Article Text
PDF
Abstract
BACKGROUND To study the transition from metered dose inhalers using chlorofluorocarbons as propellants (CFC-MDIs) to non-CFC containing devices, a systematic review was conducted of clinical trials which compared the delivery of salbutamol and terbutaline via CFC-MDIs and non-CFC devices.
METHODS Papers were selected by searching electronic databases (Medline, Cochrane, and BIDS) and further information and studies were sought from pharmaceutical companies. The studies were assessed for their methodological quality.
RESULTS Fifty three relevant trials were identified. Most were scientifically flawed in terms of study design, comparison of inappropriate doses, and insufficient power for the determination of therapeutic equivalence. Differences between inhaler devices were categorised according to efficacy and potency. Most trials claimed to show therapeutic equivalence, usually for the same doses from the different devices. Two commercially available salbutamol metered dose inhalers using a novel hydrofluorocarbon HFC-134a as propellant were equally as potent and efficacious as conventional CFC-MDIs, as were the Rotahaler and Clickhaler dry powder inhalers (DPIs). Evidence suggests that a dose of 200 μg salbutamol via CFC-MDI may be substituted with 200 μg and 400 μg of salbutamol via Accuhaler and Diskhaler DPIs, respectively. Terbutaline delivered via a Turbohaler DPI is equally as potent and efficacious as terbutaline delivered via a conventional CFC-MDI.
CONCLUSIONS When substituting non-CFC containing inhalers for CFC-MDIs, attention must be given to differences in inhaler characteristics which may result in variations in pulmonary function.
- inhaled bronchodilators
- salbutamol
- terbutaline
- chlorofluorocarbons (CFCs)
- metered dose inhalers
- CFC-free
- hydrofluorocarbons
- therapeutic equivalence
Statistics from Altmetric.com
- inhaled bronchodilators
- salbutamol
- terbutaline
- chlorofluorocarbons (CFCs)
- metered dose inhalers
- CFC-free
- hydrofluorocarbons
- therapeutic equivalence
The forthcoming conversion from metered dose inhalers using chlorofluorocarbons as propellants (CFC-MDIs) to non CFC-containing inhalers1 will affect millions of patients with respiratory diseases in the UK alone. A part of the conversion will be the selection of appropriate alternative inhaler devices, which is dependent upon several factors such as the determination of therapeutically equivalent alternatives, their handling and acceptability by patients, and their cost. Doctors and patients in Europe will be faced with a choice of up to 30 different metered dose inhalers using a novel hydrofluorocarbon HFC-134a as propellant (HFC-MDIs) by the year 2000.2 At the same time many doctors may take the opportunity to increase their use of dry powder inhalers (DPIs).
It is necessary to define the clinical effectiveness of the newer HFC-MDIs compared with the existing CFC-MDIs. This should be based on properly conducted trials with relevant clinical end points in preference to surrogate markers of efficacy such as drug deposition or pharmacokinetic parameters. This point is reinforced by a previous review3 which considered the relationship between clinical efficacy and lung deposition, and concluded that differences in drug deposition alone did not always explain corresponding differences in bronchodilatory responses among inhaler devices.
Important clinical differences might therefore be missed by studies with such end points. To investigate the comparability of CFC-containing and CFC-free devices we undertook a systematic review of the evidence from trials which compared the bronchodilator effects of the short acting β2 adrenoceptor agonists salbutamol and terbutaline delivered via CFC-free inhalers (DPIs and HFC-MDIs) and CFC-MDIs.
Methods
Studies for inclusion in the review were selected by searching the Medline, BIDS, and Cochrane databases. The search strategy included the use of the following “free text” terms: {salbutamol or albuterol or terbutaline}and {inhaler device(s)} and{clinical trial*}and {compar* or equivalen* or bioequivalen* or versus}and {English language}. Additional clinical trials, published or unpublished, were obtained from the medical information departments of 3M Health Care, Glaxo-Wellcome, Astra Pharmaceuticals, and Medeva pharmaceutical companies.
Of the studies identified, only those which compared two or more inhaler devices and evaluated clinical (bronchodilator) end points were included. Some studies which described the transfer from CFC-MDIs for both inhaled bronchodilators and steroids were not considered eligible for inclusion.
EVALUATION OF METHODOLOGICAL QUALITY
Studies were evaluated for methodological quality by considering the following factors: subjects, study design, treatment interventions, and clinical outcomes. Trials in healthy volunteers were excluded because of the influence of the disease state upon the response to inhaled therapy. Most studies evaluated compared the effects of drug/inhaler combinations in asthmatic patients, which enabled the severity of lung disease to be classified as mild, moderate, severe, or life threatening according to the British Thoracic Society guidelines for the management of asthma.4
With regard to study design, patient and sequence (in crossover studies) randomisation was considered essential for unbiased trial conduct. Those which were double blind and included a placebo control group were considered superior to single blind or open studies. Most trials conformed to one of three study designs: single doses taken on separate days, cumulative dosing, and chronic treatment. Some single dose studies determined changes in PD20 FEV1values (logarithm of the dose of bronchoconstrictor required to reduce the forced expiratory volume in one second (FEV1) by 20%) following the administration of bronchoprovocative agents such as histamine.
Various pulmonary function tests were used to assess the comparative efficacy of inhaler devices. Most studies measured FEV1and/or peak expiratory flow rate (PEFR) as primary outcome measures. PEFR, however, is not only associated with a greater degree of variability than FEV1 (an observation independent of the measuring device), but is also not generally regarded as the most sensitive indicator of airflow obstruction.5
To show that the test product is of comparable efficacy to a standard therapy requires that trialists validate the statistical power of the study to reduce the likelihood of falsely concluding that two inhaler devices are therapeutically equivalent. For studies where details of sample size calculations were absent, estimations of power were made according to standard methods.6 ,7 In addition, studies were scrutinised to ensure that the effect of treatment was greater than the predefined therapeutic limit (maximally tolerated clinical differences) in order to exclude trials which reported two treatments as equivalent when both were ineffective.
Results
SALBUTAMOL DELIVERING DEVICES
Six studies comparing one HFC-MDI (Airomir®, 3M) with a CFC-MDI8-13 and four comparing another (Evohaler®, Allen and Hanbury) with a CFC-MDI14-17 were found and are presented in table 1(a) and (b), respectively. More studies comparing dry powder devices with CFC-MDIs were found: 14 studies used a Rotahaler® (Allen and Hanburys),18-29 three used a Diskhaler®(Allen and Hanburys),30-32 two used an Accuhaler® (Allen and Hanburys),33 ,34 and three a Clickhaler® (Medeva).35-37 The results of these studies are presented in table 1(c)–(f).
Summary of trials comparing salbutamol delivered via CFC-free devices and via CFC-MDI
TERBUTALINE DELIVERING DEVICES
Twenty one studies were found which compared a dry powder device (Turbohaler®, Astra) with a CFC-MDI38-58 and the results are presented in table 2.
Summary of trials comparing terbutaline delivered via Turbohaler (TBH) and via CFC-MDI
QUALITY OF STUDIES
Most studies identified were of inferior methodological quality, mainly because of inadequate blinding, absence of a placebo control group, and failure to randomise. Many trials were reported as abstracts or not published at all and so were devoid of substantial details for critical appraisal. This may partly be due to the reluctance of many journals to publish such evaluations. The most frequently encountered flaw was that many were designed as comparative or (superiority) trials with null hypotheses of equal efficacy and were therefore underpowered to detect inequivalence, the real concern of equivalence studies. The absence of high quality evidence and substantial variation in the nature of many of the studies prevented any attempt at more formal statistical analysis.
The summary of evidence for alternatives to CFC-MDIs has been divided into two categories: (1) those which, when administered at the same dose, are interchangeable and (2) those where there is only sufficient evidence for substitution at given dose combinations. These are presented in table 3.
Therapeutically equivalent alternatives to short acting β2 adrenoceptor agonists delivered via CFC-MDIs
Discussion
Salbutamol and terbutaline are the most widely used short acting β2 agonist bronchodilators with CFC-based MDIs of these drugs accounting for 83% of all bronchodilator delivery devices sold in the UK. Most patients will change device but not drug, and it is important that it should be clear which CFC-free devices are clinically equivalent. Other issues important in ensuring a smooth transition such as patient acceptability, patient education, and cost are not addressed here.
The practitioner of evidence based medicine will have considerable difficulty in identifying studies of adequate quality to assist in the choice of device and will be forced to depend on the often unpublished evidence presented to national licensing agencies, and on the interpretation of these studies by the licensing agencies. In the UK the Medicines Control Agency requires new non-CFC containing products to show therapeutic equivalence to existing products containing CFC. The guidance59 states that this is best obtained from pharmacodynamic, single dose, short term studies by, for example, demonstrating equivalent dose and time-dependent increases in pulmonary function following single inhaled doses in asthmatic patients. Doses used in the trial should ensure that clinically relevant differences are shown. If therapeutic equivalence is not demonstrated, dose ranging studies are required. A final concern for the evidence based practitioner will be that the only widely available source of information about the outcome of the licensing procedures is indirect in the form of pharmaceutical industry advertising.
A key issue in the studies presented here is that they are often not capable of demonstrating the equivalence which they claim to address. If there appeared to be no differences between treatment groups, for instance, the null hypothesis was not rejected and the investigators interpreted the data as showing that the two treatments were equivalent. In properly designed equivalence trials, however, the conventional significance test has little relevance: failure to detect a difference does not imply equivalence. The null hypothesis should not be “equivalence”—that is, that there is no difference between the treatments—but rather “inequivalence”—that is, that there is a difference. Rejecting this hypothesis then leads to a correct interpretation of both treatments being statistically and clinically equivalent.6
Such trials require an increased sample size to provide appropriate statistical power and many of the trials evaluated in this review were underpowered. As an example, one study claimed that the Turbohaler was an effective alternative to CFC-MDI for the delivery of an identical dosage of terbutaline.43 Twelve subjects completed the crossover trial. To estimate the true sample size required for 80% statistical power to deem that both inhalers were therapeutically equivalent, however, a total of 138 subjects were required. This is based upon a sample size formula for a one sided interval7with the therapeutic equivalence limit for FEV1 taken as ±0.3 l (a maximally tolerated clinical difference) and the intra-subject standard deviation in FEV1 as 1.0 l (taken from the study). The inadequacy of including only 12 subjects is clear.
Therapeutic equivalence is determined by the observation of equal effects with two inhaler/drug combinations independent of dose, drug (except for pharmacological class), or inhaler type. If two combinations show therapeutic equivalence at whatever dose of each, this result applies strictly to those doses and cannot be extrapolated to all doses. A common misinterpretation by authors in the studies reviewed was to claim equal potency of two drug/device combinations on the basis of these studies looking at therapeutic equivalence.
A further issue is the use of non-comparable doses. If, for example, subjects were given salbutamol via inhaler A at a higher dose than actually required (on the plateau of the dose-response curve), then a smaller dose administered via inhaler B may appear to be equally efficacious and the interpretation of equivalence of potency will be erroneous.60 To this end, dose ranging studies are more informative for establishing potency differences between drug/inhaler combinations than chronic treatment studies, even though the latter may be a better reflection of the therapeutic use of these drugs.
Dose recommendations derived from the comparative clinical trial data differ somewhat from those in current practice. Recommendations in the British National Formulary61 suggest that, for salbutamol, the doses of DPIs should be twice those of CFC-MDIs. This does not seem to be the case for the Rotahaler or the Accuhaler, although there is a shortage of evidence for the latter.
For the purpose of substituting non-CFC containing inhalers for current CFC-MDIs, we found no evidence to suggest that HFC-MDIs are inappropriate. The quantity of evidence in favour of substituting HFC-MDIs, however, is rather limited compared with the more established dry powder devices. In addition to establishing therapeutically equivalent alternatives, the other factors identified above which are beyond the scope of this review must also be considered when deciding on alternatives.
Acknowledgments
The authors thank 3M Health Care for financial support.