INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Inhaled corticosteroids are the most potent and effective antiinflammatory agents available for the treatment of persistent asthma (1). Early intervention with inhaled corticosteroids may prevent the airway remodeling that results from inflammation, and may prevent loss of pulmonary function (2). However, inhaled corticosteroids are often underutilized because of questions about their safety and because patients do not perceive immediate symptom relief with inhaled corticosteroids as compared with bronchodilator medications (6, 7).

The addition of a long-acting bronchodilator to inhaled corticosteroids for the treatment of asthma has been shown to significantly improve pulmonary function and asthma symptom control compared with increasing the dose of inhaled corticosteroids (8). The National Institutes of Health (NIH) guidelines for the diagnosis and management of asthma now recognize that many patients with persistent asthma need treatment with both a long-acting bronchodilator and an inhaled corticosteroid for optimal management of symptoms (1). However, multiple medication regimens can be confusing to patients, and can lead to poor adherence to treatment schedules. The combination of a long-acting bronchodilator with an inhaled corticosteroid in a single inhalation device would simplify the management of asthma and provide treatment with two effective medications that have complementary mechanisms of action (11).

Twice-daily administration of salmeterol, a long-acting bronchodilator, has been shown to be more effective than albuterol administered four times daily in improving pulmonary function and control of symptoms in asthma (14). Fluticasone propionate (FP) is an inhaled corticosteroid with a better efficacy-to-safety ratio than other inhaled corticosteroids (17). Several studies have demonstrated the efficacy and safety of salmeterol and FP given as individual agents through the Diskus inhaler (Glaxo Wellcome, Inc., Research Triangle Park, NC), an inhalation device that delivers premeasured, sealed doses of drug contained in 60 blisters, providing treatment for 30 d at recommended doses (21). Some patients who have difficulty with the hand and breath coordination required with metered dose inhalers (MDIs) may benefit from the simplicity of the Diskus device, which is easy to use and contains a built-in dose counter that displays the number of doses remaining in the device (25).

The objective of the study reported here was to compare the efficacy and safety of salmeterol 50 µg and FP 250 µg in a combination dry-powder product administered twice daily through the Diskus device with that of FP and salmeterol alone in patients previously treated with low to medium doses (as defined by NIH guidelines for the treatment of asthma) of inhaled corticosteroids.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Male and female patients were eligible for the study if they were at least 12 yr of age and had a medical history of asthma (as defined by the American Thoracic Society [28]) of at least 6 mo duration that required pharmacotherapy over the 6 mo preceding the study. Patients were required to have an FEV₁ between 40% and 85% of the predicted value based on European Community for Coal and Steel standards for persons aged 18 yr and older (readjusted for African- Americans), or based on Polgar standards for persons aged 12 to 17 yr (29). Patients were required to have a  15% increase in FEV₁ at 30 min after two puffs (180 µg) of inhaled albuterol, and to have received inhaled corticosteroids continuously for at least 12 wk before the study. Patients screened for the study had been treated with beclomethasone dipropionate (462 to 672 µg/d), triamcinolone acetonide (1,100 to 1,600 µg/d), flunisolide (1,250 to 2,000 µg/d), or FP (440 µg/d) for at least 4 wk prior to screening.

Female patients had negative pregnancy tests and were surgically sterile, postmenopausal for at least 1 yr, or using an acceptable birth control method for at least 1 mo prior to participation in the study. Exclusion criteria included a history of life-threatening asthma; hypersensitivity reaction to sympathomimetic drugs or corticosteroids; smoking within the year previous to the study or a smoking history of > 10 pack-years; use of oral or injectable corticosteroid therapy within the month preceding the study; use of intranasal corticosteroid therapy (except for FP [Flonase; Glaxo Wellcome Inc.]) during the study; use of daily oral corticosteroid treatment within the 6 mo preceding the study; use of any other prescription or over-the-counter medication that could have affected the course of asthma or interacted with sympathomimetic amines; abnormal chest radiographs; clinically significant abnormal 12-lead electrocardiograms (ECGs); or a history of significant concurrent disease (e.g., glaucoma, diabetes, hypertension). All patients (or guardians) gave informed consent before entry into the study.

Study Design and Interventions

The 12-wk, randomized, double-blind, parallel-group study (Protocol SFCA3003) was approved by the institutional review boards for each of the 42 investigative sites. Eligible patients entered a 2-wk, single-blind, placebo-controlled screening period to evaluate eligibility, assess compliance with therapy, obtain baseline data, and confirm asthma stability. During the screening period, patients continued to take their inhaled corticosteroid in addition to placebo delivered from a Diskus device.

Patients were instructed in the use of a peak flow meter (MiniWright; Clement Clark, London, UK), MDI, and the Diskus inhaler. Patients were instructed to complete daily diary cards, attend scheduled study visits, and report medication compliance. Compliance was measured with the dose counter on the Diskus device. Stable asthma was confirmed via diary cards at the end of the screening period, and was a requirement for continuation in the study. Patients were not eligible if, during the screening period, they had more than three nights with awakenings caused by asthma that required treatment with albuterol during the 7 d immediately preceding randomization. Patients using more than 12 puffs of albuterol daily for more than 3 d were also ineligible. At the end of the screening period, FEV₁ had to be between 40% and 85% of the predicted value, and had to be within ± 15% of the FEV₁ obtained at the beginning of the screening period.

Patients entering the double-blind phase of the study were randomly allocated to receive one of the following treatments, delivered from the Diskus device, twice daily for 12 wk: salmeterol 50 µg and FP 250 µg in combination, salmeterol 50 µg, FP 250 µg, or placebo. Patients were permitted to take albuterol as needed for relief of symptoms.

Assessment of Efficacy and Safety

Primary efficacy endpoints included area under the 12-h serial FEV₁ curve relative to baseline (AUC), morning predose FEV₁, and the probability of patients remaining in the study without being withdrawn because of worsening asthma. Because the salmeterol/FP combination product comprised two different drugs with different therapeutic properties in terms of asthma control, three primary efficacy variables were established. The selected primary efficacy variables were used to evaluate the effects of the combination product that reflected both bronchodilatory properties (serial FEV₁) and the increased asthma stability associated with antiinflammatory therapy (assessed through the probability of patient's remaining in the study and through the morning predose FEV₁). The combination product was expected to be significantly more efficacious than salmeterol alone in terms of asthma stability (probability of patients' remaining in the study and morning predose FEV₁), and similar to or better than salmeterol alone in terms of bronchodilatory effects (AUC serial 12-h FEV₁). The combination product was expected to be the same as or better than FP alone in terms of asthma stability (probability of patients' remaining in the study and morning predose FEV₁), and significantly more efficacious than FP alone in terms of bronchodilatory effects (AUC for serial 12-h FEV₁).

Secondary efficacy measures included peak expiratory flow (PEF), patient-rated daily diary card symptom scores, albuterol use, and nighttime awakenings caused by asthma and requiring albuterol. At each clinic visit, pulmonary function tests were performed in triplicate before the morning dose of medication (only the best effort was recorded), using spirometric equipment that exceeded the minimal performance recommendations (31). Throughout the study, patients measured PEF in triplicate (only the best effort was recorded) twice daily; rated daily symptoms (wheezing, shortness of breath, and cough) on a six-point scale ranging from 0 = no symptoms to 5 = symptoms that severely interfered with daily activities; recorded the number of nighttime awakenings caused by asthma symptoms and requiring albuterol; and recorded the frequency of supplemental albuterol use.

Patients returned to the clinic (between 6:00 A.M. and 9:00 A.M.) on a weekly basis for the first 4 wk of the study and then every 2 wk for 2 mo, for a total of 12 wk of double-blind treatment. At each clinic visit, diary card information was assessed and pulmonary function tests were performed to determine whether patients met predetermined asthma stability criteria. Patients were withdrawn from the study because of worsening asthma if they met any of the following criteria: a clinical exacerbation requiring emergency treatment, hospitalization, or use of asthma medication not allowed by the study protocol; a decrease in FEV₁ of more than 20% from the predose FEV₁ at the randomization visit; more than a 20% decrease from the mean morning baseline PEF on more than 3 of 7 d immediately preceding a visit; 12 or more albuterol puffs per day on more than 2 of 7 d immediately preceding a visit; or more than two nights with awakenings caused by asthma symptoms that required albuterol during the 7 d immediately preceding a clinic visit.

Safety was assessed by examining 12-lead ECGs, morning plasma cortisol concentrations, plasma cortisol response to a short infusion of synthetic adrenocorticotropic hormone (ACTH), routine laboratory tests, physical examinations, and 24-h Holter monitoring at selected sites. Adverse events were recorded at each clinic visit, and the investigator rated the relationship of the event to the study drug. Standard clinical laboratory tests and physical examinations were conducted at screening and at Week 12.

Twelve-lead ECGs were obtained at screening and both before and 90 min after dosing on Day 1, Week 1, and Week 12, during the 12-h serial pulmonary function tests. A clinically significantly abnormal ECG was defined as a 12-lead tracing consistent with ischemic changes, ventricular hypertrophy, intraventricular conduction abnormalities (e.g., bundle-branch block, Wolff-Parkinson-White syndrome), or clinically significant arrhythmias (e.g., atrial fibrillation, ventricular tachycardia). Twenty-four-hour Holter monitors were connected and tapes started in the clinic within 2 d after screening and within 2 d before Week 12 in a subset of the patients.

Morning plasma cortisol concentrations and response to synthetic corticotropin stimulation tests were evaluated at screening and at the study endpoint. Blood samples for the determination of plasma cortisol concentrations were collected before the morning dose of drug and 30 to 60 min after administration of synthetic corticotropin (250 µg) on synthetic corticotropin test days. A normal morning plasma cortisol concentration was defined as  5 µg/dl. A normal response to synthetic corticotropin stimulation was defined as a plasma cortisol concentration of at least 18 µg/dl and an increase from baseline of at least 7 µg/dl.

Statistical Analyses

The primary measures of efficacy were serial FEV₁ and the probability of patients' remaining in the study over time. Previous studies indicate that a reasonable assumption for the standard deviation of FEV₁ is 0.55 L. The sample size of 80 subjects per treatment was chosen to provide 80% power to detect a difference of 0.25 L for any pairwise treatment comparison, using a t test and a significance level of p  0.05.

Regarding the probability of patients' remaining in the study over time, it was estimated that the proposed sample size of 80 subjects per treatment group would provide 85% power to detect a difference in dropout rates of 20%, assuming that 10% of subjects taking the combination product and 30% of subjects taking salmeterol would be withdrawn for lack of efficacy before the end of the study.

All statistical tests were two-sided and based on a significance level of p  0.05. Pairwise treatment comparisons were viewed solely as descriptive in cases in which the overall treatment comparison was not statistically significant.

Descriptive statistics were calculated for demographic and other baseline variables. Overall treatment group differences were evaluated with analysis-of-variance (ANOVA) F tests and Cochran-Mantel-Haenszel tests for continuous and categorical variables, respectively, with each variable controlled for investigator cluster. (Groups or clusters of investigators were used for statistical analyses, rather than individual investigators, because of the large number of investigators participating in the study. Clusters were defined before unblinding of the clinical trial, and were defined to be of similar size. Investigators who screened fewer than 20 patients were grouped into clusters based on the geographical proximity of sites. Investigators who screened 20 or more patients were defined as stand-alone clusters.)

The population used for all efficacy analyses was the intent-to-treat population minus 13 patients at one site who were excluded because their data did not meet study standards. Serial FEV₁ data were analyzed by clinic visit. Morning predose FEV₁ and all diary measures were analyzed at each clinic visit and at the study endpoint. Endpoint analyses included data from the final visit during treatment for patients completing the study, and data from the last visit for patients who were withdrawn before the end of the study. The probability of patients' remaining in the study at each week without being withdrawn because of worsening asthma was estimated by using the Kaplan- Meier product limit method, based on the number of efficacy failures in each treatment group. Log-rank tests were used to compare the homogeneity of the resulting survival estimates.

ANOVA F tests, with control for investigator cluster, were used to assess treatment differences in morning predose FEV₁, serial FEV₁ measures, and morning and evening peak flow. Serial FEV₁ measures were also analyzed with a repeated-measures ANOVA. Patient- recorded diary data were analyzed with van Elteren's tests, stratified by investigator cluster.

Safety analyses were based on the intent-to-treat population. Assessment of the incidence of adverse events was done with Fisher's exact test; cardiac rates measured by Holter monitoring were analyzed with ANOVA F tests, with control for investigator cluster. Abnormalities in cardiac rhythm were assessed with van Elteren's tests, stratified by investigator cluster.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Of 484 patients screened, 349 were randomly assigned to treatment. The major reason for screening failure was lack of reproducible lung function (i.e., FEV₁ not within 15% of the FEV₁ measurement obtained at the beginning of screening). Randomization resulted in comparable treatment groups at baseline with respect to patient demographics and pulmonary function (Table 1). Mean treatment compliance rates ranged from 91% to 95% across treatment groups. From 8% to 14% of patients in each group had compliance rates < 80%. No patient was withdrawn from the study because of poor compliance with study medication. The disposition of patients is shown in Table 1. Overall, comparable results were obtained with or without data from 13 patients at one site who were excluded because of significant deviations from standards of good clinical practice.

Efficacy

Spirometric data. Mean change from baseline at endpoint in morning predose FEV₁ is shown in Table 2. Improvements in FEV₁ at endpoint were significantly (p  0.028) greater in patients treated with the combination product than in those given placebo, who experienced a decline in FEV₁. Patients treated with the combination product experienced a 23% improvement from baseline in FEV₁ at endpoint, compared with 4% and 13% improvements from baseline in patients treated respectively with salmeterol or FP, alone (p  0.001); the mean percent decline from baseline in FEV₁ was 5% with placebo (p < 0.001 versus combination product).

Twelve-hour serial FEV₁ results on treatment Day 1 and at the end of Week 1 and Week 12 revealed greater sustained mean improvements from baseline (weighted averages) over 12 h with the combination product than with salmeterol, FP, or placebo (p  0.025; Figure 1). A majority of patients treated with the combination product achieved a  15% increase in FEV₁ within 30 min on Day 1 of treatment that was maintained during the 12-h dosing interval. The improvement in FEV₁ during the 12-h dosing interval increased progressively during the trial. At 12 wk, the improvement in FEV₁ with the combination product ranged from 0.54 L to 0.69 L during the 12-h dosing interval (ranging from high to low at each of the measurement time points over 12 h), representing a 26% to 33% improvement from pretreatment baseline FEV₁ values. The mean AUCs at Day 1, Week 1, and Week 12 are given in Table 2. The combination product significantly increased the mean AUC over that with placebo, salmeterol, and FP at Day 1 (p  0.025), Week 1 (p < 0.001), and Week 12 (p  0.003).

View larger version (21K): [in this window] [in a new window]   Figure 1.   Mean change from baseline in serial FEV1 over a 12-h period at Day 1, Week 1, and Week 12.

At endpoint, the mean change from baseline in the predose FEV₁ % predicted was significantly (p < 0.001) greater with the combination product (14%) than with placebo (4%), salmeterol (1%), or FP (7%). The mean FEV₁% predicted values at endpoint were 83%, 64%, 69%, and 74% in the combination-product, placebo, salmeterol, and FP treatment groups, respectively.

Probability of patients' remaining in the study without being withdrawn because of worsening asthma. Only 29% of patients in the placebo group remained in the study for the entire 12-wk study period, compared with 84%, 48%, and 73% of patients treated with the combination product, salmeterol, and FP, respectively. The percentage of patients withdrawn from the study because of worsening asthma was lowest with the combination product (4%), highest with placebo (62%), and next highest with salmeterol (38%) and FP (22%), respectively (Table 1). Survival analysis revealed that patients in the combination-product group had a significantly greater probability of remaining in the study over time without being withdrawn because of worsening asthma than did those in the placebo, salmeterol, or FP treatment groups (p  0.002; Figure 2). Similarly, patients treated with either salmeterol or FP alone had a significantly greater probability of remaining in the study over time without being withdrawn because of worsening asthma than did those given placebo (p < 0.001).

View larger version (19K): [in this window] [in a new window]   Figure 2.   Probability of remaining in the study (Kaplan-Meier survival curve).

Patient-recorded data. Values at baseline and for the mean change from baseline at endpoint for PEF, asthma symptom scores, percentage of days with no asthma symptoms, percentage of nights with no awakenings, and supplemental albuterol use are listed in Table 3. Significant improvements in PEF were noted during the trial, after treatment with the combination product as compared with placebo, salmeterol, and FP. After treatment with the combination product, improvements at endpoint in morning PEF (Figure 3) were significantly (p < 0.001) greater (53.5 L/min; 16% improvement from baseline) than the changes noted after administration of placebo (14.1 L/min; 2% decline from baseline), salmeterol (11.6 L/min; 3% decline from baseline), and FP (15.2 L/min; 4% improvement from baseline). Similar changes were noted in percent predicted morning and evening PEF. Patients treated with the combination product experienced significant improvements in PEF within the first day (data not shown) of treatment as compared with the changes that occurred placebo and FP, with further improvements noted during the course of the trial (Figure 3).

View larger version (26K): [in this window] [in a new window]   Figure 3.   Mean change from baseline in PEF.

Reductions in asthma symptom scores (with decreased scores indicating improvement) were significantly (p  0.011) greater after treatment with the combination product (0.8; 57% improvement from baseline) than after treatment with FP (0.4; 25% improvement from baseline); mean asthma symptom scores increased with placebo (0.4; 25% deterioration from baseline) and salmeterol (0.1; 6% deterioration from baseline). The mean percentage of days with no asthma symptoms was significantly (p  0.004) greater after treatment with the combination product than after placebo or after salmeterol, or FP treatment alone. The mean percentage of nights with no awakenings increased significantly (p  0.016) over baseline with the combination product (7.2%) as compared with the increase observed with FP (2.8%) and with a decline in the placebo and salmeterol treatment groups. Patients treated with the combination product significantly reduced their supplemental albuterol use (based on the number of puffs per 24 h), by 66% (2.3 puffs/d), compared with a reduction of 28% (0.9 puffs/d), in the FP treatment group; supplemental albuterol use increased by 24% (0.9 puffs/d) in the placebo group and remained the same in the salmeterol treatment group (p  0.002, combination product versus salmeterol, FP, and placebo).

Safety

No clinically relevant differences were noted among treatment groups with respect to clinically significant abnormalities in physical examinations or clinical laboratory tests.

Adverse events. Each of the three treatment options was well tolerated over the 12-wk study period, with no serious drug-related adverse events. Two patients treated with salmeterol withdrew from the study because of adverse events; however, these adverse events were considered by the investigator to be unrelated to study drug (bilateral subcapsular cataracts and postsurgical infection). The most commonly occurring drug-related adverse events ( 2%) were candidiasis (unspecified sites and oropharyngeal) and cough. Unspecified candidiasis occurred in two (2%) patients in the combination-product group and three (4%) patients in the FP group. Oropharyngeal candidiasis occurred in three (4%) patients in the combination-product group and two (2%) patients in the FP group. Cough occurred in two (2%) patients in the combination-product group and one (1%) patient in the salmeterol group.

Holter monitor, 12-lead ECG. Continuous 24-h ambulatory electrocardiography revealed no significant differences among treatment groups in mean HR or in the occurrence of ventricular or supraventricular ectopy. Results of 12-lead electrocardiography did not indicate any significant unfavorable changes from baseline at the end of 12 wk of treatment.

HPA axis assessments. No clinically significant differences were noted among treatment groups with respect to morning plasma cortisol abnormalities or response to synthetic corticotropin stimulation. At baseline, one patient each in the placebo (3%), salmeterol (3%), and FP (3%) treatment groups had morning plasma cortisol concentrations < 5 µg/dl. At endpoint, the number of patients with morning plasma cortisol concentrations < 5 µg/dl was similar in the placebo (two; 6%), combination-product (one; 3%), salmeterol (none), and FP (two; 6%) treatment groups.

At baseline, one patient each in the placebo (3%) and FP (3%) treatment groups had poststimulation cortisol levels < 18 µg/dl. The numbers of patients with poststimulation increases in cortisol levels of < 7 µg/dl after 12 wk of treatment were three (8%), four (11%), five (15%), and three (9%) in the placebo, combination-product, salmeterol, and FP treatment groups, respectively. The numbers of patients with poststimulation cortisol levels < 18 µg/dl after 12 wk of treatment were similar in the placebo (two; 6%), combination-product (one; 3%), salmeterol (none), and FP (two; 6%) treatment groups.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients with asthma suboptimally controlled with medium doses of inhaled corticosteroids experienced statistically and clinically significant improvements in FEV₁, PEF, and asthma symptoms, and reductions in nighttime awakenings and supplemental albuterol use after twice-daily treatment with the combination product compared with treatment with the individual agents. Treatment with the combination product significantly improved FEV₁ within 30 min after administration of the first dose, and this was maintained during the 12-h dosing interval. During the 12 wk of treatment, a further improvement was observed in FEV₁, most likely from the influence of FP in the combination product.

In addition to the improvements in FEV₁ noted within 30 min after the first dose, treatment with the combination product was associated with significant improvements in PEF and other diary card measures within the first day. The expected onset of action of inhaled long-acting bronchodilators is generally within 1 h, whereas the expected onset of action of most inhaled corticosteroids is 1 wk or more. However, significant improvements in PEF and asthma symptoms, and reductions in supplemental albuterol, have been observed as early as 1 d after treatment with FP in patients not previously treated with inhaled corticosteroids (32). The combination of a long-acting bronchodilator and an inhaled corticosteroid with a relatively quick onset of action could provide patients with significant clinical benefits that may occur relatively soon after its administration. The simplicity of this combination regimen (i.e., both drugs in one delivery device), coupled with the potential for the patient's perception of rapid onset of action, may increase patient adherence to asthma treatment with this regimen.

Because patients randomized to receive placebo and salmeterol in our study were required to discontinue their baseline inhaled corticosteroid therapy, predefined asthma stability criteria were used to withdraw patients whose asthma control deteriorated during the study. Investigators also had the discretion of withdrawing patients because of clinical exacerbations. The asthma stability criteria used in the study are consistent with those in recent trials that utilized these types of criteria to characterize mild exacerbations of asthma (33). Only 4% of patients withdrew because of worsening asthma with the combination product, compared with 62%, 38%, and 22% of patients given placebo, salmeterol, and FP, respectively. Of the patients who were withdrawn for worsening asthma, only 2% of these treated with the combination product were withdrawn because of clinical exacerbations of their asthma, compared with 12%, 7%, and 17% of patients given salmeterol, FP, and placebo, respectively. These differences in withdrawal rates resulted in a significantly greater probability of patients treated with the combination product remaining in the study without being withdrawn because of worsening asthma as compared with patients in the other treatment groups.

The higher withdrawal rate because of worsening asthma with placebo than with salmeterol or FP may explain the small differences in PEF and serial FEV₁ observed among patients given salmeterol, FP, and placebo toward the end of the study (i.e., the patients remaining in the placebo group were those with less severe asthma). The combination product resulted in the lowest rate of withdrawal because of worsening asthma, indicating that the combined use of inhaled long-acting ₂-agonists and inhaled corticosteroids minimizes the risk of asthma exacerbations as compared with the use of these agents individually (33).

Although treatment with salmeterol was more effective than placebo for some clinical measures, many patients in the salmeterol treatment group did not improve, and in some cases experienced worsening of their asthma control as a result of discontinuing their inhaled corticosteroids at randomization. These findings support current treatment guidelines that advocate concurrent use of salmeterol with inhaled corticosteroids, rather than monotherapy with salmeterol as a substitute for inhaled corticosteroids (1).

Although inhaled corticosteroids are regarded as the most effective antiinflammatory medications for treating persistent asthma, there remain concerns about their safety at higher doses. Several investigators have reported significant improvements in FEV₁ and PEF, as well as reductions in asthma symptoms, number of exacerbations, and supplemental albuterol use after the addition of salmeterol or formoterol in patients previously treated with inhaled corticosteroids as opposed to increasing the dose of inhaled corticosteroids (8, 9, 10, 33). On the basis of these results, the approach of adding an inhaled long-acting ₂-agonist to inhaled corticosteroids as an alternative to higher doses of inhaled corticosteroids is being advocated in treatment guidelines for asthma. Despite the consistent improvements in asthma control demonstrated with this approach, there remain concerns about undertreatment of inflammation. However, there is increasing evidence that long-acting ₂-agonists and inhaled corticosteroids have complementary mechanisms of action in the treatment of asthma (11- 13). Regular use of inhaled corticosteroids has been shown to modulate ₂-agonist receptor function (11). In addition, Eickelberg and colleagues have recently shown that ₂-agonists are potent activators of the glucocorticoid receptor in vitro (13). Although the exact mechanism of interaction of these two classes of drugs in the treatment of asthma remains unknown, the magnitude and consistency of clinical benefits observed when inhaled long-acting ₂-agonists and inhaled corticosteroids are used together indicates that many patients can benefit from using both classes of drugs.

In summary, combination treatment with salmeterol 50 µg and FP 250 µg administered twice daily through a Diskus device significantly improves pulmonary function and asthma symptom control, and reduces the complexity of asthma therapy without additional safety risk in patients previously treated with medium doses of inhaled corticosteroids. Administration of the combination product through the Diskus device, which patients have found easy to use, and infrequent use of rescue albuterol as a result of combination therapy, may further simplify the asthma treatment regimen. In addition, the availability of multiple strengths of FP in the combination product will allow flexibility in titrating the dose of inhaled corticosteroid according to asthma severity (34, 35). These two effective, complementary treatment options combined in one delivery device provide a superior level of asthma control and improvement in pulmonary function to that achieved with the individual agents alone, and represent a new treatment option for optimizing asthma control.