Effect of Clopidogrel on the Rate and Functional Severity of Stroke Among High Vascular Risk Patients
A Prespecified Substudy of the Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management and Avoidance (CHARISMA) Trial
Background and Purpose— Disabling stroke is costly and considered by some patients a fate worse than death. We aimed to determine whether clopidogrel reduces the rate and functional severity of stroke among high vascular risk patients, including patients with previous transient ischemic attack or ischemic stroke, who were enrolled in the Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management and Avoidance (CHARISMA) trial.
Methods— We randomly assigned 15 603 high vascular risk patients to receive clopidogrel (75 mg daily) or placebo in addition to background acetylsalicylic acid and followed them for a median of 28 months. The main outcome of this prespecified substudy was the functional severity of stroke outcome events as measured by the modified Rankin Scale (mRS) score at 3 months after the stroke outcome.
Results— During follow-up, 436 (2.8%) patients had a definite adjudicated stroke and a follow-up assessment of the mRS at 3 months poststroke, of whom 202 had been randomly assigned clopidogrel and 234 placebo (relative risk reduction 14%, 95% CI: −4% to 29%, P=0.12). There was no significant difference between the mean mRS scores at 3 months after stroke among patients assigned clopidogrel compared with placebo (mean mRS 3.6 [SD 2.4] clopidogrel versus 3.3 [SD 2.1] placebo; P=0.15). There was also no significant difference between the various categories of the mRS score at 3 months after stroke among patients assigned to clopidogrel compared with placebo. Among 4320 patients with a qualifying diagnosis of transient ischemic attack or ischemic stroke, 233 (5.4%) experienced a stroke during follow-up, of whom 103 were randomly assigned clopidogrel and 130 placebo (relative risk reduction 20%, 95% CI: −3% to 38%). There was no significant difference between the mean mRS scores at 3 months after stroke among patients with a qualifying transient ischemic attack or ischemic stroke who were assigned clopidogrel compared with placebo (3.4 [SD 2.1] clopidogrel versus 3.3 [SD 1.9] placebo; P=0.48).
Conclusion— The addition of clopidogrel to acetylsalicylic acid did not significantly alter the rate and functional severity of stroke outcome events among high vascular risk patients enrolled in the CHARISMA trial.
Much of the burden of stroke can be attributed to its high rates of case fatality (20%) and dependency (50% of survivors).1–3 Strategies for stroke prevention such as antiplatelet therapy are likely to be optimally effective if they can reduce fatal and disabling strokes at least as effectively as they prevent nondisabling strokes.
The Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management and Avoidance (CHARISMA) trial was designed to explore the hypothesis that long-term treatment with a combination of clopidogrel and acetylsalicylic acid (ASA) may be more effective than ASA alone in preventing the occurrence of the composite of stroke, myocardial infarction, or death due to vascular causes among a broad population of patients at high risk for vascular events.4,5 The rate of stroke, myocardial infarction, or vascular death was 6.8% among patients assigned clopidogrel and 7.3% among patients assigned placebo (relative risk: 0.93, 95% CI: 0.83 to 1.05).5
A secondary a priori hypothesis was that the addition of clopidogrel to ASA may reduce the severity of stroke outcome events, perhaps by minimizing the size of thrombus formed on ruptured/eroded atherosclerotic plaque and thus the size of emboli to the brain and resultant brain infarction. However, it was also possible that adding clopidogrel to aspirin could increase the severity of any hemorrhagic strokes.
In this prespecified substudy of the CHARISMA trial, we aimed to determine the effect of clopidogrel on the functional severity of stroke as measured by the modified Rankin Scale (mRS)6,7 among all high vascular risk subjects randomized in the CHARISMA trial and followed-up for the occurrence of stroke events. In addition to the prespecified analysis, we also assessed the rate and severity of stroke events in those patients recruited after a transient ischemic attack (TIA) or stroke.
The methods of the CHARISMA trial have been described in detail elsewhere.4,5 Briefly, CHARISMA was a multicenter, multinational, randomized, parallel-group, double-blind trial of clopidogrel versus placebo in high-risk patients at risk of atherothrombotic events and who were receiving low dose ASA at the time of randomization. A total of 15 603 patients with either clinically established cardiovascular disease or multiple risk factors were randomly assigned to receive clopidogrel (75 mg per day) plus low-dose ASA (75 to 162 mg per day) or placebo plus low-dose ASA. Patients with a previous disabling condition such that they were bedridden or demented were excluded from the trial. Patients were followed-up prospectively at 1, 3, and 6 months, and every 6 months thereafter until the trial end. The median follow-up period was 28 months. The primary efficacy end point for the CHARISMA trial was a composite of stroke, myocardial infarction, or death from cardiovascular causes.
The population for the substudy on stroke severity consisted of all high vascular risk patients randomized in the CHARISMA trial who experienced at least 1 adjudicated stroke end point during follow-up and underwent an assessment of the functional severity of the stroke outcome event at 3 months after the event.
The primary outcome measure was functional severity of the first stroke outcome event as defined by the mRS score at 3 months after the stroke.6,7
All randomized subjects who experienced at least 1 stroke end point as determined by the investigator were evaluated using the mRS approximately 3 months (or at study end date or death if these occurred earlier) after the stroke to determine the severity of the stroke and the patient’s functional outcome. Assessments were made after each stroke end point, but only the assessment after the initial adjudicated stroke was used for the primary analysis.
The assessment and scoring of the mRS was done by the same observer immediately after the stroke and at the 3 months poststroke follow-up assessment. The mRS was assessed incorporating all activities that the patients could actually do, not what they thought they could do. The difference between a mRS score of 2 (independent) and 3 (dependent) was the need for assistance in any activities of daily living (eg, walking, bathing, dressing, grooming).
Data were analyzed on an intention-to-treat basis. Kaplan–Meier curves were generated to show time to first adjudicated stroke in each treatment group. The mRS score was summarized by treatment group using counts and percentage as well as mean, SD, minimum, and maximum.
In the patients who had an assessment of the mRS at 3 months after their stroke, analysis of variance was used to test the difference in mean scores between treatment groups. An adjusted analysis was also performed using analysis of variance because the population was chosen based on an end point and could not be assumed to be random. Adjustment factors included whether the patient was on the study drug at the time of stroke (yes/no), time from randomization to stroke (continuous days), and the initial mRS score. The poststroke mRS score was also assessed on patients who did not permanently discontinue the study drug before the functional outcome assessment.
Only observed values were used in the analysis and presentations, that is, no attempt was made to impute missing values.
Among the 15 603 high vascular risk patients randomized to placebo (n=7801) or clopidogrel (n=7802), in addition to background ASA, and followed for a median of 28 months, 503 were reported by the investigator to have had a stroke. Of these, 441 strokes were adjudicated as a stroke. The Figure is a Kaplan–Meier curve showing the rate of stroke over the first 30 months of follow-up according to the random treatment allocation at Time 0. Among the 441 patients with adjudicated stroke, 436 had an assessment of the mRS 3 months after their stroke.
Stroke Outcomes According to Randomized Treatment Group
Among the 436 patients who had an adjudicated definite stroke during trial follow-up and underwent an assessment of functional stroke severity at 3 months poststroke, 202 had been randomly assigned clopidogrel and 234 assigned aspirin (relative risk reduction 14%, 95% CI: −4% to 29%, P=0.12, 2-tailed).
Baseline Demographic and Other Prognostic Factors
Table 1 shows that there were no significant differences in the prevalence or level of baseline demographic and other prognostic factors between the two treatment groups of patients who experienced a stroke during follow-up.
Duration of Study Drug Treatment
The 234 patients assigned placebo (who subsequently experienced a stroke) were treated with study drug for 17.1 months (median) compared with 16.1 months (median) for the 202 patients assigned clopidogrel (P=0.43).
Time Between First Adjudicated Stroke and Follow-Up Assessment of Functional Stroke Severity
The median time between stroke and follow-up assessment of functional stroke severity was 163 days for patients on clopidogrel versus 130 days for patients on placebo (P=0.03).
Pathological Subtype of Stroke Outcomes
Among the 436 patients who had a stroke outcome, 369 (85%) were ischemic (171 [84.7%] among patients assigned clopidogrel versus 198 [84.6%] among patients assigned placebo), 41 (9%) were hemorrhagic (20 [9.9%] clopidogrel versus 21 [9.0%] ASA), and 26 (6%) were of uncertain pathological type (11 [5.4%] clopidogrel versus 15 [6.4%] ASA).
Functional Outcome 3 Months After the First Stroke Outcome Event
Table 2 shows that there was no significant difference between the mean mRS scores at 3 months after stroke among patients assigned clopidogrel compared with placebo (3.6 [SD 2.3] clopidogrel versus 3.3 [SD 2.1] placebo; P=0.15). After adjusting for treatment group, on study drug at the time of stroke, time from randomization to stroke, and initial mRS score, similar results remained with no difference between treatment groups.
Table 3 shows that there was also no significant difference between the various categories of the mRS score at 3 months after stroke among patients assigned to clopidogrel compared with placebo. Although patients assigned clopidogrel showed a small trend toward more severe stroke than placebo in each comparison, it is highly probable that this finding reflects random error (chance). These results were consistent in patients who did not permanently discontinue the study drug (n=261; clopidogrel: n=122, placebo: n=139) before the poststroke functional outcome assessment (P=0.09).
Patients With Qualifying TIA or Ischemic Stroke
Among the 15 603 high vascular risk patients randomized in the main CHARISMA trial, 4320 patients were enrolled with a qualifying diagnosis of documented cerebrovascular disease (TIA [n=1233] or ischemic stroke [n=3245]; 158 patients with ischemic stroke also had a history of TIA), of whom 2163 were assigned placebo and 2157 clopidogrel.
An adjudicated first stroke during follow-up (ie, a recurrent stroke for patients who qualified with stroke) occurred in 233 patients, of whom 103 were randomly assigned clopidogrel and 130 to placebo (relative risk reduction 20%, 95% CI: −3% to 38%). Most strokes were ischemic (n=202 of 233 [87%]; 91 patients assigned clopidogrel versus 113 on placebo). A few strokes were hemorrhagic (n=19 of 236 [8%]; 10 clopidogrel versus 9 placebo). Only 12 strokes were of unknown type.
Table 4 shows that there was no significant difference between the mean mRS scores 3 months after (recurrent) stroke among patients with a qualifying TIA or ischemic stroke who were assigned clopidogrel compared with placebo (3.4 [SD 2.1] clopidogrel versus 3.3 [SD 1.9] placebo; P=0.48).
The principal finding of this substudy of the CHARISMA trial was that there was no significant difference in the rate of stroke, or severity of stroke, as measured by mRS score between high vascular risk patients randomly assigned clopidogrel (plus ASA) or placebo (plus ASA) and followed-up for a median of 28 months.
A strength of this study is that is was a large randomized controlled trial in which a large number of strokes occurred in a broad range of high vascular risk patients and were evaluated by assessors who were blind to the treatment allocation.
A weakness of the study is that although the hypotheses and analyses were prespecified, it was a substudy and not the primary aim of the trial. Patients were included in this analysis of stroke severity on the basis of an end point, and consequently it is not possible to assume that there was an equal balance in important prognostic factors between the treatment groups (ie, the comparison is not random). However, Table 1 indicates there was no significant difference in the prevalence of measured baseline demographic and other important prognostic factors for stroke severity between the treatment groups.
A second potential weakness is that the mRS was not assessed at the time of randomization (ie, at baseline), introducing the possibility that there was some imbalance in functional status at baseline due to chance among patients randomized to the 2 treatment groups. However, patients were not eligible for study entry if they were severely disabled, thus minimizing the potential for important imbalance in stroke severity at baseline.
Third, the mRS was assessed later after the stroke end point among patients assigned clopidogrel than those assigned placebo (163 days median for patients assigned clopidogrel versus 130 days for patients assigned placebo, P=0.03), introducing the possibility of bias favoring the outcome of patients assigned clopidogrel (ie, longer time to recover between stroke end point and functional assessment).
Fourth, the mRS is quite insensitive to small, but sometimes clinically meaningful, changes in functional status.6
Fifth, the mRS was assessed by a wide range of health professionals, some of whom may not have been well trained in, or familiar with, its use. Many were not stroke physicians. However, many studies have shown that the mRS has high interobserver reliability.6
Finally, the study is likely to have been underpowered to reliably identify or exclude a modest but clinically important effect of combination antiplatelet therapy on stroke severity compared with monotherapy.
This trial fails to provide evidence that adding clopidogrel to ASA significantly reduces the severity of stroke outcome events, as measured by the mRS, 3 months after stroke.
Sources of Funding
The CHARISMA trial was sponsored and funded by Sanofi-Aventis and Bristol-Myers Squibb.
G.J.H. has received consulting fees from Sanofi-Aventis, Bristol-Myers Squibb, Bayer, and Boehringer Ingelheim and lecture fees from Sanofi-Aventis, Bristol-Myers Squibb, Bayer, and Boehringer Ingelheim. W.H. is a member of the executive Committee of CHARISMA and has received consulting fees from Sanofi-Aventis and Bristol-Myers Squibb. J.D.E. has received consulting fees from Sanofi-Aventis and Bristol-Myers Squibb. J.-L.M. has received consulting fees from Sanofi-Aventis and Bristol-Myers Squibb. D.L.B. has received research grants (directly to the institution) from Astra Zeneca, Bristol Myers Squibb, Eisai, Ethicon, Heartscape, Sanofi Aventis, and The Medicines Company. K.A.A.F. has received consulting fees from Sanofi-Aventis; lecture fees from Sanofi-Aventis and Bristol-Myers Squibb; and grant support from Sanofi-Aventis. E.J.T. has received research support from Sanofi-Aventis and Bristol-Myers Squibb.
- Received April 6, 2010.
- Accepted May 14, 2010.
Hankey GJ, Jamrozik K, Broadhurst RJ, Forbes S, Anderson CS. Long-term disability after first-ever stroke and related prognostic factors in the Perth Community Stroke Study. Stroke. 2002; 33: 1034–1040.
Hankey GJ, Spiesser J, Hakimi Z, Bego G, Carita P, Gabriel S. Rate, degree and predictors of recovery from disability following ischemic stroke. Neurology. 2007; 68: 1583–1587.
Hankey GJ, Spiesser J, Hakimi Z, Carita P, Gabriel S. Time frame and predictors of recovery from disability following recurrent ischemic stroke. Neurology. 2007; 68: 202–205.
Bhatt DL, Fox KA, Hacke W, Berger PB, Black HR, Boden WE, Cacoub P, Cohen EA, Creager MA, Easton JD, Flather MD, Haffner SM, Hamm CW, Hankey GJ, Johnston SC, Mak K-H, Mas J-L, Montalescot G, Pearson TA, Steg G, Steinhubl SR, Weber M, Brennan DM, Fabry-Ribaudo L, Booth J, Topol EJ; for the CHARISMA Investigators. Clopidogrel and aspirin versus aspirin alone for the prevention of atherothrombotic events. N Engl J Med. 2006; 354: 1706–1717.
de Haan J, Horn J, Limburg M, van der Meulen J, Bossuyt P. A comparison of stroke scales with measures of disability, handicap and quality of life. Stroke. 1993; 24: 1178–1181.