Unrestricted Use of Drug-Eluting Stents Compared with Bare-Metal Stents in Routine Clinical Practice: Findings From the National Heart, Lung, and Blood Institute Dynamic Registry J. Dawn Abbott, MD, FACC,* Matthew R. Voss, MD,* Mamoo Nakamura, MD,† Howard A. Cohen, MD, FACC,‡ Faith Selzer, PHD,† Kevin E. Kip, PHD, Helen A. Vlachos, MSC,†Robert L. Wilensky, MD, FACC,§ David O. Williams, MD, FACC* Providence, Rhode Island; Pittsburgh and Philadelphia, Pennsylvania; and New York, New York Abbott, J. D. et al. J Am Coll Cardiol 2007;50:
Drug-eluting vs. Bare Metal Stents: Background n Stents are utilized almost uniformly in patients undergoing percutaneous coronary intervention n Until recently, drug-eluting stents have been the preferred treatment n Some have raised concerns that drug-eluting stents may be associated with more frequent stent thrombosis leading to excess death and MI n Robust comparisons of drug-eluting and bare metal stents have been limited to highly selected, simple patient subgroups. n Comparisons between DES and BMS for non-protocol usage in North America have been limited
Drug-eluting vs. Bare Metal Stents: Purpose of Investigation In the setting of routine clinical practice to: n Determine and compare the baseline clinical and angiographic features, procedural strategies and clinical outcomes of patients treated with a bare metal and drug-eluting stent n Compare outcomes following adjustments for baseline imbalances
Dynamic Registry Prospective observational multicenter study Enrollment of sequential “waves” of consecutive patients having coronary intervention Exclusion: Refusal or inability to provide written informed consent 2000 patients per wave separated by 18 months Trained research coordinators collect data Oversampling of women and minorities Follow-up obtained by phone at 30 days, 6 months, 1 year Funding just received to follow patients for five years
Study Design n Identified patients treated with DES in Wave 4 (2004) and compared them to patients treated with a BMS in Wave 3 (2002) n Wave 3 represented the last wave before the routine availability of DES n Intent of excluding wave 4 BMS patients was to eliminate selection bias seen in Wave 4 n Patients treated with BMS in Wave 3 would likely have been treated with DES had one been available n One-year follow-up data was available in 93% of wave 3 patients and 96% of wave 4 patients
DES vs. BMS: Study Population BMS Era October 2001 to March 2002 N=2047 DES Era February to May 2004 N=2112 Exclude if no stent N=284 Exclude if only BMS or no stent N=652 BMS Group N= lesions DES Group N= lesions SES 59.8% PES 31.5%
Statistical Analysis n Univariate differences between BMS and DES u Categorical variables: chi-square test u Continuous data: Wilcoxon rank-sum test n Cumulative one-year event rates u Kaplan-Meier approach and compared by log-rank statistic n Multivariable analysis u Cox proportional hazards regression used to estimate unadjusted and adjusted hazard ratios of adverse clinical outcomes n Probability values <0.05 were considered significant
VariableBMS n=1763 DES n=1460 p-value Mean Age (years) % Female Diabetes, % Hypertension, % Hypercholesterolemia, % < Current smoking, % Prior myocardial infarction, % Prior coronary bypass, % Prior angioplasty, % Renal insufficiency DES vs. BMS: Baseline Characteristics
VariableBMS n=1763 DES n=1460 p-value Vessel Disease0.098 Single Double Triple Indication for procedure, % < Acute MI Unstable Angina Stable Angina Other Ejection Fraction, mean Cardiogenic Shock2.10.5< DES vs. BMS: Baseline Characteristics
VariableBMS n=2551 DES n=1995 p-value Mean Reference Vessel Diameter Mean Lesion Length < Lesion Types, % Total Occlusion Thrombus < Calcified Bifurcation Ostial Lesion Tortuosity, % None/Mild Moderate/Severe Tortuosity DES vs. BMS:Attempted Lesion Characteristics
DES vs. BMS: DES vs. BMS: Procedural Characteristics BMS n= 2551 DES n=1995 p-value Lesion Complication, % Abrupt Closure Dissection Side Branch Occlusion Persistent flow reduction Procedural Success, %0.16 Complete Partial Failure Angiographic success, %
VariableBMS n=176 3 DES n=1460 p- value Death MI CABG MACE (Death, Any MI, Any CABG) Bleeding Requiring Transfusion DES vs. BMS: DES vs. BMS: In-hospital Unadjusted Event Rates
VariableBMS n=1763 DES n=1460 p-value Death MI CABG <0.001 Target Vessel Revascularization <0.001 Repeat PCI MACE (Death, MI, Repeat Revascularization) <0.001 DES vs. BMS: DES vs. BMS: Cumulative Unadjusted One-Year Event Rates
Stent Thrombosis n Definite stent thrombosis reported n Over 1 year 1.0% of DES patients (n=14) u 0.8% (n=11) subacute u 0.2% (n=3) late n Data were not collected during wave 3 u not available for BMS patients
DES vs. BMS: Adjusted and Unadjusted Events at One-Year
Copyright ©2007 American College of Cardiology Foundation. Restrictions may apply. Abbott, J. D. et al. J Am Coll Cardiol 2007;50: Adjusted Kaplan-Meier Event Curves at 1 Year Freedom from (A) Death/MI (B) Repeat Revascularization
One-year Event Rates for complex lesion types According to Stent Type
Copyright ©2007 American College of Cardiology Foundation. Restrictions may apply. Abbott, J. D. et al. J Am Coll Cardiol 2007;50: Adjusted HRs and 95% CIs for 1 Year (A) Death/MI (B) TVR
DES vs. BMS: Conclusions n DES as used in routine clinical practice in patients with complex lesions resulted in a substantial reduction in clinically driven target vessel revascularization compared with BMS-treated patients n Overall risk of repeat revascularization after adjustment was 43% lower in DES patients n No excess hazard of death/MI and a low rate of stent thrombosis observed in DES patients at one-year n These findings support the use of DES in routine clinical practice n Extended follow-up is needed to determine incidence of very late stent thrombosis