Asthma is a serious global health issue that affects all age groups, with a reported 339 million sufferers worldwide, presenting a number of challenges for primary care physicians1. Many patients with asthma remain symptomatic, despite treatment, for multiple different reasons2,3,4,5,6. It has been suggested that patients may overestimate and thus inaccurately report their level of disease control, because they accept and tolerate a certain level of symptoms, assuming them to be an inevitable consequence of asthma7,8. Physicians may underestimate the prevalence and severity of symptoms and overestimate the degree to which the patient’s asthma is controlled, meaning the patient may not receive adequate medication to achieve control of their disease2,7. Reducing asthma symptoms and future risk through correct add-on therapy and management in patients who remain uncontrolled despite treatment is a major challenge for those working in both secondary and primary care. The Global Initiative for Asthma (GINA) strategy recommends a stepwise approach to asthma management in order to achieve symptom control and prevent future risks, including exacerbations, loss of lung function, and side effects of medication (Fig. 1)9. Inhaled corticosteroids (ICS) are considered an effective long-term controller treatment in the management of asthma10. However, if asthma remains uncontrolled despite medium-dose ICS, increasing the dose of ICS may not be appropriate due to an increased risk of local and systemic side effects and variation in individual ICS dose-responsiveness between patients. In addition, most of the clinical benefit of ICS use is seen at low doses. Add-on treatments may therefore be required9. In addition, evidence suggests that the ICS dose–response curve is relatively flat, with 80–90% of the maximum achievable therapeutic effect in adult asthma obtained at 200–250 μg of fluticasone propionate or equivalent (Fig. 2); therefore, addition of an add-on therapy may be considered to be a more effective and safer treatment strategy11,12,13.

Fig. 1: GINA treatment recommendations for patients aged ≤5 years, 6–11 years and ≥12 years9.
figure 1

© 2020, Global Initiative for Asthma, reproduced with permission. FEV1 forced expiratory volume in 1s, GINA Global Initiative for Asthma, ICS inhaled corticosteroid, Ig immunoglobulin, IL interleukin, LABA long-acting β2-agonist, LTRA leukotriene receptor antagonist, OCS oral corticosteroid, SABA short-acting β2-agonist.

Fig. 2: Schematic dose–response curves for different outcomes for efficacy and adverse effects with inhaled corticosteroids, expressed as fluticasone propionate in µg/day.
figure 2

Reprinted with permission from the American Thoracic Society. Copyright © 2020 American Thoracic Society. Beasley et al.12. FEV1 forced expiratory volume in 1 s, PEF peak expiratory flow.

Long-acting β2-agonist (LABA; e.g. salmeterol, formoterol, vilanterol, indicaterol14,15,16), leukotriene receptor antagonist (LTRA; e.g. montelukast and zafirlukast [discontinued]15,17) and long-acting muscarinic antagonist (LAMA; tiotropium18 [the only LAMA currently indicated for use in patients with asthma]) add-on controller therapies have been shown to improve lung function and asthma control and reduce exacerbations in asthma patients, have safety profiles similar to placebo, and are currently indicated for use in patients with asthma. More information on these classes of agents, including indications and their modes of action, is detailed in Table 1.

Table 1 Drug names, indications and mode of action of LABAs, LTRAs and LAMAs.

With multiple add-on therapies available for the management of asthma, there have been several systematic reviews published that evaluate the efficacy and safety of add-on therapies compared with either placebo or another add-on therapy14,15,16,17,19,20. However, none compare LABA, LTRA and LAMA as add-on treatments to ICS in a single consolidated review, and there are no head-to-head trials evaluating all three treatments within the same trial. Here we systematically analyse and review the literature to explore the challenges of asthma management, the impact of poor asthma control on patients’ lives and compare outcomes from published studies. We examine the effect of three add-on treatments on lung function, asthma control, exacerbations and safety, with the aim of assisting primary care physicians in selecting the most appropriate add-on treatment to ICS.


Search results

The literature search identified 14 relevant publications that met the inclusion criteria for this review: 2 Cochrane reviews and 12 additional randomised controlled trials (RCTs) that were not included within the Cochrane reviews.

  • The search strings for LABA studies generated 164 publications, of which 1 meta-analysis and 4 additional RCTs met the criteria for inclusion in this review16,21,22,23.

  • The search strings for LTRA studies generated 54 publications, of which 3 RCTs met the criteria for inclusion in this review24,25,26.

  • The search strings for LAMA studies generated 106 publications, of which 8 RCTs met the criteria for inclusion in this review21,22,27,28,29,30,31.

Asthma control in adult patients

LABA (salmeterol) significantly improved asthma control when added to ICS compared with placebo (measured by Asthma Control Questionnaire [ACQ])21. The LTRA (montelukast) did not have an effect on asthma control when compared with placebo (measured by ACQ) (Table 2)32. Data for LAMA (tiotropium) are more varied (Table 2)21,27,28,29,30,31. Paggiaro et al. reported that there was no difference of effect between tiotropium (5 µg and 2.5 µg) and placebo on ACQ score. Four papers compared the effect of LABAs directly with tiotropium. There was no significant difference between LABAs and tiotropium on asthma control as measured by ACQ (Table 2)21,33,34. One study included difference in Mini-Asthma Quality of Life Questionnaire (Mini-AQLQ) response scores as a secondary efficacy endpoint. At study endpoint at 16 weeks, salmeterol (50 µg) significantly improved overall Mini-AQLQ score compared with placebo, but there was no significant difference in response scores between the tiotropium (5 µg) and placebo groups. When directly compared, there was no difference in treatment response between salmeterol (50 µg) and tiotropium (5 µg) at study endpoint22. No studies compared asthma control, measured by ACQ, in LABA vs LTRA or tiotropium vs LTRA.

Table 2 ACQ-7 responder rates, exacerbations and AEs.

Lung function in adult patients

LABAs significantly improved forced expiratory volume in 1 s (FEV1), morning and evening peak expiratory flow (PEF) (salmeterol/formoterol) and forced vital capacity (FVC) (salmeterol) compared with placebo or ICS alone (Figs. 35)16,21,22,23. The literature reporting the effect of LTRAs as add-on to ICS on lung function is varied (Figs. 35), suggesting no beneficial effect of montelukast on FEV1 or FVC compared with ICS alone24. Results for the effect of montelukast on PEF are conflicting. Only one study of zafirlukast (now discontinued) was identified, with the authors reporting significant improvements in FEV1 and both morning and evening PEF compared with placebo26.

Fig. 3: Mean difference in FEV1.
figure 3

ALAACRC American Lung Association Asthma Clinical Research Centers, AUC area under curve, BID twice daily, CI confidence interval, FEV1 forced expiratory volume in 1 s, LABA long-acting β2-agonist, LAMA long-acting muscarinic agonist, LTRA leukotriene receptor antagonist, NR not reported, NS non-significant, QD once daily.

Fig. 4: Mean difference in PEF.
figure 4

BID twice daily, CI confidence interval, LABA long-acting β2-agonist, LAMA long-acting muscarinic agonist, LTRA leukotriene receptor antagonist, NR not reported, NS non-significant, PEF peak expiratory flow, QD once daily.

Fig. 5: Mean difference in FVC.
figure 5

ALAACRC American Lung Association Asthma Clinical Research Centers, AUC area under curve, BID twice daily, CI confidence interval, FVC forced vital capacity, LABA long-acting β2-agonist, LAMA long-acting muscarinic agonist, LTRA leukotriene receptor antagonist, NR not reported, NS non-significant, QD once daily.

Of the seven available studies of LAMAs (tiotropium), all but two trial arms from two studies reported significant improvements in FEV1 (peak, trough and area under the curve [AUC]) and FVC compared with placebo (Figs. 3, 5)21,22,27,28,29,30,31. In all the published studies in adults21,22,27,28,29,30,31, tiotropium significantly improved morning and evening PEF compared with placebo or ICS alone, except for one tiotropium 2.5 µg trial arm, which reported no significant difference between tiotropium and placebo on trough PEF response (Fig. 4)29.

A systematic review comparing the use of LABAs with LTRAs as add-on to ICS reported that LABAs (salmeterol or formoterol) have a significantly greater effect on FEV1 and PEF compared with LTRAs (montelukast or zafirlukast) (Figs. 3, 4)15. There are no data available comparing the effect of LABAs and LTRAs on FVC. Four papers compared the effect of tiotropium with the LABAs salmeterol and formoterol on lung function parameters (Figs. 3, 4)33.

Exacerbations in adult patients

LABAs provided numerical improvement in the number of patients reporting at least one exacerbation compared with placebo (Table 2)21,22,23. Data for LTRAs (montelukast and zafirlukast [now discontinued]) are varied (Table 2). LAMA (tiotropium) both significantly and non-significantly reduced the number of patients experiencing at least one exacerbation (Table 2)27. A meta-analysis comparing LABA with LTRA reported a 2% reduction in risk of exacerbations in patients using LABA+ICS vs LTRA+ICS combination therapy (Table 2)15. Both tiotropium and LABAs had a comparable effect on the risk of exacerbations21,22,33,34.

Safety in adult patients

Overall, comparable proportions of patients report adverse events (AEs) with LABA, LTRA and tiotropium treatment as add-on to ICS with both placebo and with one another (Table 2)15,21,22,23,25,26,27,28,29,30,31,34.

Efficacy and safety of LABAs, LTRAs and LAMAs in paediatric patients

A recent systematic review by Vogelberg et al. compared the efficacy and safety of LABAs, LTRAs and LAMAs (tiotropium) in paediatric patients aged 4–17 years with asthma35. LABA treatment as add-on to ICS improved lung function when compared with placebo, as measured by FEV1 and FEV1 % predicted. There was no difference in risk of exacerbations requiring oral corticosteroid (OCS) between LABAs plus ICS compared with ICS alone, although it should be noted that not all trials were powered to assess exacerbations. The proportion of patients experiencing AEs or serious AEs (SAEs) with the addition of LABA to ICS was broadly similar35. An additional RCT of 512 patients aged 5–12 years with persistent asthma reported improvements in lung function and asthma control, and no differences in risk of exacerbations and AEs, in patients receiving LABAs (formoterol) compared with those receiving placebo as add-on to ICS36. However, in a systematic review comparing LABA plus ICS vs higher-dose ICS in children with asthma, combination therapy led to a trend towards an increased risk of oral steroid-treated exacerbations and hospital admissions16.

For LTRA (montelukast), a study by Simons et al. described a greater improvement from baseline FEV1 in patients receiving montelukast compared with placebo37. In addition, a systematic review found an improvement in baseline FEV1 and FEV1 % predicted in patients receiving ICS plus montelukast compared with those receiving ICS plus placebo, but these differences were not significant19. There was no difference between montelukast and placebo as add-on to ICS in the risk of exacerbations19. Limited available data suggest that the proportion of patients experiencing AEs with the addition of montelukast to ICS is comparable with those receiving placebo as add-on to ICS37.

Tiotropium improved FEV1 and FEV1 % predicted as add-on to ICS with or without additional controller therapies21,27,28,38,39,40,41,42. The proportion of paediatric patients with exacerbations requiring OCS was low in all studies included within the review by Vogelberg et al.35. The review authors also concluded that 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 controllers15,21,22,23,25,26,27,28,29,30,31,34. An additional study of 102 patients aged 1–5 years with persistent asthma symptoms reported similar findings, with the number of patients reporting AEs similar in those who received tiotropium as add-on to ICS to those who received placebo as add-on42.

There were fewer published studies on the efficacy and safety of LABAs, LTRAs and LAMAs as add-on to ICS in patients aged <5 years compared with studies in older age groups42,43,44,45. An RCT of 12 patients with asthma aged 2–5 years reported that LABA (formoterol) as add-on to ICS provided rapid and sustained bronchodilation for ≥8 h compared with placebo45. A 12-week RCT of 689 patients with persistent asthma (≥3 episodes of asthma symptoms during the previous year) aged 2–5 years reported that LTRA (montelukast) as add-on to ICS (in at least 50% of participants) improved multiple parameters of asthma control, including daytime and overnight asthma symptoms and the percentage of days without asthma symptoms or asthma compared with placebo. There were no reported differences in the frequency of reported AEs44. Similarly, Bisgaard et al. reported that, in patients aged 2–5 years with intermittent asthma, montelukast significantly reduces the rate of asthma exacerbations and delayed the median time to first exacerbation compared with placebo over 12 months. However, patients in this trial did not receive montelukast as add-on to ICS43. A 12-week RCT of 102 children aged 1–5 years by Vrijlandt et al. reported that tiotropium as add-on to ICS with or without additional controller medications was associated with fewer reported AEs or asthma exacerbations compared with placebo. There was no significant difference in adjusted weekly mean combined daytime asthma symptom score between baseline and Week 12 between the tiotropium and placebo groups42.


The long-term aims of asthma management are symptom control, reduction of the future risk of exacerbations and airflow limitation, while at the same time minimising treatment side effects9. Although major advances have been made in asthma treatment and management, there still remain many patients who have poor asthma control and maintain the potential risk of worsening of their symptoms, as well as an increased risk of exacerbations, and unscheduled urgent and emergency care visits and hospitalisations46,47. For adults, adolescents and children, there is a need for effective add-on treatments as an alternative to increasing the ICS dose alone, as long-term, high-dose ICS use is associated with an increased risk of side effects9.

The findings from this literature review suggest that LABAs, LTRAs and tiotropium have similar safety profiles in both adult and paediatric populations (Table 2). Therefore, comparing the reported efficacy of the three add-on treatments in each patient population could assist with decision-making. Greater improvements in lung function have been reported with LABAs and LAMAs vs LTRAs in adults (Figs. 35). In addition, there appears to be greater improvements in asthma control and exacerbations with LABAs and LAMAs as add-on therapies than with LTRAs in this population.

Much of the available evidence for asthma management is based on research carried out in adults, which leads to a greater restriction of licensing of medication in children48, creating additional difficulties in selecting the most appropriate treatment option for paediatric patients with asthma49. Despite advances in care, asthma still presents a burden within this population, with many children remaining symptomatic and uncontrolled50. Data in patients aged <18 years are currently limited due to inherent difficulties in the study of this population; however, available evidence suggests that LABAs and tiotropium have comparable effects with respect to lung function, asthma control and exacerbations. Of the LABAs, LTRAs and LAMAs reviewed here, the LTRA montelukast is the only add-on treatment that is indicated for use in patients aged <4 years (as a chewable tablet51) and the only add-on treatment recommended for use in patients aged ≤5 years as an optional controller treatment9. Clinical trials of LTRAs in children aged ≤5 years have not demonstrated any safety concerns52. However, in 2020, the U.S. Food and Drug Administration (FDA) determined that a boxed warning for the LTRA montelukast was appropriate due to the risk of mental health side effects, and advised that healthcare professionals (HCPs) consider the benefits and risks of mental health side effects before prescribing montelukast53. Despite less published evidence regarding use of add-on therapies in paediatric patients (aged <18 years) than in adult patients (aged ≥18 years), current available data suggest that all three add-on therapies have comparable safety profiles, with LABAs and LAMAs providing greater improvements in lung function than LTRAs. When selecting the most appropriate add-on therapy for paediatric patients, it is important to consider the reported efficacy, safety data and subsequent post-marketing safety warnings (if applicable) and the indications of these add-on therapies, as not all are appropriate for all age ranges (Table 1).

When stepping up asthma therapy and considering add-on therapy, it is important to review the options available, to involve patients in decisions about their treatment and to keep a dialogue between patients and HCPs46. An up-to-date individualised asthma action plan can help to keep a record of any attempted treatment approaches and help the patient to self-manage54. The action plans should be discussed and agreed with patients and reviewed at regular intervals to make sure that they remain up to date and are fit for purpose54.

Poor asthma control leads to unfavourable outcomes, more frequent exacerbations, irreversible loss of lung function and even asthma-related deaths. Add-on therapy with LABA, LAMA or LTRA should be considered when asthma symptoms remain uncontrolled with at least medium–high ICS. Primary care practitioners can and should regularly assess symptom control following assessment of adherence, triggers, device technique and comorbidities. Therapy should be stepped up as recommended in GINA guidelines in order to attain optimal control, considering individual symptoms, lung function, comorbidities, inhaler technique, adherence and patient preference as important parameters for a personalised choice.

In conclusion, in adults, LAMAs and LABAs appear to provide a greater improvement in lung function than LTRAs as add-on to ICS, although there are no individual studies that directly compare LAMAs with LTRAs. LAMAs appear to be an effective alternative to LABAs for attaining asthma control, optimising lung function and preventing exacerbations, with a possible higher lung function benefit of LAMAs compared with LABAs. Data in patients aged <18 years are currently limited due to inherent difficulties in the study of this population. Current available evidence from clinical trials suggests that LABAs, LTRAs and LAMAs have comparable safety profiles, with LABAs and LAMAs providing greater improvements in lung function than LTRAs, yet it should be noted the FDA have advised a boxed warning for the LTRA montelukast due to risk of mental health side effects. Asthma should be treated in accordance with current guidelines, with regular checks made to ensure symptoms are controlled, as well as ensuring optimal strategies are in place to prevent exacerbations and achieve best lung function. If control is not achieved, treatment should be stepped up, ensuring that factors that may influence control (such as adherence, administration technique, allergic triggers and comorbidities) are addressed for each individual patient.


Our literature search was conducted in PubMed. For the comparison of add-on therapies, we identified RCTs and meta-analyses that compared the LAMA tiotropium with LABAs or LTRAs (directly or with placebo) as add-on to ICS in patients with asthma.

Data from RCTs of ≥4 weeks’ duration in all patients with asthma of all age groups, reporting change in FEV1, asthma control, exacerbations and AEs were included. Data were extracted from published manuscripts and publicly available online data. We checked the reference lists of the systematic reviews for references with any additional data for endpoints that were not described in the systematic reviews and to ensure that all trials met the inclusion criteria77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116. Search strings are detailed in Supplementary Methods.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.