Morphologic characteristics of severe basilar artery atherosclerotic stenosis on 3D high-resolution MRI

Intracranial arterial atherosclerotic stenosis (ICAS) is an important and common cause of ischemic stroke in the world [1]. Patients with intracranial arterial severe stenosis (≥ 70%) face high risk of stroke recurrence, despite best medical treatment [2]. Intracranial percutaneous transluminal angioplasty and stenting (PTAS) has been increasingly used in clinical practice for medically refractory patients [3]. Neurointerventionists generally enrolled symptomatic patients for PTAS based on stenosis severity. However, intracranial arteries have unique vessel wall characteristics (with plenty of penetrating arteries, exhibiting a thin media and adventitia, lacking external elastic lamina), which are different from extracranial arteries [4, 5]. PTAS for ICAS is associated with high rates of periprocedural complications including ischemic and hemorrhagic stroke [6]. The disappointing results of two major randomized controlled trials, the Stenting and Aggressive Medical Management for Preventing Recurrent Stroke in Intracranial Arterial Stenosis (SAMMPRIS) trial and the Vitesse Intracranial Stent Study for Ischemic Stroke Therapy (VISSIT) trial [7, 8], reminded us that lesion characteristics may be one of the important factors affecting the incidence of periprocedural complications.

Previous studies [9‐12] showed that morphological characteristics, including pre- existing remodeling and plaque location, were closely related to periprocedural complications. Coronary intervention study found that negative remodeling (NR) was associated with high incidence of in-hospital complications, including post-interventional dissection; in contrast, the rate of major adverse cardiac events in patients with positive remodeling (PR) was higher [9]. In addition, plaque around the branch vessel was associated with branch occlusion following stent placement, which may be due to “snow-plowing” effect (pushing the plaque into branch or perforating arteries in the process of angioplasty or stenting) [10]. Similar periprocedural complications had been shown in the treatment of ICAS with PTAS [7, 8, 11]. Moreover, eccentric lesions are vulnerable to rupture and may lead to distal embolization [12].

Imaging modalities such as digital subtract angiography (DSA), computer tomography angiography (CTA) and magnetic resonance angiography (MRA) mainly show the luminal stenosis and could not provide high quality images for evaluating the morphology of atherosclerotic arterial wall [13]. High-resolution magnetic resonance imaging (HRMRI) has emerged as a valuable tool for evaluating intracranial vessel wall, which can give information about remodeling patterns, plaque location and components [14, 15]. Most of previous studies were performed with 2D black blood technique, in which cross-sectional images perpendicular to the artery long axis may not be acquired for angled lesions or tortuous intracranial artery [15]. Recently, a 3D black blood technique, volumetric isotropic turbo spin echo acquisition (VISTA), is beginning to be used in clinical research [16]. In our study, we aimed to investigate artery remodeling patterns and plaque distribution in symptomatic patients with severe basilar artery (BA) stenosis on 3.0 T MRI scanner.

This cross-sectional study was approved by the hospital ethics committees. Each patient signed a written informed consent before examination. All patients fulfilled the following criteria: 1) ischemic stroke or transient ischemic attack(TIA) in the BA territory within 90 days; 2) 70–99% BA stenosis on MRA; and 3) 2 or more vascular risk factors including hypertension, diabetes mellitus, hyperlipidemia, obesity and smoking. Patients were excluded, if having the following conditions: 1) MRI contraindications, for instance, ferromagnetic implants or claustrophobia; 2) non-atherosclerotic vascular disease, such as vasculitis, dissection or moyamoya disease; 3) vertebral artery stenosis (≥ 50%); 4) evidence of cardiogenic cerebral embolism.

Patients were consecutively enrolled in this study from September 2014 to February 2018 if diagnosed as having symptomatic atherosclerotic BA stenosis in the neurology department of our hospital.

In our study, HRMRI with 3D VISTA can cover major intracranial arteries in a clinical acceptable scanning time. After MPR of 3D images, we can observe intracranial plaques in any planes with high spatial resolution and acquire cross-sectional images at the MLN, distal and proximal reference sites, which were perpendicular to the BA long axis. 3D technique is able to avoid obliqueness artifacts and more suitable for intracranial artery examination than 2D scan which may overestimate the true wall thickness and vessel area [21].

Most researches show that PR is related to unstable plaques that are vulnerable to rupture, while NR is the process of vessel shrinkage, which is relatively stable [22‐24]. During interventional therapy, the repeated crossing with a microwire or stent might be prone to plaque rupture. Coronary intervention study showed PR was associated with high incidence of major adverse cardiac events [9]. The previous study also showed that patients with negative remodeling had a high rate of post-interventional dissection [9]. In the process of PTAS, the vessel size is an important factor affecting prognosis [25]. Usually, the balloon or stent size is selected according to the diameter of proximal reference vessel during DSA. For patients with NR, neuroninterventionists should focus on the reduced vessel diameter in the selection of balloon size. Otherwise, it may lead to selecting oversized balloon or stent and increasing the risk of vessel injury. So, a small-sized balloon may be a good choice for patients with NR. The remodeling pattern is an important factor influencing perioperative complications. The preoperative HRMRI can fully assess artery remodeling and guide intracranial PTAS.

In this study, most plaques were distributed eccentrically and predominantly located at the right, left and dorsal walls, where most of perforating arteries originate [26]. Chen et al. [27] found that the most severe plaques of BA were commonly located at the left and right walls, which was similar to our results. Nevertheless, a previous study showed that the majority of middle cerebral artery plaques were located opposite to the orifice of perforating artery [28], different from our study. In the SAMMPRIS trial, patients with BA stenosis had a significantly higher rate of perforator infarct versus other intracranial arteries [29]. Another study reported that the rate of perforator occlusion after middle cerebral artery stenting was very low [30]. The difference in distribution of intracranial plaques may be the important reason of different incidence of perforator stroke in intracranial arteries. PTAS restores the vascular lumen by compressing atheromatous debris into the orifice of perforating arteries (snow-plowing effect). Another study found perforating artery occlusion was the most common cause of ischemic stroke after BA PTAS [10]. Plaques near the orifice of perforating artery increase the risk of perforator infarct [31]. So, for such patients, a small-sized balloon had been suggested to dilate BA stenosis before placing a stent [11]. Eccentric plaques are vulnerable and prone to distal embolism [12]. Vessel dissection or perforation is easy to occur with eccentric lesions during stenting [12]. HRMRI could identify the location of eccentric plaques and may distinguish the true and false lumen, which is helpful to avoid iatrogenic dissection [11]. For patients with severe basilar artery stenosis, PTAS is not a good choice if the eccentric plaques are distributed at bilateral walls. HRMRI’s findings are helpful to select the right patient for treatment, and choose the appropriate stent and balloon.

There are some limitations in our study. First, the sample size was relatively small, and there was no comparative study of pathology. Second, the patient’s clinical follow-up has been ongoing by neurologists, not yet completed. Follow-up information was not included in this cross-sectional study. Third, the results were not applicable to all patients, because our study only included patients with BA stenosis ≥70%.

After multiple planar reconstruction of 3D VISTA image, high-quality cross-sectional image is able to evaluate the remodeling pattern and plaque distribution of basilar artery atherosclerotic stenosis, which might be used to guide intracranial PTAS.