Endovascular treatment of acute basilar artery occlusion caused by vertebral artery stump syndrome: A clinical analysis of 37 cases

Abstract

Introduction

Acute basilar artery occlusion (ABAO) caused by vertebral artery stump syndrome (VASS) has a low incidence and is always underestimated. Due to the occlusion of the origin of the vertebral artery (VA), it is often combined with basilar artery (BA) endovascular diseases or non-dominant contralateral vertebral artery, making the endovascular treatment (EVT) challenging to implement.

Objective

This article focuses on whether EVT and two interventional route options could bring clinical benefits to this group of patients: basilar artery thrombectomy through the occluded lateral vertebral artery and implementing revascularization of the occluded vertebral artery (dirty-road-path); thrombectomy through the non-occluded lateral vertebral artery (clean-road-path).

Methods

We collected six cases of acute embolic basilar artery occlusion (ABAO) due to VASS from January 2020 to December 2021 at our hospital and retrospectively analyzed 31 patients previously reported in the literature and applied statistical analysis to investigate the treatment options and clinical prognosis of these patients.

Results

The clean-road-path surgical protocol was applied in 4 of 37 patients, the dirty-road-path protocol was applied in 29 patients, and 4 patients did not recanalized the basilar artery. By statistical analysis we found that successful recanalization of the basilar artery was clinically significant in reducing the modified Rankin Scale (mRS) scores in these patients, the statistical difference in the benefit of the two surgical protocols was negative. There was a significant positive correlation between preoperative National Institute of Health Stroke Scale (NIHSS) and postoperative 90-day mRS scores.

Conclusion

Endovascular treatment can benefit patients with ABAO due to VASS, and patients with higher preoperative NIHSS scores are more vulnerable to getting a poor prognosis. Comparison between the two endovascular options did not yield statistically significant results, but the dirty-road-path option may be superior to using the clean-road-path.

1. Introduction

ABAO is rare, accounting for 1% of all ischemic strokes. However, the ABAO-related mortality or disability rate reaches to almost 80% [[1], [2], [3]]. Even though EVT has been widely used, the mortality in ABAO remains high. Among patients with acute large vessel occlusion treated with thrombectomy, the prevalence of acute ABAO was 10.4%. VASS accounts for 1.4% of acute posterior circulation ischemic strokes [4]. Due to its low incidence, VASS is often underestimated in acute ABAO. However, the presence of VA origin occlusion often makes EVT difficult to perform, especially when the origin of bilateral VA or contralateral unoccluded VA cannot be visualized by angiography, the posterior communicating artery and the collateral access from the cervical trunk of the thyroid becomes the only approach for EVT. And endovascular procedures through these accesses require relatively high technical skills for a performer. Moreover, the endovascular procedure is difficult, which in turn leads to long operation time and increases the risk of postoprocedure fatal complications. At present, dirty-road-path and clean-road-path were reported as the main EVT schemes for ABAO caused by VASS [[5], [6], [7]], and there is no comparison between the two programs. Thus, we aim to present an investigation of the clinical benefit of EVT and these two endovascular options in patients with ABAO due to VASS

2. Materials and method

2.1. Materials

This is a retrospective case series and clinical analysis. 78 inpatients received thrombectomy for acute basilar artery occlusion (ABAO) at the Stroke Center of our hospital from January 2020 to December 2021. According to the diagnostic criteria of VASS: (1) acute ischemic stroke in the posterior circulation; (2) occlusion of the origin of the vertebral artery; (3) presence of distal forward flow in the ipsilateral vertebral artery; (4) lack of other causes of ischemic stroke [8]. Finally, 6 patients with ABAO caused by VASS were included in the case series.

The clinical information of the patients was collected, including gender, age, area of infarction, lesion arteries, collateral compensatory pathways, endovasclar procedures, preoperative mRS (Pre-mRS), preoperative and postoperative NIHSS, mRS scores at discharge, and 90-day mRS scores. According to the 90-day mRS scores, 0–3 was defined as good prognosis, 4–6 was regarded as a poor prognosis. 5 of the 6 patients received endovascular treatment, 4 patients of whom were treated with a similar plan to endovasclar procedure A (described below), and 1 patient was treated with endovascular procedure B (described below). We also retrospectively analyzed 31 cases in previous literature, and collected patients' preoperative NIHSS (Pre-NIHSS), endovascular procedures, pre-mRS scores and 90-day mRS scores.

2.2. Methods

First, we divided all these patients into four groups，patients with good prognosis after the successful recanalization was defined as group A, and patients with the poor prognosis after the successful recanalization was defined as group B; Similarly, in the non-recanalization patients, those with good prognosis was defined as group C and those with poor prognosis was defined as group D. Then we do the heterogeneity test through subgroup meta-analyses by statistical software Stata 15 (Stata Corp, College Station, TX) (Fig. 3). After the test was passed, the cases were divided into the successful recanalization of basilar artery group and the non-recanalization group; Successful recanalization patients was divided into two groups according to Procedure A and B. Then the statistical software IBM SPSS Statistics 26 was applied to obtain statistical data based on the difference value in mRS score (pre-mRS minus 90-day mRS) index by independent sample t-test between groups, and a Pearson's correlation analysis of Pre-NIHSS score and 90-day mRS score in all these 37 patients was performed to indicate the relevance between Pre-NIHSS score and prognosis of all these patients. All statistical tests were two-sided with a P value < 0.05.

Fig. 3.

Forest plot of meta-analysis.

To ensure the homogeneity of the included patient groups, we performed the heterogeneity test by meta-analyses. Two independent samples t-test was used to test the difference value of mRS score (pre-mRS scores minus 90-day mRS scores) between the successful recanalization patients and the non-recanalization patients, and the difference between Procedure A and Procedure B. All statistical tests were two-sided with a P value < 0.05 and performed using Stata 15 (Stata Corp, College Station, TX).

All patients in the collection had tandem vertebral artery and basilar artery lesions. Procedure A (dirty-road-path) was chosen when the occluded vertebral artery was easy to access: basilar artery embolization was performed via the occluded side of the vertebral artery, followed by revascularization of the occluded vertebral artery. When the occluded side of the vertebral artery is difficult to access and the contralateral VA can be used as the endovascular route, Procedure B (clean-road-path) was chosen: basilar artery embolization is performed through the unoccluded VA without dealing with the occluded VA.

Procedure A (case five in our stroke center): The cerebral angiography showed that it was type III aortic arch, the origin of the right vertebral artery (RVA) was occluded and the stump can be seen locally, the right thyroid neck trunk was anastomosed with the V2 segment of the RVA through the ascending carotid artery supplying blood to the distal RVA, the distal end of the BA was not visualized and the left VA terminated in the left posterior inferior cerebellar artery (PICA), the left VA supplied blood to the right VA through the anterior spinal artery. Because the patient's RVA was the dominant artery and the stump was obvious, we decide to perform Procedure A. First, we placed a coronary 3.0*12 mm balloon to dilate the origin and V1 segment of RVA to re-open the VA stump. Then the XT-27 microcatheter was placed in the P2 segment of the left PCA through the Synchro (0.014 mm*200 mm) guide wire, and the Solitaire-AB stent (4.0 mm*20 mm) was released under the microcatheter to perform BA thrombectomy. In the end, 3 vertebral artery balloon stents (4.0 *18 mm*2, 4.0*15 mm*1) were released at the origin of the RVA to the proximal end of V2 segment. (Fig. 1A–E).

Fig. 1.

Surgical Protocol A. Protocol A: A-E, A: The right VA was occluded at the beginning and the stump was visible; the right metacervical trunk was anastomosed to the right VA at the V2 segment through the ascending carotid artery and supplied blood to the distal end; the distal part of the BA was not visualized. B: the left VA supplied blood to the right VA via the anterior spinal artery. C: 5F-Navien intermediate catheter passed over the occlusion segment to the V2 segment, angiography showed the BA occluded. D: After thrombectomy via the SAB (4.0*20 mm) stent, BA, bilateral PCA and superior cerebellar arteries were all visualized. E: 3 VA stents was released at the origin of the RVA to the proximal end of V2 segment, mTICI3.

Procedure B (Case three in our stroke center): cerebral angiography showed that the right VA and the right PICA were slender and the distal BA was not visualized, from the origin of the left VA to the V2 segment was occluded, and there was no obvious stump at the beginning, the left external carotid artery anastomoses with the left V3 segment supplied blood to the VA through the occipital artery but the left V4 segment was not visible, the PCA was open bilaterally and supplied by the internal carotid artery. Because the patient's left VA had no obvious stump, the antegrade and retrograde plan was difficult to perform, we decide to perform Procedure B. We placed the 5F intermediate catheter at the proximal V4 segment of the left VA through the 8F guide catheter, and the Rebar-18 microcatheter was selected into the P2 segment of the left PCA under the guidance of Synchro (0.014in*200 cm) micro-guide wire. After confirming that the distal end was unobstructed, a Solitaire-AB stent (4mmX20mm) was released at the BA through the microcatheter to perform the BA thrombectomy. Then the BA, anterior inferior cerebellar arteries, and superior cerebellar arteries were visualized which indicates that the thrombectomy was successful (Fig. 2F-I).

Fig. 2.

Surgical Protocol B. Protocol B: F–I F: Left vertebral artery occluded from the beginning to the V2 segment, with no obvious stump at the beginning. G: The left external carotid artery anastomosed with the left V3 segment of the VA through the lateral branch of the occipital artery, and the left V4 segment of left VA was not visible. H: The 5F intermediate catheter was guided to proximal V4 segment of the right VA via the Synchro (0.014in*200 cm) microguide wire through 8F guiding catheter, then the microcatheter a Solitaire AB (4mmX20mm) stent was peterlaced to remove the embolus. Re-angiography showed BA, the AICA and superior cerebellar arteries were visualized. I: after thrombectomy, the left occipital artery supplied blood to the left VA through collateral branches, and V4 segment of the left VA, BA and bilateral superior cerebellar arteries were all revascularized.

3. Results

From January 2020 to December 2021, 487 patients received emergency endovascular treatment at our Stroke Center. Among them, 78 patients were diagnosed as ABAO, and 6 patients were ABAO caused by VASS. ABAO patients accounted for 16% of all the cerebral thrombectomy patients, and VASS patients accounted for 7.6% of ABAO.

Detailed clinical data of 6 patients were collected. The infarct sites included the cerebellum, thalamus, pons, and occipital lobe; occluded vessels were the vertebral artery, left posterior inferior cerebellar artery, basilar artery, and posterior cerebral artery. The mean recanalization time in the 5 patients was 148.8 (range 78 to 200) minutes without technical complications such as perforation, bleeding, or re-embolization. Four of 6 cases were successfully applied with the dirty-road-path protocol, three patients had a good prognosis with no recurrence, one patient died due to a complication of malignant cerebellar edema. One case applied with the clean-road-path protocol achieved the BA recanalization and obtained a good prognosis with no recurrence during the 90-day follow-up. Another case died due to his family's refusal of surgery (Table 1).

Table 1.

Information and clinical data of the 6 patients from the stroke center of our hospital.

Patient number	A/G	Pre-NIHSS	Pre-mRS	Site of infarct	Collateral	Arterial lesion	Procedure	Postoperative	NIHSS/mRS at Discharge	TPTR（min）	90d-mRS	Recurrence	IVT	Cause of Death
1	66/M	22	4	Bil-Brainstem R-Occipital lobe,R-Cerebellum	ACA	LVA,BA	A	Aspirin, Clopidogrel	5/3	146	2	None	None	/
2	86/M	26	5	L-Cerebellum,R-Thalamus,L-mesencephalon	ACA	LVA, BA	A	Aspirin, Clopidogrel	0/0	154	0	None	None	/
3	57/F	17	4	L-Cerebellum	DCA,OA	LVA,BA	B	Aspirin, Clopidogrel	2/1	78	0	None	None	/
4	72/M	22	4	Bil-Cerebellum	ACA	LVA,PICA,BA	A	Aspirin, Clopidogrel	Death	200	6	Death	None	circulatory collapse，Severe pneumonia
5	80/M	18	4	L-pons	ACA,ASA	RVA,BA	A	Aspirin, Clopidogrel	0/0	166	0	None	None	/
6	85/M	26	5	R-Pons	ACA	Bil-VA,BA	–	–	Death	–	6	Death	None	Cerebral hernia

NIHSS: National Institute of Health Stroke Scale; mRS: modified Rankin Scale; ACA: Ascending cervical artery; DCA: Deep carotid artery; OA: Occipital artery; ASA: anterior spinal artery; TPTR: Time from puncture to revascularization; Bil: bilateral; R: right; L: left; M: male; F: female; LVA: left vertebral artery; RVA: right vertebral artery; BA: basilar artery; PICA: posterior inferior cerebellar artery; Pre-NIHSS: preoperative NIHSS; Pre-mRS: preoperative mRS; post-NIHSS: postoperative NIHSS; A/G: age/gender. IVT:intravenous thrombolysis.

The results of heterogeneity test of Meta-analysis suggested that there was good homogeneity among the studies (Q = 1.42, p = 0.9649) (Fig. 3). The mean score of preoperative NIHSS was 22.89 (range 4 to 35, n = 37), and the mean score of 90-day MRS was 3.57 (range 0 to 6, n = 37), the mean score of preoperative MRS was 4.27 (range 2 to 5, n = 37). For the non-recanalization group the mean numerical value of pre-mRS scores minus 90-day MRS scores was −0.75 (range −1 to 0, n = 4), and the successful recanalization group was 1.00 (range −3 to 4, n = 33). Independent sample t-test in both groups suggested that the numerical value was higher in the successful -recanalization group (p < 0.01), which indicated that successful recanalization could reduce the mRS score. The preoperative NIHSS score was positively correlated with the postoperative 90-day mRS score (r = 0.594, p < 0.001). 29 patients were applied with Procedure A, 16 cases had good prognosis (55.2%, 16/29), and another 13 cases showed poor prognosis (44.8%, 13/29). In 4 patients applied with protocol B, 2 had good prognoses and 2 had poor prognoses. The independent t-test between protocol A and protocol B based on the difference value in MRS score (pre-mRS scores minus 90-day MRS scores) showed no statistically difference (p = 0.469). 4 patients did not have successful basilar artery recanalization, 3 died and 1 had an mRS score of 5 at 90 days after arterial thrombolysis (Table 2, Table 3). In the pooled analysis of 37 patients, 33 patients were successfully revascularized, 18 (54.5%，18/33) cases had a good prognosis, 15 (44.5%, 15/33) cases had a poor prognosis, 4 patients who were not successfully revascularized all showed a poor prognosis.

Table 2.

Summary of the patients.

Quote	Number-Of -patients	Protocol A	Protocol B	Non-recanalization	90d-mRS 0-3	90d-mRS 4-6
Wenbin Zhang, et，al [10]	10	8	–	2	3	7
Ryosuke Maeok, et，al [5]	2	1		1	1	1
Akinari Y amano, et，al [7]	1	–	1	–	–	1
José E. Cohen, et，al [6]	7	5	2	–	4	3
Robert D Ecker, et，al [12]	6	6	–	–	4	2
Bradley A. Gross,et,al [13]	5	5	–	–	2	3
Our cases	6	4	1	1	4	2
Summary	37	29	4	4	18	19

90d-mRS: 90 day modified Rankin scale.

Table 3.

Statistical results.

	Successful recanalization n = 33	Non-recanalization n = 4	P Value	Procedure A (Dirty-road-path) n = 29	Procedure B (Clean-road-path) n = 4	P Value
Pre-mRS	4.18 (2–5)	5 (5–5)	<0.001	4.21（2–5）	4（3–5）	0.667
90-day mRS	3.3 (0–6)	5.75 (5–6)	<0.001	3.24（0–6）	3.75（0–6）	0.695
difference value	1 (-3–5)	−0.75 (-1–0)	<0.001	1.10 (-2–5)	0.25 (-3–4)	0.469

Pre-mRS: preoperative modified Rankin scale; 90-day mRS: 90-day modified Rankin scale; difference value: the value of pre-mRS scores minus 90-day mRS scores.

4. Discussion

Independent sample t-tests based on difference value in mRS score (pre-mRS scores minus 90-day mRS scores) between the successful recanalization group and non-recanalization group showed that the successfully revascularized group had decreased mRS scores (p < 0.001). Compared to the nearly 80% mortality and disability rate described in the previous literature [1,2], it is obvious that successfully revascularization of the occluded BA could significantly reduce the mortality and disability in patients with ABAO. This similar findings has been reported in previous case series and small sample clinical studies [7,[9], [10], [11]], but they did not perform a higher level of statistical analysis to make it clear until this data was published.

Meanwhile, a previous single-center, small-sample clinical study of posterior circulation tandem lesions had suggested that preoperative NIHSS score and time from onset to lesion vascularization were independent predictors of prognosis [11].ABAO due to VASS also belongs to the category of posterior circulation tandem lesions, and our results of a positive correlation between preoperative NIHSS score and 90-day mMRS (r = 0.594, P < 0.001) were coincideded with this study, Which indicated that patient with higher preoperative NIHSS scores may be more vulnerable to a poor outcome.

What is more, how to choose endovascular approaches is a problem that clinicians need to consider when dealing with this type of disease. The endovascular procedure B is less difficult to perform and can rapidly revascularize the occluded BA through the patent VA. For the ABAO patients with difficult endovascular approach via occluded VA, it is undoubtedly the optimal option to revascularize the occluded BA via patent VA, but retrospective analysis shows that this option has the risk of recurrent posterior circulation stroke [7].However, compared with lean-road-path procedure, the dirty-road-path procedure requires more techniques and the potential risks are also greater. Identifying the origin of the occluded VA and passing Micro-guide-wire after identification are often time consuming, which may eventually lead to the failure of the operation. Choosing the antegrade approach or the reverse approach is also an open question. Anterograde approach is to perform occlusive VA angioplasty and stent placement first, and then perform BA thrombectomy. The disadvantage is that the BA revascularization time is delayed, which may lead to a poor long-term prognosis of patients. Reverse approach is to perform occlusive VA angioplasty first, then perform BA thrombectomy, and finally perform occlusive VA stenting. The advantage is that the occluded BA can be revascularizaed more quickly to recovery of blood flow, decreasing the nerve cells death in brain. But the VA may re-occlude during the procedure, making later endovascular plan difficult to implement [2,7]. Some authors advocated that reverse-approach for such posterior circulation tandem lesions provided better long-term benefits to patients [12], nevertheless, prospective studies or large samples studies are still lacking. Four of the five patients whom we applied with the reverse approach had a good outcome and no recurrent posterior circulation strokes at 3-month follow-up, and one patient with successful BA revascularization died of malignant post-ischemic cerebellar edema.

In addition, according to Poiseuille's law blood flow is proportional to the quadratic of the vessel radius and the dominant VA may represent a much larger percentage of the total basilar blood flow. As a channel to occluded BA, occlusion or stenosis of the dominant VA can affect the speed and degree of vascular recanalization in intra-arterial thrombectomy with ABAO. A clinical study of a endovascular procedure for acute basilar artery occlusion noted that occlusion of the dominant vertebral artery is often associated with poor prognosis [3]. Therefore, when choosing the endovascular plan, it is not only necessary to consider whether the occlusion VA is easy to pass through, the time of initial symptoms and possible intraoperative complications, and which side is the dominant vertebral artery also needs to be included in the evaluation scope [3,11,13]. When the occluded dominant VA is not easy overtake, and the contralateral hypoplasic VA is unfeasible to perform BA thrombectomy, some scholars proposed the application of the SHERPA scheme for surgery [14]. The implementation of this scheme is similar to that of Procedure A, the main difference between them is that the identification of the culprit VA is through a micro-guide wire from the contralateral VA. This scheme provides new endovascular opportunities for such special patients, but both reported cases presented with a poor prognosis, so more clinical practice is still needed to corroborate its effectiveness.

Finally, there is no consensus on the exact etiology of stroke in the posterior circulation due to VASS, the pathogenetic mechanism of VASS might be the distal limit of the expanded thrombus and emboli caused by the stagnating clot fragment in occluded VA or a low-flow state supplied by the collateral flow via the deep cervical arteries [4,8,15]. From the pathological perspective, revascularization of the occluded vertebral artery according to Procedure A reduced the risk of reoccurrence of ABAO or other ischemic stroke in the posterior circulation, which might bring a good long-term outcome for patients. Unfortunately, based on the present data, we did not draw a statistical conclusion that protocol A is better than protocol B. The main reason may be that the sample size of protocol B is too small to reach more reasonable conclusions.

Our studies has several limitations. Since the information of all these 37 patients is uncomplete, only the preoperative NIHSS score was verified as a factor affecting prognosis. The time from a puncture to recanalization of the occluded BA, whether thrombolysis before endovascular procedure, and the treatment plan after endovascular procedure, could be independent factors affecting prognosis. The sample size of non-recanalization group and Procedure B are relatively small, so the statistical bias and error are unavoidable. And another significant limitation is that no mRS scores before stroke and at admission were found in most of these reviewed cases, but as a compensation, the Pre-mRS Scores (the mRS score before thrombectomy) were compared between them, for it can to some extent reflect the status of the patient before thrombectomy.

5. Conclusion

Since the first vertebral artery stump syndrome reported in 2008 [8], there have been only sporadic small sample retrospective analysis and case reports in literature. At present, large sample researches of this disease are still lacking, however, ABAO caused by VASS is a clinical problem which need to be confronted with. In this study, Endovascular treatment was proposed to be performed in patients with ABAO due to VASS, and in terms of path strategy, dirty-road-path option or clean-road-path option is dependent on performer's consideration, although the former may provide a good long-term outcome. Yet, further studies are needed to explore the optimal treatment strategy for ABAO caused by VASS.

Author contribution statement

Keqi Lei: Conceived and designed the experiments; Analyzed and interpreted the data; Wrote the paper.

Weiping Chen: Contributed reagents, materials, analysis tools or data; Analyzed and interpreted the data.

Zhijuan Cheng and Kai Wang: Analyzed and interpreted the data; Wrote the paper.

Fang Li: Performed the experiments; Wrote the paper.

Min Yin: Performed the experiments; Analyzed and interpreted the data.

Xingen Zhu: Conceived and designed the experiments; Analyzed and interpreted the data.

Hua Guo: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data.

Jianglong Tu: Conceived and designed the experiments; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Funding statement

Zhijuan Cheng was supported by Jiangxi Provincial Department of Science and Technology key R&D program - general project: Research and application of key technologies for predicting the death risk of stroke patients by using artificial intelligence technology [20203BBGL73129].

Weiping Chen was supported by Science and Technology Project of Jiangxi Provincial Health Commission 20201043: Research on the pathophysiological role and mechanism of TLR4 activation in alpha-synuclein transgenic Parkinson's disease model mice 2020-2022 [20201043].

Xingen Zhu was supported by National Clinical Research Center for Geriatrics--Jiangxi branch center [2021ZDG02001].

Data availability statement

Data included in article/supplementary material/referenced in article.

Declaration of interest’s statement

The authors declare no conflict of interest.