A clinical comparative study of dual LVIS devices and single flow diversion stenting for the treatment of unruptured V3–V4 vertebral artery dissection

Abstract

Purpose

This study aims to compare the efficacy and safety of using overlapping low-profile visualized intraluminal support (LVIS) devices and flow diversion (FD) for the treatment of unruptured vertebral artery dissection (VAD) in the V3–V4 segments.

Methods

The clinical and imaging data of 71 patients with unruptured VAD in the V3–V4 segments who underwent either dual LVIS stenting (d-LVIS group) or single FD stenting (FD group) at our center from September 2014 to December 2021 were retrospectively analyzed.

Results

Immediate postoperative angiography revealed no significant difference in the degree of occlusion between the two groups in treating vertebral artery dissecting aneurysms (with or without noncompact coiling). However, the d-LVIS group had significantly higher fluoroscopy exposure time and total radiation exposure dose compared to the FD group. During the perioperative period, two cases of pontine infarction and one case of acute thrombosis were encountered. One patient died from subarachnoid hemorrhage during the follow-up period. For dissecting the aneurysm, angiographic follow-up (8.56 ± 1.96 months) showed similar healing outcomes between the two groups (with or without noncompact coiling). However, seven patients (7/40, 17.5%) showed poor healing and one patient showed mild in-stent stenosis. For simple dissection, angiographic follow-up (8.78 ± 1.83 months) showed patent lumens in both groups, with all dissections healing well, and two patients having mild in-stent stenosis.

Conclusion

Both methods could effectively treat unruptured VAD in V3–V4 segments. Nevertheless, simple FD implantation is relatively easier to perform and involves lower radiation exposure.

Introduction

Vertebral artery dissection (VAD) is caused by the disruption of the internal elastic lamina, which may occur as a consequence of trauma or spontaneously. ¹ The pathological changes are as follows: blood enters the vessel wall along the torn intima to form intramural hematoma, which expands the vessel wall and leads to lumen compression and stenosis, resulting in posterior circulation ischemic symptoms, and even further develop to form a dissecting aneurysm resulting in a mass effect or subarachnoid hemorrhage (SAH).^1,2 The V3–V4 segments of the vertebral artery are particularly susceptible to dissection owing to their unique vascular tortuosity and altered arterial blood flow, unlike the V1–V2 segments. ³

In addition to conservative treatment such as medical anticoagulation, commonly used invasive treatment methods mainly include microneurosurgery and endovascular intervention. Surgical treatment of VAD mainly depends on the sacrifice of the diseased vessel. Based on the relative location of posterior inferior cerebellar artery (PICA) and dissecting lesions, it mainly includes complete trapping, incomplete trapping, and revascularization of the PICA when necessary. ⁴ Such strategies are more commonly used in patients with acute SAH. In addition, their location at the base of the skull and proximity to the brain stem, along with their origin from the posterior inferior cerebellar artery and some perforating vessels of the brain stem in the neck region, pose considerable challenges for neurosurgeons. ⁵

At present, endovascular treatment has become the main treatment for VAD. The main strategies of endovascular treatment include occlusion of the pathological vessel and vascular remodeling. The remodeling of the lumen depends on the flow diverter of the stent. ⁶ There is increasing evidence that endovascular vascular remodeling via overlapping stent implantation or intracranial blood flow diversion (FD) is effective and safe in the treatment of VAD.^6–11 However, there are still no clear clinical studies comparing multiple stents versus single FD implantation or those determining the most effective stent. Therefore, this study retrospectively analyzed the clinical and imaging data of patients with V3–V4 VAD treated using the low-profile visualized intraluminal support (LVIS) device (MicroVention-Terumo, California, USA) and Tubridge FD (MicroPort Medical, Shanghai, China), to investigate their efficacy and potential complications.

Methods

Study population

Clinical and imaging data were collected from patients diagnosed with V3–V4 segment VAD via MRI and/or angiography at the First Affiliated Hospital of Zhengzhou University between September 2014 and December 2021 using the Picture Archiving and Communication System. The inclusion criteria were as follows: (1) confirmation of VAD of V3–V4 segments through high-resolution MRI and/or angiography; (2) endovascular treatment was performed through our center; (3) utilization of either dual LVIS stents overlapping implantation or single FD implantation (with or without coiling). The exclusion criteria included: (1) acute SAH was present; (2) presence of other cerebrovascular diseases, tumors, or hematological diseases.

The eligible patients who met the inclusion criteria were divided into the d-LVIS group or the FD group according to the different interventional treatment methods employed. Each patient provided written informed consent for the publication of this article and any related photos. The First Affiliated Hospital of Zhengzhou University Ethics Committee approved this work. This study complied with the 2013 revision of the Declaration of Helsinki requirements (www.wma.net/en/30publications/10policies/b3/index.html).

Stent placement strategy

The general principles of endovascular treatment are as follows: For patients with simple VAD who have no improvement in symptoms or imaging findings after six months of strict antiplatelet or anticoagulant therapy, active intervention is considered in this study; for vertebral artery dissecting aneurysms with symptoms, ugly changes in shape, aneurysm enlargement, aneurysm diameter exceeding 7 mm, mass effect or involvement of the basilar artery, active treatment should be considered.

In addition, for dissecting aneurysms involving PICA, the treatment option of stent placement within PICA was excluded from this study, and the principles for the treatment of patients included in this study were as follows: if only stent implantation without coil embolization is performed, the operation is not special; if coil embolization was needed, real-time angiography was performed during embolization, and the PICA was deliberately protected. Contralateral vertebral artery blood flow was also assessed, and balloon occlusion tests were performed if necessary.

Stent placement procedure

All patients took a daily dosage of 75 mg of clopidogrel and 100 mg of aspirin for more than five days before the procedure. Notably, after seven days of dual antiplatelet therapy, all patients received an analysis to test drug sensitivity. The drug dosage or category was changed based on the thromboelastogram analysis if necessary. Overall, patients were prescribed 75 mg of clopidogrel daily for six months and 100 mg of aspirin daily for 12 months after the procedure. Antiplatelet therapy was discontinued or extended according to angiographic findings.

The procedures for single FD placement (with or without coiling) or the first LVIS stent placement (with or without coiling) were described previously.^7,10–12 Balloon dilation was considered as appropriate for dissection with stenosis. The implantation of the second LVIS stent is the difficulty of dual LVIS device stenting. We released the first stent across the dissection lesion while retaining the delivery guide wire. The microcatheter was then delivered through the lumen of the first stent. If the tip of the stent microcatheter encountered resistance in the lumen of the stent or failed to successfully pass through the stent despite repeated attempts, a 300-cm microguide wire was shaped into a “pigtail” or “J” shape, passed through the first stent to the distal end, and then inserted into the stent microcatheter along the microguide wire. If it was unclear whether the stent microcatheter had entered the distal end through the lumen of the first stent, locally magnified high-resolution C-arm CT was performed to verify the positioning, and the microcatheter was adjusted accordingly. Finally, the second LVIS stent was delivered through the stent microcatheter and partially overlapped with the first stent to ensure maximum coverage of the dissection lesion. For single FD implantation, precautions were taken to avoid the distal end of the stent from entering the basilar artery or causing minimal impact on the blood flow in the basilar artery perforator or contralateral vertebral artery.

After stent placement, working position angiography was performed to assess the immediate occlusion of the dissecting aneurysm using the modified Raymond–Roy occlusion classification (mRRC) ¹³ and the O’Kelly–Marotta (OKM) grading scale. ¹⁴ During the procedures, the fluoroscopy time, fluoroscopy radiation dose, and total radiation dose were recorded for both groups using the automatic X-ray dose acquisition program of the workstation.

Perioperative complications and follow-up

Perioperative neural-related complications included acute stent thrombosis, new or aggravated stroke symptoms that occurred after surgery, such as new headache, limb weakness, or dysphagia, which were further confirmed by magnetic resonance imaging.

Angiographic follow-up was conducted for 6 to 12 months, whenever possible. For patients with simple dissection, good healing was determined by the following criteria: the lumen of the diseased vessel returning to normal or the patient's symptoms disappearing or improving without further stenosis or progression to a dissecting aneurysm. In the case of patients with dissecting aneurysms, healing was evaluated using the mRRC or OKM grading scale to assess aneurysm occlusion. Good healing was defined as follows: the mRRC grade was either I or II, indicating improvement compared to the initial state, and the OKM grade was either D or C, also showing improvement compared to the initial state.

The degree of in-stent stenosis was categorized as follows: mild stenosis with a range of 25% to less than 50% and moderate-to-severe stenosis when the stenosis was 50% or more. ¹⁵ Patients were followed up either through outpatient visits or by telephone, and their neurological function was scored using the modified Rankin Scale (mRS).

Statistical analysis

All statistical analyses were performed using SPSS 26.0 software (IBM, Armonk, NY, USA). Quantitative variables were expressed as mean ± standard deviation and were compared between groups using the independent sample t-test. Categorical variables were presented as numbers with percentages and were compared between groups using the chi-square test or Fisher's exact test, depending on the sample size. A p-value <.05 was considered statistically significant.

Results

Study cohort characteristics

As shown in Figure 1, a total of 367 patients were diagnosed with dissection in the V3–V4 segments of the vertebral artery, using high-resolution MRI and/or angiography. Of these, 52 patients were not treated in our hospital. In addition, six patients underwent open microsurgery, 109 patients were treated using conservative medical approaches such as anticoagulation, four cases were complicated with cerebrovascular malformations, blood diseases, tumors. The remaining 196 patients underwent endovascular treatment. Among these 196 patients, 47 underwent pathological artery occlusion with coils, 49 underwent single LVIS stent-assisted coil embolization, and 24 received other stent placements. A total of 76 patients were treated with dual LVIS stent implantation or single FD implantation, including five patients with acute SAH. Finally, 71 patients were enrolled in this study, comprising 33 patients who underwent overlapping implantation of dual LVIS stents, and 38 patients who underwent FD implantation. The general data for both groups of patients are shown in Table 1.

Figure 1.

Flow chart of inclusion and exclusion.

Table 1.

General patient data for both groups.

	d-LVIS group (N, %)	FD group (N, %)
Total number of patients	33	38
Total number
Simple dissection	9	12
Dissecting aneurysm	24	26
With coiling	12	9
Without coiling	12	17
Gender
Male	20 (60.6)	26 (68.4)
Female	13 (39.4)	12 (31.6)
Hypertension	20 (60.6)	21 (55.3)
Diabetes mellitus	7 (21.2)	7 (18.4)
Hyperlipidemia	5 (15.2)	4 (10.5)
Smoking	8 (24.2)	16 (42.1)
Alcohol	6 (18.2)	7 (18.4)
Location of lesion
V3 segment	11 (33.3)	13 (34.2)
V4 segment	22 (66.7)	25 (65.8)
Average diameter of dissecting aneurysm (mm)	3.76 ± 1.68	4.31 ± 2.09
Mean length of dissection (mm)
Simple dissection	6.12 ± 1.85	6.27 ± 1.56
Dissecting aneurysm	6.83 ± 2.40	7.16 ± 2.82
Classification of dissecting aneurysms a
Classic dissecting aneurysm	20 (83.3)	19 (73.1)
Segmental ectasia	4 (16.7)	7 (26.9)
The main accompanying symptoms
Headache or dizziness	13 (39.4)	14 (36.8)
Nausea or vomiting	6 (18.2)	7 (18.4)
Cerebral infarction or transient ischemic attack	4 (12.1)	3 (7.9)
Other symptoms	1 (3.0)	6 (15.8)
No symptoms	9 (27.3)	8 (21.1)
Pretreatment mRS
0	9 (27.3)	8 (21.1)
1	23 (69.7)	27 (71.1)
2	1 (3.0)	3 (7.9)
Follow-up mRS
0	25 (78.1)	24 (72.7)
1	6 (18.8)	6 (18.2)
2	1 (3.1)	1 (3.0)
≥3	0 (0.0)	2 (6.0)

^aClassification of dissecting aneurysms ¹⁶ : I. classic dissecting aneurysm, II. segmental ectasia, III. dolichoectatic dissecting aneurysm, and IV. large mural bleeding ectasia. No cases of the latter two more complex types were observed in this study. mRS: modified Rankin Scale.

Surgical outcomes

Both groups underwent successful operations. In the FD group, 29 patients were treated with FD implantation alone (Figure 2), and nine patients were treated with FD combined with noncompact coil embolization. In the d-LVIS group, 21 patients were treated with overlapping dual LVIS stents, and 12 patients were treated with dual LVIS stents combined with noncompact coil embolization (Figure 3). Coil embolization was not performed in cases of simple dissection (Table 1).

Figure 2.

Tubridge flow diversion for V4 segment dissecting aneurysm of the right vertebral artery. A 35-year-old man was admitted to the hospital with complaints of persistent headache and dizziness for a week. (a) Preoperative high-resolution MRI showed a dissection in the V4 segment of the right vertebral artery (arrow). (b) Angiography revealed a dissecting aneurysm in the V4 segment of the right vertebral artery (arrow). (c) A Tubridge flow diverter (FD) was placed (head and tail indicated by arrows). (d–e) High-resolution flat panel CT demonstrated complete opening of the FD and good adherence to the vessel wall. f. Follow-up angiography at 12 months showed unobstructed blood flow in the V4 segment of the right vertebral artery, and the dissecting aneurysm had healed.

Figure 3.

Dual LVIS stenting combined with coil embolization for the treatment of the V4 segment of the left vertebral artery dissected aneurysm. A 57-year-old man was admitted to the hospital with complaints of dizziness, nausea, and vomiting persisting for three days and worsening over the past day. (a) Angiography revealed a dissecting aneurysm in the V4 segment of the left vertebral artery (arrow); (b) preoperative high-resolution MRI showed a dissection in the V4 segment of the right vertebral artery (arrow); (c) the patient underwent dual LVIS stents implantation combined with noncompact coil embolization (white arrow shows the head and tail of the first stent, black arrow shows the head and tail of the second stent); (d–e) high-resolution flat panel CT demonstrated complete opening of the LVIS stents and good adherence to the vessel wall; (f) follow-up angiography at eight months revealed unobstructed blood flow in the V4 segment of the left vertebral artery, and the dissecting aneurysm had healed. LVIS: low-profile visualized intraluminal support.

No statistically significant difference was observed between the two groups in the immediate efficacy of stent placement for dissecting aneurysms (Table 2). Compared with the d-LVIS group, the FD group had significantly less fluoroscopy time, lower fluoroscopy radiation dose, and a reduced total radiation dose (Table 2).

Table 2.

Angiographic results of the two groups.

	d-LVIS group	FD group	Total number/p-value
Immediate angiographic results
VADA (without coiling)	12	17	29
A	4	8	.438*
B	7	9
C	1	0
VADA (with coiling)	12	9	21
I	1	0	1.00*
II	3	2
III	8	7
Follow-up angiography results
Simple dissection	8	9	17
Good healing	8	9	a
Poor healing	0	0
VADA (without coiling)	9	14	23
Good healing	7	11	1.00*
Poor healing	2	3
VADA (with coiling)	10	7	17
Good healing	9	6	1.00*
Poor healing	1	1
Fluoroscopy time (min)	22.88 ± 9.48	13.15 ± 5.63	<.001
Fluoroscopy radiation dose (mGy)	191.31 ± 77.85	111.84 ± 45.49	<.001
Total radiation dose (mGy)	1652.18 ± 358.30	1311.79 ± 436.88	<.001

^aThe number of patients who showed poor healing in both groups was 0, owing to which the P-value could not be calculated; *Fisher's exact test; VADA: vertebral artery dissecting aneurysm.

Perioperative complications and follow-up

Perioperative complications occurred in 3 (4.2%) patients, including two cases of brainstem infarction and one case of acute thrombosis. One patient with FD implantation developed dysphagia 48 h after the operation, and MRI revealed left pontine infarction. The symptoms significantly improved after symptomatic treatment, and the mRS score was 0 at the last follow-up. Another patient who underwent FD implantation combined with coil embolization developed new-onset dizziness 33 h after the operation, and MRI showed mild acute pontine infarction. The symptoms did not significantly improve before discharge, and the mRS score at the last follow-up was 1. A patient who underwent overlapping implantation of LVIS stents developed unilateral hemiplegia immediately after the operation. Reangiography showed acute thrombosis in the stent. The patient underwent intraarterial tirofiban infusion, and the diseased vessel was successfully reopened. However, the patient still had residual limb weakness, and the mRS score at the last follow-up was 2.

Out of the 71 patients, 58 were followed up by angiography at 6–12 months, the mean follow-up time was 8.64 ± 1.91 months. The overall good healing rate of dissecting aneurysm was 82.5% (33/40 cases). No significant difference was observed in the short-term efficacy between the two treatment methods (with or without coiling), as detailed in Table 2. All cases of simple dissection healed completely.

With regard to the effect of noncompact coiling in the treatment of dissecting aneurysms, the rate of poor healing for dissecting aneurysms without coiling (21.7%, 5/23 cases) was higher than that for dissecting aneurysms with noncompact coiling (11.8%, 2/17 cases). However, this difference was not statistically significant (p = .677). In addition, among the patients with dissection aneurysms, one in the FD group and two in the d-LVIS group showed mild stenosis in the stent. None of the three patients developed symptoms of posterior circulation ischemia during the continuous follow-up. In total, 65 patients received outpatient or telephone follow-up for a period ranging from 6 to 21 months. Among them, two patients died owing to SAH and severe pneumonia (unrelated to the lesions). The remaining patients had no complications related to the lesions. At the last follow-up, the mRS scores were as follows: 49 patients scored 0, 12 patients scored 1, two patients scored 2, and two patients scored 6.

Discussion

Numerous clinical studies have demonstrated the effectiveness of endovascular treatment for VAD.^{3,7,9,10,17,18} For ruptured dissections, the main endovascular treatment methods include occlusion of the parent artery, intracranial-covered stent implantation, or stent-assisted coil embolization. On the other hand, for unruptured dissections, revascularization is mainly considered, which includes stent-assisted coil embolization, overlapping stent implantation, or flow diverter implantation. ¹⁸ Because dissecting aneurysms differ from true aneurysms, dense-compact coiling can easily lead to enlargement of the false lumen and even bleeding, whereas noncompact coiling is prone to relapse. Overlapping stent placement increases metal coverage and provides a good flow-diverting effect.^11,17,19 Both methods can change the hemodynamics, weaken the blood flow impact at the tear site, promote blood retention in the dissection lesion to form thrombosis, and finally achieve stent endothelialization, leading to complete reconstruction of the affected vessels.^9–11

However, overlapping stents may be strongly associated with a higher risk of ischemic events in the perioperative period and could trigger more severe neointimal hyperplasia, potentially leading to stenosis or even occlusion of the parent artery.^11,12,20 In addition, although the vertebral artery has fewer perforating vessels compared to the basilar artery, the relatively high metal coverage in both methods carries the risk of occlusion of important branches.^11,21 In our center, one FD stent is more expensive than two LVIS stents, posing a significant economic burden for patients (125,000 RMB vs. 34,000*2 RMB). Therefore, it is important to explore whether the overlapping implantation of two LVIS stents or the implantation of an FD stent is more suitable for treating VAD. However, there are few clinical reports on the efficacy and safety of these two treatments, and relevant data are currently lacking.

For dissecting aneurysm, no significant difference was observed between the two groups in terms of immediate poststent implantation angiographic results and follow-up angiographic results. This similarity could be attributed to the comparable metal coverage achieved by both methods. The metal coverage of two overlapping LVIS stents can reach 23%–46%, ¹¹ while that of a single FD stent is 30%–35%. ²² Moreover, in this study, the healing rate of dissecting aneurysm treated with a combination of stent and noncompact coiling was higher than that of stent alone (11.8% vs. 21.7%), but the difference was not statistically significant, possibly owing to the small sample size. Therefore, non-compact coiling in the dome of dissecting aneurysm may be more conducive to the formation of intracranial thrombosis and to promote the healing of aneurysm.

The high rate of good healing in the dissecting aneurysms (33/40, 82.5%) observed during the angiographic follow-up in this study could be attributed to the small size of the lesions and the fact that all dissecting aneurysms were classic or segmental aneurysms, as well as the combination of two LVIS stents or the high metal coverage of the FD stent itself. Zhang et al. ¹⁶ showed that endovascular treatment of long dilated and giant hemorrhagic dissecting aneurysms was more likely to result in recanalization or incomplete occlusion compared to classical dissecting aneurysms. For patients with simple dissection that did not respond to medical therapy, the study showed favorable results in all cases treated with FD or dual LVIS overlapping stent placement. However, two patients developed asymptomatic mild in-stent stenosis. Therefore, it may be necessary to further explore whether endovascular intervention is necessary for simple dissection.

During the perioperative period, two cases of brain stem infarction and one case of acute thrombosis occurred, indicating a low incidence of perioperative complications (3/71, 4.2%). This favorable outcome may be attributed to several factors. Firstly, the appropriate stent mesh ratio used in the treatment effectively protected the patency of the involved branch vessels while successfully obliterating the dissecting aneurysm. Secondly, the application of scientific and standardized clinical guidance for antiplatelet and anticoagulation therapy after the operation likely contributed to the low incidence of complications. Moreover, it is essential to consider the specific type of dissecting aneurysms included in this study. All patients with dissecting aneurysms were classified into classic and segmental ectasia dissecting aneurysms. The incidence of complications of this type of dissecting aneurysm is theoretically much lower than that of complex dissecting aneurysms, such as dolichoectatic dissecting aneurysms and giant bleeding aneurysms.

The fluoroscopy radiation dose and total radiation dose in the d-LVIS group were significantly higher than those in the FD group. This discrepancy can be attributed to the complexity of the operation involved in dual LVIS stent implantation. In the FD group, the procedure for FD implantation is relatively straightforward, which likely contributes to reduced fluoroscopy time and radiation exposure. However, in the d-LVIS group, the process of accurately and smoothly passing the delivery catheter of the second LVIS stent through the lumen of the first stent is challenging. This step can sometimes be difficult owing to factors such as the softness of the delivery guide wire and the obstruction caused by the stent mesh. In addition, when attempting to repenetrate the stent body using a microguidewire and microcatheter, there is a risk of inadvertently entering between the stent and the vessel wall from the proximal stent orifice mesh, which increases the difficulty of the operation and may lead to unnecessary radiation exposure, which further contributes to the higher radiation dose observed in the d-LVIS group.

To sum up, compared with dual LVIS devices overlapping implantation for the treatment of unruptured V3-4 VAD, single FD implantation has a simpler surgical procedure, shorter fluoroscopy time, lower radiation dose, and similar efficacy. In view of this, unless the patient is financially unable to afford it, Single FD stenting will be a priority for treatment.

This study has several limitations, which should be considered when interpreting the results. First, the study design was retrospective, which introduces inherent biases in the selection of surgical procedures. Second, the operation time and radiation exposure time are influenced by the proficiency and experience of the operator. Skilled operators may be more efficient and may help to reduce the operation time and unnecessary radiation exposure. Third, this study included the Tubridge FD as the flow diverter option. Not including other types of flow diverters limits the generalizability of the results to a broader range of devices. Fourth, the number of patients in each treatment group was small, and the follow-up duration was relatively short. Finally, the lack of randomization in treatment assignment might have introduced selection bias and confounding factors. The efficacy and safety of the treatment need to be confirmed by prospective, large-sample, randomized controlled trials.

Conclusion

In conclusion, both dual LVIS stents overlapping implantation and single FD implantation have been shown to be effective in treating VAD in V3–V4 segments. However, single FD implantation is relatively simple to operate and is associated with lower radiation exposure compared to dual LVIS stents overlapping implantation.