Long-acting Beta-Agonists with and without Inhaled Corticosteroids and Catastrophic Asthma Events

Article Outline

• Abstract
• Materials and Methods
  • Trial Inclusion
  • Outcome Measures
  • Statistical Analysis
  • Role of the Funding Source
• Results
  • Search Results
  • Trial Characteristics
  • Asthma-related Intubations and Deaths
• Discussion
• Study Limitations
• Conclusions
• Acknowledgment
• References
• Copyright

Abstract 

Background

It is unclear whether long-acting β-agonists with concomitant inhaled corticosteroids increase asthma-related intubations and deaths. We pooled data on long-acting β-agonists with variable and concomitant inhaled corticosteroids to evaluate the risk for catastrophic asthma events.

Methods

We conducted searches of electronic databases, the US Food and Drug Administration website, clinical-trials registries, and selected references through December 2008. We analyzed randomized controlled trials in patients with asthma, which lasted at least 3 months, evaluated long-acting β-agonists compared with placebo or long-acting β-agonists with inhaled corticosteroids compared with corticosteroids alone, and included at least 1 catastrophic event, defined as asthma-related intubation or death.

Results

In pooled trial data that included 36,588 participants, long-acting β-agonists increased catastrophic events 2-fold (Peto odds ratio [OR] 2.10; 95% confidence interval [CI], 1.37-3.22). Statistically significant increases were seen for long-acting β-agonists with variable corticosteroids compared with placebo (OR 1.83; 95% CI, 1.14-2.95) and for concomitant treatment with corticosteroids compared with corticosteroids alone (OR 3.65; 95% CI, 1.39-9.55). Similar increases in risk were seen for variable and concomitant corticosteroid use, salmeterol and formoterol, and children and adults. When the analysis was restricted to trials with controlled corticosteroid use, given as part of the study intervention, concomitant treatment still increased catastrophic events compared with corticosteroids alone (OR 8.19; 95% CI, 1.10-61.18).

Conclusion

Long-acting β-agonists increase the risk for asthma-related intubations and deaths, even when used in a controlled fashion with concomitant inhaled corticosteroids.

Keywords: Asthma, Inhaled corticosteroids, Intubation, Long-acting beta-agonists, Meta-analysis, Mortality

There has been growing concern about asthma-related morbidity and mortality associated with the long-acting β-agonists salmeterol and formoterol, given with or without concomitant inhaled corticosteroids.¹, ², ³ Pooled trial data have consistently shown that long-acting β-agonists, when given with variable inhaled corticosteroid use, increase the risk for asthma-related hospitalizations, intubations, and deaths.⁴, ⁵, ⁶, ⁷, ⁸ In July 2005 an advisory committee to the US Food and Drug Administration (FDA) concluded that Boxed Warnings of an increased risk for asthma-related mortality should be placed on all products containing long-acting β-agonists, with recommendations for treatment only after other asthma drugs have failed.¹

The FDA subsequently requested data from sponsors of long-acting β-agonists (GlaxoSmithKline, Brentford, London, England; AstraZeneca, Wilmington, Del; and Novartis, Basel, Switzerland) on asthma-related intubations and deaths that occurred during published and unpublished randomized trials, as of January 2008.⁹, ¹⁰, ¹¹ Several meta-analyses have been performed, most with the cooperation or sponsorship of the pharmaceutical industry, which evaluated long-acting β-agonists with concomitant inhaled corticosteroids.⁷, ⁸, ⁹, ¹⁰, ¹¹, ¹², ¹³, ¹⁴, ¹⁵ These analyses consistently reported more asthma intubations and deaths for combined treatment compared with inhaled corticosteroids alone, but none reached statistical significance. However, each of these studies restricted their analyses to a subset of the total available data, for example, including only trials with salmeterol,⁹, ¹⁴ formoterol,⁷, ¹¹, ¹³ adults,¹² asthma deaths,⁷ published reports of events,¹⁵ single drug sponsors,⁷, ⁹, ¹⁰, ¹¹, ¹⁴ or FDA-approved dosages⁸, ¹⁴, ¹⁵ of long-acting β-agonists.

In December 2008 a follow-up FDA advisory committee meeting concluded that the risks of salmeterol (Serevent, GlaxoSmithKline) and formoterol (Foradil, Novartis) outweighed the benefits and should be banned for use in asthma for all ages.¹⁶ The committee separately evaluated long-acting β-agonists combined with a corticosteroid in a single inhaler, such as salmeterol with fluticasone (Advair, GlaxoSmithKline) and formoterol with budesonide (Symbicort; AstraZeneca), and concluded that further safety studies were needed to assess risk.¹, ¹⁶

The objective of this meta-analysis is to pool all the available data on long-acting β-agonists with variable and concomitant inhaled corticosteroids, evaluating the risk for asthma-related intubations or deaths. Subgroup analyses will compare results between variable and concomitant corticosteroids, salmeterol and formoterol, adults and children, and fatal and nonfatal events.

Back to Article Outline

Materials and Methods 

Trial Inclusion 

We performed a search of the MEDLINE, EMBASE, and Cochrane databases; the US FDA website; clinical-trials registries of drug manufacturers; and previous meta-analyses to identify trials on long-acting β-agonist use in patients with asthma published through December 2008. Studies were included if they were randomized controlled trials of long-acting β-agonists compared with placebo or long-acting β-agonists with inhaled corticosteroids compared with an equal or higher dose of inhaled corticosteroids alone of at least 3 months duration that reported at least 1 asthma-related intubation or death. We included reports of events that were provided by the investigator in the published trial or by the drug manufacturer in a subsequent account. In addition, we included pooled trial data from GlaxoSmithKline, because we were unable to obtain individual trial-level information for those events.

Outcome Measures 

Two reviewers independently extracted data from the selected articles, reconciling differences by consensus. Outcomes assessed were catastrophic asthma events, defined as asthma-related intubations or deaths that were reported by the investigator or the industry sponsor. Attempts were made to contact investigators of previous meta-analyses and the industry sponsors to obtain additional information concerning trials and events.

Statistical Analysis 

The proportions of patients with a catastrophic event to those without events from each trial were pooled using the fixed-effects method expressed as a Peto odds ratio (OR) with corresponding 95% confidence intervals.¹⁷ This method was chosen because we noted low event rates and minimal interstudy heterogeneity in the analysis.¹⁷ The results were then compared with those found with the random-effects method.¹⁸ Evidence of interstudy heterogeneity was evaluated, with statistical significance set at alpha=0.1. The analysis was performed using Cochrane Review Manager 4.2 (Cochrane Library Software, Oxford, UK).

Only trials that reported at least 1 event could be used in the estimation of odds ratios. A subsequent analysis included all trials, published and unpublished, with and without reported events, to estimate an absolute increase in risk; these results are reported in the “Discussion” section.

Subgroup analyses, chosen a priori, were performed to evaluate the difference in results between trials with variable corticosteroids (use in<100% of participants) versus concomitant corticosteroids (use in 100% of participants), salmeterol versus formoterol, children (aged<12 years) versus adults, and fatal versus nonfatal events. Further subgroup analysis compared results between trials with no corticosteroid use at all and trials with the controlled use of corticosteroid as a study drug in combination with a long-acting β-agonist (either in a single inhaler or separate inhalers). The results of the subgroups were compared with each other using the test of interaction.¹⁹

A sensitivity analysis was performed to evaluate the effect of excluding pooled trial data from GlaxoSmithKline, because individual trial-level information was not available for these events.

Role of the Funding Source 

The funding for this analysis came from salary support from Santa Clara Valley Medical Center for Drs Salpeter and Wall. The institution had no role in the design, conduct, or reporting of the study. The investigators all had complete access to the data, and no sponsorship from the institution or the pharmaceutical industry was provided to conduct this analysis.

Back to Article Outline

Results 

Search Results 

Figure 1 shows the results of the search for articles. The search identified approximately 6500 articles, of which 211 were potentially relevant trials. Of these, 10 individual trials met the inclusion criteria.²⁰, ²¹, ²², ²³, ²⁴, ²⁵, ²⁶, ²⁷, ²⁸, ²⁹ GlaxoSmithKline provided additional pooled trial data for long-acting β-agonists with and without concomitant inhaled corticosteroids.⁹ For the meta-analysis, these data were considered to be 2 pooled trials of long-acting β-agonists, one without any concomitant corticosteroid use and one with background corticosteroid use in all patients.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 1. 

  Flow chart of trials search.

Trials were excluded for the following reasons: Two trials were not randomized, 2 trials were of asthma and chronic obstructive pulmonary disease, 19 trials did not have the appropriate comparator groups, 70 trials were of less than 3 months duration, 23 trials provided duplicate data on participants from other trials, and 83 trials did not report adverse events.

Trial Characteristics 

The meta-analysis included a total of 36,588 participants followed for 21,343 patient-years (Table 1, online). The mean trial duration was 7.0 months (range, 3-12 months), with a mean study size of 3183 participants (248-26,353). The mean (standard deviation) age of participants at baseline was 38.3 (1.7) years (45.3% were men) in the β-agonist group and 38.7 (1.1) years (44.8% were men) in the control group. The dropout rate was 19.2% in the β-agonist group and 21.4% in the control group. All trials were randomized, double-blind trials that performed analysis according to intention-to-treat.

Table 1 (online). Studies Included in Meta-Analysis

Study, Reference Duration		Intervention	No. (n)	Mean Age	Dropout (%)	Percent on ICS	Comments
Long-acting β-Agonist Compared with Placebo, with Variable Corticosteroid Use
Foradil 040 trial20 12 wk	β-agonist	Formoterol 12 μg or 24 μg BID	269	35.5	5.9	Not stated	Phase 3 study sponsored by Novartis (Basel, Switzerland), NDA20-831, Protocol 040. There was 1 nonfatal asthma intubation in patient aged>12 y. Reported in Novartis FDA briefing material.10 Not reported in original trial.
	Control	Placebo	135	35.5	6.7
Foradil 041 trial21 12 wk	β-agonist   Control	Formoterol 12 μg and 24 μg BID Placebo	275   141	32.6   33.5	6.2   6.4	45   49	Phase 3 study sponsored by Novartis, NDA20-831, Protocol 041 One respiratory arrest with nonfatal intubation in patient aged>12 y. One asthma death in 66-year-old woman not on ICS, receiving formoterol 24 μg BID. Events reported in Novartis FDA briefing material.10 Asthma death reported in original trial but thought not to be related to study drug.
Foradil 230729 trial 12 wk	β-agonist   Control	Formoterol 12 μg and 24 μg BID Placebo	1054   514	38.8   37.8	13.8   15.2	65   66.7	Phase 4 safety study sponsored by Novartis, For258D-2307. One asthma intubation in a 53-year-old African-American man receiving formoterol 12 μg BID and 1 asthma intubation in a 51-year-old Caucasian man receiving formoterol 24 μg BID. Both events were reported in a clinical safety review on FDA website in 2005, but they were not reported in the Novartis FDA briefing material in 2008.
GSK pooled trials9 Not stated	β-agonist   Control	Salmeterol, dose not stated Placebo	296   298	Not stated   Not stated	Not stated   Not stated	0   0	Pooled trial data in GSK (Brentford, London, England) FDA briefing material. Three asthma intubations in the salmeterol group and 1 asthma intubation in the control group. All events occurred in randomized placebo-controlled trials without ICS use. After communication with Steven Yancey of GSK on 2/2/09, we were informed that GSK declined to provide individual trial-level information.
SMART22 28 wk	β-agonist   Control	Salmeterol 50 μg BID Placebo	13,174   13,179	39.2   39.1	22.5   23.8	49   49	Phase 4 safety study, sponsored by GSK, Trial SLGA5011. There were 13 asthma deaths and 24 intubations in the salmeterol group, and 3 asthma deaths and 19 intubations in the placebo group. All events reported in original trial.
Long-acting β-Agonist with ICS Compared with ICS Alone
GSK pooled trials9 Not stated	β-agonist   Control	Salmeterol, dose not stated Placebo	633   642	Not stated   Not stated	Not stated   Not stated	100   100	Pooled trial data provided in GSK FDA briefing material. There were 8 asthma intubations in the salmeterol group and 3 asthma intubations in the placebo group. All events occurred in randomized placebo-controlled trials with concomitant ICS use in GSK database. After communication with Steven Yancey of GSK on 2/2/09, we were informed that GSK declined to provide individual trial-level information
Ind et al23 24 wk	β-agonist	Salmeterol 50 μg+fluticasone 250 μg BID	173	44.8	15.7	100	Sponsored by GSK, Trial SLGB4010. There was 1 asthma death in the salmeterol/fluticasone group. Event reported in GSK briefing material9 and GSK-sponsored meta-analysis,14 so we removed the event from GSK pooled data. Not reported in original trial. Unpublished information received by Steven Yancey of GSK on 11/11/08.
	Control	Fluticasone 250 or 500 μg BID	329	44.8	11.4	100
Kelsen et al24 24 wk	β-agonist	Salmeterol 42 μg + beclomethasone 168 μg BID	239	42.4	20	100	Sponsored by GSK, Trial SLGA5017. There was 1 asthma intubation in the salmeterol/beclomethasone group. Reported in GSK briefing material9 and GSK-sponsored meta-analysis,14 so we removed the event from GSK pooled data. Unpublished information received by Steven Yancey of GSK on 11/11/08.
	Control	Beclomethasone 336 μg BID	244	42.0	20	100
Kemp et al25 12 wk	β-agonist   Control	Salmeterol 42 μg BID Placebo	252   254	42.0   41.6	9.9   18.5	100   100	Sponsored by GSK, Trial SLGA5001. All patients on ICS, with 1 asthma intubation in the salmeterol group. Reported in original trial as respiratory failure, judged to be possibly related to study drug. Asthma intubation reported in meta-analysis by Jaeschke et al,12 but not reported in GSK-sponsored meta-analysis.9 After communication with Steven Yancey of GSK on 4/22/09, he informed us that this event was not reported in their pooled data.
O'Byrne et al26 52 wk	β-agonist       Control	Formoterol 4.5 μg+budesonide 200 μg or 400 μg BID Budesonide 200 μg or 400 μg BID	869       862	34.8       35.4	16       16	100       100	OPTIMA study sponsored by AstraZeneca (Wilmington, DE), Trial SD-037-0345. One asthma intubation due to status asthmaticus in formoterol/budesonide group, with subsequent death due to septic shock. Reported in AstraZeneca FDA briefing material as asthma intubation;11 reported in AstraZeneca-sponsored meta-analysis as asthma death.7 No event reported in original trial. We report event as intubation in our analysis. Unpublished information received from Anders Ottosson of AstraZeneca on 11/5/08.
O'Byrne et al27 52 wk	β-agonist     Control	Formoterol 4.5+budesonide 80 μg BID Budesonide 320 μg BID	1834     926	35.5     36.0	3.2     3.1	100     100	Sponsored by AstraZeneca, Trial SD-039-0673. One asthma death in formoterol/budesonide group. Reported in AstraZeneca-sponsored meta-analysis as death due to asthma in 65-year-old woman.7 Not reported as a drug-related asthma event in AstraZeneca FDA briefing material11 or original trial. Unpublished information received by Anders Ottosson of AstraZeneca on 11/5/08.
Von Berg et al28 12 wk	β-agonist   Control	Formoterol 4.5 μg and 9 μg BID Placebo	165   83	11.4   11.3	6.1   3.6	100   100	Sponsored by AstraZeneca, Trial SD-037-0003(sponsorship not disclosed in trial). All patients receiving inhaled or oral corticosteroids. One death reported in AstraZeneca-sponsored meta-analysis as asthma-related respiratory failure and death in 13-year-old boy who was taking ICS.7 Death reported in trial as due to subarachnoid hemorrhage and judged unlikely to be due to study drug. Not reported as a drug-related asthma event in AstraZeneca FDA briefing material.11 Unpublished information received by Anders Ottosson of AstraZeneca on 11/5/08.

BID=2 times per day; FDA=Food and Drug Administration; GSK=GlaxoSmithKline; ICS=inhaled corticosteroid; SMART=Salmeterol Multi-center Asthma Research Trial.

Five trials evaluated β-agonists with variable inhaled corticosteroid use, with a total of 29,335 participants followed for 14,932 patient-years. Inhaled corticosteroids were used by 47.9% of the participants.

Seven trials evaluated β-agonists with concomitant inhaled corticosteroids, with a total of 7253 participants followed for 6044 patient-years. All participants received inhaled corticosteroids, either as uncontrolled background use or controlled as part of the study intervention. The combination of β-agonist and corticosteroid was compared with an equal dose of corticosteroid in 4 trials, an equal or higher dose of corticosteroid in 1 trial, and a higher dose of corticosteroid in 2 trials (Table 1, online).

Asthma-related Intubations and Deaths 

For all trials combined, there were 59 catastrophic events in the β-agonist group and 26 events in the control group (Table 1, online), with a Peto odds ratio of 2.10 (95% confidence interval [CI], 1.37-3.22; Figure 2). For trials with variable corticosteroid use, the odds ratio for catastrophic events for long-acting β-agonists was 1.83 (95% CI, 1.14-2.95) compared with placebo. For trials with concomitant corticosteroids, the odds ratio for catastrophic events for long-acting β-agonists plus inhaled corticosteroids, compared with inhaled corticosteroids alone, was 3.65 (95% CI, 1.39-9.55).

• 
  • View Large Image
  • Download to PowerPoint

• Figure 2. 

  Effect of long-acting β-agonists on asthma intubations and deaths. 1. Long-acting β-agonists with variable inhaled corticosteroids compared with placebo. 2. Long-acting β-agonists with concomitant inhaled corticosteroids compared with inhaled corticosteroids alone. n=number of participants with at least 1 asthma intubation or death; N=number of participants; SMART=Salmeterol Multi-center Asthma Research Trial; GSK=GlaxoSmithKline; OR = odds ratio; CI=confidence interval.

Minimal interstudy heterogeneity was noted in the analyses (P>.8). Similar results were seen when the analysis was performed using the random effects method, with significant increases in catastrophic events seen for variable corticosteroid use (OR 1.76; 95% CI, 1.08-2.89) and concomitant corticosteroid use (OR 2.7; 95% CI, 1.06-6.88).

In subgroup analysis, there was no statistically significant difference in results between concomitant and variable corticosteroids, with a P value for interaction of 0.24 (Table 2). Similar results were seen for trials with controlled corticosteroid use as part of the study intervention (OR 8.19; 95% CI, 1.1-61.18) and trials that used no corticosteroids at all (OR 2.2; 95% CI, 1.1-4.36, P for interaction=.23). In further subgroup analysis, there was no significant difference in results for formoterol compared with salmeterol, children compared with adults, and asthma deaths compared with intubations (Table 2).

Table 2. Subgroup Analysis

Analysis	Subgroup	Odds Ratio (95% CI)	P Value for Interaction
ICS use: variable compared with concomitant use	Variable ICS	1.83(1.14-2.95)	.24
	Concomitant ICS	3.65(1.39-9.55)
ICS use: no use compared with controlled study drug	No ICS	2.2(1.1-4.36)	.23
	Controlled ICS	8.19(1.1-61.18)
β-agonist used	Salmeterol	1.94(1.24-3.04)	.28
	Formoterol	4.81(1.12-20.68)
Age of participants	Adults	2.08(1.35-3.20)	.69
	Children	2.58(0.96-6.94)
Asthma event	Intubations	1.76(1.08-9.55)	.42
	Deaths	4.03(1.7-9.55)

ICS=inhaled corticosteroid; CI=confidence interval.

In sensitivity analysis, when the pooled data from GlaxoSmithKline were excluded, significant increases in catastrophic events were seen for variable corticosteroid use (OR 1.79; 95% CI, 1.09-2.92) and concomitant corticosteroid use (OR 7.34; 95% CI, 1.42-37.97).

Back to Article Outline

Discussion 

Our pooled data show that the use of long-acting β-agonists, with and without concomitant inhaled corticosteroids, was associated with a significant increase in risk for asthma-related intubations and deaths. The magnitude of risk was, in fact, higher for trials with controlled concomitant corticosteroid use (OR 8.2) than for trials with no corticosteroid use at all (OR 2.2), although the difference between the 2 subgroups was not statistically significant. The results of this meta-analysis are based on a relatively small number of events, so they should be interpreted with caution. However, these findings provide evidence that long-acting β-agonists are associated with significant risk, even when used with concomitant inhaled corticosteroids.

This is the first meta-analysis to show that concomitant treatment with long-acting β-agonists and inhaled corticosteroids increased catastrophic asthma events. This is because our analysis was based almost entirely on reports of events provided by the drug manufacturers to the FDA, which were not released until December 2008.⁹, ¹⁰, ¹¹ Of note, only 2 trials reported asthma-related intubations or death in their published article.²², ²⁵ We chose to use the composite outcome of asthma-related intubation or death because both of these catastrophic events are increased with long-acting β-agonist use.⁴, ⁵, ⁶, ⁷, ⁸ Similar increases in risk were seen for fatal and nonfatal events. We also chose to pool the data on salmeterol and formoterol, and on adults and children. We thought this was reasonable because little heterogeneity was seen in results between these subgroups. A consistent finding was a greater number of catastrophic events in the β-agonist group, compared with the control group, in each of the trials included in the analysis.

This meta-analysis was not designed to assess the absolute increase in catastrophic events associated with the addition of long-acting β-agonists to inhaled corticosteroids, because only those trials with at least 1 event are included in the calculation of odds ratios. To estimate the total number of patients at risk, we can use the cumulative trial data provided in the briefing material from the drug manufacturers, which includes published and unpublished trials with and without reported events.⁹, ¹¹ If all trials with and without events are included in the analysis, there were 14 events in 35,000 patients treated with combined therapy (0.04%) and 3 events in 29,000 patients treated with corticosteroids alone (0.01%) over an average trial duration of 5 months. This would indicate an absolute increase in risk of 3 catastrophic asthma events per 10,000 patients treated over a 5-month period. The magnitude of the potential problem is highlighted by the fact that several million patients are treated with long-acting β-agonists in the United States each year, the majority of whom also are receiving inhaled corticosteroids.⁹, ³⁰

One proposed mechanism for the increase in catastrophic asthma events associated with long-acting β-agonists is the development of tolerance to their bronchoprotective effects over time.³¹, ³² Regular use of short- and long-acting β-agonists is associated with down-regulation and desensitization of β-receptors, an increase in airway hyperreactivity, and an increase in asthma deaths and near-deaths.³¹, ³³, ³⁴, ³⁵ The use of inhaled corticosteroids, on the other hand, is associated with reduced airway reactivity and a reduction in life-threatening and fatal events.²⁶, ³⁶, ³⁷ Although there is some evidence that corticosteroids might partially protect against the adverse effects of β-agonists, regular β-agonist use still results in substantial tolerance to its effects over time, despite concomitant treatment with corticosteroids.³¹, ³⁴

The sole purpose of this meta-analysis was to evaluate the risk for catastrophic events associated with long-acting β-agonists. The study did not evaluate the benefits of long-acting β-agonists and therefore could not assess the risk-to-benefit ratio of these agents. A recent advisory committee to the FDA reviewed the available evidence and concluded that the risks of salmeterol (Serevent) and formoterol (Foradil) for asthma outweighed the benefits and should be banned for use in asthma for all ages.¹⁶ The committee agreed that further safety studies were still needed to clarify the risk of combination products, such as salmeterol with fluticasone (Advair) and formoterol with budesonide (Symbicort).¹⁶ If long-acting β-agonists were not used in the treatment of asthma, other treatment options include inhaled and oral corticosteroids, leukotriene inhibitors, short-acting anticholinergic agents, and as-needed short-acting β-agonists.³⁸, ³⁹, ⁴⁰, ⁴¹

In the United States, the asthma mortality rate peaked in the 1990s and has now been steadily declining over the past several years.⁴² This reduction in mortality, despite the continued use of long-acting β-agonists, has been used as evidence against an adverse mortality effect of these drugs.⁹, ¹⁴ However, one possible explanation for this trend could be that the asthma mortality rates are related to the ratio of β-agonist use to inhaled corticosteroid use over time.⁴² The sales of long-acting β-agonists were increasing over the past several years, but the sales of inhaled corticosteroids increased even faster during this time period, so that the ratio of β-agonist to inhaled corticosteroid use has gradually decreased at a rate that is proportional to the decreasing trend in asthma mortality rates.⁹, ⁴², ⁴³ If there were, in fact, a causal relationship between the ratio of β-agonist to corticosteroid use and asthma mortality in the United States, then we would expect a further decrease in asthma deaths if the sales of long-acting β-agonists were curtailed.

Back to Article Outline

Study Limitations 

Our analysis has several limitations. Standard meta-analytic results can be uncertain when the numbers of events per study are small, as is the case with catastrophic asthma events. The assessment of intubations and deaths was further hindered by the fact that many trials did not report the events even when they occurred and by the difficulty in ascertaining the true cause of respiratory failure or death. We chose to include only those events reported by the investigator or drug manufacturer to be asthma-related intubations or deaths. If we had chosen to include other life-threatening events thought to be due to “status asthmaticus” or “severe acute asthma” without documentation of intubation, or to include respiratory arrests and sudden deaths of unclear cause, there would be 12 more events, with 9 occurring in the long-acting β-agonist group and 3 occurring in the control group.

We chose to compare the combination of β-agonist and corticosteroid with an equal or higher dose of corticosteroid to assess whether the addition of long-acting β-agonists to inhaled corticosteroids carried a greater risk than the use of corticosteroids alone. We also chose to exclude from the analysis trials that used an active-comparator drug, such as a short-acting β-agonist, in the control group. However, our analysis of the pooled data provided by GlaxoSmithKline indicates that the long-acting β-agonist salmeterol increases the risk for asthma intubation or death by 2-fold compared with active-comparator controls (OR 2.09; 95% CI, 1.04-4.21]).⁹

Back to Article Outline

Conclusions 

Despite these limitations, our pooled data suggest that long-acting β-agonists increase the risk for asthma-related intubations and deaths, even when used in a controlled fashion with concomitant inhaled corticosteroids. At present, guidelines suggest that long-acting β-agonists be added to inhaled corticosteroids in patients with moderate or severe asthma that is not well controlled with inhaled corticosteroids alone,³⁸ but this approach might not in fact confer a meaningful health benefit compared with the use of other standard therapies.¹ The results of this meta-analysis suggest that long-acting β-agonists carry significant risk for catastrophic asthma events, even when used concomitantly with inhaled corticosteroids.

Back to Article Outline

Acknowledgment 

The authors thank Hau Liu, MD, for assistance during the preparation of this manuscript. Dr Liu received no compensation for this contribution.

Back to Article Outline

References 

1. US Food and Drug Administration Office of Surveillance and Epidemiology. Risks and benefits of long-acting beta adrenergic agonists in the treatment of asthma. http://www.fda.gov/ohrms/dockets/ac/08/briefing/2008-4398b1-01-FDA.pdfAccessed December 6, 2008
   • View In Article
2. Kramer JM. Balancing the benefits and risks of inhaled long-acting beta-agonists–the influence of values. N Engl J Med. 2009;360:1592–1595
   • View In Article
   • CrossRef
3. Drazen JM, O'Byrne PM. Risks of long-acting beta-agonists in achieving asthma control. N Engl J Med. 2009;360:1671–1672
   • View In Article
   • CrossRef
4. Salpeter SR, Buckley NS, Ormiston TM, Salpeter EE. Long-acting beta-agonists increase severe asthma exacerbations and asthma-related deaths: meta-analysis of randomized controlled trials. Ann Intern Med. 2006;144:904–912
   • View In Article
5. Cates CJ, Cates MJ. Regular treatment with salmeterol for chronic asthma: serious adverse events. Cochrane Database Syst Rev. 2008;CD006363
   • View In Article
6. Cates CJ, Cates MJ, Lasserson TJ. Regular treatment with formoterol for chronic asthma: serious adverse events. Cochrane Database Syst Rev. 2008;(4):CD006923
   • View In Article
7. Sears MR, Ottosson A, Radner F, Suissa S. Long-acting beta-agonists: a review of formoterol safety data from asthma clinical trials. Eur Respir J. 2009;33:21–32
   • View In Article
   • CrossRef
8. Levenson M. Long-acting beta-agonists and adverse asthma events meta-analysis. Statistical briefing package for Joint Meeting of the Pulmonary-Allergy Drugs Advisory Committee, Drug Safety and Risk Management Advisory Committee and Pediatric Advisory Committee on December 10-11, 2008 http://www.fda.gov/ohrms/dockets/ac/08/briefing/2008-4398b-01-FDA.pdfAccessed December 6, 2008
   • View In Article
9. GlaxoSmithKline. Sponsor briefing information (Benefit risk assessment of salmeterol for the treatment of asthma in adults and children). http://www.fda.gov/ohrms/dockets/ac/08/briefing/2008-4398b1-04-GSK.pdfAccessed December 6, 2008
   • View In Article
10. Novartis. Foradil. FDA Advisory Committee meeting on the safety of long-acting beta2-agonists for the treatment of asthma in adults and children (December 10-11, 2008). Briefing document http://www.fda.gov/ohrms/dockets/ac/08/briefing/2008-4398b1-02-Novartis.pdfAccessed December 6, 2008
    • View In Article
11. AstraZeneca. AstraZeneca briefing materials: Review of the benefits and risks of formoterol-containing products. http://www/fda.gov/ohrms/dockets/ac/08/briefing/2008-4398b1-03-AstraZeneca.pdfAccessed December 6, 2008
    • View In Article
12. Jaeschke R, O'Byrne PM, Mejza F, et al. The safety of long-acting beta-agonists among patients with asthma using inhaled corticosteroids: systematic review and metaanalysis. Am J Respir Crit Care Med. 2008;178:1009–1016
    • View In Article
    • CrossRef
13. Jaeschke R, O'Byrne PM, Nair P, et al. The safety of formoterol among patients with asthma using inhaled corticosteroids (Systematic review and meta-analysis). Pol Arch Med Wewn. 2008;118:627–635
    • View In Article
14. Bateman E, Nelson H, Bousquet J, et al. Meta-analysis: effects of adding salmeterol to inhaled corticosteroids on serious asthma-related events. Ann Intern Med. 2008;149:33–42
    • View In Article
15. Rodrigo GJ, Moral VP, Marcos LG, Castro-Rodriguez JA. Safety of regular use of long-acting beta agonists as monotherapy or added to inhaled corticosteroids in asthma (A systematic review). Pulm Pharmacol Ther. 2009;22:9–19
    • View In Article
    • CrossRef
16. US Food and Drug Administration. Summary minutes of the joint Pulmonary-Allergy Drugs Advisory Committee, Drug Safety and Risk Management Advisory Committee, and Pediatric Advisory Committee, December 10-11, 2008. http://www.fda.gov/ohrms/dockets/AC/08/minutes/2008-4398m1-final.pdfAccessed February 6, 2009
    • View In Article
17. Bradburn MJ, Deeks JJ, Berlin JA, Russell Localio A. Much ado about nothing: a comparison of the performance of meta-analytical methods with rare events. Stat Med. 2007;26:53–77
    • View In Article
    • MEDLINE
    • CrossRef
18. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–188
    • View In Article
    • MEDLINE
    • CrossRef
19. Altman DG, Bland JM. Interaction revisited: the difference between two estimates. BMJ. 2003;326:219
    • View In Article
    • CrossRef
20. U.S Food and Drug Association. Center for Drug Evaluation and Review. Foradil (Formoterol Fumarate inhalation) aerolizer. Part 4 http://www.fda.gov/cder/foi/nda/2001/20831_Foradil_medr_P4.pdfAccessed September 3, 2005
    • View In Article
21. U.S Food and Drug Association. Center for Drug Evaluation and Review. Foradil (Formoterol Fumarate inhalation) aerolizer. Part 5 http://www.fda.gov/cder/foi/nda/2001/20831_Foradil_medr_P5.pdfAccessed September 3, 2005
    • View In Article
22. Nelson HS, Weiss ST, Bleecker ER, et al. The Salmeterol Multicenter Asthma Research Trial: a comparison of usual pharmacotherapy for asthma or usual pharmacotherapy plus salmeterol. Chest. 2006;129:15–26
    • View In Article
    • MEDLINE
    • CrossRef
23. Ind PW, Dal Negro R, Colman NC, et al. Addition of salmeterol to fluticasone propionate treatment in moderate-to-severe asthma. Respir Med. 2003;97:555–562
    • View In Article
    • Abstract
    • Full-Text PDF (149 KB)
    • CrossRef
24. Kelsen SG, Church NL, Gillman SA, et al. Salmeterol added to inhaled corticosteroid therapy is superior to doubling the dose of inhaled corticosteroids: a randomized clinical trial. J Asthma. 1999;36:703–715
    • View In Article
    • MEDLINE
    • CrossRef
25. Kemp JP, Cook DA, Incaudo GA, et al. Salmeterol improves quality of life in patients with asthma requiring inhaled corticosteroids (Salmeterol Quality of Life Study Group). J Allergy Clin Immunol. 1998;101(2 Pt 1):188–195
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (241 KB)
    • CrossRef
26. O'Byrne PM, Barnes PJ, Rodriguez-Roisin R, et al. Low dose inhaled budesonide and formoterol in mild persistent asthma: the OPTIMA randomized trial. Am J Respir Crit Care Med. 2001;164(8 Pt 1):1392–1397
    • View In Article
    • MEDLINE
27. O'Byrne PM, Bisgaard H, Godard PP, et al. Budesonide/formoterol combination therapy as both maintenance and reliever medication in asthma. Am J Respir Crit Care Med. 2005;171:129–136
    • View In Article
    • MEDLINE
    • CrossRef
28. Von Berg A, Papageorgiou Saxoni F, Wille S, et al. Efficacy and tolerability of formoterol Turbuhaler in children. Int J Clin Pract. 2003;57:852–856
    • View In Article
    • MEDLINE
29. U.S Food and Drug Administration. NDA 020831 Formoterol fumarate clinical review. http://www.fda.gov/ohrms/dockets/ac/05/briefing/2005-4148B1_03_03-FDA-Clinical-Review.pdfAccessed September 3, 2005
    • View In Article
30. Patterns of medication use in the United States, 2006. A Report from the Slone Survey. http://www.bu.edu/slone/SloneSurvey/AnnualRpt/SloneSurveyWebReport2006.pdfAccessed February 21, 2009
    • View In Article
31. Salpeter SR, Ormiston TM, Salpeter EE. Meta-analysis: respiratory tolerance to regular beta2-agonist use in patients with asthma. Ann Intern Med. 2004;140:802–813
    • View In Article
32. Lipworth BJ. Long-acting beta(2)-adrenoceptor agonists: a smart choice for asthma?. Trends Pharmacol Sci. 2007;28:257–262Epub 2007 Apr 26
    • View In Article
    • MEDLINE
    • CrossRef
33. Sears MR, Taylor DR, Print CG, et al. Regular inhaled beta-agonist treatment in bronchial asthma. Lancet. 1990;336:1391–1396
    • View In Article
    • Abstract
    • CrossRef
34. Kalra S, Swystun VA, Bhagat R, Cockcroft DW. Inhaled corticosteroids do not prevent the development of tolerance to the bronchoprotective effect of salmeterol. Chest. 1996;109:953–956
    • View In Article
    • MEDLINE
    • CrossRef
35. Spitzer WO, Suissa S, Ernst P, et al. The use of beta-agonists and the risk of death and near death from asthma. N Engl J Med. 1992;326:501–506
    • View In Article
    • MEDLINE
    • CrossRef
36. Kerrebijn KF, van Essen-Zandviet EEM, Neijens JJ. Effects of long-term treatment with inhaled corticosteroids and beta-agonists on the bronchial responsiveness in children with asthma. J Allergy Clin Immunol. 1987;79:653–659
    • View In Article
    • MEDLINE
    • CrossRef
37. Suissa S, Ernst P. Inhaled corticosteroids: impact on asthma morbidity and mortality. J Allergy Clin Immunol. 2001;107:937–944
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (107 KB)
    • CrossRef
38. Expert Panel Report 3 (EPR-3): Guidelines for the Diagnosis and Management of Asthma-Summary Report 2007. J Allergy Clin Immunol. 2007;120(5 Suppl):S94–S138
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (9656 KB)
    • CrossRef
39. Gross NJ. Anticholinergic agents in asthma and COPD. Eur J Pharmacol. 2006;533:36–39
    • View In Article
    • MEDLINE
    • CrossRef
40. Currie GP, Lee DK, Srivastava P. Long-acting bronchodilator or leukotriene modifier as add-on therapy to inhaled corticosteroids in persistent asthma?. Chest. 2005;128:2954–2962
    • View In Article
    • MEDLINE
    • CrossRef
41. Kanazawa H. Anticholinergic agents in asthma: chronic bronchodilator therapy, relief of acute severe asthma, reduction of chronic viral inflammation and prevention of airway remodeling. Curr Opin Pulm Med. 2006;12:60–67
    • View In Article
    • MEDLINE
42. Salpeter SR, Buckley NS. Author's response: Asthma-related death rate http://www.annals.org/cgi/eletters/144/12/904#3689Accessed February 22, 2009
    • View In Article
43. Wijesinghe M, Perrin K, Harwood M, et al. The risk of asthma mortality with inhaled long acting beta-agonists. Postgrad Med J. 2008;84:467–472
    • View In Article
    • CrossRef