The Registry Experience at the Washington Hospital Center with Drug-Eluting Stents (REWARDS) study includes unselected patients with coronary artery disease treated with sirolimus-eluting stents (SESs; n = 2,392) or paclitaxel-eluting stents (PES; n = 1,119). This study aimed to examine the long-term safety profile of the 2 stents in a “real-world” population, especially in relation to stent thrombosis, and to compare differences in the diabetic cohort. Patients were followed for 3 years with regard to major adverse cardiac events (MACEs), including death, Q-wave myocardial infarction, and target lesion revascularization. Rates of stent thrombosis were also studied. Baseline characteristics were similar between stents. Although MACE rates at 3 years were similar (SES 28.1% vs PES 28.9%, p = 0.62), there was a significant difference in unadjusted rates of target lesion revascularization (SES 15.6% vs PES 12.6%, p = 0.03), death (SES 15.7% vs PES 19.0%, p = 0.02), and Q-wave myocardial infarction (SES 0.8% vs PES 2.1%, p = 0.003). After multivariable Cox regression to adjust for confounders, there was no significant difference in overall MACEs. Incidence of stent thrombosis was higher in the SES group (SES 2.2% vs PES 1.6%, p = 0.22), but this was not statistically significant (hazard ratio 1.6, 95% confidence interval 0.8 to 2.9, p = 0.17). Overall, diabetics had a higher MACE rate, but there was no difference between insulin- and noninsulin-dependent diabetics. In conclusion, at 3 years, PES and SES achieved similar results in MACEs and stent thrombosis. This should foster confidence that SES or PES can be compared to second-generation drug-eluting stents without concerns for safety or efficacy.
It has been established from pooled data on safety and efficacy of sirolimus-eluting stent (SES) and paclitaxel-eluting stent (PES) trials that the 2 drug-eluting stents are associated with significantly lower rates of angiographic and clinical restenosis compared to bare metal stents for up to 4 years in selected patients. Several head-to-head analyses of SESs and PESs have reported conflicting results regarding the clinical outcomes of these 2 stent platforms, specifically in reference to the incidence of revascularization. Another safety concern posed by drug-eluting stents is the higher incidence of late stent thrombosis. Delayed healing characterized by persistent fibrin deposition, poorer endothelialization, and local hypersensitivity reaction are some mechanisms that have been postulated to explain the late occurrence of thrombosis-related events. Current long-term data comparing the incidence of stent thrombosis between these 2 stents, especially after 1 year, have been reported in only a few studies. Concerns regarding the long-term safety of drug-eluting stents therefore linger. An important subpopulation treated with drug-eluting stents is comprised of diabetic patients. Patients with diabetes are known to have an accelerated and more aggressive form of atherosclerosis and tend to develop substantially higher rates of restenosis compared to nondiabetics. Some recent randomized trials and registry studies have suggested a lower incidence of target lesion revascularization in diabetic patients treated with SESs compared to those treated with PESs, but controversy still remains. The aim of this registry analysis was therefore twofold: to examine the long-term safety profile of SESs and PESs in a “real-world” population, especially in relation to stent thrombosis, and to compare differences in outcomes between diabetic and nondiabetic patients.
Methods
An unselected cohort of patients with coronary artery disease undergoing percutaneous coronary intervention with SESs or PESs was enrolled in the Registry Experience at Washington Hospital Center with Drug-Eluting Stents (REWARDS) registry. This is a single-center registry of patients treated exclusively with SESs or PESs. Demographic, clinical, and procedural characteristics were collected at the time of the index procedure. The study cohort included patients undergoing drug-eluting stent implantation for indications not approved by the Food and Drug Administration. The overall study population included 3,511 patients (6,514 lesions), of whom 2,392 patients (4,538 lesions) were treated with SESs and 1,119 patients (1,976 lesions) were treated with PESs. All patients gave written informed consent for the procedure, and the study was approved by the institutional review board at Washington Hospital Center (Washington, DC).
All patients were treated with aspirin 325 mg before percutaneous coronary intervention and loaded with clopidogrel 300 to 600 mg. Dual antiplatelet therapy was recommended for a minimum duration of 12 months. During percutaneous coronary intervention, anticoagulation was accomplished with bivalirudin (bolus of 0.75 mg/kg, followed by an intravenous infusion of 1.75 mg/kg/hour) or unfractionated heparin (bolus of 40 U/kg and additional heparin to achieve an activated clotting time of 250 to 300 seconds). Choice of drug-eluting stent type was at the discretion of the operator, as was the use of adjunctive devices and glycoprotein IIb/IIIa inhibitors.
Death was defined as all-cause mortality. Q-wave myocardial infarction was defined as an increase in creatine kinase-MB ≥2 times the upper normal value in the presence of new Q waves on electrocardiogram in ≥2 contiguous leads. Target lesion revascularization was clinically driven and defined as subsequent percutaneous or surgical revascularization of the stented lesion. The composite end point of major adverse cardiac events (MACEs) included death, Q-wave myocardial infarction, and target lesion revascularization. Stent thrombosis included definite and probable stent thromboses, according to the Academic Research Consortium definition. Early stent thrombosis was defined as occurring from day 0 to 30, and late stent thrombosis was defined as occurring beyond day 30. Angiographic success was defined as residual stenosis <30% with Thrombolysis In Myocardial Infarction grade 3 flow.
Demographic, clinical, and procedural data and in-hospital outcomes were collected from hospital chart review by independent research personnel blinded to the study objectives. All data management and analyses were performed by a dedicated data coordinating center. Clinical follow-up was performed on all patients at 1 month and 6, 12, 24, and 36 months by trained personnel using telephone contact or office visits. Clinical events were adjudicated by independent physicians not involved in the index procedure.
Statistical analysis was performed using SAS 9.1 (SAS Institute, Cary, North Carolina) and R 2.7.2 (R Foundation for Statistical Computing, Vienna, Austria). Continuous variables were expressed as mean ± SD and compared using Student’s t test. Categorical variables were expressed as percentages and compared to chi-square test or Fisher’s exact test, as appropriate. Statistical significance was defined as a p value <0.05. Univariable analysis of a MACE and its individual end points was performed using Kaplan-Meier curves, using log-rank test. Multivariable Cox proportional hazards regressions were used to examine adjusted rates of MACEs, death, Q-wave myocardial infarction, target lesion revascularization, and stent thrombosis. Variables for each multivariable model were selected based on univariable p values and overall clinical relevance. We also constructed a competing risk model using cumulative incidence functions to display the proportion of patients with the event of interest of the competing event as time passed. To analyze the effect of baseline predictors on cumulative incidence functions, the regression model of Fine and Gray was applied and the following baseline characteristics were included in the model: age, male gender, African-American race, acute myocardial infarction presentation, previous myocardial infarction, previous coronary artery bypass graft, hypercholesterolemia, chronic renal insufficiency, any smoking history, and type C lesions. This strategy of competing risk has been frequently used in analysis of competing causes of heart failure and valvular heart disease but has not been applied to drug-eluting stent research. We used this method to estimate the cumulative incidence function for each competing risk.
Results
Baseline clinical and procedural characteristics are presented in Tables 1 and 2 . Patients in the SES group had more dyslipidemia (87.5% vs 84.1%, p = 0.01) and ostial lesions (5.2% vs 4.0%, p = 0.002). This cohort also presented more often with stable angina pectoris (34.1% vs 26.1%, p <0.001). Patients in the PES group had an older average age (66 vs 63 years) and were more frequently men (63.0% vs 67.0%, p = 0.02). They presented more often with unstable angina pectoris (41.2% vs 47.7%, p <0.001) and acute myocardial infarction (21.2% vs 23.7%, p = 0.09). Type C lesions were more frequent in the PES cohort (19.3% vs 23.7%, p <0.001). Patients in the PES cohort were treated with a larger number of stents (1.4 ± 0.7 vs 1.5 ± 0.7, p <0.001), which on average were shorter (21.3 ± 6.6 vs 19.6 ± 6.2 mm, p <0.001) and of a smaller diameter (3.1 ± 1.9 vs 3.0 ± 0.8 mm, p <0.001). Use of glycoprotein IIb/IIIa inhibitors was more frequent in patients treated with SESs (14.1% vs 9.1%, p <0.001).
| Variable | SES | PES | p Value |
|---|---|---|---|
| (n = 2,392) | (n = 1,119) | ||
| Men | 1,548 (63.0%) | 763 (67.0%) | 0.02 |
| Age (years), mean ± SD | 63.3 ± 11.7 | 66.1 ± 12.1 | 0.07 |
| Diabetes mellitus | 881 (36.2%) | 411 (36.2%) | 0.99 |
| Insulin-requiring diabetes mellitus | 284 (11.7%) | 135 (11.9%) | 0.85 |
| Current smoker | 444 (18.1%) | 197 (17.3%) | 0.58 |
| Systemic hypertension ⁎ | 2,015 (82.4%) | 920 (81.0%) | 0.31 |
| Dyslipidemia † | 2,122 (87.5%) | 952 (84.1%) | 0.01 |
| Family history of preventive coronary disease | 1,315 (56.1%) | 570 (52.6%) | 0.06 |
| Heart failure | 374 (16.3%) | 157 (14.2%) | 0.12 |
| Chronic renal insufficiency ‡ | 317 (13.0%) | 166 (14.7%) | 0.15 |
| Dialysis | 82 (3.4%) | 40 (3.6%) | 0.77 |
| Previous myocardial infarction | 768 (33.2%) | 385 (35.9%) | 0.12 |
| Previous coronary artery bypass grafting | 435 (17.9%) | 228 (20.2%) | 0.10 |
| Previous percutaneous coronary intervention | 625 (26.4%) | 300 (27.5%) | 0.49 |
| Peripheral vascular disease | 396 (16.3%) | 196 (17.4%) | 0.43 |
| Clinical presentation | |||
| Stable angina pectoris | 834 (34.1%) | 296 (26.1%) | <0.001 |
| Unstable angina pectoris | 1,009 (41.2%) | 541 (47.7%) | <0.001 |
| Acute myocardial infarction | 504 (21.2%) | 263 (23.7%) | 0.09 |
| Cardiogenic shock | 70 (3.0%) | 29 (2.6%) | 0.58 |
| Left ventricular ejection fraction, mean ± SD | 48 ± 14 | 48 ± 14 | 0.73 |
⁎ History of hypertension diagnosed and/or treated with medication or currently being treated with diet and/or medication by a physician.
† Includes patients with a previously documented diagnosis of dyslipidemia. The patient may be treated with diet or medication. A new diagnosis can be made during this hospitalization with an increased total cholesterol > 160 mg/dl. Does not include increased triglycerides.
‡ Previously diagnosed or treated with medication, diet, or dialysis by a physician. Diagnosis at admission if a baseline creatinine level >2.0 mg/dl is found.
| Variable | SES | PES | p Value |
|---|---|---|---|
| (n = 4,538) | (n = 1,976) | ||
| Coronary narrowing location | |||
| Left main | 89 (2.0%) | 37 (1.9%) | 0.81 |
| Left anterior descending | 1,782 (39.3%) | 736 (37.2%) | 0.12 |
| Left circumflex | 1,041 (22.9%) | 452 (22.9%) | 0.95 |
| Right | 1,369 (30.2%) | 625 (31.6%) | 0.24 |
| Saphenous vein graft | 238 (5.2%) | 116 (5.9%) | 0.31 |
| Location | |||
| Ostial | 230 (5.2%) | 76 (4.0%) | 0.03 |
| Proximal | 2,151 (48.9%) | 747 (39.0%) | <0.001 |
| Mid | 1,452 (33.0%) | 798 (41.7%) | <0.001 |
| Distal | 542 (12.3%) | 285 (14.9%) | 0.006 |
| Lesion type (ACC/AHA classification) | |||
| Type A | 335 (7.8%) | 133 (7.1%) | 0.37 |
| Type B | 3,134 (73.0%) | 1,288 (69.2%) | 0.002 |
| Type C | 827 (19.3%) | 441 (23.7%) | <0.001 |
| In-stent restenosis | 240 (5.3%) | 99 (5.0%) | 0.67 |
| Procedural details | |||
| Number of lesions treated, mean ± SD | 1.8 ± 2.2 | 1.7 ± 0.9 | 0.001 |
| Number of stents, mean ± SD | 1.4 ± 0.7 | 1.5 ± 0.7 | <0.001 |
| Stent diameter (mm), mean ± SD | 3.1 ± 1.9 | 3.0 ± 0.8 | <0.001 |
| Stent length (mm), mean ± SD | 21.3 ± 6.6 | 19.6 ± 6.2 | <0.001 |
| Intra-aortic balloon pump | 78 (3.2%) | 38 (3.3%) | 0.85 |
| Intravascular ultrasound | 2,996 (68.0%) | 1,222 (64.8%) | 0.01 |
| Direct stenting | 1,546 (38.0%) | 612 (33.6%) | 0.001 |
| Rotational atherectomy | 145 (3.2%) | 55 (2.8%) | 0.38 |
| Brachytherapy | 36 (0.8%) | 20 (1.0%) | 0.38 |
| Glycoprotein IIb/IIIa inhibitor use | 342 (14.1%) | 103 (9.1%) | <0.001 |
| Angiographic success | 4,403 (97.7%) | 1,887 (97.9%) | 0.53 |
| No reflow | 20 (0.5%) | 12 (0.7%) | 0.29 |
At 3 years, unadjusted MACE was similar between SESs and PESs (SES 28.1% vs PES 28.9%, p = 0.62). Univariable analysis showed significant differences in rates of target lesion revascularization (SES 15.6% vs PES 12.6%, p = 0.03), death (SES 15.7% vs PES 19.0%, p = 0.02), and Q-wave myocardial infarction (SES 0.8% vs PES 2.1%, p = 0.003). Incidence of stent thrombosis was numerically higher in the SES group, but this did not reach statistical significance (SES 2.2% vs PES 1.6%, p = 0.22). After adjusting for baseline differences using Cox proportional hazard regression, there was no significant difference in overall MACEs and only a borderline significant difference in target lesion revascularization ( Figures 1 and 2 ) . Cox proportional hazard models could not be constructed for Q-wave myocardial infarction due to the small number of events (n = 33).
Analysis of patients with and without diabetes mellitus showed a significant difference in unadjusted rates of death that favored SESs in the diabetic cohort (SES 24% vs PES 30%, p = 0.046; Table 3 ). Also, there was a significant difference in target lesion revascularization that favored PESs in the nondiabetic cohort. After adjusting for baseline differences using Cox proportional hazard models, there was no significant difference in MACEs or any individual end points ( Table 4 ). Adjustment was made for age, male gender, African-American race, acute myocardial infarction presentation, previous myocardial infarction, previous coronary artery bypass grafting, hypercholesterolemia, chronic renal insufficiency, any smoking history, and type C lesions. In an analysis of insulin-dependent versus noninsulin-dependent diabetics, there was still no significant difference in MACEs or any individual end points, even after adjusting for baseline differences using Cox proportional hazard models ( Table 4 ). Stent thrombosis was numerically higher in the 2 cohorts of diabetics in the SES group, but this was not statistically significant, even after adjustment.
| Variable | Diabetics (n = 1,253) | Nondiabetics (n = 2,230) | NIDDM (n = 845) | IDDM (n = 408) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| SES (%) | PES (%) | p value | SES (%) | PES (%) | p value | SES (%) | PES (%) | p value | SES (%) | PES (%) | p value | |
| Major adverse cardiac events ⁎ | 36.0 | 39.4 | 0.26 | 23.4 | 22.9 | 0.80 | 32.8 | 36.2 | 0.33 | 42.9 | 45.9 | 0.57 |
| Death | 24.1 | 29.5 | 0.046 | 11.0 | 13.0 | 0.17 | 21.1 | 26.9 | 0.07 | 30.5 | 34.6 | 0.41 |
| Target lesion revascularization | 17.5 | 16.0 | 0.54 | 14.4 | 11.0 | 0.03 | 24.7 | 19.6 | 0.14 | 20.1 | 21.9 | 0.71 |
| Stent thrombosis † | 2.6 | 2.2 | 0.65 | 1.9 | 1.2 | 0.24 | 2.5 | 1.1 | 0.17 | 2.8 | 4.4 | 0.40 |
⁎ Composite of death, Q-wave myocardial infarction, and target lesion revascularization.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
