Although clopidogrel pretreatment benefits patients undergoing percutaneous coronary intervention (PCI) for acute coronary syndromes, these benefits are less well established among patients undergoing elective PCI—in particular, when they are treated with the direct thrombin inhibitor, bivalirudin. We used data from the multicenter Evaluation of Drug Eluting stents and ischemic Events registry to assess the association between clopidogrel pretreatment and PCI-related complications among patients undergoing elective PCI with bivalirudin as the antithrombotic regimen. The primary end point was the composite of in-hospital death or myocardial infarction. From January 2005 and December 2007, 4,681 patients underwent elective PCI at 55 United States centers, and 1,913 (41%) received bivalirudin as the planned anticoagulant. Clopidogrel pretreatment was used in 923 patients (48%). The incidence of in-hospital death or myocardial infarction was similar among patients who did and did not receive clopidogrel pretreatment (5.5% vs 5.8%, p = 0.83). This result was unchanged in propensity-adjusted analyses (adjusted odds ratio for pretreatment 0.91, 95% confidence interval 0.60 to 1.39, p = 0.66). Also, no differences were seen in the in-hospital bleeding events (1.0% vs 1.0%, p = 0.94) or 1-year ischemic complications between the 2 treatment groups (7.5% vs 8.3%, p = 0.26). In conclusion, among unselected patients undergoing elective PCI with bivalirudin as the planned anticoagulant, clopidogrel pretreatment was common but was not associated with a reduced risk of ischemic complications.
Although the benefits of aggressive antiplatelet therapy before percutaneous coronary intervention (PCI) are generally accepted for patients with acute coronary syndromes, the benefits of pretreatment are less certain for patients undergoing elective PCI. The only clinical trial to address this specific issue coupled early clopidogrel loading with a longer duration of therapy and was underpowered to study the benefits of these 2 components separately. Moreover, virtually all studies of clopidogrel pretreatment have used unfractionated heparin (with or without a glycoprotein IIb/IIIa inhibitor) as the foundation antithrombotic regimen. Thus, the benefits of clopidogrel pretreatment among patients undergoing elective PCI and receiving newer antithrombotic regimens, such as bivalirudin (which does not activate platelets in vivo or in vitro ), are unknown. In the present study, we used data from the Evaluation of Drug Eluting stents and ischemic Events (EVENT) registry, a multicenter study of the practice and outcomes of PCI, to address this gap in current knowledge. The goals of the present study were to examine the association between clopidogrel pretreatment and both acute and long-term ischemic complications among patients undergoing elective PCI with bivalirudin as the anticoagulant regimen. We also sought to determine whether differences in the timing of clopidogrel loading altered this relation.
Methods
The methods and population of the EVENT registry have been previously published. From 2004 and 2007, patients undergoing attempted implantation of ≥1 approved coronary stents at 55 participating centers were sequentially enrolled in 4 “waves” of ∼2,500 patients each (wave 1, August 2004 to March 2005; wave 2, June to September 2005; wave 3, February to June 2006; wave 4, July to December 2007). Specific efforts were made to enroll patients consecutively during each enrollment period (e.g., on predetermined days of the week) to minimize selection bias. The ethical review committees at all participating institutions approved the study protocol, and all patients provided written informed consent before participation. Data on patient characteristics, clinical presentation, and treatment were collected prospectively using standardized case report forms and submitted to the data coordinating center. Creatine kinase (CK) and CK-MB levels were assessed at baseline and every 8 hours for a minimum of 2 samples after the procedure using each site’s clinical laboratory and reference values. If a myocardial infarction (MI) was suspected at a later point, additional biomarkers were obtained as clinically indicated.
The population for our study consisted of all EVENT patients who underwent elective PCI in waves 2, 3, or 4. Patients enrolled in wave 1 were excluded because the details of the timing and precise dose of clopidogrel loading were not recorded. Patients with unstable angina were excluded because of the possibility that elevation of postprocedure biomarkers might have reflected ongoing MI before PCI rather than a true procedure-related event.
The main exposure variable of interest was clopidogrel pretreatment. Because the antiplatelet effect of clopidogrel varies substantially with both the dose and timing of drug administration, for the purposes of our study, clopidogrel pretreatment was prospectively defined as a loading dose of ≥600 mg given at least 2 hours prior to PCI, ≥300 mg given at least 6 hours prior to PCI, or chronic clopidogrel therapy 75 mg/day for at least one week prior to PCI. Those who received lower loading doses or no clopidogrel before PCI were included in the “no pretreatment group.”
The primary end point was the composite of in-hospital death or MI. The key secondary end point was the composite of death or nonfatal MI at 1 year. Additional end points included bleeding and stent thrombosis. Periprocedural MI was defined as elevation of CK-MB of ≥3× the local upper limit of normal. Bleeding was defined as any major or minor hemorrhage according to the Thrombolysis In Myocardial Infarction criteria and also included any access site bleeding requiring clinical intervention or transfusion. Stent thrombosis was defined as definite or probable stent thrombosis according to the Academic Research Consortium definitions. Lesion complexity and other angiographic characteristics were determined by the PCI operator. All suspected clinical events (e.g., death, MI, stent thrombosis) were reviewed by 2 independent cardiologists who were unaware of the patient characteristics and treatments.
Continuous variables are described as the mean ± SD and were compared using Student’s t test (for normally distributed variables) or the Wilcoxon rank-sum test (for non-normally distributed variables). Categorical variables are described as proportions and were compared using the chi-square test or Fisher’s exact test, as appropriate.
We used the median odds ratio (MOR) to quantitatively assess the variability in receiving clopidogrel pretreatment across the study sites. The MOR is defined as the median value of the odds ratio for clopidogrel pretreatment for all possible pairs of patients between 2 randomly selected hospital sites and is independent of the overall prevalence of clopidogrel pretreatment in the study population. The MOR can be conceptualized as the increased likelihood of receiving clopidogrel pretreatment that a patient would have if presenting to another hospital. As such, it demonstrates the extent to which the probability of receiving clopidogrel pretreatment varied by hospital site.
Because clopidogrel pretreatment was not randomly assigned, we used hierarchical multivariable logistic regression models (adjusting for site as a random effect) for the in-hospital outcomes and Cox proportional hazards regression analysis for the 1-year outcomes to estimate the independent association between clopidogrel pretreatment and each clinical outcome (including the primary and secondary end points, their components, and stent thrombosis). These models were adjusted for differences in patient characteristics between the pretreatment and no pretreatment groups by including a propensity score that reflected a patient’s likelihood of receiving clopidogrel pretreatment.
The propensity score was developed using multivariable logistic regression analysis and included an extensive list of sociodemographic (age, gender, body mass index), clinical (systolic blood pressure on admission, diastolic blood pressure on admission, diabetes, hypertension, hyperlipidemia, smoker, end-stage renal disease requiring dialysis, congestive heart failure, peripheral arterial disease, history of MI, history of PCI, glomerular filtration rate [using the Modification of Diet in Renal Disease equation), history of bypass surgery), angiographic (lesion length and reference vessel diameter), and procedural (number of vessels stented, number of stents deployed, PCI of saphenous vein graft, PCI of left main coronary artery, use of glycoprotein IIb/IIIa inhibitor in hospital, worst American College of Cardiology/American Heart Association lesion class, PCI of bifurcation lesion) factors. For these models, the total implanted stent length was used as a proxy for lesion length and the minimum stent diameter as a proxy for vessel diameter. The model C-statistic was adequate at 0.71, and all variables used in creating the propensity score had a standardized difference <10 ( Supplementary Figure 1 ), indicating adequate balance between the 2 treatment groups.
In addition to our main comparison, we tested whether an interaction was present between the primary outcome and the timing of clopidogrel administration (600 mg at ≥2 hours before PCI, 300 mg ≥6 hours before PCI, or 75 mg/day for ≥1 week vs no pretreatment as the reference category) on the primary outcome of in-hospital death or MI. Statistical significance was defined as p <0.05. All statistical analyses were performed using SAS, version 9.2 (SAS Institute, Cary, North Carolina).
Results
The EVENT registry included 7,607 patients in waves 2 to 4. Of these, 2,926 underwent PCI for an acute coronary syndrome (ST-segment elevation MI, non–ST-segment elevation MI, or unstable angina without biomarker elevation), and 4,681 underwent elective PCI (stable angina or positive stress test findings). Of the elective PCI population, 1,913 patients received bivalirudin alone as the antithrombotic regimen and constituted the study population.
Of the study population, clopidogrel pretreatment was given to 923 patients (48%). The variation was marked in the practice of clopidogrel pretreatment across sites, ranging from 0% to 92% ( Figure 1 ), with a median odds ratio across sites of 2.24 (95% confidence interval 1.72 to 2.74), implying a 224% median variation in the likelihood of receiving clopidogrel pretreatment among identical patients if they presented to 1 of 2 randomly selected hospitals. Differences in the baseline demographic, clinical, angiographic, and procedural characteristics between the subjects who received clopidogrel pretreatment and those who did not are listed in Table 1 . The patients who received clopidogrel pretreatment had a slightly greater prevalence of risk factors, such as hypertension, dyslipidemia, peripheral arterial disease, previous MI, previous bypass surgery, previous PCI, and previous stroke, but generally similar angiographic and procedural characteristics, with the exception of a greater lesion complexity ( Tables 1 and 2 ).
| Variable | Clopidogrel Pretreatment | p Value | |
|---|---|---|---|
| Yes (n = 923) | No (n = 990) | ||
| Age (years) | 65.5 ± 10.8 | 65.5 ± 10.8 | 0.918 |
| Men | 623 (67.5%) | 686 (69.3%) | 0.398 |
| Diabetes mellitus | 343 (37.2%) | 361 (36.5%) | 0.738 |
| Hypertension ⁎ | 773 (83.8%) | 791 (80.1%) | 0.036 |
| Hyperlipidemia † | 795 (86.3%) | 739 (75.3%) | <0.001 |
| Current smoker/quit smoking <1 year | 194 (21.1%) | 195 (19.9%) | 0.505 |
| Peripheral arterial disease | 136 (14.9%) | 103 (10.5%) | 0.004 |
| Previous myocardial infarction | 295 (32.5%) | 183 (18.9%) | <0.001 |
| Previous stroke | 125 (13.6%) | 71 (7.2%) | <0.001 |
| Previous percutaneous coronary intervention | 535 (58.3%) | 278 (28.5%) | <0.001 |
| Previous bypass surgery | 240 (26.1%) | 214 (21.8%) | 0.031 |
| Ejection fraction <50% | 288 (37.7%) | 246 (29.0%) | <0.001 |
| Indication for percutaneous coronary intervention | <0.001 | ||
| Chronic stable angina or abnormal stress test finding | 763 (82.7%) | 876 (88.5%) | |
| Other | 160 (17.3%) | 114 (11.5%) | |
| Hemoglobin (g/dl) | 13.8 ± 4.2 | 14.3 ± 7.8 | 0.088 |
| Platelet count (×10 9 /L) | 228.2 ± 67.6 | 231.8 ± 69.5 | 0.261 |
| Creatinine (mg/dl) | 1.2 ± 1.7 | 1.4 ± 4.7 | 0.290 |
⁎ Defined as documented systolic or diastolic hypertension in patient’s medical record or current receipt of antihypertensive medication.
† Defined as documented elevation of total or low-density lipoprotein cholesterol according to medical record or current receipt of any lipid-lowering medication.
| Variable | Clopidogrel Pretreatment | p Value | |
|---|---|---|---|
| Yes (n = 923) | No (n = 990) | ||
| Intervened vessel | |||
| Left main | 24 (2.6%) | 18 (1.8%) | 0.240 |
| Left anterior descending | 400 (43.3%) | 428 (43.2%) | 0.963 |
| Right coronary | 322 (34.9%) | 337 (34.0%) | 0.697 |
| Left circumflex | 257 (27.8%) | 280 (28.3%) | 0.831 |
| Ramus intermedius | 18 (2.0%) | 25 (2.5%) | 0.396 |
| Saphenous vein graft | 76 (8.3%) | 60 (6.1%) | 0.063 |
| Lesion class | 0.037 | ||
| A | 96 (10.6%) | 109 (11.2%) | |
| B1 | 339 (37.3%) | 331 (34.0%) | |
| B2 | 307 (33.8%) | 385 (39.6%) | |
| C | 166 (18.3%) | 148 (15.2%) | |
| Preintervention Thrombolysis In Myocardial Infarction flow grade | 0.247 | ||
| 0 | 28 (3.1%) | 38 (3.9%) | |
| 1 | 25 (2.7%) | 41 (4.2%) | |
| 2 | 89 (9.7%) | 96 (9.8%) | |
| 3 | 774 (84.5%) | 804 (82.1%) | |
| Bifurcation lesion | 102 (11.1%) | 88 (8.9%) | 0.112 |
| Number of vessels treated | 0.579 | ||
| 1 | 772 (84.0%) | 846 (85.7%) | |
| 2 | 137 (14.9%) | 131 (13.3%) | |
| 3 | 10 (1.1%) | 10 (1.0%) | |
| Number of lesions treated | 0.169 | ||
| 1 | 626 (68.1%) | 712 (72.1%) | |
| 2 | 224 (24.4%) | 219 (22.2%) | |
| ≥3 | 69 (7.5%) | 56 (5.7%) | |
| Number of stents | 0.224 | ||
| 1 | 535 (58.6%) | 618 (62.7%) | |
| 2 | 270 (29.3%) | 241 (24.4%) | |
| ≥3 | 108 (11.7%) | 125 (12.6%) | |
| Total stent length (mm) | 26.7 ± 16.5 | 27.5 ± 18.2 | 0.310 |
| Minimum stent diameter (mm) | 2.9 ± 0.5 | 2.9 ± 0.4 | 0.344 |
The in-hospital and 1-year clinical outcomes are listed in Table 3 . The incidence of the primary end point (i.e., in-hospital death or MI) did not differ between the clopidogrel pretreatment and no-pretreatment groups (5.5% vs 5.8%, p = 0.83). These results were also confirmed in our propensity-adjusted analysis (adjusted odds ratio 0.91, 95% confidence interval 0.60 to 1.39). Also, no difference was seen in the incidence of the individual end point components or in protocol-defined bleeding between the 2 treatment groups.
| Outcome | Clopidogrel Pretreatment | Univariate p Value | Adjusted Odds ratio (95% CI) | Multivariable p Value ⁎ | |
|---|---|---|---|---|---|
| Yes (n = 923) | No (n = 990) | ||||
| In-hospital outcomes | |||||
| Death/myocardial infarction | 51 (5.5%) | 57 (5.8%) | 0.826 | 0.91 (0.60–1.39) | 0.657 |
| Death | 3 (0.3%) | 3 (0.3%) | 1.000 | NA † | |
| Myocardial infarction | 48 (5.2%) | 56 (5.7%) | 0.660 | 0.89 (0.57–1.35) | 0.608 |
| Composite bleeding ‡ | 9 (1.0%) | 10 (1.0%) | 0.938 | 1.38 (0.49–3.91) | 0.542 |
| 1-Year outcomes | |||||
| Death/myocardial infarction | 69 (7.5%) | 82 (8.3%) | 0.263 | 0.82 (0.58–1.16) | 0.261 |
| Death | 13 (1.4%) | 19 (1.9%) | 0.381 | 0.65 (0.31–1.35) | 0.254 |
| Myocardial infarction | 59 (6.4%) | 65 (6.6%) | 0.879 | 0.84 (0.57–1.20) | 0.378 |
| Definite or probable stent thrombosis | 3 (0.3%) | 3 (0.3%) | 0.999 | NA † | |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
