Reproduction of this document for non-commercial purposes is permitted provided appropriate credit is given to CCOHTA.

Transcription

1 Technology Report Issue 32 November 2002 The Role of Clopidogrel in the Secondary Prevention of Recurrent Ischemic Vascular Events After Acute Myocardial Ischemia: A Critical Appraisal of the CURE Trial

2 Publications can be requested from: CCOHTA , Green Valley Crescent Ottawa, Ontario, Canada K2C 3V4 Tel. (613) Fax. (613) or download from CCOHTA s web site: Cite as: Boucher M, Armstrong P, Pharand C, Skidmore B. The role of clopidogrel in the secondary prevention of recurrent ischemic vascular events after acute myocardial ischemia: a critical appraisal of the CURE trial. Ottawa: Canadian Coordinating Office for Health Technology Assessment; Technology report no 32. Reproduction of this document for non-commercial purposes is permitted provided appropriate credit is given to CCOHTA. Legal Deposit National Library of Canada ISBN: (print) ISBN: (electronic version) Publications Mail Agreement Number:

3 Canadian Coordinating Office for Health Technology Assessment The Role of Clopidogrel in the Secondary Prevention of Recurrent Ischemic Vascular Events After Acute Myocardial Ischemia: A Critical Appraisal of the CURE Trial Michel Boucher, B Pharm Dipl Bus Adm M Sc 1 Paul Armstrong, MD FACC FRCPC 2 Chantal Pharand, Pharm D BCPS 3 Becky Skidmore, BA MLS 1 November Canadian Coordinating Office for Health Technology Assessment, Ottawa, Ontario Division of Cardiology, Department of Medicine, University of Alberta and University of Alberta Hospital, Edmonton, Alberta Faculty of Pharmacy, University of Montréal and Hôpital du Sacré-Coeur de Montréal, Montréal, Québec

4

5 Reviewers These individuals kindly provided comments on this report. External Reviewers Jafna L Cox, MD FRCPC FACC Acting Director of Research, Division of Cardiology, Dalhousie University Halifax, Nova Scotia Thao Huynh, MD FRCPC McGill University Health Centre Montréal General Hospital Montréal, Québec The report was also reviewed by the manufacturers and distributors of Plavix TM in Canada; Bristol-Myers Squibb and Sanofi- Synthelabo Canada Inc. Jack Hirsh, CM MD FRCPC FRACP FRSC D Sc Director, Henderson Research Centre, Professor Emeritus, McMaster University, Hamilton, Ontario Peter J. Zed, BSc BSc (Pharm) Pharm D Pharmacotherapeutic Specialist Emergency Medicine, CSU Pharmaceutical Sciences, Vancouver General Hospital, Vancouver, British Columbia CCOHTA Scientific Advisory Panel Reviewers Ruth L. Collins-Nakai, MD MBA FRCPC FACC Chair, Scientific Advisory Panel Cardiologist, University of Alberta Edmonton, Alberta Robert Coté, MD Neurologist McGill University Division of Neurology The Montréal General Hospital Montréal, Québec This report is a review of existing public literature, studies, materials and other information and documentation (collectively the source documentation ) which are available to CCOHTA. The accuracy of the contents of the source documentation on which this report is based is not warranted, assured or represented in any way by CCOHTA and CCOHTA does not assume responsibility for the quality, propriety, inaccuracies, or the reasonableness of any statements, information or conclusions contained in the source documentation. CCOHTA takes sole responsibility for the final form and content of this report. The statements and conclusions in this report are those of CCOHTA and not of its Panel members or reviewers. i

6 Authorship Michel Boucher was the primary author, and as such was involved in all aspects of the project. He was also responsible for the primary writing of the report. Paul Armstrong and Chantal Pharand provided comments on the design of the review, approved the literature search strategy, critically revised all drafts of the report and contributed relevant content information. They were also available as advisors to the primary author during the preparation of the report. Becky Skidmore was responsible for the design and execution of the literature search strategy, for writing the methods section and associated appendix on literature searching, and for verifying and formatting bibliographic references. Disclosure of Conflicts of Interest Robert Côté has been involved in the past with Sanofi-Synthelabo Canada Inc. and Bristol-Myers Squibb Canada as a consultant as well as a speaker for continuing medical education activities. Jack Hirsh received a consultant stipend from a third party to prepare comments on the initial draft version of the current report. Disclaimer This CCOHTA report is a critical appraisal of the CURE trial. Its intent is to determine the role of clopidogrel in the management of vascular events after an episode of acute coronary syndromes. Accordingly, it does not claim to assess and discuss all efficacy and safety properties of clopidogrel. ii

7 REPORT IN BRIEF November 2002 A Critical Appraisal of the CURE Trial Technology Name Clopidogrel (Plavix TM ) Disease/Condition Acute coronary syndromes (ACS) include a type of angina (chest pain of cardiac origin) called unstable angina and a type of heart attack called non-stsegment elevation myocardial infarction. It is estimated that 50% of all heart attacks are caused by ACS. Patients with ACS may be classified as being at high-, intermediate- or low-risk of non-fatal heart attack or death from cardiac causes. The 30-day risk of these cardiac complications after an episode of ACS is >10% in the high-risk group and < 2% in the low-risk group. Technology Description Clopidogrel belongs to a newer class of antiplatelet medications called thienopyridines that work differently than the traditionally used acetylsalicylic acid (ASA), also known as aspirin. Clopidogrel, in combination with ASA, was recently approved in Canada for the treatment of ACS. The Issue A recent clinical study, (the CURE trial) compared the use of the clopidogrel/asa combination to ASA alone in 12,562 patients with ACS. This treatment was used to prevent vascular events such as stroke, heart attack or death from cardiac causes. The results of the CURE trial showed a lower rate of a cluster of these vascular events in patients treated with the clopidogrel/asa combination after an episode of ACS, compared to patients treated with ASA alone. This suggests a possible first-line role for the more expensive combination of clopidogrel/asa in patients with ACS. Prior to the release of the CURE trial, ASA was considered the first choice of treatment for most patients with ACS and clopidogrel as an alternative. This was partially based on the results of a clinical study released in 1996, the CAPRIE trial. Assessment Objectives To examine the use of the clopidogrel/asa combination in the prevention of vascular events after an ACS; 1. to determine the population most likely to benefit from this intervention, and 2. to assess the efficacy and safety of this intervention in this population Methodology A critical appraisal of the CURE trial was conducted. However, to ensure that no other potentially relevant clinical studies that addressed the objectives of our report were available, a literature search was performed. No such clinical studies were identified, other than the previously mentioned CAPRIE trial. A Canadian representative of the manufacturer of clopidogrel was also invited to submit relevant information. Conclusions The population studied in the CURE trial largely included ACS patients at high risk of cardiac complications. The results provide evidence that the early addition of clopidogrel to ASA reduces subsequent cardiovascular morbidity, compared to ASA alone, in this specific group of patients. In the CURE trial, this clinical benefit was mainly due to a reduced number of non-fatal heart attacks. However, patients experienced a higher risk of bleeding complications. This summary is based on a comprehensive health technology assessment available from CCOHTA s web site ( Boucher M, Armstrong P, Pharand C, Skidmore B. The role of clopidogrel in the secondary prevention EXECUTIVE of recurrent ischemic SUMMARY vascular events after acute myocardial ischemia: a critical appraisal of the CURE trial. Background Canadian Coordinating Office For Health Technology Assessment (CCOHTA) Green Valley Crescent, Ottawa, ON, Canada K2C 3V4 Tel: Fax: Ex CCOHTA is an independent, non-profit health research agency funded by the federal, provincial and territorial governments. iii

8 EXECUTIVE SUMMARY Background Clopidogrel belongs to a relatively new class of antiplatelet medications called thienopyridines. These medications inhibit platelets through a mechanism different to that of acetylsalicylic acid (ASA). Clopidogrel, in combination with ASA, was recently approved in Canada for the treatment of acute coronary syndromes (ACS), including unstable angina (UA) and non-stsegment elevation myocardial infarction (MI). Non-ST elevation ACS patients can be classified as being at high, intermediate or low risk of fatal or non-fatal MI. While the 30-day risk of these events is greater than 10% in the high-risk group, it is less than 2% in the low-risk group. It is estimated that 50% of MIs are non-st-segment elevation MIs. The combination of ASA and clopidogrel for the treatment of ACS stems from the results of the recent Clopidogrel in Unstable angina to prevent Recurrent Events (CURE) trial. This trial reported a lower rate of vascular events with the clopidogrel/asa combination compared to placebo/asa treatment, when these medications were used to prevent vascular events after an ACS. This suggests a possible first-line role for clopidogrel in this clinical setting. Prior to the CURE trial, ASA was recommended as the antiplatelet agent of choice for most patients with ACS, while clopidogrel was considered as an alternative. This was partially based on the results of the Clopidogrel versus Aspirin in Patients at Risk of Ischaemic Events (CAPRIE) trial. As the population groups and treatment strategies in the two trials differed somewhat, it is important to clearly identify whether specific patient groups would benefit from the clopidogrel/asa combination treatment for the prevention of subsequent vascular events. Objectives The present analysis examines the use of the clopidogrel/asa combination in the secondary prevention of recurrent ischemic vascular events after an ACS. The two objectives in this clinical setting are: Methods 1) to determine the population most likely to benefit from the use of the combination of clopidogrel and ASA, and 2) to assess the clinical benefits and harm of clopidogrel/asa. A literature search was performed to confirm that no clinical trials other than the CAPRIE and CURE trials have been conducted to assess the efficacy and safety of clopidogrel using the prevention of vascular events after myocardial ischemia as a primary endpoint. BIOSIS Previews, EMBASE, MEDLINE, PASCAL and ToxFile were searched on the DIALOG system. No other such trials were identified. A Canadian representative of the manufacturer of clopidogrel was also invited to submit relevant information. iv

9 Results The CURE trial was critically appraised. CURE was a double-blind, randomized trial which compared the outcomes of 12,562 patients with non-st-segment elevation ACS treated with either clopidogrel 75 mg daily or placebo, in addition to ASA (recommended dose, 75 to 325 mg daily) for 3 to 12 months (mean duration of treatment, 9 months). There were two primary outcomes. The first was a composite outcome of cardiovascular death, non-fatal MI, or stroke. The second was the composite of the first primary outcome or refractory ischemia. The majority of the patients recruited in the CURE trial were considered to be at high risk of cardiac ischemia and necrosis, based on electrocardiogram (ECG) changes and/or elevation of cardiac enzymes. There was a 2.1% (95% confidence interval (CI): 1%, 3.2%) absolute risk reduction (ARR) in the occurrence of the first primary outcome in the clopidogrel/asa group. The clopidogrel/asa combination was also associated with a 2.3% (95% CI: 1%, 3.6%) ARR in the occurrence of the second primary outcome. These outcomes were largely driven by a statistically significant reduction in the incidence of non-fatal MIs. There were no statistically significant differences between the two groups in the incidence of stroke, cardiovascular death and non-cardiovascular death. However, these single endpoints should be interpreted with caution, as the CURE trial was only powered to detect a significant difference for the two primary composite outcomes. There was a 1% (95% CI: 0.4%, 1.6%) absolute risk increase (ARI) in major bleeding episodes in the clopidogrel/asa group, compared to the control group. When only major bleeding complications requiring the transfusion of two or more units of blood were considered, the clopidogrel/asa combination was associated with a 0.7% ARI (95% CI: 0.1%, 1.2%). In decreasing order of magnitude, higher incidences of major bleeding were observed for gastrointestinal hemorrhages, followed by bleeding at arterial puncture sites and, to a lesser extent, surgical sites. Although serious, these would be expected to be reversible and non-fatal in most cases. The dose of ASA used may also be a factor. Finally, a nearly significant higher rate of major bleeding complications was observed in patients undergoing coronary artery bypass graft (CABG) surgery if clopidogrel was not stopped at least five days before the procedure (relative risk (RR), 1.53, P = 0.06). Conclusion The population studied in the CURE trial mainly included patients with non-st elevation ACS who were at high risk of cardiac ischemia and necrosis, based on electrocardiogram (ECG) changes and/or elevation of cardiac enzymes. The results of the trial provide evidence that the early addition of clopidogrel to ASA in this population reduces subsequent cardiovascular morbidity, compared to use of ASA alone. In the CURE trial, this clinical benefit was mainly driven by a reduction in incidence of non-fatal MIs. Patients receiving the clopidogrel/asa regimen, however, were exposed to a higher risk of bleeding complications. v

10 TABLE OF CONTENTS EXECUTIVE SUMMARY... iv ABBREVIATIONS... vii 1 INTRODUCTION Technology Overview Clinical Background Therapeutic Considerations The CURE and CAPRIE Trials The CURE trial The CAPRIE trial OBJECTIVES METHODS RESULTS Assessment of the Quality of the Methods Used in the CURE Trial Randomization Sample size Outcomes Drop-outs and withdrawals Review of the Results of the CURE Trial Primary outcomes Secondary outcomes Safety DISCUSSION CONCLUSION REFERENCES Appendix 1: Literature Search Strategy Appendix 2: Jadad Scale vi

11 ABBREVIATIONS ACC: ACS: AHA: ARI: ARR: ASA: CABG: CAD: CEA: CI: CK-MB: ECG: ER: GP: HF: ITT: MI: NNH: NNT: PAD: PCI: RR: RRI: RRR: TnT: UA: American College of Cardiology Acute coronary syndromes American Heart Association Absolute risk increase Absolute risk reduction Acetylsalicylic acid Coronary artery bypass graft Coronary artery disease Cost-effectiveness analysis Confidence interval Creatine phosphokinase (MB isoenzyme) Electrocardiogram Emergency room Glycoprotein Heart failure Intention-to-treat Myocardial infarction Number needed to harm Number needed to treat Peripheral arterial disease Percutaneous coronary intervention Relative risk Relative risk increase Relative risk reduction Troponin T Unstable angina vii

12 1 INTRODUCTION 1.1 Technology Overview Clopidogrel is a member of a relatively new class of antiplatelet medications called thienopyridines, as is the drug ticlopidine. 1 These drugs exert their platelet inhibition effect through a mechanism different from that of acetylsalicylic acid (ASA), which inhibits platelet aggregation through irreversible acetylation and inactivation of the cyclooxygenase enzyme. 1 Thienopyridines irreversibly inhibit platelet aggregation by preventing adenosine diphosphatemediated structural alterations in the platelet glycoprotein (GP) IIb/IIIa receptor, thereby inhibiting platelet binding to fibrinogen. 1 Clopidogrel is manufactured and distributed in Canada through a joint venture between Sanofi- Synthelabo Canada Inc. and Bristol-Myers Squibb, under the trade name Plavix TM. It is approved for the secondary prevention of atherothrombotic events [myocardial infarction (MI), stroke and vascular death] in patients with atherosclerosis documented by stroke, MI, or established peripheral arterial disease (PAD). Recently, Health Canada approved clopidogrel (in association with ASA and other standard therapies) for the reduction of atherothrombotic events (MI, ischemic stroke, cardiovascular death and/or refractory ischemia) in patients with acute coronary syndromes (ACS), unstable angina (UA) or non-q-wave MI, without ST-segment elevation Clinical Background Atherosclerotic heart disease is a common cause of cardiovascular disability and death. Epidemiological studies have identified several risk factors for coronary artery disease (CAD) including, among several others, serum lipid abnormalities. 3 Lipids contribute to the formation of atherosclerotic plaques on vessel walls. Many atherosclerotic plaques remain stable or only progress gradually, leading to a progressive narrowing of the vessel lumen. This contributes to reduced oxygen delivery to the heart, which may lead to angina pectoris and potentially MI. 3 However, coronary artery stenosis occurs slowly and is associated with the development of collateral blood vessels over time, which may protect against MI. More frequently, MIs are associated with the development of a coronary artery thrombus. 4 In most cases, this occurs when an atherosclerotic plaques fissures, or ruptures, leading to the formation of a mural thrombus at the site of rupture and ultimately to coronary artery occlusion. 4 Patients with retrosternal chest pain may present with or without ST-segment elevation on an electrocardiogram (ECG). ST-segment elevation is caused by the total occlusion of the coronary artery, which leads to an acute MI. Most patients with this presentation will have a Q-wave MI, while a few will develop a non-q-wave MI. When the obstructing thrombus is not totally occlusive, obstruction is transient, or when a rich collateral network of blood vessels is present, ST-segment elevation is not seen on the ECG. Such patients present with UA or non-st-segment elevation MI. 4 Since these disorders share the common physiology of atheromatous plaque rupture or fissure with thrombus formation, but typically without complete vessel occlusion as is 1

13 seen with ST-segment elevation MI, and because they are often indistinguishable in the acute setting, UA and non-st-segment elevation MI are often considered together. 1 They are also known as non-st elevation ACS. 3 A distinction between UA and non-st-segment elevation MI is their respective degree of association with myocardial ischemia and necrosis. Patients with non-st elevation ACS may be considered to have experienced UA in the absence of an elevation in biochemical markers of myocardial necrosis such as the MB isoenzyme of creatine phosphokinase (CK-MB) or troponin T (TnT). When such markers of myocardial injury are released, the diagnosis of non-st-segment elevation MI is considered. 5 For most cases of non-st elevation ACS, progression to MI leads to a non-q wave MI, with only a minority progressing to Q-wave MI. 5 Q-wave MIs and non Q-wave MIs differ in the extent of myocardial necrosis they cause; the former are usually considered more severe than the latter. 4 It has been estimated that as many as 50% 6 and possibly more than 85% 7 of acute MIs involve thrombus formation precipitated by atherosclerotic plaque rupture or fissure. Other data indicate that 40.3% of all MIs are non-st-segment elevation MIs. This information was obtained from a chart review of 4,856 patients admitted for an MI in 44 hospitals in Ontario between 1994 and [MIs associated with left bundle branch block and patients with paced cardiac rhythms were excluded (Chau Tran, Institute for Clinical Evaluative Sciences, Toronto: personal communication, 2002 Jun)]. Based on the information available, it may be projected that at least 50% of MIs are non-st-segment elevation MIs. Non-ST elevation ACS carries a high burden. UA and non-st-segment elevation MI are lifethreatening disorders and a major cause of emergency visits and hospitalization. In 1997, there were 5,315,000 visits to emergency rooms (ERs) in the US for the evaluation of chest pain and related symptoms. In 1996, there were 1,433,000 hospitalizations in the US for UA and non-stsegment elevation MI. 8 In Canada, there were 200 hospital separations (end of hospital stay due to discharge or death) per 100,000 population attributable to MI between 1989 and Because non-st elevation MIs account for 50% of all MIs, this means approximately 100 hospital separations per 100,000 population may be attributed to this type of MI in Canada. 1.3 Therapeutic Considerations Although MI patients with ST-segment depression have a better early survival (5 days) than those with ST-segment elevation, their long-term (6 months) mortality may be worse. 10 This observation justifies aggressive therapy. 1 Such therapy includes acute management and longterm management. Long-term management is based partially on interventions including smoking cessation, lipid management, hypertension control and treatment of diabetes. 10 Short-term management of patients with ST-segment elevation MI differs from the management of patients with UA or non-st-segment elevation MI. In the first case, patients may be considered for immediate reperfusion therapy with thrombolytic agents or percutaneous coronary intervention (PCI). 11 In the latter, patients are routinely investigated to assess whether myocardial injury has occurred. 10 2

14 Proper stratification of patients with ACS, based on particular risk factors, is important. Generally speaking, ACS patients are stratified as being at high, intermediate or low risk of acute ischemia (Figure 1). 8,10 The 30-day risk of fatal or non-fatal MI is 12 to 30% in the first group, 4 to 8% in the second and less than 2% in the third group. 10 The intensity of drug therapy can be related to risk stratification. 8,10 ASA monotherapy and investigation of myocardial ischemia over a six to eight hour observation period in the ER are recommended for the lower-risk group, or patients with possible ACS. However, the use of ASA and antithrombin agents, either in the form of unfractionated heparin or low-molecular-weight heparin, along with possible consideration for cardiac catheterization, are recommended for the intermediate-risk group. 10 For the high-risk group, the combination of ASA and antithrombin agents, in conjunction with intravenous GP IIb/IIIa inhibitors, such as abciximab, eptifibatide or tirofiban, as well as early cardiac catheterization, have been recommended. 8,10 Figure 1: Risk stratification of patients presenting with non-st elevation ACS 10 High-risk: Prolonged chest pain (> 20 min) or ongoing, with at least one high-risk feature such as: transient ST-segment elevation or depression, or sustained ST-segment depression positive biomedical markers (abnormal troponin/ck-mb serum levels) recurrent myocardial ischemia with ST-segment shift with or without chest pain acute MI in past four weeks hemodynamic compromises (heart failure or hypotension) with chest pain Intermediate-risk: No high-risk features, but one or more of the following: ongoing chest pain but no high-risk features crescendo angina preceding rest pain borderline positive troponin serum levels previous intervention such as coronary artery bypass graft (CABG) increased baseline risk: diabetes, age Low-risk: No high- or intermediate-risk features: single episode of chest pain at rest, or crescendo exertional angina normal or non-specific abnormalities or unchanged ECG from previous one Experts recommend that patients with definite ACS should also be considered for anti-ischemic therapy, including oxygen therapy, nitrates and beta-blockers. Clopidogrel is considered an alternative to ASA for patients unable to take ASA, but may be added to ASA for hospitalized patients, unless these patients are scheduled for CABG surgery. 8 3

15 1.4 The CURE and CAPRIE Trials The CURE trial Combined use of ASA and clopidogrel stems from the results of the recent Clopidogrel in Unstable angina to prevent Recurrent Events (CURE) trial. 8,10,12,13 Prior to the release of the results of the CURE trial, the American College of Cardiology/American Heart Association (ACC/AHA) guidelines for the management of patients with UA and non-st-segment elevation MI recommended ASA as the antiplatelet agent of choice for most patients. Clopidogrel was only recommended in patients unable to take ASA due to hypersensitivity to this drug or due to major gastrointestinal intolerance. 5 The purpose of the CURE trial was to evaluate the efficacy and safety of the clopidogrel/asa combination, compared to ASA monotherapy, in 12,562 patients with ACS without ST-segment elevation. 14 Patients were eligible if they had been hospitalized within 24 hours after onset of symptoms. They were followed for 3 to 12 months (mean duration of treatment, 9 months). The CURE trial reported a 20% relative risk reduction (RRR) in the rate of vascular events (defined as a composite outcome of cardiovascular death, non-fatal MI, or stroke) when the combination therapy was used, as opposed to patients treated with ASA alone. 14 The results of the CURE trial indicated that clopidogrel, in combination with ASA, may be considered as a first-line agent for the secondary prevention of recurrent ischemic vascular events in patients with non-st elevation ACS, i.e. myocardial ischemia presumably mediated by thrombosis, rather than being considered as an alternative to ASA. Also, the potential population for such combination therapy may be considered relatively large, given that at least 50% of all MIs may be non-st-segment elevation MIs. It is therefore important to clearly identify whether specific patient groups would benefit from the clopidogrel/asa combination treatment for the prevention of subsequent vascular events The CAPRIE trial To provide some perspective to the CURE study, results from the Clopidogrel versus Aspirin in Patients at Risk of Ischaemic Events (CAPRIE) trial were examined as well. Results of this trial were published in One of its primary objectives was to evaluate the efficacy and safety of clopidogrel in the secondary prevention of recurrent vascular events after a recent MI, an objective similar to that of the CURE trial. The CAPRIE trial recruited 19,185 patients and compared clopidogrel 75 mg daily and ASA 325 mg daily. Patients were divided in three different subgroups: those with ischemic stroke, MI and atherosclerotic PAD. For the MI subgroup, randomization had to occur within 35 days of the onset of symptoms. Only 17% of the patients recruited into each group with MI as a qualifying event had UA. Patients were followed for one to three years for recurrent vascular events (mean follow up 1.91 years). Accordingly, both the population groups and treatment strategies in the CAPRIE trial were somewhat different from those studied in the CURE trial. 4

16 The intention-to-treat (ITT) analysis of the primary outcome cluster of ischemic stroke, MI, or vascular death showed an overall RRR of 8.7% (95% confidence interval (CI): 0.3%, 16.5%) in favour of clopidogrel. 15 This represents an absolute risk reduction (ARR) of 0.5% (95% CI: 0.0%, 1.0%) and a number needed to treat (NNT) of 199 patients (95% CI: 102 to 4,198). No statistically significant differences were observed in the subgroup analysis of patients who suffered an MI (clopidogrel: 5.03%, ASA: 4.84%, p: 0.66). 15 However, the CAPRIE trial was only powered to detect a realistic treatment effect in the whole cohort, not in individual subgroups, 15 therefore interpretation of subgroup analyses conducted in this trial should be cautious. Several safety parameters were evaluated in the CAPRIE trial. These are summarized in Table 1. Table 1: CAPRIE trial Adverse events 15 Adverse event Clopidogrel (%) ASA (%) Rash * Diarrhoea * Indigestion, nausea, * vomiting Any bleeding disorder Intracranial bleeding Gastrointestinal * haemorrhage Abnormal liver function * * Statistically significant, p < 0.05 For other cardiology-related indications, clopidogrel has also been used in an effort to prevent thrombosis in patients undergoing PCI with stenting. 1 However, evaluation of the value of clopidogrel in this clinical setting is outside the scope of our report. 5

17 2 OBJECTIVES The present analysis examines the use of the clopidogrel/asa combination in the secondary prevention of recurrent ischemic vascular events after an ACS. The two objectives in this clinical setting are: 1) to determine the population most likely to benefit from the use of the combination of clopidogrel and ASA, and 2) to assess the clinical benefits and harm of clopidogrel/asa. 3 METHODS A literature search was performed to confirm that no clinical trials other than the CAPRIE and CURE trials have been conducted to assess the efficacy and safety of clopidogrel using the prevention of vascular events after myocardial ischemia as a primary endpoint. Published literature was obtained by searching a number of databases (Appendix 1). On the DIALOG system, BIOSIS Previews, EMBASE, MEDLINE, PASCAL and ToxFile were searched, resulting in 599 unique records. No language restrictions were applied. Searches were performed and updated on the CD ROM version of The Cochrane Library. Grey literature was obtained through searching a number of clinical trial registries as well as the web sites of health technology assessment and near-technology assessment agencies and their associated databases. Further information was sought by hand searching the bibliographies of selected papers. The manufacturers and distributors of clopidogrel in Canada, Sanofi-Synthelabo Canada Inc. and Bristol-Myers Squibb, were contacted to obtain the most recent product monograph of Plavix TM. They were also invited to submit relevant information that could assist CCOHTA in the current review. Since no other trials were found, the authors undertook a critical appraisal of the CURE trial. This included an overall assessment of the quality of the trial, using the Jadad scale (Appendix 2). This scale is composed of three items related directly to the reduction of bias (randomization, double-blinding and study withdrawals and drop-outs). A score is given for each of the three items, for a maximum of five points. Allocation concealment is also considered in the assessment with ratings of adequate, unclear or inadequate. 16 To facilitate the interpretation of the results of the CURE trial, values for ARR, ARI, NNT and number needed to harm (NNH) were calculated for statistically significant results. Consistent with current methodology, 95% CIs were also calculated for each of these values. 17,18 These were calculated using the CIA statistical package (version 2.0.0) from Altman et al. 19 6

18 4 RESULTS Apart from the CAPRIE and CURE trials, and excluding revascularization studies, no other completed clinical trials were identified which addressed the objectives of our report. However, the literature search identified one ongoing trial. This trial, known as the COMMIT study, is currently being carried out in China and is randomizing 30,000 participants to either ASA or a combination of ASA and clopidogrel. 1 This trial is also known as the second Chinese Cardiac Study (CCS-2). 20,21 The study aims to determine whether adding clopidogrel to ASA for up to four weeks in hospital after suspected acute MI further reduces the risk of major vascular events, compared to using ASA alone. The effect of metoprolol, a beta-blocker, is also studied, following a 2 x 2 factorial design. To be randomized, patients must present within 24 hours of the onset of symptoms of suspected acute MI and have supporting ECG abnormalities (STsegment elevation, ST-segment depression or bundle branch block). The trial began in July 1999 and is expected to be completed by the year Assessment of the Quality of the Methods Used in the CURE Trial The CURE trial compared the outcomes of patients with non-st-segment elevation ACS treated with either clopidogrel 75 mg daily or placebo, in addition to ASA (recommended dose, 75 to 325 mg daily), for 3 to 12 months (mean duration of treatment, 9 months). At trial entry, patients were given a loading dose of clopidogrel 300 mg, followed by clopidogrel 75 mg daily (or placebo) Randomization The CURE trial was randomized, double-blind and placebo-controlled. We determined the randomization procedure was well described and appropriate. It was accomplished with a central, 24-hour, computerized randomization service. Permutated-block randomization, stratified according to clinical centers, was used. Double-blinding was maintained by using matching placebo for both the initial loading dose of 300 mg of clopidogrel and the daily maintenance dose of 75 mg. The study was coordinated by the Canadian Cardiovascular Collaboration Project Office located at McMaster University in Hamilton, Ontario. A steering committee consisting of national coordinators oversaw the study. Investigators were supported by a research grant from Sanofi- Synthelabo and Bristol-Myers Squibb Sample size The initial sample size of 9,000 participants for this trial was calculated based on an expected rate of events for the control group of 12 to 14%. However, because the rate of events appeared to be lower than had originally been expected, the sample size of the study was increased after its initiation. Sample size was recalculated assuming a 10% rate of events for the placebo/asa 7

19 group for the first primary outcome and a rate of 14% for the second primary outcome. This led to a new sample size of 12,500 patients, which provided 90% power to detect a 16.9% reduction in risk for the first primary outcome and 16.4% for the second primary outcome. In addition to the recalculation of the sample size, a further modification of the study protocol occurred. Initially, patients older than 60 years of age with no new ECG changes but with a history of CAD were enrolled. However, after a review of the overall rate of events among the first 3,000 patients, the steering committee recommended that only patients with ECG changes or elevation of biomedical markers be enrolled. Trial participants were recruited between December 1998 and September 2000 at 482 centres in 28 countries. The ethics review board of each institution approved the study Outcomes There were two co-primary outcomes. The first was the composite of cardiovascular death, nonfatal MI, or stroke. The second was the composite of the first primary outcome or refractory ischemia. The secondary outcomes were severe ischemia, HF and the need for revascularization. 14 These, either considered as composite or individual outcomes, are relevant to the main study research question, i.e. to determine the role of long-term clopidogrel/asa therapy in ACS patients. The safety-related outcomes available in the published article reporting the results of the CURE trial were limited to bleeding complications and some hematological parameters. 14 Bleeding complications were categorized as life-threatening, major or minor. Major bleeding episodes were defined as substantially disabling bleeding, intraocular bleeding leading to the loss of vision, or bleeding necessitating the transfusion of at least two units of blood. Major bleeding was classified as life-threatening if (a) the bleeding episode was fatal or led to a reduction in the hemoglobin level of at least five grams per deciliter or to substantial hypotension requiring the use of intravenous inotropic agents, (b) it necessitated a surgical intervention, (c) it was a symptomatic intracranial hemorrhage, or (d) it necessitated the transfusion of four or more units of blood. Minor bleeding episodes included other hemorrhages that led to the discontinuation of the study medication. 14 Hematological parameters monitored during the trial were thrombocytopenia and neutropenia Drop-outs and withdrawals The analysis for all endpoints was appropriately based on the ITT principle. However, no detailed information was provided regarding study drop-outs and withdrawals other than reporting in the statistical analysis section of the publication that six patients in the clopidogrel/asa group and seven in the placebo/asa group were lost to follow-up. Regarding withdrawals, it is reported that 21.1% of patients in the clopidogrel/asa group and 18.8% in the control group permanently discontinued their study medication. 14 The reasons for dropping out or withdrawing from the study were not specified in the manuscript. This information was however clarified as part of the current review. The main 8

20 reason for drop-out was patient refusal to continue participation in the trial. The main reason for permanently discontinuing study medications was also patient refusal. The next most common reason was adverse events (Dr. Salim Yusuf, Canadian Cardiovascular Collaboration Project Office, McMaster University, Hamilton, ON: personal communication, 2002 Sep). Accordingly, application of the quality assessment tool, the Jadad scale (Appendix 2), assigned a quality rating of 5/5 for this trial. It was not possible to assess whether treatment allocation was concealed for all patients. 4.2 Review of the Results of the CURE Trial A review of the baseline characteristics of the CURE trial participants determined that both treatment groups were comparable Primary outcomes The first primary outcome, which included death from cardiovascular causes, non-fatal MI, or stroke, occurred in 582 of the 6,259 patients (9.3%) on the clopidogrel/asa combination as compared with 719 of the 6,303 patients (11.4%) on the placebo/asa regimen. This corresponds to a RR of 0.80 (95% CI: 0.72, 0.90) in favour of the clopidogrel/asa combination. 14 It also translates to an ARR of 2.1% (95% CI: 1%, 3.2%), which corresponds to an NNT of 47 patients (95% CI: 32, 96) to prevent one occurrence of the first primary composite outcome over an average period of nine months. The results of the second primary outcome, which included either the first primary outcome or refractory ischemia, also favoured the clopidogrel/asa combination. This composite outcome was observed in 1,035 patients in the clopidogrel/asa group (16.5%) and in 1,187 patients (18.8%) in the control group, leading to a RR of 0.86 (95% CI: 0.79, 0.94) in favour of the clopidogrel/asa combination. 14 This corresponds to a 2.3% ARR (95% CI: 1%, 3.6%) and an NNT of 44 patients (95% CI: 28, 104) to prevent one occurrence of the second primary composite outcome over an average period of nine months. The rate of occurrence of most of the components of the primary composite outcomes tended to be lower in the clopidogrel/asa group. The largest effect was observed for non-fatal MIs, with a RR of 0.77 (95% CI: 0.67, 0.89) in favour of the clopidogrel/asa combination. 14 This corresponds to a 1.5% ARR (95% CI: 0.6%, 2.3%) and an NNT of 68 patients (95% CI: 44, 155) to prevent one non-fatal MI over an average period of nine months. There were no statistically significant differences between the two groups for the other single components of the primary outcomes, or in the incidence of death from non-cardiovascular causes. However, given that the CURE trial was not powered to detect differences between the two treatment groups for individual endpoints, interpretation should be cautious. Finally, the rate of the first primary outcome was lower in the clopidogrel/asa group within the first 30 days after randomization (RR: 0.79; 95% CI: 0.67, 0.92). This effect sustained between 30 days and the end of the study (RR: 0.82, 95% CI: 0.70, 0.95). 14 Table 2 presents the incidence of the main study outcomes. 9

21 Table 2: CURE trial Incidence of the main study outcomes 14 Outcome Clopidogrel/ASA (%) ASA (%) RR (95% CI) Primary composite outcomes First primary (0.72, 0.90) outcome Second primary outcome (0.79, 0.94) Individual components of the primary composite outcomes Cardiovascular (0.79, 1.08) death Non-fatal MI (0.67, 0.89) Stroke (0.63, 1.18) Refractory ischemia (0.82, 1.04) Other endpoints Non-cardiovascular death (0.60, 1.39) Secondary outcomes As was found when examining the primary outcomes, there was also a trend in favour of the clopidrogrel/asa group for the secondary outcomes (severe ischemia, recurrent angina, coronary revascularization procedures, HF). Potential additional clinical benefits may be seen with the clopidogrel/asa combination in terms of reducing the number of coronary revascularization procedures during the initial period of hospitalization, as well as the incidence of HF following ACS. However, additional research is required to confirm such potential benefits. Also, subgroup analysis showed that there was a tendency toward a greater benefit from the clopidogrel/asa combination among patients who had previously undergone revascularization [RR (first primary outcome): 0.56, 95% CI: 0.43, 0.72] than among those who had not (RR: 0.88, 95% CI: 0.78, 0.99; p for interaction = 0.002). 14 However, authors of the CURE trial indicate that this result should be interpreted with caution, given the large number of subgroup analyses performed. Table 3 presents the incidence of secondary outcomes. Table 3: CURE trial Incidence of secondary outcomes 14 Outcomes Clopidogrel/ASA (%) ASA (%) RR (p value) Severe ischemia (p = 0.003) Recurrent angina (p = 0.01) Revascularization procedures during: a) the entire study b) the initial hospitalization N/A* (p = 0.03) HF (p = 0.03) * N/A: not available 10

22 4.2.3 Safety There was no statistically significant difference in the proportion of life-threatening bleeding episodes in the clopidogrel/asa group (2.2%), compared to the placebo/asa group (1.8%), RR of 1.21 (95% CI: 0.95, 1.56). However, major bleeding episodes were significantly more common in the clopidogrel/asa group (3.7%), compared to the control group (2.7%), RR of 1.38 (95% CI: 1.13, 1.67). 14 This corresponds to an ARI of 1% (95% CI: 0.4%, 1.6%) for the clopidogrel/ ASA group, compared to ASA monotherapy. The NNH in this case was 99 patients (95% CI: 61, 253). If only the patients with major bleeding requiring the transfusion of two or more units of blood are considered, the rates were 2.8% in the clopidogrel/asa group and 2.2% in the control group, leading to a significant RR of 1.30 (95% CI: 1.04, 1.62) for patients on the clopidogrel/asa combination. 14 This corresponds to an ARI of 0.7% (95% CI: 0.1%, 1.2%) and an NNH of 153 patients (95% CI: 83, 927) to cause one additional serious adverse bleeding event requiring the transfusion of two or more units of blood within an average of nine months. The incidence of minor bleeding was also higher in the clopidogrel/asa group (5.1%), compared to the ASA monotherapy group (2.4%), leading to a significant RR of 2.12 (95% CI: 1.75, 2.56). 14 This corresponds to an ARI of 2.7% (95% CI: 2.1%, 3.4%) and a NNH of 37 patients (95% CI: 29, 49) for the clopidogrel/asa combination. In total, 533 patients in the clopidogrel/asa group (8.5%) and 317 patients in the control group (5%) suffered from a bleeding complication. 14 This corresponds to an ARI of 3.5% (95% CI: 2.6%, 4.4%) and an NNH of 29 patients (95% CI: 23, 38) for the clopidogrel/asa combination. Table 4 summarizes the bleeding complications in the CURE trial. Table 4: CURE trial Incidence of bleeding complications 14 Outcome Clopidogrel/ASA ASA (%) RR (95% CI) (%) Any bleeding (1.48, 1.94) complications Minor bleeding (1.75, 2.56) Major bleeding (1.13, 1.67) Major bleeding (1.04, 1.62) requiring > 2 units of blood Life-threatening bleeding (0.95, 1.56) A nearly significant higher rate of major bleeding complications was observed in patients undergoing CABG surgery if clopidogrel was not stopped at least five days before the procedure (9.6% in the clopidogrel/asa group versus 6.3% in the control group; relative risk (RR): 1.53, p = 0.06). There was however no apparent excess in major bleeding within seven days after surgery in patients who interrupted their clopidogrel treatment more than five days before surgery (4.4% in the clopidogrel/asa group versus 5.3% in the control group). 14 The number of patients with thrombocytopenia or neutropenia was similar in both groups. In the clopidogrel/ 11

23 ASA group, there were 26 cases of thrombocytopenia and eight cases of neutropenia whereas 28 and 5 cases were reported respectively, in the placebo/asa group. 14 The published version of the CURE trial limited safety assessment reporting to the monitoring of bleeding complications and hematological abnormalities (thrombocytopenia and neutropenia). However, unpublished data from the CURE trial reveal that there were a higher proportion number of patients in the clopidogrel/asa combination group (5.2%) than in the ASA/placebo group (3.5%) who permanently discontinued study medications because of other adverse events. (Dr. Salim Yusuf: personal communication, 2002 Sep). 12

24 5 DISCUSSION The initial CURE study protocol was modified on two occasions. Such modifications are usually acceptable, provided they are clearly reported in the publication, which was the case for CURE. The first modification, which involved recalculating the sample size, led to a larger sample size than originally planned. This should have intrinsically provided additional validity to the results. However, the second modification, which involved a tightening of the entry criteria for the patients 60 years of age and older, carries the potential for different consequences. Indeed, because the entry criteria were revised after the first 3,000 participants (24% of the total study population) had been recruited, it is possible that some patients were enrolled without having ACS. This may have resulted in an increase in the clinical heterogeneity of the trial population, which could reduce the internal validity of the results. On the other hand, this may also have led to an underestimation of the potential efficacy of the clopidogrel/asa combination since patients who may not have had an ACS were exposed to similar risk while deriving less benefit from the intervention. Fortunately, the entry criteria were revised relatively early so the remainder of the trial population was not affected by the initial more flexible selection criteria. In total, there were 624 patients in the study, out of 12,562 enrollees, who had normal ECGs (Dr. Salim Yusuf: personal communication, 2002 Sep). As these patients represent less than five percent of the ITT study population, they should have minimal impact on the overall results. It may however be derived that, since over 95 percent of the recruited participants were at higher risk of cardiac ischemia and necrosis, based on the ECG and/or cardiac enzyme elevation, the results of the CURE trial mainly apply to a higher-risk population. The magnitude of the benefit (composite outcome of cardiovascular death, non-fatal MI, or stroke) observed in the population studied in the CURE trial corresponds to an NNT of 47 patients (95% CI: 32, 96). When refractory ischemia was considered, in addition to the first primary endpoint, the magnitude of benefit was comparable, with an NNT of 44 patients (95% CI: 28, 104). Although the rate of occurrence of most components of this composite endpoint tended to be lower in the clopidogrel/asa group, statistical significance was only reached for non-fatal MIs. Although the CURE trial was not powered to detect a realistic effect on single components of the primary outcomes, there were no statistically significant differences in the incidence of stroke, cardiovascular death, or non-cardiovascular death despite recruiting over 12,500 participants. Accordingly, based on the CURE trial, the morbidity benefit associated with the clopidolgrel/asa combination in higher-risk non-st elevation ACS patients mainly consists of a reduction in the number of non-fatal MIs. Because MIs were about four to five times more frequent than strokes in the CURE trial, it may be observed that non-st elevation ACS patients are at higher risk of having future MIs than they are of having subsequent strokes. Consequently, the prevention of non-fatal MI in this population is an important clinical benefit. The clinical benefit observed however, for the clopidogrel/asa combination, needs to be interpreted in light of the associated increased bleeding risk. In particular, for every 99 patients treated with the clopidogrel/asa combination, one patent suffered a major bleeding episode, 13

25 when compared to the control group (NNH: 99; 95% CI: 61, 253). In order to capture both the efficacy and harm associated with the clopidolgel/asa combination, Albers and Amarenco combined major bleeding episodes and major vascular events into a single endpoint. A RRR of 8%, favouring the clopidogrel/asa combination over ASA, was found with such an evaluation. 22 This was reported to correspond to an absolute net benefit of 1.1%, with a calculated odds ratio (OR) of 1.10 (95% CI: 0.99, 1.22). 22 Another way of considering both the morbidity benefit and the bleeding complications of the clopidogrel/asa combination, compared to ASA alone, is the following. About one non-st elevation ACS patient in 50 (NNT: 47) will avoid a vascular event (cardiovascular death, nonfatal MI, or stroke). However, about one in 100 (NNH: 99) will suffer a major bleeding episode. This means for every two high-risk non-st elevation ACS patients who will benefit from the clopidogrel/asa combination, one will suffer a major bleeding episode. It may be concluded that the clopidogrel/asa combination provides increased efficacy with respect to the prevention of vascular events compared to ASA alone, but this benefit is partly offset by an increased risk of bleeding. However, in making such a determination, it is important to also consider the clinical consequence of the bleeding complications. The sites where a higher incidence of major bleeding was observed with the clopidogrel/asa combination, compared to ASA, were gastrointestinal (1.3% vs 0.7%), followed by arterial puncture sites (0.6% vs 0.3%) and surgical sites (0.9% vs 0.8%). 14 Although serious, these would be expected to be reversible and non-fatal in most cases. For example, upper gastrointestinal bleeding is generally associated with a mortality rate of 1 to 10% in the US and Europe. 23 Another consideration is that the incidence and severity of gastrointestinal bleeding associated with ASA is generally considered to be dose-related, 24 although a recent meta-analysis showed no relation between gastrointestinal hemorrhage and ASA dose. 25 A retrospective analysis of the relationship between ASA doses used in the CURE trial and bleeding complications was recently conducted. It reported that, for major or life-threatening bleeding, adjusted ORs for highest (> 300 mg) versus lowest (< 100 mg) daily dose of ASA were 1.7 (95% CI: 0.9, 3.0) for the ASA/placebo group and 1.5 (95% CI: 0.9, 24) for the clopidgrel/asa group, suggesting that bleeding risks in the CURE trial were also related to the dose of ASA. The incidence of minor bleeding did not show a clear ASA dose response. 26 Finally, in the CURE trial, a higher proportion of patients on the clopidogrel/asa combination permanently discontinued their study medication because of adverse events other than bleeding or hematological abnormalities, compared to the control group. It was not possible to obtain additional information regarding these other adverse events. In the CAPRIE trial, a higher proportion of patients in the clopidogrel group reported rash and diarrhea, compared to patients on ASA (Table 1). The ACC/AHA guidelines for administration of antiplatelet therapy in patients with UA and non- ST-segment elevation MI were revised in March The increased morbidity benefit as well as the bleeding concerns described above were considered in these revised recommendations: 14