Abstract

Asthma patients who continue to experience symptoms despite being on regular inhaled corticosteroids (ICS) represent a management challenge. Long-acting beta2-agonists (LABA) or anti-leukotrienes (LTRA) are two treatment options that could be considered as add-on therapy to ICS.ObjectivesWe compared the efficacy and safety profile of adding either daily LABA or LTRA in adults and children with asthma who remain symptomatic on ICS.Search strategyWe searched the Cochrane Airways Group Specialised Register (up to and including March 2010). We consulted reference lists of all included studies and contacted authors and pharmaceutical manufacturers for other published or unpublished studies.Selection criteriaWe included randomised controlled trials (RCTs) conducted in adults or children with recurrent asthma that was treated with ICS and where a fixed dose of a long-acting beta2-agonist or leukotriene agent was added for a minimum of 28 days.Data collection and analysisTwo authors independently assessed the risk of bias of included studies and extracted data. We sought unpublished data and further details of study design, where necessary.Main resultsWe included 17 RCTs (7032 participants), of which 16 recruited adults and adolescents (6850) and one recruited children aged 6 to 17 years (182). Participants demonstrated substantial reversibility to short-acting beta-agonist at baseline. The studies were at a low risk of bias. The risk of exacerbations requiring systemic corticosteroids was lower with the combination of LABA and ICS compared with LTRA and ICS, from 11% to 9% (RR 0.83, 95% CI 0.71 to 0.97; six studies, 5571 adults). The number needed to treat (NNT) with LABA compared to LTRA to prevent one exacerbation over 48 weeks was 38 (95% CI 22 to 244). The choice of LTRA did not significantly affect the results. The effect appeared stronger in the trials using a single device to administer ICS and LABA compared to those using two devices. In the absence of data from the paediatric trial and the clinical homogeneity of studies, we could not perform subgroup analyses. The addition to ICS of LABA compared to LTRA was associated with a statistically greater improvement from baseline in several of the secondary outcomes, including lung function, functional status measures and quality of life. Serious adverse events were more common with LABA than LTRA, although the estimate was imprecise (RR 1.35, 95% CI 1.00 to 1.82), and the NNT to harm for one additional patient to suffer a serious adverse event on LABA over 48 weeks was 78 (95% CI 33 to infinity). The risk of withdrawal for any reason in adults was significantly lower with LABA and ICS compared to LTRA and ICS (RR 0.84, 95% CI 0.74 to 0.96).Authors' conclusionsIn adults with asthma that is inadequately controlled on low doses of inhaled steroids and showing significant reversibility with beta2-agonists, LABA is superior to LTRA in reducing oral steroid treated exacerbations. Differences favouring LABA in lung function, functional status and quality of life scores are generally modest. There is some evidence of increased risk of SAEs with LABA. The findings support the use of a single inhaler for the delivery of LABA and inhaled corticosteroids. We are unable to draw conclusions about which treatment is better as add-on therapy for children.PLAIN LANGUAGE SUMMARYWhat are the effects of long-acting beta2-agonists compared with anti-leukotrienes when added to inhaled steroids?People who continue to experience asthma symptoms despite regularly taking inhaled corticosteroids are a challenge for management. It is not clear whether the addition of a long-acting beta2-agonist (LABA) such as formoterol or salmeterol would provide more benefit in comparison with an oral anti-leukotriene agent (LTRA), for example zafirlukast or montelukast.Seventeen trials (16 in adults and one in children) were included in this review and were of good quality. We found that the addition of a LABA provides significantly greater protection against exacerbations requiring oral steroids when compared with a LTRA for adults. Based on the results of our analyses, approximately 38 adults (with a range of between 22 and 244) would need to be treated with a LABA rather than a LTRA for 48 weeks to prevent one experiencing an exacerbation needing a course of oral steroids. The trial on children did not contribute data on the main outcome and therefore we could not draw any conclusions for children.LABAs also led to a greater improvement in lung function, improvement in symptoms, use of rescue medication, quality of life and symptoms compared to the use of LTRAs. The magnitude of the improvements was modest. Serious adverse events were more frequent with LABA than with LTRAs although this result was imprecise. Based on our analyses, around 78 people would need to be treated for 48 weeks with a LABA rather than a LTRA for one of them to experience a serious adverse event. However, due to the lack of precision around our result, the true number could be between 33 and infinity. There are currently insufficient data to draw any conclusions about the effects of these drugs in children.