Lay summary

It can be difficult for doctors to decide which treatment is best to prescribe to children and adolescents with asthma to help reduce their symptoms. In this review, we weigh up the available evidence on three asthma treatments that work in different ways. We looked at two types of inhalers and one type of medicine that is either swallowed as a tablet or granules. The two inhalers helped to improve lung function more than the oral medication, which may be due to their different modes of action. All three treatments were found to be as safe as a placebo.

Introduction

Asthma is one of the most prevalent chronic diseases in childhood [1], yet diagnosing and treating asthma in children remains challenging. Poor control of asthma in children and adolescents is common and represents a considerable cause of morbidity [2, 3]. In addition to its physical effects, the disease can have an emotional impact on the patient and cause a great burden for patients’ families and the community [1]. There is, therefore, a need for more pharmacological options to improve asthma control in children and adolescents whose symptoms are not fully treated with inhaled corticosteroids (ICS).

Selecting the most appropriate add-on treatment to manage and reduce asthma symptoms in children and adolescents whose asthma remains uncontrolled despite treatment can be challenging. The Global Initiative for Asthma (GINA) recommends that patients with asthma who continue to experience symptoms and/or exacerbations on low-dose ICS have their ICS dose increased and combined with long-acting β2-agonists (LABAs) or other controllers in a step-wise fashion (Fig. 1). Further controller medications include long-acting muscarinic antagonists (LAMAs; e.g. tiotropium), leukotriene receptor antagonists (LTRAs), theophylline and biologics [4]. GINA also recommends as-needed low-dose ICS/formoterol as reliever therapy in all patients > 12 years of age, with short-acting β2-agonists (SABAs) recommended as an alternative reliever medication [4], although it should be noted that the recommendation for children is to ensure additional ICS is taken whenever the SABA reliever is given [4]. The goals of asthma management are aligned across all age groups: namely, to achieve good symptom control, maintain normal activity levels, lung function and development, and minimise future risk of exacerbations and side effects associated with medication [4].

Fig. 1
figure 1

GINA treatment recommendations for patients aged ≥ 5 years, 6–11 years and ≥ 12 years [4]. FEV1, forced expiratory volume in 1 s; GINA, Global Initiative for Asthma; HDM, house dust mite; ICS, inhaled corticosteroid; LABA, long-acting β2-agonist; LAMA, long-acting muscarinic antagonist; LTRA, leukotriene receptor antagonist; OCS, oral corticosteroids; SABA, short-acting β2-agonist; SLIT, sublingual immunotherapy

Full size image

Previous studies have demonstrated the efficacy and safety of LABAs as add-on to ICS compared with placebo [5, 6]. LABAs are available both as single therapy to be taken as add-on to ICS, or as dual therapy, where ICS and LABA are delivered in the same device. Single-therapy LABAs are indicated as add-on treatment to ICS for patients aged from 4 years in Europe and the USA [7,8,9,10].

Tiotropium, an alternative add-on treatment to ICS, is a LAMA that is efficacious in clinical trials in adolescents and children with asthma as add-on to ICS [11, 12] or to ICS with other controllers [13, 14]. In the European Union, it is now indicated as add-on maintenance treatment in patients aged 6 years and older with severe asthma who experienced one or more severe asthma exacerbations in the past year [15]. In the USA, tiotropium is indicated in the long-term, once-daily maintenance treatment of asthma in patients aged 6 years and older [16].

The LTRA montelukast is indicated in the treatment of asthma as an add-on therapy in paediatric patients with mild-to-moderate persistent asthma who are inadequately controlled on ICS and in whom SABAs provide inadequate control [17]. It can also be tried as an alternative to ICS in patients with mild-to-persistent asthma who do not have a history of asthma attacks and have trouble using inhaled medications, and is indicated for the prophylaxis of asthma in patients aged at least 2 years [18]. Montelukast oral granules are indicated in patients aged between 6 months and 5 years [19].

Despite the availability of these controller medications, few studies have directly compared their efficacy in adolescents and children with asthma. A number of systematic reviews have compared the effects of LAMAs, LABAs and LTRAs as add-on to ICS in patients with asthma [6, 20,21,22], although reviews in children aged < 12 years or adolescents aged 12–18 years are limited. Moreover, none have been published that compare the efficacy and safety of all three add-on treatments within one review in patients aged ≤18 years. More systematic reviews and treatment recommendations have been published for patients aged ≥12 years than those for younger patients. As such, there is a need for an up-to-date review of the literature related to the treatments available as add-on to ICS in paediatric patients with asthma.

The aim of this literature review is to compare the efficacy and safety of three controller options (LAMA, LABA and LTRA) as add-on to ICS in adolescents and children aged 4–17 years with asthma. We compare the magnitude of forced expiratory volume in 1 s (FEV1) improvements with each drug class, their effects on exacerbations, and the proportion of patients with adverse events (AEs) and serious AEs (SAEs).

Methods

We carried out an electronic literature search of the Cochrane Database of Systematic Reviews in December 2018 to identify any previously published systematic reviews, which were then manually checked for relevance. We then searched PubMed for articles published since the search date detailed within the systematic review.

The inclusion criteria for this review were randomised controlled trials (RCTs) of at least 4 weeks in duration in children and adolescents aged 4–17 years. The types of intervention included LABA, LAMA or LTRA versus placebo, or versus each other, added onto ICS, compared with the same dose of ICS alone. The primary outcome of interest was lung function, measured using FEV1. For FEV1, we included percent predicted as well as absolute values, as this has the advantage of removing physical confounding factors, particularly when comparing studies with different age groups of children. Secondary outcomes included exacerbations requiring oral corticosteroids (OCS), and proportion of patients reporting AEs and SAEs.

Data were extracted from published articles in PubMed and publicly available data online. We also checked the reference lists of the systematic reviews for any additional data for endpoints that were not described in the systematic reviews. Results were compared with data from tiotropium trials in paediatric patients (PensieTinA- [NCT01277523], VivaTinA- [NCT01634152], RubaTinA- [NCT01257230] and CanoTinA-asthma® [NCT01634139]).

We used the following search strings:

Studies of LABA as add-on to ICS

((((((((((clinical trial[MeSH Terms]) OR clinical trial) OR clinical study)))))

AND asthma[MeSH Terms]))

AND ((((((((Asthma Control Questionnaire) OR ACQ)) OR ((forced expiratory volume) OR FEV)) OR ((exacerbation) OR worsening)) OR adverse event)))))

AND ((((((((((((((((((child*) OR paediat*) OR pediat*) OR adolesc*) OR infan*) OR young*) OR preschool*) OR “pre school*”) OR pre-school*))))

AND (((((((((seretide) OR symbicort) OR advair) OR viani) OR flutiform))

OR (((((((((((glucocorticoids[MeSH Terms]) OR inhaled corticosteroid*) OR budesonide) OR beclomethasone) OR beclometasone) OR fluticasone) OR triamcinolone) OR flunisolide) OR ciclesonide))

AND (((((((((adrenergic beta 2 receptor antagonists[MeSH Terms]) OR ((((beta*) AND agonist*)) AND ((long-acting) OR “long acting”))) OR ((((beta*) AND adrenergic*)) AND ((long-acting) OR “long acting”))) OR ((bronchodilat*) AND ((long-acting) OR “long acting”))) OR salmeterol) OR serevent) OR *formoterol) OR foradil) OR vilanterol))))))

AND (“2015/02/01”[Date - Publication]: “2018/12/19”[Date - Publication])

Studies of LTRA as add-on to ICS

((((((((((clinical trial[MeSH Terms]) OR clinical trial) OR clinical study)))))

AND asthma[MeSH Terms]))

AND ((((((((forced expiratory volume) OR FEV)) OR ((exacerbation) OR worsening)) OR adverse event)))))

AND ((((((((((((((((((child*) OR paediat*) OR pediat*) OR adolesc*) OR infan*) OR young*) OR preschool*) OR “pre school*”) OR pre-school*))))

AND (((((((((((((glucocorticoids[MeSH Terms]) OR inhaled corticosteroid*) OR budesonide) OR beclomethasone) OR beclometasone) OR fluticasone) OR triamcinolone) OR flunisolide) OR ciclesonide))) AND ((((((((((((leukotriene antagonists[MeSH Terms]) OR LTRA) OR leukotriene*) OR leucotriene*) OR anti-leukotriene*) OR anti-leucotriene*) OR montelukast) OR singulair) OR zafirlukast) OR accolate) OR pranlukast) OR azlaire))))

AND (“2014/07/01”[Date - Publication]: “2018/12/19”[Date - Publication])

Studies of LAMA as add-on to ICS

(((((((((clinical trial[MeSH Terms]) OR clinical trial) OR clinical study)))))

AND asthma[MeSH Terms]))

AND ((((((((Asthma Control Questionnaire) OR ACQ)) OR ((forced expiratory volume) OR FEV)) OR ((exacerbation) OR worsening)) OR adverse event)))))

AND ((((((((((((((((((child*) OR paediat*) OR pediat*) OR adolesc*) OR infan*) OR young*) OR preschool*) OR “pre school*”) OR pre-school*))))

AND (((((((((((((glucocorticoids[MeSH Terms]) OR inhaled corticosteroid*) OR budesonide) OR beclomethasone) OR beclometasone) OR fluticasone) OR triamcinolone) OR flunisolide) OR ciclesonide))) AND (((((((((((((((((((((((muscarinic) AND antagonist*)))) AND (((long-acting) OR “long acting”)))))) OR ((antagonists, muscarinic[MeSH Terms]) AND (((long-acting) OR “long acting”)))))) OR LAMA) OR glycopyrronium) OR aclidinium) OR tiotropium) OR umeclidinium) OR NVA237) OR seebri) OR LAS34273) OR turdorza) OR pressair) OR eklira) OR genuair) OR spiriva) OR GSK573719)))

The literature searches were reviewed from the title, abstract or descriptors, and all studies that were not RCTs or that clearly did not fit the inclusion criteria were excluded. Data were analysed from the articles deemed appropriate for inclusion. Where appropriate, we performed a meta-analysis using the Cochrane statistical package RevMan 5, assuming equivalence if the risk ratio estimate and its confidence interval (CI) were between 0.9 and 1.1. The risk of bias was assessed using a domain-based evaluation, in line with recommendations provided in the Cochrane Handbook for Systematic Reviews of Interventions [23]. Various domains, including allocation concealment and blinding, were judged as being low, unclear or high. Studies were deemed to be of high methodological quality when the reported randomisation and blinding procedures were adequate and at a low risk of bias, with balanced group attrition.

Results

Identification of relevant articles

A literature search identified four systematic reviews (Fig. 2). Of these, one compared RCTs of LABAs as add-on to ICS, published up to February 2015, and was included in the review [24]. Three of the systematic reviews compared LTRAs with placebo as add-on to ICS. Of these, two were included in this review [25, 26], with the most recent studies published up to July 2014. One systematic review comparing LTRAs with placebo [27] was excluded as data from the included studies were already covered in the 2010 systematic reviews. No systematic reviews were identified that compared LAMAs with placebo, or LABAs, LTRAs or LAMAs directly with one another. We reviewed the three systematic reviews and analysed the relevant studies for inclusion in this review.

Fig. 2
figure 2

Study selection flow diagram. LABA, long-acting β2-agonist; LAMA, long-acting muscarinic antagonist; LTRA, leukotriene receptor antagonist

Full size image

Additional literature searches identified 73 articles, published since February 2015, comparing LABAs with placebo, of which two met the inclusion criteria for this review [28, 29]. Twenty-three articles published since July 2014 were identified comparing LTRA with placebo, of which one met the inclusion criteria for this review [30]. An additional 16 articles comparing LAMAs with placebo were identified, of which two met the inclusion criteria for this review [11, 31]. We also included two studies in which patients received tiotropium as add-on to ICS plus other controllers, which were not identified in the literature search as the search strings excluded additional controller medications to LAMA [13, 14]. There were no additional studies identified that compared LABAs, LTRAs or LAMAs directly with one another. In total, 29 studies were included in this review.

The designs of all included studies are summarised in Table 1. All studies were randomised, and most were double-blinded and parallel-group in design, ranging from 4 to 54 weeks in duration. Participants were 4–18 years of age. Primary outcomes included safety and lung function.

Table 1 Details of the trials included
Full size table

An overview of judgements on domains related to risk of bias is reported in Table 2. Most bias items were deemed to be of low or unclear risk.

Table 2 Risk of bias summary: review authors’ judgements about each risk of bias item for each included study
Full size table

FEV1 results

The LABA studies included in the Cochrane meta-analysis present a combination of peak and trough FEV1 measurements, and some articles do not specify at what time point the measurement was taken [24]. For this reason, we present both peak and trough FEV1 response data where available.

FEV1: absolute difference in litres

We performed a meta-analysis of nine LABA studies. There was a treatment difference in FEV1 of 0.07 L (95% CI 0.05, 0.08) (Fig. 3). Excluding the two outliers (a vilanterol study that found no improvement [− 0.06 to 0.02 L] [28] and a very small [n = 21] salmeterol study [0.42 L (95% CI 0.21, 0.63)] [46]), mean treatment differences were 0.04–0.13 L (Fig. 3). None of the included LTRA studies presented data for change from baseline in litres.

Fig. 3
figure 3

Treatment difference in FEV1 response between LABA added to ICS and ICS alone. BUD, budesonide; CI, confidence interval; FEV1, forced expiratory volume in 1 s; form, formoterol; ICS, inhaled corticosteroid; LABA, long-acting β2-agonist; VI, vilanterol

Full size image

For the LAMA studies, we pooled the data for studies where tiotropium was the only add-on therapy (no additional LABA add-on therapy permitted) (RubaTinA-asthma® and CanoTinA-asthma®) [11, 31] and presented both peak and trough results for tiotropium Respimat® 5 μg and 2.5 μg (Fig. 4). Peak FEV1 was defined as the maximum FEV1 within 3 h after dosing and trough FEV1 was defined as the pre-dose FEV1 measured 24 h after the previous drug administration and 10 min prior to the evening dose of the patient’s usual asthma medication. We did the same for studies where tiotropium Respimat® was the third or even fourth controller (PensieTinA-asthma® and VivaTinA-asthma®) (Fig. 4). None of the included studies investigated tiotropium delivered via the HandiHaler® device [13, 14].

Fig. 4
figure 4

Pooled treatment difference in peak (a) and trough (b) FEV1 response between tiotropium Respimat® and placebo added to ICS for patients with symptomatic moderate asthma and patients with symptomatic severe asthma. CI, confidence interval; FEV1, forced expiratory volume in 1 s; ICS, inhaled corticosteroid

Full size image

FEV1 improvements versus placebo with tiotropium Respimat® as add-on to ICS in studies of children and adolescents with symptomatic moderate asthma were 0.159–0.168 L for peak FEV1 and 0.105–0.118 L for trough FEV1 (Fig. 4). For studies in children and adolescents with symptomatic severe asthma, FEV1 improvements versus placebo were 0.074–0.117 L for peak FEV1 and 0.064–0.071 L for trough FEV1 (Fig. 4).

FEV1 response: percent predicted

The Cochrane analysis of LABA studies (Table 3) found an improvement in FEV1 percent predicted with LABAs added to ICS versus ICS of 2.99% (95% CI 0.86, 5.11; n = 534) [24]. Results from individual LABA studies are also detailed in Table 3. Improvements in peak FEV1 percent predicted with tiotropium added to ICS versus ICS were 4.07–7.70%, and 2.85–5.05% for trough FEV1; improvements with tiotropium added to ICS with other controllers were 1.64–6.33% for peak FEV1 and 0.83–3.85% for trough FEV1.

Table 3 Mean difference in FEV1% predicted
Full size table

The treatment difference with montelukast added to ICS compared with ICS alone varied, with the systematic review finding an improvement of 0.09% (95% CI − 0.07 to 0.25; n = 188) [25] and individual studies mostly ranging from 1.3 to 2.6%. One single-centre study found an improvement of 10.8% with montelukast compared with ICS, but this was a small, 4-week study (n = 24), and no confidence intervals or statistical comparison was available [50].

Exacerbations requiring OCS

The Cochrane analysis of LABA studies (n = 1669) found no difference in the risk of exacerbations requiring OCS between LABAs plus ICS compared with ICS alone (risk ratio 0.95; 95% CI 0.70, 1.28) (Table 4) [24]. The individual studies were quite variable, with study durations of 4–54 weeks. We found no additional studies reporting on exacerbations requiring OCS in our literature search.

Table 4 Exacerbations requiring oral corticosteroids
Full size table

Risk ratios were not available for the tiotropium studies, but the proportion of patients with exacerbations requiring OCS was low in all of the studies (Table 4). Tiotropium provided improvements in time to first exacerbation requiring OCS when added onto ICS versus placebo, with hazard ratios of 0.23–1.14, and 0.40–2.06 when added on to other controllers.

The systematic review of the LTRA studies showed no difference between montelukast and placebo on top of ICS, but the authors noted that there was evidence of statistical heterogeneity [25]. The network meta-analysis found no difference between montelukast and placebo (odds ratio 0.94; 95% CI 0.58, 1.45) [26]. One 7-month study found fewer exacerbations with montelukast than with placebo as add-on to ICS (odds ratio 0.26; 95% CI 0.09, 0.76) [30].

Adverse events and serious adverse events

The proportion of patients experiencing AEs or SAEs with the addition of LABA to ICS was broadly similar, with some variations in the proportion of patients with AEs or SAEs between studies (Table 5).

Table 5 AEs and SAEs
Full size table

There was no increase in the number of patients with AEs or SAEs with tiotropium compared with placebo as add-on to ICS or add-on to ICS plus other controllers (Table 5).

There were limited data on the number of patients with AEs in the montelukast analyses; the study that did report the proportion of patients with AEs showed no significant difference between montelukast and placebo as add-on to ICS (Table 5). There were insufficient data to make a comment on SAEs in the montelukast trials.

Efficacy and safety of tiotropium Respimat® as add-on to ICS and additional controller medications

In studies where tiotropium Respimat® was added onto ICS and additional controller medications (PensieTinA-asthma® and VivaTinA-asthma®) [13, 14], the effect size for both lung function and exacerbations requiring OCS was comparable with the studies where tiotropium was the only controller [11, 31], or where LABA or LTRA were added onto ICS [24,25,26, 28,29,30]. In addition, the studies demonstrated comparable safety with placebo [13, 14].

Discussion

In this literature review, the addition of once-daily tiotropium (with or without other controllers) and twice-daily LABAs to ICS in children and adolescents provided similar improvements in lung function [11, 13, 14, 24, 28, 29, 31], and greater improvements than with once-daily LABA vilanterol added onto ICS [28]. Data reporting on the effect of LTRAs as add-on to ICS on lung function were somewhat inconsistent, yet a previous systematic review found no improvement with montelukast compared with placebo when added to ICS [25], so it may be appropriate to suggest that twice-daily LABAs and tiotropium are more effective at improving lung function in adolescents and children as add-on to ICS. This assumption could be further clarified if future studies directly compared tiotropium, LABAs and LTRAs as add-on to ICS.

An additional endpoint that we analysed in this review was asthma exacerbations. However, the exacerbation data were more difficult to interpret, as the studies were of different durations and not necessarily powered to show a treatment difference in exacerbation frequency. Powering a study in paediatric patients to assess asthma exacerbations may present ethical considerations, with patients receiving placebo or care that is inconsistent with the best proven method, potentially being exposed to unnecessary risk and harm, especially where exacerbation events are expected [52]. In addition, not all studies included a risk ratio, making the comparison of data difficult. However, in the tiotropium trials, where exacerbations were included as a safety endpoint, it was possible to demonstrate that tiotropium provided a reduction in the risk of exacerbations requiring OCS when added onto ICS, either alone or with additional controller treatments, compared with placebo [11, 13, 14, 31]. Although the results from the individual studies of LABA as add-on to ICS varied, the previously published Cochrane review by Chauhan et al. suggested that LABAs and placebo have a comparable risk of asthma exacerbation [24]. In regards to the effect of LTRAs on asthma exacerbations, the data were more inconclusive. The one RCT included on LTRAs reported that montelukast reduced the risk of exacerbations compared with placebo. However, the sample size was small, with only 76 participants [30]. The two systematic reviews reported no reduction in the risk of exacerbations compared with placebo; however, the width of the CIs suggests a large spread of data [25, 26]. It could therefore be suggested that the highest quality of evidence was for the trials investigating LABA or LAMA as add-on to ICS.

The safety data showed no increase in the proportion of patients reporting AEs or SAEs with LABAs or with tiotropium when added to ICS [11, 13, 14, 24, 28, 29, 31]. The available data for LTRAs were limited, but suggested no increase in the proportion of patients with AEs with montelukast compared with placebo as add-on to ICS [49]. However, it should be noted that previous post-marketing studies have suggested that paediatric patients receiving montelukast are more likely to report neuropsychiatric AEs than those receiving ICS [53, 54]. Therefore, the results from this review indicate that LABAs, LTRAs and LAMAs all have a comparable safety profile to placebo, but other real-world and post-marketing evidence should also be considered.

This literature review aims to provide an up-to-date overview of the efficacy and safety of three classes of drugs that are options for adding onto ICS in adolescents and children with asthma. The strength of the study is that this is the first literature review and meta-analysis to collate and compare the efficacy and safety of LABAs, LTRAs and LAMAs in children and adolescents in one review. Previous reviews have compared the efficacy and safety of LABAs and LAMAs, or LABAs and LTRAs, in adolescents aged over 12 years and in adults, but none has compared all three therapeutic options in one review, and none has done so for this patient population in children and adolescents aged 4–17 years.

We have focused on a limited number of endpoints that are considered important in the treatment of asthma such as lung function, exacerbations and AEs. However, there is considerable variability in the methodology and definition of these endpoints between studies, making the comparison of data more difficult. There were only a limited number of montelukast studies in children that met the inclusion criteria, so LTRA data are lacking for some endpoints. For example, for the LABA studies, we were able to perform a meta-analysis of absolute change in lung function in litres, but LTRA studies only reported lung function change in percent predicted. Moreover, when extracting the FEV1 data from the various studies, the time point of the measurement in relation to drug administration (i.e. peak/trough) was not always clear. Only the LAMA studies reported whether FEV1 was peak (defined as the maximum FEV1 within 3 h after dosing) or trough FEV1 (defined as the pre-dose FEV1 measured 24 h after the previous drug administration and 10 min prior to the evening dose of the patient’s usual asthma medication). As Fig. 4 demonstrates, there are differences between the responses depending on when the measurement is taken, with peak FEV1 (Fig. 4a) values higher than the equivalent trough FEV1 (Fig. 4b) values. Therefore, it is possible that some of the between-study differences in FEV1 response for LABAs and LTRAs may be attributable to the time point at which the measurement was taken, but this cannot be confirmed.

In light of the extension of the tiotropium label and the most recent treatment guidelines for children with asthma [4], the results provide support for the use of tiotropium as add-on therapy in adolescents and children with asthma aged 4–17 years. The results are in agreement with those of a recently published systematic review that compared LABAs with LAMAs in patients aged over 12 years [22]. The authors reported that use of LAMA as add-on to ICS was associated with a lower risk of asthma exacerbations compared with placebo, and had a comparable benefit to LABA on lung function. The authors note that their review was designed and conducted in patients aged 12 years and over because tiotropium was not approved in patients aged less than 12 years at the time the study was undertaken [22]. In addition, it does not review the literature on LTRAs as an add-on treatment.

In conclusion, tiotropium and LABAs have similar efficacy, and provide greater improvements in lung function than montelukast as add-on to ICS in children and adolescents with asthma. All three controller options have comparable safety profiles. The results of our literature review in patients aged 4–17 years provide needed additional information, and further supports the use of tiotropium in children and adolescents with asthma. The clinical decision on the preferred add-on therapy should also take into account patient phenotype and comorbidities, dose regimen and frequency, the availability of combination therapy, and the delivery device, although more research is required in these younger age groups.