Efficacy of high-resolution cone-beam CT in the evaluation of perforators in vertebral artery dissection

Abstract

Introduction

A major complication of internal coil trapping for vertebral artery dissection (VAD) is medullary infarction associated with perforator occlusion. Currently, higher spatial resolution imaging can be obtained with high-resolution cone-beam computed tomography (VASO CT), and the efficacy of perforator visualization adjacent to VAD was examined.

Methods

Eight patients who underwent internal coil trapping or stent-supported coil embolization underwent VASO CT to evaluate perforators around VAD. Visualization of perforators was compared with conventional digital subtraction angiography (DSA) and three-dimensional rotational angiography (3D-RA). Postoperative MRI was performed in all patients to investigate ischemic complications. The relationship between the perforators and the infarction was analyzed.

Results

Perforator visualization was much clearer on VASO CT than on 2D DSA or 3D-RA. It was sharp enough to identify each perforating artery. Medullary infarctions were detected in two cases. In these two cases, each ischemic lesion corresponded to the territory of a perforator that was well visualized on VASO CT. The axial view with adjoining tissue structures on VASO CT was useful to detect the territories of perforators.

Conclusions

VASO CT is an efficient modality for the detection and identification of perforators in the vicinity of VAD. It provides accurate anatomical information about the vertebrobasilar system that is useful for the treatment of unruptured VAD.

Introduction

Patients with ruptured vertebral artery dissection (VAD) should be treated by surgical or endovascular methods.^1,2 The endovascular method is less invasive and is becoming the first-line therapy.^1–3 However, postoperative medullary infarction has been reported as a result of perforator occlusion by internal coil trapping.^1,2,4,5 Therefore, it is important to identify the adjacent perforators around the VAD during preoperative evaluation for internal coil trapping. With respect to unruptured VAD, the natural course is usually favorable, and observation with serial radiologic examinations is recommended.^6–8 If serial radiologic examinations indicate VAD enlargement or patients show neurological deterioration, surgical therapy, including endovascular methods, should be considered. However, the optimal treatment has not been established and indications for internal coil trapping of VAD depend on the size and the morphological changes of VAD, the patient’s age, and the symptoms. Perforators adjacent to the VAD are also critical for determining the indication for internal coil trapping of unruptured VAD to prevent ischemic complications.

Flat-detector C-arm cone-beam computed tomography (CB-CT) has been shown to be a valuable imaging technique, providing cross-sectional imaging with CT-like quality intraoperatively. High-resolution CB-CT (VASO CT) is one of the modalities on the Allura Clarity digital subtraction angiography (DSA) system (Philips Medical Systems, Best, The Netherlands) using a new flat-detector. Higher spatial resolution with a wide range of contrast obtained by this modality has the potential to detect perforators more clearly than conventional DSA. There has as yet been no study that has examined the difference in detecting perforators between VASO CT and conventional angiography.

Our preliminary experience using VASO CT technology for the detection and identification of adjacent perforating arteries is presented in terms of the endovascular management and postoperative magnetic resonance imaging (MRI) evaluation.

Patients and methods

A total of eight patients (four men, four women; mean age 59.4 years, range 40–74 years) who underwent endovascular internal coil trapping or stent-supported coil embolization underwent VASO CT to evaluate perforators around the VAD since 2013. VADs were located distal to the posterior inferior cerebellar artery (PICA) in five cases and proximal to the PICA in three cases (Table 1).

Table 1.

Baseline characteristics, radiological findings, and clinical outcome of patients with vertebral artery dissection who underwent coil embolization.

Case	Age (y)	Sex	Location of aneurysm	Detection of perforator adjacent to aneurysm			New ischemic lesion on MRI
				DSA	3D-RA	VASO CT
1	57	M	Distal to PICA	+	–	+	–
2	74	F	Distal to PICA	+	+	+	–
3	48	M	Proximal to PICA	+	–	+	+
4	60	M	Distal to PICA	+	+	+	+
5	67	F	Distal to PICA	+	–	+	–
6	40	F	Proximal to PICA	+	+	+	–
7	72	M	Distal to PICA	+	–	+	–
8	56	F	Proximal to PICA	–	–	–	–

DSA: digital subtraction angiography; 3D-RA: three-dimensional rotational angiography; VASO CT: high-resolution cone-beam computed tomography; PICA: posterior inferior cerebellar artery; y: years; M: male; F: female.

The DSA system was a flat-detector biplane angiography unit (Allura Clarity FD20/20; Philips Medical System). The 80-kV VASO CT evaluation was performed before and after treatment. The motorized frontal C-arm was used to acquire 30 projection images/s at 80 kV, the scanning time was 20 seconds, and the detector format used was 22 cm × 22 cm. Iodinated contrast medium (Optiray320; Mallinckrodt, Pointe-Claire, Canada) was diluted to a concentration of around 33% with heparinized saline and injected intra-arterially at a rate of 2.5–3.5 ml/s for 25 seconds, with a five-second start delay in order to obtain maximum contrast filling at the level of the target lesion. Contrast medium was injected at a rate of 6 ml/s with a total volume of 8 ml for two-dimensional (2D) DSA and at a rate of 2 ml/s for 18 seconds for three-dimensional rotational angiography (3D-RA).

All procedures and acquisitions were performed under general anesthesia with endotracheal intubation. During this acquisition, the ventilator for anesthesia was paused for about 20 seconds of apnea to avoid motion artifact under the monitoring and control of the anesthesiologist. The acquisition dataset was transferred to the workstation (Xtra Vision; Philips Medical System) for the reconstruction process.

All VASO CT images taken before and after treatment were reconstructed with a 512³ matrix focused on the target lesion of the VAD and with the “optimized for the stent” function that is used for enhancement of stent visualization. After reconstruction of the 3D volume of interest, anatomical details were evaluated by multi-planar reconstructions with volume rendering and manipulation of the parameters.

MRI was performed in all cases using a 1.5-T system (Magnetom Avanto; Siemens, Erlangen, Germany) equipped with a neurovascular surface coil. The diffusion-weighted imaging studies were obtained the day after the procedure.

Results

Perforators detected by conventional DSA were also detected by VASO CT in all cases (Table 1). However, perforator visualization was much clearer on VASO CT than on conventional DSA and 3D-RA (Figure 1). No perforators that were not detected by conventional DSA were detected by VASO CT (Table 1). 3D-RA could not detect perforators that were detected by conventional DSA in four cases (Table 1). Although the visualization of perforators was frequently affected by the metal artifact after coil embolization, it was detected with manual optimization of the parameters in VASO CT mode (Figure 1(e) and (f)).

Figure 1.

Case no. 5, left vertebral artery angiograms.

Conventional digital subtraction angiography (DSA) demonstrates vertebral artery dissection (VAD) and the anterior spinal artery (arrow) originating from the vertebral artery (a). Three-dimensional rotational angiography (3D-RA) does not show the anterior spinal artery (b). Conventional DSA demonstrates the spared anterior spinal artery (arrow) after stent-supported coil embolization of the VAD (c). High-resolution cone-beam computed tomography (VASO CT) shows the anterior spinal artery (arrow) more clearly than conventional DSA (d). Before and after stent-supported coil embolization images of VASO CT visualize both stent struts and the anterior spinal artery ((e), (f), arrow).

Medullary infarctions were observed in two cases. In case No. 3, the patient was a 48-year-old man. He presented with severe nuchal pain, and the MRI showed left VAD without subarachnoid hemorrhage (Figure 2(a)). The symptom worsened, and a subsequent MRI was performed two weeks after onset. This MRI showed significant changes in size and shape of the lesion with VAD (Figure 2(b)). This morphological change over such a short period was considered to indicate the potential risk of rupture or an ischemic event. Therefore, internal coil trapping using the endovascular approach was performed. Two dominant perforators originating from the proximal segment of the vertebral artery were observed (Figure 2(d)). VASO CT visualized the perforators more clearly, and the axial view showed that one perforator ran to the dorsolateral part of the medulla oblongata (Figure 2(e)), and the other ran to extracranial muscle (Figure 2(f)). In order to prevent recanalization of this fragile pathology, the sacrifice of these two perforators by internal coil trapping could not be avoided because of anatomical reasons (Figure 2(g)). The anterior spinal artery originated from the right vertebral artery (Figure 2(h)). The diffusion-weighted images on MRI at 48 hours after the operation showed some high-intensity spotty lesions in accordance with the perfusion area of this perforator (Figure 2(i)), but the patient had no neurological deficit.

Figure 2.

Case no. 3.

Magnetic resonance imaging (MRI) demonstrates left vertebral artery dissection (VAD) on admission (a). MRI shows the changes in size and shape of the VAD two weeks after onset (b). Conventional digital subtraction angiography (DSA) (anteroposterior (A-P) and lateral view) shows the left VAD located proximal to the posterior inferior cerebellar artery and two dominant perforators (arrows) originating from the proximal segment of the vertebral artery ((c), (d)). High-resolution cone-beam computed tomography (VASO CT) in the axial view shows that one perforator (arrow) from the left vertebral artery runs to the dorsolateral part of the medulla oblongata (e), and the other (arrow) runs to extracranial muscle (f). Conventional DSA in the lateral view shows the occlusion of the left vertebral artery after internal coil trapping (g). VASO CT in the coronal view shows the anterior spinal artery (arrow) originating from the right vertebral artery (h). MRI after internal coil trapping demonstrates infarction (arrow) in the dorsolateral part of the medulla oblongata (i).

In case no. 4, a 60-year-old man presented with sudden onset of severe headache and nausea. The DSA showed a left VAD that was 14 mm × 7 mm in diameter without evidence of subarachnoid hemorrhage (Figure 3(a) and (b)). One month after onset, the pathological lesion still remained with the same size and shape, so endovascular treatment with internal coil trapping was performed. Conventional DSA showed that the anterior spinal artery arose from just distal to the lumen of the VAD, and the medullary perforator arose from the dome of the VAD (Figure 3(a) and (b)), but 3D-RA could not detect the medullary perforator (Figure 3(c)). VASO CT visualized the perforators more clearly, and the axial view showed that the perforator ran to the ventrolateral part of the medulla oblongata (Figure 3(d) and (e)). Although the anterior spinal artery was preserved after internal coil trapping, one small medullary perforator was occluded with the VAD (Figure 3(f) and (g)). Although this patient had no focal neurological deficit after the operation, the MRI revealed some high-intensity spotty areas in the perfusion area of this perforator (Figure 3(h)).

Figure 3.

Case no. 4.

Conventional digital subtraction angiography (DSA) in the anteroposterior (A-P) view shows left vertebral artery dissection (VAD) located distal to the posterior inferior cerebellar artery, the anterior spinal artery (short arrow) originating from just distal to the VAD, and the medullary perforator (long arrow) originating from the VAD ((a) and (b)). Three-dimensional rotational angiography shows the anterior spinal artery (arrow), but not the medullary perforator (c). High-resolution cone-beam computed tomography (VASO CT) in the coronal view shows both the anterior spinal artery (short arrow) and the medullary perforator (long arrow) (d). VASO CT in the axial view shows that the medullary perforator (arrow) originating from the VAD runs to the ventrolateral part of the medulla oblongata (e). Conventional DSA shows the occlusion of the VAD (f). VASO CT in the coronal view after internal coil trapping shows the anterior spinal artery originating from the vertebral artery stump (g). Magnetic resonance imaging after internal coil trapping demonstrates infarction (arrow) in the ventrolateral part of the medulla oblongata (h).

Discussion

VASO CT provides more information about perforators in relation to other anatomical structures than 2D DSA or 3D-RA. In addition, the axial view on VASO CT showed the horizontal tracts of medullary perforators, which were not seen on conventional DSA images. This improvement of the spatial resolution contributes to effective strategy in the treatment of VAD. This modality provides accurate information for the endovascular management of VAD.

In four of the present cases, the perforators that were not detected on 3D-RA could be identified on VASO CT. In these cases, VASO CT was very useful and practical, because this was the only modality that provided the reconstruction images in terms of three dimensions on orthogonal planes (axial, sagittal, and coronal). 3D-RA images reconstructed in thin-slice 3D multi-planar reconstruction mode, which we did not usually make, could potentially detect the perforators. However, the tracts of perforators can be assessed along with the adjoining tissue structures, such as the skull and the brainstem, on VASO CT. From this perspective, more accurate information about functional vascular anatomy associated with VAD could be obtained from VASO CT. In case no. 3, two perforators arose from the VAD, and the axial view of VASO CT showed clearly that one perforator was supplying the area of the medulla oblongata, and the other was contributing to the extracranial muscle branch (Figure 2(e) and (f)). These perforators could be differentiated using 2D DSA. However, VASO CT, especially the axial view, gives more accurate information on perforators than 2D DSA or 3D-RA.

Postoperative medullary infarctions have been reported as a result of internal coil trapping, and some studies investigated the risk factors related to medullary infarction following internal coil trapping.^4,5,9,10 According to the literature, the incidence of MRI-confirmed medullary infarction is 19%–47%.^4,5,9,10 The main factor was considered to be the total length of the entire trapping lumen of the vertebral artery. Ikeda et al.⁵ reported that a longer length of trapped area, specifically the segment proximal to the dilated portion, was associated with a higher rate of postoperative medullary infarction. Since critical perforators such as the anterior spinal artery originate from the distal segment of the VA,¹¹ coils tend to be fewer in that area, and they tend to be more in the proximal segment to prevent recanalization. However, the medial area of the medulla oblongata has rich anastomoses coming from the contralateral vertebral artery or the lower part of the basilar artery.¹¹ On the other hand, the perforators originating from the vertebral artery just proximal to the PICA usually supply the lateral area of the medulla oblongata, and, therefore, this area has fewer anastomoses.¹² Therefore, it is very important to detect perforators originating from the proximal segment of the dilated portion, as well as the anterior spinal artery originating from the distal segment, and to confirm that these perforators run to the medulla oblongata, not to muscle or other tissues, to decide the range of coil embolization.

Since unruptured VAD tends to follow a benign clinical course, conservative treatment has been advocated.^6–8 However, some reports suggested that patients with VADs who had a lesion with a relatively large profile or showed changes during the period of observation should undergo intervention.^8,13 Naito et al. even suggested that the risk of bleeding from unruptured VAD is higher than previously considered, because, in their series, three of 21 unruptured VAD patients developed subsequent subarachnoid hemorrhage, and four of 14 ruptured VAD patients had previously experienced headaches or neck pain.¹³ Since the optimal guideline or indications for treatment of unruptured VADs have not yet been established, we must carefully determine the indications for unruptured VAD while considering the risk of bleeding with conservative treatment. The morbidity and the clinical outcome of endovascular treatment are closely associated with the risk of ischemic complications associated with the perforating artery. Therefore, it is an essential step to identify the anterior spinal artery and the medullary perforators in the management of embolization. VASO CT is today the key modality to obtain such accurate information regarding these perforators. If ischemic complications could be predicted based on VASO CT, stent-assisted coil embolization or conservative treatment could be performed. We decided to choose conservative treatment for the patient with an unruptured VAD in Figure 4 because VASO CT confirmed the medullary perforator arising from the pathological lumen of the VAD (Figure 4 (c)). In this situation, the ischemic risk was considered higher, even with the use of stent-supported coil embolization.

Figure 4.

Right vertebral angiogram.

Conventional digital subtraction angiography (DSA) shows right vertebral artery dissection (VAD) located distal to the posterior inferior cerebellar artery and a medullary perforator (arrow) originating from just distal to the aneurysm. Three-dimensional rotational angiography does not show the perforator (b). High-resolution cone-beam computed tomography (VASO CT) shows the perforator (arrow) more clearly than conventional DSA (c).

In the majority of VADs, it is still difficult to discriminate between the true lumen and the false lumen on VASO CT images. MRI has the advantage of allowing analysis of the vessel wall. High-resolution MRI is reported to be the best modality;¹⁴ it reveals mural hematoma in a false lumen, which clarifies the difference between true and false lumens. The further advancement of VASO CT will allow visualization of the smallest perforators. This technology, together with optimization of vessel wall imaging, will improve the clinical outcomes of VADs in the near future.

Compliance with ethical standards

We declare that all human studies have been approved by our ethics committee and have therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. We declare that all patients gave informed consent prior to inclusion in this study.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.