Hemodynamics and stroke risk in intracranial atherosclerotic disease
Leng X., Lan L., Ip HL., Abrigo J., Scalzo F., Liu H., Feng X., Chan KL., Fan FSY., Ma SH., Fang H., Xu Y., Li J., Zhang B., Xu Y., Soo YOY., Mok VCT., Yu SCH., Liebeskind DS., Wong KS., Leung TW.
Objective To investigate whether hemodynamic features of symptomatic intracranial atherosclerotic stenosis (sICAS) might correlate with the risk of stroke relapse, using a computational fluid dynamics (CFD) model. Methods In a cohort study, we recruited patients with acute ischemic stroke attributed to 50 to 99% ICAS confirmed by computed tomographic angiography (CTA). With CTA-based CFD models, translesional pressure ratio (PR = pressurepoststenotic/pressureprestenotic) and translesional wall shear stress ratio (WSSR = WSSstenotic − throat/WSSprestenotic) were obtained in each sICAS lesion. Translesional PR ≤ median was defined as low PR and WSSR ≥4th quartile as high WSSR. All patients received standard medical treatment. The primary outcome was recurrent ischemic stroke in the same territory (SIT) within 1 year. Results Overall, 245 patients (median age = 61 years, 63.7% males) were analyzed. Median translesional PR was 0.94 (interquartile range [IQR] = 0.87–0.97); median translesional WSSR was 13.3 (IQR = 7.0–26.7). SIT occurred in 20 (8.2%) patients, mostly with multiple infarcts in the border zone and/or cortical regions. In multivariate Cox regression, low PR (adjusted hazard ratio [HR] = 3.16, p = 0.026) and high WSSR (adjusted HR = 3.05, p = 0.014) were independently associated with SIT. Patients with both low PR and high WSSR had significantly higher risk of SIT than those with normal PR and WSSR (risk = 17.5% vs 3.0%, adjusted HR = 7.52, p = 0.004). Interpretation This work represents a step forward in utilizing computational flow simulation techniques in studying intracranial atherosclerotic disease. It reveals a hemodynamic pattern of sICAS that is more prone to stroke relapse, and supports hypoperfusion and artery-to-artery embolism as common mechanisms of ischemic stroke in such patients. Ann Neurol 2019;85:752–764