circle of Willis - carotid artery diseases - stroke - angiography
círculo de Willis - doença das artérias carótidas - acidente vascular cerebral - angiografia
Extracranial atherosclerotic disease accounts for 15–20% of all ischemic strokes[1]. In patients with carotid atherosclerosis, the risk of stroke is altered by the
presence of collateral circulation and by the varying demographic properties of the
patients[2]. The incidence of stroke in patients with asymptomatic and symptomatic internal
carotid artery stenosis is 1, 2–5, 9% and 10% respectively[1].
The protective role of collateral circulation depends on many factors such as anatomic
variations, systemic arterial pressure, age and the progression rate of the occlusive
disease. The mainstay of collateral circulation is the circle of Willis and, in the
case of a large artery occlusion, it is readily available to restore the perfusion.
The anatomic properties of the collateral circulation may determine the subtypes of
stroke in patients with occlusive carotid artery disease. In a previous study, up
to 50% of anatomical variations in the cerebral collateral of patients were detected
in the circle of Willis[2].
In patients with symptomatic carotid artery stenosis, reduction of blood flow due
to a stenotic carotid artery is compensated by the increased blood flow of the collateral
system[3]. The circle of Willis is an important collateral pathway, as it can maintain blood
flow from the contralateral carotid and basilar artery to the stenotic carotid artery
region. In a study of patients with symptomatic carotid artery disease, an incomplete
circle of Willis was associated with a higher risk of both transient ischemic attack
and ischemic stroke[4]. However, the results of certain other studies indicated that the prevalence of
discontinuity in the circle of Willis in patients with symptomatic carotid disease
delivered conflicting results. Early autopsy studies showed that the prevalence of
absent or hypoplastic segments were associated with increased stroke risk when compared
with normal subjects. Conversely, other non-invasive imaging studies showed no difference
in patients experiencing transient ischemic attack or ischemic stroke due to carotid
disease[3].
The aim of the present study was to assess the anomalies in the circle of Willis with
CT angiography in patients with symptomatic and asymptomatic carotid artery disease,
to reveal its association with ischemic stroke.
METHODS
A retrospective analysis of patients who presented at our outpatient stroke clinic
with carotid artery disease was conducted between January, 2013 and June, 2015. The
study was approved by the ethics committee, and included 175 patients who had one-sided
carotid artery disease detected by multi-detector CT angiography (Philips Healthcare,
5680 DA Best, The Netherlands). Grading of the carotid artery stenosis was performed
according to the criteria of the North American Symptomatic Carotid Endarterectomy
Trial[5]. Patients excluded from the study were those with > 50% stenosis of the contralateral
carotid artery to the symptomatic carotid artery detected by CT angiography, intracranial
aneurysm, vascular malformations, and patients with dissection.
The demographic properties of the patients (age, gender), the risk factors for stroke,
the degree of carotid artery stenosis, whether the patients were symptomatic or asymptomatic,
and the findings of the circle of Willis were recorded.
The symptomatic patients were defined as those patients who presented with the symptoms
of transient ischemic attack or ischemic stroke at the vascular territory of the internal
carotid artery and monocular blindness. Asymptomatic patients were defined as those
patients who followed up at the outpatient stroke clinic without displaying the symptoms
of large artery disease.
The circle of Willis of the patients was evaluated as the anterior communicating artery
(AcomA), right and left precommunicating arteries (ACA A1 segment), posterior communicating
artery (PcomA) and precommunicating posterior cerebral artery (PCA P1 segment). The
hypoplastic arteries, or the arteries that could not be visualized with CT angiography,
were accepted as abnormal.
The presence of A1 segment of asymptomatic ACA, Acom A, A1 segment of symptomatic
ACA, P1 segment of symptomatic PCA, symptomatic PcomA was accepted as a complete circle
of Willis, and thus the absence of any of these arteries was accepted as a circle
of Willis anomaly.
CT angiography
A CT angiography examination was performed using the Philips Brilliance 64 detector
CT (Holland) device (Philips Healthcare, 5680 DA Best, The Netherlands). Venous access
was established through the antecubital vein and then an 80 mL non-ionic contrast
agent was administered at a rate of 4.5 mL/second, while axial-plane computed tomography
images of the carotid and cerebral arteries were obtained using the tracking method.
Acquired slices were transferred to the workstation (Philips IntelliSpace Portal,
Philips Healthcare) and multiplane images, maximum intensity projection and volume
rendering 3-dimensional images were developed by postprocessing the original slices
via the appropriate software (AVA). These images were reviewed with respect to the
vascular plaques and stenosis.
Statistical analysis
The study data were analyzed in SPSS 16.0 for Windows (SPSS Inc., Chicago, Illinois,
USA). Demographic and baseline characteristics were summarized as a mean±SD for continuous
variables and as a percentage of the group for categorical variables. Non-normally-distributed
data are presented as medians (inter-quartile range). The normality analysis was performed
with the Kolmogorov–Smirnov test. Comparison of the two groups was performed by the
Mann–Whitney U test for nonparametric comparison and the χ2 test for categorical variables.
All the hypotheses were constructed as two-tailed and the α critical value was accepted
as 0.05.
RESULTS
During the study period, a total of 175 patients participated in this study, of whom
121 (69%) were males and 54 (31%) were females. The mean age of the study population
was 66.7 ± 9.2 years. Of the 121 patients, 29.7% did not have an anomaly in the circle
of Willis, while 70.3% exhibited an anomaly. [Table 1] shows the demographic properties of patients with and without a circle of Willis
anomaly.
Table 1
Demographic properties.
|
Variable
|
Anomaly (+)
|
Anomaly (-)
|
p- value
|
|
HT
|
73.2%
|
57.7%
|
0.04
|
|
DM
|
39%
|
28.8%
|
0.20
|
|
CVA
|
8.9%
|
9.6%
|
0.88
|
|
CAD
|
40.7%
|
34.6%
|
0.45
|
|
PAD
|
4.9%
|
1.9%
|
0.36
|
|
HPL
|
18.7%
|
34.4%
|
0.02
|
|
Smoker
|
26%
|
28.8%
|
0.69
|
HT: hypertension; DM: Diabetes mellitus; CVA: Cerebrovascular attack; CAD: Coronary
Artery Disease; PAD: Peripheral Artery Disease; HPL: Hyperlipidemia.
Hypertension was the most common comorbid condition in groups, followed by coronary
artery disease, diabetes mellitus, smoking, hyperlipidemia, previous stroke and peripheral
artery disease. The presence of hypertension (p = 0.044) and hyperlipidemia (p = 0.02)
was statistically significant between the circle of Willis anomaly positive and negative
patients. The numbers of patients with symptomatic and asymptomatic carotid artery
disease were 83 (47.4%) and 92 (52.6%) respectively. An anomaly in the circle of the
Willis was detected in 63 (68.7%) patients with asymptomatic and in 60 (72.3%) patients
with symptomatic carotid artery disease ([Tables 2] and [3]). The overall incidence of anomaly in the circle of Willis was not statistically
significant between the patients with symptomatic and asymptomatic carotid artery
disease (p = 0.58). Moreover, there was no statistically significant difference in
patients with symptomatic and asymptomatic carotid artery disease according to the
segments of the circle of Willis (Acoma p = 0.96, IpcomA p = 0.300, IACA A1 p = 0.394,
IPCA P1 p = 0.256, KACA A1 p = 0.194, KPCA P1 p = 0.448, KPcomA p = 0.473).
Table 2
The association between symptomatic and asymptomatic carotid artery disease and a
circle of Willis anomaly.
|
Variable
|
Anomaly negative n (%)
|
Anomaly positive n (%)
|
p-value
|
|
Asymptomatic
|
29 (31.5)
|
63 (68.5)
|
0.58
|
|
Symptomatic
|
23 (27.7)
|
60 (72.3)
|
Table 3
Circle of Willis anomaly of individual arteries between asymptomatic (n = 92) and
symptomatic (n = 83) carotid artery disease.
|
Variable
|
Asymptomatic n (%)
|
Symptomatic n (%)
|
|
AcomA
|
18 (19.6%)
|
16 (19.3%)
|
|
lACA A1
|
19 (20.7%)
|
13 (15.7%)
|
|
IPCA P1
|
7 (7.6%)
|
3 (3.6%)
|
|
IPcomA
|
46 (50%)
|
48 (57.8%)
|
|
CACA A1
|
2 (2.2%)
|
5 (6%)
|
|
CPCA P1
|
6 (6.5%)
|
8 (9.6%)
|
|
CPcomA
|
57 (62%)
|
47 (56.6%)
|
AcomA: anterior communicating artery; IACA A1: ipsilateral precommunicating arteries
A1 segment; CACA A1: contralateral precommunicating arteries A1 segment; IPcomA: ipsilateral
posterior communicating artery; CPcomA: contralateral posterior communicating artery;
IPCA P1: ipsilateral precommunicating posterior cerebral artery P1 segment; CPCA P1:
contralateral posterior communicating artery P1 segment.
There was a statistically significant difference between symptomatic carotid artery
disease and the degree of stenosis (p < 0.001) ([Table 3]). The median carotid artery stenosis in symptomatic and asymptomatic carotid artery
stenosis patients was 70-90% (25th percentile 70-90% and 75th percentile preocclusive
for the symptomatic and 25th percentile 50-69% and 75th percentile 70-90% for the
asymptomatic carotid artery stenosis group). The preocclusive and occlusive carotid
artery stenosis comprised 37.4% of the symptomatic and 10.8% of the asymptomatic carotid
artery disease patients.
DISCUSSION
In this study, no difference could be detected in patients with symptomatic and asymptomatic
carotid artery disease in terms of the circle of Willis anomaly.
The clinical presentation of an occlusive carotid artery disease is highly variable;
while some patients may be diagnosed incidentally, other patients may present with
a devastating stroke. In addition, the involvement of intracranial and/or extracranial
vessels may accompany this situation. The presence and effectiveness of the collateral
circulation may determine the variability of clinical symptoms. The collateral cerebral
circulation may not only provide sufficient blood flow to the ischemic area, but also
perfuse distal to the occluded artery. In addition, the anatomical features of collateral
circulation may determine the stroke type in cases of occlusive carotid artery disease.
However, it is reported that up to 50% of significant anatomic variations among patients
were seen in the circle of Willis[2].
The contribution of the circle of Willis to cerebral hemodynamics is not only determined
by the stenosis of vertebral and carotid arteries, but also by the presence of the
vessels that constitute the circle of Willis.
Previous studies indicated that patients with asymptomatic internal carotid artery
occlusions have a better preserved hemodynamic status when compared with symptomatic
patients[6],[7]. However, from this perspective, the role of the circle of Willis is not clear.
In the study conducted by Waajer et al., instances of the circle of Willis anomaly were significantly higher in patients
with symptomatic carotid disease when compared to the control patients. In their study,
they found significantly more hypoplastic or invisible A1 segments in the symptomatic
carotid artery stenosis group, and the compromised anterior collateral pathway and
usually-accompanying posterior pathway occurred more frequently in the symptomatic
carotid artery stenosis group. In this study, controls were retrospectively collected
from patients without carotid stenosis. As we had no control group in our study, we
could not draw this conclusion[3]. In an magnetic resonance angiography study by Hartkamp et al., a significantly higher percentage of complete anterior and posterior circulation
configuration was found in the patient group than in the control subjects[8]. The prevalence of a circle of the Willis anomaly was reported in 79.7% of the symptomatic
carotid artery disease group[9]. In our study, the anomaly in the circle of the Willis was 72.3 % and 68.5% in the
symptomatic and asymptomatic carotid artery disease groups respectively. Although
we did not recruit a healthy control group for our study, the prevalence we have reported
in this study is significantly higher than the prevalence of 50% in the normal population[9],[10]. Although the number of patients with a circle of the Willis anomaly is more significant
in the symptomatic carotid artery disease group, there was no statistically significant
difference among the symptomatic and asymptomatic carotid artery disease groups.
In a prior study, it was thought that the hypoplasia of the ACA A1 segment may be
a factor that contributes to acute ischemic stroke, due to compromise of the collateral
circulation. When compared with healthy individuals, the anomaly was detected significantly
more often in acute ischemic stroke patients[11]. In a study conducted by Shaban et al., hypoplastic and absent ACA A1 were detected
in 5.9% of acute ischemic stroke patients, although hypoplastic and absent ACA A1
did not have any effect on vascular distribution, side, or the volume of the infarct[10]. In our study, an ipsilateral or contralateral presence of the ACA A1 segment was
not significantly different between the symptomatic and asymptomatic patient population.
(IACA A1 p = 0.394, KACA A1 p = 0.194).
In patients with unilateral internal carotid artery occlusion, the presence of collateral
flow via the posterior communicating artery in the circle of the Willis was associated
with a low prevalence of border zone infarcts[12]. However, collateral function of the circle of the Willis was not increased in asymptomatic
patients with internal carotid artery occlusion. In our study, absence of ipsilateral
PcomA was not statistically significant different between the symptomatic and asymptomatic
groups (p = 0.300).
The cerebral collateral circulation maintains cerebral tissue perfusion under physiologic
and pathologic conditions, such as ischemia. Previous studies revealed the role of
collateral circulation in patients with carotid artery disease: the absence of collateral
circulation in patients with symptomatic carotid artery disease has a worse prognosis.
Evaluation of the cerebral collateral circulation, together with the cerebral perfusion
and the clinical parameters of patients with occlusive carotid artery disease, may
give the full spectrum of the cerebrovascular condition. Nevertheless, the clinical
applications of information regarding collateral circulation in patients with carotid
artery disease are still not clear[2].
Our study has several limitations, such as its retrospective design, the limited number
of patients and the lack of a control group.
In conclusion, we were able to detect a difference between the symptomatic and asymptomatic
internal carotid artery occlusion patient population in terms of a circle of the Willis
anomaly. Future prospective studies may, therefore, help to better understand the
role of collateral circulation in patients with carotid artery disease.
