The 12-month results of CREDO (clopidogrel for the Reduction of Events During Observation) were presented at AHA to coincide with the publication of this important clinical trial in the November 20, 2002, issue of JAMA.1 CREDO was designed to assess the benefit of long-term dual antiplatelet therapy with heparin and aspirin in patients undergoing elective percutaneous coronary intervention (PCI). The study also addressed the effect of the timing (3-24 hours before PCI) of a pre-procedural clopidogrel loading dose on the 28-day outcomes after PCI. The results of the 28-day analysis were previously presented at the Transcatheter Cardiovascular Therapeutics meeting in September.
The study included patients with symptomatic coronary artery disease (CAD) and objective evidence of ischemia (angina, abnormal noninvasive testing, or electrocardiogram changes) who were referred for PCI or angiography with a high likelihood of undergoing PCI with stent implantation. Patients with contraindications to antiplatelet therapy, or who had received clopidogrel within 10 days or a glycoprotein (GP) IIb-IIIa inhibitor within 7 days, were excluded from the study, as were those with significant left main CAD, ST elevation, or thrombolytic therapy within 24 hours, recent unsuccessful PCI, or lesion morphology not amenable to stenting. Patients were randomized to active therapy (clopidogrel) or placebo; however, patients in both groups received clopidogrel (75 mg per day) for 28 days after stent implantation as standard therapy for prevention of sub-acute stent thrombosis. Patients in the active therapy group received a loading dose of clopidogrel (300 mg orally) 3-24 hours prior to PCI, and postprocedure clopidogrel was continued "long-term" for 1 year. Patients in the control group received preprocedural placebo and after 28 days of oral clopidogrel, received placebo until the end of the 12-month treatment period. All patients received aspirin at a dose of 325 mg per day through day 28, after which they received 81-325 mg per day at the discretion of the investigator. GP IIb-IIIa inhibitor use was allowed in 20% of patients prespecified at the time of PCI and as "bail out" therapy for any patient at the operator’s discretion. The primary end point at 1 year was the composite of death, MI, and stroke. The primary end point at 28 days was the composite of death, MI, or target vessel revascularization (TVR) in the per-protocol population (those patients who underwent PCI). Secondary end points included the individual components of each composite, TVR or any revascularization at 1 year, and major bleeding events (TIMI bleeding criteria). The 28-day data were also evaluated with respect to the timing of clopidogrel loading dose administration (greater than vs less than 6 hours) prior to PCI.
A total of 2116 patients were enrolled from 99 centers in the United States and Canada. The groups were well matched in terms of baseline demographic and procedural characteristics. Eighty-six percent of randomized patients in each group ultimately underwent PCI as planned. There were nonsignificant trends toward lower rates of calcium channel blocker and statin use in the clopidogrel group. Overall GP IIb-IIIa inhibitor (primarily abciximab) use was somewhat higher in the clopidogrel group (47.4% vs 43.3%, P = 0.08). In both groups, roughly half of the patients receiving GP IIb-IIIa inhibitors received them for a "bail-out" indication. Sixty-three percent of clopidogrel-treated patients and 61% of placebo-treated patients had completed the full 1-year course of treatment per protocol. The most frequent reasons for treatment discontinuation in both groups were "patient choice" and "adverse event."
At 28 days, there was no significant overall difference in the composite of death, MI, or TVR between the patients pretreated with clopidogrel vs placebo. However, when stratified with respect to time from pretreatment to procedure, patients pretreated with clopidogrel more than 6 hours before PCI showed a 38.6% reduction in events when compared to placebo (5.8% vs 9.4%, P = 0.05). Subgroup analysis demonstrated a lack of benefit for late pretreatment in all subgroups analyzed. At 28 days, there was no difference in bleeding complication rates between the groups.
At 1 year, patients treated with clopidogrel experienced a significant 26.9% relative reduction and a 3% absolute reduction in the risk of death, MI, or stroke (8.5% vs 11.5%, P = 0.02). This difference was due largely to events occurring between 28 days and 1 year. Similar reductions were achieved in individual components of the end point, including a 24.6% reduction in death and a 20.8% reduction in MI, though these did not reach statistical significance. The degree of benefit for clopidogrel was also consistent (though not always statistically significant) across subgroups analyzed, including patients with and without diabetes, acute coronary syndromes, or concomitant GP IIb-IIIa inhibitor therapy. There were no differences in terms of rates of TVR or any revascularization at any time during follow-up.
There was a trend toward increased major bleeding in clopidogrel-treated patients (8.8% vs 6.7%, P = 0.07), with most bleeding events occurring in patients undergoing invasive procedures, particularly coronary artery bypass graft surgery (CABG). In fact, more than half of all patients undergoing CABG in CREDO experienced a major bleeding event, which might, in part, be attributable to the fact that clopidogrel was not routinely withheld preoperatively, as is the current recommended practice.
In their discussion, the authors acknowledge that this study was limited with respect to the issue of the effect of duration of pretreatment by the small sizes and lack of randomization of subgroups. They acknowledge that the single randomization makes it difficult to separate the effect of clopidogrel pretreatment from the effect of long-term clopidogrel administration. They also remark upon the high proportion of patients in both groups who did not complete the full course of treatment according to the protocol, which, if improved, might have strengthened the results of this study.
Comment by Sarah M. Vernon, MD
The interventional cardiology community is coming to view atherosclerosis not simply in the context of "fixing" the particular lesion in question, but as a systemic disease that carries with it long-term risk. The results of CREDO suggest that the continuation of clopidogrel long term after PCI (probably in addition to earlier oral loading of this drug preprocedure) can result in substantially improved outcomes even in a relatively low-risk (elective PCI) population with coronary artery disease. The results of this study extend the results of CURE2 and its substudy PCI-CURE,3 which evaluated a similar strategy in higher-risk patients with acute coronary syndromes. While the magnitude of benefit in the PCI-CURE, in terms of 6-month MACE (cardiovascular death, MI, or urgent TVR) was similar, its protocol was extremely conservative, with a low percentage of patients undergoing PCI and, most notably, with a 10-day (median) period of clopidogrel administration prior to delayed PCI. The study design of CREDO, with the brief duration between clopidogrel administration and PCI, makes this strategy more applicable to clinical practice in the United States in 2002.
The results of CREDO are compelling, if not convincing. CREDO would suggest that there are probably real benefits to initiating clopidogrel more than 6 hours prior to PCI. However, in the average clinical practice, the strategy of preprocedural loading carries with it some logistical implications, particularly in light of the substantive bleeding risk associated with CABG. It can be difficult, in some cases, to accurately predict who will undergo PCI (and, more importantly, who will not undergo CABG) based on clinical and noninvasive data before a diagnostic angiogram. Fortunately, most of these patients will not require emergent or urgent surgery, in which case, the risk of bleeding can be substantially reduced by discontinuing clopidogrel at least 5 days prior to scheduled surgery per the AHA/American College of Cardiology guidelines.
Mechanistically, dual antiplatelet therapy makes sense, given that the thienopyridines and aspirin block separate, independent routes of platelet activation, perhaps increasing their anti-inflammatory, as well as antithrombotic effect. An additive effect in combination with the glycoprotein IIb-IIIa inhibitors, which inhibit platelet aggregation, as was seen in this study, also seems highly plausible. It also stands to reason that chronic therapy might have a substantial impact on a chronic systemic disease. However, significant questions remain, including the best regimen (dose and timing ) for expediency and optimal antiplatelet effect at the time of PCI, and, perhaps more importantly, the optimal duration of therapy after the procedure, or for patients with established CAD in general. Given the cost and safety implications of long-term clopidogrel administration in an extremely large population of eligible patients, it is important that we address these questions. In their discussion, the authors mention 2 important ongoing trials that may address these issues: ISAR-REACT, which will evaluate a 600-mg clopidogrel loading dose prior to PCI with or without a glycoprotein IIb-IIIa inhibitor (abciximab), and CHARISMA, which will evaluate a large group of high-risk, but stable patients with CAD receiving aspirin (ASA) or ASA with clopidogrel, with anticipated 42 months of follow-up.
Dr. Vernon is Assistant Professor of Medicine, Director, VAMC Cardiac Catheterization Laboratory University of New Mexico Health Sciences Center, Albuquerque, NM.
1. Steinhubl SR, et al. JAMA. 2002;288:2411-2420.
2. The clopidogrel in unstable angina to prevent recurrent events trial investigators: Effects of clopidogrel in addition to aspirin in patients with acute coronary syndromes without ST-segment elevation. N Engl J Med. 2001;345:494-502.
3. Mehta SR, et al. Effects of pretreatment with clopidogrel and aspirin followed by long-term therapy in patients undergoing percutaneous coronary intervention: The PCI-CURE study. Lancet. 2001;358:527-533.