nature publishing group
Bronchodilatory Effect of the PPAR-γ Agonist
Rosiglitazone in Smokers With AsthmaM Spears1, I Donnelly2, L Jolly2, M Brannigan1, K Ito3, C McSharry2, J Lafferty1, R Chaudhuri1,
G Braganza1, P Bareille4, L Sweeney4, IM Adcock3, PJ Barnes3, S Wood5 and NC Thomson1
Smokers with asthma show a reduced response to inhaled
subtypes in vitro
and reduce inflammation in animal models of
corticosteroids. We hypothesized that a peroxisome
asthma and neutrophilic airways disease.12 On the basis of this
γ) agonist would be
evidence, we hypothesized that the PPAR-γ agonist rosiglitazone
superior for the clinical treatment of these asthma patients.
would have anti-inflammatory activity that would be of benefit
Forty-six smokers with asthma were randomized to inhaled
in smokers with asthma. Therefore, we undertook an exploratory
beclometasone dipropionate (200 µg per day) or rosiglitazone
clinical trial to examine the effect of rosiglitazone on lung
(8 mg per day) for 4 weeks. Rosiglitazone produced
function, Asthma Control Questionnaire (ACQ) score, and
improvements in lung function (forced expiratory volume in
inflammatory end points in a group of smokers with asthma.
1 s (FEV1) = 183 ml, P = 0.051; forced expiratory flow between
25 and 75% of the forced vital capacity (FEF25–75) = 0.24 l/s, Results
P = 0.030) as compared with inhaled beclometasone
A total of 3,895 subjects with asthma were invited to participate
dipropionate. Further trials using PPAR-
γ agonists in steroid-
in the study between August 2005 and May 2007, of whom
resistant airway disease are indicated.
294 gave positive responses. Following screening through
telephone calls, visits were arranged for 187 subjects. After a
Inhaled corticosteroids are recommended as the first line run-in period involving weaning from inhaled corticosteroids
of treatment in patients with chronic persistent asthma.1 and assessment of bronchodilator reversibility, 91 subjects
Smokers with asthma, however, exhibit an impaired response met the criteria for randomization (see Methods for further
to both inhaled and oral corticosteroids,2–5 possibly because of details). The trial contained four treatment arms, and subjects
noneosinophilic airway inflammation, impaired glucocorticoid were randomly allocated to the various treatments. Forty-
receptor function, and/or reduced histone deacetylase activity.6 five subjects were randomized to other treatments, which
Cigarette smoking in asthma patients is also associated with an are not discussed in this article.13 The other 46 subjects were
accelerated decline in lung function,7 increase in the number randomized equal y to rosiglitazone and inhaled beclometasone
of emergency department visits for asthma (with associated dipropionate. The demographic, clinical (including previous
costs)7,8 and increase in severity of symptoms, as compared with inhaled corticosteroid and long-acting β2-agonist use), and
nonsmoking asthmatic patients.9 The prevalence of smoking inflammatory baseline characteristics of the recruited subjects
in subjects with asthma reflects the prevalence in the general in each group were well matched. All the end points
population, and therefore smokers with asthma constitute a presented are the changes relative to the response in the group
large group of patients with poorly control ed disease.10 Smoking assigned to inhaled corticosteroids.
cessation is an effective therapy in this group,11 but because
sustained quitting rates are low, additional or alternative therapies lung function
are needed for individuals with asthma who continue to smoke.
At 2 weeks, rosiglitazone demonstrated a borderline
The glucocorticoid receptor is a member of the nuclear improvement in prebronchodilator forced expiratory volume
hormone receptor family, which includes the peroxisome in 1 s (FEV1) (164 ml, 95% confidence interval (CI), −1 to 329,
proliferator–activated receptor-γ (PPAR-γ). PPAR-γ agonists P
= 0.051) nda significant improvement
exert anti-inflammatory effects on multiple inflammatory cell in prebronchodilator peak expiratory flow (32.7 l/min, 95% CI
1Department of Respiratory Medicine, Faculty of Medicine, University of Glasgow, Glasgow, UK; 2Department of Immunology, Faculty of Medicine, University of Glasgow, Glasgow, UK; 3Airway Disease Section, National Heart and Lung Institute, Imperial College, London, UK; 4Discovery Medicine, GlaxoSmithKline, London, UK;
5Department of General Practice, Faculty of Medicine, University of Glasgow, Glasgow, UK. Correspondence: NC T
Received 24 January 2009; accepted 2 March 2009; advance online publication 8 April 2009.
CliniCal pharmaCology & TherapeuTiCs
table 1 Baseline demographics and induced sputum results
Comparison of treatment responses at 14 and 28 days. (a
in forced expiratory volume in 1 s (FEV1) (ml) at 14 and 28 days. (b
in forced expiratory volume in 1 s (FEF25–75) (l/s) at 14 and 28 days. The
changes presented are mean group changes from randomization to 14 and 28 days of treatment (paired t
-test; error bars represent 95% confidence
values were derived by comparing the relative changes in the
two treatment arms, using analysis of covariance. ICS, inhaled corticosteroid
(beclometasone dipropionate); Rosi, rosiglitazone.
When changes in ACQ scores were compared , there
was no difference between the rosiglitazone group and the one
on inhaled beclometasone dipropionate.
(absolute count (104 cells))Bronchial epithelial cells (%)
Induced sputum cytology.
No relative differences in sputum cell
proportions were observed between the two treatment arms
Data presented as median (interquartile range) unless stated otherwise.
95% CI; 95% confidence interval; ACQ, Asthma Control Questionnaire score (range, 0–6, with higher scores indicating worse asthma control); BMI, body mass index; FEV1, forced expiratory volume in 1 s; FEF25–75, forced expiratory flow between 25 and
A borderline reduction in sputum inter-
75% of the forced vital capacity; IgE, immunoglobulin E; IU, international units; LABA, long-acting β
leukin-8 was observed in the group treated with rosiglitazone
2-agonist; PEF, peak expiratory flow; pre-BD, prebronchodilator.
(−534.1 pg/ml, 95% CI, −1,844.4 to 36.5, P
= 0.068) relative to
the group on inhaled beclometasone dipropionat
= 0.018), and significant improvement in both forced
expiratory flow between 25 and 75% of the forced vital capacity Compliance
(FEF25–75) (0.36 l/s, 95% CI, 0.09–0.63, P
= Eighty-five percent of the subjects who completed the study
and FEF75 (0.24 l/s, 95% CI, 0.09–0.39, P
= 0.002). After 4 weeks, achieved >80% compliance with therapy.
the group treated with rosiglitazone again demonstrated a bor-
derline improvement in prebronchodilator FEV1 (183 ml 95% Adverse events
CI −1.0 to 367.0, P
= andand a sig-
No serious adverse events occurred in subjects in either of the
nificant improvement in FEF25–75 (0.24 l/s, 95% CI, 0.03–0.46, treatment arms during the trial. There were two withdrawals
= andWith respect to other meas-
due to adverse events. One occurred in the rosiglitazone group
urements of lung function, there was no difference between the (allergic reaction, periorbital edema) and one in the inhaled
beclometasone dipropionate group (diarrhea and vomiting).
table 2 Predefined and exploratory end point changes
This improvement is much larger than the effect seen
Change by 14 days
Change by 28 days
in previous trials that examined the effect of inhaled
corticosteroids in smokers with asthma2,3 and was associated
with improvements in our predefined secondary lung function
end points, suggesting that the FEV1 improvement is real. The
failure to produce a conclusive improvement may be due to an
Δ Pre-BD FEF25–75 (l/s (95% CI)) 0.36* (0.09, 0.63)
The improvement in lung function produced by rosiglitazone
was not associated with a reduction in asthma symptoms (as
detected by the ACQ score), sputum profile, or supernatant at
28 days. What can explain this? The subjects had relatively mild
asthma and did not display sputum eosinophilia or neutrophilia
at baseline, and therefore we would not have been able to detect
substantial changes in these cells or sputum cytokines. The lack
of change in ACQ score is possibly an example of dissociation
between change in lung function and ACQ score, as has
What is the mechanism by which rosiglitazone produces
the improvements that we have detected in lung function? The
reason for the poor response to corticosteroids in smokers with
asthma is currently unknown. However, one possible reason is
that cigarette smoking may induce an oxidative stress–mediated
change in the glucocorticoid receptor, resulting in a change in
its behavior and efficacy.18 Recent research has demonstrated
that rosiglitazone is able to bind to the glucocorticoid receptor
ligand-binding domain and thereby alter gene transcription.19
Therefore, our results may be a demonstration of an alternative
mode of glucocorticoid-receptor activation leading to the
detected improvements in lung function. An alternative
explanation emerges from the fact that PPAR-γ has been
Changes in lung function, Asthma Control Questionnaire (ACQ) score, and induced
shown to modulate a distinct but partially overlapping set of
sputum results following treatment with rosiglitazone (compared to response
inflammatory genes relative to corticosteroids.20 Further studies
to treatment with inhaled beclometasone alone). Lung function data and ACQ score differences are difference of adjusted means with adjustment for baseline
examining the relative effects of rosiglitazone on corticosteroid
measurement (analysis of covariance).
and PPAR-γ-specific functional outputs are indicated in smokers
Δ, change; 95% CI, 95% confidence interval; FEV1, forced expiratory volume in 1 s;
with asthma and other conditions that are associated with
FEF25–75, forced expiratory flow between 25 and 75% of the forced vital capacity;
FVC, forced vital capacity; IL-8, interleukin-8; MPO, myeloperoxidase; PEF, peak
relative corticosteroid insensitivity.
expiratory flow; pre-BD, prebronchodilator.
The improvement seen in FEF75 at 14 days and FEF25–75 at
< 0.05, †P
both 14 and 28 days is of interest, given that there are currently
few therapies available for the treatment of small-airway
The frequency of occurrence of headaches was similar in the obstruction. Smal -airway obstruction, seen in many pulmonary
two the groups (five in the beclometasone dipropionate group conditions,21–23 is associated with dynamic hyperinflation,
and four in the rosiglitazone group). Three subjects in the reduced tolerance to exercise, and increased dyspnea. Given
beclometasone dipropionate group reported pharyngitis.
the improvement seen in our patients, consideration should
be given to studying PPAR-γ agonists in subjects with small-
There exists a subpopulation of asthma patients who fail to
In conclusion, this trial—to our knowledge, the first to
respond adequately to current therapies.14 As a result, this group examine the efficacy of a PPAR-γ agonist in subjects with
has poorer asthma control and consumes a disproportionate asthma—has demonstrated modest improvements in lung
share of health-care budgets. Smokers with asthma are part of function measurements in a group of smokers with mild
this large, difficult-to-treat group. This randomized, controlled, to moderate asthma. Our conclusions are tempered by the
exploratory clinical trial examined the impact of a novel alterna-
exploratory nature of this work, reflected in the short duration
tive approach using the PPAR-γ agonist rosiglitazone in a group of treatment and the small number of subjects involved. Given
of smokers with mild to moderate asthma.
that we have detected an effect in a treatment-resistant group,
Treatment with rosiglitazone produced a trend toward further trials should be undertaken to examine PPAR-γ agonists
improvement in prebronchodilator FEV1 relative to low- in asthma and other obstructive airway conditions. Issues that
dose inhaled corticosteroids, at both 14 days and 28 days. should be addressed in future trials include dose response,
CliniCal pharmaCology & TherapeuTiCs
interaction with corticosteroids,19,24 PPAR-γ polymorphisms,25 from a previous clinical trial with oral corticosteroids in smokers with
and PPAR-γ endobronchial expression response to treatment.26 asthma.5 This led us to estimate that we needed to recruit 22 subjects per
PPAR-γ agonists may represent a new therapeutic class for group to detect a 230 ml difference in FEV1 between the treatment arms,
al owing for a 10% dropout rate. There was a slightly higher dropout rate
(13%) in the trial, resulting in a short time extension to allow a larger
number of subjects to be randomized to treatment.
The primary end point was the difference in prebronchodilator FEV1
Patients with mild to moderate1 stable asthma, aged 18–60
between the group on rosiglitazone and the one on beclometasone
years, and on ≤1,000 µg of beclometasone dipropionate (or equiva-
dipropionate at 28 days. The secondary end points were changes in
lent) per day and smokers of 5 or more cigarettes per day with at least 5
pre- and postbronchodilator peak expiratory flow, forced vital capacity,
pack-years of smoking history were eligible for enrollment. All subjects
FEF25–75, FEF75, and ACQ. Exploratory end points were changes in
demonstrated reversible airflow obstruction.27 Exclusion criteria
sputum differential, sputum supernatant, and serum cytokines. The
included diabetes, recent myocardial infarction, and other active
randomization schedule was generated in blocks using a validated
pulmonary diseases (full criteria available at http://www. clinicaltrials.
system (RandAll). Lung function changes were examined with analysis
gov; NCT00119496). Patients were recruited from general practice,
of covariance (incorporating the Kenward–Rogers method) using SAS
hospital clinics, and research databases. The West Glasgow Research
v8.2 (SAS Institute, Cary, NC). All data obtained after day 1 of treatment
Ethics Committee approved the study, and all patients gave written
were used for analysis. The remaining statistical analysis was performed
using Minitab 15 (Minitab, State College, PA). The level of statistical
significance was set at <0.05. Parametric data were examined using paired
The study was a randomized, prospective, double-blind,
-testing, two-sided t
-testing, or analysis of variance, and nonparametric
double-dummy, active comparator, parallel-group design. Subjects
data were analyzed with Mann–Whitney or Kruskal–Wallis testing, as
were monitored for asthma stability for up to 6 weeks and underwent a
appropriate. Given the exploratory nature of the trial, the secondary and
corticosteroid weaning and monitoring phase that lasted 1 month within
exploratory analyses were not corrected for type 1 errors due to multiple
this period. All the subjects were treated with inhaled β
2 weeks and were excluded from randomization if they experienced an
exacerbation of asthma at any point during this run-in phase.
If subjects were stable and met entry criteria at the end of the 2-week
This study is dedicated to the memory of S.W., who died shortly after the
corticosteroid-free period (including bronchodilator reversibility),
commencement of recruitment. The study was a collaborative project
they attended a randomization visit that entailed spirometry and peak
between the University of Glasgow, Imperial College, and GlaxoSmithKline’s
expiratory flow recordings, completion of an ACQ,28 induced sputum
(GSK) discovery arm. Sponsorship was provided by GSK. M.S. was a
expectoration for differential cell count and supernatant mediators,
beneficiary of research training fellowships from Chest, Heart & Stroke
and routine blood tests for safety (full blood count and renal and
Scotland and the Chief Scientist’s Office (Scotland) during the course of
liver function tests) and characterization (levels of total and specific
the study. An employee of GSK, blinded to the group allocations of the
immunoglobulin E; total, low-density lipoprotein, and high-density
individual patients, performed the lung function statistical analysis. We
lipoprotein cholesterol; and triglycerides).
specifically express our gratitude to Brian Rae, Greater Glasgow and Clyde
The subjects were then randomized with equal bias to one of four
Primary Care R&D, for his help and support with this project. We also thank
groups. Two of the treatment arms and their corresponding results
the participants and the participating general practices from the Greater
are not discussed in this article.13 The subjects discussed in this article
Glasgow and Lanarkshire region; without their help this study would not
received either 4 mg twice a day of oral rosiglitazone maleate ( Avandia;
GlaxoSmithKline, Greenford, UK) or 100 µg twice a day of inhaled
ConFliCt oF inteRest
hydrofluoroalkane beclometasone dipropionate (Qvar; IVAX, Runcorn,
P.B. and L.S. are employees of and own shares in GlaxoSmithKline. The other
UK) (equivalent to ~400 µg per day chlorofluorocarbon beclometasone
authors declared no conflict of interest.
dipropionate).29 The subjects returned for prebronchodilator lung
function tests at 2 weeks and repeated the assessments carried out at
2009 American Society for Clinical Pharmacology and Therapeutics
1. Global Initiative for Asthma. Global Strategy for Asthma Management and
Lung function assessments conformed to consensus
Prevention (2008). <http:/www.ginasthma.com/>.
guidelines.27 Sputum induction, differential count, and supernatant
2. Chalmers, G.W., Macleod, K.J., Little, S.A., Thomson, L.J., McSharry, C.P. &
analysis were performed as previously described.11,30 Sputum
Thomson, N.C. Influence of cigarette smoking on inhaled corticosteroid
supernatants were collected for determination of interleukin-8 and
treatment in mild asthma. Thorax 57
, 226–230 (2002).
myeloperoxidase (interleukin-8; R&D Systems, Abingdon, UK;
3. Tomlinson, J.E., McMahon, A.D., Chaudhuri, R., Thompson, J.M., Wood, S.F.
myeloperoxidase; Immundiagnostik, Oxford Biosystems, Oxford, UK).
& Thomson, N.C. Efficacy of low and high dose inhaled corticosteroid in
Local hospital laboratories performed the full blood counts, urea and
smokers versus non-smokers with mild asthma. Thorax 60
, 282–287 (2005).
electrolyte determinations, liver function tests, total and specific serum
4. Lazarus, S.C. et al
. Smoking affects response to inhaled corticosteroids or
immunoglobulin E tests, and total and differential cholesterol tests.
leukotriene receptor antagonists in asthma. Am. J. Respir. Crit. Care Med
Continuation of smoking during the study was confirmed by history
5. Chaudhuri, R., Livingston, E., McMahon, A.D., Thomson, L., Borland, W. &
and the detection of urinary nicotine metabolites, using the SmokeScreen
Thomson, N.C. Cigarette smoking impairs the therapeutic response to
system (GFC Diagnostics, Bicester, UK). Subjects were regarded as
oral corticosteroids in chronic asthma. Am. J. Respir. Crit. Care Med
current smokers if their category was “mild smoker” or greater and their
urine cotinine level was >1.1 mg/ml. Treatment compliance was assessed
6. Thomson, N.C. & Spears, M. The influence of smoking on the treatment
response in patients with asthma. Curr. Opin. Allergy Clin. Immunol
7. Lange, P., Parner, J., Vestbo, J., Schnohr, P. & Jensen, G. A 15-year follow-up
Because of the reduced response to inhaled cortico-
study of ventilatory function in adults with asthma. N. Engl. J. Med
steroids in smokers with asthma and the lack of published information
on the effect of thiazolidinediones in asthma, we were unable to perform
8. Eisner, M.D. & Iribarren, C. The influence of cigarette smoking on adult asthma
standard power calculations. The study was informed using FEV1 changes
outcomes. Nicotine Tob. Res
, 53–56 (2007).
9. Chaudhuri, R. et al
. Role of symptoms and lung function in determining
19. Matthews, L., Berry, A., Tersigni, M., D’Acquisto, F., Ianaro, A. & Ray, D.
asthma control in smokers with asthma. Allergy 63
, 132–135 (2008).
Thiazolidinediones are partial agonists for the glucocorticoid receptor.
10. Thomson, N.C. Smokers with asthma: what are the management options?
, 75–86 (2009).
Am. J. Respir. Crit. Care Med
, 749–750 (2007).
20. Ogawa, S. et al
. Molecular determinants of crosstalk between nuclear
11. Chaudhuri, R. et al
. Effects of smoking cessation on lung function and airway
receptors and toll-like receptors. Cell 122
, 707–721 (2005).
inflammation in smokers with asthma. Am. J. Respir. Crit. Care Med
21. Hamid, Q. et al
. Inflammation of small airways in asthma. J. Allergy Clin.
, 44–51 (1997).
12. Spears, M., McSharry, C. & Thomson, N.C. Peroxisome proliferator-
22. Hogg, J.C. et al
. The nature of small-airway obstruction in chronic obstructive
activated receptor-γ agonists as potential anti-inflammatory agents in
pulmonary disease. N. Engl. J. Med
, 2645–2653 (2004).
asthma and chronic obstructive pulmonary disease. Clin. Exp. Allergy 36
23. Shaw, R.J., Djukanovic, R., Tashkin, D.P., Millar, A.B., du Bois, R.M. & Orr, P.A. The
role of small airways in lung disease. Respir. Med
, 67–80 (2002).
13. Spears, M. et al
. Effect of theophylline plus beclometasone on lung function in
24. Nie, M., Corbett, L., Knox, A.J. & Pang, L. Differential regulation of chemokine
smokers with asthma-a pilot study. Eur. Respir. J
. (2009); e-pub ahead of print 5
expression by peroxisome proliferator-activated receptor-γ agonists:
interactions with glucocorticoids and β2-agonists. J. Biol. Chem
14. Chung, K.F. et al
. Difficult/therapy-resistant asthma: the need for an integrated
approach to define clinical phenotypes, evaluate risk factors, understand
25. Palmer, C.N. et al
. PPARG locus haplotype variation and exacerbations in
pathophysiology and find novel therapies. ERS Task Force on Difficult/
asthma. Clin. Pharmacol. Ther
, 713–718 (2007).
Therapy-Resistant Asthma. European Respiratory Society. Eur. Respir. J
26. Benayoun, L. et al
. Regulation of peroxisome proliferator-activated receptor-γ
expression in human asthmatic airways. Relationship with proliferation,
15. Juniper, E.F., Price, D.B., Stampone, P.A., Creemers, J.P., Mol, S.J. & Fireman, P.
apoptosis, and airway remodeling. Am. J. Respir. Crit. Care Med
Clinically important improvements in asthma-specific quality of life, but
no difference in conventional clinical indexes in patients changed from
27. Standardization of spirometry—1987 update. Statement of the American
conventional beclomethasone dipropionate to approximately half the
Thoracic Society. Am. Rev. Respir. Dis
, 1285–1298 (1987).
dose of extrafine beclomethasone dipropionate. Chest 121
28. Juniper, E.F., O’Byrne, P.M., Guyatt, G.H., Ferrie, P.J. & King, D.R. Development
and validation of a questionnaire to measure asthma control. Eur. Respir. J
16. Cox, G. et al
. Asthma control during the year after bronchial thermoplasty.
, 902–907 (1999).
N. Engl. J. Med
, 1327–1337 (2007).
29. Vanden Burgt, J., Busse, W.W., Martin, R.J., Szefler, S.J. & Donnell, D.
17. Morjaria, J.B., Chauhan, A.J., Babu, K.S., Polosa, R., Davies, D.E. & Holgate, S.T.
Efficacy and safety overview of a new inhaled corticosteroid, QVAR
The role of a soluble TNF-α receptor fusion protein (etanercept) in
(hydrofluoroalkane-beclomethasone extrafine inhalation aerosol), in
corticosteroid refractory asthma: a double blind, randomised, placebo
asthma. J. Allergy Clin. Immunol
, 1209–1226 (2000).
controlled trial. Thorax 63
, 584–591 (2008).
30. Pavord, I.D., Pizzichini, M.M., Pizzichini, E. & Hargreave, F.E. The use of
18. Ito, K., Chung, K.F. & Adcock, I.M. Update on glucocorticoid action and
induced sputum to investigate airway inflammation. Thorax 52
resistance. J. Allergy Clin. Immunol
, 522–543 (2006).
CliniCal pharmaCology & TherapeuTiCs
In September last year Merck & Co. pulled its popular and effective arthritis painmedication, Vioxx, from the market because researchers investigating the possibleefficacy of heightened doses of Vioxx in the treatment of problems other than arthritisdiscovered that, at those heightened doses, the drug increased the probability ofcardiovascular harms (heart attacks and strokes). In December
• Neue Juristische Wochenschrift • 2002 • Heft 3 • Rechtsprechung • Zivilgerichte • BGH • BGH: Umverpacken von Arzneimitteln - Adalat Fundstelle BGH: Umverpacken von Arzneimitteln - Adalat NJW 2002, 221 Umverpacken von Arzneimitteln - Adalat Wird im Zusammenhang mit dem Parallelimport von Arzneimitteln einer Originalpackung ein Beipackzettel in deutscher Sprache eingefügt,