Abstract
Long-term oral corticosteroid (OCS) use in patients with severe asthma is associated with significant adverse effects.
This 40-week, randomised, double-blind trial evaluated the OCS-sparing potential of tralokinumab in patients with severe, uncontrolled asthma requiring maintenance OCS treatment plus inhaled corticosteroids/long-acting β2-agonists. Overall, 140 patients were randomised to tralokinumab 300 mg or placebo (n=70 in each group) administered subcutaneously every 2 weeks. The primary end-point was percentage change from baseline in average OCS dose at week 40, while maintaining asthma control. Secondary end-points included proportion of patients with a prescribed maintenance OCS dose of ≤5 mg, those with a ≥50% reduction in prescribed maintenance OCS dose and asthma exacerbation rate. Safety was also assessed.
At week 40, the percentage reduction from baseline in the final daily average OCS dose was not significantly different between tralokinumab and placebo (37.62% versus 29.85%; p=0.271). There were no significant between-treatment differences for any secondary end-point. Overall, reporting of adverse events and serious adverse events were similar for the tralokinumab and placebo groups. Although a greater proportion of tralokinumab-treated patients reported upper respiratory tract infections (35.7% versus 14.3%), there were no reported cases of pneumonia.
Overall, tralokinumab did not demonstrate an OCS-sparing effect in patients with severe asthma.
Abstract
Tralokinumab did not demonstrate significant OCS-sparing benefits versus placebo in severe asthma http://ow.ly/kzCt30mC43j
Introduction
Asthma is a heterogeneous disease characterised by chronic airway inflammation and hyperresponsiveness that affects ∼334 million people worldwide [1]. Of these, an estimated 5–10% of patients have severe asthma that remains inadequately controlled despite treatment with high-dose inhaled corticosteroids (ICS)/long-acting β2-agonists (LABA) [2]. These patients may be prescribed maintenance oral corticosteroids (OCS) to manage their symptoms and/or to prevent exacerbations [2]. However, cumulative use (including frequent and intermittent use) of OCS is associated with significant adverse effects [3, 4], which in turn diminish the patient's health-related quality of life [4]. Therefore, new therapies that reduce the need for frequent OCS exposure are required in the setting of severe asthma.
Interleukin (IL)-13, a pleiotropic cytokine that can induce inflammation [5, 6], has been identified as a potential therapeutic target in patients with severe, uncontrolled asthma [7, 8]. Tralokinumab, an immunoglobulin G4 human monoclonal antibody, potently and specifically neutralises IL-13, thereby inhibiting signalling through the IL-13 receptor [9, 10]. In phase 2 trials in patients with severe, uncontrolled asthma, tralokinumab improved lung function, but did not reduce the annual asthma exacerbation rate (AAER) or improve measures of asthma control [11, 12]. However, in a post hoc subgroup analysis of a subpopulation of patients with evidence of an activated IL-13 axis, such as elevated levels of serum periostin or dipeptidyl peptidase-4, tralokinumab improved the AAER [11]. This observation suggested that certain subpopulations of patients with severe asthma might respond to tralokinumab treatment. Two large phase 3 trials, STRATOS 1 and 2, reported ahead of TROPOS, confirmed that tralokinumab did not improve the AAER in the all-comers population with severe asthma [13]. In addition, in contrast to the phase 2 trial, periostin and dipeptidyl peptidase-4 were not shown to predict response to tralokinumab treatment in either STRATOS 1 and 2. However, these trials did suggest that in a subpopulation of patients with severe asthma with fractional exhaled nitric oxide (FeNO) concentrations ≥37 ppb, there might be an enhanced benefit with tralokinumab [13].
Given the need to reduce the requirement of OCS in patients with severe asthma, treatments that may allow tapering of OCS without loss of disease control are needed. However, at the beginning of this trial, no clinical trial had been conducted in patients with severe asthma on maintenance OCS with agents that attenuate IL-13 signalling. Consequently, the purpose of this phase 3 TROPOS trial (NCT02281357) was to evaluate the ability of tralokinumab to reduce OCS use in patients with severe asthma requiring maintenance OCS treatment in combination with ICS/LABA. The primary objective was to determine whether add-on treatment with tralokinumab provided OCS-sparing benefits compared with placebo. Secondary objectives were to assess the effect of tralokinumab on the proportion of these patients with a prescribed OCS maintenance dosage ≤5 mg at the end of the treatment period, and the proportion of patients with ≥50% reduction in prescribed OCS maintenance dosage, both compared with placebo. In addition, the effect of tralokinumab on the AAER was assessed.
Methods
Trial design and patients
Full details of the trial design have been published previously [14]. In brief, this was a randomised, double-blind, placebo-controlled, parallel-group, multicentre, phase 3 trial in patients with severe, uncontrolled asthma requiring maintenance ICS/LABA and OCS. Male and female patients aged 12–75 years were eligible and were required to have the following criteria for inclusion: asthma for ≥12 months, with a daily requirement of medium- or high-dose ICS (total daily dose ≥500 µg fluticasone propionate dry powder or an equivalent delivered dose) for ≥6 months of the 12 months prior to enrolment; physician-prescribed ICS (total daily dose ≥500 µg fluticasone propionate dry powder formulation equivalent) and a LABA for ≥3 months prior to enrolment; OCS treatment for 6 months prior to visit 1; and a stable OCS daily dose between ≥7.5 mg and ≤30 mg (prednisone or prednisolone equivalent) daily or daily equivalent for ≥1 month prior to enrolment. Full inclusion and exclusion criteria are provided in section 1 of the online supplementary appendix.
After initial enrolment (visit 1), patients entered either a 2-week run-in period (if there had been a documented failure of OCS dose reduction ≤6 months prior to visit 1) or a 2-week run-in period plus an up to 8-week optimisation period (the optimisation period was shorter in some patients (i.e. those in whom the optimal dose was reached earlier)) to establish a minimum effective OCS dose (established by dose titration every 2 weeks (Q2W); figure 1).
Eligible patients were randomised in a 1:1 ratio at week 0 (baseline) to receive either tralokinumab 300 mg or placebo s.c. Q2W and entered a 40-week treatment period comprising three phases: a 12-week induction phase, a 20-week OCS dose-reduction phase and an 8-week maintenance phase. Randomised patients were stratified by age group (adults versus adolescents) and baseline OCS dose (adults only; ≤10 mg versus >10 mg prednisone or prednisolone). The follow-up period consisted of two visits at weeks 44 and 54 for safety assessments. Details on randomisation and blinding (including processes used to administer the trial drug) and full trial procedures are provided in section 2 of the online supplementary appendix.
This trial was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice (International Conference on Harmonisation) and the AstraZeneca policy on Bioethics. The trial was approved by the independent ethics committees at all participating centres and all patients provided written informed consent.
End-points
The primary efficacy end-point was the percentage change from baseline in the final daily average OCS dose at week 40, while maintaining asthma control. Secondary end-points included the proportion of patients with a prescribed maintenance OCS dose of ≤5 mg, the proportion of patients with a ≥50% reduction in prescribed maintenance OCS dose and the AAER up to week 40. An asthma exacerbation was defined as worsening of asthma that required a temporary increase in systemic corticosteroids for ≥3 days or that resulted in an emergency-room or urgent-care visit resulting from asthma that led to a temporary increase in systemic corticosteroids for ≥3 days to treat symptoms or an inpatient hospitalisation due to asthma.
Exploratory end-points included the proportion of patients with a decrease from baseline in their final daily average OCS dose, percentage and least squares (LS) mean absolute change from baseline in prebronchodilator forced expiratory volume in 1 s (FEV1), change from baseline in the Asthma Control Questionnaire (ACQ)-6 and Asthma Quality of Life Questionnaire (AQLQ; standardised for patients aged ≥12 years) scores and an assessment of the relationship between baseline biomarker (FeNO) and the effect of tralokinumab on OCS dose reduction and clinical efficacy. Safety end-points included the incidence and frequencies of adverse events and serious adverse events (SAEs) and collection of blood samples for determination of clinical chemistry, haematology and urinalysis. Safety aspects of the trial were monitored by an independent Data and Safety Monitoring Board. All hospitalisations, emergency-room and urgent-case visits, malignancy events and cardiovascular/cerebrovascular events were adjudicated by an independent committee of experts. Potential anaphylaxis events were evaluated by a blinded external evaluator. Additional details on the assessment of trial end-points are provided in section 3 of the supplementary appendix.
Statistical analysis
It was estimated that ≥55 patients per group would be required for the trial to detect a difference of 50% in the primary end-point between tralokinumab and placebo, with 90% power, using a two-sided test at a 5% significance level. Based on the findings from an analysis of the STRATOS 1 trial, which identified FeNO high (≥37 ppb) as a biomarker-positive population [14], the trial protocol was amended to include patients with FeNO ≥37 ppb as the primary analysis population, provided that >50% of the trial population had high FeNO levels (≥37 ppb). If <50% of the trial population had FeNO ≥37 ppb, the primary analysis population would instead include patients with FeNO ≥30 ppb. If <50% of the population had FeNO ≥30 ppb, the all-comers population (i.e. all patients with any level of FeNO) would become the primary population for analysis. To account for multiplicity, a hierarchical testing strategy was used for the primary and secondary outcomes. The statistical analyses were performed by Biometrics and Information Sciences, AstraZeneca, using SAS (version 9.4; SAS Institute, Cary, NC, USA) and additional validated software (where appropriate). Further details of the statistical analysis methodology, analysis of the primary, secondary and exploratory end-points can be found in section 4 of the supplementary appendix. The impact of missing data on primary, secondary and exploratory end-points were explored using sensitivity analyses (section 5 of the supplementary appendix).
Results
Patient demographics and baseline disease characteristics
The trial was conducted from February 2015 through September 2017. A total of 218 patients were enrolled from 56 participating sites in Europe and USA. Of these, 140 patients from 44 sites were randomised to treatment with either tralokinumab (n=70) or placebo (n=70), of whom 129 (92%) completed the trial and 124 (89%) completed treatment (figure 2). Reasons for trial withdrawal are summarised in supplementary table S1. Patient baseline demographics and clinical characteristics were well balanced across treatment groups (table 1 and supplementary table S2). At baseline, patients had a mean age of 54.7 years, with the median time since asthma diagnosis being 24.0 years. The majority of patients were female (62.1%). The mean±sd baseline FeNO level was 35.22±29.50 ppb; ∼51% of patients in each treatment group had FeNO levels <30 ppb. Thus, the primary population for efficacy analysis was the all-comers population.
Overall, 48 (69%) patients in each treatment group entered the 8-week OCS dose-optimisation phase prior to randomisation, with an average dose reduction of 1.36 mg in the tralokinumab group and 0.99 mg in the placebo group (supplementary table S3).
Efficacy
Primary end-point
At week 40, the percentage reduction from baseline in the final daily average OCS dose (LS means) was not significantly different between the tralokinumab and placebo groups (37.62% versus 29.85%; p=0.271 (table 2 and figure 3). The change in OCS dose over time suggested that there was no treatment effect with tralokinumab at earlier time points (figure 3).
Secondary end-points
There were no statistically significant differences observed between the tralokinumab and placebo groups for any secondary end-point (table 2). The proportion of patients with a final daily average OCS dose of ≤5 mg at week 40 was 45.7% in the tralokinumab group versus 40.0% in the placebo group (OR 1.33, 95% CI 0.65–2.73; p=0.442). Overall, 44.3% of patients had a ≥50% reduction from baseline in the final daily average OCS dose at week 40 in the tralokinumab group compared with 37.1% of patients in the placebo group (OR 1.38, 95% CI 0.70–2.74; p=0.356).
Some 67.1% of patients treated with tralokinumab experienced at least one asthma exacerbation during the 40-week treatment period, with 925 total exacerbation days, compared with 75.7% of patients treated with placebo, with 1201 total exacerbation days (supplementary table S4). The AAER was 20% lower in the tralokinumab group through week 40 (1.84 versus 2.31; rate ratio 0.80, 95% CI 0.57–1.12; p=0.186).
Exploratory end-points
The proportion of patients with an OCS dose reduction of ≥75% to <90% was 21.4% in the tralokinumab group compared with 8.6% in the placebo group (supplementary table S5). Although no statistically significant difference was noted between the tralokinumab and placebo groups in the LS mean percentage change from baseline in prebronchodilator FEV1 at week 40 (12.1% versus 4.3%; figure 4a), a nominally statistically significant and clinically meaningful improvement in LS mean absolute change from baseline in prebronchodilator FEV1 (litres) was noted at week 40 for tralokinumab compared with placebo (LS mean difference 0.17, 95% CI 0.05–0.29; p=0.0053; figure 4b). No statistically significant improvements were observed in mean changes from baseline in the ACQ-6 and AQLQ scores at week 40 for tralokinumab compared with placebo (supplementary figures S1 and S2).
In the population of patients with either FeNO high (≥37 ppb) or FeNO low (<37 ppb), no significant differences were observed between the tralokinumab and placebo groups in the LS mean percentage reduction in the final daily average OCS dose at week 40 (FeNO high 37.73% versus 32.27%, 95% CI –31.03–20.12; FeNO low 37.56% versus 28.62%, 95% CI –26.15–8.28; supplementary table S5). Likewise, there were no significant differences between tralokinumab and placebo groups for any of the secondary end-points in the FeNO high or FeNO low populations (supplementary table S6).
Safety
127 (90.7%) patients experienced at least one adverse event during the 40-week treatment period (65 (92.9%) patients in the tralokinumab group and 62 (88.6%) patients in the placebo group; table 3). The most frequently reported adverse events (>5%), which differed between the tralokinumab and placebo treatment groups, were viral upper respiratory tract infection (URTI; 35.7% versus 14.3%), bronchitis (15.7% versus 24.3%), asthma (11.4% versus 22.9%) and back pain (10.0% versus 2.9%). During treatment, 14 (10.0%) patients experienced injection site reactions (11 (15.7%) patients in the tralokinumab group and three (4.3%) patients in the placebo group). The most common injection site reactions were erythema (six (8.6%) patients in the tralokinumab group) and pain (six (8.6%) patients in the tralokinumab group and two (2.9%) patients in the placebo group).
Overall, 25 (17.9%) patients had at least one SAE (nine (12.9%) patients in the tralokinumab group and 16 (22.9%) patients in the placebo group; supplementary table S7). Asthma was the most common SAE, being reported in five (7.1%) patients in the tralokinumab group and eight (11.4%) patients in the placebo group. A total of eight (5.7%) patients discontinued treatment due to adverse events (six (8.6%) patients in the tralokinumab group and two (2.9%) patients in the placebo group). All adverse events were single events. One patient in each treatment group discontinued due to worsening asthma and one female patient in the tralokinumab group discontinued due to breast cancer. Adverse events leading to treatment discontinuation were judged to be related to treatment in three patients receiving tralokinumab (two injection site events and one event of angio-oedema). 11 (7.9%) patients experienced severe infections during the treatment period (six (8.6%) patients in the tralokinumab group and five (7.1%) patients in the placebo group). There were no reported cases of pneumonia during the trial. Information on severe infections is provided in supplementary table S8.
At week 40, the mean eosinophil count increased from 243 cells·µL−1 at baseline to 382 cells·µL−1 in the tralokinumab group; in the placebo group, the mean eosinophil count changed from 210 cells·µL−1 at baseline to 246 cells·µL−1 at week 40. A total of eight patients (seven patients in the tralokinumab group and one in the placebo group) experienced blood eosinophilia (>1500 cells·µL−1). The patient in the placebo group with this increase had an adverse event of chronic eosinophilic pneumonia of moderate intensity 225 days post-first dose of the investigational product. The event was nonserious and considered not related to the investigational product by the investigator. The patient eventually recovered from the event. No major cardiovascular or cerebrovascular events or cases of anaphylaxis were reported in this trial. No patients died during the treatment period. There were no detectable anti-drug antibodies observed in either the tralokinumab 300 mg or placebo group.
Discussion
Results from this phase 3 TROPOS trial did not demonstrate a significant OCS-sparing effect of tralokinumab compared with placebo in patients with severe asthma. Tralokinumab did not result in a statistically significant reduction in the final daily average OCS dose at week 40 compared with placebo (primary end-point). In addition, no significant effects were observed between tralokinumab and placebo in the proportion of patients with a final daily average OCS dose of ≤5 mg, the proportion of patients with a ≥50% reduction in their final daily average OCS dose from baseline and the AAER (secondary end-points). There was a 20% reduction in the AAER with tralokinumab treatment compared with placebo, but this was not statistically significant.
In contrast to TROPOS, several other trials that examined the OCS-sparing effect of monoclonal antibodies (benralizumab, mepolizumab and dupilumab) in patients with severe asthma have reported positive findings [15–17]. The ZONDA trial reported that two benralizumab dosing regimens (30 mg administered s.c. every 4 or 8 weeks) significantly reduced the median final OCS dose from baseline by 75% (p<0.001 for both comparisons) in patients with severe asthma [15]. Likewise, results from the VENTURE trial reported that dupilumab treatment reduced OCS use while decreasing the rate of severe exacerbations and increasing FEV1 (percentage change in OCS dose was −70.1% in the dupilumab group versus −41.9% in the placebo group; p<0.001) [17]. The failure of tralokinumab to provide OCS-sparing benefits or reduce asthma exacerbations in TROPOS is probably explained by a lack of anti-inflammatory effect and the functional redundancy of IL-13 and IL-4 cytokines in asthma [18]. Indeed, results from a recent tralokinumab phase 2 MESOS trial, in which patients with moderate to severe asthma underwent bronchial biopsies before and after 12 weeks of treatment with tralokinumab reported no effect on airway eosinophil inflammation [19]. This lack of anti-inflammatory effect probably explains why previous trials of therapies targeting IL-13, such as lebrikizumab, did not demonstrate consistent reductions in asthma exacerbations [20, 21].
Notably, and in line with results from the phase 2 trials [11, 12], treatment with tralokinumab improved lung function (exploratory end-point), with a significant and clinically meaningful improvement in prebronchodilator FEV1 at week 40. Interestingly, results from the MESOS phase 2 trial [19] and a post hoc analysis of data from a subset of patients in the tralokinumab phase 2b trial suggest that tralokinumab may improve lung function by affecting airway smooth muscle tone [22]. However, tralokinumab treatment did not lead to significant improvements in other exploratory end-points relating to asthma control or quality of life in TROPOS.
FeNO is a well-established biomarker for airway inflammation, which is induced through IL-13 axis-mediated pathways [23, 24]. In the pivotal STRATOS 1 trial a subpopulation of patients with severe asthma with FeNO ≥37 ppb was identified as most likely to have an enhanced response to tralokinumab [13]. For this reason, prior to unblinding of trial results and following consultation with the United States Food and Drug Administration, the statistical analysis plan for TROPOS was revised to perform the primary analysis in a FeNO high subpopulation (FeNO ≥37 ppb) if there were a sufficient proportion of patients. Given the small sample size for this subpopulation, contingency plans were devised to ensure that the primary analysis would only be performed in a FeNO high subpopulation if ≥70 patients were included (thresholds of ≥37 ppb then ≥30 ppb). In TROPOS, more than half of the patients had FeNO levels below the lower 30 ppb threshold, and therefore, the primary and secondary analyses were performed in the all-comers population. However, exploratory analyses of the primary and secondary end-points based on FeNO levels did not reveal that higher FeNO levels (≥37 ppb) were predictive of a greater treatment effect with tralokinumab. The pivotal phase 3 trial STRATOS 2 also failed to show meaningful reductions in the AAER in FeNO high patients [13].
Tralokinumab demonstrated an acceptable safety profile, which was consistent with previous reports [11, 12]. Overall, the incidence of adverse events was comparable between the tralokinumab and placebo groups, although patients treated with tralokinumab reported a higher incidence of URTIs and injection site reactions. The explanation for the disparity in rates of URTIs remains unclear, since there were no reported cases of pneumonia in this trial and the proportion of patients with severe infections were similar across treatment groups. Interestingly, similar observations were observed in the VENTURE trial, where the most frequent adverse event was URTI (9% of patients in the dupilumab group and 18% of patients in the placebo group) [17]. Likewise, results from the LIBERTY ASTHMA trial, in which patients received add-on dupilumab at a dose of 200 or 300 mg or placebo for 52 weeks, also reported that URTI was the most commonly reported adverse event [25].
A small increase in mean blood eosinophil counts was observed for the tralokinumab group versus the placebo group, with more tralokinumab-treated patients having an increase in eosinophil counts above 1500 cells·µL−1 than placebo-treated patients (seven versus one). However, despite this finding, no associated adverse events were reported in those tralokinumab-treated patients with increased blood eosinophil counts. This finding is in line with that reported in the VENTURE trial, where transient blood eosinophilia was observed in more patients in the dupilumab group than the placebo group (14% versus 1%), without any clinical consequences or associated adverse events [17]. In the LIBERTY ASTHMA trial, eosinophilia was reported as an adverse event in 4.1% of patients who received dupilumab and 0.6% of patients who received placebo; however, in this instance, in 0.2% of the overall population, these adverse events were accompanied by clinical symptoms, with two events reported as SAEs (worsening of hypereosinophilia and chronic eosinophilic pneumonia) [25].
A strength of the current trial was the trial design, which to a large extent, replicated methods previously used in other OCS-sparing trials [15, 16, 26, 27], but with two notable exceptions. Firstly, based on findings from the tralokinumab development programme [11, 12], a 12-week induction phase was incorporated into the TROPOS trial design to ensure maximal effect on FEV1. Secondly, the maintenance phase was extended to 8 weeks to assess longer term efficacy compared with previously reported OCS-sparing trials [16, 26–28]. Notably, the reduction in OCS use observed in the placebo group in TROPOS was similar to that reported in the ZONDA trial [15], indicating that the trial design was adequate to demonstrate an OCS-sparing effect. However, this trial had limitations, including the usual restrictions of selectivity and a relatively small sample size, which meant that several analyses, such as the impact of increased eosinophil levels, could not be performed.
In summary, tralokinumab did not demonstrate clinically significant OCS-sparing effects or improvements in asthma control compared with placebo in patients with severe, uncontrolled asthma. Higher levels of FeNO (≥37 ppb) were not predictive of a greater treatment effect with tralokinumab. Hence, the findings in this trial, along with the previously reported results of the STRATOS 1 and 2 trials, suggest that targeting IL-13 alone is not an effective strategy to manage severe asthma.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material ERJ-00948-2018_Supplement
Acknowledgements
We thank the healthcare providers, research staff, patients and caregivers who participated in the trial. Medical writing support was provided by Khund Sayeed, Mohammed Najeeb and Frances Gambling of Cactus Communications (Mumbai, India), funded by AstraZeneca (Cambridge, UK), in accordance with Good Publication Practice (GPP3) guidelines (www.ismpp.org/gpp3). Presented as an abstract at the 28th International Congress of the European Respiratory Society, Paris, France; September 15–19, 2018.
Footnotes
This article has supplementary material available from erj.ersjournals.com
This study is registered at clinicaltrials.gov with identifier NCT02281357. Data underlying the findings described in this manuscript may be obtained in accordance with AstraZeneca's data sharing policy described at https://astrazenecagrouptrials.pharmacm.com/ST/Submission/Disclosure
Conflict of interest: W.W. Busse reports personal fees for consultancy from AstraZeneca, during the conduct of the study; personal fees for consultancy from 3M, Boehringer Ingelheim, GlaxoSmithKline, Novartis, Sanofi-Regeneron and Teva, personal fees for data and safety monitoring board work from Boston Scientific and Genentech, personal fees (royalties) from Elsevier, personal fees for educational videos from Medscape, personal fees for joint oversight committee work from ICON Clinical Research Limited, and personal fees from Pfizer, Roche, AstraZeneca and Circassia, and grants from NIH-NIAID and NIH-NHLBI, outside the submitted work.
Conflict of interest: G.G. Brusselle reports honoraria for advisory boards and/or lectures from AstraZeneca, Boehringer-Ingelheim, Chiesi, GlaxoSmithKline, Novartis, Pfizer, Sanofi/Regeneron, Teva, UCB and Zambon, during the conduct of the study.
Conflict of interest: S. Korn reports consulting and lecture fees from Almirall, AstraZeneca, Boehringer Ingelheim, Chiesi, Teva and Roche, and grants and personal fees from GlaxoSmithKline and Novartis, during the conduct of the study.
Conflict of interest: P. Kuna reports personal fees from Adamed, Allergopharma, ALK, AstraZeneca, Bayer Celon Pharma, Chiesi, FAES, HAL Allergy, Lekam, Polpharma, Pfizer, Sandoz and Teva, during the conduct of the study; and personal fees from Berlin Chemie, Boehringer Ingelheim and Novartis, outside the submitted work.
Conflict of interest: A. Magnan reports personal fees and non-financial support from GlaxoSmithKline, Novartis, Boehringer Ingelheim, AstraZeneca, Stallerg nes, ALK, MundiPharma, Teva, Menarini and Meda Pharma, during the the past 5 years.
Conflict of interest: D. Cohen is an employee of AstraZeneca.
Conflict of interest: K. Bowen is an employee of and own shares in AstraZeneca.
Conflict of interest: T. Piechowiak is an employee of and owns stock options in AstraZeneca.
Conflict of interest: M.M. Wang is an employee of and owns shares in AstraZeneca.
Conflict of interest: G. Colice is an employee of and owns shares and stock options in AstraZeneca.
Support statement: This study was sponsored by AstraZeneca. Funding information for this article has been deposited with the Crossref Funder Registry.
- Received May 21, 2018.
- Accepted October 30, 2018.
- Copyright ©ERS 2019