Evidence-based medicine and research regarding posterior circulation ischemia (PCI) has consistently lagged behind anterior circulation ischemia (ACI) due to presumed diminished prevalence of the disease and surgical inaccessibility. However, posterior circulation infarcts account for roughly one-quarter of all ischemic strokes,1 and they remain a significant cause of patient morbidity and mortality. In this article, we review pertinent vertebrobasilar anatomy, along with background research of posterior circulation atherosclerotic pathology. Further, we illustrate case examples of vertebrobasilar disease, highlighting endovascular treatment options.
The clinical diagnosis of vertebrobasilar insufficiency (VBI) and PCI is generally more complex than that of ACI, especially for a non-neurologist. Whereas aphasia and weakness associated with ACI is typically obvious, VBI and PCI present with symptomatology that can be somewhat more subtle. Typical findings are vague and may be confused with inner ear disease. Based on the New England Medical Center-Posterior Circulation Registry (NEMC-PCR),2 the most common symptoms include dizziness and syncope (47%), unilateral limb weakness (41%), dysarthria (31%), headache (28%), and nausea or vomiting (27%). The most frequent signs are unilateral limb weakness (38%), gait ataxia (31%), unilateral limb ataxia (30%), dysarthria (28%), and nystagmus (24%). Fewer than 1% of patients with VBI presented with only one complaint, signifying pattern recognition in diagnosis of such patients. In general, a combination of cranial nerve palsies and long tract signs localizes the lesion to the brain stem and should, therefore, trigger a posterior circulation vascular evaluation.
For many years, patients with suspected ACI have been interrogated and treated with a vastly different approach than patients with suspected PCI. Although researchers have questioned this double standard in the 1980s and 1990s, it was not until 2004 when Caplan et al drastically changed this paradigm by demonstrating that stroke mechanisms responsible for PCI and ACI were more alike than previously thought.3 In their study of 407 patients from the NEMC-PCR, data collected from 1988 to 1996 demonstrated that cardiogenic emboli (24% vs. 38%) were less common than large artery occlusive disease (32% vs. 9%) in PCI compared to ACI, respectively. The difference in the cardiogenic embolism rate was explained by the fact that the posterior circulation receives approximately one-fifth of the total brain circulation, leading to fewer emboli based on this hemodynamic phenomenon. Thus, by factoring in this inherent hemodynamic disparity, the researchers concluded that the pathophysiology of PCI is indeed very similar to ACI. This landmark trial signified that just like ACI, PCI should also be evaluated thoroughly to select an appropriate therapy based on the causative stroke mechanism.
With respect to vascular imaging, noninvasive angiography was not widely available until the 1990s. Prior to that, catheter angiography of the vertebrobasilar system was performed infrequently due to the inherent risk associated with this procedure, specifically the infarction risk of vital brainstem structures. Therefore, while the anterior circulation was investigated frequently and thoroughly, angiography of the vertebrobasilar circulation was only performed in severe clinical cases, contributing to more uncertainty regarding PCI.4
It was generally believed that PCI had low recurrence and mortality rates, making primary and secondary prevention ineffective at a population level. The small recurrence rate was based upon multiple small scaled studies performed in the 1960s and 1970s, which demonstrated recurrence rates of 2%-6%.5 It was not until 1998 that the Warfarin-Aspirin Symptomatic Intracranial Disease (WASID) trial demonstrated a PCI stroke recurrence rate of 22% (15% per year) during a 14-month mean follow-up of 59 symptomatic patients with more than 50% stenosis in the intracranial posterior circulation.6 In 2003, Qureshi et al retrospectively studied 102 patients diagnosed with symptomatic vertebrobasilar stenosis.7 Over a mean follow-up of 15 months, 14% (11% per year) had a recurrent stroke with a total mortality of 21%. In the same year, a meta-analysis of 36 cohorts from 46 articles demonstrated that the recurrence event rate of PCI was not lower than the recurrence rate of ACI; interestingly, it raised the possibility that this rate might be even higher in the first month following a posterior circulation event.5 A prospective study in 2009 supported this theory by showing a 30% recurrence rate for PCI.8 These recurrence rates were significantly higher than previously thought, igniting new interest in this topic.
Multiple treatment options have been available for ACI with the mainstay of treatment including carotid endarterectomy (CEA) and anticoagulation for noncardioembolic and cardioembolic strokes, respectively. The single treatment option for PCI was traditionally anticoagulation, since surgical intervention is technically challenging due to difficult access to the posterior circulation. In contrast, the cervical carotid arteries are surgically accessible with the first reported case of CEA published in 1954.9 Anticoagulation therapy for PCI was primarily based on a few uncontrolled retrospective studies performed in the 1950s and 1960s.10,11 While multiple randomized and controlled trials in early the 1990s showed the effectiveness of anticoagulation therapy for recurrent PCI of cardioembolic etiology, there was no clear advantage of warfarin therapy for PCI due to noncardioembolic causes, which led to further clumping of these various stroke mechanisms into one group.12-15 This practice was further validated by a retrospective study in 1995 that demonstrated superiority of warfarin over aspirin, even in PCI due to intracranial stenosis.16 In 2001, however, the Warfarin–Aspirin Recurrent Stroke Study (WARSS), a randomized, double-blind, multicenter clinical trial, showed equal efficacy of these 2 treatments for noncardioembolic PCI.17 In 2005, the WASID investigation also indicated equal efficacy of warfarin vs. aspirin for preventing strokes due to intracranial atherosclerosis.18 Since both ACI and PCI were included in this trial, in 2006, subgroup analysis of the PCI patients was published indicating no significant difference in recurrent stroke rates between the two therapies.19 The trial was terminated prematurely due significantly higher rates of adverse effects in the warfarin-treated arm than the antiplatelet-treated arm. Therefore, empiric anticoagulation therapy not only failed to demonstrate increased efficacy in the setting of noncardiogenic phenomenon, but it also was shown to carry a worse adverse effects profile.
With growing research and acceptance of coronary artery stenting for treating coronary artery disease, coupled with the advancing knowledge and techniques of carotid artery stenting (CAS), there was a natural outgrowth of potential endovascular treatment options for PCI in the setting of vertebral and basilar artery atherosclerotic disease. While numerous randomized trials have shown effectiveness of CEA and CAS, no same scaled trial for PCI has been conducted. In 2009, it was shown that there is a higher percentage of 50% or more stenosis in PCI versus ACI patients (26% vs. 11.5), which in concert with a high recurrence rate in PCI, demanded endovascular consideration.20 However, in 2007 the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS) subgroup analysis failed to show any advantage of endovascular therapy over medical management for proximal vertebral artery stenosis.21 The study was nevertheless underpowered with only 8 patients in each arm. In 2012, a meta-analysis revealed very low complication rates for endovascular treatment of proximal vertebral artery stenosis (stroke and death rate of 1.1%).22 A restenosis rate of 23% with almost half requiring repeat interventions was also reported. In 2013, it was shown that intracranial vertebrobasilar (VB) stenosis carried a much higher 90-day recurrent event rate than extracranial VB stenosis (33% vs. 16%).23 Therefore, endovascular treatment of intracranial VB stenosis needed to be evaluated. The SAMMPRIS trial, however, had already been terminated due to unexpectedly very high rates of strokes in the endovascular treatment arm of the study compared to the medical arm.24 At that point, the FDA restricted the use of the particular stent to patients with 70%-99% stenosis and 2 or more strokes despite maximized medical therapy. In asymptomatic patients, studies have shown no significant difference in stroke rates with or without vertebral artery stenosis.25
The vertebral arteries (VAs) are the first superoposterior branch of the subclavian arteries and are divided into 4 anatomic segments. These paired arteries course medially and posteriorly between the longus colli and anterior scalene muscles (V1, preforaminal or extraosseous segment) and travel superiorly through the foramina transversaria, most commonly from C6 to C2 (V2, foraminal segment). The VAs exit the transverse foramina of the axis and course around the lateral masses of the atlas (V3, extraspinal segment), and finally pierce the dura at the atlantoccipital space and ascend superiorly through the foramen magnum (V4, intradural or intracranial segment). At the pontomedullary junction, the two VAs unite to form a nonpaired basilar artery, which courses cephalad and divides into the posterior cerebral arteries (PCAs) at the pontomesencephalic junction or, more accurately, in the interpeduncular cistern.26
The major branches of the vertebrobasilar system can be divided into perforant and circumferential arteries. The posterior inferior cerebellar arteries (PICAs), anterior inferior cerebellar arteries (AICAs) and superior cerebellar arteries (SCAs) are the lateral circumferential arteries. The perforant branches directly supply the brainstem and are typically not seen on imaging due to their small caliber. Medial circumferential arteries fill in the gap between the 2 mentioned territories.
The VAs typically measure 3-5 mm; however, there is significant variability in terms of size.27 The left VA is dominant in half of the population. Right dominant and codominant VAs each compose 25% of the population. It has been reported that 15% of individuals have an atretic (< 2 mm) VA. The VAs are typically the first superior branch of the subclavian arteries. Occasionally, they may be the second branch, if the thyrocervical trunk branches more proximally as the first superior branch. On the right side, the thyroidea ima artery may occasionally arise as the first superior branch of the subclavian artery. In 6% of cases, the left VA may originate directly from the aortic arch, between the left common carotid artery (CCA) and left subclavian artery origins. This particular variation represents the second most common aortic arch anomaly, following the common origin of the innominate and left common carotid arteries, commonly but incorrectly referred to as a bovine arch.28 The VAs may enter the foramina transversaria at levels other than C6 in 10% of cases. Occasionally, only one VA is the sole contributor of the basilar artery with the other VA terminating as PICA. Duplications and other rare aberrant origins have also been reported in the literature.29
The initial work-up for VBI and PCI typically starts with a nonenhanced CT (NECT), which is an excellent modality for detecting various intracranial pathologies. However, ischemic stroke remains a clinical diagnosis, and NECT should not be used to diagnose such an entity due to a low sensitivity of 31% and 81% at 3 and 5 hours, respectively. The sensitivities are even lower in PCI due to inherent artifact associated with the posterior fossa, primarily due to streak artifact from adjacent osseous structures. The primary goal of NECT is to exclude intracranial hemorrhage and other stroke mimics. Confirmation of ischemic stroke on NECT is a secondary goal, and a normal examination does not exclude ischemia. MRI is superior in detecting ischemia. Diffusion-weighted imaging (DWI) has sensitivity of at least 90% and could aid in diagnosis.
Ultrasound examination is of limited utility in the evaluating the posterior circulation, predominantly due to lack of an acoustic window. Although the most common location of atherosclerosis of VAs is at their origin, only the V2 segments of the VAs are routinely evaluated. Direction of blood flow is easily determined with color Doppler, which is crucial in diagnosis of subclavian steal syndrome. In this syndrome, there is critical stenosis of the subvclavian artery proximal to origin of the VA with reversal of blood flow in the ipsilateral VA perfusing the upper extremity vasculature. In short, the vertebral artery steals blood from the cerebral circulation to supply the upper extremity. Spectral Doppler waveforms only show changes in severe cases of stenosis. A normal VA waveform is a low-resistant arterial waveform. Spectral broadening, high-resistant waveform and elevated peak systolic velocities may indicate stenosis; however, these findings are not sensitive enough, and no criteria resembling carotid stenosis have been established for staging VA stenosis.
CT angiography (CTA) and MR angiography (MRA) are the 2 most common imaging techniques for evaluating cerebral vasculature. In comparison to CTA, MRA has no ionizing radiation and may be performed without intravenous contrast by using time-of-flight (TOF) sequences. This is especially important in patients with chronic kidney disease where intravenous contrast is problematic. TOF sequences provide excellent images of the circle of Willis. Imaging the cervical carotid and vertebral arteries with this technique is somewhat limited due to a large field of view, and the V1 segments of VAs are typically not seen. Complete evaluation of the vasculature may be accomplished with CTA or contrast-enhanced MRA. CTA is quick and widely accessible. Ionizing radiation and iodinated contrast are the downsides of this modality. Although CTA and MRA techniques have vastly improved from the early days, artifacts are associated with both of these modalities. Therefore, catheter angiography remains the gold standard for imaging the cerebral vasculature. Figure 1 demonstrates all 4 segments of the vertebral arteries and the basilar artery by CTA; postprocessed image demonstrates the underlying bony anatomy of the head and neck.
The goal of each imaging modality is to clearly demonstrate each segment of the posterior circulation, separating normal from diseased sections. The clinical use of ultrasound, CTA, and MRA, as well as cerebral angiography are further illustrated in the following cases.
A 67-year-old man with past medical history (PMH) significant for hypertension and chronic kidney disease presented with one month of intermittent vertigo, dizziness, light-headedness, blurry vision, double vision, headache and unsteadiness. On admission, noncontrast head CT and brain MRI demonstrated subacute infarcts in the bilateral cerebellar hemispheres, a subacute to chronic infarct in the left frontal lobe, and left temporal lobe encephalomalacia (Figure 2). MRA was significant for occlusion of the proximal segment of the right vertebral artery and critical stenosis of the left vertebral artery at the ostium (Figures 3A, 4A, 4B). Digital subtraction angiogram (DSA) performed on hospital day 6 demonstrated critical stenosis of the left vertebral artery ostium (Figure 4C). The right vertebral artery was occluded with poor collateral flow to the V4 segment (Figure 3B). The right vertebral artery supplied some flow to the right PICA territory but no flow was seen distally.
The patient was started on anticoagulation therapy with aspirin and clopidogrel. Two days later, on hospital day 8, the left vertebral artery was treated with angioplasty and stented using a 4-x-18-mm bare-metal stent (VISION stent; Abbott Laboratories, Abbott Park, Illinois). Postintervention angiogram demonstrated no residual stenosis and no significant stent encroachment within the left subclavian artery (Figure 4D). There was excellent runoff to the basilar artery with new retrograde flow within V4 and V3 segments. The postprocedure hospital course was otherwise uneventful and the patient was discharged on aspirin and clopidogrel. He continued to do well postprocedure without posterior circulation insufficiency symptomatology.
A 57-year-old man with PMH of stroke, hypertension, and dyslipidemia on chronic aspirin therapy presented with 2 days of intermittent difficulty walking, dizziness and right-sided weakness. Neurological examination revealed a fluctuating mental status, right-sided cranial nerve weaknesses, decreased right-sided extremity strength, and an upward Babinski reflex on the right.
Initial head CT and brain MRI demonstrated several areas of subacute infarction in the right cerebellum, bilateral pons, left thalamus, and portions of the right occipital lobe (Figures 5, 6). The right vertebral artery was densely calcified on CT and demonstrated abnormal signal on MRI, concerning for occlusion MRA demonstrated occlusion of the right vertebral artery and critical stenosis of the V4 segment of the left vertebral artery (Figure 7A). There was stenosis of the basilar artery, nonvisualization of the PCAs and multiple areas of stenosis in the intracranial ICAs and MCAs. Diagnostic DSA confirmed the areas of stenosis seen on MRA but did show some flow in the right vertebral artery (Figures 7B, C). The appearance of the basilar artery on DSA raised concern for thrombosis of the stenotic vessel and treatment with aspirin and clopidigrel was initiated. MRA and CTA performed after 13 days of anticoagulation demonstrated improved flow to the right vertebral artery and basilar artery (Figure 8). There was persistent severe stenosis of the distal intracranial portions of both vertebral arteries and mild stenosis of the mid basilar artery. The next day, stent-supported angioplasty of the mid right vertebral artery stenosis was performed using a 4.5-x-20-mm Gateway balloon-Wingspan stent system (Boston Scientific, Fremont, California). Postprocedure angiography demonstrated approximately 50% residual narrowing at the level of the stenosis with markedly improved luminal diameter and flow (Figure 9). The patient continued to do well clinically, maintained on an anti-platelet regimen.
A 58-year-old man with PMH significant for hypertension, diabetes, and hyperlipidemia was admitted for altered mental status. An unenhanced CT (not shown) demonstrated no acute finding. US revealed findings consistent with stenosis of the right vertebral artery (Figure 10). MRI performed the same day revealed multiple areas of restricted diffusion in the occipital lobes, compatible with ischemic strokes of embolic etiology (Figure 11). Old lacunar infarcts in the cerebellum were also evident. MRA of the brain suggested at least a 70% stenosis of the mid basilar artery, which was upgraded to 90% stenosis upon DSA (Figure 12). Since the patient had experienced multiple strokes despite aggressive medical management, endovascular management of the mid-basilar artery stenosis was considered. On day 9 of admission, successful angioplasty was performed (Figure 13). Following intervention, the patient remained stable with no additional episodic posterior circulation ischemic episodes or stroke.
Primary atherosclerotic disease of the vertebrobasilar system remains a significant subset of ischemic vascular disease and stroke affecting the population-at-large. With increasing clinical awareness, investigators have advanced noninvasive imaging capabilities, which primarily affect the vertebral and/or basilar arteries, and precisely diagnose areas of pathology. Concomitant neurointerventional advances in catheter-based therapies have advanced the treatment options for this often debilitating or deadly disease. Continued research remains paramount in an effort to recognize symptomatology, perform appropriate medical imaging tests, and formulate appropriate treatment plans.
Namini A, Naylor M, Koenigsberg RA. Vertebrobasilar Insufficiency and Stroke — A Review of Posterior Circulation Diagnostic Imaging and Endovascular Treatment Optio. J Am Osteopath Coll Radiol. 2015;4(3):15-23.