Carotid artery hemodynamics and geometry as predictors of cerebral small vessel disease: insights from 4D flow imaging.

BACKGROUND AND PURPOSE

Cerebral small vessel disease (cSVD) is a common condition with complex pathophysiology, resulting in poor clinical outcomes. This study uses four-dimensional flow (4D flow) imaging to examine how carotid artery hemodynamics and geometry impact cSVD development.

METHODS

Combining ultrasound, 4D flow, and three-dimensional time-of-flight magnetic resonance angiography (3D-TOF MRA), we analyzed and measured internal carotid artery (ICA) hemodynamics and geometric parameters. Based on cranial magnetic resonance imaging (MRI) findings, the characteristics, counts, and scores of small vessel diseases were assessed. Single-factor analyses affecting cSVD were conducted, followed by multivariable logistic regression to pinpoint independent risk factors. A P-value of < 0.05 was considered statistically significant. ROC curves were used to evaluate the model's performance and explanatory power.

RESULTS

Multivariable analysis revealed differences in white matter hyperintensities (WMH) related to maximum pressure gradient (PG), with an OR of 1.602. Maximum velocity(velocity) negatively correlated with enlarged perivascular space (PVS) count (P = 0.025). ICA-C5/ICA initial segment diameter ratio, velocity, and PG were independent risk factors for lacune (Lac) (P < 0.01). Circumferential wall shear stress (WSS) and PG were positively correlated with cerebral microbleed (CMB) count (P = 0.019 and P = 0.002), while the ICA-C5/ICA ratio was negatively correlated (P = 0.003). The MCA/ICA-C5 diameter ratio was positively associated with SVD score (P = 0.021). ROC curves demonstrated strong model performance for WMH and Lac.

CONCLUSION

Diameter variations in the ICA and MCA may indicate changes in vascular compliance and potential remodeling abnormalities. Hemodynamic anomalies in the ICA's initial segment could suggest distal vascular disease, offering new insights into cSVD pathogenesis.