Relative oral corticosteroid-sparing effect of 7 inhaled corticosteroids in chronic asthma: a meta-analysis

Relative oral corticosteroid-sparing effect of 7 inhaled corticosteroids in chronic asthma: a meta-analysis

Relative oral corticosteroid–sparing effect of 7 inhaled corticosteroids in chronic asthma: a meta-analysis Anwar K. Abdullah, MD, FRCPC, FCCP,* and S...

748KB Sizes 3 Downloads 30 Views

Relative oral corticosteroid–sparing effect of 7 inhaled corticosteroids in chronic asthma: a meta-analysis Anwar K. Abdullah, MD, FRCPC, FCCP,* and Salman Khan, MD†

Background: The relative efficacy of various inhaled corticosteroids (ICSs) for oral corticosteroid (OCS)–sparing effect in asthma is not known. To our knowledge, no randomized controlled trial directly comparing 2 ICSs has been reported, but several randomized controlled trials have reported comparison of various ICSs with placebo. Objective: To conduct an adjusted indirect comparison of 7 ICSs for their OCS-sparing effect. Methods: PubMed and bibliographies of relevant articles. Eighteen placebo-controlled randomized trials of 7 ICSs were analyzed using a random-effect model. Pooled benefit ratios (BRs) (ICS/placebo) for elimination of OCS and weighted mean differences (ICS ⫺ placebo) for OCS dose change by each ICS vs placebo were determined. Pairwise adjusted indirect comparisons of various ICSs were then made. Results: For OCS elimination, all ICSs were more effective than placebo (BR: mometasone, 17.2; budesonide, 8.2; beclomethasone and fluticasone, 5.4; triamcinolone, 4.6; ciclesonide, 2.8; and flunisolide, 2.2). On pairwise adjusted indirect comparison, the BR of mometasone was significantly higher than that of triamcinolone (P ⫽ .02), ciclesonide (P ⫽ .01), and flunisolide (P ⫽ .01) and that of budesonide was significantly higher than that of ciclesonide (P ⫽ .02) and flunisolide (P ⫽ .03). For OCS dose change, beclomethasone achieved a significantly lower final mean OCS dose than fluticasone or flunisolide (P ⬍ .001). In all other comparisons, the differences were not statistically significant. Conclusions: All ICSs studied were significantly more effective than placebo for OCS sparing, but mometasone seemed to be more effective than others. However, because of very few trials for some ICSs, more placebo-controlled trials for adjusted indirect comparison or randomized trials for direct comparison of these ICSs are needed for definitive conclusions. Ann Allergy Asthma Immunol. 2008;101:74–81.

INTRODUCTION Inhaled corticosteroids (ICSs) are preferable to oral corticosteroids (OCSs) in the treatment of asthma whenever possible because they produce fewer adverse effects.1 Different ICS preparations are known to differ in their effectiveness.2 A review of OCS-sparing effects of 5 ICSs (beclomethasone, budesonide, flunisolide, fluticasone, and triamcinolone) published in 2004 showed that all ICSs reduced OCS requirements in patients with asthma,3 but no comparison of the various ICSs for their OCS-sparing effect was reported. Since then, 2 more ICSs, mometasone and ciclesonide, have become available for clinical use. A number of randomized controlled trials (RCTs) comparing various ICSs with placebo for their OCS-sparing effect have been published, and 1 double-blind crossover study4 has reported a head-to-head comparison of beclomethasone and budesonide, but, to our knowledge, no RCT directly comparing any 2 ICSs for this effect measure has been reported. Direct comparison by RCTs remains the gold standard to estimate the relative effects of multiple competing interven-

Affiliations: * Virginia Center for Behavioral Rehabilitation, Petersburg, Virginia; † Central State Hospital, Petersburg, Virginia. Disclosures: Authors have nothing to disclose. Received for publication October 12, 2007; Received in revised form December 4, 2007; Accepted for publication January 2, 2008.

74

tions. In the absence of sufficient direct evidence, adjusted indirect comparison using the random-effects model (considering the increase in heterogeneity) has been suggested as an acceptable alternative.5 In contrast to the naı¨ve indirect method that pools findings only from the treatment arms of multiple trials, in the adjusted indirect method the comparison of the interventions of interest is adjusted by the results of their direct comparison with a common control group (eg, placebo) in separate RCTs, thus respecting the randomizations used in the original trials. This meta-analysis compares the OCS-sparing effect of 7 ICSs (beclomethasone, budesonide, ciclesonide, flunisolide, fluticasone, mometasone, and triamcinolone) based on adjusted indirect comparison, using data from published RCTs that have compared these ICSs with placebo and have included the OCS-sparing effect as one of the end points. METHODS We have followed the Quality of Reporting of Meta-analyses guidelines6 for reporting our methods and results. Data Sources and Study Selection We conducted a systematic literature search for placebocontrolled randomized trials of all ICSs that included in their outcome elimination of OCSs in patients with OCS-dependent chronic asthma while maintaining or improving their asthma control. We searched PubMed (up to September

ANNALS OF ALLERGY, ASTHMA & IMMUNOLOGY

VOLUME 101, JULY, 2008

75

28

1977

2006

27

2

3

3

3

5

4 4 4

3

3

3 2 4 3 3

4 4 3

NR

Indu

Indu

Indu

Indu

NR Indp Indu

Indu

Indu

Indu Indu NR Indu Indu

NR Indu Indu

15–77 5–16 18–85

12–74

22–77

22–75 27–77 6–13 52 (11) 20–69

NA NA ⬎16

CFCmdi

HFAmdi

DPI

DPI

21–69

13–83

13–83

12–77

Nebulizer 19–83

CFCmdi CFCmdi CFCmdi

HFAmdi

DPI

CFCmdi CFCmdi CFCmdi CFCmdi DPI

CFCmdi CFCmdi CFCmdi

12

12

12

16

12

14 12 16

12

26

12 4b 12 12 20

Until 0a 8 28

Yes

Yes

Yes

Yes

NA

NA NA Yes

Yes

NA

Yes Yes Yes NA NA

NA Yes Yes

100⫻4 100⫻4 100⫻4 200⫻4 100⫻4 100⫻4 2 puffs⫻4 250⫻4 400⫻2 800⫻2 400⫻2 800⫻2 320⫻2 640⫻2 500⫻2 500⫻2 750⫻2 1000⫻2 500⫻2 2000⫻2 500⫻2 1000⫻2 400⫻2 800⫻2 400⫻2 800⫻2 200⫻4

ICS dose ␮g/d

33

123

132

111

301

78 32 96

141

113

34 25 27 124 159

20 25 108

0

1

1

2

50

5 0 4

5

18

1 0 0 9 20

0 3 7

10 11 35 33 17 13 15 61 47 41 33 32 45 47 40 17 31 31 79 94 40 36 45 43 41 43 16 17

38

43

33

78

33 15 30

44

30

16 12 12 54 51

10 11 33

Randomized Excluded ICS PLC

No. of patients Baseline OCS, mean (SD) mg/d

10.4 7.9 (3.2) 16.5 (9.5) 17.1 (7.5) 19.5 18.7 12.1 (5.3) 11.9 (5.1) 13.6 (7.4) 11.5 (6.1) 15.2 15.4 (8.3)c 9.5 (4.2) 10.2 (5.0) 9.5 (5.6) 10.3 (5.8) 15.4 (9.4) 13.6 (7.6) 11.9 12.0 12.4 (6.9) 13.0 (6.6) 8.3

10.9 (1.7) 9.6 (2.8) ⱖ16

ICS group

9.1

10.0 (7.0)

11.6

13.0 (8.8)

10.3 (5.9)

12.7 13.3 (5.2)c 10.2 (5.1)

12.0 (6.6)

1.2 (5.0)

10.3 12.1 (3.8) 20.4 (15.3) 15.6 (7.7) 19.7

9.8 (2.4) 10.0 (2.2)

PLC group

OCS, mean (SD) mg/d

NS ⬍.05 NA

NS NS NS NS NS NS NA

NS NS NS NS NA

NA ⬍.05 NS NS NA

NS NS NA

P value

1 1

⬍3d

1

1

2

1–5

1.25–5

Clinical

2.5–5

2 2 1

1

1.25–5d 2.5–10 5 1.25–2.5

2

1 1 1 2 2

4 1 4

time, wk

Clinical

2.5–5 2.5 5d 5 2.5–5

1 2.5 1⫺2.5

mg/d

OCS dose reduction rate

Abbreviations: BDP, beclomethasone; BUD, budesonide; CFC, chlorofluorocarbon; CIC, ciclesonide; DPI, dry powder inhaler; FH, flunisolide; FP, fluticasone; HFA, hydrofluoroalkane; ICS, inhaled corticosteroid; Indp, independent funding; Indu, industry; mdi, metered-dose inhaler; MF, mometasone; NA, data not available; NR, not reported; NS, not significant; OCS, oral corticosteroid; PLC, placebo; TA, triamcinolone. P value is for OCS mean. a Until dose reduced to 0 or exacerbation occurred. b When asthma remained stable for 4 weeks on a new regimen. c Alternate day. d Double for alternate day regimens.

TA

MF

2000

1999

25

26

1999

24

FP

1980 1981 1995

21 22 23

FH

2000

19

2006

1976 1977 1978 1991 1998

14 15 16 17 18

20

1973 1974 1974

11 12 13

CIC

BUD

BDP

Study

Minimum ICS Age Study OCS delivery range, period, dose ICS Reference Publication Jadad Funding device y wk determined? type No. year score

Table 1. Characteristics of the Studies Included in Meta-analysis

Table 2. Meta-analysis for Frequency of OCS Elimination by Different ICSa

Abbreviations: BDP, beclomethasone; BUD, budesonide; CI, confidence interval; CIC, ciclesonide; FH, flunisolide; FP, fluticasone; ICS, inhaled corticosteroid; MF, mometasone; N1 and N2, total number of patients in the ICS and placebo groups, respectively; n1 and n2, number of patients in the ICS and placebo groups, respectively; OCS, oral corticosteroid; RCT, randomized controlled trial; and TA, triamcinolone. (t∧2 is a measure of heterogeneity, and fail-safe N and tolerance level are measures of publication bias.) a The various ICSs are arranged in descending order of their pooled benefit ratios and within each ICS in descending order of the weight assigned to each study. b The benefit ratio is the ratio of benefit rate in the ICS group to that in the placebo group ([n1/N1]/[n2/N2]).

76

ANNALS OF ALLERGY, ASTHMA & IMMUNOLOGY

Table 3. Change in Oral Corticosteroid Doses by Various Inhaled Corticosteroidsa

Abbreviations: BDP, beclomethasone; BUD, budesonide; CI, confidence interval; CIC, ciclesonide; FH, flunisolide; FP, fluticasone; ICS, inhaled corticosteroid; MDD, mean dose difference (ICS-placebo); OCS, oral corticosteroid; RCT, randomized controlled trial; and WMD, weighted mean difference. (t∧2 is a measure of heterogeneity, and fail-safe N and tolerance level are measures of publication bias.) a Meta-analysis for final OCS mean dose and OCS mean dose change in milligrams and percentages is given by different ICSs compared with placebo, arranged in descending order of their pooled WMD for each effect measure and within each ICS in descending order of the weight assigned to each study.

VOLUME 101, JULY, 2008

77

2007) using the terms asthma and oral corticosteroid or oral corticosteroid sparing and inhaled corticosteroid or beclomethasone or budesonide or ciclesonide or flunisolide or mometasone or triamcinolone and limited our search to “humans, randomized controlled trials, reviews.” We scanned bibliographies of relevant RCTs and review articles for additional trials. We excluded trials that were not placebo controlled, duplicate reports of the same trials, and reports that did not include data from which at least 1 of the effect measures we used for comparison could be calculated. The study selection process is summarized in Figure 1. Data Extraction and Quality Evaluation Baseline characteristics, presented in Table 1, were extracted from the selected RCT reports. In each trial, patients who dropped out after randomization because of exacerbation of asthma were included in the meta-analysis and counted as failure of the test ICS, but to prevent dropouts due to reasons other than lack of efficacy of the test ICS affecting our results, we excluded those not completing the study due to reasons other than exacerbation of asthma. Mean OCS doses (prednisolone equivalents) and their SDs at baseline and at the end of the trial, and number of patients who achieved complete elimination of OCS, were extracted or derived from data reported in each trial. Mean changes from baseline in OCS dose in milligrams and percentage of baseline dose were calculated. Data were extracted by one of us (A.K.A.) and double checked for accuracy by one of us (S.K.). The quality of each trial was evaluated by the score of Jadad et al.7 Data Synthesis and Analysis We used the following effect measures for comparison of various ICSs: complete elimination of OCS, final OCS dose, change (in milligrams) in OCS dose, and percentage change in OCS dose achieved at the end of the trial period. All selected RCTs were grouped according to the ICS under study. The overall treatment effect of each ICS relative to

Figure 1. Flow diagram of the study selection process for meta-analysis. ICS indicates inhaled corticosteroid; RCT, randomized controlled trial.

78

Table 4. Adjusted Indirect Comparison of ICSs for Causing Complete Elimination of Oral Corticosteroid ICS1/ICS2

RBR (95% CI)a

P value

MF/BUD MF/BDP MF/FP MF/TA MF/CIC MF/FH BUD/BDP BUD/FP BUD/TA BUD/CIC BUD/FH BDP/FP BDP/TA BDP/CIC BDP/FH FP/TA FP/CIC FP/FH TA/CIC TA/FH CIC/FH

2.10 (1.01–7.75) 3.19 (0.80–12.74) 3.19 (0.76–13.33) 3.74 (1.18–11.82) 6.14 (1.62–23.22) 7.82 (1.73–35.30) 1.52 (0.56–4.09) 1.52 (0.53–4.35) 1.78 (0.96–3.31) 2.93 (1.18–7.27) 3.73 (1.18–11.81) 1.00 (0.32–3.15) 1.17 (0.54–2.54) 1.93 (0.70–5.34) 2.45 (0.71–8.50) 1.17 (0.50–2.75) 1.93 (0.65–5.68) 2.45 (0.67–8.94) 1.64 (0.84–3.20) 2.09 (0.79–5.53) 1.27 (0.39–4.14)

.27 .10 .11 .02 .01 .01 .41 .44 .07 .02 .03 ⬎.99 .68 .21 .16 .71 .23 .17 .14 .14 .69

Abbreviations: BDP, beclomethasone; BUD, budesonide; CI, confidence interval; CIC, ciclesonide; FH, flunisolide; FP, fluticasone; ICS, inhaled corticosteroid; MF, mometasone; RBR, relative benefit ratio, TA, triamcinolone. a The RBR is the BR of one ICS/BR of the other ICS.

placebo with regard to these effect measures was determined by pooling its estimates from each group of ICSs using the random-effects model of DerSimonian and Laird.8 For complete elimination of OCSs, a pooled benefit ratio (BR) (ie, benefit rate for ICS/benefit rate for placebo) was derived from each group of ICS trials. For other outcomes, the pooled mean OCS dose difference (ICS-placebo) was calculated for each group of ICS trials. Adjusted indirect comparisons4 of different ICSs were then made pairwise using pooled values of each of these effect measures using the test of interaction.9 Statistical analyses and graphic generation of forest plots were performed using computer software (MIX).10 ␶2 was used for testing heterogeneity, and “fail-safe N” and “tolerance level” were used for estimating publication bias and small study effect. L’Abbe plot was generated using computer software (Microsoft Excel; Microsoft Corporation, Redmond, Washington). A significance level of .05 (2-directional alternative hypothesis) was used in all tests. RESULTS Search Results and Study Characteristics The results of the study selection process are summarized in Figure 1. Characteristics of the 18 RCTs11–28 meeting the selection criteria for this analysis are presented in Table 1. In

ANNALS OF ALLERGY, ASTHMA & IMMUNOLOGY

9 trials, 2 different doses of the ICS were compared with the same placebo group. For this meta-analysis, a trial of each dose of ICS was considered a separate study. Thus, the 18 selected trials resulted in 27 studies, which included 996 patients in ICS groups and 560 patients in placebo groups. Quantitative Data Synthesis Elimination of OCSs. Table 2 summarizes the results of meta-analysis for frequency of complete elimination of OCSs by different ICSs arranged in descending order of their pooled BRs and within each ICS in descending order of the weight assigned to each study. The overall effect of every ICS except flunisolide was significantly in favor of ICS compared with placebo (P ⬍ .003). Flunisolide was also numerically better than placebo, but the difference was not statistically significant (P ⫽ .12). The beclomethasone and fluticasone groups showed significant heterogeneity (␶2 ⫽ 0.45 and 0.78, respectively). The mometasone, ciclesonide, and flunisolide groups included only 4, 2, and 2 studies, respectively, and had a fail-safe N value lower than the tolerance level, indicating small study effect. Only one study of triamcinolone was available. L’Abbe plot (Figure 2) also showed higher benefit rates for every ICS compared with placebo in all trials and relatively more heterogeneity in the beclomethasone and fluticasone groups. In a secondary analysis using the fixed-effect model, our results did not change materially. Heterogeneity evaluated

Figure 2. L’Abbe plot. The benefit rate for oral corticosteroid elimination by inhaled corticosteroid (ICS) (n1/N1) is plotted against that by placebo (n2/N2) for each study included in the meta-analysis, stratified by ICS type. Benefit rates for ICS in all studies are above the diagonal (benefit ratio, 1.0), but there is wide scatter around the dashed line representing the estimated overall treatment effect (average of all pooled benefit ratios, 6.53). BDP indicates beclomethasone; BUD, budesonide; CIC, ciclesonide; FH, flunisolide; FP, fluticasone; MF, mometasone; and TA, triamcinolone.

VOLUME 101, JULY, 2008

with the I2 statistic remained high for beclomethasone (50.8% [95% CI, 12.4%– 81.5%]) and fluticasone (83.5% [95% CI, 65.6%–92.1%]) but was low for all other ICSs (mometasone, 0% [95% CI, 0%– 84.7%]; budesonide, 0% [95% CI, 0%84.7%]; ciclesonide, 0%; and flunisolide, 0%). OCS mean dose reduction. Only trials reporting means and SDs of final OCS doses, change (in milligrams) in OCS dose, and percentage change in OCS dose or providing data from which these could be calculated were included in metaanalysis for mean OCS dose reduction by various ICSs. Results are summarized in Table 3. For final OCS mean dose, based on 4 RCTs of beclomethasone, 1 of fluticasone (using 2 different doses), and 1 of flunisolide, the final OCS dose achieved was significantly lower in the ICS group than in the placebo group in all studies (overall P ⬍ .05). For OCS mean dose change (in milligrams), 3 RCTs of fluticasone (each using 2 different doses) and 2 of beclomethasone were used for this meta-analysis. The ICS group showed significantly more decrease in OCS mean doses than placebo for beclomethasone and fluticasone in all studies (overall P ⬍ .001). For OCS mean dose change (percentage of baseline), 1 trial each of budesonide and ciclesonide (both using 2 different doses of ICS) was included in this meta-analysis, and both showed significantly more percentage reduction in OCS mean doses by the ICS compared with placebo (overall P ⬍ .001). Adjusted indirect comparison of ICS. Table 4 presents a summary of pairwise adjusted indirect comparison of pooled BRs of all ICSs for complete elimination of OCS. Mometasone and budesonide had significantly higher BRs relative to placebo than ciclesonide and flunisolide (P ⬍ .05), indirectly suggesting that the former were more effective than ciclesonide or flunisolide for this effect measure. Mometasone had a significantly higher BR relative to placebo than triamcinolone also (P ⬍ .05). In all other pairwise comparisons of the various ICSs for this outcome, there was no statistically significant difference. For OCS mean dose reduction, only few pairwise comparisons were possible (Table 5), because SDs needed for these calculations were not given in many reports. From the few studies giving SDs, beclomethasone achieved significantly lower final OCS mean dose than fluticasone and flunisolide (P ⬍ .001), but there was no significant difference when fluticasone was compared with flunisolide. There was also no significant difference between beclomethasone and fluticasone for OCS mean dose change from baseline (in milligrams) or between budesonide and ciclesonide for OCS mean dose change as percentage of baseline dose. DISCUSSION Comparative data regarding the efficacy of different ICSs for their OCS-sparing effect are scarce. We found only 1 report of a head-to-head comparison of 2 ICSs for this effect. It was a double-blind crossover trial4 that found beclomethasone

79

Table 5. Adjusted Indirect Comparison of Inhaled Corticosteroids for Mean OCS Dose Reduction Variable Final OCS dose, mg/d BDP-FP BDP-FH FP-FH OCS dose change, mga BDP-FP OCS dose change, %a BUD-CIC

WMDd (95% CI)

P value

⫺6.38 (⫺11.4 to ⫺1.4) ⫺5.06 (⫺9.9 to ⫺0.2) 1.32 (⫺0.7 to 3.3)

⬍.001 ⬍.001 .19

0.72 (⫺3.2 to 4.6)

.65

18.5 (⫺12.5 to 49.5)

.09

Abbreviations: BDP, beclomethasone; BUD, budesonide; CI, confidence interval; CIC, ciclesonide; FH, flunisolide; FP, fluticasone; OCS, oral corticosteroid; WMDd, difference between the weighted mean difference in OCS doses of 2 ICSs. a Final ⫺ baseline.

marginally more potent than budesonide. In our indirect comparison, these 2 ICSs were not significantly different. Several studies comparing these 2 ICSs for other measures of efficacy have also found their differences to be statistically insignificant.2 In our meta-analysis for complete elimination of OCS in OCS-dependent patients with chronic asthma, the pooled BR of 17.2 for mometasone, indicating that mometasone is 17.2 times more effective than placebo, was the highest among all ICSs, and that of flunisolide (2.2) was the lowest, with the remainder falling in between (budesonide, 8.2; beclomethasone, 5.4; fluticasone, 5.4; triamcinolone, 4.6; ciclesonide, 2.8; and flunisolide, 2.2) (Table 2). For indirect comparison of the various ICSs, we determined their relative BR (RBR) (ie, BR of ICS1/BR of ICS2). We found that the RBR of mometasone/triamcinolone (3.74), mometasone/ciclesonide (6.14), and mometasone/flunisolide (7.82) and the RBR of budesonide/ciclesonide (2.93) and budesonide/flunisolide (3.73) were significantly higher (P ⫽ .01 to P ⫽ .03) than those of others (RBR, 1.00 –3.19; P ⫽ .07 to P ⬎ .99) (Table 4). Only 10 RCTs could be included in the meta-analysis for OCS mean dose reduction (final OCS dose achieved, change [in milligrams] in OCS, and percentage change in OCS dose) and presented in Table 5. These results may not be reliable because of only a few studies available for analysis, resulting in significant small-study effects (Table 3), and also because the methods for indirect comparison of continuous data (eg, final OCS dose achieved, change [in milligrams] in OCS, and percentage change in OCS dose) are not as well validated as for binary data5 (eg, complete elimination of OCS). To our knowledge, there is no other study reporting comparison of OCS-sparing effect of various ICSs, but a review of the literature comparing various ICSs for all measures of efficacy2 found mometasone better than beclomethasone or budesonide and equal to fluticasone, and ciclesonide better than or equal to budesonide and fluticasone; no study comparing the efficacy of mometasone with triamcinolone or

80

flunisolide has been reported. Because flunisolide has been less effective than most other ICSs in other measures of efficacy,2 it is not surprising that flunisolide was less effective than mometasone and budesonide in its OCS-sparing effect also (Table 4). But it is not clear why ciclesonide, which has been equal to or more effective than mometasone and budesonide in other measures of efficacy,2 was found less effective in its OCS-sparing effect (Table 4). It may partly be because the analysis for ciclesonide is based on only 1 trial. Additional trials comparing ciclesonide with other ICSs for their OCS-sparing effect will be needed to clarify this. In summary, using the adjusted indirect method, we compared 7 ICSs for their OCS-sparing effect in OCS-dependent chronic asthma. Based mainly on the effect measure of complete elimination of OCS, we found the following: (1) all ICSs are more effective than placebo, (2) mometasone and budesonide are significantly more effective than ciclesonide or flunisolide, and mometasone is significantly more effective than triamcinolone, and (3) for all other pairwise comparisons, the differences are not statistically significant. The robustness of these conclusions is decreased because of the marked heterogeneity in the beclomethasone and fluticasone groups of studies, differences in dosing and formulations, and considerable small-study effects evident in the mometasone, triamcinolone, ciclesonide, and flunisolide groups of studies. Randomized trials for direct comparison of these ICSs or more and larger placebo-controlled trials for adjusted indirect comparison will be needed to make more definitive conclusions regarding their relative OCS-sparing effects. REFERENCES 1. Global Initiative for Asthma. GINA report: Global Strategy for Asthma Management and Prevention. Available at: http://www.ginasthma.org. Accessed September 4, 2007. 2. Abdullah AK, Khan S. Evidence-based selection of inhaled corticosteroid for treatment of chronic asthma. J Asthma. 2007;44:1–12. 3. Ververeli K, Chipps B. Oral corticosteroid-sparing effects of inhaled corticosteroids in the treatment of persistent and acute asthma. Ann Allergy Asthma Immunol. 2004;92:512–522. 4. Rafferty P, Tucker LG, Frame MH, Fergusson RJ, Biggs BA, Crompton GK. Comparison of budesonide and beclomethasone dipropionate in patients with severe chronic asthma: assessment of relative prednisolone-sparing effects. Br J Dis Chest. 1985;79:244 –250. 5. Glenny AM, Altman DG, Song F, et al; International Stroke Trial Collaborative Group. Indirect comparisons of competing interventions. Health Technol Assess. 2005;9:1–134, iii-iv. 6. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement: Quality of Reporting of Meta-analyses. Lancet. 1999;354:1896 –1900. 7. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996;17:1–12. 8. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–188. 9. Altman DG, Bland JM. Interaction revisited: the difference between two estimates. BMJ. 2003;326:219. 10. Bax L, Yu LM, Ikeda N, Tsuruta N, Moons KG. Development and validation of MIX: comprehensive free software for meta-analysis of causal research data. BMC Med Res Methodol. 2006;6:50. 11. Cameron SJ, Cooper EJ, Crompton GK, Hoare MV, Grant IW. Substi-

ANNALS OF ALLERGY, ASTHMA & IMMUNOLOGY

12.

13.

14.

15. 16. 17. 18.

19.

20.

tution of beclomethasone aerosol for oral prednisolone in the treatment of chronic asthma. BMJ. 1973;4:205–207. Hodson ME, Batten JC, Clarke SW, Gregg I. Beclomethasone dipropionate aerosol in asthma: transfer of steroid-dependent asthmatic patients from oral prednisone to beclomethasone dipropionate aerosol. Am Rev Respir Dis. 1974;110:403– 408. Double-blind trial comparing two dosage schedules of beclomethasone dipropionate aerosol in the treatment of chronic bronchial asthma: preliminary report of the Brompton Hospital-Medical Research Council Collaborative Trial. Lancet. 1974;2:303–307. Harvey LL, Nair SV, Kass I. Beclomethasone dipropionate aerosol in the treatment of steroid-dependent asthma: a 12-week double-blind study comparing beclomethasone dipropionate and a vehicle aerosol. Chest. 1976;70:345–350. Kerigan AT, Pugsley SO, Cockcroft DW, Hargreave FE. Substitution of inhaled beclomethasone dipropionate for ingested prednisone in steroiddependent asthmatics. Can Med Assoc J. 1977;116:867– 871. Richards W, Platzker A, Church JA, Yamamoto F, Foster S. Steroiddependent asthma treated with inhaled beclomethasone dipropionate in children. Ann Allergy. 1978;41:274 –277. Lacronique J, Renon D, Georges D, Henry-Amar M, Marsac J. Highdose beclomethasone: oral steroid-sparing effect in severe asthmatic patients. Eur Respir J. 1991;4:807– 812. Nelson HS, Bernstein IL, Fink J, et al; Pulmicort Turbuhaler Study Group. Oral glucocorticosteroid-sparing effect of budesonide administered by Turbuhaler: a double-blind, placebo-controlled study in adults with moderate-to-severe chronic asthma. Chest. 1998;113:1264 –1271. Miyamoto T, Takahashi T, Nakajima S, et al; Japanese Pulmicort Turbuhaler study group. A double-blind, placebo-controlled steroid-sparing study with budesonide Turbuhaler in Japanese oral steroid-dependent asthma patients. Respirology. 2000;5:231–240. Bateman E, Karpel J, Casale T, Wenzel S, Banerji D. Ciclesonide reduces the need for oral steroid use in adult patients with severe, persistent asthma. Chest. 2006;129:1176 –1187.

VOLUME 101, JULY, 2008

21. Slavin RG, Izu AE, Bernstein IL, et al. Multicenter study of flunisolide aerosol in adult patients with steroid-dependent asthma. J Allergy Clin Immunol. 1980;66:379 –385. 22. Shapiro GG, Izu AE, Furukawa CT, Pierson WE, Bierman CW. Shortterm double-blind evaluation of flunisolide aerosol for steroid-dependent asthmatic children and adolescents. Chest. 1981;80:671– 675. 23. Noonan M, Chervinsky P, Busse WW, et al. Fluticasone propionate reduces oral prednisone use while it improves asthma control and quality of life. Am J Respir Crit Care Med. 1995;152(pt 1):1467–1473. 24. Westbroek J, Saarelainen S, Laher M, O’Brien J, Barnacle H, Efthimiou J. Oral steroid-sparing effect of two doses of nebulized fluticasone propionate and placebo in patients with severe chronic asthma. Respir Med. 1999;93:689 – 699. 25. Nelson HS, Busse WW, deBoisblanc BP, et al. Fluticasone propionate powder: oral corticosteroid-sparing effect and improved lung function and quality of life in patients with severe chronic asthma. J Allergy Clin Immunol. 1999;103(pt 1):267–275. 26. Fish JE, Karpel JP, Craig TJ, et al. Inhaled mometasone furoate reduces oral prednisone requirements while improving respiratory function and health-related quality of life in patients with severe persistent asthma. J Allergy Clin Immunol. 2000;106:852– 860. 27. Karpel JP, Nayak A, Lumry W, et al. Inhaled mometasone furoate reduces oral prednisone usage and improves lung function in severe persistent asthma. Respir Med. 2007;101:628 – 637. 28. Chervinsky P. Treatment of steroid-dependent asthma with triamcinolone acetonide aerosol. Ann Allergy. 1977;38:192–197.

Requests for reprints should be addressed to: Anwar K. Abdullah, MD, FRCPC, FCCP 12909 Scrimshaw Cir Chester, VA 23836 E-mail: [email protected]

81