A Comparative Study of the Etiology of Intracranial Vertebral Artery Dissection and Carotid Artery Dissection

Abstract

Background: 

Both extrinsic (trauma or violent movement) and intrinsic (structural abnormality, atherosclerosis, or hemodynamic instability) factors may result in arterial dissection. The role of these factors in the origin or progression of dissection remains unclear. This study aimed to characterize the clinical features, radiologic features, and outcomes of intracranial vertebral artery dissection compared with carotid artery dissection and to determine the major causative factors.

Methods: 

Consecutive patients with craniocervical dissection (n=127) were retrospectively analyzed. Patients with intracranial vertebral artery dissection (n=77) and those with carotid artery dissection (n=35) were compared with respect to patient age, sex, cerebrovascular risk factors, laboratory indices, and radiologic features.

Results: 

Intracranial vertebral artery dissection was the most common craniocervical arterial dissection in our cohort (n=77, 60.6%). Body mass index in the intracranial vertebral artery dissection group was significantly greater than that in carotid artery dissection group. Clinical manifestations of intracranial vertebral artery dissection included ischemic stroke (37.7%), dizziness or vertigo (39.0%), and headache or neck pain (44.2%). Two patients had a definite history of trauma. The frontal and lateral tortuosity ratios of the vertebral basilar artery were significantly greater while the vertex angle was smaller in the intracranial vertebral artery dissection group compared with carotid artery dissection group. A positive correlation between the tortuosity ratios and subarachnoid hemorrhage and a significant inverse correlation between the tortuosity ratios and lipid parameters (high-density lipoprotein; apolipoprotein A1) were identified.

Conclusions: 

Intrinsic causes may play a more important role in the development of intracranial vertebral artery dissection than carotid artery dissection.

Arterial dissection refers to the development of a tear in the intimal layer of the arterial wall, resulting in extravasation of blood between the layers of the wall. Vertebral artery dissection (VAD) is a common cause of stroke in young patients. VAD can be either extracranial or intracranial. Extracranial dissections usually occur on the distal extracranial segment near the atlas and axis. Intracranial dissections are usually associated with a worse prognosis. The annual incidence of intracranial VAD (iVAD) is ~1.0 to 1.5 per 100,000 individuals.^1,2 iVAD may result in subarachnoid hemorrhage and ischemic stroke. Early diagnosis and treatment may prevent further damage to the vascular wall and prevent potentially fatal neurological complications.³ Carotid artery dissection (CAD) can also occur extracranially or intracranially, leading to subarachnoid hemorrhage and ischemic stroke. The annual incidence of symptomatic spontaneous internal CAD is 2.5 to 3 per 100,000, while that of CAD resulting from blunt injuries ranges from 1% to 3%.^4,5

The etiology of iVAD may include extrinsic and intrinsic causes; however, there is no clear consensus on the major cause of iVAD. Several reports have described trauma or violent movement-induced iVAD^6–11 as a common extrinsic reason for iVAD. However, some patients ignore transient minor injuries (such as quickly overrotating or overstretching the neck). Moreover, the role of this movement in the origin or progression of iVAD remains unclear. The precise etiology of CAD is also not clear. The most common extrinsic cause is trauma, such as sports injury, surgical trauma, car accident, or chiropractic neck manipulation.

A structural defect or abnormality of the vessel wall is an intrinsic cause of artery dissection and includes connective tissue disorders (such as fibromuscular dysplasia^12,13), atherosclerosis,¹⁴ and vasculitis.¹⁵ These intrinsic factors may result in arterial wall weakness. Traditional vascular risk factors for ischemic strokes, such as hypertension, smoking, and migraines, have been reported in relation to artery dissection¹⁶; however, the role of these factors in dissections is not well characterized.

To the best of our knowledge, few reports have profiled the intrinsic and extrinsic causes of iVAD and CAD. Therefore, the present study aimed to compare the etiology of iVAD and CAD. We investigated the cerebrovascular risk profiles, clinical characteristics, imaging features, and outcomes of iVAD and CAD to determine the different roles of intrinsic and extrinsic factors between dissections occurring in different sites and to explore the association between these factors and clinical outcomes.

METHODS

This case-control study included patients from 2 major medical centers (Chinese PLA General Hospital and Peking University Third Hospital, including the Departments of Neurology, Neurosurgery, Interventional Radiology, and Vascular Surgery). Consecutive inpatients with craniocervical dissection (n=127) who sought treatment at these centers between April 1, 2007, and September 30, 2014, were retrospectively analyzed. The dissections were confirmed by cerebral angiography and magnetic resonance imaging.

The clinical records of patients in each group were retrospectively analyzed. Data pertaining to the following variables were recorded: age at disease onset, symptoms at disease onset, sex, body mass index (BMI), personal history of cerebrovascular disease, family history of ischemic heart disease, other cerebrovascular risk factors (including hypertension, diabetes mellitus, smoking, and alcohol use), and laboratory indices, including routine blood and serologic parameters [eg, homocysteine, hemoglobin A1c, total cholesterol (TC), low-density lipoprotein, high-density lipoprotein (HDL), triglycerides, apolipoprotein AI (ApoAI), ApoB, and uric acid]. The atherogenic index (TC/HDL) was also calculated.

The diagnosis of artery dissection was based on the following diagnostic criteria:¹⁷ (1) compatible clinical signs and symptoms of VAD or CAD with definitive angiographic, computed tomography angiography, or magnetic resonance angiography findings of dissection in the vertebral artery or carotid artery; (2) no evidence of luminal irregularities, stenosis, or occlusion in vertebral arteries or the other cervical and intracranial arteries that would be suggestive of atherosclerosis.

Three-dimensional imaging of the vertebral basilar artery (VBA) using digital subtraction angiography (DSA) was not required for all patients; in contrast, the frontal and lateral views of the VBA are necessary in standard DSA examinations. Therefore, we adopted standard frontal and lateral imaging of the VBA in DSA to analyze the morphology of the VBA. Classically, the VBA pierces the dura mater and courses superiorly over the anterior surface of the medulla oblongata. Similar to the imaging approach in our study, the limitation of space resulted in tortuosity mainly toward the left or right direction and less toward the anterior or posterior direction.

The VBA was measured using a 2-point method (the starting point, ie, the point where the vertebral artery exits the first cervical intervertebral foramen; the terminal point, ie, the top of the basilar artery) and it was analyzed using a rectangular space coordinate system. The frontal image of the VBA in the DSA examination was subsequently viewed as its projection in the xz plane, and the lateral image was viewed as its projection in the yz plane. Thus, the VBA projections in the xz and yz planes were assessed to indirectly reflect the vascular morphology of the VBA.

The tortuosity ratio and the vertex angle to the VBA vascular morphologic description of the parameters were determined as follows:

$\begin{array}{c}\mathrm{<span\; class="ej-keyword"\; data-value="Tortuosity\; ratio">Tortuosity\; ratio</span>}=\mathrm{absolute\; length\; of\; the\; VBA}\\ \left(\mathrm{between\; the\; starting\; point\; and\; the\; terminal\; point}\right)/\mathrm{the\; straight}-\mathrm{line\; distance\; between\; two\; points}\times 100\%\end{array}$

GetData Graph Digitizer digital tools (software version 2.26, http://get data-graph-digitizer.com/) were used to obtain the point coordinates of the center line of the frontal and lateral VBAs (Fig. 1). Point coordinates were recorded for xz (frontal) and yz (lateral) and then output to a text file. The xz and yz point coordinates were entered in MATLAB software 7.6 (MathWorks, Natick, MA). The length of the VBA in the picture and the length between the starting and ending points were calculated using a definite integral method. The tortuosity ratios of the xz and yz planes were calculated according to the previously discussed formula.

FIGURE 1: Schematic illustration of the methodology for determining the tortuosity ratio of the frontal and lateral VBA. Tortuosity ratio=absolute length of the VBA (between the starting point and the terminal point)/the straight-line distance between 2 points ×100%. The starting point (A); the terminal point (B). Line OB is the sagittal line across the top point of the basilar artery (Fig. 1, line OB), and line OA is the line between the top point of the basilar artery and the point at which the vertebral artery exits the first cervical intervertebral foramen (Fig. 1, line OA). The angle between the 2 lines was named vertex angle (the basal artery deviation degree).

We subsequently drew 2 lines. The first line was the sagittal line across the top point of the basilar artery (Fig. 1, line OB), and the second line was the line between the top point of the basilar artery and the point at which the vertebral artery exited the first cervical intervertebral foramen (Fig. 1, line OA). The angle between the 2 lines was recorded and named vertex angle (the basal artery deviation degree). The basilar artery was elevated compared with the brain stem because of the vascular tortuosity of the VBA; the top point of the basilar artery was increased, whereas the angle was smaller. This angle was used to reflect the degree of deviation of the basal artery (Fig. 1).

Statistical analysis was performed using SPSS software (version 17.0). Shapiro-Wilk normality test was used to assess normality of distribution of continuous variables. Between-group differences with respect to normally distributed continuous variables were assessed using t test, while those with respect to normally distributed continuous variables were assessed using the Wilcoxon–Mann–Whitney test. Categorical variables were compared using the χ² test. Correlation analyses were performed with Pearson or Spearman rank order correlation coefficients. P-values <0.05 were considered indicative of statistical significance. Data are presented as mean ± SD, unless otherwise noted.

RESULTS

Baseline Characteristics and Potential Risk Factors

iVAD (n=77, 60.6%) was the most common dissection among the 127 radiologically confirmed craniocervical arterial dissections. The other sites of craniocervical dissections (n=50, 39.4%) included extracranial vertebral dissections (n=12, 9.4%), CADs (n=35, 27.5%), posterior cerebral artery dissections (n=2, 1.6%), and posterior inferior cerebellar artery dissection (n=1, 0.8%).

The iVAD and CAD groups were included in the analysis. The characteristics of patients in each group are shown in Table 1. The BMI, height, weight, red blood cell count, hemoglobin level, and uric acid level in the iVAD group were significantly increased compared with the CAD group.

TABLE 1 - Epidemiological, Anthropometric, and Clinical Characteristics, as well as Cerebrovascular and Biochemical Risk Factors in the iVAD and CAD Groups

Parameter	iVAD group	CAD group	P
N	77	35	—
Male, n (%)	60 (77.9)	25 (71.4)	0.456
Age, y	53.24±11.27	56.74±9.66	0.116
BMI, kg/m2	26.18±3.28	24.53±2.81	0.011
Height, cm	169.14±6.24	166.34±8.11	0.048
Weight, kg	75.12±11.77	67.97±10.03	0.002
Hypertension, n (%)	57 (74.0)	21 (60.0)	0.135
Diabetes mellitus, n (%)	14 (18.2)	7 (20.0)	0.819
Personal history of cardiovascular disease, n (%)	11 (14.3)	2 (5.7)	0.339
Family history of cerebrovascular or cardiovascular disease, n (%)	11 (14.3)	3 (8.6)	0.397
Alcohol abuse, n (%)	18 (23.4)	8 (22.9)	0.952
Current smoker, n (%)	27 (35.1)	9 (25.7)	0.326
Hemoglobin, g/L	143.66±15.92	134.35±14.01	0.006
RBC, ×1012/L	4.67±0.53	4.41±0.48	0.025
HCY, µmol/L	18.68±10.20	16.52±10.04	0.345
TC, mmol/L	4.18±0.82	3.97±0.94	0.249
TG, mmol/L	1.64±0.88	1.58±0.76	0.728
LDL, mmol/L	2.48±0.78	2.29±0.83	0.271
HDL, mmol/L	1.04±0.26	1.06±0.23	0.809
TC/HDL	4.19±1.55	3.85±0.96	0.257
ApoAI, mmol/L	1.18±0.27	1.20±0.24	0.751
ApoB, mmol/L	0.88±0.23	0.88±0.23	0.081
UC, μmol/L	344.32±95.06	296.96±73.08	0.011

ApoAI indicates apolipoprotein AI; ApoB, apolipoprotein B; BMI, body mass index; HCY, homocysteine; HDL, high-density lipoprotein; LDL, low-density lipoprotein; RBC, red blood cell count; TC, total cholesterol; TG, triglycerides; UC, uric acid.

Clinical Presentation

Clinical Characteristics of iVAD

The clinical manifestations of iVAD included ischemic stroke (29, 37.7%), headache or neck pain (34, 44.2%), dizziness or vertigo (30, 39.0%), subarachnoid hemorrhage (5, 6.5%), transient ischemic attack (7, 9.1%), and no symptoms (3, 3.9%) (Table 2). Two patients had a definite history of trauma.

TABLE 2 - Clinical Presentation at Admission in the iVAD and CAD Groups

Clinical presentation	iVAD group, n (%)	CAD group, n (%)	P
Ischemic stroke	29 (37.7)	11 (31.4)	0.523
Isolated headache/neck pain	12 (15.6)	1 (2.9)	0.060
All patients with headache/neck pain	34 (44.2)	8 (22.9)	0.031
Acute thunderclap headache	17 (22.1)	4 (11.4)	0.205
Chronic headache/neck pain	13 (16.9)	2 (5.7)	0.140
Isolated dizziness/vertigo	11 (14.3)	2 (5.7)	0.339
All patients with dizziness/vertigo	30 (39.0)	13 (37.1)	0.854
Subarachnoid hemorrhage	5 (6.5)	4 (11.4)	0.457
Transient ischemic attack	7 (9.1)	5 (14.3)	0.410
All patients with tinnitus	5 (6.5)	1 (2.9)	0.663
No symptoms	3 (3.9)	2 (5.7)	—
Total	77 (100)	50 (100)	—

iVAD and Headache/Neck Pain

The frequency of headache or neck pain in the iVAD group was significantly greater than that in the CAD group (P=0.031). The frequency of acute thunderclap headache and chronic headache/neck pain were not significantly different between the iVAD and CAD groups (P=0.205 and 0.140, respectively).

Among the patients with iVAD, 34 patients (34/77, 44.2%) had a history of headache and 11 patients (11/77, 14.3%) had an isolated headache. Headache was the initial symptom in all VBA dissection patients with headache. Seventeen patients (50.0%) presented with a thunderclap headache. The localization of pain included the temporoparietal area (6/34, 17.6%), retro-orbital area (1/34, 2.9%), occipitonuchal area (21/34, 61.8%), cervical area (1/34, 2.9%), and pancerebralgia (5/34, 14.7%). The patients complained of continuous (16/34, 47.1%) or intermittent (18/34, 52.9%) pain. Most patients experienced acute onset of pain (17/34, 50.0%), whereas 13 patients had a gradual onset of pain (13/34, 38.2%). The remaining patients had a gradual onset and acute exacerbation of pain (4/34, 11.8%). In majority of the patients, the pain intensity was severe (25/34, 73.5%).

There was no difference between the iVAD group patients with or without headache/neck pain with respect to sex distribution (P=0.405). However, patients with headache/neck pain were significantly younger than patients without headache/neck pain (49.06±9.56 vs. 56.56±11.51, P=0.003).

Radiologic Features

Isolated unilateral dissection of the vertebral artery was found in 22 of 77 patients (28.6%). Isolated dissection of the proximal basilar artery was found in 17 of 77 patients (22.1%). Approximately two thirds of the patients experienced proximal basilar artery involvement (52/77, 67.5%). Bilateral dissection of the vertebral artery was present in 23 of 77 patients (29.9%; Table 3).

TABLE 3 - Distribution and Extent of Vertebral Basilar Artery Dissection on Radiologic Imaging

Extent of artery	n (%)
Isolated bilateral vertebral artery	3 (3.9)
Isolated unilateral vertebral artery	22 (28.6)
Unilateral vertebral artery + proximal basilar artery	15 (19.5)
Bilateral vertebral artery + proximal basilar artery	20 (26.0)
Isolated proximal basilar artery	17 (22.1)
Total	77 (100)

There was no clear relationship between the deviation of the proximal basilar artery and the dominant hand (Table 4).

TABLE 4 - Deviation of the Proximal Basilar Artery From the Midline

Deviation From the Midline	Consistent with dominant hand, n (%)	Consistent with dominant vertebral artery, n (%)
No basilar artery deviation	18 (23.4)	23 (29.9)
Inconsistent with the direction of the basilar artery deviation	28 (36.4)	45 (58.4)
Consistent with the direction of the basilar artery deviation	31 (40.3)	9 (11.7)

No deviation between the basilar artery deviation and dominant hand indicates that there was no deviation of the basilar artery deviation from the midline. No deviation between the basilar artery deviation and the dominant vertebral artery indicates that there was no deviation of the basilar artery deviation from the midline or there was no obvious dominant blood supply of the vertebral artery.

Approximately one third of all iVAD patients (30/77, 39.0%) had other concomitant craniocervical artery lesions including vascular stenosis (14/77, 18.2%), aneurysm (8/77, 10.4%), aneurysm and stenosis (3/77, 3.9%), stenosis and dissection on the contralateral site (3/77, 3.9%), arteriovenous malformation (1/77, 1.3%), and carotid cavernous fistula (1/77, 1.3%).

The frontal and lateral tortuosity ratios in the iVAD group were significantly greater compared with the CAD group (frontal tortuosity ratio, 1.45±0.17 vs. 1.35±0.12, respectively, P=0.013; lateral tortuosity ratio, 1.18±0.12 vs. 1.11±0.07, respectively, P=0.001). The mean vertex angle (basal artery deviation degree) in the iVAD group was significantly smaller than that in the CAD group (31.67±7.27 degrees vs. 36.03±8.63 degrees, respectively, P=0.021).

A positive correlation between the tortuosity ratios and subarachnoid hemorrhage was identified (r=0.233, P=0.042). Furthermore, there was a significant inverse correlation between tortuosity ratios and lipid parameters (HDL, r=−0.292, P=0.013; ApoAI, r=−0.256, P=0.034; Table 5).

TABLE 5 - Correlation of Vascular Tortuosity Ratios With the Clinical Characteristics, Cerebrovascular, and Biochemical Risk Factors

	Frontal tortuosity ratio		Lateral tortuosity ratio
Parameter	Correlation coefficient	P	Correlation coefficient	P
Hypertension	0.019	0.873	−0.112	0.338
Diabetes mellitus	0.101	0.387	−0.005	0.968
Alcohol abuse	−0.058	0.620	−0.128	0.269
Current smoking	−0.169	0.146	0.018	0.876
Headache/neck pain	0.041	0.725	−0.033	0.780
Subarachnoid hemorrhage	0.233	0.042	0.137	0.239
Ischemic stroke	−0.047	0.685	0.019	0.873
Age	−0.069	0.551	0.018	0.875
Sex	−0.093	0.425	0.022	0.848
Disease course	−0.076	0.514	−0.106	0.360
BMI	−0.119	0.306	0.024	0.834
HCY	0.004	0.975	0.109	0.378
TC	−0.125	0.291	−0.099	0.406
TG	−0.025	0.836	−0.033	0.779
LDL	−0.055	0.643	−0.010	0.936
HDL	−0.292	0.013	−0.204	0.085
TC/HDL	0.158	0.188	0.113	0.348
ApoAI	−0.256	0.034	−0.186	0.126
ApoB	0.019	0.874	0.033	0.786
Lp(a)	0.031	0.806	−0.012	0.922
UC	0.077	0.515	0.058	0.627
Hemoglobin	0.071	0.550	−0.079	0.506
RBC	0.095	0.422	−0.125	0.292

ApoA1 indicates apolipoprotein AI; ApoB, apolipoprotein B; BMI, body mass index; HCY, homocysteine; HDL, high-density lipoprotein; LDL, low-density lipoprotein; Lp(a), lipoprotein (a); RBC, red blood cell count; TC, total cholesterol; TG, triglycerides; UC, uric acid.

Outcome of VBA Dissection

The bloodstream of the dominant vertebral artery was clearly affected in 65 patients (65/77, 84.4%), and these patients were offered interventional treatment. Thirteen patients (16.9%) could not be reached to obtain data for the follow-up analysis. The median follow-up duration was 33.72 months (range: 6.41 to 95.18). Most patients (49/77, 63.7%) showed complete recovery (28/77, 36.4%) or a stable state (without a recurrent cerebrovascular event; 21/77, 27.3%). Seven patients experienced recurrent cerebrovascular events (7/77, 9.1%). Six patients died of cerebrovascular disease (6/77, 7.8%). Two patients died from other reasons (2/77, 2.6%).

DISCUSSION

iVAD comprises a unique group of dissections. This uniqueness is primarily with respect to the different roles of intrinsic and extrinsic causes regarding the onset of dissections, as discussed below.

First, based on our study, iVAD is the most common type among craniocervical dissections. In our cohort, iVAD patients accounted for more than half of all the dissection patients (77/127, 60.6%). According to data from Asia regarding symptomatic VAD, the V4 segment (intracranial vertebral artery) is the most common site of dissection.^18–20 Other studies conducted in western countries have demonstrated an increased frequency of dissection in the V2 and V3 segments,^21,22 as the artery in the V2 segment is susceptible to longitudinal stretching within the transverse foramen, and the area between the vertebrae and the artery in the V3 segment is vulnerable to injury with rotation and torsion (which are mainly attributed to extrinsic causes).²² Our study is based on patients who received treatment at 2 large medical centers in the capital of China, which cater to patients from all regions of the country. Thus, these findings may reflect the special features of dissection in China. Furthermore, the differential role of intrinsic and extrinsic causes regarding the onset of dissection may explain the differences in the frequencies between the data from Asia and western countries.

Second, there were no significant differences between the 2 groups of dissection patients with respect to the main traditional risk factors (including age, hypertension, diabetes mellitus, smoking, TC, low-density lipoprotein, or the atherogenic index). The role of atherosclerosis in dissection remains controversial; however, there is histopathological evidence of the relationship between atherosclerosis and dissection. Frösen et al²³ reported that the saccular intracranial aneurysm walls bear several histopathological similarities with early atherosclerotic lesions with lipid accumulation in the wall. Moreover, nonspecific atherosclerotic lesions have been demonstrated in dissection lumen.²⁰ Many previous studies have demonstrated that vascular wall weakness is a key factor in artery dissection.²⁴ Atherosclerosis may play a partial role in the vascular wall weakness; however, this finding requires further confirmation.

To the best of our knowledge, the relationship between BMI and VAD has not been previously reported. In our cohort, patients with iVAD had higher body weight and BMI compared with patients in the CAD group. Patients with spontaneous cervical artery dissection have been reported to have increased height and lower body weight and BMI compared with age-matched and sex-matched healthy controls.²⁵ Other studies have shown that obesity is associated with an increased risk of aortic complications (dissection) in adults with Marfan syndrome²⁶ and is a risk factor for early onset of acute aortic dissection.²⁷ Potential explanations include redundancies of the cervical arteries in taller subjects or increased neck mobility in taller subjects with a longer neck.²⁵ However, the reverse result in iVAD may be explained by the increase in intrinsic causes of atherosclerosis.¹⁶

Third, headache is an important symptom of iVAD, especially in the occipitonuchal area. In our study, the frequency of headache or neck pain in iVAD group was significantly greater than that in the CAD group (P=0.031). These findings are similar to a previous report in stroke.²⁸ Headache in patients with iVAD may mimic a cluster headache or migraine.^29,30 Furthermore, the increased frequency of headache or neck pain may be related to the character of the anatomic structure and structural changes.

Fourth, the most important difference in VBA dissection is concerned with its special imaging characteristics. There is a relatively free space (eg, lack of other tissues or bones) for the basilar artery and vertebral artery before the ventral brain stem and after the dorsal of clivus. This space provides conditions for vascular tortuosity, which is an important profile manifestation of iVAD. In our study, the frontal and lateral tortuosity ratios were significantly increased in the iVAD group compared with the CAD group, and the basal artery deviation degree in the iVAD group was larger than that in the CAD group. These findings suggest that the tortuosity ratios may play a more important role in the development of iVAD as an intrinsic factor compared with CAD. In addition, approximately one third of all iVAD patients in our cohort (30/77, 39.0%) also exhibited other craniocervical artery lesions, suggesting that the factors that induce iVAD may also be risk factors for other craniocervical artery lesions. The increased vascular tortuosity subsequently results in vascular damage (vascular dissection) and possibly subarachnoid hemorrhage (a positive association between SAH and frontal tortuosity ratio was identified, r=0.233, P=0.042), indicating that increased vascular tortuosity may increase the risk of rupture of vascular dissection. In a previous study, basilar artery tortuosity showed a positive correlation with intracranial aneurysm size, suggesting that tortuosity may increase hemodynamic instability and aneurysm growth.³¹ Vertebral artery tortuosity has also been found to be an independent risk factor for cervical artery dissection, indicating that arterial tortuosity may weaken the cervical vascular structure because of its influence on hemodynamics.³² Therefore, maintaining hemodynamic stability might be beneficial for patients with high VBA tortuosity. Previous hemodynamic research has suggested that S-type bypass improves the hemodynamics in bypassed arteries (ie, reduces the low wall shear stress zone),³³ as well as accelerates atherosclerosis via a reduction in the shear stress and stimulation of an atherogenic phenotype.³⁴

An inverse correlation was observed between the tortuosity ratios and the serum levels of HDL and ApoAI. As athero-protective factors, lower HDL and ApoAI levels influence cholesterol efflux capacity and have a strong inverse association with carotid intima-media thickness.³⁵ The current novel findings suggest that HDL and ApoAI may play a role in the tortuosity of the VBA as chronic intrinsic metabolic factors. However, the association between tortuosity and ApoA1 and tortuosity and HDL levels warrant further investigation in a much larger series.

In the absence of protection from the skull, other extracranial vertebral arteries (V1, V2, and V3 segments) are vulnerable to direct lesions after trauma or violent movement. Despite the protection from the skull, trauma or violent movement may also result in iVAD.^5–14 In most cases, it is difficult to define the clinical relationship between symptom onset and recent mild head trauma. In our cohort, only 2 patients had a definite recent history of trauma. However, some patients may have overlooked mild trauma or transient over rotation or overstretching of the neck. It remains unknown whether dissection must necessarily be caused by trauma. Our findings indicated no clear relationship between the deviation of the proximal basilar artery and the dominant hand; more movements of the dominant hand did not influence the vascular morphology of the VBA. Therefore, the extrinsic cause comprises a minor etiological cause or a predisposing cause of VBA dissection.

There was no correlation between the disease course and the tortuosity ratios, which indicated that there was an uncertain subclinical time course of vascular damage. Vascular tortuosity may be present before dissection. Patients with partial dissection may have no clinical symptoms, and the number of iVAD is most likely underestimated. Most patients exhibit a good prognosis. However, dissection can lead to basilar artery occlusion and disastrous outcomes. Thus, clinicians should master iVAD and facilitate early diagnosis and treatment.

Some limitations of our study must be considered while interpreting the findings. This was a retrospective study with a relatively small sample size. A large, prospective study is required to confirm our findings. In addition, the causal relationship between anatomic findings and VA dissection could not be verified in the present study and needs to be further studied in basic research in the future.

CONCLUSIONS

iVAD is a unique dissection entity with distinct clinical characteristics and radiologic features. External causes seem to play less of a role in the development of iVAD and intrinsic vessel-based differences (excess tortuosity, concurrent lesions such as aneurysms and stenosis) may contribute more to the formation of iVAD.