Session Information
Session Type: Abstract Submissions (ACR)
Background/Purpose:
Concerns regarding Adverse Effects (AEs) often dominate decisions on applying Glucocorticoid (GC) therapy. Evidence of AEs is mainly based on observational data without proper control groups. Thus, an examination of Randomized Controlled Trials (RCTs) for short to medium term AEs from GC would be useful. Our aim was to assess the risk of selected AEs, ranked as most worrisome among patients and rheumatologists (1), in relation to dose.
Methods:
A systematic search in MEDLINE, EMBASE, and CENTRAL was performed. Inclusion criteria were RCTs in patients with Rheumatoid Arthritis (RA) with any Prednisone Equivalent Dose Differences (PEDD) between trial arms, independent of type of administration, type of GC, or study duration. RCTs of GCs without a prednisone equivalent conversion factor were excluded. One reviewer extracted the data from the included trials and a second reviewer checked the extracted data. Analyses were based on differences in dose between trial arms (more vs. less) and therefore an assumption for linearity in response, e.g. 10 mg/d PEDD in trials of 10 vs. 0 mg was equal to 70 vs. 60 mg. For the meta-analyses, a random-effects (REML) model was used to estimate the pooled Risk Ratios (RR) with 95% Confidence Intervals (CI). The average daily and accumulated PEDD in mg between trial arms were estimated and categorized in following dose-groups: low (≤7.5 mg), medium (>7.5, ≤30 mg), and high (>30 mg) for each RCT. Meta-regression and stratified analyses were conducted to explore dose-response, using daily and accumulated PEDD as covariates and dose-groups as strata for each outcome.
Results:
Of 2821 references identified, 524 were reviewed in detail, and 59 RCTs (66 Randomized Comparisons with a total of 4831 patients) were eligible for inclusion. Overall, the risks of the selected AEs over 24 to 104 weeks were not increased when comparing more vs. less GC (Table). Although none of the analyses showed statistical significance, some of the meta-regression and stratified analyses suggested a trend for a dose-response for GC given up to doses of maximum 75 mg/d PEDD. There was a trend towards an increasing risk of osteoporosis (from RR 1.07 to 3.41 when applied in medium dose), cardiovascular diseases (from RR 1.19 to 2.75 when applied in high dose). Interestingly, with data for the high dose relatively limited, the dose-response pattern for both non-viral infections and diabetes showed a tendency to decreasing risk with increasing GC dose.
|
Outcome |
No of Studies(RC) |
No of Pt. |
Pt. years |
Overall RR 95%(CI) |
Duration Median(weeks) |
Daily PEDD Median(mg) |
Accum. PEDD Median(mg) |
Dose-Response RR (95%CI) |
|||
|
Low dose Stratum |
Medium dose stratum |
High dose stratum |
P-value for |
||||||||
|
Diabetes |
18 |
1880 |
2196 |
1.18 (0.56,2.49) |
78 |
8.8 |
4612 |
1.72(0.29,10.18) |
1.03(0.13,8.36) |
1.03(0.29,3.67) |
0.84 |
|
Osteoporosis |
12 |
1464 |
2166 |
1.31 (0.74,2.32) |
104 |
5.0 |
3650 |
1.07 (0.51,2.25) |
3.41(0.68,7.13) |
NA |
0.17 |
|
Cardiovascular diseases |
20 |
2180 |
2865 |
0.88 (0.59,1.32) |
38 |
5 |
2730 |
1.19(0.44,3.26) |
2.41(0.61,9.50) |
2.75(0.18,42.76) |
0.27 |
|
Hypertension |
21 |
2178 |
3047 |
1.24(0,83,1,86) |
104 |
5 |
3600 |
1.41(0.85,2.35) |
0.77(0.40,1.50) |
1.67(0.35,7.84) |
0.28 |
|
Infections (non-viral) |
19 |
2107 |
2402 |
0.90 (0.78,1,04) |
24 |
7.5 |
1425 |
0.95(0.78,1.16) |
0.82(0.63,1.07) |
0.36(0.01,11.01) |
0.53 |
|
Renal Dysfunction |
15 |
2044 |
2724 |
1.41 (0.97,2.05) |
78 |
5 |
1890 |
1.25(0.64,2.43) |
1.99(0.11,35.44) |
NA |
0.72 |
|
Weight gain |
18 |
1967 |
2626 |
1.19 (0.98,1.46) |
78 |
7.3 |
3625 |
1.23(0.83,1.83) |
1.25(0.92,1.71) |
0.91(0.49,1.70) |
0.60 |
|
#PEDD: Prednisone Equivalent Dose Differences; RC: Randomized Comparisons; RR Risk Ratio; CI: Confidence Intervals; pt: patients |
|||||||||||
Conclusion:
This study presents empirical evidence, based on RCTs, that: 1) there is no statistical evidence to support concerns of an overall increased risk of the selected side effects of GCs, when the GC is used over a period of 24 to 104 weeks; 2) a tendency to a dose-dependent increased risk was noted for osteoporosis and cardiovascular side effects, while there may be decreasing risk for diabetes and non-viral infections.
References:
(1) van der Goes MC, et al. Ann Rheum Dis 2010; 69(6):1015-21.
Disclosure:
S. Tarp,
MundiPharma,
2;
D. Furst,
Abbott, Actelion, Amgen, BMS, Gilead, GSK, NIH, Novartis, Pfizer, Roche/Genentech, UCB ,
2,
Abbott, Actelion, Amgen, BMS, Biogenldec, Centocor, Gilead, GSK, NIH, Novartis, Pfizer, Roche/Genentech, UCB,
5,
Abbott, Actelion, UCB (CME ONLY) ,
8;
J. R. Kirwan,
Nitec, Mundipharma, Horizon,
2;
M. Boers,
Horizon and Mundipharma,
5;
H. Bliddal,
MundiPharma,
2;
T. Woodworth,
None;
E. M. Bartels,
MundiPharma,
2;
B. Danneskiold-Samsoe,
MundiPharma,
2;
L. E. Kristensen,
Mundipharma,
5,
Mundipharma,
8;
S. Thirstrup,
None;
M. Rasmussen,
None;
M. Kaldas,
None;
R. Christensen,
MundiPharma,
2.
« Back to 2012 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/short-to-medium-term-safety-of-glucocorticoid-therapy-in-rheumatoid-arthritis-a-systematic-review-and-dose-response-analysis-of-randomized-controlled-trials/