In this retrospective observational study, we enrolled a consecutive series of 5,393 patients into the Clinical Research Center for Stroke program; these patients were admitted to one of 9 nationwide university or community-based hospitals in the Republic of Korea between September 2008 and March 2010 because of acute ischemic stroke. The Clinical Research Center for Stroke program was designed to prospectively collect clinical data from patients who experienced a stroke and to build a web-based multicenter stroke registry. The Clinical Research Center for Stroke plans to investigate the clinical evidence in the field of stroke prevention, management, and prognosis; to provide practical information; and to eventually aid decision making based on evidence. The registry also includes information on management performance measures as part of a quality improvement program. Each participating hospital received human research approval to enroll cases without individual patient consent under a common rule as well as exemption from subsequent review by their institutional review boards.

In total, 3,451 (64.0%) of the enrolled patients who visited the hospital within 24 hours after stroke onset were included. We excluded 371 patients (10.8%) who did not undergo diffusion-weighted imaging or cerebrovascular workup by MRA to evaluate the symptomatic lesion. After further excluding 23 patients who did not undergo assessment of functional outcome at discharge, 3,057 patients were included in the study.

We extracted the following data from the registry: demographic profile, medical history, smoking, detected risk factors, baseline stroke severity assessed by the National Institutes of Health Stroke Scale (NIHSS) score, stroke subtype by Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification, thrombolysis, imaging findings, and functional outcome at discharge. A trained physician performed routine assessment of functional outcome with the modified Rankin Disability Scale (mRS) at discharge or transfer to a rehabilitation unit. Scores were categorized as good (mRS, 0-2) or poor (mRS, 3-6) outcome.

All images were obtained using whole body magnetic resonance imaging. Two-dimensional time-of-flight MRA was used for angiography. The imaging study was conducted on the day of admission, and the procedure was performed using the stroke protocol of each participating hospital.

Steno-occlusion of a cerebral artery was identified by the on-site neuroradiologist and reviewed by a neurologist. The stenosis was defined as a reduction in diameter or transient signal void with distal flow and categorized as either a mild (<50%) or a moderate to severe (≥50%) stenosis with some degree of modification.20 The occlusion denoted the loss of signal without distal flow. The investigated vessels were divided into 7 groups according to the location of the arterial lesion: extracranial internal carotid artery (EICA), intracranial internal carotid artery, middle cerebral artery (MCA), anterior cerebral artery, posterior cerebral artery, basilar artery (BA), and vertebral artery (VA). Ischemic lesions on diffusion-weighted magnetic resonance imaging with compatible clinical findings were classified according to the distribution of arterial territory. SYSO was diagnosed when stenosis or occlusion of the cerebral artery was observed with a relevant ischemic lesion. The number of SYSOs was categorized as the involvement of a single vessel or multiple vessels (≥2). Involvement of multiple vessels was categorized as ipsilateral multiple anterior circulation, multiple posterior circulation, bilateral anterior circulation, and both anterior and posterior circulation. For involvement of a single vessel, the severity of SYSO was categorized as mild, moderate to severe, and occlusion. Location was investigated by 2 methods: involvement of the anterior versus posterior circulation and individual arterial location.

For univariate statistical analysis, χ² test and Student t test were used to study demographic characteristics and clinical variables and to study the relationship of variables with functional outcome. We chose the variables that were associated with the presence of SYSO (P<0.2) for covariates of the multivariate model and considered the clinical plausibility.

The χ² test and Student t test were used to analyze the relationship between functional outcome and SYSO characteristics, including existence, number, severity, and location. Multivariate logistic regression models were developed to measure the adjusted odds ratios (ORs) of SYSO and the characteristics of poor functional outcome. For comparison of effects among each group on the basis of the number and location of SYSOs, we applied a contrast method using a combination of estimated coefficients from the logistic model. For evaluation of the severity of SYSO, a likelihood ratio test was applied to compare the dose-response trends with the outcome. Statistical significance was defined as P<0.05. All statistical analyses were performed using SPSS version 17.0 software (Chicago, IL, USA).

The mean age of the 3,057 study subjects was 67.3±12.5 years (male, 58.9%), and their median baseline NIHSS score was 4 (interquartile range, 2-10). In total, 15.1% of these subjects received thrombolytic therapy. At discharge, 1,401 subjects (45.8%) were assessed as having a poor functional outcome. Poor outcome was associated with age, sex, diabetes, high risk of a cardioembolic source, smoking, history of stroke, baseline NIHSS score, and thrombolysis (P<0.2) (Supplementary Table 1).

A total of 1,929 subjects (63.1%) had SYSO with more than mild stenosis. The detection rate of SYSO from stroke onset to hospital arrival within 3 hours, 6 hours, 12 hours, and 24 hours was 66%, 65.1%, 60.2%, and 58.0%, respectively.

A comparison of the baseline characteristics of the study subjects based on the presence of SYSO is shown in Table 1. Patients with SYSO were older and had significantly higher proportions of diabetes, high risk of a cardioembolic source, history of stroke, and more severe baseline NIHSS score. In total, 50.8% of strokes were attributable to large artery disease and 25.6% were attributable to cardioembolic stroke.

In patients with a single SYSO (n=1,339), MCA was the most affected vessel, followed by EICA, VA, and BA (Figure 1). Of the single SYSOs, 70.6% were located in anterior vessels. The characteristics of old age, female sex, and high risk of a cardioembolic source were significantly greater in those with an anterior vessel SYSO compared with a posterior vessel SYSO.

In terms of severity, 25.5% of patients with a single SYSO had mild stenosis, 36.1% had moderate to severe stenosis, and 38.4% had occlusion. As shown in Figure 1, symptomatic arterial occlusion was frequently observed in the MCA and EICA and moderate to severe stenosis was most frequently observed in the BA.

A total of 590 subjects (30.6%) with SYSO had involvement of multiple vessels. Compared with patients with a single SYSO, involvement of multiple vessels was related to old age and a more severe baseline NIHSS score but not with high risk of a cardioembolic source. Among the patients with multiple SYSOs, involvement of both the anterior and posterior circulation (44.4%) was the most commonly observed pattern, followed by involvement of the ipsilateral multiple anterior circulation (34.6%), multiple posterior circulation (13.7%), and bilateral anterior circulation (7.3%).

In total, 55.3% of patients with SYSO had a poor functional outcome, which was almost twice as high as that of patients without SYSO (Table 2). The multiple logistic model confirmed that the presence of SYSO was independently associated with a poor outcome at discharge (adjusted OR, 1.77; 95% confidence interval [CI], 1.46-2.15) (Table 2 and Supplementary Table 2).

When the involvement of single and multiple vessels was analyzed separately in the multiple logistic model, both single SYSO (adjusted OR, 1.49; 95% CI, 1.22-1.83) and multiple SYSOs (adjusted OR, 2.71; 95% CI, 2.09-3.53) increased the risk of a poor outcome (Table 2). Additional analysis of the involvement of single and multiple vessels showed that the impact on functional outcome differed depending on the number of SYSOs (P for contrast <0.001).

Analysis of the relationship between outcome and anterior versus posterior location of SYSO revealed that a poor outcome was more prevalent when SYSO was in the anterior circulation than in the posterior circulation (55.2% vs. 37.1%). For individual arterial locations, the EICA, intracranial internal carotid artery, MCA, and BA showed a statistically significant crude OR for functional outcome (Table 2). However, the effect of the location on outcome disappeared in the multivariable logistic model (P for contrast=0.52), and an additional regression model introducing the individual arterial location showed similar results (P for contrast=0.21).

Analysis of the relationship between the severity of a single SYSO and outcome showed that a moderate to severe degree of stenosis (OR, 1.41; 95% CI, 1.07-1.85) and occlusion (OR, 2.00; 95% CI, 1.51-2.65) of the cerebral vasculature significantly increased the risk of poor outcome compared with those without SYSO in the multivariable model. In addition, there was a significant dose-response relationship (P for trend <0.001). There was no additional effect of severity and location of a single SYSO (P for interaction effect=0.45).