and LOCF analyses (data not shown) were entirely consistent with those presented here.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Four hundred and five patients with COPD from 27 clinical centers were randomized into three treatment groups: 132 were treated with salmeterol, 138 with ipratropium, and 135 with placebo. The treatment groups were similar for demographic characteristics and baseline pulmonary function (Table 1).

Lung Function

Pulmonary function, as measured by FEV₁ and FVC, was significantly improved with both salmeterol and ipratropium compared with placebo on Day 1 of treatment (Figures 1A and 1B, respectively). Salmeterol had a significantly longer duration of action than ipratropium, whereas the onset of response to ipratropium was significantly faster. The area under the curve over 12 h was similar after salmeterol (42 µg twice daily) and ipratropium (36 µg four times daily) for both FEV₁ (Figure 2A) and FVC (Figure 2B). Both salmeterol and ipratropium improved lung function in patients responsive and patients nonresponsive to albuterol. The FEV₁ AUC and FVC AUC responses for the albuterol-responsive group were greater than those in the albuterol-nonresponsive group, but no significant difference (p  0.097) was observed between salmeterol and ipratropium. Responses to both salmeterol and ipratropium were significantly greater than placebo (p < 0.001) for both the albuterol-responsive and albuterol-nonresponsive groups.

View larger version (23K): [in this window] [in a new window]   Figure 1.   Effect of salmeterol on lung function. Spirometric measures were made sequentially on the first day of treatment. Patients inhaled salmeterol at t = 0 and placebo immediately after the 6-h spirometry (triangles), patients inhaled ipratropium at t = 0 and again after the 6-h spirometry (circles), or patients inhaled placebo at both times (squares).

View larger version (31K): [in this window] [in a new window]   Figure 2.   Effect of bronchodilators on lung function (A, FEV2, AUC; B, FVC, AUC) by reversibility stratification. Patients were classified as responsive or nonresponsive after albuterol challenge and before randomization. Lung function was determined spirometrically and is expressed as the 12-h "area under the curve" change from baseline. Patients treated with placebo, open columns; salmeterol, solid columns; ipratropium, shaded columns. *p < 0.05 versus placebo.

At the end of 12 wk of therapy, treatment with salmeterol resulted in significant bronchodilation with no evidence of loss of FEV₁ or FVC response (Figures 3A and 3B). Similarly, no loss of response was observed with ipratropium. In the salmeterol group, the baseline lung function measured in the morning before administration of any bronchodilator was significantly increased after 4 wk of continuous therapy and continued throughout the study. A similar pattern of improvement from baseline in pulmonary function was observed in morning PEF. Salmeterol and ipratropium both resulted in significant increases in PEF compared with placebo. The number of puffs/d of rescue MDI also decreased in both the salmeterol and ipratropium groups, with the greatest decline in albuterol use in the salmeterol group (from 3.6 to 1.8 puffs/d).

View larger version (26K): [in this window] [in a new window]   Figure 3.   Effect of 12 wk of therapy with salmeterol on lung function (A, FEV1; B, FVC). Patients inhaled salmeterol at t = 0, and spirometry was monitored sequentially over 12 h after the first dose (squares) and after 12 wk of continuous treatment (triangles).

Symptoms

At baseline, the distance walked in 6 min was 427, 420, and 417 yd in the placebo, salmeterol, and ipratropium groups, respectively. There were no significant differences between treatment groups (p > 0.2) at baseline. The mean distance walked did not increase by more than 10 yd for any treatment group during the 12-wk treatment period. Pairwise dyspnea comparisons of salmeterol versus placebo demonstrated that salmeterol significantly reduced (p = 0.013) the prewalk dyspnea score at the last assessment (Week 10). Borg scores increased from prewalk values of 1.08, 1.32, and 1.18, to postwalk scores of 2.83, 3.08, and 2.79 for placebo, salmeterol, and ipratropium, respectively. There were no significant differences between treatment groups in postwalk scores at baseline or at the end of the study period.

The BDI scores were 6.01, 5.96, and 6.27, indicating moderate impairment attributable to dyspnea with no significant differences for patients in the three treatment groups (p > 0.24). During the early weeks of treatment, significant improvement (p < 0.005) of dyspnea as measured by the TDI occurred for both the salmeterol and ipratropium groups compared with placebo. However, the statistical significance was lost by Week 6 because of improvement of dyspnea in the placebo group.

Dyspnea was also assessed daily by the patient on a five-point scale and recorded on diary cards (Table 2). Both treatments were significantly better than placebo for the first 2 wk of treatment, but significance was lost in both groups over the course of the study.

Patients rated their daytime chest tightness and cough on a five-point scale, with 0 equal to no symptoms (Table 2). Baseline chest tightness and cough were low in all treatment groups. These symptoms decreased with all three treatments, with no overall significant differences observed.

The mean number of awakenings was low in all groups. The number of awakenings in the placebo group was significantly lower (p = 0.019) than for the salmeterol group during the baseline period (Table 2). Over the 12-wk treatment period, there were no significant differences between the groups.

Exacerbations

The percentages of patients experiencing one or more COPD exacerbations over the 12-wk treatment period were 30.4, 28.8, and 26.8% for the placebo, salmeterol, and ipratropium groups, respectively. Twenty patients (14.8%) in the placebo group experienced their first exacerbation during Week 1 compared with 6 patients (4.6%) in the salmeterol group and 6 patients (4.4%) in the ipratropium group (p  0.005 for actives versus placebo).

Health Status Measures

Quality of life was assessed by the Chronic Respiratory Disease Questionnaire (CRDQ) (Figure 4). Baseline CRDQ scores were similar for each treatment group. At Week 12, the overall scores increased by 6.8, 10.3, and 9.2 for the placebo, salmeterol, and ipratropium groups, respectively, with the difference between the salmeterol and placebo groups approaching significance (p = 0.078). The proportion of patients who achieved a clinically significant change of 10 from baseline was 38, 46, and 41% for placebo, salmeterol, and ipratropium, respectively.

View larger version (17K): [in this window] [in a new window]   Figure 4.   Effect of quality of life. Quality of life was assessed by the CRDQ. Vertical axis: Change in CRDQ score from baseline. Horizontal axis: Time. Patients treated with placebo, open columns; salmeterol, solid columns; ipratropium, shaded columns. *p < 0.05 versus placebo.

Withdrawals

There were 29/135 (21.5%), 22/132 (16.7%), and 25/138 (18.1%) patients withdrawn from the study in the placebo, salmeterol, and ipratropium groups, respectively. There were no significant differences between treatment groups in the proportion of patients who withdrew because of adverse events (9 patients each in the placebo and salmeterol groups and 10 patients in the ipratropium group), failure to return for study visits (3 patients in the placebo group and 4 patients each in the salmeterol and ipratropium groups), or other reasons (6, 8, and 5 patients in the placebo, salmeterol, and ipratropium groups, respectively). "Other" reasons for withdrawal included noncompliance with study medication, not meeting inclusion criteria, and steroid use. Significantly fewer patients withdrew because of lack of efficacy in the salmeterol group compared with placebo, (1 patient versus 11 patients; p = 0.007). There were no differences between ipratropium and placebo in withdrawals due to lack of efficacy (6 patients versus 11 patients; p = 0.38), and no differences between salmeterol and ipratropium in the number of patients withdrawn because of lack of efficacy (p = 0.10).

Adverse Events

A total of 302 patients reported at least one adverse event during the study. There were no clinically significant differences in the incidence of adverse events across the treatment groups with the exception of events in the ear, nose, and throat category (e.g., sore throat and upper respiratory tract infection). The incidence of these events was greater for both salmeterol (n = 58, p = 0.011) and ipratropium (n = 58, p = 0.031) compared with placebo (n = 39). Nine patients in the placebo group, 9 patients in the salmeterol group, and 10 patients in the ipratropium group were withdrawn because of an adverse event. One death, a patient in the placebo group who experienced a ruptured aortic aneurysm, occurred in the study. There were no clinically important changes in clinical chemistry, routine blood counts, vital signs, ECGs, 24-h Holter monitoring, or physical examinations.

There were no significant differences between treatment groups in the total number of ventricular and supraventricular ectopic cardiac beats or in the mean number of events per 1,000 beats. No life-threatening or symptomatic arrhythmias were observed. At baseline, nine (3%), seven (3%), and one (< 1%) patient had asymptomatic nonsustained ventricular tachycardia (NSVT) on 24-h ECG in the placebo, salmeterol, and ipratropium groups, respectively. During the study, three (1%), seven (3%), and seven (2%) patients in the placebo, salmeterol, and ipratropium groups, respectively, had a significant change from baseline in asymptomatic NSVT. One patient in the ipratropium group experienced NSVT after presenting to the emergency room for chest pain.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The current study demonstrates that salmeterol is an effective bronchodilator in patients with COPD. Salmeterol improved both expiratory airflow and vital capacity, and, consistent with its expected pharmacodynamic properties, the benefit of salmeterol persisted for at least 12 h. The bronchodilation achieved with salmeterol administered every 12 h was comparable to that achieved with ipratropium administered every 6 h. No tachyphylaxis to salmeterol was observed over the 12 wk of therapy. Over this interval, acute treatment was associated with improved pulmonary function and with reduced dyspnea and rescue albuterol use compared with placebo. No significant safety concerns were noted during the course of this study. Taken together, these observations suggest that salmeterol is an effective therapeutic agent in COPD.

Patients suffering from COPD are characterized by chronic limitation of expiratory airflow (1, 2, 15). Airflow is always limited and, by definition, does not vary dramatically from day to day. Nevertheless, many patients with COPD benefit from the administration of bronchodilators, currently a mainstay of therapy.

Three classes of bronchodilators are currently used to improve airflow in patients with COPD. Anticholinergic agents are believed to reduce smooth muscle contraction by antagonizing endogenous cholinergic tone (16, 17). -agonist bronchodilators can directly relax airway smooth muscle through the stimulation of ₂ receptors with consequent increases in intracellular cyclic AMP (18, 19). Although somewhat controversial, theophylline may also cause bronchodilatation by relaxing airway smooth muscle through an increase in cyclic AMP due to inhibition of phosphodiesterase (20).

Several previous studies have suggested that salmeterol is beneficial for patients with COPD (9, 21). The current study confirms and extends these findings, providing a direct comparison with ipratropium for both magnitude and duration of action. The study was not designed to demonstrate equivalence of these bronchodilators, but rather to show their comparability. Some differences were observed. Consistent with its known pharmacodynamics in asthma (7, 8), improved lung function in COPD after salmeterol administration persisted for at least 12 h. The effect of ipratropium was waning by 6 h, although retreatment was effective. Ipratropium had a slightly greater peak benefit on FVC than did salmeterol, although as with FEV₁, the effects on FVC were of shorter duration. Using the AUC technique to estimate the time-averaged improvement in lung function, salmeterol administered twice daily was comparable in effect to ipratropium administered four times daily. The dose of ipratropium used in the current study (36 µg) was that indicated in the product's labeling. Although it is common practice to administer higher doses of ipratropium to selected individuals (1, 24, 25), a comparison of higher dosages of ipratropium with salmeterol cannot be made from the results of the current study.

Patients with COPD are frequently classified as "responders" or "nonresponders" to bronchodilators on the basis of lung function after acute challenge in the pulmonary function laboratory. A variety of criteria have been used to categorize individuals into one class or the other. Often this categorization is used to guide therapeutic intervention. The bronchodilator response, however, is not categorical, but continuous (26, 27). Moreover, many individuals will vary in their day-to-day responsiveness (27). The current study enrolled patients who were both responders and nonresponders to the short-acting -agonist albuterol. The individuals who were responders had a larger benefit from salmeterol. Although the nonresponders also responded to salmeterol, the magnitude of the effect was less, consistent with a continuous rather than a categorical response. Interestingly, the response to ipratropium was also less in the nonresponsive group. This suggests that bronchodilator testing of patients with COPD may identify a group that may respond better to either therapy, but will not predict which will be the most effective drug in an individual patient.

Airflow is traditionally used to assess both severity of COPD and response to treatment despite poor correlation with symptoms. The converse, that symptomatic improvement may result when modest improvements occur in airflow, suggests that parameters other than airflow can affect the patient with COPD. Moreover, patients with COPD often adapt their lifestyle to their disease states (28, 29). Thus, many patients will be relatively asymptomatic despite severe physiologic compromise that has resulted in limitation of daily activities. Bronchodilators may improve lung function, but for improved lung function to have a beneficial effect on quality of life, changes in behavior may also be required (1, 2, 30). The beneficial effect of salmeterol on health status (i.e., quality of life) in COPD is supported by a report using a different quality of life measure, the St. George's Questionnaire (21). The current study differs from that of Jones and Bosh (22) in two important respects: (1) a different instrument was used to assess quality of life, and (2) the current trial evaluated changes over a 3-mo period whereas Jones and Bosh assessed patients over 4-mo. The longer treatment duration may have allowed for greater changes. Although significant differences were not observed in the current trial, salmeterol improved quality of life scores, both on average and when assessed as the percentage of patients who achieved an improvement generally regarded as clinically meaningful. The results of the current study, therefore, generally support those of Jones and Bosh (22) and are also consistent with Mahler and coworkers (9).

In summary, salmeterol is an effective bronchodilator for patients with COPD. The 12-h duration of action and lack of tachyphylaxis suggest that salmeterol is ideally suited for use where chronic bronchodilatation is required. Because the majority of patients with COPD require chronic bronchodilatory therapy, salmeterol is effective for the treatment of these patients.