A Markov model to compare the long-term effect of aspirin, clopidogrel and clopidogrel plus aspirin on prevention of recurrent ischemic stroke due to intracranial artery stenosis
Correspondence Address: Source of Support: National Science and Technology Major Project of Key Drug Innovation and Development (2011ZX09307-303-03), Conflict of Interest: None DOI: 10.4103/0028-3886.128290
Source of Support: National Science and Technology Major Project of Key Drug Innovation and Development (2011ZX09307-303-03), Conflict of Interest: None
Background: Given the importance of intracranial stenosis as a cause of recurrent ischemic stroke and the lack of evidence supporting a clear choice for prevention of recurrent ischemic events, a computer simulation model for prognostic prediction could be used to improve decision making. Aims: The aim of the following study is to compare the long-term effect of aspirin, clopidogrel and clopidogrel plus aspirin for prevention of recurrent stroke due to atherosclerotic intracranial artery stenosis. Setting and Design: The cohort consisted of 206 patients from 2006 to 2011. Materials and Methods: A two-state Markov model was used to predict the prognosis of patients with stroke or transient ischemic attack (TIA) caused by angiographically verified 50-99% stenosis of a major intracranial artery to receive aspirin, clopidogrel, or dual therapy. Statistical Analysis: Two tests were used: Pearson Chi-square test or Fisher's exact test (for percentages) and Kruskal Wallis test (for rank order data). Results: In the 10-year Markov cohort analysis, 36.24% of patients who were treated with clopidogrel plus aspirin developed to recurrent stroke while the probability for patients in the aspirin group and clopidogrel group was 42.60% and 48.39% respectively. Patients with clopidogrel plus aspirin had the highest quality-adjusted life years, followed by aspirin and clopidogrel. Conclusion: To prevent recurrent stroke in patients with intracranial artery stenosis, especially in those patients with a history of TIA or coronary artery disease, medical therapy with clopidogrel plus aspirin should be considered in preference to aspirin alone.
Keywords: Aspirin, clopidogrel, Markov model, recurrent ischemic stroke
Atherosclerotic stenosis of a major intracranial artery is an important cause of stroke, especially in Asians, Africans and Hispanics. ,,,, The risk of recurrent stroke in patients with intracranial atherosclerotic stenosis may be as high as 15% per year. , Given the importance of intracranial stenosis as a cause of recurrent ischemic stroke and the lack of evidence supporting a clear choice for prevention of recurrent ischemic events,  a computer simulation model for prognostic prediction could be used to improve decision making. In this study, we conducted a Markov cohort analysis to predict the prognosis in patients with atherosclerotic intracranial artery stenosis after treated with clopidogrel plus aspirin, aspirin or clopidogrel alone for prevention of recurrent stroke.
The study protocol was approved by the institutional review board. Data collection and analysis were supervised by the operation committee. It was retrospective cohort analysis. The inclusion criteria included initial non-disabling ischemic stroke confirmed by magnetic resonance imaging (MRI) and transient ischemic attack (TIA) within 48 h of onset, which was attributable to angiographically verified 50-99% stenosis of a major intracranial artery (carotid, middle cerebral, vertebral, or basilar) confirmed by digital subtraction angiography, a modified Ranking scale of 3 or less (indicating a non-disabling stroke) and an age of at least 40 years. Exclusion criteria included a contraindication to aspirin or clopidogrel therapy, any antiplatelet agents taken before initial stroke, any stents placed before initial stroke, non-atherosclerotic stenosis of an intracranial artery, a cardiac source of embolism and a co-existing condition that limited survival to <5 years. A total of 271 patients were collected according to inclusion and exclusion criteria between May 2006 and April 2011.
Among these patients, treatment assignments were stratified. Patients in Group A, Group C and Group D were given aspirin, clopidogrel and clopidogrel plus aspirin respectively. The initially prescribed dose of enteric-coated aspirin was 100 mg daily and that of clopidogrel was 75 mg daily. Atorvastatin calcium was prescribed basically with a dose of 20 mg daily in each patient. Medical therapy was initiated within 48 h of onset. All enrolled patients were assessed by the investigators.
Patients were followed monthly at outpatient clinic to determine whether any events had occurred. If a stroke was suspected, patients underwent brain MRI. The components of the primary end points (ischemic stroke, brain hemorrhage, or death from vascular causes other than stroke) were adjudicated. Ischemic stroke was defined as a new focal neurologic deficit of sudden onset that lasted at least 24 h and that was not associated with a hemorrhage on brain computed tomography (CT) scanning. Intracerebral hemorrhage was defined as evidence of parenchymal blood on CT scanning. Death from vascular causes other than stroke was defined as sudden death or death within 30 days after a myocardial infarction. All patients were to be followed until any single component of the primary end points or death occurred, or a common termination date was reached. During the follow-up period, however, 55 patients were lost or dropped out of the study. A total of remaining 206 patients were finally analyzed on a statistical basis. The follow-up period of those 206 patients ranged from 7 to 59 months with a mean duration of 25 months.
The Markov model
The Markov model is a multistate transition model that allows patients in a hypothetical cohort to make transitions among various health states, at different rates, over extended periods.  In this study, we constructed a Markov model to predict the prognosis of patients with stroke or TIA caused by angiographically verified 50-99% stenosis of a major intracranial artery to receive aspirin, clopidogrel, or clopidogrel plus aspirin. After diagnosis and initial treatment, the Markov model was used to trace long-term effects on prevention of recurrent ischemic stroke. We modeled two Markov states, which were "recovery" and "recurrent stroke". The health status of patients after diagnosis and initial treatments was represented as "recovery", which was the original state of the Markov chain. All patients with each treatment strategy began the process in the health state of recovery.
The duration of each cycle was one year and the model terminated after 10 cycles. During each cycle of the process, the hypothetical patients were at risk of recurrent stroke. The transition probabilities of the Markov models were derived from our clinical data by using the exponential transformation, P[t]=1-e -rt where r denotes the annual rate of recurrent stroke, t is the cycle length.  Meanwhile, quality-adjusted life years (QALYs) were used to measure the outcome of each treatment strategy. We assigned a value of 1 to the health state of recovery and a value of 0 to the state of recurrent stroke.
A sensitivity analysis was done by varying the transition probabilities for a plausible range (varied by ±25%) to investigate the influences of variations in the transition probabilities on results. In addition, simulations of the model for patients who were >60-year-old and patients with history of TIA or coronary artery diseases (CAD) were run separately to examine the influence on preferred strategy.
Baseline features of the three groups (Group A: Aspirin; Group C: Clopidogrel; Group D: Clopidogrel plus aspirin) were compared with the use of a one-way analysis of variance (for means), Pearson Chi-square test or Fisher's exact test (for percentages) and Kruskal Wallis test (for rank order data). All reported P values are two-sided, with Bonferroni's method to adjust the minimum achieved P value to account for the multiple tests; P < 0.05 was considered statistically significant. Statistical analysis was performed using the SPSS software system (SPSS for Windows, Version 17.0, SPSS Inc., Chicago, IL).
A total of remained 206 patients had been finally analyzed. Most of the baseline characteristics had no significant difference among the three treatment groups except for patients with history of CAD and history of post-extracranial internal carotid artery (ICA) or vertebral artery (VA) stent [Table 1]. Multivariant analysis showed that Group D (61.91%) had significantly more patients with history of CAD than Group A (3.70%) and Group C (20.97%) (P < 0.05); Group C had significantly more patients with history of CAD than Group A (P < 0.05). Group D (23.81%) also had significantly more patients with history of post-extracranial ICA or VA stent than Group A (2.47%) (P < 0.05); there is no significant difference between Group A (2.47%) and Group C (9.68%), Group C (9.68%) and Group D (23.81%) (P > 0.05).
In the 10-year Markov cohort analysis, the possibility of recurrent stroke was 36.24% for patients in clopidogrel plus aspirin group, compared with 42.60% in the aspirin group and 48.39% in the clopidogrel group. Patients with clopidogrel plus aspirin gain the highest QALYs (8.24 years), followed by aspirin (7.89 years) and clopidogrel (7.56 years).
[Table 2] lists the results of sensitivity analysis. On average, when the baseline value of transition probability increased by 25%, the QALYs decreased by 0.38, 0.49 and 0.46 for patients in Group D, Group C and Group A respectively. Alternatively, the decreased transition probability resulted in an increased QALYs of 0.40, 0.53 and 0.45 for the above three groups. When we took a look at patients who were >60-year-old, the QALYs fell slightly to 8.20, 7.07 and 7.40 for Group D, Group C and Group A respectively. In addition, simulation of the model for patients with history of TIA or CAD widened the differences among the three groups. The results showed that the additional QALYs gained with clopidogrel plus aspirin compared to aspirin alone ranged from 3.89 to 7.35 while the difference between clopidogrel plus aspirin and clopidogrel alone ranged from 0.52 to 1.73. Likewise, the benefits gained with clopidogrel compared to aspirin ranged from 2.69 to 6.53. As a result, clopidogrel plus aspirin gained the highest QALYs when using most plausible scenarios in our analysis.
Intracranial stenosis is one of the most common causes of recurrent stroke world-wide. , Even in patients with minor stroke or TIA, intracranial stenosis has a high early risk of recurrent stroke.  Patients at greatest risk for recurrent stroke secondary to intracranial stenosis are those with both high-grade stenosis and hemodynamic compromise.  In patients with intracranial stenosis, anticoagulation with warfarin or low-molecular-weight heparin, associated with significantly higher rates of adverse events, was no better than aspirin in reducing risk of recurrent stroke  or improving disability.  Despite medical therapy with aspirin, among patients with >70% stenosis, approximately 23% had a recurrent ipsilateral ischemic stroke over the next 12 months. , Combination treatment with cilostazol and aspirin did not show any clinical benefit compared with aspirin and placebo.  In the management of atherothrombosis with clopidogrel in high-risk patients (MATCH) study,  the combination of clopidogrel and aspirin was no more effective than clopidogrel alone in preventing recurrent ischemic events in patients with stroke, but data in the acute setting suggest this combination might be more effective than aspirin alone for the prevention of early recurrent stroke and TIA. , In the clopidogrel and aspirin for reduction of emboli in symptomatic carotid stenosis trial,  combination therapy with clopidogrel and aspirin was more effective than aspirin alone in reducing asymptomatic embolization. In patients with intracranial stenosis, the presence of microembolic signals is associated with risk of early recurrent TIA, stroke  and new silent cerebral infarctions.  In the clopidogrel plus aspirin versus aspirin alone for reducing embolization in patients with acute symptomatic cerebral or carotid artery stenosis (CLAIR) study,  combination therapy with clopidogrel and aspirin is more effective than aspirin alone in reducing the proportion of patients with at least one microembolic signal in a population of patients who have cerebral or carotid artery stenosis, most of whom had stenosis of the intracranial carotid artery.
A consensus conferences dedicated to intracranial atherosclerotic disease (ICAD) were considered timely due to their importance. ,, Recently, the large clinical and observational studies ,,,,,,,,, were completed. Patients who have had a stroke or TIA attributed to stenosis (50-99%) of a major intracranial artery face a 12-14% risk of subsequent stroke during the 2-year period after the initial ischemic event, despite treatment with antithrombotic medications. The annual risk of subsequent stroke may exceed 20% in high-risk groups. The medical treatment of patients with symptomatic ICAD is directed toward: (1) Prevention of intraluminal thrombo-embolism; (2) plaque stabilization and regression; and (3) management of atherogenic risk factors. Aspirin monotherapy, the combination of aspirin and extended release dipyridamole and clopidogrel monotherapy are all acceptable options in patients with non-cardioembolic ischemic stroke and TIA. Daily treatment with statin had been recommended for patients with a recent stroke or TIA which could reduce the incidence of fatal or non-fatal stroke, whether patients met their low-density lipoprotein goal or not. There was no ischemic stroke, brain hemorrhage, or death from other vascular causes after treatment with statin. When combined with dual antiplatelet therapy, there would be more effective for the prevention of recurrent vascular events. , However, some issues regarding medical management cannot be addressed by clinical trials due to practical and logistical considerations.
The results of our Markov model have important implications for prevention of recurrent stroke in patients with intracranial artery stenosis: Dual therapy with clopidogrel plus aspirin seemed to have the best long-term effect in reducing rates of recurrent ischemic events, especially in those patients with a history of TIA or CAD. The sensitivity analysis yielded robust results, which demonstrated that clopidogrel plus aspirin was the optimal strategy on prevention of recurrent ischemic stroke.
Patients of intracranial artery stenosis with history of TIA, CAD, or extracranial artery stenosis should be treated aggressively. Medical therapy with clopidogrel plus aspirin should be considered in preference to aspirin alone especially for those patients with a history of TIA or CAD. It seems that 100 mg of aspirin and 75 mg of clopidogrel per day is reliably safe for Chinese patients in this study. Future randomized controlled trails should help to further refine clinical decision making for patients with this condition.
There is no conflict of interest with any financial organization regarding the material discussed in the manuscript.
[Table 1], [Table 2]