Summary of findings
When asthma is not controlled by low-dose inhaled corticosteroids alone, many asthma guidelines recommend the use of additional long-acting beta
However, long-standing controversy over the regular use of beta
Concern remains that the symptomatic benefit derived from treatment with long-acting beta
Two long-acting beta
Much debate has focused on the interaction between inhaled corticosteroids and long-acting beta
Counfounding by severity has been shown by Sears 2008 in data from the RELIEF study, where the rate of asthma-related SAEs was significantly higher in both arms of the study among participants taking ICS in comparison with those not taking ICS. This is a serious threat to any conclusions drawn from observational data when the interaction between ICS and formoterol is assessed. Therefore, there is a need to systematically review all available data from controlled trials that randomly assigned participants to regular formoterol in combination with inhaled corticosteroids, and to consider all serious adverse events (fatal and non-fatal), whether or not these are deemed by the investigators to be related to trial medication.
The focus of this review is therefore on regular formoterol randomly assigned in combination with inhaled corticosteroids (in a single inhaler or in separate inhalers) and compared with inhaled corticosteroids alone. Because of the difficulty involved in deciding whether adverse events are asthma-related, this review focusses on studies that capture mortality and serious adverse events and records both all-cause outcomes and those considered by trial investigators to be asthma-related events.
A review comparing regular salmeterol randomly assigned in combination with inhaled corticosteroids (in a single inhaler or in separate inhalers) versus inhaled corticosteroids alone is currently being updated (Cates 2009), and an overview of the safety of regular formoterol or salmeterol in children has been published (Cates 2012a).
To assess the risks of mortality and non-fatal serious adverse events in trials that randomly assign participants with chronic asthma to regular formoterol and inhaled corticosteroid versus the same dose of inhaled corticosteroid.
Criteria for considering studies for this review
Types of studies
Controlled parallel-design clinical trials, with or without blinding, in which formoterol and inhaled corticosteroid were randomly assigned to participants with chronic asthma for comparison with outcomes in those given the same dose of inhaled corticosteroid alone. Studies on acute asthma and exercise-induced bronchospasm have not been included.
Types of participants
Participants with a clinical diagnosis of asthma of any age group, unrestricted by disease severity or previous or current treatment.
Types of interventions
Inhaled corticosteroids and formoterol given regularly once or twice daily for a period of at least 12 weeks at any dose and delivered by any single or separate devices (CFC-MDI, HFA-MDI, DPI). Studies that randomly assigned participants to formoterol and inhaled corticosteroids for intermittent use as a reliever have not been included in this review, and studies that compared different doses of formoterol or different delivery devices or propellants without a placebo arm were also not included. Studies in which formoterol was randomly assigned without an inhaled steroid have been considered in a separate review (Cates 2012). Studies that use comparison groups given the same dose and type of inhaled corticosteroid in the control arm will be included in this review, and co-intervention with leukotriene-receptor antagonists, cromones or theophylline will be allowed as long as they are not part of the randomly assigned intervention and therefore are not systematically different between groups. Studies comparing formoterol with salmeterol will be subject to another review and were not included in this review. We have also excluded from this review studies in which inhaled corticosteroids were used in all participants as background treatment (rather than as a randomised intervention).
Types of outcome measures
- All-cause mortality.
- All-cause non-fatal serious adverse events.
- Asthma-related mortality.
- Asthma-related non-fatal serious adverse events.
- Respiratory-related mortality.
- Respiratory-related non-fatal serious adverse events.
- Cardiovascular-related mortality.
- Cardiovascular-related non-fatal serious adverse events.
- Asthma-related non-fatal life-threatening events (intubation or admission to intensive care).
- Respiratory-related non-fatal life-threatening events (intubation or admission to intensive care).
Outcomes were not subdivided according to whether the trial investigators considered them related to trial medication.
Search methods for identification of studies
Trials were identified using the Cochrane Airways Group Specialised Register of trials, which is derived from systematic searches of bibliographic databases, including the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE, CINAHL, AMED and PsycINFO and from handsearching of respiratory journals and meeting abstracts (see Appendix 3 for additional details). All records in the Specialised Register coded as 'asthma' were searched using the following terms:
(((beta* and agonist*) and (long-acting or "long acting")) or ((beta* and adrenergic*) and (long-acting or "long acting")) or (bronchodilat* and (long-acting or "long acting")) or (salmeterol or formoterol or eformoterol or advair or symbicort or serevent or seretide or oxis)) AND (serious or safety or surveillance or mortality or death or intubat* or adverse or toxicity or complications or tolerability)
Searches were conducted up to August 2012 with no restriction on language of publication.
Searching other resources
Reference lists of all primary studies and review articles were checked for additional references. Web sites of clinical trial registers were checked for unpublished trial data; Food and Drug Administration (FDA) submissions in relation to formoterol were also checked.
Data collection and analysis
Selection of studies
Two review authors independently assessed studies identified in the literature searches by examining title, abstract and keywords fields. Studies that potentially fulfilled the inclusion criteria were obtained in full text. These were independently assessed for inclusion by CJC, with a second assessment by Toby Lasserson for the original review and by Marta Oleszczuk for the 2012 update. Disagreements were resolved by consensus.
Data extraction and management
Data were extracted using a prepared checklist before they were entered into Rev Man 5.0 by one reviewer (CJC), and data extraction and entry were checked by a second reviewer (Toby Lasserson,MF or SS). Outcome data were independently extracted by the third reviewer (RJ, MF or SS) and discrepancies resolved by correspondence with the sponsors, when necessary. Data included characteristics of included studies (methods, participants, interventions, outcomes) and results of the included studies. Sponsors of included studies were contacted for unpublished adverse event data, and the sponsor's Web site was searched for further details of adverse events. All-cause serious adverse events (fatal and non-fatal) were recorded, and in view of the difficulty involved in deciding whether events were asthma related, details of the cause of death and of serious adverse events were noted when available. The definition of serious adverse events was recorded, and further information was sought if this was not clear (particularly in relation to hospital admissions and serious adverse events).
Assessment of risk of bias in included studies
One review author (CJC) assessed all included studies for bias protection (including sequence generation for randomisation, allocation concealment, blinding of participants and assessors, loss to follow-up and selective outcome reporting) with assistance from Susan Hansen for the original version of the review and from MF or SS for the 2012 update.
Unit of analysis issues
We confined our analysis to participants with one or more serious adverse event, rather than focusing on the number of events that occurred (as the latter are not independent when one patient suffers multiple events).
Assessment of heterogeneity
Heterogeneity was assessed using I
The outcomes of this review were dichotomous, and we recorded the numbers of participants with at least one outcome event by allocated treated group. Pooled odds ratio (ORs) and risk difference (RDs) were calculated. The Peto OR provides advantages when events are rare as no adjustment for zero cells is required (Bradburn 2007). We considered this specific property to be more important than potential problems with unbalanced treatment arms and large effect sizes associated with this method in view of the low number of events and the high proportion of zero cells. The primary analysis of results for serious adverse event outcomes was conducted in RevMan 5.2 using the Peto method, and the Mantel-Haenszel method was used as a sensitivity analysis. Funnel plots were inspected for assessment of possible publication bias.
Subgroup analysis and investigation of heterogeneity
Subgroup analyses were conducted on the basis of participant age (adults vs children) and the dose of formoterol used (usual dose vs high dose). Subgroups were compared with the use of tests for interaction (Altman 2003).
Sensitivity analysis was carried out to assess the impact of the method used to combine study events (risk difference, Peto OR and Mantel-Haenszel OR). The degree of bias protection included in the study design was also included in the of sensitivity analysis.
Summary of findings tables
We assessed the quality of the evidence for all-cause mortality, all-cause non-fatal serious adverse events and asthma-related serious adverse events. Assessments were conducted according to recommendations put forth by the GRADE working group and are presented in our reviews in separate Summary of Findings tables for adults and children.
Description of studies
Results of the search
A total of 523 abstracts were found from the search of the Cochrane Airways Group Specialised Register of trials in October 2008. For this review, 59 abstracts were identified as potentially relevant, and of these, 48 were subsequently excluded (see Characteristics of excluded studies) and 11 were included in the review. Ten additional trials were identified primarily from the AstraZeneca register of controlled trials (Buhl 2003; D5896C00001; Morice 2007; Morice 2008; Peters 2008; SD-039-0714; SD-039-0718; SD-039-0719; SD-039-0725; SD-039-0726). No additional trials were found from the Novartis Website or the FDA Website. A submission from Novartis to the FDA in December 2008 indicated that no trials in the Novartis database were trials of Foradil in which participants had also been randomised to inhaled corticosteroids (Novartis 2008).
The August 2012 search update found another 199 abstracts, from which 55 were identified as potentially relevant to this review. The full papers identified 25 reports of six new trials of 2550 adults and adolescents given regular formoterol in combination with budesonide or mometasone (Brown 2012; Meltzer 2012; Nathan 2010; Spector 2012; Weinstein 2010; Zangrilli 2011). Twelve additional citations were identified for six trials already included in the review (D5896C00001; Noonan 2006; Peters 2008; SD-039-0718; SD-039-0719; SD-039-0725). Twelve studies (with 18 reports) were excluded as documented in the Characteristics of excluded studies.
No new studies on children were identified.
The 27 trials included in this review are described in detail in Characteristics of included studies, and a summary of the daily dose of budesonide and formoterol that was used in each trial is provided in Table 1). To avoid confusion, all delivered doses have been converted to an equivalent metered dose (so Symbicort 320/9 mcg is a delivered dose that is equivalent to a metered dose of budesonide 400 mcg and formoterol 12 mcg).
Table 1 also indicates whether each study randomly assigned participants to once- or twice-daily formoterol, used combined or separate inhalers and delivered the medication using dry powder inhaler (DPI) or pressurised metered-dose inhalers (pMDIs). Some trials had more than two arms so featured more than one option in each of these cases. Because OPTIMA (O'Byrne 2001; O'Byrne 2001a) and FACET (Pauwels 1997; Pauwels 1997a) randomly assigned participants to higher and lower doses of budesonide, each has been considered as two separate comparisons and has been given two identifiers. The review therefore lists a total of 29 studies, drawn from the 27 trials that have been conducted.
All the trials on budesonide and formoterol have been sponsored or supported by AstraZeneca, and all the trials on mometasone and formoterol have been sponsored by Merck or Schering-Plough (Meltzer 2012; Nathan 2010; Weinstein 2010).
After the 2012 update, a total of 10,578 adults and adolescents were randomly assigned in 13 trials enrolling participants over the age of 12 years (Brown 2012; Corren 2007; D5896C00001; Jenkins 2006; Meltzer 2012; Morice 2007; Nathan 2010; Noonan 2006; Peters 2008; Price 2002; Spector 2012; Weinstein 2010; Zangrilli 2011), a further six trials that enrolled adults over the age of 18 years (Buhl 2003; Chuchalin 2002; Kuna 2006; O'Byrne 2001; Pauwels 1997; Zetterstrom 2001) and a single trial that enrolled those older than 16 years of age (SD-039-0726). All participants in these studies had a mean age of greater than 18 years.
The weighted mean duration of the adult and adolescent studies was 28 weeks. The daily metered dose of formoterol used was 12 to 24 mcg, with the exception of Jenkins 2006 and Peters 2008, who used 48 mcg daily (which remains within the licensed daily dose range). The daily metered dose of budesonide ranged from 200 to 1600 mcg, and for mometasone from 200 to 800 mcg (see Table 1).
The seven trials in children included 2788 participants in the following age ranges: Morice 2008, 6 to 11 years old; Pohunek 2006, 4 to 11; SD-039-0714, 11 to 17; SD-039-0718, 6 to 15; SD-039-0719, 6 to 11; SD-039-0725, 6 to 15 and Tal 2002 (4 to 17). In all studies, the mean age of participants was younger than 18 years.
The weighted mean duration of the studies of children and adolescents was 13 weeks. The daily metered dose of formoterol was 12 to 24 mcg. The daily metered dose of budesonide was 200 to 400 mcg (see Table 1).
Risk of bias in included studies
|Figure 1. Methodological quality summary: review authors' judgements about each methodological quality item for each included study.|
Very little information is available from the paper publications or from Web reports on sequence generation or allocation concealment, but this is unlikely to be a source of bias in view of the fact that all the studies are sponsored, and standard methodology is likely to have been used to minimise the risk of selection bias. We therefore regarded the risk of selection bias as low, although sequence generation and allocation concealment are marked as unclear in most studies in Figure 1.
All of the studies were double-blind with the exception of SD-039-0719, which was an open study. We regard the overall risk of performance and detection bias as low for the all-cause events. We were concerned that bias might have been introduced in the attribution of asthma as the cause of serious events, as this was not independently assessed. Although the trials were double-blind, formoterol can have a big impact on asthma symptoms, and those who decided on the cause of the events may have guessed which treatment was being given.
Incomplete outcome data
The rate of withdrawals and dropouts was clearly reported and was generally less than 20% for randomly assigned participants; these rates were similar in the arms of each study. However, Spector 2012 reported more withdrawals on budesonide alone (34% compared with 24% on combination treatment), so we judged this study to be at high risk of attrition bias.
Data have been found or provided from the sponsor for fatal and non-fatal serious adverse events by treatment group and causation for all studies, except for Brown 2012, which does not include details of asthma-related serious adverse events. We have therefore obtained data from all trials in relation to the primary outcomes of all-cause mortality and all-cause serious adverse events, and we regard the overall risk of reporting bias as low.
Effects of interventions
See: Summary of findings for the main comparison Regular formoterol and ICS compared to ICS in adults with asthma; Summary of findings 2 Regular formoterol and ICS compared to ICS in children with asthma
No deaths were reported in the trials on children and adolescents (2788 participants). In the adult and adolescent studies (10,578 participants), seven deaths occurred. Six deaths were reported in 6163 participants taking formoterol with inhaled corticosteroids, and one death occurred out of 4415 participants taking inhaled corticosteroids alone. These trials were combined with the use of the Peto odds ratio (OR; as no continuity correction for zero cells is required). The increased odds of all-cause mortality with formoterol did not reach statistical significance (Peto OR 3.56, 95% confidence interval (CI) 0.79 to 16.03, P = 0.1) and I
|Figure 2. Forest plot of comparison: 1 Formoterol and ICS versus same dose ICS, outcome: 1.1 All-cause mortality.|
Reports on the cause of each death are documented in Table 2.
Serious adverse events (non-fatal all-cause)
A non-fatal serious adverse event is defined as an event that falls into any of the following categories:
- Is life-threatening.
- Requires inpatient hospitalisation or prolongation of existing hospitalisation.
- Results in persistent or significant disability/incapacity.
- Is a congenital anomaly/birth defect.
This is further explained in Appendix 4, and AstraZeneca has confirmed that this definition was used in its included trials (even though this often was not made explicit in the paper reports).
Adults and adolescents
The number of participants experiencing one or more non-fatal serious adverse events was very similar when formoterol was randomly assigned with inhaled corticosteroids (ICS) in comparison with ICS alone. In 142 out of 6163 (2.3%) participants on regular formoterol with ICS and in 113 out of 4415 (2.6%) on ICS alone, such events occurred. The Peto OR was close to 1 (0.98, 95% CI 0.76 to 1.27, P = 0.87), and I
|Figure 3. Forest plot of comparison: 1 Formoterol and ICS versus same dose ICS, outcome: 1.2 All-cause non-fatal serious adverse events.|
Children and adolescents
In trials of participants who were younger than 18 years of age, the results were more heterogeneous, and more non-fatal serious adverse events occurred with formoterol. A total of 25 such events were reported amongst young people out of 1719 (1.5%) on regular formoterol with ICS, and nine events occurred amongst 1069 (0.8%) participants taking ICS alone. The increased odds of serious adverse events with formoterol did not reach statistical significance: Peto OR: 1.62 (95% CI 0.80 to 3.28, P = 0.18); I
The test for interaction between adults and children did not find a significant impact of age on treatment effect during analysis as Peto OR (test for subgroup differences: Chi² = 1.72, df = 1, P = 0.19, I² = 41.9%; see Figure 3).
Mortality by cause of death
None of the seven deaths in adults were reported as being due to asthma in the original trial reports, but the death in O'Byrne 2001 (OPTIMA) was subsequently attributed to status asthmaticus and septic shock in a recent meta-analysis (Sears 2008). The full report on the cause of death provided by the sponsors stated, "One of the deaths occurred in a 35 year old female after an 8 day hospitalisation for a severe asthma attack leading to intubation, ventilation, and nosocomial pneumonia with septic shock." This is the only death that has been reported as related to asthma and it occurred in a patient taking budesonide/formoterol Figure 4. Two deaths were reported as due to suicide, one as homicide, and one as cerebrovascular accident; one patient died of a uterine leiomyosarcoma and one from cardiac arrest (see Table 2).
|Figure 4. Forest plot of comparison: 1 Formoterol and ICS versus same dose ICS, outcome: 1.3 Asthma mortality.|
Serious adverse events related to asthma
Adults and adolescents
The number of adults experiencing one or more asthma-related non-fatal serious adverse events was lower when formoterol was randomly assigned with ICSs in comparison with ICSs alone, and after the 2012 update, the difference reached statistical significance when both the Peto OR and the RD were used. A total of 17 out of 5981 (0.3%) participants on regular formoterol and ICS suffered an asthma-related serious adverse event, as well as 30 out of 4227 (0.7%) on ICS alone. The Peto OR was 0.49 (95% CI 0.28 to 0.88), and I
|Figure 5. Forest plot of comparison: 1 Formoterol and ICS versus same dose ICS, outcome: 1.4 Asthma-related non-fatal serious adverse events.|
Children and adolescents
In trials in participants who were younger than 18 years of age, the results were again more heterogeneous. Nine young people out of 1719 (0.5%) on regular formoterol and ICS suffered an asthma-related serious adverse event (SAE), as did four out of 1069 (0.4%) on ICS alone. The increased odds of SAEs related to asthma was imprecise and was not statistically significant (Peto OR 1.49, 95% CI 0.48 to 4.61, P = 0.49), and I
The difference between children and adults was not statistically significant (test for subgroup differences: Chi² = 2.94, df = 1, P = 0.09, I² = 66.0%). The data retrieved were insufficient to allow assessment of the other proposed secondary outcomes (such as intensive care unit (ICU) admission and intubation). One intubation (ICS only group; O'Byrne 2001) was reported.
Risk of bias
No deaths occurred in children in the unblinded study (SD-039-0719), so exclusion of this study resulted in no difference in mortality outcomes. When this study was excluded for non-fatal SAEs, the Peto OR for children was 1.69 (95% CI 0.81 to 3.54) and I
|Figure 6. Funnel plot of comparison: 1 Formoterol and ICS versus same dose ICS, outcome: 1.2 All-cause non-fatal serious adverse events.|
Methods of analysis
Primary outcomes were also analysed using Mantel-Haenszel fixed-effect and random-effects models. The result of a fixed-effect model for mortality was OR 1.95, 95% CI 0.56 to 6.82. This method uses a correction for zero cells, which means that the pooled OR is smaller than the Peto OR, because the addition of 0.5 to all cells when the arms have similar numbers randomly assigned will generate an OR of 3 when only one event is reported. When outcomes are very sparse (as for mortality), the results are entirely dependent on the size of the zero cell adjustment and whether the treatment arms are balanced. For all-cause SAEs in adults, the Mantel-Haenszel fixed-effect (OR 0.97, 95% CI 0.75 to 1.26) and random-effects models (OR 0.93, 95% CI 0.71 to 1.21) yielded results almost identical to those obtained by the Peto method. Similarly for asthma-related SAEs in adults, the fixed-effect (OR 0.52, 95% CI 0.30 to 0.90) and random-effects Mantel-Haenszel models (OR 0.53, 95% CI 0.30 to 0.95) provided very similar to results to those obtained when the Peto model was used.
Dose of formoterol
The dose of formoterol used in all studies was within the licensed daily dose, so no sensitivity analysis was required to exclude unlicensed doses.
Mortality data were too sparse for any subgroup analyses to be carried out.
Although the results for adults and children showed opposite directions of effect for non-fatal SAEs (both all-cause and asthma-related), the test for interaction did not show a significant interaction of treatment effect and age.
Summary of main results
The CIs for all-cause mortality in adults indicate that for every thousand patients treated with regular formoterol and ICS in comparison with ICS alone, we can expect something between three additional deaths and one less death in adults over 28 weeks of treatment, and at most four additional deaths to four fewer deaths in children over 13 weeks of treatment (the average duration of treatment in the respective trials). The pooled Peto OR for adults was 3.56 (95% CI 0.79 to 16.03) and could not be calculated for children because no deaths in children were reported.
All-cause non-fatal SAEs
For non-fatal SAEs, the limits of the pooled CI are six more to seven fewer adults and ten more to three fewer children for every thousand treated over the period of time represented in the trials. The Peto OR was 0.98 (95% CI 0.76 to 1.27) for adults, and 1.62 (95% CI 0.80 to 3.28) for children, with heterogeneity noted in the results of studies in children.
Only one death was attributed to asthma, so evidence was insufficient to allow assessment of impact on asthma-related mortality in adults or children.
Asthma-related non-fatal SAEs
Since six new trials in adults were added, a significant reduction in asthma-related non-fatal events has been reported in adults (Peto OR 0.49, 95% CI 0.28 to 0.88). This represents a reduction of three adults per thousand treated with combination therapy over 28 weeks, with a 95% CI of six fewer to less than one per thousand fewer. The results in children do not show a significant difference (Peto OR 1.49, 95% CI 0.48 to 4.61), and although the direction of the effect in children contrasts with that in adults, the difference between results in children and those in adults is not significant (test for subgroup differences: Chi² = 2.94, df = 1, P = 0.09, I² = 66.0%).
Overall completeness and applicability of evidence
Two large studies have examined the use of regular salmeterol (SMART 2006, SNS 1993), but the only large surveillance study on formoterol (Pauwels 2003; RELIEF) investigated its use as a reliever rather than as maintenance therapy and therefore has not been included in this review. This means that data in this review are insufficient to allow investigation of the impact of formoterol on SAEs in comparison with our previous review on salmeterol (Cates 2008).
The small number of events in this review results in low precision of the estimates of relative risk between formoterol and control. However, outcome data for all-cause events were obtained from all included studies, and a funnel plot did not suggest publication bias (Figure 6).
Two of the new trials included in the 2012 update focused on African American adults (Brown 2012; Spector 2012), and one studied Hispanic adults (Zangrilli 2011), so the diversity of ethnic groups represented is now greater. However, there remains very little in the way of separate data on adolescent participants recruited in any of the adult or adolescent trials. Separate data on adolescents must be reported in large ongoing trials initiated by the FDA (Chowdhury 2011).
Quality of the evidence
Risks of bias in the studies included in this review are thought to be low, as almost all of the studies were double-blind, and although allocation concealment was not well reported, it is likely to have been adequate, as all trials were sponsored or supported by product manufacturers. Because the trials were carried out for regulatory purposes, the collection of SAE data will have been assessed with the use of uniform definitions across studies.
Whilst we judged the risks of bias to be low in relation to all-cause fatal and non-fatal SAEs, this is not necessarily the case for asthma-related events. No independent assessment of the causation of events was undertaken, so bias may have been introduced if the investigators had a high threshold for classifying events as asthma-related. However, we would expect such a bias to decrease any differences observed in asthma-related events.
The level of heterogeneity within the subgroup of paediatric trials is significant (I
The 2012 update of this review includes three new trials that are examining the combination of formoterol and mometasone in 1248 adults (Meltzer 2012; Nathan 2010; Weinstein 2010). The number of participants is insufficient to allow investigators to compare and contrast the safety of this new combination versus the combination of formoterol and budesonide used in the other studies. However, a large ongoing trial on formoterol and mometasone is designed to recruit 11,000 adults and adolescents (NCT01471340) and will enable better safety comparison of various products in the future.
Despite the addition of new evidence to this updated review, we have downgraded the quality of evidence across a number of outcomes for imprecision ( Summary of findings for the main comparison; Summary of findings 2). The very low power of studies in children to detect any differences in SAEs means that we will have to await the results of the large ongoing study in children (NCT01471340) before we can determine whether the different directions of effect in asthma-related SAEs between children and adults are confounded or conflicting for another reason.
Potential biases in the review process
Selection of the best method to combine studies with rare events is contentious when event rates are low, not least because of the corrections required to calculate ORs with zero events (Sweeting 2004). It became apparent in the course of the review that the pooled ORs were heavily dependent on the zero adjustment used in the Mantel-Haenszel and inverse variance methods; therefore, we used the Peto OR and RDs to report results of this review. The imbalance between trial arms is never greater than two to one; therefore, the likely bias with use of the Peto OR is small (Sweeting 2004).
Similarly, the included studies were influenced by the decision to restrict the review to trials that randomly assigned participants to formoterol and ICS, but this decision reduces the risk of bias that arises when patients discontinue their usual inhaled steroid medication if they feel better while receiving the randomly assigned treatment. This presupposes a similar risk of SAEs when formoterol and budesonide are delivered via a single inhaler, and when formoterol is introduced to ICS therapy via a separate inhaler, when both are randomly assigned treatments in a controlled trial.
Agreements and disagreements with other studies or reviews
Comparison of the results of this review with those of the review on regular formoterol without randomly assigned ICS (Cates 2012) indicates that nine out of 10 deaths in the trials comparing formoterol with placebo or comparing formoterol with ICS versus the same dose of ICS were performed in participants who were randomly assigned to formoterol (with or without ICS). This is a cause for concern because although it may seem that many of the deaths were not related to asthma, it is often difficult to be sure of the exact cause of death, and the classification of cause of death is not straightforward. For example, the participant who died during the OPTIMA trial (O'Byrne 2001) was recorded by authors as dying from septic shock but was listed in Sears 2008 as dying from status asthmaticus and septic shock, whereas the 13-year-old boy who died in Von Berg 2003 is listed in Sears 2008 as dying of respiratory failure, although the article reported that the cause of death was subarachnoid haemorrhage. Sears 2008 does not report all-cause mortality in the subgroup of trials in participants receiving regular formoterol and maintenance ICS; the primary analysis on all-cause mortality included the RELIEF study, which allowed regular long-acting beta
Only one asthma-related death was reported in this review, but the overview of Sears 2008 identified two additional asthma-related deaths from the AstraZeneca database of trials in which participants were receiving maintenance ICS; all three deaths occurred amongst participants who had been randomly assigned to regular formoterol.
Six additional deaths were reported when formoterol and ICS were compared with higher doses of ICS－three deaths in each arm (Jaeschke 2008).
We agree with the conclusion of Sears 2008 that "the power is insufficient to conclude no increased mortality with formoterol" when regular formoterol is used in conjunction with ICS.
All-cause non-fatal SAEs
Sears 2008 did not present data on all-cause SAEs, but Jaeschke 2008 reported that a reduction in asthma-related SAEs and in hospitalisations does not seem to translate into similar reductions in all-cause SAEs (which are about four times more common). Jaeschke 2008 and Jaeschke 2008a did not include trials in children. Information derived from trials in children in this review is insufficient to allow us to determine whether the increased risk of non-fatal SAEs found with formoterol alone in Cates 2012 (Peto OR 2.48, 95% CI 1.27 to 4.83) is abolished by the addition of randomised ICS (Peto OR 1.62, 95% CI 0.80 to 3.28), as a large degree of overlap in CIs led to negative test findings for interaction (test for subgroup differences: Chi² = 0.74, df = 1, P = 0.39, I² = 0%) (see Figure 7). This is discussed more fully in the overview of the safety of regular formoterol or salmeterol in children (Cates 2012a) and is in agreement with the findings of McMahon 2011, who reported a significant association between younger age and increased risk of formoterol or salmeterol monotherapy, but no significant age association with combination inhalers.
|Figure 7. All-cause serious adverse events in children given regular formoterol (with or without ICS).|
Implications for practice
From the evidence provided in this review, it is not possible to reassure people with asthma that regular use of ICS with formoterol carries no risk of increasing mortality in comparison with ICS alone. On the other hand, we have found no conclusive evidence of harm, and only one asthma-related death was registered over more than 4200 patient-years of observation of individuals taking formoterol. In adults, the decrease in asthma-related SAEs seen amongst those receiving regular formoterol with ICS was not accompanied by a similar decrease in all-cause SAEs. In children, the number of events was too small to allow determination of whether the increase in all-cause non-fatal SAEs previously found among those taking regular formoterol alone is abolished by the additional use of ICS. Clinical decisions and information provided to patients regarding regular use of formoterol must take into account the balance between known symptomatic benefits of formoterol and the degree of uncertainty associated with its potential harmful effects.
Implications for research
Future research should clearly specify the number of patients with fatal and non-fatal SAEs by treatment group and cause. New large surveillance studies on combination therapies that are licensed in the United States have been mandated by the FDA; these aim to recruit 11,000 adults and adolescents to compare regular formoterol with budesonide against budesonide alone, and a similar number to compare regular formoterol with mometasone against mometasone alone, but trial investigators will not report results until 2017.
We thank Susan Hansen, Elizabeth Stovold and Emma Jackson of the Cochrane Airways Group for assistance in searching for trials and obtaining the abstracts and full reports and for extracting data on trial characteristics. We also thank Toby Lasserson for his contribution to previous versions of the review. We acknowledge the assistance of Matthew Cates in relation to the physiology of beta-agonist receptors and co-writing of the protocol, and of Marta Oleszczuk for assessing studies for inclusion in the 2012 update. We thank Joe Gray, Finn Radner and Anders Ottosson of AstraZeneca, and Davis Gates from Merck, for obtaining data on file for SAEs in these studies.
Data and analyses
- Top of page
- Summary of findings [Explanations]
- Authors' conclusions
- Data and analyses
- What's new
- Contributions of authors
- Declarations of interest
- Sources of support
- Differences between protocol and review
- Index terms
Appendix 1. Pharmacology of beta
Beta-agonists are thought to cause bronchodilatation primarily through binding of beta
The in vivo effect of any beta
Appendix 2. Possible mechanisms of increased asthma mortality with beta-agonists
This hypothesis states that direct adverse effects of beta
During the 1960s epidemic, most deaths occurred in patients with severe asthma, and it was originally assumed that asthma and its sequelae, including hypoxia, were the primary cause of death. However, mucus plugging and hypoxia do not preclude a cardiac event as the final cause of death, and one might expect those with severe asthma to take additional doses of a prescribed inhaler. As noted by Speizer and Doll, most deaths in the 1960s were in the 10- to 19-year age group, and “at these ages children have begun to act independently and may be particularly prone to misuse a self-administered form of treatment” (Speizer 1968). If toxicity were related to increasing doses of beta
Additional concerns about a possible toxic effect of beta
In patients with mild asthma and without a bronchoconstrictor challenge, salmeterol and salbutamol cause a similar degree of near-maximal bronchodilation at low doses (Bennett 1994). However, whilst as a one-off dose salbutamol is typically used at 2 to 4 times the concentration of salmeterol, the dose equivalences for salmeterol versus salbutamol in increasing heart rate and decreasing potassium concentration and diastolic blood pressure were 17.7, 7.8 and 7.6, respectively (i.e. salmeterol had a greater effect across all parameters). Given the lower intrinsic efficacy of salmeterol (Table 2), these results highlight the importance of in vivo factors; one possible explanation for the difference is the increased lipophilicity of salmeterol compared with salbutamol, contributing to higher systemic absorption (Bennett 1994).
When increasing actuations of standard doses of formoterol and salmeterol inhalers are compared in stable asthmatic patients, relatively similar cardiovascular effects are seen at lower doses (Guhan 2000). However, at highest doses (above those recommended by the manufacturers), trends towards an increase in systolic blood pressure were noted with formoterol; in comparison, a trend towards a decrease in diastolic blood pressure and an increase in QTc interval was seen with salmeterol, although no statistical analysis of the difference was performed. In contrast, in asthmatic patients with methacholine-induced bronchoconstriction, no significant difference was noted between salmeterol and formoterol in causing increased heart rate and QTc interval, although formoterol caused significantly greater bronchodilatation and hypokalaemia (Palmqvist 1999). Whilst good evidence of cardiovascular and metabolic side effects has been observed with increasing doses of beta
In this setting, the term tolerance refers to an impaired response to beta
Studies comparing salmeterol and formoterol have shown that both cause tolerance compared with placebo, but no significant difference was noted between the drugs (van der Woude 2001). There also appears to be little difference in the tolerance induced by regular formoterol and regular salbutamol treatment (Hancox 1999; Jones 2001). To the review authors' knowledge, no studies have looked specifically at the degree of tolerance caused by isoprenaline and fenoterol in the setting of acute bronchoconstriction. Tolerance to bronchodilatation has clearly been shown to occur with addition of inhaled corticosteroids to salmeterol and formoterol (Lee 2003) and terbutaline (Yates 1996). Evidence as to whether high-dose steroids can reverse tolerance in the acute setting is conflicting (Lipworth 2000; Jones 2001).
At first glance, the toxicity and tolerance hypotheses might appear incompatible, as systemic and cardiovascular tolerance ought to protect against toxicity in the acute setting, and good evidence suggests that such tolerance occurs in stable asthmatic patients (Lipworth 1989). However, whilst this study showed that changes in heart rate and potassium levels were blunted by previous beta
Whilst the tolerance hypothesis is often cited as contributing towards the asthma mortality epidemics, it is difficult to argue that reduced efficacy of a drug can cause increased mortality relative to a time when that drug was not used at all. However, tolerance to the bronchodilating effect of endogenous circulating adrenaline is theoretically possible, and evidence has revealed rebound bronchoconstriction when fenoterol is stopped (Sears 1990), which may be detrimental. Furthermore, it appears that regular salbutamol treatment can actually increase airway responsiveness to allergen (Cockcroft 1993)－a potentially important effect that could produce a variant of the toxicity hypothesis. Differences between beta
Confounding by severity
Historically, this hypothesis has been used extensively to try to explain the association between mortality and the use of fenoterol during the 1970s New Zealand epidemic (see Pearce 2007), and it is still quoted today. The hypothesis essentially relies on the supposition that patients with more severe asthma are more likely to take higher doses of beta
The delay hypothesis
This hypothesis accepts that beta
Reduced corticosteroid treatment
A slight but significant variation of the delay hypothesis suggests that patients who have separate beta
Appendix 3. Sources and search methods for the Cochrane Airways Group Specialised Register (CAGR)
Electronic searches: core databases
Handsearches: core respiratory conference abstracts
MEDLINE search strategy used to identify trials for the CAGR
1. exp Asthma/
3. (antiasthma$ or anti-asthma$).mp.
4. Respiratory Sounds/
6. Bronchial Spasm/
8. (bronch$ adj3 spasm$).mp.
10. exp Bronchoconstriction/
11. (bronch$ adj3 constrict$).mp.
12. Bronchial Hyperreactivity/
13. Respiratory Hypersensitivity/
14. ((bronchial$ or respiratory or airway$ or lung$) adj3 (hypersensitiv$ or hyperreactiv$ or allerg$ or insufficiency)).mp.
15. ((dust or mite$) adj3 (allerg$ or hypersensitiv$)).mp.
Filter to identify RCTs
1. exp "clinical trial [publication type]"/
2. (randomised or randomised).ab,ti.
11. 9 not (9 and 10)
12. 8 not 11
The MEDLINE strategy and RCT filter are adapted to identify trials in other electronic databases.
Appendix 4. Definition of Serious Adverse Events
The Expert Working Group (Efficacy) of the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) defines serious adverse events as follows (ICHE2a 1995):
"A serious adverse event (experience) or reaction is any untoward medical occurrence that at any dose:
- Results in death.
- Is life threatening.
- Requires inpatient hospitalisation or prolongation of existing hospitalisation.
- Results in persistent or significant disability/incapacity.
- Is a congenital anomaly/birth defect.
NOTE: The term 'life threatening' in the definition of 'serious' refers to an event in which the participant was at risk of death at the time of the event; it does not refer to an event that hypothetically might have caused death if it were more severe."
Last assessed as up-to-date: 1 August 2012.
Contributions of authors
CJC: Conception of the idea and co-writing of protocol with MJC. Trial selection, data extraction and co-writing the original review and the 2012 update.
RJ: Trial selection, data extraction and co-writing the review.
MF & SS: Data-extraction and co-writing the 2012 update.
Declarations of interest
None known for CJC. RJ received on one occasion honorarium and travel support from GlaxoSmtihKline for a lecture related to the topic of this review and is a deputy editor of a medical journal that is financed in part by advertising of drugs, including medications for asthma.
Sources of support
- NHS R&D, UK.
- NIHR, UK.Programme Grant (10/4001/01)
- European Union (FP7) Health, Not specified.ASTROLAB project (EC HEALTH-F5-2011-282593)
Differences between protocol and review
Peto OR was used for primary meta-analysis of ORs, as otherwise the results are largely dependent on the zero correction adopted. Single-inhaler therapy and adjustable maintenance dosing were not included in the review, nor was comparison with higher-dose ICS. This was done because we decided to restrict our attention to the question of regular use of formoterol, in addition to the same ICS regimen, in both active and control arms. Subgroup analysis was not attempted on the basis of asthma severity or dose of ICS.
Medical Subject Headings (MeSH)
Administration, Inhalation; Adrenal Cortex Hormones [*adverse effects; therapeutic use]; Adrenergic beta-Agonists [administration & dosage; *adverse effects]; Anti-Asthmatic Agents [administration & dosage; *adverse effects]; Asthma [drug therapy; *mortality]; Ethanolamines [administration & dosage; *adverse effects]; Randomized Controlled Trials as Topic
MeSH check words
Adolescent; Adult; Child; Humans
* Indicates the major publication for the study