Detection of venous angiomas on susceptibility enhanced magnetic resonance imaging in patients with seizures

ABSTRACT

Objectives:

To evaluate the use of susceptibility-enhanced sequence with conventional magnetic resonance imaging (MRI) of the brain for identification of venous angiomas in epilepsy work up.

Methods:

A record-based study was performed retrospectively in Radiology department at our Hospital in Eastern region of Saudi Arabia, from Jan. 2019-2024. Adult patients for whom MRI brains were conducted for epilepsy work-up with added SWAN (susceptibility weighted angiography) sequence were considered. Imaging proven cases of space occupying lesions in the brain, post-injury, and post-interventional cases were not taken. A venous angioma (or malformation) was documented when a tuft of veins drained to a larger vein (traversing through cortex or reaching under ependymal layer), appeared low signal structure on SWAN image, and enhanced on contrast-enhanced sequence. Consensus reporting was made by 2 experienced neuroradiologists. The usefulness of the SWAN sequence in the detection of venous malformation determined if the visualized abnormality was found to be related to a focus resulting in abnormal waves on the brain electroencephalography. This observation was compared to the accidently found such malformations that were seen in epileptic patients with normal EEGs (control group). Fisher’s Exact test was applied and a P-value of <0.05 was taken as statistically significant for an association.

Results:

Total number of patients was 114; 65 females (57%) and 49 males (43%), with an average age of 31.4 (range, 15-50). The SWAN found venous angiomas in 34 (29.8%), and 8 were responsible for abnormal electroencephalograms while neither of the 3 accidently detected venous malformations in the control with normal EEGs (p-value=0.001).

Conclusion:

An added SWAN sequence with conventional MRI brain imaging for patients with seizures can assist to visualize symptomatic venous angiomas leading to focal seizures.

Venous angioma is not an uncommon congenital venous malformation observed on brain imaging and autopsies. They may be detected incidentally and can be asymptomatic. However, venous anomalies have been reported to cause seizures.^1,2 particularly if associated with other parenchymal abnormalities or if get complicated. Seizures or epilepsy can be due to abnormal or excessive neural activity in the brain. Many causes of epilepsy are noted, from tumors, trauma, or infection to congenital abnormalities or malformations, and DVA is a cause of such lesional epilepsy.^3,4 For example, cavernous angiomas can trigger epilepsy by irritating surrounding brain tissue probably because of repeated hemorrhages, while venous angiomas when complicated (associated with thrombosis leading to adjacent infarction).^5-7

Getting a diagnosis is not always easy because there is no single test available that can diagnose epilepsy. Once an epileptic seizure is seen or documented, clinical assessment, laboratory, neurophysiology (EEG), and neuroimaging (CT, MRI) are required to investigate the cause of such a problem. The MRI is the recommended imaging test of choice and is preferred because it provides exquisite details of the brain parenchyma and is sensitive in identifying even small abnormalities or masses. Susceptibility-weighted sequence accentuates the paramagnetic properties of haem products.^1,3 Paramagnetic deoxyhemoglobin in a vein acts as a contrast entity by which venous malformation can be visualized without need or administration of contrast agent.^3,4

Therefore, we aim to highlight significance of adding a SWAN (susceptibility-weighted angiogram) sequence in brain MRI for the detection of such lesions contributing to seizure activity in epileptic patients.

Methods

This record-based cross-sectional study was conducted in our hospital from January 2019 to January 2024. All patients (N=114) for whom MRI brain imaging was carried out for seizure work up with added susceptibility sequences were taken. Already-known cerebral tumors, patients with traumatic brain injuries, and post-operative cases were not considered. The research got ethical approval from the radiology research committee and in conducted in accordance with Helsinki Declaration. Medical records of the patients were acquired through the Hospital Information System for clinical findings, while MR imaging features documented through Radiology Information System/Picture Archiving and Communication System. All patient related clinical and imaging data were kept confidential.

Brain MR studies were carried out on a 1.5 Tesla GE scanner. Brain MRI involved conventional T1 and T2 weighted sequences, FLAIR (Fluid Attenuation and Inversion Recovery), DWI (Diffusion Weighted Imaging)/ADC (Apparent Diffusion Coefficient) and added SWAN sequence [TE (time to echo), 50 ms; TR (time to repeat), minimum; flip angle, 15 degrees; matrix, 320 x 192; FOV (field of view), 24 cm; section thickness, 2 mm; bandwidth, 41.67 kHz, flow compensated; number of slices, 40; acquisition time, 4 minutes]. Contrast-enhanced T1W images were also acquired if needed.

The presence of a blooming artifact (i.e., susceptivity artifact caused by the presence of a paramagnetic substance, seen as an area focal hypointensity/black area), when seen as a linear or curvilinear structure, was taken as a vascular structure containing blood within. Either contrast-enhanced (axial, sagittal, and coronal) T1W imaging or cerebral angiographies were performed where those structures were seen opacified with contrast. Two experienced neuroradiologists interpreted the brain MRIs, blinded to clinical information and final diagnoses, and consensus reporting was made. Any vascular malformation suspected on the SWAN sequence was confirmed in subsequent contrast-enhanced studies or cerebral angiography. A venous malformation visualized on SWAN if related to an abnormal focus on electroencephalography was documented. High-frequency discharges and focal sharp waves acquired on EEGs were considered epileptiform foci. The observations were compared to patients with accidently visualized venous angiomas on MRIs with normal EEGs (control group). In cases of detection of venous angioma in normal or asymptomatic consecutive patients (for whom MRI brain imaging was requested for non-specific causes other than seizures or epilepsy-like headaches) during the same study period by the SWAN imaging, EEGs were acquired to document any abnormal findings. Fisher’s exact test was used to find out any association.

Results

From a study population comprising 114 cases, females were 65 (57%), and 49 males (43%), with an average age of 31.4 (range, 15-50) (Table 1). The SWAN found venous angiomas (Figure 1) in 34 (29.8%), and 8 were responsible for abnormal electroencephalograms while neither of the 3 accidently detected venous malformations in the control with normal EEGs (p-value=0.001) (Table 2).

Table 1.

- Demographic, imaging, and electrophysiologic characteristics of study population.

Characteristics	n (%)
Gender
Male	49 (43.0)
Female	65 (57.0)
SWAN detected anomaly
Not present	80 (70.2)
Present	34 (29.8)
Epileptiform Focus
Corresponded	8 (7.0)
Found other abnormality	3 (2.6)
Not correspond	3 (2.6)
Not found	100 (87.7)
EEG
Normal	49 (43.0)
Abnormal	65 (57.0)

EEG - electroencephalogram, SWAN - Susceptibility-weighted angiogram

Figure 1.

- Selected MR brain images of a patient showing A) axial SWAN image and B) contrast-enhanced axial T1W image. The vascular malformation (in this case a venous angioma) is seen as a linear structure reaching the subcortical parietal lobe.

Table 2.

- SWAN-detected and electrophysiologic findings in the study population.

Characteristics	EEG			P–value
	Normal	Abnormal	Total
SWAN detected anomaly
Not present	46 (93.9)	34 (52.3)	80	0.0000008 (Using Fisher’s Exact Test)
Present	3 (6.1)	31 (47.7)	34
Epileptiform focus
Corresponded	0	8 (12.3)	8	0.001 (Using Fisher’s Exact Test)
Found other abnormality	0	3 (4.6)	3
Did not correspond	3 (6.1)	0	3
Not found	46 (93.9)	54 (83.1)	100
Total	49	65	114

EEG - electroencephalogram,

Most of these venous anomalies (6/8) had adjacent subtle white matter high signal signifying a possible sequel of complication within the venous angioma (infarct-related to like thrombosis). Five of these with abnormal EEGs were seen in the right parietal, 2 at right frontal, and 1 at left frontal region. (Figure 1)

Discussion

Susceptibility-weighted angiogram (SWAN), also called susceptibility-weighted imaging (SWI), is a venographic imaging (T2* weighted angiography) that is very sensitive to slow flow venous blood vessel, hemorrhage/blood products or microbleed, and iron/met-hemoglobin.^8-10 It is a gradient-echo sequence that utilizes tissue susceptibility differences to produce a specific contrast.⁸ Short TE provides a time-of-flight effect allowing high-resolution visualization of cerebral vessels.^8,9

Congenital venous angiomas are usually incidentally detected on imaging.¹⁰ However, they can lead to lesional hemorrhage/bleeding (in 1-5%), and can be associated with ischemic stroke and seizures. A venous angioma is an intertwined vascular formation that undergoes enlargement, inciting inflammatory process in the affected brain tissue.¹¹ MRI shows a tangle of small vessels (may look like the spokes of a wheel) with a prominent cortical draining vein (giving a caput medusae sign).⁸ Although catheter angiography remains the gold standard in evaluation of vascular malformations, however, SWI/SWAN imaging offers improved sensitivity in detecting low-flow vascular malformations that were invisible on routine gradient-recalled echo (GRE) sequences. Also, SWI/SWAN imaging is found to be helpful in separating nidus from hemorrhage and calcification.¹² Developmental venous anomalies when isolated require no treatment, and complication rate is extremely low (about 0.15% in a year) mostly due to sudden thrombosis or occlusion of collecting vein, resulting in venous infarction and bleeding.⁶ If part of a mixed vascular malformation (i.e., associated with a cavernoma), then treatment is predicated on the other component.

Currently, susceptibility-sensitive imaging is in use for traumatic injury, stroke, vascular abnormalities, hemorrhagic cerebral lesions, demyelinating diseases, and congenital conditions.^13-20 In demyelinating diseases like MS, white matter lesions are seen to be located close to small veins (peri-venular/central vein sign).¹⁴ Also, in cases of polymicrogyria (a developmental cortical abnormality), abnormalities of cortical veins have been demonstrated by this imaging.¹⁹ We strongly feel that SWI/SWAN imaging needs to be incorporated as a regular sequence in conventional brain imaging to identify small vascular malformations that can sometimes be a cause of focal seizures particularly when an epileptic focus is suspected on electroencephalogram (EEG). The MR imaging detected and symptomatic DVA can be treated surgically.²¹ However, our study was focused on the identification of DVA being a cause of seizures that might not be associated with adjacent parenchymal abnormality or surgically treated. Kim Br and colleagues²² focused on the presence of venous anomaly (VA) in children, the distribution of VA, and forms of epilepsy.Although they included gradient imaging in their MRI studies, however, did not specifically indicate the role of this sequence in isolation for the identification of such lesions in screening or first MRI.

Limitations to our study include a retrospective small sample size and a single-center study. Not every patient who was found to have venous angioma in the brain underwent routine angiography possibly because of the smaller size and relatively less severity of symptoms. Larger prospective studies are needed to incorporate subgroups of epilepsy in both pediatric and adult age groups presenting with seizures while evaluating SWI/SWAN for identifying vascular malformations and simultaneously correlating these findings with the clinical and EEG patterns.

Conclusion

Susceptibility sequence when added to conventional MRI of brain for seizure workup can assist to identify venous angioma causing abnormal encephalographic abnormality.