The circle of Willis (CW) is an important anastomotic arterial polygon at the base of the brain that connects the carotid and vertebrobasilar systems [
Various methods have been proposed for classification of the morphology of the CW due to complex anterior and posterior circulation variations [
The goal of the current study was to establish the proportion of variant anatomy of the CW and associated anomalies in patients with suspected cerebrovascular disease referred for cerebral CTA in two tertiary hospitals in Kenya.
This was a cross-sectional descriptive study carried out on patients referred for cerebral CTA at Kenyatta National Hospital and Nairobi Hospital, both located in Nairobi County, Kenya.
CT image acquisition was done by SIEMENS SOMATOM definition AS +128 and PHILLIP’S BRILLIANCE 64 SLICE at Kenyatta and Nairobi hospitals, respectively.
Nonenhanced scans were acquired first to demonstrate hemorrhage or calcification.
80 mls of LOCM (300mg/ml) via pump injector at injection rate of 5mls/second through an antecubital vein cannula (at least gauge 20) with a delay time of 7 seconds was used for acquisition of images in adult patients at Kenyatta National Hospital.
20 mls saline push before contrast administration, followed by 60 mls of LOCM at an injection rate of 4-5 mls/s via pump injector and 30 mls of saline chase with a scan delay time of 4.2 seconds, was used at Nairobi Hospital.
For young children and infants, a 22- or 24-gauge IV catheter was used and LOCM used at an injection rate of 2 mls/second.
A total of 94 patients (53 males and 41 females) referred for cerebral CTA at Kenyatta National Hospital and Nairobi Hospital from August 2017 to February 2018 were included. MIP and 3D cerebral CTA images were analyzed for anatomic variants and pathology. A cut-off value 0.8 mm vessel diameter was used. Any vessel less than 0.8 mm was considered hypoplastic/absent. Final configuration and pathology were confirmed by 2 senior radiologists. Patient’s age, gender, anatomic variants of the circle of Willis, and presence of aneurysms and AVMs were recorded in the patient data sheet.
Statistical analysis using SPSS version 20.0 was done to determine association between the variants with aneurysms and arteriovenous malformations. P value was set at 0.05.
Male patients accounted for 56.4% while 43.6% were female patients with and approx. male to female ratio of 1:1.
The age distribution of patients referred for cerebral CTA is depicted in Figure
Age distribution of patients referred for cerebral CTA.
Majority of patients were referred for cerebral CTA due to clinically suspected intracranial hemorrhage accounting for 30.85%. Other clinical indications included suspected AVMs, aneurysms, cerebral sinus thrombosis, stroke, trauma, and tumor.
A complete CW was seen in 37.2% of the patients with no statistically significant difference between males and females (p=0.9) as shown in Table
Percentage distribution of complete vs incomplete CW in males vs female patients.
GENDER | PERCENTAGE DISTRIBUTION OF CW CONFIGURATION | |
---|---|---|
COMPLETE | INCOMPLETE | |
MALE | 37.7% ( | 62.3% ( |
FEMALE | 36.6% ( | 63.4% ( |
COMBINED(MALE AND FEMALE) | 37.2% ( | 62.8% ( |
Figure
Axial cerebral CTA MIP image showing a complete CW configuration.
Type A variant accounted for 78.7% of the anterior circulation variants and was significantly higher in females than male patients (p=0.016) as demonstrated in Table
Anterior circulation variant percentage distribution.
ANTERIOR CW | FREQUENCY (IN BOTH MALE AND FEMALES) | FREQUENCY IN MALES | FREQUENCY IN FEMALES |
---|---|---|---|
TYPEA | 78.7% (n=74) | 69.8% (n=37) | 90.2% (n=37) |
TYPE B | 0% (n=0) | ||
TYPE C | 1.1% (n=1) | ||
TYPE D | 4.3% (n=4) | ||
TYPE E | 0% (n=0) | ||
TYPE F | 0% (n=0) | ||
TYPE G | 4.3% (n=4) | ||
TYPE H | 11.7% (n=11) | ||
TYPE I | 0% (n=0) | ||
TYPE J | 0% (n=0) |
Figure
Axial MIP image showing combined type A anterior circulation variant and type E posterior circulation variant.
Type E posterior circulation variant was the commonest seen accounting for 41.5% of the posterior circulation variants (Table
Proportion of posterior circulation variants.
POSTERIOR CW | FREQUENCY |
---|---|
TYPE A | 27.7% (26) |
TYPE B | 4.3% (4) |
TYPE C | 7.4% (7) |
TYPE D | 4.3% (4) |
TYPE E | 41.5% (39) |
TYPE F | 2.1% (2) |
TYPE G | 7.4% (7) |
TYPE H | 0% (0) |
TYPE I | 0% (0) |
TYPE J | 5.3% (5) |
A complete anterior CW configuration was seen in 84%. The proportion of this configuration was significantly higher in females than males (92.7% vs. 77.4%, p=0.044).
A complete posterior CW configuration was seen in 39.4% (39.6% vs. 39% in males and females, respectively, p=0.953).
There was a significantly higher proportion of adult PCA configuration in females than males (95%vs 81% p=0.044).
Fetal PCA (Figure
Axial cerebral CTA image showing bilateral fetal PCA with right P1 segment aplasia.
Aneurysms were seen in 24.5% of the patients referred for cerebral CTA (Table
Distribution of intracranial aneurysms.
Location of aneurysms | ||
---|---|---|
ACOA | 10 | 43.60% |
MCA | 5 | 21.70% |
ICA | 5 | 21.70% |
Other vessels than | 3 | 13.00% |
Figure
3D cerebral CTA showing a saccular aneurysm at the ACoA.
There was no association between CW configuration and presence of aneurysms as shown in Table
Presence and number of aneurysms in complete and incomplete CW.
Total | Circle of WILLS anatomy | P value | ||
---|---|---|---|---|
Characteristics | Incomplete | Complete | ||
| ||||
No | 71(75.5%) | 43(72.9%) | 28(80.0%) | 0.438 |
Yes | 23(24.5%) | 16(27.1%) | 7(20.0%) | |
| ||||
| ||||
1 Aneurysm | 18(78.3%) | 12(75.0%) | 6(85.7%) | 0.567 |
More than 1 Aneurysm | 5(21.7%) | 4(25.0%) | 1(14.3%) |
AVMs were seen in 8 (8.5%) of patients undergoing cerebral CTA (Figure
3D cerebral CTA image showing an AVM with feeding artery arising from the left MCA.
Fenestrations, duplications, azygous ACA, and persistent TA were not observed in this study.
Previous anatomic, CTA, and MRA studies have demonstrated significant variations in the configuration of the CW in various population groups.
The purpose of this study was to determine the proportion of variant anatomy of the circle of Willis (CW) and associated anomalies in patients with suspected cerebrovascular disease referred for cerebral CTA in 2 tertiary Kenyan hospitals. No similar studies on the CW configuration have been done in Kenya.
Variations in the CW configuration have been associated with cerebrovascular diseases [
Various methods have been proposed for classification of variant anatomy of CW. The classification of CW configuration in this study is based on the Chen et al. classification which classified anterior and posterior variants into 10 groups (A-J) [
The proportion of patients with complete CW configuration was 37.2% (n=35). This was slightly higher in males than females: 37.7% vs. 36.6%. The findings in the current study are consistent with previous studies which have demonstrated a prevalence of between 12.24%-60% [
An incomplete CW configuration was seen in 62.8% of patients, 62.3% in males versus 63.4% in females. Previous studies have demonstrated a correlation between an incomplete CW and stroke severity and prognosis [
According to the Chen et al., classification type A-F anterior variants are considered to be complete [
The commonest anterior circulation variant was type A variant accounting for 78.7% (n=74). This configuration was seen in 69.8% (n=37) and 90.2% (n=37) of males and females, respectively. This finding is similar to studies done by Naveen SR et al. and Dr. Arjun Bhaddur et al. which showed type A variant being the most common anterior circulation variant [
Type E posterior variant has been associated with increased risk of ischaemic stroke [
In our study, type E posterior circulation variant was the commonest posterior circulation variant accounting for 41.5%. This study finding is similar to findings on an MRA-TOF study by Reddy Ravikanth et al. which demonstrated a predominance of type E posterior circulation at 53% [
A cerebral CTA study done by Zhang Ning Jin et al. on Chinese population with family history of stroke showed dominance of type E posterior variant with a prevalence of 52.3% [
In the current study, fetal PCA was seen in 30.9% (n=29). Complete PCA was demonstrated in 6.4% while partial PCA was seen in 24.5%. Unilateral fetal PCA was more common than bilateral fetal PCA. 12.8% were right sided, 10.6% left sided, and 5.6% bilateral fetal PCA.
These findings are similar to a publication by Dimmick SJ Faulder et al. which show a prevalence of 15%-32% for fetal PCA with bilateral fPCA at 8%, right sided at 10%, and left sided fPCA at 10% [
In the current study, aneurysms were seen in 24.5% (n=23) of patients referred for cerebral CTA. 69.6% of the aneurysms were located in the anterior CW while 4.3% were located in the posterior CW. 78.3% of patients had one aneurysm while 21.7% had more than 1 aneurysm.
26.1% (n=6) were located in other vessels other than the CW. The prevalence of aneurysms in this study is lower than that in an earlier Kenyan study by Chepsiror et al. based on conventional angiography which showed a higher prevalence of aneurysms at 40.9% and anterior circulation aneurysms at 95.2% [
In the current study, ACoA aneurysms were the commonest at 43.6% (n=10).
21.7% were located in the MCA, 21.7% were located in the ICA, and 13% were seen in other vessels than the CW. These findings are similar to a multimodality study on size and location of ruptured intracranial aneurysms by Young Gyun et al. which showed that a majority of aneurysms were located in the ACoA [
AVMs were seen in 8.5% (n=8) in the current study.
62.5% (n=5) had incomplete CW while 37.5% (n=3) had a complete CW configuration.
These study findings are similar to an earlier Kenyan conventional angiographic study by Chepsiror et al. which showed a prevalence of 8% [
Association between duplication and fenestration of intracranial arteries with aneurysmal formation has been documented in previous studies [
In the study by Roger M et al., patients with intracranial pathology such as subarachnoid hemorrhage, stroke, tumor, and aneurysms were excluded from the study. In a previous Kenyan cadaveric study by Hassan Said et al. on 36 adult brains, fenestration and duplication of ACoA were seen in 26% and 13%, respectively [
Persistent trigeminal artery was not demonstrated in the current study. There is a general low prevalence of persistent trigeminal artery at 0.1-0.6% of all cerebral angiograms [
In our study, no case of azygous anterior cerebral artery was identified. In a previous publication on aneurysms of the azygous anterior cerebral artery by Rajesh Kumar et al., prevalence of azygous ACA was quoted at <1% with a range of 0.21%-3.73% [
In the current study, no association was found between the CW configuration and presence of aneurysms (p=0.567). These findings are similar to a retrospective cerebral CTA study by P. Brzegowy et al. which showed no association between anatomic variations of the MCA and CW with aneurysms [
There was no significant correlation between CW configuration and occurrence of AVMs (P=0.987).
The findings in this study demonstrate slight differences in the CW configuration. A higher proportion of complete anterior CW was seen in female patients. Type A anterior and type E posterior circulation variants were the commonest in both males and females. No significant association was demonstrated between CW configuration and occurrence of aneurysms/AVMs in this study. Various methods of classification of CW configuration have been proposed with no standardized method to date, thus the need for consensus building between neuroanatomists, neurologists, neurosurgeons, and neuroradiologists for ease of comparison between similar future studies on the CW.
Original data can be accessed upon request.
This manuscript was submitted to the University of Nairobi as part of fulfillment for master’s degree in diagnostic radiology.
The authors declare that there are no conflicts of interest regarding the publication of this manuscript.