Basilar artery occlusion: A review of clinicoradiologic features, treatment selection, and endovascular techniques

Abstract

Basilar artery occlusion (BAO) is an infrequent but often fatal subtype of stroke. Predicting outcomes and selecting patients for endovascular therapy (EVT) remains challenging. Advances in neuroimaging and the development of prognostic scoring systems have augmented clinical decision-making over time. Recent randomized trials, BEST (Basilar Artery Occlusion Endovascular Intervention vs. Standard Medical Treatment), BASICS (Basilar Artery International Cooperation Study), BAOCHE (Basilar Artery Occlusion CHinese Endovascular Trial) and ATTENTION (Endovascular Treatment for Acute Basilar Artery Occlusion), compared EVT and medical management for patients with BAO. These trials yielded mixed results. The former two suggested unclear benefit while the latter two supported a benefit of EVT. While all had limitations, most providers agree caution should be exercised when excluding patients from EVT who may stand to benefit. Further studies are therefore needed to determine the effectiveness, safety, selection criteria, and optimal technical approach for EVT among patients with BAO. Hyperacute-phase advanced imaging can offer several benefits to aid decision making. It is reasonable to exclude patients with low National Institutes of Health Stroke Scale (NIHSS), large imaging-proven cores, and evidence of perforator occlusion by branch atheromatous disease. Herein, we review the clinical presentation, imaging work-up, treatments, and clinical outcomes for BAO, while highlighting knowledge gaps in treatment selection and technique.

Mesh terms

• -Basilar Artery / diagnostic imaging
• -Thrombectomy
• -Treatment Outcome
• -Humans
• -Endovascular Procedures / methods*

Introduction

Basilar artery occlusion (BAO) is the most devastating stroke subtype with a high rate of morbidity and mortality. Fortunately, it is less common, accounting for 10% of large vessel occlusions (LVO) and approximately 1% of all ischemic strokes. ¹ BAO results from either in situ atherosclerosis or embolism from a cardiac or proximal arterial source. In this narrative review, we summarize the clinical presentation, imaging work-up, treatments, and clinical outcomes for BAO. Throughout, we highlight important knowledge gaps in treatment selection and technique.

Methods

The following databases were searched: MEDLINE, Scopus, Google Scholar, and Web of Science. The following search terms were used: “basilar artery occlusion”, “basilar artery occlusion and medical therapy”, “basilar artery occlusion and mechanical thrombectomy”, “basilar artery occlusion and endovascular thrombectomy”, “basilar artery occlusion and clinical presentation”, “basilar artery occlusion and imaging”, “basilar artery occlusion and outcomes”. We included randomized trials and observational studies in humans. We also reviewed relevant references from the literature identified. Studies were excluded if they met predetermined exclusion criteria, including preclinical studies or those involving animal models of stroke, and studies of obsolete endovascular devices/techniques (e.g., MERCI).

Clinical presentation

Recognizing BAO is, of course, the first step in clinical management. Risk factors for BAO include hypertension, diabetes mellitus, hyperlipidemia, coronary artery disease, and smoking. Common symptoms include bulbar and motor symptoms with decreased consciousness. Prodromal symptoms, including paresthesias, ataxia, headaches, double vision, and vision loss, may precede disabling symptoms. The onset can be abrupt or stuttering, often related to etiology. Sudden symptoms may suggest an embolic etiology, while stuttering symptoms are often associated with atherosclerosis. Furthermore, branch atheromatous basilar artery disease can also cause occlusion of paramedian perforators. ²

The clinical diagnosis of BAO is often delayed due to multiple factors, including shortcomings of pre-hospital scales, early intrahospital assessments, nonspecific symptoms, sometimes limited exam findings, and presentations that mimic other entities. ³ The Face Arm Speech Test (FAST) score was shown to miss about 40% of posterior circulation strokes as compared to 10% of anterior circulation strokes. ⁴ Although the National Institutes of Health Stroke Scale (NIHSS) scale is widely used in the acute setting, it performs poorly in predicting posterior circulation strokes. ⁵ Examining patients with BAO effectively is challenging as ∼40% are comatose on initial presentation. ⁶ A study that included 85 patients with either BAO or bilateral distal vertebral artery occlusions revealed that the most common symptoms occurring 2 weeks before stroke were nausea, headache, and vertigo. ⁷ In a retrospective study comparing BAO and left middle cerebral artery (MCA) occlusions, only 38% of patients were clinically diagnosed accurately in the BAO group as compared to 90% in the left MCA occlusion group. Furthermore, the BAO group was found to have a longer median time from ED arrival to diagnosis as compared to the left MCA group (8 h 24 min vs. 1 h 23 min; p < 0.001). Incorrect preliminary diagnoses in the BAO group included entities such as anterior circulation stroke, delirium, dementia, ethanol intoxication, conversion disorder, seizure, Bickerstaff encephalitis, Charcot- Marie-Tooth syndrome, electrolyte derangements, Guillain-Barré syndrome, presyncope, cerebral venous sinus thrombosis, sepsis, and intracranial hypotension. ⁸

One posterior circulation stroke registry analysis showed decreased level of consciousness, tetraparesis, and pupillary abnormalities were associated with a 3.5 fold increased risk of poor outcome. ² Embolic occlusions were also noted to be associated with worse outcomes compared to atherosclerotic occlusions. Embolism may be associated with less developed collaterals, and occlusions from embolism are often at the rostral basilar tip where deficits are more severe.^2,9 In another study, predictors of poor outcome after BAO included higher NIHSS score, older age, absence of hyperlipidemia, presence of prodromal minor stroke, longer time to treatment, proximal basilar occlusion location, and prior posterior circulation stroke. ¹⁰ Untreated BAO has a very high mortality rate of up to 90%. Intra-arterial thrombolysis decreases mortality to ∼50%, and the most important prognostic factor for good outcome is early recanalization. ¹¹

While scores and prognostic models can offer guidance, caution should be used when using them to exclude patients from treatment. Indeed, predicting long term outcomes represents an important knowledge gap.

Imaging

Neuroimaging is an essential component of the work-up and treatment of BAO. Technological advances in imaging over the last decade have provided more information for endovascular thrombectomy (EVT) patient selection and prognostication. ¹² Detection of early ischemic changes for posterior circulation strokes is limited with non-contrast computed tomography (NCCT), the first-line modality for most patients with suspected stroke. Hwang et al. revealed that the sensitivity for detecting acute infarction in the posterior fossa was only 42%. ¹³ This is attributed to beam hardening artifacts and the additional time often required for visualization of infarction in white matter. Therefore, older NCCT-based assessments and scores have less clinical relevance today.

Computed tomography perfusion (CTP) has several advantages for EVT patient selection. Using a prespecified Critical Area Perfusion Score, Cereda et al. showed that BAO patients presenting with limited regions of severe hypoperfusion (T max >10 s) had better outcomes following EVT. ¹⁴ Indeed, better functional outcomes have been related to lower perfusion deficit volumes on all CTP parameter maps. ¹⁵ However, there are significant knowledge gaps that limit the use of CTP to select BAO patients for EVT given the lack of validated thresholds for defining penumbra and ischemic core.^15,16 In our opinion, diffusion-weighted MRI is the ideal modality to assess early ischemic changes, especially for potential EVT candidates with NIHSS>10.^13,17 However, it must be acknowledged that timely emergent MRI is unavailable at many centers. Computed tomographic angiography (CTA) can aid decision making, beyond simple LVO identification, through evaluation of the extent of thrombus burden and the quality of posterior circulation collaterals.^18,19 That stated, caution should be exercised when excluding patients based solely on any imaging modality. ²⁰

Numerous imaging-based prognostic scoring systems have been developed for posterior circulation stroke. They are based on parenchymal ischemic changes, location of thrombus, thrombus burden, and collateral status. Scoring systems are more reliable at predicting outcomes when they emphasize factors such as early infarction of brainstem corticospinal tracts and large regions of the cerebellum. Large cerebellar infarctions for example can lead to cerebral edema and fourth ventricle compression, resulting in poor outcomes independent of brainstem status. When compared to structures in the lateral brain stem, involvement of the corticospinal tracts may lead to more severe functional deficits. The posterior communicating artery (PCOM) is the main collateral pathway central to many of the scoring systems. The PCOM is beneficial in cases where thrombus does not fully occlude the basilar tip or extend into the P1 segment, allowing retrograde flow to the rostral basilar artery. The location of the thrombus is also a key point in scoring systems since occlusion of the distal segment of the basilar artery may respond better to thrombolysis when compared to proximal and mid basilar occlusions. However, retrograde thrombosis has also been noted to cause occlusion of the brainstem paramedian perforators in basilar apex syndromes. ¹²

CT/CTA-based scoring systems include the Posterior Circulation Acute Stroke Prognosis Early CT Score (pc-ASPECTS), Basilar Artery on Computed Tomography Angiography (BATMAN), Posterior Circulation CT Angiography (pc-CTA), Posterior Circulation Collateral Score (PC-CS), and pretreatment Collateral Score (CS) (Table 1).^{18,19,21–24} MRI-based scoring systems include the Renard DWI, Brainstem DWI, Bern DWI and pc-ASPECT (Table 2).^25–28 The two most common scoring systems are the pc-ASPECTS and the BATMAN score. The pc-ASPECTS can be applied to CT, DWI MRI, and CTP.^22,25,29 Puetz et al. reported, in their retrospective study of 46 patients, that good 90-day functional outcome (modified Rankin Scale, mRS ≤3) was observed in 52% of patients with pc-ASPECTS ≥8 but only 4% in those with pc-ASPECTS <8. Furthermore, pc-ASPECTS ≥8 was associated with a 60% reduction in mortality.^22,30 Another study of the Basilar Artery International Cooperation Study (BASIC) registry revealed higher mortality among patients with CBV pc-ASPECTS <8. ²⁹ Yet another study showed that BATMAN <7 was associated with a poor outcome among patients treated with EVT. ¹⁹ Sensitivity for detecting infarct location and volume is higher in DWI-pc-ASPECTS when compared to NCCT pc-ASPECTS. A cut off value of 6 or less on DWI-pc-ASPECTS usually indicates poor outcome. ³¹ CTP -ASPECTS may offer a better status of the ischemic region when compared to NCCT-ASPECTS since CTP provides quantitative parameters that may better define the ischemic penumbra and infarct core. ³² Caruso et al. noted that pc-ASPECTS on CTP appeared to be a powerful marker for predicting functional outcome. They applied pc-ASPECTS to NCCT, CTA, and CTP. Mean transit time-defined pc-ASPECTS was the most reliable parameter in one study; it was associated with discharge NIHSS, discharge mRS, and 3-month mRS. ³³

Table 1.

CT and CTA prognostic scoring systems of acute basilar artery occlusion.

Posterior Circulation Acute Stroke Prognosis Early CT Score (pc-ASPECTS)22,30	-Score allots the posterior circulation 10 points. One point is subtracted for hypoattenuation in the left or right of the following regions: thalamus, cerebellum, or PCA territory. Two points are subtracted for hypoattenuation in any part of the midbrain or pons. -A pc-ASPECTS of 10 indicates absence of visible posterior circulation ischemia, and a score of 0 indicates hypoattenuation in all pc-ASPECTS territories.
Basilar Artery on Computed Tomography Angiography (BATMAN) 19	-The vertebrobasilar system is divided into 6 segments, including vertebral arteries (considered as 1 segment), each PCA (considered separately), and the 3 segments of the basilar artery. -For opacification of each PCA, three points are allotted for a fetal origin, two points for a posterior communicating artery > 1 mm, and one point for <1 mm. One point is allotted for opacification of either one or both vertebral arteries and each of the three basilar artery segments giving a maximum score of 10. -The BATMAN score assesses both collateral status and thrombus burden.
Posterior Circulation CT Angiography (pc-CTA) 23	-6-point score evaluating collateral flow, based on antegrade or retrograde contrast opacification of vessels within the occluded territory. -The posterior circulation is divided into 6 segments, each receiving a score of one if occluded: one or both the two intracranial vertebral arteries; the proximal basilar segment; the middle basilar segment; the rostral basilar segment; the right PCA; the left PCA. -Score of 0 indicates patency of the whole arterial tree, whereas a score of 6 indicates there is complete occlusion of the posterior circulation, including at least 1 VA and both PCAs.
Posterior Circulation Collateral Score (PC-CS) 24	-10-point grading system to quantify the collateral flow in the posterior communicating arteries and the cerebellar arteries. -One point is allotted for opacification of each PICA, AICA, SCA, patent PCom with a caliber smaller than the ipsilateral P1; Two points are allotted for each PCom with a caliber equal to or larger than the ipsilateral P1. -A score of 10 reflects that all branches are patent, while a score of zero signifies that no branches opacify. Severity is classified into three groups for the pc-CS: Poor 0-3, intermediate 4–5, and good 6–10.
Pretreatment Collateral Score (CS) 18	-2-point scoring system based only on the posterior communicating artery. Zero points are given if there is no opacification. One point is given if there is unilateral opacification. Two points are given if there is bilateral opacification. -Patients with good collaterals had a CS of 2, while poor collateral status was defined as CS of 0 or 1.

PCA: posterior cerebral artery; PCom: posterior communicating artery; SCA: superior cerebellar artery; AICA: anterior inferior cerebellar arteries; VA: vertebral arteries; BA: basilar artery.

Table 2.

MRI prognostic scoring systems of acute basilar artery occlusion.

Renard DWI Score 26	-10-point semiquantitative score. -One point is allotted for unilateral (two for bilateral) involvement of each brain stem level (medulla, pons, and mesencephalon) or cerebellar infarction. Half of a point is allotted for unilateral (one for bilateral) infarction of the thalamus or temporo-occipital lobe. -A total score of more than 3 was considered elevated.
Brainstem DWI (BS DWI) 27	-The BS DWI is a semiquantitative 22-point scale. -The number of arterial territories with abnormal DWI is assessed at each brain stem level: 0–8 in the medulla, 0–6 in the pons, and 0–8 in the midbrain. -Cerebellar DWI lesions are dichotomized as minor (less than one-third of a cerebellar hemisphere) or major (more than one-third of a cerebellar hemisphere).
Bern DWI score 28	-Lesions in the brainstem are assigned 1 and 2 points (uni- and bilateral). If there is pyramidal tract involvement, scores are doubled due to the greater impact on the clinical status. -Cerebellar lesions are included because cerebellar infarction is commonly associated with brainstem compression and hydrocephalus. One point is allotted for infarcts involving less than one-third of a hemisphere, and two points for the involvement of more than one-third. -Involvement of the thalamus and/or the temporo-occipital cortex is allotted one point for each side.

MRI: Magnetic Resonance Imaging; DWI: Diffusion Weighted Imaging.

However, all of these scores may have limited generalizability since they had small sample sizes, retrospective designs, and different definitions for outcomes.

Endovascular treatment

Patients with BAO who have severe strokes clinically but not large established infarcts on imaging may be ideal candidates for EVT. The benefits of EVT over medical therapy for anterior circulation LVO are well established with robust evidence. ³⁴ This is not the case for posterior circulation LVO, and there are several knowledge gaps particularly in patient selection. BASICS, a multicenter open-label international randomized controlled trial, compared EVT and standard medical therapy for patients with BAO. Three hundred patients were randomly assigned within 6 h of symptom onset in a 1:1 ratio. Most patients (∼80% in both groups) received intravenous thrombolysis. Comparing groups, there was no significant difference in the proportion that achieved a good outcome, defined as a score of mRS ≤3. There was also no difference in symptomatic intracranial hemorrhage (sICH). However, there were several limitations of this trial. Twenty-nine percent (124 of 424) of eligible patients were not enrolled, and 79% of those (98 of 124) ultimately received EVT, suggesting the possibility of selection bias. Outcomes for patients not enrolled were not presented. Furthermore, the trial was underpowered given patient recruitment was lower than anticipated. ³⁵

Similar results were presented in the Endovascular Treatment versus Standard Medical Treatment for Vertebrobasilar artery occlusion (BEST) randomized controlled trial. One hundred thirty-one patients were enrolled within 8 h of symptom onset at 28 centers in China. Patients were randomly assigned (1:1) to EVT plus standard medical therapy or standard medical therapy alone. High crossover and poor recruitment lead to the trial being terminated early. In the intention-to-treat analysis, there was no difference in the proportion with mRS ≤3 at 90 days (42% in the EVT group vs. 32% in the control group; adjusted odds ratio 1.74; 95% CI 0.81, 3.74). However, there was a significant improvement in those treated with EVT in the per-protocol and as-treated analyses. Limitations of this study include a small sample size, poor adherence to the assigned study arm, and a progressive drop in recruitment. A third of patients who qualified for enrollment declined trial participation. Similar to BASICS, this may have introduced selection bias. ³⁶ A systematic review and meta-analysis by Katsanos et al. identified five studies (three additional observational studies) comparing EVT and medical management; 1098 patients were included. There was no significant difference in functional outcome nor all-cause mortality at 3 months. Furthermore, there was a higher risk of sICH in patients who underwent EVT when compared to those who received medical management (RR 5.42; 95% CI 2.74, 10.71). ³⁷

Two recently-presented unpublished BAO randomized trials in China (BAOCHE: Basilar Artery Occlusion CHinese Endovascular Trial and ATTENTION: Endovascular Treatment for Acute Basilar Artery Occlusion) have shed new light on EVT for BAO. These are the first trials to show clear efficacy of EVT when compared to best medical management. In the BAOCHE trial, 217 patients with confirmed BAO on vascular imaging who presented within 6-24 hours from last seen well were enrolled. 73% of patients were male, and the overall mean age was 64 years. The median NIHSS score was 20. Only patients with small to moderate infarct burden assessed by non-contrast CT were enrolled. Pre-specified efficacy boundaries were crossed in favor of EVT in a pre-planned interim analysis. 46% of patients in the EVT arm achieved an mRS 0-3 at 90 days as compared to 24% in the control group (adjusted OR 2.92 , p = 0.001). The mortality rate in the EVT arm was 31% as compared to 42% in the control group. 5.9% of patients in the EVT arm had symptomatic hemorrhage when compared to 1.1% in the control group (p-value 0.13). The number needed to treat was 4.5. These results indicate that patients with BAO presenting in the late window stand to benefit from EVT, similar to those with anterior circulation occlusions as shown in DAWN.^38–40 Furthermore, the ATTENTION trial enrolled 340 patients with BAO confirmed on vascular imaging within 12 hours from stroke onset (226 were treated with EVT and 114 with best medical management). The mean age was 67 years. All patients had an NIHSS ≥10, and the median was 24. A 90-day mRS of 0-3 was noted in 46% in the EVT arm as compared to 23% in the best medical management arm (adjusted risk ratio 2.1, p <0.001). The number needed to treat was 4. The 90-day mortality rate with EVT was also lower when compared to the control arm.^41,42 While promising, perhaps the greatest limitation of both trials is limited generalizability since both included only Asians.

There have been several other non-randomized studies. In a one-year follow-up study of patients enrolled in the original Acute Basilar Artery Occlusion Study (BASILAR), the beneficial effect of EVT was similar to that observed at 90 days, and the mortality rate was significantly improved with EVT (55% vs. 84%). ⁴³ Mbroh et al. conducted a meta-analysis to study the risk factors, safety, efficacy, and outcomes after EVT between posterior circulation and anterior circulation LVO. Posterior (1172 patients) and anterior (7726 patients) LVO from 16 studies were included. Analyses revealed a higher NIHSS in the posterior group. The likelihood of obtaining successful reperfusion and favorable 90-day outcome were comparable between the two groups. Posterior LVOs were associated with less sICH (OR 0.56; 95% CI 0.37, 0.85), possibly related to less intravenous thrombolysis, but higher mortality (OR 1.92; 95% CI 1.46, 2.53), possibly related to natural history, higher NIHSS, and/or longer onset-to-puncture time. ⁴⁴ Wu et al. examined the relationship of stroke etiology and 90-day outcomes among 167 patients with BAO treated with EVT. 47% had intracranial atherosclerotic disease (ICAD), and 53% had embolic occlusions. There were no differences in successful reperfusion (achieved in 89%) or 90-day functional outcomes. However, among the younger subgroup (<60 years), embolic BAO was associated with greater functional independence. ⁴⁵ A meta-analysis by Kai Xun et al. assessed prognostic factors in patients with posterior circulation strokes after EVT. Worse 90-day mRS was associated with diabetes mellitus, hypertension, and a history of prior stroke. ⁴⁶ Interestingly, Meinel et al. showed that futile recanalization was more likely to occur with BAO compared to anterior LVO (aOR = 2.146; CI 1.267, 3.633). Older age, higher stroke severity, intracranial stenosis, coronary artery disease, higher number of passes, and intracranial stenting were found to be predictors of futile recanalization. ¹

Despite the mixed data, we believe that for patients with clinically severe stroke symptoms and small imaging-defined infarcts, reperfusion offers the best chance for a good outcome. The challenges facing trials for BAO include slow recruitment rates, small sample sizes, and ethical dilemmas when randomizing patients to control arms. Nonetheless, there is sufficient equipoise to reconsider selection criteria for BAO after the recent publication of BASICS, BEST, and soon to be published BAOCHE and ATTENTION trials. In our opinion, advanced imaging, preferably MRI when available, can be used to assess early ischemic changes more reliably and in a timely fashion with appropriate protocols. It is reasonable to exclude patients with large, established brain stem infarcts unlikely to benefit from EVT. Future trials could also consider the exclusion of patients with mild symptoms. As above, data suggest that there may be less benefit for those with NIHSS <10. ³⁵ This is especially true if minor symptoms are related to perforator occlusion from branch atheromatous disease. Furthermore, it is important to interpret clinical scales with the posterior circulation in mind. The posterior NIHSS (POST-NIHSS) was developed to identify patients who would benefit from EVT. Extra points for gait/truncal ataxia (3 points), dysphagia (4 points), and abnormal cough (5 points) were added. Compared to the traditional NIHSS, POST-NIHSS has a higher prognostic accuracy. ⁴⁷ Similarly, the mRS may be less well suited for outcomes after BAO. Since the brain stem is a compact structure, a small infarct can lead to substantial disability. ⁴⁸ There is growing support for the utility-weighted mRS or other outcomes more posterior circulation specific and patient centered. ⁴⁹ In our opinion, a comprehensive approach to patient selection should be undertaken to maximize benefits, minimize risks, and optimize resources. Our approach includes clinical presentation, mechanism of BAO, location of occlusion, and neuroimaging findings (Table 3).

Table 3.

Recommended endovascular patient selection characteristics and technical considerations.

Clinical	-Patients with age <75 years may benefit more from EVT.1,10 -Patients with exam findings including decreased level of consciousness, tetraplegia/tetraparesis, and locked-in syndrome may benefit more from EVT. 2 -Patients with severe stroke syndromes with NIHSS >10 may benefit more from EVT.17,35
Mechanism	-Patients with BAO secondary to embolism may be more readily reperfused than those with BAO secondary to ICAD.45,77 -Patients with perforator occlusion from branch atheromatous disease may benefit less from EVT.
Neuroimaging	-Patients with BAO may benefit more from EVT than those with isolated posterior cerebral or vertebral artery occlusions. 15 -Patients with distal basilar tip occlusions may benefit more than proximal occlusions.2,9 -Patients with lower infarct burdens and better collaterals, such as pc-ASPECTS ≥8, and BATMAN score ≥ 8 may benefit more from EVT.19,22
Technical considerations	-First-line aspiration may offer improved reperfusion and safety over first-line stent retriever EVT.50,56 -Aspiration may provide a more effective and safer option for embolic BAO. 55 -While reasonable to attempt aspiration or stent retriever EVT first, direct angioplasty may also be considered for ICAD-related BAO.59,60 -Angioplasty for residual stenosis thought to be related to ICAD after EVT may be considered.52,53,60 -For posterior circulation tandem lesions, there may be a benefit to navigating through the contralateral vertebral artery when the vertebral arteries are codominant.66,67

EVT: Endovascular therapy; NIHSS: National Institutes of Health Stroke Scale; ICAD: Intra-Cranial Atherosclerotic Disease; BAO: Basilar Artery Occlusion; BATMAN: Basilar Artery on Computed Tomography Angiography; pc-ASPECTS: Posterior Circulation Acute Stroke Prognosis Early CT Score.

Endovascular technical considerations

After determining whether the benefits of treatment outweigh the risks, there are several technical considerations for endovascular approaches to BAO. These considerations relate to differences in posterior circulation anatomy and multiple possible BAO etiologies, including cardioembolism, artery-to-artery embolism, and basilar artery ICAD. ² Not surprisingly, these differences may underlie several current knowledge gaps.

One consideration is whether the first-line approach should be with aspiration alone, stent retriever alone, or a combination. While there is clinical equipoise in many cases, we believe that an aspiration first approach is reasonable. Gengfan Ye et al. conducted a meta-analysis of two prospective and three retrospective studies comparing the efficacy and safety of first-line aspiration (n = 193) and first-line stent retriever (n = 283) EVT for BAO. Successful TICI 2b-3 reperfusion was significantly more likely in the aspiration group (OR = 2.0; 95% CI 1.1, 3.5). Furthermore, there were fewer new territory embolic events (OR = 0.2; 95% CI 0.05, 0.83) and procedure time was shorter (mean difference −27.6 min, 95% CI −51.0, −4.3) with this approach. There were no significant differences in complete TICI 3 reperfusion, need for rescue therapy, hemorrhagic complications, 90-day favorable outcomes, or 90-day mortality between the two groups. ⁵⁰ Interestingly, another study supported that a combined stent retriever and aspiration approach has a similar safety profile, but higher rates of revascularization when compared to stent retriever or aspiration alone. ⁵¹

We believe that the endovascular approach should be tailored to the presumed etiology when possible. While it can be difficult to determine the etiology acutely, there are clinical and imaging features that can help guide decision making. Proximal/mid basilar occlusion with tip patency, absence of convex edge at occlusion, longer occlusion length, and vessel wall calcification on CTA can suggest ICAD.^52,53 Absence of thalamic infarctions on MRI can also suggest ICAD, likely related to sparing the basilar tip and P1 segments supplying the thalamoperforators.^53,54 Aspiration may provide a more effective and safer option for patients with embolic BAO. ⁵⁵ These thrombi tend to have higher erythrocyte and lower fibrin content.^56,57 In contrast, ICAD BAO poses unique challenges since EVT itself can cause intimal injury and further plaque rupture, likely accounting for higher rates of re-occlusion. ⁵⁸ The ideal treatment approach represents a substantial knowledge gap. In our opinion, it is reasonable to perform a staged approach with aspiration, followed by stent retriever, then angioplasty if necessary. However, others advocate for a direct angioplasty approach. ⁵⁹ Ma et al. compared the safety and efficacy between direct angioplasty and stent retriever first EVT for ICAD BAO among 108 patients. They showed that direct angioplasty shortened procedure time and reduced the number of passes. However, both approaches had similar safety and efficacy, with no differences in reperfusion, emergent stent placement, embolization to distal territories, re-occlusion, or outcomes. ⁶⁰

Anatomical considerations are also important to optimize technique. The vertebral arteries originate at nearly right angles from the subclavian arteries. For comparison, the internal carotid arteries are more of a direct extension of their parent common carotid arteries. Furthermore, the diameter of vertebral arteries usually measures ∼3 mm. They have a relatively smaller caliber when compared to internal carotid arteries (4–5 mm) and common carotid arteries (6–7 mm).^61,62 For EVT, large bore aspiration systems afford benefits compared to smaller systems, but the reduced diameter of vertebral arteries may pose a limit. Many prefer an 8 Fr guide catheter, but smaller guides of 6 Fr in the distal vertebral artery can still prove adequate to support advancing a 5 Fr aspiration catheter into the basilar artery. Furthermore, some advocate for the use of a 0.014-inch buddy wire or a large caliber coronary guide catheter placed in the subclavian artery for support if the subclavian artery is torturous. ⁶³

Although transfemoral access is more often used as an upfront approach for EVT, transradial access is increasingly being used as an alternative at many centers. This approach may offer particular advantages in BAO since catheters are naturally navigated through the subclavian allowing access into the vertebral artery without maneuvering through the aortic arch. Factors favoring transradial over transfemoral access include great vessel tortuosity, arch developmental variants, and marked calcification of the abdominal aorta and descending thoracic aorta. One study that included 9 patients with BAO showed that radial access may be a viable option and technically feasible. The average puncture-to-reperfusion time was 35.8 min. Furthermore, 89% of patients achieved technical success which was defined as TICI score of 2B-3. The one unsuccessful case had severe proximal stenosis of the basilar artery and a microcatheter could not be advanced through the lesion. ⁶⁴ A single center retrospective study comparing transfemoral versus transradial access in posterior circulation EVT revealed similar results; 6 patients with BAO and 4 patients with P1-P2 segment occlusion were included in the transradial group. Shorter puncture-to-reperfusion time was noted in the transradial group when compared to the transfemoral group (29.2 ± 17.6 min vs. 63.9 ± 56.7). ⁶⁵ These studies had several limitations including small sample sizes and potential bias given their retrospective nature. Although transradial access may be a good alternative, it is important to maintain a working knowledge of transradial techniques, technical expertize in this approach, and stock the preferred equipment to avoid delays in EVT.

Another consideration regrading approach is which vertebral artery to navigate en route to EVT. It stands to reason that the vertebral artery with the largest diameter, least tortuosity, and least stenosis is ideal. There can be unique challenges with tandem lesions, involving occlusion of the basilar artery and occlusion or severe stenosis of one or both vertebral arteries. Piechowiak et al. compared BAO with tandem lesions (n = 15), isolated ICAD BAO (n = 14), and isolated intracranial embolic BAO (n = 23). Access was achieved via the dominant vertebral or, in the case of tandem lesions, via the contralateral vertebral when possible. Successful reperfusion was achieved in 15/15 patients with tandem lesions and 35/37 patients with isolated BAO. Ninety-day mRS ≤2 was achieved in 8/15 with tandem lesions compared to 4/14 with isolated ICAD BAO and 5/23 with isolated embolic BAO (p = 0.18). ⁶⁶ Baik et al. showed that up to 25% (55/224) of BAO patients have tandem lesions. In 60% (33/55), the ipsilateral vertebral artery with the lesion was utilized for intracranial access. The contralateral was more frequently selected when the vertebral arteries were codominant (41% vs. 15%; p = 0.032). In their cohort, 16 patients underwent angioplasty or stenting of the proximal lesion, which was more often performed when the access route was the dominant lesioned vertebral artery (88% vs. 41%; p = 0.002). Since there was short-term recurrent BAO for two patients in which angioplasty of the vertebral lesion was not performed, the authors suggest that proximal angioplasty may have benefits. There were no significant differences in procedure time, successful reperfusion, or clinical outcome in comparisons of the vertebral artery access route or angioplasty of the tandem vertebral lesion; 87% (48/55) achieved successful reperfusion and 31% (17/55) achieved mRS ≤2 at 90 days. ⁶⁷

Intracranial stenting, both acutely at the time of EVT for persistent stenosis secondary to ICAD and subacutely for severe stenosis refractory to medical management, has gained growing acceptance. Wang et al. described 97 patients with severe symptomatic intracranial vertebrobasilar stenosis (70–99%): 30 patients had BA stenosis, 55 patients had V4 stenosis, and 12 patients with V4-BA stenosis. In their cohort, 96 patients underwent successful stent deployment. Two deaths, two intracranial hemorrhages, and 7 ischemic events were noted. To differentiate lesion morphology, they used the Mori classification of intracranial artery stenosis: type A (<5 mm in length, concentric or moderately eccentric lesions not totally occlusive), type B (tubular, 5–10 mm in length, extremely eccentric or totally occluded lesions), and type C (10–15 mm in length, extremely angulated lesions with excessive tortuosity of the proximal segment, or totally occluded lesions). Self-expanding stents were more frequent for Mori C lesions, while Apollo stents were used more frequently for Mori A lesions. A higher rate of complications were noted with Mori C lesions compared to Mori A (p = 0.008) and Mori B (p = 0.047).^68,69 Another multi-center prospective registry by Miao et al. assessed the safety and efficacy of stenting within 30 days for patients with severe symptomatic stenosis (defined as 70–99%). Among the 159 patients treated with balloon mounted stents, 48 patients had basilar artery stenosis and 48 patients had intracranial vertebral artery stenosis. Among the 141 treated with balloon angioplasty followed by self-expanding stents, 46 patients had basilar artery stenosis and 25 patients with intracranial vertebral artery stenosis. The overall 30-day rate of stroke, transient ischemic attack, and death was noted to be 4.3%. Successful revascularization was achieved in 97.3% of patients. ⁷⁰ Acute stenting has also been shown to be a feasible option for patients who fail BAO thrombus retrieval. A study of 93 patients performed in China evaluated the safety and efficacy of rescue stenting following EVT failure in patients with BAO. The stroke etiologies were 76.2% atherosclerotic, 7.1% cardioembolic, and 16.7% other/unknown. 75 (92.6%) patients in the rescue stenting group achieved TICI 2b/3. Furthermore, rescue stenting was associated with improved 90-day mRS 0-3 compared to patients who did not undergo rescue stenting (51.9% vs. 16.7%; p = 0.023) without an increase in sICH. Rescue stenting was also associated with a lower mortality rate (18.5% vs. 58.3%; p = 0.006). Limitations of this study include its observational design, small number of patients in the non-stented group, enrollment of only patients with NIHSS ≥10, and regional differences in ICAD prevalence since it was performed in China. ⁷¹

While there is mixed data, the anesthesia strategy used in BAO treatment may affect outcomes. A retrospective analysis by Du et al. suggested that general anesthesia (GA) may result in fewer patients achieving mRS 0-3 at 90 days compared to conscious sedation (CS). One hundred thirty-nine patients (78.1%) underwent GA and 39 patients (21.9%) underwent CS. GA was associated with less mRS 0-2 (OR 0.28; 95% CI 0.13–0.59) and less mRS 0-3 (OR, 0.23; 95% CI, 0.10–0.52) when compared to CS in univariate analyses. ⁷² Similar results were noted in another study of 298 patients from the CICAT (Codi Ictus Catalunya) registry with posterior circulation strokes undergoing EVT. Patients who underwent GA were less likely to have a good outcome (19.7% vs. 45.1%, p < 0.001) at 3 months. ⁷³ However, there may be selection bias since patients more severe stroke syndromes may be more likely to undergo GA. Li et al. examined at the anesthesia strategy of 639 patients undergoing EVT in the BASILAR registry. Two hundred fifty-seven patients (40.2%) were in the GA group, 250 patients (39.1%) in the local anesthesia (LA) group, and 132 patients (20.7%) in the CS group. They found that there were no differences in mortality, hemorrhagic transformation, or 90-day outcomes between the three groups. Of note, 6.4% of patients were converted from LA/CS to GA. ⁷⁴ There remains no consensus regarding the ideal anesthesia strategy in EVT for BAO. However, extrapolating from anterior LVO data, it stands to reason that whatever approach yields the fastest time to reperfusion may be reasonable. If GA is pursued, attention should be paid to minimize reductions in blood pressure.^75,76

Conclusion

Despite revolutionary advances in stroke care, the morbidity and mortality of BAO remains high. Recent randomized trials have yielded mixed results and questions remain. Prospective studies using standardized clinical assessments and imaging protocols are needed to improve our understanding of the efficacy, safety, selection criteria, and optimal technical approaches for EVT among patients with BAO. At present, a reasonable approach is to select patients with more severe stroke symptoms who have lower infarct burdens. Posterior circulation-specific clinical measures of stroke severity and outcomes may bring some clarity, and hyperacute-phase advanced imaging may offer several benefits to aid decision making. While several knowledge gaps remain, most practitioners in our field agree that a reperfused basilar artery is better than an occluded one for patients with viable brain to save.