- Email: [email protected]
Inhaled Corticosteroids vs Placebo for Preventing COPD Exacerbations
CHEST
Original Research COPD
Inhaled Corticosteroids vs Placebo for Preventing COPD Exacerbations A Systematic Review and Metaregression of Randomized Controlled Trials Ritesh Agarwal, MD, DM, FCCP; Ashutosh N. Aggarwal, MD, DM, FCCP; Dheeraj Gupta, MD, DM, FCCP; and Surinder K. Jindal, MD, FCCP
Background: Inhaled corticosteroids (ICS) have been shown to decrease the occurrence of COPD exacerbations. However, the relationship of baseline lung function and reduction of exacerbations with the use of ICS remains unknown. Herein, we perform a metaregression to evaluate the efficacy of ICS in preventing COPD exacerbations. Methods: We searched the PubMed, EmBase, and Cochrane Central Database of Controlled Trials databases (1988-2008) for studies that have reported the efficacy of ICS vs placebo in preventing COPD exacerbations. We pooled the risk ratio (RR) and 95% CIs from individual studies using a random-effects model to assess the exacerbations in the two groups. We also performed a weighted random effects metaregression using baseline FEV1 values. Results: Our search yielded 11 studies (8,164 patients). The use of ICS was associated with reduction in the occurrence of exacerbations (RR, 0.82; 95% CI, 0.73-0.92). There was the presence of significant statistical heterogeneity but no evidence of publication bias. Sensitivity analysis revealed benefit of ICS only in patients with FEV1 , 50% (RR, 0.79; 95% CI, 0.69-0.89) with persistence of statistical heterogeneity. Metaregression showed that the percentage risk reduction in exacerbations with the use of ICS is invariant across the severity of COPD (assessed by FEV1). Conclusion: There is only a modest benefit of ICS in preventing COPD exacerbations, which is not related to the level of baseline lung function on metaregression analysis. The benefits of ICS in preventing COPD exacerbations thus seem to be overstated. CHEST 2010; 137(2):318–325 Abbreviations: ICS 5 inhaled corticosteroids; RR 5 risk ratio
is primarily a smoking-related inflammatory COPD disease of the lungs clinically manifesting as
cough, sputum production, and dyspnea. It is a major cause of mortality and morbidity worldwide and results in significant economic and social burden.1 COPD is also the fourth leading cause of death in the world (sixth in low-income, third in middle-income, and fifth in high-income countries),2 and further Manuscript received June 3, 2009; revision accepted August 17, 2009. Affiliations: From the Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research, Chandigarh, India. Correspondence to: Ritesh Agarwal, MD, DM, FCCP, Assistant Professor, Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research, Sector-12, Chandigarh 160012, India; e-mail: [email protected]
increases in its prevalence and mortality are predicted in the coming decades.3 Apart from smoking cessation and long-term oxygen therapy (in selected patients), no other pharmacologic intervention has been shown to improve survival in patients with COPD. Thus, the aim of current medical therapy in COPD is focused on reduction of symptoms and complications associated with the disease. One major focus of most pharmacological therapies is to prevent exacerbations of COPD as they not only adversely impact health status but are also a source of huge financial burden. © 2010 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.09-1305 Original Research
318
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
In fact, COPD exacerbations requiring hospitalization account for one-half or more of the total medical costs for treating this disease.4,5 Inhaled corticosteroids (ICS) are antiinflammatory agents, and thus have been widely used for the treatment of COPD. The effectiveness of ICS in patients with COPD remains controversial.6,7 The current international guidelines recommend ICS to be used in patients with COPD with an FEV1 value less than 50% predicted and repeated exacerbations, as they have been shown to decrease exacerbations in this subgroup of patients with COPD.1 However, ICS are also associated with numerous adverse effects, including oropharyngeal candidiasis,8 cataracts,9 and fractures.10 In patients with COPD, they have also been shown to significantly increase the risk of pneumonia.11 Given the possible morbidity associated with the use of ICS, it is important to reevaluate the benefit of ICS, especially with its claim of reducing exacerbations. Five systematic reviews and/or metaanalyses have suggested beneficial effects of ICS in preventing exacerbations.12-16 It has also been suggested that larger ICS-related reductions in exacerbations tend to be seen in trials that enrolled patients with lower mean baseline values of FEV1.8 In this study, we perform a systematic review and metaanalysis of randomized controlled trials that have evaluated the efficacy of ICS vs placebo in prevention of COPD exacerbations. We also perform a metaregression of trials to evaluate the risk reduction with every percentage decline in baseline values of FEV1 across studies, apart from the simpler stratified method used by previous analyses to compare the efficacy of ICS in preventing COPD exacerbations across varying severity of COPD.
Materials and Methods
patients at risk for COPD based on accelerated FEV1 decline and studies including patients with both asthma and COPD were excluded) irrespective of the severity of COPD, (2) follow-up duration of the study for at least 1 year, (3) randomized controlled trials comparing ICS vs placebo (studies with combination of ICS plus LABA were excluded) and reporting the exacerbation rates in the two groups, and (4) peer-reviewed fully published studies. Data were recorded on a data extraction form, and discrepancies were resolved by consensus between the authors. The following items were extracted: (1) publication details: citation, title, author(s); (2) details on randomization, blinding, and withdrawals/ dropouts in the study; (3) dosages and devices of ICS used in the various studies; (4) the mean age, gender, current smoking status, and baseline FEV1 of the study participants; and (5) the definition of exacerbations and the rate ratio of exacerbations in the individual studies. The definitions for exacerbation were those used by the authors in the original studies. Trial Quality The methodologic quality of each trial was evaluated using the six-point scale (0 5 worst to 5 5 best) as described by Jadad et al.17 This instrument assesses the adequacy of randomization, blinding, and the handling of withdrawals and dropouts, with a score of one point for each “yes” or zero points for each “no.” One additional point is given if the method to generate the sequence of randomization was described and it was appropriate (table of random numbers, computer generated, and so forth) or the method of double-blinding was described and it was appropriate (identical placebo, active placebo, dummy, and so forth). On the other hand, one point is deducted if the method to generate the sequence of randomization was described and it was inappropriate (patients were allocated alternately, or according to date of birth, hospital number, and so forth) or the study was described as double-blind but the method of blinding was inappropriate (eg, comparison of tablet vs injection with no double dummy). The studies are said to be of low quality if the Jadad score is 2 and high quality if the score is 3.17,18 Statistical Analysis The statistical packages Comprehensive Metaanalysis (version 2.2.046 for MS-Windows; Biostat; Englewood, NJ; http://www. meta-analysis.com) and StatsDirect (StatsDirect version 2.7.2 for MS-Windows; StatsDirect Ltd; Cheshire, UK; http://www. statsdirect.com) were used to perform all statistical analyses. Determination of the Pooled Treatment Effect
Search Strategy and Selection Criteria To identify the studies for inclusion in this review, all the authors independently searched the computer databases PubMed, Cochrane Central Database of Controlled trials, and EmBASE for relevant studies published from 1988 to 2008 using free text terms: inhaled corticosteroids, ICS, chronic obstructive pulmonary disease, and COPD, in various combinations, and limiting the search to English literature. Bibliographies of all selected articles and review articles that included information on the role of inhaled corticosteroids in preventing exacerbations of COPD were reviewed. Data Abstraction The abstracts of the studies were independently reviewed by two authors (RA and ANA), without blinding to study the details. Any disagreement was resolved by discussion between the authors. We included studies that met the following criteria: (1) studies including confirmed cases of COPD (studies recruiting www.chestjournal.org
The results of RR with their 95% CIs from individual studies were pooled using both fixed effects model of Mantel-Haenszel and DerSimonian-Laird random effects model.19,20 However, the results and ensuing discussion have been done only with the random effects model. Assessment of Heterogeneity The impact of heterogeneity on the pooled estimates of the individual outcomes of the metaanalysis was assessed by the I2 test and the Cochran Q statistic. The I2 test measures the extent of inconsistency among the results of the studies, which are interpreted as the approximate proportion of total variation in study estimates that is due to heterogeneity rather than sampling error. An I2 value greater than 50% indicates significant heterogeneity.21 The Cochran test calculates the weighted sum of squared differences between individual study effects and the pooled effect across studies, with the weights being those used in the pooling CHEST / 137 / 2 / FEBRUARY, 2010
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
319
method. The P value level at which heterogeneity should be diagnosed is unclear, given that the Q statistic has low power, and Fleiss have recommended a value of at least 0.1. Assessment of Publication Bias We checked for the presence of publication bias using the Begg funnel plot.23 The funnel plot is a measure of the log RR (in the X-axis) against the standard error of the log RR (in the Y-axis). Each open circle represents each study in the metaanalysis. The line in the center indicates the summary proportion and the other two lines indicate the 95% CI. In the absence of publication bias, the proportion estimates from smaller studies are expected to be scattered above and below the summary estimate, producing a triangular or funnel shape. We also checked for publication bias using two statistical tests: (1) the Egger test,24 which is a test for asymmetry of the funnel plot. This is a test for the Y intercept 5 0 from a linear regression of normalized effect estimate (estimate divided by its standard error) against precision (reciprocal of the standard error of the estimate); and (2) Begg and Mazumdar test,25 which tests the interdependence of variance and effect size using rank correlation method. Sensitivity Analyses We performed sensitivity analyses to examine effect sizes when only the following types of studies were included: studies including patients with mild to moderate severity of COPD (ie, FEV1 . 50%), and studies including patients with severe to very severe COPD (ie, FEV1 50%).1 Metaregression We used unrestricted maximum likelihood weighted linear random-effects metaregression to evaluate variation between studies, to model the rate ratio as a function of FEV1 (percentage predicted), and to graph the response curve between FEV1 and reduction of exacerbations with the use of ICS.26 The dependent variable for the regression was the natural log of each studyspecific rate ratio for exacerbations, weighted by the inverse of its variance. The FEV1 for each study was then treated as a continuous, independent variable. The coefficient of the FEV1 in the regression model estimates the slope of the linear FEV1 doseresponse effect. Solving the regression equation estimates the percentage risk reduction in exacerbation predicted at any given percentage predicted FEV1. An institutional review board clearance was not required for this study as this was a metaanalysis of published studies.
Results Our initial database search retrieved 1,095 citations (Fig 1). Of these, 177 studies were excluded as they did not involve COPD. Finally, 11 studies (8,164 study participants; 4,241 ICS, 3,923 placebo) that met our inclusion criteria were included in the final analysis.27-37 All the studies were prospective, double-blind, randomized, placebo-controlled trials. The Jadad score was greater than three for all the studies, indicating high quality of the individual studies (Table 1). Three trials had a systemic steroid runin phase of 2 weeks.28,30,33 The baseline characteristics
Figure 1. Flow diagram showing the trial selection process for this systematic review. ICS 5 inhaled corticosteroids.
of the study population are shown in Table 2. The study populations in all the trials were preponderantly men with a mean age of 62 years. The FEV1 values were less than 50% in seven of the 11 studies.29,30,32-35,37 The definition of COPD exacerbation varied among trials but in general was the definition used in day-to-day practice (Table 3). The pooled results of from the 11 included studies showed an 18% relative risk reduction (Fig 2) in the occurrence of exacerbations (RR, 0.82; 95% CI, 0.73-0.92) with the use of ICS in patients with COPD (over a mean follow-up period of 2.1 years). However, there was presence of significant heterogeneity with an I2 value of 55% (95% CI, 0-75.6) and Cochran Q statistic of 22.24 (P 5 .01). There was no evidence of publication bias on visual examination of the funnel plot (Fig 3) or on the statistical tests (BeggMazumdar: Kendall t 5 20.054545, P 5 .76; Egger: bias 5 20.550623, P 5 .46). Original Research
320
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
Table 1—Quality of the 11 Trials as Assessed by the Jadad Score Description of Withdrawals Randomization Blinding and Dropouts
Study/Year Renkema et al27/1996 Vestbo et al28/1999 Weir et al29/1999 Burge et al30/2000 Lung Health Study31/2000 Calverley et al32/2003 Calverley et al33/2003 Szafranski et al34/2003 Calverley et al35/2007 Choudhury et al36/2007 Calverley et al37/2008
2 1 1 2 2
2 2 2 2 2
1 1 1 1 1
2 2 1 2 2 2
2 2 2 2 2 2
1 1 1 1 1 1
erbation reduction by ICS; that is, the RR (or the percentage risk reduction) is invariant across the severity of COPD (as assessed by FEV1) included in the trials. The coefficient for the FEV1 in the regression model is the natural logarithm of the change in percentage risk reduction for each percentage predicted increase in FEV1. Discussion
Sensitivity Analysis and Metaregression Sensitivity analysis was performed based on the severity of COPD (Table 4). The analysis revealed that the treatment effect was derived largely from studies that included patients with severe to very severe COPD (FEV1 less than 50% predicted). Excluding patients with mild to moderate COPD yielded an estimate effect almost similar to the overall metaanalysis with similar degrees of heterogeneity. In fact, there was no heterogeneity in the summary statistic of patients with mild to moderate COPD (Table 4). The results of the linear regression model (Table 5) confirmed the visual impression of the scatter plot (Fig 4). There is no linear response of FEV1 on exac-
The results of this systematic review suggest that there is a modest relative risk reduction in COPD exacerbations with the use of ICS. In the sensitivity analysis, the beneficial effect was noted only in patients with FEV1 values less than 50%. There was presence of significant statistical heterogeneity with no evidence of publication bias. A metaregression analysis also failed to demonstrate a linear relationship between FEV1 values and decline in COPD exacerbations with the use of ICS. The current systematic review is different from previous reviews in many aspects. We added metaregression to the simple metaanalytic technique to explore among-study differences and strengthen the validity of our analysis. Metaregression is a method of combining results from multiple studies using basic metaanalysis in combination with weighted regression analysis to predict study effects using study-level covariates.38 Further, we used a randomeffects metaregression, which takes into account both within-trial variances of treatment effects and
Table 2—Baseline Characteristics of the Study Population in Various Studies Mean Age, y
Study/Year Renkema et al27/1996 Vestbo et al28/1999 Weir et al29/1999
Male Current Sex, % Smoking, %
No. ICS
No. Placebo
Study Duration, y
ICS, Dose in mg
Device Used pMDI with spacer DPI, Turbuhaler
Baseline FEV1, % Predicted
55
…
40.5
21
24
2
Budesonide, 800
59
60.4
76.6
145
145
3
Budesonide, 800
59
74.5
38.7
49
49
2
Beclomethasone, 2,000
Burge et al30/2000
63.8
74.5
37.5
376
375
3
Fluticasone, 1,000
Lung Health Study31/2000 Calverley et al32/2003 Calverley et al33/2003 Szafranski et al34/2003 Calverley et al35/2007 Choudhury et al36/2007 Calverley et al37/2008 Overall
56.3
63
90.2
559
557
4.5
63.5
72.5
50
358
361
1
Fluticasone, 1,000
pMDI with spacer pMDI with spacer pMDI with spacer DPI, Accuhaler
64.5
74.5
34.5
257
256
1
Budesonide, 800
DPI, Turbuhaler
36.0
64.5
81.5
35
198
205
1
Budesonide, 800
DPI, Turbuhaler
36.5
65
75.5
43
1534
1524
3
Fluticasone, 1,000
DPI, Accuhaler
40.4
67.5
52
38.1
128
132
1
Fluticasone, 1,000
DPI, Accuhaler
51.3
65
68.5
…
616
295
1
Mometasone, 800
DPI, Accuhaler
42.3
62.1
69.7
48.4
4241
3923
…
…
49.2
2.1
Triamcinolone, 1,200
63.5 78.9 40.7 46.3 64.2 40.8
ICS 5 inhaled corticosteroid. www.chestjournal.org
CHEST / 137 / 2 / FEBRUARY, 2010
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
321
Table 3—Definition of Exacerbation Used in the Various Studies Included in the Metaanalysis Study/Year
Definition of Exacerbation
Renkema et al27/1996 Vestbo et al28/1999 Weir et al29/1999 Burge et al30/2000 Lung Health Study31/2000 Calverley et al32/2003 Calverley et al33/2003 Szafranski et al34/2003 Calverley et al35/2007 Choudhury et al36/2007 Calverley et al37/2008
Increased complaints of dyspnea and/or cough and/or sputum production, with or without fever Affirmative answer to the question, “Have you since your last visit experienced more cough and phlegm than usual?” Not defined Worsening of respiratory symptoms that required treatment with oral corticosteroids or antibiotics, or both, as judged by the general practitioner Not defined Worsening of COPD symptoms that required treatment with antibiotics, oral corticosteroids, or both Worsening of symptoms requiring medical intervention (oral antibiotics and/or corticosteroids or hospitalization) Use of oral steroids and/or antibiotics and/or hospitalization due to respiratory symptoms Symptomatic deterioration requiring treatment with antibiotic agents, systemic corticosteroids, hospitalization, or a combination of these Use of oral steroids and/or antibiotics and/or hospitalization due to respiratory symptoms Clinically significant worsening of COPD symptoms requiring treatment with antibiotics and/or systemic steroids
the residual between-trial heterogeneity that is not explained by the covariates in the regression.26 We focused primarily on COPD exacerbations rather than other outcomes as other outcomes have already been properly evaluated and published.39-41 To strengthen the analysis, we chose a minimum arbitrary duration of the study of at least 1 year so that the treatment is given for an appropriate period, and exacerbations may be properly evaluated. Although previous reviews have included studies with duration of 6 months, we excluded these studies as authors themselves have acknowledged difficulty in evaluating exacerbations in such brief periods.42 Unlike other reviews, we did not include data that have not been fully published, as inclusion of unpublished data in metaanalyses remains controversial because of the lack of quality control and associated peer review of
these data.43 Also, we included only studies that evaluated ICS alone compared with placebo as combination therapy of ICS with other drugs precludes the evaluation of the efficacy of ICS alone in preventing COPD exacerbations. What are the implications of this study in clinical practice? The results of our study show that the perceived benefits of ICS (in preventing COPD exacerbations) in patients with severe to very severe COPD seem to be overstated. We did observe a modest benefit of ICS in preventing COPD exacerbations in patients with FEV1 , 50%. However, this finding was not further substantiated on a metaregression analysis wherein we did not observe a significant effect of baseline values of FEV1 (percentage predicted) on the benefits from ICS compared with placebo. Thus, ICS is likely to have only modest benefits in
Figure 2. Forest plot showing the occurrence of exacerbations in patients receiving ICS vs placebo. There was 18% relative risk reduction in the occurrence of exacerbations as determined by the randomeffects model. See Figure 1 legend for expansion of abbreviation. Original Research
322
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
Figure 3. Funnel plot comparing log RR vs the SE of log RR. Open circles represent trials included in the metaanalysis. The line in the center indicates the summary log RR. The other lines represent the 95% CIs. There is no evidence of publication bias. RR 5 risk ratio; SE 5 standard error.
preventing COPD exacerbations, if at all, and should be judiciously used in patients with COPD keeping in mind the risk-benefit ratio. These results could also be potentially explained by a low statistical power of our metaanalysis because of the inclusion of only 11 trials. Further, the metaregression analysis is dependent on the mean FEV1 values, and does not take into account the standard deviation of FEV1. Hence, extreme caution should be exercised in the interpretation of these results in that they cannot be extended to daily practice. Although based on the results of this study, an individual patient data metaanalysis should be performed to properly define the role of ICS in preventing COPD exacerbations. Two recent metaanalyses suggest a higher risk of serious pneumonia in patients receiving ICS with no
improvement in survival with the use of ICS.40,41 In fact, the risk of pneumonia was increased in patients with the lowest baseline FEV1,40 meaning that in situations where ICS is indicated for prevention of exacerbations, its use is likely to be associated with higher risk of pneumonia. Finally, in patients with COPD, higher doses of ICS are required for clinical efficacy compared to asthma, and this also may increase the propensity for steroid related side effects. It is widely believed that patients with COPD with lower levels of lung function are prone to develop more frequent and more severe exacerbations.44,45 In one study, the hazards for the time to first exacerbation was greater among patients with an FEV1 less than 50% predicted than in patients with a higher FEV1.46 A retrospective analysis of the ISOLDE study suggested that fluticasone caused a greater decline in exacerbation rates in patients with worse lung function compared with better lung function.47 However, a recent study showed that although baseline FEV1 was an independent predictor of exacerbation, the relationship was not particularly strong.5,48 It may also be possible that the benefit of ICS in COPD is not dependent on baseline FEV1 values but on certain other parameters, such as patients with COPD who suffer one or more exacerbations (irrespective of FEV1), and this fact should also be kept in mind in future COPD trials. The other limitations of our study include the presence of clinical heterogeneity in that the different studies have included different doses of ICS and varying severity of patients with COPD. We also included RR from individual studies rather than the number of patients with COPD who suffered an exacerbation. In analyzing exacerbation, the length of follow-up in clinical trials should also be considered. Most trials have estimated the mean rate
Table 4—Sensitivity Analysis on the Reduction of Exacerbations With the Use of ICS Stratified by the Severity of COPD Severity of COPD
Included Studies
RR (95% CI)
I2
Cochran Q Statistic (P Value)
Mild to moderate (FEV1 . 50%)
Renkema et al27 Copenhagen study28 Inhaled Steroids in Obstructive Lung Disease in Europe (ISOLDE) study30 Choudhury et al36 Weir et al29 Lung Health study31 Trial of Inhaled Steroids and Long-acting 2 Agonist (TRIAL) study32 Calverley et al33 Szafranski et al34 Towards a Revolution in COPD Health (TORCH) study35 Calverley et al37
1.026 (0.859-1.226), fixed 1.026 (0.859-1.226), random
0
2.91 (.41)
0.824 (0.775-0.876), fixed 0.785 (0.692-0.891), random …
57.4
14.09 (.03)
…
…
Severe to very severe (FEV1 , 50%)
RR 5 risk ratio. See Table 2 for expansion of other abbreviations. www.chestjournal.org
CHEST / 137 / 2 / FEBRUARY, 2010
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
323
Acknowledgments
Table 5—Metaregression of Reduction of Exacerbations With the Use of ICS and FEV1 Overall
Coefficient
Standard Error
95% CI
P Value
FEV1 Constant
0.005702 20.45984
0.008229 0.367822
20.01043-0.021831 21.18076-0.261076
.36 …
The variance in the FEV1 coefficient between the individual studies t2 5 0.02239194. See Table 2 for expansion of abbreviation.
of exacerbation for the group by using an average that was unweighted for the follow-up duration, and could potentially lead to an exaggerated beneficial effect of ICS.6,49 Thus the benefit of ICS may be even less than what has been suggested from this analysis. Finally, metaregression analysis has its own limitations in that it is an observational association across trials, and does not have the benefit of randomization. Thus, as mentioned previously, the results of our study need to be confirmed by an individual patient data metaanalysis. In conclusion, the results of our study suggest a modest benefit of ICS in preventing COPD exacerbations. In a subgroup analysis, the benefit was observed only in patients with FEV1 values , 50% predicted. However, on a metaregression analysis, there was no demonstrable linear relationship between efficacy of ICS and prevention of COPD exacerbations that is related to the level of stable lung function. Thus the role of ICS in preventing COPD exacerbations needs reappraisal.
Figure 4. Metaregression of the effect of FEV1 on the log RR for the risk of exacerbations associated with the use of inhaled corticosteroids. Each circle represents one study. Circle size reflects the weight of the study in the metaregression based on inverse variances. The line represents metaregression with an unrestricted maximum effects likelihood random-effects model (slope 5 0.0057, intercept 5 20.4598, Q 5 10.06, P 5 .36). The numbers represent the citation number of the studies included in the metaregression. See Figure 3 legend for expansion of abbreviation.
Author contributions: Dr Agarwal: contributed to literature search, data analysis, and drafting of the manuscript. Dr Aggarwal: contributed to literature search, data analysis, and drafting of the manuscript. Dr Gupta: contributed to literature search and drafting of the manuscript. Dr Jindal: contributed to literature search and drafting of the manuscript. Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.
References 1. Global strategy for the diagnosis, management and prevention of chronic obstructive pulmonary disease. (Updated 2008). http://www.goldcopd.com. Accessed September 2, 2009. 2. World Health Organization. The Global Burden of Disease— 2004 Update. Geneva, Switzerland: WHO Press; 2008. 3. Lopez AD, Shibuya K, Rao C, et al. Chronic obstructive pulmonary disease: current burden and future projections. Eur Respir J. 2006;27(2):397-412. 4. Strassels SA, Smith DH, Sullivan SD, Mahajan PS. The costs of treating COPD in the United States. Chest. 2001;119(2): 344-352. 5. Niewoehner DE. Relation of chronic obstructive pulmonary disease exacerbations to FEV(1)—an intricate tango. Respiration. 2009;77(2):229-235. 6. Suissa S, McGhan R, Niewoehner D, Make B. Inhaled corticosteroids in chronic obstructive pulmonary disease. Proc Am Thorac Soc. 2007;4(7):535-542. 7. Cazzola M, Matera MG. To add, or not to add an inhaled corticosteroid in moderate COPD: that is the question. Chest. 2008;134(2):223-225. 8. Sin DD, McAlister FA, Man SF, Anthonisen NR. Contemporary management of chronic obstructive pulmonary disease: scientific review. JAMA. 2003;290(17):2301-2312. 9. Ernst P, Baltzan M, Deschênes J, Suissa S. Low-dose inhaled and nasal corticosteroid use and the risk of cataracts. Eur Respir J. 2006;27(6):1168-1174. 10. Suissa S, Baltzan M, Kremer R, Ernst P. Inhaled and nasal corticosteroid use and the risk of fracture. Am J Respir Crit Care Med. 2004;169(1):83-88. 11. 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. 12. Alsaeedi A, Sin DD, McAlister FA. The effects of inhaled corticosteroids in chronic obstructive pulmonary disease: a systematic review of randomized placebo-controlled trials. Am J Med. 2002;113(1):59-65. 13. Wilt TJ, Niewoehner D, Kim C, et al. Use of spirometry for case finding, diagnosis, and management of chronic obstructive pulmonary disease (COPD). Evid Rep Technol Assess (Summ). 2005;121(121):1-7. 14. Gartlehner G, Hansen RA, Carson SS, Lohr KN. Efficacy and safety of inhaled corticosteroids in patients with COPD: a systematic review and meta-analysis of health outcomes. Ann Fam Med. 2006;4(3):253-262. 15. Wilt TJ, Niewoehner D, MacDonald R, Kane RL. Management of stable chronic obstructive pulmonary disease: a systematic review for a clinical practice guideline. Ann Intern Med. 2007; 147(9):639-653. 16. Puhan MA, Bachmann LM, Kleijnen J, Ter Riet G, Kessels AG. Inhaled drugs to reduce exacerbations in patients with Original Research
324
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
17. 18.
19. 20. 21. 22. 23. 24. 25. 26. 27. 28.
29.
30.
31. 32.
33.
chronic obstructive pulmonary disease: a network meta-analysis. BMC Med. 2009;7:2. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996;17(1):1-12. Kjaergard LL, Villumsen J, Gluud C. Reported methodologic quality and discrepancies between large and small randomized trials in meta-analyses. Ann Intern Med. 2001;135(11): 982-989. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst. 1959;22(4):719-748. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7(3):177-188. Armitage P, Berry G, Matthews JNS. Analysing Means and Proportions. Statistical Methods in Medical Research. Oxford: Blackwell Science; 2002:83-146. Fleiss JL. Analysis of data from multiclinic trials. Control Clin Trials. 1986;7(4):267-275. Dear K, Begg C. An approach to assessing publication bias prior to performing a meta-analysis. Stat Sci. 1992;7(2):237-245. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629-634. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50(4): 1088-1101. Thompson SG, Higgins JP. How should metaregression analyses be undertaken and interpreted? Stat Med. 2002; 21(11):1559-1573. Renkema TE, Schouten JP, Koëter GH, Postma DS. Effects of long-term treatment with corticosteroids in COPD. Chest. 1996;109(5):1156-1162. 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. Weir DC, Bale GA, Bright P, Sherwood Burge P. A doubleblind placebo-controlled study of the effect of inhaled beclomethasone dipropionate for 2 years in patients with nonasthmatic chronic obstructive pulmonary disease. Clin Exp Allergy. 1999;29(Suppl 2):125-128. 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. Lung Health Study Research Group. Effect of inhaled triamcinolone on the decline in pulmonary function in chronic obstructive pulmonary disease. N Engl J Med. 2000;343(26):1902-1909. 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. 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.
www.chestjournal.org
34. 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. 35. Calverley PM, 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. 36. 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. 37. Calverley PM, Rennard S, Nelson HS, et al. One-year treatment with mometasone furoate in chronic obstructive pulmonary disease. Respir Res. 2008;9:73. 38. Sutton AJ, Higgins JP. Recent developments in meta-analysis. Stat Med. 2008;27(5):625-650. 39. Soriano JB, Sin DD, Zhang X, et al. A pooled analysis of FEV1 decline in COPD patients randomized to inhaled corticosteroids or placebo. Chest. 2007;131(3):682-689. 40. Drummond MB, Dasenbrook EC, Pitz MW, Murphy DJ, Fan E. Inhaled corticosteroids in patients with stable chronic obstructive pulmonary disease: a systematic review and metaanalysis. JAMA. 2008;300(20):2407-2416. 41. 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. 42. 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. 43. Cook DJ, Guyatt GH, Ryan G, et al. Should unpublished data be included in meta-analyses? Current convictions and controversies. JAMA. 1993;269:2749-2753. 44. Garcia-Aymerich J, Monsó E, Marrades RM, et al; EFRAM Investigators. Risk factors for hospitalization for a chronic obstructive pulmonary disease exacerbation. EFRAM study. Am J Respir Crit Care Med. 2001;164(6):1002-1007. 45. Garcia-Aymerich J, Farrero E, Félez MA, Izquierdo J, Marrades RM, Antó JM; Estudi del Factors de Risc d’Agudització de la MPOC investigators. Risk factors of readmission to hospital for a COPD exacerbation: a prospective study. Thorax. 2003;58(2):100-105. 46. 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. 47. Jones PW, Willits LR, Burge PS, Calverley PM; Inhaled Steroids in Obstructive Lung Disease in Europe study investigators. Disease severity and the effect of fluticasone propionate on chronic obstructive pulmonary disease exacerbations. Eur Respir J. 2003;21(1):68-73. 48. Niewoehner DE, Lokhnygina Y, Rice K, et al. Risk indexes for exacerbations and hospitalizations due to COPD. Chest. 2007;131(1):20-28. 49. Suissa S. Statistical treatment of exacerbations in therapeutic trials of chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2006;173(8):842-846.
CHEST / 137 / 2 / FEBRUARY, 2010
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
325
Original Research COPD
Inhaled Corticosteroids vs Placebo for Preventing COPD Exacerbations A Systematic Review and Metaregression of Randomized Controlled Trials Ritesh Agarwal, MD, DM, FCCP; Ashutosh N. Aggarwal, MD, DM, FCCP; Dheeraj Gupta, MD, DM, FCCP; and Surinder K. Jindal, MD, FCCP
Background: Inhaled corticosteroids (ICS) have been shown to decrease the occurrence of COPD exacerbations. However, the relationship of baseline lung function and reduction of exacerbations with the use of ICS remains unknown. Herein, we perform a metaregression to evaluate the efficacy of ICS in preventing COPD exacerbations. Methods: We searched the PubMed, EmBase, and Cochrane Central Database of Controlled Trials databases (1988-2008) for studies that have reported the efficacy of ICS vs placebo in preventing COPD exacerbations. We pooled the risk ratio (RR) and 95% CIs from individual studies using a random-effects model to assess the exacerbations in the two groups. We also performed a weighted random effects metaregression using baseline FEV1 values. Results: Our search yielded 11 studies (8,164 patients). The use of ICS was associated with reduction in the occurrence of exacerbations (RR, 0.82; 95% CI, 0.73-0.92). There was the presence of significant statistical heterogeneity but no evidence of publication bias. Sensitivity analysis revealed benefit of ICS only in patients with FEV1 , 50% (RR, 0.79; 95% CI, 0.69-0.89) with persistence of statistical heterogeneity. Metaregression showed that the percentage risk reduction in exacerbations with the use of ICS is invariant across the severity of COPD (assessed by FEV1). Conclusion: There is only a modest benefit of ICS in preventing COPD exacerbations, which is not related to the level of baseline lung function on metaregression analysis. The benefits of ICS in preventing COPD exacerbations thus seem to be overstated. CHEST 2010; 137(2):318–325 Abbreviations: ICS 5 inhaled corticosteroids; RR 5 risk ratio
is primarily a smoking-related inflammatory COPD disease of the lungs clinically manifesting as
cough, sputum production, and dyspnea. It is a major cause of mortality and morbidity worldwide and results in significant economic and social burden.1 COPD is also the fourth leading cause of death in the world (sixth in low-income, third in middle-income, and fifth in high-income countries),2 and further Manuscript received June 3, 2009; revision accepted August 17, 2009. Affiliations: From the Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research, Chandigarh, India. Correspondence to: Ritesh Agarwal, MD, DM, FCCP, Assistant Professor, Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research, Sector-12, Chandigarh 160012, India; e-mail: [email protected]
increases in its prevalence and mortality are predicted in the coming decades.3 Apart from smoking cessation and long-term oxygen therapy (in selected patients), no other pharmacologic intervention has been shown to improve survival in patients with COPD. Thus, the aim of current medical therapy in COPD is focused on reduction of symptoms and complications associated with the disease. One major focus of most pharmacological therapies is to prevent exacerbations of COPD as they not only adversely impact health status but are also a source of huge financial burden. © 2010 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.09-1305 Original Research
318
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
In fact, COPD exacerbations requiring hospitalization account for one-half or more of the total medical costs for treating this disease.4,5 Inhaled corticosteroids (ICS) are antiinflammatory agents, and thus have been widely used for the treatment of COPD. The effectiveness of ICS in patients with COPD remains controversial.6,7 The current international guidelines recommend ICS to be used in patients with COPD with an FEV1 value less than 50% predicted and repeated exacerbations, as they have been shown to decrease exacerbations in this subgroup of patients with COPD.1 However, ICS are also associated with numerous adverse effects, including oropharyngeal candidiasis,8 cataracts,9 and fractures.10 In patients with COPD, they have also been shown to significantly increase the risk of pneumonia.11 Given the possible morbidity associated with the use of ICS, it is important to reevaluate the benefit of ICS, especially with its claim of reducing exacerbations. Five systematic reviews and/or metaanalyses have suggested beneficial effects of ICS in preventing exacerbations.12-16 It has also been suggested that larger ICS-related reductions in exacerbations tend to be seen in trials that enrolled patients with lower mean baseline values of FEV1.8 In this study, we perform a systematic review and metaanalysis of randomized controlled trials that have evaluated the efficacy of ICS vs placebo in prevention of COPD exacerbations. We also perform a metaregression of trials to evaluate the risk reduction with every percentage decline in baseline values of FEV1 across studies, apart from the simpler stratified method used by previous analyses to compare the efficacy of ICS in preventing COPD exacerbations across varying severity of COPD.
Materials and Methods
patients at risk for COPD based on accelerated FEV1 decline and studies including patients with both asthma and COPD were excluded) irrespective of the severity of COPD, (2) follow-up duration of the study for at least 1 year, (3) randomized controlled trials comparing ICS vs placebo (studies with combination of ICS plus LABA were excluded) and reporting the exacerbation rates in the two groups, and (4) peer-reviewed fully published studies. Data were recorded on a data extraction form, and discrepancies were resolved by consensus between the authors. The following items were extracted: (1) publication details: citation, title, author(s); (2) details on randomization, blinding, and withdrawals/ dropouts in the study; (3) dosages and devices of ICS used in the various studies; (4) the mean age, gender, current smoking status, and baseline FEV1 of the study participants; and (5) the definition of exacerbations and the rate ratio of exacerbations in the individual studies. The definitions for exacerbation were those used by the authors in the original studies. Trial Quality The methodologic quality of each trial was evaluated using the six-point scale (0 5 worst to 5 5 best) as described by Jadad et al.17 This instrument assesses the adequacy of randomization, blinding, and the handling of withdrawals and dropouts, with a score of one point for each “yes” or zero points for each “no.” One additional point is given if the method to generate the sequence of randomization was described and it was appropriate (table of random numbers, computer generated, and so forth) or the method of double-blinding was described and it was appropriate (identical placebo, active placebo, dummy, and so forth). On the other hand, one point is deducted if the method to generate the sequence of randomization was described and it was inappropriate (patients were allocated alternately, or according to date of birth, hospital number, and so forth) or the study was described as double-blind but the method of blinding was inappropriate (eg, comparison of tablet vs injection with no double dummy). The studies are said to be of low quality if the Jadad score is 2 and high quality if the score is 3.17,18 Statistical Analysis The statistical packages Comprehensive Metaanalysis (version 2.2.046 for MS-Windows; Biostat; Englewood, NJ; http://www. meta-analysis.com) and StatsDirect (StatsDirect version 2.7.2 for MS-Windows; StatsDirect Ltd; Cheshire, UK; http://www. statsdirect.com) were used to perform all statistical analyses. Determination of the Pooled Treatment Effect
Search Strategy and Selection Criteria To identify the studies for inclusion in this review, all the authors independently searched the computer databases PubMed, Cochrane Central Database of Controlled trials, and EmBASE for relevant studies published from 1988 to 2008 using free text terms: inhaled corticosteroids, ICS, chronic obstructive pulmonary disease, and COPD, in various combinations, and limiting the search to English literature. Bibliographies of all selected articles and review articles that included information on the role of inhaled corticosteroids in preventing exacerbations of COPD were reviewed. Data Abstraction The abstracts of the studies were independently reviewed by two authors (RA and ANA), without blinding to study the details. Any disagreement was resolved by discussion between the authors. We included studies that met the following criteria: (1) studies including confirmed cases of COPD (studies recruiting www.chestjournal.org
The results of RR with their 95% CIs from individual studies were pooled using both fixed effects model of Mantel-Haenszel and DerSimonian-Laird random effects model.19,20 However, the results and ensuing discussion have been done only with the random effects model. Assessment of Heterogeneity The impact of heterogeneity on the pooled estimates of the individual outcomes of the metaanalysis was assessed by the I2 test and the Cochran Q statistic. The I2 test measures the extent of inconsistency among the results of the studies, which are interpreted as the approximate proportion of total variation in study estimates that is due to heterogeneity rather than sampling error. An I2 value greater than 50% indicates significant heterogeneity.21 The Cochran test calculates the weighted sum of squared differences between individual study effects and the pooled effect across studies, with the weights being those used in the pooling CHEST / 137 / 2 / FEBRUARY, 2010
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
319
method. The P value level at which heterogeneity should be diagnosed is unclear, given that the Q statistic has low power, and Fleiss have recommended a value of at least 0.1. Assessment of Publication Bias We checked for the presence of publication bias using the Begg funnel plot.23 The funnel plot is a measure of the log RR (in the X-axis) against the standard error of the log RR (in the Y-axis). Each open circle represents each study in the metaanalysis. The line in the center indicates the summary proportion and the other two lines indicate the 95% CI. In the absence of publication bias, the proportion estimates from smaller studies are expected to be scattered above and below the summary estimate, producing a triangular or funnel shape. We also checked for publication bias using two statistical tests: (1) the Egger test,24 which is a test for asymmetry of the funnel plot. This is a test for the Y intercept 5 0 from a linear regression of normalized effect estimate (estimate divided by its standard error) against precision (reciprocal of the standard error of the estimate); and (2) Begg and Mazumdar test,25 which tests the interdependence of variance and effect size using rank correlation method. Sensitivity Analyses We performed sensitivity analyses to examine effect sizes when only the following types of studies were included: studies including patients with mild to moderate severity of COPD (ie, FEV1 . 50%), and studies including patients with severe to very severe COPD (ie, FEV1 50%).1 Metaregression We used unrestricted maximum likelihood weighted linear random-effects metaregression to evaluate variation between studies, to model the rate ratio as a function of FEV1 (percentage predicted), and to graph the response curve between FEV1 and reduction of exacerbations with the use of ICS.26 The dependent variable for the regression was the natural log of each studyspecific rate ratio for exacerbations, weighted by the inverse of its variance. The FEV1 for each study was then treated as a continuous, independent variable. The coefficient of the FEV1 in the regression model estimates the slope of the linear FEV1 doseresponse effect. Solving the regression equation estimates the percentage risk reduction in exacerbation predicted at any given percentage predicted FEV1. An institutional review board clearance was not required for this study as this was a metaanalysis of published studies.
Results Our initial database search retrieved 1,095 citations (Fig 1). Of these, 177 studies were excluded as they did not involve COPD. Finally, 11 studies (8,164 study participants; 4,241 ICS, 3,923 placebo) that met our inclusion criteria were included in the final analysis.27-37 All the studies were prospective, double-blind, randomized, placebo-controlled trials. The Jadad score was greater than three for all the studies, indicating high quality of the individual studies (Table 1). Three trials had a systemic steroid runin phase of 2 weeks.28,30,33 The baseline characteristics
Figure 1. Flow diagram showing the trial selection process for this systematic review. ICS 5 inhaled corticosteroids.
of the study population are shown in Table 2. The study populations in all the trials were preponderantly men with a mean age of 62 years. The FEV1 values were less than 50% in seven of the 11 studies.29,30,32-35,37 The definition of COPD exacerbation varied among trials but in general was the definition used in day-to-day practice (Table 3). The pooled results of from the 11 included studies showed an 18% relative risk reduction (Fig 2) in the occurrence of exacerbations (RR, 0.82; 95% CI, 0.73-0.92) with the use of ICS in patients with COPD (over a mean follow-up period of 2.1 years). However, there was presence of significant heterogeneity with an I2 value of 55% (95% CI, 0-75.6) and Cochran Q statistic of 22.24 (P 5 .01). There was no evidence of publication bias on visual examination of the funnel plot (Fig 3) or on the statistical tests (BeggMazumdar: Kendall t 5 20.054545, P 5 .76; Egger: bias 5 20.550623, P 5 .46). Original Research
320
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
Table 1—Quality of the 11 Trials as Assessed by the Jadad Score Description of Withdrawals Randomization Blinding and Dropouts
Study/Year Renkema et al27/1996 Vestbo et al28/1999 Weir et al29/1999 Burge et al30/2000 Lung Health Study31/2000 Calverley et al32/2003 Calverley et al33/2003 Szafranski et al34/2003 Calverley et al35/2007 Choudhury et al36/2007 Calverley et al37/2008
2 1 1 2 2
2 2 2 2 2
1 1 1 1 1
2 2 1 2 2 2
2 2 2 2 2 2
1 1 1 1 1 1
erbation reduction by ICS; that is, the RR (or the percentage risk reduction) is invariant across the severity of COPD (as assessed by FEV1) included in the trials. The coefficient for the FEV1 in the regression model is the natural logarithm of the change in percentage risk reduction for each percentage predicted increase in FEV1. Discussion
Sensitivity Analysis and Metaregression Sensitivity analysis was performed based on the severity of COPD (Table 4). The analysis revealed that the treatment effect was derived largely from studies that included patients with severe to very severe COPD (FEV1 less than 50% predicted). Excluding patients with mild to moderate COPD yielded an estimate effect almost similar to the overall metaanalysis with similar degrees of heterogeneity. In fact, there was no heterogeneity in the summary statistic of patients with mild to moderate COPD (Table 4). The results of the linear regression model (Table 5) confirmed the visual impression of the scatter plot (Fig 4). There is no linear response of FEV1 on exac-
The results of this systematic review suggest that there is a modest relative risk reduction in COPD exacerbations with the use of ICS. In the sensitivity analysis, the beneficial effect was noted only in patients with FEV1 values less than 50%. There was presence of significant statistical heterogeneity with no evidence of publication bias. A metaregression analysis also failed to demonstrate a linear relationship between FEV1 values and decline in COPD exacerbations with the use of ICS. The current systematic review is different from previous reviews in many aspects. We added metaregression to the simple metaanalytic technique to explore among-study differences and strengthen the validity of our analysis. Metaregression is a method of combining results from multiple studies using basic metaanalysis in combination with weighted regression analysis to predict study effects using study-level covariates.38 Further, we used a randomeffects metaregression, which takes into account both within-trial variances of treatment effects and
Table 2—Baseline Characteristics of the Study Population in Various Studies Mean Age, y
Study/Year Renkema et al27/1996 Vestbo et al28/1999 Weir et al29/1999
Male Current Sex, % Smoking, %
No. ICS
No. Placebo
Study Duration, y
ICS, Dose in mg
Device Used pMDI with spacer DPI, Turbuhaler
Baseline FEV1, % Predicted
55
…
40.5
21
24
2
Budesonide, 800
59
60.4
76.6
145
145
3
Budesonide, 800
59
74.5
38.7
49
49
2
Beclomethasone, 2,000
Burge et al30/2000
63.8
74.5
37.5
376
375
3
Fluticasone, 1,000
Lung Health Study31/2000 Calverley et al32/2003 Calverley et al33/2003 Szafranski et al34/2003 Calverley et al35/2007 Choudhury et al36/2007 Calverley et al37/2008 Overall
56.3
63
90.2
559
557
4.5
63.5
72.5
50
358
361
1
Fluticasone, 1,000
pMDI with spacer pMDI with spacer pMDI with spacer DPI, Accuhaler
64.5
74.5
34.5
257
256
1
Budesonide, 800
DPI, Turbuhaler
36.0
64.5
81.5
35
198
205
1
Budesonide, 800
DPI, Turbuhaler
36.5
65
75.5
43
1534
1524
3
Fluticasone, 1,000
DPI, Accuhaler
40.4
67.5
52
38.1
128
132
1
Fluticasone, 1,000
DPI, Accuhaler
51.3
65
68.5
…
616
295
1
Mometasone, 800
DPI, Accuhaler
42.3
62.1
69.7
48.4
4241
3923
…
…
49.2
2.1
Triamcinolone, 1,200
63.5 78.9 40.7 46.3 64.2 40.8
ICS 5 inhaled corticosteroid. www.chestjournal.org
CHEST / 137 / 2 / FEBRUARY, 2010
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
321
Table 3—Definition of Exacerbation Used in the Various Studies Included in the Metaanalysis Study/Year
Definition of Exacerbation
Renkema et al27/1996 Vestbo et al28/1999 Weir et al29/1999 Burge et al30/2000 Lung Health Study31/2000 Calverley et al32/2003 Calverley et al33/2003 Szafranski et al34/2003 Calverley et al35/2007 Choudhury et al36/2007 Calverley et al37/2008
Increased complaints of dyspnea and/or cough and/or sputum production, with or without fever Affirmative answer to the question, “Have you since your last visit experienced more cough and phlegm than usual?” Not defined Worsening of respiratory symptoms that required treatment with oral corticosteroids or antibiotics, or both, as judged by the general practitioner Not defined Worsening of COPD symptoms that required treatment with antibiotics, oral corticosteroids, or both Worsening of symptoms requiring medical intervention (oral antibiotics and/or corticosteroids or hospitalization) Use of oral steroids and/or antibiotics and/or hospitalization due to respiratory symptoms Symptomatic deterioration requiring treatment with antibiotic agents, systemic corticosteroids, hospitalization, or a combination of these Use of oral steroids and/or antibiotics and/or hospitalization due to respiratory symptoms Clinically significant worsening of COPD symptoms requiring treatment with antibiotics and/or systemic steroids
the residual between-trial heterogeneity that is not explained by the covariates in the regression.26 We focused primarily on COPD exacerbations rather than other outcomes as other outcomes have already been properly evaluated and published.39-41 To strengthen the analysis, we chose a minimum arbitrary duration of the study of at least 1 year so that the treatment is given for an appropriate period, and exacerbations may be properly evaluated. Although previous reviews have included studies with duration of 6 months, we excluded these studies as authors themselves have acknowledged difficulty in evaluating exacerbations in such brief periods.42 Unlike other reviews, we did not include data that have not been fully published, as inclusion of unpublished data in metaanalyses remains controversial because of the lack of quality control and associated peer review of
these data.43 Also, we included only studies that evaluated ICS alone compared with placebo as combination therapy of ICS with other drugs precludes the evaluation of the efficacy of ICS alone in preventing COPD exacerbations. What are the implications of this study in clinical practice? The results of our study show that the perceived benefits of ICS (in preventing COPD exacerbations) in patients with severe to very severe COPD seem to be overstated. We did observe a modest benefit of ICS in preventing COPD exacerbations in patients with FEV1 , 50%. However, this finding was not further substantiated on a metaregression analysis wherein we did not observe a significant effect of baseline values of FEV1 (percentage predicted) on the benefits from ICS compared with placebo. Thus, ICS is likely to have only modest benefits in
Figure 2. Forest plot showing the occurrence of exacerbations in patients receiving ICS vs placebo. There was 18% relative risk reduction in the occurrence of exacerbations as determined by the randomeffects model. See Figure 1 legend for expansion of abbreviation. Original Research
322
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
Figure 3. Funnel plot comparing log RR vs the SE of log RR. Open circles represent trials included in the metaanalysis. The line in the center indicates the summary log RR. The other lines represent the 95% CIs. There is no evidence of publication bias. RR 5 risk ratio; SE 5 standard error.
preventing COPD exacerbations, if at all, and should be judiciously used in patients with COPD keeping in mind the risk-benefit ratio. These results could also be potentially explained by a low statistical power of our metaanalysis because of the inclusion of only 11 trials. Further, the metaregression analysis is dependent on the mean FEV1 values, and does not take into account the standard deviation of FEV1. Hence, extreme caution should be exercised in the interpretation of these results in that they cannot be extended to daily practice. Although based on the results of this study, an individual patient data metaanalysis should be performed to properly define the role of ICS in preventing COPD exacerbations. Two recent metaanalyses suggest a higher risk of serious pneumonia in patients receiving ICS with no
improvement in survival with the use of ICS.40,41 In fact, the risk of pneumonia was increased in patients with the lowest baseline FEV1,40 meaning that in situations where ICS is indicated for prevention of exacerbations, its use is likely to be associated with higher risk of pneumonia. Finally, in patients with COPD, higher doses of ICS are required for clinical efficacy compared to asthma, and this also may increase the propensity for steroid related side effects. It is widely believed that patients with COPD with lower levels of lung function are prone to develop more frequent and more severe exacerbations.44,45 In one study, the hazards for the time to first exacerbation was greater among patients with an FEV1 less than 50% predicted than in patients with a higher FEV1.46 A retrospective analysis of the ISOLDE study suggested that fluticasone caused a greater decline in exacerbation rates in patients with worse lung function compared with better lung function.47 However, a recent study showed that although baseline FEV1 was an independent predictor of exacerbation, the relationship was not particularly strong.5,48 It may also be possible that the benefit of ICS in COPD is not dependent on baseline FEV1 values but on certain other parameters, such as patients with COPD who suffer one or more exacerbations (irrespective of FEV1), and this fact should also be kept in mind in future COPD trials. The other limitations of our study include the presence of clinical heterogeneity in that the different studies have included different doses of ICS and varying severity of patients with COPD. We also included RR from individual studies rather than the number of patients with COPD who suffered an exacerbation. In analyzing exacerbation, the length of follow-up in clinical trials should also be considered. Most trials have estimated the mean rate
Table 4—Sensitivity Analysis on the Reduction of Exacerbations With the Use of ICS Stratified by the Severity of COPD Severity of COPD
Included Studies
RR (95% CI)
I2
Cochran Q Statistic (P Value)
Mild to moderate (FEV1 . 50%)
Renkema et al27 Copenhagen study28 Inhaled Steroids in Obstructive Lung Disease in Europe (ISOLDE) study30 Choudhury et al36 Weir et al29 Lung Health study31 Trial of Inhaled Steroids and Long-acting 2 Agonist (TRIAL) study32 Calverley et al33 Szafranski et al34 Towards a Revolution in COPD Health (TORCH) study35 Calverley et al37
1.026 (0.859-1.226), fixed 1.026 (0.859-1.226), random
0
2.91 (.41)
0.824 (0.775-0.876), fixed 0.785 (0.692-0.891), random …
57.4
14.09 (.03)
…
…
Severe to very severe (FEV1 , 50%)
RR 5 risk ratio. See Table 2 for expansion of other abbreviations. www.chestjournal.org
CHEST / 137 / 2 / FEBRUARY, 2010
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
323
Acknowledgments
Table 5—Metaregression of Reduction of Exacerbations With the Use of ICS and FEV1 Overall
Coefficient
Standard Error
95% CI
P Value
FEV1 Constant
0.005702 20.45984
0.008229 0.367822
20.01043-0.021831 21.18076-0.261076
.36 …
The variance in the FEV1 coefficient between the individual studies t2 5 0.02239194. See Table 2 for expansion of abbreviation.
of exacerbation for the group by using an average that was unweighted for the follow-up duration, and could potentially lead to an exaggerated beneficial effect of ICS.6,49 Thus the benefit of ICS may be even less than what has been suggested from this analysis. Finally, metaregression analysis has its own limitations in that it is an observational association across trials, and does not have the benefit of randomization. Thus, as mentioned previously, the results of our study need to be confirmed by an individual patient data metaanalysis. In conclusion, the results of our study suggest a modest benefit of ICS in preventing COPD exacerbations. In a subgroup analysis, the benefit was observed only in patients with FEV1 values , 50% predicted. However, on a metaregression analysis, there was no demonstrable linear relationship between efficacy of ICS and prevention of COPD exacerbations that is related to the level of stable lung function. Thus the role of ICS in preventing COPD exacerbations needs reappraisal.
Figure 4. Metaregression of the effect of FEV1 on the log RR for the risk of exacerbations associated with the use of inhaled corticosteroids. Each circle represents one study. Circle size reflects the weight of the study in the metaregression based on inverse variances. The line represents metaregression with an unrestricted maximum effects likelihood random-effects model (slope 5 0.0057, intercept 5 20.4598, Q 5 10.06, P 5 .36). The numbers represent the citation number of the studies included in the metaregression. See Figure 3 legend for expansion of abbreviation.
Author contributions: Dr Agarwal: contributed to literature search, data analysis, and drafting of the manuscript. Dr Aggarwal: contributed to literature search, data analysis, and drafting of the manuscript. Dr Gupta: contributed to literature search and drafting of the manuscript. Dr Jindal: contributed to literature search and drafting of the manuscript. Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.
References 1. Global strategy for the diagnosis, management and prevention of chronic obstructive pulmonary disease. (Updated 2008). http://www.goldcopd.com. Accessed September 2, 2009. 2. World Health Organization. The Global Burden of Disease— 2004 Update. Geneva, Switzerland: WHO Press; 2008. 3. Lopez AD, Shibuya K, Rao C, et al. Chronic obstructive pulmonary disease: current burden and future projections. Eur Respir J. 2006;27(2):397-412. 4. Strassels SA, Smith DH, Sullivan SD, Mahajan PS. The costs of treating COPD in the United States. Chest. 2001;119(2): 344-352. 5. Niewoehner DE. Relation of chronic obstructive pulmonary disease exacerbations to FEV(1)—an intricate tango. Respiration. 2009;77(2):229-235. 6. Suissa S, McGhan R, Niewoehner D, Make B. Inhaled corticosteroids in chronic obstructive pulmonary disease. Proc Am Thorac Soc. 2007;4(7):535-542. 7. Cazzola M, Matera MG. To add, or not to add an inhaled corticosteroid in moderate COPD: that is the question. Chest. 2008;134(2):223-225. 8. Sin DD, McAlister FA, Man SF, Anthonisen NR. Contemporary management of chronic obstructive pulmonary disease: scientific review. JAMA. 2003;290(17):2301-2312. 9. Ernst P, Baltzan M, Deschênes J, Suissa S. Low-dose inhaled and nasal corticosteroid use and the risk of cataracts. Eur Respir J. 2006;27(6):1168-1174. 10. Suissa S, Baltzan M, Kremer R, Ernst P. Inhaled and nasal corticosteroid use and the risk of fracture. Am J Respir Crit Care Med. 2004;169(1):83-88. 11. 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. 12. Alsaeedi A, Sin DD, McAlister FA. The effects of inhaled corticosteroids in chronic obstructive pulmonary disease: a systematic review of randomized placebo-controlled trials. Am J Med. 2002;113(1):59-65. 13. Wilt TJ, Niewoehner D, Kim C, et al. Use of spirometry for case finding, diagnosis, and management of chronic obstructive pulmonary disease (COPD). Evid Rep Technol Assess (Summ). 2005;121(121):1-7. 14. Gartlehner G, Hansen RA, Carson SS, Lohr KN. Efficacy and safety of inhaled corticosteroids in patients with COPD: a systematic review and meta-analysis of health outcomes. Ann Fam Med. 2006;4(3):253-262. 15. Wilt TJ, Niewoehner D, MacDonald R, Kane RL. Management of stable chronic obstructive pulmonary disease: a systematic review for a clinical practice guideline. Ann Intern Med. 2007; 147(9):639-653. 16. Puhan MA, Bachmann LM, Kleijnen J, Ter Riet G, Kessels AG. Inhaled drugs to reduce exacerbations in patients with Original Research
324
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
17. 18.
19. 20. 21. 22. 23. 24. 25. 26. 27. 28.
29.
30.
31. 32.
33.
chronic obstructive pulmonary disease: a network meta-analysis. BMC Med. 2009;7:2. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996;17(1):1-12. Kjaergard LL, Villumsen J, Gluud C. Reported methodologic quality and discrepancies between large and small randomized trials in meta-analyses. Ann Intern Med. 2001;135(11): 982-989. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst. 1959;22(4):719-748. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7(3):177-188. Armitage P, Berry G, Matthews JNS. Analysing Means and Proportions. Statistical Methods in Medical Research. Oxford: Blackwell Science; 2002:83-146. Fleiss JL. Analysis of data from multiclinic trials. Control Clin Trials. 1986;7(4):267-275. Dear K, Begg C. An approach to assessing publication bias prior to performing a meta-analysis. Stat Sci. 1992;7(2):237-245. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629-634. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50(4): 1088-1101. Thompson SG, Higgins JP. How should metaregression analyses be undertaken and interpreted? Stat Med. 2002; 21(11):1559-1573. Renkema TE, Schouten JP, Koëter GH, Postma DS. Effects of long-term treatment with corticosteroids in COPD. Chest. 1996;109(5):1156-1162. 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. Weir DC, Bale GA, Bright P, Sherwood Burge P. A doubleblind placebo-controlled study of the effect of inhaled beclomethasone dipropionate for 2 years in patients with nonasthmatic chronic obstructive pulmonary disease. Clin Exp Allergy. 1999;29(Suppl 2):125-128. 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. Lung Health Study Research Group. Effect of inhaled triamcinolone on the decline in pulmonary function in chronic obstructive pulmonary disease. N Engl J Med. 2000;343(26):1902-1909. 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. 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.
www.chestjournal.org
34. 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. 35. Calverley PM, 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. 36. 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. 37. Calverley PM, Rennard S, Nelson HS, et al. One-year treatment with mometasone furoate in chronic obstructive pulmonary disease. Respir Res. 2008;9:73. 38. Sutton AJ, Higgins JP. Recent developments in meta-analysis. Stat Med. 2008;27(5):625-650. 39. Soriano JB, Sin DD, Zhang X, et al. A pooled analysis of FEV1 decline in COPD patients randomized to inhaled corticosteroids or placebo. Chest. 2007;131(3):682-689. 40. Drummond MB, Dasenbrook EC, Pitz MW, Murphy DJ, Fan E. Inhaled corticosteroids in patients with stable chronic obstructive pulmonary disease: a systematic review and metaanalysis. JAMA. 2008;300(20):2407-2416. 41. 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. 42. 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. 43. Cook DJ, Guyatt GH, Ryan G, et al. Should unpublished data be included in meta-analyses? Current convictions and controversies. JAMA. 1993;269:2749-2753. 44. Garcia-Aymerich J, Monsó E, Marrades RM, et al; EFRAM Investigators. Risk factors for hospitalization for a chronic obstructive pulmonary disease exacerbation. EFRAM study. Am J Respir Crit Care Med. 2001;164(6):1002-1007. 45. Garcia-Aymerich J, Farrero E, Félez MA, Izquierdo J, Marrades RM, Antó JM; Estudi del Factors de Risc d’Agudització de la MPOC investigators. Risk factors of readmission to hospital for a COPD exacerbation: a prospective study. Thorax. 2003;58(2):100-105. 46. 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. 47. Jones PW, Willits LR, Burge PS, Calverley PM; Inhaled Steroids in Obstructive Lung Disease in Europe study investigators. Disease severity and the effect of fluticasone propionate on chronic obstructive pulmonary disease exacerbations. Eur Respir J. 2003;21(1):68-73. 48. Niewoehner DE, Lokhnygina Y, Rice K, et al. Risk indexes for exacerbations and hospitalizations due to COPD. Chest. 2007;131(1):20-28. 49. Suissa S. Statistical treatment of exacerbations in therapeutic trials of chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2006;173(8):842-846.
CHEST / 137 / 2 / FEBRUARY, 2010
Downloaded from chestjournal.chestpubs.org by Kimberly Henricks on February 9, 2010 © 2010 American College of Chest Physicians
325