Oral Montelukast, Inhaled Beclomethasone, and Placebo for Chronic Asthma

For patients with persistent asthma, the National Asthma Education and Prevention Program recommends the regular use of controller medications to provide long-term control, together with as-needed use of rescue medications (1). Inhaled corticosteroids are currently the most commonly used controller medications (1). Although inhalation into the airway has advantages in terms of both safety and efficacy, problems with compliance with inhaled agents may limit their real-world effectiveness, especially in elderly and pediatric patients (2, 3). Cysteinyl leukotrienes are important pro-inflammatory mediators of asthma (4). Recent clinical studies of leukotriene receptor antagonists (5-7) and a 5-lipoxygenase inhibitor (8) have shown these agents to have clinical benefit in patients with chronic asthma, and treatment guidelines now consider these agents alternative, first-line controller medications (1). Recent studies showed that montelukast, a potent and specific leukotriene receptor antagonist (9), had efficacy in adult (over a 12-week period) and pediatric (over an 8-week period) patient with chronic asthma and had a tolerability profile similar to that of placebo (10-12). To date, no comparisons of leukotriene receptor antagonists and inhaled corticosteroids have been published. Our placebo-controlled, parallel-group study, which was primarily designed to compare the effect of montelukast with that of placebo in a phase III clinical development trial, also compared montelukast and inhaled beclomethasone in patients with chronic asthma who require the daily use of a controller medication. Methods Patients Healthy, nonsmoking, male and female patients 15 years of age and older were eligible to participate if they had had asthma for at least 1 year before the initial study evaluation. Each patient also had to have 1) an FEV1 between 50% and 85% of predicted value, 2) an increase of at least 15% in absolute FEV1 after the use of inhaled -agonist on at least two of three visits during period 1 [see below], 3) a daytime asthma symptom score of at least 64 [of a possible 336], and 4) average daily use of at least one puff of as-needed, short-acting, inhaled -agonist (salbutamol). Patients were excluded from the study if they had used inhaled and oral corticosteroids, cromolyn, or nedocromil within 4 weeks before the initial evaluation; had used long-acting -agonists, antimuscarinics, and newly instituted theophylline within 2 weeks before the initial evaluation; or had used long-acting antihistamines (for example, they could not have used astemizole within 3 months of the initial evaluation, and they could not have used terfenadine or loratadine within 2 weeks of the initial evaluation). Intermittent use of short-acting antihistamines was allowed, and immunotherapy was permitted if it had been started at least 6 months before the initial study evaluation and if the monthly dose remained constant. Protocol Our randomized, double-dummy, placebo-controlled, parallel-group trial had a 2-week, single-blind placebo run-in period (period 1); a 12-week, double-blind treatment period (period 2); and a 3-week, double-blind placebo washout period [period 3]. Eligible patients were randomly assigned to one of three treatment regimens: 1) montelukast, 10 mg once daily in the evening; 2) inhaled beclomethasone, 200 g twice daily; or 3) placebo. The ratio of montelukast recipients to beclomethasone recipients to placebo recipients was 3:2:2, with a blocking factor of 7, according to a single, computer-generated allocation schedule. Patients, investigators, and coordinating center staff were blinded to the treatment schedule until all corrections to the database were completed. During period 3, a subset of patients originally assigned to receive active treatment (approximately 40, as determined by the allocation schedule) was switched to placebo in a blinded manner. The other patients originally assigned to receive active treatment continued to receive that treatment. This allowed us to assess withdrawal from therapy. The study was done at 36 clinical centers in 19 countries in Europe, Africa, Australia, Central America, and South America. The study protocol and informed consent were approved by local ethics review committees. All participants gave written informed consent before participation; consent of parents or guardians was obtained for patients younger than 18 years of age. Patients were recruited from the patient pool of each clinical study center and through local newspaper advertisements. Medication The study medication consisted of 10-mg montelukast film-coated tablets, placebo tablets that were identical in appearance to the montelukast tablets, beclomethasone (100 g/puff) in inhalers (Allen & Hanburys, Stockley Park, United Kingdom), and placebo in inhalers identical to those used for beclomethasone. Patients were instructed to take one tablet at bedtime and to take two puffs from the inhaler (using the AeroChamber spacer device [Clement Clark, Columbus, Ohio]) at bedtime and in the morning. Short-acting, inhaled -agonist (salbutamol, 100 g/puff) (Allen & Hanburys) was to be used as needed. Patients with worsening episodes of asthma that required additional therapy were treated with oral corticosteroids according to a standard protocol. Patients who had more than two worsening episodes of asthma requiring corticosteroid therapy were dropped from the study. Measurements Centralized spirometry training was done according to standard American Thoracic Society criteria before the start of the study. Spirometry was done at each visit after inhaled -agonist therapy had been withheld for at least 6 hours, theophylline therapy had been withheld for at least 24 hours, and antihistamine therapy had been withheld for at least 48 hours. At least three spirometry maneuvers were done, and the largest FEV1 was reported. Spirometry data were transmitted electronically to a central database and monitored continuously for quality, and feedback was given to the study centers (13). If quality was not maintained, sites were visited by coordinating center personnel. Answers to four questions on daytime asthma symptoms and one question on nocturnal awakenings were collected on a daily diary card, as were morning and evening peak expiratory flow rate and daily use of as-needed salbutamol. With respect to daytime asthma symptoms, patients used a 7-point scale (on which 0 means best and 6 means worst) to rate the severity of symptoms, their frequency, the degree to which they were bothersome, and their impact on daily activities. These ratings were combined into a mean daily score. Nocturnal awakenings were evaluated by the patient's response (on a four-point scale) to a single question (14). The validation of these questions has been published elsewhere (14). Peak expiratory flow was measured in the morning and in the evening immediately before study medication was taken. The best of three measurements was recorded. Prespecified end points other than peak expiratory flow included peripheral blood eosinophil counts, global evaluations by physicians and patients (on a 7-point scale, on which 6 means very much worse and 0 means very much better) (7), and asthma-specific quality of life (15). Asthma outcome end points, including asthma attacks (defined as worsening asthma requiring oral corticosteroid treatment or an unscheduled visit to a physician, emergency department, or hospital), percentage of days with asthma exacerbations, and percentage of asthma-control days (as defined elsewhere [11, 16]), were also evaluated. The diary card, the Asthma Quality-of-Life Questionnaire (15), and the global evaluation questions were translated from English into appropriate local languages and were validated for linguistic equivalence and cultural differences (17). Patient compliance with study therapy was determined by weighing inhalers and counting tablets. The weight of a full canister was determined by averaging the weights of at least 50 canisters of beclomethasone and placebo. A standard puff weight was determined by repeatedly actuating and weighing 10 canisters. Laboratory safety tests, including hematologic tests, serum biochemistry analysis, and urinalysis, were done at prespecified visits, and the results were analyzed by a central laboratory. The central laboratory also determined eosinophil counts with an automated cell counter. Statistical Analysis Our primary objective was to compare montelukast with placebo with respect to two prespecified primary end points: FEV1 and daytime symptom score. Other objectives were to compare beclomethasone with placebo and to compare montelukast with beclomethasone in a stepwise manner in accordance with the Dunnet-Tamhane approach, with the comparison of montelukast and beclomethasone prespecified as an estimation of the difference using 95% CIs (18). No multiplicity adjustments were made for each of the secondary end points, which were prespecified as supportive. The analysis for each efficacy end point included all patients who had a measurement taken at baseline and at least one measurement taken after randomization. Treatment responses were calculated as the average change or percentage change from baseline to period 2; the baseline value was the average value for period 1. With SAS software (SAS Institute, Inc., Cary, North Carolina), we used an analysis of variance (ANOVA) model to estimate treatment group means and between-group differences and to construct 95% CIs by using the least-square means and the differences in least-square means. The model contained factors for treatment, study center, and stratum (theophylline use). The consistency of the treatment effect across study centers, strata, and subgroups (which were formed according to sex, age, ethnicity, history of allergic rhinitis, and history of exercise-induced asthma) was assessed by interaction testing done using the ANOVA model with an intera