Randomised, placebo controlled trial of effect of a leukotriene receptor antagonist, montelukast, on tapering inhaled corticosteroids in asthmatic patients

Abstract

Objective

To determine the ability of montelukast, a leukotriene receptor antagonist, to allow tapering of inhaled corticosteroids in clinically stable asthmatic patients.

Design

Double blind, randomised, placebo controlled, parallel group study. After a single blind placebo run in period, during which (at most) two inhaled corticosteroids dose decreases occurred, qualifying, clinically stable patients were allocated randomly to receive montelukast (10 mg tablet) or matching placebo once daily at bedtime for up to 12 weeks.

Setting

23 academic asthma centres in United States, Canada, and Europe.

Participants

226 clinically stable patients with chronic asthma receiving high doses of inhaled corticosteroids (113 randomised to montelukast and 113 to placebo).

Interventions

Every 2 weeks, the inhaled corticosteroids dose was tapered, maintained, or increased (rescue) based on a standardised clinical score.

Main outcome measures

Last tolerated dose of inhaled corticosteroids.

Results

Compared with placebo, montelukast allowed significant (P=0.046) reduction in the inhaled corticosteroid dose (montelukast 47% v placebo 30%; least square mean difference 17.6%, 95% confidence interval 0.3 to 34.8). Fewer patients on montelukast (18 (16%) v 34 (30%) placebo, P=0.01) required discontinuation because of failed rescue.

Conclusions

Montelukast reduces the need for inhaled corticosteroids among patients requiring moderate to high doses of corticosteroid to maintain asthma control.

Key messages

• Leukotriene receptor antagonists have complementary action to inhaled corticosteroids in asthma

• Many patients receive higher doses of inhaled corticosteroids than clinically required

• In this placebo controlled trial, montelukast allowed significant reduction of inhaled corticosteroid doses

• Fewer patients receiving montelukast had failed rescue than patients receiving placebo

Introduction

People with persistent asthma often require daily treatment to control symptoms, usually with inhaled corticosteroids.¹ Recent publications have shown that in patients whose asthma is not completely controlled by inhaled corticosteroids, adding a second drug rather than increasing the dose of corticosteroid achieves better control.²,³ Leukotriene receptor antagonists have complementary effects to corticosteroids on inflammatory events in asthma,⁴,⁵ providing a basis for these additive clinical effects. Montelukast is a potent, specific, cysteinyl leukotriene receptor antagonist which, given daily, improves asthma control in adults⁶,⁷ and children,⁸ protects against exercise induced bronchoconstriction,⁹ and decreases sputum eosinophil counts.¹⁰ Subgroup analyses of previously reported clinical trials have suggested that montelukast has an additive effect with inhaled corticosteroids.⁶,¹¹,¹²

We performed a placebo controlled, double blind, randomised trial to investigate whether daily treatment with montelukast (10 mg) allowed reduction of inhaled corticosteroids while maintaining asthma control.

Participants and methods

Participants

We recruited 226 non-smoking adults with a clinical history of asthma for at least 1 year. Their forced expiratory volume in 1 second had to be ⩾70% of the predicted value (after withholding β agonist for ⩾6 hours) at the prestudy visit and to improve by ⩾15% (absolute value) after inhaled β agonist. Patients were excluded if they had had emergency treatment for an asthma exacerbation within 1 month, been in hospital for asthma within 3 months, or had an unresolved upper respiratory tract infection within 3 weeks of the prestudy visit. Patients must have been treated with a stable, twice daily dose of inhaled corticosteroids for at least 3 weeks before the prestudy visit.

The protocol was approved by the ethics review committees or institutional review boards of all participating centres. Written informed consent was obtained from all participants.

Study design

The study was conducted at 23 academic medical centres or large group practices in the United States, Canada, and Europe. We used a randomised, double blind, placebo controlled, two period, parallel group design to investigate the effect of montelukast (10 mg tablet once daily at bedtime, irrespective of food) on the ability to taper the dose of inhaled corticosteroids over 12 weeks. Patients were recruited over 3 months. A prestudy visit preceded a single blind, placebo run in period, which lasted at most 7 weeks. The purpose of the run in was to achieve or approach the minimum inhaled corticosteroid dose necessary to maintain clinical stability. After, at most, two dose reductions, clinically stable patients were eligible for randomisation if they demonstrated during the 7-10 days before or at the allocation visit a forced expiratory volume in 1 second ⩾90% of the run in baseline value, less than pre-established levels of asthma symptoms and β agonist use, peak flow at least 65% of the maximum, and required a prespecified minimum dose of inhaled corticosteroids. Allocation was according to a computer generated allocation schedule with a blocking factor of 4. Three blocks were initially sent to each centre. Patients remained on their randomised treatment (10 mg montelukast or placebo) throughout the study. Follow up was stopped at the end of 12 weeks’ treatment.

Methods

Participants returned to the clinic every 2 weeks, and their inhaled corticosteroids dose was adjusted (decreased, increased, or maintained) based on a previously validated composite clinical score (box).¹³ The dose of inhaled corticosteroids was tapered or increased by about 25% (according to pre-established criteria). Unstable patients who failed to regain clinical stability after an increase in dose of corticosteroids were withdrawn from the trial and classified as “failed rescue.” Patients were also withdrawn from the study if they interrupted their study drug for more than 5 consecutive days or had worsening asthma requiring treatment with oral, intravenous, or intramuscular corticosteroids.

Criteria for tapering dose of inhaled corticosteroids: composite clinical score

Before β agonist FEV1 ⩾90% of value at allocation
Daytime symptom score ⩽120% of preallocation* baseline	Yes=1 point	No=0 points
β Agonist use ⩽135% of preallocation*	Yes=1 point	No=0 points
baseline	Yes=1 point	No=0 points

 Score

3 Points: patient stable=taper inhaled corticosteroid

2 Points: patient stable=maintain dose of inhaled corticosteroid

0 or 1 point: patient unstable=increase dose of inhaled corticosteroid

*Established 7-10 days before allocation visit to clinic

Diary cards—As part of the taper criteria, patients recorded daytime symptoms,¹⁴ “as needed” use of β agonists, and morning and afternoon peak expiratory flow on a daily diary card.

Spirometry—Standard equipment and quality control procedures were used across all study sites. Participants performed at least three forced vital capacity manoeuvres achieving American Thoracic Society standards.¹⁵ The best forced expiratory volume was recorded from each set of measurements.

Statistical analysis

The primary analysis was by intention to treat. All patients with postallocation data were included in the analysis. The last tolerated dose of inhaled corticosteroids was the primary end point and was defined as the last dose during the treatment period at which the subject had a composite clinical score of 2 or 3. An endpoint committee, blind to subject allocation, determined all last tolerated doses for all allocated patients based on pre-established criteria. This end point was compared between treatment groups as the percentage change from preallocation baseline (to adequately summarise changes in corticosteroids of different potencies) with an analysis of variance model. This model was used to construct 95% confidence intervals for the least square means and differences of means. All statistical tests were two tailed, and a P value ⩾0.05 was considered significant.

The number (percentage) of patients with failed rescue and tapering were prespecified as a secondary end point and were compared between treatment groups by the Cochran-Mantel-Haenszel test. The time to failed rescue was compared with a log rank test.

The study was designed with a sample size of 100 patients in each group to have 90% power to detect (α=0.050, two tailed test) a mean difference between the two treatment groups of about 30 percentage points in the percentage change from baseline in dose of inhaled corticosteroid.¹³

Results

Randomisation and withdrawals

Of the 226 patients (113 randomised to placebo, 113 to montelukast), one could not be included in the analysis (montelukast) because of insufficient postallocation data. A further 48 patients did not complete 12 weeks’ treatment: 31 in the placebo group (16 for “failed rescue,” nine for clinical adverse experiences, three for protocol deviations, and three withdrew consent) and 17 in the montelukast group (seven for “failed rescue,” four for clinical adverse experiences, four for protocol deviations, one withdrew consent, and one lost to follow up). The two treatment groups did not differ in baseline values (table 1).

Table 1.

Characteristics of patients before randomisation. Values are mean (SD) unless stated otherwise

Variable	Placebo	Montelukast
Sample size*	113	113
Mean (range) age (years)	41 (16-68)	40 (17-70)
No of men	61	47
Duration of asthma (years)	19 (14.0)	18 (13.3)
FEV1 (% predicted)	82.3 (12.9)	84.8 (11.1)
Daytime symptom score†	1.38 (0.8)	1.29 (0.7)
β Agonist use (puffs/day)	2.8 (2.6)	2.4 (2.2)

^*Includes all patients randomly allocated in period two. One patient in the montelukast treatment group did not have any postallocation data available. ^†Daily average of four questions, 0-6 point scale. FEV₁= forced expiratory volume in 1 second. 

The mean inhaled corticosteroid dose for the 226 allocated patients was reduced 37.3% before randomisation (tables 2 and 3). The preallocation doses of inhaled corticosteroids, daily symptoms, β agonist use, and forced expiratory volumes were comparable between treatment groups (tables 1 and 3). Compared with placebo, montelukast significantly (P=0.046) reduced the last tolerated dose of inhaled corticosteroids. Mean percentage changes from preallocation baseline were 47% and 30% for the montelukast and placebo groups, respectively (least square mean difference 17.6%, 95% confidence interval 0.3 to 34.8; table 3). Tapering was consistent across all inhaled corticosteroids.

Table 2.

Mean (range) doses^* of inhaled corticosteroids before study and before allocation

	Prestudy (μg/day)	Preallocation (μg/day)
Fluticasone:
Placebo	1000.0 (500 to 1500)	714.3 (300 to 1500)
Montelukast	1156.3 (500 to 2000)	687.5 (300 to 1500)
Beclomethasone:
Placebo	1253.3 (800 to 2000)	753.3 (500 to 1500)
Montelukast	1165.0 (800 to 2000)	715.0 (400 to 1500)
Budesonide:
Placebo	1288.9 (800 to 2400)	770.4 (400 to 1600)
Montelukast	1078.3 (800 to 2400)	591.3 (400 to 1200)
Flunisolide:
Placebo	2027.8 (1500 to 3000)	1333.3 (500 to 3000)
Montelukast	1911.8 (1500 to 3000)	1264.7 (500 to 3000)
Triamcinolone:
Placebo	2017.4 (1200 to 3200)	1321.7 (400 to 2400)
Montelukast	1991.1 (1200 to 3600)	1217.8 (400 to 2400)

^*Dose actuated, not delivered at mouthpiece. 

Table 3.

Mean (SD) changes in inhaled corticosteroid dose (μg/day) during study

Treatment group	Prestudy	Pre-allocation	Last tolerated	% reduction from allocation
Placebo	1681 (670)	1079 (557)	727 (743)	30%
Montelukast	1588 (646)	976 (553)	526 (716)	47%*

^*P=0.046 compared with placebo. 

Forty five (40%) patients on montelukast and 33 (29%) on placebo tapered completely off inhaled corticosteroids. Seventy (62%) patients on montelukast compared with 57 (50%) on placebo had their dose tapered by 50% or more of the prerandomisation baseline; 31 (28%) on montelukast and 41 (36%) on placebo could not taper at all (P=0.055).

Thirty four (30%) patients on placebo had failed rescues compared with 18 (16%) on montelukast (P=0.01). In addition, time to failed rescue was better with montelukast compared with placebo (P=0.042, log rank test).

No significant differences occurred between treatment groups in the change in spirometry measurements (% change in forced expiratory volume 2.7% montelukast v 5.4% placebo), asthma symptom score (0.07 v 0.12), and β agonist use (0.29 v 0.36 puffs/day) at the last tolerated dose.

There were no significant differences in the frequency of clinical or laboratory adverse experiences between placebo and montelukast patients. Nine patients on placebo and four on montelukast stopped the trial because of adverse effects (worsening asthma for all montelukast patients).

Discussion

We found that, compared with placebo, montelukast (10 mg, administered once daily at bedtime) allowed reduction of moderate to high doses of inhaled corticosteroids while maintaining clinical stability in patients with chronic asthma. These data demonstrate that montelukast has an additive benefit to inhaled corticosteroids in this setting. Another leukotriene receptor antagonist, pranlukast, has been shown to decrease requirements for inhaled corticosteroids, although the study used a different design.¹⁶ Given the potential limitations of prolonged exposure to high doses of inhaled corticosteroids,^17–20 therapeutic strategies providing clinical control of asthma with the lowest corticosteroid dose would be valuable and are consistent with present treatment guidelines.¹

Patients often receive higher doses of inhaled corticosteroids than required.¹⁷ For example, during the placebo run in period of this trial, patients reduced their inhaled corticosteroids dose by about 37%, consistent with previous publications.²¹ Additionally, a pilot study for our trial that carefully evaluated the taper and rescue protocols also found appreciable reductions of inhaled corticosteroids. Because we expected a large placebo response we included a 5-7 week, single blind, placebo run in period.

At allocation, patients appeared well matched. Although the inhaled corticosteroids dose was slightly lower in montelukast patients, the similarity in forced expiratory volume, symptoms, and use of β agonists supports similar disease severity in these patients. Furthermore, a lower inhaled corticosteroid dose at allocation might provide less opportunity for tapering in montelukast patients.

A second, important finding of this study was that the number of failed rescues was lower with montelukast than placebo. Patients taking montelukast were more likely to be successfully rescued after an episode of worsening asthma and therefore more likely to remain clinically stable.

Since the study was time based (12 weeks after randomisation) not event based, it had certain limitations. In some patients, tapering (based on clinical stability) could have continued beyond 12 weeks; therefore, the true number of patients who could have stopped inhaled corticosteroids is unknown. Additionally, since patients were not followed beyond the 12 weeks we do not know how many required the reintroduction of inhaled corticosteroids to maintain clinical stability. A study with longer follow up would allow more complete understanding of the clinical importance of the effect of montelukast observed in this study.

Montelukast was generally well tolerated in this asthmatic population taking high doses of inhaled corticosteroids. Double blind, placebo controlled, clinical trials have confirmed the short term (⩽3 month) safety profile in adults and children (ages 6-14).^6–9 Additional long term clinical trials and real world clinical experience are necessary to understand the long term safety profile of leukotriene receptor antagonists.

In conclusion, our results suggest that leukotriene receptor antagonists such as montelukast may be useful for long term treatment of asthmatic patients requiring high doses of inhaled corticosteroids.