Knorr B, Matz J, Bernstein JA, Nguyen H, Seidenberg BC, Reiss TF, Becker A, for the Pediatric Montelukast Study Group . Montelukast for Chronic Asthma in 6- to 14-Year-Old ChildrenA Randomized, Double-blind Trial. JAMA. 1998;279(15):1181-1186. doi:10.1001/jama.279.15.1181
From the Departments of Pulmonary-Immunology (Drs Knorr, Seidenberg, and Reiss) and Biostatistics (Dr Nguyen), Merck Research Laboratories, Rahway, NJ; Atlantic Asthma and Allergy Center, Inc, Baltimore, Md (Dr Matz); Bernstein Allergy Group, Inc, Cincinnati, Ohio (Dr Bernstein); and Section of Allergy and Clinical Immunology, Department of Pediatrics, University of Manitoba, Winnipeg (Dr Becker).
Context.— Leukotrienes are important mediators of asthma by causing bronchoconstriction,
mucous secretion, and increased vascular permeability. Studies using compounds
that block leukotrienes have demonstrated improvement in asthma control in
adults and adolescents, but children younger than 12 years, for whom asthma
is the most common chronic disease, have not been studied.
Objective.— To determine the clinical effect of montelukast, a leukotriene receptor
antagonist, in 6- to 14-year-old children with asthma.
Design.— Eight-week, multicenter, randomized, double-blind study.
Setting.— Forty-seven outpatient centers at private practices and academic medical
centers in the United States and Canada.
Patients.— A total of 336 children with forced expiratory volume in 1 second (FEV1) between 50% to 85% of the predicted value, at least 15% reversibility
after inhaled β-agonist administration, a minimal predefined level of
daytime asthma symptoms, and daily β-agonist use. Concomitant inhaled
corticosteroids at a constant daily dose were used by 39% of patients receiving
montelukast and 33% receiving placebo.
Intervention.— After a 2-week placebo run-in period, patients received either montelukast
(5-mg chewable tablet) or matching-image placebo once daily at bedtime for
Main Outcome Measure.— Morning FEV1 percent change from baseline.
Results.— Mean morning FEV1 increased from 1.85 L to 2.01 L in the
montelukast group and from 1.85 L to 1.93 L in the placebo group. This represents
an 8.23% (95% confidence interval [CI], 6.33% to 10.13%) increase from baseline
in the montelukast group and a 3.58% (95% CI, 1.29% to 5.87%) increase from
baseline in the placebo group (P<.001 for montelukast
Conclusion.— Montelukast improves morning FEV1 in 6- to 14-year-old children
with chronic asthma.
ASTHMA IS the most common chronic illness of childhood, affecting approximately
10% of children.1 In the United States alone,
approximately 2.2 million ambulatory care visits per year are made by children
for the treatment of asthma.2 Worldwide, the
prevalence of childhood asthma and hospitalizations for it are increasing.3 Current therapies for the treatment of asthma in children
have limitations (eg, requiring inhalations, multiple daily administrations,
or plasma drug level monitoring). Accordingly, new therapies that are effective,
well tolerated, and easily administered would be advantageous in the treatment
of childhood asthma.
Leukotrienes are important mediators of asthma. Leukotrienes are produced
and released from inflammatory cells, including eosinophils and mast cells.
They cause bronchoconstriction, mucous secretion, and increased vascular permeability.4- 6 Studies using compounds
that block leukotrienes (receptor antagonists and 5-lipoxygenase inhibitors)
have demonstrated improvement in asthma control in patients aged 12 years
and older.7,8 We know of no studies
that have addressed the effect of leukotriene blockers in children with asthma
younger than 12 years.
Montelukast (MK-0476) is an orally administered, specific leukotriene
receptor antagonist. In recent adult studies, montelukast (10 mg once daily
at bedtime) demonstrated improvement in parameters of asthma control, including
forced expiratory volume in 1 second (FEV1), daytime and nighttime
symptom scores, and as-needed β-agonist use.9- 11
The purpose of this 8-week study was to determine the effect of montelukast
(5-mg chewable tablet administered once daily at bedtime), compared with placebo,
on parameters of asthma control, including measurements of airways obstruction,
patient-reported end points, and asthma outcomes, as well as to determine
the safety profile in 6- to 14-year-old patients with asthma.
This was a multicenter, double-blind, randomized, placebo-controlled,
2-period, parallel-group study comparing the clinical effect of oral montelukast
(5-mg chewable tablet) with matching-image placebo once daily at bedtime in
6- to 14-year-old children with asthma. The study consisted of a 2-week, single-blind,
placebo run-in period and an 8-week, double-blind, active treatment period
The study was conducted at 46 study centers in the United States and
1 study center in Canada between August 1995 and April 1996. All patients,
study sites, and the coordinating center (Merck Research Laboratories) were
blinded to treatment allocation. Unblinding of the database occurred on June
25, 1996. Patients were randomly allocated according to a computer-generated
schedule in blocks of 5 (3 montelukast, 2 placebo) to receive either montelukast
or placebo. Since airway patency is circadian, bedtime dosing was selected
to provide higher plasma levels of montelukast coinciding with the time of
maximal airways narrowing in the early morning hours.12
All patients used short-acting, inhaled β-agonists as needed to treat
their asthma. A percentage of patients (not to exceed 40%) were allowed concomitant
inhaled corticosteroids at a constant dose and dosing interval, beginning
at least 4 weeks before the prestudy visit. Oral corticosteroid rescue was
permitted for worsening asthma during the study according to a predefined
rescue plan. Patients requiring more than 1 course of oral corticosteroids
were discontinued from the study.
Written informed consent approved by the respective institutional review
boards (in the United States) or ethical review committee (in Canada) was
obtained from the parents or guardians of all patients. Additionally, informed
assent approved by the respective institutional review boards and ethical
review committee was obtained from each patient.
Male and female outpatients, aged 6 to 14 years with a history of intermittent
or persistent asthma symptoms, were enrolled. Eligibility for randomization
included FEV1 between 50% to 85% of the predicted value and an
increase in FEV1 of 15% or greater 20 to 30 minutes after inhalation
of β-agonist at least twice during the prestudy visit and placebo run-in
period. Patients were also required to have a minimum biweekly daytime asthma
symptom score of 21 (see description of diary card below) and to have used,
on average, at least 1 puff of albuterol per day during the 2-week run-in
period. At the prestudy visit, patients received a peak flowmeter (Mini-Wright
model, Clement Clark, Columbus, Ohio) and a practice diary card. To become
eligible for the active treatment period, patients were required to demonstrate
adequate understanding of and competence with these instruments as well as
the ability to perform reproducible spirometry.
Study exclusions included active upper respiratory tract infection within
3 weeks, acute sinus disease requiring antibiotic treatment within 1 week,
emergency department treatment for asthma within 1 month, prior intubation
for asthma, or hospitalization for asthma within 3 months before the prestudy
(screening) visit. Excluded medications included astemizole within 3 months;
oral, inhaled (if not already using concomitantly), or parenteral corticosteroids
within 1 month; cromolyn, nedocromil, β-agonists (oral or long-acting),
antimuscarinics, cimetidine, metoclopramide, phenobarbital, phenytoin, terfenadine,
loratadine, or anticholinergic agents within 2 weeks; and theophylline within
1 week before the prestudy visit. Patients receiving immunotherapy had to
maintain therapy at a constant dosage during the study, and therapy had to
have begun at least 6 months before the prestudy visit. The use of new or
changing doses of concomitant asthma medications by a patient, other than
short-acting, inhaled β-agonists, resulted in discontinuation.
The FEV1 was the prespecified primary end point. Other prespecified
end points were daytime asthma symptoms; AM and PM peak expiratory flow rates
(PEFs); daily use of inhaled, short-acting, as-needed β-agonist; nocturnal
awakenings; pediatric asthma–specific quality-of-life questionnaire;
global evaluations (physician, parent, patient, and combined); change in peripheral
blood eosinophil counts; school loss; and asthma outcome end points, including
episodes of severe worsening asthma (percentage of days and percentage of
patients with an asthma exacerbation), use of rescue oral corticosteroids
(percentage of patients), discontinuations because of worsening asthma (determined
by whether additional asthma medications were required), and asthma-control
days. Clinic-measured PEF was analyzed as a post hoc end point.
Spirometry (FEV1 and PEF) was performed at each clinic visit
between 6 AM and 9 AM. Inhaled β-agonists and short-acting antihistamines
were withheld for at least 6 and 48 hours, respectively, prior to spirometry.
Patients receiving scheduled concomitant inhaled corticosteroids withheld
their morning dose until completion of a clinic visit. The largest FEV1 from a set of 3 acceptable maneuvers at each clinic visit was recorded
as the true value.13 Airway reversibility (evaluated
by measuring FEV1 20 to 30 minutes after administration of 2 puffs
of albuterol) was tested at 2 visits during the run-in period and 4 and 8
weeks after randomization. Spirometry measurements were collected with a standard
spirometer (Puritan Bennett PB 100/PB110, Wilmington, Mass). Each clinic center
transmitted the spirometry data electronically to the central spirometry quality
control center where the data were reviewed to ensure uniform adherence to
American Thoracic Society standards of acceptability and reproducibility.13
A daily diary card (validated for use in 6- to 14-year-old patients)
that contained daytime asthma symptom and nighttime awakening scales was used
in the study.14 The 3 daytime asthma symptom
scales (regarding the frequency and bother of asthma symptoms and activity
limitations due to asthma on a 6-point scale of 0 to 5, where 0 is best) were
combined into a mean daily score. Nighttime awakenings were evaluated by the
response to a single question.14 Diary card
questions were read verbatim by caregivers to patients aged 6 to 8 years and
their responses recorded; patients aged 9 to 14 years answered and recorded
diary card questions under adult supervision. Daytime symptoms were recorded
on the diary card in the evening at bedtime and nocturnal awakenings in the
morning on arising. The change in nocturnal awakenings was determined for
a prespecified group of patients with 2 or more nights with awakenings per
week during the run-in period.
The PEF was measured by the patient in the morning upon arising (AM
PEF) and in the evening at bedtime (PM PEF) before taking study medication.
The largest of 3 measurements was recorded on the diary card. The PEF measurements
performed within 4 hours of β-agonist use were identified on the diary
card. The patients also recorded as-needed β-agonist use during the day
and at night, the use of oral corticosteroid rescue, and an unscheduled visit
to a doctor's office or hospitalization due to worsening asthma. On completion
of the double-blind treatment period, parents, physicians, and patients independently
evaluated the change in the patient's asthma (global evaluations) by selecting
the most appropriate response using a 7-point scale (very much better [score
of 0], moderately better, a little better, unchanged, a little worse, moderately
worse, very much worse [score of 6]). Patients aged 9 to 14 years completed
a validated, self-administered, pediatric asthma–specific quality-of-life
questionnaire at the randomization visit before receiving study medication
and at the last visit of the double-blind, active treatment period (week 8).15 The questionnaire contained 3 asthma-specific quality-of-life
domains—activity, symptoms, and emotions. In response to the questions,
patients identified an answer on a 7-point scale that ranged from 1 (worst)
to 7 (best).
An asthma-exacerbation day was defined as a day when any one of the
following occurred: a decrease of more than 20% from baseline in AM PEF; an
increase of more than 70% from baseline in β-agonist use (a minimum increase
of 2 puffs); an increase more than 50% from baseline in symptom score; "awake
all night" with asthma; or worsening asthma requiring oral corticosteroid
rescue, unscheduled visit to a doctor's office, or hospitalization. An asthma-control
day was defined as any day when none of the following occurred: worsening
asthma requiring oral corticosteroid rescue, unscheduled visit to a doctor's
office, or hospitalization; nocturnal awakenings; and use of more than 2 puffs
Laboratory tests (hematology, chemistry, and urinalysis) were obtained
at the prestudy visit, at the randomization visit (before the patient took
study medication), and after 2, 4, and 8 weeks of treatment. As part of the
hematology assessment, blood samples were analyzed for eosinophil counts by
an automated cell counter in a central laboratory (Covance Laboratories, Indianapolis,
Ind). A complete physical examination (including height and weight) and 12-lead
electrocardiograms were performed before and at completion of the double-blind,
active treatment period. At each clinic visit, adverse experiences and vital
sign measurements were recorded.
An intention-to-treat analysis, including all patients with prerandomization
baseline values and at least 1 treatment period measurement, was performed.
The FEV1 and total daily β-agonist use were analyzed as percent
change from baseline. All other end points were analyzed as change from baseline.
For all end points, the average percent change or change during the treatment
period was compared between the 2 treatment groups using an analysis of variance
(ANOVA) model that included terms for treatment, study center, and stratum
(concomitant inhaled corticosteroid use). Treatment by stratum, age, sex,
Tanner stage, race, allergic rhinitis, and exercise-induced asthma interactions
was evaluated by including each term separately.16
The Shapiro-Wilks test statistic was used to test for normality. The relationship
between baseline FEV1 and treatment effect (percent change from
baseline in FEV1) was examined by a weighted regression. In addition,
an analysis of covariance on FEV1 was performed with change in
height as the covariate. Ordinal data were also analyzed using the Cochran-Mantel-Haenszel
test to corroborate the ANOVA results. Global evaluations were analyzed using
a 7-point scale and by collapsing the scores into 3 response options: better
(0, 1, 2), unchanged (3), and worse (4, 5, 6).
A 95% confidence interval (CI) was calculated for the difference between
the montelukast and placebo treatment groups based on the ANOVA model. All
statistical tests were 2-tailed, and a P value of
.05 or less was considered statistically significant.
The onset of action of montelukast was examined by evaluating the daily
scores or measurements for daytime asthma symptoms, AM PEF, and total daily
as-needed β-agonist use over the first 21 days of the treatment period.
A slope analysis via a mixed-model approach was used to evaluate the treatment
response over time and compare the time course among the treatment groups.
An overall test of equal intercepts and slopes between the treatment groups
All randomized patients were included in the safety evaluations. The
number and percentage of patients reporting adverse experiences and clinical
laboratory abnormalities were summarized by treatment group.
The study was designed with a sample size of 240 patients (96 in the
placebo group and 144 in the montelukast treatment group) to have 90% power
(2-sided test at α=.05) to detect a 7.1-percentage-point difference
in FEV1 percent change from baseline between the 2 treatment groups.
A total of 336 patients entered the double-blind, active treatment period;
201 were assigned to montelukast and 135 to placebo treatment (Figure 1). There were no clinically meaningful differences between
the montelukast and placebo treatment groups in their baseline characteristics
(Table 1). Discontinuations from
the study were similarly distributed between the montelukast and placebo treatment
groups (12 [6%] and 10 [7%], respectively). Seven patients withdrew consent:
2 (1%) in the montelukast group and 5 (3.7%) in the placebo group. One patient
(0.7%) in the placebo group was discontinued because of a protocol deviation.
Two patients were unavailable for follow-up: 1 patient (0.5%) in the montelukast
group and 1 (0.7%) in the placebo group. Eleven patients discontinued from
the study because of an adverse experience: 8 (4%) in the montelukast group
and 3 (2%) in the placebo group. One patient (0.5%) in the montelukast group
discontinued because of a laboratory adverse experience.
Nine patients (5 in the montelukast group and 4 in the placebo group)
were excluded in the intention-to-treat analysis of the primary end point
(FEV1) because of missing baseline or treatment data or significant
deviations from good clinical practice standards.
Montelukast, compared with placebo, caused significant (P<.001) improvement in the primary end point, FEV1 percent
change from baseline. Averaged over the 8-week treatment period, the least
squares mean±SD percent change from baseline in FEV1 was
3.58%±13.33% and 8.23%±13.52% for the placebo and montelukast
groups, respectively. The least squares mean difference between the 2 treatment
groups was 4.65% (95% CI, 1.92%-7.38%). The analysis of the percent change
from baseline in FEV1 with height as a covariate demonstrated similar
results: least squares mean difference of 4.90% (95% CI, 2.15%-7.64%). Furthermore,
the effect of montelukast on FEV1 was consistent (no loss of effect)
over the 8-week treatment period (Figure 2).
Secondary outcomes are summarized in Table 2 and Table 3.
Montelukast, compared with placebo, demonstrated significant improvement in
percent change in total daily as-needed β-agonist use (P=.01), percentage of days (P=.049) and percentage
of patients (P=.002) with asthma exacerbations, all
domains (symptom [P=.007], activity [P<.001], and emotions [P=.002]) of a pediatric
asthma–specific quality-of-life questionnaire, parental (P=.049) and combined (P=.04) global evaluations,
and clinic-measured AM PEF (P=.03). Figure 3 demonstrates that montelukast, compared with placebo, caused
a significant decrease (P=.02) in peripheral blood
eosinophil levels over the 8-week active treatment period. Other secondary
outcomes (daytime asthma symptoms, patient-reported AM and PM PEFs, physician's
global evaluation, patient's global evaluation, nocturnal awakenings, discontinuations
because of worsening asthma, rescue oral corticosteroid use, asthma-control
days, and school loss) did not reach statistical significance; the study,
however, was not powered to detect a difference between treatment groups in
an end point other than FEV1.
The onset of action of montelukast was analyzed using predefined patient-reported
diary card parameters, including daytime asthma symptom scores, total daily
as-needed β-agonist use, and patient-reported AM PEF measurements. Montelukast
had a rapid onset of action (within 1 day of dosing). Figure 4 depicts the time response profile over the first 21 days
of therapy for total daily as-needed β-agonist use. The difference in
average values after the first dose between the montelukast and placebo treatment
groups was significant (P=.02). Similar results over
the first 21 days of therapy were observed for patient-reported AM PEF (P=.03); the results with daytime asthma symptom scores
were not statistically significant but numerically favored montelukast (P=.06).
Of note, the effects of montelukast on FEV1 and total daily
as-needed β-agonist use were consistent (ie, the interactions were not
significant) across sex (P=.77 and .88, respectively),
ethnic groups (P=.23 and .78, respectively), Tanner
stage (P=.32 and .06, respectively), history of allergic
rhinitis (P=.11 and .19, respectively), history of
exercise-induced asthma (P=.80 and .25, respectively),
and concomitant inhaled corticosteroid use (P=.82
and .53, respectively).16 Importantly, the
montelukast treatment effect in patients aged 6 to 11 years and 12 to 14 years
was comparable (mean percent change in FEV1 of 7.7% and 9.8%, respectively).
The most common adverse experiences were headache, asthma, and upper
respiratory tract infection (Table 4).
Overall, there were no significant differences between the montelukast and
placebo treatment groups in the frequency of any adverse experience, with
the exception of allergic rhinitis, which occurred significantly (P=.01) more frequently in the placebo group than in the montelukast
Eleven patients were discontinued from the study because of an adverse
experience: 8 (4%) in the montelukast group and 3 (2%) in the placebo group.
Of the montelukast-treated patients, 5 patients were discontinued because
of asthma, 1 patient was discontinued because of pneumonia, 1 because of dehydration,
and 1 because of an upper respiratory tract infection. Of the 3 placebo-treated
patients, 2 were discontinued because of asthma and 1 was discontinued because
Eleven patients (5.5%) in the montelukast group and 2 (1.5%) in the
placebo group had laboratory values considered adverse experiences during
treatment, the majority of which were transient and self-limited. Importantly,
there were no significant differences between treatment groups in the frequency
of patients with serum transaminase (alanine aminotransferase and aspartate
aminotransferase) elevations. One montelukast-treated patient discontinued
treatment because of a decreased neutrophil count, an abnormality that had
also been present at prerandomization testing.
This study demonstrates the therapeutic benefit of montelukast, a leukotriene
receptor antagonist, in 6- to 14-year-old patients with chronic asthma. Patients
treated with either as-needed β-agonist alone or inhaled corticosteroids
had significant improvement in their asthma control when they received montelukast,
5-mg chewable tablet once daily at bedtime. Though the magnitude of the changes
observed appeared modest, they were consistent with those reported in other
pediatric trials using currently available therapies.17- 20
Montelukast, compared with placebo, demonstrated significant improvements
in FEV1 (primary end point), clinic-measured AM PEF, total daily
as-needed β-agonist use, all domains of a pediatric asthma–specific
quality-of-life questionnaire, parental and combined global evaluations, percentage
of days and percentage of patients with asthma exacerbations, and peripheral
blood eosinophil levels. Other outcomes (daytime asthma symptoms, patient-reported
AM and PM PEFs, physician's global evaluation, patient's global evaluation,
nocturnal awakenings, discontinuations because of worsening asthma, rescue
oral corticosteroid use, asthma-control days, and school loss) did not reach
statistical significance; the study, however, was not powered to detect a
difference between treatment groups in an end point other than FEV1.
The onset of action of montelukast was rapid; treatment effects occurred
within 1 day after the first dose, as assessed by diary card parameters: total
daily as-needed β-agonist use and patient-reported AM PEF. Other controller
agents for asthma, including cromolyn and inhaled corticosteroids, appear
to require a longer treatment duration before their effects become evident.
With inhaled corticosteroids, a treatment period of approximately 1 week may
be needed before pediatric patients with moderate asthma demonstrate improvement
in lung function.21,22 With cromolyn
sodium, a treatment effect generally may require 1 week to 3 weeks of therapy.23
Montelukast not only demonstrated a rapid onset of action, but its treatment
effects were maintained consistently over time. There was no evidence of tolerance
in this or a prior adult study, suggesting that montelukast continues to be
effective in the long-term treatment of asthma.11
Montelukast demonstrated a consistent effect across all subgroups (age,
sex, race, Tanner stage, history of allergic rhinitis, history of exercise-induced
bronchoconstriction, and concomitant inhaled corticosteroid use) and baseline
FEV1 similar to adult studies.11
These findings suggest that a broad range of patients with asthma benefit
from montelukast treatment. The similarity of effect in corticosteroid- and
noncorticosteroid-using patients suggests that the treatment effect of montelukast
may be additive to that of inhaled corticosteroids. Furthermore, in a recent
adult study, treatment with montelukast permitted significant tapering of
inhaled corticosteroids compared with placebo.24
Asthma affects the physical, social, and emotional aspects of the lives
of children.15 Therefore, it was important
to ask children in the study about the impact of the disease and the interventions
they received on their quality of life. Patients receiving montelukast, compared
with patients receiving placebo, reported significant improvements from baseline
for each of the 3 domains (activity, symptoms, and emotions) of a pediatric
asthma–specific quality-of-life questionnaire.
Montelukast improved several other asthma control end points. Montelukast
demonstrated a statistically significant change in the parental and combined
global evaluations of response to study therapy over the active treatment
period. Importantly, montelukast demonstrated a statistically significant
improvement in the mean percentage of days with an asthma exacerbation (asthma-exacerbation
days) as well as in the percentage of patients who experienced at least 1
The eosinophil is an asthma inflammatory effector cell that plays a
critical role in the pathogenesis of asthma. This cell and its mediators are
found in increased quantities in bronchial tissue and are associated with
asthma severity.25 Montelukast, compared with
placebo, also significantly decreased peripheral blood eosinophil counts,
suggesting that montelukast may have significant effects on parameters of
asthmatic inflammation. Inhaled corticosteroids have been shown to similarly
affect peripheral blood eosinophil counts in asthmatic patients while β-agonists
Daytime asthma symptoms reported at home using a diary card improved
(although not statistically significantly); a larger than anticipated placebo
response masked the treatment effect. However, the symptoms domain of the
quality-of-life questionnaire was statistically significant, suggesting that
montelukast does improve asthma symptoms in children.
The rate of clinical adverse experiences between the montelukast and
placebo groups was similar. Laboratory adverse experiences were infrequent,
mild, transient, and similar in frequency between the montelukast and placebo
groups. Furthermore, the incidence of elevated serum transaminase levels was
similar between the montelukast and placebo groups. Overall, montelukast was
generally well tolerated over the 8-week treatment period in 6- to 14-year-old
patients. Future studies will be needed to determine the long-term safety
profile of montelukast. However, no mechanism-based toxic effects have been
identified to date.
In summary, this study demonstrated that montelukast once daily is effective
therapy in 6- to 14-year-old patients with asthma. Montelukast was well tolerated
and demonstrated a safety profile generally similar to placebo. These results
are consistent with and confirm the results seen in adult studies with montelukast.
Overall, the results of this study suggest that montelukast would be a well-tolerated
and effective therapeutic option to current asthma therapies in 6- to 14-year-old