0
Original Research: COPD |

Use of Inhaled Corticosteroids in Patients With COPD and the Risk of TB and InfluenzaTB/Influenza and Inhaled Corticosteroids in COPD: A Systematic Review and Meta-analysis of Randomized Controlled Trials FREE TO VIEW

Yaa-Hui Dong, PhD; Chia-Hsuin Chang, MD, ScD; Fe-Lin Lin Wu, PhD; Li-Jiuan Shen, PhD; Peter M. A. Calverley, MD; Claes-Göran Löfdahl, MD, FCCP; Mei-Shu Lai, MD, PhD; Donald A. Mahler, MD, FCCP
Author and Funding Information

From the National Taiwan University Health Data Research Center (Dr Dong), Taipei, Taiwan; Graduate Institute of Clinical Pharmacy (Drs Dong, Wu, and Shen) and Department of Pharmacy (Drs Wu and Shen), College of Medicine, National Taiwan University, Taipei, Taiwan; Center of Comparative Effectiveness Research, National Center of Excellence for Clinical Trial and Research (Drs Dong and Lai), Department of Internal Medicine (Dr Chang), and Department of Pharmacy (Drs Wu and Shen), National Taiwan University Hospital, Taipei, Taiwan; Graduate Institute of Epidemiology and Preventive Medicine (Drs Chang and Lai), College of Public Health, National Taiwan University, Taipei, Taiwan; Clinical Science Center (Dr Calverley), University Hospital Aintree, Liverpool, England; Department of Respiratory Medicine and Allergology (Dr Löfdahl), Lund University Hospital, Lund, Sweden; Section of Pulmonary and Critical Care Medicine (Dr Mahler), Geisel School of Medicine at Dartmouth, Hanover, NH; and Dartmouth-Hitchcock Medical Center (Dr Mahler), Lebanon, NH.

Correspondence to: Mei-Shu Lai, MD, PhD, Graduate Institute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, 17 Xuzhou Rd, Taipei 10055, Taiwan; e-mail: mslai@ntu.edu.tw


Part of this article was presented in abstract form at the 29th International Conference on Pharmacoepidemiology & Therapeutic Risk Management, August 25-28, 2013, Montréal, QC, Canada.

Drs Dong and Chang contributed equally to the manuscript.

References 8, 9, 21, 26, 27, 30, 33-36, 38-40, 42, 44, 46.

†References 8, 9, 20, 21, 26, 27, 29-31, 33-35, 38-41, 44, 46.

‡References 8, 9, 20, 21, 26, 27, 29, 33-35, 38-40, 43-45.

Funding/Support: This study was supported in part by the Taiwan Department of Health [Grant DOH101-TD-B-111-01].

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details.


Chest. 2014;145(6):1286-1297. doi:10.1378/chest.13-2137
Text Size: A A A
Published online

Background:  The use of inhaled corticosteroids (ICSs) is associated with an increased risk of pneumonia in patients with COPD. However, the risks of other respiratory infections, such as TB and influenza, remain unclear.

Methods:  Through a comprehensive literature search of MEDLINE, EMBASE, CINAHL, Cochrane Library, and ClinicalTrials.gov from inception to July 2013, we identified randomized controlled trials of ICS therapy lasting at least 6 months. We conducted meta-analyses by the Peto, Mantel-Haenszel, and Bayesian approaches to generate summary estimates comparing ICS with non-ICS treatment on the risk of TB and influenza.

Results:  Twenty-five trials (22,898 subjects) for TB and 26 trials (23,616 subjects) for influenza were included. Compared with non-ICS treatment, ICS treatment was associated with a significantly higher risk of TB (Peto OR, 2.29; 95% CI, 1.04-5.03) but not influenza (Peto OR, 1.24; 95% CI, 0.94-1.63). Results were similar with each meta-analytic approach. Furthermore, the number needed to harm to cause one additional TB event was lower for patients with COPD treated with ICSs in endemic areas than for those in nonendemic areas (909 vs 1,667, respectively).

Conclusions:  This study raises safety concerns about the risk of TB and influenza associated with ICS use in patients with COPD, which deserve further investigation.

Figures in this Article

COPD carries a substantial disease burden and has a tremendous impact on mortality and morbidity worldwide.1 Patients with COPD usually are older, have smoking histories, and exhibit considerable impairment in innate respiratory mechanisms that predispose this population to a higher risk of respiratory infection.2

GOLD (Global Initiative for Chronic Obstructive Lung Disease) guidelines recommend inhaled corticosteroids (ICSs) in combination with long-acting bronchodilators for patients with severe airflow limitation (FEV1 < 50% predicted) and with repeated exacerbations.2 Such patients are estimated to make up < 10% of all patients with COPD.3 However, survey data indicate that ICS therapies are increasingly prescribed for up to 70% of patients with COPD.4 Because of widespread ICS use, the biologic mechanism of immunomodulation, and the partially systematic absorption, it is important to consider the detrimental effects of ICS on host immunity, which may precipitate the onset of infection.4,5

Although an increased risk of pneumonia associated with ICS has been reported in observational studies,6,7 large randomized controlled trials,8,9 and meta-analyses,10,11 few studies have addressed the safety of ICS treatment with respect to other respiratory infections, including TB and influenza. Several observational studies have indicated a positive association between ICS and TB,1216 but these studies were susceptible to confounding by indication, unmeasured confounding, or selection bias. For example, COPD is probably a risk factor for TB.17 Furthermore, COPD severity and cigarette smoking could confound the relation between ICS use and TB.18,19 Absence of information on these variables has prevented further adjustment in research analyzing health-care claims databases.12,14,15 Several hospital-based studies have extracted detailed patient characteristics from medical records; however, these studies recruited only a limited number of patients from a single medical center, and, thus, the results cannot be generalized to the entire COPD population.13,16 To our knowledge, no systematic evaluation has examined the relation between influenza and ICS use. The large Toward a Revolution in COPD Health (TORCH) trial suggested that ICS use increases the likelihood of influenza,8 whereas two other trials indicated that ICS use is not associated with an increased risk.20,21 Because of the potential drawbacks in observational studies and the inconclusive findings reported in individual trials, we aimed to conduct a systematic review and meta-analysis of randomized controlled trials to examine the risk of TB and influenza associated with ICS use in patients with COPD.

Data Sources and Searches

We searched the databases of MEDLINE, EMBASE, CINAHL, Cochrane Library, and ClinicalTrials.gov from inception to July 2013. Full-text terms and specific thesaurus terms (eg, Medical Subject Headings for MEDLINE and EMTREE for EMBASE), including “COPD” and “ICS,” were used for the search strategy (see e-Appendix 1 for details). We examined the bibliographies of relevant articles for eligible trials. To identify complete trial information and unpublished trials, we searched the clinical study registers of drug manufacturers (GlaxoSmithKline plc [www.gsk-clinicalstudyregister.com] and AstraZeneca [www.astrazenecaclinicaltrials.com]) and the US Food and Drug Administration website (www.fda.org).

Study Selection

The study inclusion criteria were as follows: (1) randomized, double-blind, active, or placebo-controlled trials; (2) studies of patients with COPD of any severity; (3) studies involving patients who received ICS treatment as an intervention vs non-ICS treatment as a control in which ICS treatment included ICS alone or ICS in combination with long-acting β2-agonists (LABAs) and non-ICS treatment included tiotropium, LABA, or placebo; (4) trials providing data on TB and influenza (including zero-event trials); and (5) trials lasting ≥ 6 months. We excluded trials that included patients with asthma; did not involve predefined intervention or control treatments; or were published only in protocols, abstracts, or non-English languages.

Outcome Measures

The outcomes of interest were TB and influenza. For eligible trials, we retrieved event numbers for TB (including pulmonary TB) and influenza, regardless of severity, from the listings of adverse events reported in the original articles, ClinicalTrials.gov, and the clinical study registers of the drug manufacturers. If the outcomes were not available from these resources, we contacted the authors or manufacturers for additional information.

Data Extraction and Assessment of Risk of Bias

Two investigators (Y.-H. D., C.-H. C.) independently assessed identified references and extracted relevant characteristics and outcomes from eligible trials. To assess the risk of bias of individual trials, we applied the Cochrane risk of bias tool.22 We also documented how adverse events were monitored. Any discrepancy was resolved by discussion and consensus (e-Appendix 1).

Statistical Analysis

We used the intention-to-treat analytic strategy. For the comparison of ICS treatment vs non-ICS treatment, we estimated the risk of TB and influenza with the Peto OR and 95% CI. The Peto OR does not require a continuity correction and provides the best CI coverage when events are rare.23 To account for the potential imbalance of sample size between treatment groups within trials and interpret the results more intuitively, we also computed the Mantel-Haenszel OR and risk ratio with various continuity correction factors (0.001, 0.01, and 0.1) for trials with zero events in ICS treatment or non-ICS treatment groups.24 For trials with multiple treatment groups, we obtained an overall estimate by collapsing data on ICS and LABA-ICS into an intervention group and data on tiotropium, LABA, and placebo into a control group. We used both fixed-effects and random-effects models. Statistical heterogeneity was evaluated using the I2 statistic, with a value ≥ 50% indicating a substantial level of heterogeneity.

Given the inherent limitations that relative statistics are not easily defined for trials without any event, we calculated the pooled Mantel-Haenszel risk difference.23 Furthermore, we applied the Bayesian and Markov chain Monte Carlo methods. Compared with the pooled Peto and Mantel-Haenszel relative statistics, the summary Bayesian OR and 95% credible interval were estimated according to trial data and prior distributions. This approach is more flexible in the analysis of trials with zero events in both treatment groups.24

To address various safety profiles across different ICS treatments,5 we conducted stratified analyses to individually examine the risk of TB and influenza in fluticasone, mometasone, and budesonide trials. We also compared the pooled estimates of the risk of TB in trials conducted in endemic areas (Asia and Africa) with those conducted in nonendemic areas (America, Europe, and Oceania). A meta-regression analysis using the variance-weighted least squares estimator was applied to examine whether the pooled risk difference with ICS varied significantly with area. The pooled risk difference was converted to the number needed to harm (NNH), which is the number of patients who needed to be treated with an ICS rather than with placebo or active comparisons for an additional patient to be harmed by an adverse event of TB.

STATA 9.0 (StataCorp LP) and WinBUGS version 1.4.3 (MRC Biostatistics Unit) were used for the frequentist (Peto and Mantel-Haenszel methods) and Bayesian meta-analyses, respectively. A two-sided α of 0.05 was defined as statistically significant.

Eligible Trials

We identified 32 studies that fulfilled the eligibility criteria8,9,20,21,2553 and contacted 21 authors for outcome ascertainment.27,3034,36,3844,4652 Nine authors (42.9%) provided outcome information; five (23.8%) did not respond, but information was confirmed by the clinical study registers of drug manufacturers. Seven4652 and six trials4752 did not provide clear data on TB and influenza events, respectively, even after we contacted the authors or manufacturers for further information. Therefore, 25 eligible trials for TB8,9,20,21,2545,53 and 26 for influenza8,9,20,21,2546,53 were included in the meta-analysis (Fig 1, e-Appendix 1).

Figure Jump LinkFigure 1. Flow diagram of the literature search process. aNumber of references identified through each database was 316 (MEDLINE), 1,261 (EMBASE), 13 (CINAHL), 680 (Cochrane), and 49 (ClinicalTrials.gov). bReferences were identified through bibliographies of eligible trial and systematic review articles and the clinical study registers of drug manufacturers. cReferences were usually excluded for more than one reason.Grahic Jump Location

For TB, 25 trials enrolled 22,898 subjects (mean age, 63.2 years; male sex, 70.1%; mean smoking history, 45.0 pack-years). Of these trials, 71.4% included patients with severe airflow limitation, and 32.0% enrolled populations in endemic areas. Approximately 60% of trials were ≥ 1 year in study duration, and 68.0% involved a high-dose ICS treatment arm (fluticasone, > 500-1,000 μg/d; mometasone, ≥ 800 μg/d; budesonide, > 800-1,600 μg/d).54 For influenza, 26 trials enrolled 23,616 subjects, with similar characteristics to those included in the analysis for TB events (Table 1, e-Tables 1, 2). The included trials8,9,20,21,2546 enrolled fewer patients from endemic areas and had a shorter study period but involved more patients treated with high-dose ICSs than trials not included in the meta-analysis (e-Table 3).4752

Table Graphic Jump Location
Table 1 —Summary of Trial and Patient Characteristics at Baseline, ICS Treatment,a and Withdrawal Rates and Fractions of Loss to Follow-up

Data are presented as mean (range) or No. (%). Represented are the total number of trials and the number of trials reporting on at least one TB or influenza event or on zero events. ICS = inhaled corticosteroid; LABA = long-acting β2-agonist.

a 

ICS treatment and concomitant ICS use at baseline include ICS alone and LABA-ICS.

b 

Nineteen, 21, 13, and 17 trials for TB and 20, 22, 14, and 18 trials for influenza reported on smoking history, FEV1, concomitant ICS use at baseline, and loss to follow-up.

Of the included trials, five reported at least one TB event,8,9,20,21,29 and 20 reported zero TB events;2528,3045 nine trials reported at least one influenza event,8,9,20,21,2528,46 and 17 reported zero influenza events.2945 Trials reporting at least one TB or influenza event enrolled more male patients than those reporting zero events. Among trials with at least one TB or influenza event, more included patients with severe airflow limitation, enrolled populations from endemic areas, and had a study duration of ≥ 1 year. Furthermore, for trials with at least one TB event, more involved a high-dose ICS treatment arm than trials with zero TB events (Table 1).

Risk of Bias of Included Trials

Of the 26 included trials, 16 addressed adequate sequence generation and allocation concealment (e-Table 4). All the trials reported withdrawal rates, which varied across trials and tended to be lower in the ICS treatment group (28.2%) than in the non-ICS treatment group (33.7%). Eighteen trials described the proportion of patients lost to follow-up, which was similar between ICS treatment (2.4%) and non-ICS treatment (2.7%) groups. Twenty-one trials provided information on COPD exacerbations, of which 16 showed a numerically or significantly lower use of oral corticosteroids in the ICS treatment group than in the non-ICS treatment group (e-Table 5). None of the trials included TB or influenza as a predefined outcome. The risk of bias was generally similar in trials reporting at least one TB or influenza event and in those reporting zero TB or influenza events.

Risk of TB and Influenza Associated With ICS Treatment

For 25 included trials, the crude risk of TB was 0.15% (18 of 12,062 patients) in the ICS treatment group and 0.06% (seven of 10,836 patients) in the non-ICS treatment group. According to the Peto approach, ICS treatment was associated with a significantly increased risk of TB vs non-ICS treatment (Peto OR, 2.29; 95% CI, 1.04-5.03; I2 = 0.4%) (Fig 2, Table 2). Results of the Mantel-Haenszel and Bayesian approaches also demonstrated a significantly increased risk of TB with ICS treatment, although the estimates tended to be less precise for larger continuity correction factors and risk differences and in the random-effects model (Table 2).

Figure Jump LinkFigure 2. Risk of TB associated with ICS treatment (ICS alone and LABA-ICS) compared with non-ICS treatment (TIO, LABA, or PL) using the Peto approach. bOnly trials reporting on at least one TB event were included in the meta-analysis. ICS = inhaled corticosteroid; LABA = long-acting β2-agonist; PL = placebo; TIO = tiotropium dry powder delivered through HandiHaler (Boehringer Ingelheim Pharmaceuticals, Inc).Grahic Jump Location
Table Graphic Jump Location
Table 2 —Risk of TB and Influenza Associated With ICS Treatment Using Various Meta-analysis Pooling Methods and Statistics

Risk was computed using ICS treatment relative to non-ICS treatment. ICS treatment included ICS alone and LABA-ICS; non-ICS treatment included tiotropium, LABA, and placebo. CrI = credible interval; RD = risk difference; RR = risk ratio. See Table 1 legend for expansion of other abbreviations.

a 

A vague prior distribution with 50,000 Markov chain Monte Carlo iterations and a thin parameter of 5 was used. The posterior inference was undertaken after discarding the initial results of 9,999 iterations.

b 

Only trials reporting at least one TB or influenza event were included in the estimation process.

For 26 included trials, the crude risk of influenza was 1.16% (146 of 12,541 patients) in the ICS treatment group and 0.73% (81 of 11,075 patients) in the non-ICS treatment group. According to the Peto approach, ICS treatment showed a marginally but nonsignificantly increased risk of influenza vs non-ICS treatment (Peto OR, 1.24; 95% CI, 0.94-1.63; I2 = 14.3%) (Fig 3, Table 2). Various meta-analysis pooling methods and summary statistics yielded similar findings (Table 2).

Figure Jump LinkFigure 3. Risk of influenza associated with ICS treatment (ICS alone and LABA-ICS) compared with non-ICS treatment (TIO, LABA, or PL) using the Peto approach. bOnly trials reporting on at least one influenza event was included in the meta-analysis. See Figure 2 legend for expansion of abbreviations.Grahic Jump Location

Of the eligible trials, 19, two, and five assessed fluticasone,8,9,25,2740,44,45 mometasone,20,21 and budesonide,26,4143,46 respectively. In fluticasone trials, patients tended to be older with more severe airflow limitation and were treated with high-dose ICSs, but fewer patients were studied for longer periods vs mometasone or budesonide trials (e-Table 6). Based on the Peto approach, fluticasone was observed with an increased risk of TB (Peto OR, 2.50; 95% CI, 1.12-5.79) and influenza (Peto OR, 1.60; 95% CI, 1.05-2.45). Other pooling methods demonstrated consistent findings. For mometasone or budesonide, however, no excess risk of TB and influenza was found (e-Table 7).

Table 3 shows the results of the stratified analysis comparing risk of TB between trials conducted in endemic vs nonendemic areas. The risk difference with ICS was numerically higher in endemic (0.11%) than in nonendemic areas (0.06%), but the difference was not significant (P = .76). Because the crude baseline risk of TB in the non-ICS arm was substantially higher in endemic areas (0.12%), the NNH to cause one TB event in patients with COPD treated with ICSs was lower for those in endemic areas than in those in nonendemic areas (909 vs 1,667, respectively).

Table Graphic Jump Location
Table 3 —Risk of TB Associated With ICS Treatment in Various Geographic Areas

Risk was computed using ICS treatment relative to non-ICS treatment. ICS treatment included ICS alone and LABA-ICS; non-ICS treatment included tiotropium, LABA, and placebo. Represented are the total number of trials and the number of trials conducted in endemic (Asia and Africa) and nonendemic (America, Europe, and Oceania) areas. NNH = number needed to harm. See Table 1 and 2 legends for expansion of other abbreviations.

a 

The pooled statistics were based on the fixed-effects model.

This meta-analysis of data from randomized controlled trials demonstrated that patients with COPD receiving ICSs had an approximately twofold increased risk of TB, particularly in endemic areas such as Asia and Africa. Meanwhile, a marginally, but nonsignificantly increased risk of influenza associated with ICSs was observed. This study highlights the potential harm of ICSs in increasing respiratory infections. These findings should be weighed against benefits and risks of ICS in the management of COPD.55

Use of corticosteroids has profound effects on immunomodulation. Several experiments suggest that corticosteroids inhibit macrophage functions,56 induce the apoptosis of dendritic cells,57 and suppress the activation of T cells58,59 in the airways. Given that the phagocytic activities of macrophages and the T-helper type 1 adaptive response involve dendritic cells and that T cells are crucial defense mechanisms against Mycobacterium tuberculosis,60 these complex effects may attenuate host immune response and predispose patients to TB infection. Over the past decades, evidence has shown an interaction between epithelial receptors and bacterial invasion, such as platelet-activating factor receptor, for Streptococcus pneumoniae.61 Further studies are warranted to clarify whether a relation exists between upregulation of specific receptors and TB risk with corticosteroids.

Administration of substantial doses of systemic corticosteroids for prolonged periods has been recognized, with an increased risk of TB in clinical settings.62 Five observational studies in Canada, Taiwan, and Korea examined the incidence of TB associated with ICS use (Table 4).1216 Four of these studies indicated that ICS predisposed patients with respiratory medications or respiratory diseases to a relative risk ranging from 1.20 to 15.43.12,13,15,16 The methodologic differences in study type, data source, study population, location, follow-up period, and outcome ascertainment may explain the variations in TB incidence and risk estimate among these studies. There are several explanations for the differences between findings from the observational studies and the present meta-analysis. The observational studies may have enrolled more patients with COPD with severe airflow limitation and comorbid diseases. Their average study durations (3.1-8.6 years) were longer than those of the included trials (1.3 years). For TB case ascertainment, randomized controlled trials may be more complete because trial participants were receiving regular follow-up and close monitoring. Moreover, our approach addressed the potential challenge of residual confounding by COPD severity or cigarette smoking. Considering that TB incidence is higher in endemic areas, further research maximizing the benefit-risk ratio for patients with COPD in these countries is warranted.

Table Graphic Jump Location
Table 4 —Observational Studies of ICS Treatment and TB Risk

Risk was computed using ICS treatment relative to non-ICS treatment. HIRA = Health Insurance Review and Assessment Service; HR = hazard ratio; ICD-9 = International Classification of Diseases, Ninth Revision; ICD-10 = International Classification of Diseases, Tenth Revision; LHID = Longitudinal Health Insurance Database; NA = not available. See Table 1 legend for expansion of other abbreviation.

a 

Risk of TB associated with ICS is separately reported for patients with normal chest radiographs and for those with radiologic sequelae of prior pulmonary TB in the original study. We, thus, calculated the summary risk estimate based on the inverse variance method and the random-effects model.

In contrast to the results of the TORCH trial suggesting that ICS is associated with an apparently increased risk of influenza (Peto OR, 7.37; 95% CI, 1.49-36.55),8 the present pooled analysis indicates that the increased risk of influenza was not statistically significant among patients receiving ICS treatment (Peto OR, 1.24; 95% CI, 0.94-1.63). Several possible explanations can be considered. First, unlike the opportunistic TB infection in the lower respiratory tract, influenza is an acute upper respiratory tract illness, and interferon type I innate response is essential to limit virus replication and spread.63 One study indicated that corticosteroids might exert less effect on the innate response in airway epithelium compared with its potent suppression of adaptive immunity.57 This might partially contribute to the present nonsignificant findings of influenza risk with ICS, although further studies are needed to elucidate comprehensive biologic hypotheses. Second, although we included trials lasting ≥ 6 months to mitigate the effect of seasonality on influenza occurrence, confounding due to variable study periods across trials cannot be fully excluded. Moreover, because of nonspecific signs and symptoms of influenza, underdiagnosis or misclassification of influenza with other respiratory illnesses is possible. Despite nonsignificant findings in the present study, it should be acknowledged that increasing outbreaks of influenza globally have caused substantial death and illness. Because patients with respiratory diseases are highly susceptible to influenza infections that may contribute to morbidity and mortality,63 further studies are required to specifically evaluate the safety of ICS in terms of an increased risk of influenza and influenza-related illness.

Data from meta-analyses have indicated that fluticasone poses an increased risk of pneumonia, whereas budesonide does not present an apparent risk.11,64 In line with the potential intraclass differences in safety profiles due to different potency,5,11,64 we suggest that fluticasone (high potency ICS) rather than mometasone or budesonide (weak potency ICSs) treatment leads to an increased risk of TB and influenza. However, this finding needs to be interpreted with caution because the number of mometasone and budesonide trials was limited and the demographic characteristics varied across individual ICSs. Additionally, all trials with low-dose ICS treatment reported zero TB events (six for fluticasone and two for budesonide), which prevented us from examining the dose-response relation or clearly clarifying the effect of ICS dose and potency on TB risk. Further studies of head-to-head comparisons of individual ICSs are needed to explore the hypothesis of intraclass differences while taking both important demographic characteristics and ICS dose into account.

The limitations of this study mainly reflect the challenges of assessing drug safety in clinical trials.65,66 First, the number of trials reporting on TB events was limited, which prevented us from conducting further subgroup analyses or meta-regressions to test possible assumptions such as the effect modification mediated by important demographic characteristics. Additionally, the absence of information on timing of TB occurrence in relation to ICS use precluded us from calculating a risk estimate based on exposure duration. Second, in these clinical trials, a normal chest radiograph was not a consistent study requirement, so occult TB may have been present or reactivated when an ICS was used. In addition, because TB and influenza were not predefined outcomes and no homogeneous definitions of these end points (eg, microbiologic confirmation by cultures, acid-fast statin smears, rapid antigen tests) existed across trials, misclassification of these adverse events is possible. However, because all the included trials were double blind, such misclassification is likely to be nondifferential, and the direction of bias may be toward the null. Third, we observed that use of oral corticosteroids for COPD exacerbations was not balanced between treatment groups and, thus, may have blunted the risk estimates of ICS treatment. Fourth, the withdrawal rate in the included trials was lower in the ICS treatment group than in the non-ICS treatment group, which may raise concerns of overestimating the relative risk due to underestimating the incidence of TB and influenza in the non-ICS group.66 However, proportions of patients lost to follow-up were similar across treatment groups; thus, the unfavorable bias for ICS treatment should be limited. Finally, although we synthesized data from multicountry and multicenter trials to facilitate the application of the findings across countries and ethnicities, all the trials excluded patients with significant diseases, and most excluded patients with specific pulmonary morbidities. This may limit the generalizability of the present results to the frail populations in real practice.

In view of the balance between benefits and risks of ICS treatment in patients with COPD, the results have substantial implications for clinical practice because use of ICS is shown to carry a significantly (twofold) increased risk of TB and a borderline risk of influenza. Until more evidence is available, ICS should only be prescribed as an add-on treatment to long-acting bronchodilators for patients with severe airflow limitation and with repeated exacerbations. TB screening prior to the onset of ICS treatment and regular monitoring thereafter may be indicated in endemic countries. Considering various rates of occurrence of influenza-associated illness across seasons and increasing outbreaks of influenza worldwide in recent years, the potential risk of influenza associated with ICS treatment deserves special attention and further investigation.

Author contributions: Dr Lai had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Dr Dong: contributed to the study conception and design; study selection; assessment of risk of bias; extraction, analysis, and interpretation of the data; drafting of the manuscript; critical revision of the manuscript for important intellectual content; and final approval of the manuscript.

Dr Chang: contributed to the study conception and design, study selection, assessment of risk of bias, extraction and interpretation of the data, drafting of the manuscript, critical revision of the manuscript for important intellectual content, and final approval of the manuscript.

Dr Wu: contributed to the data interpretation, critical revision of the manuscript for important intellectual content, and final approval of the manuscript.

Dr Shen: contributed to the data interpretation, critical revision of the manuscript for important intellectual content, and final approval of the manuscript.

Dr Calverley: contributed to the provision of additional data, data interpretation, critical revision of the manuscript for important intellectual content, and final approval of the manuscript.

Dr Löfdahl: contributed to the provision of additional data and data interpretation, critical revision of the manuscript for important intellectual content, and final approval of the manuscript.

Dr Lai: contributed to the study conception and design, data interpretation, critical revision of the manuscript for important intellectual content, final approval of the manuscript, and obtaining funding.

Dr Mahler: contributed to the provision of additional data anddata interpretation, critical revision of the manuscript for important intellectual content, and final approval of the manuscript.

Financial/nonfinancial disclosures: The authors have reported to CHEST the following conflicts of interest: Dr Calverley has led studies sponsored by several pharmaceutical companies, including GlaxoSmithKline plc, AstraZeneca, Takeda Pharmaceutical Company Limited, and Boehringer Ingelheim GmbH. Dr Mahler has consulted for Boehringer Ingelheim GmbH, Forest Laboratories Inc, GlaxoSmithKline plc, Novartis AG, and Sunovion Pharmaceuticals Inc; has participated in advisory Boards for Forest Laboratories Inc, GlaxoSmithKline plc, Merck Sharp & Dohme Corp, Novartis AG, Pearl Therapeutics Inc, and Sunovion Pharmaceuticals Inc; and has received royalties from Mapi Research Institute and Taylor & Francis Group. The clinical trials office at Dartmouth-Hitchcock Medical Center has received grant support from Boehringer Ingelheim GmbH, GlaxoSmithKline plc, Novartis AG, and Sunovion Pharmaceuticals Inc for which Dr Mahler was principal investigator. Drs Dong, Chang, Wu, Shen, Löfdahl, and Lai have reported that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.

Role of the sponsors: The sponsor did not play any role in the study design, literature search, study selection, collection and analysis of data, interpretation of results, or drafting of the manuscript.

Other contributions: The authors thank Chun-Yao Lu, MPH, for assisting with the database search.

Additional information: The e-Appendix and e-Tables can be found in the “Supplemental Materials” area of the online article.

ICS

inhaled corticosteroid

LABA

long-acting β2-agonist

NNH

number needed to harm

The global burden of disease: 2004 update. WHO website. http://www.who.int/healthinfo/global_burden_disease/GBD_report_2004update_full.pdf. Accessed July 1, 2013.
 
Global strategy for the diagnosis, management and prevention of chronic obstructive pulmonary diseases. Updated 2013. GOLD website. http://www.goldcopd.org/uploads/users/files/GOLD_Report_2013_Feb20.pdf. Accessed July 28, 2013.
 
Lange P, Marott JL, Vestbo J, et al. Prediction of the clinical course of chronic obstructive pulmonary disease, using the new GOLD classification: a study of the general population. Am J Respir Crit Care Med. 2012;186(10):975-981.
 
Suissa S, Barnes PJ. Inhaled corticosteroids in COPD: the case against. Eur Respir J. 2009;34(1):13-16.
 
Barnes NC. The properties of inhaled corticosteroids: similarities and differences. Prim Care Respir J. 2007;16(3):149-154.
 
Ernst P, Gonzalez AV, Brassard P, Suissa S. Inhaled corticosteroid use in chronic obstructive pulmonary disease and the risk of hospitalization for pneumonia. Am J Respir Crit Care Med. 2007;176(2):162-166.
 
Joo MJ, Au DH, Fitzgibbon ML, Lee TA. Inhaled corticosteroids and risk of pneumonia in newly diagnosed COPD. Respir Med. 2010;104(2):246-252.
 
Calverley PMA, Anderson JA, Celli B, et al; TORCH Investigators. Salmeterol and fluticasone propionate and survival in chronic obstructive pulmonary disease. N Engl J Med. 2007;356(8):775-789.
 
Wedzicha JA, Calverley PM, Seemungal TA, Hagan G, Ansari Z, Stockley RA; INSPIRE Investigators. The prevention of chronic obstructive pulmonary disease exacerbations by salmeterol/fluticasone propionate or tiotropium bromide. Am J Respir Crit Care Med. 2008;177(1):19-26.
 
Drummond MB, Dasenbrook EC, Pitz MW, Murphy DJ, Fan E. Inhaled corticosteroids in patients with stable chronic obstructive pulmonary disease: a systematic review and meta-analysis. JAMA. 2008;300(20):2407-2416.
 
Singh S, Amin AV, Loke YK. Long-term use of inhaled corticosteroids and the risk of pneumonia in chronic obstructive pulmonary disease: a meta-analysis. Arch Intern Med. 2009;169(3):219-229.
 
Lee CH, Kim K, Hyun MK, Jang EJ, Lee NR, Yim JJ. Use of inhaled corticosteroids and the risk of tuberculosis. Thorax. 2013;68(12):1105-1113.
 
Kim JH, Park JS, Kim KH, Jeong HC, Kim EK, Lee JH. Inhaled corticosteroid is associated with an increased risk of TB in patients with COPD. Chest. 2013;143(4):1018-1024.
 
Lee CH, Lee MC, Shu CC, et al. Risk factors for pulmonary tuberculosis in patients with chronic obstructive airway disease in Taiwan: a nationwide cohort study. BMC Infect Dis. 2013;13:194.
 
Brassard P, Suissa S, Kezouh A, Ernst P. Inhaled corticosteroids and risk of tuberculosis in patients with respiratory diseases. Am J Respir Crit Care Med. 2011;183(5):675-678.
 
Shu CC, Wu HD, Yu MC, et al; Taiwan Anti-Mycobacteria Investigation (TAMI) Group. Use of high-dose inhaled corticosteroids is associated with pulmonary tuberculosis in patients with chronic obstructive pulmonary disease. Medicine (Baltimore). 2010;89(1):53-61.
 
Inghammar M, Ekbom A, Engström G, et al. COPD and the risk of tuberculosis—a population-based cohort study. PLoS ONE. 2010;5(4):e10138.
 
den Boon S, van Lill SWP, Borgdorff MW, et al. Association between smoking and tuberculosis infection: a population survey in a high tuberculosis incidence area. Thorax. 2005;60(7):555-557.
 
Horne DJ, Campo M, Ortiz JR, et al. Association between smoking and latent tuberculosis in the US population: an analysis of the National Health and Nutrition Examination Survey. PLoS ONE. 2012;7(11):e49050.
 
Doherty DE, Tashkin DP, Kerwin E, et al. Effects of mometasone furoate/formoterol fumarate fixed-dose combination formulation on chronic obstructive pulmonary disease (COPD): results from a 52-week phase III trial in subjects with moderate-to-very severe COPD. Int J Chron Obstruct Pulmon Dis. 2012;7:57-71.
 
Tashkin DP, Doherty DE, Kerwin E, et al. Efficacy and safety of a fixed-dose combination of mometasone furoate and formoterol fumarate in subjects with moderate to very severe COPD: results from a 52-week phase III trial. Int J Chron Obstruct Pulmon Dis. 2012;7:43-55.
 
Higgins JPT, Green S., eds. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0. Updated March 2011. The Cochrane Collaboration website. http://handbook.cochrane.org. Accessed July 1, 2013.
 
Bradburn MJ, Deeks JJ, Berlin JA, Russell Localio A. Much ado about nothing: a comparison of the performance of meta-analytical methods with rare events. Stat Med. 2007;26(1):53-77.
 
Sweeting MJ, Sutton AJ, Lambert PC. What to add to nothing? Use and avoidance of continuity corrections in meta-analysis of sparse data. Stat Med. 2004;23(9):1351-1375.
 
GlaxoSmithKline Clinical Study Register. A multicentre, randomised, double-blind parallel group placebo controlled study assessing the efficacy and safety of inhaled salmeterol/fluticasone 50/500mcg twice daily, inhaled fluticasone 500mcg twice daily and placebo all administered via MDI in the treatment of patients with COPD. SFCT01/SCO30002. GlaxoSmithKline Clinical Study Register website. http://www.gsk-clinicalstudyregister.com/search?study_ids=SCO30002#rs. Accessed July 1, 2013.
 
Sharafkhaneh A, Southard JG, Goldman M, Uryniak T, Martin UJ. Effect of budesonide/formoterol pMDI on COPD exacerbations: a double-blind, randomized study. Respir Med. 2012;106(2):257-268.
 
Ferguson GT, Anzueto A, Fei R, Emmett A, Knobil K, Kalberg C. Effect of fluticasone propionate/salmeterol (250/50 μg) or salmeterol (50 μg) on COPD exacerbations. Respir Med. 2008;102(8):1099-1108.
 
GlaxoSmithKline Clinical Study Register. A multicentre, randomised, double-blind, parallel group, 24 week study to compare the effect of the salmeterol/fluticasone propionate combination product 50/250mcg, with salmeterol 50mcg both delivered twice daily via the DISKUS®/ACCUHALER®inhaler on lung function and dyspnoea in subjects with chronic obstructive pulmonary disease (COPD). SCO100470. GlaxoSmithKline Clinical Study Register website. http://www.gsk-clinicalstudyregister.com/search?study_ids=SCO100470#rs. Accessed July 1, 2013.
 
Zheng JP, Yang L, Wu YM, et al. The efficacy and safety of combination salmeterol (50 μg)/fluticasone propionate (500 μg) inhalation twice daily via Accuhaler in Chinese patients with COPD. Chest. 2007;132(6):1756-1763.
 
Calverley P, Pauwels R, Vestbo J, et al; TRial of Inhaled STeroids ANd long-acting beta2 agonists study group. Combined salmeterol and fluticasone in the treatment of chronic obstructive pulmonary disease: a randomised controlled trial. Lancet. 2003;361(9356):449-456.
 
Hanania NA, Darken P, Horstman D, et al. The efficacy and safety of fluticasone propionate (250 μg)/salmeterol (50 μg) combined in the Diskus inhaler for the treatment of COPD. Chest. 2003;124(3):834-843.
 
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(8):1084-1091.
 
Aaron SD, Vandemheen KL, Fergusson D, et al; Canadian Thoracic Society/Canadian Respiratory Clinical Research Consortium. Tiotropium in combination with placebo, salmeterol, or fluticasone-salmeterol for treatment of chronic obstructive pulmonary disease: a randomized trial. Ann Intern Med. 2007;146(8):545-555.
 
Anzueto A, Ferguson GT, Feldman G, et al. Effect of fluticasone propionate/salmeterol (250/50) on COPD exacerbations and impact on patient outcomes. COPD. 2009;6(5):320-329.
 
Kardos P, Wencker M, Glaab T, Vogelmeier C. Impact of salmeterol/fluticasone propionate versus salmeterol on exacerbations in severe chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2007;175(2):144-149.
 
Wouters EF, Postma DS, Fokkens B, et al; COSMIC (COPD and Seretide: a Multi-Center Intervention and Characterization) Study Group. Withdrawal of fluticasone propionate from combined salmeterol/fluticasone treatment in patients with COPD causes immediate and sustained disease deterioration: a randomised controlled trial. Thorax. 2005;60(6):480-487.
 
GlaxoSmithKline Clinical Study Register. A randomized, double-blind, parallel-group clinical trial evaluating the effect of the fluticasone propionate/salmeterol combination product 250/50mcg bid via DISKUS versus salmeterol 50mcg bid via DISKUS on bone mineral density in subjects with chronic obstructive pulmonary disease (COPD). SCO40041. GlaxoSmithKline Clinical Study Register website. http://www.gsk-clinicalstudyregister.com/search?study_ids=SCO40041#ps. Accessed July 1, 2013.
 
Reid DW, Wen Y, Johns DP, Williams TJ, Ward C, Walters EH. Bronchodilator reversibility, airway eosinophilia and anti-inflammatory effects of inhaled fluticasone in COPD are not related. Respirology. 2008;13(6):799-809.
 
van der Valk P, Monninkhof E, van der Palen J, Zielhuis G, van Herwaarden C. Effect of discontinuation of inhaled corticosteroids in patients with chronic obstructive pulmonary disease: the COPE study. Am J Respir Crit Care Med. 2002;166(10):1358-1363.
 
Burge PS, Calverley PM, Jones PW, Spencer S, Anderson JA, Maslen TK. Randomised, double blind, placebo controlled study of fluticasone propionate in patients with moderate to severe chronic obstructive pulmonary disease: the ISOLDE trial. BMJ. 2000;320(7245):1297-1303.
 
Pauwels RA, Löfdahl CG, Laitinen LA, et al; European Respiratory Society Study on Chronic Obstructive Pulmonary Disease. Long-term treatment with inhaled budesonide in persons with mild chronic obstructive pulmonary disease who continue smoking. N Engl J Med. 1999;340(25):1948-1953.
 
Vestbo J, Sørensen T, Lange P, Brix A, Torre P, Viskum K. Long-term effect of inhaled budesonide in mild and moderate chronic obstructive pulmonary disease: a randomised controlled trial. Lancet. 1999;353(9167):1819-1823.
 
Bourbeau J, Rouleau MY, Boucher S. Randomised controlled trial of inhaled corticosteroids in patients with chronic obstructive pulmonary disease. Thorax. 1998;53(6):477-482.
 
Paggiaro PL, Dahle R, Bakran I, Frith L, Hollingworth K, Efthimiou J; International COPD Study Group. Multicentre randomised placebo-controlled trial of inhaled fluticasone propionate in patients with chronic obstructive pulmonary disease. Lancet. 1998;351(9105):773-780.
 
GlaxoSmithKline Clinical Study Register. A randomized, double-blind, parallel-group, comparative trial of inhaled fluticasone propionate 250mcg bid, 500mcg bid, and placebo bid via the DISKUS in subjects with chronic obstructive pulmonary disease (COPD). FLTA3025. GlaxoSmithKline Clinical Study Register website. http://www.gsk-clinicalstudyregister.com/search?study_ids=FLTA3025#rs. Accessed July 1, 2013.
 
Calverley PM, Kuna P, Monsó E, et al. Beclomethasone/formoterol in the management of COPD: a randomised controlled trial. Respir Med. 2010;104(12):1858-1868.
 
Rennard SI, Tashkin DP, McElhattan J, et al. Efficacy and tolerability of budesonide/formoterol in one hydrofluoroalkane pressurized metered-dose inhaler in patients with chronic obstructive pulmonary disease: results from a 1-year randomized controlled clinical trial. Drugs. 2009;69(5):549-565.
 
Shaker SB, Dirksen A, Ulrik CS, et al. The effect of inhaled corticosteroids on the development of emphysema in smokers assessed by annual computed tomography. COPD. 2009;6(2):104-111.
 
Tashkin DP, Rennard SI, Martin P, et al. Efficacy and safety of budesonide and formoterol in one pressurized metered-dose inhaler in patients with moderate to very severe chronic obstructive pulmonary disease: results of a 6-month randomized clinical trial. Drugs. 2008;68(14):1975-2000.
 
Choudhury AB, Dawson CM, Kilvington HE, et al. Withdrawal of inhaled corticosteroids in people with COPD in primary care: a randomised controlled trial. Respir Res. 2007;8:93.
 
Calverley PM, Boonsawat W, Cseke Z, Zhong N, Peterson S, Olsson H. Maintenance therapy with budesonide and formoterol in chronic obstructive pulmonary disease. Eur Respir J. 2003;22(6):912-919.
 
Szafranski W, Cukier A, Ramirez A, et al. Efficacy and safety of budesonide/formoterol in the management of chronic obstructive pulmonary disease. Eur Respir J. 2003;21(1):74-81.
 
Tashkin DP, Doherty DE, Kerwin E, et al. Efficacy and safety characteristics of mometasone furoate/formoterol fumarate fixed-dose combination in subjects with moderate to very severe COPD: findings from pooled analysis of two randomized, 52-week placebo-controlled trials. Int J Chron Obstruct Pulmon Dis. 2012;7:73-86.
 
Global strategy for asthma management and prevention. Updated 2012. GINA website. http://www.ginasthma.org/local/uploads/files/GINA_Report_March13.pdf. Accessed July 1, 2013.
 
Price D, Yawn B, Brusselle G, Rossi A. Risk-to-benefit ratio of inhaled corticosteroids in patients with COPD. Prim Care Respir J. 2013;22(1):92-100.
 
Ek A, Larsson K, Siljerud S, Palmberg L. Fluticasone and budesonide inhibit cytokine release in human lung epithelial cells and alveolar macrophages. Allergy. 1999;54(7):691-699.
 
Schleimer RP. Glucocorticoids suppress inflammation but spare innate immune responses in airway epithelium. Proc Am Thorac Soc. 2004;1(3):222-230.
 
Hogg JC, Chu FS, Tan WC, et al. Survival after lung volume reduction in chronic obstructive pulmonary disease: insights from small airway pathology. Am J Respir Crit Care Med. 2007;176(5):454-459.
 
Fauci AS, Dale DC, Balow JE. Glucocorticosteroid therapy: mechanisms of action and clinical considerations. Ann Intern Med. 1976;84(3):304-315.
 
Gupta A, Kaul A, Tsolaki AG, Kishore U, Bhakta S. Mycobacterium tuberculosis: immune evasion, latency and reactivation. Immunobiology. 2012;217(3):363-374.
 
Iovino F, Brouwer MC, van de Beek D, Molema G, Bijlsma JJ. Signalling or binding: the role of the platelet-activating factor receptor in invasive pneumococcal disease. Cell Microbiol. 2013;15(6):870-881.
 
American Thoracic Society/Centers for Disease Control and Prevention. Targeted tuberculin testing and treatment of latent tuberculosis infection. Am J Respir Crit Care Med. 2000;161(suppl):S221-S247.
 
Hale BG, Albrecht RA, García-Sastre A. Innate immune evasion strategies of influenza viruses. Future Microbiol. 2010;5(1):23-41.
 
Sin DD, Tashkin D, Zhang X, et al. Budesonide and the risk of pneumonia: a meta-analysis of individual patient data. Lancet. 2009;374(9691):712-719.
 
Singh S, Loke YK. Drug safety assessment in clinical trials: methodological challenges and opportunities. Trials. 2012;13:138.
 
Kesten S, Plautz M, Piquette CA, Habib MP, Niewoehner DE. Premature discontinuation of patients: a potential bias in COPD clinical trials. Eur Respir J. 2007;30(5):898-906.
 

Figures

Figure Jump LinkFigure 1. Flow diagram of the literature search process. aNumber of references identified through each database was 316 (MEDLINE), 1,261 (EMBASE), 13 (CINAHL), 680 (Cochrane), and 49 (ClinicalTrials.gov). bReferences were identified through bibliographies of eligible trial and systematic review articles and the clinical study registers of drug manufacturers. cReferences were usually excluded for more than one reason.Grahic Jump Location
Figure Jump LinkFigure 2. Risk of TB associated with ICS treatment (ICS alone and LABA-ICS) compared with non-ICS treatment (TIO, LABA, or PL) using the Peto approach. bOnly trials reporting on at least one TB event were included in the meta-analysis. ICS = inhaled corticosteroid; LABA = long-acting β2-agonist; PL = placebo; TIO = tiotropium dry powder delivered through HandiHaler (Boehringer Ingelheim Pharmaceuticals, Inc).Grahic Jump Location
Figure Jump LinkFigure 3. Risk of influenza associated with ICS treatment (ICS alone and LABA-ICS) compared with non-ICS treatment (TIO, LABA, or PL) using the Peto approach. bOnly trials reporting on at least one influenza event was included in the meta-analysis. See Figure 2 legend for expansion of abbreviations.Grahic Jump Location

Tables

Table Graphic Jump Location
Table 1 —Summary of Trial and Patient Characteristics at Baseline, ICS Treatment,a and Withdrawal Rates and Fractions of Loss to Follow-up

Data are presented as mean (range) or No. (%). Represented are the total number of trials and the number of trials reporting on at least one TB or influenza event or on zero events. ICS = inhaled corticosteroid; LABA = long-acting β2-agonist.

a 

ICS treatment and concomitant ICS use at baseline include ICS alone and LABA-ICS.

b 

Nineteen, 21, 13, and 17 trials for TB and 20, 22, 14, and 18 trials for influenza reported on smoking history, FEV1, concomitant ICS use at baseline, and loss to follow-up.

Table Graphic Jump Location
Table 2 —Risk of TB and Influenza Associated With ICS Treatment Using Various Meta-analysis Pooling Methods and Statistics

Risk was computed using ICS treatment relative to non-ICS treatment. ICS treatment included ICS alone and LABA-ICS; non-ICS treatment included tiotropium, LABA, and placebo. CrI = credible interval; RD = risk difference; RR = risk ratio. See Table 1 legend for expansion of other abbreviations.

a 

A vague prior distribution with 50,000 Markov chain Monte Carlo iterations and a thin parameter of 5 was used. The posterior inference was undertaken after discarding the initial results of 9,999 iterations.

b 

Only trials reporting at least one TB or influenza event were included in the estimation process.

Table Graphic Jump Location
Table 3 —Risk of TB Associated With ICS Treatment in Various Geographic Areas

Risk was computed using ICS treatment relative to non-ICS treatment. ICS treatment included ICS alone and LABA-ICS; non-ICS treatment included tiotropium, LABA, and placebo. Represented are the total number of trials and the number of trials conducted in endemic (Asia and Africa) and nonendemic (America, Europe, and Oceania) areas. NNH = number needed to harm. See Table 1 and 2 legends for expansion of other abbreviations.

a 

The pooled statistics were based on the fixed-effects model.

Table Graphic Jump Location
Table 4 —Observational Studies of ICS Treatment and TB Risk

Risk was computed using ICS treatment relative to non-ICS treatment. HIRA = Health Insurance Review and Assessment Service; HR = hazard ratio; ICD-9 = International Classification of Diseases, Ninth Revision; ICD-10 = International Classification of Diseases, Tenth Revision; LHID = Longitudinal Health Insurance Database; NA = not available. See Table 1 legend for expansion of other abbreviation.

a 

Risk of TB associated with ICS is separately reported for patients with normal chest radiographs and for those with radiologic sequelae of prior pulmonary TB in the original study. We, thus, calculated the summary risk estimate based on the inverse variance method and the random-effects model.

References

The global burden of disease: 2004 update. WHO website. http://www.who.int/healthinfo/global_burden_disease/GBD_report_2004update_full.pdf. Accessed July 1, 2013.
 
Global strategy for the diagnosis, management and prevention of chronic obstructive pulmonary diseases. Updated 2013. GOLD website. http://www.goldcopd.org/uploads/users/files/GOLD_Report_2013_Feb20.pdf. Accessed July 28, 2013.
 
Lange P, Marott JL, Vestbo J, et al. Prediction of the clinical course of chronic obstructive pulmonary disease, using the new GOLD classification: a study of the general population. Am J Respir Crit Care Med. 2012;186(10):975-981.
 
Suissa S, Barnes PJ. Inhaled corticosteroids in COPD: the case against. Eur Respir J. 2009;34(1):13-16.
 
Barnes NC. The properties of inhaled corticosteroids: similarities and differences. Prim Care Respir J. 2007;16(3):149-154.
 
Ernst P, Gonzalez AV, Brassard P, Suissa S. Inhaled corticosteroid use in chronic obstructive pulmonary disease and the risk of hospitalization for pneumonia. Am J Respir Crit Care Med. 2007;176(2):162-166.
 
Joo MJ, Au DH, Fitzgibbon ML, Lee TA. Inhaled corticosteroids and risk of pneumonia in newly diagnosed COPD. Respir Med. 2010;104(2):246-252.
 
Calverley PMA, Anderson JA, Celli B, et al; TORCH Investigators. Salmeterol and fluticasone propionate and survival in chronic obstructive pulmonary disease. N Engl J Med. 2007;356(8):775-789.
 
Wedzicha JA, Calverley PM, Seemungal TA, Hagan G, Ansari Z, Stockley RA; INSPIRE Investigators. The prevention of chronic obstructive pulmonary disease exacerbations by salmeterol/fluticasone propionate or tiotropium bromide. Am J Respir Crit Care Med. 2008;177(1):19-26.
 
Drummond MB, Dasenbrook EC, Pitz MW, Murphy DJ, Fan E. Inhaled corticosteroids in patients with stable chronic obstructive pulmonary disease: a systematic review and meta-analysis. JAMA. 2008;300(20):2407-2416.
 
Singh S, Amin AV, Loke YK. Long-term use of inhaled corticosteroids and the risk of pneumonia in chronic obstructive pulmonary disease: a meta-analysis. Arch Intern Med. 2009;169(3):219-229.
 
Lee CH, Kim K, Hyun MK, Jang EJ, Lee NR, Yim JJ. Use of inhaled corticosteroids and the risk of tuberculosis. Thorax. 2013;68(12):1105-1113.
 
Kim JH, Park JS, Kim KH, Jeong HC, Kim EK, Lee JH. Inhaled corticosteroid is associated with an increased risk of TB in patients with COPD. Chest. 2013;143(4):1018-1024.
 
Lee CH, Lee MC, Shu CC, et al. Risk factors for pulmonary tuberculosis in patients with chronic obstructive airway disease in Taiwan: a nationwide cohort study. BMC Infect Dis. 2013;13:194.
 
Brassard P, Suissa S, Kezouh A, Ernst P. Inhaled corticosteroids and risk of tuberculosis in patients with respiratory diseases. Am J Respir Crit Care Med. 2011;183(5):675-678.
 
Shu CC, Wu HD, Yu MC, et al; Taiwan Anti-Mycobacteria Investigation (TAMI) Group. Use of high-dose inhaled corticosteroids is associated with pulmonary tuberculosis in patients with chronic obstructive pulmonary disease. Medicine (Baltimore). 2010;89(1):53-61.
 
Inghammar M, Ekbom A, Engström G, et al. COPD and the risk of tuberculosis—a population-based cohort study. PLoS ONE. 2010;5(4):e10138.
 
den Boon S, van Lill SWP, Borgdorff MW, et al. Association between smoking and tuberculosis infection: a population survey in a high tuberculosis incidence area. Thorax. 2005;60(7):555-557.
 
Horne DJ, Campo M, Ortiz JR, et al. Association between smoking and latent tuberculosis in the US population: an analysis of the National Health and Nutrition Examination Survey. PLoS ONE. 2012;7(11):e49050.
 
Doherty DE, Tashkin DP, Kerwin E, et al. Effects of mometasone furoate/formoterol fumarate fixed-dose combination formulation on chronic obstructive pulmonary disease (COPD): results from a 52-week phase III trial in subjects with moderate-to-very severe COPD. Int J Chron Obstruct Pulmon Dis. 2012;7:57-71.
 
Tashkin DP, Doherty DE, Kerwin E, et al. Efficacy and safety of a fixed-dose combination of mometasone furoate and formoterol fumarate in subjects with moderate to very severe COPD: results from a 52-week phase III trial. Int J Chron Obstruct Pulmon Dis. 2012;7:43-55.
 
Higgins JPT, Green S., eds. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0. Updated March 2011. The Cochrane Collaboration website. http://handbook.cochrane.org. Accessed July 1, 2013.
 
Bradburn MJ, Deeks JJ, Berlin JA, Russell Localio A. Much ado about nothing: a comparison of the performance of meta-analytical methods with rare events. Stat Med. 2007;26(1):53-77.
 
Sweeting MJ, Sutton AJ, Lambert PC. What to add to nothing? Use and avoidance of continuity corrections in meta-analysis of sparse data. Stat Med. 2004;23(9):1351-1375.
 
GlaxoSmithKline Clinical Study Register. A multicentre, randomised, double-blind parallel group placebo controlled study assessing the efficacy and safety of inhaled salmeterol/fluticasone 50/500mcg twice daily, inhaled fluticasone 500mcg twice daily and placebo all administered via MDI in the treatment of patients with COPD. SFCT01/SCO30002. GlaxoSmithKline Clinical Study Register website. http://www.gsk-clinicalstudyregister.com/search?study_ids=SCO30002#rs. Accessed July 1, 2013.
 
Sharafkhaneh A, Southard JG, Goldman M, Uryniak T, Martin UJ. Effect of budesonide/formoterol pMDI on COPD exacerbations: a double-blind, randomized study. Respir Med. 2012;106(2):257-268.
 
Ferguson GT, Anzueto A, Fei R, Emmett A, Knobil K, Kalberg C. Effect of fluticasone propionate/salmeterol (250/50 μg) or salmeterol (50 μg) on COPD exacerbations. Respir Med. 2008;102(8):1099-1108.
 
GlaxoSmithKline Clinical Study Register. A multicentre, randomised, double-blind, parallel group, 24 week study to compare the effect of the salmeterol/fluticasone propionate combination product 50/250mcg, with salmeterol 50mcg both delivered twice daily via the DISKUS®/ACCUHALER®inhaler on lung function and dyspnoea in subjects with chronic obstructive pulmonary disease (COPD). SCO100470. GlaxoSmithKline Clinical Study Register website. http://www.gsk-clinicalstudyregister.com/search?study_ids=SCO100470#rs. Accessed July 1, 2013.
 
Zheng JP, Yang L, Wu YM, et al. The efficacy and safety of combination salmeterol (50 μg)/fluticasone propionate (500 μg) inhalation twice daily via Accuhaler in Chinese patients with COPD. Chest. 2007;132(6):1756-1763.
 
Calverley P, Pauwels R, Vestbo J, et al; TRial of Inhaled STeroids ANd long-acting beta2 agonists study group. Combined salmeterol and fluticasone in the treatment of chronic obstructive pulmonary disease: a randomised controlled trial. Lancet. 2003;361(9356):449-456.
 
Hanania NA, Darken P, Horstman D, et al. The efficacy and safety of fluticasone propionate (250 μg)/salmeterol (50 μg) combined in the Diskus inhaler for the treatment of COPD. Chest. 2003;124(3):834-843.
 
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(8):1084-1091.
 
Aaron SD, Vandemheen KL, Fergusson D, et al; Canadian Thoracic Society/Canadian Respiratory Clinical Research Consortium. Tiotropium in combination with placebo, salmeterol, or fluticasone-salmeterol for treatment of chronic obstructive pulmonary disease: a randomized trial. Ann Intern Med. 2007;146(8):545-555.
 
Anzueto A, Ferguson GT, Feldman G, et al. Effect of fluticasone propionate/salmeterol (250/50) on COPD exacerbations and impact on patient outcomes. COPD. 2009;6(5):320-329.
 
Kardos P, Wencker M, Glaab T, Vogelmeier C. Impact of salmeterol/fluticasone propionate versus salmeterol on exacerbations in severe chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2007;175(2):144-149.
 
Wouters EF, Postma DS, Fokkens B, et al; COSMIC (COPD and Seretide: a Multi-Center Intervention and Characterization) Study Group. Withdrawal of fluticasone propionate from combined salmeterol/fluticasone treatment in patients with COPD causes immediate and sustained disease deterioration: a randomised controlled trial. Thorax. 2005;60(6):480-487.
 
GlaxoSmithKline Clinical Study Register. A randomized, double-blind, parallel-group clinical trial evaluating the effect of the fluticasone propionate/salmeterol combination product 250/50mcg bid via DISKUS versus salmeterol 50mcg bid via DISKUS on bone mineral density in subjects with chronic obstructive pulmonary disease (COPD). SCO40041. GlaxoSmithKline Clinical Study Register website. http://www.gsk-clinicalstudyregister.com/search?study_ids=SCO40041#ps. Accessed July 1, 2013.
 
Reid DW, Wen Y, Johns DP, Williams TJ, Ward C, Walters EH. Bronchodilator reversibility, airway eosinophilia and anti-inflammatory effects of inhaled fluticasone in COPD are not related. Respirology. 2008;13(6):799-809.
 
van der Valk P, Monninkhof E, van der Palen J, Zielhuis G, van Herwaarden C. Effect of discontinuation of inhaled corticosteroids in patients with chronic obstructive pulmonary disease: the COPE study. Am J Respir Crit Care Med. 2002;166(10):1358-1363.
 
Burge PS, Calverley PM, Jones PW, Spencer S, Anderson JA, Maslen TK. Randomised, double blind, placebo controlled study of fluticasone propionate in patients with moderate to severe chronic obstructive pulmonary disease: the ISOLDE trial. BMJ. 2000;320(7245):1297-1303.
 
Pauwels RA, Löfdahl CG, Laitinen LA, et al; European Respiratory Society Study on Chronic Obstructive Pulmonary Disease. Long-term treatment with inhaled budesonide in persons with mild chronic obstructive pulmonary disease who continue smoking. N Engl J Med. 1999;340(25):1948-1953.
 
Vestbo J, Sørensen T, Lange P, Brix A, Torre P, Viskum K. Long-term effect of inhaled budesonide in mild and moderate chronic obstructive pulmonary disease: a randomised controlled trial. Lancet. 1999;353(9167):1819-1823.
 
Bourbeau J, Rouleau MY, Boucher S. Randomised controlled trial of inhaled corticosteroids in patients with chronic obstructive pulmonary disease. Thorax. 1998;53(6):477-482.
 
Paggiaro PL, Dahle R, Bakran I, Frith L, Hollingworth K, Efthimiou J; International COPD Study Group. Multicentre randomised placebo-controlled trial of inhaled fluticasone propionate in patients with chronic obstructive pulmonary disease. Lancet. 1998;351(9105):773-780.
 
GlaxoSmithKline Clinical Study Register. A randomized, double-blind, parallel-group, comparative trial of inhaled fluticasone propionate 250mcg bid, 500mcg bid, and placebo bid via the DISKUS in subjects with chronic obstructive pulmonary disease (COPD). FLTA3025. GlaxoSmithKline Clinical Study Register website. http://www.gsk-clinicalstudyregister.com/search?study_ids=FLTA3025#rs. Accessed July 1, 2013.
 
Calverley PM, Kuna P, Monsó E, et al. Beclomethasone/formoterol in the management of COPD: a randomised controlled trial. Respir Med. 2010;104(12):1858-1868.
 
Rennard SI, Tashkin DP, McElhattan J, et al. Efficacy and tolerability of budesonide/formoterol in one hydrofluoroalkane pressurized metered-dose inhaler in patients with chronic obstructive pulmonary disease: results from a 1-year randomized controlled clinical trial. Drugs. 2009;69(5):549-565.
 
Shaker SB, Dirksen A, Ulrik CS, et al. The effect of inhaled corticosteroids on the development of emphysema in smokers assessed by annual computed tomography. COPD. 2009;6(2):104-111.
 
Tashkin DP, Rennard SI, Martin P, et al. Efficacy and safety of budesonide and formoterol in one pressurized metered-dose inhaler in patients with moderate to very severe chronic obstructive pulmonary disease: results of a 6-month randomized clinical trial. Drugs. 2008;68(14):1975-2000.
 
Choudhury AB, Dawson CM, Kilvington HE, et al. Withdrawal of inhaled corticosteroids in people with COPD in primary care: a randomised controlled trial. Respir Res. 2007;8:93.
 
Calverley PM, Boonsawat W, Cseke Z, Zhong N, Peterson S, Olsson H. Maintenance therapy with budesonide and formoterol in chronic obstructive pulmonary disease. Eur Respir J. 2003;22(6):912-919.
 
Szafranski W, Cukier A, Ramirez A, et al. Efficacy and safety of budesonide/formoterol in the management of chronic obstructive pulmonary disease. Eur Respir J. 2003;21(1):74-81.
 
Tashkin DP, Doherty DE, Kerwin E, et al. Efficacy and safety characteristics of mometasone furoate/formoterol fumarate fixed-dose combination in subjects with moderate to very severe COPD: findings from pooled analysis of two randomized, 52-week placebo-controlled trials. Int J Chron Obstruct Pulmon Dis. 2012;7:73-86.
 
Global strategy for asthma management and prevention. Updated 2012. GINA website. http://www.ginasthma.org/local/uploads/files/GINA_Report_March13.pdf. Accessed July 1, 2013.
 
Price D, Yawn B, Brusselle G, Rossi A. Risk-to-benefit ratio of inhaled corticosteroids in patients with COPD. Prim Care Respir J. 2013;22(1):92-100.
 
Ek A, Larsson K, Siljerud S, Palmberg L. Fluticasone and budesonide inhibit cytokine release in human lung epithelial cells and alveolar macrophages. Allergy. 1999;54(7):691-699.
 
Schleimer RP. Glucocorticoids suppress inflammation but spare innate immune responses in airway epithelium. Proc Am Thorac Soc. 2004;1(3):222-230.
 
Hogg JC, Chu FS, Tan WC, et al. Survival after lung volume reduction in chronic obstructive pulmonary disease: insights from small airway pathology. Am J Respir Crit Care Med. 2007;176(5):454-459.
 
Fauci AS, Dale DC, Balow JE. Glucocorticosteroid therapy: mechanisms of action and clinical considerations. Ann Intern Med. 1976;84(3):304-315.
 
Gupta A, Kaul A, Tsolaki AG, Kishore U, Bhakta S. Mycobacterium tuberculosis: immune evasion, latency and reactivation. Immunobiology. 2012;217(3):363-374.
 
Iovino F, Brouwer MC, van de Beek D, Molema G, Bijlsma JJ. Signalling or binding: the role of the platelet-activating factor receptor in invasive pneumococcal disease. Cell Microbiol. 2013;15(6):870-881.
 
American Thoracic Society/Centers for Disease Control and Prevention. Targeted tuberculin testing and treatment of latent tuberculosis infection. Am J Respir Crit Care Med. 2000;161(suppl):S221-S247.
 
Hale BG, Albrecht RA, García-Sastre A. Innate immune evasion strategies of influenza viruses. Future Microbiol. 2010;5(1):23-41.
 
Sin DD, Tashkin D, Zhang X, et al. Budesonide and the risk of pneumonia: a meta-analysis of individual patient data. Lancet. 2009;374(9691):712-719.
 
Singh S, Loke YK. Drug safety assessment in clinical trials: methodological challenges and opportunities. Trials. 2012;13:138.
 
Kesten S, Plautz M, Piquette CA, Habib MP, Niewoehner DE. Premature discontinuation of patients: a potential bias in COPD clinical trials. Eur Respir J. 2007;30(5):898-906.
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).
Supporting Data

Online Supplement

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543