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

CLINICAL RESEARCH STUDY
Long-acting Beta-Agonists with and without Inhaled
Corticosteroids and Catastrophic Asthma Events
Shelley R. Salpeter, MD, FACP,a,b Andrew J. Wall, MD,a,b Nicholas S. Buckleyc
a
Stanford University School of Medicine, Stanford, Calif; bSanta Clara Valley Medical Center, San Jose, Calif; cCalifornia Institute for
Technology, Pasadena.
ABSTRACT
BACKGROUND: It is unclear whether long-acting ␤-agonists with concomitant inhaled corticosteroids
increase asthma-related intubations and deaths. We pooled data on long-acting ␤-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 ␤-agonists
compared with placebo or long-acting ␤-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 ␤-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 ␤-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 ␤-agonists increase the risk for asthma-related intubations and deaths, even
when used in a controlled fashion with concomitant inhaled corticosteroids.
© 2010 Elsevier Inc. All rights reserved. • The American Journal of Medicine (2010) 123, 322-328
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 ␤-agonists salmeterol and formoterol, given with or without concomitant inhaled corticosteroids.1-3 Pooled trial data have
consistently shown that long-acting ␤-agonists, when given
Funding: Santa Clara Valley Medical Center, San Jose, Calif.
Conflict of Interest: None of the authors have any conflicts of interest
associated with the work presented in this manuscript. None of the authors
have had any relationships with a pharmaceutical company that manufactures a ␤-agonist or other respiratory medications. Dr Salpeter has provided
expert testimony on a litigation case involving a long-acting ␤-agonist and
was paid on an hourly basis.
Authorship: All authors had access to the data and played a role in
writing this manuscript.
Reprint requests should be addressed to Shelley Salpeter, MD, 751 S.
Bascom Ave, San Jose, CA 95128.
E-mail address: [email protected]
0002-9343/$ -see front matter © 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjmed.2009.07.035
with variable inhaled corticosteroid use, increase the risk for
asthma-related hospitalizations, intubations, and deaths.4-8
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 ␤-agonists,
with recommendations for treatment only after other asthma
drugs have failed.1
The FDA subsequently requested data from sponsors of
long-acting ␤-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 2008.9-11 Several meta-analyses have been performed, most with the cooperation or
sponsorship of the pharmaceutical industry, which evalu-
Salpeter et al
Safety of Long-acting ␤-Agonists
323
ated long-acting ␤-agonists with concomitant inhaled corOutcome Measures
ticosteroids.7-15 These analyses consistently reported more
Two reviewers independently extracted data from the seasthma intubations and deaths for combined treatment comlected articles, reconciling differences by consensus. Outpared with inhaled corticosteroids alone, but none reached
comes assessed were catastrophic asthma events, defined as
statistical significance. However, each of these studies reasthma-related intubations or deaths that were reported by
stricted their analyses to a subset
the investigator or the industry
of the total available data, for exsponsor. Attempts were made to
ample, including only trials with
contact investigators of previous
CLINICAL SIGNIFICANCE
salmeterol,9,14 formoterol,7,11,13
meta-analyses and the industry
adults, 12 asthma deaths, 7 pubsponsors to obtain additional in● There is a 3-fold increase in asthmalished reports of events,15 single
formation concerning trials and
related
intubations
and
deaths
in
those
drug sponsors,7,9-11,14 or FDA-apevents.
␤
-agonists
with
contaking
long-acting
proved dosages8,14,15 of long-actcomitant corticosteroids compared with
ing ␤-agonists.
Statistical Analysis
corticosteroids alone.
In December 2008 a follow-up
The proportions of patients with a
FDA advisory committee meeting
● Similar increases in risk are seen for
catastrophic event to those withconcluded that the risks of salmevariable and concomitant corticosteroid
out events from each trial were
terol (Serevent, GlaxoSmithKline)
use, salmeterol and formoterol, children
pooled using the fixed-effects
and formoterol (Foradil, Novartis)
and adults, and fatal and nonfatal
method expressed as a Peto odds
outweighed the benefits and
ratio (OR) with corresponding
events.
should be banned for use in
95% confidence intervals.17 This
● This information can help assess the
asthma for all ages.16 The commethod was chosen because we
risks and benefits of the addition of
mittee separately evaluated longnoted low event rates and minimal
acting ␤-agonists combined with a
long-acting ␤-agonists to inhaled corinterstudy heterogeneity in the
corticosteroid in a single inhaler,
ticosteroids in the treatment of asthma.
analysis.17 The results were then
such as salmeterol with fluticacompared with those found with
sone (Advair, GlaxoSmithKline)
the random-effects method.18 Evand formoterol with budesonide
idence of interstudy heterogeneity
(Symbicort; AstraZeneca), and concluded that further safety
was
evaluated,
with
statistical
significance set at alstudies were needed to assess risk.1,16
pha ⫽ 0.1. The analysis was performed using Cochrane ReThe objective of this meta-analysis is to pool all the
view Manager 4.2 (Cochrane Library Software, Oxford,
available data on long-acting ␤-agonists with variable and
UK).
concomitant inhaled corticosteroids, evaluating the risk for
Only trials that reported at least 1 event could be used in
asthma-related intubations or deaths. Subgroup analyses
the estimation of odds ratios. A subsequent analysis inwill compare results between variable and concomitant corcluded all trials, published and unpublished, with and withticosteroids, salmeterol and formoterol, adults and children,
out reported events, to estimate an absolute increase in risk;
and fatal and nonfatal events.
these results are reported in the “Discussion” section.
Subgroup analyses, chosen a priori, were performed to
MATERIALS AND METHODS
evaluate the difference in results between trials with variable corticosteroids (use in ⬍ 100% of participants) versus
Trial Inclusion
concomitant corticosteroids (use in 100% of participants),
We performed a search of the MEDLINE, EMBASE, and
salmeterol versus formoterol, children (aged ⬍ 12 years)
Cochrane databases; the US FDA website; clinical-trials
versus adults, and fatal versus nonfatal events. Further subregistries of drug manufacturers; and previous meta-analygroup analysis compared results between trials with no
ses to identify trials on long-acting ␤-agonist use in patients
corticosteroid use at all and trials with the controlled use of
with asthma published through December 2008. Studies
corticosteroid as a study drug in combination with a longwere included if they were randomized controlled trials of
acting ␤-agonist (either in a single inhaler or separate inlong-acting ␤-agonists compared with placebo or long-acthalers). The results of the subgroups were compared with
ing ␤-agonists with inhaled corticosteroids compared with
each other using the test of interaction.19
an equal or higher dose of inhaled corticosteroids alone of at
A sensitivity analysis was performed to evaluate the effect of
least 3 months duration that reported at least 1 asthmaexcluding pooled trial data from GlaxoSmithKline, because indirelated intubation or death. We included reports of events
vidual trial-level information was not available for these events.
that were provided by the investigator in the published trial
or by the drug manufacturer in a subsequent account. In
Role of the Funding Source
addition, we included pooled trial data from GlaxoSmithKline,
The funding for this analysis came from salary support from
because we were unable to obtain individual trial-level inSanta Clara Valley Medical Center for Drs Salpeter and
formation for those events.
324
The American Journal of Medicine, Vol 123, No 4, April 2010
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.
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.20-29 GlaxoSmithKline provided
additional pooled trial data for long-acting ␤-agonists with
and without concomitant inhaled corticosteroids.9 For the
meta-analysis, these data were considered to be 2 pooled
trials of long-acting ␤-agonists, one without any concomitant corticosteroid use and one with background corticosteroid use in all patients.
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.
Figure 1
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 ␤-agonist
group and 38.7 (1.1) years (44.8% were men) in the control
group. The dropout rate was 19.2% in the ␤-agonist group
and 21.4% in the control group. All trials were randomized,
double-blind trials that performed analysis according to
intention-to-treat.
Five trials evaluated ␤-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 ␤-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 ␤-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).
Flow chart of trials search.
Salpeter et al
Safety of Long-acting ␤-Agonists
325
Figure 2 Effect of long-acting ␤-agonists on asthma intubations and deaths. 1. Long-acting
␤-agonists with variable inhaled corticosteroids compared with placebo. 2. Long-acting ␤-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.
Asthma-related Intubations and Deaths
For all trials combined, there were 59 catastrophic events in
the ␤-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 ␤-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 ␤-agonists plus inhaled corticosteroids, compared with inhaled corticosteroids alone, was
3.65 (95% CI, 1.39-9.55).
Table 2
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
Subgroup Analysis
Analysis
Subgroup
Odds Ratio (95% CI)
P Value for Interaction
ICS use: variable compared with concomitant use
Variable ICS
Concomitant ICS
No ICS
Controlled ICS
Salmeterol
Formoterol
Adults
Children
Intubations
Deaths
1.83
3.65
2.2
8.19
1.94
4.81
2.08
2.58
1.76
4.03
.24
ICS use: no use compared with controlled study drug
␤-agonist used
Age of participants
Asthma event
ICS ⫽ inhaled corticosteroid; CI ⫽ confidence interval.
(1.14-2.95)
(1.39-9.55)
(1.1-4.36)
(1.1-61.18)
(1.24-3.04)
(1.12-20.68)
(1.35-3.20)
(0.96-6.94)
(1.08-9.55)
(1.7-9.55)
.23
.28
.69
.42
326
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).
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).
DISCUSSION
Our pooled data show that the use of long-acting ␤-agonists,
with and without concomitant inhaled corticosteroids, was
associated with a significant increase in risk for asthmarelated 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 ␤-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 ␤-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 2008.9-11 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
␤-agonist use.4-8 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
␤-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 ␤-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.
The American Journal of Medicine, Vol 123, No 4, April 2010
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 ␤-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 ␤-agonists is the
development of tolerance to their bronchoprotective effects
over time.31,32 Regular use of short- and long-acting ␤-agonists is associated with down-regulation and desensitization
of ␤-receptors, an increase in airway hyperreactivity, and an
increase in asthma deaths and near-deaths.31,33-35 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 ␤-agonists, regular ␤-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
␤-agonists. The study did not evaluate the benefits of longacting ␤-agonists and therefore could not assess the risk-tobenefit 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 ␤-agonists were not used in the
treatment of asthma, other treatment options include inhaled and oral corticosteroids, leukotriene inhibitors, shortacting anticholinergic agents, and as-needed short-acting
␤-agonists.38-41
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 ␤-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 ␤-agonist use to inhaled corticosteroid use over
time.42 The sales of long-acting ␤-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 ␤-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
␤-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 ␤-agonists were
curtailed.
Salpeter et al
Safety of Long-acting ␤-Agonists
327
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 ␤-agonist group and 3
occurring in the control group.
We chose to compare the combination of ␤-agonist and
corticosteroid with an equal or higher dose of corticosteroid
to assess whether the addition of long-acting ␤-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 ␤-agonist, in the control group. However, our
analysis of the pooled data provided by GlaxoSmithKline
indicates that the long-acting ␤-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
CONCLUSIONS
Despite these limitations, our pooled data suggest that longacting ␤-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 ␤-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 ␤-agonists carry significant risk for catastrophic asthma events, even when used concomitantly with
inhaled corticosteroids.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
ACKNOWLEDGMENT
The authors thank Hau Liu, MD, for assistance during the
preparation of this manuscript. Dr Liu received no compensation for this contribution.
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. Available at: http://www.fda.gov/ohrms/
17.
18.
19.
dockets/ac/08/briefing/2008-4398b1-01-FDA.pdf. Accessed December 6, 2008.
Kramer JM. Balancing the benefits and risks of inhaled long-acting
beta-agonists–the influence of values. N Engl J Med. 2009;360:15921595.
Drazen JM, O’Byrne PM. Risks of long-acting beta-agonists in achieving asthma control. N Engl J Med. 2009;360:1671-1672.
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.
Cates CJ, Cates MJ. Regular treatment with salmeterol for chronic
asthma: serious adverse events. Cochrane Database Syst Rev. 2008:
CD006363.
Cates CJ, Cates MJ, Lasserson TJ. Regular treatment with formoterol
for chronic asthma: serious adverse events. Cochrane Database Syst
Rev. 2008(4):CD006923.
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.
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. Available at: http://www.fda.gov/ohrms/
dockets/ac/08/briefing/2008-4398b-01-FDA.pdf. Accessed December
6, 2008.
GlaxoSmithKline. Sponsor briefing information. Benefit risk assessment of salmeterol for the treatment of asthma in adults and children.
Available at: http://www.fda.gov/ohrms/dockets/ac/08/briefing/20084398b1-04-GSK.pdf. Accessed December 6, 2008.
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. Available at:
http://www.fda.gov/ohrms/dockets/ac/08/briefing/2008-4398b1-02Novartis.pdf. Accessed December 6, 2008.
AstraZeneca. AstraZeneca briefing materials: Review of the benefits
and risks of formoterol-containing products. Available at: http://www/
fda.gov/ohrms/dockets/ac/08/briefing/2008-4398b1-03-AstraZeneca.
pdf. Accessed December 6, 2008.
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.
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.
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.
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.
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. Available at: http://www.fda.gov/ohrms/
dockets/AC/08/minutes/2008-4398m1-final.pdf. Accessed February 6,
2009.
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.
DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin
Trials. 1986;7:177-188.
Altman DG, Bland JM. Interaction revisited: the difference between
two estimates. BMJ. 2003;326:219.
328
20. U.S Food and Drug Association. Center for Drug Evaluation and
Review. Foradil (Formoterol Fumarate inhalation) aerolizer. Part
4. Available at: http://www.fda.gov/cder/foi/nda/2001/20831_
Foradil_medr_P4.pdf. Accessed September 3, 2005.
21. U.S Food and Drug Association. Center for Drug Evaluation and
Review. Foradil (Formoterol Fumarate inhalation) aerolizer. Part
5. Available at: http://www.fda.gov/cder/foi/nda/2001/20831_
Foradil_medr_P5.pdf. Accessed September 3, 2005.
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.
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.
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;36:703715.
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.
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):13921397.
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.
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.
29. U.S Food and Drug Administration. NDA 020831 Formoterol fumarate clinical review. Available at: http://www.fda.gov/ohrms/dockets/
ac/05/briefing/2005-4148B1_03_03-FDA-Clinical-Review.pdf. Accessed September 3, 2005.
30. Patterns of medication use in the United States, 2006. A Report from
the Slone Survey. Available at: http://www.bu.edu/slone/SloneSurvey/
The American Journal of Medicine, Vol 123, No 4, April 2010
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
AnnualRpt/SloneSurveyWebReport2006.pdf. Accessed February 21,
2009.
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.
Lipworth BJ. Long-acting beta(2)-adrenoceptor agonists: a smart
choice for asthma? Trends Pharmacol Sci. 2007;28:257-262. Epub
2007 Apr 26.
Sears MR, Taylor DR, Print CG, et al. Regular inhaled beta-agonist
treatment in bronchial asthma. Lancet. 1990;336:1391-1396.
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.
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.
Kerrebijn KF, van Essen-Zandviet EEM, Neijens JJ. Effects of longterm treatment with inhaled corticosteroids and beta-agonists on the
bronchial responsiveness in children with asthma. J Allergy Clin Immunol. 1987;79:653-659.
Suissa S, Ernst P. Inhaled corticosteroids: impact on asthma morbidity
and mortality. J Allergy Clin Immunol. 2001;107:937-944.
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.
Gross NJ. Anticholinergic agents in asthma and COPD. Eur J Pharmacol. 2006;533:36-39.
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.
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.
Salpeter SR, Buckley NS. Author’s response: Asthma-related death
rate. Available at: http://www.annals.org/cgi/eletters/144/12/904#3689.
Accessed February 22, 2009.
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.
(online) Studies Included in Meta-Analysis
Study,
Reference Duration
Intervention
No. (n)
Mean Age
Dropout
(%)
Percent
on ICS
Comments
Foradil 040 trial20
12 wk
␤-agonist
Control
Foradil 041 trial21
12 wk
␤-agonist
Control
Foradil 230729 trial
12 wk
␤-agonist
Control
GSK pooled trials9
Not stated
␤-agonist
Control
SMART22
28 wk
␤-agonist
Control
Formoterol 12 ␮g or
24 ␮g BID
Placebo
269
35.5
5.9
135
35.5
6.7
Formoterol 12 ␮g
and 24 ␮g BID
Placebo
275
32.6
6.2
45
141
33.5
6.4
49
Formoterol 12 ␮g
and 24 ␮g BID
Placebo
1054
38.8
13.8
65
514
37.8
15.2
66.7
Salmeterol, dose
not stated
Placebo
296
Not stated
Not stated
0
298
Not stated
Not stated
0
13,174
39.2
22.5
49
13,179
39.1
23.8
49
633
Not stated
Not stated
100
642
Not stated
Not stated
100
Salmeterol 50 ␮g
BID
Placebo
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.
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 ␮g BID. Events reported in Novartis FDA briefing
material.10 Asthma death reported in original trial but thought not
to be related to study drug.
Phase 4 safety study sponsored by Novartis, For258D-2307.
One asthma intubation in a 53-year-old African-American man
receiving formoterol 12 ␮g BID and 1 asthma intubation in a
51-year-old Caucasian man receiving formoterol 24 ␮g 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.
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 triallevel information.
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.
Safety of Long-acting ␤-Agonists
Long-acting ␤-Agonist Compared with Placebo, with Variable Corticosteroid Use
Salpeter et al
Table 1
Long-acting ␤-Agonist with ICS Compared with ICS Alone
GSK pooled trials9
Not stated
␤-agonist
Control
Salmeterol, dose
not stated
Placebo
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 placebocontrolled 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
328.e1
328.e2
Table 1
Continued
Study,
Reference Duration
23
Ind et al
24 wk
Intervention
␤-agonist
Control
Kelsen et al24
24 wk
␤-agonist
Control
Kemp et al25
12 wk
␤-agonist
Control
␤-agonist
Control
O’Byrne et al27
52 wk
␤-agonist
Control
Von Berg et al28
12 wk
␤-agonist
Control
Mean Age
Dropout
(%)
Percent
on ICS
Salmeterol 50
␮g ⫹ fluticasone
250 ␮g BID
Fluticasone 250 or
500 ␮g BID
Salmeterol 42 ␮g ⫹
beclomethasone
168 ␮g BID
Beclomethasone
336 ␮g BID
Salmeterol 42 ␮g
BID
Placebo
173
44.8
15.7
100
329
44.8
11.4
100
239
42.4
20
100
244
42.0
20
100
252
42.0
9.9
100
254
41.6
18.5
100
Formoterol 4.5
␮g ⫹ budesonide
200 ␮g or 400
␮g BID
Budesonide 200 ␮g
or 400 ␮g BID
869
34.8
16
100
862
35.4
16
100
Formoterol
4.5 ⫹ budesonide
80 ␮g BID
Budesonide 320 ␮g
BID
1834
35.5
3.2
100
926
36.0
3.1
100
Formoterol 4.5 ␮g
and 9 ␮g BID
Placebo
165
11.4
6.1
100
83
11.3
3.6
100
Comments
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.
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.
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.
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.
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.
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.
The American Journal of Medicine, Vol 123, No 4, April 2010
O’Byrne et al26
52 wk
No. (n)