Safetyandefficacy by Asociación Latinoamericana de Tórax ALAT

Safety and Efficacy of Combined Long-Acting β-Agonists and Inhaled Corticosteroids vs Long-Acting β-Agonists Monotherapy for Stable COPD Gustavo J. Rodrigo, José A. Castro-Rodriguez and Vicente Plaza Chest 2009;136;1029-1038; Prepublished online July 24, 2009; DOI 10.1378/chest.09-0821

The online version of this article, along with updated information and services can be found online on the World Wide Web at: http://chestjournal.chestpubs.org/content/136/4/1029.full.html

CHEST is the official journal of the American College of Chest Physicians. It has been published monthly since 1935. Copyright 2009 by the American College of Chest Physicians, 3300 Dundee Road, Northbrook, IL 60062. All rights reserved. No part of this article or PDF may be reproduced or distributed without the prior written permission of the copyright holder. (http://chestjournal.chestpubs.org/site/misc/reprints.xhtml) ISSN:0012-3692

CHEST

Original Research COPD

Safety and Efficacy of Combined Long-Acting ␤-Agonists and Inhaled Corticosteroids vs Long-Acting ␤-Agonists Monotherapy for Stable COPD A Systematic Review Gustavo J. Rodrigo, MD; Jose´ A. Castro-Rodriguez, MD, PhD; and Vicente Plaza, MD

Background: Current guidelines recommend the use of inhaled corticosteroids (ICSs) added to long-acting ␤2-agonists (LABAs) for treatment of symptomatic patients with severe and very severe COPD. However, the evidence has been inconclusive. The aim of this review was to assess the safety and efficacy of LABAs/ICSs compared with LABA monotherapy for patients with moderate-to-very severe COPD. Methods: Systematic searches were conducted on MEDLINE, EMBASE, the Cochrane Controlled Trials Register, and the trial registers of manufacturers, without language restriction. Primary outcomes were COPD exacerbations and mortality. Secondary outcomes included lung function, health-related quality of life, and adverse effects. Results: Eighteen randomized controlled trials (12,446 participants) were selected. Therapy with LABAs/ICSs did not decrease the number of severe exacerbations (relative risk [RR], 0.91; 95% CI, 0.82 to 1.01; I2 ⴝ 1%), or all-cause mortality (RR, 0.90; 95% CI, 0.76 to 1.06; I2 ⴝ 0%), respiratory mortality (RR, 0.80; 95% CI, 0.61 to 1.05; I2 ⴝ 0%), and cardiovascular mortality (RR, 1.22; 95% CI, 0.88 to 1.71; I2 ⴝ 0%). To the contrary, the number of moderate exacerbations (RR, 0.84; 95% CI, 0.74 to 0.96; I2 ⴝ 50%) and the St. George respiratory questionnaire total score (weighted mean difference, ⴚ1.88; 95% CI, ⴚ2.44 to ⴚ1.33; I2 ⴝ 29%) were significantly reduced with LABA/ICS therapy. Although therapy with LABAs/ICSs increases FEV1 significantly (0.06 and 0.04 L, respectively), they were associated with an increased risk of pneumonia (RR, 1.63; 95% CI, 1.35 to 1.98; I2 ⴝ 20%). Conclusions: Compared with LABA monotherapy, the magnitude of the benefits of LABA/ICS therapy did not reach that of the criteria for predefined clinically important effects and were associated with serious adverse effects. (CHEST 2009; 136:1029 –1038) Abbreviations: ICS ⫽ inhaled corticosteroid; LABA ⫽ long-acting ␤2-agonist; MI ⫽ myocardial infarction; NNTB ⫽ number needed to treat for benefit; NNTH ⫽ number needed to treat for harm; RR ⫽ relative risk; SABA ⫽ short-acting ␤2-agonist; SGRQ ⫽ St. George respiratory questionnaire; WMD ⫽ weighted mean difference

is a preventable and treatable disease that C OPD is characterized by airflow limitation that is not fully reversible.1,2 The main therapeutic goals are to prevent and control symptoms, reduce the frequency and severity of exacerbations, and improve health status and exercise tolerance. Current guidelines1,2 recommend a stepwise increase in treatment, depending on the severity of the disease. Short-acting inhaled bronchodilators (short-acting ␤2-agonists [SABAs] and anticholinergic agents) are recomwww.chestjournal.org

mended for the relief of symptoms on an as-needed basis, whereas long-acting inhaled bronchodilators (long-acting ␤2-agonists [LABAs] or tiotropium) in a regularly scheduled regimen are recommended as first-line therapy in symptomatic patients with moderate-to-very severe COPD. Evidence3 shows that LABA monotherapy is associated with significant improvements regarding COPD exacerbations, pulmonary function, quality of life, and use of rescue medication, with a low CHEST / 136 / 4 / OCTOBER, 2009

1029

incidence of adverse effects. However, in view of the multicomponent nature of COPD, the use of inhaled corticosteroids (ICSs), particularly in combination with a LABA has obtained widespread acceptance among clinicians. In fact, the addition of ICSs in patients with severe or very severe disease (stages III or IV) with repeated exacerbations is recommended for decreasing exacerbation rates, and improving lung function and health status.1,2 Nevertheless, the evidence of the superiority of combination therapy (LABAs/ICSs) over LABA monotherapy has been inconclusive. While a 2007 review4 failed to demonstrate the superiority of combination therapy over LABA monotherapy in reducing COPD exacerbations, others have reported5–7 some benefits of therapy with LABAs/ICSs in terms of COPD exacerbations, pulmonary function, and quality of life. Even so, these conclusions might be questionable because they are based on a reduced number of selected studies and outcomes. Consequently, we performed a systematic review to assess the safety and efficacy of the use of LABAs/ICSs in COPD patients compared with LABA monotherapy. The following two specific questions were identified: (1) what are the risks of adding an ICS to a LABA compared with LABAs monotherapy? and (2) does therapy with LABAs/ICSs provide significant clinical benefits compared with LABA monotherapy?

Materials and Methods Search Strategy and Selection Criteria We identified studies from MEDLINE, EMBASE (January 1980 to May 2009), and the Cochrane Controlled Trials Register (second quarter of 2009) databases by using the following MeSH, full text, and keywords terms: (long-acting ␤2 adrenoceptor agonist OR salmeterol OR formoterol OR inhaled corticosteroids OR fluticasone OR budesonide OR beclomethasone) AND (COPD OR chronic bronchitis OR emphysema). Also, we performed a search of relevant files from AstraZeneca Manuscript received April 2, 2009; revision accepted July 2, 2009. Affiliations: From the Departamento de Emergencia (Dr. Rodrigo), Hospital Central de las Fuerzas Armadas, Montevideo, Uruguay; the School of Medicine (Dr. Castro-Rodriguez), Pontificia Universidad Cato´lica de Chile, Santiago, Chile; and Servei de Pneumologia (Dr. Plaza), Hospital de la Santa Creu i Sant Pau, Universitat Autonoma de Barcelona, Barcelona, Spain. Funding/Support: The funding for this study came from salary support for Dr. Rodrigo. No sponsorship from institutions or pharmaceutical industry was provided to conduct this study. Correspondence to: Gustavo J. Rodrigo, MD, Departamento de Emergencia, Hospital Central de las Fuerzas Armadas, Av 8 de Octubre 3020, Montevideo 11600, Uruguay; e-mail: gurodrig@ adinet.com.uy © 2009 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal.org/site/ misc/reprints.xhtml). DOI: 10.1378/chest.09-0821

(www.astrazenecaclinicaltrials.com) and GlaxoSmithKline (www. gsk-clinicalstudyregister.com) databases. Trials published solely in abstract form were excluded because the methods and results could not be fully analyzed. The specific inclusion criteria were as follows: (1) stable adult patients aged ⬎ 40 years with COPD satisfying the diagnostic criteria of the American Thoracic Society/European Respiratory Society1 or the Global Initiative for Chronic Obstructive Lung Disease2; (2) therapy with inhaled LABAs plus ICSs (delivered via metered-dose inhaler or dry powder inhaler) as the intervention arm compared with therapy with a LABA; (3) study durations of ⬎ 1 month; (4) randomized controlled trials (parallelgroup design) without language restriction; and (5) primary outcomes “severe COPD exacerbation” (requiring hospitalization or withdrawals) and “moderate COPD exacerbations” (requiring systemic corticosteroids or antibiotic use), all-cause mortality (deaths for any cause), respiratory deaths (deaths due to a respiratory event such as COPD exacerbation or pneumonia), and cardiovascular mortality (including sudden death) during the treatment period. Secondary outcome measures were as follows: mean change in FEV1 (pre-bronchodilator therapy and postbronchodilator therapy); mean change from baseline in the St. George respiratory questionnaire (SGRQ) total score8; end-oftreatment dyspnea score; withdrawals from the study during the treatment period (overall, due to adverse effects, and due to lack of efficacy); and adverse effects (pneumonia, oropharyngeal candidiasis, viral respiratory infections, and myocardial infarctions [MIs]). Data Abstraction and Validity Assessment Titles, abstracts, and citations were independently analyzed by all reviewers. From full text, they independently assessed studies for inclusion based on the criteria for population, intervention, study design, and outcomes. Three reviewers (G.J.R., J.C.R., and V.P.) were independently involved in all stages of study selection, data extraction, and quality assessment. Any disagreement was resolved by consensus. In case of multiple published or unpublished reports for a particular study, data from the most recent version were extracted. Statistical Analysis Binary outcomes were pooled by using common relative risks (RRs) and 95% CIs. If pooled effect estimates were significantly different between groups, we calculated the number needed to treat for benefit (NNTB) or the number needed to treat for harm (NNTH). For continuous outcomes, the standardized mean difference or weighted mean difference (WMD) and 95% CIs were calculated. Heterogeneity was further measured by using the I2 test.9 With low heterogeneity (I2 ⬍ 40%), data were combined by mean of a fixed-effects model10; otherwise, a random-effects model was used. Publication bias of primary outcomes was evaluated by means of the visual inspection of funnel plots.11 A predefined sensitivity analysis of the primary outcome of severe COPD exacerbations was conducted to explore the influence of the following factors: concealment allocation12 (adequate vs unclear); trial duration (long-term [ⱖ 52 weeks] vs short-term [⬍ 52 weeks]); reversibility to SABA (poorly reversible patients or FEV1 ⬍ 15% from baseline vs reversible patients or FEV1 ⱖ 15% from baseline); choice of LABAs (salmeterol vs formoterol); and the use of ICSs before the patients were enrolled (ⱖ 50% of patients vs ⬍ 50% of patients). Subgroups were compared by using the interaction test.13 A p value of ⬍ 0.05 using a two-tailed test was considered to indicate significance. Metaanalyses were performed with using a

1030

Original Research

weeks)16 –18,20,21,23–25,27–29 and 7 short-term trials (ie, ⬍ 52 weeks).14,15,19,22,26,30,31 Studies enrolled mostly stable patients with COPD who met the Global Initiative for Chronic Obstructive Lung Disease criteria for moderate-to-very severe COPD exacerbations.2 The mean age of patients was 64 years (72% of patients were male), with an average baseline FEV1 of 40% of predicted normal values. Allocation concealment was adequate in only 5 studies,16,20,21,23,24 and was unclear in the remaining 13 studies. Primary Outcomes

Figure 1. Flowchart for identification of studies.

statistical software package (Review Manager, version 5.0.20; the Nordic Cochrane Centre, the Cochrane Collaboration; Copenhagen, Denmark).

Results Of 164 potential relevant citations, 18 randomized, controlled trials14 –31 fulfilled the inclusion criteria (Fig 1). Five trials were unpublished.27–31 Data analysis was restricted to the LABAs/ICSs and LABAs arms of those trials (12,446 patients) [Table 1]. Five studies used formoterol/budesonide combination therapy,17,20,26,27 and 13 studies used salmeterol/fluticasone combination therapy.14 –16,18,19,21–31 Seven studies14 –16,21,23,24,31 evaluated therapy with inhaled fluticasone in combination with salmeterol at a dosage of 500 ␮g twice daily, seven studies18,19,22,25,28 –30 assessed therapy with fluticasone at a dosage of 250 ␮g twice daily, and four studies17,20,26,27 evaluated therapy with budesonide at a dosage of 320 ␮g twice daily. All trials used single inhalers containing both ICSs and LABAs to deliver the combined therapy. Eleven trials15–21,23–26 reported that a mean of 31% of patients (range, 0 to 55% of patients) had received ICSs before they were enrolled in the study. There were 11 long-term trials (ie, ⱖ 52 www.chestjournal.org

Compared with LABA monotherapy, combination therapy with LABAs/ICSs did not significantly decrease the risk of severe COPD exacerbations (11.3% vs 12.5%, respectively) [Table 2]. The post hoc subgroup analysis did not show significant differences in COPD exacerbations regarding concealment allocation, trial duration, reversibility to SABA use, LABA choice, and use of ICSs before the patients were enrolled in the study (Table 3). To the contrary, the use of LABAs/ICSs was associated with a significantly reduced risk of moderate COPD exacerbations when compared with LABAs alone (17.5% vs 20.1%, respectively), with evidence of statistical heterogeneity among trials. The NNTB was 31 (95% CI, 20 to 93). Patients receiving therapy with LABAs/ICSs were not associated with a significant decrease of overall mortality when compared with those receiving LABAs alone (4.5% vs 5.5%, respectively) [Fig 2]. In the same way, the metaanalysis did not show significant differences between groups regarding the risk of respiratory deaths (1.8% in the LABAs/ ICSs group vs 2.4% in the LABAs group) and cardiovascular mortality (1.6% vs 1.4%, respectively), without evidence of statistical heterogeneity (Fig 2). On visual inspection of the funnel plots (Fig 3), publication bias could be ruled out for all-cause, respiratory, and cardiovascular mortality. To the contrary, for severe COPD exacerbations, the plot presented an asymmetrical shape with an absence of small studies showing a group benefit for therapy with LABAs. Secondary Outcomes Regarding pulmonary function, patients treated with LABAs/ICSs showed significantly greater increases in the mean change in FEV1 from baseline (pre-bronchodilator therapy and post-bronchodilator therapy) compared with patients treated with LABAs alone (Table 4); however, both comparisons showed statistical heterogeneity. Patients receiving therapy with LABAs/ICSs showed a significantly greater reduction in the SGRQ total score. Finally, at the CHEST / 136 / 4 / OCTOBER, 2009

1031

Table 1—Characteristics of Included Studies

Studies Cazzola et al14

Mahler et al15

Mean Age, yr

Mean Baseline FEV1

Current Smoker

1.15 L

45 pack-yr

⬍ 12

SF: 40 (92)

1.16 L

44 pack-yr

⬍ 12

S: 160 (64)

40% predicted

46%

SF: 165 (62)

41% predicted

46%

S: 372 (70)

46% predicted

51%

3.7

SF: 358 (75)

49% predicted

52%

4.0

F: 255 (75)

36% predicted

36%

FB: 254 (78)

36% predicted

33%

Single-center/52 wk S: 6 (100)

55–78

48% predicted

SF: 5 (83)

53–77

50% predicted

3.5

S: 177 (58)

42% predicted

51%

SF:178 (61)

41% predicted

43%

F: 201 (76)

36% predicted

38%

FB:208 (76)

36% predicted

30%

S: 184 (75)

1.41 L

35%

SF: 189 (73)

1.41 L

39%

S: 59 (75)

40% predicted

34%

SF: 62 (69)

42% predicted

42%

44% predicted

43%

SF: 1,533 (75)

44% predicted

43%

S: 487 (78)

40% predicted

37%

SF: 507 (74)

40% predicted

37%

S: 388 (52)

33% predicted

38%

SF: 394 (58)

33% predicted

40%

F: 284 (66)

39% predicted

42%

FB: 845 (69)

39% predicted

44%

F: 495 (65)

34% predicted

41%

FB: 988 (63)

34% predicted

36%

Study Location/ Duration

Single-center/12 wk S: 20 (90)

69 centers/24 wk

Calverley et al16 196 centers/52 wk

Calverley et al17 109 centers/52 wk

Dal Negro et al18

Hanania et al19

75 centers/24 wk

Szafranski et al20 89 centers/52 wk

Wouters et al21

O’Donnell et al22

Patients: No. (% male)

39 centers/52 wk

21 centers/8 wk

Calverley et al23 444 centers/156 wk S: 1,521 (76)

Kardos et al24

95 centers/ 44 wk

Ferguson et al25 94 centers/52 wk

Tashkin et al26

D-5899C0 000127

194 centers/26 wk

237 centers/52 wk

Bronchodilator Response, %

Intervention: Dose S: 50 ␮g twice MDI SF: 50/250 or 50/500 ␮g twice MDI S: 50 ␮g twice DPI SF: 50/500 ␮g twice DPI S: 50 ␮g twice DPI SF: 50/500 ␮g twice DPI F: 9 ␮g twice DPI FB: 9/320 ␮g twice DPI S: 50 ␮g twice DPI SF: 50/ 250 ␮g twice DPI S: 50 ␮g twice DPI SF: 50/250 ␮g twice DPI F: 9 ␮g twice DPI FB: 9/320 ␮g twice DPI S: 50 ␮g twice DPI SF: 50/500 ␮g twice DPI S: 50 ␮g twice DPI SF: 50/250 ␮g twice DPI S: 50 ␮g twice DPI SF: 50/500 ␮g twice DPI S: 50 ␮g twice DPI SF: 50/500 ␮g twice DPI S: 50 ␮g twice DPI SF: 50/250 ␮g DPI twice F: 9 ␮g twice DPI FB: 9/160 ␮g or 320/9 ␮g twice MDI F: 9 ␮g twice DPI FB: 9/320 ␮g or 160/9 MDI twice

Allocation Concealment Unclear

Unclear

Adequate

Unclear

Adequate

Unclear

Adequate

Unclear

(Continued)

1032

Original Research

Table 1—(Continued)

Studies

Study Location/ Duration

SCO4004128

31 centers/156 wk

SCO10025029

SCO10047030

SCO10492531

98 centers/52 wk

135 centers/24 wk

11 centers/12 wk

Mean Age, yr

Mean Baseline FEV1

Current Smoker

Bronchodilator Response, %

S: 94 (63)

⬍ 70% predicted

SF: 92 (63)

⬍ 70% predicted

S: 403 (57)

⬍ 50% predicted ⬎ 10 pack-yr

SF: 394 (51)

⬍ 50% predicted ⬎ 10 pack-yr

S: 532 (78)

1.68 L

44%

SF: 518 (82)

1.65 L

42%

S: 38 (79)

⬍ 50% predicted ⬎ 10 pack-yr

SF: 39 (82)

⬍ 50% predicted ⬎ 10 pack-yr

Patients: No. (% male)

Intervention: Dose

Allocation Concealment

S: 50 ␮g twice DPI SF: 50/250 ␮g DPI twice S: 50 ␮g twice DPI SF: 50/250 ␮g DPI twice S: 50 ␮g twice DPI SF: 50/250 ␮g DPI twice S: 50 ␮g twice DPI SF: 50/500 ␮g DPI twice

Unclear

NA ⫽ not applicable; S ⫽ salmeterol; F ⫽ formoterol; SF ⫽ salmeterol/fluticasone; FB ⫽ formoterol/budesonide; MDI ⫽ metered-dose inhaler; DPI ⫽ dry powder inhaler.

end of the protocol, patients treated with combination therapy had a significantly lower dyspnea score compared with patients treated with LABAs alone, although with evidence of statistical heterogeneity among the trials (Table 4). LABAs/ICSs therapy was associated with a significant decrease of overall withdrawals from the study (Table 4). Furthermore, patients receiving therapy with LABAs/ICSs were associated with a significantly lower rate of withdrawals from the study due to lack of efficacy, but not with significant withdrawals from the study due to adverse effects (Table 4). However, the use of LABAs/ICSs was also associated with significantly increased rates of pneumonia (63% increase in RR), viral respiratory infections (22% increase in RR), and oropharyngeal candidiasis (59%

increase in RR) compared with the use of LABAs alone. Finally, combined therapy did not show a significant difference in the rate of MI, compared with LABA monotherapy.

Discussion In the present study, which is the largest systematic review designed to evaluate the safety and efficacy of the regular use of LABAs/ICSs compared with the use of LABAs alone in stable patients with moderate-to-very severe COPD, we found that treatment with LABAs/ICSs did not modify the risks of overall mortality, respiratory deaths, and cardiovascular mortality (primary outcomes) compared with

Table 2—Analysis of Primary Outcomes (LABAs/ICSs vs LABA)

Outcomes COPD exacerbations (requiring hospitalization or withdrawal) COPD exacerbations requiring systemic corticosteroids All-cause mortality Respiratory deaths Cardiovascular mortality

www.chestjournal.org

References

LABAs ⫹ ICS, No./Total No. (%)

LABAs, No./ Total No. (%)

15–17,19–30

757/6,685 (11.3)

704/5,612 (12.5)

RR ⫽ 0.91 (0.82–1.01)

⫺1.2 (⫺2.4 to 0.0)

15–26,29

794/4,532 (17.5)

1,015/5,058 (20.1)

⫺2.5 (⫺1.0 to ⫺4.1)

16,17,20,21, 23–26,28–30 16,17,20,21, 23–26,28–30 16,21,23–26, 28–30

240/5,292 (4.5)

261/4,721 (5.5)

RR ⫽ 0.84 (0.74–0.96); p ⫽ 0.008/NNTB ⫽ 31 (20–93) RR ⫽ 0.90 (0.76–1.06)

⫺1.0 (⫺1.8 to 0.0)

94/5,292 (1.8)

114/4,721 (2.4)

RR ⫽ 0.80 (0.61–1.05)

⫺0.6 (⫺1.2 to 0.0)

72/5,856 (1.6)

63/5,299 (1.4)

RR ⫽ 1.22 (0.88–1.71)

0.2 (⫺0.3 to 0.7)

Measure (95% CI)

Absolute Risk Reduction

CHEST / 136 / 4 / OCTOBER, 2009

I2, % 1

1033

Table 3—Sensitivity Analysis. Comparisons Between RR in COPD Exacerbations Requiring Hospitalization or Withdrawal) Stratified by Concealment Allocation (Adequate vs Unclear), Trial Duration (Long-Term > 52 Weeks vs Short-Term < 52 Weeks), Reversibility to SABAs (Poorly Reversible Patients or FEV1 < 15% From Baseline vs Reversible Patients or FEV1 > 15%), Baseline Severity (Moderate to Severe vs Severe to Very Severe), LABAs choice (Salmeterol vs Formoterol), and Use of ICSs Before the Patients Were Enrolled (< 50% of Patients vs < 50% of Patients) Subgroup Comparisons 16,20,21,23,24

Adequate vs unclear concealment14,15,17–19,22,25–31 Long-term16,17,20,21,23,25,27–29 vs short-term15,19,22,24,26,30 Poorly reversible16,17,20,21,23,24 vs reversible15,19,22,25–30 Salmeterol15,16,19,21–25,28–30 vs Formoterol17,20,26,27 Previous use of ICS (ⱖ 50%16, 24,26 vs ⬍ 50% of patients15,17–21,23–25)

Interactive Test17 RR (95% CI)

p Value

0.93 (0.82–1.05) 关35%兴 vs 0.90 (0.77–1.05) 关0%兴

0.96 (0.79–1.18)

0.74

0.93 (0.84–1.03) 关1%兴 vs 0.80 (0.59–1.11) 关0%兴

0.86 (0.61–1.19)

0.37

0.90 (0.80–1.01) 关0%兴 vs 0.94 (0.74–1.20) 关0%兴

1.04 (0.79–1.36)

0.75

0.94 (0.84–1.05) 关0%兴 vs 0.87 (0.73–1.04) 关41%兴

0.92 (0.75–1.14)

0.46

0.84 (0.65–1.14) 关28%兴 vs 0.92 (0.88–1.04), 关31%兴

1.06 (0.79–1.43)

0.65

RR (95% CI) 关I2兴

treatment with LABAs alone. On the contrary, the analysis of secondary outcomes showed that therapy with LABAs/ICSs significantly increased the risk of pneumonia, oropharyngeal candidiasis, and viral respiratory infections (question 1). However, it is interesting to point out that these adverse effects were not accompanied by a concomitant and proportional increase in respiratory-related mortality or overall mortality. Concerning the benefits (question 2), we found that therapy with LABAs/ICSs significantly decreased the frequency of moderate COPD exacerbations independently of concealment allocation, trial duration, reversibility to SABA therapy, LABA choice, and previous use of ICSs. In the same way, therapy with LABAs/ICSs was associated with significant increases in the mean change in pre-bronchodilator therapy and postbronchodilator therapy FEV1, the mean change in SGRQ total score, and with a significant decrease in the end-of-treatment dyspnea score compared with treatment with LABAs alone. However, given that the size of these benefits did not reach the suggested clinically important minimal differences (FEV1, 0.10 to 0.14 L; SGRQ score, 4-unit decrement),32 the relevance of these improvements seems uncertain. In the same way, the 16% decrease in the rate of moderate COPD exacerbations was smaller than the suggested threshold value of 22% for clinical significance.32 Conversely, treatment with LABAs/ICSs failed to significantly reduce the risk of severe COPD exacerbations. These facts suggest a limited extra efficacy when ICSs were added to LABAs for COPD treatment. The relative benefits of therapy with LABAs/ICSs must be weighed against the risks. Thus, the most concerning side effect was the increase in the risk of pneumonia associated with the administration of

ICSs added to LABAs. The precise mechanism is uncertain, but it could be related to the fact that ICSs achieve locally high concentrations in the lung, increasing the risk of pneumonia due to their immunosuppressive effects.33 Thus, inhaled fluticasone at dosages of 1,000 ␮g/d exerts effects on serum cortisol levels that are equivalent to 10 mg of prednisone, a dose that may double the risk of pneumonia in patients with arthritis.34 Our analysis showed an increase in the risk of pneumonia with both moderate dosages of fluticasone (500 ␮g/d; RR, 1.75; 95% CI, 1.16 to 2.64; I2 ⫽ 30%) and high dosages of fluticasone (1,000 ␮g/d; RR ⫽ 1.64; 95% CI, 1.32 to 2.06, I2 ⫽ 22%). To the contrary, therapy with LABAs/ICSs was associated with significant decreases in overall withdrawals and withdrawals due to lack of efficacy compared with LABA monotherapy. This finding could be associated with the fact that therapy with LABAs/ICSs significantly reduces dyspnea with a greater clinical effectiveness perception. Also, this fact could be associated with better control of the disease. When we compared the results of the present metaanalysis with those of previous reviews (based on a limited number of published trials), we found some similarities and few differences. For example, a Cochrane review5 of 10 studies, reported a significant RR reduction of moderate COPD exacerbations of 18% with the combined treatment with LABAs/ ICSs compared with treatment with LABAs alone. Also, the combination therapy was more effective than that with LABAs alone in improving quality of life as measured by the SGRQ (⫺1.64 points), and predose and postdose FEV1 (0.06 and 0.05 L, respectively). While there was no significant difference in terms of overall mortality, pneumonia occurred more commonly in patients receiving

1034

Original Research

Figure 2. Pooled RR for overall mortality, respiratory deaths, and cardiovascular mortality (with 95% CIs) of eligible studies comparing inhaled LABAs/ICSs with LABAs.

combined therapy (58% increase in RR). However, another metaanalysis4 that was limited to five English-language trials failed to demonstrate the superiority of combination treatment over LABA monotherapy in reducing COPD exacerbations and overall mortality. More recently, Sobieraj et al7 on the basis of seven studies reported that therapy with LABAs/ICSs decreased the risk of moderate COPD exacerbations (relative reduction risk of 18%), decreased the SGRQ total score (⫺1.98 points), and increased the risk of pneumonia (32% increase in RR) compared www.chestjournal.org

with therapy with LABAs alone. Additionally, combined therapy also showed a reduced risk in overall withdrawals (17% increase in RR), and without difference in the rate of overall mortality between groups. Another systematic review35 has found no evidence that therapy with LABAs alone is more effective than combined therapy with LABAs/ICSs. In the same study, a subgroup analysis showed that, when added to LABAs, ICSs significantly reduced the number of exacerbations in patients with FEV1 ⱕ 40% predicted. However, these conclusions are uncertain because the review was based on only CHEST / 136 / 4 / OCTOBER, 2009

1035

Figure 3. Funnel plots of LABAs/ICSs vs LABAs comparing COPD severe exacerbations (A), all-cause mortality (B), respiratory deaths (C), and cardiovascular mortality (D).

seven studies15–17,19,20,23,24 (6,376 patients with COPD). Finally, Singh et al36 showed a significantly increased risk of pneumonia with combined therapy compared with LABA monotherapy (32% increase in RR) without a significantly increased risk of death. Interestingly, the risk of pneumonia could be specifically attributed to the use of ICSs because the risk for pneumonia associated with ICS use was similar when therapy with ICSs were compared with placebo or when ICSs were added to therapy with LABAs and compared with LABA therapy alone. In the same way, a recent trial37 comparing therapy with inhaled tiotropium with therapy with salmeterol/fluticasone in patients with COPD showed an increased hazard ratio for time to reported pneumonia for combined therapy of ⬎ 2 years (94% increase in RR). This review was performed according to the methodological criteria suggested for scientific reviews.38 The fact that there was low evidence of clinical and statistical heterogeneity between studies increased the confidence of our findings. Furthermore, there was no evidence of publication bias in the majority of primary outcomes. However, our metaanalysis had

several potential limitations that came from the quality of the reported data. Thus, the trials did not consistently use similar definitions of COPD exacerbations or pneumonia. In particular, we recognized that the severity of exacerbations is a complex concept constituted by several factors, and that this fact could modify our results. Also, most of the studies were not specifically designed to monitor outcomes as all-cause, respiratory, or cardiovascular mortality. Additionally, the risk of bias was unclear in 13 trials in the analysis. Also, the fact that 80% of the reviewed patients were men limits the applicability of the results since COPD is suspected to affect men and women equally. This metaanalysis confirms and extends data from previous reviews. The main results of our review are as follows: LABAs/ICSs did not decrease the risk of all-cause, respiratory, and cardiovascular mortality; however, therapy with LABAs/ICSs increases the risk of pneumonia, oropharyngeal candidiasis, and viral respiratory infections. The use of LABAs/ICSs was associated with a lower incidence of moderate COPD exacerbations (but not of severe exacerbations), increased pulmonary function, improved dys-

1036

Original Research

Table 4 —Analysis of Secondary Outcomes (LABAs/ICSs vs LABAs) LABAs ⫹ ICSs, No./Total No. (%)

LABAs, No./ Total No. (%)

Absolute Risk Reduction

Measure (95% CI) 关p Value兴

I2, %

Outcomes

References

Mean change in pre-bronchodilator therapy FEV1, L Mean change in post-bronchodilator therapy FEV1, L Mean change in SGRQ End-of-treatment dyspnea score Total withdrawals

14–16,19,21–27, 29–30

5,613

5,082

WMD ⫽ 0.06 (0.04–0.07) 关0.0001兴

14–16,19,22, 23,26,27

3,455

2,501

WMD ⫽ 0.04 (0.02–0.05) 关0.0001兴

Withdrawals due to adverse effects Withdrawals due to lack of efficacy Pneumonia Oropharyngeal candidiasis Viral respiratory infections MI

16,17,20, 23–26,30 15,16,19,21,22, 24–26,30 14–17,19–31 15,16, 19–21, 23–26,28–30 15,16,23,25, 29,30 15,16,19,21, 23–26,28–31 16,19,21,22, 24,26,27,29 15–17,19,22,23, 25,27,29,30 16,23–25,28,29

WMD ⫽ ⫺1.88 (⫺2.44 to ⫺1.33) 关0.0001兴 3,216 2,643 SMD ⫽ ⫺0.20 (⫺0.25 to ⫺0.15) 关0.0001兴 1,731/5,919 (29.2) 1,731/5,919 (29.2) RR ⫽ 0.87 (0.82–0.92) 关0.0001兴; ⫺1.6 (⫺0.2 to ⫺3.0) NNTB ⫽ 25 (18–41) 680/5,381 (12.6) 654/4,803 (13.6) RR ⫽ 0.93 (0.84–1.03) 关0.26兴 ⫺1.0 (⫺0.3 to 2.3) 4,617

4,040

63/3,362 (1.9)

117/3,376 (3.5)

263/5,212 (5.0)

153/4,540 (3.4)

292/3,521 (8.4)

200/2,741 (7.2)

441/4,844 (9.1)

342/4,362 (7.8)

34/3,278 (1.0)

33/3,265 (1.0)

RR ⫽ 0.54 (0.40–0.72) 关0.0001兴; NNTB ⫽ 73 (56–120) RR ⫽ 1.63 (1.35–1.98) 关0.0001兴; NNTH ⫽ 40 (26–72) RR ⫽ 1.59 (1.07–2.37) 关0.002兴; NNTH ⫽ 22 (10–179) RR ⫽ 1.22 (1.07–1.39) 关0.004兴; NNTH ⫽ 57 (33–179) RR ⫽ 1.03 (0.64–1.64) 关0.91兴

⫺1.6 (⫺0.8 to ⫺2.4)

29 82 0 0 0

1.7 (0.9 to 2.5)

1.2 (0.1 to 2.3)

1.3 (0.1 to 2.4)

0 (⫺0.5 to 0.5)

SMD ⫽ standardized mean difference.

pnea, and health-related quality-of-life total scores. However, the magnitude of these benefits did not reach the recent predefined criteria to be clinical important.32 Current guidelines1,2 recommend the use of ICSs in combination with LABAs to reduce the frequency of exacerbations in symptomatic patients with severe and very severe COPD. Furthermore, some authors39 have suggested that, in COPD patients as in asthma patients, concomitant ICS therapy is preferable over LABA monotherapy. Nevertheless, this review suggests that combination therapy with LABAs/ICSs presents a borderline statistical and limited clinical significance compared with LABA monotherapy. Moreover, combination therapy offers no statistically significant additional survival benefit and increased the risk of serious adverse effects. Even so, this last issue requires further prospective evaluation in large studies using objective definitions of pneumonia. It is likely that most patients with COPD with these levels of severity should be treated only with LABA monotherapy. However, it is possible also that a future definition of different COPD phenotypes will allow us to know which patients can benefit from ICSs, and which should only be treated with LABAs. Thus, patients who benefit from combination therapy with LABAs/ ICSs could be those patients with steroid-responsive eosinophilic bronchitis. www.chestjournal.org

Acknowledgments Author contributions: Dr. Rodrigo (1) has made substantial contributions to conception and design, acquisition of data, and analysis and interpretation of data; (2) has drafted the submitted article and revised it critically for important intellectual content; and (3) has provided final approval of the version of the article to be published. Dr. Castro-Rodriguez (1) has made substantial contributions to conception and design, and interpretation of data; (2) has revised the article critically for important intellectual content; and (3) has provided final approval of the version to be published. Dr. Plaza (1) has made substantial contributions to conception and design, and interpretation of data; (2) has revised the article critically for important intellectual content; and (3) has provided final approval of the version to be published. Financial/nonfinancial disclosures: Dr. Rodrigo has participated as a lecturer and speaker in scientific meetings and courses under the sponsorship of Boehringer Ingelheim, GlaxoSmithKline, AstraZeneca, Dr. Esteve SA, and Merck Sharp and Dome. Dr. Castro-Rodriguez has participated as a lecturer and speaker in scientific meetings and courses under the sponsorship of Merck Sharp and Dohme, GlaxoSmithKline, and Grunenthal; and as member of the advisory board for GlaxoSmithKline. Dr. Plaza has participated as a lecturer and speaker in scientific meetings and courses under the sponsorship of AstraZeneca, GlaxoSmithKline, Dr. Esteve SA, and Merck Sharp and Dohme.

References 1 Celli BR, MacNee W, ATS/ERS Task Force. Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J 2004; 23:932–946 2 Global Initiative for Chronic Obstructive Lung Disease. Global strategy for the diagnosis, management and prevention of CHEST / 136 / 4 / OCTOBER, 2009

1037

3 4

6 7 8 9 10 11 12

13 14 15

17 18

20 21

COPD. Available at: www.goldcopd.org. Accessed August 27, 2009 Rodrigo GJ, Nannini LJ, Rodriguez-Roisín R. Safety of long-acting ␤-agonists in stable COPD: a systematic review. Chest 2008; 133:1079 –1087 Wilt TJ, Niewoehner D, MacDonald R, et al. Management of stable chronic obstructive pulmonary disease: a systematic review for a clinical practice guideline. Ann Intern Med 2007; 147:639 – 653 Nannini LJ, Cates CJ, Lasserson TJ, et al. Combined corticosteroid and long-acting ␤-agonist in one inhaler versus long-acting ␤-agonists for chronic obstructive pulmonary disease. Cochrane Database Syst Rev (database online). Issue 4, 2007 Hanania NA. The impact of inhaled corticosteroid and longacting ␤-agonist combination therapy on outcomes in COPD. Pulm Pharmacol Ther 2008; 21:540 –550 Sobieraj DM, White M, Coleman CI. Benefits and risks of adjunctive inhaled corticosteroids in chronic obstructive pulmonary disease: a meta-analysis. Clin Ther 2008; 30:1416 –1425 Jones PW, Quirk FH, Baveystock CM. The St. George’s respiratory questionnaire. Respir Med 1991; 85:25–31 Higgins JPT, Thompson SG, Deecks JJ, et al. DG. Measuring inconsistency in meta-analyses. BMJ 2003; 327:557–560 Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959; 22:719 –748 Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997; 315:629 – 634 Higgins JPT, Green S, eds. Cochrane handbook for systematic reviews of interventions version 5.0.1 [updated September 2008]. Available at: www.cochrane-handbook.org. Accessed August 27, 2009 Altman DG, Bland JM. Interaction revisited: the difference between two estimates. BMJ 2003; 326:219 Cazzola M, Di Lorenzo G, Di Perna F, et al. Additive effects of salmeterol and fluticasone or theophylline in COPD. Chest 2000; 118:1576 –1581 Mahler DA, Wire P, Horstman D, et al. Effectiveness of fluticasone propionate and salmeterol combination delivered via the diskus device in the treatment of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2002; 166: 1084 –1091 Calverley P, Pauwels R, Vestbo J, et al. Combined salmeterol and fluticasone in the treatment of chronic obstructive pulmonary disease: a randomized controlled trial. Lancet 2003; 361:449 – 456 Calverley PM, Boonsawat W, Cseke Z, et al. Maintenance therapy with budesonide and formoterol in chronic obstructive pulmonary disease. Eur Respir J 2003; 22:912–919 Dal Negro RW, Pomari C, Tognella S, et al. Salmeterol and fluticasone 50 microg/250 microg bid in combination provides a better long-term control than salmeterol 50 microg bid alone and placebo in COPD patients already treated with theophylline. Pulm Pharmacol Ther 2003; 16:241–246 Hanania NA, Darken P, Horstman D, et al. The efficacy and safety of fluticasone propionate (250 ␮g)/salmeterol (50 ␮g) combined in the discus inhaler for the treatment of COPD. Chest 2003; 124:834 – 843 Szafranski A, Cukier A, Ramirez A, et al. Efficacy and safety of budesonide/formoterol in the management of chronic obstructive pulmonary disease. Eur Respir J 2003; 21:74 – 81 Wouters EFM, Postma DS, Fokkens B, et al. Withdrawal of fluticasone propionate from combined salmeterol/fluticasone

22 23 24

25 26

27 28 29 30 31 32 33 34

38 39

treatment in patients with COPD causes immediate and sustained disease deterioration: a randomized controlled trial. Thorax 2005; 60:480 – 487 O’Donnell DE, Sciurba F, Celli B, et al. Effect of fluticasone propionate/salmeterol on lung hyperinflation and exercise endurance in COPD. Chest 2006; 130:647– 656 Calverley PM, Anderson JA, Celli B, et al. Salmeterol and fluticasone propionate and survival in chronic obstructive pulmonary disease. N Engl J Med 2007; 356:775–789 Kardos P, Wencker M, Glaab T, et al. Impact of salmeterol/ fluticasone propionate versus salmeterol on exacerbations in severe chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2007; 175:144 –149 Ferguson GT, Anzueto A, Fei R, et al. Effect of fluticasone propionate/salmeterol (250/50 mg) or salmeterol (50 mg) on COPD exacerbations. Respir Med 2008; 102:1099 –1108 Tashkin DP, Rennard SI, Martin P, et al. Efficacy and safety of budesonide and formoterol in one pressurized metereddose inhaler in patients with moderate to very severe chronic obstructive pulmonary disease: results of a 6-month randomized clinical trial. Drugs 2008; 68:1975–2000 AstraZeneca. Clinical trial register: D-5899C00001 trial. Available at: http://www.astrazenecaclinicaltrials.com/clinicaltrials/. Accessed August 27, 2009 GlaxoSmithKline. Clinical trial register: SCO40041 trial. Available at: http://www.gsk-clinicalstudyregister.com/. Accessed August 27, 2007 GlaxoSmithKline. Clinical trial register: SCO100250 trial. Available at: http://www.gsk-clinicalstudyregister.com/. Accessed August 27, 2007 GlaxoSmithKline. Clinical trial register: SCO100470 trial. Available at: http://www.gsk-clinicalstudyregister.com/. Accessed August 27, 2007 GlaxoSmithKline. Clinical trial register: SCO104925 trial. Available at: http://www.gsk-clinicalstudyregister.com/. Accessed August 27, 2007 Cazzola M, MacNee W, Martinez FJ, et al. Outcomes for COPD pharmacological trials: from lung function to biomarkers. Eur Respir J 2008; 31:416 – 468 Suissa S, McGhan R, Niewoehner D, et al. Inhaled corticosteroids in chronic obstructive pulmonary disease. Proc Am Thorac Soc 2007; 4:535–542 Wolfe F, Caplan L, Michaud K. Treatment for rheumatoid arthritis and the risk of hospitalization for pneumonia: associations with prednisone, disease modifying antirheumatic drugs, and anti-tumor necrosis factor therapy. Arthritis Rheum 2006; 54:628 – 634 Puhan MA, Bachmann LM, Kleijnen J, et al. Inhaled drugs to reduce exacerbations in patients with chronic obstructive pulmonary disease: a network meta-analysis. BMC Med 2009; 7:2 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; 169:219 –229 Wedzicha JA, Calverley PMA, Seemungal TA, et al. The prevention of chronic obstructive pulmonary disease exacerbations by salmeterol/fluticasone propionate or tiotropium bromide. Am J Respir Crit Care Med 2008; 177:19 –26 Moher D, Cook DJ, Eastwood S, et al. Improving the quality of reports of meta-analyses of randomized controlled trials: the QUOROM statement. Lancet 1999; 354:1896 –1900 Sears M. Long-acting bronchodilators in COPD [editorial]. Chest 2008; 133:1057–1058

1038

Original Research

Safety and Efficacy of Combined Long-Acting β-Agonists and Inhaled Corticosteroids vs Long-Acting β-Agonists Monotherapy for Stable COPD Gustavo J. Rodrigo, José A. Castro-Rodriguez and Vicente Plaza Chest 2009;136; 1029-1038; Prepublished online July 24, 2009; DOI 10.1378/chest.09-0821 This information is current as of November 1, 2009 Updated Information & Services

Updated Information and services, including high-resolution figures, can be found at: http://chestjournal.chestpubs.org/content/136/4/1029.ful l.html

References

This article cites 31 articles, 20 of which can be accessed free at: http://chestjournal.chestpubs.org/content/136/4/102 9.full.html#ref-list-1

Open Access

Freely available online through CHEST open access option

Permissions & Licensing

Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.chestjournal.org/site/misc/reprints.xhtml

Reprints

Information about ordering reprints can be found online: http://www.chestjournal.org/site/misc/reprints.xhtml

Email alerting service

Receive free email alerts when new articles cite this article. Sign up in the box at the top right corner of the online article.

Images in PowerPoint format

Figures that appear in CHEST articles can be downloaded for teaching purposes in PowerPoint slide format. See any online article figure for directions