Overview of Reviews
The Cochrane Library and Long-Acting Beta-agonist Treatment for Childhood Asthma: An Overview of Reviews
Asthma is characterized by chronic airway inflammation and affects many children. One treatment option used to control asthma symptoms is the class of medications known as the long acting β2-agonists (LABA) and they are often used in combination with inhaled corticosteroids (ICS). The use of LABA in children has been widely debated in the literature.
To synthesize the evidence currently in the Cochrane Library of systematic reviews related to the question: ‘In children with symptomatic asthma, does treatment with a LABA as a primary or add-on therapy to ICS improve objective measures of lung function and asthma symptom control, reduce exacerbations, and improve quality of life?’.
The Cochrane Database of Systematic Reviews was searched using the term ‘asthma’ in the title for all systematic reviews examining LABA vs any other pharmacological agent for the treatment of asthma. Data were extracted and entered into tables; syntheses occurred using qualitative and quantitative methods.
Seven reviews were identified; however, only four reviews contained data for children. There was no significant difference in the asthma exacerbations among those treated with LABA or any other treatment. For those treated with LABA with no/varied ICS vs placebo with no/varied ICS, there was a significant improvement in percent predicted forced expiratory volume in the first second (FEV1) and a significant number of children had a ≥ 15% increase in FEV1. There was also a significant reduction in the change in whole day rescue medication, fewer asthma related night-time wakenings, and lower symptom scores for LABA with no/varied ICS. Change in FEV1 (Litres (L) or percent predicted) and improvement in morning and evening peak expiratory flow (PEF) favoured LABA and ICS vs fixed dose ICS and placebo. There was a significantly higher risk of adverse events among those receiving LABA and ICS compared to ICS with or without placebo.
There is no clear benefit to LABA in preventing asthma exacerbations in children. While there is a paucity of evidence in this area, ICS should remain the controller therapy of choice; however, the addition of LABA to ICS may result in improvement in pulmonary function measures, reduce the use of rescue medications and improve quality of life measures in children with chronic asthma. Copyright © 2008 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd. The Cochrane Collaboration
Editors' note: Overviews of reviews, compiling evidence from multiple Cochrane reviews into one accessible and usable document, will be a regular feature of this journal. Our aim for each overview is to focus on the treatment question, ‘which treatment should I use for this condition?’, and to highlight the Cochrane reviews and their results in doing so. It is our hope that the overview will serve as a ‘friendly front end’ to the Cochrane Library, allowing the reader a quick overview (and an exhaustive list) of Cochrane reviews relevant to the clinical decision at hand.
Description of the condition
Asthma is one of the most prevalent chronic health conditions worldwide. It is conservatively estimated that 300 million people suffer from asthma worldwide1. Globally, asthma accounts for 15 million disability-adjusted life years (DALYs) lost, and for about 1 in every 250 deaths1. Although asthma may present at any age, the incidence is highest among children2 and in developed countries. Among children in the developed world, the prevalence of physician-diagnosed asthma is estimated to be over 10%3.
Asthma is defined as a disorder of airway inflammation producing paroxysmal or persistent symptoms (dyspnea, chest tightness, wheeze, and cough) associated with variable airflow limitation and airway hyperresponsiveness4. In adults and older children, pulmonary function testing provides objective measures of airflow limitation and airway hyperresponsiveness that can be used to support the diagnosis of asthma. These measures are not easily obtained in preschool children and thus the diagnosis for young children rests on history and physical findings, so the disease may be under-reported. The diagnosis of asthma in this age group is supported by a history of severe wheezing episodes, wheezing after 1 year of age, more than three episodes of wheezing, chronic cough, or a positive response to anti-asthma medications4.
Description of the interventions
The goal of asthma management is to achieve and maintain control of symptoms. While education, trigger avoidance, and environmental improvements can reduce the symptoms, most children with asthma will need medical treatments to achieve these goals. Asthma medications can be broadly divided into relievers and controllers. Relievers are medications used to reverse bronchoconstriction and reverse the associated symptoms. In the past, most relievers were short-acting (e.g. salbutamol, ipratropium bromide); however, longer duration beta-agonist (e.g. long acting ß2-agonists (LABA) such as salmeterol and formoterol) and anticholinergic (e.g. tiotropium) reliever therapy now exist. Controller medications are taken daily on a long-term basis to keep asthma under clinical control and include inhaled corticosteroid (ICS) and leukotriene receptor antagonists (LTRA).
The most common reliever medications are short acting inhaled ß2-agonists (SABA) such as salbutamol and terbutaline. These medications are hydrophilic and interact directly with ß2-receptors in airway smooth muscle resulting in smooth muscle relaxation and relief of bronchoconstriction. They have a fast onset of action and duration of effect of six hours or less5. LABA agents such as salmeterol and formoterol were developed to provide longer symptom control and have duration of action up to 12 hours. Such properties can be particularly useful feature for preventing night time symptoms; however, the role of long-acting ß2-agonists in the management of asthma has been debated. There has been concern about the regular use of LABA on bronchial hyperresponsiveness6, development of tolerance7, progressively reduced protection against provoking stimuli8, and masking the deterioration of asthma9. Frequent use of SABA or LABA agents indicate a significant inflammatory process that should be controlled with anti-inflammatory drugs such as ICS.
Why is it important to do this overview?
While monotherapy with LABA is not recommended, guidelines differ regarding when to add LABA to ICS treatment. All recommend LABA as add-on to ICS therapy; however, the dose of ICS at which this step is recommended varies among guidelines as well as by patient age4, 10–12. Given the high prevalence of asthma in the population, and the burden it imposes on the child, their family, and society, it would be valuable for clinicians to have a strong evidence base to support treatment decisions regarding the use of LABA agents. This overview of reviews summarizes the evidence from existing Cochrane Library reviews regarding the use of LABA in the treatment of asthma in children.
Objectives of the Overview
The objective of this overview is to synthesize the evidence currently in The Cochrane Library of Systematic Reviews related to the following questions:
In children with symptomatic asthma, does treatment with LABA as a primary or add-on therapy to ICS improve objective measures of lung function and asthma symptom control, reduce exacerbations, and improve quality of life?;
Is there any evidence of harm associated with the use of LABA in children with symptomatic asthma?
The Cochrane Database of Systematic Reviews was searched for all systematic reviews examining LABA agents vs any other pharmacological agent for the treatment of asthma in children. The term ‘asthma’ was entered and restricted to the record title; this resulted in 139 systematic reviews, seven of which examined LABA treatments. Two reviewers examined the seven systematic reviews for eligibility. Upon further assessment, three systematic reviews were excluded because the review did not include any trials in children13–15. The four remaining reviews form the basis of this overview.
Data extraction and synthesis
One reviewer extracted the following data from each review: search strategies, inclusion criteria (including population, intervention, comparison(s) and outcomes), methodological quality assessment, and results. We did not approach the authors of the reviews or of any of the individual studies included within the reviews for unpublished data. Dichotomous data were presented as Relative Risks (RR) and 95% Confidence Intervals (CI) and continuous data were summarized as a weighted mean difference (WMD) and 95% CIs. When more than two studies were included in the comparison, an I2 was calculated to quantify the degree of statistical heterogeneity. I2 values of 25%, 50%, and 75% would typically represent, low, moderate and high levels of heterogeneity respectively16. Statistical analysis was performed in RevMan using fixed effects modelling to assign study weights.
Description of Cochrane Reviews included
Criteria for including studies
The included reviews are described in Table 1. The inclusion criteria were fairly comparable. All four reviews included randomized controlled trials conducted in adults and children; however, only results for children are included in this overview of reviews. When a lower age range was reported, the reviews included children ≥ 2 years of age15, 17. The reviews examined those with chronic asthma (15; definition: patients having been on ICS daily for at least 30 days prior to study entry), chronic or recurrent asthma (17; no definition provided) and those with a clinical diagnosis of asthma (18–19; definition: clinical diagnosis of asthma for 3 or 6 months, respectively). Four comparisons were identified:
LABA twice a day via any inhalation device vs daily regular SABA;
LABA vs placebo in chronic asthma in addition to fixed dose ICS;
LABA twice a day with ICS vs higher dose ICS with or without placebo;
LABA twice a day via any inhalation device vs placebo administered in the same way as the LABA among those receiving varied or no ICS.
Table I. Characteristics of included reviews
|Regular treatment with long acting ß2-agonists vs daily regular treatment with short acting beta agonists in adults and children with stable asthma||Walters EH, Walters JAE, Gibson PW||April 2002||3||22–874||Adults and children with a clinical diagnosis of asthma for at least 6 months||LABA (salmeterol or formoterol) twice a day via any inhalation device||Short acting inhaled beta agonist *co-interventions included ICS||Day and night symptom scores; Bronchodilator use for symptom relief; Daily PEF; Airway function; Exacerbation rates; Adverse events; Airway hyper-reactivity; Quality of life; Global assessment of efficacy (patient and investigator)|
|Long-acting ß2-agonists vs placebo in addition to inhaled corticosteroids in children and adults with chronic asthma||Ni Chroinin M, Greenstone IR, Danish A, Magdolinos H, Masse V, Zhang X, Ducharme FM||June 2005||9||16–286||Children aged ≥ 2 years and adults with chronic asthma and having received daily ICS for at 30 days prior to study entry||LABA and ICS||Placebo and ICS *co-interventions allowed providing they remained unchanged throughout the study period||Asthma exacerbations requiring short course ICS; Hospital admissions; Asthma control (pulmonary function tests, days and nights without symptoms, functional status, quality of life and rescue short- acting beta2agonists)|
|Combination of inhaled long-acting ß2-agonists and inhaled steroids vs higher dose of inhaled steroids in children and adults with persistent asthma||Greenstone IR, Ni Chroinin MN, Masse V, Danish A, Magdalinos H, Zhang X, Ducharme FM||July 2005||3 (2 contributed to meta-analysis)||20–120||Children aged ≥ 2 years and adults with recurrent or chronic asthma||LABA (salmeterol or formoterol) twice a day with ICS||Higher dose ICS with or without placebo *co-interventions allowed providing they remained unchanged throughout the study period||Asthma exacerbations of moderate intensity (requiring a short course of systemic corticosteroids); Asthma exacerbations requiring hospital admissions; Asthma control (pulmonary function, days with symptoms, symptom scores, quality of life, functional status, rescue SABA); Withdrawal rates; Changes in inflammation; Clinical and biochemical adverse effects|
|Long-acting ß2-agonists for chronic asthma in adults and children where background therapy contains varied or no inhaled corticosteroid*||Walters EH, Gibson PG, Lasserson TJ, Walters JAE||Nov 2006||11||24–518||Children and adults with a clinical diagnosis of asthma for at least 3 months||LABA (salmeterol or formoterol) twice a day via any inhalation device where some of the children are receiving ICS||Placebo administered in the same way as LABA where some of the children are receiving ICS||Day and night symptom scores; Bronchodilator use for symptom relief; Daily PEF; Lung function; Exacerbation rates; Adverse events; Airway hyper-reactivity; Quality of life; Global assessment of efficacy (patient and investigator); Reduction in other asthma medication use; Tolerance to beta-2 agonists|
One review18 failed to report primary outcomes and the remaining three reviews assessed asthma control19 and asthma exacerbations15, 17.
All four reviews searched the Cochrane Airways Group (CAG) Asthma and Wheez* register of randomized controlled trials (RCT) to identify studies examining long acting β2-agonists. The CAG register represents a comprehensive search of MEDLINE, CINHL, EMBASE, and hand-searching of more than 20 of the world's leading respiratory journals. All reviews examined the reference lists for other potentially relevant trials. Two reviews contacted other researchers18–19 and one review contacted pharmaceutical companies manufacturing the agent15. Three reviews specifically stated that they did not impose any language restriction15, 17, 19.
Description of studies
The four reviews included a minimum of three and a maximum of 11 trials in children; the smallest trial involved 16 children and the largest 874 children. All the trials compared LABA to either a placebo or a second active treatment. When the inclusion criteria for age range were reported, the minimum age of children was 2 years of age15, 17. Similar outcomes or variations of those outcomes were reported in all four reviews.
Methodological quality of studies included in the reviews
All four of the reviews concluded that the methodological quality of the included studies was good and the majority of the included trials were of high quality. The reviews used the same methodological tools: the 5-point Jadad score and the Cochrane approach to assessing concealment of allocation. Of the 26 trials, one trial received a Jadad score of 1 and two trials received a score of 2. The remaining 23 studies scored a 3 (23%), 4 (38%), or 5 (27%). The majority of trials used an adequate method for concealing allocation (54%); however, the proportion of trials that used adequate allocation concealment within each review was variable (18%19 to 100%15).
Effects of interventions
Regardless of the comparison, LABA did not result in clinically or statistically significantly different rates of asthma exacerbations requiring steroids or hospitalizations (Table 2).
Table II. Exacerbations
|Requiring||LABA and ICS||174 (2)||10/87||9/87||RR||1.11 (0.49, 2.52)|
|steroids||vs higher dose ICS|| || || || ||I2 = 36.3%|
| ||LABA and ICS||402 (4)||29/202||32/200||RR||0.90 (0.57, 1.42)|
| ||vs fixed dose ICS|| || || || ||I2 = 0%|
|Requiring||LABA and ICS||120 (1)||1/60||0/60||RR||3.00 (0.12, 72.20)|
|hospitalization||vs fixed dose ICS|| || || || || |
| ||LABA and ICS vs||641 (4)||16/328||11/313||RR||1.49 (0.69, 3.21)|
| ||fixed dose and ICS|| || || || ||I2 = 6.7%|
|Major||LABA vs SABA||568 (1)||89/290||82/278||RR||1.04 (0.81, 1.34)|
|exacerbation*||(+/− ICS)|| || || || || |
|> 1 major||LABA with||1229 (4)||224/620||202/609||RR||1.07 (0.94, 1.23)|
|exacerbation||no/varied ICS vs|| || || || ||I2 = 7.0%|
| ||placebo with|| || || || || |
| ||no/varied ICS|| || || || || |
Airway obstruction or airway inflammation
FEV1 and PEF were used to measure lung function (Table 3).
Table III. Airway obstruction or airway inflammation
|FEV1 (L) at endpoint||LABA and ICS vs higher dose ICS||120 (1)||4.36 (1.51)||4.08 (1.47)||MD||0.28 (−0.25, 0.81)|
|Change in FEV1(L)||LABA with no/varied ICS vs placebo with no/varied ICS||347 (1)||0.32 (0.40)||0.24 (0.39)||MD||0.08 (0.00, 0.16)|
|Change in FEV1 (%predicted)||LABA with no/varied ICS vs placebo with no/varied ICS||347 (1)||11.00 (15.00)||6.00 (4.00)||MD||5.00 (1.95, 8.05)|
|%predicted FEV1||LABA vs SABA (+/− ICS)||22 (1)||88 (17.68)||76 (13.27)||MD||12.00 (−1.06, 25.06)|
|Change in FEV1 (L or % predicted)||LABA and ICS vs fixed dose ICS||340 (4)||NR||NR||SMD||0.36 (0.15, 0.58)|
| || || || || || ||I2 = 52.2%|
|Number of children with ≥ 15% increase in FEV1||LABA with no/varied ICS vs placebo with no/varied ICS||344 (2)||124/181||87/163||RR||1.28 (1.08, 1.52)|
| || || || || || ||I2 = 0.0%|
|Change in morning PEF (L/min)||LABA and ICS vs higher dose ICS||120 (1)||30.92 (33)||22.57 (33)||MD||8.35 (−3.46, 20.16)|
|Change in morning PEF (L/min)||LABA and ICS vs fixed dose ICS||587 (4)||25.80 (32.33)||12.68 (38.82)||WMD||12.35 (7.02, 17.68)|
| || || || || || ||I2 = 31.6%|
|Change in morning PEF (%predicted)||LABA with no/varied ICS vs placebo with no/varied ICS||351 (1)||17.00 (13.42)||15.00 (13.08)||MD||2.00 (−0.77, 4.77)|
|Change in evening PEF |(L/min)||LABA and ICS vs fixed dose ICS||307 (2)||21.94 (30.59)||4.90 (33.56)||WMD||12.96 (6.59, 19.34)|
| || || || || || ||I2 = 89.1%|
|Change in evening PEF (L/min)||LABA with no/varied ICS vs placebo with no/varied ICS||556 (2)||NR||NR||WMD||23.04 (−3.54, 49.61)|
| || || || || || ||I2 = 93.7%|
|Change in evening PEF (%predicted)||LABA with no/varied ICS vs placebo with no/varied ICS||349 (1)||16.00 (13.38)||12.00 (13.04)||MD||4.00 (1.23, 4.77)|
|Evening PEF at endpoint (L/min)||LABA and ICS vs higher dose ICS||48 (1)||290 (73.5)||286 (73.5)||MD||4.00 (−35.79, 45.59)|
|Number of children with ≥ 15% increase in PEF||LABA with no/varied ICS vs placebo with no/varied ICS||199 (1)||79/103||46/96||RR||1.60 (1.27, 2.02)|
LABA with no/varied ICS vs placebo with no/varied ICS had a significant increase in change in percent predicted FEV1 (MD: 5.00 [95% CI: 1.95, 8.05]), and the number of children with ≥ 15% increase in FEV1 (RR: 1.28 [95% CI: 1.08, 1.52]). There was no significant difference in change in FEV1 (MD: 0.08 [95% CI: 0.00, 0.16]). There was a significant improvement in change in FEV1 (L or %predicted) for LABA and ICS vs fixed ICS (SMD: 0.36 [95% CI: 0.15, 0.58]). There was no significant change in FEV1 outcomes among the other comparisons.
Compared to placebo and ICS, LABA and ICS treatment also resulted in a significant improvement in change in morning PEF (WMD: 12.35 [95% CI: 7.02, 17.68]) and change in evening PEF (WMD: 12.96 [95% CI: 6.59, 19.34]). There was a significant increase in evening PEF among children treated with LABA with no/varied ICS vs placebo with no/varied ICS (MD: 4.00 [95% CI: 1.23, 4.77]) and a greater number of children had a ≥ 15% increase in PEF (RR: 1.60 [95% CI: 1.27, 2.02]). There was no significant difference in PEF outcomes for the other comparisons.
Use of rescue medication
There was a significant reduction in the change in whole day rescue medication (puffs/day) for LABA with no/varied ICS (WMD: − 0.50 [95% CI: − 0.94, − 0.06]). The remaining comparisons and outcomes were not significant for the use of rescue medication (Table 4).
Table IV. Use of rescue medication
|Daytime use||LABA and ICS vs higher dose ICS||48 (1)||0.25 (0.34)||0.46 (1.02)||MD||− 0.21 (−0.64, 0.22)|
|Change in daytime rescue medication (puffs/day)||LABA and ICS vs fixed dose ICS||159 (1)||− 0.75 (1.78)||− 0.38 (1.73)||MD||− 0.37 (−0.92, 0.18)|
|Change in evening rescue medication (puffs/day)||LABA and ICS vs fixed dose ICS||153 (1)||− 0.15 (1.05)||− 0.11 (0.97)||MD||− 0.04 (−0.36, 0.28)|
|Change in overall daily rescue inhalation (puffs/day)||LABA and ICS vs fixed dose ICS||286 (1)||− 0.11 (0.75)||− 0.09 (0.70)||MD||− 0.02 (−0.18, 0.14)|
|Change in whole day rescue medication (puffs/day)||LABA with no/varied ICS vs placebo with no/varied ICS||207 (1)||− 0.80 (2.02)||− 0.30 (1.03)||MD||− 0.50 (−0.94, − 0.06)|
Indicators of quality of life
LABA with no/varied ICS vs placebo with no/varied ICS had significantly higher percent of nights without asthma related awakening (MD: 6.40 [95% CI: 2.11, 10.69]) and had significantly lower symptom score (MD: − 1.01 [95% CI: − 1.76, − 0.26]). There was no significant difference in other indicators of quality of life for the other comparisons and outcomes (Table 5).
Table V. Indicators of quality of life
|% free symptom free days||LABA and ICS vs fixed dose ICS||286 (1)||77.5 (22)||75.1 (20)||MD||0.12 (−0.11, 0.35)|
|Symptom score (whole day)||LABA with no/varied ICS vs placebo with no/varied ICS||32 (1)||5.87 (1.21)||6.88 (0.90)||MD||− 1.01 (−1.76, − 0.26)|
|Change in number of nights without asthma symptoms||LABA with no/varied ICS vs placebo with no/varied ICS||207 (1)||9.10 (22.22)||4.10 (20.49)||MD||5.00 (−0.83, 10.83)|
|% of nights without asthma awakening||LABA with no/varied ICS vs placebo with no/varied ICS||210 (1)||94.80 (11.00)||88.40 (19.50)||MD||6.40 (2.11, 10.69)|
|Change in number of nights wakening||LABA and ICS vs fixed dose ICS||286 (1)||− 1.7 (10.4)||− 1.9 (10.4)||MD||0.20 (−2.21, 2.61)|
|Global assessment of patient efficacy - good or very good||LABA with no/varied ICS vs placebo with no/varied ICS||184 (1)||56/91||46/93||RR||1.24 (0.96, 1.62)|
|Day time symptom score at endpoint||LABA and ICS vs higher dose ICS||48 (1)||0.24 (0.49)||0.18 (0.49)||MD||0.06 (−0.22, 0.34)|
Total adverse events
There was a significantly higher risk of adverse events among those receiving LABA and ICS compared to higher dose ICS (RR: 1.13 [95% CI: 1.02, 1.26]). There was no significant difference in the occurrence of adverse events for LABA vs the other comparisons (Table 6).
Table VI. Total adverse events
|LABA and ICS vs. higher dose ICS||120 (1)||59/60||52/60||RR||1.13 (1.02, 1.26)|
|LABA vs. SABA (+/− ICS)||80 (1)||27/40||32/40||RR||0.84 (0.65, 1.10)|
|LABA and ICS vs. fixed dose ICS||344 (3)||153/170||155/174||RR||1.01 (0.94, 1.08)|
| ||I2 = 55.6%|
|LABA with no/varied ICS vs. placebo||1268 (5)||446/641||409/627||RR||1.06 (0.99, 1.15)|
|with no/varied ICS|| ||I2 = 5.4%|
Asthma is one of the most common chronic diseases affecting children. Most children need medical therapy to achieve and maintain control of their symptoms. While anti-inflammatory control with ICS is the cornerstone of therapy, some children remain poorly controlled and may require additional therapy. LABA, which can provide symptom relief for up to 12 hours, have been suggested as one such therapy. This overview of reviews examined the role of LABA in the control of childhood asthma
The outcome measures extracted from the four Cochrane Reviews included within this overview of reviews were grouped as follows:
Surrogate measures of severity of airway inflammation such as asthma exacerbation requiring systematic steroids, hospital admissions, and use of reliever medications;
Objective measures of severity of airway inflammation such as FEV1 and peak expiratory flow rates;
Quality of life indicators such as symptom scores, symptom free days, and change in night time symptoms;
Summary of the Main Results
From the four included Cochrane Reviews, there was no clear benefit to LABA in any comparison in preventing asthma exacerbations in children. Overall, there appears to be small improvement in pulmonary function measures, a possible reduction in use of rescue medications and some improvement in quality of life measures in patients who received LABA in addition to ICS. There was a suggestion of a higher risk of adverse events among those patients who received LABA. Overall, caution is advised in applying these results since they are based on multiple comparisons, small number of trials and the clinical significance of some pooled results are questionable.
Overall completeness and applicability of the evidence
Overall there were a small number of heterogeneous trials identified, most with small sample sizes and with a short duration of intervention, making it difficult to comment on the sustainability of improvements seen with short term treatment and on differences in side effects or adverse events between treatment groups. There was no data identified pertaining to young children (less than 4 years of age) thus no conclusion can be drawn for this age group. There were no data identified that compared LABA as monotherapy vs placebo without some background use of ICS among the study population; the current guidelines4, 10–12 do not recommend it as mono-therapy and these reviews provide no counter evidence.
Quality of the evidence
There were 26 trials identified in the 4 reviews. The overall methodological quality of the included studies was ‘good’; however, just over half used an adequate method for concealment of allocation.
LABA vs SABA
Three studies within the relevant Cochrane Review included a comparison of regular LABA use to regular SABA use in children with asthma18. These studies were heterogeneous in design, intervention, length of treatment, co-interventions allowed and outcomes examined. Two trials were parallel design RCTs20–21 while one was a cross-over study that used no washout period between treatment groups22. Two studies21–22 used salmeterol as the LABA while the remaining study used formoterol20. All studies used salbutamol as the SABA. ICS were allowed as a co-intervention in two studies with 60% and 100% of patients using ICS respectively21–22 while in the remaining study, patients used only β2-agonists20. Two studies were small, single centre (22 and 45 patients)20, 22 and one was a larger, multicentre study with 847 patients enrolled21. The duration of two studies was short at 4 and 12 weeks respectively20, 22 while one study followed patients for 52 weeks21. The mean age of the participants ranged from 9–10.2 years, limiting the generalizability of this comparison to younger children. Only one trial examined the risk of major exacerbations (defined as a worsening of asthma symptoms requiring treatment in addition to the study drug and rescue SABA) and found no difference between LABA and SABA groups. As well, although 847 were randomized in the trial, this specific outcome is reported for only 568 patients. One study examined change in FEV1 and found no difference between LABA and SABA20. The summaries of two studies21–22 included in the review list quality of life scores (symptoms scores) as outcome measures but these are not reported in the review with no comment being made on why they are not reported.
In summary, from these few studies there is no evidence that treatment with regular LABA confers any benefit to regular treatment with SABA.
ICS plus LABA vs higher dose ICS
The relevant Cochrane Review17 identified only three studies in which higher dose ICS were compared to ICS plus LABA23–25. Of these three studies, only two provided useable data23, 25. These studies were also heterogeneous in design, interventions, length of treatment, and outcomes measured. One trial used a cross-over design to randomize 27 children to treatment with formoterol and budesonide 100 mcg twice/day and then to budesonide 200 mcg twice/day for a total of 12 weeks23. The other, a multicentre RCT randomized 120 patients to salmeterol twice/day and beclomethasone 200 mcg twice/day or to beclomethasone 400 mcg/day and followed patients for 54 weeks25. The mean age of the patients in these two studies ranged from 4.5–11.4 years. Neither study found a difference between groups in exacerbations requiring steroids or a change in any objective measure of pulmonary function. The studies also found no difference in admissions to hospital25, use of SABA rescue medications23, or asthma symptom score25. Overall adverse events were reported in one study and were found to be slightly higher among those patients receiving LABA25.
Thus, within the limitations of the two studies identified by this review, there is no evidence that the addition of LABA to steroids provides any benefit over higher dose ICS dose for children with asthma.
ICS plus LABA vs fixed dose ICS
The relevant Cochrane Review15 identified nine studies in which LABA were compared to fixed dose ICS25–33. Of the nine studies, all were parallel design RCTs, six were multicentre studies25–26, 29, 31, 32–33, sample sizes ranged from 16–286 patients, and none enrolled children less than 5 years of age. In three trials26, 31–33 formoterol was the LABA examined while the remaining studies examined salmeterol. Most of the trials used short duration (<16 weeks) interventions, although one25 used a 54 week intervention period. The ICS used in four studies was beclomethasone at a dose range of 100–200 mcg twice/day25, 28, 30, 31, budesonide was used in one study27, and four studies reporting maintaining the patient on their ‘usual dose of ICS’ but did not specify specific ICS examined26, 29, 32–33. Of the five studies25, 27, 29–31 that examined exacerbations (requiring steroids or admission), none found a benefit to the use of LABA. Four studies reported an objective measure of pulmonary function25, 27–29. Both change in FEV1, morning PEF and evening PEF were improved in patients receiving LABA and ICS compared to ICS and placebo. No benefit to LABA was found in the two studies that examined quality of life measures29, 31, nor in the two studies that examined rescue SABA29, 31 and no difference was found in three studies that reported total adverse events between the groups25, 27, 29.
In summary, although there were nine studies for this comparison, they were heterogeneous, did not include young children, and found only small improvements in pulmonary function tests with the addition of LABA to a fixed dose of ICS agent.
LABA with varied ICS or no ICS vs placebo with varied ICS or no ICS
The relevant Cochrane Review19 identified 11 studies in which LABA were compared to placebo in populations with variable background use of ICS34–44. Of the 11 studies, all but one44 were parallel design RCTs, 9 were multicentre studies34–39, 41–44 and sample sizes ranged from 24–518 patients. The mean age of the children in these studies ranged from 8–11.9 years with no study enrolling children less than 4 years of age. In four trials formoterol was the LABA examined34–35, 40, 42 while the remaining studies examined salmeterol36–39, 41, 43–44. Most trials used a short-duration (<12 weeks) intervention, although three examined one year of intervention34, 39, 41 and one 26 weeks44. The background ICS use ranged from 0–82% of participants. Of the four studies34–35, 41, 43 that examined asthma exacerbations, none found a benefit to the use of LABA. Four studies reported an objective measure of pulmonary function35, 37, 41, 43. Change in FEV1, number of children with > 15% increase in FEV1, evening PEF, and number of children with > 15% increase in PEF were improved in patients receiving LABA. Two studies examined quality of life measures40, 43, only one43 found any benefit to LABA. Only one study found a reduction in the use of rescue medications43. No difference was found in the five studies that reported total adverse events between the groups37–38, 41–43.
In summary, in this population with varied ICS use, the only benefits from the addition of LABA was small improvement in pulmonary function testing, a suggestion of improvement in quality of life and reduced use of rescue medications.
Agreements and disagreements with other studies or reviews
Patients with persistent asthma symptoms require maintenance therapy with ICS12, 44. Those patients who continue to have symptoms while on moderate-dose ICS therapy often have a second or third treatment added. These additions may include LABAs or LTRA; however, LABA/ICS combination therapy is recommended and used most commonly in adults with asthma14. In addition, there is relatively clear evidence that they should not be used as monotherapy in adults, as long-acting beta2-agonist monotherapy has been shown to increase severe and life-threatening asthma exacerbations, as well as asthma-related deaths46. The evidence in adults is based on a number of systematic reviews in the Cochrane Library which include a large number of trials and study patients. Our overview does suggest that there may be a higher risk of adverse events among children who received LABA although clearly, adult evidence is more robust than what we were able to find in this overview of reviews.
Implications for practice
There is no evidence that LABA should be used as monotherapy in children with asthma.
There is no evidence that LABA as add-on therapy to ICS reduces risk of exacerbations requiring admission or steroids in children with asthma.
There is insufficient evidence to comment on the role of LABA in children less than 4 years of age.
LABA as add-on to ICS may improve pulmonary function measures and some quality of life outcomes.
ICS should remain the therapy of choice in children with asthma until contrary evidence emerges and, as in other chronic disease, adherence should be reviewed prior to consideration of newer agents.
Despite the paucity of evidence, consensus guidelines recommend LABA as add-on therapy to ICS in children with poorly controlled asthma.
Implications for research
Future research in children with asthma should address the gap in specific age groups of children with asthma.
Long-term clinical trials are needed to adequately assess the benefits and risks to LABA in children.
The authors would like to acknowledge Denise Thomson for her assistance with this review. Dr Rowe's research is supported by the 21st Century Canada Research Chairs Program through the Government of Canada (Ottawa, Ontario).
Contributions of Authors
All authors contributed to the methods of the review. KR extracted the data for the tables and constructed the results section. AP is the primary author of this report; all other authors contributed to edits and take responsibility for the manuscript.
Declaration of Interest
In the past 2 years, Dr Rowe has received speaking fees and research funding from AstraZeneca and GlaxoSmithKline for asthma research. He is neither a paid employee nor a share holder in either of these companies. No other authors have conflicts of interest to declare.