Review

Respiration 2015;89:66–75 Received: August 22, 2014

Accepted: September 8, 2014
DOI: 10.1159/000368371
Published online: November 27, 2014

Current Perspectives on the Contribution of

Inhaled Corticosteroids to an Increased Risk
for Diabetes Onset and Progression in Patients
with Chronic Obstructive Pulmonary Disease
Felix J.F. Herth a, b Peter Bramlage c Dirk Müller-Wieland d
a
Department of Pneumology and Critical Care Medicine, Thoraxklinik, University of Heidelberg, and b Translational
Lung Research Center Heidelberg, Heidelberg, c Institute for Pharmacology and Preventive Medicine, Mahlow,
and d Department of Internal Medicine, Endocrinology, Diabetes and Metabolism, Asklepios Klinik St. Georg,
Hamburg, Germany

Key Words Introduction

Chronic obstructive pulmonary disease · Diabetes ·
Hyperglycemia · Inhaled corticosteroids Chronic obstructive pulmonary disease (COPD),
which is characterized by progressive airflow limitation,
is predicted to become the third leading cause of death
Abstract globally in 2030 [1]. COPD results from long-term expo-
Recent studies have suggested that inhaled corticosteroids sure to inhaled irritants (e.g. air pollution or smoking),
(ICS) play a role in the development of hyperglycemia and which cause robust inflammatory responses in the lungs
type-2 diabetes in patients with chronic obstructive pulmo- and several prominent symptoms, including shortness of
nary disease (COPD). Nevertheless, this corticosteroid-asso- breath, coughing, and sputum production. Moreover,
ciated adverse effect remains controversial. Moreover, the COPD is known to be associated with several important
pharmacokinetic properties and patient characteristics that chronic comorbid diseases [2]. In particular, recent evi-
might contribute to an increased risk for diabetes upon ICS dence has supported that COPD may constitute an im-
exposure have not been thoroughly investigated. In the portant risk factor for the development of type-2 diabetes
present review, we critically discuss current evidence re- [2, 3], and several studies have identified an increased
garding the relationship between ICS therapy in COPD pa- prevalence of diabetes in COPD patients [4–8].
tients and an increased risk for the incidence and progres- The treatment of COPD involves lifestyle modifica-
sion of type-2 diabetes. In addition, we address therapeutic tions as well as various pharmacotherapies. In fact, due to
conditions, clinical implications, and future perspectives re- the presence of inflammation in COPD, corticosteroids
lated to this potentially important ICS-associated adverse ef- have been used for decades to treat acute exacerbations of
fect in COPD patients. © 2014 S. Karger AG, Basel COPD (AECOPD) [9]. Current estimates suggest that
115.127.13.234 - 3/20/2017 3:28:34 AM

© 2014 S. Karger AG, Basel Prof. Dr. Felix J.F. Herth, FCCP
0025–7931/14/0891–0066$39.50/0 Department of Pneumology and Critical Care Medicine
Thoraxklinik, University of Heidelberg
Downloaded by:

E-Mail [email protected]
Amalienstrasse 5, DE–69126 Heidelberg (Germany)
www.karger.com/res
E-Mail felix.herth @ med.uni-heidelberg.de
corticosteroids are administered to over 70% of COPD structive pulmonary disease’ and ‘COPD’. These system-
patients in the USA and Europe [10]. These drugs can be atic searches were then extended through manual screen-
delivered either systemically or through the respiratory ing of the references included in the selected papers in or-
tract as inhaled corticosteroids (ICS). In this regard, the der to identify further supporting information.
most commonly used ICS for the treatment of COPD are
fluticasone propionate, budesonide, and beclometasone Study Selection
dipropionate. Although these corticosteroids are routine- The included references fulfilled the following criteria:
ly employed and generally considered to be safe and ef- (1) published in a peer-reviewed journal, (2) reported re-
fective for the short-term treatment of severe COPD-as- sults that indirectly or directly yielded insight into the re-
sociated symptoms [11–13], it is known that prolonged lationship between ICS and diabetes in COPD patients,
exposure to corticosteroids can lead to substantial side and (3) written in English. Relevant full-text articles were
effects [14]. Thus, there is ongoing controversy surround- evaluated and discussed by two reviewers in order to as-
ing the use of ICS [10, 15, 16], especially at high doses, sess quality. A flow diagram outlining our systematic lit-
which are routinely used in the clinic for the treatment of erature search is shown in figure 1.
COPD [17, 18].
Several recent studies have raised concern regarding
potential adverse effects associated with the use of ICS, Evidence Linking ICS to Diabetes Risk in COPD
including enhanced susceptibility to pneumonia, influ- Patients
enza, cataracts, and fractures as well as tuberculosis [19–
22]. In addition, investigations have supported the role of ICS-Induced Hyperglycemia
ICS in increasing the risk of hyperglycemia and type-2 Recent evidence supports the association of ICS ther-
diabetes in COPD patients [23, 24]. Nevertheless, this apy with insulin resistance, hyperglycemia, and diabetes
side effect of ICS therapy remains controversial [25, 26], in COPD patients (table 1). In fact, patients with type-2
and the mechanisms by which ICS may contribute to dia- diabetes, who were predominantly taking ICS for lung-
betes development remain ill defined. In particular, the related conditions, were found to exhibit poorer glycemic
roles of ICS pharmacokinetics and patient characteristics control [i.e. higher glycated hemoglobin (HbA1c) levels]
have not been thoroughly investigated. than those who did not receive ICS [27]. Furthermore,
In this review, we critically discuss current evidence Slatore et al. [23] conducted a prospective cohort study
regarding the relationship between ICS therapy in COPD involving 1,698 US veterans (approximately 85% with
patients and an increased risk for the incidence and pro- COPD) and found a dose-dependent effect of ICS therapy
gression of type-2 diabetes. In addition, we discuss thera- (beclometasone, flunisolide, and fluticasone) on serum
peutic conditions (e.g. dosing and pharmacokinetics), glucose concentrations in patients with self-reported dia-
clinical implications, and future perspectives related to betes [23]. Notably, the mean ICS daily doses used in the
this potential ICS-associated adverse effect in COPD pa- study were 621 μg (SD = 555) and 610 μg (SD = 553) for
tients. subjects with and without diabetes, respectively (reported
in triamcinolone equivalents). Among patients with dia-
betes, it was found that for each 100-μg increment in ICS
Methods dose there was an associated 1.82 mg/dl increase in serum
glucose concentration. Moreover, subjects treated with
Systematic Search of the Literature antidiabetic drugs showed an increase of 2.65 mg/dl in
Electronic literature searches were performed using serum glucose for every additional 100 μg of ICS. Thus,
PubMed in June 2014. To identify information related to when considering published data regarding the conver-
the association between ICS use and diabetes in COPD pa- sion of serum glucose to HbA1c values (i.e. every 29 mg/
tients, the term ‘inhaled corticosteroids’ was searched in dl increase in glucose equals a 1% rise in HbA1c) [28],
combination with each of the following terms: ‘diabetes’ these reported changes correspond to respective HbA1c
and ‘hyperglycemia’. Also, additional searches were per- augmentations of around 0.06 and 0.1% for every 100 μg
formed to identify factors that might contribute to ICS- of triamcinolone. Also, based on these findings it was es-
induced side effects, such as diabetes, in COPD patients. timated that subjects with diabetes taking 500 μg of fluti-
For this, the term ‘inhaled corticosteroids’ was searched in casone twice daily would experience an overall increase
combination with each of the following terms: ‘chronic ob- in serum glucose concentration of 72.8 mg/dl [23]. Strik-
115.127.13.234 - 3/20/2017 3:28:34 AM

ICS and Diabetes in COPD Respiration 2015;89:66–75 67

DOI: 10.1159/000368371
Downloaded by:
 Identification
Total records identified related to an
association between ICS use and diabetes or
factors that might impact diabetes risk in
COPD patients: 151

Screening

Number of articels further Excluded based on

examined: 49 relevance: 102

Identification

Relevant articles further Additional supporting Excluded based on

examined: 41 references identified: 28 quality: 8

Included
Fig. 1. Flow diagram showing our system-
atic literature search on the relationship be- Total included: 69
tween ICS and diabetes in COPD patients.

Table 1. Key studies suggesting that ICS therapy can increase diabetes risk and/or progression

Population Principal findings Reference

1,698 patients with and without Dose-dependent effect of ICS therapy on serum glucose concentrations in patients 23
diabetes (approximately 85% with with diabetes; each 100-μg increment in ICS dose led to a serum glucose increase of
COPD) 1.82 mg/dl, while subjects treated with antidiabetic drugs showed an increase of
2.65 mg/dl for every 100 μg of ICS; no ICS-related blood sugar changes were found
in nondiabetic patients
388,584 patients treated for Dose-dependent effect of ICS therapy on diabetes risk and progression; 34% 24
respiratory disease (i.e. asthma and increased risk of diabetes onset with ICS and 64% increase with high-dose ICS
COPD patients), of whom 30,167 (fluticasone-equivalent doses ≥1,000 μg/day). Also, ICS therapy led to a 34%
developed diabetes during the increased likelihood of progression to insulin use, whereas high-dose ICS therapy
5.5-year follow-up period led to a 54% increase
18,226 subjects with diabetes, of Patients with both diabetes and COPD who received a total corticosteroid DDD of 33
whom 5.9% had COPD (67.2% were ≥0.83/day displayed a 94% increased risk for diabetes-related hospitalization
given corticosteroids during the compared to those who did not receive corticosteroids; lower corticosteroid doses
12 months after study entry) (DDD <0.83/day) were not associated with an increased risk of diabetes-related
hospitalization

ingly, this correlates with a 2.5% elevation in HbA1c [28]. tion that systemic corticosteroids can induce hyperglyce-
Nevertheless, this study did not find similar ICS-associ- mia in COPD patients in the absence of diabetes. Indeed,
ated blood sugar changes in nondiabetic patients. That it was reported that prednisolone-treated COPD patients
being said, an investigation examining the effects of in- displayed a circadian cycle of hyperglycemia that oc-
haled budesonide on insulin sensitivity in nondiabetics curred in the afternoon and evening [30]. Also, the De-
with asthma and COPD found that glucose rose signifi- partment of Veterans Affairs Cooperative Study Group
cantly following ICS treatment in COPD patients [29]. In found that hyperglycemia occurred significantly more of-
line with this, recent studies have also supported the no- ten in subjects with AECOPD who were treated with sys-
115.127.13.234 - 3/20/2017 3:28:34 AM

68 Respiration 2015;89:66–75 Herth/Bramlage/Müller-Wieland

DOI: 10.1159/000368371
Downloaded by:
temic glucocorticoids compared to controls [31]. In this of diabetes complications in patients with both diabetes
regard, investigations have begun to elucidate the mecha- and COPD who were treated with corticosteroids (sys-
nisms by which corticosteroids might induce hyperglyce- temic and inhaled) [33]. A total of 18,226 subjects with
mia, including direct effects on hepatic gluconeogenesis diabetes were examined, of whom 5.9% had COPD.
[32]. Taken together, these data suggest that hyperglyce- Among these COPD patients, 67.2% were given cortico-
mia could represent an adverse effect associated with the steroids during the 12 months after study entry. It was
use of high-dose ICS treatments in COPD patients, espe- found that subjects with both diabetes and COPD dis-
cially when type-2 diabetes is comorbid. played an increased risk of diabetes-related hospitaliza-
tion upon high-dose corticosteroid therapy. In fact, there
ICS and an Increased Risk for Incidence and was a 94% increase in the likelihood of hospitalization for
Progression of Diabetes diabetes-related complications in those who received a
In addition to the possible short-term danger of ICS- total corticosteroid defined daily dose (DDD) of ≥0.83/
induced hyperglycemia, Suissa et al. [24] recently con- day in comparison to those who did not receive cortico-
ducted a population-based cohort study to examine steroid therapy. In contrast, lower corticosteroid doses
whether ICS use and dose could increase the risk of de- (DDD <0.83/day) were not associated with an increased
veloping diabetes in COPD patients. Overall, the cohort risk for diabetes-related hospitalization. Therefore, it was
included 388,584 subjects treated for respiratory disease suggested that routine use of high-dose corticosteroids
(i.e. asthma and COPD patients), and 7.8% were found to might need to be avoided in individuals with COPD and
develop diabetes during the 5.5-year follow-up period. It comorbid diabetes, and when corticosteroids are used
was determined that treatment with ICS (i.e. beclometa- they should be administered in the context of close blood
sone, budesonide, triamcinolone, fluticasone, and fluni- glucose monitoring. Also, it was proposed that efficacy
solide; all doses converted to fluticasone equivalents) was should be reviewed within 4–8 weeks after commencing
associated with a 34% increased risk of diabetes onset, as ICS therapy in patients with COPD and diabetes.
defined by initiation of an oral antidiabetic medication.
Moreover, the diabetes risk was greatest in those patients Controversy Surrounding the Role of ICS in Diabetes
treated with high ICS doses. Indeed, there was a 64% in- In spite of the above findings, major clinical trials ex-
creased risk upon administration of fluticasone-equiva- amining the safety and efficacy of various ICS therapies
lent doses ≥1,000 μg/day. Moreover, the effect of ICS in COPD patients did not report increased rates of diabe-
therapy on diabetes progression was examined among a tes [17, 18, 34, 35]. However, it has been suggested that
subset of patients with respiratory disease and newly di- this may be due to the fact that these trials were not large
agnosed diabetes treated with oral hypoglycemic agents enough to detect the increased diabetes risk based on the
[24]. Notably, ICS therapy was associated with a 34% in- reported incidence of diabetes in COPD patients
creased likelihood of progression to insulin use, whereas (14.2/1,000/year) [24]. In addition, a recent retrospective
high-dose ICS led to a 54% increase. Therefore, ICS treat- analysis of double-blind control trials indicated that ICS
ment may not be ideal for patients with diabetes due to therapy in patients with asthma or COPD was not associ-
the potential for dose-dependent augmentation of blood ated with the increased risk of new onset diabetes or hy-
glucose levels, an increased risk of diabetes progression, perglycemia [26]. Furthermore, while another study re-
and a requirement for initiation/intensification of diabe- ported that elderly patients given oral corticosteroids
tes treatments [23, 24]. In this respect, it currently re- showed an increased risk of diabetes, ICS users did not
mains unclear whether ICS therapy in patients with [36]. Similarly, a further investigation observed no link
COPD and diabetes might also worsen long-term diabe- between ICS and diabetes risk in elderly patients [25].
tes-related complications, such as microvascular and That being said, studies analyzing data from the 1990s do
macrovascular events. not allow for the assessment of very high ICS doses (i.e.
similar to those currently used to treat COPD) [24]. Nev-
ICS and Diabetes-Related Hospitalization ertheless, a more recent small randomized trial examin-
Furthermore, ICS-mediated effects on serum glucose ing whether inhaled fluticasone propionate (440 μg twice
levels may contribute to diabetes-related hospitalizations. daily for 6 weeks) led to changes in HbA1c levels in pa-
Indeed, a recent retrospective study of claims data from tients with type-2 diabetes and asthma or COPD also ob-
the Australian Government Department of Veterans’ Af- served no effect of ICS treatment [37]. Finally, another
fairs examined whether there was a dose-dependent risk small study investigating the effects of inhaled beclometa-
115.127.13.234 - 3/20/2017 3:28:34 AM

ICS and Diabetes in COPD Respiration 2015;89:66–75 69

DOI: 10.1159/000368371
Downloaded by:
Table 2. Various randomized clinical trials examining the efficacy and safety of distinct doses of ICS in the treatment of COPD

ICS and dose* Population (study length) Reduction in AECOPD Reference

Budesonide/formoterol
320/9 μg 812 adults with moderate to severe COPD (12 months) 24% vs. placebo and 23% vs. formoterol 66
320/9 μg 1,022 COPD patients (12 months) 23% vs. placebo 67
320/9 μg 1,704 patients moderate to very severe COPD (6 months) 20–25% vs. placebo and formoterol 68
320/9 μg 1,964 adults with moderate to very severe COPD (12 months) 37% vs. placebo and 25% vs. formoterol 43
320/9 μg 1,293 adults with moderate to severe COPD (12 months) 36% vs. formoterol 69
320/9 μg 1,219 adult patients with a history of COPD (12 months) 35% vs. formoterol 44
160/9 μg 1,219 adult patients with a history of COPD (12 months) 26% vs. formoterol 44
160/9 μg 1,964 adults with moderate to very severe COPD (12 months) 41% vs. placebo and 29% vs. formoterol 43
Fluticasone p./salmeterol
500/50 μg 994 clinically stable patients with severe COPD and a history 35% vs. salmeterol 34
of repeated exacerbations (44 weeks)
500/50 μg 1,465 COPD patients (12 months) 25% vs. placebo 41
500/50 μg 6,112 COPD patients (3 years) 25% vs. placebo and 12% vs. salmeterol 17
250/50 μg 782 COPD patients (12 months) 30% vs. salmeterol 42
250/50 μg 797 COPD patients (12 months) 30% vs. salmeterol 40

* All doses were delivered twice daily. p. = Propionate.

sone dipropionate (2,000 μg/day for 2 weeks) on normal the safety and efficacy of various doses of inhaled
and elderly subjects found no effects on glucose metabo- budesonide and fluticasone propionate (in combination
lism [38]. with the β2-adrenergic agonists formoterol and salmeter-
Taken together, recent evidence indicates that high- ol) in COPD patients and have not observed large differ-
dose ICS therapy in COPD patients can lead to an in- ences with regard to reductions in acute exacerbations
creased risk for insulin resistance and hyperglycemia, as with lower ICS doses (table  2). Also, data suggest that
well as diabetes incidence and progression. However, this high-dose fluticasone (500 μg, twice daily) was not found
topic remains controversial and will require further in- to be advantageous in terms of lung function and quality
vestigation. Nevertheless, there is reason to consider the of life when compared to a 250-μg dose [17, 34, 40–42].
dangers associated with high-dose ICS in COPD patients Moreover, both high- and low-dose budesonide/for-
in the clinical setting. In fact, in order to reduce the risk moterol (320/9 or 160/9 μg; twice daily) improved pul-
of potential ICS-related side effects, the efficacy and phar- monary function and reduced AECOPD symptoms over
macokinetics of these drugs at low/moderate doses may a 1-year period in patients with moderate to severe COPD
need to be more thoroughly evaluated. [43, 44]. Therefore, future validation and implementa-
tion of such reduced doses could eliminate potential ICS-
ICS Dosing and Diabetes Risk related effects on hyperglycemia and diabetes risk. Thus,
Although ICS therapies are widely used to treat COPD, systematic identification of the lowest possible effective
optimal doses and lengths of treatment regimens remain ICS doses may represent a priority in improving thera-
to be adequately defined. This issue has recently become peutic safety in COPD patients. For this, head-to-head
a focus due to increased awareness regarding the substan- comparisons of low and high doses for existing and
tial risk for side effects associated with the use of high- emerging treatment strategies will be required.
dose ICS [19–24]. Dosing and pharmacokinetic proper-
ties (e.g. oral bioavailability, lung retention, and clear-
ance) play a critical role in the clinical efficacy and safety ICS Pharmacokinetics and Diabetes Risk
of ICS therapies. Although the use of high-dose ICS (i.e.
≥500 μg, twice daily) has become routine in the clinic, ICS Activity and Bioavailability
there is evidence to support the efficacy of more moderate The ICS dose is not the only variable that should be
ICS doses in the treatment of COPD [39]. For example, taken into account with regard to diabetes risk. Indeed,
recent randomized clinical trials have already evaluated distinct ICS therapies for COPD also display significant
115.127.13.234 - 3/20/2017 3:28:34 AM

70 Respiration 2015;89:66–75 Herth/Bramlage/Müller-Wieland

DOI: 10.1159/000368371
Downloaded by:
 Pharmacokinetic properties of ICS

Pulmonary
bioavailability

Long pulmonary
residence

Long half-life

Rapid clearance Glucocorticoid

receptor binding
On-site activation Lipid conjugation affinity Oral bioavailability

Enhanced ICS safety Increased/prolonged Systemic side effects

antiinflammatory (e.g. risk of diabetes
Fig. 2. How various pharmacokinetic vari- effects: onset/progression)
ables influence the safety and efficacy of ICS efficacy
ICS therapy.

variability in key pharmacokinetic properties (table  3), Table 3. Key pharmacokinetic characteristics of several ICS com-
which can result in differential activity or a risk for sys- monly used to treat COPD
temic side effects (fig. 2). In general, although different
ICS Key pharmacokinetic properties
ICS are rapidly cleared from the body, they display varied
levels of glucocorticoid receptor binding, oral/lung bio- Fluticasone High relative receptor binding (1,800)
availability, and absorption rates [45, 46]. For example, propionate Low oral availability (<1%)
considering the receptor binding of drugs commonly Absorption rate (4.8 h)
used to treat COPD, fluticasone propionate shows a high High half-life (14 h)
relative receptor affinity, whereas budesonide and fluni- Budesonide Intermediate relative receptor binding (935)
solide show intermediate and lower affinities, respective- Intermediate oral availability (11%)
ly [46]. However, this factor alone does not determine the Absorption rate (1.9 h)
efficacy of a given ICS and can even contribute to sys- Intermediate half-life (2.8 h)
Lipid conjugated
temic side effects. Indeed, a high pulmonary bioavailabil-
ity is also required for these drugs to function locally, and Beclometa-sone Low relative receptor binding (53)
it has been reported that pulmonary residence times are dipropionate High oral availability (15%)
high for fluticasone propionate but short for budesonide Low half-life (0.1 h)
On-site activation in lungs
and flunisolide [47]. In addition, variability in lung reten-
tion and mean absorption times must also be considered Flunisolide Low/intermediate relative receptor binding
in relation to adverse effects (e.g. fluticasone propionate (180)
Intermediate oral availability (7%)
and budesonide show absorption times of 4.9 and 1.8 h,
Low half-life (1.6 h)
respectively) [46].
With regard to nonlung absorption, increased oral bio- Triamcino-lone Low/intermediate relative receptor binding
acetonide (223)
availability (i.e. the proportion of the dose that is swal-
High oral availability (23%)
lowed) can lead to heightened systemic exposure and the Low half-life (2.0 h)
potential for side effects. However, when considering bio-
availability, measuring the systemic exposure to an inhaled Mometasone High relative receptor binding (2,200)
furoate Low oral availability (<1%)
drug can be challenging and requires a temporal compari- Intermediate half-life (4.5 h)
son of corticosteroid plasma concentrations for inhaled
115.127.13.234 - 3/20/2017 3:28:34 AM

ICS and Diabetes in COPD Respiration 2015;89:66–75 71

DOI: 10.1159/000368371
Downloaded by:
dosing with those obtained after intravenous dosing. Also, Intersubject Variability
the administered doses need to be large enough to produce In addition to the unique pharmacokinetic character-
measurable plasma concentrations (i.e. within the limit of istics of the various ICS, it has also been suggested that
reliable detection) [48]. In this regard, peak plasma levels interindividual variability may exist with regard to ICS
of budesonide occurred much earlier and were approxi- pharmacokinetics (e.g. rate of systemic lung absorption).
mately 20-fold higher than those of fluticasone propionate Notably, these differences could also contribute to ICS
[49]. In addition, ICS plasma concentrations can vary sig- dosing requirements and side effects. Indeed, as part of a
nificantly based on the type of delivery system used (e.g. small randomized clinical study, 9 normal subjects were
dry powder vs. metered-dose inhalers) [50]. Nevertheless, administered high-dose budesonide and fluticasone pro-
it has been demonstrated that ICS display a wide range of pionate (1,600 and 1,000 μg, respectively; both twice dai-
oral bioavailability, with fluticasone propionate showing ly) [59]. Blood samples were sequentially collected prior
<1%, budesonide having 11%, and beclometasone mono- to and at various times after the delivery of an ICS dose
propionate being the highest at 26% [46]. Thus, beclometa- in order to determine plasma drug concentrations and
sone monopropionate may be more likely to induce ad- pharmacokinetics. Notably, it was demonstrated that
verse effects based solely on the level of systemic exposure. there was considerable intersubject variability with re-
However, that being said, factors such as drug metabolism gard to the rate of lung absorption for both drugs. Also,
can also influence bioavailability and the likelihood of an some individuals displayed a more sustained plasma drug
ICS producing negative effects. In this regard, ICS are gen- level than would be predicted for budesonide and higher-
erally thought to be inactivated and metabolized via the than-expected plasma drug concentrations in the morn-
cytochrome P450 system [51]. Interestingly, inflammation ing compared to the evening for fluticasone propionate.
has been found to be associated with the downregulation Nevertheless, these results may need to be interpreted
of cytochrome P450 enzymes, which could lead to pro- with care as it has been suggested that the study of ICS
longed ICS activity in COPD patients [52]. absorption rates and systemic effects of budesonide and
fluticasone propionate in healthy individuals may not be
ICS Half-Life and Modification relevant to diseased subjects [49, 60].
Half-life also represents an important ICS characteris- Overall, current evidence indicates that the standard
tic. A low-dose ICS present for a prolonged period of time of care could be effectively altered in order to reduce the
is more likely to have a better safety profile than a high- ICS dose, thereby diminishing therapeutic side effects.
dose ICS with a short half-life. In this regard, fluticasone Considering the unique pharmacokinetic properties of
propionate shows a relatively long half-life (14 h) com- each ICS may help to determine which drugs are less
pared to other ICS, while the newer fluticasone furoate likely to contribute to the development and/or progres-
displays a half-life of 17–24 h [53]. For this reason, recent sion of diabetes and other adverse effects in COPD pa-
clinical trials have begun to evaluate the efficacy and safe- tients.
ty of fluticasone furoate for the treatment of COPD [54,
55]. Additionally, some ICS require bioactivation pro-
cesses in the lung, which can contribute to their safety Clinical Implications and Future Perspectives
profiles by locally restricting their activity [56]. Finally, it
is known that certain ICS have the potential to become Strikingly, recent ICS use rates for the treatment of
lipid modified within the lung, which can facilitate reten- COPD have been reported to be over 70% [10, 39]. How-
tion and slow release from the target tissue, thereby re- ever, these rates do not coincide with treatment guide-
ducing systemic side effects. In this respect, budesonide lines, which only recommend the use of ICS in late-stage
can be lipid conjugated [57, 58], while fluticasone propio- disease (approximately 20% of patients) [13]. In line with
nate cannot [58]. this, it was suggested that at least 25% of COPD patients
Taken together, consideration of these various phar- given ICS therapies could be considered as overtreated
macokinetic properties might help to optimize the ICS [61]. Indeed, a recent study analyzing 10,711 COPD pa-
therapy for COPD patients (fig. 2). Indeed, efficient sys- tients found an inappropriate ICS use rate of 18% and
temic clearance, a high receptor affinity, a low oral bio- reported that this lack of adherence to the recommended
availability, a high lung deposition, a long pulmonary res- criteria for ICS therapy led to a lower self-reported patient
idence time, a long half-life, and lipid modification might health status and higher healthcare costs [62]. Thus, cur-
contribute to the enhanced efficacy and safety of an ICS. rent clinical practices regarding the use of ICS therapy for
115.127.13.234 - 3/20/2017 3:28:34 AM

72 Respiration 2015;89:66–75 Herth/Bramlage/Müller-Wieland

DOI: 10.1159/000368371
Downloaded by:
the management of COPD may be contributing to un- and incorporation of novel treatment strategies could ef-
necessary increases in ICS-related adverse event rates. fectively reduce potential ICS side effects, such as diabe-
Therefore, efforts may be required to raise awareness tes, in COPD patients.
among healthcare professionals regarding recommenda-
tions for ICS use when treating COPD patients. Notably,
such modifications in the management of COPD patients Conclusion
have the potential to reduce not only costs related to the
administration of ICS therapies but also those associated Although recent studies have suggested that high-dose
with subsequent requirements that might arise for oral ICS therapy might contribute to the development of type-
antidiabetic medications [24]. 2 diabetes in COPD patients, this idea remains controver-
In this regard, adequate analysis of risk-to-benefit ra- sial and requires further investigation. Nevertheless, there
tios and increased selectivity of patients receiving ICS is reason to consider the clinical implications associated
therapy has been encouraged [63]. However, future stud- with high-dose ICS therapy in COPD patients. In fact, in
ies will be required to firmly establish which patient char- order to reduce the potential risk of ICS-induced diabe-
acteristics might be associated with an increased risk for tes, the efficacy and pharmacokinetics of these drugs at
specific ICS-related side effects. Although it remains un- low or moderate doses may need to be more thoroughly
clear whether unique factors contribute to the suscepti- evaluated. Also, ICS overuse represents a critical issue
bility for an increased risk of diabetes incidence and pro- that must be addressed. In addition, as we learn more re-
gression upon ICS therapy, current evidence may support garding the adverse effects associated with ICS therapy,
avoidance of the routine use of high-dose corticosteroids adequate patient selection and monitoring will be neces-
in individuals with COPD and comorbid diabetes. In- sary to improve the safety and efficacy of these treat-
deed, it has already been suggested that high-dose corti- ments. In this regard, current evidence may suggest that
costeroids be used in the context of close blood glucose care should be taken when administering high-dose ICS
monitoring in these patients [33]. However, a further un- therapies to COPD patients with comorbid diabetes. Tak-
derstanding of the mechanisms that contribute to ICS- en together, future research into the potential role of ICS
related effects on glycemic control could allow for in- in the development and progression of diabetes, as well as
formed selection and improved monitoring of at-risk pa- improved therapeutic regimens to reduce side effects (i.e.
tients, ultimately improving the safety and efficacy of ICS optimal dosing), can lead to improved management of
treatment of COPD. COPD patients.
Moreover, the addition of certain add-on therapies
may prove to be useful for diminishing hyperglycemia
in ICS-treated patients. For example, the phosphodies- Financial Disclosure and Conflicts of Interest
terase-4 inhibitor roflumilast is an anti-inflammatory
All authors are consultants to Novartis and have received re-
medication that appears to enhance lung function in
search funding.
COPD patients. Interestingly, roflumilast treatment was
also found to improve fasting blood glucose and HbA1c
levels in subjects with comorbid type-2 diabetes [64].
Therefore, this drug could be beneficial with regard to References 1 World Health Organization: Chronic ob-
glycemic control in ICS-treated COPD patients. Never- structive pulmonary disease (COPD): burden
of COPD. http://www.who.int/respiratory/
theless, clinical studies are only now underway to assess copd/burden/en/ (accessed March 13, 2014).
combination therapies involving roflumilast and ICS 2 Mannino DM, Thorn D, Swensen A, Holguin
for the management of COPD [65]. Thus, emerging F: Prevalence and outcomes of diabetes, hy-
pertension and cardiovascular disease in
therapeutic strategies could also be pivotal for reducing COPD. Eur Respir J 2008;32:962–969.
the risk of diabetes onset and progression in COPD pa- 3 Jones PW, Nadeau G, Small M, Adamek L:
tients. Characteristics of a COPD population catego-
rised using the gold framework by health sta-
Taken together, the overuse of ICS therapy regimens, tus and exacerbations. Respir Med 2014; 108:
the lack of patient selectivity, and inadequate treatment 129–135.
options may all contribute to ICS-related side effects. 4 Cazzola M, Bettoncelli G, Sessa E, Cricelli C,
Biscione G: Prevalence of comorbidities in
Therefore, improved clinical management based on cur- patients with chronic obstructive pulmonary
rent therapeutic options as well as the future development disease. Respiration 2010;80:112–119.
115.127.13.234 - 3/20/2017 3:28:34 AM

ICS and Diabetes in COPD Respiration 2015;89:66–75 73

DOI: 10.1159/000368371
Downloaded by:
 5 Feary JR, Rodrigues LC, Smith CJ, Hubbard salmeterol/fluticasone propionate or tiotro- coids on exacerbations of chronic obstructive
RB, Gibson JE: Prevalence of major comor- pium bromide. Am J Respir Crit Care Med pulmonary disease. Department of Veterans
bidities in subjects with COPD and incidence 2008;177:19–26. Affairs Cooperative Study Group. N Engl J
of myocardial infarction and stroke: a com- 19 Kew KM, Seniukovich A: Inhaled steroids Med 1999;340:1941–1947.
prehensive analysis using data from primary and risk of pneumonia for chronic obstruc- 32 Ferris HA, Kahn CR: New mechanisms of glu-
care. Thorax 2010;65:956–962. tive pulmonary disease. Cochrane Database cocorticoid-induced insulin resistance: make
6 Lee CT, Mao IC, Lin CH, Lin SH, Hsieh MC: Syst Rev 2014;3:CD010115. no bones about it. J Clin Invest 2012; 122:
Chronic obstructive pulmonary disease: a risk 20 Loke YK, Cavallazzi R, Singh S: Risk of frac- 3854–3857.
factor for type 2 diabetes – a nationwide pop- tures with inhaled corticosteroids in COPD: 33 Caughey GE, Preiss AK, Vitry AI, Gilbert AL,
ulation-based study. Eur J Clin Invest 2013; systematic review and meta-analysis of ran- Roughead EE: Comorbid diabetes and COPD:
43:1113–1119. domised controlled trials and observational impact of corticosteroid use on diabetes com-
7 Rana JS, Mittleman MA, Sheikh J, Hu FB, Man- studies. Thorax 2011;66:699–708. plications. Diabetes Care 2013;36:3009–3014.
son JE, Colditz GA, Speizer FE, Barr RG, Ca- 21 Weatherall M, Clay J, James K, Perrin K, 34 Kardos P, Wencker M, Glaab T, Vogelmeier
margo CA Jr: Chronic obstructive pulmonary Shirtcliffe P, Beasley R: Dose-response rela- C: Impact of salmeterol/fluticasone propio-
disease, asthma, and risk of type 2 diabetes in tionship of inhaled corticosteroids and cata- nate versus salmeterol on exacerbations in se-
women. Diabetes Care 2004;27:2478–2484. racts: a systematic review and meta-analysis. vere chronic obstructive pulmonary disease.
8 Sode BF, Dahl M, Nordestgaard BG: Myocar- Respirology 2009;14:983–990. Am J Respir Crit Care Med 2007; 175: 144–
dial infarction and other co-morbidities in 22 Dong YH, Chang CH, Wu FL, Shen LJ, Cal- 149.
patients with chronic obstructive pulmonary verley PM, Lofdahl CG, Lai MS, Mahler DA: 35 Lung Health Study Research G: Effect of in-
disease: a Danish nationwide study of 7.4 mil- Use of inhaled corticosteroids in patients with haled triamcinolone on the decline in pulmo-
lion individuals. Eur Heart J 2011; 32: 2365– chronic obstructive pulmonary disease and nary function in chronic obstructive pulmo-
2375. the risk of tuberculosis and influenza: a sys- nary disease. N Engl J Med 2000; 343: 1902–
9 Pelaia G, Vatrella A, Cuda G, Maselli R, Mar- tematic review and meta-analysis of random- 1909.
sico SA: Molecular mechanisms of corticoste- ized controlled trials. Chest 2014; 145: 1286– 36 Blackburn D, Hux J, Mamdani M: Quantifica-
roid actions in chronic inflammatory airway 1297. tion of the risk of corticosteroid-induced dia-
diseases. Life Sci 2003;72:1549–1561. 23 Slatore CG, Bryson CL, Au DH: The associa- betes mellitus among the elderly. J Gen Intern
10 Suissa S, Barnes PJ: Inhaled corticosteroids in tion of inhaled corticosteroid use with serum Med 2002;17:717–720.
COPD: the case against. Eur Respir J 2009;34: glucose concentration in a large cohort. Am J 37 Faul JL, Wilson SR, Chu JW, Canfield J, Kus-
13–16. Med 2009;122:472–478. chner WG: The effect of an inhaled cortico-
11 Gartlehner G, Hansen RA, Carson SS, Lohr 24 Suissa S, Kezouh A, Ernst P: Inhaled cortico- steroid on glucose control in type 2 diabetes.
KN: Efficacy and safety of inhaled corticoste- steroids and the risks of diabetes onset and Clinical Med Res 2009;7:14–20.
roids in patients with COPD: a systematic re- progression. Am J Med 2010;123:1001–1006. 38 Ebden P, McNally P, Samanta A, Fancourt GJ:
view and meta-analysis of health outcomes. 25 Dendukuri N, Blais L, LeLorier J: Inhaled cor- The effects of high dose inhaled beclometha-
Ann Fam Med 2006;4:253–262. ticosteroids and the risk of diabetes among the sone dipropionate on glucose and lipid pro-
12 Han MK, Martinez FJ: Pharmacotherapeutic elderly. Br J Clin Pharmacol 2002;54:59–64. files in normal and diet controlled diabetic
approaches to preventing acute exacerbations 26 O’Byrne PM, Rennard S, Gerstein H, Radner subjects. Respir Med 1989;83:289–291.
of chronic obstructive pulmonary disease. F, Peterson S, Lindberg B, Carlsson LG, Sin 39 Izquierdo Alonso JL, Rodriguez Glez-Moro
Proc Am Thorac Soc 2011;8:356–362. DD: Risk of new onset diabetes mellitus in pa- JM: The excessive use of inhaled corticoste-
13 Rabe KF, Hurd S, Anzueto A, Barnes PJ, Buist tients with asthma or COPD taking inhaled roids in chronic obstructive pulmonary dis-
SA, Calverley P, Fukuchi Y, Jenkins C, Rodri- corticosteroids. Respir Med 2012; 106: 1487– ease. Arch Bronconeumol 2012;48:207–212.
guez-Roisin R, van Weel C, Zielinski J; Global 1493. 40 Anzueto A, Ferguson GT, Feldman G, Chin-
Initiative for Chronic Obstructive Lung Dis- 27 Reynolds RM, Labad J, Sears AV, Williamson sky K, Seibert A, Emmett A, Knobil K, O’Dell
ease: Global strategy for the diagnosis, man- RM, Strachan MW, Deary IJ, Lowe GD, Price D, Kalberg C, Crater G: Effect of fluticasone
agement, and prevention of chronic obstruc- JF, Walker BR; Edinburgh Type 2 Diabetes propionate/salmeterol (250/50) on COPD ex-
tive pulmonary disease: GOLD executive Study Investigators: Glucocorticoid treat- acerbations and impact on patient outcomes.
summary. Am J Respir Crit Care Med 2007; ment and impaired mood, memory and me- COPD 2009;6:320–329.
176:532–555. tabolism in people with diabetes: the Edin- 41 Calverley P, Pauwels R, Vestbo J, Jones P,
14 Groenewegen KH, Schols AM, Wouters EF: burgh Type 2 Diabetes Study. Eur J Endocri- Pride N, Gulsvik A, Anderson J, Maden C;
Mortality and mortality-related factors after nol 2012;166:861–868. Trial of Inhaled Steroids and Long-Acting
hospitalization for acute exacerbation of 28 Nathan DM, Kuenen J, Borg R, Zheng H, Beta2 Agonists Study Group: Combined sal-
COPD. Chest 2003;124:459–467. Schoenfeld D, Heine RJ; Group A1c-Derived meterol and fluticasone in the treatment of
15 Rabe KF, Wedzicha JA: Controversies in Average Glucose Study Group: Translating chronic obstructive pulmonary disease: a ran-
treatment of chronic obstructive pulmonary the A1C assay into estimated average glucose domised controlled trial. Lancet 2003; 361:
disease. Lancet 2011;378:1038–1047. values. Diabetes Care 2008;31:1473–1478. 449–456.
16 Postma DS, Calverley P: Inhaled corticoste- 29 Canis R, Demirkok SS, Osar Z, Balci H, Can 42 Ferguson GT, Anzueto A, Fei R, Emmett A,
roids in COPD: a case in favour. Eur Respir J G: Effects of inhaled budesonide on insulin Knobil K, Kalberg C: Effect of fluticasone pro-
2009;34:10–12. sensitivity in nondiabetic patients with asth- pionate/salmeterol (250/50 microg) or salme-
17 Calverley PM, Anderson JA, Celli B, Ferguson ma and chronic obstructive pulmonary dis- terol (50 microg) on COPD exacerbations.
GT, Jenkins C, Jones PW, Yates JC, Vestbo J, ease. Adv Ther 2007;24:560–570. Respir Med 2008;102:1099–1108.
investigators T: Salmeterol and fluticasone 30 Burt MG, Roberts GW, Aguilar-Loza NR, 43 Rennard SI, Tashkin DP, McElhattan J, Gold-
propionate and survival in chronic obstruc- Frith P, Stranks SN: Continuous monitoring man M, Ramachandran S, Martin UJ, Silkoff
tive pulmonary disease. N Engl J Med 2007; of circadian glycemic patterns in patients re- PE: Efficacy and tolerability of budesonide/
356:775–789. ceiving prednisolone for COPD. J Clin Endo- formoterol in one hydrofluoroalkane pres-
18 Wedzicha JA, Calverley PM, Seemungal TA, crinol Metab 2011;96:1789–1796. surized metered-dose inhaler in patients with
Hagan G, Ansari Z, Stockley RA; INSPIRE In- 31 Niewoehner DE, Erbland ML, Deupree RH, chronic obstructive pulmonary disease: re-
vestigators: The prevention of chronic ob- Collins D, Gross NJ, Light RW, Anderson P, sults from a 1-year randomized controlled
structive pulmonary disease exacerbations by Morgan NA: Effect of systemic glucocorti- clinical trial. Drugs 2009;69:549–565.
115.127.13.234 - 3/20/2017 3:28:34 AM

74 Respiration 2015;89:66–75 Herth/Bramlage/Müller-Wieland

DOI: 10.1159/000368371
Downloaded by:
44 Sharafkhaneh A, Southard JG, Goldman M, 54 Dransfield MT, Bourbeau J, Jones PW, Hana- 62 de Miguel-Diez J, Carrasco-Garrido P, Rejas-
Uryniak T, Martin UJ: Effect of budesonide/ nia NA, Mahler DA, Vestbo J, Wachtel A, Gutierrez J, Martin-Centeno A, Gobartt-
formoterol pMDI on COPD exacerbations: a Martinez FJ, Barnhart F, Sanford L, Lettis S, Vazquez E, Hernandez-Barrera V, Gil de MA,
double-blind, randomized study. Respir Med Crim C, Calverley PM: Once-daily inhaled Jimenez-Garcia R: Inappropriate overuse of
2012;106:257–268. fluticasone furoate and vilanterol versus inhaled corticosteroids for COPD patients:
45 Derendorf H, Hochhaus G, Meibohm B, vilanterol only for prevention of exacerba- impact on health costs and health status. Lung
Mollmann H, Barth J: Pharmacokinetics and tions of COPD: two replicate double-blind, 2011;189:199–206.
pharmacodynamics of inhaled corticoste- parallel-group, randomised controlled trials. 63 Anton E: How and when to use inhaled cor-
roids. J Allergy Clinmmunol 1998; 101:S440– Lancet Respir Med 2013;1:210–223. ticosteroids in chronic obstructive pulmo-
S446. 55 Martinez FJ, Boscia J, Feldman G, Scott-Wil- nary disease? Exp Rev Respir Med 2013; 7:
46 Derendorf H, Nave R, Drollmann A, Cerasoli son C, Kilbride S, Fabbri L, Crim C, Calverley 25–32.
F, Wurst W: Relevance of pharmacokinetics PM: Fluticasone furoate/vilanterol (100/25; 64 Field SK: Roflumilast, a novel phosphodies-
and pharmacodynamics of inhaled cortico- 200/25 mug) improves lung function in terase 4 inhibitor, for COPD patients with a
steroids to asthma. Eur Respir J 2006; 28: COPD: a randomised trial. Respir Med 2013; history of exacerbations. Clin Med Insights
1042–1050. 107:550–559. Circ Respir Pulm Med 2011;5:57–70.
47 Derendorf H: Pharmacokinetic and pharma- 56 Mutch E, Nave R, McCracken N, Zech K, Wil- 65 Calverley PM, Martinez FJ, Fabbri LM, Goeh-
codynamic properties of inhaled corticoste- liams FM: The role of esterases in the metabo- ring UM, Rabe KF: Does roflumilast decrease
roids in relation to efficacy and safety. Respir lism of ciclesonide to desisobutyryl-cicle- exacerbations in severe COPD patients not
Med 1997;91(suppl A):22–28. sonide in human tissue. Biochem Pharmacol controlled by inhaled combination therapy?
48 Derendorf H, Daley-Yates PT, Pierre LN, 2007;73:1657–1664. The react study protocol. Int J Chron Ob-
Efthimiou J: Systemic bioavailability of in- 57 Miller-Larsson A, Jansson P, Runstrom A, struct Pulmon Dis 2012;7:375–382.
haled steroids: the importance of appropriate Brattsand R: Prolonged airway activity and 66 Szafranski W, Cukier A, Ramirez A, Menga G,
and comparable methodology. Eur Respir J improved selectivity of budesonide possibly Sansores R, Nahabedian S, Peterson S, Olsson
2001;17:157–158. due to esterification. Am J Respir Crit Care H: Efficacy and safety of budesonide/for-
49 Harrison TW, Tattersfield AE: Plasma con- Med 2000;162:1455–1461. moterol in the management of chronic ob-
centrations of fluticasone propionate and 58 Miller-Larsson A, Mattsson H, Hjertberg E, structive pulmonary disease. Eur Respir J
budesonide following inhalation from dry Dahlback M, Tunek A, Brattsand R: Revers- 2003;21:74–81.
powder inhalers by healthy and asthmatic ible fatty acid conjugation of budesonide: 67 Calverley PM, Boonsawat W, Cseke Z, Zhong
subjects. Thorax 2003;58:258–260. novel mechanism for prolonged retention of N, Peterson S, Olsson H: Maintenance thera-
50 Whelan GJ, Blumer JL, Martin RJ, Szefler SJ; topically applied steroid in airway tissue. py with budesonide and formoterol in chron-
Asthma Clinical Research Network; Pediatric Drug Metab Dispos 1998;26:623–630. ic obstructive pulmonary disease. Eur Respir
Pharmacology Research Unit Network: Fluti- 59 Minto C, Li B, Tattam B, Brown K, Seale JP, J 2003;22:912–919.
casone propionate plasma concentration and Donnelly R: Pharmacokinetics of epimeric 68 Tashkin DP, Rennard SI, Martin P, Ramach-
systemic effect: effect of delivery device and budesonide and fluticasone propionate after andran S, Martin UJ, Silkoff PE, Goldman M:
duration of administration. J Allergy Clin Im- repeat dose inhalation – intersubject variabil- Efficacy and safety of budesonide and for-
munol 2005;116:525–530. ity in systemic absorption from the lung. Br J moterol in one pressurized metered-dose in-
51 Ernst P, Suissa S: Systemic effects of inhaled Clin Pharmacol 2000;50:116–124. haler in patients with moderate to very severe
corticosteroids. Curr Opin Pulm Med 2012; 60 Singh SD, Whale C, Houghton N, Daley- chronic obstructive pulmonary disease: re-
18:85–89. Yates P, Kirby SM, Woodcock AA: Pharma- sults of a 6-month randomized clinical trial.
52 Renton KW: Cytochrome P450 regulation cokinetics and systemic effects of inhaled Drugs 2008;68:1975–2000.
and drug biotransformation during inflam- fluticasone propionate in chronic obstructive 69 Fukuchi Y, Samoro R, Fassakhov R, Tanigu-
mation and infection. Curr Drug Metab 2004; pulmonary disease. Br J Clin Pharmacol 2003; chi H, Ekelund J, Carlsson LG, Ichinose M:
5:235–243. 55:375–381. Budesonide/formoterol via Turbuhaler®
53 Allen A, Bareille PJ, Rousell VM: Fluticasone 61 White P, Thornton H, Pinnock H, Georgop- versus formoterol via Turbuhaler® in pa-
furoate, a novel inhaled corticosteroid, demon- oulou S, Booth HP: Overtreatment of COPD tients with moderate to severe chronic ob-
strates prolonged lung absorption kinetics in with inhaled corticosteroids – implications structive pulmonary disease: phase III multi-
man compared with inhaled fluticasone pro- for safety and costs: cross-sectional observa- national study results. Respirology 2013; 18:
pionate. Clin Pharmacokinet 2013;52:37–42. tional study. PLoS One 2013;8:e75221. 866–873.

115.127.13.234 - 3/20/2017 3:28:34 AM

ICS and Diabetes in COPD Respiration 2015;89:66–75 75

DOI: 10.1159/000368371
Downloaded by: