Copyright ©ERS Journals Ltd 2003 Formoterol as relief medication in asthma: a worldwide safety and effectiveness trial1 Dept of Respiratory Diseases, Ghent University Hospital, Ghent, Belgium, 2 Firestone Institute for Respiratory Health, St. Joseph's Healthcare and McMaster University, Hamilton, Canada, 3 Dept of General Practice, University of Glasgow, Glasgow, UK, 4 Dept ofRespiratory Diseases, Hospital Universitario LaPaz, Madrid, Spain, 5 Pulmonary Dept, Rui Jin Hospital, Shanghai Second Medical University, Shanghai, China, 6 Husläkarna Österåker, Åkersberga, Sweden, 7 Dept of Internal Medicine, Brüderkrankenhaus, Paderborn, Germany, 8 Dept of Pneumology, Hôpital Bichat Claude Bernard, Paris, France, and 9 Clinical Development, AstraZeneca R&D Lund, Lund, Sweden CORRESPONDENCE: R. Pauwels, Dept of Respiratory Diseases, University Hospital, De Pintelaan 185, B9000, Ghent, Belgium. Fax: 32 92402341. E-mail: Romain.Pauwels@UGent.be Keywords: age, asthma, effectiveness, formoterol, safety, salbutamol
Received: May 19, 2003
The aim of the study was to compare the safety and effectiveness of as-needed formoterol with salbutamol in a large international real-life asthma study. Children and adults (n=18,124) were randomised to 6 months as-needed treatment with open-label formoterol 4.5 µg Turbuhaler® or salbutamol 200 µg pressurised metered dose inhaler or equivalent. Primary safety variables were asthma-related and nonasthma-related serious adverse events (SAE)s and adverse events (AE)s resulting in discontinuation (DAE)s. The primary efficacy variable was time to first asthma exacerbation. The incidences of AEs, SAEs and DAEs arising from SAEs were not significantly different between treatments. DAEs for nonserious AEs were higher with formoterol. Asthma-related AEs decreased with formoterol (1,098 (12.3%) versus 1,206 (13.5%)), asthma-related SAEs were similar (108 (1.2%) versus 121 (1.4%)) but more asthma-related DAEs occurred in the formoterol group (89 (1.0%) versus 48 (0.5%)). Time to first exacerbation was prolonged (hazard ratio 0.86) and less as-needed and maintenance medication was used with formoterol. Reductions of exacerbations with as-needed formoterol versus salbutamol increased with increasing age and asthma medication level. This real-life study demonstrates that formoterol as-needed has a similar safety profile to salbutamol, and its use as a reliever therapy is associated with fewer asthma symptoms and exacerbations. Guidelines for treatment of asthma recommend regular use of anti-inflammatory therapy for any form of persistent asthma and a minimal use of reliever medication as-needed 1, 2. Over the last decade, long-acting ß2-agonists (LABA) have become widely used as regular maintenance treatment in conjunction with inhaled corticosteroids (ICS) 3, 4. Rapid-acting ß2-agonists, such as salbutamol and terbutaline, are established reliever medications, but have a relatively short duration of bronchodilator effect. Formoterol is a unique ß2-agonist, being both rapid and long acting 5, 6. In mild, moderate and severe persistent asthma, the use of formoterol as maintenance therapy or as-needed, has resulted in improvements in asthma control 79. The 3-month study by Tattersfield et al. 9 demonstrated that the use of formoterol as reliever medication reduced exacerbations compared with terbutaline, without compromising safety in patients with moderate-persistent asthma who used frequent doses of reliever medication despite regular ICS. However, that study raised questions as to whether these observations of safety and efficacy could be extended to a wider population of patients with asthma. The safety and efficacy of ß2-agonists may vary according to age, severity of disease, and concomitant medication 10. With respect to formoterol, a key question is safety and efficacy when used as-needed in conjunction with regular LABA. The safety of formoterol as-needed, in the absence of ICS, is also of relevance given the potential for LABA to mask an underlying deterioration in airway inflammation 11. This 6-month "real-life" study was designed to assess the safety and effectiveness of formoterol as reliever medication, compared with the most widely used reliever therapy, salbutamol, in asthmatics over a wide age range, with different degrees of asthma severity and receiving a variety of other maintenance medications. As safety was the primary focus and in order to mimic clinical practice, the study was run open-label to allow each of the participating countries to compare formoterol with salbutamol by the most appropriate inhaler. Data collected by the patient and investigator focussed on adverse events (AE)s and exacerbations whereas collection of other efficacy data was minimised.
Study subjects Outpatients from general practice and specialist centres, aged 6 yrs, with a clinical diagnosis of asthma and using, or candidates for receiving, a ß2-agonist as reliever medication, were eligible. Females who were pregnant, breastfeeding or not using an acceptable method of contraception were excluded. To mimic a normal prescribing situation, patients with concomitant cardiovascular diseases were included at the discretion of the treating physician. Prescribing information indicated a need for caution in patients with thyrotoxicosis, ischaemic heart disease, tachyarrhythmias, severe heart failure or prolonged Q-T interval corrected for heart rate. Written informed consent was obtained from all adult patients and from the parent or legal guardian of all children. Written or oral consent was obtained from all children. The open, randomised, parallel-group study was carried out at 1,139 centres in 24 countries. Approval was obtained from regulatory agencies and ethics committees at all centres.
Study design
The primary efficacy variable was time to first exacerbation. An exacerbation was defined by one or more of the following: 1) any increase in maintenance asthma medication; 2) a course of oral corticosteroids lasting At entry, patients were randomised in chronological order at each site, according to a computer-generated code, and treatment communicated via code envelope. Patients were assigned to one of two treatment as-needed regimens, i.e. either formoterol 4.5 µg per dose, via Turbuhaler® (Oxis®, AstraZeneca, Södertälje, Sweden), or salbutamol 200 µg per dose via pressurised metered dose inhaler (pMDI) or equivalent dose via dry powder inhaler (DPI). Eighteen countries used salbutamol via pMDI, one via Diskhaler® (200 µg per dose; VentolinTM, GlaxoSmithKline, Uxbridge, UK), two via Diskus® (200 µg per dose; VentolinTM, GlaxoSmithKline, Uxbridge, UK), and three via Turbuhaler® (100 µg per dose; Inspiryl®, AstraZeneca, Sweden). The open-label design allowed each of the 24 participating countries to compare formoterol Turbuhaler® with salbutamol by the most appropriate inhaler. Formoterol 4.5 µg via Turbuhaler® is an equipotent bronchodilator dose to salbutamol 200 µg via pMDI 1214, which was selected as the comparator in 18 countries. In six countries salbutamol was administered via DPIs ("DPI countries") at a dose equivalent to 200 µg via pMDI 15. The use of double-dummy placebo inhalers to blind the study was considered to pose an unacceptable risk that patients in need of reliever medication during an acute attack could inadvertently use a placebo. Patients attended the clinic at entry to the study and after 1, 3 and 6 months of treatment. Asthma maintenance treatment was recorded at entry and at the final visit. During the study, the investigators could change the maintenance treatment according to their clinical judgment. Patients were instructed to contact the investigator if their use of reliever medication exceeded 12 inhalations per day in adults and eight inhalations per day in children. Patients on regular treatment with a LABA (formoterol or salmeterol) were instructed to contact the investigator if any of the following occurred: 1) daily use of the study medication exceeded 10 and six inhalations in adults and children, respectively, if using regular treatment with formoterol 4.5 µg b.i.d.; 2) eight and four inhalations in adults and children, respectively, if using regular treatment with formoterol 9 µg b.i.d. or salmeterol 50 µg b.i.d.; and 3) four inhalations in adults using either formoterol 18 µg b.i.d. or salmeterol 100 µg b.i.d. The investigator could then decide on appropriate action. The patient or parent/legal guardian filled in a notebook distributed to each patient at entry to the study. Any unscheduled healthcare contacts due to asthma, the number of days incapacity due to asthma, and changes in concomitant asthma medication were recorded in the notebook for the entire 6 months. Patients were contacted by telephone to remind them to record daily symptoms and use of study medication during the 2 weeks preceding each scheduled clinicvisit. At each clinic visit, the investigator recorded spontaneously reported and/or observed AEs, including deterioration of any pre-existing medical condition, such as asthma. The number, duration and first occurrence of hospitalisations, the number and first occurrence of emergency treatments, courses of oral corticosteroids, and increases in asthma maintenance medication, and the number of days on which the patient was incapable of performing normal activities were recorded. In addition, over the previous 2 weeks, the total number of inhalations of as-needed study medication and the total number of days with asthma symptoms were recorded. At visit four or at discontinuation from the study, additional information on AEs was collected by means of a standard question "Have you (Has your child) had any health problems since visit one?"
Analysis The primary safety variables were asthma-related and nonasthma-related SAEs and DAEs. All events were characterised on a preferred term level, counting patients only once for a particular AE, even if the subject experienced multiple occurrences of that AE during the treatment period. The numbers of patients experiencing at least one AE, SAE, DAE or subcategory thereof were compared between the treatment groups using a Chi-squared test. However, in the safety evaluation, p-values were used as flags to indicate possible findings. The overall evaluation of safety was based on all aspects of AEs, not just the primary variables. Time to first exacerbation was analysed using a Cox proportional hazards model adjusting for treatment, asthma medication level at baseline, age and geographical region. The average use of study drug per day and the percentage of days with asthma symptoms were compared between treatments using a Linear Mixed Effects Model adjusting for treatment, period and interaction treatment by period. The distribution of asthma medication levels at end of the study was compared using a Generalised Linear Model (proportional odds) adjusting for treatment and baseline asthma medication level. The number of days when subjects were unable to conduct normal activities due to asthma was compared using an analysis of variance model with treatment as factor and days in study as a covariate.
A priori defined analyses were performed in patient subgroups classified by age and by asthma medication level at entry. The age categories were children (611 yrs), adolescents (1217 yrs), adults (1864 yrs) and the elderly (>64 yrs). Asthma severity (intermittent, mild, moderate and severe) was defined by the use of maintenance treatment at entry, classified according to recommendations of the Global Initiative for Asthma (GINA) guidelines (table 1
All analyses were undertaken according to intention to treat. A p-value of <0.05 was considered statistically significant. All tests were two-sided (where applicable).
Of 18,132 patients (aged 491 yrs) enrolled, 18,124 were randomised to receive formoterol (n=9,064) or salbutamol (n=9,060). Randomised patients (formoterol 140, salbutamol 122) who did not receive any study treatment or had no data recorded were not included in the analysis. A total of 1,189 discontinued the study (formoterol: 664 (7.3%), salbutamol: 525 (5.8%); p<0.001) due to the following: 1) lost to follow-up (formoterol: 211 (2.3%), salbutamol: 204 (2.3%)); 2) AEs (formoterol: 213 (2.4%), salbutamol: 119 (1.3%)); 3) eligibility criteria not fulfilled (formoterol: 12 (0.1%), salbutamol: 21 (0.2%)); or 4) other reasons (formoterol: 228 (2.5%), salbutamol: 181 (2.0%)). High numbers of patients completed the 6-month study (formoterol: 8260 (93%), salbutamol: 8413 (94%)) and the mean treatment duration was comparable between the groups (formoterol: 173 versus salbutamol: 175 days).
Demographical data were well balanced between the treatment groups (table 1
Safety
In all, 305 SAEs occurred in 278 patients (3.1%) using formoterol, including 13 deaths (0.1%), compared with 327 SAEs in 299 patients (3.3%) using salbutamol, including 11 deaths (0.1%) (table 2 DAEs in the study were few. There were more DAEs in the formoterol group (213 (2.4%) versus 119 (1.3%); p<0.001) but this difference in DAEs was not due to SAEs (40, 0.4% formoterol versus 37, 0.4% salbutamol; p=0.73) but rather to nonserious AEs (173, 1.9% formoterol versus 82, 0.9% salbutamol; p<0.001). The number of asthma-related DAEs was significantly higher with formoterol (89, 1.0% versus 48, 0.5%; p<0.001). A statistically significant treatment interaction for asthma-related DAEs by inhaler type (countries with DPI or pMDI) was present (p=0.029). In the six DPI countries (n=5,056) the incidence of asthma-related DAEs was similar with formoterol and salbutamol (28, 1.1% versus 24, 1.0%; p=0.61), whereas a significant treatment difference was seen in the 18 pMDI countries (n=12,806) (61, 1.0% versus 24, 0.4%; p<0.001). However, also within the pMDI countries there was a large variation, indicating that the increased rate of asthma-related DAEs in a few countries was not representative of the whole study population. There were significantly more nonasthma-related DAEs in the formoterol group (124, 1.4% versus 71, 0.8%; p<0.001). No treatment interaction by inhaler type was seen (p=0.86), with about the same increase in DPI countries (formoterol 48, 1.9% versus 28 1.1%) as in pMDI countries (formoterol 76, 1.2% versus 43, 0.7%). Among symptoms, statistically significant differences were found for tremor (formoterol 19, 0.2% versus six, 0.1%; p=0.009), headache (formoterol 14, 0.2% versus three, 0.1%; p=0.008) and tachycardia (formoterol nine, 0.1% versus two, 0.1%; p=0.034).
There were no differences between the study treatments for safety variables related to age, asthma medication levels, or concomitant ICS or LABA treatment (table 3
Efficacy Fewer patients using formoterol experienced at least one exacerbation of any category (2,549 (28.6%) versus 2,893 (32.4%)), or at least one severe exacerbation (1,181 (13.2%) versus 1,343 (15.0%)). The time to first asthma exacerbation of any category (primary efficacy variable) was significantly longer in the formoterol group compared with the salbutamol group (fig. 2a
Patients in each age group and in each level of baseline asthma medication had longer times to first exacerbation with formoterol compared with salbutamol (table 3
During the study, the use of reliever medication in the overall population decreased with both treatments, with a significant difference in favour of formoterol at each time point (table 4
The percentage of days with asthma symptoms decreased in both groups during the study. Patients using formoterol had a significant reduction in days with asthma symptoms versus salbutamol in all periods (table 4
The majority of patients (79% in both groups) had the same asthma medication level (table 1
This international real-life study performed in a widely diverse population of >18,000 asthmatics showed that using the rapid and long-acting ß2 aganist formoterol as reliever medication was as safe as using the rapid and short-acting ß2-agonist salbutamol. In addition, use of formoterol as reliever medication resulted in a prolonged time to a first asthma exacerbation and reduced medication requirements. The real-life study design had several unique features to maximise recruitment of a diverse population and to approach normal clinical practice. These included minimal entry criteria, no run-in period, and no lung function or compliance measurements. Daily records of symptoms and reliever use were only collected for 2 weeks before each post-randomisation visit. Classification of patient's asthma severity was based only on levels of maintenance treatment at entry in relation to GINA guidelines 16, but asthma control was not assessed before randomisation. The salbutamol dose for comparison, 200 µg via pMDI, was selected to provide equivalent bronchodilator effect to formoterol 4.5 µg via Turbuhaler® 1214. The open-label trial design allowed formoterol Turbuhaler to be compared with salbutamol via any delivery device, especially pMDI, which is the most widely used delivery device for reliever medication. Blinding using double-dummy placebo relief medication was excluded for safety reasons. An open design was deemed appropriate, as the primary study focus was safety. The use of formoterol as reliever medication was not associated with any increase in AEs, SAEs, cardiovascular side-effects or discontinuations due to SAEs compared with salbutamol. No differences were observed in the incidence ofasthma-related and cardiovascular-related deaths. Well-known side-effects of ß2-agonists, such as headache and tremor, were more frequent in the formoterol group but the difference compared to salbutamol was very small (increased incidence affecting around one in 250 patients). In the salbutamol group, 1.2% more patients had an asthma-related AE and 3.8% more patients had asthma exacerbations as defined. In contrast, fewer (0.5%) salbutamol patients discontinued the study due to asthma-related AEs. No difference in rates of asthma-related DAEs was found in DPI countries, whereas an unevenly distributed difference was found in the pMDI countries, making it possible that the open study design contributed to this difference in asthma-related DAEs in favour of salbutamol. As there were no differences in the total numbers of AEs or SAEs, and the difference in DAEs was only due to nonserious AEs, as-needed formoterol can be considered to have a similar safety profile to salbutamol in asthmatic patients. The current study confirms that formoterol, when used as-needed, reduces exacerbations of asthma as demonstrated previously in a double-blind trial 9. Furthermore, in the present study formoterol as-needed also reduced exacerbations when added to maintenance LABA. In contrast to several double-blind trials 79, only clinical criteria (asthma events) were used to define exacerbations in this real-life study. An obvious limitation of the study was that the investigators were not blinded to treatment and this may have influenced study outcomes, e.g. excessive reliever use or symptoms recorded in the notebook could make the investigator increase maintenance treatment or prescribe an oral steroid course to a greater extent in one or other group depending onexpectations or experience. However, relative risks of experiencing severe exacerbations, such as hospitalisation or emergency treatment, were reduced by formoterol as-needed to a similar extent as exacerbations classified by an increase in maintenance therapy or the need for a course of oral steroids. Furthermore, whilst increases in maintenance therapy often coincided with scheduled clinic visits, reflecting normal clinical practice, the occurrence of severe exacerbations was not temporally associated with scheduled visits. The analyses by age and asthma maintenance level were performed in subgroups larger than many previous trials with formoterol 79, and are therefore useful additions to the overall analysis. Any differences in safety profiles between formoterol and salbutamol in the subgroups were consistent with the overall findings. There were no signs of loss of asthma control with as-needed formoterol in any subgroup, including patients not using concomitant ICS. When LABAs were first introduced there were concerns that they could have similar adverse effects to regular use of short-acting inhaled ß2-agonists 17. However, a large body of clinical studies have established the safety of LABA use, especially in combination with ICS therapy 1821. Data from the current study indicate that maintenance use of ICSor LABA do not affect the safety or efficacy profiles of formoterol compared with salbutamol. The subgroup analyses provide evidence of good safety and efficacy of formoterol as-needed versus salbutamol in mild-intermittent to severe-persistent asthma, in all age groups from 6-yr-olds to the elderly and in patients treated with or without maintenance LABA and ICS therapy. These are important findings, since using a rapid and long-acting ß2-agonist, both as reliever and as maintenance therapy, could increase the simplicity and convenience of bronchodilator treatment for many patients. Asthma guidelines currently advocate stepwise increases inmaintenance therapy to control asthma. In this study, clinicians could alter maintenance therapy as judged appropriate. Nevertheless, the increased effectiveness of as-needed formoterol versus salbutamol in reducing all exacerbations, severe exacerbations and achieving greater reductions in maintenance and as-needed therapy, suggests that the choice of reliever therapy may also be important in optimising asthma control. In conclusion, this real-life study has shown that formoterol as-needed has a similar safety profile to salbutamol, and in this open study its use as a reliever therapy was associated with fewer asthma symptoms and exacerbations.
The authors would like to thank all the investigators for their active participation in the RELIEF (REal LIfe EFfectiveness of Oxis® Turbuhaler® as-needed in asthmatic patients) study.
For editorial comments see page 723.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||