ACC 2006
BackgroundBackground
• There is ongoing debate to identify the most effective There is ongoing debate to identify the most effective
drug-eluting stent between the two currently most used drug-eluting stent between the two currently most used
devices (ie, sirolimus- [SES] and paclitaxel-eluting devices (ie, sirolimus- [SES] and paclitaxel-eluting
stents [PES])stents [PES])
• In order to overcome the limitations and apparently In order to overcome the limitations and apparently
conflicting results of single randomized trials, directconflicting results of single randomized trials, direct11
and indirectand indirect22 comparison meta-analyses have been comparison meta-analyses have been
performedperformed
• Most recently, methods have been developed to Most recently, methods have been developed to
combine direct and indirect meta-analyses,combine direct and indirect meta-analyses,33 but to date but to date
they have been employed rarelythey have been employed rarely
1Kastrati et al, JAMA 2005; 2Biondi-Zoccai et al, Int J Cardiol 2005; 3Bucher et al, J Clin Epidemiol 1997
ACC 2006
RationaleRationale
1Bucher et al, J Clin Epidemiol 1997; 2Song et al, BMJ 2003; 3Psaty et al, JAMA 2003
TREATMENT TREATMENT AA
TREATMENT TREATMENT BB
Direct comparisonDirect comparison
TREATMENT TREATMENT AA
TREATMENT TREATMENT BBIndirect comparisonIndirect comparison
TREATMENT TREATMENT CC
TREATMENT TREATMENT CC
Direct and indirect comparison meta-analyses can beDirect and indirect comparison meta-analyses can be
then combined using a generic inverse variance approachthen combined using a generic inverse variance approach1-31-3
ACC 2006
Aim of the studyAim of the study
• We aimed to search, appraise and then combine We aimed to search, appraise and then combine
previously published direct and indirect previously published direct and indirect
comparison meta-analyses of SES vs PEScomparison meta-analyses of SES vs PES
• The goal was to maximize statistical precision The goal was to maximize statistical precision
and power, as well as increasing robustness and and power, as well as increasing robustness and
validity of such SES vs PES comparisonvalidity of such SES vs PES comparison1-31-3
1Bucher et al, J Clin Epidemiol 1997; 2Song et al, BMJ 2003; 3Psaty et al, JAMA 2003
ACC 2006
MethodsMethods• Direct and indirect comparison meta-analyses of Direct and indirect comparison meta-analyses of
SES vs PES were systematically searched in SES vs PES were systematically searched in electronic databaseselectronic databases
• Pertinent results for binary angiographic restenosis Pertinent results for binary angiographic restenosis and target lesion revascularization at 6-12 and target lesion revascularization at 6-12 months were abstracted and pooled by means of months were abstracted and pooled by means of fixed effects generic inverse variance weightingfixed effects generic inverse variance weighting
• We then computed odds ratios (OR) with 95% We then computed odds ratios (OR) with 95% confidence intervals, and inconsistency (Iconfidence intervals, and inconsistency (I22) as a ) as a measure of statistical heterogeneitymeasure of statistical heterogeneity
• Two-tailed statistical significance was set at the Two-tailed statistical significance was set at the 0.05 level0.05 level
ACC 2006
ResultsResults• We identified a full report of direct comparison We identified a full report of direct comparison
meta-analysis,meta-analysis,11 which pooled 6 head-to-head which pooled 6 head-to-head randomized clinical trials comparing SES vs PES randomized clinical trials comparing SES vs PES in 3669 patients (1845 treated with SES and 1824 in 3669 patients (1845 treated with SES and 1824 with PES)with PES)
• We also retrieved an indirect comparison meta-We also retrieved an indirect comparison meta-analysis,analysis,22 which compared SES vs PES from 9 which compared SES vs PES from 9 additional trials enrolling 2185 patients (1007 additional trials enrolling 2185 patients (1007 treated with SES and 1178 treated with PES), treated with SES and 1178 treated with PES), using as benchmark the bare-metal-stent control using as benchmark the bare-metal-stent control group for each studygroup for each study
1Kastrati et al, JAMA 2005; 2Biondi-Zoccai et al, Int J Cardiol 2005
ACC 2006
ResultsResultsTrialTrial SES PtsSES Pts PES PtsPES Pts BMS PtsBMS PtsCORPAL
ISAR-DESIRE
ISAR-DIABETES
REALITY
SIRTAX
TAXI
C-SIRIUS
E-SIRIUS
RAVEL
SES-SMART
SIRIUS
TAXUS I
TAXUS II MR
TAXUS II SR
TAXUS IV
TAXUS VI
331
100
125
684
503
102
50
175
120
129
533
-
-
-
-
-
321
100
125
669
509
100
-
-
-
-
-
31
135
131
662
219
-
-
-
-
-
-
50
178
118
128
525
30
134
136
652
227
DIR
EC
TD
IREC
T
META
-AN
ALY
SIS
META
-AN
ALY
SIS
IND
IREC
TIN
DIR
EC
T
META
-AN
ALY
SIS
META
-AN
ALY
SIS
ACC 2006
ResultsResults
• Pooling these two systematic reviews, by Pooling these two systematic reviews, by means of a fixed effects generic inverse means of a fixed effects generic inverse variance method lead to the inclusion of a variance method lead to the inclusion of a total of 5854 patients (2852 treated with SES total of 5854 patients (2852 treated with SES and 3002 with PES)and 3002 with PES)
• In this population the use of SES in In this population the use of SES in comparison to PES was associated with comparison to PES was associated with significant reductions in binary angiographic significant reductions in binary angiographic restenosis (p<0.0001) and target lesion restenosis (p<0.0001) and target lesion revascularization (p=0.0008), even if the revascularization (p=0.0008), even if the former end-point was statistically former end-point was statistically inhomogeneousinhomogeneous
1Kastrati et al, JAMA 2005; 2Biondi-Zoccai et al, Int J Cardiol 2005
ACC 2006
Binary angiographic restenosisBinary angiographic restenosis
ACC 2006
Target lesion revascularizationTarget lesion revascularization
ACC 2006
ConclusionsConclusions
• This review combining previous direct and This review combining previous direct and indirect meta-analyses of SES vs PES indirect meta-analyses of SES vs PES definitively establish the superiority of SES definitively establish the superiority of SES in comparison to PES in patients undergoing in comparison to PES in patients undergoing percutaneous coronary interventionpercutaneous coronary intervention
• These data, notwithstanding their inherent These data, notwithstanding their inherent limitations, should be taken into account limitations, should be taken into account together with costs in clinical decision together with costs in clinical decision makingmaking
ACC 2006
For further slides on these topics please feel For further slides on these topics please feel free to visit the metcardio.org website:free to visit the metcardio.org website:
http://www.http://www.metcardiometcardio..orgorg//slidesslides..htmlhtml