reviews

The Role of Long-Acting

Bronchodilators in the Management of
Stable COPD*
Donald P. Tashkin, MD, FCCP; and Christopher B. Cooper, MD, FCCP

Bronchodilators form the foundation of symptomatic treatment of COPD. Several long-acting

bronchodilators are now available for use in COPD, but publications of large-scale studies of their
efficacy have, for the most part, postdated the publication of major clinical guidelines. This article
provides a critical review of large (> 50 patients), double-blind, clinical trials of three long-acting
bronchodilators in COPD (the once-daily anticholinergic tiotropium, and the twice-daily ␤2-
agonists formoterol and salmeterol) within the context of the objectives of treatment defined by
the Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines. Fourteen
published studies were identified, of which 12 studies were published since the release of the
GOLD guidelines. All three long-acting bronchodilators were found to effectively improve lung
function; however, they differed in their effects on outcomes other than bronchodilation, with
salmeterol demonstrating inconsistent efficacy compared with placebo in preventing exacerba-
tions and improving health status, and only tiotropium demonstrating consistent superiority to
the short-acting bronchodilator ipratropium. Based on this review, a treatment algorithm for the
introduction of long-acting bronchodilators to patients with COPD is proposed, which includes
the use of long-acting bronchodilators early in the treatment algorithm.
(CHEST 2004; 125:249 –259)

Key words: bronchodilators; COPD; exacerbations; formoterol; Global Initiative for Chronic Obstructive Lung Disease;
health status; lung function; salmeterol; tiotropium

Abbreviations: GOLD ⫽ Global Initiative for Chronic Obstructive Lung Disease; ICS ⫽ inhaled corticosteroid;
SGRQ ⫽ St. George Respiratory Questionnaire; TDI ⫽ transition dyspnea index

C expiration
OPD is characterized by reduced airflow on
due to airway obstruction that is
partially reversible and usually worsens over time.
Patients with COPD have reduced lung function and
*From the Division of Pulmonary and Critical Care Medicine
and the UCLA Pulmonary Function and Exercise Physiology
Laboratory, David Geffen School of Medicine at UCLA,
Los Angeles, CA.
Dr. Tashkin has received consulting fees from Boehringer In-
gelheim, and Dr. Cooper has received consulting fees from
Boehringer Ingelheim and is affiliated with the Boehringer
Ingelheim Speakers’ Bureau.
Manuscript received October 28, 2002; revision accepted April
14, 2003.
Reproduction of this article is prohibited without written permis-
sion from the American College of Chest Physicians (e-mail:
permissions@chestnet.org).
Correspondence to: Donald P. Tashkin, MD, FCCP, Division of
Pulmonary and Critical Care Medicine, UCLA School of Medi-
cine, PO Box 951690, 10833 Le Conte Ave, Los Angeles, CA
90095-1690; e-mail: DTashkin@mednet.ucla.edu
worsening of symptoms (exacerbations). The sepa-
rate or combined effects of dyspnea, reduced exer-
cise capacity, and repeated exacerbations cause im-
pairment of the health-related quality of life of many
patients with COPD,1 as well as causing a greater
burden on health-care resources.2
Currently, COPD is the second most common
noninfectious disease in the world, causing some
2.75 million deaths annually,3 and global mortality is
predicted to more than double by 2030. The increas-
ing burden of COPD has stimulated research into
For editorial comment see page 9
novel treatment approaches (including new pharma-
cotherapies) and optimized management strategies.
An international collaboration has been convened,
the Global Initiative for Chronic Obstructive Lung

www.chestjournal.org CHEST / 125 / 1 / JANUARY, 2004 249

Downloaded From: http://journal.publications.chestnet.org/ on 02/16/2014

 Disease (GOLD), which seeks, among other objec-
tives, to provide evidence-based recommendations
for an integrated COPD management strategy. In
2001, the GOLD committee published guidelines
for the diagnosis, management, and prevention of
COPD, including recommendations for pharmaco-
logic management.4
The GOLD guidelines4 provide a staging system
for the classification of disease severity based on a
combination of spirometric lung function and symp-
toms. In a departure from earlier guidelines, the
GOLD guidelines identify a stage 0, which includes
patients who may not have symptoms, and still have
normal lung function, but are at risk of acquiring
COPD by virtue of tobacco smoking or certain
environmental exposures. Beyond stage 0, COPD is
identified by the presence of airflow obstruction as
primarily defined by a reduction in the forced expi-
ratory ratio (FEV1/FVC) below 70%. The severity of
COPD is defined in four stages—mild, moderate,
severe, and very severe—according to the severity of
impairment in the FEV1.
The GOLD guidelines recommend a stepwise
approach to disease management, with broncho-
dilators being the mainstay of treatment. Short-
acting bronchodilators are recommended initially, as
needed, in mild disease. As the disease progresses,
regular maintenance treatment is indicated; for this,
the guidelines note that long-acting bronchodilators
are more convenient. In more severe disease, poly-
pharmacy becomes common and a trial of inhaled
corticosteroids is recommended.
The bronchodilators currently available for use in
COPD can be broadly categorized into three classes:
anticholinergics, ␤2-sympathomimetic agonists, and
methylxanthines. All three have proven effective in
improving lung function in patients with COPD,
although the narrow therapeutic range of methyl-
xanthines and their relatively weak bronchodilator
effect make them a less attractive initial therapeutic
option.
The GOLD guidelines do not, however, distin-
guish between anticholinergic and ␤2-agonist bron-
chodilators, regarding both as equally effective based
on evidence then available. Nor do they provide
specific guidance on the choice between inhaled
long-acting and short-acting bronchodilators (simply
noting that long-acting bronchodilators are more
convenient). This position was perhaps inevitable
given that, at the time the guidelines were being
developed, only two studies5–7 of long-acting bron-
chodilators in COPD were available, and only one of
these5 included an active comparator.
However, since initial publication of the GOLD
guidelines in 2001, there have been 17 publications
featuring data from large, comparative, double-blind
250
Downloaded From: http://journal.publications.chestnet.org/ on 02/16/2014
studies of long-acting bronchodilators. These new
studies provide a valuable body of evidence on the
relative efficacy of short- and long-acting broncho-
dilators, as well as the relative efficacy of long-acting
␤2-agonists and anticholinergics. This review seeks to
examine these new data within the context provided
by the GOLD guidelines.
Rationale for the Use of Bronchodilators
in COPD
Impaired lung function in COPD is caused by
structural narrowing of the airways, combined with the
effects of cholinergic vagal bronchoconstrictive tone
and decreased lung elastic recoil.8,9 Bronchodilators
improve the airflow limitation observed in patients with
COPD by producing airway smooth-muscle relaxation,
although ␤2-agonists and anticholinergics achieve this
effect through different mechanisms. Anticholinergic
bronchodilators (in particular, tiotropium) produce re-
laxation of airway smooth muscle through antagonism
of acetylcholine at M3-muscarinic receptors on airway
smooth muscle,10 whereas ␤2-agonists induce bron-
chodilation through stimulation of ␤2-receptors, lead-
ing to an increase in cyclic adenosine monophosphate
(as also occurs with phosphodiesterase inhibitors, such
as oral methylxanthines).11
The mechanistic differences between these two
classes of inhaled bronchodilator are reflected in the
relative utility of each for the management of
COPD. Short-acting ␤2-agonists, such as albuterol,
have a more rapid onset but shorter duration of
action than anticholinergics, and thus are commonly
prescribed as a “rescue” medication to help relieve
acute bronchospasm. Inhaled anticholinergic medi-
cations, such as ipratropium, have a slower onset and
slightly longer duration of action. One consequence
of the differences in their modes of action is that the
effects of combining anticholinergic and ␤2-agonist
bronchodilators are additive, providing greater effi-
cacy than either agent alone.12
The two classes of agent also differ in their
nonbronchodilator effects. For example, muscarinic
receptors play a role in regulating mucus secretion,
although the effect of antimuscarinics on sputum
volume is variable.13,14 Viral infections increase cho-
linergic tone, an effect that may be directly counter-
acted by anticholinergics.15 Ex vivo, long-acting ␤2-
agonists have shown cellular effects beyond those
exerted on bronchial smooth muscle, including ef-
fects on mucociliary transport and neutrophils.16 –18
These nonbronchodilator effects may provide addi-
tional benefits, although the relevance (if any) of
these effects to their clinical efficacy in COPD is
currently unknown.
Reviews
 Long-Acting Bronchodilators
Three long-acting, inhaled bronchodilators have
been approved in the United States or Europe for
use in COPD: tiotropium (the most recently li-
censed), formoterol, and salmeterol. An overview of
published, double-blind studies of long-acting bron-
chodilators in COPD involving at least 50 patients is
given in Table 1.
Tiotropium, which became available in many parts of
Europe in 2002, is a once-daily, long-acting, inhaled
anticholinergic that, unlike the short-acting anticholin-
ergic ipratropium, acts through prolonged M3-receptor
blockade. Salmeterol, the first long-acting ␤2-agonist to
be licensed for COPD, has been available for this
indication since 1997 in Europe and since 1998 in the
United States. Formoterol, approved for use in COPD
in the United States and some European countries in
2001, is a long-acting ␤2-agonist that is administered,
like salmeterol, twice daily.
Only one published trial has directly compared
two long-acting bronchodilators; this study com-
prised two 6-month trials, both placebo controlled,
comparing tiotropium with salmeterol.21,22 Although
no large-scale, direct comparisons of the clinical
efficacy of salmeterol and formoterol have been
made, smaller studies36,37 have found that formoterol
has an onset of action that is faster than salmeterol
and similar to salbutamol.

Table 1—Overview of Published Double-Blind Studies of Long-Acting Bronchodilators in COPD Conducted in

> 50 Patients

Source Long-Acting Bronchodilator Control Trial Duration

Casaburi et al (2002)19 Tiotropium 18 ␮g/d Placebo 1 yr

Vincken et al (2002)20 Tiotropium 18 ␮g/d Ipratropium 40 ␮g qid 1 yr
Donohue et al (2002)21 Tiotropium 18 ␮g/d Placebo 6 mo
Salmeterol 50 ␮g bid 24 wk
Brusasco et al (2003)22 Tiotropium 18␮g/d Placebo 6 mo
Salmeterol 50 ␮g/d 24 wk
Aalbers et al (2002)23 Formoterol 6 ␮g bid Placebo 12 wk
Formoterol 12 ␮g bid
Formoterol 24 ␮g bid
Dahl et al (2001)24 Formoterol 12 ␮g bid Ipratropium 40 ␮g qid 12 wk
Formoterol 24 ␮g bid Placebo
Rossi et al (2002)25 Formoterol 12 ␮g bid Theophylline (individualized 52 wk
dosage)
Formoterol 24 ␮g bid
D’Urzo et al (2001)26 Formoterol 12 ␮g bid/ipratropium 40 ␮g qid combination Salbutamol 200 ␮g 3 wk
qid/ipratropium 40 ␮g
qid combination
Szafranski et al (2003)27 Formoterol 4.5 ␮g bid Placebo 12 mo
Formoterol 4.5 ␮g/budesonide 160 ␮g bid combination Budesonide 200 ␮g bid
Boyd et al (1997)6 Salmeterol 50 ␮g bid Placebo 16 wk
Salmeterol 100 ␮g bid
Jones and Bosh (1997)7 Salmeterol 50 ␮g bid Palcebo 16 wk
Salmeterol 100 ␮g bid
Mahler et al (1999)5 Salmeterol 50 ␮g bid Ipratropium 40 ␮g qid 12 wk
Placebo
Rennard et al (2001)28 Salmeterol 50 ␮g bid Ipratropium 40 ␮g qid 12 wk
Placebo
van Noord et al (2000)29 Salmeterol 50 ␮g bid Placebo 12 wk
Salmeterol 50 ␮g bid/ipratropium 40 ␮g qid combination
Rutten-van Mölken et al (1999)30 Salmeterol 50 ␮g bid Placebo 12 wk
Salmeterol 50 ␮g bid/ipratrotropin qid combination
Chapman et al (2002)31 Salmeterol 50 ␮g bid/ipratropium 40 ␮g qid combination Ipratropium 40 ␮g qid 24 wk
ZuWallack et al (2001)32 Salmeterol 50 ␮g bid Theophylline 12 wk
Salmeterol 50 ␮g bid/theophylline combination
Cazzola et al (2000)33 Salmeterol 50 ␮g bid 3 mo
Salmeterol 50 ␮g/fluticasone 250 ␮g bid combination
Salmeterol 50 ␮g/fluticasone 500 ␮g bid combination
Salmeterol 50 ␮g/theophylline bid combination
Mahler et al (2002)34 Salmeterol 50 ␮g bid Placebo 24 wk
Salmeterol 50 ␮g/fluticasone 500 ␮g bid combination Fluticasone 500 ␮g bid
Calverley et al (2003)35 Salmeterol 50 ␮g bid Placebo 52 wk
Salmeterol 50 ␮g/fluticasone 500 ␮g bid combination Fluticasone 500 ␮g bid

www.chestjournal.org CHEST / 125 / 1 / JANUARY, 2004 251

Downloaded From: http://journal.publications.chestnet.org/ on 02/16/2014

 Long-Acting Bronchodilators and GOLD
Treatment Goals
The GOLD guidelines4 specify the following prac-
tical objectives that define effective disease manage-
ment: to relieve symptoms, improve exercise toler-
ance, prevent and treat exacerbations, prevent and
treat complications, improve health status, reduce
mortality, and prevent disease progression. The
GOLD guidelines also note that these objectives
should be achieved with a minimum of side effects.
In the following sections of this article we will
compare published clinical trial data for each long-
acting bronchodilator in the context of the objectives
recommended by the GOLD committee. While
preventing disease progression, preventing and
treating complications, and reducing mortality are
important goals of treatment, there have as yet been
no published studies that have investigated the effi-
cacy of long-acting bronchodilators with respect to
these outcomes. We have, therefore, confined our
review to the effects of these treatments on relieving
symptoms, increasing exercise tolerance, preventing
exacerbations, and improving health status. We have
also reviewed the effects on lung function data since,
although improved lung function is not explicitly
stated by the GOLD as an objective of treatment, it
is a fundamental measure of bronchodilator efficacy.
A summary of the findings is shown in Table 2.
Lung Function
All the long-acting bronchodilators have shown
substantial improvements in lung function compared
with placebo, although comparisons are made diffi-
cult by the multitude of end points employed. All
three long-acting bronchodilators produce acute im-
provements in lung function that are similar to those
observed with ipratropium. These are sustained for
24 h in the case of tiotropium, and 12 h in the cases
of salmeterol and formoterol,5,19,20,24,25,28,32 although
the improvement in FEV1 area under the curve from
0 to 12 h with salmeterol compared with ipratropium
Table 2—Summary of Long-Acting Bronchodilator Efficacy in Lung Function and GOLD Treatment Goals*
Improve Lung Relieve
Drugs Function Symptoms
Salmeterol ⫹⫹ ⫹ ⫺
Formoterol ⫹⫹ ⫹ ⫺†
Tiotropium ⫹⫹⫹ ⫹⫹
*⫺ ⫽ no demonstrated efficacy; ⫹/⫺ ⫽ equivocal efficacy; ⫹ ⫽ superior to placebo; ⫹⫹ ⫽ superior to placebo and ipratropium; ⫹⫹⫹ ⫽ su-
perior to placebo, ipratropium, and salmeterol.
†One study with ⬍ 50 patients found an improvement in exercise tolerance with formoterol.
‡No published data.
252
Downloaded From: http://journal.publications.chestnet.org/ on 02/16/2014
was seen at only half the time points in one of two
studies.5,28 With long-term administration, both
tiotropium and salmeterol significantly improved
morning trough (predose) FEV1 when compared
with both placebo and ipratropium at all time points
over 1 year.5,19,20
Combination therapies appear to improve the
bronchodilator effects of long-acting ␤2-agonists.
The simultaneous administration of salmeterol and
ipratropium produced an additive effect on postdose
FEV1,29 and a combination with either theophylline
or fluticasone increased the effect of salmeterol on
trough FEV1.33 The combination of formoterol twice
daily and ipratropium four times daily also improved
morning trough FEV1 compared with a combination
of salbutamol and ipratropium.26
All the long-acting bronchodilators, therefore,
have shown efficacy over periods of either 12 h
(salmeterol, formoterol) or 24 h (tiotropium). The
only direct comparison of different classes of long-
acting bronchodilators found that salmeterol and
tiotropium had similar effects on day 1, but peak and
average FEV1 were significantly higher with tiotro-
pium from day 15 onwards, and trough FEV1 was
significantly higher at 24 weeks.21,22 It appears that
the difference at the end of the study (24 weeks) may
be due, in part, to the fact that the response to
tiotropium was maintained, whereas the response to
salmeterol decreased over the time course of the
study,21 suggesting the development of tachyphy-
laxis.
Other studies have not demonstrated tachyphy-
laxis to the bronchodilator effect of salmeterol. The
latter observation may reflect the relatively shorter
(12 to 16 weeks) duration of most trials, since
another 24-week trial showed that the bronchodila-
tor efficacy of salmeterol relative to placebo dimin-
ished over time, suggesting tachyphylaxis.31 Such
tachyphylaxis would be consistent with the reduction
in bronchodilator efficacy seen in long-term (90-day)
studies of albuterol (since both albuterol and salme-
terol are partial ␤-agonists).38 In contrast, a 12-
month study of formoterol (a full ␤-agonist) did not
Improve Exercise Reduce Exacerbation
Tolerance Frequency Improve Health Status
⫹/⫺ ⫹/⫺
⫹ ⫹⫹
‡ ⫹⫹ ⫹⫹
Reviews
 reveal evidence of tachyphylaxis, although the earli-
est time point assessed in this study was 3 months
after initiation of study drug, so that the possibility of
early onset tachyphylaxis cannot be excluded.25
Symptoms
One of the most important symptoms in COPD is
dyspnea, or the perception of breathlessness.39 Most
of the studies examined here used the transition
dyspnea index (TDI), a validated instrument for the
measurement of dyspnea that is incurred by activities
of daily living in COPD.40 The TDI focal score is a
total of three component scales, and a 1-U change is
considered to be clinically meaningful.41
Tiotropium has been shown to improve TDI focal
scores compared with both placebo and ipratropium
at all time points over a 12-month period.19,20 Sal-
meterol did not improve TDI focal scores compared
with theophylline,32 and it does not provide consis-
tent benefit compared with placebo.5,22,28,34 Formot-
erol, 24 ␮g bid, but not 12 ␮g bid, improved TDI
focal scores compared with placebo over 3 months,23
but formoterol has not yet been compared on this
measure with other active treatments.
A second measure of COPD symptoms is patient
use of “reliever” short-acting ␤2-agonists, since a
treatment that effectively reduces symptoms should
also reduce the need for rescue medication. Tiotro-
pium reduced the use of rescue medication com-
pared with both placebo and ipratropium,19,20
whereas salmeterol at 50 ␮g and 100 ␮g reduced
rescue medication use compared with placebo but
not compared with ipratropium or theophyl-
line.5,6,29,32,34 The effect of formoterol on the use of
rescue medication has not been reported.
In many of these studies,5,6,19,27–29,32 COPD symp-
toms were also assessed by means of a patient diary
record. The details of this assessment varied from
study to study, but always involved assigning a score
either to overall symptoms or to a series of symptoms
commonly associated with COPD, such as cough,
breathlessness, and chest tightness; however, there
was no consistent effect of any long-acting broncho-
dilator on these outcomes, which may suggest that
this instrument is not a sensitive outcome measure in
COPD.
The evidence suggests, therefore, that tiotropium
is superior to ipratropium in relieving dyspnea, and it
also appears to have a more consistent effect com-
pared with placebo than does salmeterol. There are
fewer data for formoterol, but the data that do exist
suggest that formoterol is superior to placebo.23–25
www.chestjournal.org
Downloaded From: http://journal.publications.chestnet.org/ on 02/16/2014
Exercise Tolerance
There have been few publications describing
large-scale studies of the effects of long-acting bron-
chodilators on exercise tolerance in COPD, and
those that have been published have yielded largely
negative results. Formoterol did not improve shuttle
walking test distances compared with placebo,23 and
salmeterol did not improve the distance walked in
the 6-min walking test.5,6,28 A more elaborate test of
exercise performance is cycle ergometry, and a small
study42 has found a significant effect of formoterol
on time to exhaustion during an incremental work
rate protocol. Recent presentations at national meet-
ings indicate that larger studies (with ⬎ 50 patients)
using cycle ergometry have been performed compar-
ing the effects of tiotropium and placebo on exer-
tional dyspnea (Borg scale) and endurance time.
While results of these studies in abstract form43,44
suggest that tiotropium improves exercise tolerance,
definitive peer-reviewed publications have not yet
appeared.
Exacerbations
Although some small studies have investigated the
role of long-acting bronchodilators in the treatment
of exacerbations, this area is beyond the scope of the
current article. Therefore, we will focus here on the
evidence that long-acting bronchodilators can reduce
the incidence and/or severity of exacerbations in
COPD.
Assessing the incidence or severity of exacerba-
tions relies on a robust operational definition, and as
yet no consensus has been reached on how exactly to
define an exacerbation in COPD.45 Commonly, ex-
acerbations are defined on the basis of an increase in
symptoms using a modification of the criteria of
Anthonisen et al46 (increased cough and sputum,
sputum purulence, and increased dyspnea), although
a consensus panel47 has proposed a definition based
also on a need for increased health-care resources
(eg, prescription of additional medications, or physi-
cian or hospital visits).
The studies reviewed here have used differing
definitions of exacerbations, thus inevitably making
comparisons more difficult. The studies of salmet-
erol have used a definition based on a need for a
change in drug therapy and/or hospitalization,
whereas the definition of exacerbations in the tiotro-
pium studies was based on an increase in two or
more respiratory symptoms for at least 3 days (which,
for the majority of events, required additional med-
ication). The formoterol studies defined exacerba-
tions using both symptoms and health-care use. The
symptom-based definition, “bad days,” was broadly
CHEST / 125 / 1 / JANUARY, 2004 253
 based (worsening in two or more symptoms on any 1
day, coupled with a ⬎ 20% reduction in peak expi-
ratory flow rate), and so is likely to have detected
events that would not conventionally be thought of as
exacerbations. The health-care– based definition was
similar to that used in the salmeterol studies.
In addition to employing varying definitions of
exacerbations, the studies reviewed here are of dif-
ferent lengths. The differences in study duration are
important because exacerbations, in general, occur
with an average annual frequency of one to two times
and at most only a few times per year in any
individual patient, so that longer studies are more
likely to reveal significant effects.48
The 1-year studies19,20,22 of tiotropium found that
it reduced the incidence of symptom-based exacer-
bations, the time to first exacerbation, and the time
to first hospitalization compared with either placebo
or ipratropium. However, although tiotropium also
reduced the incidence of hospitalizations for COPD
compared with placebo, it did not significantly re-
duce the incidence of hospitalizations compared with
ipratropium.
Formoterol has been shown to reduce the number
of bad days compared with placebo and theophylline,
but not with ipratropium.23–25 Formoterol at 24 ␮g,
but not lower doses, also reduced the number of
exacerbations that were serious enough to require
additional therapy compared with placebo.26,27
Salmeterol alone did not reduce the exacerbation
rate compared with placebo in six of seven stud-
ies,5,6,22,28,29,31,34 although the combination of salme-
terol plus ipratropium did29; in one study35 salmet-
erol did reduce the exacerbation rate in patients with
a history of exacerbations. Salmeterol also does not
seem to have a consistent effect on exacerbations
compared with ipratropium,5,28 and did not reduce
the incidence of exacerbations compared with the-
ophylline.32
In summary, tiotropium has been shown to reduce
the incidence of exacerbations over 12 months com-
pared with both placebo and ipratropium. Salmet-
erol did not significantly reduce the incidence of
exacerbations compared with placebo in five of six
12- to 16-week studies, although a combination of
salmeterol and ipratropium may provide a protective
effect against exacerbations. Formoterol at 24 ␮g,
but not 12 ␮g, appears to offer a protective effect
against moderate-to-severe exacerbations over 12
months, while lower doses reduce the incidence of
mild exacerbations.
Health Status
Health status is a broad term that encompasses the
patient’s overall health, with particular emphasis on
254
Downloaded From: http://journal.publications.chestnet.org/ on 02/16/2014
the impact of impaired health on his or her quality of
life. Measures of health status can be either generic
(ie, applicable across a range of diseases) or disease
specific. In COPD, the most commonly used mea-
sure of health status is the disease-specific St.
George Respiratory Questionnaire (SGRQ), which
consists of three components (symptoms, activity,
and impacts).49 The Chronic Respiratory Disease
Questionnaire is also commonly used.50 For both
scales, thresholds of change have been defined that
are regarded as clinically meaningful.51
Tiotropium improved health status as measured by
SGRQ compared with both placebo and ipratropi-
um.19,20 Tiotropium, but not salmeterol, improved
SGRQ scores compared with placebo, and more
patients achieved a clinically meaningful improve-
ment in SGRQ scores with tiotropium vs placebo as
compared with salmeterol vs placebo.22 Formoterol
was superior on the SGRQ total score to placebo and
ipratropium over 12 weeks, but not to placebo over 1
year or theophylline over 12 months.24,25,27
Most studies of salmeterol have not shown any
significant effect on health status compared with
either placebo,28,30,31,34,35 ipratropium,5,28 or theoph-
ylline.32 The combination of salmeterol and ipratro-
pium also did not improve overall health-related
quality of life compared with placebo, although the
combination did provide clinically and statistically
significant improvements in the SGRQ symptoms
domain.30
Only one study6,7 has found a significant effect of
salmeterol on health status compared with placebo,
finding a significant benefit of low-dose (50 ␮g)
salmeterol compared with high-dose (100 ␮g bid)
salmeterol. The latter interesting result was hypoth-
esized to be due to greater CNS stimulation as
evidenced by the higher incidence of tremor in the
high-dose group. These effects may have led to sleep
disturbances and worse overall health status, thereby
counteracting the improvements that would be ex-
pected due to a reduction in respiratory symptoms.
Studies published to date, therefore, show that
both tiotropium and formoterol have demonstrated
improvements in health status compared with pla-
cebo and ipratropium. However, while there may be
some benefit on health status with salmeterol com-
pared with placebo, no consistent effect has been
demonstrated.
Safety and Tolerability
All long-acting bronchodilators appear to have
good safety and tolerability profiles. Tiotropium was
associated with a higher incidence of dry mouth (a
class effect of anticholinergics) in all the studies
Reviews
 reviewed, although most cases were mild and there
were few resultant withdrawals. Formoterol ap-
peared to be associated with increased tachycardia
and tremor in one study.23 Both tremor and tachy-
cardia are a class effect of ␤2-agonist use and are
unlikely to be problematic at recommended doses
unless the patient suffers from preexisting arrhyth-
mia and hypoxemia.52 Rennard et al28 reported that
salmeterol was associated with an increased inci-
dence of adverse events related to the ear, nose, and
throat, although this was not found in other trials,
and salmeterol 100 ␮g was associated with a higher
incidence of tremor.6
Discussion
This review has sought to examine the evidence
for the efficacy of long-acting bronchodilators in
COPD from within the framework provided by the
GOLD guidelines. To this end, we have examined
the mechanistic differences between ␤2-agonists and
anticholinergics, and reviewed the results of impor-
tant studies of three long-acting bronchodilators:
a once-daily M3-antagonist (tiotropium), and two
twice-daily ␤2-agonists (salmeterol and formoterol).
The trials reviewed here used different patient
inclusion criteria, employed somewhat different end
points, and were of varying duration. These differ-
ences in study design and subject characteristics
likely influenced the results; therefore, it is difficult
to judge the relative efficacy of different pharmaco-
therapeutic agents based on these trials, underscor-
ing the need for more head-to-head clinical trials.
The studies reviewed have shown that all the
available long-acting bronchodilators are effective in
improving lung function, but that they vary in their
effects on other clinical outcomes. Salmeterol, in
particular, has demonstrated inconsistent efficacy in
symptom relief, prevention of exacerbations, and
improvement of health status. This conclusion is in
accord with a formal meta-analysis of the use of
salmeterol in COPD.53 Formoterol, also, has dem-
onstrated inconsistent effects on symptoms com-
pared with placebo, although high-dose formoterol
(24 ␮g) does appear to reduce moderate-to-severe
exacerbations. Of the three long-acting bronchodila-
tors, only tiotropium has demonstrated superiority to
both placebo and ipratropium on the outcomes
reviewed, except for exercise tolerance (for which
published data are not yet available). These different
effects on outcomes may reflect differences in either
the mechanism or the duration of action of these
agents, or a combination of both.
As discussed earlier, the GOLD guidelines do not
provide much guidance on the role of long-acting
www.chestjournal.org
Downloaded From: http://journal.publications.chestnet.org/ on 02/16/2014
bronchodilators in COPD, an inevitable conse-
quence of the paucity of data available when the
guidelines were drafted; however, based on the
review of data available from clinical trials now
published, it is reasonable to propose some specific
recommendations with regard to the relative role of
tiotropium, formoterol, and salmeterol in the main-
tenance treatment of COPD. An algorithm for the
introduction of the various treatment alternatives,
based on the framework of the GOLD guidelines, is
suggested in Table 3 and Figure 1. Some of the
principles that led to the development of this algo-
rithm will be described briefly below.
The introduction of bronchodilator and other
treatments is stratified according to the classification
and staging of COPD proposed in the GOLD guide-
lines. Table 3 illustrates this classification and staging
along with the defining pulmonary function impair-
ments at each level, and what are regarded as typical
symptoms for each stage. The fact that individual
patients might not exactly fit all categories shown in
Table 3 is noted. Instead, the stratification chosen is
meant as a general guide to disease severity. A typical
age of presentation is assigned to each level of
disease severity, and again this is meant only as a
general expectation recognizing that individual pa-
tients may differ in age of presentation due to
varying disease susceptibility and rates of disease
progression. Also, patients often consciously or un-
consciously avoid symptoms by adjusting their life-
style, thus choosing to limit their range and magni-
tude of physical activities to prevent experiencing
dyspnea. In these patients, maintenance bronchodi-
lator therapy might permit an increase in physical
activity while not necessarily reducing breathless-
ness. The list of investigations in Table 3 is also
meant as a guide to what might be deemed an
appropriate degree of investigation for each stage.
While Table 3 separates bronchodilator therapy from
other treatments, Figure 1 focuses on inhaled ther-
apies, including corticosteroids, in order to illustrate
a proposed three-step approach.
Figure 1 illustrates two alternative pathways for
the staged introduction of long-acting bronchodilator
therapy with increasing COPD severity. Both path-
ways lead from as-needed, short-acting bronchodila-
tors alone in mild disease (stage I) to a single
long-acting bronchodilator plus an as-needed, short-
acting agent in moderate disease (stage II), to the
combination of long-acting anticholinergic and long-
acting ␤2-adrenergic therapy with progression of
disease severity with or without the addition of
inhaled corticosteroids for patients with frequent
exacerbations or inadequate symptom control de-
spite optimal treatment with bronchodilators alone
(stage III/IV). However, available evidence, already
CHEST / 125 / 1 / JANUARY, 2004 255
 256
Table 3—COPD Staging by Symptoms, Pulmonary Function, and Management Algorithms*

GOLD Typical
Classifications Stage Age, yr Typical Symptoms Appropriate Investigations FEV1/FVC, % FEV1, % Predicted Bronchodilator Therapy Other Treatments

At risk O ⬎ 20 None None Normal Normal None Smoking cessation

Mild I ⬎ 35 Productive cough Screening spirometry ⬍ 70 ⱖ 80 As needed, short-acting Smoking cessation
bronchodilator (eg,
ipratropium, albuterol
or combination
therapy)
Moderate II ⬎ 50 Productive cough; Spirometry before and after ⬍ 70 ⬍ 80 and ⱖ 50 Tiotropium (algorithm Smoking cessation;
reduced physical bronchodilator step 1); with or rehabilitative exercise
activity with or without without short-acting

Downloaded From: http://journal.publications.chestnet.org/ on 02/16/2014

exertional dyspnea, bronchodilator (eg,
episodes of acute albuterol)†
bronchitis
Moderately III ⬎ 60 Productive cough; Annual pulmonary function ⬍ 70 ⬍ 50 and ⱖ 30 Tiotropium; long-acting Smoking cessation;
severe dyspnea with moderate testing; annual arterial ␤-adrenergic rehabilitative
exertion; occasional blood gas analysis; chest bronchodilator exercise;
exacerbations radiograph (algorithm step 2) [eg, supplemental oxygen
salmeterol or if indicated; with or
formoterol]‡§ without trial of ICS㛳
Severe IV ⬎ 70 Productive cough; Six-monthly pulmonary ⬍ 70 ⬍ 30 Tiotropium; long-acting Smoking cessation;
dyspnea with mild function testing; annual ␤-adrenergic rehabilitative
exertion or at rest; arterial blood gas bronchodilator (eg, exercise; with or
frequent exacerbations; analysis; chest radiograph salmeterol or without supplemental
with or without ankle formoterol); with or oxygen; ICS
swelling without (algorithm step 3)㛳
methylxanthine‡§
*This is a general schema. Individual patients may not exactly fit the classifications of COPD shown.
†Alternatively, the right-hand pathway shown in Figure 1 could be selected at this step.
‡A short-acting bronchodilator can be used for rescue medication.
§Low-dose methylxanthines can be prescribed if the response to inhaled bronchodilator therapy is insufficient.
㛳ICS indicated if repeated exacerbations requiring treatment with antibiotics and/or oral glucocorticoids.

Reviews
 Figure 1. The proposed three-step algorithm (1, 2, and 3; bold arrows) for mild, moderate, and severe
COPD. #Alternatively, the right-hand pathway (arrows) could be selected at this stage.§A short-acting
bronchodilator can be used for rescue medication. †ICS indicated if repeated exacerbations requiring
treatment with antibiotics or oral glucocorticoids; or if favorable response (decreased symptoms,
increased lung function, and/or decreased health-care utilization). ‡Low-dose methylxanthines can be
prescribed if the response to inhaled bronchodilator therapy is insufficient.

reviewed, suggests certain advantages in choosing
the left-hand pathway (in bold in figure) in less
severe disease (stage II). The rationale for preferring
the three-step algorithm (presented in the left of Fig
1) over alternatives is presented below.
All the long-acting bronchodilators have demon-
strated superior efficacy on at least some outcomes
compared with short-acting bronchodilators. There-
fore, the use of short-acting bronchodilators should
be restricted to symptomatic relief for patients with
infrequent, intermittent symptoms, and long-acting
bronchodilators should be used for maintenance
treatment.
Patients whose symptoms are not adequately con-
trolled with short-acting bronchodilators, or whose
use of relief medication becomes more frequent or
regular rather than intermittent, should be given a
long-acting bronchodilator. The proposed algorithm
provides a three-step schedule for introducing these
to the patient, depending on the level of disease
severity.
The evidence reviewed above suggests that, of the
long-acting bronchodilators, tiotropium provides the
most consistent improvements on the widest range
of outcomes. Therefore, the recommended step 1 is
the introduction of tiotropium. As with other long-
acting bronchodilators, tiotropium should be given in
combination with a reliever medication (short-acting
rescue ␤2-agonist medication was available to all
patients in the studies reviewed).
Studies of the combination of long-acting anticho-
linergics and long-acting ␤2-agonists have not been
published. However, both salmeterol and formoterol
in combination with ipratropium have additive ef-

www.chestjournal.org CHEST / 125 / 1 / JANUARY, 2004 257

Downloaded From: http://journal.publications.chestnet.org/ on 02/16/2014

 fects on lung function and other outcomes, so that it
seems reasonable to expect that the combination
of long-acting anticholinergics and long-acting ␤2-
agonists will provide additive benefits similar to
those observed with a combination of short-acting
bronchodilators. In patients with more severe symp-
toms, it is reasonable, therefore, to add salmeterol or
formoterol to tiotropium.
As suggested in the GOLD guidelines, other
pharmacotherapeutic agents, such as an inhaled
corticosteroid (ICS) and theophylline, should be
reserved for patients whose symptoms are not ade-
quately controlled and/or have frequent exacerba-
tions despite treatment with inhaled bronchodilators,
although these recommendations may change as
clinical trials of combinations of ICS and long-acting
bronchodilators are published.27,34,35 The placement
of these in Table 3 and Figure 1 follows the GOLD
precedent.
The treatment algorithm shown represents the
first attempt to provide an evidence-based rationale
for the appropriate use of long-acting bronchodila-
tors in COPD. In future, new studies comparing
combinations of long-acting bronchodilators with
one another and with ICS will help to define further
the role of combination therapy in COPD. More-
over, since some of the treatment goals listed by
GOLD have not been adequately studied, it is
important for future studies to address the potential
effects of long-acting bronchodilators on long-term
outcomes, such as disease progression and mortality.
However, a substantial body of data on these long-
acting bronchodilators has already been published
and, as reviewed here, the evidence for the utility of
long-acting bronchodilators, alone and in different
combinations (with or without ICS), in COPD is well
established.
References
1 Jones PW. Issues concerning health-related quality of life.
Chest 1995; 107:187S–193S
2 Oostenbrink JB, Rutten-van Mölken MPMH, Koch P, et al.
Costs of COPD exacerbations [abstract]. Am J Respir Crit
Care Med 2001; 163:A642
3 World Health Report 2000. Health systems: improving per-
formance. Annex table 3. Geneva, Switzerland: World Health
Organization, 2000
4 Pauwels RA, Buist AS, Calverley PMA, et al. Global strategy
for the diagnosis, management, and prevention of chronic
obstructive pulmonary disease: NHLBI/WHO Global Initia-
tive for Chronic Obstructive Lung Disease (GOLD) Work-
shop Summary. Am J Respir Crit Care Med 2001; 163:1256 –
1276 (updated 2003; available at www.goldcopd.com.
Accessed December 19, 2003)
5 Mahler DA, Donohue JF, Barbee RA, et al. Efficacy of
salmeterol xinafoate in the treatment of COPD. Chest 1999;
115:957–965
6 Boyd G, Morice AH, Pounsford JC, et al. An evaluation of
258
Downloaded From: http://journal.publications.chestnet.org/ on 02/16/2014
salmeterol in the treatment of chronic obstructive pulmonary
disease (COPD). Eur Respir J 1997; 10:815– 821
7 Jones PW, Bosh TK. Quality of life changes in COPD patients
treated with salmeterol. Am J Respir Crit Care Med 1997;
155:1283–1289
8 Gross NJE, Skorodin MS. Cholinergic bronchomotor tone in
COPD: estimates of its amount in comparison with that in
normal subjects. Chest 1989; 96:984 –987
9 Barnes PJ. Managing chronic obstructive pulmonary disease.
London, UK: Science Press, 1999; 1– 80
10 Barnes PJ. Bronchodilators: basic pharmacology. In: Calver-
ley PMA, Pride, N, eds. Chronic obstructive pulmonary
disease. London, UK: Chapman & Hall, 1995; 391– 417
11 Lulich KM, Goldie RG, Paterson JW. ␤-Adrenoceptor func-
tion in asthmatic bronchial smooth muscle. Gen Pharmacol
1988; 19:307–311
12 Combivent Aerosol Study Group. In chronic obstructive
pulmonary disease, a combination of ipratropium and albu-
terol is more effective than either agent alone: an 85-day
multicenter trial. Chest 1994; 105:1411–1419
13 Barnes PJ. Muscarinic receptor subtypes in airways. Life Sci
1993; 52:521–527
14 Disse B. Antimuscarinic treatment for lung diseases from
research to clinical practice. Life Sci 2001; 68:2257–2564
15 Jacoby DB, Fryer AD. Interaction of viral infections with
muscarinic receptors. Clin Exp Allergy 1999; 29:59 – 64
16 Dowling RB, Johnson M, Cole PJ, et al. Effect of salmeterol
on Haemophilus influenzae infection of respiratory mucosa in
vitro. Eur Respir J 1998; 11:86 –90
17 Dowling RB, Johnson M, Cole PJ, et al. Effect of fluticasone
propionate and salmeterol on Pseudomonas aeruginosa infec-
tion of the respiratory mucosa in vitro. Eur Respir J 1999;
14:363–369
18 Johnson M, Rennard S. Alternative mechanisms for long-acting
␤2-adrenergic agonists in COPD. Chest 2001; 120:258–270
19 Casaburi R, Mahler DA, Jones PW, et al. A long-term
evaluation of once-daily inhaled tiotropium in chronic ob-
structive pulmonary disease. Eur Respir J 2002; 19:217–224
20 Vincken W, van Noord JA, Greefhorst APM, et al. Improved
health outcomes in patients with COPD during 1 yr treat-
ment with tiotropium. Eur Respir J 2002; 19:209 –216
21 Donohue JF, van Noord JA, Bateman ED, et al. A 6-month
placebo-controlled study comparing lung function and health
status changes in COPD patients treated with tiotropium or
salmeterol. Chest 2002; 122:47–55
22 Brusasco V, Hodder R, Miravitlles M, et al. Health outcomes
in a 6-month placebo controlled trial of once-daily tiotropium
compared with twice-daily salmeterol in patients with COPD.
Thorax 2003; 58:399 – 404
23 Aalbers R, Ayres J, Backer V, et al. Formoterol in patients
with chronic obstructive pulmonary disease: a randomized,
controlled, 3-month trial. Eur Respir J 2002; 19:936 –943
24 Dahl R, Greefhorst LAPM, Nowak D, et al. Inhaled formot-
erol dry powder versus ipratropium bromide in chronic
obstructive pulmonary disease. Am J Respir Crit Care Med
2001; 164:778 –784
25 Rossi A, Kristufek P, Levine BE, et al. Comparison of the
efficacy, tolerability, and safety of formoterol dry powder and
oral, slow-release theophylline in the treatment of COPD.
Chest 2002; 121:1058 –1069
26 D’Urzo AD, De Salvo MC, Ramirez-Rivera A, et al. In
patients with COPD, treatment with a combination of for-
moterol and ipratropium is more effective than a combination
of salbutamol and ipratropium: a 3-week, randomized,
double-blind, within-patient, multicenter study. Chest 2001;
119:1347–1356
27 Szafranski W, Cukier A, Ramirez A, et al. Efficacy and safety
Reviews
 of budesonide/formoterol in the management of chronic
obstructive disease. Eur Respir J 2003; 21:74 – 81
28 Rennard SI, Anderson W, ZuWallack R, et al. Use of a
long-acting inhaled ␤2-adrenergic agonist, salmeterol xin-
afoate, in patients with chronic obstructive pulmonary dis-
ease. Am J Respir Crit Care Med 2001; 163:1087–1092
29 van Noord JA, de Munck DRAJ, Bantje T, et al. Long-term
treatment of chronic obstructive pulmonary disease with
salmeterol and the additive effect of ipratropium. Eur Respir J
2000; 15:878 – 885
30 Rutten-van Mölken MPMH, Roos B, van Noord JA. An
empirical comparison of the St George’s Respiratory Ques-
tionnaire (SGRQ) and the Chronic Respiratory Disease Ques-
tionnaire (CRQ) in a clinical trial setting. Thorax 1999;
54:995–1003
31 Chapman KR, Arvidsson P, Chuchalin AG, et al. The addition
of salmeterol 50 ␮g bid to anticholinergic treatment in
patients with COPD: a randomized, placebo controlled trial.
Can Respir J 2002; 9:178 –185
32 ZuWallack RL, Mahler DA, Reilly D, et al. Salmeterol plus
theophylline combination therapy in the treatment of COPD.
Chest 2001; 119:1661–1670
33 Cazzola M, Di Lorenzo G, Di Perna F, et al. Additive effects
of salmeterol and fluticasone or theophylline in COPD. Chest
2000; 118:1576 –1581
34 Mahler DA, Wire P, Horstman D, et al. Effectiveness of
fluticasone propionate and salmeterol combination delivered
via the diskus device in the treatment of chronic obstructive
pulmonary disease. Am J Respir Crit Care Med 2002; 166:
1084 –1091
35 Calverley P, Pauwels R, Vestbo J, et al. Combined salmeterol
and fluticasone in the treatment of chronic obstructive pul-
monary disease: a randomised controlled trial. Lancet 2003;
361:449 – 456
36 Kottakis J, Cioppa GD, Creemers J, et al. Faster onset of
bronchodilation with formoterol than with salmeterol in
patients with stable, moderate to severe COPD: results of a
randomized, double-blind clinical study. Can Respir J 2002;
9:107–115
37 Benhamou D, Cuvelier A, Muir JF, et al. Rapid onset of
bronchodilation in COPD: a placebo-controlled study com-
paring formoterol (Foradil Aerolizer) with salbutamol (Ven-
todisk). Respir Med 2001; 95:817– 821
38 Rennard S, Serby C, Ghafouri M, et al. Extended therapy
with ipratropium is associated with improved lung function in
patients with COPD: a retrospective analysis of data from
seven clinical trials. Chest 1996; 110:62–70
39 Harver A, Schwartzstein RM, Killian KJ, et al. Understanding
dyspnoea in COPD [abstract]: mechanisms, measurement
and management. Eur Respir Rev 2002; 12:25–55
www.chestjournal.org
Downloaded From: http://journal.publications.chestnet.org/ on 02/16/2014
40 Mahler DA, Weinberg DH, Wells CK, et al. The measure-
ment of dyspnea: contents, interobserver agreement, and
physiologic correlates of two new clinical indexes. Chest 1984;
85:751–758
41 Witek TJ, Mahler DA. Validity and patterns of response of
dyspnea measures in a 6 month bronchodilator intervention
trial in COPD [abstract]. Am J Respir Crit Care Med 2002;
165:A266
42 Liesker JJ, Van De Velde V, Meysman M, et al. Effects of
formoterol (Oxis Turbuhaler) and ipratropium on exercise
capacity in patients with COPD. Respir Med 2002; 96:559 –
566
43 O’Donnell DE, Magnussen H, Aguilaniu B, et al. SPIRIVA®
(tiotropium) improves exercise tolerance in COPD [abstract].
Am J Respir Crit Care Med 2002; 165:A227
44 O’Donnell DE, Magnussen H, Aguilaniu B, et al. SPIRIVA®
(tiotropium) reduces exertional dyspnea in COPD [abstract].
Am J Respir Crit Care Med 2002; 165:A265
45 Jeffery PK, MacNee W, Mahler DA, et al. COPD exacerba-
tions: causes, impacts and implications for management. Eur
Respir Rev 2002; 18:1–24
46 Anthonisen NR, Manfreda J, Warren CPW, et al. Antibiotic
therapy in exacerbations of chronic obstructive pulmonary
disease. Ann Intern Med 1987; 106:196 –204
47 Rodriguez-Roisin R. Toward a consensus definition for
COPD exacerbations. Chest 2000; 117:398 – 401
48 Burge PS, Calverley PMA, Jones PW, et al. Randomised,
double blind, placebo controlled study of fluticasone propi-
onate in patients with moderate to severe chronic obstructive
pulmonary disease: the ISOLDE trial. BMJ 2000; 320:1297–
1303
49 Jones PW, Quirk FH, Baveystock CM, et al. A self-complete
measure of health status for chronic airflow limitation: the St
George’s Respiratory Questionnaire. Am Rev Respir Dis
1992; 145:1321–1327
50 Guyatt GH, Berman LB, Townsend M, et al. A measure of
quality of life for clinical trials in chronic lung disease. Thorax
1987; 42:773–778
51 Jones PW. Interpreting thresholds for a clinically significant
change in health status in asthma and COPD. Eur Respir J
2002; 19:398 – 404
52 Cazzola M, Imperatore F, Salzillo A, et al. Cardiac effects of
formoterol and salmeterol in patients suffering from COPD
with preexisting cardiac arrhythmias and hypoxemia. Chest
1998; 114:411– 415
53 Appleton S, Poole P, Smith B, et al. Long-acting ␤2-agonists
for chronic obstructive pulmonary disease patients with
poorly reversible airflow limitation. Cochrane Database Syst
Rev 2002; 3:CD001104
CHEST / 125 / 1 / JANUARY, 2004 259