Abstract
Objective:
Exercise intolerance is a hallmark of chronic obstructive pulmonary disease (COPD).
Methods:
Patients with COPD were randomized in two multicentre, double-blind, incomplete block crossover studies. Patients received two of six treatments in sequence (12 weeks each): placebo, umeclidinium (UMEC)/vilanterol (VI) (125/25 mcg or 62.5/25 mcg), VI (25 mcg) or UMEC (62.5 mcg or 125 mcg). Exercise endurance time (EET) and trough forced expiratory volume in 1 second (FEV1) (Week 12) were co-primary endpoints. Safety was monitored throughout.
Results:
Both studies showed similar 3-hour post-dose EET improvements from baseline for UMEC/VI (Week 12). Significant EET improvements were observed with both UMEC/VI doses versus placebo at Week 12 in Study 418 (UMEC/VI 125/25 mcg: 65.8 s; p = 0.005; UMEC/VI 62.5/25 mcg: 69.4 s; p = 0.003), but not in Study 417, where a placebo effect was evident. Post hoc integrated data analysis showed significant but smaller EET improvements for both UMEC/VI doses versus placebo at Week 12 (UMEC/VI 125/25 mcg: 47.5 s; p = 0.002; UMEC/VI 62.5/25 mcg: 43.7 s; p = 0.001). Both studies showed trough FEV1 improvements at Week 12 for both UMEC/VI doses. The incidence of adverse events was similar between treatment groups within each study.
Conclusions:
UMEC/VI improved lung function and EET.
Background
The pharmacological management of stable chronic obstructive pulmonary disease (COPD) primarily aims to improve symptoms and quality of life, optimize lung function, reduce COPD exacerbations and improve exercise tolerance [Celli et al. 2004; GOLD, 2014]. Inhaled bronchodilators such as muscarinic antagonists and β2-agonists are central to this approach. Muscarinic antagonists block acetylcholine-mediated bronchoconstriction by binding to M3 receptors, while β2-agonists induce smooth muscle relaxation when bound to β2-adrenergic receptors [Brusasco, 2006; GOLD, 2014].
Although forced expiratory volume in 1 second (FEV1) is commonly used for the assessment of responses to pharmacotherapeutic interventions in COPD, changes in FEV1 may not fully predict symptomatic responses. Patients with COPD exhibit expiratory airflow limitation, leading to increased breathing effort, dyspnoea and exercise intolerance [O’Donnell and Webb, 1993; O’Donnell et al. 1997]. Therefore, direct measures of lung hyperinflation, exercise tolerance and exertional dyspnoea may be more sensitive to therapeutic intervention and/or more clinically relevant than FEV1 [Bauerle et al. 1998; O’Donnell et al. 1998; Officer et al. 1998]. The endurance shuttle walking test (ESWT) has been used to evaluate exercise capacity in patients with COPD and studies have validated a minimal clinically important difference (MCID) of 45–85 seconds in response to a bronchodilator [Brouillard et al. 2007; Pepin et al. 2011; Borel et al. 2013].
The long-acting muscarinic antagonist umeclidinium (UMEC, GSK573719) and the combination of UMEC with the long-acting β2-agonist vilanterol (VI) are approved maintenance treatments for COPD in the US and EU; they are not indicated for treatment of asthma [US Food and Drug Administration, 2013]. UMEC/VI is associated with improvements in FEV1 [Donohue et al. 2013, Celli et al. 2014], but effects on lung hyperinflation, exercise tolerance and exertional dyspnoea have not previously been well characterized. Here we present the results of two large, multicentre, randomized, double-blind crossover studies to examine the effect of the inhaled combination of UMEC/VI at two different doses (125/25 mcg [delivering 113 mcg UMEC and 22 mcg VI, respectively] and 62.5/25 mcg [delivering 55 mcg and 22 mcg, respectively]) on exercise performance, expiratory flow and lung volumes.
Methods
Study design and patients
The two studies were multicentre, double-blind, randomized crossover studies involving an incomplete treatment block. Study 418 (GlaxoSmithKline study number: DB2114418) (Clinicaltrials.gov identifier: NCT01323660] was performed at 42 centres in seven countries. Study 417 (GlaxoSmithKline study number: DB2114417) (Clinicaltrials.gov identifier: NCT01328444) was performed at 31 centres in six countries.
Eligible patients were: current or former smokers; ⩾40 years of age; had a smoking history of ⩾10 pack-years; had a clinical diagnosis of moderate-to-severe stable COPD (post-bronchodilator FEV1/forced vital capacity [FVC] <70% and FEV1 ⩾35% and ⩽70% predicted) [Hankinson et al. 1999; Celli et al. 2004]; a score of ⩾2 on the Modified Medical Research Council Dyspnoea Scale at Visit 1; and a resting functional residual capacity (FRC) ⩾120% of predicted (to ensure patients were hyperinflated, as hyperinflation is associated with exercise intolerance) [Stocks and Quanjer, 1995]. Patients were excluded if they had comorbid respiratory conditions or a current diagnosis of asthma. Further information about criteria for inclusion, exclusion and randomization are provided in the online Supplementary Methods. (http://tar.sagepub.com/content/by/supplemental-data)
Both studies were approved by the Medical Ethical Committee of the participating centres and conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. All participants provided signed informed consent before study participation.
Randomization
The randomization code was generated by GlaxoSmithKline using a validated computerized system (RandAll version 2.5). Allocation of treatments was controlled with RAMOS (Randomization and Medication Ordering System), a telephone-based interactive voice response system used by the investigator or designee. The study drugs were administered in a double-blind fashion. The treatment sequences were selected to optimize the power for comparisons between UMEC/VI and placebo (Supplementary Table 1) and therefore the number of subjects in each treatment group was unbalanced.
Procedures
Patients who met eligibility criteria at screening (Visit 1) completed a 12 to 21 day run-in period (including Visits 2 and 3) followed by two 12-week treatment periods separated by a 14-day washout period (Supplementary Figure 1). The incremental shuttle walk test (ISWT) was explained to the patients at Visit 1, and performed at Visit 2 and again at Visit 8, to define an appropriate speed of walking in the ESWT for each treatment period as it produces symptom-limited maximal performance (further details provided in Supplementary Methods) [Singh et al. 1992; Revill et al. 1999; Pepin et al. 2005]. An ESWT was performed on Visit 3 to familiarize patients with the procedure during the run-in period and the subsequent Visit 4 ESWT was used as baseline for Period 1. Visit 8 ESWT was used as baseline for Period 2. All patients were provided with salbutamol for use on an ‘as-needed’ basis throughout the run-in, washout and treatment periods. Stable/regular doses of inhaled corticosteroids (ICS) were permitted.
Patients were randomized in a 7:2:2:3:6:6 ratio to receive 2 of 6 treatments in sequence: placebo, UMEC 62.5 mcg, UMEC 125 mcg, VI 25 mcg (delivering 22 mcg), UMEC/VI 62.5/25 mcg or UMEC/VI 125/25 mcg, all delivered orally by the ELLIPTA™ dry power inhaler (DPI).
Outcomes
The two co-primary endpoints in both studies were exercise endurance time (EET) in an ESWT at 3 hours post-dose and trough FEV1 at Week 12 of each treatment period (obtained 24 hours after dosing on Day 84). The modified Borg scale was measured at 2-minute intervals during the ESWT. Data were reported at isotime, which was the time recorded for the shortest post-dose ESWT during that period.
Spirometry assessments were conducted at screening (including assessments of reversibility to salbutamol and ipratropium) and at each treatment clinic visit (further details provided in Supplementary Methods).
Secondary endpoints included measures of lung volume (inspiratory capacity [IC], FRC and residual volume [RV]) at Week 12 of each treatment period (trough and 3 hours post-dose) and FEV1 3 hours post-dose at Week 12. Lung volumes were obtained using a constant volume body plethysmograph (body box) at each clinic visit [Wanger et al. 2005]. Baseline plethysmography was conducted for each period at Visit 4 prior to randomization and at Visit 9 prior to dosing.
Rescue salbutamol use and ease of DPI use were also assessed. Baseline salbutamol use was determined for 7 day periods during the run-in for Treatment Period 1 and during the washout for Treatment Period 2. Salbutamol use was also recorded by patients during each treatment period using a paper diary card. Inhaler use was observed and captured at each centre twice during the study (Visits 5 and 6). A patient questionnaire was used to determine the ease of inhaler use and how many doses were left in the inhaler (further details provided in Supplementary Methods).
Safety assessments included adverse events (AEs) and the incidence of exacerbations. Vital signs, clinical chemistry parameters, haematology parameters and 12-lead electrocardiograms (ECGs) were also evaluated.
Statistical analysis
The studies were powered assuming an EET residual within-patient standard deviation (SD) of 114 s. The studies had 94% power to detect a 70 s difference between UMEC/VI and placebo, and 92% power to detect a difference of 0.100 L in trough FEV1 between UMEC/VI and placebo, assuming a residual within patient SD of 0.168 L for a crossover study (see Supplementary Methods for further details).
A step-down testing hierarchy was applied to account for multiplicity across comparisons and was dependent upon statistical significance being achieved for previous tests in the hierarchy (i.e. if a comparison failed to meet statistical significance then inference could not be drawn for subsequent statistical analyses in the testing hierarchy). The co-primary comparisons of interest were 3-hour post-dose EET and trough FEV1 for UMEC/VI 125/25 mcg versus placebo, followed by UMEC/VI 62.5/25 mcg versus placebo at Week 12.
Secondary analyses assessed UMEC/VI 125/25 mcg versus placebo for IC, FRC, RV and 3-hour post-dose FEV1, followed by UMEC/VI 62.5/25 mcg versus placebo. Other comparisons of interest were carried out for VI and the two doses of UMEC versus placebo, as well as for UMEC/VI versus the individual components, although these analyses were not powered to detect treatment differences as defined above for UMEC/VI versus placebo.
A post hoc integrated analysis of co-primary endpoints data for both studies was carried out. The two component bronchodilators were also characterized in an exploratory analysis.
EET at Week 12 was analyzed for the intent-to-treat (ITT) population using a mixed model repeated measures analysis, including covariates of mean walking speed (the mean of the walking speeds in each of the two treatment periods), period walking speed (the difference between a patient’s walking speed for each treatment period, and the mean overall walking speed for that patient), period, treatment, visit, smoking status, centre group, visit by period walking speed interaction, visit by mean walking speed interaction, and visit by treatment interaction, where visit is nominal. Subject was included as a random effect. The variance–covariance matrix was assumed to be unstructured. The model used all available 3-hour post-dose EET values recorded on Day 2, Week 6 and Week 12. Missing data were not directly imputed in this analysis; however, all non-missing data for a subject was used within the analysis to estimate the treatment effect for 3-hour post-dose EET at Week 12. The response variable was change from baseline in 3-hour post-dose EET. Inference was restricted by the stepdown closed testing procedure.
All programming for statistical analysis was carried out in a HARP™ environment using SAS Version 9 or a later release and S-Plus Version 8.1 or later release, as detailed in the Supplementary Methods.
Results
Patients
In Study 418, 634 patients were enrolled, 308 were randomized and 307 patients were included in the ITT population across 42 centres in the US (45%), the Czech Republic (13%), South Africa (13%), Denmark (9%), Canada (7%), Ukraine (7%), and the UK (5%). A total of 217 (71%) patients in the ITT population completed the study (Figure 1). Of the 596 patients enrolled in Study 417, 349 patients were randomized and 348 were included in the ITT population across 31 centres in the US (56%), Germany (19%), Russian Federation (12%), Estonia (7%), the UK (4%) and Bulgaria (2%). A total of 258 (74%) patients completed the study. Patient demographics and the majority of baseline characteristics were similar between the two studies; most patients were current smokers, were Global Initiative for Chronic Obstructive Lung Disease (GOLD) Stage II or III, and slightly more were male than female in each study (Table 1). However, the proportion of patients using ICS at screening was greater in Study 418 (39.4% compared with 28.2% in Study 417), as were the proportion of patients reversible to salbutamol plus ipratropium (65.6% in Study 418, 55.0% in Study 417; Table 1).
Summary of patient disposition for (a) Study 418 and (b) Study 417.
Demographics and baseline characteristics of ITT populations in Studies 417 and 418.
n = 348 in Study 417 and n = 303 in Study 418.
n = 340 in Study 417 and n = 302 in Study 418.
EET, exercise endurance time; FEV1, forced expiratory volume in one second; FVC, forced vital capacity; GOLD, Global Initiative for Chronic Obstructive Lung Disease; ICS, inhaled corticosteroid; ISWT, incremental shuttle walk test; ITT, intent-to-treat; SD, standard deviation.
Exercise endurance
In Study 418, both UMEC/VI doses were associated with statistically significant improvements in EET in comparison with placebo (Table 2). Furthermore, these improvements were clinically significant at Week 12, reaching the MCID of 45–85 s [Brouillard et al. 2007; Pepin et al. 2011; Borel et al. 2013]. The improvements were seen at Day 2 and were maintained over the 12-week study period (Figure 2a). In Study 417, there were improvements in EET with UMEC/VI 125/25 mcg at Day 2 which approached the MCID of 45–85 s [Brouillard et al. 2007; Pepin et al. 2011; Borel et al. 2013] (42.0 s, 95% confidence interval (CI) 17.3, 66.7; p < 0.001) compared with placebo. These were sustained at Week 6 (42.0 s, 95% CI 12.2, 71.8; p = 0.006) but not at Week 12. For UMEC/VI 62.5/25 mcg, there was an improvement of 38.2 s (95% CI 14.0, 62.3; p = 0.002) compared with placebo at Day 2, but this was not sustained at Week 6 or 12 (Table 2; Figure 2b). This appeared to be related to an increase in EET over the 12-week treatment period in the placebo group, rather than a decline in efficacy in the active treatment arms: in Study 417 the improvement from baseline with placebo at Week 12 was 36.7 s (standard error [SE] 13.17 s) compared with 0.1 s (SE 16.66 s) in Study 418.
EET (s) and trough FEV1 (L) change from baseline and placebo in ITT populations of Studies 417 and 418 at Week 12.
Analysis performed using a repeated measures model with covariates of period walking speed, mean walking speed, period, treatment, visit, smoking status, centre group, visit by period walking speed, visit by mean walking speed, and visit by treatment interactions. For FEV1, walking speed is replaced by baseline.
p < 0.01; $p < 0.001.
n is the number of patients with analysable data at Week 12.
CI, confidence interval; EET, exercise endurance time; FEV1, forced expiratory volume in 1 second; ITT, intent-to-treat; LS, least squares; SE, standard error; UMEC, umeclidinium; VI, vilanterol.
Change from baseline in 3-hour post-dose EET (s) in ITT population of: (a) Study 418; (b) Study 417; and (c) post hoc integrated analysis of both studies.
The least squares (LS) mean change from baseline in EET for the two UMEC/VI doses was similar across the two studies at Week 12 (Table 2). The post hoc integrated analysis of EET from both studies demonstrated that both doses of UMEC/VI resulted in an increase in the LS mean change from baseline in 3-hour post-dose EET at Week 12 compared with placebo (Figure 2c).
In exploratory analyses, VI also produced improvements in EET from baseline across both studies but these were less than those observed with the UMEC/VI combination. UMEC 125 mcg and UMEC 62.5 mcg also provided improvements in EET from baseline, but these were inconsistent between studies, e.g. UMEC 62.5 mcg provided greater improvement than UMEC 125 mcg in Study 417, but vice versa in Study 418 (Supplementary Table 2).
Trough FEV1
In Study 418, statistically significant improvements in trough FEV1 were demonstrated for UMEC/VI 125/25 mcg and UMEC/VI 62.5/25 mcg compared with placebo (p < 0.001) (Table 2). The improvements in trough FEV1 were observed from Day 2 and were maintained over 12 weeks in both studies (Figure 3(a)). In Study 417, both doses of UMEC/VI were associated with numerical improvements in trough FEV1 compared with placebo (Table 2; Figure 3b), but statistical significance could not be inferred because the prior comparison for EET in the testing hierarchy failed to demonstrate significance. The post hoc integrated analysis including data from both studies demonstrated greater LS mean changes from baseline in trough FEV1 for both doses of UMEC/VI compared with placebo at Week 12 (Figure 3c).
LS mean change from baseline in trough FEV1 (L) in ITT population of: (a) Study 418); (b) Study 417; and (c) post hoc integrated analysis of both studies.
VI produced consistent improvements in trough FEV1 across the two studies, but responses to UMEC 62.5 mcg and UMEC 125 mcg were variable across the studies (Supplementary Table 2). It was noted that both doses of UMEC/VI demonstrated greater LS mean changes from baseline in trough FEV1 (and thus superior bronchodilation) compared with the individual components (data not shown).
Lung volume
Improvements in trough and post-dose IC, FRC, RV and FVC were observed in both studies with both doses of UMEC/VI at Week 12 (Figure 4a, b; Supplementary Table 3). In Study 418, the improvements were statistically significant compared with placebo. However, in Study 417, statistical significance could not be inferred because the prior comparison for EET in the testing hierarchy failed to demonstrate significance.
LS mean change from placebo in: (a) baseline trough and 3-hour post-dose FEV1 (L), FVC and IC for UMEC/VI; and (b) LS mean change from placebo in baseline trough and 3-hour post-dose FRC (L) and RV for UMEC/VI (ITT population; Studies 418 and 417).*
Results on the mean number of puffs of rescue medication per day over Weeks 1 to 12 and exercise dyspnoea scale (Borg) at Week 12 are shown in Supplementary Results and Supplementary Table 4.
Safety
The incidence of AEs was similar between all treatment groups within each study, with an overall higher incidence in Study 418 across all treatments including placebo (Study 418, 30–54%; Study 417, 12–35%; Supplementary Table 5). The most common AEs (⩾3% incidence) in both studies were nasopharyngitis and headache (Table 3). There was one death in Study 418 in the UMEC/VI 62.5/25 mcg group (reported as ‘cancer’) and one death in Study 417 in the UMEC 125 mcg group (reported as ‘death’). Neither death was considered related to study medication by the reporting investigator. An overview of all AEs is provided in Supplementary Table 4.
Summary of on-treatment AEs reported by ⩾3% of patients (ITT populations in Studies 417 and 418).
AE, adverse event; ITT, intent-to-treat; UMEC, umeclidinium; VI, vilanterol.
In both studies, vital sign assessments indicated that none of the treatments had a clinically-meaningful effect on pulse rate or systolic or diastolic blood pressure prior to the walk. The proportion of patients with abnormal, clinically significant ECG assessments at any time post-baseline was similar across all treatments, including placebo, in both studies.
Discussion
The results of these two double-blind, randomized, crossover studies demonstrated similar improvements from baseline in walking exercise tolerance with both doses of UMEC/VI that met the MCID of 45–85 s [Brouillard et al. 2007; Pepin et al. 2011; Borel et al. 2013]. In Study 418, the improvements in EET with both doses of UMEC/VI reached statistical significance compared with placebo. In contrast, the improvements in Study 417 with UMEC/VI were not statistically significant at Week 12. Despite this, post hoc integrated analysis of data from both studies showed improvements in EET compared with placebo at Week 12 for both doses of UMEC/VI, at or near the lower MCID limit.
The differences in the trial results is explained by the unexpected improvements in EET observed in the placebo group in Study 417. The improvements in the placebo group were noted early in the study (at Day 2) and were maintained throughout. Investigations were undertaken to determine if this could be due to a learning effect, differences in the patient populations, differences in exercise limitations or aspects of the study methodology. While the population in Study 418 was more reversible to both salbutamol and ipratropium + salbutamol, and included a higher proportion of patients able to walk at the top three levels of the ISWT (45–58% compared with 24–34% in Study 417), there was no definitive explanation that could account for this large placebo effect. It is possible that there were differences in the execution of the exercise tests across study centres and countries, although this risk was minimized as investigators from participating centres in both studies attended the same training sessions and patients were recruited predominantly (45% of patients) from the US.
In addition to similar improvements from baseline in EET, both studies demonstrated similar improvements from baseline in trough FEV1 with UMEC/VI. The improvements were maintained from Day 2 through to Week 6 and Week 12; at Week 12 the improvements were statistically significant in Study 418, and in Study 417, were numerically superior to placebo. Across the two studies, the improvements in trough FEV1 with both doses of UMEC/VI were clinically meaningful compared with placebo. In contrast to EET, there was no difference between the placebo groups in trough FEV1. Improvements were also observed with UMEC and VI monotherapy, although this did not reach the levels observed with the combination therapy.
Improvements were also observed for both doses of UMEC/VI versus placebo for trough and post-dose FVC, RV, IC, FRC and post-dose FEV1 in each study. Indeed, the lung volume changes obtained with bronchodilation in the present study are similar to those obtained with lung volume reduction procedures [Snell et al. 2012]. The 3-hour post-dose improvements with UMEC/VI were consistently higher than the trough measurements in both studies. These data suggest that bronchodilation was associated with improvements in static hyperinflation, as has been reported previously with use of tiotropium [Celli et al. 2003].
Improving exercise capacity, and walking in particular, is an important objective in the treatment of COPD [Vestbo et al. 2013]. Previous exercise studies using the ESWT have been performed in single centres. A single-dose crossover study in patients with FEV1 <70% predicted (52% mean; n = 20) showed a significant change in lung function and a 117 ± 208 s improvement in walking performance with salmeterol compared with placebo [Brouillard et al. 2008]. We have also reported a 128 ± 141 s improvement with tiotropium alone compared with placebo in COPD [Bedard et al. 2012]. Although UMEC/VI was associated with bronchodilation and lung deflation in the present studies, the smaller and more variable improvements in walking endurance compared with the single-centre studies is likely to be related to the multicentre study design. Nevertheless, improvements in walking endurance in a multicentre study provides support for the rationale that bronchodilation can improve walking endurance in COPD, potentially via improvements in static hyperinflation, although we were unable to establish a relationship between improvements in dynamic hyperinflation and exercise tolerance in the present study. Further studies would be needed to explore any differential effects of treatment on exercise capacity, as the studies were not powered to determine treatment differences between UMEC/VI combination therapy and the monotherapies.
In addition to the improvements in EET lung function and lung volumes, UMEC/VI was generally well tolerated across the two studies, and no significant safety concerns were identified with treatment with UMEC/VI or the individual components compared with placebo.
Conclusions
These results demonstrated that UMEC/VI 125/25 and 62.5/25 produced effective bronchodilation thereby improving lung function and reducing lung volumes, which translated into improvements in exercise capacity. All active treatments were well tolerated at the doses assessed in the two studies.
Footnotes
Acknowledgements
Our thanks to the GlaxoSmithKline study team of Eric Lamb and Susan Davis, to the investigators and to the patients who consented to carry out this research. A full list of the investigators is available in the ONLINE SUPPLEMENTARY MATERIALS. GlaxoSmithKline was involved in the study design, the conduct of the study and the collection, analysis and interpretation of data. Stuart Wakelin, PhD, CMPP, of Fishawack Indicia Ltd, provided editorial and formatting assistance in the preparation of the manuscript, which was also funded by GlaxoSmithKline.
Funding
This study was funded by GlaxoSmithKline.
Conflict of interest statement
F.M. has received fees for speaking at conferences sponsored by Boehringer Ingelheim, Pfizer, GlaxoSmithKline and Grifols, and has served on advisory boards for GlaxoSmithKline and Boehringer Ingelheim. He has also received research grants for participating in multicentre trials sponsored by GlaxoSmithKline, Boehringer Ingelheim, Altana Pharma, Merck, AstraZeneca, Nycomed and Novartis, and has received an unrestricted research grant from Boehringer Ingelheim and GlaxoSmithKline. He holds a CIHR/GSK research chair on COPD. A.C., A.C.D., A.H.G. and J.H.R. are employees of GlaxoSmithKline and have stock options in GlaxoSmithKline. G.C. was an employee of GlaxoSmithKline at the time of the study and manuscript preparation and held stocks/shares in GlaxoSmithKline. At the time of the manuscript submission, he was an employee of Aerocrine, Morrisville, NC, USA, and did not receive any fees relating to the study or manuscript development. S.S. developed the ISWT, but has no financial interest. She has served on advisory boards for GlaxoSmithKline and Pfizer. She has also received fees for speaking from AstraZeneca and GlaxoSmithKline.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
