A perfusion territory shift attributable solely to the secondary collaterals in moyamoya patients: a potential risk factor for preoperative hemorrhagic stroke revealed by t-ASL and 3D-TOF-MRA.

OBJECTIVE:The authors conducted a study to noninvasively and nonradioactively reveal moyamoya disease (MMD) intracerebral perfusion and perfusion territory supplied by the unilateral internal carotid artery (ICA) and external carotid artery (ECA) and bilateral vertebral arteries (VAs) before surgery and to further identify risk factors for preoperative hemorrhage in adult MMD. METHODS:Forty-three consecutive adult patients with bilateral MMD underwent unenhanced T1-weighted MRI, territorial arterial spin labeling (t-ASL), and unenhanced 3D time-of-flight MRA (3D-TOF-MRA). Clinical factors, including age, sex, hypertension, diabetes mellitus, hyperlipidemia, current smoking status, and history of taking aspirin, were gathered and stratified. Univariate logistic regression analyses were used to examine the relationship between various risk factors and the occurrence of preoperative hemorrhage. Stepwise multivariate logistic regression analyses were used to determine independent risk factors of preoperative hemorrhage in MMD. RESULTS:Among the 86 MMD hemispheres, t-ASL revealed 137 perfusion territory shifts in 79 hemispheres. Five distinct categories of perfusion territory shifts were observed on t-ASL maps. The subtypes of perfusion territory shift on t-ASL maps were further subdivided into 2 different categories, group A and group B, in combination with findings on 3D-TOF-MRA. A perfusion territory shift attributable solely to the secondary collaterals was a potential independent risk factor for preoperative hemorrhage (p = 0.026; 95% CI 1.201-18.615; OR 4.729). After eliminating the influence of the secondary collaterals, the primary collaterals had no significant effect on the risk of preoperative hemorrhage (p = 0.182). CONCLUSIONS:t-ASL could reveal comprehensive MMD cerebral blood perfusion and the vivid perfusion territory shifts fed by the unilateral ICA and ECA and bilateral VAs in a noninvasive, straightforward, nonradioactive, and nonenhanced manner. 3D-TOF-MRA could subdivide t-ASL perfusion territory shifts according to their shunt arteries. A perfusion territory shift attributable to the secondary collaterals is a potential independent risk factor for preoperative hemorrhage in MMD patients. A perfusion territory shift fed by the primary collaterals may not have a strong effect on preoperative hemorrhage in MMD patients. These findings make the combined modalities of t-ASL and 3D-TOF-MRA a feasible tool for MMD disease assessment, management, and surgical strategy planning.