Background: Epidemiological evidence has suggested a link between beta2-agonists and increased asthma mortality. There has been much debate about possible causal links for this association, and whether regular (daily) long-acting beta2-agonists are safe. This is an updated systematic review.
Objectives: To assess the risk of mortality and non-fatal serious adverse events in trials which randomised patients with chronic asthma to regular salmeterol and inhaled corticosteroids in comparison to the same dose of inhaled corticosteroids.
Search methods: We identified randomised trials using the Cochrane Airways Group Specialised Register of trials. We checked websites of clinical trial registers for unpublished trial data. Food and Drug Administration (FDA) submissions in relation to salmeterol were also checked. The date of the most recent search is August 2012.
Selection criteria: We included parallel design controlled clinical trials on patients of any age and severity of asthma if they randomised patients to treatment with regular salmeterol and inhaled corticosteroids (in separate or combined inhalers), and were of at least 12 weeks duration.
Data collection and analysis: We conducted the review according to standard procedures expected by the Cochrane Collaboration. We obtained unpublished data on mortality and serious adverse events from the sponsors, and from FDA submissions. We assessed the quality of evidence according to GRADE recommendations.
Main results: We have included 35 studies (13,447 participants) in adults and adolescents, and 5 studies (1862 participants) in children in this review. We judged that the overall risk of bias was low, and we obtained data on serious adverse events from all studies. All except 542 adults (and none of the children) who were randomised to salmeterol were given fluticasone in the same (combination) inhaler.Seven deaths occurred in 6986 adults on regular salmeterol with inhaled corticosteroids (ICS), and seven deaths in 6461 adults on regular inhaled corticosteroids at the same dose. The difference was not statistically significant (Peto odds ratio (OR) 0.90; 95% confidence interval (CI) 0.31 to 2.60, moderate quality evidence). The risk of dying from any cause in adults on ICS was 10 per 10,000, and on salmeterol and ICS we would expect between 3 and 26 deaths per 10,000. No deaths were reported in 1862 children, and no deaths were reported to be asthma-related in adults or children.Non-fatal serious adverse events of any cause were reported in 167 adults on regular salmeterol with ICS, compared to 135 adults on regular ICS; again this was not a statistically significant increase (Peto OR 1.15; 95% CI 0.91 to 1.44, moderate quality evidence). The frequency of serious adverse events was 21 per 1000 in the adults treated with ICS and 24 per 1000 in those treated with salmeterol and ICS. The absolute difference in the risk of non-fatal serious adverse events was an increase of 3 per 1000, that was not statistically significant (risk difference (RD) 0.003; 95% CI -0.002 to 0.008).There were 6 of 930 children with serious adverse events on regular salmeterol with ICS, compared to 5 out of 932 on regular ICS: there was no significant difference between treatments (Peto OR 1.20; 95% CI 0.37 to 3.91, moderate quality evidence).Asthma-related serious adverse events were reported in 29 and 23 adults in each group respectively, a non-significant difference (Peto OR 1.12; 95% CI 0.65 to 1.94, moderate quality evidence), and only 1 asthma-related event was reported in children in each treatment group.
Authors' conclusions: We found no statistically significant differences in fatal or non-fatal serious adverse events in trials in which regular salmeterol was randomly allocated with ICS, in comparison to ICS alone at the same dose. Although 13,447 adults and 1862 children have now been included in trials, the frequency of adverse events is too low and the results are too imprecise to confidently rule out a relative increase in all cause mortality or non-fatal adverse events with salmeterol used in conjunction with ICS. However, the absolute difference between groups in the risk of serious adverse events was very small. We could not determine whether the increase in all cause non-fatal serious adverse events reported in the previous meta-analysis on regular salmeterol alone is abolished by the additional use of regular ICS. We await the results of large ongoing surveillance studies mandated by the FDA to provide more information. There were no asthma-related deaths and few asthma-related serious adverse events. Clinical decisions and information for patients regarding regular use of salmeterol have to take into account the balance between known symptomatic benefits of salmeterol and the degree of uncertainty and concern associated with its potential harmful effects.