Role of Plavix in Hemorrhagic and Ischemic Complications of Catheterization.

SIGNIFICANCE AND BACKGROUND FOR THE STUDY The benefits of dual anti-platelet therapy with full dose aspirin and a thienopyridine have been firmly established in patients undergoing coronary stenting. Leon showed reduction in the combined end-point of death, myocardial infarction, target vessel revascularization and stent thrombosis from 3.5% with aspirin alone to 2.3% with aspirin and ticlopidine. This benefit was associated, however, with an increased rate of bleeding and vascular complications of 5.5% with aspirin+ticlopidine versus 1.8% with aspirin alone. Multiple other studies have evaluated the safety and efficacy of the thienopyridine clopidogrel in percutaneous coronary intervention (PCI). In patients at intermediate risk (without elevation of troponin-T prior to PCI), pretreatment with clopidogrel several hours before the procedure reduced the event rate and obviated the need for additional glycoprotein IIb/IIIa receptor inhibition (ISAR-REACT). In the Clopidogrel for the Reduction of Events During Observation (CREDO) Study, investigators showed that the benefit of pretreatment with 300 mg of clopidogrel was only seen when the pretreatment was instituted at least 12 hours pre-procedure (optimally 24 hours before the PCI). However, survival analysis in this study showed a cumulative event rate at 30 days of 8.3% in the placebo arm, 7.8% in those administered clopidogrel within 15 hours of the procedure, and 3.5% in those administered clopidogrel more that 15 hours before the index PCI.

Subsequently, the benefit of higher doses of clopidogrel was studied both ex vivo and in clinical studies (PRONTO). A 600 mg dose of clopidogrel achieved adequate platelet inhibition within 2 hours of administration, unlike 300 mg that required at least 3 to 6 hours to achieve full effect Subsequent data from the ISAR-REACT investigators in 2,159 patients showed that the benefit of 600 mg of clopidogrel was not time dependent. The trial focused on the possible additional benefit of glycoprotein IIb/IIIa inhibition in these patients who had all been pre-treated with 600 mg of clopidogrel. More recently, ISAR-REACT-2 showed that clopidogrel pretreatment in moderate to high risk PCI patients reduced the combined endpoint event rate (relative to placebo) from 11.7% to 5.7% in patients treated with glycoprotein IIb/IIIa inhibitors and from 15.5% to 7.6% in those not receiving glycoprotein IIb/IIIa inhibition. The recently published Assessment of the Best Loading Dose of Clopidogrel to Blunt Platelet Activation, Inflammation and Ongoing Necrosis (ALBION) Trial showed in a low risk group of patients that 600 mg and 900 mg doses of clopidogrel provide faster and greater platelet inhibition than 300 mg as assessed by biochemical assays. The study was underpowered to demonstrate clinical benefits, but it showed a trend towards reduction in troponin-I release as a marker of necrosis from 58% to 42% in the higher dose clopidogrel groups. There was no demonstrable increase in major bleeding complications (2.9% in both 300 and 900 mg groups), but there was a statistically insignificant increase in minor bleeding in the 900 mg dose group.

Our investigation will evaluate the safety and efficacy of pre-treatment (≥2 hours) with 600 mg of clopidogrel in 600 outpatients undergoing elective cardiac catheterization with coronary angiography in our institution who have no evidence of high risk features such as ischemic ST segment deviations, unstable angina, recent myocardial infarction, or abnormal troponin-T levels. The benefit of clopidogrel in patients who undergo interventions may be offset by the bleeding complications encountered, especially in patients who do not undergo coronary angioplasty. Our trial is designed to assess this potential benefit-risk trade-off. Unlike the ALBION trial, we will focus on clinical outcomes using a larger sample size. Our hypothesis is that pretreatment with 600 mg of clopidogrel will significantly reduce the incidence of ischemic complications in patients undergoing PCI, but with possible increase in hemorrhagic and vascular complications in the overall population.

In addition, we will also examine whether the use of 5 French arterial sheaths in cases that do not go onto intervention offsets the increase in bleeding risk, if any, associated with pretreatment with 600 mg of clopidogrel. Very large diameter arterial sheaths (e.g., 10 French) have been associated with higher vascular complication rates, but whether 5 French sheaths offer lower vascular complication rates than 6 French sheaths is uncertain. Nearly all PCIs at our institution are performed using 6 French sheaths (which have larger lumens, more accommodating of angioplasty equipment and higher injection rates of iodinated contrast). Patients who undergo PCI following diagnostic angiography using a 5 French sheath will require exchange for a larger 6 French arterial sheath. On the other hand, patients who do not need angioplasty after coronary angiography will not require any exchange of their arterial sheath, and those with smaller sheath sizes may well have less vascular and bleeding complications.

DESCRIPTION OF RESEARCH PROTOCOL Design: This is a prospective randomized clinical trial comparing ischemic, hemorrhagic and vascular outcomes in outpatients undergoing elective cardiac catheterization with coronary angiography following 1:1 randomization to pretreatment with 600 mg clopidogrel vs. no pretreatment (and clopidogrel dosing post-procedure if a PCI is performed). Patients will provide written informed consent prior to enrollment and randomization. Since some operators may feel strongly about using glycoprotein IIb/IIIa inhibition in patients who did not receive clopidogrel well in advance of the PCI, blinding of the loading strategy and use of a placebo are not feasible. A separate 1:1 randomization will assign patients to undergo cardiac catheterization using 5 or 6 French arterial introducing sheaths. Since the arterial sheath sizes are color-coded, catheter sizes are imprinted on each catheter, and the operator must know the arterial sheath size in case alternate catheters are required, blinding of sheath size assignment is also not feasible. The investigator adjudicating and abstracting the outcomes data such as bleeding and ischemic complications for analysis will be blinded, however, to the clopidogrel reimen and randomization in order to avoid any bias.

Hypotheses to be tested: The primary hypothesis being examined is that clopidogrel 600 mg given ≥2 hours prior to cardiac catheterization will reduce the frequency of post-procedural ischemic events in the subset of patients who undergo ad hoc coronary interventions compared with those who do not receive clopidogrel pre-procedure. The risk of hemorrhagic and vascular complications by preloading strategy will be evaluated in all patients randomized (encompassing those who do and those who do not undergo PCI). Additional analyses will be performed to see if the size of the initial arterial introducing sheath also has an effect on the rate of hemorrhagic and vascular complications and if there is any interaction of sheath size with preloading strategy on these outcomes.

Patient population: All outpatients undergoing elective cardiac catheterization with planned coronary angiography with possible ad hoc PCI (if lesion anatomy suitable) and who have not received clopidogrel or ticlopidine during the 14 days prior to the procedure are eligible. Patients with high-risk clinical features, contraindications to clopidogrel, or planned upcoming invasive procedures will be excluded (vide infra).

Procedural factors: The decision to perform ad hoc PCI after diagnostic coronary angiography (vs. medical therapy or bypass surgery) will be made by the interventional cardiologist performing or overseeing the cardiac catheterization, in consultation with the patient's clinical cardiologists. The specific angioplasty techniques, devices, equipment, anti-thrombin and anti-platelet pharmacological regimen to be utilized are NOT specified by this protocol and are at the discretion of the interventional cardiologist.

Patients randomized to 5 French arterial sheaths and NO clopidogrel preloading will be allowed to ambulate 2 hours after arterial hemostasis is achieved (with delayed ambulation if bleeding or orthostatic hypotension occurs). All other participants (i.e., 6 French arterial sheaths, or 5 French arterial sheaths with clopidogrel preloading) will be allowed to ambulate no earlier than 4 hours after hemostasis occurs.

Variables of interest include baseline demographic and clinical characteristics (including age, gender, left ventricular ejection fraction, diabetes mellitus, hypertension, dyslipidemia, chronic renal insufficiency, serum creatinine level, tobacco use, extent of coronary disease, symptom severity, prior myocardial infarction, prior PCI, prior CABG), angiographic lesion characteristics (e.g., bifurcation location, lesion class), procedural factors (e.g., pharmacological antithrombin regimen, glycoprotein IIb/IIIa inhibitor use), and post-procedural tests (including serum creatinine, serum creatine kinase and its MB isoform, troponin-T, and radiological imaging studies). Information will be gleaned from the Cardiac Catheterization Laboratory Quality Assurance Database, from electronic medical records, from review of the procedural angiograms, and conversations with the patient (or if the patient cannot be located, their family or health care providers).

Endpoints: All endpoints will be assessed at 14 days post-procedure (and divided for analytical purposes into in-hospital and post-discharge events). See Endpoint section of this document for details.

Follow-up: Written and electronic medical records will be reviewed to identify clinical events occurring during the hospitalization (through the 1st 14 days post-procedure if the length of stay is prolonged). Participants will have provided written informed consent for clinical follow-up at 14 days post-procedure prior to entry into this study, and contact information for the patient, his/her primary care provider, his/her cardiologist, and his/her next of kin will be collected before discharge. The patient, his/her physicians, or his/her next of kin will be contacted by telephone at 14 or more days post-procedure and asked if the patient was treated after discharge for any heart-related, bleeding, or vascular access site problems (i.e., any of the clinical events detailed above). If yes, attempts will be made, with the patient's authorization, to obtain more information about the event, including requests for documentation from other health care facilities and providers.

Data Monitoring and Safety Plan: A independent data safety monitoring committee (DSMC) comprised of Murray Mittleman, MD, DrPH (BIDMC cardiovascular epidemiologist, serving as Chair), David J. Cohen, MD, MSc (interventional cardiologist at the Mid America Heart Institute), and Shiva Gautam, PhD (BIDMC biostatistician) will review the rates of the 4 composite endpoints enumerated above after 14 day follow-up has been completed on the 1st 300 subjects enrolled. The DSMC will advise the principal investigator whether the study should terminated because of concerning and striking differences in the ischemic or vascular/bleeding complication rates between treatment strategies or if the study should be allowed to continue to the planned enrollment of 600 participants. Study recruitment will be suspended during the DSMC's review and deliberations.

Data Management: Much of the data required for this study is routinely collected as part of current clinical care, quality assurance and regulatory processes and stored in the Cardiac Catheterization Laboratory Quality Assurance Database (residing a commercial application named Apollo and manufactured by Lumedx). The fields of interest for patients randomized in this study will be queried form the Apollo database and transferred directly to the Access database that serves as the primary electronic data repository for this study. This database will be housed on a network server within the BIDMC firewall with restricted access (password and valid user identification required). Additional information (from medical records or patient contacts) will be collected on paper case report forms prior to entry into the Access database. All case report forms and paper clinical information with patient identifiers will be stored in a locked file drawer within a locked office. Angiograms will be reviewed using clinical workstations in physically secured areas, and electronic copies of angiograms will be made with all patient identifiers electronically removed. Medical record numbers and names will be removed prior to creation of final analytical dataset. All paper and electronic information will be destroyed in a HIPAA-compliant fashion after completion and publication of this study.

Statistical Considerations Randomization: A randomization scheme will be prepared using 1:1:1:1 ratios (for the 2 clopidogrel dosing strategies combined independently with 5 vs. 6 French arterial sheaths) with permuted blocks of random sizes of 4 or 8. The treatment assignment will be designated in opaque, tamper-resistant randomization envelopes provided by the Statistical Department of the Harvard Clinical Research Institute.

Statistical Analysis: All statistical analyses will be performed using SAS for Windows v9. The primary analysis will compare the cumulative incidence at 14 days of the primary composite endpoint (any death, myocardial infarction, or myocardial necrosis) in patients with PCI attempted between those randomized to clopidogrel loading vs. those randomized to standard clopidogrel administration using the Fisher's exact test. A p value of 0.05 or less will be required for statistical significance. The secondary endpoint of stent thrombosis among those who undergo stent deployment will also be evaluated using Fisher's exact test. The other endpoints (including vascular and bleeding) will be examined in all patients randomized (and not just subsets who had PCI attempted or stents deployed). Two group comparisons (clopidogrel loading vs. standard dosing; 5 French vs. 6 French arterial sheaths) and 4 group comparisons (the 4 combinations of clopidogrel dosing strategy and arterial sheath size; or subgroups with PCI or no PCI by clopidogrel dosing strategy) will be performed using Fisher's exact test, t-tests, ANOVA, or analogous non-parametric tests as appropriate. Since multiple exploratory analyses will be performed, p-values for everything other than the primary analysis delineated above must be interpreted with caution. Logistical regression models will also be constructed to identify pre-procedural predictors of ischemic and of bleeding or vascular complications.