PDF: CICAS
(2014)Objective
Determine the prevalence and 1-year outcomes of intracranial atherosclerosis (ICAS) in Chinese stroke patients using MRI/MRA.
Study Summary
Intervention
Prospective observational study of 2864 patients with cerebral ischemia <7 days from symptom onset. All underwent MRI/MRA. Degree and location of ICAS were assessed centrally.
Study Design
Arms: Array
Outcome
• Recurrent stroke by 12 months: 3.27% (no stenosis), 3.82% (50–69%), 5.16% (70–99%), 7.27% (occlusion)
• Independent predictors of recurrence: ICAS severity, age, family history of stroke, history of ischemia or heart disease, complete circle of Willis, higher admission NIHSS
• Antiplatelet use was high (85%), but recurrence remained significant in high-risk patients
Bottom Line
ICAS was present in 46.6% of 2,864 Chinese acute cerebral ischemia patients — the most common vascular lesion. At 12 months, recurrent stroke was 3.27% (no stenosis), 3.82% (50-69%), 5.16% (70-99%), and 7.27% (occlusion). The highest-risk subgroup — occlusion + ≥3 risk factors — had 19.05% recurrence. Seven independent predictors identified: stenosis severity (HR 1.29/grade), age (HR 1.03/year), family history (HR 2.01), prior cerebral ischemia (HR 2.37), heart disease (HR 1.98), complete circle of Willis (HR 2.36), and admission NIHSS (HR 1.05/point).
Major Points
- ICAS prevalence 46.6% (1,335/2,864) in Chinese acute stroke/TIA — most common vascular etiology, far exceeding extracranial-only (4.9%).
- Recurrent stroke graded by stenosis: 3.27% (none) → 3.82% (50-69%) → 5.16% (70-99%) → 7.27% (occlusion) at 12 months. Occlusion vs none HR 2.392 (95% CI 1.615-3.543; P<0.001).
- Highest-risk subgroup: occlusion + ≥3 risk factors → 19.05% recurrence at 12 months.
- Rates substantially lower than SAMMPRIS (12.2%) and WASID (23% for �≥70%), likely due to MRA vs DSA and observational setting.
- MCA most commonly affected: occlusion 14.18%, severe stenosis 6.04%, moderate 9.39%.
- 7 independent multivariate predictors: stenosis severity (HR 1.286/grade; P=0.0008), age (HR 1.033/year, 95% CI 1.014-1.053; cut-off point 63 years; P=0.0007), family history (HR 2.008; P=0.0047), prior cerebral ischemia (HR 2.374; P=0.0015), heart disease (HR 1.981; P=0.0087), complete circle of Willis (HR 2.359; P=0.0145), NIHSS (HR 1.049/point; P=0.009).
- Complete circle of Willis paradoxically increases risk (HR 2.36) — likely reflects hemodynamic compromise requiring collateral recruitment.
- By ICAS distribution, coexisting anterior + posterior circulation involvement carried the highest recurrence risk: HR 2.883 (1.805-4.602; P<0.001); posterior alone HR 2.073 (1.239-3.467; P=0.006).
- 96.2% received antithrombotics, 75.8% statins — yet recurrence remained high in severe stenosis groups.
- Prospective, 22-center Chinese cohort with 93.9% follow-up at 12 months.
Study Design
- Study Type
- Prospective, multicenter, hospital-based observational cohort
- Randomization
- No
- Blinding
- Central MRA readers blinded to clinical data; disagreements >10% resolved by third reader. Interrater κ=0.815.
- Sample Size
- 2864
- Follow-up
- 12 months (assessments at discharge, 3, 6, 12 months)
- Centers
- 22
- Countries
- China, Hong Kong
Primary Outcome
Definition: 12-month recurrent stroke (ischemic or hemorrhagic)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| No ICAS: 50/1,529 (3.27%) | Any ICAS: 80/1,335 (5.99%). By grade: 50-69%: 3.82%; 70-99%: 5.16%; Occlusion: 7.27% | 2.392 (occlusion vs none) (1.615-3.543) | <0.001 |
Limitations & Criticisms
- MRA used instead of DSA (gold standard) — TOF-MRA prone to flow artifacts; may over/underestimate stenosis.
- Cannot fully exclude recanalized cardioembolic embolus mimicking in-situ stenosis.
- No repeated MRA during follow-up to track stenosis progression.
- BP and medication changes during follow-up not documented.
- Did not exclude intracranial vasculitis.
- Cannot determine if recurrent stroke was in territory of stenotic artery (no mandatory DWI at recurrence).
- Different setting/imaging/ethnicity vs WASID/SAMMPRIS — not directly comparable.
Citation
Stroke. 2014;45:663-669.