Prevalence and Outcomes of Symptomatic Intracranial Large Artery Stenoses and Occlusions in China
The Chinese Intracranial Atherosclerosis (CICAS) Study
Background and Purpose—We aimed to establish the prevalence, characteristics, and outcomes of intracranial atherosclerosis (ICAS) in China by a large, prospective, multicenter study.
Methods—We evaluated 2864 consecutive patients who experienced an acute cerebral ischemia <7 days after symptom onset in 22 Chinese hospitals. All patients underwent magnetic resonance angiography, with measurement of diameter of the main intracranial arteries. ICAS was defined as ≥50% diameter reduction on magnetic resonance angiography.
Results—The prevalence of ICAS was 46.6% (1335 patients, including 261 patients with coexisting extracranial carotid stenosis). Patients with ICAS had more severe stroke at admission and stayed longer in hospitals compared with those without intracranial stenosis (median National Institutes of Health Stroke Scale score, 3 versus 5; median length of stay, 14 versus 16 days; both P<0.0001). After 12 months, recurrent stroke occurred in 3.27% of patients with no stenosis, in 3.82% for those with 50% to 69% stenosis, in 5.16% for those with 70% to 99% stenosis, and in 7.27% for those with total occlusion. Cox proportional hazards regression analyses showed that the degree of arterial stenosis, age, family history of stroke, history of cerebral ischemia or heart disease, complete circle of Willis, and National Institutes of Health Stroke Scale score at admission were independent predictors for recurrent stroke at 1 year. The highest rate of recurrence was observed in patients with occlusion with the presence of ≥3 additional risk factors.
Conclusions—ICAS is the most common vascular lesion in patients with cerebrovascular disease in China. Recurrent stroke rate in our study was lower compared with those of previous clinical trials but remains unacceptably high in a subgroup of patients with severe stenosis.
Extracranial carotid atherosclerotic stenosis is the most common vascular lesion found in stroke patients who are white. In contrast, atherosclerotic stenosis of the major intracranial arteries is found commonly among stroke patients of Asian, black, and Hispanic ancestry.1–5 In China, stroke became the number 1 cause of death in 2008, and the prevalence is increasing steadily.6 Previous single-center studies reported that the prevalence of intracranial atherosclerotic stenosis accounts for 33% to 67% of stroke or transient ischemic attack (TIA) cases in China and other countries in Asia, but there have been no large multicenter prospective studies to establish the prevalence of intracranial atherosclerosis (ICAS) among acute stroke patients in clinical practice.7–11 Moreover, the rate of recurrent stroke among patients with ICAS has not been well-established outside of the setting of a clinical trial.12 The Stenting and Aggressive Medical Management for Preventing Recurrent stroke in Intracranial Stenosis (SAMMPRIS) study recently reported a recurrent stroke risk of 12.2% at 12 months among the medical treatment group.13 Yet, the prognosis of patients with intracranial large artery occlusive disease in usual clinical practice, especially in high-prevalence countries such as China, remains unclear.
The Chinese IntraCranial AtheroSclerosis (CICAS) study group aimed to document the prevalence and risk of recurrent stroke in patients with intracranial large artery occlusive disease in China using the results of magnetic resonance angiography (MRA), which is commonly available in many stroke centers worldwide.
Study Design and Patient Eligibility
The CICAS study is a prospective, multicenter, hospital-based, cohort study. Twenty-two general hospitals covering a wide geographic area in China participated. The study had been approved by the Institutional Review Board at the participating hospitals. In addition, all participants provided written informed consent.
Data were prospectively collected from patients with cerebral ischemia, including ischemic stroke and TIA, using a standardized case report form. Inclusion criteria were onset of symptom <7 days, and age between 18 and 80 years. We excluded patients who were clinically unstable or required close monitoring or were moribund, were disabled before admission (modified Rankin Scale score >2), were physically or subjectively unable to comply with MR examination, or had severe comorbidity. We excluded patients with known source of cardioembolism, including history of atrial fibrillation, valvular heart disease or cardiac valve replacement, or had atrial fibrillation/atrial flutter or valvular heart disease diagnosed by electrocardiography/transesophageal, echocardiography/transthoracic, or echocardiography/Holter echocardiography during hospitalization. On admission, baseline data, including age, sex, medical history, and physical examination, were collected. All patients underwent specified clinical evaluation, including laboratory tests, 12-lead electrocardiography, MRI, and 3-dimensional time-of-flight MRA, for cerebral circulation. Extracranial carotid vessels were examined by duplex color Doppler ultrasound or contrast-enhanced MRA.
Risk Factors Definition
The clinical information included age, sex, hypertension (defined by a history of hypertension, or being treated with an antihypertensive agent before admission, or diagnosed at discharge), diabetes mellitus (defined by a history of diabetes mellitus, or being treated for diabetes mellitus or glycosylated hemoglobin ≥7%, or diagnosis at discharge), and hyperlipidemia (defined as low-density lipoprotein cholesterol 2.6 mmol/L at admission, or a history of hyperlipidemia or receiving lipid-lowering treatment, or diagnosis at discharge), a history of cerebral ischemia (including a history of ischemic stroke and TIA), a history of hemorrhagic stroke (including intracerebral hemorrhage and subarachnoid hemorrhage), and heart disease (defined as a history of myocardial infarction, angina pectoris, and congestive heart failure). In addition, current smoker (defined as a patient who had smoked continuously for 6 months with ≥1 cigarette per day), heavy drinker (drinking >2 units per day on average for men or >1 unit per day on average for women), and hyperhomocystinemia (serum level of total homocysteine ≥15 μmol/L) status were noted.
All patients underwent conventional MRI on a 3.0- or 1.5-T MR scanner, including a sequence of 3-dimensional time-of-flight MRA, T2/T1-weighted imaging, fluid-attenuated inversion recovery sequences, and diffusion-weighted imaging. The protocol used 3-dimensional time-of-flight MRA in an axial plane with the following parameters for 3.0-T scanner: repetition time/echo time (TR/TE), 20 to 25/3.3 to 3.9 ms; flip angle, 15° to 20°; and slice thickness, 0.65 to 1.00 mm. For 1.5-T scanner, the parameters were: TR/TE, 25 to 35/6.9 ms; flip angle, 25° to 35°; and slice thickness, 0.65 to 1.00 mm. All MRI/MRA images were stored in digital format and were read centrally by 2 readers (X.Z. and Y.O.Y.S.) who were blinded to subjects’ clinical information or outcome. Intrarater and interrater reliability were 0.793 and 0.815, respectively. Disagreements involving >10% degree of stenosis were resolved by a third reader (K.S.L.W.) who decided the final value.
The degree of intracranial stenosis on MRA was calculated using the published method from the WASID study.14 All measurements were made with Wiha DigiMax Digital Calipers 6″ (Germany) with a resolution of 0.01 to 0.03 mm for 0 to 100 mm. We assessed the following arterial segments: bilateral intracranial internal carotid artery, anterior cerebral artery A1/A2, middle cerebral artery M1/M2, posterior cerebral artery P1/P2, and basilar artery. For the internal carotid artery, an intracranial location was defined when the stenotic lesion was distal to the ophthalmic artery. According to the severity of stenosis, we classified the patients into 4 groups: <50% or no stenosis, 50% to 69% stenosis, 70% to 99% stenosis, and occlusion groups. We regarded the focal flow void found on MRA with distal filling as severe stenosis (70% to 99%).15 The absence of distal filling on MRA would be regarded as 100% occlusion. According to the distribution of stenosis, we categorized vessels with >50% stenosis or occlusion as presence of intracranial stenosis, extracranial stenosis, or tandem group (both intracranial and extracranial stenosis). The extracranial part of internal carotid artery was estimated with ultrasonography examination according to the published diagnostic criteria16 or by contrast-enhanced MRA.
White matter changes were classified as punctate and confluent lesions. Punctuate lesion was defined as subcortical or periventricular hyperintensities ≥5 mm on either T2 or fluid-attenuated inversion recovery images, and confluent lesions were defined as diffuse involvement of the deep white matter or periventricular regions on either T2 or fluid-attenuated inversion recovery images. A complete circle of Willis was considered as exact presence of each part of the circle, including anterior communicating artery, bilateral A1, terminal internal carotid artery, posterior communicating artery, and P1. Any missing segment in the circle of Willis was considered incomplete.
Follow-Up and Assessment of End Points
Trained research personnel at Beijing Tiantan Hospital and Hong Kong Prince of Wales Hospital who used standard scripts to collect study data at discharge and at 3, 6, and 12 months after stroke onset contacted patients via telephone. The enrollee was the targeted respondent for the questionnaire. When it was impossible to speak to the patient or when the information provided by the patient was deemed unreliable by the interviewer, the caregiver was contacted and interviewed. All suspected end points or safety events were sent to the central adjudicating committee for final decision. Each case fatality was either confirmed on a death certificate from the local citizen registry or from the attended hospital. In case of lack of local citizen registry information or death without hospitalization, case fatality was deemed reliable if death was reported on 2 consecutive follow-up periods from different proxies. End points and safety events included the following: stroke or TIA, death, other vascular events (including myocardial infarction, angina pectoris, hemorrhagic events, and other ischemic vascular events), and disability. Death was assessed by vascular death (including fatal stroke, other cardiovascular death, or death by any causes). Disability was measured by the modified Rankin Scale score, ranging from 0 to 5 (death was rated as 6). Stroke recurrence was defined as sudden functional deterioration in neurological status with decrease of National Institutes of Health Stroke Scale (NIHSS) score of 4 or more, or a new focal neurological deficit of vascular origin lasting >24 hours, including recurrent ischemic or hemorrhagic stroke.
All analyses were performed with SAS software version 9.1.3 (SAS Institute, Cary, NC). Continuous variables were summarized as median (interquartile range [IQR]) or mean (SD). Categorical variables such as male sex, vascular risk factors, and location of stenotic arteries were presented as percentages. Independent samples t-test or Wilcoxon test were used for comparison of continuous variables, and χ2 test or Fisher exact test was used for comparison of categorical variables. Adjusted differences of NIHSS score at admission and hospital stay for patients with and without ICAS were analyzed by general linear model. For the analyses of predictors of long-term clinical outcomes, univariate and multivariate Cox proportional hazards regression were used to estimate the impact in terms of hazard ratios of possible determinants of recurrent stroke, taking the time variable into consideration. Hazard ratio (HR), 95% confidence interval (CI), and P value were observed. All tests were 2-sided, with a significance level fixed at 5%.
From October 2007 to June 2009, 3580 patients were registered. We excluded 325 patients with cardioembolism. In addition, we excluded 391 patients because of incomplete cerebrovascular work-up, such as MRI/MRA examination. We included a total of 2864 consecutive patients as our cohort. We followed-up the cohort patients for 12 months; 176 (6.15%) patients were lost to follow-up.
The clinical features of the participants are summarized in Table 1. The mean age at admission for index stroke was 61.9±11.2 years (IQR, 19–80), and 67.9% of the patients were men. Hypertension (78.1%), hyperlipidemia (75.9%), and history of cerebral ischemia (70.5%) were the 3 most common vascular risk factors. The common treatments administered during hospital stay included antithrombotic therapy (96.2%, including antiplatelet therapy for 85.44%), statins (75.8%), and antihypertensive drugs (51.3%). There were 108 patients not using antithrombotics during hospitalization (32 for contraindication, 5 for patient refusal, 71 for undocumented reasons). Among them, 61 patients were prescribed with antithrombotics at discharge. The prevalence of intracranial large artery occlusive disease was 46.6% (1335 patients, including 261 patients with coexisting extracranial carotid occlusive disease). Patients with ICAS were older and had significantly more vascular risk factors, such as diabetes mellitus, hypertension, family history of stroke, and history of cerebral ischemia.
Patients with intracranial occlusive disease had more severe stroke at admission and stayed longer in hospitals compared with those without intracranial occlusive disease (median NIHSS score, 3 versus 5; median length of stay, 14 versus 16 days; both P<0.0001). After adjusting for sex, age, history of diabetes mellitus, history of hypertension, family history of stroke, and history of cerebral ischemia, by general linear model, patients with ICAS still had more severe stroke (NIHSS score [IQR], 5 [2–9] versus 3 [1–5]) at admission and longer hospital stay (16 [IQR, 13–21] days versus 14 [IQR, 10–18] days).
Of 2864 patients, 1074 (37.5%) patients had intracranial lesions only, 261 (9.1%) had both intracranial and extracranial lesions, and 141 (4.9%) had extracranial lesions only, whereas 1388 (48.5%) had no significant occlusive disease. Occlusive lesion was found in 109 patients (3.81%) in anterior cerebral artery, in 406 patients (14.18%) in middle cerebral artery, in 225 patients (7.86%) in posterior cerebral artery, in 66 patients (2.30%) in basilar artery, and in 148 patients (5.17%) in intracranial internal carotid artery. Severe (70%–99%) stenosis was found in 43 patients (1.50%) in anterior cerebral artery, in 173 patients (6.04%) in middle cerebral artery, in 116 patients (4.05%) in posterior cerebral artery, in 36 patients (1.26%) in basilar artery, and in 12 patients (0.42%) in intracranial internal carotid artery. Moderate (50% to 69%) stenosis was found in 67 patients (2.34%) in anterior cerebral artery, in 269 patients (9.39%) in middle cerebral artery, in 192 patients (6.70%) in posterior cerebral artery, in 79 patients (2.76%) in basilar artery, and in 15 patients (0.52%) in intracranial internal carotid artery.
Data were available for analysis for all 2864 patients at discharge from hospital, for 2776 (96.9%) patients at 3 months, for 2774 (96.9%) patients at 6 months, and for 2688 (93.9%) patients at 12 months. By 12 months, recurrent stroke occurred in 50 (3.27%) patients with no intracranial occlusive disease and in 80 (5.99%) patients with intracranial occlusive disease. Among patients with intracranial occlusive disease, recurrent stroke occurred in 13 patients (3.82%) for 50% to 69% stenosis, in 13 patients (5.16%) for 70% to 99% stenosis, and in 54 patients (7.27%) for occlusion. Among 130 recurrent strokes, 86 (66.2%) were nonfatal ischemic, 14 (10.8%) were nonfatal hemorrhagic, 17 (13.1%) were fatal ischemic, 5 (3.8%) were fatal hemorrhagic, and 8 (6.2%) were fatal but of unknown stroke type. Only 5 patients were using oral anticoagulation at discharge, and 43 patients underwent intervention such as stenting. Among them, only 1 had a recurrent stroke.
Univariate analysis showed that many factors were associated with recurrent stroke (Table 2). In multivariate analysis, the presence of ICAS (only or associated with extracranial atherosclerosis) was an independent predictor of stroke recurrence. The degree of arterial stenosis (HR, 1.286; 95% CI, 1.111–1.488; P=0.0008), age (HR, 1.033; 95% CI, 1.014–1.053; P=0.0007; cut-off point, 63 years), family history of stroke (HR, 2.008; 95% CI, 1.239–3.255; P=0.0047), history of cerebral ischemia (HR, 2.374; 95% CI, 1.391–4.053; P=0.0015) or heart disease (HR, 1.981; 95% CI, 1.189–3.302; P=0.0087), complete circle of Willis (HR, 2.359; 95% CI, 1.186–4.691; P=0.0145), and NIHSS score at admission (HR, 1.049; 95% CI, 1.012–1.087; P=0.009) were independent predictors for recurrent stroke (Table 3). The highest rate of recurrence was observed in patients with occlusion with the presence of ≥3 risk factors (Figure).
The burden of stroke in China is likely the highest in the world, and stroke is the most common cause of death, affecting 2.5 million people each year.6,17 Moreover, stroke mortality outnumbers ischemic heart disease mortality in China, but the reasons behind the observed differences remain unclear.18 The propensity for intracranial large artery disease may partly explain the higher prevalence of stroke in China. CICAS is the first large, multicenter, prospective study to document the prevalence and outcomes of intracranial large artery occlusive disease in various parts of China. Our finding that 46.6% patients with stroke had intracranial occlusive disease could be extrapolated into an annual incidence of ≈810 000 of the 2.5 million strokes in China, assuming 70% of strokes in China are ischemic. Despite its heavy burden, especially in Asia, there is no specific treatment for ICAS. This clear evidence of the huge burden of ICAS hopefully will stimulate further research into effective specific treatments for this disease.
Our study has clearly established that patients with intracranial large artery disease had more severe stroke, stayed longer in the hospital, and had higher risk of recurrent stroke. The risks of recurrent stroke were documented systematically mostly in randomized clinical trials, such as Warfarin–Aspirin Symptomatic Intracranial Disease (WASID)12 and SAMMPRIS,13 but not in usual clinical practice. Most of the studies of ICAS were small retrospective case series or were performed among select patients.19 In the WASID study, the severity of symptomatic intracranial stenosis was independently associated with a higher risk of subsequent stroke in the territory of stenosed artery. The risk of recurrent stroke in the territory of symptomatic stenotic artery was as high as 23% during the first year in patients with stenosis ≥70%.12 In the SAMMPRIS study, the risk was 12.2% in the medical arm. However, the risk of recurrent stroke among patients with 70% to 99% stenosis was only 5% in our cohort. The differences might have been caused by different settings of the studies (randomized clinical trial versus cohort study), different diagnostic methods (digital subtraction angiography [DSA] versus MRA), and different ethnicities studied. Our finding is more in line with the recurrent stroke rate found in the Trial of Cilostazol in Symptomatic Intracranial Arterial Stenosis II (TOSS II) study that reported a rate of 3.8% in 7 months.20 More importantly, we found that the degree of stenosis is not the only independent predictor for recurrent stroke. Other factors such as diastolic blood pressure, no use of antithrombotic drug, complete circle of Willis, and history of cerebral ischemia or heart disease, and family history of stroke were also independent predictors.21 High blood pressure, no use of antithrombotic drugs, and significant history of stroke are well-known risk factors for recurrent stroke.22 Complete circle of Willis on MRA might indicate the need for opening of the collateral circulation.
This study has several limitations. First, intracranial large artery disease may be diagnosed by transcranial Doppler ultrasound, MRA, CT angiography, or DSA. DSA remains the gold standard of diagnosis of luminal stenosis for ICAS. MRA is flow-sensitive, but not as accurate as DSA. However, DSA is expensive, invasive, and not readily available in most community hospitals, and thus is not well-suited for a large-scale cohort study that more closely reflects common clinical practice. However, MRA is noninvasive and more easily accessible compared with DSA.23 However, time-of-flight MRA is prone to artifacts because of flow abnormalities. Low flow velocities may easily mock stenosis, whereas high flow velocities within the stenosis may underestimate the degree of stenosis.24 Second, although we excluded patients with presumed cardioembolism, we could not accurately exclude a recannalizing embolus from an in situ stenosis. Because the study aimed to include a wide geographic representation and various levels of hospitals, and was not just limited to top-tier hospitals in China, we did not request investigations such as echocardiography, 24-hour Holter monitoring, event recorder, or others, in the participating hospitals. We cannot completely rule out the possibility of some proportion of patients with cardioembolic strokes in our study population. Thus, some of the intracranial recurrent ischemic strokes may not be attributable to the high atherosclerotic burden but rather to insufficient medical treatment (platelet inhibitor instead of oral anticoagulation). Most hospitals in China are located in the low-income regions. Because of limited resources, we did not include repeated MRAs to document persistent occlusive lesion. Third, we did not document blood pressure readings and changes of medications during the 12-month follow-up period and relied solely on the data from hospital inpatient records. Fourth, we did not specifically exclude intracranial vasculitis or primary angiitis of the central nervous system (PACNS). However, it is a rare disease worldwide25 and is diagnosed by brain biopsy, which is seldom performed in China. The possibility of PACNS is usually raised in the setting of an abnormal cerebral angiogram. Nevertheless, PACNS usually affects small cerebral arteries rather than intracranial large arteries. Finally, one-third of patients with recurrent stroke had multiple intracranial stenoses (Table 2), so it was difficult to identify which one was the symptomatic lesion. Moreover, because this multicenter study involved different levels of hospital centers, we did not have the neuroimaging data, such as diffusion-weighted imaging or CT scan, to confirm the exact location of the new infarct. It was difficult to pinpoint whether a recurrent stroke occurred in the territory of stenotic or occluded intracranial artery, which is the usual practice for stenting trials, but it is difficult to perform outside of clinical trials.
In conclusion, to our knowledge, CICAS is the first large, prospective, multicenter, cohort study to provide comprehensive data about the distribution and prognosis of ICAS in Asia. Despite the prevalent use of antithrombotic drugs, antihypertensive drugs, and statins, the risk of recurrent stroke remains high among those with severe stenosis and multiple risk factors.
Sources of Funding
This study was funded by the Ministry of Science and Technology and the Ministry of Health of the People’s Republic of China, National S&T Major Project of China (2008ZX09312-008), State Key Development Program of Basic Research of China (2009CB521905), and in part by the S.H. Ho Cardiovascular Disease and Stroke Center of the Chinese University of Hong Kong.
Guest Editor for this article was Bo Norrving, MD, PhD.
- Received September 12, 2013.
- Revision received December 19, 2013.
- Accepted December 27, 2013.
- © 2014 American Heart Association, Inc.
- Sacco RL,
- Kargman DE,
- Gu Q,
- Zamanillo MC
- Gorelick PB,
- Wong KS,
- Bae HJ,
- Pandey DK
- Liu L,
- Wang D,
- Wong KS,
- Wang Y
- Wong KS,
- Li H,
- Chan YL,
- Ahuja A,
- Lam WW,
- Wong A,
- et al
- De Silva DA,
- Woon FP,
- Lee MP,
- Chen CP,
- Chang HM,
- Wong MC
- Kasner SE,
- Chimowitz MI,
- Lynn MJ,
- Howlett-Smith H,
- Stern BJ,
- Hertzberg VS,
- et al
- Samuels OB,
- Joseph GJ,
- Lynn MJ,
- Smith HA,
- Chimowitz MI
- Nederkoorn PJ,
- van der Graaf Y,
- Eikelboom BC,
- van der Lugt A,
- Bartels LW,
- Mali WP
- Kim AS,
- Johnston SC
- Kwon SU,
- Hong KS,
- Kang DW,
- Park JM,
- Lee JH,
- Cho YJ,
- et al
- Holmstedt CA,
- Turan TN,
- Chimowitz MI